Target Specificity of Cas9 Nuclease via DNA ... - ACS Publications

Jun 6, 2018 - Department of Chemistry, Seoul National University, Seoul 08826, ... Department of Research and Development, LumiMac, Inc., Seoul 05805,...
0 downloads 0 Views 1MB Size
Communication pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. 2018, 140, 7778−7781

Target Specificity of Cas9 Nuclease via DNA Rearrangement Regulated by the REC2 Domain Keewon Sung,† Jinho Park,† Younggyu Kim,‡ Nam Ki Lee,† and Seong Keun Kim*,† †

Department of Chemistry, Seoul National University, Seoul 08826, Republic of Korea Department of Research and Development, LumiMac, Inc., Seoul 05805, Republic of Korea



Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on July 2, 2018 at 17:21:36 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

molecular mechanism of Cas9-mediated regulation of NTS conformational dynamics and its functional role in the target specificity of CRISPR/Cas9 using single-molecule fluorescence assays. To monitor the intrinsic dynamics of the NTS following the binding of DNA to Cas9:gRNA, we employed single-molecule Förster resonance energy transfer (smFRET) spectroscopy. Once a DNA substrate was immobilized on a glass surface, the FRET donor (Cy3) and acceptor (Alexa 647) labeled respectively at the NTS and the target strand (TS) (Figure 1a and Table S1) reported a single FRET state (E ∼ 0.57) (Figure 1b, upper panel). When we added Cas9:gRNA to the fully complementary “on-target” DNA, we observed two distinct FRET states: a major component, low-FRET state (E ∼ 0.44) and a minor component, high-FRET state (E ∼ 0.74) (Figure 1b, lower panel). Because our dual-labeled DNA substrate is well targeted by Cas9:gRNA (Figure S1), the two FRET states are likely to represent two subconformations of the NTS within the Cas9:gRNA:DNA ternary complex. To investigate the subconformational dynamics in detail, we examined the individual single-molecule time trajectories. Immediately after the injection of Cas9:gRNA, a sudden jump in the FRET efficiency from the DNA-only value to the high-FRET value was observed (Figure 1c). The change in the FRET value was accompanied by an increase in total fluorescence intensity from the donor and acceptor, which is caused by the binding of Cas9:gRNA (Figure S2), i.e., proteininduced fluorescence enhancement (PIFE).20 The FRET efficiency then dropped from the high-FRET to the lowFRET value, and exhibited repetitive transitions between them. When we employed catalytically inactivated “dead” Cas9 (dCas9), we obtained the same FRET states with similar dynamic transitions (Figure S3), which implies that the subconformational dynamics is related to a process leading to the DNA cleavage rather than one following it. We also used DNA constructs with mismatched base-pairs in the PAM-distal end from 2-bp (M19-20) to 4-bp (M17-20). We found that, as we go from M19-20 to M17-20, there occur an abrupt inversion of the relative population between the lowvs high-FRET states (Figure 1d) and more frequent dynamic transitions (Figure S4a). We also found a rapid drop in the cleavage efficiency (Figure S4b) that accompanies the above changes, all of which indicates that the low-FRET state represents a cleavage-competent conformation whereas the

ABSTRACT: Understanding the underlying principles for the target-specific nuclease activity of CRISPR/Cas9 is a prerequisite to minimize its off-target DNA cleavage for genome engineering applications. Here, we show that the noncatalytic REC2 domain of Cas9 nuclease plays a crucial role in off-target discrimination. Using single-molecule fluorescence methods, we investigate conformational dynamics of the non-target strand (NTS) of DNA interacting with Cas9 and find that REC2 regulates the NTS rearrangement for cleavage reaction with the help of positively charged residues on its surface. This mechanistic model for the target specificity of Cas9 provides molecular insights for the rational approach to Cas9 engineering for highly specific genome editing.

