Cloning and Insertional Inactivation of the dye - Journal of Bacteriology

Vol. 154, No. 3

JOURNAL OF BACTERIOLOGY, June 1983, p. 1309-1314

0021-9193/83/061309-06$02.00/0 Copyright C 1983, American Society for Microbiology

Cloning and Insertional Inactivation of the dye (sfrA) Gene, Mutation of Which Affects Sex Factor F Expression and Dye Sensitivity of Escherichia coli K-12 ROGER S. BUXTON* AND LUCY S. DRURY Division of Microbiology, National Institute for Medical Research, London NW7 JAA, England Received 24 January 1983/Accepted 29 March 1983

Deletions of the Escherichia coli K-12 chromosome between trpR and thr render the bacterium sensitive to the dye toluidine blue (Dye-), and if male (Hfr or F'), the strain is sterile (Fex-), failing to donate F' or chromosomal markers and resistant to male-specific phages as a consequence of its inability to elaborate F pili. A 6-kilobase Sail fragment of E. coli chromosomal DNA cloned into the plasmid pBR322 has been shown to complement both the Dye- and Fexphenotypes. Insertion of the transposon yb (TnlO00) into a specific part of this plasmid invariably results in both the Dye- and Fex- phenotypes, indicating that these phenotypes derive from mutation in a single gene. Complementation tests between such insertions and sfrA4, a previously isolated mutation resulting in a Fex- phenotype and reported to code for a transcriptional control factor for F (L. Beutin, P. A. Manning, M. Achtman, and N. Willetts, J. Bacteriol. 145:840844, 1981), indicated that dye and sfrA4 were mutations in a single cistron. It is proposed that the dye (sfrA) gene product is necessary not only for efficient transcription of the F factor genes, but also for some component(s) of the bacterial envelope, loss of which results in sensitivity to toluidine blue.

Conjugal transfer of DNA in Escherichia coli is dependent upon the presence in the donor cell of a sex factor (14, 17), and for the sex factor F a number of tra genes have been identified coding for F pilus synthesis, mating aggregation, and DNA transfer (for review, see ref. 29). Nevertheless, the expression of such donor properties has been shown to be strongly dependent upon the physiological state of the donor cell (7, 8, 24), and hence, cellular functions and cell components determined by chromosomal genes are also likely to be important in conjugation. Since the F factor is a dispensable genetic element, it seems likely that it utilizes chromosomal genes which also function in the general physiology of the bacterium. The isolation of chromosomal mutations resulting in male sterility could therefore be a way of identifying new cellular functions. A number of such chromosomal mutations which result in male sterility but do not affect the inheritance of the sex factor have been identified (1, 4, 18, 20). We have previously isolated deletions at 99 to 100 min on the E. coli genetic map, extending from the deo genes to thr, which confer such a Fex- (F expressionless) phenotype. These deletions define a gene lying between trpR and thr which we termed msp for

male-specific phage resistance since we identified this gene by this mutant phenotype (4). In Hfr or F' strains, this deletion results in the loss of the ability to transfer genetic markers. Subsequently, others have isolated mutants with similar properties, probably deriving from point mutations, which also map in the same region of the chromosome, variously termed fex, for F expression (18), sfrA, for sex-factor regulation (1), and cpxC, for conjugative plasmid expression (20), by methods specifically designed to isolate male-sterile mutants. In this paper we use the mnemonic Fex. Similar deo-thr deletions were also later isolated by Roeder and Somerville (25). These authors defined the position of a gene, dye, important for cellular resistance to methylene blue and toluidine blue, suggesting an alteration in the permeability properties of the cell consequent upon changes in the cell envelope. Since the Dye- and Fex- phenotypes do not represent readily selectable unique phenotypes, we have studied the relationship between these phenotypes after first identifying a plasmid clone carrying these genes and then inserting the transposon yb (TnJO00) into the cloned genes. This is particularly easy because insertions of -yb are readily isolated (12) and occur in any particular

