An Escherichia coli plasmid vector system for high-level production and purification of heterologous peptides fused to active chloramphenicol acetyltransferase

Gene, 126 (1993) 109-113 0 1993 Elsevier Science Publishers B.V. All rights reserved. 0378-l 119/93/$06.00

109

GENE 06999

An Escherichia coli plasmid vector system for high-level production and purification of heterologous peptides fused to active chloramphenicol acetyltransferase (Fusion proteins; cut; ColEl ori; tat promoter; synthetic gene; affinity purification; chloramphenicol rubella)

caproate agarose;

Johan Robben”, Gaby Massie”?, Eugkne Bosmansb, Bernadette Wellensb and Guido Volckaert” “Laboratory of Gene Technology, K.U.Leuven, B-3001 Leuven, Belgium; and b Eurogenetics N. V., Transportstraat 4, Industriezone West, B-3980 Tessenderlo. Belgium. Tel. (32-13) 66 88 30

Received by A.J. Podhajska: 25 April 1992; Revised/Accepted: 11 December/l4 December 1992; Received at publishers: 17 December 1992

SUMMARY

A very small plasmid vector system is described for construction and high-level production of C-terminal chloramphenicol acetyltransferase (CAT) fusion proteins in Escherichia cd. The only functional elements of the plasmid are a minimal region of the ColEl origin of DNA replication and the Tn9 cat gene, both under control of a tat promoter. Since C-terminal fusion to CAT does not interfere with chloramphenicol (Cm) resistance, plasmids are maintained under Cm selection. Because of its small size (1392 bp), the system is especially convenient for building and expression of synthetic genes and gene fragments. This concept was utilized to generate a fusion with a synthetic gene encoding the multipleepitope fragment from the rubella virus El membrane protein. Affinity-purified fusion proteins were obtained in mg amounts from lOO-ml batches of culture fluid, and incorporated as a specific antigen in a rubella immunoglobulin G enzyme-linked immunosorbent assay.

INTRODUCTION

Translational gene fusion is a frequently used strategy for expression and purification of heterologous peptides and proteins. In recent years, several high-level expression Correspondence to: Dr. J. Robben, Laboratory of Gene Technology, K.U.Leuven, Willem de Croylaan 42, B-3001 Leuven, Belgium. Tel. (3216) 28 66 11; Fax (32-16) 22 07 61. TDeceased.

Abbreviations: Asor,, absorbance at 600 nm wavelength; aa, amino acid (s); bp, base pair(s); CAT, Cm acetyltransferase; cat, gene encoding CAT; Cm, chloramphenicol; ELISA, enzyme-linked immunosorbent assay; Ig, immunoglobulin; IPTG, isopropyl-B-o-thiogalactopyranoside; LB, Luria-Bertani (medium); MCS, multiple cloning site (polylinker); ori, origin of DNA replication; PAGE, polyacrylamide-gel electrophoresis; P,,,, B-lactamase promoter; P,,,, tat promoter; RV, rubella virus; SDS, sodium dodecyl sulfate; ’ (prime), denotes a truncated gene at the indicated side.

systems have been developed in which the target peptide or protein is almost unvaryingly fused to an affinity purifiable ‘carrier’. This carrier can be a whole protein such as the maltose-binding protein (di Guan et al., 1988) and glutathione S-transferase (Smith and Johnson, 1988), or a protein fragment such as the protein A IgG-binding domain (Lowenadler et al., 1987). Also oligopeptide tails have been used successfully (Hochuli et al., 1988; Hopp et al., 1988). A similar approach has been to fuse target proteins to the N terminus (Friefeld et al., 1985) or the C terminus (Dykes et al., 1988) of CAT. We introduced an alternative approach of cat gene fusion by using cat both as the selectable marker and as the gene fusion partner, taking advantage of the fact that C-terminal fusion to CAT does not affect Cm resistance. This has allowed the development of a plasmid fusion expression vector of a minimal size, particularly convenient to the construction of synthetic genes.

110 EXPERIMENTAL

AND DISCUSSION

(a) Vector encoding CAT protein fusions The prototype cat fusion vector pJRtac99

(Fig. 1) is

derived from the multicopy chemical degradation sequencing vector pGV451 (Volckaert, 1987) in which, essentially, the constitutive Pbla promoter is replaced by the IPTG-inducible P,,, . Besides P,,, , the vector carries the Tn9 cut gene as the only selectable marker, and the artificial HyRep2 ori (Volckaert, 1987). The multiple cloning site (MCS) inserted

into the natural

ScaI site near the 3’

end of the cut gene is used to clone target

sequences,

or

new polylinkers for specific applications. C-terminal peptide or protein fusion to CAT does not destroy CAT activity (Dykes et al., 1988), hence, addition of Cm to the culture medium ensures selection on CAT fusion protein synthesis. By combining functions of both resistance marker and fusion carrier in one gene, a vector of minimal size (1392 bp) has been realized, facilitating manipulations and cloning, since the number of restriction sites is much reduced within the vector. This is highly desirable for construction of synthetic genes by assembly of oligonucleotide cassettes, as well as for cassette mutagenesis.

