239, No. 2, June,
Cloning and Expression
of the metE Gene in Escherichia
JENNIFER CHU,* ROBERT SHOEMAN, JEAN HART,2 TIMOTHY ANTHONY MAZAITIS,3** NORMAN KELKER,3,* NATHAN AND HERBERT WEISSBACH Roche
Biology, Roche Research Center, Nutley, New Jersey, 07110 and *Department New York University Medical School, New York, New York, 10016 Received
A X-transducing phage was isolated that contains the w&E gene. This gene codes for N5-methyl-H4-folate:homocysteine methyltransferase (EC 126.96.36.199), an enzyme that catalyzes the terminal reaction in methionine biosynthesis. A 9.1-kb EcoRl fragment of this phage, containing the w&E gene, was then cloned into pBR325. This plasmid, pJ19, was used to transform Escherichia coli strain 2276, a metE mutant, and restore the MetE+ phenotype. Although the transformed cells produced large amounts of the metE protein in vivo, in vitro studies using pJ19 as template showed low synthesis of the metE protein. o 19% Academic press, IN.
The final step in the biosynthesis of methionine involves a methyl transfer from N5-methyl-H4-folate5 to homocysteine. In Escherichia coli there are two enzymes that can carry out this conversion [see (l-3) for reviews]. One of these, termed the Blz-dependent methyltrans* This paper is dedicated to Dr. B. L. Horecker on the occasion of his 70th birthday, ’ Present address: Bridgewater-Raritan East High School, Martinsville, N. J. 08836. 3 Present address: Enzo Biochem., Inc., 325 Hudson Street, New York, N. Y. 10013. * To whom correspondence should be addressed. ’ Abbreviations used: N5-methyl-H4-folate-Glu$, DL-N5-methyl-H,-tetrahydrofolate containing three glutamic acid residues; Ap’, ampicillin resistant. SAM, S-adenosylmethionine; NaDodSO,, sodium dodecyl sulfate; Hft, high-frequency transducing; Lft, low-frequency transducing, X, bacteriophage lambda; thi, thiamine; leu, leucine; val’, valine resistant; metE, N’-methyl-H,-folate; homocysteine methyltransferase (EC 188.8.131.52); metF, N5,“-methylene-H,folate reductase (EC 184.108.40.206); metH, E&dependent N’-methyl-H,-folate: homocysteine methyltransferase (EC 220.127.116.11); W&J, methionine regulon repressor; metK, methionine adenosyltransferase (EC 18.104.22.168); ret, recombination gene. 467
ferase (EC 22.214.171.124), contains a tightly bound cobalamin prosthetic group (4-6) and uses N5-methyl-H4-folate derivatives containing one or more glutamate residues as substrate. S-Adenosylmethionine (SAM) and a reducing system (l-3) are also required, and there is good evidence that a methyl-Blz enzyme is an intermediate in the reaction (7, 8). The other enzyme, referred to as the non-Blz-methyltransferase (EC 126.96.36.199), requires as substrate N5-methyl-H4-folate with three or more glutamates, and is stimulated by phosphate ions (9, 10). The latter enzyme does not contain a cobalamin prosthetic group and the reaction does not require SAM or a reducing system. The B12-dependent methyltransferase, which is the product of the metH gene, is constitutive in E. coli, although the enzyme is not active unless the organism is grown in the presence of the vitamin or an appropriate analog (11). The synthesis of the non-Blz methyltransferase, which is coded for by the metE gene, is known to be repressed by either methionine or vitamin B12and involves two distinct mechanisms 0003-9861/85 Copyright All rights
0 1985 by Academic Press, Inc. of reproduction in any form reserved.
(12, 13). The effect of methionine on the Mucta Surviving lysogens were subjected to penicillin m&E mutant expression of the metE gene is similar to enrichment and a temperature-sensitive was isolated. MetE+ survivors of heat induction were that seen with other genes in the methiisolated and characterized. Isolation of survivors onine biosynthetic pathway (metH being an exception) (12), since the expression of was greatly facilitated by carrying out heat induction in the presence of rabbit antibodies to phage Mu all of these genes is repressed when the (21). The growth media used have been previously organism is grown in the presence of high described (20). levels of methionine (14). Two regulatory The isolation of Mu24 AmetE+-1, -2, -7, and -11 genes are involved in this process, the lysogens was as described by Leisinger et al (22), metJ and metK genes. The former is be- based on the Schrenk and Weisberg procedure (23), lieved to code for a repressor protein (13- except the Lft lysate was transduced into Mu24, a 16) whereas the metK gene product is m&E mutant, and selection was for MetE+ phenotype. Measurement of lytic and transdu.&ng titers. The methionine adenosyltransferase (EC transducing and lytic titers were determined as 188.8.131.52) (17, 18). Thus, it has been postu(21). The indicator strains used lated that SAM and the metJ protein are described previously for determining the transducing and lytic titers the components involved in the repression were JEN7 and RW262, respectively. A helper phage, of the methionine biosynthetic pathway XcI857Sam7, was also required for the transducing by methionine. The effect of vitamin BIz titer. on the expression of the m&E gene is Preparation of phage DNA and analysis. Mu24 more specific since only one other enzyme, Xm.etE+ lysogens were grown and harvested, and the N5*‘o-methylene-H,-folate reductase (EC phage DNAs were isolated according to Zubay (24). 184.108.40.206), the product of the m&F gene, is Phage and plasmid DNAs were analyzed in 0.8-1.0% repressed when E. coli is grown in the agarose gels as described (21). Heteroduplex formation and analysis were performed as described presence of vitamin B1z (13). Other studies lengths were deterhave shown that this effect of vitamin B12 (25, 26). The double-stranded mined by comparison with the replicative form II requires the presence of the holoenzyme form of the Bzl-methyltransferase, i.e., of #X174, which was used as an internal standard. Construction of pJl9, a metIP plasmid The 9.1-kb with a bound cobalamin (12). EcoRI fragment bearing the m&E gene and present Recently, the Salmonella typhimurium on both the AmetE+and XmotE+-ll transducing metE gene has been cloned (19). The pres- phages [EcoRI D’; (21)] was cloned into plasmid ent report describes the construction of a pBR325 as follows. After restricting km&E+-2 and plasmid containing the E. coli m&E gene pBR325 with EcoRI, the fragments were ligated with T, DNA ligase. Strain JEN6 was transformed with and some characteristics of the expression the ligated DNA. JEN6 is a metE derivative of C600, of the gene in viva and in vitro. MATERIALS
Strains, plasmids, and phages E. coli strains RW262, RW592, and Mu24 (m&E), and phages X199 and h248 were obtained from R. Weisberg, National Institutes of Health (Bethesda, Md.); E. wZi strains MC4100, 2276 (metE, thi) and AT753 (wal’, Zeu, thi) were supplied by M. Casadaban, University of Chicago (Chicago, Ill.); A. L. Taylor, University of Colorado (Denver, Colo.); and T. Eckhardt, Smith Kline and French, Inc. (Philadelphia, Pa.); respectively. Phage XcI857Sam7 was supplied by E. McFall, New York University Medical School (New York, N. Y.). JEN6 (metE, va6: Zou, thi) was constructed by Pl(AT753) transduction (20) and was a recipient for plasmid DNA transformations. JEN7 (thi, m&E), derived from MC4100, was used to determine the transducing titer of XmetE+-2. To construct this strain, MC4100 was mutated by treatment with a temperature-sensitive mutant of bacteriophage Mu,
which was constructed by Pl transduction. No Met+ transformants appeared after several days of incubation on minimal agar, while Apr transformants appeared after only 24 h of incubation on Ap-containing glucose-tryptone-yeast extract agar. These Ap’ transformants were analyzed for Met+ phenotypes. From the 890 colonies screened, one Met+ was found. A purified colony of this transformant was examined using the Eckhardt (27) plasmid-screening gel and a 15.4-kb plasmid, designated pJ19, was found. A crude DNA lysate of JENG/pJlS was prepared and used to transform JEN6 with selection for Met+. All colonies picked were ampicillin resistant and Met+ phenotype. EcoRI digestion of this hybrid plasmid, pJ19, showed the presence of a 9.1-kb fragment identical in size to the EcoRI D’ fragment from Xmet+-2. An EcoRI and BamHI double digest of pJ19 confirmed that the cloned 9.1-kb metE+ fragment is identical to the EooRI D’ fragment from AmetE+-2. DNA from plasmid pJ19 was prepared from cleared
