SI Yufeng , HE Feng, , WEN Haishen, LI Siping, and HE Huiwen

1) The Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Ministry of Education, Shanghai Ocean University, Shanghai 201306, China

2) The Key Laboratory of Mariculture, Ministry of Education, Ocean University of China, Qingdao 266003, China

Abstract Pituitary adenylate cyclase activating polypeptide (PACAP) and growth hormone-releasing hormone (GHRH) play important roles in the GH/IGF growth axis in fishes. To determine whether epigenetic change is involved in the regulation of pacap and ghrh responses to low salinity stress in Cynoglossus semilaevis, the correlation between growth traits, DNA methylation status and gene expression level in low salinity (15, S15) and optimal salinity (30, S30) at day 7 (D7) and day 60 (D60) were analyzed. Results showed that exposure to low salinity for 60 days attenuated C. semilaevis growth rate. Under low salinity, DNA methylation level of pacap promoter increased in females and decreased in males at day 7, but inverted at day 60. Additionally, pacap expression was up-regulated in both males and females. The pacap promoter methylation level was higher and expression level was lower in female than in male. The results suggest that pacap promoter methylation level is negatively correlated to mRNA level and positively correlated to body weight, while gene expression level is negatively related with body weight. With the decrease of salinity, DNA methylation level of ghrh promoter and exon1, as well as its gene expression displayed minor changes. Overall, pacap gene seems to play an important role in fish growth, contributing to female growth superiority, while ghrh gene seems not pertinent under salinity stress.The results indicate that low salinity potentially affects fish growth through regulating DNA methylation in pacap promoter. This study expands the understanding of the molecular mechanism of how salinity modulates fish growth from the epigenetic perspective.

Key words DNA methylation; Cynoglossus semilaevis; salinity stress; pacap; growth difference

1 Introduction

Salinity is one of the vital environmental factors in marine fish aquaculture (Callawayet al., 2012). Fluctuations in salinity severely affect the development, growth and reproduction of fish (Zhanget al., 2017). Sometimes salinity variation can result in heavy economic losses (Liet al., 2020). It has been studied in many researches due to its economic importance. However, most studies mainly focused on physiological and biochemical aspects, such as the measurement of the haematological and serum biochemical indexes (Alliotet al., 1983; Koshiishi, 1986; Watanabeet al., 1993). Only a few researches have been done on molecular aspects, such as osmoregulation genes cloning (Huanget al., 2010; Wanget al., 2012; Thanhet al.,2014) and transcriptome sequencing (Zhanget al., 2017;Siet al., 2018). Growth is a particularly important physiological index in fish culture. Fish growth is influenced by ecological factors, such as temperature, salinity, photoperiod, ammonia, oxygen and pH (Boeuf and Payan,2001). Of which, salinity is a critically important factor of the aquatic environment. Many studies have demonstrated the effect of external salinity on growth capacities in fish(Boeuf and Payan, 2001).

GH/IGF growth axis or hypothalamus-pituitary growth axis plays a major role in growth regulation in aquaculture animals (Foxet al., 2009; Fuet al., 2019). Several hormones are involved in the growth regulation process,such as the pituitary adenylate cyclase activating polypeptide (PACAP) and growth hormone-releasing hormone(GHRH) from hypothalamus, growth hormone (GH) from pituitary gland and growth hormone receptor (GHR) from liver. PACAP and GHRH are widely distributed in the central nervous system and peripheral organs and control the synthesis and secretion of GH (Vaudryet al., 2009). Thepacapandghrhgenes have been cloned in many fishes,such asCynoglossus semilaevis,Epinephelus coioides,Ictalurus punctatusandCoilia nasus(Maet al., 2011; Qianet al., 2012; Zhanget al., 2016). These two genes are predominantly expressed in the brain ofC. semilaevis(Jiet al., 2011; Maet al., 2011). Maet al. (2011) found thatghrhexpression levels in females were significantly higher than in males during the majority ofC. semilaevisearly development. While Jiet al. (2011) reported that there was no significant difference in the expressions ofpacapandghrhbetween the females and males before 8 months.However, theghrhmRNA level in males was significantly higher than in females between 9 and 12 months. PACAP and GHRH have close relationship and can be indicators of growth status, but their roles in fish growth and how they respond to salinity fluctuation still remain unclear.

