Androgen receptor distribution in adult human testis.

The distribution of the androgen receptor (AR) in archival human testes was determined immunocytochemically using an affinity-purified peptide-specific rabbit antibody, PG21, and employing a modified biotin-streptavidin-immunoperoxidase method that incorporated a biotin amplification step. In combination with microwave epitope retrieval, the biotin amplification step increased the sensitivity of the immunostaining assay approximately 20-fold. Thus, the useful range at which PG21 rendered a robust, specific immunostaining signal without also increasing nonspecific background was extended dramatically. Broadening the useful range of the PG21 antibody made it possible to resolve the relative amounts of immunopositive AR in different cell types of the human testis. At a high PG21 concentration, for example, all AR-positive cells exhibited a robust immunostaining intensity, but it was not possible to distinguish between nuclei exhibiting either high or moderate immunostaining intensities. In contrast, as the concentration of PG21 was decreased, distinct populations of testicular cells exhibited differential AR immunostaining intensities in their nuclei. AR immunostaining of Sertoli cell nuclei was present at low PG21 concentrations at which no immunostaining of peritubular myoid cells or Leydig cells could be detected. In turn, AR immunostaining of peritubular myoid cells was detected at PG21 concentrations that did not immunostain Leydig cells. Moreover, within the seminiferous epithelium, Sertoli cell nuclear AR staining intensity was less at stages V and VI of the cycle of the seminiferous epithelium than that at stage III, and stage III staining intensity was greater than that at stages I and II. This AR immunostaining pattern in human Sertoli cell nuclei as a function of the cycle of the seminiferous epithelium is reminiscent of the pattern observed in rodent species. Finally, no AR immunostaining of germ cells was observed at any of the PG21 concentrations examined.