comparative genomics and evolution of genes encoding … · encoding bacterial (p)ppgpp...
TRANSCRIPT
Rel RelA and SpoT Proteins 585
Comparative Genomics and Evolution of GenesEncoding Bacterial (p)ppGpp SynthetasesHydrolases(the Rel RelA and SpoT Proteins)
For correspondenceEmail GerhardMittenhuberbiologieuni-greifswalddegmittenhuberpaulde Tel +49-3834-864218 Fax +49-3834-864202
J Mol Microbiol Biotechnol (2001) 3(4) 585-600
copy 2001 Horizon Scientific Press
JMMB Research Article
Gerhard Mittenhuber
Institut fuumlr Mikrobiologie und Molekularbiologie Ernst-Moritz-Arndt-Universitaumlt F-L Jahnstr 15 D-17487Greifswald Germany
Abstract
In the gram-negative model organism Escherichia colithe effector molecule of the stringent response(p)ppGpp is synthesized by two different enzymesRelA and SpoT whereas in the gram-positive modelorganism Bacillus subtilis only one enzyme named Relis responsible for this activity Rel and SpoT alsopossess (p)ppGpp hydrolase activity BLAST searcheswere used to identify orthologous genes in databasesThe construction and bootstrapping of phylogenetictrees allowed classification of these orthologs Fourgroups could be distinguished With the exception ofNeisseria and Bordetella (β subdivision) the RelA andSpoT groups are exclusively found in the γ subdivisionof proteobacteria Two Rel groups representing theactinobacterial and the BacillusClostridium groupwere also identified The SpoT proteins are related tothe gram positive Rel proteins RelA proteins carrysubstitutions in the HD domain (Aravind and Koonin1998 TIBS 23 469-472) responsible for ppGppdegradation A theory for the evolution of thespecialized paralogous relA and spoT genes ispresented After gene duplication of an ancestral rel-like gene the spoT and relA genes evolved from theduplicated genes The distribution pattern of theparalogous RelA and SpoT proteins supports a newmodel of linear bacterial evolution (Gupta 2000 FEMSMicrobiol Rev 24 367-402) This model postulates thatthe γ subdivision of proteobacteria represents the mostrecently evolved bacterial lineage However twoparalogous closely related genes of Porphyromonasgingivalis (Cytophaga-Flavobacterium-Bacteroidesphylum) encoding proteins with functions probablyidentical to the RelA and SpoT proteins do not fit inthis model Completely sequenced genomes of severalobligately parasitic organisms (Treponema pallidumChlamydia species Rickettsia prowazekii) and theobligate aphid symbiont Buchnera sp APS as well asarchaea do not contain rel-like genes but they are
present in the Arabidopsis genome In crosslinkingexperiments using different analogs of ppGpp ascrosslinking reagents and RNA polymerasepreparations of Escherichia coli binding of ppGpp todistinct regions at the C-terminus of the β subunit (theRpoB gene product) andor at the N-terminus of the βsubunit (the RpoC gene product) was observedpreviously RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibitspecific insertions or deletions in the ppGpp bindingregions
Introduction
The stringent response is an adaptive global mechanismfor control of gene expression in bacterial cells subjectedto starvation for amino acids or carbon sources (for reviewsee Cashel 1996) Initiallly discovered forty years ago inthe gram-negative model organism Escherichia coli (Stentand Brenner 1961) this response is characterized byprofound changes in the transcriptome of starved cells (egcessation of rRNA biosynthesis and activation of genesencoding amino acid biosynthesis genes activation of thestationary sigma factor σS (Gentry et al 1993) andtranscription at σS-dependent promoters (Kvint et al 2000)The stringent response is mediated through the synthesisof the effector molecule (p)ppGpp (Cashel and Gallant1969)
In E coli two different proteins are involved in(p)ppGpp synthesis RelA is bound to ribosomes sensesthe amount of uncharged tRNArsquos and synthesizes (p)ppGppfollowing amino acid limitation (Haseltine and Block 1973)The SpoT is a cytosolic protein (Gentry and Cashel 1995)and functions as (p)ppGpp synthetase after carbon(Hernandez and Bremer 1991 Murray and Bremer 1996)and fatty acid (Seyfzadeh et al 1993) starvation SpoT isalso a (p)ppGpp hydrolase (Hernandez and Bremer 1991Murray and Bremer 1996) The relA and spoT genes of Ecoli are related resulting in the hypothesis that these genesevolved by gene duplication (Metzger et al 1989) Residual(p)ppGpp synthesis in a relA mutant is abolished in a relAspoT double mutant (Xiao et al 1991) relAspoT doublemutants show a complex phenotype (morphologicalalterations loss in the ability to grow on amino acid-freeminimal media) (Xiao et al 1991 Hernandez and Cashel1995) The relA and spoT genes are listed as paralogousin the COG (cluster of orthologous groups) database(Tatusov et al 1997 Tatusov et al 2000) as COG0317
Several gram-positive bacterial species (ie Bacillussubtilis (Wendrich and Marahiel 1997) Corynebacteriumglutamicum (Wehmeier et al 1998) Mycobacteriumtuberculosis (Avarbock et al 1999) Staphylococcus
bull MALDI-TOF Mass Spectrometry in Microbiology
Edited by M Kostrzewa S Schubert (2016) wwwcaistercommalditof
bull Aspergillus and Penicillium in the Post-genomic Era
Edited by RP Vries IB Gelber MR Andersen (2016) wwwcaistercomaspergillus2
bull The Bacteriocins Current Knowledge and Future Prospects
Edited by RL Dorit SM Roy MA Riley (2016) wwwcaistercombacteriocins
bull Omics in Plant Disease Resistance
Edited by V Bhadauria (2016) wwwcaistercomopdr
bull Acidophiles Life in Extremely Acidic Environments
Edited by R Quatrini DB Johnson (2016) wwwcaistercomacidophiles
bull Climate Change and Microbial Ecology Current Research and Future Trends
Edited by J Marxsen (2016) wwwcaistercomclimate
bull Biofilms in Bioremediation Current Research and Emerging Technologies
Edited by G Lear (2016) wwwcaistercombiorem
bull Microalgae Current Research and Applications
Edited by MN Tsaloglou (2016) wwwcaistercommicroalgae
bull Gas Plasma Sterilization in Microbiology Theory Applications Pitfalls and New Perspectives
Edited by H Shintani A Sakudo (2016) wwwcaistercomgasplasma
bull Virus Evolution Current Research and Future Directions
Edited by SC Weaver M Denison M Roossinck et al (2016) wwwcaistercomvirusevol
bull Arboviruses Molecular Biology Evolution and Control
Edited by N Vasilakis DJ Gubler (2016) wwwcaistercomarbo
bull Shigella Molecular and Cellular Biology
Edited by WD Picking WL Picking (2016) wwwcaistercomshigella
bull Aquatic Biofilms Ecology Water Quality and Wastewater Treatment
Edited by AM Romaniacute H Guasch MD Balaguer (2016) wwwcaistercomaquaticbiofilms
bull Alphaviruses Current Biology
Edited by S Mahalingam L Herrero B Herring (2016) wwwcaistercomalpha
bull Thermophilic Microorganisms
Edited by F Li (2015) wwwcaistercomthermophile
bull Flow Cytometry in Microbiology Technology and Applications
Edited by MG Wilkinson (2015) wwwcaistercomflow
bull Probiotics and Prebiotics Current Research and Future Trends
Edited by K Venema AP Carmo (2015) wwwcaistercomprobiotics
bull Epigenetics Current Research and Emerging Trends
Edited by BP Chadwick (2015) wwwcaistercomepigenetics2015
bull Corynebacterium glutamicum From Systems Biology to Biotechnological Applications
Edited by A Burkovski (2015) wwwcaistercomcory2
bull Advanced Vaccine Research Methods for the Decade of Vaccines
Edited by F Bagnoli R Rappuoli (2015) wwwcaistercomvaccines
bull Antifungals From Genomics to Resistance and the Development of Novel Agents
Edited by AT Coste P Vandeputte (2015) wwwcaistercomantifungals
bull Bacteria-Plant Interactions Advanced Research and Future Trends
Edited by J Murillo BA Vinatzer RW Jackson et al (2015) wwwcaistercombacteria-plant
bull Aeromonas
Edited by J Graf (2015) wwwcaistercomaeromonas
bull Antibiotics Current Innovations and Future Trends
Edited by S Saacutenchez AL Demain (2015) wwwcaistercomantibiotics
bull Leishmania Current Biology and Control
Edited by S Adak R Datta (2015) wwwcaistercomleish2
bull Acanthamoeba Biology and Pathogenesis (2nd edition)
Author NA Khan (2015) wwwcaistercomacanthamoeba2
bull Microarrays Current Technology Innovations and Applications
Edited by Z He (2014) wwwcaistercommicroarrays2
bull Metagenomics of the Microbial Nitrogen Cycle Theory Methods and Applications
Edited by D Marco (2014) wwwcaistercomn2
Caister Academic Press is a leading academic publisher of advanced texts in microbiology molecular biology and medical research Full details of all our publications at caistercom
Further Reading
Order from caistercomorder
586 MittenhuberTa
ble
1 A
lpha
betic
al li
st o
f spe
cies
ana
lyze
d fo
r enz
ymes
invo
lved
in (p
)ppG
pp m
etab
olis
m a
nd in
form
atio
n ab
out t
he p
hylo
gene
tic s
tatu
s A
bbre
vatio
ns o
f nam
es a
re u
sed
in th
e ph
ylog
enet
ic tr
ees
and
the
Tabl
es
Org
anis
mPh
ylog
enet
ic p
ositi
onAb
brev
iatio
n
Actin
obac
illus
actin
omyc
etem
com
itans
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Aac
Arab
idop
sis th
alia
naEu
kary
ota
Viri
dipl
anta
e E
mbr
yoph
yta
Tra
cheo
phyt
a S
perm
atop
hyta
Mag
nolio
phyt
a e
udic
otyl
edon
s R
osid
ae e
uros
ids
II B
rass
ical
es B
rass
icac
eae
Ath
Aqui
fex
aeol
icus
Aqui
fical
es A
quifi
cace
aeAa
eBa
cillu
s an
thra
cisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
up B
acillu
s ce
reus
gro
upBa
nBa
cillu
s ha
lodu
rans
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bha
Bacil
lus
stea
roth
erm
ophi
lus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bst
Bacil
lus
subt
ilisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
upBs
uBo
rdet
ella
bro
nchi
sept
icaPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBb
rBo
rdet
ella
per
tuss
isPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBp
eBo
rrelia
bur
gdor
feri
Spiro
chae
tale
s S
piro
chae
tace
aeBb
uBr
adyr
hizo
bium
japo
nicu
mPr
oteo
bact
eria
α s
ubdi
visi
on B
rady
rhiz
obiu
m g
roup
Bja
Buch
nera
sp
APS
Prot
eoba
cter
ia γ
sub
divi
sion
Bsp
Cam
pylo
bact
er je
juni
Prot
eoba
cter
ia ε
sub
divi
sion
Cam
pylo
bact
er g
roup
Cje
Caul
obac
ter c
resc
entu
sPr
oteo
bact
eria
α s
ubdi
visi
on C
aulo
bact
er g
roup
Ccr
Chla
myd
ia tr
acho
mat
isBa
cter
ia C
hlam
ydia
les
Chl
amyd
iace
aeC
trCh
lam
ydia
pne
umon
iae
Bact
eria
Chl
amyd
iale
s C
hlam
ydia
ceae
Cpn
Chlo
robi
um te
pidu
mG
reen
sul
fur b
acte
riaC
teCl
ostri
dium
ace
tobu
tylic
umBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cac
Clos
tridi
um d
iffici
leBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cdi
Cory
neba
cter
ium
dip
hter
iae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e C
oryn
ebac
teria
ceae
Cdi
pCo
ryne
bact
eriu
m g
luta
micu
mFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es C
oryn
ebac
terin
eae
Cor
yneb
acte
riace
aeC
glDe
halo
cocc
oide
s et
heno
gene
sG
reen
non
-sul
fur b
acte
ria D
ehal
ococ
coid
es g
roup
Det
Dein
ococ
cus
radi
odur
ans
Ther
mus
Dei
noco
ccus
gro
up D
eino
cocc
ales
Dra
Desu
lfovib
rio v
ulga
risPr
oteo
bact
eria
δ s
ubdi
visi
onD
vuEn
tero
cocc
us fa
ecal
isFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up E
nter
ococ
cace
aeEf
aEs
cher
ichia
col
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Eco
Geo
bact
er s
ulfu
rredu
cens
Prot
eoba
cter
ia δ
sub
divi
sion
Geo
bact
eria
ceae
Gsu
Haem
ophi
lus
ducr
eyi
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Hdu
Haem
ophi
lus
influ
enza
ePr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aeH
inHe
licob
acte
r pylo
riPr
oteo
bact
eria
ε s
ubdi
visi
on H
elic
obac
ter g
roup
Hpy
Legi
onel
la p
neum
ophi
laPr
oteo
bact
eria
γ s
ubdi
visi
on L
egio
nella
ceae
Lpn
Myc
obac
teriu
m a
vium
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
avi
um c
ompl
ex (M
AC)
Mav
Myc
obac
teriu
m b
ovis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mbo
Myc
obac
teriu
m le
prae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
aeM
leM
ycob
acte
rium
tube
rcul
osis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mtu
Myc
opla
sma
geni
taliu
mFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up M
ollic
utes
Mge
Myc
opla
sma
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esM
pnM
yxoc
occu
s xa
nthu
sPr
oteo
bact
eria
δ s
ubdi
visi
on M
yxob
acte
ria M
yxco
cocc
ales
Cys
toba
cter
inae
Mxa
Neiss
eria
gon
orrh
oeae
Prot
eoba
cter
ia β
sub
divi
sion
Nei
sser
iace
aeN
goNe
isser
ia m
enin
gitid
isPr
oteo
bact
eria
β s
ubdi
visi
on N
eiss
eria
ceae
Nm
ePa
steu
rella
mul
tocid
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aePm
uPo
rphy
rom
onas
gin
giva
lisC
FB g
roup
Bac
terio
dace
aePg
iPs
eudo
mon
as a
erug
inos
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
seud
omon
as g
roup
Pae
Pseu
dom
onas
put
ida
Prot
eoba
cter
ia γ
sub
divi
sion
Pse
udom
onas
gro
upPp
uSa
lmon
ella
typh
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Sty
Shew
anel
la p
utre
facie
nsPr
oteo
bact
eria
γ s
ubdi
visi
on A
ltero
mon
adac
eae
Spu
Spiro
plas
ma
citri
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
es S
piro
plas
mat
acea
eSc
iSt
aphy
loco
ccus
aur
eus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Sau
Stre
ptoc
occu
s eq
uiFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSe
qSt
rept
ococ
cus
equi
similis
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Sequ
Stre
ptoc
occu
s m
utan
sFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSm
uSt
rept
ococ
cus
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Spn
Stre
ptoc
occu
s py
ogen
esFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSp
ySt
rept
omyc
es a
ntib
iotic
usFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
San
Stre
ptom
yces
coe
licol
orFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
Sco
Syne
choc
ystis
sp
PC
C68
03C
yano
bact
eria
Chr
ooco
ccal
esSs
pTh
erm
otog
a m
aritim
aTh
erm
otog
ales
Tma
Thio
bacil
lus
ferro
oxid
ans
Prot
eoba
cter
ia γ
sub
divi
sion
Tfe
Trep
onem
a pa
llidum
Spiro
chae
tale
s S
piro
chae
tace
aeTp
aUr
eapl
asm
a ur
ealyt
icum
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esU
urVi
brio
cho
lera
ePr
oteo
bact
eria
γ s
ubdi
visi
on V
ibrio
nace
aeVc
hVi
brio
sp
Prot
eoba
cter
ia γ
sub
divi
sion
Vib
riona
ceae
Vsp
Xyle
lla fa
stid
iosa
Prot
eoba
cter
ia γ
sub
divi
sion
Xan
thom
onas
gro
upXf
aYe
rsin
ia p
estis
Prot
eoba
cter
ia γ
sub
divi
sion
Ent
erob
acte
riace
aeYp
e
Rel RelA and SpoT Proteins 587
