PDGF-induced proliferation in human arterial and venous smooth muscle cells: molecular basis for differential effects of PDGF isoforms

Abstract

Platelet-derived growth factor (PDGF) has been implicated in the pathogenesis of arterial atherosclerosis and venous neointimal hyperplasia. We examined the effects of PDGF isoforms on smooth muscle cells (SMCs) from arterial and venous origins in order to further understand the differential responsiveness of these vasculatures to proliferative stimuli. Serum-starved human arterial and venous SMCs exhibited very different proliferative responses to PDGF isoforms. Whereas, proliferation of arterial SMCs was strongly stimulated by PDGF-AA, venous SMCs showed no proliferative response to PDGF-AA, but instead demonstrated a significantly greater proliferative response to PDGF-BB than arterial SMCs. Part of this difference could be attributed to differences in PDGF receptors expression. There was a 2.5-fold higher (P < 0.05) density of PDGF receptor-α (PDGF-Rα) and a 6.6-fold lower (P < 0.05) density of PDGF-Rβ expressed on arterial compared to venous SMCs. Concomitant with an increased proliferative response to PDGF-AA in arterial SMCs was a marked PDGF-Rα activation, enhanced phosphorylation of ERK1/2 and Akt, a transient activation of c-Jun NH2-terminal kinase (JNK), and a significant reduction in expression of the cell-cycle inhibitor p27(kip1). This pattern of signaling pathway changes was not observed in venous SMCs. No phosphorylation of PDGF-Rα was detected after venous SMC exposure to PDGF-AA, but there was enhanced phosphorylation of ERK1/2 and Akt in venous SMCs, similar to that seen in the arterial SMCs. PDGF-BB stimulation of venous SMC resulted in PDGF-Rβ activation as well as transactivation of epidermal growth factor receptor (EGF-R); transactivation of EGF-R was not observed in arterial SMCs. These results may provide an explanation for the differential susceptibility to proliferative vascular diseases of arteries and veins.

Fig. 1

PDGF-induced proliferation of arterial smooth muscle cells (ASMC, open circles; n = 6) and venous smooth muscle cells (VSMC, closed circles; n = 6), as assessed by the bromo-deoxyuridine incorporation assay. All three PDGF isoforms significantly stimulated the proliferation of arterial SMC. In contrast, PDGF-AA exerted no mitogenic effect on venous SMC (A), although PDGF-AB (B), or PDGF-BB (C) stimulated greater DNA synthesis in venous SMCs at high concentrations (50 and 100 ng/ml), compared to arterial SMCs. ^*P <0.05, arterial SMCs compared to venous SMCs.

Fig. 2

Expression of PDGF receptor isoforms (PDGF-Rα and PDGF-Rβ) on arterial smooth muscle cells (ASMCs, open bars) and venous smooth muscle cells (VSMCs, solid bars). Cells were cultured in full medium, without additional growth factors. Receptors were quantified by flow cytometry. Each bar represents the means ± SD of four observations. The number of PDGF-Rα receptors was similar to that of PDGF-Rβ receptors on arterial SMCs. The number of PDGF-Rα receptors on venous SMCs was twofold lower (^*P <0.05) than that on arterial SMCs, while PDGF-Rβ expression on venous SMCs was 6.6-fold greater (^*P <0.05) than that on arterial SMCs.

Fig. 3

Differential phosphorylation of PDGF receptors (PDGF-R) and EGF receptors (EGF-R) in arterial smooth muscle cells (ASMCs) and venous SMCs (VSMCs) in response to two PDGF isoforms. Cells were incubated for 5 min with either PDGF-AA (50 μg/ml) or PDGF-BB (50 μg/ml), then lysed. The cell lysate was subsequently subjected to analysis in duplicate in a phospho-RTK immunoblot array containing 42 RTKs. A: Representative image of a phospho-RTK immunoblot array. The regions on the blot where various phospho-receptors were visualized are indicated by arrows and specific receptor labels (phosphorylated PDGF-Rα labeled as p-PDGF-Rα, phosphorylated PDGF-Rβ labeled as p-PDGF-Rβ, and phosphorylated EGF-R labeled as p-EGF-R). The unlabeled dots shown in the four corners of each sub-panels in panel A are internal controls. B: Means ± SD of chemiluminescent signals (quantified as image pixel density) from three separate phospho-RTK array experiments. The asterisk indicates a statistically significant (^*P <0.05) difference between ASMC versus VSMC. Phospho-PDGF-Rα increased in ASMC, but not in VSMC, in response to PDGF-AA. PDGF-AA had no effect on phosphorylation of PDGF-Rα or PDGF-Rβ in VSMC. Upon PDGF-BB stimulation, the phosphorylation of PDGF-Rα increased to a similar degree in ASMC and VSMC. The increase in phosphorylation of PDGF-Rβ was high in ASMC, but was even greater in VSMCs (^*P <0.05). PDGF-BB transactivated EGF-R in VSMC but not in ASMC.

