Time for primary review 23 days.
In newts, excision of up to 50% of the ventricular myocardium results in spontaneous regeneration and rebuilding of the heart, a process that involves de-differentiation of the remaining cardiomyocytes, de-novo synthesis of DNA followed by proliferation and mitotic cell division of both cardiomyocytes and connective tissue, and subsequent replacing of the lost tissue [1–3]. In contrast, the mammalian heart loses its capability of a quantitatively sufficient proliferative response to various injuries soon after birth [4–6]. Interestingly, recent studies reported a certain degree of cardiomyocyte regeneration in various pathologic conditions, such as heart failure , orthotopic heart transplantation [8,9], or myocardial infarction [10,11], and even surmised the existence of resident cardiac stem cells , or extracardiac cardiomyocyte progenitor cells  in the mammalian heart. However, these findings are still controversial . In some of these studies, the presence of the Y-chromosome in hearts from female donors after gender-incompatible heart transplantation was used as evidence for population of the hearts by extracardiac progenitors. However, recent in vitro studies reported the occurrence of spontaneous cell fusions of putative stem cells, such as bone-marrow  or neural stem cells , with embryonic stem cells when cultured together, which challenges the concept of transdifferention and requires additional investigations in this field.
Nevertheless, the mechanisms of re-activated or preserved cardiomyocyte proliferation and regeneration in humans (for review see ) are obviously not sufficient to prevent extensive scar formation and the relentless disease progression of heart failure after myocardial infarction or loss of contractile function with pressure or volume overload [18,19]. Despite progress in pharmacological treatment of the syndrome of heart failure , mortality and curtailment of quality of life in advanced stages of congestive heart failure are significant , and the question remains, whether additional approaches can be developed for repairing human myocardial tissue.
Taking the newt as a model of regenerating irreversibly damaged myocardium, the features of de-differentiation, proliferation, and re-differentiation, might be the basis for developing strategies for repairing and rebuilding the failing human heart: Presumably, the most desirable future goal would be the identification of the cellular and molecular mechanisms that prevent the adult mammalian cardiomyocyte from re-entering the cell cycle after loss of contractile tissue leading to therapeutic strategies that force surviving cardiomyocytes to de-differentiate, proliferate, and re-differentiate after myocardial damage [5,22]. However, another innovative approach, probably within nearer reach, is the exogenous delivery of less differentiated cells to the damaged heart, that retain to a certain degree their capability of proliferation, and subsequently may colonise the scar and differentiate into cells with contractile properties.
This review will try to characterise the results of recent attempts to rebuild injured myocardial tissue by transplanting cells of various origin and stage of differentiation, such as fetal and neonatal cardiomyocytes, skeletal myoblasts, and embryonic and mesenchymal stem cells. Advantages, disadvantages, and major obstacles of using cells of various levels of differentiation are discussed. In an attempt to initiate a new discussion on the future direction of cellular cardiomyoplasty, we endeavour to delineate the value of some new techniques and approaches that are looming at the apparent (sometimes hazy) horizon for resolving problems and improving results in cellular cardiomyoplasty.
2 Fetal and neonatal cardiomyocytes—what can be achieved?
Transplanting fetal or neonatal myocytes can be regarded as the primary model of cellular cardiomyoplasty. Being of myocardial origin, these cells differentiate towards an adult cardiomyocyte phenotype, when remaining in an appropriate environment, but still retain a certain capability of proliferation [4,23]. Less differentiated cells, delivered to scarred myocardial tissue, always must undergo differentiation into a contractile phenotype to serve as a substitute for injured myocardial tissue.
