## Introduction

Today’s growing world population has led to increasing demand for aquaculture as a luxury and cheap protein source1. Correspondingly, the current aquaculture practice has shifted from extensive to semi- or intensive systems. In the year 2016, the global aquatic food production has exceeded 171 million tons2. Fish production in the first two quarters of 2017 and 2018 increased to 5.80 million tons3. Among the Indian major carps (IMCs) Catla catla is the most commonly farmed freshwater fish due to its size, good flavor, high protein content, omega-3 fatty acids, yet with fewer triglycerides, which promote brain function4; besides species like C. catla are also a cheap source of aqua-protein (about 2 US dollars/kg in countries like India). However, an intensive aquaculture system triggers a highly stressful environment that adversely affects the immune system, making the cultivated fish more vulnerable to infectious agents5. Besides, any culture system with maximum rearing density triggers frequent outbreaks of several infectious diseases, increasing the host susceptibility, virulence of the pathogen, and health-related problems6,7,8. Like other IMCs, Catla suffers from several infections, including aeromoniasis, edwardsiellosis, and epizootic ulcerative syndrome (EUS)9. Among these, Aeromonas hydrophila is a leading bacterial pathogen known to cause symptoms like haemorrhagic septicaemia, infectious dropsy, ulcerative lesion, and fin rot resulting in mass mortality10,11 affecting the quality and quantity of the size of harvest significantly. To manage these diseases, fish farmers conventionally use broad-spectrum antibiotics and chemotherapeutics, which often lead to frequent outbreaks and spread of resistant strains and environmental threats, creating further problems12,13. Vaccines are an effective prophylactic measure in aqua-practice to inhibit or control of infectious diseases; however, its success rate varies since, as they are pathogen specific14. Furthermore, the vaccines developed for intracellular fish pathogens are yet to become successful15,16.

In this regard, natural immunostimulants have become an attractive alternative. In fish disease management, they are widely accepted to sustain aquaculture, as they are biocompatible, biodegradable, often readily available, eco-friendly, and safe for human health17,18. Several immune-stimulants are known to afford disease protection by boosting non-specific and specific immune systems19,20,21,22. The use of herbals and their bioactive constituents can circumvent the use of traditional chemotherapies. In plants, polyphenols like anthraquinones, lignans, flavonoids, and aromatic acids are widely distributed; they exhibit antioxidant properties23; in human health management, the herbal medicine also play a progressive role in controlling cardiac disease24, cancer25, and viral infection26. Chrysophanic acid or chrysophanol (1, 8-dihydroxy-3-methyl-anthraquinone) that come under the anthraquinone family is widely distributed in Chinese herbs (Rheum officinale and Polygpnum cuspidatum), and exhibit anti-bacterial and anti-fungal properties27,28. Chrysophanic acid (CA) induces reactive oxygen species (ROS) production, dysfunction of mitochondria, and damage of ATP and DNA, causing necrotic cell death in human liver J5 cancer cells29,30,31. CA is also known to stimulate cytosolic Ca21 production, and cause a decrease in Dilated Cardiomyopathy (DCM) levels32. However, no detailed research has been conducted on CA’s effect with reference to immunity and cytokine gene modulations in aquatic species. This work aims to find out the effect of diets containing CA on the innate and adaptive immunity and expression of immune-related gene expression in C. catla against A. hydrophila.

## Results

### Immunological response

Both healthy and infected groups fed with the 2 CA enriched diet had increased phagocytic and hemolytic activity and SOD generation (P < 0.05) between the 6th and 8th weeks, but not in the 2nd or 4th week. Fish fed with 4 CA diet in both groups exhibited these activities at the 4th, 6th, and 8th weeks. However, the 8 CA administration significantly enhanced these parameters in the 2nd week only; no differences were observed in either the 1st week or after the 6th week (Figs. 1a, 2a, 3a). In both 2 CA groups, enhanced LP index, MPO content, bactericidal action, and antibody levels were observed between the 6th and 8th weeks. A similar trend existed in both groups fed with 4 CA from week 2 to 8. The maximum values of the immune parameters were obtained when both groups were fed the 8 CA diets between weeks 2 and 4; however, no increases were observed in the initial study period (week 1) or after prolonged periods (after week 8) (Figs. 4a,b, 5a, 6b).

Low respiratory burst (RB) activity was observed in both 2 CA and 4 CA groups in the 2nd and 4th weeks; however, the highest (P < 0.05) values were noted after week 6. In both 8 CA groups, the highest RB activity (P < 0.05) manifested between weeks 4 and 6, but these values were insignificant in the 1st week and after week 8 (Fig. 1b). The nitric oxide (NO) synthase was low in both 2 CA groups (weeks 2–6); increased significantly after the 8th week. Similarly, both 4 CA groups had slightly higher levels (P > 0.05) of NO synthase in the 2nd and 4th weeks, and significantly (P < 0.05) higher levels between weeks 6 and 8. NO synthase levels varied moderately (P > 0.05), except in week 2 (Fig. 2b).

Hydrogen peroxide (H2O2) production did not differ (P > 0.05) significantly in the 2 CA fed groups from weeks 2 to 6 but attained significant levels (P < 0.05) in the 8th week. The highest H2O2 production was observed in both 4 CA groups between the 4th and 8th weeks, and the lowest production activity (P > 0.05) occurred in the 8 CA groups (Fig. 3b). Lysozyme levels and TP levels decreased slightly (P > 0.05) in the 2nd week in the 2 CA and 4 CA groups. These values rose significantly (P < 0.05) in both 8 CA groups from 4 to 8 weeks, though increases did not vary in either the 1st or after the 8th weeks (Figs. 5b, 6a).

### Immune gene expression

The mRNA transcripts of IL-1β and 10, TLR-4, as well as MHC-I in head kidney leucocytes, were significantly low between weeks 2 and 4 when fed with 2 CA in both groups, whereas its expression was up-regulated between the 6th and 8th weeks. Feeding with 4 CA could induced IL-1β, IL-10, TLR-4, and MHC-I mRNA expression in both groups significantly from weeks 4 to 8, compared with that of the control. However, the IL-1β, IL-10, TLR-4, and MHC-I mRNA expression were significant in both groups in the 4th and 6th weeks in the 8 CA groups; yet no up-regulation was observed in the first or after the 8th week (Figs. 7a,b, 8b, 10c). The expression levels of TRAF6 and TNF-α were not significantly up-regulated in the 2 CA groups from week 2 to 6, while significant up-regulation was observed in the 8th week. Both groups fed with 4 CA showed down-regulation in the 2nd and 4th weeks, whereas up-regulation was observed in the 6th and 8th weeks. Nevertheless, the TRAF6 and TNF-α expression levels were slightly higher in the groups fed with 8 CA from week 4 to 8 (Fig. 9a,b).

The expression of INF-γ, TLR3, TLR5, Lyz-C, and Lyz-G genes showed no up-regulation in both 2 CA groups from weeks 2 to 6; however, mRNA gene expressions were up-regulated in the 8th week. The 4 CA groups had a significant induction of INF-γ, MyD88, TLR3, TLR5, Lyz-C, and Lyz-G genes expression between weeks 4 and 8, but not in the 1st week (Figs. 8a,c, 10b,d, 11a,b). Interestingly, a slight up-regulation of NF-kB and TLR2 expression was observed in weeks 2 and 4 in the 2 CA group. However, there was a significant up-regulation between weeks 6 and 8. Similar results were found with the 4 CA and 8 CA groups in the 2nd and 4th weeks; however, all the values decreased after the 6th week (Figs. 8d, 10a). The expression of NOD1 and NOD2 were not significantly improved in the 2 CA group (weeks 2–6), yet there was a sudden up-regulation in the 8th week. Significant NOD1 and NOD2 mRNA expressions were also observed in the 4 CA group after the 4th week, as well as in the 8 CA groups (weeks 4 and 6), but not in the 1st week or after the 8th week (Fig. 12a,b).

### RPS

Survival was 100% in group I (H-0 mg). In both post-challenged and un-challenged groups (groups III and IV), treated with 4 CA and 8 CA resulted in higher survival rates between 80.5 and 89%. On the other hand, post-challenged or un-challenged (groups II and VI) treated with 2 CA had low survival between 72.5 and 79%. However, post-challenged group V treated with the control diet had the least survival of 64% (Fig. 13).

## Discussion

The ability of phagocytic cells to kill the invading pathogens is an essential innate defense mechanism33. In the present study, phagocytic activity significantly improved in both groups (P < 0.05) with doses of 2 mg kg−1 CA (2 CA) between weeks 6 and 8; and in the 4 CA group between weeks 4 and 8. The phagocytic leucocytes harvested from HK of the 2 CA and 4 CA groups revealed high RB activity, attributed as a metabolic function due to CA in the 6th and 8th weeks, resulting in higher amounts of SOD synthesis. The HK leukocytes RB activity is another important innate immune function widely considered a bio-indicator of immune-competence, exclusively triggered by immunostimulants33,34. Extracellular O2 production was highly significant in the 2 CA groups (6 and 8 weeks) and the 4 CA groups (4–8 weeks). Similar results were found in rohu, in which a higher level of O2 production was reported (Days 4, 6, and 14) post-vaccination with A. hydrophila35. The O2 production in macrophages exhibits immunity after the initiation of phagocytes in fish that generate reactive oxygen products; such as H2O2 and OH, during a period of robust O2 intake, termed as RB36. This ability to kill microorganisms by phagocytes has an essential role in anti-microbicidal mechanisms in fish37. The trout HK leucocytes collected after zymosan exposure showed maximum oxygen consumption or RB activity that produced 12 nM or 13 nM superoxide anion for 107 cells38. In the present study, both groups fed 2 and 4 mg kg−1 CA diets (2 CA and 4 CA) showed no significant change in RB activity in weeks 4–6. These results were in agreement with a recent study in which a significant observance of RB activity was reported in Atlantic salmon and rainbow trout against Lepeophtheirus salmonis39. However, significantly high RB activity was observed in Pangasianodon hypophthalmus against pathogen40. Similarly, Munoz, et al.41 reported that a higher O2 level afforded protection to Dicentrarchus labrax against Sphaerospora dicentrarchi. Natural immunostimulants do not significantly affect the innate defense response in fish42,43; numerous studies have reported a higher RB activity in fish through various immunostimulants44,45. H2O2 production was significantly high in the 8th week of the 4 CA feeding, whereas the RB activity in the 8 CA group was manifested in the 2nd week. Similar increases in H2O2 production were observed in Psetta maxima and Sparus aurata via the administration of glucan46. High ROS production due to glucan may be associated with augmented bactericidal and phagocytosis, as reported in Atlantic salmon and rainbow trout47,48. The ability to generate H2O2 was reported in neutrophils due to O2 breakdown during RB in channel catfish49. However, in the present study, the H2O2 production was low in both groups fed 2 mg kg−1 CA (2 CA) from week 2 to 6. Similarly, the native or heat-denatured lectin of Abrus precatorius induced low phagocytosis levels, bactericidal, and H2O2 production in mice macrophages50.

A significantly higher proliferative response was observed in both 2 CA groups after week 6, while in the 4 CA group, the same effect was observed after week 4. Consequently, with the administration of the A. hydrophila vaccine, the proliferative response in fish increased51. A lower response was observed in the earlier stage, week 2, in the 2 CA group, suggesting the induction of low memory. However, the potential influence of leukocyte function on enhancing specific immunity related to immunostimulant is yet to be explored. Notably, NO production was directly related to the granulocytes associated with the innate immune system52. In this study, NO synthase production was low in both 2 CA groups between weeks 2 and 6, which became significant in the 8th week. However, the 4 CA groups produced a significantly higher level of NO synthase in weeks 6 and 8. Das et al.35 reported similar effects of NO production in leukocytes in rohu against A. hydrophila. Consequently, the present results suggest that CA potentially influences NO synthesis of ROS and RNS in stimulated leukocytes as an immediate response to pathogens.

