With the advent of genome editing using CRISPR-Cas9, researchers have been excited by the possibilities of precisely placed edits in cellular DNA. Any double-stranded break in DNA, like that induced by CRISPR-Cas9, is repaired by one of two pathways: Non-homologous end joining (NHEJ) or homology-directed repair (HDR). Using the NHEJ pathway results in short insertions or deletions (indels) at the break site, so the HDR pathway is preferred. However, the low efficiency of HDR recombination to insert exogenous sequences into the genome hampers its use. There have been many attempts at boosting HDR frequency, but the methods compromise cell growth and behave differently when used with various cell types and gene targets. The strategy employed by the authors of an article in Communications Biology tethered the DNA donor template to Cas9 complexed with the ribonucleoprotein and guide RNA, increasing the local concentration of the donor template at the break site and enhancing homology-directed repair.
Based on previous work by this same lab, specific sequences of unmodified single-stranded DNA (ssDNA) could covalently bond with HUH endonucleases, enzymes that cleave and join ssDNA, and HUH could function when fused with protein partners including Cas9. For this research, the porcine circovirus 2 (PCV) Rep protein was fused to Cas9 at either the amino or carboxy terminus. Both Cas9-PCV fusion proteins bound a fluorescent oligo that contained the PCV recognition sequence; unfused Cas9 did not. Changing the catalytic tyrosine in PCV to phenylalanine (Y96F) also prevented binding of the single-stranded oligodeoxynucleotide (ssODN). A gel shift assay confirmed a change in mobility when an equimolar amount of ssDNA and Cas9-PCV protein formed a protein complex; Cas9 alone did not. Even with PCV appended to Cas9, the enzyme could still cleave DNA.
To monitor HDR efficiency, the small tag HiBiT was inserted into the targeted loci. When complemented with the subunit LgBiT, HiBiT is able to form the NanoBiT® enzyme and generate a luminescent signal in the presence of its substrate. The more often HiBiT is successfully knocked into locus of interest, the greater the intensity of the light signal, giving a relative measure of HDR. When GAPDH was targeted in HEK 293T cells, all versions of Cas9 were able to insert the ssODN without a PCV recognition sequence at similar levels in the cells. However, when the ssODN contained PCV sequence, the Cas9 fusions with PCV showed a two- to threefold increase in luminescence. Cas9 alone and the Cas9-(Y96F)PCV fusion did not change their light signal. These data suggest that having the ssODN attached to Cas9 via PCV increased HiBiT insertion. When using U2-O2 cells and targeting GAPDH, the luminescent profile was similar to that observed for HEK 293T cells. In addition, when targeting vinculin, a gene expressed at lower levels than GAPDH, in HEK 293T cells, the Cas9-PCV fusions showed an increase in luminescence compared to Cas9 alone. Thus, cell type and locus target changed but Cas9-PCV fusions were better able to insert the ssODN with HiBiT.
Can the luminescent signal increase if a higher concentration of Cas9 or donor DNA are added? Regardless of the concentration of the RNP complex, the HDR efficiency was higher for the Cas9-PCV fusions compared to Cas9 alone. The authors noted that there was a greater change in HDR at lower concentrations of Cas9-PCV, seeing a 15- to 30-fold enhancement. When increasing the concentration of the ssODN with PCV recognition sequence, HDR efficiency did not change for Cas9, but maximal signal was seen with a 1:1 ratio of RNP to ssODN for the Cas9-PCV fusions. Again, this supports the idea that the covalent bond between the ssDNA and the PCV fused to Cas9 increases the homology-directed repair of the targeted gene.
To confirm the genomic DNA would show the same effects as those expressed in the proteins, the GAPDH insertion sites were sequenced. HDR efficiency was 5- to 11-fold greater when using ssODN + PCV recognition sequence with Cas9-PCV fusions. One example cited by the authors indicated Cas9 inserted HiBiT into GAPDH correctly with a 0.7% rate, but the carboxy Cas9-PCV fusion was inserted at a rate of nearly 8%. In addition, the HDR to indel formation ratio was two- to threefold higher for the Cas9-PCV fusions compared to Cas9 alone, another indication of greater HDR efficiency.
The data have shown that a covalent bond between the ssDNA template and the Cas9-PCV fusions were able to enhance HiBiT insertion into targeted loci. Could this tethered RNP system also repair a defect in a fluorescent reporter gene integrated in a cell line? A dual fluorescent GFP-mCherry reporter HEK 293T cell line that had a frameshift mutation in the mCherry gene was introduced to an ssODN + PCV recognition sequence that would correct the mutation. Fluorescence was assessed using flow cytometry. HDR efficiency was higher for the Cas9-PCV fusions versus Cas9 alone. The background correction level without any ssODN was the same for all three Cas9 proteins, fused and unfused. Only in the presence of the ssODN was there a significant difference among the RNPs, with greater mCherry expression seen for the Cas9 fusions.
By bringing the ssDNA template and the RNP complex with Cas9 in closer proximity through a covalent bond, Aird et al. were able to improve the HDR efficiency of CRISPR-Cas9 genomic editing by an order of magnitude. In addition, the greatest efficiencies were found at lower concentrations of Cas9. This technique offers a method to improve the precision of gene editing by CRISPR-Cas9 with minimal modifications to template and proteins. Further experimentation may refine this enhancement into a truly beneficial tool for realizing the potential of CRISPR-Cas9.
Aird, E.J., Lovendahl, K.N., St. Martin, A., Harris, R.S. and Gordon, W.R. (2018) Increasing Cas9-mediated homology-directed repair efficiency through covalent tethering of DNA repair template. Communications Biology 1, 54. DOI: 10.1038/s42003-018-0054-2