Experimental studies using animal models and clinical trials involving intracoronary infusion of bone marrow-derived cells in patients demonstrated that stem/progenitor cells have the potential to improve the function of the myocardium after ischaemic injury.1 2 Two hypotheses explain the role of adult progenitor cells in tissue regeneration. Stem cell plasticity is associated with homing of the stem cells to the area of tissue injury and transdifferentiation into functional endothelial cells or cardiomyocytes. An alternative hypothesis involving the concept of tissue-committed stem cells is based on the observation that bone marrow harbours a heterogeneous population of cells which express tissue-specific markers—for example, genes characteristic for cardiomyocytes or endothelial cells. These cells are present in peripheral blood and are attracted by cytokines to the place of tissue injury. Both mechanisms require the presence of a mobile pool of progenitor cells which are mobilised in endothelial injury or myocardial ischaemia and subsequent homing and engraftment to the site of injury.3^–8

In recent years much more has become known about the role of autologous stem/progenitor cells which can be found circulating in peripheral blood and reflect the reparatory mechanism of endothelium and myocardium. Some data suggest also that measurement of the stem cell number in peripheral blood may be a surrogate marker to assess the endothelial damage regardless of the nature of noxious stimuli. This review summarises the data on mobilisation of stem/progenitor cells in coronary artery disease (CAD), including stable coronary heart disease and acute coronary syndromes (ACS), examining cell types, mechanisms and relevance to clinical outcomes.

Evidence that migratory circulating cells participate in the formation of virtually all cardiac structures in a transplanted heart comes from pivotal work of Quaini et al who documented the systemic chimerism in cases in which a male patient received a heart from a woman using the FISH technique to detect the presence of Y chromosome. Up to 18% of cardiomyocytes, small coronary arteries and capillaries in the donor heart were Y-chromosome positive, suggesting that mobilised recipients primitive cells positive for c-kit and MDR-1 antigens took part in the formation of these structures.9

POPULATIONS OF MOBILISED CELLS

The populations mobilised in ACS consist predominantly of committed lineages such as polymorphonuclear granulocytes, T and B lymphocytes, monocytes and highly heterogeneous subpopulations characterised by the ability of endothelial differentiation (endothelial progenitor cells (EPC)), haematopoietic progenitors (haematopoietic stem cells (HSC)), stromal (mesenchymal) cells (MSC) and other poorly defined progenitors.5 10 It is noteworthy that adult bone marrow contains a limited number of “true” stem cells (self-replicating, capable of generating, sustaining and replacing terminally differentiated cell lineages) and a majority of already committed lineages. The population of EPC represents <2% and MSC <0.001% of the total number of bone marrow mononuclear cells. In the peripheral blood EPC are extremely rare (0.0001%).11 Stem and progenitor cells are identified using flow cytometry (fluorescence activated cell sorting (FACS)) by the presence of surface markers (receptors) classified according to the cluster of differentiation nomenclature. The distribution of markers overlaps the different population of cells and no single marker identifies the particular cell type. Therefore a constellation of markers is used to identify the progenitor cells and there is a considerable debate on which set of markers is sufficient to identify the cells.10 12

Endothelial precursors capable of transforming into mature, functional endothelial cells are present in the circulation in the pool of mononuclear cells and can be derived from either bone marrow or peripheral blood monocytes. The most widely used immunophenotype for EPC identification is the presence of following surface markers: CD34, vascular endothelial growth factor (VEGF) receptor 2 (VEGFR-2) (kinase insert domain containing receptor (KDR) or fetal liver kinase-1 (flk-1)) and CD133, which is no longer present as the EPC mature into endothelial cell (EC).10 13 EPC are also characterised by their ability to uptake acetylated low-density lipoprotein (diI-Ac-LDL), to bind plant lectins (Ulex europaeus) and to form endothelial colonies. During maturation into endothelial cells EPC became CD133 negative, but start expressing endothelial markers, such as von Willebrand factor and CD31.10 14

The bone marrow-derived haemangioblasts are not the only precursors of EPC. A growing body of evidence supports the hypothesis that peripheral blood-derived monocytes can undergo differentiation into EPC. Also, a recently identified population of peripheral blood-derived CD34⁻/CD133⁺/VEGFR-2⁺ cells may differentiate into CD34⁺/CD133⁺/VEGFR-2⁺ and eventually mature EC. The bone marrow myelomonocytic lineage can have common properties as endothelial lineage—for example, expression of endothelial genes and proteins when cultured under angiogenic conditions.15 16 This population is mobilised during experimental limb ischaemia in humans, expresses higher levels of CXCR4 and sromal-derived factor-1 (SDF-1)-mediated adhesion than more mature CD34⁺/CD133⁺/VEGFR-2⁺ and is negative for monocyte marker CD14.17 Therefore it seems that the population of circulating EPC consists of at least three subpopulations—less mature CD34⁻/CD133⁺ cells, which can differentiate into CD34⁺/CD133⁺ and subsequently more mature CD34⁺/CD133⁻ cells.

