Abstract
Purpose:
We have recently identified the retinal pigment epithelium (RPE) as the site of pathogenesis in three mutant mouse models of RNA splicing retinitis pigmentosa (RP). However, these results need to be replicated in the human RPE. With the advent of CRISPR/Cas9 genome editing techniques, we can now develop human cell line models for these forms of disease. Here, we present the use of CRISPR/Cas9 genome editing to knockout PRPF31 and characterize the effect on RPE function.
Methods:
We designed a guide RNA to exon 7 of PRPF31, and co-transfected it with a wild type Cas9 co-expressing GFP into ARPE-19 cells. GFP positive cells were single cell sorted into a 96 well plate and grown to confluence. DNA was isolated from each clone and the region flanking the predicted cut site was Sanger sequenced to identify those that exhibit correct cutting and non-homologous end joining (NHEJ). Five NHEJ lines were selected for further characterization using both qRT-PCR and phagocytosis assays to quantify FITC-labeled photoreceptor outer segment uptake with flow cytometry.
Results:
Approximately 25% of the individual clones validated following transfection showed NHEJ with deletions between 2 and 11 bases and one clone had a 1 base insertion. We identified only heterozygous indels, consistent with previous reports that mutations in PRPF31 cause disease via haploinsufficiency. Expression of PRPF31 in 4 of the 5 genome edited clones was significantly reduced by 50-80%, as compared to the wild-type control. To confirm these changes were a result of genome editing, we determined expression levels of the PRPF31 modifier CNOT3. We found that one line had a 2-fold increase in expression, which may explain reduced levels of PRPF31 in that line. Flow cytometry analysis of POS uptake demonstrated phagocytosis was reduced by 10-60-fold in the genome edited lines.
Conclusions:
Currently, it is difficult to study the disease mechanism of RNA splicing factor RP in human models. We have created a human cell line model for PRPF31-associated disease that mimics findings in mouse models. These lines will allow us to study the disease in a more relevant model, affording us the capability to interrogate splicing more deeply. Further, we can begin to study the effect of AAV-mediated gene augmentation of PRPF31 on disease pathogenesis and rescue of functional deficiencies.