aureus (Gentry et al 2000) Streptococcus equisimilis(Mechold et al 1996 Mechold and Malke 1997)Streptomyces coelicolor (Chakraburtty et al 1996Martinez-Costa et al 1996 1998) Streptomycesantibioticus (Hoyt and Jones 1999)) however possess onlyone relAspoT paralog which exhibits both (p)ppGppsynthetase and hydrolase activity In these differentspecies ppGpp fulfills a variety of regulatory functions InB subtilis mutant strains defective in the relAspoT geneexhibit also pleiotropic phenotypes (amino acidauxotrophies (Wendrich and Marahiel (1997) defects inthe expression of several genes controlled by thesporulation sigma factor σH and reduced sporulationefficiency (Eymann et al 2001)) A rel mutant ofMycobacterium tuberculosis displays slower aerobic growthrate and is less able to survive extended anaerobicincubation (Primm et al 2000) The relAspoT gene is even
essential in S aureus (Gentry et al 2000) ppGpp plays arole in antibiotic and pigment production (Martinez-Costaet al 1996 Chakraburtty and Bibb 1997 Hoyt and Jones1999 Hesketh et al 2001) and morphologicaldifferentiation (Chakraburtty and Bibb 1997) inStreptomyces species
In the genomes of the gram-negative speciesMyxococcus xanthus (Harris et al 1998) andPorphyromonas gingivalis (Sen et al 2000) both relA andspoT genes seem to be present Harris et al (1998) clonedand sequenced the relA gene of M xanthus anddemonstrated that its gene product is essential for fruitingbody formation The other paralogous gene is not yetcloned Sen et al (2000) performed a search of theldquounfinished microbial genomerdquo sequence database of Pgingivalis at TIGR identified two homologues ORFrsquos andshowed that ORF1 encodes RelA of P gingivalis The otherORF is not yet characterized experimentally
Table 2 RelA orthologs The sequence of the RelA protein of E coli was used as query sequence inthe BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 The asterisk marks sequence fragments which were not included in thealignments because of their shortness Designations of proteins SpoT Pfr SpoT of Phlomobacterfragariae (Foissac et al 2000) SpoT SpoT sequence from an unidentified bacterium (Foissac et al2000) CsrS Vsp CsrS of Vibrio sp (Oumlstling et al 1995)
Designation Length Accession number E-value
RelA Eco 744 J04039 00RelA Vch 738 Q9S3S3 11e-263RelA Vsp 744 P55133 67e-262RelA Hin 743 P44644 35e-256RelA Pae 747 AAG04323 36e-183RelA Bha 728 BAB04961 11e-140RelA Bsu 734 U86377 24e-138RelA Xfa (XF1316) 718 AE003964 12e-136RelA Nme 769 CAB85211 76e-135Rel Ssp 760 P74007 8e-131Rel Sau 736 O32419 19e-129Rel Sco 847 X92520 75e-128Rel Cgl 760 O87331 41e-127Rel Mtu 790 Q50638 60e-126Rel Mle 787 Q49640 33e-125Rel Mxa 757 O52177 42e-125Rel Seq 739 Q54089 88e-125Rel San 841 O85709 23e-124Rel Aae 696 O67012 12e-123Rel Dra (DR1838) 787 Q9RTC7 26e-111SpoT Vch (VC2710) 705 AAF95850 3e-108SpoT Eco 702 P17580 17e-101SpoT Pae 701 AAG08723 17e-101SpoT Nme 725 CAB85138 23e-94SpoT Hin 677 P43811 87e-94Rel Tma (TM0729) 751 Q9WZI8 54e-93Rel Bja 779 Q9RH69 72e-84Rel Bbu 667 O51216 11e-80SpoT Xfa (XF0352) 735 AE003887 87e-77Rel Cje 731 AL139077 59e-74Rel Sci 749 O34098 62e-69RshA Sco (SC4H215) 725 O69970 86e-67Rel HpyJ99 776 Q9ZL68 76e-66F15I123 Ath 715 Q9SYH1 33e-59RSH2 Ath 710 AAF37282 34e-52SpoT Pfr 388 (fragment) AAG00076 30e-45T2H39 Ath 615 O81418 21e-40RSH1 Ath 883 AAF37281 24e-38Rel Uur 718 AE002125 33e-37Rel Mpn 733 P75386 71e-32Rel Mge 720 P47520 20e-29SpoT 283 (fragment) AAF73865 13e-27CsrS Vsp 119 (fragment) Q56730 1e-20
588 Mittenhuber
RelASpoT homologs have also been discovered inthe model plant Arabidopsis thaliana (van der Biezen etal 2000) These authors used the yeast two-hybrid assayin order to identify proteins that interact with RPP5 ThisArabidopsis protein which possesses a protein-proteininteraction module called ldquoNB-ARCrdquo (van der Biezen andJones 1998 Aravind et al 1999) is part of a signaltransduction cascade involved in sensing the fungalpathogen Peronospora parasitica (Noel et al 1999) Inthis two-hybrid assay the RSH1 (RelASpoT homolog)protein was identified a predicted plasma membrane-anchored cytoplasmic molecule with significant homologyto RelASpoT In a BLAST search two other unlinked rel-like genes were detected in the Arabidopsis genome theRSH2 and RSH3 genes (van der Biezen et al 2000)Functional complementation of appropriate E coli and Scoelicolor mutants by the RSH1 gene was also achievedin this study
Very recently the cloning and characterization of agene encoding a relAspoT homolog in Bacillusstearothermophilus was reported (Wendrich et al 2000)The authors include a phylogenetic tree of 32 proteinshomologous to RelASpoT of B subtilis van der Biezen etal (2000) also present a phylogenetic tree The RelA andSpoT proteins of E coli and closely related bacteria formdistinct clusters A cluster of actinobacterial RelASpoTsequences and a cluster of sequences representing theBacillusClostridium group can be observed Wendrich et
al (2000) also propose a nomenclature for enzymesinvolved in (p)ppGpp metabolism (p)ppGpp synthetase I(represented by RelA of E coli) (p)ppGpp synthetase II(represented by SpoT of E coli) and (p)ppGpp synthetaseIII (represented by synthetases of gram-positive origin thename Rel is proposed for this type of enzyme) Thisnomenclature is adopted in this paper althoughbiochemical evidence has not yet been reported for severalof the proteins named ldquoRelrdquo
However Wendrich et al (2000) and van der Biezenet al (2000) do not address the interesting question ofevolutionary relationships between these three classes ofenzymes to some detail In this paper these relationshipsare investigated using a larger number of RelA SpoT andRel sequences for construction of phylogenetic trees (fora list of species analyzed see Table 1) and by performingstatistical tests of tree topologies (ie bootstrapping(Felsenstein 1985)) During these studies we noted thatgenes encoding Rel-like proteins are absent in completelysequenced genomes of bacteria adopted to intracellularlifestyles Therefore we also investigated whether thetarget regions for binding of (p)ppGpp to the β and βsubunits of RNA polymerase (RNAP) in these speciesexhibit insertions or deletions (indels)
Results
Four General Groups of ppGpp SynthetaseHydrolaseProteinsOur initial phylogenetic analysis of the dataset of Wendrichet al (2000) showed a different tree topology than the onereported by the authors (data not shown) The reliability ofthe branching pattern of these tree data was then testedusing bootstrapping (Felsenstein 1985) Bootstrappinginvolves a randomization process in a multiple sequencealignment where all residues in a given column of thealignment are preserved in that column Pseudosamplesare created by random selection of sites from the data setwith replacement This is repeated a large number of times(eg 1000) and a phylogenetic tree is derived for eachpseudosample The percentage of pseudosamples whichsupports a unique clustering topology provides an estimateof the support for the pattern examined Two shortsequence fragments (ie CsrS from Vibrio sp (GenBankaccession number AAA85717 length 119 amino acidsOumlstling et al 1995) and a fragment derived from a gene ofPseudomonas denitrificans (GenBank acc No P29941length 175 aa Crouzet et al 1991) which were includedin the dataset of Wendrich et al (2000) were excluded fromthe dataset used for creation of the alignment becausebootstrapping of a dataset containing short sequencefragments is not informative A phylogenetic tree wasconstructed by the neighbor-joining (NJ) method (Saitouand Nei 1987) on the basis of the Poisson-corrected aminoacid distance (daa) Bootstrapping was performed with 1000replicates (Figure 1) This tree is mainly multifurcating (orldquocondensedrdquo) because many of the interior branchesexhibit a bootstrap value lower than 50 Only branchesseparating representatives of closely related species aswell as the RelA and SpoT proteins are supported by highbootstrap values
Figure 1 Rectangular cladogram of 30 RelA SpoT and Rel proteinsNumbers on the branches are percentages of 1000 bootstrap samplesthat support the branch only values gt50 are shown Bootstrapping wasperformed by the neighbor-joining (NJ) method (Saitou and Nei 1987) onthe basis of the Poisson-corrected amino acid distance (daa) The sequenceF15I1-23 of Arabidopsis was defined as outgroup
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
bull MALDI-TOF Mass Spectrometry in Microbiology
Edited by M Kostrzewa S Schubert (2016) wwwcaistercommalditof
bull Aspergillus and Penicillium in the Post-genomic Era
Edited by RP Vries IB Gelber MR Andersen (2016) wwwcaistercomaspergillus2
bull The Bacteriocins Current Knowledge and Future Prospects
Edited by RL Dorit SM Roy MA Riley (2016) wwwcaistercombacteriocins
bull Omics in Plant Disease Resistance
Edited by V Bhadauria (2016) wwwcaistercomopdr
bull Acidophiles Life in Extremely Acidic Environments
Edited by R Quatrini DB Johnson (2016) wwwcaistercomacidophiles
bull Climate Change and Microbial Ecology Current Research and Future Trends
Edited by J Marxsen (2016) wwwcaistercomclimate
bull Biofilms in Bioremediation Current Research and Emerging Technologies
Edited by G Lear (2016) wwwcaistercombiorem
bull Microalgae Current Research and Applications
Edited by MN Tsaloglou (2016) wwwcaistercommicroalgae
bull Gas Plasma Sterilization in Microbiology Theory Applications Pitfalls and New Perspectives
Edited by H Shintani A Sakudo (2016) wwwcaistercomgasplasma
bull Virus Evolution Current Research and Future Directions
Edited by SC Weaver M Denison M Roossinck et al (2016) wwwcaistercomvirusevol
bull Arboviruses Molecular Biology Evolution and Control
Edited by N Vasilakis DJ Gubler (2016) wwwcaistercomarbo
bull Shigella Molecular and Cellular Biology
Edited by WD Picking WL Picking (2016) wwwcaistercomshigella
bull Aquatic Biofilms Ecology Water Quality and Wastewater Treatment
Edited by AM Romaniacute H Guasch MD Balaguer (2016) wwwcaistercomaquaticbiofilms
bull Alphaviruses Current Biology
Edited by S Mahalingam L Herrero B Herring (2016) wwwcaistercomalpha
bull Thermophilic Microorganisms
Edited by F Li (2015) wwwcaistercomthermophile
bull Flow Cytometry in Microbiology Technology and Applications
Edited by MG Wilkinson (2015) wwwcaistercomflow
bull Probiotics and Prebiotics Current Research and Future Trends
Edited by K Venema AP Carmo (2015) wwwcaistercomprobiotics
bull Epigenetics Current Research and Emerging Trends
Edited by BP Chadwick (2015) wwwcaistercomepigenetics2015
bull Corynebacterium glutamicum From Systems Biology to Biotechnological Applications
Edited by A Burkovski (2015) wwwcaistercomcory2
bull Advanced Vaccine Research Methods for the Decade of Vaccines
Edited by F Bagnoli R Rappuoli (2015) wwwcaistercomvaccines
bull Antifungals From Genomics to Resistance and the Development of Novel Agents
Edited by AT Coste P Vandeputte (2015) wwwcaistercomantifungals
bull Bacteria-Plant Interactions Advanced Research and Future Trends
Edited by J Murillo BA Vinatzer RW Jackson et al (2015) wwwcaistercombacteria-plant
bull Aeromonas
Edited by J Graf (2015) wwwcaistercomaeromonas
bull Antibiotics Current Innovations and Future Trends
Edited by S Saacutenchez AL Demain (2015) wwwcaistercomantibiotics
bull Leishmania Current Biology and Control
Edited by S Adak R Datta (2015) wwwcaistercomleish2
bull Acanthamoeba Biology and Pathogenesis (2nd edition)
Author NA Khan (2015) wwwcaistercomacanthamoeba2
bull Microarrays Current Technology Innovations and Applications
Edited by Z He (2014) wwwcaistercommicroarrays2
bull Metagenomics of the Microbial Nitrogen Cycle Theory Methods and Applications
Edited by D Marco (2014) wwwcaistercomn2
Caister Academic Press is a leading academic publisher of advanced texts in microbiology molecular biology and medical research Full details of all our publications at caistercom
Further Reading
Order from caistercomorder
586 MittenhuberTa
ble
1 A
lpha
betic
al li
st o
f spe
cies
ana
lyze
d fo
r enz
ymes
invo
lved
in (p
)ppG
pp m
etab
olis
m a
nd in
form
atio
n ab
out t
he p
hylo
gene
tic s
tatu
s A
bbre
vatio
ns o
f nam
es a
re u
sed
in th
e ph
ylog
enet
ic tr
ees
and
the
Tabl
es
Org
anis
mPh
ylog
enet
ic p
ositi
onAb
brev
iatio
n
Actin
obac
illus
actin
omyc
etem
com
itans
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Aac
Arab
idop
sis th
alia
naEu
kary
ota
Viri
dipl
anta
e E
mbr
yoph
yta
Tra
cheo
phyt
a S
perm
atop
hyta
Mag
nolio
phyt
a e
udic
otyl
edon
s R
osid
ae e
uros
ids
II B
rass
ical
es B
rass
icac
eae
Ath
Aqui
fex
aeol
icus
Aqui
fical
es A
quifi
cace
aeAa
eBa
cillu
s an
thra
cisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
up B
acillu
s ce
reus
gro
upBa
nBa
cillu
s ha
lodu
rans
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bha
Bacil
lus
stea
roth
erm
ophi
lus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bst
Bacil
lus
subt
ilisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
upBs
uBo
rdet
ella
bro
nchi
sept
icaPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBb
rBo
rdet
ella
per
tuss
isPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBp
eBo
rrelia
bur
gdor
feri
Spiro
chae
tale
s S
piro
chae
tace
aeBb
uBr
adyr
hizo
bium
japo
nicu
mPr
oteo
bact
eria
α s
ubdi
visi
on B
rady
rhiz
obiu
m g
roup
Bja
Buch
nera
sp
APS
Prot
eoba
cter
ia γ
sub
divi
sion
Bsp
Cam
pylo
bact
er je
juni
Prot
eoba
cter
ia ε
sub
divi
sion
Cam
pylo
bact
er g
roup
Cje
Caul
obac
ter c
resc
entu
sPr
oteo
bact
eria
α s
ubdi
visi
on C
aulo
bact
er g
roup
Ccr
Chla
myd
ia tr
acho
mat
isBa
cter
ia C
hlam
ydia
les
Chl
amyd
iace
aeC
trCh
lam
ydia
pne
umon
iae
Bact
eria
Chl
amyd
iale
s C
hlam
ydia
ceae
Cpn
Chlo
robi
um te
pidu
mG
reen
sul
fur b
acte
riaC
teCl
ostri
dium
ace
tobu
tylic
umBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cac
Clos
tridi
um d
iffici
leBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cdi
Cory
neba
cter
ium
dip
hter
iae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e C
oryn
ebac
teria
ceae
Cdi
pCo
ryne
bact
eriu
m g
luta
micu
mFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es C
oryn
ebac
terin
eae
Cor
yneb
acte
riace
aeC
glDe
halo
cocc
oide
s et
heno
gene
sG
reen
non
-sul
fur b
acte
ria D
ehal
ococ
coid
es g
roup
Det
Dein
ococ
cus
radi
odur
ans
Ther
mus
Dei
noco
ccus
gro
up D
eino
cocc
ales
Dra
Desu
lfovib
