The role of miRNA regulation in fetal cardiomyocytes, cardiac maturation and the risk of heart disease in adults

Abstract

Myocardial infarction is a primary contributor towards the global burden of cardiovascular disease. Rather than repairing the existing damage of myocardial infarction, current treatments only address the symptoms of the disease and reducing the risk of a secondary infarction. Cardiac regenerative capacity is dependent on cardiomyocyte proliferation, which concludes soon after birth in humans and precocial species such as sheep. Human fetal cardiac tissue has some ability to repair following tissue damage, whereas a fully matured human heart has minimal capacity for cellular regeneration. This is in contrast to neonatal mice and adult zebrafish hearts, which retain the ability to undergo cardiomyocyte proliferation and can regenerate cardiac tissue after birth. In mice and zebrafish models, microRNAs (miRNAs) have been implicated in the regulation of genes involved in cardiac cell cycle progression and regeneration. However, the significance of miRNA regulation in cardiomyocyte proliferation for humans and other large mammals, where the timing of heart development in relation to birth is similar, remains unclear. miRNAs may be valuable targets for therapies that promote cardiac repair after injury. Therefore, elucidating the role of specific miRNAs in large animals, where heart development closely resembles that of humans, remains vitally important for identifying therapeutic targets that may be translated into clinical practice focused on tissue repair.

Introduction

Cardiovascular disease is the largest cause of death in most developed countries, and therefore there is a need to understand the pathogenesis of heart disease to discover new therapies to reduce this burden (WHO, 2017). The adult human heart has limited capacity to repair damage because adult cardiomyocytes lack the ability to proliferate (Zak, 1973; Bergmann et al. 2009). This poor response to injury is in stark contrast to the zebrafish, whose cardiomyocytes maintain the capability to proliferate throughout life, and thus an adult zebrafish can regenerate heart tissue even after a substantial injury (Jopling et al. 2010). Thus, humans after birth have lost an evolutionarily ancient ability to regenerate heart tissue after injury.

The absence of effective regenerative treatments in humans highlights the need for new approaches for healing heart damage. microRNAs (miRNAs) have been implicated in the regulation of genes involved in cardiac cell cycle progression and regeneration in several species. However, the significance of miRNA regulation of cardiomyocyte proliferation for humans and other large mammals, where the cardiomyocyte proliferation capacity after birth is diminished compared to that of zebrafish, mice and rats, remains unclear. Therefore, to define the role of miRNAs in the modulation of cardiomyocyte proliferation, we have reviewed evidence for the gene and protein pathways modulated by the miRNAs that underpin its regulation in different animal models.

Myocardial infarction and the need for new therapeutic approaches

Myocardial infarction (MI) is a key component of the burden of cardiovascular disease world‐wide, causing growing hospital admissions and financial impact in most developed countries (Wong et al. 2012; Benjamin et al. 2017; Wilkins et al. 2017). Increased community awareness of early signs of a heart attack has reduced the time between the event and hospitalization, thereby increasing patient survival (Rubart & Field, 2006). Additionally, shorter door‐to‐needle times once in hospital, with prompt revascularization of the occluded artery has resulted in significantly reduced infarct size and reduced hospital mortality (Yaylali, 2010). Notwithstanding this progress, many patients still suffer significant irreversible myocardial injury, with chronic damage to the heart muscle resulting in poor clinical outcomes and a greater likelihood of heart failure in later years (Krum et al. 2011; Fig. 1). Current secondary prevention approaches after acute coronary syndromes focus on addressing the ongoing symptoms and risk factor modification to prevent re‐infarction. Heart regeneration treatments, such as stem and progenitor cell therapy, have significant challenges in production and directed administration (Michler, 2013; Nowbar et al. 2014; Nature, 2017; Quyyumi et al. 2017). To date, clinical trials of these treatments lack effectiveness, and although they can successfully promote revascularization, they fail at promoting cardiomyocyte proliferation and regeneration of heart tissue (Michler, 2013; Nowbar et al. 2014; Nature, 2017; Quyyumi et al. 2017). The current lack of effective treatments for cardiac repair necessitates the development of new approaches to repairing heart damage. One strategy is the targeted utilization of miRNAs as a tool to promote cardiomyocyte proliferation and cardiac repair, by attempting to reproduce the regulatory conditions in a fetal heart.

Figure 1. Chronic heart disease timeline.

(1) Hypertension and MI account for 75% of the total risk for chronic heart failure. (2) Using a combination of risk factors an absolute 5‐ or 10‐year individual risk for chronic heart disease can be calculated. While those patients at high risk can be readily identified using these criteria, the majority of MIs occur in a much larger pool of people who have intermediate risk with only one major risk factor (Krum et al. 2011). Most of the patients at intermediate risk do not experience a cardiac event until they are at an older age, and therefore lifestyle modifications as early as possible are important in reducing cardiovascular disease risk. (3) Cardiac remodelling occurs during chronic heart failure. (4) Treatments utilized and effectiveness varies from patient to patient, often requiring multiple treatments before stabilization of chronic heart failure. Some clinically stabilized patients will return to an unstable chronic heart failure state (red arrow). (5) Quality of life and survival rates for severe chronic heart failure are similar to the most common forms of cancer, with a case‐fatality rate of 75% over 5 years overall. CHF, chronic heart failure; MI, myocardial infarction. Adapted from Krum et al. (2011).

