Evolutionary and functional analyses of variants of the toxin-coregulated pilus protein TcpA from toxigenic Vibrio cholerae non-O1/non-O139 serogroup isolates

E. Fidelma Boyd1 and Matthew K. Waldor2

Department of Microbiology, National University of Ireland, University College Cork, Cork, Ireland1
Howard Hughes Medical Institute and Division of Geographic Medicine and Infectious Diseases, Tufts-New England Medical Center and Tufts University School of Medicine, 750 Washington Street, Boston, MA 02111, USA2

Author for correspondence: E. Fidelma Boyd. Tel: +353 21 4903624. Fax: +353 21 4903624. e-mail: f.boyd{at}ucc.ie


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The toxin-coregulated pilus (TCP) is a critical determinant of the pathogenicity of Vibrio cholerae. This bundle-forming pilus is an essential intestinal colonization factor and also serves as a receptor for CTX{phi}, the filamentous phage that encodes cholera toxin (CT). TCP is a polymer of repeating subunits of the major pilin protein TcpA and tcpA is found within the Vibrio pathogenicity island (VPI). In this study genetic variation at the tcpA locus in toxigenic isolates of V. cholerae was investigated and three novel TcpA sequences from V. cholerae strains V46, V52 and V54, belonging to serogroups O141, O37 and O8, respectively, were identified. These novel tcpA alleles grouped into three distinct clonal lineages. The polymorphisms in TcpA were predominantly located in the carboxyl region of TcpA in surface-exposed regions of TCP fibres. Comparison of the genetic diversity among V. cholerae isolates at the tcpA locus with that of aldA, another locus within the VPI, and mdh, a chromosomal locus, revealed that tcpA sequences are far more diverse than these other loci. Most likely, this diversity is a reflection of diversifying selection in adaptation to the host immune response or to CTX{phi} susceptibility. An assessment of the functional properties of the variant tcpA sequences in the non-O1 V. cholerae strains was carried out by analysing whether these strains could be infected by CTX{phi} and colonize the suckling mouse. Similar to El Tor strains of V. cholerae O1, in vitro CTX{phi} infection of these strains required the exogenous expression of toxT, suggesting that in these strains ToxT regulates TCP expression and that these TcpA variants can serve as CTX{phi} receptors. All the V. cholerae non-O1 serogroup isolates tested were capable of colonizing the suckling mouse small intestine, suggesting that the different TcpA variants could function as colonization factors.

Keywords: pathogenesis, intestinal colonization, CTXphi receptor

Abbreviations: CT, cholera toxin; TCP, toxin-coregulated pilus; VPI, Vibrio pathogenicity island

The GenBank accession numbers for the sequences reported in this paper are AY078355AY078358.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Cholera is a severe and sometimes lethal human diarrhoeal disease caused by the Gram-negative bacterium Vibrio cholerae. It remains a significant health problem in many parts of the world today. Of the nearly 200 recognized serogroups of V. cholerae, only two serogroups, O1 and O139, are associated with epidemic cholera (Albert et al., 1993 ). However, several V. cholerae non-O1/non-O139 serogroup isolates also encode all of the known V. cholerae virulence factors and have given rise to sporadic cholera outbreaks (Bik et al., 1995 ; Faruque et al., 1998 ; Ghosh et al., 1997 ). V. cholerae colonizes and multiplies within the small intestine after ingestion of contaminated food or water. Colonization requires the elaboration of a type IV bundle-forming pilus, the toxin-coregulated pilus (TCP) (Taylor et al., 1987 ), which has substantial sequence similarity to other type IV pili from a range of enteric bacteria (Shaw & Taylor, 1990 ). TCP-deficient mutants of V. cholerae are unable to colonize the suckling mouse small bowel as well as the human intestine (Herrington et al., 1988 ; Taylor et al., 1987 ). Within the intestine, V. cholerae secretes cholera toxin (CT), an A-B type exotoxin that elicits a secretory response in epithelial cells, resulting in the voluminous secretory diarrhoea characteristic of cholera.

The biogenesis and regulation of TCP production involves at least 15 genes encoded in the tcp gene cluster as well as several unlinked genes (Kaufman et al., 1993 ; Manning, 1997 ; Peek & Taylor, 1992 ). The tcp gene cluster is part of a large pathogenicity island termed the TCP-ACF element, or the Vibrio pathogenicity island (VPI) (Brown & Taylor, 1995 ; Karaolis et al., 1998 ; Kovach et al., 1996 ). The V. cholerae O1 serogroup is divided into two biotypes, classical and El Tor, on the basis of a number of phenotypic differences. At most loci, the two V. cholerae O1 biotypes are thought to be nearly identical (Beltran et al., 1999 ; Byun et al., 1999 ); however, the amino acid sequences of the major pilin protein TcpA are known to differ considerably between the El Tor and classical biotypes (75% similarity at the nucleotide level) (Iredell & Manning, 1997 ; Rhine & Taylor, 1994 ). Indeed, reports have identified several variant TcpA sequences among non-O1/non-O139 serogroup isolates (Ghosh et al., 1997 ; Nandi et al., 2000 ; Chakraborty et al., 2000 ; Mukhopadhyay et al., 2001 ). Furthermore, it is known that there are marked differences in in vitro growth conditions under which optimal expression of TCP occurs in each biotype (Manning, 1997 ). For example, using the most common in vitro growth conditions, El Tor strains do not produce TCP or CT. Expression of TCP and CT are co-ordinately regulated by ToxR, a transmembrane DNA-binding protein (Skorupski & Taylor, 1997 ). Direct activation of transcription of the genes encoding these virulence factors requires ToxT, a ToxR-regulated cytoplasmic transcription factor that is encoded within the VPI (Skorupski & Taylor, 1997 ).