lysates of JENG/pJlS by two successive CsCl-ethidium bromide equilibrium centrifugations. This DNA was used to transform strains RR1 and 22’76, using a procedure previously described (28,29). pJ19 DNA was digested with EcoRI and the 9.1-kb fragment, which contains the 5.6 kb of E. coli genomic DNA and 3.5 kb of EDNA, was purified by preparative agarose gel electrophoresis. Grcwth of E. coli strains, assay and isolation of the metE protein. E. coli strains K12, 2276, and 2276/ pJ19 were grown at 37°C to the mid-log phase of growth (ODGoo = 0.7) in M-9 salts medium containing 10 pg/ml methionine and 0.4% glucose. The cells were harvested by centrifugation, washed in M-9 medium, and sonicated in a buffer containing 50 mM Tris-Cl, pH 7.6, 1 mM dithiothreitol (1 ml/g wet cells). The unbroken cells and cell debris were removed by centrifugation at 10,000~ for 10 min at 4°C. The supernatant was assayed for metE enzyme activity essentially as described by Whitfield et al. (9). w&E protein was purified by a modification of the procedure of Whitfield et aL (9) from E. coli B, grown in minimal M-9 salt medium containing 0.4% glucose. Pteroyl-y-glutamyl-y-glutamylglutamic acid (folate-Glua) was a kind gift of Dr. Leon Ellenbogen, Lederle Laboratories. m-&-Folate-Glus was prepared by reduction of the above compound with NaBHI (30) and DL-N5-[‘4C]-methyl-&-folate-Glus was synthesized as previously described (31) and had a specific activity of 6600 cpm/nmol. Protein concentrations were determined using the Bio-Rad Coomassie blue reagent, using bovine serum albumin as a standard. DNA-directed in vitro protein [email protected]
E. coli DNA-directed in vitro protein-synthesizing systems, using either highly purified E. coli factors (32, 33) or crude fractions consisting of a ribosomal salt wash and 0.25 M and 1.0 M salt eluates from a DEAE column fractionation of an S200 extract, were prepared as previously described (34). The incubations contained as template either 0.5 pg of pBR322 DNA, 2 pg of pJ19 DNA, or 1 pg of the purified 9.1-kb DNA insert from pJ19. Unless stated otherwise, the reactions were performed for 60 min at 37’C and [%S]methionine (20,000 cpm/pmol) was used as the label. Where appropriate, hot Cl&COOH-precipitable radioactivity was measured and aliquots of the incubations were applied to 12% NaDodSO1-polyacrylamide gels to separate the protein products (35). After electrophoresis, the gels were soaked in Enhance (New England Nuclear Corp.), dried, and fluorographed at -70°C. RESULTS
Density gradient XmetF-2 lysates.
and titers of
XmetE+ transducing phages were isolated by the Schrenk and
Weisberg procedure (23) as described under Materials and Methods. Twelve MetE+ colonies were purified by single-colony isolation and tested for their ability to yield high-frequency MetE+-transducing (Hft) lysates using Mu24 as a recipient. Four lysogens were isolated (21) but only XmetE+-2 is characterized here. Fractions from a cesium chloride gradient of lysates from AmetE+- were analyzed for transducing and lytic titers and Azso. This phage preparation showed a single AZ60 peak but distinct lytic and transducing peaks. However, the number of transducing particles greatly exceeded the number of lytic particles, 6 X 10’ transducing particles and 5.6 X lo7 lytic particles. These data indicate that the XmetE+-2 phage is defective. While the lytic titers are an accurate representation of the actual number of lytic phage particles present, the transducing titers are generally 10’ to lo3 times lower than the actual number of transducing particles (36). Therefore, the XmetE+-2 lysate has between lo* to lo5 more transducing than lytic particles. Because of this ratio, it was not necessary to further purify the transducing particles for DNA characterization. Heteroduplex analysis of AmetE+-2. An estimate of the lengths of the chromosomal insertion of the hmetE+-2 phage was done by heteroduplex analysis between XmetE+-2 and X (Fig. 1). The heteroduplex contained a double-stranded region equal in length to the right arm of X (19.5 kb), a single-stranded region of nonhomology representing the chromosomal substitution and, at the terminus of the left arm, a homologous region which was 3.4 kb. Since length measurements of single-stranded DNA are unreliable under these conditions, the length of the chromosomal substitution was determined by subtracting the total length of homologous DNA in the heteroduplex from the total measured length of X phage DNA. The value for the chromosomal substitution was 22.1 kb for XmetE+-2. Restriction endonuclease analysis of XmetEt-2. Restriction endonuclease anal-
CHU ET AL.