Fig. 4

Phosphorylation of proteins in the MAPK and Akt pathways in arterial smooth muscle cells (ASMC) and venous smooth muscle cells (VSMC) in response to various PDGF isoforms. Cells were incubated for 5 min with either PDGF-AA (“AA”, 50 μg/ml), AB (“AB”, 50 μg/ml), or PDGF-BB (“BB”, 50 μg/ml), then lysed. The cell lysate was subsequently subjected to SDS–PAGE and immunoblotting using the respective specific antibodies against the phosphorylated proteins. A: Representative blots from three separate experiments are shown. “C” indicates control or unstimulated quiescent cells. B: Means ± SD of chemiluminescent signals (quantified as image pixel density) from three separate experiments. PDGF-AA increased the level of phosphorylated JNK1 (p-JNK1) in ASMC, but not in VSMC. Other than JNK-1, there were no substantial differences in the levels of phosphorylated ERK1/2, p38, or Akt upon stimulation by PDGF-AA, PDGF-AB, or PDGF-BB between ASMC and VSMC. Phosphorylation of JNK 2/3 was not observed in either cell type after any PDGF isoform treatment.

Fig. 5

Time course of changes in phosphorylation of JNK1/2/3 in arterial smooth muscle cells (ASMC) and venous smooth muscle cells (VSMC) in response to two PDGF isoforms. Cells were incubated for the indicated durations with either PDGF-AA (“AA”, 50 μg/ml) or PDGF-BB (“BB”, 50 μg/ml), then lysed. The cell lysate was subsequently subjected to SDS–PAGE and immunoblotting using a specific antibody against phosphorylated JNK (p46/p-JNK1 and p54/p-JNK2/3). A: Representative blots from three separate experiments are shown. “C” indicates control or unstimulated quiescent cells. B: Means ± SD of chemiluminescent signals (quantified as image pixel density) from three separate experiments. PDGF-AA stimulated transient activation of JNK1 in ASMC, but not in VSMC. PDGF-BB stimulated JNK1 phosphorylation in both ASMC and VSMCs, although the activation of JNK1 in VSMC was sustained longer than that in ASMC. There was no JNK2/3 phosphorylation observed in either ASMC or VSMC at any time point.

Fig. 6

Protein expression of cell-cycle inhibitors, p27^kip1 and p21^cip1 in arterial smooth muscle cells (ASMC) and venous smooth muscle cells (VSMC) in response to various PDGF isoforms. Cells were incubated for 48 h with either PDGF-AA (“AA”, 50 μg/ml) or PDGF-BB (“BB”, 50 μg/ml), then lysed. The cell lysate was subsequently subjected to SDS–PAGE and immunoblotting using specific antibodies against either p27^kip1 or p21^cip1. A: Representative blots from two separate experiments are shown. “C” indicates control or unstimulated quiescent cells. B: Means ± SD of chemiluminescent signals (quantified as image pixel density) from two separate experiments. Protein levels of p27^kip1 were high in quiescent ASMC and down-regulated by PDGF-AA, PDGF-AB, or PDGF-BB. Protein levels of p27^kip1 were also high in quiescent VSMC and down-regulated by PDGF-AB or PDGF-BB, but not significantly affected by PDGF-AA. The level of p21^cip1 expression was similar in the control and ASMC treated with various PDGF isoforms. In VSMC, the expression of p21^cip1 was barely detectable in quiescent cells, which was not altered by PDGF-AA. However, p21^cip1 expression was increased substantially by PDGF-AB and PDGF-BB in VSMC.