Identifying kinematic biomarkers of the dystrophic phenotype in a zebrafish model of Duchenne muscular dystrophy.

Abstract

Background: Dystrophin-deficient zebrafish larvae are a small, genetically tractable vertebrate model of Duchenne muscular dystrophy that is well suited for early-stage therapeutic development. However, current approaches for evaluating their mobility, a physiologically relevant therapeutic outcome, yield data of low resolution and high variability that provides minimal insight into potential mechanisms responsible for their abnormal locomotion.

Methods: To address these issues, we used high speed videography and deep learning-based markerless motion capture to quantify escape response (ER) swimming kinematics of two dystrophic zebrafish strains (sapje and sapje-like). Each ER was partitioned into an initiating C-start, a subsequent power stroke, and a final burst of undulatory swimming activity.

Results: Markerless motion capture provided repeatable, high precision estimates of swimming kinematics. Random forest and support vector machine prediction models identified overall ER distance and peak speed, the instantaneous speed conferred by the power stroke, and the average speed and distance covered during burst swimming as the most predictive biomarkers for differentiating dystrophic from wild-type larvae. For each of these predictors, mutant and wild-type larvae differed markedly with effect sizes ranging from 2.4 to 3.7 standard deviations. To identify mechanisms underlying these performance deficits, we evaluated the amplitude and frequency of propulsive tail movements. There was little evidence that tail stroke amplitude was affected by the absence of dystrophin. Instead, temporal aspects of tail kinematics, including tail maximal angular velocity during the C-start and power stroke and tail stroke frequency during burst swimming, were slowed in mutants. In fact, tail kinematics were as effective as direct, non-survival in vitro assessments of tail muscle contractility in differentiating mutant from wild-type larvae.

Conclusions: ER kinematics can be used as precise and physiologically relevant biomarkers of the dystrophic phenotype, may serve as non-lethal proxies for skeletal muscle dysfunction, and reveal new insights into why mobility is impaired in the absence of dystrophin. The approach outlined here opens new possibilities for the design and interpretation of studies using zebrafish to model movement disorders.