Vibrio splendidus virulence to Apostichopus japonicus is mediated by hppD through glutamate metabolism and flagellum assembly

ABSTRACT Vibrio splendidus is the main opportunistic pathogen that causes skin ulcer syndrome in Apostichopus japonicus. hppDIn the present study, mutant V. splendidus with an in-frame deletion of hppDV.s. (MTVs) was constructed. The median lethal doses of wild-type V. splendidus (WTVs) and MTVs were 5.129 × 106 and 2.606 × 1010 CFU mL−1, respectively. RNA-Seq was performed using WTVs and MTVs cells at different growth stages to explore the mechanisms of the pathogenesis mediated by hppDV.s. Gene Ontology analysis showed that the expression levels of 105 genes involved in amino acid metabolism and protein binding were remarkably different between MTVs and WTVs. Kyoto Encyclopedia of Genes and Genomes analysis showed that the pathways of glutamate metabolism and flagellum assembly involved in biofilm formation and swarming motility were suppressed in MTVs. Correspondingly, the swarming motility, biofilm formation and colonisation of MTVs were remarkably decreased compared with those of WTVs. The results showed that 4-hppD catalyses tyrosine into fumarate, which could enhance glutamate metabolism and ATP production; promote flagellum assembly through the TCA cycle and lead to higher swarming, biofilm formation and colonisation abilities, to contribute to the pathogenesis of V. splendidus.


Introduction
The Vibrio genus causes serious infectious diseases that affect marine animals and humans [1,2]. The bacterium Vibrio splendidus naturally inhabits aquatic environments as the dominant vibrioplankton and is also abundant in and on many animal species [3]. Over the past few decades, different V. splendidus strains have been demonstrated to be associated with the mortality of cod larvae, turbot and shellfish [4,5]. Sea cucumber Apostichopus japonicus (Echinodermata, Holothuroidea) frequently suffer from "skin ulcer syndrome" (SUS) outbreak caused by bacterial infection. The SUSs affecting sea cucumbers probably correspond to different diseases depending on species or geographic location [ [6][7][8][9]. The main pathogens include Vibrio sp., Pseudomonas sp. and even spherical virus; among which, V. splendidus was considered the major pathogen that infects A. japonicus [10,11]. Compared with human pathogenic Vibrio species, little is known about Vibrio pathogenesis in aquatic animals. Despite its virulence factors and lethality have been described, few reports have focused on the pathogenic pathway of V. splendidus until now.
In the pathogenic process, the production of virulence factors results in metabolism and biological function disorders, which eventually destroy host cells and tissues. The virulence factors of Vibrio sp. contain adhesin factor, extracellular metalloprotease, hemolysin, siderophore, lipopolysaccharide and type III secretion system [12]. The main virulence factors of V. splendidus, including extracellular metalloprotease Vsm [13,14], quorum sensing [15], the siderophore receptors of LutA and FepA [16] and a hemolysin 4-hydroxyphenylpyruvate dioxygenase (4-hppD) encoded by the hppD gene of V. splendidus (hppDV.s.) [17], have been demonstrated to be related to the fulminant infectious diseases of sea cucumber.
The transformation catalyzed by 4-hppD has vital physiological importance and occurs in almost all aerobic organisms [18]. In aerobic metabolism, 4-hppD catalyses the tyrosine catabolism (that is the transformation of 4-hydroxyphenylpyruvate to homogentisate [HGA]). High HGA concentrations accumulate, selfpolymerise and autoxidise into pyomelanin via non-enzymatic reactions; this oxidation process can damage the host and result in a hemolytic phenotype [18][19][20][21]. Our previous study confirmed the dual functions of hppDV.s. for V. splendidus survival and cytotoxicity [22]. On the one hand, hppDV.s. promotes pyomelanin production and provides competitiveness for V. splendidus in unfavorable environments [22]. On the other hand, the purified recombinant hppDV.s. possesses cytotoxic effect and hemolytic activity [17]. Thus, pyomelanisation may benefit the survival and perhaps exacerbate the virulence of Vibrio sp. under this host environment. However, the regulation of global cellular metabolism and bacterial pathogenicity caused by 4-hppD is not known.
The infectivity and pathogenicity of Vibrio sp. are caused by the collective effect of various virulence factors. Recently, the microbial transcriptome sequencing (RNA-seq) technique has dramatically accelerated analytical capacity and has been widely applied to construct global networks of virulence factors [23]. In the present study, a mutant V. splendidus with an in-frame deletion of hppDV.s. (MTVs) was constructed. The growth curve, hemolytic activity, UV resistance and pathogenicity of wild-type V. splendidus (WTVs) and MTVs were analyzed and compared. RNA-seq was used to analyze the RNA from WTVs and MTVs cells collected at different growth stages to reveal the pathogenesis mechanism mediated by hppDV.s.

