AcuC, a histone deacetylase, contributes to the pathogenicity of Aeromonas hydrophila

Abstract The interactions of pathogens and phagocytes are complex. Our study demonstrated that Aeromonas hydrophila B11 can survive in the macrophagocytes of Tilapia mossambica. To explore the regulatory processes of A. hydrophila survival in the macrophagocytes, we used the mini‐Tn10 transposon mutagenesis system to build a mutant library by mixing Escherichia coli Sm10 (pLOFKm) and A. hydrophila B11. In total, 102 mutant colonies were detected, and 11 of them showed reduced survival in macrophagocytes. The mutant with the most severe phenotype, AM73, was chosen for further research. The ORF interrupted by mini‐Tn10 in AM73 was approximately 960 bp and was deposited in GenBank with the accession number SRP049226. The 319 amino acid protein encoded by the ORF showed a high degree of identity (89%) with proteins in the histone deacetylase/AcuC/AphA family of A. hydrophila subsp. hydrophila ATCC7966. A strain (AC73) in which the acuC mutation was complemented was constructed by generating the recombinant expression plasmid pACYC184‐acuC and introducing it into the AM73 mutant strain. Our experiments revealed that strain AM73 was deficient in biofilm formation, adhesion, survival in macrophagocytes, and virulence compared with A. hydrophila B11, and all of these biological properties were improved in strain AC73. The expression of 10 significant virulence genes was significantly inhibited in strain AM73. The results indicated that AcuC was an important regulatory protein contributing to the pathogenicity of A. hydrophila.

A. hydrophila is reported to be the causative agent of several diseases (da Silva et al., 2012;Peyghan, 2010) in farmed and feral fish as well as gastrointestinal and extraintestinal diseases in humans (Janda & Abbott, 2010). A. hydrophila causes high mortality in aquaculture throughout the world and results in extensive economic losses (Reyes-Becerril, López-Medina, Ascencio-Valle, & Esteban, 2011).
Therefore, an increasing amount of research is being focused on the virulence of A. hydrophila.
The process of infection includes adhesion, colonization, invasion, proliferation in the host, and the secretion of toxins Wang et al., 2015). Bacterial adhesion to host surfaces is one of the key steps in the initial infection process (Chen, Yan, Wang, Zhuang, & Wang, 2008;Luo et al., 2016), and many pathogens have been shown to develop the ability to adhere to their hosts Lin et al., 2017;Qin, Lin, Chen, Xu, & Yan, 2016). During infection, pathogens are able to neutralize bactericidal and bacteriostatic mechanisms to survive and replicate in various host cells, including macrophagocytes (Beaz-Hidalgo & Figueras, 2013). The interior of cells can protect pathogens from phagocytes and various antibiotics, which can lead to recrudescent infection (Garzoni & Kelley, 2009;Kaufmann, 2011). Therefore, the ability to survive in host phagocytes is an important virulence factor of pathogens (Qin et al., 2014). Biofilms can not only protect bacteria from host defense mechanisms, including phagocytosis, but also serve as a recalcitrant source of bacteria during antimicrobial therapy (Qin et al., 2014); therefore, biofilm formation is considered to be a virulence factor of pathogens. A. hydrophila has conventionally been considered an extracellular pathogen, but previous evidence has demonstrated that it can survive intracellularly (Qin et al., 2014). Other extracellular pathogens, including Staphylococcus aureus, have been shown to survive intracellularly (Foster, 2005;Thwaites & Gant, 2011). Currently, research on the genes that are conducive to the intracellular survival of nonobligate intracellular pathogens is limited.
The functions of the three-gene operon acuABC are involved in acetoin catabolism (Grundy, Waters, Takova, & Henkin, 1993). Acetoin utilization A (AcuA) was deduced to be an acetyltransferase of the GNAT family (Sterner & Berger, 2000). The function of acetoin utilization B (AcuB) is unknown, but the acetoin utilization C (AcuC) protein and class I histone deacetylases (HDACs) were shown to be homologs (Thiagalingam et al., 2003). AcuA and AcuC are encoded by the acuABC operon, which comprises a protein acetylation/deacetylation posttranslational modification system to control the activity of acetyl-coenzyme A (Ac-CoA) synthetase in Bacillus subtilis (Leipe & Landsman, 1997). Acetyl-CoA synthetase, a ubiquitous enzyme, is responsible for the reversible conversion of acetate to Ac-CoA. The protein acetylation/deacetylation posttranslational modification is an efficient mechanism for controlling the activity of structural proteins, gene expression regulators, and enzymes in response to rapidly changing physiological conditions (Gardner, Grundy, Henkin, & Escalante-Semerena, 2006). In Salmonella enterica, Acs activity is modulated by the protein acetylation/deacetylation system, which is critical to the synthesis of the acetyl-AMP intermediate from acetate and ATP (Starai, Celic, Cole, Boeke, & Escalante-Semerena, 2002). Studies of Escherichia coli demonstrated that the abundant lysine acetylation may modify or regulate the activities of many enzymes in pivotal metabolic processes and the synthesis of biological building blocks in response to changes in the environment (Gulick, Starai, Horswill, Homick, & Escalante-Semerena, 2003). The conservation of AcuC suggested the possibility that the acuC gene might encode a protein acetylation/ deacetylation posttranslational modification system in A. hydrophila. However, no studies of the acuC gene in A. hydrophila have been reported.
In this study, a mini-Tn10 transposon mutagenesis system was used to build a mutant library, and a mutant defective in intracellular survival was chosen for further phenotypic analysis. The aim of this study is to explore the possible mechanisms by which A. hydrophila survives in macrophages.

