Increased antibody expression from Escherichia coli through wobble-base library mutagenesis by enzymatic inverse PCR

Gene, 123 (1993) l-7 ‘0 1993 Elsevier Science Publishers

B.V. All rights reserved.

0378-l 119/93/$06.00

GENE 06839

Increased antibody expression from Escherichia coli through wobble-base library mutagenesis by enzymatic inverse PCR (Recombinant DNA; screening; protein production level; signal peptide; ribosome binding; LEIPCR; mRNA folding; class-X5 restriction enzymes)

Willem P.C. Stemmer, Hybritech,

Inc., Therapeutics

Received by F. Barany:

Suzanne

K. Morris, Curtis R. Kautzer

Department, P.O. Box 269006, San Diego, CA 92196-9006,

27 May 1992; Revised/Accepted:

18 August/24

August

and Barry S. Wilson

USA. Tel. (619) 455-6700; Fax (619) 453-4124

1992; Received

at publishers:

11 September

1992

SUMMARY

We tested the value of a new library mutagenesis approach, called library enzymatic inverse PCR (LEIPCR), for expression-level enhancement of antibody Fv fragments produced in Escherichia coli. The production level of active, metal chelate-specific antibody from our constructs is limited by a low expression level of the second, heavy-chain cistron. To increase the production level, LEIPCR was applied to the wobble bases of the second cistron leader peptide. In LEIPCR mutagenesis, the entire plasmid is amplified using mutagenic primers with class-IIS restriction endonuclease (ENase) sites at their 5’ ends. The PCR product is digested with the class-X3 ENase (here, BsaI; GGTCTCN’NNNN,), which removes its own recognition sequence, and the ends are self-ligated. Thus, LEIPCR can be used to make plasmid mutant libraries regardless of the nucleotide sequence, and independent of available ENase sites. The resulting library of lo7 wobble mutants was screened for active Fv by a colony filter lift. A selected mutant was shown to produce between four- and elevenfold more active Fv than the wild type (wt), and fivefold more heavy chain. Mutations outside of the leader peptide were shown not to be involved. The mutated areas of the mRNAs of two different up-mutants may have less secondary structure than the wt. Thus, the sequence of the mRNA of the second leader peptide was limiting to the expression level of heavy-chain and active Fv.

INTRODUCTION

The level of foreign proteins that E. coli can be made to secrete into the periplasm or especially into the medium is usually very low (Better et al., 1988; Skerra and Pluckthun, 1988; Skerra et al., 1991b; Takkinen, 1991). Secretion into the medium of foreign proteins in E. coli is determined by a large number of factors and optimizaCorrespondence to: Dr. W.P.C. Stemmer at his present address, Affymax Research Institute, 4001 Miranda Ave., Palo Alto, CA 94304, USA. Tel. (415) 496.2300/2329; Fax (415) 424-0832. Abbreviations: aa, amino acid(s); Ap, ampicillin; bp, base pair(s); cfu, colony-forming unit(s); dNTP, deoxyribonucleotide triphosphate; EIPCR, enzymatic inverse PCR, ENase, restriction endonuclease; Fv,

tion has been very tedious. Only after extensive fermentation development has it been possible to obtain high expression levels of secreted protein, but with such a finely tuned system even a minor change in aa sequence can give a lOO-fold decrease in expression level (Carter et al., 1992). Before time-consuming fermentation development, the construct is optimized by molecular biology techniques. 25.kDa binding fragment of mAb; IPTG, isopropylthiogalactoside; kb, kilobase or 1000 bp; LEIPCR, library EIPCR, mAb, monoclonal antibody(ies); nt, nucleotide(s); oligo, oligodeoxyribonucleotide; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; PolIk, Klenow (large) fragment of E. coli DNA polymerase I; RBS, ribosome-binding site(s); SDS, sodium dodecyl sulfate; Tc, tetracycline; wt, wild type; [ 1, denotes plasmid-carrier state.

pMCHAFv1 (1833 bp)

Fig. 1. Expression

vectors.

Plasmid

pMIN1

is a I-kb expression

vector

which contains

a synthetic

lnc promoter,

supF (Huang

et al., 1987) as the

selectable marker, and a rep- ColEI origin. The vector must be used with commercially available chemically or electro-competent E. coli MC1061[P3] cells (Invitrogen Inc., San Diego, CA). These cells contain amber mutations in both the Ap- and Tc-resistance genes, located on a P3 plasmid, which is too big to interfere with plasmid purification. Transformants are selected for on plates with 25 lg Ap/ml and 7.5 ,ug Tc/ml. Plasmid pMCHAFvlis the 1.8.kb expression vector that was used for EIPCR mutagenesis and Fv expression. The light chain of the Fv fragment is followed by the heavy chain, each chain having its own OmpA signal peptide, and both chains are driven by a single lac promoter. For nonexpression work, the cells are grown in M9CA medium (Sambrook extract/24 g tryptone/3 g NaH,P04/

et al., 1989). For colony lift screening assays, the cells are plated on 23 x 23 cm plates with CS agar (48 g yeast 3 g Na,HP04/15 g agar, all per liter) with 0.5 pg IPTG/ml for induction. For expression level determination,

clones are grown in CS broth with 0.2 mM IPTG in baffled shaker flasks at 250 rpm for 30 h at 30°C with a boost of 0.2 vol. of 240 g/l yeast extract and 120 g/l tryptone after 18 h. The Fv expressing constructs are always grown at 30°C. CS medium was developed in our lab especially to obtain greater expression levels. Using CS broth, rather than M9 or LB, we can use higher levels of IPTG before overexpression lethality occurs, and most of the Fv ends up in the media rather

than in the periplasm.

B, BarrlHI, Bs, BssHII,

Bsp, BspHII,

E, EcoRI: H, NindIII,

N, NheI, Sp, SphI, X,

XbaI

The antibody used here, CHA255, reacts with indiumEDTA-containing haptens. Such haptens can use “IIn or can be more complex structures which include indiumEDTA on one end with “Y chelated to another part of the hapten (Reardan et al., 1985; Phelps et al., 1990). The approx. 25-kDa binding domains of an antibody are called the Fv fragments and consist of a noncovalently associated light and heavy chain. Our plasmid construct is dicistronic, with the light chain followed by the heavy chain. With the CHA255 Fv, the chains comigrate on SDS-PAGE and a chain ratio could not be obtained, A different Fv sequence in an otherwise identical vector showed that the heavy chain was expressed at a much lower level than the light chain. The heavy chain in this case could be separated from the light chain by SDS-PAGE. From published studies even minor changes in the RBS are known to have up to a lOOO-fold effect on expression level and this site is frequently implicated in expression optimization (Curry and Tomich,1988; Olins et al., 1988; Min et al., 1988; Jespers et al., 1991). However, our con-

struct already had the strongest known RBS, from T7 gene ZO ( Olins et al., 1988). The sequence of the mRNA of the leader peptide is known to affect the RBS (Scherer et al., 1980; Coleman et al., 1985; Stanssens et al., 1985; Altuvia and Oppenheim, 1986), although it is not part of the traditional AGGA motif. In one case it was shown that this effect was due to obstruction of the RBS by the mRNA secondary structure (Schauder and McCarthy, 1989). Therefore, we hypothesized that the structure of the mRNA of our leader peptide was limiting the heavychain expression level. In similar situations, library mutagenesis combined with a screening assay has been an efficient approach to expression level improvement without delving in detail into the regulatory relationships (Min et al., 1988; Bucheler et al., 1990; Jespers et al., 1991). Therefore, we decided to mutagenize the wobble bases of the signal peptide of the heavy chain, without changing the aa sequence, in the belief that this would enable us to identify improvements in the amount of heavy-chain and active antibody expressed.

3

. PCR Template: uncut plasmid, maximum about 4 kb.