rio v
ulga
risPr
oteo
bact
eria
δ s
ubdi
visi
onD
vuEn
tero
cocc
us fa
ecal
isFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up E
nter
ococ
cace
aeEf
aEs
cher
ichia
col
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Eco
Geo
bact
er s
ulfu
rredu
cens
Prot
eoba
cter
ia δ
sub
divi
sion
Geo
bact
eria
ceae
Gsu
Haem
ophi
lus
ducr
eyi
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Hdu
Haem
ophi
lus
influ
enza
ePr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aeH
inHe
licob
acte
r pylo
riPr
oteo
bact
eria
ε s
ubdi
visi
on H
elic
obac
ter g
roup
Hpy
Legi
onel
la p
neum
ophi
laPr
oteo
bact
eria
γ s
ubdi
visi
on L
egio
nella
ceae
Lpn
Myc
obac
teriu
m a
vium
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
avi
um c
ompl
ex (M
AC)
Mav
Myc
obac
teriu
m b
ovis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mbo
Myc
obac
teriu
m le
prae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
aeM
leM
ycob
acte
rium
tube
rcul
osis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mtu
Myc
opla
sma
geni
taliu
mFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up M
ollic
utes
Mge
Myc
opla
sma
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esM
pnM
yxoc
occu
s xa
nthu
sPr
oteo
bact
eria
δ s
ubdi
visi
on M
yxob
acte
ria M
yxco
cocc
ales
Cys
toba
cter
inae
Mxa
Neiss
eria
gon
orrh
oeae
Prot
eoba
cter
ia β
sub
divi
sion
Nei
sser
iace
aeN
goNe
isser
ia m
enin
gitid
isPr
oteo
bact
eria
β s
ubdi
visi
on N
eiss
eria
ceae
Nm
ePa
steu
rella
mul
tocid
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aePm
uPo
rphy
rom
onas
gin
giva
lisC
FB g
roup
Bac
terio
dace
aePg
iPs
eudo
mon
as a
erug
inos
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
seud
omon
as g
roup
Pae
Pseu
dom
onas
put
ida
Prot
eoba
cter
ia γ
sub
divi
sion
Pse
udom
onas
gro
upPp
uSa
lmon
ella
typh
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Sty
Shew
anel
la p
utre
facie
nsPr
oteo
bact
eria
γ s
ubdi
visi
on A
ltero
mon
adac
eae
Spu
Spiro
plas
ma
citri
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
es S
piro
plas
mat
acea
eSc
iSt
aphy
loco
ccus
aur
eus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Sau
Stre
ptoc
occu
s eq
uiFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSe
qSt
rept
ococ
cus
equi
similis
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Sequ
Stre
ptoc
occu
s m
utan
sFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSm
uSt
rept
ococ
cus
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Spn
Stre
ptoc
occu
s py
ogen
esFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSp
ySt
rept
omyc
es a
ntib
iotic
usFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
San
Stre
ptom
yces
coe
licol
orFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
Sco
Syne
choc
ystis
sp
PC
C68
03C
yano
bact
eria
Chr
ooco
ccal
esSs
pTh
erm
otog
a m
aritim
aTh
erm
otog
ales
Tma
Thio
bacil
lus
ferro
oxid
ans
Prot
eoba
cter
ia γ
sub
divi
sion
Tfe
Trep
onem
a pa
llidum
Spiro
chae
tale
s S
piro
chae
tace
aeTp
aUr
eapl
asm
a ur
ealyt
icum
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esU
urVi
brio
cho
lera
ePr
oteo
bact
eria
γ s
ubdi
visi
on V
ibrio
nace
aeVc
hVi
brio
sp
Prot
eoba
cter
ia γ
sub
divi
sion
Vib
riona
ceae
Vsp
Xyle
lla fa
stid
iosa
Prot
eoba
cter
ia γ
sub
divi
sion
Xan
thom
onas
gro
upXf
aYe
rsin
ia p
estis
Prot
eoba
cter
ia γ
sub
divi
sion
Ent
erob
acte
riace
aeYp
e
Rel RelA and SpoT Proteins 587
aureus (Gentry et al 2000) Streptococcus equisimilis(Mechold et al 1996 Mechold and Malke 1997)Streptomyces coelicolor (Chakraburtty et al 1996Martinez-Costa et al 1996 1998) Streptomycesantibioticus (Hoyt and Jones 1999)) however possess onlyone relAspoT paralog which exhibits both (p)ppGppsynthetase and hydrolase activity In these differentspecies ppGpp fulfills a variety of regulatory functions InB subtilis mutant strains defective in the relAspoT geneexhibit also pleiotropic phenotypes (amino acidauxotrophies (Wendrich and Marahiel (1997) defects inthe expression of several genes controlled by thesporulation sigma factor σH and reduced sporulationefficiency (Eymann et al 2001)) A rel mutant ofMycobacterium tuberculosis displays slower aerobic growthrate and is less able to survive extended anaerobicincubation (Primm et al 2000) The relAspoT gene is even
essential in S aureus (Gentry et al 2000) ppGpp plays arole in antibiotic and pigment production (Martinez-Costaet al 1996 Chakraburtty and Bibb 1997 Hoyt and Jones1999 Hesketh et al 2001) and morphologicaldifferentiation (Chakraburtty and Bibb 1997) inStreptomyces species
In the genomes of the gram-negative speciesMyxococcus xanthus (Harris et al 1998) andPorphyromonas gingivalis (Sen et al 2000) both relA andspoT genes seem to be present Harris et al (1998) clonedand sequenced the relA gene of M xanthus anddemonstrated that its gene product is essential for fruitingbody formation The other paralogous gene is not yetcloned Sen et al (2000) performed a search of theldquounfinished microbial genomerdquo sequence database of Pgingivalis at TIGR identified two homologues ORFrsquos andshowed that ORF1 encodes RelA of P gingivalis The otherORF is not yet characterized experimentally
Table 2 RelA orthologs The sequence of the RelA protein of E coli was used as query sequence inthe BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 The asterisk marks sequence fragments which were not included in thealignments because of their shortness Designations of proteins SpoT Pfr SpoT of Phlomobacterfragariae (Foissac et al 2000) SpoT SpoT sequence from an unidentified bacterium (Foissac et al2000) CsrS Vsp CsrS of Vibrio sp (Oumlstling et al 1995)
Designation Length Accession number E-value
RelA Eco 744 J04039 00RelA Vch 738 Q9S3S3 11e-263RelA Vsp 744 P55133 67e-262RelA Hin 743 P44644 35e-256RelA Pae 747 AAG04323 36e-183RelA Bha 728 BAB04961 11e-140RelA Bsu 734 U86377 24e-138RelA Xfa (XF1316) 718 AE003964 12e-136RelA Nme 769 CAB85211 76e-135Rel Ssp 760 P74007 8e-131Rel Sau 736 O32419 19e-129Rel Sco 847 X92520 75e-128Rel Cgl 760 O87331 41e-127Rel Mtu 790 Q50638 60e-126Rel Mle 787 Q49640 33e-125Rel Mxa 757 O52177 42e-125Rel Seq 739 Q54089 88e-125Rel San 841 O85709 23e-124Rel Aae 696 O67012 12e-123Rel Dra (DR1838) 787 Q9RTC7 26e-111SpoT Vch (VC2710) 705 AAF95850 3e-108SpoT Eco 702 P17580 17e-101SpoT Pae 701 AAG08723 17e-101SpoT Nme 725 CAB85138 23e-94SpoT Hin 677 P43811 87e-94Rel Tma (TM0729) 751 Q9WZI8 54e-93Rel Bja 779 Q9RH69 72e-84Rel Bbu 667 O51216 11e-80SpoT Xfa (XF0352) 735 AE003887 87e-77Rel Cje 731 AL139077 59e-74Rel Sci 749 O34098 62e-69RshA Sco (SC4H215) 725 O69970 86e-67Rel HpyJ99 776 Q9ZL68 76e-66F15I123 Ath 715 Q9SYH1 33e-59RSH2 Ath 710 AAF37282 34e-52SpoT Pfr 388 (fragment) AAG00076 30e-45T2H39 Ath 615 O81418 21e-40RSH1 Ath 883 AAF37281 24e-38Rel Uur 718 AE002125 33e-37Rel Mpn 733 P75386 71e-32Rel Mge 720 P47520 20e-29SpoT 283 (fragment) AAF73865 13e-27CsrS Vsp 119 (fragment) Q56730 1e-20
588 Mittenhuber
RelASpoT homologs have also been discovered inthe model plant Arabidopsis thaliana (van der Biezen etal 2000) These authors used the yeast two-hybrid assayin order to identify proteins that interact with RPP5 ThisArabidopsis protein which possesses a protein-proteininteraction module called ldquoNB-ARCrdquo (van der Biezen andJones 1998 Aravind et al 1999) is part of a signaltransduction cascade involved in sensing the fungalpathogen Peronospora parasitica (Noel et al 1999) Inthis two-hybrid assay the RSH1 (RelASpoT homolog)protein was identified a predicted plasma membrane-anchored cytoplasmic molecule with significant homologyto RelASpoT In a BLAST search two other unlinked rel-like genes were detected in the Arabidopsis genome theRSH2 and RSH3 genes (van der Biezen et al 2000)Functional complementation of appropriate E coli and Scoelicolor mutants by the RSH1 gene was also achievedin this study
Very recently the cloning and characterization of agene encoding a relAspoT homolog in Bacillusstearothermophilus was reported (Wendrich et al 2000)The authors include a phylogenetic tree of 32 proteinshomologous to RelASpoT of B subtilis van der Biezen etal (2000) also present a phylogenetic tree The RelA andSpoT proteins of E coli and closely related bacteria formdistinct clusters A cluster of actinobacterial RelASpoTsequences and a cluster of sequences representing theBacillusClostridium group can be observed Wendrich et
al (2000) also propose a nomenclature for enzymesinvolved in (p)ppGpp metabolism (p)ppGpp synthetase I(represented by RelA of E coli) (p)ppGpp synthetase II(represented by SpoT of E coli) and (p)ppGpp synthetaseIII (represented by synthetases of gram-positive origin thename Rel is proposed for this type of enzyme) Thisnomenclature is adopted in this paper althoughbiochemical evidence has not yet been reported for severalof the proteins named ldquoRelrdquo
However Wendrich et al (2000) and van der Biezenet al (2000) do not address the interesting question ofevolutionary relationships between these three classes ofenzymes to some detail In this paper these relationshipsare investigated using a larger number of RelA SpoT andRel sequences for construction of phylogenetic trees (fora list of species analyzed see Table 1) and by performingstatistical tests of tree topologies (ie bootstrapping(Felsenstein 1985)) During these studies we noted thatgenes encoding Rel-like proteins are absent in completelysequenced genomes of bacteria adopted to intracellularlifestyles Therefore we also investigated whether thetarget regions for binding of (p)ppGpp to the β and βsubunits of RNA polymerase (RNAP) in these speciesexhibit insertions or deletions (indels)
Results
Four General Groups of ppGpp SynthetaseHydrolaseProteinsOur initial phylogenetic analysis of the dataset of Wendrichet al (2000) showed a different tree topology than the onereported by the authors (data not shown) The reliability ofthe branching pattern of these tree data was then testedusing bootstrapping (Felsenstein 1985) Bootstrappinginvolves a randomization process in a multiple sequencealignment where all residues in a given column of thealignment are preserved in that column Pseudosamplesare created by random selection of sites from the data setwith replacement This is repeated a large number of times(eg 1000) and a phylogenetic tree is derived for eachpseudosample The percentage of pseudosamples whichsupports a unique clustering topology provides an estimateof the support for the pattern examined Two shortsequence fragments (ie CsrS from Vibrio sp (GenBankaccession number AAA85717 length 119 amino acidsOumlstling et al 1995) and a fragment derived from a gene ofPseudomonas denitrificans (GenBank acc No P29941length 175 aa Crouzet et al 1991) which were includedin the dataset of Wendrich et al (2000) were excluded fromthe dataset used for creation of the alignment becausebootstrapping of a dataset containing short sequencefragments is not informative A phylogenetic tree wasconstructed by the neighbor-joining (NJ) method (Saitouand Nei 1987) on the basis of the Poisson-corrected aminoacid distance (daa) Bootstrapping was performed with 1000replicates (Figure 1) This tree is mainly multifurcating (orldquocondensedrdquo) because many of the interior branchesexhibit a bootstrap value lower than 50 Only branchesseparating representatives of closely related species aswell as the RelA and SpoT proteins are supported by highbootstrap values
Figure 1 Rectangular cladogram of 30 RelA SpoT and Rel proteinsNumbers on the branches are percentages of 1000 bootstrap samplesthat support the branch only values gt50 are shown Bootstrapping wasperformed by the neighbor-joining (NJ) method (Saitou and Nei 1987) onthe basis of the Poisson-corrected amino acid distance (daa) The sequenceF15I1-23 of Arabidopsis was defined as outgroup
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
586 MittenhuberTa
ble
1 A
lpha
betic
al li
st o
f spe
cies
ana
lyze
d fo
r enz
ymes
invo
lved
in (p
)ppG
pp m
etab
olis
m a
nd in
form
atio
n ab
out t
he p
hylo
gene
tic s
tatu
s A
bbre
vatio
ns o
f nam
es a
re u
sed
in th
e ph
ylog
enet
ic tr
ees
and
the
Tabl
es
Org
anis
mPh
ylog
enet
ic p
ositi
onAb
brev
iatio
n
Actin
obac
illus
actin
omyc
etem
com
itans
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Aac
Arab
idop
sis th
alia
naEu
kary
ota
Viri
dipl
anta
e E
mbr
yoph
yta
Tra
cheo
phyt
a S
perm
atop
hyta
Mag
nolio
phyt
a e
udic
otyl
edon
s R
osid
ae e
uros
ids
II B
rass
ical
es B
rass
icac
eae
Ath
Aqui
fex
aeol
icus
Aqui
fical
es A
quifi
cace
aeAa
eBa
cillu
s an
thra
cisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
up B
acillu
s ce
reus
gro
upBa
nBa
cillu
s ha
lodu
rans
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bha
Bacil
lus
stea
roth
erm
ophi
lus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Bst
Bacil
lus
subt
ilisFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
upB
acillu
sSt
aphy
loco
ccus
gro
upBs
uBo
rdet
ella
bro
nchi
sept
icaPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBb
rBo
rdet
ella
per
tuss
isPr
oteo
bact
eria
β s
ubdi
visi
on A
lcal
igen
acea
eBp
eBo
rrelia
bur
gdor
feri
Spiro
chae
tale
s S
piro
chae
tace
aeBb
uBr
adyr
hizo
bium
japo
nicu
mPr
oteo
bact
eria
α s
ubdi
visi
on B
rady
rhiz
obiu
m g
roup
Bja
Buch
nera
sp
APS
Prot
eoba
cter
ia γ
sub
divi
sion
Bsp
Cam
pylo
bact
er je
juni
Prot
eoba
cter
ia ε
sub
divi
sion
Cam
pylo
bact
er g
roup
Cje
Caul
obac
ter c
resc
entu
sPr
oteo
bact
eria
α s
ubdi
visi
on C
aulo
bact
er g
roup
Ccr
Chla
myd
ia tr
acho
mat
isBa
cter
ia C
hlam
ydia
les
Chl
amyd
iace
aeC
trCh
lam
ydia
pne
umon
iae
Bact
eria
Chl
amyd
iale
s C
hlam
ydia
ceae
Cpn
Chlo
robi
um te
pidu
mG
reen
sul
fur b
acte
riaC
teCl
ostri
dium
ace
tobu
tylic
umBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cac
Clos
tridi
um d
iffici
leBa
cter
ia F
irmic
utes
Bac
illus
Clo
strid
ium
gro
up C
lost
ridia
ceae
Cdi
Cory
neba
cter
ium
dip
hter
iae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e C
oryn
ebac
teria
ceae
Cdi
pCo
ryne
bact
eriu
m g
luta
micu
mFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es C
oryn
ebac
terin
eae
Cor
yneb
acte
riace
aeC
glDe
halo
cocc
oide
s et
heno
gene
sG
reen
non
-sul
fur b
acte
ria D
ehal
ococ
coid
es g
roup
Det
Dein
ococ
cus
radi
odur
ans
Ther
mus
Dei
noco
ccus
gro
up D
eino
cocc
ales
Dra
Desu
lfovib
rio v
ulga
risPr
oteo
bact
eria
δ s
ubdi
visi
onD
vuEn
tero
cocc
