Fig. 1: Prediction of SaCas9 and AsCas12a epitopes with reduced binding to MHC I molecules.
From: Rational engineering of minimally immunogenic nucleases for gene therapy
a Schematic of MAPPs analysis to identify epitopes from SaCas9 and AsCas12a that bind to MHC I molecules. b Computational workflow to nominate mutations predicted to abrogate epitope binding to MHC I molecules while maintaining nuclease function. Crystal structures were used to create all-atom protein models in Rosetta. Epitope regions identified in MAPPs were targeted for mutational analysis, along with adjacent N-terminal and C-terminal subsequence frames to ensure that new epitopes were not created for any overlapping peptide subsequences. A computational protein design method utilized 14 MHC Class I PSSM models to introduce mutations predicted to eliminate MHC binding of epitope peptides while avoiding the creation of new predicted epitopes and maintaining predicted protein stability. Final models were evaluated using NetMHCpan and Rosetta. c Location of immunogenic epitopes on SaCas9 (left) and AsCas12a (right). d Sequences of immunogenic epitopes. Domain architecture of SaCas9 (left) and AsCas12a (right) with catalytic sites shown in red above and location of immunogenic epitopes indicated below. Sequences of immunogenic epitopes and proposed single amino acid mutations for each epitope are listed below R-I RuvC-I, REC recognition domain, R-II RuvC-II, HNH HNH nuclease, R-III RuvC-III, WED wedge domain, PI PAM-interacting domain.
