In this study, Nile tilapia (Oreochromis niloticus) were acclimatized in recirculating aerated freshwater tanks at 26 °C under a natural photoperiod before use. All animal experiments were conducted per the regulations of the Guide for Care and Use of Laboratory Animals approved by the Committee of Laboratory Animal Experimentation of Southwest University, China.

The recombinant construct of Cyp17a2 was prepared by cloning the 1566 bp cyp17a2 ORF sequences into a pCold I expression vector. The recombinant plasmid with His-tag at its N-terminal was transformed into Escherichia coli, and expressed with the induction of isopropyl β-D-l-thiogalactopyanoside (IPTG, 1 mmol/L). The His-Cyp17a2 recombinant proteins (25-30 µg) was purified with a Ni-NTA super flow cartridge (Qiagen, Hilden, Germany) and used as an antigen to immunize female rabbits three times at 15-day intervals. Ten days after the last immunization, rabbit serum was collected and purified by affinity chromatography on Sepharose 4B Fast Flow resin (Sigma, Darmstadt, Germany) combined with the Cyp17a2 recombinant protein. To confirm the specificity of the polyclonal antibody, total proteins extracted from testis, ovary, head-kidney, brain, and muscle from 180 dah tilapia were separated using 12.5% SDS-PAGE under reducing conditions. Western blotting was performed according to the methods described previously (29) and the antibody against tilapia Cyp17a2 was diluted at 1:1000.

To determine the cellular localization of Cyp17a2 in the testis and head-kidney of tilapia, the testes and head-kidneys from wild-type males at different development stages were dissected and fixed in Bouin’s solution for 12 hours at room temperature, then dehydrated and embedded in paraffin. The tissue blocks were then sliced into 5-μm-thickness sections, then the sections were treated in a blocking solution (5% BSA diluted in 1x PBS) (Sangon Biotech, China), incubated with the primary antibody against Cyp17a2 overnight at 4°C and rinsed with 1x PBS five times for 5 min per wash. In this study, the anti-Cyp17a2 polyclonal antibody was diluted at 1:1000 before use. As a negative control, the primary antibody was replaced with normal rabbit serum. The slides were then incubated with anti-rabbit immunoglobulin G (diluted at 1:1000) at room temperature for 1 h, and then rinsed with 1x PBS three times for 5 min per wash. Immunoreactive signals were visualized using diaminobenzidine tetrachloride (DAB) (Sigma, Germany) as the substrate. Sections were then counterstained with hematoxylin. Finally, all the images for these sections were acquired with an Olympus BX5 light microscope (Olympus, Japan).

A homozygous mutation line of cyp17a2 was constructed in tilapia to elucidate the functions of Cyp17a2 in fish steroidogenesis, spermatogenesis, and male fertility by using CRISPR/Cas9 technology. The guide RNA (gRNA) target site was selected from sequences corresponding to GGN18NGG on the sense strand of DNA (http://zifit.partners.org/ZiFiT/). The synthesis of gRNA and Cas9 mRNA, and the screening and establishment of homozygous mutation line were carried out according to previous report (25).

Testes at different developmental stages (75, 90 and 180 dah) and head-kidneys (180 dah) from wild-type and cyp17a2 ^-/- fish were dissected and embedded in paraffin. IHC analysis was conducted as described above. Anti-Vasa (germ cells marker, diluted at 1:1000), -PCNA (Proliferating cell nuclear antigen, diluted at 1:500, Cusabio, Wuhan, China), -StAR1 (Steroidogenic acute regulatory protein 1, diluted at 1:500), -Cyp17a1 (cytochrome P450 family 17 subfamily a 1, Leydig cell marker, diluted at 1:1000), -Cyp11b2 (Cytochrome P450, family 11, subfamily B, polypeptide 2, Leydig cell marker, diluted at 1:1000), and -Cyp17a2 polyclonal antibodies were used to assess the impacts of cyp17a2 mutation on spermatogenesis and head-kidney development.

Gonads and head-kidneys were collected from wild-type and cyp17a2 ^-/- XY fish (n≥3 fish/genotype). Total RNA was isolated for each sample using RNAiso Plus (Takara, Dalian, China), and reverse transcribed using the Prime Script RT Master Mix Perfect Real Time Kit according to the manufacturer’s instructions (Takara, Dalian, China). All qRT-PCRs were carried out in an ABI-7500 fast Real-time PCR machine (Applied Biosystems, USA), and all experiments were performed according to the manufacturer’s instructions. The relative abundance of target genes was normalized to β-actin using R=2^-ΔΔCt formula (30). The primer sequences used for the PCR reactions were listed in Table 1 .

