Slow-release gonadotropin-releasing hormone (GnRH) agonist implants are frequently used for contraception in male dogs. Although the effects are fully reversible, there is still concern about the safety of the implant's mode of action. Addressing this, we investigated cellular stress and androgen receptor (AR) signaling during downregulation and recovery. Testicular tissues were sampled from dogs castrated at different time points after GnRH implant removal and compared with untreated controls. AR, hypoxia-inducible factor 1 (HIF1A), heat shock proteins heat shock protein 72 (HSP72), heat shock protein 73 (heat shock cognate, HSPA8) (HSP73), heat shock protein A2 (HSPA2), heat shock protein 90 alpha (inducible isoform) (HSP90AA1), and heat shock protein 90 beta (constitutive isoform) (HSP90AB1) were investigated by quantitative real-time polymerase chain reaction and AR, HSP72, HSP73, and HSP90 immunohistochemically. While AR, HIF1A, and HSP70 were upregulated at gene expression level, HSPA8, HSPA2, and HSP90AA1 expression were downregulated during spermatogenic arrest
HSP90AB1 expression did not change. Immunohistochemistry verified AR-expression in Sertoli, peritubular, and Leydig cells, occasionally also in spermatogonia. Stress-inducible HSP72 was occasionally detected, while constitutive HSP73 and HSP90 were abundantly expressed by germ cells. Our results were similar to studies on seasonal breeders such as pine voles, geese, fish, and soft-shelled turtles. Accordingly, GnRH implants did not impose additional cellular stress on testicular cells when compared with natural recrudescence. Since comparative data on HIF1α are scarce, we cannot draw conclusions about hypoxic conditions.