Quorum Sensing and c-di-GMP-Dependent Alterations in Gene Transcripts and Virulence-Associated Phenotypes in a Clinical Isolate of Aeromonas hydrophila

Abstract

Recently, we demonstrated that the LuxS-based quorum sensing (QS) system (AI-2) negatively regulated the virulence of a diarrheal isolate SSU of Aeromonas hydrophila, while the ahyRI-based (AI-1) N-acyl-homoserine lactone system was a positive regulator of bacterial virulence. Thus, these QS systems had opposing effects on modulating biofilm formation and bacterial motility in vitro models and in vivo virulence in a speticemic mouse model of infection. In this study, we linked these two QS systems with the bacterial second messenger cyclic diguanosine monophosphate (c-di-GMP) in the regulation of virulence in A. hydrophila SSU. To accomplish this, we examined the effect of overproducing a protein with GGDEF domain, which increases c-di-GMP levels in bacteria, on the phenotype and transcriptional profiling of genes involved in biofilm formation and bacterial motility in wild-type (WT) versus its QS null mutants. We provided evidence that c-di-GMP overproduction dramatically enhanced biofilm formation and reduced motility of the WT A. hydrophila SSU, which was equitable with that of the ΔluxS mutant. On the contrary, the ΔahyRI mutant exhibited only a marginal increase in the biofilm formation with no effect on motility when c-di-GMP was overproduced. Overall, our data indicated that c-di-GMP overproduction modulated transcriptional levels of genes involved in biofilm formation and motility phenotype in A. hydrophila SSU in a QS-dependent manner, involving both AI-1 and AI-2 systems.

1. INTRODUCTION

At least two modified autoinducer (AI) systems, which are known to date in other bacteria, exist in a diarrheal isolate SSU of A. hydrophila [1, 2]. One of them is the S-ribosylhomocysteinase (LuxS)-based autoinducer AI-2 system [3]. LuxS plays a major role in AI-2 synthesis, and is also an integral component of the activated methyl cycle [4]. The bacterial phenotypes that are dependent on the QS systems include surface attachment and/or biofilm formation, motility, and in vivo virulence [5–7].

Recently, we characterized the AI-2-mediated QS in A. hydrophila SSU and showed that the ΔluxS mutant formed denser biofilm in vitro, had increased virulence in a septicemic mouse model of A. hydrophila infections, and exhibited decreased motility compared to that of the wild-type (WT) bacterium [1]. However, limited information was available on the AI-1 [N-acyl-homoserine lactone (AHL)-based QS (ahyRI)] system in the regulation of bacterial virulence. Consequently, we demonstrated that several virulence traits were altered in the ΔahyRI mutant [2]. However, there are at least three ahyR homologs that exist in the genome of A. hydrophila SSU, and ahyR2 is genetically linked to the type 6 secretion system (T6SS) (unpublished data). The contribution of these three ahyR homologs in A. hydrophila SSU AI-1 QS is currently unknown.

Like QS, cyclic diguanosine monophosphate (c-di-GMP), the bacterial intracellular second messenger, has also been implicated in the regulation of cell surface properties of several bacterial species [8]. The cellular level of c-di-GMP was shown to be controlled through the opposite activities of diguanylate cyclases (DGCs) and phosphodiesterases (PDEs), which contained functional GGDEF and EAL domains, respectively. Subsequently, it was found that GGDEF and EAL domain proteins were involved in c-di-GMP synthesis and degradation (hydrolysis), respectively [9, 10]. Recently, it was shown that QS modulated the c-di-GMP signaling pathway to control bacterial virulence [11, 12], and the role of LuxS in controlling c-di-GMP production at the transcriptional level was described in Vibrio species [13]. Indeed, c-di-GMP activated biofilm formation in a variety of bacteria, including Pseudomonas aeruginosa, Salmonella enterica serovar Typhimurium, Vibrio spp., and Yersinia pestis [14, 15] and correspondingly repressed motility of these bacteria [14, 16, 17].

In V. cholerae, an accumulation of AI-2 induced the expression of a gene encoding the master QS-regulator HapR [18]. HapR negatively affected bacterial biofilm formation and exopolysaccharide (EPS) biosynthesis [19] by lowering the intracellular level of c-di-GMP [13]. Further, HapR negatively regulated the transcription of vpsR, vpsT, and the vps genes, while VpsR and VpsT positively regulated the transcription of vps genes in V. cholerae [20]. The contribution of the AI-2 extracellular signal to this process is not clearly understood. Although LuxS was initially identified by its contribution to the expression of the hapR gene in a light production bioassay [18], deletion of the luxS gene in a later study did not affect HapR-dependent biofilm formation in V. cholerae [21].

The mechanism of coexistence and co-regulation of the above-mentioned two QS systems (AI-1 and AI-2) in A. hydrophila SSU is not clear. Since QS and c-di-GMP signaling regulate some of the same complex processes, like biofilm formation, motility, and virulence of bacteria, and since we observed that a GGDEF domain protein encoding gene is genetically linked to the luxS gene of A. hydrophila SSU [1], it is possible that these two signaling pathways converge in A. hydrophila.

In this study, we provided data showing an effect of c-di-GMP overproduction on the phenotypic and transcriptional profiling of virulence-associated genes in WT and QS mutants of A. hydrophila SSU. To our knowledge, our study is the first that illustrated an interplay between AI-1, AI-2, and c-di-GMP in A. hydrophila.

