Multiple biomarker responses (serum biochemistry, oxidative stress, genotoxicity and histopathology) in Channa punctatus exposed to heavy metal loaded waste water

Experiments were conducted to investigate the health of fish Channa punctatus inhabiting heavy metal-loaded waste water. Heavy metals in the order of Fe > Mn > Zn > Co > Ni > Cu = Cr were present in the waste water. Gills had high metal load followed by liver and then kidney. Albumin, albumin to globulin (A:G) ratio, triglyceride, high density lipoprotein (HDL) and very low density lipoprotein (VLDL) were found to be lower but phospholipid, low density lipoprotein (LDL), total protein, lipid and cholesterol were higher as compared to the reference. Oxidative stress markers such as superoxide dismutase (SOD), catalase (CAT), glutathione S transferase (GST) and lipid peroxidation (LPO) were significantly higher in all tissues, whereas reduced glutathione (GSH) levels were comparatively low. Damage to DNA was observed with significantly higher mean tail length of comets in the exposed fish gill cells (30.9 µm) followed by liver (24.3 µm) and kidney (20.6 µm) as compared to reference fish (5.2, 4.8 and 5.9 µm respectively). Histopathology in gill, liver and kidney also showed marked damage. Integrated biochemical, oxidative stress, genotoxicity and histopathological findings are valuable biomarkers for native fish adaptive patterns, and monitoring of water quality/pollution of freshwater ecosystems.

Accumulation of heavy metals in gill, liver and kidney of C. punctatus is given as Supplementary Table S2. In gills of exposed fish, Fe (17619 mg/kg d.w) accumulation was the highest and Ni (31.59 mg/kg d.w) the lowest. In liver also, Fe (1501 mg/kg d.w) was the maximum and Cr (23 mg/kg d.w) the minimum. Similarly, kidney has highest Fe (3554 mg/kg d.w) and lowest Cr (10 mg/kg d.w). In the reference fish Fe accumulation followed by Cu in gills, liver and kidney. In both the exposed and the reference fish, it was observed that gills had a high metal load followed by liver and kidney respectively.
Data on glucose content as well as protein and lipid profile in the serum is given in Table 1. Low glucose content was found in fishes obtained from Panethi reservoir as compared to the reference site. It has been reported by other investigators that under prolonged exposure to heavy metals serum glucose level first elevates and then declines until it attains a depleted level 13,14 . This could be due to depletion of energy reserves to cope up with stress caused by high accumulation of heavy metals. Low glucose content in chronically exposed fishes observed in polluted waters could also be due to improper gluconeogenesis. Fishes inhabiting at contaminated site showed relatively higher total protein as compared to the reference whereas albumin was low. Lower albumin may be due to liver damage as evidenced in histopathological study. Moreover, a higher serum globulin value but lower A:G ratio was observed. These observations agree well with Gopal et al. 15 . In another field study, conducted by our group, significantly higher levels of total protein and globulin but low albumin and A:G ratios were reported in C. punctatus dwelling in polluted canal with thermal power plant effluents 6 . Increase in globulin content might be due to increased production of vital proteins combating metal toxicity. The protein synthesis could also be needed to meet out the demand for the repair of damaged tissues and heightened immune response. As albumin and globulin are among the most abundant proteins in the animal kingdom 6 . Albumin is produced by liver and globulin by many parts of body 16  pressure between blood and tissue fluids 17 . In the current study globulin level was comparatively high which may reflect the heightened immunological defense response due to toxicants. A:G ratio is an index used to track changes in the composition of serum or plasma and its normal value lies between 0.8 and 2.0 in case of humans 17 . At the test site, A:G ratio was below 0.8 value which would probably indicates the onset of pathological processes. Albumin, globulin and their ratio (A:G) are used to predict the liver and kidney disorders. Furthermore, increased protein levels would not necessarily reflect the healthy status of fish rather the ratio of albumin to globulin gives the clear picture as can be seen in these chronically exposed fishes. Estimation of total protein alone to check the environmental effect would therefore, be misguiding. Lipid is also an important energy source and an essential component of cell membrane (phospholipids and cholesterol). Besides this they also play a significant role as messengers in signal transduction pathways and molecular recognition processes 18 . Hence, any changes in lipid metabolism would signal to impairment of these crucial pathways. In these chronologically exposed fishes, total lipids, total cholesterol and phospholipids levels were comparatively high. Similar observations have also been made by other investigators 19,20 . Cholesterol, phospholipids, and triglycerides combine to form total lipids; hence elevation in these components (phospholipids and cholesterol) is directly proportional to total lipids. High cholesterol content in the blood/serum could also be due to the transport of lipid from the synthesis site for subsequent utilization either through oxidation or a process of gradual instauration of lipid molecules 20 . Liver dysfunction and disturbance of lipid metabolism also favors elevation in these components. Being an important structural component of cell membrane lipids maintain fluidity, so membrane degeneration could be another possible cause of their elevation. The values for triglycerides and VLDL were low in exposed fish as compared to those in the reference. Corroborating values of triglycerides were reported in Labeo rohita 21 . Abalaka 22 also reported low triglycerides values in Clarias gariepinus under wild conditions. It has been reported that constant energy demand leads to mobilization of triglycerides since they serve as lipid depots 18,23 . A decline in triglycerides could also be correlated to their utilization in membrane biogenesis 18,23 . Decreased triglyceride may also be due to low feed intake or low absorption due to gut damage or improper synthesis in the liver. VLDL is the triglyceride-rich lipoprotein and its concentration depends on the triglyceride fraction 18,23 . HDL level was also low in the heavy metal exposed fishes. Low HDL content was also reported in Oreochromis niloticus on exposure to heavy metals 24 . It could be due to lipid peroxidation. HDL is also known to help scavenge cholesterol from extra hepatic tissues. Hence, decrease in HDL content could be correlated with increasing cholesterol levels 25 . However, LDL levels were high relative to those in reference fishes. Metwally 24 also reported a heavy metal induced rise in serum LDL. Kojima et al. 26 attribute it to changes in gene expression of some hepatic enzymes like HMG-CoA reductase (hydroxyl-methylglutaryl-CoA), which would suppress LDL-receptor gene expression.
The data of SOD, CAT, GST, GSH and LPO in gill, liver and kidney of C. punctatus is summarized in Fig. 1. There were significantly higher levels of SOD, CAT, GST but lower GSH levels in the test tissues of fishes chronologically exposed to heavy metals compared to fish from reference site. Activities of SOD (38.5 U/mg protein), CAT (42.2 nmol/mg protein/min), GST (387.3 nmol/mg protein/min) were found to be elevated in gills while GSH level (131.7 nmole/mg protein) was low compared to that in reference fish. Similarly the activities of SOD (20.3 U/mg protein), CAT (23.9 nmol/mg protein/min), GST (254.3 nmol/mg protein/min) were also elevated in the liver while GSH level (93 nmole/mg protein) was lower than in the reference fish. Similar to gill and liver, kidney also showed similar trend. Activities of SOD (54.6 U/mg protein), CAT (7.61 nmol/mg protein/min) and GST (117.6 nmol/mg protein/min) were relatively high in kidney whereas a lower levels of GSH (64.8 nmol/mg protein) were recorded in the fishes of the contaminated than those in fresh water (reference). Similar observations have also been recorded in gill, liver and kidney of other and same species of fishes such as Carassius auratus 27 , B. bocagei 28 and C. punctatus 29 . The heavy metals (Cr, Mn, Fe, Co, Ni, Cu and Zn) accumulated in these tissues are all potentially redox active suggesting an imbalance between production of ROS and their neutralization, and bearing fish is said to be under the influence of oxidative stress. This imbalance may lead to damage of tissues and cellular components, which in turn would trigger induction of antioxidant defense mechanisms 30,31 . Hence, SOD, CAT, GST and GSH can serve as sensitive biomarkers of environmental pollution in aquatic organisms. SOD converts the superoxide radical anion (O .− 2 · ) to H 2 O 2 , its activity in gill, liver and kidney was elevated compared to the reference fish. Likewise, CAT activity also increases to destroy the H 2 O 2 which could otherwise penetrate through the biomembranes and may inactivate several enzymes 32 . GST is the phase II biotransformation enzyme which catalyzes conjugation of electrophile to glutathione (GSH). We found an enhanced level of GST activity in fishes inhabiting in effluent contaminated water reservoir. GSH is also capable of scavenging ROS. Contrary to increased activity of GST, GSH levels were found to be low in gill, liver and kidney of exposed fishes. Depletion of GSH could be due to its oxidation to GSSG 31 .