FIG. 1. Heteroduplex analyses between X and M&E+-2. (A) Electron micrograph of a heteroduplex between h and XmetE+-2 DNA. The replicative form II of 6x174 DNA, 5.38 kb in length, was included as an internal standard. (B) Tracing of the electron micrograph in (A). Double-stranded DNA is represented as a heavy line, while single-stranded DNA is represented as a light line. L and R indicate the left and right ends of XDNA. Arrows 1 and 2 indicate the junction points between the double- and single-stranded DNA. Arrow 2 also indicates the chromosomal insertion at the Xatt+ site.
by showing that in XmetE+-Z the chromosomal substitution is in the left arm. AmetE+contained all of the EcoRI, BarnHI, and Hind111 restriction sites (3’739) to the right of the hatt+ site at 57.4 map units. In addition, the BumHI junction at 58.1 map units was also present,
whereas the Hind111 site at 57.0 map units was absent. Thus, as would be expetted from the Campbell model (40), the substitution is adjacent to and to the left of Xatt+. From further restriction analyses a restriction map of the XmetE+-2-transducing phage was deduced (Fig. 2).
9.1 kb & 10 map unit3 (0.456kbl
FIG. 2. Partial restriction map of XmetE+-2-transducing phage. (A) XmetE+-2 DNA, and (B) an enlargement of the EcoRI to Hind111 region, which contains the 9.1-kb fragment that was subcloned in plasmid pJ19.
In vivo studies with
strain 2276 and Strain 2276, a stable metE mutant, was transformed with plasmid pJ19. In methionine-free medium, strain 2276 did not grow whereas strains 2276/ pJ19 and K12 both had generation times of about 60 min. The 2276/pJ19 transformant acquired a MetE+ phenotype and cell-free extracts had a metE enzyme specific activity several-fold greater than wild-type E. coli K12 (Table I). Since the metE protein is thought to comprise 35% of the total soluble protein in wildtype E. coli (9), one could expect that the metE protein would be a major component in the 2276/pJ19 extracts. Aliquots of extracts from strains 2276 and 2276/pJ19 were subjected to electrophoresis on NaDodSO*-polyacrylamide gels, along with purified metE protein (Fig. 3). It can be 2276/pJ19.
E. coli STRAINS Strain
K12 (MetE*) 2276 (MetE-) 2276/pJ19 (MetE+)
550 co.5 3200
Note. Growth medium contained 10 pg/ml methionine and extracts were assayed for metE enzyme activity as described in the text. Specific activity is defined as nmol methionine formed mg protein-’ 15 min-‘.
2276 t WRIFIEDMT EMET E PROTEIN
FIG. 3. Plasmid directed in viva synthesis of metE gene product. Purified metE protein (2.5 cg) and aliquots of the extracts of 2276 and 2276/pJ19 (see Table I) containing 2.5 pg protein were subjected to electrophoresis on 12% NaDodSO1-polyacrylamide gels and stained with Coomassie blue. The position of m&E protein is marked with an arrow.
clearly seen that in extracts of 2276/pJ19 the metE protein (&f, - 90,000) is a major protein component and that this protein is essentially lacking in the 2276 extract.