Bacterial strains, culture conditions and chemicals
Pathogenic V. splendidus strain was separated from the tissues of sea cucumbers with SUS from the indoor farms of Jinzhou Hatchery in May 2013. The strain's pathogenicity to A. japonicus was demonstrated by reinfection experiment [24]. The strain was identified by 16S rDNA sequencing analysis. Escherichia coli DH5α (Takara, China) and S17λpir were cultured in Luria-Bertani broth medium at 37°C with shaking at 200 rpm. V. splendidus was cultured in Zobell's 2216E medium (1 g yeast extract and 5 g tryptone in 1 L seawater, pH 7.6-8.0) at 28°C with shaking at 200 rpm. Ampicillin (Ap, 100 μg mL −1 ), chloramphenicol (Cm, 50 μg mL −1 ) and kanamycin (Kn, 50 μg mL −1 ) were added to the medium when needed. Taq DNA polymerase and pMD19-T vector were purchased from Takara (Beijing, China). Vectors pDM4 and pBT3 were preserved in our laboratory. Restriction endonucleases were purchased from New England Biolabs and used in accordance with the manufacturer's specifications.

MTVs construction and complementation
The plasmid preparation, the extraction of DNA fragments from agarose gels and the purification of PCR products were performed using the respective kits from Omega Bio-Tek (GA) according to the manufacturer's instructions. The primers used in the present research are listed in Table 1. The primers were designed according to the genomic DNA of V. atlanticus strain LGP32 with accession number FM954973.2. MTVs was constructed as previously described by Hu et al [25]. The 700 bp upstream fragment and 677 bp downstream fragment of hppDV.s. were amplified by polymerase chain reaction (PCR) using KohppDV.s.F1/KohppDV. s.R1 and KohppDV.s.F2/KohppDV.s.R2 and fused by overlap PCR. The 1376 bp fragments were inserted into the pMD19-T vector and then transformed into E. coli DH5α. Three positive clones were collected for sequencing at Genewiz Biotechnology (Suzhou, China). The overlap fragment was digested and inserted into the Bgl II site of suicide vector pDM4 containing the sacB gene to construct plasmid pDM4KohppDV.s.. The plasmid was transformed into S17λpir. The recombinant plasmid pDM4KohppDV.s. was transformed into WTVs by conjugation. Bacterial conjugation was performed as described previously [26]. The conjugants were placed on 2216E agar with Cm and Ap, and the positive conjugants were confirmed by PCR (KohppDV.s.F1/ KohppDV.s.R2) and sequencing. The second cross-over recombination was carried out in 2216E agar with 10% sucrose. The MTVs was verified by PCR (KohppDV.s. F1/KohppDV.s.R2) and sequencing. Plasmid pET28ahppDV.s. was generated to complement the MTVs with the hppDV.s. sequence. The DNA fragment of hppDV.s. was amplified by hppD C F/ hppD C R and ligated into pMD19-T for the convenience of subsequent enzyme digestion. Plasmid pBT3 with constitutive promoter P Trc was digested with BamH I and Xho I, and plasmid pBT3hppDV.s. was constructed by ligating hppDV.s. into pBT3 to reach a higher hppDV.s. expression efficiency. The DNA fragment named P Trc -fused hppDV.s., which contained the Trc promoter and the functional hppDV.s. gene, was obtained after pBT3hppDV.s. was digested using Swa I. Then, the DNA fragment of the P Trc -hppDV.s. fragment was inserted into the EcoR V site of pET28a, and the recombinant plasmid was transformed into S17λπ to construct S17λπ/pET28ahppDV.s. The recombinant plasmid pET28ahppDV.s. was transformed into MTVs by conjugation. The conjugants were placed on 2216E agar with Kn and Ap, and the positive conjugants were confirmed by PCR and sequencing.

