Number and Function of Endothelial Progenitor Cells as a Marker of Severity for Diabetic Vasculopathy
Objective— Peripheral arterial disease (PAD) is a threatening complication of diabetes. As endothelial progenitor cells (EPCs) are involved in neovasculogenesis and maintenance of vascular homeostasis, their impairment may have a role in the pathogenesis of diabetic vasculopathy. This study aimed to establish whether number and function of EPCs correlate with PAD severity in type 2 diabetic patients.
Methods and Results— EPCs were defined by the expression of CD34, CD133 and KDR, and quantified by flow cytometry in 127 diabetic patients with and without PAD. PAD severity has been assessed as carotid atherosclerosis and clinical stage of leg atherosclerosis obliterans. Diabetic patients with PAD displayed a significant 53% reduction in circulating EPCs versus non-PAD patients, and EPC levels were negatively correlated with the degree of carotid stenosis and the stage of leg claudication. Moreover, the clonogenic and adhesion capacity of cultured EPCs were significantly lower in diabetic patients with PAD versus patients without.
Conclusions— This study demonstrates that EPC decrease is related to PAD severity and that EPC function is altered in diabetic subjects with PAD, strengthening the pathogenetic role of EPC dysregulation in diabetic vasculopathy. EPC count may be considered a novel biological marker of peripheral atherosclerosis in diabetes.
Diabetic vasculopathy is characterized by high prevalence, early development and rapid progression. Peripheral arterial disease (PAD) is indeed a striking source of morbidity and disability in diabetic subjects and is one leading cause of nontraumatic amputations in western countries.1 Mechanisms accounting for the aggressiveness of diabetic vasculopathy are poorly understood but may depend on a profoundly impaired collateral vascular development.2
Endothelial progenitor cells (EPCs) are a subtype of bone marrow–derived progenitor cells expressing surface antigens of both hematopoietic stem cells and endothelial cells3: they are involved in adult neovasculogenesis and maintainance of vascular integrity.4,5 A growing amount of data suggest that EPCs are altered in clinical conditions characterized by high cardiovascular risk, including type 1 and type 2 diabetes.6–8 Furthermore, we have demonstrated that circulating EPCs are decreased in subjects with PAD, especially in the presence of diabetes.9 Therefore, EPC reduction has been hailed as a novel concept in the pathogenesis of diabetic vascular complications. Additionally, recent data have shown that levels of circulating EPC predict cardiovascular events in patients with coronary artery disease.10,11 However, it is not definitely clear whether EPC alterations have a pathogenic role in the development of cardiovascular disease or simply represent its epiphenomenon. Meanwhile, it has been suggested that EPCs may constitute a novel prototype of cardiovascular biomarkers.12 Again, strong correlations between EPC alterations and disease severity are still lacking. In fact, if dysfunction of progenitor cells had a causative role in vascular disease, one would expect that patients with more severe disease had more profound EPC decrease or dysfunction.
With this background, we aimed to assess the relationships between number and function of EPCs and the severity of atherosclerotic disease in type 2 diabetic patients.
Materials and Methods
The study was approved from the local ethics committee and informed consent was obtained from all subjects. A total of 127 type 2 diabetic subjects, recruited from our Metabolic outpatient clinic and divided into diabetic patients with (n=72) and without (n=55) PAD were included. All patients were screened for other classic risk factors for atherosclerosis: smoking, arterial hypertension, obesity and dyslipidemia. Patients underwent metabolic evaluation including fasting glucose, glycohemoglobin, and lipid profile. Atherosclerosis of carotid vessels was assessed by ultrasonography according to the Society of Radiologists in Ultrasound Consensus Conference.13 The percentage of vessel obstruction was measured first as a continuous variable, and then classified as stenosis of up to 30%, stenosis of 31% to 50%, stenosis of 51% to 70%, or stenosis of 71% to 100%. Diagnosis of atherosclerotic involvement of the lower extremities was assessed noninvasively, and patients were classified according to the Leriche/Fontaine clinical classification of lower-limb atherosclerosis obliterans and according to the revised version of the recommended standards for reports dealing with lower extremity ischemia.14 The presence of coronary artery disease (CAD) and diabetic retinopathy were also assessed. For more details on patient selection criteria and evaluation please see the supplemental data, available online at http://atvb.ahajournals.org.
Quantification of Peripheral Blood Progenitor Cells by Flow Cytometry
Peripheral blood progenitor cells were analyzed for the expression of surface antigens with direct 2- or 3-color flow cytometry (fluorescence-activated cell sorter; Calibur, Becton Dickinson Biosciences) as previously reported.15,16 Progenitor cells were defined by the surface expression of CD34, CD133 and KDR (supplemental Figure I, available online at http://atvb.ahajournals.org). Expression of CD34 and KDR was studied in all subjects, whereas the complete assay, including assessment of CD133 expression, was performed in a subset of 55 (43% of total). For more details on flow cytometry, please see supplemental data.
EPC isolation and culture were performed as previously described.17,18 Briefly, peripheral blood mononuclear cells were plated on fibronectin-coated dishes (Becton Dickinson Biosciences) and grown in supplemented endothelial cell growth medium (Clonetics) for 15 days. Clusters of attaching cells were counted every 3 days starting from day 3. For more details on the culture methods, please see supplemental data.
Characterization of Cultured EPCs
At the end of the growth curve, cultured cells were characterized to confirm their endothelial phenotype in 6 randomly selected subjects. Cells were stained with DiI-acetylated low-density lipoproteins (Molecular Probes) and fluorescein isothiocyanate-lectin (Sigma-Aldrich).19 For further characterization we stained putative EPCs with the endothelial markers von Willebrand factor (vWf), KDR and CD31. To have further methodological confirmation of the endothelial phenotype, cultured EPCs were detached using EDTA and analyzed by flow cytometry for the expression of KDR, CD31, CXCR4 and CD18. For more details on EPC characterization, please see the supplemental data.
Besides studying the clonogenic capacity of patients’ EPCs, we also evaluated in vitro the ability of DiI-LDL-labeled EPCs to adhere to a human umbilical vein endothelial cell monolayer, as previously described.8 For more details on this procedure, please see supplemental data.
In Vivo Functional Characterization of Cultured EPCs
To confirm that isolated cells functionally correspond to EPCs and take part in new vessel growth, CMTMR (5-(and-6)-(4-chloromethyl-benzoyl-amino)-tetramethylrhodamine)–labeled cells were injected into rat ischemic hindlimbs. Muscle capillaries were stained with vWf and nuclei counterstained with Hoescht. Elongated cellular structures double positive for vWf and CMTMR were considered EPC-bearing neovessels. For more details on this procedure, please see supplemental data.
Data are expressed as mean±SEM. All results from flow cytometry are expressed as number of cells/106 cytometric events. Differences between 2 or more groups were analyzed using 2-tailed Student t test and ANOVA, respectively. The Hochberg procedure was applied to account for α-inflation attributable to multiple testing. The χ2 test was used for dichotomous variables. Simple linear regression analysis (Pearson r) was used to assess correlations between severity of PAD and progenitor cell counts. Spearman ρ was also calculated to assess correlation of progenitor cell numbers with severity of lower extremity atherosclerosis obliterans. Statistical associations between progenitor cell counts and clinical conditions or risk factors were examined by multivariate analysis using multiple linear regression. Statistical association between risk factors or clinical conditions and the presence of PAD was assessed by multiple stepwise logistic regression analysis. The discriminatory capacity of EPC count for the presence of PAD was investigated using the receiver operating characteristic curve: cut-off values were obtained by optimizing the sum of sensitivity and specificity. Statistical significance was accepted at P≤0.05.
Subject characteristics are resumed in the Table. As expected, PAD patients had also a higher prevalence of CAD than non-PAD patients.
Characterization of Circulating EPCs by Flow Cytometry
Flow cytometry was used to identify and quantify peripheral blood CD34+ cells and CD34+KDR+ cells (supplemental Figure IA through IF). On average, 23% of circulating CD34+ cells were KDR+. The expression of VEGF-R2 (KDR) on CD34+ cell surface has been related to the endothelial differentiation of generic progenitors. In a subset of 55 subjects (43.3% of total) we also determined the expression of CD133, so that 3 subpopulations can be identified as representative of the total circulating EPC pool: CD34+CD133−KDR+, CD34−CD133+KDR+ and CD34+CD133+KDR+ cells (supplemental Figure IG and IH). We show that total CD34+KDR+ cells are the main constituent of this pool (88.0%), and their level is much more strictly correlated to the total EPC pool (defined as [CD34+] or [CD133+] and [KDR+] cell count) than CD133+KDR+ and CD34+CD133+KDR+ cells (r=0.98; 0.46 and 0.55, respectively).
