Gene, 97 (1991) 113-117

113

Elsevier

GENE

03839

A vector for directional cloning and expression of polymerase chain reaction products in Esciherichia coli (Expression

vector;

ferredoxin;

recombinant

DNA)

Mark E. Brandt, Ann H. Gabrik and Larry E. Vickery Department of Physiology and Biophysics, Universityqf Cal$ornia, Irvine, CA 92717 (U.S.A.) Received by G. Wilcox: 17 July 1990 Rev%sed: 27 August 1990 Accepted: 29 August 1990

SUMMARY

This paper describes the construction of a modified vector for the cloning and expression of protein-encoding genes in Escherichia coli. The vector, pfXblue, is derived from the system originally developed by Nagai and Thogerson [Nature 309 (1984) SIO-812], but contains a modified multiple cloning site (MCS) from M13mp18 to allow directional insertion of foreign coding sequences. The MCS is located within the Ml3mpI8 ZacZ’ gene and thus allows blue/w~te screening of colonies for inserts. The inserted gene is expressed as a fusion protein, which, when cleaved by the coagulation factor Xa protease, yields the mature product. This vector was successfully used for the production of a mitochondrial[2Fe-2Slferredoxin using polymerase chain reaction products generated from a chick kidney cDNA library.

INTRODUCTION

The synthesis of eukaryotic proteins in a prokaryotic system such as E. co& is commonly used to produce large quantities of protein for enzymatic and/or st~ctur~ studies. The accumulation of a foreign gene product, however, may have deleterious effects on the growth of the bacterial host. In addition, the eukaryotic protein may be prone to degradation by endogenous proteases. The approach of Nagai and Thargerson (1984) avoids some of these problems and has been used successfully by a number of laboratories for the production of foreign proteins (cf.,

Correspondenceto: Dr. L.E. Vickery, Department of Physiology Biophysics, University of California, Irvine, CA 92717 (U.S.A.) Tel. (714)856-6580; Fax (714)856-8540.

and

aa, amino acid(s); Ap, ampicillin; bp, base pair(s); cFd, chick mt ferredoxin; CIAP, calf intestinal alkaline phosIPTG, isopropyl-/?-D-thiogalactopyranoside; Km, kanamycin;

Abbreviations:

gene encoding phatase;

MBN, mung-bean nuclease; MCS, multiple cloning chondrial; nt, nucleotide(s); oligo, o~g~eox~bonucieotide; DNA replication; PCR, poiymerase chloro-3-indolyl-@-D-galactoside.

0378-l 119/91/$03.50

0

1991 Elsevier

chain reaction;

Science Publishers

site;

XGal,

mt, mitoori, origin of 5-bromo4

B.V. (Biomedical

Division)

Nagai and Thogerson, 1987). In this system, the expression vector is transformed into a defective bacteriophage A lysogenie E. coli strain. The synthesis of the exogenous protein, under the control of the /1pL promoter and the temperaturesensitive cl857-encoded repressor, occurs only after inactivation of the repressor by temperature shift; use of the repressor prevents synthesis of the potentially toxic exogenous product during the growth phase of the cells. The exogenous protein is then produced as a fusion with the first 31 aa of the I CII protein followed by a heptapeptide that includes a recognition sequence for the coagulation factor Xa protease. The fusion protein produced by this method tends to be more stable in the E. coli than the foreign protein alone (Hellebust et al, 1989); the protein of interest can be cleaved from the leader by the factor Xa protease. In addition, the vector supplies the AUG translation start site, allowing the expression in mature form of proteins in which N-terminal sequences would normally be removed during post-translational processing. The method of Nagai and Thogerson (1984) requires a number of steps for insertion of coding sequences into the expression vector (cf., Coghlan and Vickery, 1989). We have constructed a modified vector that allows insertion of

114 EXPERIMENTAL

a variety of sequences in a directional manner in frame with the a cII and factor Xa-coding region in a single step and which allows blue/white screening of the colonies for the presence of the insertion. Our aim was to construct a vector

(a) Construction of the expression vector Construction ofthe expression vector pfXblue is outlined in Fig. 1. It was first necessary to remove the BamHI site from the M13mp18 multiple cloning site to allow correct processing of a later intermediate with BamHI. M13mp18

that has general usefulness; however, it is designed as a system for expressing genes amplified from cDNA libraries by PCR. Hi&III

HindI

AND DISCUSSION

Be111

HindI ;

StiI

Ml3mp18

multiple

Synthetic

linker

Factor

El

Ezl

cloning site

Xa sequence

xcn sequence Modified 2ac.Z’

