Compositions and methods of improving specificity in genomic engineering using RNA-guided endonucleases

Abstract

To provide methods of generating optimized guide RNAs (gRNAs).SOLUTION: Disclosed herein are optimized guide RNAs (gRNAs) and methods of designing and using the optimized gRNAs that have increased target binding specificity and reduced off-target binding.SELECTED DRAWING: None

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 62 / 209,466, filed Aug. 25, 2015, the entire contents of which are incorporated herein by reference.

[0002] STATEMENT OF GOVERNMENT RIGHTS This invention was made with Government support under Federal Grants MCB1244297 and CBET1151035 awarded by the National Science Foundation, and Federal Grants F32GM11250201, R01DA036865, and DP2OD008586 awarded by the National Institutes of Health. The Government has certain rights in this invention.

[0003] The present disclosure is directed to optimized guide RNAs (gRNAs) and methods for designing and using said gRNAs with increased target-binding specificity and reduced off-target binding. [Background technology]

[0004] RNA-guided endonucleases, particularly the protein Cas9, have been hailed as a potential "perfect genome engineering tool" because it can be guided by a single-stranded "guide RNA" molecule to cleave DNA with almost any sequence. This ability has been exploited in recent years for several emerging biological and medical applications, generating great excitement and promise for its future use. However, real genome engineering requires extremely precise control of the ability to selectively target and cleave precise DNA sequences, so that off-target DNA does not become unintentionally damaged and mutated.

[0005] Cas9 is the endonuclease of prokaryotic type II CRISPR (clustered regularly interspaced short palindromic repeats)-CRISPR-associated (Cas) response to invading foreign DNA. During this response, Cas9 is first bound by the CRISPR RNA (crRNA):transactivating crRNA (tracrRNA) duplex and then guided to cleave DNA containing a 20 base pair (bp) "protospacer" site complementary to a variable 20 bp segment of the crRNA (Figure 1A). In combination with a single-stranded guide RNA (sgRNA), the Cas9-sgRNA complex binds to a 20 bp "protospacer" sequence in the targeted DNA, provided that the protospacer is immediately followed by a protospacer adjacent motif (PAM, herein "TGG"). After binding, the Cas9 endonuclease creates a double-stranded break (triangle) within the protospacer. Essentially, the only constraint on the sequences that Cas9 can target is that the protospacer site in the foreign DNA molecule must be immediately followed by a short protospacer adjacent motif (PAM), such as "NGG" in the case of S. pyogenes Cas9. Analysis of crystallographic and biochemical experiments suggests that specificity in protospacer binding and cleavage is conferred first through recognition of the PAM site by the Cas9 protein itself, followed by strand invasion by the bound RNA complex and direct Watson-Crick base pairing with the protospacer (Figure 1A).

[0006] The ability of Cas9 to be modularly "programmed" by single-stranded RNA hairpins to target almost any DNA site has generated great excitement in recent years after the CRISPR-Cas9 system has been repurposed for several heterologous biotechnology applications. In particular, a single-stranded guide RNA (sgRNA) hairpin has been designed that combines the essential components of the crRNA:tracrRNA duplex into a single functional molecule. With this sgRNA, Cas9 can be introduced into various organisms to generate targeted double-strand breaks in vivo for significantly easier genome engineering. Nuclease-null Cas9 (D10A / H840A, also known as "dCas9") and chimeric dCas9 derivatives have also been used to alter gene expression through targeted binding in vivo at or near promoter sites and to introduce targeted epigenetic modifications.

[0007] In practice, off-target binding and cleavage by Cas9 is problematic as it may adversely affect its potential use. Considerable efforts have been made to improve the specificity of Cas9 / dCas9 activity. First, the broadest efforts are largely achieved by judicious selection of target sequences with no other similar sequences in the genome, but recent investigations have found that these methods are performed with poor ability to predict off-target cleavage. In addition, efforts have been made to directly engineer the protein itself through the introduction of point mutations that have been found to modulate or increase specificity at the PAM or protospacer binding. Cas9 derivatives that do not perform double-stranded DNA cleavage but only nick a single strand of DNA are also used in pairs ("paired nickases") under the assumption that the chance of off-target nicking at multiple sites close enough to each other to produce double-stranded breaks would be extremely rare. Finally, some work has been done to produce guide RNA variants themselves, aiming to achieve greater specificity. Previous efforts in which 5'-extensions were added to guide RNAs to complement additional nucleotides after the protospacer did not show increased Cas9 cleavage specificity in vivo. Instead, they were digested in living cells back to nearly their standard length (Figure 1A). For applications in genome engineering, and especially for therapeutic applications, extreme specificity in gene targeting is required so that off-target DNA is not damaged and erroneous mutations are not introduced. However, in practice, there have been several reports of off-target binding and cleavage by Cas9 that may adversely affect its potential use. Summary of the Invention [Problem to be solved by the invention]

[0008] There remains a need to reduce off-target binding and increase nuclease specificity using the CRISPR / Cas9 system. [Means for solving the problem]

[0009] The present invention is directed to a method of producing an optimized guide RNA (gRNA), comprising the steps of: a) identifying a target region of interest, the target region of interest comprising a protospacer sequence; b) determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; and d) producing a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of the protospacer-targeting sequence or segment, the RNA segment being at the 5' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprising a linker between the 5' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker being a polynucleotide having a length of N nucleotides. a nucleotide sequence, wherein the first gRNA is capable of invading the protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading the off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; e) calculating or computationally simulating an estimate of the invasion kinetics and the lifetime that the first gRNA remains invaded within the protospacer and off-target site duplex, where the invasion kinetics is estimated nucleotide by nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site with the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site;g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with the second gRNA; h) identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and i) testing the optimized gRNA in vivo to determine binding specificity.

[0010] The present invention is directed to a method of producing an optimized guide RNA (gRNA), comprising the steps of: a) identifying a target region of interest, the target region of interest comprising a protospacer sequence; b) determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; and d) producing a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of the protospacer-targeting sequence or segment, the RNA segment being at the 3' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprising a linker between the 3' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker being a polynucleotide having a length of N nucleotides. a nucleotide sequence, wherein the first gRNA is capable of invading the protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading the off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; e) calculating or computationally simulating an estimate of the invasion kinetics and the lifetime that the first gRNA remains invaded within the protospacer and off-target site duplex, where the invasion kinetics is estimated nucleotide by nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site with the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site;g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with the second gRNA; h) identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and i) testing the optimized gRNA in vivo to determine binding specificity.

[0011] The present invention is directed to optimized gRNAs produced by the methods described above.

[0012] The present invention is directed to isolated polynucleotides encoding the optimized gRNAs described above.

[0013] The present invention is directed to a vector comprising the isolated polynucleotide described above.

[0014] The present invention is directed to a cell comprising the above-described isolated polynucleotide or the above-described vector.

[0015] The present invention is directed to a kit comprising an isolated polynucleotide as described above, a vector as described above, or a cell as described above.

[0016] The present invention is directed to a method of epigenome editing in a target cell or subject, comprising contacting a cell or subject with an effective amount of the optimized gRNA molecule described above and a fusion protein, the fusion protein comprising a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcription activation activity, transcription repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0017] The present invention is directed to a method for site-specific DNA cleavage in a target cell or subject, comprising contacting a cell or subject with an effective amount of the optimized gRNA molecule described above and a fusion protein or Cas9 protein, the fusion protein comprising a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcription activation activity, transcription repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0018] The present invention is directed to a method of genome editing in a cell, comprising administering to the cell an effective amount of the optimized gRNA molecule described above and a fusion protein, the fusion protein comprising a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcriptional terminator activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0019] The present invention is directed to a method of regulating gene expression in a cell, comprising contacting the cell with an effective amount of the optimized gRNA described above and a fusion protein, the fusion protein comprising a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcriptional terminator activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity. [Brief description of the drawings]

[0020] [Figure 1A-1D]Atomic force microscopy (AFM) images of dCas9-sgRNA bound at a protospacer sequence within a single streptavidin-labeled DNA molecule derived from the human AAVS1 locus (Figure 1B). The percentage of bound DNA occupied by Cas9 / dCas9-sgRNA along AAVS1-derived (Figure 1C) or genetically engineered DNA substrates (Figure 1D) designed with a series of fully and partially complementary protospacer sequences are shown. The vertical axis represents the (23 bp) segment where each key feature is located on each substrate. [Figure 2A-2B] Modulation of binding affinity and specificity by guide RNA variants is shown. Figure 2A shows a schematic of dCas9 bound to a single-stranded guide RNA (tru-gRNA, purple) that is truncated by two nucleotides from its 5' end. Figure 2B shows a schematic proposed mechanism of dCas9 bound to a single-stranded guide RNA (hp-gRNA, blue) that forms a hairpin with a 5'-end extension and a PAM-distal binding segment of its targeting region. [Fig. 2C-2D] Figure 2C shows the single-site binding affinity (KA) for dCas9 with tru-gRNA (purple, n=257) along the genetically engineered DNA substrate (see Figure 1D). The dashed line shows the single-site affinity of dCas9-sgRNA for comparison. Figure 2D shows the single-site binding affinity (KA) for dCas9 with guide RNAs that have a 5'-hairpin that overlaps with nucleotides complementary to the last 6 (hp6-gRNA, blue) or 10 (hp10-gRNA, green) PAM-distal nucleotides of the protospacer. [Figure 3A-3B]We show that Cas9 undergoes progressive conformational transitions as it binds to sites matching the protospacer sequence. Figure 3A shows the percentage of bound DNA occupied by Cas9 / dCas9 along the DNA substrate, with colors representing populations of Cas9 / dCas9 clustered according to their structure (by mean squared difference after alignment (see text)). Different features on the DNA used for site-specific analysis of Cas9 / Cas9 structural properties were labeled as non-specific sequences (α; "20MM"), sites containing 10 PAM distal mismatches in the protospacer (β, "10MM"), sites containing 5 PAM distal mismatches in the protospacer (γ, "5MM"), or complete protospacer sites (δ or ε; "0MM" for dCas9 or Cas9, respectively). Figure 3B shows the volume versus height of observed Cas9 / dCas9, color-coded by the cluster to which each protein is assigned. Dashed lines delineate regions likely consisting of aggregates (top right) or nearby DNA with adsorbed streptavidin labels (bottom left). For comparison - average height of streptavidin end labels: 0.92 nm ± 0.006 nm (SEM); average volume of streptavidin end labels: 0.110χ104 nm3 ± 0.002χ104 nm3 (SEM); n = 1941. [Figure 3C-3D]The ensemble averages of the major clusters are shown in Figure 3C and color-coded according to the clustered structures they represent. Figure 3D shows the average volumes and heights of Cas9 / dCas9 with sgRNA (red circles, red label for Cas9 and blue label for dCas9) or tru-gRNA (purple circles) bound to each feature on the substrate. Note that dCas9 with tru-gRNA is only expected to interact with the first 3 or 8 PAM-distal mismatches of the 5MM and 10MM sites (here labeled "3MM" and "8MM", respectively). See Table 2 for standard errors of the average volumes and heights. For Cas9 / dCas9 with sgRNA, its structural properties at each feature are statistically different (δ-ε, α-ε: p<0.05; α-β: p<0.005; β-γ, γ-δ: p<<0.0005. Hotelling's T2 test). [Figure 4A-4C]Kinetic Monte Carlo (KMC) experiments reveal differences in the stability of R-loops or structures formed by the protospacer duplex with the invading guide RNA within stably bound Cas9 for different guide RNA variants. Figure 4A shows a schematic of strand invasion of the protospacer (green) by the guide RNA (red) for a KMC experiment. The R-loop is highlighted. The transition rates for invasion (vf, for rate, m→m+1, where m is the extent of strand invasion, i.e., equivalently, the length of the R-loop), or duplex reannealing (vr, for rate, m→m-1) are functions of nearest neighbor DNA:DNA and RNA:DNA hybridization energies. For details, see the main text and supplemental methods. Figure 4B shows the time fraction that an R-loop is of size m "at equilibrium" for sg-RNA (red) or tru-gRNA (purple) resulting from a KMC experiment (simulations started at m = 20 or 18, respectively). Simulations run until t > 10,000 (arbitrary units). Figure 4C shows the kinetic Monte Carlo time course of R-loop "breathing" for sgRNA (red) and tru-gRNA (purple) after complete invasion (simulations started at m = 20 or 18, respectively). Asterisks highlight the start position of the simulation. (Inset) Histograms of the respective lifetimes where R-loops are > 16 bp long. [Figure 4D]FIG. 4D shows a proposed model for the mechanism governing Cas9 / dCas9 specificity, based on the results of AFM imaging and kinetic Monte Carlo (KMC) experiments (see text). Cas9 / dCas9 binds to the PAM and the guide RNA invades the PAM-adjacent protospacer duplex. During this strand invasion, the guide RNA will displace the complementary strand of the protospacer. Competition between invasion and duplex reannealing results in a dynamic ("wobble") R-loop structure. The stability of the protospacer-guide RNA interaction at sites 14-17 is dramatically increased by binding at sites 19 and 20, promoting a conformational change in Cas9 / dCas9 that permits DNA cleavage in Cas9. [Figure 5A-5B] We show that kinetic Monte Carlo (KMC) experiments reveal differences in the ability to cross mismatches (MMs) and invade protospacers depending on the guide RNA structure. Figures 5A-5B show the occupancy ratios of R-loop lengths m over time for invading sgRNAs (Figure 5A) or tru-gRNAs (Figure 5B) from KMC experiments (starting at m=10, highlighted by an asterisk). White X's indicate the location of mismatches. Simulations are run until t≧10,000 (arbitrary units) and results are averaged over 100 trials. [Figure 5C] Figure 5C shows a representative KMC time course for strand invasion (starting at m=10) with a mismatch site at m=14 (arrow) for sgRNA (red) and tru-gRNA (purple). While the sgRNA is highly stably invaded after bypassing the mismatch, the tru-gRNA is repeatedly recaptured behind the mismatch as a result of the inherent instability of its R-loop (see Figure 4). [Figure 6A-6B]We show that experimental (Hsu et al. (2013) Nature biotechnology, 31, 827-832) cleavage frequencies at target sites containing single rG·dG, rC·dC, rA·dA, and rU·dT mismatches in the PAM-distal region (≥10 protospacer sites) correlate with R-loop stability determined from kinetic Monte Carlo experiments. Figure 6A shows the log10 (p-value) of the correlation between Cas9 cleavage frequency and R-loop stability at site m (fraction of time that guide RNA remains bound to the protospacer at site m, see text) during strand invasion initiated at site m. (i) The stability at sites m=10 to m=14 is highly inversely correlated with the probability that the guide RNA will detach from the protospacer before crossing the mismatch (Figure 15B), whereas (ii) sites m=14 to m=17 are associated with conformational changes that induce cleavage activity (from AFM images). Colors correspond to correlation coefficients. Figure 6B shows that the experimental cleavage frequency does not significantly correlate with the estimated binding free energy (ΔG°37) of the guide RNA-protospacer equilibrium (left), but does correlate with the stability of site m=14 during strand invasion (right). Error bars are standard errors of the mean occupancy time at site m=14. For these kinetic Monte Carlo experiments, max(t)=100 (arbitrary units). Color bars are used to indicate the position of the mismatch (MM) sites. [Figure 7A-7C]We outline a proposed mechanism by which guide RNA structure influences Cas9 / dCas9 specificity. Figure 7A shows that for single-stranded guide RNAs (sgRNAs), only the first few nucleotides of the RNA (which binds to sites 18-20 of the protospacer) stabilize the R-loop wobble and binding at sites 14-17 of the protospacer, allowing an efficient conformational transition to the active state to allow cleavage. However, this increased stability conferred by these bases allows for transient stabilization at the mismatch site and a conformational change that allows cleavage. In many cases, across the mismatch, the R-loop remains stably fully invaded. Figure 7B shows that for guide RNAs truncated by only the first few (here 2) nucleotides (tru-gRNAs), the reduced stability of the R-loop (characterized by significant instability) reduces the probability of maintaining an active conformation. If a mismatch site is present within the protospacer, the instability of the R-loop ensures that it is rapidly and repeatedly "recaptured" behind the mismatch, becoming highly impeded at this site. Figure 7C shows that while "simple" extensions at the 5' end of guide RNAs for targeting the protospacer and adjacent sites beyond the protospacer were found to be digested back to approximately the length of the sgRNA in vivo (Figure 7A), guide RNAs with 5'-hairpins complementary to the "PAM-distal" targeting segment (hp-gRNA) are expected to remain protected within the Cas9 / dCas9 structure prior to invasion. After binding to the PAM site by the hp-gRNA and initiating strand invasion, when bound to the complete protospacer, the hairpin opens and full strand invasion can occur. If a PAM-distal mismatch is present at the target site, it is energetically more favorable for the hairpin to remain closed and strand invasion is impeded. The ability of Cas9-hp-gRNA to cleave RNA remains to be confirmed. [Figure 8A-8B]The purity of expressed Cas9 and dCas9 in SDS gels of purified Cas9 (Figure 8A) and dCas9 (Figure 8B) products (nominal molecular weight: 160 kDa) is shown. The eluted band indicates that the products are approximately 95% pure. [Figure 9A-9C] Additional images of Cas9 / dCas9 bound to DNA are shown. A) dCas9 binding distribution (compare with FIG. 1) to substrates that do not contain homology to the AAVs1 protospacer sequence (n=443). Superimposed are the cumulative distribution (CDF) of PAM sites (CDFPAM, black) and the CDF of bases bound by dCas9 (red, CDFCas9). To avoid artifacts introduced by overlap with streptavidin tags (criterion for DNA selection) and binding of DNA to exposed blunt ends (leading to an expected increase in non-specific binding), the comparison starts 100 bases from each end. B) Absolute difference Dn between the CDF of protein binding and the CDF of PAM sites. The dashed line is the Kolmogorov-Smirnov criterion for the goodness of fit of the two distributions. C) Using MATLAB, the CDF of binding was compared to the CDF of the PAM distribution from 100,000 randomly generated sequences with equal probabilities of G, A, T, and C. The vertical red line is the experimental Sup(Dn), indicating that the experimental dCas9 binding matched the experimental PAM distribution more closely than 71.20% of the generated sequences. [Figure 10A-10C]Showing binding to a "nonsense" substrate (>3 bp) that contains no homology to the protospacer sequence. (A) Image of dCas9 alone. (B) Volume (left) and height (right) histograms (n=423) of dCas9 imaged using a Gaussian fit to the first peak alone. From the Gaussian fit: mean height is 1.746 nm (95% confidence: 1.689 nm - 1.802 nm), standard deviation 0.441 nm; mean volume is 1302 nm3 (95% confidence: 1266 nm3 - 1337 nm3), standard deviation 259.1 nm3 (note that here dCas9 does not have DNA in its binding channel, so its recorded volume may appear artificially low due to reduced mechanical resistance to the AFM probe). Heights were measured relative to the median of a 10-pixel region surrounding each protein, and volumes were recorded as adjacent features greater than two times the standard deviation of the local background height. (C) Additional representative images of dCas9 bound to DNA that has been labeled at one end with monovalent streptavidin. [Figures 11A-11D] Representative diagrams of dCas9-sgRNA bound to RNA and examples of processing of protein structural features are shown. Figure 11A shows a representative wide-field image of dCas9 bound to genetically modified DNA. Figure 11B shows a close-up of the boxed area. The white arrow is monovalent streptavidin, and the red arrow is dCas9 protein. Figures 11C-11D show examples of extraction from the original image (Figure 11C) and isolation of Cas9 / dCas9 structures (Figure 11D). This extraction was repeated for each isolated protein bound to DNA, which was then aligned pairwise through repeated translation, rotation, and inversion to minimize their mean squared morphological differences. From these minimized mean squared differences, we constructed a distance matrix and clustered each protein according to the method of Laio and Rodriguez (2014) Science (New York, NY), 344, 1492-1496, and then mapped the ensemble of structures by clustering them back to their sites on the DNA (Figure 2A, Figures 10A-10C). [Figure 12A] Characterization of Cas9 / dCas9-sgRNAs mapped to their respective binding sites is shown. Top: Stacked histograms of aligned structures (right, see text) clustered by volume (left), maximum height (middle), and mean square difference (MSD) for all experimental conditions. Populations are colored according to binned volume, height, or structural clusters, similar to the scatter plots below. The binding distributions of extracted Cas9 / dCas9 molecules (Figures 10A-10C) closely match those of the full dataset (Figures 1C-1D, Figures 8A-8B), indicating that the selection procedure is unbiased and the selected proteins are representative of the entire dataset. Bottom: Scatter plots of volumes for (left) maximum heights, (middle) maximum heights, and (right) structural clusters for all Cas9 / dCas9 molecules colored by binned volumes. [Figure 12B] Characterization of Cas9 / dCas9-sgRNAs mapped to their respective binding sites is shown. Top: Stacked histograms of aligned structures (right, see text) clustered by volume (left), maximum height (middle), and mean square difference (MSD) for all experimental conditions. Populations are colored according to binned volume, height, or structural clusters, similar to the scatter plots below. The binding distributions of extracted Cas9 / dCas9 molecules (Figures 10A-10C) closely match those of the full dataset (Figures 1C-1D, Figures 8A-8B), indicating that the selection procedure is unbiased and the selected proteins are representative of the entire dataset. Bottom: Scatter plots of volumes for (left) maximum heights, (middle) maximum heights, and (right) structural clusters for all Cas9 / dCas9 molecules colored by binned volumes. [Figure 13]Structural characteristics of Cas9 / dCas9 with tru-gRNA and hp-gRNA at their respective binding sites. Proportion of bound DNA occupied by Cas9 / dCas9 along the genetically modified DNA substrate with colors representing the population of Cas9 / dCas9 clustered according to their structures (see Figure 3C). Protein structures were classified according to the dCas9 / Cas9 with the sgRNA to which it most closely resembles (by mean squared difference after alignment (see text)). For reference, positions of the complete protospacer site on the genetically modified DNA substrate: 144-167 bp; positions of the 10MM (8MM) site: 452-465 bp; positions of the 5MM (3MM) site: 592-610 bp. A similar trend was observed as seen with dCas9 / Cas9 with sgRNA: as dCas9 binds to mismatched and matching sites, the proportion of the population that clusters with the largest (yellow) group increases, but this effect is suppressed for tru-gRNA, where even at the perfect protospacer site, a significant proportion of the population clusters with the smaller (green and blue) groups. The effect for hp10-gRNA is particularly pronounced, highlighting its poor affinity for off-target sites. [Figures 14A-14C]A model of strand invasion of DNA protospacers by guide RNA and the estimated binding stability of RNA invaded into protospacers with PAM-distal mismatches is shown. Figure 14A shows a schematic model of strand invasion of DNA protospacers by guide RNA. See also Figure 4A. Guide RNA is assumed to dissociate when m=1. Figure 14B shows the calculated probability distribution of dissociation times for guide RNAs that first invade to m=5 for protospacers with different numbers of adjacent PAM-distal mismatches. The length of these dissociation times can be considered as an approximation of dCas9 binding propensity at that site. Asterisks highlight the dissociation times for the population of guide RNAs that cannot invade completely the first time after the first invasion to m=5. Invaded RNA is highly unstable at protospacer sites with 15 PAM-distal mismatches (15MM), and in experiments we rarely observe Cas9 / dCas9 bound to these sites (Figure 1D). Invaded RNA (before dissociation) at protospacer sites with 10 or 5 PAM-distal mismatches (10MM and 5MM) is expected to remain significantly longer (albeit within two orders of magnitude) than RNA at the 15MM site; we find that its binding propensity is nearly equal to or lower than that of the perfect protospacer site (0MM) in AFM experiments. The probability density function was calculated using the Q-matrix method as described (Sakmann et al. (1995) Single-channel recording, Springer; 2nd ed.) using sequence-specific transition rates between m states (vf and vr, see Supplementary Methods). Figure 14C shows that examination of the estimated half-lives of RNA-protospacer binding at protospacers with different numbers of PAM distal mismatches suggests that there are approximately three types of invaded RNA with similar stability: those with >11 PAM distal mismatches (low stability); those with 3 to 11 PAM distal mismatches (moderate stability); and those with <3 PAM distal mismatches (high stability).These results are qualitatively similar to the distribution of dCas9 on genetically modified substrates observed via AFM (Figure 1D). [Figure 15] Figure 1 shows the average simulated first arrival times across the mismatch site during strand invasion by sgRNA and tru-gRNA. Average simulated (Kinetic Monte Carlo) first arrival times across the mismatch site during strand invasion by sgRNA (blue) and tru-gRNA (red) for different positions of the mismatch site. Error bars are the standard deviation of the recorded first arrival times. Sequence of the protospacer in the box (AAVS1 site). [Figure 16A-16B] Correlation between Cas9 cleavage frequency (Hsu et al. (2013) Nature Biotechnology, 31, 827-832) and R-loop stability measurements from kinetic Monte Carlo. Figure 16A shows that the statistical power and strength of the correlation between the stability of the R-loop site (from kinetic Monte Carlo, see text) and the cleavage frequency of the experiment by Hsu et al. (2013) Nature Biotechnology, 31, 827-832 decreases with increasing simulation length (max(t)=100 to max(t)=1000, arbitrary units). This result suggests that the kinetics of strand invasion may be an important predictor of off-target cleavage rates. Figure 16B shows the correlation between the fraction of time that the R-loop is of size m and the probability that the kinetic Monte Carlo study predicts that the invading strand will dissociate before crossing the mismatch. Binding at sites 10 to 14-15 correlates very strongly inversely (-0.5-0.85) with the probability of dissociation before crossing a mismatch, whereas from AFM imaging experiments we find that binding at sites -> 16 is associated with a conformational change in Cas9 / dCas9. [Figure 17] An overview of deep sequencing data comparing on-target activity is shown. [Figure 18] 1 shows an overview of deep sequencing data comparing specificity enhancement. [Figure 19]Lane 1 shows protospacer 1, dystrophin; lane 1 shows GFP control; lane 2 shows the complete gRNA; lane 3 shows Tru-gRNA 19nt; lane 4 shows Tru-gRNA 18nt; lane 5 shows Tru-gRNA 17nt; lane 6 shows Tru-gRNA 16nt; lane 7 shows Hp-gRNA 4bp; lane 8 shows Hp-gRNA 5bp; lane 9 shows Hp-gRNA 6bp; lane 10 shows Hp-gRNA 7bp; lane 11 shows Hp-gRNA 8bp; lane 12 shows Hp-gRNA 9bp, hairpin 1 (lane 12, 9nt hp) -GtgagtaggttcgCCTACTCAGACTGTTACTC (SEQ ID NO: 335) (where tgagtagg is part of the hairpin and the underlined part is the hairpin loop). [Figure 20] Protospacer 1, dystrophin, internal loops shown. [Figure 21A] Figure 1 shows the calculated secondary structure of the 5' end of the protospacer-targeting segment of the hp-gRNA used for deep sequencing experiments (using the NuPack software suite). The colors are the probability of each nucleotide being present in the equilibrium secondary structure. [Figure 21B] Figure 1 shows the calculated secondary structure of the 5' end of the protospacer-targeting segment of the hp-gRNA used for deep sequencing experiments (using the NuPack software suite). The colors are the probability of each nucleotide being present in the equilibrium secondary structure. [Figure 22] Dystrophin, indel rate, (all sites) shown. [Diagram 23] Dystrophin, on-target / total (off-targets) shown. [Figure 24]Lane 1 shows the GFP control; lane 2 shows the complete gRNA; lane 3 shows the Tru-gRNA; lane 4 shows the 10 bp hp-gRNA; lane 5 shows the 6 bp hp-gRNA, Hairpin 1. Transformations - Surv_OT1=DS_OT2; Surv_OT53=DS_OT3. [Figure 25A] Protospacer 2, EMX1, tru-hps, and internal loops are indicated. [Figure 25B] Protospacer 2, EMX1, tru-hps, and internal loops are indicated. [Figure 26A-26B] Hairpin structures are shown. Figure 26A shows Hairpin 1, a 6 bp 5'-hairpin. Figure 26B shows Hairpin 2, a 5 bp 5'-hairpin on an 18 nt (truncated) gRNA. [Figure 26C] Hairpin structures are shown in Figure 26C, which shows hairpin 3, a 3 bp 5'-hairpin. [Figure 27] EMX1, indel rate (all sites) shown. [Figure 28] EMX1, showing indel rate (low rate / off-target). [Figure 29] EMX1, showing on-target / total (off-target) [Diagram 30] Protospacer 3, VEGFA1 is shown. Lane 1 shows GFP control; lane 2 shows full gRNA; lane 3 shows Tru-gRNA; lane 4 shows 10 bp hp-gRNA; lane 5 shows 6 bp hp-gRNA. [Figure 31A] Protospacer 3, VEGFA1:pam proximal hairpin is shown. Lane 1 shows GFP control; lane 2 shows complete gRNA; lane 3 shows hp-gRNA1; lane 4 shows hp-gRNA2; lane 5 shows hp-gRNA3; lane 6 shows hp-gRNA4; lane 7 shows hp-gRNA5; lane 8 shows hp-gRNA6. [Figure 31B]Protospacer 3, VEGFA1:pam proximal hairpin is shown. Lane 1 shows GFP control; lane 2 shows complete gRNA; lane 3 shows hp-gRNA1; lane 4 shows hp-gRNA2; lane 5 shows hp-gRNA3; lane 6 shows hp-gRNA4; lane 7 shows hp-gRNA5; lane 8 shows hp-gRNA6. [Diagram 32] Protospacer 3, VEGFA1:pam proximal hairpin is shown. [Diagram 33] Protospacer 3, VEGF1, internal loop are shown. Lane 1 shows control; lane 2 shows complete; lane 3 shows 2nt hp; lane 4 shows 3nt hp, hairpin 5; lane 5 shows 4nt hp. [Fig. 34A-34B] Deep sequencing experiments for hairpins 1, 2, and failed 3 are shown. Figure 25A shows hairpin 4 - a computationally derived hairpin designed to identify off-target site 2 while maintaining on-target activity. Figure 25B shows hairpin 5 - a 4 bp 5'-hairpin (gRNAs usually have significant 3' secondary structure). [Diagram 35] VEGF1, indel rate, (all sites) shown. [Diagram 36] VEGF1, indel rate, (low rate, off-target). [Figure 37] VEGF1, on-target / total (off-target) indicated. [Figure 38] Protospacer 4, VEGFA3 is shown. Lane 1 shows GFP control; lane 2 shows full gRNA; lane 3 shows Tru-gRNA; lane 4 shows 3 bp hp-gRNA; lane 5 shows 4 bp hp-gRNA; lane 6 shows 5 bp hp-gRNA; lane 7 shows 6 bp hp-gRNA, and lane 8 shows 10 bp hp-gRNA. [Figure 39A]gRNA4, VEGFA3:pam proximal hairpin. Lane 1 shows GFP control; lane 2 shows complete gRNA; lane 3 shows hp-gRNA1; lane 4 shows hp-gRNA2; lane 5 shows hp-gRNA3; lane 6 shows hp-gRNA4; lane 7 shows hp-gRNA5; lane 8 shows hp-gRNA6. [Figure 39B] gRNA4, VEGFA3:pam proximal hairpin. Lane 1 shows GFP control; lane 2 shows complete gRNA; lane 3 shows hp-gRNA1; lane 4 shows hp-gRNA2; lane 5 shows hp-gRNA3; lane 6 shows hp-gRNA4; lane 7 shows hp-gRNA5; lane 8 shows hp-gRNA6. [Fig. 40A-40B] Figure 40A shows the 1-4 bp hairpin targeting 3'-region of a hairpin.Figure 40B shows the 2-4 bp hairpin targeting 3'-region of a hairpin with a GU wobble pair. [Figure 40C] The 3-4 bp hairpin targeting 3'-region of the hairpin with a GU wobble pair (variant design) is shown. [Diagram 41] VEGF3, indel rates (all sites) shown. [Diagram 42] VEGF3, showing indel rate (low rate / off-target). [Diagram 43] VEGF3, on-target / total (off-target) indicated. [Fig. 44A-44B] Figure 44A shows a hairpin designed to target the EMX1 gene. Figure 44B shows the EMX1-sg1 sequence of the hairpin in Figure 44A. [Figure 44C] The effect of decreasing protospacer length and increasing hairpin length on specificity is shown. [Figure 45A] The DNA / RNA sequence is shown. [Figure 45B] The DNA / RNA sequence is shown. [Figure 45C] The DNA / RNA sequence is shown. [Figure 45D]The DNA / RNA sequence is shown. [Diagram 46] FIG. 1 shows a diagram illustrating the Surveyor assay. [Figure 47] Figure 1 shows tolerance of AsCpf1 and LbCpf1 to mismatched or truncated crRNAs, and endogenous gene modification by AsCpf1 and LbCpf1 using crRNAs containing a single mismatched base. Activity determined by T7E1 assay; error bars, sem; n=3 (adapted from Kleinstiver et al., Nat. Biotech. 34:869-875). [Figure 48] Surveyor assay results for hp-gRNA used with V-type CRISPR system, in which a hairpin is added to the 3' end of the full-length gRNA to eliminate off-target activity, are shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] Disclosed herein are compositions and methods for site-specific DNA targeting and epigenomic gene editing and / or transcriptional regulation, such as DNA cleavage and gene activation or suppression.The present invention is directed to a modular method for designing and using optimized guide RNAs with hairpin structures (hpgRNAs), which can be easily integrated into existing biotechnology infrastructure and provide controlled reduction of off-target activity while still retaining the ability to specifically target precise DNA sequences.The methods described herein provide a novel approach to designing optimized gRNAs to perform methods that are significantly superior to other available methods, and can be used in particular in combination with other increasingly improved protein-specific means for highly improved performance.

