|Home | About | Journals | Submit | Contact Us | Français|
To characterize and compare primary and restenotic lesions of the superficial femoral artery and analyze the contribution of TGF-β/Smad3 signaling to the pathophysiology of peripheral artery occlusive disease.
Immunohistochemical studies were performed on specimens retrieved from the superficial femoral artery of patients undergoing either atherectomy for primary atherosclerotic or recurrent disease after stenting and/or prior angioplasty. Immunohistochemical analysis revealed a significantly higher smooth muscle cell (SMC) content (α-actin) and expression of Smad3 in restenotic lesions while primary lesions contained significantly more leukocytes (CD45) and macrophages (CD68). Further studies demonstrated colocalization of Smad3 with α-actin and PCNA, suggesting a role for Smad3 in the proliferation observed in restenotic lesions. To confirm a role for Smad3 in SMC proliferation, we both upregulated Smad3 via adenoviral mediated gene transfer (AdSmad3) and inhibited Smad3 through transfection with siRNA in human aortic SMCs, assessing proliferation with tritiated thymidine. Overexpression of Smad3 enhanced whereas inhibition of Smad3 decreased cell proliferation.
Differences in cellular composition and cell proliferation in conjunction with the finding that Smad3 is expressed exclusively in restenotic disease suggest that TGF-β, through Smad3 signaling, may play an essential role in SMC proliferation and the pathophysiology of restenosis in humans.
Intimal hyperplasia leading to restenosis significantly limits the long-term success of percutaneous interventions for lower extremity vascular occlusive disease. It has been reported that in patients undergoing superficial femoral artery (SFA) revascularization, restenosis occurs within six months in one-fourth of patients who receive stents and nearly one-half of those who receive angioplasty alone.1 This equates to over 50% of patients requiring re-intervention following angioplasty of the SFA.2 It has been suggested that the mechanisms leading to the development of restenosis may be different from those that produce primary atherosclerosis. The growing use of endovascular atherectomy in peripheral vascular disease provides the opportunity to compare the pathophysiology of restenotic versus primary atherosclerotic plaque and to better define the cellular events that lead to recurrent disease following vascular intervention.
Primary atherosclerotic plaque has been shown to be hypocellular, consisting primarily of ambiguous ground substance and collagen with a variable lipid component.3 In contrast, restenotic lesions are typically hypercellular with foci of vascular smooth muscle cells (SMCs) and extracellular matrix.4 Vascular SMC proliferation has been implicated in the pathogenesis of intimal hyperplasia and ultimately restensosis,5-9 and has been studied in both primary atheromata and restenotic lesions. The vast majority of these studies have been conducted in specimens derived from coronary arteries. Moreover, the data vary widely; in some studies there was little evidence of cellular proliferation in restenotic lesions10-13 and in others, proliferation was a prominent feature of this process.14, 15 It has been suggested that peripheral vascular lesions differ from central or coronary artery disease. However, the composition of atherosclerosis or intimal hyperplasia in peripheral arterial or superficial femoral artery disease has not been well characterized.
Among the many growth factors implicated in the development of intimal hyperplasia, transforming growth factor-β(TGF-β), specifically TGF-β1, has been well established as a primary contributor to this process. TGF-β1 expression is increased in injured and restenotic vessels and in neointimal hyperplasia after balloon injury in animals.8, 10, 17, 18 Furthermore, exposure of arteries to exogenous TGF-β1 after injury results in increased neointima formation in animal models.7 Although it has been shown that TGF-βsignals through several pathways, it is believed that the main signaling mechanism is through the Smad family of proteins. TGF-βsignals by binding to type I receptors followed by formation of a heteromeric complex with the type II receptor resulting in phosphorylation and activation of the TGF-βdownstream signaling molecules Smad2 and Smad3. Activated Smad2 or Smad3 then heterodimerize with Smad4 and the resulting complex translocates to the nucleus where it acts as a transcription factor to regulate the expression of target genes.19 Interestingly, our previous work in an animal model of restenosis revealed that Smad3 is upregulated in injured arteries and that in the rat model of intimal hyperplasia, increased levels of Smad3 are associated with increased SMC proliferation. Based on these previous findings, we hypothesized that the mechanism behind the formation of restenotic versus primary atherosclerotic plaque is different and that expression of the TGF-protein Smad3 may in part be responsible for these differences.
