We selected exon 8 to exon 11 as target sites to generate the Thada founder mice by the CRISPR-Cas9 technology. We used the heterozygous mutants mice (Thada^+/− ) intercrossed, and successfully obtained homozygous mice (Thada^-/- ). Littlermate wild type mice (WT) were included as controls. Genotyping of Thada knockout mice was done by PCR on tail genomic DNA.

All mice described were maintained on the C57BL/6J background. All animals were fed in an ad libitum mode, using a commercial pelleted diet (Beijing Keao Xieli Feed Co.). Mice were housed in groups (5 adult mice per cage) under specific pathogen-free conditions in a temperature controlled room (21-22°C) with a 12 h-12 h light/dark cycle. Beginning at Pnd 21, vaginal opening was monitored daily. Mice used for HFHS experiments were fed with normal chew diet until 8 weeks of age, after which they were maintained on HFHS diet (Research Diet 12451) for another 10 weeks. Female mice were sacrificed between 3 and 6 months.

At 8 weeks of age, the estrous cycle of WT and KO mice was identified by vaginal smear once a day for 2 weeks. The percentage of different stages (diestrus, proestrus, estrous, and metestrus) of the estrous cycle was determined and recorded. The fertility measurement of mice started at the age of 8 weeks and lasted for six months. WT and Thada KO female mice were mated with fertile male mice of the same age using a ratio of 2:1 and checked daily for signs of pregnancy. The number of pups per litter and the gender ratio were calculated.

Serum samples were collected from mice in basal resting condition during diestrus. After coagulation, the serum samples were extracted after centrifugation 2500rpm for 10min (4°C) and stored at −80°C until used. Serum testosterone, progesterone, and estradiol were measured with radioimmunoassay (Beijing North Institute of Biological Technology).

Thada knockout and wild type mice were first injected intraperitoneally with 7.5 IU of pregnant mare serum gonadotropin (PMSG) and after 44 h with 7.5 IU of human chorionic gonadotropin (hCG) (Ningbo Sansheng Pharmaceutical Co., Ltd., China). Sixteen hours later, animals were sacrificed by decapitation. The cumulus-oocyte-complexes were extracted from the fallopian tubes and the oocytes were counted after hyaluronidase treatment. The mature oocytes were assessed by the extrusion of the first polar body (PB1).

Ovaries were fixed overnight with 4% PFA, then dehydrated and embedded in paraffin. For hematoxylin and eosin (H&E) staining, the sections were stained according to standard histological protocols. The images were captured by an inverted microscope.

For follicle counting, the ovaries of normal diet mice were serially sectioned (5μm) and stained with H&E. Every fifth section of each ovary was observed under a microscope. Follicles were classified as healthy or atretic by morphological criteria described previously (10). Atretic preantral follicles consisted of the oocyte that had degenerated with either a disorganized granulosa layer with apoptotic cells (more than 5% of the cells showed apoptosis signs) and/or a hypertrophied theca layer surrounded. Atretic antral follicles were usually present as an intact oocyte, whereas the layer of granulosa cells contained more than 5% apoptotic cells, and the theca layer showed signs of hypertrophy. All kinds of atretic follicles were excluded during the count. Follicles were classified as described previously (11). Only primordial, primary, secondary, and antral follicles were counted when the oocyte with nucleus and nucleolus were present in every fifth section. The adjacent sections were examined, and the follicle with the same nucleus was counted only once during the counting of larger follicles. The number of primordial, primary, secondary, and antral follicles was multiplied by five to estimate the total number of follicles within one ovary. Every tenth section of each ovary was observed to determine the number of corpus luteum. In order to prevent double-counting of corpus luteum, adjacent sections were traced and compared.

Genotyping of Thada knockout mice was done by PCR on tail genomic DNA. The primers for PCR genotyping are:

Thada-allele forward primer: 5’-CAGATGGGATACAGAGGCTAGGCG-3’; reverse primer: 5’-AATGACTGGAGATAATTGCTCATGCC-3’;

WT-allele forward primer: 5’-CAGATGGGATACAGAGGCTAGGCG-3’; reverse primer: 5’-CGCACTCTTTCCCACCTGTCTTC-3’.

Ovarian tissues were collected, snap-frozen in liquid nitrogen, and stored at -80°C. Total RNA was extracted with Trizol, 1μg of RNA were reverse transcribed using the PrimeScript RT Reagent Kit (Takara, Shiga, Japan). Real-time PCR was performed using the LightCycler 480 Probes Master kit (Roche Diagnostics, Meylan, France).

