Molecular Signatures Determining Coronary Artery and Saphenous Vein Smooth Muscle Cell Phenotypes

【摘要】  Objective- Phenotypic differences between vascular smooth muscle cell (VSMC) subtypes lead to diverse pathological processes including atherosclerosis, postangioplasty restenosis and vein graft disease. To better understand the molecular mechanisms underlying functional differences among distinct SMC subtypes, we compared gene expression profiles and functional responses to oxidized low-density lipoprotein (OxLDL) and platelet-derived growth factor (PDGF) between cultured SMCs from human coronary artery (CASM) and saphenous vein (SVSM).

Methods and Results- OxLDL and PDGF elicited markedly different functional responses and expression profiles between the 2 SMC subtypes. In CASM, OxLDL inhibited cell proliferation and migration and modified gene expression of chemokines (CXCL10, CXCL11 and CXCL12), proinflammatory cytokines (IL-1, IL-6, and IL-18), insulin-like growth factor binding proteins (IGFBPs), and both endothelial and smooth muscle marker genes. In SVSM, OxLDL promoted proliferation partially via IGF1 signaling, activated NF- B and phosphatidylinositol signaling pathways, and upregulated prostaglandin (PG) receptors and synthases. In untreated cells, -chemokines, proinflammatory cytokines, and genes associated with apoptosis, inflammation, and lipid biosynthesis were higher in CASM, whereas some ß-chemokines, metalloproteinase inhibitors, and IGFBPs were higher in SVSM. Interestingly, the basal expression levels of these genes seemed closely related to their responses to OxLDL and PDGF. In summary, our results suggest dramatic differences in gene expression patterns and functional responses to OxLDL and PDGF between venous and arterial SMCs, with venous SMCs having stronger proliferative/migratory responses to stimuli but also higher expression of atheroprotective genes at baseline.

Conclusions- These results reveal molecular signatures that define the distinct phenotypes characteristics of coronary artery and saphenous vein SMC subtypes.

We compared gene expression profiles and functional responses of arterial and venous smooth muscle cells (SMCs) with oxidized LDL and PDGF and identified the molecular signatures that define the distinct phenotypes characteristic of arterial and venous SMC subtypes.

【关键词】  gene expression profiling smooth muscle cells coronary arteries saphenous vein

Introduction

Smooth muscle cells (SMCs) consist of heterogeneous subtypes among various vascular beds and at different vascular developmental stages. 1-3 SMCs from veins and arteries have different embryonic origins and are exposed to different hemodynamic environments. However, primary SMC cultures isolated from veins and arteries share many common features, including similarities in morphology and responses to mitogens and chemoattractants. 3,4 Atherosclerosis occurs in arteries but rarely in veins under normal anatomic conditions, despite the exposure of all vessels to systemic risk factors such as smoking, hyperlipidemia, and hyperglycemia. However, after implantation during arterial bypass surgery saphenous veins become prone to accelerated atherosclerosis. 5

SMC proliferation and migration induced by cytokines, growth factors like platelet-derived growth factor (PDGF), modified lipoproteins like oxidized low-density lipoprotein (OxLDL), and other agents from both vascular cells and infiltrating immune cells play important roles in many disease processes including atherosclerosis, failure of vascular bypass grafts, and postangioplasty restenosis. 6,7 However, the molecular features that determine the different SMC subtypes and unique functional responses are still poorly elucidated.

In this study, we used microarray gene expression profiling to identify the unique molecular signatures that distinguish human vascular SMC subtypes from saphenous vein (SVSM) and coronary artery (CASM), and further characterized the molecular features that determine different proliferative and migratory responses to OxLDL, a key atherogenic factor, 6 and PDGF, a potent mitogen and chemoattractant. 4 We found that SMCs from arteries and veins in culture maintained distinct cell lineage gene expression programs and responses. OxLDL and PDGF induced dramatically different gene expression and functional responses between the 2 SMC subtypes with stronger proliferative/migratory responses produced in venous SMCs that are at least partially via IGF1 signaling. Remarkably, in the basal state, -chemokines and genes related to lipid biosynthesis were expressed at higher levels in arterial than in venous SMCs, whereas some ß-chemokines, metalloproteinase inhibitors (TIMP1, 2, and 3), insulin-like growth factor binding proteins (IGFBPs), and many atheroprotective genes were more highly expressed in venous SMCs. This is the first study to our knowledge identifying global molecular signatures that define coronary artery and saphenous vein SMC subtypes. The results provide clues to the molecular basis of the varying susceptibility to atherosclerosis between veins and arteries in different states and in different vascular beds. These data are complementary to data from similar experiments documenting inherent differences in endothelial cells from different vascular beds and correlations of gene expression to disease relevant pathways. 8

