DNA methyltransferases and stress-related genes expression in zebrafish larvae after exposure to heat and copper during reprogramming of DNA methylation

DNA methylation, a well-studied epigenetic mark, is important for gene regulation in adulthood and for development. Using genetic and epigenetic approaches, the present study aimed at evaluating the effects of heat stress and copper exposure during zebrafish early embryogenesis when patterns of DNA methylation are being established, a process called reprogramming. Embryos were exposed to 325 μg Cu/L from fertilization (<1 h post fertilization - hpf) to 4 hpf at either 26.5 °C or 34 °C, followed by incubation in clean water at 26.5 °C till 96 hpf. Significant increased mortality rates and delayed hatching were observed following exposure to combined high temperature and Cu. Secondly, both stressors, alone or in combination, significantly upregulated the expression of de novo DNA methyltransferase genes (dnmt3) along with no differences in global cytosine methylation level. Finally, Cu exposure significantly increased the expression of metallothionein (mt2) and heat shock protein (hsp70), the latter being also increased following exposure to high temperature. These results highlighted the sensitivity of early embryogenesis and more precisely of the reprogramming period to environmental challenges, in a realistic situation of combined stressors.

differences in genome DNA methylation status have been found in zebrafish early embryos and germ cells. The oocyte DNA is globally hypomethylated compared to the one of sperm. After fertilization, zebrafish exhibit moderate bulk DNA demethylation, with de novo methylation starting at the 32-cell stage at 1 3/4 h post-fertilization (hpf). The methylation level is then restored by sphere stage at 4 hpf to levels observed in sperm [9][10][11][12] . It is also noteworthy that in zebrafish the paternal DNA methylation pattern is maintained throughout early embryogenesis while the relatively hypomethylated maternal DNA is reprogrammed to a pattern similar to that of the sperm (i.e., zebrafish inherit the DNA methylome from sperm) 11,12 . Currently, alterations in DNA methylation due to environmental exposures during reprogramming have received little attention. Martin et al. reported severe phenotypical changes and lower global DNA methylation levels in zebrafish exposed to 5-azacytidine, a demethylating agent, from 1 to 24 hpf 13 . Shorter windows of exposure from 1 to 4 hpf appeared to be critical and resulted in similar results, whereas exposure from 6 to 24 hpf showed almost no effects on morphology. Interestingly, this sensitive early stage window corresponds with the dynamic demethylation and remethylation events that occur during development and indicates that this time window must be included in toxicity testing 7 . More recently, McGee et al. exposed zebrafish embryos to different flame retardant chemicals during early development. They reported a critical window of exposure from 0.75 hpf to 2 hpf, which induced mortality and malformations at 96 hpf and delayed remethylation of the zygotic genome 14 .
Accordingly, the present study aimed at evaluating the effects of two widespread environmental stress factors that frequently co-occur in aquatic systems, i.e., temperature and copper exposure, during the DNA methylation reprogramming, a presumed sensitive period of zebrafish development. Copper (Cu) is an essential micronutrient for all organisms, but, in excess, it can cause detrimental effects on fish from the cellular to population level 15 . In zebrafish, Johnson et al. reported that exposure of embryos to a range of copper concentrations from 50 to 1000 μ g/L resulted in mortality, hatching inhibition, impairment of larval development and lateral line dysfunction 16 . It is also admitted that metal toxicity generally increases with rising temperature in aquatic species 17,18 . For instance, rising temperature was found to increase nickel toxicity in zebrafish embryos 19 , copper susceptibility in fathead minnow Pimephales promelas 20 and cadmium toxicity in the European bullhead Cottus gobio 21 . Besides, although epigenetic changes following exposure to multiple stressors remain to date unknown, several studies suggested that thermal stress [22][23][24] 27 .
