reprint microbiology-uk aug 2014

Genetic and mutational characterization of the smallalarmone synthetase gene relV of Vibrio cholerae

Shreya Dasgupta, Pallabi Basu, Ritesh Ranjan Pal, Satyabrata Bagand Rupak K. Bhadra

Correspondence

Rupak K. Bhadra

[email protected]

Received 25 March 2014

Accepted 26 June 2014

Infectious Diseases and Immunology Division, CSIR-Indian Institute of Chemical Biology,Kolkata-700 032, India

In Vibrio cholerae, the causative agent of cholera, products of three genes, relA, spoT and relV,

govern nutritional stress related stringent response (SR). SR in bacteria is critically regulated by

two intracellular small molecules, guanosine 39-diphosphate 59-triphosphate (pppGpp) and

guanosine 39,59-bis(diphosphate) (ppGpp), collectively called (p)ppGpp or alarmone. Evolution of

relV is unique in V. cholerae because other Gram-negative bacteria carry only relA and spoT

genes. Recent reports suggest that RelV is needed for pathogenesis. RelV carries a single

(p)ppGpp synthetase or RelA-SpoT domain (SYNTH/RSD) and belongs to the small alarmone

synthetase (SAS) family of proteins. Here, we report extensive functional characterizations of the

relV gene by constructing several deletion and site-directed mutants followed by their controlled

expression in (p)ppGpp0 cells of Escherichia coli or V. cholerae. Substitution analysis indicated

that the amino acid residues K107, D129, R132, L150 and E188 of the RSD region of RelV are

essential for its activity. While K107, D129 and E188 are highly conserved in RelA and SAS

proteins, L150 appears to be conserved in the latter group of enzymes, and the R132 residue was

found to be unique in RelV. Extensive progressive deletion analysis indicated that the amino acid

residues at positions 59 and 248 of the RelV protein are the functional N- and C-terminal

boundaries, respectively. Since the minimal functional length of RelV was found to be 189 aa,

which includes the 94 aa long RSD region, it seems that the flanking residues of the RSD are also

important for maintaining the (p)ppGpp synthetase activity.

INTRODUCTION

Vibrio cholerae, the causative agent of the fatal diarrhoealdisease cholera, resides in various aquatic bodies andinfects humans through contaminated food and water.While staying outside and inside of the human intestine,the pathogen faces numerous physico-chemical stressesamongst which nutritional starvation is a crucial para-meter. The global changes in gene expression associatedwith nutritional deprivation in bacteria, called stringentresponse (SR), are triggered by intracellular accumulationof two small molecules, guanosine 39-diphosphate 59-triphosphate (pppGpp) and guanosine 39,59-bis(dipho-sphate) (ppGpp), together called (p)ppGpp or alarmone.

SR is characterized by repression of rRNA transcription,positive regulation of amino acid biosynthesis, readjust-ment of metabolic pathways according to physiologicalrequirements and induction of stationary phase genesneeded for survival (Cashel et al., 1996; Chatterji & Ojha,2001; Choy, 2000; Jishage et al., 2002; Srivatsan & Wang,2008; Stephens et al., 1975; Traxler et al., 2011). InEscherichia coli and in other Gram-negatives, the productsof relA and spoT genes synthesize and degrade (p)ppGpp,respectively (Atkinson et al., 2011; Cashel et al., 1996;Mittenhuber, 2001). RelA is a ribosome-associated proteinand synthesizes (p)ppGpp under amino acid starvation(Cashel et al., 1996; Haseltine & Block, 1973; Wendrich et al.,2002). On the other hand, SpoT is a bifunctional enzymewith weak synthetase and strong (p)ppGpp hydrolaseactivities (Gentry & Cashel, 1996; Murray & Bremer, 1996;Xiao et al., 1991). SpoT synthesizes (p)ppGpp under carbonand fatty acid starvation (Battesti & Bouveret, 2006;Seyfzadeh et al., 1993; Xiao et al., 1991). Interestingly, thegenomes of Gram-positive organisms carry a single rel genecoding for a bifunctional enzyme Rel with strong (p)ppGppsynthesis and hydrolase activities (Avarbock et al., 2005;Hogg et al., 2004; Mechold et al., 2002; Mittenhuber, 2001).

Abbreviations: ACT, acetolactate synthetase-chorismate mutase-tyrR;pppGpp, guanosine 39-diphosphate 59-triphosphate; ppGpp, guanosine39,59-bis(diphosphate); HD, ppGpp hydrolase domain; PEI, polye-thyleneimine; RSD, RelA-SpoT domain; RSH, RelA/SpoT homologue;SAS, small alarmone synthetase; SDM, site-directed mutagenesis; SR,stringent response; SYNTH, (p)ppGpp synthetase domain; TGS,threonyl-tRNA synthetase-GTPase-SpoT.

One supplementary table and two supplementary figures are availablewith the online version of this paper.

Microbiology (2014), 160, 1855–1866 DOI 10.1099/mic.0.079319-0

079319 G 2014 The Authors Printed in Great Britain 1855

One important feature of the RelA/SpoT/Rel enzymesis that they carry N-terminally located hydrolase and(p)ppGpp synthetase/RelA-SpoT (SYNTH/RSD) domains,which are responsible for enzymic activities. On the otherhand their C-terminally located conserved domains TGS(this domain is conserved in threonyl t-RNA synthetase,GTPases and SpoT proteins) and ACT (this domain isconserved in acetolactate synthase, chorismate mutaseand TyrR receptor proteins) (Aravind & Koonin, 1998;Atkinson et al., 2011; Potrykus & Cashel, 2008), areinvolved in regulation of enzymic activities (Battesti &Bouveret, 2006; Das et al., 2009; Gropp et al., 2001;Mechold et al., 2002; Schreiber et al., 1991). In SpoT/Relenzymes, the hydrolase domain carries highly conservedhistidine-aspartic acid (HD) residues (Aravind & Koonin,1998; Das et al., 2009; Potrykus & Cashel, 2008). However,in RelA, the HD residues are substituted with other aminoacids leading to impaired (p)ppGpp hydrolase activityfunction (Aravind & Koonin, 1998).

Recent reports indicate that apart from canonical relA/rel/spoT genes certain Gram-positive and Gram-negativebacterial genomes may carry additional RSD coding genes(Das et al., 2009; Lemos et al., 2007; Murdeshwar &Chatterji, 2012; Nanamiya et al., 2008). These newly dis-covered genes code for a small protein (usually within~30 kDa in size), hence they are also called small alarmonesynthetase or SAS (Nanamiya et al., 2008). We havereported that the genome of V. cholerae codes for a 259 aaSAS protein, called RelV (Das et al., 2009). It was shownthat RelV could synthesize (p)ppGpp under glucose orfatty acid starvation under DrelADspoT double mutantbackground. However, V. cholerae DrelADspoTDrelV triplemutant was unable to accumulate (p)ppGpp under aminoacid, glucose or fatty acid starvation and thus, they werephenotypically (p)ppGpp0 cells (Das et al., 2009).

