Crucial Role of Stromal Cell–Derived Factor-1α in Neointima Formation After Vascular Injury in Apolipoprotein E–Deficient Mice
Background— Recent evidence indicates that stromal cell–derived factor-1α (SDF-1α) is expressed in human atherosclerotic plaques, whereas high plasma levels are clinically associated with stable coronary artery disease. Herein, we investigate the involvement of SDF-1α in neointimal formation after vascular injury.
Methods and Results— SDF-1α was detected by immunohistochemistry in carotid arteries of apolipoprotein E–deficient (apoE−/−) mice at various stages of neointima formation after wire-induced injury. Double immunofluorescence revealed that SDF-1α staining was mostly confined to smooth muscle cells (SMCs). Furthermore, SDF-1α plasma levels peaked 1 day after vascular injury. Treatment of apoE−/− mice after carotid injury with a neutralizing SDF-1α monoclonal antibody for 3 weeks reduced neointimal lesion area by 44% (n=5, P<0.05) compared with isotype control. In SDF-1α antibody–treated apoE−/− mice, neointimal SMC content was decreased (21.7±2% versus 39.4±4%, n=5, P=0.005), whereas the relative content of neointimal macrophages remained unchanged. As shown by flow cytometry, carotid injury resulted in a marked expansion of circulating Sca-1+lineage− progenitor cells (PBPCs) in the peripheral blood of apoE−/− mice after 1 day, which was mediated by SDF-1α. Systemic injection of isolated PBPCs after vascular injury demonstrated their recruitment to neointimal lesions, where they can adopt an SMC-like phenotype.
Conclusions— SDF-1α plays an instrumental role in neointimal formation after vascular injury in apoE−/− mice by regulating neointimal SMC content. This contribution appears to be attributable to SDF-1α–dependent recruitment of circulating SMC progenitor cells.
Received August 27, 2003; revision received September 19, 2003; accepted September 22, 2003.
Intimal fibrous proliferation and inward remodeling are the main features of a disease process caused by excessive vascular repair after mechanical injury, eg, by balloon angioplasty, stent implantation, or atherectomy to treat patients with obstructive atherosclerosis.1,2 This phenomenon is termed restenosis and limits therapeutic revascularization. Initial endothelial denudation leads to an exposure of subendothelial matrix, which can precipitate the adhesion of activated platelets and fibrin deposition, thereby supporting the inflammatory recruitment of leukocytes. The development of neointimal hyperplasia is attributable to the accumulation of dedifferentiated smooth muscle cells (SMCs).3,4
Proinflammatory CC chemokines like monocyte chemotactic protein-1 and RANTES have been shown to aggravate neointima formation and neointimal monocyte recruitment after vascular injury in hypercholesterolemic animals.5,6 The CXC chemokine stromal cell–derived factor-1α (SDF-1α), which is essential for stem cell mobilization/homing and organ system vascularization,7–10 is also highly expressed in human atherosclerotic plaques and effectively activates platelets in vitro.11 Therefore, a participation of SDF-1α in human atherothrombotic disease has been assumed. Conversely, reduced SDF-1α plasma levels have been found to be associated with symptomatic coronary artery disease, implicating an anti-inflammatory role for SDF-1α.12
To clarify the role of SDF-1α in neointimal plaque formation, we investigated lesional SDF-1α protein expression and SDF-1α plasma levels after wire-induced arterial injury in hypercholesterolemic mice. Furthermore, neutralization of SDF-1α after carotid injury in apolipoprotein E–deficient (apoE−/−) mice by application of a blocking antibody was evaluated for its effect on neointimal lesions. Bone marrow–derived cells have been shown to contribute to neointimal SMC content,13–15 and circulating SMC progenitors have been demonstrated in humans.16 We therefore determined whether SDF-1α is involved in the mobilization of progenitor cells into the peripheral blood and whether these cells can be recruited to neointimal lesions and adopt an SMC phenotype.
