Gene, 43 (1986) 281-286

Elsevier

281

GENE 1587

Short Communications

Plasmid vectors useful in the study of translation initiation signals

(Recombinant DNA; start codon; cloning vectors; /_?-galactosidase hybrid proteins; phages M13, I, P22)

Elizabeth Wyckoff *, Laura Sampson, Melody Hayden *, Ryan Parr, Wai Mun Huang and Sherwood Casjens * *

Department of Cellular Viral and Molecular Biology. University of Utah Medical Center, Salt Lake City, UT 84132 (U.S.A.) Tel. (801)581-5980

(Received July 12th, 1985)

(Revision received January 31st, 1986)

(Accepted February 4th, 1986)

SUMMARY

The construction and characterization of plasmid vectors useful in the analysis of translation initiation signals in Escherichiu coli and in the construction of 1acZ gene hybrids are described. Transcription on the vectors proceeds from a CAMP-independent lac promoter through several restriction sites into a truncated IucZ structural gene lacking its first eight codons. Because this gene has no translation initiation signal, its level of expression is extremely low. A DNA fragment containing a translation start signal can be inserted between the promoter and truncated lacZ gene to produce a hybrid protein with functional /.&galactosidase activity. The vectors described here differ in sequence between the EcoRI cloning site and the ZacZ gene to allow easy, in-frame joining of DNA containing a translation initiation signal to the 1ucZ gene. Cells containing plasmids can be screened directly for in-frame inserts by colony color on indicator plates.

INTRODUCTION

* Present addresses: (E.W.) Department of Biochemistry, Duke

University Medical Center, Durham, NC 27710 (U.S.A.)

Tel. (919)684-6501; (M.H.) National Jewish Hospital, Depart-

ment of Cytogenetics, 906 Goodman Building, 3800 East Colfax,

Denver, CO 80206 (U.S.A.) Tel. (303)388-4461.

** To whom correspondence and reprint requests should be

addressed.

Abbreviations: aa, amino acid(s); /#Gal, pgalactosidase; bp,

base pair(s); CAMP, cyclic AMP; nt, nucleotide(s); ORF, open

reading frame; PA, polyacrylamide; SDS, sodium dodecyl

sulfate; XGal, 5-bromo-4-chloro-3-indolyl-P-D-galactopyrano-

side.

It is often useful in the study of complex genetic systems to utilize genetic fusions to well-defined genes. The IacZ gene of E. coli has been particularly suitable for such fusions (Weinstock et al., 1983; Casadaban et al,, 1983). We report here the con- struction and properties of a set of plasmid vectors in which transcription initiates from the lacUV5 (CAMP-independent) promoter and is directed to- ward several restriction sites preceding a truncated 1acZ gene which begins within its eighth codon. DNA framents containing the translation-initiation

0378-l 119/86/$03.50 0 1986 Elsevier Science Publishers B.V. (Biomedical Division)

282

region of genes from E. coli and coliphage operons can be used to make and express hybrid genes. The construction of other vectors for the study of trans- lation start sites has been described (Casadaban et al., 1980; 1983; Minton, 1984; Shapira et al., 1983; Thomas et al., 1982; Looman et al., 1985). However, only those described here contain a pro- moter directing transcription across the hybrid gene, lack any possible 1ucZ translation start and allow insertion of foreign translation starts in all three reading frames. This allows direct screening for clones which contain translation start signals joined so that a pGal hybrid protein is synthesized.

EXPERIMENTAL AND DISCUSSION

(a) Construction and characterization of plasmid pTSV-1

The pTSV plasmids were designed to be useful in the analysis of prokaryotic translation initiation se- quences. Their structures are summarized in Fig. 1. The plasmid pMC1403 contains a modified lac

