Optimization of an Isolation Method for Ichthyobacterium seriolicida, the Causative Agent of Bacterial Hemolytic Jaundice

Abstract

We optimized a method for isolating Ichthyobacterium seriolicida, which is difficult to isolate even from freshly dead diseased fish. The bacterium was isolated from the blood of diseased yellowtail Seriola quinqueradiata up to 3 h after death, but not from the spleen, even when sampled immediately after death. The bacterium remained culturable from blood samples collected in medium-filled syringes and stored at 8°C for at least 72 h. Therefore, collecting blood within 3 h after death using a medium-filled syringe and storing it at low temperature is considered effective for preserving the viability of I. seriolicida.

Bacterial hemolytic jaundice (BHJ) is a bacterial infection caused by Ichthyobacterium seriolicida. BHJ outbreaks were first reported in the early 1980s, and by the 1990s, the disease had spread to Japanese amberjack aquaculture operations along the coasts of Kyushu Island, Shikoku Island, and the Kii Peninsula of Japan (Sorimachi et al., 1993; Takano and Matsuyama, 2020). BHJ is now the third most damaging fish disease in yellowtail aquaculture, after streptococcosis and nocardiosis, causing significant economic harm in fish aquaculture (2019 Annual Reports on Fish Diseases, Food Safety and Consumer Affairs Bureau, Ministry of Agriculture, Forestry and Fisheries, https://www.maff.go.jp/j/syouan/suisan/suisan_yobo/disease/attach/pdf/gyobyou_higai_jyoukyou-2.pdf). The causative bacterium of BHJ does not grow in culture media for general bacteria and can be cultured only in antibiotic-free Leibovitz L-15 medium supplemented with fetal bovine serum. Moreover, I. seriolicida grows very slowly even in this medium (Sorimachi et al., 1993; Iida and Sorimachi, 1994). For these reasons, I. seriolicida is prone to contamination and difficult to isolate. Furthermore, empirical evidence indicates that I. seriolicida is difficult to isolate from dead fish, possibly because the bacterium is quickly lost or becomes unculturable within dead fish. Thus, successful isolation of I. seriolicida requires that samples be collected at the aquaculture farm and promptly returned to the laboratory for aseptic manipulation using a laminar flow cabinet. These challenges impede the collection of clinical strains, disease diagnosis, and epidemiologic studies. Hence, an optimized method for isolating I. seriolicida is needed.

Aseptic blood sampling using a syringe is currently the only method available for isolating I. seriolicida (Sorimachi et al., 1993). However, it remains unclear how many hours after the death of a diseased fish it becomes impossible to isolate the bacterium using this method. Although the spleen commonly exhibits distinct lesions in Japanese amberjack affected by BHJ (Takano and Matsuyama, 2020), there are no reports on the feasibility of isolating I. seriolicida from the organ. In this study, therefore, we determined the postmortem time during which I. seriolicida can be isolated using the conventional method from blood samples collected in fish infection experiments. We also tested the feasibility of using the spleen for isolation, as this organ is easy to collect and shows typical lesions of BHJ, such as swelling (Sorimachi et al., 1993) and necrosis (Maeno et al., 1995). In addition, to maximize the preservation time of blood samples, we compared the “shelf life” of I. seriolicida in blood stored directly in a tube and in blood collected and stored in a syringe filled with liquid culture medium.

