“Quorum Non-Sensing”: Social Cheating and Deception in Vibrio cholerae

Abstract

Quorum sensing (QS) is widely used by bacteria to coordinate behavior in response to external stimuli. In Vibrio cholerae, this process is important for environmental survival and pathogenesis, though, intriguingly, a large percentage of natural isolates are QS deficient. Here, we show that QS-deficient mutants can spread as social cheaters by ceasing production of extracellular proteases under conditions requiring their growth. We further show that mutants stimulate biofilm formation and are over-represented in biofilms compared to planktonic communities; on this basis, we suggest that QS-deficient mutants may have the side effect of enhancing environmental tolerance of natural populations due to the inherent resistance properties of biofilms. Interestingly, high frequencies of QS-deficient individuals did not impact production of QS signaling molecules despite mutants being unable to respond to these inducers, indicating that these variants actively cheat by false signaling under conditions requiring QS. Taken together, our results suggest that social cheating may drive QS deficiency emergence within V. cholerae natural populations.

INTRODUCTION

Social cheating remains an important obstacle to understanding the evolution of cooperation in both inter- and intraspecific contexts (1). Within cooperative systems, cheaters can spread under conditions that allow benefit from social actions performed or goods produced by other individuals while not incurring the cost of these traits themselves. Because bacteria lack complex social systems which can allow for the suppression of cheating, cheaters in bacterial populations may be primed for persistence within the group or rise to frequencies high enough to disrupt the social dynamics of the population entirely.

One particular bacterial system susceptible to exploitation is quorum sensing (QS), a cell-cell communication process which allows bacteria to coordinately alter gene expression in response to changes in cell density (2). Here, individuals may incur metabolic costs of producing signals, enzymes, or proteins while presumably benefiting from similar production by other group members. For instance, it is predicted that QS plays a role in bacterial pathogens utilizing QS-controlled virulence networks, since coordinated expression of toxins could be advantageous when the population has reached sufficiently high numbers during infection (3–5). In addition, QS may be similarly useful in environmental biofilm populations, since QS control of extracellular enzymes allows for efficient degradation of macromolecules in the surrounding area only when the population is large enough to scavenge the resources (3).

Vibrio cholerae is a Gram-negative bacterium responsible for causing cholera. In its environmental states, V. cholerae primarily resides in aqueous reservoirs such as rivers, estuaries, and coastal waters, where it associates with both abiotic and biotic (primarily chitinous) surfaces in the form of biofilms. The infectious cycle of V. cholerae begins with contaminated food and water, followed by propagation in the host intestinal tract and exit through an acute, watery diarrhea (6). QS plays a vital role in V. cholerae as this organism transits between environmental and host reservoirs. QS regulatory systems are complex in V. cholerae. At low cell densities, the components of the QS pathway act as kinases to phosphorylate LuxO, which in turn activates the transcription of small RNAs (qrr1 to qrr4) that destabilize mRNA of hapR, encoding the QS master regulator (7). At high cell densities, two sets of autoinducers, CAI-1 and AI-2, bind to cognate sensors on the bacterial surface and induce conformational changes in the sensors (8), which results in dephosphorylation of LuxO. The small RNAs are therefore no longer transcribed, the hapR transcript is stabilized, and HapR protein is expressed in high quantities so that it is able to mediate its downstream effects. In order to adapt to diverse environmental niches, V. cholerae uses QS mechanisms to coordinate its gene expression and control a variety of physiological functions (2). For example, it has been shown that QS represses virulence gene expression and biofilm formation while activating production of extracellular proteases (HapA and PrtV) through the QS master regulator, HapR, at high cell densities, suggesting the importance of QS in entering and exiting the host, as well as in environmental survival (9–12). The QS regulon also consists of a number of additional genes involved in chitin-induced natural competence, stress responses, and hemolysin production, among others (13–16).