[0022] The disclosed method and optimized gRNA have the great advantage of being easily applied to current methodologies and infrastructure already in place for performing RNA-guided genome engineering. In some embodiments, Cas9, dCas9, or Cpf1 is delivered to cells using a viral vector together with a vector encoding a transcript of the optimized gRNA in the cell. The present invention would require only a few additional nucleotides to the vector encoding the optimized gRNA, which can be easily adapted by current standard techniques. Like truncated guide RNAs (tru-gRNAs), the optimized gRNA or hpgRNA can be used in combination with, for example, paired nickases, or other modified forms of the endonuclease itself, to further improve specificity. A series of experiments were performed in vitro, which showed that the use of the optimized gRNA produced using the methods described herein increased specificity in DNA binding compared to the best available gRNA options (see Figure 2). The use of this optimized gRNA eliminates or significantly weakens the activity at targets that contain only a few mismatched DNA sequences, which tend to be the sites where off-target activity by RNA-guided endonucleases occurs.This optimized gRNA also provides specificity of cleavage activity in mammalian cells at sites where even the best known improvements to guide RNA are known to induce off-target activity.The present invention is a generally applicable method for reducing the off-target activity by RNA-guided endonucleases, especially Cas9, by technically changing the structural design of guide RNA.

[0023] 1.Definition The terms "comprise," "including," "having," "having," "can," "containing," and variations thereof, as used herein, are intended to be open-ended transitional phrases, terms, or words that do not exclude additional actions or structural possibilities. The singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. The present disclosure also contemplates other embodiments that "comprise," "consist of," and "consist essentially of" the embodiments or elements provided herein, whether or not expressly described.

[0024] For the recitation of numerical ranges herein, each intervening number is specifically contemplated to the same degree of precision. For example, for the range 6 to 9, the numbers 7 and 8 are specifically contemplated in addition to 6 and 9, and for the range 6.0 to 7.0, the numbers 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, and 7.0 are specifically contemplated.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art. In case of conflict, the present document, including definitions, will control. Preferred methods and materials are described below, but methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. The materials, methods, and examples disclosed herein are illustrative only and are not intended to be limiting.

[0026] "Adeno-associated virus" or "AAV", as used interchangeably herein, refers to a small virus belonging to the Dependovirus genus of the Parvoviridae family that infects humans and some other primate species. AAV is not currently known to cause disease, and therefore the virus induces a very mild immune response.

[0027] As used herein, "binding region" refers to the region within a nuclease target region that is recognized and bound by a nuclease, such as Cas9.

[0028] As used herein, "chromatin" refers to the organized complex of chromosomal DNA associated with histones.

[0029] "Cis-regulatory element" or "CRE", as used interchangeably herein, refers to a region of non-coding DNA that regulates the transcription of nearby genes. CREs are found in the vicinity of the gene(s) they regulate. CREs generally regulate gene transcription by functioning as binding sites for transcription factors. Examples of CREs include promoters, enhancers, super-enhancers, silencers, insulators, and locus control regions.

[0030] "Clustered Regularly Interspaced Short Palindromic Repeats" and "CRISPR", used interchangeably herein, refer to loci containing multiple short direct repeats found in approximately 40% of sequenced bacterial and 90% of sequenced archaeal genomes.

[0031] As used herein, "coding sequence" or "encoding nucleic acid" refers to a nucleic acid (RNA or DNA molecule) that comprises a nucleotide sequence that codes for a protein. The coding sequence may further comprise a start and stop signal operably linked to regulatory elements, including a promoter and a polyadenylation signal, capable of directing expression in cells of an individual or mammal to which the nucleic acid is administered. The coding sequence may be codon-optimized.

[0032] As used herein, "complement" or "complementary" means that a nucleic acid can undergo Watson-Crick (e.g., AT / U and CG) or Hoogsteen base pairing between nucleotides or nucleotide analogs of a nucleic acid molecule. "Complementarity" refers to the property shared between two nucleic acid sequences such that when aligned antiparallel to one another, the nucleotide bases at every position are complementary.

[0033] As used herein, "correcting", "genome editing" and "repairing" refer to changing a mutant gene that codes for a truncated protein or no protein at all to obtain full-length functional or partially full-length functional protein expression. Correcting or repairing a mutant gene can include replacing a region of the gene that has a mutation, or replacing the entire mutant gene with a copy of the gene that does not have the mutation using a repair mechanism such as homology-directed repair (HDR). Correcting or repairing a mutant gene can also include repairing a frameshift mutation that results in a premature stop codon, an ectopic splice acceptor site, or an ectopic splice donor site by creating a double-strand break in the gene and then repairing the gene using non-homologous end joining (NHEJ). NHEJ can add or delete at least one base pair during repair, which can restore the proper reading frame and remove the premature stop codon. Correcting or repairing a mutant gene can also include destroying the ectopic splice acceptor site or splice donor sequence. Correcting or repairing a mutated gene can also include deleting a non-essential gene segment by the simultaneous action of two nucleases on the same DNA strand to restore the proper reading frame by removing the DNA between two nuclease target sites, and repairing the DNA break by NHEJ.

[0034] As used herein, "demethylase" refers to an enzyme that removes methyl (CH3-) groups from nucleic acids, proteins (especially histones), and other molecules. Demethylases are important in epigenetic modification mechanisms. Demethylase proteins alter the transcriptional regulation of the genome by controlling the level of methylation that occurs on DNA and histones, which in turn regulates the chromatin state at specific loci within an organism. "Histone demethylase" refers to a methylase that removes methyl groups from histones. There are several families of histone demethylases that act on different substrates and play different roles in cellular functions. The Fe(II)-dependent lysine demethylase can be a JMJC demethylase. The JMJC demethylase is a histone demethylase that contains a JumonjiC (JmjC) domain. The JMJC demethylase can be a member of the KDM3, KDM4, KDM5, or KDM6 family of histone demethylases.

[0035] "DNase I hypersensitive site" or "DHS," as used interchangeably herein, refers to a docking site for transcription factors and chromatin modifiers, including p300, which regulate distal target gene expression.

[0036] "Donor DNA," "donor template," and "repair template," as used interchangeably herein, refer to a double-stranded DNA fragment or molecule that contains at least a portion of a gene of interest. The donor DNA can encode a fully functional protein or a partially functional protein.

[0037] "Endogenous gene" as used herein refers to a gene originating in an organism, tissue, or cell. An endogenous gene is native to a cell in its normal genomic and chromatin context and is not heterologous to the cell. Such cellular genes include, for example, animal genes, plant genes, bacterial genes, protozoan genes, fungal genes, mitochondrial genes, and chloroplast genes. "Endogenous target gene" as used herein refers to the endogenous gene targeted by optimized gRNA and CRISPR / Cas9-based or CRISPR / Cpf1-based systems.

[0038] "Enhancer" as used herein refers to a non-coding DNA sequence that contains multiple activator and repressor binding sites. Enhancers can range from 50bp to 1500bp in length and can be proximal, i.e., 5' upstream to the promoter, within any intron of the regulated gene, or distal, i.e., within the intron or intergenic region of an adjacent gene away from the locus, or on a different chromosome. More than one enhancer can interact with a promoter. Similarly, enhancers can regulate more than one gene without being limited to linkage, and can "skip" adjacent genes to regulate more distant genes. Transcriptional regulation can involve elements located on a different chromosome than the chromosome on which the promoter resides. The proximal enhancer or promoter of an adjacent gene can serve as a platform for recruiting more distal elements.

[0039] "Duchenne muscular dystrophy" or "DMD", used interchangeably herein, refers to a recessive, lethal, X-linked disorder that results in muscle degeneration and ultimately death. DMD is a common, inherited, monogenic disease that affects 1 in 3500 males. DMD is the result of inherited or spontaneous mutations that result in nonsense or frameshift mutations in the dystrophin gene. The majority of dystrophin mutations that cause DMD are exon deletions that disrupt the reading frame in the dystrophin gene, causing premature translation termination. DMD patients typically lose the ability to support themselves during childhood, develop progressive muscle weakness during their teenage years, and die in their twenties.

[0040] "Dystrophin" as used herein refers to a rod-shaped cytoplasmic protein that is part of a protein complex that connects the cytoskeleton of muscle fibers to the surrounding extracellular matrix through the cell membrane. Dystrophin provides structural stability to the dystroglycan complex of the cell membrane, which is responsible for regulating muscle cell integrity and function. The dystrophin gene or "DMD gene", as used interchangeably herein, is 2.2 megabases at the locus Xp21. The primary transcript is approximately 2,400 kb, and the mature mRNA is approximately 14 kb. 79 exons code for a protein that is more than 3500 amino acids.

[0041] "Exon 51" as used herein refers to the 51st exon of the dystrophin gene. Exon 51 is frequently adjacent to frame-breaking deletions in DMD patients and has been targeted in clinical trials for oligonucleotide-based exon skipping. A clinical trial for the exon 51 skipping compound eteplirsen recently reported significant functional benefit over 48 weeks, with an average of 47% dystrophin-positive fibers compared to baseline. Mutations in exon 51 are ideally suited for permanent correction by NHEJ-based genome editing.

[0042] As used herein interchangeably, "frameshift" or "frameshift mutation" refers to a type of genetic mutation in which the addition or deletion of one or more nucleotides results in a shift in the reading frame of a codon in an mRNA. The shift in reading frame can result in a change in the amino acid sequence in protein translation, such as a missense mutation or a premature stop codon.

[0043] A "full-length gRNA" or "standard gRNA," used interchangeably herein, refers to a gRNA that includes a "backbone," which is generally 20 nucleotides in length, and a protospacer-targeting sequence or segment.

[0044] As used herein, "functional" and "fully functional" describe a protein that has biological activity. A "functional gene" refers to a gene that is transcribed into mRNA that is translated into a functional protein.

[0045] As used herein, a "fusion protein" refers to a chimeric protein created through the joining of two or more genes that originally encoded separate proteins. Translation of the fusion gene results in a single polypeptide possessing functional properties derived from each of the original proteins.

[0046] "Genetic construct" as used herein refers to a DNA or RNA molecule that comprises a nucleotide sequence that codes for a protein.The coding sequence comprises start and stop signals that are operably linked to regulatory elements, including promoters and polyadenylation signals, that can induce expression in the cells of the individual to which the nucleic acid molecule is administered.As used herein, the term "expressible form" refers to a genetic construct that contains the necessary regulatory elements that are operably linked to the coding sequence that codes for a protein, such that the coding sequence is expressed when present in the cells of an individual.

[0047] "Genetic disease" as used herein refers to a disease, especially a condition present from birth, caused in part or in whole, directly or indirectly, by one or more abnormalities in the genome. The abnormality may be a mutation, an insertion, or a deletion. The abnormality may affect the coding sequence of a gene or its regulatory sequence. The genetic disease may be, but is not limited to, DMD, hemophilia, cystic fibrosis, Huntington's disease, familial hypercholesterolemia (LDL receptor deficiency), hepatoblastoma, Wilson's disease, congenital hepatic porphyria, inherited disorders of hepatic metabolism, Lesch-Nyhan syndrome, sickle cell anemia, thalassemia, xeroderma pigmentosum, Fanconi anemia, retinitis pigmentosa, ataxia telangiectasia, Bloom's syndrome, retinoblastoma, and Tay-Sachs disease.

[0048] As used herein, "genome" refers to the complete set of genes or genetic material present in a cell or organism. Genomes include DNA, or RNA in RNA viruses. Genomes include both genes, (coding regions), non-coding DNA, and the genomes of mitochondria and chloroplasts.

[0049] "Guide RNA", "gRNA", "single-stranded gRNA", and "sgRNA", as used interchangeably herein, refer to short synthetic RNAs consisting of a "backbone" sequence essential for Cas9- or Cpf1-binding and a user-defined "spacer" or "targeting sequence" (also referred to herein as a protospacer-targeting sequence or segment) that defines the genomic target to be modified. "hpgRNA", "hp-gRNA", and "optimized gRNA", as used interchangeably herein, refer to gRNAs with additional nucleotides at either the 5' or 3' end that can form a secondary structure with all or a portion of the protospacer-targeting sequence or segment.

[0050] "Histone acetyltransferase" or "HAT" are used interchangeably herein to refer to enzymes that acetylate conserved lysine amino acids on histone proteins by transferring an acetyl group from acetyl-CoA to form ε-N-acetyl-lysine. DNA wraps around histones and can turn genes on and off by transferring an acetyl group to the histone. In general, histone acetylation leads to transcriptional activation and, as it relates to euchromatin, increases gene expression. Histone acetyltransferases can also acetylate non-histone proteins such as nuclear receptors and other transcription factors to promote gene expression.

[0051] "Histone deacetylase" or "HDAC", as used interchangeably herein, refers to an enzyme that catalyzes the deacetylation of ε-N-acetyl lysine amino acids on histones by removing acetyl groups (O=C-CH 3 HDACs are also called lysine deacetylases (KDACs), referring to their function rather than their targets, which also include non-histone proteins.

[0052] "Histone methyltransferase" or "HMT," as used interchangeably herein, refers to histone modifying enzymes (e.g., histone-lysine N-methyltransferase and histone-arginine N-methyltransferase) that catalyze the transfer of one, two, or three methyl groups to lysine and arginine residues of histone proteins. Attachment of methyl groups occurs primarily at specific lysine or arginine residues on histones H3 and H4.

[0053] " Homologous recombination repair " or " HDR " used interchangeably herein refers to the mechanism in cells to repair double-stranded DNA damage when homologous fragments of DNA are present in the nucleus, usually in the G2 and S phases of the cell cycle. HDR uses donor DNA template to guide repair, and can be used to create specific sequence changes to genomes, including targeted addition of whole genes. When donor template is provided with site-specific nuclease, for example with CRISPR / Cas9-based system or CRISPR / Cpf1-based system, the cellular machinery will repair the break by homologous recombination, which is enhanced by several orders of magnitude in the presence of DNA break. In the absence of homologous DNA fragments, non-homologous end joining can occur instead.

[0054] As used herein, "genome" refers to the complete set of genes or genetic material present in a cell or organism. Genomes include DNA, or RNA in RNA viruses. Genomes include both genes, (coding regions), non-coding DNA, and the genomes of mitochondria and chloroplasts.

[0055] "Genome editing" as used herein refers to changing genes. Genome editing can include correcting or repairing mutated genes. Genome editing can include knocking out genes, such as mutated or normal genes. Genome editing can be used to treat disease or enhance muscle repair by changing targeted genes.

[0056] "Identical" or "identity" as used herein in the context of two or more nucleic acid or polypeptide sequences means that the sequences have a certain percentage of residues that are identical over a particular region. The percentage can be calculated by optimally aligning the two sequences, comparing the two sequences over a particular region, determining the number of positions where identical residues exist in both sequences to give the number of matched positions, dividing the number of matched positions by the total number of positions in the particular region, and multiplying the result by 100 to give the percentage of sequence identity. If the two sequences are of different lengths, or the alignment results in one or more sticky ends such that only a single sequence is included in a particular region of comparison, the residues of the single sequence are included in the denominator of the calculation, but not in the numerator. When comparing DNA and RNA, thymine (T) and uracil (U) can be considered equivalent. Identity can be performed manually or by using a computer sequence algorithm such as BLAST or BLAST 2.0.

[0057] "Insulators" as used herein refer to gene boundary elements that prevent interactions between enhancers and promoters. By being present between enhancers and promoters, insulators can inhibit subsequent interactions. Insulators can determine the set of genes that enhancers can affect. Insulators are needed when two adjacent genes on a chromosome have very different transcription patterns and inducing or suppressing one mechanism does not inhibit the adjacent gene. Insulators have also been found to cluster at the boundaries of topological association domains (TADs) and may play a role in distributing the genome into "chromosomal neighborhood"-genomic regions where regulation exists. Insulator activity is thought to arise primarily through the 3D structure of DNA mediated by proteins including CTCF. Insulators likely function through multiple mechanisms. Many enhancers form DNA loops that are placed in close physical proximity to promoter regions during transcriptional activation. Insulators can promote the formation of DNA loops that prevent promoter-enhancer loops from forming. Barrier insulators can prevent the spreading of heterochromatin from silenced genes to actively transcribed genes.

[0058] "Invasion" as used herein refers to a DNA duplex break at a protospacer region within a target region of a target gene, such as by a gRNA binding to a DNA sequence that is complementary to the protospacer.

[0059] "Invasion kinetics" as used herein refers to the rate at which invasion proceeds. Invasion kinetics can refer to the rate at which the guide RNA invades the duplex until "full invasion" where the protospacer is fully invaded, or the rate at which the segment of the protospacer DNA bound to the guide RNA is displaced from its complementary strand and extends from its PAM-proximal region to full invasion as it binds to the guide RNA, nucleotide by nucleotide.

[0060] "Lifespan" as used herein refers to the period during which a gRNA remains inserted into a region within the target region of a target gene.

[0061] As used herein, "locus control region" refers to a global cis-regulatory element that enhances the expression of linked genes at distal chromatin sites. It functions in a copy number-dependent manner and is tissue-specific, as seen in the selective expression of the β-globin gene in erythroid cells. The expression level of a gene can be modified by LCRs and gene proximal elements, such as promoters, enhancers, and silencers. LCRs function by recruiting chromatin modifications, coactivators, and transcription complexes. Their sequences are conserved in many vertebrates, and the conservation of certain sites may suggest importance in function.

[0062] "Mismatched" or "MM", as used interchangeably herein, refers to mismatched bases, including G / T or A / C pairings. The mismatch is usually due to tautomerization of the bases during G2. The damage is repaired by recognizing the deformation caused by the mismatch, determining the template and non-template strands, and excising the incorrectly incorporated base and replacing it with the correct nucleotide.

[0063] As used herein, "modulate" can mean any alteration of activity, such as modulating, down-regulating, up-regulating, decreasing, inhibiting, increasing, decreasing, deactivating, or activating.

[0064] As used herein interchangeably, "mutated gene" or "mutated gene" refers to a gene that has undergone a detectable mutation. A mutant gene has undergone an alteration, such as loss, gain, or exchange of genetic material, that affects the normal transmission and expression of the gene. As used herein, a "disrupted gene" refers to a mutant gene that has a mutation that results in a premature stop codon. The disrupted gene product is truncated compared to the full-length, undisrupted gene product.

[0065] As used herein, the term "non-homologous end joining (NHEJ) pathway" refers to a pathway that repairs double-strand breaks in DNA by directly ligating the broken ends without the need for a homologous template. Template-independent religation of DNA ends by NHEJ is a stochastic, error-prone repair process that introduces random microinsertions and microdeletions (indels) at the DNA breakpoint. This method can be used to intentionally disrupt, delete, or change the reading frame of targeted gene sequences. NHEJ generally uses short homologous DNA sequences, called microhomologies, to guide repair. These microhomologies are often present in single-stranded overhangs at the ends of double-strand breaks. When the overhangs are perfectly matched, NHEJ usually repairs the break accurately, while imprecise repairs resulting in loss of nucleotides can also occur, which is much more common when the overhangs are mismatched.

[0066] As used herein, a "normal gene" refers to a gene that has not undergone alteration, such as the loss, gain, or exchange of genetic material. A normal gene undergoes normal gene transmission and gene expression.

[0067] As used herein, "nuclease-mediated NHEJ" refers to NHEJ that is initiated after a nuclease, such as cas9, makes a double-stranded DNA break.

[0068] As used herein, "nucleic acid" or "oligonucleotide" or "polynucleotide" refers to at least two nucleotides that are covalently linked together. The depiction of a single strand also defines the sequence of the complementary strand. Thus, nucleic acid also encompasses the complementary strand of the depicted single strand. Many variants of nucleic acid can be used for the same purpose as a given nucleic acid. Thus, nucleic acid also encompasses substantially identical nucleic acids and their complements. A single strand provides a probe that can hybridize to a target sequence under stringent hybridization conditions. Thus, nucleic acid also encompasses probes that hybridize under stringent hybridization conditions.

[0069] Nucleic acids can be single-stranded or double-stranded, and can contain portions of both double-stranded and single-stranded sequences. Nucleic acids can be DNA, both genomic and cDNA, RNA, or hybrids, where the nucleic acids can contain combinations of deoxyribonucleotides and ribonucleotides, and combinations of bases including uracil, adenine, thymine, cytosine, guanine, inosine, xanthine hypoxanthine, isocytosine, and isoguanine. Nucleic acids can be obtained by chemical synthesis methods or by recombinant methods.

[0070] As used herein, "on-target site" refers to the target region or sequence in the genome that the gRNA is intended to target. Ideally, the on-target site has perfect homology (100% identity or homology) to the target DNA sequence and no homology exists elsewhere in the genome.

[0071] As used herein, an "off-target site" refers to a region of the genome that has partial homology or partial identity to the on-target site or target region of the gRNA, but that is not intended or designed to be targeted by the gRNA.

[0072] As used herein, "operably linked" means that the expression of a gene is under the control of a promoter that is spatially linked to the gene. The promoter can be located 5' (upstream) or 3' (downstream) of the gene under its control. The distance between the promoter and the gene can be approximately the same as the distance between the promoter and the gene that the promoter controls in the gene from which the promoter is derived. As is known in the art, variations in this distance can be accommodated without loss of promoter function.

[0073] As used interchangeably herein, "p300 protein", "EP300", or "E1A binding protein p300" refers to the adenovirus E1A-associated cellular p300 transcriptional coactivator protein, encoded by the EP300 gene. p300 is a highly conserved acetyltransferase involved in a wide range of cellular processes. p300 functions as a histone acetyltransferase that regulates transcription through chromatin remodeling and is involved in the processes of cell proliferation and cell differentiation.

[0074] As used herein, "partially functional" describes a protein that is encoded by a mutated gene and that has less biological activity than a functional protein, but more biological activity than a non-functional protein.

[0075] As used interchangeably herein, a "premature stop codon" or "out-of-frame stop codon" refers to a nonsense mutation in the sequence of DNA that results in a stop codon at a position not normally found in the wild-type gene. A premature stop codon can result in a truncated or shortened protein compared to the full-length form of the protein.

[0076] "Primary cells" as used herein refers to cells taken directly from living tissue (e.g., biopsy material). Primary cells can be established for growth in vitro. These cells have undergone few population doublings and therefore better represent the main functional elements of the tissue from which they are derived, compared to passaged (tumor or artificially immortalized) cell lines, and therefore are a more representative model for in vivo conditions. Primary cells can be taken from various species, such as mice or humans.

[0077] " Protospacer sequence " or " protospacer segment " used interchangeably herein refers to the DNA sequence targeted by Cas9 nuclease or Cpf1 nuclease in CRISPR bacterial adaptive immune system. In CRISPR / Cas9 system, protospacer sequence is generally followed by protospacer adjacent motif (PAM); PAM is at 5' end. In CRISPR / Cpf1 system, PAM is followed by protospacer sequence; PAM is at 3' end.

[0078] "Protospacer-targeting sequence" or "protospacer-targeting segment," as used interchangeably herein, refers to a nucleotide sequence of a gRNA that corresponds to a protospacer sequence and facilitates targeting of a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system to the protospacer sequence.

[0079] "Promoter" as used herein means a synthetic or naturally derived molecule capable of conferring, activating or promoting expression of a nucleic acid in a cell. A promoter can contain one or more specific transcriptional regulatory sequences to further promote expression and / or to alter the spatial expression and / or temporal expression of the same. A promoter can also be located as far away as several thousand base pairs from the start site of transcription or contain any distal enhancer or repressor element in the genome. Promoters can be obtained from sources including viruses, bacteria, fungi, plants, insects and animals. A promoter can constitutively or variably regulate the expression of genetic components with respect to the cell, tissue or organ in which expression occurs, or with respect to the developmental stage in which expression occurs, or in response to external stimuli such as physiological stress, hormones, toxins, drugs, chemicals, metal ions or inducers. Representative examples of promoters include the bacteriophage T7 promoter, the bacteriophage T3 promoter, the SP6 promoter, the lac operator-promoter, the tac promoter, the SV40 late promoter, the SV40 early promoter, the RSV-LTR promoter, the CMV IE promoter, the SV40 early promoter or the SV40 late promoter, and the CMV IE promoter.

[0080] As used herein, "protospacer adjacent motif" or "PAM" refers to the DNA sequence immediately following the DNA sequence targeted by Cas9 or immediately preceding the DNA sequence targeted by Cpf1 nuclease in the CRISPR bacterial adaptive immune system. PAM is a component of the invading virus or plasmid, but not of the bacterial CRISPR locus. Cas9 and Cpf1 will not successfully bind and cleave the target DNA sequence if it is not followed or preceded, respectively, by the PAM sequence. PAM is an essential targeting component (not found in the bacterial genome) that distinguishes the bacteria itself from non-self DNA, thereby preventing the CRISPR locus from being targeted and destroyed by nucleases.

[0081] The term "recombinant" when used with respect to a cell, or a nucleic acid, protein, or vector, indicates that the cell, nucleic acid, protein, or vector has been modified by the introduction of a heterologous nucleic acid or protein or the alteration of a naturally occurring nucleic acid or protein, or that the cell is derived from a cell so modified. Thus, for example, a recombinant cell expresses a gene not found in the native (naturally occurring) type of cell, or expresses a second copy of a naturally occurring gene that is instead normally or abnormally expressed, or under-expressed or not expressed at all.

[0082] "Silencer" or "repressor", used interchangeably herein, refers to a DNA sequence that can bind to a transcriptional regulator and prevent a gene from being expressed as a protein. Silencers are sequence-specific elements that induce a negative effect on the transcription of that particular gene. There are many locations where silencer elements can be located in DNA. The most common location is found upstream of the target gene, where they can help to suppress the transcription of the gene. This distance can vary widely, from approximately -20 bp upstream to -2000 bp upstream of the gene. Some silencers can be found downstream of the promoter, located within the introns or exons of the gene itself. Silencers have also been found within the 3' untranslated region (3'UTR) of mRNA. There are two main types of silencers in DNA: typical silencer elements and atypical negative regulatory elements (NREs). In typical silencers, the gene is actively repressed by the silencer element, mostly by inhibiting general transcription factor (GTF) assembly. NREs usually passively repress genes by inhibiting other elements that are upstream of the gene.

[0083] "Skeletal muscle" as used herein refers to a type of striated muscle that is under the control of the somatic nervous system and is attached to bones by bundles of collagen fibers known as tendons. Skeletal muscle is composed of individual components known as myocytes, or "muscle cells," sometimes colloquially called "muscle fibers." Muscle cells are formed from the fusion of developmental myoblasts (a type of embryonic precursor cell that gives rise to muscle cells) in a process known as myogenesis. These long, cylindrical, multinucleated cells are also called myofibers.

[0084] As used herein, "skeletal muscle condition" refers to conditions associated with skeletal muscle, such as muscular dystrophies, aging, muscle degeneration, wound healing, and muscle weakness or atrophy.

[0085] "Subject" and "patient", used interchangeably herein, refer to any vertebrate, including, but not limited to, mammals (e.g., cows, pigs, camels, llamas, horses, goats, rabbits, sheep, hamsters, guinea pigs, cats, dogs, rats, and mice, non-human primates (e.g., monkeys such as cynomolgus or rhesus monkeys, chimpanzees), and humans). In some embodiments, the subject can be human or non-human. The subject or patient may be undergoing other forms of therapy.

[0086] "Super enhancer" as used herein refers to a region of the mammalian genome that contains multiple enhancers that are collectively bound by a set of transcription factor proteins to drive the transcription of genes involved in cell identity. Super enhancers are often identified with nearby genes that are important for controlling and defining cell identity, and can be used to quickly identify key focal points that regulate cell identity. Enhancers have some quantifiable traits with a range of values, and these traits are generally elevated in super enhancers. Super enhancers are bound by higher levels of transcriptional regulatory proteins and are associated with more highly expressed genes. The expression of genes associated with super enhancers is particularly sensitive to perturbations, which can explain the sensitivity of super enhancer-associated genes to small molecules that facilitate cell state transitions or target transcription.

[0087] As used herein, a "targeted enhancer" refers to an enhancer that is targeted by a gRNA and CRISPR / Cas9-based system. The targeted enhancer may be within the target region.

[0088] As used herein, "target gene" refers to any nucleotide sequence that encodes a known or putative gene product. A target gene may be a mutated gene involved in a genetic disease.

[0089] "Target region," "target sequence," "protospacer," or "protospacer sequence," as used interchangeably herein, refer to the region of a target gene that is targeted by a CRISPR / Cas9-based or CRISPR / Cpf1-based system.

[0090] As used herein, the term "transcribed region" refers to a region of DNA that is transcribed into a single-stranded RNA molecule known as messenger RNA, which results in the transfer of genetic information from a DNA molecule to messenger RNA. During transcription, RNA polymerase reads the template strand in a 3' to 5' direction and synthesizes RNA from 5' to 3'. The mRNA sequence is complementary to the DNA strand.

[0091] As used herein, "target regulatory element" refers to a regulatory element targeted by a gRNA and CRISPR / Cas9-based system. The target regulatory element may be within the target region.

[0092] As used herein, the term "transcribed region" refers to a region of DNA that is transcribed into a single-stranded RNA molecule known as messenger RNA, which results in the transfer of genetic information from a DNA molecule to messenger RNA. During transcription, RNA polymerase reads the template strand in a 3' to 5' direction and synthesizes RNA from 5' to 3'. The mRNA sequence is complementary to the DNA strand.

[0093] "Transcription start site" or "TSS," as used interchangeably herein, refers to the first nucleotide of a transcribed DNA sequence at which RNA polymerase begins synthesis of an RNA transcript.

[0094] As used herein, "transgene" refers to a gene or genetic material that contains a gene sequence that is isolated from one organism and introduced into another organism. This non-native segment of DNA can retain the ability to produce RNA or protein in the transgenic organism, or it can change the normal function of the genetic code of the transgenic organism. The introduction of a transgene has the potential to change the phenotype of the organism.

[0095] As used herein, "tru gRNA" refers to a full-length guide RNA with nucleotides typically truncated by 2 nucleotides from its 5' end.

