2025 年 60 巻 2 号 p. 76-87
The histopathological features of yellowtail Seriola quinqueradiata experimentally infected with Lactococcus garvieae (strain Lg-Aq-33) and L. formosensis (Lg-Aq-55 and Lg-Aq-103) were compared to elucidate the histopathology of lactococcosis caused by L. formosensis. Experimental infections were performed in duplicate for each bacterial strain using injection (1.0 × 104 cfu/100 g body weight) and immersion (1 h, 1.0 × 106 cfu/mL) methods. The cumulative mortality trend after experimental infection in each experimental group did not differ between the two experiments. All infected fish showed lesions typical of lactococcosis, such as inflammation of the inner operculum, ulceration of the caudal peduncle, epicarditis, and encephalitis. Histopathological observations confirmed previously reported lesions of lactococcosis, and revealed new lesions of the disease and those that are characteristic of L. formosensis infections. New histopathological features of lactococcosis in yellowtail included bacterial invasion and granulomas in the parenchyma of the heart, as well as bacterial invasion in the spleen and kidney. Lactococcosis caused by L. formosensis had more frequent granulomatous inflammation and verrucos endocarditis than L. garvieae infection.
Lactococcosis caused by Lactococcus garvieae was first identified in 1974 at a fish farm of yellowtail Seriola quinqueradiata in the Kochi Prefecture (Yoshida, 2016). During the 1990s, this disease significantly damaged yellowtail aquaculture; however, the widespread use of injectable vaccines and antimicrobials that were approved around 2000 significantly reduced this problem. In 2012, a different bacterial serotype, which did not agglutinate with the previously used anti-L. garvieae serum, was isolated from yellowtail in the Oita Prefecture (Oinaka et al., 2015). The bacterium was considered homologous to L. garvieae, based on the sequence homology of the 16S rRNA gene; thus, L. garvieae was classified by antisera into two serotypes, serotype I and the newer serotype II (Fukuda et al., 2015). However, a subsequent study classified serotype II bacteria as L. formosensis and not as L. garvieae (Mahmoud et al., 2023). The disease caused by L. formosensis has not been controlled owing to the low efficacy of serotype I-targeting vaccines and pharmaceuticals against L. formosensis outbreaks. Diseased fish with lactococcosis caused by L. formosensis present gross features, including protruding eyes, inflammation of the medial operculum, and ulcers on the caudal peduncle, with epicarditis and encephalitis observed on autopsy (Yoshida, 2016). Although these manifestations are similar to those caused by L. garvieae, no studies have compared the signs of these two bacteria. On the other hand, histopathological observations of fish infected with L. garvieae have revealed the presence of granulomatous inflammation in the eyes, operculum, caudal peduncle, heart, and brain tissue, with hyaline drop degeneration of renal tubular epithelial cells (Miyazaki, 1980, 1982). In this study, we compared the histopathology of yellowtail artificially infected with L. garvieae and L. formosensis using the above histopathological findings as a control.
Yellowtail, artificially raised in the Wakayama and Ehime prefectures, were reared in tanks at the National Fisheries University using commercial formula feed (Hayashikane Sangyo) and used in the following experiments.
Bacteria strains and growth conditionL. garvieae serotype I and L. formosensis strains were isolated from diseased fish at farms in Ehime Prefecture between 2017–2021. To confirm the serotypes, a slide agglutination test using anti-rabbit serum (Japan Fisheries Resource Conservation Association) and PCR-mediated identification were performed, as previously described (Ohbayashi et al., 2017). PCR amplification was performed using the set of PCR primers LGD-F and LGD-R. The predicted sizes of the PCR-amplified products were 258 bp for L. garvieae serotype I and 1,258 bp for L. formosensis; the L. garvieae Lg-Aq-33 was used as the serotype I strain and the Lg Aq 55 and Lg Aq 103 were used as L. formosensis strains. All bacterial strains were cultivated on brain heart infusion (BHI) agar medium (Eiken Chemical) at 25°C for 24 h.
Bacterial suspensions were prepared by suspending the bacterial colonies in sterile physiological saline (0.85% sodium chloride solution) to a turbidity of OD600. The turbidity of the bacterial suspension was measured using a spectrophotometer, and the concentrations were calculated using the turbidimetric method.
