Abstract
Anti-programmed death ligand 1 (PD-L1) therapy has been beneficial in treating patients with certain cancers. Here, we tested whether anti-PD-L1 therapy is effective for controlling different types of tumors using animal models of TC-1, MC38 and B16. We found that, despite PD-L1 expression, anti-PD-L1 therapy showed little and some antitumor activity in the TC-1 and MC38 models. However, anti-PD-L1 therapy exhibited a more dramatic antitumor effect in the B16 model. This difference in antitumor responses was likely associated with the CD8 + T cell infiltration status of tumor tissues. In the B16 model, CD8 + T cells and to a lesser degree NK cells were found to be responsible for the antitumor response of anti-PD-L1 therapy, as determined by immune cell subset depletion. In particular, CD8 + T cells from B16-bearing mice produced an IFN-γ in response to B16 cells and citrate phosphate buffer-treated B16 cell peptide elutes but not to an immunodominant class I epitope, Trp2180-188, suggesting that CD8 + T cells that recognize neoantigens were induced in B16 tumor-bearing mice and then reactivated by anti-PD-L1 for tumor control. When B16 tumor-bearing mice were treated with anti-PD-L1 in combination with Trp2180-188 peptide vaccines, they displayed significantly more tumor control than either single therapy. Taken together, these studies show that B16 melanomas are more effectively controlled through reactivation of tumor-infiltrating lymphocytes by anti-PD-L1 therapy. Moreover, combined therapy using anti-PD-L1 and Trp2 peptide vaccines is more beneficial for controlling B16 melanomas through reactivation of neoantigen-specific CD8 + T cells and induction of Trp2-specific CD8 + T cells.