Knockout of sws2a and sws2b in Medaka (Oryzias latipes) Reveals Their Roles in Regulating Vision-Guided Behavior and Eye Development

The medaka (Oryzias latipes) is an excellent vertebrate model for studying the development of the retina. Its genome database is complete, and the number of opsin genes is relatively small compared to zebrafish. Short wavelength sensitive 2 (sws2), a G-protein-coupled receptor expressed in the retina, has been lost in mammals, but its role in eye development in fish is still poorly understood. In this study, we established a sws2a and sws2b knockout medaka model by CRISPR/Cas9 technology. We discovered that medaka sws2a and sws2b are mainly expressed in the eyes and may be regulated by growth differentiation factor 6a (gdf6a). Compared with the WT, sws2a−/− and sws2b−/− mutant larvae displayed an increase in swimming speed during the changes from light to dark. We also observed that sws2a−/− and sws2b−/− larvae both swam faster than WT in the first 10 s of the 2 min light period. The enhanced vision-guided behavior in sws2a−/− and sws2b−/− medaka larvae may be related to the upregulation of phototransduction-related genes. Additionally, we also found that sws2b affects the expression of eye development genes, while sws2a is unaffected. Together, these findings indicate that sws2a and sws2b knockouts increase vision-guided behavior and phototransduction, but on the other hand, sws2b plays an important role in regulating eye development genes. This study provides data for further understanding of the role of sws2a and sws2b in medaka retina development.


Introduction
Visual pigments in photoreceptor cells are very important in light detection and signal transduction [1,2]. Visual pigments consist of opsin receptor proteins that are covalently bound to vitamin A-derived photochromophores, and opsins mediate light absorption in cones and rods [3]. In vertebrates, rod cells responsible for dark or low-light vision express only rhodopsin (rh1), while cone cells for bright or well-lit vision fall into four main categories: short-(sws1, sws2), medium-(rh2), and long-(lws) wavelength-sensitive opsins [4]. Several studies have shown that opsins play a crucial part in feeding, behavior, and color sensitivity in fish [5][6][7][8]. In the process of fish evolution, some copies of the opsin gene may be lost, but all opsins are basically present, while sws2 and rh2 are lost in most mammals [9,10]. Therefore, the function of sws2 and rh2 in fish is worth studying. RH2 mainly performs its visual function in fish that are active at night or dusk [11]. Much less is known about sws2, with only Percomorpha fish having received extensive research [12].
Visual pigment proteins show great diversity under natural selection, and many fish genomes contain one to three copies of sws2 genes [13][14][15]. In rainbow trout (Oncorhynchus mykiss), sws2 is switched from sws1, a process that begins before the yolk sac is fully absorbed and continues throughout the juvenile period [16,17]. Atlantic Cod (Gadus morhua) have lost sws1 and lws opsins in evolution, and only uses sws2 and rh2 opsins to detect 2 of 13 prey, avoid predators, and adapt to light response [18]. Similarly, the threespine stickleback (Gasterosteus aculeatus) adapts its visual perception to blackwater habitats by tuning spectra at sws2 sites [14]. There is a group of weakly electric teleost fishes in South America (Gymnotiforms) that are nocturnal, many living in muddy streams or deep rivers, and which have lost the sws1 and sws2 opsin genes through evolution [19]. In addition, changes in opsin expression patterns and gene expression regulation also affect opsin function.
The expression of opsin varies greatly among different fish species, but each species may change according to ontogeny and changes in the light environment [20]. On the other hand, the regulation of sws2 gene expression is also controversial. Both gdf6a and forkhead box Q2 (foxq2) have been proven to regulate the expression and differentiation of sws2 [21,22]. In addition, some studies have shown that changes in the opsin genes affected by transcriptional regulators or drug exposure can also cause changes in visual guidance behavior [23,24].
The medaka (Oryzias latipes) is an ideal model organism for studying visual development, and its retina is rich in various types of cone photoreceptors that mediate color perception [25,26]. Medaka and zebrafish (Danio rerio) retinas exhibit distinct patterns of photoreceptor cell mosaics, with each blue-sensitive cone (sws2) surrounded by four double cones (lws and rh2) [27]. Despite the complexity of the spatiotemporal pattern of cone opsin expression in fish, recent genome-editing technologies have made the use of mutants in genome studies possible [28,29].
In this study, we generated sws2a and sws2b knockout medaka models using CRISPR/ Cas9. We discovered that medaka sws2a and sws2b are mainly expressed in the eyes and may be regulated by gdf6a. The enhanced vision-guided behavior in sws2a −/− and sws2b −/− medaka larvae may be related to the enhanced photoconductive signaling. Additionally, we also found that sws2b affects the expression of eye development genes, while sws2a does not. These findings indicate that sws2a and sws2b knockouts increase vision-guided behavior and upregulated phototransduction-related genes, but on the other hand, sws2b plays an important role in regulating eye development genes.

