fast purification of a functional elongator trnamet expressed from a

Volume 16 Number 16 1988 Nucleic Acids Research

Fast purification of a functional etongator tRNA"1 expressed from a synthetic gene in vivo

Thierry Meinnel, Yves Mechulam and Guy Fayat

Laboratoire de Biochimie, Ecole Polytechnique, UA DO. 240 du CNRS, 91128 Palaiscau Cedex,France

Received June 30, 1988; Accepted July 18, 1988

To study the interaction of an aminoacyl-tRNA synthetase with its cognate tRNAs, it is of interest to beable to design synthetic tRNA mutant genes, and to rapidly purify sufficient amounts of the corresponding tRNAs.A prerequisite is to show that the synthetase retains fall specificity towards a wild type tRNA expressed from asynthetic gene. For this purpose, we have used an expression vector to overproduce and purify elongator tRNA1116*.To date, the most widely used rapid purification procedure, extraction from polyacrylamide gels, does not provideenough pure material to achieve fine characterizations, such as affinity constants measurements by the analysis ofthe intrinsic enzyme fluorescence quenching (1). We propose a rapid method for the purification of a functionaltRNA1™ based on the construction of a synthetic gene, the obtention of an overproducing E. coli strain and atwo-step FPLC driven purification from a crude extract

A shuttle vector, pBSTNAV, (derived from Btuescript M13+KS) carrying a multisite cloning sequenceEcoRl/Smal/Pstl flanked upstream with a synthetic lipoprotein promoter and downstream with the rrnCtranscription terminator, both prepared from the pGFIB-I phasmid (2), was designed (figure 1). The syntheticelongator tRNAmet gene was constructed by assembling six different overlapping oligonucleotides (figure 1)synthetised with the Pharmacia gene assembler and purified on a Mono Q column. A mixture of the 6oligonucleotides (l.S uM each) was heated to 95 °C, slowly cooled and submitted to ligation during 1 hour at 37°C.Oligonucleotides 1 and 4 were kept unphosphorylated, to avoid the concatemerization of assembled genes duringligation. Agarose gel electrophoresis showed that the major ligation product corresponded to the tRNA gene.Cloning between the EcoV.1 and Pst\ sites of the shuttle vector was realised by performing ligation at 14°C, underthe conditions defined by Dardel (3). To eliminate reconstructed vector molecules, the ligation mixture was thenrestricted by Sml. After transformation of JMlOlTr cells (4), the screening of overproducing clones was achievedby determining the methionine accepting specific activity in crude tRNA minipreparations (5). Among the 20clones randomly tested, 8 displayed a 10- to 20-fold higher activity, corresponding to a 30 to 50-fold overproductionof elongator tRNA, since initiator tRNA accounts for 70% of the total methionine accepting capacity in E. coli.Sequencing of the corresponding single-stranded DNAs showed that in each case, the construction was exactly asexrjected.

lpp rrnCpromoter t a n d m t o i

(Hccl)/(»«rl) EaofI ! • • ! FvnII F.tl BUuJIIIpBSIN&V

CCCCATCCATCGAOTCA«CAATCTC(;TGTACTGACTATTACTACCCCAGTOTCCAAGCTTACCCCACCATCGGTCCTC

T T

rigura 1

Crude stripped tRNA was prepared from a 1 liter culture of the overproducing cells in LB medium(50ng/ml ampicillin), according to Zubay (6). 500 Awn were usually obtained, with a specific activity of 370 to600 pmoie tRNAmet/A260 ""'• 0- e. 22 to 35 % purity). It was then loaded on a DEAE-Trisacryl M coiumn (1.1x 5 cm) equilibrated with i 0.2 M NaCl, 20 mM Tris-HCl (pH = IS), 8 mM MgCl2. tRNA was eluted byperforming a 0.2 to 0.4 M NaCl linear gradient (2 ml/min; 0.24 M/h). The methionine accepting fraction contained120 to 240 A26<) units with a specific activity of 670 to 1050 pmote/A2fio units (70* yield). After ethanol

© IRL Press Limited, Oxford, England. 8 0 9 5

Downloaded from https://academic.oup.com/nar/article-abstract/16/16/8095/2378352by gueston 13 February 2018

Nucleic Acids Research

precipitation, 60 A250 units were dried, redissolved in 0.S ml 10 mM ammonium acetate (pH =• 6.6), 1.8 Mammonium sulfate, 10 mM MgCl2 ""d loaded on a TSK-phenyl column (7.5 x 0.75 cm). tRNAs were eluted byperforming 1 1.58 M to 1.22 M reverse ammonium sulfate gradient (0.54 M/h; 0.7 ml/min). A typicalchroenstogram is shown below.

- 8 , 5

0 - m10 20

Elution TOIUMI* (Ml)

The pooled fractions were desalted by filtration through a GF05M-Trisacryl column (2.5 x 25 cm; 4 ml/min)equilibrated in 10 mM Tris-HCl (pH=7_5).

The total procedure allows purification of 1 to 2 mg tRNA"101 (overall yield=56%) to a more than 1400pmote/A26o u n ' t specific activity in less than 48 hours (bacteria as starting material). The overproduced tRNA wasfully aminoacylatable, as demonstrated by the methionine acceptance profile of the TSK-phenyl eluate. Indeed, thefraction corresponding to the maximal absorbance reached 1800 pmoley^go The Michaelis constant i s well as thedissociation constant (1) of the final product for mooomeric methionyl-tRNA synthctase have been determined(Km=0.8 jiM; KQ=0A (iM) and compared to those of elongator tRNA"161 purified from a non-overproducing strain(Km=0.25 nM: KD=0.1 |jM)). In spite of the slight differences, possibly reflecting undermodification of theoverproduced specie*, these results confirmed that methionyl-tRNA synthetase has retained full ipecificity towardsthe tubstrate of synthetic origin.

