2025 年 60 巻 3 号 p. 130-137
The mass mortality episodes of the Pacific oyster Crassostrea gigas larvae, which occurred in a Japanese hatchery between 2013 and 2014, were investigated bacteriologically. A total of 77 bacterial isolates were obtained from dead larvae, seawater from rearing tanks, and other sources. Immersion infection challenge experiments revealed that 34 of the 77 isolates were highly virulent towards C. gigas larvae, with the LD50 of less than 105 CFU/mL. These virulent isolates were identified as Vibrio coralliiyticus (n = 28), V. tubiashii (n = 3), and unknown Vibrio species (n = 3), based on PCR and partial sequence analyses of three genes (16S rRNA, recA, rpoA). This indicates that both V. coralliilyticus and V. tubiashii, which have long been known worldwide as important pathogens of hatchery-reared bivalve mollusk larvae, are involved in mass mortality of C. gigas larvae cultured in the hatchery. Furthermore, some virulent isolates from diseased oyster larvae in the same hatchery in 1997, which were previously identified as V. splendidus biovar II using conventional phenotypic characterization tests, were re-identified as V. coralliilyticus. These results imply that V. coralliilyticus is a major pathogen of C. gigas larvae in the hatchery.
Several Vibrio species have been reported to be the cause of mass mortality in oyster hatcheries (Travers et al., 2015; Dubert et al., 2017). The intermittent high larval mortality caused by Vibrio spp. has resulted in shortages of oyster larvae and seeds needed by the commercial shellfish industry in the USA (Richards et al., 2021). V. tubiashii and V. coralliilyticus are among the pathogens that are particularly infectious to oyster larvae and other bivalves (Richards et al., 2021). V. tubiashii has been associated with mortality in shellfish hatcheries in the USA and other locations worldwide (Richards et al., 2015). It was first reported as a pathogen of oyster larvae in 1965 (Tubiash et al., 1965) and was later named V. tubiashii (Hada et al., 1984). V. coralliilyticus was first identified as a coral pathogen and has been associated with significant losses in coral reefs worldwide (Ben-Haim et al., 2003). V. coralliilyticus has also been reported to be a pathogen of the commercially important Eastern oyster Crassostrea virginica and Pacific oyster C. gigas larvae in hatcheries (Richards et al., 2015). Recent DNA sequencing has revealed that some isolates previously identified as V. tubiashii were in fact V. coralliilyticus (Richards et al., 2014), indicating that these two species have a close phylogenetic relationship (Ben-Haim et al., 2003; Richards et al., 2015). This misidentification has complicated efforts to understand the roles of these pathogens in larval shellfish mortality and their potential involvement in coral bleaching events (Richards et al., 2014).
The causes of mass mortality in triploid Pacific oyster larvae were investigated in the late 1990s at a hatchery in Hiroshima Prefecture, Japan (Sugumar et al., 1998a; Sugumar et al., 1998b). Conventional characterization tests revealed that the bacteria isolated from the diseased larvae were V. splendidus biovar II, which demonstrated high pathogenicity to larvae in experimental challenges (Sugumar et al., 1998a; Sugumar et al., 1998b).
In 2013 and 2014, the same hatchery experienced another series of mass mortality of Pacific oyster larvae. Several bacteria were isolated from both the affected larvae and the rearing seawater. Therefore, this study aimed to perform pathogenicity and identification tests on these bacterial isolates to determine whether they were responsible for the observed mass mortality events.
Bacteria were isolated from triploid larvae and seawater in a hatchery where mass mortality occurred in 2013 and 2014. For isolation of bacteria, dead or live larvae were homogenized in sterile seawater (9 times the weight of larvae) using a glass homogenizer, and the homogenates were further diluted tenfold with sterile seawater. Then, 0.1 mL of these diluted homogenates were added to Marine agar 2216 medium (MA, Difco) and TCBS agar (Difco), and cultured at 25°C. The dominant colonies were fished on MA after two days of culture. Bacteria from the rearing and injection seawater were isolated by applying seawater, diluted stepwise with sterile seawater, on each medium. The rearing seawater was sampled from the larvae tank, and the injection seawater was sampled from the sand-filtered seawater entering the tank. These isolates were suspended in marine broth 2216 medium (Difco) with 10% glycerin and stored at −80°C until analysis.
Pathogenic examinationsPathogenic examinations were conducted using 34 and 43 isolates from samples in 2013 and 2014, respectively. Bacteria were grown on seawater nutrient agar (SWNA: 0.5% Lab-Lemco powder (Oxoid), 1% Bacto peptone (BD), and 1% Bacto agar (BD) adjusted using seawater) at 25°C for 24 h and suspended in sterile seawater. The suspended bacteria were centrifuged at 10,000 ×g (1 min), and the bacterial pellet was resuspended in sterile seawater. The suspended bacteria were diluted in six steps (10-fold dilutions) using sterile seawater. The bacterial counts ranged from approximately 102 to 108 CFU/mL.
Diploid Pacific oyster larvae (7–10 days old larvae, mean shell height: 150.3–198.5 μm) produced at Fisheries and Marine Technology Center, Hiroshima Prefectural Technology Research Institute were used. Prior to the experiment, there was no mass mortality of unknown cause in the larvae. The larvae were placed into 24-well plastic tissue culture plates each using 1 mL of filtered (TCW-1N-PPS, Advantec) seawater at a density of 40–80 larvae/well, and 0.1 mL of six-step prepared bacterial suspensions were added. Wells without bacteria served as controls. These assays were performed in duplicates at 25°C, and larval survival was recorded after 24 h. Larvae were considered dead when they sank to the bottom of the well in a cloudy state without transparency or apparent movement. The LD50 (lethal dose) value (CFU/mL) of the isolates was calculated using the probit method. Isolates with an LD50 value of 105 CFU/mL or less were defined as virulent in this study.
Bacteriological and proteolytic examinationsMorphological, physiological, and biochemical characterization tests were conducted on the ten selected virulent isolates using standard methods. The cultivation-related tests were all assessed two days later at 25°C.
The proteolytic activities of all the isolates were assessed using 1% azocasein. Bacteria were cultured in seawater nutrient broth (SWNB: 0.5% Lab-Lemco powder and 1% Bacto peptone adjusted with seawater) at 25°C for 24 h. Following centrifugation at 12,000 ×g for 1 min (4°C), the resulting supernatant was utilized for analysis. This test was performed as described by Hasegawa et al. (2008) and Hasegawa et al. (2009).
Genetic analysisEleven virulent and ten non-virulent isolates were identified using the partial sequence of 16S rRNA (1,331 bases), recA (581 bases), and rpoA (452 bases). The genomic DNA of the isolates was extracted using the boiling method. One loop of bacterial pellet was suspended in 500 μL TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) and boiled for 10 min. After centrifugation at 12,000 ×g for 10 min (4°C), the supernatant was used for DNA analyses. The 16S rRNA fragment was amplified via PCR using the universal primer pair, 20F and 1500R (Weisburg et al., 1991). The fragment of recA was amplified using recA130F and recA720R primers (Sawabe, 2010), and that of rpoA was amplified using rpoA1F and rpoA500R primers (Sawabe, 2010). The amplified fragments were purified using the MinElute PCR purification kit (Qiagen) and sequenced using the BigDye Terminator Cycle Sequencing kit (v.3.1, Applied Biosystems) and the ABI 310 automated DNA sequencer (Applied Biosystems). The sequence data obtained from the isolates were compared with the sequence data in the BLAST program (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Phylogenetic analyses based on the concatenated sequences were performed using the neighbor-joining (NJ) method with DNAsis (Hitachi software).
In addition, all 77 isolates were subjected to a PCR analysis using the primers vcpAF and vcpAR (Wilson et al., 2013) to determine whether they were V. coralliilyticus.
Re-identification of previous bacterial isolatesIsolates No. 58, No. 59, and No. 60 (Sugumar et al., 1998a) were identified using the genetic analyses mentioned above. These bacteria were isolated from inactive oyster larvae in the same hatchery in 1997 and identified as V. splendidus biovar II using simple biochemical characterization tests (Sugumar et al., 1998a). Subsequently, their pathogenicity against Pacific oyster larvae was confirmed (Sugumar et al., 1998a). The isolates were preserved at Hiroshima University through freeze-drying method.