[0096] As used herein, "transregulatory element" refers to a region of non-coding DNA that regulates the transcription of a gene away from the gene it is transcribed from. Transregulatory elements can be on the same or different chromosome as the target gene. Examples of transregulatory elements include enhancers, super enhancers, silencers, insulators, and locus control regions.

[0097] "Variant," as used herein with respect to a nucleic acid, means (i) a portion or fragment of a reference nucleotide sequence (including a nucleotide sequence that has an insertion or deletion when compared to the reference nucleotide sequence); (ii) the complement of a reference nucleotide sequence, or a portion thereof; (iii) a nucleic acid that is substantially identical to a reference nucleic acid, or its complement; or (iv) a nucleic acid that hybridizes to a reference nucleic acid under stringent conditions, its complement, or a sequence substantially identical thereto.

[0098] A "variant" with respect to a peptide or polypeptide differs in amino acid sequence by insertion, deletion, or conservative substitution of amino acids, but retains at least one biological activity. A variant can also refer to a protein having an amino acid sequence that is substantially identical to a reference protein having an amino acid sequence that retains at least one biological activity. Conservative substitution of amino acids, i.e., replacing an amino acid with another amino acid of similar properties (e.g., hydrophilicity, degree, and distribution of charged regions), is recognized in the art as generally involving minor changes. These minor changes can be determined to some extent by considering the hydropathic index of an amino acid, as understood in the art. Kyte et al., J. Mol. Biol. 157:105-132 (1982). The hydropathic index of an amino acid is based on consideration of its hydrophobicity and charge. It is known in the art that amino acids of similar hydropathic index can be substituted and still retain protein function. In one embodiment, amino acids with hydropathic indices of ±2 are substituted. The hydrophilicity of an amino acid can also be used to identify substitutions that will result in a protein that retains biological function. Consideration of the hydrophilicity of an amino acid in the context of a peptide allows for the calculation of the maximum local average hydrophilicity of the peptide. Substitutions can be made with amino acids that have hydrophilicity values ​​within ±2 of each other. Both the hydrophobicity index and hydrophilicity value of an amino acid are influenced by the particular side chain of that amino acid. Consistent with this finding, it will be appreciated that amino acid substitutions that are compatible with biological function depend on the relative similarity of the amino acids, particularly the side chains of those amino acids, as revealed by hydrophobicity, hydrophilicity, charge, size, and other properties.

[0099] As used herein, "vector" refers to a nucleic acid sequence containing an origin of replication. The vector can be a viral vector, a bacteriophage, a bacterial artificial chromosome, or a yeast artificial chromosome. The vector can be a DNA or RNA vector. The vector can be a self-replicating extrachromosomal vector, preferably a DNA plasmid. For example, the vector can encode Cas9 and at least one optimized gRNA nucleotide sequence of any one of SEQ ID NOs: 149-315, 321-323, and 326-329.

[0100] Unless otherwise defined herein, scientific and technical terms used in connection with this disclosure shall have the same meaning as commonly understood by those skilled in the art. For example, all academic terms used in connection with cell and tissue culture, molecular biology, immunology, microbiology, genetics, and protein and nucleic acid chemistry, and hybridization described herein, and these techniques, are well known and commonly used in the art. The meaning and scope of terms should be clear; however, in case of any potential ambiguity, the definitions provided herein take precedence over any dictionary or external definitions. Furthermore, unless otherwise required by context, singular terms shall include the plural, and plural terms shall include the singular.

[0101] 2. CRISPR system The CRISPR system is a microbial nuclease system involved in defense against invading phages and plasmids, providing a form of adaptive immunity. CRISPR loci in microbial hosts can contain a combination of CRISPR-associated (Cas) genes and non-coding RNA elements that can program the specificity of CRISPR-mediated nucleic acid cleavage. A short piece of foreign DNA, called a spacer, is integrated between the CRISPR repeats in the genome and acts as a "memory" of past exposure. Cas9 forms a complex with the 3' end of a single guide RNA ("sgRNA"), and this protein-RNA pair recognizes its genomic target by complementary base pairing between the 5' end of the sgRNA sequence and a predefined 20 bp DNA sequence known as a protospacer. This complex is guided to the homologous locus of pathogen DNA, i.e., the protospacer and protospacer-adjacent motif (PAM) in the pathogen genome, via an encoded region in the CRISPR RNA ("crRNA"). The non-coding CRISPR array is transcribed and cleaved within the direct repeats to produce short crRNAs containing individual spacer sequences that guide the Cas nuclease to the target site (protospacer). By simply replacing the 20bp recognition sequence of the expressed chimeric sgRNA, the Cas9 nuclease can be directed to a new target on the genome. The CRISPR spacer is used to recognize and silence exogenous genetic elements in a manner similar to RNAi in eukaryotes.

[0102] Three classes of CRISPR systems are known: type I, type II, and type III effector systems. Type II effector systems use a single effector enzyme, Cas9, to perform targeted DNA double-strand breaks in four sequential steps to cleave dsDNA. Compared to type I and type III effector systems, which require multiple different effectors acting as a complex, type II effector systems can function in alternative contexts, such as eukaryotic cells. Type II effector systems consist of a long precursor-crRNA, which is transcribed from a spacer-containing CRISPR locus, Cas9 protein, and tracrRNA, which is involved in precursor-crRNA processing. The tracrRNA hybridizes to the repeat region that separates the spacer of the precursor-crRNA, thus initiating dsRNA cleavage by endogenous RNase III. This cleavage is followed by a second cleavage event within each spacer by Cas9, producing the tracrRNA and mature crRNA that remains associated with Cas9, forming the Cas9:crRNA-tracrRNA complex.

[0103] A genetically engineered form of the Streptococcus pyogenes type II effector system has been shown to function in human cells for genome engineering, in which the Cas9 protein is guided to a target site in the genome by a synthetically reconstituted "guide RNA" ("gRNA", also used interchangeably herein with chimeric sgRNA, which for Cas9 is generally a crRNA-tracrRNA fusion that obviates the need for RNase III and crRNA processing).

[0104] The Cas9:crRNA-tracrRNA complex unwinds the DNA duplex and searches for a sequence match with the crRNA to cleave it. Target recognition occurs upon detection of complementarity between the "protospacer" sequence in the target DNA and the remaining spacer sequence in the crRNA. Cas9 mediates cleavage of the target DNA if the correct protospacer adjacent motif (PAM) is also present at the 3' end of the protospacer. For protospacer targeting, the sequence must immediately follow the protospacer adjacent motif (PAM), a short sequence recognized by the Cas9 nuclease required for DNA cleavage. Another type II system has a different PAM requirement. The S. pyogenes CRISPR system can have a PAM sequence for this Cas9 (SpCas9) as 5'-NRG-3' (where R is either A or G) and characterizes the specificity of this system in human cells. The unique ability of CRISPR / Cas9-based systems is the straightforward ability to simultaneously target multiple different genomic loci by co-expressing a single Cas9 protein with two or more sgRNAs. For example, in genetically modified systems, the Streptococcus pyogenes type II system naturally prefers to use the "NGG" sequence (where "N" can be any nucleotide), but also accepts other PAM sequences such as "NAG" (Hsu et al. (2013) Nature Biotechnology, 31, 827-832). Similarly, the Cas9 from Neisseria meningitidis (NmCas9) usually has a natural PAM of NNNNGATT, but is active across a range of PAMs, including the highly degenerate NNNNGNNN PAM (Esvelt et al. Nature Methods (2013) doi:10.1038 / nmeth.2681).

[0105] 3. CRISPR / Cas9-based systems Provided herein is a CRISPR / Cas9 system that includes an optimized gRNA, such as a hairpin gRNA (also referred to herein as "hpgRNA" or "hp-gRNA"), that allows for improved DNA targeting for use in epigenome editing and transcriptional regulation, such as specifically cleaving a target region of interest, such as a target gene, or activating or suppressing gene expression of a target gene. The optimized gRNA provides increased target binding specificity while reducing the off-target binding and off-target activity of CRISPR / Cas9-based and CRISPR / Cpf1-based systems by adjusting the lifetime at off-target locations to minimize any activity at the off-target sites.

[0106] Optimized gRNAs can modulate Cas9-fusion protein activity by modulating Cas9 lifetime at these locations and modulating overall invasion kinetics, independent of second domain activity. In addition, gRNA binding to the protospacer at the 5' end of the protospacer-targeting segment may also participate in Cas9 cleavage. Reduced binding to off-target sites would limit the possibility of complete invasion / cleavage at these off-target sites. A genetically engineered form of the type II effector system of Streptococcus pyogenes has been shown to function in human cells for genome engineering. In this system, the Cas9 protein was guided to the target site in the genome by a synthetically reconstituted "guide RNA" ("gRNA", also used interchangeably herein as chimeric single-stranded guide RNA ("sgRNA"), which for Cas9 is generally a crRNA-tracrRNA fusion that obviates the need for RNase III and crRNA processing). Provided herein is a CRISPR / Cas9-based system for use in genome editing and the treatment of genetic diseases.CRISPR / Cas9-based systems can be designed to target any gene, including genes involved in genetic diseases, aging, tissue regeneration, or wound healing.CRISPR / Cas9-based systems can include Cas9 protein or Cas9 fusion protein and at least one optimized gRNA as described below.Cas9 fusion protein can include a domain that has an activity different from that endogenous to Cas9, such as a transactivation domain.

[0107] The target gene can have a mutation, such as a frameshift or nonsense mutation. If the target gene has a mutation that results in a premature stop codon, an ectopic splice acceptor, or an ectopic splice donor, the CRISPR / Cas9-based system can be designed to recognize and bind to the nucleotide sequence upstream or downstream of these premature stop codons, ectopic splice acceptors, or ectopic splice donors. The CRISPR-Cas9-based system can also be used to disrupt normal gene splicing by targeting splice acceptors and donors to induce skipping of premature stop codons or repair broken reading frames. The CRISPR / Cas9-based system may or may not mediate off-target changes to protein-coding regions of the genome.

[0108] i.Cas9 CRISPR / Cas9-based systems can include Cas9 protein or Cas9 fusion protein. Cas9 protein is an endonuclease that cleaves nucleic acid, and is encoded by CRISPR locus and is included in type II CRISPR system. Cas9 protein can be from any bacterial or archaeal species, such as Streptococcus pyogenes. Cas9 protein can be mutated to inactivate nuclease activity. Inactivated Cas9 protein from Streptococcus pyogenes (iCas9, also called "dCas9"), which does not have endonuclease activity, has recently been targeted to genes in bacteria, yeast, and human cells by gRNA to suppress gene expression through steric hindrance. As used herein, both "iCas9" and "dCas9" refer to Cas9 protein with amino acid substitutions D10A and H840A and inactivated its nuclease activity. In some embodiments, an inactivated Cas9 protein from Neisseria meningitides, such as NmCas9, can be used. For example, a CRISPR / Cas9-based system can include iCas9 of SEQ ID NO:1.

[0109] ii. Cas9 fusion proteins The CRISPR / Cas9-based system can include a fusion protein of a Cas9 protein that does not have nuclease activity, such as dCas9, and a second domain. The second domain can include a transcription activation domain, such as a VP64 domain or a p300 domain, a transcription repression domain, such as a KRAB domain, a nuclease domain, a transcription terminator domain, a histone modification domain, a nucleic acid association domain, an acetylase domain, a deacetylase domain, a methylase domain, such as a DNA methylase domain, a demethylase domain, a phosphorylation domain, a ubiquitination domain, or a sumoylation domain. The second domain can be a modifier of DNA methylation or chromatin looping.

[0110] In some embodiments, the fusion protein can include a dCas9 domain and a transcriptional activator. For example, the fusion protein can include the amino acid sequence of SEQ ID NO: 2. In other embodiments, the fusion protein can include a dCas9 domain and a transcriptional repressor. For example, the fusion protein can include the amino acid sequence of SEQ ID NO: 3. In further aspects, the fusion protein can include a dCas9 domain and a site-specific nuclease that is distinct from the Cas9 nuclease activity.

[0111] The fusion protein can comprise two heterologous polypeptide domains, where the first polypeptide domain comprises a Cas protein and the second polypeptide domain does not have nuclease activity. The fusion protein can comprise a Cas9 protein or a mutated Cas9 protein fused with a second polypeptide domain having nuclease activity as described above. This second polypeptide domain can have a nuclease activity different from the nuclease activity of the Cas9 protein. Nucleases, or proteins with nuclease activity, are enzymes capable of cleaving phosphodiester bonds between nucleotide subunits of nucleic acids. Nucleases are usually further classified into endonucleases and exonucleases, although some enzymes can belong to both categories. Well-known nucleases are deoxyribonucleases and ribonucleases.

[0112] (1) CRISPR / Cas9-based gene activation system The CRISPR / Cas9-based system can be a CRISPR / Cas9-based gene activation system that can activate regulatory element functions using the exceptional specificity of epigenetic editing. The CRISPR / Cas9-based gene activation system can be used to screen for enhancers, insulators, silencers, and locus control regions that can be targeted to increase or decrease target gene expression. This technology can be used to assign functions to putative regulatory elements identified through genomic studies such as the ENCODE and Roadmap Epigenomics projects.

[0113] CRISPR / Cas9-based gene activation systems can activate gene expression by modifying DNA methylation, forming chromatin loops, or catalyzing the acetylation of histone H3 lysine 27 at their target sites, resulting in robust transcriptional activation of target genes from promoters and proximal and distal enhancers. CRISPR / Cas9-based gene activation systems are highly specific and can be directed to target genes using as little as one guide RNA. CRISPR / Cas9-based gene activation systems can activate the expression of a gene or a family of genes by targeting enhancers at distant locations in the genome.

[0114] (a) Histone acetyltransferase (HAT) proteins The CRISPR / Cas9-based gene activation system can include histone acetylase proteins, such as p300 protein, CREB binding protein (CBP; an analog of p300), GCN5, or PCAF, or fragments thereof. Acetylation of histones at regulatory elements using programmable CRISPR / Cas9-based fusion proteins is an effective strategy to increase expression of target genes. CRISPR / Cas9-based histone acetylases, which can be targeted to any site in the genome, have the unique ability to activate distal regulatory elements. The histone acetylase protein can include human p300 protein or a fragment thereof. The histone acetylase protein can include wild-type human p300 protein or a mutant human p300 protein, or a fragment thereof. The histone acetylase protein can include the core lysine-acetyltransferase domain of human p300 protein, i.e., p300 HAT Core (also known as "p300 Core").

[0115] (b) CRISPR / dCas9 p300 Core activation system The p300 protein regulates the activity of many genes in tissues throughout the body. It regulates cell growth and division, promotes cell maturation and specialized functions (differentiation), and plays a role in preventing the growth of cancerous tumors. The p300 protein can activate transcription by linking transcription factors to the transcriptional protein complex in the nucleus of the cell. The p300 protein also functions as a histone acetylase that regulates transcription through chromatin remodeling.

[0116] dCas9 p300 Core Fusion proteins are a powerful and easily programmable means to synthetically engineer acetylation at targeted endogenous loci, resulting in regulation of genes regulated by proximal and distal enhancers. p300 Core can acetylate lysine 27 on histone H3 (H3K27ac) and provide H3K27ac enrichment. Fusion of the catalytic core domain of p300 with dCas9 can result in substantially higher transactivation of downstream genes than direct fusion of the full-length p300 protein, despite robust protein expression. dCas9 p300 Core Fusion proteins can also be used, for example, in the context of the Nm-dCas9 scaffold, particularly in the distal enhancer region (where dCas9 VP64 exhibited little, if any, measurable downstream transcriptional activity), dCas9 VP64 In addition, dCas9 may exhibit increased transactivation ability compared to p300 Core Exhibits precise and robust genome-wide transcriptional specificity. p300 Core may be capable of strong transcriptional activation and simultaneous enrichment of acetylation at promoters targeted by epigenetically modified enhancers.

[0117] dCas9 p300 Corecan activate gene expression through a single gRNA that targets and binds to promoters and / or characterized enhancers. This technology also provides the ability to artificially transactivate genes distal to putative and known regulatory regions, facilitating transactivation through the application of a single programmable effector and a single target site. These capabilities allow multiplexing to target several promoters and / or enhancers simultaneously. The mammalian origin of p300 may offer advantages over viral-derived effector domains for in vivo applications by minimizing the potential for immunogenicity.

[0118] dCas9 p300 Core Gene activation by p300 Core is highly specific to the target gene. In some embodiments, p300 Core comprises amino acids 1048 to 1664 of SEQ ID NO:2 (i.e., SEQ ID NO:4). In some embodiments, the CRISPR / Cas9-based gene activation system comprises dCas9 of SEQ ID NO:2. p300 Core Fusion protein or Nm-dCas9 of SEQ ID NO:5 p300 Core Includes fusion proteins.

[0119] (2) CRISPR / Cas9-based gene silencing system The CRISPR / Cas9-based system can be a CRISPR / Cas9-based gene suppression system that can use the exceptional specificity of epigenome editing to inhibit regulatory element function. In some embodiments, the CRISPR / Cas9-based gene suppression system, such as that comprising dCas9KRAB, can inhibit distal enhancer activity by highly specific remodeling of the epigenetic state of the targeted locus.

[0120] (a) CRISPR / dCas9 KRAB Gene Suppression Systems dCas9 KRABRepressors are highly specific epigenetic editing tools that can be used in loss-of-function screens to study gene function and to discover targets for drug development. KRAB dCas9- has exceptional specificity, targeting a specific enhancer and silencing only the gene targeted by that enhancer, creating a repressive heterochromatin environment at that site. KRAB can be used to screen for novel regulatory elements within the endogenous genomic context by silencing proximal or distal regulatory elements and corresponding gene targets. The specificity of the dCas9-KRAB repressor allows it to be used for transcriptome-wide specificity to silence endogenous genes. Epigenetic mechanisms for disruption at targeted loci include histone methylation.

[0121] The KRAB domain, a typical heterochromatin-forming motif within naturally occurring zinc finger transcription factors, is genetically associated with dCas9 and is the basis for the RNA-guided synthetic repressor, dCas9. KRAB Kruppel-associated box ("KRAB") recruits heterochromatin-forming factors: Kap1, HP1, SETDB1, and NuRD. It induces H3K0 trimethylation, histone deacetylation. KRAB-based synthetic repressors can efficiently silence the expression of single genes and have been used to silence oncogenes, inhibit viral replication, and treat dominant-negative diseases.

[0122] 4. CRISPR / Cpf1-based system The disclosed optimized gRNAs can be used with the "Clustered Regularly Interspaced Short Palindromic Repeats" from Prevotella and Francisella 1 or ("CRISPR / Cpf1") systems. The CRISPR / Cpf1 system, a DNA editing technique similar to the CRISPR / Cas9 system, is found in Prevotella and Francisella bacteria and prevents virus-derived genetic damage. Cpf1 is a class II CRISPR / Cas system RNA-guided endonuclease that contains a 1,300 amino acid protein. The Cpf1 gene is associated with the CRISPR locus, encoding an endonuclease that uses guide RNA to find and cut viral DNA. Because functional Cpf1 does not require tracrRNA, only crRNA, Cpf1 is a smaller and simpler endonuclease than Cas9 and has a smaller sgRNA molecule (almost half the number of nucleotides of Cas9). Examples of Cpf1 that can be used with optimized gRNAs include Cpf1 from Acidaminococcus and Lachnospiraceae bacteria.

[0123] The Cpf1 locus encodes Cas1, Cas2, and Cas4 proteins, which are more similar to type I and III than to type II systems. The Cpf1 locus contains a mixed α / β domain, RuvC-I, followed by a helical region, RuvC-II, and a zinc finger-like domain. The Cpf1 protein has a RuvC-like endonuclease domain that is similar to the RuvC domain of Cas9. Cpf1 does not have the HNH endonuclease domain, and the N-terminus of Cpf1 does not have the α-helical recognition lobe of Cas9. The Cpf1 CRISPR-Cas domain structure indicates that Cpf1 is functionally unique and classified as a class 2, type V CRISPR system.

[0124] The CRISPR / Cpf1 system consists of the Cpf1 enzyme and a guide RNA that finds the correct location on the double helix and positions the complex there to cleave the target DNA. The CRISPR / Cpf1 system activity has three phases: adaptation, crRNA formation, and interference. During the adaptation phase, Cas1 and Cas2 proteins facilitate the adaptation of small pieces of DNA to the CRISPR array. The crRNA formation phase involves processing of pre-cr-RNA to produce mature crRNA to guide Cas proteins. In the interference phase, Cpf1 is bound to the crRNA to form a binary complex to identify and cleave the target DNA sequence.

[0125] Whereas G-rich PAMs are targeted by Cas9, Cpf1-crRNA complexes cleave target DNA or RNA by specifying the protospacer adjacent motif 5'-YTN-3' (where "Y" is pyrimidine and "N" is any nucleobase) or 5'-TTN-3'. Whereas PAMs targeted by Cas9 are at the 3' side of the guide RNA, PAMs targeted by Cpf1 are at the 5' side of the guide RNA. After PAM specification, in contrast to the blunt end cleavage of Cas9, Cpf1 introduces sticky end-like DNA double strand breaks with 4 or 5 nucleotide overhangs, thereby enhancing the efficiency of gene insertion and specificity during NHEJ or HDR. Because the human genome is more T-rich than G-rich, TTN PAM sites are more useful for human genome engineering than GGN PAM sites. The protospacer-targeting segment of the gRNA for Cpf1 is at its 3'-most end, whereas the Cas9 gRNA is at its 5'-most end.

[0126] 5. gRNA A CRISPR / Cas9-based system or a CRISPR / Cpf1-based system can include at least one gRNA, such as an optimized gRNA as described herein, that targets a nucleic acid sequence. The gRNA provides specific targeting of the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system to a target region or gene. For a CRISPR / Cas9-based system, the gRNA is a fusion of two non-coding RNAs: crRNA and tracrRNA. The gRNA or sgRNA can target any desired DNA sequence by exchanging a sequence that encodes a 20 bp protospacer that confers targeting specificity through complementary base pairing with the desired DNA target. The gRNA mimics the naturally occurring crRNA:tracrRNA duplex contained in a type II effector system. This duplex can include, for example, a 42 nucleotide crRNA and a 75 nucleotide tracrRNA, which acts as a guide for Cas9. The gRNA can target and bind to a target region of a target gene. For a CRISPR / Cpf1 based system, the gRNA is a crRNA.

[0127] A CRISPR / Cas9-based system or a CRISPR / Cpf1-based system can include at least one gRNA, such as an optimized gRNA described herein, where the gRNAs target different DNA sequences. The target DNA sequences can be overlapping. The target sequence or protospacer is followed by a PAM sequence at the 3' end of the protospacer. Different type II systems have different PAM requirements. For example, the Streptococcus pyogenes type II system uses the "NGG" sequence, where "N" can be any nucleotide.

[0128] 6. Methods for Producing Optimized Guide RNAs (gRNAs) The present disclosure is directed to a method for producing an optimized gRNA, such as a hairpin gRNA (also referred to herein as "hpgRNA" or "hp-gRNA"). The optimized gRNA comprises the nucleotide sequence of a full-length gRNA and nucleotides added to the 5' or 3' end of the full-length gRNA. In some embodiments, the full-length gRNA can be designed using a program such as SgRNA designer, CRISPR MultiTargeter, or SSFinder. The nucleotides added to the 5' end of the full-length gRNA for CRISPR / Cas9 systems or the 3' end for CRISPR / Cpf1 systems can form a secondary structure by hybridizing or partially hybridizing with nucleotides in the protospacer-targeting sequence of the full-length gRNA. This secondary structure regulates DNA binding or cleavage by impeding the invasion of the DNA duplex by the gRNA. This secondary structure affects the invasion kinetics of the gRNA but not the binding energy of the gRNA with a complementary DNA strand. As described in the Examples below, the guide RNA of the Type II CRISPR-Cas system binds to the protospacer through a Cas9-facilitated process known as "strand invasion," in which the Cas9 protein itself first binds and melts the protospacer adjacent motif (PAM) via direct interaction, and then base-pairs with the 3'-end of the gRNA, the PAM-adjacent nucleotide (the "seed" region), proceeding nucleotide-by-nucleotide from the 3' of the gRNA to the 5'-end base-pairing with the protospacer. A similar mechanism is used in the CRISPR / Cpf1 system.

[0129] The nucleotides added to the 5' or 3' end of the full-length gRNA are not simply added to hybridize with the protospacer-targeting segment of the guide RNA (hairpin) to prevent access to the protospacer at thermodynamic equilibrium.As described in the Examples, equilibrium thermodynamic secondary structure properties (such as the melting temperature of the gRNA secondary structure) do not all correlate with the specificity of the guide RNA.Rather, in the case of cleavage, and in subsequent computational studies of Cas9 binding (as measured through ChIP-Seq in cells (see doi:10.1038 / nbt.2916; doi:10.1038 / nbt.2889)), there is a significant and substantial correlation between these and the estimated strand invasion kinetics and the structure, design, and function of the guide RNA that regulates strand invasion into the protospacer, which is necessarily different from the hairpin designed to thermodynamically compete for binding at equilibrium with on- and off-target sites. For example, secondary structure elements designed to be stable at equilibrium (such as RNAs that form hairpin-like structures containing an internal rG-rU wobble pair in the stem) may be rapidly destabilized during strand invasion (e.g., because the rG-rU wobble pair is the terminal base pair of the stem, it incurs significant energy loss on the RNA secondary structure as the adjacent nucleotide invades the protospacer, and also regulates strand invasion and binding kinetics by a mechanism entirely separate from simply blocking access to the protospacer at thermodynamic equilibrium). Secondary structures that are stable at equilibrium but rapidly destabilized during strand invasion can be designed using the methods described herein in a manner that distinguishes on-target and off-target sites with a minimal thermodynamic energy difference between sites (e.g., the result of a single internal mismatch), which cannot be substantially distinguished by cis-blocking or thermodynamic competition. If the invasion of an on-target site destabilizes a hairpin containing a GU wobble pair, this site is kinetically distinguished by invasion. For example, the VEGFA1 site described in the Examples below (target site is GGGTGGGGGGAGTTTGCTCC and off-target site 2 is [ka] (mismatches are underlined) were able to reduce off-target cleavage by 93% and 98%, respectively, compared to standard or full-length guide RNAs or truncated guide RNAs, using computationally designed secondary structures that account for strand invasion.

[0130] Furthermore, to enhance the initiation of strand invasion by the guide RNA, nucleotides can be added to the 5' or 3' end of the full-length gRNA to disrupt the "naturally occurring" secondary structures on the protospacer-targeting segment of the gRNA in the "seed" region. Thus, the addition of these nucleotides represents a different class of guide RNA modification, forming secondary structures that alter strand invasion by partially hybridizing with nucleotides in the protospacer-targeting sequence to modulate DNA binding or cleavage.

[0131] An optimized gRNA is designed to minimize binding at off-target sites and to bind to a protospacer sequence. In some embodiments, the off-target site is a known or predicted off-target site. In some embodiments, the method includes the steps of: identifying a target region of interest, the target region of interest comprising a protospacer sequence; determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; determining at least one or more off-target sites for the full-length gRNA; generating a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment being at the 5' end of the polynucleotide sequence of the full-length gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment being at the 5' end of the polynucleotide sequence of the full-length gRNA, and the RNA segment, wherein the linker comprises a polynucleotide sequence having a length of N nucleotides, wherein the first gRNA is capable of invading the protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading the off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; calculating an estimate or computationally simulating the invasion kinetics and the lifetime that the first gRNA remains invaded within the protospacer and off-target site duplex, wherein the invasion kinetics is estimated nucleotide by nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to the DNA sequence that is complementary to the protospacer sequence;comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site with the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site; randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with the second gRNA; identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and testing the optimized gRNA in vivo to determine binding specificity.

[0132] In some embodiments, the method includes the steps of: identifying a target region of interest, the target region of interest comprising a protospacer sequence; determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; determining at least one or more off-target sites for the full-length gRNA; generating a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment being a 3' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprising a linker between the 3' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker having a length of N nucleotides. a polynucleotide sequence that represents a first gRNA capable of invading a protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and a first gRNA capable of invading an off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; calculating or computationally simulating an estimate of the invasion kinetics and the lifetime that the first gRNA remains invaded within the protospacer and off-target site duplex, where the invasion kinetics is estimated on a nucleotide-by-nucleotide basis by determining the energy difference between further invasion of a different gRNA and re-annealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence; comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site to the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site;randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with the second gRNA; identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and testing the optimized gRNA in vivo to determine specificity of binding.

[0133] In some embodiments, the energetics of further invasion of a different gRNA are determined by determining at least one of the energetics of (I) breaking DNA-DNA base pairing, (II) forming RNA-DNA base pairing, (III) the energy difference resulting from breaking or forming a different secondary structure in the uninvaded guide RNA, and (IV) forming or breaking an interaction between the displaced DNA strand that is complementary to the protospacer and any unpaired guide RNA nucleotides that are not included in the secondary structure. In some embodiments, the energetics of reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence are determined by determining at least one of the energetics of (I) forming DNA-DNA base pairing, (II) breaking RNA-DNA base pairing, (III) the energy difference resulting from breaking or forming a different secondary structure in the newly uninvaded guide RNA, and (IV) forming or breaking an interaction between the displaced DNA strand that is complementary to the protospacer and any unpaired guide RNA nucleotides that are not included in the secondary structure. In some embodiments, the method further comprises determining energetic considerations from at least one of: (V) base pairing across the mismatch; (VI) interaction with the Cas9 protein; and / or (VII) additional heuristics (wherein the additional heuristics relate to binding lifetime, degree of invasion, stability of the invading guide RNA, or other calculated / simulated properties of gRNA invasion versus Cas9 cleavage activity).

[0134] The CRISPR / Cas9-based or CRISPR / Cpf1-based systems can use gRNAs of various sequences and lengths, such as the optimized gRNAs described herein. In some embodiments, the full-length gRNA can include a protospacer-targeting segment that corresponds to the polynucleotide sequence (i.e., the protospacer) of the target DNA sequence. In some embodiments, the protospacer-targeting segment can have at least 10 nucleotides, at least 11 nucleotides, at least 12 nucleotides, at least 13 nucleotides, at least 14 nucleotides, at least 15 nucleotides, at least 16 nucleotides, at least 17 nucleotides, at least 18 nucleotides, at least 19 nucleotides, at least 20 nucleotides, at least 21 nucleotides, at least 22 nucleotides, at least 23 nucleotides, at least 24 nucleotides, at least 25 nucleotides, at least 30 nucleotides, or at least 35 nucleotides. The gRNA can target at least one of the following regions of the target gene: a promoter region, an enhancer region, a repressor region, an insulator region, a silencer region, a region involved in DNA looping with the promoter region, a gene splicing region, or a transcription region. In some embodiments, the full-length gRNA comprises a protospacer-targeting segment having about 15 to 20 nucleotides.

[0135] In some embodiments, the RNA segment comprises 2 to 20 nucleotides, 3 to 10 nucleotides, or 5 to 8 nucleotides. In some embodiments, the RNA segment comprises 2 to 20 nucleotides, 3 to 10 nucleotides, or 5 to 8 nucleotides that are complementary to the protospacer-targeting sequence. In some embodiments, M is 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6 , 2 to 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to M can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. For example, M can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20.In some embodiments, the RNA segments are selected from the group consisting of 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 16, 2 to 17, 2 to 18, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 15, 2 to 16, 2 to 17, 2 to 18, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 18, 2 to 19, 2 to 19, 2 to 20 ... 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, The RNA segment can have 7 to 16, 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10 nucleotides. In some embodiments, the RNA segment can have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 nucleotides.