Western blotting was performed to detect the expression of steroidogenic enzymes in the testes of cyp17a2 ^-/- XY tilapia. Total proteins were extracted from the testes of cyp17a2 ^+/+ XY (n=3) and cyp17a2 ^-/- XY (n=3) fish at 180 dah, diluted to a final concentration of 20 mg/mL and stored at -20 °C until use. Western blotting was performed as described previously (29). The primary antibodies against steroidogenic enzymes (anti-Cyp17a1 and -Cyp11b2) diluted at 1:1000. α-tubulin was used as a reference protein and its abundance was used to normalize protein sample loading. Densitometry was quantified and normalized using α-tubulin as a reference protein using the Fusion-CAPT software.

The cell count of spermatogenic cells at various stages (i.e. spermatogonia, primary spermatocyte and secondary spermatocyte) were calculated according to the methods described previously (31). Briefly, ten sections were randomly selected from each cyp17a2 ^+/+ (n=3) and cyp17a2 ^-/- (n=3) XY fish at 90 dah, and the sections were stained with conventional hematoxylin and eosin (H&E) staining. Then the image of each section at 40x magnification was acquired with an Olympus BX5 light microscope (Olympus, Japan) for germ cell counting.

Blood samples were collected from the caudal veins of cyp17a2 ^+/+ XY (n≥3) and cyp17a2 ^-/- XY (n≥3) fish after anesthesia (250 mg/L, MS-222, Sigma, St Louis, USA) at 180 dah and kept at room temperature for 1 h. After centrifugation at 3000 rpm for 10 min, the supernatant was transferred to a clean centrifuge tube. The supernatant was then centrifuged at 12000 rpm for 10 min at 4 °C and frozen at -80 °C until use. Serum steroid levels were measured using the EIA (enzyme-linked immunosorbent assay) kits (Cayman, Michigan, USA) or detected by MetWare (http://www.metware.cn/) based on the AB Sciex QTRAP^® 6500 LC-MS/MS system (AB Sciex, Framingham, USA). The steroid information was listed in Table 2 .

To detect the impacts of cyp17a2 gene mutation on spermatogenesis and fertility of male tilapia, changes of sperm morphology, sperm motility, and fertility were compared between cyp17a2 ^-/- and cyp17a2 ^+/+ XY fish at 300 dah. Papanicolaou staining was used to detect the morphology of sperms according to the method which were used in previous reports (25, 31). Sperm motility, VCL (curvilinear velocity), VSL (straight line velocity) and sperm quality were examined via computer-assisted sperm analysis using the Sperm Quality Analyzer software according to the manufacturer’s instructions (Zoneking Software, China). The fertility of cyp17a2 ^+/+ and cyp17a2 ^-/- XY fish was assessed via artificial insemination. Eggs from wild-type XX female fish (n=3) were divided into six groups (approximately 300 eggs/group) and artificial insemination was performed using sperm obtained from the cyp17a2 ^+/+ and cyp17a2 ^-/- XY fish. Moreover, the survival rate of embryos was calculated from 1-9 days after fertilization.

All the data for qRT-PCR, Western blotting, cell count, LC-MS/MS and EIA assay are presented as the mean ± SD. And Student’s t-test was performed to determine the difference between the two groups. P<0.05 was considered statistically significant differences.

Total proteins extracted from gonads (ovary and testis), head-kidney, brain, and muscle were used to check the specificity of the Cyp17a2 polyclonal antibody by Western blotting. Western blotting showed that a single and specific band could be recognized in testis, ovary, and head-kidney, but not in brain and muscle by using tilapia Cyp17a2 antibody ( Figure 1A ). By IHC, positive Cyp17a2 immunostaining signals were observed in the Leydig cells of the testes in XY fish at 90, 180 and 210 dah ( Figure 1B ). In the head-kidney, Cyp17a2 was predominantly expressed in interrenal cells from 5 to 180 dah ( Figure 1C ).

To investigate the potential roles of Cyp17a2 in fish steroidogenesis, a cyp17a2 homozygous mutation line was produced using CRISPR/Cas9 technology by targeting the first exon ( Figure 2A ). A frame-shift mutation with 11-bp (TTGGAGGAGGC) deletion in cyp17a2 homozygous mutant was obtained by Sanger sequencing ( Figure 2B ). A heteroduplex motility assay showed that heterozygous cyp17a2 ^+/− individual was identified as possessing both heteroduplex and homoduplex amplicons, while the cyp17a2 ^-/- and cyp17a2 ^+/+ individuals were found to be single band profile with only homoduplex amplicons ( Figure 2C ). The comparison of the predicted amino acid sequences demonstrated that the 11-bp deletion in the cyp17a2 gene resulted in the production of truncated protein due to frame shift and the occurrence of a premature stop codon ( Figure 2D ). IHC and Western blotting analyses also demonstrated that Cyp17a2 protein completely vanished in the testes of cyp17a2 ^-/- XY fish ( Figures 2E, F ).