A total of 77 bacteria, 34 from samples in 2013 and 43 from samples in 2014, were isolated. These bacteria grew well on both MA and TCBS agars. Of the 34 isolates from 2013, 15 were virulent, with LD50 values of 105 CFU/mL or less, and of the 43 isolates from 2014, 19 were virulent. The mortality rates of larvae caused by the three representative virulent isolates and three non-virulent isolates are shown in Fig. 1. The LD50 values of the virulence isolates were 1.3 × 104 CFU/mL (HCG-14001), 2.0 × 104 CFU/mL (HCG-14034), and 2.5 × 104 CFU/mL (HCG-14052), whereas those of the non-virulent isolates were 1.0 × 107 CFU/mL (HCG-14002), 4.0 × 106 CFU/mL (HCG-14004), and 7.9 × 107 CFU/mL (HCG-14012).
Table 1 shows the 11 virulent and ten non-virulent bacterial isolates selected based on the sequencing analysis. The virulent isolates were estimated as V. coralliilyticus (five isolates), V. tubiashii (three isolates), and unknown species of the genus Vibrio (three isolates). The non-virulent isolates were estimated as V. rotiferianus (two isolates), V. chagasii (four isolates), V. fortis (two isolates), V. mediterranei (one isolate), and V. splendidus (one isolate). The closest bacterial species to the three unknown Vibrio isolates (HCG-13006, HCG-13038, and HCG-14011) based on BLAST analysis are shown in Table 2. The three isolates exhibited different closely related bacterial species depending on the gene type. HCG-13006 and HCG-14011 were closely related to V. tubiashii, V. coralliilyticus, and V. neptunius, while HCG-13038 showed close affiliation with V. neptunius and V. ponticus. The bacterial sequence data have been deposited in the DNA Data Bank of Japan (DDBJ) under the accession numbers LC663538–LC663547, LC663709–LC663718, and LC663796–LC663815.
| Isolation | Virulence (LD50: Log10 CFU/mL) | Proteolytic activity (A450/A620) | Identification of Vibrio coralliilyticus by PCR | Estimation of bacterial species based on 16S rRNA, recA, and rpoA sequences | ||
|---|---|---|---|---|---|---|
| Year | Isolate | Source | ||||
| 2013 | HCG-13011 | Dead larvae | 4.9 | 1.2 | + | V. coralliilyticus |
| HCG-13013 | Dead larvae | 4.1 | 2.5 | + | nt*1 | |
| HCG-13018 | Rearing water | 4.8 | 2.0 | + | nt | |
| HCG-13021 | Rearing water | 4.4 | 2.6 | + | nt | |
| HCG-13024 | Dead larvae | 4.5 | 2.4 | + | nt | |
| HCG-13026 | Rearing water | 4.6 | 2.2 | + | nt | |
| HCG-13027 | Dead larvae | 4.4 | 2.5 | + | nt | |
| HCG-13035 | Dead larvae | 4.9 | 1.4 | + | V. coralliilyticus | |
| HCG-13042 | Dead larvae | 4.6 | 2.7 | + | nt | |
| HCG-13045 | Dead larvae | 4.6 | 1.3 | + | nt | |
| HCG-13003 | Dead larvae | 4.7 | 1.1 | − | V. tubiashii | |
| HCG-13012 | Dead larvae | 4.4 | 0.9 | − | V. tubiashii | |
| HCG-13016 | Rearing water | 4.9 | 1.7 | − | V. tubiashii | |
| HCG-13006 | Dead larvae | 5.0 | 1.3 | − | Vibrio sp. | |
| HCG-13038 | Dead larvae | 4.0 | 1.3 | − | Vibrio sp. | |
| HCG-13002 | Dead larvae | 7.8 | 0.2 | − | V. rotiferianus | |
| HCG-13005 | Dead larvae | 6.4 | 0.9 | − | V. chagasii | |
| HCG-13010 | Dead larvae | 6.6 | 0.7 | − | V. fortis | |
| HCG-13023 | Live larvae | 7.0 | 0.2 | − | V. mediterranei | |
| HCG-13008 | Feed organism | 7.2 | 0.3 | − | nt | |