2.1 Six months, and still alive
The first pioneer studies reported survival of fetal or neonatal heart tissue over the first weeks, when transplanted into a non-physiological environment, such as the anterior chamber of the eye, the pinna of the ear, or the subcutaneous tissue of the rat hind limb [24–27]. Soon after, the formation of stable grafts over 2 months in normal hearts was demonstrated after injection of a suspension of cultured fetal cardiomyocytes in mice , and fetal or neonatal cardiomyocytes in rat hearts . A recent quantitative analysis, using detection of the Sry gene of the Y-chromosome after injection of male neonatal cardiomyocytes into female normal rat hearts, suggested a survival rate of 15% of the initially injected number of cells after 12 weeks . Similarly, in damaged myocardial tissue, such as scars created by cryoinjury , temporary , or permanent coronary occlusion [32,33], stable transplants were described after intramyocardial injection of fetal or neonatal cardiomyocytes or fragments of fetal ventricular tissue, but not after injection of adult cardiomyocytes , that showed signs of coagulation necrosis already 1 day after transplantation . The notable exception to the growing body of evidence supporting survival in the damaged heart, reported in a porcine model of chronic myocardial infarction by Watanabe et al.  remains unexplained: In this study, fetal as well as neonatal cells only survived in normal myocardium, but surprisingly not in the infarct. Viability of engrafted cells in infarcted myocardium now has been demonstrated for up to 6–7 months after transplantation of fetal and neonatal cardiomyocytes [34,32] with a survival rate of 60% of the initial number of injected neonatal cells in a permanent coronary occlusion model in the rat after 6 months . Reasons for greater survival in the infarct model than the non-infarcted model from Müller-Ehmsen's  work may be related to less washout of cells from an infarct with compromised blood supply than from viable myocardium with an intact vasculature. Also, a lack of contractile force within the scar may have led to better cell survival by less trauma, or less pushing of the cells out of the region of initial injection.
2.2 Early declining proliferative activity
The proliferative activity of transplanted cells may be essential for colonising the scar tissue after transplantation. In the investigations by Reinecke et al. neonatal grafts were positive for PCNA (proliferating cell nuclear antigen) from 3 days to 2 weeks after transplantation with a maximum at day six after transplantation . Labelling with tritiated thymidine in Soonpaa's mice transplantation study, demonstrated a decline from 29% 3H-thymidine-positive cells at gestational day 15 to 0.6% 19 days after transplantation . After injection of neonatal cardiomyocytes at three sites into the ischemic risk zone, Scorsin et al. described the cells as remaining grouped around the injection site . Even 5–7 months after transplanting fetal cardiomyocytes, Connold et al. reported that the majority of cells was detected around the previous injection site  with a limited degree of migration.
2.3 Incomplete differentiation and integration
Function of the heart depends crucially on its microanatomic structure as an electrical and mechanical syncytium. Therefore, integration into the host and differentiation toward the adult phenotype are critical for the transplant to successfully contribute to cardiac function. In the fundamental investigations by Soonpaa et al., transplanting fetal cardiomyocytes into normal mice hearts led to a close juxtaposition of graft and host cells with circumferential alignment of graft cells parallel to the host myocardium, and importantly intercalated disks between host and graft cells, as demonstrated by electron microscopy . After transplantation of neonatal cells in cryoinjured hearts, graft cells demonstrated an increasingly polar alignment of N-cadherin, an adherens junction protein, and connexin-43 (, also ). Albeit close spatial approximation between host and graft cells early after transplantation in the study by Reinecke et al., most grafts were separated from the host by scar tissue by 8 weeks after transplantation with only few contact sites and gap junctions between host and graft, visible by confocal microscopy. This tendency of separation between host and graft was even more pronounced in the long-term follow-up by Müller-Ehmsen et al. , who described the graft as ‘confluent clusters of myocardial cells in the scar’ 6 months after transplantation, and importantly did not find any gap junctions between graft and host, but frequently between grafted cells. Transplanted immature cardiomyocytes, characterized by their small cell size, high nuclear to cytoplasm ratio and loose myofibrillar organization early after transplantation , showed an increase in graft-cell diameters up to 8 weeks post-transplantation and myofibrillae, forming complete sarcomeres, junctional complexes and abundant mitochondria [28,29], indicating a certain differentiation and switch to hypertrophic growth patterns. Fetal cardiomyocytes in culture appeared spherical with disorganised myofilaments and sarcomeric structures, while 4 weeks after transplantation the graft had developed sarcomeres, desmosomes and fasciae adherentes . However, 65 days after engrafting of fetal myocardial tissue in Leor's experiments, the transplant was still positive for α-actin on immunostaining, the actin-isoform typical of fetal and smooth muscle cells [31,33], indicating incomplete differentiation at the level of protein expression.
