Final Report Summary - CARDIOCELL (Development of cardiomyocyte replacement strategy for the clinic)
Executive summary:
Heart diseases represent a major health problem in Europe in general. From a clinical standpoint it has been clear that the adult human heart has an insufficient if any potential to repair itself after injury. The current strategies in the treatment of myocardial infarctions are to reduce the injury but no regenerative efforts have succeeded. The incomplete regenerative capacity of the adult heart has together with the high prevalence of cardiac diseases in Europe created a great need for regenerative therapies.
Despite a large number of clinical trials with the clear aim to regenerate heart muscle with the use of bone-marrow cells the overall experience is that no new heart muscle is generated and the functional improvement is of such low magnitude, if any at all, that alternative treatment strategies, relying on the regeneration/replacement of myocardium needs to be explored. The work within CARDIOCELL consortium focused on opening up the understanding of the normal development of heart muscle cells, finding out the inherent regenerative capacity to pave the way to future clinical studies based on regeneration/replacement of myocardium rather than supportive pleiotropic effects.
The consortium was managed through the fourth work package.
We have established mechanisms by which the earliest development of the heart involves important fate decisions and mechanisms guiding, governing the development of the earliest cells that will later develop into a heart muscle cell. The understanding of these mechanisms enables scientists to guide cells to develop into heart muscle cells in cultures. Cells in cultures are however never enriched for only one cell type. Even worse cell cultures from embryonic stem cells do in their non-enriched form hold a potential to form tumours. We therefore put an enormous effort in finding ways to select ou/purify the cells of heart muscle cell potential. We did that by identifying surface proteins on cells which in combination were unique to heart muscle cells (like bar codes) so that they could purified in from all populations of diverse types of cells which often is the case when stem cells are stimulated to grow and differentiate.
In contrast to what we anticipated the adult heart was shown to harbour an inherent generative mechanism even in adults. This finding opened up a completely new field in medicine; How to repair the heart by stimulating the endogenous repair mechanisms rather than transplanting cells?
Project Context and Objectives:
Myocardial infarction and ischemic heart disease are major health problems in Europe. They cause the loss of functional heart muscle cells and consequently lead to a large number of patients with impaired heart function. Patients with severe heart failure have a very poor prognosis comparable to those with a malignant disease because of the lack of contractile cardiomyocytes. Medical treatment is only supportive whereas the only heart muscle cell replacement therapy for the moment is heart transplantation. Due to mismatch between supply of donor organs and the need for them, patients waiting for heart transplantation die before they get transplanted. Cell replacement has therefore emerged as an attractive alternative treatment strategy. This may be why the first reports on the bone marrow-derived cells switching fate from blood into functional cardiomyocytes in mice almost immediately initiated clinical trials. The initial clinical trials based on bone-marrow cells (BMC) introduced into the heart have been rather disappointing because the improvement of left ventricular function has been very modest, if significant at all. Of the first two fully blinded randomised trials to date on BMC into infarcted hearts, the German REPAIR-AMI (Assmus et al., 2006; Schachinger et al., 2004) showed a minor effect compared to placebo, evaluated with a radiography-based method to determine left ventricular systolic function, and the Norwegian ASTAMI which was negative based on findings with magnetic resonance imaging, the latter method being a more sensitive and specific method to evaluate cardiac function than radiography (Lunde et al., 2006). These less than encouraging results are not surprising since recent experimental evidence has seriously questioned the concept of 'trans-differentiation' for the heart and other organs (Balsam et al., 2004; Murry et al., 2004; Nygren et al., 2004). Alternative explanations have been aadopted e.g. That the bone-marrow cells might control the remodelling process of the infarcted heart. These effects are currently under investigation and this strategy has been supported by the EU in the last call of FP6 and represents an important strategy to improve cardiac function.
The overriding rationale to study progenitor and stem cells from the adult heart is both fundamental and applied. First, these newly appreciated cells (Beltrami et al., 2003; Meilhac et al., 2004a; Oh et al., 2003) raise wholly novel research questions regarding the underlying biology of cardiac homeostasis and self-repair. For example, the details of cardiac myogenesis in the adult context are not understood in relation to the proteins, pathways and molecular partners that operate in embryos; the lineage relations are unclear; and the contribution of latent progenitor cells to cardiac muscle cell number largely remains to be proven, although very recent fate-mapping studies suggest their likely importance in disease (Hsieh et al., 2007) Second, heart-derived progenitor and stem cells have been substantiated by multiple independent laboratories as predisposed to cardiogenic differentiation, and a second distinguishing feature from hematopoietic stem cells or other bone marrow derivatives is the presence of many of the pivotal regulators that execute cardiac myogenesis in vivo. Thus, there exists a uniquely compelling reason to investigate heart-derived progenitor and stem cells, to undertake a clonal analysis of these populations, to establish improved markers and procedures for their purification, and to identify optimal conditions for their commitment to entering a cardiac muscle lineage, in relation to the fundamental principles of cardiac development that inform this FP7 project overall.
In the context of stem cells for cardiac therapy, the characteristics of those found in the embryo that will contribute the myocardium, the vasculature, and other cellular components of the developing heart, provide important insights. An interesting development in recent years has been the reports on resident stem cells in the adult heart (Beltrami et al., 2003; Oh et al., 2003). It is quite likely that stem cells isolated from the adult heart are "relics" of embryogenesis, stem cells that were sequestered and remained quiescent. Regulatory factors and signalling pathways that govern the behaviour of cardiac stem cells in the embryo may well prove to be important in the manipulation of adult cardiac stem cells. These considerations also apply to cardiogenic derivatives of ES cells (Moretti et al., 2006). These may open new possibilities to find cell populations, which could be expanded for possible use in humans for cell therapeutic purposes.
The use of adult and embryonic stem cells in potential therapies for heart disease incites the imagination of clinicians, basic researchers, and the lay public alike. CARDIOCELL is philosophically committed to the premise that the clinical deployment of cell therapies for heart disease will occur most safely and effectively only once a better substrate of enabling scientific knowledge is established.
The major objective of the CARDIOCELL consortium was to identify new sources of cells with the potential to replace heart muscle cells. The work was organised to retrieve basic science knowledge about immature heart muscle cells (with regenerative potential) and local stem cells to enable the development of new strategies for heart muscle cell replacement therapies towards ischemic heart disease. The consortium was formed as a to addresses the objectives of Topic HEALTH-2007-2.4.2-5: Cell therapies for the treatment of heart ischemia 'optimising cell preparations for use in cell transplantation studies'.
Project Results:
Every search for cell therapy against myocardial disease should involve specific interest in cardiomyocytes. The rational to establish WP1 was that understanding the development of cardiomyocytes will be a crucial step to identify the phenotype of early cardiomyocytes and to identify possible sources for new cardiomyocytes.
Lineage analysis of stem cell precursor populations
During the period of the CARDIOCELL project, the Buckingham group have built on our initial finding that two myocardial cell lineages contribute to the heart. We have correlated this with gene expression in the progenitor cells that contribute to different myocardial subdomains. This information provides a basis for generating cardiomyocytes from pluripotent stem cells (ES or iPS) or from endogeneous heart stem cells for repair of different parts of the heart.
At the arterial pole of the heart, we have examined the lineage relationship between heart muscle and the skeletal muscles of the head. Further we showed clonality between right ventricular myocardium and the temporalis and jaw skeletal muscles derived from first branchial arch of the neck. Eye muscles; also showed relationship with right ventricular myocardium. Heart muscle at the base of the aorta and pulmonary trunk shows relationship with facial expression muscles derived from the mesodermal core of the second branchial arch. Furthermore, these muscles on the left or right side of the head are related to pulmonary trunk or aortic heart muscle, respectively. This analysis [9] therefore demonstrates that a common immature gives rise to arterial pole heart muscle and skeletal muscles of the head, with important implications for cardiomyocyte progenitor populations in the context of ES or iPS derivatives and also for congenital heart defects. In the latter context, we are collaborating with pediatric cardiologist colleagues to examine linkage with head muscle defects, such as that seen in diGeorge syndrome patients.
This clonal analysis [16] therefore establishes a lineage tree for the venous pole of the heart. The unexpected clonality with pulmonary trunk myocardium raises the question of how progenitor cells that contribute posteriorly to the venous pole, also contribute anteriorly to the arterial pole of the developing heart. Cell tracking experiments after injection of fluorescent dyes and mouse embryo culture, by Nigel Brown's laboratory, have demonstrated that some cells situated in the posterior second heart field move anteriorly, acquiring expression of marker genes of the anterior second heart fields, to contribute to the outflow region of the heart [17].
We investigated the mammal heart muscle development with the use of a mouse model which generated green fluorescence in it's heart muscle cells. This model enabled us to sort out differentiated heart muscle cells from fetal, newborn as well as adult hearts and by comparing the non-heart muscle cells to the heart muscle cells the genes that are 'turned on' in heart muscle cells can be identified as heart muscle genes. Interesting enough these were then identified three crucially different stages with distinct different types of heart muscle cells; fetal phase involving only mono-nucleated and normal two copies of each gene. The newborn phase with exponentially increasing proportion of heart muscle cells with double nuclei and increasing proportion of multiple copies of DNA. The adult phase with greater than 95% bi-nucleated heart muscle cells as well significant proportion of nuclei with multiple copies of DNA [3].
To understand the interplay for early fate decisions of cells to develop into the heart muscle pathway gene regulators, transcription factors, were manipulated and the effects were elaborated. Two important factors Islet1, being a factor in the earliest cells important in heart development and Fgf10 being a factor important to the second heart field.
The Fgf10 gene is expressed in the anterior part of the mouse second heart field (SHF) and it was a transgene insertion at -117 kb upstream of Experiments confirm Islet1 binding to the 1.7 kb Fgf10 cardiac enhancer in the SHF, with continued weak binding in the heart, where Nkx2-5 now binds very strongly. This leads us to propose the following model for transcriptional regulation of Fgf10, as an example of genes implicated in cardiac myogenesis.
In contrast to Islet1, which is a transcriptional activator in the cardiac context, we have also investigated the role of Prdm1 (Blimp1), a transcriptional repressor present in the anterior SHF, and also in the surrounding ectoderm and endoderm. The Prdm1 null mutation leads to loss of most of the branchial arches and to heart defects, however it is not clear whether these are due to a primary effect on the SHF. We have therefore gone on to examine a mesoderm specific deletion of Prdm1 (Prdm1flox/flox; Mesp1Cre/+). We show that this results in myocardial malformations of the arterial pole of the heart and defects in the 4th pharyngeal arch arteries. This is probably mainly due to the defect in proliferation of SHF progenitors that we observe. When crossed onto a Tbx1+/- heterozygote background the mesodermal Prdm1 mutant has much more serious defects demonstrating genetic interaction between Tbx1 and Prdm1, through the FGF signalling pathway (see Task 1.3). It is planned to submit a manuscript on the role of Prdm1 by the end of 2012. CARDIOCELL support will of course be acknowledged. This work corresponds to the work predicted in deliverable
Recently, the Patient lab has shown that the anterior blood/vessel 'mother cell' and heart precursors share a common genetic control, both requiring gata 4, 5 and 6 function for proper development [10]. However, the signals that determine cardiac versus blood/endothelial fates in the anterior lateral plate mesoderm (ALPM) are currently unknown. A role for fibroblast growth factor (Fgf) signalling in heart development has been demonstrated but whether it controls cell fate, survival or proliferation is unknown (reviewed in [18]. We therefore compared the effects of Fgf signalling on blood, vessel and heart development in the zebrafish. We found that the loss of cardiac tissue seen when Fgf signalling is inhibited is accompanied by an increase in blood and endothelium, and that this reflects a stable change of fate rather than an effect on survival or proliferation, monitored by apoptosis and proliferation assays. Measurement of expression of genes of lycat, which lies upstream of both etsrp and scl genes in the blood/vessel precursor genetic cascade [19] and nkx2.5 one of the earliest regulatos of heart muscle cell development, revealed that both genes are expressed in early embryos. Furthermore, on treatment with SU5402, a potent inhibitor of Fgf receptor the expression of lycat was enhanced whilst nkx2.5 levels were diminished, thus strongly suggesting that these two genes could be mediating the effects of Fgf signalling to establish cardiac and blood/vessel identity. These observations, together with our previous experiments where we have shown that nkx2.5 alone cannot mediate the effects of Fgf signalling [11], led us to hypothesize that Fgf acts simultaneously to repress the blood/vessel development via lycat whilst positively regulates the cardiac programme via nkx2.5. To test this hypothesis, we overexpressed nkx2.5 in blood/vessel precursor deficient embryos. Due to the technical difficulties of over expressing mRNA in mutant embryos (only 25% of the offspring carry the mutation) we utilised the scl and etsrp inhibitor in combination. These experiments showed that, by an Fgf-independent repression of the blood/vessel precursor regulators together with an overexpression of the cardiac regulator nkx2.5 a bigger heart is achieved, with both atrial and ventricular gene expression being upregulated. We therefore show that Fgf signalling is simultaneously and independently regulating the expression of nkx2.5 and repressing the expression of blood/vessel precursor regulator, lycat, a mechanism indispensable for the development of a proper sized heart field. Over all, our observations indicate that the ratio of cardiac to blood/endothelial cells in the developing embryo is determined in part by the magnitude of Fgf signalling, and that an elevation of Fgf signalling represents a mechanism by which the anterior blood/vessel precursor population could have been recruited into the HF during evolution. This work identifies Fgf and Fgf signalling pathways as important mediators in heart development.
