Determining the role of androgen receptor and glucocorticoid receptor in the rodent adrenal cortex through conditional gene targeting

Abstract

Androgens are well documented as important regulators of male health, primarily in the maintenance and development of male sexual characteristics. However, a decline in circulating androgens has also been associated with co-morbidities such as obesity, cardiac disease and metabolic syndrome. Previous research has focussed upon the body wide impact of adrenal androgens, however whilst androgen receptor (AR) is abundantly expressed in the adrenal cortex of both rodents and humans, surprisingly little is known about androgen action on the adrenal cortex itself. This gap in our understanding is at least in part due to the perceived lack of suitable animal models. Rodents have largely been overlooked as a model system as their adrenals are unable to produce androgens due to lack of 17α Hydroxylase and 17, 20 lyse activity and they therefore do not have a zona reticularis. However, historical studies using castrated mice showed that removal of androgens leads to the redevelopment of an additional cortex zone known as the transient X-zone. The foetal adrenal is thought to give rise the adult adrenal cortex in human and rodents. These foetal cells are maintained for a period postnatally and regress differently depending on species and sex. In the human this zone is known as the ‘foetal zone’, and the rodent homologue termed the ‘X-zone’. The mechanisms underpinning the regression of the X-zone and its purpose and maintenance postnatally still aren’t clearly understood.

To provide a comprehensive overview of androgen signalling in the adrenal cortex, multiple mouse models were utilised. First, Cre/loxP technology was used to ablate AR specifically from the adrenal cortex. Further androgen manipulation was achieved through castration (removal of androgens) and human chorionic gonadotropin (hCG) treatment (increased androgens). The initial study investigates the impacts on the male mouse adrenal. Histology analysis revealed the presence of an X-zone in all experimental cohorts following loss of AR or circulating androgens, confirmed by 20- α-hydroxysteroid dehydrogenase (20 alpha-HSD) expression. These data demonstrate that androgens signalling via AR is required for X-zone regression during puberty. However, interrogation of morphology of hCG treated cohorts revealed no phenotypic changes compared to controls, this demonstrates that hyper stimulation with androgens does not negatively impact the adrenal cortex or influence X-zone morphology.

Differences in X-zone morphology and 20 alpha-HSD localization prompted cortex measurements which revealed significant differences in X-zone depth and cell density depending on ablation of AR, circulating androgens or both. This suggests that androgens and androgen receptor are working together and also independently to regulate the adrenal cortex. This result was strengthened through analysis of steroid enzyme genes and cortex markers, which revealed that normal AKR1B7 expression was absent following loss of androgens but not androgen receptor. A final part of this study examined the impacts long term androgen receptor ablation and long term castration in ageing animals.

A final part of this study examined the impacts long term androgen receptor ablation and long term castration in ageing animals. These results demonstrate that following prolonged loss of androgens that there is no major disruption to the adrenal cortex. Morphology analysis and X-zone measurements revealed that X-zone regression was occurring in mice with long term castration, characterized by a reduction in size and pockets of vacuolization throughout the X-zone. This phenotype is also observed in ageing females with X-zone regression via vacuolization. These data suggest that following prolonged loss of androgens, the male adrenal is feminized and behaves as such. In contrast, AR ablation only, results in an enlarged adrenal with large spindle cell lesions and X-zone expansion confirmed by X-zone measurements. Initial experiments have demonstrated that androgens can work independently of AR to regulate the adrenal cortex. Together these data suggests that AR is required to control the appropriate action of circulating androgens in the adrenal cortex, with loss of AR resulting in off target signalling from circulating androgens in the adrenal leading to spindle cell hyperplasia, X-zone expansion and X-zone mislocation.

A second set of studies were carried out to determine the role of androgen signalling in the female adrenal, specifically, if loss of AR leads to the absence of normal X-zone regression during pregnancy. To answer this question the same selective AR ablation model was used. Analysis of litters comparing observed and expected genetic distribution revealed significantly fewer females being born carrying complete ablation of adrenal AR. Morphology analysis of these mice revealed severe cortex disruption and spindle cell hyperplasia similar to that observed in mutant males. Investigation of adrenals following pregnancy revealed that X-zone regression still occurred despite loss of AR. This result shows that X-zone regression in the female is under different regulation compared to male adrenal and occurs via an androgen-independent signalling mechanism. However, loss of AR still leads to anatomical dysregulation of the adrenal cortex.

AR ablation revealed changes in glucocorticoid receptor (GR) expression in the adrenal cortex. To dissect this relationship further a final study was conducted, attempting to ablate GR from the adrenal cortex also using the Cyp11a1 Cre. Initial observations of these mice revealed excessive hair loss through barbering, curved spines and stressed behaviour when monitored in the cage under normal conditions. Immunohistochemistry was used to confirm GR ablation in the adrenal cortex, however, to our surprise, GR expressing cells were not steroidogenic and thus were not targeted by the Cre recombinase. Despite no GR ablation in the adrenal, morphology analysis revealed severe disruption to the adrenal cortex. The Cyp11a1 Cre not only targets the adrenal but is expressed in the hindbrain. To determine if GR ablation in the hindbrain explains the phenotype, we next used PCR analysis interrogating hindbrain genomic DNA to determine if there was recombination of GR. Results confirmed GR recombination in the hindbrain. Due to the observation of stressed behaviour and adrenal cortex disruption, we wanted to determine if this was a result of hyperactivity of the adrenal cortex. Serum corticosterone was analysed and was elevated in these animals. These data revealed that GR ablation in the hindbrain results in adrenal cortex disruption and an elevated stress response, potentially highlighting a new model to investigate stress disorders and their impact on the hypothalamic-pituitary-adrenal axis.

Together this data defines new roles for AR signalling in the adrenal cortex and the role of the hindbrain GR signalling in regulating adrenal morphology and function.