In Vitro Antibiofilm Activity of Resveratrol against Aeromonas hydrophila

Aeromonas hydrophila is a Gram-negative bacterium that widely exists in various aquatic environments and causes septicemia in fish and humans. Resveratrol, a natural polyterpenoid product, has potential chemo-preventive and antibacterial properties. In this study, we investigated the effect of resveratrol on A. hydrophila biofilm formation and motility. The results demonstrated that resveratrol, at sub-MIC levels, can significantly inhibit the biofilm formation of A. hydrophila, and the biofilm was decreased with increasing concentrations. The motility assay showed that resveratrol could diminish the swimming and swarming motility of A. hydrophila. Transcriptome analyses (RNA-seq) showed that A. hydrophila treated with 50 and 100 μg/mL resveratrol, respectively, presented 230 and 308 differentially expressed genes (DEGs), including 90 or 130 upregulated genes and 130 or 178 downregulated genes. Among them, genes related to flagellar, type IV pilus and chemotaxis were significantly repressed. In addition, mRNA of virulence factors OmpA, extracellular proteases, lipases and T6SS were dramatically suppressed. Further analysis revealed that the major DEGs involved in flagellar assembly and bacterial chemotaxis pathways could be regulated by cyclic-di-guanosine monophosphate (c-di-GMP)- and LysR-Type transcriptional regulator (LTTR)-dependent quorum sensing (QS) systems. Overall, our results indicate that resveratrol can inhibit A. hydrophila biofilm formation by disturbing motility and QS systems, and can be used as a promising candidate drug against motile Aeromonad septicemia.


Introduction
Aeromonas hydrophila is a typically zoonotic pathogen with a worldwide distribution in aquatic environment, and has different pathogenic potential between aquatic animals and mammals [1,2]. It can not only cause motile aeromonad septicemia in fish, leading to severe economic losses in aquaculture, but also cause various diseases in humans, such as gastroenteritis, skin infections, meningitis, and necrotizing fasciitis [3,4]. Traditionally, antibiotics are widely used to treat A. hydrophila infections. However, the wide and frequent use of antibiotics in fish farming has led to the emergence of antibiotic resistance in pathogens and treatment fail ure [5,6]. This phenomenon threatens the healthy development of aquaculture and safety of aquatic products. Therefore, novel strategies are required to combat drug resistant A. hydrophila.
A. hydrophila possesses a variety of virulence factors, among which biofilm has been well characterized and considered one of the most critical virulence factors for establishing infections [7][8][9]. Biofilms are sessile microbial communities that attach to surfaces using a self-produced extracellular polymeric matrix [10]. The bacteria in biofilms become much more resistant to antibiotic treatments and host defense compared to their free-floating state [11]. Biofilm formation greatly enhances the survival of pathogens in hosts, and results in the persistence of bacterial infections. The beginning of biofilm formation is the

Effects of Resveratrol on Biofilm Structure
Microscopic observations matched the quantitative biofilm data well ( Figure 2). In the absence of resveratrol, the biofilm formed by A. hydrophila NJ-35 had a uniform distribution with a dense coverage of the coverslip. When treated with resveratrol, the biofilm formed by A. hydrophila NJ-35 was sparser than that in the control group, and the biofilm structure became looser as the concentration of resveratrol increased.

Effects of Resveratrol on Motility
The effects of resveratrol on the swimming and swarming motilities of A. hydrophila NJ-35 were also examined. As shown in Figure 3, resveratrol at 50 μg/mL did not affect the swimming motility, whereas resveratrol at 100 μg/mL significantly reduced the swimming motility of A. hydrophila NJ-35 (p < 0.05). In contrast, both 50 and 100 μg/mL resveratrol noticeably decreased the swarming motility of A. hydrophila NJ-35 in a dose-dependent manner (p < 0.05). These results suggested that the inhibition of biofilm formation by resveratrol at 50 μg/mL was more closely related to swarming rather than swimming,

Effects of Resveratrol on Biofilm Structure
Microscopic observations matched the quantitative biofilm data well ( Figure 2). In the absence of resveratrol, the biofilm formed by A. hydrophila NJ-35 had a uniform distribution with a dense coverage of the coverslip. When treated with resveratrol, the biofilm formed by A. hydrophila NJ-35 was sparser than that in the control group, and the biofilm structure became looser as the concentration of resveratrol increased. The data were presented as the mean ± SD (n = 3) of three independent experiments. * p < 0.05, ** p < 0.01.