Female, 8-week-old apoE−/− mice (C57BL/6 background, M&B, Ry, Denmark) were fed an atherogenic diet containing 21% fat for 1 week before and up to 4 weeks after injury. Mice were anesthetized with ketamine hydrochloride (80 mg/kg IP) and xylazine (5 mg/kg IP), a 0.36-mm flexible angioplasty guidewire was advanced by 1 cm via a transverse arteriotomy of the external carotid artery, and endothelial denudation of the common carotid artery was achieved by 3 rotational passes.6,17 At 1 day and 2 weeks after wire injury, carotid arteries and plasma samples were collected (n=4 per group). In subgroups (n=5 per group), blocking SDF-1α monoclonal antibody (mAb; clone 79014.111, R&D Systems) or IgG1 isotype control was injected 1 day before injury (100 μg IP) and subsequently twice weekly (50 μg IP) for 3 weeks. Serum levels of total cholesterol and triglycerides were determined after 3 weeks of mAb treatment. Furthermore, male peripheral blood progenitor cells (PBPCs; 100 000±25 000 per mouse, n=6) were administered 30 minutes after wire injury by direct percutaneous intracardial injection, and carotid arteries were harvested after 4 weeks. In a separate set of experiments, apoE−/− mice received a single injection of SDF-1α mAb (100 μg; n=3 per group) or isotype control 2 weeks after wire-induced carotid injury. Fluorescence-labeled PBPCs were then administered intracardially 12 hours after mAb administration and 30 minutes before collection of arteries and blood for flow cytometric analysis. After in situ perfusion fixation (4% paraformaldehyde) and paraffin embedding, tissue sections from both common carotid arteries were obtained. In some animals (n=8), carotid tissue was snap-frozen in liquid nitrogen 2 and 4 weeks after arterial injury and PBPC application. Animal experiments were approved by the local authorities and complied with the German animal protection law.
Histomorphometry, Immunohistochemistry, and Immunofluorescence
Serial 5-μm cross sections of left common carotid arteries were stained with modified Movat’s pentachrome.18 Within a standardized distance (0 to 500 μm) from the bifurcation, areas within lumen and internal and external elastic laminae were measured by planimetry in 10 sections per mouse with Diskus Software (Hilgers) to determine neointimal and medial areas.
Quantitative immunohistochemistry of α-smooth muscle actin (α-SMA; clone 1A4, Dako) and Mac-2 (clone M3/38, Cedarlane) was performed with Vectastain ABC-AP kit. The neointimal areas stained for SMCs or macrophages were determined in digitized images (5 sections per mouse) and were expressed as percentage of total neointimal area.
Immunohistochemistry for SDF-1α was performed in sections reacted with mouse anti-m/hSDF-1α mAb (clone 79018.111; R&D Systems) or isotype control. Primary mAbs were detected after antigen retrieval (citrate buffer, pH 6.0, 0.05% Tween-20, in a microwave oven) and blocking of unspecific protein binding (5% horse serum) with biotin-conjugated secondary antibody and a preformed avidin/biotinylated alkaline phosphatase complex (Vectastain ABC-AP kit) visualized with Vector Red Substrate kit (all from Vector Labs). The fluorescent Vector Red product also served for double immunofluorescence staining of SDF-1α and smooth muscle myosin heavy chain (SMMHC; polyclonal rabbit antibody, X/80, provided by Dr D. Drenckhahn, University of Würzburg, Germany) or Mac-2 reacted with a secondary FITC-conjugated antibody. Digital images were recorded with a fluorescence microscope (Leica DMLB microscope) connected to a CCD camera (JVC).
Enzyme-Linked Immunosorbent Assay
Mouse SDF-1α standard or EDTA plasma samples were incubated in 96-well plates coated with SDF-1α mAb (clone 79018.111) and reacted with biotinylated SDF-1α antibody (all from R&D Systems), streptavidin–horseradish peroxidase, and tetramethyl benzidine peroxidase substrate for detection. Absorbance was read at 450 nm and background corrected, and SDF-1α concentrations were calculated with standards.
