How can the adult zebrafish and neonatal mice teach us about stimulating cardiac regeneration in the human heart?

During myocardial infarction (MI), blocked coronary arteries cause individual cells to become deprived of oxygenated blood, which could lead to death of up to one billion cardiomyocytes [1]. In the adult human heart, instead of regeneration of damaged myocardium, a fibrotic scar forms after injury, impairing cardiac function and thereby predisposing to incurable heart failure (HF) [2]. Most treatments for HF aim to either ameliorate symptoms or slow the pathological remodelling which underlies the progressive nature of the disease [3,4]. Although some HF patients undergo transplantation, the number of available donors is limited while demand continues to rise due to an increase in the mean age of the population and increasing survival following MI [4]. Additionally, although allogeneic transplantation theoretically improves outcomes, prognosis remains poor due to chronic rejection and the detrimental effects of long-term immunosuppression [4]. In contrast, treatment strategies that aim to replace lost cardiomyocytes via endogenous regeneration may prevent HF post-MI.

There is a divergence between species in the response to myocardial injury [4,5]. In stark contrast to the adult mammalian heart, as outlined in Figure 1, both the adult zebrafish and neonatal mouse have been shown to regenerate healthy new cardiomyocytes to replace damaged myocardium [5,6]. Indeed, following a diverse range of injury protocols, the adult zebrafish and neonatal mouse employ endogenous regenerative pathways that involve the proliferation of preexisting cardiomyocytes without the use of a progenitor cell population, ultimately facilitating scarless regeneration [5–9]. However, recent studies have illustrated that the role of different regeneration pathways is mode-of-injury dependent [10] and, as summarized in Table 1, different methods of cardiomyocyte injury have unique advantages and disadvantages that merit discussion [5,7,9,11–20]. One of the major differences between the cryoinjury and resection protocols is the amount of fibrotic tissue produced as a by-product. After resection injury, a relatively small amount of collagen localizes superficially on the cut surface of the myocardium [5], whereas in cryo-injured hearts, a significant amount of collagen transiently accumulates in place of the dead tissue mass [7]. Nevertheless, it has been demonstrated that zebrafish are capable of efficiently regenerating cardiomyocytes removed by either genetic ablation [9], where up to 60% of cardiac muscle cells are removed, or by severe cryoinjury [13]. In murine cardiomyocytes, there is a transition from hyperplastic to hypertrophic growth at postnatal day 7 (P7) that correlates with cell cycle arrest and loss of regenerative capacity [21]. In contrast, the adult human heart was long considered to be a postmitotic organ [22]. However, a recent pulse-chase experiment by Bergmann et al. showed the integration of ¹⁴C into the DNA of cardiac cells from the atmosphere, proving they turnover at a rate of approximately 1% annually, with turnover gradually decreasing with age [23]. Although the meager rate of renewal remains insufficient for generating a meaningful response to injury, this finding nevertheless led to a paradigm shift, reinvigorating interest in regenerative strategies for treatment of cardiac disease. Despite this, the precise proliferative capacity, and the programmed timing of cell cycle arrest of cardiomyocytes in humans remains unknown. Here, mechanisms underlying the regenerative capacity of neonatal mice and adult zebrafish are reviewed and the potential for their translation to humans is discussed.

Table 1. The different cardiac injury methods used in zebrafish and neonatal mice.

