Three-Dimensional Analysis of Busulfan-Induced Spermatogenesis Disorder in Mice

We examined if the distribution of impaired or normal spermatogenesis differs along the length of seminiferous tubules in disorders of spermatogenesis. For this purpose, three-dimensional (3D) reconstruction of seminiferous tubules was performed in mice with experimental spermatogenesis disorder induced by intraperitoneal injection of busulfan, and the areas of impaired and normal spermatogenesis were analyzed microscopically. The volume of the testis and length of seminiferous tubules decreased, and the proportion of tubule areas with impaired spermatogenesis increased depending on the dose of busulfan. With the highest dose of busulfan, although the proportion of impaired spermatogenesis was similar among individual seminiferous tubules, it was slightly but significantly higher in shorter tubules and in tubule areas near branching points. The tubule areas with impaired and normal spermatogenesis consisted of many segments of varying lengths. With increasing doses of busulfan, the markedly impaired segments increased in length without changing in number, whereas normal segments, although reduced in number and length, remained even with the highest dose of busulfan. Individual remaining normal segments consisted of several different stages, among which stage I and XII were found at higher frequencies, and stage VI at a lower frequency than expected in normal seminiferous tubules. We also examined if the distribution of impaired or normal spermatogenesis differs among different 3D positions in the testis without considering the course of seminiferous tubules. Although the proportions of impaired spermatogenesis with the minimum dose of busulfan and normal spermatogenesis with the highest dose of busulfan greatly varied by location within a single testis, there were no 3D positions with these specific proportions common to different testes, suggesting that the factors influencing the severity of busulfan-induced spermatogenesis disorder are not fixed in location among individual mice.