Neutrophil granules release MPO enzymes during oxidative RB that produce toxic hypochlorous acids that react with pathogenic microorganisms53. MPO content was highly significant (P < 0.05) in both 2 CA groups at 6 and 8 weeks, similar to that of the 4 CA groups from weeks 2 to 8. A significantly elevated level of MPO content was reported in P. hypophthalmus against monogenean infection40. Conversely, a recent study in rohu found that MPO activity was not significantly different between healthy and infected fish54. The present results indicate that a relatively lower concentration (2 mg kg−1 CA) stimulates MPO after 6 weeks, whereas a medium concentration (4 mg kg−1 CA) stimulated MPO earlier, at week 4. Hemolytic activity was statistically influenced (P < 0.05) in the 2 CA groups in weeks 6 and 8, whereas the 4 CA groups induced MPO between weeks 4 and 8. Among various immune mechanisms, only the complement pathway has the potential to avert microbial infection. Therefore, the hemolytic exertion in fish serum is recognized as an intersperse complement pathway, which plays a vital role against infectious pathogens55. In this study, a significant hemolysin titer was obtained with the 4 mg kg−1 CA, suggesting the influence of complement pathways against pathogens56. However, the groups fed with a low dose (2 mg kg−1 CA) produced a high hemolysin titre in the later stages, whereas a high dose (8 mg kg−1 CA) produced similar results in the earlier stages. Additionally, a more recent study reported less hemolysin titre in rohu against dactylogyrid monogenean54. However, these changes may be dependent upon fish species, the pathogenicity of the microbes, or the blocking of certain microbe epitopes, capable of influencing the alternate complement cascade.

Apart from the RB mechanism, lysozyme is a lysis enzyme produced by granulocytes during non-specific oxygen-independent pathways that play an essential role in innate immunity57,58,59. In vertebrates, lysozyme is an imperative humoral component in the systemic and mucosal immune systems; it further acts as a defensive factor against pathogenic microorganisms57; lysozyme secreted by human granulocytes attach to the hyphae of Candida albicans60. These studies strongly suggest the existence of a lysozyme-conflict defence mechanism in fish against pathogens. TP is also an important compound involved in the immune system. Both lysozyme activity and TP concentration reached significant levels (P < 0.05) in both groups fed with 2 CA and 4 CA diets between the 4th and 8th weeks. Elevated serum lysozyme activity was also reported in the immunostimulant administration in rainbow trout and rohu61,62.

Bactericidal and antibody levels were statistically improved (P < 0.05) in both 2 CA groups after week 6 and in the 4 CA groups from weeks 4 to 8. The relationship between antibody production and protection against pathogens has been demonstrated in Catla against furunculosis51. However, the significance of protection or resistance to pathogen within antibodies further suggests cellular immunity following immunostimulants. Antibody synthesis depends primarily on effector-cell proliferation and its specific antibody secretion or memory cell differentiation, which representing the complex progression needed for cytokine secretion and cell cooperation63.

In the present study, immune defense system changes when infected with A. hydrophila led to both up- and down-regulation of selected immune genes. The immune-related gene expression patterns in fish have been investigated only recently; and, consequently, the understanding of immunity to pathogen is limited. IL-1β is an important inflammatory mediator in microbial infections64. In the 2 mg kg−1 CA diets, the IL-1β and IL-10 expression levels in HK were significantly low between weeks 2 and 4, and relatively high in weeks 6 and 8. The gene’s significant up-regulation was also present in the 4 CA groups in the 6th and 8th weeks. IL-1β was up-regulated in the skin of trout against Gyrodactylus derjavini65. However, in the present study, IL-1β showed a significant down-regulation in Catla infected with A. hydtophila; suggesting the release of pathogen toxins in different phases of pathogen attachment, which play a significant role in arbitrating inflammation, thereby leading to down-regulation62; an increase in IL-1β and IL-10 expression. The exact actions of CA involvement in fish against A. hydtophila may be contained within the immune system. IL-10 is involved in the synthesis of the multifunctional cytokine inhibitory factor, which promotes immune-suppressive function66, and plays a major role in undermining the proinflammatory cytokines responses; such as IL-1β, TNFα, and IFN-γ; as well as preventing tissue damage67,68. Increased IL-10 mRNA expression levels have been reported in fish due to different immunostimulants related to the transcription arising in innate immune cells, such as macrophages69,70.

TLRs facilitate inflammatory immune responses when binding to molecules with PAMPs, and trigger an inflammation-related genes expression through signal transduction71,72, also referred to as endogenous TLR ligands, which acts as an ominous early alarm to innate and adaptive immunities73. The TLR2, TLR3, TLR4, and TLR5 expressions were significantly up-regulated in the 4 CA groups after 4 weeks and in the very early and late stages in the 2 CA and 8 CA groups. Modulation of negating innate immune gene receptors, like TLRs and NLRs, plays a fundamental role in forming an Ig repository74. A recent study reported that similar functions were demonstrated in the embryo stages in different organs or tissues of L. rohita TLR2 and TLR5 genes75. The TLR4 has been recently demonstrated with the ability to recognize viral or bacterial motifs76.

TRAF6 and TNF-α were up-regulation in the HK in both 4 CA groups after the 6th week and after the 8th week in the 2 CA groups. A similar expression pattern of TNF-α was recorded in zebrafish embryos and adults against Edwardsiella tarda infection77 and in carp HK, due to LPS78. However, TRAF6 and TNF-α expression presented low in the HK in both 4 CA groups between the 2nd and 4th weeks and between the 4th and 6th weeks in the 8 CA groups. Similarly, a low TNF-α expression was reported in rohu HK against E. tarda79. The TRAF6 response to different PAMPs treatments in Epinephelus tauvina revealed its contribution in influencing immune responses80. MHC receptors are associated with antigen-presenting cells (APCs), which assist T-cells in initiating the immune system81. In HK, leucocyte MHC-I expression was up-regulated in the 2 CA groups in the 6th and 8th weeks and the 4 CA groups on or after the 4th week. The high level of MHC expression in Catla, due to A. hydrophila, suggests that MHC contains cells (macrophages) within the HK that mediate the inflammation. However, no such change in MHC-I expression was observed in the 2 CA groups between weeks 2 and 4. Gharbi et al.82 reported the MHC receptors susceptibility in Atlantic salmon to L. salmonis.

The up-regulation of NF-kB occurred in both 2 mg kg−1 CA groups in weeks 6 and 8, whereas MyD88 expression significantly was up-regulated in the same period, in the 4 CA groups in weeks 4–8, and the 8 CA in weeks 4–6. The IL-10 mechanism was described in Catla by blocking the NF-kB-signals in the HK. Association of PAMP-TLR induced oligomerization, triggered intracellular signaling cascade via recruitment of the myeloid differentiation factor 88 (MyD88)-dependent or -independent pathways83. Combined stimulation of intracellular MAP kinase pathway and Iκβ deprivation irritates the triggering of transcription factors, like AP-1 and NF-κB; which triggers proinflammatory cytokine secretion in the B and T cells leading to B cell proliferation84. In zebrafish, the TLR4-induced MyD88 dependent signals triggered the regulation of anti-inflammation85. The results herein suggest that both MyD88-dependent and independent TLR-induced signaling pathways together to regulate the activation of catla lymphocytes and lymphoid organs. NF-κB is capable of binding to Ig kappa light-chain of pathogens encountered within the B cells86. Therefore, the tight NF-κB phosphorylation and adhesion regulation are necessary to inhibit dysfunction of immune function87. Thus, the sudden increase in NF-κB levels observed in this study indicates their vital role in the pathophysiology linked to TLR signaling.

When fed with 4 mg kg−1 CA, both NOD1 and NOD2 demonstrated significant expression on or after week 4. In addition to TLRs, NLRs is also responded to microbial components88 and endogenous ligands, as a result of tissue or cellular injuries89. The TLR2 and TLR5, NOD1, and NOD2 genes were up-regulated carp’s embryonic stages90,91. The TLR and NOD constitutive expression suggest the possibility of innate immune receptors in the early stage of CA treatment in Catla.

The chicken (Lyz-C) and goose (Lyz-G) gene expressions were not up-regulated in the 2 CA groups in the 8th week and in the 4 CA groups between weeks 4 and 8. Similar Lyz-C and Lyz-G up-regulation was recorded in Japanese flounder kidney, intestine, heart, whole blood, spleen, and ovary against E. tarda92,93. Hence, the Lyz gene up-regulation may be responsible for the rise in serum lysozyme concentration in Catla fed with CA diets against A. hydrophila infection.

There was 100% survival in H-0 mg, but only 64% of survival rate was observed in the infected group fed with the control diet. Both post-challenged and un-challenged groups treated with median dose (4 mg kg−1) and high doses (8 mg kg−1) of CA had high survival (between 80.5 and 89%), but the low survival rate was observed in both groups treated with low dose (2 mg kg−1) of CA resulting in low survival (between 72.5 and 79%). Similar reports in different fishes when fed with diets enriched with various active constituents indicate the same trends94,95,96.

In conclusion, to the best of our knowledge, this is the first report on the positive influence of CA in C. catla on innate and adaptive immune responses. The optimum level of CA is 4 mg kg−1, which manifested during the 2nd week of treatment. The pathway related to the inflammation and immunomodulation of genes also triggered a similar response. The 8 mg kg−1 of CA resulted in such response much later, after the 6th week. Further comprehensive investigations are necessary to explain immune gene expressions’ ability to elucidate the mechanism of action in other fish species, with different chrysophanol doses, against different pathogens.

## Materials and methods

### Formulation of experimental diet

The basal/control diet contained maize grain, fish meal, finger millet, and pearl millet as protein sources; rice-bran and wheat-flour as carbohydrate sources; and groundnut oil cake and vegetable oil as lipid sources (Table 1). Each ingredient was finely pounded and mixed with the required water volume to make a soft bread. The prepared feed was kept in an aluminum vessel and steamed in a pressure cooker for 15 min (at 15 psi). The formulated blend was cooled at room temperature (RT); then incorporated with the pre-mix of vitamins and minerals. Four experimental diets were prepared, by mixing chrysophanic acid (CA) in four different concentrations: (1) 0 mg kg−1, (2) 2 mg kg−1 (2CA), (3) 4 mg kg−1 (4CA), and (4) 8 mg kg−1 (8CA). Pellet feeds were prepared using a manual pelletizer (2 mm). The prepared diets were immediately oven-dried at 60 °C for 12 h. and tightly packed, stored in appropriate containers, and labeled. The experimental pellet feed constitutions were analyzed using regular procedures.

### Aeromonas hydrophila

Aeromonas hydrophila (MTCC 1739) was acquired at Himedia (India), and sub-cultured at 37 °C for 24 h. in a nutrient broth. The bacterial suspension obtained was centrifuged at 3000 × g for 10 min, and the supernatant was then discarded. The remaining pellet bacteria was re-dissolved in phosphate buffered saline (PBS, pH 7.4); until an optical density (OD) of the suspension [0.5 at 456 nm at 1 × 106 colony forming unit (CFU)] was achieved using a microplate reader and stored in a deep freezer for further use.