Effects of Resveratrol on Biofilm Structure
Microscopic observations matched the quantitative biofilm data well ( Figure 2). In the absence of resveratrol, the biofilm formed by A. hydrophila NJ-35 had a uniform distri bution with a dense coverage of the coverslip. When treated with resveratrol, the biofilm formed by A. hydrophila NJ-35 was sparser than that in the control group, and the biofilm structure became looser as the concentration of resveratrol increased.

Effects of Resveratrol on Motility
The effects of resveratrol on the swimming and swarming motilities of A. hydrophila NJ-35 were also examined. As shown in Figure 3, resveratrol at 50 μg/mL did not affec the swimming motility, whereas resveratrol at 100 μg/mL significantly reduced the swim ming motility of A. hydrophila NJ-35 (p < 0.05). In contrast, both 50 and 100 μg/mL resvera trol noticeably decreased the swarming motility of A. hydrophila NJ-35 in a dose-depend ent manner (p < 0.05). These results suggested that the inhibition of biofilm formation by resveratrol at 50 μg/mL was more closely related to swarming rather than swimming

Effects of Resveratrol on Motility
The effects of resveratrol on the swimming and swarming motilities of A. hydrophila NJ-35 were also examined. As shown in Figure 3, resveratrol at 50 µg/mL did not affect the swimming motility, whereas resveratrol at 100 µg/mL significantly reduced the swimming motility of A. hydrophila NJ-35 (p < 0.05). In contrast, both 50 and 100 µg/mL resveratrol noticeably decreased the swarming motility of A. hydrophila NJ-35 in a dose-dependent manner (p < 0.05). These results suggested that the inhibition of biofilm formation by resveratrol at 50 µg/mL was more closely related to swarming rather than swimming, while resveratrol at 100 µg/mL was involved in both swimming and swarming motilities.
biotics 2023, 12, x FOR PEER REVIEW 4 of Figure 3. Effects of resveratrol on the swimming and swarming motility of A. hydrophila. The abilit of swimming (A,B) and swarming (C,D) motility were assessed by examining the migration of b teria through the agar, from the center toward the periphery of the plate. The results were rep duced in three independent experiments, and the error bars represent SDs. * p < 0.05.

Cytotoxicity of Resveratrol on J774A.1, with or without A. hydrophila NJ-35 Infection
The cytotoxic effect of resveratrol, at sub-MIC concentrations on J774A.1 cells, w detected by LDH activity. As shown in Figure 4, the cytotoxicity of 1% DMSO on J774A cells was less than 1%, and there was no significant difference in cytotoxicity between t 50 μg/mL resveratrol and 1% DMSO. However, the cytotoxicity of resveratrol on J774A cells remarkably increased when co-cultured with 100 μg/mL resveratrol. To directly e amine the antibacterial activity, J774A.1 cells were exposed to A. hydrophila NJ-35, treat with or without resveratrol. Compared to the drug-free group (only A. hydrophila NJinfection), 50 μg/mL resveratrol significantly decreased the LDH release of J774A.1 ce but no significant difference was observed between the 100 μg/mL resveratrol and dru free groups. The results indicate that resveratrol within 50 μg/mL is not cytotoxic J774A.1 cells and has a good protective effect on the J774A.1 cells infected by A. hydroph NJ-35.  The cytotoxic effect of resveratrol, at sub-MIC concentrations on J774A.1 cells, was detected by LDH activity. As shown in Figure 4, the cytotoxicity of 1% DMSO on J774A.1 cells was less than 1%, and there was no significant difference in cytotoxicity between the 50 µg/mL resveratrol and 1% DMSO. However, the cytotoxicity of resveratrol on J774A.1 cells remarkably increased when co-cultured with 100 µg/mL resveratrol. To directly examine the antibacterial activity, J774A.1 cells were exposed to A. hydrophila NJ-35, treated with or without resveratrol. Compared to the drug-free group (only A. hydrophila NJ-35 infection), 50 µg/mL resveratrol significantly decreased the LDH release of J774A.1 cells, but no significant difference was observed between the 100 µg/mL resveratrol and drugfree groups. The results indicate that resveratrol within 50 µg/mL is not cytotoxic to J774A.1 cells and has a good protective effect on the J774A.1 cells infected by A. hydrophila NJ-35.  The data were presented as the mean ± SD (n = 6). * p < 0.05, ** p < 0.01.

Transcriptome Analysis of A. hydrophila NJ-35 Treated with Resveratrol
As demonstrated above, A. hydrophila NJ-35 showed reduced biofilm formation and motility in a dose-dependent manner when treated with resveratrol. In order to determine the molecular mechanism of these changes, RNA-seq was performed on A. hydrophila NJ-35 with 0, 50 and 100 μg/mL resveratrol (coded as Res 0, Res 50 and Res 100 groups). The differentially expressed genes (DEGs) were analyzed to infer the candidate genes related to biofilm and motility affected by resveratrol treatment, and to reveal the possible function of these differential genes and related molecular mechanisms.

Screening and Functional Enrichment Analysis
Using the criteria of fold-change ≥ 2 and p-value < 0.05, we identified 230 (90 up-and 140 downregulated) and 308 DEGs (130 up-and 178 downregulated) in the presence of 50 and 100 μg/mL resveratrol compared to cultures grown without resveratrol, respectively, of which 120 DEGs were common in both datasets ( Figure    The data were presented as the mean ± SD (n = 6). * p < 0.05, ** p < 0.01.

Transcriptome Analysis of A. hydrophila NJ-35 Treated with Resveratrol
As demonstrated above, A. hydrophila NJ-35 showed reduced biofilm formation and motility in a dose-dependent manner when treated with resveratrol. In order to determine the molecular mechanism of these changes, RNA-seq was performed on A. hydrophila NJ-35 with 0, 50 and 100 µg/mL resveratrol (coded as Res 0, Res 50 and Res 100 groups). The differentially expressed genes (DEGs) were analyzed to infer the candidate genes related to biofilm and motility affected by resveratrol treatment, and to reveal the possible function of these differential genes and related molecular mechanisms.

Screening and Functional Enrichment Analysis
Using the criteria of fold-change ≥ 2 and p-value < 0.05, we identified 230 (90 up-and 140 downregulated) and 308 DEGs (130 up-and 178 downregulated) in the presence of 50 and 100 µg/mL resveratrol compared to cultures grown without resveratrol, respectively, of which 120 DEGs were common in both datasets ( Figure  The data were presented as the mean ± SD (n = 6). * p < 0.05, ** p < 0.01.

Transcriptome Analysis of A. hydrophila NJ-35 Treated with Resveratrol
As demonstrated above, A. hydrophila NJ-35 showed reduced biofilm formation and motility in a dose-dependent manner when treated with resveratrol. In order to determine the molecular mechanism of these changes, RNA-seq was performed on A. hydrophila NJ-35 with 0, 50 and 100 μg/mL resveratrol (coded as Res 0, Res 50 and Res 100 groups). The differentially expressed genes (DEGs) were analyzed to infer the candidate genes related to biofilm and motility affected by resveratrol treatment, and to reveal the possible function of these differential genes and related molecular mechanisms.

