Evaluation of Different Functional Antioxidant Groups and Protein Responses in Asian Seabass (Lates calcarifer Bloch, 1970) as an Early Indication of Fish Health in Aquaculture Farms.

The study was conducted to investigate the organ-specic antioxidants and protein damages responses in Lates calcarifer inhabiting different aquaculture farms that susceptible to different threat of pollutions. Enzymes at the front line of antioxidant system superoxidase dismutase (SOD) and catalase (CAT) evidenced to work together but respond differently in different body compartment. High SOD responses were followed with lower CAT responses observed in muscle (p < 0.05) with opposite responses exhibited by both gill and liver. The responses of SOD and CAT in muscle also showed a signicant strong correlation to each other (Setiu Wetland: 0.91, Semerak: 0.79) (p < 0.01). The glutathione-dependent enzymes, glutathione-s-transferase (GST) and glutathione reductase (GR) in body compartments responded with a strong correlation to each other especially in muscle (Tumpat; muscle: 0.89, gill: 0.95 and liver: 0.54 (p < 0.01, p < 0.05), (Setiu Wetland; muscle: 0.72 (p < 0.01) and Semerak; muscle: 0.79 (p < 0.01). Opposite results were found for both protein damages biomarkers, thiols (-SH-) and carbonyl (-CH-) in comparison to biomarkers responses. In contrary to (-SH-) inconsistent results were observed for the (-CH-) with muscle found to be most oxidised. The responses of SOD, CAT, GST, GR, thiols and carbonyl were all computed into the same scale through Integrated Biomarker Score (IBR) to classify each aquaculture farms based on biomarkers responses. There was a slight variation in deviation score (A) between organs within the range of (5-7). However, Tumpat overall showed the highest IBR score and signicantly higher (p < 0.05) than Setiu Wetland with a total score of combining responses score in all three organs (IBR: 84, muscle: 31, gill: 24, and liver: 29) followed by intermediate score in Semerak (IBR: 72, muscle: 23, gill: 23 and liver: 26) and Setiu Wetland (IBR: 59, muscle: 18, gill: 23 and liver: 18). These results indicated that the responses of antioxidant enzymes and protein damages in L.calcarifer from different organs are heterogeneous. Therefore, biomarkers should be selected based on their sequential groups in the antioxidant system for a better explanation of the oxidative stress evaluation in sh.

Introduction and the number kept increasing showing high demand for this species (FAO, 2016). Therefore, there is a need for a sustainable production of this highly commercial sh to meet the population demands.
Various studies have been conducted on the subject of sh diseases for the betterment of Asian seabass production such as vibriosis (Nor et  ). However, a study on the oxidative stress as biomarkers for the sh that acts as precursors to other diseases has rarely been the subject of interest among researchers in L.calcarifer. Antioxidant biomarkers have been widely explored in human study to detect various arrays of diseases such as diabetes, Alzheimer's, cancer, tumour and corneal diseases (Giacco & Brownlee, 2010;Galasko et al., 2012;Vallabh et al., 2017). Fish possess the same defending system as human in preventing the attack of reactive and dangerous radicals in the body system triggered by pollutants in their ecosystem (Poljšak & Fink, 2014).
The same application could be used in the aquaculture system to access the health of the sh before it becomes untreatable. However, a good understanding in the response of the antioxidant system in their body is important for further development of a good biomarker. Therefore, there is a need to identify the responses of antioxidant enzymes and protein damages in L.calcarifer that can provide early oxidative stress indication in sh and prevent the disease outbreaks in aquaculture farms.
This study emphasized on the responses of antioxidant system in the extent of their smaller functional classi cation in their defending system toward active radicals. The organs-speci c antioxidant biomarker responses was quanti ed and observed in the rst-line defense of antioxidant system and the glutathione dependent enzymes system. Further the oxidative damages that occurred in the sh were measured through protein damages biomarker. This hopes for a better understanding on these biomarkers responses in L.calcarifer for future potential use as early measurable indicators of sh health in aquaculture farm industry for a sustainable production of this species.

Sample collection
Market ready sizes of (± 500 g) L.calcarifer samples were obtained from three main aquaculture farms on the east side coast of peninsular Malaysia as shown in Fig. 1. Twenty biological replicates (n = 20) of L. calcarifer were selected from each aquaculture farm in Setiu Wetland, Semerak and Tumpat. The sh were directly selected from the cage farms and only physically healthy condition sh was selected to be used for experiment analysis. Prior to dissection, the sampled sh were anaesthetized after the selection from the cage to avoid stress during handling. After weighing and measuring, each sampled sh was immediately dissected for muscle gill and liver. All organs were weighed for extraction estimation ratio with extracting buffer and immediately stored at -80 ºC to avoid protein degradation. All animals used were strictly in accordance to EU Directive 2010/63/EU for animal experiments.

