TEXT

Top Abstract Text References

Cholera is a life-threatening diarrheal disease caused by the gram-negative bacterium Vibrio cholerae. V. cholerae enters humans orally through ingestion of contaminated food or water, swims toward and penetrates the mucus gel lining of the small intestine, and eventually adheres to the apical surface of the intestinal epithelial cells. V. cholerae bacteria respond to the specific environment encountered within the host intestine and express a number of virulence factors which allow the organisms to colonize the surface of the epithelial cells and cause disease; under normal in vitro growth conditions, virulence factors are not expressed (3). Several major virulence factors have been identified, most notably, cholera toxin (CT) and toxin coregulated pilus (TCP) (9, 16). However, it has become clear that other, unidentified factors are produced which aid in colonization and induce disease. Identification of such factors remains problematic, given that they are probably expressed only within a specific host environment. Although in vitro conditions have been developed which elicit V. cholerae CT and TCP production, these conditions are clearly not those encountered within the host (for example, CT and TCP are expressed optimally at 30°C in the Classical biotype). Thus, the identification of genes required for V. cholerae pathogenesis and the environmental signals that regulate their expression remains an active area of research.

RNA polymerase containing the alternate sigma factor ⁵⁴ (⁵⁴-holoenzyme) transcribes genes with diverse physiological roles in different bacteria, including pilin genes in pathogenic Neisseria and Pseudomonas spp., which are important for colonization by these organisms (8). ⁵⁴ is likewise required for some aspect of V. cholerae colonization. A V. cholerae strain containing a mutation in the gene encoding ⁵⁴ (rpoN) has a significant defect in the ability to colonize, but this defect is distinct from two other phenotypes associated with the rpoN strain, namely, nonmotility and low levels of glutamine synthetase expression (7). The rpoN strain also produces normal amounts of TCP under laboratory inducing conditions, so the nature of the ⁵⁴-dependent gene(s) which enhances colonization remain unknown. ⁵⁴-Holoenzyme has a requirement for an activator protein to initiate transcription; ⁵⁴-dependent transcriptional activators generally bind to DNA enhancer elements located within the promoter region and activate transcription by direct contact with ⁵⁴-holoenzyme (8, 14).

We have previously identified two ⁵⁴-dependent transcriptional activators, FlrA and FlrC, which are required for V. cholerae flagellar synthesis (7). However, strains containing mutations in either flrA or flrC do not demonstrate the severe colonization defect observed in a V. cholerae rpoN strain, thus implying that an additional ⁵⁴-dependent activator(s) is involved in colonization. We utilized PCR with degenerate oligonucleotides to identify six additional ⁵⁴-dependent transcriptional activators of V. cholerae. We have tentatively assigned roles in quorum sensing and nitrogen assimilation to two of these newly discovered ⁵⁴-dependent transcriptional activators, based on sequence and phenotype, but none appears to be involved in colonization. Our results suggest the presence of yet more unidentified ⁵⁴-dependent transcriptional activators in V. cholerae and demonstrate the involvement of ⁵⁴ in the expression of multiple sets of genes in this pathogen.

Identification of multiple V. cholerae ⁵⁴-dependent transcriptional activators. The V. cholerae rpoN mutant strain is defective in the ability to colonize a host and is, additionally, nonmotile (nonflagellated) and expresses low levels of glutamine synthetase (7); these phenotypes are distinct from the colonization defect. We thus predicted the presence of more ⁵⁴-dependent transcriptional activators in addition to the two flagellar activators FlrA and FlrC. All ⁵⁴-dependent activators are modular proteins, and they contain a highly conserved catalytic domain required for activation of transcription by ⁵⁴-holoenzyme (17). Kaufman and Nixon (5) demonstrated that ⁵⁴-dependent transcriptional activators could be isolated by PCR utilizing degenerate primers designed to recognize this conserved catalytic domain. We utilized their technique with identical primers recognizing the conserved amino acids (W/F)PGNV and ELFGH(V/A/D/E/G) and chromosomal DNA from V. cholerae Classical biotype strain O395 (11); the primers also incorporated restriction sites for EcoRI and BamHI. The PCR with Taq DNA polymerase consisted of 30 cycles of 92°C for 45 s, 42°C for 1 min, and 72°C for 1.5 min. The resulting fragments were first digested with EcoRI and BamHI and then ligated into pTZ19U (10) that had been similarly digested.

