Cell therapy for myocardial regeneration is an exciting new field of medical research that has the potential to revolutionise cardiovascular medicine. Despite significant improvements in emergency treatment, myocardial infarction (MI) leads to a net loss of contractile tissue in many patients with coronary artery disease (CAD). Often, this is the beginning of a downward spiral towards congestive heart failure (CHF) and life-threatening arrhythmia. Other than heart transplantation with its obvious limitations, current therapeutic means aim at preventing further episodes of myocardial ischaemia and at enabling the patient to survive with a heart that is working at a fraction of its original capacity. They are far from representing a cure. In this situation, it is understandable that cardiac stem cell therapy attracts considerable attention and raises many hopes. In order to adequately judge both the potential benefits and the limitations of cardiac cell therapy, some understanding of the mechanism and the consequences of MI and its current treatment concepts is needed.
In the setting of acute MI, several studies have shown a functional benefit of intracoronary infusion of bone marrow cells compared with the standard treatment alone, but patients with chronic ischaemic heart disease and impaired heart function may require a different approach. Therefore, the authors' group developed a protocol for injection of purified CD133+ bone marrow stem cells directly into the diseased myocardium at the time of coronary artery bypass graft (CABG) surgery. Based on the encouraging results in the first six patients, the authors completed a dose-escalation safety trial and then conducted a controlled study to determine efficacy compared with the standard CABG operation.
Bone Marrow Cells and Angiogenesis
During embryonic development the primary vascular plexus is formed by haemangioblasts, stem cells capable of generating both haematopoietic progeny and endothelial cells, in a process termed vasculogenesis. Further blood vessels are generated by both sprouting and non-sprouting angiogenesis, finally leading to the complex functional adult circulatory system. Until recently, only two mechanisms of post-embryonic vascular remodelling have been recognised. Angiogenesis, the proliferative outgrowth of local capillaries, is one way to reinforce perfusion. Angiogenesis can occur under various conditions, including ischaemia. In case of myocardial ischaemia due to the occlusion of a coronary artery, pre-existing small collateral vessels also bear the capacity to enlarge in a process termed arteriogenesis. It has long been assumed that both mechanisms are mainly due to local proliferation of resident cells.
The advent of cellular therapy of ischaemic organ damage has introduced neoangiogenesis (sometimes also termed vasculogenesis) due to immigrating stem cells and progenitors as a third possible mechanism operating to improve perfusion of the adult damaged heart. Accumulating evidence indicates that immigrating (stem) cells can truly differentiate along endothelial lineage but also provide paracrine support in these three courses of action during regenerative vascular remodelling.
Putative progenitors for therapeutic angiogenesis have been isolated from adult human peripheral blood based on their expression of CD34, a marker molecule shared by microvascular endothelial cells and haematopoietic stem cells. The same group provided the proof of concept by transplantation of genetically marked mouse bone marrow into recipient mice that were subsequently subjected to five distinct models of vascular remodelling including myocardial ischaemia. In this particular system, transgenic mice constitutively expressing beta-galactosidase under the transcriptional regulation of an endothelial-cell-specific promoter were used as donors to replace the bone marrow in the recipient animals. Definitively bone marrow-derived endothelial (progenitor) cells were found in reproductive organ tissues as well as in healing cutaneous wounds one week after punch biopsy. Marrow-derived endothelial progenitor cells were found to incorporate into capillaries among skeletal myocytes in an additional test for peripheral post-ischaemic regeneration after hindlimb ischaemia, as well as into foci of neovascularisation at the border of an infarct after permanent ligation of the anterior descending artery. Most importantly, direct injection of the bone marrow mononuclear cell fraction in rat models of myocardial ischaemia increased the capillary density. Analysis of the effects of blood and bone marrow mononuclear cell implantation into ischaemic myocardium in pigs further revealed that the stem cell effects are not limited to angiogenesis and improved collateral perfusion, but also include the supply of regulatory cytokines. However, concerns exist regarding limited efficiency owing to the minute numbers of stem cells in small sample volume of non-enriched blood and bone marrow that are delivered intramyocardially and the risk of foreign tissue differentiation following local stromal cell injections. Kocher et al. circumvented this problem by using positively selected CD34+/133+ cells from human donors after stem cell mobilisation with granulocyte colony stimulating factor (G-CSF) for intravenous injection after permanent ligation of the left anterior descending coronary artery in nude rats, resulting in a five-fold increase in the number of capillaries compared with control. As a result of the stem-cell-mediated angiogenesis, which was attributed to the content of marrow-derived angioblasts, the authors also found an approximately 20% increase in left ventricular ejection fraction (LVEF) and cardiac index together with a reduced severity of ventricular remodelling in CD34-treated humans compared with control ischaemic animals.
Another candidate cell population for the regeneration of ischaemic cardiac muscle and vascular endothelium is CD45+ haematopoietic CD34LOW/-/c-Kit+, so-called side population stem cells with a specific Hoechst 33342 DNA dye efflux pattern. Orlic et al. used an alternative method to enrich putative regenerative stem cells for local application by depleting unwanted cell lineages prior to enrichment for the expression of the stem cell factor receptor c-Kit from murine bone marrow. Thus concentrated, cells considered to represent haematopoietic stem cells were observed to incorporate not only into vascular structures but dominantly led to myocardial regeneration. Subsequent experiments by this group employed mobilisation of stem cells by G-CSF prior to experimental MI, which also led to a significant increase in vascular density within the scar, a reduction in mortality and a significant reduction in infarct size.
Although the evidence that angiogenesis occurs in ischaemic myocardium is convincing, this new therapeutic option also has a potential for serious side effects. Most importantly, bone marrow-derived endothelial cells were found as part of the tumour neovasculature in experimental colon cancer. This finding might suggest a risk of triggering the growth of silent tumours by systemic use of pro-angiogenic stem cell therapy.
Bone Marrow Cells and Myogenesis
While the pro-angiogenic effect of marrow-derived stem cells appears to be well established, stem-cell-mediated myogenesis remains a matter of debate. The traditional view implies that ischaemic damage to the myocardium can only be compensated by hypertrophy, not hyperplasia, of surrounding cardiomyocytes. This has recently been challenged, and intramyocardial as well as extramyocardial sources of regenerating contractile cells have been suggested. Cardiomyocyte proliferation has been described, although only with minute frequency. The existence of cardiomyocytes of non-cardiac origin has been suggested by chimerism analyses after transplantation, but the biological relevance of some of these data has been questioned.
The notion that bone marrow cells can regenerate infarcted myocardium led to great excitement. In their landmark paper, Orlic et al. described that injection of genetically labelled murine LinNEG/c-Kit+ stem cells (isolated from mouse bone marrow by depletion of committed cells, and further enriched for expression of c-Kit) led to the formation of new myocardium, occupying two thirds of the infarct region within nine days. This initiated a wave of enthusiasm and critical discussion. The data were interpreted to indicate trans-differentiation of adult haematopoietic stem cells by crossing lineage boundaries.
But the fact that cells are derived from bone marrow is not necessarily proof that they are haematopoietic in origin, especially in the light of growing knowledge about mesenchymal, non-haematopoietic stem cells within the marrow. The recognition of cell fusion as a common phenomenon in some artificial transplant models for regeneration of ischaemic tissue has added to the controversy. From the clinician's point of view this was no surprise, since cell fusion is an intrinsic characteristic of contractile cells. Multinucleated skeletal myotubes are a classic example of cell fusion, and cardiomyocytes have long been known to form a large syncytial union.
More serious concerns were produced by two publications that could not reproduce the promising in vivo trans-differentiation data. Using a modified Lin+ depletion protocol for stem cell enrichment in an otherwise similar myocardial ischaemia model, Balsam et al. found abundant green fluorescent protein (GFP)+ cells in the myocardium after 10 days, which nearly disappeared until day 30. The remaining donor cells lacked cardiac tissue-specific markers, and instead adopted only haematopoietic fates as indicated by the expression of CD45. Murry et al. used both cardiomyocyte-restricted and ubiquitously expressed reporter transgenes to follow murine LinNEG/c-Kit+ stem cells after transplantation into healthy and injured mouse hearts, and could not find evidence for relevant differentiation into cardiomyocytes. In defence of the initial paper some have argued that the cell isolation protocols were not completely identical, and both groups observed some functional improvement in cell-treated hearts, but it cannot be denied that the evidence for myogenesis based on haematopoietic adult stem cell myogenesis is extremely controversial. Very recently, a direct comparison of human CD133+ bone marrow cells and human skeletal myoblasts in a myocardial ischaemia model in immunoincompetent rats demonstrated similar functional improvement in both groups, although only the myoblasts reached robust engraftment. The authors' studies underline the angiogenic capacity of CD133+ stem cells from adult human bone marrow and chord blood in a Scid-mouse MI model. Moreover, both cell preparations had a beneficial effect on post-infarction mortality and apoptosis. Adult bone marrow preparations contained a higher c-Kit population and caused cardiac functional restoration in echocardiography. These findings underscore our limited understanding of how stem cells can elicit an improvement of heart function.
In contrast, the myogenic potential of stromal cell-derived mesenchymal stem cells is much better documented. Stroma cells are usually isolated based on their ability to adhere to plastic, not by selection for expression of certain surface markers. Their number in primary marrow aspirates is low, but they readily multiply for numerous cycles in culture, without apparent genotypic and phenotypic changes. Wakitani et al. reported the in vitro development of myogenic cells from rat bone marrow mesenchymal stem cells exposed to the DNA-demethylating agent 5-azacytidine, and Makino et al. isolated a cardiomyogenic cell line from murine bone marrow stromal cells that were treated with 5-azacytidine and screened for spontaneous beating. Those cells connected with adjoining cells, formed myotube-like structures, and beat spontaneously and synchronously. They expressed various cardiomyocyte-specific proteins, had a cardiomyocytelike ultrastructure, and generated several types of sinus node-like and ventricular cell-like action potentials. When isogenic marrow stromal cells are implanted in rat hearts, they appear to become integrated in cardiac myofibres, assume the histologic phenotype of cardiomyocytes, express connexins, and form gap junctions with native cardiomyocytes. Again, epigenetic modification with 5-azacytidine is believed to facilitate differentiation towards a cardiomyocyte phenotype in vivo. Human mesenchymal stem cells derived from the marrow of volunteers have also been injected into hearts of immunodeficient mice, and again it was observed that they assume cardiomyocyte morphology and express various cardiomyocyte-specific proteins.