J. x01. Rid. (1991) 218; 119-128

Regulation

of the F Plasmid tra Y Promoter in Escherichia cob by Host and Plasmid Factors

Philip M. Silverman?, Elizabeth Wickersham and Robin Harris Program in Molecular and Cell Biology Oklahoma Medical Research Foundation, 825 NE 13th #treet Oklahoma City, 73104, U.S.A. (Received

1 August

1990: accepted 8 ,Vovem,ber 1990)

F plasmid DNA transfer (tra) gene expression in Escherichia coli is regulated t)J chromosomeand F-encoded gene products. To study the relationship among these regulatory factors, we constructed low-copy plasmids containing a @(traYI-‘1acZ)hyb gene that couples fl-galactosidase and Lac permease synthesis to the F plasmid traY promoter. Wild-type transformants maintained high levels of /$galactosidase over a broad range of cult,ure densities. Primer extension analysis of tra mR?r’A from F’lac and Q(traY’-‘1acZ)hyh strains indicated very similar, though not identical, transcription initiation sites. Moreover. @(tra Y’-‘2acZ)hyb gene expression required both TraJ and SfrA, as does tra gene expression activity was reduced s30-fold in the absence of TraJ, which in F+ strains. /?-Galactosidase could be supplied in cis or in trans. In a two-plasmid system in which TraJ was supplied in expression was a linear. saturable trans by a lac-traJ operon fusion, @(tra Y’-‘1acZ)hyb function of traJ expression. Enzyme activity was reduced z tenfold in @A mutants. That reduction could not be attributed to an effect on the TraJ level. Several ot’her cellular or environmental variables had only a modest effect on @(traY’-‘ZacZ)hyb expression. Hyperexpression was observed at high cell density (twofold) and in anaerobic cultures (1% to I.&fold). Tn contrast, expression was reduced twofold in integration host factor mutants.

1. Introduction Transcript,ional regulation of the F plasmid tra (DNA transfer) genes involves both host- and plasmid-encoded components (for a review, see 1985: Tppen-Thler & Minkley, 1986; Silverman, Willetts & Skurray, 19Si). Among the latter, the TraJ protein is thought to activate transcription initiation at the tray and possibly the traM promoters (Fowler $ Thompson, 1986; Mullineaux & Willetts, 1985; Gaffney et al., 1983). There are additional tra promoters (Manning et al.. 1984; Jalajakumari et al., 1987; Ham et al., 1989), but many and perhaps all of these promoters are expressed to a significant degree in the absence of TraJ (Mullineaux & Willetts, 1985; Grossman & 1989). Regulation of braY promoter Silverman, activity by TraJ is therefore the critical element in regulating conjugal DNA transfer overall. The TraJ level in F+ cells and in those containing F-like R factors is itself regulated by the plasmidspecified fin system (Finnegan & Willetbs, 1973: Willetts. 1977: for a review, see Willetts & Skurray tAuthor

to whom

correspondence

should

be addressed.

(1987) and Paranchych & Frost (1988)). In this system. the plasmid-specific Jin1’ gene product, an anti-sense RNA, represses traJ expression in a process that also requires the fin0 gene product,. F tra gene expression is derepressed owing to an 153 insertion in$nf&. AJinO+ gene from an appropriate R factor restores repression. In addition to TraJ, at’ least three bacterial gene product)s are required to maintain maximal levels of tra mRNA (Beutin et aZ., 1981: Sambucetti et al., 1982; Gaffney et aE., 1983). These are SfrA, CpxA and CpxB (for a review, see Silverman. 1985). Like the Fin system. these gene product,s have also been thought) of as acting by altering the TraJ level (Reutin et al., 1981; Cuozzo & Silverman. 19X6), thus preserving TraJ as the only genetic regulat,ory component acting at’ the tra Y promoter. Other studies have suggested that integration host factor (IHF$) and the tray gene product itself play a role in regulating tra gene expression (Gamas it al.. 1987:

SAhbreviations used: IHF, integration host-f’act’or; kh. IO3 bases or base-pairs; bp. base-pairs: TPT(;. isopropylB-u-thio~alac~tosit-le.