Regulation of normal cardiomyocyte proliferation, hypertrophy and differentiation

The heart is the first functional organ to develop in most species, and also one of the organs whose temporal development during pregnancy is easiest to disrupt (Abdulla et al. 2004). Congenital heart defects generally develop at the beginning of pregnancy during early development of the functional four‐chambered heart. At this time, changes to the intrauterine environment have the largest impact upon overall fetal heart health, and are the largest cause of miscarriage (Allan et al. 1986). The human heart completes organogenesis with the development of conduction and circulatory systems during the eighth week of human pregnancy, after which the heart continues to grow via controlled cardiomyocyte proliferation and hypertrophy (Soonpaa et al. 1996; Anderson et al. 2003). Cardiomyocytes make up the majority of the cell volume in the heart, followed by non‐myocytes, such as fibroblasts, vascular smooth muscle cells, endothelial cells and mast cells (Nag, 1980; Zak, 1984). During much of fetal development, the heart grows by the proliferation of mononucleated cardiomyocytes, which is known as a period of hyperplastic or proliferative growth (Soonpaa & Field, 1998). Binucleation occurs when the cardiomyocytes undergo karyokinesis (nuclear division) in the absence of cytokinesis (cell division) (Ahuja et al. 2007). Binucleated cardiomyocytes have a limited capacity for cell division, hence they contribute to the mass of the heart by increasing their size by hypertrophic growth (Oparil et al. 1984).

Animal models for heart development and cardiac regeneration

The zebrafish is one of the most important models for both developmental and regenerative biology as it can regrow many injured or amputated tissues such as fins, retinae, optic nerves, spinal cord, brain, hair cells, pancreas, liver, kidney and heart muscle (Kikuchi, 2014). Physiologically, the zebrafish heart is dissimilar to the human heart, with a two‐chamber system, mostly mononucleated cardiomyocytes, and it functions in a relatively hypoxic environment throughout life (Wills et al. 2008; Singleman & Holtzman, 2012). Mice and rats have been the focus of many miRNA heart development studies but their hearts are dissimilar to humans in both the timing of cardiac maturation and their cardiovascular physiology (Black et al. 2012; Botting et al. 2012; Porrello & Olson, 2014). The developmental timing of quiescence of cardiomyocytes in mice and rats occurs during the period just after birth (Fig. 2), when there is substantial cardiac remodelling initiated by the large change in circulation (due to the initiation of air breathing) and a switch in cardiac metabolism from mainly glycolysis in utero to fatty acid oxidation in the newborn (Lopaschuk & Jaswal, 2010). The changes in cardiac non‐coding RNAs and target gene expression in rodents during the perinatal period are difficult to interpret due to the large cardiac metabolic changes occurring alongside the cardiac developmental changes. In contrast, at birth the process of cardiomyocyte quiescence is nearing completion in human and ovine cardiomyocytes and is not confounded by perinatal metabolic changes (Burrell et al. 2003; Mollova et al. 2013). Guinea pigs have also been used to study the end‐stage failing myocardium and the response to heart failure, as, much like the ventricular myocardium of humans and other large animals, guinea pig hearts have structurally similar cardiomyocytes (cardiac muscle fibres comprising predominantly β‐myosin heavy chain and small amounts of α‐myosin) and similar calcium cycling (Hasenfuss, 1998). Thus this model animal is well suited to defining the transition from cardiac hypertrophy to failure (Hasenfuss, 1998). In addition, the fetal guinea pig heart is more mature at birth in regards to mitochondrial structure and function as well as thyroid hormone signalling than mice or rats (Barrie & Harris, 1977; Castro et al. 1986; Hoerter et al. 1994). Although the timing of cardiomyocyte quiescence in guinea pigs is proposed to be similar to the human and sheep, this has not yet been proven. The sheep has been widely used as an animal model for developmental biology and cardiac programming and the development of the cardiovascular system of humans is more like sheep than rodents, including the maturation of cardiomyocytes and cardiac autonomic nervous system (Lebowitz et al. 1972; Burrell et al. 2003; Jonker et al. 2007). The similarity in developmental timing makes the sheep an appropriate animal model for studies of the role of miRNA in heart development and regeneration in humans (Burrell et al. 2003).

Figure 2. The transition of cardiomyocyte proliferation (green) to quiescence and hypertrophy (red) across gestation in zebrafish rodents, sheep and humans.

Guinea pigs have not been included because the timing of the transition is not fully known. Adapted from Morrison et al. (2015).

Issues faced when translating cardiac regeneration models to humans

Although all of the outlined animal models are useful for studying heart regeneration, each species has differences from the human, which may potentially impair translation to the clinic. In zebrafish, the cardiomyocytes are all mononucleated and maintain the ability to proliferate throughout life, and thus their heart is able to regenerate after significant damage (Poss et al. 2002; Yin et al. 2012). In mice and rats, binucleation begins 3–4 days after birth (term, 21 days) (Clubb & Bishop, 1984), whereas in sheep this process begins after 110 days’ gestation (term, 150 days) and is completed just after birth, resulting in a limited capacity for regeneration in postnatal life (Adler & Costabel, 1980; Burrell et al. 2003; Jonker et al. 2007; Botting et al. 2012). Interestingly, most of the cardiomyocytes in the human heart are mononucleated, and although they generally don't become binucleated, most lose the capacity to proliferate in late gestation (Schmid & Pfitzer, 1985; Woodcock & Matkovich, 2005; Mollova et al. 2013), and thus there is limited cardiomyocyte proliferation after birth (Xin et al. 2013). Furthermore, in the human heart, 90% of cardiomyocyte nuclei are diploid (2c) shortly after birth, whereas the majority of adult cardiomyocyte nuclei are tetraploid (4c) (Adler, 1991). The purpose of polyploidy in cardiomyocyte nuclei and multinucleation of cardiomyocytes is not well understood and although the mechanisms by which multinucleation occurs have become more clear (Paradis et al. 2014), the biological importance of this mechanism remains elusive. The limited regenerative capacity of adult cardiomyocytes in most species is not sufficient to replace lost cardiomyocytes as a result of injury, disease or ageing (Bergmann et al. 2009; Barnett & van den Hoff, 2011). Consequently, the number of cardiomyocytes a human has at birth is an important factor in an individual's risk of developing cardiovascular disease later in life.

