Imaging studies of peripheral nerve regeneration induced by porous collagen biomaterials

Abstract

There is urgent need to develop treatments for inducing regeneration in injured organs. Porous collagen-based scaffolds have been utilized clinically to induce regeneration in skin and peripheral nerves, however still there is no complete explanation about the underlying mechanism. This thesis utilizes advanced microscopy to study the expression of contractile cell phenotypes during wound healing, a phenotype believed to affect significantly the final outcome. The first part develops an efficient pipeline for processing challenging spectral fluorescence microscopy images. Images are segmented into regions of objects by refining the outcome of a pixel-wide model selection classifier by an efficient Markov Random Field model. The methods of this part are utilized by the following parts. The second part extends the image informatics methodology in studying signal transduction networks in cells interacting with 3D matrices. The methodology is applied in a pilot study of TGFP signal transduction by the SMAD pathway in fibroblasts seeded in porous collagen scaffolds. Preliminary analysis suggests that the differential effect of TGFP1 and TGFP3 to cells could be attributed to the "non-canonical" SMADI and SMAD5. The third part is an ex vivo imaging study of peripheral nerve regeneration, which focuses on the formation of a capsule of contractile cells around transected rat sciatic nerves grafted with collagen scaffolds, 1 or 2 weeks post-injury. It follows a recent study that highlights an inverse relationship between the quality of the newly formed nerve tissue and the size of the contractile cell capsule 9 weeks post-injury. Results suggest that "active" biomaterials result in significantly thinner capsule already 1 week post-injury. The fourth part describes a novel method for quantifying the surface chemistry of 3D matrices. The method is an in situ binding assay that utilizes fluorescently labeled recombinant proteins that emulate the receptor of , and is applied to quantify the density of ligands for integrins a113, a2p1 on the surface of porous collagen scaffolds. Results provide estimates for the density of ligands on "active" and "inactive" scaffolds and demonstrate that chemical crosslinking can affect the surface chemistry of biomaterials, therefore can affect the way cells sense and respond to the material.