Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Sep;28(9):1471-86.
doi: 10.1210/me.2014-1060. Epub 2014 Jul 16.

Wnt Signaling Inhibits Adrenal Steroidogenesis by Cell-Autonomous and Non-Cell-Autonomous Mechanisms

Affiliations
Free PMC article

Wnt Signaling Inhibits Adrenal Steroidogenesis by Cell-Autonomous and Non-Cell-Autonomous Mechanisms

Elisabeth M Walczak et al. Mol Endocrinol. .
Free PMC article

Erratum in

  • Corrigendum.
    Mol Endocrinol. 2015 Dec;29(12):1805. doi: 10.1210/me.2015-1259. Mol Endocrinol. 2015. PMID: 26623962 Free PMC article. No abstract available.

Abstract

Wnt/β-catenin (βcat) signaling is critical for adrenal homeostasis. To elucidate how Wnt/βcat signaling elicits homeostatic maintenance of the adrenal cortex, we characterized the identity of the adrenocortical Wnt-responsive population. We find that Wnt-responsive cells consist of sonic hedgehog (Shh)-producing adrenocortical progenitors and differentiated, steroidogenic cells of the zona glomerulosa, but not the zona fasciculata and rarely cells that are actively proliferating. To determine potential direct inhibitory effects of βcat signaling on zona fasciculata-associated steroidogenesis, we used the mouse ATCL7 adrenocortical cell line that serves as a model system of glucocorticoid-producing fasciculata cells. Stimulation of βcat signaling caused decreased corticosterone release consistent with the observed reduced transcription of steroidogenic genes Cyp11a1, Cyp11b1, Star, and Mc2r. Decreased steroidogenic gene expression was correlated with diminished steroidogenic factor 1 (Sf1; Nr5a1) expression and occupancy on steroidogenic promoters. Additionally, βcat signaling suppressed the ability of Sf1 to transactivate steroidogenic promoters independent of changes in Sf1 expression level. To investigate Sf1-independent effects of βcat on steroidogenesis, we used Affymetrix gene expression profiling of Wnt-responsive cells in vivo and in vitro. One candidate gene identified, Ccdc80, encodes a secreted protein with unknown signaling mechanisms. We report that Ccdc80 is a novel βcat-regulated gene in adrenocortical cells. Treatment of adrenocortical cells with media containing secreted Ccdc80 partially phenocopies βcat-induced suppression of steroidogenesis, albeit through an Sf1-independent mechanism. This study reveals multiple mechanisms of βcat-mediated suppression of steroidogenesis and suggests that Wnt/βcat signaling may regulate adrenal homeostasis by inhibiting fasciculata differentiation and promoting the undifferentiated state of progenitor cells.