In the current study, we demonstrate that the pathophysiology of the restenotic lesion is distinct from that of primary atheromatous plaque in the superficial femoral artery. Furthermore, we demonstrate that these differences may be related to changes in Smad3 expression and SMC proliferation. Together, our findings contribute to a broader understanding of the pathophysiology of peripheral artery restenosis in humans.
Cell culture reagents were from Gibco BRL Life Technologies. Chemicals, if not specified, were purchased from Sigma Chemical Co.
Specimens were retrieved from the superficial femoral artery of patients undergoing percutaneous atherectomy for primary atherosclerotic lesions (n=5) or restenotic lesions after stenting and/or prior angioplasty (n=3). All specimens were collected under the approval of the Institutional Review Board. Cores of tissue in no specific orientation were obtained from percutaneous atherectomy procedures using the Silver Hawk atherectomy device (ev3 Inc., Plymouth, MN). Between 4-6 tissue cores were paraffin embedded into one block, then cut int sample, at least 3 sections, through multiple tissue cores, were analyzed that were separated by over 100μm.
Paraffin embedded tissues were cut into 7 μm sections for analysis. Quantitative immunohistochemistry using mouse anti-PCNA 1:50 (Santa Cruz), mouse anti-α-actin 1:800 (Sigma), mouse anti-CD68 1:50 (Dako), mouse anti-CD45 1:50 (Dako), and rabbit anti-Smad3 1:50 (Zymed) was performed as described elsewhere.20 Staining was compared to the appropriate negative controls using rabbit or mouse IgG at a concentration (in with a Nikon Eclipse E800 upright microscope and digital images were acquired using a RetigaEXi CCD digital camera. Percent positivity was calculated as (# positive cells/# total nuclei) per section by identifying positive cells (brown) versus hematoxylin stained cells by pixel color on Adobe Photoshop (Version 9.0.2), then counting each subset of cells on NIH image software (ImageJ 1.36b).
Immunofluorescent staining was performed with donkey anti-mouse Alexa 555 or donkey anti-rabbit Alexa 488 (Molecular Probes). TO-PRO 3 (Molecular Probes) was used to identify nuclei. Fluorescent staining was visualized and digital images were taken on a Zeiss LSM 510 Laser Scanning Confocal imaging system with the appropriate argon beam lasers. For each specimen, cells were counted in 4 fields at 882x magnification. Digital images were analyzed using Zeiss Image Browser software.
Smooth muscle cells and media were purchased from ScienCell and maintained at 37°C with 5% CO2.
The sequence for the human Smad3 gene was obtained from the Pubmed Nucleotide database. Using the BLOCK-iT™ RNAi Designer (Invitrogen), potential siRNA sequences were generated. Two sequences were ultimately tested for their ability to inhibit Smad3 expression by both Western blot and reverse transcriptase PCR. The sequence used for the following experiments was 5′-GCUCCAUCUCCUACUACGAGCUGAA-3′.
Primary human aortic smooth muscle cells (HAoSMC) (passages 2-6) were infected with either AdLacZ or AdSmad3 (3×104 particles/cell in 2% FBS for 4 hours). Following overnight recovery in 2% FBS and 48 h of serum starvation (in 0.5% FBS), tritiated thymidine incorporation, as a surrogate for DNA synthesis, was assessed as previously described.21
Primary HAoSMCs (P2-6), were transfected with either siRNA to Smad3 or a scramble control (20 pmoles/well in a 24-well plate). Following 48 h of serum starvation (in 0.5% FBS), tritiated thymidine incorporation was assessed as previously described.21
Vascular SMCs were infected with AdLacZ or AdSmad3 as described above. After 48 hours of serum starvation, cells were lysed. The lysate was then subjected to apoptosis ELISA (Roche) following manufacturer’s instruction.
