Background
Individuals with insulin resistance states and elevated levels of circulating insulin, the prototype of which is type II diabetes, are more prone to develop vascular disease and less likely to benefit from available treatments compared to non-diabetic individuals[
1]. Abciximab and eptifibatide, two widely used integrin inhibitors, improve mortality in diabetics undergoing percutaneous coronary intervention (PCI). In a pooled analysis of three large clinical trials, abciximab was associated with a 44% reduction in one year mortality in diabetics (4.5% in patients receiving placebo and 2.5% in patients receiving abciximab)[
2]. Similarly, eptifibatide was associated with a reduction in one year mortality in diabetics (3.5% in patients receiving placebo and 1.3% in patients receiving eptifibatide) in the Enhanced Suppression of the platelet IIb/IIIa Receptor with Integrilin Therapy (ESPRIT) trial[
3].
Abciximab and eptifibatide, in addition to inhibiting platelet aggregation via antagonism of fibrinogen binding to α
IIbβ
3 integrins, also antagonize ligand binding to α
vβ
3 integrins on vascular cells[
4,
5]. Recent studies in cultured cells have revealed considerable cross-talk between α
vβ
3 integrins and insulin receptor-mediated signals. Vuori and Ruoslahti[
6] found that α
vβ
3 integrins associate with insulin-receptor substrate-1 (IRS-1), a docking protein that phosphorylates on tyrosine following insulin-receptor activation and binds SH2 domain-containing proteins that propagate the insulin signal. Moreover, α
vβ
3 integrins associated with tyrosine phosphorylated insulin receptors and other, as yet unidentified, tyrosine phosphorylated proteins in insulin-treated fibroblasts[
7]. These associations were specific for α
vβ
3 integrins and proliferative responses to insulin were enhanced by extracellular matrices that ligated α
vβ
3 integrins. More recently, Lopez-Alemany et al. reported that plasminogen activator inhibitor-1 (PAI1) competes with α
vβ
3 integrins for binding to vitronectin and by this mechanism blocks insulin-induced migration in NIH3T3 cells and human umbilical vein endothelial cells[
8].
Given the important role of smooth muscle cell (SMC) proliferation in atherosclerosis progression and in revascularization failures, the present studies were performed to explore the hypothesis that abciximab and eptifibatide inhibit proliferative responses of human aortic SMC (HASMC) to insulin via antagonizing αvβ3 integrins.
Discussion
Antagonism of α
vβ
3 integrins markedly inhibits proliferative responses of HASMC to insulin. This conclusion is based on our findings that insulin-induced proliferation was inhibited by anti-β
3 integrin monoclonal antibody (m7E3), chimeric antigen binding fragment of 7E3 (c7E3), anti-α
vβ
3 monoclonal antibody (LM609), anti-β
3 peptides (eptifibatide) and anti-α
v peptides (cRGD). These studies add to the wealth of data that α
vβ
3 antagonists have profound effects on SMC proliferation and migration, two mechanisms that play central roles in vascular pathology. In various studies, α
vβ
3 antagonists have been shown to inhibit proliferative responses of SMC to insulin, thrombospondin, thrombin, IGF-1, osteopontin, Del1 and transforming growth factor-β and to inhibit migratory responses of SMC to insulin-like growth factor-1 – (IGF-1), PDGF, vitronectin, thrombospondin and osteopontin (reviewed in[
22]). In animal models, α
vβ
3 antagonists have been shown to reduce SMC migration, SMC proliferation and neointima formation following vascular injury[
14,
23].
In this study, insulin stimulated a proliferative response in quiescent HASMC maintained in 0.5% FBS that was similar in magnitude to that observed with PDGF-BB. These results are consistent with previous studies which showed that insulin can stimulate proliferation of cultured SMC and also enhance proliferative responses to other mitogens[
24‐
26]. Insulin has also been found to stimulate neointimal formation and proliferation of organ cultures of saphenous veins and internal mammary arteries[
27]. Saphenous vein responses to insulin were similar in magnitude to those observed with PDGF-BB whereas insulin was a less efficacious mitogen than PDGF-BB in internal mammary cultures. In contrast to these studies, Obata et al.[
28] found that low concentrations of insulin stimulated insulin receptor substrate-1 phosphorylation and amino acid uptake but not thymidine incorporation into DNA in rat aortic SMC. The concentrations of insulin used by Obata et al (1–10 nmol/L) were approximately 100 fold lower than used in the present studies.
We found that focal adhesions as delineated by anti-vinculin staining formed rapidly following treatment of quiescent HASMC with insulin and that α
vβ
3 antagonists partially inhibited this response. Vinculin is one of the first actin binding proteins recruited into focal adhesions[
29]. It is widely used as a marker of 'classic' focal adhesions, which are usually located at the cell periphery, are highly tyrosine phosphorylated, and contain α
vβ
3 integrins. Vinculin is also found in focal complexes, which are short-lived structures that mature into focal adhesions.
Recent studies have highlighted the existence of two major signaling pathways that are initiated by insulin binding to the insulin receptor and which mediate insulin action[
15,
16]. One pathway, which involves IRS proteins and phosphatidylinositol 3-kinase, appears to be responsible for most, if not all, of the metabolic aspects of insulin action. The second signaling pathway, involving Ras and mitogen activated protein kinase (MAPK), is responsible for proliferative responses to insulin. Doronzo et al reported that insulin increased ERK 1/2 and JNK1 phosphorylation in HASMC in a time dependent manner beginning within 5 minutes of treatment[
19]. In the present studies, we found that antagonism of α
vβ
3 inhibited insulin-induced activation of JNK1, but not ERK 1/2, in HASMC. JNK1 is a member of the mitogen activated protein kinase superfamily that is activated by dual phosphorylation at a Thr-Pro-Tyr motif and once activated, functions to phosphorylate c-jun at amino terminal serine regulatory sites which increases activity of the transcription factor AP-1. A clear link between JNK1 activation and proliferation in insulin-treated HASMC has not been established although JNK antagonists has been reported to inhibit proliferation of cultured SMC in other systems[
30,
31]. In studies utilizing the rat carotid balloon injury model, transfection of a dominant negative JNK prior to injury prevented neointimal formation and markedly suppressed SMC proliferation in both the intima and the media after rat carotid artery injury[
32].
The inhibitory effect of α
vβ
3 antagonists on JNK1 activation occurs in response to a variety of stimuli. In addition to inhibiting JNK1 activation in response to insulin, α
vβ
3 antagonists inhibit α-thrombin-induced[
9] and TFB-induced[
33] JNK1 activation in SMC. Several potential mechanisms that might explain these results are suggested by recent studies. Activation of the small G protein Rho is mediated by integrin engagement[
34] and recently Ohtsu et al reported that activation of Rho, and its effector Rho-kinase/ROCK was required for angiotensin II-induced JNK activation in SMC. Alternatively, studies in myocytes[
35] and HEK cells[
21,
36] have implicated the formation of protein complexes involving focal adhesion kinase in activation of JNK.
α
vβ
3 antagonists inhibited insulin-induced proliferation without blocking ERK 1/2 phosphorylation. The biochemical steps involved in signal transduction through the ERK pathway are well established but less is known about how these signals are implemented into specific biological responses, and in particular the role of intracellular localization of members of this pathway. Following activation, ERK localizes to different subcellular compartments, including focal adhesions, and phosphorylates specific proteins leading to cellular responses. The specificity of the biological response is likely to be at least partially controlled by the localization of signaling, which enables ERK activity to be directed towards specific targets. Integrin engagement is necessary for active ERK localization to focal adhesions suggesting that a potential mechanism whereby integrin antagonism could inhibit growth but not ERK phosphorylation is via interrupting ERK targeting[
37].
Tirofiban had no effect on insulin-induced proliferation, consistent with prior studies showing that tirofiban does not antagonize α
vβ
3[
5]. The different affinities of eptifibatide and tirofiban for α
vβ
3 are not surprising given that eptifibatide is a synthetic, cyclic peptide with a Lys-Gly-Asp (KGD) sequence whereas tirofiban is a nonpeptide derivative of tyrosine. Tirofiban was used at a concentration of 30 μmol/L and eptifibatide was used at various concentrations between 5 and 100 μmol/L in our studies. This concentration of tirofiban is approximately 700 fold greater than peak plasma concentration observed in patients (40 ng/ml) receiving a continuous infusion[
38]. There is wide inter-individual variation in plasma levels however, and it is also unknown whether tirofiban concentrations are higher within the vessel wall than in plasma.
Intimal thickening due to abnormal proliferation of vascular smooth muscle cells is the major cause of revascularization failures in diabetics. Since αvβ3 integrin expression is upregulated in atherosclerotic lesions and at sites of balloon angioplasty, these results suggest that one way to potentially regulate insulin effects on SMC at sites of vascular healing is via antagonism of αvβ3.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AP performed the flow cytometry, proliferation and binding assays. He also assisted in interpreting the data and writing the manuscript. RZ performed the focal adhesion assays and immunoflouresence studies. JH performed the western blots and JNK assays. GAS conceived of the study, participated in the design, coordinated the interpretation of the data and drafted the manuscript. All authors read and approved the final manuscript.