It was reported previously that the C-terminal domain ofRelA/Rel proteins may be involved in regulating the N-terminal (p)ppGpp synthetase activity (Gropp et al., 2001;Mechold et al., 2002). However, there are no such distinctN- and C-terminal domains in RelV and currently noinformation is available about the role, if any, of theseregions in maintaining (p)ppGpp synthetase activity. Inthis study, we have done extensive genetic and mutationalanalyses to characterize relV gene function. Site-directedmutagenesis of the RSD region of RelV allowed us toidentify five essential amino acid residues for (p)ppGppsynthetase activity. Through progressive deletions of theN- or C-terminal amino acids of the RelV protein, thefunctional boundaries were determined. Extensive deletionanalysis ultimately helped in determining the 189 aminoacid residues long minimal functional region of the RelVenzyme.

METHODS

Bacterial strains, plasmids and growth conditions. The bacterialstrains and plasmids used are listed in Table 1. In this study a

spontaneous deletion mutant of relV under DrelADspoT background,

strain BS1.1, was isolated by repeated subculturing as done before

(Das et al., 2009), and was designated BSRD10 (Table 1). Sequencing

of the PCR amplified relV region of BSRD10 indicated that there is a

10 bp deletion within the relV ORF, which should express a 172 aa

residue-containing truncated RelV protein (last nine amino acids

originated due to a frame shift mutation) and BSRD10 showed

(p)ppGpp0 phenotype (data not shown). Bacterial strains were

routinely grown in Luria broth (LB; Difco) at 37 uC with vigorous

shaking and for plate culture Luria agar (LA; Difco) was used

(Haralalka et al., 2003). Antibiotics (Sigma-Aldrich) were used at the

following concentrations unless stated otherwise: ampicillin, 100 mg

ml21; kanamycin, 40 mg ml21; spectinomycin, 50 mg ml21; strep-

tomycin, 100 mg ml21; chloramphenicol, 30 mg ml21 for E. coli and

3 mg ml21 for V. cholerae; tetracycline, 10 mg ml21 for E. coli. Bacterial

cells were also grown in M9 minimal (M9M) solution (Sigma-

Aldrich) or agar (1.5 %; Difco) containing 0.4 % glucose as a carbon

source (Das & Bhadra, 2008; Das et al., 2009) and growth of bacterial

cells was usually checked after 24–30 h after incubating the plates at

37 uC. Bacterial strains were preserved at 270 uC in LB containing

20 % sterile glycerol (Haralalka et al., 2003). Before initiation of any

experiment, bacterial strains including mutants were always freshly

inoculated from their –70 uC stock to avoid development of any

suppressor mutations.

Molecular biological methods. Standard molecular biological

methods (Ausubel et al., 1989) were followed unless stated otherwise.

Restriction and nucleic acid-modifying enzymes were purchased from

New England Biolabs, and were used essentially as directed by the

manufacturer. Electrocompetent V. cholerae cells were prepared

essentially as reported earlier (Das & Bhadra, 2008) and transformants

were selected on LA plates containing appropriate antibiotics.

Plasmid constructions. The plasmid pRelVD10 was constructed by

PCR, amplifying only the relV ORF carrying 10 bp deletion (size

802 bp) using the primers VCRVorf-F/VCRVorf-R (Table S1,

available in the online Supplementary Material) and genomic DNA

of the V. cholerae strain BSRD10 (Table 1) followed by digestion with

the enzymes EcoRI/HindIII and cloning in similarly digested vector

pBAD24. For transformation E. coli DH5a cells were used. For

construction of plasmids carrying deletion of the relV ORF at 59 and/

or 39 end, the desired fragment was PCR amplified using different sets

of primers (Table S1) and genomic DNA of the V. cholerae strain

N16961 (Heidelberg et al., 2000) followed by digestion of the

fragment with EcoRI/HindIII and cloning in the similarly digested

vector DNA pBAD24. Forward primers were always designed with an

artificially inserted ATG start codon to amplify 59 end deleted

fragments of relV ORF. Similarly, 39 end deleted relV ORF was always

amplified with the reverse primers carrying an artificially inserted

TAG stop codon. Authenticity of each construct was verified by DNA

sequencing using relevant primers (Table S1).

Site-directed mutagenesis. Site-directed mutagenesis (SDM) of

the relV gene was done using the QuikChange II SDM kit (Stratagene)

and the recombinant plasmid pRelVBAD as the template. Pairs of

complementary oligonucleotides (Table S1) were designed with single

or double position of nucleotide degeneracy to insert the desired

amino acid substitutions at specific positions (Avril et al., 2005; Wu

et al., 2008). Authenticity of desired site-directed mutation was

confirmed by DNA sequencing using relevant primers (Table S1).

Determination of intracellular (p)ppGpp. Intracellular accumula-

tion of (p)ppGpp in V. cholerae and E. coli strains was determined by

labelling glucose-starved cells with [32P]-H3PO4 (100 mCi ml21; BRIT)

followed by TLC essentially as described previously (Das & Bhadra,

2008; Das et al., 2009).

S. Dasgupta and others

1856 Microbiology 160

Reverse transcriptase-PCR assay. For reverse transcriptase (RT)-

PCR assay of the relV gene, total cellular RNA was prepared from

bacterial cells grown in LB medium to an OD600 value of ~1.0 using

TRI Reagent (Sigma) essentially as described by the supplier. RT-PCR

was done with the OneStep RT-PCR kit (Qiagen) essentially as

described previously (Pal et al., 2012) using purified RNA and the

primers 1224int-F/1224int-R (Table S1).

DNA sequencing and computational analyses. DNA sequencing

was done using an ABI3130 genetic analyser (Applied Biosystems

Inc.) as described previously (Das et al., 2009). DNA sequence data

were analysed by using Chromas 1.45 (http://technelysium.com.au/

?page_id=13). DNA sequences were obtained from JCVI (http://www.

jcvi.org/cms/home/) and protein domain information was obtained

from KEGG (www.genome.jp/kegg/). BLASTN and BLASTP programs

were used to search for homologous nucleotide or protein sequences,

respectively, in the database (www.ncbi.nlm.nih.gov). Genomatix

(www.genomatix.de/cgi-bin/dialign/dialign.pl) and CLUSTAL W2 soft-

ware (http://www.ebi.ac.uk/Tools/msa/clustalw2/) were used for the

alignment of protein sequences and further modified through GeneDoc

software version 2.7.000 (http://genedoc.sharewarejunction.com). For

designing primers, Primer3 software (http://bioinfo.ut.ee/primer3-0.4.

0/) was used.