Interestingly, TCP is also a receptor for CTX{phi}, the temperate filamentous phage that carries the CT genes (ctxAB) (Waldor & Mekalanos, 1996 ). Lysogenic conversion of non-toxigenic strains to toxigenicity by CTX{phi} infection appears to be a crucial step in the evolution of fully pathogenic V. cholerae. After infection, CTX{phi} can replicate as a plasmid or integrate at a specific attachment site, attRS, within the V. cholerae genome (Pearson et al., 1993 ; Waldor & Mekalanos, 1996 ). The O1 classical strains of V. cholerae do not produce CTX{phi}, although they produce CT and they contain CTX prophages integrated at two sites (Davis et al., 2000 ). It is believed that there are several similar steps in the pathways by which CTX{phi} infects V. cholerae and the F-pilus-specific filamentous phages, Ff phages, infect E. coli. Both phages are thought to first bind to a pilus receptor, the F pilus for Ff phages and TCP for CTX{phi} (Jacobson, 1972 ; Waldor & Mekalanos, 1996 ). Then, both types of phages require a second receptor, a complex of the periplasmic and inner-membrane proteins TolQ, TolR and TolA for infection to occur (Heilpern & Waldor, 2000 ; Sun & Webster, 1986 , 1987 ). For the Ff phages, a minor coat protein, pIII, located at one end of the phage particle, binds to both the F pilus and TolA (Riechmann & Holliger, 1997 ). The OrfU protein of CTX{phi} is thought to be functionally equivalent to pIII of Ff (Boyd et al., 2000 ; Heilpern & Waldor, 2000 ; Holliger & Riechmann, 1997 ; Waldor & Mekalanos, 1996 ). Therefore, CTX{phi} infection of V. cholerae is believed to involve OrfU binding to TcpA and to the V. cholerae orthologue of TolA (Heilpern & Waldor, 2000 ).

The discovery of variable CTX{phi} orfU sequences between the two O1 biotypes and among a number of toxigenic non-O1/non-O139 V. cholerae isolates (Boyd et al., 2000 ) prompted us to examine the tcpA gene in these isolates. Here we show the presence of highly variable tcpA sequences among a number of non-O1/non-O139 serogroup isolates. Functional studies with these strains revealed that the expression of the variant tcpA sequences was ToxT-dependent and that the variant TcpA proteins can act as receptors for CTX{phi}. Interestingly, all the non-O1/non-O139 serogroup isolates were capable of colonizing the suckling mouse intestine.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacterial strains.
The bacterial strains used in this study are shown in Table 1. Bacterial strains were stored at -70 °C in Luria–Bertani (LB) broth containing 20% glycerol. Antibiotics were used at the following concentrations: ampicillin (Ap) 100 µg ml-1, kanamycin (Kn) 50 µg ml-1 and streptomycin (Sm) 200 µg ml-1.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
PCR and nucleotide sequencing of tcpA and aldA genes.
We amplified by PCR the complete tcpA gene from several V. cholerae strains, including the classical biotype strains C1 and CA401, El Tor biotype strain SM115, O139 serogroup strain MO2, and the non-O1/non-O139 serogroup strains CO130, 151, 208, V46, V52 and V54 that all were previously shown to contain variable orfU sequences at the CTX{phi} locus (Table 1) (Boyd et al., 2000 ). The primer pair tcpH1 (5'-AGCCGCCTAGATAGTCTGTG-3') and tcpA4 (5'-TCGCCTCCAATAATCCGAC-3') was designed from previously published sequence (Heidelberg et al., 2000 ). A part of the aldA gene was PCR-amplified from the same set of strains using primer pair aldA1 and aldA2 as previously described (Boyd et al., 2000 ). The cycling conditions for PCR included an initial denaturation step at 96 °C for 1 min, followed by 30 cycles of 1 min of denaturation at 94 °C, 1 min of primer annealing at 52 °C (for tcpA) or 51 °C (for aldA) and 2 min of primer extension at 72 °C. The aldA and tcpA genes were sequenced directly from the products of PCR. Nucleotide sequencing of PCR products was performed with an Applied Biosystems model 373A automated DNA sequencing system using a DyeDeoxy terminator cycle sequencing kit. The sequences of both DNA strands were determined in all cases. Strains with polymorphic nucleotide and amino acids sites were sequenced at least twice. Sequences were assembled and edited using the MacVector program.

Sequence analysis.
Comparisons and phylogenetic analysis of the V. cholerae aldA and tcpA sequences were carried out using the MEGA program (S. Kumar, K. Tamura & M. Nei; http://wwwmegasoftware.net/). Multiple sequence alignment was performed on the tcpA gene sequences and the deduced amino acid sequences using CLUSTAL W (Higgins et al., 1996 ). Gene trees were constructed from pairwise comparisons by the neighbour-joining algorithm (Saitou & Nei, 1987 ). The predicted secondary structure of the TcpA protein was obtained from http://www.biochem.ucl.ac.uk/bsm/pdbsum/1qqz/main.html and used to map the position of polymorphic sites along the protein. The theoretical three-dimensional structures of TcpA and the TCP fibre (accession nos PDB ID 1QQZ and RCSB 001169, respectively) were analysed and coloured using RASMOL (Sayle & Milner-White, 1995 ; http://www.umass.edu/microbio/rasmol/).

CTX{phi} transduction assay V. cholerae strains with variant tcpA sequence.
Filtered cell-free supernatants from a strain [O395(pCTX-Kn)] producing a Kn-marked El Tor CTX{phi} (denoted here as CTXET-Kn{phi}) and O139 Calcutta CTX{phi} (denoted here as CTXCalc-Kn{phi}) were used to transduce various potential recipients to KnR using a previously described assay (Waldor & Mekalanos, 1996 ; Davis et al., 1999 ; Heilpern & Waldor, 2000 ). Briefly, 75 µl sterile supernatant from overnight cultures of O395(pCTXET-Kn) was mixed with 75 µl recipient cells, each of which harboured pMT5 and had been grown overnight at 30 °C in Ap and IPTG. This plasmid (pMT5) contains an IPTG-inducible toxT (DiRita et al., 1996 ). The phage and recipient cells were gently mixed for 30 min at room temperature. Appropriate dilutions of the mixture were then plated on LB agar containing Kn to enumerate the transductants. Also, dilutions of the recipient strains were plated on LB agar containing Sm to determine the total number of potential recipient bacteria. The frequency of infection was determined by dividing the number of transductants ml-1 (Knr c.f.u.) by the number of recipients ml-1 (Smr c.f.u.). All strains were tested using the same phage lysate.