| HCG-13009 | Rearing water | 7.6 | 1.7 | − | nt | |
| HCG-13014 | Rearing water | 7.6 | 0.8 | − | nt | |
| HCG-13015 | Rearing water | 7.6 | 1.6 | − | nt | |
| HCG-13017 | Rearing water | 6.8 | 0.8 | − | nt | |
| HCG-13019 | Rearing water | 7.5 | 0.6 | − | nt | |
| HCG-13020 | Rearing water | 7.3 | 0.2 | − | nt | |
| HCG-13022 | Rearing water | 7.3 | 2.0 | − | nt | |
| HCG-13025 | Dead larvae | 7.2 | 2.3 | − | nt | |
| 2013 | HCG-13033 | Dead larvae | 7.5 | 2.2 | − | nt |
| HCG-13034 | Dead larvae | 7.3 | 2.1 | − | nt | |
| HCG-13036 | Dead larvae | 6.6 | 2.4 | − | nt | |
| HCG-13039 | Dead larvae | 7.2 | 0.4 | − | nt | |
| HCG-13040 | Dead larvae | 7.1 | 0.7 | − | nt | |
| HCG-13041 | Dead larvae | 6.7 | 1.1 | − | nt | |
| 2014 | HCG-14001 | Dead larvae | 4.1 | 1.4 | + | V. coralliilyticus |
| HCG-14003 | Dead larvae | 4.1 | 1.1 | + | nt | |
| HCG-14015 | Dead larvae | 4.0 | 1.1 | + | nt | |
| HCG-14019 | Rearing water | 4.4 | 2.4 | + | nt | |
| HCG-14023 | Rearing water | 4.2 | 2.0 | + | nt | |
| HCG-14024 | Live larvae | 4.6 | 2.9 | + | nt | |
| HCG-14025 | Dead larvae | 4.2 | 2.4 | + | nt | |
| HCG-14026 | Dead larvae | 4.5 | 1.8 | + | nt | |
| HCG-14027 | Rearing water | 4.8 | 1.8 | + | nt | |
| HCG-14029 | Rearing water | 4.7 | 1.6 | + | nt | |
| HCG-14030 | Rearing water | 4.3 | 1.6 | + | nt | |
| HCG-14031 | Dead larvae | 4.3 | 1.7 | + | nt | |
| HCG-14034 | Dead larvae | 4.3 | 1.7 | + | V. coralliilyticus | |
| HCG-14038 | Rearing water | 4.9 | 1.1 | + | nt | |
| HCG-14048 | Dead larvae | 4.9 | 1.6 | + | nt | |
| HCG-14050 | Dead larvae | 4.9 | 1.3 | + | nt | |
| HCG-14052 | Injection water | 4.4 | 1.2 | + | V. coralliilyticus | |
| HCG-14054 | Dead larvae | 4.5 | 1.4 | + | nt | |
| HCG-14011 | Live larvae | 4.2 | 1.4 | − | Vibrio sp. | |
| HCG-14002 | Dead larvae | 7.0 | 0.2 | − | V. rotiferianus | |
| HCG-14004 | Dead larvae | 6.6 | 1.1 | − | V. chagasii | |
| HCG-14005 | Dead larvae | 7.1 | 1.1 | − | V. chagasii | |
| HCG-14033 | Dead larvae | 7.2 | 1.0 | − | V. chagasii | |
| HCG-14010 | Live larvae | 7.3 | 1.2 | − | V. splendidus | |
| HCG-14012 | Live larvae | 7.9 | 0.9 | − | V. fortis | |
| HCG-14006 | Dead larvae | 7.6 | 0.7 | − | nt | |
| HCG-14013 | Injection water | 6.2 | 0.4 | − | nt | |
| HCG-14014 | Rearing water | 5.9 | 0.9 | − | nt | |
| HCG-14016 | Dead larvae | 7.6 | 1.5 | − | nt | |
| HCG-14017 | Dead larvae | 7.7 | 0.4 | − | nt | |
| HCG-14018 | Injection water | 7.2 | 0.6 | − | nt | |
| HCG-14020 | Live larvae | 7.8 | 0.9 | − | nt | |
| HCG-14021 | Injection water | 7.2 | 1.9 | − | nt | |
| HCG-14022 | Injection water | 6.9 | 0.7 | − | nt | |
| HCG-14028 | Rearing water | 7.7 | 0.8 | − | nt | |
| HCG-14032 | Dead larvae | 7.4 | 0.5 | − | nt | |
| HCG-14035 | Injection water | 6.6 | 0.4 | − | nt | |
| HCG-14036 | Injection water | 7.4 | 0.5 | − | nt | |
| HCG-14037 | Injection water | 6.6 | 0.5 | − | nt | |
| HCG-14039 | Live larvae | 7.0 | 0.2 | − | nt | |
| HCG-14047 | Injection water | 6.9 | 0.4 | − | nt | |
| HCG-14049 | Injection water | 6.6 | 0.6 | − | nt | |
| HCG-14051 | Injection water | 6.9 | 0.3 | − | nt | |
| 1997 | No.58*2 | Inactive larvae*2 | 4.6*2 | 0.8 | + | V. coralliilyticus |
| No.59*2 | Inactive larvae*2 | 4.5*2 | 0.4 | + | V. coralliilyticus | |