us fa
ecal
isFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up E
nter
ococ
cace
aeEf
aEs
cher
ichia
col
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Eco
Geo
bact
er s
ulfu
rredu
cens
Prot
eoba
cter
ia δ
sub
divi
sion
Geo
bact
eria
ceae
Gsu
Haem
ophi
lus
ducr
eyi
Prot
eoba
cter
ia γ
sub
divi
sion
Pas
teur
ella
ceae
Hdu
Haem
ophi
lus
influ
enza
ePr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aeH
inHe
licob
acte
r pylo
riPr
oteo
bact
eria
ε s
ubdi
visi
on H
elic
obac
ter g
roup
Hpy
Legi
onel
la p
neum
ophi
laPr
oteo
bact
eria
γ s
ubdi
visi
on L
egio
nella
ceae
Lpn
Myc
obac
teriu
m a
vium
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
avi
um c
ompl
ex (M
AC)
Mav
Myc
obac
teriu
m b
ovis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mbo
Myc
obac
teriu
m le
prae
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
aeM
leM
ycob
acte
rium
tube
rcul
osis
Firm
icut
es A
ctin
obac
teria
Act
inob
acte
ridae
Act
inom
ycet
ales
Cor
yneb
acte
rinea
e M
ycob
acte
riace
ae M
ycob
acte
rium
tube
rcul
osis
com
plex
Mtu
Myc
opla
sma
geni
taliu
mFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up M
ollic
utes
Mge
Myc
opla
sma
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esM
pnM
yxoc
occu
s xa
nthu
sPr
oteo
bact
eria
δ s
ubdi
visi
on M
yxob
acte
ria M
yxco
cocc
ales
Cys
toba
cter
inae
Mxa
Neiss
eria
gon
orrh
oeae
Prot
eoba
cter
ia β
sub
divi
sion
Nei
sser
iace
aeN
goNe
isser
ia m
enin
gitid
isPr
oteo
bact
eria
β s
ubdi
visi
on N
eiss
eria
ceae
Nm
ePa
steu
rella
mul
tocid
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
aste
urel
lace
aePm
uPo
rphy
rom
onas
gin
giva
lisC
FB g
roup
Bac
terio
dace
aePg
iPs
eudo
mon
as a
erug
inos
aPr
oteo
bact
eria
γ s
ubdi
visi
on P
seud
omon
as g
roup
Pae
Pseu
dom
onas
put
ida
Prot
eoba
cter
ia γ
sub
divi
sion
Pse
udom
onas
gro
upPp
uSa
lmon
ella
typh
iPr
oteo
bact
eria
γ s
ubdi
visi
on E
nter
obac
teria
ceae
Sty
Shew
anel
la p
utre
facie
nsPr
oteo
bact
eria
γ s
ubdi
visi
on A
ltero
mon
adac
eae
Spu
Spiro
plas
ma
citri
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
es S
piro
plas
mat
acea
eSc
iSt
aphy
loco
ccus
aur
eus
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Bac
illus
Stap
hylo
cocc
us g
roup
Sau
Stre
ptoc
occu
s eq
uiFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSe
qSt
rept
ococ
cus
equi
similis
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Sequ
Stre
ptoc
occu
s m
utan
sFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSm
uSt
rept
ococ
cus
pneu
mon
iae
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Stre
ptoc
occa
ceae
Spn
Stre
ptoc
occu
s py
ogen
esFi
rmic
utes
Bac
illus
Clo
strid
ium
gro
up S
trept
ococ
cace
aeSp
ySt
rept
omyc
es a
ntib
iotic
usFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
San
Stre
ptom
yces
coe
licol
orFi
rmic
utes
Act
inob
acte
ria A
ctin
obac
terid
ae A
ctin
omyc
etal
es S
trept
omyc
inea
e S
trept
omyc
etac
eae
Sco
Syne
choc
ystis
sp
PC
C68
03C
yano
bact
eria
Chr
ooco
ccal
esSs
pTh
erm
otog
a m
aritim
aTh
erm
otog
ales
Tma
Thio
bacil
lus
ferro
oxid
ans
Prot
eoba
cter
ia γ
sub
divi
sion
Tfe
Trep
onem
a pa
llidum
Spiro
chae
tale
s S
piro
chae
tace
aeTp
aUr
eapl
asm
a ur
ealyt
icum
Firm
icut
es B
acillu
sC
lost
ridiu
m g
roup
Mol
licut
esU
urVi
brio
cho
lera
ePr
oteo
bact
eria
γ s
ubdi
visi
on V
ibrio
nace
aeVc
hVi
brio
sp
Prot
eoba
cter
ia γ
sub
divi
sion
Vib
riona
ceae
Vsp
Xyle
lla fa
stid
iosa
Prot
eoba
cter
ia γ
sub
divi
sion
Xan
thom
onas
gro
upXf
aYe
rsin
ia p
estis
Prot
eoba
cter
ia γ
sub
divi
sion
Ent
erob
acte
riace
aeYp
e
Rel RelA and SpoT Proteins 587
aureus (Gentry et al 2000) Streptococcus equisimilis(Mechold et al 1996 Mechold and Malke 1997)Streptomyces coelicolor (Chakraburtty et al 1996Martinez-Costa et al 1996 1998) Streptomycesantibioticus (Hoyt and Jones 1999)) however possess onlyone relAspoT paralog which exhibits both (p)ppGppsynthetase and hydrolase activity In these differentspecies ppGpp fulfills a variety of regulatory functions InB subtilis mutant strains defective in the relAspoT geneexhibit also pleiotropic phenotypes (amino acidauxotrophies (Wendrich and Marahiel (1997) defects inthe expression of several genes controlled by thesporulation sigma factor σH and reduced sporulationefficiency (Eymann et al 2001)) A rel mutant ofMycobacterium tuberculosis displays slower aerobic growthrate and is less able to survive extended anaerobicincubation (Primm et al 2000) The relAspoT gene is even
essential in S aureus (Gentry et al 2000) ppGpp plays arole in antibiotic and pigment production (Martinez-Costaet al 1996 Chakraburtty and Bibb 1997 Hoyt and Jones1999 Hesketh et al 2001) and morphologicaldifferentiation (Chakraburtty and Bibb 1997) inStreptomyces species
In the genomes of the gram-negative speciesMyxococcus xanthus (Harris et al 1998) andPorphyromonas gingivalis (Sen et al 2000) both relA andspoT genes seem to be present Harris et al (1998) clonedand sequenced the relA gene of M xanthus anddemonstrated that its gene product is essential for fruitingbody formation The other paralogous gene is not yetcloned Sen et al (2000) performed a search of theldquounfinished microbial genomerdquo sequence database of Pgingivalis at TIGR identified two homologues ORFrsquos andshowed that ORF1 encodes RelA of P gingivalis The otherORF is not yet characterized experimentally
Table 2 RelA orthologs The sequence of the RelA protein of E coli was used as query sequence inthe BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 The asterisk marks sequence fragments which were not included in thealignments because of their shortness Designations of proteins SpoT Pfr SpoT of Phlomobacterfragariae (Foissac et al 2000) SpoT SpoT sequence from an unidentified bacterium (Foissac et al2000) CsrS Vsp CsrS of Vibrio sp (Oumlstling et al 1995)
Designation Length Accession number E-value
RelA Eco 744 J04039 00RelA Vch 738 Q9S3S3 11e-263RelA Vsp 744 P55133 67e-262RelA Hin 743 P44644 35e-256RelA Pae 747 AAG04323 36e-183RelA Bha 728 BAB04961 11e-140RelA Bsu 734 U86377 24e-138RelA Xfa (XF1316) 718 AE003964 12e-136RelA Nme 769 CAB85211 76e-135Rel Ssp 760 P74007 8e-131Rel Sau 736 O32419 19e-129Rel Sco 847 X92520 75e-128Rel Cgl 760 O87331 41e-127Rel Mtu 790 Q50638 60e-126Rel Mle 787 Q49640 33e-125Rel Mxa 757 O52177 42e-125Rel Seq 739 Q54089 88e-125Rel San 841 O85709 23e-124Rel Aae 696 O67012 12e-123Rel Dra (DR1838) 787 Q9RTC7 26e-111SpoT Vch (VC2710) 705 AAF95850 3e-108SpoT Eco 702 P17580 17e-101SpoT Pae 701 AAG08723 17e-101SpoT Nme 725 CAB85138 23e-94SpoT Hin 677 P43811 87e-94Rel Tma (TM0729) 751 Q9WZI8 54e-93Rel Bja 779 Q9RH69 72e-84Rel Bbu 667 O51216 11e-80SpoT Xfa (XF0352) 735 AE003887 87e-77Rel Cje 731 AL139077 59e-74Rel Sci 749 O34098 62e-69RshA Sco (SC4H215) 725 O69970 86e-67Rel HpyJ99 776 Q9ZL68 76e-66F15I123 Ath 715 Q9SYH1 33e-59RSH2 Ath 710 AAF37282 34e-52SpoT Pfr 388 (fragment) AAG00076 30e-45T2H39 Ath 615 O81418 21e-40RSH1 Ath 883 AAF37281 24e-38Rel Uur 718 AE002125 33e-37Rel Mpn 733 P75386 71e-32Rel Mge 720 P47520 20e-29SpoT 283 (fragment) AAF73865 13e-27CsrS Vsp 119 (fragment) Q56730 1e-20
588 Mittenhuber
RelASpoT homologs have also been discovered inthe model plant Arabidopsis thaliana (van der Biezen etal 2000) These authors used the yeast two-hybrid assayin order to identify proteins that interact with RPP5 ThisArabidopsis protein which possesses a protein-proteininteraction module called ldquoNB-ARCrdquo (van der Biezen andJones 1998 Aravind et al 1999) is part of a signaltransduction cascade involved in sensing the fungalpathogen Peronospora parasitica (Noel et al 1999) Inthis two-hybrid assay the RSH1 (RelASpoT homolog)protein was identified a predicted plasma membrane-anchored cytoplasmic molecule with significant homologyto RelASpoT In a BLAST search two other unlinked rel-like genes were detected in the Arabidopsis genome theRSH2 and RSH3 genes (van der Biezen et al 2000)Functional complementation of appropriate E coli and Scoelicolor mutants by the RSH1 gene was also achievedin this study
Very recently the cloning and characterization of agene encoding a relAspoT homolog in Bacillusstearothermophilus was reported (Wendrich et al 2000)The authors include a phylogenetic tree of 32 proteinshomologous to RelASpoT of B subtilis van der Biezen etal (2000) also present a phylogenetic tree The RelA andSpoT proteins of E coli and closely related bacteria formdistinct clusters A cluster of actinobacterial RelASpoTsequences and a cluster of sequences representing theBacillusClostridium group can be observed Wendrich et
al (2000) also propose a nomenclature for enzymesinvolved in (p)ppGpp metabolism (p)ppGpp synthetase I(represented by RelA of E coli) (p)ppGpp synthetase II(represented by SpoT of E coli) and (p)ppGpp synthetaseIII (represented by synthetases of gram-positive origin thename Rel is proposed for this type of enzyme) Thisnomenclature is adopted in this paper althoughbiochemical evidence has not yet been reported for severalof the proteins named ldquoRelrdquo
However Wendrich et al (2000) and van der Biezenet al (2000) do not address the interesting question ofevolutionary relationships between these three classes ofenzymes to some detail In this paper these relationshipsare investigated using a larger number of RelA SpoT andRel sequences for construction of phylogenetic trees (fora list of species analyzed see Table 1) and by performingstatistical tests of tree topologies (ie bootstrapping(Felsenstein 1985)) During these studies we noted thatgenes encoding Rel-like proteins are absent in completelysequenced genomes of bacteria adopted to intracellularlifestyles Therefore we also investigated whether thetarget regions for binding of (p)ppGpp to the β and βsubunits of RNA polymerase (RNAP) in these speciesexhibit insertions or deletions (indels)
Results
Four General Groups of ppGpp SynthetaseHydrolaseProteinsOur initial phylogenetic analysis of the dataset of Wendrichet al (2000) showed a different tree topology than the onereported by the authors (data not shown) The reliability ofthe branching pattern of these tree data was then testedusing bootstrapping (Felsenstein 1985) Bootstrappinginvolves a randomization process in a multiple sequencealignment where all residues in a given column of thealignment are preserved in that column Pseudosamplesare created by random selection of sites from the data setwith replacement This is repeated a large number of times(eg 1000) and a phylogenetic tree is derived for eachpseudosample The percentage of pseudosamples whichsupports a unique clustering topology provides an estimateof the support for the pattern examined Two shortsequence fragments (ie CsrS from Vibrio sp (GenBankaccession number AAA85717 length 119 amino acidsOumlstling et al 1995) and a fragment derived from a gene ofPseudomonas denitrificans (GenBank acc No P29941length 175 aa Crouzet et al 1991) which were includedin the dataset of Wendrich et al (2000) were excluded fromthe dataset used for creation of the alignment becausebootstrapping of a dataset containing short sequencefragments is not informative A phylogenetic tree wasconstructed by the neighbor-joining (NJ) method (Saitouand Nei 1987) on the basis of the Poisson-corrected aminoacid distance (daa) Bootstrapping was performed with 1000replicates (Figure 1) This tree is mainly multifurcating (orldquocondensedrdquo) because many of the interior branchesexhibit a bootstrap value lower than 50 Only branchesseparating representatives of closely related species aswell as the RelA and SpoT proteins are supported by highbootstrap values
Figure 1 Rectangular cladogram of 30 RelA SpoT and Rel proteinsNumbers on the branches are percentages of 1000 bootstrap samplesthat support the branch only values gt50 are shown Bootstrapping wasperformed by the neighbor-joining (NJ) method (Saitou and Nei 1987) onthe basis of the Poisson-corrected amino acid distance (daa) The sequenceF15I1-23 of Arabidopsis was defined as outgroup
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 587
aureus (Gentry et al 2000) Streptococcus equisimilis(Mechold et al 1996 Mechold and Malke 1997)Streptomyces coelicolor (Chakraburtty et al 1996Martinez-Costa et al 1996 1998) Streptomycesantibioticus (Hoyt and Jones 1999)) however possess onlyone relAspoT paralog which exhibits both (p)ppGppsynthetase and hydrolase activity In these differentspecies ppGpp fulfills a variety of regulatory functions InB subtilis mutant strains defective in the relAspoT geneexhibit also pleiotropic phenotypes (amino acidauxotrophies (Wendrich and Marahiel (1997) defects inthe expression of several genes controlled by thesporulation sigma factor σH and reduced sporulationefficiency (Eymann et al 2001)) A rel mutant ofMycobacterium tuberculosis displays slower aerobic growthrate and is less able to survive extended anaerobicincubation (Primm et al 2000) The relAspoT gene is even
essential in S aureus (Gentry et al 2000) ppGpp plays arole in antibiotic and pigment production (Martinez-Costaet al 1996 Chakraburtty and Bibb 1997 Hoyt and Jones1999 Hesketh et al 2001) and morphologicaldifferentiation (Chakraburtty and Bibb 1997) inStreptomyces species
In the genomes of the gram-negative speciesMyxococcus xanthus (Harris et al 1998) andPorphyromonas gingivalis (Sen et al 2000) both relA andspoT genes seem to be present Harris et al (1998) clonedand sequenced the relA gene of M xanthus anddemonstrated that its gene product is essential for fruitingbody formation The other paralogous gene is not yetcloned Sen et al (2000) performed a search of theldquounfinished microbial genomerdquo sequence database of Pgingivalis at TIGR identified two homologues ORFrsquos andshowed that ORF1 encodes RelA of P gingivalis The otherORF is not yet characterized experimentally
Table 2 RelA orthologs The sequence of the RelA protein of E coli was used as query sequence inthe BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 The asterisk marks sequence fragments which were not included in thealignments because of their shortness Designations of proteins SpoT Pfr SpoT of Phlomobacterfragariae (Foissac et al 2000) SpoT SpoT sequence from an unidentified bacterium (Foissac et al2000) CsrS Vsp CsrS of Vibrio sp (Oumlstling et al 1995)
Designation Length Accession number E-value