In order to test whether the reprogramming event is sensitive to the effects of a combination of environmental stressors, zebrafish embryos were exposed to Cu from fertilization (before complete hardening of the protective chorion) to 4 hpf (sphere stage) at either 26.5 °C or 34 °C, followed by incubation in clean water at 26.5 °C to 96 hpf. We then quantified global DNA methylation and the expression of the 6 dnmt3 genes in 96 hpf larvae. The analysis was completed by measuring the expression of genes encoding for well-known stress-induced proteins, namely metallothionein (mt2) and heat shock proteins (hsp70 and hsp90). Using genetic and epigenetic approaches, this research aimed to refine the current knowledge on the sensitivity of early embryogenesis and more precisely of the reprogramming period to environmental challenges.

Results
Mortality rate and hatching efficiency. Figure 1 shows the proportion of eggs and embryos that died during the experiment (black bars), the proportion of embryos that stayed alive (dark grey bars), that hatched (i.e., larvae, light grey bars) and finally proportion of larvae that died (spotted dark grey bars). Mortality typically occurred within the first 24 hpf. After that period, the survival rate was stable until 96 hpf. Temperature, Cu exposure and their interaction significantly (P < 0.001) influenced mortality rate at 4 hpf and 24 hpf (Table 1). At 4 hpf, no mortality was recorded in groups held at control temperature, whereas we observed an increased mortality rate in groups held at 34 °C (4.4 ± 2.3% in unexposed group and 24.2 ± 7.0% in Cu-exposed group). At 24 hpf, mortality rate was increased following exposure to combined stressors with values reaching 48.0 ± 10.4%.
A peak of hatching was observed at 72 hpf in the control group (average hatching rate of 91.1 ± 3.0%), although some embryos started to hatch at 48 hpf. No interaction effect on the hatching success was observed. However, hatching rate was significantly affected at 72 hpf and 96 hpf by temperature (P < 0.001 at 72 hpf and P < 0.01 at 96 hpf) and Cu exposure (P < 0.001) ( Table 1). Both factors significantly reduced hatching rate in an independent way. Overall, delayed embryo hatching was observed following exposure to high temperature and Cu from < 1 to 4 hpf. In fact, at 72 hpf, average hatching rates of 91.1 ± 3.0% were observed in control group compared to 41.3 ± 11.2% in Cu-exposed group at 34 °C; at 96 hpf, average hatching rates of 99.1 ± 1.2% were observed in control group compared to 89.5 ± 4.6% in Cu-exposed group at 34 °C. DNA methylation. The global DNA methylation levels in 96 hpf larvae following exposure to 325 μ g Cu/L at either 26.5 °C or 34 °C from < 1 to 4 hpf is shown in Fig. 2. The average percentage of methylated cytosines ranged between 1.12 ± 0.16% and 1.35 ± 0.21% over all treatments. There was no significant difference in global cytosine methylation levels between the four groups.
The expression profiles of the 6 dnmt3 genes in 96 hpf larvae following exposure to 325 μ g Cu/L at either 26.5 °C or 34 °C from < 1 to 4 hpf are depicted in Fig. 3. In our study, we followed the new nomenclature based on molecular evolution and phylogenetic analysis of dnmt3 paralogues proposed by Campos et al. 28 . Temperature and Cu exposure, alone or in interaction, significantly affected the mRNA expression of dnmt3 genes (Table 2). First, the mRNA expression of all dnmt3 was significantly increased by exposure to high temperature, regardless of Cu exposure. The increase ranged from 22% for dnmt3a2 and dnmt3b4 to 68% for dnmt3b1. Second, the mRNA expression of dnmt3b1, dnmt3b2 and dnmt3b3 was significantly influenced by exposure to Cu, independently of heat stress. Exposure to 325 μ g Cu/L significantly increased the mRNA expression of dnmt3b1, dnmt3b2 and dnmt3b3 by 33%, 20% and 29%, respectively. Finally, we observed an interaction effect of temperature and Cu exposure for the expression of dnmt3a1 and dnmt3b4. Overall, their expressions were increased following  . Different letters (a-c) mean significant (P < 0.001) interaction effect between treatments within a given time period. exposure to both stressors. For instance, we observed an increased dnmt3b4 expression in groups held at 34 °C (3.71 ± 0.40 in unexposed group compared to 4.56 ± 0.41 in Cu-exposed group), whereas no significant changes were recorded in groups held at 26.5 °C.
