Immune responses and protective immunity in Pangasius pangasius (Hamilton, 1822) as induced by outer membrane proteins of Edwardsiella tarda and aluminium hydroxide adjuvant complex

Edwardsiella tarda is considered one of the important bacterial fish pathogens. The outer membrane proteins (OMPs) of E. tarda are structurally and functionally conserved, and immunogenic. This study assessed the effects of the OMPs of E. tarda CGH9 as a vaccine without aluminium hydroxide [AH] (T1) and with AH adjuvant (T2) on the respiratory burst (ROB) activity, lymphocyte proliferation of head kidney (HK) leukocytes, and serum antibody production in pangas catfish Pangasius pangasius. The ROB activity and lymphocyte proliferation of HK leukocytes increased in both vaccinated groups compared to the control. Nonetheless, the T2 group showed a gradual increase in ROB activity and lymphocyte proliferation of HK leukocytes up to 3-weeks post-vaccination (wpv). The serum antibody production in the T1 group decreased initially for up to 2-wpv and increased from 3-wpv; whereas, in the T2 group, the serum-specific antibody levels were significantly high from 1-wpv compared to control. Simultaneously, the protective efficacy in terms of relative percentage survival in the T2 group after injecting with a lethal dose of E. tarda CGH9 was high (89.00±15.56) compared to the T1 group (78.00±0.00). Furthermore, the catfish administered with a booster dose of E. tarda OMPs with or without AH adjuvant showed no additional increase in immune response or protective immunity. These results suggested that E. tarda OMPs and AH adjuvant complex has a higher potential to induce protective immunity, which may be a good choice as a vaccine to combat E. tarda infection in catfish.


Introduction
Edwardsiella tarda, a Gram-negative bacterial pathogen of the family Enterobacteriaceae, causes edwardsiellosis in many cultured fish species. It is considered one of the important bacterial pathogens in global aquaculture notably in Nile tilapia (Oreochromis niloticus), Japanese flounder (Paralichthys olivaceus), turbot (Scophthalmus maximus), carp (Labeo rohita), channel catfish (Ictalurus punctatus), striped bass (Morone saxatilis), rainbow trout (Oncorhynchus mykiss) and pangas catfish (Pangasius pangasius) (Kodama et al. 1987;Park et al. 2012;Xu and Zhang 2014;Adikesavalu et al. 2016a). Pangasius pangasius is one of the important farmed freshwater catfish species with high economic values, especially in Asian countries including India (Mugaonkar et al. 2019). Control of E. tarda infection mostly relies on the use of antimicrobial compounds, mainly antibiotics. Earlier studies reported that members of the genus Edwardsiella were highly susceptible to tetracyclines, aminoglycosides, cephalosporins, quinolones, antifolates, chloramphenicol, nitrofurantoin, fosfomycin, and most β-lactam antibiotics (Stock and Wiedemann 2001). In recent years, resistance to antibiotics such as amoxicillin, chloramphenicol, levofloxacin, streptomycin, lincomycin, norfloxacin, and nalidixic acid has been reported in E. tarda (Ogbonne et al. 2018;Ezzat et al. 2021). This selection and spread of antibiotic-resistant bacteria have raised concerns regarding antibiotic-based therapeutic procedures to control bacterial infection. Therefore, the best way to combat bacterial diseases is through vaccination (Assefa and Abunna 2018). Currently, most aquaculture vaccines are aimed at different bacterial diseases, like vibriosis, furunculosis, and yersiniosis, although with varying degrees of success (Sommerset et al. 2005;Dadar et al. 2017;Assefa and Abunna 2018). However, the existence of antigenic diversity within the same bacterial species poses difficulty in developing an effective vaccine (Park et al. 2012). Numerous vaccine strain candidates such as formalin killed cells, lipopolysaccharides, extracellular proteins, live attenuated, avirulent bacterial, ghost cells, outer membrane proteins (OMPs), recombinant proteins, and DNA vaccines have been developed in the past and have been tested in the different fish models (Park et al. 2012). The two vaccine types namely killed bacterins and modified live (attenuated) bacteria are available against Edwardsiella ictaluri, which primarily affects catfish (Assefa and Abunna 2018). The live attenuated vaccine is presently sold commercially against edwardsiellosis (Dadar et al. 2017;Assefa and Abunna 2018). Recently, Wise et al. (2020) evaluated the efficacy of a live-attenuated E. ictaluri vaccine in I. punctatus fry administered through an oral vaccine delivery system. According to them, it improved the total harvest, survival, and FCR significantly in pond trials.
To date, several vaccine candidates have been reported for the edwardsiellosis disease Jiao et al. 2009Jiao et al. , 2010Sun et al. 2010Sun et al. , 2011Tang et al. 2010). However, the immunogenic nature of the prepared vaccines was still debatable because inactivation procedures may alter the antigenicity of protective epitopes, attenuated vaccine candidate may revert to their virulent form, and live avirulent vaccine candidate has the chance of acquiring virulent features through the transformation of genes that encode virulence factors (Dadar et al. 2017). Earlier reports by different research groups highlighted the fact that OMP homologues of bacterial species are both structurally and functionally conserved and therefore, could be immunogenic with broad protective coverage (Sun et al. 2011;Dubey et al. 2019). Jalava et al. (2003) reported that the immunoprotective efficacy of a vaccine candidate against pathogens was improved by combining it with adjuvants. Although the use of oil-based adjuvants in fish vaccines has long-term immune responses, side effects such as growth reduction, formation of granulomas, and serious intra-abdominal tissue lesions have been reported in different fish species (Jiao et al. 2010;Noia et al. 2014;Li et al. 2020). As the aluminium hydroxide (AH) adjuvant is considered safe for human vaccines, it has also been reported as a vaccine adjuvant in I. punctatus against enteric septicemia of catfish (Tyler and Klesius 1994), P. olivaceus against E. tarda infection (Jiao et al. 2010) and viral hemorrhagic septicemia (Vinay et al. 2013) and S. maximus against viral reddish body syndrome (Fan et al. 2012). The aluminium adjuvant-antigen complex could effectively increase the antigen exposure time by acting as slow-release depots, which, in turn, leads to prolonged innate signalling that stimulates T helper-2-type (Th-2) lymphocyte response almost exclusively and thus, considered effective in promoting protective humoral immunity (Gupta et al. 1995;Jiao et al. 2010). Besides, increased specific antibody titers and NADPH-dependent reactive oxygen species (ROS) production were documented with aluminium salts as adjuvants (Angosto et al. 2018). The pangas catfish is one of the important farmed freshwater catfish species with high economic values that are also affected by E. tarda infection (A'yunin et al. 2020). Additionally, the reports on the usage of E. tarda OMPs along with AH adjuvant as a vaccine in the catfish model were scanty. Therefore, the present study aimed to assess the immunoprotective efficacy of E. tarda CGH9 OMPs with and without AH adjuvant in P. pangasius concerning respiratory burst (ROB) activity, in-vitro lymphocyte proliferation, and serum antibody production.