RESULTS AND DISCUSSION

Growth phenotypes of (p)ppGpp0 cells undercontrolled expression of relV

Earlier it was reported that expression of the relV genethrough its natural promoter is toxic in E. coli DrelADspoT(p)ppGpp0 strain CF1693 (Das et al., 2009). We have

Table 1. Strains and plasmids used in this study

Strain or plasmid Relevant genotype and/or phenotype Source/reference

V. cholerae

N16961 Wild-type, lacking hapR function, O1 serogroup, biotype El Tor, Smr Heidelberg et al. (2000)

RRV1 N16961 DrelA DrelV; Kmr Smr Spr Das et al. (2009)

BS1.1 N16961 DrelA DspoT; Cmr Kmr Smr Das et al. (2009)

BRV1 N16961 DrelA DspoT DrelV; Cmr Kmr Smr Spr Das et al. (2009)

BSRD10 BS1.1 carrying spontaneous chromosomal 10 bp deletion within relV ORF; Cmr Kmr Smr This study

E. coli

CF1648 Wild-type MG1655 Xiao et al. (1991)

CF1652 CF1648 DrelA; Knr Xiao et al. (1991)

CF1693 CF1652 DspoT; Cmr Knr Xiao et al. (1991)

DH5a F9 endA1 hsdR17 supE44 thi-1 recA1 gyrA96 relA1 D(argF-lacZYA) U169 (w80dlacZDM15) Lab stock

XL1Blue recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F9 proAB lacIqZDM15 Tn10 (Tet)]; Tcr Stratagene

Plasmids

pBAD24 pBR322 ori, L-arabinose inducible vector; Apr Guzman et al. (1995)

pRelVBAD 0.81 kb relV ORF of N16961 (coding for full-length 259 aa RelV protein)

cloned in EcoRI-HindIII digested pBAD24; Apr

Pal et al. (2012)

pRelVD10 0.80 kb mutant relV ORF containing internal 10 bp deletion of BSRD10

cloned in EcoRI-HindIII digested pBAD24; Apr

This study

pRelV-K107A pRelVBAD with Lys107Ala mutation; Apr This study

pRelV-D129G pRelVBAD with Asp129Gly mutation; Apr This study

pRelV-R132G pRelVBAD with Arg132Gly mutation; Apr This study

pRelV-L150I pRelVBAD with Leu150Ile mutation; Apr This study

pRelV-G170A pRelVBAD with Gly170Ala mutation; Apr This study

pRelV-Y171A pRelVBAD with Tyr171Ala mutation; Apr This study

pRelV-E188G pRelVBAD with Glu188Gly mutation; Apr This study

pRelV-Q190A pRelVBAD with Gln190Ala mutation; Apr This study

pBRelV1-245 As pRelVBAD ; RelV (aa 1–245 or DC14); Apr This study





pBRelV50-259 As pRelVBAD ; RelV (aa 50–259 or DN49); Apr This study

pBRelV55-259 As pRelVBAD ; RelV (aa 55 –259 or DN54); Apr This study

pBRelV59-259 As pRelVBAD ; RelV (aa 59– 259 or DN58); Apr This study

pBRelV60-259 As pRelVBAD ; RelV (aa 60–259 or DN59); Apr This study

pBRelV59-248 As pRelVBAD ; RelV (aa 59–248 or DC11–DN58); Apr This study

pBRelV59-247 As pRelVBAD ; RelV (aa 59–247 or DC12–DN58); Apr This study

Ap, ampicillin; Cm, chloramphenicol; Km, kanamycin; Sm, streptomycin; Sp, spectinomycin.

V. cholerae relV gene function

http://mic.sgmjournals.org 1857

recently reported cloning of the relV ORF under anarabinose-inducible promoter PBAD of the vector pBAD24and designated pRelVBAD (Pal et al., 2012). In this study,we have further examined the toxic effect of relV geneexpression through pRelVBAD in E. coli background. E. coliDrelA strain CF1652 or DrelADspoT mutant CF1693 (Table 1)were transformed with pRelVBAD and plated on LA con-taining appropriate antibiotics with 0.2, 0.1, 0.01, 0.001 or0.0001 % arabinose or without it. Several transformantswere obtained using the DrelA strain both in the presence orabsence of arabinose. A representative plate containing 0.01 %arabinose along with the control strain CF1652(pBAD24) isshown in Fig. 1(a). However, transformants of DrelADspoTstrain grew in plates containing 0.001 %, 0.0001 % or noarabinose. Results obtained with 0.01, 0.001 % or no arabi-nose are shown in Fig. 1(a). As expected, the control strainCF1693(pBAD24) gave numerous transformants in thepresence or absence of arabinose (Fig. 1a). Thus, it appearsthat the expression of relV using high concentrations ofarabinose, i.e. 0.2, 0.1 or 0.01 % is indeed toxic for E. coliDrelADspoT (p)ppGpp0 cells. This might be due to over-production of (p)ppGpp through expression of relV in trans.Further, the colony size of CF1693(pRelVBAD) was markedlysmall compared with the control strain CF1693(pBAD24) orthe SpoT+ strain CF1652 carrying either the plasmidpBAD24 or pRelVBAD (Fig. 1a).

To provide further evidence that overexpression of relVleads to toxic accumulation of (p)ppGpp, we cloned an

allele of relV ORF with mutation in its RSD encodingregion in the vector pBAD24 and designated pRelVD10. Asexpected, pRelVD10 gave numerous transformants usingthe E. coli DrelADspoT strain CF1693 irrespective of arabi-nose concentration used, and colony size was similar tothat of the control strain CF1693(pBAD24) (data not shown).The result suggests that the mutated relV allele most likelyexpressed a non-functional mutant RelV protein without(p)ppGpp synthetase activity, and thus it has no effect on theviability of the DrelADspoT strain CF1693 irrespective ofpresence of arabinose. This conclusion was further con-firmed by radioactive phosphorus labelling of intracellularnucleotides of CF1693(pRelVBAD) or CF1693(pRelVD10) cellsfollowed by TLC analysis. As expected, only CF1693(pRelVBAD)showed accumulation of (p)ppGpp as shown in Fig. 1(b).

During the course of this study Mechold et al. (2013)reported the successful use of arabinose to express differentconstructs with (p)ppGpp synthetase activity in E. coli.They also reported that ppGpp has more growth inhibitoryeffect in E. coli than pppGpp. Since our TLC analysis (Fig.1b) indicates that the major component of the synthesizedalarmone is ppGpp, we believe that toxic effects ofinduction of RelV expression in E. coli (p)ppGpp0 cellscould be ppGpp mediated.