T

he CRISPR (clustered regularly interspaced short palindromic repeats)/Cas (CRISPR-associated) system is a vital machinery for adaptive immunity in prokaryotes.1 Among diverse types of the system, type-II CRISPR/Cas utilizes an effector complex that consists of Cas9 endonuclease and two guide RNAs (gRNAs) called CRISPR RNA (crRNA) and trans-activating crRNA (tracrRNA) to degrade invading DNA.2 The Cas9-gRNA effector complex (Cas9:gRNA) identifies its target using base-complementarity between the 20-nucleotide “spacer” region in crRNA and the “protospacer” sequence in the target DNA with additional recognition of a short sequence called protospacer-adjacent motif (PAM), before embarking on double-strand break of DNA.2,3 Due to its RNA-guided target-programmability, CRISPR/ Cas9 has been extensively studied for genome editing, but its application has been hampered by off-target cleavage.4 With a goal to minimize this undesirable side effect, many structural studies have been performed to extract mechanistic insights for the target recognition and cleavage processes.5−9 In particular, single-molecule studies including our previous work10,11 have uncovered conformational dynamics for the target-specific activation of the HNH nuclease domain12,13 as well as the target-specific expansion of the R-loop between crRNA and DNA.10,11,14−16 Meanwhile, structural rearrangement of the non-target strand (NTS), the displaced DNA strand upon the dsDNA−crRNA heteroduplexation, has been newly reported to be strongly coupled to the DNA-cleavage reaction by simulations17,18 and experiments.12,19 However, the regulatory mechanism as well as the structural characteristics of the NTS rearrangement remain poorly understood. Here, we report the © 2018 American Chemical Society

Received: March 20, 2018 Published: June 6, 2018 7778

DOI: 10.1021/jacs.8b03102 J. Am. Chem. Soc. 2018, 140, 7778−7781

Communication

Journal of the American Chemical Society

Figure 1. Subconformational dynamics of the NTS within Cas9:gRNA:DNA. (a) Schematic of dual-labeled DNA for smFRET experiment. (b) FRET histograms for DNA alone (upper) and for postincubation complex between on-target DNA and Cas9:gRNA (lower) with Gaussian fits. (c) Representative time-trajectory recorded after the injection of Cas9:gRNA to on-target DNA. (d) Steady-state FRET histograms for off-target DNAs with PAM-distal mismatches.

Figure 3. Regulation of the NTS dynamics alters Cas9 nuclease specificity. Fraction of the D conformation with WT (blue) and REC2#1 Cas9 (red) for on-target and off-target DNAs with 2-bp mismatch in the PAM-distal region (upper) and the corresponding cleavage activity of Cas9 (lower) (mean ± s.e.m., n = 3); the position of the mismatched base is numbered from the PAM-proximal end (top).

the TS.17,18 However, it is not the HNH movement that induces the NTS rearrangement because they exhibit different behaviors toward Mg2+ ion. The HNH transition is known to occur only in the presence of divalent cations,12 but our FRET data for the NTS rearrangement shows that the relative population of the D vs I conformations is governed by the PAM-distal mismatch even in the absence of Mg2+ ion (Figure S5). We thus turned our attention to looking for the regulator domain other than HNH that mediates the NTS rearrangement. On the basis of the measured FRET efficiencies of the NTS conformations and the available crystal structure of Cas9:gRNA:DNA (Figure S6), we recognized that the REC2 domain located near the PAM-mid region of crRNA-DNA heteroduplex may be a putative mediator for the NTS transition. We first examined whether the NTS would approach REC2, using site-specific dye-labeling on REC2. Just as in our earlier DNA dual-labeling scheme, we used the same donor for the same labeling position of the NTS but a new acceptor (LD650, a Cy5 derivative) to label the D257C residue of REC2 that hardly affects the Cas9 activity (Figure S7a,b). The relative population of the high-FRET state grew dramatically as the number of the mismatched base-pairs increased (Figure S7c), suggesting that the NTS in the I conformation indeed positions at a short distance from REC2. Then, to test the functional relationship between REC2 and the NTS, we introduced REC2-deleted Cas9 mutant (ΔREC2 Cas9) (Figure S8a). Interestingly, ΔREC2 Cas9 showed a higher on-target cleavage activity than the wild-type (WT Cas9) (Figure S8b), which may reflect a structural role of REC2 in sterically blocking the HNH docking.13 The smFRET measurement for the dual-labeled DNA complexed with ΔREC2 Cas9 shows that the deletion of REC2 markedly reduces the relative population of the I conformation particularly for off-targets, shifting the subcon-

Figure 2. REC2 domain regulates the NTS transition with positively charged residues. (a) Three positively charged clusters are exposed on the REC2 surface of the Cas9:gRNA:DNA crystal structure (PDB ID: 4UN3), colored by the polarity of electrostatic potential from red (−5 kcal/mol·e) to blue (+5 kcal/mol·e). (b) Fraction of the NTS in the D conformation with WT (blue) and REC2-#1 Cas9 (red) for on- and off-target DNAs (mean ± s.e.m., n = 3). (c) Time it takes for the I conformation from binding to initial transition toward the D conformation. (d) Dwell time for the I and D conformations during the repetitive transitions that occur after the first displacement. In panels c and d, the error bars represent s.d.

high-FRET state corresponds to a cleavage-incompetent conformation. Considering that the NTS should be displaced from the TS toward the RuvC nuclease domain for cleavage,9,21 we designate the former FRET state in the “displaced” (D) conformation and the latter FRET state in the “intermediate” (I) conformation. MD simulations have predicted that, for the concerted cleavage of the NTS and TS,8 the displacement of the NTS would be highly coupled to the conformational activation of the HNH nuclease domain that is responsible for the cleavage of 7779

DOI: 10.1021/jacs.8b03102 J. Am. Chem. Soc. 2018, 140, 7778−7781

Communication

Journal of the American Chemical Society

Figure 4. Mechanistic model for REC2-directed regulation of the NTS dynamics for DNA-cleavage activity and specificity.

S9a), suggesting that these amino acids may also be involved in the NTS displacement in addition to other processes for DNA cleavage such as the HNH activation.9 Collectively, these results indicate that REC2 drives the NTS into the I conformation electrostatically via two different mechanisms, thereby raising the threshold for the NTS displacement that is thermodynamically favorable even in the presence of a few PAM-distal mismatches without the REC2 functioning. Because the NTS displacement is closely related to the DNAcleavage activation (Figure S4b), we reasoned that the surface charge on REC2 should manage target-specific cleavage by Cas9 nuclease. Correspondingly, for various off-target DNAs constructed by the systematic introduction of mismatched basepairs in the PAM-distal region, WT Cas9 showed a substantial reduction in the off-target cleavage activity compared to the REC2 mutant species in accordance with the decrease in the population of the D conformation (Figures 3 and S12). This result highlights the significance of the regulatory role of the surface charge on REC2 for the off-target rejection beyond the NTS rearrangement. In summary, our study elucidates the regulatory mechanism for the intrinsic conformational dynamics of the NTS within the Cas9 complex. Along the reaction pathway toward the NTS displacement and subsequent cleavage, REC2 stabilizes the initial-I conformation using the positively charged residues on its surface and destabilizes the displaced NTS in the D conformation allosterically, thereby inhibiting DNA cleavage particularly in the presence of PAM-distal mismatches (Figure 4). This model is consistent with a recent cryo-EM study for the type-I CRISPR/Cascade, which found that a captured intermediate structure during the R-loop expansion supports the interaction between the NTS and lysine residues right next to crRNA-DNA heteroduplex.22 This correspondence may suggest the electrostatic regulation of the NTS as a conserved mechanism among diverse CRISPR systems. Furthermore, the inhibitory role of REC2 in the NTS displacement is reminiscent of the recently proposed regulation model for the HNHnuclease activation, in which REC2 sterically blocks the HNH movement toward the DNA-cleavage position.13 Because this model involves another domain called REC3 that is responsible for sensing the PAM-distal base-pairing that also governs the NTS dynamics (Figure 1d), it is likely that REC2 would regulate both HNH and the NTS in response to the PAM-distal sensing of REC3. Although further studies are required to correlate the two REC2-directed regulation processes in molecular detail, our result provides strong evidence that REC2 plays a critical role in the off-target discrimination. We project that the newly identified key residues on REC2 as well as the concept of the electrostatic NTS-regulation from our

formational equilibrium toward the D conformation compared to the case of WT Cas9 (Figure S8c). To further investigate the underlying molecular mechanism of the REC2-associated NTS rearrangement, we again referred to the crystal structure of Cas9:gRNA:DNA, whose electrostatic charge distribution on REC2 suggests three distinct clusters of positively charged amino acids (Figure 2a). By noting that the electrostatic interaction may mediate the REC2NTS coupling, we introduced three Cas9 mutants (termed REC2-#1, -#2, and -#3) whose positively charged residues were point-mutated to alanines. Among these mutants, only REC2#1 retained the full on-target activity (Figure S9). For this mutant, the conformational distributions of the NTS also shift toward the D conformation as for ΔREC2 Cas9 (Figure S10a and Figure 2b). For detailed kinetics, we separated the time required for initial NTS displacement following DNA binding and the dwell times during the subsequent transitions (Figure S10b). Given the two different inactive states of HNH that form right after the binding vs generally during repetitive transitions with a catalytically active state,12 the separate analysis allows characterization of different kinetic behaviors due to the HNHconformational discrepancy. Our dwell time data shows that WT Cas9 has a longer dwell time between the binding and initial displacement than REC2-#1 Cas9, and the difference becomes larger as the number of PAM-distal mismatches increases (Figures 2c and S10c), supporting that the positive charge on the REC2 surface raises the kinetic barrier for the initial NTS displacement by trapping the NTS in the “initial-I” conformation. On the other hand, during the repetitive transitions after the initial displacement, dwell time distributions of both I and D conformations were well described by a biexponential fit, indicating HNH-conformational heterogeneity among the active and inactive states in each NTS conformation (Figure S10d,e). The amplitude-weighted average dwell time in this “repetitive-I” conformation shows marginal differences between the two Cas9 species, whereas the average dwell time in the D conformation becomes shorter for WT Cas9 than for REC2-#1 Cas9 (Figure 2d). This suggests that the charged residues allosterically extract the NTS from the D conformation once the NTS is displaced. Because crystallographic and cryo-EM structures of the D conformation indicate that HNH is placed between REC2 and the displaced NTS,9,21 the allosteric regulation of the NTS in the D conformation by REC2 could be mediated by the HNH movement. Unexpectedly, our smFRET data for the other mutants (REC2-#2 and -#3) shows that they too share a similar behavior with REC2-#1 Cas9 (Figure S11) despite their impaired cleavage activities (Figure 7780

DOI: 10.1021/jacs.8b03102 J. Am. Chem. Soc. 2018, 140, 7778−7781

Communication

Journal of the American Chemical Society

(20) Hwang, H.; Myong, S. Chem. Soc. Rev. 2014, 43, 1221−1229. (21) Huai, C.; Li, G.; Yao, R.; Zhang, Y.; Cao, M.; Kong, L.; Jia, C.; Yuan, H.; Chen, H.; Lu, D.; Huang, Q. Nat. Commun. 2017, 8, 1375. (22) Xiao, Y.; Luo, M.; Hayes, R. P.; Kim, J.; Ng, S.; Ding, F.; Liao, M.; Ke, A. Cell 2017, 170, 48−60.

study may shed light on rational design of highly specific Cas9 variants for genome editing.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b03102. Experimental methods, supplementary data, and sequence list for DNA and RNA constructs (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Seong Keun Kim: 0000-0001-9891-0931 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Research Foundation of Korea (NRF) grants to N.K.L. (NRF-2017R1A2B3010309) and to S.K.K. (2018R1A2B2001422).



REFERENCES

(1) Marraffini, L. A.; Sontheimer, E. J. Nat. Rev. Genet. 2010, 11, 181−190. (2) Jinek, M.; Chylinski, K.; Fonfara, I.; Hauer, M.; Doudna, J. A.; Charpentier, E. Science 2012, 337, 816−821. (3) Sternberg, S. H.; Redding, S.; Jinek, M.; Greene, E. C.; Doudna, J. A. Nature 2014, 507, 62−67. (4) Hsu, P. D.; Lander, E. S.; Zhang, F. Cell 2014, 157, 1262−1278. (5) Nishimasu, H.; Ran, F. A.; Hsu, P. D.; Konermann, S.; Shehata, S. I.; Dohmae, N.; Ishitani, R.; Zhang, F.; Nureki, O. Cell 2014, 156, 935−949. (6) Anders, C.; Niewoehner, O.; Duerst, A.; Jinek, M. Nature 2014, 513, 569−573. (7) Jiang, F.; Zhou, K.; Ma, L.; Gressel, S.; Doudna, J. A. Science 2015, 348, 1477−1481. (8) Sternberg, S. H.; LaFrance, B.; Kaplan, M.; Doudna, J. A. Nature 2015, 527, 110−113. (9) Jiang, F.; Taylor, D. W.; Chen, J. S.; Kornfeld, J. E.; Zhou, K.; Thompson, A. J.; Nogales, E.; Doudna, J. A. Science 2016, 351, 867− 871. (10) Lim, Y.; Bak, S. Y.; Sung, K.; Jeong, E.; Lee, S. H.; Kim, J.-S.; Bae, S.; Kim, S. K. Nat. Commun. 2016, 7, 13350. (11) Cromwell, C. R.; Sung, K.; Park, J.; Krysler, A. R.; Jovel, J.; Kim, S. K.; Hubbard, B. P. Nat. Commun. 2018, 9, 1448. (12) Dagdas, Y. S.; Chen, J. S.; Sternberg, S. H.; Doudna, J. A.; Yildiz, A. Sci. Adv. 2017, 3, eaao0027. (13) Chen, J. S.; Dagdas, Y. S.; Kleinstiver, B. P.; Welch, M. M.; Sousa, A. A.; Harrington, L. B.; Sternberg, S. H.; Joung, J. K.; Yildiz, A.; Doudna, J. A. Nature 2017, 550, 407−410. (14) Szczelkun, M. D.; Tikhomirova, M. S.; Sinkunas, T.; Gasiunas, G.; Karvelis, T.; Pschera, P.; Siksnys, V.; Seidel, R. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 9798−9803. (15) Singh, D.; Sternberg, S. H.; Fei, J.; Doudna, J. A.; Ha, T. Nat. Commun. 2016, 7, 12778. (16) Singh, D.; Wang, Y.; Mallon, J.; Yang, O.; Fei, J.; Poddar, A.; Ceylan, D.; Bailey, S.; Ha, T. Nat. Struct. Mol. Biol. 2018, 25, 347−354. (17) Palermo, G.; Miao, Y.; Walker, R. C.; Jinek, M.; McCammon, J. A. ACS Cent. Sci. 2016, 2, 756−763. (18) Palermo, G.; Miao, Y.; Walker, R. C.; Jinek, M.; McCammon, J. A. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 7260−7265. (19) Räz, M. H.; Hidaka, K.; Sturla, S. J.; Sugiyama, H.; Endo, M. J. Am. Chem. Soc. 2016, 138, 13842−13845. 7781

DOI: 10.1021/jacs.8b03102 J. Am. Chem. Soc. 2018, 140, 7778−7781