Bacterial strain
The bacterial strain Edwardsiella tarda CGH9 (Accession number KX159725) was originally isolated from African catfish (Clarias gariepinus) with typical symptoms of dropsy (Adikesavalu et al. 2016a). The bacterial strain was grown in tryptone soya broth (TSB) at 30°C and stock cultures of 5 ml aliquots containing 20% glycerol were maintained frozen at −20°C until further use.

Collection and maintenance of experimental fish
The experimental pangas catfish, Pangasius pangasius (61.71±2.36 g and 20.21±0.76 cm) were procured from Kantipota, South 24 Parganas district, West Bengal, India (Lat. 22°27′49″ N; Long. 88°24′ 41″ E), and disinfected by immersion in a 5 ppm potassium permanganate (KMnO 4 ) solution for 15 min. The fish were stocked in 500 L capacity circular fibreglass reinforced plastic (FRP) tanks at 65 numbers/tank and acclimatized for 21 days with continuous aeration before experimental vaccination. The fish were fed with a balanced basal dry pellet feed containing 30% protein, 2% fat, and 8% fiber (CP9931, CP Pvt. Ltd., India) twice daily at 2% body weight. The wastes and faecal matter were siphoned out daily and 50% of water was exchanged once in 3 days.

Preparation of anti-pangas serum globulin (ASG) horseradish peroxidase (HRPO) immunoconjugate (ASG-HRPO immunoconjugate)
Anti-pangas serum globulin enzyme (HRPO) immunoconjugate was developed as described by Adikesavalu et al. (2016b). Briefly, rabbit immune serum, with specific antibodies against pangas serum globulin protein fraction, was separated by salt fractionation method using saturated ammonium sulphate. Then, 1 ml HRPO solution (4 mg/ml in distilled water) and 200 μl of freshly prepared sodium periodate (0.1 M) were added and stirred gently for 20 min at room temperature. The mixture was dialyzed overnight against sodium acetate buffer (0.001 M, pH 4.4) at 4°C and 20 μl of sodium carbonate buffer (0.1 M, pH 9.5) was added to the mixture to raise the pH to 9. After that, the fractionated rabbit antiserum globulin protein fraction at 8 mg/ml concentration was added along with 100 μl of freshly prepared sodium boro-hydride solution (4 mg/ml in distilled water) and incubated at 4°C for 2 h. The mixture was dialyzed against PBS for 48 h and an equal volume of 60% glycerol in PBS was added to the fresh ASG-HRPO immunoconjugate and stored at -20°C.