Epigenetic processes, such as DNA methylation, may modulate environment-induced phenotypic variations by altering gene expression (Angerset al., 2010). DNA methylation is affected by an array of external factors, such as temperature (Varriale and Bernardi, 2006; Navarro-Martínet al., 2011; Metzger and Schulte, 2017; Wanget al.,2017), water quality (Zhouet al., 2001; Wanget al., 2009)and hormones (Contractoret al., 2004; Aniaguet al., 2008).To date, it is well known that DNA methylation is a major mechanism for gene silencing (Dinget al., 2012; Siet al.,2016; Liet al., 2017b). One of the most phenotypic influential environmental factors is salinity, which may contribute to both non-heritable and heritable phenotypic variations through DNA methylation changes. In half smooth tongue sole,igf2has a complex function in response to low salinity stress through changing regional DNA methylation and mRNA expression in the liver (Liet al.,2020).

Half smooth tongue sole (Cynoglossus semilaevis) belongs to flatfish and is an important commercial mariculture species (Chenet al., 2010; Wenet al., 2014). The females of the species grow one to two times faster than the males (Chenet al., 2007). It can survive in a wide range of salinities ranging from 14 to 37 parts per thousand (Siet al., 2018). The optimal salinity is 30 in aquaculture (Xuet al., 2017). Thus,C. semilaevisis an excellent model to study the mechanism behind the salinities’function on growth characteristics. The present study explored the changes of DNA methylation and gene expression patterns, with respect to growth rate ofC. semilaevissubjected to low salinity stress. The results provided a basic understanding of how salinity alterations influence the fish growth through epigenetic modifications.

2 Materials and Methods

2.1 Salinity Stress Experiment and Fish Sampling

A total of 240 female and male half smooth tongue sole were collected from a local fish farm in Lijin, China. These fish were cultured with prepared diet for 10 months in 5 m× 5 m × 1 m commercial fish ponds. The aquaculture conditions were controlled as 20 ± 0.5℃; 14 h:10 h light/dark cycle; O2≥ 4 ng mL?1; 30 salinity. The experimental group(Salinity 15, S15) and control group (Salinity 30, S30) both comprised of three tanks per treatment populated with randomly selected fish (n= 40/tank). The fish were acclimated at an optimal salinity of 30 for a period of 7 d.Before the start of the experiment, the salinity in the experimental group was decreased by 5 per day over a period of three days towards a target of 15 (S15). The control group was maintained at 30 before and during the experiment (S30). The experiment lasted for 60 d. At day 7 (D7) and day 60 (D60) of the experiment, six individuals (three males and three females) from each replicate were anesthetized using 0.1% tricaine methanesulfonate(MS-222). The brains were dissected and immediately frozen in liquid nitrogen and stored at ?80℃ for DNA and RNA extraction. Body weight was measured and the weight gain rate (WGR) was calculated as WGR = (terminal weight – original weight) / original weight × 100%.All animal handling was carried out in accordance with the ethical guidelines and protocols of Animal Care Committee of Ocean University of China.

2.2 Genomic DNA Isolation and DNA Bisulfite Modification and Methylation Analysis

Genomic DNA was extracted from brain using Marine Animals DNA Kit according to the manufacturer’s protocol (TransGen Biotech, Beijing). The DNA purity and concentration were measured using Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, USA), and their integrity was evaluated by 1% agarose gel electrophoresis.The DNA was stored at ?20℃ until further analyses. DNA samples (200 ng) were sodium bisulfite-modified using the MethylampTMDNA Modification Kit (QIAGEN, CA) following the manufacturer’s instructions. The CpG-rich regions ofpacap(Accession No. HQ334202.1) andghrh(Accession No. JF773679.1) were identified by online MethPrimer design software (http://www.urogene.org/methprimer/). Primers were designed using Oligo 6.0 software(Table 1). Cycling conditions were 94℃ for 5 min followed by 40 cycles of 94℃ for 30 s, Tm for 30 s, and 72℃for 30 s with a final extension of 10 min at 72℃. The 25 μL reaction mixture included TaKaRa EpiTaqTMHS 0.125 μL, 10 × EpiTaq PCR buffer 2.5 μL, MgCI22.5 μL, dNTP mixture 3 μL, genomic DNA template 3 μL (＜100 ng), F(forward) and R (reverse) primers 1 μL, respectively. PCR product was purified and cloned into pEasy-T1 vector, sequenced, and analyzed using DNAMAN software. Non-CpG cytosines of each sequence served as internal controls to verify bisulfite DNA modification efficiency (P＞95% in all samples).

