Twenty diseased coral trout (0.5-year-old, 70.42±15.50 g in weight), which were reared in concrete tanks, were obtained from an infected fish farm in Wenchang, Hainan, China in March 2022, and immediately placed into ice-cold sterile bags and sent to the laboratory according to Hossain et al. (2011). The water quality indicators measured during the disease outbreak were 28 °C, pH 8.0, salinity of 24, 7.8 mg L^-1 dissolved oxygen, 0.20 mg L^-1 nitrite, and 0.68 mg L^-1 total ammonia. Healthy coral trout (31.04±5.17 g) were purchased from a fish farm in Changjiang, Hainan, China, and maintained good health with no contamination with L. anguillarum, P. damsela, V. harveyi pathogens by sampling and assessment according to Xu et al. (2008); Sun et al. (2015), and Carraro et al. (2018).

The potential pathogens were examined according to the previously described method (Wang et al., 2020). Firstly, the organs (liver, kidney, intestine, muscle, and gill) were dissected from the diseased coral trout in the laboratory, compressed manually between two glass slides to prepare the thin sections of organs, and then subjected to a observation for potential parasites using microscopy (Yang, 2018). Secondly, to determine whether this disease was caused by virus, bacteria-free organ filtrates were prepared according to Gong et al. (2010), and two replicate aquaria of healthy coral trout (10 fish per aquarium) were artificially infected with 0.2 mL of the freshly-prepared bacteria-free organ filtrates by intramuscular injection. Another two replicate aquaria of control coral trout (10 fish per aquarium) were treated intramuscularly with the same volume (0.2 mL) of sterile normal saline. All the test fish were stocked in aquaria containing 100 L natural aerated seawater at 28 °C and observed for fifteen days to record fish mortalities and any pathological signs. Thirdly, in order to determine if this disease was due to bacterial infection, liver samples were streaked onto thiosulfate citrate bile salt sucrose (TCBS) agar plate (Sinopharm Chemical Reagent Co., Ltd) for bacterial isolation according to Gu et al. (2015), and incubated at 28°C for 24h, then the uniform isolates were subjected to purification by repeated streaking onto nutrient agar (NA) plates amended with 15% NaCl. After the observation of the isolates’ colony and cell morphologies for purity assessment, the suspensions of pure isolates were further prepared by washing the inoculated NA plates after 24 h of incubation at 28 °C, and the colony forming units (CFU) in the suspensions were further estimated by calculating CFU on NA plates from a series of 10-fold dilutions in sterile normal saline. Afterwards, two replicate aquaria of healthy coral trout (10 fish per aquarium) were artificially infected by intramuscular injection with 0.2 mL of each pure isolate with 3.0 × 10⁷ CFU mL^-1 (Yao et al., 2015). Another two replicate aquaria of control coral trout (10 fish per aquarium) were treated with the same volume (0.2 mL) of sterile normal saline intramuscularly. All the test fish were stocked in aquaria containing 100 L natural aerated seawater at 28 °C and observed carefully for seven days to record fish mortalities and any pathological signs. The challenge isolate was re-isolated from freshly dead fish to confirm the cause of death, and histopathological changes in the liver of infected and healthy fish were also examined as described by Phrompanya et al. (2021).

The identification of the pathogenic isolate was performed using 16S rRNA gene sequencing analysis and biochemical tests (Liu et al., 2018). The total DNA was extracted from the pathogenic isolate by the TIANamp DNA Kit (Tiangen Biotech. Co., Ltd., Beijing, China) following the manufacturer’s instructions. Then, the amplification of the 16S rRNA gene was conducted as described by Liu et al. (2018) using the universal primer pair 27F (5’-AGAGTTTGATCCTGGCTCAG-3’) and 1492R (5’-GGTTACCTTGTTACGACTT-3’), and the amplified product was subjected to sequencing through the ABI 3730 XL DNA Sequencer (Applied Biosystems, Waltham, MA, USA). Finally, the 16S rRNA gene sequence of the pathogenic isolate was subjected to the Basic Local Alignment Search Tool (BLAST) to search the closest related sequences in GenBank, and the construction of a phylogenetic tree was carried out using the neighbor-joining (NJ) method. Besides, the pathogenic isolate was phenotypically identified using API 20E identification kit (Biomerieux, France) following the guidance of manufacturer (Xie et al., 2007). The phenotypic features of V. ponticus previously described by Kim et al. (2007); Liu et al. (2018) and You (2018) were used as controls.

The virulence of the pathogenic isolate was examined by testing the mean lethal dose (LD₅₀) in the coral trout according to Gu et al. (2015). Three replicate aquaria of healthy coral trout (10 fish per aquarium) were artificially infected by intramuscular injection with 0.2 mL of the pathogenic isolate with 2.1 × 10⁴, 2.1 × 10⁵, 2.1 × 10⁶, and 2.1 × 10⁷ CFU mL^-1. Another three replicate aquaria of control coral trout (10 fish per aquarium) were treated intramuscularly with the same volume (0.2 mL) of sterile normal saline. All the test fish were stocked in aquaria containing 100 L aerated seawater at 28 °C without water being changed and observed carefully for seven days to record fish mortalities and any pathological signs. The challenge isolate was re-isolated from freshly dead fish to confirm the cause of death. Based on the logarithms of challenge doses and cumulative mortalities of challenged fish, the medium lethal dose (LD₅₀) value was estimated using the Kochi method (Li et al., 2012) to evaluate the virulence of the pathogenic isolate.

