Molecular mechanisms of heart regeneration

Abstract

The adult mammalian heart is incapable of clinically relevant regeneration. The regenerative deficit in adult mammalian heart contrasts with the fetal and neonatal heart, which demonstrate substantial regenerative capacity after injury. This deficiency in adult mammals is attributable to the lack of resident stem cells after birth, combined with an inability of pre-existing cardiomyocytes to complete cytokinesis. Studies of neonatal heart regeneration in mammals suggest that latent regenerative potential can be re-activated. Dissecting the cellular and molecular mechanisms that promote cardiomyocyte proliferation is key to stimulating true regeneration in adult humans. Here, we review recent advances in our understanding of cardiomyocyte proliferation that suggest molecular approaches to heart regeneration.

1. Introduction

Cardiovascular diseases and, in particular, heart failure remain dominant causes of death worldwide [1]. Heart failure can arise from many pathological remodeling processes, but the most common form of systolic heart failure primarily results from a substantial loss of differentiated cardiomyocytes from infarction. With insufficient cardiomyocyte proliferation capacity, the mammalian heart responds to infarction with repair as opposed to regeneration, leading to replacement of the injured muscle tissue with a meshwork of extracellular matrix (ECM) that forms a scar [2, 3]. The loss of muscle tissue eventually compromises contractility of the remaining myocardium, sometimes leading to heart failure and death [4]. Many patients can have their myocardial infarction aborted by catheter and other reperfusion treatments, but after the first day of infarction, this does not preserve myocardium [5]. Thus, there is a worldwide unmet clinical need for therapeutic strategies for heart regeneration in the days to years after an infarction.

Prenatal mammals have a remarkable ability to regenerate their heart following injury; however, this regenerative ability lasts for a brief window of time shortly after birth. Studies have demonstrated a transient cellular and molecular response from pre-existing cardiomyocytes that leads to substantial myocardial regeneration in neonatal mammalian models [6]. Over the past decade, investigators in the field have invested considerable effort and resources for new strategies to achieve effective cardiac regeneration. These therapeutic strategies range from engineered heart patches, exogenous delivery of cardiac progenitor cells, and cell reprogramming strategies to small molecule delivery [7-9]. While many cell-based therapies prompted initial excitement, they have met considerable clinical challenges such as lack of an ideal cell type and no evidence of true cardiomyocyte regeneration in the heart [10, 11].

Amongst the myriad of potential therapies for cardiac regeneration, no clear clinical strategy has emerged thus far, underscoring the need for further insights into the fundamental biology of the cardiomyocyte and its interaction with neighboring cells during the regenerative process. Several signaling pathways- Hippo/YAP [12, 13], Hedgehog (Hh) [14], and most recently thyroid hormone signaling [15], have been investigated as potential therapeutic pathways for adult heart regeneration. In this Review, we will explore recently-defined cellular and molecular signaling pathways that mediate cardiomyocyte cell cycle re-entry as well as emerging concepts in mammalian heart regeneration. We highlight the fundamental biology of cardiomyocyte renewal in designing strategies to potentially achieve adult mammalian heart regeneration.

1.1. Heart homeostasis

Prenatally, the mammalian heart grows through cardiomyocyte cell division (hyperplasia) as opposed to hypertrophic growth seen in adult mammalian heart [16]. Shortly after birth, cardiomyocytes undergo additional DNA synthesis without cytokinesis, resulting in many becoming binucleated and tetraploid [16]. Strong lines of evidence from ¹⁴C isotope decay measurements in humans, and complementary stable isotope labeling in combination with fate mapping studies in rodents, demonstrated a modest level of cardiomyocyte renewal in the adult heart. About 1% of cardiomyocytes are replaced per year in adult mammals [17-20], In mammals, the low rate of cardiomyocyte mitosis was traced back to division of pre-existing cardiomyocytes that were also more likely to be mononucleated and diploid [6, 18, 21-23]. Thus, in the absence of injury, adult mammalian heart homeostasis is maintained by the slow turnover of cardiomyocytes at approximately 1% per year, and this rate declines to 0.5% or less with age [17, 18, 20, 24]. The reason cardiomyocyte turnover declines with age in adult mammals is unknown.