Model	Diagram	Protocol	Relation to mammalian physiology	Advantages	Limitations	Ref.
Resection		∼20% ventricular myocardium surgically removed Tissue recovers within ∼60 days	Thrombosis and collagen deposition superficially on wound	Fast and complete recovery of cardiac function	Time-consuming Imprecise/tedious	[5,11]
Cryoablation		Extensive cell death over ∼20% of ventricular wall Apex of ventricle is exposed, and a precooled cryoprobe and dry ice is applied to freeze the ventricle Scar tissue replaced with new myocardium within ∼2 months	More similar to localized damage seen after MI in mammals Inflammatory response	Extensive cell death Well tolerated	Injury observed in all cell types Morphological alterations postrecovery	[7,12,13]
Genetic ablation		∼60% ventricular cardiomyocytes depleted Cardiomyocytes undergo genetic manipulation and express toxins → production of cytotoxic metabolites Tissue recovers within ∼30 days	Signs of heart failure	Reversible Target specific cell lineage Reproducible Practical for large-scale analyses Noninvasive Temporal control	Complete ablation not always achieved Non-specific ‘bystander effect’ on neighboring cell types Does not reproduce pathological changes seen following MI	[9,14,15]
Chemoptogenetic ablation		Zebrafish express FAP in cardiomyocytes A fluorogenic dye administered binding FAP Localized near-infrared exposure results in excited photosensitizer-mediated ROS generation, which damages 30% of cardiac tissues	In humans, post-MI and particularly postreperfusion, there is an increase in ROS production, which further damages the myocardium; similarly, this chemoptogenic ablation model generates ROS that damage cardiomyocytes	This injury method did not activate the epicardium and endocardium May help identify specific cardiomyocyte responses to injury without whole organ activation	This model produces milder injury and only affects cardiomyocytes Post-MI, other cells in addition to cardiomyocytes are involved, which this model does not address	[16]
LAD ligation		LAD is permanently ligated with a single suture Infarct size ∼10–15% of the left ventricle Complete cardiac regeneration within ∼7 days	Mimics pathophysiological changes seen after MI	Reproducible Possibility for minimally invasive methods that minimize postsurgical mortality	Does not always stimulate complete regeneration Only feasible in mice Invasive, time-consuming Microsurgery – extensive training required	[17–20]

Different methods of cardiomyocyte injury (resection, cryoablation, genetic ablation, chemoptogenetic ablation and LAD ligation) are used in regenerative studies in zebrafish and neonatal mice. This table summarizes the advantages and limitations of each methodology, as well as their relevance to human physiology.

A: Atrium; BA: Bulbus arteriosus; LAD: Left anterior descending artery; MI: Myocardial infarction; ROS: Reactive oxygen species; SN: Sinus venosus; V: Ventricle.

Images created with BioRender.com.

Cardiomyocyte-intrinsic mechanisms affecting cardiac regeneration

Cardiomyocyte polyploidization reduces their regenerative ability

Differences in cell cycle activity and ploidy levels between neonatal and adult mammalian cardiomyocytes may cause loss of regenerative ability in adult cardiomyocytes [24]. In mice, cardiomyocytes exit the cell cycle due to failed cytokinesis or karyokinesis in the absence of cytokinesis, thus terminally differentiating into bi-/multi-nucleated or polyploid cells [25]. In neonatal mice, the transition from diploid to polyploid cardiomyocytes coincides with the loss of regenerative potential at P7 [25]. Around this time period, mice cardiomyocytes also exhibit a shift toward binucleation. At P2, 93.2% of cardiomyocytes are mononucleated, whereas 78.4% of them are binucleated at P11 [25]. Furthermore, 99% of the adult zebrafish cardiomyocytes are normally mononucleated and diploid [26]. Induction of polyploidization in >45% of cardiomyocytes in the adult zebrafish resulted in loss of regenerative capacity postresection injury [26]. Accordingly, cardiomyocytes of nonregenerative species are often polyploid. However, patterns of ploidy and multinucleation vary significantly across mammalian species [24,27], thus limiting translatability of these findings to humans. Across experiments, 22.5 to 33.4% of human cardiomyocytes were found to be diploid, with the rest being polyploid and the majority of those being tetraploid [28,29]. Most of those cardiomyocytes are mononucleated cells that exhibit increasing polyploidization with age and after myocardial infarction [29–31]. Nevertheless, although ploidy does not hinder regeneration in other tissues, such as the liver [32], modulating cardiomyocyte polyploidy may be a useful therapeutic target in cardiac regenerative medicine.

