Fig 2: Evidence of TGFBR1 activation in ovarian granulosa cells of TGFBR1-CAG9Cre mice. a-d Real-time PCR analysis of expression of TGFBR1 CA, Smad7, Inha, and Zp3 in ovaries from control and TGFBR1-CAG9Cre mice at PD3 and PD7. Real-time PCR was performed using ??CT method. Data are mean ± s.e.m. n = 4-5. * P < 0.05, ** P < 0.01, and *** P < 0.001. Ns, not significant. e Western blotting analysis of TGFBR1CA, phospho-SMAD2/3, and HSD3B using ovaries from 2-month-old control and TGFBR1-CAG9Cre mice. TGFBR1CA was detected using an anti-HA antibody. n = 3. Each lane represents an independent sample. f-i X-gal staining using ovaries from Rosa26/Gdf9-iCre mice (f-h) and Rosa26 control mice (i). Panels g and h are higher magnification images of two different fields of panel (f). At least 3 independent samples per group were used. j Real-time PCR analysis of the expression of Gli1, Gli2, and Tgfbr3 using ovaries from 8-week-old control and TGFBR1-CAG9Cre mice. Data are mean ± s.e.m. n = 4-5. * P < 0.05 and *** P < 0.001. k-n RNAscope in situ hybridization analysis of Gli1 mRNA distribution using 8-week-old control (k) and TGFBR1-CAG9Cre ovaries (l). n = 4. Positive and negative controls using TGFBR1 CA flox/+ ovaries were shown in (m) and (n), respectively. Sections were counterstained with hematoxylin. Scale bar is representatively shown in (f) and equals 25 µm (g, h, and k-n) and 100 µm (f and i)

Fig 3: MIS maintains granulosa cell quiescence and prevents normal development of activated follicles. (A) Expression levels and percentages of quiescence-related markers in control- and MIS-treated granulosa cells based on their state of activation. (B and C) RT-qPCR validation of quiescence genes performed on PND6 CTL and MIS mouse (B) and rat (C) ovaries. y axis shows the fold change normalized expression levels relative to control ovary (*<0.05, **<0.01; unpaired Student’s t test, n > 3 for CTL and MIS). (D) Micrographs of brightfield, and DAPI, DDX4, or INHA immunofluorescence of representative primordial, primary and secondary follicles dissected from adult mouse ovaries. Percentage of primordial and activated follicles following 24 h of culture with control or 10 µg/mL MIS. (E) Double immunofluorescence DDX4 (red)/INHA (green) of PND6 CON and MIS rat ovaries. Examples in Insets of normal INHA+ growing preantral follicles with one or several layers of cuboidal granulosa cells (Up Middle, Right, circled with white dashes), and MIS-treated abnormal follicles with large oocytes surrounded by inhibin negative squamous granulosa cells or singular layer of weakly positive INHA+ granulosa cells (Down Middle, Right, circled with white dashes). (F) Oocyte diameter in follicles of CTL- and MIS-treated PND6 rat ovaries classified according to the INHA immunofluorescence status of their surrounding granulosa cells (Up). A population of large oocytes (diameter, >66 µm; smallest activated primary follicles in CTL) surrounded by INHA-granulosa cells in MIS-treated ovaries are circled in blue. Percentage of INHA+ granulosa cells surrounding growing large oocytes (diameter, >66 µm) (Down). (G) RNAish showing colocalization of Misr2 (red)/Nr5a2 (green) and Nr5a2 (green)/Inha (red) in PND6 CTL- and MIS-treated mouse ovaries. (Scale bars, 50 µm.) Insets magnify follicles from CTL and MIS-treated tissues. (H) Histology of H&E-stained 5-µm section of CTL- and MIS-treated rat ovaries at PND15. Representative images of primordial, primary and secondary follicles are shown in Insets. (Scale bars, 50 µm.) (I) Scatterplot of oocyte area by function of follicle area with linear regression in PND6 MIS-treated and CTL mouse ovaries (n = 3). (J) Average oocyte diameter in PND6 mice ovaries (Up). Average linear regression slope from Fig. 7I (Down).

Fig 4: Immunohistochemical analysis of ovarian tumor markers. a-j Expression of granulosa cell and germ cell markers in 8-week-old control and TGFBR1-CAG9Cre ovaries. Representative images from immunohistochemical analysis of FOXL2 (a and b), INHA (c and d), FOXO1 (e and f), AMH (g and h), and DDX4 (i and j) are shown. k and l Negative controls using isotype-matched rabbit and goat IgGs. Experiment was performed using ABC method, and signals were developed using NovaRED Peroxidase Substrate Kit. Sections were counterstained with hematoxylin. Five independent samples per group were utilized in this analysis. Scale bar is representatively depicted in (a) and equals 50 µm (a-l)