The regulation of cardiomyocyte proliferation is important for heart development and regeneration. The proliferation patterns of cardiomyocytes are closely related to heart morphogenesis, size, and functions. The proliferation levels are high during early embryogenesis; however, mammalian cardiomyocytes exit the cell cycle irreversibly soon after birth. The cell cycle exit inhibits cardiac regeneration in mammals. On the other hand, cardiomyocytes of adult zebrafish and probably newts can proliferate after cardiac injury, and the hearts can be regenerated. Therefore, the ability to reproliferate determines regenerative ability. As in other cells, the relationship between proliferation and differentiation is very interesting, and is closely related to cardiac development, regeneration and homeostasis. In this review, these topics are discussed.
The heart begins to function in the early stages of embryonic development, supporting embryonic and postnatal lives. The regulation of proliferation in cardiomyocytes (CMs) is important for development, functions and regeneration of the heart.
The regulation of proliferation in CMs is required for normal morphogenesis and determines the appropriate heart size, which is necessary for pumping appropriate volume of blood during development. In mammals, cell proliferation and hypertrophy in CMs increase the heart size before and after birth, respectively. Therefore, the regulation of proliferation in CMs is important at least during the embryonic stages of mammals. In addition, the sizes at birth probably influence the sizes in subsequent stages. How is the proliferation of CMs regulated? How is heart size determined? CMs exit from their cell cycle after birth. Recently, evidence that adult mammalian CMs can proliferate has been increasing. However, the ratio is very low. How do cells exit the cell cycle and is the exit maintained in the majority of CMs? What is the significance?
The hearts of neonatal mice and adult zebrafish can regenerate after injury through proliferation of pre-existing CMs (Jopling et al. 2010; Kikuchi et al. 2010; Porrello et al. 2011b). However, adult mammals are not capable of cardiac regeneration because cell cycle arrest is strictly maintained in a majority of CMs. What determines the differences among species?
In general, cell proliferation and differentiation show an inverse relationship during development. Is there such a relationship in CMs? If yes, how is the relationship regulated?
In this review, these topics are discussed mainly from the point of view of cell cycle regulation, and with consideration of recent studies in addition to future perspectives.
The cell cycle system
The cell cycle, by which cells can divide, is composed of four phases (Fig. 1): S-phase, in which DNA replication occurs; M-phase, where karyokinesis and cytokinesis occur; and G1 and G2-phases, in which cells prepare for subsequent phases, for example, RNA and protein synthesis, and check the cell status for entry to subsequent phases. The cell cycle is tightly regulated by the cell cycle machinery.
The central cores of the cell cycle machinery are cyclins and cyclin-dependent kinases (CDKs) (Fig. 1). CDKs are activated by complex formation with cyclins, following which they phosphorylate complex-specific substrates. The sequential activation of CDKs advances the cell cycle. In mammals, the complexes are cyclin D-CDK4/6 (G1-phase), cyclin E-CDK2 (G1/S-phase), cyclin A-CDK2/1 (S/G2-phase), and cyclin B-CDK1 (M-phase). The protein synthesis and degradation of these main cyclins determine the timing of activation of the partner CDKs. In addition, CDK activities are regulated by CDK inhibitors; INK4 family (p15, p16, p18 and p19) and Cip/Kip family (p21, p27 and p57).
Proliferation pattern of CMs during development and adult stages
The proliferation patterns of CMs have been studied in various animals. In mammals, these patterns in mice and rats have been reported in several reports (Rumyantsev 1977). Here, we explain the pattern in mouse (Fig. 2).
Proliferation levels of mouse CMs are high during early embryogenesis and start to decrease around embryonic day 10–12 (E10–12) (Erokhina 1968; Toyoda et al. 2003). Many CMs undergo mitosis without cytokinesis, resulting in binucleated CMs during first the 2 weeks in postnatal stages (Soonpaa et al. 1996), following which CMs exit the cell cycle.
The expression, complex formation, and activity of main cyclins and CDKs changed in a synchronous manner during embryonic and postnatal stages (Ikenishi et al. 2012). These levels decreased from midgestation to birth, and then showed one wave in which the peak was around postnatal day 5 (P5). These analyses and analysis of the cell cycle distribution patterns in mono- and binucleated CMs indicated that the wave of cell cycle regulator expression or activities during postnatal stages mainly produced binucleated cells from mononucleated cells (Ikenishi et al. 2012).