Expression of sws2a and sws2b in Medaka
There are two orthologous genes of sws2 in medaka, namely sws2a and sws2b, and sequence alignment showed the amino acid residues of medaka sws2a and sws2b had 77.27% homology ( Figure 1A). In addition, the expression of sws2a and sws2b mRNA in different tissues was detected by RT-PCR. Both sws2a and sws2b were expressed mainly in the eyes among the adult tissues ( Figure 1B). The expression levels of sws2a and sws2b were low in larval fish, while sws2b expression was increased in adult fish ( Figure 1C).

Establishment of sws2a and sws2b Mutant Medaka
To understand the importance of sws2a and sws2b in visual performance, we generated sws2a and sws2b mutant medaka by CRISPR/Cas9 technology. The CRISPR/Cas9 guide-RNAs close to the translation start codon were selected to change the base sequences of sws2a and sws2b, thereby blocking protein translation. Compared with the WT medaka, sws2a and sws2b homozygous mutants showed 4 bp and 274 bp deletions, respectively, and the deletions both resulted in the early termination of translation of the entire seventransmembrane domain (7tm_1) in SWS2 (Figure 2A

Establishment of sws2a and sws2b Mutant Medaka
To understand the importance of sws2a and sws2b in visual performance, we generated sws2a and sws2b mutant medaka by CRISPR/Cas9 technology. The CRISPR/Cas9 guide-RNAs close to the translation start codon were selected to change the base sequences of sws2a and sws2b, thereby blocking protein translation. Compared with the WT medaka, sws2a and sws2b homozygous mutants showed 4 bp and 274 bp deletions, respectively, and the deletions both resulted in the early termination of translation of the entire seven-transmembrane domain (7tm_1) in SWS2 (Figure 2A-D). Protein structure prediction further elucidated the knockout results ( Figure 2E,F).

Establishment of sws2a and sws2b Mutant Medaka
To understand the importance of sws2a and sws2b in visual performance, we ge ated sws2a and sws2b mutant medaka by CRISPR/Cas9 technology. The CRISPR/C guide-RNAs close to the translation start codon were selected to change the base quences of sws2a and sws2b, thereby blocking protein translation. Compared with the medaka, sws2a and sws2b homozygous mutants showed 4 bp and 274 bp deletions, res tively, and the deletions both resulted in the early termination of translation of the en seven-transmembrane domain (7tm_1) in SWS2 (   In order to determine whether the sws2a −/− and sws2b −/− mRNA was decayed, we analyzed the transcript levels of sws2a and sws2b in WT, sws2a −/− , and sws2b −/− medaka by RT-qPCR. The sws2a −/− and sws2b −/− medaka both had no significant decrease in sws2a and sws2b mRNA levels, respectively, while transcriptional compensation of sws2b deletion was observed sws2a −/− ( Figure 3A,B). Moreover, we designed two pairs of total-length primers to distinguish the expression differences between the two transcripts of medaka sws2 ( Figure S1). Next, we sequenced the total lengths of sws2a and sws2b in sws2a −/− and sws2b −/− medaka, which showed 4 bp and 274 bp deletions in sws2a and sws2b mRNA levels in sws2a −/− and sws2b −/− mutants ( Figure S2). sws2a −/− and sws2b −/− showed no morphological difference from the WT medaka ( Figure S3).

Feeding and the Behavioral Tests
The food intake of Artemia in sws2a −/− medaka larvae was signific that of WT and sws2b −/− ( Figure 4A, p < 0.05), but there was no signifi growth ( Figure 4B, p > 0.05). Because gdf6a and foxq2 were reported to determine the expression of sws2 [21,22], we analyzed the mRNA expression levels of gdf6a and foxq2 in sws2a −/− and sws2b −/− mutants. GDF6A mRNA expression levels were significantly increased in both sws2a −/− and sws2b −/− compared to WT ( Figure 3C, p < 0.05). However, the mRNA expression levels of foxq2 exhibited no significant difference in the fish ( Figure 3D, p > 0.05).

Feeding and the Behavioral Tests
The food intake of Artemia in sws2a −/− medaka larvae was significantly higher than that of WT and sws2b −/− ( Figure 4A, p < 0.05), but there was no significant difference in growth ( Figure 4B, p > 0.05). We next examined the visual responsiveness in sws2a −/− and sws2b −/− mutants at 6 dph. Compared with the WT, sws2a −/− and sws2b −/− mutant larvae displayed an increase in swimming speed during the changes from light to dark ( Figure 4C, p < 0.05). We also observed that sws2a −/− and sws2b −/− larvae both swam faster than WT in the dark and the first 10 s of the 2 min light period ( Figure 4D, p < 0.05). They all had a peak in the first 10 s of the 2 min light period, while the swimming speed was detected to increase by 48.1% and 57.3% in sws2a −/− and sws2b −/− larvae, respectively.

Transcript Levels of Phototransduction-Related Genes in Larvae of sws2a −/− and sws2b −/− Mutants
We examined the mRNA levels of phototransduction-related genes in larvae of the WT, sws2a −/− , and sws2b −/− medaka by RT-qPCR. Surprisingly, the deletion of sws2a or sws2b did not affect the expression of other cone and rod genes ( Figure 5A, p > 0.05). We further measured the mRNA levels of cone phototransduction genes in WT, sws2a −/− , and sws2b −/− We next examined the visual responsiveness in sws2a −/− and sws2b −/− mutants at 6 dph. Compared with the WT, sws2a −/− and sws2b −/− mutant larvae displayed an increase in swimming speed during the changes from light to dark ( Figure 4C, p < 0.05). We also observed that sws2a −/− and sws2b −/− larvae both swam faster than WT in the dark and the first 10 s of the 2 min light period ( Figure 4D, p < 0.05). They all had a peak in the first 10 s of the 2 min light period, while the swimming speed was detected to increase by 48.1% and 57.3% in sws2a −/− and sws2b −/− larvae, respectively.

Transcript Levels of Eye Development Gene in Larval sws2a and sws2b Knockout Medaka
We next tested whether sws2a and sws2b knockout affected the expression of the eye development genes in the larvae. Compared to WT, sws2a −/− larvae displayed an increased mRNA expression pattern of SIX homeobox 7 (six7) ( Figure 6A, p < 0.05), while no alteration was observed in the sws2b −/− . In addition, the transcriptional expression levels of paired box 6 (pax6), SIX homeobox 3a (six3a), and six3b were significantly decreased in the sws2b −/− larvae ( Figure 6B-D, p < 0.05).

