Fish treatments were carried out according to the recommendations in the Guidelines for the Care and Use of Laboratory Animals of the National Advisory Committee for Laboratory Animal Research in China and approved by the Animal Care Committee of Hunan Normal University(Permit Number: 4237). Individual 2nCC and 3nCC were randomly selected. All samples were cultured in open pools (0.067 ha) with suitable pH (7.0-8.5), water temperature (22-24°C), dissolved oxygen content (5.0-8.0 mg/L) and adequate forage at the State Key Laboratory of Developmental Biology of Freshwater Fish, Hunan Normal University, China. All fish were anesthetized with 100 mg/L MS-222 (Sigma-Aldrich, St Louis, MO, USA) before blood collection. The study was conducted in accordance with the local legislation and institutional requirements.

All the C. auratus used in this study were from the Dongting Lake water system. After detection by flow cytometry, 2nCC and 3nCC were raised separately for subsequent collection. Two different ploidy C. auratus with good growth conditions during the breeding season (April) were selected for sampling. Before dissection, the experimental fish were anesthetized with 100 mg/L MS-222 (Sigma-Aldrich St. Louis MO USA), and the brains of diploid and triploid Carassius auratus were removed under sterile conditions. Subsequently, the pituitary, gonad, and other tissues were rapidly sectioned for RNA and DNA extraction. The RNA and DNA were then cut into small pieces, placed in EP tubes (Axygen), frozen with liquid nitrogen, and transferred to a -80°C refrigerator for histology. Sectional experimental materials were fixed with Bouin’s solution. The samples used in the experiments contained three and more biological replicates.

Total RNA was extracted from the brain, pituitary, and gonad of 2nCC and 3nCC in April (breeding season) under sterile and non-enzyme-contaminated experimental conditions using Total RNA Kit II from Omega. The OD value, concentration, and integrity of the extracted RNA were detected, and standard RNA was stored at -80°C. Total brain, pituitary, and gonad RNA extracted from 2nCC and 3nCC individuals were used as templates, and the PrimeScript™ RT Reagent Kit with gDNA Eraser (perfect kit for Real-Time PCR, TaKaRa RR047A) was used to generate cDNA using the reverse transcription. The cDNA obtained was stored in a -20°C refrigerator. The samples used in the experiments contained three and more biological replicates.

C. auratus is closely related to C. auratus red var. We used previous research as a reference for cloning the coding regions of HPG axis-related genes in C. auratus red var. and autotetraploid C. auratus (21). The cDNA obtained by reverse transcription from the brain, pituitary, and gonad tissues of 2nCC and 3nCC was used as a starting template for cloning the coding region of the corresponding gene. PCR amplification was initiated using the primers prepared in Table 1 , and each gene was amplified. Coding region sequences of Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes from two types of fish were obtained. The total reaction system was combined int 20 μL, 10 μL Premix Taq (LA Taq™ Version 2.0 plus dye) (TaKaRa Company, RR903), 1 μL of forward and reverse primers, 7 μL of sterilized double-distilled water, and 1 μL of cDNA template. The samples used in the experiments contained three and more biological replicates.

The amplification reaction procedures for each gene were generally similar. PCR products obtained were detected with electrophoresis. When fragments of the desired size were found, they were recovered using gel. For detailed experimental procedures, refer to the attached instruction manual for the DNA gel recovery kit (Shanghai Sangong, B518131-0100). The products were ligated and transformed, and a single colony was selected from the ampicillin-resistant petri dish (30 per petri dish) for PCR. The total reaction system was 10 μL, and 1 μL of the bacterial solution was taken as the starting point Template, 5 μL of 2× Rapid Taq Master Mix (Vazyme, P222-AA), 3 μL of sterilized double-distilled water, 0.5 μL of forward and reverse primers. The primers are specific for the corresponding genes in Table 1 . The product was detected, and the bacterial liquid similar to the target fragment was screened to complete the detection. Twenty samples for each gene were sent for detection to reduce the error caused by the small number of sequences.