1.2. Adult scarring after infarction

The number of cardiomyocytes entering the cell cycle increases after myocardial injury, but cell division (cytokinesis) does not occur sufficiently to allow regeneration in the adult mammalian heart [18]. Instead an injury to the adult heart results in repair by fibrotic scar formation [6]. The repair process leads to deposition of connective tissue (cardiac fibrosis) and only partial restoration of tissue structure [4]. The wound healing process after a myocardial injury relies on the instructive role of an inflammatory microenvironment [25]. The inflammatory reaction and the proliferative response by nondamaged stromal cells that become activated in situ or get recruited from the bloodstream are essential, as dead heart muscle is prone to rupture, and myocardial rupture is frequently instantaneously fatal [25, 26]. Remaining cardiomyocytes develop compensatory hypertrophy, presumably to maintain organ function [16].

1.3. True heart regeneration: Zebrafish, axolotl, newborn mammals.

Lower vertebrates, such as newts and zebrafish, have significantly higher cardiomyocyte renewal rates compared to mammals [23, 27, 28]. They generally maintain > 95 % diploid cardiomyocytes even as adults and exhibit full regeneration of heart muscle by proliferation of pre-existing cardiomyocytes, even after up to 20 % of the heart has been surgically resected [29, 30]. Interestingly, direct experimental evidence for a higher regenerative capacity in neonates of up to ~ 7 days after birth has been shown in mice after injury by ventricular resection [6], infarction [31], cryoinjury [32, 33], or clamping [34]. The apical resection model was reported to increase cardiomyocytes in M-phase and cytokinesis in neonatal mice [6]. Studies examining proliferative cardiomyocytes between species show that the frequency of mononucleated cardiomyocytes of neonatal mice with retained regenerative potential up to ~ 7 days of birth, resemble that of adult lower vertebrates [35] (Figure 1). Furthermore, case reports from corrective human heart surgeries [36] and myocardial infarction in a newborn child [31] show full regeneration of heart muscle. This suggests that there is a window of high regenerative capacity in the neonatal human heart.

Figure 1: Response to cardiac injury in neonatal and adult mammals.

In this figure, we summarize the response to injury in neonatal and adult mammalian hearts. Following injury, neonatal mice have the ability to successfully regenerate their heart by cardiomyocyte proliferation and neovascularization to replace lost tissue and restore cardiac function. On the other hand, adult mammals are not capable of regenerating significant amounts of heart tissue. Instead, injury in the adult mammalian heart leads to scarring and loss of function. Due to the low cell turnover rate, the adult mammalian heart is incapable of replacing the lost cardiomyocytes. Instead, injury results in cardiomyocyte hypertrophy, replacement fibrosis and chronic inflammation, compromising cardiac function.

1.4. Neonatal and adult cardiomyocytes are different

Since the discovery of the transient regenerative capacity of the neonatal mouse heart, many studies have focused on identifying pathways to re-activate the neonatal regenerative program in the adult heart after injury. Neonatal and adult cardiomyocytes are structurally different. For example, adult cardiomyocytes are elongated with highly aligned uniform sarcomeres and dense myofibrillar structures [37, 38]. In mice, the adult cardiomyocytes switch in contractile gene isoforms from the fetal beta myosin heavy chain (β-MHC or MYH7) to the adult alpha (α-MHC or MYH6) isoform [38]. In addition, there are differences in fuel utilization: glycolysis in neonatal cardiomyocytes vs. primarily fatty acid oxidation in adult cardiomyocytes, which allows more efficient production of ATP [39]. Interestingly, a study comparing the transcriptome of neonatal and adult cardiomyocytes revealed that neonatal cardiomyocytes were more transcriptionally related to endothelial cells and adult leukocytes than to the adult cardiomyocytes. The authors argue that neonatal and adult cardiomyocytes could be considered as two distinct cell types [40], suggesting that neonatal cardiomyocytes are primed from the start to maintain an active cell-cycle program that allows them to continue to proliferate after injury. This raises the question of whether the adult heart has retained a subpopulation of cardiomyocytes with a permissive developmental program that allows the heart to maintain the annual 1% cardiomyocyte renewal.