Preparation of outer membrane proteins (OMPs) of Edwardsiella tarda CGH9
The suspension of E. tarda CGH9 OMPs was prepared by following Adikesavalu et al. (2016a). In brief, the strain was revived on tryptone soya agar (TSA) and cultured in TSB at 30°C for 24 h. The harvested bacterial cell pellet was washed thrice using sterile phosphate buffer saline (PBS) by centrifugation at 8000 rpm for 25 min at 25°C. The cell pellet was resuspended in 20 mL PBS, followed by treatment with 2% sodium dodecyl sulphate (SDS) and 2% mercaptoethanol for 20 min at 60°C for solubilization. The supernatant was collected by centrifuging the solubilized extract at 8000 rpm for 30 min at 4°C. The supernatant was dialyzed for 48 h against sterile PBS, followed by filter sterilization (φ 0.22 μ) and stored at −20°C until further use. The estimation of protein content was as per Lowry et al. (1951).

Preparation of aluminium hydroxide (AH) adjuvant
The aluminium hydroxide (AH) adjuvant was prepared following the method of Jiao et al. (2010). Briefly, 5% sodium hydroxide (NaOH) (HiMedia, India) and 5% aluminium sulphate (Al 2 (SO 4 ) 3 ) (HiMedia, India) were prepared in distilled water. The two solutions were filter sterilized (φ 0.45 μ) and incubated at 60 o C for 30 min. Two volumes of 5% NaOH and five volumes of 5% Al 2 (SO 4 ) 3 were mixed with stirring, followed by centrifugation at 10, 000 rpm for 5 min. After washing twice with sterile PBS, a diluted solution of aluminium hydroxide at 4 mg/ml concentration was prepared and stored at 4 o C for further use.

Vaccine formulation
The crude suspension containing E. tarda CGH9 OMPs at 500 μg/ml was used directly as a vaccine. The other formulation contained a mixture of crude E. tarda CGH9 OMPs at 500 μg/ml and AH adjuvant at 0.2 mg/ml concentration. A volume of 100 μl from the mixture will contain 50 μg of E. tarda CGH9 OMPs and 0.02 mg of AH as an adjuvant.

Experimental vaccination
The vaccination experiment was performed for 4-weeks and the catfish were divided randomly into five groups (65 fish/ group) labelled as control (C), T1 (E. tarda CGH9 OMPs without AH adjuvant), T1B (booster E. tarda CGH9 OMPs without AH adjuvant), T2 (E. tarda CGH9 OMPs and AH adjuvant complex) and T2B (booster E. tarda CGH9 OMPs and AH adjuvant complex). For experimenting with duplicates, the 65 fish from each group were divided into two 300 L FRP tanks (33 fish/tank 1 and 32 fish/tank 2) per group. Therefore, a total of 10 FRP tanks of 300 L capacity with either 33 fish/tank or 32 fish/tank were used. The feeding was stopped 24 h before vaccination and the fish were anesthetized using clove oil (50 μl/l) to reduce their physical activity before vaccination. Each fish of the groups T1 and T1B was administered with 0.1 ml (50 μg) of E. tarda CGH9 OMPs without AH adjuvant intramuscularly. Similarly, the T2 and T2B groups were injected intramuscularly with 0.1 ml of E. tarda CGH9 OMPs and AH adjuvant complex, i.e., 50 μg OMPs and 0.02 mg AH. The C group received 0.1 ml of sterile PBS. The fish were maintained in their respective tanks with continuous aeration and the feeding was resumed 12 h post-vaccination. The fish were observed for mortality, external signs of infections, and behavioural changes. One day after 2-weeks post-vaccination (wpv), fish from the two boosters (T1B and T2B) groups were injected intramuscularly with a second dose of 0.1 ml each of the respective vaccine formulation and maintained again in their respective tanks with continuous aeration and feeding in the same way as mentioned above. Experimental protocols were approved by the Indian Council of Agricultural Research, New Delhi under the Niche Area of Excellence programme and fulfilled the ethical guidelines including adherence to the legal requirements of India (CPCSEA 2018).

Blood sampling and isolation of head kidney (HK) leukocytes
On each sampling, i.e., 1-, 2-, 3-and 4-wpv, four fish from each group were randomly sampled for blood collection and head kidney isolation to study the effect of the vaccine on the immune parameters such as ROB activity, in-vitro lymphocyte proliferation, and serum antibody levels. Likewise, on 1-and 2-weeks post-booster vaccination (wpbv), the fish from T1B and T2B groups were sampled to assess the effect of the booster dose on the immune parameters. This 1-and 2-wpbv corresponded with samples that were collected on 3-and 4-wpv from C, T1, and T2 groups. The assessment of serum antibody production alone was done up to 6-wpv in C, T1, and T2 groups. The fish serum collected up to 6-wpv corresponded with samples that were collected up to 4-wpbv in T1B and T2B groups.
The blood was collected from the anesthetized fish using clove oil at 50 μl/l water by caudal vein puncture and an aliquot of pooled blood was mixed with 2.7% EDTA to prevent coagulation. The aliquot of blood without anti-coagulant was allowed to clot at room temperature (≈30°C) and kept at 4°C overnight. The serum samples were collected by centrifugation at 2500 rpm for 15 min and pooled before storage at -20°C. After blood collection, the fish were euthanized by increasing the clove oil dose to 100 μl/l water. The leukocytes from the head kidney and spleen were collected as per Adikesavalu et al. (2016a). In brief, the head kidney and spleen were extirpated and prepared the cell suspension in PBS (pH 7.4) containing antibiotics, such as penicillin 100 IU/ml (HiMedia, India) and streptomycin 100 μg/ml (HiMedia, India) by squeezing the tissue pieces through sterile stainless steel screens. The cell suspension was layered over a density gradient, Histopaque (Sigma, USA) at a 1:3 ratio, and centrifuged at 1200 rpm for 30 min. The mononuclear cells present between plasma and density gradient were collected and washed with PBS. The cells were finally suspended in RPMI-1640 medium with 2.05 mM L-glutamine (Hyclone, USA) supplemented with 10% fetal calf serum (HiMedia, India) and antibiotics (penicillin 100 IU/ ml and streptomycin 100 μg/ml). Trypan blue dye exclusion method using a haemocytometer was done for enumeration and determining the viability of the purified leukocytes.

Respiratory burst (ROB) activity
The ROB activity by neutrophils was determined by the reduction of nitroblue tetrazolium (NBT) to formazan following Mohanty and Sahoo (2010) with minor modification. Briefly, an equal volume of freshly collected heparinized blood and 0.2% NBT (Amresco, USA) in PBS was mixed and incubated at room temperature (30°C) for 30 min. After incubation, 50 μl of this mixture was mixed with 1 ml of dimethylformamide (HiMedia, India) to solubilize the reduced formazan product and centrifuged at 2000 rpm for 5 min to collect the supernatant. The extent of NBT reduction was measured at an optical density of 540 nm using di-methyl formamide as blank.

In-vitro lymphocyte proliferation assay
The proliferative response of the HK leukocytes was determined by the tetrazolium-based colourimetric assay as described by Adikesavalu et al. (2016a). Briefly, the HK leukocytes were distributed into the wells of a 96-well cell culture plate at 1.50×10 6 cells/well. Mitogen concanavalin-A [Con-A] (Genei, Bangalore) stock solution was prepared at the concentration of 80 μg/mL using a growth medium. To each well, 100 μl with the final Con-A concentration of 40 μg/ml was added. Similarly, 100 μl of the suitably diluted stock solution of crude E. tarda CGH9 OMPs was added to the respective wells to achieve a final concentration of 40 μg/ml. Meanwhile, a stock suspension of E. tarda CGH9 of approximately 1.50×10 8 cells/ml was prepared using the growth medium. A multiplicity of infection of a 1:10 (phagocytes: bacteria) ratio was used for infecting the HK leukocytes as described by Nakhro et al. (2013). This was achieved by adding 100 μl of stock suspension of E. tarda CGH9 into appropriate test wells. The remaining wells were filled up to 200 μl with the medium (negative control). The plate was incubated for 24 h at 30°C. After incubation, 20 μl of filter-sterilized MTT [3-(4, 5-dimethyl thiazol-2-yl)-2, 5-diphenyl tetrazolium bromide] (HiMedia, India) solution (10 mg MTT/ml) was added into all the wells and incubated again at 30°C for 6 h. After incubation, 150 μl of the supernatant was carefully removed without disturbing the cells or the formazan precipitate. This was followed by the addition of 150 μl of dimethyl sulphoxide (DMSO) and 20 μl of glycine buffer (0.1 M glycine, 0.1M NaCl, pH 10.5) into all the wells. The contents of the wells were mixed thoroughly by pipetting and incubated for 10 min at 30°C. The cell culture plate was read on a Microtiter plate reader at 595 nm.

Serum antibody production
Indirect ELISA was performed in a 96-well plate (Nune, Denmark) as per Mishra and Shekhar (1997). Briefly, the plate was coated in duplicate with a coating buffer (carbonate bicarbonate buffer, pH 9.6) containing antigen (2.5 μg E. tarda CGH9 OMPs/well) and kept at 4 o C overnight. After the antigen coating and blocking (5% skimmed milk powder in PBS) for 2 h at 37 o C, the plate was washed thoroughly 4 times with PBS containing 0.05% Tween-20. Then 100 μl of diluted (1:200 in PBS) test and control fish sera were added into the appropriate antigen-coated wells and incubated at 37 o C for 2 h. The plate was washed again as above and 100 μl each of diluted (1:100 in PBS) anti-pangas serum globulin enzyme (HRPO) immunoconjugate was added into the wells and left for 2 h at 37 o C. The plate was then thoroughly washed and 100 μl of substrate solution (15 μl of 6% H 2 O 2 , 0.025 g ortho-phenylenediamine dihydrochloride in 25 ml citrate buffer) was added to the wells in dark. The colour development was arrested after 20 min by adding 3N HCl (35% pure). The colour development was read at 492 nm using a Microplate reader.