It is well established that (p)ppGpp0 cells are unable to growin nutritionally poor M9M medium because (p)ppGpppositively regulates several amino acid biosynthesis operonsin bacteria that ultimately help in survival and growth of

(a)

-Ara

-Ara

+0.01%Ara

+0.01%Ara

ΔrelA(pBAD24) ΔrelA(pBAD24) ΔrelA(pRelVBAD) ΔrelA(pRelVBAD)

ΔrelAΔspoT(pBAD24)

ΔrelAΔspoT(pRelVBAD)



ΔrelA

Δspo

T(pBAD24)

ΔrelA

Δspo

T(pRelV

BAD)

ΔrelA

Δspo

T(pRelV

Δ10)

+0.001%Ara +0.01%Ara

CTP

ATP

GTP

ppGpp

Origin

-Ara +0.01%Ara

Wt Ec

(b)

Fig. 1. (a) Phenotypic analysis through controlled expression of the relV gene using the plasmid pRelVBAD in E. coli. Genotypesof strains are as indicated. (b) Autoradiogram showing ppGpp production by expressing the relV gene in E. coli (p)ppGpp0

(DrelADspoT) cells. As controls DrelADspoT(pBAD24) and DrelADspoT(pRelVD10) cells were used. Glucose-starved cellswere labelled with 32P-orthophosphoric acid, extracted, resolved by TLC followed by autoradiography.



micro-organisms under severe nutrient deficient conditions(Durfee et al., 2008; Eymann et al., 2002). We found that ata 0.1 % arabinose concentration, the growth of E. coli(p)ppGpp0 strain CF1693 carrying the plasmid pRelVBAD onM9M agar was like the control E. coli wild-type strainCF1648 (WtEc) as shown in Fig. 2(a). E. coli DrelA strainCF1652, which is (p)ppGpp+ due to the presence of thespoT gene (Xiao et al., 1991) used as a positive control, alsoshowed growth (Fig. 2a). As expected, E. coli (p)ppGpp0

strain CF1693 carrying the empty vector pBAD24 orpRelVD10 used as negative control was unable to grow onM9M agar (Fig. 2a). A similar result was obtained when WtV. cholerae strain N16961 (WtVc), its mutant derivativesRRV1 (DrelADrelV) and the (p)ppGpp0 strain BRV1(DrelADspoTDrelV) (Table 1) were used for M9M growthassay (Fig. 2b). Radioactive phosphorus labelling of theBRV1(pRelVBAD) cells and different control strains followedby TLC analysis further confirmed the synthesis of (p)ppGppthrough pRelVBAD-mediated RelV expression (Fig. 2c).

Thus, the logical conclusion is that production of(p)ppGpp by RelV in the absence of the SpoT hydrolaseis toxic unless cells are grown under conditions that requirehigh levels of (p)ppGpp. This is in agreement with thereports on other SAS genes, e.g. relP, relQ and yjbM (Lemoset al., 2007; Nanamiya et al., 2008).

Essential amino acid residues of the RSD regionof RelV

We have shown that controlled expression of RelV in E. colior V. cholerae (p)ppGpp0 cells allowed their growth on M9Magar plates. This assay was further utilized to identify theessential amino acid residues of the RSD region of RelV.KEGG (http://www.kegg.jp/ssdb-bin/ssdb_motif?kid=vch:VC1224) analysis indicated that the RSD of RelV is 94 aaresidues long (amino acid positions 107 to 201). Similarly,RSDs of different SAS proteins including RelP and RelQ ofStreptococcus mutans, YjbM and YwaC of Bacillus subtiliswere bioinformatically identified and aligned with that ofRelV as shown in Fig. 3(a). For comparison, we alsoincluded RSD regions of the canonical RelA proteins ofV. cholerae (RelAVc), E. coli (RelAEc) and Streptococcusdysgalactiae subspecies equisimilis (RelSeq). We also alignedamino acid sequences of RelV and its homologues present indifferent species of the genus Vibrio as shown in Fig. S1.Such alignments revealed conservation of several aminoacids in the RSD region of RelV (Fig. 3a) and based on thiswe did SDM of some of these amino acids as shownschematically in Fig. 3(b).

Hogg et al. (2004) previously reported the essential role ofaspartic acid at position 264 (D264) in coordinating thebound Mg2+ with the glutamic acid at position 323 (E323)


ΔrelAΔspoTΔrelV(pRelVBAD)


ΔrelAΔspoT(pRelVΔ10)

ΔrelAΔspoTΔrelV(pRelVΔ10)

ΔrelAΔspoTΔrelV(pBAD24)

WtEc

WtVc

(a) (c)

(b)ΔrelAΔrelV

ΔrelAΔrelA

Δspo

TΔrelV(pBAD24)

ΔrelA

Δspo

TΔrelV(pRelV

BAD)

Wt Vc

M9M+Ara

M9M+Ara

CTP

ATP

GTP

ppGpp

Origin

Fig. 2. Growth of E. coli (p)ppGpp0 (DrelADspoT) (a) and V. cholerae (p)ppGpp0 (DrelADspoTDrelV) (b) cells carrying differentplasmid constructs was assayed on M9M agar containing 0.1 % arabinose. E. coli DrelA and V. cholerae DrelADrelV cells wereused as controls. (c) Autoradiogram showing accumulation of (p)ppGpp in V. cholerae (p)ppGpp0 (DrelADspoTDrelV) cellscarrying the plasmid pRelVBAD. DrelADspoTDrelV(pBAD24) was used as a control. Glucose-starved cells were labelled with32P-orthophosphoric acid, extracted and resolved by TLC followed by autoradiography.



within the RSD of RelSeq in the presence of the donor ATPmolecule during the (p)ppGpp synthesis reaction. Wefound that both amino acids are conserved in canonicalRelA/Rel and SAS proteins (Fig. 3a). In fact, alignment ofamino acid sequences of the RSD regions of RelA/Rel/SASproteins indicated that the glutamic acid is present in aconserved motif EXQX, where X could be any amino acid(Fig. 3a). Therefore, aspartic acid at position 129 (D129)and glutamic acid at position 188 (E188) of RelV weresubstituted with glycine (Fig. 3b). We observed that whenthe mutant protein RelVD129G or RelVE188G was expressedthrough the plasmid pRelV-D129G or pRelV-E188G in E.coli or V. cholerae (p)ppGpp0 cells then they failed to growon M9M agar medium (Fig. 4a, b) and to accumulate(p)ppGpp (Figs 4c and S2). The results suggest that boththe mutant proteins RelVD129G and RelVE188G had lost their(p)ppGpp synthetase activity. Thus, we conclude that likecanonical RelA/Rel enzymes, D129 and E188 of the RSDregion of RelV are essential for (p)ppGpp synthetaseactivity. Hogg et al. (2004) previously reported that thepositively charged amino acids, arginine residue at position

241 (R241) and two lysine residues at positions 243 (K243)and 251 (K251), in the RSD region of the RelSeq proteincoordinate with the pyrophosphate group of the donorATP molecule during synthesis of (p)ppGpp. Interestingly,our analysis indicated that the RelV protein carries twoconserved lysine residues at positions 107 (K107) and 115(K115), which correspond to K243 and K251 of the RelSeq

protein (Fig. 3a). Substitution of the K107 with alanine(Fig. 3b) led to complete loss of synthetase activity ofRelVK107A mutant protein since both E. coli CF1693 or V.cholerae BRV1 (p)ppGpp0 cells carrying pRelV-K107Afailed to show any growth on M9M agar or accumulate(p)ppGpp (Figs 4a–c and S2). We believe that substitutionof these corresponding amino acids in other SAS proteinsmost likely leads to loss of (p)ppGpp synthetase activity.Similarly, it was found that the hydrophobic amino acidleucine at position 150 (L150) of RelV and other SASproteins is conserved (Fig. 3a), and in canonical RelA/Relproteins, the corresponding amino acid is either valine (inRelA) or isoleucine (in RelSeq). Interestingly, substitutionof L150 of RelV with the similar amino acid isoleucine