(b) IPTG-induction and plasmid compatibility The pJRtac99 ori is a derivative of the pBR327 ori (ColEl ori) of which the natural RNA11 promoter and the major part of the region coding for the antisense RNA I control element had been deleted (Volckaert, 1987). Initiation of replication relies on read-through transcription from the cat-controlling

P,,,. However,

in all pJRtac99-

bearing E. coli strains tested, even /UC repressor overproducing (l~c1~) strains BMH71-I 8 and JM 109 (YanischPerron et al., 1985), repression of the P,,, is apparently inadequate, since high amounts of plasmid DNA (estimated up to 10000 copies per cell) and of CAT protein (up to 30% of total cell proteins) accumulate. Addition of IPTG had little or no visible effect on plasmid or CAT production. A single exception is lacZq strain WK6 (Zell and Fritz, 1987) which efficiently blocks the P,,,. When grown in the absence of IPTG, production of vector and CAT protein were very low, but sufficient to confer Cm resistance. On the other hand, cultures grown in the presence of 0.1 mM of IPTG showed both vector DNA and CAT protein production at high levels (Fig. 2), confirming the anticipation of P,,,- controlled plasmid replication. In the same strain, high-level expression of (hybrid) CAT protein could be induced equally well by adding IPTG in the mid-exponential phase. Fusion proteins were harvested 2 to 4 h after induction. The use of the HyRepZ ori in pJRtac99 has one drawback: insertion of a transcription terminator or of a large

AGTACTGCAGCGTCGACCAGGGACCCGGGTAATGAAGAGTCTAGATCTACT YCSVDOGPG* 213

B

KdSI NarI EheI EagI Asp1 (PstI] GCCGTCTGGTTGGCGCCACCCCGGAACGGCCGAGACTGCGTCTGG ACGTCGGCAGACCAACCGCGGTGGGGCCTTGCCGGCTCTGACGCAGACCAGCT CSRLVGATPERPRLRLV 242 EP2

BsiWI

[Sal11

[SalI]

[AvaI]

TCGACGCAGACGACCCGCTGCTGCGTACGGCACCGGGTC GCGACTGCTGGGCGACGACGCAAGCCGTGGCCCAGGGCC DADDPLLRTAPGP

EP3 BstEII BdmHI [AvaI] [Pa-I] CCGGGGAAGTTTGGGTTACCCCGGTTATCGGATCCCAGGCGCGC CCTTCAAACCCAATGGGGCCAATAGCCTAGGGTCCGCGCGATT GEVWVTPVIGSQAR* EPl 286

Fig. I. Structure of CAT fusion prototype vector pJRtac99 and of synthetic sequences encoding rubella virus El epitopes used in the construction of CAT-El epitope fusion plasmids. (A) Plasmid pJRtac99 consists of the HyRep2 orifrom pGV451 (Volckaert, 1987), the P,,, promoter from pDR540 (Pharmacia) and the cat gene from Tn9 The original MCS, inserted into the natural ScaI site of car, was designed to accommodate synthetic oligodeoxynucleotides encoding specific RV El epitopes (see panel B). The MCS can be easily replaced by other MCS so as to enable the construction of other synthetic genes (unpublished results). Unique HindIII, NcoI (containing the cat start codon) and &HI sites are available for subcloning of the cat gene fusion with or without P,,, into other vectors. The total sequence of pJRtac99 will appear in the EMBL, GenBank and DDBJ Sequence Databases under the accession No. X69551).(B) Synthetic oligodeoxynucleotide cassettes encoding one of the previously identified epitopes, EP2, EPI and EP3 (Terry et al., 1988). The indicated internal restriction sites do not occur in pJRtac99. The individual epitopes were cloned in frame into the MCS of pJRtac99. Enzymes used in these constructions are indicated in bold. Subsequently, the epitopes were fused to each other in different combinations (EP2 + EP3; EP3 + EPI; EP2 + EPI and EP2 + EP3 + EPI) by exchange of the appropriate restriction fragments. The plasmid containing all three epitope sequences encodes the entire antigenic cluster from CYSTS* to Arg285 of the El protein, without redundant aa sequences generated by translation of MCS sequences. Details on the construction of these plasmids can be obtained from the authors upon request.

111 duced absence

(unpublished

the regulatory -

+ IPTG

___----. - IPTG

results).

in the pJRtac99

RNAI-RNA11

- IPTG

Mr

1

2

3

5

6

7

8

9

Mr

+-

C

- IPTG

other

interacting

hand, coding

domain

in tram in strains

(c) Fusion with rubella virus (RV) El epitopes The immunodominant RV antigen is the envelope

+ IPTG 4

the

puc19 pJRtac99

+ IPTG

123456789

the for pur-

ports an advantage by making the vector compatible with any vector carrying the intact ColEl ori, such as the pBR and pUC vectors. This can be useful, e.g., to provide plasmid-encoded lac repressor able of repressing the P,,,.

B

On

ori of the sequence

tein El (Wolinsky

et al., 1991). A segment

aa

harbours

245

to

285,

clonal antibody-defined neutralizing epitopes,

a cluster

incap-

pro-

of it, comprising of three

erythrocyte-binding EP2, EP3 and EPl

mono-

and virusin sequential

order (Terry et al., 1988). By means of in-house developed computer software, a corresponding gene with codon usage optimized for E. coli high-level expression was designed to maintain even bp composition and to include unique restriction sites at regular intervals along the DNA sequence. Three synthetic DNA cassettes, each encoding one of the three characterized epitopes were cloned individually into the pJRtac99 MCS, and subsequently fused to each other by rearrangement of appropriate restriction fragments (Fig. 1). The expressed fusion proteins showed the expected molecular weights on SDS-PAGE (Fig. 3). Expression levels of the individual fusions were equal to or only

Fig. 2. IPTG-induction of pJRtac99 plasmid replication and CAT protein expression. Cultures started from single colonies from freshly transformed E. coli WK6 cells (Zell and Fritz, 1987) were grown to A,,,, = 0.5, frozen at -70°C in 8% glycerol, and used (100 ~1) to inoculate 100 ml of LB supplemented with Cm (25 pg/ml) in Erlenmeyer flasks. IPTG (0.1 mM, final concentration) was added at the time of inoculation. Growth was followed by measuring of the absorbance at Ahoo (A). Samples corresponding to 6 x 10’ cells were taken during different growth phases (dashes), and analysed for plasmid DNA (B) and total cell proteins (C). To determine pJRtac99 plasmid contents, cells were pelleted, resuspended and mixed with 2 pg of pUC18 as internal standard. Plasmid DNA was subsequently extracted according to Birnboim and Doly (1979), and uniquely cut with PstI prior to separation by agarose gel electrophoresis. Lane M,, molecular weight markers (PstIdigested phage h DNA). Lane numbers in panels B and C correspond to samples indicated in panel A.

target sequence between the cat gene and the ori may cause negative effects on plasmid copy number and, consequently, on the expression level. Nevertheless, genes of more than 500 bp were cloned with minor reduction of the fusion protein yield. Insertion of the pho.4 gene from E. co/i JM83 as a promoterless, terminatorless 1433-bp fragment downstream from the cat reading frame did not significantly influence the amount of plasmid DNA pro-

Fig. 3. 1% SDS-15% PAGE analysis of total cellular proteins from E. coli cultures over-producing CAT proteins fused with RV El epitopes EPI, EP2 and EP3. Size markers (in kDa) are indicated at the right cells promargin. CAT,, , cellular proteins from pJRtac99-harbouring ducing unfused CAT.

112

slightly lower than that of unfused CAT. The larger the hybrid protein produced, however, the more culture growth was delayed, indicating toxicity of the RV epitopic peptides to E. coli cells. Western blot analysis confirmed specific reactivity of the entire RV El antigenic cluster with human rubella IgG-positive antisera, and with a mouse monoclonal RV antibody, lE7 (data not shown). Fusions with only one or two of the epitopes showed no reaction. One reason might be that the El antigenic determinants are conformational - rather than sequential epitopes for which a particular structured domain is a prerequisite to be recognized as a RV antigen by human antibodies. CAT-antigen hybrid proteins were single step purified by affinity chromatography on commercial Cm-agarose, yielding up to 3 mg of fusion proteins from 100 ml of batch culture fluid. Fig. 4 shows an SDS-PAGE analysis of the purified fusion protein carrying all three epitopes. The 29-kDa major band corresponds to size predicted from the gene fusion sequence. A faint lower band corre-

sponds to 25 kDa, and runs at about the same position as unfused CAT. This indicated that the hybrid protein was slightly sensitive to host proteases at the fusion junction. Purified CAT-antigen fusion proteins bearing all three epitopes were used as such in a RV IgG ELISA. CAT or CAT-antigen fusion proteins were coated onto microtitration wells and incubated with human RV IgG antiserum. Standard curves were obtained from a rubella IgG positive antiserum pool from 16 patients, diluted in IgGnegative serum. Bound antibodies were detected with horseradish peroxidase conjugated anti-human IgG antiserum (Dakopatts), and the difference in absorbance at 429 nm was determined between reactions of human antiserum with CAT and CAT-antigen fusion proteins. Although a high background signal was generated which was not reduced by addition of unfused CAT during serum incubation, standard curves showed a proper linear relation between RV antibody titer, and specific antibodies bound to the recombinant antigen (data not shown). A population study is to be carried out to test the applicability of the CAT antigen fusion proteins as a diagnostic reagent.