2. RESULTS

2.1. Identification of genes in A. hydrophila SSU which are homologs of QS-dependent c-di-GMP regulated genes in Vibrio and Pseudomonas

Although our recent studies showed the roles of AHL-mediated AI-1 QS system [2] and that of the luxS gene [1] in A. hydrophila virulence, we were interested in identifying additional genes which could be involved in the QS system of A. hydrophila SSU. Consequently, we took advantage of our annotation of the genome sequence of A. hydrophila subsp. hydrophila ATCC 7966^T [22]. By using specific primers (Table 1) and sequence analysis of the resulting polymerase chain reaction (PCR) product, we identified the presence of two more ahyR homologs on the genome of A. hydrophila SSU. The gene ahyR2 was genetically linked to the T6SS.

Table 1.

Oligonucleotides used and PCR products

Primer	Oligonucleotide sequence	Use	PCR product (bp)
ggdefF1 ggdefR1	CCGCTCGAGATGATCAAGTATCAAATCCCCA GCTCTAGATCAGGATGGCAGCGGCATC	GGDEF domain encoded gene of A. hydrophila cloned into pBAD/Myc-HisB ara Kmr Apr	966
litRF1 litRR1	GAACCACGCGCAGAAATAAG GGGCCGTCACGTACTTGA	litR	595
vpsT (csgAB)F1 vpsT (csgAB)R1	TCAGAGATACTCCTTGGCCCA CGCTTCATGATCACCCCATA	vpsT (csgAB)	645
vpsR(fleQ)F1 vpsR(fleQ)R1	TGCTTCCTTGCTCATCCCGTA TGTTGATTGTTGACGCCGAT	vpsR(fleQ)	1403
fleN F1 fleN R1	TGGTCTGCGCAAAATGCGT TTATTCACGGGAACCTTCCTG	fleN	856
luxSF2 luxSR2	ACCTCCAAGTGGGATGCGTAT CGGGCCATCGAAAAAATGT	luxS	971
ahyRF1 ahyRR1	TATTGCATCAGCTTGGGGAA TGAAACAAGACCAACTGCTTG	ahyR	781
ahyIF1 ahyIR1	AGGAAAATTAAAAGAACACCC TTATTCTGTGACCAGTTCGC	ahyI	610
ahyR2F2 ahyR2R2	TCCTCATACCCCGGAAATGA TCACTGCTCCTCATCCACAT	ahyR2	790
ahyR2F1 ahyR2R1	ATGTTCAACGATGCGATCTG TTATGACAGCCGCAGGTTGGT	ahyR3	612
aha4220F1 aha4220R1	CAACCTTACCCATCAAGGCTA TCTATACCCACGCCTCATTCA	aha4220 homolog	1101
aha1018F1 aha1018R1	AGAGAGGGGGGAGGGCAT GCATTTCAAGAAGAGTACCCA	aha1018 homolog	720
aha0145F aha0145R	AGAGTTTGATCATGGCTCAG GTATTACCGCGGCTGCTG	aha0145 homolog	531

We also identified in its genome the homologs of luxQ, luxU and luxO genes, which are involved in AI-2-dependent phosphorylation cascade [23], and exhibited 95, 96, and 92% homology with the corresponding genes found in the genome of A. hydrophila 7966^T strain [22]. However, we were unable to amplify the luxP gene which encodes an autoinducer binding protein in V. harveyi [24, 25]. The luxP gene is also not present in the genome of A. hydrophila 7966^T strain [22].

We noted the presence of a litR gene in A. hydrophila SSU which is a homolog of hapR gene in V. cholerae [19]. The LitR of A. hydrophila SSU was highly homologous (~98%) to the corresponding gene found in the annotated genome sequences of A. hydrophila 7966 and A. salmonicida subsp. salmonicida A449 strains [22, 26], and exhibited 35–40% identity with HapR protein sequences of vibrios. HapR is a DNA-binding transcription factor that initiates a program of gene expression, switching the cells from individual, low-cell-density state, to the high-cell density state [13]. HapR in V. cholerae regulates c-di-GMP metabolizing genes that lead to a net reduction in the level of cellular c-di-GMP, and thereby, to the negative control of biofim formation [13]. By directly down-regulating the expression of the vpsT gene [13], HapR is also shown to control biofilm formation in a c-di-GMP-independent manner. VpsT is a transcriptional activator which together with VpsR is required for the expression of the V. cholerae polysaccharide biosynthesis operon [27, 28]. We found the vpsT homolog, named csgAB (Fig. 7), in the A. hydrophila SSU genome which encoded a protein with 37–53% identity to VpsT of Vibrio species.

Fig. 7.

Integrated schematic based on the published information illustrating QS-dependent regulatory network in Vibrio, Pseudomonas, and A. hydrophila. LuxR/HapR/LitR represents a major regulator of QS target genes [23, 50–53] involved in the AI-2 phosphorylation cascade [19, 45–47, 72]. HapR negatively affects biofilm formation and exopolysaccharide biosynthesis [19] by lowering the intracellular level of c-di-GMP [13] and repressing the positive regulator VpsT [73]. VpsT from V. cholerae, a master regulator for biofilm formation [15], directly senses c-di-GMP to inversely control extracellular matrix production and motility. VpsR and FleQ are the transcriptional regulators containing an AAA-type ATP-ase domain and a DNA- binding domain [29]. The orthologous systems CsrA LetA/LetS exist in many gram-negative bacteria, such as S. enterica (BarA/SirA), Erwinia carotovora (ExpA/ExpS), V. cholerae (VarA/VarS), Pseudomonas spp. (GacA/GacS), Photorhabdus luminescens (UvrY/BarA) and E. coli (UvrY/BarA) [74]. The central regulator CsrA is antagonized by sRNA and CsrC, whose expression is positively regulated by the BarA-UvrY two-component system.[48, 75]. An unconventional GG-DEF-EAL protein CsrD regulates sRNAs at the level of RNA stability [76–78]. The global regulator CsrA contributes to the control of c-di-GMP levels by modulating selected panels of GGDEF and EAL genes in response to internal signals [79]. Also CsrA, a post-transcriptional regulator, can activate LuxO, possibly through an additional regulatory factor (denoted X) [80].

Interesting was also the presence of FleQ in A. hydrophila SSU (Fig. 7), which showed 51% identity with its homolog in P. aeruginosa [29], and 35% identity with VpsR in V. cholerae [28]. Likewise, the FleN in A. hydrophila SSU demonstrated 69% identity with FleN of P. aeruginosa [29]. FleQ, the first example of a c-di-GMP effector involved in transcription control, is a master regulator of flagellar gene expression in P. aeruginosa [29]. FleN, an anti-activator of FleQ, down-regulates FleQ activity through direct interactions and thereby plays a crucial role in maintaining a single flagellum in P. aeruginosa [30, 31].

Our genome sequencing of A. hydrophila 7966 strain suggested the presence of 23 proteins containing GGDEF domain, 7 proteins with an EAL domain, and 15 proteins that had both GGDEF and EAL domains [22]. One of the GGDEF domain protein (AHA0701h) encoding genes was located immediately downstream of the luxS gene in A. hydrophila SSU. The AHA0701h protein exhibited 93% identity with its homolog found in A. hydrophila 7966 strain [22]. A homolog of A. hydrophila SSU AHA0701h, referred to as VCA0956 (46/68), was also found in V. cholerae (Fig. 1).

Fig. 1.

A homology of A. hydrophila AHA0701 protein with its VCA0956 homolog found in V. cholerae.

Therefore, A. hydrophila SSU QS systems have characteristics of both Vibrio and Pseudomonas, in addition to having unique signaling cascades which makes it an ideal candidate to study QS-dependent c-di-GMP regulatory network. Such studies on A. hydrophila are important as the organism leads to both gastroenteritis and septicemia depending on the route of infection [2].

2.2. Overproduction of GGDEF domain protein regulates virulence-associated phenotypes of WT A. hydrophila SSU

The GGDEF domain and an allosteric binding site for c-di-GMP (I site) [32] are typically found as part of modular DGC enzymes, which catalyze the synthesis of the signaling molecule c-di-GMP [33]. The presence of AHA0701h with GGDEF domain, the location of its gene immediately downstream of the luxS gene in A. hydrophila SSU, together with the homology of AHA0701h with VCA0956 (Fig. 1), prompted us to test the effect of its over-production on WT A. hydrophila virulence-associated phenotype(s). Indeed, overproduction of VCA0956, a cytoplasmic protein composed of a single DGC domain, was successfully used to increase c-di-GMP levels in V. cholerae [34, 35] and V. vulnificus [36]. Studies have shown that the modulation of c-di-GMP level by overproducing proteins with either the DGC or the PDE domain of V. cholerae and A. veronii biovar sobria resulted in an altered biofilm formation and motility [35, 37].

For these studies, we cloned the gene encoding AHA0701h into the pBAD/Myc-HisB vector. Subsequently, the empty vector and pBAD/Myc-His::aha0701h (Table 2) were transformed into WT A. hydrophila SSU and its ΔahyRI mutant, and the phenotype of these strains was studied after arabinose induction.

Table 2.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Source or reference
A. hydrophila SSU		CDCa, Atlanta
SSU-Rifr	Rifampin resistance (Rifr)	Laboratory stock
SSUΔluxS	luxS mutant of A. hydrophila SSU-R; Rifr Kmr	This study
SSUΔluxS/pBR322::luxS	luxS mutant complemented Rifr Kmr Apr	This study
ΔahyRI	ΔahyRI deletion mutant of hydrophila SSU-R; Rifr Smr Spr	This study
ΔahyRI/pBR322:: ΔahyRI	ΔahyRI mutant complemented Rifr Smr Spr Apr
E. coli	TOP10	Invitrogen
Plasmids
pBAD/Myc-HisB	vector, ara BAD promoter Apr	Laboratory stock
pBAD:: ggdef	GGDEF domain encoded gene of A. hydrophila cloned into pBAD/Myc-HisB ara Kmr Apr	Laboratory stock

Overproduction of AHA0701h resulted in a dramatically increase in biofilm formation, as measured by quantitative crystal violet (CV) staining, in WT A. hydrophila SSU compared to that of bacteria harboring only the vector (Fig. 2A & B). The biofilm formed by WT A. hydrophila with vector alone, visualized by light microscopy, appeared as uniform sparse cell attachment on the surface of Thermanox plastic cover slips after 12 h of incubation, with a greater number of bacterial cells adherent at the line of the edge of the cover slips after 24 h of incubation (Fig. 3A). In contrast, biofilm formed by the WT A. hydrophila with pBAD/Myc-His B::aha0701h plasmid appeared as a mature, dense three-dimensional structure after 24 h of incubation, uniformly distributed all along the cover slip’s surface (Fig. 3B). Thus, overexpression of the gene encoding AHA0701h in WT A. hydrophila dramatically increased the number of attached cells and accelerated the dynamics of three-dimensional structured biofilm formation.

Fig. 2.

Measurement of biofilm mass formed on polystyrene plastic after crystal violet staining: A – WT A. hydrophila SSU with pBAD/Myc-HisB vector alone; B – WT A. hydrophila SSU with pBAD/Myc-His::aha0701h; C - ΔluxS mutant.

Fig. 3.

Light microscopy images of biofilm formation on Thermanox cover slips by WT A. hydrophila with uniform sparse cell attachment on the surface of plastic cover slip (A), and WT A. hydrophila with overproduction of GGDEF domain protein (AHA0701h), forming much thicker and denser biofilm (B), after 24 h of cultivation.

The number of biofilm forming cell-units of WT A. hydrophila SSU with overproduced GGDEF protein (AHA0701h) was greater compared to that of the ΔluxS mutant (Fig. 2B & C). Further, the ΔluxS mutant formed a matured biofilm after 48 h of incubation [1], while a mature biofilm of WT A. hydrophila SSU with overproduced GGDEF protein (AHA0701h) appeared after 24 h (Fig. 3B).

While over-expression of the gene encoding AHA0701h in WT A. hydrophila SSU positively regulated biofilm formation, motility of the bacteria was decreased compared to that of the WT bacteria (Fig. 4A-3 versus A-1). Additionally, the WT A. hydrophila, which overproduced GGDEF protein (AHA0701h), exhibited minimal motility similar to that of the A. hydrophila ΔluxS mutant (Fig. 4A-3 versus B-3).

Fig. 4.

Effect of the overproduction of AHA0701h on the motility of A. hydrophila inoculated onto the surface of 0.35% LB agar plates containing 0.2 % arabinose: A 1 – WT A. hydrophila SSU; A 2 – WT A. hydrophila SSU with pBAD/Myc-HisB vector alone; A 3 – WT A. hydrophila SSU with pBAD/Myc-His::aha0701h; B 1 - ΔahyRI mutant with pBAD/Myc-HisB vector alone; B 2 - ΔahyRI mutant with pBAD/Myc-His::aha0701h; B 3 - ΔluxS mutant with pBAD/Myc-HisB vector alone; C - the bar graph shows distances in millimeters by which these strains migrated from the point of inoculation in soft agar plates. Data from three independent experiments were plotted with standard deviations. Statistical significance by Student’s t test (p ≤ 0.05) was found for A3 and B3 compared to other genetic background strains.

Thus, over-expression of the gene encoding AHA0701h in WT A. hydrophila led to enhanced biofilm formation and decreased motility, and these data correlated with the observations of Rahman et al. (2007) who showed that overproduction of the GGDEF domain protein AdrA increased c-di-GMP concentration and biofilm formation and reduced motility in A. veronii biovar sobria [37].

2.3. The effect of over-production of GGDEF domain protein on biofilm formation and motility in the ΔahyRI mutant of A. hydrophila SSU

We transformed empty vector and pBAD/Myc-HisB::aha0701h plasmid in the ΔahyRI mutant of A. hydrophila. Recently, we demonstrated that the ahyRI deletion resulted in decreased biofilm formation but the motility of the mutant was not affected compared to that of the parental bacteria [2]. In contrast to WT A. hydrophila with overproduced GGDEF protein (AHA0701h), which resulted in increased biofilm formation (Figs. 2B &3B), biofilm formed by the ΔahyRI mutant with over-production of GGDEF protein was marginally increased after 48 h of incubation (data not shown).

Importantly, both WT A. hydrophila and its ΔahyRI mutant had similar swarming and swimming motility [2] (also Fig. 4 A-2 versus B-1). However, overproduction of the GGDEF protein (AHA0701h) in the ΔahyRI mutant did not alter its motility compared to that of the parental mutant strain (Fig. 4 B-1 versus B-2). This was in contrast to decrease in motility that was observed with the WT A. hydrophila overproducing the GGDEF protein (AHA0701h) (Fig. 4 B-2 versus A-3). Thus, overproduction of GGDEF protein in WT versus its ΔahyRI mutant exhibited contrasting phenotypic alterations.

These data prompted us to study alteration in the transcription of major genes involved in biofilm formation and motility of the WT A. hydrophila and its ΔahyRI mutant after overproduction of GGDEF protein (AHA0701h), as compared to the parental background strains and the luxS mutant. Such studies were performed to evaluate a possible correlation between the presence and expression of QS system-encoding genes and c-di-GMP overproduction in A. hydrophila.