Multiple miRNAs regulate cardiac regeneration

The cell cycle of cardiomyocytes is also regulated by certain miRNAs that control cardiomyocyte proliferation via mRNA degradation [33]. Important relevant miRNAs identified in murine hearts include miR-17-92, which can enhance cardiomyocyte proliferation in neonatal mice and in adult mice post-coronary artery occlusion [33], and miR-195, which limits cardiac regeneration in neonatal mice after left anterior descending artery (LAD) ligation [34]. Upregulation of miR-195, part of the miR-15 family, leads to cardiomyocyte binucleation via inhibition of Check1, a mitosis gene controlling cytokinesis and chromosomal segregation [35]. Also, various miRNAs from the miR-15 family were upregulated during P7 and P14 in mice, the time period associated with the loss of their cardiomyocyte regenerative capacity [35]. In addition, when zebrafish regenerate their cardiomyocytes post-resection injury, miR-133 levels are suppressed [36]. However, elucidating the role of different miRNAs in cardiac regeneration is limited by the vast number of targets they each possess [36]. For example, multiple downstream mediators of miR-133 have been identified in zebrafish, such as the cell cycle regulator mps1 and the cell junction cx43 [36]. A recent study on 96 miRNAs on human-derived induced pluripotent stem cell cardiomyocytes showed that they impact the regulation of the Hippo pathway and specifically its mediator YAP, but their role was not essential for cardiac proliferation to occur [37]. Overall, despite uncontrollable cell division being a potential risk, promoting cell cycle re-entry through the modulation of miRNA remains a potential therapeutic option for cardiac regeneration that merits further research [33,34,36].

C-Myc expression may induce cardiomyocyte regeneration

C-Myc is a proto-oncogene that controls differentiation and proliferation in multiple tissues [38]. Under physiological conditions, c-Myc is not expressed in the adult murine heart, but it can be upregulated in the presence of pathological signals, such as ischemia and pressure overload [39]. Wang et al. found that upregulating the c-Myc/FoxM1 pathway increased cardiomyocyte proliferation after apical resection injury to the neonatal murine heart and post-LAD ligation in adult mice [40]. Kisby et al. also increased regeneration by inducing cardiomyocytes in mice to a pluripotent dedifferentiated state via c-Myc and three other transcription factors [41]. However, their administration via an adenoviral vector resulted in limited cardiac repair in the myocardium following LAD ligation [41], highlighting the necessity of improving available delivery methods. A recent study identified a positive correlation between c-Myc activity and P-TEFβ, thus proposing increasing P-TEFβ levels as a way of achieving c-Myc-induced cardiomyocyte proliferation [42]. P-TEFβ can act as a mediator of CDK9, with the levels of the latter remaining relatively high throughout the adult life of the zebrafish [43]. However, they decrease significantly in adult mice [43], further suggesting a possible role of P-TEFβ, and hence c-Myc, in cardiac regeneration. Research on the role of c-Myc in zebrafish cardiomyocyte proliferation has been limited and has not yet yielded definitive, clear-cut findings. Miklas et al. suggested that c-Myc targets, including oxidative phosphorylation pathways, are overexpressed in zebrafish hearts 7 days after chemical ablation injury and generally in young zebrafish hearts [44]. Another study, however, demonstrated downregulation of c-Myc targets during cardiac regeneration in zebrafish after cryoinjury [45]. It is important to highlight that overexpression of c-Myc is carcinogenic [46] because its targets are upregulated in many tumors [45,46]. Therefore, its potential therapeutic expression in future studies should be transient to limit malignancy [46].

The Hippo-YAP pathway modulates cardiac regeneration

The Hippo pathway has been extensively studied in regenerative medicine because it regulates cell cycle activity and interacts with the NRG1 and ErbB2 pathway [47]. The Hippo pathway cascade consists of core kinases – namely, MST1/2, MAP4Ks and LATS1/2 [48]. Activation of this pathway leads to inactivation of the most important downstream effectors – namely, YAP and TAZ [49]. Hence, many targets exist in the pathway to affect the ultimate downstream mediators. When the YAP gene, a critical regulator of the Hippo pathway, was directly knocked out in neonatal mice, minimal cardiac regeneration and extensive fibrosis were observed after MI-induced injury [50]. Furthermore, Salvador, an upstream inhibitor of YAP, when knocked out of mice, promoted cardiomyocyte proliferation [51]. Because studies in mice found that YAP decreases with age and is undetectable by week 12 [52], more research is needed to determine whether a similar pattern occurs in humans and whether potentiating YAP expression is safe therapeutically, given that YAP is a proto-oncogene [53]. A study in zebrafish, however, demonstrated that YAP deficiency did not reduce cardiomyocyte proliferation in cryoinjured myocardium [53,54]. It may be that other or additional proteins contribute to cardiac proliferation in zebrafish via the Hippo pathway [48,54]; these differences emphasize the need for further research in large mammals to extrapolate accurately the therapeutic potential of modulating the Hippo pathway in humans.