l)t~mpsey, 19X7: lnamoto $ Ohtsubo. 1990: Bowit, & Salwr, 1990). Analyses of tra mRNA levels and of /Ggalaetosidasr levels in F tra : : la& fusion strains indicated t,hat) F trrr gene expression is reduced at least 20-fold in tht> ahsent>r ofTra,J (Willetts. 1977; Gaffney rt al., 1983). In several reports small plasmids containing a tra Y promoter situated so as t.o drive expression of a reporter gene have been described; however. results wit,h these plasmids have generally been disappointing hetsause reporter gene expression was stimulated only modestly in traJ+ cells (Fowler rt nl., 1983; Finlay et al., 1986). Mullineaux & Willet’ts (1985). however. using a fm l’-galK operon fusion plasmid. reported a 1S-fold st’imulation of GalK synt’hesis in tmJ+ cells. Similarly, the drpendent~c of trrrA/ expression on TraJ is unresolved (Fowler c!+ Thompson. 1986: Mullineaux & \\Tillet ts. 1985; Cs#ney rt al., 1983). The precise location of the F plasmid t,ro Y promoter is also unclear. Studies with expression plasmids indicat’ed that it is at, or near a BstEIl sit’e in t)hc int,ergeniu segment between traJ and tml (Fowler it al.. 1983; Mullineaux & Willetts. 198.5: Fowler & Thompson. 1986). One study suggested that there are several regulatable promoters within t,his segment (Fowler bz Thompson. 1986). It is not known which. if not all, of t’hem are used in F+ c4ls. M:e describe here t’he construction and properties of tm I’ promot’er plasmids that, differ in several respet+s from those previously described. Most important, they conserve the sequence context of t ht. tra Y promot’er. and transcription initiation f’rom the plasmid tra Y promot’er is qualitatively vrhry similar t)o that from F itself. We have used these plasmids to study quantit,atively the effects of’ factors known (or suspected) t’o affect fro 1’ promoter funt+ion

2. Materials and Methods (a) Bacterial

Strains

strains

and

plasmids

used in these studies were all derivat’ives of coli K-l 2 from our laboratory st,ocks. linreferenced strains were constructed in this laborator> by standard methods. DH.5cr (Bethesda Researcsh 1,aboratories) and JC3272 (his, lys, trp Alac. Achtman rt al., 1971) have been described. AE2086 is a gyrA21 (nalR) derivative of JC3272. M1174: Ml163 and Ml164 art’. respectivel,v. isogenita his+ @A ’ , I&+ sj’rA4 and his+ $A5 derivatives of JC3272 (Beutin & Acht’man. 1979). AE2280, AE2276 and AE2278 are rPcA1 derivatives of 111174, M 1163 anti 111164. respect’ivrly. Strains K5302 (himAA Sma,T : :TnknnR; Cranston et al., 1988) and K2704 (hipA : : cnmR: Flamm &. \Veisberg. 1985) were obtained t’rom Dr I). Friedman. The IHF mutations were moved by bacteriophage Pl transduction into strain J(Z3272, in which the) were tested for their effects on tra Y promoter at+vity. i’hloramphenicolor kanamycin-resistant transduotan& were screened for their abilitv t)o maintain a pN1101 replicon (pL(i339; see below), which requires IHF (Biek & Cohen, 1989. and references therein): transformation frequentairs with pLG339 were reduced > 103-fold hy &her THF mutation. Ir:sch&;hia

I’lasmids .l(‘FI,O (I?’ (1.~1 /UC ). .l(‘FL!N) (l~‘/rtrJ!NIr\m / ltrc ‘). pRS25. RIO0 (Irl. ,$,,0’ ) and It 100-I i/P/ ,/ir,O-1 ) have all been tlt~st~ribrtl (;1c~htman V/ trl.. I!17 I 1!)7X: Egawa & Hirota. 1962). pRW300 was drrivtbtl t’ronl pLG339, a pS(‘lO1 replic~on that c~onfc~rs kananiycin ant1 t,rtracyclinr rrsistanc.r (Ntokrr, c,t (11.. 19X2). 1)). rr~tlloving an k;coRI ~k:coR\ fragment that includrs t)hcx frt pronlc~tc~r rt~sirlts). The uniquf rcbsl ric? iotl (F. Lvrbel~, utipublishrd sit,es in thr t~1 gene. including the Snll sittx. ~~~1~1~1 Ithat fro Y promoter Table 1 TraJ-dependent expression of a @(tray’-‘1acZ)hyb gene: Tra,J complementation in trans Plasmid(s)t pLW403 (k&/ +) pLW404 (traJAl5)/JCFLO (traJ+) pLW404 (tmJA15)j.JCFLWO (traJW[AM])

tray rspr~ssion$ 1x54. 2110 w4. MT, 93. 97

tThe bacterial drain was JC3272. f/G(:alactosidase activities of duplicate cultures ww measurpd as drsl*ribed in Mat,erials and M&hods.

125

I’.

M.

