A metabolic switch from OXPHOS to glycolysis is essential for cardiomyocyte proliferation in the regenerating heart

Hessel Honkoop, Dennis de Bakker, Alla Aharonov, Fabian Kruse, Avraham Shakked, Phong Nguyen, Cecilia de Heus, Laurence Garric, Mauro Muraro, Adam Shoffner, Federico Tessadori, Joshua Peterson, Wendy Noort, George Posthuma, Dominic Grun, Willem van der Laarse, Judith Klumperman, Richard Jaspers, Kenneth Poss, Alexander van Oudenaarden, Eldad Tzahor, Jeroen Bakkers

Preprint posted on December 18, 2018

Heart regeneration: Cardiomyocytes need to undergo a metabolic switch in order to re-enter the cell cycle

Selected by Andreas van Impel


During myocardial infarctions the occlusion of a blood vessel leads to an undersupply of oxygen in the cardiac muscle tissue, resulting in the loss of heart muscle cells. After such cardiac injuries the mammalian heart fails to replace the ischemic areas with novel functional contractile cardiomyocytes. This inability to regenerate results in cardiac dysfunction and often even in heart failure which makes heart attacks a major cause of death in humans.

In contrasts to mammals, fish and amphibians possess remarkable regenerative properties which have made these animals important models for regeneration in general and heart regeneration in particular. Previous work has shown that upon injury of the zebrafish heart cardiomyocyte proliferation is induced in the so-called border zone area surrounding the affected tissue; this ultimately results in the complete replacement of the apoptotic cardiomyocytes within the injury area by novel functional heart muscle cells [reviewed in 1]. However, it is currently not fully understood which processes and signalling events drive this regenerative response in the zebrafish heart (in contrast to the responses induced in mammalian cardiac tissues).


Key findings

In order to more closely characterize the small number of differentiated cardiomyocytes that start to re-enter the cell cycle after heart injury, Honkoop, de Bakker and colleagues performed single cell mRNA-sequencing on border zone tissue isolated from cryo-injured adult zebrafish hearts. Subsequent transcriptome analysis revealed that cardiomyocytes cluster into four different subgroups with normal, adult cardiomyocytes in one cluster and dedifferentiated, embryonic-like cardiomyocytes expressing various known border zone marker genes in a separate cluster. The remaining two subgroups reflect intermediated populations of dedifferentiating cardiomyocytes. Further analysis indicated that many genes involved in energy producing metabolic pathways were differentially expressed among the individual clusters, implicating fundamental metabolic changes in cardiomyocytes during their dedifferentiation and re-entry into the cell cycle. In particular, differentiated, adult cardiomyocytes show increased expression of genes involved in the oxygen-dependent mitochondrial oxidative phosphorylation while the proliferative population within the border zone seemed to have increased expression of enzymes required for glycolysis accompanied by a reduction in mitochondrial activity.

To test the functional importance of this finding, the authors inhibited glycolysis and found that under these conditions cardiomyocyte proliferation in the border zone was significantly reduced, suggesting that the proposed metabolic switch during the dedifferentiation of cardiomyocytes is involved in cell cycle re-entry and therefore the regeneration of heart tissue. Furthermore, the authors show that Neuregulin1, a potent injury-induced mitogen that is known to trigger cardiomyocyte dedifferentiation [2], acts upstream of this metabolic switch as it induces the expression of glycolysis genes in cardiomyocytes. Finally, the authors turn to a murine model with an improved cardiac regenerative capacity (mice overexpressing the Neuregulin1 receptor ErbB2) to address the importance of increased glycolysis for cardiomyocyte proliferation and heart regeneration in the mammalian system. As one would predict from the zebrafish results, Honkoop, de Bakker et al. find that ErbB2 overexpressing mice (in contrast to wild type mice) show an enhanced expression of glycolysis related genes in cardiomyocytes after myocardial infarction and that this induction is essential for their cell cycle re-entry and therefore the increased heart regeneration in this mouse model. Taken together, the here presented results demonstrate that cardiomyocytes undergo an important metabolic reprogramming (from mitochondrial oxidative phosphorylation to increased glycolysis) during the response to cardiac injury, which is essential for the induction of cell proliferation and the regeneration of cardiac tissue.


