Gene, 166 (1995) 43-48 © 1995 Elsevier Science B.V. All rights reserved. 0378-1119/95/$09.50
43
GENE 09324
Regulation of tcp genes in classical and E1Tor strains of Vibrio cholerae 0 1 (Toxin co-regulated pilus; activation; promoter; fimbriae)
Sonia Thomas, Susan G. Williams and Paul A. Manning Microbial Pathogenesis Unit, Department of Microbiology and Immunology, The University of Adelaide, Adelaide S.A. 5005, Australia
Received by R. Morona: 17 February 1995; Revised/Accepted: 7 April/10 April 1995; Received at publishers: 4 September 1995
SUMMARY
Expression of genes encoding the toxin-co-regulated pilus (TCP) varies between the two biotypes of Vibrio cholerae O1. Sequence analysis of the tcp locus from the classical and E1 Tor strains has revealed differences in the intergenic regions between tcpI and tcpP, and tcpH and tcpA, which may be involved in regulation. To investigate this possibility, transcription of tcpA, and the predicted upstream promoters for tcpI and tcpP, has been analysed in the classical and E1 Tot strains using promoter-cat (chloramphenicol acetyltransferase) fusions. Together with primer extension analyses, these studies indicate that the tcpA and tcpP promoters are toxR-dependent and suggest that TcpP may be involved in activation of both the tcpI and tcpP promoters. We conclude that differences in the level of tcpA expression in a classical and an E1 Tor strain are likely to be due to the effect of sequence variation on the ability of control factors to act on these regulatory regions.
INTRODUCTION
In order to cause disease Vibrio cholerae (Vc) O1 must first colonize the intestinal mucosa. This involves a number of surface associated and extracellular determinants as the organism must traverse the mucus layer and adhere to the gut epithelium (Manning, 1994). Since attachment and colonization are essential steps in the pathogenesis of this organism, the components involved in these processes represent potential vaccine targets. Correspondence to: Dr. P.A. Manning, Microbial Pathogenesis Unit, Dept. of Microbiology and Immunology, The University of Adelaide, Adelaide, S.A. 5005, Australia. Tel. (61-8) 303-5974; Fax (61-8) 303-4362; e-mail: pmanning@micr°bscience'adelaide'edu'au
Abbreviations: A, absorbance (lcm); aa, amino acid(s); Ab, antibody(ies); bp, base pair(s); CAT, Cm acetyltransferase; eat, gene encoding CAT; CFA, colonization factor agar; CFB, colonization factor broth; Cm, chloramphenicol; kb, kilobase(s) or 1000 bp; NB, nutrient broth; nt, nucleotide(s); P, promoter; PhoA, alkaline phosphatase; RBS, ribosome-binding site(s); TCP, toxin-coregulated pilus; tcp, gene encoding TCP-related function; u, unit(s); Vc, Vibrio cholerae; [], denotes plasmid-carrier state. SSDI 0378-1119(95)00610-9
One of these, TCP, has been implicated as a potentially important immunogen (Taylor et al., 1987; 1988). The role of TCP in colonization and its significance as a protective antigen has been assessed in humans as well as animal models (Herrington et al., 1988; Sharma et al., 1989a,b). One feature of the study of colonization factors, in particular TCP, in Vc is that the expression is very dependent upon the cultural conditions including medium, temperature, osmolarity and degree of aeration (Taylor et al., 1987; Jonson et al., 1990; Voss and Attridge, 1993). This is most marked, especially with E1 Tor strains, which do not produce TCP under the same conditions as classical strains (Hall et al., 1988; Sharma et al., 1989a; Jonson et al., 1990; Voss and Attridge, 1993). This implies that there are differences in the regulation of tcp expression between these biotypes. The 5-kb XbaI region of the tcp operon has been sequenced for both the classical strain Z17561 and the E1 Tot strain H1 (Ogierman et al., 1996) and has revealed two additional ORFs not previously identified by TnphoA mutagenesis (Taylor et al., 1988). These genes