Growth curve measurement
The WTVs and MTVs were placed in 2216E agar at 28°C overnight. Single colonies were inoculated into tubes with 10 mL of fresh 2216E medium and cultured at 28°C with shaking at 200 rpm. Overnight cultures were diluted to the same concentration, and 200 μL aliquots of WTVs and MTVs were transferred into flasks with 100 mL of fresh 2216E medium. The flasks were incubated at 28°C with shaking at 200 rpm. The optical density at 600 nm (OD 600 ) was measured at different time points. Three independent experiments were performed in this manner.

Hemolytic activity measurement
The hemolytic activity of WTVs and MTVs were measured as described by Liang et al [17]. The same concentrations of WTVs and MTVs cells were spread on 2216E solid medium complemented with 5% sterilised sheep blood and cultured at 28°C to qualitatively detect bacterial hemolytic activity. A hemolytic ring was observed after 24 h. To quantitatively detect hemolytic activity, 10 μL of the strains preserved with glycerol at a ratio of 1:1 (v/v) were inoculated into 10 mL of 2216E medium and incubated at 28°C with shaking at 200 rpm. The WTVs and MTVs cultured to an OD 600 of 1.0 were centrifuged at 8000 × g for 10 min to collect the supernatant, which was then sterile filtered through a 0.22 μm sterile filter membrane (Millipore, USA). Blood cell preparation was conducted as described by Lee et al [21]. Briefly, cells from sterile defidrinated sheep blood was washed in sterilized Dulbecco's modified Eagle medium (DMEM) for three times, and then resuspended in sterilized DMEM to a concentration of 5% (v/v). Equal volume (500 μL) of preheated sheep blood cell suspension was amended with 0, 10, 50 and 100 μL of WTVs and MTVs cell-free supernatants. The 2216E medium served as the control. The released protohemes after incubation at 28°C for 5 h were measured at OD 570 under a Spectramax 190. Three biological replicates were obtained from the WTVs and MTVs groups.

Anti-UV activity measurement
The WTVs and MTVs were cultured in 2216E medium at 28°C. The same concentration of bacteria was poured into glass petri plates, which were then exposed to UV light with shaking regularly. Bacterial culture was collected and diluted to approximately 10 4 CFU mL −1 at 0, 5, 10, 15 and 30 min. Then, the dilutions collected at different time points were spread onto the plates. The single colonies of viable WTVs and MTVs were counted after 24 h. The experiment was repeated at least three times.

Artificial infection
The A. japonicus infection experiments were conducted as described by Zhang et al [24]. Healthy sea cucumbers (weight, 3 ± 1 g) were purchased from Shandong Oriental Ocean Company (Yantai, China) and reared in aerated natural seawater (salinity, 28 psu) at 16°C for 3 days. Then, the sea cucumbers were randomly divided into 13 groups with 20 individuals in each group. WTVs and MTVs strains were cultured to an OD 600 of 1.0 in 2216E medium (28°C 24 h), centrifuged at 8000 × g, washed with PBS (28 °C) and resuspended with seawater. For artificial infection, weight-matched A. japonicus were continuously immersed with 5 × 10 7 , 1 × 10 7 , 5 × 10 6 , 1 × 10 6 , 5 × 10 5 and 1 × 10 5 CFU mL −1 V. splendidus (WTVs or MTVs). A. japonicus infected with PBS served as the negative control. The daily mortality of infected A. japonicus was recorded. Dead A. japonicus were picked out immediately, and the observed symptoms were recorded.