Purify 700 bp fragment

B&II ECORI .&OR1

ECORI Synthetic

’ Hi&III

B&II

linker 1. Ligate

5’-CCTCCGCGGCCATGG 3’-GGAGGCGCCGGTACCTTAA

2.HindIII Sk41

EcoRI

I

Hin,dII I Hi&III BamHI

Ligate HmdIII HtidIII

HindI

HindI

BUIAI ShII -L$ small fragment

Bar&II

9

Ligate h HindI

PCR

ZucZ’ 5’

ModifiedlaG

PCR Primers:

Primer

5’-CCTACCATGGTTACGAATTCGAGC

S&l Ligate

__)

3’ Primer

BamHI

5’-TCATATGTACCGCGGTTGAT

Fig. 1. Construction

of the expression

vector ptXblue.

Restriction

endonucleases

and other enzymes

used for DNA manipulation

were obtained

from

Boehringer-Mannheim, New England Biolabs, Stratagene, or United States Biochemicals. The plasmids pfX8 and pMb3 were gifts from Charles Glabe (University of California, Irvine, CA) (Coghlan and Vickery, 1989); pMb3-EcoRI was produced by digesting with EcoRI followed by MBN, and then ligating the resulting blunt ends. Oligos for construction or for use as PCR primers were purchased from either the Biotechnology Instrumentation Facility (University of California, Riverside, CA), or Operon Technologies, Inc. (Alameda, CA). The E. coli strain DHSaF’IQ (Bethesda Research Laboratories, Gaithersburg, MD) was used for propagation of the intermediate products techniques (Sambrook et al., 1989). PCR (Saiki et al., 1988) was carried

during construction. out using methods

DNA manipulations adapted from those

were performed using standard included with the United States

Biochemicals GeneAmp Kit (United States Biochemicals, Cleveland, OH). Intermediate products were partially characterized by restriction mapping. The final construct was verified by double-stranded dideoxy sequencing (Sambrook et al., 1989) using an oligo primer corresponding to a segment of the I CII protein.

115 was

cleaved

with

BamHI

and

the

overhanging

ends

digested with MBN. The resulting blunt ends were ligated to yield an M 13mp 18 lacking the Barn HI site. The ligation mixture from this and all subsequent steps generating circular plasmids was used to transform competent DHSaF’IQ cells, followed by purification of the plasmids from the cells. The MCS of the modified M13mp18 was excised and purified as a 700-bp EcoRI-BgZII fragment. Oligos comprising a linker were synthesized, added in excess to the EcoRIBgZII fragment, and ligated. A 62-bp StuI-Hind111 fragment was released with HindIII, purified, and ligated to the large fragment of S&I + HindIII-digested pfX8. The resulting plasmid (pfX8mcs) was digested with BamHI + Hind111 to yield the 79-bp fragment required for the next step. The vector pMb3 contains an EcoRI site (Remaut et al., 1983; Varadarajan et al., 1985). This site was removed using the method described above for the removal of the BamHI site from M13mpl8 to allow future use of the unique EcoRI site from the MCS. EcoRI-deleted pMb3 was digested with BamHI + HindIII, and then used to accept the 79-bp BamHI-Hind111 fragment to yield pc2fX.

BamHI stu1 --NcoI GGATCCATCGAGGGTA-ACCATGGTTACGAATCC

Fig. 2. The structure the factor

EcoRI

of pfXblue. The underscored

Xa recognition

site. The pMb3-derived

. . . MCS

aa residues sequences

comprise (i.e., the

entire vector excluding the lacZ’ and the factor Xa region) consist of (clockwise from the IacZ’): the pUC19 lad gene, the ori and ApR gene from pBR322, an incomplete segment of a KmR gene, and the stated sequences from 1. This structure (determined by sequencing some of the intermediate

product

plasmids)

does not entirely

match

the structure

expected from the description ofthe construction ofpMb3 (Remaut et al., 1981; 1983; Varadarajan et al., 1985). The nt sequence data for pfXblue have been submitted to GenBank and assigned the accession number M36415.