ACKNOWLEDGEMENTS

Dr J-M MASSON is acknowledged for the gracious gift of pGFIB-I.

REFERENCESOLANQUET S., IWATSUBO M. and WALLER JP. (1973) Eur. J. Bwchem. 36, 213-2262.NORMANLY J., MASSON J-M, KLEINA L.G., ABELSON J. and MILLER J.H. (1986) Proc. NatL Acad. Sci.(USA) 83. 6548-65523.DARDEL F. (1988) Nucleic Acids Res. 16, 1767-17784.MREL P. H., LEVEQUE F , MELLOT P., DARDEL F., PANVERT M., MECHULAM Y. and FAYAT G.(1988) Biochimie in press

5JAYAT G., FROMANT M., KALOGERAXOS T. and BLANQUET S. (1983) Biochimie 65,221-2256.ZUBAY G. (1962) J. MoL Biol. 4, 347-356

8 0 9 6


Volume 16 Number 16 1988 Nucleic Acids Research

Extranudear gene expression in yeast: evidence for a plasmid-encoded RNA polymerase ofunique structure

Duncan W.Wilson"1" and Peter A.Meacock*

Leicester Biocentre, University of Leicester, Leicester, LEI 7RH, UK

Received June 24, 1988; Revised and Accepted July 26, 1988 Accession DO. X07946

ABSTRACTStrains of the yeast Kluyveroayces lactis that produce killer-toxin have beenfound to contain two linear dsDNA plasaids, kj (8.9 Kb) and k2 (13.4 8b). Thefour transcribed open reading fraaes of plasaid kj contain no recognisableyeast nuclear expression signals. Moreover, a toxin subunit gene fused withthe lacZ gene of Bscherichia coli is not detectably expressed when introducedto K.lactia or Saccharoaycea cerevlalae on a nuclear vector, even when nativek} and k2 are present in the cell. This and other evidence is consistent withthe hypothesis that kj and k2 reside in an extranuclear location, and do notutilise the nuclear RNA polyaerases I, II or III for transcription of theirgenes. Sequencing of plasaid k2, which is thought to encode factors necessaryfor the Maintenance or expression of kj, reveals an open reading fraaepredicted to encode a 974 aaino acid polypeptide with hoaology to severalDHA-directed RNA polyaerases. He suggest that this is a component of a novelplasaid-specific extranuclear gene expression system.

INTRODUCTION

Many strains of the budding yeast Kluyyeroaycea lactis contain two linear

double-stranded A/T rich DNA plasaids, kx (8.9 Kb) and k2 (13.4 Kb)1 that are

associated with secretion of a aulti-subunit protein toxin2 which inhibits

the growth of certain sensitive yeast species. Structurally, the plasaids

belong to a class of extrachroaoaoaal linear eleaents which includes

adenovirus3, bacteriophage phi 29* plasaids of Streptoaycea rochel^ and the

Sl/82 aitochondrial plasaids associated with Bale sterility in Baize6- All of

these aolecules have inverted terminal repeat sequences (2OZbp and 184bp in

the case of k} and k27) and proteins covalently attached to the 5' end of

each DNA strand (for kj and k 2 terminal proteins see reference 8). These Bay

function as priaers in the initiation of terminal DNA replication. In support

of this hypothesis one of the four transcribed open reading fraaes of k}9>10

encodes a product with hoaology to the DNA polyaerases of adenovirus and phi

29, and to a protein encoded by an ORF of the aaixe plasaid SI11.

Plasaids kj and k2 can be transferred to p° (rho zero) strains of

© IR L Press Limited, Oxford, England. 8°97Downloaded from https://academic.oup.com/nar/article-abstract/16/16/8095/2378352by gueston 13 February 2018


Bwyharoayces cerevisiae (which lack aitochondrial DNA) where they are

maintained and express the killer phenotype12'13. However, they do not becoae

eatabliahed in the presence of aitochondrial DNA14. These observations,

together with the high A/T content of the plasaid DNA, fluorescence staining

of S.cerevisiae P° derivatives containing the plasaids 13 and fractionation of

yeast nuclei and cytoplasa by centrifugation techniques (reference 15, D.H.

Wilson and P.A. Meacock unpublished observations) are all consistent with a

cytoplassie location for the plasaids. Plasaid curing experiments have shown

that plasaid k2 can be maintained in the absence of k^, but plasaid k^ is

unable to exist independentIyl6 suggesting dependence upon k2-encoded

products. Such factors aay be concerned with the replication or segregation

of kj, or for the expression of k^-encoded genes essential for kj aaintenance.

A variety of data suggest that these plasaids say utilise a novel systea

for gene transcription; viz. none of the ORFs of )q is preceded by

recognisable yeast nuclear promoter eleaents, although all four are preceded

by a actif identical with, or closely related to, the sequence

ACT(A/T)AATATATQA. This has been teraed the Upstreaa Conserved Sequence or

XXXP, and transcription is initiated approxiaately 14 bp downstreaa of this

eleaent (Roaanos and Boyd, aanuscript subaitted). Toxin production cannot be

detected froa p° strain? of S.cerevisiae which have been transforaed with the

coding region of plasaid kj cloned into yeast 2 fM (aicron)-based vectors

(D.H. Wilson and P.A. Meacock, unpublished observations, reference 15).

Furthermore, Northern blotting reveals that when kj DNA is introduced into

K.lactis on nuclear vectors transcription of the kj gene ORF2 (which encodes

two toxin subunita) is initiated at a nuaber of sites distinct froa those

used by native linear k\, and the transcript is preaaturely terainated

(Rcmanos and Boyd, aanuscript subaitted). Thus it appears that the yeast

nuclear UNA polyaerases I, II and III are unable to recognise and correctly

transcribe the genes of these plasaids.