Both Fgf8 and Fgf10 are expressed in the cardiac precursor cells of the anterior SHF, however Fgf10 mutants do not have a cardiac phenotype, whereas Fgf8 mutants had been shown to have arterial pole defects. The Buckingham group have analysed Fgf8/Fgf10 double mutants where a conditional mutation of Fgf8 alleles was directed to the mesodermal cells that contribute to the heart by using an Mesp1Cre. We show that when Fgf gene dosage is reduced progressively severe defects are observed at the arterial pole of the heart. These include defects due to a reduction of myocardium at the base of the great arteries as well as malformation of the associated arteries due to defects in the contribution of endothelial cells from the mesodermal SHF core of the posterior branchial arches. These defects, which reveal a role for Fgf10 as well as Fgf8, can be explained by reduced proliferation of SHF progenitors as FGF signalling is reduced.
This work is being prepared for publication.
The Patient group and others have shown that the cardiac and blood/vessel precursor programmes are under common genetic control yet antagonistic to each other [10, 11, 21, 24]. The Patient group has postulated that during evolution this antagonism has been resolved in favour of the cardiac programme, recruiting more mesoderm into the heart, possibly representing the SHF, at the expense of blood/vessel precursor output. A candidate for resolving this antagonism in favour of the heart was Islet1, a bona fide marker of the SHF. Using the frog, Xenopus, as an evolutionary midpoint between fish and mammals, we found that Islet1, one of the earliest regulators of heart development, is actually required early in both cardiac and Blood/vessel precursor induction. Isl1 is first expressed at the end of the earlier embryo development (gastrulation) in an anterior domain that includes lineages of both first and second heart field as well as primitive blood tissue. Isl1 is a critical component of the early cardiac transcriptional network as, when it is knocked down, the expression of early markers of heart development are decreased before the onset of terminal myocardial marker expression and cardiac tube formation. Our analysis has also shown that Isl1 is actually instructive in inducing cardiac muscle, as opposed to an earlier proposed role in mouse mutants as a survival and proliferative factor [25].This finding agrees with embryonic stem cell models whereby Isl1 promotes cardiac development at the expense of other lineages [26, 27]. We have also shown that Isl1 is required for primitive red blood cell development. BMP signalling has been shown to be critical for differentiation of cardiac and blood lineages by us and others [28-31]. In the SHF there is a complex relationship between BMP signalling and Isl1 in which BMP signalling first activates Isl1 which then feeds back in a positive manner into BMP signalling. Hence, this study provides clues as to how Islet1 might govern the multi lineage potential of candidate cardiac stem/progenitor cells in the adult. Thus completing the deliverable 1.7.
The Patient group have shown that the cardiac and blood/vessel precursor programmes are under common genetic control yet antagonistic to each other [10, 11, 21, 24]. The common genetic control is provided by the transcription factors, gata4, 5 and 6 [10]. These discoveries were made by studying the roles of regulators expressed either in the cardiac or blood/vessel precursor territories or in both. One of the regulators expressed in the precursors of both tissues in zebrafish is a member of the NK2 family, nkx2.7. This gene has previously been shown to regulate, together with nkx2.5 zebrafish heart development.
The Patient group showed that disruption of nkx2.7 function causes defects in cardiac laterality. Indeed, we found that embryonic laterality was affected generally in nkx2.7 deficient embryos and was accompanied by random expression of left-right asymmetry associated genes. We went on to show that these findings in appearance of morphology. In Further, the Patient group showed disruption of nkx2.7 function causes defects in cardiac laterality. Indeed, we found that embryonic laterality was affected generally in nkx2.7 deficient embryos and was accompanied by random expression of left-right asymmetry associated genes. We went on to show that these phenotypes are preceded by disruption of the formation of the organ of assymetry, Kupffer's vesicle (KV)
We have optimized a protocol for the generation of high numbers of cardiomyocytes as published earlier [37], as well as a protocol with high reproducibility of cardiogenic differentiation based on Wobus et al. [38]. The hanging drop method takes advantage of a defined number of starting ESCs for differentiation into embryoid bodies resulting in high (differentiated cell clusters from ESC) reproducibility of the differentiation into the cardiac lineage. The mass culture method is used for the production of large numbers of cardiomyocytes but lacks the better reproducibility of differentiation of the hanging drop method. The hanging drop protocol on the other hand is more time-consuming and cost-intensive in relation to total cell numbers.
Changes to the published protocols mentioned above include the optimization of the setup of tissue culture dishes containing from ESC differentiated cell aggregations, embryoid bodies on a horizontal shaker and the optimization of shaking frequency. For the mass culture protocol we optimized the timing of the plating of embryoid bodies on culture dishes for enhanced cardiogenic differentiated. In addition we determined the number of passages for ESCs without a decrease in heart muscle differentiation potential which is important for long term culture of ESCs. Our results have been published in Current Protocols in Stem Cell Biology in the form of a detailed protocol [12].
To optimise heart muscle differentiation one need to identify them with high certainty. We have sorted out heart muscle cells from hearts on the basis of a genetically modified cells which fluorescence when the have taken the heart muscle fate decision. In these arrays eGFP positive and negative cells have been compared between fetal, new-born and adult time-points. These arrays have been published with extensive comparison identifying several important genes involved in the transition between the phases in development and thus identifying key therapeutic targets to address to unlock the resistance to cell division, and thus lack of reparative potential of adult heart muscle cells [3]. The overall problem, of having a mixture of cells and thus also cells bearing a tumour potential, when differentiating heart muscle cells from undifferentiated cells such as the ESC is one of the major obstacles to the field of regenerative heart therapy. The only field that has taken stem cell based therapy into clinical routine is the field of blood diseases, hematology. One of the major reasons is that this field of medicine has taken the effort to characterize the identity of a set of proteins on the cell surface of each cell type. This enables the purification of cell populations, containing only a specific cell type. These proteins on the cell surface thus function as a bar code system to identify cells. By sorting out heart muscle cells on the basis on the fluorescence gene (that is only activated in cells that have taken fate to decision to become heart muscle cells) the heart muscle cells could be compared with the cells, which haven't taken this decision (all other cells than heart muscle cells). In this way we could identify genes activated and proteins uniquely expressed in heart muscle cells. Of these the surface marker proteins were especially studied and thus these were characterised. In early fetal heart muscle cells the combination of having a marker VCAM1 and not having another one PECAM-1 was unique to heart muscle cells . These are of special interest since they have the capacity to divide and be expanded in vitro.
We have expanded this work to also involve the culturing of these fetal heart muscle cells making it possible to study these in clonal studies by co-culturing them with fetal cardiac connective tissue cells. This work has been completed and is under review currently [13]. In these studies we developed a sorting protocol with VCAM-1 and PECAM-1 as positive and negative markers to sort out fetal cardiomyocytes with high accuracy.
These were fully competent with regard to beating capacity and had an action potential, which to a high degree was of ventricular (72.2%) to lesser degree atrial (16.6%) and to even lesser degree (11.1%) of un-different type .
Moreover did these cells show the characteristics of heart muscle cells based on the action potentials retrieved from individual cells (ECG like recordings)
The work on early heart muscle cells was expanded further since the fundamental traits of these heart muscle cells (mono-nucleated, having a normal amount of DNA and thus readably undergoing cell division) where so favourable. One of the earliest regulators of gene expression in early heart muscle development, Nkx2.5 is not uniquely specific for the heart muscle cells. However the cardiac enhancer element is uniquely active in cardiomyocytes but only up to embryonic day 12.0 [39]. With the use of mice trans-genic with a cassette of this enhancer element driving eGFP (the gene that makes the cells fluorescence) we sorted out heart muscle cells on the basis of Nkx2.5-eGFP. Since fewer cells can be retrieved in this way a completely different strategy had to be adopted. The Jovinge group therefore in collaboration with the Jacobsen group at Oxford University established the FluidigmTM platform. Thereby single cells were isolated on the basis of Nkx2.5 expression and was screened for a set of genes identified in previous micro-arrays. These will represent early cardiomyocytes and their progenitors and to make it possible not only to identify fetal phenotype cardiomyocytes but also identify possible early progenitors.
These analyses were therefore done on a single cells level. These results identified new surface markers on cells expressing the most specific gene for heart muscle cells, cTropT at these early stages which were compared with the signals identified from the gene expression analyses performed on a-MHC-eGFP heart muscle cells. The surface markers identified by this comparison the combination of CD36, N-cadherin, ATP1b1[40], which enables the sorting of these in vivo as well as in vitro. One of the major obstacles in the work with massive gene-screens such as mRNA-based micro-arrays is the vast amount of information retrieved. We used the standard technique of triple micro-arrays comparing isolated cells from hearts of the a-mhc-eGFP mouse described above and made them comparative between the eGFP positive and negative populations. These were then ranked in order of expression ration between the comparators. From these data the surface markers were singled out. Still a high number of candidates were retrieved. In a strategy established in the Jovinge group under this grant; the murine-human conservative gene approach. In this strategy the significantly up-regulated surface markers were identified in the mouse were taken to a human protein data base (Human Protein Atlas) where they were tested. Surface markers identified as positive markers on human as well as in the mouse were tested as positive surface markers on human as well in mouse heart muscle cells (a conserved gene expression between species has a higher likelihood of being truly positive). Thus enhancing the process of finding true positive markers as well as murine marker.
In collaboration with several German research groups the Fleischmann group could identify a novel mechanism whereby mouse ESCs are pushed into pacemaker-like heart muscle cells. Thus, these studies provide an elegant and specific tool to direct the fate of pluripotent cells into a specific cardiomyocyte lineage. These studies are providing novel insights into the developmental and signalling pathways of the biology of pacemaker cells and are providing also insights and tools for bioengineering and therapeutic purposes [6].
The cultures of human ESC to differentiate into cardiomyocytes were tested on the basis of substrate; plastic matrigel and the most efficient protocol was found to be the one based on co-culturing differentiated cardiomyocytes on irradiated fetal cardiac fibroblasts [13]. The most efficient way to differentiate cardiomyocytes from ESC of all tested methods was shown to be already published by Christine Mummery's group (not as part of CARDIOCELL's work; EB cultures in co-cultures with the endoderma cell line END2 [41].
By the planning of CARDIOCELL consortium an intense conflict of the field has been whether there is any regeneration in the adult mammal heart at all. Some groups have identified extensive regeneration while the traditional view has been that adult mammal heart has no regenerative capacity.
Analysis of cells with regenerative potential within the adult heart
From other tissues cells, based on their features in an flow-cytometric analysis cells with a special special feature to pump out fluorochrome fast are referred to sipe population (SP) cells. This feature has been shown to enrich for cells with stem cell like properties.
From the adult heart of mice SP cells were propagated for more than 10 months and 300 population doublings, without obvious crisis or replicative senescence evident by the rates of growth. At passage 15-16, less than 5% of the cells showed features of aging. Four cells expanded from an individual cell isolated, so-called clones, were tested at later stages, with less than 1% positive at passages 25 and 0.2% positive at passage 33-42. Cloned cells were highly enriched for surface marker Sca-1 and the SP phenotype and, at 14 passages, generated secondary clones with an efficiency of greater than 40%.
Cloned cardiac SP cells are self-renewing and can maintain their phenotype after long-term propagation. The SP cells were self-renewing cardiac Sca-1+ cells. Cloned cardiac SP cells resembled an incomplete form of eaqrliest heart anlage, the cardiac mesoderm.
Next, it was crucial to determine whether cloned cardiac SP cells retained the many cardiogenic transcription factors that are measurable in the starting Sca-1+ population[45] The most commonly expressed genes for cardiogenic transcription factors were Gata4, Gata6, Hand2, Mef2a, Mef2c, Tbx5, Tbx20, with infrequent representation of Hand1, Isl1, and Nkx2-5. The isolated cells from the adult heart were, despite of the adult environment of very immature origin, lacking the most mature markers of heart muscle cells.