These studies raised the following questions about the mechanisms of the CM proliferation pattern (Fig. 2). (i) How do CMs slow down their proliferation? (ii) how do CMs exit their cell cycle? (iii) how do CMs maintain their cell cycle exit? and (iv) how many CMs can proliferate in adult stages? Answers to these questions are important to create and maintain the CM proliferation pattern. In addition, other interesting questions are: (v) what is the biological significance of the cell cycle exit in many CMs ? and (vi) how are binucleated CMs produced and what is the biological significance? These are discussed in the following sections.
How do CMs slow down their proliferation?
Cars cannot stop suddenly, and must brake in enough time before stopping at their goal. Similarly, cells must slow down their cell proliferation in sufficient time prior to cell cycle arrest or exit. The proliferation level of mouse CMs starts to decrease around E10–12 (Erokhina 1968; Toyoda et al. 2003). What triggers the decrease?
jumonji (jarid2) mutant mice showed hyperproliferation and overexpression of cyclin D1 in CMs at E10 (Takeuchi et al. 1999; Toyoda et al. 2003), and molecular and genetic analyses indicated that jumonji brakes CM proliferation through repression of cyclin D1 expression (Toyoda et al. 2003). The jumonji gene was originally identified by a mouse gene trap strategy (Takeuchi et al. 1995), and has essential roles in the development of multiple tissues (Takeuchi et al. 2006). The Jumonji protein has several conserved domains. As such domains, a jmjC domain catalyzes histone demethylation, and many histone demethylases have been identified in the jmjC family proteins (Takeuchi et al. 2006). Interestingly, Jumonji would not have a histone demethylase activity; however, Jumonji shows transcriptional repressor activity by recruiting histone H3–K9 methyltransferases, G9a and GLP, to the cyclin D1 promoter (Shirato et al. 2009). We also showed that a Jumonji–cyclin D1 pathway is required for the precise coordination of cell cycle exit and migration during neurogenesis in the mouse hindbrain (Takahashi et al. 2007).
Intriguingly, mutant mice of Tsc1 and Tsc2, whose gene products are factors in the TOR pathway, exhibit similar phenotypes to those in CMs of jumonji mutant mice (Kobayashi et al. 1999, 2001), suggesting that Tsc1/2 and the TOR pathway function as the triggers. It is very interesting and important to know what determines the timing when jumonji and Tsc1/2 work, and the interaction between these molecules.
How do CMs exit their cell cycle?
After deceleration, cars stop at the goal by applying the brakes hard. Similarly, some systems are necessary for cell cycle arrest or exit. Almost all mouse CMs exit their cell cycle at P5–14. What triggers the cell cycle exit? Analysis of expressions and activities of the cyclin E-CDK2 and cyclin A-CDK1/2 complexes showed that CDK activities start to be inhibited at P5 before the protein levels start to decrease at P10 (Ikenishi et al. 2012), suggesting that factors such as CKIs inhibit CDK activities, and contribute to the cell cycle exit. In fact, the expression of two CKIs, p21Cip1 and p27Kip1, showed a peak around P5, when cyclin E- and cyclin A-CDK activities start to decrease. In addition, postnatal CMs in p21Cip1 and p27Kip1 knockout mice showed failure in the cell cycle exit at G1-phase (Tane et al. 2014).
A transcriptional factor, Meis 1, also regulates the cell cycle exit of CMs (Mahmoud et al. 2013). Proliferation CMs in Meis 1 KO mice are prolonged at postnatal stages. Interestingly, Meis 1 activates expression of CKIs, p15ink4b, p16ink4a and p21Cip1 (Mahmoud et al. 2013). Together with the results for p21Cip1 KO mice, these results suggest that Meis 1 triggers the cell cycle exit through activation of p21Cip1 transcription. The requirement for p15ink4b and p16ink4a for the cell cycle exit of CMs, the detailed molecular mechanisms of Meis1 for p21Cip1, and the upstream system for p27Kip1 are still unknown.
Recently, the regulation of cell cycle exit by micro RNAs has been reported. For example, miR-195 has been suggested to downregulate proliferation through repression of genes such as Chek1 (Porrello et al. 2011a). Eulalio et al. performed functional screening for human micro RNAs, which promote proliferation. They showed 40 micro RNAs promoted mouse neonatal CM proliferation. Interestingly, miR-590 and miR-199a promoted cell cycle re-entry in adult CMs as well (Eulalio et al. 2012). These micro RNAs might be applicable as drugs for the treatment of CM loss, because they are activators but not repressors of CM proliferation. On the other hand, the contributions of these micro RNAs to cell cycle exit of CMs are interesting. The expression patterns of these micro RNAs during development and loss of function analysis will be helpful to know the contribution.
How do CMs maintain their cell cycle exit?
