Introduction

Vibrio cholerae, the causative agent of cholera, is an opportunistic pathogen that is well adapted for survival in the marine environment during interepidemic periods (Faruque et al., 1998). Environmental surveillance for V. cholerae in Bangladesh revealed that V. cholerae persists in the aquatic environment predominately as viable but nonculturable cells in the planktonic phase and as culturable cells in biofilms formed on biotic and abiotic surfaces (Alam et al., 2006). V. cholerae biofilms associated with phytoplankton and zooplankton in water columns are suggested to be important for cholera transmission, based on the reported correlation between cholera disease outbreaks and seasonal aquatic plankton blooms (Huq et al., 1984, 1995). In addition, it has been shown that simple filtration of water through sari cloth removes up to 99% of V. cholerae cells by removing the zooplankton-associated V. cholerae (Huq et al., 1996), and that this practice reduces cholera infections by 48% (Colwell et al., 2003).

Chitin is the main component of the cell walls of fungi and the exoskeletons of crustaceans and insects. Crustaceans are the majority of zooplankton, with krill and copepods comprising the largest animal biomass in the marine environment (Murphy et al., 2007). Chitin is probably the most abundant biopolymer in the marine environment, and chitin utilisation is important for C and N recycling. The chitinolytic bacteria, especially surface-associated bacteria that have higher metabolic efficiency, are essential in this process (Yu et al., 1991).

The colonisation of chitinous surfaces of zooplankton is important for the persistence of V. cholerae in the aquatic environment (Huq et al., 1983; Pruzzo et al., 2008), where V. cholerae has been shown to utilise chitin as a sole nutrient source, allowing for the maintenance of metabolic activity for up to 6 months (Nahar et al., 2011). V. cholerae has several mechanisms that facilitate chitin colonisation, including the mannose-sensitive haemagglutinin (Chiavelli et al., 2001; Meibom et al., 2004), chitin-regulated pilus (Meibom et al., 2004) and two chitin-binding proteins (36 and 53 kDa) (Tarsi and Pruzzo, 1999; Kirn et al., 2005; Zampini et al., 2005). The toxin coregulated pilus has also been shown to mediate biofilm differentiation on chitin surfaces (Reguera and Kolter, 2005). The chitin utilisation pathway is conserved in the Vibrionaceae and many of these genes are upregulated upon attachment of V. cholerae to chitin (Meibom et al., 2004; Hunt et al., 2008).

In addition to nutrient accessibility, colonisation of chitinous zooplankton has also been shown to increase the resistance of V. cholerae to low pH (Nalin et al., 1979) and alum/chlorine treatment (Chowdhury et al., 1997). Furthermore, association with chitin surfaces promotes horizontal gene transfer by inducing natural competence (Meibom et al., 2005), allowing V. cholerae to acquire the genes useful for adaption to various stresses both in vivo and in vitro (Bartlett and Azam, 2005).

Protozoan grazing has been identified as one of the major mortality factors faced by bacteria in the environment, and predation has a strong top–down control on V. cholerae populations (Worden et al., 2006). In response to protozoan grazing, V. cholerae has been shown to exhibit an increase in biofilm formation, where V. cholerae biofilms are protective against grazing pressure, whereas the planktonic counterparts are more readily consumed (Matz et al., 2005). The high cell density within biofilms also allows for the quorum sensing (QS) regulated production and accumulation of antiprotozoal factors (Sun et al., 2013). QS has also been reported to control the expression of an extracellular protease PrtV (Vaitkevicius et al., 2006) and an inhibitory factor, VasX, secreted by type VI secretion system (Ishikawa et al., 2009; Miyata et al., 2011), which are important for resistance to predation.

