Abstract
Edwardsiellosis in the red seabream Pagrus major is one of the most significant bacterial infections in Japanese marine aquaculture. In this study, we investigated whether fish carcasses harboring Edwardsiella anguillarum (the causative agent of edwardsiellosis) serve as a source of infection using three experimental infection trials. In Experiment 1, healthy red seabream exhibited scavenging behavior on infected carcasses, resulting in increased E. anguillarum concentrations (viable counts and environmental DNA) in the rearing water and the bacterial detection in the intestines. This bacterium was subsequently isolated from the trunk kidney of fish in the same tank, and the fish developed edwardsiellosis. In Experiment 2, fish were infected with E. anguillarum through cohabitation with infected carcasses. Higher infection ratios were observed in the group with scavenging access to the carcasses than in the group without. In Experiment 3, oral administration of either an E. anguillarum suspension or a kidney homogenate from diseased fish successfully induced infection. In conclusion, infected fish carcasses may serve as a reservoir of E. anguillarum, with scavenging behavior playing a critical role in edwardsiellosis transmission in red seabream farms by contributing to both the release of the bacterium into surrounding water and the oral consumption of infected tissues.
Edwardsiellosis in the red seabream Pagrus major is one of the most significant bacterial infections in Japanese marine aquaculture. The disease was first identified in cultured red seabream in 1975 (Yasunaga, 1977; Yasunaga et al., 1982) and has caused considerable economic losses in red seabream farming. Affected fish often exhibit ulceration and/or hemorrhage at the base of the fins and on the head, along with the formation of white nodules in the kidney and spleen. These lesions are considered characteristic signs of edwardsiellosis (Yasunaga, 1977; Iida et al., 2016). The causative agent was initially identified as atypical Edwardsiella tarda (Yasunaga et al., 1982). However, Shao et al. (2015) reclassified the bacterium isolated from red seabream as a novel species, E. anguillarum, based on its bacteriological and phylogenetic characteristics, similarly to the reclassification of E. tarda strains isolated from Japanese eel Anguilla japonica in China. Although red seabream strains are non-motile and differ from the definition of E. anguillarum proposed by Shao et al. (2015), recent reports of edwardsiellosis in red seabream have referred to the causative agent as E. anguillarum (Endo et al., 2022; Sugiura et al., 2022). Therefore, we designated the causative agent of the disease as E. anguillarum in this study. As no effective countermeasures, such as vaccines or disease-resistant fish strains, have been developed yet, identifying the source of E. anguillarum infection and interrupting its transmission cycle are crucial steps toward the effective prevention and control of edwardsiellosis in red seabream aquaculture.
We focused on infected fish carcasses as a major source of E. anguillarum. Previous studies on other Edwardsiella species demonstrated that healthy fish can be infected by bacteria shed from infected carcasses (Klesius, 1994; Matsuoka, 2004). Matsuoka (2004) also reported that E. piscicida released from dead Japanese flounder Paralichthys olivaceus exhibited higher virulence than bacteria cultured on agar medium. Furthermore, because red seabreams are carnivorous, scavenging behavior has frequently been observed under both experimental and natural conditions (e.g., Tsukamoto et al., 1989; Nguyen et al., 2008). These findings led us to hypothesize that red seabream succumbing to edwardsiellosis may act as incubators for highly virulent E. anguillarum and that healthy individuals can become infected by scavenging these infected carcasses.
In this study, we investigated whether fish carcasses infected with E. anguillarum serve as a source of infection in three experimental infection trials (Experiments 1–3). In Experiment 1, we analyzed the temporal dynamics of E. anguillarum infection intensity in fish cohabiting with the infected individuals. In Experiment 2, we assessed the effects of scavenging behavior on edwardsiellosis transmission. In Experiment 3, we evaluated the potential for transmission of E. anguillarum through the oral consumption of tissues from infected carcasses.
Materials and Methods
Fish and bacteria
The red seabream used for experimental infections was bred and maintained in-house at the Ehime Fisheries Research Center. The fish were reared in 1000-L tanks with UV-irradiated running seawater at the Ehime Fisheries Research Center (Experiment 1) or South Ehime Fisheries Research Center, Ehime University (Experiments 2 and 3). They were fed daily with a commercial compound feed. The bacterial strain E. anguillarum E15-21, isolated from diseased red seabream (identified by Endo et al., 2022), was used as the model strain. The stock culture was preserved in Tryptic Soy Broth (TSB; Becton, Dickinson and Company) containing 10% glycerol at –80°C. For experimental use, the bacterium was first grown on Trypto-Soya Agar (TSA; Shimadzu Diagnostics) plates at 28°C for 24 h, then inoculated into 10 or 500 mL of TSB and incubated at 28°C with shaking for 12–14 h. The resulting culture was used as the E. anguillarum suspension for experimental infection. The concentrations of E. anguillarum were adjusted using phosphate-buffered saline (PBS) for intraperitoneal injection and oral administration. In contrast, red seabream rearing water was used for immersion. Unless otherwise stated, the colony-forming unit (CFU) counts were determined using TSA plates.
Cohabitation experiments between E. anguillarum-infected and non-infected fish (Experiment 1)
Experiment 1 was conducted twice (Trials 1 and 2) at the Ehime Fisheries Research Center. A total of 90 fish in each trial (mean body weight ± SD: Trial 1, 70.64 ± 11.97 g; Trial 2, 46.84 ± 6.10 g) were acclimated to experimental conditions in a 1000-L tank at 28°C for 1 wk. The fish were divided into an infection and a control group (n = 45 per group). In each group, 30 fish (designated as “non-injected fish”) were transferred to a 250-L tank with running seawater (Trial 1, 1.69 ± 0.60 L/min; Trial 2, 1.78 ± 0.66 L/min) maintained at 28°C. The remaining 15 fish (“injected fish”) were marked by clipping a portion of the pelvic fin and intraperitoneally injected with 100 μL of either E. anguillarum suspension (infection group) or PBS (control group). Fish in the infection group received E. anguillarum at a dose of 5.70 × 105 CFU/fish (= 5.76 log10 CFU/fish) in Trial 1 and 2.36 × 105 CFU/fish (= 5.37 log10 CFU/fish) in Trial 2. The injected fish were then cohabited with non-injected fish in the same tank.
Mortality of both injected and non-injected fish was monitored daily. Compound feed was provided ad libitum once daily. Before feeding, dead non-injected fish were promptly collected (dead injected fish were left in the tank), and 4–6 surviving non-injected fish were sampled at 1, 2, 3, and 5 days post-challenge (dpc), as well as at 14 dpc in Trial 1 and 9 dpc in Trial 2. Surviving injected fish were also sampled at the end of each trial. All sampled fish were anesthetized with 0.1% FA100 (Bussan Animal Health) and euthanized via spinal cord transection. The presence or absence of external symptoms, such as ulcerative and/or hemorrhagic lesions at the base of the fins (Fig. 1A and B) and on the head (Fig. 1C), was recorded. The fish were then aseptically dissected, and the trunk kidney was subjected to gross examination, focusing on white nodules. The kidneys were classified as type I (white nodules only, Fig. 1D) and type II (white nodules with overall kidney enlargement, Fig. 1E).
