Limits...
Targeted and genome-wide sequencing reveal single nucleotide variations impacting specificity of Cas9 in human stem cells.

Yang L, Grishin D, Wang G, Aach J, Zhang CZ, Chari R, Homsy J, Cai X, Zhao Y, Fan JB, Seidman C, Seidman J, Pu W, Church G - Nat Commun (2014)

Bottom Line: However, potential off-target effects of the Cas9 nuclease represent a major safety concern for any therapeutic application.We combine whole-genome sequencing and deep-targeted sequencing to characterise the off-target effects of Cas9 editing.Whole-genome sequencing of Cas9-modified hiPSC clones detects neither gross genomic alterations nor elevated mutation rates.

View Article: PubMed Central - PubMed

Affiliation: 1] Department of Genetics, Harvard Medical School, Boston, Massachusetts 02115, USA [2] Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, Massachusetts 02115, USA.

ABSTRACT
CRISPR/Cas9 has demonstrated a high-efficiency in site-specific gene targeting. However, potential off-target effects of the Cas9 nuclease represent a major safety concern for any therapeutic application. Here, we knock out the Tafazzin gene by CRISPR/Cas9 in human-induced pluripotent stem cells with 54% efficiency. We combine whole-genome sequencing and deep-targeted sequencing to characterise the off-target effects of Cas9 editing. Whole-genome sequencing of Cas9-modified hiPSC clones detects neither gross genomic alterations nor elevated mutation rates. Deep sequencing of in silico predicted off-target sites in a population of Cas9-treated cells further confirms high specificity of Cas9. However, we identify a single high-efficiency off-target site that is generated by a common germline single-nucleotide variant (SNV) in our experiment. Based on in silico analysis, we estimate a likelihood of SNVs creating off-target sites in a human genome to be ~1.5-8.5%, depending on the genome and site-selection method, but also note that mutations might be generated at these sites only at low rates and may not have functional consequences. Our study demonstrates the feasibility of highly specific clonal ex vivo gene editing using CRISPR/Cas9 and highlights the value of whole-genome sequencing before personalised CRISPR design.

Show MeSH

Related in: MedlinePlus

In silico analyses estimating the rates at which genome-variations reduce the specificity of Cas9.(a) Linear relations govern reduction of Cas9 specificity due to genome variations. We analyzed the effect of PGP1 SNVs determined from whole-genome sequencing on all unique Cas9 target sites in the human exome (‘unbiased’), and in all exomic-sites identified by a Cas9 site-selection algorithm (‘CasFinder’). Shown is the fraction of sites for which SNVs converted a three-mismatch off-target to a two-mismatch off-target, where sites are binned by number of three-mismatch off-target sites. Dots represent fractions of sites acquiring two-mismatch off-targets for each bin; lines represent linear regressions in the range of 0–100 off-targets assuming zero y- intercepts. The smaller regression slope for CasFinder (0.08%, R2=0.4795) versus that for the unbiased site selection (0.15%, R2=0.9975) indicates that Cas9 site-selection algorithms can reduce but do not eliminate SNV-generated off-targets. (b) Distribution of three-mismatch off-target counts of the Cas9 targets sites in (a). CasFinder-selected sites show a distribution of three-mismatch off-target counts that is highly concentrated to lower values (16.29±13.84, mean±st.dev) compared with the unbiased site selection (46.83±25.76; P=0, 2-sided Z-test). However, the low CasFinder regression rate in (a) suggests that the reduction of SNV-generated off-target rates by site-selection algorithms does not depend solely on filtering out sites with high numbers of off-targets. (See Supplementary Notes 1 and 2 for details).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4352754&req=5

f3: In silico analyses estimating the rates at which genome-variations reduce the specificity of Cas9.(a) Linear relations govern reduction of Cas9 specificity due to genome variations. We analyzed the effect of PGP1 SNVs determined from whole-genome sequencing on all unique Cas9 target sites in the human exome (‘unbiased’), and in all exomic-sites identified by a Cas9 site-selection algorithm (‘CasFinder’). Shown is the fraction of sites for which SNVs converted a three-mismatch off-target to a two-mismatch off-target, where sites are binned by number of three-mismatch off-target sites. Dots represent fractions of sites acquiring two-mismatch off-targets for each bin; lines represent linear regressions in the range of 0–100 off-targets assuming zero y- intercepts. The smaller regression slope for CasFinder (0.08%, R2=0.4795) versus that for the unbiased site selection (0.15%, R2=0.9975) indicates that Cas9 site-selection algorithms can reduce but do not eliminate SNV-generated off-targets. (b) Distribution of three-mismatch off-target counts of the Cas9 targets sites in (a). CasFinder-selected sites show a distribution of three-mismatch off-target counts that is highly concentrated to lower values (16.29±13.84, mean±st.dev) compared with the unbiased site selection (46.83±25.76; P=0, 2-sided Z-test). However, the low CasFinder regression rate in (a) suggests that the reduction of SNV-generated off-target rates by site-selection algorithms does not depend solely on filtering out sites with high numbers of off-targets. (See Supplementary Notes 1 and 2 for details).

