New Vessel Formation in the Context of Cardiomyocyte Regeneration – the Role and Importance of an Adequate Perfusing Vasculature Katherine C

SCR-00429; No. of pages: 17; 4C: 3, 9, 10, 11, 12 Stem Cell Research (2014) xx, xxx–xxx

Available online at www.sciencedirect.com ScienceDirect

www.elsevier.com/locate/scr

New vessel formation in the context of cardiomyocyte regeneration – the role and importance of an adequate perfusing vasculature Katherine C. Michelis a, Manfred Boehm b, Jason C. Kovacic a,⁎ a The Zena and Michael A. Wiener Cardiovascular Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA b Center for Molecular Medicine, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, USA

Received 28 January 2014; received in revised form 16 March 2014; accepted 18 April 2014

Abstract The history of revascularization for cardiac ischemia dates back to the early 1960’s when the first coronary artery bypass graft procedures were performed in humans. With this 50 year history of providing a new vasculature to ischemic and hibernating myocardium, a profound depth of experience has been amassed in clinical cardiovascular medicine as to what does, and does not work in the context of cardiac revascularization, alleviating ischemia and adequacy of myocardial perfusion. These issues are of central relevance to contemporary cell-based cardiac regenerative approaches. While the cardiovascular cell therapy field is surging forward on many exciting fronts, several well accepted clinical axioms related to the cardiac arterial supply appear to be almost overlooked by some of our current basic conceptual and experimental cell therapy paradigms. We present here information drawn from five decades of the clinical revascularization experience, review relevant new data on vascular formation via cell therapy, and put forward the case that for optimal cell-based cardiac regeneration due attention must be paid to providing an adequate vascular supply. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Introduction the fore for the purpose of cardiovascular regeneration. Examples would include angiogenic approaches that aim to Cardiovascular regenerative medicine can be arbitrarily augment new vessel formation and which draw upon the divided into a number of sub-disciplines or fundamentally fields of vascular medicine and hematology (e.g. endothelial different approaches based on the underlying principle that progenitor cell-based therapies); myogenic approaches that is being utilized to regenerate cardiovascular tissue. While aim to form new cardiac muscle (e.g. myoblast therapy or there is considerable overlap, each sub-discipline has direct fibroblast reprogramming to a cardiomyocyte fate via evolved somewhat independently and often calls upon defined factors); or more process-driven approaches such as a specific additional field of research that is brought to bone marrow cell therapy, targeted epicardial stimulation (e.g. by thymosin β4) or attempts to harness/recapitulate dedifferentiation of somatic cells toward a stem cell state. ⁎ Corresponding author at: Mount Sinai Hospital, One Gustave L. Among these approaches, the great majority are geared Levy Place, Box 1030, New York, NY 10029. Fax: +1 212 534 2845. toward producing new cardiomyocytes, with angiogenic E-mail address: [email protected] (J.C. Kovacic). approaches being notably less common. Indeed, in most http://dx.doi.org/10.1016/j.scr.2014.04.009 1873-5061/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 2 K.C. Michelis et al. situations where cell-based regenerative approaches have dynamic process that will likely lead to worsening cardiac attempted to generate cardiomyocytes, the need for an ischemia over time. Moreover, any therapy which targets adequate perfusing vasculature is often neglected or, at regeneration of (injured) myocardium without in some way best, an afterthought. This may be an important oversight in repairing the coronary vasculature is incomplete, in that it our thinking, as the most prevalent cause of congestive does not address the primary problem. heart failure (CHF) is myocardial ischemia due to occlusive atherosclerotic coronary artery disease (CAD). Indeed, it would appear somewhat naive to presume that any cell- Hybernating myocardium based regenerative therapy will perform optimally in the setting of occluded epicardial coronary arteries, or that a The condition of hibernating myocardium is especially therapy that is directed toward generating new cardio- illustrative of what we believe would be likely to occur if myocytes will serendipitously also produce adequate new cardiac tissue were successfully regenerated in the absence vessels. While investigators have been aware of these of an adequate blood supply. Hibernating myocardium is a concerns, given the nascent nature of the field there was well-defined clinical condition that arises when persistently previously little attention paid to these aspects. With the impaired myocardial blood flow and ischemia result in recent impressive strides that have been made in cardiovas- dysfunctional (but viable) cardiac tissue with reduced cular regenerative medicine (many of which are reviewed in cellular metabolic activity and impaired left ventricular this issue of Stem Cell Research) we believe it is timely and contractility (Rahimtoola, 1985, 1989; Fallavollita et al., appropriate to now also focus on the “vascular realities” of 2003). A similar but less severe condition is “stunned achieving heart regeneration and rejuvenation in contem- myocardium,” where perfusion returns to normal at rest, porary patients. although there is persistent myocardial dysfunction. Espe- cially in the case of hibernating myocardium, restoring Heart failure, hibernating myocardium and the normal blood flow is critical since cardiomyocytes can regain function after revascularization (Vanoverschelde et al., importance of an adequate perfusing vasculature 2000; Kalra and Zoghbi, 2002). However, cardiomyocytes cannot survive indefinitely in this “hibernating” state, and Heart failure myocyte apoptosis is an ongoing process that appears to begin as soon as hypoperfusion occurs (Chen et al., 1997). Readers will be aware that one of the major focuses of Hibernating myocardium is characterized by a number cardiovascular cell-therapy to date has been to improve of adaptations at the cellular and molecular level, which cardiac contractility in patients with reduced left ventricu- include alterations in cell morphology, apoptosis, and lar ejection fraction and CHF. Importantly, the most protein expression. Microscopic analysis of hibernating common cause of CHF and left ventricular systolic dysfunc- myocardium in both human and animal models reveals tion is occlusive atherosclerotic disease of the epicardial dedifferentiated cells with loss of myofilaments and accu- coronary arteries (Hunt et al., 2009; Lloyd-Jones et al., mulation of glycogen (Fig. 1)(Maes et al., 1994; Bito et al., 2002). By pooling data from 24 multi-center trials studying 2007). In a swine model, apoptosis of myocytes was seen to CHF in over 43,000 patients, it has been estimated that CAD occur in combination with cellular dedifferentiation, leading is responsible for CHF in over 60% of patients (Gheorghiade to hypertrophy of the remaining functional cardiomyocytes et al., 2006). Furthermore, CHF patients with CAD have a (Lim et al., 1999). Based on large animal models and human worse prognosis than patients with non-ischemic cardiomy- studies, hibernating myocardium is understood to activate opathy, with the extent of CAD inversely related to survival a cell survival program that entails upregulation of genes (Bart et al., 1997; Felker et al., 2002). In addition, there is and proteins involved in anti-apoptosis, growth, and cyto- substantial evidence that atherosclerotic lesions progress protection (Depre et al., 2004), such as inhibitor of apoptosis over time in a predictable fashion. Both retrospective and (IAP), vascular endothelial growth factor (VEGF), H11 prospective studies have shown that initially mild plaques on kinase, heat shock protein 70 (HSP70), hypoxia inducible angiography can become severe over 1–3 years and lead to factor-1α (HIF-1α), and glucose transporter 1 (GLUT1) acute coronary events including death from cardiac causes, (Fig. 2). Mitochondria also adopt a protective phenotype, cardiac arrest, myocardial infarction (MI), or rehospitaliza- with depressed oxygen utilization mediated by downregulation due to unstable or progressive angina (Stone et al., tion of key proteins involved in the electron transport 2011; Ambrose et al., 1988; Glaser et al., 2005). chain (Page et al., 2008; Kelly et al., 2011). Hibernating Patients with established CHF also are at high risk for cardiomyocytes also have reduced expression of Ca2+- suffering adverse outcomes due to further ischemic insults. handling proteins, such as sarcoplasmic reticulum Ca2+- For example, Uretsky et al found that in CHF patients with ATPase (SERCA) and phospholamban, decreased levels of CAD who die suddenly, over 55% have evidence of an acute cAMP (3′-5′cyclic adenosine monophosphate), and increased coronary event at autopsy (Uretsky et al., 2000). Addition- expression of the negatively inotropic cytokines tumor ally, the most common structural abnormalities leading to necrosis factor-α and nitric oxide synthase-2 (Fallavollita sudden cardiac death are features of CAD, including plaque et al., 1999; Kalra et al., 2002; Luss et al., 2002). Myocardial rupture, thrombosis, or infarct (Davies, 1992). biopsies from patients with reversible LV dysfunction We believe it is of importance that any cell-based also demonstrate an increase in α-adrenergic receptor strategy for CHF and impaired left ventricular contraction density and a decrease in β-adrenergic receptor density should be designed with an understanding that CAD is the (Shan et al., 2000). Overall, these changes provide a likely most common underlying cause, and that atherosclerosis is a explanation for the reduced contractility seen in hibernating

Figure 1 Hybernating versus normally perfused cardiomyoctes. (Left) Electron microscopy of normally perfused and functioning cardiomyocytes, with abundant sarcomeres (sm) and mitochondria (m) in the cytoplasm. (Right) Electron microscopy of hibernating cardiomyocytes is notable for disordered structure, with sarcomeres (sm) seen only at the cell periphery. Glycogen (gl) accumulation is present in the cytosol and only small mitochondria are present (arrow). Reproduced with permission from Maes et al. (1994).

β α -adrenergic -adrenergic GLUT1 = upregulated receptor receptor = downregulated

Gs glycolysis gluocse pyruvate ATP cAMP

PKA

ETC TCA

NOS2 contraction machinery L-type Ca2+ channel TNFα Ca2+

transverse tubule

nucleus PKA

cell survival P program PLB HSP SERCA2a VEGF 70 RyR H11K CSQ α sarcoplasmic HIF1 IAP reticulum

Figure 2 Altered signaling pathways in hibernating myocardium. Abbreviations not previously defined: AC, adenylate cyclase; CSQ, calsequestran; Gs, stimulatory G protein; NOS2, nitric oxide synthase-2; PLB, phospholamban; P, phosphate; RyR, ryanodine receptor; TNFα, tumor necrosis factor-α.