2.4 Do they contract?
Summarizing these basic findings, one has to state that complete regeneration of the damaged heart by transplantation, i.e. replacement of the scar by contractile tissue, has not yet been accomplished. In longer follow-up studies, the graft appeared mainly surrounded by collagen and scar tissue, meaning incomplete mechanical and electrical integration, although the principle development of intercalated disks between host and graft cell was convincingly demonstrated in normal hosts. To date, no study has been able to show synchronous beating of the graft within the infarcted myocardium in vivo. Thus, does the graft contribute to systolic force development, a process that requires adequate electrical and mechanical synchronisation? Four weeks after transplantation of fetal cardiomyocytes in cryoinjured hearts, the graft beat spontaneously when excised en bloc, as reported by Sakai et al. ; and the graft obtained after injection of fetal and neonatal cells in the subcutaneous tissue by Li et al., demonstrated a local electrogram resembling an ‘idioventricular rhythm’ , showing the principally preserved contractile capacity after engraftment.
A substantial number of investigations demonstrated improved left ventricular function after cellular cardiomyoplasty: Four weeks after transplantation of fetal cardiomyocytes in cryoinured hearts, Li et al. demonstrated higher developed left-ventricular pressures in the transplanted group in an ex vivo Langendorff-preparation . In vivo assessment of left ventricular function by transthoracic echocardiography in a model of reperfused myocardial infarction in rats revealed higher ejection fractions, as assessed from two-dimensional images, and cardiac output, calculated from Doppler profile in the pulmonary outflow tract 1 month after transplantation in a study by Scorsin et al. . A long-term follow-up over 6 months after transplantation in infarcted rat hearts, created by permanent coronary occlusion, similarly showed higher developed pressures in a Langendorff-preparation, along with smaller end-diastolic left ventricular diameters and increased fractional shortening, measured by transthoracic echocardiography .
Are the observed improvements in left ventricular performance really due to an active synchronous contraction, triggered by rare host cell-to-graft cell contacts or mechanical stretch? do the grafts beat independently, similar to Sakai's excised tissue? or does cellular cardiomyoplasty favourably influence the remodelling process after myocardial infarction by altering the mechanic and elastic properties of the infarcted myocardium by thickening and stiffening the scar with subsequent reduction of systolic wall stress?
Six months after intramyocardial injection of neonatal cardiomyocytes, left ventricular function, as assessed by intravenous angiography by Müller-Ehmsen et al. , was improved in terms of higher ejection fractions, less dyskinesia in the infarcted zone, along with thickening of the scar on autopsy, despite the excessive separation of graft and host tissue by connective tissue. Similarly, engraftment of fetal cardiomyocytes in Etzion's study, using a permanent coronary occlusion model in rats, showed that transplantation prevented scar thinning and progressive left ventricular dilation over 53 days after transplantation, hinting at favourable effects on the remodelling process as the main mechanism of improved cardiac performance . A recent study, in which fetal cardiomyocytes were injected 4 weeks after permanent coronary artery ligation, emphasized that cellular cardiomyoplasty did not reverse already developed left ventricular dilation, indicating that the effects on cardiac remodelling depend on the time the cells are transplanted  (see also ). However, it has to be emphasized that the type of transplanted cells is important: Sakai et al. compared the results after transplantation of fetal cardiomyocytes, enteric smooth muscle cells and skin fibroblasts, and convincingly demonstrated a higher degree of improvement of ex vivo parameters of left ventricular function in the cardiomyocyte group, indicating that elastic and contractile properties of the transplanted cells are not negligible [39,45].