cDNA library preparation, Illumina sequencing and mapping of reads
For RNA-Seq, the WTVs and MTVs strains were cultured in 2216E agar plates for 24 h. Single colonies of both strains were inoculated into flasks with fresh 2216E medium and cultured at 28°C with shaking at 200 rpm. The WTVs and MTVs cells were collected when OD 600 reached 0.6, 1.0 and 1.5.
rRNA was removed using the Ribo-zero Kit to enrich the mRNA. Subsequently, mRNA was broken into short fragments using a fragmentation reagent. First-strand cDNA was synthesised using a random hexamer primer and using mRNA as a template. Buffers, DNA polymerase I, RNase H and dNTPs were added to synthesise the double-stranded cDNA. The resulting cDNAs were purified using AMPure XP beads to obtain the final library. Sequencing was conducted on the Illumina HiSeq/MiSeq sequencing platform at Novogene Biotech (Beijing, China). The raw data were submitted to the NCBI Sequence Read Archive (SRA) database.
The trimming and quality control of raw Illumina reads were conducted to obtain clean reads, which were mapped to the genome of V. atlanticus strain LGP32 (FM954973.2) using Bowtie2 [27].

Analysis of differential gene expression
The reads were aligned to the bacterial genome (NCBI FM954973.2) for V. splendidus transcriptome analysis. Transcript abundance quantification (in fragments per kilobase of transcript sequence per millions base pairs sequenced) was performed using RNA-seq by expectation maximisation with default parameters. Differential expression was determined using the following thresholds: q < 0.005 and |log2 fold change | ≥ 1.

Quantification of mRNA expression
The employed specific primers are listed in Table 1. Melting curve analysis and agarose gel electrophoresis were conducted to ensure the specificity of the PCR products. The total RNA of V. splendidus was collected from WTVs and MTVs cells with different OD 600 values. The primers used for the reference control of 16S rDNA were 933F and 16SRTR1. Real-time polymerase chain reaction (RT-PCR) was carried out in an ABI 7500 RT-PCR detection system (Applied Biosystems) with a total volume of 20 μL (8 μL  for 20 s, 72°C for 20 s. The 2 −ΔΔCT method was used to analyze the relative expression level of the candidate genes [28].

Biofilm assay
The biofilm assay for WTVs, MTVs and hppD C was conducted as described by Luo et al [29]. WTVs, MTVs and hppD C strains were grown overnight at 28°C in 2216E medium. A 2 μL of bacterial culture was added to a 96-well plate with 198 μL of 2216E medium per well and incubated at 28°C until OD 600 reached 1.0. Then, the plate was washed three times with sterile PBS, dyed with 200 μL of crystal violet (1%) for 30 min, then washed with sterile PBS and air-dried. The attached biofilm was dissolved with 200 μL of ethanol and then quantified at OD 570 . Five independent biological replicates were conducted.

Motility analysis
Swarming motility assay was set up as previously described by Wang et al [30]. WTVs, MTVs and hppD C strains were separately cultured overnight and diluted to 1:100 in fresh 2216E broth. Then, 2 µL of diluted bacterial cultures (OD 600 = 1.0) were dropped on 2216E plates containing 0.3% agar at 28°C for 12 h to observe the colony size. All the experiments were repeated three times.