Circulating EPCs Are Reduced in the Presence of PAD
Diabetic patients with either carotid or lower extremity atherosclerosis had a mean 53% reduced level of circulating CD34+KDR+ cells when compared with diabetic subjects free from PAD (40.9±2.9 versus 87.4±6.9; P<0.001). There was also a significant 22% reduction in CD34+ cells and a 41% reduction in CD133+ cells, as determined in a subset of 29 patients with and 26 patients without PAD (Figure 1A). To quantify the discriminatory capacity of progenitor cell count in identifying PAD, we built up a receiver operating characteristic curve, which revealed higher accuracy and sensitivity for CD34+KDR+ than for CD34+ cell count. (Figure 1B).
In order to establish whether risk factors or clinical conditions other than PAD might have explained in part the reduced CD34+KDR+ and CD34+ cell counts, a multiple linear regression analysis was performed: among all clinical conditions, only PAD, but not CAD, was independently associated with CD34+KDR+ cell reduction, whereas among the classic risk factors, increasing age was statistically associated with decreasing CD34+KDR+ cells. On the other hand, retinopathy and male gender were negatively correlated with CD34+ cell count (supplemental Figure II and supplemental Table I, available online at http://atvb.ahajournals.org). Statistical analyses revealed that medications had no influence on progenitor cell counts (not shown). Moreover, in a multiple stepwise logistic regression analysis, CD34+KDR+ cell count, age and male gender were independently associated with the presence of PAD (supplemental Table II, available online at http://atvb.ahajournals.org).
The Amount of Circulating CD34+KDR+ EPCs Is Negatively Correlated With Severity of Carotid and Lower Extremity Atherosclerosis
CD34+KDR+ cell count was strongly negatively correlated with the extent of carotid atherosclerosis (r=−0.51) and with the clinical severity of lower extremity atherosclerosis according to both Leriche/Fontain (ρ=−0.63) and Rutherford (ρ=−0.64) classifications (Figure 2A and 2B). Remarkably, those correlations remained significant even after adjustment for age, gender, duration of diabetes and HbA1c (supplemental Table III, available online at http://atvb.ahajournals.org). Even if many patients (n=37) had both forms of PAD, carotid and lower extremity atherosclerosis were simultaneously and independently negatively correlated with CD34+KDR+ cells (not shown). However, the presence of both carotid and lower extremity atherosclerosis was not synergically characterized by further CD34+KDR+ cell reduction (Figure 2C).
The correlations between CD34+ cell count and indicators of PAD severity were always weaker than CD34+KDR+ cells and were no more significant after statistically adjustment.
EPCs Cultured From PAD Patients Display Reduced Clonogenic and Adhesion Capacity
We isolated EPCs using a validated culture method.18 During growth in endothelial medium, a subset of peripheral blood mononuclear cells form colonies of endothelial cells and after 2 weeks display function and phenotype of endothelial cells: >90% of survived cells bind lectin, uptake LDL and are positive for the surface expression of vWf, CD31 and KDR. In support to these data, flow cytometry analysis showed that 93.5±7.8% of cultured EPCs were KDR+ and 94.8±4.3 were CD31+. In addition, we show that isolated EPCs were positive for CXCR4 (SDF-1α receptor) and CD18 (intercellular adhesion molecule receptor), which have been shown to be important for EPC function (Figure 3). Finally, we tested in vivo the ability of EPCs to take part in new vessel formation, and show their presence in vascular structures of muscles subjected to IR injury (supplemental Figure III, available online at http://atvb.ahajournals.org), whereas no injected cell was recognized in contralateral nonischemic muscles.
To have a functional demonstration of EPC alterations in vascular complications of type 2 diabetic subjects, we cultured EPCs from 15 patients with and 15 patients without PAD, comparable for age, sex and concomitant risk factors. We show that, from day 3 to 15, peripheral blood from PAD patients gave rise to a lower number of cell clusters than patients without PAD. Specifically, area under curve, which may represent the clonogenic capacity, and cell clusters after 15 days of culture, which represent true outgrown EPCs, were lower in patients with PAD than in patients without PAD (Figure 4A and 4B). Then, we studied the adhesion property of patients’ EPCs to a human umbilical vein endothelial cell monolayer and show that EPCs from PAD patients have a 35% reduced capacity to adhere to mature endothelial cells than EPCs from patients without PAD (Figure 4C through 4E). The number of adherent EPCs/high power field was significantly correlated to the number of cell clusters at day 3 (r=0.41; P=0.01).
Diabetes mellitus is characterized by a widespread endothelial dysfunction20 and a 2- to 3-fold increased risk of developing cardiovascular diseases. As previously highlighted by our group, circulating blood cells in diabetes are subjected to many biochemical alterations attributable to the high oxidative stress and the unfavorable vascular environment.21,22
The present study demonstrates that number and function of circulating EPCs are profoundly altered in type 2 diabetic patients with PAD compared with diabetic patients without PAD. Moreover, we show strong correlations between circulating EPC levels and the severity of carotid and lower extremity arterial disease.
In this work we have used 2 independent methods to study endothelial progenitors: flow cytometry of fresh blood and ex vivo culture. Flow cytometry is considered the gold standard for quantitative enumeration of EPC, being sensitive, precise and reproducible23: using this technique, we have defined EPCs by the surface expression of at least 1 marker of stemness and immaturity (eg, CD34 or CD133) plus the expression of the endothelial marker VEGFR2 (KDR). With the use of 3 surface antigens we have identified 3 subpopulations of undifferentiated progenitor cells (CD34+, CD133+ and CD34+CD133+ cells) and the 3 corresponding subpopulations of endothelial-committed progenitors (CD34+KDR+, CD133+KDR+ and CD34+CD133+KDR+ cells) which form the circulating EPC pool. In our study sample, there was a net trend toward decrease of all progenitor cell subpopulations in PAD versus non-PAD patients, but statistical significance was reached only for CD34+, CD133+ and CD34+KDR+ cell counts after adjusting for multiple testing. This result is consistent with the demonstration that CD34+KDR+ cell count predicted cardiovascular events in coronary patients.10,11 Moreover, total CD34+KDR+ cells are the major (88%) constituent of the circulating EPC pool, whereas CD133+KDR+ cells are more immature cells, such those recently mobilized from bone marrow,24 and are much rarer in peripheral blood (12%) in steady-state conditions (ie, without acute bone marrow stimulation). An obvious limitation is that CD133 expression was determined in a subset of patients, thus limiting statistical power. Nonetheless, we would suggest that CD34+KDR+ is the most appropriate phenotype to identify EPCs, because those cells are more strictly linked to cardiovascular damage, at least in diabetes.
Different from flow cytometry, EPC culture is not fully standardized, as many protocols have been proposed.19 Moreover, the resulting cell population has been shown to be spurious and its phenotype not entirely defined.25 In this work, we have used a prolonged culture system to allow for positive selection of the so-called “late outgrowing” EPCs.18 Putative EPCs have been extensively characterized: >90% of cells have shown to be positive for lectin binding and LDL uptake and for the expression of the endothelial antigens CD31, vWf, and KDR. Remarkably, cultured EPCs were also positive for CXCR4 and CD18 antigens, which have been shown to be fundamental for cell trafficking and recruitment.26,27 Finally, we have demonstrated that isolated cells are incorporated into vascular structures of rat ischemic hindlimbs. Taken together, these data indicate a high quality of the culture method used to isolate EPCs which allows for showing a 38% decrease in EPC clusters from patients with versus patients without PAD, confirming our flow cytometry data.
Experimental studies have shown that EPCs constitute a circulating pool of cells able to form a patch that actively repairs the denuded or dysfunctional endothelium.28 Notably, EPC number has been negatively correlated to parameters of endothelial function in humans.29 As endothelial damage is considered the first step in the development of the atherosclerotic plaque, EPC depression can be causally linked to the atherogenetic process. On the other side, EPCs are also actively recruited at sites of new vessel growth and take part in compensatory angiogenesis in ischemic tissues.4 The ability to form collaterals represents a crucial response to vascular occlusive diseases, because it determines the severity of residual ischemia, and whether clinical manifestations of atherosclerosis will develop.
Consequently, depletion of circulating EPCs may contribute to both endothelial dysfunction, as an early event in the atherogenetic process, and to poor collateralization, as a late event leading to the clinical manifestations of atherosclerosis and cardiovascular disease progression. In this study, carotid stenosis and the stage of lower extremity arterial disease have been considered, respectively, as representative of the anatomic atherosclerotic burden and of the late clinical outcome of local atherosclerosis obliterans.
We report that, in both conditions, CD34+KDR+ cell count was closely negatively correlated to disease severity: higher degrees of carotid stenosis, as well as worse stages of leg claudication and ischemic lesions were associated with lower levels of CD34+KDR+ EPCs. Moreover, we have demonstrated that EPCs from PAD patients display a significantly impaired adhesion to mature endothelium, which is a fundamental step for EPC function in both maintenance of endothelial homeostasis and angiogenic processes. Interestingly, in the culture experiments, PAD patients had a lower number of cell clusters as early as day 3, suggesting also impairment in adhesion to extracellular matrix, such as the fibronectin coating culture dishes.