1.5. Barriers to cardiomyocyte proliferation

In the context of cardiac regeneration, the term “terminally differentiated cardiomyocyte” describes a cardiomyocyte that lacks proliferative capacity. This permanent withdrawal from the cell-cycle in majority of postnatal cardiomyocytes correlates with an extensive change in the epigenome and transcriptome, as well as an increase in abundance of multinucleated and polyploid cardiomyocytes [35, 41-43]. Although the degree of cardiomyocyte polyploidization varies across mammalian species, all the mammalian species studied so far have cardiomyocyte cell-cycle exit shortly after birth [24]. As a result, the majority of murine adult ventricular cardiomyocytes (around 80% or more) [44] are binucleated and tetraploid (2Nx2, where N represents the haploid DNA content per nucleus). Recent evidence demonstrates that the percentage of mononuclear diploid cardiomyocytes in the adult heart depends on the genetic background, that varies substantially among different murine strains and is positively correlated with the potential for cardiac regeneration after injury [45]. There is sparse evidence showing that binucleated cardiomyocytes (2nx2) can undergo division [46]. At the time of birth, the neonatal human heart comprises of primarily mononucleated cardiomyocytes and, approximately 30 % binucleated cardiomyocytes. This proportion of mono – and binucleated cardiomyocytes does not change significantly after birth. However, in humans, cardiomyocyte polyploidization increases between birth and adulthood, and the majority of cardiomyocytes become mononucleated tetraploid (ca. 66 %) [17, 24].

Perinatal oxidative stress and consequent DNA damage due to higher oxygen levels may be a major contributor to the established cardiomyocyte cell-cycle withdrawal [47]. The alternative cell cycle routes to cytokinesis may be more advantageous during oxidative stress, since multiple sets of chromosomes may reassure gene expression simultaneous to gene-inactivating mutations [48]. Furthermore, alternative cell cycle routes to cytokinesis are also more favorable from a contractile and energetic standpoint. For example, dividing cardiomyocytes must disassemble their sarcomeres during M-phase to undergo cytokinesis, resulting in a loss of contractile function. Cardiomyocyte endoreduplication increases cardiomyocyte DNA content to enhance expression of contractile proteins and a higher pump strength, at the expense of cell renewal [49, 50]. Interestingly, Huang and colleagues have shown that mononucleated cardiomyocytes isolated from left atrium and pulmonary vein have different resting membrane potential, pacemaker activity and beating rate than binucleated cardiomyocytes [51] and may exert different functions [52]. In summary, it is suggested that loss of regenerative capacity in the hearts of adult mammals may be a trade-off in order for the adult heart to achieve higher contractility and efficient calcium handling to meet increased metabolic and hemodynamic demands, but sacrificing cytokinesis potential [53].

1.6. Manipulating the barriers to adult cardiomyocyte proliferation

The overall arrest in cardiomyocyte division has been attributed to downregulation of positive cell cycle regulators and centrosome disassembly, resulting in a block in the cardiomyocyte cell cycle [16, 35, 54, 55]. Many of the emerging signaling pathways with the potential to reverse cardiomyocyte cell cycle arrest and promote proliferation, were first discovered during heart development when cardiomyocyte proliferation is substantial. The hypothesis is that developmental signaling pathways may also be present in differentiated cardiomyocytes but are kept suppressed until re-activated with the right stimuli.

Cell cycle factors

Molecular interventions such as over-expression of cardiomyocyte cell cycle regulators have been attempted, but overcoming the multiple cell cycle barriers is a challenge, since cell cycle factors do not universally activate cell division [56-58]. For example, over-expression of cyclin B1-CDC2, and concomitant inhibition of p21/p27 induced cardiomyocyte multinucleation instead of proliferation [56-58]. Furthermore, manipulation of the adult cardiomyocyte cell cycle by overexpression of oncogenes has proven less efficient due to ensuing apoptosis [59]. Recently however, Mohamed et al. used a combination of four cell cycle regulators (4F: cyclin-dependent kinase 1 (CDK-1), CDK4, cyclin CCNB, and CCND) and detected a substantial increase in histone H3-phosphorylation in > 10 % of adult and > 20 % of P7 cardiomyocytes [9]. Next, they utilized intramyocardial delivery of adenoviruses encoding the 4F for 2 weeks post-MI and noted a successful induction of cardiomyocyte cell cycle activity and cardiac repair [9]. Similar results, with a rescue of heart function post MI, were also obtained with a simplified cocktail consisting of CDK4, Cyclin D1 and chemical inhibitors of TGF-β and Weel (2F2i) [9] (Figure 2, strategies for adult mammalian heart regeneration). The authors raise the possibility that the high percentage of pHH3 positive cardiomyocytes after 4 F treatment in vitro may be a result of aberrant phosphorylation of pHH3 or a G2/M arrest [9]. The observed repair in vivo warrants further investigation on the beneficial pro-mitotic effects of the 4Fs on cardiomyocyte proliferation.

Figure 2: Strategies for adult mammalian heart regeneration.

In this figure, we highlight therapeutic strategies to augment cardiac regeneration in adult mammalian hearts after injury. Several therapeutic approaches have been described to achieve cardiac regeneration either by stimulation of cardiomyocyte cell cycle re-entry by stimulation of endogenous pathways or by cell-therapy to deliver in-vitro produced induced pluripotent stem cell (iPSC) – derived cardiomyocytes possibly with engineered scaffolds. Stimulation of endogenous pathways of cardiomyocyte proliferation involve administration of mitogens like miR-199a, cyclins, cyclin dependent kinases (CDKs) to induce cardiomyocyte proliferation. Cell therapy comprises of application of 3D engineered scaffolds or extracellular matrix scaffolds to the heart at the site of injury and delivery of iPSC-derived cardiomyocytes at the site of injury.

Developmental factors

Hippo/Yap pathway

Among developmental signaling pathways, the Hippo/Yap pathway has been studied extensively in context of cardiac regeneration [60, 61]. The Hippo pathway comprises of a series of adaptors and kinases that promote phosphorylation of the transcriptional co-activator Yes-associated protein (YAP) [62], and the transcriptional coactivator with PDZ-binding motif (TAZ). YAP/TAZ remain transcriptionally inactive until phosphorylated, when YAP enters the nuclei and activates gene expression by interacting with DNA-binding transcription factors, such as TEAD (TEA domain family members), SMAD1, RUNX4 (RUNT-related transcription factor 4), T-box 5 (TBX5) and p73 [62] in the heart. Genetic deletion of YAP in mice resulted in underdeveloped hearts and embryonic lethality, while embryonic overexpression of YAP induced hyperproliferation of cardiomyocytes and severely disproportional ventricles and death [61, 63, 64]. Furthermore, deletion of a Hippo signaling component (Salv;Sav1) in cardiomyocytes showed evidence of regenerated myocardium after apical resection at P8 [60]. Subsequently, when YAP1 was over-expressed in the adult heart, cardiomyocyte proliferation was detected. In depth experiments revealed that Yap/TEAD complexes induce activation of the insulin-like growth factor pathway and upregulation of nuclear WNT target gene ß-catenin to induce cardiomyocyte cell cycle re-entry and regeneration post injury [64, 65]. Studies in mice have revealed that targeting the Hippo-YAP pathway has potential for increasing cardiomyocyte proliferation; however, it is important to point out that YAP is also a potent oncogene and stimulation of YAP may simultaneously modulate a maladaptive response. Recently, it has been shown that YAP activation induces cardiomyocyte dedifferentiation and heart failure when pressure overload is present [66].

Meis1

Developmental transcription factor Meis1 plays a key role in the cell cycle by inducing expression of p21 and thereby blocking cardiomyocyte cell cycle activity [67]. Conditional deletion of Meis1 in neonatal and adult mouse cardiomyocytes is sufficient to induce cell cycle reentry. However, direct deletion of Meis1 downstream targets, p21 or p27 or both, does not induce cardiomyocyte proliferation [68], but instead increases tumorigenesis. Interestingly, overexpression of the essential transcription factor Tbx20 in adult cardiomyocytes promotes cardiomyogenesis by directly acting on cell-cycle repressors p21, Meis1 and Btg2 [69]. This provides further evidence of the important role of Meis1 in cardiomyocyte cell cycle activity.

