Climate change affects the parasitism rate and impairs the regulation of genes related to oxidative stress and ionoregulation of Colossoma macropomum

Global climate change represents a critical threat to the environment since it influences organismic interactions, such as the host-parasite systems, mainly in ectotherms including fishes. Rising temperature and CO2 are predicted to affect this interaction other and critical physiological processes in fish. Herein, we investigated the effects of different periods of exposure to climate change scenarios and to two degrees of parasitism by monogeneans in the host-parasite interaction, as well as the antioxidant and ionoregulatory responses of tambaqui (Colossoma macropomum), an important species in South American fishing and aquaculture. We hypothesized that temperature and CO2 changes in combination with parasite infection would interfere with the host’s physiological processes that are related to oxidative stress and ionoregulation. We experimentally exposed C. macropomum to low and high levels of parasitism in the current and extreme climate scenarios (4.5 °C and 900 ppm CO2 above current levels) for periods of seven and thirty days and we use as analyzed factors; the exposure time, the climate scenario and parasitism level in a 2 × 2 × 2 factorial through a three-way ANOVA as being fish the experimental unit (n = 8). An analysis of gill enzymatic and gene expression profile was performed to assess physiological (SOD, GPx and Na+/K+-ATPase enzymes) and molecular (Nrf2, SOD1, HIF-1α and NKA α1a genes) responses. A clear difference in the parasitism levels of individuals exposed to the extreme climate scenario was observed with a rapid and aggressive increase that was higher after 7 days of exposure though showed a decrease after 30 days. The combination of exposure to the extreme climate change scenario and parasitism caused oxidative stress and osmoregulatory disturbance, which was observed through the analysis of gene expression (Nrf2, SOD1, HIF-1α and NKA α1a) and antioxidant and ionoregulatory enzymes (SOD, GPx and Na+/K+-ATPase) on the host, possibly linked to inflammatory processes caused by the high degree of parasitism. In the coming years, these conditions may result in losses of performance for this species, and as such will represent ecological damage and economical losses, and result in a possible vulnerability in relation to food security.

www.nature.com/scientificreports/ monogeneans live freely on the fish's skin and feed on mucus, epithelial cells, and gills resulting in damage to the host tissue such as deep wounds in the epidermis, basement membrane and into the dermis 7,8 . Diseases in highly commercialized species cause significant economic losses every year, especially in the fish farming sector where high densities of stressed animals and poor water quality provides ideal conditions for outbreaks of disease, including those caused by monogeneans 5,9 . For this reason, in recent years, the topics of climate change and infectious diseases in aquatic animals have gained special attention, but the evidence regarding the connection is still scarce [10][11][12] . Some studies have shown that in fish climate change aggravates infectious diseases caused by bacteria 13,14 and parasites [15][16][17] . However, information is lacking when it comes to Neotropical fish, especially those from the Amazon.
The oxidative stress appears to play a major role in the pathogenesis and progression of many infectious diseases 18 . In our preliminary study 16 , we observed that the degree of parasitism by Monogenea in tambaqui (Colossoma macropomum), a highly commercialized species in the Amazon region, is differentially affected by the climate change scenario; fish kept in the extreme climate scenario show a significant increase in the level of parasitism, which is accompanied by an inflammatory response, as well as by overexpression of pro-inflammatory cytokines such as IL-1β and negative regulation of anti-inflammatory cytokine such as IL- 10. Taken together, the data suggest that the activation of inflammatory mechanisms seems to be an adaptive response of the species to try to deal with the increase in parasitism imposed by climate change, which evidences a key role of the genes that regulate inflammation.