Here we demonstrate that expression of an 0HF2/lacZ fusion gene, cloned

into a yeast nuclear vector, cannot be detected in cells of K.lactis or

S.cerevisiae which contain plasaids kj and k2- This suggests that

transcription of 0RF2, and probably all plasaid-borne genes, occurs in an

extranuclear cellular compartment. The provision of novel cytoplasaic

expression factors could be one of the aaintenance functions of kj, and to

investigate this we have begun to determine the nucleotide sequence of this

plasaid. The predicted product of one of the OBFs encoded by k2 has hoaology

to two different subunits found within DNA-directed RNA polyaerases, whilst

8098



another has homology to a vaccinia virus hellease necessary for specific

viral transcription.

MATERIALS AND METHODS

Strains and Media

Kluyveromyces lactis IFO1267 (prototrophic [lq+k2+)) was obtained from the

National Collection of Yeast Cultures, Food Research Institute, Colney Lane,

Norwich, UK. K.lactis SDH (K. lactis Iac4 trpl [kj^*]) was kindly provided

by Prof. C.Hollenberg, and K.lactis ABK802 (prototrophic, [k1°,k2+]) by Dr.

A.Boyd. Saccharosiyces cerevlslae SPK103 (3 hl»3-A tn>l-289 ura3~52 leu2-3

[k\+ k2+ L+] p°) was prepared by cytoduction of a p° derivative of strain

S15O-2B (obtained froa J.Hicks, Cold Spring Harbor Laboratory, New York) with

JC25K (a ade2-l hla4-15 karl-1 [kj+.kg*]) obtained fro* Dr. M.A.Romanos.

Yeast were propagated at 30'C using either YPD Bedim (1 X yeast extract, 2 *

peptone, 2 X glucose) or in minimal aediua (0.67 * yeast nitrogen base, 2 X

glucose) supplemented with nutrients as required. Plasmid constructions,

bacterial assays for />-galactosidase activity, and saperinfection for

preparation of single stranded DNA made use of B.coli NM622 Adac-proAB) thl~

BUPE hsdA5 [II proAB lacIQ £ZM15117 (provided by Dr. A.Mileham), maintained

on M9 minimal medium^ supplemented with 0.001 X thiamine, in order to ensure

maintenance of the F' episome and susceptibility to N13 amd M13K07 phage

infection. N13K07 helper phage was from Pharmacia, Uppsala, Sweden.

Bngymea

The Klenow fragment of DNA polymerase I, exonuclease III, exonuclease VII,

and all restriction endonucleases used in this study were purchased from

Bethesda Research Laboratories (BRL), Bethesda, Maryland, USA. T4 DNA ligase

was purchased from Pharmacia, and Proteinase K from Boehringer (BCL),

Mannhelii, FRQ. Zymolyase 100-T was from the Kirin brewery, Tokyo, Japan. All

ensymes were used as recommended by the supplier.

Materials for Sequence Analysis

Deoxynucleotides and dideoxynucleotides were purchased from BCL. Acrylamide

and bisscrylamide were "Eloctran" grade, purchased from BDH chemicals, Poole,

UK. Ultrapure enzyme grade urea was from BRL. Slgmacote and

N,N,N',N'-Tatramethylethylenediamine (TEMED) were obtained from Sigma, St

Louie, Missouri, USA. Sequencing primers were 17 bp "universal primer"

supplied by Pharmacia and custom-synthesised 17 bp oligodeoxynucleotides

prepared by J.Keyte, Biochemistry department, Leicester University, using an

Applied Biosystems DNA synthesiser. [of-^JdATP, at 10 iCi /il~l and 660 Ci

8099



•Mol-1 wore obtained from Anershom International, UK.

Preparation of Plasmid k; DMA

K.loctia ABK802 was grown in 1 litre of YPD broth to a density of 3xlO7 cells

•1~1, cells were pelleted and washed once in water, then resuspended in 50 ml

of 3RD buffer (1.2 M Sorbito 1, 20 sM BDTA, 50 mM Dithiothreitol) and

incubated at 37'C for 20 sins. To the suspension was added 1 ml of 10 mg ml"1

ZymoLyase 100-T in 3RD buffer and incubation continued until spheroplasting

was complete. Sphaeroplasts were pelleted and resuspended in 20 nl 10 nM

BDTA, 50 »M Tris.BCl pfi 8.0, then Sodium N-lauryl Sarcosinate added to a

final concentration of 2 X. To the lysate was added 1 ml of 10 mg al"1

Proteinase K, th« mixture incubated at 37'C for 1 h, then the lysate chilled

on ice, mixed with 5 ml of 5 H NaCl and left on ice for 1 h. After

centrifugation at 18000 rpm for 30 mins at 4'C in a Sorvall SS34 rotor the

supernatant was extracted three tines with an equal volume of phenol, then

twice with chloroform. Extracted supernatant was mixed with 2 ml 3 M sodium

acetate pH 5.5 and 50 ml absolute ethanol, placed at -80'C for 30 mins,

nucleic acid pelleted and the pellet rinsed with 10 ml of 70 * ethanol. The

pellet was briefly dried under vacuum, resuspended in 5 ml TB buffer (1 mM

KDTA, 10 mM Tris.BCl pH 7.6) then nucleic acid fractionated according to site

in a 10 K-40 * sucrose gradient, containing 1 H NaCl, 5 mM KDTA, 20 mM

Tris.BCl pB 8.0, by centrifugation at 26000 rpm for 24 h at 26'C in a Sorvall

ABB27 rotor. Fractions containing plasmid k2 were identified by agarose gel

electrophoresIs, pooled and diluted with 2 volumes of water. DNA of plasmid

k2 was precipitated for 1 h at -80'C after the addition of l/10th volume of

3 H sodium acetate pH 5.5 and 2.5 volumes of absolute ethanol, then pelleted,

the pellet rinsed with 0.6 ml of 70 * ethanol, briefly vacuum dried and

reauspended in 200 Ml TB buffer.

Yeast transformants were grown to between 1x10? and 5x10? cells ml"1 in 25 ml

minimal medium, imposing selection for presence of plasmid-boroe markers.

Optical density of the cultures at 600 nm was determined, and cultures placed

on ice for 20 miss. An aliquot was removed and plated onto complete media in

order to determine, by subsequent replica plating onto complete and selective

media, the percentage of cells within the population which contained

plasmids. Chilled cells were pelleted and resuspended in 1 ml 60 mM KHjP04.

To 0.4 ml of the suspension was added an equal volume of

60 mM NasHP04, 40 mM MaHaP04, 10 mM KC1, 1 mM MgSO4, 60 mM P mercaptoethanol

pB 7.0, then 10 id 1 * SDS and Phenylmethylsulphonyl fluoride (FWSF) to a

8100



final concentration of 0.5 aM. The Mixture Mas vortexed for 5 sees, incubated

at 28'C for 10 Bins then 30 ul chlorofora added, the aixture vortexed again

and incubation continued at 28'C for 10 Bins. A 200 Ul aliquot of 4 ag al"1

ONPQ (freshly dissolved in 100 aM KH2P04) was waned to 28'C then nixed with

the treated cells, a tiner started and incubation continued at 28'C. At the

first appearance of a yellow colour within the Mixture the reaction was

stopped, by addition of 0.5 ml I N NajCO,, and tiae of incubation recorded.

Cells were removed frcsi the mixture and optical density of the supernatant

recorded at a wavelength of 420 na. Units of 0-galactoaidase activity were

expressed as 0D42O ain~l OD^o"1, then Multiplied by a factor of 1000. When

the percentage of plasBid-containing yeast cells was known, assay values were

corrected to that expected if 100 * of cells had contained plasaid at the

tiae of assay. Bacterial transforaants were assayed in exactly the sane

•anoer, but using 10 al cell cultures. Since bacterial transforaants were

grown under antibiotic selection, all cells contained plasaids.

Generation of Nested Overlapping Deletions Froa Cloned ky DMA

DMA (5 Ug) of a plasaid containing the 7.7 Kb gas M.S3Q I fragaent of k2

cloned within p&(BL19+ was linearised by digestion with the restriction

endonuclease* Baa 31 and Kpn I- The aixture was extracted with equal voluaes

of phenol and chlorofora, and ethanol precipitated. The fragaent was

resuspended in TE buffer to a concentration of 1 Ug DNA (Jl"1-

For deletion of the 3'recessed strand, 5 All of linearised plasaid fragaent

were aixed with 27 ul water, 4 Ul lOxBxoIII buffer (10 aH NgCl2, 660 aM

Tris.HCl pH 8.0) and brought to reaction temperature by incubating at 37'C

for 5 ains. Then 4 Ul of 10 units ul~* exonuclease III were added. A 2 Ul

saaple was iaaediately reaoved and aixed with 2 ul IOXBXOVII buffer (300 aH

KC1, 100 aM BDTA, 100 BM Tris.HCl pH 7.5) on ice. Samples were siailarly

taken at 30 second intervals for 5 Bins, at which tiae 20 ul of prewaraed

IXBXOIII buffer containing 40 units Bxonuclease III were added to the

incubation. Ten subsequent samples of 4 ul each were reaoved at 30 a

intervals into 2 Ul of ice-cold IOxBxoVII buffer. The twenty samples were

each diluted to 19 ul with water and aixed with 1 ul exonuclease VII diluted

to 1.5 units ul~l in lxBxoVII buffer. The reactions were incubated at 37'C

for 2 h to remove single-stranded regions of DNA. Samples were phenol

extracted, chlorofora extracted and ethanol-predpitated, resulting pellets

were drained, vacuum dried and resuspended in 15 ul of 10 aH MgCl,, 10 aM

Dithiothreitol, laM ATP, 100 Ug al"1 Bovine Serua Albumin, 50 aM

Trifl.BCl pH 7.4, 50 Urn dATP, 60 Urn dTTP, 50 urn dCTP, 60 Ua dOTP, 3 units T4

8101



DNA Ligase, 0.5 unit Klenow fragment. Ligation was allowed to proceed

overnight at 30'C. Twenty aliquots of competent B.coll IM522 were transformed

with 6 ill of each ligation dilated by addition of 100 *d of 0.1 N CaCl,.

Transforaeuts were picked and plasaid DNA prepared froa 1.6 ml cultures by

•tandard methods. Degree of deletion within each plasaid was determined by

agaroae gel electrophoresis of restriction digestion products.

Buperlofoctioo of POCBL—containing K.coli

A 111 aliquot of M9 asdia, suppleaanted with 0.001* thiaaine, 50 Hg al"1

aapicillin, 0.2 % glucose, was inoculated with 10 Ml of stationary phase

culture of K.coli NH522 transforaod with a deletion-derivative of a pEHBL/k2

clone. The culture was grown at 37'C overnight, and a 20 Ml aliquot used to

inoculate 2 al of 2xYT broth, 0.001 * thiaaine, 150 MS al"1 aapicillin

prewaraed to 37'C. This culture was shakes at 37'C until reaching an optical

density of between 0.6 and 1.0 at a wavelength of 660 na. A 1 al aliquot of

this culture was infected with N13K07 helper-pbage at a Multiplicity of 10

M13K07 plaque foraing units/bacterial cell. The cell/phage mixture was shaken

at 37'C for 1 h, then 400 Ml added to 10 al prewaraed 2xYT broth, containing

0.001 * thiaaine, 150 Mg ml"1 aapicillin, 70 Ug "I" 1 kanaaycin. The culture

was grown overnight at 37'C, cells pelleted and cell-free supernatant* stored

until required for template preparation.