(a)

Relseq : 226

: 42

::::::::

: 47: 57: 45: 107: 250: 251

: 308

: 114: 120: 131: 116: 171: 318: 319

: 385

: 161: 170: 178: 163: 201: 360: 360

RelAEcRelAVcRelVYjbMYwaCRelPRelQ

RelseqRelAECRelAVCRelVYjbMYwaCRelPRelQ

(b)

107129 132 150

224

201

188170171 190

106

Fig. 3. (a) Alignment of amino acid sequences of the RSD regions of RelV and other (p)ppGpp synthetases. Dark and lightshading indicates identical or similar amino acids, respectively. Numbers in the left and right margins represent the positions ofthe corresponding amino acid residues in the full-length protein. The aligned protein sequences include Rel of S. equisimilis

(RelSeq), RelA of E. coli (RelAEc), RelA of V. cholerae (RelAVc), RelV of V. cholerae, YjbM and YwaC of B. subtilis, RelP andRelQ of S. mutans. The other conserved motifs present in different (p)ppGpp synthetases are EXQX and GY, as indicated. (b)Schematic diagram showing the amino acid sequence of the RSD along with its flanking region (amino acid residues 106 to224) of the RelV protein. Upward and downward arrows indicate the bioinformatically identified RSD (amino acid residues 107to 201). Natural amino acids are in bold face and underlined, and substituted residues are shown in large font size above eachof the natural residues. Numbers above each substituted amino acid residue indicate its position.



(L150I) followed by expression of the mutant proteinRelVL150I through the plasmid pRelV-L150I in E. coli or V.cholerae (p)ppGpp0 neither supported growth of bacterial

cells on M9M agar nor accumulated intracellular (p)ppGpp(Fig. 4a–c), indicating functional loss of the mutant protein.We have also substituted arginine at position 132 (R132) of


ΔrelAΔspoT(pRelV-E188G)

ΔrelAΔspoT(pRelV-L150I)

ΔrelAΔs

poT(

pRel

V-E

188G

)

ΔrelAΔs

poT(

pRel

V-L

150I

)

ΔrelAΔspoT(pRelV-R132G)

ΔrelAΔspoT(pRelV-D129G)

ΔrelAΔspoT(pRelV-K107A)

ΔrelAΔs

poT(

pRel

V-R

132G

)

ΔrelAΔs

poT(

pRel

V-D

129G

)

ΔrelAΔs

poT(

pRel

V-K

107A

)ΔrelAΔspoT(pRelVBAD)

WtEc

ΔrelAΔs

poT(

pBA

D24

)

ΔrelAΔs

poT(

pRel

VB

AD)

ΔrelAΔs

poT(

pRel

V-D

129G

)

ΔrelAΔs

poT(

pRel

VB

AD)

Wt E

c

M9+AraM9+Ara


ΔrelAΔspoTΔrelV(pRelV-E188G)

ΔrelAΔspoTΔrelV(pRelV-L150I)

ΔrelAΔspoTΔrelV(pRelV-R132G)

ΔrelAΔspoTΔrelV(pRelV-D129G)

ΔrelAΔs

poTΔ

relV

(pR

elV

-D12

9G)

ΔrelAΔspoTΔrelV(pRelV-K107A)


ΔrelAΔs

poTΔ

relV

(pR

elV

BA

D)

WtVc

(b)(a)

(d)

M

1.35

CTP

ATP

GTP

ppGpp

Origin

1.070.870.60

0.31

0.23

0.110.19

0.28/0.27

R N R N R N R N

(c)

Fig. 4. Analysis of site-directed mutants of relV in E. coli (p)ppGpp0 (DrelADspoT) (a) and V. cholerae (p)ppGpp0

(DrelADspoTDrelV) (b) cells. Growth of bacterial cells was assayed on M9M agar containing 0.1 % arabinose. Bacterial strainscarrying different plasmids are as indicated. WtEc and WtVc cells were used as positive controls. (c) Autoradiogram showingaccumulation of (p)ppGpp in E. coli (p)ppGpp0 (DrelADspoT) cells expressing different relV mutant alleles through plasmids asindicated. WtEc and E. coli DrelADspoT(pRelVBAD) cells were used as positive controls and E. coli DrelADspoT(pBAD24) wasused as a negative control. (d) RT-PCR analysis showing transcripts of relV or its site-directed mutant allele in E. coli (p)ppGpp0

(DrelADspoT) and V. cholerae (p)ppGpp0 (DrelADspoTDrelV) cells. Bacterial cells carried the plasmid pRelVBAD or pRelV-D129Gas indicated. ‘M’ denotes QX174 DNA digested with HaeIII used as DNA markers and their sizes (in kb) are indicated in the leftmargin. R indicates RT with Taq DNA polymerase. N denotes use of only Taq DNA polymerase as a negative control.



RelV to glycine (R132G) (Fig. 3b) since the correspondingposition in other SAS proteins has a glycine residue (Fig. 3a).Expression of the mutant protein RelVR132G through thepRelV-R132G in E. coli or in V. cholerae (p)ppGpp0 cells

failed to grow on M9M agar or accumulate intracellular(p)ppGpp, which strongly suggests loss of (p)ppGpp syn-thetase activity of RelVR132G protein (Figs 4a–c). Ourbioinformatics analysis of the RSD regions of RelA/Rel/

ΔrelA

Δspo

T(pBRelV1-249)

ΔrelA

Δspo

T(pBRelV1-248)

ΔrelA

Δspo

T(pBRelV1-247)

ΔrelA

Δspo

T(pBRelV1-245)

ΔrelA

Δspo

T(pBRelV1-250)

ΔrelA

Δspo

T(pBAD24)

ΔrelA

Δspo

T(pRelV

BAD)

Wt Ec

ΔrelAΔspoT(pBRelV1-249)







WtEc

M9M+Ara

ΔrelAΔspoTΔrelV(pBRelV1-249)







WtVc

M9M+Ara

(b)(a)

RelV1-259

RelV50-259

RelV55-259

RelV59-259

RelV60-259

RelV59-248

RelV59-247

RelV1-250

50

1

107 201

259

250

249

245

247

247

248

248

RSD

55

59

60

59

59

RelV1-249

RelV1-248

RelV1-247

RelV1-245

(c)

(d)