[0136] In some embodiments, N is 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6 , 2 to 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to N can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. For example, N can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20. In some embodiments, the linker comprises 1 to 20 nucleotides, 3 to 10 nucleotides, or 5 to 8 nucleotides.For example, the linkers may be any of the following, some or all of which are complementary to the protospacer-targeting sequence: 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 16, 2 to 18, 2 to 19, 2 to 20 ... 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7 , 4 to 6, 4 to 5, 5 to 20, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to The linker may have 16, 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10 nucleotides. In some embodiments, the linker may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 nucleotides. In some embodiments, the linker may include a stabilizing linker, such as a tetraloop.Examples of tetraloops include, but are not limited to, ANYA, CUYG, GNRA, UMAC, and UNCG.

[0137] In some embodiments, the RNA segment and / or the protospacer-targeting sequence provide a secondary structure. In some embodiments, the secondary structure is formed by partially hybridizing the protospacer-targeting sequence with the RNA segment. In some embodiments, the secondary structure regulates DNA binding or cleavage by Cas9 by preventing invasion of the protospacer duplex or off-target duplex by the optimized gRNA. In some embodiments, the secondary structure stably maintains the 5' end of the gRNA within the protein and protects the optimized gRNA within Cas9 from degradation.

[0138] In some embodiments, the secondary structure is formed by hybridizing all or a portion of the RNA segment to the 5'-terminal nucleotide of the protospacer-targeting sequence or segment, the central nucleotide of the protospacer-targeting sequence or segment, and / or the 3'-terminal nucleotide of the protospacer-targeting sequence or segment. In some embodiments, a contiguous segment of the RNA segment hybridizes to the protospacer-targeting sequence or segment. In some embodiments, a non-contiguous segment of the RNA segment hybridizes to the protospacer-targeting sequence or segment. In some embodiments, the secondary structure is a hairpin.

[0139] In some embodiments, the secondary structure is stable at room temperature or 37° C. In some embodiments, the total equilibrium free energy of the secondary structure is less than about 2 kcal / mol at a temperature of about 4° C. to about 50° C., e.g., at room temperature or 37° C. For example, the total equilibrium free energy of the secondary structure is less than about 2 kcal / mol at a temperature of about 4° C. to about 50° C., about 4° C. to about 40° C., about 4° C. to about 37° C., about 4° C. to about 30° C., about 4° C. to about 25° C., about 4° C. to about 20° C., about 4° C. to about 10° C., about 5° C. to about 50° C., about 5° C. to about 40° C., about 5° C. to about 37° C., about 5° C. to about 30° C., about 5° C. to about 25° C., about 5° C. to about 20° C., about 5° C. to about 10° C., about 10° C. to about 50° C., about 10° C. to about 40° C., about 10° C. to about 37° C., about 10° C. to about 30° C., about 10 ... It may be less than about 10 kcal / mol, less than about 5 kcal / mol, less than about 4 kcal / mol, less than about 3 kcal / mol, less than about 2 kcal / mol, less than about 1 kcal / mol, or less than about 0.5 kcal / mol at a temperature of about 25° C., about 10° C. to about 20° C., about 20° C. to about 50° C., about 20° C. to about 40° C., about 20° C. to about 37° C., about 20° C. to about 30° C., about 25° C. to about 50° C., about 25° C. to about 40° C., about 25° C. to about 37° C., or about 25° C. to about 30° C. In some embodiments, the RNA segment hybridizes to or forms a non-canonical base pair with at least two nucleotides of the protospacer-targeting sequence or segment. In some embodiments, the non-canonical base pair is rU-rG.

[0140] In some embodiments, 1 to 20 nucleotides are randomized within the linker, e.g., 1 to 20, 1 to 15, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 1 to 4, 1 to 3, 1 to 2, 2 to 20, 2 to 15, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 5, 2 to 4, 3 to 20, 3 to 15, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 3 to 4, 4 to 20, 4 to 15, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6 , 4 to 5, 5 to 20, 5 to 15, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 15, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 15, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 15, 8 to 10, 8 to 9, 9 to 20, 9 to 15, or 9 to 10, 10 to 20, 10 to 15, or 15 to 20 nucleotides can be randomized within the linker.

[0141] In some embodiments, 1 to 20 nucleotides are randomized within the RNA segment, e.g., 1 to 20, 1 to 15, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 1 to 4, 1 to 3, 1 to 2, 2 to 20, 2 to 15, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 5, 2 to 4, 3 to 20, 3 to 15, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 3 to 4, 4 to 20, 4 to 15, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 From 5, 5 to 20, 5 to 15, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 15, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 15, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 15, 8 to 10, 8 to 9, 9 to 20, 9 to 15, or 9 to 10, 10 to 20, 10 to 15, or 15 to 20 nucleotides can be randomized within the RNA segment.

[0142] In some embodiments, step (g) is repeated X times, thereby producing X gRNAs, and step (e) is repeated with each of the X gRNAs, where X is 0 to 20. In some embodiments, X is 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6 , 2 to 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 2 0, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to 16 , 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. For example, X can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20.

[0143] In some embodiments, the invasion kinetics and lifetime are calculated using the kinetic Monte Carlo method or the Gillespie algorithm. In some embodiments, the invasion kinetics and lifetime can be determined using "deterministic" methods, such as differential equations that model strand invasion, known to those skilled in the art. The kinetic Monte Carlo (KMC) method is a Monte Carlo computer simulation that aims to simulate the time evolution of some process that occurs in nature. This process is generally a process that occurs with a known transition rate between states. This known transition rate is the input to the KMC algorithm. The Gillespie algorithm (also known as the Doob-Gillespie algorithm) produces statistically plausible trajectories (possible solutions) of probability equations. The Gillespie algorithm can be used to simulate increasingly complex systems. This algorithm is particularly useful for simulating reactions in cells, where the number of reactants is generally several tens of molecules (or less). Mathematically, it is a variation of the kinetic Monte Carlo method and is similar to the kinetic Monte Carlo method. The Gillespie algorithm allows for discrete and stochastic simulation of systems with small numbers of reactants, since every reaction is explicitly simulated. The trajectories corresponding to a single Gillespie simulation represent exact samples from the probability mass function that is the solution of the master equation.

[0144] In some embodiments, the design criteria can be specificity, binding lifetime regulation, and / or predicted cleavage specificity. For example, an optimized gRNA can be designed to have a binding lifetime equal to or greater than that of the full gRNA at the on-target site, and / or a binding lifetime equal to or less than that of the full gRNA at the off-target site. In some embodiments, an optimized gRNA is selected to have a binding lifetime equal to or less than that of the full-length gRNA to at least three off-target sites, where these off-target sites are predicted to be the closest off-target sites or have the highest identity to the on-target site. In some embodiments, the design criteria include a lifetime or cleavage rate at the off-target site that is equal to or less than that of the full-length or truncated gRNA at the off-target site, and / or a predicted on-target activity rate that is greater than 10% of the predicted on-target activity rate of the full-length or truncated gRNA.

[0145] In some embodiments, the optimized gRNA is tested in step i) using a mismatch-sensitive nuclease to determine CRISPR activity, such as surveyor assay or T7 endonuclease I (T7E1) assay, or next-generation sequencing technology, such as Illumina MiSeq or GUIDE-Seq. In some embodiments, the optimized gRNA is tested in step i) using a reporter assay, in which Cas9-fusion protein activity alters the expression of a reporter protein, such as GFP. GUIDE-Seq is an assay designed to analyze off-target cleavage.

[0146] In some embodiments, target regions can be determined based on sequence proximity to the PAM sequence using programs such as CRISPR design (Ran, et al. Nature Protocols (2013) 8:2281-2308) and CCTop (Stemmer, PLoS One (2015) 10:e0124633) tools. In some embodiments, target sites can include promoters, DNAse I hypersensitive sites, Transposase-Accessible chromatin sites, DNA methylation sites, transcription factor binding sites, epigenetic marks, expression quantitative trait loci, and / or regions associated with human traits or phenotypes in genetic association studies. Target sites were identified using a variety of techniques including DNase-sequencing (DNase-seq), Assay for Transposase-Accessible Chromatin using high-throughput sequencing (ATAC-seq), ChIP-sequencing, self-transcribing active regulatory region sequencing (STARR-Seq), single molecule real time sequencing (SMRT), Formaldehyde-Assisted Isolation of Regulatory Elements sequencing (FAIRE-seq), micrococcal nuclease sequencing (MNase-seq), reduced representation bisulfite sequencing (RRBS-seq), whole genome bisulfite sequencing (WBS-seq), and genomic DNA sequencing (GDNA-seq). This can be determined by methyl-DNA immunoprecipitation (MDI-seq), methyl-linked DNA immunoprecipitation (MEDIP-seq), or gene association studies.In some embodiments, off-target sites can be determined using CasOT (PKU Zebrafish Functional Genomics group, Peking University), CHOPCHOP (Harvard University), CRISPR Design, (Massachusetts Institute of Technology), CRISPR Design tool (The Broad Institute of Harvard and MIT), CRISPR / Cas9 gRNA finder (University of Colorado), CRISPRfinder (Universite Paris-Sud), E-CRISP (DKFZ German Cancer Research Center), CRISPR gRNA Design tool (DNA 2.0), PROGNOS (Emory University / Georgia Institute of Technology), ZiFiT (Massachusetts General Hospital). Examples of tools that can be used to determine target regions and off-target sites are described in International Publication No. WO2016109255, the entire contents of which are incorporated herein by reference.

[0147] 7. Target Gene As disclosed herein, the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system can be designed to target and cleave the expression of any target gene. For example, a gRNA, such as the optimized gRNA described herein, can target and bind to a target region within a target gene. The target gene can be an endogenous gene, a transgene, or a viral gene in a cell line. In some embodiments, the target gene can be a known gene. In some embodiments, the target gene is an unknown gene. The gRNA can target any nucleic acid sequence. The nucleic acid sequence target can be DNA. The DNA can be any gene. For example, the gRNA can target a gene such as DMD, EMX1, or VEGFA.

[0148] In some aspects, the target gene is a disease-associated gene. In some embodiments, the target cell is a mammalian cell. In some embodiments, the genome comprises a human genome. In some embodiments, the target gene can be a prokaryotic gene or a eukaryotic gene, such as a mammalian gene. For example, the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system can target mammalian genes, such as DMD (dystrophin gene), EMX1, VEGFA, IL1RN, MYOD1, OCT4, HBE, HBG, HBD, HBB, MYOCD (Myocardin), PAX7 (Paired box protein Pax-7), FGF1 (fibroblast growth factor-1) genes, such as FGF1A, FGF1B, and FGF1C. Other target genes include, but are not limited to, Atf3, Axud1, Btg2, c-Fos, c-Jun, Cxcl1, Cxcl2, Edn1, Ereg, Fos, Gadd45b, Ier2, Ier3, Ifrd1, Il1b, Il6, Irf1, Junb, Lif, Nfkbia, Nfkbiz, Ptgs2, Slc25a25, Sqstm1, Tieg, Tnf, Tnfaip3, Zfp36, Birc2, Ccl2, Ccl20, Ccl7, Cebpd, Ch25h, CSF1, Cx3cl1, Cxcl10, Cxcl5, Gch, Icam1, Ifi47 , Ifngr2, Mmp10, Nfkbie, Npal1, p21, Relb, Ripp2, Rnd1, S1pr3, Stx11, Tgtp, Tlr2, Tmem140, Tnfaip2, Tnfrsf6, Vcam1, 1110004C05Rik (GenBank accession number BC010291), Abca1, AI561871 (GenBank accession number BI143915), AI882074 (GenBank accession number BB730912), Arts1, AW049765 (GenBank accession number BC026642).1), C3, Casp4, Ccl5, Ccl9, Cdsn, Enpp2, Gbp2, H2-D1, H2-K, H2-L, Ifit1, Ii, Il13ra1, Il1rl1, Lcn2, Lhfpl2, LOC677168 (GenBank Accession No. AK019325), Mmp13, Mmp3, Mt2, Naf1, Ppicap, Prnd, Psmb10, Saa3, Serpina3g, Serpinf1, Sod3, Stat1, Tapbp, U90926 (GenBank Accession No. NM_020562), Ubd, A2AR (Adenosine A2A Receptor), B7-H3 (also known as CD276), B7-H4 (also known as VTCN1), BTLA (B and T Lymphocyte Attenuator) Attenuator; also known as CD272), CTLA-4 (Cytotoxic T-Lymphocyte-Associated protein 4; also known as CD152), IDO (Indoleamine 2,3-dioxygenase), KIR (Killer-cell Immunoglobulin-like Receptor), LAG3 (Lymphocyte Activation Gene-3), PD-1 (Programmed Death 1 (PD-1) receptor), TIM-3 (T cell immunoglobulin domain and mucin domain 3), and VISTA (V-domain Ig suppressor of T cell activation. In some embodiments, the target gene is DMD (dystrophin), EMX1, or VEGFA gene.

[0149] 8. Compositions for genome editing The present invention is directed to a composition for genome editing, genome modification, or modification of gene expression of a target gene. The composition comprises an optimized gRNA produced by the disclosed method together with a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. In some embodiments, the gRNA can discriminate between on-target and off-target sites with the minimum thermodynamic energy difference between these sites, providing increased specificity. In some embodiments, the optimized gRNA regulates strand invasion into the protospacer.

[0150] Increased specificity is achieved by adding an extension, e.g., a protospacer-targeting sequence, to the 5' or 3' end of a full-length or standard gRNA that is self-complementary to the segment of the full-length or standard gRNA that targets the protospacer, such that it forms a "hairpin" structure. See Figures 1B and 2B. The hairpin acts as a kinetic barrier to strand invasion of the protospacer, but during strand invasion of the complete target site, the hairpin can be displaced and full invasion can occur.

[0151] As shown in Figure 2D, binding to the complete protospacer by dCas9 occurs preferentially, strongly suggesting that the hairpin is indeed displaced during invasion. The disclosed optimized gRNA, a hairpin, was designed to increase specificity in binding to targeted sites by inhibiting invasion when there is a mismatch between the target and the PAM-distal targeting region of the guide RNA. In this case, it is energetically more favorable for the hairpin to remain closed, and the presence of the hairpin likely promotes melting and detachment of Cas9 / dCas9 from these sites.

[0152] Optimized gRNAs (hpgRNAs) with 5'-hairpins or 3'-hairpins significantly enhanced the specificity of binding compared to both standard guide RNAs and the best available guide RNA variants (see Examples), eliminating or significantly weakening binding at protospacer sites containing mismatches. Increasing the length of the hairpin increased the specificity of dCas9 binding. Optimized gRNAs and hpgRNAs can be used to tune the affinity and specificity of Cas9 / dCas9 or Cpf1 binding. Based on the size and structure of the hairpin, the hairpin of the hpgRNA could fit into the DNA binding channel of the Cas9 / dCas9 molecule and protect it from degradation. In some embodiments, the hairpin length, loop length, and loop composition can be varied to allow finer control of these properties. In some embodiments, the hairpin length can be about 1 to about 20 nucleotides or about 3 to about 10 nucleotides.For example, the hairpin length may be 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20 , 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to 16 , 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. For example, the hairpin length can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20, or about 5 to about 8 nucleotides.

[0153] In some embodiments, the loop length can be from about 1 to about 20 nucleotides, from about 3 to about 10 nucleotides, or from about 5 to about 8 nucleotides. For example, the loop length can be 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 8, 2 to 9, 2 to 10, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 10, 2 to 12, 2 to 11, 2 to 10, 2 to 12, 2 to 13, 2 to 14, 2 to 15, 2 to 16, 2 to 17, 2 to 18, 2 to 19, 2 to 12, 2 to 11, 2 to 10, 2 to 12 ... et al. 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20, 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to 16, It can be 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. In some embodiments, the loop length can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20, or about 5 to about 8 nucleotides.

[0154] In some embodiments, the loop composition can be from about 1 to about 20 nucleotides, from about 3 to about 10 nucleotides, or from about 5 to about 8 nucleotides. For example, the loop composition can be 1 to 20, 1 to 19, 1 to 18, 1 to 17, 1 to 16, 1 to 15, 1 to 14, 1 to 13, 1 to 12, 1 to 11, 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 2 to 20, 2 to 19, 2 to 18, 2 to 17, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 8, 2 to 9, 2 to 10, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 10, 2 to 12, 2 to 11, 2 to 10, 2 to 12, 2 to 15, 2 to 16, 2 to 15, 2 to 14, 2 to 13, 2 to 12, 2 to 11, 2 to 10, 2 to 9, 2 to 8, 2 to 7, 2 to 6, 2 to 10, 2 to 12, 2 to 11, 2 to 10, 2 to 12, 2 to 15, 2 to 15, 2 to 16, 2 to 15, 2 to 1 5, 3 to 20, 3 to 19, 3 to 18, 3 to 17, 3 to 16, 3 to 15, 3 to 14, 3 to 13, 3 to 12, 3 to 11, 3 to 10, 3 to 9, 3 to 8, 3 to 7, 3 to 6, 3 to 5, 4 to 20, 4 to 19, 4 to 18, 4 to 17, 4 to 16, 4 to 15, 4 to 14, 4 to 13, 4 to 12, 4 to 11, 4 to 10, 4 to 9, 4 to 8, 4 to 7, 4 to 6, 4 to 5, 5 to 20 , 5 to 19, 5 to 18, 5 to 17, 5 to 16, 5 to 15, 5 to 14, 5 to 13, 5 to 12, 5 to 11, 5 to 10, 5 to 9, 5 to 8, 5 to 7, 5 to 6, 6 to 20, 6 to 19, 6 to 18, 6 to 17, 6 to 16, 6 to 15, 6 to 14, 6 to 13, 6 to 12, 6 to 11, 6 to 10, 6 to 9, 6 to 8, 6 to 7, 7 to 20, 7 to 19, 7 to 18, 7 to 17, 7 to 16 , 7 to 15, 7 to 14, 7 to 13, 7 to 12, 7 to 11, 7 to 10, 7 to 9, 7 to 8, 8 to 20, 8 to 19, 8 to 18, 8 to 17, 8 to 16, 8 to 15, 8 to 14, 8 to 13, 8 to 12, 8 to 11, 8 to 10, 8 to 9, 9 to 20, 9 to 19, 9 to 18, 9 to 17, 9 to 16, 9 to 15, 9 to 14, 9 to 13, 9 to 12, 9 to 11, or 9 to 10. In some embodiments, the loop composition can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20, or about 5 to about 8 nucleotides.

[0155] The composition can include a viral vector and a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system with at least one gRNA, such as an optimized gRNA described herein. In some embodiments, the composition includes a modified AAV vector and a nucleotide sequence encoding a CRISPR / Cas9-based system with at least one gRNA, such as an optimized gRNA described herein. The composition can further include donor DNA or a transgene. These compositions can be used in genome editing, genome engineering, and correcting or reducing the effects of genetic mutations involved in genetic diseases.

[0156] The target gene may be involved in cell differentiation or any other process in which gene activation, repression or destruction may be desired, or may have mutations such as deletions, frameshift mutations, or nonsense mutations. If the target gene has a mutation that causes a premature stop codon, an ectopic splice acceptor, or an ectopic splice donor, the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, can be designed to recognize and bind to the nucleotide sequence upstream or downstream of these premature stop codons, ectopic splice acceptors, or ectopic splice donors. The CRISPR / Cas9-based system or the CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, can also be used to disrupt normal gene splicing by targeting splice acceptors and donors to induce skipping of premature stop codons or repair broken reading frames. CRISPR / Cas9-based or CRISPR / Cpf1-based systems involving at least one gRNA, such as the optimized gRNAs described herein, may or may not mediate off-target changes to protein-coding regions of the genome.

[0157] In some embodiments, the CRISPR / Cas9 based system increases or decreases gene expression of a target gene by at least about 1 fold, at least about 2 fold, at least about 3 fold, at least about 4 fold, at least about 5 fold, at least about 6 fold, at least about 7 fold, at least about 8 fold, at least about 9 fold, at least about 10 fold, at least 15 fold, at least 20 fold, at least 30 fold, at least 40 fold, at least 50 fold, at least 60 fold, at least 70 fold, at least 80 fold, at least 90 fold, at least 100 fold, at least about 110 fold, at least 120 fold, at least 130 fold, at least 140 fold, at least 150 fold, at least 160 fold, at least 180 fold, at least 190 fold, at least 210 fold, at least 220 fold, at least 230 fold, at least 240 fold, at least 250 fold, at least 260 fold, at least 270 fold, at least 280 fold, at least 290 fold, at least 300 fold, at least 310 fold, at least 320 fold, at least 330 fold, at least 340 fold, at least 350 fold, at least 360 fold, at least 370 fold, at least 380 fold, at least 390 fold, at least 400 fold, at least 410 fold, at least 420 fold, at least 430 fold, at least 440 fold, at least 450 fold, at least 460 fold, at least 470 fold, at least 480 fold, Induce or suppress at least 160-fold, at least 170-fold, at least 180-fold, at least 190-fold, at least 200-fold, at least about 300-fold, at least 400-fold, at least 500-fold, at least 600-fold, at least 700-fold, at least 800-fold, at least 900-fold, at least 1000-fold, at least 1500-fold, at least 2000-fold, at least 2500-fold, at least 3000-fold, at least 3500-fold, at least 4000-fold, at least 4500-fold, at least 5000-fold, at least 600-fold, at least 7000-fold, at least 8000-fold, at least 9000-fold, at least 10000-fold, at least 100000-fold. The control level of gene expression of the target gene can be the level of gene expression of the target gene in a cell that is not treated with any CRISPR / Cas9-based system.

[0158] A. Modified lentiviral vector The composition for genome editing, genome modification, or modification of gene expression of target gene can include modified lentiviral vector. The modified lentiviral vector includes a first polynucleotide sequence that encodes a system and a second polynucleotide sequence that encodes at least one sgRNA. The first polynucleotide sequence can be operably linked to a promoter. The promoter can be a constitutive promoter, an inducible promoter, a repressible promoter, or a regulatable promoter.

[0159] The second polynucleotide sequence encodes at least one gRNA, such as an optimized gRNA described herein. For example, the second polynucleotide sequence can encode at least one gRNA, at least two gRNAs, at least three gRNAs, at least four gRNAs, at least five gRNAs, at least six gRNAs, at least seven gRNAs, at least eight gRNAs, at least nine gRNAs, at least ten gRNAs, at least eleven gRNAs, at least twelve gRNAs, at least thirteen gRNAs, at least fourteen gRNAs, at least fifteen gRNAs, at least sixteen gRNAs, at least seventeen gRNAs, at least eighteen gRNAs, at least nineteen gRNAs, at least twenty gRNAs, at least twenty-five gRNAs, at least thirty gRNAs, at least thirty-five gRNAs, at least forty gRNAs, at least forty-five gRNAs, or at least fifty gRNAs.The second polynucleotide sequence may be selected from 50 gRNAs from 1 gRNA, 45 gRNAs from 1 gRNA, 40 gRNAs from 1 gRNA, 35 gRNAs from 1 gRNA, 30 gRNAs from 1 gRNA, 25 different gRNAs from 1 gRNA, 20 gRNAs from 1 gRNA, 16 gRNAs from 1 gRNA, 8 different gRNAs from 1 gRNA, 50 different gRNAs from 4 different gRNAs, 45 different gRNAs from 4 different gRNAs, 40 different gRNAs from 4 different gRNAs, 35 different gRNAs from 4 different gRNAs, 30 different gRNAs from 4 different gRNAs, 25 different gRNAs from 4 different gRNAs, 20 different gRNAs from 4 different gRNAs, 16 different gRNAs from 4 different gRNAs, 8 different gRNAs from 4 different gRNAs. The polynucleotide sequences may encode different gRNAs, 8 different gRNAs to 50 different gRNAs, 8 different gRNAs to 45 different gRNAs, 8 different gRNAs to 40 different gRNAs, 8 different gRNAs to 35 different gRNAs, 8 different gRNAs to 30 different gRNAs, 8 different gRNAs to 25 different gRNAs, 8 different gRNAs to 20 different gRNAs, 8 different gRNAs to 16 different gRNAs, 16 different gRNAs to 50 different gRNAs, 16 different gRNAs to 45 different gRNAs, 16 different gRNAs to 40 different gRNAs, 16 different gRNAs to 35 different gRNAs, 16 different gRNAs to 30 different gRNAs, 16 different gRNAs to 25 different gRNAs, or 16 different gRNAs to 20 different gRNAs. Each of the polynucleotide sequences encoding the different gRNAs may be operably linked to a promoter. The promoters operably linked to the different gRNAs may be the same promoter. The promoters operably linked to different gRNAs can be different promoters. The promoters can be constitutive, inducible, repressible, or regulatable promoters.At least one gRNA can bind to a target gene or locus.When two or more gRNAs are included, each of the gRNAs binds to a different target region in one target locus, or each of the gRNAs binds to a different target region in a different locus.

[0160] B. Adeno-associated virus vector AAV can be used to deliver the composition to cells using various construct structures.For example, AAV can deliver CRISPR / Cas9-based system or CRISPR / Cpf1-based system and gRNA expression cassette on another vector.Alternatively, when the small Cas9 protein from species such as Staphylococcus aureus or Neisseria meningitidis is used, both Cas9 and up to two gRNA expression cassettes can be combined in a single AAV vector within the packaging limit of 4.7kb.

[0161] The composition comprises a modified adeno-associated virus (AAV) vector as described above.The modified AAV vector can be capable of delivering and expressing CRISPR / Cas9-based system or CRISPR / Cpf1-based system in mammalian cells.For example, the modified AAV vector can be AAV-SASTG vector (Piacentino et al. (2012) Human Gene Therapy 23:635-646).The modified AAV vector can be based on one or more of several capsid types, including AAV1, AAV2, AAV5, AAV6, AAV8 and AAV9. Modified AAV vectors can be based on AAV2 pseudotypes with alternative muscle-tropic AAV capsids, such as AAV2 / 1, AAV2 / 6, AAV2 / 7, AAV2 / 8, AAV2 / 9, AAV2.5, and AAV / SASTG vectors, which efficiently transduce skeletal or cardiac muscle by systemic and local delivery (Seto et al. Current Gene Therapy (2012) 12:139-151).

[0162] 9.Target cells As disclosed herein, gRNA, such as the optimized gRNA described herein, can be used with any type of cell together with CRISPR / Cas9 system.In some embodiments, the cell is a bacterial cell, a fungal cell, an archaeal cell, a plant cell, or an animal cell, such as a mammalian cell.In some embodiments, it can be an organ or an animal microorganism. In some embodiments, the cells are selected from the group consisting of, but not limited to, 293-T cells, 3T3 cells, 721 cells, 9L cells, A2780 cells, A2780ADR cells, A2780cis cells, A172 cells, A20 cells, A253 cells, A431 cells, A-549 cells, ALC cells, B16 cells, B35 cells, BCP-1 cells, BEAS-2B cells, bEnd.3 cells, BHK-21 cells, BR 293 cells, BxPC3 cells, C2C12 cells, C3H-10T1 / 2 cells, C6 / 36 cells, Cal-27 cells, CHO cells, COR-L23 cells, COR-L23 / CPR cells, COR-L23 / 5010 cells, COR-L23 / R23 cells, COS-7 cells, COV-434 cells, CML cells, and the like. T1 cells, CMT cells, CT26 cells, D17 cells, DH82 cells, DU145 cells, DuCaP cells, EL4 cells, EM2 cells, EM3 cells, EMT6 / AR1 cells, EMT6 / AR10.0 cells, FM3 cells, H1299 cells, H69 cells, HB54 cells, HB55 cells, HCA 2 cells, HEK-293 cells, HeLa cells, Hepa1c1c7 cells, HL-60 cells, HMEC cells, HT-29 cells, Jurkat cells, J558L cells, JY cells, K562 cells, Ku812 cells, KCL22 cells, KG1 cells, KYO1 cells, LNCap cells, Ma-MeI 1, 2, 3...48 cells, MC-38 cells, MCF-7 cells, MCF-10A cells, MDA-MB-231 cells, MDA-MB-468 cells, MDA-MB-435 cells, MDCK II cells, MDCK II cells, MG63 cells, MOR / 0.2R cells, MONO-MAC 6 cells, MRC5 cells, MTD-1A cells, MyEnd cells, NCI-H69 / CPR cells, NCI-H69 / LX10 cells, NCI-H69 / LX20 cells, NCI-H69 / LX4 cells, NIH-3T3 cells, NALM-1 cells, NW-145 cells, OPCN / OPCT cells, Peer cells, PNT-1A / PNT 2 cells,Raji cells, RBL cells, RenCa cells, RIN-5F cells, RMA / RMAS cells, Saos-2 cells, Sf-9 cells, SiHa cells, SkBr3 cells, T2 cells, T-47D cells, T84 cells, THP1 cells, U373 cells, U87 cells, U937 cells, VCaP cells, Vero cells, WM39 cells, WT-49 cells, X63 cells, YAC-1 cells, YAR cells, GM12878, K562, H1 embryonic stem cells, HeLa-S3, HepG2, HUVEC, SK-N-SH, IMR90, A549, MCF7, HMEC, etc. LHCM, C D14+, CD20+, primary heart and liver cells, differentiated H1 cells, 8988T, Adult_CD4_naive, Adult_CD4_Th0, Adult_CD4_Th1, AG04449, AG04450, AG09309, AG09319, AG10803, A oAF, AoSMC, BC_Adipose_UHN00001, BC_Adrenal_Gland_H12803N, BC_Bladder_01-11002, BC_Brain_H11058N, BC_Breast_02-03015, BC_Colon_01-1 1002, BC_Colon_H12817N, BC_Esophagus_01-11002, BC_Esophagus_H12817N, BC_Jejunum_H12817N, BC_Kidney_01-11002, BC_Kidney_H12817N, BC _Left_Ventricle_N41, BC_Leukocyte_UHN00204, BC_Liver_01-11002, BC_Lung_01-11002, BC_Lung_H12817N, BC_Pancreas_H12817N, BC_Penis_H1 2817N, BC_Pericardium_H12529N, BC_Placenta_UHN00189, BC_Prostate_Gland_H12817N, BC_Rectum_N29, BC_Skeletal_Muscle_01-11002, BC_Sk eletal_Muscle_H12817N, BC_Skin_01-11002, BC_Small_Intestine_01-11002, BC_Spleen_H12817N, BC_Stomach_01-11002, BC_Stomach_H12817N,BC_Testis_N30, BC_Uterus_BN0765, BE2_C, BG02ES, BG02ES-EBD, BJ, bone_marrow_HS27a, bone_marrow_HS5, bone_marrow_MSC, Breast_OC, Caco-2, CD20+_RO01778, CD20+_RO01794, CD34+_Mobilized, CD4+_Naive_Wb11970640, CD4+_Naive_Wb78495824, Cerebellum_OC, Cerebrum_frontal_OC, Chorion, CLL, CMK, Colo829, Colon_BC, Colon_OC, Cord_CD4_naive, Cord_CD4_Th0, Cord_CD4_Th1, Decidua, Dnd41, ECC-1, Endometrium_OC, Esophagus_BC, Fibrobl, Fibrobl_GM03348, FibroP, FibroP_AG08395, FibroP_AG08396, FibroP_AG20443, Frontal_cortex_OC, GC_B_cell, Gliobla, GM04503, GM04504, GM06990, GM08714, GM10248, GM10266, GM10847, GM12801, GM12812, GM12813, GM12864, GM12865, GM12866, GM12867, GM12868, GM12869, GM12870, GM12871, GM12872, GM12873, GM12874, GM12875, GM12878-XiMat, GM12891, GM12892, GM13976, GM13977, GM15510, GM18505, GM18507, GM18526, GM18951, GM19099, GM19193, GM19238, GM19239, GM19240, GM20000, H0287, H1-neurons, H7-hESC, H9ES, H9ES-AFP-, H9ES-AFP+, H9ES-CM, H9ES-E, H9ES-EB, H9ES-EBD, HAc, HAEpiC, HA-h, HAL, HAoAF, HAoAF_6090101.11, HAoAF_6111301.9, HAoEC, HAoEC_7071706.1, HAoEC_8061102.1, HA-sp, HBMEC, HBVP, HBVSMC, HCF, HCFaa, HCH, HCH_0011308.2P, HCH_8100808.2, HCM, HConF, HCPEpiC, HCT-116, Heart_OC, Heart_STL003, HEEpiC, HEK293, HEK293T, HEK293-T-REx, hepatocytes, HFDPC, HFDPC_0100503.2, HFDPC_0102703.3, HFF, HFF-Myc, HFL11W, HFL24W, HGF, HHSEC, HIPEpiC, HL-60, HMEpC, HMEpC_6022801.3, HMF, hMNC-CB, hMNC-CB_8072802.6, hMNC-CB_9111701.6, hMNC-PB, hMNC-PB_0022330.9, hMNC-PB_0082430.9, hMSC-AT, hMSC-AT_0102604.12, hMSC-AT_9061601.12, hMSC-BM, hMSC-BM_0050602.11, hMSC-BM_0051105.11, hMSC-UC, hMSC-UC_0052501.7, hMSC-UC_0081101.7, HMVEC-dAd, HMVEC-dBl-Ad, HMVEC-dBl-Neo, HMVEC-dLy-Ad, HMVEC-dLy-Neo, HMVEC-dNeo, HMVEC-LBl, HMVEC-LLy, HNPCEpiC, HOB, HOB_0090202.1, HOB_0091301, HPAEC, HPAEpiC, HPAF, HPC-PL, HPC-PL_0032601.13, HPC-PL_0101504.13, HPDE6-E6E7, HPdLF, HPF, HPIEpC, HPIEpC_9012801.2, HPIEpC_9041503.2, HRCEpiC, HRE, HRGEC, HRPEpiC, HSaVEC, HSaVEC_0022202.16, HSaVEC_9100101.15, HSMM, HSMM_emb, HSMM_FSHD, HSMMtube, HSMMtube_emb, HSMMtube_FSHD, HT-1080, HTR8svn, Huh-7, Huh-7.5, HVMF, HVMF_6091203.3, HVMF_6100401.3, HWP, HWP_0092205, HWP_8120201.5, iPS, iPS_CWRU1, iPS_hFib2_iPS4, iPS_hFib2_iPS5, iPS_NIHi11, iPS_NIHi7, Ishikawa, Jurkat, Kidney_BC, Kidney_OC, LHCN-M2, LHSR, Liver_OC, Liver_STL004, Liver_STL011, LNCaP, Loucy, Lung_BC, Lung_OC, lymphoblastoid cell line, M059J, MCF10A-Er-Src, MCF-7, MDA-MB-231, Medullo, Medullo_D341, Mel_2183, Melano, Monocytes-CD14+, Monocytes-CD14+_RO01746, Monocytes-CD14+_RO01826, MRT_A204, MRT_G401, MRT_TTC549, Myometr, naive B cell, NB4, NH-A, NHBE, NHBE_RA, NHDF, NHDF_0060801.3, NHDF_7071701.2, NHDF-Ad, NHDF-neo, NHEK, NHEM.f_M2, NHEM.f_M2_5071302.2, NHEM.f_M2_6022001, NHEM_M2, NHEM_M2_7011001.2, NHEM_M2_7012303, NHLF, NT2-D1, Olf_neurosphere, Osteobl, ovcar-3, PANC-1, Pancreas_OC, PanIsletD, PanIslets, PBDE, PBDEFetal, PBMC, PFSK-1, pHTE, Pons_OC, PrEC, ProgFib, Prostate, Prostate_OC, Psoas_muscle_OC, Raji, RCC_7860, RPMI-7951, RPTEC, RWPE1, SAEC, SH-SY5Y, Skeletal_Muscle_BC, SkMC, SKMC, SkMC_8121902.17, SkMC_9011302, SK-N-MC, SK-N-SH_RA, Small_intestine_OC, Spleen_OC, Stellate, Stomach_BC, T cell CD4+, T-47D, T98G, TBEC, Th1, Th1_Wb33676984, Th1_Wb54553204, Th17, Th2, Th2_Wb33676984, Th2_Wb54553204, Treg_Wb78495824, Treg_Wb83319432, U2OS, U87, UCH-1, Urothelia, WERI-Rb-1, and WI-38. In some embodiments, the target cells can be any cell, such as primary cells, HEK293 cells, 293Ts cells, SKBR3 cells, A431 cells, K562 cells, HCT116 cells, HepG2 cells, or K-Ras-dependent and K-Ras-independent cell populations.

