CpxR-Dependent Thermoregulation of Serratia marcescens PrtA Metalloprotease Expression and Its Contribution to Bacterial Biofilm Formation

ABSTRACT

PrtA is the major secreted metalloprotease of Serratia marcescens. Previous reports implicate PrtA in the pathogenic capacity of this bacterium. PrtA is also clinically used as a potent analgesic and anti-inflammatory drug, and its catalytic properties attract industrial interest. Comparatively, there is scarce knowledge about the mechanisms that physiologically govern PrtA expression in Serratia. In this work, we demonstrate that PrtA production is derepressed when the bacterial growth temperature decreases from 37°C to 30°C. We show that this thermoregulation occurs at the transcriptional level. We determined that upstream of prtA, there is a conserved motif that is directly recognized by the CpxR transcriptional regulator. This feature is found along Serratia strains irrespective of their isolation source, suggesting an evolutionary conservation of CpxR-dependent regulation of PrtA expression. We found that in S. marcescens, the CpxAR system is more active at 37°C than at 30°C. In good agreement with these results, in a cpxR mutant background, prtA is derepressed at 37°C, while overexpression of the NlpE lipoprotein, a well-known CpxAR-inducing condition, inhibits PrtA expression, suggesting that the levels of the activated form of CpxR are increased at 37°C over those at 30°C. In addition, we establish that PrtA is involved in the ability of S. marcescens to develop biofilm. In accordance, CpxR influences the biofilm phenotype only when bacteria are grown at 37°C. In sum, our findings shed light on regulatory mechanisms that fine-tune PrtA expression and reveal a novel role for PrtA in the lifestyle of S. marcescens.

IMPORTANCE We demonstrate that S. marcescens metalloprotease PrtA expression is transcriptionally thermoregulated. While strongly activated below 30°C, its expression is downregulated at 37°C. We found that in S. marcescens, the CpxAR signal transduction system, which responds to envelope stress and bacterial surface adhesion, is activated at 37°C and able to downregulate PrtA expression by direct interaction of CpxR with a binding motif located upstream of the prtA gene. Moreover, we reveal that PrtA expression favors the ability of S. marcescens to develop biofilm, irrespective of the bacterial growth temperature. In this context, thermoregulation along with a highly conserved CpxR-dependent modulation mechanism gives clues about the relevance of PrtA as a factor implicated in the persistence of S. marcescens on abiotic surfaces and in bacterial host colonization capacity.

INTRODUCTION

Serratia marcescens belongs to the Enterobacteriaceae family and can be isolated from a wide variety of environmental niches, ranging from water and soil to air. In addition to its environmental ubiquity, S. marcescens is an emergent health-threatening nosocomial pathogen. In recent years, outbreaks of multidrug-resistant strains and high incidence in intensive and neonatal care units have increasingly been reported (1–3). The World Health Organization recently declared S. marcescens, together with other enterobacteria, a research priority target to develop alternative antimicrobial strategies given the high frequency of carbapenem-resistant clinical isolates (4). In a recent work, Hoarau et al. (5) also identified S. marcescens as one of the three most abundant microbial species that colonize the dysbiotic gut of Crohn's disease patients, in detriment to beneficial bacteria. S. marcescens can also develop either symbiotic or pathogenic interactions with plants and insects (6).

The ability of S. marcescens to adapt to and survive in both hostile and changing environments also relates to the bacterial capacity to express a wide range of secreted enzymes, including lipases, phospholipases, chitinases, proteases, and nucleases (6). Typically, the catalytic action of secreted effectors allows bacteria to perform vital tasks outside or within a host, such as conversion of environmental compounds in usable nutrient sources, adherence to surfaces, and breaching of host protective barriers along colonization pathways or manipulation of intracellular traffic pathways within invaded cells. The expression of these multiple effectors needs to be coordinated at the right time and in the right place to allow bacteria to thrive and shift between different lifestyles.

PrtA, also named serralysin or PrtS (7), is a 50.2-kDa repeats-in-toxin (RTX) alkaline zinc metalloprotease that has been well characterized structurally (8) and was shown to depend on the type I LipBCD system for secretion to the extracellular milieu (9, 10). Similarly to proteases that belong to the serralysin family in other bacteria, such as Photorhabdus, Erwinia, or Pseudomonas, S. marcescens PrtA has been implicated in cytotoxicity, immunomodulation, or virulence traits using different experimental models, including nematodes, insects, and mice (11–17). In addition, the potential application of the enzymatic properties of PrtA, for instance, as a detergent additive, has attracted industrial interest, and different purification strategies and the optimization of its catalytic activity have been reported (15, 18–22). In several countries, purified serralysins are also commercially available as potent analgesic and anti-inflammatory drugs (23).

One of the most conspicuous mechanisms that pathogenic bacteria display to tightly regulate the expression of bacterial effectors in response to either extra- or intrahost challenges are the so-called two-component systems (TCS). In these signal transduction systems, the activation of a sensor histidine kinase leads to autophosphorylation followed by transfer of the phosphoryl group to a cognate response regulator in an aspartate residue. CpxAR is a canonical TCS, broadly conserved among many pathogenic and nonpathogenic bacteria. In most of these organisms, CpxAR regulates gene expression to counteract stressful conditions that menace bacterial envelope intactness. CpxA, a histidine kinase, can detect a variety of stimuli that range from pH alterations and overexpression of envelope proteins (such as NlpE or pilus subunits) to toxic concentrations of metal ions (24–27). Upon phosphate transfer from the CpxA sensor, the phosphorylated cognate response regulator CpxR generates the output response by driving the transcriptional expression of numerous genes. The Cpx regulon comprises both evolutionarily conserved genes, such as those for the protease/chaperone DegP (25), the disulfide bond oxidoreductase DsbA (24), and CpxP, which functions as both a chaperone and a repressor of the Cpx response (28), as well as species-specific genes involved in diverse phenotypes, including antibiotic resistance, motility, and biofilm formation (29, 30).

Biofilm is an orchestrated collective form of bacterial life that relies on the timely production of an intricate extracellular mesh that structurally and functionally supports its formation. This extracellular matrix can be composed of self-produced exopolysaccharides, proteins, lipids, DNA, and bacterial appendices such as fimbria and flagella. Previous work has shown that S. marcescens is able to develop biofilm associated with either biotic or abiotic surfaces (31–34). This ability is associated with the capacity of Serratia to colonize and persist in medical devices, such as catheters or prostheses (35), and to enhance bacterial resistance to antibiotics (36).

In this work, we characterized regulatory aspects that govern S. marcescens PrtA production. We found that PrtA expression is subject to transcriptional regulatory mechanisms that involve the bacterial growth temperature and the action of the Cpx signal transduction system. In addition, we reveal that PrtA expression influences the ability of Serratia to develop a mature biofilm.

RESULTS AND DISCUSSION

The expression of PrtA, the major S. marcescens secreted protease, is thermoregulated at the transcriptional level.

Phenotypes influenced by temperature have been previously observed in Serratia, such as flagellum-dependent swimming or swarming motility (37, 38), the generation of outer membrane vesicles (OMVs) (39), or the production of metalloproteases and serine proteases in certain strains, such as Serratia sp. strain SCBI (38, 40). To examine whether exoprotease expression is affected by bacterial growth temperature in S. marcescens clinical strain RM66262 (41), we first determined total proteolytic activity in the culture supernatant after growing S. marcescens at different temperatures in the range of 18 to 37°C. Proteolytic activity was quantified in S. marcescens culture supernatants by using azocasein as the substrate, as described previously (42). As shown in Fig. 1A, a 68.3% decrease of total proteolytic activity was observed when bacterial growth temperature was increased from 30°C to 37°C. According to previous work, PrtA is responsible for the main proteolytic activity that S. marcescens secretes into the culture medium (43). Therefore, we constructed a prtA null mutant and evaluated protease activity in this strain. As shown in Fig. 1B, in comparison with wild-type (wt) levels, proteolytic activity dramatically decreased in the mutant strain at the two temperatures tested (83.6% at 30°C and 58.7% at 37°C). Neither the wild-type nor the prtA strain showed alterations in growth capacity at the tested temperatures (data not shown). The decreased proteolytic activity of the prtA strain and the observed modulation by temperature were also qualitatively detected by the width of clear halos around the colony that result from proteolytic activity when strains are grown in LB agar-skim milk plates (Fig. 1B). The relative abundance of PrtA among proteins secreted by Serratia and its temperature-dependent production were examined by SDS-PAGE analysis of the bacterial culture supernatant (Fig. 1C). The identity of the protein labeled PrtA was confirmed by excision of the band from the polyacrylamide gel, followed by enzymatic digestion with trypsin and tandem mass spectrometry (MS/MS), as detailed in Materials and Methods. The absence of PrtA expression in the prtA mutant strain could be also observed by SDS-PAGE analysis at both temperatures tested (Fig. 1C). The remnant protease activity detected in the prtA background (Fig. 1B) can be attributed to the expression of other proteases. Indeed, when we performed an in silico analysis of the S. marcescens RM66262 genome, in addition to prtA, we found that it includes slpC, slpD, and slpE homologues (44, 45), all of them containing the N-terminal HEXXHXXGXXH domain and RTX repeats, the main features of serralysin family proteins.

FIG 1.

Dependence of S. marcescens RM66262 PrtA on the bacterial growth temperature. (A) Secreted protease activity from the S. marcescens wild-type strain grown in LB medium for 16 h at the indicated temperatures. (B) Secreted protease activities from wild-type, prtA, and lipB strains grown for 16 h at the indicated temperatures. Below the graph, representative LB agar-skim milk plate images show protease degradation halos for each strain. For both panels A and B, protease activity was measured by the azocaseinase assay. Activity results are expressed as percentages relative to the value of the wild-type strain grown at 30°C. Means ± SDs from three independent experiments are shown. Significant differences in activity by unpaired t test between the indicated strains are shown as follows: ***, P < 0.001, and nd., not detected. (C) SDS-PAGE analysis of extracellular proteins from the wild-type, prtA, and lipB strains grown in LB medium for 16 h at 30°C or 37°C, as indicated. The arrowheads indicate the positions of the PrtA protein. Molecular mass standards are indicated to the right.

It has been previously shown that the LipBCD type I secretion system mediates the extracellular export of PrtA (9). As shown in Fig. 1B, inactivation of lipB resulted in the abrogation of secreted proteolytic activity when bacteria were grown either at 30°C or at 37°C. As expected, the absence of the PrtA band was observed in the lipB strain culture supernatant (Fig. 1C). These results indicate that all detectable extracellular proteolytic activities of S. marcescens RM66262 strain rely on a functional LipBCD type I secretion system for their secretion.

At either 30°C or 37°C, complementation of the prtA mutant defect in protease secretion was achieved by expression of prtA under the control of a constitutive promoter from the pBB2::prtA plasmid (Fig. 2A). Temperature dependence on secreted protease activity was still observed in the prtA complemented strain. In order to further analyze this result, we determined prtA transcriptional levels by quantitative real-time PCR (qRT-PCR). As shown in Fig. 2B, prtA transcript levels were 29-fold higher at 30°C versus 37°C in the wild-type strain, indicating that thermoregulation occurs at the transcriptional level. In contrast to secreted protease activity results, the complemented prtA/pBB2::prtA strain reached the same transcriptional expression values irrespective of the bacterial growth temperature. These results suggest that (i) the difference in steady-state levels of prtA transcript between 30°C and 37°C is not related to differential, temperature-dependent mRNA stability, as equivalent levels were measured from the constitutive pBB2::prtA expression plasmid, and (ii) the failure to lose thermal modulation of the secreted activity by constitutive expression of prtA indicates additional thermoregulated factors that influence PrtA expression.

FIG 2.

(A) Complementation of the prtA mutant by in trans expression of prtA. Protease activity was measured by the azocaseinase assay. The results of proteolytic activity are expressed as percentages relative to the value of the wt/pBB2 strain grown in LB for 16 h at 30°C. Below the graph, representative LB agar-skim milk plate images show protease degradation halos for each strain. Means ± SDs from three independent experiments are shown. Significant differences in activity relative to the wt/pBB2 strain grown at 30°C calculated by one-way analysis of variance (ANOVA) with Bonferroni's multiple-comparison test are indicated as follows: ***, P < 0.001. (B) qRT-PCR analysis of prtA transcriptional expression. The value obtained for the wt/pBB2 strain grown in LB medium for 16 h at 30°C was taken as the reference value. mRNA levels were normalized to 16S rRNA expression. Relative expression was calculated using the 2^−ΔΔCT method. Means ± SDs from three independent experiments are shown. Significant difference versus reference value calculated by paired t test is indicated as follows: ***, P < 0.001. (C) Transcriptional expression of prtA and lipBCD. Bacteria were grown for 16 h in LB medium at the indicated temperatures. Transcriptional activity was calculated as the ratio of GFP fluorescence and OD₆₀₀ (FU/OD₆₀₀) measured from the wild-type strain carrying the PprtA-gfp or PlipBCD-gfp reporter plasmids. Means ± SDs from three independent experiments are shown. Significant differences between growth temperatures calculated by unpaired t test are indicated as follows: ***, P < 0.001.

To further examine the mechanism that underlies thermoregulation of PrtA expression, we constructed the PprtA-gfp reporter plasmid harboring the gfp gene, which encodes green fluorescent protein (GFP) under the transcriptional control of the prtA promoter region (443 bp upstream of the translational ATG start codon of prtA). As shown in Fig. 2C, transcriptional activity was 246% higher when bacteria were grown at 30°C versus 37°C. This result indicates that a thermoregulatory mechanism targets the prtA promoter region. To analyze if expression of the dedicated transporter LipBCD was also affected by growth temperature at the transcriptional level, we measured fluorescence from a PlipBCD-gfp transcriptional reporter, which harbors 500 bp corresponding to the putative lipBCD promoter region fused to the gfp gene. We found that in the wild-type strain, after 16 h of growth at 30°C, fluorescence reached a value 31% higher than the one obtained at 37°C (Fig. 2C).

Therefore, we can infer that increased PrtA expression at low temperatures needs to be coordinated with an enhanced secretion capacity, achieved by higher levels of expression of the Lip transporter system. This result also indicates that LipBCD expression would differentially limit PrtA secretion at 30°C or 37°C and explains why even when the transcript is constitutively expressed in the prtA/pBB2::prtA complemented strain, we still observe a temperature-dependent effect in secreted protease activity levels. Overall, these results show that expression of PrtA, which is responsible for the major secreted proteolytic activity in S. marcescens, is modulated by the bacterial growth temperature at the transcriptional level.

CpxR directly binds to a conserved motif within the prtA promoter.

To gain further insight into the mechanism that underlies prtA regulated expression, we performed a bioinformatic search in the putative promoter region of prtA, comprising 443 bp upstream of the translational ATG. The use of the MEME/FIMO (46, 47) predictive tools showed a putative CpxR-binding site 216 bp upstream of the prtA ATG translational initiation codon. This CpxR-binding sequence showed 14 out of 16 conserved bases in comparison with the consensus motif constructed by the search engine using a series of previously identified promoter regions that contain CpxR bona fide recognition sequences (Fig. 3). Extending the search to annotated Serratia marcescens genomes, we found that the CpxR-binding motif was highly conserved in all examined prtA promoter regions irrespective of the clinical, enthomopathogenic, or environmental origin of the isolate, and it was even detected in the Serratia liquefaciens and Serratia proteamaculans genomes (Fig. 3B).

FIG 3.

In silico analysis of CpxR-binding site. (A) The logo shows the DNA consensus motif for the CpxR-binding site obtained by MEME/FIMO bioinformatics tools. (B) CpxR-binding motif is conserved in Serratia prtA promoter regions. The predicted CpxR-binding site sequence in the putative promoter region of S. marcescens RM66262 prtA is underlined. DNA sequences deposited in the NCBI database under the following numbers were used for the analysis of putative CpxR-binding sites within the prtA promoter region in Serratia strains: NZ_JWLO00000000.1 (S. marcescens RM66262), NZ_HG738868.1 (S. marcescens SMB2099), NZ_CP012685.1 (S. marcescens UNAM836), NZ_AP013063.1 (S. marcescens SM39), NZ_CP016032.1 (S. marcescens U36365), NZ_CP011642.1 (S. marcescens CAV1492), NZ_HG326223.1 (Serratia marcescens subsp. marcescens Db11), NC_020211.1 (S. marcescens WW4), NZ_CP005927.1 (Serratia sp. FS14), NC_021741.1 (Serratia liquefaciens ATCC 27592), and NC_009832.1 (Serratia proteamaculans 568). The type of isolate is indicated as follows: asterisk, clinical; solid dot, enthomopathogenic; open dot, environmental.

To assess whether CpxR is able to directly interact with the putative regulatory region upstream of prtA, electrophoretic mobility shift (EMSA) and DNase I footprinting assays were performed. For EMSA, a PCR-amplified fragment derived from the prtA promoter region encompassing 443 bp upstream of the translational ATG and purified 6×His-tagged CpxR (CpxR-6×His) protein were used. Phosphorylation of CpxR was achieved by use of acetyl phosphate as a phosphoryl donor. A retarded band was detected when 20 pmol of phosphorylated CpxR was used (Fig. 4A), indicating that, as predicted, CpxR is able to recognize a CpxR-binding motif within the fragment. A 10- to 162-fold excess of competing nonspecific nucA DNA fragment (a 436-bp DNA region that codes for the S. marcescens NucA nuclease) did not affect the interaction (Fig. 4A, left), while the shifted band was progressively lost when increasing amounts of unlabeled prtA promoter fragment were included in the mixture (Fig. 4A, right), indicating that the interaction of CpxR with the prtA promoter region is specific. To determine whether CpxR phosphorylation status affected the capacity for DNA binding, we compared the mobility shift abilities of CpxR when it was preincubated in the absence and presence of acetyl phosphate. At least 10 pmol of nonphosphorylated CpxR was required to shift the mobility of the DNA probe, while 5 pmol of the phosphorylated form was enough to exert the effect, indicating that phosphorylation enhances CpxR binding affinity for its target sequence in the DNA (Fig. 4B). For the DNase I protection assay, CpxR-6×His previously incubated with acetyl phosphate and labeled DNA fragments containing either coding or noncoding sequences 443 bp upstream of the translational ATG start of prtA were used. As shown in Fig. 4C, CpxR-6×His protected an overlapping region from nucleotides −216 to −241 relative to the prtA translational ATG start site. This protected region overlapped the in silico-predicted CpxR consensus binding motif sequence (Fig. 3).

FIG 4.

CpxR interaction with a CpxR-binding motif within the S. marcescens prtA promoter region. (A) Electrophoretic mobility shift assays (EMSAs) were performed using 20 pmol of purified CpxR-6×His. Target DNA was a ³²P-labeled PCR fragment that included the prtA promoter region (PprtA). Binding specificity was assessed by competition reactions in which increasing amounts (42, 85, 170, 340, 510, and 680 ng) of nonspecific (nucA; left) or specific (PprtA; right) unlabeled DNA template competed with labeled DNA for binding to CpxR-6×His. (B) EMSAs were performed using a nonlabeled PCR fragment containing PprtA and purified CpxR-6×His (at increasing amounts of 2.5, 5, 10, 20, or 40 pmol, as indicated) preincubated (+) or not preincubated (−) with 25 mM acetyl phosphate (AcP). (C) DNase I footprinting analysis was performed on coding and noncoding strands of the prtA promoter region. The DNA fragments were incubated with 0 (−) or 20 (+) pmol of purified CpxR-6×His. DNA ladder sequences (A, G, and T) are shown. The gels were sliced (noncontiguous lanes from a single gel are indicated by dashed lines) to exclude nonoptimal protein concentrations assayed or ladder lanes that resulted in smeared patterns. The nucleotide sequences of the CpxR-protected regions are indicated, and the protected DNA regions are underlined. A dotted line indicates the overlap of the protected sequences in each strand.

CpxR transcriptionally regulates PrtA expression in a temperature-dependent manner.

In light of our results and to analyze CpxR influence on PrtA expression, a cpxR mutant strain was used and secreted proteolytic activity was assayed. At 30°C, the activity was not significantly affected by cpxR inactivation (Fig. 5A, left side of graph). However, at 37°C, a 40% increase in secreted proteolytic activity was obtained in the cpxR mutant strain compared to that in the wild-type strain (Fig. 5A, right side of graph). Complementation of cpxR inactivation in trans by expression of CpxR from the pBB5::cpxR plasmid restored the activity of the mutant strain to wild-type levels at 37°C, corroborating the cpxR-dependent phenotype. No significant differences were obtained at 30°C (Fig. 5A). These results were also reflected in the proteolytic halos detected when bacteria were grown in LB agar-skim milk plates (Fig. 5A, bottom). In agreement, the analysis of secreted protein profiles of the wild-type and cpxR strains by SDS-PAGE showed that only at 37°C was cpxR inactivation able to affect PrtA expression levels, being increased in the cpxR background (Fig. 5B). To verify that CpxR modulates PrtA expression at the transcriptional level, we measured fluorescence from wild-type and cpxR strains harboring the PprtA-gfp reporter plasmid. In good correlation with the results described above, inactivation of cpxR caused a 76% increase—in comparison with those of the wild-type strain—in fluorescence levels when bacteria were cultured at 37°C, while no significant difference was observed after growth at 30°C (Fig. 5C, left graph). In addition, lipBCD transcriptional activity was not influenced by cpxR inactivation (Fig. 5C, right graph). These results indicate that at 37°C, PrtA expression levels achieved in the cpxR strain are accomplished by derepression of prtA and not by variation in lipBCD expression. These results also indicate that thermal modulation of lipBCD transcriptional activity (as shown in Fig. 2C) might be under the control of a cpxR-independent mechanism.

FIG 5.

CpxR-dependent modulation of PrtA expression. (A) Secreted proteolytic activity from the wild-type or cpxR strains was measured by the azocaseinase assay. Results are expressed as percentages relative to the value obtained for the wt/pBB5 strain grown at 30°C. Below the graph, representative LB agar-skim milk plate images show protease degradation halos for each strain. Bacteria were grown for 16 h in LB medium supplemented with 15 μg/ml of gentamicin at the indicated temperatures. (B) SDS-PAGE analysis of extracellular proteins from wild-type and cpxR and prtA mutant strains, grown for 16 h at 30°C or 37°C. The arrowhead indicates the position of PrtA protein in the 37°C sample. Molecular mass standards are indicated to the right. (C) Transcriptional expression of prtA and lipBCD. Strains were grown in LB medium for 16 h at the indicated temperatures. Transcriptional activity was calculated as the ratio of GFP fluorescence and OD₆₀₀ (FU/OD₆₀₀) measured from wild-type or cpxR strains carrying the PprtA-gfp or PlipBCD-gfp reporter plasmids. For panels A and C, means ± SDs from three independent experiments are shown. Significant differences by one-way ANOVA calculated with Bonferroni's multiple-comparison test are indicated as follows: ***, P < 0.001; **, P < 0.01; and ns., no significant difference.

Together, these results allow us to postulate a regulatory model for PrtA expression in which the CpxAR system is activated by temperature, being upregulated when bacteria are grown at 37°C. By direct interaction with the CpxR recognition motif located in the prtA promoter region, activated CpxR exerts a repressing action over prtA transcriptional activity.

Bacterial growth temperature defines the activity status of the Serratia CpxAR system.

Our results indicate that CpxR-mediated repression over prtA is physiologically exerted at 37°C but not at 30°C, suggesting that the activity status of the CpxAR system is influenced by the bacterial growth temperature. Because there were no previous reports about the temperature impact on CpxAR system activity, and to further explore this phenotype, we examined by qRT-PCR the transcriptional level of cpxP from Serratia grown at 37°C or at 30°C. cpxP is a well-known member of the CpxAR regulon, which is transcriptionally activated by CpxR and is present in a group of Enterobacteriaceae, including Salmonella, Yersinia, and Serratia (41, 48). CpxP also serves as an auxiliary protein to regulate the CpxAR system in a negative-feedback loop by preventing autophosphorylation of CpxA through CpxP binding (28).

As shown in Fig. 6A, transcriptional levels of cpxP increased 68-fold when Serratia was grown at 37°C compared to the values obtained at 30°C, while low expression levels could be detected in the cpxR background. This result reveals that the CpxAR-dependent induction over cpxP transcription is favored at 37°C, indicating that the S. marcescens CpxAR system is activated under this growth condition. In light of this result, we also determined cpxP transcriptional activity in a Salmonella enterica serovar Typhimurium 14028s strain that chromosomally harbors a cpxP-lacZ transcriptional reporter. Bacterial growth at 37°C mildly increased Salmonella cpxP expression in comparison with that at 30°C (Fig. 6B), suggesting that in addition to S. marcescens, temperature could be an environmental cue for CpxAR systems in other enterobacteria.

FIG 6.

Growth temperature influences the activity status of the CpxAR system. (A) qRT-PCR analysis of cpxP expression. The wild-type and cpxR strains were grown in LB medium for 16 h at the indicated temperatures. Values obtained for the wild-type strain grown at 30°C were taken as the reference values. mRNA levels were normalized to the 16S rRNA gene, and relative expression was calculated using the 2^−ΔΔCT method. Means ± SDs from three biological replicates are shown. Significant differences versus reference condition calculated by paired t test are indicated as follows: *, P < 0.05, and ***, P < 0.001. (B) β-Galactosidase activity from a cpxP-lacZ transcriptional fusions expressed in wild-type and ΔcpxR Salmonella enterica serovar Typhimurium 14028s strains. Bacterial cultures were grown in LB medium for 16 h at the indicated temperatures. Means ± SDs from three biological replicates are shown. Significant differences versus wild-type strain grown at 30°C by one-way ANOVA with Bonferroni's multiple-comparison test are indicated as follows: ***, P < 0.001.

