Water pH and hardness alter ATPases and oxidative stress in the gills and kidney of pacu ( Piaractus mesopotamicus )

ABSTRACT This study aimed to assess the effects of low and high water hardness in interaction with different water pH in pacu (Piaractus mesopotamicus). Pacu juveniles were subjected to low (50 mg CaCO3 L-1 - LWH) or high water hardness (120 mg CaCO3 L-1 - HWH) at water pH of 5.5 (acidic), 7.5 (circumneutral) or 9.0 (alkaline) for 15 days. Gills and kidneys were collected (days 1, 5 and 15). Gill Na+/K+-ATPase (NKA) and vacuolar-type H+-ATPase (V-ATPase) activities were higher in alkaline pH with HWH on day 1. Gill and kidney NKA and V-ATPase activities were higher in acidic pH with LWH on day 15. Gill NKA activity of pacus under alkaline pH with LWH was higher than those exposed to HWH. Reduced antioxidant capacity in the gills and kidney and enhanced thiobarbituric acid reactive substances (TBARS) levels were demonstrated in fish exposed to acidic or alkaline pH, mainly with LWH. HWH increased glutathione-S-transferase (GST) activity and reduced TBARS levels in the gills and kidney. On day 15, GST activity was increased at acidic pH with LWH. In conclusion, circumneutral pH presents less oxidative stress and fewer variations in ATPases and HWH reduced deleterious effects in fish exposed to acidic or alkaline pH.


Introduction
The relationship between water pH and hardness influences fish development and welfare (Baldisserotto, 2011;Copatti et al., 2011a,b). Water hardness and pH are interrelated water quality parameters and they present profound effects on fish productivity and physiology and therefore deserve special attention . Hardness is important since the dissolved minerals (more specifically calcium and magnesium) are essential in the biological processes of aquatic animals, for example, bone and scale formation in fish (Guerreiro et al., 2004;Canario, Flik, 2007). Calcium 2 e190032 [2] is crucial for osmoregulation and maintenance of heart, muscle and nerve function (Parra, Baldisserotto, 2007;Boyd et al., 2016). Freshwater fish survival is greater at very acidic or alkaline water pH when fish are in high rather than in low water hardness (Baldisserotto, 2011). Calcium is important at stabilizing tight junctions in the gills and reduces their permeability, consequently contributing to osmoregulatory homeostasis and survival of fish by limiting diffusive branchial ion loss to the water by paracellular route (Baldisserotto, 2011;Chasiotis et al., 2012). Apparently, one of the survival mechanisms of species living in waters of very low hardness is a high affinity for calcium, which would contribute to closing paracellular junctions in the gills (Gonzalez et al., 1998).
In addition to the potential detrimental effects of low water hardness on fish osmoregulation, conditions leading to water acidification or alkalinisation may result in the decline of fish populations due to changes in osmoregulatory homeostasis (Gonzalez et al., 1998;Parra, Baldisserotto, 2007;Zahangir et al., 2015). Osmoregulation in freshwater teleosts is essential to maintain cell volume and prevent loss of ions (Evans et al., 2005). However, several biological or environmental processes may modify water pH and compromise cellular osmoregulation. For instance, photosynthesis and pond fertilisation increase water pH, while acidic rainfall decreases water pH (Boyd et al., 2016). In addition, acidic waters can promote ionic imbalance due to high concentrations of H + , which causes the efflux of Na + and K + ions (Aride et al., 2007). This results in a greater energetic expenditure, which compromises energy destined for growth to the osmoregulatory mechanisms (Bolner, Baldisserotto, 2007;Lemos et al., 2018a). Exposure to very alkaline pH may inhibit Na + and Cltransport by the gills, causing disturbances in ion influx (Yesaki, Iwama, 1992). In addition, the main problems in alkaline waters seem to be the inhibition of ammonia excretion and increase of CO 2 excretion in gills (Wood, 2001).
The transfer of ions is driven by highly specialized cells of the branchial epithelium called ionocytes (or chloride cells or mitochondrion-rich cells) (Evans et al., 2005), which, if damaged, impair the transport of ions in fish (Christensen et al., 2012). Numerous mitochondria in these cells are thought to supply the ATP for ion-transport proteins to drive the vectorial transport of ions as part of ion and acid base regulation (Marshall, 2002). Therefore, to maintain high ion levels in the blood relative to the environment, freshwater teleosts take up ions by active transport driven by ATP located in the apical and basolateral membranes of ionocytes and pavement cells (Wilson, Laurent, 2002).
Na + K + -ATPase (NKA)and vacuolar-type H + -ATPase (V-ATPase) are considered as indexes of the ionoregulatory capacity status in gills and kidney of teleosts (Evans, 1993;Marshall, 2002;Perry et al., 2003). NKA is an ionoregulatory enzyme expressed in animal cells to create an electrochemical gradient providing the driving force for ion transport in osmoregulatory organs, including fish gills and kidney (Whittamore, 2012). V-ATPase is a multitasking enzyme which provides the driving force for NaCl uptake in freshwater teleosts and is essential to this the linkage indirectly with an epithelial Na + channel through which Na + can be driven down its electrochemical gradient into the cell (Marshall, 2002). Ionocytes participate in response regulation when fluids become alkaline (alkalosis), and it is possible that exposure to alkaline pH also stimulates the proliferation of ionocytes (Goss et al., 1994). During respiratory acidosis, there is an increase in V-ATPase activity in kidney, which ensures that HCO 3 levels accumulate in the body fluids to restore pH (Perry et al., 2003). An increase in the rates of O 2 uptake led to even greater ion losses probably due to distortion and widening of the gill tight junctions (Gonzalez, McDonald, 1994). Thus, during osmoregulation alterations, macromolecules may experience damage from oxidative stress as a consequence of the changes in cell volume and membrane rearrangement (Rosas-Rodríguez, Valenzuela-Soto, 2010).
Oxidative stress, i.e. the imbalance between pro-and anti-oxidants, is characterised by environmental variables in fish in which antioxidant defence systems are ineffective at neutralising reactive oxygen species (ROS) formation, thus affecting cellular function via the oxidation of proteins, nucleic acids and lipids (Jones, 2006;Stoliar, Lushchak, 2012). Additionally, ROS can break down into free radicals that can perpetuate the destructive cycle of lipid peroxidation (LPO) chain reactions in cell membranes of various organ systems (Martínez-Álvarez et al., 2005), which may affect osmoregulation. To prevent tissue damage in situations when there is the danger of inordinate accumulation of ROS, animals have developed a defence system to rapidly break down ROS (Birnie-Gauvin et al., 2017). In this sense, fish tissues have antioxidant defence systems (e.g., Glutathione-S-transferase -GST; Total antioxidant capacity against peroxyl radicals -ACAP) to protect them from oxidative stress and to maintain the cellular redox balance (Do Carmo et al., 2018).
The gills and kidney are complex organs that have a central role in fish physiology and contribute to maintain osmotic balance and acid-base regulation in freshwater teleosts (Wood et al., 1999;Duffy et al., 2011). As the ionic composition of freshwater fish blood is always different from that of the surrounding water, the transcellular transport of solutes across the gills and kidney contributes significantly to the overall maintenance of salt and water balance in fish (Evans et al., 2005). The gills are the first target of acidic or alkaline water, since they are intimately in contact with the environment (Atli, Canli, 2011). The parenchyma of gills is a heterogeneous epithelium that is directly exposed to the surrounding aqueous environment. The gill epithelium presents a large surface area for the movement of biological material between blood and water, and is simultaneously involved in respiration, iono/osmoregulation, acid-base balance and nitrogen waste excretion (Evans et al., 2005). The kidney is the primary organ for the elimination of water and is particularly important for freshwater species due to the efficient ion reabsorption mechanisms to minimise ion loss (Atli, Canli, 2011). The kidney contributes to the maintenance of hydromineral balance by assisting in the elimination of excess water and the reabsorption of systemic salts (Duffy et al., 2011), complementing the function of the gills. Obviously, the main function of the kidney in freshwater teleosts is to produce dilute urine, allowing for the excretion of the excess body water that has entered by osmosis through the gills (Bolner, Baldisserotto, 2007). Considering the relevance of these organs for osmoregulation and antioxidant capacity, the gills and kidney were the focal points of study.
