Background

Plesiomonas shigelloides, a gram-negative, rod-shaped bacterium that causes foodborne intestinal infections [1], can cause gastroenteritis, including acute secretory gastroenteritis, an invasive shigellosis-like disease, and a cholera-like illness [2,3,4]. Escobar et al. found that co-infections of P. shigelloides with either rotavirus or pathogenic Escherichia coli were 16.2-fold (95% confidence interval (CI) 5.5–62.3) and 13.8-fold (95% CI 3.3–69.3) more likely to result in diarrhea, respectively [5]. Extra intestinal infections, such as meningitis, bacteremia, and pseudoappendicitis, including skin and soft tissue infections, are also associated with P. shigelloides infection [6,7,8]. Fresh and estuarine water are considered the natural environments of P. shigelloides, which is often isolated from fish and other seafood [9].

P. shigelloides can grow under both aerobic and anaerobic conditions [10, 11]. The enzymes required for catabolism under aerobic and anaerobic conditions are substantially different; therefore, at the same time, to respond to the availability of oxygen, it is necessary to regulate the expression of genes related to cell functions, such as nutrient absorption and excretion systems, biosynthetic pathways, and macromolecule synthesis [12]. The Arc two-component signal transduction system, comprising the kinase sensor ArcB and its cognate response regulator ArcA, is one of the mechanisms that enable E. coli to adapt to changing oxygen availability [13, 14]. ArcB is activated in the form of a simplified electron carrier under conditions of hypoxia and energy provided by ATP. It has three cytoplasmic domains, and the autophosphorylation of His292 in the H1 domain, followed by transfer of the phosphate group to Asp576 in the D1 domain, then to His717 in the H2 domain [15], and finally to Asp54 in ArcA results in phosphorylation of ArcA [16], which activates ArcA to promote or repress the expression of Arc-regulated genes.

A previous study indicated that about 1139 genes in the E. coli K-12 genome are regulated either directly or indirectly by ArcA [17]. Under anaerobic conditions, ArcA inhibits the expression of genes required for aerobic metabolism, energy generation, amino acid transport, and fatty acid transport [18]. Another transcription factor involved in controlling anaerobic gene expression and facilitating bacterial adaptation to anaerobic conditions is FNR (fumarate and nitrate respiration) [19]. A comparison of the ArcA and FNR regulons showed that 303 genes were regulated by both proteins [17]. Jiang et al. found that citrate utilization in an anaerobic environment in E. coli is under direct control of FNR via the CitA-CitB system and under indirect control by ArcA [20]. A recent study showed that ArcA overexpression in aerobic conditions results in downregulation of respiratory pathways and enhanced growth rates on glycolytic substrates of E. coli, coinciding with acetate excretion and increased carbon uptake rates [21].

ArcA also controls chemotaxis and motility, contributing to the pathogenicity of E. coli [22]. Kato et al. determined that the ΔarcA mutant displayed a motility-defective phenotype and ArcA is necessary for the expression of fliA [23]. Furthermore, in Salmonella enterica sv. Typhimurium, the ΔarcA mutant was also non-motile and lacked flagella [24]. Biofilms are sessile bacterial communities that predominate in nature, and may form wherever a solid surface is in contact with a liquid [25]. Many opportunistic pathogens are capable of biofilm formation. E. coli dominates biofilms found on urethral catheters, and has also been isolated from percutaneous trans-hepatic catheters [26, 27]. Previous studies on certain enterobacteria and non-enterobacteria have also reported the relationship between ArcA and biofilms. For example, Hengge proposed that ArcA has a regulatory role between the sigma factor RpoS and biofilm formation [28]. Xi et al. found that the response regulator ArcA enhances biofilm formation in a vpsT-dependent manner under anaerobic conditions in Vibrio cholerae [29]. In addition, studies on Actinobacillus pleuropneumoniae and Haemophilus parasuis also suggested that ArcA regulates the formation of biofilms positively [30, 31]. However, in Porphyromonas gingivali, Wu et al. showed that ArcA inhibits FimA production and inhibits biofilm formation [32].