In this study, we investigated the role of chitin in V. cholerae grazing resistance by exposing V. cholerae biofilms grown on chitin flakes and crustacean surfaces to the bacteriotrophic surface-feeding flagellate, Rhynchomonas nasuta. Results show that biofilms on chitinous surfaces were more resistant to predation than those on abiotic surfaces. Furthermore, we demonstrated that chitin metabolism was QS-regulated and that the ammonium produced as a by-product of chitin metabolism is toxic to R. nasuta.

Materials and methods

Strains and culture conditions

V. cholerae A1552 wild type (WT) and its isogenic mutant ΔhapR, a master response regulator of QS (Yildiz et al., 2004), were routinely maintained in Luria–Bertani broth and on agar plates. R. nasuta, a surface-feeding flagellate, was isolated from the Sydney Institute of Marine Science (Erken et al., 2011) and routinely grown on heat-killed Pseudomonas aeruginosa PAO1 (final concentration 107 cells ml−1) in 50% nine-salt solution (NSS) medium (NSS is an artificial seawater medium that contains 17.6 g NaCl, 1.47 g Na2SO4, 0.08 g NaHCO3, 0.25 g KCl, 0.04 g KBr, 1.87 g MgCl2•6H2O, 0.45 g CaCl2•2H2O, 0.01 g SrCl2•6H2O and 0.01 g H3BO3 in 1 l of distilled water; 50% NSS contains half of the salts of NSS medium) (Vaatanen, 1976) grown statically at room temperature (RT). Ampicillin (200 μg ml−1) and gentamycin (200 μg ml−1) were added to eradicate indigenous bacteria. R. nasuta was subcultured in antibiotic-free medium 3 days prior to grazing experiments.

Artemia cysts from the Great Salt Lake, UT, USA (INVE Aquaculture, Dendermonde, Belgium) were decapsulated and hatched as previously described with modification (Sung et al., 2008). Briefly, decapsulated Artemia cysts were incubated in 50% NSS at 26 °C for 18 h for hatching. The axenic Artemia were killed at the Instar I stage by freezing at −20 °C overnight. To rule out bacterial contamination in grazing experiments, aliquots from cultures of R. nasuta and Artemia were plated on Luria–Bertani agar.

Microtitre plate grazing assay

To compare the grazing resistance of V. cholerae biofilms formed on abiotic and chitinous surfaces, batch experiments were performed in 24-well tissue culture plates as previously reported with modification (Weitere et al., 2005). V. cholerae overnight cultures were co-incubated at a final concentration of 106 cells ml−1 with 104 cells ml−1 of R. nasuta in 1 ml of 50% NSS medium. Chitin flakes (2% w/v, C9213, Sigma-Aldrich, St Louis, MO, USA) or glucose (0.01% w/v) or both were supplemented as carbon sources. The microtitre plates were incubated at RT with shaking at 60 r.p.m. for 3 days. After incubation, R. nasuta was enumerated by direct inspection with an inverted light microscope (× 200 magnification, CKX41, Olympus, Tokyo, Japan). V. cholerae biofilm biomass was determined by MTT staining as per manufacturer’s recommendation (Cell Proliferation Kit I, Roche, Basel, Switzerland). The experiments were run in replicates of four and repeated twice.

Flow cell grazing assay

V. cholerae A1552 and R. nasuta were incubated in continuous flow cells (1 × 4 × 40 mm), assembled and prepared as previously described (Moller et al., 1998). One millilitre of 50% NSS containing 106 cells ml−1 of V. cholerae A1552 and 104 cells ml−1 of R. nasuta was injected into the cell and allowed to settle for 2 h before the flow was restored by pumping 50% NSS into the cell at a rate of 3 ml h−1. Chitin flakes were glued to the bottom of flow cells using glass silicon, and glucose (0.01% w/v) was supplemented in the medium flow. The flow cells were incubated at RT for 3 days. R. nasuta was enumerated daily by inverted microscopy, and the effluents were collected and plated to determine the V. cholerae colony-forming units at the end of the experiment. The experiments were run in replicates of three and repeated twice.