Pacu (Piaractus mesopotamicus (Holmberg, 1887)) is a Neotropical Freshwater Characin native to the Paraná, Paraguay and Uruguay basins with herbivorous/omnivorous habits, and is widely cultured in fish farms in South America (Valladão et al., 2016). A recent study found that high water hardness (HWH) has a protective effect on pacus exposed to acid or alkaline waters because HWH improved biochemical and physiological variables . However, the oxidative stress (thiobarbituric acid reactive substances (TBARS), glutathione-S-transferase (GST) and total antioxidant capacity against peroxyl radicals (ACAP)) and ATPase responses (NKA and V-ATPase) to variations in water pH and hardness are still unknown for pacu. Therefore, the present study was designed to investigate the effects of LWH and HWH in interaction with circumneutral, acidic and alkaline water pH in the pacu through the analysis of oxidative stress and ATPase activity in the gills and kidney. We hypothesised that an increase in water hardness may reduce ATPase activity and improve antioxidant capacity in pacu juveniles exposed to acidic or alkaline pH.

Material and Methods
Experimental conditions. Pacu juveniles (39.56 ± 2.64 g) were purchased from a fish farmer in Ajuricaba, Brazil and brought to the Continental Aquaculture Laboratory at the Universidade Federal do Rio Grande (FURG), Rio Grande, Brazil. Juveniles were maintained for ten days in 18 continuously aerated 250-L tanks in six recirculating aquaculture system (12 fish per tank, 216 fish in total), with biological and mechanical filters, water hardness of 50 mg CaCO 3 L -1 and water pH 7.5. The same system used for maintenance was used for the experimental procedures, except for water pH and hardness, which were maintained according to the specified treatments.
The design was organized into six treatments and three replicates for 15 days, under the following experimental conditions: low water hardness (50 mg CaCO 3 L -1 -LWH) or high water hardness (120 mg CaCO 3 L -1 -HWH) at water pH of 5.5 (acidic), 7.5 (circumneutral) or 9.0 (alkaline) according with Copatti et al. (2019). Experiments were run with catholyte pH automatically controlled (Tekna TPR 603, USA) via feedback by using 50% NaOH or 98% H 2 SO 4 (Copatti et al., 2011a). In HWH, water hardness was increased by the addition of di-hydrated calcium chloride (CaCl 2 .2H 2 O) according to Copatti et al. (2011a). We used this compound because its solubility is much higher than CaCO 3 , for example.
These water pH values were chosen because in silver catfish (Rhamdia quelen) exposed to water pH 5.5, 7.0 or 9.0 with LWH (30 or 60 mg CaCO 3 L −1 ) or HWH (120 or 180 mg CaCO 3 L −1 ) the survival was higher than 92.7% (Copatti et al., 2011a). Previous study verified that at zero water hardness silver catfish mortality was superior to 62 and 40% at pH 7.0 and 8.0 (Copatti et al., 2001b). This demonstrates that zero water hardness can be lethal to fish, and then it was not evaluated in our study.
The fish were fed commercial food (Supra Acqua Line ® , São Leopoldo, Brazil, 320 g kg -1 crude protein; 3,600 kcal digestible energy) twice a day (7:30 a.m. and 5:30 p.m.) until apparent satiety. The water that was lost due to evaporation or cleaning of the tanks to remove excess residues (faeces and feed) was replaced with fresh water under the respective treatment conditions. The methodology of this experiment was approved by the Ethical and Animal Welfare Committee of the Instituto de Biologia at the Universidade Federal da Bahia, Salvador, Brazil (number 29/2015). Vouchers were deposited in the ichthyological collection of the Museu Oceanográfico of FURG (catalog number 2929).

Sample collection.
For the gills and kidney analyses, three fish from each tank (n = 9 per treatment) were randomly sampled on days 1, 5 and 15 to verify possible deleterious effects on acute, medium and longer exposure under the experimental conditions. Fish were immediately euthanised with a lethal concentration of hydrochloride benzocaine (50 ppm), because benzocaine did not cause oxidative stress (Stringhetta et al., 2017), and samples of the gills and kidney were collected. Prior to tissue collection, animals underwent 24 h of fasting.
Samples of the gills and kidney were stored in an ultrafreezer (-80°C) and then homogenised (1 : 5; w/v) in a Tris-HCl (100 mM, pH 7.75) buffer with EDTA (2 mM) and Mg 2+ (5 mM) (Da Rocha et al., 2009). Afterwards, the homogenate was centrifuged for 20 min (10,000×g, 4°C); the supernatant obtained was used for all analysis. The protein concentration was determined using the Coomassie blue staining method, where bovine serum albumin was used as standard protein, according to Bradford (1976).

e190032[4]
Analysis. NKA and V-ATPase activities were measured simultaneously. NKA activity was analysed using the method described by Gibbs, Somero (1989) and V-ATPase activity was analysed according to Toni et al. (2014). The data are expressed as mmol ADP mg protein -1 h -1 .
Total antioxidant capacity against peroxyl radicals (ACAP) was determined through the detection of ROS in samples treated or not with a peroxyl radical generator according to Amado et al. (2009). All samples were first diluted with homogenisation buffer to 2.0 mg protein mL -1 and then exposed to peroxyl radicals generated by the thermal (37°C) decomposition of 2,2′-azobis (2 methylpropionamidine) dihydrochloride (ABAP, 4 mM). Peroxyl radicals reacted with a fluorescent substrate (2′,7′ dichlorofluorescein diacetate-H 2 DCF-DA) and fluorometry (excitation 485 nm; emission 520 nm) were measured on a microplate reader (Victor 2, Perkin Elmer) with readings every 10 min for 70 min. The data were expressed as the relative area, where a higher relative area (the difference between the ROS area with and without ABAP relative to that without ABAP) means a lower antioxidant capacity.
The lipid peroxidation levels were measured according to Oakes, Van Der Kraak (2003) by thiobarbituric acid reactive substances (TBARS) content. The TBARS content was determined by the quantification of MDA (malondialdehyde) following the protocol described in detail by Maltez et al. (2018). The data are expressed as nmol TMP (tetramethoxypropane (ACROS Organics)) mg wet tissue −1 .
Statistical analysis. The results are expressed as the mean ± standard error of the mean (SEM). Levene's test demonstrated the homogeneity of data variances. Data were then compared using a two-way ANOVA (pH vs. hardness) followed by Tukey's post hoc test within the same sampling day. Differences were considered significant at P < 0.05.

Results
ATPase activity. No mortality occurred in any group. On day 1, gills NKA and V-ATPase activities in pacus exposed to alkaline pH with HWH were significantly higher than those subjected to other pH values at the same hardness (P < 0.05). On the same day (day 1), gill V-ATPase activity in pacus subjected to acidic pH with LWH was significantly higher than those exposed to HWH at the same pH (P < 0.05) (Tab. 1).
Tab. 1. Na + /K + -ATPase (NKA) and vacuolar-type H + -ATPase (V-ATPAse) activities (mmol ADP mg protein -1 h -1 ) in the gills and kidney of pacu (Piaractus mesopotamicus) juveniles submitted to different pH and water hardness at different times. Note. LWH = low water hardness (50 mg CaCO 3 L −1 ); HWH = high water hardness (120 mg CaCO 3 L −1 ). Data are presented as the means ± SEM (n = 9 fish per treatment). Different uppercase letters indicate statistically significant differences between pH at the same hardness (P < 0.05). Different lowercase letters indicate statistically significant differences between hardness at the same pH (P < 0.05). On day 5, kidney V-ATPase activity of fish subjected to alkaline pH with LWH was significantly higher than those exposed to the other treatments at the same water pH or hardness (P < 0.05) (Tab. 1).
On day 15, in general, pacus exposed to acidic pH with LWH presented gill and kidney NKA and V-ATPase activities significantly higher than other treatments at the same water pH or hardness (P < 0.05). On the same day (day 15), gills NKA activity of pacus exposed to alkaline pH with LWH was significantly higher than those subjected to alkaline pH with HWH (P < 0.05) (Tab. 1).
Oxidative stress. On days 1 and 5, gill GST activity in fish exposed to circumneutral pH with LWH was significantly higher than in those exposed to the other treatments at the same water pH or hardness (P < 0.05) (Fig. 1a). In addition, on day 1, gill GST activity in juveniles subjected to alkaline pH with LWH was significantly lower than those subjected to HWH at the same pH (P < 0.05) (Fig. 1a). On day 5, gill GST activity in juveniles subjected to acidic pH with HWH was significantly higher than those subjected to LWH at the same pH (P < 0.05) (Fig. 1a). At the end of the exposure period (day 15), gill GST activity in pacus exposed to acidic pH was significantly higher than those exposed to circumneutral or alkaline water pH (P < 0.05) (Fig. 1a).
On day 1, kidney GST activity in pacus subjected to acidic pH with LWH was significantly lower than those subjected to acidic pH with HWH (P < 0.05) (Fig. 1b). On day 5, kidney GST activity in pacus exposed to alkaline pH was significantly lower than those exposed to other water pH levels (except for circumneutral pH with HWH) (P < 0.05) (Fig. 1b). At the end of the exposure period (day 15), kidney GST activity in fish subjected to acidic pH with HWH was significantly higher than those subjected to circumneutral pH (P < 0.05) (Fig. 1b).
Fish exposed to alkaline pH with HWH on day 1 showed significantly lower gill ACAP (higher relative area) compared to acidic pH with HWH (P < 0.05) (Fig. 2a). Fish exposed to acidic pH with LWH on day 5 showed significantly higher kidney ACAP (lower relative area) compared to those exposed to other water pH levels at the same hardness (P < 0.05) (Fig. 2b). No significant differences were registered among treatments for gill ACAP on days 5 and 15 or kidney ACAP on days 1 and 15 (Fig. 2 a and b).
On day 1, gill TBARS levels in pacus subjected to alkaline pH with LWH were significantly higher than in fish subjected to other treatments at the same water pH or hardness (P < 0.05) (Fig. 3a). On day 5, gill TBARS levels in pacu exposed to alkaline pH, with HWH, were significantly higher than in pacus exposed to other treatments at the same water pH or hardness (P < 0.05) (Fig. 3a). No significant differences were observed among treatments for gill TBARS levels on day 15 (Fig. 3a).
Pacus exposed to alkaline pH with LWH on day 1 showed significantly higher kidney TBARS levels than those exposed to other water pH levels at the same hardness (P < 0.05) (Fig. 3b). On day 5, juveniles subjected to acidic or alkaline pH with LWH had significantly higher kidney TBARS levels than those subjected to circumneutral pH at the same hardness (P < 0.05) (Fig. 3b). In addition, at the same day (day 5), kidney TBARS levels in fish exposed to acidic pH with LWH were significantly higher than of those exposed to the same pH with HWH (P < 0.05) (Fig.  3b). On day 15, pacus exposed to alkaline pH with LWH showed significantly higher kidney TBARS levels than those exposed to circumneutral pH with LWH or alkaline pH with HWH (P < 0.05) (Fig. 3b).

Discussion
In the present study, pacus apparently showed tolerance to changes in water pH, because no fish exposed to the experimental pH (5.5, 7.5 or 9.0) died throughout the experiment. In addition, in the current study, both acidic and alkaline pH were harmful to juveniles, mainly with LWH, as shown by deleterious effects in relation to ATPase and antioxidant levels in fish kept under these conditions. Copatti et al. (2011a) stated the protective effect of HWH occurs because the high ion loss is reduced by increasing the water hardness in acidic or alkaline pH, which could cause an improvement in the osmoregulation and antioxidant capacity of the fish in these water pH values. In this sense, our initial hypothesis was that an increase in water hardness would reduce ATPase activity and improve antioxidant capacity in pacu juveniles exposed to acidic or alkaline pH. This hypothesis was confirmed, because the ATPase levels and oxidative stress variables in pacus exposed to acidic (mainly) or alkaline pH under HWH were reduced or showed fewer changes compared to fish subjected to circumneutral pH. The effect of HWH on increased fish survival in acidic and alkaline pH has been reported in previous studies because their ion loss was reduced with HWH (Townsend, Baldisserotto, 2001;Parra, Baldisserotto, 2007;Copatti et al., 2011a).

Fig. 3.
Thiobarbituric acid reactive substances (TBARS) content (nmol TMP mg wet tissue -1 ) in a. gills and b. kidney of pacu (Piaractus mesopotamicus) juveniles under different water hardness and pH at different times. LWH = low water hardness (50 mg CaCO 3 L -1 ); HWH = high water hardness (120 mg CaCO 3 L -1 ). Data are presented as the means ± SEM (n = 9 fish treatment -1 ). Different uppercase letters indicate statistically differences between pH at the same hardness (P < 0.05). Different lowercase letters indicate statistically differences between hardness at the same pH (P < 0.05).

e190032[7]
An increase in NKA and V-ATPase activities in the gills and kidney may be attributed to the specialised structure located in these tissues in fish. So, this increase in ATPase activities could then lead to physiological and ionoregulatory alterations (Copatti et al., 2015). In the kidney of teleosts, ATPase provides energy for ion transport, because solute and water absorption usually occurs via the retention of Na + , driven by ATPases in the basolateral membrane (Evans, 1993); an increase in ATPases in the kidney can be the result of its pivotal role in both detoxification and osmoregulation (Atli, Canli, 2011). This is consistent with the increase in NKA and V-ATPase activities in the gills and kidney of pacu juveniles observed in the current study at different times during acidic or alkaline pH exposure. This has been verified mainly in LWH in comparison to HWH, because the greater ion loss would induce higher NKA and V-ATPase activities in an attempt to regulate ion levels. Thus, higher ATPase activities would also contribute to reduce ion loss. In addition, this mechanism of ion transport could be better elucidated through morphological analysis. Therefore, it is suggested that in future studies, morphological studies should be carried out to better understand the ion transport mechanisms with the ATPases activity.
The gills and kidney of fish tend to have normal functioning affected under exposition to very acidic or alkaline waters (Wood et al., 1999;Bolner, Baldisserotto, 2007;Parra, Baldisserotto, 2007). Wilkie et al. (1999) showed that alkaline water causes transient decreases in the capacity of the ion transport system by directly acting on the gill's respective Cl − and Na + transport sites. It seems probable, therefore, that exposure to alkaline pH with LWH stimulated the proliferation of ionocytes or increased their activity in the gills and kidney in pacu, preventing the impaired active transport of ions through the epithelium of these organs.
The primary effects of acid water exposure on fish is increased rates of passive Na + losses (Kwong et al., 2014). Na + uptake in freshwater fish is primarily coupled to H + secretion through the actions of Na + /H + exchanger (not evaluated in the current study) and V-ATPase (Evans et al., 2005), so a reduction in water pH would reduce the gradient to drive Na + influx. The increased activity of NKA and V-ATPase in gills and kidney could increase acid excretion, contributing to the necessary enhancement of HCO 3 − reabsorption (Perry et al., 2006). In addition, in acidic waters, there is a reduction in NKA and V-ATPase activities in the gills and kidney, probably due to the generation of ionocyte lesions, which could increase ion loss (Wood, 2001;Evans, 2011).
The increased ion losses during acid exposure are thought to be largely associated with the disruption of paracellular tight junctions (Kwong, Perry, 2013), which apparently is caused by calcium displacement from these junctions (Kwong et al., 2014). So, it would be expected that the increase in water hardness would minimize changes in osmoregulation on pacus exposed to acid waters, as demonstrated by the reduction in NKA and V-ATPase activities. Thus, decreased NKA and V-ATPase activities could also occur due to the minor disturbance in ion regulation and the positive effects of HWH. In this context, HWH could reduce diffusive ion losses in silver catfish (Townsend, Baldisserotto, 2001).
The alterations of NKA activity, along with the increasing level of ROS causes functional disorders in osmoregulation as well as tissue damage (Atli, Canli, 2007). In the current study, we found that, in pacus, LPO and antioxidant defences may be increased or inhibited under acidic or alkaline pH depending on water hardness and the duration of exposure, and gills and kidney showed variability in antioxidant competence. The kidney demonstrated greater antioxidant capacity compared to the gills, since GST activity was higher in this tissue, although ACAP was slightly higher in the gills (lower relative area). TBARS levels, in turn, were higher in the gills, indicating more LPO end products compared to the kidney. Peroxidative damage to gills may result from oxidative deterioration of polyunsaturated fatty acids, thereby impacting the osmoregulatory functions of the gills (Evans, 1987) and, consequently, increasing the production of ROS and thereby increasing lipid damage.
Enhanced ROS formation during an environmental stress situation causes the activation of redox-sensitive transcription factors (Nrf2, for example) and ROS-mediated covalent modification of antioxidant proteins leading to enhanced antioxidant defences (Hermes-Lima et al., 2015). In addition, previous evidence indicates that acidic or alkaline pH can trigger a pro-oxidant scenario in different fish organs (Mai et al., 2010;Maqsood, Benjakul, 2011). The higher LPO (TBARS) and antioxidant defences (GST and ACAP) in the gills and kidney of pacu juveniles exposed to acidic or alkaline pH, mainly with LWH, observed in our study was expected and demonstrates that the function of these organs was affected. Under these conditions, fish were not able to neutralise ROS in these tissues. The capacity to neutralise ROS was greater in fish subjected to LWH at circumneutral pH. Despite the negative effects of acid and alkaline waters, our results demonstrate that oxidative stress variables are early indicators of acidic or alkaline pH-induced toxic effects and suggest that HWH presented positive effects, reducing LPO and increasing GST activity and ACAP (lower relative area). In this way, Copatti et al. (2019) verified that HWH improved biochemical and physiological responses in pacu under acidic or alkaline pH.
GST activity is involved in detoxification, and elevated GST activity demonstrates that an attempt to neutralise elevated levels of ROS has occurred (Monteiro et al., 2009) and serves to re-establish the balance between pro-oxidants and antioxidants to alleviate ROS-induced oxidative damage (Wilhelm Filho et al., 2005). In the present study, HWH increased GST activity in kidney on day 1 and gills on day 5 in pacus subjected to acidic and/or alkaline pH. Interestingly, in circumneutral pH, an elevation of water hardness reduced GST activity in the gills on these same days (1 and 5), which suggests that, in this tissue, HWH had no positive effect when pacus were exposed to circumneutral pH. The ACAP assessment sheds light on the overall antioxidant status of the fish against peroxyl radicals (Maltez et al., 2018), which was compromised in the kidney of pacus through alkaline pH exposure on day 5. In addition, a possible explanation for lower gills ACAP (higher relative area) in alkaline pH with HWH on day 1 was due to a compensatory effect, because an increase in GST activity and a reduction in TBARS levels were observed under these conditions.
