Towards defining fibroblast phenotypes in cutaneous scarring: CD90 (Thy-1), CD34 and SMA expression in dermal scar fibroblasts
Ho, Jonathan Dale
MetadataShow full item record
BACKGROUND: Cutaneous scarring is a reparative response to wounding in an attempt to restore homeostasis. Pathologic scars include hypertrophic and keloidal scars. No defined dermal scar fibroblast phenotype has been described. This study examines for such a phenotype, looking at expression patterns and spatial relationships of CD90, CD34 and smooth muscle actin (SMA) expressing fibroblasts in cutaneous scars. Additionally, this work investigates for evidence of scar fibroblast transition from the background CD34+ stromal cell network. It also delineates a timeline for the appearance/disappearance of this phenotype in physiologic scarring. Finally, it assesses the relative contributions of CD90+ and SMA+ fibroblasts to scar collagenization. METHODS: 117 scars were classified as reparative (n=47), hypertrophic (n= 40) or keloidal (n=30). Where possible, scar age was calculated. Immunohistochemistry with CD90, CD34 and SMA was performed on all scars. Double-staining immunohistochemistry for CD90/CD34 was applied to all scars assessing for the presence of dual CD90+/CD34+ transitioning cells. Double-color immunofluorescence was also performed to further identify transition. A subset of scars was double stained with CD90/SMA to evaluate spatial relationships. Additional scars were double-stained with CD90/procollagen-1 or SMA/procollagen-1 to assess for active collagen synthesis. Expression was graded as diffuse, focal/rare (i.e. minority) and negative. RESULTS: A CD90diffuse/CD34negative/minority pattern was the most commonly observed phenotype among all scars. SMA expression was variable. Transitioning CD90+/CD34+ fibroblasts were observed in 90.6% of scars. In reparative scars, a CD90diffuse/CD34negative/minority phenotype was time-limited, developing within 48 hours and reverting to a CD34diffuse state at 160-180 days. Many pathologic scars exhibited prolonged CD90diffuse expression. Both CD90+ fibroblasts and myofibroblasts express procollagen-1. CD90+ fibroblasts contributed more cells to scar mass than myofibroblasts. When spatial relationships were examined, myofibroblasts exclusively localized to CD90+ areas and exhibited CD90 double-positivity. CD90 expression was not limited to SMA+ zones. CONCLUSIONS: Scar fibroblasts predominantly exhibit a CD90diffuse/CD34negative/minority phenotype. These CD90+ fibroblasts likely transition from the background CD34+ network. This phenotype is reversible in reparative scars, but is prolonged in some pathologic scars. Both CD90+ fibroblasts and myofibroblasts collagenize scars. The co-localization of myofibroblasts to CD90-rich areas and CD90 dual-positivity may suggest a common origin.