The isolate of Ph. d. subsp. piscicida (isolate I752) used in this study came from the bacterial collection of the Dipartimento di Scienze della Produzione Animale, Universita di Udine, Italy, and was recovered from sea bream (Scophthalmus maximus) infected with pasteurellosis in 1996.

The bacteria were cultured in four different media: (1) TSB with 1.5% NaCl; (2) iron-restricted TSB with 1.5% NaCl and 0.175 mM 2,2'-dipyridyl (TSB + IR^-); (3) GRM (D-glucose 20 g/l, special peptone (Oxoid) 10 g/l, yeast extract 5 g/l, KH₂PO₄ 0.25 g/l, MgSO₄ · 7H₂O 0.01 g/l, CaCl₂ 0.6 mg/l, FeSO₄ · 7H₂O 2.4 mg/l, MnSO₄ · 7H₂O 0.45 mg/l, NaCl 10 g/l; pH 7.2); and (4) iron-limited GRM which had the same constituents as (3), excluding FeSO₄ · 7H₂O, but with the addition of 0.25 mM 2,2'-dipyridyl (GRM+IR^-). The bacteria were first cultured in TSB at 22℃ for 18 h, and were then diluted 1 to 10 in the chosen culture medium. The bacteria were cultured in either TSB or GRM for 18 h or in TSB + IR^- or GRM+ IR^- for 36 h, respectively, without agitation. The bacteria were then transferred into fresh media (again using a 1 to 10 dilution) and cultured to log phase prior to use.

Bacteria were grown and harvested by centrifugation at 2,900 × g for 30 min at 4℃, and were washed twice with sterile phosphate-buffered saline (PBS: 0.02 M NaH₂PO₄ · 2H₂O, 0.02 M Na₂HPO₄ · 2H₂O, 0.15 M NaCl; pH 7.2). The concentration of the washed bacteria was adjusted to an absorbance of 1.0 at 610 nm with PBS, and colony-forming units (CFU/ml) of the suspension were determined retrospectively. A portion of the washed bacterial preparation was heat-killed at 60℃ for 60 min, and aliquots of the killed bacteria were stored at -70℃ and used for immunizing fish and screening sea bass antisera.

Sea bass weighing 40 g or 350 g were bought from a commercial fish farm in Italy. The fish were maintained in a land-based flow-through seawater aquarium belonging to the Dipartimento di Scienze della Produzione Animale, Universita di Udine. The tanks were equipped with a system for sterile drainage water, and both the water and the fish were determined to be Ph. d. subsp. piscicida-free. The water temperature was maintained between 25 and 26℃, and the water salinity was 24 ‰.

Eight fish (350 g) were injected intraperitoneally (i.p.) with 5 × 10³ CFU/fish (1 ml) of live Ph. d. subsp. piscicida. The bacteria were grown in TSB at 22℃ for 18 h, and were washed twice with sterile PBS before adjusting the concentration of the suspension prior to injection. Four fish were injected with PBS as a negative control. The fish were bled 3 weeks later from their caudal vein, and the sera were used for Western blot analysis.

Sixty-three fish (40 g), which were also maintained as described above, were injected i.p. with a live preparation of Ph. d. subsp. piscicida. The bacteria were grown in TSB as described above, washed twice with PBS, and adjusted to a concentration of 1 × 10³ CFU/fish (0.2 ml). Another 44 fish were injected with 0.5 mg/fish of heat-killed bacteria in 0.2 ml PBS. The bacteria were grown in TSB, washed twice with PBS, and heat-killed prior to use as described above. Ten fish were injected with PBS (0.2 ml/fish) and used as a negative control. Fish were bled 3 and 4 weeks post-injection, and the antibody levels present in their sera were examined by ELISA.

Antibody levels produced by the fish were measured using ELISA according to the method of Bakopoulos et al. [5], with slight modifications. ELISA plates (Immulon; Dynatech, USA) were coated with 0.01% poly-L-lysine in carbonate-bicarbonate buffer, pH 9.6 (100 ml/well), and were incubated for 1 h at 20℃. The plates were washed three times with low salt wash buffer (LSW: 0.02 M Trizma base, 0.38 M NaCl, 0.05% (v/v) Tween-20, pH 7.4), and 100 ml/well of bacterial suspension was added to the plates. Bacteria cultured in the four different media were adjusted to a concentration of 2 × 10⁸ bacteria/ml. The plates were incubated at 22℃ for 1.5 h, and bacteria were then fixed to the wells by the addition of 50 ml/well glutaraldehyde (0.05% v/v) diluted in PBS for 20 min. The plates were again washed by three LSW buffer washes. Non-specific binding sites were blocked with 100 ml/well of a 3% (v/v) solution of H₂O₂ for 1 h, and were then incubated with 250 ml/well 1% (w/v) gelatine solution in LSW. The ELISA plate was washed three times with LSW buffer. Fish sera, which were diluted 1 : 100 in PBS containing 0.01% (v/v) Tween-20, were added in duplicate (100 ml/well). Sera from fish injected with PBS were used as a negative control, along with a negative control of LSW. The plates were incubated at 20℃ for 1.5 h, washed 5 times with high salt wash buffer (0.02 M Tris, 0.5 M NaCl, 0.1% Tween-20, pH 7.8), and allowed to stand for 5 min during the last wash prior to the removal of the buffer. Anti-sea bass IgM monoclonal antibody (MAb) (Aquatic Diagnostics, UK) was added to the wells (100 ml/well) for 1 h at 20℃. The plates were again washed with the HSW as described above prior to incubation for 1 h with horseradish peroxidase conjugated to anti-mouse-IgG (Diagnostics Scotland, UK) diluted 1 : 1000 in LSW. Plates were washed with HSW as described above, and chromogen/substrate [120 ml of 43 mM tetramethylbenzidine dihydrochloride in 2M acetic acid added to 12 ml of substrate buffer (0.1 M citric acid, 0.1 M sodium acetate, pH 5.4, containing 0.33% v/v H₂O₂)] was added to each well (100 ml/well). The reaction was stopped after 7 min by the addition of 50 ml/well of 2M H₂SO₄, and was read spectrophotometrically at 450 nm using an ELISA reader (Dynatech, USA). Optical densities exceeding or equal to 3 times the mean background absorbance were considered positive.

Some of the bacteria used in Western blot analysis were freshly isolated from infected fish onto tryptone soya agar (TSA), and were then immediately sub-cultured in one of the four different culture media described above (Fig. 1a-e), while bacteria used in Fig. 1 (f-h) were subcultured several times (more than ten) on TSA prior to culturing in the four different culture media. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed on Ph. d. subsp. piscicida. In brief, a 12% acrylamide separating gel was prepared, onto which a 4% stacking gel was layered. Bacteria that had been grown in the four different media and washed in PBS as described above were then adjusted to an OD of 1.0 at 610 nm. The bacteria were prepared in sample buffer as described by Bollag et al. [8], with boiling for 5 min. Samples were placed on the gel (15 ml/well), and 180 V was applied to the gel for 45 min. Pre-stained molecular standard markers (Bio-Rad, USA) were used as standards for each gel. Gels containing whole bacteria were transferred onto nitrocellulose membranes by applying 60 V for 70 min using a Hoeffer transblotter. Non-specific binding sites on the nitrocellulose membrane were blocked with 1% w/v bovine serum albumin in Tris-buffered saline (TBS: 10 mM Tris, 0.5M NaCl, pH 7.5) for 1 h at 20℃. The membranes were then washed three times with TBS containing 0.1% (v/v) Tween-20 (TBST); each wash lasted 10min. The membranes were then incubated in anti-Ph. d. subsp. piscicida sea bass sera (diluted 1 to 10 with TBST) overnight at 4℃ with gentle agitation, after which they were washed as described above. Sera collected from fish injected with PBS were used as a negative control. The membranes were incubated with anti-sea bass IgM MAb for 3 h at 20℃. They were again washed three times with TBST, and were then incubated with anti-mouse IgG-HRP conjugate diluted 1 to 500 with TBST. The membranes were incubated with the conjugate for 1 h at 20℃. Blots were washed three times with TBST, 10 min per wash, and were then washed once with TBS and once with PBS. The reaction was then developed by the addition of 20% chromogen (4-chloro-naphthol, 3 mg/ml in methanol) in PBS and 0.01% H₂O₂ until bands appeared. The reaction was stopped with distilled H₂O.

Data from ELISA were analyzed using a one-way ANOVA. Multiple comparisons using a Tukey pairwise comparison made with the Minitab 11 program. Results were considered significant when p < 0.05.

Sera from sea bass infected with live bacteria exhibited significant differences in their antibody responses at 3 weeks post-infection when screened against non-iron-restricted and iron-restricted bacteria using ELISA, with the latter producing greater antibody responses compared to those of non-iron limited bacteria (Table 1). When heat-killed bacteria were used to immunize sea bass, no significant differences were observed in the antibody responses of their sera against non-iron-limited and iron-limited bacteria, as indicated by screening with ELISA at 3 weeks post-infection (data not shown). However, significant differences were found in these sera between bacteria cultured in TSB and bacteria cultured under the other three culture conditions at 4 weeks post-immunization (Table 1). No antibodies were detected against Ph. d. subsp. piscicida in control fish injected with PBS.

The response of anti-sea bass Ph. d. subsp. piscicida sera against bacteria grown under different culture conditions, compared using Western blot analysis.

Each gel represents serum collected from an individual fish at 3 weeks after injection with Ph. d. subsp. piscicida. The patterns of staining in the blots presented in Fig. 1 (A-E), where the bacteria had been freshly isolated from an infected fish, showed differences between bacteria cultured under iron-restriction (TSB + IR^-, GRM+ IR^-) and those grown under non-restricted conditions (TSB, GRM). However, when bacteria had been passaged several times on TSA prior to culturing in the four different media (Fig. 1F-H), a similar pattern of staining was obtained with the antisera against bacteria grown under all four culture conditions. This is clearly shown in Fig. 1A and 1F, where the same serum was used but the staining pattern changed between the freshly-isolated bacteria (Fig. 1A) and bacteria passaged several times on TSA (Fig. 1F).

The reaction of sea bass sera to the freshly-isolated bacteria (Fig. 1A-E) resulted in staining of a 22 kDa band in the TSB + IR^- and GRM+ IR^- bacterial profiles. The sea bass sera, however, recognized a band at 24 kDa in bacteria cultured in TSB and GRM, which was approximately 2 kDa higher than that observed with TSB + IR^-- and GRM+ IR^--cultured bacteria. Two further bands were detected at 47 and/or 57 kDa in some of the profiles of TSB- and GRM-cultured bacteria using freshly-isolated bacteria (Fig. 1A-D). When bacteria had been passaged several times in vitro, two main bands were identified at 24 and 47 kDa; the 22 kDa band was only observed in two repeatedly passaged samples grown under iron-restriction (Fig. 1F & G). Generally, when freshly-isolated bacteria were used, less staining of low molecular weight material was evident with the TSB + IR^- and GRM+ IR^- bacterial samples compared with TSB and GRM bacteria. However, a similar level of staining of low molecular weight material was observed with the bacteria subcultured several times, as shown in Fig. 1F-H.