Antioxidant Activity
One individual sampled L.calcarifer represents one biological replicate for every targeted antioxidant activity, superoxidase dismutase (SOD), catalase (CAT), glutathione-s-transferase (GST) and glutathione reductase (GR). Twenty biological replicates with three technical replicates (triplicates) for each organ were used for antioxidant analysis to avoid bias and errors in accordance to Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines.
All samples were extracted using homogenizer (ULTRA TURREX) in 10 mM Tris -HCL buffer (pH 7.2) and the post-mitochondrial supernatants (PMS) was obtained after centrifugation at 15000 X g and 4ºC for one hour (Jaafar et al, 2015). Protein mass determination was done for all samples following Bradford assay using Bovine Serum Albumin (BSA) as a standard (Bradford, 1976).
All analysis was done using spectrophotometric method with different wavelength used for each antioxidant enzyme depending on the nature of each bioassay. SOD assay with xanthine oxidase as based was used in quantifying the SOD activity following the method by (McCord & Fridovich, 1969) with modi cations. 62.5 mM of potassium phosphate (pH 7.8) was prepared together with 0.125 mM EDTA, 83.3 µM hypoxanthine and 16.7 µM Cytochrome C added freshly to the buffer on the day of the assay being performed. Xanthine oxidase was added to each prepared well and absorbent readings were measured kinetically at 550 nm by using multi-mode Microplate reader (Molecular Devices model SpectraMax iD3) for every 30 seconds for 5 minutes. The absorbency readings were then used to calculate the activity of SOD by taking the reading of 50% inhibition from each sample read. CAT activity was assayed following (Aebi, 1974) by observing the decrease of absorbance due to the consumption of hydrogen peroxide (H 2 O 2 ) at 240 nm (extinction coe cient: = 0.04 mM − 1 cm − 1 ) using spectrophotometer. The decrease of absorbance was followed kinetically for 5 minutes with 30 seconds' interval.
The activity of GST was measured following the method of Habig et al. (1974) in 0.15M potassium phosphate buffer (pH 6.5) and 1-chloro-2,4-dinitrobenzene as substrate (extinction coe cient: = 9.6 mM − 1 cm − 1 ). The increase of absorbance was followed at 340 nm using multi-mode Microplate reader (Molecular Devices model SpectraMax iD3) for every 30 seconds for 5 minutes. The increase in absorbance was observed due to the conjugation of 2,4-Dinitrochlorobenzene (CDNB) with reduced glutathione (GSH). The GR activity was measured in 0.48M potassium phosphate buffer (pH 7.2) and βnicotinamide-adenine dinucleotide phosphate reduced form tetrasodium salt (β-NADPH) as substrate (extinction coe cient: = 6.2 mM − 1 cm − 1 ; (Carlberg & Mannervick, 1985). The absorbance was followed at 340 nm using multi-mode Microplate reader (Molecular Devices model SpectraMax iD3) for 5 minutes with 30 seconds' interval. The decrease of absorbance was observed due to the consumption of β NADPH to convert oxidized glutathione (GSSG) to glutathione (GSH).

Page 5/25
The replicates concept for the analysis of protein damages, thiols (-SH-) and carbonyls (-CH-) followed the concept of antioxidant activity with the same sample from biological replicates used for antioxidant activity and followed with 2 gels replicates as technical replicates for each sample. All samples were extracted using homogeniser (ULTRA TURREX) in 10 mM Tris -HCL buffer (pH 7.2) and the postmithochondrial supernatants (PMS) was obtained after centrifugation at 15000 X g and 4ºC for one hour (Jaafar et al., 2015). Protein mass determination was done for all samples following Bradford assay using Bovine Serum Albumin (BSA) as a standard (Bradford, 1976).
The thiols-containing protein and carbonyl-containing protein were quanti ed following the method of (Jaafar et al., 2015) through western blotting technique. Both thiols and carbonyl followed a similar method with a difference in the uorescent tagged used. The thiols containing protein was labelled using uorescent tagging 5-iodoacetamido-uorescein (IAF) and carbonyl-containing protein was labelled using uorescein-5-thiosemicarbazide (FTSC) in dimethyl sulfoxide (DMSO) to a nal concentration of 0.2 mM.
The samples were incubated in the dark for two hours after being tagged. 20% Trichloroacetic acid (TCA) was used to aid precipitation of protein and the tagged samples were centrifuged at 11000 X g for 3 minutes to obtain the protein precipitates. The protein precipitates were then thoroughly crushed with the aid of sonicator in ice. The crushed protein precipitates were added with 500 ul of cold acetone and incubated overnight to wash away the content of the TCA. The protein precipitates were further centrifuged at 11000 X g for 3 minutes and the excess acetone was discarded and 10 ul of sample buffer with β-mercaptoethanol was added for every 20ug of protein. The tagged protein samples were then boiled at 98 ºC for 3 minutes to separate the protein molecules into a single strand protein. 12 percent resolution gel and 6 percent stacking gel were used to resolve the protein using BioRad Electrophoresis system at a constant voltage (150V) .Two gels were assigned for each sample as replicates and the gel was scanned after completed. The gels were subsequently stained with colloidal coomasie staining with CBB G-250 overnight and preceded with gel view again after destaining. The viewed gels for both uorescent and 1DE gels stained with Coomasie blue G-250 were analyzed using Quantity one image analysis software (Bio-Rad, Hercules, CA, USA) measuring the total intensity of each lane and quanti ed as an arbitrary unit. Total optical densities for each lane were normalized with Coomasie staining for the same gel.