[0163] 10. Methods of epigenome editing The present disclosure relates to a method of epigenome editing in a target cell or subject using a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. The method can be used to activate or repress a target gene. The method includes contacting a cell or subject with an effective amount of an optimized gRNA molecule as described herein and a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. In some embodiments, the optimized gRNA is encoded by a polynucleotide sequence and packaged into a lentiviral vector. In some embodiments, the lentiviral vector includes an expression cassette comprising a promoter operably linked to a polynucleotide sequence encoding an sgRNA. In some embodiments, the promoter operably linked to a polynucleotide encoding an optimized gRNA is inducible.

[0164] 11. Methods for site-specific DNA cleavage The present disclosure relates to a method of site-specific DNA cleavage in a target cell or subject using a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. The method comprises contacting a cell or subject with an effective amount of an optimized gRNA molecule as described herein and a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. In some embodiments, the optimized gRNA is encoded by a polynucleotide sequence and packaged into a lentiviral vector. In some embodiments, the lentiviral vector comprises an expression cassette comprising a promoter operably linked to a polynucleotide sequence encoding an sgRNA. In some embodiments, the promoter operably linked to a polynucleotide encoding an optimized gRNA is inducible.

[0165] The number of gRNAs administered to a cell or sample can be at least 1 gRNA, at least 2 different gRNAs, at least 3 different gRNAs, at least 4 different gRNAs, at least 5 different gRNAs, at least 6 different gRNAs, at least 7 different gRNAs, at least 8 different gRNAs, at least 9 different gRNAs, at least 10 different gRNAs, at least 11 different gRNAs, at least 12 different gRNAs, at least 13 different gRNAs, at least 14 different gRNAs, at least 15 different gRNAs, at least 16 different gRNAs, at least 17 different gRNAs, at least 18 different gRNAs, at least 18 different gRNAs, at least 20 different gRNAs, at least 25 different gRNAs, at least 30 different gRNAs, at least 35 different gRNAs, at least 40 different gRNAs, at least 45 different gRNAs, or at least 50 different gRNAs.The number of gRNAs administered to a cell can be at least 50 different gRNAs from at least one gRNA, at least 45 different gRNAs from at least one gRNA, at least 40 different gRNAs from at least one gRNA, at least 35 different gRNAs from at least one gRNA, at least 30 different gRNAs from at least one gRNA, at least 25 different gRNAs from at least one gRNA, at least 20 different gRNAs from at least one gRNA, at least 16 different gRNAs from at least one gRNA, at least 12 different gRNAs from at least one gRNA, at least 8 different gRNAs from at least one gRNA, at least 4 different gRNAs from at least one gRNA, at least 50 different gRNAs from at least 45 different gRNAs from at least 40 different gRNAs from at least 4 different gRNAs, at least 35 different gRNAs from at least 4 different gRNAs, at least 4 ... different gRNAs, at least 30 different gRNAs from at least 4 different gRNAs, at least 25 different gRNAs from at least 4 different gRNAs, at least 20 different gRNAs from at least 4 different gRNAs, at least 16 different gRNAs from at least 4 different gRNAs, at least 12 different gRNAs from at least 4 different gRNAs, at least 8 different gRNAs from at least 8 different gRNAs, at least 50 different gRNAs from at least 8 different gRNAs, at least 45 different gRNAs from at least 8 different gRNAs, at least 40 different gRNAs from at least 8 different gRNAs, at least 35 different gRNAs from at least 8 different gRNAs, at least 30 different gRNAs from 8 different gRNAs, at least 25 different gRNAs from at least 8 different gRNAs, at least 20 different gRNAs from 8 different gRNAs, at least 16 different gRNAs from at least 8 different gRNAs, or at least 12 different gRNAs from 8 different gRNAs.

[0166] The gRNA can comprise a complementary polynucleotide sequence of the target DNA sequence followed by a PAM sequence. The gRNA can comprise a "G" at the 5' end of the complementary polynucleotide sequence. The gRNA can comprise at least 10 base pairs, at least 11 base pairs, at least 12 base pairs, at least 13 base pairs, at least 14 base pairs, at least 15 base pairs, at least 16 base pairs, at least 17 base pairs, at least 18 base pairs, at least 19 base pairs, at least 20 base pairs, at least 21 base pairs, at least 22 base pairs, at least 23 base pairs, at least 24 base pairs, at least 25 base pairs, at least 30 base pairs, or at least 35 base pairs of the complementary polynucleotide sequence of the target DNA sequence followed by a PAM sequence. The PAM sequence can be "NGG", where "N" can be any nucleotide. The gRNA can target at least one of the promoter region, enhancer region, or transcription region of the target gene. In some embodiments, the gRNA targets a nucleic acid sequence having at least one polynucleotide sequence of SEQ ID NOs: 13-148, 316, 317, or 320. The gRNA can include at least one nucleic acid sequence of SEQ ID NOs: 149-315, 321-323, or 326-329.

[0167] 12. Methods of correcting mutant genes and treating subjects The present disclosure is also directed to a method of correcting a mutated gene in a subject. The method comprises administering a composition as described above to a cell of the subject. The use of the composition to deliver a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, to a cell can restore the expression of a fully functional or partially functional protein together with a repair template or donor DNA (which can replace the entire gene or the region containing the mutation). The CRISPR / Cas9-based system or a CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, can be used to introduce a site-specific double-strand break at a targeted genomic locus. The site-specific double-strand break is produced when the CRISPR / Cas9-based system or a CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, binds to a target DNA sequence, thereby allowing the target DNA to be cut. This DNA break can stimulate natural DNA repair mechanisms, resulting in one of two possible repair pathways: homology-directed repair (HDR) or non-homologous end joining (NHEJ).

[0168] The present disclosure is directed to genome editing using a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, without a repair template, which can efficiently correct the reading frame and restore the expression of a functional protein involved in a genetic disease. The disclosed CRISPR / Cas9-based system or CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, can include using a homology-directed repair or nuclease-mediated non-homologous end joining (NHEJ)-based correction approach, which allows for efficient correction in growth-limited primary cell lines where homology-directed or selection-based gene correction may not be applicable. This strategy integrates rapid and robust assembly of an active CRISPR / Cas9-based or CRISPR / Cpf1-based system, involving at least one gRNA, such as the optimized gRNA described herein, with efficient gene editing methods for the treatment of genetic diseases caused by mutations in non-essential coding regions resulting in frameshifts, premature stop codons, ectopic splice donor sites, or ectopic splice acceptor sites.

[0169] A nuclease-mediated nonhomologous end joining Restoration of protein expression from endogenous mutated gene can be by template-free NHEJ-mediated DNA repair.In contrast to the transient method of targeting target gene RNA, the correction of target gene reading frame in genome by transiently expressed CRISPR / Cas9-based system or CRISPR / Cpf1-based system with at least one gRNA, such as the optimized gRNA described herein, can lead to permanently restored target gene expression by each modified cell and all its progeny.

[0170] Nuclease-mediated NHEJ gene correction corrects mutated target genes and can offer several potential advantages over HDR pathways. For example, NHEJ does not require a donor template, which can lead to non-specific insertion mutations. In contrast to HDR, NHEJ operates efficiently in all phases of the cell cycle, and can therefore be efficiently utilized in both cycling and post-mitotic cells (such as muscle fibers). This provides a robust and permanent gene repair alternative to oligonucleotide-based exon skipping or drug-imposed read-through of stop codons, and may theoretically require only one drug treatment. NHEJ-based gene correction using CRISPR / Cas9-based or CRISPR / Cpf1-based systems, as well as other genetically engineered nucleases, including meganucleases and zinc finger nucleases, can be combined with other existing ex vivo and in vivo platforms for cell- and gene-based therapy, in addition to the plasmid electroporation approach described herein. For example, delivery of CRISPR / Cas9- or CRISPR / Cpf1-based systems by mRNA-based transfection or as purified cell-permeable proteins can enable a DNA-free genome editing approach that would avoid any possibility of insertional mutagenesis.

[0171] B homology-directed repair Restoring protein expression from endogenous mutated gene can include homologous recombination repair.The method as described above further includes administering a donor template to the cell.The donor template can include a nucleotide sequence encoding a fully functional or partially functional protein.For example, the donor template can include a miniaturized dystrophin construct (called "minidys"), a fully functional dystrophin construct for repairing mutated dystrophin gene, or a fragment of the dystrophin gene that results in the repair of mutated dystrophin gene after homologous recombination repair.

[0172] 13. Genome editing methods The present disclosure also relates to genome editing using the above-described CRISPR / Cas9-based system or CRISPR / Cpf1-based system to restore the expression of fully functional or partially functional proteins using repair template or donor DNA (which can replace the whole gene or the region containing mutation). The CRISPR / Cas9-based system or CRISPR / Cpf1-based system can be used to introduce site-specific double-strand breaks at targeted genomic loci. Site-specific double-strand breaks are produced when the CRISPR / Cas9-based system or CRISPR / Cpf1-based system uses gRNA to bind to target DNA sequence, thereby allowing target DNA to be cut. The CRISPR / Cas9-based system and the CRISPR / Cpf1-based system have the advantage of the advancement of genome editing due to their rapid, successful and efficient gene modification. This DNA break can stimulate natural DNA repair mechanisms, resulting in one of two possible repair pathways: homology-directed repair (HDR) or non-homologous end joining (NHEJ).

[0173] The present disclosure is directed to genome editing using a repair template-free CRISPR / Cas9-based or CRISPR / Cpf1-based system that can efficiently correct reading frames and restore the expression of functional proteins involved in genetic diseases. The disclosed CRISPR / Cas9-based or CRISPR / Cpf1-based systems and methods can include using homology-directed repair or nuclease-mediated non-homologous end joining (NHEJ)-based correction techniques, which allow efficient correction in growth-limited primary cell lines where homology-directed or selection-based gene correction may not be applicable. This strategy integrates the rapid and robust assembly of an active CRISPR / Cas9-based or CRISPR / Cpf1-based system with an efficient gene editing method for the treatment of genetic diseases caused by mutations in non-essential coding regions that result in frameshifts, premature stop codons, ectopic splice donor sites, or ectopic splice acceptor sites.

[0174] The present disclosure provides a method for correcting mutant genes in cells and treating subjects suffering from genetic diseases such as DMD. The method can include administering a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, a polynucleotide or vector encoding said CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, or a composition of said CRISPR / Cas9-based system or a CRISPR / Cpf1-based system to a cell or a subject as described above. The method can include administering a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, for example, administering a Cas9 protein, a Cpf1 protein, a Cas9 fusion protein containing a second domain, a nucleotide sequence encoding said Cas9 protein, Cpf1 protein, or a Cas9 fusion protein, and / or at least one gRNA, where the gRNAs target different DNA sequences. These target DNA sequences can be overlapping. The number of gRNAs administered to the cells can be at least one gRNA, at least two different gRNAs, at least three different gRNAs, at least four different gRNAs, at least five different gRNAs, at least six different gRNAs, at least seven different gRNAs, at least eight different gRNAs, at least nine different gRNAs, at least ten different gRNAs, at least fifteen different gRNAs, at least twenty different gRNAs, at least thirty different gRNAs, or at least fifty different gRNAs, as described above. The gRNAs can include at least one nucleic acid sequence of SEQ ID NOs: 149-315, 321-323, or 326-329. The method can include homology directed repair or non-homologous end joining.

[0175] 14. Constructs and Plasmids The composition may comprise a genetic construct encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system as disclosed herein, as described above. The genetic construct, such as a plasmid, may comprise a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, such as a Cas9 protein, a Cpf1 protein, and a Cas9 fusion protein, and / or at least one of the optimized gRNAs described herein. The composition may comprise a genetic construct encoding a modified AAV vector, as described above, with at least one gRNA, such as an optimized gRNA, as described herein, and a nucleic acid sequence encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. The genetic construct, such as a plasmid, may comprise a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, as described herein, with at least one gRNA, such as an optimized gRNA, as described herein. The composition may comprise a genetic construct encoding a modified lentiviral vector, as described above, with at least one gRNA, such as an optimized gRNA, as described herein. The gene construct, such as a plasmid, can include the nucleic acid encoding Cas9-fusion protein and at least one sgRNA.The gene construct can exist in the cell as a functional extrachromosomal molecule.The gene construct can be a linear minichromosome (including centromere), a telomere, or a plasmid or cosmid.

[0176] Gene constructs can also be part of the genome of recombinant virus vectors, including recombinant lentiviruses, recombinant adenoviruses, and recombinant adenovirus-associated viruses. Gene constructs can be part of the genetic material in attenuated live microorganisms or recombinant microbial vectors that live in cells. Gene constructs can include regulatory elements for gene expression of the coding sequence of nucleic acid. Regulatory elements can be promoters, enhancers, start codons, stop codons, or polyadenylation signals.

[0177] The nucleic acid sequence may constitute a gene construct, which may be a vector. The vector may be capable of expressing a fusion protein, such as a Cas9-fusion protein, in a mammalian cell. The vector may be recombinant. The vector may contain a heterologous nucleic acid encoding a Cas9-fusion protein. The vector may be a plasmid. The vector may be useful for introducing a nucleic acid encoding a Cas9-fusion protein into a cell, where the transformed host cell is cultured and maintained under conditions in which expression of the Cas9-fusion protein system occurs.

[0178] Coding sequences can be optimized for stability and high levels of expression. In some cases, codons are selected to reduce secondary RNA structure formation, such as those formed due to intramolecular binding.

[0179] The vector can include a heterologous nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, and can further include a start codon that can be upstream of the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence, and a stop codon that can be downstream of the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The start and stop codons can be in frame with the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The vector can also include a promoter operably linked to the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The promoter operably linked to the CRISPR / Cas9-based or CRISPR / Cpf1-based system coding sequence can be a promoter from simian virus 40 (SV40), a mouse mammary tumor virus (MMTV) promoter, a human immunodeficiency virus (HIV) promoter, such as the bovine immunodeficiency virus (BIV) long terminal repeat (LTR) promoter, a Moloney virus promoter, an avian leukosis virus (ALV) promoter, a cytomegalovirus (CMV) promoter, such as the CMV immediate early promoter, an Epstein-Barr virus (EBV) promoter, or a Rous sarcoma virus (RSV) promoter. The promoter can also be a promoter from a human gene, such as human ubiquitin C (hUbC), human actin, human myosin, human hemoglobin, human muscle creatine, or human metallothionein. The promoter can also be a tissue-specific promoter, such as a muscle- or skin-specific promoter, natural or synthetic. Examples of such promoters are described in US Patent Publication No. 20040175727, the contents of which are incorporated herein in their entirety.

[0180] The vector can also include a polyadenylation signal that can be downstream of the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system.The polyadenylation signal can be SV40 polyadenylation signal, LTR polyadenylation signal, bovine growth hormone (bGH) polyadenylation signal, human growth hormone (hGH) polyadenylation signal, or human β-globin polyadenylation signal.The SV40 polyadenylation signal can be the polyadenylation signal from pCEP4 vector (Invitrogen, San Diego, CA).

[0181] The vector can also include an enhancer upstream of a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, i.e., a Cas9 protein, a Cpf1 protein, or a Cas9 fusion protein coding sequence, or an sgRNA, such as an optimized gRNA, as described herein. The enhancer can be essential for DNA expression. The enhancer can be human actin, human myosin, human hemoglobin, human muscle creatine, or a viral enhancer, such as those derived from CMV, HA, RSV, or EBV. Polynucleotide function enhancers are described in U.S. Pat. No. 5,593,972, U.S. Pat. No. 5,962,428, and WO 94 / 016737, the contents of each of which are fully incorporated by reference. The vector can also include a mammalian origin of replication to maintain the vector extrachromosomally, and can produce multiple copies of the vector in the cell. The vector can also include regulatory sequences, which can be sufficiently adapted for gene expression in mammalian or human cells into which the vector is administered. The vector can also include a reporter gene, such as green fluorescent protein ("GFP"), and / or a selectable marker, such as hygromycin ("Hygro").

[0182] The vector can be an expression vector or a system for producing the protein by conventional techniques and readily available starting materials (Sambrook et al., Molecular Cloning and Laboratory Manual, Second Ed., Cold Spring Harbor (1989), fully incorporated by reference). In some embodiments, the vector can comprise a nucleic acid sequence encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system (including a nucleic acid sequence encoding a Cas9 protein, a Cpf1 protein, or a Cas9 fusion protein), and a nucleic acid sequence encoding at least one gRNA comprising at least one nucleic acid sequence of SEQ ID NOs: 149-315, 321-323, or 326-329.

[0183] 15. Pharmaceutical Compositions The composition may be in a pharmaceutical composition. The pharmaceutical composition may comprise about 1 ng to about 10 mg of DNA encoding the CRISPR / Cas9-based system, the CRISPR / Cpf1-based system, or the protein components of the CRISPR / Cas9-based system, i.e., Cas9 protein, Cpf1 protein, or Cas9 fusion protein. The pharmaceutical composition may comprise about 1 ng to about 10 mg of DNA of a modified AAV vector and a nucleotide sequence encoding the CRISPR / Cas9-based system, together with at least one gRNA, such as the optimized gRNA described herein. The pharmaceutical composition may comprise about 1 ng to about 10 mg of DNA of a modified lentiviral vector. The pharmaceutical composition according to the present invention is formulated according to the mode of administration to be used. When the pharmaceutical composition is an injectable pharmaceutical composition, it is sterile, pyrogen-free, and particulate-free. It is preferred that an isotonic formulation is used. In general, additives for isotonicity may include sodium chloride, dextrose, mannitol, sorbitol, and lactose. In some cases, an isotonic solution such as phosphate buffered saline is preferred. Stabilizers include gelatin and albumin. In some embodiments, a vasoconstrictor is added to the formulation.

[0184] The composition can further comprise a pharmaceutically acceptable excipient. The pharmaceutically acceptable excipient can be a functional molecule as a vehicle, adjuvant, carrier, or diluent. The pharmaceutically acceptable excipient can be a gene transfer promoter (which can include surfactants), such as immune stimulating complexes (ISCOMS), Freund's incomplete adjuvant, LPS analogs (including monophosphoryl lipid A), muramyl peptides, quinone analogs, vesicles such as squalene and squalene, hyaluronic acid, lipids, liposomes, calcium ions, viral proteins, polyanions, polycations, or nanoparticles, or other known gene transfer promoters.

[0185] The gene transfer promoter is a polyanion, a polycation (including poly-L-glutamic acid (LGS)), or a lipid. The gene transfer promoter is poly-L-glutamic acid, and more preferably, poly-L-glutamic acid is present in the composition for genome editing at a concentration of less than 6 mg / ml. The gene transfer promoter can also include surfactants, such as immune stimulating complexes (ISCOMS), Freund's incomplete adjuvant, LPS analogs (including monophosphoryl lipid A), muramyl peptides, quinone analogs, and vesicles, such as squalene and squalene, and hyaluronic acid can also be used and administered with the gene construct. In some embodiments, the DNA vector encoding the composition may also include a gene transfer facilitating agent, such as a lipid, a liposome (including lecithin liposomes or other liposomes known in the art), a DNA-liposome mixture (see, e.g., WO 09324640), calcium ions, viral proteins, polyanions, polycations, or nanoparticles, or other known gene transfer facilitating agents. Preferably, the gene transfer facilitating agent is a polyanion, polycation (including poly-L-glutamic acid (LGS)), or a lipid.

[0186] 16. Constructs and Plasmids The composition may comprise a genetic construct encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system as disclosed herein, as described above. The genetic construct, such as a plasmid or expression vector, may comprise a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, and / or at least one gRNA, such as an optimized gRNA, as described herein. The composition may comprise a genetic construct encoding a modified lentiviral vector, as described above, and a nucleic acid sequence encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system as disclosed herein. The genetic construct, such as a plasmid, may comprise a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system. The composition may comprise a genetic construct encoding a modified lentiviral vector, as described above. The genetic construct, such as a plasmid, can include a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system and at least one sgRNA, such as an optimized gRNA described herein. The genetic construct can be present in the cell as a functional extrachromosomal molecule. The genetic construct can be a linear minichromosome (including a centromere), a telomere, or a plasmid or cosmid.

[0187] Gene constructs can also be part of the genome of recombinant virus vectors, including recombinant lentiviruses, recombinant adenoviruses, and recombinant adenovirus-associated viruses. Gene constructs can be part of the genetic material in attenuated live microorganisms or recombinant microbial vectors that live in cells. Gene constructs can include regulatory elements for gene expression of the coding sequence of nucleic acid. Regulatory elements can be promoters, enhancers, start codons, stop codons, or polyadenylation signals.

[0188] The nucleic acid sequence can constitute a gene construct, which can be a vector. The vector can be capable of expressing a fusion protein, such as a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, in mammalian cells. The vector can be recombinant. The vector can include a heterologous nucleic acid encoding a fusion protein, such as a CRISPR / Cas9-based system. The vector can be a plasmid. The vector can be useful for introducing a nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system into a cell, where the transformed host cell is cultured and maintained under conditions in which expression of the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system occurs.

[0189] Coding sequences can be optimized for stability and high levels of expression. In some cases, codons are selected to reduce secondary RNA structure formation, such as those formed due to intramolecular binding.

[0190] The vector can include a heterologous nucleic acid encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system, and can further include a start codon that can be upstream of the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence, and a stop codon that can be downstream of the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The start and stop codons can be in frame with the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The vector can also include a promoter operably linked to the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence. The CRISPR / Cas9-based system or CRISPR / Cpf1-based system can be under light-inducible or chemical-inducible control to allow dynamic control in space and time. The promoter operably linked to the CRISPR / Cas9-based system or CRISPR / Cpf1-based system coding sequence can be a promoter from simian virus 40 (SV40), a mouse mammary tumor virus (MMTV) promoter, a human immunodeficiency virus (HIV) promoter, such as the bovine immunodeficiency virus (BIV) long terminal repeat (LTR) promoter, a Moloney virus promoter, an avian leukosis virus (ALV) promoter, a cytomegalovirus (CMV) promoter, such as the CMV immediate early promoter, an Epstein-Barr virus (EBV) promoter, or a Rous sarcoma virus (RSV) promoter. The promoter can also be a promoter from a human gene, such as human ubiquitin C (hUbC), human actin, human myosin, human hemoglobin, human muscle creatine, or human metallothionein. The promoter can be a tissue-specific promoter, such as a muscle- or skin-specific promoter, natural or synthetic. Examples of such promoters are described in US Patent Publication No. 20040175727, the contents of which are incorporated herein in their entirety.

[0191] The vector can also include a polyadenylation signal that can be downstream of the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system.The polyadenylation signal can be SV40 polyadenylation signal, LTR polyadenylation signal, bovine growth hormone (bGH) polyadenylation signal, human growth hormone (hGH) polyadenylation signal, or human β-globin polyadenylation signal.The SV40 polyadenylation signal can be the polyadenylation signal from pCEP4 vector (Invitrogen, San Diego, CA).

[0192] The vector can also include an enhancer upstream of the CRISPR / Cas9-based system or the CRISPR / Cpf1-based system and / or the sgRNA, such as the optimized gRNA described herein. The enhancer can be essential for DNA expression. The enhancer can be human actin, human myosin, human hemoglobin, human muscle creatine, or a viral enhancer, such as those derived from CMV, HA, RSV, or EBV. Polynucleotide function enhancers are described in U.S. Pat. No. 5,593,972, U.S. Pat. No. 5,962,428, and WO 94 / 016737, the contents of each of which are fully incorporated by reference. The vector can also include a mammalian origin of replication to maintain the vector extrachromosomally and produce multiple copies of the vector in the cell. The vector can also include a regulatory sequence, which can be fully adapted for gene expression in the mammalian or human cell to which the vector is administered. Vectors may also include a reporter gene, such as green fluorescent protein ("GFP"), and / or a selectable marker, such as hygromycin ("Hygro").

[0193] The vector can be an expression vector or a system for producing a protein by conventional techniques and readily available starting materials (Sambrook et al., Molecular Cloning and Laboratory Manual, Second Ed., Cold Spring Harbor (1989), fully incorporated by reference). In some embodiments, the vector can include a nucleic acid sequence encoding a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system and at least one gRNA, such as an optimized gRNA described herein.

[0194] In some embodiments, a gRNA, such as an optimized gRNA described herein, is encoded by a polynucleotide sequence and packaged into a lentiviral vector. In some embodiments, the lentiviral vector comprises an expression cassette. The expression cassette can comprise a promoter operably linked to a polynucleotide sequence encoding a gRNA, such as an optimized gRNA described herein. In some embodiments, the promoter operably linked to a polynucleotide encoding a gRNA is inducible.

[0195] i. Adeno-associated virus vector The composition comprises a modified adeno-associated virus (AAV) vector as described above. The modified AAV vector can have enhanced cardiac and skeletal muscle tissue tropism. The modified AAV vector can be capable of delivering and expressing a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system in mammalian cells with at least one gRNA, such as the optimized gRNA described herein. For example, the modified AAV vector can be an AAV-SASTG vector (Piacentino et al. (2012) Human Gene Therapy 23:635-646). The modified AAV vector can deliver nuclease to skeletal and cardiac muscle in vivo. The modified AAV vector can be based on one or more of several capsid types, including AAV1, AAV2, AAV5, AAV6, AAV8, and AAV9. Modified AAV vectors can be based on AAV2 pseudotypes with alternative muscle-tropic AAV capsids, such as AAV2 / 1, AAV2 / 6, AAV2 / 7, AAV2 / 8, AAV2 / 9, AAV2.5, and AAV / SASTG vectors, which efficiently transduce skeletal or cardiac muscle by systemic and local delivery (Seto et al. Current Gene Therapy (2012) 12:139-151).

[0196] 17. Method of Service Provided herein is a method for delivering CRISPR / Cas9-based system or CRISPR / Cpf1-based system and optimized gRNA described herein to provide gene constructs and / or proteins of CRISPR / Cas9-based system or CRISPR / Cpf1-based system.Delivery of CRISPR / Cas9-based system or CRISPR / Cpf1-based system and optimized gRNA described herein can be transfection or electroporation of CRISPR / Cas9-based system or CRISPR / Cpf1-based system and optimized gRNA described herein as one or more nucleic acid molecules expressed in cells and delivered to the surface of cells.CRISPR / Cas9-based system or CRISPR / Cpf1-based system proteins can be delivered to cells.These nucleic acid molecules can be electroporated using BioRad Gene Pulser Xcell or Amaxa Nucleofector IIb device or other electroporation device. A number of different buffers can be used, including BioRad electroporation solution, Sigma phosphate buffered saline (product #D8537) (PBS), Invitrogen OptiMEM I (OM), or Amaxa Nucleofector solution V (NV). Gene transfer can include a gene transfer reagent such as Lipofecamine 2000.

[0197] The vector encoding the CRISPR / Cas9-based system or CRISPR / Cpf1-based system protein can be delivered to the modified target cells in tissue or subject by DNA injection (also called DNA vaccination) with or without in vivo electroporation, liposome-mediated, nanoparticle-facilitated, and / or recombinant vector. The recombinant vector can be delivered by any virus type. The virus type can be recombinant lentivirus, recombinant adenovirus, and / or recombinant adeno-associated virus.

[0198] Nucleotides encoding CRISPR / Cas9-based or CRISPR / Cpf1-based system proteins can be introduced into cells to induce gene expression of target genes. For example, one or more nucleotide sequences encoding CRISPR / Cas9-based or CRISPR / Cpf1-based systems induced in target genes can be introduced into mammalian cells. Upon delivery of the CRISPR / Cas9-based or CRISPR / Cpf1-based system to cells, and upon delivery of the vector to mammalian cells, the transgenic cells will express the CRISPR / Cas9-based or CRISPR / Cpf1-based system. The CRISPR / Cas9-based or CRISPR / Cpf1-based system can be administered to mammals to induce or regulate gene expression of target genes in mammals. The mammal may be a human, non-human primate, cow, pig, sheep, goat, antelope, bison, water buffalo, bovid, deer, hedgehog, elephant, llama, alpaca, mouse, rat, or chicken, preferably a human, cow, pig, or chicken.

