An improved Escherichia coli expression vector for the construction and identification of full-length cDNA clones

Gene, 44 (1986) 347-351

347

Elsevier GENE

1651

An improved Escherichia coli expression vector for the construction and identification of full-length cDNA clones (Recombinant DNA; tandem promoter; primer; tailed linker; multiple cloning sites; sporamin A)

Kenzo Nakamura*,

Yukimoto Iwasaki and Tsukaho Hattori

Laboratory of Biochemistry, Faculty of Agriculture, Nagoya University, Chikusa, Nagoya 464 (Japan) Tel. (052)781-5111, Ext. 6314 (Received

March

(Revision

received

(Accepted

3rd, 1986) April 13th, 1986)

April 18th, 1986)

SUMMARY

An Escherichiu coli expression vector designed for the efficient synthesis and identification of a full-length cDNA clone is constructed. The vector allows the synthesis of double-stranded cDNAs downstream from the tandem lac control regions employing the vector-primer and linker procedure of Okayama and Berg [Mol. Cell Biol. 2 (1982) 161-1701. Full-length cDNA clones carrying the 5’noncoding region in addition to the entire coding and 3’-noncoding regions can be expressed in E. colt’ cells without fusing their coding region to that of E. coli proteins; these clones are identified by colony immunoassay. The entire cDNA insert can be easily excised from the plasmid, since the multiple cloning sites in the vector are duplicated at both ends of the cDNA insert during its synthesis.

INTRODUCTION

Cloning of the full-length cDNA covering the entire coding region of the mRNA is especially useful for the determination of primary structures of poly* To whom

correspondence

and

reprint

requests

should

be

addressed. Abbreviations: complementary

Ap, ampicillin; to mRNA;

propyl-/?-D-thiogalactopyranoside; SDS, sodium

dodecyl

bp, base pair(s); ds, double

cDNA,

stranded;

DNA

II’TG,

iso-

LB, Luria broth; kb, 1000 bp;

sulfate.

0378-l 119/86/$03.50 0 1986 Elsevier Science Publishers

B.V.

peptides and for the study of their synthesis using appropriate expression vectors. The major difficulty in cDNA cloning lies in the identification of a cDNA clone of particular interest from the cDNA library constructed from poly(A) + RNA. The immunological screening of a cDNA ‘expression’ library (Broome and Gilbert, 1978; Helfman et al., 1983; Young and Davis, 1983) is a method generally applicable for screening a large number of colonies in a relatively short time, and only a specific antibody .is a prerequisite for this screening. In this method, a cDNA preparation is cloned into an expression vec-

(Biomedical Division)

!-lind

Eco Multiple Cloning Site Vector Primer

Tailed Linker

Fig. 1. Construction and structure of pKEN602 (upper part), preparation of vector-primer and tailed linker (middle part), and structure of full-length cDNA clone after synthesis (lower part). An 182-bp PvuII-SmaI fragment carrying the lac control region prepared from pUC8 was cloned into Hind111 site ofpUC19dEco using 8-mer Hind111 linkers to produce pKEN601. The pKEN601 DNA was cleaved with SphI and treated with Sl nuclease. After IO-mer EcoRI linkers were attached to both ends, the DNA was digested with EcoRI and recircularized with T4 ligase. By this procedure, a 13-bp sequence of the pKEN601 which contained the Hind111 site located downstream of the tandem lac control regions was deleted to produce pKEN602. The (dT)-tailed vector-primer and the (dG)-tailed linker can be prepared from pKEN602 as described in EXPERIMENTAL

tor, and colonies expressing an antigenic structure are identified by an immunological assay procedure using specific antibody. Since it is generally believed that fusion of the cDNA coding sequence to the E. coli structural gene sequence on the vector is necessary for its expression in E. coli, the ds cDNA synthesized by the conventional Sl nuclease method, which often results in the loss of 5’ side of the mRNA sequence, is usually used to construct cDNA expression libraries. The major drawback of this approach is that cDNA clones identified usually lack important N’-terminal coding and 5’-noncoding sequence of the mRNAs. We have previously shown (Hattori et al., 1985; Nakamura et al., 1986) that full-length cDNA clones carrying the 5’-noncoding region in addition to the entire coding and 3’-noncoding regions can be identified directly by immunological screening of the cDNA expression library constructed by the application of the vector-primer and linker procedure for the efficient synthesis of full-length cDNAs (Okayama and Berg, 1982) to an E. coli expression vector pUC8 (Vieira and Messing, 1982). The translation of antigen-related polypeptides in cells harboring these cDNA clones apparently starts from the initiation codon inside cDNA (Hattori et al., 1985; Nakamura et al., 1986). This approach has been successfully used for direct immunological identification of full-length cDNA clones for patatin and two other tuber-specific proteins of potato (Nakamura et al., 1986; K.N., unpublished results), a major storage protein (sporamin) (Hattori et al., 1985) and catalase (S. Sakajo, T.H., K.N. and T. Asahi, manuscript in preparation) from sweet potato tuberous roots, and several other plant proteins. AND DISCUSSION, section a. Reaction conditions for homopolymer tailing have been described previously (Nakamura et al., 1986). Most of the multiple cloning sites in the vector are duplicated at both ends of the ds cDNA synthesized in the correct orientation downstream of the tandem lac control regions. The following abbreviations are used for the restriction enzymes. B, BumHI; E, EcoRI; H, HindHI; K, KpnI; Ps, PstI, Pv, PvuII; Sa, SalI; Sm, SmaI; Sp, Sphl; Ss, SstI; X, XbaI. Amp’, ampicillin-resistance gene; Klenow, large fragment of E. cob DNA polymerase I. Parentheses indicate restriction enzyme sites eliminated by homopolymer tailing. Underlined restriction enzyme sites are duplicated at both ends of the ds cDNA. Thick open arrow indicates the [UCcontrol region. The filled-in and open boxes in the ds cDNA indicate coding region and noncoding region, respectively, of mRNA.