Screening and Functional Enrichment Analysis
Using the criteria of fold-change ≥ 2 and p-value < 0.05, we identified 230 (90 up-and 140 downregulated) and 308 DEGs (130 up-and 178 downregulated) in the presence of 50 and 100 μg/mL resveratrol compared to cultures grown without resveratrol, respectively, of which 120 DEGs were common in both datasets ( Figure    To gain insight into the functions of the DEGs that were altered by resveratrol treatment, all of the DEGs were mapped to terms in the GO and KEGG databases. GO analysis showed that among the up-and down-regulated genes of A. hydrophila NJ-35 in Res  50 and Res 100 groups, "metabolic process", and "cellular process" in biological process, "cell", "cell part", and "membrane" in the cellular component, "catalytic activity", "binding", and "transporter activity" in molecular function were the major enriched groups ( Figure 6). In addition, we noted that the number of downregulated genes distributed in the biological process of "response to stimulus" and "locomotion" was much higher than the number of upregulated genes. The top 5 KEGG pathways were as follows: "phenylalanine metabolism", "tyrosine metabolism", and "bacterial chemotaxis" in both Res 50 and Res 100 groups, "phenylalanine, tyrosine and tryptophan biosynthesis" and "histidine metabolism" in Res 50 group, "plant-pathogen interaction" and "selenocompound metabolism" in Res 100 group (Figure 7). It can be found that "bacterial chemotaxis" was the key enrichment pathway of the DEGs in both Res 50 and Res 100 groups, which indicated that the treatment of resveratrol had a significant impact on the motility of A. hydrophila NJ-35. To gain insight into the functions of the DEGs that were altered by resveratrol treatment, all of the DEGs were mapped to terms in the GO and KEGG databases. GO analysis showed that among the up-and down-regulated genes of A. hydrophila NJ-35 in Res 50 and Res 100 groups, "metabolic process", and "cellular process" in biological process, "cell", "cell part", and "membrane" in the cellular component, "catalytic activity", "binding", and "transporter activity" in molecular function were the major enriched groups ( Figure 6). In addition, we noted that the number of downregulated genes distributed in the biological process of "response to stimulus" and "locomotion" was much higher than the number of upregulated genes. The top 5 KEGG pathways were as follows: "phenylalanine metabolism", "tyrosine metabolism", and "bacterial chemotaxis" in both Res 50 and Res 100 groups, "phenylalanine, tyrosine and tryptophan biosynthesis" and "histidine metabolism" in Res 50 group, "plant-pathogen interaction" and "selenocompound metabolism" in Res 100 group (Figure 7). It can be found that "bacterial chemotaxis" was the key enrichment pathway of the DEGs in both Res 50 and Res 100 groups, which indicated that the treatment of resveratrol had a significant impact on the motility of A. hydrophila NJ-35.  In this study, we focused our analysis on the major genes which were related to biofilm formation and motility. The KEGG classifications associated with "cell motility" and "cellular community prokaryotes" can be found in both Res 50 and Res 100 groups, which

Analysis of DEGs Related to Biofilm Formation and Motility
In this study, we focused our analysis on the major genes which were related to biofilm formation and motility. The KEGG classifications associated with "cell motility" and "cellular community prokaryotes" can be found in both Res 50 and Res 100 groups, which contained 12 and 2 DEGs in the Res 50 group, 13 and 5 DEGs in the Res 100 group, respectively ( Figure S1). Notably, most of them were downregulated after resveratrol treatment. Not surprisingly, the expressions of the DEGs in "bacterial chemotaxis (ko02030)" and "flagellar assembly (ko02040)" pathways were almost repressed in resveratrol-treated groups. Interestingly, two genes associated with type IV pilus were downregulated, while the other two genes, encoding Flp family proteins, were upregulated. In addition, two GGDEFdomain-containing proteins (encoded by U876_RS00255 and U876_RS04630), that are required for the production of the second messenger cyclic-di-guanosine monophosphate (c-di-GMP) and a helix-turn-helix transcriptional regulator (encoded by U876_RS01620) belonging to LysR-Type transcriptional regulators (LTTRs), were decreased. Besides the genes mentioned above, resveratrol also inhibited the expression of other virulence factors, including OmpA, T6SS, lipase and extracellular proteases (protease, elastase and collagenase) ( Table 1).

Validation of Differentially Expressed Genes by qRT-PCR
A total of five genes were randomly selected to be measured the relative mRNA transcript levels using real-time quantitative PCR. As shown in the Figure 8, the qRT-PCR results showed similar expression tendency as the RNA-seq data, despite some quantitative differences at the expression level. The result suggested that the transcription abundance of DEGs in transcriptome analysis was highly reliable.