[0199] Methods for introducing nucleic acids into host cells are known in the art, and any known method can be used to introduce a nucleic acid (e.g., an expression construct) into a cell. Suitable methods include, for example, viral or bacteriophage infection, gene transfer, conjugation, protoplast fusion, lipofection, electroporation, calcium phosphate precipitation, polyethyleneimine (PEI)-mediated gene transfer, DEAE-dextran-mediated gene transfer, liposome-mediated gene transfer, particle gun technology, calcium phosphate precipitation, direct microinjection, nanoparticle-mediated nucleic acid delivery, and the like. In some embodiments, the compositions can be delivered by mRNA delivery and ribonucleoprotein (RNP) complex delivery.

[0200] 18. Route of Administration The composition can be administered to a subject by various routes, including orally, parenterally, sublingually, transdermally, rectally, transmucosally, topically, via inhalation, via buccal administration, intrapleurally, intravenously, intraarterially, intraperitoneally, subcutaneously, intramuscularly, intranasally, intrathecally, and intraarticularly, or combinations thereof. For veterinary use, the composition can be administered in a suitably acceptable formulation according to normal veterinary practice. Veterinarians can easily determine the most suitable dosing regimen and route of administration for a particular animal. The composition can be administered by conventional syringes, needleless injection devices, "microprojectile bombardment gone guns", or other physical methods, such as electroporation ("EP"), "hydrodynamic methods", or ultrasound.

[0201] The composition can be delivered to mammals by several techniques, including DNA injection (also called DNA vaccination) with or without in vivo electroporation, liposome-mediated, nanoparticle-facilitated, recombinant vectors such as recombinant lentivirus, recombinant adenovirus, and recombinant adenovirus-associated virus.The composition can be injected into skeletal muscle or cardiac muscle.For example, the composition can be injected into tibialis anterior muscle.

[0202] 19. Kit Provided herein is a kit that can be used for site-specific DNA binding. The kit includes a composition as described above and instructions for using said composition. The instructions included in the kit can be affixed to the packaging material or included as a package insert. The instructions are generally, but are not limited to, written or printed materials. Any medium capable of storing such instructions and communicating them to the end user is contemplated by the present disclosure. Such media include, but are not limited to, electronic storage media (e.g., magnetic disks, tapes, cartridges, chips), optical media (e.g., CD ROM), and the like. As used herein, the term "instructions" can include the address of an internet site that provides the instructions.

[0203] The composition can include a modified lentiviral vector, a nucleotide sequence encoding a CRISPR / Cas9-based system, and an optimized gRNA, as described above. The CRISPR / Cas9-based system can be included in a kit for specifically binding and targeting a particular regulatory region of a regulatory region, as described above. EXAMPLES

[0204] 20. Example The foregoing can be better understood by reference to the following examples, which are given for purposes of illustration and not to limit the scope of the invention.

[0205] Example 1 material and method Materials. Tris-HCl (pH 7.6) buffer was obtained from Corning Life Sciences. L-glutamic acid monopotassium salt monohydrate, dithiothreitol (DTT), and magnesium chloride were obtained from Sigma Aldrich Co., LLC.

[0206] Cloning of Cas9, dCas9, and sgRNA expression plasmids; Plasmids encoding Cas9, dCas9, and sgRNA targeting the AAVS1 locus of human chromosome 19 were cloned, expressed, and purified using standard techniques. The DNA substrates used for imaging - (i) a 1198 bp substrate derived from a segment of the AAVS1 locus of human chromosome 19; (ii) a "genetically engineered" 989 bp DNA substrate containing a series of six complete, partial, or mismatched target sites; and (iii) a 1078 bp "nonsense" substrate containing no homology to the protospacer (>3 bp) - were also produced using standard techniques. Plasmids encoding wild-type Cas9 and dCas9 were obtained from Addgene (plasmid 39312 and plasmid 47106). Plasmids for expression of Cas9 and dCas9 in bacteria were cloned using Gateway Cloning (Life Technologies). Briefly, PCR was used to amplify Cas9 and dCas9 genes and add flanking attL1 and attL2 sites. BP recombination was performed to introduce these genes into a shuttle vector, and then LP recombination was performed to introduce these genes into pDest17 and add an N-terminal hexa-histidine tag (Life Technologies). Plasmids encoding chimeric sgRNAs and sgRNA variants (described below) were cloned as previously described (Perez-Pinera et al., (2013) Nature methods, 10, 973-976).

[0207] Expression and purification of Cas9, dCas9. Plasmids encoding Cas9 or dCas9 were transformed into SoluBL21 competent cells (Genlantis) according to standard techniques (Sambrook, J., Fritsch, EF, and Maniatis, T. (1989) Molecular cloning. Cold spring harbor laboratory press New York.). A single colony was used to inoculate a 25 mL starter culture. The 25 mL starter culture was grown overnight and used to inoculate a 1 L culture. The inoculated 1 L culture was grown at 25°C for 5 hours, after which the temperature was reduced to 16°C and protein expression was induced by the addition of 0.1 mM IPTG. The induced culture was grown at 16°C for an additional 12 hours. Cells were harvested by centrifugation at 4000×g and stored at −80°C for long-term storage.

[0208] The cell pellet was resuspended in 30 mL of Lysis buffer (50 mM Tris-HCl, 500 mM NaCl, 10 mM MgCl2, 10% v / v glycerol, 0.2% Triton-1000, and 1 mM PMSF). The cell suspension was lysed by sonication at 30% duty cycle for 5 min. The suspension was then centrifuged at 12,000×g for 30 min. The supernatant was then removed and incubated with Ni-NTA resin (Qiagen) for 30 min under gentle agitation. The resin was then packed into a column, washed with wash buffer (35 mM imidizole, 50 mM Tris-HCl, 500 mM NaCl, 10 mM MgCl2, 10% v / v glycerol) and eluted with elution buffer (120 mM imidizole, 50 mM Tris-HCl, 500 mM NaCl, 10 mM MgCl2, 10% v / v glycerol). The solvent was then exchanged into storage buffer (50 mM Tris-HCl, 500 mM NaCl, 10 mM MgCl2, 10% v / v glycerol) using an Ultracel-30k centrifugal filter. Samples were then aliquoted and frozen at -80°C. A representative polyacrylamide SDS gel of purified Cas9 and dCas9 is shown in Figure S1, indicating approximately >95% purity.

[0209] Expression and purification of sgRNA and guide RNA variants. Guide RNA was transcribed in vitro using MEGAshortscript T7 Transcription Kit (Life Technologies). DNA templates with T7 promoters were generated via PCR from guide RNA plasmids, and reactions were set up according to the manufacturer's instructions. T7 templates for guide RNAs truncated by two nucleotides from their 5' ends (tru-gRNAs) and those with 5' extensions forming hairpins (hp-gRNAs) were generated by PCR of standard gRNA plasmids. RNA was then purified using phenol-chloroform extraction using standard techniques (Sambrook et al. (1989) Molecular cloning. Cold spring harbor laboratory press New York).

[0210] Production of DNA substrates. Genomic DNA was extracted and purified from HEK293T cell line using DNeasy kit (Qiagen) according to the manufacturer's protocol. PCR was then used to amplify the AAVS1 locus. A 1198 bp substrate from AAVS1 was constructed via direct PCR from genomic DNA using the primers 5'-\Bt\-CCAGGATCAGTGAAACGCAC-3' and 5'-GAGCTCTACTGGCTTCTGCG-3' (where \Bt\ represents the biotin labeling of the primer at the 5' end) from Integrated DNA Technologies (IDT). The "genetically modified" DNA substrates containing a series of PAMs and full or partial protospacer sites were ordered as two gBlock fragments, each containing an EcoRI restriction site at one end. The substrates were digested, ligated together, and then enriched via PCR with primers (Integrated DNA Technologies, IDT): 5'-\Bt\-CATGACGTGCAGCAAGC-3' and 5'-CGACGATGCGCTGAATC-3'. To construct a "nonsense" substrate that does not contain sites showing homology (>3 bp) to the protospacer: a 690 bp DNA construct containing a series of restriction sites was synthesized (GeneScript, Inc.) and an additional length of DNA from lambda DNA (New England Biolabs) was subcloned into this construct; a 1078 bp substrate was then PCR amplified using primers (IDT): 5'-\Bt\-GACCTGCAGGCATGCAAGCTTGG-3' and 5'-CAGCGTCCCCGGTTGTGAATCT-3'. All DNA was gel purified, diluted to 25 nM in working buffer (20 mM Tris-HCl (pH 7.6), 100 mM potassium glutamate, 5 mM MgCl2, and 0.4 mM DTT) and incubated with 40× excess monomeric streptavidin for 10 min (Howarth et al., (2006) Nature Methods, 3, 267-273) and then incubated with Cas9 / dCas9.

[0211] Sodium dodecyl sulfate-polyacrylamide gels of purified Cas9 and dCas9 are shown in Figures 8A-8B, indicating approximately >95% purity.

[0212] Atomic force microscopy. Atomic force microscopy (AFM) was performed in air using a Bruker (nee Veeco) Nanoscope V Multimode with a RTSEP (Bruker) probe (nominal spring constant 40 N / m, resonant frequency 300 kHz). Prior to the experiment, protein and guide RNA were mixed in a 1:1.5 ratio for 10 min. Protein and DNA were mixed in a solution of working buffer for at least 10 min (maximum 35 min) at room temperature and deposited for 8 s on freshly cleaved mica (Ted Pella, Inc.) that had been treated with 3-aminopropylsiloxane prepared (as previously described (24)), rinsed with ultrapure water (>17 MΩ), and air-dried. Protein was briefly centrifuged and then incubated with DNA. At least four preparations were imaged for each experimental condition when standard sgRNA was used and at least two preparations for experiments with other guide RNA variants. Typically, images were acquired at 1-1.5 lines / s for each sample with a pixel resolution of 1024 x 1024 for an area of ​​2.75 sq. microns or 2048 x 2048 for an area of ​​5.5 sq. microns. For each experimental condition, images of several thousand (approximately 2500-6000) DNA molecules were resolved.

[0213] DNA Tracking and Refinement with Sub-Pixel Resolution. The obtained AFM images were smoothed and leveled (plane-wise, line-wise, and third-order polynomial leveling) using Gwyddion (http: / / gwyddion.net / ), an open source image analysis software for scanning probe microscopy, and then exported to MATLAB (Mathworks, Inc.). 151×151 pixel (405 nm×405 nm) regions containing each DNA molecule were visually screened for clearly identifiable streptavidin labeling, the presence of at least one bound Cas9 / dCas9 molecule, and a clear end-to-end path to ensure no clumping or overlap with other DNA molecules. The DNA contour was manually traced and the putative boundaries of streptavidin and Cas9 / dCas9 were marked. This tracing was then refined algorithmically using a method based on Wiggins et al. (2006) Nature nanotechnology, 1, 137-141. Starting from the highlighted center of streptavidin (x1), the position of the next element of the backbone (x2) is estimated by stepping 2.5 nm towards the closest hand traced point that crosses the estimated boundary of streptavidin. For a two-fold linear interpolation of the image of DNA, draw an 11 pixel line perpendicular to the (x1-x2) line segment at x2. Relocate x2 to the position on the normal with the maximum topographical height, then adjust 2.5 nm from x1 on the new (x1-x2) line. The position of x3... is then x n is iteratively estimated using the nearest hand-traced points to give an initial guess for the next backbone position, which is then refined as before to find the point x n Continue the refinement process until the traced DNA molecule is less than 2.5 nm from its edge. If the refined trace falls within the predicted boundary of the Cas9 / dCas9 molecule at xi, then recalculate the DNA position at x instead. i A point on the cubic Hermite spline (point x i-1 , x i , xj , and x j+1 Here, x j is the first hand-traced point that crosses the putative Cas9 / dCas9 boundary).

[0214] Upon completion of the tracing, the height of the DNA along the contour (relative to the median pixel height of the local region) is extracted. The estimated boundary of streptavidin and Cas9 / dCas9 is calculated as this (μ d +σ d ), where μ d and σ d are the mean and standard deviation of the height of the traced DNA above the predicted location of the bound protein and where the estimates converge.

[0215] To account for any instrumental hysteresis that may distort the apparent length of the DNA, the DNA length was normalized and only DNA molecules originally measured to be 20% of their expected length (assuming a known number of base pairs, 0.33 nm per base pair) were used for further analysis (for AAVS1 substrate - number of traces: 804; nominal length: 1198 bp, average recorded length: 1283 bp, standard deviation: 154 bp; for genetically modified substrates - number of traces: 1520, nominal length: 986 bp, average recorded length: 1071 bp, standard deviation: 124 bp; for "nonsense" substrate - number of traces: 616, nominal length: 1078 bp, average recorded length: 1217 bp, standard deviation: 135 bp). This step prevents us from inappropriately analyzing, for example, two DNA molecules that appear to be collinear, DNA that may have been fragmented, or DNA that may have been cut by Cas9 and separated (which was rare, see text).

[0216] The binding histograms in Figures 1C-1D, 2C-2D, and 9 were generated by mapping the relative position of each bound protein to the bases overlapped by that protein (nearest neighbor interpolation) and summing the total number of proteins bound to each site (if a single Cas9 / dCas9 can be interpreted as contacting multiple (k) sites, then in the binding histograms, each contact region was weighted by 1 / k). Peaks in the binding histograms were weighted by an empirical Gaussian exp(-((x-μ) / w) 2 ) where μ is the mean peak position and w is the peak width parameter (w=√2σ with standard deviation σ) using MATLAB.

[0217] Determination of the apparent dissociation constant of dCas9. The apparent dissociation constant of dCas9 with different guide RNA variants was determined as previously described (Yang et al. (2005) Nucleic Acids Res., 33, 4322-4334). Briefly, known solution concentrations of dCas9-guide RNA ([dCas9]0) and DNA molecules ([DNA]0) were analyzed by the reaction of bound protein (protein Θ dCas9 The number of "modified" DNA molecules with and without bound proteins (the ratio of DNA bound by proteins) was counted. After tracking the DNA with bound proteins (see above), the average number of bound proteins per DNA molecule (n dCas9 The total dissociation constant, K d,DNA =[DNA][dCas9] / [DNA·dCas9]=(1-Θ dCas9 )([dCas9]0-n dCas9 [DNA]0) / (Θ dCas9 ) was calculated as

[0218] Protospacer-specific dissociation constant K d,プロトスペーサー is the site-specific binding constant K a,ss =K d,ss -1 (Bound dCas9 Θ dCas9,ss Instead of Θ,dCas9,プロトスペーサー , i.e., the proportion of DNA with dCas9 bound within one peak width of a Gaussian fit in its respective binding histogram (i.e., see Table 1).

[0219] Protein alignment and clustering. Images of Cas9 and dCas9 proteins that are isolated and appear to only contact DNA at a single location were extracted. These features were grouped into a set of 134 nm × 134 nm bounding boxes (μ d +2σ d ) (where μ d and σ d are the mean and standard deviation of protein-bound DNA heights); this step essentially had the effect of removing most of the aggregated / clustered Cas9 / dCas9 from the ensemble as well as its proteins from images with large extrinsic noise. After 4-fold nearest neighbor interpolation, (μ d +σ d Protein features with topographical heights greater than 0.01 were aligned by repeated translation, rotation, and inversion to minimize the mean squared difference between their topographical heights, respectively. A distance matrix was constructed from these minimized mean squared differences, and proteins with standard sgRNAs were then clustered according to this criterion using the method of Rodriguez and Laio (27); proteins with guide RNA variants were clustered according to the closest Cas9 / dCas9 structure with a standard sgRNA. Ensemble average structures were extracted by performing a reference-free alignment across each member of the individual clusters according to the method of Penczek, Radermacher, and Frank (28). The properties of the Cas9 / dCas9 population at each feature on the DNA (such as the protospacer site) were determined using proteins bound within one peak width of a Gaussian fit to the binding histogram (i.e., see Table 1).

[0220] Dynamic Monte Carlo (KMC) of guide RNA strand invasion and R-loop "wobble". Dynamic Monte Carlo (KMC) experiments to simulate strand invasion by guide RNA at protospacer sites were performed using a Gillespie-type (continuous time, discrete state) (Gillespie (1976) Journal of computational physics, 22, 403-434) algorithm implemented in MATLAB. Strand invasion is modeled as a one-dimensional random walk in a position-dependent potential determined by the relative nearest-neighbor dependent DNA:DNA and RNA:DNA binding free energies. See, for example, FIG. 4A. That is, the guide RNA is base-paired with the protospacer up to the protospacer site m (1≧m≧20 for sgRNAs and 1≧m≧18 for truncated sgRNAs (tru-gRNAs)) and initially moves at a forward velocity (rate of further guide RNA invasion) v f exp(-(ΔG°(m+1) RNA:DNA -ΔG°(m+1) DNA:DNA ) / 2RT), where R is the Boltzmann constant, T is the temperature (here we used 37°C to match the parameter set), and ΔG°(m+1) RNA:DNA is the free energy of base pairing between the RNA and the protospacer at site m+1, and ΔG°(m+1) DNA:DNA is the free energy of base pairing between the protospacer and its complementary DNA strand. (A correction term of 1 / 2 is included to achieve fine balance.) v f was set to 0. The reverse rate (the rate of rehybridization between the protospacer and its complementary DNA strand) v r is exp(-(ΔG°(m) DNA:DNA -ΔG°(m) RNA:DNA) / 2RT); the simulation is stopped when state m=1 (signifying guide RNA-protospacer dissociation). For each iteration of the algorithm, starting at time t=0 (arbitrary time units), a rate that depends on m is determined, and two random numbers r1 and r2 are derived from a uniform distribution between 0 and 1. t is calculated as Δt=log(r1) / (v f +v r ) and state m is r2 ≧ vf / (v f +v r ), m was increased to m+1, otherwise decreased to m-1. For "equilibrium" measurements of R-loop fluctuations, m started at m=20 (or 18 for tru-gRNA) and the algorithm was iterated until t≧10,000. For measurements of the dynamics of "invasion" kinetics (e.g., in the presence of mismatched base pairs), m started at m=10 (until t=1000).

[0221] Free energy parameters are taken from the literature from experiments in 1M NaCl at 37°C. Sequence-dependent DNA:DNA hybridization free energy ΔG°(x) DNA:DNA is taken from SantaLucia et al. (1996) Biochemistry, 35, 3555-3562; sequence-dependent RNA:DNA hybridization free energy ΔG°(x) RNA:DNA is taken from Sugimoto et al. (1995) Biochemistry, 34, 11211-11216; ΔG°(x) for the introduced point mismatches rG·dG, rC·dC, rA·dA, and rU·dT RNA:DNAValues ​​were obtained from Watkins et al. (2011) Nucleic Acids research, 39, 1894-1902 (under slightly high salt conditions). The sequence of the protospacer used is "ATCCTGTCCCTAGTGGCCCC" (SEQ ID NO: 336), i.e., the AAVS1 target site similar to the AFM experiment; the sequence of the protospacer complementary DNA is "GGGGCCACTAGGGACAGGAT" (SEQ ID NO: 337), and the sequence of the guide RNA is "GGGGGCCACUAGGGACAGGAU" (SEQ ID NO: 338) for sgRNA, or "GGCCACUAGGGACAGGAU" (SEQ ID NO: 339) for truncated RNA.

[0222] Correlation between KMC-derived R-loop stability and experimental Cas9 cleavage rate. To analyze the correlation between guide RNA-protospacer interactions and in vivo Cas9 cleavage rate, we extracted the maximum likelihood estimates (MLEs) of the guide RNA and the targeted DNA sequences by Hsu et al. (2013) Nature biotechnology, 31, 827-832 and their experimentally determined Cas9 cleavage frequencies. The guide RNA and the targeted DNA sequences by Hsu et al. (2013) Nature biotechnology, 31, 827-832 with single nucleotide PAM-distal (≥10bp from the PAM site) mismatches of rG·dG, rC·dC, rA·dA, and rU·dT (n=136) and their experimentally determined maximum likelihood estimates (MLEs) of the Cas9 cleavage frequencies at the sites were imported into the KMC script. Simulations of strand invasion starting at m=10 were repeated 1000 times for each sequence (until t=100) to obtain an average m≧16 of the ratio of times to correlate with the empirical cleavage rate. The average occupancy fraction, along with the MLE cleavage frequency, was bootstrapped through 100,000 permutations and significance was then determined by recalculating the correlation coefficient and p-value. The guide RNA-protospacer binding free energy was estimated by summing the nearest neighbor energies using the parameter set listed above and was found to be −3.1 kcal mol-1 The starting coefficient was adjusted to .

[0223] dCas9-tru-gRNA and dCas9-hp-gRNA data for comparison with dCas9-sgRNA structural properties. When comparing protein height and volume measurements across experiments, AFM imaging conditions should be kept largely consistent so as not to introduce artifacts. This generally does not pose a problem when, for example, comparing the height and volume of dCas9 bound to different sites on a genetically modified DNA molecule, but does pose a challenge when comparing the structural properties of dCas9 / Cas9 when using different guide RNAs or DNA substrates. As a control, the height and volume of the streptavidin protein used to label the ends of the tracked DNA molecules were used. This should remain unchanged across all experimental conditions for the different experiments. For experiments with sgRNA, the average height of streptavidin differed by less than 0.1 nm (mean difference: 0.087 nm; standard deviation of the difference: 0.052 nm) and its average volume (1098 nm 3 ) is 15 nm 3 The mean difference was 14.461 nm. 3 ;Standard deviation of difference: 10.419 nm 3 However, the average height and volume of experiments using tru-gRNA and hp-gRNA were up to 0.14 nm and 225 nm, respectively, higher than those using sgRNA. 3 To directly compare the results of these experiments, the heights of dCas9 with tru-gRNA and hp-gRNA on the genetically modified DNA were shifted by their average height difference relative to that with sgRNA, and the volumes were adjusted by the percentage of the average volume difference.

[0224] Example 2 Atomic force microscopy captures specific and non-specific Cas9 / dCas9 binding along engineered DNA substrates at high resolution Analysis of crystallographic and biochemical experiments suggests that specificity in protospacer binding and cleavage is conferred through initial recognition of the PAM site by Cas9 itself, followed by strand invasion by the bound RNA complex and direct Watson-Crick base pairing with the protospacer (Figure 1A), although the complete mechanistic picture remains to be seen.To directly probe the relative propensities for binding to protospacers and off-target sites with single-molecule resolution, 50 nM Cas9-sgRNA or dCas9-sgRNA complexes targeting the AAVS1 locus on human chromosome 19 were imaged by AFM in air after incubation with one of three DNA substrates (2.5 nM):

[0225] (i) An 1198-bp segment of the AAVS1 locus containing the complete target site after the PAM (herein "TGG") (Figure 1C).

[0226] (ii) A 989 bp engineered DNA substrate (Figure 1D) containing a series of six complete, partial, or mismatched target sites, each separated by approximately 150 bp. Mismatches at these sites could span both the "seed" (PAM-proximal, approximately 12 bp) and "non-seed" (PAM-distal) regions of the protospacer. The only PAM sites in this engineered substrate were at these explicitly designed positions.

[0227] (iii) A 1078 bp "nonsense" DNA with no homology (more than 3 bp sequence) to the target sequence (Figures 9A-9C).

[0228] FIG. 1C shows that dCas9 and Cas9 show nearly identical binding distributions for AAVS1 substrates (n=404 and n=250, respectively). FIG. 1D shows that for genetically engineered substrates (n=536), dCas9 is most likely to bind to complete protospacers with no mismatch (MM) sites (peak 1, hereafter referred to as complete or "0MM" sites) and, with reduced affinity, to sites with 5 or 10 mismatched bases distal to the PAM site (third and fourth features from streptavidin labeling, respectively, hereafter referred to as "5MM" or "10MM" sites). Sites containing a greater number of mismatches (second and fifth features), or with two PAM-proximal mismatched nucleotides (sixth feature), are bound at significantly lower rates. (Below) Distribution of PAM ("TGG") sites in each substrate.

[0229] Structurally, S. pyogenes Cas9 is a 160 kDa monomeric protein, approximately 10 nm × 10 nm × 5 nm (from the crystal structure), roughly divided into two lobe-like halves, each containing a nuclease domain. Consistent with the X-ray structure, dCas9-sgRNA imaged via AFM appears as a large egg structure (Figures 10A-10C), and these structures bound along DNA were observed after incubation of Cas9 or dCas9 with DNA and are assigned to Cas9 or dCas9, respectively (Figure 1B, Figures 10A-10C, and Figures 11A-11D). To unambiguously determine the sequences of the sites bound by Cas9 and dCas9, biotin-labeled DNA molecules were labeled at one end with a monovalent streptavidin tag prior to AFM imaging. Observed DNA molecules with bound Cas9 or dCas9 proteins were selected for further analysis and tracked at subpixel resolution and sites bound by Cas9 / dCas9 were extracted following a modified protocol adapted from that of Wiggins et al. (25) (see Supplementary Methods for details).

[0230] This method proved to be remarkably robust (Table 1): on DNA bound by Cas9 or dCas9, a peculiar enrichment of protein was observed, centered precisely on the location of the protospacer site with the adjacent PAM (within the expected 23 bp, Figure 1C-D), appearing as a sharp peak. No such obvious peak was observed on DNA substrates that did not contain the target site (Figure 9A-9C). The standard deviation of the peak widths ranged from 36 to 60 bp, a significant improvement compared to binding experiments using single-molecule fluorescence, which yielded a peak width standard deviation σ of approximately 1000 bp. The average apparent Cas9 / dCas9 "footprint" on DNA covered 78.1 bp ± 37.9 bp; this broadening of the apparent footprint, beyond the approximately 20 bp footprint of Cas9 on DNA determined by biochemical and crystallographic methods, is a well-established consequence of imaging convolutions with the width of the AFM tip. Previously, it was observed in vitro that Cas9, after putative DNA cleavage as a single-turnover endonuclease, remained bound to the targeted DNA for extended periods (>10 min) and could not be released from the cut strand without strong chemical treatment. Most of the DNA molecules observed with bound Cas9 appeared as full-length AAVS1-derived substrates, with only a small percentage (~5%) of substrates being cleaved and separated. After tracking these DNA molecules, it was observed that Cas9 bound to these "full-length" substrates with a nearly identical distribution as dCas9 (Kolmogorov-Smirnov two-tailed test, significance level 5%) (Figure 1C).

[0231] [Table 1]

[0232] By examining the occupancy of dCas9 bound at different positions along the genetically engineered substrate, we were able to determine the relative binding propensity of dCas9 to various mismatched and partial target sites (Figure 1D, Table 1). The overall dissociation constant between dCas9 and the entire DNA substrate was estimated to be 2.70 nM (± 1.58 nM, 95% confidence, Table 2). In particular, the dCas9 dissociation constant at a perfect (perfectly matched) protospacer site located in the substrate (within one peak width in the binding histogram) was estimated to be 44.67 nM (± 1.04 nM, 95% confidence). Previous electrophoretic mobility shift analysis (EMSA) had estimated dCas9-sgRNA binding to protospacer sites on short DNA molecules (~50 bp) to be 0.5 nM to 2 nM. Although the observed increase in dissociation constants at protospacer sites can be associated with the presence of multiple off-target sites on the genetically engineered DNA substrate, dissociation constants determined by AFM are typically nearly an order of magnitude larger than those determined by conventional assays. (26) This difference is often due to nonspecific interactions of the protein with the shorter blunt ends of DNA that are not accounted for by EMSA.

[0233] [Table 2]

[0234] On genetically engineered substrates, dCas9 is relatively tolerant of distal mismatches (showing a 50-60% binding propensity to the perfect target site, Figure 1D and Table 1), with the same apparent affinity (within confidence) for target sites containing 5 and 10 distal mismatches (MM). However, binding to protospacer sites containing only two PAM-adjacent mismatches occurs with a similar propensity as for sites with 15 or even 20 (PAM site alone) distal mismatches (approximately 5-10% binding propensity to the perfect target, nearly background binding signal), a finding consistent with previous biochemical studies. On genetically engineered substrates, there are no PAM sites other than those adjacent to the protospacer site, whereas on AAVS1-derived substrates there is a distinct "shoulder peak" of enhanced Cas9 and dCas9 binding near the AAVS1 target, where PAM sites are particularly abundant. On the "nonsense" substrate and on segments of the AAVS1-derived substrate distant from the target site, the slight enrichment of dCas9 closely mirrored the distribution of PAM sites (Kolmogorov-Smirnov two-tailed test, significance level 5%), with dCas9 distribution on the "nonsense" substrate more closely reflecting the experimental PAM distribution than 71.20% of 100,000 randomly generated sequences with the same dA, dT, dC, and dG distribution (Figures 9A-9C). dCas9 binding along the "nonsense" substrate (with 879 PAM sites within 1079 bp) corresponds so well with the PAM site distribution that this was interpreted as a measure of actual dCas9-PAM interactions. The average single-site dissociation constant for dCas9 binding along the "nonspecific" substrate was estimated to be approximately 867 nM (standard deviation ± 209 nM). This can be understood as an estimate of the dissociation constant for dCas9 binding on DNA with no protospacer homology.

[0235] Example 3 sgRNAs with two-nucleotide truncations at their 5' ends (tru-gRNAs) do not increase the binding specificity of dCas9 in vitro It has been found that Cas9 still exhibits cleavage activity even when up to four nucleotides of the guide (protospacer-targeting) segment of the sgRNA or crRNA are truncated from their 5'-ends, and Fu et al. (21) recently showed that the use of sgRNAs with these 5'-truncations (optimally of 2-3 nucleotides) can indeed result in an order of magnitude increase in Cas9 cleavage fidelity in vivo. It was suggested that the increased sensitivity to mismatch sites (MMs) using these truncated sgRNAs (termed "tru-gRNAs", Figure 2A) is the result of a decrease in the binding energy between the guide RNA and the protospacer site. This implies that the binding energy conferred by the additional 5'-nucleotides on the sgRNA can compensate for any mismatched nucleotides and stabilize Cas9 at the incorrect site, whereas the tru-gRNA would be relatively less stable on DNA when a mismatch is present.

[0236] As a test of this proposed mechanism, we imaged dCas9 with a tru-gRNA that had a 2-nucleotide 5'-truncation relative to the previously used sgRNA. The dCas9-tru-gRNA complex was incubated with genetically engineered substrates containing a series of complete and partial protospacer sites. Again, a unique peak at the exact protospacer site was observed (Figure 2C and Table 1), but the apparent binding constant compared to dCas9 with a complete sgRNA at this site is significantly reduced (i.e., the dissociation constant is increased, see Table 2). However, off-target binding by dCas9 with a tru-gRNA at a protospacer site with a PAM-distal mismatch, relative to binding at the complete protospacer site, is actually increased when compared to dCas9 with a sgRNA (Figure 2C and Table 1). Similar to dCas9 with sgRNA, dCas with tru-gRNA binds to protospacers with 10 or 5 PAM-distal mismatch sites with roughly equal propensity (note that tru-gRNA is only expected to interact with the first 8 or 3 mismatches at these sites, respectively). These results suggest that the increased cleavage fidelity using tru-gRNA is not necessarily conferred by a relative decrease in binding propensity at off-target sites or a relative decrease in stability in the presence of mismatches; rather, it is approximately 4-5 × 10 6 Since there may be some "threshold" effect where a decrease in the binding constant below M effectively abolishes cleavage activity in vivo, these results and further results provided below suggest that the increased specificity exhibited by tru-gRNA may be influenced by discrimination in the cleavage mechanism itself. Furthermore, these findings would suggest that tru-gRNA can improve the specificity of active Cas9 in cleavage but not in its binding activity for applications using dCas9 (or chimeric derivatives) in vivo.