Quantitative reverse transcription PCR (RT-qPCR) was used to detect mRNA transcript levels of HPG axis-related genes in 2nCC and 3nCC. Specific RT-qPCR primers were designed to code region sequences of HPG axis-related genes in 2nCC and 3nCC. The primer sequences are shown in Table 2 . β-actin is the internal reference gene. The cDNA of each tissue obtained in 2.2 was diluted with sterile water at a ratio of 1:5 and used as the starting template for RT-qPCR. The total reaction system was 10 μL, of which 2x Power SYBR Green PCR Master Mix (ABI) 5 μL sterilized water 3 μL, forward and reverse primers 0.5 μL each, and diluted cDNA 1 μL. Fluorescence assay was performed using the Prism7500 Sequence Detection System (ABI), and biological replicates were performed to ensure reliable experimental results.

The DNA of each prepared tissue was treated with an EZ DNA Methylation-Gold kit (D5005, ZYMO RESEARCH, USA). The CT Conversion Reagent solution was prepared to collect the total DNA (400 ng) from each tissue. Subsequently, 20 μL of sterilized water was added to a 200 μL PCR tube, and then 130 μL of the prepared CT Conversion Reagent solution was added to the PCR tube and mixed well. The above-mentioned mixed PCR samples were put into the PCR machine, and the set program was 98°C for 10 min and 64°C for 150 min. After PCR, the product was temporarily stored at 4°C for 20 h. Subsequently, 600 μL of M-Binding Buffer was first added to the adsorption column, followed by the processed PCR samples, and they were mixed by inversion after closing the tube cap. The samples were then centrifuged, and the filtrate was poured off. Then 200 μL M-wash Buffer was drawn and added to the adsorption column, centrifuged, and the filtrate was poured off. Lastly, 200 μL M-Desulphonation Buffer was drawn and put in the adsorption column, centrifuged, and the filtrate was poured off. Subsequently, M-Desulphonation Buffer was drawn and put into M-Elution Buffer heated in a water bath at 56°C. Then 16 μL of the solution was pipetted into the adsorption column, incubated at room temperature for 15 min, centrifuged, and the obtained DNA was stored at -20°C. The samples used in the experiments contained three and more biological replicates.

Based on the mRNA sequences of C. auratus red var. Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes in the GenBank database, the complete DNA sequences of the above genes were searched from the C. auratus red var. genome. Then, the mRNA and DNA sequences of these genes were compared. When the first exon was found, the upstream 2000 bp sequence was used as the promoter region of the gene, and the sequence of the selected region was entered into the Meth Primer online website and predicted CpG-rich regions in the promoter regions of Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes. The CpG-enriched region of each gene (if a single gene has multiple CpG-enriched regions, randomly select a region from them) was selected as the original target sequence to design primers. The primer sequences are shown in Table 3 . They are used to amplify CpG-enriched promoters of the corresponding 2nCC and 3nCC genes. The bisulfite PCR (BS-PCR) primers were designed with Primer Premier 5.0 software using the CpG-enriched region of the HPG axis-related gene promoter cloned from C. auratus with a different ploidy than the original target sequence. BS-PCR was performed using DNA from 2.2 as a template and the specific primers in Table 4 . The total reaction system was 10 μL of LA DNA polymerase 5 μL, 0.5 μL forward and reverse primers, 3.5 μL sterilized water each, and 0.5 μL of template DNA. The resulting cloned products were subjected to electrophoresis analysis, and strips containing the expected fragment length were subjected to gel recovery followed by ligation, transformation, and sequencing.

The pituitary, ovary, and testis tissues of 2nCC and 3nCC in the breeding season (April) were removed and soaked in Bouin’s solution for 48 h. After 48 h, small tissues were separated (pituitary tissues do not need to be separated), and the excess fixative was blotted on the surface with filter paper. The separated tissues were placed in gradient alcohol for paraffin embedding. The dehydration time and alcohol gradient are 70% alcohol for 1 h (2-3 times, observable 80% alcohol, overnight; 90%, 95%, 100% alcohol, 30 min each in turn). After alcohol gradient dehydration, the materials were transferred to xylene solution and allowed to stand for 5-10 min. Subsequently, the paraffin was melted and put in an oven at about 60°C for 2-3 h to ensure the paraffin was in a molten liquid state. The tissues soaked in wax were moved to an embedding rack, and paraffin was allowed to cool. Subsequently, the tissue block was removed from the packaging. The wax was removed from the embedded rack and trimmed, and serial sections with a thickness of about 5-7 mm were prepared using a microtome. The cut wax slice was flattened with warm water, removed from the glass slide, and placed in a drying machine (42°C, 24-48 h). After drying, they were put in the slide box and prepared for subsequent experiments. The samples used in the experiments contained three and more biological replicates.

