Biophysical Impact Of Β-Cardiac Myosin Cardiomyopathy Mutations And Development Of Artificial Intelligence-Based In Vitro Motility Analysis

Hypertrophic and Dilated Cardiomyopathies (HCM and DCM, respectively) are among the most common causes of heart failure, with 1 in 500 and 1 in 2500 affected people in the US. Interestingly, a significant percentage of HCM and DCM are rooted in genetic mutations in β-cardiac myosin, the molecular motor that powers heart contractions. However, the mechanisms by which the numerous HCM and DCM mutations alter myosin function remain poorly understood. Recently, it has been discovered that the myosin molecular motor does not simply remain idle waiting to participate in a cardiac contraction. Instead, myosin can adopt an auto-inhibited state that prevents motors from engaging during a contraction. The auto-inhibited state of myosin is in an equilibrium with an active motor state capable of generating force, and alterations to such an equilibrium present a unifying mechanism that could explain the effects of HCM (hypercontractility, more active motors) and DCM (hypocontractility, less active motors). The auto-inhibited state of myosin can be studied by coating a surface with myosin motors and observing fluorescently labeled actin filaments glide over them in the presence of ATP at low ionic strength (termed the in vitro motility assay or IVMA). Due to challenges in the traditional method of analyzing IVMA data (time spent, human bias), we established the parallel goals of 1) creating an accurate and high-throughput motility analysis routine based on machine learning and 2) using the IVMA to study wild type (WT) human myosin constructs of variable tail length (long and short). This allowed us to define the mechanical effect of the auto-inhibited state of myosin on the velocity of actin filament displacement, only possible in the long-tail constructs, which can adopt the auto-inhibited state. Next, we investigated the impact of one HCM mutation (R723G) and two DCM mutations (F764L, E525K) on the auto-inhibited state of myosin using a combination of IVMA, biochemical assays, and analytical modeling. Interestingly, in the IVMA, the R723G mutation showed no effect on short-tail constructs, which are incapable of adopting the auto-inhibited state, but increased actin filament velocity of long-tail constructs. F764L decreased velocity in long-tail constructs and, to a much lesser extent, in the short-tail constructs, too. Conversely, the E525K mutation showed no velocity changes in IVMA. Biochemical assays interrogating the auto-inhibited state of myosin and ATPase activity showed an equivalent effect for R723G and F764L to that of the IVMA. Moreover, the E525K mutation significantly increased biochemical auto-inhibition at physiological ionic strength while paradoxically increasing ATPase activity, suggesting that competing mechanisms underline the net phenotype of the mutation. These findings establish auto-inhibition as a critical determinant of myosin function in cardiomyopathy mutations, providing a novel analytical framework for studying motor protein dysfunction with direct implications for the development of targeted cardiac therapies.