Recruitment of coregulator G9a by Runx2 for selective enhancement or suppression of transcription

Abstract

Runx2, best known for its role in regulating osteoblast-specific gene expression, also plays an increasingly recognized role in prostate and breast cancer metastasis. Using the C4-2B/Rx2(dox) prostate cancer cell line that conditionally expressed Runx2 in response to doxycycline treatment, we identified and characterized G9a, a histone methyltransferase, as a novel regulator for Runx2 activity. G9a function was locus-dependent. Whereas depletion of G9a reduced expression of many Runx2 target genes, including MMP9, CSF2, SDF1, and CST7, expression of others, such as MMP13 and PIP, was enhanced. Physical association between G9a and Runx2 was indicated by co-immunoprecipitation, GST-pulldown, immunofluorescence, and fluorescence recovery after photobleaching (FRAP) assays. Since G9a makes repressive histone methylation marks and is primarily known as a corepressor, we further investigated the mechanism by which G9a functioned as a positive regulator for Runx2 target genes. Transient reporter assays indicated that the histone methyltransferase activity of G9a was not required for transcriptional activation by Runx2. Chromatin immunoprecipitation assays for Runx2 and G9a showed that G9a was recruited to endogenous Runx2 binding sites. We conclude that a subset of cancer-related Runx2 target genes require recruitment of G9a for their expression, but do not depend on its histone methyltransferase activity.

Fig. 1. Selective positive and negative coregulator effects by G9a on Runx2 target genes

C4-2B/Rx2^dox cells were infected with lentiviral vectors encoding shRNA targeting G9a mRNA (shG9a) or a non-specific sequence (shNS), and the infected populations were selected with puromycin. The two infected cell populations were cultured in medium supplemented with charcoal-stripped serum (CSS) containing dox (250 ng/ml) to induce Runx2 expression or equal volume of vehicle (distilled water) for 24 h before harvest. A: Immunoblots (left) were performed using antibodies against G9a, FLAG epitope (to detect Runx2), and actin as a control. mRNA levels were assessed by qRT-PCR (right). B & C: The mRNA levels for the indicated target genes of Runx2 were analyzed by qRT-PCR as described in Materials and Methods. All experiments were repeated at least three times and the representative results are shown. Abbreviations used: PIP, Prolactin-induced protein; MMP13 and MMP9, Matrix metalloproteinsase-13 and -9, respectively; PGC, Progastricsin-C; CSF-2, Colony-stimulating factor-2; SDF-1, Stromal differentiating factor-1; CST7, Cystatin-7.

Fig. 2. G9a associates with Runx2

A: Cos-7 cells were transfected with expression vectors for FLAG-Runx2 and HA-G9a (500 ng each in 6-well dishes). Cell extracts were immunoprecipitated with FLAG antibodies, and the immunoprecipitates were analyzed by immunoblotting using antibodies against HA epitope. B: GST or GST-G9a on glutathione agarose beads was incubated with in vitro transcribed and translated ³⁵S-labeled Runx2. The bound protein fraction was analyzed by SDS-PAGE and autoradiography.

Fig. 3. Runx2 colocalizes with G9a and enhances its intranuclear mobility

A: C4-2B/Rx2^dox cells were treated for 24 h with dox to induce Runx2 expression or vehicle (veh) as control, and immunofluorescence analysis was performed to detect G9a (green) or Runx2 (red) proteins. DAPI staining (blue) indicates dense chromatin organization in the cell nuclei. Runx2/G9a, overlay of red and green images (yellow) indicates overlap; Runx2/G9a/dapi, overlay of red, green, and blue images. B: Fluorescence recovery after photobleaching of G9a-GFP was assessed as described in Materials and Methods in Cos7 cells transfected with plasmid encoding GFP-G9a alone (blue curve) or together with plasmid encoding Runx2 (red curve).

Fig. 4. Enhancement of Runx2-mediated transcription by G9a in a transient reporter assay

CV1 cells in 12-well plates were transfected with the 6XOSE2-Luciferase reporter plasmid illustrated in A (200 ng/well) alone or together with expression vectors for Runx2 (1 ng) and either HA-tagged G9a full length (G9a) or HA-G9a(H/K) methyltransferase-deficient mutant (H/K) (50, 100 and 200 ng). After transfection the cells were grown for 48 h before they were subjected to luciferase assays (B) and immunoblot analysis using antibodies against HA and actin (C).

Fig. 5. Runx2 recruits G9a to the regulatory elements of its target genes

C4-2B/Rx2^dox cells were plated in 15-cm dishes and cultured in media supplemented with CSS for two days and then treated with dox or vehicle for an additional 16 h before ChIP analysis using antibodies against FLAG to detect Runx2 (A) or against G9a (B). Immunoprecipitated DNA was analyzed with primers (Table 1) designed to amplify the Runx2-occupied regions of the indicated target genes.