The prepared pituitary tissue sections were placed in xylene solution for dewaxing. After deparaffinization, they were rehydrated in gradient alcohol at 100%, 95%, 90%, 80%, and 70%, each for 5 min. The 0.5% periodic acid solution was oxidized for 5 min. After oxidation, the sections were stained with Schiff reagent (Shanghai Soleibao Biological Co., Ltd.) for 15 min and rinsed with running water for about 10 min. Subsequently, the sections were stained with 1% orange-yellow G Solution and 5% phosphotungstic acid solution for 10-15 s and rinsed with running water for 15 s. The sections were then stained with 1% methyl blue solution for 2-5 s and washed with 1% acetic acid solution for 1-5 s. Finally, the sections were dehydrated in gradient alcohol 70%, 80%, 90%, 95%, and 100% for 5 min each. The xylene was transparent for 5 min, mounted with gum, and then observed and photographed with a microscope (Olympus CX41).

The prepared ovary and testis tissue sections were put in xylene solution for dewaxing. After Deacetylation and rehydration were consistent with the above experiments; After rehydration, the tissues were stained with the hematoxylin staining solution for 25 min-1 h. The tissues were then rinsed and soaked in the glacial acetic acid solution for about 5-10 s, and color changes were observed. When the tissues turned red, they were removed from the stain and washed with running water. Subsequently, the slices were soaked in an alkaline aqueous solution for about 5-10 s, and color changes were observed. Tissues were removed from the stain once they turned blue, washed with running water, and dehydrated in 70%, 80%, and 90% graded alcohol for 5 min each. After dehydration, the sections were put in an eosin staining solution for 5-20 min. The sections were then transferred to 95% and 100% alcohol solution and allowed to stand for 5 min for dehydration. Subsequently, the sections were transferred to xylene for about 4-5 min and sealed with gum. The residual mounting medium was wiped with xylene, observed, and photographed with an Olympus CX41 microscope.

The immunofluorescence technique analyzed the localization of Fshβ and Lhβ in the pituitary tissue of crucian diploid carp. Immunofluorescence technology detects the expression and localization of proteins and polypeptides in biological cells or tissues using the principle that antigens can be specifically linked to antibodies (21). The reagents required included rabbit anti-Lhβ (primary antibody), rabbit anti-Fshβ (primary antibody), donkey anti-rabbit IgG labeled with FITC (fluorescent secondary antibody), immuno-highlighter, 20xPBS (phosphate buffered saline, purchased from BBI Bio), anti-degreasing slides and polylysine-treated anti-fluorescence quenchers, Triton-X 100 purchased from Shanghai Sangon Bio Co., Ltd., blocking buffer BSA and peroxidase blocking solution. The other reagents needed were prepared in the laboratory. The negative control was prepared using diluted PBS instead of the primary antibody. The pituitary tissue section obtained in 2.6 were washed three times with 1xPBS, 5 min/time. Subsequently, excess 1xPBS on the section was removed, and peroxidase blocking solution was added dropwise.

Sections were placed in a wet box at room temperature for 30 mins and then washed thrice with 1xPBS for 5 minutes each. The surrounding tissue material was wiped, and Triton-X 100 (0.5%) was added dropwise to soak the material and placed in the wet box at room temperature for 40 min. The tissues were then washed with 1xPBS three times, 5 min/time. After washing, microwave high-temperature treatment was used to repair the antigen of the test object. Subsequently, the test tissue was placed in citric acid antigen retrieval solution for high-temperature treatment for 15 min and then placed at room temperature for 1-2 min. After incubation, the slices were washed thrice with 1xPBS, 5 min/time. Then BSA blocking solution was added to the tissue material and stored in a humid box at room temperature for about 1-2 h. After 2 h, the slices were removed from the box, wiped, and used as an immunohighlight pen.

The tissue on the same slice was divided into different sections. The diluted primary antibody (the primary antibody was diluted with 1xPBS at a ratio of 1:150) and 1xPBS buffer (negative control) were added dropwise. The slice was placed in the wet box and stored light shield box. The slices were hybridized at 4°C overnight (12-14 h) to recover the primary antibody. The recovered antibody was incubated at 37°C for 30 min and washed thrice with 1xPBS, 5 min/time. Subsequently, diluted fluorescent two antibody (dilute the secondary antibody at a ratio of 1:150 with 1xPBS) was dropped into the antibody in a dark environment. The slices were then hybridized at room temperature for 1-2 h in a humidified dark box, washed thrice with 1xPBS, 5 min/time (washing was performed in a dark environment). Excess buffer and added anti-fluorescence quencher were removed under dark conditions. Subsequently, the slices were placed under a fluorescence microscope (Olympus BX63) to observe and photograph the tissue fluorescence site. The immunofluorescence technique was used to observe the localization of Fshr in the ovaries and testis of 2nCC.