2.4. Alteration in the transcription of genes involved in biofilm formation and motility in luxS and ahyRI mutants of A. hydrophila SSU

The analysis of transcripts by reverse-transcription PCR demonstrated increases in the levels for litR and fleQ (vpsR), and decreases in the levels of csgAB (vpsT) and fleN in the ΔluxS mutant compared to that of WT A. hydrophila (Fig. 5A; Supplemental data, Table I). On the contrary, the ΔahyRI mutation in WT A. hydrophila led to down-regulation of the litR (Figs. 5B and 6A) and an up-regulation of the fleN gene transcript (Fig. 5B), but the levels of expression for the csgAB (vpsT) (Fig. 6B) and fleQ (vpsR) genes were not altered compared to that of the WT A. hydrophila (Supplemental data, Table I). The ΔahyRI mutation also caused an increase in the luxS gene transcript as compared to that of the WT A. hydrophila (Fig. 5B). The expression level of the aha0701h gene in the ΔahyRI mutant was comparable to that in the WT bacteria (Supplemental data, Fig. I, lanes 1 & 3) in contrast to an over-expression of this gene which we noticed in the ΔluxS mutant (Supplemental data, Fig. I, lanes 1 & 2).

Fig. 5.

Schematic representation of fold-changes in the expression of selected genes in various genetic backgrounds of A. hydrophila SSU: the luxS mutant (A); the ahyRI mutant (B); WT with AHA0701h over-production (C); the ahyRI mutant with AHA0701h over-production (D). The data used to generate fold changes (arithmatic mean ± SD) with statistical analysis is shown in Table I; Supplemental data.

Fig. 6.

Reverse-transcription PCR analysis of the litR (A) and vpsT (B) gene expression in different genetic backgrounds of A. hydrophila SSU.

It is known that the elevation of c-di-GMP in the low-cell-density state of V. cholerae induces the transcriptional activator vpsT involved in HapR/LitR-dependent regulation of biofilm formation [27]. Further, regulation by VpsR has also been linked to signal transduction by bacterial c-di-GMP [38, 39]. These studies prompted us to overexpress GGDEF protein (AHA0701h) encoding gene for studying the effect of high level of intracellular c-di-GMP on the transcriptional expression of selected genes mentioned above in WT A. hydrophila and its ΔahyRI mutant.

2.5. The effect of overproduction of GGDEF domain protein on genes involved in biofilm formation and regulation of motility in WT and ΔahyRI mutant of A. hydrophila SSU

We found that luxS, litR, csgAB (vpsT), fleQ, and fleN gene transcripts were significantly increased by AHA0701h overproduction in WT bacteria (Fig. 5C; Supplemental data, Table I). It should be noted that the expression levels of these genes in WT and the WT with vector alone were similar. When GGDEF protein was overproduced in the ΔahyRI mutant, the litR gene transcript was increased compared to its level noted in the ΔahyRI mutant with vector alone (Fig. 6A). At the same time, this level of litR gene transcript in the ΔahyRI mutant with overproduced GGDEF protein was comparable to that in the WT bacteria with vector alone (Fig. 6A; Supplemental data, Table I). Most likely, AHA0701h overproduction normalized the transcriptional level of litR gene in the ΔahyRI mutant, making it similar to that in the WT bacteria.

The same trend of transcript level normalization was also noted for the luxS gene in the ΔahyRI mutant with AHA0701h over-production (Supplemental data, Table I) but occurred as a result of down-regulation of this gene (Supplemental data, Table I). While the transcripts for csgAB (vpsT) and fleQ (vpsR) were not altered in the ΔahyRI mutant compared to that of WT bacteria (Supplemental data, Table I), the AHA0701h overproduction led to decreased levels of csgAB (vpsT) (Figs. 5D and 6B) and the increased levels of fleQ (vpsR) (Fig. 5D) in the ΔahyRI mutant.

Thus, the ahyRI deletion affected on the differentiation in the transcriptional responses of selected genes under AHA0701h over-production conditions compared to that in the WT A. hydrophila.

2.6. Co-modulation of QS gene transcripts and the effects of GGDEF domain protein overproduction on the expression of gene transcript levels in A. hydrophila SSU

Our data showed that the luxS gene transcript modulation was dependent on the presence and expression of AI-1 genes and on the c-di-GMP levels (Figs. 5B, C & D). This prompted us to study the effects of luxS mutation and c-di-GMP level modulation on the expression of ahyR, ahyI, and two additional ahyR (ahyR2 and ahyR3) homologs which exist in A. hydrophila SSU. The ahyR and ahyI gene transcripts were down-regulated in the ΔluxS mutant, while the ahyR2 and ahyR3 gene transcripts were over-expressed compared to that in the WT bacteria (Fig. 5A, Supplemental data, Table I). The GGDEF protein when overproduced in WT A. hydrophila resulted in the up-regulation of luxS, ahyI and all of the three homologs of ahyR genes as compared to that in the WT bacteria with vector alone (Fig. 5C, Supplemental data, Table I).

In the ahyRI mutant, the expression of ahyR2 and ahyR3, together with the luxS gene was increased as compared to that in the WT A. hydrophila (Fig. 5B, Supplemental data, Table I). However, overproduction of GGDEF protein in the ahyRI mutant resulted in down-regulation of luxS, ahyR2, and ahyR3 gene transcripts compared to that in the ΔahyRI mutant with vector alone. The level of expression of these genes returned to the level of WT bacteria (Supplemental data, Table I). Thus, our results demonstrated that AI-1 and AI-2 QS systems were linked at least at the transcriptional level in A. hydrophila SSU. The GGDEF domain protein overproduction modulated the transcriptional levels of genes involved in AI-1 and AI-2, and this modulation was dependent on the presence of both of them in the cells.