44 have been designated tcpP and tcpQ but the roles of their protein products have not been established (Ogierman et al., 1996). However, based on structure and sequence homologies putative roles have been suggested (Iredell and Manning, 1994a; Ogierman et al., 1996). Sequence comparison between the classical and E1 Tor tcp regions indicates greater than 98% homology in the tcpB, tcpH, tcpI, tcpP and tcpQ genes (Fig. 1). The tcpA gene is only 77% homologous between the biotypes, with most variation at the C terminus probably associated with biotypespecific antigenic variability (Sun et al., 1990; Iredell and Manning, 1994a; Ogierman et al., 1996). The intergenic regions between tcpI-tcpP and tcpH-tcpA also differ between the biotypes and show 89 and 87% identity, respectively. The introduction of the 5-kb XbaI region from the classical strain Z17561 into the E1 Tor strain H1 leads to the expression of tcpA when cultured on CFA medium (Evans et al., 1979) although H1 itself does not produce TCP under the same conditions (data not shown). These data suggest that sequence variations which occur within the intergenic regions may be responsible for the differences in tcpA expression. Moreover, the predicted aa sequences for TcpH and TcpI imply that it is unlikely that they act, as originally proposed (Taylor et al., 1988), as typical transcriptional activator and repressor of TCP biosynthesis, respectively (Harkey et al., 1994; Iredell and Manning, 1994b; Ogierman et al., 1996). Although genes required for the biosynthesis of TCP have been identified, little is known about the transcriptional organization of this region. We have undertaken these studies to gain an insight into this organization within the 5-kb XbaI region of the tcp gene cluster, and identify the predicted start of the operon, in both classical and E1 Tor strains.
RESULTS AND DISCUSSION
(a) Construction of tcpA, tcpP and tcpl promoter-cat fusions To analyse tcpA, tcpI and tcpP promoter activity the relevant regions were cloned into the promoter detection vector pPM3024 which contains the promoterless cat and galK genes (Williams and Manning, 1991). The CAT activity from strains harbouring the promoter fusions was compared. DNA upstream from tcpA and the intergenic region between tcpI and tcpP were subcloned into pPM3024 using pPM2114 (classical) and pPM3290 (El Tor), as sources of DNA (Fig. 1). In order to evaluate the effects of the Vc background, all of the plasmids were introduced into wild type and isogenic toxR mutant strains. An inter-
7~ pPM2114 0
1.0
2.0
3.0
4.0
5.0
5.027 kb pPM2635 pPM2639 pPM2640 pPM2641 pPM,?.642 pPM2643
--1 tcpl
tcpP tcpH
tcpA
tepB
tcpQf.cpC
P pPM2644 pPM2645 pPMP,646 pPM2647 pPM2648 pPM2649 pPM,?.650 pPM2651
pPM3290 0
1.0
2.0
3.0
4.0
5.0
5.023 kb
Fig. 1. Subclones of the promoter fragments. The various fragments were subcloned into the promoter detection vector pPM3024 (Williams and Manning, 1991) with respect to the promoterless cat gene as shown. The top part of the figure shows a restriction map of the 5-kb XbaI fragment cloned from Z17561 (classical, Inaba) present in plasmid pPM2114 (Faast et al., 1989). The subclones present in pPM3024 are shown. The bottom part of the figure shows the fragments of pPM3290 containing the 5-kb XbaI fragment from H1 (El Tor, Ogawa; Ogierman et al., 1995) which were cloned into pPM3024. A physical map of the genes contained within the 5-kb XbaI fragment and the direction of their transcription is shown in the centre (Ogierman et al., 1996) All of the plasmids were constructed in E. coli DH5 and DNA was extracted from these transformants and electroporated into Vc strains as described by Stoebner and Payne (1988). The toxR mutants, V796 and V798, derived from strains H1 and Z1756, respectively, have been previously described (Williams and Manning, 1991).
esting observation was made during these constructions: the toxR mutants invariably yielded at least a 10-fold higher level of transformants than their isogenic toxR + parents. The basis for this is unknown. No significant CAT activity was detected when all tcp promoter constructs were assessed in E. coli strain DH5 grown in either NB or CFB (data not shown), suggesting that activating
45 factors, present in Vc but not in E. coli, are required for tcp gene expression.