| Bacteria and culture conditions
The bacteria and plasmids used in this study are listed in Table 1. E. coli and A. hydrophila were cultured at 37°C and 28°C, respectively, in Luria-Bertani medium. The bacteria were washed in phosphate-buffered saline (PBS; pH 7.4) after 24 hr growth. The bacterial density was determined by OD 550 . Antibiotics were added to the medium at the following concentrations: 50 μg/ml ampicillin (Ap); 25 μg/ml chloromycetin (Cm); 100 μg/ml kanamycin (Km); and 50 μg/ml streptomycin (Sm).

| Preparation of Tilapia mossambica macrophage suspensions
Healthy Tilapia mossambica individuals were obtained from the market.
The macrophages were prepared as previously described (Leung, Low, Lam, & Sin, 1995). In brief, the fish were incubated on ice to reduce activity, and head kidneys were removed, pooled under sterile conditions and then filtered with a 100 μm filter membrane. Leibovitz-15 medium (Biological Industries, Israel) was used to suspend the samples, and heparin (10 IU/ml), streptomycin/penicillin (100 IU/ml), and 2% fetal calf serum were added. The cell suspension was layered over a 34%/51% discontinuous Percoll gradient. The cell suspension was centrifuged at 400g for 30 min at 4°C. The cells above the 34%/51% Our study interface were collected, washed and resuspended in Leibovitz-15 medium with 10% fetal calf serum, 100 IU/ml streptomycin/penicillin, and 10 IU/ml heparin. After the cell suspension was adjusted to 1 × 10 7 cells/ml, samples were transferred to six-well plates (1 ml/well).

| Preparation of mucus
Healthy Anguilla japonica (297.5 ± 17.6 g) were obtained from the aquaculture farm. Briefly, the skin mucus was prepared by scraping the surface of A. japonica (Balebona et al., 1998). The gill mucus was obtained from the gill arches of A. japonica by scraping the surface (Lumsden, Ostland, Byrne, & Ferguson, 1993). The intestinal mucus was obtained by scraping the surface of the intestine from A. japonica (Yan, Chen, Ma, Zhuang, & Wang, 2007). Mucus was mixed with sterile PBS after incubation at 4°C for 3 or 4 hr. The mixtures were centrifuged for 30 min at 20,000g at 4°C. The supernatants were filtered through 0.45 μm and 0.22 μm filters. Then, the protein concentrations of the mucus mixtures were adjusted to 1 mg/ml with sterile PBS using the method of Bradford to measure protein concentration (Bradford, 1976).

