From genome to vaccine: in silico predictions, ex vivo verification

Bioinformatics tools enable researchers to move rapidly from genome sequenc e to vaccine design. EpiMer and EpiMatrix are computer-driven pattern-match ing algorithms that identify T cell epitopes. Conservatrix. BlastiMer, and Patent-Blast permit the analysis of protein sequences for highly conserved regions, for homology with other known proteins, and for homology with prev iously patented epitopes, respectively. Two applications of these tools to epitope-driven vaccine design are described in this review. Using Conservat rix and EpiMatrix, we analyzed more than 10000 HIV-1 sequences and identifi ed peptides that were potentially immunostimulatory and highly conserved ac ross HIV-1 clades. MHC binding assays and CTL assays have been carried out: 50 (69%) of the 72 candidate epitopes bound in assays with cell lines expr essing the corresponding MHC molecule, 15 of the 24 B7 peptides (63%) stimu lated gamma-interferon release in ELISpot assays. These results lend suppor t to the bioinformatics approach to selecting novel. conserved, HIV-1 CTL e pitopes. EpiMatrix was also applied to the entire 'proteome' derived from t wo Mycobacterium tuberculosis (Mtb) genomes. Using EpiMatrix. BlastiMer, an d Patent-Blast, we narrowed the list of putative Mtb epitopes to be tested in vitro from 1 600000 to 3000. a 99.8%, reduction. The pace of vaccine des ign will accelerate when these and other bioinformatics tools are systemati cally applied to whole genomes and used in combination with in vitro method s for screening and confirming epitopes. (C) 2001 Published by Elsevier Sci ence Ltd.