ACKNOWLEDGEMENTS

30

J. R. was supported by a post-doctoral fellowship from the Instituut tot Aanmoediging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw-I.W.O.N.L.. This research was funded in part by contract ETC-007 in the framework of the Vlaamse Actieprogramma Biotechnologie of the Flemish governement. This paper is dedicated by G.V. and J.R. to the memory of Gaby Massie, who died on April 15, 1991.

20 REFERENCES

14.4 Fig. 4. 1% SDS-15% PAGE of CAT hybrid protein bearing the three epitopes EPl, EP2 and EP3 (Terry et al., 1988) affinity purified from E. coli over-producing cultures. Cells grown in 100 ml batch cultures were centrifuged, the pellet resuspended in 4 ml of lysis buffer (50 mM TrisHCl pH S/2 mM EDTA/O.I mM 2-mercaptoethanol) and incubated with lysozyme (3O”C, I5 min) in the presence of phenylmethylsulfonyi fluoride (0.1 mM) and DNase I (Iug/ml) (Boehringer), prior to lysis by repeated freeze-thawing (3 to 5 cycles). The supernatant (16 000 x g, 30 mm) containing the soluble protein fraction was applied on a Cm caproate agarose (Sigma) cofumn (2 ml bed volume) and washed with 20 ml of 50 mM TrisHCl pH 7.8/2 mM EDTA/l M NaCI/O.l mM 2-mercaptoethanol. CAT fusion proteins were subsequently eluted with 5 mM Cm prepared in the above buffer. CAT , purified unfused CAT produced from pJRtac99.

Bimboim, H.C. and Daly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7 (1979) 1513-1523. di Guan, C., Li, P., Riggs, P. D. and Inouye, H.: Vectors that facilitate the expression and purification of foreign peptides in Escherichia co/i by fusion to maltose-binding protein. Gene 67 (1988) 21-30. Dykes, C.W., Bookless, A.B., Coomber, B.A., Noble, S.A., Humber, DC. and Hobden A.N.: Expression of atria1 natriuretic factor as a cleavable fusion protein with chloramphenicol acetyltransferase in Escherichia coli. Eur. J. Biochem. 174 (1988) 411.-416. Friefeld, B.R., Korn, R., de Jong, P.J., Sninsky, J.J. and Horwitz, MS.: The 140-kDa adenovirus DNA polymerase is recognized by antibodies to Escherichia c&synthesized determinants predicted from an open reading frame on the adenovirus genome. Proc. Natl. Acad. Sci. USA 82 (1985) 2652-2656. Hochuli, E., Bannwarth, W., Dobeli, H., Gentz, R. and Stiiber, D.: Genetic approach to facilitate purification of recombinant proteins

113 with

a novel

metal

chelate

adsorbent.

Bio/Technology

61 (1988)

1321-1325.

expressed

in Escherichia

coli as fusions with glutathione

S-transfer-

ase. Gene 67 (1988) 31-40.

Hopp, T.P., Prickett, KS., Price, V.L., Libby, R.T., March, C.J., Cerretti, D.P., Urdal, D.L. and Conlon, P.J.: A short polypeptide marker

Volckaert, G.: A systematic approach to chemical DNA sequencing by subcloning in pGV451 and derived vectors. Methods Enzymol. 155

sequence useful for recombinant protein identification and purification. Bio/Technology 6 (1988) 120441210. Liiwenadler, B., Jansson, B., Paleus, S., Holmgren, E., Nilsson, B., Moks, T., Palm, G., Josephson, S., Philipson, L. and Uhlen, M.: A gene

(1987) 231-250. Yanisch-Perron, C., Vieira, J. and Messing, J.: Improved Ml3 phage cloning vectors and host strains: nucleotide sequences of the

fusion system for generating 58 (1987) 87-97. Smith, D.B. and Johnson,

antibodies

against

short peptides.

K.S.: Single step purification

Gene

of polypeptides

Ml3mpl8 Zell, R. and

and pUC19 vectors. Gene 33 (1985) 103-119. Fritz, H.-J.: DNA mismatch-repair in Escherichin coli

counteracting dues. EMBO

the hydrolytic deamination J. 6 (1987) 1809-1815.

of 5-methyl-cytosine

resi-