RelA Eco 744 J04039 00RelA Vch 738 Q9S3S3 11e-263RelA Vsp 744 P55133 67e-262RelA Hin 743 P44644 35e-256RelA Pae 747 AAG04323 36e-183RelA Bha 728 BAB04961 11e-140RelA Bsu 734 U86377 24e-138RelA Xfa (XF1316) 718 AE003964 12e-136RelA Nme 769 CAB85211 76e-135Rel Ssp 760 P74007 8e-131Rel Sau 736 O32419 19e-129Rel Sco 847 X92520 75e-128Rel Cgl 760 O87331 41e-127Rel Mtu 790 Q50638 60e-126Rel Mle 787 Q49640 33e-125Rel Mxa 757 O52177 42e-125Rel Seq 739 Q54089 88e-125Rel San 841 O85709 23e-124Rel Aae 696 O67012 12e-123Rel Dra (DR1838) 787 Q9RTC7 26e-111SpoT Vch (VC2710) 705 AAF95850 3e-108SpoT Eco 702 P17580 17e-101SpoT Pae 701 AAG08723 17e-101SpoT Nme 725 CAB85138 23e-94SpoT Hin 677 P43811 87e-94Rel Tma (TM0729) 751 Q9WZI8 54e-93Rel Bja 779 Q9RH69 72e-84Rel Bbu 667 O51216 11e-80SpoT Xfa (XF0352) 735 AE003887 87e-77Rel Cje 731 AL139077 59e-74Rel Sci 749 O34098 62e-69RshA Sco (SC4H215) 725 O69970 86e-67Rel HpyJ99 776 Q9ZL68 76e-66F15I123 Ath 715 Q9SYH1 33e-59RSH2 Ath 710 AAF37282 34e-52SpoT Pfr 388 (fragment) AAG00076 30e-45T2H39 Ath 615 O81418 21e-40RSH1 Ath 883 AAF37281 24e-38Rel Uur 718 AE002125 33e-37Rel Mpn 733 P75386 71e-32Rel Mge 720 P47520 20e-29SpoT 283 (fragment) AAF73865 13e-27CsrS Vsp 119 (fragment) Q56730 1e-20
588 Mittenhuber
RelASpoT homologs have also been discovered inthe model plant Arabidopsis thaliana (van der Biezen etal 2000) These authors used the yeast two-hybrid assayin order to identify proteins that interact with RPP5 ThisArabidopsis protein which possesses a protein-proteininteraction module called ldquoNB-ARCrdquo (van der Biezen andJones 1998 Aravind et al 1999) is part of a signaltransduction cascade involved in sensing the fungalpathogen Peronospora parasitica (Noel et al 1999) Inthis two-hybrid assay the RSH1 (RelASpoT homolog)protein was identified a predicted plasma membrane-anchored cytoplasmic molecule with significant homologyto RelASpoT In a BLAST search two other unlinked rel-like genes were detected in the Arabidopsis genome theRSH2 and RSH3 genes (van der Biezen et al 2000)Functional complementation of appropriate E coli and Scoelicolor mutants by the RSH1 gene was also achievedin this study
Very recently the cloning and characterization of agene encoding a relAspoT homolog in Bacillusstearothermophilus was reported (Wendrich et al 2000)The authors include a phylogenetic tree of 32 proteinshomologous to RelASpoT of B subtilis van der Biezen etal (2000) also present a phylogenetic tree The RelA andSpoT proteins of E coli and closely related bacteria formdistinct clusters A cluster of actinobacterial RelASpoTsequences and a cluster of sequences representing theBacillusClostridium group can be observed Wendrich et
al (2000) also propose a nomenclature for enzymesinvolved in (p)ppGpp metabolism (p)ppGpp synthetase I(represented by RelA of E coli) (p)ppGpp synthetase II(represented by SpoT of E coli) and (p)ppGpp synthetaseIII (represented by synthetases of gram-positive origin thename Rel is proposed for this type of enzyme) Thisnomenclature is adopted in this paper althoughbiochemical evidence has not yet been reported for severalof the proteins named ldquoRelrdquo
However Wendrich et al (2000) and van der Biezenet al (2000) do not address the interesting question ofevolutionary relationships between these three classes ofenzymes to some detail In this paper these relationshipsare investigated using a larger number of RelA SpoT andRel sequences for construction of phylogenetic trees (fora list of species analyzed see Table 1) and by performingstatistical tests of tree topologies (ie bootstrapping(Felsenstein 1985)) During these studies we noted thatgenes encoding Rel-like proteins are absent in completelysequenced genomes of bacteria adopted to intracellularlifestyles Therefore we also investigated whether thetarget regions for binding of (p)ppGpp to the β and βsubunits of RNA polymerase (RNAP) in these speciesexhibit insertions or deletions (indels)
Results
Four General Groups of ppGpp SynthetaseHydrolaseProteinsOur initial phylogenetic analysis of the dataset of Wendrichet al (2000) showed a different tree topology than the onereported by the authors (data not shown) The reliability ofthe branching pattern of these tree data was then testedusing bootstrapping (Felsenstein 1985) Bootstrappinginvolves a randomization process in a multiple sequencealignment where all residues in a given column of thealignment are preserved in that column Pseudosamplesare created by random selection of sites from the data setwith replacement This is repeated a large number of times(eg 1000) and a phylogenetic tree is derived for eachpseudosample The percentage of pseudosamples whichsupports a unique clustering topology provides an estimateof the support for the pattern examined Two shortsequence fragments (ie CsrS from Vibrio sp (GenBankaccession number AAA85717 length 119 amino acidsOumlstling et al 1995) and a fragment derived from a gene ofPseudomonas denitrificans (GenBank acc No P29941length 175 aa Crouzet et al 1991) which were includedin the dataset of Wendrich et al (2000) were excluded fromthe dataset used for creation of the alignment becausebootstrapping of a dataset containing short sequencefragments is not informative A phylogenetic tree wasconstructed by the neighbor-joining (NJ) method (Saitouand Nei 1987) on the basis of the Poisson-corrected aminoacid distance (daa) Bootstrapping was performed with 1000replicates (Figure 1) This tree is mainly multifurcating (orldquocondensedrdquo) because many of the interior branchesexhibit a bootstrap value lower than 50 Only branchesseparating representatives of closely related species aswell as the RelA and SpoT proteins are supported by highbootstrap values
Figure 1 Rectangular cladogram of 30 RelA SpoT and Rel proteinsNumbers on the branches are percentages of 1000 bootstrap samplesthat support the branch only values gt50 are shown Bootstrapping wasperformed by the neighbor-joining (NJ) method (Saitou and Nei 1987) onthe basis of the Poisson-corrected amino acid distance (daa) The sequenceF15I1-23 of Arabidopsis was defined as outgroup
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
588 Mittenhuber
RelASpoT homologs have also been discovered inthe model plant Arabidopsis thaliana (van der Biezen etal 2000) These authors used the yeast two-hybrid assayin order to identify proteins that interact with RPP5 ThisArabidopsis protein which possesses a protein-proteininteraction module called ldquoNB-ARCrdquo (van der Biezen andJones 1998 Aravind et al 1999) is part of a signaltransduction cascade involved in sensing the fungalpathogen Peronospora parasitica (Noel et al 1999) Inthis two-hybrid assay the RSH1 (RelASpoT homolog)protein was identified a predicted plasma membrane-anchored cytoplasmic molecule with significant homologyto RelASpoT In a BLAST search two other unlinked rel-like genes were detected in the Arabidopsis genome theRSH2 and RSH3 genes (van der Biezen et al 2000)Functional complementation of appropriate E coli and Scoelicolor mutants by the RSH1 gene was also achievedin this study
Very recently the cloning and characterization of agene encoding a relAspoT homolog in Bacillusstearothermophilus was reported (Wendrich et al 2000)The authors include a phylogenetic tree of 32 proteinshomologous to RelASpoT of B subtilis van der Biezen etal (2000) also present a phylogenetic tree The RelA andSpoT proteins of E coli and closely related bacteria formdistinct clusters A cluster of actinobacterial RelASpoTsequences and a cluster of sequences representing theBacillusClostridium group can be observed Wendrich et
al (2000) also propose a nomenclature for enzymesinvolved in (p)ppGpp metabolism (p)ppGpp synthetase I(represented by RelA of E coli) (p)ppGpp synthetase II(represented by SpoT of E coli) and (p)ppGpp synthetaseIII (represented by synthetases of gram-positive origin thename Rel is proposed for this type of enzyme) Thisnomenclature is adopted in this paper althoughbiochemical evidence has not yet been reported for severalof the proteins named ldquoRelrdquo
However Wendrich et al (2000) and van der Biezenet al (2000) do not address the interesting question ofevolutionary relationships between these three classes ofenzymes to some detail In this paper these relationshipsare investigated using a larger number of RelA SpoT andRel sequences for construction of phylogenetic trees (fora list of species analyzed see Table 1) and by performingstatistical tests of tree topologies (ie bootstrapping(Felsenstein 1985)) During these studies we noted thatgenes encoding Rel-like proteins are absent in completelysequenced genomes of bacteria adopted to intracellularlifestyles Therefore we also investigated whether thetarget regions for binding of (p)ppGpp to the β and βsubunits of RNA polymerase (RNAP) in these speciesexhibit insertions or deletions (indels)
Results
Four General Groups of ppGpp SynthetaseHydrolaseProteinsOur initial phylogenetic analysis of the dataset of Wendrichet al (2000) showed a different tree topology than the onereported by the authors (data not shown) The reliability ofthe branching pattern of these tree data was then testedusing bootstrapping (Felsenstein 1985) Bootstrappinginvolves a randomization process in a multiple sequencealignment where all residues in a given column of thealignment are preserved in that column Pseudosamplesare created by random selection of sites from the data setwith replacement This is repeated a large number of times(eg 1000) and a phylogenetic tree is derived for eachpseudosample The percentage of pseudosamples whichsupports a unique clustering topology provides an estimateof the support for the pattern examined Two shortsequence fragments (ie CsrS from Vibrio sp (GenBankaccession number AAA85717 length 119 amino acidsOumlstling et al 1995) and a fragment derived from a gene ofPseudomonas denitrificans (GenBank acc No P29941length 175 aa Crouzet et al 1991) which were includedin the dataset of Wendrich et al (2000) were excluded fromthe dataset used for creation of the alignment becausebootstrapping of a dataset containing short sequencefragments is not informative A phylogenetic tree wasconstructed by the neighbor-joining (NJ) method (Saitouand Nei 1987) on the basis of the Poisson-corrected aminoacid distance (daa) Bootstrapping was performed with 1000replicates (Figure 1) This tree is mainly multifurcating (orldquocondensedrdquo) because many of the interior branchesexhibit a bootstrap value lower than 50 Only branchesseparating representatives of closely related species aswell as the RelA and SpoT proteins are supported by highbootstrap values
Figure 1 Rectangular cladogram of 30 RelA SpoT and Rel proteinsNumbers on the branches are percentages of 1000 bootstrap samplesthat support the branch only values gt50 are shown Bootstrapping wasperformed by the neighbor-joining (NJ) method (Saitou and Nei 1987) onthe basis of the Poisson-corrected amino acid distance (daa) The sequenceF15I1-23 of Arabidopsis was defined as outgroup
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 589
We assumed that a larger set of sequences used asinput file for the alignment might result in a more informativetree Therefore BLAST searches using the RelA (Table 2)and SpoT (Table 3) sequences of E coli and the Relsequence of B subtilis (Table 4) as query sequences wereperformed The Rel sequence of B subtilis was also usedto search the ldquounfinished microbial genomesrdquo database atTIGR (see Table 5) A dataset comprising 80 relatedsequences was obtained in this way An alignment of thesesequences was performed and an unrooted tree wasobtained (Figure 2) Several designations have beenreplaced by numbers in this tree A clear separation of theRelA and SpoT families on one hand and of two Rel familiesof gram-positive origin (BacillusClostridium group and aactinobacterial group) can be observed Bootstrapping ofthis dataset also resulted in a largely condensed tree withinternally supported branches (Figure 3) Interestingly as
also indicated by Wendrich et al (2000) and as depictedin Figure 1 mycoplasmal Rel proteins form a clearoutgroup
It seemed likely that exclusion of certain sequencesfrom the dataset used to produce the alignment mightincrease the resolution of the condensed part of the treeIn a first attempt the outgroup sequences of Arabidopsis(RSH2 Ath F15I1-23 Ath) and of the Mycoplasma (Rel MpnRel Mge Rel Uur) group as well as the Arabidopsissequences T2H3-9 and RSH1 which are located in thecondensed part of the tree were removed from the dataset(total 73 sequences) subjected to the alignment Theresulting bootstrapped tree did not exhibit increasedresolution in the condensed part of the tree (data notshown) In a second attempt internal sequences of thecondensed region of the tree in Figure 3 (ie Rel Dra RelSsp Rel Det Rel Aae Rel Tma Rel Mxa Rel Gsu Rel
Table 3 SpoT orthologs The sequence of the SpoT protein of E coli was used as query sequencein the BLAST search Individual proteins are sorted according to ascending E-values For designationsof species see Table 1 For explanation of the asterisk see Figure 2 Designation of the proteinSpoT Pde SpoT of Pseudomonas denitrificans (Crouzet et al 1991)
Designation Length Accession number E-value
SpoT Eco 702 P17580 00SpoT Vch (VC2710) 705 AAF95850 19e-280SpoT Pae 701 AAG08723 66e-207SpoT Hin 677 P43811 23e-204SpoT Pfr 388 (fragment) AAG00076 11e-165SpoT Xfa (XF0352) 735 AE003887 37e-165SpoT Nme 725 CAB85138 13e-164Rel Bha 728 BAB04961 58e-148Rel Bsu 734 U86377 25e-147Rel Seq 739 Q54089 42e-143Rel Ssp 760 P74007 38e-142Rel Sau 736 O32419 8e-140Rel Mxa 757 O52177 19e-138Rel Aae 696 O67012 16e-134Rel San 841 O85709 17e-133Rel Mtu 790 Q50638 2e-132Rel Cgl 760 O87331 23e-131Rel Mle 787 Q49640 16e-130Rel Sco 847 X92520| 38e-127Rel Dra (DR1838) 787 Q9RTC7 55e-126Rel Tma (TM0729) 751 Q9WZI8 13e-125SpoT 283 (fragment) AAF73865 91e-123Rel Bja 779 Q9RH69 47e-116RelA Pae 747 AAG04323 95e-112RelA Nme 769 CAB85211 1e-105RelA Hin 743 P44644 94e-103Rel Bbu 667 O51216 51e-102RelA Eco 744 J04039 52e-100RelA Vsp 744 P55133 76e-97RelA Xfa (XF1316) 718 AE003964 29e-94Rel Cje 731 AL139077 34e-91Rel Sci 749 O34098 16e-90SC4H215 Sco 725 O69970 1e-88RelA Vch (VC249-51) 738 Q9S3S3 25e-88Rel Hpy99 776 Q9ZL68 15e-85RSH1 Ath 883 AAF37281 13e-73F15I123 Ath 715 Q9SYH1 96e-72Rel Uur 718 AE002125 12e-67RSH2 Ath 710 AAF37282 15e-67T2H39 Ath 615 O81418 81e-62Rel Mge 720 P47520 41e-56Rel Mpn 733 P75386 6e-55CsrS Vsp 119 (fragment) Q56730 11e-45SpoT Pde 175 (fragment) P29941 24e-10