Silverman -

et

kll

M/u1 BamHI

HindIII(Taq1 lppP IacPO

traJ

CGACTGlCC~TAG~ATCClTiGTGAGGAGGTTCCTATG CGCGTT~ATAAGGTACCGGAc -

5

0

IO

20

15

[IPTG] tP~) (bi

Figure 4. Trans-activation of tra Yp by TraJ. (a) The l~q/Zac-tm.1 operon fusion of pRH103. The sequences indicate, respectively, the ribosome binding site and initiating ATG codon of TraJ (Cuozzo & Silverman. 1986) in relation to the 5’ terminus of the P DPljA of pRHl03 (wnfirmed by sequence analysis). and the synthetic M~u-~B~~HI linker used to reconstruct the 3’ truJ codons. (b) Bet’a-galaetosidase activity as a function of traJ induction. The recA1 strain AE2280 containing

pR,H103 and pLW405 was grown in the presence of IPTU at t,he indicated concentrations. The inset shows a second rxperiment over a narrower range of inducer concent’rations. Open cultures.

and filled

symbols

are data

from

duplicate

initiation at the traJ promoter (Beutin et al.. 1981): since tra I’ promoter activity in pLW405/pltH 103 transformant’s was still @A-dependent, (Table :3), whereas traJ transcription in that syst’ern was fr(JrrI expression vector promoters (see above). Furthermore, direct assay by immuno-overlay blotting showed that the sfrA mutations did not. reduce the TraJ level by any mechanism. If anything, the .$+A mutants accumulated somewhat more TraJ than did the otherwise isogenic sfrA + st’rain (Fig. 5). The extracts used in this assay were also measured for P-galactosidase activity. with results comparable to t’hose in Table 2. (d) Anaerobic

expression

of traYp

During anaerobic growth. SfrA represses the expression of genes encoding proteins that only function aerobically; this property of SfrA has been designated Arc (aerobic respiration control: Iuchi & Lin. 1988). An obvious possibility is that negative Arc and positive Sfr functions are mutually rxclusive activities of the same protein. This simple

activity w-as reduced in the sfrA mutants, though not t’o t’he degree observed with cells lacking TraJ. INoreover. Western blots (not shown) indicated that residual LacZ polypeptide in sfrA mutant cells is the same size as that in sfr.4+ cells. in contrast to LacZ polypeptide synthesized in t,he absent of TraJ (see above). Thus. the effects of the sj?;l mutations on tm y promoter function are not its extreme as t’hat) of TraJ deprivation. Previous studies attributed the effect, of sj’rd mutations to an effect on TraJ levels. owing to a,n effect on either fraJ expression or TraJ protein turnover (Beutin rt al.. 1981: (luozzo di Silverman. 1986). U’e could rule out an effect on tranwript,ion I

Table 2 Ejffect of chromosomal sfrA mutations on expre0ressiwnC$ the @(tray’-‘1acZ)hyh gene of pLW403 I3acterial strain Ml174 Ml 16’3 JlIl&

tru Y expressiont

,xfrA+ @A4 @A5

1648. 16.59 123, 123 115, 122

2

3

Figure 5. In viva TraJ accumulation.

4

5

Immuno-overlay blots were prepared as described in Materials and Methods. Host strains and plasmid content are indicated above the data; the @A+ and sfrA mutant strains were Ml174 (lane 3), Ml163 (lane 4). and Ml164 (lane 5). The mobilities of proteins with the indicated molecular masses (M, x 103) and of purified TraJ are indicated on t,he left. The unidentified immunoreartive species at about 30 x lo3 M, serves as a loading standard ((‘uozzo & Silverman. 1986).

Regulation of the F Plasmid tray

Promoter

125

Table 4 Aerobic and anaerobic expression of the @(tray’-‘1acZ)hyb

gene

Ira Y expression? Straiqplasmid

Relevant

Ml 174/pLW403 Ml 163/pLW403 M1164/pLW403 K!3272/pLW405

@A+ sfrA4 sfrA5 sfrA+

genotype

traJ+ traJ+ traJ + traJA15

Anaerobic 2603f495 381+68 345+73

83+-11

(5)

Aerobic

Anaer/aer$

1916k260 240+49

144

1.45

195+ 14 (4) 66&6

The data are the mean + standard deviation of six cultures, except as indicated ttra Y expression was measured as described in Table 1. IRatio of enzyme activities in anaerobic and aerobic cultures. $Aerobic and anaerobic means are different (98 to 99o/0 confidence level). Jl.l\erohic and anaerobic means are different (>99?/, confidence level).