| No.60*2 | Inactive larvae*2 | < 4.0*2 | 0.3 | + | V. coralliilyticus | |
| Genes | Vibrio sp. | ||
|---|---|---|---|
| HCG-13006 | HCG-13038 | HCG-14011 | |
| 16S rRNA | V. tubiashii ATCC 19109 (CP009354.1*1) 1,323/1,328 (99%*2) | V. neptunius KCTC 12702 (CP079859.1) 1,328/1,328 (100%) | V. coralliilyticus ATCC BAA-450 (CP156656.1) 1,323/1,329 (99%) |
| recA | V. neptunius KCTC 12702 (CP079859.1) 597/607 (98%) | V. ponticus DSM 16217 (AP019657.1) 491/567 (87%) | V. neptunius KCTC 12702 (CP079859.1) 571/581 (98%) |
| rpoA | V. neptunius KCTC 12702 (CP079859.1) 461/463 (99%) | V. neptunius KCTC 12702 (CP079859.1) 450/452 (99%) | V. neptunius KCTC 12702 (CP079859.1) 449/452 (99%) |
The molecular phylogenetic tree constructed from the concatenated sequences of 16S rRNA, recA, and rpoA is shown in Fig. 2. The three isolates of unknown species within the genus Vibrio (HCG-13006, HCG-13038, and HCG-14011) were distinct from the other isolates of V. coralliilyticus and V. tubiashii.
Table 1 presents the 77 isolates identified through genetic analysis, including PCR testing. Of the 15 virulent isolates obtained in 2013, ten were identified as V. coralliilyticus. Similarly, of the 19 virulent isolates collected in 2014, 18 were identified as V. coralliilyticus. Of the 28 V. coralliilyticus isolates, 18 were obtained from larvae (dead or alive), nine from the seawater where the larvae were reared, and one from the injection seawater. The LD50 values of the larval-derived isolates ranged from 1.0 × 104 to 7.9 × 104 CFU/mL, those of the isolates from the rearing seawater ranged from 1.6 × 104 to 7.9 × 104 CFU/mL, and that of the isolate from the injection seawater was 2.5 × 104 CFU/mL.
Re-identification of previous bacterial isolatesThree isolates (No.58, No. 59, and No. 60) from the same hatchery, initially identified as V. splendidus biovar II using conventional characterization tests (Sugumar et al., 1998a), were subsequently identified as V. coralliilyticus through genetic analysis (Table 1). In the molecular phylogenetic tree, they were positioned within a cluster slightly distinct from V. coralliilyticus isolates from 2013 and 2014 (Fig. 2).
Characterizations of the isolatesThe results of the conventional characterization tests for the virulent isolates are shown in Table 3. The characteristics of the 28 V. coralliilyticus isolates were identical, except for acid production from sucrose in three isolates. The characteristics of the three V. tubiashii isolates were identical, except for acid production from galactose in one isolate. The characteristics of the V. coralliilyticus isolates from 2013 and 2014 were similar to those of V. tubiashii, except for acid production from cellobiose: V. tubiashii produced acid from cellobiose, whereas V. coralliilyticus did not. The two virulent Vibrio spp. isolates (HCG-13006 and HCG-14011) differed from V. coralliilyticus and V. tubiashii in their sugar utilization patterns. However, the characteristics of the Vibrio sp. isolate (HCG-13038) matched those of the V. coralliilyticus isolates from 2013 and 2014. Additionally, acid production from galactose and glycerol differed between V. coralliilyticus isolates from 2013 and 2014 and those from 1997.