Discussion
In contrast to most mammals that have lost sws2 and rh2 genes over e time, all four opsin gene types are present in whole-genome duplication in although there is a phenomenon of gene copy loss [30,31]. SWS2A and sws2b oped by the tandem duplication of sws2 in the common ancestor of Neotele [32]. For instance, despite having highly conserved foxl2a and foxl2b domains i these two genes have divergent functions and synergies. Disruption of foxl2 led to premature ovarian failure and partial sex reversal, respectively, and foxl2 jointly regulate the development and maintenance of zebrafish ovaries [33]. Th domain of sws2a and sws2b proteins in medaka was also highly conserved, and similarity between the two proteins reached 77.27%. Medaka sws2a and sws2b expressed in the eye during adulthood, while sws2b expression is significantly adult fish relative to larva. This is consistent with the expression pattern o sws2b in bluefin killifish (Lucania goodei) [34].
In this study, we produced and characterized sws2a and sws2b medaka m with 4 bp and 274 bp deletions that caused the loss of seven transmembrane SWS2A and SWS2B. We detected that the transcription levels of sws2a and sws affected by the mutation. Although we did not find specific antibodies for med or SWS2B that confirm them at the protein level, we further amplified the tota the gene using cDNA as a template, and sequenced it. The results showed tha sws2b mutants had 4 bp and 274 bp deletions of sws2a and sws2b mRNA, r These results indicated that the knockouts of sws2a and sws2b in this study a Our results are in accordance with the recent work on lca5 knockout zebrafish gene and protein expressions are regulated by upstream signals [36]. GDF6A i of the bone morphogenetic protein family that induces dorsal retinal differen ing ocular morphogenesis [37]. GDF6A deletion zebrafish larvae displayed

Discussion
In contrast to most mammals that have lost sws2 and rh2 genes over evolutionary time, all four opsin gene types are present in whole-genome duplication in teleost fish, although there is a phenomenon of gene copy loss [30,31]. SWS2A and sws2b were developed by the tandem duplication of sws2 in the common ancestor of Neoteleostei fishes [32]. For instance, despite having highly conserved foxl2a and foxl2b domains in zebrafish, these two genes have divergent functions and synergies. Disruption of foxl2a and foxl2b led to premature ovarian failure and partial sex reversal, respectively, and foxl2a and foxl2b jointly regulate the development and maintenance of zebrafish ovaries [33]. The structural domain of sws2a and sws2b proteins in medaka was also highly conserved, and the protein similarity between the two proteins reached 77.27%. Medaka sws2a and sws2b are mainly expressed in the eye during adulthood, while sws2b expression is significantly elevated in adult fish relative to larva. This is consistent with the expression pattern of sws2a and sws2b in bluefin killifish (Lucania goodei) [34].
In this study, we produced and characterized sws2a and sws2b medaka mutant lines with 4 bp and 274 bp deletions that caused the loss of seven transmembrane domains in SWS2A and SWS2B. We detected that the transcription levels of sws2a and sws2b were not affected by the mutation. Although we did not find specific antibodies for medaka SWS2A or SWS2B that confirm them at the protein level, we further amplified the total lengths of the gene using cDNA as a template, and sequenced it. The results showed that sws2a and sws2b mutants had 4 bp and 274 bp deletions of sws2a and sws2b mRNA, respectively. These results indicated that the knockouts of sws2a and sws2b in this study are effective. Our results are in accordance with the recent work on lca5 knockout zebrafish [35]. SWS2 gene and protein expressions are regulated by upstream signals [36]. GDF6A is a member of the bone morphogenetic protein family that induces dorsal retinal differentiation during ocular morphogenesis [37]. GDF6A deletion zebrafish larvae displayed fewer blue cone photoreceptor cells, and gdf6a could develop and maintain the sws2 [21]. Disruption of foxq2 in zebrafish showed the loss of sws2 cone expression in 5-day post-fertilization (dpf) larvae, and further results indicated that foxq2 was an activator of sws2 transcription and inhibited sws1 expression [22]. In addition, the thyroid hormone (TH) may induce the transition of sws1 to sws2 opsin by binding to its receptor thrβ2 [38]. Growth hormone (GH) can accelerate the process of sws1 to sws2 opsin conversion [39]. Here, we observed that gdf6a expression was significantly upregulated in sws2a −/− and sws2b −/− mutants, while foxq2 expression was unaffected. We speculated that sws2a and sws2b are mainly regulated by gdf6a in medaka, mainly because when sws2a or sws2b was knocked out, its upstream gene gdf6a expression increased through negative feedback regulation. Similar to foxl2 and sox9, foxl2 is downstream of sox9, and sox9 transcript level significantly increased in foxl2a −/− and foxl2b −/− zebrafish mutants [33,40].
Feeding assessment showed increased food intake in sws2a −/− mutant zebrafish, while sws2b −/− was unaffected. In a previous report, tbx2b zebrafish mutants (reduced UV (SWS1) cones) showed declined foraging performance [41]. Another study demonstrated that sws2 and rh2 opsins decreased, and visual prey capture was impaired in six6a/six6b/six7 triple-knockout zebrafish at 6 dpf [23]. It has also been proposed that in the case of sws1 regulation, the increase in sws2 may be related to prey search [5,42]. Therefore, we think that the increased intake of sws2a −/− may be related to the deletion of sws2a. Interestingly, the feeding rate and survival rate of larvae of haddock (Mellanogrammus aeglefinus) increased under blue and green light compared with other light colors [43,44]. Light/dark motion tests and responses to light stimuli are effective methods to reflect the integrated function of visual pathways [45][46][47]. In this study, compared with the control group, sws2a −/− and sws2b −/− mutants both significantly increased swimming speed when they changed from light to darkness. When the larvae were suddenly stimulated by light, sws2a −/− and sws2b −/− mutant larvae both reacted strongly within the first 10 s. The swimming speed increased by 48.1% and 57.3% in sws2a −/− and sws2b −/− larvae, respectively. These results are consistent with our finding that cone phototransduction pathway-related genes are further activated in sws2a −/− and sws2b −/− larvae. GRK7A is mainly expressed in the outer cone segment; grk7 knockdown in larval zebrafish caused a delay in dark adaptation and impaired cone response recovery [48]. Moreover, grk7a has been shown to be involved in the recovery of cone light response in larval zebrafish [49]. Additionally, GNB3 is a G-protein beta subunit located in the outer segment of cones and mainly plays a role in the phototransduction cascade of cones [50]. We concluded that the enhanced vision-guided behavior in sws2a −/− and sws2b −/− medaka larvae may be the result of upregulated phototransduction genes.
In vertebrates, opsins are G-protein-coupled receptors expressed in the retina, responsible for promoting eye sensitivity to light. Each cone opsin covers a distinct part of the visible spectrum, with corresponding non-absorption maxima [51]. Our RT-PCR analysis indicated that the loss of sws2a and sws2b did not affect the normal expression of other opsins. In addition to being absorbed by sws2, blue light may also be absorbed by neighboring sws1 and rh2. Thus, the sensitivity of blue light is not controlled by sws2 [52]. Fish such as the southern catfish (silurus meridionalis), which are mainly nocturnal and live in underground or even burrowing freshwater environments, lost sws2 [53,54]. Some recently studied cartilaginous fishes also show no expression of sws2, further supporting the hypothesis that sws2 was lost early in the cartilaginous lineage [55][56][57]. However, sws2 plays an irreplaceable role in some species [14]. The Japanese flounder (Paralichthys olivaceus) adjusted their sws2a and rh2 genes as their light environment changed during development [58]. Therefore, diverse visual systems are adaptive responses to varying environments in different species.
In this study, there was no significant retinal histological damage observed during the retinal histological examination, indicating that its effect may be at the molecular level. SIX3, six6, and six7 are important regulators of fish retinal development and differentiation [23,59]. Knock-down six3b and six7 resulted in the loss of eyes, whereas disruption of six3a and six7 had no significant eye phenotypes [60]. A previous study demonstrated that six6 deletion mice showed severe retinal abnormalities [61]. PAX6 is necessary for the normal development of fish eyes [62]. Deletion of a single copy of pax6 in mice showed microphthalmia, while mutation with a double copy showed anophthalmia [63]. In the present study, we found that six3a, six3b, and pax6b were downregulated in sws2b −/− medaka larvae, suggesting that the loss of sws2b may affect the retinal development of medaka larvae. In addition, the upregulated expression of six7 in sws2a −/− medaka larvae might be due to the regulation of sws2a by six7 [64]. Paradoxically, sws2b −/− medaka larvae showed increased motor capacity, but decreased transcription levels of eye development and regulatory genes. Our explanation is that sws2b may delay the movement of medaka larvae, but on the other hand, sws2b plays an important role in regulating eye development genes. The mechanism of medaka sws2a and sws2b double knockout in regulating eye development needs to be determined in future studies.
In summary, this study established sws2a and sws2b knockout medaka by CRISPR/Cas9 technology. We speculated that the enhanced vision sensitivity in regulating vision-guided behavior in sws2a and sws2b knockout medaka larvae might be related to the upregulation of phototransduction-related genes. Additionally, sws2b affects eye development gene expression, implying different mechanisms of sws2a and sws2b. This study provides data for further understanding of the role of sws2a and sws2b in medaka retina development.

Medaka Lines and Maintenance
The wild-type (WT) medaka are an orange strain, and were maintained in an environment of 26~28 • C and 14 light/10 h dark cycle. Medaka embryos were cultured at 27~28 • C in medaka embryo medium (MEM) [65]. The 6 dph (day post-hatching) larvae were fed with live Artemia twice daily when the yolk sac was almost completely consumed. All the fish were anesthetized with tricaine methanesulfonate (MS-222) before the tissue collection.

Generating sws2a −/− and sws2b −/− Mutants by CRISPR/Cas9 Technology
Medaka sws2a (ENSORLG00000040205) and sws2b (ENSORLG00000028370) genes were targeted using CRISPR/Cas9 technology. The sequencing of single-guide RNAs (sgR-NAs) and PCR primers are shown in Supplementary Table S1. sgRNAs were cloned into pMD-18T vector and were synthesized using TranscriptAid T7 High Yield Transcription kit (ThermoFisher Scientific, Waltham, MA, USA). The compounds of sgRNAs (50 ng/µL) and Cas9 protein (New England Biolabs, Ipswich, MA, USA) were co-injected into one-or two-cell stage wild-type embryos. The F0 medaka were raised to adulthood and outcrossed with wild-type to produce F1 medaka. A T7 endonuclease 1 assay (Vazyme, Nanjing, China) was used to detect sws2a heterozygous mutant individuals according to the manufacturer's instructions. The heterozygous individuals of sws2b mutation with large fragment deletion were distinguished by PCR detection and then sequenced. The F1 heterozygous individuals in-crossed to generate F2 homozygous individuals, and all experiments were conducted with F3 homozygous individuals. Unless otherwise mentioned, the homozygous mutant lines of sws2a and sws2b in subsequent experiments were addressed as sws2a −/− and sws2b −/− , respectively. The SWISS_MODEL (https://swissmodel.expasy.org/interactive) (1 March 2023) was used to predict the protein tertiary structures. The protein tertiary structures of sws2a and sws2b were predicted using the SWISS_MODEL (https://swissmodel.expasy.org/).

Larvae Feeding Assays
For larvae food intake, 6 dph larvae were fed with Artemia in wells of a 6-well plate with 8 mL MEM (diameter 3.48 cm wells) for 30 min (6 larvae per 6-well plate). The density of the Artemia is adjusted to 150 Artemia per ml. Then, larvae were anesthetized with MS-222 (Argent Chemical Laboratories, Redmond, WA, USA) and fixed with 4% paraformaldehyde (PFA) (Servicebio, Wuhan, China) overnight. Photographs of the orange area of Artemia in the digestive tract were taken by a stereomicroscope and measured with Image J1 software. The amount of food ingested by medaka larvae was developed by the procedure described previously [66].

Growth Performance and Survival Rate
For growth performance and survival rate, 6 dph larvae were fed with abundant Artemia twice daily for 10 days. Twenty WT, sws2a −/− , and sws2b −/− medaka larvae were randomly selected, anesthetized, and fixed with 4% PFA for total length measurement. The experiment was repeated 3 times.

Behavioral Tests
Behavioral tests were conducted between 15:00 and 17:00 using the DanioVision Observation Chamber (Noldus Information Technology, Wageningen, The Netherlands) linked to the EthoVision XT13 software. The 6 dph larvae were plated onto 24-well plates (diameter 15.6 mm wells) with 1 mL MEM (individual larvae per 24-well plate). Further analysis was performed using custom Open Office Org 2.4 software.
Light response: The larvae were acclimated for 30 min in the dark at 28 • C and then tracked the movement of larvae for 4 min, with 2 min of the dark period and 10 s of the light stimulation period [46]. The average swimming speed (cm/s) for dark and light were collected every 2 min and 10 s, respectively.
Light/dark behavior analysis: The larvae were acclimated for 10 min at 28 • C, and the larval locomotor activity was tested in response to dark-light conversion (3 min light/3 min dark/3 min light/3 min dark) based on the protocol by Huang et al. [67], with modifications to the transition stimulation time. The average swimming speed (cm/s) for each individual larva was collected every 60 s.

RNA Isolation and Quantitative RT-PCR
All fish were sampled in the light phase of the light/dark cycle. The adult fish tissues (n = 3) and the two larval eyes of the 6 dph medaka (n = 6) were collected and frozen in liquid nitrogen. An equal amount of RNA was extracted from each sample according to the RNAiso instruction, and the cDNA was reversed transcribed by the HiScript ® III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). The reaction system (20 µL) contained 1 µL cDNA template, 10 µL SYBR (Vazyme, China), 0.4 µL of each primer, and 8.2 µL ddH 2 O. The cycling parameters were 95 • C for 30 s, 40 cycles at 95 • C for 10 s, 58 • C for 30 s, and melting curve from 65 • C to 95 • C (gradually increasing 0.5 • C s −1 ), with data acquired every 6 s. The results were normalized to β-actin, and relative transcript abundances of genes were performed using the 2 −∆∆Ct value method [68]. All primers are shown in Supplementary Table S2.

Statistical Analysis
All results are presented as means ± S.E.M (standard error of the mean), and the normality of the data was first tested by the Shapiro-Wilk test. The differences among three groups were analyzed by one-way ANOVA and Duncan's multiple-range test, and (p < 0.05) was considered a significant difference. The differences between the two groups were determined with Student's t-test, and statistical significance was determined at p < 0.05.

Supplementary Materials:
The following supporting information can be downloaded at https://www. mdpi.com/article/10.3390/ijms24108786/s1. Author Contributions: K.L. designed and performed the experiment, analyzed experimental data, and wrote the original draft. J.W., S.T. and X.J. raised fish and participated in some of the experiments. X.-F.L. designed and supervised the experiment, and wrote and edited the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This study was supported by the Development Project of Hubei Province (2020BBA035) and the National Natural Science Foundation of China (31972809).
Institutional Review Board Statement: All experimental protocols were approved by the Institutional Animal Care and Use Ethics Committee of Huazhong Agricultural University (an approval reference number HZAUFI-2020-0024).

Informed Consent Statement: Not applicable.
Data Availability Statement: All data are available from the corresponding author by request.