V. cholerae biofilms formed on chitinous surfaces of zooplankton and exposed to predation

To compare biofilms on chitin flakes with those formed on natural chitinous surfaces, for example, exoskeletons of Artemia, the microtitre plate assay was modified to include 150 ml−1 killed Artemia instead of chitin flakes. Owing to the difficulties encountered in quantifying the V. cholerae biofilm biomass formed on Artemia surfaces, only R. nasuta was enumerated microscopically with a haemocytometer. The experiments were run in replicates of four and repeated twice.

Supernatant toxicity assay

The cell-free supernatants of V. cholerae biofilms from the above assays were collected, filtered through 0.22 μm filters (Millex-GP, Millipore, Billerica, MA, USA) and stored at −20 °C. The antiprotozoal activity of collected supernatants was tested as previously described (Matz et al., 2005). R. nasuta (at a final concentration 104 cells ml−1) was added to biofilm supernatants supplemented with heat-killed P. aeruginosa PAO1 in 24-well tissue culture plates and incubated at RT with shaking at 60 r.p.m. for 3 days. R. nasuta was enumerated microscopically. The experiments were run in replicates of four and repeated twice. The ammonium concentration of supernatants was determined by using the Ammonia Kit (TNT832, Hach, Lakewood, CO, USA) as per manufacturer’s recommendation.

Transcriptomic analysis

V. cholerae A1552 WT and ΔhapR were grown in 50% NSS supplemented with 2% w/v chitin flakes at RT with shaking at 60 r.p.m. for 3 days. After removing the planktonic phase, the total RNA was extracted from biofilms formed on chitin flakes using the RNeasy Protect Bacteria Mini Kit (Qiagen, Hilden, Germany) as per the manufacturer’s recommendation, and paired-end sequencing was performed on the Illumina Hi-Seq 2000 platform (San Diego, CA, USA) with 237 million reads of 100 bp length.

Reads from Illumina were processed with Cutadapt (version 1.3) (Martin, 2011) with a quality trimming threshold of 20 and a minimum read length of 50 bp. The trimmed reads were subsequently depleted of ribosomal RNA with SortMeRNA version 1.8 (Koylova and Touzete, 2012), and the nonribosomal RNA reads mapped to the genome of V. cholerae (RefSeq accession numbers NC 002505 and NC 002506) using Bowtie version 2.1.0 (Langmead et al., 2009). The transcript abundance was estimated with Cufflinks version 2.1.0 (Langmead et al., 2009) using the default geometric method. Calculations of the log2 fold changes and false discovery rates were performed with Cuffdiff (Langmead et al., 2009). The differentially expressed transcripts were determined using the selection criteria of more than twofold change and q-value <0.05.

Results

Grazing resistance of V. cholerae biofilms grown on chitin flakes in batch culture

Owing to the importance of chitin utilisation and resistance to protozoan grazing for the survival of V. cholerae in the environment, the grazing resistance of chitin-associated V. cholerae and the role of QS in this process were assessed. Hence, biofilms of V. cholerae WT and ΔhapR strains were grown on chitin flakes and abiotic control surfaces, and exposed to predation by R. nasuta.

In the absence of a carbon source, the biofilms of V. cholerae formed on the surface of microtitre plates were not extensive. In the presence of glucose, there was a slight, but not significant, increase in biofilm biomass (228%, P>0.05), whereas the biomass of biofilms formed on chitin flakes was significantly greater (3053%, P<0.001) (Figure 1). There was no significant removal of biofilm biomass by R. nasuta under any of the above batch biofilms and no significant difference in biofilm biomass of V. cholerae WT and ΔhapR strains (Figure 1).

Figure 1
figure 1

Grazing resistance of biofilms of V. cholerae grown on abiotic and chitinous surfaces. Biofilm biomass of V. cholerae WT (a) and ΔhapR (b) strains with no carbon source (NSS), with glucose (Glu) and/or chitin flakes (Chi) in the presence (Graz+) and absence (Graz−) of R. nasuta.

