Region-specific cellular ecosystems govern divergent remodelling trajectories in hypertrophic cardiomyopathy

Background

Hypertrophic cardiomyopathy (HCM) exhibits marked regional heterogeneity despite shared sarcomere gene mutations. The molecular basis underlying preferential septal hypertrophy and relative apical preservation remains poorly understood.

Purpose

To define region-specific cellular programs and intercellular signalling networks underlying spatially distinct remodelling in HCM.

Methods

We performed integrated single-nucleus RNA sequencing and single-nucleus ATAC sequencing of hypertrophic septum (VS), mildly hypertrophic apex (VA) and non-hypertrophic anterolateral wall (AL) tissues from HCM patients, together with matched peripheral immune profiling. Gene regulatory networks, trajectory inference, and ligand-receptor analyses were applied. Functional validation was performed using transcription factor overexpression in neonatal cardiomyocytes and recombinant THBS1 treatment in patient-derived iPSC-cardiomyocytes. A neural network classifier was constructed to evaluate the discriminative capacity of region-specific transcriptional signatures.

Results

Cardiomyocytes segregated into divergent region-enriched states. Septal cardiomyocytes preferentially activated stress-associated transcriptional programmes, whereas apical cardiomyocytes displayed antioxidant and metabolic signatures. Pseudotime reconstruction revealed bifurcating trajectories from a homeostatic state toward either stress-dominant septal remodelling or an adaptive apical programme.

Immune profiling demonstrated a spatial gradient of CD4⁺ T cell activation, with cytotoxic subsets enriched in the septum. Ligand-receptor analysis identified CD44-HBEGF signalling between cytotoxic CD4⁺ T cells and cardiomyocytes as a prominent septal interaction axis. Cardiomyocyte HBEGF expression correlated with adverse functional parameters in an external cohort.

In contrast, a distinct fibroblast subset was preferentially enriched in the apex and engaged cardiomyocytes via THBS1-CD36 signalling. THBS1 exposure partially restored homeostatic transcriptional programmes in patient-derived iPSC-cardiomyocytes.

Neural network modelling demonstrated that region-specific transcriptional signatures were sufficient to predict anatomical origin with high accuracy and identified a principal component axis separating adaptive apical states from stress-dominant septal remodelling.

Conclusion

HCM remodelling is organised by spatially distinct cellular ecosystems rather than a uniform sarcomere-driven process. The septum is characterised by immune-coupled stress programmes, whereas the apex harbours a fibroblast-centred compensatory network. These findings provide a mechanistic framework for regional heterogeneity in HCM and suggest that spatially targeted therapeutic strategies may complement sarcomere-directed interventions.graphical abstract