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
590 Mittenhuber
Figure 2 Radial phylogenetic tree of 80 RelA SpoT and Rel proteins Due to space limitations the designation of individual Rel proteins was replaced bynumbers 1 refers to Rel Sco 2 to Rel San 3 to Rel Mav 4 to Rel Mle 5 to Rel Mtu 6 to Rel Mbo and 7 to Rel Cdiph
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 591
Dvu) were removed from the dataset (total 65 sequences)used for production of the alignment However use thisdataset in the construction of a bootstrapped tree also didnot result in increased resolution of the consensed part onthe tree (data not shown) In a third attempt seven otheroutgroup sequences (ie Rel Cte Rel1 Pgi Rel2 Pgi RelBbu Rel Sci Rel Cje and Rel Hpy) were removed fromthe dataset The alignment of this dataset (58 sequences)finally resulted in improved resolution of the resultingbootstrapped tree (Figure 4)
A statistically insignificant branching (53) separatesthe RelA proteins from the SpoT proteins and two groupsof Rel proteins from gram-positive species A statisticallysignificant branching (99 ) separates the Rel proteins ofthe BacillusClostridium group from the SpoT proteins andthe actinobacterial Rel proteins The SpoT proteins areseparated from the actinobacterial Rel proteins by astatistically less significant branching (78) An unrooted
radial tree originating from this alignment is presented inFigure 5 It should be noted that on the other hand whenanother dataset comprising only the sequencesrepresenting the condensed part of the tree of Figure 2was used as input file for an alignment this procedure alsodid not increase the resolution in the condensed part ofthe resulting bootstrapped tree (data not shown)
Substitutions in the HD Domain of RelA ProteinsIt was recognized by Aravind and Koonin (1998) that theHD domain defines a new superfamily of metal-dependentphosphohydrolases Aravind and Koonin (1998) define fivemotifs within the HD domain Three of them (motifs I IIand V) contain highly conserved histidine or aspartateresidues Site-directed mutagenesis experiments of genesencoding cGMP-phosphodiesterases have shown thatmutations of the histidine residue in the HD signature inmotif II or of the aspartate residue in motif V resulted in the
Table 4 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST searchIndividual proteins are sorted according to ascending E-values For designations of species see Table 1 For explanation of theasterisk see Figure 2
Designation Length Accession number E-Value
Rel Bsu 734 U86377 00Rel Bha 728 BAB04961 2e-278Rel Sau 736 O32419 12e-241Rel Seq 739 Q54089 99e-197Rel Ssp 760 P74007 75e-167Rel Mxa 757 O52177 75e-167Rel Sco 847 X92520 26e-164Rel San 841 O85709 31e-161Rel Mtu 790 Q50638 29e-158Rel Mle 787 Q49640 23e-156Rel Cgl 760 O87331 24e-154Rel Aae 760 O67012 95e-148SpoT Eco 702 P17580 3e-140SpoT Pae 701 AAG08723 31e-138RelA Pae 747 AAG04323 35e-137Rel Tma (TM0729) 751 Q9WZI8 93e-137SpoT Vch (VC2710) 705 AAF95850 45e-135RelA Eco 744 J04039 68e-134RelA Vch 738 Q9S3S3 18e-133RelA Vsp 744 P55133 11e-131RelA Hin 743 P44644 57e-130SpoT Nme 725 CAB85138 37e-129SpoT Hin 677 P43811 22e-122RelA Nme 769 CAB85211 13e-121SpoT Xfa (XF0352) 735 AE003887 17e-120Rel Bbu 667 O51216 12e-104Rel Sci 749 O34098 14e-103RelA Xfa (XF1316) 718 AE003964 6e-103Rel Bja 779 Q9RH69 68e-102Rel Cje 731 AL139077 48e-93Rel Dra (DR1838) 787 Q9RTC7 27e-91RshA Sco (SC4H215) 725 O69970 98e-88Rel Hpy 776 Q9ZL68 13e-82Rel Uur 718 AE002125 28e-78T2H39 Ath 615 AAC28176 19e-71Rel Mge 720 P47520 53e-69Rel Mpn 733 P75386 23e-68F15I123 Ath (identical to RSH3) 715 AAD25787 12e-67RSH3 712 AAF37283 e-67SpoT Pfr 388 (fragment) AAG00076 33e-65RSH1 Ath 883 AAF37281 8e-64RSH2 Ath 710 AAF37282 11e-62SpoT 283 (fragment) AAF73865 39e-43CsrS Vsp 119 (fragment) Q56730 13e-25SpoT Pde 175 (fragment) P29941 18e-8
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
592 Mittenhuber
most severe effect on the catalytic activity (Turko et al1996) Aravind and Koonin (1998) showed that the RelAand SpoT proteins are also members of this superfamilyBut the RelA protein of E coli contains substitutions in thepredicted catalytic residues of this domain Aravind andKoonin (1998) suggest that the HD domain of RelA isinactivated but still retains its native structure An inactivephosphohydrolase domain helps to explain why RelAproteins are not able to degrade ppGpp On the other handthe N-terminus of SpoT encompassing the HD domain issuffient for ppGpp hydrolase activity (Gentry and Cashel1996) An alignment of the N-termini of the SpoT Rel andRelA proteins is presented in Supplementary Material (httpjmmbnetsupplementary) This alignment shows theregion around conserved motifs I and II and the regionaround conserved motif V Importantly all the proteins
identified in this study as putative RelA proteins carrysubstitutions in all conserved regions whereas Rel andSpoT proteins and the Arabidopsis paralogs possess afunctional HD domain (httpjmmbnetsupplementary)Interestingly in most substitutions of the HD signaturesequence insertion of a proline residue took place anamino acid known to influence protein folding (eg Eylesand Gierasch 2000) Nevertheless the SpoT Rel and RelAproteins share some overall similiarity in this region Thisanalysis also suggests that the protein designated RelA ofM xanthus by Harris et al (1998) is in fact a Rel or SpoTprotein
Comparison of Putative ppGpp Binding SitesDuring this analysis we noted that genes encoding(p)ppGpp synthetaseshydrolases are absent in thecompletely sequenced genomes of the obligate parasitesChlamydia pneumoniae and C trachomatis (Kalman et al1999) Rickettsia prowazekii (Andersson et al 1998) andTreponema pallidum (Fraser et al 1998) Similarly thegenome of an intracellular symbiont of aphids Buchnerasp APS (Shigenobu et al 2000) does not contain a RelAor SpoT gene although this extremely adapted bacteriumis a close relative of E coli It seems likely that the relgenes were deleted during the adaptation to the intracellularenvironment and the establishment of specialized lifestylesin a process called ldquoreductive evolutionrdquo (Andersson andKurland 1998)
Genetic studies (Glass et al 1986) using strainscarrying amber mutations in rpo genes and biochemicalinvestigations (Reddy et al 1995 Chatterji et al 1998Toulokhonov et al 2001) using various ppGpp analogs ascrosslinking agents identified RNAP of E coli as target forppGpp Glass et al (1986) concluded that ppGpp binds tothe β subunit encoded by rpoB and also identified acommon motif in purine-nucleotide binding proteins(GXXXXGK la Cour et al 1985) in the amino acidsequence of RpoB at residues 880 to 886 By a combinationof a variety of methods (SDS-PAGE analysis of the[32P]N3ppGpp (azido-ppGpp)-bound enzyme trypticdigestion and Western blots) Chatterji et al (1998)identified a 45 kDa fragment of the β subunit as bindingsite for ppGpp This fragment is located at the C-terminusand consists of residues 802-1211 1216 or 1223 (Chatterjiet al 1998) Using the smaller ppGpp analog 6-thio-ppGpphowever the β subunit encoded by rpoC was very recentlyidentified as target for ppGpp (Toulokhonov et al 2001)The binding site for ppGpp was determined at the N-terminus (between residues 29-102) of the β subunit Theauthors explain these conflicting results by the propertiesof the different crosslinking reagents and by the 3D modelof RNAP (Zhang et al 1999) This model shows that theN-terminus of β and the C-terminus of β are spatially closeand constitute an intertwined interface Toulokhonov et al(2001) postulate that binding of ppGpp to RNAP isallosteric that the binding site is modular and is locatedclose to the intersubunit interface of the N-terminal and C-terminal ends of β and β
In order to reveal the presence or absence of thedifferent identified ppGpp binding regions in the β and βsequences alignments of RpoB (Table 6 httpjmmbnetsupplementary) and RpoC (Table 7 httpjmmbnet
Figure 3 Rectangular cladogram of 80 RelA SpoT and Rel proteinsNumbers on the branches indicated are bootstrap values as described inthe legend to Figure 1
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 593
residues 1140 to 1260 of the RpoB alignment are presentin the proteobacterial and chlamydial sequences and againaround residues 1340 to 1420 in the proteobacterialmycobacterial and mycoplasmal sequences Mostinterestingly the RpoB and RpoC sequences of the specieswhich do not possess a rel like gene do not exhibit specific
Table 5 Rel orthologs The sequence of the Rel protein of B subtilis was used as query sequence in the BLAST search of the ldquounfinishedmicrobial genomesrdquo database at TIGR For designations of species see Table 1 Individual proteins are sorted according to ascending E-values
Designation Length Contig designation E-value
Rel Ban 727 gnl|TIGR_1392|banth_1489 00Rel Bst 707 (fragment) gnl|UOKNOR_1422|bstear_Contig408 00Rel Efa 737 gnl|TIGR|gef_10492 00Rel Cdi 735 gnl|Sanger_1496|cdifficile_Contig876 00Rel Spn 740 gnl|TIGR|Spneumoniae_3836 00Rel Smu 740 gnl|UOKNOR_1309|Smutans_Contig98 00Rel Spy 739 gnl|OUACGT_1315|Spyogenes_Contig1 00Rel Cac 740 gnl|GTC|Caceto_gnl 00Rel Sequ 719 (fragment) gnl|Sanger_1336|sequi_Contig474 00Rel Gsu 716 gnl|TIGR_35554|gsulf_57 00Rel Det 728 gnl|TIGR_61435|deth_1541 00Rel Mbo 738 gnl|Sanger_1765|mbovis_Contig403 e-171Rel Cdiph 759 gnl|Sanger_1717|cdiph_Contig2 e-166Rel Dvu 717 gnl|TIGR_881|dvulg_159 e-166Rel Mav 647 gnl|TIGR|Mavium_223 e-162SpoT Tfe 759 gnl|TIGR|t_ferrooxidans_6160 e-149SpoT Ype 707 (fragment) gnl|Sanger_632|Ypesits_Contig1008 e-148SpoT Sty 709 (fragment) gnl|Sanger_601|Styphi_CT18 e-148RelA Sty 744 gnl|Sanger_601|Styphi_CT18 e-142RelA Ype 744 gnl|Sanger_632|Ypesits_Contig854 e-146RelA Ppu 746 gnl|TIGR|pputida_10724 e-145Rel Cte 731 gnl|TIGR|Ctepidum_3499 e-143SpoT Spu 712 (fragment) gnl|TIGR_24|sputre_6408 e-142RelA Spu 735 gnl|TIGR_24|sputre_6426 e-142RelA Lpn 734 gnl|CUCGC_446|lpneumo_MF18110397 e-141RelA Pmu 739 gnl|CBCUMN_747|Pmultocida e-140SpoT Pmu 711 (fragment) gnl|CBCUMN_747|Pmultocida e-139SpoT Aac 712 (fragment) gnl|OUACGT_714|Aactin_Contig253 e-139SpoT Bbr 759 gnl|Sanger_518|bbronchi_Contig2526 e-140SpoT Bpe 759 gnl|Sanger_520|Bpertussis_Contig236 e-140RelA Aac 743 gnl|OUACGT_714|Aactin_Contig233 e-136SpoT Ngo 718 gnl|OUACGT_485|Ngon_Contig1 e-134RelA Hdu 735 gnl|HTSC_730|ducreyi e-133RelA Ngo 737 gnl|OUACGT_485|Ngon_Contig1 e-129RelA Bpe 737 gnl|Sanger_520|Bpertussis_Contig301 e-120SpoT Hdu 569 (fragment) gnl|HTSC_730|ducreyi e-111Rel1 Pgi 762 gnl|TIGR|Pgingivalis_GPGcon e-114Rel Ccr 743 gnl|TIGR|Ccrescentus_12574 e-113SpoT Lpn 530 (fragment) gnl|CUCGC_446|lpneumo_MF91102397 e-94Rel2 Pgi 746 gnl|TIGR|Pgingivalis_GPGcon e-81
Table 6 RpoB sequences compared in this study For designations ofspecies see Table 1
Designation Length Accession number
RpoB Eco 1342 RNEBCRpoB Bsu 1193 P37870RpoB Tpa 1178 O83269RpoB Bsp 1342 BAB12761RpoB Rpr 1374 O52271RpoB Ctr 1252 O84317RpoB Cpn 1252 BAA98291RpoB Hin 1343 P43738RpoB Sau 1182 P47768RpoB Mpn 1391 P78013RpoB Bbu 1155 AAB91501RpoB Mtu 1172 CAB09390
Table 7 RpoC sequences compared in this study
Designation Length Accession number
RpoC Eco 1407 RNECCRpoC Bsu 1199 P37871RpoC Sau 1057 P47770RpoC Mtu 1316 P47769RpoC Bsp 1407 BAB12760RpoC Hin 1415 P43739RpoC Rpr 1372 Q9ZE20RpoC Ctr 1396 O84316RpoC Cpn 1393 Q9Z999RpoC Tpa 1416 O83270RpoC Bbu 1377 O51349RpoC Mpn 1290 P75271
supplementary) amino acid sequences were performedOur Supplementary Material (httpjmmbnetsupplementary) shows that the common motif in purinenucleotide-binding proteins described above is present alsoin RpoB sequences of obligately parasitic bacteria and inthe symbiont Buchnera sp APS Large insertions around
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
594 Mittenhuber
alterations This might indicate that the binding sites forppGpp are still present in the β and β subunits of thesespecies
Discussion
Absence of ppGpp in Archaea Presence of ppGpp inPlantsGenes encoding (p)ppGpp synthetases are absent in allcompletely sequenced genomes of Archaea Sinceeubacterial RNA polymerase is the target for the regulatoryaction of (p)ppGpp and since archaea possess atranscriptional apparatus similar to that of eukaryotesinstead (Langer et al 1995) this difference betweenbacteria and archae might explain this absence Starvationstudies in halobacteria demonstrated stringent control ofstable RNA biosynthesis but both growth rate control andstringent control are probably governed by mechanismsthat operate in the absence of ppGpp (Cimmino et al 1993Scoarughi et al 1995)
The finding that Arabidopsis possesses RelASpoThomologs (van der Biezen et al 2000) might indicate thatplants have retained some aspects of the bacterialtranscription apparatus Indeed in Arabidopsis proteinssimilar to bacterial sigma factors (σ70) have been identified
as products of genes specifically expressed in leafs anddestined for chloroplasts (Isono et al 1997) Alsochloroplast genomes encode subunits of bacterial-typeRNA polymerases (Sugiura et al 1998 Turmel et al 1999Allison 2000) Furthermore ppGpp was detected in thealga Chlamydomonas reinhardtii (Heizmann and Howell1978)
Syntheny ConsiderationsSynteny studies (comparison of the relative positions ofgenes in the genome of different organisms) might be asource for auxiliary information to identify orthologousgenes in genome sequencing projects (Huynen and Bork1998) In a systematic comparison using 256 operons ofE coli and 100 operons of B subtilis as query sequencesItoh et al (1999) tried to identify operons showing aconserved order of orthologous genes in 11 completegenome sequences These authors found that operonstructures are very rarely conserved The sequences ofthe relA and spoT operons of E coli (but not the sequenceof the corresponding rel gene of B subtilis) were also usedas query sequences This analysis showed that no genomecontains operons showing exactly the same organizationas the spoT and relA operons of E coli However due tothe complexity of the spoT operon the function of this
Figure 4 Rectangular cladogram of 58 RelA SpoT and Rel proteins Four different groups can be distinguished Numbers on the branches indicated arebootstrap values as described in the legend to Figure 1
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 595
Figure 5 Radial phylogenetic tree of 58 RelA SpoT and Rel proteins
operon was misclassified as ldquoDNARNA processingrdquo in thisstudy For this reason a short description of the relA andspoT operons of E coli and the rel gene region of B subtilisis given
In E coli the relA gene is the first gene in an operoncontaining also the genes encoding the MazEF (ma-zehebrew for ldquowhat is itrdquo) antitoxintoxin module (Aizenmanet al 1996) Two promoters of the mazEF genes are
negatively autoregulated by MazE and MazF andexpression of the module is positively regulated by FIS(Marianovsky et al 2001) In B subtilis and other gram-positive bacteria the putative mazEF locus is locatedupstream of the sigB operon (Mittenhuber 1999) encodingthe general stress response (Hecker and Voumllker 1998)sigma factor σB and its regulatory proteins (Wise and Price1995) A third gene in the relA operon mazG encodes a
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
596 Mittenhuber
protein with unknown function A BLAST search revealedthat similar genes are present in many bacterial genomes(data not shown) They are designated as similiar to ahypothetical 261 kDa protein named YBL1 (accessionnumber P33653) of Streptomyces cacaoi (Urabe andOgawara 1992)
The spoT gene of Escherichia coli is the third gene inan operon consisting of five genes The order is gmk-rpoZ-spoT-trmH-recG Gmk encodes guanylate kinase (Gentryet al 1993) rpoZ the Ω-subunit of RNAP (Gentry andBurgess 1989) trmH a tRNA modifying enzyme (tRNA(Gm18) 2-O-methyltransferase) (Persson et al 1997) andrecG a DNA helicase (Lloyd and Sharples 1991 Kalmanet al 1992) Interestingly the first three genes are linkedto guanosine metabolism whereas the trmH and recG geneproducts play a role in RNA and DNA processing Lloydand Sharples (1991) identified a putative promoter nearthe end of spoT indicating that trmH and recG might betranscribed as separate unit
The organization of the rel loci of B subtilis M lepraeM tuberculosis S equisimilis and S coelicolor is similar(Wendrich and Marahiel 1997 Avarbock et al 1999)Upstream of the rel gene the apt genes (encoding adeninephosphoryltransferase catalyzing the formation ofadenosine monophosphate from adenine andphosphoribosyl pyrophosphate in a salvage reaction) andgenes encoding components of the protein exportmachinery (secDF) are located in the same direction anddownstream of rel the cypH genes encoding cyclophilin(a peptidyl-prolyl cis-trans isomerase) are located in theopposite direction
Proposal for the Evolution of the Rel RelA and SpoTProteinsThe Mollicutes (Mycoplasma and relatives) which arenormally associated with the BacillusClostridium group ofgram-positive bacteria form a distinct outgroup in the treesThese organisms undergo faster rates of evolution andquite regularly form outgroups in phylogenetic trees oforthologous protein sequences (Eisen 1998)
With the exception of Neisseria and Bordetella speciesboth belonging to the β-proteobacteria two different genesencoding enzymes involved in (p)ppGpp synthesis anddegradation namely RelA and SpoT are only found in theszlig and γ subdivision of proteobacteria The SpoT genes ofthis group are related to genes encoding the actinobacterialgroup and the BacillusClostridium group of Rel proteins(Figures 4 and 5 see also the E-values in Table 3) Sincecomplete protein sequences were compared theseparation of the RelA proteins from the Rel and SpoTproteins is most probably due to the fact that the Rel andSpoT proteins possess ppGpp hydrolase activity whereasthe RelA proteins lack this activity It is also interesting tonote that the paralogous genes of individual species arenot closely related to each other Multiple parallel geneduplications in individual species as origin of the relA andspoT genes can therefore be excluded
Facilitated by the knowledge of the enzymatic activitiesof the Rel RelA and SpoT proteins and based on somelogical speculation a model for evolution of the RelA andSpoT proteins within the β and γ subdivision ofproteobacteria is proposed
(1) The common ancestor of the β- and γ-proteobacteriapossessed a rel gene of actinobacterial origin The spoTgene evolved from this gene under adaptation of the geneproduct to carbon and fatty acid starvation(2) Following gene duplication the additional copy of thisancestral gene is able to adopt to new functions In thiscase this adoption of new function was most probablyinactivation of the HD domain loss of (p)ppGpp hydrolyzingactivity and adaptation to amino acid starvation This copyevolved to the relA gene of β- and γ-proteobacteria
The Special Case of P gingivalisP gingivalis however represents an interesting exceptionfrom the scenario described in the previous paragraphThis organism possesses a relA and a spoT gene (Sen etal 2000 httpjmmbnetsupplementary) which arehowever not related to the proteobacterial relA and spoTgenes (Figure 3) Both paralogs which are named Rel1Pgi and Rel2 Pgi in this paper are closely related to eachother and to Rel of Chlorobium tepidum (Figure 3) In Pgingivalis the evolution of the paralogous rel1 (spoT-related) and rel2 (relA-related) occured independently ofthe proteobacterial relA and spoT genes At the moment itis impossible to decide whether this paralogy representsan example of convergent evolution Alternatively lateralgene transfer of short catalytically active domains of ppGppsynthetaseshydrolases might have been involved in theevolution of the rel1 and rel2 genes of P gingivalis In orderto solve this question more sequences encoding ppGppsynthetaseshydrolases from representatives of the CFB(Cytophaga-Flavobacterium-Bacteroides) phylum andgreen sulfur bacteria should be determined and compared(see also next paragraph) Alternatively a carefulphylogenetic analysis of the individual domains of the RelRelA and SpoT proteins (Aravind and Koonin 1998) whichis however beyond the scope of this paper might help tosolve this issue Such an analysis has been performed forthe separate domains of the conserved HSP70 (DnaK)family Striking differences in the relative rate of amino acidreplacement in different rates were observed and wereinterpreted as evidence for functional divergence (Hughes1993)
ppGpp SynthetasesHydrolases and BacterialEvolutionA drastically different view of bacterial evolution has beenrecently postulated by Gupta (2000a b) This hypothesisplaces the γ subdivision of proteobacteria at the end of aevolutionary linear scheme According to this theory the γsubdivision of proteobacteria constitutes the most recentlyevolved bacterial group Therefore specialized paralogousspoT and relA genes might be a relatively new invention ofbacterial evolution which might explain this limiteddistribution pattern among the β- and γ-proteobacteriaSimilarly the (maybe more efficient) vitamin B6 biosynthesispathway of E coli is mainly restricted to the γ subdivisionwhereas another widely conserved biochemicallyuncharacterized pathway is presumably operating in otherspecies (Mittenhuber 2001) Other unique features of theγ-proteobacteria were recently recognized by Margolin(2000) in his study on prokaryotic cell division Genesencoding the ZipA FtsL and FtsN proteins are only found
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 597
in genomes of completely sequenced γ-proteobacteriaVery recently Eisen (2000) noted that the currentlyavailable datasets of completely sequenced genomes arenot representative of evolutionary diversity It might bepredicted that the availability of completely sequencedgenomes of proteobacterial species belonging tosubdivisions other than β and γ might help to identify theprecursor of the proteobacterial relA and spoT genes
ppGpp is Not Present in Obligately IntracellularSpeciesThe species lacking a rel-like gene require living cells fortheir replication and growth and cannot be cultivated invitro The absence of rel-like genes in genomes of obligatelyparasitic organisms was also detected in a comparativegenome analysis of B burgdorferi and T pallidum(Subramanian et al 1999) R prowazekii possesses afew short pieces of a pseudogene which show strongsequence similarity to the relAspoT homologs (Anderssonet al 1998 Zomorodipour and Andersson 1999) whereasno rel-like gene is present in the C trachomatis genome(Stephens et al 1998) Interestingly some cultivableintracellular pathogens (eg B burgdorferi andMycoplasma species among others) possess rel-likegenes In most cases cultivation of bacteria implies thatthe inoculum is able to form colonies Investigations onvertical sections of E coli colonies show that the colony iscomposed of different layers of bacteria containing alsononviable bacteria (Shapiro 1994) It is very likely thatmany cells in a developed colony are starved for nutrientsand that the stringent response is switched on in thesecells It might be extremely interesting to investigatewhether a functional connection between in vitro cultivabilityof bacterial species and the absence of genes encoding(p)ppGpp synthetaseshydrolases in obligate parasites canbe established experimentally
Experimental ProceduresUsing the deduced protein sequences of the relA and spoTgenes of E coli and of the rel gene of B subtilis as querysequences BLAST searches (Altschul et al 1997) of thenon-redundant protein database nrdb95 (Holm and Sander1998) were performed at httpdoveembl-heidelbergdeBlast2 using the blastp program BLAST searches ofunfinished microbial genomes using the deduced proteinsequence of the rel gene of B subtilis as query wereperformed at httpwwwncbinlmnihgovMicrob_blastunfinishedgenomehtml using the program tblastn RawDNA sequences were translated using the programldquoTranslate toolrdquo at httpwwwexpasychtoolsdnahtmlThe COG database can be assessed at httpwwwncbinlmnihgovCOG Alignments were generatedusing CLUSTALW (Thompson et al 1994) at httpwwwebiacukclustalw Radial phylogenetic trees wereconstructed using the data from the alignments with thehelp of the program TreeView (httptaxonomyzoologyglaacukrodtreeviewhtml Page 1996) and edited usingthe program Metafile Companion (httpwwwcompanionsoftwarecom) Bootstrapping of datasets wasperformed by the neigbor-joining (NJ) method on the basisof the Poisson-corrected amino acid distance (daa) (Saitouand Nei 1987 for a discussion of different statistical
methods and their applications see Hughes 1999 Nei andKumar 2000) using MEGA 20 (httpwwwmegasoftwarenet Kumar et al 2001)
Acknowledgements
This work was performed in the lab of Michael Heckerwhose support is gratefully acknowledged I thank MichaelHecker Ulrich Zuber and an anonymous referee forcomments on the manuscript Sequences of unfinishedmicrobial genomes were obtained at the website of ldquoTheInstitute of Genomic Researchrdquo (TIGR) at httpwwwtigrorg Sequencing of unfinished genomes is inprogress at several institutions funded by a variety oforganizations This listing includes ldquoname of organism(institution(s) source of funding)rdquo and can be also viewedat httpwwwtigrorgtdbmdbmdbinprogresshtml Aactinomycetemcomitans (University of Oklahoma NationalInstitute of Dental Research (NIDR)) B anthracis (TIGROffice of Naval Research (ONR)Department of Energy(DOE)National Institute of Allergy and Infectious Diseases(NIAD)) B halodurans (Japan Marine Science andTechnology Center) B stearothermophilus (University ofOklahoma National Science Foundation (NSF) Bbronchiseptica B pertussis C difficile N meningitidis Ypestis (Sanger Centre Beowulf Genomics) C crescentusC tepidum D ethenogenes D vulgaris S putrefaciensT ferrooxidans (TIGRDOE) C acetobutylicum (GenomeTherapeutics DOE) C diphteriae (Sanger CentreWorldHealth Organization (WHO)Public Health LaboratoryBeowulf Genomics) G sulfurreducens (TIGRUniversityof Massachusetts AmherstDOE) H ducreyi (NIAID) Lpneumophila (Columbia Genome Center NIAID) Mavium V cholerae (TIGRNIAID) M bovis (Sanger CentreInstitut PasteurVLA Weybridge MAFFBeowulfGenomics) M leprae (Sanger Centre The New YorkCommunity Trust) N gonorrhoeae (University ofOklahoma NIAID) P multocida (University of MinnesotaUS Department of Agriculture - National Reserach Initiative(USDA-NRI)Minnesota Turkey Growers Association) Pgingivalis (TIGRForsyth Dental Center NIDR) Paeruginosa (University of WashingtonPathoGenesisCystic Fibrosis FoundationPathoGenesis) P putida (TIGRGerman Consortium DOEBundesministerium fuumlr Bildungund Forschung (BMBF) S typhi (Sanger CentreWellcome Trust) S aureus (TIGR NIAIDMerck GenomeResearch Institute (MGRI)) S pneumoniae (TIGR TIGRNIAIDMGRI) S coelicolor (Sanger CentreJohn InnesCentre Biotechnology and Biological Sciences ResearchCouncil (BBSRC)Beowulf Genomics)
References
Aizenman E Engelberg-Kulka H and Glaser G 1996An Escherichia coli chromosomal ldquoaddiction modulerdquoregulated by guanosine [corrected] 35-bispyrophosphate a model for programmed bacterial celldeath Proc Natl Acad Sci USA 93 6059-6063
Allison LA 2000 The role of sigma factors in plastidtranscription Biochimie 82 537-548
Altschul SF Madden TL Schaffer AA Zhang JZhang Z Miller W and Lipman DJ 1997 Gapped
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
598 Mittenhuber
BLAST and PSI-BLAST a new generation of proteindatabase search programs Nucleic Acids Res 25 3389-3402
Andersson SG and Kurland CG 1998 Reductiveevolution of resident genomes Trends Microbiol 6 263-268
Andersson SG Zomorodipour A Andersson JOSicheritz-Ponteacuten T Alsmark UC Podowski RMNaslund AK Eriksson AS Winkler HH and KurlandCG 1998 The genome sequence of Rickettsiaprowazekii and the origin of mitochondria Nature 296133-140
Aravind L and Koonin EV 1998 The HD domain definesa new superfamily of metal-dependentphosphohydrolases Trends Biochem Sci 23 469-472
Aravind L Dixit VM and Koonin EV 1999 The domainsof death evolution of the apoptosis machinery TrendsBiochem Sci 24 47-53
Avarbock D Salem J Li LS Wang ZM and RubinH 1999 Cloning and characterization of a bifunctionalrelAspoT homologue from Mycobacterium tuberculosisGene 233 261-269
Cashel M and Gallant J 1969 Two compoundsimplicated in the function of the RC gene of Escherichiacoli Nature 221 838-841
