miR-29c induction contributes to downregulation of vascular extracellular matrix proteins by glucocorticoids

Abstract

Maternal undernutrition increases maternal glucocorticoids (GCs) and alters microRNA expression in offspring. Given that the mechanisms of GC action on vascular development are not clear, this study examined the influence of GCs on microRNA 29c (miR-29c) and its predicted targets in primary rat aorta smooth muscle cells (RAOSMCs). Dexamethasone (Dex) and corticosterone (Cor) time-dependently increased miR-29c expression and reduced collagen type III (Col3A1), collagen type IV (Col4A5), elastin (ELN), and matrix metalloproteinase-2 (MMP2) protein in RAOSMCs. These effects were blocked by mifepristone. These genes were also targeted by miR-29c, as confirmed by a significant decrease in luciferase reporter activity of Col3A1 (34%), Col4A5 (45%), ELN (17%), and MMP2 (28%). In cells transfected with reporter plasmids, including the 3'-untranslated region of genes targeted by miR-29c, treatment with Dex or Cor also resulted in decreases in luciferase activity. Gain or loss of function of miR-29c significantly altered mRNA expression of Col3A1 (26% and 26%, respectively), Col4A5 (28% and 32%, respectively), and MMP2 (24% and 14%, respectively) but did not affect ELN. Gain or loss of function of miR-29c also significantly altered protein levels of Col3A1 (51% and 16%, respectively), Col4A5 (56% and 22%, respectively), ELN (53% and 71%, respectively), and MMP2 (28% and 53%, respectively). Coincubation of anti-miR-29c with Dex or Cor partially attenuated the effects of these steroids on protein expression of Col3A1 (25% and 24%, respectively), Col4A5 (26% and 44%, respectively), ELN (31% and 55%, respectively), and MMP2 (46% and 26%, respectively) in RAOSMCs compared with anti-miR negative controls. Our results demonstrate that GCs regulate the expression of Col3A1, Col4A5, ELN, and MMP2, at least in part, through induction of miR-29c.

Keywords: corticosterone; dexamethasone; microRNA; smooth muscle.

Fig. 1.

A: transfection of Cy3-labeled negative control oligonucleotides (50 nM) in rat aortic smooth muscle cells (RAOSMCs) for 48 h demonstrated widespread transfection, as revealed by Cy3 fluorescence (top) and phase contrast (bottom) images. Similar results were obtained from 3 sets of independent experiments. B: micro-RNA 29c (miR-29c) levels in RAOSMCs increased dramatically after transfection for 96 h with pre-miR-29c oligonucleotides (50 nM) or negative control (NC). snRNA, small nuclear RNA. Values are means ± SE for 3 independent experiments. *P < 0.05 vs. NC. C: miR-29c levels in RAOSMCs after transfection of anti-miR-29c oligonucleotides (50 nM) decreased significantly relative to anti-miR-29c NC (aNC). Values are means ± SE for 3 independent experiments. *P < 0.05 vs. aNC.

Fig. 2.

A: immunoblotting reveals that dexamethasone (Dex, 0.1 μM) and corticosterone (Cor, 10 μM) produced time-dependent decreases in protein abundances of collagen type III (Col3A1), collagen type IV (Col4A5), elastin (ELN), and matrix metalloproteinase-2 (MMP2) in primary cultures of RAOSMCs. All the proteins assayed are predicted targets of miR-29c. Multiple gels were used to resolve the proteins; each assay included GAPDH as a loading control; one representative GAPDH blot is shown for the group of immunoblots. DMSO, treatment with 0.2% DMSO only at 0 h. B: protein expression levels were quantified after normalization relative to their corresponding GAPDH control. DMSO, treatment with 0.2% DMSO only at 0 h. Values are means ± SE. *P < 0.05 vs. DMSO (control). C: glucocorticoid (GC) receptor antagonist mifepristone (MIF) at 1 μM blocked effects of 48 h of treatment with 0.1 μM Dex and 10 μM Cor compared with 48 h of treatment with 0.2% DMSO alone on protein levels for Col3A1, Col4A5, ELN, and MMP2 in RAOSMCs. GAPDH was used as a loading control, and all protein levels are shown relative to their appropriate GAPDH loading control. D: protein expression levels after 48 h of treatment were compared following normalization relative to their GAPDH control. Values are means ± SE from 3 sets of independent experiments. *P < 0.05 vs. DMSO. #P < 0.05 vs. Dex or Cor. E: after 24 or 48 h of treatment of primary RAOSMCs with 0.2% DMSO, 0.1 μM Dex, or 10 μM Cor, conditioned medium was collected and analyzed via immunoblotting for the presence of Col3A1, Col4A5, and ELN. A Ponceau S-stained protein band on the nitrocellulose membrane was used as a loading control. Blots are representative of results from 3 sets of independent experiments. ELN levels were undetectable at 24 and 48 h. F: treatment of primary cultures of RAOSMCs with 0.1 μM Dex and 10 μM Cor for 12 h decreased mRNA levels of Col3A1, Col4A5, ELN, and MMP2. The 18S levels in each sample were used to normalize mRNA abundance. Values are means ± SE from 3 sets of independent experiments. *P < 0.05 vs. DMSO.

