Hypoxia-induced myocardial regeneration

Abstract

The underlying cause of systolic heart failure is the inability of the adult mammalian heart to regenerate damaged myocardium. In contrast, some vertebrate species and immature mammals are capable of full cardiac regeneration following multiple types of injury through cardiomyocyte proliferation. Little is known about what distinguishes proliferative cardiomyocytes from terminally differentiated, nonproliferative cardiomyocytes. Recently, several reports have suggested that oxygen metabolism and oxidative stress play a pivotal role in regulating the proliferative capacity of mammalian cardiomyocytes. Moreover, reducing oxygen metabolism in the adult mammalian heart can induce cardiomyocyte cell cycle reentry through blunting oxidative damage, which is sufficient for functional improvement following myocardial infarction. Here we concisely summarize recent findings that highlight the role of oxygen metabolism and oxidative stress in cardiomyocyte cell cycle regulation, and discuss future therapeutic approaches targeting oxidative metabolism to induce cardiac regeneration.

the adult heart is one of the least regenerative organs in mammals, and therefore significant cardiac injury often results in heart failure. In sharp contrast to adult mammals, some vertebrate species, including zebrafish (64) and urodele amphibians (9, 55, 56), have a remarkable cardiac regeneration capacity throughout their lifetime (reviewed in 11, 43). In addition, recent reports demonstrate that mammals can regenerate their heart during embryogenesis (16, 70), and for a short period of time after birth (4, 8, 13, 42, 53, 61–63, 76, 77). Genetic lineage tracing demonstrated that the major source of newly formed cardiomyocytes is preexisting cardiomyocytes both in neonatal mice (62, 63, 76) and in zebrafish (29, 32). As expected, the limited postnatal time window of cardiac regeneration in neonatal mammals coincides with the cell cycle withdrawal and binucleation in a majority of cardiomyocytes (62, 63, 69). This temporal correlation between cardiomyocyte cell cycle withdrawal and diminishing cardiac regenerative capacity is also reported in humans (47). Although multiple studies reported genes and proteins that regulate postnatal polyploidy and binucleation of cardiomyocytes in mammals (1, 38, 54, 73, 79), upstream environmental cues that trigger the postnatal shift in regenerative capacity remain poorly understood.

Oxidative Metabolism and Postnatal Cardiomyocyte Cell Cycle Arrest

It is especially intriguing that across vertebrate species, there is a strong correlation between oxidative metabolism and proliferative capacity of cardiomyocytes (Fig. 1) given that in several tissue specific and pluripotent stem cells, mitochondrial metabolism and ROS level are critical regulators of proliferative capacity and the cell cycle (28, 33, 80). For example, fish and amphibians have two-chambered and three-chambered hearts, respectively, which result in the mixing of arterial and venous blood (20, 22). Similarly, the developing mammalian circulation carries relatively hypoxic blood, with arterial $P{a}_O{}_{{}_2}$ of 25–35 mmHg, primarily due to arteriovenous mixing in the shunt-dependent fetal circulation (12, 15, 17, 35, 36, 57). However, the $P{a}_O{}_{{}_2}$ rises rapidly to ~100 mmHg with the first breath due to shunt closure (71). Consequently, a drastic metabolic transition occurs in neonatal cardiomyocytes. Early physiological and biochemical studies showed that embryonic/neonatal cardiomyocytes rely primarily on glycolysis as an energy source, whereas postnatal cardiomyocytes utilize mitochondrial respiration as a main source of energy production to meet their high metabolic demands (3, 39–41). We recently examined enzymes related to glycolytic and mitochondrial metabolism in the early postnatal period by utilizing quantitative mass-spectrometry and showed that mitochondrial Krebs cycle enzymes and fatty acid oxidation enzymes are upregulated within 7 days after birth, with a synchronized downregulation of anaerobic glycolysis (66). This is consistent with previous findings demonstrating increased mitochondrial biogenesis in postnatal cardiomyocytes (27, 45, 58–60). As a result of the steep rise in mitochondrial energy production in the postnatal heart, there is an increase in ROS production and oxidative nuclear DNA damage in cardiomyocytes. We showed that activation of Wee1 kinase, a component of the DNA damage response pathway, directly triggers cell cycle arrest in postnatal cardiomyocytes (66). It is intriguing that mitochondrial DNA damage is also elevated in the postnatal heart (58), which may also play an important role in cardiomyocyte cell cycle regulation.