Factors linked to the transition in cardiomyocyte growth

Some factors have been singled out as having crucial roles in the switch from cardiomyocyte proliferative growth to quiescence. These factors were identified using rodent models of cardiac development and include changes in thyroid hormone concentrations and exposure to atmospheric oxygen (Naqvi et al. 2014; Puente et al. 2014; Sereti et al. 2018). There are, however, inconsistencies with these results when applied to other species.

Recent insight into the timing of cardiomyocyte cell cycle arrest in mice and rats has implications for our understanding of the proliferative capacity of binucleated cardiomyocytes and indicates that they may proliferate in a 2 + 1 manner (cytoplasmic division resulting in two mononucleated and one binucleated cardiomyocyte) (Naqvi et al. 2014). For example, in rats, there is a proliferative burst at postnatal day 15 due to a spike in thyroid hormone, which occurs 2 weeks after birth (Naqvi et al. 2014). In contrast, in precocial species such as humans, sheep and pigs, where proliferation declines in late gestation, the major spike in thyroid hormone occurs before birth (Chattergoon et al. 2007, 2012; Forhead & Fowden, 2014). In these species, the prepartum cortisol surge is responsible for stimulating deiodination of thyroxine (T4) to triiodothyronine (T3), resulting in a large rise in plasma T3 concentration (Forhead & Fowden, 2014). In sheep, thyroid hormone concentration increases approximately four times in magnitude from ∼110–125 to 135–145 days’ gestation. This is accompanied by a decrease in T3 clearance as well as a doubling in thyroid hormone production rate (Thorburn & Hopkins, 1973). The period when thyroid hormone concentration is increasing is when sheep mononucleated cardiomyocytes become binucleated and transition from proliferative to hypertrophic growth with no evidence for a burst in proliferative capacity during this period (Jonker et al. 2007; Chattergoon et al. 2012). These data reiterate and emphasize the need for animal models that lose proliferative capacity antenatally in the area of cardiomyocyte development, to better reflect human heart disease.

It has been suggested that the transition to an oxygen‐rich environment after birth in mice is a key signal that induces their cardiomyocytes to cease proliferating. One line of evidence is that this process does not occur in the zebrafish as it remains in a relatively low oxygen environment throughout life (Puente et al. 2014). This process also correlates with a neonatal mouse being capable of heart regeneration shortly after birth (Porrello et al. 2011b ; Puente et al. 2014) due to it being mildly hypoxemic, likely a result of immature lungs at birth (Pringle, 1986; Harding & Bocking, 2001; Lock et al. 2013). Withdrawal from the cell cycle in neonatal mice can be delayed by decreasing oxygen levels slightly below atmospheric levels (15% oxygen compared to atmospheric 21%) during the neonatal period (Puente et al. 2014). This hypothesis is supported by the fact that hyperoxia in mice accelerates the withdrawal of cardiomyocytes from the cell cycle shortly after birth. The metabolic environment of the fetal mouse heart has more similarities to zebrafish than the murine adult, including their relatively lower $P_{{\mathrm{O}}_2}$ for a period after birth and a greater reliance on glucose than fatty acids for energy utilization (Puente et al. 2014). Further evidence demonstrates that severe systemic hypoxia in adult mice is sufficient to promote an incomplete regenerative response after myocardial infarction as well as increase cardiomyocyte proliferation in uninjured heart tissue (Nakada et al. 2017). This seems counterintuitive as myocardial infarction, as well as causing pathological hypertrophy, causes local cardiomyocyte hypoxia and ultimately cardiomyocyte loss, which may indicate that systemic hypoxia is essential for promoting the regenerative response. This response is consistent with remote limb transient ischemia preconditioning treatments that result in smaller infarct size in animals models, though this method has been somewhat less successful in the clinic (Gleadle & Mazzone, 2016).

The neonatal period in mammals, especially mammals with altricial development such as rats and mice, is a time of considerable adjustment and physiological distress. All of the neonate's organs receive multiple stimuli as a result of exposure to atmospheric oxygen and the need to feed, move independently, secrete waste and regulate body temperature (Yutzey, 2014). Although changes in atmospheric oxygen have been singled out as the cause of cardiomyocyte cell‐cycle withdrawal (Puente et al. 2014), other factors such as postnatal glucocorticoid changes (Fowden et al. 1998) and metabolic stress resulting from periods of starvation between birth and the initiation of nursing or intermittent feeding are also likely to contribute to cardiomyocyte maturation. In addition, chicken cardiomyocytes act counterintuitively to the atmospheric oxygen theory, as they do not undergo binucleation until weeks after hatching and have little difference in oxygen tension between in ovo and post‐hatching (Svensson Holm et al. 2014). Therefore, hypoxia may play a critical role in the capacity for regeneration after birth but is not likely to be the sole stimulus. Large animals, such as sheep, have cardiomyocytes that lose the capacity to proliferate during late gestation before birth (Fig. 3; Bensley et al. 2010), and while there is some limited capacity for proliferation after birth, this is not sufficient for regeneration (Herdrich et al. 2010). This loss of proliferative capacity occurs at a time when the fetus is still hypoxaemic ($P_{\mathrm{a}{\mathrm{O}}_2}$ = 20–25 mmHg in the sheep fetus (Orgeig et al. 2010; Duan et al. 2017) and $P_{\mathrm{a}{\mathrm{O}}_2}$ = 20–35 mmHg in the human fetus (Soothill et al. 1986)), and therefore another trigger must exist that may be modulated by the switch to an oxygen‐rich environment and fatty acid metabolism at birth. In large animals, the effect of fetal hypoxemia on heart development does not change cardiomyocyte proliferation; rather, the lower oxygen and substrate/energy delivery to the fetal heart decreases the number of cardiomyocytes and induces pathological hypertrophy (Rouwet et al. 2002; Wang et al. 2011; Botting et al. 2014). This response suggests that oxygen exposure has diverse effects in different species and is dependent on a number of factors rather than on only lowered oxygen exposure.

Figure 3. Generalized comparison of oxygen availability across gestation in different species and the timing of cardiomyocyte quiescence.

Quiescence in human and sheep cardiomyocytes starts to occur before birth when fetal $P_{\mathrm{a}{\mathrm{O}}_2}$ is still low. This transition occurs after birth in mice and rats. The much slower rise of neonatal $P_{\mathrm{a}{\mathrm{O}}_2}$ in these rodent species may reflect their underdeveloped lungs during the first two weeks of life. Zebrafish remain in a relatively hypoxic environment throughout development and after birth and retain the capacity for cardiomyocyte proliferation throughout their lifespan (Bensley et al. 2010; Jopling et al. 2010; Porrello et al. 2011b ; Mollova et al. 2013).

Cell cycle regulators and miRNA involved in cardiomyocyte proliferation

Heart muscle cell proliferation is a highly regulated and finely tuned developmental process involving numerous cellular signals working in concert to ensure the correct assembly of the functional organ. Positive cell cycle regulators such as cyclins and cyclin‐dependent kinases (CDKs) are highly expressed during embryonic heart development, but downregulated in the adult heart (Fig. 4; Pasumarthi & Field, 2002; Ahuja et al. 2007), with overexpression of cell cycle activators, such as SV40 large T antigen, cyclin A2, and cyclin D2 causing proliferation and dedifferentiation of cardiomyocytes (Katz et al. 1992; Chaudhry et al. 2004; Engel et al. 2005; Pasumarthi et al. 2005; Engel et al. 2006; Kuhn et al. 2007; Bersell et al. 2009; Kubin et al. 2011). Premature exit from the cell cycle in fetal development can be caused by adverse changes to the intrauterine environment, resulting in a less structurally complete and less functional organ at birth and throughout postnatal life (Lock et al. 2017).

Figure 4. Proliferation of fetal cardiomyocytes is regulated by multiple signalling pathways that stimulate or inhibit cyclins and cytokinesis.

Promotion of cell cycle progression is indicated by blue arrows and inhibition is indicated by red arrows. AKT, protein kinase B; Ang‐II, angiontensin II; AT‐R, angiotensin receptor; CDK, cyclin dependent kinase; ERK, extracellular signal‐related kinase; FGF1, fibroblast growth factor; FGFR, fibroblast growth factor receptor; G₀, gap zero phase (quiescent); G₁, cell cycle first gap phase; G₂, second gap phase; GSK‐3β, glycogen synthase kinase‐3β; IGF, insulin‐like growth factor; IGF‐1R, insulin‐like growth factor‐1 receptor; M, mitosis; NRG1, neuregulin‐1; PI3K, phosphoinositide‐3 kinase; S, DNA synthesis phase. Adapted from Botting et al. (2012).

Proliferation and differentiation are usually conflicting cellular processes and thus molecules that promote differentiation, such as mitogen‐activated protein kinase (MAPK) p38, are also involved in the repression of cell cycle re‐entry (Liang & Molkentin, 2003). Overexpression of p38 also inhibits neonatal cardiomyocyte proliferation, whereas deletion of the gene or inhibition of its activity in cardiac tissue increases the number of cardiomyocytes undergoing mitosis (Fig. 4; Engel et al. 2005). Inhibition of p38 and treatment with fibroblast growth factor 1 (FGF1; Fig. 4) has also been implicated in the de‐differentiation of adult cardiomyocytes as well as enhanced cardiomyocyte proliferation and reduced apoptosis, resulting in improved cardiac function in post‐myocardial infarction adult rats (Cuevas et al. 1997; Engel et al. 2006). miR‐350 has been identified as an important regulator of p38 in cardiomyocyte cell culture, although this miRNA is not consistently expressed across fetal development in rats (Ge et al. 2013). Transcription factors such as GATA4 and Nkx2.5 are also involved in cardiomyocyte differentiation and hypertrophy, and their genes are strongly modulated by epigenetic mechanisms (Callis et al. 2009; Schlesinger et al. 2011). Cardiomyocyte proliferation can also be induced by neuregulin 1 (NRG1), which stimulates the epidermal growth factor receptor (ERBB2–ERBB4) heterodimer or ERBB4–ERBB4 homodimer receptors (Figs 4 and 5; Bersell et al. 2009). Activation of phosphoinositide 3‐kinase (PI3K) is a common mechanism through which NRG1 and FGF1, as well as insulin‐like growth factors (IGF1 and IGF2), promote cardiomyocyte proliferation. Treatment with NRG1 improved cardiac function and reduced scar tissue formation in adult mouse hearts 2 weeks post‐myocardial infarction (Bersell et al. 2009). FGF1 and IGF1R are two of the many targets regulated by miR‐195 (Wang et al. 2012, 2014), IGF‐1 is also regulated by miR‐1, one of the most abundant cardiac miRNAs (Elia et al. 2009). It is evident that the PI3K pathway is a powerful promoter of cellular proliferation and cytokinesis. Epigenetic manipulation of this pathway may prove a valuable strategy to improve the cardiac repair response after an infarction.