Mentions: To evaluate the scope of impact of SNVs on Cas9 specificity, we conducted an in silico analysis to calculate the likelihood of having high probable off-target sites due to SNVs in a given human genome. We took advantage of the availability of a comprehensive list of genome variations for the PGP1 cell line21, and tailored our analyses to the specificity characteristics of the TAZ-targeting site (Supplementary Note 1). First, the probability of finding a SNV-generated off-target should, in principle, increase linearly with the number of potential off-targets. We verified this by finding all three-mismatch off-targets for all unique Cas9 target sites (1,922,668 sites), computing the fraction that are converted to two-mismatch off-target sites by SNVs in PGP1 genome, and finding a very strict linear relationship (regression slope=0.145, R2=0.9975; Fig. 3a and Supplementary Note 2). However, Cas9 target sites are usually chosen using algorithms that purport to select sites with high specificity, which could potentially bias selection towards targets with less potential for SNP-mediated off-target generation. To gauge the impact of SNVs on targets chosen in this manner, we analyzed a database of 927,104 specificity-checked Cas9 target sites in the human exome that had previously been generated by the CasFinder algorithm22 and identified for each target site all three-mismatch off-target sites genome-wide. For target sites selected this way we found a probability of 1.54% that SNVs in the PGP1 genome lead to conversion of three-mismatch off-target sites to two-mismatch sites. This probability is lower than the 2.34% rate predicted by regression for the unbiased target-site selection (Supplementary Note 1). Indeed, plotting the rates of conversion of CasFinder sites by numbers of three-mismatch off-targets reveals a lower regression rate of 0.08% than the 0.145% of the unbiased sites (Fig. 3a). As expected, the CasFinder-selected sites show a distribution of three-mismatch off-target counts that is concentrated to lower values. (Fig. 3b; 16.17±12.56 versus 46.83±25.76; P=0, two-sided Z-test; Supplementary Note 1). Our results suggest that Cas9 site-selection software can reduce, but not eliminate, the occurrence of potential off-targets generated by SNVs in a genome.


Targeted and genome-wide sequencing reveal single nucleotide variations impacting specificity of Cas9 in human stem cells.

Yang L, Grishin D, Wang G, Aach J, Zhang CZ, Chari R, Homsy J, Cai X, Zhao Y, Fan JB, Seidman C, Seidman J, Pu W, Church G - Nat Commun (2014)

In silico analyses estimating the rates at which genome-variations reduce the specificity of Cas9.(a) Linear relations govern reduction of Cas9 specificity due to genome variations. We analyzed the effect of PGP1 SNVs determined from whole-genome sequencing on all unique Cas9 target sites in the human exome (‘unbiased’), and in all exomic-sites identified by a Cas9 site-selection algorithm (‘CasFinder’). Shown is the fraction of sites for which SNVs converted a three-mismatch off-target to a two-mismatch off-target, where sites are binned by number of three-mismatch off-target sites. Dots represent fractions of sites acquiring two-mismatch off-targets for each bin; lines represent linear regressions in the range of 0–100 off-targets assuming zero y- intercepts. The smaller regression slope for CasFinder (0.08%, R2=0.4795) versus that for the unbiased site selection (0.15%, R2=0.9975) indicates that Cas9 site-selection algorithms can reduce but do not eliminate SNV-generated off-targets. (b) Distribution of three-mismatch off-target counts of the Cas9 targets sites in (a). CasFinder-selected sites show a distribution of three-mismatch off-target counts that is highly concentrated to lower values (16.29±13.84, mean±st.dev) compared with the unbiased site selection (46.83±25.76; P=0, 2-sided Z-test). However, the low CasFinder regression rate in (a) suggests that the reduction of SNV-generated off-target rates by site-selection algorithms does not depend solely on filtering out sites with high numbers of off-targets. (See Supplementary Notes 1 and 2 for details).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4352754&req=5