Mouse colonization assay.
Five-day-old suckling CD1 mice were used to study the intestinal colonization properties of spontaneous Smr derivatives of strains encoding TcpA variants. In this single-strain inoculation assay, 1/100 dilutions of overnight cultures of individual test strains grown at 30 °C in LB broth were used to intragastrically inoculate suckling mice. After 24 h, the small intestines were removed and mechanically homogenized in 7 ml LB with a Tissue Tearor (Biospec Products). To enumerate the V. cholerae in the homogenates serial dilutions were plated on LB agar containing Sm. Four animals were used per group.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Analysis of tcpA and aldA sequence variation
We examined V. cholerae natural isolates, which were previously shown to contain the CTX prophage, for the presence of variant tcpA sequences. We sequenced the tcpA gene from nine V. cholerae strains chosen for their diversity at the CTX{phi} orfU locus. These included four O1 serogroup isolates (two classical and two El Tor biotype strains) and five non-O1 serogroup isolates. The total sample of V. cholerae isolates examined for sequence diversity at the tcpA locus included an additional four strains from the database: two O1 serogroup strains (O395 and N16961) and two non-O1 serogroup strains (151 and 208) (Heidelberg et al., 2000 ; Mekalanos et al., 1983 ; Novais et al., 1999 ; Rhine & Taylor, 1994 ). The total of seven non-O1 serogroup strains included three O37 serogroup strains and single isolates representing O8, O11, O139 and O141 serogroups. Among the 603 bp of the 13 tcpA sequences examined, there were 300 polymorphic sites that resulted in 86 aa replacements (Fig. 1). The classical biotype isolates, C1 and CA401, contained a tcpA sequence identical to the classical biotype tcpA sequence from strain O395 (Karaolis et al., 2001 ). V. cholerae strains SM115, C5, MO2 and CO130, representing two O1 El Tor, O139 and O37 serogroups, respectively, each contained a tcpA sequence identical to that from N16961, the El Tor biotype strain that was used for the determination of the V. cholerae genome sequence (Heidelberg et al., 2000 ) (Fig. 1). Among the V. cholerae non-O1/non-O139 serogroup isolates examined, three highly variable novel tcpA sequences were identified from non-O1/non-O139 serogroup strains V46, V52 and V54. Nucleotide sequence comparison of tcpA from V46, an O141 serogroup clinical isolate, showed it to be 21 and 23% divergent at the amino acid level compared to the classical and El Tor tcpA type sequences, respectively. V. cholerae strain V52, an O37 serogroup clinical isolate from Sudan, contained a tcpA sequence that showed 1 and 19% divergence at the amino acid level compared to those of the classical and El Tor biotypes. The most divergent tcpA sequence identified was found in V54, an O8 serogroup clinical isolate that showed 33 and 29% divergence at the amino acid level compared to the classical and El Tor biotype strains, respectively. Two strains, 151 and 208, that were previously shown to contain variant tcpA sequences were also included in this analysis (Novais et al., 1999 ). 151, an environmental isolate from Mexico, and 208, a clinical isolate from Thailand, both contained a tcpA sequence that showed 20 and 22% divergence at the amino acid level compared to the classical and El Tor biotype strains, respectively (Novais et al., 1999 ).



View larger version (42K):
[in this window]
[in a new window]
 
Fig. 1. Distribution of polymorphic amino acid sites in TcpA. Variant amino acids are indicated by single-letter abbreviations. Dots indicate amino acid identity and dashes represent gaps introduced to allow alignment of the TcpA sequences with the N16961 sequence. Numbers at the left are strain designations and those across the top are amino acid positions along the gene numbered from the start codon.

 
To further characterize the genetic variability within the VPI region, we compared the sequences of aldA from the same set of V. cholerae isolates examined at the tcpA locus. The aldA gene encodes aldehyde dehydrogenase and is located 14 kb upstream of the tcpA gene within the VPI. Of the 765 bp region of the aldA gene examined among 12 isolates, there were only 23 polymorphic sites. All of the variation found at this locus was accounted for by differences between epidemic isolates and the non-O1/non-O139 serogroup isolates and over half of this variation was attributable (14 polymorphic sites) to differences between strain V46 and the rest. These sequence comparisons indicate that the remarkable variability at the tcpA locus is not present throughout the VPI. Recent comparative sequence analysis by Karaolis et al. (2001) of the entire VPI region between a classical and an El Tor biotype strain tend to agree with our findings. They showed that most of the genetic variation occurred within the central tcp region, whereas the ends of VPI were virtually identical between the two biotypes (Karaolis et al., 2001 ). This is consistent with the hypothesis that the tcpA locus has a very different evolutionary history compared to the rest of the VPI and may have undergone multiple horizontal transfer and recombination events, generating the hypervariability at the tcpA locus.

Distribution of polymorphic sites along the TcpA protein
Some regions of the TcpA variants were conserved. In fact, there were no polymorphisms in the N-terminal 38 aa of these TcpA sequences. This part of the protein is thought to form an {alpha}-helix (H1in Fig. 2A) (http://www.biochem.ucl.ac.uk/bsm/pdbsum/1qqz/main.html). The last four predicted ß-sheets in TcpA also lacked polymorphisms (Fig. 2A). The latter four regions are all likely to be buried within the TcpA molecule (Fig. 2B). Certain amino acids were conserved in all of the TcpA sequences. Several of these were shown to be required for pilin stability by mutational analysis by Taylor and colleagues (Kirn et al., 2000 ). For example, the cysteine residues at positions 120 and 186 are present in all the TcpA sequences. These two cysteine residues are thought to form a disulphide bond that is essential for pilin stability (Kirn et al., 2000 ). Similarly, the K at position 165, which is required for pilin stability (Kirn et al., 2000 ), is conserved in all the TcpA sequences. All the TcpA sequences also contained conserved proline residues at positions 58, 87, 99 (except strain V54) and 132 (Fig. 2A).