Vaccine efficacy determination: Challenge with Edwardsiella tarda CGH9
After 4-wpv, vaccine efficacy determination was performed in duplicates by dividing 40 catfish from each of the five groups (C, T1, T2, T1B, and T2B) into two 200 L tanks per group. Therefore, a total of 10 tanks of 200 L capacity with 20 fish/tank were set and labeled properly. Each fish was challenged intramuscularly with 0.1 ml of ≈10 8 cells/ml (lethal dose) of E. tarda CGH9 and observed daily for mortality, external signs of infections, and behavioural changes for 30 days. The dying fish were randomly selected for the examination of bacterial recovery. The effectiveness of the vaccine formulation was determined from the mortality data using the relative percentage survival (RPS) as described by Bader et al. (2004). The RPS values >60 were considered protective.

Statistical analyses
The data (mean±standard deviation) were analyzed by oneway repeated-measures analysis of variance (ANOVA) using OriginPro version 9.0.0. Pairwise comparisons of mean values were determined by the Tukey test, considering a probability level of p<0.05.

Results
The immune effects of E. tarda CGH9 OMPs were evaluated with or without AH adjuvant by quantifying the levels of ROB activity, in-vitro lymphocyte proliferation, and serum antibody production. The vaccinated catfish exhibited inflammation at the site of injection, which lasted for a week when compared to control.

Respiratory burst (ROB) activity
The ROB activity was significantly (p<0.05) high on 1-wpv in both E. tarda CGH9 OMPs vaccinated groups without AH adjuvant (T1) and with AH adjuvant (T2) compared to control (Fig. 1A). A significant (p<0.05) difference in ROB activity between the control and T1 groups was noted only up to 2-wpv. Whereas, the ROB activity increased gradually up to 3-wpv compared to the control in the T2 group RPS = 1 − Mortality in vaccinated group Mortality in unvaccinated control × 100 and reduced thereafter. The differences in ROB activities between the T1 and T2 groups on 1-wpv as well as on 3-wpv were significant (p<0.05). The ROB activity in the T1 group was significantly (p<0.05) high on 1-wpv; while it was significantly (p<0.05) high in the T2 group on 3-wpv (Fig. 1A). The ROB activities in booster groups (T1B and T2B) showed no further increase (Fig. 1B). In group T1B, the ROB activity on 1wpbv was significantly (p<0.05) low compared to 2-wpv from the T1 group (Fig. 1B). The ROB activity in the T2B group showed an increase on 1-wpbv compared to 2-wpv from the T2 group (Fig. 1B). Nonetheless, the observed increase was insignificant (p>0.05). On 2-wpbv, the ROB activities in T1B and T2B groups reached near normal, which corresponded to those observed in T1 and T2 groups on 4-wpv.

In-vitro lymphocyte proliferation assay
The lymphocyte proliferation of HK leukocytes of the T1 group, when tested with a non-specific stimulator Con-A, showed a significant (p<0.05) increase on 1-wpv when compared to control. It reduced thereafter to reach near control level on 4-wpv ( Fig. 2A). In the T2 group, the lymphocyte proliferation of HK leukocytes, when tested with Con-A showed a significant (p<0.05) increase up to 3-wpv, when compared to control and reduced thereafter. The level of the proliferative response of leukocytes on 3-wpv (3.231OD) was significantly (p<0.05) higher than those observed on 1-wpv and 4-wpv ( Fig. 2A). The levels of proliferative responses in the T2 group were high on 2-, 3-and 4-wpv compared to the T1 group. However, the differences in the proliferative responses between the T1 and T2 groups were significant (p<0.05) only on 3-and 4-wpv ( Fig. 2A) In the lymphocyte proliferation of HK leukocytes of the T1B group, when tested with Con-A, no significant change was observed on 1-wpbv when compared to 2-wpv from the T1 group (Fig. 2B). The levels on 2-wpbv (0.887 OD) were significantly (p<0.05) low when compared to 1-wpbv and also to 2-wpv from the T1 group. The lymphocyte proliferation of the T2B group, when tested with Con-A, showed no significant increase on 1-wpbv and 2-wpbv when compared to 2-wpv from the T2 group. However, the levels of proliferative responses in the T2B group were significantly (p<0.05) high on 1-and 2-wpbv compared to the T1B group (Fig. 2B).
Further, the lymphocyte proliferation of HK leukocytes of the T1 group, when tested with E. tarda CGH9 OMPs, increased insignificantly (p>0.05) up to 2-wpv and reduced on 3-wpv when compared to control (Fig. 3A). In the T2 group, when tested with E. tarda CGH9 OMPs, the proliferative responses were significantly (p<0.05) high on all wpv compared to the control. The levels of proliferative responses increased up to 3-wpv (2.762 OD). The levels of proliferative responses in the T2 group were high on all wpv compared to the T1 group. However, as observed with Con-A, the differences in the levels of proliferative responses between the T1 and T2 groups were significant only on 3and 4-wpv (Fig. 3A).
The proliferative responses of the T1B group, when tested with E. tarda CGH9 OMPs, displayed a significant (p<0.05) decrease on 1-wpbv and 2-wpbv when compared to 2-wpv from the T1 group (Fig. 3B). The proliferative responses of the T2B group, when tested with E. tarda CGH9 OMPs, displayed a significant (p<0.05) increase on 1-wpbv (2.766 OD) when compared to 2-wpv from the T2 group. The responses decreased significantly on 2-wpbv and reached a level similar to that of 2-wpv from the T2 group. The differences in the proliferative responses between the T1B and T2B groups were significant (p<0.05) in both 1-and 2-wpbv (Fig. 3B).
Lymphocyte proliferation of HK leukocytes of the T1 group, when tested with E. tarda CGH9 whole cells, showed an increase from control on 1-wpv, which later reduced to the level of control on 4-wpv (Fig. 4A). The lymphocyte proliferation of the T2 group, when tested with E. tarda CGH9 whole cells, showed an increase from control on all wpv. However, significant (p<0.05) differences in the levels of proliferation between control and different wpv were observed only from 2-wpv. The levels of proliferative responses in the T2 group were high on all wpv compared to the T1 group. However, the differences in the levels of proliferative responses between T1 and T2 groups were significant (p<0.05) only on 2-and 4-wpv (Fig. 4A).
The lymphocyte proliferation of the T1B group, when tested with E. tarda CGH9 whole cells, no significant changes were observed on 1-wpbv when compared to 2-wpv from the T1 group (Fig. 4B). The levels on 2-wpbv (2.121 OD) were low when compared to 1-wpbv and also to 2-wpv from the T1 group. Besides, the lymphocyte proliferation of the T2B group, when tested with E. tarda CGH9 whole cells, showed no significant changes on 1-wpbv and 2-wpbv when compared to 2-wpv from the T2 group. The differences in the levels of proliferative responses between the T1B and T2B groups were significant (p<0.05) only on 1-wpbv (Fig. 4B).