Inflammation and oxidative stress are inextricably linked. Oxidative stress plays a key role during inflammatory reactions and is considered an important pathophysiological process promoted by reactive oxygen species (ROS) production 19,20 . Enzymes and genes related to the antioxidant system play a fundamental role as the main defensive line in the response to oxidative stress by eliminating the ROS and in the maintenance of homeostasis 21,22 . The antioxidant capacity is partly regulated by a wide variety of transcription factors, including NF-E2-related factor 2 (Nrf2), which modulates the transcription of type II detoxifying enzyme genes, such as SOD1, which encodes Cu/Zn superoxide dismutase, and is a crucial combatant to augmented oxidative stress and prevents the free radical exchange reaction of relatively toxic superoxide by accelerating its oxidation to non-toxic hydrogen peroxide 23,24 .
ROS and antioxidant mechanisms also play an important role in regulating HIF-1α (hypoxia inducible factor) 20 . HIF is a key regulator that plays a central role in O 2 homeostasis, and is able to transcriptionally control the expression of more than 1000 genes that are involved by binding to DNA sequences that contain hypoxic response elements (HRES) 25 . HIF-1α can be activated/stabilized in response to increased ROS levels and is also induced by the parasite damage to the gills 26 . It is particularly important when looking at the relationship between HIF-1α and genes related to inflammatory mechanisms. Cytokines can activate HIF-1α expression in a ROS-dependent manner 27 . For fish, the role of HIF-1α on inflammation is still unclear and may vary between different species, both promoting and inhibiting inflammation 28,29 .
Recent studies have reported not only the negative effects of parasitism on the antioxidant system, but also its negative effects on fish ionoregulation [30][31][32] . These results deserve to be highlighted, mainly when it comes to infection by monogeneans, since they promote severe damage to the gills of the hosts, which is the main organ related to ionoregulatory processes in fish 7,8 . Gills are a physical barrier that largely depend on the structural integrity of epithelial cells and intercellular tight junctions. The structural integrity of epithelial cells is related to antioxidant ability, gas exchange, ion regulation, and acid-base balance, and the damage caused by the parasite can compromise all of these functions 33,34 .
Na + /K + -ATPase (NKA) is essential to ensure efficient ion transport in osmoregulatory epithelia in order to maintain ionic homeostasis. In the osmoregulatory tissues of aquatic organisms, the electrochemical gradient generated by the NKA provides a driving energy for many transport systems via channels and transporters that are localized in the apical and basolateral membranes 33,35 . The NKA α-1a subunit is expressed in lamellar mitochondrion-rich cells in the gills and has been reported as the gill-specific gene involved in ion absorption in freshwater fish 36 . Some mammalian studies report that inhibition of NKA α-1a production increases oxidative stress and intracellular ROS formation 37 . However, there is no data regarding the role of NKA in the face of exposure to climate change scenarios and parasites that cause severe branchial tissue damage, one of the most responsive organs to ionoregulation. Monogenea may interfere with all of these crucial molecular/physiological functions in the host, as mentioned above.
Tambaqui (Colossoma macropomum) is the main native species farmed in South American continental waters 38 . Similar to several fish species cultivated worldwide, tambaqui are highly affected by monogenean infections and thus constitutes a good biological model for evaluating environment-parasite-host interaction. We previously demonstrated that climate change will favor the proliferation of monogeneans, especially in lessparasitized tambaqui, and promotes the activation of inflammatory mechanisms 16 . These results motivated us to perform this second study.
Thus, herein, we investigated whether different periods of exposure to climate change scenarios (7 and 30 days) and two degrees of parasitism (low and high) would affect the environment-parasite-host interaction and the tambaqui antioxidant and ionoregulatory responses. We hypothesized that (i) an increased rate of parasitism caused by the extreme climate scenario would be maintained during the 30 days of exposure and (ii) the inflammation caused by the high parasitism would affect the expression of genes and enzymes related to antioxidant and ionoregulatory responses.

Results
Parasitological analysis. Two degrees of parasitism were considered after the parasitological analysis: low, 1-32 monogeneans per fish, and high, over 32 monogeneans per fish. The species identified in this study were Mymarothecium boegeri 39 and Notozothecium janauachensis 40 .
