Mechanism of action of bolandiol (19-nortestosterone-3beta,17beta-diol), a unique anabolic steroid with androgenic, estrogenic, and progestational activities

Abstract

Bolandiol is a synthetic anabolic steroid that increases lean body mass and bone mineral density without significant stimulation of sex accessory glands in castrate adult male rats. Since bolandiol suppresses gonadotropins and endogenous testosterone (T) production, we investigated its mechanism of action. We compared the potency of bolandiol in vitro and in vivo with T, 5alpha-dihydrotestosterone (DHT), 19-nortestosterone (19-NT) and estradiol (E(2)). Bolandiol bound with lower affinity to the recombinant rat androgen receptor (AR) than the other androgens and had low, but measurable, affinity for recombinant human progestin receptors (PR-A, PR-B), and estrogen receptors (ERalpha and beta-1). Functional agonist activity was assessed in transcription assays mediated by AR, PR, or ER. Bolandiol was stimulatory in all these assays, but only 4-9% as potent as T, DHT, and 19-NT via AR, 1% as potent as progesterone via PR, and 3% and 1% as potent as E(2) acting through ERalpha or ERbeta, respectively. In immature castrate rats, bolandiol was equipotent to T in stimulating growth of the levator ani muscle but less potent than T in stimulating growth of the sex accessory glands. Bolandiol also stimulated uterine weight increases in immature female rats, which were partly blocked by ICI 182,780, but it was not aromatized in vitro by recombinant human aromatase. In contrast to T, stimulation of sex accessory gland weights by bolandiol was not inhibited by concomitant treatment with the dual 5alpha-reductase inhibitor dutasteride. As bolandiol exhibits tissue selectivity in vivo, it may act via AR, PR, and/or ER, utilize alternative signaling pathway(s) or transcriptional coregulators, and/or be metabolized to a more potent selective steroid.

Fig. 1

Androgenic activity of bolandiol. CV-1 cells were transiently transfected with 3XHRE-LUC and a human AR expression vector and treated with various concentrations of bolandiol (A). Inhibition of bolandiol-stimulated transactivation by the antiprogestin CDB-2914 (B) or by the antiandrogen bicalutamide (C).

Fig. 2

Progestational activity of bolandiol. T47Dco cells were transiently transfected with PRE₂-tk-LUC and treated with various concentrations of bolandiol (A). Inhibition of bolandiol-stimulated transactivation by CDB-2914 (B) or by bicalutamide (C).

Fig. 3

Estrogenic activity of bolandiol. T47Dco cells were transiently transfected with 3XERE-LUC and treated with various concentrations of bolandiol (A). Inhibition of bolandiol-stimulated transactivation by ICI 182,780 in T47Dco cells (B). HEK-293 cells were cotransfected with ERE-LUC and hERα (C) or hERβ (D) and treated with various concentrations of E₂, bolandiol, or 19-NT.

Fig. 4

Dose-response for stimulation of (A) ventral prostate (VP), (B) seminal vesicles (SV), or (C) levator ani muscle (LA) weights by various androgens or vehicle. Twenty-two day-old castrate rats were treated sc for 8 days with bolandiol (0.13, 0.5, or 2.0 mg/rat/day), 19-NT (0.13, 0.5, 1.0, or 2.0 mg/rat/day), T (0.013, 0.05, 0.2, or 0.8 mg/rat/day), or vehicle (10% ethanol in sesame oil), necropsied, and VP, SV, and LA weights determined. Organ weights in vehicle-treated animals have been displaced to the left of 0 dose on the x-axis for clarity. There were no significant differences in VP weights in animals treated with bolandiol at 1 or 4 mg total dose compared with T at 0.1 and 0.4 mg total dose (P>0.05). SV weights at 1 mg total dose of bolandiol were not different from those at 0.4 mg total dose of T, and SV weights at 4 mg total dose of bolandiol were not different from those at 1.6 mg total dose of T (P>0.05). There was no significant effect of these treatments on final body weight.

Fig. 5

Serum LH levels in castrate immature male rats treated with various doses of androgens or vehicle. Twenty-two day-old castrate rats were treated sc for 8 days with bolandiol, T, or vehicle (A) or 19-NT, T, or vehicle (B) as indicated in the legend to Fig. 4. *P< 0.05 vs vehicle-treated.

Fig. 6

Dose-response for stimulation of uterine weights by estradiol (E₂), bolandiol, or vehicle. Immature female rats were treated sc for 3 days with E₂ (0.017, 0.067, 0.267 μg/rat/day), bolandiol (0.08, 0.33, 1.25 mg/rat/day), or vehicle (10% ethanol/sesame oil). ICI 182,780 (0.4 mg/rat/day) was administered concurrently with a maximal stimulatory dose of E₂ or bolandiol, and uterine weights were determined. *P<0.05 vs E₂ alone; **P<0.05 vs bolandiol alone by Student’s t-test.

Fig. 7

Effect of dutasteride on (A) ventral prostate (VP) or (B) seminal vesicles (SV) weights in castrate immature rats. Twenty-two day-old castrate rats were treated sc for 8 days with doses of bolandiol, 19-NT, or T that had previously been shown to be stimulatory or with vehicle (10% ethanol/sesame oil), necropsied, and VP and SV weights determined. Dutasteride significantly suppressed T-stimulated VP (*P<0.05) and SV (*P<0.001) weights. There was no significant effect of dutasteride on sex accessory gland weights in rats treated with vehicle or bolandiol, whereas there was a significant increase (*P<0.05) in both VP and SV weights in rats treated with 19-NT. There was no significant effect of these treatments on final body weight.

Fig. 8

Time course of incubation of (A) testosterone (T) or (B) bolandiol with recombinant human aromatase. Aromatization of T was complete by 120 min, whereas no aromatization of bolandiol was observed during a 180 min incubation. Recovery of all steroids in this system decreased with incubation time (i.e. at 120 min, the amount of T + E₂ was 80% of that at 0 min; the amount of bolandiol was 75% of that at 0 min).