Why this is cool

The preprint from the Bakkers lab nicely demonstrates the power of single cell transcriptomics in identifying and characterizing cellular responses and events in relatively small subpopulations of a certain cell type, which might otherwise get masked in the transcriptome analysis of a mixture of the whole cell population. Using this approach, the authors find an essential role for a shift in the energy providing metabolic pathways during the induction of cardiomyocyte proliferation and heart regeneration. The here presented results therefore imply that this metabolic cardiomyocyte reprogramming represents another important piece of the ‘heart regeneration puzzle’ and that an induction of such a switch might provide a promising target for therapeutic interventions after myocardial infarction in the future.


Open questions / future directions

  • Do all cardiomyocytes in the border zone initiate or at least have the capacity to undergo this metabolic switch and to re-enter the cell cycle? If not, what makes this dedifferentiating population special?
  • What is the exact molecular connection between enhanced glycolysis and cardiomyocyte dedifferentiation / induction of proliferation?
  • Is a ‘reverse metabolic switch’ required to end the regeneration program in cardiomyocytes?
  • How can one stimulate such a metabolic switch after myocardial infarction in human patients and would that alone be sufficient for an improved prognosis?



Further reading

[1] González‐Rosa, J. M., Burns, C. E., Burns, C. G. Zebrafish heart regeneration: 15 years of discoveries. Regeneration (Oxf). 2017 Jun; 4(3): 105–123

[2] Gemberling, M., Karra, R., Dickson, A. L., Poss, K. D. Nrg1 is an injuryinduced cardiomyocyte mitogen for the endogenous heart regeneration program in zebrafish. eLife 2015;4:e05871

Tags: danio rerio, regeneration

Posted on: 30th January 2019 , updated on: 31st January 2019

Read preprint (2 votes)

  • Author's response

    Jeroen Bakkers shared

    Dear Andreas,

    Thank you for highlighting our work and starting an interesting scientific discussion. Here I would like to address your questions.

    Your first question about the identity of the cells that undergo metabolic reprogramming and re-enter the cell cycle is important and very interesting. Based on our single-cell analysis and validation we think that the dedifferentiation and metabolic programming is induced at the border zone close to the injury. Since we did not detect any expression of the genes that are indicative for the dedifferentiation or metabolic switch in the uninjured heart, we think both processes are induced by the injury. Whether all cardiomyocytes located in this border zone or only a subpopulation of cardiomyocytes can respond to the injury remains to be determined. To address this question extensive lineage tracing experiments will be required.

    The second question relates to the connection between the enhanced glycolysis and cardiomyocyte dedifferentiation and proliferation. Very little is known about this but there are some hints from work by others in the cardiac field and stem cell field. It is known from mammalian studies in the 1990’s that the embryonic heart uses mainly glycolysis for energy metabolism while mature cardiomycytes in the adult heart mainly use OXPHOS. In mice, the switch from glycolysis to OXPHOS occurs in the first week after birth which coincides with a sharp decrease in cardiomyocyte proliferation and a loss in regenerative capacity of the heart. Whether the process can be reversed has not been address but our results suggest that this is what might be happening in the zebrafish heart. Interestingly, a similar metabolic shift from mitochondrial OXPHOS to glycolysis occurs in proliferating tumor cells, and was first described by Otto Warburg in 1927. While glycolysis generates much less ATP compared to fatty acid oxidation, it is thought that glycolysis and the connected pentose-phosphate pathway provides essential metabolites that are needed to create sufficient biomass to sustain proliferation of the tumor cells. Furthermore, progenitor cells in the developing embryo as well as induced pluripotent stem cells depend on glycolysis to maintain proliferation and their potency. In addition, glycolytic enzymes such as PKM2 and PFKFB4 can also directly interact with cell cycle regulators to promote proliferation.

    The answer to your third question is probably a yes, since a fully functional and contractile myocardium needs large amounts of ATP. Since OXPHOS is much more efficient in producing ATP, the new cardiomyocytes will most likely revert back to it.

    At this point, I am not able to answer your last question. We and others in this field hope that all the work that has been done and will be done will help to better understand how cardiac regeneration works in the zebrafish and that this knowledge eventually will result in improving heart repair in the future.

    Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here