scheme would result in the aerobic expression of tra genes and genes under aerobic respiration control, and the anaerobic expression of neither. However, j?-galactosidase activities in anaerobic pLW403 cultures were no less than in aerobic cultures (Table 4); in fact, we consistently observed slight anaerobic hyperexpression, even in cells lacking SfrA or TraJ. (e) tra,Y expression in IHF

mutants

We have also tested the hypothesis that IHF is required for tra Y promoter function (Gamas et al.. 1987: Dempsey, 1987). In order to maintain a constant strain background, we constructed JC3272 derivatives of two IHF mutants, one in each of the two subunits of the protein, and transformed them with pLW301. We were unable to use pLW403 for these experiments because IHF is required for maintenance of pSClO1 replicons (Biek & Cohen, 1989, and references therein). In contrast, the transformation frequencies of the IHF mutants with the pBR322 replicon pLW301 were similar to that of JC3272 itself. Although pLW301 is nominally of higher copy number than pLW403, tra gene expression with both replicons was similar in magnitude (compare Tables 1 and 5) and similarly dependent on TraJ (/?-galactosidase activity = 75 Miller units in a pLW304 (traJA15) transformant) and SfrA (p-galactosidase activity = 160 Miller units in the sfrA strain Ml 164). The IHF mutant transformants maintained 40 to 45% of the ,%galactosidase activity of JC3272(pLW301) (Table 5).

Strain ,JC3272 AE2333 AE2335

of thr @(tray’-‘1acZ)hyb mutants IHF genotype himA ’ hip+ AhimA (SmaI : : Tnkana) hip+ himA+ hipA : : earn

gene in IHF

in parentheses

4. Discussion (a) The tray promoter region Primer extension analysis of RXS from @(tray’-‘2acZ)hyb and from F’ cells indicated transcription initiation from similar, though not identical, sites. The common sites are consistent with other data placing the tray promoter within 20 bp of the BstEII site between the trd and tra Y coding regions (Mullineaux & Willetts, 1985; Fowler & Thompson, 1986). Part of that intergenic sequence is shown in Figure 6, along with the 5’ terminus of the quantitatively predominant tra I- transcript that is evident in primer extension experiments. There is reasonable agreement between consensus features of E’. coli a7’ promoters (Mulligan & McClure. 1986; Alexandrov & Mironov. 1990) and the ha sequence at positions - 10 and - 35. While the data identify the promot,er shown in Figure 6, Mullineaux & Willetts (1985) and Fowler & Thompson (1986) noted other, plausible a7’ promot,er sites near the BstEII sit,e. Furthermore, Fowler & Thompson (1986) noted that. in the absence of t,he region immediately surrounding the B&J311 site, transcription apparently originated from sites closer to the traJ translation t’ermination codons than the B&E11 site. They inferred the existence of shadow promoters whose activities were evident only in t,he absence of the tra Y promoter or in vitro. Finally, our primer extension analyses of F’ cells indicated that tray transcripts originate from multiple

sites within

a 35 bp segment

All t,hese observations several tra Y promoters.

Table 5 Expression

1d 1.311

-50

-LO

-30

(see Fig. 3).

suggest that there may be

-20

-10

l 1

;GACTTTCCG;~TATTTA~~~TTT~TGATT;TT~TGCAAA~ATAAGTGGT~ TTGaca-17 bp-TAlaaT TTGAca-17 bp-,AtaaT

tra Y expression? 1847 1520 712 758 626

748

All strains contained pLW301, ttra Y expression was measured as described for Table 1.

Figure 6. The ,F plasmid tra Y promoter. Base + 1 is the adenosine corresponding to the 5’ terminus of the quantitatively predominant tray transcript (see Figs 2 and 3). The consensus hexamers and spacing are from Mulligan & McClure (1986) and Alexandrov & Mironov (1990) (top and bott,om. respectively). Upper-case characters indicat,e more important’ features.

126

I’. M. Silverwuzn

A complex pattern of transcription initiation sites does not imply multiple promoters. Relatively modest changes in a consensus E. coli a7’ promot,er altered transcription initiation from a simple pattern to one in which as many as five initiation sites were spread over a 25 to 30 bp segment (Jacquet & Reiss, 1990), much as we observe with R’NA from I” cells. One variable may be the base composit’ion of the spacer separating the - 10 and -35 consensus hexamers; apparently, the higher the A +T content, the more heterogeneous the pattern (Jacquet & Reiss, 1990). According to that rule, the pattern for tray expression should be heterogeneous, since the A + T content of spacer for the tra,Y promoter is 12/17 bases. Subtle differences in promoter structure owing to different sequence contexts might account for the slightly different pat’terns of transcription initiation sites that, we observe with RNA from F” and pLW403 cells. As to the shadow promoters identified by Fowler $ Thompson (1986), we note that the (A+T) content’ of bp -80 to bp - 1 (with respect to the BstEII site) is 71 “/0 (Fowler et al., 1983) and that there is a strong positive correlation between promoter sites and A+T content (Mulligan & McClure, 1986). Such sites may in fact be used in the absence of the normal promoter (Fowler $ Thompson, 1986) and may even have regulatory significance (Malan $ McClure, 1984). Finally, while our data do not’ rule out failed primer extension product’s or degraded mRNA molecules as sources of error, they generally agree with results obtained by complementary approaches (Mullineaux & Willetts. 1985: Fowler & Thompson, 1986).

(b) The role qf TraJ

in tray

promoter

activity

Of all the components suggested to affect F plasmid tru gene expression, TraJ is most clearly required (Willetts, 1977; Gaffney et al., 1983; see Fig. 3 of this paper). However, the mechanism of action of TraJ has not been elucidated (Ippen-Ihler & Minkley, 1986), nor has the relationship between TraJ and other regulatory molecules (Heutin et ul.. 1981; Cuozzo & Silverman, 1986). In vivo and in vitro data have been ambiguous as to the level of dependence of the tray promoter on Tra.J protein (Fowler et al., 1983; Mullineaux & Willett,s. 1985; Finlay et al., 1986; Fowler & Thompson, 1986). Furthermore, shadow promoters near truY showed a degree of Tra,J dependence in Z%IOcomparable to that of the tm Y promoter itself (Fowler liL Thompson, 1986). With the plasmids described above, O(truY’‘ZacZ)hyb expression was consistently stimulated 30-fold in TraJ+ cells. We believe that lesser TraJ effects reported by others resulted from elevated levels of constitutive tra Y promoter activity, owing to the sequence context of the promoter. For example, we will show elsewhere (Silverman et al., unpublished results) that alteration of the DNA sequence 5’ to the tray promoter elevates promoter

et al

activity in t,he absence of TraJ and SfrA as much as 1O-fold. TraJ could be supplied in cis or in trans. Wk~rn supplied in truns by F’lac, Q(tru Y’-‘lac%)hyh gene expression was only about half of t’he maximum. Studies in which the level of truJ expression in t,hc: cell could be varied indicated that t,hr Irvc~l of cD(tru I’-‘ZacZ)hyb gene expression is a sat.urable function of the TraJ level. Tt thus appears that TraJ was limiting in the F’ strain. However. F’ and Hfr strains contain several thousand TraJ monomers (Cuozzo & Silverman, 1986). Even if TraJ act’s as a homodimer, as zone sedimentation data suggest (Cuozzo & Silverman, 1986). there would st,ill br 1000 to 2000 active TraJ molecules per cell. \I’rt . an F’lac/pLLI’404 strain should contain at’ most about, 12 tray promoters, assuming a copy number of six for each plasmid (Stoker et ul.. 1982; Shields et al., 1986). There may be addit)ional Tra*J-deprndent promoters. but it is not, likely that they would exceed the number of TraJ molecules that, we estimated. The apparent paradox that tra Y expression is limited when cells contain a molar excess of Tra.J. probably a large ex~rss. will probably not be resolved without %n uitro analyses of thra roles of’ TraJ and SfrA in tm gene exprrssiorr. (c) Roll qf SfrA in tray promoter

activit?/

8frA is one of a family of genetic regulatory proteins whose activities are controlled by phosphorylation/dephosphorylation; homologous proteins are DNA binding proteins that in their active form regulat)e transcription initiation at appropriate promoters (summarized by Stock et al.. 1989; see also Makino et al., 1989: and Jin et al., 1990). RfrA may act in an analogous way to stimulate trrr gene expression. Two reports indicated t,hat sfr A mutations lead to reduced TraJ levels in F’ strains (Beutin et al., 1981; Cuozzo & Silverman, 1986). We do not’ find such an effect in pLW403 strains. where the effect of the ,@A mut,at,ions OII t,ru Y promoter function is nevertheless apparent,. Hence, while sfrA mutant F’ strains do not accumulat,e as much TraJ protein as ,sfrA+ strains (Cuozzo &. Silverman. 1986), the difference may not, be of major regulatory significance. Instead, our results indicate that SfrA acts directly at the tray promoter. SfrA functions anaerobically to repress chromosomal genes under aerobic respiration control (Iuchi & Lin, 1988). As tmY promoter activity is, if anything, slightly elevated in anaerobically grown cells, the repressor and activator functions of SfrA appear to be independently regulated propert>ies of the same protein, as concluded also by Iuchi rt nl. (19896). It follows that the total SfrA level must be sufficient under anaerobic conditions to allow both repression of aerobic genes and activation of trn genes (see Table 4). This notion is consistent with evidence that different membrane sensors regulate Arc and Sfr functions (Cuozzo & Silverman, 1986: Weber 8: Silverman, 19X8; Rainwater & Silverman. 1990; Tuchi et ul.. 1989a,).

