Influence of quorum sensing in multiple phenotypes of the bacterial pathogen Chromobacterium violaceum Free

Chromobacterium violaceum is a bacterial pathogen that communicates through quorum sensing (QS), via the C6-homoserine lactone signal (C6-HSL). It is well known that the production of the pigment violacein is controlled by QS in this microorganism, in fact QS-dependent violacein production is widely used as a marker to evaluate the efficiency of potential anti-QS molecules, such as those extracted from plants. In addition to violacein, the production of chitinase is also known to be controlled by QS, but besides those two phenotypes there is a lack of experimental studies aimed to discover additional process controlled by QS in this organism; therefore, in this work the production of exoprotease, aggregation, biofilm formation, swarming motility, H₂O₂ resistance as well as carbon and nitrogen utilization was determined in the wild-type strain and the QS negative mutant CVO26. Our results indicate that alkaline exoprotease activity is QS controlled in this organism, that QS increases aggregation, biofilm formation, swarming, that may increase H₂O₂ stress tolerance, and that it may influence the utilization of several carbon and nitrogen sources.

INTRODUCTION

Chromobacterium violaceum is a saprophyte Gram-negative bacterium that occasionally behaves as a pathogen of extreme virulence, causing fatal septicemia, skin lesions, liver and lung abscesses (Midani and Rathore 1998; Jitmuang 2008; Yang and Li 2011). In addition, it is biotechnologically relevant due the production of useful metabolites (Durán and Menck 2001). Moreover, it is widely used to study the inhibition of N-acyl homoserine lactone (AHLs)-dependent quorum sensing (QS) by diverse compounds and for assaying the production of short-chain AHLs (McClean et al., 1997; Steindler and Venturi 2007; Castillo-Juarez et al., 2013), due the tight AHLs QS control over the production of the pigment violacein (Durán et al., 2007). In addition, its chitinase activity was shown to be under QS control (Chernin et al., 1998), but besides these two phenotypes no other QS controlled factors are well described for this organism. The AHLs-QS system of C. violaceum consists of the protein CviI, which synthesizes the autoinducer C6-homoserine lactone (C6-HSL), and CviR, that is a cytoplasmic transcription factor that activates gene expression following binding to C6-HSL. The CviI/CviR system controls virulence, since QS inhibitors targeting CviR decrease C. violaceum-mediated killing of Caenorhabditis elegans (Swem et al., 2009; Chen et al., 2011). Recently, the consensus DNA sequence for promoter recognition by CviR was determined and 53 potential CviR binding sites were found in the C. violaceum genome; experimentally, it was determined that CviR binds and regulates transcription form six promoters including vioA, (violacein synthesis cluster), CV_4240, (chitinase) and cviI (HSL synthase), participating hence in a classical QS positive-feedback loop (Stauff and Bassler 2007). Since multiple C. violaceum genes are likely under the control of QS, we hypothesize that QS would influence several C. violaceum phenotypes. Hence, we evaluated the production of exoprotease, bacterial aggregation, biofilms, swarming motility, stress resistance as well as carbon and nitrogen sources utilization in the wild-type strain and the QS negative mutant CVO26. Our results indicate that these phenotypes are strongly influenced by QS. The identification of new QS controlled traits in C. violaceum improves our understanding of the basic biology of this pathogen and would allow researchers to assay other phenotypes besides violacein production to test the anti-QS properties of experimental compounds.

RESULTS

Exoprotease

Although it is mentioned that exoprotease production is controlled by QS in C. violaceum (Chernin et al., 1998), no experimental data are available. Hence in this work, the extracellular protease production of the wild-type strain (CV 12472) and the CV026 mutant (Sm^r mini-Tn5 Hg^rcviI::Tn5xylE Km^r) that is unable to produce C6-HSL was evaluated. Supernatants of both strains grown in LB medium were examined at 15, 24 and 48 h, concomitantly with violacein production as a positive control for QS [determined as in Castillo-Juarez et al. (2013)]. For exoprotease, collagen was used as substrate and the activity determined at pH 8.0 in Tris 20 mM and CaCl₂ 1 mM, following the protocol for the alkaline protease of Pseudomonas aeruginosa (Howe and Iglewski 1984). Collagenase activity was found in the wild-type supernatants and it increased with time, being maximal at 48 h. In contrast, the CV026 supernatants showed a very low protease activity at 15 h (∼3-fold lower than wild-type) and it remained in the same level at 24 and 48 h (10-fold lower than wild-type), while violacein was also lower than wild-type (∼3.5-fold lower). In order to corroborate that the exoprotease in C. violaceum is QS controlled, the quorum quencher furanone C-56 was added at 25 μM to the wild-type cultures assayed at 24 h and as expected protease activity was inhibited 65% and violacein by 56%; in addition, when 20 μM of exogenous HSL C-6 signal was added to the CV026 mutant at 24 h its violacein increased, reaching levels 72% higher than those of the wild-type strain and concomitantly its protease also increased to near the wild-type level (Fig. 1A and B), strongly suggesting that the exoprotease activity is regulated by QS. Moreover, in agreement with the collagenase assays, the wild-type strain developed a proteolysis halo in LB plates with 3% of skim milk after 24 h of incubation at 37°C, while for CV026 no appreciable halo was formed. To further characterize the proteolytic activity, wild-type supernatants at 24 h were assayed from pH 6 to 10, and its activity was maximal at pH 8.0 (Fig. S1, Supporting Information), being therefore an alkaline protease. Removing the divalent cations by adding 50 mM of EDTA or EGTA almost abolished the protease activity, indicating that the protease is metal dependent, similarly to other bacterial alkaline proteases (Gupta, Beg and Lorenz 2002); proteolysis was also inactivated by the serine protease inhibitor PMSF. The pH profile of the exoprotease and the effect of chelators and PMSF are shown in (Fig. S1, Supporting Information). The potential importance of the exoprotease for C. violaceum physiology was evidenced by the fact that it was able to survive and proliferate during several culture passes in minimal medium with caseinate as sole carbon source. In contrast to collagenase, elastolytic activity was not found under similar assay conditions.

Cell aggregation

Since aggregation and biofilm formation are promoted by QS in several bacteria (Waters and Bassler 2005), aggregation was determined after 15 h of incubation without shaking (Gonzalez Barrios et al., 2006). For the wild-type, 94 ± 1.5% of bacteria were aggregated, while for CV026 the aggregated cells were only 27 ± 5% and the addition of 20 μM of the C6-HSL signal to the mutant increased the aggregated cells to 63 ± 1.2%, while the addition of 25 μM of furanone C-56 to the wild-type strain reduced its aggregation to 79 ± 2% (Fig. 1B).

Biofilm formation

Similarly, biofilm formation after 48 h in glass tubes, determined by crystal violet staining (O'Toole et al., 1999), was 3.8 ± 0.9-fold greater for the wild-type strain than for CV026 and it was inhibited ∼50% by C-56 in the wild-type strain and increased to levels similar to wild-type in CV026 by the addition of C6-HSL, demonstrating that it is promoted by QS (Fig. 1C), in agreement with a recent report showing that the QS deficient strain VIR24 forms poor biofilm compared to the CV12472 wild-type strain and that biofilm is inhibited by quorum quenching (Liu et al., 2013).

Swarming motility

Similarly, swarming at 24 h was 2.3 ± 0.52-fold higher in wild-type than in CV026; it was inhibited 78 ± 6% by C-56 in the wild-type and restored close to wild-type levels by C6-HSL in the mutant (Fig. 1D).

Oxidative stress

Resistance against H₂O₂ was adapted from the protocol described in Zhang, Garcia-Contreras and Wood (2007) and it was 758 ± 369-fold higher in the wild-type than in CV026, decreased 1120 ± 891-fold by C-56 in the wild-type and increased 6 ± 4-fold by C6-HSL in the mutant (Fig. 1E).

Carbon and nitrogen metabolism

Although besides the adenosine catabolism (Heurlier et al., 2005), the utilization of some peptides as nitrogen sources, and of D-trehalose as carbon source in P. aeruginosa (Dandekar, Chugani and Greenberg 2013), few metabolic processes are known to be controlled by QS; since around half of the bacterial QS-controlled genes have not defined functions, it is possible that several metabolic processes are influenced by QS. To investigate such processes in C. violaceum, phenotypic arrays using the Biolog system were done. The plates PM1, PM2A (carbon sources), PM3B (nitrogen sources) and PM4A (phosphorus and sulfur sources) were used. Phenotypic arrays showed that for carbon sources (PM1 and PM2A), there were 8 compounds in which wild-type growth was 3-fold ≤ than CV026 and 16 compounds in which the CV026 growth was 3-fold ≤ than wild-type. For nitrogen sources (PM3B), wild-type growth was 3-fold ≤ than CV026 in six compounds and CV026 growth was 3-fold ≤ than wild-type only in one compound; in contrast, for phosphorus and sulfur sources (PM4A) no differences between wild-type and mutant were found (Table S1, Supporting Information). Suggesting that QS may influence the carbon and nitrogen metabolism in this bacterium, nevertheless, the observed differences in growth could be influenced by regulators independent of QS as well.

DISCUSSION

After demonstrating that QS influences multiple phenotypes in this bacterium, the next step would be the identification of the genes/proteins involved in such phenotypes, and the determination of the level of QS regulation. A clue for the identification of the QS regulated genes relies in the work of Stauff and Bassler (2011) which identified 53 genes containing putative CviR binding sites, among those genes is pilE2 which encodes a type IV pilus protein; hence, perhaps the regulation of these fimbriae is involved in the enhancement of QS over aggregation and biofilm formation, another QS controlled factor perhaps related to aggregation/biofilm formation is the fucose-binding lectin (CV-IIL) since its expression correlates with violacein production and is impaired in the CV026 mutant (Zinger-Yosovich et al., 2006). In addition, 20 genes with metabolic functions are putative CviR targets, suggesting that QS may strongly influence C. violaceum metabolism, as it was suggested by phenotypic microarrays here. Therefore, in this bacterium several private goods may be regulated by QS, which may be important for its ecology, since control of metabolic processes and stress resistance by QS restricts the appearance of social cheaters (Dandekar, Chugani and Greenberg 2012; García-Contreras et al., 2014). In addition, other putative CviR targets include nucleases and DNA repair proteins and genes related to competence (Stauff and Bassler 2011). If DNA repair and competence are also under QS control in this bacterium remain to be determined. In addition, the effect of QS regulated genes on C. violaceum virulence, their role in infections in animal models (we are currently performing testing its virulence in a mice acute infection model) as well as the elucidation of the factors that trigger C. violaceum virulence should be determined in order to understand better the pathogenic processes of this rare although potentially fatal pathogen.

SUPPLEMENTARY DATA

Supplementary Data.

We thank Professor Wilbert Bitter from the Department of Molecular Cell Biology, VU University Amsterdam, Amsterdam, for the provision of the strains. We thank Alejandra Guadalupe Villegas Pañeda for her assistance in some experiments.

FUNDING

RG-C research is funding by the SEP/CONACyT-México grant No. 152794 I C-J was supported by Fideicomiso COLPOS 167304 and Programa Cátedras-CONACyT 2112.

Conflict of interest statement. None declared.

REFERENCES