brief communication published online: 6 April 2015 | doi: 10.1038/nchembio.1793
Small molecule–triggered Cas9 protein with improved genome-editing specificity Directly modulating the activity of genome-editing proteins has available, cell-permeable small molecule. We previously evolved the potential to increase their specificity by reducing activity inteins that undergo protein splicing only in the presence of following target locus modification. We developed Cas9 4-hydroxytamoxifen (4-HT)21. These inteins were developed by nucleases that are activated by the presence of a cell-permeable inserting the human estrogen receptor ligand-binding domain into small molecule by inserting an evolved 4-hydroxytamoxifen– the M. tuberculosis RecA intein and evolving the resulting inacresponsive intein at specific positions in Cas9. In human cells, tive fusion protein into a conditionally active intein that requires conditionally active Cas9s modify target genomic sites with up the presence of 4-HT21–23. Subsequent evolution at 37 °C yielded a to 25-fold higher specificity than wild-type Cas9. second-generation intein, 37R3-2, with improved splicing properThe RNA-guided endonuclease Cas9 from the type II CRISPR- ties in mammalian cells23. We envisioned that inserting the 37R3-2 Cas system enables simple and efficient genome editing in a wide intein into Cas9 at a location that disrupts Cas9 activity until protein variety of organisms. Virtually any target DNA locus can be cleaved splicing has taken place could result in conditionally active Cas9 by programming Cas9 with a single guide RNA (sgRNA) that contains nucleases that are active only in the presence of 4-HT (Fig. 1a). a stretch of ~20 nucleotides complementary to the target sequence1–4. We genetically inserted the 4-HT–dependent intein at each of 15 Owing to its simplicity and robustness, the Cas9 system has been positions in Cas9 (Cys80, Ala127, Thr146, Ser219, Thr333, Thr519, widely adopted for biological research and therapeutic develop- Cys574, Thr622, Ser701, Ala728, Thr995, Ser1006, Ser1154, Ser1159 ment. The DNA cleavage specificity of Cas9 is imperfect5–9, however, and Ser1274), chosen to distribute the location of the intein across the raising concerns over off-target genome modification that may limit its usefulness in theraa Inactive Cas9 Inactive Cas9 Active Cas9–sgRNA complex peutic or research applications. Researchers 4-HT have reduced Cas9 off-target activity through Protein protein10–13 and sgRNA14 engineering, and by 3′ 5′ splicing 5′ 3′ Cas9 direct delivery of Cas9–sgRNA protein–RNA Cas9 5′ 15–17 complexes into cells . A complementary, underexplored strategy to improve Cas9 specificity is to reduce its activity once it has had sufficient opportuEvolved ligand4-HT binding dependent intein activates intein nity to modify the target DNA locus. Indeed, S219 higher concentrations of Cas9 in cells have b c 45 – 4-HT been observed to degrade specificity 5–7 + 4-HT (1 µM) C574 40 (defined as the ratio of on-target to off-target 35 DNA cleavage activity), presumably because 30 any Cas9 protein present after the target locus 25 has been modified can only process off-target 20 substrates. Unfortunately, wild-type Cas9 15 nucleases are not known to be regulated by 10 other molecules and therefore are used in their DNA 5 sgRNA constitutively active form. Although Cas9 0 Cas9 can be regulated at the transcriptional level Intein insertion sites Ligand-responsive intein insertion sites through the use of inducible promoters18,19, used in subsequent experiments transcriptional control cannot limit activity Figure 1 | Insertion of an evolved ligand-dependent intein enables small-molecule control to the short temporal windows that may of Cas9. (a) Intein insertion renders Cas9 inactive. Upon 4-HT binding, the intein undergoes be necessary to maximize genome-editing conformational changes that trigger protein splicing and restore Cas9 activity. (b) The evolved specificity17,20, in contrast with the high temintein was inserted to replace each of the colored residues. Intein-inserted Cas9 variants at poral resolution of post-translational strategies Ser219 and Cys574 (green) were used in subsequent experiments. (c) Genomic EGFP disruption that directly control protein activity. activity of wild-type Cas9 and intein-Cas9 variants in the absence or presence of 4-HT. To enable post-translational control over Intein-Cas9 variants are identified by the residue replaced by the intein. Error bars reflect Cas9 in cells, we sought to engineer variants the s.d. of three biological replicates. of Cas9 that can be controlled with a readily OH
N
GFP-negative cells (%)
O
– W Ca T s9 Ca Cy s9 s Al 80 a Th 127 r1 Se 46 r Th 219 r3 Th 33 Cy r519 s Th 574 r6 Se 22 r Al 701 a Th 728 Se r99 r1 5 Se 006 r1 Se 154 r Se 1159 r12 74
© 2015 Nature America, Inc. All rights reserved.
