CRISPR-Cas9 mediated HDR: Tips for successful experimental design

Discover how type of donor template, template design, and choice of PAM site can affect efficiency of Cas9-mediated homology-directed repair (HDR). Then read the linked application note for detailed, step-wise guidance to maximize HDR rates in your own genome editing experiments.

Nov 10, 2017

CRISPR-Cas9 revolutionizes genomics by enabling efficient site-directed genome editing in a wide variety of biological systems. After double-strand breaks are generated by CRISPR-Cas9, mammalian cells use endogenous cellular machinery to repair the broken sites, exemplified by the canonical non-homologous end joining (NHEJ) pathway and homology-directed repair (HDR) pathway [1]. The error-free HDR faithfully copies genetic information from a related DNA sequence, but this process is characteristically inefficient. As a result, while CRISPR-Cas9 mediated gene knock-out via NHEJ proves to be robust in most organisms, seamless insertion of exogenous genetic material by HDR remains a challenge [2-4].

The Alt-R® CRISPR-Cas9 genome editing system consists of chemically modified, synthetic crRNA and tracrRNA, as well as Cas9 endonuclease derived from S. pyogenes, providing a complete solution for successful gene knockout. In conjunction with a donor DNA template, the system supports rapid genome engineering using HDR. However, given the generally low baseline rate of HDR compared to the error-prone NHEJ pathway, it is critical to optimize HDR experimental conditions. HDR rate can differ greatly between cell types. In agreement with feedback from collaborators, our internal testing has revealed less efficient HDR in transformed cells, compared to immortalized cell lines derived from the same species. In addition, the efficiency of HDR can be affected by a number of experimental factors, including the selected PAM sites, the type of donor DNA, and the concentration of donor DNA present at the time of repair. These factors, and many others, contribute to the high variability of HDR efficiency observed across cell lines [5].

We address these variables here in a brief summary of considerations for successful HDR experimental design. You can also review the more detailed application note, based on our in-house research, that provides stepwise guidance to maximize HDR rates in your own genome editing experiments.

Click here to access the HDR application note.

Design considerations for HDR experiments

Use of single-stranded DNA as donor template

An example workflow for an efficient HDR experiment is shown in Figure 1. HDR exchanges DNA sequence information via sequence homology; i.e., a donor DNA template that contains a desired insert flanked by homology arms complementary to the ends of a planned sequence break is included in the CRISPR-Cas9 HDR experiment for insertion. Historically, double-stranded DNA (dsDNA) was often used, but recent studies have demonstrated the superiority of single-stranded oligo deoxynucleotides (ssODNs) as the donor templates for HDR. ssODNs make it possible to include shorter homology arms, while providing a higher efficiency of insertion than similar templates in dsDNA form [6,7]. As a global leader in providing synthetic DNA, IDT supplies custom synthesized ssDNA as Ultramer® Oligonucleotides and Megamer® Single-Stranded Fragments, both of which provide high-fidelity template sequences that are well suited for HDR [8].


Figure 1. Workflow for co-delivery of Alt-R crRNA, tracrRNA, and Cas9 endonuclease, together with a ssODN donor template for efficient HDR.

crRNA guide selection for HDR optimization

The efficiency of HDR is site-specific, therefore we recommend testing at least 2–4 CRISPR-Cas9 PAM sites close to the desired target region. HDR rates decrease significantly when the template insertion is just 5–10 bases away from the cut site, so it is important to consider the position of each Cas9 cut site relative to where the mutation is to be introduced. As a rule, we recommend selecting the most active PAM site that is as close to the intended insertion site as possible for the best chance of efficient HDR.

Rational design of a ssODN donor template

ssODN template sequences can be designed to bind to either the targeting or non-targeting strands. We recommend designing and testing templates complementary to both strands whenever possible.We have observed robust HDR with homology arms of 30–60 bases, when creating small insertions (e.g., base corrections, tag insertions) using Ultramer Oligonucleotides, for a total maximum donor length of 200 bases. In addition, we find no consistent improvement in HDR efficiency with the use of ssODN templates designed asymmetrically versus symmetrically, regardless of whether the additional length of homology was introduced on the PAM-distal or PAM-proximal side. Taken together, we recommend keeping the desired mutations/insertions approximately centered in the ssODN donor template, flanked by 30–60 nt homology arms on both sides.

Avoid recreating the PAM sequence after the repair takes place

We recommend designing the ssODN template sequence to introduce a silent mutation into the protospacer sequence or even mutating the PAM site itself, when the repaired sequence is recognizable by the crRNA guide. This will prevent the HDR template from reconstituting the PAM site near the insertion and, therefore, avoid unwanted re-cutting at the same locus after the repair takes place.

Concluding remarks

The use of synthetic guide RNAs in the Alt-R genome editing system simplifies the experimental workflow by removing time-consuming cloning steps before HDR. With careful optimization of experimental conditions and inclusion of a deliberately designed ssODN donor template, you can obtain precise incorporation of desired mutations/insertions via the HDR pathway. For further guidance in maximizing HDR with the use of Alt-R CRISPR-Cas9 products, review the complete HDR application note.


  1. Hsu PD, Lander ES, Zhang F. (2014) Development and applications of CRISPR-Cas9 for genome engineering. Cell, 157(6): 1262–1278.
  2. Agudelo D, Duringer A, et al. (2017) Marker-free coselection for CRISPR-driven genome editing in human cells. Nat Methods, 14(6): 615–620.
  3. Lieber MR. (2010) The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu Rev Biochem, 79: 181–211.
  4. Mao Z, Bozzella M, et al. (2008) Comparison of nonhomologous end joining and homologous recombination in human cells. DNA Repair (Amst), 7(10): 1765–1771.
  5. Elliott B, Richardson C, et al. (1998) Gene conversion tracts from double-strand break repair in mammalian cells. Mol Cell Biol, 18(1): 93–101.
  6. Miura H, Gurumurthy CB, et al. (2015) CRISPR/Cas9-based generation of knockdown mice by intronic insertion of artificial microRNA using longer single-stranded DNA. Sci Rep,. 5: 12799.
  7. Yoshimi K, Kunihiro Y, et al. (2016) ssODN-mediated knock-in with CRISPR-Cas for large genomic regions in zygotes. Nat Commun, 7: 10431.
  8. Jacobi AM, Rettig GR, et al. (2017) Simplified CRISPR tools for efficient genome editing and streamlined protocols for their delivery into mammalian cells and mouse zygotes. Methods, 121–122: 16–28.