Cashel M Gentry DR Hernandez VJ and Vinella D1996 The stringent response In Escherichia coli andSalmonella Cellular and Molecular Biology FCNeidhardt (Editor-in- Chief) ASM Press Washington DCp 1458-1496
Chakraburtty R White J Takano E and Bibb M 1996Cloning characterization and disruption of a (p)ppGppsynthetase gene (relA) of Streptomyces coelicolor A3Mol Microbiol 19 357-368
Chakraburtty R and Bibb M 1997 The ppGpp synthetasegene (relA) of Streptomyces coelicolor A3(2) plays aconditional role in antibiotic production and morphologicaldifferentiation J Bacteriol 179 5854-5861
Chatterji D Fujita N and Ishihama A 1998 Themediator for stringent control ppGpp binds to the β-subunit of Escherichia coli RNA polymerase Genes toCells 3 279-287
Cimmino C Scoarughi GL and Donini P 1993Stringency and relaxation among the halobacteria JBacteriol 175 6659-6662
Crouzet J Levy-Schil S Cameron B Cauchois LRigault S Rouyez MC Blanche F Debussche Land Thibaut D 1991 Nucleotide sequence and geneticanalysis of a 131-kilobase-pair Pseudomonasdenitrificans DNA fragment containing five cob genes andidentification of structural genes encoding Cob(I)alaminadenosyltransferase cobyric acid synthase andbifunctional cobinamide kinase-cobinamide phosphateguanylyltransferase J Bacteriol 173 6074-6087
Eisen JA 1998 Phylogenomics improving functionalpredictions for uncharacterized genes by evolutionaryanalysis Genome Res 8 163-167
Eisen JA 2000 Assessing evolutionary relationshipsamong microbes from whole-genome analysis CurrOpin Microbiol 3 475-480
Eyles SJ and Gierasch LM 2000 Multiple roles of prolylresidues in structure and folding J Mol Biol 301 737-
747Eymann C Mittenhuber G and Hecker M 2001 The
stringent response σH-dependent gene expression andsporulation in Bacillus subtilis Mol Gen Genet 264 913-923
Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783-791
Foissac X Danet JL Zreik L Gandar J NourrisseauJG Bove JM and Garnier M 2000 Cloning of thespoT gene of ldquoCandidatus Phlomobacter fragariaerdquo anddevelopment of a PCR-restriction fragment lengthpolymorphism assay for detection of the bacterium ininsects Appl Environ Microbiol 66 3474-3480
Fraser CM Norris SJ Weinstock GM White OSutton GG Dodson R Gwinn M Hickey EKClayton R Ketchum KA Sodergren E HardhamJM McLeod MP Salzberg S Peterson J KhalakH Richardson D Howell JK Chidambaram MUtterback T McDonald L Artiach P Bowman CCotton MD Fujii C Garland S Hatch B Horst KRoberts K Watthey L Weidman J Smith HO andVenter JC 1998 Complete genome sequence ofTreponema pallidum the syphilis spirochete Science 281375-388
Gentry D Bengra C Ikehara K and Cashel M 1993Guanylate kinase of Escherichia coli K-12 J Biol Chem268 14316-14321
Gentry D Li T Rosenberg M and McDevitt D 2000The rel gene is essential for in vitro growth ofStaphylococcus aureus J Bacteriol 182 4995-4997
Gentry DR and Burgess RR 1989 rpoZ encoding theomega subunit of Escherichia coli RNA polymerase is inthe same operon as spoT J Bacteriol 171 1271-1277
Gentry DR Hernandez VJ Nguyen LH Jensen DBand Cashel M 1993 Synthesis of the stationary-phasesigma factor σS is positively regulated by ppGpp JBacteriol 175 7982-7989
Gentry DR and Cashel M 1995 Cellular localization ofthe Escherichia coli SpoT protein J Bacteriol 177 3890-3893
Gentry DR and Cashel M 1996 Mutational analysis ofthe Escherichia coli spoT gene identifies distinct butoverlapping regions involved in ppGpp synthesis anddegradation Mol Microbiol 19 1373-1384
Glass RE Jones ST and Ishihama A 1986 Geneticstudies on the β subunit of Escherichia coli RNApolymerase VII RNA polymerase is a target for ppGppMol Gen Genet 203 265-268
Gupta RS 2000a The natural evolutionary relationshipsamong prokaryotes Crit Rev Microbiol 26 111-131
Gupta RS 2000b The phylogeny of proteobacteriarelationships to other eubacterial phyla and eukaryotesFEMS Microbiol Rev 24 367-402
Harris BZ Kaiser D and Singer M 1998 The guanosinenucleotide (p)ppGpp initiates development and A-factorproduction in Myxococcus xanthus Genes Dev 12 1022-1035
Haseltine WA and Block R 1973 Synthesis ofguanosine tetra- and pentaphosphate requires thepresence of a codon specific uncharged transferribonucleic acid in the acceptor site of ribosomes ProcNatl Acad Sci USA 70 1564-1568
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
Rel RelA and SpoT Proteins 599
Hecker M and Voumllker U 1998 Non-specific general andmultiple stress resistance of growth-restricted Bacillussubtilis cells by the expression of the σB regulon MolMicrobiol 29 1129-1136
Heizmann P and Howell SH 1978 Synthesis of ppGppand chloroplast RNA in Chlamydomonas reinhardiBiochem Biophys Acta 517 115-124
Hesketh A Sun J and Bibb M 2001 Induction of ppGppsynthesis in Streptomyces coelicolor A3(2) grown underconditions of nutritional sufficiency elicits actII-orf4transcription and actinorhodin biosynthesis MolMicrobiol 39 136-144
Hernandez VJ and Bremer H 1991 Escherichia colippGpp synthetase II activity requires spoT J Biol Chem266 5991-5999
Hernandez VJ and Cashel M 1995 Changes inconserved region 3 of Escherichia coli σ70 mediateppGpp-dependent functions in vivo J Mol Biol 252 536-549
Holm L and Sander C 1998 Removing near-neighbourredundancy from large protein sequence collectionsBioinformatics 14 423-429
Hoyt S and Jones GH 1999 RelA is required foractinomycin production in Streptomyces antibioticus JBacteriol 181 3824-3829
Hughes AL 1993 Nonlinear relationships amongevolutionary rates identify regions of functional divergencein heat-shock protein 70 genes Mol Biol Evol 10 243-255
Hughes AL 1999 Adaptive evolution of genes andgenomes Oxford University Press Oxford UK
Huynen M and Bork P 1998 Measuring genomeevolution Proc Natl Acad Sci USA 95 5849-5856
Isono K Shimizu M Yoshimoto K Niwa Y Satoh KYokota A and Kobayashi H 1997 Leaf-specificallyexpressed genes for polypeptides destined forchloroplasts with domains of σ70 factors of bacterial RNApolymerase in Arabidopsis thaliana Proc Natl Acad SciUSA 94 14948-14953
Itoh T Takemoto K Mori H and Gojobori T 1999Evolutionary instability of operon structures disclosed bysequence comparisons of complete microbial genomesMol Biol Evol 16 332-346
Kalman M Murphy H and Cashel M 1992 Thenucleotide sequence of recG the distal spo operon genein Escherichia coli K-12 Gene 110 95-99
Kalman S Mitchell W Maranthe R Lammel C FanJ Hyman RW Olinger L Grimwood J Davis RWand Stephens RS 1999 Comparative genomes ofChlamydia pneumoniae and C trachomatis NatureGenet 21 385-389
Kumar S Tamura K Jakobsen IB and Nei M 2001MEGA 2 Molecular evolutionary genetics analysissoftware Bioinformatics (submitted)
Kvint K Farewell A and Nystroumlm T 2000 RpoS-dependent promoters require guanosine tetraphosphateeven in the presence of high levels of σS J Biol Chem275 14795-14798
la Cour TF Nyborg J Thirup S and Clark BF 1985Structural details of the binding of guanosine diphosphateto elongation factor Tu from E coli as studied by X-raycrystallography EMBO J 4 2385-2388
Langer D Hain J Thuriaux P and Zillig W 1995Transcription in archaea similarity to that in eucarya ProcNatl Acad Sci USA 92 5768-5772
Lloyd RG and Sharples 1991 Molecular organizationand nucleotide sequence of the recG locus of Escherichiacoli K-12 J Bacteriol 173 6837-6843
Margolin W 2000 Themes and variations in prokaryoticcell division FEMS Microbiol Rev 24 531-548
Marianovsky I Aizenman E Engelberg-Kulka H andGlaser G 2001 The regulation of the Escherichia colimazEF promoter involves an unusual alternatingpalindrome J Biol Chem 276 5975-5984
Martinez-Costa OH Arias P Romero NM Parro VMellado RP and Malpartida F 1996 A relAspoThomologous gene from Streptomyces coelicolor A3(2)controls antibiotic biosynthesis genes J Biol Chem 27110627-10634
Martinez-Costa OH Fernandez-Moreno MA andMalpartida F 1998 The relAspoT homologous gene inStreptomyces coelicolor encodes both ribosome-dependent (p)ppGpp-synthesizing and -degradingactivities J Bacteriol 180 4123-4132
Mechold U Cashel M Steiner K Gentry D and MalkeH 1996 Functional analysis of a relAspoT gene homologfrom Streptococcus equisimilis J Bacteriol 178 1401-1411
Mechold U and Malke H 1997 Characterization of thestringent and relaxed responses of Streptococcusequisimilis J Bacteriol 179 2658-2667
Metzger S Sarubbi E Glaser G and Cashel M 1989Protein sequences encoded by the relA and the spoTgenes of Escherichia coli are interrelated J Biol Chem264 9122-9125
Mittenhuber G 1999 Occurence of MazEF-like antitoxintoxin systems in bacteria J Mol Microbiol Biotechnol1 295-302
Mittenhuber G 2001 Phylogenetic analyses andcomparative genomics of vitamin B6 (pyridoxine) andpyridoxal phosphate biosynthesis pathways J MolMicrobiol Biotechnol 3 1-20
Murray KD and Bremer H 1996 Control of spoT-dependent ppGpp synthesis and degradation inEscherichia coli J Mol Biol 259 41-57
Nei M and Kumar S 2000 Molecular evolution andphylogenetics Oxford University Press Oxford UK
Noel L Moores TL van der Biezen EA Parniske MDaniels MJ Parker JE and Jones JD 1999Pronounced intraspecific haplotype divergence at theRPP5 complex disease resistance locus of ArabidopsisPlant Cell 11 2099-2112
Oumlstling J Flardh K and Kjelleberg S 1995 Isolation ofa carbon starvation regulatory mutant in a marine Vibriostrain J Bacteriol 177 6978-6982
Page RDM 1996 TREEVIEW An application to displayphylogenetic trees on personal computers CABIOS 12357-358
Persson BC Jaumlger G and Gustafsson C 1997 ThespoU gene of Escherichia coli the fourth gene of the spoToperon is essential for tRNA (Gm18) 2-O-methyltransferase activity Nucleic Acids Res 25 4093-4097
Primm TP Andersen SJ Mizrahi V Avarbock D
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15
600 Mittenhuber
Rubin H and Barry C E III 2000 The stringentresponse of Mycobacterium tuberculosis is required forlong-term survival J Bacteriol 182 4889-4898
Reddy PS Raghavan A and Chatterji D 1995Evidence for a ppGpp-binding site on E coli RNApolymerase proximity relationship with the rifampicinbinding domain Mol Microbiol 15 255-265
Seyfzadeh M Keener J and Nomura M 1993 spoT-dependent accumulation of guanosine tetraphosphate inresponse to fatty acid starvation in Escherichia coli ProcNatl Acad Sci USA 90 11004-11008
Saitou N and Nei M 1987 The neighbor-joining methoda new method for reconstructing phylogenetic trees MolBiol Evol 4 406-425
Sato S Nakamura Y Kaneko T Asamizu E andTabata S 1999 Complete structure of the chloroplastgenome of Arabidopsis thaliana DNA Res 6 283-290
Shapiro JA 1994 Pattern and control in bacterial colonydevelopment Sci Prog 76 399-424
Scoarughi GL Cimmino C and Donini P 1995 Lackof production of (p)ppGpp in Halobacterium volcanii underconditions that are effective in the eubacteria J Bacteriol177 82-85
Sen K Hayashi J and Kuramitsu HK 2000Characterization of the relA gene of Porphyromonasgingivalis J Bacteriol 182 3302-3304
Shigenobu S Watanabe H Hattori M Sakaki Y andIshikawa H 2000 Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp APS Nature407 81-86
Stephens RS Kalman S Lammel C Fan J MaratheR Aravind L Mitchell W Olinger L Tatusov RLZhao Q Koonin EV and Davis RW 1998 Genomesequence of an obligate intracellular pathogen of humansChlamydia trachomatis Science 282 754-759
Stent GS and Brenner S 1961 A genetic locus for theregulation of ribonucleic acid synthesis Proc Natl AcadSci USA 47 2005-2014
Subramanian G Koonin EV and Aravind L 2000Comparative genome analysis of the pathogenicspirochetes Borrelia burgdorferi and Treponema pallidumInfect Immun 68 1633-1648
Sugiura M Hirose T and Sugita M 1998 Evolution andmechanism of translation in chloroplasts Ann RevGenet 32 437-459
Tatusov RL Koonin EV and Lipman DJ 1997 Agenomic perspective on protein families Science 278631-637
Tatusov RL Galperin MY Natale DA and KooninEV 2000 The COG database a tool for genome-scaleanalysis of protein functions and evolution Nucleic AcidsRes 28 33-36
Thompson JD Higgins DG and Gibson TJ 1994CLUSTAL W improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting position-specific gap penalties and weightmatrix choice Nucleic Acids Res 22 4673-4680
Toulokhonov II Shulgina I and Hernandez VJ 2001Binding of the effector ppGpp to E coli RNA polymeraseis allosteric modular and occurs near the N-terminus ofthe β subunit J Biol Chem 276 1220-1225
Turko IV Francis SH and Corbin JD 1998 Potential
roles of conserved amino acids in the catalytic domain ofthe cGMP-binding cGMP-specific phosphodiesterase JBiol Chem 273 6460-6466
Turmel M Otis C and Lemieux C 1999 The completechloroplast DNA sequence of the green algaNephroselmis olivacea insights into the architecture ofancestral chloroplast genomes Proc Natl Acad Sci USA96 10248-10253
Urabe H and Ogawara H 1992 Nucleotide sequenceand transcriptional analysis of activator-regulator proteinsfor β-lactamase in Streptomyces cacaoi J Bacteriol 1742834-2842
van der Biezen EA and Jones JD 1998 The NB-ARCdomain a novel signalling motif shared by plant resistancegene products and regulators of cell death in animalsCurr Biol 8 R226-R227
van der Biezen EA Sun J Coleman MJ Bibb MJand Jones JD 2000 Arabidopsis RelASpoT homologsimplicate (p)ppGpp in plant signaling Proc Natl AcadSci USA 97 3747-3752
Wehmeier L Schaumlfer A Burkovski A Kraumlmer RMechold U Malke H Puumlhler A and Kalinowski J1998 The role of the Corynebacterium glutamicum relgene in (p)ppGpp metabolism Microbiology 144 1853-1862
Wendrich TM and Marahiel MA 1997 Cloning andcharacterization of a relAspoT homologue from Bacillussubtilis Mol Microbiol 26 65-79
Wendrich TM Beckering CL and Marahiel MA 2000Characterization of the relAspoT gene from Bacillusstearothermophilus FEMS Microbiol Lett 190 195-201
Wise AA and Price CW 1995 Four additional genesin the sigB operon of Bacillus subtilis that control activityof the general stress factor σB in response toenvironmental signals J Bacteriol 177 123-133
Xiao H Kalman M Ikehara K Zemel S Glaser Gand Cashel M 1991 Residual guanosine 35-bispyrophosphate synthetic activity of relA null mutantscan be eliminated by spoT null mutations J Biol Chem266 5980-5990
Zhang G Campbell EA Minakhin L Richter CSeverinov K and Darst SA 1999 Crystal structure ofThermus aquaticus RNA polymerase at 3 Aring resolutionCell 98 811-824
Zomorodipour A and Andersson SGE 1999 Obligateintracellular parasites Rickettsia prowazekii andChlamydia trachomatis FEBS Lett 452 11-15