PWA Sequencing

Sequencing was by the dideoxy aethod19 but using [er-35s]dATP. All reactions

were carried out at 50'C. Dideoxy-terminated fragaents were electrophoresed

upon a 6 k polyacrylaaide/urea sequencing gel, in lxTBB buffer (5.5 g 1~1

boric acid, 0.93 g I"1 KDTA, 10.8 g I"1 Tris base). Blectrophoresis was for

2.5 h or up to 7.5 h and gels dried under vacuua (without prior fixing) at

80'C for 90 Bins then exposed to Kodak XAR-5 fila for at least 12 h.

RESULTS AND DISCP38IOH

Killer Plasaid Penes Are Rot Expressed Within The Yeast Nucleus

Previous studies (reference 15. Wilson and Meacock, unpublished

observations) have shown that kj toxin genes, cloned in circular autonomously

replicating vectors, do not confer a killer phenotype upon their host.

However, none of these experiaents could be perforaed with yeast strains that

also contained both kj and k2 as native linear plassdds. Thus plasaid-encoded

factors necessary for lq toxin gene expression aay have been absent. In order

to test, quantitatively, whether linear plasaid encoded genes contained

within nuclear vectors can be expressed when the endogenous plasaids k\ and

8102



•iiiniiiiiiMiM.i

Aap Thr Atp Pro V»l V«l L«u

OAC ACT Q AT CCC QTC QTT TTA

*b,,

<V/ A amp

Fig. 1. Structure of tbe plasaid YHpOBFZ2. Regions of tbe plaaaid which arederived froa plaaaida k\ and YRp7 are indicated. The fusion junction betweenthe saino-terainal coding region of 0BF2 and lacZ is shown. Solid box: IncZcoding region. Hatched box: Reaainder of 0HF2 coding region. Open box: k\genes and TBP1 gene of YHp7. Curved arrows: Bacterial genes. Arrow headsindicate direction of transcription of genes. A (delta): Partial gene.Bacterial origin of replication, ori. is also shown.

kg are also present in the cell, we assayed expression of a cloned kj gene

using a gene fusion.

A translational fusion was Bade between tbe aaino-terainal 36 codons of

ORF2 and the lacZ gene of Escherlchia coll. As the predicted lacZ-fusion

product retains the ORF2 signal sequence it aay becoae aeabrane associated.

However analogous gene fusions between lacZ and the QAL2 peraease of

S.cerevlsiae have been successfully assayed in yeast^O. The 0BF2-lacZ gene

fusion retained 1.8 Kb of k} DHA noraally upstrean of 0RF2 and 1.5 Kb of k}

DNA noraally downstreaa of the 0BF2 carboxy terainus. The fusion gene was

transferred to the S.cereyislae vector YHp721, to fora plasaid TRpOBFZ2. See

Figure 1. During construction of plasaid YRpOBFZ2, the AHS1 sequence,

necessary for autonoaous replication of YHp7 in S.cerevisiae. was reaoved.

However the A/T rich kj-derived DNA fortuitously provides eleaents22 enabling

8103



Table 1. Expression of an 0RF2-lecZ fusion gene in yeast and B.coli

Plaamid

pUC19

pOCOBFZ2

PL0669Z

pPA13

YHpOBFZ2

YHp7

KHp2

—

B.coliM622

2.67

3.14

11.4

0.04

0.43

0.01

—

—

S.cerev;LsiaeSPK103

—

—

776.88

39.20

0.71

0.49

—

—

I-lfffftt*8011 IFO1287

— —

— —

— —

— —

0.68 —

— —

0.62 —

— 8.36

Units of 0-galactosidase activity expressed by various plasmids in strains ofyeast and B.coli.. For calculation of activity values, and genotypes of•trains, see materials and sethods. SD11 carries an inactivating lesionwithin the endogenous K.1actla 0-galactosidase, IF01267 does not, but wasgrown under conditions which would not induce expression of the enzyme beyondbasal levels. Plasaid pOCOR!Z2 carries the 0RF2-lacZ fusion expressed fromthe lac promoter of pUC19. YRpORFZ2 carries the 0RF2-lacZ fusion cloned intothe yeast high copy nuaber vector YRp7. pLQ669Z expresses the lacZ gene tramthe S.cerevisiae CYC1 nuclear promoter. pPA13 expresses lacZ from thecauliflower mosaic virus 19S promoter. TRp7 t> KHp2 are, respectively,S.cerevisiae and K.lactie autonomously replicating plasmids which carry nolacZ gene.

autonomous replication of the plasaid in both S.cerevlsiao and K.lactis.

Planid YRpOBFZ2 was transformed into strains of E.lactie and

S.cerevisiae which harboured linear planids k^ and k2> using a whole-cell

transformation procedure^. Transformed cells were permeabilised with

chlorofora, and 0-galactosidaae activity measured by spectrophotometrlc assay

of the rate of hydrolysis of o-nitrophenyl-# (beta)-D-galactoside (ONPQ). No

ensyae activity could be detected in either yeast (Table 1). Detection of

activity in B.coli strain W622, when transformed with YHpOHFZ2 or with

P000RFZ2 (which expresses the gene fusion from the lac promoter) confirmed

that, when expressed, the fusion protein did retain normal 0-galactosidase

activity. Also, other controls indicated that, under these conditions, it was

8104



DCS

OBF 1 ACTATAATA-TATCA AAAATQ

ORF 2 ACT-TAATA-TATGA AAAAIO

OBF 3 ACT-AAATA-TATOA AAAATO

ORF 4 lAAAATAATCTCA AAAATO.