| Invasion and survival assays of the wild-type strain in macrophages in vitro
The assays were performed as described previously (Leung, Lim, Lam, & Sin, 1996). In step I, the concentration of bacteria was normalized using colony-forming units (CFUs) as described by Jin & Pancholi (2006). The exact number of CFUs was determined by plating the bacterial culture on Luria-Bertani agar plates and counting the resulting colonies. The macrophage suspension was added to six-well plates and incubated for 2 hr. In step II, the bacteria (100 bacteria/cell) were added, and the mixture was incubated at 28°C for 1 hr. The cells were fetal calf serum and 100 IU/ml streptomycin/penicillin. In step III, the cell suspension was incubated at 28°C in 5% CO 2 and sampled at 0 hr, 1 hr, 2 hr, 4 hr, 12 hr, and 24 hr. Cells were centrifuged for 5 min at 100g and 28°C. After the supernatant was removed, sterile distilled water was used to lyse the cells but not the intracellular bacteria, and the lysate was incubated for 30 min. The cell lysate was diluted 10fold with PBS. The dilutions of 10 −3 , 10 −4 , 10 −5 , 10 −6 , 10 −7 were plated on solid Luria-Bertani medium and cultivated at 28°C for 24 hr. The resulting bacterial colonies were observed and counted.

| Mutagenesis of A. hydrophila
E. coli Sm10 with pLOFKm (a type of suicide vector) was mated with A. hydrophila strain B11 to introduce the mini-Tn10Km transposon into the A. hydrophila (Herrero, de Lorenzo, & Timmis, 1990). In brief, E. coli Sm10 (mini-Tn10Km) and A. hydrophila were mixed at a ratio of 1:4 on 0.22 μm filters and cultivated on tryptone soya agar plates with 3 mmol/L isopropyl β-D-1-thiogalactopyranoside for 4 hr at 28°C.
Then, the filters were eluted with trypticase soy broth, and the tryptone soya agar plates were incubated with kanamycin and streptomycin at 28°C for 24 hr. Single colonies were selected for further study of intracellular survival. The procedure was the same as described earlier, except the incubation at the end was 1 hr.

| Southern blots
Southern blots were performed as described by Qin et al. (2014). In brief, the genomic DNA of A. hydrophila B11 and the mutant strains were extracted with a DNA extraction kit (TaKaRa, Japan). The DNA was digested with SacI (TaKaRa) restriction endonuclease, electrophoresed on a 1% agarose gel, and transferred to a nylon membrane (Liu, Mitsukawa, Oosumi, & Whittier, 1995). Using a digoxigenindeoxyuridine-triphosphate-tagged probe (Roche, Switzerland), blotting was performed to detect the mini-Tn10Km transposon. A region of the Km R gene was amplified to be the probe. The length of the probe was 176 bp, and the primers used were F Km3 : 5ˈ-CGG GGA TCG CAG TGG-3ˈ and F Km4 : 5ˈ-TGG GAA GCC CGA TGC-3ˈ with the DIG-PCR probe synthesis kit (Roche). Hybridization was performed at 42°C for 16 hr followed by washing and immunological detection with a digoxigenin detection kit (Roche). The results with single bands for the mutants were as expected.

| TAIL-PCR
TAIL-PCR was performed as previously described by Qin et al. (2014).
The random primer was provided by a genomic walking kit (TaKaRa, Japan), and the nested primers were designed on mini-Tn10 using the primer Premier Version 5.0 Tool (PREMIER Biosoft International, Palo Alto, CA) ( Table 2). The first PCR was performed with the primers L1/R1 and the random primer. The product of the first PCR and the primers L2/R2 were used for the second PCR. The third PCR was performed with the primers L3/R3 and the product of second PCR as template. The thermal cycling conditions are listed in Table 3. The DNA products were purified and cloned into pMD18-T (TaKaRa, Japan) for sequencing and then analyzed by BLAST, ClustalW and MegAlign (DNAStar) to determine the site interrupted by mini-Tn10.

| Construction of the complemented strain AC73
The acuC gene of wild-type B11 was amplified with the primers acuC- We selected the complemented AC73 strain on plates with chloromycetin and detected the expression of the AcuC protein by western blot (Merino, Rubires, Aguilar, & Tomás, 1996). In brief, the complemented AC73 strain and A. hydrophila B11 were heated to 100°C for 5 min.