Oxidative DNA damage can reduce cardiac regeneration

Adult zebrafish and neonatal mammals use anaerobic glycolysis as their energy source because they exist in a relatively hypoxic environment [55,56]. In contrast, adult mammals have a higher oxygenation status and perform aerobic respiration. The switch from glycolysis to oxidative phosphorylation causes reactive oxygen species (ROS) production, which damages proteins, lipids and DNA [57]. The detrimental effects of ROS were noted in a study on postnatal mice, which found an increase in oxidized DNA and an upregulation of the DNA damage response pathway by P7 compared with P0 [58]. Another study on neonatal mice showed that mild hypoxic conditions increased cardiomyocyte replication, whereas exposure to high oxygen levels increased oxidative stress and consequently decreased cardiomyocyte replication [59]. To further emphasize the causal role of ROS in cardiomyocyte cell cycle arrest, Puente et al. [59]. injected mice with diquat, an ROS generator, and found that cardiomyocyte replication decreased compared with control. Conversely, N-acetylcysteine, an ROS scavenger, administered on P14, increased the number of cardiomyocytes. Notably, cardiomyocyte proliferation gradually decreased over 1 month, suggesting that ROS scavenging may have limited benefits in long-term cardiac regeneration. Further research has shown that in vitro administration of HIF-1α, a transcription factor that mediates metabolic responses to hypoxia, correlated with an increase in cardiomyocyte cell cycle activity [60]. However, forced activation of HIF-1α signaling in cardiomyocytes was shown to cause dilated cardiomyopathy and HF [61]. Given the vital importance of aerobic metabolism, ways to minimize oxidative DNA damage should be the primary focus of research rather than reduction in mitochondrial oxidative phosphorylation.

In contrast to neonatal mice, recent studies on zebrafish suggest that elevated levels of ROS can promote a regenerative response [62]. Using RNA in situ hybridization, a recent study found that ROS production increased in zebrafish after metronidazole-induced cardiac injury due to the activity of duox/nox2, a family of ROS-generating NADPH oxidases [63]. ROS were found to promote cardiac myocyte proliferation because inhibiting them with apocynin resulted in decreased cardiac regeneration [63]. The study concluded that a likely cause for the pro-regenerative effect of ROS in zebrafish is its inhibitory effect on Dusp6, a PTP, resulting in derepression of MAP kinase signaling, favoring the ERK1/2 pathway and therefore promoting cardiac repair [63]. In fact, treatment of adult zebrafish with a selective inhibitor of PTP-1B increased cardiac regeneration rate by twofold to threefold [64]. Given the harmful effects of ROS on DNA, a means of increasing cardiac regeneration without inducing oxidative damage may be via modulation of substrates downstream of ROS, such as PTPs. Studies have also found that glycolysis is an important determinant of cardiac regeneration in zebrafish because blocking glycolysis decreases cardiac regeneration [56]. Both glycolysis and oxidation of PTPs may contribute to cardiac regeneration independently; however, more research is needed to confirm these findings.

The adult mammalian heart has limited cardiomyocyte replication, with annual regeneration rates of 0.5–1% [23]. Interestingly, fate mapping demonstrated that the majority of dividing cardiomyocytes in the adult mammalian heart originated from a population of hypoxic cardiomyocytes [6,65]. Additionally, studies have shown that mortality rates from ischemic heart disease decrease linearly with increasing altitudes, thus suggesting a potential link between hypoxia and cardiovascular health [66]. A study using human cardiomyocytes found that moderate hypoxia increased cardiomyocyte proliferation [67]. However, there are conflicting results in the literature, as a recent randomized controlled trial found that ischemic conditioning did not improve outcomes post-MI [68].

Fatty acid oxidation is the main source of energy in the adult human heart, whereas embryonic and neonatal cardiomyocytes produce energy mostly through glycolysis [69]. The shift toward fatty acid oxidation occurs to sustain the metabolic needs of the heart, but this may also contribute to loss of regenerative capacity in the human heart as cardiomyocytes exit the cell cycle [69]. Favoring glycolysis instead of fatty acid oxidation may improve cardiac regeneration. For instance, studies have found that in HF and dilatative cardiomyopathy, there is a shift from fatty acid oxidation to glycolysis, which may be a compensatory mechanism to improve cardiac function [57,70]. However, a different study found that glycolysis may favor cardiac fibrosis post-MI, and thus inhibiting glycolysis may result in reduced fibrosis. Promoting glycolysis to regenerate cardiomyocytes post-MI may therefore be complicated by increased fibrosis and the arrhythmogenic propensity of fibrotic substrates [71]. The risks and benefits of glycolysis upregulation should be further studied [72]. Further exploring the modulation of mitochondrial oxidative metabolism as a potential therapeutic target in cardiac regeneration may be an important step toward developing safe and effective regenerative therapies. Overall, more research is needed to define the impact of oxygen metabolism in cardiomyocyte proliferation and to elucidate potential therapeutic targets that are both safe and effective.

Cardiomyocyte-extrinsic mechanisms affecting cardiac regeneration

The autonomic nervous system regulates cardiomyocyte proliferation

Research has highlighted the importance of the sympathetic nervous system (SNS) in mediating cardiac regeneration. For instance, ablating sympathetic nerves in the subepicardium, using 6-OHDA inhibited regeneration in neonatal mice [73]. However, 6-OHDA has been shown to induce ROS production [74], which may have damaged the epicardium, an important regulator of cardiac regeneration [75]. Nevertheless, a study using a neural chemorepellent to disrupt sympathetic fibers without injuring the epicardium still impaired ventricular regeneration in zebrafish [76], thus suggesting that the sympathetic nervous system affects cardiac regeneration. Interestingly, reduction of parasympathetic supply by vagotomy in neonatal mice also impaired cardiomyocyte proliferation, indicating that the autonomic nervous system plays a broad role in cardiac regeneration [76]. Therefore, because MI results in sympathetic denervation in large mammals [77], future studies should assess if reinnervation can support cardiac regeneration.

Epicardium & cardiac regeneration

Over the past decade, there has been substantial evidence around the importance of the epicardium on cardiac regeneration via a huge array of mechanisms. Studies in zebrafish show that after ventricular resection, the epicardium becomes further activated and the epicardial cells undergoing epithelial–mesenchymal transition have further proliferation and more directly migrate toward the injured zone [78]. Much evidence exists around the array of growth factors secreted and stimulated by the activated epicardium. For example, epicardial IGF signaling in zebrafish was shown to increase cardiomyocyte proliferation after resection [79]. In zebrafish, it was shown that the subepicardial cardiomyocytes that primarily contribute to the regenerating myocardium have increased expression of the intracellular transcription factor gata4, which may be responsible for regeneration after ventricular resection [80]. Additionally, the activation and upregulation of epicardial Notch receptors in zebrafish were shown to be crucial in cardiomyocyte proliferation after ventricular resection [81].

NRG1 increases cardiac regeneration

Many molecular pathways have been found to regulate cardiac regeneration; for instance, the growth factor NRG1 was found to support cardiomyocyte proliferation by binding to ErbB2/ErbB4 receptors [82]. Activation of NRG1 was also shown to induce glycolysis, which, as previously discussed, favors cardiac regeneration [83]. Furthermore, NRG1 was shown to reverse the regenerative impairment caused by vagotomy in neonatal mice, suggesting that NRG1 also acts through the autonomic nervous system pathway [76]. A study using zebrafish identified an 11-fold increase in NRG1 after cardiomyocyte genetic ablation injury and even in the absence of injury, NRG1 treated neonatal mice had increased cardiomyocyte proliferation [84]. In humans, targeted NRG1 administration induced cardiomyocyte proliferation in cultured myocardial samples from infants with HF [85]. Older patients were unaffected, possibly due to reducing ErbB2 expression during ageing as seen in mice [85,86]. This suggests a potential therapeutic window for NRG1 administration, which may be overcome by methods of ErbB2 upregulation. Future research should explore methods to increase NRG1 and ErbB2 expression and their therapeutic potential.

IGF increases regenerative potential

IGF is an anabolic hormone contributing to the growth of tissues [87]. In zebrafish, for instance, inhibition of IGF1 receptors, resulted in impaired cardiomyocyte proliferation after ventricular resection [79,87]. Furthermore, in neonatal mice with induced MI, administration of IGFBP3 restored regenerative capacity after it was shown that knockout of IGFBP3 impaired regeneration [88]. These findings suggest the role of IGF in regeneration is conserved across species. Interestingly, when cortisol levels were increased in zebrafish with cryoinjured myocardium, there was a notable downregulation of IGF, leading to cardiac scarring [89]. When psychological stress was ameliorated, pharmacologically with propranolol/fluoxetine, regeneration was subsequently rescued [89]. Future research should therefore explore pathways modulating stress responses and whether these can be exploited therapeutically to promote regeneration.

The role of the extracellular matrix in cardiomyocyte regeneration

The regenerative ability of the zebrafish and neonatal mice heart may also be attributed to differences in extracellular matrix (ECM) components. ECM remodeling is a major contributor to scar formation and fibrosis, which hinders cardiac function [90]. Adult mammalian hearts exhibit a dominant fibrotic response of the ECM components to injury, which may contribute to their inability for cardiac regeneration [91].

Notari et al. found that in neonatal mice P2 hearts are ~50% stiffer than P1 hearts, coinciding with a loss of regenerative potential by P2 [92]. Similarly, in zebrafish a decrease in collagens is observed 7 days after amputation via resection injury, correlating with a decrease in cardiac stiffness [91]. Indeed, when cardiomyocytes from neonatal mice were cultured on less stiff matrices, they dedifferentiated [93]; cell rounding was observed, indicating cell division [94]. This evidence suggests that increase in ECM stiffness reduces the capacity for cardiac regeneration [93] and may therefore be a potential therapeutic target.

The differential composition of ECM between species influences cardiac regeneration. In mouse ECM, collagens predominate, encouraging tissue differentiation [91]. By contrast, in zebrafish ECM (zECM), elastin content is high, promoting cell proliferation whereas downregulation of fibroblasts restricts fibrosis [94]. Regeneration was stimulated when zECM was injected into adult mice hearts following acute MI via intramyocardial injection, suggesting that zECM has pro-proliferative effects [94]. However, it is unknown whether these results will be reproduced in subacute or chronic HF patients [94]. The effects of zECM did not persist beyond 6 weeks, suggesting a limited therapeutic window. Furthermore, repeated zECM injections may result in uncontrollable proliferation, via NRG1 signaling, thereby inducing HF [95].

The ECM gene Ccn2a is known to be a positive regulator of heart regeneration in zebrafish and neonatal mice. Ccn2a is highly expressed in injured zebrafish cardiac tissue, and loss of Ccn2a expression diminishes cardiomyocyte proliferation after cryoinjury [10,96]. Ccn2a also regulates expression of downstream ECM proteins such as fibronectins and collagens, which modulate heart regeneration by directing cardiomyocyte migration [10,97–99]. Further ECM molecules associated with cell proliferation include tenascin C and periostin. Upregulation of tenascin C corresponds with an increase in tissue remodeling and encourages regenerative growth in zebrafish following amputation via resection injury, whereas periostin has been shown to stimulate cardiomyocyte cell cycle reentry after cryoinjury [100,101]. Thus, a variety of ECM components work together to promote cardiac regeneration in zebrafish and neonatal mice.