In their permanent occlusion model, Etzion et al. reported a tendency toward enhanced vascularization of the scar after transplantation, that, however, did not reach statistical significance . Four weeks after fetal cell transplantation, Li et al. counted a significantly higher number of arterioles and venules in the transplanted tissue , and a recent study demonstrated a marked increase of capillary density with slight improvement of regional myocardial blood flow after cardiomyocyte transplantation into scars created by cryoinjury . One might speculate that cardiomyocyte transplantation exhibits some of its beneficial effects by enhancing scar perfusion, and thereby promoting infarct healing, or in addition, by releasing growth factors with influence on the healing process.
On the other hand, a recent study in infarcted mice hearts demonstrated pronounced improvement of echocardiographic parameters and higher force development in isolated muscle strips after transplantation, along with intact cellular electrical excitability of the transplanted cells . The authors speculated that a direct contribution of the graft to systolic force development might be the main cause of improved left ventricular function. This study, which included 153 animals, was the first to report an improvement of survival in the transplanted group, but the follow-up was relatively short, and most animals in the control group died within the first weeks.
Some studies assessed the effect of cellular cardiomyoplasty in animal models of global heart failure, in which the separation of the graft by scar tissue might be less important. The feasibility and survival was demonstrated in dystrophic mice and dog hearts by Koh et al. . Yoo et al. provided evidence for at least partial survival of transplanted cardiomyocytes in cardiomyopathic hamsters for 4 weeks. Ex vivo evaluation in a Langendorff-perfusion system demonstrated higher developed pressures in the transplanted group . Interestingly, Scorsin et al. demonstrated improved left ventricular function in doxorubicin-induced heart failure in mice 1 month after fetal cardiomyocyte-transplantation, as assessed in vivo by transthoracic echocardiography, although the transplanted cells (except for one out of 12 animals) apparently did not survive over the period of 1 month .
2.5 Translation into clinical practice?
In summary, transplantation of immature cardiomyocytes in infarcted myocardium exhibits beneficial effects on cardiac function without eye-catching differences between fetal and neonatal cardiomyocytes. Especially in long-term studies, the main effect might predominantly be due to favourable influence on the remodelling process rather than replacing contractile tissue with synchronously beating graft tissue. To date, direct translation of this increasingly recognized body of research into clinical practice is not feasible, as immature cardiomyocytes are not available for therapeutic measures in humans. If available at all, the issue of an appropriate immunosuppressive regime would have to be resolved. With progress in the insight of cell-cycle re-entry in differentiated cardiomyocytes [22,51,52], one might speculate that biopsied cardiomyocytes could be forced to transiently proliferate in vitro and then used for autologous grafting, which would also obviate the need for immunosuppression. However, now that alternative sources of cells, which might be more readily available, such as skeletal myoblasts or stem cells, are under investigation, cardiomyocyte transplantation might be regarded as a scientific model of what can be achieved.
