Haematoxylin and eosin

Histological assessment and classification of follicle structure were conducted as described previously with slight modification [18 (link),21 (link)]. Briefly, at least three pieces of ovarian tissue in each group were fixed with Bouin’s solution (FUJIFILM Wako Pure Chemical Corporation), kept at 4 °C overnight, dehydrated in a series of ethanol solutions (70–100%), and embedded in paraffin. Serial sections (5 μm thick) of ovarian tissues were prepared and stained with haematoxylin and eosin (both Muto Pure Chemicals CO. Tokyo, Japan). To avoid double counting, three sections, each at least 20 µm apart, were assessed by light microscopy (Nikon ECLIPE E600, Nikon).
Follicles within the ovarian tissues were classified as primordial (<80 um), primary (0.08–1 mm), or pre-hierarchical (>1 mm) [8 (link)]. All the follicles were further characterized as follows: ‘normal’, when the layer of granulosa cells is attached to the spherical oocyte surrounding it and the homogenous ooplasm contains a tiny granulated nucleus, or ‘abnormal’, wherein aggregation and shrinkage of nuclear chromatin and wrinkling of the nuclear membrane were regarded as signs of atresia [6 (link)]. The proportion of morphologically normal follicles per section was calculated by dividing the number of normal follicles by the total number of assessed follicles.

Paraffin-embedded sections were stained with haematoxylin and eosin (200,108, Muto Pure Chemicals, Tokyo, Japan) or Sirius Red (MKCB3138 V, Sigma–Aldrich, Tokyo, Japan). For lipid staining, frozen sections were stained with Oil Red O (M3G0644, Nacalai Tesque, Kyoto, Japan). Macrophages were detected by F4/80 (MCA497R, Bio–Rad, Tokyo, Japan) and Vectastain secondary antibodies (Vector Laboratories, Burlingame, CA, USA). To quantify the area of staining by Oil Red O and Sirius Red, images of 5 random fields from each section were processed using ImageJ software (National Institute of Mental Health, Bethesda, MD, USA). Each value was expressed as the percentage of the total area of each section that stained positively. The number of F4/80-positive cells was counted and averaged for 5 random fields in each section. The NAS and fibrosis stages were scored by Y.K. in a blinded manner according to the method of Kleiner et al.^{44 (link)}.