Materials and Methods

Bacterial culture and preparation of diseased fish

Ichthyobacterium seriolicida JBKA-6 (= ATCC BAA-2465^T = JCM 18228^T) and isolate 013821 (isolated in 2001 the blood of diseased Japanese amberjack farmed in Oita Prefecture, western Japan) were used for experimental infections. JBKA-6 and 013821 were cultured according to an established method (Iida and Sorimachi, 1994; Takano and Matsuyama, 2020). The medium for culture and isolation was prepared as follows: Leibovitz’s L-15 medium (Gibco) was supplemented with fetal bovine serum (Gibco) and NaCl (Fujifilm Wako Pure Chemical Corp.) at final concentrations of 10% and 1.6%, respectively. The prepared medium was sterilized using Nalgene™ Rapid-Flow™ sterile disposable filter units with polyethersulfone membranes (Thermo Fisher Scientific) with a pore size of 0.2 μm. The bacterial stock was inoculated in 100 mL of sterilized medium and incubated at 180 rpm and 26°C for 80–120 h as a culture. The bacterial suspension was cultured until an optical density at 600 nm (OD₆₀₀) of 0.10–0.15 was reached. The OD₆₀₀ value was measured using an Ultrospec 3300 Pro spectrophotometer (Amersham Bioscience Corp.).

In Experiments 1–3, Seriola quinqueradiata (weight: 43.6–134.6 g) produced in March 2023 at the Japan Fisheries Research and Education Agency, Goto Field Station, were used as test fish. In Experiment 4, Seriola quinqueradiata (weight: 106.2–150.6 g) produced in October 2024 at Yamasaki Giken Co., Ltd., were used as test fish. All test fish used in this study were healthy individuals that had been reared from the juvenile stage to the time of experimentation in a controlled facility at Nansei Field Station, Japan Fisheries Research and Education Agency. Test fish were injected with 100 μL of bacterial suspension per fish via the tail vein using a 1.0-mL syringe equipped with a 23G needle. After injection, the fish were transferred to a 0.5-ton panlitic tank supplied with free-flowing, sand-filtered seawater adjusted to 25°C. On 6 days after bacterial injection, all surviving fish exhibiting typical symptoms of BHJ, such as slow swimming and jaundice in the orbital and rostral areas (Wada et al., 1989; Sorimachi et al., 1993), were collected and euthanized with an excess dose of 2-phenoxyethanol (Fujifilm Wako Pure Chemical Corp.), then subjected to the following tests. A total of 20 I. seriolicida–infected diseased fish were prepared for each experiment.

Experiment 1: Determination of post-mortem time during which I. seriolicida can be isolated from blood

To determine the post-mortem time during which I. seriolicida JBKA-6 can be isolated from the blood, four diseased fish (A1–A4) were collected and euthanized, and the body surface was disinfected using 70% ethanol. During the experiment, euthanized diseased fish were kept at 8°C, and approximately 0.1 mL of blood was collected from the tail vein of each fish at 0, 1, 3, 5, and 7 h post-euthanasia using a 1.0-mL syringe equipped with a 23G needle. Bacterial isolation was performed by adding 50 μL of blood to a test tube containing 5 mL of medium and then incubating the tube for 7 days at 26°C with shaking. All manipulations except blood collection were performed aseptically in a laminar flow cabinet.

Experiment 2: Isolation from the spleen

To assess the possibility of isolating I. seriolicida JBKA-6 from the spleen, four diseased fish (B1–B4) were collected and euthanized, and the body surface was disinfected using 70% ethanol. Approximately 0.1 mL of blood was collected from the tail vein of each diseased fish immediately following euthanasia using a 1.0-mL syringe equipped with a 23G needle, and the possibility of successful bacterial isolation was evaluated using the method described above (control test). After blood collection, the whole spleen was collected from each fish and stored in a 1.5-mL tube. Appropriate amounts of spleen tissue were taken from the tube and stored at 8°C for 0, 1, 3, 5, and 7 h after collection, ground using a Biomasher V (Nippi Corp.), and homogenized to 100 mg/mL in liquid medium. Bacterial isolation was performed by adding 50 μL of homogenized suspension to a test tube containing 5 mL of liquid medium and incubated for 7 days at 26°C with shaking. All manipulations except blood collection were performed aseptically in a laminar flow cabinet.