Fig. 1. Ulcerative and/or hemorrhagic lesions at the base of fins (A and B) and on the head (C), and trunk kidney with white nodules (D and E; arrowheads) in red seabream infected with E. anguillarum observed in this study. The kidneys were classified as type I (white nodules only; D) and type II (white nodules with overall kidney enlargement; E). A kidney without white nodules was regarded as asymptomatic (F).
Samples of the intestine (approximately 1 cm from the anus) and trunk kidney tissue were collected from each fish, weighed, and homogenized in 500 μL of PBS using 5 mm stainless steel beads and a TissueLyser II (QIAGEN K.K.) at a speed of 25 Hz for 1 min. Serial tenfold dilutions (10−1 to 10−5) of the homogenate were prepared in PBS, and 100 μL of each dilution was plated in duplicate on Salmonella-Shigella Agar (SSA; Shimadzu Diagnostics) plates. Black pigmented colonies were counted as viable E. anguillarum after incubation at 28°C for 72 h. Several colonies were identified as E. anguillarum using the species-specific PCR method described by Griffin et al. (2014). The number of viable E. anguillarum in each tissue sample (CFU/g tissue) was calculated based on the average colony count from duplicate plates and tissue weight. Additionally, 1-L water samples were collected from each group in 1-L plastic bottles prior to fish sampling on all sampling days to estimate the concentration of E. anguillarum in the rearing water (see the section “Estimation of E. anguillarum concentration in rearing water” below).
E. anguillarum-transmission experiments with infected fish carcasses (Experiment 2)
Experiment 2 was conducted three times (Trials 1, 2, and 3) at the South Ehime Fisheries Research Center of Ehime University. A total of 60 fish (Trials 1 and 2) or 66 fish (Trial 3) were used, with mean body weights ± SD as follows: Trial 1, 53.55 ± 13.58 g; Trial 2, 16.59 ± 2.99 g; Trial 3, 40.53 ± 6.21 g. Fish were divided into three groups (scavenging, non-scavenging, and control), each consisting of 20 (Trials 1 and 2) or 22 (Trial 3) fish. Each group was transferred to a 100-L tank with running seawater and acclimated to experimental conditions for 1 wk at 28°C (Trials 1 and 3) or 26°C (Trial 2). An additional 30 fish were intraperitoneally injected with E. anguillarum suspension and transferred to a separate 100-L tank at the corresponding temperature. The infection dose was 5.90 × 105 CFU/fish (= 5.77 log10 CFU/fish) in Trial 1, 2.79 × 106 CFU/fish (= 6.45 log10 CFU/fish) in Trial 2, and 2.40 × 106 CFU/fish (= 6.38 log10 CFU/fish) in Trial 3. In Trial 1, 20 surviving fish were euthanized at 5 dpc using cold narcosis on ice and collected as the infection source. In Trials 2 and 3, 20 and 22 fish that died at 2 dpc were used, respectively. Of these, half were directly introduced into the scavenging group, while the remaining fish were enclosed in polyester mesh bags (29 × 19 cm; 1-mm mesh opening) to prevent scavenging and were introduced into the non-scavenging group. An equal number of non-infected fish were euthanized and introduced in the control group. All fish used as the source of infection in Trial 1 were dissected and examined prior to use. All exhibited white nodules in the trunk kidney, from which E. anguillarum was isolated.
Mortality in each group was monitored for 28 dpc in Trial 1 and 10 dpc in Trials 2 and 3. The seawater flow rates during Trials 1, 2, and 3 were 1.6 ± 0.03 L/min, 1.71 ± 0.05 L/min, and 1.08 ± 0.06 L/min, respectively. Compound feed was provided ad libitum once daily. Before feeding, dead fish (excluding those used as the infection source) were promptly collected, and 5–15 surviving fish were sampled at 7 and 28 dpc in Trial 1 and at 5 and 10 dpc in Trials 2 and 3. The presence of external symptoms and white nodules in the trunk kidney was examined as described in Experiment 1. E. anguillarum was isolated from the trunk kidney by streaking onto SSA plates, followed by incubation at 28°C for 72 h. Additionally, 1-L water samples were collected from each group at 1, 2, 3, 4, 5, 7, 9, 15, 20, and 25 dpc (the last three time points were specific to Trial 1).
In Trial 1, all fish in the non-scavenging group died at 10 dpc due to aquarium system failure. Observations for this group were terminated at that point, and dead fish were immediately examined and treated as fish that survived up to 10 dpc.
Experimental oral infection of E. anguillarum (Experiment 3)
Experiment 3 was conducted twice (Trials 1 and 2) at the South Ehime Fisheries Research Center of Ehime University. A total of 80 (26.88 ± 4.88 g; Trial 1) or 120 (30.01 ± 5.72 g; Trial 2) fish were divided into four groups (Groups A, B, and C and the control group), with each group consisting of 20 (Trial 1) or 30 (Trial 2) fish. Each group was transferred to a 100-L tank with running seawater, and acclimated to experimental conditions at 28°C for 1 wk.
E. anguillarum suspensions, diseased-kidney homogenates, and E. anguillarum-containing feed (Trial 1) were prepared for oral administration or immersion. To prepare the diseased-kidney homogenate, five fish were intraperitoneally injected with 100 μL of E. anguillarum suspension and transferred to a 100-L tank at 28°C. The infection dose was 1.32 × 107 CFU/fish (= 7.12 log10 CFU/fish) in Trial 1, 2.25 × 106 CFU/fish (= 6.35 log10 CFU/fish) in Trial 2. After two days, three dead fish were collected, and their trunk kidneys were harvested. Each kidney was homogenized in 1 mL of PBS as described in Experiment 1, and the resulting homogenates were pooled into a single 5-mL tube (total volume approximately 3 mL). The E. anguillarum-containing feed was prepared by immersing 21 g of compound feed in 9 mL of E. anguillarum suspension (3.20 × 108 CFU/mL = 8.51 log10 CFU/mL). An aliquot of each material, including E. anguillarum-containing feed homogenized in PBS, was plated in triplicate on SSA plates and the number of viable E. anguillarum was calculated based on colony counts.
After anesthetizing the fish using 0.1% FA100, the E. anguillarum suspension and diseased-kidney homogenate were orally administered to fish in Groups A and B, respectively, using 1-mL syringe fitted with silicon tube (100 μL/fish). In Trial 1, the administered doses were 1.60 × 107 CFU/fish (= 7.20 log10 CFU/fish) for the suspension and 1.34 × 106 CFU/fish (= 6.13 log10 CFU/fish) for the homogenate. In Trial 2, the doses were 6.00 × 107 CFU/fish (= 7.78 log10 CFU/fish) and 4.60 × 107 CFU/fish (= 7.66 log10 CFU/fish), respectively. In Trial 1, fish in Group C were fed E. anguillarum-containing feed (offered ad libitum), and feed consumption was calculated by measuring the difference in feed weight before and after feeding. The total dose was 5.96 × 107 CFU/20 fish (2.98 × 106 CFU/fish = 6.47 log10 CFU/fish). In Trial 2, fish in Group C were immersed for 1 h in 20 L of E. anguillarum suspension diluted with rearing water (6.00 × 104 CFU/mL = 4.78 log10 CFU/mL) in a plastic bucket under aerated conditions, and then rinsed once in another plastic bucket containing 20 L of fresh seawater. Fish in the control groups of both trials were orally administered 100 μL of PBS. After treatment, each group was transferred back to its respective 100-L tank. Mortality was monitored for 7 dpc. Seawater flow rates were 1.68 ± 0.06 L/min in Trial 1 and 1.05 ± 0.06 L/min in Trial 2. Compound feed was provided ad libitum once daily. Before feeding, dead fish were promptly collected, and 5–15 surviving fish were sampled at 1, 4, and 7 dpc and examined as described in Experiment 2. In addition, 1-L water samples were collected from each group at 5 min post-challenge (to investigate potential regurgitation of administered materials) and 1, 4, and 7 dpc.
Estimation of E. anguillarum concentration in rearing water
Rearing water collected in each experiment was plated in triplicate on SSA plates (100 μL/plate), and black pigmented colonies were counted after incubation at 28°C for 72 h. From the remaining water, 500 mL was vacuum-filtered through a cellulose mixed-ester membrane filter (47 mm diameter, 0.45 μm pore size; ADVANTEC, Toyo Roshi Kaisha). The filters were used for DNA extraction and quantitative PCR (qPCR) analysis of E. anguillarum environmental DNA (eDNA) using primers and a probe as described by Reichley et al. (2015) (see Takeuchi et al., 2024 for detailed methodology). Standard curves were generated using artificially synthesized target DNA (2 × 101–2 × 104 copies per reaction). The coefficient of determination (R2) values exceeded 0.99, and the PCR efficiencies ranged from 90.30 to 94.05%. The viable counts (CFU/mL) and eDNA concentrations (copies/mL) of E. anguillarum in the rearing water were calculated from the average colony counts on triplicate SSA plates and qPCR-quantified eDNA copy numbers, respectively.
Distilled water (500 mL) was filtered in parallel with each day’s filtration as a filtration negative control to assess potential contamination. These filters were subjected to qPCR using the same protocol as that used for the rearing water samples. Ultrapure water was used as the PCR negative control. No E. anguillarum DNA was detected in any of the filtration- or PCR-negative controls. Furthermore, prior to the introduction of fish (0 dpc), we confirmed the absence of viable colonies and eDNA of E. anguillarum in the seawater of all aquaria used in the experiments.
Statistical analysis
The infection ratios in Experiments 2 and 3 were calculated as the proportion of E. anguillarum-positive fish (including dead individuals) in each group across the sampling periods. Statistical differences between groups were assessed using Fisher’s exact test in R (version 4.4.3; https://www.r-project.org/). Statistical significance was set at p < 0.05.
Results
Experiment 1
No mortality was observed in the control groups of either Trial 1 or Trial 2, and E. anguillarum was not isolated from any control fish. The results of the infection groups are presented in Table 1. The mortality of the injected fish began at 1 dpc, with 10 injected fish (66.7%) dying within 14 dpc in Trial 1, whereas all injected fish in Trial 2 died within 2 dpc. The scavenging behavior by non-injected fish on dead injected fish was first observed at 2 dpc in both trials. Additionally, one dead non-injected fish in Trial 1 (3 dpc) and two in Trial 2 (7 and 8 dpc) exhibiting type II trunk kidneys, and E. anguillarum was isolated from all of these individuals.
Table 1. Summary of results for the infection groups from cohabitation experiments between
E. anguillarum-injected and non-injected fish (Experiment 1)
| Trial | dpc*1 | Cumulative
no. of dead-
injected fish
(n = 15) | No. of
dead/surviving
non-injected
fish | E. anguillarum concentration
in rearing water | Examination results in non-injected fish |
|---|
No. of fish
sampled | No. of fish showing clinical signs | No. of fish in which
E. anguillarum
was isolated from | Viable counts of E. anguillarum in
(log10 CFU ± SD/g tissue) |
|---|
Viable counts
(log10 CFU/mL) | eDNA conc.
(log10 copies/mL) | Gross trunk kidney | External*5
symptoms | Intestine | Trunk
kidney | Intestine | Trunk
kidney |
|---|
| Type I | Type II | Total |
|---|
| Trial 1 | 1 | 1 (6.7) | 0/30 | ND*3 | 0.09 | 6 | 0 | 0 | 0 | 0 | 0 | 0 | | |
| 2 | 4 (26.7) | 0/24 | 1.20 | 1.92 | 6 | 0 | 0 | 0 | 0 | 3 (50.0)*6 | 0 | 4.13 ± 0.63 | |
| 3 | 5 (33.3) | 1/17*2 | 3.03 | 4.04 | 6 | 0 | 0 | 0 | 0 | 3 (50.0) | 1 (16.7) | 4.99 ± 0.98 | 3.76 |
| 5 | 6 (40.0) | 0/11 | 2.75 | 3.69 | 6 | 3 (50.0)*4 | 0 | 3 (50.0) | 0 | 3 (50.0) | 3 (50.0) | 5.30 ± 1.73 | 5.32 ± 1.34 |
| 14 | 10 (66.7) | 0/5 | 2.45 | 3.19 | 5 | 1 (20.0) | 4 (80.0) | 5 (100) | 3 (60.0) | 4 (80.0) | 5 (100) | 5.05 ± 1.52 | 7.65 ± 0.26 |
| Trial 2 | 1 | 0 | 0/30 | 0.33 | 0.47 | 6 | 0 | 0 | 0 | 0 | 1 (16.7) | 0 | 2.26 | |
| 2 | 15 (100) | 0/24 | 0.83 | 2.68 | 6 | 0 | 0 | 0 | 0 | 2 (33.3) | 0 | 2.77 ± 1.40 | |
| 3 | | 0/18 | 4.16 | 5.02 | 6 | 0 | 0 | 0 | 0 | 6 (100) | 2 (33.3) | 4.22 ± 1.80 | 3.19 ± 1.22 |
| 5 | | 0/12 | 2.45 | 3.49 | 6 | 3 (50.0) | 3 (50.0) | 6 (100) | 1 (16.7) | 2 (33.3) | 6 (100) | 4.18 ± 0.56 | 6.91 ± 1.85 |
| 7 | | 1/5*2 | | | | | | | | | | | |
| 8 | | 1/4*2 | | | | | | | | | | | |
| 9 | | 0/4 | ND | 1.63 | 4 | 2 (50.0) | 2 (50.0) | 4 (100) | 1 (25.0) | 3 (75.0) | 4 (100) | 5.33 ± 0.51 | 9.29 ± 0.24 |
Low levels of E. anguillarum eDNA (−1.05 – −0.30 log10 copies/mL) were detcted in pre-infection seawater (0 dpc) and sporadically thereafter in control water samples. No mortality or E. anguillarum-infected fish were observed in control groups.
Shaded areas highlight the day on which scavenging behavior on the carcasses was first observed.
*
2 Dead non-injected fish showed gross type II trunk kidneys, and
E. anguillarum were isolated from the kidney.
*
4 The ratio of fish showing the clinical signs.
*
5 All fish with external symptoms (ulcerative and/or hemorrhagic lesions) had type I or II kidneys.
*
6 Isolation rate of
E. anguillarum.
eDNA of E. anguillarum was detected in seawater from both control and infection groups at 0 dpc (−0.60 – −0.30 log10 copies/mL), as well as in rearing water from the control groups at 2 and 14 dpc in Trial 1, and at 2 and 3 dpc in Trial 2 (−1.05 – −0.44 log10 copies/mL). This could be explained by the introduction of seawater contaminated with E. anguillarum from a nearby red seabream farm where edwardsiellosis has been found (see Takeuchi et al., 2024 for details). However, the concentrations of suspected contaminating E. anguillarum DNA were markedly lower than those in water samples from the infection groups. Moreover, the introduced seawater was UV irradiated, and no viable E. anguillarum was isolated from these samples. Therefore, we concluded that introducing E. anguillarum-contaminated seawater did not affect the results.
Viable E. anguillarum was isolated from rearing water in the infection group from 2 to 14 dpc in Trial 1 (1.20–3.03 log10 CFU/mL) and from 1 to 5 dpc in Trial 2 (0.33–4.16 log10 CFU/mL). eDNA was detected in the infection groups on all days after 1 dpc in both trials (Trial 1, 0.09–4.04 log10 copies/mL; Trial 2, 0.47–5.02 log10 copies/mL). The viable count and eDNA concentration in the rearing water peaked at 3 dpc in both trials. Among the surviving non-injected fish, individuals exhibiting characteristic signs of edwardsiellosis were observed at 5 dpc in both trials. In Trial 1, half of the examined fish had white nodules in the trunk kidney at 5 dpc. By 14 dpc, all examined fish had either type I or type II kidneys, with type II kidneys being predominant (80.0%). In Trial 2, half of the fish examined at 5 and 9 dpc displayed type I kidneys, whereas the other half exhibited type II kidneys. Fish with external symptoms were observed at 14 dpc in Trial 1 (60.0%) and at 5 dpc (16.7%) and 9 dpc (25.0%) in Trial 2, all of which also exhibited either type I or type II kidneys.
Viable E. anguillarum was first isolated from the intestines of non-injected fish at 2 dpc in Trial 1 (50.0%) and at 1 dpc in Trial 2 (16.7%). Viable bacteria in the trunk of the kidney were first detected at 3 dpc in both trials (Trial 1, 16.7%; Trial 2, 33.3%). Thereafter, viable E. anguillarum was continuously isolated from both the intestine and trunk kidney of non-injected fish. The proportion of fish from which the bacteria were isolated from the kidneys reached 100% at 14 dpc in Trial 1 and at 5 dpc in Trial 2. The viable counts of the E. anguillarum in the intestine ranged from 3.17 to 6.88 log10 CFU/g in Trial 1 and 1.79 to 7.05 log10 CFU/g in Trial 2. In the trunk kidney, the counts ranged from 3.76 to 7.85 log10 CFU/g in Trial 1 and 2.33 to 9.56 log10 CFU/g in Trial 2, showing an increasing trend until the end of the experiment. E. anguillarum was also isolated from the type II kidneys of all surviving fish injected with E. anguillarum at the end of Experiment 1 (6.22–7.09 log10 CFU/g), whereas the bacterium was not isolated from the kidneys of PBS-injected fish in Trials 1 and 2.
Experiment 2
Although scavenging behavior on uninfected fish carcasses was observed in the control groups in all trials, no dead or E. anguillarum-infected fish were observed in any control fish. Furthermore, neither viable E. anguillarum nor eDNA was detected in seawater at 0 dpc or in rearing water from the control group in either Trial.
The viable counts and eDNA concentrations of E. anguillarum in the rearing water of the scavenging and non-scavenging groups are shown in Fig. 2 and Table 2. Scavenging behavior on infected fish carcasses was observed exclusively in the scavenging groups, beginning at 1 dpc in all trials, and the edible portions of the carcasses appeared to have been completely consumed by 5 dpc. Viable E. anguillarum was consistently isolated from rearing water in the scavenging group after 1 dpc (Trial 1, 0.82–3.35 log10 CFU/mL; Trial 2, 1.52–3.40 log10 CFU/mL; Trial 3, 0.52–4.14 log10 CFU/mL), except at 5 and 7 dpc in Trial 1, and 7 dpc in Trial 3. In the non-scavenging group, viable E. anguillarum were isolated from 1–5 dpc in Trials 1 and 2 (Trial 1, 1.56–2.60 log10 CFU/mL; Trial 2, 1.00–2.85 log10 CFU/mL), except at 4 dpc in Trial 1. In Trial 3, viable bacteria were consistently isolated after 1 dpc (0.52–3.77 log10 CFU/mL). eDNA of E. anguillarum was detected in all rearing water samples collected after 1 dpc in both scavenging (Trial 1, 2.55–3.82 log10 copies/mL; Trial 2, 3.38–4.09 log10 copies/mL; Trial 3, 2.81–5.19 log10 copies/mL) and non-scavenging groups (Trial 1, 1.59–3.44 log10 copies/mL; Trial 2, 2.12–3.70 log10 copies/mL; Trial 3, 2.97–4.66 log10 copies/mL). Although the maximum viable counts and eDNA concentration of E. anguillarum were generally observed at 1–2 dpc, the eDNA concentrations peaked at 4 dpc in the scavenging group in Trial 2 and at 5 dpc in the non-scavenging group in Trial 1. The peak values of both the viable counts and eDNA concentrations were higher in the scavenging groups than in the non-scavenging groups in all trials, with fold differences ranging from 2.3 to 5.6 for viable counts and from 2.4 to 3.4 for eDNA concentrations.
Fig. 2. Changes in viable counts (A, B, C) and eDNA concentrations (D, E, F) of E. anguillarum in the rearing water of scavenging and non-scavenging groups from E. anguillarum-transmission experiments using infected carcasses (Experiment 2). Arrows indicate the day on which scavenging behavior on the carcasses was first observed. ND; not detected. *Water sampling in the non-scavenging group in Trial 1 was limited to 9 dpc, as all fish in this group died at 10 dpc due to an aquarium system failure. No viable E. anguillarum or eDNA was detected in control groups.