Clearly, cell-cell communication mediated by QS is important for transitions between host and aquatic environments in V. cholerae. Surprisingly, however, ca. 50% of natural V. cholerae isolates are QS deficient, most of which bear mutations in the hapR gene, implying a selective advantage of QS⁻ mutants in nature (17–19). Similar findings have been noted in the opportunistic pathogen Pseudomonas aeruginosa, where strains containing mutations in LasR, the regulator of QS in P. aeruginosa, have been isolated from a variety of infection origins (20). Recent work has suggested that these strains may spread as a result of social cheating through ceasing production of community-available “public goods” (1, 21). It is predicted that P. aeruginosa growth under protein-rich conditions selects for QS-deficient strains capable of utilizing protein breakdown products but not contributing to this pool of extracellular proteases. These QS⁻ mutants may also act as social cheats within biofilms when cocultured with wild-type cells, compromising antibiotic resistance and overall productivity of the biofilm (22). Interestingly, QS in P. aeruginosa appears to be self-regulating, since genes necessary for the production of QS signaling molecules are contained in the ca. 6% of the genome predicted to be regulated by QS (1). This suggests that most P. aeruginosa QS⁻ mutants passively cheat since they neither respond to nor produce signals themselves. In this study, we show in vitro growth conditions in which high frequencies of QS-deficient mutants arise in V. cholerae. In addition, we show that populations with high frequencies of QS-deficient mutants produce QS signaling molecule levels similar to those of wild-type populations, suggesting that these mutants actively cheat through signaling others to produce QS-dependent “public goods.” We hypothesize that this observed social cheating may explain the abundance of QS mutants in natural populations of V. cholerae.

MATERIALS AND METHODS

Bacterial strains and growth.

All strains used in the present study were derivatives of V. cholerae El Tor C6706 (23). In-frame deletion strains were constructed using double-homologous recombination by cloning the flanking regions of the genes of interest into a suicide vector, pWM91, containing a sacB counterselectable marker (24). These plasmids were then introduced into V. cholerae by conjugation, and colonies were selected for recombination events. PCR amplification of the flanking region from the resulting strains was then applied to confirm the deletion constructs. For all liquid cultures and plating, streptomycin was added to a final concentration of 100 μg/ml unless otherwise noted. Bacterial cultures were grown as indicated in Luria broth (LB) or M9-minimal medium (25) containing either casein sodium salts (1% [wt/vol]) or glucose (1% [wt/vol]) as a carbon source.

Competition experiments.

Single colonies of V. cholerae C6706 wild-type and QS-deficient deletion mutants (ΔhapR) were inoculated and grown independently with shaking (180 rpm) at 30°C overnight in LB containing streptomycin. Inocula were prepared by normalizing the optical density at 600 nm (OD₆₀₀) of independent cultures, serially dilution through sterile phosphate-buffered saline (PBS), mixing strains at indicated ratios, and inoculation 1:100 into either LB or M9 medium containing 1% casein protein, glucose, glycerol, or succinate. Cultures were grown at 30°C with shaking for ∼24 h. Daily transfers were performed by diluting overnight growth 1:100 into fresh medium. At the indicated time points, cultures were plated onto LB agar containing streptomycin and grown at 37°C for 24 h. ΔhapR strains were identified by rugose colony formation after an additional 48 h of growth at room temperature. Each experiment was performed in triplicate.

Measurement of QS signaling molecule concentrations.

To determine whether QS-deficient strains are capable of signaling, independent cultures of wild-type and ΔhapR strains were grown in LB without antibiotics at 37°C with shaking. The concentrations of QS signaling molecules CAI-1 and AI-2 in culture supernatants were measured at OD₆₀₀ values of 0.3, 0.8, and 1.2, as described previously (18). To confirm that mixed populations containing both wild-type and ΔhapR strains in casein medium still produced QS molecule levels similar to samples with the wild type alone, strains were grown independently in LB at 30°C. After overnight growth, the OD₆₀₀ was normalized, and each strain was diluted 1:100 into M9 minimal medium containing 1% casein (wt/vol) at increasingly lower ΔhapR/wild-type ratios of 1:1, 1:20, and 1:200. Cultures were grown overnight at 30°C with shaking. Daily inocula were prepared by diluting overnight cultures 1:100 into fresh medium. After 7 days of growth, QS molecule concentrations were measured as previously described.

De novo evolution of QS deficiencies.

To test the QS-deficient mutant appearing de novo within populations grown on casein protein and the dependency of this on protease production, ten individual overnight cultures of indicated strains were inoculated 1:100 into M9 minimal medium containing either 1% casein protein (wild type, ΔhapA mutant [VCA0865], and ΔprtV mutant [VCA0223]) or 1% glucose (wild type only) and grown at 30°C with shaking. Daily propagation was performed by diluting overnight cultures 1:100 into fresh growth medium. At various points, populations were screened for QS-deficient strains through serially diluting and plating onto LB agar. QS-deficient phenotypes were screened as described below.

Growth and measurement of bacterial biofilms.

Single colonies of either wild-type, ΔhapA, ΔprtV, or ΔepsF (VC0920) strains were inoculated into separate borosilicate glass tubes containing 1 ml of minimal medium with 1% casein protein and grown at room temperature (22°C) without shaking for the time indicated. After either 10 or 20 days of growth, planktonic cells were removed from the biofilms, serially diluted, and plated onto LB agar containing streptomycin. Biofilms were washed to remove remaining planktonic cells and resuspended in sterile PBS using sterile borosilicate glass beads. This suspension was then serially diluted and plated onto LB agar containing streptomycin. Plates were incubated at 37°C for 24 h, and colonies were scored for QS-deficient mutants.

Screening and confirmation of QS-deficient phenotypes.

To quantify QS-deficient mutants within the populations, cultures were plated onto LB agar as indicated, and colonies were scored for rugose phenotypes indicative of QS-deficient mutations. A subset of candidate mutants were then confirmed by introducing a QS-activated luxCDABE plasmid and examining them for luminescence production (26). The sequences of the hapR locus of a subset of mutants were determined by PCR amplification of the hapR coding region and sequencing. Mutations in the hapR gene were compared to previously published hapR sequences.

Fluorescence microscopy.

P_lac-rfp and P_lac-gfp were cloned into the pPZP201 vector (27) and introduced into wild-type and hapR mutant strains, respectively. Each strain was grown overnight in LB, and the OD₆₀₀ was normalized. Strains were mixed at the indicated ratios and coinoculated 1:100 into M9 minimal medium containing 1% casein (wt/vol). Cultures were grown at 30°C for 24 h and inoculated 1:100 into fresh medium for the duration of the experiment. Initial and final time points (7 days) were examined by fluorescence microscopy.

RESULTS

Selective advantage and de novo emergence of QS-deficient mutants under protein-rich conditions.

QS regulates a number of important physiological functions in V. cholerae, but previous work has shown that >50% of V. cholerae natural isolates are QS-deficient mutants, most of which contain mutations in the QS master regulator gene hapR (17-19), indicating that QS-deficient isolates may have a selective advantage under certain conditions. To test this hypothesis, we mixed wild-type cells with hapR in-frame deletion mutants representing QS⁻ strains and propagated this mixture under various growth conditions. We found that the ratio of hapR mutants within populations remained virtually unchanged when grown in LB, artificial seawater, or natural pond water (see Fig. S1 in the supplemental material). Interestingly, when mCherry-labeled wild-type cells and green fluorescent protein (GFP)-labeled ΔhapR cells were mixed at 200:1 ratios and grown in a minimal medium containing casein as a sole carbon source, the ΔhapR mutant could reach high relative ratios after 10 days (Fig. 1A). The initial inoculum ratio of wild-type and ΔhapR strains did not affect this outcome (Fig. 1B). Taken together, these results indicate a strong selective advantage of QS-deficient strains when cocultured with wild-type individuals under conditions in which proteins are used as the sole carbon source. Interestingly, total cell number within the populations did not change much throughout the duration of the experiment (see Fig. S2 in the supplemental material), suggesting that low frequencies of wild-type individuals are sufficient to support the entire population.