CTP

ATP

GTP

ppGpp

Origin

Fig. 5. Functional analysis of the C-terminal deletion mutants of relV. (a) Schematic representation (not drawn to scale) showingprogressive deletions of truncated C-terminal, N-terminal or both ends of RelV. Grey rectangular box represents thebioinformatically identified RSD region of RelV; amino acid positions covering this domain are shown below the box. Black anddotted lines represent intact and deleted amino acid sequences, respectively. Amino acid sequence lengths of intact RelV or itstruncated versions are indicated in the margins. Growth of E. coli (p)ppGpp0 (DrelADspoT) (b) and V. cholerae (p)ppGpp0

(DrelADspoTDrelV) (c) cells carrying different relV C-terminal deletion constructs including the plasmids pBAD24 andpRelVBAD was assayed on M9M agar containing 0.1 % arabinose. WtEc and WtVc cells were also used as positive controls. (d)Autoradiograms showing (p)ppGpp accumulation in E. coli (p)ppGpp0 (DrelADspoT) strain carrying different relV C-terminaldeletion constructs as indicated. WtEc and E. coli DrelADspoT(pRelVBAD) strains were used as positive controls and E. coli

DrelADspoT(pBAD24) served as a negative control.



SAS proteins also indicated a glycine-tyrosine (GY) motif asshown in Fig. 3(a). Interestingly, there was no loss of(p)ppGpp synthetase activity when the G170 or Y171residue of RelV was substituted with alanine (Fig. S2)although Hogg et al. (2004) previously showed that thesubstitution of the corresponding Y308 residue of the RelSeq

protein may lead to defective (p)ppGpp synthetase activity.Similarly, there was no effect when the conserved glutamineresidue at position 190 (Q190) of the EXQX motif of RelVwas substituted with alanine (Fig. S2).

We confirmed that the observed growth phenotype of E.coli or V. cholerae cells carrying each of the above-mentioned constructs using M9M agar assay was due tothe expression of a mutant allele of relV by determiningtheir transcripts using RT-PCR assay. As expected, theassay indicated a similar amount of expression of relV allelein each case. We used the cells of CF1693(pRelV-D129G)or BRV1(pRelV-D129G) to measure the transcript level ofmutated relV alleles along with different controls (Fig. 4d).

Determination of functional C-terminal boundaryof RelV

To understand the role, if any, of the C-terminal region ofRelV in maintaining the (p)ppGpp synthetase function,we developed several progressive deletion constructs(Fig. 5a) by cloning each truncated relV ORF under thePBAD promoter of the plasmid pBAD24, and each clonewas tested in E. coli or V. cholerae (p)ppGpp0 cells byM9M growth assay as well as accumulation of intracellular(p)ppGpp by TLC analysis. We observed that the E. coli orV. cholerae (p)ppGpp0 strain CF1693 or BRV1, respect-ively, carrying the plasmid pBRelV1-250, pBRelV1-249or pBRelV1-248 could grow on M9M agar (Fig. 5b, c)and was also able to accumulate (p)ppGpp (Fig. 5d)suggesting that each of these expressed truncated pro-teins has retained its (p)ppGpp synthetase activity. Insharp contrast, the CF1693 or BRV1 strain carryingthe recombinant plasmid pBRelV1-247 or pBRelV1-245failed to grow on M9M agar or accumulate intracellular

ΔrelA

Δspo

T(pBRelV55-259)

ΔrelA

Δspo

T(pBRelV59-259)

ΔrelA

Δspo

T(pBRelV60-259)

ΔrelA

Δspo

T(pBRelV50-259)

ΔrelA

Δspo

T(pBAD24)

ΔrelA

Δspo

T(pRelV

BAD)

Wt Ec







WtEc

M9+Ara







WtVc

M9+Ara

CTP

ATP

GTP

ppGpp

Origin

(c)(a)

(b)

Fig. 6. Functional analysis of the N-terminal deletion mutants of relV. Growth of E. coli (p)ppGpp0 (DrelADspoT) (a) and V.

cholerae (p)ppGpp0 (DrelADspoTDrelV) (b) cells carrying different relV deletion constructs including the plasmids pBAD24 andpRelVBAD was assayed on M9M agar containing 0.1 % arabinose. WtEc and WtVc cells were also used as positive controls.(c) TLC analysis showing (p)ppGpp accumulation in E. coli (p)ppGpp0 (DrelADspoT) cells carrying different N-terminalrelV deletion constructs. WtEc and E. coli DrelADspoT(pRelVBAD) strains served as positive controls and E. coli

DrelADspoT(pBAD24) was used as a negative control.



(p)ppGpp (Fig. 5b–d). Thus it appears that deletion of 12but not 11 amino acid residues from the C-terminal endof the RelV protein completely impaired its (p)ppGpp

synthetase function. Hence, the amino acid at position248 of RelV appears to be its functional C-terminalboundary.

ΔrelA

Δspo

T(pBRelV59-248)

ΔrelA

Δspo

T(pBRelV59-247)

ΔrelA

Δspo

T(pBAD24)

ΔrelA

Δspo

T(pRelV

BAD)

ΔrelA

Δspo

TΔrelV(pBRelV59-248)

ΔrelA

Δspo

TΔrelV(pBRelV59-247)

ΔrelA

Δspo

TΔrelV(pRelV

BAD)

Wt Ec





WtEc





WtVc

M9M+Ara M9M+Ara

(c)(a)

(b) (d)

CTP

1.35

M R N R N R N

1.07

0.87

0.60

0.31

0.190.23

0.28/0.27

ATP

GTP

ppGpp

Origin

Fig. 7. Determination of the minimal functional region of RelV. (a) Growth of E. coli (p)ppGpp0 (DrelADspoT) cells carryingdifferent relV deletion constructs including the plasmids pBAD24 and pRelVBAD was assayed on M9M agar containing 0.1 %arabinose. WtEc cells were used as a positive control. (b) TLC analysis showing (p)ppGpp accumulation in E. coli (p)ppGpp0

(DrelADspoT) cells carrying different relV deletion constructs including the plasmids pBAD24 and pRelVBAD. WtEc cells wereused as a positive control. (c) Growth of V. cholerae (p)ppGpp0 (DrelADspoTDrelV) cells carrying different relV deletionconstructs including the plasmids pBAD24 and pRelVBAD was assayed on M9M agar containing 0.1 % arabinose. WtVc cellswere used as a positive control. (d) RT-PCR analysis showing transcripts of the relV gene or its deletion mutant alleles in V.

cholerae (p)ppGpp0 (DrelADspoTDrelV) cells. Bacterial cells carried the plasmids pRelVBAD, pBRelV59-248 and pBRelV59-247 as indicated. M denotes QX174 DNA digested with HaeIII used as DNA markers and their sizes (in kb) are indicated in theleft margin. R indicates RT with Taq DNA polymerase. N denotes use of only Taq DNA polymerase as a negative control.