Table 2. Results of the examination of red sea bream in the scavenging and non-scavenging groups from Experiment 2
| Trial | Group | Maximum E. anguillarum-
concentration in rearing water | dpc*2 | No. of fish
sampled | Status | No. of fish showing clinical signs | No. of fish in which
E. anguillarum
was isolated |
|---|
| Gross trunk kidney | External*5
symptoms |
|---|
Viable counts
(log10 CFU/mL) | eDNA conc.
(log10 copies/mL) | Type I | Type II | Total |
|---|
| Trial 1 | Scavenging | 3.35 (1)*1 | 3.82 (1) | 7 | 5 | Surviving | 1 (20.0)*4 | 3 (60.0) | 4 (80.0) | 2 (40.0) | 4 (80.0)*6 |
| | | | 24 | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100) |
| | | | 28 | 14 | Surviving | 5 (35.7) | 7 (50.0) | 12 (85.7) | 4 (28.6) | 12 (85.7) |
| Non-scavenging | 2.60 (3) | 3.44 (5) | 7 | 5 | Surviving | 0 | 1 (20.0) | 1 (20.0) | 0 | 1 (20.0) |
| | | | 10 | 15 | Dead*3 | 1 (6.7) | 1 (6.7) | 2 (13.3) | 0 | 6 (40.0) |
| Trial 2 | Scavenging | 3.40 (2) | 4.09 (4) | 5 | 5 | Surviving | 1 (20.0) | 3 (60.0) | 4 (80.0) | 1 (20.0) | 5 (100) |
| | | | 8 | 1 | Dead | 0 | 1 (100) | 1 (100) | 1 (100) | 1 (100) |
| | | | 10 | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100) |
| | | | 10 | 13 | Surviving | 4 (30.8) | 8 (61.5) | 12 (92.3) | 6 (46.2) | 12 (92.3) |
| Non-scavenging | 2.85 (1) | 3.70 (2) | 5 | 5 | Surviving | 0 | 0 | 0 | 0 | 0 |
| | | | 10 | 15 | Surviving | 2 (13.3) | 0 | 2 (13.3) | 0 | 2 (13.3) |
| Trial 3 | Scavenging | 4.14 (1) | 5.19 (1) | 5 | 2 | Dead | 0 | 2 (100) | 2 (100) | 2 (100) | 2 (100) |
| | | | 5 | 10 | Surviving | 3 (30.0) | 7 (70.0) | 10 (100) | 0 | 10 (100) |
| | | | 10 | 10 | Surviving | 4 (40.0) | 6 (60.0) | 10 (100) | 5 (50.0) | 10 (100) |
| Non-scavenging | 3.77 (1) | 4.66 (1) | 3 | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100) |
| | | | 5 | 1 | Dead | 0 | 1 (100) | 1 (100) | 1 (100) | 1 (100) |
| | | | 5 | 10 | Surviving | 3 (30.0) | 3 (30.0) | 6 (60.0) | 3 (30.0) | 6 (60.0) |
| | | | 10 | 1 | Dead | 0 | 1 (100) | 1 (100) | 1 (100) | 1 (100) |
| | | | 10 | 9 | Surviving | 2 (22.2) | 3 (33.3) | 5 (55.6) | 2 (22.2) | 5 (55.6) |
No mortality or E. anguillarum-infected fish were observed in control groups.
*
1 Days post challenge when the maximum value was obtained.
*
3 Fish died at 10 dpc due to aquarium system failure were examined as surviving fish for up to 10 dpc.
*
4 The ratio of fish showing the clinical signs.
*
5 All fish with external symptoms (ulcerative and/or hemorrhagic lesions) had type I or II kidneys.
*
6 Isolation rate of
E. anguillarum.
The results of the red seabream examination of the scavenging and non-scavenging groups are summarized in Table 2. Over 80% of the surviving fish in the scavenging groups exhibited either type I or type II kidneys in all trials (Trial 1, 80.0–85.7%; Trial 2, 80.0–92.3%; Trial 3, 100%), with more than half of these kidneys classified as type II (Trial 1, 50.0–60.0%; Trial 2, 60.0–61.5%; Trial 3, 60.0–70.0%). Among these fish, 20.0–50.0% displayed external symptoms. One to two dead fish with type II kidneys were observed in all trials. Among them, one or two individuals in Trials 2 and 3 exhibited external symptoms. Less than 20% of the fish in the non-scavenging groups in Trials 1 and 2 showed white nodules (Trial 1, 13.3–20.0%; Trial 2, 0–13.3%), and the isolation ratio of E. anguillarum from the surviving fish remained below 40% (Trial 1, 20.0–40.0%; Trial 2, 0–13.3%). One fish in Trial 1 exhibited type II kidneys; nevertheless, no fish displayed external symptoms or mortality. In contrast, although the proportion was lower than that in the scavenging group, more than half of the surviving fish in the non-scavenging group in Trial 3 exhibited either type I or type II kidneys (55.6–60.0%), with 22.2–30.0% showing external symptoms. Three dead fish with type II kidneys were observed in this trial, two of which also displayed external symptoms. E. anguillarum was isolated from all the fish in all trials that exhibited white nodules. The infection ratio throughout the sampling periods was significantly higher in the scavenging group than in the non-scavenging group in all trials (p < 0.01).
Experiment 3
No viable colonies or E. anguillarum eDNA were detected in the seawater at 0 dpc or in the control groups in Trials 1 and 2. Additionally, no dead or E. anguillarum-infected fish were observed in any control group.
The results for the infection groups in Experiment 3 are summarized in Table 3. Viable E. anguillarum was isolated exclusively from the rearing water at 5 min post-challenge in Group A of Trial 1 (1.52 log10 CFU/mL), in Group B of Trial 2 (1.56 log10 CFU/mL), and at 1 dpc in Group C in Trial 1 (3.65 log10 CFU/mL). eDNA of E. anguillarum was detected in all rearing water samples collected after the introduction of infected materials in both trials. The eDNA concentrations were as follows for Trial 1: Group A, 0.91–3.63 log10 copies/mL; Group B, -0.20–2.93 log10 copies/mL, and Group C, 1.58–4.84 log10 copies/mL. For Trial 2, it was as follows: Group A: 1.60–3.61 log10 copies/mL; Group B: 1.39–3.75 log10 copies/mL; and Group C: 1.85–2.44 log10 copies/mL. The concentrations in Groups A and B peaked at 5 min post-challenge in both trials, whereas in Group C, they peaked at 1 dpc in Trial 1 and at 7 dpc in Trial 2. The highest eDNA concentration was recorded at 1 dpc in Group C in Trial 1 and at 5 min post-challenge in Group B in Trial 2.
Table 3. Summary of results for the infection groups from experimental oral infection with
E. anguillarum (Experiment 3)
| Trial | Group: infection method
(material administered or
immersed, concentration) | Periods
post
challenge | E. anguillarum concentration
in rearing water | No. of
fish
sampled | Status | No. of fish showing clinical signs | No. of fish
in which
E. anguillarum
was isolated |
|---|
Viable counts
(log10 CFU/mL) | eDNA conc.