Experimental infectionTwo experimental infection trials were conducted with L. garvieae and L. formosensis using injection and immersion methods. In the first experimental infection, fish from the Wakayama Prefecture (four experimental groups; n = 40; mean weight 113.0 g) were challenged with Lg-Aq-33 (L. garvieae serotype I) and Lg-Aq-55 (L. formosensis). In the second experimental infection, fish from the Ehime Prefecture (four experimental groups; n = 40; mean weight, 197.1 g) were similarly infected with Lg-Aq-33 and Lg-Aq-103 (L. formosensis). For habituation, 10 fish were reared in separate 500 L round tank with an external filtration system and maintained at 25°C for 1 week prior to experiments. During the experimental period, fish were fed the formula based on 2% of their body weight.
In the injection method, experimental fish were anesthetized with quinaldine (2-methylquinoline; final concentration, 20 ppm), weighted, and inoculated intraperitoneally with 100 μL/100 g body weight of the bacterial suspension (1.0 × 105 cfu/mL) using a syringe (1 mL, TERUMO) and needle (26-gauge, 0.5 inch, NIPRO).
In the immersion method, 10 fish were kept in a 50 L round tank (40 L of seawater) with 4 mL bacterial suspension (1.0 × 1010 cfu/mL; final bacterial concentration, 1.0 × 106 cfu/mL). The fish were then immersed in the solution for 1 h.
Post infection, the fish were returned to the same tank and followed up for 14 days. Moribund and dead fish were collected, and cumulative mortalities were calculated. Moribund fish were those that had capsized or were lying on their sides with only the operculum moving and were certain to die within a few hours; therefore, the fish were considered dead and included in the cumulative mortality. The collected fish were observed for gross features and were anesthetized if necessary. Fish collected within 1 h of death or in a moribund state were used as histopathological specimens (3–5 fish from each experimental group). At autopsy, re-isolation was attempted from the kidneys by streaking a platinum loop on a BHI agar plate, and the isolated bacteria were confirmed using the antisera. In preliminary experiments, a saline solution control group was established, confirming the absence of dead fish and any histopathological changes.
HistopathologyPieces of the lateral muscle, gill, brain, heart and intestinal organs, including the liver, kidney, spleen, intestine, and peritonea of the diseased fish were fixed in Bouin’s solution and processed for histopathological examination. Paraffin-embedded tissue sections (2–4 μm) were stained with Mayer’s hematoxylin and eosin, azan stain, Berlin blue, and May–Grünwald Giemsa.
Statistical analysisData were analyzed using Fisher’s exact test. P < 0.05 was used to indicate statistical significance.
The cumulative mortality of the two experimental infections is shown in Fig. 1. In the first experiment, the cumulative mortality of the L. garvieae injected group was 90%, with one fish (five fish in total) sampled daily from 2–6 days post infection (dpi); in the L. garvieae immersed group, the cumulative mortality was 50%, with one fish sampled at 3, 6, 9, and 10 dpi (four fish in total); in the L. formosensis injected group, the cumulative mortality was 80%, with one fish sampled daily from 3–5 dpi and two fish sampled at 6 dpi (five fish in total); in the L. formosensis immersed group, the cumulative mortality rate was 40%, with one fish sampled at 7, 10, and 11 dpi (three fish in total).
In the second experiment, all fish in the L. garvieae injected group died by 6 dpi, with one fish sampled at 2, 4, and 5 dpi and two at 3 dpi (five fish in total). In the L. garvieae immersed group, the cumulative mortality was 60%, and one fish was sampled daily from 6–9 dpi (four fish in total). In the L. formosensis injected group, the cumulative mortality was 70%, with one fish sampled at 3, 4, and 8 dpi and two at 6 dpi (five fish in total). In the L. formosensis immersed group, cumulative mortality was 40%, with one fish sampled at 4 dpi and two at 8 dpi (three fish in total).
Colonies of the same serotype and color as the inoculated bacteria were isolated from the kidneys of all moribund and dead fish and aggregated with the corresponding antiserum.
Gross and anatomical featuresIn both experiments, the sampled fish exhibited typical lactococcosis symptoms, including skin darkening, inner operculum inflammation (Fig. 2A), liver congestion (Fig. 2B), encephalitis (Fig. 2C), and epicarditis (Fig. 2D). After 5 dpi, the most frequently observed symptoms included caudal peduncle ulceration (Fig. 2E), hyperemia and opacity of the eyeballs (Fig. 2F), and hemorrhage of the ventral, pectoral, and caudal fins. No differences in gross and anatomical features were observed between L. garvieae and L. formosensis or between the strains of L. formosensis.
