Bevacizumab promotes T-cell–mediated collagen deposition in the mouse model of conjunctival scarring

PURPOSE. We determine the effects of bevacizumab on collagen production in a mouse model of conjunctival scarring. METHODS. Experimental surgery was performed as described for the mouse model of conjunctival scarring, and bevacizumab was introduced by conjunctival injection. The capacity of bevacizumab to recognize conjunctival VEGF-A was determined by ELISA. Col1a1 was measured by real-time PCR and immunoblotting. T cells and collagen were visualized by immunofluorescence and picrosirius red staining of bleb cryosections. Conjunctival CD4þ or CD8aþ T cells were counted by flow cytometry. Mouse splenic T cells were cultured with bevacizumab/IgG and their numbers, cell cycle, and collagen production were measured using a cell counter, flow cytometry, and sircol soluble collagen assay, respectively. Reconstitution experiments in severe combined immunodeficiency (SCID) mice were performed by injection of freshly isolated T cells on day 2 postoperatively. RESULTS. Bevacizumab recognized approximately 20% of endogenous murine VEGF-A. Injection of bevacizumab raised Col1a1 expression in the blebs at mRNA and protein levels. Bevacizumab did not induce collagen in conjunctival fibroblasts, but increased CD4þ and CD8aþ cell numbers as well as collagen production by these cells. Collagen appeared to accumulate in the vicinity of T cells in the bevacizumab-treated blebs. While SCID blebs did not show elevated collagen levels, reconstitution with CD4þ or CD8aþ cells resulted in increased Col1a1 expression at mRNA and protein levels. CONCLUSIONS. Bevacizumab increased collagen production in the mouse model of conjunctival scarring. This collagen induction was mediated by T cells that were also stimulated by bevacizumab to increase in numbers.