To further verify the regulatory action of the CpxAR system on PrtA expression, we also assayed a CpxAR-activating condition previously described for other enterobacteria. NlpE is an outer membrane lipoprotein that, when overexpressed, leads to CpxAR activation (27). Therefore, we transformed the S. marcescens strains with the pSU::nlpE expression plasmid and we measured proteolytic activity from the resultant strains grown either at 30°C or at 37°C. As shown in Fig. 7A, either by azocasein hydrolysis assay or by detection of proteolytic halos, NlpE overexpression resulted in a severe decrease of secreted proteolytic activity at both temperatures. At 30°C, NlpE-dependent repression was CpxR dependent, as we obtained a 63.5% reduction in the wild type but no significant effect in the cpxR strain, indicating that the inhibition is due to an activated CpxAR pathway. At 37°C, NlpE overexpression caused an 89.7% decrease in relation to the secreted proteolytic activity in the wild-type strain. However, we observed that NlpE overexpression also lowered proteolytic activity, by 51.7%, in the cpxR strain. In addition to potential CpxR-independent pathways that can be responsible for this last result, NlpE overexpression at 37°C caused a growth defect in all tested strains (Fig. 7B), indicating that at this temperature, NlpE expression also exerts a stressful, detrimental effect over S. marcescens growth capacity.

FIG 7.

Effect of NlpE overexpression on CpxAR activation. (A) The indicated strains were grown in LB medium for 16 h at 30°C or 37°C. Secreted protease activity was measured by the azocaseinase assay. For each temperature, activity results are expressed as percentages relative to the value obtained for the wt/pSU strain. Below the graphs, representative LB agar-skim milk plate images show protease degradation halos for each strain. (B) Optical density from the bacterial cultures used for panel A was measured at 600 nm (OD₆₀₀). Means ± SDs from three independent experiments are shown. Significant differences in activity calculated by unpaired t test between the indicated strains are shown as follows: ***, P < 0.001, and ns., no significant difference.

Together, our results allow us to propose that in S. marcescens, high temperature activates CpxAR signal transduction. This would result in increased levels of phosphorylated CpxR, a condition that enhances CpxR affinity for its recognition motif within prtA promoter. CpxAR system activation is also achieved by NlpE overexpression. Phospho-CpxR binding would result in repression of prtA transcription. In contrast, lower temperatures would decrease the levels of phosphorylated CpxR, leading to derepression of the transcriptional activity from the prtA promoter (Fig. 8 shows a scheme that depicts our working model). As shown in Fig. 5, in the absence of a CpxAR-inducing condition, overexpression of CpxR is not able compensate for the requirement of activated, phosphorylated CpxR to downregulate prtA expression.

FIG 8.

Proposed model for the regulation of PrtA expression. At 30°C (left), the CpxAR system is noninduced, yielding low levels of CpxR-P. High prtA transcript levels correlate with large amounts of PrtA detected in the extracellular space, and hence, high proteolytic activity levels are achieved. A growth temperature of 37°C (right) represents an inducing cue for CpxAR, increasing CpxR-P levels, which, in turn, repress PrtA expression by direct binding to prtA promoter region. PrtA expression decreases in comparison to that at 30°C. Still-uncovered regulatory mechanisms (represented by “X”) might contribute to activate PrtA gene transcription at 30°C and/or repress its expression at 37°C.

PrtA expression influences S. marcescens biofilm formation capacity.

S. marcescens displays biofilm formation capacity that has been shown to be related to its ability to colonize, persist, and proliferate on either biological or inert surfaces. S. marcescens biofilms form successfully on periodontal tissues (49), contact lenses (50), neonatal feeding tubes (51), and catheters, prostheses, and other medical devices (3, 6). S. marcescens also forms biofilms on corals (52, 53) and over zygomycete mycelium (54). More recently, the capacity of S. marcescens to generate mixed biofilms together with Escherichia coli and Candida tropicalis was correlated with the abundance of these three microorganisms in dysbiotic Crohn's disease patients (5). Taking into account that in other enterobacteria, the CpxAR pathway is involved in modulating the ability of biofilm generation (55–57), we examined whether PrtA could influence S. marcescens biofilm formation capacity.

To that aim, we performed in vitro biofilm assays in polystyrene microwell plates, followed by biofilm quantitation using crystal violet staining as detailed in Materials and Methods. As shown in Fig. 9A, when the strains were grown at 30°C in either SLB medium (peptone at 10 g/liter and yeast extract at 5 g/liter)—a low-osmolarity condition that has been shown to enhance biofilm formation in other Enterobacteriaceae (57)—or LB medium, the lack of PrtA expression in the prtA strain reduced 56.9% or 38.2%, respectively, the capacity of the bacteria to form biofilm compared with that of the wild-type strain. In agreement with PrtA expression being downregulated at 37°C versus 30°C, prtA deficiency in biofilm formation was more attenuated at 37°C, showing a reduction of 30.2% in SLB medium or 31.2% in LB medium (Fig. 9A). Under all conditions, the prtA defective biofilm phenotype was restored to wild-type levels by PrtA expression in trans from the pBB2::prtA plasmid (Fig. 9A). These results demonstrate that PrtA expression contributes to the ability of S. marcescens to structure a biofilm community.

FIG 9.

(A) PrtA involvement in S. marcescens biofilm phenotype; (B) CpxR influence on biofilm formation. Strains were grown in SLB or LB culture medium at 30 or 37°C for 48 h in 96-well microtiter plates. The adhered biofilm was measured by crystal violet staining. Results are expressed as percentages relative to the values obtained for the wt/pBB2 (A) or wt/pBB5 (B) strain. Means ± SDs from three independent experiments are shown. Significant differences in activity calculated by one-way ANOVA with Bonferroni's multiple-comparison test are indicated as follows: ***, P < 0.001, and ns., no significant difference.

In addition, and in correlation with CpxR inhibitory action over PrtA expression at 37°C, cpxR inactivation enhanced 37.9% (in SLB medium) and 77.0% (in LB medium) the capacity of the wild-type strain to form biofilm at 37°C. In trans expression of CpxR from pBB5::cpxR reestablished the biofilm formation ability of the cpxR mutant to wild-type levels. As expected, at 30°C, cpxR inactivation did not show any effect on the biofilm phenotype when S. marcescens strains were grown either in LB medium or in SLB medium (Fig. 9B). This result reinforces the conclusion that CpxR-regulated PrtA expression is involved in Serratia biofilm development. Because we observed that PrtA expression was not associated with S. marcescens autolysis (data not shown) as was shown to be the case for secreted GelE metalloprotease in Enterococcus faecalis (58), we can discard the possibility that PrtA-dependent liberation of extracellular DNA could act as the modulator of the formation or maintenance of S. marcescens biofilm. A recent work by Selan and colleagues (59) showed that the S. marcescens ATCC 21074 PrtA homologue was able to impair the capacity of Staphylococcus aureus to attach to an inert matrix and develop biofilm. However, this ability was retained in a single-amino-acid-mutant protein that annuls the protease hydrolytic capacity, showing that the biofilm interference phenotype was independent of the catalytic activity of the protein. Together with our results, these findings allow us to hypothesize alternative scenarios that are the object of our ongoing research: (i) PrtA can convert an inert substrate into a functional one that enhances biofilm formation, (ii) PrtA could favor S. marcescens biofilm formation by exerting its hydrolytic activity over a substrate that is detrimental for biofilm formation, or (iii) PrtA harbors structural properties that promote adherence of S. marcescens to the attachment surfaces or aid the stability of the extracellular matrix that allows cell-cell interactions in the biofilm assembly process.

Concluding remarks.

We here show novel insights into the regulation of PrtA expression at the transcriptional level. An increase in PrtA expression in response to temperature decrease allows us to hypothesize that metalloprotease activity is required for the interaction with either the external ambient temperature or unregulated-body-temperature hosts. Downregulation of PrtA expression levels at 37°C would favor Serratia in the transition to mammalian niches. Thermoregulation would allow colonization of regulated-body-temperature hosts by limiting the cytotoxic capacity of PrtA toward mammalian cells (60). At this temperature, CpxR-dependent repression would strengthen PrtA downregulation in the presence of conditions related to CpxAR-mediated surface sensing or the detection of threats to bacterial envelope integrity.

We showed that involvement of PrtA in biofilm formation is not constrained by environmental temperature, which is indicative of a relevant role for this protein in the capacity of Serratia to form multicellular communities in a diversity of extra- or intrahost niches. We have previously demonstrated that in S. marcescens, the biogenesis of outer membrane vesicles (OMVs) is thermoregulated and that PrtA forms part of OMV cargo (39). Like PrtA expression, OMV production is enhanced at temperatures below 30°C. Because it has been observed that OMVs can integrate bacterial biofilm structures (61), we can conjecture that PrtA could enhance bacterial community formation not only as a secreted soluble protein but also by exerting its action as a component of OMVs.

Previous findings showed that PrtA is able to breach host barriers or alter immune defenses: it was reported to be an agent capable for the hydrolysis of several host substrates, such as the Hageman factor/kallikrein-kinin coagulation system (62–65), the components of human complement (60), the heavy chains of IgG and IgA immunoglobulins (66), and the rip-trialysin antimicrobial peptide present in insect salivary fluids (67). Conceivably, PrtA displays a dual role as biofilm enhancer factor that, at the same time, leads to attainment of a localized threshold concentration efficacious to exert the enzymatic action over host components, facilitating colonization and invasion processes.

MATERIALS AND METHODS

Bacterial strains and plasmids.

Serratia marcescens RM66262 is a nonpigmented clinical isolate from a patient with a urinary tract infection (41). The strains and plasmids used in this study are listed in Table 1.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Genotype and/or comments	Reference or source
Strains
S. marcescens
Wild type	RM66262; clinical isolate	41
prtA strain	prtA::pKNOCK-Cm	This work
lipB strain	lipB::pKNOCK-Gm	This work
cpxR strain	cpxR::pKNOCK-Cm	78
wt/pBB2	Wild type/pBBR1-MCS2 Kmr	76
prtA/pBB2 strain	prtA::pKNOCK-Cm/pBBR1-MCS2 Kmr	This work
prtA/pBB2::prtA strain	prtA::pKNOCK-Cm/pBBR1-MCS2::prtA Kmr	This work
wt/pBB5	Wild type/pBBR1-MCS5 Gmr	This work
wt/PprtA-gfp	Wild type/pPROBE-NT [ASV]::PprtA	This work
cpxR/PprtA-gfp strain	cpxR/pPROBE-NT [ASV]::PprtA	This work
wt/PlipBCD-gfp	Wild type/pPROBE-OT::PlipBCD	This work
cpxR/PlipBCD-gfp strain	cpxR/pPROBE-OT::PlipBCD	This work
cpxR/pBB5 strain	cpxR::pKNOCK-Cm/pBBR1-MCS5 Gmr	This work
cpxR/pBB5::cpxR strain	cpxR::pKNOCK-Cm/pBBR1-MCS5::cpxR Gmr	This work
wt/pSU	Wild type/pSU36 Kmr	This work
wt/pSU::nlpE	Wild type/pSU36::nlpE Kmr	This work
cpxR/pSU strain	cpxR::pKNOCK-Cm/pSU36 Kmr	This work
cpxR/pSU::nlpE strain	cpxR::pKNOCK-Cm/pSU36::nlpE Kmr	This work
E. coli
One Shot TOP10	F− mcrA Δ(mrr-hsdRMS-mcrBC) ϕ80lacZΔM15 ΔlacX74 nupG recA1 araD139 Δ(ara-leu)7697 galE15 galK16 rpsL endA1 Smr	Invitrogen
SM10 λpir	thiJ thr leu tonA lacY 61lic recA::RP4-2-Tc::Mu λpir Kmr
M15/pRep4	F− ϕ80lacZΔM15 thi lac mtl recA+ placI Kmr	Qiagen
Salmonella enterica serovar Typhimurium
Wild type	14028s cpxP-lacZ	79
ΔcpxR strain	14028s ΔcpxRA cpxP-lacZ	80
Plasmids
pGEM-T::prtA	PCR-amplified prtA coding sequence and flanking regions cloned in pGEM-T; Ampr	This work
pBB2	pBBR1-MCS2 Kmr; broad host range	81
pBB2::prtA	pBBR1-MCS2::prtA Kmr	This work
pGEM-T::PprtA	PCR-amplified prtA promoter region cloned in pGEM-T; Ampr	This work
PprtA-gfp	pPROBE-NT [ASV]::PprtA	This work
PlipBCD-gfp	pPROBE-OT::PlipBCD	This work
pGEM-T::cpxR	PCR-amplified cpxR coding sequence cloned in pGEM-T; Ampr	This work
pBB5	pBBR1-MCS5 Gmr; broad host range	81
pBB5::cpxR	pBBR1-MCS5::cpxR Gmr	This work
pSU	pSU36 Kmr; derived from pACYC184	82
pSU::nlpE	pSU36::nlpE	This work
pQE32::cpxR	Expression vector for CpxR-6×His	This work

Insertion mutations in prtA and lipB were constructed by cloning internal regions of each gene into pKNOCK suicide plasmids (68). For constructing the prtA strain, primers prtA left-Fw and prtA right-Rv were used to PCR amplify from the chromosome a 2,556-bp DNA region encompassing the prtA gene, which was subsequently cloned into pGEM-T. The resulting plasmid, pGEM-T::prtA, was afterwards digested with NotI and SmaI restriction enzymes to obtain an internal 560-bp region of prtA. For the lipB strain, an internal 621-bp region was PCR amplified from the chromosome and subsequently digested with BamHI and XhoI enzymes. These internal DNA fragments were purified and ligated into the respective sites of pKNOCK-Cm (prtA mutant) or pKNOCK-Gm (lipB mutant). The resulting plasmids were introduced into competent E. coli SM10 (λpir) cells by chemical transformation and then mobilized into the S. marcescens wild-type strain by conjugation. Insertional mutants were confirmed by PCR analysis.

To construct the pBBR1-MCS2::prtA plasmid, the prtA gene was amplified from the chromosome by PCR using primers prtA ATG-Fw and prtA-Rv. The 1,500-bp fragment obtained was purified and cloned into KpnI-HindIII-digested plasmid pBBR1-MCS2, and the resulting plasmid was mobilized into S. marcescens prtA by conjugation.

To analyze the transcriptional activity of prtA, the putative promoter region was PCR amplified from the chromosome using primers prom prtA-Fw and prom prtA-Rv and cloned into pGEM-T. Afterwards, pGEM-T::PprtA was digested with EcoRI enzyme, yielding a 443-bp fragment which was subsequently cloned into the same site of pPROBE-NT [ASV] gfp-reporter vector (69). To study lipBCD transcriptional expression, the promoter region of the operon was amplified by PCR using the primers prom lipBCD-Fw2 and prom lipBCD-Rv (Table 2). The purified PCR product was digested with the HindIII and XbaI restriction enzymes and was ligated into the same sites of pPROBE-OT (69). Either pPROBE::PprtA-gfp or pPROBE::PlipBCD-gfp was mobilized by conjugation into the S. marcescens wild-type or cpxR strain.

TABLE 2.

Primers used in this study

Primer	Sequence (5′→3′)
prtA left-Fw	AGGCTCGCTGCCGTTAG
prtA right-Rv	CCTGATCGTGCGTTCGC
prtA ATG-Fw	GGGGTACCGTTATGTCTATCTGTCTG
prtA-Rv	CCCAAGCTTTTACACGATAAAGTCAGTG
lipB-Fw.BamHI	CGCGGATCCAAAGGCGATGCGGTATTGC
lipB-Rv.XhoI	CCGCTCGAGGAACGCGTTGACGTTGCC
cpxR-Fw	ACGGGATCCATATGAACAAGATTCTGTTAG
cpxR-Rv	AGCAAGCTTTCATGTTGCAGATACCATC
prom prtA-Fw	CGGAATTCAGGCTCGCCGCCGATAG
prom prtA-Rv	CGAAGCTTAACCTCCCCGTAAGCCAG
prom lipBCD-Fw2	CCCAAGCTTCCATAGCCGTGCCAGGAA
prom lipBCD-Rv	TGCTCTAGACCGCAATTTCATTGCGCG
16S RT-Fw	AAACTGGAGGAAGGTGGGGATGAC
16S RT-Rv	ATGGTGTGACGGGCGGTGTG
prtA RT-Fw	TTACCCGTGAGAACCAAACC
prtA RT-Rv	TGTAGTTGCCGAAGGTGATG
cpxP RT-Fw	TGGAAGCCATGCATAAACTG
cpxP RT-Rv	TACGCTGCTGATGTTTCTGG
nucA-Fw	GCTCTAGAGGCAAGACGCGCAACTGG
nucA-Rv	CCGCTCGAGGAAATCGGCGCCCTTCGG
nlpE-Fw.SalI	ACGCGTCGACCTATGAAAAAAATTACGGTAGC
nlpE-Rv.HindIII	AGCAAGCTTTTATTTGCTGCTGCAGTTC

For the complementation of the cpxR mutant, the cpxR gene was first PCR amplified from the chromosome using primers cpxR-Fw and cpxR-Rv, and the DNA product was cloned into the pGEM-T vector. Then plasmid pGEM-T::cpxR was digested using BamHI and SacI enzymes, yielding a 764-bp fragment corresponding to the coding region of cpxR, which was purified and cloned into BamHI-SacI-digested pBBR1-MCS5. The plasmid was then mobilized into S. marcescens cpxR by conjugation.

Construction of the pSU36::nlpE plasmid was done by amplifying the coding region of nlpE from the chromosome with primers nlpE-Fw.SalI and nlpE-Rv.HindIII, yielding a 703-bp fragment which was subsequently purified, digested with SalI and HindIII enzymes, and cloned into previously SalI-HindIII-restricted pSU36. The construction was then introduced into the S. marcescens wild-type or cpxR strain by electroporation.

Media and growth conditions.

Strains were routinely cultured in Miller's Luria-Bertani (LB) medium at the desired temperature. For biofilm assays, SLB medium (peptone at 10 g/liter and yeast extract at 5 g/liter) was also used. The antibiotics used for selection in E. coli or S. marcescens were tetracycline, chloramphenicol, kanamycin, and ampicillin at concentrations of 4, 20, 50, and 100 μg/ml, respectively.

Proteomic analysis.

The excised band of interest was submitted to the CEQUIBIEM proteomic facility in Argentina. Mass spectrometric data were obtained using an Ultraflex II (Bruker) matrix-assisted laser desorption ionization–time of flight (MALDI-TOF)/TOF spectrometer. The resulting mass spectra were used for the identification of the protein by the Mascot search engine using the preliminary gene sequence of S. marcescens DB11 on the SEED server (http://www.theseed.org).