FIG 1.

Selective advantage of QS mutants in casein medium. (A) Representative fluorescence microscopy images of competition between mCherry-labeled wild-type and GFP-labeled ΔhapR mutant strains inoculated into minimal medium containing casein (1% [wt/vol]) in a 200:1 wild-type/ΔhapR mutant ratio. (B) Coinoculation of wild-type and ΔhapR strains in minimal medium containing casein (1% [wt/vol]) shows strong benefit and stable persistence of QS-deficient mutants. QS mutant phenotypes were determined at various time points by examining rugose colony formation. Starting inoculum ratios of ΔhapR mutants to the wild-type strain were designed to approximate 1:1, 1:20, and 1:200. Each experiment was performed in triplicate. Values shown are mean percentages at indicated time points ± the standard error (error bars).

Because clinical and environmental QS-deficient strains display a variety of mutations within the hapR reading frame (17–19), we reasoned that natural QS⁻ isolates are likely due to separate spontaneous mutational events and subsequent selection rather than migration of mutants into new populations. To test whether hapR mutants could emerge de novo within a population, we propagated 10 populations of wild-type V. cholerae separately under conditions utilizing either protein or glucose as a carbon source and sampled the dynamics of QS deficiency mutants within the population over 25 days of growth (∼170 generations). As shown in Fig. 2A, QS⁻ lineages arose after a delay and swept to relatively high proportions over time within populations utilizing casein protein as a carbon source, although virtually no QS-deficient mutants were detected in cultures grown on glucose. It has been reported that catabolite repression affects QS regulation in V. cholerae (28). We examined hapR mutant competitiveness in a minimal medium containing either glycerol or succinate. We found that, similar to when grown in medium containing glucose, hapR mutants and the wild type grew equally well in these carbon sources (see Fig. S3 in the supplemental material), suggesting that the absence of QS mutants in the presence of glucose is not due to catabolite repression. Sequencing of the hapR coding region of 37 in vitro QS-deficient mutants isolated from separate populations grown on casein protein indicated that a majority of these mutants (35 of 37) had defects in the HapR protein, some of which contained identical mutations seen in clinical and environmental isolates previously reported (17, 18) (Fig. 2B, red symbols). These findings not only confirmed the advantage of dysfunctional QS networks under these conditions but also suggested a possible cause of V. cholerae QS deficiency isolates persisting in natural settings where protein might serve as an important nutrient source.

FIG 2.

De novo evolution of QS-deficient mutants in vitro. (A) Wild-type cultures were grown in minimal medium containing either casein (triangles) or glucose (rhombuses) added in final concentrations of 1% (wt/vol). QS mutants were surveyed through plating onto LB agar and scoring for rugose colony formation and confirmed by testing bioluminescence output using a QS-controlled luminescence reporter plasmid (pBB1). The data points represent n = 10 replicates. The values shown are mean percentages at indicated time points ± the standard error (error bars). (B) hapR sequence comparison of QS⁻ mutants from de novo evolution and natural isolates. De novo mutants were recovered from casein medium and the reading frame of hapR was amplified by PCR using the primers CCCAACAGAGATTGACCTTGA (forward) and CTGGAAGTGATACCCATCTG (reverse). PCR products were sequenced, and the reading frame was analyzed for amino acid deletions (circles), frameshift nucleotide insertions or deletions (triangles), missense mutations (squares), and nonsense mutations (diamonds). Previously published data are shown as black-bordered, open shapes. Strains obtained and sequenced through the present study are indicated by red, filled shapes. (C) Growth of wild-type and mutants in 1% casein and LB. Each strain was inoculated independently and grown overnight in either minimal medium containing 1% casein or in LB at 30°C for 16 h. The OD₆₀₀ was then measured. Each experiment was performed in triplicate. (D) Propagation of protease-deficient strains mitigates selective advantage. Individual ΔprtV and ΔhapA mutants were generated and propagated for 25 days (∼170 generations) in minimal medium containing casein. Experiments were performed with 10 replicates. Horizontal lines show the mean percentages at the indicated time points.