Fig. 3.

A: treatment of primary cultures of RAOSMCs with 0.1 μM Dex and 10 μM Cor produced a time-dependent increase in miR-29c abundance. Level of miR-29c was normalized relative to U6 snRNA. B: Dex and Cor stimulated dose-dependent effects on miR-29c expression at 48 h in RAOSMCs. C: treatment with 1 μM mifepristone blocked effects of 0.1 μM Dex and 10 μM Cor on miR-29c abundance. Values are means ± SE from 3 independent experiments. *P < 0.05 vs. control. #P < 0.05 vs. Dex and #P < 0.05 vs. Cor.

Fig. 4.

A: sequence alignments and coordinated positions of miR-29c seed regions with its target sites in the 3′-untranslated region (UTR) of Col3A1, Col4A5, ELN, and MMP2. B: for luciferase assay, RAOSMCs were transfected with firefly luciferase reporter constructs containing a 3′-UTR fragment of Col3A1, Col4A5, ELN, or MMP2. Cells were also cotransfected with pre-miR-29c (50 nM) or pre-miR NC. Ratio of firefly to Renilla was determined after 48 h and reported as relative luciferase activity normalized to NC values. C: Dex (0.1 μM) and Cor (10 μM) treatment for 48 h in RAOSMCs decreased activity of firefly luciferase reporter constructs containing a 3′-UTR fragment of Col3A1, Col4A5, ELN, or MMP2. Ratio of firefly to Renilla was determined and reported as relative luciferase activity compared with NC. Values are means ± SE from 3 sets of independent experiments. *P < 0.05 vs. control.

Fig. 5.

Relative to effects of a control transfection (NC), transfection of RAOSMCs with 50 nM pre-miR-29c (miR-29c) significantly depressed mRNA levels for Col3A1, Col4A5, and MMP2. Conversely, transfection with anti-miR-29c increased mRNA levels for Col3A1, Col4A5, and MMP2 relative to control (aNC). Values are means ± SE for 3 independent experiments. *P < 0.05 vs. control.

Fig. 6.

A: relative to effects of a control transfection (NC), a 96-h transfection of RAOSMCs with 50 nM pre-miR-29c (miR-29c) significantly depressed protein levels for Col3A1, Col4A5, ELN, and MMP2. Immunoblots were prepared from 3 separate gels, and GAPDH was used as a loading control for each. B: when normalized relative to their corresponding GAPDH controls, all protein abundances were depressed significantly by transfection with pre-miR-29c. C: transfection with anti-miR-29c increased protein levels for Col3A1, Col4A5, ELN, and MMP2 relative to control (aNC). GAPDH was used as a loading control. D: when normalized relative to their corresponding GAPDH controls, all protein abundances were increased significantly by transfection with anti-miR-29c. Values are means ± SE from 3 independent experiments. *P < 0.05 vs. control.

Fig. 7.

A: protein expression of Col3A1, Col4A5, ELN, MMP2, and cyclooxygenase 2 (COX2) were determined after transfection of anti-miR-29c (anti-miR-29c) or NC (aNC) for 96 h and treatment with 0.1 μM Dex or 10 μM Cor for the last 48 h. COX2 was used as a NC that was responsive to GC, but not miR-29C. GAPDH was used as loading control. B: when protein abundances were normalized relative to their corresponding GAPDH controls, values were significantly decreased by treatment with Dex or Cor. C: inhibition of protein expression by Dex was calculated relative to changes in corresponding DMSO controls in cells transfected with aNC or anti-miR-29c. ANOVA indicated that anti-miR-29c transfection significantly attenuated Dex-induced inhibition for all 4 proteins. Post hoc comparisons further revealed that magnitudes of Dex-induced inhibition were decreased significantly by anti-miR-29c for each individual protein. D: inhibition of protein expression by Cor was calculated as described for Dex. ANOVA indicated that anti-miR-29c transfection significantly attenuated Cor-induced inhibition for all 4 proteins. Post hoc comparisons revealed that magnitudes of Cor-induced inhibition were decreased significantly by anti-miR-29c for Col4A5 and ELN. Values are means ± SE from 3 independent experiments. *P < 0.05 vs. control (aNC, DMSO). #P < 0.05 vs. aNC/Dex or aNC/Cor.