Experiment 3: Culturability of I. seriolicida collected from blood using the medium-filled syringe method

The culturability of I. seriolicida JBKA-6 in blood collected with a medium-filled syringe was evaluated over time. Four diseased fish (C1–C4) were collected and euthanized, and the body surface was disinfected using 70% ethanol. Approximately 1.0 mL of blood was collected from the tail vein of each diseased fish immediately after euthanasia using a 1.0-mL syringe equipped with a 23G needle and transferred to a 1.5-mL tube (control sample). Supernatant (serum) was added to the medium as a sample because the control sample coagulated quickly in the tube. Simultaneously, approximately 0.1 mL of blood was repeatedly collected from the tail vein of the same fish using a 1.0-mL syringe equipped with a 23G needle pre-filled with 0.5 mL of liquid medium, making a total of six syringes containing a mixture of blood and liquid medium per fish (test sample). At 0, 1, 3, 6, 24, and 72 h after blood collection, 100 μL of control sample or test sample was inoculated into 5 mL of medium and incubated for 7 days at 26°C with shaking. All collected samples were kept at 8°C during the experiment. All manipulations except blood collection were performed aseptically in a laminar flow cabinet.

Experiment 4: Evaluation of the medium-filled syringe method for a low-growth isolate of I. seriolicida

The applicability of the medium-filled syringe method for isolating I. seriolicida from blood was evaluated using isolate 013821, which exhibited slower growth than JBKA-6 (Supplementary figure 1). Nine diseased fish infected with 013821 were collected and euthanized. The body surfaces of each fish were then disinfected with 70% ethanol. Approximately 1.0 mL of blood was collected from the tail vein of each diseased fish immediately after euthanasia using a 1.0-mL syringe equipped with a 23G needle and transferred into a hematocrit tube for hematocrit measurements. Hematocrit values were determined following the protocol of Haematological Examination of Fish (Z. Svobodová and B. Vykusová, 1991). The hematocrit tubes loaded with sampled blood were centrifuged at 10,000 rpm for 10 minutes using a MRX 152 (TOMY Digital Biology CO., LTD.). Simultaneously, approximately 0.1 mL of blood was collected from the tail vein using a 1.0-mL syringe equipped with a 23G needle pre-filled with 0.5 mL of liquid medium, resulting in a syringe containing a mixture of blood and liquid medium for each fish. Blood samples collected in medium-filled syringes were stored at 8°C until inoculation into fresh medium. One hour after blood collection, 100 μL of the mixture was added to 5 mL of medium and incubated for 7 days at 26°C with shaking. All manipulations except for blood collection and hematocrit measurements were performed aseptically in a laminar flow cabinet. As a positive control, the same procedures were also performed on five diseased fish infected with JBKA-6. As a negative control, the same procedures were also performed on nine naïve fish prior to the infection experiment.

Method for determining successful isolation

The success of bacterial isolation was evaluated at 7 days after the start of culture. Successful isolation was determined based on visual and microscopic examination for growth and contamination in the culture medium. The criterion for successful isolation of I. seriolicida is a turbidity (OD₆₀₀) of approximately 0.1–0.2 at the time of maximum growth (Takano et al., 2020), and contamination was determined if growth clearly exceeded this level. In addition, the culture medium pH becomes alkaline as I. seriolicida proliferates; if a yellowish change in the indicator was observed, indicating acidification of the culture medium, the culture was deemed contaminated. Cultures determined to be successful isolation visually were fluorescently stained using a LIVE/DEAD™ BacLight™ bacterial viability kit (Thermo Fisher Scientific), and pure cultures of I. seriolicida were confirmed by mirror inspection using a BX51-P polarizing microscope (Olympus Corp.).

Storage conditions for fish and blood samples

All fish and blood samples examined in this study were stored in a standard domestic refrigerator, simulating refrigerated transport to a laboratory for diseases diagnosis under suboptimal conditions encountered at certain aquaculture farms. The temperature within the domestic refrigerator used in the experiment was 8°C.

Ethics statement

All fish used for rearing and sampling in this study were handled in accordance with policies specified by the Animal Experimentation Committee of the Fisheries Technology Institute. All experiments involving fish were approved by the same committee (no. 2300).