N-terminal deletion analysis of RelV

After determining the C-terminal functional boundary ofthe RelV, we wanted to know the role, if any, of the N-terminal region in regulating the synthetase activity. Likedetermination of the functional C-terminal boundary, asimilar strategy was followed to identify the functionalN-terminal boundary by developing several progressivedeletion constructs using the arabinose-inducible expres-sion vector pBAD24 (Fig. 5a). We found that E. coli or V.cholerae (p)ppGpp0 cells carrying the recombinant plasmidpBRelV60-259 only, failed to grow on M9M agar and werealso unable to accumulate intracellular (p)ppGpp (Fig. 6a–c), confirming that the amino acid residue at position 59 ofthe N-terminal region is the functional boundary of RelVkeeping the C-terminal region intact. The results suggestthat like the C-terminal region, deletion of a certain stretchof amino acids (here amino acid residues 1–58) of the N-terminal region is tolerable for functional activity of RelV.Thus, it appears that both N- and C-terminal amino acidsequences are involved in regulating the (p)ppGppsynthetase activity of RelV, and most likely these sequencesare needed to maintain the proper structure of the RelVenzyme, but this needs further study.

Determination of minimal functional length ofRelV

Having determined that amino acid residues 59 and 248 actas the functional N- and C-terminal boundaries, respect-ively, of the RelV, it was of interest to determine theminimal functional length of this SAS protein. Ideally, afragment of amino acids consisting of residues 59 to 248 ofRelV should be functional, and to confirm this we cloned aregion of the relV ORF that encoded amino acids 59 to 248in pBAD24 as shown schematically in Fig. 5(a). Expressionof this N- and C-termini-truncated RelV protein in anarabinose-dependent manner through the recombinantplasmid pBRelV59-248 in E. coli (p)ppGpp0 CF1693 cellsindicated that this protein was indeed functional, becausethe strain CF1693(pBRelV59-248) showed growth on M9Mmedium and also accumulated intracellular (p)ppGppas analysed using a TLC method (Fig. 7a, b). However,this strain contributed to a decreased amount of (p)ppGppcompared with the CF1693(pRelVBAD) control strain,indicating attenuation in the (p)ppGpp synthetic activityof the truncated RelV protein consisting of amino acidresidues 59 to 248. In sharp contrast, when we expressedthe truncated RelV protein fragment consisting of aminoacid residues 59 to 247 through the plasmid pBRelV59-247in CF1693 cells, it failed to grow in M9M agar indicatingthat deletion of a single amino acid residue led to completeloss of (p)ppGpp synthetase activity of the mutant protein(Fig. 7a, b). Thus, the minimal functional fragment of RelVis only 189 aa long and it carries the predicted RSD regionof 94 residues along with the flanking N- and C-terminal48 and 47 aa residues, respectively. Similar results wereobtained when each of the construct pBRelV59-248 or

pBRelV59-247 were expressed in the V. cholerae (p)ppGpp0

cells of the strain BRV1 as shown in Fig. 7(c). We confirmedplasmid-driven expression of Wt or truncated ORF ofthe relV gene by RT-PCR assays using total cellular RNAisolated from BRV1(pRelVBAD), BRV1(pBRelV59-248) andBRV1(pBRelV59-247) strain as shown in Fig. 7(d).

CONCLUSION

The present study attempted to dissect the function ofthe relV gene of V. cholerae by genetic and mutationalapproaches under in vivo conditions using E. coli as wellas V. cholerae (p)ppGpp0 cells. The relV gene is highlyconserved in other Vibrio species, suggesting its probableimportance in physiology/niche adaptation in this group oforganisms of marine origin. SDM analysis carried out inthis study allowed us to identify five amino acid residues(K107, D129, R132, L150 and E188) of the RSD region ofRelV that are essential for maintaining (p)ppGpp synthe-tase function. Additionally, progressive deletion analysishas helped us in underpinning the functional N- and C-terminal boundaries of RelV. To our knowledge this is thefirst report to establish the minimal functional region ofan SAS protein, and the results further suggest that thisbioinformatically identified RSD region indeed providesvaluable information but its practical value extends beyondthe predicted region. We hope that the results presentedhere will help in the future to understand further thefunctions of other SAS enzymes and reveal the role of thesesmall proteins including RelV in regulation of the SR indiverse bacteria.

ACKNOWLEDGEMENTS

We thank Professor Siddhartha Roy for his constant support and

encouragement. We are grateful to Dr M. Cashel, National Institute of

Health, Bethesda, MD, for the generous gift of E. coli strains CF1648,

CF1652 and CF1693. We thank Pratap Koyal and Shibprasad Sharma

for their excellent technical assistance in this work. The work was

supported by a research grant from the Council of Scientific and

Industrial Research (CSIR), Government of India. S. D., P. B. and

R. R. P. are grateful for research fellowships from CSIR and S. B. is

grateful for a research fellowship from ICMR, New Delhi.

REFERENCES

Aravind, L. & Koonin, E. V. (1998). The HD domain defines a new

superfamily of metal-dependent phosphohydrolases. Trends Biochem

Sci 23, 469–472.

Atkinson, G. C., Tenson, T. & Hauryliuk, V. (2011). The RelA/SpoT

homolog (RSH) superfamily: distribution and functional evolution of

ppGpp synthetases and hydrolases across the tree of life. PLoS ONE 6,

e23479.

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman,J. G., Smith, J. A. & Struhl, K. (1989). Current Protocol in Molecular

Biology. NY: John Wiley and Sons.

Avarbock, A., Avarbock, D., Teh, J. S., Buckstein, M., Wang, Z. M. &Rubin, H. (2005). Functional regulation of the opposing (p)ppGpp



synthetase/hydrolase activities of RelMtb from Mycobacterium tuber-culosis. Biochemistry 44, 9913–9923.

Avril, T., Freeman, S. D., Attrill, H., Clarke, R. G. & Crocker, P. R.(2005). Siglec-5 (CD170) can mediate inhibitory signaling in theabsence of immunoreceptor tyrosine-based inhibitory motif phos-phorylation. J Biol Chem 280, 19843–19851.

Battesti, A. & Bouveret, E. (2006). Acyl carrier protein/SpoTinteraction, the switch linking SpoT-dependent stress response tofatty acid metabolism. Mol Microbiol 62, 1048–1063.

Cashel, M., Gentry, D. R., Hernandes, V. J. & Vinella, D. (1996). Thestringent response. In Escherichia coli and Salmonella typhimurium:cellular and molecular biology, pp. 1458–1496. Edited by F. C.Neidhardt. Washington, DC: American Society for Microbiology.

Chatterji, D. & Ojha, A. K. (2001). Revisiting the stringent response,ppGpp and starvation signaling. Curr Opin Microbiol 4, 160–165.

Choy, H. E. (2000). The study of guanosine 59-diphosphate 39-diphosphate-mediated transcription regulation in vitro using acoupled transcription-translation system. J Biol Chem 275, 6783–6789.

Das, B. & Bhadra, R. K. (2008). Molecular characterization of Vibriocholerae DrelA DspoT double mutants. Arch Microbiol 189, 227–238.

Das, B., Pal, R. R., Bag, S. & Bhadra, R. K. (2009). Stringent responsein Vibrio cholerae: genetic analysis of spoT gene function andidentification of a novel (p)ppGpp synthetase gene. Mol Microbiol72, 380–398.

Durfee, T., Hansen, A. M., Zhi, H., Blattner, F. R. & Jin, D. J. (2008).Transcription profiling of the stringent response in Escherichia coli.J Bacteriol 190, 1084–1096.

Eymann, C., Homuth, G., Scharf, C. & Hecker, M. (2002). Bacillussubtilis functional genomics: global characterization of the stringentresponse by proteome and transcriptome analysis. J Bacteriol 184,2500–2520.

Gentry, D. R. & Cashel, M. (1996). Mutational analysis of theEscherichia coli spoT gene identifies distinct but overlapping regionsinvolved in ppGpp synthesis and degradation. Mol Microbiol 19,1373–1384.

Gropp, M., Strausz, Y., Gross, M. & Glaser, G. (2001). Regulation ofEscherichia coli RelA requires oligomerization of the C-terminaldomain. J Bacteriol 183, 570–579.

Guzman, L. M., Belin, D., Carson, M. J. & Beckwith, J. (1995). Tightregulation, modulation, and high-level expression by vectors contain-ing the arabinose PBAD promoter. J Bacteriol 177, 4121–4130.

Haralalka, S., Nandi, S. & Bhadra, R. K. (2003). Mutation in the relAgene of Vibrio cholerae affects in vitro and in vivo expression ofvirulence factors. J Bacteriol 185, 4672–4682.

Haseltine, W. A. & Block, R. (1973). Synthesis of guanosine tetra- andpentaphosphate requires the presence of a codon-specific, unchargedtransfer ribonucleic acid in the acceptor site of ribosomes. Proc NatlAcad Sci U S A 70, 1564–1568.

Heidelberg, J. F., Eisen, J. A., Nelson, W. C., Clayton, R. A., Gwinn,M. L., Dodson, R. J., Haft, D. H., Hickey, E. K., Peterson, J. D. & otherauthors (2000). DNA sequence of both chromosomes of the cholerapathogen Vibrio cholerae. Nature 406, 477–483.

Hogg, T., Mechold, U., Malke, H., Cashel, M. & Hilgenfeld, R. (2004).Conformational antagonism between opposing active sites in abifunctional RelA/SpoT homolog modulates (p)ppGpp metabolismduring the stringent response [corrected]. Cell 117, 57–68.

Jishage, M., Kvint, K., Shingler, V. & Nystrom, T. (2002). Regulationof sigma factor competition by the alarmone ppGpp. Genes Dev 16,1260–1270.

Lemos, J. A., Lin, V. K., Nascimento, M. M., Abranches, J. & Burne,R. A. (2007). Three gene products govern (p)ppGpp production byStreptococcus mutans. Mol Microbiol 65, 1568–1581.

Mechold, U., Murphy, H., Brown, L. & Cashel, M. (2002).Intramolecular regulation of the opposing (p)ppGpp catalyticactivities of Rel(Seq), the Rel/Spo enzyme from Streptococcusequisimilis. J Bacteriol 184, 2878–2888.

Mechold, U., Potrykus, K., Murphy, H., Murakami, K. S. & Cashel, M.(2013). Differential regulation by ppGpp versus pppGpp inEscherichia coli. Nucleic Acids Res 41, 6175–6189.

Mittenhuber, G. (2001). Comparative genomics and evolution ofgenes encoding bacterial (p)ppGpp synthetases/hydrolases (the Rel,RelA and SpoT proteins). J Mol Microbiol Biotechnol 3, 585–600.

Murdeshwar, M. S. & Chatterji, D. (2012). MS_RHII-RSD, a dual-function RNase HII-(p)ppGpp synthetase from Mycobacteriumsmegmatis. J Bacteriol 194, 4003–4014.

Murray, K. D. & Bremer, H. (1996). Control of spoT-dependent ppGppsynthesis and degradation in Escherichia coli. J Mol Biol 259, 41–57.

Nanamiya, H., Kasai, K., Nozawa, A., Yun, C. S., Narisawa, T.,Murakami, K., Natori, Y., Kawamura, F. & Tozawa, Y. (2008).Identification and functional analysis of novel (p)ppGpp synthetasegenes in Bacillus subtilis. Mol Microbiol 67, 291–304.

Pal, R. R., Bag, S., Dasgupta, S., Das, B. & Bhadra, R. K. (2012).Functional characterization of the stringent response regulatory genedksA of Vibrio cholerae and its role in modulation of virulencephenotypes. J Bacteriol 194, 5638–5648.

Potrykus, K. & Cashel, M. (2008). (p)ppGpp: still magical? Annu RevMicrobiol 62, 35–51.

Schreiber, G., Metzger, S., Aizenman, E., Roza, S., Cashel, M. &Glaser, G. (1991). Overexpression of the relA gene in Escherichia coli.J Biol Chem 266, 3760–3767.

Seyfzadeh, M., Keener, J. & Nomura, M. (1993). spoT-dependentaccumulation of guanosine tetraphosphate in response to fatty acidstarvation in Escherichia coli. Proc Natl Acad Sci U S A 90, 11004–11008.

Srivatsan, A. & Wang, J. D. (2008). Control of bacterial transcription,translation and replication by (p)ppGpp. Curr Opin Microbiol 11,100–105.

Stephens, J. C., Artz, S. W. & Ames, B. N. (1975). Guanosine 59-diphosphate 39-diphosphate (ppGpp): positive effector for histidineoperon transcription and general signal for amino-acid deficiency.Proc Natl Acad Sci U S A 72, 4389–4393.

Traxler, M. F., Zacharia, V. M., Marquardt, S., Summers, S. M.,Nguyen, H. T., Stark, S. E. & Conway, T. (2011). Discretely calibratedregulatory loops controlled by ppGpp partition gene induction acrossthe ‘feast to famine’ gradient in Escherichia coli. Mol Microbiol 79,830–845.

Wendrich, T. M., Blaha, G., Wilson, D. N., Marahiel, M. A. & Nierhaus,K. H. (2002). Dissection of the mechanism for the stringent factorRelA. Mol Cell 10, 779–788.

Wu, J., Hassan, K. A., Skurray, R. A. & Brown, M. H. (2008).Functional analyses reveal an important role for tyrosine residues inthe staphylococcal multidrug efflux protein QacA. BMC Microbiol 8,147.

Xiao, H., Kalman, M., Ikehara, K., Zemel, S., Glaser, G. & Cashel, M.(1991). Residual guanosine 39,59-bispyrophosphate synthetic activityof relA null mutants can be eliminated by spoT null mutations. J BiolChem 266, 5980–5990.

Edited by: E. Hartland



reprint microbiology-uk aug 2014

Documents