Materials and Methods

In Vitro Cell Culture

Three separate primary cultures of CASM and SVSM were obtained from Cambrex (Walkersville, Md). Cells had tested positive for smooth muscle cell markers like -actin and negative for von Willebrand factor (vWF). CASM and SVSM were cultured 100-mm or 150-mm culture dishes in SmGM-2 media (Cambrex) containing 5% fetal sera, human epidermal growth factor, human fibroblast growth factor, insulin, and antibiotics. Cells cultured at passage 6 and 80% confluence were deprived of serum for 12 hours, and then exposed to OxLDL (40 µg/mL), PDGF (10 ng/mL), or vehicle for 24 hours. Fully oxidized LDL was purchased from BTI (Stoughton, Mass). PDGF-BB was obtained from CalBioChem.

Microarray Hybridization and Data Analysis

Details of RNA isolation and microarray hybridization protocols have been described previously. 8 Briefly, total RNA isolated from primary cell cultures and human universal RNA (Stratagene, La Jolla, Calif) used as common reference for all samples were labeled with the Agilent Direct Fluorescent labeling kit (Palo Alto, Calif) and hybridized to the Agilent human 1A oligo array representing 16,391 unique gene sequences (http://www.chem.agilent.com). Data analysis was performed using Resolver System (V4.0.1.0.10.RSPLIT; Rosetta, Inc, Seattle, Wash), by which the "raw ratios" (sample RNA versus common reference RNA for each condition and cell type) were first re-ratioed, and then the ratios were combined using an error-weighted approach in which expression log ratios with less error contribute more to the combined result than those with greater error (www.ROSETTABIO.com). Fold changes of gene expression change 1.5 and corresponding ANOVA P 0.001 were considered significant and used in the following analyses.

Statistical analyses of gene ontology (GO) and pathways (PA) terms were applied to the lists of differentially expressed genes and have been described in detail previously. 8 Briefly, given a list of differentially expressed genes, we scored each GO or PA term by comparing the number of genes in the list annotated by this term to the expected number of such genes based on the total number of genes annotated by this term in the whole array using the hypergeometric distribution. GO term scores with P 0.005 were considered significant. The pathway analysis database contains 360 curated pathways collected at Stanford from various sources such as KEGG, BioCarta, and SPAD. PA terms with Z scores 2.6 (corresponding P =0.0047) were considered significant.

Cell Proliferation and Migration Assays

CASM and SVSM were cultured in 96-well plates. At 80% confluence, cells were treated with OxLDL (40 µg/mL), PDGF (10 ng/mL), 10% fetal bovine serum (FBS), or vehicle for 24 hours in the presence of BrdU for proliferation assays after incubating overnight in serum-free medium. The Cell Proliferation ELISA kit (Roche, Germany) was used according to the manufacturer?s instructions. For the antibody immunoneutralization experiment, SVSM were first incubated with 25 µg/mL of polyclonal anti-IGF1 receptor (Upstate, Charlottesville, Va) in serum-free medium for 24 hours before adding indicated agents.

Cell migration assays were done using the Innocyte Cell Migration Assay kit (Oncogene, San Diego, Calif) in a 96-well plate. SMCs plated on the upper chambers migrated through the 8-µm pore-size of the membrane in response to chemoattractant molecules added in the lower chambers. Migrated cells attached on the lower side of the membrane were stained with a fluorescent dye, dislodged from the membrane and quantified in a fluorescent reader. Three separate experiments each with 6 replicate wells for each condition were performed for both cell proliferation and migration assays.

Quantitative Real-Time Polymerase Chain Reaction

Total RNA (5 µg) from the same preparations used for microarray experiments was converted to cDNA using MMLV reverse transcriptase. The cDNA was amplified in triplicate on the ABI PRISM 7900HT with primers and probes (Applied Biosystems, Foster City, Calif). Gene expression levels were normalized to corresponding 18S internal controls.