2.3 Total RNA Extraction and Reverse Transcription

Total RNA was extracted from all tissues ofC. semilaevisusing RNAiso reagent (TaKaRa, Japan). The concentration of total RNA was measured. Reverse transcription was carried out using two-step method with Prime-ScriptTMRT reagent Kit with gDNA Eraser (Perfect Real Time) (TaKaRa, Japan). A 10 μL reverse transcriptase reaction mixture containing 5 × gDNA Eraser Buffer 2 μL,gDNA Eraser 1 μL, total RNA 1 μg, and RNase-free ddH2O up to 10 μL at 42℃ was incubated for 2min, and all the first-step reaction solution was added to the second-step reaction solution with 1 μL PrimeScriptRTEnzyme Mix I,1 μL RT Primer Mix, 4 μL 5 × PrimeScript Buffer 2 (for RealTime),4μLRNase-free ddH2Oandwasincubated at 37℃for15minandthen at 85℃for5s. Thesynthesized cDNA was stored at ?20℃ for later use.

2.4 Quantitative Real-Time PCR (qPCR) Analyses

The relative levels ofpacapandghrhmRNA were determined using brain cDNA from half smooth tongue sole.Gene specific primers were designed using Oligo 6.0 software (Table 1). Quantitative PCR was performed using the StepOne Plus Real-Time PCR system (Applied Biosystems) and SYBR?Premix Ex Taq? (Tli RNaseH Plus)Kit (TaKaRa, Japan, Code No. RR420A) according to the manufacturer’s protocols. The 20 μL mixture of PCR consisted of 10 μL SYBR?Premix Ex Taq? (2×), 0.4 μL both primers, 0.4 μL of ROX and 2 μL DNA template, and RNasefree water was added to the final volume. PCR amplification was performed in a 96-well optical plate at 95℃ for 30 s, followed by 40 cycles of 95℃ for 5 s, 58℃ for 30 s,and a final extension at 72℃ for 2 min. 18S ribosomal RNA (rRNA) gene was taken as a reference gene, and was amplified under the same conditions. Each experiment included three repetitions. Relative gene expression was calculated using the 2?ΔΔCTmethod (Livak and Schmittgen,2001).

2.5 Statistical Analysis

All the data were calculated by Microsoft Excel. The results were expressed as mean ± SE (standard error). The differences between different salinity stresses, or different sexes, or different sampling time were determined by oneway ANOVA followed by Duncan’s multiple comparison tests using SPSS 20.0 (SPSS Co. Ltd., Chicago). Multiway ANOVA was performed to evaluate the combined effects of salinity, sex and sampling time using SPSS 20.0.Associations between methylation status and mRNA expression or growth traits were examined by Pearson’s tests using SPSS 20.0. Statistical significance was accepted atP＜ 0.05.

Table1 Primers used for BS-PCR and quantitative PCR of Cspacap and Csghrh

3 Results

3.1 Growth Traits Under Salinity Stress

Fig.1 The weight gain rate of female and male individuals of C. semilaevis in S15 and S30 groups at D7 and D60.Blue represents female and red represents male. Different lowercase letters indicate the statistical significant difference between groups (P ＜ 0.05).

The growth rate under different salinities is presented in Fig.1. During the first 7 days, weight gain rate (WGR)remained unchanged under low salinity treatment. At the end of the study (D60), WGR of both female and maleC.semilaevissignificantly slowed down (P＜ 0.05). In the experiment (S15) and control (S30) groups, WGR of both female and maleC. semilaevissignificantly increased with time elapsed (P＜ 0.05). When comparing males and females in every single group, female WGR was significantly higher than male WGR (P＜ 0.05).

3.2 The DNA Methylation Levels as a Whole and at Single CpG Site in pacap Promoter Under Different Salinities

Fig.2 Diagrams of DNA methylation patterns in female and male pacap promoters in S15 and S30 at D7 and D60. Filled and open circles denote methylated and unmethylated sites, respectively. Each line represents one sequenced clone. The first line indicates the localization of studied CpG sites related to the sequence of pacap promoter.