operon in which several restriction sites precede a

IacZgene which lacks its translation start and its first eight codons (Casadaban et al., 1980). This plasmid was cleaved with SalI, and DNA complementary to the protruding 5’ ends was synthesized by phage T4 DNA polymerase. The products of this reaction were cut with EcoRI to generate molecules with one blunt and one EcoRI end. The plasmid pOP95-15 contains a lacUV5 promoter, but not the lacZ initia- tor ATG, adjacent to an EcoRI site (Fuller, 1982). To place this promoter adjacent to the partial ZucZ gene of pMC1403, the plasmid pOP95-15 was cleaved with PvuI and the 3’-protruding ends were removed with T4 DNA polymerase. This DNA was cleaved with EcoRI and ligated to the pMC1403 fragments (above) with T4 DNA ligase. Bacterial strain HBlOl (Bolivar et al., 1977) was transformed according to the procedure of Cohen et al. (1972). Plasmid DNA was screened by the methods of Klein et al. (1980) or Holmes and Quigley (198 1). Our use of restriction enzymes, DNA ligase, and other DNA enzymes has been previously described (Casjens and Huang, 1982; Casjens et al., 1983; Wyckoff and Casjens, 1985; Hayden et al., 1985). A plasmid was

isolated which contains the Zac promoter oriented so that it directs transcription across the truncated IucZ structural gene and was named pTSV-1.

(b) Vectors for cloning translation initiation signals in other reading frames

To modify pTSV-1 so that any fragment which contains a translation initiation signal and an asso- ciated truncated downstream ORF could be inserted into the EcoRI site with its truncated ORF fused in-frame to IucZ gene, DNA containing such a trans- lation start was inserted into the pTSV EcoRI site so that it initiated translation out-of-frame with the ZucZ gene. Short deletions or insertions were then made at the opened SmaI site. After closure and transfor- mation into a fat deletion strain, MC1061 (Casada- ban and Cohen, 1980), mutant clones with trans- lation extending in-frame into the 1acZ gene were identified as blue colonies on XGal plates. The EcoRI insert was then removed to give the altered cloning vector (Fig. 2, legend). These plasmids were named pTSV-2a and pTSV-2b, which contain a -1 shift, and pTSV-3a and pTSV-3b which contain a + 1 shift in reading frame relative to pTSV-1. All of

these vectors produce very low levels of /?Gal activity in liquid assays, and no plasmid-specific proteins of appropriate M, could be visualized on SDS-PA gels. The properties of these plasmids are summa- rized in Table I.

The pTSV plasmids were trimmed to a length of 7200 bp by removal of the sequences between the DruI site 66 bp 3’ to the ZacZ gene (position 116 in the sequence of Buchel et al., 1980) and the DruI site at 3232 in the pBR322 sequence (Sutcliffe, 1979). The resulting plasmids each contain a single DraI site. This deletion caused no change in the rate of /?Gal production from plasmids containing a phage P22 gene 8- or a phage I gene D-IacZ hybrid gene (not shown). The resulting plasmids, pTSV-11, -12a, -12b, -13a and -13b, were named by adding 10 to the parent’s name. Of possible use in the manipulation of these plasmids is the fact that they contain no ApaL BglII, BstEII, KpnI, PstI, ScaI, @I, StuI, %a1 or XhoI (except pTSV-12b) sites.

283

Smal EcoRI , BamHl

EcoRl

Hind Ill lac UV5

BamHl

EcoR I Sma’ Barn HI

IZI lac operon

0 pBR322

IIIUI pMB9 H lac UV5 promoter

Fig. 1. Structure of the plasmids pOP95-15, pMC1403, and pTSV-1. The plasmids pOP95-15 and pMC1403 were used in the construction of pTSV-1. Their nt sequences are largely known from the sequence of the luc operon (Hediger et al., 1985; Fuller, 1982; Kalnins et al., 1982; Buchel et al., 1980) and pBR322 (Sutcliffe, 1979; Peden, 1983). Only the sequence of tetracycline promoter region, derived from pMB9 (Bolivar et al., 1977), is not known.