349

Although this method considerably simplifies the establishment of a cDNA expression library and the cloning of full-length cDNAs, we encountered several difficulties with pUC8 in the application of the vector-primer and linker procedure. In this paper, we describe the construction of an improved E. coli expression vector specially designed for this purpose.

EXPERIMENTAL

AND DISCUSSION

(a) Construction of pKEN602 and preparation vector-primer and tailed linker

of

We constructed an expression vector pKEN602 (Fig. 1) from pUC8 and pUC19 (Yanisch-Perron et al., 1985). This improved vector carries the multiple cloning sites downstream of the tandem luc control regions, and a Hind111 site located between these two Zuccontrol regions. Previously, we used a short EcoRI-PstI fragment of pUC8 (26 bp) for the preparation of (dG)-tailed linker fragment, and it was difficult to prepare in large quantities because of its small size. Within the multiple cloning sites of pKEN602, there are KpnI and PstI sites which produce 3’-protruding termini upon cleavage. By adding homopolymer (dG) tails to the KpnI termini of pKEN602 followed by cleavage with HindIII, (dG)-tailed linker of approx. 220 bp can be prepared (Fig. 1). The (dT)-tailed vector-primer can be prepared by adding homopolymer (dT) tails to the PstI termini followed by cleavage with EcoRI (Fig. 1). After synthesis of the first strand of cDNA on the vector-primer, homopolymer (dC) tails are attached to the 3’-termini of DNAs. To remove the (dC) tail attached to the EcoRI terminus of the vector-primer, the cDNA: vector-primer is digested with Hind111 which does not cleave DNA substrates forming DNA-RNA hybrids (Okayama and Berg, 1982). Previously, (dC) tail attached to the HincII terminus of pUC8 was removed by EcoRI (Nakamura et al., 1986). (b) Direct expression tandem fuc promoters

of full-length

cDNA under

With the vector-primer and tailed linker prepared from pKEN602, the ds cDNA can be synthesized in

the correct orientation downstream of the tandem luc control regions (Fig. 1). Since the tandem duplication of promoters on the expression vector has been reported to enhance the expression of downstream foreign genes (Goeddel et al., 1980), we compared the expression of a full-length cDNA clone under the Zac promoter and the tandem luc promoters. For this purpose, we chose a cDNA insert of pIM023, which is a full-length cDNA clone of sporamin A identified by immunological screening of cDNA expression library constructed with the pUC8 vector-primer and tailed linker (Hattori et al., 1985). Since a stop codon is located in the 5’-noncoding region of the cDNA in the same coding frame as that of the sporamin precursor, the translation initiated from the 1ucZ gene of the vector can not yield a fusion polypeptide with the sporamin precursor. After the Hind111 site immediately following the poly(dA) tract of the pIM023 cDNA insert was changed to the BumHI site by Sl nuclease trimming and BumHI synthetic linker, an 1.05kb PstI-BumHI cDNA insert fragment was excised and cloned between the PstI and BamHI sites of pKEN601, an intermediate for the construction of pKEN602 (Fig. l), to produce pIM023-601. E. coli JM 103recA - cells harboring pIM023-60 1 gave stronger signal than cells harboring pIM023 by colony radioimmunoassay with anti-sporamin A serum (Fig. 2(A)). When cell extracts from these cells were examined by immtmoblot analysis, a polypeptide migrating slower than mature sporamin was detected (Fig. 2(B), lanes 4 and 6). This polypeptide migrated at the same position as an intermediate formed by the in vitro co-translational processing of a sporamin precursor (preprosporamin) with dog pancreas microsomal membrane (T.H., S. Ichihara and K.N., manuscript in preparation), suggesting that the signal peptide of the preprosporamin expressed from the correct initiation codon is efficiently processed by signal peptidase of the E. coli cells. The amount of prosporamin in cells harboring pIM023-601 was larger than that in cells harboring pIM023, indicating that tandem duplication of the kzc promoter in pKEN602 allows elevated expression of full-length cDNA. (c) Excision of cDNA insert One of the prominent features of the cDNA cloning using the vector-primer and linker fragment