Full size image

The antiprotozoal activity of V. cholerae biofilms was determined by enumeration of R. nasuta. In V. cholerae-free controls, the number of R. nasuta increased with the supplementation of glucose and/or chitin flakes. The presence of V. cholerae WT biofilms did not significantly affect the number of R. nasuta on abiotic surfaces, but R. nasuta was killed by WT biofilms grown on chitin flakes (P<0.001). In contrast, the ΔhapR biofilms supported significantly more R. nasuta individuals on abiotic surfaces compared with the controls and the WT biofilms (P<0.001). There was no significant difference in the number of R. nasuta co-incubated with ΔhapR biofilms grown on chitin flakes compared with the controls. Compared with V. cholerae WT strain, ΔhapR biofilms were less toxic in all nutrient conditions (P<0.001) (Figure 2).

Figure 2
figure 2

Effect of V. cholerae biofilms on survival of R. nasuta. Number of R. nasuta feeding on V. cholerae WT and ΔhapR biofilms with no carbon source (NSS), with glucose (Glu) and/or chitin flakes (Chi). R. nasuta concentrations in media without V. cholerae were used as controls.

Full size image

Grazing resistance of V. cholerae biofilms grown on chitin flakes in flow cells

To determine if grazing resistance was similar under continuous flow conditions where there is dilution of extracellular components to that observed in batch cultures, V. cholerae WT stain and R. nasuta were co-incubated in flow cells with chitin or glucose. The grazing effects were determined by quantification of V. cholerae colony-forming units in the biofilm effluent and the numbers of R. nasuta.

The effluents of V. cholerae biofilms grown on chitin flakes contained more V. cholerae cells (295%, P<0.001) than the effluents from biofilms developed in flow cells supplemented with glucose, indicating that the biofilm biomass was higher when grown on chitinous surfaces. Similar to the above grazing assays performed in microtitre plates, predation by R. nasuta did not significantly reduce the biofilm biomass as the colony-forming units of effluents were similar to nongrazed controls (Figure 3).

Figure 3
figure 3

Biofilms formed on chitin surfaces in a flow-through system. Colony-forming units of V. cholerae WT biofilm effluents grown on glucose (Glu) and chitin flakes (Chi). Experiments were performed in flow cells in the presence (Graz+) and absence (Graz−) of R. nasuta.

Full size image

Direct microscopic enumeration revealed that the number of R. nasuta increased when feeding on V. cholerae biofilms grown in glucose, whereas the growth of R. nasuta was inhibited by V. cholerae biofilms grown on chitin flakes, by 80% on day 3 (P<0.001). In control flow cells lacking V. cholerae, the number of R. nasuta was below the detection limits, which was most likely due to nutrient deprivation (Figure 4).

Figure 4
figure 4

Effect of biofilms on survival of R. nasuta in a flow-through system. Number of R. nasuta feeding on V. cholerae WT biofilms grown on glucose or chitin flakes. Experiments were performed in flow cells for 3 days and R. nasuta was quantified by microscopy daily.

Full size image

Grazing resistance of V. cholerae biofilms grown on Artemia

In order to determine if biofilms formed on the chitinous surfaces of zooplankton were also resistant to predation, V. cholerae biofilms were grown first on Artemia in microtitre plates. In preliminary experiments, live Artemia fed on and depleted the R. nasuta population (data not shown). Thus, in the following experiments, Artemia was killed by freezing prior to addition of V. cholerae WT in batch cultures and the number of R. nasuta grazing on V. cholerae biofilms was compared with media controls.

In the absence of V. cholerae, the number of R. nasuta increased more in wells containing Artemia (1710%) than in those containing glucose (224%), indicating that R. nasuta benefited from the nutrients released by freeze-killed Artemia (P<0.001). The V. cholerae biofilms grown on abiotic surfaces did not affect the number of R. nasuta significantly, whereas V. cholerae biofilms formed on Artemia inhibited R. nasuta (P<0.001) (Figure 5).