Transcriptional expression of stress-related genes. The expression profiles of mt2, hsp70 and hsp90 genes in 96 hpf larvae following exposure to 325 μ g Cu/L at either 26.5 °C or 34 °C from < 1 to 4 hpf are depicted in Fig. 4. We did not observe any interaction between temperature and Cu exposure for the expression of studied stress genes. However, the mRNA expression of mt2 and hsp70 was significantly affected by Cu exposure (P < 0.01 and P < 0.05, respectively), regardless of heat stress. Exposure to 325 μ g Cu/L significantly increased the mRNA expression of mt2 and hsp70 by 44% and 37%, respectively. Further, hsp70 mRNA expression was significantly influenced by temperature (P < 0.01), independently of Cu exposure. Exposure to 34 °C significantly increased hsp70 mRNA expression by 59%. The mRNA expression of hsp90 was not significantly affected by temperature or Cu exposure and ranged between 1.06 ± 0.38 and 1.42 ± 0.33 over all treatments.

Discussion
Our study was designed to investigate whether the DNA methylation reprogramming that occurs during zebrafish early development is sensitive to environmental stress factors. According to Kamstra et al. 7 , this is an extremely important time window to include in studies dealing with the effects of environmental challenges. Epigenetic modifications early in life are proposed to underlie the development of an adverse adult phenotype. Overall, both heat stress and copper exposure applied during reprogramming were particularly stressful and resulted in adverse developmental consequences on zebrafish, as indicated by high mortality, delayed hatching and increased expression of several stress-related genes. Further, we observed changes in the expression of de novo DNA methyltransferases (dnmt3a and 3b) but no differences in global methylation level. Aquatic organisms such as fish are, in most cases, exposed to multiple stressors that are either natural or anthropogenically introduced into their environment. Temperature, a potent physical stressor, and copper, a toxic chemical stressor when in excess, were addressed in the present study. Several studies have investigated the possible interaction between temperature and copper, but these were mainly conducted with adult fish. For instance, Perschbacher found an inverse relationship between temperature and copper toxicity in channel catfish, Ictalurus punctatus 29 .
In his experiments, copper was less toxic to catfish if applied at increasing temperatures. In contrast, Furuta et al. observed no significant effect of temperature on copper toxicity in Japanese flounder, Paralichthys olivaceus, and red sea bream, Pagrus major 30 . More recently, Lapointe et al. reported synergistic effects between heat stress and copper exposure in fathead minnows Pimephales promelas 20 . In our study, both heat stress and copper exposure resulted in increased mortality rates and delayed embryo hatching, suggesting adverse developmental consequences on zebrafish. We highlighted increased mortality of Cu-exposed embryos with increasing temperature (significant interaction), while both stressors independently influenced the hatching success. To date, there is no study considering the interaction effects of these two stressors on zebrafish embryogenesis, although a few studies have investigated temperature effects on the toxicity of other metals, i.e., cadmium 31 and nickel 19 . Looking first at the effects of copper, a previous study reported that exposure of zebrafish embryos to copper from fertilization to 72 hpf resulted in increased mortality between 5 and 24 hpf as well as decreased hatching success 16 . These observations are quite consistent with our results. During embryonic development, a strong influence of metals is often observed immediately after fertilization, particularly before hardening of the protective chorion. This structure is highly permeable before hardening and does not fully protect the embryo against metal penetration; thus, metals may exert direct toxic influence upon developing embryos 32,33 . Further, several mechanisms have been proposed to explain how copper delays the hatching process, e.g., inhibition of the hatching enzyme activity (chorionase), interference with the energy-requiring muscular movements necessary for the embryo to rupture the chorion, and alteration of oxygen uptake by the embryo 33,34 . However, these mechanisms are still unclear, and more than one or all of these processes may contribute to the observed delay in hatching. Secondly, zebrafish embryos normally develop within a temperature range of approximately 25-33 °C, and it was suggested that exposure to temperatures above or below these limits may produce abnormalities 35,36 . For instance, an increase in the developmental rates of zebrafish with increasing temperatures has been reported in several papers 37   rate dropped. Another study reported increased mortality and a higher number of developmental anomalies in embryos exposed for 2-3 h to 34-35 °C during early cleavage stages, which is similar to our results 39 . Increasing evidence shows that environmental challenges can affect normal development of organisms by altering DNA methylation patterns. Additionally, the DNA methylation reprogramming after fertilization is a dynamic mechanism that is essential for early development and which is considered to be sensitive to environmental stressors 6,7,40 . Dnmt3a and dnmt3b are required for this process, and the inactivation of both genes causes a complete failure in the genome-wide methylation 41 . In this context, the present study investigated whether heat stress and copper exposure applied during reprogramming affect DNA methylation patterns by altering the expression of dnmt3 genes in 96 hpf larvae. Altogether, we demonstrated that both stressors, alone or in combination, alter dnmt3 gene expression without changes in global cytosine methylation. The absence of effects on the global scale does not mean that DNA methylation modifications do not occur. The DNA methylation assay used in this study only tested for a global shift in methylated cytosines content and does not detect changes at specific loci. This is in line with several recent studies that have reported locus-specific methylation modifications and no changes in global methylation level in zebrafish embryos [42][43][44] . For instance, Bouwmeester et al. observed changes in methylation levels in the promoter regions of three selected genes vasa, vtg1 and cyp19a2 but no differences in global methylation level after exposure of embryos to several environmental contaminants from 0 to 72 hpf 43 .
In addition to measuring the global DNA methylation levels, we measured the expression of all dnmt3 genes aiming to associate their expression profiles with DNA methylation changes. Previous studies have indicated that zebrafish dnmt3 genes exhibit distinct spatio-temporal expression patterns during embryonic development 28,42,[45][46][47] . For instance, Aluru et al. have recently reported that dnmt3b genes are highly expressed in early stages of development and dnmt3a genes are more abundant in later stages, suggesting distinct roles for dnmt3a and 3b genes during development 42 . Takayama et al. observed tissue-specific expression of dnmt3a genes during later stages of development in zebrafish, suggesting a role for these genes in tissue differentiation 47 . Other studies reported localization of dnmt3b genes to specific regions of the developing zebrafish, as for instance dnmt3b1 in the aorta-gonad-mesonephros and caudal hematopoietic tissue regions 47 and dnmt3b2 in the brain, particularly in the tectum 3 , suggesting important roles in hematopoietic stem cell differentiation and neurogenesis, respectively. However, the specific roles of these genes in establishing de novo methylation patterns remain to date unknown. In the present study, we observed heat-induced upregulation of all dnmt3 genes and copper-induced upregulation of dnmt3b1, dnmt3b2 and dnmt3b3. Further, transcription levels of dnmt3a1 and dnmt3b4 were affected by copper exposure in a temperature-dependent way (significant interaction). The effects of heat stress and copper exposure on dnmt3 gene expression levels observed in this study suggest that the establishment of DNA methylation patterns could potentially be impacted. A few recent studies have shown differences in the expression patterns of dnmt3 genes following heat stress or pollutant exposure during zebrafish embryogenesis. Campos et al. observed that embryonic temperature influenced dnmt3a and dnmt3b transcript levels in a developmental stage-specific manner 28 . The authors have further reported differential responses of dnmt3 paralogues to embryonic temperature, suggesting that they play different roles in thermal epigenetic regulation of gene expression during early development. A direct comparison with our data is tricky since the experimental conditions were different and the dnmt3 expression during early development was reported to be very dynamic. Moreover, the oldest embryos were sampled at 80.8 hpf (protruding mouth stage) in Campos' experiment, while we sampled zebrafish after 96 hpf. At 80.8 hpf, they didn't observe significant differences between embryos exposed to 27 °C and 31 °C for any dnmt3 genes. More recently, Aluru et al. reported that exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) during zebrafish early development selectively alters the expression of dnmt genes, and these effects were developmental stage-specific 42 . The authors also reported changes in methylation levels in promoters of aryl hydrocarbon receptor (AHR) target genes but no differences in global methylation or hydroxymethylation levels.
