SUMMARY Millions of Americans currently have some degree of tendon tear that leaves the tissue with abnormal collagen quantity and alignment, and reduced mechanical properties. Collagen remodeling depends on mechanical loading, and therefore therapeutic interventions to restore normal tendon structure can have varied effects across diverse loading environments, i.e. patient-specific geometries, motions, or injury severities. Our ultimate, long- term objective is to design therapies tailored specifically for tendons across different loading environments. Collagen remodeling is governed by a complex system of interactions between matrix proteins, matrix metalloproteinases (MMPs), tissue inhibitors of metalloproteinases (TIMPs), degradation products, and growth factors, with mechanical loading affecting many of these interactions. Herein, we propose to experimentally elucidate unknown mechano-sensitivities of the collagen-MMP-growth factor network, and develop a computational model that integrates the multi-faceted network interactions as a tool for screening potential therapeutic interventions. Specifically, we aim to 1) test the effect of tensile loading on MMP-specific degradation of collagens by subjecting collagen I and III gels to various levels of strain with or without the addition of tendon- relevant MMPs, 2) test the hypothesis that tensile loading can release active TGFβ from collagenous matrix by subjecting collagen I and III gels to various levels of strain and measuring levels of latent and active TGFβ in the gels and media, and 3) build and test (ex vivo and in vivo) a computational model of load-dependent tendon matrix turnover that captures collagens, MMPs, TIMPs, degradation products, and TGFβ interactions as a system of ordinary differential equations. Collectively, these aims will immediately impact the field's basic knowledge of loading effects on matrix turnover and also produce the first large-scale model of the collagen- MMP-growth factor network, immediately impacting the field's ability to prospectively design therapeutic interventions (e.g., physical therapy regimens, MMP- or TIMP-targeting drugs, etc.) to control tendon matrix content and alignment given any specific loading and geometry. Such predictive capability will greatly support the proposed COBRE focus of patient-specific modeling for virtual human trials.