Figures

Figure 1.
Figure 1.
The Wnt-responsive population is heterogeneous in vivo. Adrenal glands from 6-week-old male Tcf/Lef:H2B-GFP;ShhLacZ mice were evaluated on paraffin sections. A, Representative coimmunofluorescence image for LacZ, GFP, PCNA, and nuclear counterstain DAPI. The inset on the right is an overlay without DAPI. B, Coimmunofluorescence images for BrdU, GFP, and LacZ staining. Mice were harvested 2 hours after a single BrdU injection. The inset at the bottom is an overlay without DAPI. C, Quantification of data from A from 3 sections each of n = 7 animals; error bars represent SEM. Diagram on the right represents the relative size and overlap of each population graphically. D, Coimmunofluorescence images of Cyp11b2 and LacZ. E, Coimmunofluorescence images of Cyp11b1 and LacZ. F, Coimmunofluorescence images of Cyp11b2 and GFP. G, Coimmunofluorescence images of Cyp11b1 and GFP. H, Quantification of data from D and F from 3 sections each of n = 4 animals; error bars represent SEM. Diagram on the right represents the relative size and overlap of each population graphically. Scale bars, 100 μm.
Figure 2.
Figure 2.
Adrenocortical cell model of βcat activity. ATCL7 cells were treated with 0.5μM BIO or vehicle (DMSO) in low-serum media for 24 hours. A, Subcellular localization of βcat was assessed using immunocytochemistry with βcat antibodies. B, ATCL7 cells were separated into nuclear and cytoplasmic fractions. Fractionation of βcat was determined by SDS-PAGE and immunoblotting. Nuclear lamin A/C and cytoplasmic β-tubulin served as loading and fractionation controls. C, qRT-PCR assessing Axin2 expression 8 and 24 hours after treatment. Values represent the mean ± SD, where vehicle-treated cells were normalized to 1 (n = 4). **, P < .005; ***, P < .0005.
Figure 3.
Figure 3.
Stimulation of βcat activity inhibits basal and ACTH-induced zF steroidogenesis in vitro. ATCL7 cells were pretreated with 0.5μM BIO or vehicle for 12 hours in low-serum media followed by 100nM ACTH stimulation for 6 hours and harvested. A, Harvested media were subjected to ELISA to measure corticosterone release (n = 4). Experimental values were calculated as nanograms corticosterone per milligram protein and normalized to double vehicle-treated cells. B–F, qRT-PCR assessment of changes in expression of Axin2 (B), Cyp11b1 (C), Star (D), Cyp11a1 (E), and Mc2r (F). Values represent the mean ± SD, where vehicle-treated cells were normalized to 1 (n = 4). ****, P < .00005; ***, P < .0005; **, P < .005; *, P < .05; n.s., not significant. G, Protein level changes were assessed by immunoblotting for pCreb (Ser133), total Creb, and β-actin (n = 3, representative data shown).
Figure 4.
Figure 4.
βcat activity affects Sf1 levels, promoter occupancy, and transcriptional activity. ATCL7 cells were pretreated with 0.5μM BIO or vehicle for 12 hours in low-serum media followed by 100nM ACTH stimulation for 6 hours and harvested. A, Changes in Sf1 mRNA levels were assessed by qRT-PCR. Values represent the mean ± SD (n = 4), where vehicle-treated cells were normalized to 1. B, Sf1 protein levels were assessed by immunoblot (n = 4, representative data shown). Quantification of changes in Sf1 protein level is normalized to β-actin levels. C–E, Changes in Sf1 DNA occupancy was measured by ChIP assays using Sf1 antisera. Immunoprecipitates were analyzed by qPCR using primers for the proximal promoters of Star (C), Cyp11a1 (D), and Mc2r (E). Preimmune IgG antisera were used as a negative control. Experimental values are normalized to 2% input (n = 4). *, P < .05. F, 293T cells were transfected with Star-luciferase reporter construct, Renilla-luciferase internal control construct, and empty vector or Sf1 expression construct for 24 hours. Transfected cells were treated with 2μM BIO or vehicle for 24 hours before passive lysis. Luciferase activity was measured and normalized to Renilla luciferase for transfection efficiency. Values represent the mean ± SD (n = 4). Protein levels of active βcat (ABC), Sf1, and β-actin were measured by immunoblotting (lower panels, n = 4, representative data shown). G, 293T cells were transfected with Star-luciferase reporter construct, Renilla-luciferase internal control construct, and empty vector, Sf1, or flag-tagged βcatS33Y alone or in combination for 48 hours before passive lysis. Luciferase activity was measured and normalized to Renilla luciferase for transfection efficiency. Values represent the mean ± SD (n = 4). Protein levels of Flag-βcatS33Y, Sf1, and β-actin were measured by immunoblotting (lower panels, n = 4, representative data shown). H, JEG-3 cells were transfected as in F and stimulated with 1μM BIO or vehicle for 24 hours before passive lysis. I, JEG-3 cells were transfected as in G (n = 4). ***, P < .0005; **, P < .005; *, P < .05.
Figure 5.
Figure 5.
Assessment of the gene expression profile of Wnt-responsive cells in vitro and in vivo. Schematic representation of microarray workflow. Top left, GFP+ adrenocortical cells were obtained from 6-week-old male TCF/Lef:H2B-GFP mice and Sf1:eGFP mice independently via FACS. RNA was extracted from 4 independent sorts per genotype. Using 2-sample t tests, we asked that probe sets give P < .05 and average fold changes of at least 1.5, which selected 636 increased and 907 decreased probe sets in Tcf/Lef:H2B-GFP+ cells, 15% of which were expected to be false-positives based on analysis of datasets where the sample labels were randomly permuted. Top right, ATCL7 cells were treated for 24 hours with 0.5μM BIO or vehicle in low-serum media before harvesting and RNA extraction. Four pairs of treated cells were used for expression profiling. We asked that paired t tests give P < .05 and average fold changes or at least 1.5, which selected 499 upregulated and 746 downregulated probe sets in BIO-treated cells, of which we expect 0.1% to be false-positives based on analysis of datasets where sample labels were permuted within the pairs of samples. The intersection of upregulated probe sets in the 2 experiments was 25 probe sets, representing the 24 distinct genes shown in the heat map on the bottom.
Figure 6.
Figure 6.
Ccdc80 is a novel βcat-regulated gene in adrenocortical cells. A, Section in situ hybridization for Ccdc80 expression in the adult mouse adrenal. B, ATCL7 cells were treated with 0.5μM BIO or vehicle for 18 hours in low-serum media and harvested. Changes in Ccdc80 expression were assessed by qRT-PCR. Values represent the mean ± SD, where vehicle-treated cells were normalized to 1 (n = 4). **, P < .005. C, ATCL7 cells were electroporated with empty vector, βcatS33Y, or dnTCF4E alone or in combination. At 48 hours after transfection, cells were harvested and changes in Axin2 expression (left) and Ccdc80 expression (right) were measured by qRT-PCR. Values represent the mean ± SD (n = 4), where vehicle-treated cells were normalized to 1. D, Diagram of Ccdc80-luciferase reporter construct; −1957 to +814 upstream of the transcription start site of Ccdc80 was cloned into pGL3b luciferase reporter plasmid. Four putative Tcf/Lef binding sites identified using Genomatix MatInspector are labeled #1 to #4. E, 293T cells were transfected with wild-type Ccdc80-luciferase construct or constructs with each individual Tcf binding site mutated (#1–4) and Renilla luciferase internal control construct for 24 hours. Cells were treated with 2μM BIO or vehicle for 24 hours before passive lysis. Luciferase activity was measured and normalized to Renilla luciferase for transfection efficiency. Values represent the mean ± SD (n = 4), where wild-type vehicle-treated cells were normalized to 1. F, ATCL7 cells were electroporated with wild-type Ccdc80-luciferase construct or constructs with each individual Tcf binding site mutated (#1–4) and Renilla luciferase internal control construct for 24 hours. Cells were treated with 0.5μM BIO or vehicle for 24 hours before passive lysis. Luciferase activity was measured and normalized to Renilla luciferase for transfection efficiency. Values represent the mean ± SD, where wild-type vehicle-treated cells were normalized to 1 (n = 4). ***, P < .0005; **, P < .005; *, P < .05; n.s., not significant.
Figure 7.
Figure 7.
Ccdc80 suppresses basal and ACTH-induced zF steroidogenesis in vitro. A, 293T cells were transfected with Ccdc80-myc or empty vector for 24 hours and harvested. Media were collected at the time of harvest. Concentrated CM and cell lysates were subjected to immunoblotting for Myc. B, ATCL7 cells were treated with Mock or Ccdc80-CM derived as in A for 18 hours followed by 100nM ACTH stimulation for 6 hours and harvested. Media collected at harvest were subjected to ELISA to measure corticosterone release (n = 4). Experimental values were calculated as nanograms corticosterone per milligram protein and normalized to untreated cells. C–H, qRT-PCR assessment of changes in expression of Star (C), Cyp11a1 (D), Cyp11b1 (E), Mc2r (F), Sf1 (G), and Axin2 (H) (n = 4). ****, P < .00005; ***, P < .0005; **, P < .005; *, P < .05; n.s., not significant. I, Protein level changes were measured by immunoblotting for pCreb (Ser133), total Creb, Sf1, βcat, and β-actin (n = 3, representative data shown).
Figure 8.
Figure 8.
Integration of Wnt and endocrine signals specify zG vs zF gene expression. The Wnt-responsive population (green) is heterogeneous, consisting of differentiated Cyp11b2-expressing cells (yellow), Shh-producing progenitor cells (red), and few actively proliferating cells (blue) but no differentiated Cyp11b1-expressing cells of the zF (orange). Angiotensin II (AngII) and ACTH promote steroidogenesis in the zG and zF, respectively, defining the differentiated state of the two adrenocortical cell types. Wnt signals stimulate βcat, which inhibits the transcription of zF-associated steroidogenic genes and transcriptionally activates Ccdc80, a secreted protein that in turn also inhibits zF-associated steroidogenic gene expression. The rules that govern the balance the paracrine (Wnt/Ccdc80) and endocrine (ACTH/AngII) signals in this model are predicted to involve temporal and spatial cues. Coupled with the known role of βcat in the regulation of pro-zG genes, these data collectively support a role of Wnt signaling in facilitating the unidirectional centripetal differentiation of adrenocortical progenitor cells. Specifically, Wnt-mediated inhibition of zF-associated steroidogenesis, maintenance of the progenitor cell pool, and priming of zG cell fate assures the differentiation of adrenocortical progenitor cells into Cyp11b2-expressing zG cells before becoming Cyp11b1-expressing zF cells as supported by recent work. See Discussion for details.

Similar articles

See all similar articles

Cited by 26 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback