Fig 1.
c-Myb-regulates expression of miR-143 and miR-145 in mouse embryonic stem cell-derived VEGFR2+ progenitors and adult mouse carotid vascular smooth muscle cells.
(A) Flow cytometry of cardiovascular-directed d3.75 embryoid bodies (EBs), derived from wild-type (wt) or c-myb-/- (knockout) mouse embryonic stem cells, was used to isolate progenitors with cell surface markers VEGFR2 (V) and PDGFRα (P). The known vascular smooth muscle cell (VSMC) progenitor V+/P- derived from c-myb-/- EB had lower expression levels of miR-143 (p<0.05*) and miR-145 (p<0.01**) than V+/P- derived from wt EB. The non-SMC progenitor cells V-/P+ showed no such difference in miR-143 and miR-145 expression, compared to V+/P-. (B) Primary mouse carotid VSMCs isolated from c-myblx/lx mice have reduced expression levels of c-myb (p<0.05*), as well as miR-143 (p<0.01**) and miR-145 (p<0.001***) as compared to VSMCs isolated from wt mice. (C) Carotid VSMCs isolated from mice homozygous for a hypomorphic c-myb allele (c-mybh/h) have no differences in c-myb expression, but their known reductions in c-Myb activity were associated with reduced expression levels of miR-143 (p<0.01**) and miR-145 (p<0.001***) versus wt. qPCR analysis was performed with n = 3 biological samples and three technical repeats. (D) VSMC from c-myblx/lx mice show reduced expression levels of VSMC marker, Sm22α versus wt. (E) Densitometry reveals c-myblx/lx mice have significantly reduced expression of Sm22a relative to GAPDH compared to wt (p<0.05*, n = 4).
Fig 2.
miR-143/145 promoter activity requires c-Myb binding sites (MBS).
Promoter activity was determined by co-transfection with promoter-luciferase constructs and renilla control plasmid. Site-directed mutagenesis revealed MBS to be important for reporter gene activation when compared to intact proximal promoter. (A) Proximal promoter (PP) with c-Myb binding site (MBS) positions shown relative to predicted transcriptional start. (B) Promoter luciferase constructs were transfected in c-myblx/lx VSMC or wt controls. Compared to proximal promoter, MBS1 (p<0.001+++) and MBS4 (p<0.001+++) were required for c-Myb-dependent transcription in c-myblx/lx VSMC (n = 3), both MSB1 (p<0.01**) and MBS4 (p<0.05*) were also required for transcriptional activity in wt VSMC (n = 3). Importantly, activation in wt was always greater than in c-myblx/lx VSMC (p<0.05). (C) In EOMA, an endothelial cell line that does not express c-Myb, MBS1 (p<0.001***/+++), MBS2 (p<0.05*/+), and MBS4 (p<0.001***/+++) were relevant to transcriptional activity when compared to promoter, regardless of c-Myb expression (n = 12). Co-transfection with exogenous c-Myb is however, associated with increased transcriptional activity (p<0.05, n = 3). (D) The human nasopharyngeal carcinoma cell line (CNE2) is known to express c-Myb. Intact MBS1 (p<0.05*) and MBS4 (p<0.05*, n = 3) were also necessary for miR-143/145 promoter activity in this cell line.
Fig 3.
Chromatin immunoprecipitation (ChIP) of c-Myb on the miR-143/145 promoter of immortalized mouse carotid artery vascular smooth muscle cells (VSMC).
Cells were fixed, lysed and chromatin was fragmented into oligonucleosomes (200-1500bp) then immunoprecipitated overnight at 4°C with rabbit anti-c-Myb antibody (2mg, SC-517X, Santa Cruz), anti-histone H3 (2mg, Santa Cruz) or control rabbit serum (2mg, Santa Cruz). (A) The miR-143 and miR-145 promoter contains four c-Myb binding sites (MBS). (B) ChIP and PCR revealed specificity of c-Myb binding to MBS in carotid VSMC. As a negative control, ChIP was performed on a promoter situated ~15kb upstream of miR-143/145 with no known or predicted MBS. (C) qPCR reveals differential binding of c-Myb on predicted MBS of the miR-143/145 promoter. c-Myb preferentially binds MBS2 and MBS4 in carotid VSMC (p<0.01, one-way ANOVA). When normalized to histone H3 density, c-Myb preferentially binds MBS4 (p<0.001***, one-way ANOVA). ChIP was performed in n = 3 biological samples. qPCR analysis had three technical repeats for each biological sample.
Fig 4.
c-Myb represses Elk-1 expression via miR-143/145.
(A) Elk1 has previously been validated as a miR-143/145-repressed gene. Elk1-3’UTR-Luciferase construct was transfected into wt and c-myblx/lx carotid artery VSMC. Luciferase activity was assayed on PHERASTAR and found to be augmented in c-myblx/lx VSMC, which have known reductions in c-Myb expression (p<0.01**). (B) Time course analysis of c-Myb and Elk-1 expression. After serum starvation and cell cycle arrest, c-Myb levels are diminished. Following addition of 10% fetal bovine serum, c-Myb expression peaks at 12 h. As c-Myb expression level rise, Elk-1 expression levels decrease and are undetectable at 12 h. Note that c-Myb expression is not detectable in c-myblx/lx VSMC released from serum starvation, and Elk-1 expression remains stable. (C) miR-143 mediates c-Myb repression of Elk1. The miR-143 antagomir augments Elk-1 protein expression in wt cells that express c-Myb, while the mimic abrogates protein expression in c-myblx/lx VSMC.