CARM1 is an important determinant of ERα-dependent breast cancer cell differentiation and proliferation in breast cancer cells

Abstract

Breast cancers with estrogen receptor α (ERα) expression are often more differentiated histologically than ERα-negative tumors, but the reasons for this difference are poorly understood. One possible explanation is that transcriptional cofactors associated with ERα determine the expression of genes which promote a more differentiated phenotype. In this study, we identify one such cofactor as coactivator-associated arginine methyltransferase 1 (CARM1), a unique coactivator of ERα that can simultaneously block cell proliferation and induce differentiation through global regulation of ERα-regulated genes. CARM1 was evidenced as an ERα coactivator in cell-based assays, gene expression microarrays, and mouse xenograft models. In human breast tumors, CARM1 expression positively correlated with ERα levels in ER-positive tumors but was inversely correlated with tumor grade. Our findings suggest that coexpression of CARM1 and ERα may provide a better biomarker of well-differentiated breast cancer. Furthermore, our findings define an important functional role of this histone arginine methyltransferase in reprogramming ERα-regulated cellular processes, implicating CARM1 as a putative epigenetic target in ER-positive breast cancers.

Figure 1

Overexpression of CARM1 inhibits growth and colony formation of MCF7 while exhibiting no effect on MDA-MB-231. (A) The CARM1-overexpressing MCF7 cell line (MCF7-CARM1) grew at a slower rate than the MCF7-vector control as measured by MTT. Error bars denote the standard error deviation (STDEV) (n=3). (B) p21^cip1 and p27^kip1 were expressed at higher levels in MCF7-CARM1 than in MCF7-vector cells in the presence of 10 nM E2. (C) Overexpression of CARM1 inhibited colony formation of MCF7-CARM1 in soft agar in the presence of 10 nM E2. The inset shows colonies under higher amplification in MCF7-vector cells. (D) Neither overexpressing nor knocking-down CARM1 has growth effects on MDA-MB-231. Each assay was performed in triplicate and the error bars represent the STDEV. (E) Western blotting showed that CARM1 was overexpressed in MDA-MB-231-CARM1 and knocked down in MDA-MB-231-shCARM1 cells.

Figure 2

CARM1 inhibits E2-dependent MCF7 cell growth and S-phase entry. MCF7-tet-on-CARM1 (A) and MCF7-tet-on-shCARM1 (B) tetracycline-inducible cell lines were constructed as gain-of-function and loss-of-function cell-based models to modulate the endogenous level of CARM1. CARM1 was detected by Western blotting and mRNA was determined by Q-RT PCR. * indicates statistically significant p value. (C) Inhibition of E2-dependent cell growth by overexpressing CARM1 (+Dox) in MCF7-tet-on-CARM1. Bars denote the STDEV (n=9). (D) Knocking down CARM1 in MCF7-tet-on-shCARM1 (+Dox) did not affect the basal and E2-dependent growth of MCF7 cells in MTT assays. Bars denote the STDEV (n=9). (E) E2 increased the S-phase entry, an event inhibited by CARM1 overexpression in MCF7-tet-on-CARM1 cells. (F) The effects of CARM1 overexpression or knock-down on E2-dependent S-phase entry (p>0.05) measured by BrdU labeling. Error bars indicate STDEV (n=3).

Figure 3

Overexpression of CARM1 induces morphological changes in MCF7 characteristic of a differentiated phenotype. (A) MCF7-CARM1 and MCF7-vector cells exhibit different morphology. (B) DSP1 mRNA level was significantly repressed by E2 treatment and overexpression of CARM1 could reverse E2-mediated DSP1 repression. Error bars indicate STDEV from triplicate experiments. (C) Western blots showed that overexpression CARM1 restored protein levels of E2-repressed E-cadherin and GATA3 at 12 hrs after E2 treatment.