12

3

4

(6) 123456

derived from pKEN602 is that most of the multiple cloning sites in the vector is duplicated at both ends of the ds cDNA during its synthesis (Fig. 1). Using these cleavage sites, the entire ds cDNA insert can be easily excised from the plasmid. A ds cDNA insert with different terminal structures at each end can also be prepared. The ds cDNA insert with asymmetric terminal structure can be inserted into vectors carrying a strong phage promoter (Melton et al., 1984) in desired orientation to produce a large quantity of mRNA by in vitro transcription, which can be used to identify the polypeptide coded by the cDNA by in vitro translation assay. This approach is especially useful for the cloning of less abundant mRNAs where techniques such as hybrid-select translation are hardly applicable for the final identification of cDNA clone selected by immunological screening. (d) Application for cDNA cloning The expression vectors, pKEN601 and pKEN602, have been successfully used for the construction and identification of full-length or nearly full-length cDNA clones for rat serum albumin (N. Iwatsuki, M. Nakano, Y.I., T.H. and K.N., unpublished results) and pea chloroplast 33-kDa protein of the oxygen-evolution system (M. Murase, Y.I., K.N., and A. Watanabe, unpublished results).

Fig. 2. Direct expression of full-length sporamin A cDNA clone in E. coli. (A) Colony radioimmunoassay. E. coli JM103recA cells harboring pUC8 (I), pIM023 (2). pKEN601 (3) and pIM023-601 (4) were grown on a nitrocellulose filter placed on LB-plate containing 50 pg Ap/ml and 2 mM IPTG, and colony radioimmunoassay with anti-sporamin A serum was carried out by the method of Helfman et al. (1983) except that [“‘Ilprotein A and skim milk were used instead of second antibody and bovine serum albumin, respectively. (B) Immuno-blot analysis of cell extracts. E. coli JM103recX cells harboring pUC8 (lane 3), pIM023 (lane 4), pKEN601 (lane 5) or pIM023-601 (lane 6) were grown in LB containing 50 pg Ap/ml and 2 mM IPTG, collected by centrifugation and lysed in boiling 2% SDS. Cellular proteins from equal numbers of the cells were separated by a 12.5% SDS-polyacrylamide slab gel and analyzed by immunoblotting with anti-sporamin A serum and [“‘l]protein A. Lanes 1 and 2 contain 150 and 15 ng, respectively, of purified sporamin. The closed and open arrowheads indicate positions of mature sporamin and prosporamin, respectively.

ACKNOWLEDGEMENT

We are grateful to Dr. S. Mizushima for critical reading of the manuscript.

REFERENCES Broome, S. and Gilbert, W.: Immunological screening method to detect specilic translation pruducts. Proc. Natl. Acad. Sci. USA 75 (1978) 2746-2749. Goeddel, D.V., Shepard, H.M., Yelverton, E., Leung, D., Crea, R., Sloma, A. and Pestka, S.: Synthesis of human fibroblast interferon by E. coli. Nucl. Acids Res. 8 (1980) 4057-4074. Hattori, T., Nakagawa, T., Maeshima, M., Nakamura, K. and Asahi, T.: Molecular cloning and nucleotide sequence of cDNA for sporamin, the major soluble protein of sweet potato tuberous roots. Plant Mol. Biol. 5 (1985) 313-320.

351 Helfman, D.M., Feramisco, J.R., Fiddes, J.C., Thomas, G.P. and Hughes, S.H.: Identification of clones that encode chicken tropomyosin by direct immunological screening of a cDNA expression library. Proc. Natl. Acad. Sci. USA 80 (1983) 3 l-35. Melton, D.A., Krieg, P.A., Rebagliati, M.R., Maniatis, T., Zinn, K. and Green, M.R.: Efficient in vitro synthesis ofbiologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucl. Acids Res. 12 (1984) 7035-7056. Nakamura, K., Hattori, T. and Asahi, T.: Direct immunological identification of full-length cDNA clones for plant protein without gene fusion to E. coli protein. FEBS Lett. 198 (1986) 16-20.

Okayama, H. and Berg, P.: High-efficiency cloning of full-length cDNA. Mol. Cell Biol. 2 (1982) 161-170. Vieira, J. and Messing, J.: The pUC plasmids, an M13mp7derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 9 (1982) 259-268. Yanisch-Perron, C., Vieira, J. and Messing, J.: Improved Ml3 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene33 (1985) 103-l 19. Young, R.A. and Davis, R.W.: Efficient isolation of genes by using antibody probes. Proc. Natl. Acad. Sci. USA 80 (1983) 1194-l 198. Communicated by H. Yoshikawa.