Text

Objectives and questions: Allogeneic skin coverage is often required after extensive burns/trauma when donor sites are limited, but immune rejection shortens graft survival. BMSCs are immunomodulatory yet may be insufficient at standard dosing. COX-2 enhances PGE2 production and supports Treg-mediated tolerance. We investigated whether COX-2 gene engineering enhances BMSC immunomodulation to suppress allo-reactive T-cell responses, shift the Treg/Th17 balance toward tolerance, and prolong skin allograft survival after scald injury.

Material and methods: Murine BMSCs (Balb/c) were isolated and genetically modified for COX-2 overexpression or siRNA knockdown; COX-2 expression was verified by qPCR/Western blot and PGE2 secretion by multiplex assays. Allo-reactive CD4+CD62L+ T-cell proliferation was assessed by flow cytometry in a dose-response co-culture system (BMSC:effector ratios 1:20, 1:10, 1:5; n=5/group). Mixed lymphocyte reaction (MLR) quantified Tregs and cytokines (PGE2, IL-10, IL-6, IL-17A). In vivo, C57BL/6J recipients received IV BMSCs for 4 consecutive days (total 1×10^6 cells/mouse), followed by a standardized 90°C scald injury (7 s) and skin grafting (1×1 cm). Groups: autograft, allograft control (Balb/c), allograft + unmodified BMSCs, allograft + COX-2–BMSCs, and COX-2–BMSCs + third-party allograft (C3H) (n=20/group). Graft survival was analyzed by log-rank testing; day-8 samples underwent histology, flow cytometry, and multiplex cytokines (p<0.05).

Results: COX-2 overexpression increased PGE2 secretion, whereas knockdown reduced it. In allo-reactive proliferation assays, stimulated T-cell proliferation was 54.20±2.23%. Unmodified BMSCs reduced proliferation to 42.21±1.23%, 33.53±3.26%, and 25.23±2.68% (1:20–1:5), while COX-2–BMSCs further suppressed proliferation to 36.15±1.56%, 25.66±2.32%, and 15.13±2.75% at matched ratios (all p<0.05 vs unmodified). In MLR, Tregs increased from 1.22±0.25% (control) to 5.77±1.45% (BMSCs) and 8.76±1.56% (COX-2–BMSCs) (p<0.05), accompanied by increased IL-10 and reduced IL-6/IL-17A. In the scald-plus-grafting model, mean graft survival was 10.6±0.55 days (allograft control), 13.4±1.14 days (BMSCs), and 14.6±1.82 days (COX-2–BMSCs), while autografts survived throughout the 30-day observation. COX-2–BMSCs increased Tregs (20.8±2.5% vs 15.5±3.5% with unmodified BMSCs) and shifted systemic cytokines toward tolerance (IL-10: 23.06±3.32 vs 14.22±2.35; IL-6: 37.07±8.95 vs 87.36±5.78; IL-17A: 19.23±2.56 vs 44.57±1.05; all p<0.05).

Discussion and conclusions: COX-2 gene engineering enhances BMSC-mediated immunomodulation, suppressing allo-reactive T-cell proliferation and promoting a tolerance-favoring Treg/Th17 shift, consistent with a PGE2/IL-10–linked mechanism. In a clinically relevant scald-plus-grafting model, COX-2–BMSCs significantly prolonged skin allograft survival and reduced rejection-associated inflammation. This cell-and-gene therapy strategy may extend temporary allograft coverage time after severe burns/trauma and serve as a bridge to definitive autologous reconstruction.