The TBARS assay quantifies the end products of LPO, which is a very sensitive marker of oxidative damage in the gills and kidney. TBARS are also thought to be involved in triggering the up-regulation of antioxidant enzymes (Lushchak, Bagnyukova, 2006). Acidic or alkaline pH with LWH cause oxidative stress in pacu, because the elevated TBARS concentration observed in the gills and kidney indicates high levels of LPO. Concomitantly, the results of antioxidant enzymes and ATPases in the gills and kidney can be also interpreted as an indication of a more oxidative environment in cells, leading to protein oxidation in acidic or alkaline waters with LWH. In addition, after the increase in TBARS in the gills on day 1 in fish exposed to alkaline pH with LWH, the gills showed lower TBARS levels on day 5. This result may indicate a decrease in polyunsaturated fatty acid levels due to early oxidative degradation (Sun et al., 2014). Finally, HWH reduced LPO in fish kidney subjected to acidic (day 5) or alkaline pH (day 15), according to the TBARS results.
Another important aspect that we observed in the present study was that alkaline pH presented lower antioxidant capacity and/or higher LPO than acidic pH. Similarly, in tambaqui (Colossoma macropomum), Nile tilapia (Oreochromis niloticus) and Amazon catfish (Pseudoplatystoma reticulatum x Leiarius marmoratus), alkaline pH present more deleterious effects than acidic pH (Aride et al., 2007;Lemos et al., 2018a,b). In addition, high pH levels (> 9.0) can harm fish by denaturing cellular membranes (Zahangir et al., 2015). Nevertheless, synergism between exposure to acidic water and ROS can cause DNA damage (Phillips, Ward, 2001) as a result of increased cations (Jolly et al., 2004). If cations are responsible for the changes in pH, they can react with biological molecules, and/or act as signalling molecules which will play a role in mediating extensive DNA damage (Leonard et al., 2004). Our results suggest that the occurrence of LPO in acidic pH with LWH might be initiated by ROS.
In this sense, HWH could decrease cell permeability and potentially ameliorate the effects of high pH on passive ion loss (Yesaki, Iwama, 1992;Wilkie, Wood, 1996). For instance, calcium ions appear to be important in mediating white shrimp responses to acidic or alkaline pH stress, decreasing oxidative stress (Wang et al., 2009). Saglam et al. (2014) showed that HWH has a protective role in Nile tilapia exposed to metals (Cd, Cu), and HWH was also found to trigger antioxidant enzyme responses. In the current study, we observed increased antioxidant defence capacity in pacus under acidic or alkaline pH with HWH, which shows that enzymatic defence mechanisms are used for ROS scavenging. In addition, to our knowledge, this is the first study that evaluated oxidative stress in fish under different water pH at LWH and HWH.
The oxidative stress produced in tissues (e.g., gills and kidney) are maintained within the physiological levels when the antioxidant defences are in a normal physiological state (Wilhelm Filho, 1996). Our results demonstrated that, even in acidic (mainly) or alkaline waters, oxidative stress in pacu gills and kidney is less intense after a longer exposure period (15 days), because, in acidic pH, fish were able to increase their antioxidant defences and, in alkaline pH (except to kidney TBARS at LWH), oxidative stress variables were stable independently of hardness on day 15. Maltez et al. (2018) suggests that, in some cases, a recovery period (10 days or more) is necessary to complete restoration of homeostasis after the end of the stressful conditions. Thus, we believed that pacu requires 15 days or more to acclimate to waters with acidic or alkaline pH and to restore the activity of osmoregulatory and oxidative stress enzymes in gills and kidney.
In conclusion, circumneutral pH is indicated to raise pacu juveniles because at this pH there is less oxidative stress and fewer variations in ATPase activity. In addition, an increase in NKA and V-ATPase activities in the gills and kidney in acidic or alkaline water with LWH could indicate important osmoregulatory changes for the adjustment of pacu juveniles in these water pH; HWH exposure reduced ATPase activity in both organs. In an integrated view with our observation that acidic or alkaline pH induced oxidative stress and antioxidant responses primarily under HWH, we can conclude that pacu exposed to acidic or alkaline pH are less affected by the deleterious effects with water hardness of 120 mg CaCO 3 L -1 .