The susceptibility of the pathogenic isolate to amoxicillin, ampicillin, azithromycin, bacitracin, cefotaxime, cefradine, ceftizoxime, cotrimoxazole, chloramphenicol, doxycycline, enoxacin, erythromycin, florfenicol, gentamycin, kanamycin, kitasamycin, netilmicin, nalidixic acid, novobiocin, oxacillin, penicillin, pipemidic acid, polymyxin B, rifampicin, roxithromycinum, and tobramycin was tested in triplicate by the Kirby-Bauer disk diffusion method (Joseph et al., 2011). Briefly, the pathogenic isolate was spread onto NA plates amended with 15% NaCl and antibiotic discs were then immediately placed on the inoculated plates. Afterward, the plates with antibiotic discs were incubated at 28 °C for 24 h to measure the diameters of inhibition zones surrounding the antibiotic discs. The susceptibility of the pathogenic isolate to twenty-six antimicrobials was evaluated by the instruction of the manufacturer (Hangzhou Binhe Microorganism Reagent Co., Ltd., Hangzhou, China). The antibiotic susceptibilities of V. ponticus previously reported by Kim et al. (2007); You (2018); Liu et al. (2018) and Kumari et al. (2020) were used as references.

No parasites were detected in the naturally diseased coral trout, and no mortality or visible disease signs were observed in all tested fish challenged with the bacteria-free organ filtrate (data not shown), revealing that the disease did not result from parasites or viruses. In addition, 5 different bacterial strains, temporarily named DX1, DX2, DX3, DX4, and DX5, were isolated from the liver of naturally-diseased coral trout, and no disease signs or mortalities were noted in the control or challenged fish with isolates DX1, DX3, DX4 and DX5. Only the test fish challenged with the most dominant isolate DX2 at 3.0 × 10⁷ CFU mL^-1 were found to exhibit a cumulative mortality of 100% (data not shown), indicating that isolate DX2 was pathogenic to the coral trout. The test fish challenged with isolate DX2 displayed the liver necrosis sign, similar to that noted in the naturally-infected fish ( Figure 1 ), and the same strain (DX2), confirmed by phenotypic and molecular identification, was re-isolated from the experimental diseased fish. Furthermore, the vacuolar degeneration of liver cells was observed in the artificially and naturally infected coral trout ( Figure 2 ). Yet no pathological symptoms or mortalities were observed in the control and treatment fish challenged with other bacterial isolates (data not shown). Thus, according to Koch’s postulates (Fredericks and Relman, 1996), isolate DX2 was identified as the causative pathogen of the coral trout.

A similarity of 99% to 100% was observed between the DX2 isolate (GenBank accession no. OP630658) and other V. ponticus strains in the GenBank database and further demonstrated as a V. ponticus strain through the phylogenetic tree ( Figure 3 ). In addition, the DX2 isolate possessed identical phenotypic features to V. ponticus strains reported previously (Kim et al., 2007; Liu et al., 2018; You, 2018) ( Table S1 ). It was positive for oxidase, β-galactosidase, lysine decarboxylase, and indole, and could utilize mannitol and sucrose, but was negative for arginine dihydrolase, ornithine decarboxylase, tryptophan deaminase, and Voges-Proskauer reaction, and could not utilize adonitol, amygdalin, arabinose, citrate, gelatine, glucose, inositol, melibiose, rhamnose, sodium thiosulfate, sorbitol, and urea. Thus, the DX2 isolate was identified molecularly and phenotypically as V. ponticus.

Cumulative mortalities of 6.7%, 33.3%, 60.0% and 100.0% was respectively reached in the coral trout challenged with the DX2 isolate at the cell densities of 2.1 × 10⁴, 2.1 × 10⁵, 2.1 × 10⁶, and 2.1 × 10⁷ CFU mL^-1 ( Figure 4 ) during the 7-day challenge, which showed the typical disease sign of liver necrosis. No mortality was observed in the control coral trout. Furthermore, the DX2 isolate was re-isolated from the experimental diseased fish, which was confirmed through phenotypic and molecular identification. These findings indicated that the LD₅₀ value for the DX2 isolate was 6.64×10⁵ CFU mL^-1 in the coral trout.

The DX2 isolate was highly susceptible to chloramphenicol, doxycycline, enoxacin, florfenicol, netilmicin, nalidixic acid, pipemidic acid, and polymyxin B, intermediately susceptible to gentamycin, kanamycin, novobiocin, and rifampicin, and showed multiple resistances to amoxicillin, ampicillin, azithromycin, bacitracin, cefotaxime, cefradine, ceftizoxime, cotrimoxazole, erythromycin, kitasamycin, oxacillin, penicillin, and roxithromycinum ( Table S2 ). These findings indicated that the phenicols and tetracyclines antimicrobials in aquaculture use such as doxycycline and florfenicol could be chosen for the control of isolate DX2.