Trilostane exerts antidepressive effects among wild-type, but not estrogen receptor β knockout mice

Abstract

Women with estrogen receptor (ER) positive breast cancer, who are treated with the ER blocker, tamoxifen, have an increased risk of depression. Trilostane, a 3β-hydroxysteroid dehydrogenase inhibitor, is now being used to treat tamoxifen-insensitive breast cancer. In-vitro assays show that trilostane may have actions through ERβ. Results of in-vivo research shows that actions at ERβ may underline some antidepressant effects of estrogen. We hypothesized that trilostane may exert antidepressive effects in the forced swim in part due to actions through ERβ. Trilostane (25 mg/kg, intraperitoneally), compared with vehicle, had significant antidepressant-like effects but only when administered to wild-type, not ERβ knockout, mice. Thus, actions of trilostane through ERβ may underlie some of its antidepressant-like effects.

Introduction

Breast cancer directly affects one out of seven women. Many (approximately 70%) breast cancer patients (BCPs) have estrogen (E₂) receptor (ER) positive metastases. However, among some BCPs, some of them (approximately 30%) do not benefit from the primary antihormone therapy utilized, tamoxifen. Insensitivity and/or resistance can develop to tamoxifen. Trilostane, a 3β-hydroxysteroid dehydrogenase inhibitor, traditionally has been used to treat Cushing syndrome, but more recently, higher dosages have been used to treat tamoxifen-insensitive breast cancer. Thus, trilostane represents an important alternative therapy.

Some of trilostane’s actions may be through the β isoform of the ER (ERβ). Trilostane’s effect on ERs may shift the balance of ERα and ERβ subtypes [1], Trilostane increases binding to ERβ by altering kinetics of ERβ. Changes in ERβ conformation induced by trilostane may increase ERβ stability [2]. ERβ is more highly expressed in BCPs with better disease prognosis and outcomes [3]. Thus, ERβ may be an important therapeutic target, and/or prognostic factor, in the treatment of breast cancer.

Actions at ERβ may also be important in mediating depression. In support, among some BCPs, tamoxifen (which can block ERα and β isoforms) treatment increases negative affective symptomology [4,5]. Expression of ERβ is greater than ERα in the hippocampus, a brain area that is important for depression and affective function [6,7]. Moreover, actions at ERβ have beneficial trophic effects in the brain for affect and cognition, and these effects occur without potential proliferation in peripheral tissues. Among some women and rodents, administration of E₂ produces antidepressant effects [7]. In addition, administration of ERβ-specific, but not ERα-specific, selective ER modulators (SERMs), systemically or to the hippocampus, can have antidepressant effects in the forced swim task (FST) among rats or mice [8], but these effects are not observed when ERβ is inactivated [7,8]. An important question is whether trilostane’s actions are through ERβ. To investigate this, we examined effects on depressive behavior in the forced swim test of trilostane administration to wild-type (WT) and ERβ knockout (βERKO) mice. We hypothesized that if ERβ is necessary for trilostane’s antidepressant effects, then WT, but not βERKO, mice will show less immobility in the forced swim test.

Materials and methods

All methods utilized were preapproved by the Institutional Animal Care and Use Committee at University at Albany-SUNY.

Animals and housing

Gonadally intact male mice, WTor βERKO mice bred on a C57BL/6 background (breeder pairs obtained from Jackson Laboratory, Bar Harbor, Maine, USA), were used for the study. Males were used to minimize potential confounds of E₂ among females. In experiment 1, a total of 50 WT mice, 10 mice per group were tested. In experiment 2, 30 WT (15 per group) and 30 βERKOs (15 per group) were tested. Mice were group-housed (4/5 per cage) in polycarbonate cages, containing woodchip shavings for bedding, in a temperature-controlled room (21 ± 1°C) in the Laboratory Animal Care Facility. Mice were maintained on a 12/12-h reversed light cycle (lights off at 8 : 00 h) with Purina Mice Chow (Scott’s Distributors, Boston, Massachusetts, USA) and tap water in their home cages ad libitum.

Genotyping

Given that the genetic background of this strain of mice cannot be determined by phenotypic characteristics, genotypes of experimental mice were determined with a typical polymerase chain reaction analyses. The detailed methods used to genotype these mice have been published [9].

