In this project we are probing the molecular and cellular coupling of energy metabolism and cardiomyocyte mechanics.
The passive stiffness of heart muscle governs how the ventricle fills between beats. Experiments show that calcium increases the muscle's viscous, rate-dependent stiffness more than six-fold, through a mechanism distinct from cross-bridge cycling. A model in which serial globular titin domains of titin unfold under stress, acting in parallel with a nonlinear elastic element, explains the power-law stress relaxation, hysteresis, preconditioning, and calcium dependence of passive cardiac muscle mechanics.
Myocardial ATP levels fall in heart failure, but it has been unclear whether this drop is large enough to impair the contractile proteins. Experiments on demembranated cardiac muscle show that lowering MgATP from 8 mM to 4 mM, the levels found in non-failing and failing hearts, does not change maximal force or maximal shortening velocity, but reduces maximal mechanical power by 20-25% and slows cross-bridge cycling. These results suggest that reduced ATP in the failing heart directly contributes to impaired pump function.
The goal of this project is to gain a quantitative predictive understanding of muscle metabolism under a range of dynamic conditions.
Experiments on purified cardiac mitochondria inform a kinetic model of substrate transport, the TCA cycle, and oxidative phosphorylation (diagram).
Preprint: Collins et al., bioRxiv (2026)
The goals of this project are to build a model of cardiomyocyte energy metabolism by joining our mitochondrial model to the main pathways of the cytoplasm (glycolysis, lactate exchange, the malate-aspartate shuttle, the purine nucleotide cycle, and creatine kinase energy buffering), and to use this model to understand how the heart chooses between glucose and lactate as fuel, how its energy metabolism responds to increasing work, and how metabolic intermediates such as succinate build up during loss of oxygen and clear during recovery.
In this project we are investigating how the biochemical/metabolic state of the heart in heart failure limits reserve cardiac power capacity.
The goals of this project are to build patient-specific computer models of the heart and circulation (digital twins) from routine clinical data, and to combine them with machine learning to guide diagnosis, prognosis, and treatment in heart failure. Digital twins reveal the underlying physiology of each patient, identify groups of patients with different mechanisms of disease and different risk, improve prediction of outcomes, and help predict which patients are likely to benefit from specific therapies.
Gu et al., npj Digital Medicine (2025)
Gu et al., medRxiv preprint (2026)