Wound healing is one of biology's most conserved responses: across nearly every tissue in the body, injury triggers a coordinated sequence of inflammation, cell recruitment, and matrix remodeling that closes the wound and restores tissue integrity. The central nervous system (CNS) mounts this same fundamental response — but in the CNS, wound healing rarely resolves cleanly. Instead, hematogenous immune cells and fibroblasts recruited to the injury site drive the formation of a dense, permanent scar that blocks tissue repair and axon regeneration. Our laboratory studies CNS injury as a wound healing problem: we ask why the normal healing program that closes wounds elsewhere in the body becomes stalled and fibrotic in the CNS, with the goal of identifying points in that process where healing can be redirected toward genuine tissue repair.

Our early work established that perivascular fibroblasts, not meningeal fibroblasts as classically assumed, are a principal source of the fibrotic scar after spinal cord injury, and that this fibroblast activation is orchestrated by hematogenous macrophages. In a normal wound, macrophages transition from an early inflammatory phenotype to a pro-resolving one that clears debris and permits repair to proceed. We have focused on why this transition fails in the CNS: after injury, macrophages accumulate cholesterol and lipid debris from damaged myelin and cell membranes, becoming lipid-laden "foamy" macrophages that persist at the lesion core, never resolve, and instead sustain the surrounding fibroblasts in a chronic, scar-forming state. We think this intracellular lipid burden is not unique to macrophages: microglia and astrocytes within and around the lesion take on a similar lipid-laden phenotype, and we are investigating how this shared metabolic state shapes the behavior of each cell type and their collective contribution to scar formation. Using single-cell and spatial transcriptomics, high-content imaging, and targeted signaling studies, we are mapping the pathways — including mTOR- and kinase-driven lipid handling programs — that lock these cells into this stalled, pro-fibrotic phenotype.

Beyond dissecting individual mechanisms, we use single-cell and spatial transcriptomics to build a molecular atlas of the injured spinal cord — capturing how every major cell population changes in identity and behavior over the course of injury and repair. We see this atlas not just as a mechanistic resource but as a foundation for therapeutic discovery, helping identify and prioritize candidate targets for intervention.