284

Lac mFiNA b

La& cistron I I b

I A 5LGGAATTGTGAGCGGATMCAATTTCACACAGGAAACAGCT ATG ACC ATG 3'CCTTAACACTCGCCTATTGTTAAAGTGTGTCCTTTGTCGA TAC TGG TAC

I ATT ACG GAT TCA CTG TAA TGC CTA AGT GAC

GCC GTC GTT TTA CAA CGT CGT GAC-3; C,GG CAG CAA AAT GTT GCA GCA CTG-5

'TTAA

I TTAA

I TTAA

I TTAA

I TTAA

AAT TCC GG

AA

SmaI A ATT CCC GGG GAT

GGG CCC CTA

SmaI AAT TCC CCG GGG GAT

GG GGC CCC CTA

XhoI CCC TCG AGG GGG GAT GGG AGC TCC CCC CTA

SmaI TTC CCC CGG GGG GAT

G GGG GCC CCC CTA

AA TTC CCG GAT G GGC CTA

I E pTSV-1

I g pTSV-2a

I ;’ pTSV-2b

I i pTSV-3a

I $ pTSV-3b

Fig. 2. Nucleotide sequence of the pTSV cloning regions. The nucleotide sequence of the E. co/i lac operon in this region is shown above.

The left vertical line below the sequence is the point at which Fuller (1982) created an EcoRI site and the right vertical line is the point

at which Casadaban et al. (1980) created a BarnHI site. The pTSV-1, -2a and -3b nt sequences shown were determined as described

in EXPERIMENTAL AND DISCUSSION, section b. These regions ofpTSV-2b and-3a were not completely sequenced, due to a severe

compression in the sequencing gels which prevented unambiguous determination of the number of GC bp in the lower strand between

the EcoRI and XhoI or SmaI site, respectively. The number of GC bp shown in this position for the later plasmids was deduced from

a number of experiments in which translation starts of known sequence were cloned into the various vectors (see section c). The pTSV

sequences are shown opened (see gaps) at the EcoRI site (GAATTC). The codons are shown separated according to the 1ucZ

translational reading frame. The plasmids with alternate reading frames were created from pTSV-1 as follows. The 5’ portion of P22

gene 8 was isolated from plasmid pP22-508 (Wyckoff, 1984) by (i) cleavage with PstI, (ii) removal of the protruding 3’ ends with T4

DNA polymerase, (iii) ligation to EcoRI linkers, and (iv) cleavage with EcoRI. The 532-bp fragment containing the gene 8 start was

gel-purified and ligated to EcoRI-cleaved pTSV-I. A product of transformation by this DNA which contained a single insert in the

correct orientation, but with the truncated 8 and IacZ genes out-of-frame, was named pTSV-101. This DNA was then cleaved with the

SmaI isoschizomer XmaI and mung bean nuclease. A plasmid resulting from closure of this DNA, which programs high levels of BGal

synthesis (Table I), was named pTSV-102. The phage P22 EcoRI fragment was removed to form the vector pTSV-3b. The remaining

pTSV family members were constructed in an analogous fashion. The phage I 509-bp Suu3A fragment thought to contain the Nu3

translation start (Sanger et al., 1982) was cloned into the BumHI site (which has adjacent flanking EcoRI sites) of plasmid pJB8

(Ish-Horowitz and Burke, 1981). The resulting plasmid (pJB8-509) was cleaved with EcoRI, and the fragment containing the phage

sequence was cloned into the EcoRI site of pTSV-1 (creating plasmid pTSV-509-l), an eight bp _?%I synthetic linker was inserted in

the SmaI site, and the phage EcoRI insert was removed from this plasmid (pTSV-509-2) to generate pTSV-2b. pTSV-2a was made from

pTSV-509-2 by cleaving with XhoI, digesting with mung bean nuclease, ligating and removing the EcoRI insert. Finally the I 478-bp

Sau3A fragment containing the FII gene translation start was cloned into the BarnHI site of pJB8 (forming pJB8-478), and the resulting

EcoRI fragment containing the I sequence was inserted out-of-frame into the EcoRI site ofpTSV-2b. This plasmid (pTSV-478) was then

cleaved with XhoI and mung bean nuclease, ligated and the EcoRI fragment was removed to create pTSV-3a.