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 4 K.C. Michelis et al. myocardium. In an effort to further elucidate the mecha- complexity, greater SYNTAX scores are related to worse nisms underlying decreased contractility, Bito et al simulat- clinical outcomes including MI and death (Kovacic et al., ed excitation-contraction coupling in a porcine model of 2013). In other words, clinical outcomes are related to the hibernating myocardium (Bito et al., 2004, 2005). They complexity and burden of CAD (as classified by the SYNTAX found that hibernating cardiomyocytes have a prolonged score) and the extent of myocardial ischemia. action potential, decreased L-type Ca2+ current, and This tenet is further supported by clinical trials that reduced Ca2+ transient, which is the cytosolic surge in Ca2+ investigated outcomes following “complete” over “incom- that typically triggers myocardial contraction. plete” revascularization. In this situation, CAD patients are In patients with chronic CAD and resultant hibernating classified as having had “complete” revascularization if myocardium, revascularization has been shown to signifi- a blood supply was restored to each of the 3 epicardial cantly improve cardiac performance and clinical outcomes coronary vessels and their major branches. This restoration compared to medical treatment alone (Allman et al., 2002). of blood supply may be achieved either by opening of Specifically, revascularization can promote reverse remod- obstructive epicardial coronary artery lesions by angioplasty eling and halt progression of ventricular scar (Senior et al., with stenting, or by bypassing these lesions with coronary 2001; Pirolo et al., 1986). Importantly and likely related artery bypass graft surgery. Germane to cell therapy to ongoing cardiomyocyte apoptosis (Chen et al., 1997), approaches, it does not include forming capillaries, small growing evidence suggests that early rather than delayed arterioles or collateral vessels, which are typically only revascularization is associated with a further survival sufficient to support a hibernating state in the adult human, benefit (Ling et al., 2013; Bax et al., 2003; Beanlands et but which fail to provide normal arterial perfusion. On the al., 1998). The benefit of revascularization is also notable in other hand, “incomplete” revascularization is deemed to patients with severe CAD. In a subset of patients with CHF have occurred if major vessels or their branches remain and 3-vessel disease in the Coronary Artery Surgery Study non-revascularized after bypass surgery or angioplasty with (CASS), the rate of sudden cardiac death was over 3 times stenting. In a meta-analysis of nearly 90,000 patients, higher in the medically treated group versus patients who complete revascularization was associated with reduced were surgically revascularized (Holmes et al., 1986). long-term mortality, MI and repeat coronary revasculariza- Collectively, this clinical data should guide translational tion (Fig. 3)(Garcia et al., 2013). This mortality benefit was research, as it indicates that any innovative therapy for also consistent regardless of the method of revascularization cardiomyocyte (re)generation must be coupled with ade- (bypass or stenting). Similarly, a recent randomized con- quate arterial vascularization for optimal results. This trolled trial showed that in patients with acute MI under- applies not only to attempts at cardiomyocyte regeneration going infarct-artery percutaneous coronary intervention, in patients with CAD, but also to patients with non-ischemic concurrent additional intervention on non-infarct coronary cardiac disease. It is a simple extrapolation of the above arteries that were significantly diseased lead to a reduction experience with hibernating myocardium to appreciate that in death from cardiac causes, nonfatal MI, and refractory the creation of new myocardial cells that do not have an angina (Wald et al., 2013). Therefore, on the basis of these adequate vascular supply will be an ultimately futile clinical data indicating superiority of complete revascularization, it exercise. As a further corollary with respect to myocardial is possible to make the argument that novel therapeutic hibernation, it is important to note that in many of the modalities should be designed with the intention of regene- potential patient groups that are being considered for rating or supplanting all occluded or severely stenosed cell-based regenerative therapies, and especially those coronary vessels, whether or not they are responsible for with CAD and obstructive arterial lesions, the provision of myocardial impairment. In any case, it is clear that the adequate arterial vascularization alone would likely deliver provision of “complete” and adequate myocardial perfusion benefit related to ventricular remodeling or scar formation, has been demonstrated as critically important across a broad irrespective of any new cardiomyocytes. spectrum of patients and differing clinical scenarios. We suggest that any cell-based therapy will be likely to even- tually fail if it ignores these well documented clinical facts. Burden of ischemia and complete revascularization

Another clinical scenario that is important to consider in the Endothelial cells may orchestrate tissue repair context of cell-based regenerative therapies is that of the overall burden of myocardial ischemia and “complete” While on its own the above clinical experience regarding the versus “incomplete” revascularization. At present, in con- importance of large conduit coronary vessels is compelling, a sidering the clinical treatment of a patient with significant growing body of evidence has indicated that local organ- CAD, clinicians employ what is called the SYNTAX score specific vessels may play additional ‘angiocrine’ roles which to guide their plan for revascularization (i.e. bypass graft are necessary for optimal tissue and organ regeneration. The surgery or angioplasty with stenting) (Kovacic et al., 2013). origins of this concept that ‘angiocrine’ signaling regulates The SYNTAX score is calculated for each patient based on the tissue and organ regeneration lie in developmental studies, location, complexity, and characteristics of their coronary which over several decades have shown extensive cross- atherosclerotic lesions (Sianos et al., 2005). As highlighted communication between various embryonic cell popula- by Head et al, one of the principal aspects of the SYNTAX tions that is required for correct organ specification. More score is that it estimates the extent of ischemic myocardium recently, studies emerged that specifically indicated the (Head et al., 2014). Our group has shown, as have many angiocrine role played by endothelial cells during deve- others, that with an increasing number of lesions and their lopment, with endothelial-endoderm interactions being of

Figure 3 Mortality following complete versus incomplete revascularization. Pooled analysis of clinical studies involving 89,883 patients, investigating mortality following coronary revascularization for those deemed to have had complete revascularization (CR) versus incomplete revascularization (IR) by either coronary artery bypass graft surgery or percutaneous coronary intervention (angioplasty and stenting). Presented here is the risk ratio and 95% confidence interval (CI) for occurrence of total mortality. Boxes are the relative risk estimates from each study; the horizontal bars are 95% CIs. The size of the box is proportional to the weight of the study in the pooled analysis. ACUITY = Acute Catheterization and Urgent Intervention Triage Strategy; ARTS = Arterial Revascularization Therapies Study; BARI = Bypass Angioplasty Revascularization Investigation; CABG = coronary artery bypass graft; CABRI = Coronary Angioplasty Versus Bypass Revascularization Investigation; MASS II = Second Medicine, Angioplasty or Surgery Study; NHLBI = National Heart, Lung, and Blood Institute; RR = risk ratio(s); SYNTAX = Synergy Between PCI With Taxus and Cardiac Surgery. Reproduced from Garcia et al., 2013, with permission from Elsevier.

importance for the induction of endocrine pancreatic contention that cardiac function is ultimately dictated by differentiation (Lammert et al., 2001) and early hepatic the relative balance between cardiac and coronary vascu- morphogenesis (Matsumoto et al., 2001). This latter finding lature growth, rather than cardiac growth or regeneration was subsequently confirmed by Takebe et al. (2013), who alone. demonstrated using induced pluripotent stem cells that Building on these studies, the Rafii laboratory has immature endodermal cells destined to track to a hepatic published several studies which confirm the importance of cell fate could self-organize into three-dimensional liver vascular-tissue paracrine cross-talk. First in the liver, this buds by recapitulating endothelial-mesenchymal develop- group showed that vascular endothelial growth factor mental interactions. receptor (VEGFR)-2-expressing hepatic sinusoidal endo- In the heart, Shiojima et al. (2005) utilized an elegant thelial cells provide angiocrine cues that initiate and sustain transgenic murine model in which cardiac-specific Akt1 liver regeneration induced by partial hepatectomy (Ding et expression promoted physiological hypertrophy in the al., 2010). Key downstream signaling factors from hepatic short-term, but pathological hypertrophy and cardiomyop- endothelial cells included Wnt2 and hepatocyte growth athy with longer term transgene induction. Supporting the factor (HGF), with ablation of this endothelial-derived concept of vascular-cardiac paracrine cross-talk, coronary angiocrine signaling attenuating liver regeneration. Subse- angiogenesis was enhanced in the initial phase of physio- quently, the same team showed that divergent angiocrine logic cardiac hypertrophy but impaired in the chronic signals from hepatic sinusoidal endothelial cells, regulated phase, which was directly related to the ensuing contrac- by the balance between CXCR7 and CXCR4 signaling, tile dysfunction. While potential angiocrine mediators stimulate either regeneration after immediate injury or emanating from the vasculature were not identified, it fibrosis after chronic insult (Ding et al., 2014). Moving was shown that Akt1 signaling induced mammalian target of to other organs, Raffi and colleagues also demonstrated a rapamycin (mTOR)-dependent production of VEGF and similar phenomenon in the lung, with resection of one Angiopoietin (Ang)-2 by cardiomyocytes, which was respon- lung triggering activation of VEGFR-2 and fibroblast growth sible for coronary angiogenesis during initial physiologic factor (FGF) receptor 1 in pulmonary capillary endothelial cardiac hypertrophy (Shiojima et al., 2005). Central to cells, leading to the production of angiocrine growth factors the theme of this review, these data support our broader including matrix metalloproteinase (MMP) 14. In turn, this