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 2. (A) Position of conserved and polymorphic amino acid sites along the predicted secondary structure of TcpA. The linear sequence of TcpA is numbered 1–197 with amino acids indicated by single-letter abbreviations. Above the linear sequence, the six predicted {alpha}-helices are represented by purple coil structures and thick purple arrows represent the six predicted ß strands. ß and {gamma} turns are also indicated on the linear structure. The green vertical arrows indicate polymorphic sites. (B) Predicted secondary folding of TcpA (http://www.biochem.ucl.ac.uk/bsm/pdbsum/1qqz/main.html). The conserved regions of the TcpA subunit are coloured purple.

 
Most of the polymorphisms in TcpA were present in the carboxyl region of the protein. This part of the protein is thought to be surface-exposed along the pilus fibre (Kirn et al., 2000 ). Clustered polymorphic sites occurred between amino acids 61 and 71, 89 and 96, and 133 and 180 (Fig. 2A). To better understand the structural implications of these polymorphisms in the pilin protein, we mapped the variable residues onto a theoretical three-dimensional structure of the TCP fibre developed by R. Chattopadhyaya & A. C. Ghose [accession no. 1QQZ in the Protein Data Bank (http://www.rcsb.org/pdb/)] (Fig. 3). Each pilin subunit, which makes up the TCP fibre, is a different colour (TcpA subunit 1 is dark blue and TcpA subunit 5 is gold). The red subunit represents the first subunit on the second turn of the helix. Polymorphic amino acid sites are shown in green, and as Nandi et al. (2000) recently suggested, this modelling indicates that most of the polymorphic residues are located along the surface of the fibre (Fig. 3). It is likely that these are the regions most accessible to the products of the host immune response as well as to CTX{phi} and perhaps to additional bacteriophages as well.



View larger version (80K):
[in this window]
[in a new window]
 
Fig. 3. Positions of variable amino acids on the theoretical three-dimensional structure of the polymeric structure of TCP developed by R. Chattopadhyaya & A. C. Ghose [accession no. 1QQZ in the Protein Data Bank (http://www.rcsb.org/pdb/)]. Each TcpA subunit is a different colour. The subunits are arranged in a pentameric right-handed helix starting with TcpA subunit 1 (dark blue) and finishing with TcpA subunit 5 (gold). TcpA subunit 6 (red) represents the first monomer in the second turn of the helix. Polymorphic amino acids sites occurring in more that one strain are coloured green on each TcpA subunit.

 
Analysis of the genetic relationships among tcpA alleles
Phylogenetic analysis of the tcpA locus based on pairwise comparisons of the deduced amino acid sequences revealed five distinct lineages designated I–V (Fig. 4A). V. cholerae strain V54, an O8 serogroup clinical isolate from Thailand, had the most divergent TcpA sequence and is the sole representative of lineage V (Fig. 4A). The two V. cholerae classical biotype isolates, CA401 and C1, contained a TcpA sequence identical to that from O395, a classical biotype isolate from the database (Karaolis et al., 2001 ), and are grouped in lineage I along with V52, an O37 serogroup clinical isolate. This strain was responsible for a large outbreak of cholera in Sudan in 1968, one of the largest outbreaks caused by a non-O1 strain other than O139 Bengal (Zinnaka & Carpenter, 1972 ). Multilocus enzyme electrophoresis, which assays genetic variation at a range of chromosomal loci, and IS1004 fingerprinting have shown that the O37 Sudan strains are closely related to the epidemic O1 isolates (Beltran et al., 1999 ; Bik et al., 1996 ). We have shown near identity at the mdh locus between V52 and epidemic O1 serogroup isolates, which suggests that, similar to epidemic O139 strains, toxigenic O37 strains may have arisen by horizontal gene transfer of the O37 rfb cluster into an O1 progenitor cell. This hypothesis is further supported by the finding that the O37 Sudan strains lack most of the O1 cell wall polysaccharide genes similar to the O139 Bengal strain (Bik et al., 1996 ).



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 4. (A) Phylogenetic tree for predicted TcpA sequences from 13 V. cholerae strains constructed by the neighbour-joining algorithm based on the gamma distance (alpha=2). Strain, serogroup and biotype are listed. Seven TcpA alleles cluster into five major groups I–V. Classical and El Tor biotype strains are indicated in bold type. (B) Phylogenetic tree for predicted OrfU sequences from 12 V. cholerae strains constructed by the neighbour-joining algorithm based on the gamma distance (alpha=2).

 
Two non-O1/non-O139 serogroup strains clustered in lineage II: strains 151 and 208. V46, a clinical isolate, was the sole representative of lineage III. Lineage IV contained five strains with an identical TcpA sequence: three El Tor biotype strains, N16961, SM115 and C5, an O139 serogroup strain, MO2, and an O37 serogroup strain, CO130. The CO130 O37 serogroup strain is an environmental isolate recovered in India in 1993 whose mdh sequence is very similar to V52 and epidemic O1 V. cholerae isolates (Fig. 5). If horizontal transfer of the 037 rfb cluster into O1 progenitor cells accounts for the origin of toxigenic O37 isolates, then our detection of classical and El Tor-related TcpA sequences in O37 isolates suggests that both classical and El Tor O1 strains served as progenitors of toxigenic O37 strains.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 5. Genetic relationships among V. cholerae strains based on synonymous site variation in the housekeeping gene malate dehydrogenase (mdh) constructed by the neighbour-joining algorithm (Boyd et al., 2000 ). TcpA alleles identified for each strain are shown bold type and serogroup designations are shown in parentheses.

 
Similar to the variability seen at the TcpA locus among classical and El Tor isolates, we have recently shown that the orfU gene also shows a high degree of sequence variability between the two O1 biotypes (Boyd et al., 2000 ). Comparison of the gene tree based on the TcpA sequence with that based on the OrfU sequence shows a similar topology among the same set of strains. Comparable to the TcpA gene tree, the OrfU gene tree groups the classical and El Tor V. cholerae strains into two distinct lineages (Fig. 4B). However, this correspondence in the clustering pattern of epidemic strains at the TcpA and OrfU loci does not extend to non-O1/non-O139 serogroup strains. Previously, we found that the orfU gene contains both variable and invariable regions, whose locations correlate with their role in TcpA binding, hence the two hypervariable regions within OrfU probably interact directly with TcpA (Boyd et al., 2000 ). It is tempting to speculate that the interbiotype polymorphism in OrfU may reflect selective pressures to change with TcpA, its receptor.