In addition to ArcA being related to cell metabolism, biosynthesis, and motility, many studies have provided evidence that ArcA is related to virulence. For example, a recent study found that ArcA of E. coli K12, which causes human meningitis, downregulates the expression of sRNA-17 to benefit bacterial survival in blood and the penetration of the blood-brain barrier [33]. Moreover, ArcA is also required for the toxicity of Salmonella typhimurium, Vibrio cholerae, Haemophilus influenzae, and Actinobacillus pleuropneumoniae [34,35,36,37,38].

The effects of ArcA in P. shigelloides are unknown; therefore, the present study aimed to determine the correlation between ArcA and growth, motility, biofilm formation, and virulence in P. shigelloides.

Results

Phylogenetic analysis of ArcA

The two-component system response regulator ArcA of P. shigelloides is comprised of 238 amino acids. A phylogenetic tree based on ArcA amino acid sequences was constructed using the neighbor-joining method and plotted by MEGA 6.0. Bootstrap analysis was carried out based on 1000 replicates. The RopD protein of P. shigelloides was selected as the outgroup control. A dendrogram consisting of 17 species of bacteria, including some common human gut bacteria, was constructed. The comparison results showed that ArcA is conserved in all the selected bacteria. ArcA of P. shigelloides is relatively close to those from Proteus and Aeromonas, but far from those from Actinobacillus and Pseudomonas (Fig. 1).

Fig. 1
figure 1

An unrooted phylogenetic tree constructed using the neighbor joining method based on ArcA amino acid sequences. Bootstrap values were based on 1000 replications and only values greater than 50% are shown

Full size image

Identification of the deletion and complementation of arcA

A schematic illustration of the overlap-extension PCR method used for deletion of arcA is shown in Fig. 2A. The deletion and identification of arcA is showed in Fig. 2B, in which SX (800 bp) and S-arcA-X (1517 bp) are the controls for ArcA− and ArcA+, respectively. The ΔarcA isogenic deletion mutant strain was obtained (Lane 4 in Fig. 2B). To further confirm the result, we designed arcA identification primers, arcA-F and arcA-R, to amplify the arcA gene from the genomes of ΔarcA and the WT, respectively. The PCR reaction generated a negative signal with ΔarcA and a positive one with the WT (717 bp).

Fig. 2
figure 2

Confirmation of the deletion and complementation of arcA in P. shigelloides. A Graphical process of the deletion of the arcA gene. B PCR detection of the product. 1, DL2000 DNA marker; 2, PCR fragment of SX; 3, PCR amplicon of S-arcA-X from the WT genomic DNA; 4, PCR amplificon of SX from the ΔarcA genome DNA; 5, PCR amplification of arcA from the ΔarcA genome DNA; 6, PCR amplification of arcA from the WT genome DNA. C 1, DL2000 DNA marker; 2, PCR amplification of pBAD33-UD from the pBAD33 plasmid; 3, PCR amplification of pBAD33-U-arcA-D from the arcA+ complementation strain; 4, PCR amplification of arcA from the genomic DNA of the complementation strain

Full size image

The complementation of arcA is shown in Fig. 2C. The pBAD33-UD (529 bp) is a negative control. After complementation, pBAD33-U-arcA-D (1246 bp) and arcA (717 bp) were both amplified with the correct sizes.

ArcA affected the microaerobic growth of P. shigelloides

In this study, we used LB liquid medium and M9 minimal medium with only glucose as a carbon source to verify the role of ArcA in the growth and reproduction of P. shigelloides. When grown in LB liquid, the growth of ΔarcA slightly lagged behind that of the WT in the lag and log phases before 6 h, and the growth was completely restored to the WT level upon complementation with arcA (Fig. 3A). When grown in the M9 minimal medium with only glucose as the carbon source, the growth difference between ΔarcA and WT were obvious, and the ΔarcA mutants lagged behind the growth of WT before WT entered the stable phase at 12 h. Growth in M9 plus glucose was completely restored to the WT level upon complementation with arcA (Fig. 3B), which indicated that ArcA affects the uptake and utilization of glucose by P. shigelloides. In addition, the colony forming units were counted for the WT, ΔarcA and pBAD33-arcA+ strains at OD600 = 0.6, which showed that there was a 2.6-fold reduction for ΔarcA compared with that for the WT (Fig. 3C).

