To investigate the mechanism of androgen-induced thymolysis, the effec
ts of various androgens, including testosterone (T), 19-nortestosteron
e, and 7 alpha-methyl-19-nortestosterone (MENT), were compared with th
ose of estradiol and dexamethasone (DEX) in intact, castrated, and adr
enalectomized male rats. The potency comparisons on thymus regression,
based on mass of steroids, showed DEX to be the most potent, followed
by estradiol and the androgens. Among the androgens, MENT was the mos
t potent, followed by nortestosterone and T, an order similar to their
anabolic potency on muscle. As the thymolytic effects of T and MENT w
ere not altered by the concomitant administration of an aromatase inhi
bitor or a 5-reductase inhibitor, it was concluded that the effects of
androgens were not mediated by their conversion to estrogens or 5 alp
ha-reduced metabolites. involvement of glucocorticoid receptors in and
rogen action was excluded because mifepristone (an antiglucocorticoid)
blocked DEX-induced, but not T-or MENT-induced, thymus regression. Fl
utamide, an antiandrogen, significantly blocked the thymolytic effect
of T and MENT, providing further support for this conclusion. This sug
gested that the thymolytic action of androgens is an intrinsic propert
y mediated via androgen receptors (AR). The occurrence of AR in the th
ymus was demonstrated by binding assays and the presence of AR messeng
er RNA (mRNA) by reverse transcriptase-polymerase chain reaction. Quan
titative reverse transcriptase-polymerase chain reaction for AR mRNA i
n the thymus showed 6-fold more AR mRNA in the thymic epithelial cells
than in the thymocytes. However, epithelial cells represent only a sm
all fraction of the thymus. Hence, it is hypothesized that the androge
ns produce their thymolytic effects by stimulating the secretion of a
factor(s) by the thymic epithelial cells that, in turn, causes regress
ion of the thymus.