Integrative genomic analysis reveals a conserved role for prolactin signalling in the regulation of adrenal function

Abstract

We report here that the pituitary hormone prolactin (PRL) has an important, conserved role in regulating adrenal gland function. The adrenal plays a pivotal role in endocrine homeostasis and stress response, which vary across lifespan and have different features in the two sexes.[1] Furthermore, most adrenal disorders have a higher prevalence in women.[2] To get deeper insight into the mechanisms of age- and sex-dependent adrenal function, we performed an integrative analysis of the mouse adrenal transcriptome by RNA-seq and active enhancer (as defined by H3K27ac ChIP-seq) usage at different ages (from E18.5 to P12 weeks) in both sexes. Multivariate analysis showed that age, but not sex, had a significant effect on adrenal global gene expression profiles (Figure and Supporting information S1; Tables S1, S2). Expression of genes related to lipid metabolism and immunity was progressively enriched with age (Figure 1C and Supporting information Table S3), with the proportion of macrophages increasing and neutrophils/mast cells decreasing at older ages (Figure 1D). The expression of different classes of secreted proteins, ion channels and enzymes showed a lower level of sex-dependent enrichment (Figure 1E and Supporting information Table S4). Age-and sex-dependent long non-coding RNA differentially expressed genes (DEG) followed the same pattern as coding DEG, being mostly specific for each age and sex (Supporting information Figure S2 and Table S5). Remarkably, P12 weeks adrenals of both sexes showed increased percentages of novel gene transcripts compared to previous ages (Supporting information Figure S2E). Sexually dimorphic expression pervades all cell populations of the mouse adrenal gland (Supporting information Figure S3 and Table S6). We identified a few thousand enhancers active in the adrenal gland at each temporal stage and in each sex (Supporting information Table S7). A subset among those belong to the super-enhancer class (SEC), genomic regions highly enriched in active chromatin, which regulate developmental and tissue-specific programs and often encompass disease-associated genomic loci.[3] At all ages and in both sexes, adrenal SEC have a much higher level of overlap than typical-enhancer class (TEC) elements (Figure 2A), while adrenal TEC are significantly more conserved than SEC (Figure 2B). Adrenal SEC-associated genes are enriched in genes involved in transcriptional regulation, cell-cell adhesion, protein phosphorylation, biological rhythms and steroid hormone receptor function at all ages and in both sexes (Figure 2C and Supporting information Table S8). Mouse adrenal SEC-associated genes are expressed at significantly higher levels in the adrenal gland (Supporting information Figure S4A) and include a higher percentage of tissue-specific genes (Supporting information Figure S4B) than TEC-associated genes. Human adrenal SEC are also preferentially associated to genes encoding proteins with adrenal-enriched expression than to genes with higher expression in other tissues (Supporting information Figure S4C) and SEC-associated genes are enriched with age-dependent DEG at most times and sex-dependent DEG at P12 weeks compared to TEC (Supporting information Figure S4D,E). Examples of SEC associated to genes important in adrenal physiology are shown in Supporting information Figure S5. Of note, some adrenal SEC-associated genes are expressed preferentially in the adrenal and are involved in blood pressure regulation (Supporting information Table S9). In particular, non-coding SNPs associated with blood pressure traits with very high significance are located inside a conserved SEC encompassing the KCNK3 gene (Supporting information Figure S6 and Table S10). By comparing our list of adrenal sexually dimorphic DEG with data in mouse[4, 5] and rat,[6] we could highlight a core conserved sexually dimorphic gene expression program (Figure 3A and Supporting information Table S11). We focused on the role of the PRL receptor (PRLR), because of its conserved sexually dimorphic expression in the human adrenal (Figure 3B) and the well-known role of PRL signalling in physiological adaptations and response to stress.