E2F

An additional example of developmental transcription factors that act as key regulators of cardiomyocyte proliferation is the E2F transcription factor family. The E2F family controls the synthesis of genes that regulate cell cycle entrance and passage. E2F1, 2 and 3, in particular, are considered “transactivators” that activate transcription of target genes for DNA replication and G1/S transition, thus cell proliferation [70]. E2F activity is regulated by a family of proteins, which includes the retinoblastoma protein (pRB) tumor suppressor, p107 and p130. Overexpression of E2F-1 increases cardiomyocyte apoptosis, while overexpression of the transcription factor E2F4 stimulates cardiomyocyte cycling in vivo [71, 72].

Hedgehog pathway

The role of Hedgehog (Hh) signaling pathway in adult heart homeostasis has been unclear, probably in part due to low expression of this signaling pathway in adult cardiomyocytes. Ischemic heart injury, however, reactivates the Hh pathway [73]. Recently, Singh and colleagues, compared the apical resection heart model in newts and mice and identified a novel HH-Gli1-Mycn gene regulatory network that promotes heart regeneration [14]. Fate-mapping experiments showed that HH-Gli1-Mycn induces cardiomyocyte proliferation of pre-existing cardiomyocytes. Furthermore, chromatin-immunoprecipitation sequencing (ChIP-Seq) analysis revealed that HH-Gli1-Mycn acts regulatory by directly regulating the expression of cyclin-dependent kinases including cyclinD2, cyclinE1 and Cdc7[14].

Neuregulin

The growth factor neuregulin-1 (NRG-1) and its receptors (ERBB2 and ERBB4), are critical for cardiac development and maintenance of adult heart [74]. NRG-1 is an extracellular factor with multiple effects on cell survival, metabolism, angiogenesis, and sarcomere disassembly in addition to cardiomyocyte division [75]. The mode of action of NRG1 seems to be age dependent. During embryogenesis, BMP10 - NRG1 signaling between endocardium and myocardium initiates differentiation and maturation of cardiomyocytes [76]. However, in the adult heart, NRG1 signals via ERBB2 and ERBB4 receptor to stimulate cardiomyocyte proliferation in vivo [77, 78] (Figure 3). EGF/NRG/ERBB signaling in the cardiovascular system is clearly important but has yet to lead to a regenerative or heart failure therapy.

Figure 3: Signaling pathways that modulate cardiomyocyte proliferation.

Here, we outline the major signaling pathways that have been identified to modulate adult cardiomyocyte cell-cycle re-entry. Well-characterized endogenous pathways regulating cardiomyocyte proliferation include Wnt – ß-catenin signaling, neuregulin and Hippo-YAP signaling. Activation of Wnt signaling by the binding of Wnt to its receptor frizzled, results in the translocation of ß-catenin to the nucleus, activating transcriptional activity for cell proliferation. Binding of neuregulin to its receptors ErbB2 and ErbB4 mediate translocation of Yap to the nucleus and activation of transcriptional events that lead to cell proliferation. On the other hand, reactive oxygen species – mediated DNA damage results in cell cycle arrest via Atm phosphorylation and Weel activation.

Hypoxia Inducible Factors and the hypoxic response

Mounting evidence suggests that hypoxic conditions can induce cardiomyocyte proliferation and cardiac regeneration in lower vertebrates and in the adult murine heart after an MI [79-81]. In a study by Nakada et al. [79], adult mice were subjected to chronic severe hypoxemia, with observed reduction in oxidative DNA and altered mitochondrial metabolism. This was sufficient to induce reactivation of cardiomyocyte mitosis and heart repair after induced injury [79, 80]. This reactivation of cell cycle activity was associated with hypoxic activation of hypoxia-inducible factor 1α (HIF1α) [79]. Scavenging of hydrogen peroxide (H₂O₂) (e.g. catalase overexpression) was suggested to cause a delayed cardiomyocyte cell cycle arrest, whereas hyperoxic or ROS-generating conditions accelerate it [82]. It should be noted, however, that scavenging of H₂O₂ by transgenic overexpression in zebrafish was shown to impair heart regeneration [83]. Thus, increased ROS levels can lead to activation of the DNA-damage-response pathway, which can induce cell-cycle arrest of postnatal cardiomyocytes, and it will be important to determine how hypoxic signaling may be exploited in adult mammals and humans.

These important discoveries reveal that the cardiomyocyte cell cycle is tightly controlled and highlights that there is a preserved cellular program in place to prevent loss of organ function by uncontrolled proliferation. The functional relevance and the regulatory mechanism that drive post-natal multinucleation and physiological polyploidization in the heart, remain incompletely understood. What is clear is that both in homeostasis and after injury, the majority of adult mammalian cardiomyocytes undergo DNA synthesis but do not complete cytokinesis. A key to enhanced heart regeneration could be to define a subpopulation of cardiomyocytes that might be more amenable and responsive to mitotic stimuli.

2. Secreted factors induce cardiomyocyte proliferation

Although the mass of the heart is primarily cardiomyocytes, cardiomyocytes are outnumbered compared to the total of fibroblast, endothelial, smooth muscle, and inflammatory cells [84]. These cells create a microenvironment that most likely contribute to the regulation of cardiomyocyte proliferation, through paracrine signaling of secreted molecules. A number of secreted factors, cytokines and growth factors appear to play a role in cardiomyocyte cell cycle reentry and cardiac repair [85-87]. The paracrine signaling effects could be important as some cell therapies appear to benefit the heart even in the absence of new cardiomyocyte formation [88-90]. Cardiac regeneration may thus be stimulated by exogenous factors produced by non-cardiomyocytes (Figure 3, signaling pathways that modulate cardiomyocyte proliferation).

2.1. Reactivation of cardiomyocyte cell cycle with secreted factors

Normal thyroid hormone secretion is necessary for optimal cardiac performance [91] and thyroid hormone supplementation has been proposed as treatment for heart failure patients [92]. Recently, Hirose et al. demonstrated that impairing thyroid hormone T3 signaling by either pharmacological inhibition or gene-editing approaches caused an increase in diploid cardiomyocytes by 21 % in mice at postnatal day 14. Moreover, blocking thyroid hormone signaling increased cardiomyocyte proliferative rates, such that the hearts of these mice were 37% larger than controls and contained significantly higher number of cardiomyocytes. This also resulted in a 62 % reduction of scar area and improved cardiac function after myocardial injury [15]. Thyroid (T3) administration in vivo has been previously shown to induce activation of PI3K/Akt/mTOR pathway and stimulate cardiomyocyte protein synthesis [93]. More detailed understanding of how thyroid signaling regulates regenerative capacity is warranted.

Several growth factors have also been described as capable of stimulating cardiomyocyte cell-cycle reentry. Fibroblast growth factor 2 (FGF2) was one of the first growth factors shown to promote neonatal cardiomyocyte DNA synthesis via protein kinase C activation [94]. Phase I trials have shown that intramyocardial injection of fibroblast growth factor 1 (FGF1) during coronary artery bypass graft surgery improves collateral artery growth and capillary proliferation by stimulating anti-apoptotic activity [95]. Treatment with FGF1 and a p38 mitogen-activated protein kinase inhibitor (p38i) in rats after myocardial injury showed promising effects on cardiomyocyte proliferation [96]. Other hormones with similar outcome are Growth hormone-releasing hormone agonist (GHRH) and Hepatocyte growth factor (HGF). Studies in swine models showed that delivery of IGF1/hepatocyte growth factor (HGF) or Growth hormone-releasing hormone agonist (GHRH-A) after ischemic injury significantly reduced infarct size by increasing formation of smaller cardiomyocytes and increased capillary density [97] [98]. At this time, several clinical trials (Phase I-III) using HGF gene therapy for peripheral arterial occlusive disease and coronary heart disease are undergoing [99].

Among the angiogenic secreted factors that have potential to stimulate cardiac regeneration is Follistatin-like 1 (Fstl1) [100, 101]. Fstl1 protects cardiomyocytes from cell death and hypertrophy in vitro, but it also suppresses cardiomyocyte differentiation from stem cells by inhibiting BMPs. In vivo, Fstl1 is normally present in healthy epicardium, but re-distributes to the myocardium upon injury [8]. Surprisingly, transgenic overexpression of Fstl1 in the adult heart does not elicit a regenerative response [8]. Instead, local delivery of Fstl1 through the epicardium, in both mice and swine, induced repair after an MI [8]. The authors conclude that Fstl1 must be synthesized, glycosylated and secreted by epicardium to act like a cardiomyogenic factor and emphasize that the source of the signaling molecule as well as the post translational modifications has a crucial impact.