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 6 K.C. Michelis et al. endothelial-derived angiocrine signaling led to epithelial strain has inferior recovery from hindlimb ischemia with 54% progenitor cell proliferation and alveologenesis (Ding et al., less collateral remodeling, increased ischemia and reduced 2011). Again, ablation of this endothelial-derived angiocrine angiogenesis compared to C57BL/6 (Chalothorn et al., signaling resulted in attenuated lung regeneration. 2007). Studies have since demonstrated that these differ- While at the present time many unanswered questions ences are not limited to the hindlimb, but likely involve most remain about endothelial and vascular angiocrine signaling tissue beds in the body (Chalothorn et al., 2007). Related to and the importance of this program in the heart, these this, there are significant inter-species differences in the data represent additional and provocative support to our local tissue or skeletal muscle expression of a range of argument for due attention to the vasculature in cardiac angiogenic and other factors, including VEGF, VEGFR-2 regenerative strategies. (Chalothorn et al., 2007), Ang-2, Tie1, chloride intracellular channel 4 and other genes/proteins. This may partially account for baseline differences in collaterals and inter- Insights from inter-strain differences in the arterial connections, but the differing expression of these murine response to ischemia factors may also favorably regulate any local angiogenic response that arises after an ischemic insult in C57BL/6 mice Much attention has recently been paid to the realization that (Chalothorn et al., 2007, 2009; Chalothorn and Faber, 2010; among differing mouse strains, an identical vascular injury Chen et al., 2010; Wang et al., 2010; McClung et al., 2012). or stimulus elicits a strain-specific response that varies The net result of these strain-specific differences is that significantly in the extent of ischemia, new vessel formation while C57BL/6 mice achieve almost full relative perfusion and vascular remodeling. These differences are prominent recovery in the thigh at 14 days after femoral artery between C57BL/6 and BALB/c mice, with C57BL/6 mice ligation, BALB/c mice experience severe hindlimb ischemia noted to develop greater neointimal hyperplasia (i.e. which may culminate in limb loss. Even at 28 days after occlusive vascular remodeling – a detrimental outcome), femoral ligation, only about a 40% restoration of hindlimb but also improved recovery after ischemic insults with perfusion toward baseline levels is observed in BALB/c mice increased collateral vasculature (de Vries et al., 2013; (Helisch et al., 2006). Helisch et al., 2006). While intrinsic differences in resident These findings offer major insights with respect to or circulating vascular progenitor cells have not been achieving new vessel formation in the context of cardio- thoroughly explored, much of this appears attributable myocyte and cardiac regeneration. Perhaps foremost is the to the pre-existing extent of interarterial connections and fact that the recruitment, remodeling and arterialization collateral vessels, and strain-specific immune functionality. (including smooth muscle cell investment) of pre-existing With respect to immune function, in broad terms it capillaries and microvascular connections (arteriogenesis) appears that C57BL/6 mice have a more activated immune is the predominant mechanism whereby perfusion is either system that is biased toward a pro-inflammatory response partially (BALB/c) or fully (C57BL/6) restored following (de Vries et al., 2013; Schulte et al., 2008). For example, the injury (Helisch et al., 2006; Chalothorn et al., 2007, 2009; biology of Natural Killer (NK) cells differs significantly Chalothorn and Faber, 2010; Wang et al., 2010; Mac between mouse strains. This is at least partly due to Gabhann and Peirce, 2010). The corollary of this is that the differences in the NK gene Complex (NKC) that encodes de novo formation of new vessels, sometimes referred to as multiple NK cell-related receptors including the polymorphic adult vasculogenesis, either does not occur or plays a minor Ly49 receptor family. Compared to C57BL/6 mice, BALB/c role in the typical injury response (Kovacic et al., 2008a). mice lack a 200 kb region encoding numerous members of While it may be possible that with specific stimulation or the Ly49 receptor family. As a result, while C57BL/6 mice therapies adult vasculogenesis may arise (Kocher et al., possess 15 Ly49 genes, BALB/c mice have only 8 Ly49 genes 2001; Smart et al., 2010), it is important to carefully (Proteau et al., 2004; Higuchi et al., 2010). In turn, in a consider that this process does not appear to be a major series of NK cell depletion experiments it was shown that NK aspect of the normal adult healing response. A second cells in C57BL/6 mice promote vascular remodeling via important insight is that in addition to recruiting and Interferon (IFN)-γ leading to increased intimal hyperplasia remodeling pre-existing arterial interconnections, a range compared to BALB/c mice (de Vries et al., 2013). While of diverse but powerful factors contribute to improving the numerous other differences in immune function have been vascular supply and vascular healing, including immune noted among various mouse strains, other aspects that may function and the expression of numerous angiogenic ligands be of relevance to vascular formation and remodeling and their receptors. It is critical that the complexity of include a bias toward Th1-type immune response in C57BL/ these vascular programs is not underestimated because, as 6 mice, with BALB/c tending to exhibit a Th2-type response also demonstrated by the failure of numerous angiogenic (Schulte et al., 2008). gene therapy studies over the past 2 decades (Henry et al., The extent of pre-existing microvascular connections and 2007; Gupta et al., 2009), the simple administration of a collateral vessels is also of importance for determining the single gene or factor is unlikely to promote clinically response following an ischemic insult. Helisch et al. (2006) meaningful vascular regeneration. A third important point were among the first to show that hindlimb interarterial arising from these studies of C57BL/6 versus BALB/c is that connections of C57BL/6 mice, compared to BALB/c, are thesamevascularregenerativeapproachmaynotwork better able to be recruited, and help to re-establish in all individuals. Indeed, whereas the mice described perfusion following femoral artery occlusion. Other investi- above differed only in their genetic makeup, the clinical gators have shown that there are 35% fewer collateral landscape of patients requiring cardiovascular regenera- vessels in the thigh of BALB/c mice, and correspondingly this tive therapies contains dramatic differences in genetics,

Human arteriogensis and angiogenesis As readers of this article will be aware, a plethora of cell-based cardiovascular regenerative studies have been As in the mouse, the limited human data available strongly published over the last 15 years. Regardless of the specific support the paradigm that remodeling and adaptation of cell therapy under evaluation or the experimental model, pre-existing vascular interconnections (arteriogenesis) is the many investigators have taken steps to evaluate the primary mechanism responsible for providing a new vascular potential vascularization effects of their specific therapy. supply in human cardiovascular disease states. Seminal However, this has often been by fairly crude measures such studies from the 1960’s were the first to demonstrate that as counting the number of capillaries or endothelial cells. coronary collateral arterioles existing in the normal human Few studies have put significant efforts into adequately heart can be recruited to grow into larger structures to defining the physiologic adequacy of any new vessel for- prevent myocardial necrosis (Fulton, 1964). This work is mation. While we cannot review each study that has supported by contemporary studies (Zbinden et al., 2004, examined vessel formation, we present here what we 2007), including the finding that after a period of exercise perceive to be some of the learning points and messages training the heart is able to recruit sufficient collateral from these endeavors, which might inform our objective of vessels to avoid short term ischemia with target vessel achieving functional vascularization in the context of occlusion (Zbinden et al., 2007). In addition to arterio- cardiac and cardiomyocyte regeneration. genesis, there is a wealth of data to support the occurrence of angiogenesis in the adult, that is, the formation of new capillaries from preexisting vessels by the sprouting or Early cell-therapy studies – bone marrow-derived splitting of preexisting vessels by transcapillary pillars or and circulating progenitor cells posts of extracellular matrix and/or cells (Kovacic et al., 2008a). A myriad of proteins and signaling pathways In adult humans, it is an interesting historical note that the are involved in angiogenesis and arteriogenesis, including first reported description of a putative progenitor endothe- HIF-1α, HGF, VEGF, platelet-derived growth factor (PDGF), lial cell in 1997 (Asahara et al., 1997) predated what is FGF, monocyte chemoattractant protein (MCP-1), nitric considered the first rigorous description of a resident adult oxide and many more. While we do not propose to review cardiac stem cell population by 6 years (Beltrami et al., this entire field, interested readers are referred to the 2003). As a result, early efforts at cell-based regenerative many excellent articles on this topic (Kovacic et al., 2008a; therapies focused on harnessing progenitor endothelial and Cochain et al., 2013). In addition to angiogenesis and other ‘angiogenic’ cells to increase vascularity. In what was arteriogenesis, other mechanisms of new vessel formation a landmark paper at the time, in 2001 Itescu and co-workers in the adult have been proposed such as ‘adult vasculo- studied the intravenous administration of human CD34+ cells gensis’ (de novo formation of new vessels from progenitor into athymic nude rats following ligation of the left anterior cells) (Tepper et al., 2005), However, as discussed a coronary artery (Kocher et al., 2001). It was reported that significant role for this process remains to be convincingly this therapy led to de novo new blood vessel formation in the shown in the normal healing response (Ziegelhoeffer et al., infarct bed (i.e. adult vasculogenesis) and proliferation of 2004; Simons, 2005). preexisting vasculature (angiogenesis), which contributed to Therefore, what emerges in the adult human is a picture the salvage of viable myocardium (Kocher et al., 2001). similar to that in the mouse, where the additional demand Remarkably, in the same year, it was also described that for increased blood supply and/or oxygenation is met by bone marrow-derived progenitor cells could not only give the adaptation, remodeling and growth of preexisting rise to new vessels, but also cardiomyocytes (Orlic et al., vascular structures. The available data indicate that de 2001). These and other studies triggered a wave of novo appearance of new vessels where none existed enthusiasm for the use of bone marrow derived and/or previously, from either progenitor or other cells, is not a circulating cell-based therapies aiming to augment vascu- commonly executed adult biologic program (if it occurs larity (and even generate cardiomyocytes). However, in the at all). We believe that this fact must be respected in ensuing years several critical lessons have been learned that cardiovascular regenerative medicine strategies, as it have somewhat dampened the initial enthusiasm for bone suggests that therapeutic de novo vessel formation will marrow-derived or circulating angiogenic cells. Foremost require a very sophisticated strategy, rather than merely has been that the therapeutic translation of this approach upregulating normal biologic programs. However, while the to humans has not been as spectacularly successful as above processes of new vessel formation in the adult human these early animal studies promised. Many groups, are consistent with observations in mice, we are yet to ourselves among them (Kovacic et al., 2008b; Chih et al., reconcile the fact that clinical revascularization studies 2012), embarked on studies using bone marrow-derived cells, have shown that these smaller sized vessels created during circulating progenitor endothelial cells or ‘hemangioblasts’ the normal repair process are inadequate to avoid the aiming to improve vascularity. While animal studies using

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 8 K.C. Michelis et al. these cells showed almost overwhelming efficacy, pivotal MI. The question of durable cell engraftment versus clinical studies have often failed to deliver positive results. paracrine effects was not rigorously studied, but in Notably, three consecutive randomized studies conducted addition to various beneficial effects on left ventricular under the banner of the US National Institutes of Health ejection fraction and other cardiac indices, the investiga- (NIH)–sponsored Cardiovascular Cell Therapy Research tors observed a N 50% increase in capillary density with Network (CCTRN) recently reported negative findings mesenchymal cell therapy. Perhaps more impressive, using various bone marrow-derived cell populations in arteriolar density in the infarct zone was doubled by cardiovascular disease patients (Traverse et al., 2011, active cell therapy. 2012; Perin et al., 2012). While clinical meta-analyses Despite these recent promising basic and translational have suggested small to modest effects (Delewi et al., studies using bone marrow-derived cells and ongoing 2013; Jeevanantham et al., 2012) and further phase III clinical studies, much research remains to be done to human studies are ongoing (Sheridan, 2013), meaningful achieve meaningful vascular formation in patients using improvements in clinical outcomes and even surrogate these cells. To this point, van der Laan et al. (2011) measures remain to be convincingly demonstrated. Al- recently published their findings in patients that received though a possible exception may be bone-marrow derived either an intracoronary infusion of bone marrow mono- mesenchymal stem cell therapy (Heldmanetal.,2014;Hare nuclear cells (n = 23), peripheral blood mononuclear cells et al., 2012), potential beneficial effects appear predom- (n = 18) or sham therapy (control; n = 19) following inantly due to paracrine signaling from the administered reperfused ST-segment elevation MI. Evaluating micro- cell population, which often fails to engraft and survive circulatory flow in the infarct-related artery measured beyond 1–2 weeks (Fang et al., 2011; Deuse et al., 2009). invasively in the catheterization laboratory by intra- While paracrine signaling might help hypoxic or metaboli- coronary Doppler flow wire at 4 months after MI, there cally challenged cardiomyocytes survive an acute insult, it was no difference between control patients and those in is difficult to conceptually envision how short-term para- either of the active treatment groups. As the authors state, crine effects could help improve vascularization in the this data does not support the hypothesis of enhanced long-term, especially when the clinical experience with neovascularization after this mode of cell therapy. hibernating myocardium and revascularization suggests As a further cautionary note regarding these cells, that large, epicardial-like vessels are required for adequate while paracrine and other effects may augment new vessel myocardial perfusion. formation, a series of well conducted studies have now convincingly demonstrated that bone marrow-derived cells do not make a direct contribution to the vascular endothe- Contemporary studies of bone marrow-derived cells lium (Hagensen et al., 2010; Ohle et al., 2012; Purhonen et al., 2008). These studies have investigated bone marrow cell The above comments notwithstanding, it is encouraging that contributions to diverse biologic models (atherosclerosis, several recent studies using bone marrow-derived cells have tumor growth, hyperoxia-induced pulmonary injury, aging) shown promise, at least for the generation of small vessels. using rigorous lineage tracking systems, and have dispelled While, as we have reviewed, this is suboptimal from a earlier notions of a bone marrow-derived cell contribution to clinical perspective and large, epicardial-like vessels should endothelial populations (Hagensen et al., 2010; Ohle et al., remain the long-term goal of this field, these studies offer 2012; Purhonen et al., 2008). hope for the future development of a more robust cell-therapy derived vasculature. Among these contemporary studies, Duran et al. (2013) The existence of a bipotent hemangioblast? demonstrated that a novel c-kit+Sca-1+ “cortical bone- derived stem cell” population improved survival and cardiac The putative but controversial (Kovacic and Boehm, 2009) function when injected into mice following MI. Salubrious existence of a bipotent hemangioblast has been part of the effects were greater than those seen with c-kit+Sca-1+ rationale for much of the abovementioned bone marrow cardiac-derived stem cell (CSC) injection. Cortical bone- cell-based research. As a somewhat unexpected finding from derived stem cells engrafted and survived until at least direct reprogramming and other related studies spawned 6 weeks after injection, and differentiated into mature from the discovery of induced pluripotent stem cells, we cardiac tissue, including endothelial cells, smooth muscle have recently moved somewhat closer to unravelling the cells and small-sized (arteriolar) vessels (Fig. 4). Cardiac- complex cellular biology that surrounds the origins of early derived stem cells did not show these effects. Though hematopoietic and endothelial cells. Specifically, we appear somewhat surprising that control CSCs did not show an effect to be slowly understanding that a truly bipotent hemangio- (see section below on CSCs), this nonetheless remains an blast is perhaps less likely to exist, or at least is less common encouraging finding. While yet to be harnessed to generate in the adult, than previously thought. Recent studies based larger coronary vessels, we expect c-kit+Sca-1+ cortical on cellular reprogramming and related lines of investigation bone-derived stem cells will receive heightened attention have independently shown that during progressive com- in the coming years. mitment toward a hematopoietic cell fate, cells pass Serving as another example of incremental translation- through a transitory period when they have an endothelial- al gains being made with bone marrow-derived cells, like molecular signature. In our studies of directly re- Houtgraaf et al. (2013) conducted a fairly large-scale programming murine fibroblasts to hematopoietic stem study of mesenchymal precursor cell administration via cells, fibroblasts initially lose their mesenchymal signature, the intracoronary route in a large animal (ovine) model of then adopt a transitory endothelial-like gene expression

Figure 4 Vascularization by cortical bone marrow-derived stem cells. Six weeks after MI and injection of enhanced green fluorescent protein (EGFP)-marked cortical bone-derived c-kit+Sca-1+ stem cells, mice were harvested and evaluated for engraftment and differentiation into cardiac and vascular cells. Shown here is staining for α-smooth muscle actin (αSMA; white), von Willebrand Factor (vWF; red) and EGFP (green). Nuclei were labeled with 4’, 6-diamidino-2-phenylindole (DAPI; blue). Scale bars = 10 μm. Reproduced with permission from Duran et al. (2013). pattern, which is then progressively lost as the cells adopt a hematopoietic transplant techniques to the generate new hematopoietic stem cell fate (Fig. 5)(Pereira et al., 2013). vessels. Gene expression patterns were not supportive of the existence of a bipotent hemangioblast population. This finding has been corroborated by other groups. For example, Can resident adult cardiac stem cells form physio- Rafii et al. (Rafii et al., 2013) demonstrated that the logically significant new vessels? hematopoietic derivatives of human embryonic stem cell populations originate from endothelial cells via endothelial Over a decade ago, the first formal characterization of to hematopoietic transition (a specialized type of epithelial resident adult CSCs described that, in addition to cardio- to mesenchymal transition). Further supporting evidence myocytes, c-kit+ CSCs are capable of differentiating into has also been provided by in vivo experiments where it has endothelial and smooth muscle cells, and forming vascular been shown that ‘hemogenic endothelial cells’ (rather than structures (Beltrami et al., 2003). In an important paper bipotent hemangioblasts) undergo endothelial to hemato- titled “Formation of large coronary arteries by cardiac poietic transition to give rise to primitive hematopoietic progenitor cells”, the same investigators later demonstrated cells (Zhen et al., 2013). This shift in our understanding of that rat c-kit+ CSCs could be activated ex vivo with the developmental origins of endothelial and hematopoietic insulin-like growth factor-1 (IGF-1) and HGF, and then cells is another of the many factors that need to be injected into the MI border zone in a rodent model considered in the design of future cell therapy studies (Tillmanns et al., 2008). These functionally enhanced CSCs aiming to improve vascularity. Moreover, the fact that adult engrafted within the host myocardium, proliferated and hemangioblasts may be a rare biologic entity (if they exist differentiated into endothelial cells, smooth muscle cells at all) may help to explain the inconsistent results that and cardiomyocytes. Vascular lineage differentiation was have arisen from many of the clinical studies conducted to mediated by HIF-1α upregulation, which promoted stromal- date which, at their core, were founded on adaptations of derived factor-1 (SDF-1) secretion from hypoxic coronary

2009). These two provocative and back-to-back papers from the Anversa group are the first we are aware of that provide evidence to suggest that a cell-based therapy could potentially regenerate medium or even large caliber vessels that have the potential to significantly augment coronary flow. The latter of these two studies was also specific in addressing the clinically relevant situation of hibernating myocardium. Since this time others have also reported on the likely existence of c-kit+ vascular progenitor cells that reside in the adult heart. Indeed, it has been suggested that in the adult c-kit+ CSCs do not support post-MI cardio- myogenesis, but rather, are only able to form vascular structures (Jesty et al., 2012). However, based on post-MI large animal studies (Bolli et al., 2013) it would seem that administration of c-kit + CSCs alone may be insufficient for large vessel formation, and that achieving this objective will require other measures such as biologic stimulation (e.g. ex vivo CSC stimulation using cytokines prior to administration) (Tillmanns et al., 2008) or further cell purification (e.g. into c-kit+VEGFR-2+ vascular progenitors) (Bearzi et al., 2009). Nevertheless, we believe that these data regarding c-kit+ CSCs serve as important proof of principle for the possibility Figure 5 Schematic illustration of reprogramming to hema- of generating large vessels within the heart, and point the topoietic stem cells. Reproduced from Pereira et al. (2013),with way for future studies. Furthermore, there is additional permission from Elsevier. Abbreviations not previously defined: cause for excitement in the arena of resident vascular HSC, hematopoietic stem cell; hu, human; LT-HSC, long-term- progenitor cells within the heart, with the Chien laboratory hematopoietic stem cell. recently describing a population of endothelial progenitor cells located in the outflow tract of fetal hearts, identified by expression of Isl1 and VE-Cadherin (Lui et al., 2013). vessels, with SDF-1 signaling being a key event in CSC These cells also express VEGF-A, VEGFR-1 and VEGFR-2, and commitment to a vascular fate. As one of the few papers we are highly responsive to VEGF stimulation. While their fate are aware of to demonstrate the formation of more and role in the adult remain to be defined, the identification substantive vessels, these activated CSCs formed conductive of these and other resident stem/progenitor cells within the and intermediate-sized coronary arteries together with heart with vessel forming capacity (Chong et al., 2011) capillaries and smaller vessels (Fig. 6)(Tillmanns et al., opens the door to further regenerative possibilities for 2008). These vessels were connected with the host coronary robust cardiac vascular formation. circulation, and this increase in vascularization more than doubled blood flow in the MI zone. This significant vascularization effect involving formation of medium-sized Epicardial-derived cells and vessel formation vessels, together with CSC-mediated myocardial regeneration, attenuated postinfarction remodeling, reduced infarct There is certain intuitive appeal in investigating the size and improved cardiac function (Tillmanns et al., 2008). possibility that epicardial-derived cells might harbor vessel As somewhat of a follow-up study to the above, forming capacity, because in the adult the large coronary 18 months later the same group published a second paper arteries (left anterior descending, circumflex, right coro- demonstrating the formation of larger sized vessels (Bearzi nary) are epicardial structures. In 2007 Smart et al. (2007) et al., 2009). Rather than relying on ex vivo cytokine published a major paper suggesting that epicardial-derived stimulation of c-kit+ CSCs, on this occasion the approach precursor cells could be stimulated to induce coronary was to isolate a population of human heart-derived vascular vasculogenesis and angiogenesis by treatment with thymosin progenitor cells, defined by cells co-positive for c-kit and β4, an actin monomer binding protein. Mutant mice with VEGFR-2. These c-kit+VEGFR-2+ cells were found to occupy attenuated thymosin β4 signaling demonstrated impaired vascular niches in the walls of coronary vessels, and in vitro smooth muscle differentiation and migration with defective were self-renewing, clonogenic and able to differentiate coronary vascular formation, while thymosin β4 treatment predominantly into endothelial and smooth muscle cells, but was able to rejuvenate adult epicardial explants and trigger also cardiomyocytes. In a canine myocardial ischemia model outgrowth of smooth muscle and endothelial cells (Smart et that recapitulated the situation of hibernating myocardium, al., 2007). This group then went on to show that in control injected c-kit+VEGFR-2+ cells were shown to regenerate mice subjected to MI there was an endogenous endothelial large, intermediate, and small human coronary arteries that response with formation of a capillary network, how- were connected to the host circulation. This regenerative ever, mice treated with thymosin β4 at the time of MI process included differentiation of the injected cells showed more mature vessel formation that included smooth into endothelial and smooth muscle cells (Fig. 7), and a muscle cell recruitment (Smart et al., 2010). These new corresponding increase in coronary blood flow with func- coronary vessels in thymosin β4-treated mice occupied an tional improvement of ischemic myocardium (Bearzi et al., expanded subepicardial space and the immediate underlying

Figure 6 Large vessel formation by activated c-kit+ CSCs. Large vessel formation by c-kit+ CSCs activated ex vivo with IGF-1 and HGF, and injected into the MI border zone in rats. (A) Epimyocardium of an infarcted heart at 2 weeks after cell administration: three new coronary arteries (Upper, EGFP, green) are present in the spared myocardium (SM) and border zone (BZ). The arrow points to a branching vessel. Colocalization of EGFP and αSMA is shown (Lower, orange). Preexisting coronary branches are EGFP-negative (Upper) and αSMA-positive (Lower, red, asterisks). A cluster of EGFP positive cells is present at the injection site (IS). Myocytes, αSA: white. (B) Large regenerated arteries (Upper, EGFP, green) in spared myocardium at 2 weeks. Colocalization of EGFP and αSMA is shown (Lower, orange). Inset illustrates endothelial cells (vWF, white). Reproduced with permission from Tillmanns et al. (2008), Copyright (2008) National Academy of Sciences, U.S.A.

Figure 7 Large vessel formation by activated cardiac c-kit+VEGFR-2+ cells. Generation of large coronary arteries by the injection of EGFP-labeled c-kit+VEGFR-2+ cells into canine hearts with coronary stenosis. Arteries 0.9 and1.3 mm in diameter were detected. All smooth muscle cells are positive for αSMA (red), EGFP (green), αSMA/EGFP (yellowish), and Alu probe (white; a probe marking human-derived cells). Preexisting coronary vessels (arrows) are αSMA positive and EGFP and Alu probe negative. Reproduced with permission from Bearzi et al. (2009). is yet to be performed of the size and quality of new vessel cellular grafts. Using a rodent MI model with skeletal myoblasts formation that arises from VEGF-A modRNA therapy, it is administered directly into the infarct zone at the time of notable that this therapy improved heart function and coronary artery ligation, Weyers et al found that skeletal enhanced long-term survival (Zangi et al., 2013). myoblast grafts developed an extensive vascular network Investigation of the potential of epicardial-directed (Weyers et al., 2013). Micro computed tomographic imaging of therapies for cardiac regeneration and vessel formation is vessels larger than 25 μm diameter revealed that vascular in its infancy. However, these studies appear to hint at the structures extended throughout the graft regions, including potential for a major, albeit latent vessel forming capacity large and small vessels, and both arteries and veins. Quantifi- of epicardial cells and/or epicardial-directed therapies. As cation of the extent of branching of the left coronary artery previously mentioned this approach has intuitive appeal as (which was the ligated vessel) revealed that myoblast therapy this is the location of the large arteries of the heart. We induced a significant increase in the number of major secondary anticipate much focus on the epicardium and its capacity for and tertiary arterial branches. Important limitations of this vascular formation going forwards. work were that neither the molecular mechanisms of vessel formation nor the cellular origins (host versus donor) were explored, and a great deal remains to be done in this line of Skeletal myoblast therapy research. However, despite the fact that this therapy has been almost abandoned in clinical studies, it appears reasonable to Perhaps somewhat surprisingly, investigators are continuing to use skeletal myoblast therapy as a model of cellular engraft- pursue the potential beneficial effects of skeletal myoblast ment to understand the processes facilitating the formation and therapy, yielding remarkably informative results. Importantly, inter-connection of large perfusing coronary vessels. injected skeletal myoblasts do not couple to or therefore contract in synchrony with native cardiomyocytes (Leobon et al., 2003), and a phase II randomized placebo-controlled clinical trial failed to show efficacy (Menasche et al., 2008). Neverthe- Cardiac development, embryonic and induced less, the rationale for this study was that skeletal myoblasts pluripotent stem cells likely have superior survival compared to other cell populations (as described above) in a post-MI model, and that they therefore The discovery of induced pluripotent stem cells by Shinya may be used to explore new vessel formation in the context of Yamanaka was a milestone in cellular biology that had a

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 The cardiac stem cell compartment 13 profound impact on the direction of cardiovascular stem cell Perspectives on current reprogramming and research. Prior to this, investigators working with embryonic differentiation approaches stem cells were already engaged in the task of driving pluripotent cells toward differentiated cardiovascular cell fates. As reviewed in this issue of Stem Cell Research, these As readers will be aware, enormous efforts are now being differentiation programs are intimately related to embryon- made in directing pluripotent and other cells toward cardio- ic cardiovascular developmental pathways. vascular cell fates. Recently, this research has begun to focus on the prospects for direct in situ differentiation of resident fibroblasts toward differentiated cell fates, and cardiomyocytes in particular (Song et al., 2012; Qian et al., Developmental origins of the coronary vasculature 2012). While incredibly promising, a critical element lacking in this approach of direct differentiation to cardiomyocytes The developmental origins of the coronary circulation have is that of vascularization. By definition, the application been extensively reviewed elsewhere (Kovacic et al., 2012) of specific reprogramming factors selected to trigger the and are beyond the scope of this article. However, there are reprogramming of fibroblasts to a cardiomyocyte fate, while several key principles of cardiac vascular development that are potentially effective at creating new cardiomyocytes and of relevance to our review, particularly in the context of cell myocardium, ignores the need for vascular formation and reprograming, induced pluripotent and embryonic stem cells. perfusion. Even though newly formed cardiomyocytes may Foremost is the fact that bi- or multi-potent progenitor cells be able to stimulate modest capillary ingrowth, as we have are strongly implicated in cardiac vascular formation. shown here by numerous examples this is almost certain to Multipotent cardiovascular progenitor cells are specified be inadequate for proper cardiomyocyte functioning and from early mesoderm and several studies have suggested instead, would seem destined to create a situation of that bi- or multipotent progenitor cells are responsible for hibernating myocardium involving the newly reprogrammed both cardiomyocyte and cardiac vascular development. cardiomyocytes. The same concerns can be raised for the For example, Kattman et al. (2006) suggested that a multi- application of embryonic stem cell-derived cardiomyocytes, potent VEGFR-2-expressing cell, identified in head-fold or any therapy that primarily aims to replenish cardio- stage embryos, gives rise to cardiomyocytes, endothelial myocytes without attention to vascular formation. As a case and smooth muscle cells. On the other hand, Wu et al. (2006) in point, when human embryonic stem cell-derived cardio- suggested that Nkx2.5-expressing bipotent cells may give myocytes were very recently delivered into the infarcted rise to cardiomyocytes and smooth muscle cells, while heart of non-human primates, it has been presented that the Moretti et al. (2006) provided data to indicate that a grafts were perfused by a “sluggish vascular plexus” and that multipotent islet-1-expressing cell gives rise to cardio- “developing a brisk arterial supply is a substantial remaining myocytes, endothelial and smooth muscle cells derived hurdle.” (Christoffels and Pu, 2013) While progress has also from the second heart field. Later in development, while been made in reprogramming toward vascular cell fates there may be species-specific differences that remain to be (Pereira et al., 2013; Xiong et al., 2013; Ginsberg et al., fully reconciled (Kovacic et al., 2012), it is likely that that 2012; Margariti et al., 2012), much work remains to be done the coronary vessels derive additional contributions from before this approach can be expected to deliver robust epicardial derived progenitor cells which, depending on vascular formation. the specific species, appear potent for the formation While these early efforts remain encouraging, looking of endothelial cells, vascular smooth muscle cells, and ahead we foresee that the developmental paradigm of pericytes (see review) (Kovacic et al., 2012). Alternatively, multipotent stem cells being responsible for embryonic in mice at least, cardiac endothelial cells may derive from cardiac formation will increasingly influence this field. the sinus venosus (Red-Horse et al., 2010). Although it is appropriate that initial efforts focus on While many more studies have addressed cardiac vascular differentiation or reprogramming to only a single cell fate origins, the above summary indicates that, according to (e.g. cardiomyocyte, endothelial), as we contend through- Nature’s blueprint, cardiac vascular formation likely arises out this article we believe that robust cardiac repair will in parallel with cardiomyocyte specification with a signifi- require broader regeneration of both cardiomyocytes and a cant contribution from common multipotent progenitor perfusing vasculature. It is plausible that the embryonic and cells. Indeed, this is underscored by the fact that we are induced pluripotent stem cell fields will be critical in this unaware of any genetically manipulated mice with signifi- regard, as they hold the potential to permit differentiation cant cardiac vascular defects that do not also manifest or reprogramming to multiple concurrent cardiovascular cell perturbed gross heart development. We believe that the fates. Despite the considerable technical obstacles that dependency of cardiac vascular formation on multipotent need to be overcome, we envision several possible ways cardiac progenitor cells is an important developmental in which this field may evolve. For example, a potential program that must be respected in efforts with contempo- area for research would be to develop differentiation and rary cell-based regenerative therapies. For example and as reprogramming protocols to generate multipotent cardio- we explore further below, although there may be appeal in vascular progenitor cells with restricted differentiation restricting delivered cell therapies to a differentiated abilities toward only cardiomyocytes, smooth muscle and cardiomyocyte fate to minimize the risk of teratoma and endothelial cells. This would limit tumor forming potential, tumor formation, it must be understood that differentiated while providing the cellular substrate for comprehensive cardiomyocytes lack the primary capacity to generate cardiac repair. Another possible line of investigation vessels. would be to develop in situ reprogramming protocols that

Please cite this article as: Michelis, K.C., et al., New vessel formation in the context of cardiomyocyte regeneration – theroleandimportanceof an adequate perfusing vasculature, Stem Cell Res. (2014), http://dx.doi.org/10.1016/j.scr.2014.04.009 14 K.C. Michelis et al. redirected fibroblasts toward all 3 differentiated cardio- Banerjee, I., Zhang, J., Moore-Morris, T., et al., 2012. Thymosin vascular cell types (cardiomyocytes, smooth muscle and Beta 4 is dispensable for murine cardiac development and endothelial cells). Certainly, these and other exciting function. Circ. Res. 110, 456–464. prospects would appear cause for cautious optimism in this Banerjee, I., Morris, T.M., Evans, S.M., Chen, J., 2013. Thymosin beta4 is not required for embryonic viability or vascular still nascent area of investigation. development. Circ. Res. 112, e25–e28. Bart, B.A., Shaw, L.K., McCants Jr., C.B., et al., 1997. Clinical Conclusions determinants of mortality in patients with angiographically diagnosed ischemic or nonischemic cardiomyopathy. J. Am. Coll. Cardiol. 30, 1002–1008. At least five decades of experience presently informs clinical Bax, J.J., Schinkel, A.F., Boersma, E., et al., 2003. Early versus practice for CAD and cardiac revascularization strategies. delayed revascularization in patients with ischemic cardiomyop- While there is much that can be learned from this and athy and substantial viability: impact on outcome. Circulation translated to our pursuit of cell-based regenerative ap- 108 (Suppl. 1), II39–II42. proaches, perhaps the most obvious message is that a Beanlands, R.S., Hendry, P.J., Masters, R.G., deKemp, R.A., suboptimal vascular supply leads to hibernating or ischemic Woodend, K., Ruddy, T.D., 1998. Delay in revascularization is myocardium and poor clinical outcomes. At the present associated with increased mortality rate in patients with severe time, although our understanding of the salient cellular and left ventricular dysfunction and viable myocardium on fluorine molecular processes is growing rapidly, the requirement of 18-fluorodeoxyglucose positron emission tomography imaging. Circulation 98, II51–II56. providing an adequate and interconnected vasculature in Bearzi, C., Leri, A., Lo Monaco, F., et al., 2009. Identification of a the context of cardiac regenerative therapy remains a coronary vascular progenitor cell in the human heart. Proc. Natl. significant obstacle yet to be overcome. Nevertheless, as Acad. Sci. U. S. A. 106, 15885–15890. we have reviewed, data is beginning to accumulate that is Beltrami, A.P., Barlucchi, L., Torella, D., et al., 2003. Adult cardiac shaping a platform of knowledge to propel this field forward. stem cells are multipotent and support myocardial regeneration. While rudimentary, a degree of early success has been seen Cell 114, 763–776. for the generation of capillaries and at least small vessels Bito, V., Heinzel, F.R., Weidemann, F., et al., 2004. Cellular using various cell-based approaches. Other emerging fields mechanisms of contractile dysfunction in hibernating myocardi- – not reviewed here, such as tissue engineering and the um. Circ. Res. 94, 794 801. possibility of engineered vascular grafts, may also come into Bito, V., Heinzel, F.R., Claus, P., et al., 2005. Alterations in excitation-contraction coupling in chronically ischemic or hiber- play in the future. Therefore, while we may be only beginning nating myocardium. Exp. Clin. Cardiol. 10, 142–145. to appreciate the importance of an adequate vasculature for Bito, V., van der Velden, J., Claus, P., et al., 2007. Reduced force cardiac regeneration, at least some of the tools required to generating capacity in myocytes from chronically ischemic, work toward this objective may already be within our grasp. hibernating myocardium. Circ. Res. 100, 229–237. While readers of this Stem Cell Research review series will be Bolli, R., Tang, X.L., Sanganalmath, S.K., et al., 2013. Intracoronary tantalized by many new and stimulating research ideas, delivery of autologous cardiac stem cells improves cardiac we offer the simple but critical axiom that due attention function in a porcine model of chronic ischemic cardiomyopathy. to concurrent vascularization will be required if we are Circulation 128, 122–131. to succeed in our collective, long-term goal of achieving Chalothorn, D., Faber, J.E., 2010. Strain-dependent variation in meaningful cardiomyocyte and cardiac regeneration. collateral circulatory function in mouse hindlimb. Physiol. Genomics 42, 469–479. Chalothorn, D., Clayton, J.A., Zhang, H., Pomp, D., Faber, J.E., Acknowledgements 2007. Collateral density, remodeling, and VEGF-A expression differ widely between mouse strains. Physiol. Genomics 30, 179–191. Katherine Michelis is supported by National Institutes of Chalothorn, D., Zhang, H., Smith, J.E., Edwards, J.C., Faber, J.E., Health (NIH) grant T32HL007824. Jason Kovacic acknowl- 2009. Chloride intracellular channel-4 is a determinant of native edges research support from the NIH (K08HL111330), The collateral formation in skeletal muscle and brain. Circ. Res. 105, Leducq Foundation (Transatlantic Network of Excellence 89–98. Award) and AstraZeneca. Fig. 2 was created using Servier Chen, C., Ma, L., Linfert, D.R., et al., 1997. Myocardial cell death medical art. and apoptosis in hibernating myocardium. J. Am. Coll. Cardiol. 30, 1407–1412. Chen, Z., Rubin, J., Tzima, E., 2010. Role of PECAM-1 in References arteriogenesis and specification of preexisting collaterals. Circ. Res. 107, 1355–1363. Allman, K.C., Shaw, L.J., Hachamovitch, R., Udelson, J.E., 2002. Chih, S., Macdonald, P.S., McCrohon, J.A., et al., 2012. Granulocyte- Myocardial viability testing and impact of revascularization on Colony Stimulating Factor in Angina Patients to Stimulate prognosis in patients with coronary artery disease and left Neovascularization (GAIN II) Trial: A randomized placebo- ventricular dysfunction: a meta-analysis. J. Am. Coll. Cardiol. controlled crossover study. Heart 98, 282–290. 39, 1151–1158. Chong, J.J., Chandrakanthan, V., Xaymardan, M., et al., 2011. Adult Ambrose, J.A., Tannenbaum, M.A., Alexopoulos, D., et al., 1988. cardiac-resident MSC-like stem cells with a proepicardial origin. Angiographic progression of coronary artery disease and the Cell Stem Cell 9, 527–540. development of myocardial infarction. J. Am. Coll. Cardiol. 12, Christoffels, V.M., Pu, W.T., 2013. Developing insights into cardiac 56–62. regeneration. Development 140, 3933–3937. Asahara, T., Murohara, T., Sullivan, A., et al., 1997. Isolation of Cochain, C., Channon, K.M., Silvestre, J.S., 2013. Angiogenesis putative progenitor endothelial cells for angiogenesis. Science in the infarcted myocardium. Antioxid. Redox Signal. 18, 275, 964–967. 1100–1113.

Davies, M.J., 1992. Anatomic features in victims of sudden coronary Hare, J.M., Fishman, J.E., Gerstenblith, G., et al., 2012. Compar- death. Coronary artery pathology. Circulation 85, I19–I124. ison of allogeneic vs autologous bone marrow-derived mesen- de Vries, M.R., Seghers, L., van Bergen, J., et al., 2013. C57BL/6 NK chymal stem cells delivered by transendocardial injection in cell gene complex is crucially involved in vascular remodeling. J. patients with ischemic cardiomyopathy: the POSEIDON random- Mol. Cell. Cardiol. 64, 51–58. ized trial. JAMA 308, 2369–2379. Delewi, R., Andriessen, A., Tijssen, J.G., Zijlstra, F., Piek, J.J., Head, S.J., Farooq, V., Serruys, P.W., Kappetein, A.P., 2014. The Hirsch, A., 2013. Impact of intracoronary cell therapy on left SYNTAX score and its clinical implications. Heart 100, 169–177. ventricular function in the setting of acute myocardial infarc- Heldman, A.W., DiFede, D.L., Fishman, J.E., et al., 2014. tion: a meta-analysis of randomised controlled clinical trials. Transendocardial mesenchymal stem cells and mononuclear Heart 99, 225–232. bone marrow cells for ischemic cardiomyopathy: the TAC-HFT Depre, C., Kim, S.J., John, A.S., et al., 2004. Program of cell randomized trial. JAMA 311, 62–73. survival underlying human and experimental hibernating myo- Helisch, A., Wagner, S., Khan, N., et al., 2006. Impact of mouse cardium. Circ. Res. 95, 433–440. strain differences in innate hindlimb collateral vasculature. Deuse, T., Peter, C., Fedak, P.W., et al., 2009. Hepatocyte growth Arterioscler. Thromb. Vasc. Biol. 26, 520–526. factor or vascular endothelial growth factor gene transfer Henry, T.D., Grines, C.L., Watkins, M.W., et al., 2007. Effects of maximizes mesenchymal stem cell-based myocardial salvage Ad5FGF-4 in patients with angina: an analysis of pooled data after acute myocardial infarction. Circulation 120, S247–S254. from the AGENT-3 and AGENT-4 trials. J. Am. Coll. Cardiol. 50, Ding, B.S., Nolan, D.J., Butler, J.M., et al., 2010. Inductive 1038–1046. angiocrine signals from sinusoidal endothelium are required for Higuchi, D.A., Cahan, P., Gao, J., et al., 2010. Structural variation of liver regeneration. Nature 468, 310–315. the mouse natural killer gene complex. Genes Immun. 11, 637–648. Ding, B.S., Nolan, D.J., Guo, P., et al., 2011. Endothelial-derived Holmes Jr., D.R., Davis, K.B., Mock, M.B., et al., 1986. The effect of angiocrine signals induce and sustain regenerative lung medical and surgical treatment on subsequent sudden cardiac alveolarization. Cell 147, 539–553. death in patients with coronary artery disease: a report from the Ding, B.S., Cao, Z., Lis, R., et al., 2014. Divergent angiocrine signals Coronary Artery Surgery Study. Circulation 73, 1254–1263. from vascular niche balance liver regeneration and fibrosis. Houtgraaf, J.H., de Jong, R., Kazemi, K., et al., 2013. Intracoronary Nature 505, 97–102. infusion of allogeneic mesenchymal precursor cells directly after Duran, J.M., Makarewich, C.A., Sharp, T.E., et al., 2013. Bone- experimental acute myocardial infarction reduces infarct size, derived stem cells repair the heart after myocardial infarction abrogates adverse remodeling, and improves cardiac function. through transdifferentiation and paracrine signaling mecha- Circ. Res. 113, 153–166. nisms. Circ. Res. 113, 539–552. Hunt, S.A., Abraham, W.T., Chin, M.H., et al., 2009. 2009 focused Fallavollita, J.A., Jacob, S., Young, R.F., Canty Jr., J.M., 1999. update incorporated into the ACC/AHA 2005 Guidelines for the Regional alterations in SR Ca(2+)-ATPase, phospholamban, and Diagnosis and Management of Heart Failure in Adults: a report of HSP-70 expression in chronic hibernating myocardium. Am. J. the American College of Cardiology Foundation/American Heart Physiol. 277, H1418–H1428. Association Task Force on Practice Guidelines: developed in Fallavollita, J.A., Malm, B.J., Canty Jr., J.M., 2003. Hibernating collaboration with the International Society for Heart and Lung myocardium retains metabolic and contractile reserve despite Transplantation. Circulation 119, e391–e479. regional reductions in flow, function, and oxygen consumption at Jeevanantham, V., Butler, M., Saad, A., Abdel-Latif, A., Zuba- rest. Circ. Res. 92, 48–55. Surma, E.K., Dawn, B., 2012. Adult bone marrow cell therapy Fang, J., Chen, L., Fan, L., et al., 2011. Enhanced therapeutic improves survival and induces long-term improvement in cardiac effects of mesenchymal stem cells on myocardial infarction by parameters: a systematic review and meta-analysis. Circulation ischemic postconditioning through paracrine mechanisms in rats. 126, 551–568. J. Mol. Cell. Cardiol. 51, 839–847. Jesty, S.A., Steffey, M.A., Lee, F.K., et al., 2012. c-kit + precursors Felker, G.M., Shaw, L.K., O'Connor, C.M., 2002. A standardized support postinfarction myogenesis in the neonatal, but not definition of ischemic cardiomyopathy for use in clinical adult, heart. Proc. Natl. Acad. Sci. U. S. A. 109, 13380–13385. research. J. Am. Coll. Cardiol. 39, 210–218. Kalra, D.K., Zoghbi, W.A., 2002. Myocardial hibernation in coronary Fulton, W.F., 1964. The Time Factor in the Enlargement of artery disease. Curr. Atheroscler. Rep. 4, 149–155. Anastomoses in Coronary Artery Disease. Scott. Med. J. 9, 18–23. Kalra, D.K., Zhu, X., Ramchandani, M.K., et al., 2002. Increased Garcia, S., Sandoval, Y., Roukoz, H., et al., 2013. Outcomes after myocardial gene expression of tumor necrosis factor-alpha and complete versus incomplete revascularization of patients with nitric oxide synthase-2: a potential mechanism for depressed multivessel coronary artery disease: a meta-analysis of 89,883 myocardial function in hibernating myocardium in humans. patients enrolled in randomized clinical trials and observational Circulation 105, 1537–1540. studies. J. Am. Coll. Cardiol. 62, 1421–1431. Kattman, S.J., Huber, T.L., Keller, G.M., 2006. Multipotent flk-1+ Gheorghiade, M., Sopko, G., De Luca, L., et al., 2006. Navigating cardiovascular progenitor cells give rise to the cardiomyocyte, the crossroads of coronary artery disease and heart failure. endothelial, and vascular smooth muscle lineages. Dev. Cell 11, Circulation 114, 1202–1213. 723–732. Ginsberg, M., James, D., Ding, B.S., et al., 2012. Efficient direct Kelly, R.F., Cabrera, J.A., Ziemba, E.A., et al., 2011. Continued reprogramming of mature amniotic cells into endothelial depression of maximal oxygen consumption and mitochondrial cells by ETS factors and TGFbeta suppression. Cell 151, proteomic expression despite successful coronary artery bypass 559–575. grafting in a swine model of hibernation. J. Thorac. Cardiovasc. Glaser, R., Selzer, F., Faxon, D.P., et al., 2005. Clinical progression Surg. 141, 261–268. of incidental, asymptomatic lesions discovered during culprit Kocher, A.A., Schuster, M.D., Szabolcs, M.J., et al., 2001. vessel coronary intervention. Circulation 111, 143–149. Neovascularization of ischemic myocardium by human bone- Gupta, R., Tongers, J., Losordo, D.W., 2009. Human studies of marrow-derived angioblasts prevents cardiomyocyte apoptosis, angiogenic gene therapy. Circ. Res. 105, 724–736. reduces remodeling and improves cardiac function. Nat. Med. 7, Hagensen, M.K., Shim, J., Thim, T., Falk, E., Bentzon, J.F., 2010. 430–436. Circulating endothelial progenitor cells do not contribute to Kovacic, J.C., Boehm, M., 2009. Resident vascular progenitor cells: plaque endothelium in murine atherosclerosis. Circulation 121, an emerging role for non-terminally differentiated vessel- 898–905. resident cells in vascular biology. Stem Cell Res. 2, 2–15.

Kovacic, J.C., Moore, J., Herbert, A., Ma, D., Boehm, M., Graham, involve contributions from marrow-derived endothelial precur- R.M., 2008a. Endothelial progenitor cells, angioblasts, and sor cells. Am. J. Respir. Cell Mol. Biol. 47, 11–19. angiogenesis–old terms reconsidered from a current perspec- Orlic, D., Kajstura, J., Chimenti, S., et al., 2001. Bone marrow cells tive. Trends Cardiovasc. Med. 18, 45–51. regenerate infarcted myocardium. Nature 410, 701–705. Kovacic, J.C., Macdonald, P., Feneley, M.P., et al., 2008b. Safety Page, B., Young, R., Iyer, V., et al., 2008. Persistent regional and efficacy of consecutive cycles of granulocyte-colony stimu- downregulation in mitochondrial enzymes and upregulation of lating factor, and an intracoronary CD133+ cell infusion in stress proteins in swine with chronic hibernating myocardium. patients with chronic refractory ischemic heart disease: the G- Circ. Res. 102, 103–112. CSF in angina patients with IHD to stimulate neovascularization Pereira, C.F., Chang, B., Qiu, J., et al., 2013. Induction of a (GAIN I) trial. Am. Heart J. 156, 954–963. hemogenic program in mouse fibroblasts. Cell Stem Cell 13, Kovacic, J.C., Mercader, N., Torres, M., Boehm, M., Fuster, V., 2012. 205–218. Epithelial-to-mesenchymal and endothelial-to-mesenchymal tran- Perin, E.C., Willerson, J.T., Pepine, C.J., et al., 2012. Effect of sition: from cardiovascular development to disease. Circulation transendocardial delivery of autologous bone marrow mononu- 125, 1795–1808. clear cells on functional capacity, left ventricular function, and Kovacic, J.C., Limaye, A.M., Sartori, S., et al., 2013. Comparison of perfusion in chronic heart failure: the FOCUS-CCTRN trial. JAMA six risk scores in patients with triple vessel coronary artery 307, 1717–1726. disease undergoing PCI: Competing factors influence mortality, Pirolo, J.S., Hutchins, G.M., Moore, G.W., 1986. Infarct expansion: myocardial infarction, and target lesion revascularization. pathologic analysis of 204 patients with a single myocardial Catheter. Cardiovasc. Interv. 82, 855–868. infarct. J. Am. Coll. Cardiol. 7, 349–354. Lammert, E., Cleaver, O., Melton, D., 2001. Induction of pancreatic Proteau, M.F., Rousselle, E., Makrigiannis, A.P., 2004. Mapping of differentiation by signals from blood vessels. Science 294, 564–567. the BALB/c Ly49 cluster defines a minimal natural killer cell Leobon, B., Garcin, I., Menasche, P., Vilquin, J.T., Audinat, E., receptor gene repertoire. Genomics 84, 669–677. Charpak, S., 2003. Myoblasts transplanted into rat infarcted Purhonen, S., Palm, J., Rossi, D., et al., 2008. Bone marrow-derived myocardium are functionally isolated from their host. Proc. Natl. circulating endothelial precursors do not contribute to vascular Acad. Sci. U. S. A. 100, 7808–7811. endothelium and are not needed for tumor growth. Proc. Natl. Lim, H., Fallavollita, J.A., Hard, R., Kerr, C.W., Canty Jr., J.M., Acad. Sci. U. S. A. 105, 6620–6625. 1999. Profound apoptosis-mediated regional myocyte loss and Qian, L., Huang, Y., Spencer, C.I., et al., 2012. In vivo compensatory hypertrophy in pigs with hibernating myocardium. reprogramming of murine cardiac fibroblasts into induced Circulation 100, 2380–2386. cardiomyocytes. Nature 485, 593–598. Ling, L.F., Marwick, T.H., Flores, D.R., et al., 2013. Identification of Rafii, S., Kloss, C.C., Butler, J.M., et al., 2013. Human ESC-derived therapeutic benefit from revascularization in patients with left hemogenic endothelial cells undergo distinct waves of endothe- ventricular systolic dysfunction: inducible ischemia versus lial to hematopoietic transition. Blood 121, 770–780. hibernating myocardium. Circ. Cardiovasc. Imaging 6, 363–372. Rahimtoola, S.H., 1985. A perspective on the three large multicen- Lloyd-Jones, D.M., Larson, M.G., Leip, E.P., et al., 2002. Lifetime ter randomized clinical trials of coronary bypass surgery for risk for developing congestive heart failure: the Framingham chronic stable angina. Circulation 72, V123–V135. Heart Study. Circulation 106, 3068–3072. Rahimtoola, S.H., 1989. The hibernating myocardium. Am. Heart J. Lui, K.O., Zangi, L., Silva, E.A., et al., 2013. Driving vascular 117, 211–221. endothelial cell fate of human multipotent Isl1+ heart progen- Red-Horse, K., Ueno, H., Weissman, I.L., Krasnow, M.A., 2010. itors with VEGF modified mRNA. Cell Res. 23, 1172–1186. Coronary arteries form by developmental reprogramming of Luss, H., Schafers, M., Neumann, J., et al., 2002. Biochemical venous cells. Nature 464, 549–553. mechanisms of hibernation and stunning in the human heart. Rossdeutsch, A., Smart, N., Dube, K.N., Turner, M., Riley, P.R., Cardiovasc. Res. 56, 411–421. 2012. Essential role for thymosin beta4 in regulating vascular Mac Gabhann, F., Peirce, S.M., 2010. Collateral capillary smooth muscle cell development and vessel wall stability. Circ. arterialization following arteriolar ligation in murine skeletal Res. 111, e89–e102. muscle. Microcirculation 17, 333–347. Schulte, S., Sukhova, G.K., Libby, P., 2008. Genetically pro- Maes, A., Flameng, W., Nuyts, J., et al., 1994. Histological grammed biases in Th1 and Th2 immune responses modulate alterations in chronically hypoperfused myocardium. Correlation atherogenesis. Am. J. Pathol. 172, 1500–1508. with PET findings. Circulation 90, 735–745. Senior, R., Lahiri, A., Kaul, S., 2001. Effect of revascularization on Margariti, A., Winkler, B., Karamariti, E., et al., 2012. Direct left ventricular remodeling in patients with heart failure from reprogramming of fibroblasts into endothelial cells capable of severe chronic ischemic left ventricular dysfunction. Am. J. angiogenesis and reendothelialization in tissue-engineered ves- Cardiol. 88, 624–629. sels. Proc. Natl. Acad. Sci. U. S. A. 109, 13793–13798. Shan, K., Bick, R.J., Poindexter, B.J., et al., 2000. Altered Matsumoto, K., Yoshitomi, H., Rossant, J., Zaret, K.S., 2001. Liver adrenergic receptor density in myocardial hibernation in organogenesis promoted by endothelial cells prior to vascular humans: A possible mechanism of depressed myocardial func- function. Science 294, 559–563. tion. Circulation 102, 2599–2606. McClung, J.M., McCord, T.J., Keum, S., et al., 2012. Skeletal Sheridan, C., 2013. Cardiac stem cell therapies inch toward clinical muscle-specific genetic determinants contribute to the differ- litmus test. Nat. Biotechnol. 31, 5–6. ential strain-dependent effects of hindlimb ischemia in mice. Shiojima, I., Sato, K., Izumiya, Y., et al., 2005. Disruption of Am. J. Pathol. 180, 2156–2169. coordinated cardiac hypertrophy and angiogenesis contributes to Menasche, P., Alfieri, O., Janssens, S., et al., 2008. The Myoblast the transition to heart failure. J. Clin. Invest. 115, 2108–2118. Autologous Grafting in Ischemic Cardiomyopathy (MAGIC) trial: Sianos, G., Morel, M.A., Kappetein, A.P., et al., 2005. The SYNTAX first randomized placebo-controlled study of myoblast trans- Score: an angiographic tool grading the complexity of coronary plantation. Circulation 117, 1189–1200. artery disease. EuroIntervention 1, 219–227. Moretti, A., Caron, L., Nakano, A., et al., 2006. Multipotent Simons, M., 2005. Angiogenesis: where do we stand now? Circulation embryonic isl1+ progenitor cells lead to cardiac, smooth muscle, 111, 1556–1566. and endothelial cell diversification. Cell 127, 1151–1165. Smart, N., Riley, P.R., 2013. Thymosin beta4 in vascular Ohle, S.J., Anandaiah, A., Fabian, A.J., Fine, A., Kotton, D.N., development response to research commentary. Circ. Res. 2012. Maintenance and repair of the lung endothelium does not 112, e29–e30.

Smart, N., Risebro, C.A., Melville, A.A., et al., 2007. Thymosin Vanoverschelde, J.L., Depre, C., Gerber, B.L., et al., 2000. Time beta4 induces adult epicardial progenitor mobilization and course of functional recovery after coronary artery bypass graft neovascularization. Nature 445, 177–182. surgery in patients with chronic left ventricular ischemic Smart, N., Risebro, C.A., Clark, J.E., et al., 2010. Thymosin beta4 dysfunction. Am. J. Cardiol. 85, 1432–1439. facilitates epicardial neovascularization of the injured adult Wald, D.S., Morris, J.K., Wald, N.J., et al., 2013. Randomized trial heart. Ann. N. Y. Acad. Sci. 1194, 97–104. of preventive angioplasty in myocardial infarction. N. Engl. J. Song, K., Nam, Y.J., Luo, X., et al., 2012. Heart repair by Med. 369, 1115–1123. reprogramming non-myocytes with cardiac transcription factors. Wang, S., Zhang, H., Dai, X., Sealock, R., Faber, J.E., 2010. Genetic Nature 485, 599–604. architecture underlying variation in extent and remodeling of Stone, G.W., Maehara, A., Lansky, A.J., et al., 2011. A prospective the collateral circulation. Circ. Res. 107, 558–568. natural-history study of coronary atherosclerosis. N. Engl. J. Weyers, J.J., Schwartz, S.M., Minami, E., et al., 2013. Effects of cell Med. 364, 226–235. grafting on coronary remodeling after myocardial infarction. J. Takebe, T., Sekine, K., Enomura, M., et al., 2013. Vascularized and Am. Heart Assoc. 2, e000202. functional human liver from an iPSC-derived organ bud trans- Wu, S.M., Fujiwara, Y., Cibulsky, S.M., et al., 2006. Developmental plant. Nature 499, 481–484. origin of a bipotential myocardial and smooth muscle cell Tepper, O.M., Capla, J.M., Galiano, R.D., et al., 2005. Adult precursor in the mammalian heart. Cell 127, 1137–1150. vasculogenesis occurs through in situ recruitment, proliferation, Xiong, Q., Ye, L., Zhang, P., et al., 2013. Functional consequences and tubulization of circulating bone marrow-derived cells. Blood of human induced pluripotent stem cell therapy: myocardial ATP 105, 1068–1077. turnover rate in the in vivo swine heart with postinfarction Tillmanns, J., Rota, M., Hosoda, T., et al., 2008. Formation of large remodeling. Circulation 127, 997–1008. coronary arteries by cardiac progenitor cells. Proc. Natl. Acad. Zangi, L., Lui, K.O., von Gise, A., et al., 2013. Modified mRNA Sci. U. S. A. 105, 1668–1673. directs the fate of heart progenitor cells and induces vascular Traverse, J.H., Henry, T.D., Ellis, S.G., et al., 2011. Effect of regeneration after myocardial infarction. Nat. Biotechnol. 31, intracoronary delivery of autologous bone marrow mononuclear 898–907. cells 2 to 3 weeks following acute myocardial infarction on left Zbinden, R., Zbinden, S., Windecker, S., Meier, B., Seiler, C., 2004. ventricular function: the LateTIME randomized trial. JAMA 306, Direct demonstration of coronary collateral growth by physical 2110–2119. endurance exercise in a healthy marathon runner. Heart 90, Traverse, J.H., Henry, T.D., Pepine, C.J., et al., 2012. Effect of the 1350–1351. use and timing of bone marrow mononuclear cell delivery on left Zbinden, R., Zbinden, S., Meier, P., et al., 2007. Coronary collateral ventricular function after acute myocardial infarction: the TIME flow in response to endurance exercise training. Eur. J. randomized trial. JAMA 308, 2380–2389. Cardiovasc. Prev. Rehabil. 14, 250–257. Uretsky, B.F., Thygesen, K., Armstrong, P.W., et al., 2000. Acute Zhen, F., Lan, Y., Yan, B., Zhang, W., Wen, Z., 2013. Hemogenic coronary findings at autopsy in heart failure patients with endothelium specification and hematopoietic stem cell mainte- sudden death: results from the assessment of treatment with nance employ distinct Scl isoforms. Development 140, lisinopril and survival (ATLAS) trial. Circulation 102, 611–616. 3977–3985. van der Laan, A.M., Hirsch, A., Haeck, J.D., et al., 2011. Recovery Ziegelhoeffer, T., Fernandez, B., Kostin, S., et al., 2004. Bone of microcirculation after intracoronary infusion of bone marrow marrow-derived cells do not incorporate into the adult growing mononuclear cells or peripheral blood mononuclear cells in vasculature. Circ. Res. 94, 230–238. patients treated by primary percutaneous coronary intervention the Doppler substudy of the Hebe trial. JACC Cardiovasc. Interv. 4, 913–920.