Serum antibody production
In T1 and T2 groups, indirect ELISA was performed to assess the antibody levels in the immune serum collected on different wpv. The results, expressed in OD at 492 nm, are shown in Fig. 5A. The antibody production in T1 group was reduced insignificantly up to 2-wpv (p>0.05), which then increased from 3-wpv (0.283 OD) to 6-wpv (0.364 OD) compared to control. However, the increase was significant (p<0.05) only from 4-wpv when compared to control. In the T2 group, the levels of antibodies were significantly (p<0.05) high on all wpv compared to control (Fig. 5A). Though the antibody production was increasing continuously, significant (p<0.05) differences existed only on alternative weeks up to 4-wpv. In the T1 and T2 groups, the antibody production on 4-and 5-wpv were significantly (p<0.05) high when compared individually with 1-and 2-wpv. The levels of serum antibody on 6-wpv were significantly (p<0.05) high when compared individually with 1-, 2-, and 3-wpv in both groups. The antibody production observed in the T2 group was significantly (p<0.05) high on all wpv compared to the T1 group.
In T1B and T2B groups, antibody production was observed on different wpbv. In the T1B group, the antibody production increased on all wpbv when compared to 2-wpv from the T1 group. However, the observed increase in different wpbv was significant (P<0.05) only on 3-and 4-wpbv. In the T2B group, a slight but insignificant (p>0.05) reduction in antibody production was observed on 1-wpbv when compared to 2-wpv from the T2 group, followed by a significant (p<0.05) increase on 2-wpbv. The serum antibody levels on 2-and 3-wpbv were significantly (p<0.05) high when compared individually with 1-and 4-wpbv. The serum antibody levels in the T2B group were significantly (p<0.05) high on all wpv compared to the T1B group (Fig. 5B).

Immunoprotective efficacy
After 4-wpv, the catfish from all five groups (C, T1, T2, T1B, and T2B) were challenged with a lethal dose of E. tarda CGH9. On day 1 post-challenge, the control group exhibited hyperemic skin, anorexia, abnormal movement, haemorrhages with peeled skin and exposed muscle at the site of injection, and petechial haemorrhages on the fin and skin. The bacterium E. tarda CGH9 was re-isolated from the kidney of freshly dead fish on Edwardsiella ictaluri agar and confirmed phenotypically. The challenged fish from the vaccinated groups showed comparatively less hyperemic skin and anorexic condition. The movement was normal when compared to control. Externally, inflammation with reddening at the site of injection alone was observed, which subsided and healed progressively. The immunoprotective efficacy in terms of RPS was markedly higher in T2 (89.00±15.56) followed by T1 (78.00±0.00) groups. In T1B and T2B groups, the RPS was 55.50±31.82 and 78.00±31.11, respectively (Table 1).