Isolation and Quantification of Murine Sca-1+lin−/lo PBPCs
Sca-1+ cells were isolated from pooled peripheral blood of male C57BL/6 mice (M&B, Ry, Denmark) by magnetic separation with an Sca-1 MultiSort-Kit (Miltenyi Biotec) and depleted of lin+ cells with FITC-conjugated lineage marker antibodies (anti-mCD11b, anti-CD45R, and anti-Gr-1, Pharmingen; anti-Ter-119, MoBiTec; anti-mCD3, Serotec) and anti-FITC microbeads (Miltenyi). Sca-1+lin−/lo cells (purity >95%) were labeled with CM-DiI (Molecular Probes) and injected within 1 hour after isolation.
PBPCs were quantified in anticoagulated blood collected from apoE−/− mice after wire injury or sham operation. After erythrocyte lysis, cells reacted with phycoerythrin-conjugated anti-mSca-1 (clone E13–161.7, Pharmingen) and the FITC-conjugated lineage panel were analyzed by flow cytometry (FACSCalibur, BD Biosciences).
In Situ Hybridization and Polymerase Chain Reaction
A biotinylated DNA probe (pY353/B, provided by Dr M. Guttenbach, University of Würzburg, Germany) specifically hybridizing to the murine Y-chromosome was used for the detection of injected PBPCs by in situ hybridization. Deparaffinized sections were permeabilized and digested with proteinase K (Sigma) and after postfixation, were denatured and incubated in hybridization solution at 80°C. Biotinylated probe was detected with streptavidin–horseradish peroxidase conjugate and a tyramide-based amplification kit (TSA, Perkin Elmer) with DAB substrate (Vector Labs). Carotid artery sections from uninjected female mice and untreated male mice served as negative and positive controls. In situ hybridization was combined with immunohistochemistry for α-SMA and SMMHC.
Polymerase chain reaction (PCR) for the Sry gene was used to detect male cells in the left carotid artery of female mice 2 and 4 weeks after injury and injection of male PBPCs or in uninjured male and female control mice (n=4 each). After genomic DNA extraction (QIAamp Mini kit), equal amounts of DNA were incubated with QuantiTect Mix containing SYBR Green (Qiagen) and an Sry gene-specific primer pair (forward 5′-CGTGGTGAGAGGCACAAGT-3′; reverse 5′-AACAGGCTGCCAATAAAAGC-3′). PCR amplification of a 360-bp product was performed after initial denaturation at Tann=55°C with a thermocycler (Opticon 2, MJ Research). The PCR products were resolved by 2% agarose gel electrophoresis.
Data represent mean±SEM and were compared by either 2-tailed Student’s t test or 1-way ANOVA followed by Newman-Keuls posttest if appropriate (InStat software, GraphPad). Differences with P<0.05 were considered statistically significant.
We first studied the expression of SDF-1α in injured carotid arteries of apoE−/− mice. Although SDF-1α was not detectable in uninjured arteries of apoE−/− mice on a high-cholesterol diet (Figure 1A), a distinct expression of SDF-1α was evident in medial SMCs directly adjacent to the site of injury as early as 24 hours after injury (Figure 1A). At 2 weeks after injury, a more robust staining of SDF-1α emerged in medial SMCs (Figure 1A). Although luminal cells did not express SDF-1α, a subset of neointimal cells exhibited intense SDF-1α expression at 2 and 4 weeks (Figure 1). Double immunofluorescence staining confirmed that although Mac-2–positive macrophages did not serve as a source of SDF-1α (data not shown), cells expressing SDF-1α were of an SMC phenotype (Figure 1B), as evidenced by the colocalization of SDF-1α with the definitive SMC marker SMMHC, which resulted in yellow staining in the overlay (Figure 1B).
We next tested whether the expression of SDF-1α at sites of vascular injury is accompanied by an elevation of SDF-1α in peripheral plasma. Indeed, SDF-1α plasma levels were increased as early as 24 hours after injury (793±83 versus 357±47 pg/mL in uninjured mice, n=4, P<0.01), and returned to almost baseline levels 2 weeks after injury (Figure 2).
Treatment of apoE−/− mice with the SDF-1α mAb for 3 weeks clearly inhibited neointima formation after arterial injury, as evidenced by the significantly diminished neointimal plaque area (29 020±3470 versus 51 540±6835 μm2, n=5, P<0.05; Figure 3). In contrast, the medial area was hardly affected (Figure 3). No differences in weight or serum lipid levels in mAb-treated mice were observed (data not shown). Moreover, quantitative immunohistochemistry revealed that the content of neointimal SMCs was reduced in SDF-1α mAb–treated mice (21.7±2 versus 39.4±4%, n=5, P=0.005), whereas relative macrophage content in the neointima did not change significantly in SDF-1α mAb–treated mice (Figure 4).
Because a major part of neointimal SMCs can be derived from bone marrow cells,11–13 and circulating SMC progenitors have been demonstrated in humans,14 we investigated whether SDF-1α affects the percentage of peripheral blood Sca+lin− progenitor cells after carotid injury. Arterial injury triggered the rapid expansion of PBPCs in the circulation after 1 day (Figure 5A), which was prevented by the administration of a blocking SDF-1α mAb (4.9±0.8% versus 1.7±0.3% of mononuclear cells, n=4, P<0.01; Figure 5A), with the percentage remaining at levels of uninjured mice (data not shown). Lower levels of PBPCs were observed at 3 weeks after injury, which were not affected by SDF-1α mAb treatment (Figure 5A).
To determine whether SDF-1α can also be implicated in local recruitment of PBPCs to neointimal lesions, we injected DiI-labeled PBPCs 2 weeks after carotid injury and 12 hours after SDF-1α mAb or isotype control application. Whereas PBPCs were found in injured arteries 30 minutes after injection in isotype control–treated mice (Figure 5B), neutralization of SDF-1α abolished the recruitment of injected PBPCs to arterial lesions and resulted in retention of DiI-labeled PBPCs in the circulation (Figure 5B).
To demonstrate the capability of at least a portion of PBPCs to adopt an SMC phenotype in vivo, isolated male PBPCs were injected into apoE−/− mice 30 minutes after carotid injury. The presence of PBPC-derived cells in arterial lesions was evidenced by in situ hybridization (Figure 6A) of the Y-chromosome at 4 weeks after injury. The proportion of Y-positive cells among neointimal cells was determined to be 47.2±6.9%. PBPCs were not detectable in uninjured right carotid arteries (not shown). In addition, PCR amplification of the Sry gene revealed a specific 360-bp product in genomic DNA isolated from injured carotid arteries of female mice 2 and 4 weeks after injection of male PBPCs and in uninjured arteries of male but not female control mice (data not shown). Combined immunohistochemistry revealed that a proportion of neointimal cells originating from male PBPCs expressed α-SMA (40.8±10.8% of Y-positive cells) and SMMHC as a more definitive marker for a terminally differentiated SMC phenotype (Figure 6B). This suggests that the PBPC population recruited to neointimal lesions by SDF-1α contained SMC progenitors.
We investigated the role of the CXC chemokine SDF-1α in neointima formation after carotid injury in hypercholesterolemic mice. SDF-1α protein is detected primarily in SMCs of neointimal lesions in different stages of fibromuscular plaque development. Additionally, plasma levels of SDF-1α increase modestly 1 day after injury. Application of a blocking SDF-1α mAb reduces neointimal area and neointimal SMC content, which indicates an important role during the progression of accelerated atherosclerosis. Because SDF-1α is important in the homeostasis of bone marrow homing and mobilization of progenitor cells,7,19 and neointimal SMCs originate from a bone marrow source of progenitors,13–15 we studied the effect of SDF-1α on progenitor cell mobilization and recruitment after carotid injury. We found that the expansion of PBPCs in the circulation and the accumulation of injected PBPCs in established neointimal lesions were mediated by SDF-1α. Furthermore, PBPCs recruited to injured arteries after systemic administration can adopt an SMC phenotype in vivo, which confirms that SDF-1α–responsive PBPCs contain a subpopulation of SMC progenitor cells.
In line with recent results showing that SDF-1α protein is highly expressed in human atherosclerotic plaques but not in normal vessels,11 we observed an upregulation of SDF-1α protein at different time points in the course of neointima formation after wire-induced carotid injury. As opposed to human atherosclerotic plaques, where SMCs, endothelial cells, and macrophages express SDF-1α,11 neointimal SMCs appear to be the main source of SDF-1α in apoE−/− mice after wire-induced injury. This may be due to the predominance of SMC accumulation in acute vascular repair, which leads to a fibromuscular plaque.
An anti-inflammatory and plaque-stabilizing function has been attributed to SDF-1α in atherosclerotic vascular diseases,12 because symptomatic coronary artery disease, especially unstable angina, was associated with reduced plasma SDF-1α levels. Our findings that vascular injury in apoE−/− mice results in a transient increase of plasma SDF-1α and that treatment with a blocking SDF-1α mAb resulted in a significant decrease of neointimal SMC content, whereas neointimal macrophage content was unchanged, support the concept of a plaque-stabilizing function of SDF-1α in atherosclerotic vascular disease. In addition, our observation that neointimal area decreases in mice treated with the SDF-1α mAb implies a major role in neointimal growth in vascular repair after acute injury, presumably through the effect on neointimal SMCs.
The present data reveal that arterial injury in apoE−/− mice fed a high-cholesterol diet causes the rapid expansion of Sca-1+lin−/lo PBPCs into the circulation. Although this cell population is known to contain hematopoietic stem cells with long-term repopulating capacity,20,21 it may also comprise mesenchymal or other progenitor cells that also express Sca-1. We found that this expansion of PBPCs after carotid injury is mediated by SDF-1α, because plasma SDF-1α levels are increased at the same time, and blocking of SDF-1α with a mAb abolished this expansion. Hattori et al19 reported that plasma SDF-1α elevation after application of an adenoviral vector expressing SDF-1 resulted in the mobilization of hematopoietic cells with repopulating capacity, progenitor cells, and precursor cells. Whereas SDF-1α protein is still highly expressed in neointimal lesions 2 and 4 weeks after injury, plasma levels were elevated at 1 day but not at 2 weeks. SDF-1α binding to proteoglycans has been described recently to enhance SDF-1α–induced migration of hematopoietic progenitor cells.22 Because proteoglycan synthesis is increased after balloon angioplasty and therefore proteoglycans contribute significantly to extracellular matrix in neointimal lesions,23 SDF-1α may be released into the circulation early after endothelial denudation but could be bound to proteoglycans in the developing neointimal matrix, thereby shifting its contribution from early mobilization toward a more continuous neointimal recruitment of PBPCs. Beyond its effects on recruitment within the first weeks, local SDF-1α may further affect the neointimal SMC architecture, thus contributing to arterial remodeling of injured vessels.
To assess the propensity of SDF-1α to mediate progenitor cell recruitment after mobilization, we administered PBPCs while blocking SDF-1α with a mAb 2 weeks after vascular injury. At this time point, plasma SDF-1α levels are no longer elevated, and PBPC expansion in the circulation has vanished (data not shown). We found that injected PBPCs accumulate in injured arteries and contribute to the lesional area. The administration of PBPCs might thus accelerate the natural course of neointima formation. Blocking SDF-1α reduced PBPC accumulation in injured arteries and thus prolonged transit time of PBPCs in the circulation after injection. The ongoing recruitment mediated by SDF-1α at later stages may also account for the transience of PBPC expansion by establishing a novel equilibrium between mobilization and peripheral consumption.
After arterial injury, SDF-1α expression is elicited in medial SMCs, preceding its expression in a subset of neointimal SMCs at later stages. This would be compatible with the concept that resident SMCs initially establish SDF-1α expression, migrate into the intima, and constitute the subpopulation of SDF-1α–producing neointimal SMCs. Although SDF-1α expression has not been detected in hematopoietic cells, including multilineage progenitors,24 it cannot be excluded that PBPCs recruited to neointimal lesions may serve as an alternative source of SDF-1α once they have adopted an SMC phenotype.
Previous studies have convincingly shown that neointimal lesions after vascular injury contain 50% to 60% bone marrow–derived SMCs14 and that 50% of all α-SMA–positive cells originate from the bone marrow.13 Progenitor cells for SMCs also exist in the peripheral blood of healthy humans.16 We therefore investigated whether PBPCs, which are mobilized and recruited to sites of vascular injury by SDF-1α, comprise SMC progenitors in vivo. Our observations confirm the existence of SMC progenitors in the PBPC population and support the concept that SDF-1α mediates SMC progenitor accumulation and thereby contributes to lesion development.
In summary, SDF-1α, a CXC chemokine with an important role in hematopoietic stem cell homing to the bone marrow and vasculogenesis, is abundantly produced by vascular SMCs after vascular injury and is an important regulator in neointima formation and plaque composition. Part of this effect of SDF-1α on neointimal lesions may be explained by SDF-1α–dependent mobilization and recruitment of circulating SMC progenitors into neointimal plaques.
This study was supported by Deutsche Forschungsgemeinschaft grants WE1913/2-3 and WE1913/5-1 to Dr Weber and by the Interdisciplinary Center for Clinical Research on Biomaterials (BMBF grant No. 01 KS 9503/9) to Drs Weber and Schober. We thank Dr P. Hanrath for continuous support and M.S. Roller for technical assistance.
This article originally appeared Online on October 27, 2003 (Circulation. 2003;108:r108–r114).
Mintz GS, Popma JJ, Pichard AD, et al. Arterial remodeling after coronary angioplasty: a serial intravascular ultrasound study. Circulation. 1996; 94: 35–43.
Schwartz SM. The intima: a new soil. Circ Res. 1999; 85: 877–879.
Mori E, Komori K, Yamaoka T, et al. Essential role of monocyte chemoattractant protein-1 in development of restenotic changes (neointimal hyperplasia and constrictive remodeling) after balloon angioplasty in hypercholesterolemic rabbits. Circulation. 2002; 105: 2905–2910.
Schober A, Manka D, von Hundelshausen P, et al. Deposition of platelet RANTES triggering monocyte recruitment requires P-selectin and is involved in neointima formation after arterial injury. Circulation. 2002; 106: 1523–1529.
Peled A, Petit I, Kollet O, et al. Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. Science. 1999; 283: 845–848.
Abi-Younes S, Sauty A, Mach F, et al. The stromal cell-derived factor-1 chemokine is a potent platelet agonist highly expressed in atherosclerotic plaques. Circ Res. 2000; 86: 131–138.
Damas JK, Waehre T, Yndestad A, et al. Stromal cell-derived factor-1α in unstable angina: potential antiinflammatory and matrix-stabilizing effects. Circulation. 2002; 106: 36–42.
Simper D, Stalboerger PG, Panetta CJ, et al. Smooth muscle progenitor cells in human blood. Circulation. 2002; 106: 1199–1204.
Manka D, Collins RG, Ley K, et al. Absence of P-selectin, but not intercellular adhesion molecule-1, attenuates neointimal growth after arterial injury in apolipoprotein E-deficient mice. Circulation. 2001; 103: 1000–1005.
Movat H. Demonstration of all connective tissue elements in a single section. Arch Pathol Med. 1955; 60: 289–295.
Hattori K, Heissig B, Tashiro K, et al. Plasma elevation of stromal cell-derived factor-1 induces mobilization of mature and immature hematopoietic progenitor and stem cells. Blood. 2001; 97: 3354–3360.
Yamamoto Y, Yasumizu R, Amou Y, et al. Characterization of peripheral blood stem cells in mice. Blood. 1996; 88: 445–454.
Goodman JW, Hodgson GS. Evidence for stem cells in the peripheral blood of mice. Blood. 1962; 19: 702–714.
Netelenbos T, Zuijderduijn S, Van Den Born J, et al. Proteoglycans guide SDF-1-induced migration of hematopoietic progenitor cells. J Leukoc Biol. 2002; 72: 353–362.
Strauss BH, Chisholm RJ, Keeley FW, et al. Extracellular matrix remodeling after balloon angioplasty injury in a rabbit model of restenosis. Circ Res. 1994; 75: 650–658.
Tashiro K, Tada H, Heilker R, et al. Signal sequence trap: a cloning strategy for secreted proteins and type I membrane proteins. Science. 1993; 261: 600–603.