Values were expressed as means ± S.E. Unpaired Student’s t test was used to evaluate the statistical differences between control and treated groups. Multiple comparisons were made using ANOVA and Dunn’s post-hoc correction. Values of p < 0.05 were considered significant. All experiments were repeated at least three times.
Samples were collected from the superficial femoral artery of patients undergoing atherectomy for primary atherosclerotic or restenotic lesions. Immunohistochemical analysis revealed a significantly higher expression of α-actin (a marker for SMCs) and PCNA (a marker for proliferating cells) in restenotic lesions when compared to primary atherosclerotic lesions (Fig. 1A, B, & E). In contrast, primary lesions expressed significantly more inflammatory cell markers CD68 (macrophages) and CD45 (leukocytes) (Fig. 1C, D, & E). Thus, cell proliferation appears to be a major feature of human restenotic lesions. Furthermore, restenotic lesions are comprised primarily of SMCs, whereas primary atheromata contain significantly more inflammatory cells. Consistent with prior reports, we also found that restenotic plaques were overall more cellular than primary atherosclerotic plaques (647±175 cells/hpf for restenotic lesions vs. 222±35 cells/hpf for primary lesions, p<0.05).
These observations prompted us to explore the mechanism responsible for differences between primary and restenotic lesions. Studies have shown that TGF-βis a ubiquitous cytokine associated with restenotic plaque. Moreover, we have previously shown that the TGF-βsignaling molecule Smad3 is upregulated after carotid balloon arterial injury in the rat. We therefore explored whether there might be a differential in Smad3 expression in human primary and restenotic atherectomy samples. We found that Smad3 was expressed exclusively in restenotic lesions (Fig. 2). Whereas 25.9±2.6% of cells in restenotic lesions were Smad3 positive, no primary atherosclerotic plaque cells were identified as expressing Smad3. Immunohistochemistry data from all 5 primary atherosclerotic plaques and 3 restenotic plaques have been included in the Supplemental Data.
We next attempted to localize the Smad3 expression in restenotic plaque by double immunostaining of Smad3 with either α-actin (Fig. 3A) or PCNA (Fig. 3B), followed by visualization with confocal microscopy. We found that all visualized Smad3+ cells were α-actin positive (nuclear Smad3, green, is surrounded by cytoplasmic α-actin, red) (Fig. 3A), suggesting that enhanced Smad3 expression is confined to SMCs. We also found that that all Smad3 cells were PCNA positive (nuclear Smad3, green, and nuclear PCNA, red, merge to depict yellow nuclei) (Fig. 3B). Thus, Smad3 is exclusively associated with proliferating SMCs within restenotic lesions.
Our in vivo findings suggest a role for Smad3 in mediating the increased cell proliferation observed in human restenotic lesions. To further test this hypothesis, Smad3 was overexpressed in human aortic SMCs (HAoSMCs) via adenovirus mediated gene transfer (AdSmad3) while control cells were infected with a control virus (AdLacZ) (Fig. 4A). 48 hours later, DNA synthesis was measured using a tritiated thymidine incorporation assay. Consistent with our in vivo findings, upregulation of Smad3 resulted in a significant increase in cell proliferation as compared to the AdLacZ control (Fig. 4B). To confirm that this observation is due to overexpression of Smad3 rather than changes in TGF-17, we have shown by Western blotting in the Supplemental Data that infection with either AdSmad3 or AdLacZ does not affect levels of TGF-define the role of Smad3 in cell proliferation, Smad3 expression in HAoSMCs was downregulated using an siRNA (Fig. 4A). As shown in Figure 4B, downregulation of Smad3 resulted in a significant decrease in cell proliferation compared to a scramble-treated HAoSMC control, further substantiating the importance of Smad3 in human SMC proliferation.
We next asked the question whether Smad3 overexpression had any effect on vascular SMC apoptosis. Since we had already shown that restenotic plaques, which overexpressed Smad3, were more cellular than primary plaques, we initially hypothesized that Smad3 overexpression may be associated with decreased SMC apoptosis. However, as demonstrated in Figure 5, although the level of apoptosis in cells overexpressing Smad3 was significantly higher than the negative control and cells infected with AdLacZ, it is still very low and several fold lower than the positive control. This therefore suggested that Smad3 expression does not play a major role in regulation of apoptosis in vitro.
We have found that there is significantly more cell proliferation in restenotic plaque compared to primary atheromata in patients with lower extremity vascular disease. In addition, restenotic plaque is comprised primarily of SMCs whereas atheromata contain significantly more inflammatory cells. These differences in cellular composition and cell proliferation in conjunction with the novel finding that Smad3 is expressed exclusively in restenotic disease suggest that Smad3 expression may play a role in the mechanism underlying restenotic, as opposed to primary atherosclerotic disease.
It was surprising that we found almost no white blood cells (WBCs) in restenotic plaque, since WBCs and inflammation have been shown to play an integral role in the genesis of atherosclerosis. However, our own immunohistochemical analysis of rat carotid arteries after balloon angioplasty also revealed very few neointimal inflammatory cells, as evidenced by immunohistochemical staining for CD45 and CD68. Moreover our observations are supported by similar findings in other animal models of intimal hyperplasia.22, 23 This differential localization of WBCs predominately in primary but not restenotic lesions highlights another distinction between these two processes. Whereas the inflammatory response is critically important in atherosclerosis; inflammation does not appear to be as essential in the development of restenosis.
While there is a wealth of data regarding the characterization of atherectomy specimens derived from coronary arteries,10-16 there has been only a limited evaluation of peripheral lesions and specifically cell proliferation in peripheral lesions. In several studies rates of apoptosis in peripheral versus central restenotic lesions have been described.3, 24 Others have pooled peripheral and central atherectomy specimens for the characterization of restenotic versus primary atherosclersosis.14, 25, 26 Recently studies with drug-coated stents have implied that the pathophysiology of restenotic disease in coronary versus peripheral vascular disease is quite distinct. Stents coated with rapamycin significantly inhibit restenosis in the coronary circulation whereas rapamycin coated stents were not successful in diminishing the rate of restenosis when used in the peripheral circulation.27 These findings underscore the importance of studying the pathophysiology of restenosis specifically in the peripheral arterial system.
SMC proliferation has been identified as an invariable component of intimal hyperplasia in animals, but previous studies have yielded conflicting results regarding the frequency of cell proliferation in human restenotic lesions.10-16 Pickering et al. reported a significantly higher percentage of proliferating cells in restenotic as opposed to primary atherosclerotic plaque (59% vs. 20%).14 Our findings support those of Pickering in that we identified that almost 70% of cells in restenotic lesions were PCNA positive whereas slightly over 20% of cells in primary atherosclerotic lesions were proliferating. Our data also suggest that SMC proliferation is a more dominant process in peripheral versus coronary restenotic lesions. Skowasch et al. were unable to identify proliferation in restenotic coronary lesions.13 Moreover, Glover et al. found proliferation in only one of twenty coronary restenotic samples as measured by in situ hybridization for histone 3 mRNA expression.11 O’Brien et al. found PCNA positivity in 26% of restenotic coronary lesions, but in each of these restenotic lesions <1% of cells were actually PCNA+.10 Similarly, Marek et al. found only 1-3% of cells expressed PCNA in 41% of restenotic carotid lesions.12 Alternatively, we have found a proliferative index of near 70% in restenotic lesions removed from the superficial femoral artery. Schwartz et al.hypothesized that restenosis is characterized by proliferation only in the early phase of its development; thus lesions derived from human coronary arteries if they are sampled at a later stage of their development, are not typified by proliferation.26 Our data, however, do not support this hypothesis since all of our peripheral restenotic specimens were taken at a point greater than six months following the initial intervention and as previously noted, proliferation in these lesions was still quite robust.
Having identified proliferating SMCs in the restenotic lesion, we next focused on potential signaling mechanisms that might account for increased proliferation. It is well established that TGF-βligands are upregulated following vascular injury.28 We have previously shown in a rat balloon injury model that the TGF-βsignaling molecule Smad3 is also upregulated after vascular injury. Furthermore, overexpression of Smad3 was associated with increased SMC proliferation in our animal model. This is a surprising finding since TGF-βactivity is traditionally thought to be associated with inhibition of SMC proliferation. In fact, our previous work suggests that while TGF-βcan inhibit SMC proliferation through a non-Smad3 pathway, in the presence of high levels of Smad3, which develop after arterial injury, TGF-βappears to have the converse effect of stimulating SMC proliferation. Our new and similar findings in human vessels and cells serve to confirm this hypothesis and make Smad3 a relevant target in the treatment of human restenotic disease.
There is conflicting data regarding the role of Smad3 in intimal hyperplasia. Kobayashi et al. have previously shown that SMCs derived from a Smad3 knockout mouse (i.e. cells that are lacking Smad3) have increased baseline proliferation compared to wild type cells.18 These findings would imply that Smad3 is involved in endogenous suppression of SMC proliferation. Mimicking this experimental situation we inhibited Smad3 expression with siRNA and found an opposite result: inhibition of SMC proliferation. The explanation for the discrepancy between ours and Kobayashi’s findings is not entirely clear.
In addition to stimulation of cell proliferation, elevated levels of Smad3 may induce other processes that lead to intimal hyperplasia. Extracellular matrix production has been shown to be an important component of intimal hyperplasia, and we have previously shown that Smad3 is essential for vascular SMC production of fibronectin.29 Alternatively, Smad3 may influence SMC apoptosis, however our in vitro data suggest that the effect of Smad3 may be minimal. Finally, overexpression of Smad3 may stimulate SMC migration, another process that has been shown to play a role in the development of intimal hyperplasia. This would be one area for future in vitro studies.
One of the limitations of current study is the use of human aortic SMCs for in vitro studies. As the aortic bed does not clinically succumb to occlusive lesions as compared to the SFA, it is possible that SMCs from these two different regions may have distinct responses to intracellular levels of Smad3. Another limitation is the small number of patient samples included. However, our findings even with an n of 3 were quite dramatic and all of our findings were statistically different, suggesting that Smad3-mediated SMC proliferation could be an important mechanism unique to restenosis.
The present study confirms that the mechanism underlying the development of primary atherosclerotic plaque in the superficial femoral artery is quite distinct from that of restenotic disease. White blood cells contribute significantly to primary plaque while restenotic lesions are highly cellular, comprised primarily of SMCs; in addition, SMC proliferation is quite robust. Moreover, the TGF-βsignaling protein Smad3 is upregulated in restenotic lesions and cellular localization studies suggest that this might account for the highly proliferative nature of these lesions. A better understanding of the relationship between Smad3 and cell proliferation will provide us with further insight into the mechanisms by which TGF-βinduces intimal hyperplasia as well as enable us to target specific aspects of this signaling mechanism for therapy to prevent restenosis after treatment of peripheral arterial disease.
We would like to thank Drs. Rishi Kundi and James McKinsey for their help in obtaining surgical specimen. This work was supported by a Public Health Service Grant R01 HL-68673 (K. Kent and B. Liu) from the National Heart, Lung, and Blood Institute, an American Heart Association grant-in-aid 0455859T (B. Liu), National Institutes of Health F32 HL088818-01 (S. Tsai), Society of University Surgeons-Ethicon Scholarship Grant Award (S. Tsai), Uehara Memorial Foundation Research Fellowship Award (D. Yamanouchi), and Howard Hughes Medical Institute Medical Research Training Fellowship (R. Edlin).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.