In all four chambers, the majority of cardiomyocytes were Nkx2-5-fated, while with Isl1-Cre the LV labeling was patchy and preferential labeling of second heart field derivatives was seen. By flow cytometry, approximately 50% of the Lin-/Sca-1+ cells were Nkx2-5- or Isl1-derived; no difference was seen between the SP and non-SP fractions, suggesting these two populations share a similar embryological origin.
In summary, cardiac SP and non-SP cells alike are derived from the earliest heart anlage, pre-cardiac mesoderm.
Interesting enough these cells weren't completely restricted to heart muscle cell potential. The endothelial markers CD31 and vWF were detected in 4-8% of cells, and SM-MHC in 5-15%. Thus, tri-lineage (heart muscle, endothelial, smooth muscle cell) potential was confirmed in each of the clones.
A critical feature of this FP7 consortium was our ability to collaborate with the Jacobsen lab at Oxford (also a CARDIOCELL beneficiary) and undertake single-cell quantitative RT-PCR profiling of freshly isolated cardiac SP cells, whose rarity confounds alternative approaches. In addition, cell-by-cell resolution enables one to resolve patterns of co-expression that could be obscured within a population, no matter how rigorously purified. Single cell QRT-PCR (Fluidigm) on greater than 50 single fresh cardiac SP cells confirmed both the overall transcriptional signature of the ensemble of clones and heterogeneous expression of 'core' cardiac factors, suggesting a mechanistic explanation for the block to spontaneous differentiation. Expression of key cardiogenic transcription factors in mutually exclusive patterns may represent a mechanism to maintain the stem cell pool and prevent precocious differentiation. In addition, this analysis has unmasked differences between the SP and non-SP fractions, as well as a strong association of cardiogenic transcription factors with Pdgfra.
Task 2.5: Adult Generation of Cardiomyocytes
In our initial study on cardiomyocytes isolated from mice with the transgenic cassette of cardiac a-MHC promoter driving the eGFP and thus enabling the sorting of cardiomyocytes. In a method established in line of this work we sorted nuclei from cardiomyocytes enabling not only the micro-array on the cells sorted for gene expression but also to quantify the bi-nucleation rate and the ploidy of each nuclei. This study involved a micro-array of the genes expressed in cardiomyocytes comparing fetal, neonatal to adult cardiomyocytes [3]. This study indicated that, in mice, cardiomyocytes are bi-nucleated, each nucleus is diploid and the cardiomyocytes do divide with full mitosis during fetal life. In the neonatal phase there is a rapid expansion in the proportion of bi-nucleated nuclei and the degree of ploidy and frequency of poly-ploid nuclei expand rapidly in the neonatal phase. In the adult phase > 95% of cardiomyoctes are bi-nucleated.
The mitotic capacity thus rapidly decreased at the neonatal phase to non-detected levels at the third post natal week. This therefore spoke against any adult regenerative capacity in the mammal heart. Moreover could micro-arrays comparing these important developmental shifts between the fetal, neonatal and adult phase. Several blocks in the cytokinesis structural genes involved in the contractile ring and cytokinesis could be identified such as anillin and Histone H3-like centromeric protein A. However the mouse heart expands relatively much less during the life span, which also is relatively shorter.
In humans the heart expands from the fetal stage to the adult individual to a much higher degree. Human life span is typically much longer than the mice life span of two years. Thus, the techniques used in the mouse studies aren't as accurate and feasible. A technique developed to date cell birth by the use of accelerator mass spectrometry to establish the C-14 content in the DNA was used to date the birth of human cardiomyocytes. In short lead the extensive above surface nuclear bomb testing to an increase in the C-14 content in the atmosphere. After the above surface tests a ban treaty was signed by Secretary General Chrustjev and President Johnson the immediate stop in the above surface tests led to a sharp decrease in the atmospheric C-14 content. This has increased the resolution of dating of organic molecules down to one year. The target molecule to study cell birth would be the DNA. One major obstacle in doing such studies in the heart is that heart muscle cells do only constitute 20-30% of the cells. Thus it is crucial that the nuclei of heart muscle cells are purified to a high degree to establish the age of the heart muscle cells. We thus developed a technique based on our nuclear studies in mice to sort out nuclei based on the cTropT and cTropI positivity of these nuclei.
This enabled the Jovinge group, in collaboration with the Frisén group and the Livermore laboratory, to date the cardiomyocytes retrieved from forensic autopsies from cases with a non-cardiovascular death cause. These studies also included the estimation of the amount of DNA in heart muscle cell nuclei. As in line with earlier published cases (unfilled circles) there is a certain degree of increase DNA (multiple of copies, more than the ordinary two) early in life as in mice. However, while the mouse DNA amount in heart muscle cells increases over time the first week and then become. By knowing the time pattern of this, it could be taken into consideration when estimates of heart muscle cell age were made. The average age of heart muscle cell was six years younger than the individual indicating adult generation of heart muscle cells. The best fit of estimates in mathematical modelling is with annual generation of heart muscle cells of 2% of the total population in younger ages but with a rapid decrease so that after seven decades only fractions of a per cent are renewed annually.
These studies were developed further and a new method identifying a unique heart muscle nucleus surface marker was identified enabling the retrieval of cardiomyocyte nuclei with high accuracy/purity even from frozen tissue [47]. With this technique different regions have been/are under investigation to possibly identify a regenerative zone in the adult heart. The right ventricle heart muscle cell have a very similar renewal rate kinetics (higher at younger ages with an exponential decrease in rate over the years) but at a lower level. This could reflect the higher load on the left heart.
Functional Single Cell Characterisation of Candidate Embryonic and Adult Stem Cell Precursor Populations
The single cell assay was established at the later part of the CARDIOCELL period and thus all identified progenitors have been characterised at the single cell level by phenotype and gene expression see below. However functional characterisation of the candidates has not been done yet.
A key feature in the characterisation of regenerative capacities is the mitotic capacity (potential and capacity to divide and thus expand) of the cells. In heart muscle cells this is especially complicated since they have been identified by the Jovinge group and others to contain multiple copies of DNA in mice [3] as well as in humans [4], thus traditional techniques are hard to apply in these studies.
Since heart muscle cells frequelntly do duplicate their nuclei and also multiply their DNA the traditional techniques of identifying cell division by DNA marking not feasible to estimate cell expansion through cell division in heart muscle cells. The only definitive proof for proliferative activity followed by cell division is the observation of a contractile ring prior to separation of daughter cells. To visualize these cell cycle specific events we developed a new in vivo proliferation marker that indicates this phase in great detail. This was achieved by fusing eGFP to the scaffolding protein Anillin, a component of the contractile ring, which then could be followed by its fluorescence.
A transgenic G4 ESC line was generated stably expressing the eGFP-Anillin fusion protein under the control of the ubiquitous CAG promoter. Undifferentiated ESCs of this line displayed a high overlap of eGFP-Anillin expression with the mitotic marker Ki-67 in immunofluorescent stainings (97%) and flow cytometry analysis (94%) respectively, proving its functionality. For easier detection of cardiomyocytes we generated a double transgenic ESC-line expressing RFP under control of the cardiac specific aMHC promoter. This ESC system is particularly powerful for screening assays for small substances bringing cardiomyocytes back in the cell cycle.
An alternative method to quantitatively assess proliferation in cardiomyocytes was established by the Jovinge group. In this workflow cytometry of isolated nuclei was used to determine percentages of proliferating cardiomyocytes during embryonic and postnatal development [3].
Characterisation of Plasticity of Cardiac Progenitors and Stem Cells in vivo
To elucidate the plasticity and the potential regenerative effect of resident adult cardiac stem/progenitor cells in the infracted heart we used Sca-1+ cells from the laboratory of Cardio Cell beneficiary Nr. 7, Dr. Michael Schneider. Sca1+ cells were shown to home to injured cardiac tissue when given intravenously and start to differentiate into cardiac-like cells [45]. These cells were transplanted into the border zone of mouse hearts after applying cryolesions, a model of myocardial injury resulting in highly reproducible lesions. We used 200.000 cells each of the clones TOT-AF10 or TOT-EG6 for transplantation per mouse, which could be tracked by either expression of eGFP or labelling with the membrane dye CM-Dil. Clone TOT-AF10 was transplanted into 6 mice while clone TOT-EG6 was transplanted into 4 mice. After dissection we could demonstrate robust engraftment of the transplanted cells into the cryoinjured hearts at 14- and 28 days after transplantation. In a second round of experiments two different clones of Sca1+ cells from a Hoechst negative side population (SP, dye efflux phenotype) were compared with nonSP Sca1+ cells in terms of their engraftment potential. The Sca1+ subclones SP3 and SP16 were transplanted into 5 mice each with cryo-lesions and into 3 mice each without lesions (control group). Hearts were dissected 14 days after transplantation and the degree of engraftment was assessed. As a control, cell culture medium was transplanted into 3 mice with cryo-lesion. Engrafted cells could be easily detected by red-orange fluorescence due to prior incubation with CM-Dil dye or transfection with an mOrange expressing lentivirus. After dissection we could demonstrate robust engraftment of the transplanted cells into the cryoinjured hearts at 14 days after transplantation, while in medium controls no cells could be detected. Multiple cells did stably engraft into the border zone as well as the infarcted area. There was better engraftment in cryo-lesioned hearts than in non-infarcted hearts while there was no obvious difference between engraftment of Sca1+ cells compared to SP-Sca1+ cells. These experiments have proved that engraftment of clonal Sca-1+, SP-cells is feasible and reliable.
Cardiac infarction results in permanent heart dysfunction due to insufficient myogenesis. This myogenic failure has been attributed to the absence of multipotent cardiovascular progenitor cells (CPC) in the heart, or alternatively to a context-related failure of extant CPC to adopt a cardiac fate within the infarct. We could show that CPC partially underlie myogenic infarct repair in neonatal mice, but not in adult mice subjected to the identical lesion. Infarction of postnatal day 1-3 mouse hearts induced the localized expansion of (c-kit)EGFP+ cells within the infarct, expression of early cardiac transcription factors, formation of striated cardiac myocytes, and prominent cardiac regeneration. Myogenic and vascular fates were also adopted by (c-kit)EGFP+ cells in explanted neonatal heart tissue, confirming the cardiac source of regenerative cells. Cre-recombinase mediated fate mapping and in vivo BrdU labeling experiments indicated that induction of undifferentiated precursors underlies localized myogenesis. By contrast, in the infarcted adult heart (c-kit)EGFP+ cells are induced in vivo and in explanted tissue, but adopt vascular fates; bone marrow reconstitution confirmed the cardiac source of vascular precursors. Thus adult post-infarct myogenic failure is not due to a context-dependent restriction of precursor differentiation, but rather to the limited potential of induced adult precursors Manuscript in preparation.
Characterisation of Immune Response after myocardial infarction to possibly optimise cell transplantation
In relation to myocardial injury such as infarction, there is an inflammatory response. Moreover can the immune response be directed towards extra-cellular pathogens which is TH2 and TH3 directed and stimulates the B-lymphocyte response/differentiation and thus the serological response to antigens. The defence against intra-cellular pathogens is however more directed towards cell death and thus more tissue destructive. NK cells play a pivotal role in this defence. The post inflammatory response is important to understand to optimise cell transplant survival. To evaluate how tissue destructive an infarction is in relation to its size we hypothesized that the more tissue aggressive an infarction (and its inflammatory component is) the larger component of the ischemic volume will be infarcted if only the infarction is handled appropriately fast enough.
To establish a method to screen for ischemic volume threat we compared the golden standard technique of single-photon emission computed tomography (SPECT) with Technetium isotope with a previously identified oedema sequence T2STIR. We screened > 700 acute ST-elevation infarction patients to get a population revascularised within six hours and with no previous infarction who, at catherisation still are occluded and only in one vessel. These inclusion criteria were strict to get well-defined infarctions to study the T2-STIR sequence performance.
This study showed a high compliance between the two techniques T2-STIR and SPECT.
By using these two techniques together we could establish a continuous variable of the hostility of the environment by comparing the ischemic volume to the truly infarcted volume. The proportion of the ischemic volume that in the end didn't infarct we denominated the salvage index and thus the more hostile infarction environment the lesser salvage index [50].
We this technique established we continued this task by running a study whereby we sampled STEMI patients for inflammatory cells to phenotype them with regard to their inflammatory cell response and in special to their tissue aggressive response and regulatory T-cell pattern and to relate these to the salvage index. These studies have been completed and are now being analysed and the manuscript is in preparation. Interesting enough the two inflammatory cell patterns that fall out significantly to the salvage index are the NK-cells (CD56+ CD3-) as well as CD28- lymphocytes.
Both these cell types have been identified as part of the tissue aggressive response. Other T-cell patterns examined so far (TH1, TH2 haven't shown such significant and strong correlations with the salvage index). These results are currently compiled in a manuscript and open up for new targets to optimize cell transplantation survival in myocardial infarction patients [50, 51]. This work has been published with regard to the MR technique needed to be developed for these studies [50, 52-54].
In the field of rheumatology new biological treatments have been a major break- through to the field. These are tailored to address specific inflammatory mechanisms. In the search to find treatment options to optimize infarction environment for cell grafting the effect on specific inflammatory cell patterns. In the first line of studies the TNF? antagonist Adalimimab was used in patients with previously not treated with biological anti-inflammatory medications [55]. It has the advantage of being administrated sc like insulin. The treatment was clinically effective as estimated by the clinical rheumatoid arthritis score DAS28. In this study it showed to decrease TH1 cells in peripheral blood just like the TH17 cells both being reported to be active in the tissue aggressive inflammatory response. However not identified as correlated to salvage index in the studies above. These studies have been completed and are being prepared for publication. However, future efforts will include other biological anti-inflammatory agents to hopefully identify compounds that address the inflammatory patterns identified above as unfavourable for infarction environment.
In the CARDIOCELL research portfolio human hESC derived cardiomyocytes were predicted to be tested in immuno deficient mice. At the time of writing the proposal this made sense. With our current experience these do not survive in the mouse, most likely due to the basal heart rate in the mouse to be 450-500 beats per minute (bpm) in contrast to the human 60-70. Another group reported 2012 a completely new model, the guinea-pig as suitable for this [56]. Thus it wasn't time-wise possible to set-up a completely new animal model within this consortium.
Prosthetic Myocardium: A Tissue Engineering Approach
An essential element in tissue engineering is the development of innovative types of scaffolds, which have appropriate physical and biological properties for the cells that are used to populate it. The scaffold should create an environment or 'niche' that allows the cells to mimic or adopt the function of the organ or tissue that is to be engineered. In this regard we have used molecular modelling and the generation of synthetic matrices, generated from the self-assembly of collagen and fibronectin mimetic peptides to develop an environment for adipose-derived stem cells to function as myocytes or valve cells. Using scaffolds derived from collagen and fibronectin mimetic peptides will eliminate concerns over the potential transmission of infectious agents or an immune response associated with animal derived materials. Extracellular matrix proteins are essential for cell adhesion, organisation and signalling and scaffolds derived from these proteins have applications for tissue engineering heart valves and myocardial tissue.
Summary of Achievements
1.Modelling of molecular interactions to predict tertiary structure of synthetic collagen mimetic peptides.
2.Provision of a blue-print for synthesis of mimetic peptides that will form a triple-helical structure.
3.Experimentally designed a suitable peptide sequence, which is able to replicate same kinds of intermediates and hydrogel formation as with natural collagen.
4.The honey-comb morphological structure of the resultant artificial hydrogel is suitable for cell-encapsulation studies.
5.The process of hydrogel formation involves the use of fully biocompatible biomaterials only which is suitable for preparation of three dimensionally cell-encapsulated hydrogels.
6.The peptide is easily scalable to any amount. Already we have designed a prototype of manual equipment which is suitable for the production of CMP-1 peptide in multigram scale quantities.
Molecular Modelling
Collagens constitute one of the main components of most tissues in the body. These supramomolecular structures perform essential functions which include maintaining the shape, physical characteristics and structural integrity of the tissue. Another vital function of the collagens is acting as a docking site for several growth factors and cytokines which regulate important functions of cells, such as differentiation, survival, polarity and motility. Collagens may therefore, play an essential role in the formulation of many biomaterials that are suitable for tissue engineering purposes.
The different sources of collagen for this purpose are from xenogenic, allogenic or recombinant materials, which have several disadvantages. These include risk of infections, evoking an immune response and lack of flexibility. These limitations have stimulated the exploration and development of collagen biomimetic peptides, from the bottom-up. Computational and experimental studies suggest that collagens show more preference for ion pairs than globular proteins, as evidenced by the frequent occurrence of Lys/Gly/Asp/Glu in the sequence of natural collagen. These studies suggested that introduction of ion pairs in short collagen-like peptides could increase the stability of the triple-helix.
Identifying such residue positions is much more complicated for a large multimeric collagen structure that contains two tropocollagen molecules. However, it has been recently shown that residues, which play a crucial role in the structural rigidity of a small protein can be efficiently identified using a graph theoretical approach implemented in the software Floppy Inclusions and Rigid Substructure Topography V6.2 (FIRST). We have used this approach in order to identify the structural hotspots, which are the positions and residues in the chain that are important for maintaining the structural integrity of multiple tropocollagen molecules by quantifying their contribution to the overall rigidity of the system. This analysis allows us to identify positions of the residues for further mutational modeling to cross-link the two tropocollagens as well to introduce a collagen cell binding (CB) motif, GFOGER. Furthermore, the stability and inter-helical cross-links of the constructed mutational models have been validated by the Framework Rigidity Optimized Dynamics Algorithm (FRODAN) and molecular dynamics (MD) simulations.
We have employed a systematic procedure from the bottom-up to encourage the triple-helix triple-helix association for enhancing the self-assembly and biofunctionalisation of collagen biomimetic peptides. A novelty of our approach comes from using a methodology based on constraints that allows us to explore how local atomistic changes affect global properties that may enhance self-assembly at a different scale. Collagen-like systems are especially suited for such a methodology since collagen exhibits such multiscale behavior as a result of the multiple molecular interactions that lead to constraint dynamics at different scales.
The mutational models we designed show how we could potentially drive building blocks of collagen-like peptides to self assembly by increasing the possibilities of their molecular level interactions. Our studies provide guidelines for the engineering of effective tropocollagen, with and without the common CB motif. Starting from atomistic descriptions at the smallest scale in terms of their constraint flexibility, allows inferring information about global structural integrity. The cross-links between ion pairs are further valuable for self-assembly while the formation of additional hydrogen bonds between inter-helical terminals are essential. They could favor the organization of building blocks by stitching their edges strongly using hydrogen bonds and avoiding positional overlaps of their edges, thereby assisting a proper spatial arrangement in the process of assembly. This hypothetical phenomenon attempts to mimic one of the characteristics of natural collagen (tapered tips at termini) that was previously reported. However, increasing the possibilities for cross-link is an encouraging sign to design peptides for many applications such as matrix biology.
A simple molecular based biomimetic collagen model has been cautiously fine-tuned to obtain a refined model with favourable molecular interactions that promote their self-association. As a result, the present study provides mutational models that utilize only non-covalent interactions of charged and key residues to establish association between tropocollagens. The model with the CB motif unveils an anonymous contribution of GFOGER to self association of collagen molecules. With these charge decorations, this model offers extra spots for other cell surface receptors to interact and share signals in and out. Additionally, the computational methods used here offer an efficient and effective tool to map the suitable positions for mutations of multimeric proteins. The presented models here indicate the importance of requirement of a multidimensional approach in collagen research. Since the mutation-dependent helical stability in collagen structure is directly related to genetic disorders, biomaterials/nanomaterials based on these stable mutational models with natural potency to self-associate may have significant implications in biomedicine.
Synthesis of Collagen Mimetic Peptides
The primary structure of collagen molecule contains about 1050 amino acids. Mimicking all properties of collagen using CMPs is very difficult owing to its very small length between 25-40 amino acids. Several CMPs are designed, synthesised and characterised with the following properties to increase the interaction between the triple helices in order to form hydrogels:
1. CMPs with free amino and carboxyl terminus. Polymerisation using standard SPPS method will increase the peptide length.
2. Hydrogen bonding interaction between amino and carboxyl terminus.
3. High temperature annealing
4. Metal-ion co-ordination
5. Chemical cross-linking between triple helices
6. Attaching small ß-sheet forming peptides with CMPs
7. Addition of very small amount of collagen as nucleating agent for hydrogel formation
8. Composites between Hyaluronic acid and CMPs
9. Chemical cross-linking between elastin mimetic peptides with CMPs
10. Enzymatic cross-linking between CMPs
We have designed, synthesised and characterised several peptides with all above mentioned characteristics and tested its triple helix formation. Of all above mentioned strategies, we found that the enzymatic cross-linking method is suitable for the preparation of hydrogels from CMPs. Enzymatic cross-linking of the following peptide results in the formation of hydrogels under physiological temperature.
CMP-1: (Gly-(Lys)2-(Gly-Pro-Hyp)4-Gly-Phe-Hyp-Gly-Glu-Arg-(Gly-Pro-Hyp)4-Gly-(Gln)2-Gly).
The purity of CMP-1 was analysed by HPLC and found to be more than 95% as synthesised and calculated molecular weight of the peptide is 3499.7 da. Experimentally observed molecular weight by Maldi analysis is 3499 da which is in very close agreement with predicted molecular weight.
The circular dichroism spectrum of the peptide CMP-1 in TRIS buffer shows positive peak at 225nm which shows the formation of triple helical structure and the intensity at 225nm is measured at different temperature which shows sigmoidal shape of melting. The melting temperature of the peptide is found to be 47°C F.
The melting transition of the designed CMP-1 obtained from CD measurements is further confirmed by differential scanning calorimetry (DSC).
In natural collagen the D-periodic structure (D=67nm) is formed from the intermolecular interactions that occur between telopeptide regions on both carboxyl and amino termini. The D-periodicity observed for the CMP-KQ assembly is approximately 7.6nm consistent with the small length of the peptide as compared with natural collagen.
The dimensions of the nanofibrous assembly are further confirmed by tapping mode AFM. TEM and AFM examination are in agreement with each other and confirms the formation of nanofibrous assembly upon incubation at 37°C in TRIS buffer.
Cryo-SEM images of the CMP-1 hydrogel was used to provide information about the structure of this material in its native state as drying of the hydrogel may result in aggregation. The micrograph also shows a honeycomb structure that is similar to the morphology of fibrous collagen. This honeycomb structure may be a very suitable material for tissue engineering applications such as in the 3-D encapsulation of cells.
Summary
We have achieved the preparation of artificial collagen hydrogel from enzymatic cross-linking of collagen mimetic peptides. The types of intermediate structures and processes adopted by the designed CMP in higher order formation are similar to the process adopted by natural collagen. The initially formed nanofibrous assembly is cross-linked enzymatically to form hydrogel that has a structure, which may make it a suitable candidate for the preparation of 3-D cell encapsulated scaffolds. The formation of a honeycomb structure indicates that the resultant hydrogel has room to accommodate cells, which are further able to interact with the hydrogel through the integrin-binding site in its peptide sequence.
Potential Impact:
4.2 Use and dissemination of foreground
4.2.1 Scientific Communication
-Results of the work performed within the consortium was communicated through traditional routes with publications in peer reviewed journals
-presentations at European and International Scientific meetings
-direct promotion of new research results to the research groups to whom this might be valuable.
-review publications; members of the consortium are key contributors to their respective fields and do in that respect write overview reviews in their respective fields which are one of the routes of scientific communications that offer the highest impact to the scientific field.
4.2.2 Medical Professionals
-Progress of the consortium has been reported to the cardiological and thoracic surgery societies of Europe with targeted articles for professional journals aimed at physicians and users including presentations at clinical conferences in Europe.
4.2.3 Public awareness
Public awareness to CARDIOCELL's ongoing activities will mainly be addressed by the following strategies:
-Articles for the press and other media
-Broadcasting media
-Presentations at science fairs
-Web page
The webpage contains a public section which clearly explains the purpose of the consortium and which is continuously is updated with the progress and the perceived impact on the society and of the anticipated results from the work of the consortium. This section will also contain contact information whereby the public can interact with the consortium.
The societal implications of the work of CARDIOCELL are represented mainly in the fact that alternative sources to bone marrow cells are available for generative approaches towards the heart. Through Internet; own page, EuroStemcell as well as the best-known cardiologist forum theheart.org interviews as well as reports and articles have been published to the public. One of the main findings, opened a completely new field to the cardiac regeneration; the endogenous regeneration of cardiomyocytes in the adult human heart. Despite to the general view of the field, that the human actually has an inherent regenerative capacity opens up for new therapeutic strategies that don't involve transplantation of cells to the heart. This report was cited and reported by laymen media; New York Times, Daily Mail, Reuters, Scientific American and National Geographic. By these findings Europe has taken the lead towards ne regenerative approaches involving the search for the source for this inherent mechanism and how this can be induced. The direct societal implication for the individual patient is however less for this very day. However the objectives of CARDIOCELL, from the very first draft to final technical annex excluded direct clinical use of the findings, rather we wished to enable the field to focus on new reliable sources which were fulfilled by the identification of fetal/pluripotent sources as well as to open up for endogenous regeneration as a therapy.
Project website: http://www.cardiocell.org
Heart diseases represent a major health problem in Europe in general. From a clinical standpoint it has been clear that the adult human heart has an insufficient if any potential to repair itself after injury. The current strategies in the treatment of myocardial infarctions are to reduce the injury but no regenerative efforts have succeeded. The incomplete regenerative capacity of the adult heart has together with the high prevalence of cardiac diseases in Europe created a great need for regenerative therapies.
Despite a large number of clinical trials with the clear aim to regenerate heart muscle with the use of bone-marrow cells the overall experience is that no new heart muscle is generated and the functional improvement is of such low magnitude, if any at all, that alternative treatment strategies, relying on the regeneration/replacement of myocardium needs to be explored. The work within CARDIOCELL consortium focused on opening up the understanding of the normal development of heart muscle cells, finding out the inherent regenerative capacity to pave the way to future clinical studies based on regeneration/replacement of myocardium rather than supportive pleiotropic effects.
The consortium was managed through the fourth work package.
We have established mechanisms by which the earliest development of the heart involves important fate decisions and mechanisms guiding, governing the development of the earliest cells that will later develop into a heart muscle cell. The understanding of these mechanisms enables scientists to guide cells to develop into heart muscle cells in cultures. Cells in cultures are however never enriched for only one cell type. Even worse cell cultures from embryonic stem cells do in their non-enriched form hold a potential to form tumours. We therefore put an enormous effort in finding ways to select ou/purify the cells of heart muscle cell potential. We did that by identifying surface proteins on cells which in combination were unique to heart muscle cells (like bar codes) so that they could purified in from all populations of diverse types of cells which often is the case when stem cells are stimulated to grow and differentiate.
In contrast to what we anticipated the adult heart was shown to harbour an inherent generative mechanism even in adults. This finding opened up a completely new field in medicine; How to repair the heart by stimulating the endogenous repair mechanisms rather than transplanting cells?
Project Context and Objectives:
Myocardial infarction and ischemic heart disease are major health problems in Europe. They cause the loss of functional heart muscle cells and consequently lead to a large number of patients with impaired heart function. Patients with severe heart failure have a very poor prognosis comparable to those with a malignant disease because of the lack of contractile cardiomyocytes. Medical treatment is only supportive whereas the only heart muscle cell replacement therapy for the moment is heart transplantation. Due to mismatch between supply of donor organs and the need for them, patients waiting for heart transplantation die before they get transplanted. Cell replacement has therefore emerged as an attractive alternative treatment strategy. This may be why the first reports on the bone marrow-derived cells switching fate from blood into functional cardiomyocytes in mice almost immediately initiated clinical trials. The initial clinical trials based on bone-marrow cells (BMC) introduced into the heart have been rather disappointing because the improvement of left ventricular function has been very modest, if significant at all. Of the first two fully blinded randomised trials to date on BMC into infarcted hearts, the German REPAIR-AMI (Assmus et al., 2006; Schachinger et al., 2004) showed a minor effect compared to placebo, evaluated with a radiography-based method to determine left ventricular systolic function, and the Norwegian ASTAMI which was negative based on findings with magnetic resonance imaging, the latter method being a more sensitive and specific method to evaluate cardiac function than radiography (Lunde et al., 2006). These less than encouraging results are not surprising since recent experimental evidence has seriously questioned the concept of 'trans-differentiation' for the heart and other organs (Balsam et al., 2004; Murry et al., 2004; Nygren et al., 2004). Alternative explanations have been aadopted e.g. That the bone-marrow cells might control the remodelling process of the infarcted heart. These effects are currently under investigation and this strategy has been supported by the EU in the last call of FP6 and represents an important strategy to improve cardiac function.
The overriding rationale to study progenitor and stem cells from the adult heart is both fundamental and applied. First, these newly appreciated cells (Beltrami et al., 2003; Meilhac et al., 2004a; Oh et al., 2003) raise wholly novel research questions regarding the underlying biology of cardiac homeostasis and self-repair. For example, the details of cardiac myogenesis in the adult context are not understood in relation to the proteins, pathways and molecular partners that operate in embryos; the lineage relations are unclear; and the contribution of latent progenitor cells to cardiac muscle cell number largely remains to be proven, although very recent fate-mapping studies suggest their likely importance in disease (Hsieh et al., 2007) Second, heart-derived progenitor and stem cells have been substantiated by multiple independent laboratories as predisposed to cardiogenic differentiation, and a second distinguishing feature from hematopoietic stem cells or other bone marrow derivatives is the presence of many of the pivotal regulators that execute cardiac myogenesis in vivo. Thus, there exists a uniquely compelling reason to investigate heart-derived progenitor and stem cells, to undertake a clonal analysis of these populations, to establish improved markers and procedures for their purification, and to identify optimal conditions for their commitment to entering a cardiac muscle lineage, in relation to the fundamental principles of cardiac development that inform this FP7 project overall.
In the context of stem cells for cardiac therapy, the characteristics of those found in the embryo that will contribute the myocardium, the vasculature, and other cellular components of the developing heart, provide important insights. An interesting development in recent years has been the reports on resident stem cells in the adult heart (Beltrami et al., 2003; Oh et al., 2003). It is quite likely that stem cells isolated from the adult heart are "relics" of embryogenesis, stem cells that were sequestered and remained quiescent. Regulatory factors and signalling pathways that govern the behaviour of cardiac stem cells in the embryo may well prove to be important in the manipulation of adult cardiac stem cells. These considerations also apply to cardiogenic derivatives of ES cells (Moretti et al., 2006). These may open new possibilities to find cell populations, which could be expanded for possible use in humans for cell therapeutic purposes.
The use of adult and embryonic stem cells in potential therapies for heart disease incites the imagination of clinicians, basic researchers, and the lay public alike. CARDIOCELL is philosophically committed to the premise that the clinical deployment of cell therapies for heart disease will occur most safely and effectively only once a better substrate of enabling scientific knowledge is established.
The major objective of the CARDIOCELL consortium was to identify new sources of cells with the potential to replace heart muscle cells. The work was organised to retrieve basic science knowledge about immature heart muscle cells (with regenerative potential) and local stem cells to enable the development of new strategies for heart muscle cell replacement therapies towards ischemic heart disease. The consortium was formed as a to addresses the objectives of Topic HEALTH-2007-2.4.2-5: Cell therapies for the treatment of heart ischemia 'optimising cell preparations for use in cell transplantation studies'.
Project Results:
Every search for cell therapy against myocardial disease should involve specific interest in cardiomyocytes. The rational to establish WP1 was that understanding the development of cardiomyocytes will be a crucial step to identify the phenotype of early cardiomyocytes and to identify possible sources for new cardiomyocytes.
Lineage analysis of stem cell precursor populations
During the period of the CARDIOCELL project, the Buckingham group have built on our initial finding that two myocardial cell lineages contribute to the heart. We have correlated this with gene expression in the progenitor cells that contribute to different myocardial subdomains. This information provides a basis for generating cardiomyocytes from pluripotent stem cells (ES or iPS) or from endogeneous heart stem cells for repair of different parts of the heart.
At the arterial pole of the heart, we have examined the lineage relationship between heart muscle and the skeletal muscles of the head. Further we showed clonality between right ventricular myocardium and the temporalis and jaw skeletal muscles derived from first branchial arch of the neck. Eye muscles; also showed relationship with right ventricular myocardium. Heart muscle at the base of the aorta and pulmonary trunk shows relationship with facial expression muscles derived from the mesodermal core of the second branchial arch. Furthermore, these muscles on the left or right side of the head are related to pulmonary trunk or aortic heart muscle, respectively. This analysis [9] therefore demonstrates that a common immature gives rise to arterial pole heart muscle and skeletal muscles of the head, with important implications for cardiomyocyte progenitor populations in the context of ES or iPS derivatives and also for congenital heart defects. In the latter context, we are collaborating with pediatric cardiologist colleagues to examine linkage with head muscle defects, such as that seen in diGeorge syndrome patients.
This clonal analysis [16] therefore establishes a lineage tree for the venous pole of the heart. The unexpected clonality with pulmonary trunk myocardium raises the question of how progenitor cells that contribute posteriorly to the venous pole, also contribute anteriorly to the arterial pole of the developing heart. Cell tracking experiments after injection of fluorescent dyes and mouse embryo culture, by Nigel Brown's laboratory, have demonstrated that some cells situated in the posterior second heart field move anteriorly, acquiring expression of marker genes of the anterior second heart fields, to contribute to the outflow region of the heart [17].
We investigated the mammal heart muscle development with the use of a mouse model which generated green fluorescence in it's heart muscle cells. This model enabled us to sort out differentiated heart muscle cells from fetal, newborn as well as adult hearts and by comparing the non-heart muscle cells to the heart muscle cells the genes that are 'turned on' in heart muscle cells can be identified as heart muscle genes. Interesting enough these were then identified three crucially different stages with distinct different types of heart muscle cells; fetal phase involving only mono-nucleated and normal two copies of each gene. The newborn phase with exponentially increasing proportion of heart muscle cells with double nuclei and increasing proportion of multiple copies of DNA. The adult phase with greater than 95% bi-nucleated heart muscle cells as well significant proportion of nuclei with multiple copies of DNA [3].
To understand the interplay for early fate decisions of cells to develop into the heart muscle pathway gene regulators, transcription factors, were manipulated and the effects were elaborated. Two important factors Islet1, being a factor in the earliest cells important in heart development and Fgf10 being a factor important to the second heart field.
The Fgf10 gene is expressed in the anterior part of the mouse second heart field (SHF) and it was a transgene insertion at -117 kb upstream of Experiments confirm Islet1 binding to the 1.7 kb Fgf10 cardiac enhancer in the SHF, with continued weak binding in the heart, where Nkx2-5 now binds very strongly. This leads us to propose the following model for transcriptional regulation of Fgf10, as an example of genes implicated in cardiac myogenesis.
In contrast to Islet1, which is a transcriptional activator in the cardiac context, we have also investigated the role of Prdm1 (Blimp1), a transcriptional repressor present in the anterior SHF, and also in the surrounding ectoderm and endoderm. The Prdm1 null mutation leads to loss of most of the branchial arches and to heart defects, however it is not clear whether these are due to a primary effect on the SHF. We have therefore gone on to examine a mesoderm specific deletion of Prdm1 (Prdm1flox/flox; Mesp1Cre/+). We show that this results in myocardial malformations of the arterial pole of the heart and defects in the 4th pharyngeal arch arteries. This is probably mainly due to the defect in proliferation of SHF progenitors that we observe. When crossed onto a Tbx1+/- heterozygote background the mesodermal Prdm1 mutant has much more serious defects demonstrating genetic interaction between Tbx1 and Prdm1, through the FGF signalling pathway (see Task 1.3). It is planned to submit a manuscript on the role of Prdm1 by the end of 2012. CARDIOCELL support will of course be acknowledged. This work corresponds to the work predicted in deliverable
Recently, the Patient lab has shown that the anterior blood/vessel 'mother cell' and heart precursors share a common genetic control, both requiring gata 4, 5 and 6 function for proper development [10]. However, the signals that determine cardiac versus blood/endothelial fates in the anterior lateral plate mesoderm (ALPM) are currently unknown. A role for fibroblast growth factor (Fgf) signalling in heart development has been demonstrated but whether it controls cell fate, survival or proliferation is unknown (reviewed in [18]. We therefore compared the effects of Fgf signalling on blood, vessel and heart development in the zebrafish. We found that the loss of cardiac tissue seen when Fgf signalling is inhibited is accompanied by an increase in blood and endothelium, and that this reflects a stable change of fate rather than an effect on survival or proliferation, monitored by apoptosis and proliferation assays. Measurement of expression of genes of lycat, which lies upstream of both etsrp and scl genes in the blood/vessel precursor genetic cascade [19] and nkx2.5 one of the earliest regulatos of heart muscle cell development, revealed that both genes are expressed in early embryos. Furthermore, on treatment with SU5402, a potent inhibitor of Fgf receptor the expression of lycat was enhanced whilst nkx2.5 levels were diminished, thus strongly suggesting that these two genes could be mediating the effects of Fgf signalling to establish cardiac and blood/vessel identity. These observations, together with our previous experiments where we have shown that nkx2.5 alone cannot mediate the effects of Fgf signalling [11], led us to hypothesize that Fgf acts simultaneously to repress the blood/vessel development via lycat whilst positively regulates the cardiac programme via nkx2.5. To test this hypothesis, we overexpressed nkx2.5 in blood/vessel precursor deficient embryos. Due to the technical difficulties of over expressing mRNA in mutant embryos (only 25% of the offspring carry the mutation) we utilised the scl and etsrp inhibitor in combination. These experiments showed that, by an Fgf-independent repression of the blood/vessel precursor regulators together with an overexpression of the cardiac regulator nkx2.5 a bigger heart is achieved, with both atrial and ventricular gene expression being upregulated. We therefore show that Fgf signalling is simultaneously and independently regulating the expression of nkx2.5 and repressing the expression of blood/vessel precursor regulator, lycat, a mechanism indispensable for the development of a proper sized heart field. Over all, our observations indicate that the ratio of cardiac to blood/endothelial cells in the developing embryo is determined in part by the magnitude of Fgf signalling, and that an elevation of Fgf signalling represents a mechanism by which the anterior blood/vessel precursor population could have been recruited into the HF during evolution. This work identifies Fgf and Fgf signalling pathways as important mediators in heart development.
Both Fgf8 and Fgf10 are expressed in the cardiac precursor cells of the anterior SHF, however Fgf10 mutants do not have a cardiac phenotype, whereas Fgf8 mutants had been shown to have arterial pole defects. The Buckingham group have analysed Fgf8/Fgf10 double mutants where a conditional mutation of Fgf8 alleles was directed to the mesodermal cells that contribute to the heart by using an Mesp1Cre. We show that when Fgf gene dosage is reduced progressively severe defects are observed at the arterial pole of the heart. These include defects due to a reduction of myocardium at the base of the great arteries as well as malformation of the associated arteries due to defects in the contribution of endothelial cells from the mesodermal SHF core of the posterior branchial arches. These defects, which reveal a role for Fgf10 as well as Fgf8, can be explained by reduced proliferation of SHF progenitors as FGF signalling is reduced.
This work is being prepared for publication.
The Patient group and others have shown that the cardiac and blood/vessel precursor programmes are under common genetic control yet antagonistic to each other [10, 11, 21, 24]. The Patient group has postulated that during evolution this antagonism has been resolved in favour of the cardiac programme, recruiting more mesoderm into the heart, possibly representing the SHF, at the expense of blood/vessel precursor output. A candidate for resolving this antagonism in favour of the heart was Islet1, a bona fide marker of the SHF. Using the frog, Xenopus, as an evolutionary midpoint between fish and mammals, we found that Islet1, one of the earliest regulators of heart development, is actually required early in both cardiac and Blood/vessel precursor induction. Isl1 is first expressed at the end of the earlier embryo development (gastrulation) in an anterior domain that includes lineages of both first and second heart field as well as primitive blood tissue. Isl1 is a critical component of the early cardiac transcriptional network as, when it is knocked down, the expression of early markers of heart development are decreased before the onset of terminal myocardial marker expression and cardiac tube formation. Our analysis has also shown that Isl1 is actually instructive in inducing cardiac muscle, as opposed to an earlier proposed role in mouse mutants as a survival and proliferative factor [25].This finding agrees with embryonic stem cell models whereby Isl1 promotes cardiac development at the expense of other lineages [26, 27]. We have also shown that Isl1 is required for primitive red blood cell development. BMP signalling has been shown to be critical for differentiation of cardiac and blood lineages by us and others [28-31]. In the SHF there is a complex relationship between BMP signalling and Isl1 in which BMP signalling first activates Isl1 which then feeds back in a positive manner into BMP signalling. Hence, this study provides clues as to how Islet1 might govern the multi lineage potential of candidate cardiac stem/progenitor cells in the adult. Thus completing the deliverable 1.7.
The Patient group have shown that the cardiac and blood/vessel precursor programmes are under common genetic control yet antagonistic to each other [10, 11, 21, 24]. The common genetic control is provided by the transcription factors, gata4, 5 and 6 [10]. These discoveries were made by studying the roles of regulators expressed either in the cardiac or blood/vessel precursor territories or in both. One of the regulators expressed in the precursors of both tissues in zebrafish is a member of the NK2 family, nkx2.7. This gene has previously been shown to regulate, together with nkx2.5 zebrafish heart development.
The Patient group showed that disruption of nkx2.7 function causes defects in cardiac laterality. Indeed, we found that embryonic laterality was affected generally in nkx2.7 deficient embryos and was accompanied by random expression of left-right asymmetry associated genes. We went on to show that these findings in appearance of morphology. In Further, the Patient group showed disruption of nkx2.7 function causes defects in cardiac laterality. Indeed, we found that embryonic laterality was affected generally in nkx2.7 deficient embryos and was accompanied by random expression of left-right asymmetry associated genes. We went on to show that these phenotypes are preceded by disruption of the formation of the organ of assymetry, Kupffer's vesicle (KV)
We have optimized a protocol for the generation of high numbers of cardiomyocytes as published earlier [37], as well as a protocol with high reproducibility of cardiogenic differentiation based on Wobus et al. [38]. The hanging drop method takes advantage of a defined number of starting ESCs for differentiation into embryoid bodies resulting in high (differentiated cell clusters from ESC) reproducibility of the differentiation into the cardiac lineage. The mass culture method is used for the production of large numbers of cardiomyocytes but lacks the better reproducibility of differentiation of the hanging drop method. The hanging drop protocol on the other hand is more time-consuming and cost-intensive in relation to total cell numbers.
Changes to the published protocols mentioned above include the optimization of the setup of tissue culture dishes containing from ESC differentiated cell aggregations, embryoid bodies on a horizontal shaker and the optimization of shaking frequency. For the mass culture protocol we optimized the timing of the plating of embryoid bodies on culture dishes for enhanced cardiogenic differentiated. In addition we determined the number of passages for ESCs without a decrease in heart muscle differentiation potential which is important for long term culture of ESCs. Our results have been published in Current Protocols in Stem Cell Biology in the form of a detailed protocol [12].
To optimise heart muscle differentiation one need to identify them with high certainty. We have sorted out heart muscle cells from hearts on the basis of a genetically modified cells which fluorescence when the have taken the heart muscle fate decision. In these arrays eGFP positive and negative cells have been compared between fetal, new-born and adult time-points. These arrays have been published with extensive comparison identifying several important genes involved in the transition between the phases in development and thus identifying key therapeutic targets to address to unlock the resistance to cell division, and thus lack of reparative potential of adult heart muscle cells [3]. The overall problem, of having a mixture of cells and thus also cells bearing a tumour potential, when differentiating heart muscle cells from undifferentiated cells such as the ESC is one of the major obstacles to the field of regenerative heart therapy. The only field that has taken stem cell based therapy into clinical routine is the field of blood diseases, hematology. One of the major reasons is that this field of medicine has taken the effort to characterize the identity of a set of proteins on the cell surface of each cell type. This enables the purification of cell populations, containing only a specific cell type. These proteins on the cell surface thus function as a bar code system to identify cells. By sorting out heart muscle cells on the basis on the fluorescence gene (that is only activated in cells that have taken fate to decision to become heart muscle cells) the heart muscle cells could be compared with the cells, which haven't taken this decision (all other cells than heart muscle cells). In this way we could identify genes activated and proteins uniquely expressed in heart muscle cells. Of these the surface marker proteins were especially studied and thus these were characterised. In early fetal heart muscle cells the combination of having a marker VCAM1 and not having another one PECAM-1 was unique to heart muscle cells . These are of special interest since they have the capacity to divide and be expanded in vitro.
We have expanded this work to also involve the culturing of these fetal heart muscle cells making it possible to study these in clonal studies by co-culturing them with fetal cardiac connective tissue cells. This work has been completed and is under review currently [13]. In these studies we developed a sorting protocol with VCAM-1 and PECAM-1 as positive and negative markers to sort out fetal cardiomyocytes with high accuracy.
These were fully competent with regard to beating capacity and had an action potential, which to a high degree was of ventricular (72.2%) to lesser degree atrial (16.6%) and to even lesser degree (11.1%) of un-different type .
Moreover did these cells show the characteristics of heart muscle cells based on the action potentials retrieved from individual cells (ECG like recordings)
The work on early heart muscle cells was expanded further since the fundamental traits of these heart muscle cells (mono-nucleated, having a normal amount of DNA and thus readably undergoing cell division) where so favourable. One of the earliest regulators of gene expression in early heart muscle development, Nkx2.5 is not uniquely specific for the heart muscle cells. However the cardiac enhancer element is uniquely active in cardiomyocytes but only up to embryonic day 12.0 [39]. With the use of mice trans-genic with a cassette of this enhancer element driving eGFP (the gene that makes the cells fluorescence) we sorted out heart muscle cells on the basis of Nkx2.5-eGFP. Since fewer cells can be retrieved in this way a completely different strategy had to be adopted. The Jovinge group therefore in collaboration with the Jacobsen group at Oxford University established the FluidigmTM platform. Thereby single cells were isolated on the basis of Nkx2.5 expression and was screened for a set of genes identified in previous micro-arrays. These will represent early cardiomyocytes and their progenitors and to make it possible not only to identify fetal phenotype cardiomyocytes but also identify possible early progenitors.
These analyses were therefore done on a single cells level. These results identified new surface markers on cells expressing the most specific gene for heart muscle cells, cTropT at these early stages which were compared with the signals identified from the gene expression analyses performed on a-MHC-eGFP heart muscle cells. The surface markers identified by this comparison the combination of CD36, N-cadherin, ATP1b1[40], which enables the sorting of these in vivo as well as in vitro. One of the major obstacles in the work with massive gene-screens such as mRNA-based micro-arrays is the vast amount of information retrieved. We used the standard technique of triple micro-arrays comparing isolated cells from hearts of the a-mhc-eGFP mouse described above and made them comparative between the eGFP positive and negative populations. These were then ranked in order of expression ration between the comparators. From these data the surface markers were singled out. Still a high number of candidates were retrieved. In a strategy established in the Jovinge group under this grant; the murine-human conservative gene approach. In this strategy the significantly up-regulated surface markers were identified in the mouse were taken to a human protein data base (Human Protein Atlas) where they were tested. Surface markers identified as positive markers on human as well as in the mouse were tested as positive surface markers on human as well in mouse heart muscle cells (a conserved gene expression between species has a higher likelihood of being truly positive). Thus enhancing the process of finding true positive markers as well as murine marker.
In collaboration with several German research groups the Fleischmann group could identify a novel mechanism whereby mouse ESCs are pushed into pacemaker-like heart muscle cells. Thus, these studies provide an elegant and specific tool to direct the fate of pluripotent cells into a specific cardiomyocyte lineage. These studies are providing novel insights into the developmental and signalling pathways of the biology of pacemaker cells and are providing also insights and tools for bioengineering and therapeutic purposes [6].
The cultures of human ESC to differentiate into cardiomyocytes were tested on the basis of substrate; plastic matrigel and the most efficient protocol was found to be the one based on co-culturing differentiated cardiomyocytes on irradiated fetal cardiac fibroblasts [13]. The most efficient way to differentiate cardiomyocytes from ESC of all tested methods was shown to be already published by Christine Mummery's group (not as part of CARDIOCELL's work; EB cultures in co-cultures with the endoderma cell line END2 [41].
By the planning of CARDIOCELL consortium an intense conflict of the field has been whether there is any regeneration in the adult mammal heart at all. Some groups have identified extensive regeneration while the traditional view has been that adult mammal heart has no regenerative capacity.
Analysis of cells with regenerative potential within the adult heart
From other tissues cells, based on their features in an flow-cytometric analysis cells with a special special feature to pump out fluorochrome fast are referred to sipe population (SP) cells. This feature has been shown to enrich for cells with stem cell like properties.
From the adult heart of mice SP cells were propagated for more than 10 months and 300 population doublings, without obvious crisis or replicative senescence evident by the rates of growth. At passage 15-16, less than 5% of the cells showed features of aging. Four cells expanded from an individual cell isolated, so-called clones, were tested at later stages, with less than 1% positive at passages 25 and 0.2% positive at passage 33-42. Cloned cells were highly enriched for surface marker Sca-1 and the SP phenotype and, at 14 passages, generated secondary clones with an efficiency of greater than 40%.
Cloned cardiac SP cells are self-renewing and can maintain their phenotype after long-term propagation. The SP cells were self-renewing cardiac Sca-1+ cells. Cloned cardiac SP cells resembled an incomplete form of eaqrliest heart anlage, the cardiac mesoderm.
Next, it was crucial to determine whether cloned cardiac SP cells retained the many cardiogenic transcription factors that are measurable in the starting Sca-1+ population[45] The most commonly expressed genes for cardiogenic transcription factors were Gata4, Gata6, Hand2, Mef2a, Mef2c, Tbx5, Tbx20, with infrequent representation of Hand1, Isl1, and Nkx2-5. The isolated cells from the adult heart were, despite of the adult environment of very immature origin, lacking the most mature markers of heart muscle cells.
In all four chambers, the majority of cardiomyocytes were Nkx2-5-fated, while with Isl1-Cre the LV labeling was patchy and preferential labeling of second heart field derivatives was seen. By flow cytometry, approximately 50% of the Lin-/Sca-1+ cells were Nkx2-5- or Isl1-derived; no difference was seen between the SP and non-SP fractions, suggesting these two populations share a similar embryological origin.
In summary, cardiac SP and non-SP cells alike are derived from the earliest heart anlage, pre-cardiac mesoderm.
Interesting enough these cells weren't completely restricted to heart muscle cell potential. The endothelial markers CD31 and vWF were detected in 4-8% of cells, and SM-MHC in 5-15%. Thus, tri-lineage (heart muscle, endothelial, smooth muscle cell) potential was confirmed in each of the clones.
A critical feature of this FP7 consortium was our ability to collaborate with the Jacobsen lab at Oxford (also a CARDIOCELL beneficiary) and undertake single-cell quantitative RT-PCR profiling of freshly isolated cardiac SP cells, whose rarity confounds alternative approaches. In addition, cell-by-cell resolution enables one to resolve patterns of co-expression that could be obscured within a population, no matter how rigorously purified. Single cell QRT-PCR (Fluidigm) on greater than 50 single fresh cardiac SP cells confirmed both the overall transcriptional signature of the ensemble of clones and heterogeneous expression of 'core' cardiac factors, suggesting a mechanistic explanation for the block to spontaneous differentiation. Expression of key cardiogenic transcription factors in mutually exclusive patterns may represent a mechanism to maintain the stem cell pool and prevent precocious differentiation. In addition, this analysis has unmasked differences between the SP and non-SP fractions, as well as a strong association of cardiogenic transcription factors with Pdgfra.
Task 2.5: Adult Generation of Cardiomyocytes
In our initial study on cardiomyocytes isolated from mice with the transgenic cassette of cardiac a-MHC promoter driving the eGFP and thus enabling the sorting of cardiomyocytes. In a method established in line of this work we sorted nuclei from cardiomyocytes enabling not only the micro-array on the cells sorted for gene expression but also to quantify the bi-nucleation rate and the ploidy of each nuclei. This study involved a micro-array of the genes expressed in cardiomyocytes comparing fetal, neonatal to adult cardiomyocytes [3]. This study indicated that, in mice, cardiomyocytes are bi-nucleated, each nucleus is diploid and the cardiomyocytes do divide with full mitosis during fetal life. In the neonatal phase there is a rapid expansion in the proportion of bi-nucleated nuclei and the degree of ploidy and frequency of poly-ploid nuclei expand rapidly in the neonatal phase. In the adult phase > 95% of cardiomyoctes are bi-nucleated.
The mitotic capacity thus rapidly decreased at the neonatal phase to non-detected levels at the third post natal week. This therefore spoke against any adult regenerative capacity in the mammal heart. Moreover could micro-arrays comparing these important developmental shifts between the fetal, neonatal and adult phase. Several blocks in the cytokinesis structural genes involved in the contractile ring and cytokinesis could be identified such as anillin and Histone H3-like centromeric protein A. However the mouse heart expands relatively much less during the life span, which also is relatively shorter.
In humans the heart expands from the fetal stage to the adult individual to a much higher degree. Human life span is typically much longer than the mice life span of two years. Thus, the techniques used in the mouse studies aren't as accurate and feasible. A technique developed to date cell birth by the use of accelerator mass spectrometry to establish the C-14 content in the DNA was used to date the birth of human cardiomyocytes. In short lead the extensive above surface nuclear bomb testing to an increase in the C-14 content in the atmosphere. After the above surface tests a ban treaty was signed by Secretary General Chrustjev and President Johnson the immediate stop in the above surface tests led to a sharp decrease in the atmospheric C-14 content. This has increased the resolution of dating of organic molecules down to one year. The target molecule to study cell birth would be the DNA. One major obstacle in doing such studies in the heart is that heart muscle cells do only constitute 20-30% of the cells. Thus it is crucial that the nuclei of heart muscle cells are purified to a high degree to establish the age of the heart muscle cells. We thus developed a technique based on our nuclear studies in mice to sort out nuclei based on the cTropT and cTropI positivity of these nuclei.
This enabled the Jovinge group, in collaboration with the Frisén group and the Livermore laboratory, to date the cardiomyocytes retrieved from forensic autopsies from cases with a non-cardiovascular death cause. These studies also included the estimation of the amount of DNA in heart muscle cell nuclei. As in line with earlier published cases (unfilled circles) there is a certain degree of increase DNA (multiple of copies, more than the ordinary two) early in life as in mice. However, while the mouse DNA amount in heart muscle cells increases over time the first week and then become. By knowing the time pattern of this, it could be taken into consideration when estimates of heart muscle cell age were made. The average age of heart muscle cell was six years younger than the individual indicating adult generation of heart muscle cells. The best fit of estimates in mathematical modelling is with annual generation of heart muscle cells of 2% of the total population in younger ages but with a rapid decrease so that after seven decades only fractions of a per cent are renewed annually.
These studies were developed further and a new method identifying a unique heart muscle nucleus surface marker was identified enabling the retrieval of cardiomyocyte nuclei with high accuracy/purity even from frozen tissue [47]. With this technique different regions have been/are under investigation to possibly identify a regenerative zone in the adult heart. The right ventricle heart muscle cell have a very similar renewal rate kinetics (higher at younger ages with an exponential decrease in rate over the years) but at a lower level. This could reflect the higher load on the left heart.
Functional Single Cell Characterisation of Candidate Embryonic and Adult Stem Cell Precursor Populations
The single cell assay was established at the later part of the CARDIOCELL period and thus all identified progenitors have been characterised at the single cell level by phenotype and gene expression see below. However functional characterisation of the candidates has not been done yet.
A key feature in the characterisation of regenerative capacities is the mitotic capacity (potential and capacity to divide and thus expand) of the cells. In heart muscle cells this is especially complicated since they have been identified by the Jovinge group and others to contain multiple copies of DNA in mice [3] as well as in humans [4], thus traditional techniques are hard to apply in these studies.
Since heart muscle cells frequelntly do duplicate their nuclei and also multiply their DNA the traditional techniques of identifying cell division by DNA marking not feasible to estimate cell expansion through cell division in heart muscle cells. The only definitive proof for proliferative activity followed by cell division is the observation of a contractile ring prior to separation of daughter cells. To visualize these cell cycle specific events we developed a new in vivo proliferation marker that indicates this phase in great detail. This was achieved by fusing eGFP to the scaffolding protein Anillin, a component of the contractile ring, which then could be followed by its fluorescence.
A transgenic G4 ESC line was generated stably expressing the eGFP-Anillin fusion protein under the control of the ubiquitous CAG promoter. Undifferentiated ESCs of this line displayed a high overlap of eGFP-Anillin expression with the mitotic marker Ki-67 in immunofluorescent stainings (97%) and flow cytometry analysis (94%) respectively, proving its functionality. For easier detection of cardiomyocytes we generated a double transgenic ESC-line expressing RFP under control of the cardiac specific aMHC promoter. This ESC system is particularly powerful for screening assays for small substances bringing cardiomyocytes back in the cell cycle.
An alternative method to quantitatively assess proliferation in cardiomyocytes was established by the Jovinge group. In this workflow cytometry of isolated nuclei was used to determine percentages of proliferating cardiomyocytes during embryonic and postnatal development [3].
Characterisation of Plasticity of Cardiac Progenitors and Stem Cells in vivo
To elucidate the plasticity and the potential regenerative effect of resident adult cardiac stem/progenitor cells in the infracted heart we used Sca-1+ cells from the laboratory of Cardio Cell beneficiary Nr. 7, Dr. Michael Schneider. Sca1+ cells were shown to home to injured cardiac tissue when given intravenously and start to differentiate into cardiac-like cells [45]. These cells were transplanted into the border zone of mouse hearts after applying cryolesions, a model of myocardial injury resulting in highly reproducible lesions. We used 200.000 cells each of the clones TOT-AF10 or TOT-EG6 for transplantation per mouse, which could be tracked by either expression of eGFP or labelling with the membrane dye CM-Dil. Clone TOT-AF10 was transplanted into 6 mice while clone TOT-EG6 was transplanted into 4 mice. After dissection we could demonstrate robust engraftment of the transplanted cells into the cryoinjured hearts at 14- and 28 days after transplantation. In a second round of experiments two different clones of Sca1+ cells from a Hoechst negative side population (SP, dye efflux phenotype) were compared with nonSP Sca1+ cells in terms of their engraftment potential. The Sca1+ subclones SP3 and SP16 were transplanted into 5 mice each with cryo-lesions and into 3 mice each without lesions (control group). Hearts were dissected 14 days after transplantation and the degree of engraftment was assessed. As a control, cell culture medium was transplanted into 3 mice with cryo-lesion. Engrafted cells could be easily detected by red-orange fluorescence due to prior incubation with CM-Dil dye or transfection with an mOrange expressing lentivirus. After dissection we could demonstrate robust engraftment of the transplanted cells into the cryoinjured hearts at 14 days after transplantation, while in medium controls no cells could be detected. Multiple cells did stably engraft into the border zone as well as the infarcted area. There was better engraftment in cryo-lesioned hearts than in non-infarcted hearts while there was no obvious difference between engraftment of Sca1+ cells compared to SP-Sca1+ cells. These experiments have proved that engraftment of clonal Sca-1+, SP-cells is feasible and reliable.
Cardiac infarction results in permanent heart dysfunction due to insufficient myogenesis. This myogenic failure has been attributed to the absence of multipotent cardiovascular progenitor cells (CPC) in the heart, or alternatively to a context-related failure of extant CPC to adopt a cardiac fate within the infarct. We could show that CPC partially underlie myogenic infarct repair in neonatal mice, but not in adult mice subjected to the identical lesion. Infarction of postnatal day 1-3 mouse hearts induced the localized expansion of (c-kit)EGFP+ cells within the infarct, expression of early cardiac transcription factors, formation of striated cardiac myocytes, and prominent cardiac regeneration. Myogenic and vascular fates were also adopted by (c-kit)EGFP+ cells in explanted neonatal heart tissue, confirming the cardiac source of regenerative cells. Cre-recombinase mediated fate mapping and in vivo BrdU labeling experiments indicated that induction of undifferentiated precursors underlies localized myogenesis. By contrast, in the infarcted adult heart (c-kit)EGFP+ cells are induced in vivo and in explanted tissue, but adopt vascular fates; bone marrow reconstitution confirmed the cardiac source of vascular precursors. Thus adult post-infarct myogenic failure is not due to a context-dependent restriction of precursor differentiation, but rather to the limited potential of induced adult precursors Manuscript in preparation.
Characterisation of Immune Response after myocardial infarction to possibly optimise cell transplantation
In relation to myocardial injury such as infarction, there is an inflammatory response. Moreover can the immune response be directed towards extra-cellular pathogens which is TH2 and TH3 directed and stimulates the B-lymphocyte response/differentiation and thus the serological response to antigens. The defence against intra-cellular pathogens is however more directed towards cell death and thus more tissue destructive. NK cells play a pivotal role in this defence. The post inflammatory response is important to understand to optimise cell transplant survival. To evaluate how tissue destructive an infarction is in relation to its size we hypothesized that the more tissue aggressive an infarction (and its inflammatory component is) the larger component of the ischemic volume will be infarcted if only the infarction is handled appropriately fast enough.
To establish a method to screen for ischemic volume threat we compared the golden standard technique of single-photon emission computed tomography (SPECT) with Technetium isotope with a previously identified oedema sequence T2STIR. We screened > 700 acute ST-elevation infarction patients to get a population revascularised within six hours and with no previous infarction who, at catherisation still are occluded and only in one vessel. These inclusion criteria were strict to get well-defined infarctions to study the T2-STIR sequence performance.
This study showed a high compliance between the two techniques T2-STIR and SPECT.
By using these two techniques together we could establish a continuous variable of the hostility of the environment by comparing the ischemic volume to the truly infarcted volume. The proportion of the ischemic volume that in the end didn't infarct we denominated the salvage index and thus the more hostile infarction environment the lesser salvage index [50].
We this technique established we continued this task by running a study whereby we sampled STEMI patients for inflammatory cells to phenotype them with regard to their inflammatory cell response and in special to their tissue aggressive response and regulatory T-cell pattern and to relate these to the salvage index. These studies have been completed and are now being analysed and the manuscript is in preparation. Interesting enough the two inflammatory cell patterns that fall out significantly to the salvage index are the NK-cells (CD56+ CD3-) as well as CD28- lymphocytes.
Both these cell types have been identified as part of the tissue aggressive response. Other T-cell patterns examined so far (TH1, TH2 haven't shown such significant and strong correlations with the salvage index). These results are currently compiled in a manuscript and open up for new targets to optimize cell transplantation survival in myocardial infarction patients [50, 51]. This work has been published with regard to the MR technique needed to be developed for these studies [50, 52-54].
In the field of rheumatology new biological treatments have been a major break- through to the field. These are tailored to address specific inflammatory mechanisms. In the search to find treatment options to optimize infarction environment for cell grafting the effect on specific inflammatory cell patterns. In the first line of studies the TNF? antagonist Adalimimab was used in patients with previously not treated with biological anti-inflammatory medications [55]. It has the advantage of being administrated sc like insulin. The treatment was clinically effective as estimated by the clinical rheumatoid arthritis score DAS28. In this study it showed to decrease TH1 cells in peripheral blood just like the TH17 cells both being reported to be active in the tissue aggressive inflammatory response. However not identified as correlated to salvage index in the studies above. These studies have been completed and are being prepared for publication. However, future efforts will include other biological anti-inflammatory agents to hopefully identify compounds that address the inflammatory patterns identified above as unfavourable for infarction environment.
In the CARDIOCELL research portfolio human hESC derived cardiomyocytes were predicted to be tested in immuno deficient mice. At the time of writing the proposal this made sense. With our current experience these do not survive in the mouse, most likely due to the basal heart rate in the mouse to be 450-500 beats per minute (bpm) in contrast to the human 60-70. Another group reported 2012 a completely new model, the guinea-pig as suitable for this [56]. Thus it wasn't time-wise possible to set-up a completely new animal model within this consortium.
Prosthetic Myocardium: A Tissue Engineering Approach
An essential element in tissue engineering is the development of innovative types of scaffolds, which have appropriate physical and biological properties for the cells that are used to populate it. The scaffold should create an environment or 'niche' that allows the cells to mimic or adopt the function of the organ or tissue that is to be engineered. In this regard we have used molecular modelling and the generation of synthetic matrices, generated from the self-assembly of collagen and fibronectin mimetic peptides to develop an environment for adipose-derived stem cells to function as myocytes or valve cells. Using scaffolds derived from collagen and fibronectin mimetic peptides will eliminate concerns over the potential transmission of infectious agents or an immune response associated with animal derived materials. Extracellular matrix proteins are essential for cell adhesion, organisation and signalling and scaffolds derived from these proteins have applications for tissue engineering heart valves and myocardial tissue.
Summary of Achievements
1.Modelling of molecular interactions to predict tertiary structure of synthetic collagen mimetic peptides.
2.Provision of a blue-print for synthesis of mimetic peptides that will form a triple-helical structure.
3.Experimentally designed a suitable peptide sequence, which is able to replicate same kinds of intermediates and hydrogel formation as with natural collagen.
4.The honey-comb morphological structure of the resultant artificial hydrogel is suitable for cell-encapsulation studies.
5.The process of hydrogel formation involves the use of fully biocompatible biomaterials only which is suitable for preparation of three dimensionally cell-encapsulated hydrogels.
6.The peptide is easily scalable to any amount. Already we have designed a prototype of manual equipment which is suitable for the production of CMP-1 peptide in multigram scale quantities.
Molecular Modelling
Collagens constitute one of the main components of most tissues in the body. These supramomolecular structures perform essential functions which include maintaining the shape, physical characteristics and structural integrity of the tissue. Another vital function of the collagens is acting as a docking site for several growth factors and cytokines which regulate important functions of cells, such as differentiation, survival, polarity and motility. Collagens may therefore, play an essential role in the formulation of many biomaterials that are suitable for tissue engineering purposes.
The different sources of collagen for this purpose are from xenogenic, allogenic or recombinant materials, which have several disadvantages. These include risk of infections, evoking an immune response and lack of flexibility. These limitations have stimulated the exploration and development of collagen biomimetic peptides, from the bottom-up. Computational and experimental studies suggest that collagens show more preference for ion pairs than globular proteins, as evidenced by the frequent occurrence of Lys/Gly/Asp/Glu in the sequence of natural collagen. These studies suggested that introduction of ion pairs in short collagen-like peptides could increase the stability of the triple-helix.
Identifying such residue positions is much more complicated for a large multimeric collagen structure that contains two tropocollagen molecules. However, it has been recently shown that residues, which play a crucial role in the structural rigidity of a small protein can be efficiently identified using a graph theoretical approach implemented in the software Floppy Inclusions and Rigid Substructure Topography V6.2 (FIRST). We have used this approach in order to identify the structural hotspots, which are the positions and residues in the chain that are important for maintaining the structural integrity of multiple tropocollagen molecules by quantifying their contribution to the overall rigidity of the system. This analysis allows us to identify positions of the residues for further mutational modeling to cross-link the two tropocollagens as well to introduce a collagen cell binding (CB) motif, GFOGER. Furthermore, the stability and inter-helical cross-links of the constructed mutational models have been validated by the Framework Rigidity Optimized Dynamics Algorithm (FRODAN) and molecular dynamics (MD) simulations.
We have employed a systematic procedure from the bottom-up to encourage the triple-helix triple-helix association for enhancing the self-assembly and biofunctionalisation of collagen biomimetic peptides. A novelty of our approach comes from using a methodology based on constraints that allows us to explore how local atomistic changes affect global properties that may enhance self-assembly at a different scale. Collagen-like systems are especially suited for such a methodology since collagen exhibits such multiscale behavior as a result of the multiple molecular interactions that lead to constraint dynamics at different scales.
The mutational models we designed show how we could potentially drive building blocks of collagen-like peptides to self assembly by increasing the possibilities of their molecular level interactions. Our studies provide guidelines for the engineering of effective tropocollagen, with and without the common CB motif. Starting from atomistic descriptions at the smallest scale in terms of their constraint flexibility, allows inferring information about global structural integrity. The cross-links between ion pairs are further valuable for self-assembly while the formation of additional hydrogen bonds between inter-helical terminals are essential. They could favor the organization of building blocks by stitching their edges strongly using hydrogen bonds and avoiding positional overlaps of their edges, thereby assisting a proper spatial arrangement in the process of assembly. This hypothetical phenomenon attempts to mimic one of the characteristics of natural collagen (tapered tips at termini) that was previously reported. However, increasing the possibilities for cross-link is an encouraging sign to design peptides for many applications such as matrix biology.
A simple molecular based biomimetic collagen model has been cautiously fine-tuned to obtain a refined model with favourable molecular interactions that promote their self-association. As a result, the present study provides mutational models that utilize only non-covalent interactions of charged and key residues to establish association between tropocollagens. The model with the CB motif unveils an anonymous contribution of GFOGER to self association of collagen molecules. With these charge decorations, this model offers extra spots for other cell surface receptors to interact and share signals in and out. Additionally, the computational methods used here offer an efficient and effective tool to map the suitable positions for mutations of multimeric proteins. The presented models here indicate the importance of requirement of a multidimensional approach in collagen research. Since the mutation-dependent helical stability in collagen structure is directly related to genetic disorders, biomaterials/nanomaterials based on these stable mutational models with natural potency to self-associate may have significant implications in biomedicine.
Synthesis of Collagen Mimetic Peptides
The primary structure of collagen molecule contains about 1050 amino acids. Mimicking all properties of collagen using CMPs is very difficult owing to its very small length between 25-40 amino acids. Several CMPs are designed, synthesised and characterised with the following properties to increase the interaction between the triple helices in order to form hydrogels:
1. CMPs with free amino and carboxyl terminus. Polymerisation using standard SPPS method will increase the peptide length.
2. Hydrogen bonding interaction between amino and carboxyl terminus.
3. High temperature annealing
4. Metal-ion co-ordination
5. Chemical cross-linking between triple helices
6. Attaching small ß-sheet forming peptides with CMPs
7. Addition of very small amount of collagen as nucleating agent for hydrogel formation
8. Composites between Hyaluronic acid and CMPs
9. Chemical cross-linking between elastin mimetic peptides with CMPs
10. Enzymatic cross-linking between CMPs
We have designed, synthesised and characterised several peptides with all above mentioned characteristics and tested its triple helix formation. Of all above mentioned strategies, we found that the enzymatic cross-linking method is suitable for the preparation of hydrogels from CMPs. Enzymatic cross-linking of the following peptide results in the formation of hydrogels under physiological temperature.
CMP-1: (Gly-(Lys)2-(Gly-Pro-Hyp)4-Gly-Phe-Hyp-Gly-Glu-Arg-(Gly-Pro-Hyp)4-Gly-(Gln)2-Gly).
The purity of CMP-1 was analysed by HPLC and found to be more than 95% as synthesised and calculated molecular weight of the peptide is 3499.7 da. Experimentally observed molecular weight by Maldi analysis is 3499 da which is in very close agreement with predicted molecular weight.
The circular dichroism spectrum of the peptide CMP-1 in TRIS buffer shows positive peak at 225nm which shows the formation of triple helical structure and the intensity at 225nm is measured at different temperature which shows sigmoidal shape of melting. The melting temperature of the peptide is found to be 47°C F.
The melting transition of the designed CMP-1 obtained from CD measurements is further confirmed by differential scanning calorimetry (DSC).
In natural collagen the D-periodic structure (D=67nm) is formed from the intermolecular interactions that occur between telopeptide regions on both carboxyl and amino termini. The D-periodicity observed for the CMP-KQ assembly is approximately 7.6nm consistent with the small length of the peptide as compared with natural collagen.
The dimensions of the nanofibrous assembly are further confirmed by tapping mode AFM. TEM and AFM examination are in agreement with each other and confirms the formation of nanofibrous assembly upon incubation at 37°C in TRIS buffer.
Cryo-SEM images of the CMP-1 hydrogel was used to provide information about the structure of this material in its native state as drying of the hydrogel may result in aggregation. The micrograph also shows a honeycomb structure that is similar to the morphology of fibrous collagen. This honeycomb structure may be a very suitable material for tissue engineering applications such as in the 3-D encapsulation of cells.
Summary
We have achieved the preparation of artificial collagen hydrogel from enzymatic cross-linking of collagen mimetic peptides. The types of intermediate structures and processes adopted by the designed CMP in higher order formation are similar to the process adopted by natural collagen. The initially formed nanofibrous assembly is cross-linked enzymatically to form hydrogel that has a structure, which may make it a suitable candidate for the preparation of 3-D cell encapsulated scaffolds. The formation of a honeycomb structure indicates that the resultant hydrogel has room to accommodate cells, which are further able to interact with the hydrogel through the integrin-binding site in its peptide sequence.
Potential Impact:
4.2 Use and dissemination of foreground
4.2.1 Scientific Communication
-Results of the work performed within the consortium was communicated through traditional routes with publications in peer reviewed journals
-presentations at European and International Scientific meetings
-direct promotion of new research results to the research groups to whom this might be valuable.
-review publications; members of the consortium are key contributors to their respective fields and do in that respect write overview reviews in their respective fields which are one of the routes of scientific communications that offer the highest impact to the scientific field.
4.2.2 Medical Professionals
-Progress of the consortium has been reported to the cardiological and thoracic surgery societies of Europe with targeted articles for professional journals aimed at physicians and users including presentations at clinical conferences in Europe.
4.2.3 Public awareness
Public awareness to CARDIOCELL's ongoing activities will mainly be addressed by the following strategies:
-Articles for the press and other media
-Broadcasting media
-Presentations at science fairs
-Web page
The webpage contains a public section which clearly explains the purpose of the consortium and which is continuously is updated with the progress and the perceived impact on the society and of the anticipated results from the work of the consortium. This section will also contain contact information whereby the public can interact with the consortium.
The societal implications of the work of CARDIOCELL are represented mainly in the fact that alternative sources to bone marrow cells are available for generative approaches towards the heart. Through Internet; own page, EuroStemcell as well as the best-known cardiologist forum theheart.org interviews as well as reports and articles have been published to the public. One of the main findings, opened a completely new field to the cardiac regeneration; the endogenous regeneration of cardiomyocytes in the adult human heart. Despite to the general view of the field, that the human actually has an inherent regenerative capacity opens up for new therapeutic strategies that don't involve transplantation of cells to the heart. This report was cited and reported by laymen media; New York Times, Daily Mail, Reuters, Scientific American and National Geographic. By these findings Europe has taken the lead towards ne regenerative approaches involving the search for the source for this inherent mechanism and how this can be induced. The direct societal implication for the individual patient is however less for this very day. However the objectives of CARDIOCELL, from the very first draft to final technical annex excluded direct clinical use of the findings, rather we wished to enable the field to focus on new reliable sources which were fulfilled by the identification of fetal/pluripotent sources as well as to open up for endogenous regeneration as a therapy.
Project website: http://www.cardiocell.org