Mutations

- PCR: - denature

in wobble

bases of leader peptide

- anneal primers - extend with Tm or Pfu

mutations -_ ---__

I’ _’

Primer A

Bsal

4 --_

,‘-

-_

_-._

Bsal

I

Primer B

. fill in ends cut with Bsal gel purify full length DNA ligate - electroporate - colony screen - colony purify * sequence mutated area of selected clones

+

l

mutations

l

bases of leader peptide

in wobble

l

Fig. 2. Library enzymatic inverse PCR (LEIPCR). A variety of methods have been used for library mutagenesis, but none are based on inverse (whole vector) PCR (Ochman et al., 1988; Triglia et al., 1988). In LEIPCR, adapted from EIPCR (Stemmer and Morris, 1992) the entire plasmid is amplified with a library of primers with identical class-IIS ENase sites in their 5’ ends. Digestion with class-IIS ENases, which have their recognition sequence 5’ to and separated from their cut site (Szybalski et al., 1991), such as BsaI (GGTCTCN’NNNN,), then allows the removal of the entire recognition sequence while leaving the PCR DNA with compatible overhangs on each end. Efficient intramolecular ligation of this PCR product yields a large library of plasmid mutants. LEIPCR works fine with plasmids up to about 4-5 kb, but works even better with small plasmids, such as the I-kb expression vector reported here. PCR optimization has allowed amplification of fragments up to about 10 kb (Maga and Richardson, 1991; Kainz et al., 1992; Ponce and Micol, 1992).

RESULTS

AND DISCUSSION

(a) Expression system and LEIPCR mutagenesis The plasmid vector for CHA255 Fv production is shown in Fig. 1. LEIPCR was developed from EIPCR

(Stemmer and Morris, 1992) as a new library mutagenesis method (Fig. 2). EIPCR can give up to 95% correct mutants and does not depend on the nt sequence or available ENase sites, due to the nature of the class-IIS restriction sites (Szybalski et al., 1991; Fig. 2). The 3’ end of the

lac

---..

A

stop

‘-l.a

Template Sequence

ABS

GATAG~~~TATATCATGAAATLAGACAGCTATCGCCATTGCAGTGGCACTGGCTGGTTTCGCCACCGTGGCGCAGGCCGAGGTGACCCTGGTGGAGTCT CTATCTTCCTCTATATAGTACTTTTTCTGTCGATAACGCTCGGTGGCACCGCGTCCGGCTCCACTGGGACCACCTCAGA ~“#p&~~ *::x:.,.+,,:::,{, ,*: (M K K T A I A I A V A L A G
+esian

F

A

T

V

A

Q

AjJ

V.T.

X.,

0

E

S

v

E

s

Heavy Chain

1

I

primer A GTGACCGACCAAAGCGGTGGCACCGCGTCCGGCTCCACTGGGACCACCTCAGA ]M

K

KTAIAI

AGFATVAQ

E

v

,P

L

Heavy Chain

ompA Leader Peptide

i

.... l...

5

/

Selected higher expressionwobble sequence

C

mutant 2 leader

ATG~C~G~~T~GC~TTGC~~T~GC~CT~GC~GG~TT~GC~C~GT~~C~C~~GC~

IM

K

K

T

A

I

A

IAVALAG

FATVAQAl

Fig 3. EIPCR primer design. (A) The area of the original plasmid that was replaced by a library of linkers. (B) The mutagenic primers relative to the template. Sequence B is a convenient drawing to design the primers, but is not a reaction intermediate. Sequence C is that of the selected mutant. The mixed nt positions in the oligos were synthesized using a fresh 1:l:l:l molar mixture of each of the 4 nt in the fifth reservoir of the DNA synthesizer (Hermes et al., 1989), which was previously shown to result in an approx. 1:l:l:l ratio of nt in unselected clones (unpublished results). The oligos were purified with Nensorb Prep nucleic acid purification columns (NEN-DuPont, Boston, MA). PCR reactions (100 ~1) contained 0.5 PM of each purified primer/O.5 ng pMCHAFv1 template plasmid DNA/l x Taq buffer/ZOO PM of each dNTP/O.S ~1 Taq polymerase (Perkin Elmer/Cetus, Norwalk, CT). Thermal cycling was performed on a Perkin Elmer/Cetus PCR machine with the following parameters: 94”C/3 min for one cycle; 94”C/l min, 5O”Cjl min, 72”C/2 min for three cycles; 94”C/l min, 5O”C/l min, 72”C/2 min, with autoextension at 5 s/cycle for ten cycles; 94”C/l min, 5O”C/l min. 72”C/3 min, with autoextension at 8 s/cycle for twelve cycles; followed by one IO-min cycle at 72°C. The DNA obtained from two 100 ~1 PCR reactions was flushed by addition of dNTPs to 200 PM, 20 units PolIk and 30 units T4 DNA polymerase and incubated at 37°C for 30 min. After phenol/chloroform extraction and precipitation, the DNA is digested with BsaI (New England Biolabs, Beverly, MA) which cuts off the ends of the PCR fragment, inside the oligos. The digested DNA is gel-purified, ethanol-precipitated, and ligated at low concentration and without polyethylene glycol to favor intramolecular reaction, which has the potential to be very efficient, The ligation is ethanol-precipitated using ammonium acetate, washed twice with 80% ethanol, vacuum-dried and resuspended in 20 yl 0.1 x TE (10 mM TrisHCl pH S.OjO.1 mM EDTA) for electroporation. Amounts of 1~1 of the ligation reaction were electroporated into 20 ~1 of MC1061[P3] cells (Invitrogen, San Diego, CA) using the Invitrogen electroporator. Sequence C shows the sequences of the wt and the two selected up-mutants. The second line (‘Library’) shows the maximum extent of mutagenesis that still encodes the native OmpA leader sequence (‘leader’). The mutagenized wobble bases are shown in shaded boxes, and the wobble bases that were altered in the two selected mutants are underlined. The long underlines indicate stems of possible mRNA folding loops that were obtained by computer analysis, Four stems were observed in the wt, vs. one in each up-mutant.

primers is a 20-nt exact match to the template. Although small vectors work particularly well, with minimal optimization, LEIPCR works well for vectors of 44.5 kb. PCR optimization has allowed amplification of fragments up to about 10 kb (Maga and Richardson, 1991; Kainz et al., 1992; Ponce and Micol, 1992). For larger vectors, we have

successfully used a two-fragment approach, which gives a lower ligation efficiency and a smaller library. (b) Wobble base library Fig. 3 shows how LEIPCR was applied to the expression level optimization of antibody Fv fragments secreted

5

Fig. 4. Screening results. The electroporated cells were plated on 23 x 23 cm plates at 1 x lo5 colonies per plate. Nitrocellulose in 3% non-fat milk in 25 mM TrisHCl pH 7.5, washed with 25 mM Tris, followed by incubation in 25 mM Tris containing loaded

p-thioureabenzyl

EDTA (EOTUBE;

Reardan

et al., 1985) per filter for 1 h at room

temperature.

filter lifts were blocked 20 pCi of illIn or ‘OY

The filters were washed

again,

dried

and

autoradiographed for several hours. Areas of the petri plate corresponding to the strongest signals were replated, reassayed and colony-purified. The two best clones were further characterized. The sequence of a 130-bp fragment containing the mutated area was determined by double-stranded dideoxy sequencing. No mutations outside of the targeted wobble bases were observed. The dots in circles are reference points for alignment with the agar plate. The dot indicated with the arrow is up-mutant 1

into the medium by E. coli. The library was made in the wobble bases of the signal peptide of a RBS that we hypothesized to be limiting the expression of the heavy chain. The size of the library was determined to be 1 x lo7 cfu by plating. The efficiency was 2 x lo7 cfu per pg of DNA. All of the mutants in this library should encode a native OmpA signal peptide aa sequence, and the mutants should differ only in the ompA DNA and mRNA sequence. The antibodies produced by different colonies from this library were therefore identical in terms of affinity and specificity. Any increase in the signal strength of colonies in the screening assay must be attributed only to an increase in the expression level, which may caused by several mechanisms. For example, the mRNA secondary structure could be different (Schauder and McCarthy, 1989), the ribosome reinitiation could be better (Scherer et al., 1980; Stanssens et al., 1985; Coleman et al., 1985; Altuvia and Oppenheim, 1986), or the codon usage could be better (Makoff et al., 1989; Sorensen et al., 1989). Codon usage is unlikely to be the major factor, since the

leader we used is from a highly expressed,

native protein

(OmpA). The expression level is known to be influenced strongly by the sequence of the first few codons after the ATG start codon (Scherer et al., 1980; Stanssens et al., 1985; Coleman et al., 1985; Altuvia and Oppenheim, 1986). Wobble base libraries affect all of these factors, and can generally be used to increase expression level without changing the aa sequence, regardless of which of the above factors is rate limiting.

(c) Library screening Fig. 4 shows the autoradiographic result of the primary screen of a filter with radiolabeled hapten. Approximately 1 x lo5 clones were screened per 23 x 23 cm colony filter lift, and a wide range of signals was obtained. Only a fraction of this range is due to normal variation which was also observed with controls. A similar bacterial colony lift assay has also be used to detect Fv specific for protein antigens (Dreher et al., 1991; Skerra et al., 1991a).

6 (d) Clone characterization Eight colonies with the strongest signals were colonypurified and compared to the wt by filter lift assay. The two best clones gave an about tenfold greater autoradiography signal than the wt. Assay of cell-free supernatants of induced shaker cultures of these two clones also showed a tenfold increase in the amount of active Fv fragment secreted into the medium. With any mutagenesis procedure there is a risk of introducing mutations in areas other than the target. To demonstrate that the observed expression level increase is due only to the targeted nt, a 130-bp fragment containing the mutated area was cloned back into the wt vector. The constructs derived from both mutants again expressed at an about tenfold higher level than an identically constructed wt sequence, proving that the expression level increase is only due to the 130-bp fragment. The nt sequences of the 130-bp fragment of the two mutants differed from the wt sequence only at the targeted wobble bases, confirming that the aa sequences were not altered by the mutagenesis. The sequence comparison of the wt and the mutants is shown in Fig. 3C. (e) Expression quantitation The amount of active Fv secreted into the medium was quantitated by two separate assays which were based on the separation of antibody-bound [“lIn] hapten from free [ll’In]hapten. In one assay, the Fv-containing medium was bound to nitrocellulose using a dot-blot manifold, and in the other, the antibody was separated from free hapten by centrifugation in Ultrafree concentrators of IO-kDa cutoff (Millipore, Bedford, MA). In both cases, the amount of active Fv was determined by extrapolation to a standard curve generated with native CHA2.55 antibody. The results of the dot-blot showed that mutant 1 expressed about elevenfold more active Fv then the wt clone (57 vs. 5 ng/ml; the linear regression curve fitted the data of several dilutions with a least-squares factor R2 = 0.994). In the other assay format, mutant 1 showed a fourfold increase in active Fv over the wt (97 vs. 24 ng/ ml; R2 = 0.992). Finally, the level of heavy chain of the two clones was analyzed by densitometric tracing of Western blots where the heavy chain was detected using a mAb we had produced to a specific peptide (FQEAYRRFYGPV) that was added to the heavy-chain C terminus (Poser et al., 1980). The results of scanning both periplasm and medium showed that the mutant 1 clone expressed fivefold more heavy chain in the medium and about fourfold more in the periplasm than did the wt clone (data not shown). (f) Computer analysis of mRNA secondary structure Similar to the results of Bucheler et al. (1990), the observed nt changes in the two up-mutants were primarily

(11:5) from ‘strong’ bases (C or G) to ‘weak’ bases (A or T). The secondary structure of the mRNA near the mutated area was analyzed using an RNA folding program (Loops; DNAstar, Inc., Madison, WI). The wt scored four possible loops in the mutated area, whereas both upmutants scored only one loop each (Fig. 3C). These data suggest that both mutant mRNAs are likely to have a reduced secondary structure, which may have caused the increased expression level. Since none of the loops have more than a 6 nt stem, this interpretation is speculative. Alternatively, although LEIPCR mutations are limited to a small area of a plasmid, it is conceivable that rare mutations can affect an area located far away on the mRNA. Such second-site mutants or revertants are known to occur at the mRNA level (Mitraki et al., 1991). Computer analysis of the folding of our whole mRNAs or large mRNA fragments did not yield conclusive differences in the AGO’ of secondary structure. (g) Conclusions (I) LEIPCR is a quick and powerful tool for making large site-directed mutant libraries. Using only two primers, any area of a plasmid can be targeted, independent of nt sequence or available ENase sites. A limitation of LEIPCR is the size of the DNA that can be amplified. (2) Library mutagenesis and screening can be an efficient tool for expression level optimization. With a simple mutagenesis tool like LEIPCR and a reliable screen, improvements can be obtained quickly, despite limited biochemical understanding of the control mechanisms. (3) The mutagenesis results support our hypothesis that the sequence of the wobble bases of the leader peptide was limiting the expression of heavy chain. Computer analysis of the mRNA secondary structures was inconclusive, but suggested that the expression level increase is possibly due to decreased mRNA secondary structure in the mutants.

REFERENCES Altuvia, S. and Oppenheim, A.B.; Translational regulatory signals within the coding region of the bacteriophage lambda cl11 gene. J. Bacterial. 167 (1986) 415-419. Better, M., Chang, C.P., Robinson, R.R. and Horwitz, A.H.: Escherichia coli secretion of an active chimeric antibody fragment. Science 240 (1988) 1041-1043. Biicheler, U.S., Werner, D. and Schirmer, R.H.: Random silent mutagenesis in the initial triplets of the coding region: a technique for adapting human glutathione reductase-encoding cDNA to expression in Escherichia coli. Gene 96 (1990)27 l-276. Carter, P., Kelley, R.F., Rodrigues, M.L., Snedecor, B.. Covarrubias, M., Velligan, M.D., Wang, W.L., Rowland, A.M., Kotts, C.E.. Carver, M.E., Yang, M., Bourell, J.H., Shepard, H.M. and Henner. D.: High level Escherichia coli expression and production of a bivalent humanized antibody fragment. Biotechnology 10 (1992) 163-167.

7 Coleman,

J., Inouye,

the ribosome

M. and Nakamura,

binding

K.; Mutations

site affect translational

upstream

from

efficiency. J. Mol. Biol.

181 (1985) 139-143. Curry, K.A. and Tomich, C.S.C.: Effect of ribosome binding site on gene expression in Escherichia coli. DNA 7 (1988) 173-179. Dreher, M.L., Gherardi, E., Skerra, A. and Milstein, C.: Colony assays for antibody fragments expressed in bacteria. J. Immunol. Methods 139 (1991) 197-205. Hermes, J.D., Parekh, SM., Blacklow, SC., Koster, H. and Knowles, J.R.: A reliable method for random mutagenesis: the generation of mutant libraries using spiked oligodeoxyribonucleotide primers. Gene 84 (1989) 143-151. Huang, H.V., Little, P.F.R. and Seed, B.: Improved suppressor tRNA cloning vectors and plasmid-phage recombination. In: Rodriguez, R.L. and Denhardt, D.T. (Eds.), Vectors: A Survey of Molecular Cloning Vectors and Their Uses. Butterworth, London 1987, pp. 269-283. Jespers, L., Sonveaux, E., Fastrez, J., Phanapoulos, A. and Davison, J.: Overexpression of the phage lambda lysozyme cloned in Escherichia coli: use of a degenerative mixture of synthetic ribosome binding sites and increase of the protein stability in vivo. Protein Eng. 4 (1991) 4855492. Kainz, P., Schmiedlechner, A. and Strack, H.B.: In vitro amplification of DNA fragments > 10 kb. Anal. Biochem. 202 (1992) 46-49. Maga, E.A. and Richardson, T.: Amplification of a 9.0 kb fragment using PCR. Biotechniques 11 (1991) 1855186. Makoff, A.J., Oxer, M.D., Romanos, M.A., Fairweather, N.F. and Ballantine, S.: Expression of tetanus toxin fragment in E. coli: high level expression by removing rare codons. Nucleic Acids Res. 17 (1989) 10191-10202. Min, K.T., Kim, M.H. and Lee, D.S.: Search for the optimal of the ribosome binding site by random oligonucleotide mutagenesis. Nucleic Acids Res. 16 (1988) 5075. 5087.

sequence directed

Mitraki, A., Fane, B., Haase-Pettingell, ‘C., Sturtevant, J. and King, J.: Global suppression of protein folding defects and inclusion body formation, Science 253 (1991) 54-58. Ochman, H., Gerber, AS. and Hartl, D.L.: Genetic applications of an inverse polymerase chain reaction. Genetics 120 (1988) 621-623. Olins, P.O., Devine, C.S., Rangwala, S.H. and Kavka, K.S.: The T7 phage gene 10 leader RNA, a ribosome-binding site that dramatically enhances the expression Gene 73 (1988) 227-235.

of foreign

genes in Esckerickia coli.

Oliphant, A.R. and Struhl, K.: An efficient method for generating proteins with altered enzymatic properties: application to betalactamase. Proc. Natl. Acad. Sci. USA 86 (1989) 909449098. Phelps, J.L., Beidler, D.E., Jue, R.H., Unger, B.W. and Johnson, M.J.: Expression and characterization of a chimeric bifunctional antibody with therapeutic applications. J. Immunol. 145 (1990) 1200&1204.

Ponce,

M.R. and Micol, J.L.: PCR

amplification

of long DNA

ments. Nucleic Acids Res. 20 (1992) 623. Poser, J.W., Esch, F.S., Ling, N.C. and Price, P.A.: Isolation

frag-

and se-

quence of the vitamin K dependent protein from human bone. J. Biol. Chem. 255 (1980) 8685-8691. Reardan, D.T., Meares, CF., Goodwin, D.A., McTigue, M., David, G.S., Stone, M.R., Leung, J.P., Bartholomew, R.M. and Frincke, J.M.: Antibodies against metal chelates. Nature 316 (1985) 2655267. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Cloning. A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989. Schauder, 3. and McCarthy, J.E.G.: The role of bases upstream from the Shine-Dalgarno region and in the coding sequence in the control of gene expression in Esckerickia coli : translation and stability of mRNAs in vivo. Gene 78 (1989) 59-72. Scherer, G.F.E., Walkinshaw, M.D., Arnott, S. and Morre, D.J.; The ribosome binding sites recognized by E. coli ribosomes have regions with signal character in both the leader and protein coding segments. Nucleic Acids Res. 8 (1980) 389553907. Skerra, A. and Pluckthun, A.: Assembly of a functional immunoglobulin Fv fragment in Esckerickia coli. Science 240 (1988) 103881040. Skerra; A., Dreher, M.L. and Winter, G.: Filter screening of antibody Fab fragments secreted from individual bacterial colonies: specific detection of antigen binding with a two-membrane system. Anal. Biochem. 196 (1991) 151-155. Skerra, A., Pfitzinger, I. and Pluckthun, A.: The functional expression of antibody Fv fragments in E. colt improved vectors and a generally applicable purification technique. Biotechnology 9 (199 1) 2733278. Sorensen, M.A., Kurland, C.G. and Pedersen, S.: Codon usage determines translation rate in Esckerickia coli. J. Mol. Biol. 207 (1989) 365-377. Stanssens, P., Remaut, E. and Fiers, W.; Alterations upstream from the Shine-Dalgarno region and their effect on bacterial expression. Gene 36 (1985) 211-223. Stemmer, W.P.C. and Morris, S.K.M.: Enzymatic inverse polymerase chain reaction: a restriction site independent, single fragment method for high-efficiency site directed mutagenesis. Biotechniques 13 (1992) 2155220. Szybalski, W., Kim, S.C., Hasan, N. and Podhajska, A.J.: Class-IIS restriction enzymes - a review. Gene 100 (1991) 13-26. Takkinen, K., Laukkanen, M.-L., Sizmann, D., Alfthan, K., Immonen, T., Vanne, L., Kaartinen, M., Knowles, J.K.C. and Teeri, T.T.: An active single-chain antibody containing a cellulase linker domain is secreted by Esckerickia coli. Pro. Eng. 4 (1991) 837-841. Triglia, T., Gregory-Peterson, M. and Kemp, D.J.: A procedure for in vitro amplification of DNA segments that lie outside the boundaries of known sequences. Nucleic Acids Res. 16 (1988) 8186.