Kevin M Davis1,2, Vikram Pattanayak3, David B Thompson1,2, John A Zuris1,2 & David R Liu1,2*
Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts, USA. 2Howard Hughes Medical Institute, Harvard University, Cambridge, Massachusetts, USA. 3Department of Pathology, Massachusetts General Hospital, Boston, Massachusetts, USA. *e-mail: [email protected] 1
nature CHEMICAL BIOLOGY | Advance online publication | www.nature.com/naturechemicalbiology
1
brief communication
2
On-target/off-target indel frequency ratio On-target/off-target indel frequency ratio
Indel modification frequency (%)
structural domains of Cas9 (ref. 24) (Fig. 1b). a b – 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) Because intein splicing leaves behind a single + 4-HT intein-Cas9(C574) 30 20 – 4-HT intein-Cas9(C574) cysteine residue, the intein was inserted in + 4-HT WT Cas9 + 4-HT intein-Cas9(C574) 25 place of one Cas9 amino acid in each of the – 4-HT WT Cas9 15 + 4-HT WT Cas9 15 candidate constructs. In addition to replacing 20 natural cysteine amino acids, we also favored 15 10 replacing alanine, serine or threonine residues 10 to minimize the likelihood that the resulting 5 cysteine point mutation resulting from pro5 tein splicing would disrupt Cas9 activity. 0 0 On On On On/Off 1 On/Off 2 On/Off 3 On/Off 4 The 15 intein-Cas9 candidates were expressed EMX VEGF CLTA EMX in HEK293-GFP cells together with a sgRNA that targets the genomic EGFP locus in these c d 40 + 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) 10 cells. Twelve hours after transfection, cells + 4-HT intein-Cas9(C574) + 4-HT intein-Cas9(C574) were treated with or without 1 μM 4-HT. Five + 4-HT WT Cas9 + 4-HT WT Cas9 8 30 days after transfection, cells were analyzed 6 on a flow cytometer for loss of GFP expres20 sion from Cas9-mediated EGFP cleavage and 4 subsequent nonhomologous end joining. 10 2 Eight of the candidates, corresponding to intein insertion at Ala127, Thr146, Ser219, 0 0 On/Off 1 On/Off 2 On/Off 3 On/Off 4 On/Off 1 On/Off 2 On/Off 3 Thr333, Thr519, Cys574, Ser1006 and Ser1159, VEGF CLTA demonstrated 4-HT–dependent loss of GFP expression, consistent with 4-HT–triggered Figure 2 | Genomic DNA modification by intein-Cas9(S219), intein-Cas9(C574) and Cas9 activity (Fig. 1c). Notably, three inteinwild-type Cas9. (a) Indel frequency from high-throughput DNA sequencing of amplified Cas9 variants (insertion at Ala728, Thr995 genomic on-target sites in the absence or presence of 4-HT. Note that a substantial number and Ser1154) showed high DNA modification of indels were observed at the CLTA on-target site even in the absence of a targeting sgRNA rates both in the presence and absence of 4-HT, (Supplementary Table 7). (b–d) DNA modification specificity, defined as on-target/off-target suggesting that large protein insertions at these indel frequency ratio5–7, normalized to wild-type Cas9. Cells were transfected with 500 ng of the positions do not substantially inhibit nuclease Cas9 expression plasmid and incubated at transfection with medium containing 1 M 4-HT. activity or that the intein lost its 4-HT depenAfter 12 h, the medium was replaced with medium lacking 4-HT. P values are
Small molecule–triggered Cas9 protein with improved genome-editing specificity Directly modulating the activity of genome-editing proteins has available, cell-permeable small molecule. We previously evolved the potential to increase their specificity by reducing activity inteins that undergo protein splicing only in the presence of following target locus modification. We developed Cas9 4-hydroxytamoxifen (4-HT)21. These inteins were developed by nucleases that are activated by the presence of a cell-permeable inserting the human estrogen receptor ligand-binding domain into small molecule by inserting an evolved 4-hydroxytamoxifen– the M. tuberculosis RecA intein and evolving the resulting inacresponsive intein at specific positions in Cas9. In human cells, tive fusion protein into a conditionally active intein that requires conditionally active Cas9s modify target genomic sites with up the presence of 4-HT21–23. Subsequent evolution at 37 °C yielded a to 25-fold higher specificity than wild-type Cas9. second-generation intein, 37R3-2, with improved splicing properThe RNA-guided endonuclease Cas9 from the type II CRISPR- ties in mammalian cells23. We envisioned that inserting the 37R3-2 Cas system enables simple and efficient genome editing in a wide intein into Cas9 at a location that disrupts Cas9 activity until protein variety of organisms. Virtually any target DNA locus can be cleaved splicing has taken place could result in conditionally active Cas9 by programming Cas9 with a single guide RNA (sgRNA) that contains nucleases that are active only in the presence of 4-HT (Fig. 1a). a stretch of ~20 nucleotides complementary to the target sequence1–4. We genetically inserted the 4-HT–dependent intein at each of 15 Owing to its simplicity and robustness, the Cas9 system has been positions in Cas9 (Cys80, Ala127, Thr146, Ser219, Thr333, Thr519, widely adopted for biological research and therapeutic develop- Cys574, Thr622, Ser701, Ala728, Thr995, Ser1006, Ser1154, Ser1159 ment. The DNA cleavage specificity of Cas9 is imperfect5–9, however, and Ser1274), chosen to distribute the location of the intein across the raising concerns over off-target genome modification that may limit its usefulness in theraa Inactive Cas9 Inactive Cas9 Active Cas9–sgRNA complex peutic or research applications. Researchers 4-HT have reduced Cas9 off-target activity through Protein protein10–13 and sgRNA14 engineering, and by 3′ 5′ splicing 5′ 3′ Cas9 direct delivery of Cas9–sgRNA protein–RNA Cas9 5′ 15–17 complexes into cells . A complementary, underexplored strategy to improve Cas9 specificity is to reduce its activity once it has had sufficient opportuEvolved ligand4-HT binding dependent intein activates intein nity to modify the target DNA locus. Indeed, S219 higher concentrations of Cas9 in cells have b c 45 – 4-HT been observed to degrade specificity 5–7 + 4-HT (1 µM) C574 40 (defined as the ratio of on-target to off-target 35 DNA cleavage activity), presumably because 30 any Cas9 protein present after the target locus 25 has been modified can only process off-target 20 substrates. Unfortunately, wild-type Cas9 15 nucleases are not known to be regulated by 10 other molecules and therefore are used in their DNA 5 sgRNA constitutively active form. Although Cas9 0 Cas9 can be regulated at the transcriptional level Intein insertion sites Ligand-responsive intein insertion sites through the use of inducible promoters18,19, used in subsequent experiments transcriptional control cannot limit activity Figure 1 | Insertion of an evolved ligand-dependent intein enables small-molecule control to the short temporal windows that may of Cas9. (a) Intein insertion renders Cas9 inactive. Upon 4-HT binding, the intein undergoes be necessary to maximize genome-editing conformational changes that trigger protein splicing and restore Cas9 activity. (b) The evolved specificity17,20, in contrast with the high temintein was inserted to replace each of the colored residues. Intein-inserted Cas9 variants at poral resolution of post-translational strategies Ser219 and Cys574 (green) were used in subsequent experiments. (c) Genomic EGFP disruption that directly control protein activity. activity of wild-type Cas9 and intein-Cas9 variants in the absence or presence of 4-HT. To enable post-translational control over Intein-Cas9 variants are identified by the residue replaced by the intein. Error bars reflect Cas9 in cells, we sought to engineer variants the s.d. of three biological replicates. of Cas9 that can be controlled with a readily OH
N
GFP-negative cells (%)
O
– W Ca T s9 Ca Cy s9 s Al 80 a Th 127 r1 Se 46 r Th 219 r3 Th 33 Cy r519 s Th 574 r6 Se 22 r Al 701 a Th 728 Se r99 r1 5 Se 006 r1 Se 154 r Se 1159 r12 74
© 2015 Nature America, Inc. All rights reserved.
