Biomarker approach for assessing chronic toxicity of Captan® herbicide using haematological, growth, endocrine and biochemical endpoints in air breathing catsh, Clarias batrachus

This study was conducted to determine the sub-lethal toxicity of Captan® on selected hematological (Hb, HCT, MCH) growth (K-factor, HSI, SGR), biochemical (serum glucose, protein), and endocrine parameters (growth hormone, T 3 and T 4 ) in Clarias batrachus under chronic exposures. Captan® was administered at predetermined exposure concentrations (0.53 and 1.06 mg/L) and monitored at day 15, 30 and 45 of the experimental periods. The experimental groups showed signicantly lower values (p < 0.05) of haemoglobin content, haematocrit, MCH in Captan® exposed sh compared to control. Serum glucose was signicantly higher (p < 0.05) in treated sh compared to the control group; reverse was the case for serum protein concentrations (p < 0.05). Assessment of growth parameters revealed signicantly higher k-factor and SGR in control sh. HSI was however higher in treated sh which highlights the possibility of liver hypertrophy and hyperplasia of liver cells due to higher exposure and uptake of the herbicide. Endocrine responses (T 3 and T 4 ) emerged as the most sensitive biomarker category, depicting modulated responses between sub-chronic exposure at day-15 and chronic responses at day-45. In general, the study ndings using these biomarkers indicate that Captan® exposures are capable of inducing stress-specic effects at the biochemical and physiological levels negatively impacting the overall health and longevity of such animals. The use of the IBR index provided a visual and easily comprehensible depiction of toxicity effects and biomarker responses in laboratory exposed sh and we anticipate a greater applicability in biomarker data from the wild which are largely heterogenous. corpuscular haemoglobin), growth biomarkers (condition factor, hepatosomatic index, specic growth rate), endocrine biomarkers (growth hormone, tri-iodothyronine, tetra-iodothyronine) and biochemical biomarkers (total serum protein, total serum glucose) in the air breathing catsh, Clarias batrachus. The responses of each endpoint were integrated into a single index of toxicity via the integrated biomarker response (IBR) protocol.


Introduction
The increase in human population and the proportional demand for food supply has in turn justi ed the massive use of technologies such as no-tillage, chemical and biological nutrition and intensive application of biocides (herbicides, insecticides and fungicides) ( (Dalvi, 1988) and in vitro signs of oxidative damage and cytotoxicity in hepatocytes (Suzuki et al., 2004). Notwithstanding the few available studies on the Captan® toxicity on some vital physiological and biochemical markers in mammals (Dalvi, 1988), information on the chronic effects on sh are particularly sparse (WHO, 1990).
Various biomolecules, including glucose and proteins in serum proteins, antioxidant enzymes e.g., SOD, CAT and GST, endocrine parameters and hematological parameters have been used as biomarkers of measurable toxic effects of exposure to stressors (Adeogun et al., 2015;Adeogun et al., 2018;Jerome et al., 2020). However, biomarker usage in the eld can be limited by data acquisition constraints which make the use of multivariate methods impracticable (Beliaeff and Burgeot, 2002). In such situations, a simple procedure that integrates and summarize biomarker responses into simpli ed interpretation is required. Thus, the complementary use of biomarkers with multiple interpretations to validate toxic responses is evolving into the use of a more simpli ed integrated biomarker response (IBR) index for stress response assessment in organisms under pollution regimes (Samanta et al., 2018;Paul et al., 2020).
Although the environmental impact of Captan® is expected to be limited by its high rate of hydrolysis (Müller et al., 2000) its continuous bioavailability implicates continuous and widespread human-usage and a resultant replenishment of environmental levels that increases risks of chronic exposures and effects (WHO, 1990). It is with this backdrop, that chronic toxic effects of Captan® was examined using critical endpoints i.e., haematological biomarkers (haemoglobin, total red blood cell count, total white blood cell count, mean corpuscular haemoglobin), growth biomarkers (condition factor, hepatosomatic index, speci c growth rate), endocrine biomarkers (growth hormone, tri-iodothyronine, tetra-iodothyronine) and biochemical biomarkers (total serum protein, total serum glucose) in the air breathing cat sh, Clarias batrachus. The responses of each endpoint were integrated into a single index of toxicity via the integrated biomarker response (IBR) protocol.

