Heavily and fully modified RNAs guide efficient SpyCas9-mediated genome editing

Abstract

RNA-based drugs depend on chemical modifications to increase potency and to decrease immunogenicity in vivo. Chemical modification will likely improve the guide RNAs involved in CRISPR-Cas9-based therapeutics as well. Cas9 orthologs are RNA-guided microbial effectors that cleave DNA. Here, we explore chemical modifications at all positions of the crRNA guide and tracrRNA cofactor. We identify several heavily modified versions of crRNA and tracrRNA that are more potent than their unmodified counterparts. In addition, we describe fully chemically modified crRNAs and tracrRNAs (containing no 2'-OH groups) that are functional in human cells. These designs will contribute to Cas9-based therapeutics since heavily modified RNAs tend to be more stable in vivo (thus increasing potency). We anticipate that our designs will improve the use of Cas9 via RNP and mRNA delivery for in vivo and ex vivo purposes.

Fig. 1

Initial screening of chemical modifications in the crRNA. a Schematic of Cas9 RNP paired with target DNA. The secondary structure elements of crRNA and tracrRNA are labeled. RNA is shown in orange, whereas DNA is in gray. The PAM sequence is highlighted red and cleavage sites are marked with arrows. b Chemical modifications used in this study. c Bar graph showing mCherry-positive cells after electroporation of HEK293T-TLR cells with RNPs that included the indicated crRNAs and an unmodified tracrRNA. Error bars represent standard deviation (SD) resulting from at least three biological replicates

Fig. 2

Second round of chemical optimization of CRISPR RNAs. The optimized crRNA designs are shown in a, whereas chemical designs of tracrRNA are shown in b. Each crRNA was tested with the unmodified tracrRNA T0, whereas each tracrRNA was tested with the unmodified crRNA C0. The bar graphs show the percent of cells expressing mCherry, ±SD. Each RNA was tested in triplicate

Fig. 3

Cas9 tolerates heavily and fully modified crRNA:tracrRNA. a, b Each crRNA was tested with tracrRNAs T0, T2, and T6–T8 using 20 pmol of Cas9 RNP. c HEK293T-TLR cells were also electroporated with 100 pmol of the indicated RNPs to test whether heavily modified RNAs regain functionality at higher doses. Error bars show ± SD of three biological replicates

Fig. 4

Targeting endogenous genes with modified RNAs. a The C10 guide design targeting HTT exon 50 was tested using 20 pmol of RNP along with T2 and T6–T8 in HEK293T cells. b HTT-C20 and HTT-C21 designs were tested using 80 pmol of Cas9 RNP with the indicated tracrRNAs. c The therapeutically relevant HBB locus was targeted using a previously validated guide sequence incorporated into the C20 design. d, e VEGFA-targeting crRNAs C20 and C21 were tested using the indicated tracrRNAs with 80 pmol of RNP in HEK293T cells (d) or 60 pmol of RNP in hESCs (e). Indels were calculated using TIDE. Bars show averages (±SD) of at least three biological replicates

Fig. 5

Heavily and fully modified RNAs support precise genome editing. a Schematic of the Cas9-mediated homology-directed repair in HEK293T-TLR cells. Cells were electroporated with 20 pmoles of Cas9 RNP and 400 ng of 800 bp donor (PCR fragment). Precise repair of the DSB results in GFP expression (b), and NHEJ-mediated + 1 frameshifts result in mCherry expression (c). d The fully modified C21:T8 dual guides were tested using 100 pmoles of Cas9 RNP to test whether fully modified RNAs recover HDR activity at higher doses. Bars show averages (±SD) of three biological replicates