We use parking brakes to keep the cars stationary. What do cells use for the maintenance of cell cycle exit? In mice, all expression and activation levels of the main cyclin-CDK complexes become extremely low or undetectable after P14, and the levels are maintained for life (Ikenishi et al. 2012). The cell cycle of almost all CMs is arrested in G1-phase (Erokhina 1968; Ikenishi et al. 2012). These results suggest the possibility that downregulation of the expression of G1 cyclin–CDKs triggers downregulation of the expression of all other main cyclin–CDKs, and all downregulation is maintained for life by unknown mechanisms. Maintenance of the downregulation is most likely required for the maintenance of cell cycle exit of CMs. In fact, expression of only one G1 cyclin expression was found to cause cell cycle entry in many differentiated adult CMs (Tane et al., unpubl. data).
If this assumption is correct, what triggers the downregulation? Probably, p21Cip1, p27Kip1 and Meis 1, described in (2), are closely related as upstream signals; however, the mechanisms are unknown. It is also interesting how the downregulation is maintained. Because adult CMs can proliferate in mammals in very limited numbers (Soonpaa & Field 1998), the downregulation might be strictly maintained. Epigenetic regulation may be one of the mechanisms. The involvement will be elucidated by analysis such as histone modification and DNA methylation.
Interestingly, neuregulin induced cell cycle re-entry mainly in mononucleated CMs of adult mice (Bersell et al. 2009), suggesting that the downregulation of main cyclin–CDK expression can be unleashed in a portion of mononucleated CMs by activation of a neuregulin-ErbB4 pathway. Whether these CMs are major or minor is very important. This is discussed in the next section (4).
How many CMs can proliferate in adult stages?
Proliferation of adult mammalian CMs has been reported for many years; however, the exact percentages remain controversial. Soonpaa et al. surveyed a large amount of data and showed that the percentages of CMs with DNA synthesis were 0–.0006% and 0–0.0083% in normal and injured adult mouse CMs, respectively (Soonpaa & Field 1998). Determination of the percentages is dependent on the methods, and the toxic effects of radiolabelled thymidine and halogenated nucleotide analogues cannot be excluded. Senyo et al. used stable isotope labeling and multi-isotope imaging mass spectrometry to estimate the percentages and reported 0.007–0.015% and 0.06% per day in normal and injured adult mice, respectively (Senyo et al. 2013). Even considering the yearly rate, i.e. 5.5% (=0.015 × 365), these CMs are few in number and the majority of CMs cannot enter the cell cycle.
Proliferation of adult mammalian CMs has been reported also in many genetically manipulated mice (Ahuja et al. 2007). The percentages are also limited (at most, 1% in normal adult CMs, c-Myc transgenic mice, (Xiao et al. 2001)). In addition, the genetic manipulations were performed from fertilized eggs in almost all studies. Therefore, the effects during embryogenesis must be considered. For example, it is possible that expression or knockout of target genes changes the characteristics of CMs such as maintenance of immature status, and CMs can proliferate. Therefore, conditional analysis, which allows genetic manipulation only in adult differentiated CMs is required.
Another critical question is whether and how many of these adult CMs entering the cell cycle can divide in normal and genetically manipulated mice. Recent studies reported that a small number of CMs can complete mitosis at postnatal or adult stages (Bersell et al. 2009; Eulalio et al. 2012; Mollova et al. 2013; Senyo et al. 2013). The evaluation of karyokinesis and cytokinesis in some studies mainly depends on immunostaining against pH3-S10 or Aurora B in cardiac sections. However, it is impossible to definitively ascertain the completion of karyokinesis or cytokinesis based only on these stainings. Evaluation with more reliable methods such as pulse labeling with detectable nucleotides and measurement of DNA content per nucleus is required to obtain exact answers.
What is the biological significance of the cell cycle exit in many CMs?
As described above, the proliferation is very rare in adult mouse CMs. Although the percentages increase slightly in injured mice, the percentages are still very low (see ). Even though CMs can enter the cell cycle, only an extremely low number of CMs can divide. These facts indicate that the cell cycle exit is strictly maintained in the majority of adult CMs. What is the biological significance of the cell cycle exit?
One possibility is that a robust cell cycle exit is necessary for the homeostasis of mammalian cardiac functions. For example, the disassembly of myofibrils is observed in embryonic and postnatal mammalian CMs during mitosis (Rumyantsev 1977; Ahuja et al. 2004). It is possible that the deconstruction in many CMs causes heart failure in mammalian adults. This possibility could be examined and verified by canceling all inhibitory systems. If this is the case, careful manipulation of the reset of cell cycle exit is required for regenerative therapy.
