RNA editing as a therapeutic approach for retinal gene therapy requiring long coding sequences

Abstract

RNA editing aims to treat genetic disease through altering gene expression at the transcript level. Pairing site-directed RNA-targeting mechanisms with engineered deaminase enzymes allows for the programmable correction of G>A and T>C mutations in RNA. This offers a promising therapeutic approach for a range of genetic diseases. For inherited retinal degenerations caused by point mutations in large genes not amenable to single-adeno-associated viral (AAV) gene therapy such as USH2A and ABCA4, correcting RNA offers an alternative to gene replacement. Genome editing of RNA rather than DNA may offer an improved safety profile, due to the transient and potentially reversible nature of edits made to RNA. This review considers the current site-directing RNA editing systems, and the potential to translate these to the clinic for the treatment of inherited retinal degeneration.

Keywords: ADAR; CRISPR; RNA editing; base editing; gene editing; gene therapy; genome engineering; inherited retinal degeneration; retinal disease; retinitis pigmentosa.

Figure 1

Immunofluorescence section demonstrating the anatomy of a mouse retina. Inherited retinal degenerations affect the cells of the outer retina, particularly the photoreceptors (rods and cones) and the retinal pigment epithelium, highlighted in red text. The inner (green) and outer (red) segments of the photoreceptors are demonstrated with their nuclei stained above in blue.

Figure 2

(A). Demonstrates the hydrolytic deamination of adenosine to inosine. (B) Base pairing between inosine and cytosine allow inosine to be read as a guanosine during translation by the ribosome, effectively an A-G edit. (C). Evolved mutants of human ADAR allow for deamination of cytosine to uracil for C-U editing. ADAR = adenosine deaminase acting on RNA; DD = deaminase domain.

Figure 3

Schematic figures demonstrating the RNA editing approaches of (A) BoxB-λN-ADAR with a dual BoxB design to recruit the λN peptide (B) SNAP-tag-ADAR fusion with a O6-benzyl-guanine (BG) conjugated adRNA (C) Glu-adRNA approach where the GluR2 R/G hairpin recruits exogenous or endogenous full length ADAR (D) MS2-MCP-ADAR stem-loop approach with dual MS2 stem-loop hairpins recognized by the MS2 bacteriophage coat binding protein (MCP) (E) An example of endogenous ADAR recruitment using long LEAPER-arRNAs and G-A mismatches within the guide region to reduce off-target editing (F) The REPAIR system with dPspCas13b-ADAR fusions recruited by a guide RNA with a direct repeat. Note the A-C mismatch at the target adenosine in each strategy to promote editing at the target site. gRNA = guide RNA; adRNA = ADAR guiding RNAs; λN = Lambda N protein; LEAPER = Leveraging Endogenous ADAR for Programmable Editing of RNA; REPAIR = RNA Editing for Programmable A to I Replacement; dPspCas13b = deactivated Cas13b orthologue derived from Prevotella sp. P5-125.

Figure 4

The distribution of unique “editable” mutations within all variants classed as “pathogenic” or “likely pathogenic” within the ClinVar Database. This includes all mutations including insertions, deletions and copy number variation, and these are included in “other mutations”. Blue bars represent G>A variants, while pink bars represent T>C variants. Accessed 3 October 2019.

Figure 5

An approach to RNA editing in a patient with Usher Syndrome type 2 with compound heterozygous mutations in USH2A (c.2299delG / c.11864G>A). (A) Fundus photographs and (B) fundus autofluorescence images demonstrating pigmentary retinopathy and narrowing of the retinal vasculature. (C) Following expression of an USH2A cDNA fragment in HEK293T cells, the 11864G>A (p.Trp3855Ter) mutation can be partially repaired with an efficiency of 43% to restore the premature termination codon to tryptophan. The target adenosine is highlighted in grey. (D) Repair of the target is not observed without expression of the REPAIRv2 (dPspCas13b-hDAR2_DD(E488Q/T375G)) plasmid.