Fig. 3
figure 3

Deletion of arcA affected the growth of P. shigelloides in either LB medium or M9 medium under microaerobic conditions. A Bacterial strains were grown in LB and B M9 medium containing only glucose as a carbon source under microaerobic conditions, and the optical density at 600 nm (OD600) was monitored. C Bacterial strains were grown to OD600 = 0.6, and the plate colony counting method was used to count the three strains separately. The experiments were performed three times in quadruplicate. Significant differences were indicated by asterisks (***P < 0.001)

Full size image

ArcA controls the motility of P. shigelloides by directly regulating the expression of flaK, rpoN and cheV genes

In addition to ArcA being related to the growth and metabolism of P. shigelloides, we also found that ArcA is related to motility. The WT, ΔarcA and pBAD33-arcA+ strains were freshly cultured, transferred to both swimming and swarming agar plates, and incubated at 25 °C for 24–72 h. When grown in swimming agar plates, the motility of the ΔarcA strain was markedly reduced compared with the WT. There was almost no obvious movement traces after the ΔarcA strain was grown for 24 h, and it spread by 2.8 cm when cultured for 72 h (Fig. 4B). In contrast, the WT and pBAD33-arcA+ strains had overgrown the plates under the same conditions at 72 h (Fig. 4A and C). The movement data of the strains in swimming agar plates are listed in Table 1. Moreover, when grown in swarming agar plates, the motility of the ΔarcA strain was totally lost, and there was no significant change even it was cultured for 72 h. Interestingly, the WT and pBAD33-arcA+ strains showed irregular trajectories similar to radials when grown in swarming agar plates (Fig. 4D), which was rarely mentioned in previous studies. The flagella produced by the WT, ΔarcA, and pBAD33-arcA+ strains were observed by TEM. Compared to the ΔarcA mutant strain with a single flagellum, the WT and pBAD33-arcA+ strains showed the typical three-four flagella (Fig. 4E). TEM results indicated that the lack of ArcA attenuates the flagella synthesis in Plesiomonas shigelloides.

Fig. 4
figure 4

Motility of the WT, ΔarcA and pBAD33-arcA+ strains. A The WT, B ΔarcA, and C pBAD33-arcA+ strains grown in swimming agar plates for 24 h, 48 h and 72 h, respectively. D From left to right, the WT, ΔarcA and pBAD33-arcA+ strains grown in swarming agar plates for 72 h. E TEM visualization of the flagella produce by the WT, ΔarcA and pBAD33-arcA+ strains from left to right. The hollow bacterial flagella were pointed by the colored arrows

Full size image
Table 1 The movement diameter of the strains in swimming agar plates
Full size table

A previous search for the putative ArcA binding sites at the flagella gene cluster promoter region was performed using Virtual Footprint 3.0. The analysis predicted the presence of ArcA binding sites in the promoter regions of flaK, rpoN and cheV genes (see Fig. S1A to C). To confirm a direct interaction between ArcA and the predicted binding sites, ArcA-His6 fusion protein was expressed and purified (Fig. S1D), three genes promoter region were generated by PCR and used to perform EMSA with phosphorylated ArcA (ArcA-P) and non-phosphorylated ArcA (non-ArcA-P) as the negative control. The complex of protein and DNA with ArcA-P were observed when incubated with flaK, rpoN and cheV promoter fragments (Fig. 5A, B and C). The negative control (non-ArcA-P) generated no shifts even at high protein concentration (2.0 μg). Then we performed the qRT-PCR and found that the expression of flaK, rpoN and cheV decreased approximately 5.6-, 4.3-, and 2.7-fold in the ΔarcA mutant compared to the WT (Fig. 5D). The data indicated that ArcA functions as a positive regulator controls the motility of P. shigelloides by directly regulating the expression of flaK, rpoN and cheV genes.

Fig. 5
figure 5

ArcA controls the motility of P. shigelloides by directly binding the promoter regions of flaK, rpoN and cheV genes. A The EMSA of phosphorylated ArcA protein and the flaK promoter. B The EMSA of phosphorylated ArcA protein and the rpoN promoter. C The EMSA of phosphorylated ArcA protein and the cheV promoter. The concentration of phosphorylated ArcA protein was increased gradually with the non-phosphorylated ArcA as a negative control (non-ArcA-P). D The mRNA level of flaK, rpoN and cheV of the WT and ΔarcA mutant. Significant differences were indicated by asterisks (**P < 0.01)

Full size image

ArcA negatively regulates P. shigelloides biofilm formation

The biofilm formation assays were performed by both glass-tubes and 24-well plates. When the WT, ΔarcA and pBAD33-arcA+ strains were cultured in a glass-tube, the results showed that the WT could not form a biofilm. By contrast, the ΔarcA strains could form a biofilm circle at the surface of liquid, which was visible to the naked eyes. After arcA was complemented in the deletion strains, the biofilm formation ability disappeared (Fig. 6A). Furthermore, purple crystal violet staining was observed for the residue in the tubes containing the ΔarcA strain but in not the glass tubes that had contained the other two strains (Fig. 6B). In addition, we also quantitatively measured the biofilm formation ability and the results indicated that biofilm formation of ΔarcA (OD570 approximately 0.35) was 21.56-fold higher than that in the WT (Fig. 6C). In addition, for the bacteria were cultured in the 24-well culture plates, with LB only as the negative control. Compared to the ΔarcA strain, which formed an obvious biofilm at the bottom of the wells, only a small amount of residues was observed for the WT and pBAD33-arcA+ strains after being stained (Fig. 6D). The quantitative measurement results showed that biofilm formation ability of the ΔarcA (OD595 approximately 7.86) was 23.01-fold higher than that in the WT (Fig. 6E). The data of the above two biofilm formation assays indicated that ArcA fundamentally inhibits biofilm formation in P. shigelloides.

Fig. 6
figure 6

Biofilm formation assays with both glass tubes and 24-well culture plates. A The biofilm formation of the WT, ΔarcA and pBAD33-arcA+ strains cultured in the glass tubes. B The biofilm mass determined by staining surface-attached cells with crystal violet. C The biofilm formation was measured at OD570. D The biofilm formation of the WT, ΔarcA and pBAD33-arcA+ strains cultured in 24-well plates. E The biofilm formation was measured at OD595. Significant differences were indicated by asterisks (***P < 0.001)

Full size image

ArcA enhances the invasion of Caco-2 cells in P. shigelloides

Compared with the P. shigelloides WT, the ΔarcA mutant showed a 40% reduction in its capacity to invade Caco-2 cells. In contrast to the biofilm results, the pBAD33-arcA+ complementation strain could restore the invasive ability only partially, failing to reach the same level as the WT (Fig. 7). The assay was repeated four times and the difference in invasion capabilities between the WT and ∆arcA was statistically significant (p = 0.0186). The data demonstrated that ArcA could enhance the ability to invade eukaryotic cells in P. shigelloides.