Distribution of V. splendidus in A. japonicus tissues
The distribution of V. splendidus in the different A. japonicus tissues was determined as described previously [31,32]. Six A. japonicus individuals were equally divided into two tanks, and each tank was supplemented with 10 L of aerated natural seawater. A total of 0.1 L MTVs (OD 600 = 1.0) were centrifuged at 8000 × g for 10 min and used to infect A. japonicus. The control group was A. japonicus infected with WTVs (0.1 L, OD 600 = 1.0). After A. japonicus was subjected to immersion infection for 24 h, the tentacle, muscle and body wall of A. japonicus were collected using sterilised scissors and tweezers. Three specimens were collected for each A. japonicus tissue. The tissue samples were homogenised by a homogeniser and spread on 2216E medium supplemented with Ap (100 μg mL −1 ). No colony other than V. splendidus grew on the plate supplemented with Ap. Furthermore, three clones from each plate were identified by 16S rDNA sequencing analysis and confirmed to be V. splendidus.
Viable WTVs and MTVs cells were counted after incubation at 28°C for 24 h.

Statistical analyses
All data are expressed as the mean ± SD of least three sets of independent experiments. Statistical significance was determined by one-way ANOVA with Dunnett's test. P < 0.05 was considered statistically significant.

Effect of hppDV.s. on growth, hemolytic activity and anti-UV activity
In this study, hppDV.s. was knocked out by markerless in-frame deletion to construct MTVs (Supporting Information Fig. S1). The growth rate of the MTVs strain was almost the same as the growth of the WTVs strain (Figure 1a). WTVs showed obvious αhemolytic activity. MTVs also possessed α-hemolytic activity (Figure 1b). The results showed that the hemolytic activity of WTVs on solid sheep blood plate was similar to that of MTVs. The supernatants of WTVs and MTVs also showed similar hemolytic activities (data not shown). These results suggested that hppDV.s. knockout had no obvious impact on the hemolytic activity of V. splendidus.
WTVs and MTVs were cultured in 2216E medium at 28°C until OD 600 reached 0.6 and 1.5 to investigate their anti-UV activity. We observed that the color of WTVs was darker than that of MTVs in the later stages of bacterial growth at the same concentration (Figure 1c). The WTVs and MTVs cells were exposed to UV radiation. The cell counts before exposure to UV light (T = 0 min) were approximately 1.5 × 10 4 CFU mL −1 . The cell survival rate of WTVs after UV exposure was similar to that of MTVs at the OD 600 of 0.6 ( Figure 1d). The average number of surviving cells in WTVs were 10 4 , 530 and 0 CFU after exposure to UV for 5, 10 and 15 min, respectively, at the OD 600 of 1.5. However, the average number of surviving cells in MTVs were 706, 17 and 0 CFU at 5, 10 and 15 min, respectively, at the OD 600 of 1.5. The cell survival rate of WTVs after UV exposure was higher than that of MTVs at the OD 600 of 1.5 (Figure 1e). The results showed that hppDV.s. was related to melanin production, which could increase the survival rate and prolong the survival time of bacteria exposed to UV.

Global transcriptional analysis of hppDV.s. in V. splendidus
The raw data were submitted to the NCBI SRA database with accession number SUB10384008. Thirty-five mRNAs from the MTVs exhibited obvious difference compared with those from the WTVs, that is, 16 MTVs mRNAs were down-regulated and 19 MTVs mRNAs were up-regulated in the early log phase at the OD 600 of 0.6. In the mid log phase at the OD 600 of 1.0, 105 mRNAs from MTVs exhibited obvious difference compared with those from WTVs; 24 mRNAs were down-regulated and 81 mRNAs were upregulated. In the stationary phase at the OD 600 of 1.5, 26 mRNAs from MTVs exhibited obvious difference compared with those from WTVs; 23 mRNAs were down-regulated, and three mRNAs were upregulated ( Figure 3a). The cluster heat map of the differentially expressed genes (DEGs) shows that the expression profiles of genes at the OD 600 of 1.0 are clearly distinguishable (Supporting Information Fig. S2).
Twenty-four down-regulated DEGs were selected from the OD 600 = 1.0 group according to log2 fold change . The top five Gene Ontology (GO) terms were carboxylic acid metabolic process, organic acid metabolic process, oxoacid metabolic process, cellular amino acid metabolic process and alpha-amino acid metabolic process, respectively. The relationships between genes and GO terms are shown in Figure 3b. Bacterial-type flagellum organisation enriched four down-regulated genes, namely, flagellar hook assembly protein (flgD), flagellar hook-associated protein, flagellin and flagellar cap protein. According to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, all 24 downregulated DEGs were enriched in 16 KEGG pathways (Figure 3c). According to the P-value of Fisher's exact test, metabolic pathways (vsp01100) enriched the largest number of DEGs; the pathway of alanine, aspartate and glutamate metabolism (vsp00250) had the highest ratio; and the pathway of flagellar assembly (vsp02040) was remarkable. These results showed that genes related to metabolic pathways and flagellar assembly were altered in the MTVs strain. Four genes associated with the flagellar system were down-regulated in the MTVs compared with the WTVs; thus, hppDV.s. may regulate the expression of flagellar genes responsible for the swarming motility of V. splendidus.