TABLE I

Properties of the pTSV plasmid vectors

Plasmid IucZ expression a Colony color b

pTSV-I 0.024 white

pTSV-2a 0.033 white

pTSV-2b 0.649 white

pTSV-3a 0.022 white

pTSV-3b < 0.022 white

pTSV-101 0.261 white

pTSV-102 1000 blue

a BGal activities were determined (setting the value for pTSV-

102 to 1000) in liquid cultures of E. coli carrying the indicated

plasmid according to the method of Miller (1972) as modified by

Kenyon et al. (1980).

b Blue colonies on XGal(40 pg/ml) plates indicate BGal activity.

The host strain, MCI061 (Casadaban and Cohen, 1980), which

carries a deletion of the luc operon, makes white colonies.

285

(c) Nucleotide sequence analysis of the pTSV clon- ing sites

The appropriate EcoRV-BglI fragment from each of the pTSV plasmids was cloned into the M 13mp 10 vector (Messing, 1983) by insertion into vector DNA which had been opened at its AvuII site (made blunt by mung bean nuclease) and its BglI site. Nucleotide sequences from these clones were determined according to the method of Sanger et al. (1980) as modified by Biggin et al. (1983) for use with [ 35S]thiodeoxynucleotides. The primer used is com- plementary to the 1acZ sequence beginning about 60 nt from the EcoRI site (5’-CTGGCGAAAG GGGGATGTGCTGCAAGGCGA-3’, synthe- sized on an Applied Biosystems, Inc. DNA

synthesizer). Fig. 2 shows the nt sequences deter- mined for the cloning regions of the pTSV plasmids. Cloning translation starts of known sequence, namely the phage P22 gene 8 (see section d) and the phage il genes A, FZZ, D and E (L. Sampson, unpub- lished), into one or all of these vectors has confirmed the relationships of each of the live EcoRI sites to the IacZ reading frame.

(d) Characterization of pTSV-generated LacZ- fusion proteins

To test the usefulness of these vectors, several hybrid gene products were characterized in detail. The M, of the hybrid protein synthesized by E. coli carrying plasmid with the in-frame P22 gene 8 trans- lation start (plasmid pTSV-102; Table I) is consis- tent with that expected for the hybrid protein. Immuno-blots of SDS-PA gels, performed accord- ing to the method of Huang et al. (1985), confirmed that this protein is the gene 8-lucZ fusion gene pro- duct, since antibody to the gene 8 protein or to E. coli PGal react specifically with the putative hybrid pro- tein band (Wyckoff, 1984). In addition, this hybrid gene is repressed in vivo by the lac repressor, sug- gesting that it is transcribed from the lac promoter (Wyckoff, 1984).

The hybrid protein was purified by PGal affinity chromatography (Ullmann, 1984; Wyckoff, 1984), and the N-terminal aa sequence was determined with a Beckman Model 980D sequenator. The pri- mary sequence was Pro-Thr-Thr-Glu-Ile-Gln, with contamination by Met and Glu in the first and

second cycles, respectively. The sequence of authen- tic gene 8 protein is Met-Glu-Pro-Thr-Thr-Glu-Ile-

Gln (Wyckoff, 1984). The nt sequence indicates an in-frame stop codon just upstream of the translation start site (Wyckoff, 1984), so that it is virtually certain that translation of the hybrid protein begins at the normal scaffolding protein translation start, and that there is partial removal of the N-terminal dipeptide. Similar N-terminal analyses of two ad- ditional hybrid proteins, made by insertion of the 5’ portion of the phage 2 genes D and E yielded aa sequences (8 and 6 aa determined, respectively) identical to the phage-coded gene D and E proteins (Sanger et al., 1982), respectively.

