The mediator complex subunit 1 enhances transcription of genes needed for adrenal androgen production

Abstract

There are three enzymes involved in the biosynthesis of the adrenal androgen dehydroepiandrosterone (DHEA) sulfate. Cholesterol side-chain cleavage (CYP11A1) and 17alpha-hydroxylase/17,20-lyase (CYP17) metabolize cholesterol into DHEA, whereas steroid sulfotransferase family 2A1 (SULT2A1) is responsible for conversion of DHEA to DHEA sulfate. We previously examined the mechanisms regulating CYP11A1, CYP17, and SULT2A1 transcription and found that each is regulated, in part, by the transcription factor GATA-6. Previous studies suggested that mediator complex subunit 1 (MED1, also called PPARBP or TRAP220) is a cofactor involved in not only the regulation of nuclear receptors but also the activation of GATA-6 transcription. Herein we demonstrated a role for MED1 in the regulation of CYP11A1, CYP17, and SULT2A1 transcription. Transient transfection assays with SULT2A1 deletion and mutation promoter constructs allowed the determination of specific the GATA-6 binding cis-regulatory elements necessary for transactivation of SULT2A1 transcription. Binding of MED1 and GATA-6 was confirmed by coimmunoprecipitation/Western analysis and chromatin immunoprecipitation assay. We demonstrated expression of MED1 mRNA and protein in the human adrenal and determined that knockdown of MED1 expression via specific small interfering RNA attenuated CYP11A1, CYP17, and SULT2A1 expression levels in H295R cells. In addition, we demonstrated that MED1 enhanced GATA-6 stimulated transcription of promoter constructs for each of these genes. Moreover, the activity of MED1 for SULT2A1 promoter was mediated by GATA-6 via the -190 GATA-binding site. These data support the hypothesis that MED1 and GATA-6 are key regulators of SULT2A1 expression, and they play important roles in adrenal androgen production.

Figure 1

Quantification of MED1 transcript levels in human tissues and H295R cells. Levels of MED1 mRNA were compared in human adult adrenal gland, brain, heart, kidney, liver, lung, premenopausal ovary, salivary gland, skeletal muscle, thymus, and H295R cells using qPCR and normalized to 18s rRNA. Data points are expressed as the fold over the average expression levels seen in the brain (19). The level of MED1 transcript was significantly higher in H295R cells compared with any types of human tissues that were examined (P < 0.05). However, the expression level of MED1 mRNA was not significantly different among human tissues examined.

Figure 2

Immunohistochemical localization of MED1 in the human adrenal gland. MED1 immunoreactivity was detected in the nuclei of rat liver cells (positive control) (A). As a negative control, primary antibody was replaced with buffer only, and no specific immunoreactivity was detected in liver cells (B). MED1 immunoreactivity was also detected in the nuclei of cortical cells of the adrenal cortex (C). When primary antibody was replaced with buffer alone, there was no specific immunoreactivity (D). Immunohistochemical analysis of GATA-6 (E) and SULT2A1 (F) in human adult adrenal gland. Bar, 10 μm.

Figure 3

Effects of siRNA depletion of MED1 on the transcript levels of adrenal steroidogenic enzymes. A, H295R adrenal cells were transfected with or without siRNA against MED1 (MED1 siRNA) or Stealth RNAi (negative control; NTC). After 48 h, mRNA and protein for MED1 were detected by qPCR and Western analyses, respectively. 18S rRNA and β-actin protein were used for normalization. Data are presented as mean ± se for three independent experiments. *, P < 0.05. B, Transcript levels for adrenal steroidogenic enzymes were determined using qPCR. StAR, CYP11A1, HSD3B2, CYP17, CYP21, and SULT2A1 mRNAs were examined 48 h after H295R cells were transfected with or without siRNA against MED1 (MED1 siRNA) or Stealth RNAi (NTC). Data are presented as mean ± se for three independent experiments. *, P < 0.05. C and D, Production of DHEA (C) and DHEAS (D) by H295R cells 72 h after MED1 siRNA transfection. To examine the capacity to produce DHEA and/or DHEAS, cells were treated with MED1 (MED1 siRNA) or Stealth RNAi (NTC). After this treatment, the cells were incubated for 6 h with 10 μm 22(R)-hydroxycholesterol. Data are presented as mean ± se of values from three independent experiments and expressed as a percent of NTC. *, P < 0.05, compared with NTC.