The use of CD34 alone may, therefore, be insufficient to enumerate the cells because earlier populations may be CD34 negative and CD34 is expressed on mature EC sloughed from the vascular endothelium.10 In addition, another population of cells positive for monocyte marker CD14 was identified. These cells can undergo differentiation into a population of CD34⁺/CD133⁺ EPC.18 19 Again, unlike the well-characterised proliferative potential of HSC, the EPC remain less well characterised and their clonogenic capacity in vitro may not represent the actual role in endothelial repair following acute ischaemic injury.

The ability of EPC to form endothelial colonies is also an imperfect tool, because several populations of circulating cells may undergo differentiation into endothelial lineage (HSC, monocytes/macrophages). Moreover, sloughed mature EC may have proliferative potential.11 In particular, monocytes/macrophages in culture conditions can form spindle-like adherent cells and express endothelial markers, so the use of endothelial colony-forming unit (CFU-E) assays does not necessarily mean that the cells are derived from bone marrow EPC. New assays of EPC colony-forming unit activity can measure the proliferative potential of EPC under different conditions at the single-cell level. However, the studies investigating the mobilisation of progenitor cells in ACS did not use this methodology and therefore have significant limitations.11 15 16 19 20

MOBILISATION OF STEM/PROGENITOR CELLS

Stable coronary heart disease

A recently published large prospective observational study enrolling 519 patients with stable CAD confirmed by angiography showed that the low number of circulating CD34⁺/VEGFR-2⁺ EPC is associated with a significantly higher risk of death from cardiovascular disease, a first major cardiovascular event, revascularisation and hospitalisation in comparison with patients with high EPC numbers. In addition, the cumulative event-free survival increased in stepwise fashion with increasing baseline EPC number. The EPC number was, however, not predictive of acute myocardial infarction (AMI) or all-cause mortality.21 This confirmed previous studies on a smaller group of patients with CAD which also showed a negative correlation between the number of risk factors and EPC counts.

A reduction in the EPC number was particularly evident in smokers and subjects with a positive family history of CAD. Higher numbers of risk factors were also associated with impaired migratory response.22 Since the circulating EPC count reflects the mechanism for maintaining the endothelial integrity, measurement of the number of cells may help to identify patients at increased cardiovascular risk. Interestingly, also in 45 apparently healthy subjects the number of circulating EPC-derived colonies showed a significant inverse correlation with the Framingham risk score. Moreover, the EPC count was a better predictor of endothelium-dependent vasodilatation assessed by the flow-mediated brachial-artery reactivity test than conventional risk factors, and subjects with a low EPC number had reduced flow-mediated brachial reactivity independently of their Framingham risk score.

The functional capacity of EPC (CFU) was inversely correlated with levels of low-density lipoprotein cholesterol, age, diabetes, smoking and a family history of premature coronary artery disease.21 EPC number was also correlated with response to nitroglycerin, which is the endothelium-independent vasodilatory agent. Finally, the index of EPC senescence was significantly higher in subjects with a high risk score.23 24 Therefore it seems that measurement of EPC may represent the global burden of cardiovascular risk factors as well as endothelial function beyond the widely used risk markers. Interpretation of the studies involving measurement of EPC has to incorporate the patients’ history because previous ACS as well as numerous coexisting medical conditions, drugs and invasive procedures can initiate the cell mobilisation. This can lead to misclassification of subjects with low baseline EPC counts and high risk, assigning them to the low-risk group, because the measured EPC number was increased owing to vascular injury. Also as already discussed, the population of EPC is heterogeneous and the possibility of measuring sloughed mature EC instead of EPC has to be considered.