In vitro synthesis of the metE
frowt pJ19. Crude and highly defined in vitro protein-synthesis systems (see Materials and Methods) were used to study the expression of the metE gene from plasmid pJ19. A typical result showing the products synthesized in vitro using crude fractions is seen in Fig. 4. Plasmid pJ19 directs the synthesis of four major proteins: a fusion product from the chloramphenicol transacetylase gene (Mr 28,000), /3-lactamase (M, 32,000), and two additional proteins of approximately M, 34,000 and 90,000. When the purified 9.1-kb DNA insert of pJ19 is used as template, only the M, 34,000 and 90,000 bands are seen (Fig. 4). Since the M, 90,000 protein formed in vitro has the same molecular weight as the purified metE protein, it very likely is the m&E gene product. However, the amount of synthesis of this protein in vitro was quite low when compared to the high expression of the metE gene in vivo.
FIG. 4. In vitro synthesis of metE gene product from intact plasmid pJ19 and from the purified 9.1kh insert contained in pJ19. For these experiments, the in vitro system contained crude fractions [(34) see Materials and Methods] and either 2 pg pJ19, 1 fig of the 9.1-kb insert, or 0.5 fig of pBR322 DNA as templates. Aliquots of these incubations containing 100,000 counts of CCl&OOH-precipitable protein were subjected to electrophoresis on 12% NaDodSOdpolyacrylamide gels and fluorography. The position of the presumed m&E protein is indicated by an arrow. DISCUSSION
A 9.1-kb fragment bearing the E. coli K12 metE gene was cloned by first isolating X-transducing phages carrying the metE gene and then cloning the metEbearing fragments into plasmid pBR325. The first step in the purification of the metE gene was the isolation of the transducing phage, hmetE+-2. The forced integration of X199 into the secondary attachment sites (23) generates a heterogenous lysogen, from which XmetE+-transducing phages could be isolated by abberant excision. In this study, one of the lysogens, XmetE+-2, was further characterized. Usually the ratio between lytic and transducing particles in lysate preparations from these double lysogens is approximately 1. However, heat induction of the Mu24 XmetE+-2 gave rise to lysates which contained approximately lo5 more transducing particles than lytic particles. This unexpected phenomenon has been observed previously (41) and is probably
due to the mechanism by which the double lysogens are formed. A possible explanation is that during the lysogenization step in the formation of Hft lysates both the transducing and lytic phages were integrated into the same site end to end apd in a tandem insertion. Upon heat-induced excision the transducing and helper phages undergo a single recombination event, such that a “similar” transducing phage was regenerated by ret-mediated homologous recombination by the host. Thus, the helper phage was lost. This could account for the unequal distribution of lytic and transducing particles. Restriction mapping and heteroduplex analysis of AmetE+- clearly shows that the chromosomal ‘substitution is in the left arm of X starting from the attachment site. The observations are in agreement with the Campbell model (40) of X integration, excision, and the generation of transducing particles by aberrant excision. A 9.1-kb EcoRI fragment present in hmetE+-2 phage, containing a 5.6-kb fragment of E. coli chromosomal DNA, was cloned into pBR325. This 5.6-kb fragment is large enough to include the metE gene, which we estimate to be no larger than 2.7 kb based on a molecular weight of -90,000. Although the metE gene was successfully cloned into pBR325, the transformation frequency was far lower than expected from XmetE+-2 which contains only seven EcoRI fragments. Only 1 out of about 1000 recombinant plasmids selected for ampicillin resistance on rich medium carried the metE gene and no MetE+ plasmids were found when selection was made solely for Met+. This low frequency may be due to the overproduction of the metE protein [pBR325 has a copy number of about 20 (42)], so that under derepressed conditions the high levels of the metE protein impair the growth of the organism. As expected, the 2276/pJ19 transformants had a very high metE enzyme specific activity and the metE protein was the major soluble protein in these cells. As can be seen in Fig. 3, the level of metE protein is much higher than that of EFTu (the major band at about M, 44,000),