Major histopathological changes in the sampled fish are presented in Table 1. Two experimental infections were performed with different fish origins and strains of L. formosensis; however, the results were the same and no histopathological differences were observed. No significant differences in the histopathological features were observed between the injection and immersion methods, although the sampling point timing was different. Based on the above, the experimental groups with L. garvieae serotype I and L. formosensis in both methods were collectively referred to as the Lg (18 fish) and Lf (16 fish) groups, respectively, regardless of the experiment or infection method, with the histopathological features of each organ described below.
| Sn | Sp (dpi) | Heart | Brain | Spleen | Kidney | Liver | Intestine | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ep | Pa | Me | Pa | ||||||||||||||
| Hp, BI | Gr | Gr | Ve | BI | Gr | BI | CT | BI | Gr | HDD | BI | CS | BI | BI | Gr | ||
| GIj1-1 | 2 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| GIj1-2 | 3 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| GIj1-3 | 4 | + | − | − | − | + | − | − | + | − | − | + | − | + | + | − | − |
| GIj1-4 | 5 | + | + | − | − | + | − | − | + | + | − | − | − | + | − | − | − |
| GIj1-5 | 6 | + | + | + | − | + | + | + | + | + | − | + | + | + | − | − | − |
| GIm1‑1 | 3 | + | − | − | − | + | − | − | + | − | − | + | − | + | − | − | − |
| GIm1‑2 | 6 | + | + | − | − | + | − | − | + | − | − | − | − | + | + | − | − |
| GIm1‑3 | 9 | + | + | − | − | + | + | − | + | + | − | + | − | + | − | − | − |
| GIm1‑4 | 10 | + | − | + | − | + | + | − | + | − | − | + | − | + | − | − | − |
| FIj1-1 | 3 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| FIj1-2 | 4 | + | − | + | − | + | − | − | + | − | − | + | − | + | + | − | − |
| FIj1-3 | 5 | + | − | + | + | + | + | − | + | + | − | + | + | + | − | − | − |
| FIj1-4 | 6 | + | + | + | + | + | + | + | + | + | + | + | + | + | + | + | + |
| FIj1-5 | 6 | + | + | − | − | + | + | − | + | − | − | − | − | + | − | − | − |
| FIm1-1 | 7 | + | + | + | + | + | + | − | + | + | − | − | + | + | − | − | − |
| FIm1-2 | 10 | + | + | + | + | + | + | − | + | + | − | + | + | + | + | − | − |
| FIm1-3 | 11 | + | + | + | + | + | + | + | + | + | − | + | + | + | − | + | − |
Abbreviations: Sn, sample number; Sp, sampling point; dpi, days post infection; Ep, epicardium; Pa, parenchyma; Me, meninges; Hp, hyperplasia; BI, bacterial invasion; Gr, granuloma; Ve, verruca; CT, clarification of trabeculae; HDD, hyaline droplet degeneration; CS, cloudy swelling; GIj, Lactococcus garvieae injected group; GIm, L. garvieae immersed group; FIj, L. formosensis injected group; FIm, L. formosensis immersed group; +, observed; −, not observed.
| Sn | Sp (dpi) | Heart | Brain | Spleen | Kidney | Liver | Intestine | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ep | Pa | Me | Pa | ||||||||||||||
| Hp, BI | Gr | Gr | Ve | BI | Gr | BI | CT | BI | Gr | HDD | BI | CS | BI | BI | Gr | ||
| GIj2-1 | 2 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| GIj2-2 | 3 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| GIj2-3 | 3 | + | − | − | − | + | − | − | + | − | − | + | − | + | + | − | − |
| GIj2-4 | 4 | + | − | − | − | + | − | − | + | − | − | + | − | + | − | − | − |
| GIj2-5 | 5 | + | + | + | + | + | + | − | + | + | − | + | − | + | − | − | − |
| GIm2‑1 | 6 | + | + | − | − | + | + | − | + | − | − | − | − | + | + | − | − |
| GIm2‑2 | 7 | + | − | − | − | + | − | − | + | + | − | + | + | + | − | − | − |
| GIm2‑3 | 8 | + | + | + | + | + | + | − | + | − | − | − | − | + | − | − | − |
| GIm2‑4 | 9 | + | + | − | − | + | + | − | + | + | − | + | − | + | + | − | − |
| FIj2-1 | 3 | + | − | − | − | + | − | − | + | − | − | + | − | + | − | − | − |
| FIj2-2 | 4 | + | − | − | − | + | − | − | + | − | − | − | − | + | − | − | − |
| FIj2-3 | 6 | + | + | + | + | + | + | + | + | + | − | + | + | + | − | + | − |
| FIj2-4 | 6 | + | + | + | − | + | + | − | + | + | − | + | + | + | + | − | − |
| FIj2-5 | 8 | + | + | + | + | + | + | + | + | + | + | + | + | + | + | + | + |
| Fim2-1 | 4 | + | − | + | − | + | − | − | + | − | − | − | − | + | − | − | − |
| Fim2-2 | 8 | + | + | + | + | + | + | − | + | + | − | + | − | + | − | − | − |
| Fim2-3 | 8 | + | + | + | + | + | + | + | + | + | − | + | + | + | + | + | + |
Abbreviations: Sn, sample number; Sp, sampling point; dpi, days post infection; Ep, epicardium; Pa, parenchyma; Me, meninges; Hp, hyperplasia; BI, bacterial invasion; Gr, granuloma; Ve, verruca; CT, clarification of trabeculae; HDD, hyaline droplet degeneration; CS, cloudy swelling; GIj, Lactococcus garvieae injected group; GIm, L. garvieae immersed group; FIj, L. formosensis injected group; FIm, L. formosensis immersed group; +, observed; −, not observed.
In all fish, the adventitia of the bulbus arteriosus and epicardium showed hyperplasia, with increased spongiform connective tissue and infiltration of lymphocytes, neutrophils, and macrophages (Fig. 3A). Bacterial invasion was observed in most fish, accompanied by the aggregation of phagocytic cells (macrophages) and bacterial colony formation due to intracellular proliferation. The adventitia of the bulbus arteriosus and epicardium of fish sampled at 4 dpi exhibited severe hyperplasia, extensive bacterial invasion, and connective tissue hyperplasia with fibrin deposition and adhesion. Furthermore, massive granulomas were extensively observed in the epicardium of most fish sampled after 5 dpi (Fig. 3B). These histopathological features of the adventitia of the bulbus arteriosus and epicardium were shared among the experimental groups. Twelve fish in the Lf groups exhibited large bacterial colonies in the endocardium and at the bases of the atrial and ventricular valves (Fig. 3C). Numerous granulomas were observed in the bulbus arteriosus (Fig. 3D), and apparent bacterial invasion was recognized in the cardiac muscle layer of ventricles, associated with lymphocytes and large mononuclear cells. These mononuclear cells engulfed bacteria in their cytoplasm (Fig. 3E). In nine fish (all sampled after 5 dpi), the valve was replaced with huge verruca that completely obliterated the ventricular lumen (Fig. 3F). These verrucae were composed of numerous bacteria, deposited fibrin, neutrophils, macrophages, and necrotic cells and surrounded by inflammatory exudates, phagocytes, and their residues. In contrast, in the Lg group, there were few bacteria and granulomas in the heart parenchyma (four fish), and verruca was observed in only two fish.
All fish exhibited meninges with bacterial invasion as well as lymphocyte and macrophage infiltration (Fig. 4A and B). The capillaries in the meninges were dilated and the pia mater was thickened, with bacterial invasion and large mononuclear cells (Fig. 4C). After 5 dpi (seven fish in the Lg group and 11 fish in the Lf group, no significant difference), granulomas were observed beneath the surrounding connective tissue of the olfactory nerve cord and ventral pia matter of the medulla oblongata, cerebellum, and mesencephalon (Fig. 4D). Bacteria were observed in the olfactory, optic nerve, and molecular and granular layers of the mesencephalon, and cerebellum in one fish from the Lg group and five fish from the Lf group (Fig. 4E). In particular, the olfactory bulb showed severe bacterial invasion and the infiltration of lymphocytes and phagocytic cells, possibly macrophages, which were necrotic, collapsed, and partially depleted (Fig. 4F).
In fish sampled before 4 dpi, no bacteria were observed in the spleen; however, stored erythrocyte levels were significantly reduced, the splenic pulp was often atrophied, and trabeculae composed of smooth muscle fibers became clearly visible (Fig. 5A). At 5 dpi, in addition to the above lesions, bacteria and phagocytic cells were found in the sheathed artery most commonly, followed by the splenic pulp and vascular lumen (Fig. 5B). Many melanomacrophage centers were observed in some fish (Fig. 5C). These histopathological features were common in both groups; however, in the two fish of the Lf groups (after 6 dpi), extensive cellular necrosis associated with loss of normal tissue, resembling diffuse abscess, was prominent along with many granulomas in addition to the previous lesions, with bacteria and necrotic mononuclear cells observed within the lesions (Fig. 5D).
All fish had numerous melanomacrophage centers in the hematopoietic tissue, with hyaline droplet degeneration of tubular epithelial cells observed in more than half of the cells (Fig. 5E). Bacteria and phagocytes were observed in the renal tissues (serosa, vascular lumen, glomeruli, and hematopoietic tissue) of more than half the fish in the Lf group (Fig. 5F), but only two in the Lg group.
LiverNo differences in liver histopathological features were observed between the experimental groups. All fish had hepatocytes with cloudy swelling (Fig. 6A), with some showing congestion, serosal hypertrophy, and bacterial invasion (blood vessels, sinusoids, and serosa). Within the hypertrophic serosa, spongiform connective tissue was hyperplastic and filled with phagocytosed bacterial colonies (Fig. 6B). Necrotized hepatocytes, bacterial invasion, and macrophage infiltration were observed in some fish (Fig. 6C).
In the Lg group, no bacterial invasion of the intestinal tract or histopathological change was observed; however, in the Lf group, bacterial colonies were observed in the intestinal muscularis (inner circular and outer longitudinal layers) and submucosa of five fish. Bacteria were also observed in the intestinal villi (Fig. 6D). Granulomas were observed in the intestinal muscularis and had invaded the submucosa of three fish sampled at 6 dpi (Fig. 6E).
Intraperitoneal cavityAlthough there were no differences between the Lg and Lf groups, slight differences were observed between the injection and immersion methods. In the injection method, fish sampled up to 5 dpi showed numerous bacteria and infiltration of lymphocytes, neutrophils, and macrophages in the peritonea (Fig. 6F), whereas fish sampled after 6 dpi showed a significant decrease in bacteria, although leukocyte infiltration was observed. In the immersion method, all the fish showed mild bacterial and leukocyte infiltration in the peritonea.
OthersNo histopathological changes were observed in the gills and muscles of both groups.
In the present study, the histopathology of yellowtail that were experimentally infected with L. formosensis (Lf) was observed in comparison with those experimentally infected with L. garvieae (Lg). Experimental infections were performed in duplicate using injection and immersion methods.
The cumulative mortality trend after experimental infection in each experimental group did not differ between the experiments with the inoculum bacteria re-isolated from all sampled fish. Gross and anatomical features showed typical lesions of the disease, such as inflammation of the inner operculum, ulceration of the caudal peduncle, epicarditis, and encephalitis, in all experimental groups. Although the time of death differed between the injection and immersion methods, there was no difference in the symptoms of moribund or dead fish, indicating that infections in yellowtail caused by either bacteria could not be distinguished by gross and anatomical features.
Histopathological observations confirmed the previously reported lesions of lactococcosis but also revealed new lesions of the disease and characteristics of L. formosensis infection.
In the heart, the typical histopathological features of lactococcosis were observed in all experimental groups. The adventitia of the bulbus arteriosus and epicardium was hyperplastic, with bacterial invasion, fibrotic deposition, and leukocyte infiltration (lymphocytes, neutrophils, and macrophages), which ultimately led to granuloma formation. In addition to previously reported lesions, several fish in the Lg and Lf groups showed bacterial colonies and granulomas within the cardiac parenchyma, including the myocardial layer, heart valves, and endocardium. In this study, more than half of the fish in the Lf group exhibited large verruca, a characteristic of verrucous endocarditis, on the ventricular valves. These findings were also observed in several fish in the Lg group; however, in the Lf group, they were more intensively represented as cardiac lesions. Miyazaki (1980, 1982) reported that no histopathological features were observed in the hearts of yellowtails artificially infected with L. garvieae, except for bacterial pericarditis; however, it is possible to speculate that some cardiac histopathological features, such as those found in the present study, could have been overlooked in previous studies because of their low incidence. Infective endocarditis in mammals is a disease wherein the verruca formed on heart valves is caused by Streptococcus spp. (Matsuo, 1960; Fefer et al., 1998; Jensen et al., 2010; Navas et al., 2013; Abe et al., 2017). In this disease, bacterial infection scars the valve leaflet, reducing flexibility and durability, which leads to verruca formation as bacteria adhere and multiply (Matsuo, 1960; Umezu, 1996). This eventually leads to bacterial embolism and heart failure due to gigantism of the verrucae. The verruca observed in this study was attributed to the growth and enlargement of bacterial masses and granulomas, many of which were found in the ventricular valves. The growth of these verrucae filled the heart lumen, suggesting that the diseased fish had a bacterial embolism leading to circulatory disturbances. In fish, endocarditis is caused by Streptococcus spp. or parasitic infections (Chen et al., 2007; Hagiwara et al., 2009, 2010; Warren et al., 2017), with granulomatous inflammation of the endocardium and small verruca on the valves and their bases. However, no case of a huge verruca, as observed in this study, had been observed previously, suggesting that this is a characteristic symptom of L. formosensis infection but rare in fish.
The meninges of all experimental groups developed meningitis, as numerous bacterial, lymphocytic, and macrophage infiltrates were observed. Over time, these lesions were replaced by granulomas that developed into granulomatous inflammation. However, within the brain parenchyma, with the exception of the olfactory bulb, only small amounts of bacteria were observed in the mesencephalon and cerebellum. Based on these results, the main histopathological feature of the brain in yellowtail lactococcosis was meningitis. Severe lesions such as bacterial invasion and necrosis were seen only in the olfactory bulb, and lesions were also observed mainly in the olfactory bulb of greater amberjack S. dumerili injected intraperitoneally with S. dysgalactiae (Hagiwara et al., 2009). In fish, lactococcosis and streptococcosis caused by L. formosensis and S. dysgalactiae, respectively, growing in the cerebral, spinal fluid was suggested to invade the brain tissue from the olfactory bulb. The histopathological features of the brain parenchyma and meninges in this study reflected those reported by Miyazaki (1982); although not significantly different, the number of fish with granulomatous inflammation in the meninges was higher in the Lf group than in the Lg group. L. formosensis infections might be more likely to cause granulomatous inflammation of the brain than L. garvieae.
In all experimental groups, the spleen exhibited reduced splenic pulp and clarification of trabeculae composed of smooth muscle fibers and the kidney showed hyaline droplet degeneration of tubular epithelial cells and cloudy swelling of hepatocytes, congruent with the report by Miyazaki (1980, 1982). Bacterial invasion of tissues was observed in both experimental groups, which may be an overlooked histopathological feature of the disease because bacterial invasion has not been previously reported. Bacteria were particularly abundant in the sheathed arteries of the spleen. As sheathed arteries trap and filter foreign material (Suzuki and Suetake, 2013), the observed bacteria and phagocytes may have been trapped as foreign matter. Two fish in the Lf group had granulomas in the spleen, suggesting a propensity of the Lf group to form granulomas in the spleen. In the intestines, gills, and muscles, no histopathological changes were observed in Lg, as previously reported (Miyazaki, 1980, 1982). However, in the Lf group, no changes were observed in the gills and muscles, whereas bacterial invasion and granulomas were observed in the intestine. These intestinal lesions may be characteristic of L. formosensis infection. Although the intraperitoneal cavity has never been investigated, this study found that most injected bacteria were phagocytosed, processed by 6 dpi. Immersion has also been shown to induce bacterial invasion of the intraperitoneal cavity and leukocyte infiltration. In this study, the time of death differed between the injection and immersion groups; however, there were no histopathological differences, except for intraperitoneal cavity lesions. In general, the immersion method reproduces natural infections better than the injection method and is considered the ideal method for efficacy trials of agents and vaccines. However, the injection method may also be a practical method for histopathological evaluation as it infects more reliably and achieves higher mortality rates than immersion.
This histopathological study revealed that, in yellowtail lactococcosis caused by L. formosensis, the heart, brain, and other internal organs undergo granulomatous inflammation with a high incidence of verrucous endocarditis. In lactococcosis, the causes of death are congestive heart failure and circulatory disturbances due to hyperplasia of the adventitia of the bulbus arteriosus and epicardium (Miyazaki, 1982). In this study, bacterial embolization was determined to be an additional cause of death, as a huge verruca completely filling the lumen of the heart was found in both bacterial groups. Especially, this is a major cause of death in cases of L. formosensis infection. The two bacteria were experimentally infected at the same concentrations; however, in all experimental groups, the Lg group began to die earlier and had a higher mortality rate without developing granulomatous inflammation or endocarditis. The infected fish caused by L. garvieae possibly died due to congestive heart failure and circulatory disturbances before developing granulomatous inflammation or endocarditis. Based on these results, L. garvieae was speculated to be more pathogenic to yellowtail than L. formosensis in the strains used in this study; however, the cumulative mortalities were not significantly different. The pathogenicity of both bacteria should be the subject of future research.
We thank the Ehime Fisheries Research Center for donating experimental bacteria.