Fish maintenance and care
Adult specimens of the freshwater, air-breathing cat sh, Clarias batrachus (n= 150) used in the study were purchased from an aquaculture farm at Basirhat, District North 24 Parganas, West Bengal. It was then transported to the Aquatic Toxicology laboratory, Barasat Government College, West Bengal [weight 108.4 ± 2.6 g (mean ± SD); length 18.3 ± 0.9 cm (mean ± SD)]. Flow-through outdoor tanks (capacity 6000 L) was used during sh acclimation in the laboratory for a total period of four weeks. They were maintained in chlorine free tap water (pH 7.5-7.8), 12:12 h light-dark cycle and natural temperature. A continuous oxygen supply to the water tanks was made available by means of air pumps (3-Watt Sobo 548A Aquarium Air Pump, China). A commercial sh diet having 30-35% crude protein was fed to the sh ad libitum @ 3% of their body weight. Partial water renewal (30-35%) was done on every alternate day to ensure healthy water quality. The various experimental procedures that the sh were subjected to abided by the guidelines of Institutional Biosafety Committee, The University of Burdwan (approval no: BU/IBSC/20/Zo/34).

Experimental design
Chronic toxicity study Two sublethal concentrations of Captan® 0.53 and 1.06 mg/L, termed T1 and T2 respectively, were determined from fractions i.e., 1/10 th , 1/20 th respectively of predetermined Captan® 96h LC 50 (10.63 mg/l) (Saha, 2021). Captan® was measured using High Performance Liquid Chromatography (HPLC-Agilent 1260 In nit, Agilent, Santa Clara, CA, USA) after an initial pre-treatment of the test solution. Since the difference between nominal and measured concentration was less than 5%, nominal values were used during the study. The sh were divided into three sets (ten sh per replicate): two Captan®-treated sets and control as per randomized design (Gomez and Gomez, 1984). There was a total of four replicates per set. Stock solution (1000 mg/L) of Captan® was made and stored in clean glass reagent bottles. An overhead tank lled with deep tube well water was used for dilution. The water and Captan® were entirely replenished after every three days in order to ensure the exact concentration and desired water quality during the entire period. Food was given to the sh 3-4 times a day until there was notable visual satiation. All the bioassays and water quality analysis were done following the protocols of American Public Health Association (APHA, 2012). The standard water quality parameters assayed throughout the experimental tenure are summarized in supplementary le.

Blood collection and haematological assays
The anaesthetization of sh was done with 1% benzocaine solution in distilled water. Blood was collected at 15-day interval i.e., three times for the entire 45-day exposure period. Blood was drawn by caudal venipuncture of sh using a sterile 5 ml 22-gauge needle syringe (Dispovan, India), quickly transferred to EDTA vacutainer tubes (Becton Dickinson, USA) and tapped with ngers to prevent clotting of blood. A small portion of blood was collected without EDTA and spun in cooling microcentrifuge at 4000 rpm for 20 minutes at 4°C. A 20-200 µl capacity micropipette (Eppendorf, India) was used to draw the straw- The blood Haemoglobin (Hb) level was measured using Sahli's haemoglobinometer and 0.1 N HCl. The endpoint was noted when the colour in the sample tube matched with that of the standard tube following continuous dilution of the sample with distilled water (Dacie and Lewis, 2016). Capillary tubes were used to draw blood, spun in a micro-haematocrit centrifuge for 3-5 minutes at 6000 rpm. Post centrifugation, a scale was used to measure the length of packed erythrocytes and the entire column of blood (Bain et al., 2016). For the estimation of Total Red Blood Cell Count (TRBCC) and Total White Blood Cell Count (TWBCC), 0.02 ml of blood was added to 3.98 ml of Hayem's uid and Turk's uid respectively in a small glass tube. It was then slightly shaken to maintain a uniform suspension of cells in the solution.

Analyses of serum protein and serum glucose
The Protein Content in Serum Samples (PCSS) and Glucose Content in Serum Samples (GCSS) was estimated following the method of Bradford (1976) and glucose colorimetric detection kit (Invitrogen, Carlsbad, USA) respectively.