Next, we carried out comparative sequence analysis on all tcpA sequences presently available in the database. An additional 10 strains were examined to determine whether they clustered in similar phylogenetic groups based on tcpA sequences. In all cases the epidemic V. cholerae isolates clustered within the two previously identified lineages I and IV, whereas most of the non-O1/non-O139 serogroup isolates clustered within the previously identified lineages II, III and V (Fig. 6). Furthermore, the non-O1/non-O139 serogroup isolates did not form distinct lineages based on their source of isolation (environmental versus clinical) or serogroup designation.



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 6. Phylogenetic relationships among predicted TcpA sequences from 23 V. cholerae strains constructed by the neighbour-joining algorithm based on the gamma distance (alpha=2). Strain, serogroup and source of isolation are listed. GenBank accession numbers: strain Z17561, M33514 (Faast et al., 1989 ); strain HB84419, AY052830; strain XJ90006, AY056619; strain GX95065, AY056618; strain SD95001, AY052831; strain SCE188, AF208385 (Mukhopadhyay et al., 2001 ); strain 10259, AF139626 (Nandi et al., 2000 ); strain VCE22, AF414371; strain C6709, U09807 (Rhine & Taylor, 1994 ); strain H1, X74730 (Ogierman et al., 1996 ).

 
To exemplify further the hypervariability at the tcpA locus, we compared the distribution of the TcpA alleles among the clonal lineages of V. cholerae based on analysis of the housekeeping gene mdh (Fig. 5). This locus was shown previously to be a reliable estimate of genotypic divergence of the bacterial genome as a whole (Boyd et al., 1994 , 1996 , 2000 ; Byun et al., 1999 ). Two observations are evident from this comparison. First, the overall genetic diversity found at the tcpA locus is much greater than that at the mdh locus or several other V. cholerae chromosomal loci (Table 2). Second, the deduced clonal relationships based on the mdh and tcpA sequences are incongruent. That is, closely related strains on the mdh tree have divergent tcpA alleles (Figs 4A and 5). Indeed the level of sequence divergence among epidemic strains ranged from 0% for mdh to 25% for tcpA in the same set of strains. Comparisons of the level of genetic variation at the tcpA locus with that found at the major pilin loci from E. coli gives additional evidence that some form of diversifying selection is acting at this locus (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2. Nucleotide sequence polymorphism for major pilin genes (in bold), housekeeping genes and VPI genes from natural isolates of V. cholerae and E. coli

 
It is of interest to note that several major pilin proteins were shown to be hypervariable compared to housekeeping genes in E. coli (Table 2). This degree of sequence polymorphism reflects gene products that are subject to diversifying selection in adaptation to host defence systems or other variable aspects of the environment (Li et al., 1995 ; Boyd & Hartl, 1998 ; Blank et al., 2000 ). Consistent with this model, it has been demonstrated that there are biotype-specific epitopes of TCP located in the C-terminal region of TCP (Sun et al., 1991 ). This region is immunodominant and can engender biotype-specific protective immune responses (Sun et al., 1991 ; Voss et al., 1996 ). Hence, it seems reasonable to propose that the apparent diversifying selection acting at TcpA reflects the selective pressure exerted by the host immune system. The location of most of the polymorphic amino acid sites along the exterior of the theoretical three-dimensional structure of the TcpA fibre supports this hypothesis (Fig. 3).

Analysis of CTX{phi} infection of V. cholerae strains that contain variant TcpA sequence
A transduction assay was used to test whether TcpA variant strains 151, 208, V46, V52 and V54 could be infected by CTX{phi}. The test strains were examined relative to two control strains: LAC1, a lacZ mutant derivative of O395, and TCP2, a TCP-deficient mutant of O395 (Herrington et al., 1988 ; Waldor et al., 1994 ). Cell-free supernatants containing two different Kn-marked CTX{phi}s, CTXET-Kn{phi} and CTXCalc-Kn{phi} (Waldor & Mekalanos, 1996 ; Davis et al., 1999 ), were used to transduce recipient strains to Knr. CTXclass{phi} was not used in this assay since the CTX prophage arrangement in classical V. cholerae isolates does not yield extrachromosomal CTX DNA and thus does not yield virions. To help ensure production of TCP for these in vitro transduction assays, pMT5, a plasmid containing toxT under the control of an IPTG-inducible promoter, was introduced into all the recipient strains. When strains were grown in the absence of IPTG, none of the five test strains yielded Knr transductants with either phage (data not shown). However, after growth in the presence of IPTG, Knr transductants were found for all recipients with the exception of TCP2, the TCP-deficient mutant (Table 3). CTXET-Kn{phi} infected strains 151 and V46 with 4–10-fold lower frequencies than LAC1, the positive control, whereas strains V52, 208 and V54 were infected three to five orders of magnitude less efficiently than LAC1. The frequency of CTXCalc-Kn{phi} infection of 151, V46, V52 and V54 was similar to that of CTXET-Kn{phi} with the notable exception of strain LAC1, which showed a threefold decrease in CTXCalc-Kn{phi} uptake. Furthermore, 208 did not produce any Knr colonies upon repeated assays with CTXCalc-Kn{phi} (Table 3). The mechanisms accounting for the difference in infection frequency between the different strains are unknown. Some of the differences may be attributable to differing capacities of the variant TCPs to act as CTX{phi} receptors. Also, phage immunity and heteroimmunity may account for some of the differences (Davis et al., 1999 ; Kimsey & Waldor, 1998 ). Overall, these results indicate that in these non-O1 strains, ToxT can augment TCP production and that to varying degrees all of the variant TCPs can serve as receptors for CTX{phi}.