Regulation of the F Plasmid tray (d) Other factors affecting tray promoter activity TraJ and SfrA appear to be the major determinants of the level of tray promoter activity. However, we found that several other cell and environmental factors affect @(tra I”-‘ZacZ)hyb gene expression, albeit weakly. These include cell culture density, anaerobiosis and IHF. All three alter DNA structure, the first two by altering superhelical density (Dorman et al., 1988) and IHF by introducing bends in DNA (Friedman, 1988; Goodman & Nash, 1989). Increasing culture density and aerobic growth favor more relaxed DNA (Dorman et al., 1988). O(tra Y’-‘ZacZ)hyb gene expression increased with culture density, but was lower in aerobically grown than in anaerobically grown cells. Hence. superhelical density is unlikely to be a major quantitative determinant of tray promoter activity, at least in the plasmids we constructed. However. there remains the possibility that qualitative differences between transcription initiation in F’ and pLW403 cells are attributable to subtle changes in the tray promoter caused by differences in the superhelical structure of the two plasmids (Pruss $ Drilica, 1989). @(tra Y’-‘ZacZ)hyb gene expression was reduced to 40 to 45% of the maximum in IHF mutants. This effect was not as dramatic as others we have described and seems unlikely by itself to account for the effect of THF mutations on donor activity (Gamas et al., 1987; Dempsey, 1987). However, the quantitative relationship between tra gene expression and donor activity is not known; IHF mutations may affect other tra promoters more drastically than they do tra Y; more dramatic effects of IHF on tray activity might be evident with F itself; and IHF may be required for donor activity at some stage other than tra gene expression. Additional studies are necessary to resolve these uncertainties. We thank Dr David Friedman for the IHF mutant strains and Dr George Weinstock for advice on RNA isolation. This work was supported by grant GM38657 from the National Institutes of Health. P.M.S. gratefully acknowledges support from the Marjorie Nichlos Chair in Medical Research. References Achtman, M., Willetts, N. & Clark, A. J. (1971). Beginning a genetic analysis of conjugational transfer coli by determined by the F factor in Escherichia isolation and characterization of transfer-deficient mutants. J. Bacterial. 106, 52S538. Achtman, M., Kennedy, N. & Skurray, R. (1977). Cell-cell interactions in conjugating Escherichia coli: Role of the TraT protein in surface exclusion. Proc. TVat. Acad. Sci., C:.I\‘.A. 74, 5104-5108. Achtman. M., Skurray, R., Thompson, R., Helmuth, R.. Hall. S., Beutin, L. & Clark, A. J. (1978). Assignment, of tra cistrons to EcoRI fragments of F sex factor DNA. J. Bacterial. 133, 138331392. Alexandrov, N. & Mironov. A. (1990). Application of a new method of pattern recognition in DNA sequence

Promoter

analysis: a study of E. coli promoters.

127 ~I’ucl. Acids

Res. 18, 1847-1852.

Ausubel, F., Brent, R., Kingston, R., Moore, D., Smith, J.. Seidman, J. & Struhl, K. (1987). Current Protocols in Molecular Biology, John Wiley & Sons, New York. coli Beutin. L. & Achtman, M. (1979). Two Escherichia chromosomal cistrons, sfrA and sfrB, which are needed for F plasmid tra functions. .I. Bacterial. 139, 730-737. Beutin, L., Manning, P., Achtman, M. $ Willetts, N. (1981). @A and sfrB products of Escherichia coli K-12 are transcriptional control factors. J. Bacterial. 145, 840-844. Biek, I). & Cohen, S. N. (1989). Involvement of integration host factor (IHF) in maintenance of plasmid of pSClO1 in Escherichia coli: characterization pSClO1 mutants that replicate in the absence of IHF. J. Bacterial. 171, 20562065. Bowie, J. & Sauer, R. (1990). TraY proteins of F and related episomes are members of the .4rc and Mnt repressor family. J. Mol. Biol. 211. 5-6. Casadaban, M., Chou, J. & Cohen, S. (1980). In vitro gene fusions that join an enzymatically active P-galactosidaae segment to amino-terminal fragments of exogenous proteins: Escherichia coli plasmid vectors for the detection and cloning of translational initiation signals. J. Bacterial. 143, 971-980. Cuozzo, M. & Silverman, P. M. (1986). Characterization of the F plasmid TraJ protein synthesized in F’ and Hfr strains of Escherichia coli K-12. J. Biol. Chem. 261, 5175-5179. Dempsey, W. (1987). Integration host factor and conjugative transfer of the antibiotic resistance plasmid RlOO. ,I. Bacterial. 169, 43914392. Dorman. C., Bhriain: N. & Higgins. C. (1990). DNA supercoiling and environmental regulation of virulence gene expression in Shigella ,jexneri. Nature (London), 344, 789792. Egawa. R. & Hirota, Y. (1962). Inhibition of fertility by coli multiple drug resistance factor in Eschkchia K-12. .Jpn J. Genet. 37, 6&69. Finlay, B., Frost, L. & Paranchych, W. (1986). Nucleotide sequences of the Rl-19 plasmid transfer genes traM, and tray and the tra I-2 promoter. ,finP, traJ, J. Bacterial. 166, 368-374. Finnegan, D. & Willetts, N. (1973). The site of action of the F transfer inhibitor. Mol. Grn. Gwwt. 127. 307-316. Flamm. E. & Weisberg: R. (1985). Primary structure of the hip gene of Escherichia coli and its product. the / subunit, of integration host factor. .J. Mol. Biol. 183, 117.-128. Fowler. T. & Thompson, R. (1986). Shadow promoters in the F plasmid transfer operon. 12fol. Gpn. Genst. 202, 50%5 11. Fowler, T., Taylor, L. & Thompson, R. (1983). The control region of the F plasmid transfer operon: DNA sequence of the traJ and tra Y genes and rharacterization of the tray2 promoter. Gene, 26, 79-89. Friedman. D. (1988). Integration host factor: a protein for all reasons. Cell, 55. 545-554. Gaffney. D.. Skurray, R. & Willetts. N. (1983). Regulation of the F conjugation genes studied by hybridization and tra-la& fusion. J. Mol. Biol. 168, 103122. Gamas. P., Caro. L.. Galas, 1). & Chandler. )I. (1987). Expression of F transfer functions d~prnds on the Escherichia coli integration host factor Xol. Gpn. Genpt. 207. 302-305.

C:oodman & Nash (1989). Functional replacement ot a protein-induced bend in a DNA recombination sitf‘. Nature (London), 341. 251-254. Cranston. A.. Alessi. l).. Eades, I,. & Friedman. I). (IBX8). A point mutation in the Ivul gene of bacteriophage i facilitat,es phage growth in Exchrrichia coli with hiwrd and gyrB mut,at,ions. Mol. C&n. Genet. 212, 149bl.X. Cirossman. T. & Silverman. I’. M. (1989). Strufaturr and function of conjugative pili: inducible synthesis of functional F pili by &.Aerichia coli K-12 containing a lac-tra operon fusion. J. Bacturiol. 171. 650-656. Ham. L.. Firth. X;. HL Xkurray, R. (1989). Nucleotide sequence of t,hr F plasmid transfer gene. fraH: ident,ification of a new gene and a promoter within t)hr t,ransfer operon. Gene, 75, I.!% 165. Highlander. S.. Engler. M. & Weinstock, G. (1990). Secretion and expression of the Pasteurelln hnrnw lytica Irukot,oxin. .J. Kactkol. 172, 2343G2350. Inamoto. S. & Ohtsubo. E. (1990). Specific binding of thr TraY protein to oriT and the promoter region for the trrr Y gene of plasmid RI(N). .J. Riol. (‘hem. 265. 6461-6466. Ippen-Thler. K. 6t Ylinkley. E.. Jr (1986). The conjugation system of F. the fertility factor of Eschrrichio coli. dnnu.

RPP, &net.

20, Ti93-624.

Tuchi. S. Br IA. E. C’. C‘. (1988). nrcA (dye). a global regulatory gent: in Eschrrichia coli mediating rrpression of enzymes in aerobic pathways. Proc. LVat. dead. AK., i’.S.,l. 85. 1888 -1892. Tufthi. S.. Cameron. 1). & T,intt. E. (‘. C‘. (1989n). A second global regulator gene (ur~H) mediating repression of enzymes in aerobic pat~hways of Escherivhia 4i. .J. Bacfrriol. 171. X68-873. I). & Lin. E. (‘. (‘. (1989h). Iuchi. S.. Furlong. IXfferentiation of arc A, arcB, and r:p.r.-l mutant J)henotypes ot’ Escherichin coli by sex pilus formation and enzq’me regulation. .I. Ha&viol. 171. 2889p2893. .Jacquet. Mm.\. 8r Kriss. (‘. (1990). Transcription in viva directed by consensus sequences of E. coli promoters: their context heavily affects eficiencaies and start sites. Sucl. Acids Kes. 18. 11~37-1143. .Jalajakumari. M.. (:uidolin. A.. Buhk. H ,. Manning, I’.. Ham. I,.. Hodgson. A.. (Iheah. K. XL Skurray. K. (1987). Surface exclusion genes traS and traT of t,hr F six facator of Escherichia coli K-12: determination of the nucleotidt~ seyuenee and promoter and terminator ac*tivities. ,J. Mol. Kiol. 198. I 11. Jin. S.. Roitsch. T.. (‘hristie. I’. 8r X-r&r. E. (1990). The regulat,ory VirG prot,ein specifically binds to a cisacting regulatory sequence involved in transcrip tional activation of Agrobacterium tumefarie?Ls virrrlencr grnes. .I. Bacterial. 172. 531 537. Lrdrrberg, ,I.. C’avalli. I,. & Ledrrbrrg. E. (1952). Sex compatibilit! in Escherichia coli. (kn,etics. 37. iP(t730.