The methylation pattern details ofpacappromoter in the S15 and S30 groups at D7 and D60 are shown in Fig.2.The DNA methylation levels ofpacappromoter were generally low in all the groups. The methylation levels were calculated and analyzed in Fig.3A. There was no significant difference between the two salinity groups at D7 and D60 respectively. Specifically in female fish, the CpG methylation status of thepacappromoter increased, from 5.71% ± 0.83% at D7 to 7.46% ± 1.46% at D60 only under normal salinity environment (P= 0.06). When comparing the methylation level of females at the different time points, female CpG methylation status of thepacappromoter slightly decreased from 6.19% ± 0.95% to 6.03% ±0.73% in the low salinity treatment (P= 0.85). When comparing the methylation level under different salinities, female CpG methylation level ofpacapincreased when subjected to S15 at D7 (P= 0.58), while it showed a reverse trend from S30 to S15 at D60 (P= 0.11). In male fish, the methylation level of thepacappromoter decreased from 5.24% ± 0.95% to 3.49% ± 0.28% over time in the control group (P= 0.06). At low salinity, the methylation level increased from 4.13% ± 0.73% at D7 to 5.08%± 1.67% at D60 (P= 0.28). At the early stage of the experiment, male CpG methylation level decreased in low salinity (P= 0.21), while the trend was reversed at the end of the experiment (P= 0.08). Overall, the DNA methylation levels ofpacappromoter in females were higher than those in males. Specifically, there was a significant difference between males and females in S30D60 group (P=0.001). Multi-way ANOVA analysis showed that sex had significant effect onpacapDNA methylation (P= 0.001),while the other two factors, sampling time and salinity had no significant effect on DNA methylation.

Fig.3 DNA methylation levels of pacap promoter (A), every single CpG site of pacap promoter (B) and CpG18 of pacap promoter (C) of female and male C. semilaevis in low and normal salinities. Different letters above the error bars mean significant difference at P ＜ 0.05.

Furthermore, single CpG sites ofpacappromoter were studied to determine their sensitivity to low salinity. There were 21 CpG sites inpacappromoter. Of which, CpG1,CpG5, CpG6, CpG7, CpG8, CpG9, CpG11, CpG16, CpG18,CpG20 had significant differences (P＜ 0.05) (Fig.3B). Interestingly, the methylation level of CpG18 site was consistently higher than those of the other sites. This site was targeted for analysis. The methylation level of CpG18 increased with the decrease of salinity in both male and female at D7. At the end of the experiment, females had increased methylation level, while males showed decreased methylation level under low salinity stress (Fig.3C).

3.3 The Effect of Low Salinity on the DNA Methylation Levels of ghrh Promoter and Exon1

The methylation pattern details ofghrhpromoter and exon1 under short-term salinity stress and long-term salinity stress are presented in Fig.4. Theghrhpromoter and exon1 had 9 and 3 CpG sites, respectively. All the CpG sites were highly methylated. The results of the methylation analyses are presented in Fig.5. There was no significant difference between low salinity group and high salinity group inghrhpromoter methylation at D7 and D60.The methylation levels ofghrhexon1 were close to or equal to 100%. In the S15 group, the methylation status ofghrhexon1 in maleC. semilaevisincreased significantly from 94.44% ± 5.09% at D7 to 100.00% ± 0.00% at D60 (P=0.014). There was no significant difference in other groups.

Fig.4 Diagrams of DNA methylation patterns in female and male ghrh promoter (A) and exon1(B) in S15 and S30 at D7 and D60. Filled and open circles denote methylated and unmethylated sites, respectively. Each line represents one sequenced clone. The first line indicates the localization of studied CpG sites related to the sequence.

Fig.5 DNA methylation levels of ghrh promoter (A) and exon1 (B) of female and male C. semilaevis in low and optimal salinities. Different letters above the error bars mean significant difference at P ＜ 0.05.

3.4 The Gene Expression and the Relationship with DNA Methylation Level

Thepacaptranscript level had no significant change in the different salinity groups, but the relative expression level in males was markedly higher than that of females(Fig.6A). At D7,pacapgene expression was down-regulated at low salinity in both males and females. At D60,pacapgene had the same expression trend with D7 when subject to low salinity. In both S15 and S30, the expression ofpacapgene gradually decreased during the duration of the experiment. TheghrhmRNA level had no significant differences between the different experimental groups (Fig.6B). The femaleghrhexpression was higher than males at D7, while the result was reversed at D60.TheghrhmRNA level was up-regulated under low salinity stress, but not significantly.

Fig.6 Expression levels of pacap (A) and ghrh (B) genes in female and male C. semilaevis under low and normal salinities at D7 and D60. Different letters above the error bars mean significant difference at P ＜ 0.05.

Fig.7 The correlation of DNA methylation level and gene mRNA expression. (A) Correlation between brain pacap promoter methylation level and pacap mRNA expression; (B) Correlation between brain ghrh promoter methylation level and ghrh mRNA expression; (C) Correlation between brain ghrh exon1 methylation level and ghrh mRNA expression.

The correlation of gene expression level and DNA methylation level ofpacappromoter is demonstrated in Fig.7A.The DNA methylation ofpacappromoter was inversely correlated topacapgene expression (R= ?0.57,P= 0.16).The DNA methylation ofghrhpromoter was positively correlated toghrhgene expression (R= 0.18,P= 0.46)(Fig.7B). The DNA methylation ofghrhexon1 was positively correlated toghrhgene expression (R= 0.20,P=0.39) (Fig.7C).

3.5 The Correlation Between Growth Traits and DNA Methylation or Gene Expression

DNA methylation level ofpacappromoter was positively correlated with the fish weight (R= 0.55,P= 0.18)(Fig.8A). The correlation coefficients betweenghrhpromoter/exon1 methylation and fish weight were 0.33 (P=0.45) and 0.48 (P= 0.24) respectively, showing a positive correlation (Figs.8B, C).

Figs.8E, F show the association between gene expression and body weight. ThepacapmRNA level was negatively correlated with fish weight (R= ?0.83,P= 0.01).There was no significant correlation between body weight andghrhexpression level (R= ?0.01,P= 0.97).

Fig.8 The correlation between fish weight and DNA methylation or gene expression. (A) Correlation between brain pacap promoter methylation level and body weight; (B) Correlation between brain ghrh promoter methylation level and body weight; (C) Correlation between brain ghrh exon1 methylation level and body weight; (D) Correlation between brain pacap mRNA expression level and body weight; (E) Correlation between brain ghrh mRNA expression level and body weight.

4 Discussion

Salinity is one of the most important factors that affect survival and growth of fish. Half smooth tongue sole had different growth performance in different salinities. Wanget al. (2003) found that the growth rate decreased at higher or lower salinity than that at the optimal 30. The growth of fish larvae exhibited a remarkable statistical difference between salinities above 25 and below 20 (P＜ 0.05) (Xuet al., 2005). Tianet al. (2010) showed that no significant difference was found in the growth ofC. semilaevisat different salinities (22, 26, 30) during the 8-week culture. In another study, specific growth rate (SGR) of juvenileC. semilaevisfirstly increased and then decreased at high salinity. The fish grew the fast under S10-S20 during 60-d cultivation (Fanget al., 2016). In our study, WGR of fish did not change significantly under low salinity stress during the first 7 days, and decreased significantly at day 60(P＜ 0.05). This result indicates that low salinity stress affects the growth performance during long farming. The different results may be explained by fish age, experimental durations, and other aquaculture conditions such as food and feeding intervals.

The females ofC. semilaevisgrow one to two times faster than males (Chenet al., 2007). The difference in growth performance between male and female half smooth tongue sole occurred 250 days post hatching (dph). In our present study, males and females of 10 months oldC. semilaevisshowed a significant difference in growth performance (P＜ 0.05). WGR in females was much higher than that in males regardless of salinity conditions. According to this result, it was deduced that the growth advantage of females was not affected by the low salinity stress. Specifically, WGR in males decreased less than that in females under long-term low salinity stress, which indicated that low salinity had a minor effect on maleC. semilaevisgrowth. In our study, both male and female half smooth tongue sole had the best growth performance under normal salinity.

DNA methylation appeared to be an indispensable response to environmental stress by regulating gene expression (Liet al., 2017a, 2017b; Metzger and Schulte, 2018;Liet al., 2020). Congruent with previous reports (Siet al.,2015), methylation levels ofpacapgene were generally low in the present study. Low salinity had no significant effect on the DNA methylation level ofpacapgene promoter, but the changing trend of DNA methylation displayed differences in both female and maleC. semilaeviswhen subjected to low salinity. During the first 7 days, the DNA methylation increased in females and declined in males. At D60, the DNA methylation was down-regulated in females and up-regulated in males. This indicates that low salinity had negative effect on female DNA methylation in the long run (60 days), while male methylation levels were less affected by low salinity. The high expression ofpacapgene was accompanied with slow growth rate (Maet al., 2011; Siet al., 2015). In our study,pacapmRNA level was negatively correlated with fish weight although salinity did not have a significant impact on the gene expression levels ofpacapgene. The gene expression level was higher in S15 than in S30 at D7. The change in salinity in the early stage might force the fish consume more energy to regulate the osmotic balance, hence the growth slowed down. In turn, higherpacapmRNA levels resulted from the reduced expression ofghgene (data not published), which depends on the feedback regulation mechanism of the fish (Maet al., 2011). At the end of the experiment,pacapmRNA level of males and females had no big difference between S15 and S30. This indicates that by day 60, fish had adapted to the low salinity stress. ThepacapDNA methylation level was negatively correlated with the gene expression and positively correlated with fish growth in both low and optimal salinities. We infer that female growth is affected by low salinity through increasing DNA methylation ofpacapthat turns on gene expression.

Generally, DNA methylation inhibits transcription through prevent promoter binding to transcription factors (Nanet al.,1998). The putative transcription factor binding sequences ofpacapCpG sites were predicted by TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html).In silicosimulation analysis found some potential regulators, such as Sp1, AP-1, E2F-1, c-Ets-1, LEF-1, PEBP2αA1, Sp2, Sp3,Sp4, c-Myb, GR, GATA-1, HNF-3 (Fig.9). AP-1 (activating protein-1) is a component including Jun, Fos, or ATF(activating transcription factor) subunits that bind to a common DNA site, the AP-1 binding site (Karinet al., 1997).The binding site of PACAP has been reported in other species (Tao and Boulding, 2003). The AP-1 complex containing c-Fos has been demonstrated to mediate the effect of PACAP on Bcl-2 gene expression in cerebellar granule neurons (Aubertet al., 2006). In parallel to c-Fos/c-Jun induction, PACAP rapidly activated the heterodimeric transcription factor AP-1 (Sch?feret al., 1996). There were no reports describing that AP-1 transcription factors had association with the regulation ofpacapgene in teleosts. Thus,salinity may inhibit or enhance the transcription factors binding topacappromoter to modulate gene expression.

Maet al.(2011) has reported thatghrhexpression level changes during the half smooth tongue sole development.At 20–100 days post hatching (dph), theghrhexpression in females is higher than that in males. After 120 dph,ghrhexpression level is higher in males than in females.Jiet al.(2011) found that before 8 months of age,ghrhexpression showed no significant difference between males and females. In 9–12 months old half smooth tongue sole,males expressed significantly higherghrhthan females.In our study, females expressed moreghrhthan males at D7, while males expressed moreghrhthan females at D60 of the experiment. The result demonstrated thatghrhexpression levels differed between males and females in 10–12 months old fish though the difference is not significant. Monteroet al. (1998) found that PACAP and GHRH play equal roles in regulating GH secretion in amphibians. GHRH played a key role in mammals and birds.In contrast, PACAP has a strong function of stimulating GH secretion in fish, while GHRH effect is weak or nonexistent (Monteroet al., 1998). Our study on DNA methylation ofghrhpromoter and exon1 demonstrated high DNA methylation patterns, which exceed 80%, in both low salinity treatment and control groups, which indicates the low expression ofghrh. Moreover, theghrhgene expression had no correlation with fish weight (R= ?0.01). This may explain the weak role ofghrhin fish. Salinity stress had no obvious effect onghrhexpression and methylation ofC. semilaevis. Thus,ghrhdoes not seem to play a role inC. semilaevisresponse to low salinity stress.

Fig.9 Potential transcription factors of pacap promoter are shown. The ‘TF’ in the right of the nucleotide is the abbreviation of transcription factor. The dotted lines indicate the consensus binding sequence of the transcription factor. Primers used for pacap promoter amplification are shown in red font. The CpG sites are highlighted by a yellow background.

In conclusion, this research found that long term low salinity treatment had a negative effect on the growth of half smooth tongue sole. Low salinity might elevatepacapexpression through decreasingpacappromoter methylation level to inhibit fish growth. Long-term low salinity stress reduced male and female growth traits differences,but females still had a growth advantage. This may be caused by epigenetic mechanism ofpacapgene. The DNA methylation level and mRNA expression level ofghrhgene were not significantly changed under low salinity stress.This grantspacapgene a more important role in fish growth overghrh, especially with regard to female growth superiority. The results provide a basic understanding of the molecular mechanism of salinity affecting fish growth from the epigenetic perspective. More work still needs to be conducted to elucidate the functions of other gene regulation factors such as transcription factors and others.

Acknowledgements

We thank Shangdong Dongying Farm for providing the animals. This research was supported by the China Postdoctoral Science Fundation (No. 2019M651472), the National Natural Science Foundation of China (No. 316726 42), and the Key Laboratory of Mariculture of Ministry of Education, Ocean University of China (No. KLM20180 09).