(b) PtcpA-activity in classical and El Tor strains TcpA (and TCP) is produced by most classical strains when grown at 30°C on CFA or in CFB but not at 37°C (Hall et al., 1988; Sharma et al., 1989a,b). CFB medium was therefore chosen to assess the activity of the putative classical tcpA promoter. CAT activity was detected with Z17561[pPM2638] and Z17561 [pPM2640] after growth at 30°C, indicating the presence of active promoters under these conditions. These same plasmids produced only background levels of CAT when assessed in the toxR mutant, V798, or at 37°C (Table I), implying that the promoters on these plasmids are not active at 37°C and require the presence of ToxR for expression. Since these are the same conditions required for expression of the TcpA protein in Z17561, it is likely that this is the promoter at which transcription of tcpA mRNA synthesis is initiated. The same inserts in opposite orientations gave no CAT activity (pPM2639 and pPM2641; Table I). E1 Tor strains have not been demonstrated to produce TCP when cultured in CFA or CFB media (Hall et al., 1988; Sharma et al., 1989a,b). However, TcpA and TCP production by some E1 Tor strains has been demon-
TABLE I CAT activity of promoter fusionsa Plasmid
Strain growth conditions Z17561 CFB 30C
pPM3024 pPM2638 pPM2639 pPM2640 pPM2641 pPM2642 pPM2643 pPM2644 pPM2645 pPM2646 pPM2647 pPM2648 pPM2649 pPM2650 pPM2651
++++ +++
37°C
V798 CFB
H1 AKI
30°C
30°C
V796 AKI 37°C
30°C
-
strated using AKI medium with a 4-h static culture followed by vigorous shaking overnight at 30°C (Jonson et al., 1990; Attridge et al., 1993; Voss and Attridge, 1993). These conditions were consequently adopted to detect PtcpA activity in the E1 Tor strain H 1. Low CAT activity was detected from plasmids pPM2644 and pPM2646 (Table I) with no significant difference between 30°C and 37°C, but activity was diminished in the absence of ToxR. The E1 Tor PtcpA region on pPM2644 and the equivalent classical PtcpA region on pPM2638 were introduced into Vc strains Z17561 and H1. CAT activity was assessed after growth in CFB at 30°C (Table I). A high level of CAT was produced by both strain Z17561[pPM2644] and Z17561[pPM2638]; however, Hl[pPM2644] and H1[pPM2638] produced equivalent but lower levels. This suggests that the difference in PtcpA activity between the classical and E1 Tor constructs cannot be entirely due to differences at the nt level within these regions, but may be dependent on the level/quality of activating factors provided by the host cell.
(c) Detection of tcpA mRNA by Northern analysis The tcpA-cat constructs have indicated a higher level of PtcpA activity in the classical strain compared to E1 Tor. This was readily confirmed by Northern RNA analysis. Hybridization with the 1.4-kb SphI-HindIII probe encompassing tcpA revealed a transcript of approx. 800 nt in Z17561 but it was not detected in its isogenic tcpA mutant V779, nor in the environmental E1 Tor strain 1196-78 or the E1 Tor strain H1 (data not shown). Hybridization with the 5-kb XbaI probe revealed the same 800-nt transcript from Z17561 (and H1, weakly) and no additional transcripts were detected (data not shown). Thus, the tcpA mRNA is the major, persisting, transcriptional product originating within the 5-kb XbaI region. The differences in the ability to detect the tcpA mRNA with classical and E1 Tor strains is consistent with the results obtained from the tcpA-cat fusion studies.
-
+++
+
++++
+
++
++
+ +++
-
_
+ +
+ +
~Bacterial cultures were grown in CFB (Evans et al., 1979) or AKI (Iwanaga and Yamamoto, 1985) medium to an A65o of 0.8 as indicated. The cell lysates were then prepared essentially as described by Gross and Rappouli (1988), and the assay performed as described by Gorman et al. (1982).
(d) Primer extension of classical and El Tor tcpA mRNA Primer extension on RNA from classical and E1 Tot strains, using a tcpA-specific primer, produced readily detectable extension products only in the case of the classical strain (Fig. 2) consistent with the results of Northern analysis. However, prolonged exposure identified an identical fragment in the E1 Tor strain (data not shown). The largest extension product probably represents a full length extension of the tcpA mRNA and the shorter extension products may be due to premature dissociation of the reverse transcriptase. However, the presence of more than one 5' transcriptional start site cannot be ruled out, or alternatively, there may be processing of the mRNA. The 5' nt of the largest extension product corres-
46 ponds to a G residue and has been designated the + 1 site (Fig. 3A). Examination of the - 10 region reveals a sequence with homology to a cr7° promoter but the - 3 5 region shows poor homology to the consensus, with a putative c A M P - C R P binding site overlapping the - 3 5 region.
TGCA
(e) PtcpP and Ptcpl activities
J
11
g
!
i
Fig. 2. Primer extension, using a tcpA specific oligo primer, on RNA from Vc strains Z17561 (classical), H1 (El Tor), 1196-78 (El Tor, environmental isolate). Total cellular RNA was prepared from Z17561, V779 and 1196-78 following growth in CFB at 30°C and H1 following growth in AKI medium at 30°C. Total cellular RNA was prepared using the hot phenol method described by Aiba et al. (1981). Cells were grown
in CFB or AKI to an A650 of 0.8. 5 ml cells was centrifuged and the pellet was resuspended in 0.5 ml of lysis solution (0.02 mM Na-acetate pH 5.5/0.5% SDS/1 mM EDTA). This was extracted four times with hot (65°C) phenol (equilibrated with a solution containing 0.02 mM Tris.HC1 pH 7.6/0.02 mM KCI/0.01 mM MgCI2 pH 5.2), then ethanol precipitated. To remove contaminating DNA the precipitate was resuspended in water and incubated at 37°C for 20 min with DNase buffer (20 mM Tris-HCl pH 7.6/5 mM MgCI2)and 1 ~tl of DNase enzyme(10 u/~tl). This was re-extractedwith phenol, precipitated and dried in vacuo and resuspended in water. Primer extensions were performed as previously described (Williams and Manning, 1991). The T,G,C,A lanes are dideoxy sequencing reactions of pPM2114 DNA with the tcpA primer (PrA sequence). The arrow indicates the largest extension product seen only in the classical strain at this exposure. The shorter extension products may represent products of premature dissociation.
Since both T c p H and TcpI have been implicated in tcp expression, it was of interest to examine the activities of the promoters for these genes. Because tcpH appears to be translationally coupled to tcpP, this meant analysing PtcpI and PtcpP. Cell lysates of classical strains harbouring the tcpPand tcpl-cat constructs were assessed for CAT activity following growth in CFB at both 30 and 37°C (Table I). Classical strain Z17561[pPM2642] produced strong CAT activity at 30°C, which was ToxR-dependent. At 37°C, this promoter has very low activity. Thus, PtcpP has optimal activity a't 30°C in the presence of ToxR. Z17561 [ p P M 2 6 4 3 ] produced very low CAT activity at 30°C, and lysates from this strain and strain V798[pPM2643] were negative at 37°C. This construct should have detected activity of PtcpI, so these results indicate that either PtcpI is very weak or was not present within this region of DNA. Cell lysates of E1 Tor strains harbouring the putative PtcpP and PtcpI in pPM3024 were assessed for CAT activity following growth in A K I medium at 30°C or 37°C (Table I). Plasmid pPM2648 contains the putative PtcpP region in the same orientation as cat. H l [ p P M 2 6 4 8 ] exhibits CAT activity at 30°C, but this was reduced at 30°C in strain V796 and was absent at 37°C in strain H1. From these results it appears that the E1 Tor PtcpP has optimal activity at 30°C in the presence of ToxR, as was observed for the classical PtcpP. Plasmid pPM2650 contains a 1.45-kb region of D N A that includes the tcpI-tcpP intergenic region and the entire tcpP coding region. Any promoter present on pPM2648 should also be present on pPM2650. When cultured at 30°C, the level of CAT produced by H l [ p P M 2 6 5 0 ] was similar to that of H l [ p P M 2 6 4 8 ] . However, pPM2650 still mediates cat expression in strain V796 at 37°C. A second promoter may have been cloned on pPM2650 that is active at 37°C and is independent of ToxR, but this seems unlikely as no significant CAT activity was observed in D H 5 [ p P M 2 6 5 0 ] . The most likely interpretation is that the increased level of synthesis of TcpP, due to the high plasmid copy number, results in activation of its own expression in the absence of ToxR. Activity from the tcpI-cat fusion (pPM2649) was very low under all conditions examined, as observed with the
47
A CATTAGTTTAACTCTAAGTTTAAATGGTTATCAcGGAGTACTTCGTGATAATTAAAAAT~ATTATAAAATAATGATGTGAAAAATCAGCTTTTATCGTTTTAAATAGTATTTTTTTTC < tcpI +1 TTTAGGAAAATAAACTTATAAAGTTAAAAAAAAGCCCCAAACGGAAGGGGCAAAGTGTCACAGGAAAGATAATGTAACCAAGTTAATAGATATGGAATAGGCACTATAGGGGGAGTGCTA
1401
AACCGAATGAACTGTAAcGAATATTGCTTTCCGATAGCATGGTTTGCTGTTTTTTTTAATGTTATTTTATTTTTTTTAACAATTTAAATATTATGCAATCGAGTTCTcATTATCAACTGC
1641
AAAATTAGATTGCAAATAATTATATTAAAAAAAATAGACATAAAAAAATGGGTCGTTATGATTAAGAAAATGTAAAGTA~ tcpP >
1723
1521
B T A G T G T G T G A C G A G T A G G C G C TCC T C C A C A G T C A A A G T G A C TGAAAGTCATCTC TTCATC TTAC CC TAATGTTCGCAGTTGTATTGAGATCGTCAACAATATAACAAGTGTCAATTAATT
2926
GGC TATACAGC A T G C A C G A G A C G A A C A C TGTCAGTAC C G C A C C A G A T C C A C G T A G G T G G G T A T A G T G A T A A G A C G T C T TACCCAATTTTTATCGTCATTCATAATT TCGATCTCCACTCC
3046
G G A A A T A T T T T A A C G C A T T T T A T T T G C A T T A A A T T A C T T T A A A T C A T T TGAATTGAATAAGTTGGCC TT TTTTTAGTGTC TGATTTATTTTGTGCGTGAATGTTACTCGTGTTTCTTTCA -35 -I0 +i ATGCAAG TGTGTTATTAAAAA/d%TAAAAAAACACAGCAAAAAACTGACATCTGTCAATTGTAGGTGAC TTTGTGTGGTTAAATGTGCGTGTTGCTTACGTTATCTAAAAAAGACCAAGCG
3286
A C G C A T T TC C T TTAAAGAC A G T A A A A T G G T G G A G T T A C A T A A A T A T G tcpA >
3166
3333
Fig. 3. Nucleotide sequences. (A) Region between tcpI and tcpP. The + 1 position of the tcpI transcript is shown. No readily definable sequence corresponding to sigma factor consensus sequences could be identified (Helmann and Chamberlin, 1988). (B) Region between tcpH and tcpA. The end of tcpHand the beginning of tcpA and the positions of the primer used for primer-extensionare indicated. The + 1 position for the tcpA transcript and the corresponding promoter as determined (Fig. 2) are shown. A putative cAMP-CRP binding site (Ebright et al., 1984) is shown in bold overlapping the -35 region of the tcpA promoter.
classical strain. But, pPM2651 produced CAT activity at 30°C and 37°C and was independent of ToxR.
(f) Primer extension of classical and E! Tor tcpl mRNA Primer extension with a tcpl-specific primer resulted in products from both the classical strain Z17561 and the E1 Tor strain H - l , but not from the environmental E1 Tor strain 1196-78. From the nt sequence comparison of Z17561 and H1 it is apparent that extension products from the tcpI primer would result in products of different size, as was observed. The 5' nt of both transcripts is a G residue (Fig. 3B). Examination of the - 1 0 and - 3 5 regions does not reveal any homology to the ~7o consensus sequence, suggesting that this promoter is very weak, in line with the CAT assay data, or perhaps transcription from this promoter requires an alternative cy factor and/or other activating factors. In this regard, TcpI appears to be a chemotactic transducer and chemotaxisspecific cy factors have been identified in other bacteria (Helmann and Chamberlin, 1988; Harkey et al., 1994). The localization of the 5' end of tcpI m R N A indicates that the promoter is located in the tcpI-P intergenic region. The levels of tcpI extension products obtained in both Z17561 and H1 were similar, indicating that the tcpI m R N A m a y be transcribed at similar levels in the two biotypes. It is surprising that the tcpI-cat fusions gave no CAT activity although tcpI m R N A was detectable by primer extension, albeit weakly. We conclude that tcp! has a weak promoter under the conditions used here.
(g) Conclusions (1) Experimental evidence has indicated that transcription of an 800-nt m R N A transcript initiates at a ToxRdependent promoter 75 nt upstream from the tcpA start codon, and probably terminates or is processed at the terminator-like structure after tcpA. Sequence data (Ogierman et al., 1995) do not reveal the presence of a potential terminator after tcpH, and processing of a polycistronic tcpP-H-A m R N A to yield an 800-nt m R N A cannot be ruled out. (2) On the basis of the promoter fusion studies, PtcpP and PtcpI appear to be activated by TcpP. (3) The ToxR protein appears to play a role in activation of PtcpP, but this may be via ToxT since DiRita et al. (1991) have demonstrated activation of tcpI by ToxT in E. coli, and ToxT is regulated by ToxR. (4) The difference between the classical and E1 Tot strains in expression of tcpA appears to be due to differences in transcription of the tcpI and tcpP genes within the respective hosts although they seem to be expressed at comparable levels in the same host. (5) There are fundamental differences, between the two biotypes, at the level of tcpA gene transcription and expression which are modulated by different responses to environmental conditions.
ACKNOWLEDGEMENTS The authors thank the National Health and Medical Research Council of Australia, the Diarrhoeal Diseases
48 Global Vaccines Programme of WHO and the Clive and Vera Ramaciotti Foundations for their support.
REFERENCES Aiba, H., Adhya, S. and de Crombrugghe, B.: Evidence for two functional gal promoters in intact Escherichia coll. J. Biol. Chem. 256 (1981) 11905-11910. DiRita, V.J., Parsot, C., Jander G. and Mekalanos, J.J.: Regulatory cascade controls virulence in Vibrio cholerae. Proc. Natl. Acad. Sci. USA 88 (1991) 5403-5407. Ebright, R.H., Cossart, P., Gicquel-Sanzey, B. and Beckwith, J.: Mutations that alter the DNA sequence specificity of the catabolite activator protein of Escherichia coll. Nature 311 (1984) 232-235. Evans, D.G., Evans Jr., D.J., Clegg, S. and Pauley, J.A.: Purification and characterization of the CFA/I antigen of enterotoxigenic Escherichia coll. Infect. Immun. 25 (1979) 738-748. Gorman, C.M., Moffatt, L.F. and Howard, B.H.: Recombinant genomes which express chloramphenicol transacetylase in mammalian cells. Mol. Cell. Biol. 2 (1982) 1044 1051. Gross, R. and Rappuoli, R.: Positive regulation of pertussis toxin expression. Proc. Natl. Acad. Sci. USA 85 (1988) 3913-3917. Hall, R.H., Vial, P.A., Kaper, J.B., Mekalanos, J.J. and Levine, M.M.: Morphological studies on fimbriae expressed by Vibrio cholerae 01. Microb. Pathogen. 4 (1988) 257 265. Harkey, C.W., Everiss, K.D. and Peterson, K.M.: Vibrio cholerae toxincoregulated-pilus gene tcpl encodes a homolgy of methyl-accepting chemotaxis proteins. Infect. Immun. 62 (1994) 2669 2678. Helmann J.D. and Chamberlin, M.J.: Structure and function of bacterial sigma factors. Annu. Rev. Biochem. 57 (1988) 839-872. Herrington, D.A., Hall, R.H., Losonsky, G., Mekalanos, J.J., Taylor, R.K. and Levine, M.M.: Toxin, toxin-coregulated pili and the toxR regulon are essential for Vibrio cholerae pathogenesis in humans. J. Exp. Med. 168 (1988) 1487-1492. Iredell, J.R. and Manning, P.A.: Biotype-specific tcpA genes in Vibrio cholerae. FEMS Microbiol. Lett. 121 (1994a)47-54.
Iredell, J.R. and Manning, P.A.: The toxin-regulated pilus (TCP) of Vibrio cholerae O1: a model for type IV pilus biogenesis? Trends Microbiol. 2 (1994b) 187 192. Iwanaga, M. and Yamamoto, K.: New medium for the production of cholera toxin by Vibrio cholerae O1 biotype El Tor. J. Clin. Microbiol. 22 (1985) 405 408. Jonson, G., Svennerholm, A.-M. and Holmgren, J.: Expression of virulence factors by classical and El Tor Vibrio cholerae in vivo and in vitro. FEMS Microbiol. Ecol. 74 (1990) 221 228. Manning, P.A.: Surface-associated and soluble components of Vibrio cholerae involved in bacteria-host interactions. Curt. Top. Microbiol. Immunol. 192 (1994) 265-281. Ogierman, M.A., Voss, E., Meaney, C., Faast, R., Attridge, S.R. and Manning, P.A.: Comparison of the promoter proximal regions of the toxin-co-regulated tcp gene cluster in classical and El Tor strains of Vibrio cholerae O1. Gene (1996) in press. Sharma, D.P., Stroeher, U.H., Thomas, C.J., Manning, P.A. and Attridge, S.R.: The toxin coregulated pilus (TCP) of Vibrio cholerae: molecular cloning of genes involved in pilus biosynthesis and evaluation of TCP as a protective antigen in the infant mouse model. Microb. Pathogen. 7 (1989a) 437 448. Sharma, D.P., Thomas, C.J., Hall, R.H., Levine, M.M. and Attridge, S.R.: Significance of toxin-coregulated pili as protective antigens of Vibrio cholerae in the infant mouse model. Vaccine 7 (1989b) 451-456. Stoebner, J.A. and Payne, S.M.: Iron regulated hemolysin production and utilization of heme and hemoglobin by Vibrio cholerae. Infect. Immun. 56 (1988) 2891 2895. Taylor, R., Shaw, C., Peterson, K., Spears, P. and Mekalanos, J.: Safe, live Vibrio cholerae vaccines? Vaccine 6 (1988) 151-154. Taylor, R.K., Miller, V.L., Furlong, D.B. and Mekalanos, J.J.: The use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc. Natl. Acad. Sci. USA 84 (1987) 2833-2837. Voss, E. and Attridge, S.R.: In vitro production of toxin-coregulated pili by Vibrio cholerae El Tor. Microb. Pathog. 15 (1993) 255-268. Williams, S.G. and Manning, P.A.: Transcription of the Vibrio cholerae haemolysin gene, hlyA, and cloning of a positive regulatory locus, hlyU. Mol. Microbiol. 5 (1991) 2031-2038.