Modeling the Osteogenic Potential of Decellularized Human Bone Particles by Tuning their Size Distribution Through a Sonic Microfragmentation Approach

Human decellularized bone fragments are commonly used in clinics to perform allograft surgeries. To reduce the immunological response in the recipient and ensure their safety, these fragments underwent decellularization, a procedure that greatly reduces their osteogenicity (the ability to induce differentiation into osteoblast). In this work, by the introduction of an ultrasonication step to fragment the human bone, the size distribution of the resulting demineralize bone particles can be controlled, tuning their osteogenic potential. The sonication protocol is optimized by a response surface method, using 12 different sonication protocols, allowing to model the relationship between the sonication parameters and the outcoming particles properties in terms of dimensions, physical/chemical properties, and biological activity. The size distribution is extrapolated by a deep learning image segmentation while the structure is characterized by infrared and thermal analysis. The particles are combined with methacrylated silk gel to test in vitro their biological response on adipose-derived stromal cells. The ultrasonication fragmented the bone particles, revealing their internal organic matrix as proved by secondary electron microscopy and confocal microscopy. An inverse linear correlation is found between the particles' sizes and their osteogenic activity, thus proving the efficacy of the proposed ultrasonication treatment in tuning the biological response. Decellularized human bone particles coming from the musculoskeletal tissue bank are further fragmented through a probe ultrasonicator. The process is studied by a Response Surface Method allowing to building of empirical models that describe the sonicated particle's properties. The ultrasonication process decreases the size of the bone particles, revealing their internal organic matrix and improving the biological response.image