Results

Experiment 1: Determination of post-mortem time during which I. seriolicida can be isolated from blood

The turbidity (OD₆₀₀) of the inoculum used for challenge was 0.120. The experimental results of the determination of post-mortem time during which I. seriolicida JBKA-6 could be isolated from blood are shown in Table 1. I. seriolicida was successfully isolated from blood samples from a total of three fish (A1–A3); the sample from fish A4 was contaminated with other bacteria immediately after euthanasia. The bacterium could be isolated from blood for 1–3 h after death, but after 5 h, it was no longer possible to isolate I. seriolicida from any of the fish. After 7 h, all specimens exhibited contamination.

Table 1. Results of Experiment 1: Determination of post-mortem time during which I. seriolicida can be isolated from blood of experimentally infected Japanese amberjack (0–7 h)

Time	Serial number of fish
	A1	A2	A3	A4
0 h	+	+	+	co.
1 h	co.	+	+	co.
3 h	co.	+	+	co.
5 h	co.	−	−	co.
7 h	co.	co.	co.	co.

Time, elapsed time since euthanasia; co., contaminated; +, successful isolation; −, not isolated.

Experiment 2: Feasibility of isolation from spleen

The results of bacterial isolation from spleen tissue are shown in Table 2. The turbidity (OD₆₀₀) of the inoculum used for challenge was 0.135. I. seriolicida JBKA-6 was successfully isolated from blood samples from a total of 3 fish (B2–B4); the sample from fish B1 was contaminated with other bacteria immediately after euthanasia (control test). However, I. seriolicida could not be isolated from spleen samples from diseased fish, even immediately after euthanasia.

Table 2. Results of Experiment 2: Determination of post-mortem time during which I. seriolicida can be isolated from the spleen of experimentally infected Japanese amberjack (0–7 h)

Time	Serial number of fish
	B1	B2	B3	B4
0 h	co.	−	co.	−
1 h	co.	−	co.	−
3 h	co.	−	co.	−
5 h	co.	−	co.	−
7 h	co.	−	co.	−

Time, elapsed time since euthanasia; co., contaminated; +, successful isolation; −, not isolated. Ichthyobacterium seriolicida could be isolated using blood samples from a total of 3 fish (B2–B4); blood from fish B1 was contaminated with other bacteria immediately after euthanasia (control test).

Experiment 3: Culturability of I. seriolicida collected from blood using a medium-filled syringe

The culturability of I. seriolicida JBKA-6 in blood collected from medium-filled syringes is shown in Table 3. The turbidity (OD₆₀₀) of the inoculum used for challenge was 0.103. I. seriolicida JBKA-6 could be isolated from the control samples (serum obtained from blood promptly stored in 1.5-mL tubes after collection) for 1–24 h after collection; however, it could not be isolated from any of the control samples after 72 h. By contrast, I. seriolicida could be successfully isolated from test samples collected using medium-filled syringes for up to 72 h after blood collection.

Table 3. Results of Experiment 3: Culturability of I. seriolicida in blood collected using a medium-filled syringe

Time	Serial number of fish
	C1	C2	C3	C4
0 h (control)	+	+	+	+
0 h (test)	+	+	+	+
1 h (control)	+	+	+	+
1 h (test)	+	+	+	+
3 h (control)	+	+	−	+
3 h (test)	+	+	+	+
6 h (control)	+	+	−	+
6 h (test)	+	+	+	+
24 h (control)	+	+	−	+
24 h (test)	+	+	+	+
72 h (control)	−	−	−	−
72 h (test)	+	+	+	+

Time, elapsed time since blood sample collection; +, successful isolation; −, not isolated. Control samples: blood collected in an empty syringe and stored in 1.5-mL tubes after collection. Because the blood coagulated rapidly, the serum supernatant was used as the control sample by adding it to the culture medium. Test samples: blood collected in a medium-filled syringe.