The parasitological analysis revealed that the mean intensity of parasitism was significantly increased in both groups exposed to the extreme scenario in 7 and 30 days (p < 0.001; F = 175.804) with interaction between the scenarios and parasitism (p < 0.001; F = 28.993). However, after 30 days there was a decrease, which can be seen in Fig. 1. The low-parasitized animals subjected to the extreme scenario presented a greater increase in the mean intensity of parasitism as was observed in our previous study 16 and this difference was maintained in 30 days. Antioxidant enzymes. An increase of SOD enzyme in activity was observed only in fish in the high parasitism group (HG) in the current scenario when compared with the low parasitism group (LG) after 7 days (p = 0.020; F = 5.863). After 30 days, the activity remained constant and showed an increase in the LG (p = 0.006; F = 8.536) (Fig. 2a). There was no interaction of factors (p = 0.719; F = 0.131). For GPx (Fig. 2b), after 7 days, there was a decrease in activity in the HG in the extreme scenario compared to the current scenario (p = 0.030; F = 5.065) and the LG in the extreme scenario (p = 0.048; F = 3.258). The same was observed for the HG after 30 days in the extreme scenario when compared with the LG (p = 0.028, F = 5.849). There was no interaction of factors (p = 0.577; F = 0.316). The NKA activity was lower in the LG after 7 days (p = 0.034; F = 3.095), however it showed an increase in the group after 30 days (p = 0.005; F = 4.159) and interaction was observed between climate scenarios and parasitism (p = 0.014; F = 6.536) (Fig. 2c).
Gene expression analysis. There was no difference in expression in any of the conditions tested for Nrf2 gene (Fig. 3a). For SOD1 after 7 days, the expression increased in fish from the LG maintained in the extreme scenario (p = 0.014; F = 6.590) when compared with the same group maintained in the current one (Fig. 3b). After 30 days, the increase in the LG remained constant in fish subjected to the extreme scenario (p = 0.053; F = 3.984). The greatest influence was caused by the degree of parasitism, followed by the scenario. There was no interaction of factors (p = 0.147; F = 2.186). HIF-1α showed an increase in expression also in the LG of the extreme scenario after 7 days (p = 0.034; F = 4.934), and a downregulation in the same group after 30 days (p = 0.054; F = 4.023; no interaction of factors: p = 0.193; F = 1.772), but the difference was maintained between treatments, displaying the same expression pattern as SOD1 (Fig. 3c). A decrease in the expression of NKA α1a in the LG after 7 days of exposure to the extreme scenario ( Fig. 3d) was observed (p < 0.001; F = 19.057) but, in contrast, in the same group, after 30 days, there was an upregulation (p < 0.001; F = 21.036). There was also downregulation of the HG group after 30 days of exposure to the extreme scenario concerning the same group in the current scenario and interaction was observed among exposure period and scenario (p = 0.029; F = 5.253) and between the scenario and parasitism (p = 0.018; F = 6.237).