BamHI
The mixture was centrifuged at 12,000g for 15 min at 4°C, and the supernatant was carefully collected to detect the expression of the AcuC protein in cells. The extracellular proteins were obtained as described by Munro, Hastings, Ellis, & Liversidge (1980). Briefly, the cultures of the complemented AC73 strain and A. hydrophila B11 grown in Luria-Bertani medium were spread on sterile cellophane films placed on the surface of solid Luria-Bertani medium plates. After incubation at 28°C for 48 hr, the bacterial cells were washed off the cellophane sheet with PBS and removed by centrifugation at 12,000 rpm at 4°C for 15 min. Supernatants were sterilized by filtration through a 0.22 μm filter (Sartorius Stedim Biotech, Germany). The intracellular and extracellular proteins were freeze-dried, dissolved in sterile water, and stained with Coomassie brilliant blue to measure the concentration.
The protein solutions were adjusted to the same concentration and heated at 100°C for 5 min. Then, 20 μl samples were separated on a 12% gel and transferred to a polyvinylidenedifluoride membrane. The membrane was blocked with 1% bovine serum albumin and incubated with a 1:1,000 dilution of anti-hemagglutinin antibody. Membranes were washed and incubated with a 1:3,000 dilution of goat anti-rabbit horseradish peroxidase-conjugated secondary antibody, and the signal was detected using a chemiluminescent substrate (PerkinElmer, Waltham). Lysates of A. hydrophila B11 were used as controls.

| Enzyme activity of the AcuC protein
The AcuC protein is a homologue of the histone deacetylase family of proteins. The activity of AcuC in the intracellular proteins of the wildtype B11 and the mutant strain AM73 was measured. In this assay, the bacteria were washed twice with ice-cold PBS and resuspended in sterile distilled water containing protease inhibitors. After being swollen for 60 min, the bacteria were subjected to a freeze-thaw cycle. The homogenate was clarified by centrifugation at 8,000g for 30 min at 4°C; the supernatant was then lyophilized, and the proteins were dissolved in water at a concentration of 1 mg/ml. The Epigenase HDAC Activity Assay Kit (Epigentek, USA) was used to detect the enzymatic activity following the manufacturer's instructions. The proteins were incubated for 90 min with the assay buffer and with the substrate stably coated on the microplate wells. The wells were washed, and the capture antibody was added. Then, the wells were washed again, and the detection antibody and the solution for color development were added. Finally, the enzyme activity was measured colorimetrically by reading the absorbance at 450 nm using a microplate spectrophotometer. The activity of AcuC was proportional to the A 450 .

| Bacterial biofilm formation assays of wild-type strain B11, AM73, and AC73
Biofilm formation was assayed as previously described by Chan & Chua (2005)

| Bacterial adhesion assays
The assays were carried out as described by Ofek

| Bacterial survival assays of AM73, AC73, and the wild-type strain
The assays were performed according to the procedure described for the invasion and survival assays in macrophages described above except that the cell suspensions were incubated at 28°C in 5% CO 2 only for 0 hr and 1 hr. The results at 0 hr and 1 hr indicate the ability of the bacteria to invade and survive, respectively, in macrophages.

| Microscopic analyses
To construct the fluorescence labeled wild-type B11, the EX-EGFP-B01 (pReceiver-B01 with enhanced green fluorescent protein) was introduced into the bacteria and measured by direct visualization with an inverted microscope using UV for excitation (Leica DM-4000B). The invasion and survival assays were performed as described above. The cell suspension was incubated at 28°C in 5% CO 2 for 0 hr and 1 hr and observed with a confocal microscope to evaluate the status of A. hydrophila B11 in cells.