How are binucleated CMs produced and what is the biological significance?
Binucleation of postnatal CMs is observed in at least several mammals. The final percentages are 80–90%, 32%, and 25–57% in mice/rats, pigs and humans, respectively (Gräbner & Pfitzer 1974; Schmid & Pfitzer 1985; Olivetti et al. 1996; Soonpaa et al. 1996; Ikenishi et al. 2012). The binucleation occurs by mitosis without cytokinesis. How is the cytokinesis blocked in the postnatal but not embryonic stages?
Interestingly, rat CMs in culture showed a contractile ring (Li et al. 1997), suggesting an early step in cytokinesis proceeds, while later steps are blocked. It was reported that anillin, a regulator of the cleavage furrow, did not localize in the midbody region, and that p38 MAP kinase inhibited the localization (Engel et al. 2006). Moreover, inhibition of p38 in adult CMs promoted cytokinesis (Engel et al. 2005). This result is not consistent with the long-standing idea that abundant and mature myofibrils inhibit cytokinesis in terminal differentiated CMs and binucleated CMs cannot divide. Recently, it was reported that binucleated hepatocytes can divide into two mononucleated cells during liver regeneration through chromosome condensation before mitosis (Miyaoka et al. 2012). This result also supports the possibility that adult binucleated CMs can divide.
What is the biological significance of CM binucleation? Are there some differences between mononucleated and binucleated CMs in terms of functions? Or is it a mere result of cytokinesis failure? Although we do not have satisfactory answers yet, it has been postulated that twice the transcripts from two nuclei are advantageous for maintenance of functions in hypertrophic CMs. If we determine the mechanisms for binucleation, it might be possible to manipulate the percentages of binucleation, and we will be able to understand the significance of binucleation and the functional differences.
CM proliferation and size control of the heart
The size of the heart is controlled mainly by the proliferation of CMs during embryonic stages, and by hypertrophy of CMs after the cell cycle exit. The heart is an essential pump, and an appropriate size is required at each embryonic stage. In addition, the heart size at birth influences the size after birth. Therefore, regulation of CM proliferation during embryonic stages is important for not only determination of proper heart size but also heart functions. What regulates the proliferation?
Judging from the function of the jumonji-cyclin D1 pathway and Tsc1/2 (TOR pathway), these proteins probably regulate heart size. In addition, the hippo pathway controls the heart size in embryos by repression of CM proliferation (Heallen et al. 2011). The hippo pathway regulates tissue size in many animals by the control of cell proliferation and apoptosis (Pan 2010). The pathway is a kinase cascade including the tumor suppressor Hippo (Mst1/2 in mammals), Wts (Lats1/2 in mammals) and Yki (Yap/Taz in mammals). Although enhanced proliferation was observed in CMs of jumonji and Tsc1/2 KO mice that died around E10–11 (Kobayashi et al. 1999, 2001; Takeuchi et al. 1999; Toyoda et al. 2003), KO mice in the hippo pathway survived until postnatal stages (Heallen et al. 2011), suggesting that the timing and/or extent is different among these systems. Interestingly, the hippo pathway regulates the CM proliferation by inhibiting Wnt signaling (Heallen et al. 2011). It is important to understand other critical pathways, whether there is crosstalk between the hippo, jumonji-cyclin D1, TOR and other pathways, and how these pathways are regulated for appropriate heart size.
Cardiomyocytes proliferation and cardiac regeneration
Cardiomyocytes s are essential for cardiac functions. If many CMs are lost and the loss is not recovered, heart failure or lethality occurs. The neonatal mouse and adult zebrafish are able to recover the loss of CMs following injury through CM proliferation and thus regenerate their hearts (Jopling et al. 2010; Kikuchi et al. 2010; Porrello et al. 2011b). In contrast, adult mammals are not capable of substantial cardiac regeneration, because a majority of CMs cannot re-enter the cell cycle. These facts show that the mechanisms for the cell cycle exit (see “How do CM maintain their cell cycle exit?”) are strictly maintained even after injury.
Cardiomyocytess can proliferate in the neonatal mouse and adult zebrafish. In addition, adult newt CMs also can proliferate and the heart can be regenerated (Oberpriller & Oberpriller 1971, 1974), although whether newts regenerate their hearts through CM proliferation remains unknown. On the other hand, mammalian adult CM cannot proliferate again. What determines the differences among species? The answer may be obtained by ascertaining the differences in the mechanism of cell cycle exit maintenance among these species. For example, what determines the difference in cytokinesis? These studies are significant for clarifying the mechanisms for not only cardiac regeneration in neonatal mice, adult zebrafish and newts, but also strict maintenance of cell cycle exit in mammalian adult CMs.