Calculation of IBR
Integrated biomarker responses (IBRs) using various parameters (blood, growth, endocrine and biochemical) exposed to different concentrations of Captan® was estimated and multi-biomarker response in Clarias batrachus depicted as star plots. Multi-biomarker response was analyzed according to the methodology proposed by Beliaeff and Burgeot (2002) with modi cations by Guerlet et al. (2010).
Brie y, for each tissue, the general mean (m) and standard deviation (s) of a given biomarker was calculated (including data from all treatments), and subsequently standardized to obtain Y, i.e., Y = (Xm) / s, where X is the mean biomarker value of a given treatment. Thereafter Z was calculated as Z = -Y or Z = +Y according to the expected biological effect, with "-" representing an inhibition of a biological effect and "+" representing an induction (such evaluation was based on the average baseline biomarker values). Then, biomarker scores (S) were calculated as S = Z + |Min|, where Z ≥ 0 and |Min| is the absolute value of all Y calculated for a given biomarker (including all measurements). The scores (S) of all biomarkers measured in a given treatment and tissue were depicted as Star plots. The IBRs were calculated according to the following formulas: where β= Arc tan (S i+1 sinα/ S i − S i+1 cosα) and α= 2π/ n, S n+1 = S 1 .
where, Ai is the area connecting the two scores (S), S i and S i+1 are two consecutive clockwise scores (radius coordinates) of a given star plot, and n is the number of biomarkers used for calculations. The IBR index for each biochemical parameters were then standardized to calculate the mean value of each biomarker.

Statistical Analyses
The Shapiro-Wilk test examined the normality distribution of data and data transformation was applied in case of asymmetry to conform to normality while Levene's test was utilized to check homogeneity. The Graph Pad Prism 9.1.1 computer program (Prism, USA) was employed for data analysis. Comparisons between control and exposed shes were performed by two-way ANOVA followed by Dunnett's Comparison Test to determine signi cant differences among the means (p<0.05 -Gomez and Gomez 1984). Results are summarized as mean ± standard deviation (SD). Principal component analysis (PCA), as multivariate test for association between parameters was performed using Statistica®. Difference in means across quanti ed biomarker responses was also depicted using Graph Pad Prism ®version 7. The Kaplan-Meier estimator, also known as the product limit estimator, is a non-parametric statistic used to estimate the survival of test organism during exposure period. For all analysis, statistical signi cance was ascertained at no less than p < 0.05.

Results And Discussion
Acute toxicity The Kaplan-Meier curve demonstrates that Captan® had a concentration and duration-dependent adverse impact on the overall survival rates of Clarias batrachus compared with the control group (Mantel-Cox log-rank test; p < 0.05) (Figure 1a-b). From the 96h LC 50 curve, it is observed that higher concentration of Captan® is required to achieve 50% mortality within 24h. It further reveals that this concentration to achieve 50% mortality in exposed population decreased gradually with increasing duration. This is a strong indication that Captan® is more lethal with increased exposure duration. The relationship between the 24h LC 50 and 96h LC 50 indicates the toxicity of Captan® increases by a factor of 0.73 within 96 hours of exposure (Figure 1a). However, the wider error bar for 24-hour LC 50 indicates a wider possibility of responses that could lead to varied mortality patterns (higher or lower than 50% mortality). The smaller error bars for LC 50

Hematological pro les and indices
Assessment of Hb concentration in serum of sh across exposure concentrations revealed that across exposure intervals, i.e., 15days, 30days and 45days revealed a linear decrease in Hb levels with increased exposure of Captan® from control to T2 exposure group ( g 2). Similar trends were observed for haematocrit (Hct) levels, TWBCC, TRBCC, and MCH levels in the control group and across exposure concentrations. Blood parameters are reliable indicators for sh health and physiology because they directly re ect the relationship between habitat quality and health status (Adeogun et al., 2015; Burgos-Aceves et al., 2019). This study showed signi cantly lower hemoglobin and hematocrit levels in all treatments after exposure to Captan® compared to the control group (P < 0.05). This reduction in hemoglobin may be due to the destructive effects of pesticides on tissues that produce hemoglobin, which could impact the transport and distribution of dissolved oxygen to the tissues for metabolism and induce respiratory distress in affected shes (Saravanan et al., 2011).
Along with the hemoglobin (Hb) content and the leukocyte count, the hematocrit is regarded as a key indicator of the secondary stress response, and non-physiologically low hematocrit values are considered hallmarks of anemia, a speci c pathophysiological stress response (Rebl et al., 2021). From this study, the changes in the levels of white blood cells following exposure to Captan® may be due to disturbances in the process of hematopoiesis and subsequent reduction or non-speci c immune weakening in sh (Kumar et al., 2011). This study also revealed that the number of red blood cells (TRBCC) in all treatments after exposure to Captan® was less than the control group (P < 0.05). A decline in the number of red blood cells primarily indicates the severe anemia caused by herbicide exposure, which reduces the total protein in blood plasma and has signi cant implications on the total energy balance of the body (Ramesh et al., 2009). The lower values of MCH in exposed shes compared to the control group (P < 0.05) indicate a reduction in size and quantity of hemoglobin of red blood cells, which is diagnostic of anemia in sh (Rebl et al., 2021). The reduction of MCH could also re ect a large percentage of immature red blood cells in the bloodstream.