| Quantitative real-time PCR
To test the effect of the acuC mutation on other significant virulence genes, quantitative real-time reverse transcription-PCR was performed to quantify the expression of the outer membrane protein genes (ompA and ompTS), the type Ш secretion system gene (ascV), pilin and flagellar family protein genes (traA, flgE, flgL, and pilB), the serine peptidase gene (degQ), the type VI secretion system-related protein gene (vash), the lipopolysaccharide gene (rfaF), the adhesin gene (cblA), and the aerolysin and hemolysin genes (hlyA and aerA) in A. hydrophila B11 and the mutant, AM73, as described by Lü et al. (2015) with some modifications. Total RNA was isolated from A. hydrophila B11 and AM73 with Trizol. The first-strand cDNA was synthesized from the total RNA using the PrimeScript ® RT kit. The acuC gene in the wild-type strain was detected by PCR with the cDNA as the template. The sequences of 13 virulence genes were obtained from the NCBI. The primers designed using Premier 5.0 software are listed in Table 4. The analysis was performed on the Step One Plus Real-Time PCR system (ABI, USA) using SYBR green I fluorescent dye.
The reactions were performed in a 10 μl volume containing 0.2 μl

| Statistical analysis
The results were statistically analyzed by ANOVA using SPSS18.0 and are reported as the means ± SD. A value of p < .05 indicated a significant difference between samples.

| Invasion and survival assays of the wild-type bacteria in macrophages in vitro
To validate the idea that A. hydrophila could invade and survive in macrophages, we analyzed the invasion and survival of A. hydrophila B11 in macrophages. The data indicated that 2.0 × 10 5 CFU/ml A. hydrophila B11 invaded the cells, and 1.5 × 10 5 CFU/ml survived in the macrophages after 1 hr. The survival at 12 hr and 24 hr was 28% and 2.1%, respectively, of the invading bacteria ( Figure 1). Thus, we determined that A. hydrophila B11 had the ability to invade macrophages, and some bacteria survived more than 24 hr.

| The isolation of mutant strains
To screen for crucial virulence genes of A. hydrophila, a mutagenesis library was constructed by introducing the mini-Tn10Km transposon on the suicide plasmid pLOFKm into A. hydrophila B11. The mutants were selected on tryptone soya agar containing 100 μg/ml kanamycin and 50 μg/ml streptomycin. Eleven colonies that exhibited significant declines in intracellular survival were selected from 102 mutants. The survival rates of the 11 mutant strains were lower than that of the wild-type B11 (p < .01) (Figure 2a). The survival rates of AM78, AM73, AM71, AM16, AM10, and AM9 were lower than those of the other mutants. These mutants have been shown to carry mutations of flgE, acuC, asmA, glpC, ahHr, and merR, respectively. The AM73 mutant, which had the lowest survival among the mutants, was chosen for further examination, and the remaining mutants were studied in our other research.

| Southern blot analysis
Southern blots were performed to detect the number of mini-Tn10Km transposons inserted into the bacterial genome. The mutants and the positive control pLOFKm showed only one copy of the transposon, whereas the wild-type strain B11 showed no transposon signal ( Figure 2b). Thus, we determined that the insertion of a single copy of the transposon mini-Tn10Km could lead to a mutation.

| Identification of the acuC gene
To identify the gene that was mutated in AM73, a series of experiments were performed. The sequence 1,500 bp upstream of the

| Complementation
To validate the effects caused by the mutation in AM73, acuC with a hemagglutinin-tag and pACYC184 were modified to generate the pACYC184-acuC recombinant expression plasmid, which was intro-

| Enzyme activity
To assess the function of acuC in A. hydrophila, the corresponding enzyme activity was measured. The activity of the intracellular protein in the B11 strain was threefold higher than the enzyme activity in AM73 ( Figure 5). This result demonstrated that the AcuC protein had HDAC enzyme activity.

