Cardiomyocyte Identity Crisis Without Regenerative Re-entry: A Multi-Scale Atlas of Incomplete Dedifferentiation and Intercellular Communication Rewiring in the Infarcted Human Heart

Background: Adult mammalian hearts regenerate poorly after myocardial infarction (MI) because cardiomyocytes largely, though not absolutely, exit the cell cycle during postnatal maturation, yet the molecular logic connecting maturation, disease-associated dedifferentiation, and intercellular communication remains unclear. Methods: Public single-cell and single-nucleus transcriptomes of developing, adult, and infarcted human and mouse hearts (55,342 cardiomyocytes) are integrated with Visium spatial transcriptomics to reconstruct a cross-species maturation trajectory, a directional ligand-receptor communication atlas, and rare-event proliferation counts with donor-aware statistics. Results: MI cardiomyocytes show a partial identity erosion -- a stress-associated myosin isoform switch, collapse of the ventricular identity factor IRX4, and maturation-score decline -- without regenerative proliferative re-entry and without fetal transcription-factor re-activation; cycling nuclei remained exceedingly rare (<0.03%), a state corroborated by chromatin accessibility and human genetics, with only the MYC regulon passing transcriptome-wide correction. Inferred communication reorganization was confined to non-myocyte compartments and did not exceed donor-label permutation nulls; cardiomyocytes remained net receivers without gaining signalling authority. IRX4 knockdown in human iPSC-derived cardiomyocytes recapitulated this identity loss, including stress-gene induction and elevated NT-proBNP secretion, and IRX4 overexpression attenuated the adrenergic stress response in IRX4-intact cardiomyocytes (orthogonal rescue untested). Conclusions: These findings support a model of state-relationship decoupling, in which an intracellular cardiomyocyte identity crisis and an intercellular niche rewiring proceed independently -- a decoupling we define operationally and test with donor-level null models and equivalence bounds -- and may help explain the limited success of regenerative strategies that target either layer alone.