Characterization of the Small Untranslated RNA RyhB and Its Regulon in Vibrio cholerae

Bacterial strains and growth conditions.All strains are derivates of N16961, the V. cholerae El Tor genome strain (15), and are resistant to streptomycin; the fur insertion mutant is also resistant to ampicillin and was always grown in medium containing both antibiotics. Strains were cultured in Luria-Bertani (LB) or M63 minimal medium (4) containing antibiotics (200 μg/ml streptomycin with or without 100 μg/ml ampicillin); the iron chelator dipyridyl was added as indicated. Dipyridyl was added to cultures ∼1 h prior to harvesting of cells. Cultures were grown at 37°C.

Strain and plasmid construction.N16961 ryhB (strain BD1591) lacks 256 bp from the intergenic region between VC0106 and VC0107, which spans ryhB. This deletion begins ∼80 bp downstream of the upstream gene, VC0106, and ends ∼180 bp from the 3′ end of the divergently transcribed downstream gene; the deletion should not impair expression of these neighboring genes, and microarray analyses did not reveal changes in their transcript abundance. The deletion was generated by allele exchange using plasmid pBD1588, a pCVD442 derivative, as described previously (9). The insert in pBD1588 was constructed using splicing by overlap extension PCR (17) and the primers 5′GTTTTTAGCGATGGCGATTG3′ with 5′ATTGCTGGGACACTTGTTCCCTAAGACGTGG3′ and 5′GGAACAAGTGTCCCAGCAATGAGTGGACAAG3′ with 5′ATAAAGGTAAGAGGAGCCGTCG3′. Construction of N16961 hfq has been described previously (strain NHfq) (8). Strains with fur deletions (BD1803 and others) were generated by allele exchange using N16961 lac (NLAC) (8) and plasmid pBD1801. The insert in this pCVD442 derivative was constructed as described above using the primers 5′CGTCTAGACAGTGAGAAAATGCTGCCGC3′ with 5′CAGAGCGTAATTGGCTGCGTTGTCAAACCC3′ and 5′ACGCAGCCAATTACGCTCTGTAAACCGAACCAG3′ with 5′GCTCTAGAACCGCTAAACTGCTCAATGGC3′. Mutants were generated at a low frequency but could be obtained via selection of small colonies from sucrose plates. N16961 fur::Ap (strain BD1763) is an insertion mutant made by mating strain N16961 with SM10λpir(pCML13) (20). N16961 hfqH6 (strain BD1804) was made by allele exchange using plasmid pBD1800. This pCVD442 derivative was constructed as described above using primers 5′GCTCTAGATCAAGCCGTTGAAACGCTCTC3′ with 5′TTAGTGATGGTGATGGTGATGCTCTCCAGACTTCTCTGC3′ and 5′GAGCATCACCATCACCATCACTAATTCTTTGCACAATTA3′ with 5′GCTCTAGACCACGTAAACCAATCCCGC3′. Western blotting was used to screen for strains producing His₆-tagged Hfq. Mutations in all other strains were confirmed by Southern blotting and/or PCR.

Northern blot analyses.Wild-type (wt) V. cholerae and mutant V. cholerae were grown in LB medium to an optical density at 600 nm (OD₆₀₀) of ∼1 (late log phase) and then grown for an additional hour with or without 100 μM dipyridyl. Cultures were grown in M63 medium to an OD₆₀₀ of ∼0.25 (mid log phase) and then grown an additional hour with or without 25 μM dipyridyl. Cell pellets from cultures were stored prior to RNA isolation in the RNA stabilization reagent RNAlater (QIAGEN). RNA was isolated using an RNeasy Mini kit and RNase-free DNase set (QIAGEN). RNA isolated with TRIzol yielded equivalent results (not shown). RNA (6 μg) was electrophoresed on glyoxal agarose gels (Ambion) or on polyacrylamide urea gels (National Diagnostics) and transferred to BrightStar (Ambion) or HyBond (Amersham) membranes, respectively. RNA integrity and quantity was confirmed via staining of agarose gels with ethidium bromide prior to transfer. Blots were hybridized in ULTRAhyb (Ambion) to ³²P-labeled riboprobes generated by in vitro transcription. Hybridization was at 68°C for all probes except RyhB, which was hybridized at 64°C. The template used to create the ryhB riboprobe was a TOPO clone (Invitrogen) of a PCR product generated with primers 5′CGCGGATCCCGTCTTAGGGAACAAGTGAAGG3′ and 5′CCGGAATTCGCAAACGAGGTCAAAGCC3′; it contains almost all of ryhB. Other riboprobes were complementary to internal fragments of the target genes: nucleotides (nt) 214 to 477 of VC2045, nt 56 to 723 of VC1169, nt 166 to 895 of VC1255, nt 165 to 925 of VC2213, and nt 12999 to 13501 of VC1451. Blots were stripped and reprobed as recommended in Strip-EZ hybridization protocols (Ambion).