Experiment 4: Applicability of the medium-filled syringe method to a low-growth isolate of I. seriolicida

The results of analyses of I. seriolicida isolation from blood collected using medium-filled syringes are shown in Table 4. The turbidity (OD₆₀₀) of the 013821 and JBKA-6 inocula used for challenge was 0.100 and 0.118, respectively. All fish experimentally infected with either of the two I. seriolicida strains yielded successful re-isolation of the bacterium using the medium-filled syringe method. By contrast, no bacteria were isolated from any of the naïve fish samples collected using medium-filled syringes. Although hematocrit values were significantly lower in the two diseased fish groups compared with the naïve fish group, some diseased fish exhibited relatively high hematocrit values (Table 4, Fig. 1).

Table 4. Results of Experiment 4: Applicability of the medium-filled syringe method to a low-growth isolate (013821) of I. seriolicida

Group	No. of fish isolated/ No. of fish tested	Isolation rate	Range of hematocrit value
		(%)	0–10%	10–20%	20–30%	30–40%	40–50%	>50%	NA*
013821	9/9	100	−	3	1	1	−	1	3
JBKA-6	5/5	100	1	−	2	−	−	1	1
Naïve	0/9	0	−	−	−	−	1	8	0

Not available*; accurate measurement of hematocrit values was not possible due to poor separation of plasma and blood cells.

Fig. 1. Hematocrit values of naïve and infected fish in Experiment 4. Statistical significance was assessed using the Wilcoxon rank sum exact test with Bonferroni correction (*p<0.05; **p<0.005). Accurate hematocrit measurement was not possible due to poor separation of plasma and blood cells in one fish from the JBKA-6 group and three fish from the 013821 group.

Discussion

The upper limit of time during which I. seriolicida could be isolated from blood was 3 h after death of the fish, and it was confirmed under experimental conditions that this bacterium quickly becomes unculturable after the death of diseased fish (Table 1). These results suggest that transporting diseased fish from the field to the laboratory for I. seriolicida isolation may not be a reliable approach, as the bacterium would become unculturable before the isolation procedure. In addition, I. seriolicida cells have been detected in the splenic tissue of fish affected by BHJ, and distinct lesions have also been observed in the spleen (Maeno et al., 1995). However, in the present study, I. seriolicida could not be isolated from the spleen of fish immediately after death (Table 2). Although the inability to isolate the bacterium from splenic tissue is of interest, the data obtained in this study did not allow for clarification of the underlying cause. Taken together, it was considered important to collect blood samples from diseased fish for I. seriolicida isolation either at symptom onset or immediately after death.

In the blood collection method using a medium-filled syringe, samples could be stored in the syringe at 8°C for at least 72 h (Table 3). This means that the specimens can be stored for extended periods, as long as the blood is drawn from fish immediately after death. By comparison, I. seriolicida could be isolated from the control samples (serum obtained from clotted blood of diseased fish) for up to 24 h after death (Table 3, control sample), whereas the bacterium quickly became unculturable in blood remaining in the fish after euthanasia (Table 1). The reason for this difference remains unclear, but autolysis or other postmortem changes (Furnesvik et al., 2022; Huang et al., 2023) in euthanized fish may significantly affect the survival of I. seriolicida.

In a clinical setting, two possible methods are assumed for isolating I. seriolicida.; (i) transport fish specimens to the laboratory for isolation, or (ii) collect blood samples on site and return them to the laboratory. In the former case, a certain amount of time is needed to initiate I. seriolicida isolation, during which the bacteria are likely to become unculturable. In the latter case, although blood samples can be collected under good conditions, I. seriolicida may become unculturable during transport to the laboratory, depending on travel time. Even when blood was collected using syringes not pre-filled with medium, I. seriolicida was still isolated from the blood of three of four fish up to 24 h after collection. However, in sample C3, the bacterium could not be isolated as early as 3 h after blood collection (Table 3). Therefore, to successfully isolate I. seriolicida, it is important to inoculate blood samples into the medium as soon as possible after collection.