Figure 4

Comparison of the effects of MED1 and GATA-6 or SF-1 on the transcriptional activity of CYP11A1 (A), CYP17 (B), or SULT2A1 (C) reporter gene activity after transfection with reporter gene constructs. Luciferase promoter constructs containing CYP11A1, CYP17, or SULT2A1 (1 μg/well) was cotransfected in H295R cells with MED1 expression plasmid (0.1 μg/well) and GATA-6 plasmid (0.1 μg/well). H295R cells were lysed and assayed for luciferase activity 24 h after transfection. Data were normalized to cotransfected β-galactosidase expression vector, and data are expressed as fold induction over basal reporter. Results represent the mean ± se of data from at least three independent experiments, each performed in triplicate. *, P < 0.05, compared with basal level; †, P < 0.05, compared with only GATA-6-transfected group.

Figure 5

Comparison of GATA-6 isoforms on MED1-enhanced SULT2A1 gene transcription. A, Coimmunoprecipitation with nuclear extract of H295R cells using a MED1 antibody. The complex of cell lysates and MED1 antibody (IP) allowed detection of distinct bands for both short (52 kDa) and long (64 kDa) GATA-6 isoforms, which were also seen in the input sample (input). The negative control (NTC) was prepared with nuclear extract of H295R cells and rabbit IgG instead of MED1 antibody. In vitro-prepared short isoform (GATA-6 short) and long isoform (GATA-6 long) represented lysates obtained from H295R cells that had elevated expression of each isoforms after transient transfection with the respective expression vector and acted as positive controls. B–D, In vitro-prepared short isoform (GATA-6 short), long (GATA-6 long) isoform, and both short and long isoforms (GATA-6 short/long) proteins were also included as positive controls. MED1 effects on SULT2A1 reporter gene activity with GATA-6 short (B), GATA-6 long (C), or GATA-6 short/long (D) isoforms are shown. Luciferase promoter constructs containing SULT2A1 (1 μg/well) were cotransfected in H295R cells with MED1 expression plasmid (0.1 μg/well) and GATA-6 short (B), GATA-6 long (C), or GATA-6 short/long (D) isoforms plasmid (0.1 μg/well). Cells were lysed and assayed for luciferase activity 24 h after transfection. Data were normalized to cotransfected β-galactosidase expression vector, and data shown are expressed as the fold induction over basal reporter. Results represent the mean ± se of data from at least three independent experiments, each performed in triplicate. *, P < 0.05, compared with basal level; †, P < 0.05, compared with only GATA-6-trasfected group.

Figure 6

The role of the −190 GATA binding cis-element in the regulation of SULT2A1 transcription. A, A schematic representation of SULT2A1 promoter with potential GATA binding sites. Triangles represent potential binding sites, and the numbers below represent the base pair at which the site begins (based on the translational start site). B, Deletion and mutation analysis in HEK293T cells. A series of pGL3 reporter constructs containing progressively smaller amounts of SULT2A1 5′-flanking DNA (1 μg/well) were cotransfected in HEK293T cells with MED1 expression plasmid (0.1 μg/well) and GATA-6 plasmid (0.1 μg/well). Data were normalized to cotransfected β-galactosidase expression vector. Results represent the mean ± se of data from at least three independent experiments, each performed in triplicate. *, P < 0.05, compared with basal level; †, P < 0.05, compared with the GATA-6 only transfection group. C, ChIP assay using H295R nuclear lysate demonstrated a distinct band corresponding to the −190 site of the SULT2A1 promoter (input). Immunoprecipitation with either GATA-6 or MED1 antibodies showed that both proteins were associated with the −190 site in the SULT2A1 promoter. ChIP assay was performed as described in Materials and Methods.