Caveolin-1 expression by means of p38beta mitogen-activated protein kinase mediates the antiproliferative effect of carbon monoxide

Abstract

During vascular injury, the proliferation and migration of smooth muscle cells leads to characteristic neointima formation, which can be exacerbated by genetic depletion of caveolin-1 or heme oxygenase 1 (HO-1), and inhibited by carbon monoxide (CO), a by-product of heme oxygenase 1 activity. CO inhibited smooth muscle cell proliferation by activating p38 mitogen-activated protein kinase (MAPK) and p21(Waf1/Cip1). Exposure to CO increased caveolin-1 expression in neointimal lesions of injured aorta and in vitro by activating guanylyl cyclase and p38 MAPK. p38beta-/- fibroblasts did not induce caveolin-1 in response to CO, and exhibited a diminished basal caveolin-1 expression, which was restored by p38beta gene transfer. p38beta MAPK down-regulated extracellular signal-regulated protein kinase 1/2 (ERK-1/2), which can repress caveolin-1 transcription. Genetic depletion of caveolin-1 abolished the antiproliferative effect of CO. Thus, we demonstrate that CO, by activating p38beta MAPK, up-regulates caveolin-1, which acts as a tumor suppressor protein that mediates the growth inhibitory properties of this gas.

Fig. 1.

Inhibition of cell growth by CO in vitro.(A) SMC were serum-starved overnight, and then stimulated by addition of serum (10% FBS) to the culture to initiate cell growth. Cellular proliferation was assessed by measuring [³H]thymidine incorporation [counts per min (cpm) n = 3] after serum stimulation in the absence or presence of CO (250 ppm for 24 h). (B) The relative expression levels of cell cycle-related proteins p21^Waf1/Cip1, and cyclin A, were determined in SMC after CO exposure for the indicated times by Western blot analysis. (C) SMC were pretreated with CO (1 h) followed by addition of SB 203580 (10 μM) (1 h), a specific p38α/β MAPK inhibitor, followed by CO exposure in the absence or presence of SB for an additional 24 h. Expression of p21 and cyclin A was determined by Western blot analysis.

Fig. 2.

Reciprocal effects of CO and growth factor on the expression of caveolin-1 and cell cycle-related proteins. (A and B) Serum-starved SMC were treated with PDGF (10 ng/ml) for 24 h in the absence or presence of CO (250 ppm), and cellular proliferation was assessed by direct cell counting. (B) Total cell lysates were probed with caveolin-1 antibody, or antibodies to p21 and cyclin A. β-actin served as the loading control. The results represent three independent experiments.

Fig. 3.

CO increases caveolin-1 expression in vivo. LEW rats were subjected to balloon injury of the carotid artery, as described in Materials and Methods. The rats were exposed to either room air (A, B, and E–G) or CO (C, D, and H–J) before and during balloon or sham injury. Carotid arteries were analyzed by immunohistochemical staining 21 days after balloon angioplasty. Tissue samples were stained with a modified elastic tissue-Masson's trichrome stain for elastic fiber (A and C). Sections were immunostained with anti-α-SMA for the detection of SMC (B and D), or anti-caveolin-1 (E–G and H–J), which was detected in both SMC and endothelial cells (G and J). [Scale bar: 25 μm (E, F, H, and I) or 10 μm (G and J).] Samples shown are representative of three to six animals analyzed per group. Tissue sections corresponding to sham or balloon injury in the absence or presence of CO were quantified for intimal thickness, as well as the ratio of intima to media (K). Results represent the mean ± SD for three independent determinations (arbitrary units). *, P < 0.05.

Fig. 4.

Down-regulation of caveolin-1 abolishes the antiproliferative effect of CO. (A)[³H]Thymidine incorporation assay was conducted in SMC transfected with siRNA for caveolin-1 or control siRNA (Dharmacon) and treated with CO (250 ppm) or room air. Cell lysates were assayed for caveolin-1 expression in the absence or presence of CO. (B) SMC were transfected with siRNA for caveolin-1 or control siRNA (48 h), followed by an additional 24 h in the absence of presence of CO. Cell lysates were analyzed for the expression of p21, and cyclin A by Western immunoblotting. β-actin served as loading control. Results represent at three independent experiments. (C) SMC were transfected with adenoviral constructs containing HO-1 or LacZ cDNA for 48 h, followed by incubation in the absence or presence of CO (250 ppm) or SnPP for an additional 24 h before determination of cell proliferation by [³H]thymidine incorporation. Proliferation data represent three independent experiments using triplicate wells. *, P < 0.05; ns, not significant (A and C).

Fig. 5.

The stimulation of caveolin-1 expression by CO requires cGMP and p38 MAPK-dependent pathways. (A) SMC were pretreated with CO (1 h) followed by addition of SB 203580 (10 μM) (1 h), a specific p38α/β MAPK inhibitor, followed by continuous CO exposure in the absence or presence of SB for an additional 24 h. The effect of CO on caveolin-1 expression was determined by Western blot analysis. (B) Serum-starved SMC were treated with PDGF (10 ng/ml) for the indicated times in the absence or presence of CO. Total cell lysates were immunoblotted with anti-phospho p38 MAPK antibody. Total p38 MAPK served as a loading control. (C) SMC were treated with a cGMP analog, 8-Br-cGMP (20 μM) for the indicated times, or with the indicated concentrations for 24 h. Total cell lysates were immunoblotted with anti-phospho p38 MAPK antibody, or anti-caveolin-1. Total p38 MAPK or caveolin-2 served as loading controls. (D) The effect of CO on caveolin-1 expression was determined in SMC in the absence or presence of ODQ, a specific guanylate cyclase inhibitor.

Fig. 6.

p38β MAPK inhibits cellular proliferation, by increasing p21^Waf1/Cip1 and decreased cyclin A expression. (A–E) Fibroblasts were isolated from the lungs of wild-type or p38β-null mice (A–D) or from caveolin-1-null mice (E). (A) Growth curves of wild-type and p38β^–/– fibroblasts in the absence of CO (250 ppm) were determined by direct cell counting (initial cell number: 0.5 × 10⁵ cells per ml). Data represent the mean ± SD of four to six experiments. (B) Wild-type or p38β^–/– fibroblasts were treated with air or CO (250 ppm) for the indicated times (0–72 h) and assessed for caveolin-1 expression by Western analysis with anti-caveolin-1 antibody. β-actin served as loading control. Results represent at least three independent experiments. (C) Wild-type or p38β^–/– fibroblasts were treated with PDGF in the absence or presence of CO and assayed for ERK1/2 phosphorylation at the indicated times by Western analysis. (D) Wild-type and p38β^–/– fibroblasts were transfected with a p38β MAPK expression vector or control vector for 48 h (Left) or incubated for a further 24 h in the absence or presence of CO (250 ppm) (Right). Cell lysates were immunoblotted with anti-caveolin-1 antibody. β-actin served as loading control. (E) Fibroblasts derived from caveolin-1-null mice (cav^–/–) or wild-type mice were serum-starved overnight, and then stimulated by addition of serum (10% FBS) to the culture to initiate cell growth. Cellular proliferation was assessed by measuring [³H]thymidine incorporation after serum stimulation in the absence or presence of CO (250 ppm for 24 h).