Fig. 7
figure 7

The invasion capability of the WT, ∆arcA, and pBAD33-arcA+ strains into Caco-2 cells. Results were performed using analysis of variance (ANOVA) of four independent assays. Significant differences were indicated by asterisks (*P < 0.05)

Full size image

Discussion

As a facultative anaerobe, P. shigelloides can obtain energy under anaerobic or aerobic conditions through phosphorylation reactions related to electron transfer. The ArcAB binary regulatory system and the global regulatory protein FNR (ferric nitrate reductase) have been proven to play a major regulatory role in the metabolic process in response to changes in oxygen [39, 40]. Most of the known ArcA target genes of E. coli are related to aerobic respiration metabolism, and the DNA binding activity of ArcA is regulated by the reversible phosphorylation of ArcB [41]. Park et al. identified a total of 229 differentially expressed operons under anaerobic growth conditions by ChIP, among which ArcA has a direct regulatory effect on 85 of them by bioinformatic analysis [42]. At present, the role of ArcA in bacterial energy metabolism is not very clear. However, based on our comparison of the growth of P. shigelloides and the ΔarcA strain in the two media (LB and M9), it can be seen that ArcA has an impact on the metabolism of nutrients. When the ΔarcA strain was grown in M9 minimal medium with glucose as the carbon source, the glucose utilization rate was significantly lower than that of the WT before reaching the stable period. These results indicated that there is a certain connection between ArcA and the nutrition and energy metabolism of P. shigelloides.

In addition to the regulation of oxidative metabolism in bacteria, our data also confirmed that ArcA is related to bacterial motility. P. shigelloides is the unique member of the Enterobacteriaceae family that is able to produce polar flagella when grown in liquid medium and lateral flagella when grown in solid or semisolid media [43]. Previous studies have shown that P. shigelloides contained two different gene clusters, one exclusively for the lateral flagella biosynthesis and the other one containing the biosynthetic polar flagella genes [44]. The P. shigelloides polar flagella gene regions occupy higher similarity to those reported in Vibrio Parahemolyticus and Aeromonas hydrophila than the regions in E. coli or S. typhimurium [44, 45]. The primary regulatory factor of the polar flagella region of P. shigelloides is FlaK, not the FlhDC in E. coli. P. shigelloides lateral gene cluster is almost identical to the one of A. hydrophila [46]. However, no LafK ortholog could be detected in P. shigelloides even though the lafK gene has been reported in all the lateral gene clusters in the Enterobacteriaceae [46, 47]. In addition, we found that the trajectory of P. shigelloides in swarming agar plates was radial rather than circular, which was also different from the swarming motion shape of P. dendritiformis type-C [48] and Pseudomonas aeruginosa [49]. We suggest that the higher agar concentration of the swarming agar plates induced the production of lateral flagella in P. shigelloides, and resulted in a radial movement trajectory. Taken together, polar and lateral flagella transcriptional hierarchy in the P. shigelloides could represents a different Gammaproteobacteria model. Here, we provide evidence that ArcA could control the motility of P. shigelloides by directly regulating the expression of flaK, rpoN and cheV genes, and next we will focus on the flagella regulation mechanism of P. shigelloides in the future study.

Bacterial biofilms are bacteria that adhere to the surface of non-biological or active tissues in order to adapt to the living environment, and are coated in the mucus heterogeneous polymer matrix produced by themselves, forming a bacterial group that grows in a different way from planktonic bacteria [50]. Bacterial adhesion is the first step of bacterial biofilm formation. Previous studies reported that the groEL operon is related to adhesion and cell toxicity in P. shigelloides [51]. Edward et al. compared the genome sequence of 11 strains of Plesiomonas shigelloides and found that some strains contained biofilm forming proteins PgaA, PgaB and PgaC. However, subsequent experiments proved that Plesiomonas shigelloides strain EE2 can be formed even without these proteins. This indicated that P. shigelloides uses other mechanisms to regulate the formation of biofilms [52]. We found pgaC in the genome sequence of the P. shigelloides strain used in this experiment, but did not find pgaA and pgaB. At the same time, the WT showed almost no biofilm formation ability. However, after the arcA gene was deleted, the biofilm formation ability of the ΔarcA mutant strain was significantly enhanced, which indicated that ArcA has a relatively strong ability to inhibit the formation of P. shigelloides biofilms under normal conditions. Therefore, it is necessary to explore the relationship between ArcA and biofilm formation in subsequent studies. In the present study, our data also showed that ArcA is related to the virulence of P. shigelloides. Compared with the WT, the ΔarcA mutant showed a 40% reduction in infectivity of Caco-2 cells. However, the specific regulation mechanism is remains unclear. In addition, flagella [53,54,55], adhesin [56], Type 1 fimbriae [57], and curled fimbriae [58,59,60,61] are also essential for bacterial biofilm formation and virulence. They mediate the adhesion, movement, and chemotaxis of bacteria to help them seek advantages and avoid harm.