| Characteristic | V. coralliilyticus 2013, 2014 (n = 28) | V. coralliilyticus 1997 (n = 3)*1 | V. tubiashii 2013 (n = 3) | Vibrio sp. | ||
|---|---|---|---|---|---|---|
| HCG-13006 | HCG-13038 | HCG-14011 | ||||
| Gram stain | – | –*1 | – | – | – | – |
| Form | Short rod | Short rod*1 | Short rod | Short rod | Short rod | Short rod |
| Motility | + | +*1 | + | + | + | + |
| Cytochrome oxidase | + | +*1 | + | + | + | + |
| OF test | F | F*1 | F | F | F | F |
| Vp test | – | nt*2 | – | – | – | – |
| Arginine decomposition | – | nt | – | – | – | – |
| Lysine decarboxylation | – | nt | – | – | – | – |
| Ornithine decarboxylation | – | nt | – | – | – | – |
| Growth on TCBS agar | + | +*1 | + | + | + | + |
| Growth in the presence of | ||||||
| NaCl 0% | – | nt | – | – | – | – |
| NaCl 1% | + | nt | + | + | + | + |
| NaCl 3% | + | nt | + | + | + | + |
| NaCl 6% | + | nt | + | – | + | + |
| NaCl 8% | – | nt | – | – | – | – |
| Acid production from | ||||||
| Glucose | + | +*3 | + | + | + | + |
| Sucrose | 25*4 | –*3 | + | + | + | + |
| Fructose | + | +*3 | + | – | + | – |
| Maltose | + | +*3 | + | – | + | – |
| Mannose | + | +*3 | + | – | + | – |
| Cellobiose | – | –*3 | + | – | – | – |
| Galactose | – | +*3 | 2*4 | – | – | – |
| Glycerol | – | +*3 | – | – | – | – |
| Arabinose | – | –*3 | – | – | – | – |
| Lactose | – | –*3 | – | – | – | – |
The proteolytic activities of each isolate are presented in Table 1, and the correlation between LD50 values and proteolytic activity is shown in Fig. 3. Regression analysis revealed a significant negative correlation between the LD50 values and proteolytic activity (P < 0.05), indicating that isolates with lower LD50 values (virulent isolates) tended to exhibit higher proteolytic activity, whereas isolates with higher LD50 values (non-virulent isolates) showed lower proteolytic activity (Fig. 3). However, the three isolates (No. 58, No. 59, and No. 60) of V. coralliilyticus from 1997 demonstrated low proteolytic activity (Table 1).
Seventy-seven bacteria, identified to belong to the genus Vibrio based on their growth on TCBS agar, were isolated from larvae and seawater samples. Pathogenicity tests were conducted on these bacterial isolates to investigate the cause of mass mortality in triploid Pacific oyster larvae. Previous studies (Sugumar et al., 1998b) have shown that both triploid and diploid larvae are susceptible to virulent strains of V. splendidus biovar II, with only minor differences in susceptibility. Therefore, healthy diploid larvae were selected for pathogenic examinations in this study. The results revealed that 15 of 34 isolates from 2013 and 19 of 43 isolates from 2014 were virulent. These findings strongly suggest the association of the virulent isolates to mass mortality of triploid Pacific oyster larvae in the hatchery during those years.
Sequence analysis was performed on a selected set of 21 isolates, consisting of 11 virulent and ten non-virulent isolates. In addition to 16S rRNA sequences, identification was based on the sequences of rpoA and recA, which are commonly used for Vibrio analysis (Thompson et al., 2005; Sawabe, 2010). Among the 11 virulent isolates, five were identified as V. coralliilyticus, three as V. tubiashii, and three as unidentified Vibrio species. These three unidentified Vibrio species may represent distinct strains or potentially novel species within the Vibrio genus. Their genetic proximity to V. coralliilyticus and V. tubiashii and their unique phenotypic traits underscore the complexity and diversity of Vibrio populations in hatchery environments. These findings advanced our knowledge of the microbial diversity in oyster hatcheries and highlighted the potential involvement of various Vibrio species in larval mortality. Further research on these unidentified Vibrio species may provide valuable insights into their pathogenicity and ecological roles in shellfish aquaculture systems.