IHC results demonstrated that the expression of Cyp17a2 could not be detected in the head-kidney of cyp17a2 ^-/- fish ( Figure 3A ). The results of qRT-PCR showed that the expression of StAR1, cyp21a2 and cyp11b2 in the head-kidney was significantly increased in cyp17a2 ^-/- fish compared with wild-type fish ( Figure 3B ). EIA assay showed that serum cortisol level of cyp17a2 ^-/- XY fish was significantly reduced ( Figure 3C ).

EIA assay showed that cyp17a2 deficiency led to a significant decline in the biosynthesis of DHP, while no evident change of 11-KT production was detected between cyp17a2 ^-/- and wild-type XY fish ( Figures 4D, E ). Moreover, IHC analysis demonstrated that the positive signals of steroidogenic enzymes (Cyp17a1 and Cyp11b2) concerning androgen production were detected both in the testes of cyp17a2 ^-/- and wild-type XY fish ( Figure 4A ). Further analyses by qRT-PCR and Western blotting demonstrated that no significant difference in the expression level of androgen synthases (cyp17a1/Cyp17a1, cyp11b2/Cyp11b2) was found between cyp17a2 ^-/- XY fish and wild-type XY fish at 90 and 180 dah ( Figures 4B, C ).

Consistently, the results of LC-MS/MS showed that cyp17a2 gene mutation led to significant changes in C21-steroid levels in cyp17a2 ^-/- XY fish compared with wild-type XY fish. The syntheses of 17α-hydroxyprogesterone, cortisol and cortisone were significantly reduced in cyp17a2 ^-/- XY fish, while cholesterol, pregnenolone, 17α-hydroxypregnenolone, progesterone, pregnanediol, deoxycorticosterone, and corticosterone were excessive accumulated ( Figure 5 ). However, compared with wild-type XY fish, the production levels of C19-steroids (testosterone, androstenedione) and C18-steroid (17β-estradiol) were not affected ( Figure 5 ).

To detect the role of Cyp17a2 in the initiation of meiosis in the development of tilapia testes, the gonad phenotype was evaluated at 75 and 90 days, respectively. This period was proved to be the key period for the initiation of meiosis and the morphological differentiation of gonads in wild-type male tilapia. H&E staining reflected that only a few spermatogonia and few spermatocytes were found in cyp17a2 ^-/- XY fish compared with cyp17a2 ^+/+ XY fish at 75 and 90 dah ( Figure 6A ). Immunostaining showed that the number of Vasa (germ cell) and PCNA (proliferative cell) positive cells was significantly reduced in cyp17a2 ^-/- XY fish than that in wild-type XY fish at 75 and 90 dah ( Figures 6B, C ). And the relative expression level of vasa and piwil was significantly decreased in cyp17a2 ^-/- XY fish at 90 dah ( Figure 6D ). Consistently, a significant reduction of spermatogonia, primary spermatocytes, and secondary spermatocytes was detected in cyp17a2 ^-/- XY fish at 90 dah ( Figure 6E ). In addition, qRT-PCR demonstrated that compared with cyp17a2 ^+/+ XY fish, the expression of spo11, sycp3 (meiotic marker genes) were decreased while cyp26b1 (retinoic acid hydrolase) was increased significantly in cyp17a2 ^-/- XY fish at 90 dah ( Figures 6F-H ).

Surprisingly, spermatogenesis was restored at 180 dah, as indicated by the existence of all stages of spermatogenic cells and normal proliferation ( Figures 7A, B ). Papanicolaou staining showed that sperm from cyp17a2 ^-/- XY fish showed normal sperm morphology ( Figure 7C ). However, low sperm motility was indicated by the trajectory and sperm tracking parameters (VCL and VSL) ( Figures 7D, E ). Quantification of the three sperm grades showed that dramatic attenuation of PR (progressive sperm) and NP (non-progressive sperm), while a significant increase in the frequency of IM (immotile sperm) was found in cyp17a2 ^-/- XY fish compared to their control counterparts ( Figure 7F ). The fertilization rate and survival rate (D8 and D9) of cyp17a2 ^-/- XY fish were significantly lower than those of cyp17a2 ^+/+ XY fish ( Figure 7G ).