Discussion
Although the immune responses after vaccination with several E. tarda vaccine candidates have been investigated in different fish species Jiao et al. 2009Jiao et al. , 2010Sun et al. 2010Sun et al. , 2011Tang et al. 2010), and information regarding OMPs of E. tarda as a vaccine candidate in pangas catfish is limited. Previous reports suggested that OMPs of E. tarda are highly conserved, immunogenic, and could confer broad protection (Kumar et al. 2007;Tang et al. 2010;Dubey et al. 2019;LiHua et al. 2019). In this study, we used crude OMPs of E. tarda CGH9 as a vaccine candidate. Usually, OMPs of bacterial species contain many immunogenic components that elicit protective immunity (Dubey et al. 2019). In the T1 group without AH adjuvant, the ROB activity was significantly high on 1-wpv (0.455 OD) compared to the control. Whereas, in the T2 group with AH adjuvant, a gradual increase in ROB activity was observed up to 3-wpv, indicating stimulation of immune responses for longer protection. Contrarily, in an earlier report, the stimulating effect of a recombinant subunit vaccine candidate along with AH adjuvant on the ROB activity of P. olivaceus was observed to be significantly high only on 1-and 7-day post-vaccination (pv) but not on 14-and 21-day pv (Sun et al. 2011). It has been demonstrated that the vaccine, when conjugated with AH adjuvant, helps in long-term antigen exposure (Gupta et al. 1995;Gupta 1998) and increases NADPH-dependent ROS production and specific antibody titers (Angosto et al. 2018). Possibly, the crude E. tarda OMPs vaccine-AH adjuvant complex of the present study might have resulted in a gradual increase in the ROB activity of the T2 group. Surprisingly, no further increase in ROB activity of catfish was observed when administering a booster dose of E. tarda CGH9 OMPs with or without AH adjuvant at the same concentration.
The lymphocyte proliferation of HK leukocytes of the T1 group upon stimulation with Con-A showed a significant increase on 1-wpv compared to control and reduced thereafter to the level of control on 4-wpv. A similar trend was observed in the proliferative responses even with E. tarda CGH9 whole cells as the test antigen. On the other hand, when the E. tarda CGH9 OMPs were used as the antigen, the proliferative responses of HK leukocytes of the T1 group increased up to 2-wpv and were reduced thereafter. Contrastingly, the T2 group showed a significant increase in the proliferative responses up to 3-wpv in all the tested antigens, such as Con-A, E. tarda CGH9 OMPs, and E. tarda CGH9 whole cells. Our results corroborate the observations of LiHua et al. (2019), who reported an increased proliferation of whole blood cells on 21, 28, and 42 days post-immunization when Anguilla japonica was vaccinated with outer membrane protein A (OmpA) of E. anguillarum and Freund's incomplete adjuvant (FIA). In our study, the observed levels of proliferative responses in all the tested antigen groups were high on 2-, 3-and 4-wpv when compared to the T1 group. This suggested that the naïve lymphocytes of catfish administered with OMPs-adjuvant complexes are effectively stimulated for a longer period than the group-administered only the OMPs as a vaccine.
The T1B group showed no significant increase in lymphocyte proliferation with all tested antigens. Similarly, in the T2B group, the lymphocyte proliferation with all tested antigens showed no significant changes on 1-wpbv and 2-wpbv when compared to 2-wpv from the T2 group. Interestingly, the proliferative responses of the T2B group when tested with E. tarda CGH9 OMPs as the antigen displayed a significant increase on 1-wpbv (2.766 OD) compared to 2-wpv from the T2 group. However, the proliferative responses observed on 1-wpbv (2.766 OD) were almost similar to that of 3-wpv (2.762 OD) from the T2 group. Furthermore, the T1B and T2B groups showed no additional increase in lymphocyte proliferation when compared respectively to the T1 and T2 groups. Unfortunately, it was not possible to compare and discuss the observed results of the present study with the report of other authors, because similar reports assessed the immunoprotective efficacy of OMPs of E. tarda with or without AH as an adjuvant by in-vitro lymphocyte proliferation were scanty. The antibody-mediated humoral immunity, an important defence mechanism, was reported in O. niloticus, O. mykiss, P. olivaceus, and A. japonica against bacterial infections caused by Streptococcus iniae (Shelby et al. 2002), S. agalactiae (Pasnik et al. 2006), Flavobacterium psychrophilum (LaFrentz et al. 2003), Aeromonas hydrophila (LaPatra et al. 2010)and E. tarda (Jiao et al. 2010;LiHua et al. 2019). In the present study, the serum antibody production in the T1 group reduced up to 2-wpv compared to the control. In contrast, its level was significantly high from 1-wpv when compared to control in the T2 group. The observed reduction up to 2-wpv in the serum antibody levels of the T1 group could be the result of false tolerance or false suppression as vaccination may lead to the substantial decrease of natural antibodies, which cross-reacts with most antigens (Sinyakov and Avtalion 2009). On the other hand, this false tolerance or suppression was not observed in the T2 group, which might probably be the effect of the AH adjuvant. Additionally, Whyte (2007) described that the level of natural antibodies may vary among the