The fluorescence experiments mentioned in the manuscript were performed using a Prism7500 Sequence Detection System (ABI) and biological replicates were performed to ensure the reliability of the results. The results were viewed using the widely recommended relative standard curve method (2^-ΔΔCT method), and the experimental values were statistically calculated using Excel, analyzed for significance using SPSS 19.0 software, and finally plotted using Sigma Plot 12.5 software.

Coding region sequences from Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes were cloned into 2nCC and 3nCC. The Gnrh2 gene contains two diploid copies, Gnrh2-2nCC-1 and Gnrh2-2nCC-2, and three triploid copies, Gnrh2-3nCC-1 and Gnrh2-3nCC-1 and Gnrh2-3nCC-3. The length of the Gnrh2 gene encoding region in both fish is 261 bp and encodes 86 amino acids. The Fshr gene contains two diploid copies, Fshr-2nCC-1 and Fshr-2nCC-2, and three triploid copies, Fshr-3nCC-1 and Fshr-3nCC-1 and Fshr-3nCC-3. The length of coding regions of the Fshr gene in both fish is 2013 bp and encodes 670 amino acids. The Lhr gene contains two copies of the diploid called Lhr-2nCC-1 and Lhr-2nCC-2.

The length of the gene coding region is 2127 bp and encodes 708 amino acids; it contains three copies of the triploid named Lhr-3nCC-1, Lhr-3nCC-2, and Lhr-3nCC-3. The length of the gene coding regions were 2124 bp, 2127 bp, and 2127 bp, encoding 707, 708, and 708 amino acids, respectively. In both diploid and triploid, there is only one copy of the Fshβ gene, and the consensus length of the gene coding region is 393 bp, encoding 130 amino acids. There is only one copy of the Lhβ gene in two different ploidy fish, and the consensus length of the gene coding region is 423 bp and encodes 140 amino acids. Using BioEdit software to analyze the similarity of amino acid sequence and nucleotide sequence of HPG axis-related genes in 2nCC and 3nCC, it was found that the similarity of the Fshβ and Lhβ genes of the two fish was 100%. It was also observed that some base sites of the coding regions of Gnrh2, Fshr, and Lhr genes were mutated, and some bases were deleted in the Lhr gene coding region.

MEGA11 software was used to analyze the developmental relationship of Gnrh2, Fshr, and Lhr genes copies in 2nCC and 3nCC, and the genetic homology between the two was represented using evolutionary trees ( Figure 1 ). Analysis of the Gnrh2 gene showed that Gnrh2-2nCC-1 and Gnrh2-3nCC-1 had the same sequence, Gnrh2-2nCC-2 had the same sequence as Gnrh2-3nCC-2, and Gnrh2-3nCC-3 was similar to Gnrh2-2nCC-1. Closer analysis of the Fshr gene showed that the sequence similarity was greatest between Fshr-2nCC-1 and Fshr-3nCC-1 and between Fshr-2nCC-2 and Fshr-3nCC-2. The results also showed the greatest sequence similarity between Fshr-3nCC-3 and Fshr-2nCC-2. Analysis of the Lhr gene showed that the sequence similarity between Lhr-2nCC-1 and Lhr-3nCC-1 and between Lhr-2nCC-2 and Lhr-3nCC-2 was relatively close. In contrast, Lhr-3nCC-3 and Lhr-2nCC-2 had the closest sequence similarity. Table 5 shows the sequence and amino acid sequence of gene coding regions related to the HPG axis. Amino acid sequence analysis ( Figures 2 , 3 ) found that the Gnrh2 amino acid sequences of 2nCC and 3nCC both contain highly conserved biologically active decapeptides and proteolytic sites; both can be found in the Fshβ genes of two different ploidy carp. There are 12 cysteine residue sites and a highly conserved glycosylation site, “NIS”; in their Lhβ genes, there are 12 cysteine residue sites and a highly conserved glycosylation site, “NET.”

Quantitative Real-time PCR for the breeding season (April) 2nCC and 3nCC Gnrh2, Gthβ (Fshβ, Lhβ) and Gthr (Fshr, Lhr) genes were used to detect the mRNA transcription levels, and the results obtained are as shown in Figure 4 . Epigenetic expression levels of the Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes in both female and male 3nCC were significantly higher than those of the corresponding genes in 2nCC during the same period. The p-value for the expression level was P<0.05 ( Figure 4 ).

To study the relationship between the expression of Gnrh2, Fshβ, Lhβ, Fshr, and Lhr genes in Carassius auratus with different ploidy and methylation of promoter regions, methylation levels of some promoters of the above genes were detected, and the results are shown in Figure 5 . During the breeding season, methylation levels in the Gnrh2, Fshβ, and Fshr promoter regions were significantly reduced in 3nCC compared to 2nCC (P<0.05). After triploidization, the methylation levels of the Gnrh2 promoter region were 0.489 and 0.6 in 3nCC females and males, respectively, and 0.822 and 0.789 in 2nCC females and males, respectively.

Methylation levels of Fshβ gene promoter region, the methylation levels of the promoter region of the Fshr gene were 0.5 and 0.4 in 3nCC females and males, respectively, and 0.95 and 0.883 in 2nCC females and males, respectively. After triploidization, the methylation levels of the promoter regions of the Lhβ and Lhr genes were decreased (but not significantly different) in 3nCC. The methylation levels of the Lhβ promoter region were 0 and 0.04 in 3nCC females and males, respectively, 0.06 and 0.22 in 2nCC females and males in the expression and localization analysis of HPG axis genes in Carassius auratus with different ploidy. Methylation levels in the Lhr promoter region were 0.013 and 0.014 in 3nCC females and males, respectively, and 0.05 and 0.05 in 2nCC females and males, respectively. A correlation analysis was performed between the HPG axis gene expression levels and the promoter methylation levels of the two ploidy C. auratus during the breeding season ( Figure 6 ). The two were approximately negatively correlated (both females and males).

BioEdit software was used to analyze the sequences of CpG-enriched regions of partial promoters of HPG axis-related genes in 2nCC and 3nCC. The results are shown in Figure 7 . Among them, the number of CpG sites in the Gnrh2 promoter region ( Figure 7A ), Fshβ ( Figure 7B ), Lhβ ( Figure 7C ), Fshr ( Figure 7D ), and Lhr ( Figure 7E ) genes were 9, 4, 5, 6, and 8, respectively.

Figure 8 shows that both 2nCC and 3nCC have normal pituitary structures, and there are purple-red or blue-purple stained GTH cells in the pituitary region of both fish.

Figures 8A1, B1 show that the ovary structure of the two fish developed as expected, and the development levels were essentially synchronized. Many phase III and IV oocytes were found in the ovaries, and some phase II oocytes were found in the ovaries. Figures 8C1, D1 show that the testis structure of 2nCC and 3nCC also developed as expected, and several spermatocytes and sperm cells appeared in the seminiferous tubules.

The localization of Fshβ and Lshβ genes in the pituitary (Figures 9 A1, B1, C1, D1) tissues of diploid and triploid wild crucian carp was detected by immunofluorescence technique. In the GTH cells of the adenohypophysis of the diploid C. auratus, green fluorescence was observed in the area marked by the white arrow; a positive hybridization signal appeared (Figures 9 A2, B2, C2, D2). In contrast, no green fluorescence was observed in the negative control group (Figures 9 A4, B4, C4, D4); no positive hybridization signals appeared. The immunofluorescence section of the Fshβ gene showed that the expression site of the Fshβ gene was approximately the same as that of the Lhβ gene and located in GTH adenohypophysis cells of the two fish. It can be observed that the area indicated by the red arrow in the figure has green fluorescence, and a positive hybridization signal appears ( Figures 9A3, B3, C3, D3 ). No green fluorescence was seen in the control group. There was no positive hybridization signal.

Green fluorescence was observed in Sertoli cells of C. auratus diploid and follicular and granulosa cells of oocytes that developed in stages III and IV ( Figures 8A2, B2, C2, D2 , the area indicated by the white arrow), while no green fluorescence was found in the negative control group; no positive hybridization signals ( Figures 8A3, B3, C3, D3 ).