The principal components analysis (Fig. 4) demonstrated that gene expression patterns are different between climate scenarios and also between different parasitism intensities after 7 days. The extreme climate scenario, parasitism, and gene expression (SOD1 and HIF-1α) are further related by PC1 (40.4%) after 7 days. After 30 days, we observed that there is a change in the observed patterns. While climate scenario did not show much difference, it is still possible to observe that parasitism and SOD correspond to the extreme scenario and are related by PC1 (24.7%) and NKA activity to PC2 (18.2%). The cumulative explained variance of PC1 and PC2 was 57.1% after 7 days and 42.9% after 30 days. Therefore, these findings corroborate the results from the ANOVA data. www.nature.com/scientificreports/

Discussion
Several environmental stressors, mainly those caused by climatic factors, such as temperature and CO 2 , affect various molecular and physiological processes in fish [41][42][43] . Our findings are in agreement with our preliminary results 16 , which indicate that the exposure of tambaqui to the extreme climate scenario for 7 days induced a www.nature.com/scientificreports/ significant increase in the parasitic load of both groups (high and low parasitism level) with emphasis on the low group, and the same was observed in the present study. However, after 30 days, we observed a decrease in the parasitic load in both groups, but this was even greater when compared to the current scenario (see Fig. 5). This can potentially be explained by chronic thermal stress on the parasite, which can be a predominant factor that harms its survival by causing damage to its physiological functions, as observed in Poecilia reticulata 44 , when infected by the monogenean Gyrodactylus turnbulli, there was a decrease in parasitic load when exposed to a temperature of 32 ºC, which for this fish is considered extreme. Brazenor and Hutson 45 also observed that in warmer waters (31 ºC) Neobenedenia girellae parasitizing Lates calcarifer had its life cycle reduced and thus, led to fewer stages of transmission. Thus, after 30 days of exposure to the extreme climatic scenario, the decrease in the parasitic load observed here may be the result of direct effects of climatic factors on the parasite and not an adaptive response of the fish. We also observed that the tambaqui were susceptible to oxidative damage when exposed to the extreme climate scenario plus parasitism. Fish from the LG exhibit a marked increase in SOD1 mRNA levels, especially when submitted to the extreme scenario for 7 days, though no effect was observed on Nrf2 expression (Fig. 3a,  b). Our findings suggest that parasitism and the climate change scenario induce oxidative damage that influences the increase in SOD1 mRNA levels. Sun et al. 21 demonstrated that thermal stress induced both oxidative damage and positive regulation of antioxidant genes in Micropterus salmoides, including SOD1. Oxidative stress resulting from exposure of fish to parasites has also been reported 30,46,47 , but the effects of parasitism on the expression of genes related to antioxidant processes are not clear.
The literature reports increased expression of antioxidant genes through the Nrf2-ARE pathway 48 ; however, in our study SOD1 expression increases, but it is not accompanied by differential transcriptional activation of , HIF-1α (c) and NKA α1a (d) in the gills of Colossoma macropomum exposed to the extreme climate scenario as foreseen by IPCC (2014), compared to current scenario. Boxes with different letters indicate differences between scenarios and asterisks indicate differences between exposure periods (p < 0.05). Low (LG) and High (HG) indicate the degree of parasitism after seven and thirty days of exposure to the experimental conditions. A black dot in the box indicates the average of each group.   www.nature.com/scientificreports/ Nrf2. Other studies have also failed to observe a clear relationship between the levels of Nrf2 and SOD1 mRNA in fish 18,49 . Mukaigasa et al. 49 suggest that SOD may not be a direct target for Nrf2 since, when evaluating the expression of the gene in zebrafish with a mutation in Nrf2, they observed an expression similar to that described in wild fish. Another possible explanation for the maintenance of transcriptional levels of Nrf2, similar to that which occurs in non-stressed groups, is the fact that this gene not only executes via oxidative stress prevention, but also has an anti-inflammatory function (reviwed by Ma 50 ). In mammals, Nrf2 has been shown to inhibit inflammation by blocking NF-κB and consequent inhibition of pro-inflammatory cytokines (reviewed by Hennig et al. 51 ; Ma 50 ). In our previous study, we observed that the exposure of parasitized tambaqui to climate change scenarios for 7 days resulted in increased expression of IL-1β and HSP70, pro-inflammatory genes, as well as decreased expression of IL-10, anti-inflammatory cytokine, suggesting the activation of inflammation 16 . As inflammatory processes are associated with the production of reactive oxygen species (ROS) 51 , it is expected that positive differential activation of Nrf2 triggers an antioxidant response, and limits the negative effects of ROS, which would also limit the formation of the inflammasome by blocking the expression of pro-inflammatory genes. We propose that the maintenance of similar transcriptional levels of Nrf2 between the groups is an attempt to favor activation of the inflammatory process in detriment of the antioxidant response.