The results are presented as the mean ± SEM; each experiment was repeated at least 3 times. The significant differences between groups were calculated by two-tailed unpaired Student’s t-test and Chi-square test. P-value<0.05 was considered significant.

To explore the functional role of Thada in vivo, we constructed a Thada global knockout (Thada^-/- ) mouse model and designed the experiments ( Figures 1A, B ). Thada was undetectable in ovarian tissues of KO mice at the RNA level, and heterozygotes also resulted in a 50% reduction ( Figure 1C ). The genotype distribution of offsprings obtained by mating Thada^+/- conformed to Mendelian inheritance and the male/female ratio was 1.03 ( Figure 1D ). This reduction of THADA levels did not affect intrauterine development or sex selection. In addition to this, the development of KO mice appeared normal compared with WT mice from birth to adulthood. There were no significant differences in the body weight at 3 weeks and 8 weeks between WT and KO mice ( Figure 1E ). The fertility of KO and WT female mice was evaluated. The litter size and the number of males/females per litter were comparable between the two genotypes ( Figure 1F ). These results suggested that Thada with altered expression did not result in growth retardation or adverse pregnancy outcomes.

There was a closed association between THADA and hyperandrogenism in previous studies. Therefore, the development and function of the ovary have become essential aspects of our research. The anogenital distance was the evaluation index of hyperandrogenism, and the distances were similar between KO and WT mice ( Figure 2A ). The vaginal opening was considered a reliable marker for puberty onset. It did not exhibit differences between KO and WT mice ( Figure 2B ), and the time of appearance of the first estrus was also temporally coincided in KO and WT mice. In addition, the ovarian weights of the two groups were approximately the same ( Figure 2C ). These results illustrated that KO mice experienced normal pubertal development through early life.

Impaired ovulation is another common reason for patients to access the clinic. Estrous cycles, follicles development, hormone levels, and ovarian response to gonadotropin stimulation can reflect the function of the ovaries. Compared to WT mice, KO mice showed comparable estrous cyclicity and cycle length, with all phases of the estrous cycle present ( Figure 2D ). In keeping with WT mice, the ovaries of KO mice exhibited appropriate proportions of follicles at different stages ( Figures 2E, F ). Despite circulating testosterone levels were mildly elevated in adult KO mice, there was no statistical difference between the two genotypes ( Figure 3A ). Serum estrogen and progesterone levels of KO mice achieved comparable levels of WT mice, consistent with the results observed in the ovaries ( Figures 3B, C ). At the same time, we quantified the expression of hormone-related genes and detected no meaningful differences ( Figure 3D ). After being stimulated by PMSG and hCG, KO mice did not demonstrate alteration in the number of oocytes retrieved or PB1 extrusion rate ( Figures 3E, H ). The mRNA expression levels of gonadotropin receptors in the ovaries also corroborated the results of the superovulation experiment, with levels of Fshr and Lhcgr not different between the two groups ( Figures 3H, I ). These results suggested that the ovaries of Thada knockout mice may have normal function. Indeed, despite the THADA SNP rs13429458 being involved in PCOS, investigators did not find significant differences in reproductive indicators, such as serum LH level or FSH level, between genotypes of PCOS patients (12). It demonstrated that the reproductive pathogenicity of a single gene alteration was relatively limited.

Environmental factors, especially dietary factors, strongly influence PCOS patients. Animal studies also have shown that HFHS diet altered ovarian morphology and even induces the formation of polycystic ovaries (13). We performed HFHS diet induction to observe the reproductive function of Thada knockout mice. After ten weeks of HFHS treatment, the body weight increased in both genotypes, but there was no difference in change between WT and KO groups ( Figure 4A ). A subset of indicators for evaluating ovarian function was examined, such as serum hormone levels, estrous cyclicity, and ovarian morphology. Despite the period length of each estrous cycle prolonged by HFHS-fed, KO mice still exhibited similar fluctuations compared to the WT ( Figures 4B, C ). There were no notable differences in hormone levels between the two genotypes in the basal state ( Figures 4D–F ). Meanwhile, there was little change in ovarian morphology and structure corresponding to hormones ( Figure 4G ). These results suggested that Thada knockout mice maintain relatively stable ovarian function under metabolic stress.