3. DISCUSSION

In this study, we identified additional components of LuxS- and AhyRI- dependent QS and c-di-GMP networks in A. hydrophila SSU. Majority of reports on the ΔluxS mutant indicated that the lack of luxS gene in bacteria led to attenuation in biofilm formation and virulence, or retention of virulence but to a decrease in biofilm formation [40–43]. However, Ali and Benitez [44] recently showed that in V. cholerae ΔluxS mutant, there occurred an increase in biofilm formation similar to our findings in A. hydrophila SSU [1]. This finding subsequently was correlated with the model of QS regulation and biofilm formation in V. cholerae through the modulation of intracellular c-di-GMP level [13].

In Vibrio species, sensing of extracellular AI-2 involves a regulation cascade consisting of a binding protein LuxP and the response regulator LuxQ (Fig. 7) [24, 19, 47]. The VarS/A-CsrA/BCD system is an additional signaling mechanism that acts in parallel with the two QS systems to input information into LuxO [48] (Fig. 7). The global regulators CsrA and HapR both largely contribute to the control of c-di-GMP levels by modulating selected panels of GGDEF and EAL domain-encoding genes in response to internal signals [13, 48] (Fig. 7). In addition to QS-dependent HapR regulating cascade, we also identified the VarS/VarA (LetS/LetA)-CsrA regulatory system in the A. hydrophila SSU genome (unpublished data) by examining gene homologs that exist in a variety of gram-negative bacteria [48]. However, subtle differences were noted, for example, between the HapR-dependent QS systems of V. cholerae and A. hydrophila SSU. Further, how these two regulatory systems co-function in A. hydrophila is not known yet.

To date, two types of AI-2 receptors have been identified: LuxP, present in Vibrio species and LsrB, first identified in S. Typhimurium [24, 49] (Fig. 7). However, we could not identify any homolog of such receptors in A. hydrophila SSU. Moreover, no information about sRNAs involved in A. hydrophila SSU QS regulation is available from the genome sequence of A. hydrophila ATCC 7966 [22].

We did identify homologs of luxQ, luxU, luxO, and litR/hapR genes in A. hydrophila SSU genome similar to that found in Vibrio [58]. LuxR/HapR is the only TetR family member that appears to act as both activator and repressor [50–53]. While the mechanism by which LuxR (HapR) activates qrr (encoding sRNA) gene expression in V. harveyi is yet to be determined, it is known that the expression of the qrr gene requires LuxO~P acting in conjunction with the alternative sigma factor sigma-54 [23, 54]. Recently, Waters et al. [13] demonstrated that HapR altered the expression of 14 genes encoding proteins with GGDEF and/or EAL domains. The QS control of these genes resulted in increased intracellular levels of c-di-GMP at low cell density and decreased c-di-GMP levels at high cell density. The major consequence of increased c-di-GMP in the low-cell-density state is the induction of biofilm transcriptional activator vpsT (Fig. 7) [27]. We identified the homolog of vpsT in WT A. hydrophila SSU genome which is designated as csgAB. VpsT is a member of the FixJ, LuxR, and CsgD family of prokaryotic response regulators [55–57].

We did not find any candidate protein which was closely related to VpsR of V. cholerae, a key regulator of biofilm formation [28, 58], in the A. hydrophila genome, but FleQ (found in P. aeruginosa) [29], a distant VpsR homolog, encoding gene was detected (Fig. 7). In V. parahaemolyticus, the role of VpsR is reported to be strain dependent [59] while in V. fischeri, VpsR might function as a repressor of an unknown polysaccharide that could contribute to mucoidal versus non-mucoidal bacterial colonies, and together with a positive regulator of cellulose biosynthesis, might involve in the complexity of the biofilm formation [60].

VpsR and FleQ are transcriptional regulators containing an AAA-type ATPase domain and a DNA-binding domain, and belong to Fis family of prokaryotic response regulators. FleQ was identified as a c-di-GMP responsive transcriptional regulator that controls the transcription of exopolysaccharide biosynthesis genes in P. aeruginosa (Fig. 7) [29]. However, FleQ also activates expression of the two component regulatory genes fleSR (Fig. 7) as well as genes that are necessary for the assembly of the flagellar export apparatus [29, 61–64]. We demonstrated the presence of fleQ, fleS and fleR gene homologs in A. hydrophila SSU genome (unpublished data) in contrast to A. hydrophila AH-3 strain, where no master regulatory genes encoding homologs of P. aeruginosa FleQ, FleS, and FleR were found [65]. We also identified FleN which is an anti-activator of FleQ in P. aeruginosa [29–31, 38]. We believe that FleQ of A. hydrophila SSU is a significant player in biofilm formation and bacterial motility.

While searching for genes encoding c-di-GMP metabolic proteins in A. hydrophila SSU genome in addition to aha0701h, we identified the presence of aha4220 and aha1018 gene homologs. These genes encode DGC and PDE, respectively, but we could not identify, for instance, a homolog of A. hydrophila ATCC 7966 aha0996 gene that encodes DGC (unpublished data). Thus, QS-dependent c-di-GMP regulatory network in A. hydrophila SSU includes the elements of cascade related to that in Vibrio and Pseudomonas, thus making this system unique.

