Home Team Members Collaboration Attributions Description Mutation-internal Open / Structured Demonstrate Characterisation Improve Basic Parts Part Collection Human Practices Outreach Medal Criteria Judging Form

Demonstrate

Pathogenic Gene Correction in HCT116

The utility of our dCasRx-ADAR2_DD RNA editing platform lies in its ability to correct pathogenic G>A mutations and restore wild type G. Based on our sequencing and luciferase assay data, we have identified 2 dCasRx-ADAR2_DD constructs that have the highest relative efficiency and specificity out of all our constructs, dCasRx-ADAR2_DD(H460D) and dCasRx-ADAR2_DD(D338+H460D). To demonstrate that our editing platform works in a realistic setting, we examined if H460D and D338+H460D were able to correct pathogenic alleles in the colon cancer cell line HCT116. From the database, we chose 3 genes, BCOR, GNAQ and RUNX1, which harbour a G>A mutation.^[1] Apart from single gene edits, we also demonstrated the possibility of multiplex RNA editing with our dCasRx-ADAR2_DD construct. Since both CasRx and dCasRx were reported to be able to process their own crRNA, we created a gRNA array on a single plasmid, which consists of GNAQ, BCOR and RUNX1 genes arranged in tandem with their scaffold.^[2]

Pathogenic score of GNAQ, BCOR and RUNX1 from the COSMIC cancer database.^[3]

Methodology

Results

Figure 1. Representative example of editing in HCT116. Highlighted region indicates targeted editing site. Construct used was CasRx V1.

Figure 2. Percent editing of targeted sites on GNAQ, BCOR and RUNX1 with dCasRx-ADAR2_DD(D338+H460D) and dCasRx-ADAR2_DD(D338+H460D). (A) Editing on GNAQ. (B) Editing on BCOR. (C) Editing on RUNX1.

Discussion and Conclusion

For H460D, there was at least 12% of editing on GNAQ with either single or multi gRNA. It had also edited 7% of the RUNX1 gene with either single or multi gRNA. However, editing on BCOR was only achieved with a single gRNA (~7%) but not the multi gRNA. For D338+H460D, there was also at least 12% of editing on GNAQ, though the single gRNA editing was higher than H460D by ~2%. It had also slightly better editing on RUNX1 (~1.5% more) than H460D. However, editing on BCOR was not observable.

This demonstrated that our constructs were able to successfully edit pathogenic alleles in a model of human disease (colorectal cancer model instead of HEK293FT). We also showed that multiplex of at least 2 genes were possible (GNAQ and RUNX1). This holds great promise in treating diseases caused by multiple genes, such as cancer.

References