View larger version (70K): [in this window] [in a new window]   FIG. 1.   Homology of s54act gene products with the catalytic domains of 54-dependent transcriptional activators. Deduced amino acid sequences of PCR-derived fragments containing putative 54-dependent transcriptional activators from V. cholerae were aligned by GCG Pileup with the amino acid sequences of the 54-dependent activators FlrA (amino acids 154 to 302) and FlrC (amino acids 146 to 295) from V. cholerae and NtrC (amino acids 157 to 306) from S. typhimurium. Amino acids found in at least five activators are in capital letters, and amino acids common to all nine activators are designated by asterisks. Sequences recognized by degenerate primers used in PCR are designated by arrows.

s54act4 probably encodes an NtrC homologue. To determine the function of the s54act1 to s54act6 gene products, V. cholerae strains containing mutations in each gene were constructed. The internal fragments amplified by PCR and ligated into pTZ19U as described above were first digested with BamHI, made blunt ended with the Klenow fragment of DNA polymerase, digested with EcoRI, and then ligated into pGP704 (12) that had been digested with EcoRI and EcoRV. Since pGP704 requires the pir gene product for replication, mobilization of the recombinant plasmids containing internal s54act gene fragments into V. cholerae (which lacks pir) results in a chromosomal insertion into the s54act genes caused by integration of the plasmid through homologous recombination. These plasmids were mobilized into V. cholerae wild-type strain O395 (11) from E. coli SM10pir (12), as previously described (6), to form strains KKV216 (s54act1::pGP704), KKV217 (s54act2::pGP704), KKV218 (s54act3::pGP704), KKV219 (s54act4::pGP704), KKV221 (s54act5::pGP704), and KKV220 (s54act6::pGP704).

View larger version (53K): [in this window] [in a new window]   FIG. 2.   The s54act4 mutant displays a growth defect on glutamine-deficient medium that is complemented by expression of glnA. Strains were grown for 24 h on nutrient broth agar. The strains shown are KKV55 (rpoN [7]) and KKV219 (s54act4::pGP704), either without or with plasmid pKEK71 ("pglnA" [7]), which expresses S. typhimurium glnA from a 54-independent promoter.

View larger version (48K): [in this window] [in a new window]   FIG. 3.   s54act mutants are able to utilize succinate as a carbon source, unlike an rpoN mutant. Strains were grown for 48 h on minimal medium containing 10 mM ammonium, 2 mM glutamine, and 0.4% succinate. The strains shown are O395 (wild type [11]), KKV55 (rpoN [7]), KKV216 (s54act1), KKV217 (s54act2), KKV218 (s54act3), KKV219 (s54act4), KKV221 (s54act5), and KKV220 (s54act6).

None of the newly identified ⁵⁴-dependent activators is required for colonization. The ability of V. cholerae to colonize the intestines of infant mice is correlated with its ability to colonize the human intestine (15), and thus, the infant mouse model is a widely used model of V. cholerae virulence. The V. cholerae s54act mutant strains were tested for the ability to colonize infant mice by utilizing a competition assay previously described (7). Briefly, a mutant strain is coinoculated orally into infant mice along with an isogenic wild-type strain, and the amounts of mutant and wild-type strains which successfully colonize are determined after 24 h from intestinal homogenates. Any colonization defect of the mutant is reflected in a lower ratio of mutant to wild-type strains recovered from the intestine.

Nucleotide sequence accession numbers. The following sequences have been deposited in GenBank and assigned the corresponding accession numbers: s54act1, AF069055; s54act2, AF069056; s54act3, AF069387; s54act4, AF069388; s54act5, AF069389; s54act6, AF069390.