This suggestion is supported when we evaluate the activity of antioxidant enzymes. The SOD activity increases after 7 days in the HG in both scenarios, the same being observed for 30 days. However, there was an increase in the LG after 30 days (Fig. 2a) in the extreme scenario, while the GPx activity shows a decrease in highly parasitized fish after 7 and 30 days under the extreme scenario, and remains similar in the other groups (Fig. 2b). The influence of temperature and parasitism on antioxidant responses and oxidative status can vary depending on the tissue, period of exposure, level of parasitism and temperature increase 30,52,53 . In the species Seriola lalandi, infection by Neobenedenia melleni did not promote an increase in SOD activity 53 . Similar results were observed in the gills of Siganus oramin parasitized by Cryptocaryon irritans 54 . However, in the species Pangasianodon hypophthalamus, when parasitized by Thaparocleidus sp., there was an induction of oxidative stress biomarkers, including SOD activity 30 , which is similar to that observed in our study. Klein et al. 52 reported the absence of changes in SOD activity and a reduction in branchial GPx activity in Notothenia coriiceps that were exposed to temperature increases, and the authors also suggest that fish are unable to positively regulate the activity of antioxidant enzymes in gills under thermal stress conditions.
In the species Pomatoschistus micros, the combination of parasitism and an increase in temperature also did not affect SOD activity 46 . Knowing that both parasitism and increased temperature result in oxidative stress 21,46,47 , the results observed in our study indicate an attempt by the fish to eliminate ROS, which is evidenced by the increase in the SOD activity since it is the first line of enzymatic defense against oxidative stress, and whose activity results in the dismutation of the superoxide anion in H 2 O 2 and H 2 O. This was not observed for GPx, which is the peroxidase involved in the degradation of H 2 O 2 and other peroxides 55 . The changes observed in the activities of antioxidant enzymes are an indication that parasitism and climate change promoted an increase in the production of ROS and oxidative stress in tambaqui, which, in turn, can guarantee greater efficiency in the inflammatory process triggered by the combination of parasitism and climate change, as evidenced by Costa and Val 16 .
In addition, the data also shows a clear lack of integration between levels of transcripts and enzymatic catalytic activity, mainly after 30 days of exposure. This inverse pattern between mRNA levels and enzyme activity for SOD was observed in fish exposed to different environmental stressors [56][57][58] . Liu et al. 56 demonstrated that for zebrafish, the increase in transcripts of antioxidant genes was accompanied by inhibition in the activity of the respective enzyme antioxidants. Similar results have been reported for the species Ctenopharyngodon idella 58 . Regoli et al. 57 suggested that the mRNA levels of antioxidant enzymes may represent a specific cellular response at a given time, but not the outcome of the response to a stressful condition. Contradicting responses in mRNA levels and antioxidant enzyme activity may be related to the transcription's vulnerability to variation in cellular conditions imposed by metabolite status, exposure period, type of stress and even variations inherent to the species, especially because the transcription is transitory, while enzymatic catalysis is more durable and subject to additional regulatory processes 56 . Therefore, the inconsistency between enzyme activity and levels of transcripts shows the complexity of responses to oxidative stress, and these responses may not be associated with transcriptional variations, but with post-translational changes 57 .