1309

1310

BUXTON AND DRURY

Strain RB400

Mating type HfrH

RB85

F-

RB978

F-

RB979

F-

RP166

F+

52R963 ED3867 ED3868 9-9

F' pro lac(::TnS) FFF-

J. BACTERIOL.

TABLE 1. E. coli K-12 strains used Genotype Origin/reference thi tyrT+ A(gal-anX-bio-uvrB-deoR) (4) upp deoD::(A c1857 S7) thr leu thi lac Y rpsL tonA supE44 (5) (A) leu thi lacY rpsL tonA supE44 RB400 x RB85 Thr+ Srr deoD::(A c1857 S7) leu thi lacY rpsL tonA supE44 Heat-resistant derivative of RB978 A (deoD-thr) thr leu thi thyA deoC lacY R. H. Pritchard via E. F. Tresguerres. (28). H. R. Smith/(16) lac pro trp his Nalr N. Willetts/(1) thi trp lys gal malA rpsL tsx sup As ED3867, sfrA4 N. Willetts/(1) R. F. Rosenberger/CSH 9 (21) made thi trp lacZ rpsE Spr by Pl transduction

gene relatively frequently (e.g., ref. 26). In this report, dye and sfrA are shown to be almost certainly mutations in the same gene. MATERIALS AND METHODS Bacterial strais and plasmids. The bacterial strains, all derivatives of E. coli K-12, are listed in Table 1. The plasmids used were pRPG1, which is pACYC184 carrying a 19-kilobase (kb) BamHI fragment close to and including thr (11), kindly donated by R. P. Gunsalus; pBR322, kindly donated by G. T. Yarranton; and pRB1, which is pBR322 with a 6-kb SalI insert from pRPGl, described in the text. Media. Complex medium was L broth (LB) containing, per liter of distilled water, tryptone (Difco Laboratories, Detroit, Mich.), 10 g; yeast extract (Difco), 5 g; and NaCl, 10 g. Toluidine blue agar was L agar containing 200 ,ug of toluidine blue per ml (25). Antibiotics were used at the following concentrations (in micrograms per milliliter): ampicillin, 20; chloramphenicol, 25; kanamycin, 20; spectinomycin, 100; streptomycin, 200; and tetracycline, 25. Test for dye resistance. Dye resistance was tested by streaking bacteria suspended in phosphate buffer onto toluidine blue agar and incubating overnight at 35°C. Bacterial matings. Bacterial matings were performed on exponentially growing cells. For the quantitative determination of donor ability (see Tables 2 and 3), the following procedure was used: donor cells were grown overnight at 35°C in LB + kanamycin to retain the F', F' pro lac(::Tn5); the recipient strain (9-9, rpsE) was grown in LB. These cultures were diluted 1/20 into LB and grown at 37°C for approximately 2 h. Matings were performed by mixing 0.1 ml of F- and 0.9 ml of F' in small tubes for 15 min at 37°C. These were then diluted 1/10 into LB + spectinomycin and were shaken at 37°C for 1 h. Samples (0.1 ml) were plated out for Kmr Spr recombinants. The numbers of F'-carrying donor cells were determined by a viable count on L agar + kanamycin, or L agar + kanamycin + ampicillin for strains carrying plasmid pBR322 and derivatives. Rapid mating tests were either performed in small tubes followed by streaking out on selective plates, or by replicating mating mixtures from Bertani dishes

(Elesa microculture containers, Elesa SPA, Milan, Italy) with a replicator constructed from nails. Restriction endonuclease digestion and ligation of DNA fragments. Endonuclease digestions were performed in NM buffer (10 mM Tris-hydrochloride, pH 7.5, 8 mM MgCI2, -0.01% gelatin, 1 mM DL-dithiothreitol) at 37°C with added NaCl as required (9). Restriction endonucleases were mainly purchased from New England Biolabs, Beverly, Mass., and Bethesda Research Laboratories, Inc., Gaithersburg, Md. Sall was a gift of L. Harper. T4 polynucleotide ligase was from P-L Biochemicals, Inc., Milwaukee, Wis., kindly donated by G. T. Yarranton. Ligation was carried out at 15°C in 66 mM Tris-hydrochloride (pH 7.6)-6.6 mM MgCl2-10 mM dithioerythritol-0.4 mM ATP (pH 7.0). Restrictions by Sall were terminated by phenol extraction and ethanol precipitation before ligation. Gel electrophoresis. Analysis of restriction fragments was carried out on horizontal 0.7% agarose gels (19) in 89 mM Tris-89 mM borate-2.5 mM Na2EDTA, pH 8.3, for 16 h at approximately 1.5 V/cm. Samples of approximately 1 ,ug of DNA in 30 1.l were used. Gels were stained with ethidium bromide (1 ,ug/ml) for 1 h, and DNA bands were identified by UV fluorescence. EcoRI- and EcoRI-BamHI-generated fragments of lambda DNA, kindly donated by P. J. Piggot and M. G. Sargent,- were used as molecular weight standards. Plasmid DNA extraction. Plasmid DNA was extracted as described by Guerry et al. (10) after amplification with chloramphenicol (100 ,ug/ml) for pBR322-derived plasmids or with spectinomycin (300 ,ug/ml) for pACYC184-derived plasmids (6), followed by ethidium bromide-cesium chloride isopycnic density gradient centrifugation to remove chromosomal DNA. Transformation. Bacteria were transformed as described by Brown et al. (3). F pilus visualization by electron microscopy. Exponentially growing bacteria (1 ml) were mixed with approximately 1011 R17 phage and incubated for 15 min at 37°C. The addition of RNA phage prevents F pilus retraction (23). Cells were negatively stained with 1% phosphotungstic acid, pH 7.0, and were examined with an electron microscope by I. D. J. Burdett and J. Whalley.

VOL. 154, 1983

TABLE 2. Tests for donor ability of strains with deletions of dye' F' pro lac::Tn5 donor

genotype

Kmr Spr recombinants per 100 donor cells with other plasmid carried: pRB1 pRB9b pRB35c None pBR322 dye+ dye(::-yb) dye'

dye+ 19 29 92 14 38 Adye 0.00017 0.000032 41 0.00013 18 a Matings were performed as described in the text. The recipient strain used was 9-9 (rpsE). Donors used were F' pro lac::TnS derivatives of the dye+ strain RB85 and the Adye strain RB979. b pRB9 is the plasmid from strain RB2032 (see Fig. 1). c

pRB35 has yB inserted elsewhere in pRB1.

RESULTS Isolation of deo-thr deletions. We have previously inserted plasmid X cI857 S7 into the deo operon of E. coli, mapping at 99 min (4). The rare heat-resistant survivors from a A cI857 S7 lysogen plated at 41°C often derive from cells in which the phage genes involved in the lethal functions have been deleted (22, 27), and the deletion may extend into an adjacent bacterial region (27). By isolating such heat-resistant derivatives, we noticed that some, although Hfr, failed to donate markers to an F- strain, and we have shown (4) that a gene, msp, which determines resistance to the male-specific phages R17 and ,2, lies between trpR and thr. Male strains with this deletion failed to make F pili. Because of its phenotype and map position, msp seemed likely to be the same as fex, sfrA, and cpxC. Since we had difficulty determining resistance to male-specific phages in certain strains, we found it operationally easier to have this deletion in an F- strain and to test for male sterility by transfer of an F'. We therefore isolated a new deo-thr deletion in an F- strain, RB85 (thr rpsL), by crossing in the deoD::X cI857 57 from the HfrH strain RB400, selecting for Thr+ Str recombinants, and isolating heat-resistant revertants of the resulting recombinant. This strain, RB979, had apparently lost a continuous linear segment of DNA from deoD to thr, ipcluding serB, resulting in serine and threonine requirements, and was sensitive to the methylene blue analog toluidine blue (200 jig/ml) (Dye-). When F' pro lac(: :TnS) was introduced by selecting for kanamycin resistance, RB979 was unable to donate this to F' strains and hence was Fex- (Table 2), although the F factor was apparently replicated normally. Subcloning of the dye gene. Gunsalus et al. (11) have isolated a X transducing phage carrying the trpR-thr region of the chromosome, and from such a phage they subcloned a 19-kb BamHI

E. COLI dye (sfrA) GENE

1311

fragment carrying thr+ but not trpR+ into the plasmid pACYC184. With this plasmid pRPG1 as starting material, a 6-kb SalI fragment carrying the dye" gene has been subcloned into the tet gene of plasmid pBR322 by ligating a SalI digest of pRPG1 with a Sall digest of pBR322 and transforming into strain RB979, selecting for ampicillin resistance, and scoring for Dye resistance. The resulting plasmid, carrying a 6-kb SalI insert, was named pRB1. It was Tcs. Unlike strain RB979, when F' pro lac(::TnS) was introduced, the pRB1-containing strain had F pili and transferred the F' to F- strains at normal frequencies, thus being Fex+ (Table 2). The dye andfex mutations are therefore recessive to the wild-type alleles. Since the deletion in strain RB979 extends from deo to thr and since the 6-kb fragment lies within that region (it does not carry thr+), there should, therefore, be no homology between the cloned DNA fragment and the chromosome and no problems of recombination between cloned DNA and the chromosome. Restriction analysis of this plasmid confirmed the results of Gunsalus et al. (11) that there were to BamHI, SalI, or EcoRI restriction sites within the 6-kb Sall fragment. However, unlike Gunsalus et al. (11), we did find a HindIII site very close (0.1 kb) to one Sall site (Fig. 1). We presume this is the HindIlI site which these authors placed just outside the 6-kb Sall fragment, on the trpR side. If so, this would orient the 6-kb fragment on the chromosome as shown in Fig. 2. There were also two KpnI sites within the 6-kb SalI fragment.

Kpnl

SalI

FIG. 1. Restriction map of plasmid pRB1. Numbers are sizes in kb. RB strain numbers of y& insertions in dye are shown. Those inside the circle are of the opposite orientation to those shown outside. Note that yb is not shown to scale. It is 5.7 kb long.

1312

BUXTON AND DRURY SalI Ijidil

trpR

J. BACTERIOL. Sal

I

4

0.1

2-8

I

KPnI thr

27

04

FIG. 2. Orientation of the 6-kb Sall fragment on the chromosome of E. coli, based on the results in the text and the results of Gunsalus et al. (11). Numbers show sizes in kb.

Insertion of 'y8 into the dye gene. F-mediated transfer plasmid of pBR322 results in the insertion of the -yb sequence of F (Tnl000) into the plasmid at random sites (12). This technique was used to inactivate the dye gene as follows. The F+ strain RP166 was transformed with plasmid pRB1 (pBR322 plus the 6-kb Sall insert). This strain was mated for 90 min with the F- strain RB979 [A(deo-thr)], selection then being made for Apr Str recombinants. Ten separate matings were performed to isolate independent yb insertions. The recombinants were purified and then tested for Dye resistance on plates containing ampicillin to retain the plasmid and were tested for Fex by another mating with the F- strain 9-9 (rpsE), selection being made for Apr Spr recombinants. Of 400 recombinants tested, 14 were Dyes, and all of these were Fex- as judged by this test. No Fex- colonies were Dyer. The Fex phenotype was subsequently confirmed by transforming these -yb-derived plasmids into strain RB979 containing F' pro lac(::TnS) and testing for transfer of the F'. Typical results for one recombinant plasmid, pRB9, are given in Table 2. Assuming that yb insertion takes place at random (12), we conclude from this result that Dye' and Fex- are phenotypes both derived from mutation in a single gene. If, for example, dye was transcribed before fex from a single promoter so that insertions of -yb into dye were alsofex, we would expect to obtain at least some dye' fex insertions. None of these was found. Both the Dye- and Fex- phenotypes were easy to score, and we do not, therefore, think that we are overlooking mutants with only one phenotype.

One of these plasmids with a yb insertion into dye was used in a complementation test with the SfrA- mutant ED3868 (sfrA4), isolated by Beutin and Achtman (1). Strains ED3867 (sfrA+) and ED3868 (sfrA4) were each mated with the F' strain 52R963, carrying F' pro lac::TnS, selection being made for Kmr Str recombinants. The resulting F' strains were then transformed with the plasmids pBR322, pRB1, pRB9 carrying a -yb insertion in dye, and pRB35 with yb in another part of the plasmid. The F' strains were then mated with the F- strain 9-9 to test for the Fex

phenotype (Table 3). From the results it can be seen that pRB9 with yb in dye did not complement the sfrA4 mutation, whereas both pRB35 and pRB1 did complement it to some extent. We therefore conclude that dye and sfrA4 are mutations in the same cistron. Beutin and Achtman (1) have shown that all the sfrA mutations isolated (six) belong to one cistron. Strains with the sfrA4 mutation and other strains with sfrA mutations appear to be quite leaky with regard to their Fex phenotypes compared with strains having the A(deo-thr) deletions (Tables 2 and 3). Presumably, this reflects the differences between (probable) missense mutations and deletions. We have, however, tested strain ED3868 (sfrA4) for sensitivity to toluidine blue; compared with ED3867 (sfrA+), it formed slightly smaller colonies on standard toluidine blue plates. The sites of insertion of -yb into dye (sfrA) were determined by making use of the restriction sites within the yb sequence itself (12). The sites of insertion in four of these plasmids are shown in Fig. 1. These defined a small region of the plasmid where dye (sfrA) maps. We have subsequently mapped an additional 12 different sites of insertion in plasmids from the mating described above and from other matings, and in all cases the sites of insertion were in the same small region of the plasmid (data not shown). Removal of the DNA between the two KpnI sites did not affect the Dye and Fex phenotypes. DISCUSSION The results presented in this paper provide evidence that mutation of the dye gene, mapping between trpR and thr, by insertional inactivation with the transposon yb (Tnl000), causes the bacterium to be sensitive to toluidine blue and also causes male sterility. By isolating a number of such mutants, we have tried to eliminate the possibility that insertion of yb in dye is having a polar effect on a gene involved in F plasmid expression, or vice versa. TABLE 3. Complementation test between dye(::,yb) and sfrA4a F' pro

