SaCas9 base editing as a treatment strategy for Rhodopsin-associated retinitis pigmentosa

Abstract

Purpose : Gene replacement therapy has shown great promise in treating inherited retinal dystrophies, but its application to dominant diseases is limited. CRISPR-Cas9 base editing presents an alternative approach to correct dominant single nucleotide variants. We tested in vitro adenine and cytosine base editors (ABE, CBE) utilizing S. aureus Cas9 to treat the dominant RHO c.541G>A (p.E181K) mutation.

Methods : Available PAM sites were searched for the SaABE8e (Addgene #138500), KKH-SaABE8e (#138502) and the CBE SaBE4 (#100805), with single guide RNAs C[G>A]AGGGCCTGCAGTGCTCGTGtggaat and CAGGTACATCCCC[G>A]AGGGCCtgcagt; PAM sites in lower case letters. Transfections in HEK293 or patient fibroblasts with 700 ng base editor and 200 ng guide were performed using FuGene or Neon Electroporation kit. Editing efficiencies were determined 120h post-transfection by analyzing the sequencing data with EditR. Non-transfected cells and published guides were used as negative and positive controls.

Results : The RHO gene was successfully edited in HEK293 cells using SaABE8e and SaKKH-ABE8e constructs. For the SaABE8e, 45% editing efficiency was observed in guide position 12 and the SaKKH-ABE8e variant achieving 22% and 26% editing efficiencies of adenines at positions six and eight, respectively. However, editing at the adenine adjacent to the mutation site in each target sequence was not detected. The SaBE4 was confirmed to be active, achieving 19% editing with a published guide. Attempts were then made to introduce a mutant-allele specific stop codon by converting the C at guide position 11 into a T, thus converting a glutamine codon “CAG” into a stop codon “TAG”. Despite validation of both the SaBE4 construct and the guide, editing at the RHO locus was not detected in either HEK293 cells or patient fibroblasts.

Conclusions : The tested ABEs produced high editing efficiencies of the RHO gene. However, absence of editing at the adenine adjacent to the mutation site suggested the target base does not fall within the constructs’ editing window. Further optimizations for the guides or choosing alternative ABE with different PAM site are needed to correct the RHO c.541G>A mutation. The tested CBE did produce editing with a published guide, but need further optimization to efficiently edit the RHO locus.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.