The E-11 cell line (ECACC Cat. No. 01110916; European Collection of Cell Cultures, UK), developed by Iwamoto et al. [19], was cultured at 25℃ in Leibovitz (L-15) medium (Invitrogen, USA) supplemented with 1% penicillin-streptomycin (Invitrogen, USA) and 5% fetal bovine serum (FBS; Lonza, Swiss) until semi-confluent prior to virus inoculation. An RGNNV isolate (ERV378/102-5/04), kindly provided by Dr. G. Bovo (Instituto Zooprofilattico Sperimentale delle Venezie, Italy) was propagated following standard protocols at 25℃ in the E-11 cell line. Inoculated cells were maintained in L-15 medium supplemented with 1% penicillin-streptomycin and 2% FBS. Viral titers were determined according to the final dilution causing cytopathic effects in the 50% of the inoculated wells (TCID₅₀ method) [12].

Juvenile European seabass (average weight, 10 g; age, 6~7 months) from an aquaculture centre located in the South-West of the Iberian Peninsula were distributed into three tanks: 1) control non-infected animals (n = 250), 2) infected animals used for sampling (n = 250), and 3) infected animals used for cumulative mortality determination (n = 100, no sampling was conducted for this group). Water temperature was maintained between 21 and 24℃ over the course of the trial. Fish in the challenge and control groups were intramuscularly injected with RGNNV (10⁶ TCID₅₀/fish) or L-15 medium (0.1 mL), respectively. Mortality was recorded daily and the dead fish were frozen at -80℃ until virological testing. To confirm that the virus was the causal infectious agent, virological test were performed, involving the inoculation of pooled brain and eye homogenates of dead fish on E-11 cells, and the subsequent confirmation for the RGNNV presence by a previously reported RT-PCR [23]. This RT-PCR protocol amplifies a 292-pb fragment within the T4 region, located in the RNA2 segment, using the primers Noda-Fwr1 and Noda-Rev2 [23]. Tissue homogenization was performed in L-15 medium (20%, w/v) using a T10 basic Ultra-Turrax (IKA, Germany) disperser following the procedure previously described [1].

Two apparently healthy fish from the challenged (tank No. 2) and control (tank No. 1) groups were randomly collected and sacrificed, with an overdose of anesthetic (MS-222; Sigma, USA), at 3, 10, 17, 24, and 31 days post-infection (PI) and 2 months PI. These animals were dissected, and the brain, eyes, liver, spleen, kidney, and caudal fin were fixed in 4% buffered formalin (Merck, Germany) at room temperature for 24 h. The organs were then washed with distilled water for 3 h, and maintained in 70% ethanol at room temperature until use.

In addition, blood samples were collected 24 days and 2 months PI from the caudal vein of challenged and control fish (15 animals per time and group). Samples from five animals were pooled in order to obtain sufficient volume for anti-RGNNV antibody quantification. After clotting at 4℃ overnight, sera were collected by two consecutive rounds of centrifugation (400 × g at 4℃) for 15 min, and stored at -20℃ until the ELISA was performed to measure anti-RGNNV antibody levels. Three samples, each containing sera from five fish, were analyzed at each sampling time.

Organs were processed with the Shandon Excelsior ES Tissue Processor (Thermo Scientific, USA) according to the manufacturer's instructions, embedded into paraffin blocks using the Leica Histoembedder (Leica Microsystems, Germany), and 5-µm tissue sections were mounted onto slides. Sectioning was performed using the microtome RM2035 (Leica Microsystems, Germany) and Leica Disposable Microtome Blades (Leica Microsystems, Germany). The blade was carefully disinfected with ethanol between each sample. After deparaffinization and rehydrating in a graded series of ethanol, endogenous peroxidase activity was blocked by incubating at 37℃ for 1 h with a solution composed of 0.36% (w/v) β-D-glucose, 0.01% (w/v) glucose oxidase, and 0.0013% (w/v) sodium azide in PBS. Afterwards, the sections were rinsed twice with PBS for 3 min, and non-specific binding sites were blocked by incubating with normal goat serum at 1 : 10 dilution in PBS (Sigma, USA) for 30 min at room temperature (ca. 20℃). The sections were then incubated at room temperature with an anti-RGNNV monoclonal antibody (1 : 5 dilution in PBS) for 1 h (P09; Aquatic Diagnostics Ltd, UK), and the resulting antigen-antibody complex was detected using a 1 : 100 dilution of a goat anti-mouse IgG-biotin conjugate (Sigma, USA) for 1 h at room temperature (ca. 20℃). For colorimetric detection, the tissues were incubated at room temperature with 1 : 100 dilution of streptavidin-horseradish peroxidase (HRP; Sigma, USA) in PBS for 1 h. The reaction was visualized using a Vector VIP substrate kit for peroxidase (Vector Laboratories, USA). The sections were finally counter-stained with methyl green (Vector laboratories, USA) at 60℃ for 3 min, dehydrated, and mounted with Pertex mounting medium (Leica Microsystems, Germany). Antibody labeling appeared as dark purple staining in the target cells. Staining specificity in every organ analyzed was evaluated by performing IHC without the primary antibody.

Paraffin-embedded organs from challenged animals were examined for histological changes and compared to organs from non-exposed fish. After deparaffinization and rehydration, 5-µm tissue sections were stained with hematoxylin-eosin following standard histology protocols [32]. Sections were observed using an optical microscope (E800; Nikon, Japan).

Quantification of anti-RGNNV antibodies was conducted by an indirect ELISA using 96-well polystyrene plates (Immulon 4HBX; Thermo Scientific, USA) coated with 0.4 mg of lysates from RGNNV-infected and non-infected (control) E-11 cells as previously described by Grove et al. [15]. Negative controls (PBS instead of fish sera) and positive controls (anti-RGNNV monoclonal antibodies instead of fish sera) were included in each ELISA plate. After coating, the plates were washed three times with 1× low salt wash (LSW) buffer (2.4 g/L Trizma base, 22.2 g/L NaCl, 0.1 g/L merthiolate, and 0.5 mL/L Tween 20), and non-specific binding sites were blocked with 3% skimmed powder milk (Sigma, USA) in 1× LSW buffer during 2 h at room temperature (ca. 20℃). Afterwards, the plates were washed three times with 1× LSW buffer, and 100 µL of double-diluted tested pooled sera (from 1 : 32 to 1 : 262,144, in PBS) were added. Pooled samples were analyzed in triplicate. An overnight incubation was carried out at 4℃. Afterwards, plates were washed five times with 1× high salt wash (HSW) buffer (2.4 g/L Trizma base, 29.2 g/L NaCl, 0.1 g/L merthiolate, and 1 mL/L Tween 20), and each well was subsequently incubated with 100 µL of an anti-European seabass IgM monoclonal antibody (F01; Aquatic Diagnostics, UK) (1 : 33 dilution of the reconstituted antibody in 1% BSA in PBS) for 1 h at room temperature following the manufacturer's instructions. After washing five times with 1× HSW buffer, 100 µL of HRP-conjugated goat anti-mouse IgG (1 : 4,000 dilution in 1% BSA in 1× LSW buffer; Sigma, USA) were added and the plates were incubated at room temperature for 1 h. After washing five times with 1× HSW buffer, the reactions were finally visualized by adding 3,3',5,5'-tetramethylbenzidine (50 µL/well) (Sigma, USA).

The optical density (OD) of each well was measured at 450 nm using an ELISA Whittaker Microplate Reader 2001 (Anthos Labtec, Austria). The resulting OD values were normalized by subtracting the OD of the negative control wells. Based on a report by Skinner et al. [34], the mean OD value of the negative controls multiplied by three was the cut-off threshold used in the present study. The highest serum dilution above this value was the anti-RGNNV antibody endpoint titer for that sample.

RGNNV infection resulting from viral intramuscular (IM) injection induced fish mortality between 7 and 29 days PI. External signs of the disease were dark skin pigmentation, reduced feeding, hyperactivity, and erratic swimming patterns. Internally, the swim bladder was expanded and liver mass was reduced. Cumulative mortality at the end of the experiment in tank No. 3, in which no sampling was conducted, was 37%. No mortality was recorded among the uninfected fish. Analyses of dead fish by cell culture inoculation demonstrated the presence of RGNNV, which was subsequently confirmed by a previously described specific RT-PCR protocol (data not shown).

IHC revealed the presence of stained cells in tissue sections from infected fish prior to the appearance of the typical external VNN symptoms, which occurred 5 days PI. In contrast, no labeling was observed in tissues from the uninfected fish at any time (Figs. 1 and 2). Staining intensity and the number of labeled cells indicated the degree of infection in each tissue (Table 1). Staining was evident in the retina, brain, and liver throughout the entire experimental period (from 3 days to 2 months PI). In the remaining tissues, viral antigens were only detected after the onset of mortality (7 days PI; Table 1). Thus, at 10 days PI all non-nervous tissues analyzed (liver, kidney, spleen, and caudal fin) were positive for RGNNV. On the other hand, the spleen was negative 17 days PI, and the caudal fin and liver were the only non-nervous tissues showing immunostaining 24 days PI. Specifically, viral antigens were present in tubular epithelial cells of the kidney, hepatocytes, endothelial cells of the liver (Figs. 1B and D), white pulp of the spleen (Fig. 1F), and fibroblasts located in the epidermis and cartilage of the caudal fin (Fig. 1H).

When examining nervous tissues, cells in the brain showed a constant level of immunostaining throughout the experiment. In contrast, increased intensity of the IHC signal was recorded in retina as the mortality period was approached. The staining intensity then decreased to levels similar to those obtained in brain at the end of the mortality period (Table 1). The only parts of the brain that could be examined and compared among all sampled animals were the cerebellum and tectum opticum (TO). Immunolabeling was first observed in the peripheral zone of the cerebellum (i.e., the only signal observed in brain 3 days PI). Viral antigens in TO of animals sampled from 10 days PI onwards were also observed. At 17 and 24 days PI, IHC signals were predominately located in the area between the cerebellum and TO (Fig. 2B).

In retina, the ganglion layer was the only area that was stained throughout the experiment. Additionally, viral antigens also appeared in the inner nuclear layer 3 days PI as well as in the outer limiting membrane and outer plexiform layer 17 days PI (Fig. 2D). The outer plexiform layer and inner nuclear layer also exhibited IHC signals 2 months PI.

Characteristic VNN lesions, such as vacuolation, were only observed in the brain (cerebellum and TO) along with retina where degeneration was always more evident than in brain. In the cerebellum and TO, vacuolation was only observed at 3, 10, 17, and 24 days PI, coinciding with the maximum level of mortality (Fig. 3B). In retina, vacuoles appeared in the outer and inner plexiform, inner nuclear, and ganglion layers up to 10 days PI. At 17 days PI, these lesions were also observed in the outer nuclear layer (Fig. 3D). However, vacuolation was noted only in the outer plexiform, inner plexiform, and inner nuclear layers 24 days PI. The only histological alteration found in internal organs was accumulation of fat vacuoles in the liver of challenged fish 10, 17, and 24 days PI (Fig. 3F).

Three samples consisting of pooled sera from five different mocked-injected or challenged fish were screened for the presence of anti-RGNNV antibodies by an ELISA. A measurable level of specific anti-RGNNV antibodies was recorded only in challenged animals. The results indicated that intense anti-RGNNV antibody production was initiated as soon as 24 days PI when the titer was 1 : 16,384. At 2 months PI, the antibody titer was 1 : 8,192 (Fig. 4).