There were significantly (p < 0.05) higher levels of lipid peroxidation (LPO) in gill (55 nmol/mg tissue), liver (30 nmol/mg tissue) and kidney (26.7 nmol/mg tissue) of waste water exposed C. punctatus clearly indicating the membrane damage. Other investigators have also reported LPO elevations in various varieties of fish such as Goodea atripinnis 33 , B. bocagei 28 , C. punctatus 29,34 . Level of LPO mainly depends on the availability of polyunsaturated fatty acids (PUFA) and antioxidant defense. Fishes are considered as the rich source of PUFA. The metals (Cr, Mn, Fe, Co, Ni, Cu and Zn) detected herein are all redox active, accumulated significantly in fish tissues and could contribute to the generation of ROS which are responsible for damaging lipids, proteins and DNA.
In every living being, the genome governs cell functioning in response to signals from its environment. Any physical or chemical agent and xenobiotic capable of interfering with these signals may be toxic for cell cycle, growth, division and differentiation 35 . Figure 2 [I] and [II] illustrated DNA damage in gill, liver and kidney of exposed C. punctatus. A significantly (p < 0.05) higher mean tail length was observed in exposed gill (30.9 µm) as compared to that in the reference (5.2 µm) fish. Similarly, in liver (24.3 µm) and kidney (20.6 µm) of test fish showed higher DNA damage than in liver (4.8 µm) and kidney (5.9 µm) of reference fish. Other investigators have also reported higher DNA damage due to heavy metal exposure (Fe, Cu, Zn, Pb, Cd, Cr, Mn, Co, Ni and Zn) in fishes such as C. punctatus 34 , Limanda limanda 36 , Leuciscus cephalus 37 , Ahmad et al. 38 reported that DNA integrity in gills of Anguilla anguilla was lost at 1 mM Cr concentrations without any perceptible change at 100 µM. DNA plays a vital role in processes like gene transcription, gene expression, carcinogenesis and mutagenesis 39 . It have been reported that some of these metals form DNA adducts by intercalating or covalently binding with the DNA molecule. If the adducts are misrepaired or not repaired at all before DNA replication, then they may cause gene mutations and initiate carcinogenesis in animals and humans 40 .
The histological alterations observed in the gill, liver and kidney of C. punctatus are shown in Figs 3, 4 and 5 respectively. Histopathological studies of target organs along with the studies of oxidative stress and DNA damage would give the complete picture of heavy metal hazards and their overall toxic potential in aquatic animals. The gill of reference fish showed normal anatomy of secondary lamellae. The structure of heavy metal exposed fish varied from that of the reference. Exposure to waste water ingredients resulted in increased incidence of oedema and hyperplasia in the lamellae, lamellar fusion, gill bridging, epithelial lifting, and necrosis. Gills are the main respiratory organ in fish and are covered by thin epithelium which is the site of exchange of gases, regulation of ionic and acid-base balance and nitrogenous waste excretion 41 . In the current study epithelium was found to be degenerated in the exposed fishes and got separated from the lamellar tissue. The toxic exposure also induced hyperplasia, gill bridging, lamellar fusion in order to increase the distance for diffusion across the cells to reach the bloodstream. This has also been demonstrated by other investigators 29,34,42,43 . These alterations are interpreted as the defense response of fish against heavy metals. Liver of fish is the main metabolic organ where detoxification occurs. Hence, it could also have more chances of degeneration. Liver of the reference fish showed normal hepatocytes with pancrea (hepatopancrea) whereas in the exposed liver, lipid granules, vacuolation, hemorrhage, pyknotic nuclei, congestion of blood vessel and damaged pancrea were prominent. These histological changes are associated with the response of hepatocytes to toxicants and excessive xenobiotic metabolism in liver tissue 44 . Similar observations on histological responses to heavy metal pollution have been reported in the liver of other fish species 37,45 . Kidney sections of the reference group in Fig. 5(a) showed a normal structure such as glomeruli, bowman's space with uniform renal tubules and the interstices of the tubules contain hematopoietic tissue. However, increased bowmans space, damaged glomeruli and renal tubule, necrosis, hypertrophied renal tubule, large vacuolation, hyperplasia, constricted renal tubule and glomeruli and granuloma formation in chronologically exposed fish renal tissue (Fig. 5b-d) were the major hallmarks. Necrosis, epithelial tubuli contraction, glomerular injury, reduction of renal hematopoietic system, tissue damage, glomerular constriction and proliferation of connective tissue related with metals contaminated water have also been documented in other reports [46][47][48] . Therefore, the histopathological damage may affect the normal physiological functions of the organs.
Heavy metals (Cr, Mn, Fe, Co, Ni, Cu and Zn) contaminated waste water can induce alterations in physiology of C. punctatus. They are likely to be responsible for oxidative stress, genotoxicity and certain histopathological lesions in gill, liver and kidney of this fresh water fish. C. punctatus is a suitable bioindicator to monitor water quality in the waste water reservoir.  Live samples of exposed and reference Channa punctatus (n = 26; average length 14.32 ± 0.70 cm; average weight 69.6 ± 0.45 g and n = 22, average length 13.53 ± 0.82 cm; average weight 89.25 ± 1.03 g respectively) were collected and brought to the laboratory. Condition factor (K) and hepatosomatic index (HSI%) were calculated according to the equation of Fulton 51 and Bervoets et al. 52 respectively. Both fishes were euthanized for gill, liver and kidney removal and immediately after dissection they were carefully washed with phosphate buffer. These tissue samples were immediately processed for heavy metal estimation as per standard methods of APHA 49 . For instrument calibration, working standards were prepared by diluting the stock standards (1000 ppm) supplied by Wako Pure Chemical Industry Ltd., Japan. Analytical blanks were also used. The applied analytical procedure accuracy was tested using the certified reference material Dorm-2 (dogfish muscle, National Research Council, Canada) for investigated metals. Replicate analyses of the reference materials showed good accuracy, with recovery rates for Cu 99%, Zn 99%, Fe 99.2%, Ni 98.5%, Cr 98%, Mn 97.7 and Co 97%. Gill, liver and kidney were also used for oxidative stress biomarkers, comet assay and histopathology.

Ethical statement for animal experimentation.
Glucose, protein and lipid profile of serum. Blood was collected from live fishes through cardiac puncture, kept it stand for some time and thereafter, centrifuged at 3500 rpm for 10 min to obtain serum. Glucose levels were estimated using the kit (Accurex Biomedical Pvt. Ltd., India) and absorbance was read at 505 nm.
Total protein was determined as per standard method of Bradford 53 taking BSA as a standard. Albumin content was also measured using the kit (Siemens Ltd., Gujarat, India) and absorbance was read against blank at 628 nm. Globulin content was obtained after subtracting albumin from total protein content. Albumin to Globulin (A:G) ratio was also calculated.