Amino acid and protein metabolic process
In the early log phase at the OD 600 of 0.6, the GO terms that belong to catalytic activity, small molecule metabolic process and cofactor binding exhibited remarkable difference in MTVs compared with WTVs (Figure 4a). In the mid log phase at the OD 600 of 1.0, the transcripts involved in amino acid metabolism were substantially down-regulated in MTVs. These genes could encode enzymes involved in glutamate metabolism, which is critical for biofilm formation in numerous bacteria. In addition, ABC transporter-related genes were also considerably down-regulated in MTVs; this down-regulation affected the transportation of nutrients, such as amino acids and sugars, and ultimately affected cell metabolism (Figure 4b). Notably, the structural constituent of the ribosome was also affected by the deletion of hppDV.s. in the stationary phase at the OD 600 of 1.5. The genes that encode 50S ribosomal protein L27, 30S ribosomal protein S9, ribosomal subunit interface protein and 50S ribosomal protein L11 methyltransferase, were down-regulated in MTVs. This result suggests a regulatory mechanism for protein folding and modification process (Figure 4c). In summary, hppDV.s. may regulate bacterial virulence by changing the metabolic process of V. splendidus.

RT-PCR verification of DEGs
The genes selected for the RT-PCR verification of the DEGs at the OD 600 of 1.0 included flgD, ABC transporter permease (tpP), heat-shock protein (hsp20), ATPdependent chaperone (clpB), outer membrane lipoprotein (lolB) and hemagglutinin (hA). Melting curve analysis indicated that the RT-PCR product of each gene was unique. The results showed that the RT-PCR results of all genes had a consistent trend with the transcriptome results ( Figure 4d). Therefore, the transcriptome sequencing and assembly results had high stability and accuracy.

Effect on the biofilm formation and swarming motility of V. splendidus
The biofilms of WTVs, MTVs and hppD C that attached on 96-well plates were analyzed. The results demonstrated that the biofilm formation of MTVs on the surface of a 96-well plate was 81% and 65% of those in hppD C and WTVs, respectively (Figure 5a). The motility of WTVs, MTVs and hppD C strains were measured, and the results demonstrated an obvious difference in the swarming motility of WTVs, MTVs and hppD C (Figure 5(b,c)). The colony diameter of WTVs was 4-6 mm more than that of MTVs at each time point. The diameter of MTVs was 76%-83% of that of WTVs. The swarming motility of MTVs was obviously decreased compared with that of WTVs, and the motility of hppD-supplemented strains was obviously reverted. These results indicated that hppDV.s. positively regulated the swarming motility of V. splendidus.

Distribution of WTVs and MTVs in various A. japonicus tissues
The quantitative detection of cell adhesion in various tissues of A. japonicus after immersion infection with WTVs and MTVs is shown in Figure 5d. The adhesion quantities of WTVs to the body wall, tentacle and muscle tissues were 3.87 × 10 3 , 2.54 × 10 3 and 1.39 × 10 4 CFU g −1 , respectively. The adhesion quantities of MTVs to all tissues except the body wall obviously decreased. The adhesion quantities of MTVs to the body wall, tentacle and muscle tissues were 3.16 × 10 3 , 3.62 × 10 2 and 2.97 × 10 3 CFU g −1 , respectively.