The expression

vector pc2fX was improved

by the incor-

poration of a modified ZacZ’ gene from M 13mp18 to allow blue/white screening of colonies for insertions while maintaining the unique restriction sites from pc2fX. To produce the modified IacZ’, two PCR primers were synthesized. The 5’ primer contained two modifications (shown in italics) to the M13mp18 sequence. The first 3 nt (CCT), introduce one half of a St~1 site and do not anneal to the M 13mp 18. In addition, the M 13mp 18 sequence CCATGA was changed to CCATGG to introduce an NcoI site. These changes encode a LacZ’ protein that contains two modifications other than the A CII/factor Xa fusion: an extra Pro following the factor Xa sequence and the substitution of Val for the Ile at position 3 of the LacZ’ peptide (Fig. 2). The 3’ PCR primer contains a single nt mismatch to create a Sac11 site; this primer is complementary to a region downstream from the ZacZ’ stop codon. The modified ZucZ’ was amplified from an M 13mp18 template by PCR, phosphorylated with T4 polynucleotide kinase, and ligated to pc2fX that had been sequentially digested with HindIII, treated with MBN to remove the Hind111 overhang, digested with ,%I, and dephosphorylated with CIAP. Competent DHSaF’IQ cells transformed with the ligation mixture were plated on XGal and IPTG. One of the resulting blue colonies was sequenced to confirm that it contained the correct insert and was designated pfXblue. Fig. 2 depicts the structure of the completed vector and lists the 13 unique restriction sites available for cloning. Additional sites exist in the M13mp18 MCS (e.g., BamHI, PstI, and SacI) but these also exist in the region of the vector originating from pMb3 (the pL promoter and 2 cl1 sequences, and the pBR322-derived ori and the ApR gene). (b) Expression of chick ferredoxin We tested the expression vector by utilizing it to express the gene (cFd) coding for chick mt [2Fe_2S]ferredoxin. The general method is presented diagrammatically in Fig. 3. To amplify the cFd gene from the cDNA library, two PCR primers were designed based on the published sequence of the cFd gene (Kagimoto et al., 1988). The mt leader sequence was omitted from the construct by choosing the 5’ primer such that it began with the first nt of the mature protein coding sequence. The 3’ primer was chosen to be complementary to a sequence in the 3’-noncoding region containing a unique Hind111 site. A 400-bp band was amplifled from phage particles of a chick kidney cDNA library by PCR. This product, the putative cFd cDNA, was digested with Hind111 and phosphorylated with T4 polynucleotide kinase to produce double-stranded DNA suitable for ligation. This was then ligated into StuI + HindIII-cleaved pfXblue, and transformed into DHSaF’IQ cells.

116 Non-coding region 1. PCR 2. Kinase

Hind111

Hind111

\

f3’ PCR Primer Mature protein coding sequence

Leader sequence

Hind111

Hind111 .

Hin stu1

Ligate

Fig. 3. Insertion

of foreign DNA into pfXblue.

This figure illustrates

the general

method

for inserting

PCR products

into pfXblue.

The restriction

site

used in this example is the Hind111 site used for cFd. The cFd cDNA was amplified from a chick kidney library using two 23-mer PCR primers (5’ primer: 5’-TGCAGCTCAGAAGATAAAATAAC and 3’ primer: 5’-TAATTAAGCTTCTTTAAAACTCC). The chick kidney cDNA library in 1Zap (prepared by Stratagene)

was a gift of Dr. H. Henry

(University

of California,

Riverside,

Approximately 30 colonies were observed, of which six were blue. Plasmid DNA from one of the white colonies was partially sequenced to confirm that the cFd gene had inserted in-frame with the a ~11 and factor Xa region. This DNA was then used to transform competent MZ-1 cells. Production of the chick ferredoxin fusion protein was induced by incubation at 42” C for 3 h followed by incubation overnight at 37’ C. Chick ferredoxin was purified by methods developed for the human protein (Coghlan et al., 1988). As was found for the recombinant human ferredoxin produced in E. coli (Coghlan and Vickery, 1989) the

1.2 -

0.3 -

O250

300

400

350

Wavelength

450

500

(nm)

Fig. 4. Optical absorption spectrum for recombinant chick ferredoxin. The vector pfXblue/cFd was transformed into the defective 3, lysogen MZ-1 (Nagai and Thogerson, 1984); the cells were grown and induced to synthesize the chick ferredoxin protein. The protein was purified according homologue

to the methods (Coghlan

developed

et al., 1988).

for the purification

of the human

CA).

[2Fe-2S] center of the chick protein is assembled in vivo under these conditions of expression. The yield was approx. 6 mg of crude chick ferredoxin/liter of culture. Then absorption spectrum of the purified protein (Fig. 4) is similar to that of the human and bovine ferredoxins. (c) General applications of the vector for PCR products While this vector can be used for the expression of any DNA, it is of special utility for the expression of PCR products. One method, as is illustrated by the expression of the cFd, requires knowledge of nt sequences at each end of the coding region for the protein of interest. The 5’ PCR primer is designed to begin with the first nt of the coding region, and the 3’ primer is chosen complementary to a region following the stop codon. If a unique restriction site does not exist in the downstream noncoding region, the 3’ primer can be used to create one during the PCR process, either by single nt mismatches, or by additional noncomplementary sequences added to the 5’-end of this primer. In some cases this vector can also be used to express proteins for which only N-terminal sequence information is available. For this procedure, it is necessary to use a form of ‘single-sided’ PCR (Berchtold, 1989), in which a PCR primer is designed to match the known 5’ sequence (either an exact match or a degenerate primer based on the aa sequence). The 3’ primer must then be designed to complement known sequences from the library vector (e.g., sequences surrounding the insertion site in 2). In the case of J libraries, the cDNA inserts are cloned into an EcoRI site; since there is a unique EcoRI site in the cloning region of pfXblue, this enzyme can be used to guide the directional

