Glucocorticoid Receptor:MegaTrans Switching Mediates the Repression of an ERα-Regulated Transcriptional Program

Abstract

The molecular mechanisms underlying the opposing functions of glucocorticoid receptors (GRs) and estrogen receptor α (ERα) in breast cancer development remain poorly understood. Here we report that, in breast cancer cells, liganded GR represses a large ERα-activated transcriptional program by binding, in trans, to ERα-occupied enhancers. This abolishes effective activation of these enhancers and their cognate target genes, and it leads to the inhibition of ERα-dependent binding of components of the MegaTrans complex. Consistent with the effects of SUMOylation on other classes of nuclear receptors, dexamethasone (Dex)-induced trans-repression of the estrogen E₂ program appears to depend on GR SUMOylation, which leads to stable trans-recruitment of the GR-N-CoR/SMRT-HDAC3 corepressor complex on these enhancers. Together, these results uncover a mechanism by which competitive recruitment of DNA-binding nuclear receptors/transcription factors in trans to hot spot enhancers serves as an effective biological strategy for trans-repression, with clear implications for breast cancer and other diseases.

Keywords: enhancers; nuclear receptors; transcriptional regulation; transrepression.

Figure 1. Dex significantly repress the E₂-mediated activation of target genes and enhancers

(A). E₂+Dex treatment of MCF7 cells significantly represses E₂-activated genes (TFF1, FOXC1) and eRNAs expression. qRT-PCR results are presented as mean ± SD. N≥3, two-tailed Student’s t test. (B). Boxplot showing normalized GRO-seq tags (Log2) for E₂-induced genes under different treatment conditions (ICI, ICI+DEX, E₂, E₂+Dex) in MCF7 cells. (P value is calculated by Wilcoxon rank sum test) (C). Boxplots showing normalized GRO-seq tags (Log2) for eRNAs transcribed at active enhancers under different treatment conditions (ICI, ICI+DEX, E₂, E₂+Dex) in MCF7 cells. Sense and antisense eRNA transcripts are shown separately. (P value is calculated by Wilcoxon rank sum test). (D) Genome browser image showing normalized GRO-seq tag counts in MCF7 cells under different ligands treatment (ICI, ICI+DEX, E₂, E₂+Dex). TFF1 and FOXC1 loci are shown. “*” indicates the position of primers used for qPCR detection. See also Figure S1

Figure 2. GR is recruited in trans on the E₂-activated enhancers following E₂+Dex treatment

(A). ChIP-seq tag density profile plot showing ERα binding on 423 ERα-activated enhancers upon different ligands treatment (ICI, ICI+DEX, E₂, E₂+Dex). The center of the plot is based on the center of ERα binding. (B). ChIP-seq tag density profile plot showing the binding of GR to 423 ERα-activated enhancers with different ligands treatment (ICI, ICI+DEX, E₂, E₂+Dex). The center of the plot is based on the center of ERα binding. (C) De novo motif analysis of GR binding sites based on the 423 ERα-activated enhancers. (D) Heatmaps of ChIP-seq tag counts for 423 ERα-activated enhancers occupied by GR wild type or pBox mutant in MCF7 cells treated with E₂+Dex. (E) The interaction of GR with ERα is dependent on its DNA-binding domain (DBD), as shown by coimmunoprecipitation using HA-tagged WT or DBD deleted-GR. (F) ChIP-qPCR showing GR, GATA3 and RARα binding on ERα-activated enhancers (TFF1 and FOXC1 enhancers) in GR wild type and GR DBD deletion stable MCF7 cells upon E₂+Dex treatment. Data are presented as mean ± SD. N≥3, two-tailed Student’s t test. See also Figure S2

Figure 3. GR inhibits the assembly of the MegaTrans complex on E2-activated enhancers

(A) ChIP-qPCR showing the binding of MegaTrans components (GATA3, RARα, AP2ɣ), ERα and GR on E₂-activated TFF1 and FOXC1 enhancers in MCF7 cells treated with ICI, ICI+DEX, E₂, E₂+DEX. Data are presented as mean ± SD. N≥3, two-tailed Student’s t test. TFF1 e: TFF1 enhancer; FOXC1 e: FOXC1 enhancer. (B) ChIP-qPCR showing FoxA1 and P300 binding on ERα-activated enhancers in MCF7 cells treated with ICI, ICI+DEX, E₂, E₂+DEX. Data are presented as mean ± SD. N≥3, two-tailed Student’s t test. (C) ChIP-seq tag density profile plot (centered on ERα binding peaks in E₂ condition) showing the binding of GATA3 on 423 ERα-activated enhancers under different treatment conditions (ICI, E₂ or E₂+Dex). (D) ChIP-reChIP qPCR analysis showing that GR and MegaTrans complex (exemplified by GATA3, RARα, AP2ɣ) could not co-exist on the ERα-activated enhancers in MCF7 cells treated with E₂ or E₂+Dex. The GR Re-ChIP was done after E₂+Dex treatment, the RARα Re-ChIP was done after E₂ treatment. ChIP signals are presented as percentage of input. Data are shown as mean ± SD. N≥3, two-tailed Student’s t test. N.D, not detectable. See also Figure S3

Figure 4. The ability of GR to bind in trans on E2-activated enhancers depends on its SUMOylation status

(A) Western blot analysis showing immunoprecipitated wild type GR or SUMOylation mutants in MCF7 cells using different SUMO-specific antibodies upon treatment with E₂+Dex. (GR-WT:GR wild type; GR-1KR: GR C-terminal K703R mutation; GR-2KR: GR N-terminal K277R and K293R two sites mutation; GR-3KR: GR K277R, K293R and K703R all three sites mutation, SUMO2/3 PC: HA-SUMO2/3 protein as positive control) (B) RT-qPCR of GR target genes in GR knockout MCF7 cells upon over-expression of GR wild type or SUMOylation mutants. Fold change of gene expression is presented as comparison of E₂+Dex versus E₂ treatments. Data are presented as mean ± SD. N≥3, two-tailed Student’s t test. (C) ChIP-qPCR showing GR wild type and SUMOylation mutants binding on GR cis or trans-binding enhancers in MCF7 cells upon E₂+Dex treatment. Data are presented as mean ± SD. N≥3, two-tailed Student’s t test. See also Figure S4

Figure 5. The recruitment of the NCoR/SMRT complex is required for GR-mediated repression on E2-activated genes

(A) ChIP-qPCR showing GR, GATA3 and RARα binding on ERα-activated enhancers (TFF1 and FOXC1 enhancers) following siRNA-mediated knock-down of NCoR/SMRT complex in MCF7 cells treated with E₂ or E₂+Dex. ChIP signals are presented as percentage of input. Data are represented as mean ± SD. N≥3, two-tailed Student’s t test. TFF1 e: TFF1 enhancer; FOXC1 e: FOXC1 enhancer. (B) RT-qPCR showing the effect of DEX treatment on ERα-activated genes (TFF1, FOXC1) following siRNA-mediated knock down of NCoR/SMRT complex in MCF7 cells upon E₂ or E₂+Dex treatment. The gene expression changes are shown as fold change upon E₂+Dex treatment vs E₂ stimulation. Data are represented as mean ± SD. N≥3, two-tailed Student’s t test. (C) ChIP-qPCR showing NCoR/SMRT complex binding on ERα-activated enhancers upon E₂+Dex treatment compared with E₂ treatment in MCF7 cells. ChIP signals are presented as percentage of input. Data are represented as mean ± SD. N≥3, two-tailed Student’s t test. See also Figure S5

Figure 6. Working model for Dex repressing ERα-activated genes expression