Acute inflammation occurs immediately after heart injury and stimulates a regenerative response in the neonatal mouse heart [85]. Analyses of the global gene expression changes in the neonatal mouse heart following apical resection identified Interleukin 6 (IL6) [85] and Interleukin 13 (IL13) [102] as regulators of cardiomyocyte proliferation. Downstream regulators, STAT6, STAT3, and periostin were found to be critical mediators of inflammatory signaling in cardiomyocytes. When IL-13 was added to cultured neonatal and adult cardiomyocytes, ERK1/2 and Akt were activated, resulting in increased proliferation [86]. Mice lacking IL13 in the heart showed decreased cardiomyocyte proliferation and impaired heart repair after apical resection. Periostin, however, a secreted extracellular matrix protein, is highly expressed in the developing heart [103] but it is almost undetectable in the adult heart. Remarkably, periostin levels increase following ischemic injury, primarily by fibroblast activation [104]. Extracellular delivery of periostin after an MI stimulated cardiomyocyte mitosis and cardiac regeneration [105]. This finding has been challenged by others, who reported either no effect on cardiomyocyte proliferation or an increased cardiac fibrosis [106]. In summary, manipulation of signaling molecules produced by the non-cardiomyocyte cell populations in the heart might be important for overcoming post-natal cardiomyocyte cell cycle arrest to promote cardiac regeneration.

2.2. miR in cardiac repair

Accumulating evidence has shown that cardiomyocyte proliferation is under control of the miRNA network. MicroRNAs (miRNAs) are small non-coding RNAs (~22 nt in length) that play a role in posttranscriptional regulation of gene expression in the heart (reviewed elsewhere [107, 108]). For example, inhibition of the miR-15 family increased cardiomyocyte proliferation and improved cardiac function after an MI [21]. The miRNA-302-367 cluster has been shown to target the Hippo pathway to promote cardiomyocyte mitosis. Interestingly, this cluster of miRNAs can reactivate cardiomyocyte renewal in a Hippo-independent manner by stimulating homeodomain transcription factor Pitx2 expression. Pitx2 activates gene expression of electron transport chain components and ROS scavengers. Specifically, Nrf2, a regulator of the antioxidant response, is targeted, adding additional evidence of oxidative stress being a barrier to cardiomyocyte mitosis [109].

Other miRNAs, such as the miR-17-92 cluster, miR-214, and the miR-222 cluster have been reported to contribute to cardiac repair in vivo [110-113]. In an unbiased screen of more than 800 human miRNAs, multiple miRNAs were identified and shown to promote neonatal rat cardiomyocyte proliferation [114]. Among these miRNAs, cardiac overexpression of miR-199a or miR-590 induced cardiac repair of infarcted heart while reducing fibrosis. More recently, Borden et al. showed significant cardiac repair after MI, upon miR-294 treatment in mice [115]. Another study by Gabisonia et al. demonstrated that expression of human miR-199a in infarcted pig hearts can stimulate cardiac repair [116]. However, the authors also emphasize that miRNA need to be carefully dosed as repeated delivery and uncontrolled expression of the miR-199a resulted in sudden arrhythmic death of most of the treated pigs. Therefore, the positive outcomes in preclinical studies demonstrate the therapeutic potential of miRNAs, but this therapy needs to be tightly controlled.

3. Conslusion and Future directions (conluding remarks)

In conclusion, promoting de novo cardiomyocyte proliferation and differentiation, while modulating cardiac fibrosis, is a promising approach for restoring cardiac function following ischemic injury. Studies have demonstrated that cardiomyocyte mitosis can be induced with hormones, growth factors, miRNAs and also small molecules. New discoveries and promising preclinical outcomes demonstrating cardiac repair provide optimism regarding novel therapies for human systolic heart failure. However, we must also acknowledge and address the multiple challenges such as increased proliferation of non-myocyte cell types in the heart and the ubiquitous possibility of inducing cardiac arrhythmias. True heart regeneration is a complex and precisely controlled process. Therefore, a deeper understanding of cardiomyocyte biology and the crosstalk between the different cell populations in the heart is required for development of efficient approaches for repairing the mammalian heart.