ORF 9 7 4 TA-TATOA AAAATC

OHF 5 7 9 TA-TCTOA AAAATO

ORF 3 3 6 AAAATATOA ATAATO

ORF 1 5 8 ATATTTTOA AATATO

ORF 1 3 2 TA-TOTOA AAAATO

ORF 112 TA-TTTOA ATTATO

ORF 103 AAATA-TATGA AAAATQ

C o o » « i » u * TAATATGAA T T

- C

a

Fig. 2. Comparison of the upstream conserved sequence (UCS) found before eachof the OHFs of plasaid Iq with upstreaa regions of plasaid Ic2 OBFs identifiedduring this study. The predicted initiation codon (ATG) and three precedingnucleotides are shown in each case. Bach intervening micleotide isrepresented by a dot (.). Note the requirement for an adenine residue at the-3 position, and preference for an AAAATQ initiation context. A consensus UCSis shown.

possible to detect 0-galactosidase activity in yeast when expressed at high

(plasaid pIA669Z) or low (plasaid pPA13 and noo-induced IFO1267) levels.

He suggest that expression of at least OHF2, and probably all of the k}

genes, therefore requires factors unavailable within the nuclear coapartaent,

even when native linear kj and kg are also present in the cell.

Plasaid k; Encodes a Product With Boaology to Two HHA Polyaeraae Subunits

In order to investigate the novel extranuclear expression systea which

aay be encoded by kg we are determining the coaplete nucleotide sequence of

this plasaid. ONA of plasaid kg was prepared froa K.lactis ABK802 [k1°,k2+]

and digested to coapletion with the restriction endonucleases Baa HI and

Xho I. This yielded two Baa HI-Kfea I fragments, of 7.7 Kb and 5.5 Kb, which

after aodification of terainal restriction sites were cloned in both possible

orientations into the sequencing vector pEMBIl?1". An overlapping series of

deletion-derivatives were prepared froa the 7.7 Kb fragaent, in both

directions, using the exooucleose Ill/exonuclease VII deletion aethod^.

Deletion-derived clones were sequenced, and sequence data asseabled aanually

using the University of Wisconsin genetics coaputer group (UHQOQ)

Analysis of 7688 bp of continuous sequence data obtained froa plasaid kg

reveals seven OBFs, with potential to encode products of 974, 579, 33B, 158,

8105



ZQD

ySKt*uC^

M»fivtanmyMiU*IhittaftxiMnMD9iilyrUumyl4Will6«nU«tB^^

J J

TmmrnnnirniM M I U M I I I * mum w BIMX m

U.QyOji^^

Min*/Miri nu ft

as

Fig. 3. Sequence of the OBF 974 coding region aligned with that of itspredicted product, showing upatreaa DNA. The OBF waa translated using theuniversal genetic code.

132, 112 and 103 smino acids. All seven ORFs exhibit the saae codon usage as

the ORFs of plasaid kj, all are preceded by UCS-like elements and the

initiation codon of all OBFs is in a similar context (figure 2). He suggest

that this indicates a i: o—un expression Bechanisa for the genes of plasaids

k} and l<2- Alignment of lq and k2 UCS eleaents leads to definition of a new

•inijul consensus DCS; (A/T)A(A/T/-)TNTQA, where - indicates the possible

absence of a base, and N Bay be any base.

8106



aa

Ala

Arg

Asn

Asp

Cys

Gin

Glu

Gly

His

He

ftcodon

usage

GCTGCCOCAGCG

COTCOCCOACOGAOAAOGAATAACGATGAC

TOTTOC

CAACAG

GAAGAG

GOTGOCGGAGOG

CATCAC

ATTATCATA

Table 2.

kj ORFs

538372

4160882

8515

8713

8911

946

8614

523413

9010

40357

Codon Usage

ORF 974

480484

00008812

8812

8911

8813

7624

928

330644

8119

14185

of ORF

aa

Leu

Lys

Met

Phe

Pro

Ser

Thr

Trp

Tyr

Val

974 and k

ftcodon

usage

TTATTGCTTCTCCTACTG

AAAAAG

ATGTTTTTCCCTCCCCCACCG

TCTTCCTCATCGAGTAOC

ACTACCACAACG

TOG

TATTAG

GTTGTCOTAGTG

I ORFs

kx ORFs

64913194

8515

100

8317

657271

444210300

484462

100

8812

486388

OHF 974

579120175

919

100

7327

388468

389162295

457480

100

7723

392572

Percentage codon usage. Average codon usage of the four open reading frasesof plaanid k^ are shown, compared with codon usage of OHF 974.aa: Aaino acid.

The complete sequence of this region of k2 has been deposited within the

EMBL database under accession nuaber X07946. Analysis of the OBFs contained

within it will be reported elsewhere (D.M.W., PhD thesis, University of

8107



I-ali. •« rf1 »

r f i l s p » i T I T T I M I H r n i | T | p i | u L » D D I I H B » f l ] i T [ 5 1 I T I i r c T l o r a T H or i g P i n o I | P r I I L I I I L I I I O I D I I I I i t i 1 0 1 r i u i T O O F I Ir I o p i t i i i i ! H | L ] I | L | ] i g i i i n i » l » l i i l o l p r i I T T O O F I IL I I ) » I i i T | O » l c L T I L I I I L I E I I O T D P I | T O r | j i ] » o | o j » i o I I T T I I I Q P | T •

13001019

• i|L|i|a|r|o|r m i O|I|H L T I I > I B|I|I A IOIOa e i|Lji|o|i.!alA m i i|t|i n u n I[J|I i 662

r n ilqlnlB t p

o D rT O P l l l t C t l

M I I Q C O L

11 i 11 inU N i l 374I I H 1 I 403

Q A H 1 B03

530soo4T960S

Fig. 4. Horologies between product of ORF 974 and several RNA polyseraaesubunits. Origin of subunits abbreviated aa follows: Nt chl, Nicotianatabacw chloroplast. Mp chl, Marchantin polyorpha cnloroplaat. Sc,Saccfaarowyccg cereviaae. Da, Drosophila aelanogaater. Ninbers refer tolocation of final residue of nosology region within correspondingpolypoptide. Boxes and asterisks, respectively, indicate where a residue ofthe ORF 974 product is identical to, or conservatively different froa,corresponding aaino acids within two or •ore aligned RNA polyaerase subunits.By this criterion, region A of ORF 974 product has 36.3 * identity withregion A of the other polypeptides. This rises to 53.8 % if conservativesubstitutions are also taken into account. Region II shows 43.2 % identityrising to 56.8 *. Region III shows 48.3 * identity rising to 58.3 X.