(e) Conclusions

A family of cloning vectors has been constructed which facilitates the analysis of translation initiation signals in E. coli and the construction of gene fusions to the 1ucZ gene. In the absence of inserted DNA fragments these vectors express very low levels of BGal activity. Thus it is possible to screen directly for inserts containing translation starts which are joined in-frame to the 1acZ gene. Vectors have been con- structed in all three reading frames, so it is possible to choose the appropriate vector if the reading frame of the potential insert is known. These vectors contain a promoter directed toward the cloning site, making them especially useful for genes which are not adjacent to their natural promoters. DNA frag- ments containing translation starts can be cloned into these vectors at the EcoRI site (i) by using naturally occurring EcoRI sites, (ii) by cloning through a plasmid vector such as pJB8 (Ish- Horowitz and Burke, 1981) which has a unique restriction site flanked by nearby EcoRI sites, or (iii) by the use of EcoRI linkers. We have also generated hybrid genes by inserting EcoRI-blunt fragments into the different reading frames using EcoRI + SmaI digested plasmids. To date we have determined the N-terminal aa sequence of P22 gene 8, and I genes D and E initiated hybrid proteins, and all show translation initiation only at their normal start codon. This indicates a possible general use of these vectors to determine the location of naturally occur- ring translation starts.

286

ACKNOWLEDGEMENTS Huang, W.M., Ng, L. and Casjens, S.: Relationship between bacteriophage T4 and T6 topoisomerases. T6 39-protein is equivalent to the combined T4 39- and 60-protein subunits. J. Biol. Chem. 260 (1985) 8973-8977.

Ish-Horowitz, D. and Burke, J.F.: Rapid and efficient cosmid cloning. Nucl. Acids Res. 9 (1981) 2989-2998.

Kalnins, A., Otto, K., Ruther, U. and Muller-Hill, B.: Sequence of the IacZ gene of Escherichia coli. EMBO J. 2 (1983) 593-597.

Kenyon, C.J., Brent, R., Ptashne, M. and Walker, G.C.: Regu- lation of damage-inducible genes in Escherichia co&. J. Mol. Biol. 160 (1980) 445-457.

Klein, R.D., Selsing, E. and Wells, R.D.: A rapid microscale technique for isolation of recombinant plasmid DNA suitable for restriction enzyme analysis. Plasmid 3 (1980) 88-91.

Looman, A.C., De Gruyter, M., Vogelaar, A. and Van Knippen- berg, P.H.: Effects ofheterologous ribosomal binding sites on transcription and translation of the la& gene of Escherichia

co/i. Gene 37 (1985) 145-154. Messing, J.: New Ml3 vectors for cloning. Methods Enzymol.

101 (1983) 20-78. Miller, J.H.: Experiments in Molecular Genetics. Cold Spring

Harbor Laboratory, Cold Spring Harbor, NY, 1972, pp. 352-355.

Minton, N.P.: Improved plasmid vectors for the isolation of translational lac gene fusions. Gene 3 1 (1984) 269-273.

Peden, K.W.C.: Revised sequence of the tetracycline-resistance gene of pBR322. Gene 22 (1983) 277-280.

Sanger, F., Coulson, A.R., Barrell, B.G., Smith, A.J.H. and Roe, B.A.: Cloning in single-stranded bacteriophage as an aid to rapid DNA sequencing. J. Mol. Biol. 143 (1980) 161-178.

Sanger, F., Coulson, A.R., Hong, G.R., Hill, D.F. and Peterson, G.B.: Nucleotide sequence of bacteriophage lambda DNA. J. Mol. Biol. 162 (1982) 729-773.

Shapira, S.K., Chou, J., Richaud, F.V. and Casadaban, M.J.: New versatile plasmid vectors for expression of hybrid pro- teins coded by a cloned gene fused to lad gene sequences encoding an enzymatically active carboxy-terminal portion of /%galactosidase. Gene 25 (1983) 71-82.

Sutcliffe, J.G.: Complete nucleotide sequence of the Escherichia

coliplasmid pBR322. Cold Spring Harbor Symp. Quant. Biol. 43 (1979) 77-90.

Thomas, D.Y., Dubuc, G. and Narang, S.: Escherichia coli plas- mid vectors containing synthetic translational initiation se- quences and ribosome binding sites fused with the lacZ gene. Gene 19 (1982) 211-219.

Ullmann, A.: One-step purification of hybrid proteins which have /f-galactosidase activity. Gene 29 (1984) 27-3 1.

Weinstock, G., Rhys, C., Berman, M., Hampar, B., Jackson, D., Silhavy, T., Weisemann, J. and Zweig, M.: Open reading frame expression vectors: a general method for antigen pro- duction in Escherichia coli using protein fusions to p-galacto- sidase. Proc. Natl. Acad. Sci. USA 80 (1983) 4432-4436.

Wyckoff, E.: Autoregulation of phage P22 scaffolding protein synthesis. Ph. D. Dissertation, University of Utah, Salt Lake City, UT, 1984.

Wyckoff, E. and Casjens, S.: Autoregulation ofthe bacteriophage P22 scaffolding protein gene. J. Virol. 53 (1985) 192-197.

We thank Mark B. Adams for technical assist- ance, and Drs. Stanley Cohen, Forrest Fuller and Ethel Jackson for providing plasmids, host strains and unpublished results. We also thank Dr. Ray Gesteland and Matt Belhnan for oligodeoxynucleo- tide synthesis and Dr. William Gray for aa sequence analysis. This work was supported by NIH grant GM21975 to S.C., NIH grant GM21960 to W.M.H., and NIH predoctoral training grant 5T32GM07531 to E.W.

REFERENCES

Biggin, M., Gibson, T. and Hong, F.: Buffer gradient gels and % label as an aid to rapid DNA sequence determination. Proc. Natl. Acad. Sci. USA 80 (1983) 3963-3965.

Bolivar, F., Rodriguez, R.L., Greene, P.J., Betlach, M.C., Heyneker, H.L., Boyer, H.W., Crosa, J.H. and Falkow, S.: Construction and characterization of new cloning vehicles, II. A multipurpose cloning system. Gene 2 (1977) 95-113.

Buchel, D.E., Groneborn, B. and Muller-Hill, B.: Sequence of the lactose permease gene. Nature 283 (1980) 541-545.

Casadaban, M. and Cohen, S.N.: Analysis ofgene control signals by DNA fusion and cloning in Escherichia coli. J. Mol. Biol. 138 (1980) 179-207.

Casadaban, M., Chou, J. and Cohen, S.N.: In vitro gene fusions that join an enzymatically active p-galactosidase segment to amino-terminal fragments of exogenous proteins: Escherichia

cob plasmid vectors for the detection and cloning of trans- lational initiation signals. J. Bacterial. 143 (1980) 971-980.

Casadaban, M.J., Martinez-Arias, A., Shapira, S.K. and Chou, J.: B-Galactosidase gene fusions for analyzing gene expres- sion in Escherichia coli and yeast. Methods Enzymol. 100 (1983) 293-308.

Casjens, S. and Huang, W.M.: Initiation of sequential packaging ofbacteriophage P22 DNA. J. Mol. Biol. 157 (1982) 287-298.

Casjens, S., Hayden, M., Jackson, E. and Deans, R.: Additional restriction endonuclease cleavage sites on the bacteriophage P22 genome. J. Virol. 45 (1983) 864-867.

Cohen, S.N., Chang, A.C.Y. and Hsu, L.: Nonchromosomal an- tibiotic resistance in bacteria: Genetic transformation of Escherichia coli by R factor DNA. Proc. Natl. Acad. Sci. USA 69 (1972) 2110-2114.

Fuller, F.: A family of cloning vectors containing the lacUV5 promoter. Gene 19 (1982) 43-54.

Hayden, M., Adams, M. and Casjens, S.: Bacteriophage L: chromosome physical map and structural proteins. Virology 147 (1985) 431-440.

Hediger, M., Johnson, D., Nierlich, D. and Zabin, I.: DNA sequence of the lactose operon: the IacA gene and tran- scription termination region. Proc. Natl. Acad. Sci. USA 82 (1985) 6414-6418.

Holmes, D. and Quigley, M.: A rapid boiling method for the preparation of bacterial plasmids. Anal. Biochem. 114 (198 1) 193-197. Communicated by S.R. Kushner.