Protease assays.

As a qualitative approach, 2 μl of overnight-grown cultures were inoculated on LB agar plates supplemented with skim milk at 2% (wt/vol) and incubated for 16 h at 30 or 37°C. Protease activity was identified as a distinct clearing of the milk around the colony. For quantitative analysis, protease activity (here called the azocaseinase assay) was measured from culture supernatants using azocasein (Sigma) as a colorimetric substrate as previously described (42). Cultures were centrifuged and filtered to remove bacteria. A 50-μl aliquot of the filtered supernatant was mixed with 50 μl of 1% (wt/vol) azocasein and 140 μl of phosphate-buffered saline (PBS) and incubated for 1 h at 37°C. The reaction was stopped by addition of 80 μl of 10% (vol/vol) trichloroacetic acid, and the mixture was incubated for 15 min on ice. The tubes were centrifuged at 10,000 × g for 10 min. The clear supernatant was removed, and its absorbance at 340 nm (A₃₄₀) relative to that of a medium control was determined. This value was then normalized to the optical density at 600 nm (OD₆₀₀) from the original culture.

β-Galactosidase activity assays.

For β-galactosidase activity assays, bacteria were grown for 16 h in LB medium at 30 or 37°C, and the activity was determined as described previously (70).

RNA purification.

Total RNA was extracted from stationary-phase cultures grown for 16 h in LB medium at 30 or 37°C. A total of 250 μl of ice-cold 5% (vol/vol) water-saturated phenol (pH 5.5) in ethanol was added to 1 ml of the cultures to stop the degradation of RNA. Cells were centrifuged at 6,000 × g for 5 min at 4°C and resuspended in 100 μl of 10 mM Tris-HCl and 1 mM EDTA (pH 8.0). The RNA extraction was performed using the Promega SV total RNA isolation kit, following the manufacturer's instructions.

qRT-PCR.

cDNA synthesis was performed using random hexamers, 2 μg of total RNA, and 1 U of SuperScript II RNase H2 reverse transcriptase (Invitrogen). Five microliters of a 1/10 dilution of each cDNA was used as the template in quantitative real-time PCR (qRT-PCR) (reaction mixture, 20 μl), using primers prtA RT-Fw, prtA RT-Rv, cpxP RT-Fw, and cpxP RT-Rv. 16S rRNA was used as the reference gene. A 250-bp fragment was amplified in all cases. The reactions were carried out in the presence of the double-stranded-DNA-specific dye SYBR green (Molecular Probes) and monitored in real time with a Mastercyclerep Realplex real-time PCR system (Eppendorf). The relative expression was calculated using the threshold cycle (C_T) values obtained for each sample, as follows: relative expression = 2^−ΔΔCT, with ΔC_T = C_{Ttranscript of interest} − C_T16S and ΔΔC_T = ΔC_{Texperimental condition} − ΔC_{Treference condition}. The average values were calculated from triplicate samples.

Construction, expression, and purification of CpxR-6×His.

The cpxR gene was PCR amplified from S. marcescens RM66262 genomic DNA using primers cpxR-Fw and cpxR-Rv (Table 2) and cloned into a pQE32 vector as an N-terminal fusion to a 6×His tag. The fusion protein was expressed in E. coli M15/pRep4. Cells were grown in 300 ml of LB medium at 30°C to an OD₆₀₀ of 0.6. Following induction with 50 μM isopropyl-β-d-thiogalactopyranoside (IPTG) and incubation for 16 h at 30°C, cells were harvested and disrupted by sonication. The CpxR-6×His protein was purified using an Ni²⁺-nitrilotriacetic acid-agarose affinity chromatography column according to the QIAexpression purification protocol (Qiagen) and exhaustively dialyzed against 20 mM Tris-HCl (pH 7.4)–500 mM NaCl. The protein concentration was determined by a bicinchoninic acid assay (Sigma), and the protein profile of the purified CpxR-His6X protein was analyzed by SDS-PAGE.

Prediction of CpxR-binding site.

The consensus motif for the CpxR-binding site was generated by training the Multiple Expectation Maximization for motif Elicitation (MEME) tool (46) using as input cognate binding sequences in the promoter regions of the following CpxR-regulated genes from Salmonella: cpxP1, cpxP2, htrA, ppiA, yccA, spy, dsbA, amiA, amiC, cueP, gesA, and scsA (71–75). The obtained matrix was then submitted to the Find Individual Motif Occurrences (FIMO) tool to detect the presence of the motif in the promoter region of the prtA gene from S. marcescens (47).

Protein-DNA interaction analysis.

Electrophoretic gel mobility shift competition assays (EMSAs) and DNase I footprinting assays were performed using 14 fmol of a ³²P-labeled DNA fragment containing the prtA promoter (PprtA) and 20 pmol of purified CpxR-6×His protein following basically previously described protocols (76). Prior to addition of the DNA probe, CpxR-6×His protein was phosphorylated by incubation with 25 mM acetyl phosphate at 30°C for 30 min. The specificity of binding was assayed using the unlabeled PprtA probe or a 436-bp PCR fragment corresponding to the nucA gene from S. marcescens as a nonspecific competitor. The primers used to amplify the PprtA region and nucA are listed in Table 2. To evaluate the effect of CpxR phosphorylation upon binding, EMSAs were performed using 14 fmol of a nonlabeled DNA fragment containing the PprtA promoter and 2.5, 5, 10, 20, or 40 pmol of purified CpxR-6×His with or without previous incubation with 25 mM acetyl phosphate. DNase I protection assays were done for both DNA strands. The DNA sequence ladder was generated in parallel using the appropriate primers and the Sequenase DNA sequencing kit (Promega). After electrophoresis, the gels were either dried and exposed to autoradiography or stained with SYBR green (Invitrogen).

Biofilm assays.

The quantification of biofilm production was performed by following a previously established protocol (77), with slight modifications. Briefly, in a 96-well microtiter plate, single colonies were inoculated into 150 μl of LB or SLB broth in sextuplicate and grown statically at the desired temperature for 48 h. The culture was aspirated, and wells were washed with water. Each well was stained with 0.5% crystal violet for 15 min at room temperature and then washed three times with water. The wells were allowed to dry for 1 h before 200 μl of ethanol-acetone (80:20) was added, and the plate was shaken at room temperature for 1 h to dissolve crystal violet from the well walls. Finally, absorbance at 562 nm was determined using a Synergy 2 plate reader (Biotek).