Leicester). Using the algorithm of Lipaan and Pearson26 to search the

National Bioaedical Research Foundation (NBRF) protein sequence database, we

have found significant nosologies between the predicted ORF 974 product and

the sub units of several DNA-directed RNA polyaerasea. The sequence of the

predicted ORF 974 product is shown in figure 3. The codon usage of ORF 974 is

shown in table 2.

Cosrputer-assisted comparison of the predicted protein sequences of

several RNA polyaerase subunits with the ORF 974 product reveals an 80 aaino

acid region of hoaology (which we have teraed region A) in coaaon with the P

subunits of RNA polyaerases froa f.coli27, the chloroplasts of tobacco

Nicotiana tabacuu28 and the chloroplasts of the liverwort Marchantja

polyorpha^. This region ia also homologous to one of nine conserved regions

shared between the /J-subunit of B.coli RNA polyaerase and the 140 M> subunit

of S. cerevisioe HNA polyaerase 11^0. Additionally there are two other

regions, of 43 aaino acids and 59 aaiino acids, with hoaology to regions of

the B.coli RNA polyaerase $' subunit3! the M.polyorpha chloroplast RNA

8108



* ii in

• DD

, r I II III If f fl

«c polll/pollll

Ic 0

Fig. 5. Structure of predicted ORF 974 product arid several RNA polyaerasesubunits. Abbreviations as in Figure 4 legend. Hoaology regions A and I to VIare boxed. Dotted line; carboxy terminal repeat region of Sc Pol II32.

polyaerase />' •ubunit29, the 215 kD subunits of S.cerevisiae32 and Drosophila

aelanogaster33 HNA polyaeraae II and the 160 Kd subunit of S.cerevisiae HNA

polyaerase III32. These latter two regions of hoaology correspond to the RNA

polys*rane large subunit conserved regions II and III defined by Allison and

coworkers3^. See Figure 4. However regions A, II and III are arranged

differently within the ORF 974 product, as is shown in Figure 5.

He have been unable to detect any nosology between the ORF 974 product

and regions I.IV.V and VI cu—IUU to S.cerevisiae HNA Pol II, Pol III and

B.coli subunit P'. We have also failed to detect hoaology between the aaino

terminal 570 residues of the OBF 974 product and other HNA polyajerase

sutranits. The ORF 974 product bears no primary sequence nosology with the

•itochondrial RNA polyaerase of S.carevisiae34. This enzyme resembles the HNA

polyaerases of bacteriophage T3 and T7, and the predicted product of an ORF

within the •itochondrial linear plaanid 82 of Zea aays35. That the ORF 974

product is unrelated to this class of polymerases is consistent with an

extraaitocbondrial location for the killer plasaids, as would be expected

froa their maintenance and expression within P° strains ot S.cerevisiae. and

their capacity to encode proteins which enter the secretory pathway.

The apparent cytoplasaic location of k} and kj and their lack of

recognisable expression signals implies that an RNA polyaeraae of novel

properties say be required to transcribe their genes. We suggest that the ORF

974 polypeptide has a role in this task, perhaps via recognition of the UCS

elements found 5' to all kj and k2 genes. There may of course be additional

plasmid-encoded or cellular proteins required for the process of

transcription. In this regard we have detected hoaology between the product

of a second k2 ORF, ORF 579 (which is able to encode a polypeptide of 579

amino acid residues), and proteins D-669 and C-637 of the poxvirus vaccinia.

8109



Vaccinia la a cytoplasaic DNA virus which carries its own gene expression

system, and product D-569 has been identified aa the DNA-dependent ATPase36

which aay unwind the vaccinia duplex during transcription37. Perhaps the

product of ORF 579 perform a similar function for the killer plasoids. Like

ORF 974, OBF 579 is preceded by a UCS-like sequence and its predicted

initiation codon lies in an AAAATQ context (figure 2). We note that a

consequence of our interpretation of this data is that ORF 974 and ORF 679

products would be required for their own expression.

The roles performed by the various subunits of prokaryotic or eukaryotic

HNA polyserases are poorly understood, although there is soae evidence to

suggest that the P' subunit of B.coli HNA polyaerase nay have a role in DNA

binding^ whilst the P subunit may bind nucleotides^. Because the product of

OSF 974 shares nosology with soae of the regions found in both types of

subunit, its structure is unique. Comparison of its biological activity with

known UNA polyaerases of conventional structure will lead to an increased

understanding of the roles of each subunit and the conserved regions within

the*.

ACIHOWLBDQBMKNTS

We thank Michael Rcaanos and Alan Boyd for coasounicating OBF2 transcript

data prior to publication, and Michael Stark and Michael Pocklington for

advice concerning interpretation of hoaology data. Plasaid pPA13 was a gift

frost Jennifer Richardson. This work was supported by the core research

progi niii of the Leicester Biocentre, and the award of an SERC studentship to

D.W.W.

"•"Present address: Department of Molecular Biology, Lewis Thomas Laboratory, Princeton, NJ 08544,USA

•To whom correspondence should be addressed

BBMiKBM/'KS1. Ounge, N., Tanaru, A., Ozawa, F., Sakaguchi, K. (1981). J. Bact. 14C.

382-3902. Stark, M.J.R., Boyd, A. (1986). H4BO J. 5, 1995-20023. Challberg, M.D., Kelly, T.J. (1979). J. Mol. Biol. 131, 999-10124. Garcia, J., Penalva, M., Blanco, L., Salas, M. (1984). Proc. Natl.

Acad. Sci. USA. §1, 80-845. Hirochika, H., Sakaguchi, K. (1982). Plasmid 7, 59-656. Kesfcle, R.J., Thompson, R.D. (1982). Nucl. Acids. Res. 10, 8181-81907. Sor, F., Wesolowski, M., Fukuhara, H. (1983). Nucl. Acids. Res. H ,

6037-5044

8110



8. Kikucfai, Y., Hirad, K., Hishinuaa, F. (1384). Noel. Acids. Res. 12,5685-6692

9. Stark, M.J.R., Milehaa, A.J., Roaanos, N.A., Boyd, A. (1984). Kucl.Acids. Res. 12, 6011-6030

10. Sor, F., Fukuhara, H. (1985). Curr. Genet. 9, 147-15511. Jung, 0., Loavitt, N.C., Ito, J. (1987). Nucl. Acids. Res. 15, 908812. Gunge, N., Sakaguchi, K. (1981). J. Bact. 141, 155-16013. Gunge, N., Murata, K., Sakaguchi, K. (1982). J. Bact. 151, 462-46414. Gunge, N., Yaaane, C. (1984). J. Bact. 159, 533-53915. Staa, J.C., Kwakaan, J., Meijer, M., Stuitje, A.R. (1988). Nucl.

Acids. Res. 14. 6871-688416. Niwa, 0., Sakaguchi, K., Gunge, N. (1981). J. Bact. 148, 988-99017. Gough, J.A., Murray, N.B. (1983). J. Mol. Biol. 166, 1-1918. Naniatis, T., Fritsch, B.F., Saabrook, J. (1982). Molecular Cloning.

A Laboratory Manual. Cold Spring Harbor Laboratory, Cold SpringHarbor, New York

19. Sanger, F., Nicklen, S., Coulson, A.R. (1977). Proc. Natl. Acad. Sci.USA. 74, 5463-5476

20. Tscfaopp, J.F., Enr, S.D., Field, C., Schekaan, R. (1986). J. Bact.166. 313-318

21. Struhl, K., Stinchcoab, D.T., Scherer, S., Davis, R.W. (1979). Proc.Natl. Acad. Sci. USA. 76, 1035-1039

22. Thoapson, A., Oliver S.G. (1986). Yeast, g, 179-19123. Ito, H., Fukuda, Y., Murata, K., Kiaura, A. (1983). J. Bact. 153.

163-16824. Henikoff, S. (1984). Gene 25, 351-35925. Devereux J., Haeberli, P., Snithies, 0. (1984). Nucl. Acids. Res. 12,

387-39526. Lipnan, D.J., Pearson, H.R. (1985). Science g Z . 1435-144127. Ovcfainnikov, Y.A., Monastyrskaya, G.S., Gubanov, V.V., Guryev, S.O.,

Chertov, O.Y., Modyanov, N.N., Grinkevich, V.A., Makarova, I.A.,Marcfaenko, T.V., Polovnikova, I.N., Lipkin, V.M., Sverdlov, B.D.,(1981). Bur. J. Biochea. H 6 , 621-629

28. Shinosaki, K., Ohae, M., Tanaka, M., Nakasugi, T., Hayashida, N.,Matsubayashi, T., Zaita, N., dnumongse, J., Obokata, J.,YaBagucfai-Shinozaki, K., Ohto, C., Torazawa, K., Meng, B.Y., Sugita,M., Deno, H., Kaaogashira, T., Yaoada, K., Kusuda, J., Takaiwa, F.,Kato, A., Tohdoh, N., Shinada, H., Sugiura, M. (1986). EMBO J. 5,2043-2049

29. Ohyaaa, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano, T., Sano, S.,Oaesona, K., Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi,H., Oseki, H. (1986). Nature 322, 572-574

30. Sweetaer, D., Nonet, M., Young, R.A. (1987). Proc. Natl. Acad. Sci.USA. 24. 1192-1196

31. Ovchinnikov, Y.A., Monastyrskaya, G.S., Oubanov, V.V., Guryev, S.0.,Saloaatina, I.S., Shuvaeva, T.M., Lipkin, V.M., Sverdlov, B.D.(1982). Nucl. Acids. Res. 10, 4035-4044

32. Allison, L.A., Moyle, M., Shales, M., Ingles, C.J. (1986). Cell 42.,599-610

33. Biggs, J., Searlea, L.L., Greenleaf, A.L. (1985) Cell 42, 611-62134. Masters, B.S., Stohl L.L., Clayton D.A. (1987). Cell 51, 89-9935. Kuzain, B.V., Levchenko I.V., Zaitseva G.N. (1988). Nucl. Acids. Res.

!g. 417736. Broyles, S.S., Moss, B. (1987). J. Virol, gl, 1738-174237. Broyles, S.S., Moss, B. (1987). Molec. Cell. Biol. 2. 7-1438. Fukuda, H., Ishihaaa, A. (1974). J. Mol. Biol. 87, 523-540

8111



39. Armstrong, V.H., Sternbacfa, H., Bckstein, F. (1976). Biochemistry 15,2086-2091

MOTK ADDED IN PROOFSince submission of this manuscript Touasino et al. published the

complete nucleotide sequence of plasmid k2 (Nucl.Acids.Res. 16, 5863-5878).Their predictions concerning the genetic organisation of k2 ORFs are in goodagreement with our own, except that they have found ORFs 112 and 336 to bejoined into a single ORF of 453 amino acids.

8112


fast purification of a functional elongator trnamet expressed from a

Documents