Based on the retrospective observation in a relatively small number of patients it seems that a low number of circulating EPC and their impaired adhesiveness to fibronectin may be associated with a higher rate of diffuse in-stent restenosis in contrast to focal restenosis and patients without restenosis. However, the usefulness of EPC measurement for identifying patients with a high risk of restenosis must be confirmed in a large prospective study.25

Acute coronary syndromes

ACS are associated with increased levels of inflammatory and haematopoietic cytokines, which in turn can mobilise the progenitor cells from the bone marrow, although a number of studies showed no correlations between the cells and cytokines levels. Probably for both measures the timing of the blood sampling is crucial, because the cell mobilisation seems to occur very early and the inflammatory reaction is prolonged.26^–29

Shintani et al were the first to describe the rapid and significant increase of the CD34+ EPC number, which reached a maximum 7 days after the onset of ischaemia in AMI and paralleled the increase in VEGF levels.27 The increase of EPC count in AMI was subsequently confirmed by other authors. Massa et al described spontaneous mobilisation and a 5.8-fold increase of CD34⁺ progenitor cells, which peaked within a few hours (median about 3 hours) after the onset of symptoms, significantly decreased after 7 days, and reached levels comparable with those of healthy subjects within 2 months.28 The heterogeneous population of CD34+ cells was further characterised using FACS to show that not only EPC numbers are increased but also other less well-defined subpopulations, such as CD34⁺/CD117⁺, CD34⁺/CXCR4⁺, CD34⁺/CD38⁺ and CD34⁺/CD45⁺ cells which follow the same time course as EPC.26 28 The functional assays showed a significant increase of typical endothelial and haematopoietic colonies, so that EPC and HSC are both mobilised in AMI. Moreover, not only the number of EPC but also their ability to differentiate into adult EC may affect functional improvement and infarct size reduction shown by nuclear scan, left ventricular ejection fraction (LVEF) salvage and favourable remodelling in patients with AMI.30 The type of primary treatment of AMI does not seem to affect the mobilisation because a similar pattern of CD34⁺ cells increase was found in patients receiving thrombolytic therapy and in those treated with primary angioplasty.31 EPC mobilisation was also noted in patients with unstable angina, in whom an increase of EPC number showing normal adhesive properties was correlated with C-reactive protein levels. After 3 months of follow-up the number of EPC was reduced by 50% in comparison with baseline counts.32

In addition to changes in EPC and HSC, changes in the number of circulating MSC in AMI were recently reported. The data for the mobilisation of MSC in AMI are not consistent. Some studies showed changes in the number of MSC, but the significance of these findings has to be confirmed and the widely accepted methods of identification of MSC defined.33

MECHANISMS OF STEM/PROGENITOR CELL MOBILISATION

Trafficking of stem cells involves mobilisation into peripheral blood, homing, adhesion and engraftment into the target tissue. This process, which is part of inflammatory response and tissue repair, requires signalling mediated primarily by chemokines, which are the most important regulators of cell trafficking, survival and function. Numerous haematopoietic and inflammatory cytokines known to increase progenitor cell mobilisation are markedly upregulated in ACS, and also increased to lesser degree in CAD.34^–37

Most consistent data are available about EPC mobilisation and show a significant increase of VEGF levels in parallel with enhanced EPC numbers, and significant correlation between these two measures.26–28 38 Moreover, as shown by Schomig et al, interleukin 8 may together with VEGF be an additional factor associated with the mobilisation of EPC and endothelial cells in AMI.39 Data on the association between an increase of inflammatory cytokines and mobilisation of progenitor cells is not consistent. The increase of EPC is correlated with the rise in VEGF levels,27 and increased levels of SDF-1 were associated with significant mobilisation of progenitor cells26; however, these correlations were not confirmed in other studies.28 40 Importantly, the presence of significant correlations does not necessarily confirm a causal link between the mobilisation and increased cytokine levels, given the complexity of inflammatory reaction in ACS. Particular populations of stem cells in the bone marrow express the membrane receptor CXCR4, which is specific for and the only receptor for the chemokine SDF-1. In addition, SDF-1 binds exclusively to CXCR4, the G-protein-coupled seven-span transmembrane receptor, which is the major regulator of stem/progenitor cell trafficking and adhesion. The importance of this signalling axis is documented in CXCR4 and SDF-1 knockout models. Both defects lead to significant impairment of the embryonic bone marrow population by HSC, as well as significant defects in cardiogenesis and vasculogenesis and are lethal.

The population of CXCR4⁺ cells displays significant chemotactic response to an SDF-1 gradient and this chemokine plays a pivotal role in the retention and homing of haematopoietic stem cells within the postnatal bone marrow. SDF-1 expression was shown to be down-regulated by granulocyte colony-stimulating factor (G-CSF), and this is a necessary step in the mobilisation of stem/progenitor cells.

On the other hand, this action of G-CSF might explain the failure of the clinical use of this cytokine in patients with AMI, because of the impaired SDF-1-guided homing of CXCR4⁺ cells to the myocardium, although this hypothesis needs further investigation.41 A functional CXCR4 receptor was identified on embryonic pluripotent stem cells, primordial germ stem cells and a population of cells enriched in the tissue-specific markers (cardiac, endothelial, neural), therefore it can be regarded as an important stem cell marker and SDF-1 seems to be the pivotal chemoattractant for CXCR4+ stem/progenitor cells.41 In mice experiments it was shown that bone marrow contains pools of cells that express early cardiac lineage markers (Nkx2.5/Csx, GATA-4, and MEF2C) and that this population can be mobilised in experimental myocardial infarction.

This is the first proof that postnatal bone marrow contains a non-haematopoietic population of cells that express markers for cardiac differentiation (tissue-committed stem cells (TCSC)). The peak expression of cardiac markers was found at the same time as the most significant increase of stem cell numbers.42 Similar phenomenon occurs in humans with AMI.26 The SDF-1/CXCR4 axis seems particularly important in stem/progenitor cell homing, chemotaxis, engraftment and retention in ischaemic myocardium, because infarct border-zone (area adjacent to the myocardial necrosis) shows the presence of SDF-1 and the levels of the chemokine increase in AMI (fig 1).42 43 In addition, impaired CXCR4 signal transduction may be associated with reduced neovascularisation capacity of EPC.4 44 Recently published evidence shows that murine, primarily non-haematopoietic, bone marrow-derived CXCR4⁺/Sca-1⁺/lin-/CD45⁻ cells, the same population that is mobilised and undergoes chemotaxis to infarcted myocardium, express also several markers of pluripotent embryonic-like stem cells and display unique morphology. Based on their morphology and positive staining for Oct-4 and Nanog, the cells were named very small embryonic-like stem cells and might be the ideal population of stem cells for regeneration of myocardium, because of their potential for expansion and possibility of cardiogenic differentiation.45

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Circulation of tissue-committed (cardiac, endothelial, neural) stem cells (TCSC). TSCS are present in peripheral blood of healthy subjects and patients with coronary artery disease. The bone marrow is the primary source of TCSC but the cells can also hide in various niches such skeletal muscle and liver (A). Upregulation of chemokine synthesis, primarily SDF-1, HGF and LIF in acute myocardial ischaemia and stroke directs the chemotaxis of cells expressing the chemokine receptors (CXCR4, c-met and LIF-R, respectively) towards the damaged tissue, allowing TCSC homing and engraftment. The SDF-1/CXCR4 axis is the most important TCSC homing mechanism in acute coronary syndromes (B).4 87 88 HGF, hepatocyte growth factor; LIF, leukaemia inhibitory factor; LIF-R, leukaemia inhibitory factor receptor, SDF-1, stromal-derived factor-1.

Other important cytokine-receptor systems regulating stem cell mobilisation and homing are leukaemia inhibitory factor (LIF)—LIF receptor, hepatocyte growth factor—c-met axis, stem cell factor-CD117 axes. All of them are upregulated and activated in acute myocardial necrosis.42 46 47 The bone marrow enrichment and the expression of embryonic antigens in bone marrow-derived cells is significantly lower in older animals, suggesting that age-related changes play important role in the viability of the cells. Our data from humans show that the mobilisation of CD34⁺/CD117⁺ and CD34⁺/CXCR4⁺ cells in AMI is significantly lower in older than in younger patients.48 Also, survival, migration to VEGF gradient and the proliferative capacity of CD34⁺/VEGF2R⁺ and CD133⁺/VEGF2R⁺ cells were significantly better in young healthy subjects than in older people.49 This suggests that ageing may lead to decreased function and viability of EPC and other progenitor cell types and subsequent decrease of endothelial reparatory potential. The same may be true with regard to reduced number of bone marrow progenitor cells. This interesting theory needs to be proved and other factors associated with age, such as comorbidities, drugs, invasive procedures may be responsible for the changes in progenitor cell number and function.

Recent data from Fazel et al demonstrated also that the CD117(c-kit)/SCF signalling axis is crucial for the mobilisation and engraftment of c-kit bone marrow-derived progenitor cells. In addition the c-kit⁺ cells improve the collateral formation in the infarction border zone, primarily by increasing VEGF and angiopoietin-2 expression. Dysfunctional CD117 is associated with more significant myocardial damage in experimental AMI, leading to rapid development of ischaemic cardiomyopathy.50

Among other factors, concomitant drug therapy, surgical procedures and coexisting diseases may modulate the number of circulating stem cells (tables 1, 2 and 3).

Mobilisation of stem/progenitor cells and outcomes in ACS

Although measurement of circulating EPC in stable CAD may have prognostic value, much less is known about whether the same is true for ACS. So far no prospective study has shown convincingly an association between circulating stem/progenitor cells and hard clinical end points. Some data, however, suggest that cell mobilisation is correlated with clinically important measures of left ventricular function, such as LVEF and left ventricular remodelling. Leone et al showed that the number of circulating CD34⁺ cells 1 year after AMI is significantly correlated with changes of LVEF, wall motion score index, left ventricular end-diastolic volume and left ventricular end-systolic volume.29 In addition, acute mobilisation of CD34⁺, CXCR4⁺, CD117⁺ and c-met⁺ stem/progenitor cells early in AMI is significantly positively correlated with LVEF and negatively correlated with biochemical markers of left ventricular overload (pro-B-type natriuretic peptide) and cardiac necrosis markers (troponin I, maximum activity of creatine kinase MB isoenzyme). This association is, however, moderate and the link between the mobilisation and degree of myocardial necrosis has to be confirmed before using this measure as a prognostic marker.

In patients with low baseline LVEF (VEF <40%), the peak number of stem cells was significantly lower than in patients with LVEF >40%.38 Some data suggest that mobilisation of stem cells in AMI is also predictive of left ventricular function 1 year later, although these findings are confined to small populations of patients.85 Further prospective studies are needed to confirm the findings because other data showed no significant correlations between stem cell number and LVEF and cardiac necrosis markers.27 28 31

The causal relationship between stem cell mobilisation and extent of myocardial infarct size can be bidirectional because either the large necrotic area may be a potent stimulator of cell mobilisation and homing very early on in AMI, resulting in a lower number of cells detected in peripheral blood later on, or poor mobilisation and/or defective function of progenitor cells can negatively influence the postinfarct remodelling. Data from the clinical trial using a bone marrow-derived population of CD133⁺ cells given by intracoronary infusion in patients with AMI suggested that this approach may lead to an increased risk of in-stent restenosis. Indeed, data from human atherectomy specimens showed the presence of angiogenic cells positive for CD117 and coexpressing α-actin. It is not clear, however, if the cells are derived from peripheral blood EPC.86

Moreover, in another well-designed study the presence of either CD34⁺/CD133⁺/VEGFR-2⁺ and CD34⁻/CD133⁺/VEGFR-2⁺ EPC was not confirmed in human atherectomy specimens from patients with in-stent restenosis.17 George et al reported that in patients with diffuse in-stent restenosis the number of circulating EPC (CFU) was reduced in comparison with subjects with focal restenosis. However, there were no differences in EPC number between the groups with and without restenosis. Additionally, in-stent restenosis was associated with reduced adherence to fibronectin of EPC, suggesting that impaired function rather than reduced numbers may play a role in the pathogenesis of in-stent restenosis.25

CONCLUSION

In ACS significant mobilisation of a highly heterogeneous population of bone marrow-derived stem/progenitor cells occurs. Several subpopulations of cells that may be involved in endothelial and myocardial repair can be measured in peripheral blood, including lymphocytes, monocytes/macrophages, granulocytes, sloughed mature EC as well as progenitor cells such as EPC, HSC and MSC. The cell number and type can be measured by flow cytometry and additional data such as migratory capability, viability, clonogenic potential and gene expression profile can be obtained by in vitro assays. Activation of chemokines and chemokine receptors is a pivotal factor in the cell mobilisation, trafficking and engraftment. Available data correlating the mobilisation of EPC and HSC with the clinical outcome is limited and larger prospective studies are necessary before using the cells number as a prognostic markers.