Metallothioneins (MTs) are low molecular weight, cysteine-rich metal-binding proteins that play essential biological roles in metal homeostasis and detoxification 48 . Heat shock proteins (HSPs) are a set of evolutionary conserved proteins that are constitutively expressed and function as molecular chaperones in protein biogenesis and homeostasis 49 . As stress response proteins, MTs and HSPs confer cellular protection against a wide variety of stress factors and can be induced at the transcription level by diverse environmental stressors, including toxic metals and temperature 50,51 . Moreover, previous studies suggested that MTs 52 and HSPs 36 play significant roles during normal embryo development in zebrafish. In our study, we observed that exposure to copper increased the transcription level of mt2 and hsp70 in zebrafish larvae. Concomitantly, the hsp70 transcript level was also increased in response to thermal stress. Further, none of the applied stresses was able to affect the transcription level of hsp90. Similar results were previously observed in various fish species exposed either to heat or copper 20,53,54 . In zebrafish, several studies also reported that heat shock and/or copper exposure resulted in increased expression of MTs and/or HSPs in different adult tissues 55,56 or in embryos 36,57 . Overall, our gene expression results indicate that applied conditions were stressful to zebrafish larvae, thereby suggesting the susceptibility of reprogramming period to environmental challenges.
To conclude, our study is the first to report that the DNA methylation reprogramming is sensitive to the effects of environmental stressors in zebrafish embryos. As a matter of fact, we have shown that exposure to elevated temperature and/or copper during this critical developmental period resulted in high mortality, delayed hatching and increased expression of several stress genes, i.e., mt2 and hsp70. Furthermore, the expression of dnmt3 genes suggests possible impact on the establishment of DNA methylation patterns. In the future it would be interesting to focus on DNA methylation levels of selected genes (e.g., mt2 and hsp70) as well as on genome-wide methylation pattern and the relationships with gene expression. Further exploration is also needed to examine the adult phenotype of exposed embryos and to analyze molecular mechanisms that potentially link epigenetic effects and altered phenotypes. Zebrafish embryo collection and exposure. AB line wild-type zebrafish (Danio rerio) were maintained in a recirculating ZebTec housing system (Techniplast) at 28 °C with a 12:12 h (light/dark) photoperiod. Conductivity was maintained at approximately 500 μ S/cm, pH at 7.2 and dissolved oxygen at 95% saturation. Fish were fed ad libitum three times daily (twice on flake and once on Artemia nauplii). Males and females were placed together in breeding aquaria the night before spawning in a ratio of 3:2. The next morning, eggs were collected within 30 min of spawning, pooled from several breeding tanks, washed with ZebTec system water to remove debris and then randomly transferred into polypropylene beakers.
Zebrafish embryos were exposed to CuCl 2 (Sigma-Aldrich 751944) at nominal concentrations of 0 and 325 μ g/L from < 1 hpf (before complete hardening of the protective chorion) to 4 hpf (sphere stage) at either 26.5 °C or 34 °C. The tested treatments were chosen based on previous studies 16,37 . Six replicates for each treatment were used and each replicate consisted of a polypropylene beaker containing 150 ml of the respective treatment solutions and 166 ± 21 embryos. After the exposure, the embryos were rinsed thoroughly with fresh ZebTec system water and incubated at 26.5 °C until 96 hpf. Half of the water in each beaker was replaced every day to remove the wastes and dead embryos. Mortality rate and hatching efficiency were recorded daily until 96 hpf. At 96 hpf, larvae were washed, collected and stored at − 80 °C until DNA and RNA were isolated.
