Analysis of atresia in equine follicles using histology, fresh granulosa cell morphology and detection of DNA fragmentation.

Follicular atresia has been examined previously by various biochemical and histological methods. The aim of this study was to compare, for the first time, detection of granulosa cell apoptosis by biochemical DNA analysis and microscopic examination of fresh granulosa cell morphology with the established method of detecting atresia by histology in equine follicles. DNA extracted from granulosa cells was examined by staining with ethidium bromide and end-labelling with [(32)P]dideoxy-ATP, which labels the free 3'-end of DNA fragments. In 25 of 26 follicles (96%) there was agreement between end-labelling and staining of DNA with ethidium bromide (P < 0.001). Granulosa cell apoptosis was distinguished more easily in the end-labelled samples than by staining with ethidium bromide. Histological atresia and apoptosis as detected by biochemical DNA analysis were significantly correlated (P < 0.02) with 20 of 22 follicles (91%) receiving corresponding classifications with the two methods. No follicles with granulosa cell apoptosis as detected by biochemical DNA analysis were histologically viable, but some of the histologically early atretic follicles did not display DNA laddering. Stereomicroscopic evaluation of morphology of the fresh granulosa cells was significantly correlated (P < 0.001) with the histological findings, with 29 of 33 follicles (88%) receiving corresponding classifications. There was a potential error in determining follicle health by biochemical DNA analysis only, as both histologically early and late atretic follicles in some cases did not show DNA laddering. Thus, if relying solely on biochemical detection of apoptosis, severely atretic follicles could wrongly be classified as healthy follicles.