lac::TnS donor genotype

Kmr spr recombinants per 100

donor cells with other plasmid carried: None

pBR322

pRB1

dye'

pRB9

dye(::pyB)

pRB35

dye 19 sfrA + 40 100 22 79 sfrA4 1.2 0.38 25 0.48 42 a Matings were performed as described in the text. The recipient strain used was 9-9 (rpsE). The sfrA+ strain was an F' pro lac::TnS derivative of ED3867; the sfrA4 strain was an F' pro lac::TnS derivative of ED3868.

VOL. 154, 1983

A complementation test with the previously isolated SfrA- male-sterile mutant carrying the sfrA4 mutation revealed that dye and sfrA lie in the same cistron. The other male-sterile mutations (fex, cpxC), defined by point mutations and mapping close to thr, are probably also mutant alleles of the same gene. The reason why sfrAcarrying strains are not Dye sensitive is likely to be that they carry leaky missense mutations. The strain carrying sfrA4 was in fact very slightly sensitive to toluidine blue. Strains carrying the deo-thr deletion are considerably more Fexthan those with the sfrA4 mutation (Tables 2 and

3).

Mutation of one other gene, seg, mapping between serB and thr, is also known to affect properties of the F factor, resulting in temperature-sensitive replication of the F factor (13, 15). This gene must be deleted in strain RB979, but apparently F factor replication is unaffected in strains carrying the deo-thr deletion. seg-2 has been shown to be located to the left of trpR (P. L. Bergquist, personal communication) and therefore is not carried on the 6-kb Sall fragment. Experiments by Beutin et al. (2) have shown that the sfrA gene product is necessary for efficient transcription of the F factor control gene traJ, and, directly or indirectly, for transcription of the tra Y -+ Z operon. Transcription of traJ was reduced about threefold in the sfrA mutants relative to sfr+ cells. It is probable that this reduction is more dramatic in cells with deletions of sfrA. In view of the fact that mutation of dye (sfrA) causes these two very different phenotypes Dye- and Fex-, it is proposed that the dye (sfrA) gene product is involved in regulation of some component(s) of the cell envelope, a change in which causes sensitivity to dyes such as toluidine blue, and that the F factor makes use of the dye gene product to regulate transcription of the F factor tra genes. Since dye can be deleted, it does not appear to be essential for the bacterium. The isolation and study of chromosomal mutants affecting conjugation may therefore be a useful way of probing the regulation of components of the cell envelope. The phenotype of these mutants emphasizes the way in which the physiology of the F factor is bound up with components of the envelope. Lastly, this report provides an example of how important phenotypic traits may be absent in leaky missense mutants when compared with deletion or insertion mutants. ACKNOWLEDGMENTS We thank Steve Sedgwick for introducing us to yb, and Ian Burdett and Jane Whalley for electron microscopic examina-

E. COLI dye (sfrA) GENE

1313

tion of F pili. We are also very grateful to Robert Gunsalus for donating plasmid pRPG1 DNA and to Neil Willetts for the Sfr- mutant. LITERATURE CITED 1. Beutin, L., and M. Achtman. 1979. Two Escherichia coli chromosomal cistrons, sfrA and sfrB, which are needed for expression of F factor tra functions. J. Bacteriol. 139:730-737. 2. Beutin, L., P. A. Manning, M. Achtman, and N. WUlletts. 1981. sfrA and sfrB products of Escherichia coli K-12 are transcriptional control factors. J. Bacteriol. 145:840-844. 3. Brown, M. G., A. Weston, J. R. Saunders, and G. 0. Humphreys. 1979. Transformation of Escherichia coli C600 by plasmid DNA at different phases of growth. FEMS Microbiol. Lett. 5:219-222. 4. Buxton, R. S., K. Hammer-Jespersen, and T. D. Hansen. 1978. Insertion of bacteriophage lambda into the deo operon of Escherichia coli K-12 and isolation of plaqueforming Adeo+ transducing bacteriophages. J. Bacteriol.