Serum total lipid was quantitated using the diagnostic Kit (3830 Valley Centre Dr. San Diego, CA). Total cholesterol was estimated using the cholesterol diagnostic kit (Transasia Bio-Medicals, India) and high density lipoprotein (HDL) cholesterol using kit HDL-C (Siemens Ltd., Gujarat, India). Triglyceride was determined by using the Siemens kit (Siemens Ltd., Gujarat, India). Phospholipids were calculated by the method previously described in Javed and Usmani 6 . The activity of glutathione S transferase (GST) was measured by the method of Habig et al. 57 . The assay mixture contained 2.8 mL of sodium phosphate buffer (0.1 M, pH 6.5), 0.1 ml reduced glutathione and 50 µl sample. GST activity was monitored at 340 nm for 3 min at an interval of 1 min by the addition 50 µl of 1 mM 1-chloro 2, 4 dinitrobenzene (CDNB).
The level of reduced glutathione (GSH) was estimated by method of Jollow et al. 58 with minor modifications. Equal amount of homogenate sample and sulphosalicylic acid were mixed and incubated at 4 °C for 1 h and centrifugation at 12000 rpm for 15 min. The supernatant (0.2 ml) was mixed with 2.6 ml of potassium phosphate buffer (0.1 M, pH 7.4). The reaction was initiated by the addition of 0.2 ml 5, 5′-dithiobis-2-nitrobenzoic acid (DTNB) and the absorbance was monitored at 412 nm. The extent of lipid peroxidation (LPO) was measured by the standard protocols of Buege and Aust 59 involving the measurement of total malondialdehyde (MDA) which is the major product of lipid peroxidation. The reaction mixture contains 0.5 ml homogenate sample, 0.5 ml TBA (0.67%) and 0.5 ml TCA (3%). Total reaction mixture was kept in boiling water bath for 20 min, centrifuged at 4000 rpm, 4 °C for 10 min and sample was read at 530 nm. Genotoxicity (Comet Assay). Immediately after dissection, small sections of gills, liver and kidney tissues were transferred into RPMI medium and cellular dissociation was done based on Cavalcante et al. 60 . The comet assay was then performed by electrophoresis under alkaline conditions following the protocol of Singh et al. 61 , with minor modifications. Cells were embedded in an LMPA sandwich on frosted slides. To remove cellular proteins, slides were submerged in cold lysis buffer and stored at 4 °C for 1 hr in dark. The slides were then allowed to DNA unwinding in alkaline electrophoretic running buffer. Then, electrophoresis was conducted at 0.74 V/ cm and 300 mA for 40 min. After electrophoresis, the slides were neutralized and stained with ethidium bromide. The slides were visualized and scored by using Olympus fluorescent microscope (CX41) integrated CC camera with an image analysis system (Komet 5.5, Kinetic imaging, Liverpool, U.K.). The comet images of 50 cells (25 from each replicate slide) for each sample were scored at a magnification of 100× . Comet tail-length (migration of DNA in µm from its nucleus) was chosen as the parameter to assess the nuclear DNA damage.
Histopathology. For histopathological studies, gills, liver and kidney tissues of exposed and reference fishes were dissected out and fixed in Bouin's fluid. They were processed for paraffin wax embedding, cut into 4 µm thick sections, and stained with haematoxylin and eosin (H & E) exactly as described by Humason 62 . The slides were examined using a light microscope (Leica DM 2500).

Statistical analysis.
All values were given as mean ± standard error of mean (SEM). Statistical differences among the means of reference and the exposed were determined using Student's t test using SPSS (version 17) and Duncan's Multiple Range Test (DMRT) 63 . Levels of significance were established at p < 0.05 and p < 0.01. Assumptions of normality (Shapiro-Wilk test) and homogeneity (Levene's test) of data were verified 64 . material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.