Pharmacological modulation of cardiac senescence reveals divergent transcriptomic reversal patterns in human cardiomyocytes

Background Cellular senescence contributes to cardiac ageing and cardiometabolic disease [1–3], yet how pharmacological interventions modulate senescence-associated transcriptional programs in human cardiomyocytes remains insufficiently defined. We aimed to systematically quantify the transcriptomic remodelling and the potential to reverse senescence of diverse cardiometabolic and geroprotective compounds in a human stem cell–derived cardiomyocyte model. Methods Human WA09 pluripotent stem cells were differentiated into cardiomyocytes and rendered senescent using doxorubicin. Senescent cardiomyocytes were exposed to eight compounds (dasatinib, quercetin, rapamycin, metformin, bisoprolol, insulin, IGF-1, and glucagon) at three concentrations. RNA sequencing was performed, and differential expression was analysed relative to untreated senescent controls. A cardiomyocyte senescence signature was derived from senescent versus non-senescent cells. Treatment-induced log2 fold changes were integrated into this signature to calculate a senescence-reversal score based on directional concordance with the inverse senescence profile. Results Differentiation produced a clear transcriptional transition from pluripotency to cardiomyocyte identity (>12,000 DEGs), confirming model fidelity. Senescent cardiomyocytes exhibited a coherent transcriptional signature (707 DEGs) enriched for cell-cycle arrest and stress-response pathways. Pharmacological perturbations induced highly heterogeneous transcriptional responses (1–685 DEGs per condition). Metabolic hormones such as glucagon and insulin produced the largest DEG burdens but weak reversal alignment, indicating extensive yet poorly coordinated remodelling. In contrast, dasatinib combined moderate DEG counts with the strongest reversal alignment, suggesting directionally coherent modulation of senescence programs. Metformin induced minimal transcriptomic perturbation but relatively favourable reversal scores, consistent with low-amplitude yet coordinated modulation. Across all treatments, transcriptional magnitude and reversal alignment were only weakly coupled. Conclusions Pharmacological modulation of senescent human cardiomyocytes reveals a decoupling between transcriptional magnitude and biological directionality. Large transcriptomic shifts do not necessarily indicate reversal of senescence, whereas directionally coherent changes may occur with modest DEG burdens. Directionality-aware transcriptomic scoring provides a scalable framework for prioritising candidate cardiometabolic and geroprotective interventions targeting cardiac ageing.