Protease deficiencies mitigate the advantage of QS mutants.

Previous studies in P. aeruginosa have suggested that QS-deficient mutants can act as social cheaters in high-protein conditions, potentially benefiting from ceasing production of QS-controlled extracellular proteases. In V. cholerae, it has similarly been shown that the transcription of two extracellular proteases, HapA and PrtV, requires HapR-dependent QS activation (10, 12). To test whether this selective advantage of V. cholerae QS-deficient mutants in protein rich environments is due to the metabolic burden of protease production, ΔhapA and ΔprtV were propagated in casein medium and examined for the evolution of QS⁻ phenotypes. Despite a slight defect in overall growth relative to the wild type, ΔhapA and ΔprtV single mutants could grow in casein medium, whereas neither ΔhapR or ΔhapA ΔprtV double mutants were able to grow (Fig. 2C). As a control, both ΔhapR and ΔhapA ΔprtV mutants grew to wild-type levels in LB (Fig. 2C), indicating that either protease is sufficient for V. cholerae growth in protein-rich conditions. Intriguingly, no QS-deficient strains were detected in protease-deficient backgrounds after 25 days of growth in the casein medium (Fig. 2D), implying that the loss of either protease may be capable of alleviating the metabolic burden of QS. We hypothesized that either gene deletion is individually capable of mitigating the selective advantage of QS deficiencies or that these two genes likely have additive or multiplicative fitness burdens under conditions promoting cheating. To ensure that no additional extracellular proteases were produced in the ΔhapR mutants, we compared the protease production of the wild type and of the ΔhapR and ΔhapA ΔprtV mutants. Figure S4 in the supplemental material shows that at least under the tested conditions, ΔhapR mutants did not produce additional protease.

V. cholerae QS-deficient mutants do not affect QS signaling molecule production.

P. aeruginosa QS master regulator LasR has previously been shown to positively regulate lasI, responsible for the synthesis of the chemical signal (3-oxododecanoyl)-l-homoserine lactone (3-oxo-C₁₂-HSL) (29); thus, lasR mutants are likely producing fewer QS molecules in addition to being unable to respond to these signals. On the other hand, in V. cholerae, QS master regulator HapR does not appear to regulate the synthesis of QS molecules CAI-1 and AI-2 (15, 30). To confirm this, independent cultures of wild-type cells and ΔhapR mutants were grown in rich medium and assayed for QS molecule production at various phases of growth. As expected, no difference was detected in signal production between wild-type (gray bars) and ΔhapR (white bars) strains under these conditions (Fig. 3A and B).

FIG 3.

CAI-1 and AI-2 production. (A) Wild-type strain (gray bars) and ΔhapR deletion mutants (white bars) were grown in LB and assayed for autoinducer production at various phases of growth. CAI-1 and AI-2 production assays were performed as previously described (18). The fold change is reported relative to the LB control. The data represent n = 3 replicates. The values shown are the mean fold changes ± the standard errors (error bars). (B) The ΔhapR and wild-type strains were inoculated in three replicates in casein medium and competed with daily propagation under conditions previously described. After 7 days of growth, the CAI-1 and AI-2 concentrations were measured. QS signal production data are presented as the percentage of AI produced by wild-type only cultures under identical conditions. The data represents three replicate cultures. The values shown represent the mean percentages ± the standard errors (error bars).