Growth biomarkers
Assessment of growth biomarkers and indices showed that irrespective of exposure intervals, condition factor (K) was signi cantly higher in control, followed by T1 and T2. HSI also differed in a linear pattern from control, followed by T1 and T2, however HSI was higher at highest Captan® exposure concentrations at T2 and least in control group. SGR on the other hand was signi cantly higher in control group and least in T2. The condition factor is an organism-level response, with factors such as nutritional status and toxic chemical exposure causing greater-than-normal and less-than-normal weights (Adeogun and Chukwuka, 2011; Ibor et al., 2019). However, the higher condition index values in control compared to the Captan® exposed sh is considered a re ection of depletion in energy reserves because this index is positively related to muscle and liver energy content ( . Thus, the higher HSI in Captan® exposed sh compared to sh in the control group re ects an increase in size (hypertrophy) or increased number of hepatocytes (hyperplasia) due to increased xenobiotic uptake and biotransformation (Adeogun et al., 2016).
The speci c growth rate of sh is a tool mostly used in aquaculture for the estimation of sh food conversion rate over a certain period of time, and reduction in SGR can be related to the increase in pollutant-related stress (Sweilum, 2006). The signi cantly higher values of SGR in control group compared to Captan® exposure groups con rms that this herbicide impacts growth rate of exposed shes possibly by disrupted metabolism and a resultant lowered food conversion rate.

Endocrine biomarkers
Expression of the pituitary growth hormones (GH) was signi cantly higher in control group and decreased linearly from T1 to T2. Patterns of decrease from control to highest exposure group was unchanged across exposure intervals (Figure 4a). However, irrespective of exposure intervals Tri-iodothyronine (T 3 ) did not differ signi cantly between control and T1 (intermediate exposure group), but marked difference was observed between the control group and highest exposure group. At day 15, thyroxine concentration in tissue of control sh was signi cantly lower than group T1 while levels in group T2 were signi cantly higher than T1. At day 30 and 45, the trends reversed where T 4 showed highest levels in control group with decreasing levels from T1 group to T2 group.
The concentration-dependent down-regulation of the pituitary growth hormone in Captan® exposure groups is depictive of toxic effects. Studies have demonstrated negative effects of pollutants, including xenoestrogens and heavy metals on GH-mediated mechanisms, via interference with the GH receptor and/or GH transcription (Deane and Woo, 2009). Furthermore, the lower level of T 3 in sh exposed to Captan® compared to control group could be attributed to decrease in T 3 plasma concentration due to an inability of the organism to produce the optimal level T 3 or hypothalamus, pituitary and ovary interaction (Ruby et al., 1993).On day-15, the elevated levels of T 4 level recorded in exposure-group is s consistent with other species exposed to pesticide (Liu et al., 2011). The increased activity of T 4 in exposure groups compared to control group on day 15 may be indicative of sub-chronic adaptive response in exposed sh, and deployment of growth factors for repair and turnover of somatic cells. Thyroid hormones play important roles in multiple physiological functions in aquatic animals including development, growth, and reproduction (Leatherland, 2000). Another possible explanation is the disrupted synthesis of the circulating thyroid hormones, or disrupted secretion and eventual conversion of T 4 to T 3 conversion in exposed sh (Movahedinia et al., 2018). T 4 generally represents >95% of the thyroid hormone output and it is typically present in higher quantities than T 3 in the blood circulation, with the higher T 4 concentrations serving as a pool of prohormone that can be converted into T 3 by 5-iodothyronine deiodinases in target tissues (Eales, 2006;Zoeller et al., 2007).