Validation of Differentially Expressed Genes by qRT-PCR
A total of five genes were randomly selected to be measured the relative mRNA transcript levels using real-time quantitative PCR. As shown in the Figure 8, the qRT-PCR results showed similar expression tendency as the RNA-seq data, despite some quantitative differences at the expression level. The result suggested that the transcription abundance of DEGs in transcriptome analysis was highly reliable.

Discussion
A. hydrophila, one of the most common bacterial pathogens in aquaculture environments, has been reported to be resistant to a number of antibiotics [29,30]. The role of biofilm formation in A. hydrophila pathogenesis is well established, as it provides the bacterium with enhanced tolerance to antimicrobial agents and host defenses [8,9,31]. Therefore, inhibiting biofilm formation is of great significance to combat the infection of A. hydrophila. Antibiotic therapy typically alleviates the symptoms caused by planktonic bacteria, but fails to kill bacteria in biofilms [32]. During aggravation of bacterial resistance, plant-derived compounds have attracted much attention because of their safety, availability and low toxicity. Resveratrol is a natural plant polyphenol that occurs in various plants, and has been demonstrated to have antibiofilm effects on many Gram-negative and Grampositive bacterial pathogens [33]. In the present study, resveratrol, at the sub-inhibitory concentrations, can significantly inhibit the biofilm formation of A. hydrophila NJ-35 in a dose-dependent manner, which was consistent with our previous findings [28]. Thus, we further investigated the action mechanism of resveratrol on A. hydrophila biofilms.
Bacterial adhesion and colonization play an important role in the process of biofilm formation [34]. With the formation of biofilm, bacterial pathogenicity enhances significantly. In our study, the biofilms formed by A. hydrophila NJ-35 on the slides were destroyed by resveratrol; meanwhile, as the resveratrol concentration increased, the biofilms decreased gradually and became sparsely distributed on the slides, which had a similar trend to the results obtained by crystal violet staining, EPS production and the total biofilm protein.Therefore, we speculated that resveratrol decreased the adhesion of A. hydrophila NJ-35, thus inhibiting biofilm formation. This finding was also confirmed in S. aureus and avian pathogenic E. coli [24,35].
Motility has a positive influence on the development of biofilm formation, as it facilitates pathogens to colonize, adhere and invade the host cells [36,37]. Aeromonas species possess polar flagella for swimming motility and lateral flagella for swarming motility over surfaces [38]. The main antibiofilm activities of resveratrol comprised the inhibition of QS and motility [21]. Our study showed that resveratrol diminished the swimming and swarming motility of A. hydrophila NJ-35 in a dose-dependent manner. Particularly, resveratrol dramatically repressed swarming motility even at 50 µg/mL. In EHEC, trans-resveratrol inhibited both swimming and swarming motility, and suppressed the expression of several key motility and flagellar genes, including flhD, fimA, fimH, and motB [10]. Additionally, reduced swarming ability has also been reported for Proteus mirabilis and V. vulnificus [39,40]. Given the above, resveratrol might impel the biofilm formation of A. hydrophila NJ-35 via disrupting its adhesion and inhibiting swimming and swarming motility.
We further explored the molecular mechanism of these changes through transcriptome analyses ( Figure 9). With the increasing concentration of resveratrol, numbers of DEGs in the Res 50 and Res 100 groups increased. As was expected, a large number of DEGs were significantly enriched in biofilm-and motility-related pathways, especially in flagellar and type IV pilus assembly, as well as in bacterial chemotaxis. It has been reported that flagellar motility and chemotaxis exert profound influences on bacterial behaviors, including swarming, biofilm formation and auto-aggregation; the type IV pilus plays an important role in structural biofilm development, including initial formation and dispersal [41,42]. Thus, Resveratrol indeed regulated the biofilm formation and motility of A. hydrophila at the transcriptional level.
It is known that the GGDEF-domain-containing proteins contribute to the synthesis of the intracellular signaling molecule c-di-GMP. The latter coordinates the bacterial lifestyle transition from motility to sessility and vice versa [43]. Generally, low c-di-GMP levels are conducive to bacterial motility, such as swimming, twitching and swarming, and inconducive to biofilm formation. In contrast, high c-di-GMP levels promote biofilm formation and restrict bacterial motility [43,44]. Kozlova et al. [45] have demonstrated that c-di-GMP overproduction in A. hydrophila SSU dramatically increased the transcripts of luxS, litR, vpsT, fleQ, and fleN, thus enhancing biofilm formation, but reducing motility, which was consistent with that of the ∆luxS. Unlike ∆luxS, overproduction of c-di-GMP in ∆ahyRI resulted in a slight increase in biofilm formation with no effect on motility, due to the high level of c-di-GMP upregulating the transcriptional level of vpsR and downregulating the levels of vpsT and fleN. On the contrary, ∆qseB exhibited a significant reduction in biofilm formation with no effect on swimming motility when c-di-GMP was overproduced, which was correlated with altered levels of fleN and vpsT [46]. In addition, resveratrol has been demonstrated to repress the expression of ahyR and ahyI, as well as T6SS [15,47]. Thus, we indicated that resveratrol may inhibit A. hydrophila biofilm formation though AhyRI QS system in a c-di-GMP manner. However, further studies are necessary to demonstrate this.
biotics 2023, 12, x FOR PEER REVIEW 11 of Figure 9. Schematic diagram of the inhibitory effects of resveratrol on A. hydrophila. "┬" deno inhibition.

Bacterial Strains and Growth Conditions
A. hydrophila NJ-35 (CGMCC No.8319) was cultured in Luria broth (LB) or on LB ag at 28 °C [55]. Resveratrol (99%, Aladdin, Shanghai, China) was dissolved in dimethyl s foxide (DMSO) as a 10 mg/mL stock solution, and diluted to the required working co centrations depending on the assay type.

Crystal Violet Biofilm Assay
An assay of static biofilm formation was performed in 96-well polystyrene plates, previously reported [56]. An overnight culture of A. hydrophila NJ-35 was collected a normalized to 1 × 10 7 CFU/mL, then diluted 1:100 in a fresh LB medium. In our earl studies, the MIC of resveratrol against A. hydrophila NJ-35 was recorded as 1024 μg/m whereas the resveratrol concentration below 64 μg/mL did not affect the growth of hydrophila NJ-35 [28]. Resveratrol was added for experimental cultures at final concent tions of 50 and 100 μg/mL, respectively. DMSO (0.1%, v/v) was added as the control grou Two hundred microliters of the above dilutions were dispensed to the wells of microti plates and incubated at 28 °C for 48 h without agitation. Following incubation, cell grow was measured at 600 nm. Then, the suspended culture was poured out and the wells we washed three times with sterile phosphate-buffered saline (PBS). The adherent cells we The LTTRs are the largest family of diverse and well-characterized global transcriptional regulators of prokaryotes [48]. This family of proteins is involved in the regulation of various processes, including multidrug resistance, virulence, QS, motility and biofilm formation [49][50][51]. An in-frame deletion of Bcal3178 (a LysR-type regulator) caused a significant downregulation of biofilm formation and protease production, which are controlled by QS systems in Burkholderia cenocepacia [50]. Four putative LTTR family proteins (A0KIU1, A0KJ82, A0KPK0, and A0KQ63) were decreased in A. hydrophila following antibiotic treatment, and the deletion of A0KQ63 exhibited multidrug resistance properties [52]. In this study, U876_RS01620 encoding a LTTR was downregulated after adding resveratrol in A. hydrophila, while the expression of the two other genes encoding the SMR family of multidrug efflux pumps were repressed. One possible explanation is that the protein encoded by U876_RS01620 may not affect the above two genes, and resveratrol may activate other pathways to regulate SMR family proteins. In this regard, it may be of interest to evaluate whether resveratrol regulates drug resistance of A. hydrophila through the LTTR and whether the LTTR is mediated by the c-di-GMP.
The pathogenesis of A. hydrophila is multifactorial, and characterized by the involvement of a number of virulence factors, such as adhesins, outer membrane proteins, hemolysins, protease, and secretion system [53,54]. Our study showed that resveratrol dramatically decreased the gene expression of OmpA, lipase, protease, elastase, collagenase, and T6SS. In addition, resveratrol has been demonstrated to affect the QS-related gene expression, weaken the hemolytic activity in vitro, and attenuate the in vivo virulence of A. hydrophila in the crucian carp infection, supporting the protective role of resveratrol against fish disease [28,47]. Following these results, we speculated that the natural compound resveratrol might inhibit A. hydrophila biofilm formation by disturbing the c-di-GMPand LTTR-dependent QS systems, which could regulate adhesion, motility and virulence. Further study needs to be performed.

Bacterial Strains and Growth Conditions
A. hydrophila NJ-35 (CGMCC No.8319) was cultured in Luria broth (LB) or on LB agar at 28 • C [55]. Resveratrol (99%, Aladdin, Shanghai, China) was dissolved in dimethyl sulfoxide (DMSO) as a 10 mg/mL stock solution, and diluted to the required working concentrations depending on the assay type.

Crystal Violet Biofilm Assay
An assay of static biofilm formation was performed in 96-well polystyrene plates, as previously reported [56]. An overnight culture of A. hydrophila NJ-35 was collected and normalized to 1 × 10 7 CFU/mL, then diluted 1:100 in a fresh LB medium. In our earlier studies, the MIC of resveratrol against A. hydrophila NJ-35 was recorded as 1024 µg/mL, whereas the resveratrol concentration below 64 µg/mL did not affect the growth of A. hydrophila NJ-35 [28]. Resveratrol was added for experimental cultures at final concentrations of 50 and 100 µg/mL, respectively. DMSO (0.1%, v/v) was added as the control group. Two hundred microliters of the above dilutions were dispensed to the wells of microtiter plates and incubated at 28 • C for 48 h without agitation. Following incubation, cell growth was measured at 600 nm. Then, the suspended culture was poured out and the wells were washed three times with sterile phosphate-buffered saline (PBS). The adherent cells were fixed with 200 µL methanol for 15 min and dried at room temperature. Subsequently, each well was stained with 200 µL of 1% (wt/vol) crystal violet solution for 15 min and washed with PBS to remove the unbound dye. The formed biofilm was dissolved in absolute ethanol, and the absorbance was measured at 590 nm using a spectrophotometer (MultiskanGO, Thermo Scientific, Vantaa, Finland). The results were normalized to reduce the differences caused by bacterial growth rates according to the method of Niu and Gilbert [57].

Exopolysaccharides Assay
A. hydrophila NJ-35 was cultured with or without resveratrol in 6-well plates, at 28 • C for 48 h. The formed biofilms were washed three times with PBS, then resuspended in 0.85% NaCl containing 0.22% formaldehyde. After centrifugation, the supernatants were collected and the EPS were determined by the phenol-sulfuric acid method [58].

Total Biofilm Protein Assay
The biofilms formed in 6-well plates were resuspended in 1 mL PBS. After ultrasonication, the total biofilm protein concentrations were detected using the Modified Bradford Protein Assay Kit (Sangon Biotech, Shanghai, China).

Scanning Electron Microscopy
The inhibitory effect of resveratrol on the biofilm formation of A. hydrophila NJ-35 was observed using SEM. Overnight bacterial culture was normalized to 1 × 10 7 CFU/mL and diluted 1:100 in LB medium with 50 and 100 µg/mL resveratrol, respectively, and 0.1% (v/v) DMSO was added as a control. Pre-sterilized coverslips (Φ = 14 mm) were placed into 12-well plates, then 1 mL bacterial dilution was dispensed to each well and incubated at 28 • C for 48 h. The slides were rinsed well with PBS and fixed with 2.5% glutaraldehyde for 4 h. The samples were dehydrated with a series of gradient acetone (10,30,50,70,90, and 100%, v/v) for 15 min, and dehydrated in 100% acetone twice. Finally, the sample was observed under a scanning electron microscope (S-4800, Hitachi, Tokyo, Japan). Nine random positions in three independent experiments were chosen for microscopic analysis.

Lactate Dehydrogenase Assay
J774A.1 murine macrophages were cultured in DMEM medium with high glucose (HyClone, Beijing, China) containing 10% fetal bovine serum (FBS, Every green, Hangzhou, China) at 37 • C with 5% CO 2 . Cells (1 × 10 5 cells/well) were cultured in 96-well plates for 24 h, then washed three times with PBS. Cells were divided into two groups: resveratrol treatment group and A. hydrophila NJ-35 + resveratrol treatment group. DMEM medium and 1% DMSO served as negative control and solvent control, respectively. For infection, bacteria grown to logarithmic phase were collected, washed and seeded with or without resveratrol into each well at the multiplicity of infection (MOI) of 1:1. After 3 h of incubation, the supernatant was harvested and the LDH activity was measured by the LDH Cytotoxicity Assay Kit (Invitrogen, Carlsbad, CA, USA).

Swimming and Swarming Motility
An LB medium with 0.3% agar for swimming motility and 0.5% agar for swarming motility were prepared, as described previously. Resveratrol was added to the LB agar to final concentrations of 50 or 100 µg/mL. Additionally, DMSO (0.1%) was added as a control. Overnight bacterial culture was normalized to 1 × 10 7 CFU/mL with the LB medium. Then, 5 µL culture was spotted into the middle of the plate and incubated at 28 • C for 24 h. Motility was assessed by measuring the migration diameter of bacteria from the inoculation point to the periphery of the plate.

Transcriptome Analysis
A. hydrophila NJ-35 was inoculated into 100 mL LB medium, supplemented with 50 and 100 µg/mL resveratrol at an initial OD 600 of 0.05, and cultured at 28 • C until the OD 600 reached 0.8. Bacteria treated with 0.1% DMSO were set as a control group. Cells were collected and washed with PBS for RNA-seq using Illumina HiSeq TM 2500 at Shanghai OE Biotech Co., Ltd. (Shanghai, China) Three parallel samples for each group were pooled as biological replicates for transcriptome analyses. Reads were aligned to the reference genome NZ_CP006870.1 [55] using Rockhooper2 [59]. Gene transcript expression levels were calculated by RPKM [60]. Differential expression analysis was conducted using DESeq [61], then the DEGs picked out, such that p-value < 0.05 and difference of multiples >2. GO and KEGG enrichment analyses of DEGs were performed by hypergeometric distribution tests to determine the biological functions or pathways that are mainly affected by differential genes.

Real-Time Quantitative PCR (qRT-PCR) Verification
Total RNA was extracted from bacteria with RNAiso Plus (TaKaRa, Tokyo, Japan), according to the manufacturer's instructions, and quantified to 40 ng/µL with a Nanodrop 2000 (Thermo Fisher Scientific, Wilmington, MA, USA). Two microliters of diluted RNA were directly used for qRT-PCR with One Step SYBR ® PrimeScript ® Plus RT-PCR Kit (TaKaRa, Dalian, China). The primers were listed in Table 2. The expression levels of the tested genes were analyzed with rpoB as the reference gene. Fold-change was calculated using the 2 −∆∆Ct method [62].

Statistical Analysis
All experiments were repeated independently three times. The data were expressed as the mean ± SD. A one-way analysis of variance (ANOVA) was conducted to identify the significant differences, followed by Bonferroni's post-test (IBM SPSS Statistics, version 19.0, Armonk, NY, USA). A p-value of <0.05 or <0.01 was considered statistically significant.

Conclusions
Our study highlights that resveratrol at sub-MIC has an inhibitory effect on the biofilm formation and motility of A. hydrophila. Transcriptome analysis found that resveratrol significantly repressed bacterial chemotaxis and flagellar assembly pathways, disrupted type IV pilus synthesis, downregulated the c-di-GMP and LTTR levels, which all involved in QS systems. Thus, we concluded that resveratrol could decrease biofilm formation at concentrations without anti-A. hydrophila growth by inhibiting QS systems. Additionally, resveratrol also markedly suppressed the gene expression of several important virulence factors, such as OmpA, extracellular proteases, lipases, and T6SS. In conclusion, resveratrol could be considered a potential therapeutic drug by attenuating the capacity of pathogenic A. hydrophila to cause infection, and is unable to induce muti-drug resistance.