OR37 A method for assessing immunogenicity of HLA-epitopes to circumvent potential immunotoxicity of therapeutic gene editing

Aim Targeted genome editing has tremendous potential for treating monogenic diseases, especially for those that affect the hematopoietic system. Efforts to reduce the immunogenicity of administered proteins such as Cas9 and Cas12a are critical for ensuring safe CRISPR treatment. Moreover, corrections via gene editing could result in novel HLA-epitopes for host defense. We aim to expand our Minor Histocompatibility Antigen Identification Pipeline (MiHAIP) to automatically assess the immunogenicity of HLA-epitopes for optimization of therapeutic gene editing. Methods Using ClinVar as a source of pathogenic target mutations, MiHAIP simulates antigen processing given certain genetic variants and HLA allele types. NetMHCpan was used to predict the binding affinities of all possible peptides (8- to 11-mers) to the HLA alleles. Thirty-two HLA class I alleles with sufficient known binders as a training set were included in the experiment. The outputs from MiHAIP were used to generate an HLA-peptide-affinity matrix for downstream analyses. Results From 57,484 missense variants in ClinVar, 932,052 binders were returned out of 40,993,568 HLA-peptide combinations. Approximately 31.8% are considered to be the strong binders. Examination of the variants of typical monogenic diseases uncovered alterations in amino acids that commonly affect the immunogenicity of HLA-epitopes. This indicates that editing these variants may turn a non-binder into a binder. In sickle cell anemia, the homozygous allele of rs334(T) encodes valine[V] that affects the disease. However, carriers of rs334(T) and rs334(A/G/C), which encode glutamic acid [E], glycine [G] and alanine [A] respectively, are disease free. Our results show that peptides of TP[E/A/V]EKSAVTAL are strong binders of HLA-B∗07:02, and that editing on both rs334(T > C) and rs334(T > A) could introduce high immunogenicity epitopes. In contrast, editing rs334(T > G) could be a better option if patients have HLA-B∗07:02 allele. This approach enables us to assess the immunogenicity of HLA-epitopes from any variants. Conclusions This is a preliminary strategy paired with an efficient analytical pipeline to evaluate immunogenicity for therapeutic gene editing. It could be extensively applied to screen appropriate editable targets to circumvent immunotoxicity.