Figure 5. Regulation of cardiomyocyte proliferation through signalling of FGF1, NRG1 and IGF1.

Inhibition of the MAPK p38 in the presence of FGF1 or the activation of NRG1 signalling stimulates cardiomyocytes to re‐enter the cell cycle by activating PI3K causing DNA synthesis and subsequent cytokinesis. ERBB2, erb‐b2 receptor tyrosine kinase 2; ERBB4, erb‐b2 receptor tyrosine kinase 4; FGF1, fibroblast growth factor 1; FGFR, fibroblast growth factor receptor; IGF1, insulin‐like growth factor 1; IGF‐1R, insulin‐like growth factor 1 receptor; IGF2, insulin‐like growth factor 2; NRG1, neuregulin 1.

Transcriptional regulation of cardiomyocyte proliferation can also be controlled through the transcription factor Meis1 (Porrello & Olson, 2014). Meis1 is part of the TALE (three amino acid loop extension) family, and plays an important role in embryonic cardiac morphogenesis (Stankunas et al. 2008; Wamstad et al. 2012), as well as cardiac terminal differentiation (Wamstad et al. 2012). However, in the adult heart, Meis1 is associated with cell cycle arrest, and deletion of the Meis1 gene in the postnatal mouse heart causes a significant increase in cardiomyocyte proliferation (Mahmoud et al. 2013). qRT‐PCR and transcriptional analyses have also revealed that Meis1 is an upstream regulator of cell cycle exit at birth, and is required for the activation of CDK inhibitors p15, 16 and 21 (Fig. 6; Mahmoud et al. 2013). miR‐196b has been implicated as a key regulator of Meis1 in cancer cell lines (Li et al. 2012); however, the concentration of miR‐196b must be tightly controlled, with overexpression silencing secondary cell cycle regulator gene transcripts (Li et al. 2012). The ability to manipulate the Meis1 pathway to control cardiomyocyte proliferation or assist in cardiac regeneration has yet to be explored in detail.

Figure 6. Regulation of cardiomyocyte proliferation via β‐catenin signalling.

Activation of the Hippo pathway involves SAV, mammalian STE20‐like protein kinase 1 (MST1), MST2, MOB, and large tumour suppressor 1 and 2 (LATS1/2) and results in the phosphorylation and inactivation of the transcriptional co‐activator YAP impeding cardiomyocyte proliferation. Unphosphorylated YAP activates the WNT signalling pathway by interacting with β‐catenin, and consequently represses the Hippo kinases LATS2 and MST1/2 or the upstream scaffold protein SAV promotes cardiomyocyte proliferation. Suppression of YAP impedes cardiomyocyte proliferation, whereas overexpression of YAP stimulates cardiomyocyte proliferation via the triggering of pro‐growth signalling pathways such as WNT and insulin‐like growth factor 1 (IGF1). Axin, APC and GSK3β are components of the complex that phosphorylates β‐catenin to stimulate its degradation. Dashed line shows that additional molecules are involved. APC, adenomatosis polyposis coli; AXIN, axis inhibition protein; DVL, dishevelled; GPCR, G protein‐coupled receptor; GSK3β, glycogen synthase kinase 3β; IGF1R, IGF1 receptor; MOB, Mps one binder; SAV, Salvador; TCF‐LEF, T cell factor ‐ lymphoid enhancer‐binding factor; (P) represents phosphorylation. Adapted from Xin et al. (2013).

The Hippo signalling pathway is highly evolutionarily conserved and has an important role in regulating cell proliferation and organ size (Pan, 2010). This pathway involves a sequence of adaptors and kinases that promote phosphorylation of transcriptional co‐activator Yes‐associated protein (YAP) into its inactive form (Fig. 6). Suppression of YAP to its inactive form plays a major role in regulating organ size during development, and impairment of its inhibition has been implicated in cancer development (Bae et al. 2017). In its active form, YAP stimulates cell cycle progression by interacting with transcription factors, such as TEA domain family members (TEAD), RUNT‐related transcription factor 4 (RUNX4), SMAD1, T‐box 5 (TBX5) and p73 (Pan, 2010; Zhao et al. 2011), as well as upregulating proliferative genes such as CTGF and FGF1 (Zhao et al. 2010). Genetic loss of function or ablation of scaffold protein Salvador (SAV), a crucial protein in the Hippo pathway, results in perinatal lethality in mice due to extreme enlargement of the heart (Heallen et al. 2011). However, this heart enlargement is not due to cardiomyocyte hypertrophy, but rather excessive cardiomyocyte proliferation. Interestingly, WNT target genes, including nuclear β‐catenin, are also upregulated in SAV deleted hearts, suggesting a connection between the Hippo and WNT signalling pathways (Heallen et al. 2011; Fig. 6). Decreasing the concentration of β‐catenin rescues the increased cardiomyocyte proliferation that is observed in SAV knockout hearts, suggesting that the Hippo pathway negatively regulates WNT signalling to control heart growth and size (Heallen et al. 2011). Of all the components in the Hippo pathway, YAP is the most essential to regulate cardiomyocyte proliferation. Genetic loss of YAP function in the embryonic heart tissue is lethal due to reduced cardiomyocyte proliferation. Conversely, enhanced YAP function stimulates cardiomyocyte proliferation (Xin et al. 2011; von Gise et al. 2012). Consequently, YAP connects Hippo signalling and other growth promoting pathways, such as IGF and WNT signalling, to control embryonic heart development (Fig. 6). Interestingly, miR‐206 has been identified as a crucial regulator in YAP signalling, with overexpression of miR‐206 inducing cardiac hypertrophy (Yang & Sadoshima, 2011). Whether YAP can promote the proliferation of adult cardiomyocytes and enhance adult heart regeneration in response to injury remains unclear.

miRNAs are essential for regulation of cardiac development

The expression of non‐coding RNAs, such as miRNAs, plays an important role in the epigenetic machinery and is vital for normal tissue development and function (Lock et al. 2017). miRNA expression acts to repress gene expression by interfering with target gene mRNA translation or stability (Lock et al. 2017). Regulatory pathways that control cardiomyocyte proliferation in fetal development are modulated by numerous miRNAs, and have been identified by tracking expression of miRNA over the period of cardiomyocyte quiescence near birth (Liu & Olson, 2010; Small & Olson, 2011). For example, the inhibition of miR‐15 family members using locked nucleic acid (LNA) modified anti‐miRNAs leads to increased mitosis of cardiomyocytes in neonatal mice, and promotes adult cardiomyocyte proliferation, improved cardiac function after myocardial infarction and improved contractile function after ischaemia–reperfusion injury (Hullinger et al. 2012; Porrello et al. 2012). Among the other members of the miR‐15 family are miR‐15a, miR‐15b, miR‐16‐1, miR‐16‐2, miR‐195 and miR‐497, which regulate the expression of many rodent genes involved in cell cycle progression, including Crm1, Chek1, Cdc2a, Birc5, Spag5, Mps1, Pgam1, Hand1, Cdk9, c‐myc, Cyclin D1, Cyclin D2, Serum Response Factor (SRF) and Connexin‐43 (Liu et al. 2008; Porrello et al. 2011a ; Hullinger et al. 2012; Porrello et al. 2012; Yin et al. 2012). miR‐195 is a member of the miR‐15 family and is highly upregulated at postnatal day 10 in the mouse when cardiomyocytes have exited the cell cycle. Overexpression of miR‐195 in the developing murine heart also causes cardiomyocyte cell cycle arrest at the G₂ phase and leads to hypoplasia and congenital heart abnormalities (Porrello et al. 2011a ). In addition, overexpression of miR‐195 prevents heart regeneration post‐myocardial infarction in postnatal day 1 mouse hearts when regeneration of heart tissue is still possible. miR‐195 may function through suppression of checkpoint kinase 1 (Chek1) mRNA, a gene that encodes a protein kinase that promotes the G₂ to M phase transition and mitotic progression (Fig. 7; Porrello et al. 2011a ). The miR‐15 family therefore negatively regulates cardiomyocyte mitosis, and upregulation of members of this miRNA family shortly after birth may contribute to cell cycle arrest in mice (Fig. 7). Inhibition of members of the miR‐15 family in large animals may therefore garner the same improved regenerative response after infarction as demonstrated in these small animal models.

Figure 7. miRNAs can negatively (red) or positively (green) regulate cardiomyocyte proliferation.

The miR‐15 family, including miR‐195 and miR‐133, inhibit the cell cycle and therefore cardiomyocyte proliferation by suppressing genes that stimulate cell cycle progression such as cyclin‐dependent kinases, checkpoint kinase 1 and fibroblast growth factor receptor. miR‐652, miR‐25, miR‐208a and the miR‐34 family act to suppress cardioprotective genes following myocardial infarction. miR‐590‐3p, miR‐199a‐3p, and miR‐17‐92 and miR‐302‐367 clusters stimulate cardiomyocyte proliferation by suppressing the expression of genes that inhibit cell proliferation such as HOMER1, HOPX and CLIC5 and members of the Hippo pathway (Montgomery et al. 2011; Porrello et al. 2011a ; Bernardo et al. 2012, 2014; Eulalio et al. 2012; Chen et al. 2013; Wahlquist et al. 2014; Tian et al. 2015).

miRNAs that may induce DNA synthesis and increase cytokinesis in cardiomyocytes include miR‐590‐3p and miR‐199a‐3p (Fig. 7; Eulalio et al. 2012; Ali, 2013; Wang & Martin, 2014). These miRNAs increase cardiomyocyte but not cardiac fibroblast proliferation, reduce fibrotic scar size and improve cardiac function post‐myocardial infarction in neonatal mice (Eulalio et al. 2012). miR‐590‐3p and miR‐199a‐3p share several of the same targets including Homer1, which encodes a protein that modulates Ca²⁺ signalling in the heart through interaction with the Ca²⁺ release channel ryanodine receptor (RyR) (Fill & Copello, 2002). In addition, both miRNAs also target the mRNA of HOP homeobox (HOPX), which encodes a homeodomain protein that inhibits cardiomyocyte proliferation (Mariotto et al. 2016). A further target for miR‐590‐3p may also be chloride intracellular channel 5 (CLIC5), an inhibitor of cell proliferation (Ponnalagu et al. 2016). Therapeutic reagents that target the mechanisms by which these miRNAs function may therefore be utilized to promote cardiomyocyte re‐entry into the cell cycle, reduce scar formation and improve cardiac function following injury.

Altering miRNA as an intervention for myocardial infarction

miRNA repression

Agents that act as anti‐miRs of specific cardiac miRNAs that have been identified as having roles in proliferation could be utilized to promote cardiomyocyte regeneration. One of these miRNAs of interest is miR‐133, which is downregulated during the period of cardiomyocyte regeneration and proliferation in injured zebrafish myocardium implicating it in the regulation of cell cycle progression (Yin et al. 2012). Of the three genomic loci for miR‐133, only miR‐133a‐1 and miR‐133a‐2 are expressed in the heart (McCarthy, 2008), and both are vital for normal cardiogenesis in mice via the regulation of SRF (Liu et al. 2008). miR‐133 knockout mice also show increased cardiomyocyte proliferation, as well as increased Srf and Cyclin D2 expression (Liu et al. 2008). Interestingly, miR‐133 and its targets SRF and Cyclin D2, increase throughout late gestation in the sheep fetus (Morrison et al. 2015), indicating species differences in miR‐133 expression profiles. Other mRNA targets for miR‐133a include connective tissue growth factor (CTGF; Duisters et al. 2009; Gravning et al. 2012, 2013), Connexin‐43 (GJA1; Yin et al. 2012), phosphoglyceric acid mutase (PGAM1; Humphreys et al. 2012) and insulin‐like growth factor receptor 1 (IGF1R), which is a powerful factor involved in stimulating cardiomyocyte proliferation before birth (Fig. 4; Sundgren et al. 2003; Gao et al. 2014). By tracking miRNA expression across the period that mouse cardiomyocytes lose the ability to proliferate (4–10 days after birth; Porrello et al. 2011b ), miR‐195 (as previously described, another member of the miR‐15 family) was identified as the most upregulated miRNA across this period, and is likely involved in the transition to quiescence (Porrello et al. 2011a ). Inhibition of miR‐195 increases the number of mitotic cardiomyocytes in mice (Porrello et al. 2011a , 2012), indicating roles for this miRNA in the suppression of cytokinesis. In sheep, there is a species difference compared to mouse with comparatively more variable expression of many members of the miR‐15 family across gestation, though expression of their target genes decrease across late gestation, as expected (Morrison et al. 2015). This variable expression may therefore represent fine tuning of the ratio of each member of this miRNA family to target different mRNAs at particular developmental ages, as well as a reduced reliance on this single miRNA family for suppression of proliferation (Morrison et al. 2015). Using anti‐miR to inhibit the miR‐15 family, miR‐133, as well as other miRNAs that are upregulated at birth, may promote cardiomyocyte proliferation and regeneration through reduced suppression of cell cycle promoters.

A number of miRNAs have a role in modulating the response to heart failure. Inhibition of the miR‐34 family may have therapeutic potential for cardiac pressure overload and myocardial infarction. The miR‐34 family has shown increased expression in the mouse heart after myocardial infarction (Lin et al. 2010), and there was increased expression of miR‐34a, miR‐34b and miR‐34c in cardiac tissue from patients with heart disease (Thum et al. 2007; Greco et al. 2012), cementing its role as a central miRNA in cardiac disease. Inhibition of the miR‐34 family has shown therapeutic potential where cardiac remodelling was attenuated and improved cardiac function in mouse models of both pressure overload and myocardial infarction with upregulation of growth factor target genes including VEGF, VCL, Sirt1, Notch1 and Pofut1 (Bernardo et al. 2012; Boon et al. 2013). In a related study, the therapeutic potential of miR‐652 was investigated in the mouse model of cardiac pressure overload (Bernardo et al. 2014). Inhibition of miR‐652 reduced fibrosis and apoptosis, attenuated cardiac hypertrophy, was associated with improved heart function and increased the expression of the mRNA target, Jagged1, which has been implicated in promoting cell survival and angiogenesis (Bernardo et al. 2014). During heart failure, there is a reduction of cardiomyocyte calcium handling and activation of calcineurin and nuclear factor of activated T cells (NFAT) signalling (Wahlquist et al. 2014). miR‐25 has been identified as a major regulator of calcium homeostasis in cardiomyocytes via the regulation of SERCA2a mRNA (Wahlquist et al. 2014). Inhibition of miR‐25 in the mouse model of heart failure was associated with increased survival rates, improved cardiac function and decreased fibrosis (Wahlquist et al. 2014). However, this result contradicts the finding that inhibition of miR‐25 causes spontaneous cardiac dysfunction through the gene Hand2 (Dirkx et al. 2013). Another miRNA involved in the loss of cardiac contractility is miR‐208a, which when inhibited improves survival, promotes cardiac function and prevents remodelling in hypertensive rats (Montgomery et al. 2011) amongst other non‐cardiac specific effects (Grueter et al. 2012). Given the role of this subset of miRNA in hypertension, cardiac dysfunction, pressure overload and cardiac remodelling, the therapeutic potential of manipulating their expression may extend to improving the long term clinical outcomes after infarction.

miRNA mimicry

Cardiomyocyte proliferation may also be induced via the use of miRNA mimics, synthetic molecules that target gene transcripts involved in suppression of the cell cycle. miR‐199a and miR‐590 have been implicated in promoting cardiomyocyte proliferation in rodents (Eulalio et al. 2012), with adult rat cardiomyocytes treated with mimics for miR‐199a and miR‐590 showing increased cell cycle re‐entry and cardiac regeneration (Eulalio et al. 2012). Inhibition of target genes of miR‐590 and miR‐199a, such as Clic5, Hopx and Homer1 (Fig. 7), using short interfering RNA results in approximately double the number of cardiomyocytes undergoing DNA synthesis and significantly increased cytokinesis (Eulalio et al. 2012). Interestingly, unlike in rodents, these miRNAs that promote proliferation have variable expression profiles across sheep development (Morrison et al. 2015). This result indicates that there may be interplay with numerous other miRNAs to promote cardiomyocyte proliferation, such as members of the miR‐17‐92 cluster, which are also potential promoters of the cell cycle in mice and sheep that are downregulated at birth (Chen et al. 2013; Morrison et al. 2015). In addition, treatment of both miR‐199a and miR‐590 failed to have any effect on the proliferation or contractile force of human pluripotent stem cell‐derived cardiac organoids (Mills et al. 2017). The miR‐302 to miR‐367 cluster (miR‐302a, miR‐302b, miR‐302c, miR‐302d and miR‐367) are another set of miRNAs that may be involved in cardiac regeneration. This cluster, which has been demonstrated to have decreased expression after birth, is crucial for cardiomyocyte proliferation during development and targets members of the aforementioned Hippo pathway, e.g. Mst1, Lats2 and Mob1b (Fig. 7; Tian et al. 2015). Utilizing miR mimics for the miRNA in this cluster has demonstrated enhanced cardiac regeneration/proliferation and improved cardiac function in the mouse model of myocardial infarction (Tian et al. 2015). Treating heart tissue with miRNA mimics for miR‐199a, miR‐590, and members of the miR‐17‐92 and miR‐302‐367 clusters, as well as other miRNAs that are downregulated at birth, may therefore also promote improved cardiomyocyte proliferation and regeneration after injury in large animals including humans via increased suppression of cell cycle inhibitors. However, a much greater understanding of the similarities and differences in miRNA expression and their targets in small and large animals is required for potential development of therapeutic interventions.

Conclusion

Since the number of cardiomyocytes a human has for life is largely set at birth and the adult human heart has little capacity to repair, the lack of effective regenerative treatments necessitates new approaches to address cardiac damage. Given that miRNAs have an important role in the regulation of genes involved in cardiac cell cycle progression and regeneration, it is essential to study their roles in humans and large mammals, as the timing of heart development in relation to birth is different when compared to that in mice and rats. Many promising therapeutic targets have arisen from analysis of small animal models. However, follow‐up studies are required in large animals to confirm these have positive outcomes for humans, before the treatments can be translated into the clinic. Potential target miRNAs can be both inhibited (miR‐15 family, miR‐34 family, miR‐652, miR‐25 and miR‐208) or mimicked (miR‐199a, miR‐590, miR‐17‐92 cluster and miR‐302‐367 cluster) to improve the response to infarction. In addition, rapidly advancing technologies are being developed to make the quantification and targeting of small molecules such as miRNA more affordable and easily accessible. These technological advances may provide for innovative small‐RNA discoveries over the coming years allowing further identification of key miRNAs that can be targeted using mimics or inhibitors to condition an improved response to myocardial infarction.

Additional information

Competing interests

The authors have no conflict of interest to disclose.

Author contributions

M.C.L. and J.L.M. were responsible for the conception and design of the article. M.C.L., R.L.T., K.J.B., K.C.W.W., J.B.S., D.A.B., M.S. and J.L.M. were involved in analysis and interpretation of the data. M.C.L., R.L.T., D.A.B. and J.L.M. drafted the article. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

J.L.M. was funded by a NHMRC Career Development Fellowship (APP1066916) and an Australian Research Council Future Fellowship (Level 3; FT170100431). M.C.L. was funded by an Australian Government Research Training Program scholarship. K.C.W.W. was funded by a NHMRC Early Career Research Fellowship (APP1090888).

Biographies

Janna L. Morrison is Head of the Early Origins of Adult Health Research Group in the School of Pharmacy and Medical Sciences at the University of South Australia, Adelaide, Australia. She completed her PhD in 2001 at the University of British Columbia and is currently funded by an Australian Research Council Future Fellowship. Her work focuses on examining the link between low birth weight and heart disease in adulthood. She is a fellow of the Cardiovascular Section of the American Physiological Society (2015).

Mitchell C. Lock is a PhD student in the Early Origins of Adult Health Research Group in the School of Pharmacy and Medical Sciences at the University of South Australia, Adelaide, Australia. His research is focused on heart and lung development, the epigenetic regulation of cardiac tissue after heart attack and cardiac regeneration.