Kevin M Davis1,2, Vikram Pattanayak3, David B Thompson1,2, John A Zuris1,2 & David R Liu1,2*
Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts, USA. 2Howard Hughes Medical Institute, Harvard University, Cambridge, Massachusetts, USA. 3Department of Pathology, Massachusetts General Hospital, Boston, Massachusetts, USA. *e-mail: [email protected] 1
nature CHEMICAL BIOLOGY | Advance online publication | www.nature.com/naturechemicalbiology
1
brief communication
2
On-target/off-target indel frequency ratio On-target/off-target indel frequency ratio
Indel modification frequency (%)
structural domains of Cas9 (ref. 24) (Fig. 1b). a b – 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) Because intein splicing leaves behind a single + 4-HT intein-Cas9(C574) 30 20 – 4-HT intein-Cas9(C574) cysteine residue, the intein was inserted in + 4-HT WT Cas9 + 4-HT intein-Cas9(C574) 25 place of one Cas9 amino acid in each of the – 4-HT WT Cas9 15 + 4-HT WT Cas9 15 candidate constructs. In addition to replacing 20 natural cysteine amino acids, we also favored 15 10 replacing alanine, serine or threonine residues 10 to minimize the likelihood that the resulting 5 cysteine point mutation resulting from pro5 tein splicing would disrupt Cas9 activity. 0 0 On On On On/Off 1 On/Off 2 On/Off 3 On/Off 4 The 15 intein-Cas9 candidates were expressed EMX VEGF CLTA EMX in HEK293-GFP cells together with a sgRNA that targets the genomic EGFP locus in these c d 40 + 4-HT intein-Cas9(S219) + 4-HT intein-Cas9(S219) 10 cells. Twelve hours after transfection, cells + 4-HT intein-Cas9(C574) + 4-HT intein-Cas9(C574) were treated with or without 1 μM 4-HT. Five + 4-HT WT Cas9 + 4-HT WT Cas9 8 30 days after transfection, cells were analyzed 6 on a flow cytometer for loss of GFP expres20 sion from Cas9-mediated EGFP cleavage and 4 subsequent nonhomologous end joining. 10 2 Eight of the candidates, corresponding to intein insertion at Ala127, Thr146, Ser219, 0 0 On/Off 1 On/Off 2 On/Off 3 On/Off 4 On/Off 1 On/Off 2 On/Off 3 Thr333, Thr519, Cys574, Ser1006 and Ser1159, VEGF CLTA demonstrated 4-HT–dependent loss of GFP expression, consistent with 4-HT–triggered Figure 2 | Genomic DNA modification by intein-Cas9(S219), intein-Cas9(C574) and Cas9 activity (Fig. 1c). Notably, three inteinwild-type Cas9. (a) Indel frequency from high-throughput DNA sequencing of amplified Cas9 variants (insertion at Ala728, Thr995 genomic on-target sites in the absence or presence of 4-HT. Note that a substantial number and Ser1154) showed high DNA modification of indels were observed at the CLTA on-target site even in the absence of a targeting sgRNA rates both in the presence and absence of 4-HT, (Supplementary Table 7). (b–d) DNA modification specificity, defined as on-target/off-target suggesting that large protein insertions at these indel frequency ratio5–7, normalized to wild-type Cas9. Cells were transfected with 500 ng of the positions do not substantially inhibit nuclease Cas9 expression plasmid and incubated at transfection with medium containing 1 M 4-HT. activity or that the intein lost its 4-HT depenAfter 12 h, the medium was replaced with medium lacking 4-HT. P values are