(log10 copies/mL) | Gross trunk kidney | External*4
symptoms |
|---|
| Type I | Type II | Total |
|---|
| Trial 1 | Group A: forced oral administration
(E. anguillarum suspension, 7.20 log10 CFU/fish)*1 | 5 min | 1.52 | 3.63 | | | | | | | |
| 1 day | ND*2 | 1.68 | 5 | Surviving | 0 | 0 | 0 | 0 | 0 |
| 3 days | | | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100)*5 |
| 4 days | ND | 1.44 | 5 | Surviving | 0 | 1 (20.0) | 1 (20.0) | 0 | 1 (20.0) |
| 7 days | ND | 0.91 | 9 | Surviving | 5 (55.6)*3 | 1 (11.1) | 6 (66.7) | 0 | 6 (66.7) |
Group B: forced oral administration
(diseased-kidney homogenate, 6.13 log10 CFU/fish)*1 | 5 min | ND | 2.93 | | | | | | | |
| 1 day | ND | 0.93 | 5 | Surviving | 0 | 0 | 0 | 0 | 0 |
| 4 days | ND | −0.20 | 5 | Surviving | 0 | 0 | 0 | 0 | 0 |
| 7 days | ND | −0.70 | 10 | Surviving | 1 (10.0) | 1 (10.0) | 2 (20.0) | 0 | 2 (20.0) |
Group C: feeding adminitration
(E. anguillarum-containing feed, 6.47 log10 CFU/fish)*1 | 5 min | ND | 3.47 | | | | | | | |
| 1 day | 3.65 | 4.84 | 5 | Surviving | 0 | 0 | 0 | 0 | 0 |
| 4 days | ND | 2.18 | 5 | Surviving | 0 | 1 (20.0) | 1 (20.0) | 0 | 1 (20.0) |
| 7 days | ND | 1.58 | 10 | Surviving | 0 | 1 (10.0) | 1 (10.0) | 0 | 1 (10.0) |
| Trial 2 | Group A: forced oral administration
(E. anguillarum suspension, 7.78 log10 CFU/fish)*1 | 5 min | ND | 3.61 | | | | | | | |
| 1 day | ND | 1.60 | 10 | Surviving | 0 | 0 | 0 | 0 | 4 (40.0) |
| 4 days | ND | 1.79 | 10 | Surviving | 1 (10.0) | 1 (10.0) | 2 (20.0) | 0 | 3 (30.0) |
| 7 days | ND | 2.14 | 15 | Surviving | 6 (40.0) | 5 (33.3) | 11 (73.3) | 0 | 13 (86.7) |
Group B: forced oral administration
(diseased-kidney homogenate, 7.66 log10 CFU/fish)*1 | 5 min | 1.56 | 3.75 | | | | | | | |
| 1 day | ND | 1.54 | 10 | Surviving | 0 | 0 | 0 | 0 | 0 |
| 4 days | ND | 1.39 | 10 | Surviving | 2 (20.0) | 3 (30.0) | 5 (50.0) | 0 | 5 (50.0) |
| 7 days | ND | 2.27 | 15 | Surviving | 5 (33.3) | 4 (26.7) | 9 (60.0) | 0 | 10 (66.7) |
Group C: immersion
(E. anguillarum-suspension, 4.78 log10 CFU/mL)*1 | 5 min | ND | 1.85 | | | | | | | |
| 1 day | ND | 1.92 | 10 | Surviving | 0 | 0 | 0 | 0 | 10 (100) |
| 4 days | ND | 2.29 | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100) |
| | | 10 | Surviving | 6 (60.0) | 4 (40.0) | 10 (100) | 3 (30.0) | 10 (100) |
| 6 days | | | 1 | Dead | 0 | 1 (100) | 1 (100) | 0 | 1 (100) |
| 7 days | ND | 2.44 | 13 | Surviving | 3 (23.1) | 10 (76.9) | 13 (100) | 4 (30.8) | 13 (100) |
No viable E. anguillarum, its eDNA, or infected fish were detected in pre-infection or control group water; no mortality occurred in the control groups.
*
1 CFU was calculated using SSA plates.
*
3 The ratio of fish showing the clinical signs.
*
4 All fish with external symptoms (ulcerative and/or hemorrhagic lesions) had type I or II kidneys.
*
5 Isolation rate of
E. anguillarum.
White nodules in the trunk kidney were observed in all infection groups, and E. anguillarum was isolated from all fish with nodules, including dead individuals. In Trial 1, dead fish with type II kidneys were observed at 3 dpc in Group A. At 4 and 7 dpc, 20.0% and 66.7% of surviving fish in this group exhibited either type I or type II kidneys, respectively. Affected kidneys were also found in surviving fish at 7 dpc in Group B and at 4 and 7 dpc in Group C, although at lower proportions (10.0–20.0%). In this trial, E. anguillarum was isolated only from fish that exhibited affected kidneys. Although the infection ratio in Group A throughout the sampling periods did not differ significantly from those in Groups B and C, the p-value was close to the significance threshold (p = 0.0648), suggesting a possible trend toward higher infection ratios in Group A. In trial 2, E. anguillarum was isolated from 40.0% of the surviving fish without nodules at 1 dpc in Group A. At 4 and 7 dpc, 20.0% and 73.3% of surviving fish in this group had affected kidneys, and E. anguillarum was isolated from 30.0% and 86.7% of the fish, respectively. In Group B, affected kidneys were observed in 50.0% and 60.0% of surviving fish at 4 and 7 dpc, respectively, with corresponding isolation ratios of 50.0% and 66.7%. In Group C, E. anguillarum was isolated from all fish, and 100% of the surviving fish exhibited affected kidneys at both 4 and 7 dpc. Dead fish were observed at 4 and 6 dpc. The infection ratio across all sampling periods in Group C was significantly higher than those in Groups A and B (p < 0.01). Across both trials, external symptoms were observed only in Group C in Trial 2.
Discussion
The results from the experiments conducted in this study confirmed that E. anguillarum from infected fish carcasses was transmitted to healthy fish. The scavenging behavior by non-injected fish on dead injected fish was observed in the infection groups in Experiment 1, after which the non-injected fish eventually became infected with E. anguillarum and exhibited characteristic signs of edwardsiellosis. This included white nodules in the trunk kidney and external symptoms (Fig. 1 and Table 1). Fish infected with E. anguillarum from infected fish carcasses were also found in Experiment 2, regardless of the presence or absence of scavenging behavior. However, the infection ratio of E. anguillarum in the scavenging group was significantly higher than that in the non-scavenging groups (Table 2). Fish in Experiment 3 were also infected with E. anguillarum through oral administration of an E. anguillarum suspension or diseased-kidney homogenate (Table 3). These results indicated that the transmission of E. anguillarum to healthy fish was facilitated by the scavenging behavior on fish carcasses infected with the bacterium.
Among the observed symptoms, the trunk kidney with white nodules were classified into types I and II (Fig. 1) to monitor their developmental stages. Dead fish were found in all experiments, and these fish exhibited type II kidneys, characterized by white nodules with overall kidney enlargement. Toida et al. (2003) reported that the nodules observed in the kidney of E. anguillarum-infected red seabream were granulomas and that the size and proportion of granulomas in the kidney increased as the infection progressed. Based on this, we assumed that the affected trunk kidney observed in this study developed from type I to type II and that some fish with type II kidneys eventually died. Furthermore, since fish exhibiting affected trunk kidneys were observed from 3 dpc in all experiments (Tables 1, 2, and 3), it is likely that E. anguillarum rapidly caused edwardsiellosis after entering the host.
The viable counts of E. anguillarum in the rearing water in Experiments 1 and 2 peaked on the days when the scavenging behavior on fish carcasses was observed and/or the following day in each trial. The eDNA concentrations on these days were generally higher than those on the other days (Tables 1 and 2; Fig. 2). These findings indicated that E. anguillarum in the fish carcasses spreads into rearing water through scavenging behavior. Given that Edwardsiella bacteria are known to use the skin, gills, nasal cavity, and gastrointestinal tract as entry points (Morrison and Plumb, 1994; Ling et al., 2001; Kataoka et al., 2017; Endo et al., 2022), it is evident that E. anguillarum in rearing water contributes to infection via external routes. Furthermore, following the scavenging behavior in Experiment 1, the number of fish from which E. anguillarum was isolated from the intestine increased the next day, followed by the appearance of E. anguillarum-infected fish with white nodules on the trunk kidney (Table 1). This suggested that E. anguillarum from fish carcasses may infect healthy fish not only via external routes but also through oral transmission. Several fish species have been reported to be orally infected with Edwardsiella bacteria (Gutierrez and Miyazaki, 1994; Mekuchi et al., 1995; Santander et al., 2013). The proportion of fish infected with E. anguillarum in Experiment 3 increased with the concentration of orally administered E. anguillarum, regardless of its concentration in the rearing water (Table 3). This confirmed that E. anguillarum can infect red seabream orally. Additionally, Soto et al. (2013) found that E. ictaluri infecting Nile tilapia Oreochromis niloticus via immersion challenge was detected in the body mucus, stomach, and proximal intestine 3 h post-infection, concluding that cutaneous and oral routes serve as initial portals of entry. Consistent with this, our results suggested that E. anguillarum in dead infected red seabream is likely to be transmitted to healthy fish through both external and oral routes.
The final isolation ratio of E. anguillarum in the forced oral and feeding administration groups of Experiment 3 was 10.0–20.0% at an infection concentration of 106 CFU/fish and 66.7–86.7% at 107 CFU/fish. In contrast, the bacterium was isolated from all individuals in Group C in Trial 2, where fish were immersed in a bacterial suspension of 104 CFU/mL (Table 3). Although a direct comparison cannot be made because of the differences in administration methods, immersion is presumed to have infected all fish at a lower concentration than the other administration methods. Furthermore, E. anguillarum-infected fish were also found in the non-scavenging group in Trials 1 and 2 of Experiment 2, where the maximum viable counts of E. anguillarum in the rearing water was in the order of 102 CFU/mL (Fig. 2 and Table 2). Even considering that our preliminary analysis (data not shown) indicated viable counts of E. anguillarum grown on SSA plates were lower than those grown on TSA plates by a factor of 2 to 3, the viable counts in the rearing water remained lower than those used in immersion and intranasal challenges to red seabream in previous studies (105–107 CFU/fish or CFU/mL) (Matsuyama et al., 2005; Kurohara et al., 2008; Hanyu et al., 2014; Endo et al., 2022) or in this study. Our results suggest that even a small number of E. anguillarum released from fish carcasses may be capable of infecting red seabream via external routes.
External symptoms such as ulcerative and/or hemorrhagic lesions (Fig. 1A–C) were not observed in the non-scavenging groups in Trials 1 and 2 of Experiment 2 (Table 2) or in any of the forced oral administration groups in Experiment 3 (Table 3). The maximum viable counts of E. anguillarum in the rearing water of these groups ranged from below the detection limit to 2.85 log10 CFU/mL. This was lower than the counts in the water of groups where fish exhibiting the external symptoms were observed, as well as the bacterial suspension used for the immersion infection (≥3.03 log10 CFU/mL). These data point to the possibility that the development of external symptoms requires infection with a sufficient number of E. anguillarum via external routes, and that the risk increases with the spread of E. anguillarum from infected fish carcasses through scavenging behavior.
Furthermore, the findings of this study suggest that, to better evaluate the risk of infection, it is necessary to consider not only viable cell counts but also the potential presence of non-culturable pathogenic bacteria and non-pathogenic cells. E. anguillarum eDNA was consistently detected in Experiments 2 and 3, even after viable cells were no longer isolated, and in some cases, detection patterns differed from that of viable cells (Fig. 2 and Table 3). Previous studies have reported that Edwardsiella bacteria can become non-culturable in seawater or freshwater while retaining their pathogenicity to fish (Du et al., 2007; Esteve and Alcaide, 2022). Similarly, Kurohara et al. (2009) showed that E. anguillarum lost its ability to grow on agar plates within 24 h in seawater but maintained its pathogenicity to red seabream. In light of these studies, the detected eDNA likely included contributions from non-culturable bacteria that retained their pathogenicity. In contrast, although viable E. anguillarum reached 3.65 log10 CFU/mL in the water of Group C in Trial 1 of Experiment 3 at 1 dpc (the day after administration of E. anguillarum-containing feed), no fish exhibited external symptoms, and the isolation ratio of E. anguillarum was only 10.0–20.0% (Table 3). Since viable E. anguillarum was detected only at 1 dpc and not on subsequent days, it is likely that the bacteria detected in this group’s water were either low-pathogenic cells that had entered the red seabream and were excreted without establishing infection, or cells that had attenuated within the fish. Taken together, these observations indicate that the rearing water of E. anguillarum-infected fish may contain a mixture of viable and pathogenic bacteria, non-culturable but potentially infectious cells, dead cells, and non-pathogenic live cells. This complexity should be carefully considered when interpreting the relationship between eDNA detection and risk of infection.
As described above, our findings indicate that infected fish carcasses serve as a reservoir for E. anguillarum. Furthermore, scavenging behavior may play a critical role in the transmission of edwardsiellosis in red seabream farms by facilitating both the release of the bacterium into the surrounding water and the direct consumption of infected tissues. This highlights the importance of promptly removing dead fish to mitigate the impact of edwardsiellosis on red seabream farming. However, the activity and pathogenicity of E. anguillarum in seawater remain poorly understood. Moreover, although the results suggested that even a small number of E. anguillarum released from fish carcasses could infect red seabream, we were unable to evaluate whether the pathogenicity of E. anguillarum increases within the carcasses. Further studies are needed to elucidate the detailed infection cycle and of E. anguillarum on red seabream farms, as well as the mechanisms underlying enhanced pathogenicity in this bacterium.
Acknowledgements
We thank Editage (https://www.editage.jp/) for the English language editing. This study was supported by the YANMAR Co., Environmental Sustainability Support Association (YESSA) research program (grant no. KI0202016), Japan Society for the Promotion of Science (JSPS) KAKENHI (grant no. 21K14901), and Japan Science and Technology Agency (JST) A-STEP Tryout (grant no. JPMJTM20RU).
References
- Du M, Chen J, Zhang X, Li A, Li Y, Wang Y (2007) Retention of virulence in a viable but nonculturable Edwardsiella tarda isolate. Appl Environ Microbiol 73: 1349–1354.
- Endo T, Kitamura S-I, Kawakami H, Iida T (2022) The nasal cavity of red sea bream Pagrus major is an important entry site for Edwardsiella anguillarum. Fish Pathol 57: 26–29.
- Esteve C, Alcaide E (2022) Monitoring the starvation-survival response of Edwardsiella piscicida and E. tarda in freshwater microcosms, at various temperatures. Microorganisms 10: 1043.
- Griffin MJ, Ware C, Quiniou SM, Steadman JM, Gaunt PS, Khoo LH, Soto E (2014) Edwardsiella piscicida identified in the Southeastern USA by gyrB sequence, species-specific and repetitive sequence-mediated PCR. Dis Aquat Organ 108: 23–35.
- Gutierrez MA, Miyazaki T (1994) Responses of Japanese eels to oral challenge with Edwardsiella tarda after vaccination with formalin-killed cells or lipopolysaccharide of the bacterium. J Aquat Anim Health 6: 110–117.
- Hanyu K, Miyamoto A, Nakai T (2014) Carrying situation in red sea bream artificially infected with Edwardsiella tarda. Nippon Suisan Gakkaishi 80: 572–577. (In Japanese with English abstract)
- Iida T, Sakai T, Takano T (2016) Edwardsiellosis in fish. Fish Pathol 51: 87–91. (In Japanese with English abstract)
- Kataoka C, Tomiyama H, Kashiwada S (2017) Three-dimensional visualization of green fluorescence protein-labelled Edwardsiella tarda in whole medaka larvae. J Fish Dis 40: 479–484.
- Klesius P (1994) Transmission of Edwardsiella ictaluri from infected, dead to noninfected channel catfish. J Aquat Anim Health 6: 180–182.
- Kurohara K, Kimura Y, Sekiguchi Y, Kawai K (2008) Influence of infection methods on the distribution of Edwardsiella tarda in red sea bream Pagrus major tissues. Aquacult Sci 56: 551–558. (In Japanese with English abstract)
- Kurohara K, Kimura Y, Sekiguchi Y, Kawai K (2009) Virulence of fish-pathogenic bacterium Edwardsiella tarda to red sea bream Pagrus major after survival in seawater. Aquacult Sci 57: 227–232. (In Japanese with English abstract)
- Ling SH, Wang XH, Lim TM, Leung KY (2001) Green fluorescent protein-tagged Edwardsiella tarda reveals portal of entry in fish. FEMS Microbiol Lett 194: 239–243.
- Matsuoka S (2004) Discharge of Edwardsiella tarda cells from experimentally infected Japanese flounder. Fish Pathol 39: 9–13. (In Japanese with English abstract)
- Matsuyama T, Kamaishi T, Ooseko N, Kurohara K, Iida T (2005) Pathogenicity of motile and non-motile Edwardsiella tarda to some marine fish. Fish Pathol 40: 133–135.
- Mekuchi T, Kiyokawa T, Honda K, Nakai T, Muroga K (1995) Infection experiments with Edwardsiella tarda in the Japanese flounder. Fish Pathol 30: 247–250. (In Japanese with English abstract)
- Morrison EE, Plumb JA (1994) Olfactory organ of channel catfish as a site of experimental Edwardsiella ictaluri infection. J Aquat Anim Health 6: 101–109.
- Nguyen VT, Satoh S, Haga Y, Fushimi H, Kotani T (2008) Effect of zinc and manganese supplementation in artemia on growth and vertebral deformity in red sea bream (Pagrus major) larvae. Aquaculture 285: 184–192.
- Reichley SR, Ware C, Greenway TE, Wise DJ, Griffin MJ (2015) Real-time polymerase chain reaction assays for the detection and quantification of Edwardsiella tarda, Edwardsiella piscicida, and Edwardsiella piscicida-like species in catfish tissues and pond water. J Vet Diagn Invest 27: 130–139.
- Santander J, Martin T, Loh A, Pohlenz C, Gatlin DM, Curtiss R (2013) Mechanisms of intrinsic resistance to antimicrobial peptides of Edwardsiella ictaluri and its influence on fish gut inflammation and virulence. Microbiology 159: 1471–1486.
- Shao S, Lai Q, Liu Q, Wu H, Xiao J, Shao Z, Wang Q, Zhang Y (2015) Phylogenomics characterization of a highly virulent Edwardsiella strain ET080813(T) encoding two distinct T3SS and three T6SS gene clusters: Propose a novel species as Edwardsiella anguillarum sp. nov. Syst Appl Microbiol 38: 36–47.
- Soto E, Illanes O, Revan F, Griffin M, Riofrio A (2013) Bacterial distribution and tissue targets following experimental Edwardsiella ictaluri infection in Nile tilapia Oreochromis niloticus. Dis Aquat Organ 104: 105–112.
- Sugiura H, Fukunishi K, Kawakami H, Imajoh M (2022) Phenotypic differences between Edwardsiella piscicida and Edwardsiella anguillarum isolates in Japan. J Aquat Anim Health 34: 197–207.
- Takeuchi H, Kawakami H, Mano N, Yamanaka H, Shimizu S (2024) Environmental DNA-based quantification of Edwardsiella bacteria and fish-derived materials in rearing water of infected ayu Plecoglossus altivelis and red sea bream Pagrus major. Fish Sci 90: 93–103.
- Toida S, Kanai K, Yoshikoshi K (2003) The kinetics of leukocytes and histopathology of red sea bream in artificial infection of Edwardsiella tarda. Fish Pathol 38: 137–142.
- Tsukamoto K, Kuwada H, Hirokawa J, Oya M, Sekiya S, Fujimoto H, Imaizumi K (1989) Size-dependent mortality of red sea bream, Pagrus major, juveniles released with fluorescent otolith-tags in News Bay, Japan. J Fish Biol 35: 59–69.
- Yasunaga N (1977) Etiological agent from abscess and nodule formation disease of cultured red sea bream. Bull Nagasaki Prefecture Inst Fish 3: 1–4. (In Japanese)
- Yasunaga N, Ogawa S, Hatai K (1982) Characteristics of the fish pathogen Edwardsiella isolated from several species of cultured marine fishes. Bull. Nagasaki Prefect Inst Fish 8: 57–56. (In Japanese)