[0237] Furthermore, previous reports have shown that tru-gRNAs with 5'-truncations (optimally of 2-3 nucleotides) within their protospacer-targeting segments can result in an order of magnitude increase in Cas9 cleavage fidelity in vivo (Figure 2A), and these results shown in the Examples indicate that truncated gRNAs do not increase specificity in dCas9 binding (Figure 2C). Figure 2C shows the binding affinity of dCas9 with standard gRNA (dashed line) compared to that of dCas9 with tru-gRNA (trugRNA, purple line) for DNA molecules containing a complete protospacer (site i) and protospacer sites with 5 and 10 PAM-distal mismatches (sites ii and iii, respectively). Figure 2C shows that the standard guide RNA retains a significant ability to bind to these (mismatch-containing) off-target sites, and that the trugRNA does not show a relative enhancement in binding specificity at sites containing 5-10 nucleotide mismatches at the PAM-distal end of the protospacer. The binding distribution of dCas9 with tru-gRNA shows a unique peak in its affinity at protospacer sites with just 10 PAM-distal mismatches and 5 PAM-distal mismatches, demonstrating that this dCas9 does not have enhanced binding specificity compared to the full sgRNA (see Table 1). The "peaks" in the binding histogram indicate specific and stable binding at these off-target sites. In fact, the binding at off-target sites by dCas9-trugRNA is actually enhanced relative to binding to the protospacer compared to the standard guide RNA. This promiscuous binding may limit its usefulness for dCas9 and chimeric dCas9 derivatives. This may also reflect the off-target cleavage reported for this system, which, while improved compared to standard guide RNAs, was still significant at some off-target sites. For comparison, we found non-specific binding of hpgRNA at these sites with mismatches (Figure 2D).The hpgRNA bound to these sites with approximately the same affinity as it nonspecifically bound to DNA with no homology to the protospacer, and the maximum observed off-target binding affinity was reduced by approximately 22% relative to the tru-gRNA. Furthermore, based on the narrow geometry of the Cas9 binding channel, we predict that the presence of an unopened hairpin at the mismatched protospacer can inhibit the conformational change of Cas9 that is necessary to carry out cleavage (Figure 1B).

[0238] Although considerable efforts have been made to characterize this off-target activity and to improve the specificity of Cas9 / dCas9 through judicious selection of protospacer target sequences; optimization of sgRNA structure, for example by truncation of the first two 5'-nucleotides within the sgRNA; and the use of "dual-nicking" Cas9 enzymes, a clear understanding of the precise mechanism of RNA-guided cleavage in relation to the structural biology of Cas9 will be essential to develop Cas9 derivatives and guide RNAs with increased fidelity for their emerging applications in medicine and biology.

[0239] In accordance with this goal, we use atomic force microscopy (AFM) here to elucidate the individual S. pyogenes Cas9 and dCas9 proteins that bind to targets along genetically modified DNA substrates after incubation with various sgRNA variants. This technique allows us to simultaneously and directly elucidate both the binding sites and structures of individual Cas9 / dCas9 proteins, providing a wealth of mechanistic information about Cas9 / dCas9 specificity with single molecule resolution. Consistent with previous biochemical studies, we find that significant binding by Cas9 / dCas9 with sgRNA occurs at sites containing up to 10 mismatched base pairs within the target sequence. However, since the use of a guide RNA with two nucleotides truncated from its 5' end (tru-gRNA) has previously been shown to result in up to a 5000-fold reduction in off-target mutagenesis by Cas9 in vivo, we find similar specificity for dCas9 with tru-gRNA binding to mismatched targets as standard sgRNA in vitro. We find that the addition of a hairpin to the 5' end of the sgRNA that partially overlaps the target binding region of the guide RNA increases dCas9 specificity, at the expense of a reduced overall binding propensity to DNA. Our results show that when the mismatch is located more than 10 bp away from the PAM, the overall stability of the guide RNA-DNA binding does not necessarily determine the specificity in Cas9 cleavage.

[0240] Example 4 Guide RNAs with 5'-hairpins complementary to the "PAM-distal"-targeting segment (hp-gRNAs) modulate the full binding propensity and profile of dCas9 bound to DNA with mismatched protospacers in vitro dCas9 specificity can be increased by extending the 5' end of the sgRNA to form a hairpin structure that overlaps the "PAM-distal" targeting (or "non-seed") segment of the sgRNA (Figure 2B). After the PAM site is bound and strand invasion of the DNA by the guide RNA is initiated, the hairpin opens when bound to the complete protospacer and full strand invasion can occur. If a PAM-distal mismatch is present at the target site, it is energetically more favorable for the hairpin to remain closed and strand invasion is impeded. Recently, a similar spatial arrangement has been used for "dynamic DNA circuits" driven by strand invasion. In those systems, the hairpin acts as a kinetic barrier to invasion and the rate of oligonucleotide invasion is slowed down by several orders of magnitude compared to invasion attempts with mismatched targets. Here, the hairpin can be displaced during invasion of the complete target site but can inhibit invasion if a mismatch between the target and the non-seed targeting region of the guide RNA is present (Figure 2B). In this case, it is energetically more favorable for the hairpin to remain closed. Compared to previous efforts in which 5'-extensions were added to the sgRNA to complement the additional nucleotides after the protospacer, these guide RNAs did not show increased Cas9 cleavage specificity in vivo. Instead, they were digested back to nearly their standard length in living cells. Based on the size and structure of the hairpin, it can fit into the DNA binding channel of the Cas9 / dCas9 molecule and be protected from degradation.

[0241] We produced sgRNAs (hp-gRNAs) with a 5'-hairpin overlapping the last 6 (hp6-gRNA) or 10 (hp10-gRNA) nucleotides complementary to the PAM distal site of the protospacer. By mapping the observed binding positions of dCas9-hp-gRNA on the genetically modified DNA substrate (Figure 2D), a sharp peak was observed, just at the protospacer site (the PAM and protospacer were located at sites 144-167, and the binding peak was located at site 154.0 (95% confidence: 153.3-154.8) for dCas9-hp6-gRNA and 158.3 (95% confidence: 157.6-158.9) for dCas9-hp10-gRNA). The specificity peaks at sites with 5 and 10 distal mismatches were significantly smoothed, with dCas9 and hp10-gRNA showing substantially reduced affinity for off-target sites (22% reduction relative to dCas9 with tru-gRNA). The affinity peak at the complete protospacer site implies that the hairpin does indeed open upon complete invasion. n=243 for hp6-gRNA and n=212 for hp10-gRNA. dCas9 with hp-gRNA shows a similar reduction in affinity for the target site to that with tru-gRNA, but in contrast to dCas9 with tru-gRNA, dCas9 with hp-gRNA does not exhibit any sharp binding peaks at off-target sites (which would otherwise show strong specific binding). For hp6-gRNA, there was an enrichment of binding around sites of the protospacer with 5 or 10 mismatched PAM distal sites. Because they lack the sharp binding peaks observed with sgRNAs and tru-gRNAs, these enrichments are unlikely to indicate specific binding, but rather may indicate that dCas9 dissociated from these sites upon adsorption to the surface, which in the case of hp6-gRNA would indicate very weak binding at its off-target site.

[0242] In the case of hp10-gRNA, binding to these mismatched sites elsewhere on the substrate was near the level of nonspecific binding, indicating a 22% reduction in the maximum observed off-target binding affinity relative to tru-gRNA (3.18 × 10 of the maximum observed binding constant). 6 M to 2.48 x 10 6 M, Figure 2D). This increase in specificity of the hp10-gRNA is also reflected in a similar binding dissociation constant to the hp6-gRNA for the protospacer site, but a significantly increased overall dissociation constant for all (specific + non-specific) genetically engineered substrate counterparts (Table 2).

[0243] The differential enrichment at just the complete protospacer site suggests that the hairpin in the hp-gRNA is indeed open upon invasion of the complete protospacer site, since nucleotides that bind to the PAM-distal site of the protospacer would be otherwise captured in the hairpin. A putative mechanism for improved binding specificity is that the presence of the hairpin promotes melting of the guide RNA from these off-target sites when not open at protospacer sites with PAM-distal mismatches. These results suggest that hp-gRNAs can be used to tune the affinity and specificity of Cas9 / dCas9 binding, and that further manipulation of hairpin length, loop length, and loop composition may allow finer control of these properties.

[0244] Example 5 Cas9 and dCas9 undergo progressive structural transitions as they bind to DNA sites matching the protospacer sequence Using negative staining transmission electron microscopy (TEM), it was observed that the structure of dCas9 condenses and rotates upon sgRNA binding to open a putative DNA-binding channel between its two lobes. dCas9 undergoes a second structural reorientation to an extended conformation after binding to DNA containing the PAM and protospacer sequences. The role of this second transition was suggested to be related to strand invasion by the sgRNA or to align the two main Cas9 nuclease sites with the two separated DNA strands. However, these studies were only performed with or without DNA containing perfectly matched protospacer sequences, and examining the transitions between these conformations at partially matched protospacer sites could provide insight into the mechanism of off-target binding and cleavage. Thus, in addition to determining the relative binding propensities, AFM imaging was used to capture these putative conformational transitions by Cas9 and dCas9 as they bind to DNA at sites of varying complementarity to the protospacer. We extracted the volumes and maximum topographical heights of Cas9 and dCas9 proteins (n=839) with sgRNAs that appear isolated on DNA, and mapped these values ​​to their respective binding sites on DNA (Figure 3, Figures 11A-11D, and Figures 12A-12B). This binding site distribution was nearly identical to that of the full dataset, indicating that the selection was unbiased and representative. Recorded images of each of these proteins were extracted (Figures 11C-11D) and aligned pairwise by repeated rotation, flip, and translation. Protein structures were clustered according to their pairwise mean squared morphological differences (Figures 12A-12B and Table 3). A notable advantage of this technique is that it clusters any monovalent streptavidin or any aggregated Cas9 / dCas9 proteins that are co-localized on a surface with DNA in the native state separately from those assigned to the respective Cas9 / dCas9 molecules, allowing an unbiased analysis of the structural properties of these proteins on DNA.Analysis of the distribution of binding sites by either putative streptavidin molecules or aggregated proteins reveals that these did not interfere with the analysis of binding site distribution, since they are both rare and uniformly distributed along the DNA (Figures 12A-12B).

[0245] At sites that do not contain homology to the target, such as on "nonsense" DNA substrates, dCas9 molecules with sgRNA were primarily small and ovoid (Figure 3C(iii) and Table 3). However, as the dCas9 protein progressively binds to the complementary target sequence (Figure 3(α-δ)), its height and volume increase significantly compared to nonspecific binding (Figures 3D and 12A-12B, Table 2), reaching a maximum size at the protospacer sequence. This increase is also accompanied by a shift in the population of dCas9 (Figures 3A and 12A-12B, Table 2) from structures that cluster with a flatter, ovoid conformation (Figures 3C(ii) and C3(iii), blue and green) to structures that cluster progressively with slightly rounded structures with a large central bulge (Figure 3C(i), yellow). This latter observed conformation is likely an extended conformation observed previously by TEM and more recently by size-exclusion chromatography, presumably the active state (in which the nuclease domain of Cas9 is properly positioned around the DNA so that cleavage can occur most efficiently).

[0246] Catalytically active Cas9 also undergoes a significant increase in size upon binding to the protospacer sequence (Figure 3(ε)); however, there is a small but statistically significant decrease in size for dCas9, and the conformations of Cas9 at intact protospacer sites tend to cluster with a flatter (green) structure. Because we do not simultaneously observe whether the DNA is cleaved at the time of imaging, it is unclear whether this represents another conformational change following DNA cleavage or is the result of mutational differences between Cas9 and dCas9; however, since binding and strand invasion have previously been determined to be rate-limiting steps, it is likely that the DNA within Cas9 is cleaved during these measurements.

[0247] [Table 3-1] [Table 3-2]

[0248] Example 6 Interactions between the guide RNA and the target DNA at or near position 16 of the protospacer site stabilize the Cas9 / dCas9 conformational change. AFM imaging directly reveals that dCas9 / Cas9 retains a significant tendency to bind protospacer sites with up to 10 distal mismatches, but binding to DNA sites with increasing complementarity to the protospacer drives an increased shift of the population of dCas9 / Cas9 proteins to what appears to be an active conformation. Notably, we also find a similar shift in structure between off-target sites and perfectly matched sites for dCas9 with hp-gRNA (Table 2 and Figure 13). The presence of complementary PAM distal sequences is known to be associated with increased stability of Cas9 on DNA. It has also been recently found that Cas9 binding to single-stranded DNA with increased PAM distal topological complementarity to the protospacer (from 10 to 20 sites) results in an increased change in protein size. This was also further associated with a transition in Cas9 activity from nicking behavior to complete cleavage. Here, we can directly determine the volume of Cas9 / dCas9 bound to double-stranded DNA sites. Analysis of the structural properties of individual Cas9 / dCas9 proteins on double-stranded DNA reveals stable conformational transitions with matching target sequences, consistent with a "conformational gating" mechanism. Here, sgRNA base pairing with these distal sites also stabilizes the active conformation, allowing efficient cleavage to occur, while binding to sites with multiple distal mismatches shifts the equilibrium away from the active conformation (i.e., see Figure 4D).

[0249] Along these lines, we find that this effect is dramatically attenuated for dCas9 with tru-gRNA (Figure 3D and Table 3), with a smaller shift between the structural populations in which the protein clusters (Figure 13). Furthermore, we find a statistically significant difference between the height-volume properties of dCas9-tru-gRNA bound non-specifically and those bound at full or partial protospacer sites (p<0.05; Hotelling's T 2As a result of the 3D-test, the structural features at the protospacer-matching sites (10MM, 5MM, and full protospacer sites) are not statistically distinguishable (Figure 3D and Table 3). Recently, it has been argued that the invasion of the first 10 bp of the protospacer initiates the conformational change of Cas9, whereas the complete invasion of the protospacer by the guide RNA helps drive the further shift to the full active state. Therefore, we hypothesized that the observed suppression of conformational change at the matching protospacer sites for dCas9 with tru-gRNA (versus with sgRNA) is the result of the reduced stability of these guide RNAs at the PAM-distal site.

[0250] To investigate the relative stability of sgRNA and tru-gRNA at these sites, we performed a kinetic Monte Carlo (KMC) study of the dynamic structure of the R-loop - i.e., the structure formed by an invading guide RNA bound to a segment of adjacent DNA exposing a single-stranded loop of the complementary DNA of that segment (Figure 4A) - during and after strand invasion. For further details, see Supplementary Methods. Briefly, using a Gillespie-type algorithm, we calculated the sequence-dependent rates of invasion and reannealing, v, of strand invasion of a guide RNA bound to a protospacer site m, as a sequential, nucleotide-by-nucleotide competition during invasion (breaking of base pairing between the protospacer and its complementary DNA strand, followed by replacement with a protospacer-guide RNA base pair), and v, respectively. f and v r We modeled reannealing (reverse) using exp(-(ΔG°(m+1) RNA:DNA -ΔG°(m+1) DNA:DNA ) / 2RT), the transition rate v from state m to m+1 f Here, ΔG°(m+1) RNA:DNA is the free energy of base pairing between the RNA and the protospacer at site m+1, and ΔG°(m+1) DNA:DNAis the free energy of base pairing between the protospacer and its complementary DNA strand at m+1 (R is the ideal gas constant, T is the temperature, and the term 1 / 2 is added to satisfy fine balance). r is exp(-(ΔG°(m) DNA:DNA -ΔG°(m) RNA:DNA The transition rate is similarly estimated to be proportional to R-loop kinetics (RT) / 2RT. This type of transition rate has been used previously for computational studies of nucleotide base pairing and stability, and here it allows us to capture the general dynamics of R-loops in a sequence-dependent manner.

[0251] In general, RNA:DNA base pairs are energetically stronger than DNA:DNA base pairs, and at equilibrium, we find that the guide RNA is stably bound to the protospacer, as expected from the KMC trajectory (Figure 4C). However, the sgRNA is significantly more stable and remains almost completely invaded - the strand remains invaded at the 19th protospacer site during 95% of the simulated time course (Figure 4B) - while the tru-gRNA shows a significant increase or decrease in protospacer reannealing at the PAM-distal site (Figures 4B and 4C). Since the only difference between dCas9-sgRNA and dCas9-tru-gRNA is a simple truncation of two 5'-nucleotides from the guide RNA, and since we find inhibition of the conformational change by dCas9-sgRNA at sites containing five PAM-distal mismatches, these results suggest that the conformational change to the fully active state is stabilized by an interaction between the guide RNA and the protospacer near the 16th site of the protospacer, which is disrupted by the instability of the tru-gRNA in this region. Indeed, KMC experiments show that the mean lifetime between full invasion and reannealing of the DNA back to the 16th site is reduced by two orders of magnitude when the sgRNA is replaced by the tru-gRNA (Figure 4C inset). This result is consistent with previous findings that Cas9 activity with tru-gRNA variants with 2- or 3-nucleotide (nt) truncations was modulated depending on sequence context, whereas cleavage in all cases tested was dramatically reduced by approximately 90%-100% by 4-nt truncations and was abolished after 5-nt truncations. The conformational change to the protein's active state is stabilized by these interactions at or near protospacer position 16. This finding is supported by gRNA stability at protospacer positions 14-17 (which was inferred from additional KMC experiments described below and correlates with experimental in vivo off-target cleavage (see below)), but not by guide RNA stability at protospacer sites 18-20.

[0252] Example 7 Gain or loss of guide RNA-protospacer R-loop suggests a mechanism for mismatch tolerance by Cas9 / dCas9 and for increased specificity of cleavage by tru-gRNA. To investigate the mechanism by which Cas9 or dCas9 can tolerate or be sensitive to mismatches in the protospacer, we performed a series of KMC experiments using the AAVS1 protospacer site in which one or two PAM-distal (>=10 bp from the PAM) mismatches were introduced (Figure 5). Cas9 generally tolerates PAM-distal mismatches better than PAM-proximal mismatches. However, Hsu et al. (2013) Nature Biotechnology, 31, 827-832 revealed significant and variable differences in the estimated Cas9 cleavage rate at protospacers containing PAM-distal mismatches, depending on the sequence composition, type of mismatch, and site of the mismatch. Based on our AFM and previous KMC experiments, we hypothesized that the difference in cleavage rate may be a result of the differential stability of the guide RNA near the 16th site of the protospacer as well. For these simulations, we only considered sequences with protospacer-guide RNA pairs that would result in isolated rG·dG, rC·dC, rA·dA, and rU·dT mismatches, for which the sequence composition-dependent thermodynamic data are most complete and relevant for our KMC model. The effect of these mismatched base pairs is not expected to dramatically lower the overall binding energy between the sgRNA and the protospacer (Table 4); for example, single rG·dG, rC·dC, rA·dA, and rU·dT mismatches lower the RNA:DNA melting temperature by an average of 1.7°C. Rather, the effect is expected to be kinetic rather than thermodynamic in nature, by impeding strand displacement at the mismatch. Thus, we initiated a kinetic Monte Carlo experiment proceeding from the 10th protospacer site (initial R-loop length m=10), such as what would be occurring during strand invasion.

[0253]

Table 4-1

Table 4-2

Table 4-3

Table 4-4

Table 4-5

Table 4-6

[0254] KMC experiments were then performed to examine the kinetics of strand invasion in the presence of PAM-distal mismatches. In all cases (1000 trials each), the guide RNA remains fairly stably bound even when mismatches are present (i.e., not observed to melt completely) and is often able to quickly bypass these sites to complete full invasion (Figure 5C and Figures 14A-14C), although the average first arrival time of full strand invasion varies considerably, depending on the location of the mismatch site (Figures 14A-14C). The R-loop is fairly stable during invasion, as the sgRNA is often able to remain fully invaded even in the presence of multiple mismatches (Figure 5A). These results are qualitatively similar to those of previous in vitro studies of dCas9 / Cas9 binding and cleavage on mismatched targets. However, in the case of tru-gRNA (Figure 5B), the R-loop is often trapped behind the mismatch site. While the average first arrival times across the mismatch are similar for both sgRNA and tru-gRNA (Figure 14A-14C), examination of the time course for KMC reveals that due to the inherent instability of the R-loop for tru-gRNA, tru-gRNA is often quickly "recaptured" behind the mismatch (Figure 5C). For sgRNA, this recapture is much less frequent. Thus, combined with AFM imaging, the results of the KMC experiments suggest that the cause of increased tru-gRNA specificity is not discrimination during binding, but rather instability of its R-loop, which results in tru-gRNA being repeatedly captured behind the mismatch even after initially bypassing it, making it less likely for Cas9 to adopt an active conformation (Figure 4D). For sgRNA, once the mismatch is bypassed, it can remain fully penetrated with relatively little perturbation, suggesting a mechanism of mismatch tolerance.

[0255] Example 8 The stability of guide RNA interactions with protospacer positions 14-17 correlates with experimental off-target Cas9 cleavage rates, whereas the overall guide RNA-protospacer binding energy does not. To determine whether the stability of the R-loop at or near the 16th position of the protospacer, which was included in the AFM study associated with Cas9 conformational changes, is related to Cas9 activity in vivo, we performed a kinetic Monte Carlo (KMC) analysis of R-loop stability for the sequences used by Hsu et al. (2013) Nature Biotechnology, 31, 827-832. The Hsu et al. (2013) Nature Biotechnology, 31, 827-832 dataset consisted of measurements of cleavage frequency at 15 different protospacer targets containing various point mutations against the guide RNA, performed to investigate cleavage specificity by Cas9. This dataset contains 136 protospacer-guide RNA pairs with single isolated mismatches of the type rG·dG, rC·dC, rA·dA, and rU·dT in the PAM-distal region (Table 4), which we investigated using the KMC method simulating invasion starting with an R-loop size m = 10. Although the inclusion of single mismatch sites from this set only reduced the magnitude of the overall guide RNA-protospacer binding free energy by about 6% on average relative to perfectly matched targets, as noted, there was a broad distribution of Cas9 cleavage frequencies observed for these guide RNA-protospacer pairs whose origin is unclear.

[0256] The average ratio of time that the RNA was stably bound to each site in the protospacer was determined for each guide RNA across 1000 trials, which was then correlated with the maximum estimate of Cas9 cleavage activity (Table 4, Figure 6, and Figure 15). There was a moderate (0.433) but statistically significant (p<1x10) correlation between guide RNA stability at the 16th protospacer position and reported off-target cleavage activity. -6) correlation was found. Of note, no statistically significant correlation was found between cleavage rate and predicted DNA:RNA binding energy alone (0.0786; p=0.3631) (Figures 6A and 6B). In addition to R-loop stability at position 16, there was also a significant correlation between the stability of the 17th protospacer site and reported cleavage (Table 5), although this was not true for sites >= 18 (Figure 6). Since the kinetic Monte Carlo model presented here is based on a relatively simple model of strand invasion, these results further suggest that the stability of the 16-17th protospacer site, and therefore the associated conformational changes we observed, are related to Cas9 cleavage activity in vivo (Figure 4D).

[0257] [Table 5]

[0258] We limited most of our analysis to interactions with nucleotides 16-18 of the protospacer due to the observed structural differences between dCas9 with tru-gRNA and dCas9 with sgRNA. However, we also observed increased strength and statistical significance of the correlation between cleavage and stability of the 14th and 15th protospacer sites (Figure 6), with the correlation at site 14 being the most significant. Because R-loops are dynamic structures (Figure 4D), interactions with these sites may be key ones that are believed to be responsible for DNA cleavage. Truncation of the guide RNA by 4 or 5 nucleotides may abolish cleavage activity by sufficiently destabilizing the R-loop at positions 14 or 15 in much the same manner that tru-gRNA destabilizes the R-loop at sites 16-17. However, in our model, sites 14 and 15 are necessarily invaded whenever site 16 is bound by the sgRNA, so these positions are likely to be more informative because they also correlate more strongly inversely with the probability of sgRNA dissociation from the duplex before bypassing the mismatch site (Figure 6Ai and Figure 16B), another mechanism that would result in failure to produce cleavage. As of now, there is no crystallographic evidence directly linking strand invasion to the observed conformational changes that would permit cleavage. However, based on the evidence provided by the AFM experiments presented here and the results of the kinetic Monte Carlo simulations, we conclude that the stability of the guide gRNA at sites 14-17 of the protospacer during invasion is critical for this conformational change and ultimately the specificity of Cas9 cleavage.

[0259] Furthermore, the R-loop as a dynamic structure competing between strand invasion and DNA reannealing may be useful to understand the mechanism of off-target cleavage and mismatch tolerance. No statistically significant correlation was found between the cleavage rate and the predicted DNA-RNA binding energy alone (Figure 6B), suggesting that the kinetics of strand invasion may be considered when trying to determine Cas9 activity at off-target sites. If 4 or 5 nucleotides are truncated from the guide RNA, cleavage is abolished, but Cas9 is still capable of cleaving DNA with up to 6 distal mismatch sites. Transient non-specific interactions at these PAM-distal sites may be sufficient to stabilize the conformational shift essential for cleavage. We identify a minority population of dCas9-sgRNAs at partial protospacer sites with a structure similar to that at the complete protospacer (yellow, Figure 3C(i)), so this population may represent a fraction of Cas9 in a transiently stabilized active conformation. Therefore, this population may be a source of off-target cleavage.

[0260] While Cas9 / dCas9 binding specificity is largely determined by interactions with the PAM-proximal region, DNA cleavage specificity is likely governed by a conformational change to an activated structure, which is stabilized by guide RNA interactions at the 14-17 bp region of the protospacer (Figure 4D). Kinetic Monte Carlo experiments reveal that the R-loop formed during guide RNA strand invasion can be a significant dynamic structure, even when the guide RNA remains stably bound. This suggests a mechanism for improved specificity of tru-gRNA, and an origin of off-target cleavage via transient stability of the guide RNA-protospacer at critical regions around the mismatch site. The proposed mechanisms for the respective effects of sgRNA variants on Cas9 / dCas9 specificity are summarized in Figure 7.

[0261] Using AFM, it was found that hp-gRNA significantly weakened or eliminated specific binding at homeologous targets. hp-gRNA may be beneficial to modulate dCas9 binding affinity and specificity in its potential applications in biology and medicine. Specifically, based on the narrow geometry of the Cas9 binding channel, the presence of an unopened hairpin at the mismatched protospacer can inhibit the conformational change to an active state by Cas9. The opening of the hairpin in hp-gRNA upon binding could also be used as a binding-dependent signal in vivo, for example, to form a dynamic DNA / RNA structure only upon binding to a specific site.

[0262] Early guide RNA truncation studies raised the question of why the native Cas9 system uses a crRNA targeting the 20 bp protospacer site when only 16 nucleotides of the guide sequence are required for cleavage and additional nucleotides (>18) do not improve cleavage specificity in vivo. These results suggest that the presence of "extra" 5'-nucleotides binding to the 19th and 20th protospacer sites buffers this transient reannealing at the critical 14th-17th sites of the protospacer, allowing efficient conformational change to the active state and subsequent cleavage. Results from AFM and KMC experiments suggest that guide RNA stability at these sites shifts the equilibrium structure of Cas9 to the active conformation upon full invasion (Figure 4A), whereas R-loop instability for the "truncated" guide RNA reduces the force of shifting the equilibrium to the active state. The promiscuous activity of Cas9 with sgRNA versus Cas9 with tru-gRNA would also hold an evolutionary advantage in its role as an agent of adaptive immunity against invading DNA in prokaryotes, since the DNA of the invading phage would undergo rapid point mutations at the site targeted by Cas9 to avoid cleavage.

[0263] The design of guide RNA sequences for in vivo Cas9 / dCas9 applications has primarily focused on avoiding targeting multiple sites with similar sequences in the genome. However, recent studies exploring off-target cleavage have found that current methods for predicting off-target activity are largely inefficient. R-loop stability during invasion correlates significantly better with off-target cleavage rates than guide RNA-protospacer binding energy alone or mismatched positions (another important criterion used in guide RNA design, Table 3). R-loop stability at shorter times after the start of invasion correlates much better with experimental cleavage rates than long-term stability in KMC experiments (Figure 16A), suggesting that strand invasion kinetics is a factor in off-target activity prediction.

[0264] Example 9 In vivo testing To investigate dCas9 binding specificity, the optimized gRNA activity was tested in live cells. Several hairpin gRNAs (hp-gRNAs) were designed for each of the four target locations (protospacers) in the human genome (Figures 17 and 18). One was in the dystrophin gene (Figures 19-23), another was in the EMX1 gene (Figures 24-29 and 44), and two targets were in the VEGFA gene, labeled VEGFA1 (Figures 30-37) and VEGFA3 (Figures 38-43). All experiments were performed in HEK293T cells.

[0265] To modulate the binding and cleavage activity of Cas9 to the protospacer, additional nucleotides (nt) were added to the 5' end of the complete guide RNA (gRNA, total length 20 nt) and designed to form hairpins and secondary structures by hybridizing to the 5'-protospacer-targeting nucleotides, or to nucleotides in the middle or at the 3' end of the protospacer-targeting region.

[0266] Certain secondary structures of VEGFA1-targeting hp-gRNAs were computationally designed using the methods described herein to prevent binding at known off-target sites while allowing binding to the complete protospacer (Figures 44A-44C). The hp-gRNAs were selected to have binding lifetimes equal to or greater than that of the complete gRNA at the on-target site and less than or equal to that of the complete gRNA at the top three off-target sites. Other 5'-structures were designed to contain dG-rU wobble pairs to modulate the energetic properties of the hp-gRNA secondary structure or were added to the end of truncated gRNAs (tru-gRNAs, <20 nt) that have been shown to promote higher specificity of Cas9 activity by themselves.

[0267] Cell studies. For deep sequencing analysis, 293T cells were transfected with plasmids expressing Cas9 and gRNA of interest. The cells were incubated for 4 days to allow Cas9 and gRNA to exert their maximum activity. The cells were then harvested and their genomic DNA was purified. A gRNA that is very well characterized in the literature (i.e., its on-target and off-target sites are known) was used.

[0268] Surveyor assay. Compared with deep sequencing, the surveyor assay has less throughput and is less sensitive. However, the surveyor assay is faster to analyze data and less technical, and provides gel images. Therefore, the surveyor was performed as a first step, and the best conditions were analyzed in triplicate using deep sequencing. Both deep sequencing and Surveyor are methods for quantifying mutation events caused by Cas9+gRNA.

[0269] The cell study on Surveyor was the same as described above. After purifying the genomic DNA, primers were designed to amplify the targeted sites. In this experiment, a pool of 200k cells was used, each of which had a different mutation due to the stochastic nature of DNA repair. Sites across 200k cells were amplified to produce heterogeneous PCR products: some amplicons had deletions, some had insertions, and some were wild type and unmodified due to the stochastic (i.e., random, error-prone) repair of the Cas9 cut sites in each cell.

[0270] This heterogeneous PCR pool is heated and repaired, and in some cases, the different strands are annealed to each other: the wild type DNA strand may bind to the DNA with the insertion, or the insertion may bind to the deletion. When this happens, a small "bubble" is formed, and this structure is called a DNA heteroduplex (see Figure 46).

[0271] These heteroduplexes were detected by digesting and cleaving them using surveyor nuclease. DNA cleavage was then a surrogate for the mutational activity of Cas9. PCR pools were separated on a gel and the intensity of these digested bands was used to quantify the percentage of Cas9 activity.

[0272] Deep Sequencing. Primers were designed to amplify these known targets / off-targets. High fidelity polymerase was used for this PCR. Illumina adapters were also present on these primers to allow for barcoding and loading into the Illumina Mi-Seq platform. Hairpin #, target #, off-target #, sequencing coverage, etc. are listed in the figures and brief figure descriptions. Excellent coverage was obtained across samples used in the analysis. The average number of reads / sample was 20,000. The sample with the lowest # of reads was 1,700. Very few targets did not produce well-aligned reads and were not included in the analysis.