Drug administration

Trilostane (6.25, 12.5, 25, or 50 mg/kg) was suspended in vegetable oil (based upon pilot studies [10]). Drug was vortexed and spun on heat until it was thoroughly suspended. Experiment 1 was a dose-response study in which WT mice were administered 0, 6.25, 12.5, 25, or 50 mg/kg (n = 10) intraperitoneally of trilostane 16 h and 2 h before behavioral testing. Experiment 2 looked at the responsiveness of WT compared with that of βERKO mice to trilostane (the most effective dosage from experiment 1). Mice were administered with vehicle or 25 mg/kg trilostane at 16 and 2h before behavioral testing.

Forced swim test

In the forced swim test, the dependent measure of depression-like behavior is time spent immobile. In the forced swim test, mice were placed in a glass cylinder, which was 20.5 cm in diameter and 21.5 cm in depth, and was filled with 18 cm of 25°C water. A video camera and an experimenter recorded the time spent immobile, minute-by-minute [11].

Tissue collection

Immediately after testing, mice were euthanized by cervical subluxation. Whole brain and plasma were collected to measure steroid hormone levels.

Radioimmunoassay for steroid hormones

Trilostane is a 3β-hydroxysteroid dehydrogenase inhibitor, which can influence levels of 3α-hydroxy-5α-pregnan-20-one (3α-5α-THP), which can influence depression [12,13]. As such, we measured levels of 3α-5α-THP in the brain and plasma and corticosterone in plasma. The radioactive probe utilized for 3α-5α-THP (NET-1047: specific activity 65.0 Ci/mmol) was purchased from Perkin Elmer (Boston, Massachusetts, USA). The 3α-5α-THP antibody (#921412–5), obtained from Dr Robert Purdy (Veterans Medical Affairs, La Jolla, California, USA), when used in a 1 : 5000 dilution binds between 40 and 60% of [³H] 3α-5α-THP and bound 47% in this study. The corticosterone antibody (#B3-163, Endocrine Sciences, Calabasas, California, USA), which typically binds 40–60% of [³H] corticosterone, was used in a 1 : 20 000 dilution and bound 45% in this study. [³H] corticosterone (NET-182: specific activity = 48.2 Ci/mmol), was purchased from Perkin Elmer. Standard steroid extraction and radioimmunoassay techniques used by our laboratory for measurement of these steroids in plasma and whole brain were utilized [14,15].

Statistical analyses

In experiment 1, a one-way analysis of variance test was used to determine time spent immobile in the forced swim test. In experiment 2, two-way analysis of variances were used to determine effects of trilostane and genotype (WT vs. βERKO) for time spent immobile. A P value of ≤ 0.05 was considered significant.

Results

In experiment 1, Trilostane dose-dependently altered immobility

Trilostane significantly altered total time spent immobile [F(4,39) = 2.85, P ≤ 0.03, Table 1]. Mice administered with 25 mg/kg trilostane spent less time immobile than did vehicle-administered mice (P = 0.01).

Table 1.

The mean (±SEM) time (secs) spent immobile in the forced swim test

	Treatment
	Vehicle	6.25 mg/kg	12.5 mg/kg	25 mg/kg	50 mg/kg
Total immobility	412 ± 21	395 ± 25	377 ± 18	327 ± 31*	391 ± 10

^*A significant difference from all other groups (P<0.05).

In experiment 2, Trilostane to WT, but not βERKO, mice decreased depressive behavior, increased 3α-5α-THP

The significant interaction between genotype and trilostane [F(1,28) =9.43, P < 0.01] was due to trilostane reducing immobility of WT, but not βERKO, mice. Genotype and trilostane condition also interacted to influence 3α-5α-THP levels [F(1,21) = 5.27, P ≤ 0.03, Fig. 1]. Trilostane increased 3α-5α-THP levels in the hippocampus of WT, but not βERKO, mice, compared with vehicle. There were no significant differences in plasma corticosterone concentration between groups. Although nonsignificant, corticosterone levels were decreased among WT, but not βERKO (vehicle WT 13.5 ± 5.3, 25 mg/kg WT 10.3 ± 2.1, vehicle βERKO, 15.0 ±4.8, 25 mg/kg βERKO, 14.5 ± 6.4), mice administered with vehicle (13.5 ± 5.3) compared with trilostane (10.3 ± 2.1).

Fig. 1.

The mean (± SEM) time (secs) spent immobile in the forced swim test was significantly lower among the wild-type mice administered with trilostane with an increase of hippocampal 3α-5α-THP compared with estrogen receptor β knockout (βERKO) mice. *A significant difference from all other groups (P<0.05).

Discussion

Our hypothesis that ERβ is necessary for trilostane’s antidepressant effects to decrease immobility in the forced swim test was supported. In experiment 1, there was a dose-dependent effect of trilostane among WT mice to decrease depressive-like behavior. Mice administered 25 mg/kg trilostane had significantly less immobility than did vehicle-administered mice. In experiment 2, WT, but not βERKO, mice administered 25 mg/kg of trilostane, compared with vehicle, and had less immobility in the force swim task and greater hippocampal 3α-5α-THP levels. Thus, trilostane’s antidepressant effects may be due, in part, to actions at ERβ involving 3α-5α-THP.

The present results, that the administration of trilostane to WT, but not βERKO mice, had less depressive behavior in the FST, confirms previous findings. Previous data has shown that when endogenous steroid hormones are increased, exogenous administration of E₂ or administration of ERβ SERMs have decreased depressive-like behavior in the FST among WT but not βERKO mice [8]. This effect also has been shown in rats [16]. In addition, increasing immobility of βERKO mice compared with WT mice is consistent with previous reports using the FST [8,17]. Furthermore, ERβ-SERM, but not ERα-SERM, administration systemically, or to the hippocampus, have antidepressant effects in FST among rats or mice [7]. Thus, actions at ERβ may underlie some of trilostane’s effects to decrease depressive behavior.

In this study, trilostane increased 3α-5α-THP in the hippocampus of WT but not βERKO mice. An important question is whether some of the effects of trilostane through ERβ are due to effects on 3α-5α-THP levels. 3α-5α-THP has been implicated in having antidepressant effects among people [18-20] and in animal studies. Animal models have shown that when steroid hormone levels are high (proestrous), 3α-5α-THP levels are greater and depressive-like behavior is decreased [16]. Removal of the primary endogenous source of steroid hormones (ovariectomy) increases depressive-like behavior among female rats in the forced swim test and decreases 3α-5α-THP levels [21]. Furthermore, blocking formation of 3α-5α-THP prevents antidepressant behavior [22]. In addition, trilostane decreases corticosterone levels in WT but not βERKO mice. Although it is known that trilostane can influence the production of corticosteroids, such as corticosterone, which can influence depression, these effects were not seen when tissues were collected immediately after FST. The decreases in corticosterone levels may be due, in part, to the increases in 3α-5α-THP levels. Thus, 3α-5α-THP may be beneficial to decrease depressive-like behavior.

Tamoxifen has been associated with negative affect in BCPs. Animal models have shown that coadministration of E₂ and tamoxifen attenuates the effects of E₂ to reduce the adrenocorticotropin hormone and corticosterone response to restraint stress [23]. In addition, coadministration of ERβ SERMs and tamoxifen blocks antianxiety behavior [24]. Owing to the increase of negative affect associated with tamoxifen, trilostane may be a beneficial alternative therapeutic drug. Animal models have shown that trilostane inhibits E₂-stimulated proliferation that enhances the affinity of E₂ binding to ER in rat uteri [25]. Rats treated with trilostane, compared with controls, experienced an inhibition of 7,12, Dimethylbenz(a) anthracene-induced mammary tumors [25]. Findings also show in MCF-7 and T-47D breast cancer cells there are a concentration inhibition of E₂-stimulated estrogen response element activity [1]. Thus, upregulation of ERβ may underline some of trilostane’s effects associated with positive affect.

Conclusion

To date, there is no cure for breast cancer. There are alternative therapeutics, such as trilostane, for tamoxifen-resistant women, that are effective in stabilizing the disease and/or decreasing mortality. In this study, trilostane has antidepressant effects that were associated with an increase in 3α-5α-THP, among WT, but not βERKO mice. This implies that the antidepressant effects of trilostane may involve signaling through ERβ to enhance 3α-5α-THP.