Figure 4

Overexpression of CARM1 modulates E2-dependent gene signature. (A) The ratios of normalized intensities for Dox, E2, or Dox+E2 treated samples vs. that of samples treated with control vehicle (Dox vs DMSO, E2 vs DMSO, and DOX+E2 vs DMSO) were used to demonstrate the activation or repression. Heat map of gene expression calculated as fold changes compared to vehicle indicated that CARM1-induced genes (+Dox) are largely non-overlapping with E2-activated genes. Among CARM1 repressed genes, many are activated by E2 (see blow-up of the heat map), indicating that overexpressing CARM1 can inhibit some E2-activated genes. (B) Pie graphshows that among all E2-upregulated genes, 16% of them are down-regulated by CARM1 overexpression. The bottom of chart shows, among all CARM1-downregualated, E2-upregulated genes, the percentage of genes in each molecular function category.Gene ontology (GO) of the affected genes were determined by GOStat tool online (http://gostat.wehi.edu.au/cgi-bin/goStat.pl),“goa_human” database, where % represent the % of the affected genes that belong to each represented category. (C) Pie graph shows that among all E2 down-regulated genes, 56% of them are up-regulated by CARM1 overexpression. The bottom chart shows, among all CARM1-upregulated, E2-down-regulated genes, the percentage of genes in each molecular function category. (D) Q-RT PCR analyses of p21^cip1, p27^kip1, cyclin G2, MAZ, KRTAP10.12 and GATA-3 expression in MCF7-tet-on-CARM1. Error bars indicate STDEV from three independent experiments.

Figure 5

Knocking-down CARM1 shared common gene signatures with that of E2 treatment in MCF7-tet-on-shCARM1. (A) Heat map showed that CARM1 shRNA (+Dox) activated and repressed genes were largely overlapping with those regulated by E2. A blow-up of the heat map illustrated that many Dox-activated genes (knocking down CARM1) were also activated by E2. (B) Pie graph shows that ~ 65% of CARM1 shRNA activated genes were also activated by E2; ~ 35% of knocking down CARM1 activated genes was E2 non-responsive (labeled as others). Gene ontology illustrated the top affected gene categories by expressing CARM1shRNA or E2 treatment. (C) Pie graph shows that among all down-regulated genes by CARM1 knock-down, 75% of them are E2-repressed genes and 25% of them are E2-non-responsive (labeled as others). The bottom chart shows, among all down-regulated genes by E2 treatment or CARM1 knock-down, the percentage of genes in each molecular function category. (D) Quantitative RT-PCR analyses of ER target genes in MCF7-tet-on-shCARM1. Error bars indicate STDEV.

Figure 6

Knocking-down CARM1 increased E2-dependent tumor growth in xenografted mice. (A) Schematic design of the xenograft study using MCF7-tet-on-shCARM1 cell line. (B) CARM1 mRNA and protein levels were decreased in tumors from Dox recipient mice. CARM1 mRNA in tumors was quantified using Q-RT PCR (vehicle n=7, Dox n=10); protein was visualized by IHC. (C) Increased tumor volume in CARM1 knocked down mice. A representative experiment showed a higher tumor volume with knocking down CARM1 (n=10) compared to E2 alone induced tumors (n=8). The red arrows point to tumors in two representative mice. (D) BrdU and H&E staining of representative tumor samples from vehicle and Dox treated mice. (E) Mitotic index of representative tumor samples from vehicle and Dox treated mice implanted with a high dose E2 pellets. (F) Correlation of p21^cip1 and E-cadherin protein level by IHC.

Figure 7

The expression level of CARM1 positively correlates with ERα level and inversely correlates with tumor grade in ER+ breast tumors. (A) CARM1 expression level is directly correlated with ERα expression level in more than 300 human breast tumor samples using tissue microarrays (TMA). (B & C) CARM1 expression level is higher in ER+, low-grade tumors compared to high grade tumors, supporting a potential link between CARM1 and the differentiation status of ERα-dependent breast tumors. ERα+ve, GRADE 3, 4 (LOW), IHC-SCORE 270, X500 MAGNIFICATION, ERα+VE, GRADE 8, 9 (HIGH), IHC-SCORE 0, X500 MAGNIFICATION.