Figure 5
figure 5

Grazing resistance of biofilms of V. cholerae formed on Artemia. Numbers of R. nasuta in media controls and feeding on V. cholerae WT biofilms with no carbon source (NSS) or with glucose (Glu) and/or Artemia (Art).

Full size image

Ammonium produced by V. cholerae during chitin metabolism is toxic to R. nasuta

Supernatants of V. cholerae biofilms were collected from all the above assays and tested for their inhibitory effects on R. nasuta by comparing with the corresponding media controls. The supernatants of V. cholerae WT biofilms grown in 50% NSS medium were nontoxic to R. nasuta, whereas supernatants from WT biofilms grown in glucose inhibited R. nasuta by 50% (P<0.001) (Figure 6).

Figure 6
figure 6

Toxicity of supernatants from V. cholerae biofilms to R. nasuta. Toxicity of supernatants collected from V. cholerae WT and ΔhapR biofilms grown in artificial seawater (NSS), glucose (Glu) and/or on chitin flakes (Chi). R. nasuta concentrations in media without V. cholerae were used as controls.

Full size image

The supernatants of WT biofilms grown on chitin flakes were significantly more toxic to R. nasuta than those grown in glucose and media controls, resulting in a 98% decrease in R. nasuta numbers (P<0.001). In comparison, the supernatants of ΔhapR biofilms were less toxic than the supernatants of WT biofilms when grown on chitin flakes (P<0.001). Interestingly, all of the supernatants collected from V. cholerae biofilms grown in flow cells and on the surfaces of Artemia did not show significant inhibition (data not shown), which was probably owing to high dilution in the flow cells and/or low chitin abundance in both conditions as discussed below.

Chitin is a modified derivative of glucose containing nitrogen, and the metabolism of chitin produces ammonia as a by-product. Therefore, the ammonium concentration of supernatants collected from V. cholerae biofilms grown in batch cultures was measured (Figure 7). Ammonium was not detected in media controls and was low in nonchitin supernatants. In contrast, chitin utilisation by V. cholerae generated a significant amount of ammonium (3.5 mM) in the supernatants (P<0.001), whereas the ΔhapR biofilms produced 55% less than the WT (P<0.001). The ammonium concentration of supernatants collected from V. cholerae biofilms grown in flow cells and on the surface of Artemia was below the detection limits (data not shown). To confirm that ammonium is toxic to R. nasuta, a dose–response curve was plotted using NH4Cl in 50% NSS medium. The number of R. nasuta decreased with an increase in ammonium (Figure 8).

Figure 7
figure 7

Ammonium concentration of V. cholerae biofilm supernatants. Ammonium concentration of supernatants collected from V. cholerae WT and ΔhapR biofilms grown in artificial seawater (NSS), glucose (Glu) and/or on chitin flakes (Chi). Ammonium concentrations of V. cholerae-free media was measured as controls.

Full size image
Figure 8
figure 8

Effect of ammonium on survival of R. nasuta. Effect of increasing concentration of ammonium on R. nasuta survival. R. nasuta was quantified by microscopy after 3 days incubation in NSS medium supplemented with NH4Cl.

Full size image

QS regulation of V. cholerae chitin utilisation

As the V. cholerae ΔhapR strain was less toxic to R. nasuta than the WT strain and produced less ammonium when grown on chitin flakes, the role of QS regulation of chitin metabolism was investigated by transcriptomic analysis of V. cholerae WT and ΔhapR biofilms grown on chitin flakes in batch cultures. Results demonstrated that 367 transcripts were differentially expressed with more than a twofold change (Supplementary Table S1).

The chitinolytic pathway of V. cholerae (Hunt et al., 2008) begins with the degradation of chitin by chitinase and the transport of N-acetylglucosamine (GlcNAc) monomers or dimers into the periplasm. The transcripts of genes that are involved in this process were largely unchanged in the ΔhapR biofilms compared with the WT biofilms grown on chitin, with the exception of two chitinases (VC0769 and VCA0027), which were expressed at a higher level in ΔhapR biofilms. Following the transport of GlcNAc monomers and dimers, the products of degradation are delivered into central metabolism as fructose-6-phosphate, acetate and ammonium. Among the 22 genes involved in GlcNAc catabolism, transcripts of 19 genes, including the entire chitin catabolic operon (VC0611-VC0620), were expressed at a significantly lower level in the QS-negative ΔhapR biofilms, whereas the transcripts of the remaining 3 genes were not significantly different (Table 1). In addition, transcripts of the genes coding for chitin-regulated pilus (VC2423, VC2424 and VC1612) and the majority of the methyl-accepting chemotaxis genes (29 of total 45 MCPs) were repressed in the ΔhapR biofilms, whereas only 2 MCP genes were upregulated. The RNA-Seq data also confirmed QS regulation of previously reported genes, including those involved in Vibrio polysaccharide production, flagella synthesis, virulence, natural competence, type VI secretion system and c-di-GMP synthesis/degradation (Supplementary Table S1).

Table 1 Differentially regulated genes between V. cholerae WT and ΔhapR biofilms in GlcNAc catabolism
Full size table

Discussion

V. cholerae biofilms formed on chitin surfaces are grazing resistant

Results show that V. cholerae forms significantly more biofilm biomass on chitin flakes than on abiotic surfaces in both microtitre plates and continuous flow cells. It has been reported that chitin colonisation and utilisation are ubiquitous among vibrios, with all completed genomes of vibrios having multiple chitinases (Grimes et al., 2009). The association with chitin is important for the long-term persistence of V. cholerae in the environment. V. cholerae can use chitin as a sole carbon source (Nahar et al., 2011), and colonisation of chitinous zooplankton has been shown to increase the resistance of V. cholerae to low pH (Nalin et al., 1979) and alum/chlorine treatment (Chowdhury et al., 1997). Various chitin attachment mechanisms have been identified (Meibom et al., 2004), although the role of mannose-sensitive haemagglutinin in chitin association is still under debate (Watnick et al., 1999). When the V. cholerae biofilms grown on chitin flakes were exposed to predation by R. nasuta, there was little effect of grazing on biofilm biomass and this was in line with our previous observation that biofilm formation protects V. cholerae from protozoan grazing (Matz et al., 2005).

Ammonium produced by V. cholerae chitin metabolism is toxic to R. nasuta

In both microtitre plates and flow cells, V. cholerae biofilms formed on chitin flakes were more toxic to R. nasuta compared with biofilms formed on abiotic surfaces, indicating the existence of antiprotozoal factor(s) produced by V. cholerae biofilms in association with chitin. Furthermore, the chitin-dependent antiprotozoal activity was also observed when biofilms were grown on the surface of Artemia. Supernatants collected from V. cholerae biofilms grown on chitin flakes in batch cultures were inhibitory to R. nasuta and, subsequently, ammonium was identified to be responsible for the antiprotozoal activity in the supernatant. This metabolite-based grazing resistance may be a general feature of environmental bacteria. For example, biofilms of Pseudomonas fluorescens CHA0 produce the metabolites hydrogen cyanide and 2,4-diacetylphloroglucinol, which have antiprotozoal activities against bacteriotrophic protists (Jousset et al., 2006).

