Vibrio scophthalmi is considered to be an opportunistic pathogen of flat fish, primarily infecting olive flounder and turbot (Wang et al., 2004; Jo et al., 2006; Qiao et al., 2012a). Olive flounder become more sensitive to infection by V. scophthalmi under stressful conditions (Qiao et al., 2012a). The virulence factors of V. scophthalmi have been well studied. Qiao et al. (2012b) investigated the pathogenicity of V. scophthalmi using high and low virulence strains (HVS and LVS, respectively), and suggested that both strains produced a slime layer and biofilm. The level of biofilm production is positively associated with the pathogenicity of V. scophthalmi and susceptibility to antibiotics. Superoxide dismutase (SOD) activity of the HVS was higher than the LVS. Extracellular products (ECP) of the HVS showed higher pathogenicity to olive flounder as compared with the LVS. ECP demonstrated naphtol-AS-BI-phosphohydrolase, lipase, gelatinase, and leucine arylamidase activity. The extracellular O₂^- overflow and intracellular O₂^- concentration of macrophages induced by HVS were lower than those induced by LVS. Nitric oxide production was significantly higher in the HVS than in the LVS (Qiao, 2011).

Generally, variations in hematological parameters are related to the fish species, age, sexual maturity, and health conditions (Blaxhall, 1972; Hrubec et al., 2000; Jamalzadeh et al., 2009). Normal ranges for various hematological parameters in fish have been established by different studies of fish physiology and pathology (Blaxhall, 1972; Ram Bhaskar and Srinivasa Rao, 1989; Hrubec et al., 2000; Martins et al., 2008). Hematological parameters can be affected by bacterial infection (Martins et al., 2008), parasitic infection (Ghiraldelli et al., 2006), and poor water quality (Ram Bhaskar and Srinivasa Rao, 1989). Hematological tests and the analysis of serum constituents provide crucial information for monitoring the health of fish (Aldrin et al., 1982). Specifically, these approaches provide reliable information on metabolic disorders, deficiencies, and health status before they become disease problems in cultivated fish. Regular monitoring of the hematological parameters of farmed fish can prevent losses due to fish diseases in aquaculture (Ram Bhaskar and Srinivasa Rao, 1989). Currently, there is no information available regarding hematological changes in olive flounder following V. scophthalmi infection. In this study, a hematological analysis of olive flounder following V. scophthalmi infection was performed, including determination of hematocrit (Hct), hemoglobin (Hb), total protein (TP), albumin (Alb), globulin (Glo), albumin:globulin (A:G) ratio, aspatate aminotransferase (AST), alanine aminotransferase (ALT), glucose concentration, superoxide dismutase (SOD), lysozyme, and bacteriolytic activities of serum. In addition, the survival of V. scophthalmi in serum, skin mucus, and macrophages was investigated in vitro.

Two strains of V. scophthalmi (HVS A19008 and LVS A19010) with different levels of virulence were used in this study. The HVS A19008 was isolated from a kidney, and LVS A19010 was obtained from a spleen, of a diseased olive flounder at the Jeju and Pohang farms, respectively. The HVS A19008 was highly virulent and LVS A19010 had a relatively low virulence to olive flounder (Qiao et al., 2012a). These strains were preserved at -80℃ at the Fish Disease and Prevention Laboratory, Pukyong National University, Korea. Prior to experiments, bacteria were routinely grown in tryptic soy broth (TSB) or tryptic soy agar supplemented with 2% NaCl (ST) at 27℃ for 24 h and pathogenicity was checked according to challenge tests (Qiao et al., 2012a).

Normal olive flounder, with an average body weight of 47.28 g and body length of 17.01 cm, were obtained from a farm in Geoje Island and maintained in an aquaria at approximately 30 practical salinity units (psu), 18-20℃ and pH 8.0 (±0.2) for two weeks prior to the experiment. The fish were fed with commercial fish pellets (National Federation of Fisheries Cooperatives Feed, Korea) until they were used for the experiment.

Pathogenicity of these two strains was determined in vivo (Nieto et al., 1984). Bacterial suspensions were prepared by culturing the strains in TSB supplemented with 2% NaCl at 25℃ for 24 h, washing with sterilized physiological saline (PS), and then adjusting to the appropriate concentration of 10⁷ (CFU)/mL with PS. Each fish was injected intraperitoneally with 0.2 mL of bacterial suspension. Fish in the control group were injected with 0.2 mL of PS. The fish were observed daily for 14 days post bacterial challenge. All mortalities and clinical signs were recorded daily. Moribund or freshly dead fish were collected to isolate the inoculated bacteria and confirm the causative pathogen. Three fish in each group were sampled at 0, 1, 3, 9, 12, 24, 72, and 168 hours post infection (hpi) for hematological tests.

Hematological examination

The fish were sacrificed with a lethal dose of benzocaine anesthetic (CAS, Canada). Fish blood was drawn from the caudal vein with a syringe treated with an anticoagulant. Samples were then centrifuged at 300 g for 10 min at 4℃ to obtain the plasma for hematological measurements. Hb concentration was measured by a blood Hb kit (Asan Pharm Co., Seoul, Korea) and the Hct value was determined using a microcapillary reader at 1 h. The samples were kept at -70℃ until use and no samples were frozen more than once.

Biochemical examination

Peripheral blood was drawn with a syringe without anticoagulant and allowed to clot for 1 h at room temperature and 2 h at 4℃. After centrifugation at 300 g for 10 min at 4℃, serum was collected for measurement of biochemical parameters. The biochemical tests were conducted using assay kits (Asan Pharm Co., Korea) for TP, Alb, Glo, A:G ratio, AST, ALT, SOD and glucose concentration. Lysozyme activity was assayed using the turbidimetric method. The turbidimetric assay uses the lysis of Micrococcus lysodeikticus (ATCC 4698; Sigma, St. Louis, MO, USA) to determine the lysozyme activity and egg-white lysozyme (LYS 702.5; Bioshop Canada Inc., Burlington, ON, Canada) as a standard (Sankaran and Gurnani, 1972). Briefly, 100 μL of phosphate buffered saline (PBS, pH 6.2) was added to each well of a 96-well plate, and 100 μL of serum was added to the first well of each row. Twofold serial dilutions were then made in the remaining wells with PBS. One hundred microliters of 0.4 mg/mL M. lysodeikticus in PBS (pH 6.2) was added to each well. The enzyme kinetics were measured with a microtiter plate reader at 590 nm every 10 min for 40 min and the last measurement was conducted at 60 min. Lysozyme activity was recorded as 1 unit, which caused a decrease in absorbance of 0.001 per min. The bacteriolytic test followed the method of Pruzanski (1973) with Escherichia coli JM089 (ATCC 29522) was chosen to monitor the bactericidal activity in the serum of normal and challenged olive flounder individuals. E. coli JM089 was grown in 200 mL of Luria Bertani (LB) medium for 20 h at 37℃ with shaking at 200 rpm. Bacterial concentration was adjusted to an optical reading of 0.5 at 540 nm and added to the serum solution at a 1:1 ratio (bacterial suspension:serum dilution). Sterilized LB medium was added to the serum solution as a negative control. The mixture was placed on an orbital incubator for 1 h at 25℃ and 200 rpm. Results were recorded as an increase of the absorbance.

Serum and skin mucus of olive flounder were prepared following the techniques described by Bordas et al. (1996). The bacterial cells were suspended in fresh serum or skin mucus (1:1, v:v) and adjusted to 1 × 10⁶ CFU/mL bacterial cells. One hundred microlitres of each sample was removed at 0, 1, 3, 6, and 18 h after incubation at 25℃ and 10-fold serial dilutions of PBS were spread on ST plates. The survival rate of strains in serum or skin mucus was defined as the number of viable bacteria after the co-culture was divided by an initial bacterial count as described by Leung et al. (1994).

Macrophage monolayer preparation

Monolayers of head kidney macrophages from olive flounder individuals were prepared as described by Secombes (1990) with slight modifications. The head kidney was removed, ground, and filtered through a sterile nylon mesh with L-15 medium (Gibco, Grand Island, NY, USA) containing 2% fetal bovine serum (FBS, Gibco), 100 IU/mL penicillin, 100 μg/mL streptomycin, and 10 U/mL heparin (Sigma). The cell suspension was layered onto a 34-51% Percoll density gradient in siliconized tubes on ice and centrifuged at 400 g for 30 min at 4℃. The band of macrophages lying above the Percoll interface were collected and washed with L-15 medium containing 0.1% FBS. The cell pellet was resuspended in L-15 medium supplemented with 0.1% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin. The viability of cells was determined by staining with 0.5% trypan blue (Hudson and Hay, 1989). The macrophages could be used if the survival rate was more than 95% after separation. The non-adherent cells were washed off after 3 to 5 h at 18℃ and the remaining monolayers were maintained with L-15 medium (containing 5% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin) and incubated at 18℃ for 1 to 3 days before use.

Intracellular survival of V. scophthalmi in macrophages

V. scophthalmi was added to the plate with macrophage monolayers at a 10:1 ratio (bacteria cell number:macrophage number). The plates were centrifuged at 50 × g for 5 min to allow the bacteria to come in contact with the cells. Bacteria were removed and 200 μL of fresh L-15 medium containing 3 μg/mL of norfloxacin was then added. The infected cell monolayers were incubated for two more hours to completely kill residual extracellular and attached bacteria. The monolayers were then resuspended with L-15 medium containing 5% FBS and 3 μg/mL norfloxacin to prevent the growth of bacteria released from ruptured and infected cells. Plates were incubated at 20℃ and cells were lysed at 0, 3, 6, 12, and 18 h with 0.1% Triton X-100. The intracellular bacteria released were quantified by plating serial dilutions of the lysate. The numbers of CFU and the percentage invasion for each strain at different incubation times was calculated. The L-15 medium supplemented with 3 μg/mL norfloxacin without cells served as a media control.

All values for the hematological and biochemical parameters were expressed as a mean ± standard deviation (SD). Hematological parameters in fish of different treatment groups were analyzed by analysis of variance (ANOVA) using the SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). P-values of 0.05 or less were considered to be statistically significant.

Clinical signs in olive flounder were similar between fish infected both experimentally and naturally and included darkening skin, distended abdomen, hemorrhages of muscles, liver, and intestine, ascites, and hypertrophy of the spleen and kidney. The inoculated strains could be reisolated from the liver, spleen, kidney, and ascites of dead individuals.

Hematological analysis

The effects of V. scophthalmi infection on hematological parameters in olive flounder are shown in Table 1. The Hct value in fish infected by both the HVS and LVS decreased gradually post infection and was significantly different when

compared to fish in the control group. In fish infected with the HVS, the Hct value declined to 18.0% compared to the control group (23.5%) at 168 hpi. The Hb concentration increased to 2.95 g/dL at 9 hpi, and decreased to 2.80 g/dL at 168 hpi which was similar to the value at 0 h.

Biochemical examination

Serum biochemical values are shown in Table 1. The glucose concentration decreased significantly in the HVS and LVS infection groups compared to the control group from 1 to 24 hpi. TP concentration in the HVS infection group was significantly higher than in the LVS and control groups from 3 to 24 hpi. TP concentration increased to 5.86 g/dL at 12 hpi, as compared with 4.7 g/dL that in the control fish. The A:G ratio in the HVS infection group decreased significantly after 3 hpi and was significantly lower when compared to the control fish over the period from 3 to 168 hpi. Fish infected with the HVS had an A:G ratio of 0.69 at 12 hpi, significantly lower than the ratio of 1.03 A:G recorded in the control fish. The AST value and SOD activity in fish infected with the HVS also declined significantly after 12 hpi compared to those in the control group. ALT activity was not significantly different between infected and control groups. Lysozyme activity was significantly different between infected and control groups. Lysozyme activity in the HVS and LVS infected groups was

enhanced significantly from 1 to 12 hpi and then decreased from 24 to 168 hpi compared to the control fish.

Bactericidal activity of serum

As shown in Table 1, bactericidal activity of serum in the HVS and LVS infected groups increased over time post infection and reached the highest values (2.3-2.6) at 24 hpi, which was significantly higher than the corresponding value in the control fish (1.6).

The HVS could survive in fresh serum and cell numbers increased by approximately 2 log units at 18 hpi (Fig. 1). The LVS showed serum sensitivity and cell numbers decreased below the detection limit. The HVS was able to survive and replicate in skin mucus with cell numbers increasing from 7.33 × 10⁵ to 1.59 × 10⁶ CFU/mL, whereas the LVS maintained cell numbers at the original levels in skin mucus (Fig. 2).

As shown in Table 2, the HVS was internalized in macrophages and cell numbers increased significantly from 0% to 197.2%, 498.6%, 878.6%, and 1,462.7% after incubation for 3, 6, 12, and 18 h, respectively, despite the presence of norfloxacin in the media. In contrast, cell numbers in media without macrophages, but with norfloxacin, declined by 20.8% and 89.6% at 12 and 18 h, respectively, when compared to cells at 0 h. For the LVS, cell numbers increased by 208.6%, 477.4%, 762.3%, and 1042.3% after incubation for 3, 6, 12, and 18 h, respectively. Cell numbers of the LVS in medium with norfloxacin without macrophages declined 41.4% and 94.4% at 12 and 18 h, respectively, when compared to time 0 h. The HVS could survive and multiply in macrophages better than the LVS.