### Fish

Healthy, Catla catla (36.7 ± 2.1 g, 20.3 ± 2.9 cm) were procured in a local farm and kept in 500 L fiber-reinforced plastic (FRP) containers, sufficiently aerated, and filtered with de-chlorinated freshwater. The fish were immediately immersion in a KMnO4 solution for 2 min to evade any dermal infection and then kept for 2 weeks under standard laboratory conditions within natural photoperiod. During the acclimation period, fish were provided the control diet (Table 1). Fecal and unfed materials were siphoned off daily to avoid the accumulation of ammonia content in the tanks. The following measurements were recorded: pH 7.2 ± 0.05, dissolved oxygen 8.8 ± 0.02 mg l−1, temperature 24 ± 1 °C, and CaCO3 190 ± 0.2 ppm.

### Experimental setup

Fish were arbitrarily strewn into eight groups of 25 fish (8 × 25 = 200 fish), in three replicates (3 × 200 = 600 fish): Group I, healthy (non-challenged) fish fed with the basal control diet (0 mg kg−1) CA [0CA or H-0 mg]; Group II, the healthy fish fed the dietary inclusion of 2 mg kg−1 [2CA or H-2 mg], Group III: fed the 4 mg kg−1 [4CA or H-4 mg] CA diet, Group IV: fed the 8 mg kg−1 [8CA or H-8 mg] CA diet; Group V, infected (or challenged) fish fed the basal control diet (0 mg kg−1) CA [0CA or I-0 mg]; and the challenged infected fish fed dietary inclusion of 2 mg kg−1 [2CA or IC-2 mg], Group VI; 4 mg kg−1 [4CA or IC-4 mg], Group VII; and 8 mg kg−1 [8CA or IC-8 mg] CA, Group VIII. Groups I through IV represented non-infected or non-challenge fish, injected with 0.5 mL PBS; whereas groups V through VIII contained fish challenged with 0.5 mL PBS containing A. hydrophila at 1 × 106 CFU. The specified diets were provided twice daily at 10:00 and 17:00 throughout the experiment.

### Blood and tissues sampling procedure

Six fish from each group were arbitrarily chosen at the end of weeks 2, 4, 6, and 8 post-challenged with A. hydrophila. Post-anesthetized (MS-222, Sigma, USA), blood samples were drawn separately from the cardinal vein via a 1 mL plastic syringe. Each blood sample was equally divided into two separate sterile tubes with and without heparin, respectively. The heparinized blood was immediately analysed for hemato-biochemical analysis and immunological study, whereas the non-heparinized blood samples were permitted to clot at RT, then preserved at 4 °C for 4 h. All samples were centrifuged (2300 × g) for 5 min at 4 °C. The resulting serum was collected and stored in individual sterile tubes (− 80 °C) for further analyses. Lastly, the anterior kidney was dissected out aseptically from each fish, and RNAlater was added (Ambion, USA) and then preserved at − 80 °C awaiting RNS extraction.

### Immunological assays

#### Non-specific immune parameter

Phagocytosis was analysed via Aeromonas hydrophila, whereas the respiratory bursts were determined by the reduction of nitroblue tetrazolium (NBT) assay (Sigma, MO, USA) to measure neutrophils reactive oxygen radical production97. Bactericidal activity was determined using a 96-well microtiter plate with 3-(4, 5 dimethyl-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT), according to the method of Kampen et al.98. Serum lysozyme levels were measured via turbidimetric assay Ellis57. Nitric oxide synthase (NOS) was analysed as prescribed by Lee et al.99. The production of hydrogen peroxide (H2O2) was measured by phenol red oxidation method using horseradish peroxidase33. The HK leukocytes proliferative reactions were measured via MTT assay100,101. The synthesis of superoxide anion (SOD) was estimated following51. Respiratory burst (RB) in terms of oxygen radical generation was measured by phagocytes using NBT reduction assay97. Total serum myeloperoxidase (MPO) and total protein contents were measured by the technique employed by Mohanty and Sahoo79 and Gornall et al.102. Serum hemolytic activity was estimated by microtitre plate54.

#### Specific immune parameters

The serum antibody levels of fish infected with A. hydrophila was determined using an indirect enzyme linked immunosorbant assay (ELISA)103 with some modifications104.

### Immune-related gene expression study

#### Total RNA extraction

Roughly 50–100 mg of HK tissues was dissected in each fish for isolation of total RNA using TRI reagent (Sigma), per manufacturer’s instruction. The total RNA concentration was determined using spectrophotometer (Nanodrop ND-1000, Thermo Scientific, USA). The sample’s purity was achieved by quantifying the ratio of OD 260 nm/OD 280 nm (1.8–2.0). The purified RNA was then used for cDNA synthesis.

#### Expression study

Two µg of total RNA, which utilized the first strand of cDNA synthesis, were obtained via an Enhanced Avian HS RT-PCR kit (Sigma) through the use of a thermocycler (Mycycler Thermal cycler, BioRad, USA). RNA was kept for 5 min at 80 °C, containing 1 µL of 2.5 µM random hexamer, then kept another 5 min at 4 °C, allowing the primers to anneal of the RNA. A mixture of 10 × MMLV-RT buffer (2 µL), 0.4 U/µL of RNase inhibitor (1 µL), 10 mM dNTPs (1 µL), DEPC water (5 µL), and RT enzyme (1.0 µL) was softly agitated, and kept for 1 h at 42 °C, followed by 10 min at 70 °C. The resulting cDNA was synthesized and kept at 4 °C for further analyses. The constitutive expression of β-actin housekeeping gene was used in both positive control and experimental sample for normalization. The size of cloned PCR products and the primer sequences used for β-actin, as well as the immune-related genes, were investigated. The PCR mixtures contained 2 µL of 10 × PCR buffer, 13.1 µL of dH2O, 1.5 µL of MgCl2, 0.5 µL of 10 mM dNTPs, 0.5 µL of 0.05 U/µL DNA polymerase (JumpStart Accu-Taq la), 0.2 µL of 10 pmol of both primers (forward and reverse), and 2 µL cDNA. The extension parameters of the PCR as followed: 95 °C for 3 min; followed by denaturation for 30 cycles for 45 s (94 °C), and 45 s for appropriate annealing temperature (Table 2), followed by additional 45 s extension at 72 °C, and a 10-min for final extension (72 °C). The synthesized PCR products were confirmed by 1.0% agarose gel electrophoresis. The relative expression profile of β-actin and immune-relevant genes were evaluated by densitometry (Gel DOC, BIO RAD Laboratories, India).

### Relative percentage survival (RPS)

Analysis of RPS was studied in all the experimental groups as mentioned in “Experimental setup”. Each experimental group was maintained triplicate, and 20 fish were used in each group. The bacterial culture and concentration of bacterial density were the same as mentioned in “Aeromonas hydrophila”. The bacterial challenge and administration of pathogen were the same as mentioned in “Experimental setup”. The survival rate was determined at the end of the experiment. Relative percentage survival (RPS) was calculated following standard formula:

$${\text{RPS}} = 1 - \left[ {\left( {\% \;{\text{mortality}}\;{\text{in}}\;{\text{treatment}}\;{\text{group}}} \right)/\left( {\% \;{\text{mortality}}\;{\text{in}}\;{\text{control}}\;{\text{group}}} \right)} \right] \times 100.$$

### Statistical analysis

Data of each blood and tissue sample were computed as mean ± SEM in triplicate. The percentages of β-actin and the immune gene amplifying products, were later determined and further examined through one-way analysis of variance (ANOVA). Differences were subsequently calculated between the treated groups using Duncan’s Multiple Range (DMR) test, in which significance was considered at P < 0.05.

### Ethical approval

The present study follows institutional guidelines mandatory for human and animal treatment and complies with relevant legislation ethical approval from the institute for conducting the research. The animal ethical committee (approval no. 791/03/b/CPCSEA) was approved by the Tamil University, Faculty of Sciences, Department of Siddha Medicine, C-4 Quarters, Thanjavur, 613 005 Tamil Nadu.