Fig. 1.

Comparison of cardiac regeneration and oxidative metabolism across vertebrate species.

Hypoxic Proliferative Cardiomyocytes in the Adult Heart

Recent advances in in vivo lineage tracing have shown that adult mouse cardiomyocytes are renewed at a slow rate, which is mainly mediated by cell division of preexisting cardiomyocytes (2, 5, 6, 44, 68). This indicates that a limited number of cardiomyocytes are still capable of proliferation in the adult heart. But if all cardiomyocytes experience increased oxidative stress after birth, how are some cardiomyocytes able to retain proliferative competency within the adult heart? We postulated that proliferative cardiomyocytes in the adult heart are less oxidative, not unlike those in embryos or zebrafish, and thus protected from oxidative DNA damage. To test this hypothesis, we sought to identify, and trace the lineage of hypoxic cardiomyocytes in the adult heart utilizing protein expression of hypoxia inducible factor 1 alpha (Hif-1α), a master regulator of cellular hypoxic response. Hif-1α protein is stabilized during hypoxia, resulting in transcriptional activation of multiple target genes involved in glycolytic and mitochondrial metabolism, antioxidant enzymes, cell cycle regulators, etc. (30, 65). In fact, Hif-1α protein is stabilized in hypoxic stem cells and required for the maintenance of their quiescent state (72), and in addition, is required for cardiac regeneration after injury in zebrafish (29) and cardiac development in fetal hypoxic cardiomyocytes (17) through regulation of cell cycle, stress response pathways and cellular metabolism (7, 23, 46). As a result of Hif-1α stabilization-based lineage tracing, we identified a rare cardiomyocyte population that retains embryonic/neonatal features, such as smaller cell size, mononucleation, lower levels of oxidative DNA damage, and contribution to cardiomyocyte renewal (34). In addition to having a lower capillary density, RNA sequencing analysis of isolated hypoxic proliferative cardiomyocytes suggests the existence of an intrinsic mechanism that mediates the expression of Hif-1α such as an upregulation of Hif-1α mRNA and downregulation of prolyl hydroxylases (34). Therefore, our results suggest that some adult proliferative cardiomyocytes are protected from oxidative stress, either by inhabiting a hypoxic microenvironment, or activating an innate safeguard mechanism to stabilize Hif-1α or both.