Conclusions

In this work, we report the roles of ArcA in P. shigelloides, and the data showed that ArcA could control the motility of P. shigelloides by directly regulating the expression of flaK, rpoN and cheV genes. And, the phenotype experiments in this study is significant for further discovering the specific links between ArcA and P. shigelloides in terms of growth, metabolism, biofilm formation, and virulence. Our results also laid a foundation to reveal the pathogenic mechanisms of P. shigelloides.

Materials and methods

Bacterial strains, growth conditions, and plasmids

The bacterial strains, as well as the plasmids used, are listed in Table 2. Bacteria were grown in tryptic soy broth (TSB), tryptic soy agar (TSA); and Luria-Bertani (LB) liquid, solid, and semi-solid medium at 37 °C statically or in a shaking incubator, or at 25 °C statically. If necessary, media were supplemented with ampicillin (25 μg/ml), chloramphenicol (25 μg/ml) or kanamycin (50 μg/ml).

Table 2 Bacterial strains and plasmids used in this study
Full size table

Deletion and complementation studies of arcA

In this study, an effective and precise conjugate transfer process mediated by the suicide vector pRE112 was used to make deletion mutations in the arcA gene of P. shigelloides [65]. The complementation strains was constructed by introducing the recombinant vector pBAD33-arcA+ into the ΔarcA strain via electroporation. DNA sequencing were used to confirm the presence of the correct deletion mutations and complementation strains. And all primers used in this study are shown in Table 3.

Table 3 Primers used in this study
Full size table

RNA isolation and quantitative real time PCR (qRT-PCR)

Total RNA was extracted using TRIzol® Reagent (Invitrogen, Waltham, MA, USA #15596-018) according to the manufacturer’s protocol. qRT-PCR analysis was conducted on an Applied Biosystems ABI 7500 sequence detection system with SYBR green fluorescence dye. The P. shigelloides 16S rRNA gene was used as the internal control for qRT-PCR, and relative expression levels were calculated as fold change values using the 2-△△CT method. Each experiment was carried out in triplicate.

Electrophoretic mobility shift assay (EMSA)

E. coli BL21 (DE3) with pET28a-arcA+ was grown in 200 ml of LB medium for 5 h at 30 °C, and protein expression was induced by adding 0.1 mM isopropyl beta-D-1-thio-galactopyranoside (IPTG). The ArcA-His6 fusion protein was purified using an Ni-NTA-Sefinose Column (Sangon Biotech, Shanghai, China #C600791) in accordance with the protocol provided by the manufacturer. Phosphorylation reactions of ArcA were carried out as described previously [20]. EMSAs were performed by adding increasing amounts of purified and phosphorylated ArcA-His6 fusion protein (0, 0.4, 0.8, 1.2,1.6 and 2.0 μg) to the DNA probe (50 ng) in binding buffer (100 mM Tris-HCl pH 7.5, 10 mM MgCl2, 2 mM DTT, 100 mM KCl, 10% glycerol) for 30 min at 37 °C. DNA–protein complexes were separated by 6% PAGE in 0.5 × TBE buffer (44.5 mM Tris, 44.5 mM boric acid, 1 mM EDTA, pH 8.0) at 160 V for 1 h. Gels were stained with GelRed for 10 min and imaged using a gel imaging system (GE Healthcare, Chicago, IL, USA).

