dCas9 binding inhibits the initiation of base excision repair in vitro

Abstract

Cas9 targets DNA during genome editing by forming an RNA:DNA heteroduplex (R-loop) between the Cas9-bound guide RNA and the targeted DNA strand. We have recently demonstrated that R-loop formation by catalytically inactive Cas9 (dCas9) is inherently mutagenic, in part, by promoting spontaneous cytosine deamination within the non-targeted single-stranded DNA of the dCas9-induced R-loop. However, the extent to which dCas9 binding and R-loop formation affect the subsequent repair of uracil lesions or other damaged DNA bases is unclear. Here, we show that DNA binding by dCas9 inhibits initiation of base excision repair (BER) for uracil lesions in vitro. Our data indicate that cleavage of uracil lesions by Uracil-DNA glycosylase (UDG) is generally inhibited at dCas9-bound DNA, in both the dCas9:sgRNA-bound target strand (TS) or the single-stranded non-target strand (NT). However, cleavage of a uracil lesion within the base editor window of the NT strand was less inhibited than at other locations, indicating that this site is more permissive to UDG activity. Furthermore, our data suggest that dCas9 binding to PAM sites can inhibit UDG activity. However, this non-specific inhibition can be relieved with the addition of an sgRNA lacking sequence complementarity to the DNA substrate. Moreover, we show that dCas9 binding also inhibits human single-strand selective monofunctional uracil-DNA glycosylase (SMUG1). Structural analysis of a Cas9-bound target site subsequently suggests a molecular mechanism for BER inhibition. Taken together, our results imply that dCas9 (or Cas9) binding may promote background mutagenesis by inhibiting the removal of DNA base lesions by BER.

Keywords: CRISPR; DNA glycosylase; DNA targeting; R-loop; Repair.

Figure 1:. UDG cleavage in dCas9-bound substrate

(A/B) Diagram of the 45 bp DNA substrates containing uracil lesions. Each diagram shows the location of sgRNA-bound DNA region and the location of Uracil lesions relative to their distance away from the PAM motif. (A) U5-NT indicates that the Uracil lesion is 5 bp away from the PAM motif on the single stranded non-target strand. (B) U4-TS indicates that the Uracil lesion is 4 bp away from the PAM motif on the sgRNA-targeted strand. A star indicates that fluorescent labels (6-FAM) were placed on the 5’ end of the strand containing the lesion. (C/D) Denaturing Urea-polyacrylamide gel electrophoresis was used to measure UDG glycosylase activity for (C) U5-NT and (D) U4-TS. (E/F) Graph showing substrate cleavage by UDG after 15 minutes for (E) U5-NT and (F) U4-TS. Columns represent the averages of 4 independent experiments and errors bars represent standard error of mean (SEM). A single asterisk (*) denotes a significant difference (P < 0.05) between control conditions and the dCas9:sgRNA complex bound DNA (indicated by dCas9+sgRNA) with UDG as determined by t-test with Holm-Sidak correction for multiple hypothesis testing.

Figure 2:. Lesions must be within the dCas9 bound target site for dCas9+sgRNA to inhibit UDG cleavage activity

(A) Diagram of the 45 bp DNA substrate for Uracil lesions outside of the sgRNA-targeted region. U24-NT indicates that the Uracil lesion is 4 bp outside the sgRNA-targeted strand. A fluorescent label (6-FAM), indicated by a star, was placed on the 5’ end of the strand containing the lesion. (B) Denaturing Urea-polyacrylamide gel electrophoresis was used measure glycosylase activity of UDG on DNA for each reaction after 15 min at 37 °C. Inhibition of UDG activity by dCas9 can be determined by loss of UDG cleavage product. (C) Quantification of UDG activity after 15 minutes. Columns represent the average of 4 independent experiments and errors bars represent standard error of mean (SEM). A single asterisk denotes a significant difference (p < 0.05) between control conditions and the dCas9 alone condition as determined by t-test with Holm-Sidak correction for multiple hypothesis testing. dCas9+sgRNA indicates that the dCas9:sgRNA complex is bound to target dsDNA.

Figure 3:. Inhibition of UDG for lesions further away from the PAM motif

(A/B) Diagram of the 45 bp DNA substrates for Uracil lesions further away from the PAM motif. The corresponding 5’- 6-FAM fluorescent label, indicated by a star, is on the strand containing the lesion. The grey box indicates the sgRNA targeting region for (A) U11-TS and (B) U17-TS. (C/D) Graph showing substrate cleavage by UDG after 15 minutes for (C) U11-TS and (D) U17-TS. Columns represent the averages of either 4 or 5 independent experiments. Errors bars represent standard error of mean (SEM). A single asterisk (*) denotes a significant difference (P < 0.05) as determined by t-test with Holm-Sidak correction for multiple hypothesis testing. dCas9+sgRNA indicates that the dCas9:sgRNA complex is bound to target dsDNA while dCas9 + NT sgRNA indicates that dCas9:sgRNA complex has a non-targeting (NT) sgRNA which lacks sequence complementarity with the target substrate and should not bind target dsDNA.

Figure 4:. Lesions within or adjacent to the base editing window are more permissive to UDG activity

(A/B) Diagram of the 45 bp DNA substrates which place the uracil lesion within or adjacent to the base editing window on the displaced non-target strand of the dCas9-induced R-loop. A star indicates that a 5’- 6-FAM fluorescent label is on the strand containing the lesion (A) U10-NT (B) U15-NT (C/D) Quantification of UDG activity after 15 minutes for (C) U10-NT and (D) U15-NT. For U15-NT, dCas9+sgRNA indicates that dCas9 is complexed with a targeting sgRNA with full complementarity to the DNA substrate while dCas9 +NT sgRNA indicates a non-targeting (NT) sgRNA, lacking sequence complementarity to the DNA substrate, is complexed with dCas9. Columns represent the averages of either 4 or 5 independent experiments. Errors bars represent standard error of mean (SEM). A single asterisk (*) denotes a significant difference (P < 0.05) as determined by t-test with Holm-Sidak correction for multiple hypothesis testing.

Figure 5:. SMUG1 is also inhibited by RNA-directed dCas9 DNA binding.

(A) Model of the 45 bp U5-NT DNA substrate. (B) Denaturing Urea-polyacrylamide gel electrophoresis was used to show DNA cleavage activity by hSMUG1 for each dCas9+sgRNA condition after 30 min at 37 °C. dCas9+sgRNA indicates that DNA is bound by the dCas9:sgRNA complex. The efficiency of SMUG1 cleavage was used to determine the extent to which dCas9 impairs SMUG1 repair efficiency and thus accessibility to the lesion. (C) Quantification of SMUG1 activity after 30 min for various dCas9+sgRNA conditions. Columns represent the average of 4 independent experiments and errors bars represent standard error of mean (SEM). A single asterisk (*) denotes a significant difference (P < 0.05) between control conditions and the dCas9:sgRNA complex bound DNA (indicated by dsDNA + dCas9:sgRNA) as determined by t-test with Holm-Sidak correction for multiple hypothesis testing.

Figure 6:. Accessibility of uracil lesions within a crystal structure of Streptococcus pyogenes Cas9 complexed with sgRNA and dsDNA.

(A) Diagram showing a typical base editing window (in red box) around 12–18 bp away from the PAM motif (the yellow box) on the single stranded non-target strand of the dCas9-induced R-loop. (B) Crystal structure of Cas9 (grey)-sgRNA (blue)-dsDNA (light blue) complex (PDB: 5F9R) showing where U5-NT (red sphere) is located within the Cas9-induced R-loop. PAM motif (5’-NGG-3’) is shown in yellow. Hypothetical positions for U10-NT and U15-NT indicated by ? are shown in the PAM-distal region (10–20 nucleotides away from the PAM motif, light blue dashed lines) on the non-target strand (NT), as these are not shown in the crystal structure. (C) Shows where U4-TS (orange sphere), U11-TS (green sphere), and U17-TS (purple sphere) are located within the Cas9-induced R-loop on PDB: 5F9R. (D) Close-up view of each lesion individually. U5-NT (red sphere), U4-TS (orange sphere), U11-TS (green sphere), U17-TS (purple sphere), PAM motif (in yellow).

Figure 7:. Model showing UDG activity for lesions located throughout a dCas9-bound R-loop.

dCas9 is indicated by the light blue protein covering the black lines which indicate the DNA substrate. The yellow box on the DNA substrate indicates the location of the NGG PAM motif. Uracil lesion locations are color-coordinated. Green boxes and text (in top left) indicates lesions were located outside the target region (U24-NT), Red boxes and text (in bottom left) indicates that lesions were located on the sgRNA targeted strand (U4-TS, U11-TS, U17-TS). Black boxes and text (in bottom right) indicate naked DNA controls for each substrate. Whereas lesions on the displaced single-stranded DNA of the dCas9-induced R-loop (in top right) are indicated in blue boxes and text (for lesions in or adjacent to the base editing window; U15-NT and U10-NT, respectively) or brown boxes and text (for lesions outside the base editing window, U5-NT). Relative degree of UDG inhibition for each substrate region is shown: No UDG inhibition representing efficient cleavage activity (i.e. High UDG activity), Moderate UDG inhibition represents semi-efficient cleavage activity (i.e. repair is reduced or partially inhibited) and High UDG inhibition represents inefficient cleavage activity (i.e. repair activity is severely inhibited).