Purification of HfqH6-interacting RNAs.Cultures of N16961 hfqH6 (BD1804) and N16961 were grown in LB medium to an OD₆₀₀ of ∼0.8; 100 μM dipyridyl was then added, and cultures were grown for an additional 1.5 h. Cell pellets equivalent to 30 OD₆₀₀ units were frozen, and then cell extracts were prepared in lysis buffer (20 mM Tris [pH 8.0], 150 mM KCl, 1 mM MgCl₂, 5 mM β-mercaptoethanol) as described (40). His₆-tagged Hfq (HfqH6) was affinity purified on Ni-nitrilotriacetic acid (Ni-NTA) agarose (QIAGEN) from lysate generated from 7.5 OD equivalents. A control purification was done in parallel using lysate from N16961. The lysates were rocked at 4°C for 2 h, and then the Ni-NTA and “depleted” lysates were separated by centrifugation. The matrix was washed with lysis buffer containing 50 mM imidazole; adherent proteins were then eluted with lysis buffer containing 250 mM imidazole. Samples for Northern and Western analyses were generated from starting lysates, depleted lysates, 250 mM imidazole washes, and 250 mM imidazole pellets. Protein samples were prepared from 10% of each fraction, and HfqH6 purification was assessed on Western blots probed with anti His5 antibody (QIAGEN). RNA was extracted from the remaining 90% of each fraction with phenol-chloroform and precipitated with ethanol for further analyses. A total of 25% of each RNA sample was electrophoresed on an agarose gel and then blotted and probed for RyhB as described above.

RACE analysis.RNA was isolated from wt cells grown in LB-100 μM dipyridyl, polyadenylated using E. coli Poly(A) polymerase (Ambion), and reverse transcribed using the primer 5′TCACGACTCACTATAGGATCCTTTTTTTTTTTTN3′. cDNA was purified and tailed with poly(dC) using a 5′ rapid amplification of cDNA ends (RACE) system kit (Invitrogen). The RyhB 5′ end was amplified using the RACE kit AAP primer and 5′TACACTGGAAGCAATGTGAGC3′, and the 3′ end was amplified using 5′TCACGACTCACTATAGGATCC3′ and 5′CGAAACGGCCGAACTTGAGC3′. PCR products were TA cloned in pCRII (Invitrogen) and sequenced.

Sequence analyses.Homology searches were performed using the BLASTN 2.2.9, BLASTN 2.2.10, and BLASTN 2.0MP-WashU algorithms (http://www.ncbi.nlm.nih.gov/BLAST/ and http://tigrblast.tigr.org/cmr-blast/). Sequence alignments were generated using the bl2seq algorithm (http://www.ncbi.nlm.nih.gov/BLAST/) and MacVector (Accelrys). Promoter prediction was done with BPROM (Softberry, Inc., Mt. Kisco, N.Y.).

Microarray analyses.Paired cultures of wt and ryhB cells were grown in M63 to an OD₆₀₀ of ∼0.25 and then supplemented with 25 μM dipyridyl and grown an additional 1 to 1.5 h. Microarray analyses were performed as described (8). Genes were counted as differentially regulated when the ratio of mutant/wt transcripts was either >1.5 or <0.66 in three of three experiments.

Growth analyses.Cells were grown overnight in M63 medium, diluted in fresh medium, and grown to log phase and then rediluted in either M63 or M63 medium plus 25 μM dipyridyl for growth analyses. All cultures were grown in duplicate, and both members of each pair were plotted.

Intestinal colonization assays.In vivo competitions were performed as described (8). Bacteria were harvested from the suckling mouse small intestine ∼20 h after inoculation and plated on selective medium for quantitation of bacterial colonization.

Real-time RT-PCR.Quantitative reverse transcription-PCR (RT-PCR) analyses were performed basically as described (29). For analyses of ryhB expression in LB cultures, cells were grown to an OD₆₀₀ of ∼1 and then 100 μM dipyridyl was added to half of the cultures and cells were grown for an additional hour. For analysis of gene expression in the suckling mouse intestine, mice were infected with N16961 as described (8) except that mice were inoculated with 10× more bacteria. RNA was isolated from intestinal homogenates of infected mice by use of TRIzol LS (Invitrogen). Specific oligonucleotide primers (5′GAAGATGTCGCCTAGCAAACG3′) (29) were used for RT reactions, along with 2.0 μg of RNA from in vitro-grown N16961 or 12 μg of RNA from intestinal homogenates. PCRs contained 2.5 μl of cDNA and primers 5′GAAGATGTCGCCTAGCAAACG3′ and 5′GAGCAGGTTCTTTTTGACACG3′ in a total volume of 20 μl. Intestinal homogenates from five infected suckling mice and two independent cultures of N16961 grown in LB medium with or without dipyridyl were assayed.

V. cholerae encodes a ryhB homolog.Comparisons of the genomic sequences of E. coli and other gram-negative bacteria revealed that several additional species, including V. cholerae, are likely to encode a RyhB homolog (16, 23). We used Northern blotting to confirm that the putative ryhB locus in V. cholerae encodes an iron-regulated transcript. A transcript of ∼225 nt was detected in RNA from wt V. cholerae grown in LB medium-0.1 mM dipyridyl and in M63 minimal medium (which contains only 1.7 μM FeSO₄·7 H₂O) both with and without dipyridyl (Fig. 1A). This transcript was not detected when cells were grown in LB medium alone, which contains 17 μM iron (1). It also was not present in RNA from V. cholerae in which the putative ryhB locus was disrupted, and it was reduced in abundance in RNA from an hfq strain. Together, these results indicate that V. cholerae produces a RyhB homolog whose abundance in the cell is dependent upon iron levels and on Hfq.

• Open in new tab
• Download powerpoint

FIG. 1.

Northern blot analyses of ryhB expression and RyhB interactions in V. cholerae. A. RNA was isolated from wild-type N16961 (wt) and from ryhB and hfq derivatives of this strain, which were grown in the media indicated with (+D) or without dipyridyl. RNA was electrophoresed on denaturing acrylamide gels and transferred to membranes by electroblotting before hybridization to a RyhB probe. B. Effect of addition of FeSO₄·7 H₂O to cells in M63 medium. RNA was electrophoresed on an agarose gel. Total concentration of FeSO₄·7 H₂O is shown. C. Role of Fur in iron-regulated expression of ryhB. RNA from wt and fur cells grown in LB with or without 100 μM dipyridyl (Dip.) was electrophoresed on an agarose gel. D. RyhB interaction with Hfq. HfqH6 and associated RNAs were affinity purified on Ni-NTA agarose from lysate of an hfq::H6 strain (HfqH6); a control purification was performed in parallel using lysate from wt cells (wt). Protein purification was assessed on Western blots probed with an anti His5 antibody (upper panel), and RyhB purification was assessed on Northern blots of agarose gels (lower panel). Protein and RNA were extracted from total lysates (lanes 1 and 5), “depleted lysates” (lanes 2 and 6), washes containing 250 mM imidazole (lanes 3 and 7), and Ni-NTA agarose pellets after 250 mM imidazole wash (lanes 4 and 8).

RyhB transcript abundance was not notably increased when dipyridyl was added to M63, suggesting that limiting iron to 1.7 μM is sufficient to induce maximal expression (Fig. 1A and data not shown). To determine how much iron is sufficient to prevent ryhB expression, we grew the wt strain in M63 supplemented with various amounts of additional FeSO₄·7 H₂O and assessed RyhB transcript abundance. As little as 3 μM FeSO₄·7 H₂O was sufficient to reduce RyhB levels, and 6 μM FeSO₄·7 H₂O lowered its abundance to background levels (Fig. 1B). As V. cholerae requires 0.5 to 1 μM bioavailable iron for growth and survival (31), the range of conditions under which ryhB is expressed is apparently fairly narrow.

Many genes that are expressed in response to iron scarcity have been found to be regulated by the iron-dependent repressor Fur. In V. cholerae, Fur has been shown to control expression of a tonB operon, a heme transport operon, a ferrichrome utilization operon, ferrichrome and vibriobactin receptor genes, and other genes involved in iron utilization (26, 31, 37). Since we had found that ryhB is also expressed in response to iron scarcity, we assessed whether this regulation is dependent upon Fur. We used Northern blots to compare RyhB abundance in wt and fur V. cholerae grown in LB medium and LB-100 μM dipyridyl. Both fur deletion mutants, generated for this study, and fur::Ap insertion mutants, previously reported to be deficient in iron-dependent regulation of Fur targets, were tested. We found that ryhB is expressed in Δfur strains even when iron is available; unexpectedly, ryhB was repressed under this condition in the fur::Ap background (Fig. 1C). These results were observed for at least two independently derived mutants of each class (data not shown). They suggest that Fur regulates ryhB expression in V. cholerae and that Fur function is not fully abrogated in the insertion mutant, which lacks only a small portion of the protein's carboxyl terminus. Residual Fur activity in the insertion mutant could explain the small degree (∼2-fold repression) of iron-regulated expression of the vibriobactin, enterobactin, and heme receptor genes viuA, irgA, and hutA previously observed in this strain background (19, 37).

RyhB and Hfq interact in V. cholerae.The reduction in RyhB transcript abundance in the hfq mutant (Fig. 1A) is consistent with a model in which interaction between RyhB and Hfq stabilizes the sRNA. To assess directly whether RyhB-Hfq binding occurs, we determined whether an epitope-tagged Hfq (HfqH6) and RyhB could be copurified from cellular extracts. To maintain an appropriate cellular level of Hfq, sequence encoding the polyhistidine epitope tag was inserted on the V. cholerae chromosome at the 3′ end of the endogenous hfq. The resulting strain does not display the slight growth deficiency demonstrated by an hfq mutant, suggesting that the tag does not disrupt Hfq function. We have also found that a plasmid-encoded HfqH6 complements the growth deficiency of an hfq mutant, again suggesting that a tagged protein functions normally (data not shown). We affinity purified HfqH6 from cellular extracts of V. cholerae N16961 hfq::H6 grown in LB-dipyridyl and extracted copurifying RNA from the protein for analysis on Northern blots. ryhB transcripts were readily detectable among RNAs copurified with HfqH6 (Fig. 1D). RyhB was not detected in RNA samples purified in parallel from lysates of wt V. cholerae. These results indicate that there is a specific interaction between RyhB and Hfq in V. cholerae.

Mapping of RyhB transcript ends.The dominant V. cholerae RyhB transcript detected on Northern blots was significantly larger (∼225 nt versus 90 nt) than the homologous transcript from E. coli. To identify the additional sequences included in the V. cholerae transcript, we performed 5′ and 3′ RACE, using RNA from wt cells grown in the presence of an iron chelator. Sequencing the amplification products from these reactions revealed that the V. cholerae RyhB extends an additional ∼60 nucleotides at each end relative to its E. coli homolog. These results indicate that the V. cholerae RyhB is 215 nt long, which closely corresponds to the size observed on Northern blots. A stretch of approximately 60 nt within this sequence (nt 92 to 151) is also present in the genomes of other Vibrio species (e.g., V. vulnificus and V. parahaemolyticus). However, only the central region of the transcript (nt 102 to 130) is conserved among more distantly related species (Fig. 2A). The promoter prediction program BPROM detects a putative promoter upstream of the apparent transcriptional start site. This sequence overlaps with a potential Fur-binding site (Fig. 2A) that matches a consensus target at 12 of 19 bases, consistent with our observation that ryhB expression is repressed in an iron- and Fur-dependent manner (Fig. 1C).

• Open in new tab
• Download powerpoint

FIG. 2.

Boundaries, features, and potential base pairing for V. cholerae RyhB. A. Start and end sites of V. cholerae RyhB were determined by 5′ and 3′ RACE, respectively. Potential binding sites for RNA polymerase (−35 and −10) were identified using BPROM. The candidate Fur binding site (underlined) matches a consensus sequence at 12 of 19 positions. A ∼60-nt region is conserved in V. cholerae, V. vulnificus and V. parahaemolyticus. A ∼28-nt region (bold) is conserved among vibrios, Yersinia spp., E. coli, Salmonella spp., Photobacterium profundum, and other bacteria. B. Sequence complementarity between V. cholerae RyhB and the leader sequence of sodB. C. Sequence complementarity between V. cholerae RyhB and the 5′ end of sdhC. The first translated codon of sdhC is boxed.

Identification of genes regulated by RyhB.Two approaches were taken to identifying genes whose transcript abundance might be regulated by RyhB. As regulation in E. coli is thought to be dependent on base pairing between the sRNA and its target mRNAs (12, 22), we first searched the V. cholerae genome for sequences complementary to the ∼60-nt conserved core of RyhB. However, this analysis yielded only one sequence (other than ryhB itself) with an E value of less than 0.1: the putative 5′ untranslated region of sodB (VC2045), which had an E value of 0.053 in the BLASTN 2.2.9 analysis. A 16-nucleotide sequence beginning 60 bases upstream of the predicted sodB translational start site shows perfect complementarity to a region of RyhB that is conserved both among vibrios and other bacterial species (Fig. 2B). In E. coli, sodB encodes an Fe-containing superoxide dismutase previously shown to be regulated by RyhB. Interestingly, however, sequences corresponding to those that mediate SodB:RyhB RNA base pairing in E. coli (12) were not detected by the BLASTN analysis, although there is some complementarity between these regions. Instead, distinct sequences from both SodB and RyhB RNA were found to maximize potential base pairing between the two transcripts.

To identify additional candidate members of the RyhB regulon, we used DNA microarrays to compare gene expression in wt and ryhB V. cholerae. Both strains were grown in minimal medium (M63) supplemented with dipyridyl in order to maximize differential expression of RyhB-regulated genes. We identified 48 genes whose transcript abundance consistently differed between the wt and mutant strain by at least 1.5-fold under these growth conditions (Table 1). A number of the apparently up-regulated genes are known to encode iron-containing proteins, consistent with RyhB's role in other organisms. These genes include sodB (VC2045), sdhC (VC2091), a fumarate hydratase (VC1304), and gltB1 (VC2373). sodB, sdhC, and fumA (encoding fumarate hydratase) have also been found to be regulated by RyhB in E. coli, indicating that there is some conservation of the RyhB regulon among different organisms. However, no changes in transcript abundance were detected for homologs of other members of the E. coli regulon, such as acnA, ftn, and bfr. Instead, several genes encoding enzymes involved in amino acid synthesis and utilization were found to have increased transcript abundance in the V. cholerae ryhB mutant, including trpAB (VC1169/1170), hutU (VC1203), leuC (VC2492), and kbl (VCA0886). These genes and many others found to have increased transcript abundance in the V. cholerae ryhB mutant have no apparent connection to iron metabolism, though it is possible that a relationship remains to be detected. In addition, most of the genes found to have reduced transcript levels in the ryhB mutant have not been reported to encode iron-containing proteins. Down-regulation of target transcripts has not previously been reported for RyhB; thus, these genes, which lack any obvious common features, are a novel component of this regulon. However, it should be noted that genes identified in the microarray analysis may be affected indirectly or directly by RyhB and that differences in transcript abundance have only been confirmed for a subset of genes (see below).

Altered expression of several of the genes identified by microarray hybridization was confirmed using Northern blotting. Increased transcript levels for sodB, trpA, and ompA were detected in RNA from the ryhB mutant and from an hfq mutant as well, suggesting that the sRNA and Hfq act in conjunction to regulate expression of these genes (Fig. 3A-C). Northern blotting also confirmed that transcripts for nrdB and rtxA have reduced abundance in the ryhB mutant (Fig. 3D, E). Interestingly, deletion of hfq had a less-pronounced effect on the abundance of these transcripts. We were unable to detect RyhB-dependent changes in transcript abundance for VC2091 (sdhC), as expression levels were apparently too low to be detectable on Northern blots, perhaps because its promoter is not highly active when glucose is used as a carbon source (data not shown and reference 23). However, we did identify sequence near the 5′ end of VC2091 that may have the capacity to pair with RyhB (Fig. 2C), suggesting a possible means for regulating the abundance of VC2091 transcripts.

• Open in new tab
• Download powerpoint

FIG. 3.

Northern blot analyses of genes identified in microarray analyses as differentially expressed in wt and ryhB V. cholerae. RNA was isolated from wt, ryhB, and hfq V. cholerae grown in M63-dipyridyl. Blots were hybridized to riboprobes complementary to transcripts of VC2045 (A), VC1169 (B), VC2213 (C), VC1255 (D), and VC1451 (E).

Deletion of ryhB does not impair growth of V. cholerae in iron-poor media or in vivo.Since pathogenic V. cholerae may occupy environments in which iron is scarce and since RyhB is thought to contribute to efficient iron utilization, we assessed whether deletion of ryhB impaired growth of V. cholerae when the supply of iron was limited. Comparison of the growth of wt and ryhB V. cholerae in M63, which contains only 1.7 μM FeSO₄·7 H₂O and has been shown to induce RyhB expression (Fig. 1A), revealed that their levels of growth were equivalent under this condition (Fig. 4A). In addition, we found that after several (>3) hours of growth in M63-25 μM dipyridyl the ryhB mutant grew slightly more rapidly than the wt strain (Fig. 4B). Thus, disruption of RyhB-dependent gene regulation is clearly not detrimental to growth of V. cholerae under iron-limiting conditions.

• Open in new tab
• Download powerpoint

FIG. 4.

Comparative growth analyses of wt and ryhB V. cholerae in vitro and in the suckling mouse. Growth curves for wt and ryhB V. cholerae grown in M63 (A) and M63-25 μM dipyridyl (B) are shown. Cultures for each strain were grown in duplicate; both measurements were plotted. C. Competitive indices for wt and ryhB V. cholerae grown in the suckling mouse (in vivo) and in LB medium (in vitro). The competitive index is defined as follows: (mutant_ouput/wt_output)/(mutant_input/wt_input). Thus, a competitive index of <1.0 indicates a strain with a competitive disadvantage.

We also assessed whether V. cholerae requires ryhB to colonize the small intestines of suckling mice, which are an established model host for study of cholera pathogenesis. Previous analyses revealed that high-affinity iron acquisition systems are important for murine intestinal colonization by V. cholerae (34) but also suggested that multiple iron sources are available in this environment and that iron may not be in limited supply (6, 24). To determine whether RyhB-dependent regulation of iron utilization is important for intraintestinal survival in a murine host, we performed in vivo competition assays. These experiments did not reveal a statistically significant difference between growth of the ryhB and wt strains in the small intestine of the suckling mouse (Fig. 4C). Thus, proper control of the RyhB regulon is clearly not critical for survival of V. cholerae in a murine host, and inappropriate expression of this regulon cannot account for the attenuation of an hfq mutant in this model system.

The degree of iron availability in the suckling mouse intestine has not been directly assayed, and conflicting results concerning in vivo expression of iron-regulated genes have been published (6, 13). Since we had characterized V. cholerae expression of RyhB in response to various iron levels, we assessed whether the degree of RyhB expression in the murine small intestine is indicative of an iron-replete or an iron-deficient environment. We used quantitative RT-PCR to determine the abundance of ryhB transcripts relative to those of a control gene (rpoB) in RNA isolated following growth of wt V. cholerae in the murine intestine and other environments. Levels of rpoB transcripts are relatively constant; thus, changes in the transcript ratio are indicative of changes in the abundance of RyhB. In wt cells grown in LB, the mean transcript ratio was 2.35; this defines the ratio for “uninduced” levels of RyhB (Table 2). In contrast, the mean transcript ratio for V. cholerae grown in LB-100 μM dipyridyl was 4,775, a finding consistent with the RyhB induction observed under these conditions on Northern blots. For RNA obtained from five murine small intestines infected with V. cholerae, the transcript ratio ranged from 9.9 to 192.1, with a mean value of 56.5 (Table 2); thus, the highest transcript ratio measured for the intestinal samples is almost 10-fold lower than the lowest ratio for cells cultured in LB-dipyridyl. The lack of overlap between these data sets suggests that, overall, the murine small intestine is not a highly inducing environment, although it appears that a small amount of RyhB induction takes place there, perhaps in a subset of cells. The slight increase in the ryhB/rpoB ratio in intestinal samples compared to cells grown in LB medium may also reflect differences between the bacterial growth phases, as preliminary data suggest that cells grown to a lower density in LB medium also have an increased transcript ratio (data not shown). Nonetheless, it is clear that ryhB transcripts in V. cholerae isolated from infected mice are far less abundant than in cells grown in the presence of an iron chelator, suggesting that the suckling mouse model of V. cholerae infection is not highly limited for iron.