In Experiment 4, all diseased fish exhibiting typical symptoms of BHJ—namely, slow swimming and jaundice in the orbital and rostral areas—were sampled. These fish showed hematocrit values ranging from levels comparable to those of the naïve group to extremely low (Table 4, Fig. 1). Nevertheless, I. seriolicida was successfully isolated from all of the diseased fish (Table 4). These results indicate that bacterial isolation from jaundiced fish is feasible, regardless of the severity of hemolytic anemia. As the clinical diagnosis of BHJ is generally based on clinical signs (Takano and Matsuyama, 2020), this method appears suitable for isolating the bacterium from symptomatic fish.

These findings indicate that the optimal method for clinical isolation of I. seriolicida is to use syringes pre-filled with liquid medium and collect blood within 3 h of death, if possible. Furthermore, this study demonstrated that I. seriolicida remains viable for at least 72 h postmortem when blood is collected using a syringe pre-filled with liquid medium, allowing for transportation over this period.

Acknowledgements

We thank all staff members of the Marine Fisheries Research and Development Center for providing yellowtail juveniles. We also thank Dr. Yutaka Fukuda, Fisheries Research Division, Oita Prefectural Agriculture, Forestry and Fisheries Research Center, for advice regarding the experimental design. This study was funded by the Food Safety and Consumer Affairs Bureau, Ministry of Agriculture, Forestry, and Fisheries of Japan.

References

•  Furnesvik,  L.,  T.  Erkinharju,  M.  Hansen,  M. N.  Yousaf and  T.  Seternes (2022): Evaluation of histological post-mortem changes in farmed Atlantic salmon (Salmo salar L.) at different time intervals and storage temperatures. J. Fish Dis., 45, 1571–1580.
•  Huang,  X.,  H.  Wang and  Z.  Tu (2023): A comprehensive review of the control and utilization of aquatic animal products by autolysis-based processes: Mechanism, process, factors, and application. Food. Res. Int., 164, 112325.
•  Iida,  T. and  M.  Sorimachi (1994): Cultural-characteristics of the bacterium causing jaundice of yellowtail, Seriola quinqueradiata. Fish Pathol., 29, 25–28.
•  Maeno,  Y.,  K.  Nakajima,  M.  Sorimachi and  Y.  Inui (1995): Pathophysiological studies of jaundiced Yellowtail, Seriola quinqueradiata. Fish Pathol., 30, 7–14. (In Japanese with English abstract)
•  Sorimachi,  M.,  Y.  Maeno,  K.  Nakajima,  K.  Inouye and  Y.  Inui (1993): Causative agent of jaundice of yellowtail, Seriola quinquerdadiata. Fish Pathol., 28, 119–124.
•  Svobodová,  Z. and  B.  Vykusová (1991): Diagnostics, prevention and therapy of fish diseases and intoxications. In “Haematological examination of fish” (ed. by Z. Svobodová and B. Vykusová). URL: https://www.fao.org/fishery/docs/CDrom/aquaculture/a0845t/volume2/docrep/field/003/ac160e/AC160E09.htm#ch6.
•  Takano,  T. and  T.  Matsuyama (2020): Bacterial hemolytic jaundice. Fish Pathol., 55, 25–30. (In Japanese)
•  Takano,  T.,  T.  Matsuyama,  T.  Sakai,  Y.  Nakamura,  T.  Kamaishi,  C.  Nakayasu,  H.  Kondo,  I.  Hirono,  Y.  Fukuda,  M.  Sorimachi and  T.  Iida (2016): Ichthyobacterium seriolicida gen. nov., sp. nov., a member of the phylum ‘Bacteroidetes’, isolated from yellowtail fish (Seriola quinqueradiata) affected by bacterial haemolytic jaundice, and proposal of a new family, Ichthyobacteriaceae fam. nov. Int. J. Syst. Evol. Microbiol., 66, 580–586.
•  Wada,  S.,  K.  Hatai and  S. S.  Kubota (1989): Light microscopical observations of cultured yellowtail Seriola quinqueradiata, characterized by yellow discoloration of body surface. Fish Pathol., 24, 211–218. (In Japanese with English abstract)