Extended Data Fig. 1: Identification of abasic crRNAs reducing SpCas9-mediated off-target cleavage in S. pyogenes.
From: Abasic CRISPR RNAs inherently harness fidelity of SpCas9 for genome editing
a, b, Genome-wide analysis of OTs in S. pyogenes (n = 107 strains); 481 crRNAs, located < 6 kb from Cas9 loci (mean = 4.5 per strain; a). Among 171 unique spacers, OTs were ordered by off-target count (b). Observed OTs (≤3MM = 13, ≤4MM = 48, ≤5MM = 235, ≤6MM = 1,607) exceeded expected frequencies (≤3MM = 1, ≤4MM = 7, ≤5MM = 73, ≤6MM = 568). c, d, in vitro SpCas9 assays for OTs (≤4MM) using synthetic crRNAs; blue, cleaved OTs; grey triangle, substrate; black triangle, product; cleavage ratio (%). OTs were named by assigning number of mismatch followed by unique alphabetical order (lower case). e–h, In vitro SpCas9 assays as in (c, d) except for OTs (≤6MM) from SF370 (crRNA1 (Spyo1h_001), e; crRNA4 (Spyo1h_004), f; crRNA6 (Spyo1h_006), g) and from other strains harboring the same crRNA spacer sequence as the Bruno strain (crRNAb, h). i, In vitro SpCas9 assays with small RNA fractions (SF370; cold RNA extraction25 to preserve crRNA-tracrRNA complex), reacted with supplemented tracrRNA to prevent defective complex dissociation (via masking potential cross-hybridization); on-targets for crRNA1 (T1), crRNA4 (T4), and crRNA6 (T6). crRNA6 showed marginal activity (0.9%), so only crRNA1 and crRNA4 were analyzed further. j, Cleavage efficiency (%) was quantified by capillary electrophoresis (Fragment analyzer) using concentrated samples for off-targets (for example, 9x). k, Endogenous crRNA estimated at ~0.4–1.2 nM in 2.5 µg small RNA. Assays with 0.1–1.0 nM synthetic crRNAs showed that 0.25 nM synthetic crRNA matched 2.5 µg small RNA on-target activity; cleavage efficiency (%) = cleaved product/total substrate (below gels). l, Stepwise RNA oxidation from all bases to abasic modification (Ø), with preferential guanine and cytosine oxidation (o8G and h5C) leading to G > T and C > T mutations in cDNA via the A-rule (Ø•A pairing during reverse transcription). m, ELISA quantification of abasic modifications in small RNA (Bruno); anaerobic (n = 3) versus aerobic cultures (n = 4),*P = 1.7 x 10−5. n, Dot blots of abasic small RNA (ARP) performed as in Fig. 1m, except SF370. o, Relative ratio of cDNA variant (%) in in vitro oxidized (Fenton reaction) synthetic crRNAs for SF370 (crRNA 1, 4, and 6; n = 2 for 0.1 mM and 0.5 mM H2O2) and crRNAb for Bruno (n = 7 for 0.1 mM (n = 3), 0.5 mM (n = 3), and 2 mM H2O2 (n = 1)), relative to non-oxidized synthetic crRNAs (Oxi/Non); n = 7; *P = 9.9 ×10−5 and 1.5 x 10−4, respectively (Oxi/Non > 2); relative to no change (Oxi/Non = 1). p, In vitro SpCas9 assays comparing between oxidized and non-oxidized synthetic crRNAs at varying H2O2 (2 mM, upper panels; 0.5 mM, lower panels) and crRNA concentrations (50 nM, upper panel; 10 nM, lower panel); normalized cleavage (cleavage ratio normalized to on-target); ratio of normalized cleavage (oxidized/non-oxidized). Of note, similar results were observed across H2O2 and crRNA concentrations, indicating that the reduced off-target cleavage is due to a specific mechanistic effect rather than impaired crRNA function. q, Abasic-induced cDNA variants including incomplete 3’ end of cDNA sequences, analysed using synthetic RNAs with abasic modification at position 7 (7Ø; left and middle panels) and position 5 (5Ø BamHI; right panel) and o8G at position 7 (7o8G), relative to unmodified (7 G). Of note, no deletions were detected at 7o8G, in contrast to 7Ø, 5Ø BamH1 and 5Ø RNA. Unless stated otherwise, P values are two-sided (t-test), n values denote biologically independent samples, and data represent mean ± s.d.