fish species and it also depends on the environmental conditions in which they are present. The serum antibody production was observed to be the highest from 4-wpv to 6-wpv in both T1 and T2 groups. The observations of the present study confirmed many earlier reports performed in P. olivaceus, in which the antibody levels were observed to be highest at 4-8 weeks after vaccinating the fish with E. tarda antigen Eta21 (Jiao et al. 2009), E. tarda antigen Et49 (Jiao et al. 2010), attenuated E. tarda , recombinant E. tarda antigen rEta2 (Sun et al. 2011) and E. tarda OMPs (Tang et al. 2010). Furthermore, the T2 group recorded the highest level of serum antibody production on 6-wpv (0.581), which was >1.5-folds higher than those observed in the T1 group. This observation was in agreement with Jiao et al. (2010), which documented a significantly high level of specific antibodies in P. olivaceus vaccinated with E. tarda major antigen Et49 and AH adjuvant.
The vaccinated groups with or without AH adjuvant that administered a booster dose of E. tarda CGH9 OMPs, i.e., T1B and T2B, recorded high serum antibody production on all wpbv when compared to 2-wpv. The time points 1-wpbv, 2-wpbv, and 3-wpbv in T1B and T2B groups corresponded with the time point 3-wpv, 4-wpv, and 5-wpv of the T1 and T2 groups without the booster. In comparing the groups vaccinated with E. tarda CGH9 OMPs and AH adjuvant complex (T2 and T2B), the highest immune response in terms of antibody production was observed in the T2 group on 6-wpv. In the T2B group, the same level of antibody production was observed on 2-wpbv (0.587 OD), which was about two weeks earlier than in the T2 group. Our results corroborate the findings of Cheng et al. (2010) observed in the group of P. olivaceus booster-vaccinated with E. tarda ATCC 15947. Contrarily, the levels of serum antibody in T1 and T1B groups vaccinated with E. tarda CGH9 OMPs without adjuvant were in the same range. These observations suggested that a booster dose with AH adjuvant can able to increase the serum antibody levels quickly. Possibly, it effectively augmented the specific immunity induced by E. tarda CGH9 OMPs.
Many earlier reports by different research groups described the immunoprotective efficacy of different E. tarda vaccine candidates in P. olivaceus by examining RPS (Jiao et al. 2009(Jiao et al. , 2010Sun et al. 2010Sun et al. , 2011Tang et al. 2010). As shown in Table 1, the high RPS in the T2 group (89.00±15.56) could be attributed to the high levels of serum antibodies. Observations similar to the present work were made earlier in which a correlation between the serum antibody levels and protection against infection was observed in P. olivaceus, following intraperitoneal injection with OMPs of E. tarda in FIA (Tang et al. 2010) or with E. tarda antigen Et49 and AH adjuvant mixture (Jiao et al. 2010). The efficacy of E. tarda CGH9 OMPs and AH adjuvant mixture in terms of RPS of the present study was comparatively higher than the RPS (77.70) noted in A. japonica vaccinated with OmpA of E. anguillarum in FIA after the E. anguillarum challenge (LiHua et al. 2019). The present observations are also in close agreement with that of Jiao et al. (2010). The use of AH as an adjuvant is reportedly increased the immunoprotection rate by 92% in I. punctatus against enteric septicemia of catfish (Tyler and Klesius 1994) and 83.3% in S. maximus against viral reddish body syndrome (Fan et al. 2012), which are in close agreement with the RPS of the T2 group (E. tarda CGH9 OMPs and AH adjuvant complex). The observed RPS in T1B and T2B groups were low when compared respectively to the T1 and T2 groups. A similar observation with low RPS in I. punctatus vaccinated with a booster dose of 12.5 μg/100 μl of total E. ictaluri OMP was reported (Bader et al. 2004). Likewise, Tang et al. (2010) reported a low RPS in P. olivaceus vaccinated with a booster dose of 50 μg OMPs of E. tarda. Interestingly, in the present study, serum antibody levels in the T1B and T2B groups were almost similar to the T1 and T2 groups but still showed a lower RPS. This suggested that the catfish that received a booster dose of 50 μg E. tarda CGH9 OMPs with or without AH adjuvant might be too weak to resist the infection challenge. Because, this group was challenged immediately after 2-wpbv; while the vaccinated group without booster was challenged after 4-wpv.

Conclusion
The vaccine candidate of the present study E. tarda CGH9 OMPs was observed to be a potent immunogen that elicited a strong protective immunity in catfish by inducing different immune responses as assessed by ROB activity, lymphocyte proliferation of HK leukocytes, and serum antibody production. The E. tarda CGH9 OMPs, when administered with a cost-efficient AH adjuvant, enhanced the protective function of the tested immune factors in catfish. Although the serum antibody levels increased in the vaccinated group administered with a booster dose of E. tarda CGH9 OMPs and AH adjuvant complex, it yielded no additional increase in RPS, ROB activity, and lymphocyte proliferation of HK leukocytes at the test conditions. Our results suggested that E. tarda CGH9 OMPs and AH adjuvant complex at 50 μg and 0.02 mg concentrations/ fish, respectively, could be applied as a vaccine either with or without a booster to combat E. tarda infection in commercial aquaculture involving pangas catfish and related species.