Kohn Lab

A limitation in translating tissue engineering to the clinic is an inability to reproducibly achieve a significant volume of tissue. To better control cells’ response to biomaterials and guide tissue regeneration, we mimic aspects of nature’s biomineralization strategies as a basis for biomaterials design. We developed biomaterials based on the self-assembly of biological mineral onto 3D organic templates. We also demonstrated that the function of stem cells and amount and spatial distribution of bone regenerated in-vivo were enhanced by controlling the nucleation and growth of bioimimetic mineral onto an organic template. The combination of material chemistry and substrate rigidity provide substrate-mediated cues for control of cell behavior, but soluble signals from a biomaterial can also influence cell differentiation, demonstrating that a biomaterial can influence cells without being in direct contact. This finding led us to engineer cell-cell communication as a strategy to enhance regeneration. We have shown that alterations in cell-cell communication via overexpressing gap junction proteins leads to significantly greater volumes of bone formed in-vivo and more uniform spatial distribution, potentially overcoming a critical barrier in tissue engineering.

We have developed organic/inorganic hybrid biomaterials in which proteins direct biological response to the material either passively or actively. The former technique capitalizes on Nature’s mechanisms of controlling crystal growth via protein/crystal binding, while in the later technique, therapeutic agents are integrated into the mineral in a spatially controlled manner, allowing for spatial and temporal delivery of active biological molecules. We developed high throughout combinatorial approaches to discover novel peptides and investigate peptide/biomaterial interactions, Using phage display, we discovered peptides that exhibit preferential affinity to specific material chemistries and a second set of peptides with affinity to specific cells. We combined these sequences to create dual-functioning peptides that have material and cell-specific binding domains, serving as tethers between the material and biological worlds and enabling spatially directed tissue formation. Capitalizing on the peptides’ apatite-binding binding ability, we next demonstrated their ability to bind to cell-secreted mineral, and these peptides are now used to inhibit pathological calcification and bind to bone for targeted delivery of drugs.

To develop structure-function relations in mineralized tissues and define metrics for bone quality across dimensional scale, we have used acoustic emission, Raman spectroscopy, NMR and HPLC to detect and predict damage, and identify ultrastructural changes associated with compromised bone quality. In collaboration with Dr. Michael Morris, we were the first to couple mechanical and compositional analyses of tissues using Raman spectroscopy, and relate mechanical deformation at the ultrastructural level to compositional changes. Using NMR, we are amongst the first to identify the important role of water in mediating mechanical stability of bone, and the first to demonstrate that there are 3 distinct water compartments in bone, each mediating a different functional response. Most recently, we developed a new animal model of lathyrism and HPLC methods for the direct quantification of immature and mature cross-links. We identified significant positive correlations between pyridinoline cross-link content and tissue strength and we are the first to report a correlation between bone fracture toughness and cross-link maturity. A significant outcome of this work is the recognition that a critical determinant of bone quality is cross-link type, rather than total quantity, and the cross-links that exist in lower abundance are the best predictors of bone quality.

Building on our ex-vivo based structure-function findings, we developed a short-term, weight-bearing rodent exercise model. Using this model with mice of different ages, genders, background strains and genetic modifications, we identified significant location and gender-specific differences in phenotype and mechanical responsiveness of the skeletal system. We also determined that exercise can rescue deficiencies in function caused by gene deletion, suggesting that mechanical stimulation can compensate for genetically-induced deficiencies in bone quality. However, we found that the phenotype resulting from targeted deletion of bone matrix genes is dependent on the background strain on which transgenic mice are bred. Using the exercise model, we also demonstrated that in-vivo mechanical loading can increase resistance to fatigue damage, and defined compositional changes in collagen cross-linking guiding the increased resistance. Most recently, using exercise in conjunction with enhancement or inhibition of PTH signaling, we demonstrated that bone adaptation during exercise is not only a function of dynamic loading, but also PTH release, and that PTH signaling contributes differently to bone adaptation at the structural and tissue levels.

It has long been assumed that bone adapts to mechanical loading and improves its strength by turning over or adding material. We demonstrated that bone can also respond to exercise, as well as genetic alterations and other stimuli/insults, via mechanical changes that are not mediated by the addition of new bone, but via compositional changes. We have further demonstrated that these compositional changes occur predominantly in the perilacunar zone around osteocytes. We also demonstrated that exercise can shift cross-link chemistry toward the pyridinoline pathway and counter the effects of cross-link inhibition on mechanical properties. Also of significance is our finding that cross-link inhibition and rescue can result in changes in modulus without changes in mineral density, challenging the tenet that modulus is controlled by the mineral component of bone.