Discussion
The toxicity of many pathogens depends on different virulence factors [33]. Many genes have been demonstrated to be involved in the virulence regulation of aquatic pathogenic bacteria by knockout technology [34][35][36][37]. However, no studies on the contribution of the knockout of the hppDV.s. gene to the pathogenesis of V. splendidus have been reported. The 4-hppD catalytic reaction is the first committed reaction of the tyrosine catabolism pathway in essentially all aerobic organisms [38]. Homologues of 4-hppD have been identified from the bacterial pathogens Streptomyces avermitilis [39,40], Vibrio vulnificus [41], V. splendidus [17], Pseudomonas fluorescens [42] and Legionella pneumophila [43,44]. In our previous studies, the dual functions of 4-hppD for V. splendidus cytotoxicity and survival was determined by in vitro experiments [22]. However, the roles of 4-hppD in the pathogenesis processes of V. splendidus on A. japonicus are rarely reported. Thus, we launched a series of experiments to determine how the hppDV.s. gene influences the virulence of V. splendidus.
Our previous research demonstrated that hppDV.s. contributes to the α-hemolytic activity of V. splendidus [17]. However, the present result showed that hppDV. s. knockout had no substantial effect on the hemolytic activity of V. splendidus. The result demonstrated that the hemolytic activity of hppDV.s. was overlapped by other hemolysins. The mechanism of the hemolytic activity of V. splendidus is complicated. Hemolysins from Vibrio sp. have many types [1]. We also found other kinds of hemolysins, such as thermolabile hemolysin and hemolysin III family protein [45], in the V. splendidus genome. This finding means that despite hppDV.s. was knockout, alternative pathways recompense the hemolytic activity of V. splendidus. Furthermore, hppDV.s. produces pyomelanin from tyrosine via HGA [22]. A study reported that the melanin production of mutant Bacillus thuringiensis is more resistive to UV radiation than that of the wild strain [46]. In the present study, we observed that the color of WTVs was darker than that of MTVs at the same concentration as time increased, and the survival rate of MTVs exposed to UV was significantly decreased compared with that of WTVs in the later growth stages. As depicted in Supporting Information, the expression of hppDV.s. reached the highest level at the OD 600 of 1.6 and reverted to the original level at the OD 600 of 2.0 (Fig. S3). Levin et al. also found that L. pneumophila secretes HGA after the bacteria stops replicating during stationary phase [47]. Their finding is consistent with our findings. The results suggest that hppDV.s. is expressed during the stationary phase and secrete HGA to participate in selfprotection.
The A. japonicus infected with the MTVs strain exhibited a substantial delay in death, a remarkable decrease in mortality and alleviated SUS symptoms compared with those infected with the WTVs strain. Furthermore, the LD 50 was reported for the first time to show the direct pathogenicity of V. splendidus to A. japonicus. These results suggested that the knockout of hppDV.s. led to the decrease in the virulence of V. splendidus. We collected WTVs and MTVs cells at the OD 600 of 0.6, 1.0 and 1.5 for bacterium RNA-seq to explore the changes in the bacterial virulence of MTVs. The OD 600 of 0.6 and 1.5 are the early log phase and stationary phase of V. splendidus, respectively. The mRNAs from MTVs exhibited less substantial difference compared with those from WTVs during these stages. However, in the mid log phase at the OD 600 of 1.0, 105 mRNAs from MTVs exhibited remarkable difference compared with those from WTVs; thus, these genes may be associated with the pathogenicity of V. splendidus.