Dynamic growth of the WT, ΔarcA and pBAD33-arcA + strains

The WT, ΔarcA and pBAD33-arcA+ bacterial strains were cultured overnight at 37 °C with shaking into sterile LB medium and until they reached an OD600 = 0.6. Then, the bacterial solution was added to five wells of a 96-well cell plate containing 200 μl of LB at a ratio of 1:200 per well. Fresh LB was added to the surrounding wells as a control. Finally, the prepared 96-well cell plate was placed in a Molecular Devices Spectra MAX 190 full-wavelength microplate reader (Molecular Devices, San Jose, CA, USA) to carry out the Dynamic growth experiment. The dynamic growth experiment for the WT, ΔarcA and pBAD33-arcA+ strains was also carried out in M9 medium, which contains only glucose as a carbon source. The temperature was controlled at 37 °C throughout the whole process. We conducted the experiments at three time points with five repetitions for each time.

Motility assays

The motility assays were performed as described previously [66]. Freshly grown bacterial colonies were transferred using a sterile toothpick into the center of swarming agar or swimming agar plates. The swimming agar plates were incubated for 24–72 h at 25 °C and motility was examined by the migration of bacteria through the agar from the center toward the plate periphery. Additionally, according to experimental requirements, the swarming agar plates were incubated up for 72 h at 25 °C. We conducted the experiments at three time points with six repetitions for each time.

Transmission electron microscopy (TEM)

TEM and negative staining used to visualize the flagella of the WT, ΔarcA, and pBAD33-arcA+ strains was as previously described [24].

Biofilm assay

In this study, we carried out the biofilm formation assay as described previously [67, 68] with some modifications. The WT, ΔarcA, and pBAD33-arcA+ strains were grown overnight in TSB. The next day, the overnight bacterial solution was transferred to fresh TSB and the bacteria were grown to OD600 = 0.6. The bacteria were then subcultured in fresh LB liquid medium at 1:100 and inoculated into 10 × 75 mm borosilicate glass test tubes containing 3 ml of sterile LB, and incubated at 37 °C for 20 h without shaking. Subsequently, the tubes were rinsed with phosphate-buffered saline (PBS) and filled with 0.1% crystal violet stain. After 5 min, the tubes were rinsed and then photographed. The biofilm-associated crystal violet was resuspended in dimethyl sulfoxide (DMSO), and the OD570 of the resulting suspension was measured. In addition, we also applied a 24-well tissue culture plate for the biofilm formation assay [52] on the WT, ΔarcA and pBAD33-arcA+ strains. All experiments were performed at three time-points independently and each individual samples were assayed in four repetitions.

Invasion assays

The invasion assay was carried out as described previously [69], with some modifications. Briefly, approximately 5 × 107 bacterial cells were layered onto confluent monolayers of approximately 1 × 105 Caco-2 cells (suspended in Dulbecco’s modified Eagle’s medium (DMEM)) per well in 24-well plates. The plates were centrifuged at 1000×g for 30 s to promote the sinking of bacteria, followed by incubation at 37 °C in 5% CO2 for 1 h. The monolayer washed extensively with PBS, and fresh, pre-warmed DMEM containing gentamycin (100 μg/ml) was added to kill extracellular bacteria. After 1 h of incubation, the monolayer was washed with PBS twice, and the cells were lysed with 0.1% Triton X-100 for 10 min. The released intracellular bacteria were enumerated using the plate counting method. The invasive ability was expressed as the percentage of the inoculum that survived the gentamycin treatment. We conducted the assay at four time points with six repetitions for each time.

Statistical analysis

Statistical analysis of the data was performed using analysis of variance (ANOVA). A probability value (P) ≤ 0.05 was considered statistically significant (*** p ≤ .001; ** p ≤ .01; * p ≤ .05; ns indicates not significant). The construction of the ArcA evolutionary tree used the Molecular Evolutionary Genetics Analysis (MEGA 6.0) software package [70].