136:668-681. 5. Buxton, R. S., and I. B. Holland. 1973. Genetic studies of tolerance to colicin E2 in Escherichia coli K-12. I. Relocation and dominance relationship of cet mutations. Mol. Gen. Genet. 127:69-88. 6. Chang, A. C. Y., and S. N. Cohen. 1978. Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol. 134:1141-1156. 7. Curtiss, R., III, L. G. Caro, D. P. Allison, and D. R. Stallions. 1969. Early stages of conjugation in Escherichia coli. J. Bacteriol. 100:1091-1104. 8. Curtiss, R., HI, L. J. Charamella, D. R. Stallions, and J. A. Mays. 1968. Parental functions during conjugation in Escherichia coli K-12. Bacteriol. Rev. 32:320-348. 9. Davis, R. W., D. Botstein, and J. R. Roth. 1980. Advanced bacterial genetics. A manual for genetic engineering, p. 226-234. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 10. Guerry, P., D. J. LeBlanc, and S. Falkow. 1973. General method for the isolation of plasmid deoxyribonucleic acid. J. Bacteriol. 116:1064-1066. 11. Gunsalus, R. P., G. Zurawski, and C. Yanofsky. 1979. Structural and functional analysis of cloned deoxyribonucleic acid containing the trpR-thr region of the Escherichia coli chromosome. J. Bacteriol. 140:106-113. 12. Guyer, M. S. 1978. The y& sequence of F is an insertion sequence. J. Mol. Biol. 126:347-365. 13. Hathaway, B. G., and P. L. Bergquist. 1973. Temperaturesensitive mutations affecting the regulation of F-prime factors in Escherichia coli K12. Mol. Gen. Genet. 127:297-306. 14. Hayes, W. 1953. Observations on a transmissible agent determining sexual differentiation in Bacterium coli. J. Gen. Microbiol. 8:72-88. 15. Jamieson, A. F., and P. L. Bergquist. 1976. Genetic mapping of chromosomal mutations affecting the replication of the F-factor of Escherichia coli. Mol. Gen. Genet. 148:221-223. 16. Jorgensen, R. A., S. J. Rothstein, and W. S. Reznikoff. 1979. A restriction enzyme cleavage map of TnS and location of a region encoding neomycin resistance. Mol. Gen. Genet. 177:65-72. 17. Lederberg, J., L. L. Cavalli, and E. M. Lederberg. 1952. Sex compatibility in Escherichia coli. Genetics 37:720730. 18. Lerner, T. J., and N. D. Zinder. 1979. Chromosomal regulation of sexual expression in Escherichia coli. J. Bacteriol. 137:1063-1065. 19. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning. A laboratory manual, p. 150-162. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 20. McEwen, J., and P. Silverman. 1980. Chromosomal muta-

1314

21.

22. 23.

24.

BUXTON AND DRURY

tions of Escherichia coli that alter expression of conjugative plasmid functions. Proc. Nati. Acad. Sci. U.S.A. 77:513-517. Miller, J. H. 1972. Experiments in molecular genetics, p. 15. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Neubauer, Z., and E. Calef. 1970. Immunity phase-shift in defective lysogens: non-mutational hereditary change of early regulation of A prophage. J. Mol. Biol. 51:1-13. Novotny, C. P., and P. Flves-Taylor. 1974. Retraction of F pili. J. Bacteriol. 117:1306-1311. Novotny, C. P., P. F. Taylor, and K. Lavln. 1972. Effects of growth inhibitors and ultraviolet irradiation on F pili. J. Bacteriol. 112:1083-1089.

J. BACTERIOL. 25. Roeder, W., and R. L. Sonervile. 1979. Cloning the trpR gene. Mol. Gen. Genet. 176:361-368. 26. Sancar, A., and W. D. Rupp. 1979. Cloning of uvrA, lexC and ssb genes of Escherichia coli. Biochem. Biophys. Res. Commun. 90:123-129. 27. Shapiro, J. A., and S. L. Adhya. 1969. The galactose operon of E. coli K-12. II. A deletion analysis of operon structure and polarity. Genetics 62:249-264. 28. T nrre, E. F., H. G. Nandadasa, and R. H. Pritchard. 1975. Suppression of initiation-negative strains of Escherichia coli by integration of the sex factor F. J. Bacteriol. 121:554-561. 29. Willetts, N., and R. Skurray. 1980. The conjugation system of F-like plasmids. Ann. Rev. Genet. 14:41-76.