Cardiac diseases are characterized by the inability of adult mammalian hearts to overcome the loss of cardiomyocytes (CMs). Current knowledge in cardiac regeneration lacks a clear understanding of the molecular systems determining whether CMs will progress through the cell cycle to proliferate. Here, we developed a computational model of cardiac proliferation signaling that identifies key regulators and provides a systems-level understanding of the cardiomyocyte proliferation regulatory network. This model defines five regulatory networks (DNA replication, mitosis, cytokinesis, growth factor, hippo pathway) of cardiomyocyte proliferation, which integrates 72 nodes and 88 reactions. The model correctly predicts 72 of 76 (94.7%) independent experiments from the literature. Network analysis predicted key signaling regulators of DNA replication (e.g., AKT, CDC25A, Cyclin D/CDK4, E2F), mitosis (e.g., Cyclin B/CDK2, CDC25B/C, PLK1), and cytokinesis, whose functions varied depending on the environmental context. Regulators of DNA replication were found to be highly context-dependent, while regulators of mitosis and cytokinesis were context-independent. We also predicted that in response to the YAP-activating compound TT-10, the Hippo module crosstalks with the growth factor module via PI3K, cMyc, and FoxM1 to drive proliferation. This prediction was validated with inhibitor experiments in primary rat cardiomyocytes and further supported by re-analysis of published data on YAP-stimulated mRNA and open chromatin of Myc from mouse hearts. This study contributes a systems framework for understanding cardiomyocyte proliferation and identifies potential therapeutic regulators that induce cardiomyocyte proliferation.

Rationale: Heart failure is caused by the inability of adult mammalian hearts to overcome the loss of cardiomyocytes (CMs). This is due partly to the limited proliferative capacity of CMs, which exit the cell cycle and do not undergo cell division. Current knowledge in cardiac regeneration lacks an understanding of the molecular regulatory networks that determine whether CMs will progress through the cell cycle to proliferate. Our goal is to use computational modeling to understand the expression and activation levels of the core cell cycle network, specifically cyclins and cyclin-cyclin-dependent kinase (CDK) complexes. Methods: A model of core cell cycle dynamics was curated using previously published studies of CM proliferation regulators. This model incorporates those regulators known to stimulate G1/S and G2/M transitions through the core CDKs. The activity of each of the 22 network nodes (22 reactions) was predicted using a logic-based differential equation approach. The CDK model was then coupled with a minimal ODE model of cell cycle phase distributions and validated based on descriptions and experimental data from the literature. To prioritize key nodes for experimental validation, we performed a sensitivity analysis by stimulating individual knockdown for every node in the network, measuring the fractional activity of all nodes. Results: Our model confirmed that the knockdown of p21 and Rb protein and the overexpression of E2F transcription factor and cyclinD-cdk4 showed an increase in cells going through DNA synthesis and entering mitosis. A combined knockdown of p21 and p27 showed an increase of cells entering mitosis. Cyclin D-cdk4 and p21 overexpression showed a decrease and increase of Rb expression, respectively. Of the 14 model predictions, 12 agreed with experimental data in the literature. A comprehensive knockdown of the model nodes suggests that E2F (a key transcription factor driving DNA synthesis) is positively regulated by cyclin D while negatively regulated by GSK3B, SMAD3, and pRB. Conclusion: This model enables us to predict how cardiomyocytes respond to stimuli in the CDK network and identify potential therapeutic regulators that induce cardiomyocyte proliferation.