To better understand the role of LuxS and AhyRI in QS-dependent c-di-GMP regulatory network in A. hydrophila SSU, we monitored transcript levels of the selected genes involved in biofilm and motility regulation and the effects on them by overproduction of GGDEF protein.

Interestingly, the fact that the litR gene was up-regulated in the ΔluxS mutant of A. hydrophila SSU was surprising in itself because it is totally distinct compared to the QS-dependent hapR regulation which exists in Vibrio [23] (Fig. 7). Together with the absence of major components of the QS-dependent regulating cascade, such as the LuxP and the genes encoded cascade involving sRNA, suggest a different mechanism of LuxS-dependent QS gene regulation in V. cholerae [48, 58] and A. hydrophila.

Biofilm formation in V. cholerae is primarily controlled by two transcriptional activators VpsT and VpsR, which are required for the expression of Vibrio polysaccharide biosynthesis operon vps [58]. HapR represses vpsT expression both by reducing the cellular concentration of c-di-GMP via altering transcription of genes encoding GGDEF and EAL proteins and by directly binding to the vpsT promoter [13]. In addition, recently, Krasteva et al. [15] showed that the transcriptional regulator VpsT from V. cholerae directly senses c-di-GMP to inversely control biofilm and motility.

Our study showed that the ΔluxS mutant of A. hydrophila SSU formed denser biofilm and exhibited decreased motility, however, the csgAB (vpsT) transcript was down-regulated, while the fleQ (vpsR) was increased. Importantly, VpsR in A. hydrophila SSU was more closely related to FleQ in P. aeruginosa [29] thus making LitR-VpsT-FleQ system of A. hydrophila SSU unique.

Importantly, in contrast to the luxS mutant, the transcript levels for the vpsT and vpsR genes in the ahyRI mutant of A. hydrophila SSU were similar to that in the WT bacteria in spite of litR gene down-regulation, and no alteration in biofilm formation and motility was noted. However, we evidenced that ahyRI deletion sufficiently altered transcriptional response and phenotype changes when GGDEF protein was overproduced in the ahyRI mutant compared to that in the WT A. hydrophila SSU.

Notably, overproduction of GGDEF protein in WT A. hydrophila SSU resulted in increase of all selected genes compared to that in the WT bacteria (Fig. 5C). An increase in the luxS and litR transcripts could provide a decrease in c-di-GMP level that correlated with the common scheme of LuxS-dependent HapR c-di-GMP regulation in Vibrio [66] (Fig. 7). However, because GGDEF protein is overproduced, the c-di-GMP level continued to be very high that consequently could stimulate csgAB (vpsT) and fleQ (vpsR) expression levels as noted in Fig. 5C. The phenotypic alterations noted with the overproduction of GGDEF protein in the WT A. hydrophila SSU were not surprising, and corroborated with the effects of high level of c-di-GMP on biofilm formation and motility in A. veronii [37] and V. vulnificus [36]. Interestingly, these alterations were similar to that found in the ΔluxS mutant, however, the transcriptional level of expression of the selected genes involved in the biofilm formation and motility regulation were distinct.

In contrast to the up-regulation of all selected genes which were revealed in the WT A. hydrophila SSU when GGDEF protein was overproduced, in the ΔahyRI mutant, the overproduction of AHA0701h mostly resulted in adjustment of the selected gene expression levels to the level of these genes in the WT bacteria. The exceptions to that were vpsT (csgAB) and fleN, which were down-regulated, and vpsR (fleQ), which was up-regulated when compared to the same gene transcripts in the WT A. hydrophila SSU. These gene transcript level alterations were not essentially affecting motility and biofilm formation in the ΔahyRI mutant. More likely, the deletion of the ahyRI genes abrogated the effect of GGDEF protein overproduction on gene expression in the ΔahyRI mutant compared to that in the WT A. hydrophila SSU when GGDEF protein was overproduced.

In addition, we observed for the first time to our knowledge, that mutation in either the luxS gene or the ahyRI genes caused an alteration in gene transcripts of the remaining functional QS system of A. hydrophila. Interestingly, together with the luxS gene up-regulation, the expression levels of two additional copies of the ahyR gene were increased in the ahyRI mutant. Moreover, the overproduction of GGDEF protein brought down these transcripts to the level comparable to that in WT A. hydrophila SSU.

In conclusion, the transcript level of genes involved in AI-1 and AI-2 QS systems were co-regulated in A. hydrophila SSU. A balance of transcriptional level of gene expression between luxS and ahyR was regulated by modulation of c-di-GMP level. The luxS and luxR- type (ahyR) regulator of AI-1 mutually compensated each other at least at the transcriptional level. The effects of c-di-GMP modulation on a number of common genes involved in motility and biofilm formation regulation were dependent on the expression AI-1 and AI-2 systems.

4. MATERIALS AND METHODS

4.1. Bacterial strains and plasmids

The sources of A. hydrophila and E. coli strains, as well as the plasmids used in this study, are listed in Table 2. The antibiotics ampicillin (Ap), kanamycin (Km), streptomycin (Sp), and spectinomycin (Sm) were used at concentrations of 50–500, 50–100, 50–100 and 50–100 µg/ml, respectively, in Luria-Bertani (LB) medium or agar plates. Rifampicin (Rif) was utilized at a concentration of 50 µg/ml for bacterial growth and 300 µg/ml during transformation experiments. All of the antibiotics used were obtained from Sigma (St. Louis, MO). Restriction endonucleases and T4 DNA ligase were obtained from Promega (Madison, WI). The Advantages cDNA PCR Kit was purchased from Clontech (Palo Alto, CA). The digested plasmid DNA or DNA fragments from agarose gels were purified using a QIApreps Miniprep Kit (Qiagen, Inc., Valencia, CA). The medium was supplemented with L-arabinose (0.2%) when the ggdef gene was expressed from the pBAD/Myc-HisB::ggdef plasmid (Table 1) under the control of an arabinose-inducible pBAD promoter.

4.2. Primers and PCR assays

The primers used for various experiments are indicated in Table 1 and were synthesized by Sigma-Aldrich Biotechnology LP. PCR assays were performed with 100 ng of gDNA and 50 ng of plasmid DNA.

4.3. Cloning of the aha0701h gene into pBAD/Myc-HisB

The aha0701h gene was amplified with primers ggdeF1 and ggdefR1 from A. hydrophila SSU genomic DNA. The PCR product was digested with the indicated enzymes and cloned into the XhoI and XbaI sites of the conjugative expression vector pBAD/Myc-HisB. Correct clones were verified by restriction digestion and sequencing. Expression of the aha0701h gene was induced with arabinose.

4.4. Reverse transcription–polymerase chain reaction (RT-PCR)

Standard conditions for the isolation of total RNA from cells grown overnight in LB medium were used (Ambion, Austin, TX). RNA concentration was determined by optical density at 260 nm. Triple measurements were performed for each RNA sample and an average value was obtained as a final concentration. RT-PCR was performed using SuperScript™ III Platinum (Invitrogen, Carlsbad, CA). Equal amounts of DNase-treated total RNA (500 ng) were used to generate cDNA with random hexamer primers according to the manufacturer’s protocol.

The resultant cDNA was diluted eight times with Nuclease-free water and concentration was determined by optical density at 260 nm. The diluted cDNA was subsequently used in 50 µl of RT-PCR reaction. Sequences of primers used in RT-PCR are shown in Table 1. The 16S rRNA was used as the internal control gene to confirm that equal amounts of total RNA was used in each reaction [67, 68]. As a negative control and a control to detect DNA contamination, water and DNase treated RNA were used in place of cDNA template, respectively. For RT-PCR analysis, 30–35 cycles of PCR reaction was used. Standard techniques were employed to prevent and detect in vitro contamination of the PCR reactions. Amplified products were identified by size, by using agarose-gel electrophoresis and ethidium-bromide staining. Specificity of the amplified products was demonstrated by DNA sequence analysis. The relative level of the cDNAs of RT-PCR was determined by densitometric analyses using AlphaEasyFC software (AlphaInnotech, San Leandro, CA) using 16S rRNA genes as references.

4.5. in vitro growth of the biofilm

Briefly, 6-well, sterile polystyrene microtiter plates were filled with 2.5 ml of LB broth supplemented with the appropriate antibiotics. Next, a sterile, 13-mm-diameter thermanox plastic cover slip was laid in each of the wells. Medium was inoculated with 10⁶ colony forming units (cfu) of WT A. hydrophila or its mutants, and the plates were incubated at 37°C. After 12, 24 or 48 h of incubation with gentle shaking (65 rpm), those bacterial cells that were not sufficiently adherent, were removed along with the planktonic cells, by 3 washes of water. Glass or Thermanox plastic cover slips containing the attached biofilms were removed, rinsed with water, and viewed by light microscopy.

4.6. Motility assay

LB medium with 0.35% agar was used to characterize the motility phenotype of WT A. hydrophila SSU and its mutants strain. The overnight cultures grown in the presence of respective antibiotic were adjusted to the same optical density, and equal numbers of CFU (10⁶) were stabbed into 0.35% LB agar plates. Plates were incubated at 37°C overnight, and the motility was assayed by examining migration of bacteria through the agar from the center towards the periphery of the plate [69].

4.7. Crystal violet (CV) biofilm assay

The WT A. hydrophila SSU and its mutants were grown in 3 ml LB broth contained in polystyrene tubes at 37°C for 24 h with shaking. Biofilm formation was quantified according to the procedure described elsewhere [70, 71]. The biofilm formation results were normalized to 1× 10⁹ cfu to account for any differences in the growth rate of various bacterial strains used. The experiment was repeated independently three times.

4.8. Statistical analysis

All of the experiments were performed in triplicate, and wherever appropriate, the data were analyzed by using the Student’s t test, with a p value of ≤ 0.05 considered significant. The data were presented as an arithmetic mean ± standard deviation.

Supplementary Material

01. Supplemental data.

Figure I. Reverse-transcription PCR analysis of aha0701h gene encoding GGDEF domain protein (A) and 16S RNA as a reference (B) in: 1. WT A. hydrophila SSU; 2. ΔluxS mutant of A. hydrophila SSU; 3. ΔahyRI mutant of A. hydrophila SSU.