f3: In silico analyses estimating the rates at which genome-variations reduce the specificity of Cas9.(a) Linear relations govern reduction of Cas9 specificity due to genome variations. We analyzed the effect of PGP1 SNVs determined from whole-genome sequencing on all unique Cas9 target sites in the human exome (‘unbiased’), and in all exomic-sites identified by a Cas9 site-selection algorithm (‘CasFinder’). Shown is the fraction of sites for which SNVs converted a three-mismatch off-target to a two-mismatch off-target, where sites are binned by number of three-mismatch off-target sites. Dots represent fractions of sites acquiring two-mismatch off-targets for each bin; lines represent linear regressions in the range of 0–100 off-targets assuming zero y- intercepts. The smaller regression slope for CasFinder (0.08%, R2=0.4795) versus that for the unbiased site selection (0.15%, R2=0.9975) indicates that Cas9 site-selection algorithms can reduce but do not eliminate SNV-generated off-targets. (b) Distribution of three-mismatch off-target counts of the Cas9 targets sites in (a). CasFinder-selected sites show a distribution of three-mismatch off-target counts that is highly concentrated to lower values (16.29±13.84, mean±st.dev) compared with the unbiased site selection (46.83±25.76; P=0, 2-sided Z-test). However, the low CasFinder regression rate in (a) suggests that the reduction of SNV-generated off-target rates by site-selection algorithms does not depend solely on filtering out sites with high numbers of off-targets. (See Supplementary Notes 1 and 2 for details).
Mentions: To evaluate the scope of impact of SNVs on Cas9 specificity, we conducted an in silico analysis to calculate the likelihood of having high probable off-target sites due to SNVs in a given human genome. We took advantage of the availability of a comprehensive list of genome variations for the PGP1 cell line21, and tailored our analyses to the specificity characteristics of the TAZ-targeting site (Supplementary Note 1). First, the probability of finding a SNV-generated off-target should, in principle, increase linearly with the number of potential off-targets. We verified this by finding all three-mismatch off-targets for all unique Cas9 target sites (1,922,668 sites), computing the fraction that are converted to two-mismatch off-target sites by SNVs in PGP1 genome, and finding a very strict linear relationship (regression slope=0.145, R2=0.9975; Fig. 3a and Supplementary Note 2). However, Cas9 target sites are usually chosen using algorithms that purport to select sites with high specificity, which could potentially bias selection towards targets with less potential for SNP-mediated off-target generation. To gauge the impact of SNVs on targets chosen in this manner, we analyzed a database of 927,104 specificity-checked Cas9 target sites in the human exome that had previously been generated by the CasFinder algorithm22 and identified for each target site all three-mismatch off-target sites genome-wide. For target sites selected this way we found a probability of 1.54% that SNVs in the PGP1 genome lead to conversion of three-mismatch off-target sites to two-mismatch sites. This probability is lower than the 2.34% rate predicted by regression for the unbiased target-site selection (Supplementary Note 1). Indeed, plotting the rates of conversion of CasFinder sites by numbers of three-mismatch off-targets reveals a lower regression rate of 0.08% than the 0.145% of the unbiased sites (Fig. 3a). As expected, the CasFinder-selected sites show a distribution of three-mismatch off-target counts that is concentrated to lower values. (Fig. 3b; 16.17±12.56 versus 46.83±25.76; P=0, two-sided Z-test; Supplementary Note 1). Our results suggest that Cas9 site-selection software can reduce, but not eliminate, the occurrence of potential off-targets generated by SNVs in a genome.

Bottom Line: However, potential off-target effects of the Cas9 nuclease represent a major safety concern for any therapeutic application.We combine whole-genome sequencing and deep-targeted sequencing to characterise the off-target effects of Cas9 editing.Whole-genome sequencing of Cas9-modified hiPSC clones detects neither gross genomic alterations nor elevated mutation rates.

View Article: PubMed Central - PubMed

Affiliation: 1] Department of Genetics, Harvard Medical School, Boston, Massachusetts 02115, USA [2] Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, Massachusetts 02115, USA.

ABSTRACT
CRISPR/Cas9 has demonstrated a high-efficiency in site-specific gene targeting. However, potential off-target effects of the Cas9 nuclease represent a major safety concern for any therapeutic application. Here, we knock out the Tafazzin gene by CRISPR/Cas9 in human-induced pluripotent stem cells with 54% efficiency. We combine whole-genome sequencing and deep-targeted sequencing to characterise the off-target effects of Cas9 editing. Whole-genome sequencing of Cas9-modified hiPSC clones detects neither gross genomic alterations nor elevated mutation rates. Deep sequencing of in silico predicted off-target sites in a population of Cas9-treated cells further confirms high specificity of Cas9. However, we identify a single high-efficiency off-target site that is generated by a common germline single-nucleotide variant (SNV) in our experiment. Based on in silico analysis, we estimate a likelihood of SNVs creating off-target sites in a human genome to be ~1.5-8.5%, depending on the genome and site-selection method, but also note that mutations might be generated at these sites only at low rates and may not have functional consequences. Our study demonstrates the feasibility of highly specific clonal ex vivo gene editing using CRISPR/Cas9 and highlights the value of whole-genome sequencing before personalised CRISPR design.

Show MeSH
Related in: MedlinePlus