Makino. K.. Shinegawa, H.. Amemura. M., Kawamoto. T.. Yamada. M. & Nakata, A. (1989). Signal transduction in the phosphate regulon of Escherichin co/i involves phosphotransfer between PhoR, and PhoH proteins. J. Mol. Biol. 210, 551-559. Malan. T. $ McClure. W. (1984). Dual promoter control of the Escherichia coli lactose operon. C’ell, 39, 173-180. Maniatis. T., Fritsch, E. F. & Sambrook. J. (1982). Molecular C’loning: A Laboratory Manual, Cold Spring Harbor Laboratory Press. C!old Spring Harbor, NY.

Edited

(i. B Fisseau, (‘. (19X4). Manning, P.. Morelli. RNA-polymerase binding sites within the trrr region of the F factor of Escherichia coli K 1”. (kne. 27. 121 123. Masui, Y.. Mizuno, T. & Inouye. M. (1984). Novel high-level expression cloning vehicles: 104-fold amplifiEscherichia coli tninor cation of prot,rin. Rio/Technology, 2. 81-85. McEwen, J. & Silverman, P. M. (1980). (‘hromosomal of Escherichia co/i K-12 that alt’er mutations expression of conjugative plasmid functions. l’roc. iVat. Acad. AS?%..l’.N.A. 77. 513Wjl7. Miller. J. (1972). Experiw~enta in Molecular Genetica, (fold Spring Harbor Laboratory Press. (‘old Spring Harbor, KY. Mulligan, M. & McClure, W. (1986). Analysis of thtl occurrence of promoter sites in DNA. Nucl. Acids Krs. 14, lo9-~126. Mullineaux. P. & Willetts. X. (1985). Promoters in thr transfer region of the plasmid F. In Pla.smids in Bacteria (Helinski, I).. Cohen. S., ClewelI. I).. ,Jackson. 1). & Hollaender, A., eds). pp. 605 614, Plenum Publishing Co.. Xew York. Paranchyh, W. bt. Frost. I,. (1988). The physiology and biochemistry of pili. rlrl~an. Microbial. Physiol. 29. a3 114. Pruss. (:. & IMica, K. (1989). DNA supercoiling anti prokaryotic transcription. (Tell, 56. 521-523. Rainwater. S. & Silverman, P. ,M. (1990). The (!px proteins of Escherichia coli K-l 2: Evidence that, rp.rcd, ~cfH. sstl, and eu,p mutations all identify thr same g:enr. ,J. ilacteriol. 172. d45&2461. Sambucrtti. I,.. Eoyang. I,. & Silverman, I’. 31. (1982). (‘ellular co&r01 of conjugation in Eschrrichia coli K- 12: Effect of chromosomal C~X mutations on V plasmid gene expression. .J. Mol. Hiol. 161. I S31. Shields. M.. Kline, H. & Tam. .I. (1986). A rapid method for the quantitative measuremrnt of gene dosage: mini-F plasmid concentration as a function of growth rate. J. Microhiol. Neth,ods. 6, 33-46. Silverman. P. M. (1985). Host cell- plasmid intrractions in the expression of Dh’A donor activity bg F’ strains of hkcheriehia cali K- 12. BioEssays, 2. 2544259. Stock. J., Sinfa. A. & S;;tock. A. (1989). I’rot,rin phosphorylation and regulat,ion of adaptive rrsponscs in bacteria. Microbial. Reel. 53, 450-490. Stoker, X.. Fairweather. S. $ Spratt. K. (1981). Versatile low-copy-number plasmid vectors for c.loning in Escherichia coli. fkn,r. 18, 335-~341 Thompson. Jt. & Taylor. I,. (1982). Promoter mapping and DiVAA sequencing of the F plasmid transfer eentxs traN and fra.J. Mol. Grrr. (&net. 188. 513m518. Webrr. R. & Silverman. P. 11. (1988). The (‘px protrins ot Escherichia coEi K-12: st,ructure of thr C’pxA polypeptidr as an inner membrane components. .I. Mol. Viol. 203. 167-478. Willetts. N. (1977). The transcriptional control of fertilit! in F-like plasmids. .I. Mol. Biol. 112. 141. 118. Willetts. X. 8 Skurray. R. (1987). Structure and function of t,he F factor and merhanism of conjugation. In Kscherichia, coli and Salmonella typhimurium: (‘ellulur orrd Molecular Biology (Neidhardt, F.. rd-in -chief) pp. I 110-1133. American Society for Microbiology. Washington, DC’.

by J. H. Miller