The vertebrate heart has undergone a significant degree of evolutionary changes. In the earliest vertebrates, the heart was a low-pressure contractile vessel, which later developed into a two-chamber heart in early fish. Further evolution in fish, amphibians, birds and mammals later resulted in septation, chamber separation and marked thickening of the ventricles with increased ventricular pressure. In addition to the morphological changes, cardiac evolution was associated with a several fold increase in heart rate, which is needed to sustain high cardiac output. Therefore, the mammalian heart evolved from a single low-pressure tube, to the complex, high-pressure, high heart rate, multi-chamber organ of adult mammals.

In the evolution literature, the concept of space constraint is hypothesized as a major determinant of loss of the ability of cardiomyocytes to divide. While this may certainly be a factor, it does not explain the permanent loss of cardiomyocyte proliferative ability. In other words, space constraint may in fact stop cardiomyocyte proliferation, but why would it prevent any ability of cardiomyocytes to divide even after substantial cardiomyocyte loss? Another argument against that concept is the fact that the mammalian heart increases in size during postnatal development by several folds through hypertrophic cardiomyocyte growth. This occurs following the permanent cell cycle arrest of cardiomyocytes, and therefore argues against the space constraint hypothesis. In other words, if the heart still has to grow several folds during postnatal development, why does it do so through hypertrophic rather than hyperplastic growth?

One of many factors shared by organisms that are capable of heart regeneration is the oxygenation state. For example, the zebrafish's stagnant and warm aquatic environment has 1/30th oxygen capacitance compared to air, and is prone to poor oxygenation (Rees et al., 2001; Roesner et al., 2006). Moreover, the zebrafish circulatory system is relatively hypoxemic, as it has a primitive two-chambers heart, which results in mixing of arterial and venous blood. Similarly, the mammalian fetal circulation is shunt-dependent with significant mixing of arterial and venous blood. Although blood in the umbilical vein going to the fetus is 80–90% saturated with a PaO₂ of 32–35 mmHg, the blood ejected from the left ventricle is only 65% saturated with a PaO₂ of 25–28 mmHg (Dawes et al., 1954), which is quite hypoxemic compared to the postnatal circulation with a saturation above 95% and a PaO₂ of 100 mmHg. Therefore, both the zebrafish and mammalian fetal heart reside in relatively hypoxic environments; however transition from embryonic- to postnatal-circulation soon after birth drastically changes the oxygenation state of cardiomyocytes (Puente et al., 2014). In parallel to the oxygenation state, energy metabolism of the embryonic and adult heart is quite distinct. During embryonic development, when cardiomyocytes rapidly proliferate, the relatively hypoxic embryonic heart utilizes anaerobic glycolysis as a main source of energy (Fisher et al., 1980; Lopaschuk et al., 1992), whereas adult cardiomyocytes utilize the oxygen-dependent mitochondrial oxidative phosphorylation as an energy source (Wisneski et al., 1985; Gertz et al., 1988).

The energy advantage of mitochondrial oxidative phosphorylation over glycolysis is not without deleterious consequences, as the mitochondrion is considered the major source of free radical production (Harman, 1972; Nohl and Hegner, 1978; Miquel et al., 1980; Turrens, 1997, 2003). Our studies show that there is a significant increase in ROS in cardiomyocytes from day 1 to day 7 after birth. Mitochondrial ROS are generated as a consequence of electron leak by the electron transport chain (Rudolph and Heyman, 1974; Koopman et al., 2010) and can cause cellular toxicity by promoting damage of proteins, lipids or DNA, such as oxidized base, single- or double-strand breaks, resulting in cell cycle arrest, apoptosis or cellular senescence (Moos et al., 2000; Marnett et al., 2003; Hoeijmakers, 2009). We hypothesize that oxidative DNA damage might increase in cardiomyocytes postnatally and play a role in postnatal cell cycle arrest. In order to test this, we assessed oxidative base modification of DNA and showed that one of the oxidized bases in DNA, 8-oxo-7,8-dihydroguanine (8-oxoG), was undetectable at p0, but significantly increased at P7 and P14. In concert, the DNA damage response (DDR) pathway was also significantly activated as indicated by upregulation of Ser1981 phosphorylated-Ataxia Telangiectasia Mutated (p-ATM) kinase, an upstream kinase that activates multiple components of the DDR pathway, at P7 and P14. These results demonstrate that the increase in mitochondrial respiration corresponds temporally with an increase ROS in the neonatal heart and activation of DDR. Our published results also suggest that mitochondrial ROS mediated activation of the DDR is an important upstream event that mediates cell cycle arrest of postnatal cardiomyocytes, where mitochondrial catalase (mCAT) overexpression in cardiomyocytes decreases oxidative DNA damage and prolongs the postnatal window of cardiomyocyte proliferation (Puente et al., 2014). However, the factors that mediate increased mitochondrial ROS production and the subsequent oxidative DNA damage are not well understood.

Recent studies using data from the nuclear fallout during the cold war by using ¹⁴C dating to measure cardiomyocyte age indicated that turnover rates of human cardiomyocytes are about 1% per year at the age of 25 (Bergmann et al., 2009). These rate further decrease to 0.45% around 75 years of age, and approximately 45% of cardiomyocytes are replaced over the normal human lifespan (Bergmann et al., 2009). Importantly, a recent study described the dynamics of cardiomyocyte turnover in the adult mammalian heart using a novel multi isotope mass spectrometry (MIMS), and showed that the newly formed cardiomyocytes are derived from preexisting cardiomyocytes rather than a progenitor or stem cell population (Senyo et al., 2013). Intriguingly, the rate of cardiomyocyte turnover is similar to what has been previously described in humans (Bergmann et al., 2009).

Since cell cycle exit of cardiomyocytes shortly after birth is mediated by an increase in mitochondrial-derived ROS leading to oxidative DNA damage, we hypothesized that the population of cycling cardiomyocyte in the adult heart is protected from oxidative DNA damage by residing in hypoxic microenvironments, not unlike other cycling cells in the hematopoietic stem cell niche, or the hippocampus of the brain. In these hypoxic niches, stabilization of hypoxia-inducible factor-1 alpha (Hif-1α) is critical for their maintenance and function. Therefore, we reasoned that stabilization of Hif-1α identifies cycling cardiomyocytes within the uninjured adult heart. We generated a transgenic mouse that expresses a fusion protein where the oxygen dependent degradation (ODD) domain of Hif-1α is fused to the tamoxifen inducible CreERT2 under the cardiomyocyte-specific MHC promoter, thereby fate mapping hypoxic cardiomyocytes and their progeny after tamoxifen administration. Shortly after the tamoxifen pulse, we identified a rare population of hypoxic cardiomyocytes that contributed widely to new cardiomyocyte formation in the adult heart. Intriguingly, these hypoxic cardiomyocytes displayed characteristics of fetal/neonatal cardiomyocytes such as smaller cell size, mononucleation and absence of oxidative DNA damage. These findings support the hypothesis that cardiomyocyte turnover in the postnatal mammalian heart is inhibited by oxidative metabolism and the resulting oxidative DNA damage.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.