Biochemical biomarkers
The changes associated with GCSS and PCSS levels are summarised in Figure 5. Study ndings revealed that GCSS increased linearly and signi cantly from control group to T1 to T2 across all exposure concentrations. However, PCSS increased linearly and signi cantly from control group to T1 to T2 across all exposure concentrations. The signi cant differences in glucose concentrations in plasma (p < 0.05) between the control and treatment sh, following the action of Captan®, may be considered to be the manifestation of stress. Reports show that marked glucose increase is a general response of sh experiencing pollutant effects, because sh challenged by environmental pollution may have a higher turnover of glucose, and more glucose may be produced from non-carbohydrate substrates (Levesque et al., 2002). Hepatic synthesis of detoxifying enzymes under stress conditions requires high energy levels that necessitates documented depletion of hepatic glycogen stores and increased deployment of glucose (Begum and Vijayaraghavan, 1995;Ceron et al., 1997). The observed decreased protein levels may be seen in starvation, and malabsorption or malnutrition (Pagana and Pagana, 1998)

Integrated Biomarkers Response (IBR)
In order to compare the overall stress on C. batrachus, the IBR index was applied. From Table 1, control exposure group showed higher values of the IBR index for all biomarker parameters compared to T1 and T2 exposure groups. Reverse trends where IBR index was higher in Captan® exposure groups compared to sh in control group was observed for HSI and GCSS (Table 1). This is because, unlike the case for other biomarkers, higher IBR index values for HSI and GCSS indicate adverse reaction from toxicity. As such, evident from this index, the rank of toxic effects due to Captan® exposures can be ordered as: T2>T1>T 0 (control).
The transformed data of all the studied biomarkers are presented as star plot for each site in Figure 6(ad). The star plots showing progressive increase in reduction or increase in size of shapes obtained for each experimental group inherently re ects a gradient in toxic responses. From the size and shape of the grey area within the star plot, it is clear that T2 Captan® exposure group with the smallest sized star plot and grey area was the most impacted group, followed by T1, while the control group with the largest shape and grey area was the least impacted. The reverse trends explained above where large grey areas for some biomarkers indicate adverse impact is also applicable.
The IBR approach provides a simple tool for a general description of the ''health status'' of populations, by integrating different biomarker signals (Leiniö and Lehtonen, 2005 The PCA presents the association between dependent variables (haematological, endocrine, growth and biochemical parameters) and independent or effect variables (exposure interval and exposure group) within ordination space ( g 7). Two principal components were extracted with PC 1 axis accounting for 64.7% of the total variance between parameters. From PC1, we infer that irrespective of exposure interval, that sh from the control group show signi cant positive association with hematological parameters (Hb, RBC, WBC, MCH, haematocrit), biochemical (total protein), growth (condition factor, K, SGR), and endocrine parameters (T 3 , T 4 , T 3 :T 4 ratio and GH). This positive association implies that these parameters generally showed patterns of higher concentration in sh from the control group compared to shes exposed to Captan®. On the negative axis of PC 1, the T2 exposure group showed a positive association with glucose and HSI. This indicates that these parameters showed signi cantly higher values in sh from the T2 exposure group than the T1 exposure group and shes from the control group. These ndings readily corroborate depictions from the difference in means of these parameters across exposure groups and control.
From PC 2, the T2 exposure group showed a negative association with the T 3 :T 4 ratio and T 3 ( g 7). This indicates that T 3 :T 4 ratio, and concentrations of T3 were markedly lower in the highest Captan® exposure concentration compared to shes in intermediate exposure concentrations (T1). The signi cant decrease in plasma T 3 and T 3 /T 4 ratio in sh exposed at highest concentrations of Captan® suggests the inhibitory effect of this herbicide on extrathyroidal conversion of T 4 to T 3 (Eales et al., 1999;Brar et al., 2010). Quantifying T 4 and T 3 levels has been recommended as a good approach for detecting most xenobiotic effects on the peripheral thyroid system (Zoeller, 2010). The thyroid system plays a pivotal role in the body homeostasis and functioning of the nervous, cardiovascular and reproductive systems, and of body growth control (Danzi and Klein, 2012), and the effects of thyroid hormones mainly depend on intracellular concentrations of its biologically active and most circulated form T 3 (Buha et al., 2018). As such, the depleted levels of T 3 in the highest Captan® exposures portend signi cant impairments in physiological homeostasis.

Conclusion
From this study, the depiction of biomarker responses using haematological, growth, biochemical and endocrine parameters strongly re ect the toxic potential of Captan® at different levels of biological organization. The sensitivity and modulations of thyroid hormones T 4 and T 3 at sub-chronic and chronic exposure duration highlight its potential for the accurate depiction of toxicity in the wild, where toxic responses are rarely linear due to complex relationships and confounding factors in the environment. As such, the sh thyroid system may represent a valuable sentinel in aquatic ecosystems for predicting the actions of chemicals on other vertebrates, including humans. The study ndings indicate that environmental applications of Captan® necessitate regulatory steps and public health awareness due to the risks it portends to aquatic biota and possibly humans.