We then examined the QS molecule production of V. cholerae when grown in protein-rich medium. Because QS is required for growth in medium containing casein as a carbon source (Fig. 2C), ΔhapR deletion strains were mixed with the wild type in decreasing ratios and competed in casein medium under conditions previously described. After 7 days of coculture, CAI-I and AI-2 production was assayed relative to wild-type controls grown under identical conditions. Although three inoculum ratios were used (ΔhapR mutant/wild-type ratios of 1:1, 1:20, and 1:200), 7 days of competition is expected to have resulted in selection for at least 99% of the ΔhapR mutants within populations based on previous results (Fig. 1B). As shown in Fig. 3C, no significant differences in signal levels were noted between mixed populations and wild-type controls, indicating that the QS master regulator, HapR, is not necessary for autoinducer production in V. cholerae. Taken together, these results indicate that social cheating in V. cholerae QS proceeds in an active manner by signaling others to produce community available goods and yet not contributing to the pool. This may be different from the situation in P. aeruginosa, in which many QS mutants utilize public goods but make no effort to induce production. Of note, because autoinducer production of individual QS mutants grown in protein-rich medium was not tested, we cannot rule out the possibility that some mutants may have acquired additional mutations that affect CAI-1 and/or AI-2 production.

QS-deficient mutants are disproportionately enriched in biofilm communities.

Since it is likely that most bacteria found in natural and clinical settings may attach to surfaces rather than being in a free-floating state (31), we tested whether QS-deficient mutants, such as those with mutations in hapR, were differentially represented in the biofilm and planktonic communities within a population. Interestingly, we found that after 10 days of standing growth, few planktonic cells (∼0.2%) displayed QS defects, whereas a significantly higher proportion of those within the biofilm (∼10%) were QS⁻ mutants (Fig. 4A, empty symbols). This difference was even more pronounced with longer incubation times, since more than 50% of the biofilm-associated cells but only 1.2% of the planktonic cells contained dysfunctional quorum systems after 20 days of growth (Fig. 4A, closed symbols). Thus, QS-deficient mutants are disproportionately represented within biofilms, and this effect may be enhanced as the nutrient supply dwindles or metabolite concentration increases. An interesting by-product of this enrichment within biofilms may be increased tolerance to the environmental stresses and low nutrient availability inherent to natural V. cholerae populations because biofilms have such properties.

FIG 4.

Emergence of QS-deficient strains within biofilm and planktonic niches. (A) QS deficient mutants represented in either biofilm or planktonic state within a population. Individual replicates of V. cholerae C6706 were grown in medium containing casein (1% [wt/vol]) for either 10 or 17 days under standing conditions, and each community was examined for the presence of QS deficient individuals. The data represent the combination of 10 independent replicates for each time period of growth. P, planktonic cells; B, biofilm-associated cells. **, P < 0.005; ***, P < 0.0001. (B) Evolution of QS mutants within biofilm communities in ΔhapA, ΔprtV, or ΔepsF mutant strain backgrounds grown in minimal medium containing 1% casein. Each experiment was performed with 10 replicate populations. Horizontal lines represent average percent QS negative. (C) Biofilm formation in spent minimal medium containing 1% casein. Overnight cultures of wild-type, ΔhapA, ΔhapR, ΔepsF, and ΔepsF/hapR strains grown in LB were inoculated independently at 1:100 into sterilized spent minimal medium containing 1% casein and grown at 22°C for 24 h. Biofilms were stained using crystal violet and imaged.

Because HapR has been shown to repress biofilm formation, we reasoned that an interesting advantage of QS-deficient strains in natural populations may be the enhanced biofilm production intrinsic in QS⁻ backgrounds. To test this, in-frame deletion mutants of the exopolysaccharide synthase gene epsF (ΔepsF) were cultured under conditions described above and examined for the emergence of QS-deficient mutants. Although epsF mutants were significantly attenuated in biofilm formation, high numbers of QS⁻ mutants were detected within both biofilm and planktonic populations (Fig. 4B). However, we found that biofilm formation in ΔepsF ΔhapR double-deletion mutants was enhanced compared to that of ΔepsF single-deletion mutants (Fig. 4C). These data suggest that biofilm formation may play an indirect role in driving selection for QS-deficient mutants under these conditions.

To investigate the role of social cheating on the emergence of QS⁻ within biofilms, deletion strains of each extracellular protease were cultured under identical conditions and scored for characteristics of QS deficiency mutants. These protease-deficient strains formed biofilms similar to that of the wild type under this growth condition (Fig. 4C and data not shown). No QS-deficient mutants were detected in biofilms in either of the protease-deficient strains up to 20 days of growth (Fig. 4B), suggesting that social cheating may be the primary driving force for V. cholerae accumulation of QS deficiency mutants in biofilm populations.

DISCUSSION

QS plays a vital role during transitions between host and aquatic environments of V. cholerae. This cell-cell communication allows for coordinated responses within natural populations as they cope with changing external conditions while transiting between environmental and infectious states. Still, many natural isolates have been identified as lacking functional QS systems, implying a selective advantage of QS deficiencies in nature. Here, we test the relative fitness of QS mutants in a variety of growth conditions. We show that QS-deficient mutants can spread in vitro in experimental populations of V. cholerae and that these mutants are selected for when utilizing protein as a carbon source but not simple sugars. In strains lacking functionality of QS-activated extracellular proteases, this selective benefit is mitigated, suggesting that QS deficient variants may thrive as a result of social cheating under these conditions. We show that social cheating in V. cholerae QS differs from that in other systems as V. cholerae QS-deficient mutants may actively disrupt these social structures by continuing signal production despite being unable to respond to signals themselves. We propose that similar cheating may drive selection for QS deficiencies in natural cholera populations. In other bacterial QS communication systems, “liars” who produce QS molecules without responding to the signals themselves may exist and be a driving force for the evolution of cooperation in bacteria (32).

Seasonality of cholera outbreaks has long been reported. Studies have correlated cholera outbreaks between 1997 and 2000 along Peruvian coasts to increased sea surface temperatures and shown that total bacterial counts increased with elevated water temperatures and chlorophyll a concentrations (33). Cholera outbreaks in the Bay of Bengal near Bangladesh have similarly been noted to correlate with rising sea temperatures and algal blooms (34). It has been proposed that these phytoplankton blooms may serve as predictors of cholera outbreaks, particularly due to subsequent increases of zooplankton grazers, notably copepods, and detritus (35). Although copepods are likely an important vector for the transmission of V. cholerae from environmental pools, studies have additionally shown that increased concentrations of dissolved organic matter (DOM), consisting largely of proteins and carbohydrates, during intense phytoplankton blooms can support rapid growth of V. cholerae (35, 36). We speculate that DOM might provide impetus for social cheating in natural populations of V. cholerae, particularly in the setting of biofilms. In addition, it is possible that protein in host fecal matter may provide another avenue for cheating during host colonization.

Interestingly, in vitro experiments also show disproportionate representation of QS mutants in biofilm and planktonic communities within populations, likely due to derepression of biofilm formation in the QS⁻ mutant lineage. The spread of QS-deficient mutants in biofilm-attenuated strains (such as epsF mutants) suggests that enhanced biofilm production itself cannot explain the selective pressure for QS-deficient mutants but rather may serve as an evolutionary by-product in natural populations. Biofilms have long been shown to play vital roles in resistance to environmental stresses such as salinity changes, temperature variation, desiccation, and acidity, the latter of which is likely pertinent as V. cholerae transits through the host stomach during initial stages of colonization. In addition, enhanced biofilm production in QS-deficient strains may be important for transmission of V. cholerae between environmental reservoirs and from these reservoirs to human hosts as cells adhere to the chitinous exoskeletons of copepods. On the other hand, deletion of the QS regulator in V. cholerae, HapR, has recently been shown to hinder chitin-induced natural competence seen in wild-type strains (13). Taken together, we hypothesize that this enhanced transmission but decreased adaptive ability may imply an evolutionary trade-off dysfunctional QS systems and could help explain the persistence of QS deficiency isolates in nature.