hppDV.s. regulates cell metabolism through enzyme activity. HGA generates fumarate through a series of reactions, and fumarate can participate in the TCA cycle, which is the pivot of substance metabolism (Fig.  S4). α-Ketoglutarate is the carbon skeleton of glutamate and is derived from the TCA cycle. We found that the expression level of glutamate synthase remarkably decreased in MTVs according to the transcriptome data. A glutamate synthase mutant exhibited an early arrest of biofilm formation [48]. The bacterium manages to persist in the environment by associating with biofilms, which are microbial communities that attach to surfaces and encase themselves in a protective matrix [49,50]. In addition, previous research suggested that the high-density colonies of Legionella protect themselves by secreting HGA pulses, and biofilm formation is accompanied with pyomelanin production in deep-sea Pseudoalteromonas sp. SM9913 [47]. These results indicated that hppDV.s. is closely related to biofilm formation. However, no research has shown how the tyrosine metabolism mediated by hppDV.s. affects biofilm formation. The current research demonstrated that tyrosine metabolic disorder reduces a metabolic intermediate that prevents the system from regulating biofilm formation. These findings increase our understanding of the coordination between cellular metabolic status and the regulatory networks that govern biofilm formation. The results of GO and KEGG analyses showed that the GO term "bacterial-type flagellum organisation" and the flagellar assembly signaling pathway were considerably downregulated by hppDV.s. knockdown. We speculated that flagellar assembly cannot proceed normally because of the decreased TCA cycle flux, which causes the downregulation of flagellar biosynthesis genes [51]. A previous study found that flagellar systems are responsible for motility and are important for the infection and colonisation of Vibrio coralliilyticus and Vibrio parahaemolyticus [52,53]. Given the importance of the complex interaction relationship between biofilm formation and flagellar motility as depicted in Supporting Information Fig. S5, hppDV.s. knockdown may change the dispersion of V. splendidus. The results also demonstrated that the MTVs strain was less virulent and less abundant in A. japonicus with lower swarming motility.
HGA is the precursor to melanin production in bacteria and plastoquinone production in plants [54,55]. In addition, the emergence of acetyl-CoA and histone acetylation could influence pentose phosphate pathway (PPP) flux and cause a series of changes in mammal cells [56]. Our previous study demonstrated that hppDV.s. exhibits cytotoxicity to the coelomocyte of A. japonicus, and purified rhppDV.s. can cause oxidative stress in A. japonicus [22]. Excessive ROS levels can cause cell damage [57,58]. Interestingly, we also demonstrated that A. japonicus 4-hppD could downregulate ROS levels through the 4-hppD-glucose-6-phosphate dehydrogenase-PPP pathway within the A. japonicus was soaked in WTVs and MTVs (1.0 × 10 7 CFU mL −1 ) for 24 h for infection. The A. japonicus tissues were weighed and homogenised. The homogenised solution was diluted gradiently and then coated on 2216E plates. Colony count was determined on the next day. The error line represents the SD (n = 3). Asterisks represent the significance of difference (*P < 0.05, **P < 0.01).
normal range [59]. Moreover, the transcriptome of A. japonicus stimulated by WTVs and MTVs revealed that the p53 and NOD-like signaling pathways are involved in the resistance to the pathogenicity regulated by hppDV.s. (data not shown). Combined with the results of this article, we revealed the different functions of the hppD gene between hosts and pathogens (Fig. S6).
In conclusion, the deletion of the hppDV.s. gene in V. splendidus showed no effect on A. japonicus growth, but MTVs infection resulted delayed onset time, reduced mortality and alleviated SUS symptom in A. japonicus. RNA-seq indicated that glutamate metabolic pathways and flagellar assembly genes were altered in the MTVs strain. hppDV.s. regulates biofilm formation and flagellar assembly, which are responsible for the virulence of V. splendidus, by changing metabolic processes. These findings provide a theoretical basis for the research of disease prevention and control.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Data availability statement
The data that support the findings of this study are openly available in [GenBank SRA database] at https://www.ncbi. nlm.nih.gov/bioproject/?term=PRJNA763740, reference number [PRJNA763740]. Other data supporting the findings of this study are available within the article and its supplementary materials.