Integrated Biomarker Responses (ibr) Score
The deviation score (A) were all plotted as shown in (Fig. 6A, Fig. 6B and Fig. 6C) for every measured biomarker for each study site. The plot graph of the response of multi biomarkers in muscle, gill and liver for all sites only showed a slight variation between biomarkers and the sites with deviation score (A) in the range of (5-7). Tumpat overall showed the highest IBR score (muscle: 31, gill: 24, and liver: 29) for each studied organs and Setiu Wetland showed the lowest score overall (muscle: 18, gill: 23 and liver: 18). Semerak overall showed intermediate scores (muscle: 23, gill: 23 and liver: 26) as shown in Table 3.
However, overall comparison of IBR score combining all three organs showed that Tumpat was signi cantly higher (p < 0.05) compared to Setiu. Tumpat was induced with the most pollution that re ected the condition of this place to be the most polluted as compared to the other two sites based on the IBR score. Looking back at the chronology of pollution history occurred in the study sites, Tumpat was implicated with a few environmental events such as Harmful Algal Bloom (HAB) and shell sh intoxication as reported by (Lim et al., 2010: Lim et al., 2012. However, the effects of HABs are temporary and will eventually wear off. Considering the samples were retrieved in 2016, the effect that happened years before might had been worn off since it was not observed or reported happened in the same years of the sampling time. Nevertheless, observations were made during sampling time and it was found that the water system in Tumpat was blocked by a sand mound caused by erosion. This could hamper a good water circulation and trapped more pollutants induced from land considering they are the key to good water quality (Grifoll et al., 2011). Therefore, this could sensibly explain the highest response of both antioxidant and protein damages on this site.
The level of any type of pollutants was not measured or recorded from the present study. Thus on heavy metals content in bivalves showed that the metal content in the species studied was mostly below the permissible limit. Therefore, this gave a little insight into the status of these two sites from heavy metals level perspective and in line with the response of biomarkers shown by Setiu Wetland and Semerak.

Results And Discussion
First-line defense of antioxidant system SOD and CAT activity were compared according to the response level in the organs as shown in Fig. 2.
SOD activity in muscle and gill were observed to be signi cantly lower (p < 0.05) compared to the liver in Setiu Wetland. The same pattern was noticed for the response of SOD activity in Semerak with muscle and gill were signi cantly lower (p < 0.05) compared to the liver. CAT showed the same pattern of response with liver remained as the organ with the highest response. Signi cant results were observed for the response of CAT between organs for all sites (p < 0.05). As shown in Fig. 2 the response of CAT in liver was signi cantly higher compared to both muscle and gill in Setiu Wetland. Meanwhile, in Semerak and Tumpat liver showed a signi cant higher response compared to both gill and muscle with signi cant higher response of gill (p < 0.05) compared to muscle.
Further, the response level of SOD and CAT were compared as shown in Fig. 3. From the results, it showed that despite the superior SOD response compared to CAT for gill and liver in all studied sites. Muscle appeared to be in the opposite, with signi cant higher of CAT response (p < 0.05) for both Setiu and Semerak followed by a signi cant strong correlation (Setiu Wetland: 0.91, Semerak: 0.79) (p < 0.01) as shown in Table 1.  Pointing out from this, the response of SOD and CAT in muscle was evidenced to be different from sensitive organs, gill and liver.
The response in gill and liver agreed to a study reported by (Abhijith et al., 2016) in gill and liver of carp, Catla catla. From this, it explains that both enzymes at the front line of antioxidant system work together but respond differently in different body compartment as evidenced by the present study. This is further supported by the strong correlation shown in muscle. Therefore, it is very important to have these enzymes together as targeted biomarkers because their responses are correlated to each other.

Glutathione-dependent Enzymes System
The response level of both GST and GR in muscle, gill and liver was compared for each study site as shown in Fig. 4. Overall, the most response was shown in the liver for both enzymes followed by muscle being the least responded organ for all sites. A signi cant higher response of GST (p < 0.05) was observed in the liver for both sites Setiu Wetland and Semerak compared to the muscle. The response of GST in gill for Semerak also observed to be signi cantly lower (p < 0.05) compared to liver. Regarding GR, the responses in liver for all sites were signi cantly higher (p < 0.05) compared to both muscle and gill.
The response of both GST and GR were tested using Pearson correlation analysis as shown in Table 2.
There was a strong signi cant correlation observed for the response of GST and GR in all targeted tissues from Tumpat, muscle: 0.89, gill: 0.95 and liver: 0.54 (p < 0.01, p < 0.05). Strong signi cant correlations also observed in both sites Setiu Wetland (0.72) and Semerak (0.79) for muscle (p < 0.01).  GST activity was expressed in umol/min/mg protein whilst GR activity was expressed in nmol/min/mg protein. This showed that the difference of their responses was apparent. GST is amendable towards various numbers of electrophilic compounds (Habig, 1983;Armstrong, 1997;Strange et al., 2001). This allows it to catalyse non-speci c oxidants and result in a higher activity response. The presence of pollutants leads to reduction in the level of GSH and GSH-GSSG ratio and causing reduction in GR activity

Thiols Protein And Carbonyl Protein (protein Damages)
In general, a consistent pattern was observed for the total thiols containing protein for all three sites (Fig. 5). The response of oxidized thiols in the tissues of L.calcarifer was consistent with muscle remained as the least oxidized tissue followed by gill and liver for all sites. A higher signi cant result (p < 0.05) was observed in the liver as compared to gill and muscle in Setiu Wetland. However, no signi cant results were observed for Semerak and Tumpat.
In contrary to thiols, carbonyl containing protein showed inconsistent results in the targeted tissues as shown in Fig. 5. Insigni cant consistent result was observed in Setiu Wetland with liver remained as the most oxidized organ followed by gill and muscle. The same pattern of results was observed for Semerak and Tumpat with muscle as the most oxidized followed by liver and gill. Muscle was signi cantly (p < 0.05) more oxidized than gill in Semerak and gill was signi cantly (p < 0.05) less oxidized than muscle and liver in Tumpat.
The response of thiols showed that sensitive organs are more susceptible to damages as evidenced by the consistent oxidized thiols in liver for all study sites. The trend of responses for thiols is different from other targeted biomarkers in the present study considering its mechanism of action. Thiols molecules are present on the side chain of protein and they are very sensitive to the presence of ROS that tend to oxidize them (Dalle-Donne et al., 2005; Rossi et al., 2006). Therefore, the decrease in their numbers indicates damages that happened to the targeted species (Rainville, 2015 The more apparent damages indicated by high carbonyl level occurred in muscles for Semerak and Tumpat. The increase of protein damages followed by the decrease of antioxidant enzymes activity showed ine ciency of antioxidant defense (Almroth et al., 2019). Therefore, the results suggesting that the antioxidant defense in muscle is less e cient compared to gill and liver. However, enzymatic antioxidant responses were recorded in the present study and all of them were consistent respectively.

Conclusions
In conclusion, the present study managed to access the trend of biomarkers' responses in different body compartment of sh following different functionality of each enzymes and protein groups. Besides, this study also managed to classify the pollution state towards multi-biomarker responses of three main sh cage of the east side of Peninsula Malaysia. This information is bene cial to the sh farms industry for early monitoring of diseases and sustainable production of aquaculture produce. However, no speci c pollutant could be pointed out that induced responses of biomarkers. Therefore, a further study should be conducted to observe the real condition of the sites and to point out the point source pollutions to decide what pollutants might present the most in the ecosystem. Thus, this will complement the data of the biomarkers and they can further be con rmed in the laboratory to see how that certain pollution or to the extent of what concentration of that certain pollutant induces stress to a certain organism in a controlled environment.     GST and GR activity in muscle, gills and liver tissue of L. calcarifer from Setiu Wetland, Semerak and Tumpat, Each value is mean ± standard deviation, SD. n = 20*Superscripts of different letters are signi cantly different from each other at (One way ANOVA, p<0.05), n= 20