which is unchanged in the two extracts and has been reported to comprise up to 5% of the total soluble protein (43, 44). When present as a single copy in wildtype E. coli, the m&E gene expression is repressed by greater than 90% by either methionine or vitamin Bi2 (13, 14). However, preliminary results indicate that the metE gene expression in the 2276/pJ19 transformant is only partially repressed (-40%) by either of these compounds. This lack of complete repression of the expression of the metE gene in 2276/pJ19 may be related in some way to the increased copy number of metE genes per cell or could be a result of some unknown modification of the regulatory region of this gene. Another possibility is that the metJ protein becomes limiting because of the high copy number of metE genes present in 2276/pJ19. However, this seemsunlikely since, under the same conditions and in the same extracts, methionine was able to repress (>90%) the expression of the metF gene (data not shown). Although expression of the metE gene was observed in vitro using pJ19, there was a striking difference in the degree of expression observed in vitro compared to in vivo. As noted, strain 2276/pJ19 produces large amounts of the metE protein, yet in vitro studies, using pJ19 as template in a highly defined DNA-directed system, showed only a low synthesis of a protein of about M, 90,000 (presumably the metE protein). Other genes carried on this plasmid were expressed very efficiently, which indicated that the plasmid template was functional and the in vitro system was active. To test the possibility that an essential factor for metE gene expression was missing in the defined system, the purified factors in the in vitro system were replaced with crude fractions. Once again there was only a low synthesis of a M, 90,000 protein. This low expression of the metE gene was not due to the presence of an active m&J or vitamin Blz repression system, since the metF gene, which is also regulated by these components, is efficiently expressed in these in vitro systems (data not shown). It is also possible that
the poor expression of the metE gene in vitro may be due to promoter competition with the chloramphenicol acetyltransferase and P-lactamase promoters which are contained on the plasmid. Recently, Schulte et al. (19) have cloned the metE gene from S. typhimurium and have shown that when a minicell-producing strain was transformed with a plasmid containing the w&E gene, a protein of about M, 92,500 was synthesized. The present results, using a cloned E. coli metE gene, indicate that the metE protein in E. coli is very similar in molecular weight. The construction and expression of an E. coli plasmid that contains the metE gene will permit further studies on the regulation of the expression of this gene by methionine and vitamin B12. ACKNOWLEDGMENTS The authors would like to dedicate this manuscript to Dr. Bernard Horecker on the celebration of his 70th birthday. One of us (H.W.) is especially pleased to have this opportunity to express best wishes to a friend, teacher, and colleague for more than 30 years. Part of this work was supported by NIH Grant MG25319 (awarded to N.K.) and was taken from a Ph.D. dissertation submitted to the Department of Basic Medical Sciences of NYU (J.C.). N.K. was supported by an Irma T. Hirsch1 Career Scientist Award. We thank Dr. W. Maas and Dr. E. McFall for the kind use of their equipment. J.H. was the recipient of a 1934 Hoffmann-La Roche Biological Science Teacher Grant. REFERENCES 1. WEISSBACH,
mins 2. TAYLOR,
zymes 3. WOODS,
and Hormones R. T., AND 19, 121-165. D.
R. T. (1970)
28, 415-440. WEISSBACH,
(1965) in Transmethylation and Methionine Biosynthesis (Shapiro, S. K., and Schlenck, F., eds.), pp. 138-154, Univ. of Chicago Press, Chicago. 4. TAYLOR, Chem