Hypoxia Induces Cardiac Regeneration in the Adult Heart

As we have outlined earlier, postnatal mitochondrial ROS production inhibits cardiomyocyte proliferation (34, 66). Therefore, we reasoned that long-term systemic hypoxemia might decrease mitochondrial respiration and mitochondrial ROS production, and thereby activate cell cycle reentry through reduction in oxidative DNA damage. To test this hypothesis, we exposed adult mice to hypoxia (7% oxygen, equivalent to summit of Mt. Everest) for 2 wk following a gradual decrease in oxygen tension by 1% per day from 20.9% ambient oxygen to 7% over 14 days. Given that change in the fraction of inspired oxygen does not necessarily correlate with a change in oxidative metabolism (21), it is critical to directly assess mitochondrial metabolism. Following hypoxia, we observed a reduction in mitochondrial metabolism as evidenced by a decreased mitochondrial DNA copy number, cristae density, and the protein expression of Krebs cycle and fatty acid beta-oxidation enzymes (51). In parallel, we observed reduced oxidative DNA damage in cardiomyocyte nuclei and cell cycle reentry in differentiated cardiomyocytes (51). Remarkably, chronic hypoxia also induced cardiomyocyte proliferation following myocardial infarction (MI), which was accompanied by significant recovery of LV systolic function (Fig. 2). It is important to note here that the effect of hypoxia on cardiomyocyte proliferation was almost completely abolished upon injection of the ROS generator diquat, indicating that the reduction in ROS and oxidative DNA damage plays a causative role in cardiomyocyte cell cycle reentry (51). There was no increase in mortality either in the MI group, or the sham group during 2 wk of exposure to 7% O₂ following a gradual decrease in oxygen concentration by 1% per day (51). However, mortality in both MI group and in sham-operated group increased during 3 wk of 7% oxygen exposure, suggesting that prolonged exposure to severe hypoxia is poorly tolerated. Even so, the surviving MI mice showed a significant improvement in LV systolic function following recovery from hypoxia to ambient oxygen (51). Moreover, we tested whether exposure to more moderate hypoxia (10% oxygen, in combination with the administration of mitochondrial-ROS scavenger mitoTEMPO) induces an improvement in LV function in MI mice. We found that this intervention resulted in prevention of LV remodeling, which suggests that there may be some beneficial effects to moderate degrees of hypoxia, although no significant cardiomyocyte proliferation was observed. It is noteworthy here that chronic hypoxia also enhanced vascular supply as evidenced by elevated coronary collateral formation and vasodilation in cardiac capillaries, consistent with previous observations (14). In agreement, we observed an increased cell division in nonmyocyte lineages, suggesting the contribution of nonmyocyte proliferation to improvement of cardiac function (51). This may contribute to the favorable effect on cardiac remodeling following exposure to hypoxia (51). These findings may seem counterintuitive given that myocardial ischemia causes cardiomyopathy through cardiomyocyte death (reviewed in 10, 18). However, the effect of gradual prolonged systemic hypoxemia, which can reduce mitochondria-derived oxidative damage in cardiomyocytes, has not been tested, and represents a fundamentally different scenario compared with acute ischemia or hypoxia where adaptive changes do not occur. Our findings indicate that hypoxia and targeting hypoxia signaling may represent novel therapeutic strategies for heart regeneration. It is important to note here that a forced-activation of hypoxia signaling by means of genetic ablations of genes encoding prolyl hydroxylases (PHD2, -3) (50) or von Hippel Lindau (VHL) (37) in cardiomyocytes resulted in some deleterious effects including heart failure or cardiomyopathy. This is potentially due to either dysregulation of metabolic state (37, 52) or the activation of Hif-independent pathways (78). Thus it is critical to better understand the precise signaling cascades of hypoxia-induced cardiomyocyte proliferation in order to develop new therapeutic approaches.

Fig. 2.

Long-term hypoxia induced functional recovery in the infarcted heart. *P < 0.05, **P < 0.01. [Modified from Nakada et al. (51) with permission.]

Clinical Implications and Conclusion

An important question with a view to the future clinical application is how beneficial, or detrimental, hypoxia is for patients with cardiomyopathy (2a). Epidemiological data indicate significantly lower incidence of coronary artery diseases in high altitude in the United States (48, 74), although another study showed no correlation between altitude and coronary heart diseases (49). A study in Switzerland with a relatively homogeneous population in terms of ethnicity and culture revealed a protective effect of altitude in cardiovascular mortality (19). In agreement with previous studies including our own, coronary vasodilation is also found in humans at high altitude (31), which may contribute to the inverse correlation between coronary disease incidence and altitude. Nevertheless, given a few studies that have assessed the safety of exposure of patients with cardiovascular diseases to low oxygen were mainly in the setting of mountain climb or trekking (24, 26, 67, 75) which is fundamentally different from a therapeutic setting, further studies are required to evaluate the effect of hypoxia on cardiovascular disease patients.

In conclusion, mounting evidence suggests that oxidative metabolism is a key regulator of proliferative competency of mammalian cardiomyocytes (Fig. 3). Although further studies are required for evaluation of safety and efficacy, hypoxia may be a new, albeit counterintuitive, strategy for treatment of cardiomyopathy. From a mechanistic standpoint, molecular mechanisms underlying environmental oxygen-dependent metabolic switch in cardiomyocytes remain poorly understood. In particular, upstream signaling pathways, transcription factors, and epigenetic modifiers involved in modulating the metabolic shift from an energy-demanding nonproliferative state to a regenerative state are important future therapeutic targets.

Fig. 3.

Oxygenation state and cardiac regeneration in mammals.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

W.K., Y.N., and H.A.S. prepared figures; W.K., Y.N., and H.A.S. drafted manuscript; W.K., Y.N., and H.A.S. edited and revised manuscript; W.K., Y.N., and H.A.S. approved final version of manuscript.