3 Skeletal myoblasts—substitute for loss of cardiomyocytes with higher availability?
Following skeletal muscle injury, satellite cells enter the mitotic cell cycle, fuse with each other and with damaged myofibers, and subsequently restore skeletal myocyte tissue . Therefore, the properties of proliferation and differentiation, as well as the autologous source of availability, suggest that these cells might be a useful source for cardiac repair. In various species, such as dogs, rabbits and rats, engraftment of skeletal myoblasts successfully colonised injured cardiac tissue [54–57]. Survival has been shown for 12 weeks after transplantation into normal hearts , up to 18 weeks in cryoinjured myocardium , and Pouzet et al. recently reported survival, along with improved cardiac performance over a period of 1 year after transplantation . While the possibility of amplifying satellite cells in vitro, and potentially continuing proliferation after transplantation, as well as a higher tolerance of ischemia, are desirable advantages of skeletal myoblasts, the central question is whether they can function as cardiac-like myocytes, including adaptation to chronic workload and integration into the host. Several studies demonstrated that transplantation of skeletal myoblast formed multinucleated, cross-striated myofibers [54,58,61], including alignment in the cardiac fiber axis in a model of global heart failure . Around 4–7 weeks after transplantation, the expression of β-myosin heavy chain indicated a switch to low-twitch fibers [58,56]. Some studies described certain cardiac-like characteristics of the graft by histology [54,61], and the study by Chiu et al. even reported central nuclei and histologic evidence for intercalated disks . However, the systematic investigations by Reinecke et al. did not support the concept of a true transdifferentiation into cardiomyocytes, as demonstrated by the lack of α-myosin-heavy-chain, cardiac troponin I, and atrial natriuretic peptide-expression in the grafts. In culture, undifferentiated skeletal myoblast expressed N-cadherin and connexin-43 and intercalated disks in co-culture with cardiomyocytes, but with further differentiation into myotubes both proteins were markedly downregulated . Importantly, no expression of N-cadherin or connexin-43 was detected in differentiated grafts after 4 and 12 weeks . However, on a functional basis, the switch to fatigue-resistant, slow twitch fibers, may indicate an adaptation to chronic workload. In Murry's investigations, high frequency ex vivo stimulation of a 2-week-old myoblast graft induced tetanus, demonstrating the skeletal phenotype on the electrophysiological level, but by alternating tetanus and relaxation they could simulate a cardiac-like duty cycle over at least 6 min .
On the other hand, continuing proliferation after transplantation might also lead to undesirable disturbance of local left ventricular geometry, and Murry's group suggested that mechanisms of controlling the amount of proliferation after transplantation might be necessary [64,65]. Another important issue, which must carefully be investigated, is the effect of transplanted myoblasts on electrical stability of the heart, as they do not form normal electrical junctions with the host, and some animal models suggest that this could lead to a significantly reduced arrhythmic threshold .
Besides these unresolved issues, several studies reported improved left ventricular performance after myoblast transplantation: Jain et al. reported reduced left ventricular dilation, increased ex vivo systolic pressures and interestingly, improved in vivo exercise capacity after myoblast transplantation in a temporary coronary occlusion model in rats . Atkins et al. emphasized beneficial effects on diastolic left ventricular function . Pouzet et al. suggested that the effects are additive to angiotensin-converting-enzyme inhibition . In comparison with fetal cardiomyocyte transplantation, Scorsin et al. reported equal effectiveness . Again, the majority of the investigation proposed beneficial effects on the remodelling process as the main mechanism of improved left ventricular function, similar to the findings with cardiomyocyte transplantation.
The promising findings have led to the first clinical applications of skeletal myoblasts. Menasché's group and others [71,72] transplanted autologous human skeletal myoblast into myocardial scar tissue during coronary artery bypass surgery. In addition to feasibility and safety of the initial procedure, preliminary results indicated viability and thickening in the grafted scar, as evaluated by positron emission tomography and echocardiography [73,74]. Also, they reported an excellent availability of myoblasts from a single human biopsy.
4 Stem cells—an alternative source for cellular cardiomyoplasty?
The pluripotent nature and self-regenerating capacity of stem cells provide another potential source of cells that might be used for cardiomyoplasty. In vitro expansion and directed differentiation into a cardiomyocyte-like phenotype might theoretically lead to an unlimited pool of transplantable cells. Two major sources are under consideration: embryonic stem cells and mesenchymal stem cells, which are bone marrow-derived, and could be available for autologous transplantation. Whether additional sources might be available remains to be determined. For instance, endothelial cells isolated from embryonic vessels or endothelial cells in culture, as well as human umbilical vein endothelial cells were shown to express the cardiomyocyte phenotype after co-culture with rat cardiomyocytes .
4.1 Embryonic stem cells
Embryonic stem cells differentiate into a cardiomyocyte phenotype in vitro, and several groups developed different techniques for identifying or selecting cells that are committed to the cardiac lineage: Klug et al. transfected mice embryonic stem cells with a fusion gene consisting of the α-cardiac myosin heavy chain promoter and a cDNA encoding aminoglycoside phosphotransferase. After selection in an appropriate selection medium (G418), they obtained a culture of cardiomyocytes with >99% purity. In vitro synchronous contractile activity was demonstrated up to 11 months; and transplantation into dystrophic mice resulted in stable intracardiac grafts . Another in vitro study used transfection with the green fluorescent protein with subsequent Percoll gradient centrifugation and fluorescence-activated cell sorting, yielding 97% pure population of cardiomyocytes that could be characterised by immunohistochemistry and electrophysiological studies as predominantly ventricular cardiomyocytes . Cardiomyocyte-like cells, obtained by similar approaches, were characterised by immunohistology for troponin T , troponin I , gap junction protein connexin , α-cardiac-myosin heavy chain ; and functionally with regard to Ca2+ signalling . By means of patch-clamp techniques specific ion channels  and phenotypes resembling the sinus node, atrial cells or ventricular cardiomyocytes could be distinguished . These basic investigations, which were mainly obtained in mouse embryonic stem cells, were recently confirmed in cardiomyocytes originating from human embryonic stem cells: After establishing a stable stem cell line from human blastocyts , typical features of cardiomyocytes were demonstrated by immunohistology, electron microscopy and electrophysiological studies . In addition, the capability of differentiation into endothelial cells was recently demonstrated .
The findings indicate that embryonic stem cells provide an enormous potential for cellular cardiomyoplasty, as in theory the availability, even for humans, might be unlimited; any stage of differentiation, when used for transplantation, can be chosen, and even non-cardiomyocytes, such as endothelial cells might be derived from these sources. Klug reported stable engraftment of cardiomyocytes derived from embryonic stem cells . A recent study reported improvement of cardiac contractile function after intramyocardial injection of non-differentiated embryonic stem cells in a rat model of permanent coronary artery occlusion . Apparently, embryonic stem cells differentiated into a cardiomyocyte-like phenotype after engraftment, as assessed by immunostaining. Kehat et al. demonstrated electrical coupling and synchronous contractions of cardiomyocytes derived from human embryonic stem cells with rat cardiomyocytes in co-culture . These promising results will probably lead to an exploding increase in research in the next years. However, unresolved ethical issues, as well as the potential peril of introducing zoonoses into these lines when cultured on murine feeder cells  should be clarified in advance.
4.2 Bone marrow derived mesenchymal stem cells
Similarly to embryonic stem cells, bone marrow derived stem cells are regarded as multipotent, with the capability of replicating in the undifferentiated stage and differentiating into lineages of mesenchymal tissue [90,91]. With regard to cellular cardiomyoplasty, these cells seem to be optimal, as they could be transplanted autologously without need of immunosuppression (in contrast to embryonic stem cells) and expanded in vitro . After 5-azacytidine treatment of cultured murine bone marrow cells, Makino et al. were able to select cardiomyocyte-like cells on the basis of spontaneous beating . These cells were characterised by positive immunostaining for (among others) atrial natriuretic peptide and myosin, demonstrated typical sarcomeres and central nuclei on electron microscopy, and showed pacemaker-like or ventricular cell-like action potentials with a long plateau phase. The same group reported the expression of β-adrenergic and muscarinic receptors after induction of differentiation by 5-azacytidine, while α-adrenergic receptors seemed to be present even before differentiation . Toma et al. reported that human mesenchymal stem cells were able to differentiate into a cardiomyocyte phenotype after injection into the left ventricle of adult mice . However, recent in vitro studies have challenged this concept of transdifferentiation of bone marrow cells into cardiomyocytes, as they demonstrated, albeit at a low rate, the occurrence of spontaneous cell fusions between bone-marrow cells and co-cultured embryonic stem cells . With respect to these findings, one would demand the demonstration of a diploid karyotype or the absence of markers identifying host cells in these cardiomyocytes in order to prove transdifferentiation.
In addition, not all the research groups have been able to reproduce these results. Lewis et al. investigated markers of cardiac differentiation by immunofluorescence in vitro. Their extensive analysis did not support the occurrence of transdifferentiation either with or without 5-azacytidine treatment in vitro . Similarly, Murry et al. did not find evidence of cardiomyocyte differentiation of haematopoetic stem cells, transplanted either in normal or injured mice heart over a period of 35 days after transplantation .
Another proposed effect of transplanting bone marrow-derived stem cells seems to involve augmentation of angiogenesis: Shintani et al. injected bone marrow mononuclear cells into the ischemic rabbit hind limb. They reported a higher capillary density and more collateral vessels in comparision with saline-injected or fibroblast-injected animals . Similarly, in myocardial infarction created by coronary artery ligation, an increase in angiogenesis was demonstrated after intramyocardial injection of bone marrow-derived mononuclear cells [99,100]. Apparently, bone marrow-derived cells cannot only induce increased angiogenesis, but also might become a part of the vasculature and thus directly contribute to the angiogenic process, as demonstrated with different labelling techniques [101–103].
Tomita et al. reported enhanced angiogenesis after intramyocardial injection of bone marrow into 3-week-old cryoinjured myocardium of rats. Cardiac muscle-like cells, which stained positively for myosin heavy chain and troponin I, were observed in the scar tissue after 8 weeks. However, only the group with injection of 5 azacytidine treated bone marrow cells, which were shown to develop myosin heavy chain and troponin I positive myotubes in vitro, demonstrated higher systolic and developed pressures in a Langendorff-preparation 8 weeks after transplantation .
In conclusion, bone marrow-derived cells are a promising approach for cellular cardiomyoplasty, but many issues have to be resolved. For example, Orlic et al. demonstrated a marked amount of regenerated myocardium after intramyocardial injection of bone marrow derived cells [105,106]. In another publication, they suggested that systemic administration of cytokines, such as granulocyte colony stimulating factor and stem cell factor, could mobilise bone marrow cells and induce a significant translocation of bone marrow cells into damaged myocardium with subsequent cardiac repair without the need of invasive transplantation . From a scientific standpoint, it seems to be important to clearly distinguish between effects on vascularization , direct contractility or ventricular remodelling before entering the clinical realm .
5 How to improve the results
The preceding considerations might draw a general, probably incomplete picture of the current state of cellular cardiomyoplasty. Many problems that seem to be unresolved have to be addressed in future research before a clinical application is appropriate. The following paragraph lists some recent ideas and suggestions regarding how to best improve the results of cellular cardiomyoplasty (Fig. 1). Some of these findings are only published in abstract form, and the purpose of this paragraph is more the initiation of a new discussion on cellular cardiomyoplasty rather than a scientific report on established results.
5.1 The technique of transplantation
After intramyocardial injection into myocardial scar tissue, the graft remained largely separated from the host by collagen and connective tissue. Suzuki et al. developed an experimental technique of transplantation of a skeletal muscle cell line through the coronary artery with subsequent heterotopic transplantation of the heart into the abdomen . They demonstrated discrete loci of multinucleated myotubes that aligned with cardiac fibers and expressed connexin-43 after 4 weeks. Using a similar approach, this group demonstrated increased ex vivo parameters of myocardial contractility in doxorubicin-induced heart failure rats . In large animals, a transvenous approach for transplantation was proposed: Thompson et al. delivered cultured bone marrow cells of pigs via a special catheter and transvenous needle (TransAccess™), that was advanced to the interventricular vein via the coronary sinus . Similarly, a transendocardial approach, using a special device (NOGA™-system), was employed to transplant cultured human bone marrow cells  in pigs, and recently for the first time in patients . Via retrograde infusion through the great coronary intraventricular vein, Hou et al. were able to effectively distribute endothelial cells into pig myocardium .
5.2 Tissue processing and culturing technique
The early declining proliferative activity after transplantation might be a problem that hinders complete colonisation of the scar. Pouzet et al. reported a highly significant correlation between the number of injected myoblasts and the improvement in ventricular function , suggesting an important influence of the number of engrafted cells on functional outcome. Sheikh et al. suggested that overexpression of the long and short isoform of the fibroblast growth factor results in an increase in fibroblast growth factor-2 mediated proliferation of postnatal rat cardiac myocytes , thus representing an opportunity of enhancing proliferation of cardiomyocytes, and a potentially beneficial adjunct for the culturing procedure . Similarly, Soonpaa et al. created transgenic mice with overexpression of cyclin D1, resulting in a modest increase in cardiac proliferation .
In addition, a significant number of transplanted cells might undergo necrotic or apoptotic cell death after transplantation . Heat shock treatment or activation of the Akt-kinase pathway might increase survival after transplantation [116,117], leading to improved colonisation of the scar.
Pouzet et al. reported that pre-injecting the tibialis anterior muscle with bupivacaine, as a pharmacological stressor, that mimics muscle trauma, to activate muscle satellite cells, led to a significantly increased number of available myoblasts for transplantation .
Another potentially beneficial approach might be culturing the cells in a biodegradable three-dimensional scaffolds, usually referred as tissue engineering . Leor et al. were able to seed fetal cardiomyocytes into alginate sponges; and transplantation of these biografts led to attenuated left ventricular dilation . However, this technique of tissue engineering for cardiac transplantation is still in development [120–122]. Li et al. described a bioengineered graft consisting of fetal cardiac cells and a three-dimensional gelatine mesh. Spontaneous beating within the graft was demonstrated, but after engraftment in myocardial scar tissue no striking improvement of contractility was demonstrated .
5.3 Simultaneous gene therapy with cardiomyoplasty?
Transplanting the cells of various origins provides another opportunity for a therapeutic approach. As these transplanted cells survive, they might be used as a vector for delivering continuous gene expression to the infarcted heart. Koh et al. were able to use skeletal myoblasts for long-term delivery of transforming growth factor-β-1 into normal mice hearts . A number of recent studies suggested beneficial effects of combined cellular cardiomyoplasty and gene delivery with respect to cardiac perfusion [46,125,126], and even a superior effect in comparison with adenovirus-mediated gene transfer .
Another approach was suggested by Murry et al.: They transplanted a replication-defective adenovirus, containing MyoD under transcriptional control of the Rous sarcoma virus long terminal repeat. The virus converted cultured cardiac fibroblasts to skeletal muscle, indicated by expression of myogenin and skeletal myosin heavy chains in vitro. A high dose of the virus was able to induce skeletal myocyte induction within the scar tissue . This study was further supported by Etzion et al., who suggested that this approach might provide an unlimited source of cells for transplantation .
6 Conclusions and summary
Cellular myoplasty is a promising approach for treatment of congestive heart failure. Regenerating the myocardium after injury in newts requires de-differentiation, proliferation and re-differentiation of cardiomyocytes; and understanding the mechanisms and causes for the loss of this regenerative capacity in (its close relative) the mammalian heart  might be the basis for cardiac regenerative therapy in the future. Applying these basic features of regeneration, cellular cardiomyoplasty seems to be a promising step leading to regenerative strategies for the damaged heart: Transplantation of immature cardiomyocytes mainly provides a scientific model for investigating what can be achieved by different approaches of transplantation. Various sources of cells at various stages of differentiation, but with the capability of differentiating into a contractile phenotype might be available, such as skeletal myoblasts or embryonic or mesenchymal stem cells. The effects on the transition of heart failure, the best therapeutic approach, and different techniques of how to optimise results after transplantation remain to be determined.