R. T., AND WEISSBACH, 242, 1502-1508.
R. T., AND
Chem, 6. TAYLOR,
242, 1509-1516. R. T., AND
7. TAYLOR, R. T., AND WEISSBACH, H. (1969) Arch B&hem. Biophys. 129,728-744. 8. TAYLOR, R. T., AND WEISSBACH, H. (1969) Arch. Biochem, Biophys. 129, 745-766. 9. WHITFIELD, C. D., STEERS, E., JR., AND WEISSBACH, H. (1970) J. BioL Chem 245,390-401. 10. WHITFIELD, C. D., AND WEISSBACH, H. (1970) J. Biol. Chem. 245, 402-409. 11. WEISSBACH, H., REDFIELD, B., AND DICKERMAN, H. (1964) J. Biol. Chem. 239, 146-148. 12. MILNER, L., WHITFIELD, C., AND WEISSBACH, H. (1969). Arch Biochem Biophys. 133, 413-419. 13. KUNG, H. F., SPEARS, C., GREENE, R. C., AND WEISSBACH, H. (1972) Arch. B&hem Biophys. 150, 23-31. 14. GREENE, R. C., WILLIAMS, R. D., KUNG, H. F., SPEARS, C., AND WEISSBACH, H. (1973) Arch B&hem Biophys. 158,249-256. 15. HOLLOWAY, C. T., GREENE, R. C., AND Su, C.-H. (1970) J. Bucterid 104, 734-747. 16. Su, C.-H., AND GREENE, R. C. (1971) Proc. Natl. Acad Sk USA 68, 367-371. 17. GREENE, R. C., Su, C.-H., AND HOLLOWAY, C. T. (1970) B&hem. Biophys. Res. Commun. 38, 1120-1126. 18. GREENE, R. C., HUNTER, J. S. V., AND COCH, E. H. (1973) J. Bacterial. 115, 57-67. 19. SCHULTE, L. L., STAUFFER, L. T., AND STAUFFER, G. V. (1984) J. Baxterid 158, 928-933. 20. MILLER, J. (1972) in Experiments in Genetics, pp. 431-435, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 21. CHU, J. (1980) Isolation and Characterization of the MetE+ Gene in Escherichia coli, Ph.D. Dissertation, New York University, New York. 22. LEISINGER, T., HAAS, D., AND KELKER, N. (1976) J. Bacterial 125, 1217-1219. 23. SCHRENK, W. J., AND WEISBERG, R. A. (1975) Mol. Gen Genet. 137, 101-108. 24. ZUBAY, G. (1973) in Annual Reviews in Genetics (Roman, H., Sandler, L., and Campbell, A. M., eds.), Vol. 7, pp. 267-287, Annual Reviews, Palo Alto, California. 25. DAVIS, R. W., SIMON, M., AND DAVIDSON,
27. 28. 29.
30. 31. 32.
34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
N. (1971) in Methods in Enzymology (Grossman, L., and Moldave, K., eds.), Vol. 21, pp. 413-428, Academic Press, New York. MAZAITIS, A. J., PALCHAUDHURI, S., GLANSDORFF, N., AND MAAS, W. K. (1976) Md Gen Genet. 143, 185-196. ECKHARDT, T. (1978) Plusmid 1, 584-588. NORGARD, M. V., KEEM, K., AND MONAHAN, J. J. (1978) Gene 3, 279-292. VILLA-K• MAROPFF,L., EFSTRATIADIS, A., BROOME, S., LOMEDICO, P., TIZARD, R., NABER, S. P., CHICK, W. L., AND GILBERT, W. (1978) Proc Natl Acad Sci. USA 75, 3727-3731. HILLCOAT, B. L., AND BLAKLEY, R. L. (1964) B&hem Biophys. Res. Commun. 15,303-307. KERESZTESY, J. C., AND DONALDSON, K. D. (1961) Biochem. Biophys. Res. Commun 5, 286-288. KUNG, H.-F., REDFIELD, B., TREADWELL, B. V., ESKIN, B., SPEARS, C., AND WEISSBACH, H. (1977) J. Biol. Chem. 252, 6889-6894. ZARUCKI-SCHULZ, T., JEREZ, C., GOLDBERG, G., KUNG, H. F., HUANG, K.-H., BROT, N., AND WEISSBACH, H. (1979) Proc. Nutl Acad Sci USA 76,6115-6119. KUNG, H. F., SPEARS, C., AND WEISSBACH, H. (1974) .I Biol. Chem. 250, 1556-1562. LAEMMLI, U. K. (1970) Nature (London) 227,680685. SIGNER, E. R. (1969) Nature (London) 223, 158160. THOMAS, M., AND DAVIS, R. (1975) J. Mol. BioL 91, 315-328. HAGGERTY, D., AND SCHLEIF, R. (1976) J. Virol. 18, 659-663. ROBINSON, L. H., AND LANDY, A. (1977) Gene 2, 1-31. CAMPBELL, A. M. (1962) Adv. Genet. 11, 101-145. PALCHAUDHURI, S., MCFALL, E., AND CAROTHERS, A. M. (1976) J. Bucteriol 127, 998-1014. BOLIVAR, F. (1978) Gene 4, 121-136. FURANO, A. V., (1975) Proc. Natl. Acad Sci. USA 72, 4780-4784. JACOBSON, G. R., AND ROSENBUSCH, J. P. (1976) Nature (London) 261, 23-26.