Hematological characteristics are important for fish and can be used as an effective and sensitive index to monitor physiological and pathological conditions (Kori-Siakpere et al., 2005). In this study, TP concentration in the serum of olive flounder in the infected groups was significantly higher than that of control fish, which was consistent with observations made by Lee (2005). Lee (2005) found that TP concentration in fish injected with different concentrations of ECP extracted from Edwardsiella tarda, increased significantly until 168 hpi. Whereas Alb and the A:G ratio were significantly reduced in infected groups compared to a control group. Alb decreased as it is a negative acute phase protein and decreases in concentration following an inflammatory stimulus. The slight increase of glucose concentration in the infected groups was in accordance with that of olive flounder individuals infected by E. tarda ECP (Lee, 2005). The increase in glucose might be due to the increased rate of lipolysis as an alternative energy source during the infection (Ghanem and Abdel-Hamid, 2010) or the autoimmune response in the host induced by bacterial infection similar to a virus infection in mammals (Clark, 2003; Ghanem and Abdel-Hamid, 2010).

AST and ALT are jointly known as transaminases and are associated with inflammation or liver injury. AST is also found in many other organs in addition to the liver, including kidneys, muscles, and the heart. ALT is found primarily in the liver and high levels of ALT are almost always indicative of liver problems. In this study, ALT was observed to increase, indicating liver damage. Moreover, the ALT:AST ratio may also provide useful information regarding the extent and cause of liver disease, because most liver diseases are characterized by greater elevations of ALT than AST.

Lysozyme is a cationic enzyme that breaks down β-1, 4 glycosidic acids and N-acetyl-glucosamine in the peptidoglucan of bacterial cell walls. It plays an important role in the bio-defense system against Gram-positive and Gram-negative bacteria (Alexander and Ingram, 1992). In this study, there were significant changes in the lysozyme activity of infected fish compared to the control fish. The infected fish were significantly stimulated before 12 hpi and lysozyme activity reached a high level within a short period of time after infection. The bactericidal activity in infected fish also reached a high level before 72 hpi, suggesting that lysozyme levels could be correlated with phagocytic activity. The bactericidal activity of serum has been well recognized as one of the key mechanisms for clearing bacteria from fish (Ellis, 2001). In this study, the bactericidal activity in infected fish could not be maintained at a high level near the end of the experiment (168 hpi), probably due to the reduction in lysozyme secretion. Robertsen et al. (1990) demonstrated an enhanced protection against fish bacterial infection and confirmed a correlation with an increase in serum lysozyme, phagocytic, and bactericidal activity in head kidney phagocytes.

The HVS was able to survive and replicate in the serum of olive flounder individuals, whereas the LVS was unable to survive. This suggests that the survival and reproduction of V. scophthalmi in host serum plays a major role in systemic infection. According to Han et al. (2006), all virulent strains of E. tarda were able to survive and proliferate in serum, whereas the avirulent strain was susceptible to serum bactericidal activity. Similarly, survival ability in fish mucus or serum has been used as an indicator of virulence in V. harveyi, A. hydrophila, and Flavobacterium psychrophilum (Leung et al. 1994; Wiklund and Dalsgaard, 2002; Won and Park, 2008).

The results regarding the number of intracellular viable cells of bacterial strains revealed that both the HVS and LVS were capable of invasion and replication within macrophages in vitro and no significant difference in the bacterial number ingested per macrophage was observed between the HVS and LVS. Booth et al. (2006) noted that the bacterial number of the virulent E. ictaluri increased 2.6, 5.1, and 7.1 fold after 4, 8, and 12 h of incubation within channel catfish macrophages, respectively, suggesting E. ictaluri can survive and replicate within macrophages. The replication of the HVS within macrophages in vitro was faster than the replication of LVS, which might be related to the stronger resistance of HVS against macrophage-mediated killing through higher SOD activity. Therefore, the ability to survive and replicate within macrophages is a virulence factor in pathogenic bacteria.

In summary, this study evaluated hematological and biochemical alterations for olive flounder infected by V. scophthalmi and compared the survival between the HVS and LVS of V. scophthalmi in serum, skin mucus, and macrophages of olive flounder individuals. The results clearly demonstrated a significant difference in hematological characteristics between fish infected by V. scophthalmi and uninfected control fish, and between fish infected by two V. scophthalmi strains of different virulence (HVS and LVS). These changes in hematological characteristics in fish infected by V. scophthalmi could be used as an effective index to monitor the physiological and pathological conditions of fish. The survival and reproduction of V. scophthalmi in host serum, skin mucus, and macrophages could also serve as a virulence indicator for different V. scophthalmi strains.