Evaluating Pathogenicity of TPM1 Variants of Unknown Significance Using In-silico and In-vitro Models

Variants of unknown significance (VUS) pose a barrier to cascade genetic screening in families affected by inherited cardiomyopathies. Here, we investigate scalable computational methods for assessing pathogenicity of TPM1 VUS in the context of cardiomyopathy and test them against in vitro data.

Twenty TPM1 VUS (all missense) from clinical databases were computationally evaluated by inserting each into atomistic representations of tropomyosin. Molecular dynamic simulations were used to assess changes in tropomyosin flexibility, and a structural model of tropomyosin on actin allowed calculation of altered electrostatic interaction energies. Four variants representing diverse changes in these properties were carried forward for functional studies that included in vitro motility, contractility of engineered heart tissues (EHTs), and morphology of iPSC-derived cardiomyocytes.

The TPM1 variants A102D, D258E, K233N, and A239T all showed at least one significantly altered facet of contractile function. A102D was associated with increased thin filament Ca²⁺ sensitivity, increased twitch contraction force, and slowed relaxation. D258E was shown to cause a slowing of twitch relaxation and reduced ability to block myosin activity at low Ca²⁺. K233N and A239T both reduced in vitro thin filament motility, drastically decreased twitch contraction force, and triggered pronounced increases in cardiomyocyte aspect ratio.

By showing that large molecular aberrations predicted by thin filament structural models translate into meaningful functional changes (as benchmarked against clinically pathogenic mutations), this study supports the feasibility of TPM1 variant classification by means of a scalable computational pipeline.