Taken together, these data provide strong evidence for a possible role of EPC reduction in the pathogenesis of diabetic vasculopathy. Although our study was cross-sectional and does not establish cause-effect relationships, we would like to suggest that impaired collateralization leading to the clinical manifestations and complications of atherosclerosis in diabetes may be attributable to decreased and dysfunctional EPCs. Concurrently, increased carotid plaque formation may be related to the depleted reservoir of EPCs, which, also attributable to the reduced adhesiveness to the vessel wall, fail to replace successfully the damaged endothelium. In this light, ways to increase number and improve function of EPCs should be actively pursued as part of the therapeutical armamentarium against diabetic atherosclerotic complications.
It should be noted that many surrogate markers of atherosclerosis have been proposed so far, but most of them, including C-reactive protein, did not show a good correlation with the extent of the atherosclerotic burden.30 Therefore, EPC count may be proposed as a novel biomarker of diabetic atherosclerotic complications. In agreement with this hypothesis we show that PAD was independently correlated with CD34+KDR+ cell count and that, conversely, CD34+KDR+ cell count was independently associated with PAD presence. Consistently, CD34+KDR+ cell count displayed a high performance in detecting PAD among the entire population of type 2 diabetic subjects.
The present study was designed to explore specifically peripheral rather than coronary atherosclerosis. Not all patients underwent coronary angiography and CAD was defined mainly by clinical and noninvasive instrumental criteria. Therefore, we cannot exclude that some patients classified in the control group had at least some extent of CAD. This may suggest that the difference in EPC levels in patients with versus patients without macrovascular complications is underestimated. However, CAD was associated with a mild reduction in CD34+KDR+ EPCs (P=0.03, not significant after α-adjustment). It should be noted that the literature provides inconsistent data on EPC reduction in stable CAD: in one study,31 the CAD group included patients with acute coronary syndromes, that are known to be followed by EPC increase; in another study,32 CD34+CD133+ cells were not reduced in patients with chronic myocardial ischemia compared with healthy subjects. Moreover, some diabetic patients may display clinical and noninvasive instrumental features of chronic myocardial ischemia in the absence of severe atherosclerotic involvement of epicardial vessels. Indeed, heart disease in diabetes is attributable not only to coronary atherosclerosis, but also to a wide range of biochemical, metabolic and structural microvascular alterations,33 whereas PAD is the direct consequence of accelerated atherosclerosis in diabetic subjects. These reasons may explain why PAD was more closely associated to CD34+KDR+ EPC reduction than CAD in this study.
In summary, this study solidly indicates that diabetic vasculopathy is associated with EPC impairment: depletion of the EPC pool and defective adhesive capacity is probably one cause of the aggressive cardiovascular disease in these subjects. Moreover, the strong correlations between EPC reduction and the degree of peripheral atherosclerosis reveal that CD34+KDR+ EPC count represents a novel prototype of cardiovascular biomarkers and suggest a pathogenetic model by which diabetic vasculopathy worsens as EPCs decrease. We should consider EPC alterations as a therapeutic target for diabetic macroangiopathy.
Source of Funding
This work was partially supported by a grant from Consortium Agreement European Community Heart Failure and Cardiac Repair-IP 018630 (S.S).
Original received March 28, 2005; final version accepted June 27, 2006.
Waltenberger J. Impaired collateral vessel development in diabetes: potential cellular mechanisms and therapeutic implications. Cardiovasc Res. 2001; 49: 554–560.
Hristov M, Erl W, Weber PC. Endothelial progenitor cells: mobilization, differentiation, and homing. Arterioscler Thromb Vasc Biol. 2003; 23: 1185–1189.
Takamiya M, Okigaki M, Jin D, Takai S, Nozawa Y, Adachi Y, Urao N, Tateishi K, Nomura T, Zen K, Ashihara E, Miyazaki M, Tatsumi T, Takahashi T, Matsubara H. Granulocyte colony-stimulating factor-mobilized circulating c-Kit+/Flk-1+ progenitor cells regenerate endothelium and inhibit neointimal hyperplasia after vascular injury. Arterioscler Thromb Vasc Biol. 2006; 26: 751–757.
Werner N, Nickenig G. Influence of cardiovascular risk factors on endothelial progenitor cells: limitations for therapy? Arterioscler Thromb Vasc Biol. 2006; 26: 257–266.
Loomans CJ, de Koning EJ, Staal FJ, Rookmaaker MB, Verseyden C, de Boer HC, Verhaar MC, Braam B, Rabelink TJ, van Zonneveld AJ. Endothelial progenitor cell dysfunction: a novel concept in the pathogenesis of vascular complications of type 1 diabetes. Diabetes. 2004; 53: 195–199.
Tepper OM, Galiano RD, Capla JM, Kalka C, Gagne PJ, Jacobowitz GR, Levine JP, Gurtner GC. Human endothelial progenitor cells from type II diabetics exhibit impaired proliferation, adhesion, and incorporation into vascular structures. Circulation. 2002; 106: 2781–2786.
Fadini GP, Miorin M, Facco M, Bonamico S, Baesso I, Grego F, Menegolo M, de Kreutzenberg SV, Tiengo A, Agostini C, Avogaro A. Circulating endothelial progenitor cells are reduced in peripheral vascular complications of type 2 diabetes mellitus. J Am Coll Cardiol. 2005; 45: 1449–1457.
Schmidt-Lucke C, Rossig L, Fichtlscherer S, Vasa M, Britten M, Kamper U, Dimmeler S, Zeiher AM. Reduced number of circulating endothelial progenitor cells predicts future cardiovascular events: proof of concept for the clinical importance of endogenous vascular repair. Circulation. 2005; 111: 2981–2987.
Kondo T, Hayashi M, Takeshita K, Numaguchi Y, Kobayashi K, Iino S, Inden Y, Murohara T. Smoking cessation rapidly increases circulating progenitor cells in peripheral blood in chronic smokers. Arterioscler Thromb Vasc Biol. 2004; 24: 1442–1447.
Fadini GP, Sartore S, Baesso I, Lenzi M, Agostini C, Tiengo A, Avogaro A. Endothelial progenitor cells and the diabetic paradox. Diabetes Care. 2006; 29: 714–716.
Hur J, Yoon CH, Kim HS, Choi JH, Kang HJ, Hwang KK, Oh BH, Lee MM, Park YB. Characterization of two types of endothelial progenitor cells and their different contributions to neovasculogenesis. Arterioscler Thromb Vasc Biol. 2004; 24: 288–293.
Avogaro A, Toffolo G, Kiwanuka E, de Kreutzenberg SV, Tessari P, Cobelli C. L-arginine-nitric oxide kinetics in normal and type 2 diabetic subjects: a stable-labelled 15N arginine approach. Diabetes. 2003; 52: 795–802.
Friedrich EB, Walenta K, Scharlau J, Nickenig G, Werner N. A CD34-/CD133+/VEGFR-2+ endothelial progenitor cell subpopulation with potent vasoregenerative capacities. Circ Res. 2006; 98: e20–e25.
Ingram DA, Caplice NM, Yoder MC. Unresolved questions, changing definitions, and novel paradigms for defining endothelial progenitor cells. Blood. 2005; 106: 1525–1531.
Smadja D, Bieche I, Uzan G, Bompais H, Muller L, Boisson-Vidal C, Vidaud M, Aiach M, Gaussem P. PAR-1 activation on human late endothelial progenitor cells enhances angiogenesis in vitro with upregulation of the SDF-1/CXCR4 system. Arterioscler Thromb Vasc Biol. 2005; 25: 2321–2327.
Chavakis E, Aicher A, Heeschen C, Sasaki K, Kaiser R, El Makhfi N, Urbich C, Peters T, Scharffetter-Kochanek K, Zeiher AM, Chavakis T, Dimmeler S. Role of beta2-integrins for homing and neovascularization capacity of endothelial progenitor cells. J Exp Med. 2005; 201: 63–72.
Vasa M, Fichtlscherer S, Aicher A, Adler K, Urbich C, Martin H, Zeiher AM, Dimmeler S. Number and migratory activity of circulating endothelial progenitor cells inversely correlate with risk factors for coronary artery disease. Circ Res. 2001; 89: e1–e7.
Heeschen C, Lehmann R, Honold J, Assmus B, Aicher A, Walter DH, Martin H, Zeiher AM, Dimmeler S. Profoundly reduced neovascularization capacity of bone marrow mononuclear cells derived from patients with chronic ischemic heart disease. Circulation. 2004; 109: 1615–1622.