All isolates collected in 2013 and 2014 were subjected to PCR analysis to identify V. coralliilyticus. Among the 15 virulent isolates from 2013, ten were confirmed as V. coralliilyticus. Similarly, 18 of 19 virulent isolates collected in 2014 were identified as V. coralliilyticus. These findings indicate that V. coralliilyticus was the primary cause of the mass mortality in both 2013 and 2014. Prior to these incidents, there were no reports of mass mortality of oyster larvae caused by V. coralliilyticus in Japan.
A total of 28 V. coralliilyticus isolates were obtained from larvae and seawater. One of these isolates was derived from the injected seawater, and its virulence was confirmed. This suggests that seawater serves as a source of infection. As seawater is pumped from offshore and filtered using sand, sterilization may be essential to prevent the introduction of V. coralliilyticus.
Three isolates that had been previously obtained from the same hatchery in 1997 and initially identified as V. splendidus biovar II (Sugumar et al., 1998a) were re-identified using sequencing. The results revealed that the three isolates were V. coralliilyticus. This finding indicates that a mass mortality event caused by V. coralliilyticus had occurred at the hatchery in the past even though it was not recognized at that time. Interestingly, these isolates from 1997 formed phylogenetic clusters distinct from those of the V. coralliilyticus isolates obtained in 2013 and 2014. Additionally, the 1997 isolates showed differences in the utilization of sucrose, galactose, and glycerol compared to the 2013 and 2014 isolates. These observations suggest that the 1997 isolates are not identical to 2013 and 2014 isolates, highlighting the potential genetic and phenotypic diversity within V. coralliilyticus populations.
The LD50 values of V. coralliilyticus strains for Pacific oyster larvae ranged from 1.2 × 104 to 4.0 × 104 CFU/mL, according to Richards et al. (2015). Similarly, a previous study by Sugumar et al. (1998a) found that the LD50 values of three V. coralliilyticus isolates ranged from < 9.3 × 103 to 4.2 × 104 CFU/mL. In our study, the LD50 values of V. coralliilyticus isolates ranged from 1.3 × 104 to 7.9 × 104 CFU/mL, which is consistent with earlier findings.
As demonstrated by Hasegawa et al. (2009), a correlation exists between the proteolytic activity and pathogenicity of Vibrio species. In our study, the proteolytic activity of the present isolates was high in virulent isolates and low in nonvirulent isolates, and this is consistent with previous findings (Hasegawa et al., 2009). However, the proteolytic activities of the three V. coralliilyticus isolates from 1997 (No. 58, No. 59, and No. 60) were significantly low. Apart from the differences in sugar utilization described earlier, these findings suggest that the characteristics of the 1997 isolates may differ from those of the 2013 and 2014 isolates, and possibly the 1997 isolates may have different virulence factors or pathogenic mechanisms. Various Vibrio species are known to produce metalloproteases that can kill oyster larvae (Hasegawa et al., 2009). Therefore, investigating the metalloprotease activity of these virulent isolates, particularly that of the unidentified Vibrio species found in this study, is essential in future studies.
This study revealed that V. coralliilyticus is the causative agent of mass mortality in Pacific oyster larvae, and similar events had occurred previously at the same hatchery. This suggests that previously unrecognized mortalities caused by V. coralliilyticus might have had long-term effects on oyster hatcheries in Japan. V. coralliilyticus demonstrated higher pathogenicity than V. tubiashii and may have posed a persistent challenge to Pacific oyster hatcheries for several years (Richards et al., 2015). Moreover, among the diverse array of Vibrio species, the most predominant pathogenic strain identified were those of V. coralliilyticus (Kehlet-Delgado et al., 2020). It is essential to continuously monitor the mass mortality events caused by V. coralliilyticus in oyster hatcheries in Japan to better understand and mitigate their impact.
We thank the staff of the Fisheries and Marine Technology Center, Hiroshima Prefectural Technology Research Institute, for their kind assistance with the experiments.