Cardiomyocytes proliferation and differentiation
In general, cell proliferation and differentiation show an inverse relationship, and are regulated in a coordinated manner during development. Embryonic CMs must support embryonic life by functional differentiation such as by beating, and proliferate actively to increase the size of the heart. Therefore, both proliferation and functional differentiation progress simultaneously, and are indispensable for embryonic development. The situation is different from many other types of cells, such as skeletal muscle cells or neuronal cells, which are known to differentiate functionally after cell cycle arrest. Are proliferation and differentiation related to each other? Does one inhibit the other in embryonic CMs, which differ in proliferation and differentiation?
Activity for proliferation inhibits differentiation in embryonic CMs (Nakajima et al. 2011). Enhanced expression of cyclin D1 in CMs of jumonji mutant mice or transgenic mice decreased GATA4 protein expression and inhibited the differentiation of CMs. CDK4, which is activated by cyclin D1, phosphorylates GATA4 directly, and the degradation of GATA4 is promoted. GATA4 is one of the critical transcriptional factors for CM differentiation (Harvey 1999; Olson 2006). These results suggest that CDK4 activated by cyclin D1 inhibits differentiation of CMs by degradation of GATA4, and that initiation of jumonji expression unleashes the inhibition by repression of cyclin D1 expression and allows progression of differentiation, as well as repression of proliferation. Thus, a jumonji-cyclin D1 pathway coordinately regulates the proliferation and differentiation of CMs (Nakajima et al. 2011) (Fig. 3).
It is interesting but unknown whether CM proliferation inhibits the differentiation by other mechanisms, and whether CM differentiation inhibits the proliferation.
Based on the discussion above, future main subjects that need to be studied and solved are listed below.
Regulation of CM proliferation during development
Exactly how is the cell cycle regulated in CMs during embryos to adults? Exactly how many adult CMs can reenter cell cycle?
Many facts regarding cell cycle regulation have been elucidated; however, the exact details of the overall process remain unknown. For example, the mechanisms linking braking, cell cycle exit, and maintenance are unknown. In addition, why was the promotion in proliferation of CMs very limited but not in all CMs of genetically manipulated mice? One possibility is the presence of unknown redundant systems, and another is that we observed the promotion only in specific CMs. In the latter case, we do not know the mechanisms in many CMs. However, the former would be more possible because many adult and postnatal CMs can reenter the cell cycle (Tane et al., unpubl. data). In any event, we must determine whether there is heterogeneity in the ability to re-enter cell cycle among CMs.
How is expression of the cell cycle regulators and their upstream factors controlled?
It is necessary to ascertain exactly how cell cycle regulator expression or activity is up- or downregulated for precise control of CM proliferation patterns. It is also important to determine the mechanisms that control expression of upstream factors such as Meis 1 and several microRNAs. However, it will be necessary again to analyze the next upstream factors. What are the final upstream factors? Since proliferation is closely related to the differentiation, structural changes, and functions of CMs, upstream factors are most likely involved in these events. In addition, the linking might be closely related to heart size control. Epigenetic controls should also be considered.
CM proliferation and regeneration
What is the essential difference in proliferative ability of CMs between adult mammals and zebrafish/newts?
As described in 4, the ability of adult CMs to proliferate after injury is one of the critical points for cardiac regeneration, although other points are also important (Kikuchi & Poss 2012). Various possibilities can be put forward. The differences might be in cell cycle regulators (proteins and/or promoters), signaling pathways linking triggers to the cell cycle regulators, triggers, and so on. Studies comparing mammals with zebrafish/newts will be a powerful method. Will the difference be identified in the nucleotide sequences in one or some genes? Or identified in some systems? These studies are also vital to understanding the heterogeneity between species and evolution.
CM proliferation and differentiation
How are proliferation and differentiation related to each other in CMs? What is the significance in development and functional maintenance of CMs?
Proliferation and differentiation probably coordinately support the development and functional maintenance of CMs. At least it can be speculated that one influences the other in many aspects. For example, CM proliferation inhibits differentiation (Nakajima et al. 2011). It is also possible that CM differentiation or maturation blocks the proliferation, especially in the postnatal stage. Therefore, studies on the interaction would be very important for understanding the signaling mechanisms for cell cycle exit and maintenance (see also 1 [ii])
The author wishes to thank Ms. Hitomi Okayama (Tottori University) for preparing the illustrations in the figures. This work was supported by a research grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan.