Human Papillomavirus Genome-Wide Identification of T-Cell Epitopes for Peptide Vaccine Development Against Cervical Cancer: An Integration of Computational Analysis and Experimental Assay.

Human papillomavirus (HPV) has long been documented as the primary factor causing cervical cancer and other complications, and development of immunotherapeutic vaccines against HPV is thought to be an important approach in preventing women from HPV infections. It is known that the first step in vaccine development is to find potent T-cell epitopes in HPV proteins that can be effectively recognized and presented by the human leukocyte antigen (HLA) system. In the current study, we proposed a synthetic pipeline that integrates computational analysis and experimental assay to discover new peptide epitopes from HPV genome with high affinity to the HLA-A*0201, one of the most frequent HLA allele in Caucasian and Asian populations. In the procedure, a structure-based three-dimensional quantitative structure-activity relationship (3D-QSAR) methodology was described and several 3D-QSAR predictors were established using a set of activity-known HLA binders. The best predictor was then employed to perform extrapolation over the HPV genome to screen potential protein fragments with high HLA binding potency. Consequently, 10 peptides were suggested as promising candidates and their affinities toward HLA-A*0201 were assayed using a standard T2 cell surface stabilization test. Four peptides--LLITSNINA from protein E1 (BL50 = 7244 nM), VLLCVCLLI from protein E5 (BL50 = 9118 nM), VLLLWITAA from protein E5 (BL50 = 3388 nM), and LLMGTLGIV from protein E7 (BL50 = 5500 nM)--were identified as high-affinity binders. Further, the structural basis and binding mode of HLA-A*0201-LLITSNINA complex was examined in detail, revealing a complicated network of nonbonded interactions across the complex interface that should render high stability and specificity for the interaction system.