View this table:
[in this window]
[in a new window]
 
Table 3. Efficiency of CTXET-Kn{phi} infection in V. cholerae TcpA variant strains

 
Examination of intestinal colonization of TcpA variant strains
With the exception of V. cholerae O139 serogroup isolates, the intestinal colonization properties of non-O1 serogroup isolates have not been extensively studied. Similarly, outside of a limited number of O1 classical and El Tor biotype strains the colonization ability of strains with variant TcpA sequences has not been examined. To determine whether V. cholerae non-O1/non-O139 serogroup strains with variant TcpA alleles could colonize the infant mouse intestine, they were intragastrically inoculated into suckling mice. Twenty-four hours later, the number of V. cholerae cells in intestinal homogenates was determined. In these assays, strains 151, V46, V52 and V54 colonized the infant mouse intestine approximately as well as the classical biotype strain O395 (Table 4). In contrast, strain 208 exhibited deficient colonization in this assay; however, this strain was not as attenuated as TCP2 for colonization, which lacks TCP (Table 4). Thus, these non-epidemic V. cholerae isolates are capable of colonizing the suckling mouse intestine. Even though these strains are not isogenic, given the requirement for TcpA for intestinal colonization of El Tor and classical O1 strains, these findings strongly suggest that the variant TCPs can function as colonization factors.


View this table:
[in this window]
[in a new window]
 
Table 4. Colonization of the suckling mouse by V. cholerae TcpA variant strains

 
Conclusions
Like the sequence variability observed in type IV pili of several bacterial species, including E. coli (Blank et al., 2000 ), Dichelobacter nodosus, Eikenella corrodens, Neisseria species, Moraxella bovis and Pseudomonas aeruginosa (reviewed by Tennent, 1994 ), we found extensive sequence variability in TcpA, the major subunit of the V. cholerae type IV pilus TCP (Fig. 6). Since it is improbable that mutation alone resulted in the high level of sequence divergence at the tcpA locus while other loci within the same set of isolates show little or no sequence divergence (Fig. 5), our data strongly suggest that the tcpA locus has undergone frequent horizontal transfer and recombination.

To date only two V. cholerae serogroups O1 and O139 are known to cause epidemic cholera, whereas all other toxigenic V. cholerae isolates are only associated with sporadic cholera outbreaks. The lack of epidemic spread among the toxigenic non-O1/non-O139 serogroup isolates may result from a superior intestinal colonization ability of O1 and O139 serogroup strains, which needs to be examined. Furthermore, the pathogenesis of V. cholerae is not completely understood and there are undoubtedly factor(s) involved in V. cholerae persistence during inter-epidemic periods, virulence and spread that are not yet characterized and that are absent in non-epidemic isolates.


   ACKNOWLEDGEMENTS
 
We thank William S. S. Jermyn, Anne-Marie Quirke, Yvonne O’Shea and Andrew J. Heilpern for excellent technical assistance. We are grateful to F. Mooi, G. B. Nair, Y. Takeda and L. Campos for generous donations of bacterial isolates. This work was supported by an Enterprise Ireland Basic Research grant, a Higher Education Authority grant and a UCC/Food Industry Partnership Board grant to E.F.B. M.K.W. was supported by the Howard Hughes Medical Institute and NIH grant AI-42347. M.K.W. is a PEW Scholar of Biomedical Research.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Albert, M. J., Siddique, A. K., Islam, M. S., Faruque, A. S., Ansaruzzaman, M., Faruque, S. M. & Sack, R. B. (1993). Large outbreak of clinical cholera due to Vibrio cholerae non-O1 in Bangladesh. Lancet 341, 704.[Medline]

Beltran, P., Delgado, G., Navarro, A., Trujillo, F., Selander, R. K. & Cravioto, A. (1999). Genetic diversity and population structure of Vibrio cholerae. J Clin Microbiol 37, 581-590.[Abstract/Free Full Text]

Bik, E. M., Bunschoten, A. E., Gouw, R. D. & Mooi, F. R. (1995). Genesis of the novel epidemic Vibrio cholerae O139 strain: evidence for horizontal transfer of genes involved in polysaccharide synthesis. EMBO J 14, 209-216.[Abstract]

Bik, E. M., Gouw, R. D. & Mooi, F. R. (1996). DNA fingerprinting of Vibrio cholerae strains with a novel insertion sequence element: a tool to identify epidemic strains. J Clin Microbiol 34, 1453-1461.[Abstract]

Blank, T. E., Zhong, H., Bell, A. L., Whittam, T. S. & Donnenberg, M. S. (2000). Molecular variation among type IV pilin (bfpA) genes from diverse enteropathogenic Escherichia coli strains. Infect Immun 68, 7028-7038.[Abstract/Free Full Text]

Bockemuhl, J. & Meinicke, D. (1976). Value of phage typing of Vibrio cholerae biotype El Tor in West Africa. Bull W H O 54, 187-192.[Medline]

Boyd, E. F. & Hartl, D. L. (1998). Diversifying selection governs nucleotide polymorphism in fimbriae pilin genes of Escherichia coli. J Mol Evol 47, 258-267.[Medline]

Boyd, E. F., Nelson, K., Wang, F.-S., Whittam, T. S. & Selander, R. K. (1994). Molecular genetic basis of allelic polymorphism in malate dehydrogenase (mdh) in natural populations of Escherichia coli and Salmonella enterica. Proc Natl Acad Sci USA 91, 1280-1284.[Abstract]

Boyd, E. F., Wang, F.-S., Whittam, T. S. & Selander, R. K. (1996). Molecular genetic relationships of the salmonellae. Appl Environ Microbiol 62, 804-808.[Abstract]

Boyd, E. F., Heilpern, A. J. & Waldor, M. K. (2000). Molecular analysis of a putative CTX{phi} precursor and evidence for independent acquisition of distinct CTX{phi}s by toxigenic Vibrio cholerae. J Bacteriol 182, 5530-5538.[Abstract/Free Full Text]

Brown, R. C. & Taylor, R. K. (1995). Organization of tcp, acf and toxT genes within a ToxT-dependent operon. Mol Microbiol 16, 425-439.[Medline]

Byun, R., Elbourne, L. D., Lan, R. & Reeves, P. R. (1999). Evolutionary relationships of pathogenic clones of Vibrio cholerae by sequence analysis of four housekeeping genes. Infect Immun 67, 1116-1124.[Abstract/Free Full Text]

Chakraborty, S., Mukhopadhyay, A. K., Bhadra, R. K. & 8 other authors (2000). Virulence genes in environmental strains of Vibrio cholerae. Appl Environ Microbiol 66, 4022–4028.[Abstract/Free Full Text]

Davis, B. M., Kimsey, H. H., Chang, W. & Waldor, M. K. (1999). The Vibrio cholerae O139 Calcutta CTX{phi} is infectious and encodes a novel repressor. J Bacteriol 181, 6779-6787.[Abstract/Free Full Text]

Davis, B. M., Moyer, K. E., Boyd, E. F. & Waldor, M. K. (2000). CTX prophages in classical biotype Vibrio cholerae: functional phage genes but dysfunctional phage genomes. J Bacteriol 182, 6992-6998.[Abstract/Free Full Text]

DiRita, V. J., Neely, M., Taylor, R. K. & Bruss, P. M. (1996). Differential expression of the ToxR regulon in classical and E1 Tor biotypes of Vibrio cholerae is due to biotype-specific control over toxT expression. Proc Natl Acad Sci USA 93, 7991-7995.[Abstract/Free Full Text]

Dubose, F., Dykhuizen, D. E. & Hartl, D. L. (1988). Genetic exchange among natural isolates of bacteria: recombination within the phoA gene of Escherichia coli. Proc Natl Acad Sci USA 85, 7036-7040.[Abstract]

Faast, R., Ogierman, M. A., Stroeher, U. H. & Manning, P. A. (1989). Nucleotide sequence of the structural gene, tcpA, for a major pilin subunit of Vibrio cholerae. Gene 85, 227-231.[Medline]

Faruque, S. M., Albert, M. J. & Mekalanos, J. J. (1998). Epidemiology, genetics, and ecology of toxigenic Vibrio cholerae. Microbiol Mol Biol Rev 62, 1301-1314.[Abstract/Free Full Text]

Ghosh, C., Nandy, R. K., Dasgupta, S. K., Nair, G. B., Hall, R. H. & Ghose, A. C. (1997). A search for cholera toxin (CT), toxin coregulated pilus (TCP), the regulatory element ToxR and other virulence factors in non-O1/non-O139 Vibrio cholerae. Microb Pathog 22, 199-208.[Medline]

Hanne, L. F. & Finkelstein, R. A. (1982). Characterization and distribution of the hemagglutinins produced by Vibrio cholerae. Infect Immun 36, 209-214.[Medline]

Heidelberg, J. F., Eisen, J. A., Nelson, W. C. & 23 other authors (2000). DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature 2000, 477–483.

Heilpern, A. J. & Waldor, M. K. (2000). CTX{phi} infection of Vibrio cholerae requires the tolQRA gene products. J Bacteriol 182, 1739-1747.[Abstract/Free Full Text]

Herrington, D. A., Hall, R. H., Losonsky, G., Mekalanos, J. J., Taylor, R. K. & Levine, M. M. (1988). Toxin, toxin-coregulated pili and the toxR regulon are essential for Vibrio cholerae pathogenesis in humans. J Exp Med 168, 1487-1492.[Abstract]

Higgins, D. G., Thompson, J. D. & Gibson, T. J. (1996). Using CLUSTAL for multiple sequence alignments. Methods Enzymol 266, 383-402.[Medline]

Holliger, P. & Riechmann, L. (1997). A conserved infection pathway for filamentous bacteriophages is suggested by the structure of the membrane penetration domain of the minor coat protein g3p from phage fd. Structure 5, 265-275.[Medline]

Iredell, J. R. & Manning, P. A. (1997). Translocation failure in a type-4 pilin operon: rfb and tcpT mutants in Vibrio cholerae. Gene 192, 71-77.[Medline]

Jacobson, A. (1972). Role of F pili in the penetration of bacteriophage fl. J Virol 10, 835-843.[Medline]

Karaolis, D. K., Lan, R. & Reeves, P. R. (1995). The sixth and seventh cholera pandemics are due to independent clones separately derived from environmental, nontoxigenic, non-O1 Vibrio cholerae. J Bacteriol 177, 3191-3198.[Abstract]

Karaolis, D. K., Johnson, J. A., Bailey, C. C., Boedeker, E. C., Kaper, J. B. & Reeves, P. R. (1998). A Vibrio cholerae pathogenicity island associated with epidemic and pandemic strains. Proc Natl Acad Sci USA 95, 3134-3139.[Abstract/Free Full Text]

Karaolis, D. K., Lan, R., Kaper, J. B. & Reeves, P. R. (2001). Comparison of Vibrio cholerae pathogenicity islands in sixth and seventh pandemic strains. Infect Immun 69, 1947-1952.[Abstract/Free Full Text]

Kaufman, M. R., Shaw, C. E., Jones, I. D. & Taylor, R. K. (1993). Biogenesis and regulation of the Vibrio cholerae toxin-coregulated pilus: analogies to other virulence factor secretory systems. Gene 126, 43-49.[Medline]

Kimsey, H. H. & Waldor, M. K. (1998). CTXphi immunity: application in the development of cholera vaccines. Proc Natl Acad Sci USA 95, 7035-7039.[Abstract/Free Full Text]

Kirn, T. J., Lafferty, M. J., Sandoe, C. M. & Taylor, R. K. (2000). Delineation of pilin domains required for bacterial association into microcolonies and intestinal colonization by Vibrio cholerae. Mol Microbiol 35, 896-910.[Medline]

Kovach, M. E., Shaffer, M. D. & Peterson, K. M. (1996). A putative integrase gene defines the distal end of a large cluster of ToxR-regulated colonization genes in Vibrio cholerae. Microbiology 142, 2165-2174.[Abstract]

Li, J., Ochman, H., Groisman, E. A., Boyd, E. F., Solomon, F., Nelson, K. & Selander, R. K. (1995). Relationship between evolutionary rate and cellular location among the Inv/Spa invasion proteins of Salmonella enterica. Proc Natl Acad Sci USA 92, 7252-7256.[Abstract]

Manning, P. A. (1997). The tcp gene cluster of Vibrio cholerae. Gene 192, 63-70.[Medline]

Mekalanos, J. J., Swartz, D. J., Pearson, G. D., Harford, N., Groyne, F. & de Wilde, M. (1983). Cholera toxin genes: nucleotide sequence, deletion analysis and vaccine development. Nature 306, 551-557.[Medline]

Mukhopadhyay, A. K., Chakraborty, S., Takeda, Y., Nair, G. B. & Berg, D. E. (2001). Characterization of VPI pathogenicity island and CTXphi prophage in environmental strains of Vibrio cholerae. J Bacteriol 183, 4737-4746.[Abstract/Free Full Text]

Nair, G. B., Bag, P. K., Shimada, T., Ramamurthy, T., Takeda, T., Yamamoto, S., Kurazono, H. & Takeda, Y. (1994). Characterization of phenotypic, serological, and toxigenic traits of Vibrio cholerae O139 bengal. J Clin Microbiol 32, 2775-2779.[Abstract]

Nandi, B., Nandy, R. K., Vicente, A. C. & Ghose, A. C. (2000). Molecular characterization of a new variant of toxin-coregulated pilus protein (TcpA) in a toxigenic non-O1/non-O139 strain of Vibrio cholerae. Infect Immun 68, 948-952.[Abstract/Free Full Text]

Nelson, K. & Selander, R. K. (1992). Evolutionary genetics of the proline permease gene (putP) and the control region of the proline utilization operon in populations of Salmonella and Escherichia coli. J Bacteriol 174, 6886-6895.[Abstract]

Nelson, K., Wang, F.-S., Boyd, E. F. & Selander, R. K. (1997). Size and sequence polymorphism in the isocitrate dehydrogenase kinase/phosphatase gene (aceK) and flanking regions in strains of Salmonella enterica and Escherichia coli. Genetics 147, 1509-1520.[Abstract/Free Full Text]

Novais, R. C., Coelho, A., Salles, C. A. & Vicente, A. C. (1999). Toxin-co-regulated pilus cluster in non-O1, non-toxigenic Vibrio cholerae: evidence of a third allele of pilin gene. FEMS Microbiol Lett 171, 49-55.[Medline]

Ogierman, M. A., Voss, E., Meaney, C., Faast, R., Attridge, S. R. & Manning, P. A. (1996). 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 170, 9-16.[Medline]

Pearson, G. D., Woods, A., Chiang, S. L. & Mekalanos, J. J. (1993). CTX genetic element encodes a site-specific recombination system and an intestinal colonization factor. Proc Natl Acad Sci USA 90, 3750-3754.[Abstract]

Peek, J. A. & Taylor, R. K. (1992). Characterization of a periplasmic thiol:disulfide interchange protein required for the functional maturation of secreted virulence factors of Vibrio cholerae. Proc Natl Acad Sci USA 89, 6210-6214.[Abstract]

Pupo, G. M., Karaolis, D. K. R., Lan, R. & Reeves, P. R. (1997). Evolutionary relationships among pathogenic and nonpathogenic Escherichia coli strains inferred from multilocus enzyme electrophoresis and mdh sequence studies. Infect Immun 65, 2685-2692.[Abstract]

Rhine, J. A. & Taylor, R. K. (1994). TcpA pilin sequences and colonization requirements for O1 and O139 Vibrio cholerae. Mol Microbiol 13, 1013-1020.[Medline]

Riechmann, L. & Holliger, P. (1997). The C-terminal domain of TolA is the coreceptor for filamentous phage infection of E. coli. Cell 90, 351-360.[Medline]

Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406-425.[Abstract]

Sayle, R. A. & Milner-White, E. J. (1995). RASMOL: biomolecular graphics for all. Trends Biochem Sci 20, 374-376.[Medline]

Shaw, C. E. & Taylor, R. K. (1990). Vibrio cholerae O395 tcpA pilin gene sequence and comparison of predicted protein structural features to those of type 4 pilins. Infect Immun 58, 3042-3049.[Medline]

Skorupski, K. & Taylor, R. K. (1997). Control of the ToxR virulence regulon in Vibrio cholerae by environmental stimuli. Mol Microbiol 25, 1003-1009.[Medline]

Sun, D. X., Seyer, J. M., Kovari, I., Sumrada, R. A. & Taylor, R. K. (1991). Localization of protective epitopes within the pilin subunit of the Vibrio cholerae toxin-coregulated pilus. Infect Immun 59, 114-118.[Medline]

Sun, T. P. & Webster, R. E. (1986). fII, a bacterial locus required for filamentous phage infection and its relation to colicin-tolerant tolA and tolB. J Bacteriol 165, 107-115.[Medline]

Sun, T. P. & Webster, R. E. (1987). Nucleotide sequence of a gene cluster involved in entry of E colicins and single-stranded DNA of infecting filamentous bacteriophages into Escherichia coli. J Bacteriol 169, 2667-2674.[Medline]

Taylor, R. K., Miller, V. L., Furlong, D. B. & Mekalanos, J. J. (1987). Use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc Natl Acad Sci USA 84, 2833-2837.[Abstract]

Tennent, J. M. (1994). Type 4 fimbriae. In Fimbriae: Adhesion, Genetics, Biogenesis, and Vaccines, pp. 127–146. Edited by P. Klemm. Boca Raton: CRC Press.

Voss, E., Manning, P. A. & Attridge, S. R. (1996). The toxin-coregulated pilus is a colonization factor and protective antigen of Vibrio cholerae El Tor. Microb Pathog 20, 141-153.[Medline]

Waldor, M. K. & Mekalanos, J. J. (1996). Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272, 1910-1914.[Abstract]

Waldor, M. K., Colwell, R. & Mekalanos, J. J. (1994). The Vibrio cholerae O139 serogroup antigen includes an O-antigen capsule and lipopolysaccharide virulence determinants. Proc Natl Acad Sci USA 91, 11388-11392.[Abstract/Free Full Text]

Zinnaka, Y. & Carpenter, C. C. (1972). An enterotoxin produced by noncholera vibrios. Johns Hopkins Med 131, 403-411.

Received 16 November 2001; revised 18 February 2002; accepted 25 February 2002.