117 insertion of the amplified cDNA into the vector, assuming that no EcoRI sites exist in the gene. Because the 3’ primer anneals to all of the DNA in the template mixture, only the 5’ primer confers specificity to the product and must be carefully designed to avoid the amplification of incorrect sequences from the template mixture.

REFERENCES Berchtold,

M.W.: A simple method

for direct

cloning

and sequencing

cDNA by the use of a single specific oligonucleotide and oligo(dT) in a polymerase chain reaction (PCR). Nucleic Acids Res. 17 (1989) 453. Coghlan, V.M. and Vickery, L.E.: Expression of human ferredoxin and assembly of the [2Fe-2S] center in Eschenkhia coli. Proc. Natl. Acad. Sci. USA 86 (1989) 835-839. Coghlan, V.M., Cupp, J.R. and Vickery, L.E.: Purification

(d) Conclusions We have modified the E. coli expression vector developed by Nagai and Thsgerson (1984) to incorporate several new features. As with the original vector, the expressed material is synthesized as a cleavable fusion protein using the Met codon supplied by the 1 CII protein as a translation start signal. The insertion site for the foreign gene, however, now begins with the StuI recognition sequence which allows for direct ligation of blunt cDNA fragments in frame with the factor Xa peptide recognition sequence - this is especially useful for cloning of DNA products derived by PCR. In addition, twelve other unique restriction sites were introduced to guide the directional insertion of the DNA of interest. Finally, the vector contains the ZucZ’ gene and therefore allows visual selection of colonies with insertcontaining recombinant plasmids.

zation

of human

placental

ferredoxin.

and characteri-

Arch. Biochem.

Biophys.

264

(1988) 376-382. Hellebust,

H., Murby,

Different

M., Abrahmsen,

approaches

Bio/Technology Kagimoto,

J.L., Waterman,

amino acid sequence

Biochem.

Biophys.

of hybrid

Enzymol. Remaut,

lambda. Remaut,

M.:

testis ferredoxin.

protein

of p-globin

produced

by sequence

in Escherichiu coli.

and sequence-specific

pro-

in Escherichia coli. Methods

produced

153 (1987) 461-481. P. and Fiers,

expression Gene

controlled

W.: Plasmid

vectors

by the pL promoter

for high-

of coliphage

15 (1981) 81-93.

E., Stanssens,

mature

S.O.:

protein.

155 (1988) 379-383.

H.C.: Synthesis

proteins

E., Stanssens,

efficiency

fusion

and Kagimoto,

chicken

H.C.: Generation

K. and Thogerson,

teolysis

M.R.

of mature

Res. Commun.

K. and Thegerson,

specific proteolysis of a hybrid Nature 309 (1984) 810-812. Nagai,

a recombinant

7 (1989) 165-168.

K., McCarthy,

Deduced Nagai,

L., Uhlen, M. and Enfors,

to stabilize

P. and Fiers, W. Inducible

human libroblast

interferon

high level synthesis

of

in Escherichia coli. Nucleic Acids

Res. 14 (1983) 4677-4689. Saiki, R.K., Gelfand, ACKNOWLEDGEMENTS

D.H., Stoffel, S., Scharf,

G.T., Mullis, K.B. and Erlich, H.A.: amplification of DNA with a thermostable

S.J., Higuichi,

R., Horn,

Primer-directed enzymatic DNA polymerase. Science

239 (1988) 487-491.

We thank Dr. Helen Henry for providing the chick kidney cDNA library. This work was supported by National Institutes of Health research grants DK30109 and GM43548 (to LEV) and training grant T32 CA09054 (MEB).

Sambrook,

J., Fritsch,

E.F. and Maniatis,

T.: Molecular

Laboratory Manual, 2nd ed. Cold Spring Harbor Spring Harbor, NY, 1989. Varadarajan,

R., Szabo,

A. and Boxer,

Escherichia coli, and reconstitution Acad.

S.G.:

of human

Sci. USA 82 (1985) 5681-5684.

Cloning.

Laboratory,

Cloning,

expression

myoglobin.

A

Cold in

Proc. Natl.