Ammonia/ammonium at high concentrations is generally toxic to eukaryotic cells. It has been reported that the amphipod Gammarus pulex is inhibited by <0.1 mM ammonium (Berenzen et al., 2001; Prenter et al., 2004), and bacterial-derived ammonia in the intestine is largely accepted as the main source for hyperammonaemia in patients with liver disease (Sherlock, 1987). Interestingly, ammonia is also an important environmental signal in the regulation of the life cycle of Dictyostelium discoideum, where high ammonia concentrations inhibit fruiting body formation (Mahadeo and Parent, 2006). The dose–response curve of R. nasuta to ammonium demonstrated that 2 mM ammonium was lethal for 80% of R. nasuta individuals. As the supernatants of V. cholerae WT biofilms grown on chitin flakes contained 3.5 mM ammonium, these biofilms are even more toxic than R. nasuta.

R. nasuta was also inhibited when co-incubated with V. cholerae biofilms formed on chitin flakes in flow cells or on Artemia. In the marine environment, seasonal plankton blooms may provide enough chitin for V. cholerae metabolism to produce inhibitory amounts of ammonium (Huq et al., 1984). It has also been proposed that the inhibitory factors may accumulate to higher concentrations within biofilms than in supernatants owing to high cell density and retention of cell-derived compounds by exopolymeric substances (Matz et al., 2005). In addition, the antiprotozoal activity of V. cholerae biofilms in association with chitinous surfaces may also benefit from other inhibitory factors, for example, PrtV (Vaitkevicius et al., 2006), VasX (Ishikawa et al., 2009; Miyata et al., 2011) and the unknown factor secreted by V. cholerae biofilms (Matz et al., 2005).

QS regulation of chitin metabolism and antiprotozoal activity in V. cholerae biofilms

The reduced production of ammonium and decreased toxicity against R. nasuta of V. cholerae ΔhapR biofilms grown on chitin flakes indicates that QS controls V. cholerae chitin utilisation. Transcriptomic analysis revealed that 19 of the 22 genes involved in GlcNAc catabolism, as well as functionally related chitin-regulated pilus synthesis and chemotaxis, were positively regulated by QS in V. cholerae biofilms.

Interestingly, two putative chitinases, VC0769 and VCA0027, were downregulated by QS in V. cholerae biofilms, indicating that the five chitinases in V. cholerae might be redundant or functionally differentiated. The negative regulation of chitinase by QS has been reported previously, where VCA0811 was found to be repressed by QS in a microarray study of V. cholerae rugosity (Yildiz et al., 2004), and chitinase A of Vibrio harveyi was expressed at higher level in a QS-negative mutant strain (Defoirdt et al., 2010). One possible explanation is that the expression of chitinases may be a nutrient-exploring strategy used by planktonic V. cholerae cells rather than in high-density biofilms. When chitin is present in the surrounding environment, the secreted chitinases degrade chitin and generate soluble oligomers of GlcNAc, which act as signals for V. cholerae chemotaxis, chitin attachment and utilisation (Meibom et al., 2004). To date, only VC1952 and VCA0027 have been confirmed to have catabolic activity for chitin and the double-deletion mutant could not grow on chitin as the sole carbon source (Meibom et al., 2004). Thus, it is possible that the remaining three chitinases may not have catabolic activity.

It has been proposed that many mechanisms involved in V. cholerae environmental persistence are important in pathogenesis as well (Vezzulli et al., 2008). Indeed, it has been reported that six genes (VC0613, VC0614, VC0616, VC0619, VC1284 and VC2424) involved in chitin utilisation are also induced during colonisation of the infant mice small intestine (Mandlik et al., 2011). Therefore, the QS-regulated chitin utilisation genes of V. cholerae biofilms may play a role in infection by V. cholerae.

Chitin association has been reported to provide V. cholerae nutrient accessibility and to induce natural competence of V. cholerae. The data presented here show that biofilm formation of V. cholerae on chitin provides increased grazing resistance and is toxic to the heterotrophic protist, R. nasuta, owing to the production of ammonia as a by-product of chitin metabolism. Furthermore, we show that chitin metabolism is QS-regulated. These results provide an explanation for the association of V. cholerae with chitinous exoskeletons of zooplankton as well as a role for QS-regulated phenotypes in persistence and survival in the marine environment.