A final key component of the ECM associated with the heart’s regenerative capacity is Agrin, an extracellular proteoglycan [90]. In mice, postnatal reduction of its expression coincides with loss of regenerative capacity [90]. Genetic deletion of Agrin in neonatal mice increases fibrosis, while reducing cardiomyocyte proliferation, demonstrating its critical role in heart regeneration [95]. In the neonatal mouse heart, Agrin binds to Dag1 and favors cardiac regeneration through destabilization of the dystrophin–glycoprotein complex and by triggering downstream YAP signaling [90]. Postnatally, Agrin levels drop, forcing Dag1 to bind to alternative ECM molecules such as Laminin, which indirectly promotes cardiomyocyte cell cycle arrest [90]. Intramyocardial injection of recombinant mouse Agrin into adult mice stimulated cardiomyocyte cell cycle reentry [90]. Furthermore, similar studies using recombinant human Agrin improved hallmark signs of HF in porcine hearts after infarction by LAD ligation and showed additional antiinflammatory properties [102]. However, Notari et al. found no difference between the levels of Agrin and Dag1 in P1 and P2 mice, suggesting Agrin does not influence neonatal heart regeneration to a great extent [92]. Additionally, Agrin-based therapies should be considered with caution, given its carcinogenic role in the liver via YAP signaling [103].

Therapeutically, targeting ECM handling presents multiple novel approaches to induce cardiomyocyte proliferation following acute MI. Intramyocardial injection of zECM, Agrin or derivatives of periostin or fibronectins may promote cardiac regeneration while preventing progression to HF. Additionally, the success of human recombinant Agrin in larger mammals encourages further research into therapeutic avenues. Nonetheless, these studies are not entirely comparable; MI induced through ligation of the left anterior descending artery does not always stimulate complete regeneration, whereas results obtained through ventricular resection methods are often reproducible [92]. Hence, there are promising results suggesting that ECM modulation can promote cardiac regeneration; however, more research is needed.

The immune system modulates cardiac regeneration

A powerful inflammatory response occurs post-MI, triggering cell death and leading scar-formation in adult mice [6,104,105]. An equally vigorous inflammatory response occurs following apical resection injury to the myocardium in neonatal mice; however, this precipitates regeneration instead [6,104,105]. That inflammation can simultaneously be detrimental and crucial for the repair of injured cardiac tissue is a concept that was only recently heeded; the mechanism underlying this dichotomy has since been researched extensively [106]. There are two distinct subsets of macrophages that mediate inflammatory (CCR2+) and reparative (CCR2-) functions, respectively [107–112]. One study induced cardiomyocyte ablation via a diphtheria-toxin system to simulate injury in adult mice [112]. Although both types of macrophages were present within the heart [113], the resident pro-regenerative CCR2- macrophages were proportionally replaced following injury by previously circulating, scar-forming CCR2+ macrophages. However, these findings should be approached cautiously because both vascular and interstitial cells are preserved, which would have succumbed to hypoxia postinfarction, making it possible that the reparative responses observed in this model differ from what truly ensues following ischemia. Nevertheless, another study found that depleting subpopulations of CCR2+ macrophages derived from blood monocyte reservoirs in neonatal mice resulted in improved regeneration [114]. Given that CCR2- macrophages predominate in the adult myocardium, it may therefore be possible to improve cardiac repair by therapeutically inhibiting monocyte recruitment following MI. Despite this, patients on steroids with lower monocyte counts exhibit impaired wound healing post-MI [115], and macrophage-depleted neonatal mice lost their ability to regenerate [104], suggesting that a finely calibrated macrophage-mediated response is important for regeneration.

Two subsets of macrophages have also been discovered in adult zebrafish, which exhibit similar features as in mice [116]. Further insights into the role of the immune system in regeneration employed by adult zebrafish were ascertained through experiments on a phylogenetically related species: the Medaka [117]. Unlike the adult zebrafish, Medaka are unable to regenerate their myocardia [118]. Comparisons of bulk RNA-Seq data between the two fish species following resection injury revealed their respective immune systems are distinguished by a triad of diverging responses: within the Medaka, a smaller number of macrophages were activated, fewer localized to the site of injury and neutrophils remained at the resection site for significantly longer [119]. Strikingly, modulating the Medaka’s immune response to expedite macrophage recruitment and neutrophil clearance so that it more closely resembled the adult zebrafish led to increased cardiomyocyte regeneration and scar resolution [120]. Although convenient, resection injury may not recapitulate the immune response to ischemic injury because it eliminates fragmented ECM proteins, dying cells and the sources of inflammatory signals, which would have been otherwise present in the context of infarction. Other immune factors and cell types have also been shown to play important roles in mediating regeneration in both species. For instance, IL-13 knockout mice fail to regenerate after apical resection, and administration of recombinant IL-13 increases cardiomyocyte proliferation [121]. Furthermore, ablation of zebrafish Treg-like cells compromised myocardial regeneration via inhibition of IL-10-independent secretion of NRG1 [122]. However, other studies have shown perivascular cells as another source of NRG1 within the heart [84], suggesting successful regeneration may require multiple sources of growth factors. These studies provide insights into how IL-13 and human Treg cells may potentially be targeted to facilitate regeneration of injured myocardium therapeutically, and together they highlight the complexity of immune regulation of regeneration in both neonatal mice and adult zebrafish. Ultimately, however, the extent to which these findings can be used to infer knowledge about humans remains unknown.

Conclusion

The aforementioned studies have highlighted multiple cellular and molecular pathways contributing to the regenerative ability of zebrafish and neonatal mice, as illustrated in Figure 2. Metabolically, anaerobic glycolysis promotes cardiac regeneration in zebrafish, whereas in adult mice, aerobic respiration results in ROS production, inducing cardiomyocyte cell cycle arrest. Hippo-YAP and IGF signaling are essential pathways that regulate cardiomyocyte proliferation, whereas decreased NRG1 and autonomic nerve signaling are associated with reduced regenerative capacity. The activated epicardium increases cardiac regeneration via secretion of various growth factors and further activation of genes such as gata4 and Notch. Furthermore, both species have higher proportions of diploid cardiomyocytes, which favor proliferation and healing postinjury. Regulation of gene expression through miRNAs and the proto-oncogene c-Myc also play a critical role in controlling cardiomyocyte proliferation. Finally, changes in ECM handling, and the immune response – in particular, monocyte recruitment – are imperative to postnatal hypertrophic growth and adverse tissue remodeling. These mechanisms provide insight into cardiac regeneration in smaller vertebrates, but because of the lack of knowledge in larger mammals, it is still difficult to test these concepts therapeutically. Ultimately, as discussed in the future perspective section, targeting these pathways to promote cardiomyocyte proliferation remains a promising approach to regenerate adult myocardium.

Future perspective

Recent advances & clinical applications

Previously, there has been limited success in translating regenerative mechanisms for clinical application. Nevertheless, a diverse range of opportunities have recently arisen for novel therapies that could stimulate cardiomyocyte proliferation, including but not limited to drugs targeting PTP inhibition, modulation of IGF stability, increasing cardiomyocyte sensitivity to c-Myc, intramyocardial injection of Agrin or zECM and NRG1 administration [85,88,94,102,123,124]. The new therapeutic avenues explored in this review have the potential to improve recovery post-MI, thereby surpassing current treatment strategies that aim merely to minimize pathological remodeling and the progression to HF following MI. However, further research is needed to determine the extent to which these approaches can be translated to clinical practice.

Future concerns

Despite the promise of these therapies, there is concern regarding their possible detrimental effects. For instance, using c-Myc therapeutically carries a carcinogenic risk, and modulating the immune system could result in a hypo- or hyper-sensitive response, thereby compromising healing post-MI. Moreover, hypertrophic hearts in HF patients have fewer mononucleated cells and reduced capacity for proliferation and may hence not respond as favorably to potential treatments due to the barriers of polyploidization [125]. Finally, given the vast evolutionary distance between zebrafish and humans, we cannot be certain that the pathways pertaining to the adult zebrafish outlined in this review are conserved in humans.

Therefore, future studies should aim to shift from smaller vertebrates to large mammals such as pigs, dogs and sheep. Fine-tuning our understanding of the mechanisms underlying regeneration in these species will facilitate the design of studies for therapies targeting cardiac regeneration in humans.