The upregulation of HIF-1α mRNA in the LG of the extreme scenario after 7 and 30 days was possibly due to the presence of factors such as ROS and inflammatory cytokines, and this can be supported by the result we obtained with the overexpression of SOD1 in the same groups (Figs. 3b, c, 4a). Increases in HIF-1α transcripts occurred only in the LG whose level of parasitism increased in the extreme scenario, which is greater than that observed in the HG in the same scenario (Fig. 1a), for which the levels of HIF-1α mRNA are also lower (Fig. 3c). The transcriptional behavior of HIF-1α observed here is the same as that observed for the pro-inflammatory genes (HSP70 and IL-1β), and contrasts with the expression of anti-inflammatory genes (il-10) observed in our previous study after 7 days 16 . This suggests that the positive regulation of HIF-1α in the LG exposed to extreme scenario may be the result of the release of pro-inflammatory cytokines due to the parasitic load that increased considerably in these fish, contrary to what was initially observed in the HG. According to Zepeda et al. 27 , even if there is no hypoxia, HIF-1α activation can occur through cytokines in ROS-dependent mechanisms. A pattern similar to that observed after 7 days occurred after 30 days also in the extreme scenario with the expression of HIF-1α, suggesting that the inflammatory response can continue to be active in chronic conditions. In zebrafish, Ogryzko et al. 29 suggest that the stabilization of HIF-1α promotes an increase in IL-1β transcription and contributes to host protection, since it is part of an immediate pro-inflammatory response. However, as described by Guan et al. 28 for Boleophthalmus pectinirostris, activation of HIF-1α and pro and anti-inflammatory cytokines may vary between different species and, in this species, increased HIF-1α expression suppressed IL-1β expression and  59 showed that Sparicotyle chrysophrii (Monogenea) parasitizing Sparus aurata induced hypoxia in the host, which caused the expression of oxidative stress genes, and is associated with upregulation of HIF-1α. Haliotis fulgens, on the other hand, exposed to hypercapnia, hypoxia and a temperature of 32 °C displayed an upregulation of HIF-1α as well as antioxidant genes such as SOD 60 . Our results are in agreement with Wang et al. 61 who observed a relationship between the regulation of HIF-1α and defense against infection by the bacterial pathogen Streptococcus agalactiae in tilapia in the perspective of global warming. Unlike HIF-1α and SOD1, NKA α1a showed downregulation in the LG when exposed to the extreme scenario for 7 days; however, after 30 days, the expression returns to normal levels when compared to the current scenario, with the exception of the HG group that remained downregulated (Fig. 3d). As an ion transport pump, NKA is not only crucial for maintenance of the ionic homeostasis, but also plays a critical role in cellular function and signaling 62 . Thus, the downregulation observed here could be a consequence of the damage that the parasite causes to the branchial tissue and this may have affected the normal functioning of mitochondrion-rich cells that then lost their ability to induce NKA α1a expression as a result of disruption and failure in molecular machinery. We also speculate that all this downregulation could also be interpreted as a shutdown of the metabolism, as part of a hypometabolic response to concentrate the energy on the immune response. For human cells, the downregulation of NKA α1a is associated with increased oxidative stress 37 . Consequently, as already mentioned above, we observed an increase in SOD1 in the extreme scenario, which is indicative of oxidative stress. Few studies have analyzed the expression of the NKA α1a gene in fish and the majority have studied it only in relation to the effect of exposure to salinity. However, the results of Tomalty et al. 64 differ from our results, since they observed that this gene was upregulated in Oncorhynchus tshawytscha under heat stress of 25 °C. NKA α1a mRNA was also differentially expressed in blue rockfish (Sebastes mystinus) exposed to high pCO 2 and hypoxia, which suggests that the fish effectively use compensatory mechanisms to contend with climate change stressors 65 . This is the first study to analyze the expression of the NKA α1a gene in parasitized fish and exposed to synergistic effect of the forecasted 2100 climate scenario (increased temperature and CO 2 ) and parasitism.
As well as the results of gene expression, we found a decreased activity of Na + /K + -ATPase in the branchial tissue in the LG after 7 days also in the extreme scenario, but in the same group after 30 days this had increased. This reduction in activity of Na + /K + -ATPase after 7 days can be explained in the same way as for gene expression. Severe tissue damage caused by the high parasitism rate driven by the climate change scenario led to an inflammatory stage that impaired the functioning of mitochondrion-rich cells. Kumar et al. 66 reported that the Pangasius catfish infected by Monogenea reduced NKA activity and they associated this reduction with cellular stress and degradation by toxins released by the parasite itself, which impairs the synthesis. Sparus aurata exposed to an outbreak of Amyloodiniosis also showed reduced NKA activity, due to anoxia induced by parasitism that led to ionoregulatory failure 67 . The increase that we observed after 30 days in the LG in the extreme scenario may be a response to the reduced parasitism rate, which lead to a return to normal activity levels.
Evidently, acclimation of the fish to a scenario of progressive changes in temperature and CO 2 concentrations that simulates a real-life scenario may result in adjustments of responses, providing effective acclimatization to the transitioning environment, and if these predictions are confirmed, our results support the occurrence of the effects observed in animals.
In summary, this study reports the effects caused by the forecasted 2100 climate scenario (increased temperature and CO 2 ) after seven and thirty days of exposure associated with two levels of gill parasitism by monogeneans in tambaqui. The present study suggests that climate changes cause a rapid increase in parasitism during seven days, which decreases after thirty days, but remains higher than that of the control group. We also show clear evidence of regulation of genes associated with oxidative stress and inflammation, physiological stress, and ionoregulatory problems associated with a rapid and aggressive increase in parasitic infection caused by the climate change scenario. However, gene expression should not be confused with the proteins that execute the processes.  14.19 ± 1.15 cm) were obtained from a local fish farm (Fazenda Santo Antônio: 02° 44′ 802″ S; 059° 28′ 836″ W, Amazonas, Brazil) and transferred to the Laboratory of Ecophysiology and Molecular Evolution (LEEM) at INPA. Fish were acclimated for at least 3 weeks before the experiment in 310 L tanks with continuous aeration and water flow, the temperature of 26.2 ± 0.8, dissolved oxygen 6.5 ± 0.5 pH 6 ± 0.5 in order to recover from transport stress. Fish were fed once a day with a commercial diet (2-4 mm pellets, 36% of crude protein) during this period.

Material and methods
Experimental design. These fish were naturally parasitized and after the acclimation period were treated with 3 g L −1 of salt for 15 min during three consecutive days according to Schelkle et al. 69 to decrease the animals' parasitic burden and thus establish two degrees of parasitism according to the protocol used by Costa  www.nature.com/scientificreports/ After defining the two degrees of parasitism (low and high group, which we abbreviate to LG and HG, respectively), the animals were transferred to two real-time controlled environmental rooms as described by Costa and Val 16 . Each room simulated the current (current temperature and CO 2 levels) and extreme (RCP8.5) scenarios according to the Fifth IPCC Assessment Report for the year 2100 2 . The current conditions simulate the same conditions occurring in a forested area of the Amazon without human influence, with data acquisition by Fieldlogger 512k (Novus Produtos eletrônicos LTDA) every two minutes. The extreme climate room was set to 4.5 °C and 900 pmm CO 2 above the current conditions (Fig. 5). The artificial light-dark cycle was 12:12, and humidity was set as a derived condition. Prior to the experiment, the gills, skin, and fins were carefully scraped with a coverslip and observed under an optical microscope, and underlying subclinical infection and/or the presence of other parasites were assessed.
The juveniles were distributed in eight 60 L PVC tanks in four replicates per treatment (low and high levels of parasitism) containing four individuals each tank, in both scenarios. The fish were exposed to each climate room for seven and thirty days in October 2018 (Amazon dry season). After each exposure period, two fish were removed from each tank, with a total of eight fish per scenario and treatment being collected (n = 8). The volume of water was adjusted after collecting the fish. The illustration in supplementary information (Fig. S1) demonstrates the experimental setup performed in this study. Ammonia accumulation was avoided by partial water renewal throughout the experiment. The pH, O 2 and CO 2 levels and temperature of the water were measured daily ( Table 1). All animals were fed once a day with a commercial diet with 36% crude protein during the experiment. Subsequently, the fish were anaesthetized, weighed, measured and euthanized by rapidly severing their spinal cord with a scalpel for tissue sampling. Gill samples were collected using sterile tweezers and scissors, and one side of each fish was immediately immersed in liquid N 2 and stored in an ultra-freezer at − 80 °C for biochemical and genetic analysis.
Parasitological analysis. The branchial arches from one side of each fish were sampled, separated and placed individually in plastic vials of 60 mL containing formalin (1:4000). The parasites were scraped from the gills and the contents of the plastic vials were counted under a stereomicroscope in a Petri dish. After counting, the parasites were stored in microtubes in 70% alcohol for later identification according to Rawson and Rogers 70 . To estimate the total amount of parasites per fish, the result was doubled. The prevalence and mean intensity of parasites were calculated according Bush et al. 71 .
Antioxidant enzymes. The glutathione peroxidase (GPx) activity was determined based on the oxidation of NADPH in the presence of GSH (0.95 mM) and H 2 O 2 at 340 nm, according to the method described by Hopkins and Tudhope 72 and activity was calculated as nmol of oxidized NADPH min −1 mg protein −1 using the molar extinction coefficient of 6.22 mM −1 cm −1 .
Superoxide dismutase (SOD) activity was quantified based on the inhibition of the cytochrome c reduction rate by the superoxide radical at 550 nm and 25 °C, according to the method described by Turrens 73 . Enzyme activity is expressed as U SOD mg protein −1 , where 1 U of SOD corresponds to the quantity of enzyme that promoted the inhibition of 50% of cytochrome c.
Gill Na + /K + -ATPase activity. The activity of this enzyme was determined by NADH oxidation in an enzymatic reaction coupled to the hydrolysis of ATP 74 . The assay is based on the inhibition of NKA activity by ouabain (2 mM). Gills were homogenized (1:10 w/v) in buffer (pH 7.5) containing (in mM): sucrose 150, imidazole 50, EDTA 10 and deoxycholic acid 2.5, and centrifuged at 2,000 × g for 7 min at 4 °C. Supernatants were added to a reaction mixture containing (in mM): imidazole 30, NaCl 45, KCl 15, MgCl2 3.0, KCN 0.4, ATP 1.0, NADH 0.2, fructose-1,6-bisphosphate 0.1, PEP 2.0, with 3 U mL −1 pyruvate kinase and 2 U mL −1 lactate dehydrogenase. Samples were run with and without ouabain. Absorbance was read over 10 min at 340 nm in a spectrophotometer (Spectra max Plus, model 384, Molecular Devices®, USA). Na + /K + -ATPase activity was calculated by the differences between total activity and activities with the ouabain inhibitor. Gene expression analysis. Total RNA was extracted with the TRIzol reagent (Life Technologies, CA, USA), following the manufacturer's instructions, and the obtained samples were treated with an amplification grade DNase I kit (Invitrogen, CA, USA). The High-Capacity cDNA Reverse Transcription kit (Thermo Scientific, Waltham, MA, USA) was used for the complementary DNA synthesis. Quantitative PCR analysis was carried out on the Viia™ 7 Dx PCR-System (Applied Biosystem, CA, USA). The qPCR conditions were 2 min at 50 °C and 95 °C for 10 min; followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min (annealing temperature of all primers). The mRNA gene expressions were normalized to the β-actin and β-tubulin using the comparative 2 −∆∆Ct method 76 . The primers used are listed in Table 2. In all cases, each qPCR run was performed in triplicate and repeated with at least two independent samples. Data analysis. The data are presented as mean ± standard error of mean (SEM; n = 8). Prior to the comparative statistical tests, all the data were examined for normality (Shapiro-Wilk test) and homogeneity of variance (Levene test). Mean differences were evaluated by three-way ANOVA with exposure period, climate scenarios and parasitism as factors, and were discriminated using the Tukey post-hoc test and were considered significant at p < 0.05. To better fit the parametric assumptions of ANOVA, data were log-transformed. Data were analyzed by using the Agricolae package in R Software 77 and graphs were built using ggplot2 package 78 .