Total Cu concentration in the exposure water was measured using an Inductively Coupled Plasma-Mass Spectrometer (ICPMS, Elan DRC II). The accuracy of analytical methods was checked using certified reference standards (ICP/MS Multi-Element Standards ICP-MS-QC2-1, ACCUStandard). The mean concentration and standard deviation in the 325 μ g/L treatment were 280.4 ± 13.6 μ g/L.
Global DNA methylation. Genomic DNA was isolated from zebrafish larvae (96 hpf; n = 6 pools, 20 larvae/pool) using DNeasy Blood & Tissue Kit (Qiagen) according to manufacturer's protocol. DNA samples were treated with RNase A (Qiagen) to remove RNA contaminant. The DNA concentration and the quality of samples were assessed by spectrophotometry (NanoDrop 2000c Spectrophotometer, Thermo Scientific) and agarose gel electrophoresis. Quantification of global DNA methylation was examined by the colorimetric MethylFlash TM Methylated DNA Quantification Kit (Epigentek Group Inc.) following the manufacturer's instructions. The analysis was performed in duplicates with 100 ng of genomic DNA per sample. Absorption at 450 nm was determined. The percentage of methylated cytosines was calculated using the formula described by the manufacturer.
Quantitative real-time PCR (qPCR). The expression of 9 target genes was assessed by qPCR: 6 are involved in de novo DNA methylation (dnmt3a1, dnmt3a2, dnmt3b1, dnmt3b2, dnmt3b3 and dnmt3b4) and three play important roles in cell protection against environmental stressors (mt2, hsp70 and hsp90). Total RNA  was isolated from zebrafish larvae (96 hpf; n = 6 pools, 30 larvae/pool) with Extract-All Reagent (Eurobio) and DNase treated (DNA-free ™ Kit, Ambion) by following the manufacturers' protocol. The RNA concentration and the quality of samples were assessed by spectrophotometry (NanoDrop 2000c Spectrophotometer, Thermo Scientific) and agarose gel electrophoresis. Complementary DNA was synthesized from 1 μ g total RNA using the RevertAid H Minus First Strand cDNA Synthesis Kit (Thermo Scientific). Specific primers were selected from the literature or newly designed using Primer3Plus software and are listed in Table 3. Whenever possible, primers were designed to span at least one intron/exon border to avoid amplification of potential contaminating genomic DNA. Quantitative PCR was performed in a total volume of 20 μ l (5 μ l of cDNA, 2.5 μ l of each primer at 5 μ M, 10 μ l Master Mix 2x) with SYBR Green chemistry (Applied Biosystems) on a StepOnePlus Real-Time PCR System (Applied Biosystems). The PCR conditions used were 1 cycle at 95 °C for 10 min and 40 cycles at 95 °C for 15 s and 58-61.8 °C depending on gene for 1 min (Table 3). This was followed by a melting curve analysis to verify the specificity of the PCR products. Two technical replicates were used for each sample. No reverse transcriptase and no template controls were included on each plate to ensure the absence of background contamination. Gene expression data were calculated using the relative standard curve method 58 and normalized by geometric averaging of beta-actin (actb) and elongation factor alpha (ef1a). actb and ef1a were selected among several candidate housekeeping genes that remain stable over zebrafish development mentioned in the literature [59][60] , and after validation of their expression in our samples.
Statistical analysis. Data were expressed as mean ± S.E.M. Data were log-transformed when the conditions of normality and/or homogeneity of variance were not fully filled, or arcsine-square-root-transformed if expressed in percents. Differences between groups were analyzed using two-way analysis of variance (ANOVA 2) followed by a multiple comparison Tukey's HSD test at a 5% significant level. All tests were performed using the Statistica 5.5 software (StatSoft, Tulsa, OK, USA).