| Biofilm formation
To ascertain whether acuC could affect other virulence factors as well as survival, biofilm formation by wild-type B11, AM73, and AC73 was quantified by absorbance readings (A 590 ). We found that the biofilm formation by AM73 was only 18.75% of that of the wild-type B11 and that the biofilm formation of the complemented strain was 56.25% of that of the wild type (Figure 6a). The extent of biofilm formation can be determined based on the depth of color generated by adding glacial acetic acid to the slides. The color of the B11 strain and the complemented strain AC73 was darker than that of AM73 (Figure 6b). This result demonstrated that acuC affects the biofilm formation of A. hydrophila and that acuC might be involved in the regulation of multiple virulence factors.

| Bacterial adhesion
To investigate the ability of acuC to affect other virulence factors in addition to survival, the adhesive capacity of wild-type B11, AM73, and AC73 was also tested and quantified. We employed gill mucus, intestinal mucus, and skin mucus for this test. The adhesion of wildtype B11, AM73, and AC73 was 0.85, 0.43 and 0.64, respectively, in the gill mucus, 0.68, 0.28, and 0.49 in the intestinal mucus, and 1.61, 0.64, and 1.15 in the skin mucus. The results showed that the largest number of A. hydrophila adhered to the skin mucus and that the adhesive capacity of the AM73 mutant was significantly less than that of the wild type (Figure 7). These results indicated that the acuC gene plays an important role in bacterial adhesion and that acuC might be involved in the regulation of multiple virulence factors.

| Invasion and survival assays of AM73, AC73, and the wild-type strain
The ability of B11, AM73, and AC73 to invade and survive in macrophages was compared to assess the role of acuC in invasion and survival. The data showed that the numbers of wild-type B11, AM73, and AC73 that invaded macrophages were 2.1 × 10 5 CFU/ml, 2.3 × 10 5 CFU/ml, and 2.0 × 10 5 CFU/ml, respectively, at 0 hr and that 1.4 × 10 5 CFU/ml, 1.4 × 10 4 CFU/ml, and 1.0 × 10 5 CFU/ml, respectively, survived for 1 hr in macrophages. The survival rates of the wild-type B11 and AC73 were F I G U R E 4 Western blot analysis showing a band of approximately 33 kb exclusively in the intracellular protein of AC73. The intracellular and extracellular protein of the wild-type B11 was used as the control. Each lane was loaded with 20 μg protein F I G U R E 5 The HDAC enzyme activity in the intracellular proteins of the wild-type B11 and the mutant AM73 was measured. The protein concentration was 1 mg/ml. Values denoted by different numbers of asterisks were significantly different when compared by ANOVA ("**" p < .01; "*" p < .05) F I G U R E 6 (a) The biofilm-forming ability of wild-type B11, AM73, and AC73. Values denoted by different numbers of asterisks were significantly different when compared by ANOVA ("**" p < .01; "*" p < .05). (b) The picture shows the color of crystal violet dissolved in 33% acetic acid 64.5% and 48.6%, respectively, whereas the survival of AM73 was only 6.1% at 1 hr ( Figure 8). The results at 0 hr and 1 hr indicated the invasion and intracellular survival ability of the strains. The results showed that the number of intracellular AM73 was slightly, but not significantly, larger than the number of B11 at 0 hr. However, the data at 1 hr suggested a significantly impaired survival of AM73 in host macrophages.
These results showed that acuC plays a vital role in the survival of A. hydrophila in host macrophages.

| Infection of Danio rerio
The defects in biofilm formation, adherence and intracellular survival of the mutant indicated that acuC plays an important role in the patho-

| Expression of various virulence genes
Quantitative real-time reverse transcription-PCR showed that a lipopolysaccharide gene (rfaF) and one of the type VI secretion systemrelated protein genes (vash) were significantly up-regulated in AM73, while the other type VI secretion system-related protein gene (traA) showed no obvious change. The expression of two outer membrane protein genes (ompA and ompTS), the type Ш secretion system gene

| DISCUSSION
In this study, a pathogenic strain of A. hydrophila B11 was found to invade the macrophages of Tilapia mossambica in vitro, and many of the bacteria survived in the macrophages for at least 24 hr. Once the bacteria entered the phagocytes, they must resist disruption by vari- The protein acetylation/deacetylation posttranslational modification is an efficient mechanism for controlling the activity of structural proteins, gene expression regulators, and enzymes in response to rapidly changing physiological conditions (Gardner et al., 2006;Hu, Lima, & Wolfe, 2010). In Bacillus subtilis, the activity of acetylcoenzyme A (Ac-CoA) synthetase is regulated by the protein acetylation/deacetylation posttranslational modification system comprised of AcuA and AcuC (Leipe & Landsman, 1997). OatA, a peptidoglycan Oacetyltransferase, has been shown to be involved in Listeria monocytogenes immune escape and is critical for virulence (Aubry et al., 2011).
In Legionella pneumophila, histone acetylation was demonstrated to be important for virulence (Schmeck et al., 2008). In this study, the deletion of acuC was found to reduce the intracellular survival ability  (Ho, Sohel, & Schoolnik, 1992;Qin et al., 2014;Quinn et al., 1994). Flagellar motility has been demonstrated to be necessary for A. hydrophila adhesion . Outer membrane proteins are important adhesion factors and protective antigens closely related to the virulence of A. hydrophila (Kawai, Liu, Ohnishi, & Oshima, 2004). OmpA has also been reported to be correlated with the adhesion of A. hydrophila (Quinn et al., 1994).
FlgE, the structural gene encoding the flagellar hook protein, has also been reported to be correlated with the adhesion of A. hydrophila (Qin et al., 2014). It has previously been reported that the pilin gene, the gene coding for the flexible pilus subunit, was correlated with adhesion in A. hydrophila (Ho et al., 1992). In this study, the adhesion ability of A. hydrophila was significantly inhibited after deletion of acuC; this result is consistent with the gene expression results.
Biofilm formation has previously been reported to occur in three main stages: (1) attachment, specific protein-based binding to abiotic surfaces; (2) proliferation and formation of mature biofilm structures; and (3)  F I G U R E 1 1 Expression profiles of A. hydrophila virulence-related genes assayed by qRT-PCR. The expression levels of acuC and 13 virulence genes in the wild-type strain and AM73. The 16S rRNA was used as the internal control. Each bar represents the mean value of quintuplicates ± SD. Values denoted by different numbers of asterisks were significantly different when compared by ANOVA ("**" p < .01; "*" p < .05) 2000). Biofilm assays showed that the mutant strain AM73 was significantly defective in forming biofilms. This result indicated that acuC is involved in biofilm formation by A. hydrophila.
Unlike the well-studied bacterial pathogenic processes, information on the intracellular survival of A. hydrophila is still very limited. Qin et al. (2014) found that a ΔflgE mutant of A. hydrophila showed significantly reduced intracellular survival in phagocytes. In this study, the intracellular survival of AM73 was significantly lower than the survival of the wild type. In addition, the expression of motility-related genes, including flgE, traA, flgL, and pilB was significantly down-regulated in AM73. Previous studies have shown that bacterial motility affects bacterial survival in phagocytes (Qin et al., 2014). In this study, the expression of rfaF was significantly down-regulated in AM73. The rfaF gene is associated with the synthesis of lipopolysaccharide heptosyltransferase II. Guo et al. (1997) found that the proteins of lipopolysaccharide modification contribute to bacterial survival in macrophages by conferring resistance to antimicrobial peptides and nutrient scavenging. The expression of ascV, one of the type III secretion system genes, was also significantly down-regulated in AM73. Some previous reports have revealed that the secreted molecules of the type III secretion system affect bacterial internalization (Cirillo, Valdivia, Monack, & Falkow, 1998).
In this study, the deletion of acuC resulted in the reduction in mortality of infected Danio rerio. The deletion of acuC also resulted in the down-regulation of several genes, including hlyA, in the mutant strains.
A hemolytic toxin, the product of hlyA, has also been reported to induce bacterial infection (Qian, Chen, Shen, & Shen, 1995