[0273] The resulting sequencing data was analyzed using CRISPResso software (Pinello et al. Nat Biotechnol. (2016) 34(7):695-697), which aligns deep sequencing reads with specific sites of known off-target or on-target positions. The results of this software were compared with an in-house script, where a global alignment of deep sequencing reads with the human genome was performed, and correlated very well. Mutation rates were quantified using CRISPResso, and the resulting data are shown in the indicated histograms for each target gene.

[0274] The designs were first tested using the Surveyor assay to test for indels following Cas9 and hp-gRNA expression in HEK cells at target and off-target sites known to be targeted using standard gRNAs (see Table 6). Activity at these sites was compared to standard gRNAs and truncated gRNAs (tru-gRNAs). These are shown below as gels showing Surveyor nuclease cleavage of PCR-treated genomic DNA. Here, cleavage indicates mutagenesis by Cas9.

[0275] [Table 6]

[0276] The most promising hp-gRNA designs were selected for additional quantitative analysis using next-generation sequencing to evaluate Cas9 activity at on- and off-target sites in HEK cells. Specificity was defined as on-target hits / total (off-target hits).

[0277] Although Cas9 activity was generally equal or slightly reduced when using hp-gRNAs, each of the hp-gRNAs selected for deep sequencing experiments showed enhanced specificity over the full gRNAs and in most cases was equal to or greater than the tru-gRNAs in terms of specificity.

[0278] In one case, hp-gRNA hairpin targeting EMX1 showed >6000-fold improvement in specificity versus full gRNA (versus tru-gRNA with ≈100-fold improvement versus gRNA). VEGFA1-targeting hp-gRNA with computationally designed secondary structure using an in-house algorithm greatly outperformed tru-gRNA activity in terms of specificity (18-fold versus 3-fold improvement over gRNA). These hp-gRNAs were tested with S. pyogenes Cas9. Figures 44A-44C show Surveyor assays of EMX1-targeting hp-gRNA with Cas9 from S. aureous showing on-target activity, whereas no off-target activity was detectable, in contrast to tru-gRNA showing significant off-target activity.

[0279] Example 10 hp-gRNA for CRISPR / Cpf1 systems The experiment was designed to replicate the results of Kleinstiver et al., Nat. Biotech. (2016) 34:869-874. Kleinstiver et al. showed that Lachnospiraceae Cpf1 is susceptible to cleavage at off-target sites with mismatches at 8-9 nucleotides in addition to the PAM-distal site by using full-length gRNAs and gRNAs with mismatches at different positions with the target site (Figure 47). In this example, hairpin guide RNAs used with the V-type CRISPR-Cas system CRISPR-Cpf1 were designed and tested as described above using the methods of the present invention.

[0280] To test the off-target activity of Cpf1 with and without additional secondary structure elements, we cloned the DNMT1 gene [ka] was targeted for cleavage by Cpf1. "Off-target activity" can be observed when using a guide RNA with a mismatched nucleotide at position 9, e.g., a full-length guide RNA (20 nucleotides long). [ka] , or truncated gRNA (17 nucleotides long) [ka] A 9-nucleotide long secondary structure element was added to the 3' end of the Cpf1 guide RNA and hybridized to a segment of mismatched nucleotides surrounding the guide RNA. In this case, the "linker" element was a four 3'-nt protospacer targeting segment, i.e. [ka] and [ka] Surveyor assays show that inclusion of these additional 3'-elements reduces or eliminates off-target activity at the DNMT1 site exhibited by full or truncated gRNAs.

[0281] The hp-gRNA was designed using an "internal" hairpin design in which the PAM distal 4 nucleotides act as a loop. The hairpin was added to the 3' end of the gRNA. Table 7 shows the sequence of the hp-gRNA with spaces in the sequence separating the regions. Mismatches are shown in lower case.

[0282] The Surveyor results of these hp-gRNAs are shown in Figure 48, which shows that the addition of a hairpin to the 3' end eliminated off-target activity. Lane 1 shows the control; lane 2 shows the full-length gRNA containing a mismatched nucleotide at position 9; lane 3 shows the full-length gRNA containing a mismatched nucleotide at position 9 and an additional 3'-hairpin structure; lane 4 shows the truncated gRNA containing a mismatched nucleotide at position 9; lane 5 shows the truncated gRNA containing a mismatched nucleotide at position 9 and an additional 3'-hairpin structure. The Surveyor primers used are also shown in Table 7.

[0283] Cpf1 tolerates mismatches at nucleotides 8-10 when using normal guide RNA and cleaves DNA at its off-target site (Figure 47). As shown in Figure 48, Cpf1 hp-gRNA was able to abolish the off-target activity exhibited by Kleinstiver, whereas truncated gRNAs were not.

[0284] [Table 7]

[0285] It is understood that the foregoing detailed description and the accompanying examples are merely illustrative and should not be taken as limitations on the scope of the invention, which is defined solely by the appended claims and equivalents thereof.

[0286] Various changes and modifications to the disclosed embodiments will be apparent to those skilled in the art. Such changes and modifications, including but not limited to, with respect to the chemical structures, substituents, derivatives, intermediates, compounds, compositions, formulations, or methods of use of the invention, can be made without departing from the spirit and scope thereof.

[0287] For reasons of completeness, the various aspects of the invention are presented in the following numbered clauses:

[0288] Section 1. A method of producing an optimized guide RNA (gRNA), comprising the steps of: a) identifying a target region of interest, the target region of interest comprising a protospacer sequence; b) determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; and d) producing a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment being at a 5' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprising a linker between the 5' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker being an N nucleotide. a polynucleotide sequence having a length of 1 nucleotide, wherein the first gRNA is capable of invading a protospacer sequence and binding to a DNA sequence complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading an off-target site and binding to a DNA sequence complementary to the off-target site to form an off-target duplex; e) calculating or computationally simulating an estimate of the invasion kinetics and the lifetime that the first gRNA remains invaded into the protospacer and off-target site duplex, wherein the invasion kinetics is estimated on a nucleotide-by-nucleotide basis by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site with the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site;g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with the second gRNA; h) identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and i) testing the optimized gRNA in vivo to determine binding specificity.

[0289] Section 2. A method of producing an optimized guide RNA (gRNA), comprising the steps of: a) identifying a target region of interest, the target region of interest comprising a protospacer sequence; b) determining a polynucleotide sequence of a full-length gRNA that targets the target region of interest, the full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; and d) producing a polynucleotide sequence of a first gRNA, the first gRNA comprising a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprising a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment being a 3' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprising a linker between the 3' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker being an N nucleotide. a polynucleotide sequence having a length of 100 nucleotides, wherein the first gRNA is capable of invading a protospacer sequence and binding to a DNA sequence complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading an off-target site and binding to a DNA sequence complementary to the off-target site to form an off-target duplex; e) calculating or computationally simulating an estimate of the invasion kinetics and the lifetime that the first gRNA remains invaded within the protospacer and off-target site duplex, wherein the invasion kinetics is estimated nucleotide by nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site with the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site;g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the gRNA to produce a second gRNA and repeating step (e) with the second gRNA; h) identifying an optimized gRNA based on a gRNA sequence that meets the design criteria; and i) testing the optimized gRNA in vivo to determine binding specificity.

[0290] Clause 3. The method of clause 1 or 2, wherein the energetics of further invasion of the different gRNA are determined by determining the energetics of at least one of: (I) breaking DNA-DNA base pairing; (II) forming RNA-DNA base pairing; (III) the energy difference resulting from breaking or forming a different secondary structure in the uninvaded guide RNA; and (IV) forming or breaking an interaction between the displaced DNA strand, which is complementary to the protospacer, and any unpaired guide RNA nucleotides that are not included in the secondary structure.

[0291] Clause 4. The method of any one of clauses 1-3, wherein the energetics of reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence are determined by determining the energetics of at least one of: (I) forming DNA-DNA base pairs; (II) breaking RNA-DNA base pairs; (III) the energy difference resulting from breaking or forming a different secondary structure in the newly invaded guide RNA; and (IV) forming or breaking an interaction between the displaced DNA strand that is complementary to the protospacer and any unpaired guide RNA nucleotides that are not included in the secondary structure.

[0292] Clause 5. The method of clause 3 or 4, further comprising determining energetic considerations from at least one of: (V) base pairing across the mismatch; (VI) interaction with the Cas9 protein; and / or (VII) additional heuristics (wherein the additional heuristics relate to binding lifetime, degree of invasion, stability of the invading guide RNA, or other calculated / simulated properties of gRNA invasion versus Cas9 cleavage activity).

[0293] Item 6. The method of any one of items 1 to 5, wherein the full-length gRNA comprises about 15 to 20 nucleotides.

[0294] Item 7. The method of any one of items 1 to 5, wherein M is 1 to 20.

[0295] Item 8. The method of item 7, wherein M is 4 to 10.

[0296] Item 9. The method of any one of items 1 to 8, wherein the RNA segment comprises 2 to 15 nucleotides complementary to a protospacer-targeting sequence.

[0297] Item 10. The method of any one of items 1 to 9, wherein N is 1 to 20.

[0298] Item 11. The method of item 10, wherein N is 3 to 10.

[0299] Clause 12. The method of any one of clauses 1 to 11, wherein the RNA segment and / or the protospacer-targeting sequence provide a secondary structure.

[0300] Clause 13. The method of clause 12, wherein the secondary structure is formed by partially hybridizing the protospacer-targeting sequence to the RNA segment.

[0301] Clause 14. The method of clause 13, wherein the secondary structure modulates DNA binding or cleavage by Cas9 by preventing invasion of a protospacer duplex or an off-target duplex by the optimized gRNA.

[0302] Clause 15. The method of any one of clauses 12 to 14, wherein the secondary structure is formed by hybridizing all or a portion of the RNA segment to the 5'-terminal nucleotide of the protospacer-targeting sequence or segment, the central nucleotide of the protospacer-targeting sequence or segment, and / or the 3'-terminal nucleotide of the protospacer-targeting sequence or segment.

[0303] Item 16. The method according to any one of items 12 to 15, wherein the secondary structure is a hairpin.

[0304] Item 17. The method according to any one of items 12 to 16, wherein the secondary structure is stable at room temperature or 37°C.

[0305] Item 18. The method of any one of items 12 to 17, wherein the total equilibrium free energy of the secondary structure is less than about 2 kcal / mol at room temperature or 37°C.

[0306] Clause 19. The method of any one of clauses 1 to 18, wherein the RNA segment hybridizes to or forms non-canonical base pairs with at least two nucleotides of the protospacer-targeting sequence or segment.

[0307] Item 20. The method of item 19, wherein the non-canonical base pair is rU-rG.

[0308] Clause 21. The method of any one of clauses 1 to 20, wherein the optimized gRNA is used in a cell with a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system.

[0309] Clause 22. The method of any one of clauses 1 to 21, wherein the secondary structure protects the optimized gRNA in a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system to prevent degradation within the cell.

[0310] Item 23. The method of any one of items 1 to 22, wherein 1 to 20 nucleotides are randomized within the linker.

[0311] Item 24. The method of any one of items 1 to 23, wherein 1 to 20 nucleotides are randomized within the RNA segment.

[0312] Item 25. The method of any one of items 1 to 24, wherein step (g) is repeated X times, thereby producing X gRNAs, and step (e) is repeated with each of the X gRNAs, where X is 0 to 20.

[0313] Item 26. The method of any one of items 1 to 25, wherein the entry kinetics and lifetime are calculated using a kinetic Monte Carlo method or a Gillespie algorithm.

[0314] Clause 27. The method of any one of clauses 1 to 26, wherein the invasion kinetics is the rate at which the guide RNA invades the protospacer duplex until complete invasion such that the protospacer is fully invaded, and / or the rate at which a segment of protospacer DNA bound to gRNA is displaced from its complementary strand and extended as it binds to gRNA, nucleotide by nucleotide, from its PAM-proximal region, until complete invasion.

[0315] Clause 28. The method of any one of clauses 1 to 27, wherein the design criteria include specificity, modulation of binding lifetime, and / or predicted cleavage specificity.

[0316] Clause 29. The method of clause 28, wherein the design criteria include an optimized gRNA having a binding lifetime equal to or greater than the binding lifetime of a full-length gRNA to an on-target site and / or equal to or less than the binding lifetime of a full-length gRNA to an off-target site.

[0317] Clause 30. The method of clause 29, wherein the design criteria include an optimized gRNA that has a binding lifetime equal to or less than the binding lifetime of the full-length gRNA to at least three off-target sites, where these off-target sites are predicted to be the closest off-target sites or have the highest identity to the on-target site.

[0318] Clause 31. The method of clause 28, wherein the design criteria include a lifetime or cleavage rate at the off-target site that is equal to or less than the lifetime or cleavage rate of the full-length or truncated gRNA at the off-target site, and / or a predicted rate of on-target activity that is greater than 10% of the predicted rate of on-target activity of the full-length or truncated gRNA.

[0319] Clause 32. The method according to any one of clauses 1 to 31, wherein the optimized gRNA is tested in step i) using a surveyor assay, a next-generation sequencing technology, or GUIDE-Seq.

[0320] Clause 33. The method of any one of clauses 1 to 32, wherein the optimized gRNA is designed to minimize binding at off-target sites and to bind to a protospacer sequence.

[0321] Item 34. The method of any one of items 1 to 33, wherein the off-target site is a known or predicted off-target site.

[0322] Item 35. The method of any one of items 1 to 34, wherein the full-length gRNA targets a mammalian gene.

[0323] Item 36. The method of any one of items 1 to 35, wherein the target gene comprises an endogenous target gene or an introduced gene.

[0324] Item 37. The method of any one of items 1 to 36, wherein the target gene comprises a disease-associated gene.

[0325] Item 38. The method of any one of items 1 to 37, wherein the target gene is a DMD, EMX1, or VEGFA gene.

[0326] Item 39. The VEGFA gene is VEGFA1 or VEGFA3. 39. The method according to paragraph 38.

[0327] Item 40. An optimized gRNA produced by the method according to any one of items 1 to 39.

[0328] Clause 41. The optimized gRNA of clause 40, wherein the gRNA can discriminate between on-target and off-target sites using a minimum thermodynamic energy difference between the sites.

[0329] Item 42. The optimized gRNA of item 40 or 41, wherein the optimized gRNA regulates strand invasion into a protospacer.

[0330] Clause 43. The optimized gRNA described in any one of clauses 40 to 42, wherein the optimized gRNA comprises at least one nucleotide sequence of SEQ ID NOs: 149 to 315, 321 to 323, and 326 to 329.

[0331] Item 44. An isolated polynucleotide encoding the optimized gRNA described in any one of items 40 to 43.

[0332] 45. A vector comprising the isolated polynucleotide of claim 44.

[0333] Item 46. A cell comprising the isolated polynucleotide of item 44 or the vector of item 45.

[0334] Item 47. A kit comprising the isolated polynucleotide of item 44, the vector of item 45, or the cell of item 46.

[0335] Clause 48. A method for epigenome editing in a target cell or subject, comprising contacting the cell or subject with an effective amount of an optimized gRNA molecule according to any one of clauses 40 to 43 or the isolated polynucleotide and fusion protein of clause 44, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0336] Clause 49. A method for site-specific DNA cleavage in a target cell or subject, comprising contacting the cell or subject with an effective amount of an optimized gRNA molecule according to any one of clauses 40 to 43 or an isolated polynucleotide and fusion protein or Cas9 protein according to clause 44, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0337] Clause 50. A method for genome editing in a cell, comprising administering to said cell an effective amount of an optimized gRNA molecule according to any one of clauses 40 to 43 or an isolated polynucleotide and fusion protein according to clause 44, wherein said fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0338] 51. The method of claim 50, wherein the genome editing comprises correcting a mutated gene or inserting a transgene.

[0339] 52. The method of claim 51, wherein correcting the mutated gene comprises deleting, rearranging, or replacing the mutated gene.

[0340] Clause 53. The method of any one of clauses 51 or 52, wherein correcting the mutated gene comprises nuclease-mediated non-homologous end joining or homology-directed repair.

[0341] Clause 54. A method for regulating gene expression in a cell, comprising contacting said cell with an effective amount of an optimized gRNA molecule according to any one of clauses 40-43 or the isolated polynucleotide and fusion protein of clause 44, wherein said fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

[0342] Clause 55. The method of clause 54, wherein gene expression of at least one target gene is modulated when the gene expression level of said at least one target gene is increased or decreased compared to a normal gene expression level for said at least one target gene.

[0343] 56. The method of claim 54 or 55, wherein the fusion protein comprises a dCas9 domain and a transcriptional activator.

[0344] Item 57. The method of item 56, wherein the fusion protein comprises the amino acid sequence of SEQ ID NO:2.

[0345] 58. The method of claim 54 or 55, wherein the fusion protein comprises a dCas9 domain and a transcriptional repressor.

[0346] 59. The method of claim 58, wherein the fusion protein comprises the amino acid sequence of SEQ ID NO:3.

[0347] 60. The method of claim 54 or 55, wherein the fusion protein comprises a dCas9 domain and a site-specific nuclease.

[0348] Item 61. The method of any one of items 48 to 60, wherein the optimized gRNA is encoded by a polynucleotide sequence and packaged into a lentiviral vector.

[0349] 62. The method of claim 61, wherein the lentiviral vector comprises an expression cassette comprising a promoter operably linked to a polynucleotide sequence encoding the gRNA.

[0350] Item 63. The method of item 62, wherein the promoter operably linked to the polynucleotide encoding the optimized gRNA is inducible.

[0351] Clause 64. The method of any one of clauses 61 to 63, wherein the lentiviral vector further comprises a polynucleotide sequence encoding the Cas9 protein or fusion protein.

[0352] Item 65. The method of any one of items 48 to 64, wherein the at least one target gene is a disease-associated gene.

[0353] Item 66. The method of any one of items 48 to 65, wherein the target cell is a eukaryotic cell.

[0354] Item 67. The method of any one of items 48 to 66, wherein the target cell is a mammalian cell.

[0355] 68. The method according to any one of items 48 to 67, wherein the target cell is a HEK293T cell.

[0356] Appendix-Sequence Streptococcus pyogenes Cas 9 (D10A, with H840A) (SEQ ID NO: 1) [ka] dCas9 p300 Core : (Addgene Plasmid 61357) Amino acid sequence; 3X "Flag" epitope, nuclear localization sequence, Streptococcus pyogenes Cas9 (D10A, H840A), p300 Core effector, "HA" epitope (SEQ ID NO: 2) [ka] dCas9 KRAB (SEQ ID NO:3) [ka] Nm-dCas9 p300 Core: (Addgene Plasmid 61365) Amino acid sequence; Neisseria meningitidis Cas9 (D16A, D587A, H588A, N611A), nuclear localization sequence, p300 Core effector, "HA" epitope (SEQ ID NO:5) [ka]

Claims

1. 1. A method for producing an optimized guide RNA (gRNA), comprising: a) identifying a target region of interest, said target region of interest comprising a protospacer sequence; b) determining the polynucleotide sequence of a full-length gRNA that targets a target region of interest, said full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; d) producing a polynucleotide sequence of a first gRNA, wherein the first gRNA comprises a polynucleotide sequence of a full-length gRNA and an RNA segment, wherein the RNA segment comprises a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, wherein the RNA segment is the 5' end of the polynucleotide sequence of the full-length gRNA, wherein the first gRNA optionally comprises a linker between the 5' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, wherein the linker comprises a polynucleotide sequence having a length of N nucleotides, wherein the first gRNA is capable of invading the protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and wherein the first gRNA is capable of invading an off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; e) estimating or computationally simulating the invasion kinetics and the lifetime that the first gRNA remains invaded into the protospacer and off-target site duplex, where the invasion kinetics is estimated nucleotide-by-nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site to the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site; g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with said second gRNA; h) identifying optimized gRNAs based on gRNA sequences that meet the design criteria; i) testing the optimized gRNA in vivo to determine specificity of binding; The method includes:

2. 1. A method for producing an optimized guide RNA (gRNA), comprising: a) identifying a target region of interest, said target region of interest comprising a protospacer sequence; b) determining the polynucleotide sequence of a full-length gRNA that targets a target region of interest, said full-length gRNA comprising a protospacer-targeting sequence or segment; c) determining at least one or more off-target sites for the full-length gRNA; d) producing a polynucleotide sequence of a first gRNA, wherein the first gRNA comprises a polynucleotide sequence of a full-length gRNA and an RNA segment, the RNA segment comprises a polynucleotide sequence having a length of M nucleotides that is complementary to a nucleotide segment of a protospacer-targeting sequence or segment, the RNA segment is the 3' end of the polynucleotide sequence of the full-length gRNA, the first gRNA optionally comprises a linker between the 3' end of the polynucleotide sequence of the full-length gRNA and the RNA segment, the linker comprises a polynucleotide sequence having a length of N nucleotides, the first gRNA is capable of invading the protospacer sequence and binding to a DNA sequence that is complementary to the protospacer sequence to form a protospacer-duplex, and the first gRNA is capable of invading an off-target site and binding to a DNA sequence that is complementary to the off-target site to form an off-target duplex; e) estimating or computationally simulating the invasion kinetics and the lifetime that the first gRNA remains invaded into the protospacer and off-target site duplex, where the invasion kinetics is estimated nucleotide-by-nucleotide by determining the energy difference between further invasion of a different gRNA and reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence; f) comparing the estimated lifetime of the first gRNA at the protospacer and / or off-target site to the estimated lifetime of a full-length gRNA or a truncated gRNA (tru-gRNA) at the protospacer and / or off-target site; g) randomizing 0 to N nucleotides in the linker and 0 to M nucleotides in the first gRNA to produce a second gRNA and repeating step (e) with said second gRNA; h) identifying optimized gRNAs based on gRNA sequences that meet the design criteria; i) testing the optimized gRNA in vivo to determine specificity of binding; The method includes:

3. 3. The method of claim 1 or 2, wherein the energetics of the further invasion of the different gRNA are determined by determining the energetics of at least one of: (I) breaking DNA-DNA base pairing, (II) forming RNA-DNA base pairing, (III) the energy difference resulting from breaking or forming a different secondary structure in the uninvaded guide RNA, and (IV) making or breaking an interaction between the displaced DNA strand, which is complementary to the protospacer, and any unpaired guide RNA nucleotides that are not included in the secondary structure.

4. 4. The method of claim 3, wherein the energetics of reannealing of the first gRNA to a DNA sequence that is complementary to the protospacer sequence are determined by determining the energetics of at least one of: (I) forming DNA-DNA base pairs; (II) breaking RNA-DNA base pairs; (III) the energy difference resulting from breaking or forming a different secondary structure in the newly invaded guide RNA; and (IV) making or breaking interactions between the displaced DNA strand that is complementary to the protospacer and any unpaired guide RNA nucleotides that are not included in the secondary structure.

5. 5. The method of claim 4, further comprising determining energetic considerations from at least one of: (V) base pairing across the mismatch; (VI) interaction with the Cas9 protein; and / or (VII) additional heuristics relating to binding lifetime, degree of invasion, stability of the invading guide RNA, or other calculated / simulated properties of gRNA invasion versus Cas9 cleavage activity.

6. 3. The method of claim 1 or 2, wherein the full-length gRNA comprises about 15 to 20 nucleotides.

7. 3. The method of claim 1 or 2, wherein M is 1 to 20.

8. The method of claim 7, wherein M is from 4 to 10.

9. 3. The method of claim 1 or 2, wherein the RNA segment comprises 2 to 15 nucleotides complementary to a protospacer-targeting sequence.

10. 3. The method of claim 1 or 2, wherein N is 1 to 20.

11. 11. The method of claim 10, wherein N is 3 to 10.

12. The method of claim 1 or 2, wherein the RNA segment and / or the protospacer-targeting sequence provides a secondary structure.

13. The method of claim 12, wherein the secondary structure is formed by partially hybridizing the protospacer-targeting sequence to the RNA segment.

14. The method of claim 13, wherein the secondary structure modulates DNA binding or cleavage by Cas9 by preventing invasion of a protospacer duplex or an off-target duplex by the optimized gRNA.

15. 13. The method of claim 12, wherein the secondary structure is formed by hybridizing all or a portion of the RNA segment to the 5'-terminal nucleotide of a protospacer-targeting sequence or segment, to the central nucleotide of a protospacer-targeting sequence or segment, and / or to the 3'-terminal nucleotide of a protospacer-targeting sequence or segment.

16. The method of claim 12, wherein the secondary structure is a hairpin.

17. The method of claim 12, wherein the secondary structure is stable at room temperature or 37°C.

18. 13. The method of claim 12, wherein the total equilibrium free energy of the secondary structure is less than about 2 kcal / mol at room temperature or 37°C.

19. 13. The method of claim 12, wherein the RNA segment hybridizes to or forms non-canonical base pairs with at least two nucleotides of the protospacer-targeting sequence or segment.

20. 20. The method of claim 19, wherein the non-canonical base pair is rU-rG.

21. The method of claim 1 or 2, wherein the optimized gRNA is used in a cell with a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system.

22. The method of claim 1 or 2, wherein the secondary structure protects the optimized gRNA in a CRISPR / Cas9-based system or a CRISPR / Cpf1-based system to prevent degradation within the cell.

23. The method of claim 1 or 2, wherein 1 to 20 nucleotides are randomized within the linker.

24. The method of claim 1 or 2, wherein 1 to 20 nucleotides are randomized within the RNA segment.

25. 3. The method of claim 1 or 2, wherein step (g) is repeated X times, thereby producing X gRNAs, and step (e) is repeated with each of the X gRNAs, where X is from 0 to 20.

26. 3. The method of claim 1 or 2, wherein the invasion kinetics and lifetimes are calculated using a kinetic Monte Carlo method or a Gillespie algorithm.

27. 3. The method of claim 1 or 2, wherein the invasion kinetics is the rate at which the guide RNA invades the protospacer duplex until complete invasion such that the protospacer is fully invaded, and / or the rate at which a segment of protospacer DNA bound to gRNA is displaced from its complementary strand and extends from its PAM-proximal region as it binds to gRNA, nucleotide by nucleotide, until complete invasion.

28. The method of claim 1 or 2, wherein the design criteria include specificity, modulation of binding lifetime, and / or predicted cleavage specificity.

29. 29. The method of Claim 28, wherein the design criteria include an optimized gRNA that has a binding lifetime equal to or greater than the binding lifetime of a full-length gRNA to an on-target site and / or equal to or less than the binding lifetime of a full-length gRNA to an off-target site.

30. 30. The method of Claim 29, wherein the design criteria include an optimized gRNA that has a binding lifetime less than or equal to the binding lifetime of the full-length gRNA to at least three off-target sites that are predicted to be the closest off-target site or that are predicted to have the highest identity to the on-target site.

31. 29. The method of Claim 28, wherein the design criteria comprise a lifetime or cleavage rate at the off-target site that is equal to or less than the lifetime or cleavage rate of the full-length or truncated gRNA at the off-target site, and / or a predicted rate of on-target activity that is greater than 10% of the predicted rate of on-target activity of the full-length or truncated gRNA.

32. The method of claim 1 or 2, wherein the optimized gRNA is tested in step i) using a surveyor assay, a next-generation sequencing technology, or GUIDE-Seq.

33. 3. The method of claim 1 or 2, wherein the optimized gRNA is designed to minimize binding at off-target sites and to bind to a protospacer sequence.

34. 3. The method of claim 1 or 2, wherein the off-target site is a known or predicted off-target site.

35. The method of claim 1 or 2, wherein the full-length gRNA targets a mammalian gene.

36. The method of claim 1 or 2, wherein the target gene comprises an endogenous target gene or a transgene.

37. The method of claim 1 or 2, wherein the target gene comprises a disease-associated gene.

38. The method of claim 1 or 2, wherein the target gene is a DMD, EMX1, or VEGFA gene.

39. 39. The method of claim 38, wherein the VEGFA gene is VEGFA1 or VEGFA3.

40. Optimized gRNA produced by the method of claim 1 or 2.

41. 41. The optimized gRNA of Claim 40, wherein the gRNA can discriminate between on-target and off-target sites using the minimum thermodynamic energy difference between said sites.

42. 41. The optimized gRNA of Claim 40, wherein the optimized gRNA regulates strand invasion into a protospacer.

43. 41. The optimized gRNA of claim 40, wherein the optimized gRNA comprises at least one nucleotide sequence of SEQ ID NOs: 149-315, 321-323, and 326-329.

44. 41. An isolated polynucleotide encoding the optimized gRNA of claim 40.

45. 45. A vector comprising the isolated polynucleotide of claim 44.

46. 46. ​​A cell comprising the isolated polynucleotide of claim 44 or the vector of claim 45.

47. 47. A kit comprising the isolated polynucleotide of claim 44, the vector of claim 45, or the cell of claim 46.

48. 41. A method of epigenome editing in a target cell or subject, comprising contacting the cell or subject with an effective amount of the optimized gRNA molecule of claim 40 and a fusion protein, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

49. 41. A method of site-specific DNA cleavage in a target cell or subject, comprising contacting the cell or subject with an effective amount of the optimized gRNA molecule of claim 40 and a fusion protein or Cas9 protein, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

50. 41. A method of genome editing in a cell, comprising administering to the cell an effective amount of the optimized gRNA molecule of claim 40 and a fusion protein, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activation activity, transcriptional repression activity, nuclease activity, transcription termination factor activity, histone modification activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

51. 51. The method of Claim 50, wherein the genome editing comprises correcting a mutated gene or inserting a transgene.

52. 52. The method of claim 51 , wherein correcting the mutated gene comprises deleting, rearranging, or replacing the mutated gene.

53. 52. The method of claim 51 , wherein correcting the mutated gene comprises nuclease-mediated non-homologous end joining or homology directed repair.

54. 41. A method of regulating gene expression in a cell, comprising contacting the cell with an effective amount of the optimized gRNA molecule of claim 40 and a fusion protein, wherein the fusion protein comprises a first polypeptide domain comprising a nuclease-deficient Cas9 and a second polypeptide domain having an activity selected from the group consisting of transcriptional activating activity, transcriptional repressing activity, nuclease activity, transcriptional terminator activity, histone modifying activity, nucleic acid association activity, DNA methylase activity, and direct or indirect DNA demethylase activity.

55. 55. The method of claim 54, wherein gene expression of at least one target gene is modulated if the gene expression level of said at least one target gene is increased or decreased as compared to a normal gene expression level for said at least one target gene.

56. 55. The method of Claim 54, wherein the fusion protein comprises a dCas9 domain and a transcriptional activator.

57. 57. The method of claim 56, wherein the fusion protein comprises the amino acid sequence of SEQ ID NO:

2.

58. 55. The method of Claim 54, wherein the fusion protein comprises a dCas9 domain and a transcriptional repressor.

59. 59. The method of claim 58, wherein the fusion protein comprises the amino acid sequence of SEQ ID NO:

3.

60. 55. The method of Claim 54, wherein the fusion protein comprises a dCas9 domain and a site-specific nuclease.

61. The method of any one of claims 48 to 60, wherein the optimized gRNA is encoded by a polynucleotide sequence and packaged into a lentiviral vector.

62. 62. The method of Claim 61, wherein the lentiviral vector comprises an expression cassette comprising a promoter operably linked to a polynucleotide sequence encoding the gRNA.

63. 63. The method of Claim 62, wherein the promoter operably linked to the polynucleotide encoding the optimized gRNA is inducible.

64. 62. The method of Claim 61, wherein the lentiviral vector further comprises a polynucleotide sequence encoding the Cas9 protein or fusion protein.

65. The method of any one of claims 48 to 64, wherein the at least one target gene is a disease-associated gene.

66. The method of any one of claims 48 to 65, wherein the target cell is a eukaryotic cell.

67. The method of any one of claims 48 to 66, wherein the target cell is a mammalian cell.

68. The method of any one of claims 48 to 67, wherein the target cell is a HEK293T cell.

Images