Results

Arterial and Venous SMC Subtypes Have Different Proliferative and Migratory Responses

We compared CASM and SVSM proliferative and migratory responses to OxLDL, PDGF, and FBS using primary cultures of SMC from coronary artery and saphenous vein. FBS strongly induced increases in proliferation and migration in both CASM and SVSM (data not shown), suggesting that both SMC subtypes used in these experiments are fully responsive. However, the proliferative and migratory responses to OxLDL and PDGF were different. OxLDL elicited opposite proliferative responses between CASM and SVSM, moderately inhibiting both proliferation and migration in CASM ( Figure 1A and 1 B) but significantly promoting proliferation in SVSM ( Figure 1 A). PDGF also induced a greater proliferative response in SVSM than CASM ( Figure 1 A), whereas the migratory response was similar ( Figure 1 B).

Figure 1. Proliferative and migratory responses to OxLDL and PDGF in CASM and SVSM. Bar graph represents the fold change of treated versus untreated cells for cell proliferation (A) and migration (B). Results were from 3 separate cell cultures with 6 replicates for each treatment condition.

We further tested if IGF1 signaling plays a role in OxLDL-induced SVSM proliferation. We found that an antibody against the IGF1 receptor retarded SVSM proliferation induced by OxLDL and PDGF by 35% ( Figure 2 ). IGF1R antibody also reduced FBS-induced proliferation in SVSM by 45% (data not shown).

Figure 2. Inhibition of SVSM proliferation by antibody against IGF1 receptor. Percentage of inhibition was plotted for the inhibition of SVSM proliferation by IGF1R antibody. 25 µg/mL of anti-IGF1R antibody was pre-incubated in serum-free medium for 24 hours before adding OxLDL and PDGF.

OxLDL and PDGF Differentially Modulate Growth-Related Gene Expression Between CASM and SVSM

Lists of differentially expressed genes from 3 independent primary cultures of CASM and SVSM cells both in the untreated state and in response to OxLDL and PDGF were generated by the criteria of 1.5-fold change in expression level with P 0.001 as described. OxLDL and PDGF elicited strong and dramatic differences in global gene expression responses and activated distinct signaling pathways between CASM and SVSM (supplemental Tables I to V, available online at http://atvb.ahajournals.org). There were 2262 genes (13.8%) altered by OxLDL in CASM and 835 genes (5.1%) in SVSM. Statistical analysis of Gene Ontology (GO) and Pathway (PA) terms revealed distinct pathways associated with these differentially expressed genes ( Figure 3 ). In response to OxLDL, a set of 145 genes primarily related to stress, inflammatory, and immune responses was altered in the same direction in both CASM and SVSM (supplemental Table VII), whereas there were 76 genes with opposite changes between CASM and SVSM, including EDNRB, HMOX1, DCN, ABCA1, IGFB2 and IGFB4, IFI27, CTGF, and SERPINE1 ( Table ).

Figure 3. Over-represented pathways from statistical analysis of pathway (PA) and gene ontology (GO) terms. The heat map shows the significant terms for pathway analysis (A) and GO analysis (B). Red indicates terms associated with upregulated genes over control, or terms over-represented in untreated CASM; green, terms associated with downregulated genes over control, or terms over-represented in untreated SVSM. CA indicate CASM; SV, SVSM. Terms with P 0.005 are considered significant. Not all significant terms are shown. The full list of significant PA and GO terms can be found in supplemental Tables VIII and IX.

List of Selected Genes With Opposite Changes Between CASM and SVSM After OxLDL

In particular, as outlined, OxLDL differentially regulated several important growth-related genes and pathways ( Figure 4 ), providing a possible molecular mechanism that may contribute to the phenotypic differences observed between CASM and SVSM.

Figure 4. Selected genes differentially expressed between CASM and SVSM. Red indicates upregulated genes over control, or genes having a higher level in untreated CASM; green, downregulated genes over control, or genes having a higher level in untreated SVSM. Only significant genes with a P 0.001 and fold change 1.5 were shown.

IGFBPs

OxLDL upregulated IGFBP2, IGFBP3, IGFBP4, IGFBP5, and IGFBP6 in CASM, but downregulated IGFBP2 and IGFBP4 in SVSM ( Table; Figure 4 A). The direction of IGFBP changes is consistent with the observed OxLDL-induced proliferation and migration between the 2 SMC subtypes ( Figure 1 ). There was no significant change in IGF1 and IGF1R expression. Furthermore, PDGF-BB inhibited IGFBP2 but induced IGFBP3 in SVSM but induced both IGFBP2 and IGFBP3 in CASM ( Figure 4 A). Correspondingly, PDGF also caused greater cell proliferation in SVSM than in CASM ( Figure 1 A).

Cell Cycle Regulatory Genes

OxLDL in CASM significantly upregulated growth-inhibitory genes controlling cell cycling, like cyclin-dependent kinase inhibitors CDKN1A, CDKN2C, and CDKN2D, and cyclin-dependent kinase 3 (CDK3), whereas downregulating cyclin A (CCNA) and CCNA1 (supplemental Table I). In contrast, OxLDL in SVSM upregulated CDKN1A but downregulated cyclin A ( Table; supplemental Table II). These results suggest that transcriptional regulation of cell cycle gene expression may be involved in SMC proliferation. This is in agreement with the fact that IGF1R antibody only partially inhibited SVSM proliferation ( Figure 2 ).

Phosphatidylinositol 3-Kinase and NF- B Pathways

Both phosphatidylinositol 3-kinase (PI3-K) and NF- B pathway associated genes were over-represented in the set of genes induced by OxLDL in SVSM ( Figure 3 A), which agrees with our finding of OxLDL-induced cell proliferation and migration only in SVSM ( Figure 1 ). The upregulated genes associated with the PI3-K pathway in SVSM included protein tyrosine phosphatase nonreceptor type 14 (PTPN14), PTPN11, inositol polyphosphate-1-phosphatase (INPP1), and INPP5D, whereas downregulated genes included dual specific phosphatase 1 (DUSP1) and protein-tyrosine phosphatase receptor type C (PTPRC or CD45), both negative regulators of cytokine receptor signaling.

OxLDL and PDGF Modulate Genes and Molecular Pathways Involved in Atherosclerosis in SMCs

Several families of genes that may be associated with atherosclerosis were also differentially regulated by OxLDL and PDGF, including chemokines and cytokines, extracellular matrix (ECM) genes, genes related to PG pathways, and genes specific for vascular cell types.

Chemokines and Cytokines

OxLDL and PDGF differentially regulated several important cytokines and chemokines between CASM and SVSM ( Figure 4B and 4 C). For example, chemokine CXCL12 was dramatically upregulated by OxLDL (44-fold) and PDGF in CASM, whereas in SVSM it was not changed by OxLDL but downregulated by PDGF ( Figure 4 B). Both OxLDL and PDGF downregulated the expression of CXCL10 and CXCL11. Interestingly, the direction change of CXCL12 expression was opposite to its receptor CXCR4 ( Figure 4 B). OxLDL and PDGF also induced differential patterns of expression of cytokines and their regulatory proteins ( Figure 4 C).

ECM Genes

Genes from several ECM gene families were differentially expressed between CASM and SVSM in response to OxLDL and PDGF ( Figure 3 D). Among matrix metalloproteinase (MMP) genes, MMP2 was the isoform most significantly upregulated in CASM, whereas MMP10 was downregulated. Also, OxLDL and PDGF induced TIMP1 and TIMP2 expression in CASM but not in SVSM. OxLDL induced differential regulation of collagen (COL) and integrin genes, the major receptors for ECM-mediated cell adhesion, migration, proliferation, and differentiation in CASM but had effects on only a few of these genes in SVSM ( Figure 3 D). Interestingly, decorin and fibronectin, 2 important ECM genes that negatively regulate cell growth, were significantly upregulated in CASM but downregulated in SVSM ( Table ), which is consistent with the SMC proliferative and migratory responses.

ECM and SMC Marker Genes

In CASM but not in SVSM, OxLDL significantly ( P <0.001) suppressed (by 1.7- to 44-fold) several endothelial cell (EC) marker genes 9 including END1, VCAM1, PECAM1, CDH5, MMRN1, ESM1, LIPG, THBD, and C1QR1; however, the change of VWF expression was not significant (supplemental Tables I and II). Also, OxLDL in CASM upregulated several smooth muscle genes including SM22-, SM -actin, vimentin, and ADD3, but not the cardiac muscle actin-.

PG (PTG) Pathway Genes

OxLDL significantly induced expression of PG-endoperoxide synthase 1 (COX1) in both CASM and SVSM (supplemental Table VII). However, induction of PTG receptors (PTGFR, PTGER3 and PTGD2R) was only seen in SVSM (supplemental Tables I and II), suggesting that PG signaling may play an important role in OxLDL-mediated effects in SVSM.

SMCs From CASM and SVSM Have Dramatically Different In Vitro Basal Gene Expression Programs

CASM and SVSM in a basal state, in which cells were deprived of serum for 36 hours, had clearly different gene expression patterns, with 1504 genes (9.2%) more highly expressed in CASM, and 695 genes (10.3%) more highly expressed in SVSM (supplemental Table VI).

Statistical analysis of pathway (PA) and gene ontology (GO) terms revealed that upregulated genes in pathways related to apoptosis, inflammation (CD40L, IL1R, CXR4, NF- B), and lipid biosynthesis were significantly over-represented in CASM ( Figure 2 ). However, upregulated genes related to cell growth, IGF binding, oxidoreductase activity, thrombin, extracellular matrix, fibrinolysis, and complement were significantly over-represented in SVSM when compared with CASM ( Figure 2 ). Interestingly, the relative expression levels of many of these genes seemed closely related to their responses to OxLDL and PDGF ( Figure 4 ). Among these differentially expressed genes (supplemental Table VI), representatives of several particularly interesting gene families include the following.

IGFBPs

Four of the 5 IGFBPs expressed in human vascular SMCs (IGFBP2, IGFBP3, IGFBP4, and IGFBP6) were significantly higher in SVSM than in CASM ( Figure 4 A). IGFBP3 had the highest differential expression (122-fold).

Chemokines and Cytokines

Several proinflammatory cytokines (eg, IL-1, IL-6, and IL-8) and -Chemokine ligands (CXCL1, CXCL3, CXCL5, CXCL6, and CXCL8) had significantly higher expression levels in CASM than in SVSM ( Figure 4 B), with the exception of CXCL12 and its receptor CXCR4, which had markedly higher (5- to 100-fold) expression levels in CASM than in SVSM ( Figure 4 B). In contrast, 4 of the 5 ß-chemokines (CCL17, CCL22, CCL24 and CCL28) were significantly expressed more highly in SVSM than CASM, with the exception being CCL2 (MCP-1), which was higher (by 28-fold) in CASM than in SVSM ( Figure 4 B).

Endothelial and SMC Marker Genes

Endothelial-specific genes, 9 including EDN1, vascular cell adhesion molecule (VCAM), PE-CAM1, MMRN1, ICAM1, and CDH5, had a relatively higher expression level (2- to 32-fold) in CASM than in SVSM (supplemental Tables II and VI). THBD was an exception with 7-fold higher in SVSM than in CASM. However, expression of several smooth muscle marker genes, including smoothelin, smooth muscle alpha actin 2, and vimentin, were higher in SVSM than in CASM (Supplemental Tables II and VI). Although the significance of endothelial marker gene expression in SMCs is unknown, the higher expression of some typical SMC markers in SVSM may suggest that untreated venous smooth muscle cells exist in a more "differentiated" state.

ECM Genes

Notably, tissue inhibitor of metalloproteinase (TIMP) 1, TIMP2, and TIMP3 and MMP 2 were higher in SVSM, whereas MMP10, MMP3, MMP20, and MMP26 were higher in CASM. ECM genes play important roles in cell proliferation and migration. Higher levels of all 3e TIMPs in SVSM could reduce the degradation of extracellular matrix, and thus inhibit proliferation and migration ( Figure 4 D).

We selected 14 genes to perform Taqman real-time PCR for the same RNA samples used in our microarray hybridization. The relative ratios from Taqman analysis correlated well to array hybridization results, with a correlation coefficient of 0.88 ( Figure 5 ). Furthermore, the differential expression levels between CASM and SVSM of certain genes were consistent with relative expression levels in in vivo tissues between human arteries (aorta, radial, and chest arteries) and saphenous vein. For example, genes related to ECM, like decorin, lumican, TIMP3, COL6A3, and CD44, were higher in intact veins than arteries, whereas several genes related to cellular signaling, like RGS5, RAC2, RAG, DAF, G3BP, and stathmin1, were higher in arteries (data not shown).

Figure 5. Confirmation of microarray results with Taqman real-time reverse-transcription PCR. Bar graph represents the relative ratios of CASM versus SVSM for the selected genes. Real-time RT-PCR was used to quantify the selected gene expression levels for the same RNA samples hybridized on arrays.

Discussion

Differences in Gene Expression Patterns and Proliferative/Migratory Responses to OxLDL and PDGF Between CASM and SVSM

The evolution of an atheroma involves abnormal proliferation and migration of SMCs induced by cytokines and growth factors like PDGF. OxLDL plays an important role in atherogenesis, including foam cell formation and induction of adhesion molecules, chemoattractant cytokines, and growth factors involved in inflammatory process, cell proliferation, and apoptosis. It is interesting to find that OxLDL induced opposite changes in proliferation and migration between the venous and arterial SMC subtypes, with the inhibition of these responses in CASM and stimulation in SVSM. The stronger SVSM proliferative response to PDGF is consistent with a previous report of greater PDGF-induced growth in explant culture of saphenous vein than chest artery. 3 Although the contrary effects (depending on the degree of lipid oxidation and dose used) of OxLDL on cell proliferation in single type of SMCs has been reported, 10,11 direct comparison of the arterial and venous SMC subtypes to the same dose and exposed time of OxLDL has not been published.

Global gene expression profiling provides hints to the molecular mechanisms underlying the different proliferative responses to OxLDL and PDGF between the 2 SMC subtypes. OxLDL upregulated multiple isoforms of IGFBP genes in CASM and downregulated IGFBPs in SVSM, which correlates well with the functional responses. Evidence suggests that IGFBPs may play an important role in the regulation of SMC growth and in atherogenesis by the inhibition of IGF activities and direct inhibitory effects. 12-15 For example, expression levels of IGFBP2 and other IGFBPs are increased in SMCs from atherosclerotic lesions. 14,15 Overexpression of IGFBP2 and IGFBP4 in SMCs in vitro inhibits proliferation and migration. Furthermore, transgenic overexpression of IGFBP2 and IGFBP4 in in vivo smooth muscle causes apoptosis and hypoplasia. Our results showing the differential expression of IGFBPs between CASM and SVSM and the inhibition of SVSM proliferations by an IGF1R antibody support an important role for IGFBPs in determining proliferative responses between venous and arterial SMCs.

However, the fact that IGF1 receptor blockade only partially inhibited SMC proliferation suggests the direct effects of IGFBPs on SMC proliferation or/and other molecular mechanisms involved. Indeed, we found that OxLDL also regulated genes controlling cell cycles, which is consistent with the report that the growth-inhibitory effect of OxLDL in arterial SMCs is caused mainly by inhibition of nuclear translocation of cell cycle proteins, and not through apoptosis. 16 Furthermore, known pathways associated with proliferation and migration, PI-3K and NF- B pathways, were over-represented with upregulated genes only in SVSM, whereas NF- B was associated with downregulated genes in CASM. The outcome of NF- B pathway activation by OxLDL can result in cell growth or inhibition, depending on cell type, OxLDL concentration, and exposure time. 10 Taken together, these results suggest that different proliferative responses to OxLDL and PDGF between CASM and SVSM are mediated through regulation of IGF signaling (IGFBPs), genes controlling cell cycling, and PI3-K/NF- B pathways.

The finding that venous SMCs had stronger responses to OxLDL and PDGF than arterial SMCs agrees with the fact that venous SMCs under stress, such as exposure to the arterial hemodynamic environment in bypass surgery, are prone to develop accelerated atherosclerosis, an important pathological process in vein graft disease. 5

OxLDL and PDGF also regulate several gene families closely related to atherosclerosis, including cytokines and chemokines, specific markers for endothelial and smooth muscle cells, and ECM genes. Chemokines, originating mainly from leukocytes, are structurally divided into CXC ( -chemokines) and CC (ß-chemokines) subfamilies based on the arrangement of the first 2 of 4 conserved cysteine residues. Through activating G-protein coupled receptors, chemokines are potent activators and chemoattractants of leukocytes and implicated in many disease processes including atherosclerosis. 17,18 Although it has been reported that some -chemokines, CXCL10 and CXCL11 through binding to CXCR3, induce leukocyte migration and modulation of adhesion molecule expression, and CXCL12 through CXCR4 is involved in fusion of human immunodeficiency virus-1 (HIV-1) with CD4+ lymphocytes, it has not been recognized that OxLDL directly and distinctly regulates these chemokines in venous and arterial SMC subtypes.

Alterations of SMC differentiation state are a hallmark of vascular diseases including atherosclerosis and arterial re-stenosis. 19 The OxLDL-mediated suppression of endothelial genes and stimulation of SMC genes suggests phenotypic change of CASM to a more differentiated state, which is consistent with our finding that OxLDL inhibited CASM proliferation and migration. Interestingly, OxLDL suppressed EDN1 expression but dramatically induced END1B receptor (EDNRB) ( Table ). The suppression of EDN1 may be caused by OxLDL-induced upregulation (by 2.4-fold) of hepatocyte growth factor. 20 The dramatic upregulation of EDNRB in CASM by OxLDL supports the role of endothelin signaling in atherosclerosis. 21

Sukhanov et al reported that OxLDL induced changes of gene expression in human aortic smooth muscle related to apoptosis, cell-cell interaction, and lipid metabolism, 16 consistent with our findings in coronary artery SMCs. However, direct comparison of differential gene expression induced by OxLDL between these 2 SMC subtypes has not previously been reported.

Basal Gene Expression Programs in Venous SMCs Are Enriched With Atheroprotective Genes

The gene expression patterns in basal state between venous and arterial SMC subtypes were markedly different. Molecular signatures that define the venous and arterial SMC subtypes include IGFBPs, ß-chemokines, TIMPs, and some SMC marker genes, which were notably expressed higher in SVSM, and -chemokines and endothelial marker genes, which were higher in CASM. The relative expression levels of these genes between venous and arterial SMCs seem closely related to their responses to OxLDL and PDGF ( Figure 4 ).

We have previously reported complementary data from similar experiments documenting inherent differences in endothelial cells from different vascular beds, correlating gene expression to disease relevant pathways. 8 Those studies suggested that atheroprotective gene expression programs in saphenous vein endothelial cells may contribute to atherosclerosis resistance in vein. In this study, we found that several families of genes related to key atheroprotective pathways in the basal state of SMCs were relatively higher in SVSM than in CASM, including fibrinolysis, complement, inhibition of matrix metalloproteinases, and anti-growth IGFBPs and SMC markers, whereas proinflammatory -chemokines and cytokines and genes related to apoptosis were higher in CASM. These results suggest that expression of atheroprotective genes in saphenous vein SMCs also contributes to venous athero-resistance under normal anatomic conditions.

In conclusion, SMCs from coronary artery and saphenous vein in vitro exhibited distinct molecular subtypes. Venous SMCs cultured in basal conditions had relatively higher expression levels of genes related to key pathways contributing to atherosclerosis resistance. However, in response to atherogenic stimulation, venous SMCs appeared to modulate gene expression and functional responses in directions favoring atherogenesis, which could be part of the molecular basis of accelerated atherosclerosis in vein graft disease. Our results suggest molecular mechanisms underlying differential susceptibility to atherosclerosis of the 2 SMC subtypes in various pathophysiological conditions.

【参考文献】
  Manabe I, Owens GK. The smooth muscle myosin heavy chain gene exhibits smooth muscle subtype-selective modular regulation in vivo. J Biol Chem. 2001; 276: 39076-39087.

Fujita H, Shimokado K, Yutani C, Takaichi S, Masuda J, Ogata J. Human neonatal and adult vascular smooth muscle cells in culture. Exp Mol Pathol. 1993; 58: 25-39.

Yang Z, Oemar BS, Carrel T, Kipfer B, Julmy F, Luscher TF. Different proliferative properties of smooth muscle cells of human arterial and venous bypass vessels: role of PDGF receptors, mitogen-activated protein kinase, and cyclin-dependent kinase inhibitors. Circulation. 1998; 97: 181-187.

Liu B, Itoh H, Louie O, Kubota K, Kent KC. The role of phospholipase C and phosphatidylinositol 3-kinase in vascular smooth muscle cell migration and proliferation. J Surg Res. 2004; 120: 256-265.

Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation. 1998; 97: 916-931.

Ross R. Atherosclerosis-an inflammatory disease. N Engl J Med. 1999; 340: 115-126.

Libby P. Inflammation in atherosclerosis. Nature. 2002; 420: 868-874.

Deng D, Tsalenko A, Vailaya A, Ben-Dor A, Estay I, Tabibiazar R, Kincaid R, Yakhini Z, Bruhn L, Quertermous T. Differences in vascular bed disease susceptibility reflect differences in gene expression response to atherogenic stimuli. Circ Res. 2006; 98: 200-208.

Ho M, Yang E, Matcuk G, Deng D, Sampas N, Tsalenko A, Tabibiazar R, Zhang Y, Chen M, Talbi S, Ho YD, Wang J, Tsao PS, Ben-Dor A, Yakhini Z, Bruhn L, Quertermous T. Identification of endothelial cell genes by combined database mining and microarray analysis. Physiol Genomics. 2003; 13: 249-262.

Robbesyn F, Salvayre R, Negre-Salvayre A. Dual role of oxidized LDL on the NF-kappaB signaling pathway. Free Radic Res. 2004; 38: 541-551.

Bjorkerud B, Bjorkerud S. Contrary effects of lightly and strongly oxidized LDL with potent promotion of growth versus apoptosis on arterial smooth muscle cells, macrophages, and fibroblasts. Arterioscler Thromb Vasc Biol. 1996; 16: 416-424.

Scheidegger KJ, James RW, Delafontaine P. Differential effects of low density lipoproteins on insulin-like growth factor-1 (IGF-1) and IGF-1 receptor expression in vascular smooth muscle cells. J Biol Chem. 2000; 275: 26864-26869.

Delafontaine P, Song YH, Li Y. Expression, regulation, and function of IGF-1, IGF-1R, and IGF-1 binding proteins in blood vessels. Arterioscler Thromb Vasc Biol. 2004; 24: 435-444.

Grant MB, Wargovich TJ, Ellis EA, Tarnuzzer R, Caballero S, Estes K, Rossing M, Spoerri PE, Pepine C. Expression of IGF-I, IGF-I receptor and IGF binding proteins-1, -2, -3, -4 and -5 in human atherectomy specimens. Regul Pept. 1996; 67: 137-144.

Zhang QJ, Goddard M, Shanahan C, Shapiro L, Bennett M. Differential gene expression in vascular smooth muscle cells in primary atherosclerosis and in stent stenosis in humans. Arterioscler Thromb Vasc Biol. 2002; 22: 2030-2036.

Zettler ME, Prociuk MA, Austria JA, Zhong G, Pierce GN. Oxidized low-density lipoprotein retards the growth of proliferating cells by inhibiting nuclear translocation of cell cycle proteins. Arterioscler Thromb Vasc Biol. 2004; 24: 727-732.

Holm T, Damas JK, Holven K, Nordoy I, Brosstad FR, Ueland T, Wahre T, Kjekshus J, Froland SS, Eiken HG, Solum NO, Gullestad L, Nenseter M, Aukrust P. CXC-chemokines in coronary artery disease: possible pathogenic role of interactions between oxidized low-density lipoprotein, platelets and peripheral blood mononuclear cells. J Thromb Haemost. 2003; 1: 257-262.

Shin WS, Szuba A, Rockson SG. The role of chemokines in human cardiovascular pathology: enhanced biological insights. Atherosclerosis. 2002; 160: 91-102.

Kumar MS, Owens GK. Combinatorial control of smooth muscle-specific gene expression. Arterioscler Thromb Vasc Biol. 2003; 23: 737-747.

Haug C, Schmid-Kotsas A, Zorn U, Bachem MG, Schuett S, Gruenert A, Rozdzinski E. Hepatocyte growth factor is upregulated by low-density lipoproteins and inhibits endothelin-1 release. Am J Physiol Heart Circ Physiol. 2000; 279: H2865-H2871.

Haug C, Schmid-Kotsas A, Zorn U, Schuett S, Gross HJ, Gruenert A, Bachem MG. Endothelin-1 synthesis and endothelin B receptor expression in human coronary artery smooth muscle cells and monocyte-derived macrophages is up-regulated by low density lipoproteins. J Mol Cell Cardiol. 2001; 33: 1701-1712.

作者单位：David Xing-Fei Deng; Joshua M. Spin; Anya Tsalenko; Aditya Vailaya; Amir Ben-Dor; Zohar Yakhini; Phil Tsao; Laurakay Bruhn; Thomas QuertermousFrom Agilent Laboratories (D.X.-F.D., A.T., A.V., A.B.-D., Z.Y., L.B.), Palo Alto, Calif; and the Donald W. Reynolds Cardiovascular Clinical Research Center (