[7] All Prlr isoforms were upregulated in the female adrenal compared to male at P12, but not P2 weeks of age (Figure 3C). Adrenal gland weight was significantly reduced in adult female Prlr -/- mice compared to WT (Figure 3D), while total body weight was normal in all Prlr -/- animals (Supporting information Figure S7). Signalling pathway impact analysis revealed that the JAK-STAT pathway is significantly activated in female adrenals compared to male at P12 weeks of age (Supporting information Figure S8). The mean area of adrenocortical cells in female Prlr -/- adrenals was significantly decreased (Figure 3E,F). This is consistent with significantly reduced circulating corticosterone levels and a trend toward an increased ACTH/corticosterone ratio in female Prlr -/- mice compared to WT (Figure 3G,H). These results show that PRL signalling through the PRLR has a crucial role in shaping the sexually dimorphic mouse adrenal phenotype. To assess the translational relevance of the mouse model results and to investigate the role of prolactin signalling in human physiopathology, we compared circulating adrenal steroid hormone levels in patients with PRL-secreting (prolactinoma; PRLA) and non-functioning pituitary adenomas (NFPA) (Supporting information Figure S9 and Table S12). Dehydroepiandrosterone sulphate (DHEAS) levels were significantly higher in patients with PRLA compared to NFPA (Figure 4A). In the PRLA group, men had higher circulating PRL levels than women (Figure 4B) and the PRL/DHEAS ratio was significantly more elevated in men than in women (Figure 4C). DHEAS levels were significantly reduced after therapy with dopamine agonists which inhibit PRL, but not ACTH, secretion (Figure 4D-F). Overall, these data suggest that the female adrenal may be more sensitive to PRL effects also in humans. This parallels what we have shown in mice in this study and points on the sexual dimorphic expression of PRLR in the human adrenal gland as a key component of this increased response.[1, 8] In humans, the effect of high PRL to preferentially increase DHEAS levels may be related to the enriched expression of PRLR in the adrenocortical zona reticularis.[9] Modulation of adrenal steroid secretion by PRL may, thus, contribute to the positive effects of physiological concentrations of this hormone on metabolic homeostasis in basal conditions and under stress, while its deregulation in hypo- and hyperprolactinemic states may play an important role in the clinical manifestations of PRL deficiency or excess, respectively. In conclusion, we have unveiled a crucial role for PRL signalling in the sexually dimorphic phenotype of the adult adrenal gland. Our results open new perspectives for the therapy of disorders characterized by adrenal hormones hypersecretion through the use of drugs modulating prolactin release. We thank Frédéric Brau for plugin development for image analysis, Marcin Rucinski for sharing rat adrenal gland microarray data, Xarubet Ruiz-Herrera for animal breeding and genotyping, Matthias Kroiss and Bonald Figueiredo for discussions and critical reading of the manuscript. Funding: This study was supported by the Fondation ARC Project PJA 20191209289 grant to C.R., grants from the Agence Nationale de la Recherche (ANR) ANR15-CE14-0017 (LOCALDO) and ANR20-CE14-0007 (Goldilocks), CNRS EXPOGEN-CANCER International Research Project, Fondation Jérôme Lejeune #1944 to E.L and by the Deutsche Forschungsgemeinschaft (DFG) within the CRC/Transregio (project number: 314061271). The Galaxy server that was used for data analysis is in part funded by Collaborative Research Centre 992 Medical Epigenetics (DFG grant SFB 992/1 2012) and German Federal Ministry of Education and Research [BMBF grants 031 A538A/A538C RBC, 031L0101B/031L0101C de.NBI-epi, 031L0106 de.STAIR (de.NBI)]. The authors declare that they have no conflict of interest. Conceptualization: E.L.; methodology: C.R, E.A., L. S.-M., M.D.-B., N.D., M.J., M.K.; data analysis: C.R., B.A., M.D., F.C., T.D, C.C, E.L.; supervision: T.D., C.C., E.L; writing-original draft: E.L; writing-review and editing: all authors. RNA-seq data: Gene Expression Omnibus GSE173691 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE173691). ChIP-seq data: Gene Expression Omnibus GSE173704 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE173704). Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts