and Senquencing (FIND-seq)

CLAIM OF PRIORITY

This application claims priority under 35 USC §119(e) to U.S. Patent Application Serial No. 62/217,690, filed on September 11, 2015. The entire contents of the foregoing are hereby incorporated by reference.

FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

This invention was made with Government support under Grant No. DP I GM105378 awarded by the National Institutes of Health. The Government has certain rights in the invention.

TECHNICAL FIELD

Described herein are in vitro methods for defining the genome-wide cleavage specificities of engineered nucleases such as CRISPR-Cas9 Nucleases.

BACKGROUND

Engineered nuclease technology including zinc fingers, TALENs, and CRISPR-Cas9 nucleases, is revolutionizing biomedical research and providing important new modalities for therapy of gene-based diseases.

SUMMARY

At least in part, the present invention is based on the development of sensitive, unbiased methods for genome-wide detection of potential engineered nuclease (e.g., CRISPR-Cas9) off-target cleavage sites from cell type-specific genomic DNA samples. The present methods use exonuclease selection of covalently closed DNA molecules to create a population of genomic DNA molecules with very few free DNA ends, as a starting population for cleavage-specific enrichment and sequencing.

Enrichment of these cleaved fragments, estimated to be >20,0()0X from human genomic DN A, enables very sequencing-efficient discovery of in vitro cleaved DNA fragments, in contrast to methods such as Digenome-Seq (Kim et al., Nat Methods. 2015 Mar; 12(3):237-43) that rely on whole-genome sequencing. After optimization, the present in vitro assay detected 100% of off-target cleavage sites detected by the in- cell GUIDE-seq assay at the VEGFA site target site used as a test case and described in Tsai et al., Nat Biotechnol. 2015 Feb;33(2): 187-97 (in other words, the present in vitro assay detects a superset of GUIDE-seq detected cleavage sites).

Described herein are methods of enzymaticallv preparing a l ibrary of DNA fragments without DNA double-stranded breaks (DSBs) (the terms "breaks" and "cleavage" are used herein interchangeably) by ligation of stem-loop or hairpin adapters followed by exonuclease selection, and for detecting nuclease-induced cleavage of this enzymatically purified library of covalently closed DNA fragments by sequencing. Together, these two methods comprise a strategy for efficiently mining for nuclease-induced cleavage sites in complex mixtures of DNA. Thus the methods can include creating a population of molecules without ends, treating that population with a nuclease, and finding molecules in this population that have newly created ends as a result of nuclease-induced cleavage.

In a first aspect, the invention provides methods for preparing a library of covalently closed DNA fragments. The methods include providing DNA, e.g., gDNA from a cell type or organism of interest; optionally randomly shearing the DNA to a defined average length, e.g., an average length of about 200-1000 bps, e.g., about 500 bps, to provide a population of DNA fragments; preparing the fragments for end- ligation, e.g., by end-repairing and then A -tailing the sheared DNA; ligating a first hairpin adapter comprising at least a single deoxyuridme (uracil) and a first primer site compatible for use in PCR priming and/or sequencing, e.g., next generation sequencing (NGS) to the ends of the fragments, to prepare a population of ligated fragments; contacting the ligated fragments with a nuclease that induces double stranded breaks; and purifying the ligated fragments using an exonuclease, thereby preparing a library of covalently closed DNA fragments.

The methods can also include contacting the library of covalently closed DNA fragments with a nuclease to induce site-specific cleavage; ligating a second hairpin adapter comprising at least a single deoxyuridine and a second primer site compatible for use with the first primer site in PCR priming and/or sequencing; contacting the library with an enzyme, e.g., uracil DNA glycosylase (UDG) and/or endonuclease VIII, a DNA glycosylase-lyase, to nick the DNA at the deoxyuridine; and sequencing those fragments having a first and second hairpin adapter.

Also provided herein are methods for preparing a library of fragments comprising nuclease-induced double stranded breaks in DNA, e.g., genomic DNA (gDNA). The methods can include providing DNA, e.g., gDNA from a cell type or organism of interest; randomly shearing the DNA to a defined average length, e.g., an average length of about 200-1000 bps, e.g., about 100-500 bps; optionally end- repairing and then A -tailing the sheared DNA; ligating a first single-tailed hairpin adapter, preferably comprising a first region, e.g., of about 10-20, e.g., 12 nucleotides; a second region, e.g., of about 45-65, e.g., 58 nucleotides, that forms one or more hairpin loops and comprises a first primer site compatible for use in PGR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region, e.g., of about 10-20, e.g., 13 nucleotides (e.g., one longer than the first region) that is complementary to the first region, with a single deoxyuridine nucleotide between the first and second regions; contacting the sample with one or more exonucleases (e.g., bacteriophage lambda exonuclease, E. coli Exol, PlasmidSafe™ ATP -dependent exonuciease), sufficient to degrade any DNA molecules that lack the first (e.g., 5') single-tailed hairpin adapter ligated to both of their ends; treating the sample with a nuclease to induce site-specific cleavage (e.g., of on- and/or off-target sites, e.g., to induce blunt or staggered/overhanging ends) optionally end-repairing and then A- tailing the resulting ends; ligating a second (e.g., 3') single-tailed hairpin adapter comprising a first region of about 10-20, e.g., 15 nucleotides; a second region of about 40-60, e.g., 48, nucleotides that forms one or more hairpin loops and comprises a second primer compatible for use with the first primer site in PGR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region of about 10- 20, e.g., 16 nucleotides (e.g., one longer than the first region) that is complementary to the first region and that also contains a single deoxyuridine nucleotide between the second and third regions, to create a population wherein the DNA fragments that were cleaved by the nuclease have a first and second single-tailed hairpin adapter ligated to their respective ends; thereby preparing a library of fragments, e.g., wherein one end was created by a nuclease -induced double stranded break in the DNA.

In some embodiments, the methods include contacting the library with uracil DNA glycosylase (UDG) and/or endonuclease VITI, a DNA glycosylase-lyase to nick the DNA at the deoxyuridine; and sequencing those fragments bearing a first and a second hairpin adapter.

Also provided herein are methods for detecting nuclease-induced double stranded breaks (DSBs) in DNA, e.g., in genomic DNA (gDNA) of a cell. The methods include providing DNA, e.g., gDNA from a cell type or organism of interest: randomly shearing the DNA to a defined average length, e.g., an average length of about 200-1000 bps, e.g., about 500 bps; optionally end-repairing and then A-tailing the sheared DNA; ligating a first single-tailed hairpin adapter, preferably comprising a first region, e.g., of about 10-20, e.g., 12 nucleotides; a second region, e.g., of about 45-65, e.g., 58 nucleotides, that forms one or more hairpin loops and comprises a first primer site compatible for use in PCR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region, e.g., of about 10-20, e.g., 13 nucleotides (e.g., one longer than the first region) that is complementary to the first region, with a single deoxyuridine nucleotide between the first and second regions; contacting the sample with one or more exonucleases (e.g., bacteriophage lambda exonuclease, E. coli Exol, PlasrnidSafe™ ATP-dependent exonuclease), sufficient to degrade any DNA molecules that lack the first single-tailed hairpin adapter ligated to both of their ends; treating the sample with a nuclease to induce site-specific cleavage the DNA, e.g., to produse on- and/or off-target cleavage sites; optionally end- repairing and then A-tailing the resulting cleaved ends; ligating a second (e.g., 3') single-tailed hairpin adapter comprising a first region of about 10-20, e.g., 15 nucleotides; a second region of about 40-60, e.g., 48, nucleotides that forms one or more hairpin loops and comprises a second primer compatible for use with the first primer site in PCR priming and/or sequencing, e.g., next generation sequencing

(NGS); and a third region of about 10-20, e.g., 16 nucleotides (e.g., one longer than the first region) that is complementary to the first region, and that also contains a single deoxyuridine nucleotide between the second and third regions, to create a population wherein the DNA fragments that were cleaved by the nuclease have the first and second hairpin adapters ligated to their respective ends; thereby preparing a library of fragments comprising nuclease-induced double stranded breaks in the DNA, e.g., wherein one end was created by a nuclease-induced double stranded break in the DN A; contacting the library with uracil DNA glycosyiase (UDG) and/or

endonuclease VIII, a DNA glycosylase-lyase to nick the DNA at the deoxyuridine; and sequencing those fragments bearing a first and a second hairpin adapter; thereby detecting DSBs induced by the nuclease.

In some embodiments, the engineered nuclease is selected from the group consisting of meganucleases, MegaTALs, zinc-finger nucleases, transcription activator effector-like nucleases (TALEN), and Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas RNA-guided nucleases (CRISPR/Cas RGNs).

In some embodiments, treating the sample with a nuclease to induce site- specific cleavage, e.g., at on- and off-target sites, comprises contacting the sample with a Cas9 nuclease complexed with a specific guide RNA (gRNA).

Further, provided herein are methods for determining which of a plurality of guide RNAs is most specific, i.e., induces the fewest off-target DSBs. The methods include, for each of the plurality of guide RNAs: providing gDNA from a cell type or organism of interest; randomly shearing the gDNA to a defined average length, e.g., an average length of about 200-1000 bps, e.g., about 500 bps; optionally end-repairing and then A -tailing the sheared gDNA; ligating a first single-tailed hairpin adapter, preferably comprising a first region, e.g., of about 10-20, e.g., 12 nucleotides; a second region, e.g., of about 45-65, e.g., 58 nucleotides, that forms one or more hairpin loops and comprises a first primer site compatible for use in PGR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region, e.g., of about 10-20, e.g., 13 nucleotides (e.g., one longer than the first region) that is complementary to the first region, with a single deoxyuridine nucleotide between the first and second regions; contacting the sample with one or more exonucleases (e.g., bacteriophage lambda exonuclease, E. coli Exol, PiasmidSafe™ ATP -dependent exonuclease), sufficient to degrade any DNA molecules that lack the first hairpin adapter ligated to both of their ends; treating the sample with a Cas9 nuclease compatible with the guide RNA to induce site-specific cleavage (e.g., of on- and/or off-target sites, e.g., to produce blunt or staggered/overhanging ends); optionally end- repairing and then A-tailing the resulting cleaved ends; ligating a second (e.g., 3') single-tailed hairpin adapter comprising a first region of about 10-20, e.g., 15 nucleotides; a second region of about 40-60, e.g., 48, nucleotides that forms one or more hairpin loops and comprises a second primer compatible for use with the first primer site in PCR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region of about 10-20, e.g., 16 nucleotides (e.g., one longer than the first region) that is complementary to the first region, and that also contains a single deoxyuridine nucleotide between the second and third regions, to create a population wherein the DNA fragments that have a first and second hairpin adapter ligated to their ends are those that were cleaved by the nuclease; thereby preparing a library of fragments comprising nuc lease-induced double stranded breaks in DNA, e.g., wherein one end was created by a nuciease-induced double stranded break in the DNA; contacting the library with uracil DNA glycosylase (UDG) and/or

endonuclease VIII, a DNA glycosylase-lyase to nick the DNA at the deoxyuridine; and sequencing those fragments bearing a first and a second hairpin adapter, thereby detecting DSBs induced by the nuclease in each sample; optionally identifying whether each DSB is on-target or off-target; comparing the DSBs induced by the nuclease in each sample; and determining which of the plurality of guide RNAs induced the fewest off-target DSBs.

In some embodiments, the DNA is isolated from, a mammalian, plant, bacterial, or fungal cell (e.g., gDNA).

In some embodiments, the DNA is synthetic.

In some embodiments, the engineered nuclease is a TALEN, zinc finger, meganuciease, megaTAL, or a Cas9 nuclease.

In some embodiments, the engineered nuclease is a Cas9 nuclease, and the method also includes expressing in the cells a guide RNA that directs the Cas9 nuclease to a target sequence in the genome.

In some embodiments, the primer site in the first or second hairpin adapter comprises a next generation sequencing primer site, a randomized DNA barcode or unique molecular identifier (UMI).

The present methods have several advantages. For example, the present methods are in vitro; in contrast, GUTDE-seq is cell-based, requiring the introduction of double stranded oligodeoxynucleotides (dsODN) as well as expression or introduction of nuclease or nuclease-encoding components into cells. Not all cells will allow the introduction of dsODNs and/or nuclease or nuclease-encoding components, and these reagents can be toxic in some cases. If a cell type of particular interest is not amenable to introduction of dsODNs, nucleases, or nuclease-encoding components, a surrogate cell type might be used, but cell-specific effects might not be detected. The present methods do not require delivery of the dsODN, nuclease, or nuclease-encoding components into a cell, are not influenced by chromatin state, do not create toxicity issues, and enable interrogation of specific cellular genomes by cleavage of genomic DN A that is obtained from the cell-type of interest. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Methods and materials are described herein for use in the present invention; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting. All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

Other features and advantages of the invention will be apparent from the following detailed description and figures, and from the claims.

DESCRIPTION OF DRAWINGS

Figures 1A-B. Overview of an exemplary in vitro assay for genome-wide identification of CRISPR/Cas9 cleavage sites from, complex mixtures of DNA called FIND-seq (Full Interrogation of Nuclease DSBs). (a) Genomic DNA is isolated from human cells and sheared to an average of -500 bp using a Covaris S220 AFA instrument. Sheared DNA is end-repaired, A-taiied, and ligated with a 5' single-tailed hairpin adapter containing a single deoxyuradine. Successful, ligation of the hairpin adapter to both ends of the DNA fragments will result in covaiently closed DNA with no free end. Lambda Exonuclease and E. coli Exol or PiasmidSafe ATP -dependent exonuclease can then be used to dramatically reduce the background of DNA with free ends, leaving a predominantly uniform population of covaiently closed DNA molecules. Cas9 cleavage of adapter-ligated gDNA will produce ends with free 5' phosphates for the ligation of a second 3' single-tailed hairpin adapter that also contains a single uracil base. Treatment with USER, enzyme mixture and PCR enriches for DNA molecules that have been ligated with both 5' and 3' single-tailed adapters. The use of single-tailed adapters reduces background, as only molecules that contain both 5 ' and 3' adapters can be amplified and sequenced, (b) Detailed schematic of adapter ligation steps, with sequence and predicted DNA secondary structure of hairpin adapters. Ligation of hairpin adapters at both ends creates covaiently closed DNA molecules that can be subsequently reopened at uracil- containing sites. Figures 2A-B. Reads mapped at on-target and off-target cleavage sites, (a) Visualization of 'on-target' genomic regions where bidirectionaliy mapping uniform-end read signatures of Cas9 cleavage can be identified. Reads are indicated by rectangles; coverage is displayed above. The arrow indicates the predicted cut site and data is shown for 3 target sites in the VEGFA locus. A sample treated with Streptococcus pyogenes Cas9 compiexed with EMX1 gRNA is also shown as a negative control, (b) Visualization of example 'off-target' genomic region

20:56175349-56175372 in VEGFA site 1. Bidirectionaliy mapping reads with 'uniform' ends can be observed in tins region and are typical of sites detected by this in vitro assay. Reads generally start within 1-bp of the predicted cut-site and originate outward in both directions from this position.

Figures 3A-D. Analysis of FIND-seq in vitro cleavage selection assay results using different exonuclease treatment conditions, (a) Mapped read counts at genome- wide in vitro cleavage assay detected sites for a under different exonuclease treatments, (b) Number of sites detected by genome-wide in vitro cleavage assay with different exonuclease treatments, (c) Average read counts at genome-wide in vitro cleavage assay detected sites under different exonuclease treatments, (d) Percentage of GUTDE-seq detected sites that are detected by genome-wide in vitro cleavage assay. Note there are three conditions where 100% of GUIDE-seq sites are detected using the simple in vitro assay.

Figure 4. Venn diagram of overlap between GUIDE-seq and FIND-seq genomic in vitro cleavage assay performed with 1 hour of lambda exonuclease and E. coli exonuclease I, and 1 hour of PlasmidSafe exonuclease treatment at the VEGFA target site 1. In tins condition, 100% of GUIDE-seq sites are also detected with the in vitro assay.

Figures 5A-B. Analysis of FIND-seq genomic in vitro cleavage assay testing the adapter ligation order with different exonuclease treatment conditions. The condition that detected the highest number of sites was 1 hr PlasmidSafe, 1 hr Lambda exonuclease/E. Coli exonuclease I treatment, with the 3' adapter added first. (a) Number of sites by FIND-seq genome-wide in vitro cleavage assay with different exonuclease treatments and adapter ligation order, (b) Read counts at FIND-seq detected off-target sites using different exonuclease treatments and adapter ligation order. DETAILED DESCRIPTION

An important consideration for the therapeutic deployment of CRJSPR-Cas9 nucleases is having robust, comprehensive, unbiased, and highly sensitive methods for defining their off-target effects. Recently, a number of methods for defining the genome-wide off-target effects of CRISPR-Cas9 and other customizable nucleases have been described. For example, the GUIDE-seq method, which relies on uptake of a short double-stranded oligonucleotide "tag" into nuclease-induced DSBs in living cells, has been shown to define off-target sites on a genome-wide scale, identifying sites that are mutagenized with frequencies as low as 0.1% of the time in a population of cells (Tsai et a!., Nat Biotechnol. 2015). Other cell-based methods for defining nuclease-induced off -target breaks include a method that maps translocation fusions to the on-target site and another that relies on uptake of integration-deficient lenti virus (IDLV) genomes into sites of DSBs.

Despite this recent progress, cell-based methods for off-target determination have a number of limitations including: (i) a requirement to be able to introduce both the nuclease components and a tag such as the dsODN or IDLV genome into cells; (2) biological selection pressures that might favor or disfavor the growth of cells harboring certain types of off-target mutations; and/or (3) the potential confounding effects of cell -type-specific parameters such as chromatin, DNA methylation, gene expression, and nuclear architecture on nuclease off-target activities/effects.

In vitro methods using purified genomic DNA provide an attractive alternative because they would sidestep these various limitations of cell-based approaches. However, in vitro methods face the challenge that isolated genomic DNA is by experimental necessity randomly sheared (or broken) into smaller pieces. This poses a challenge because it is not easy to differentially identify DSBs induced by shearing from those induced by treatment of the genomic DNA in vitro by nucleases. The recently described Digenome attempted to use alignment of common ends induced by nucleases in genomic sequence but the signal for this type of event can be challenging to discern relative to the background of random DSBs from shearing of genomic DNA induced during genomic DNA isolation and by deliberate shearing of the DNA to smaller size pieces required for DNA sequencing methods. In addition, and perhaps more importantly, because there is no enrichment for the nuclease-induced DSBs, nearly all of the sequencing data generated with Digenome is just re-sequencing of random genomic DNA, a factor that further limits the sensitivity of the method since very few reads generated contribute the desired information about nuclease-induced DSBs. Indeed, Digenome failed to identify a number of off-target sites found by GUIDE-seq for a particular gRNA, although the caveat must be added that the two experiments were performed in different cell lines.

Described herein is an in vitro method that enables comprehensive determination of DSBs induced by nucleases on any genomic DNA of interest. This method enables enrichment of nuclease-induced DSBs over random DSBs induced by shearing of genomic DNA. An overview of how the method works can be found in Figures 1A and IB. In brief, genomic DNA from a cell type or organism of interest is randomly sheared to a defined average length . In the present experiments, an average length of 500 bps worked well for subsequent next-generation sequencing steps; however, shorter or longer lengths can also be used. The random broken ends of this genomic DNA are end repaired (i.e., the random, overhanging ends are filled m/blunted, e.g., using T4 polymerase, Klenow fragment, and T4 Polynucleotide

Kinase (PNK)) and then A-tailed (i.e., an A (adenosine) is enzymaticaily added to the 3' end of a blunt, double-stranded DN A molecule). This then enables the ligation of a hairpin adapter, preferably bearing a 5' end next-generation sequencing primer site, that also contains a single deoxyuridine nucleotide at a specific position (Figure IB) and a I -nucleotide (thymidine) overhang at the 3 ' end; hereafter this adapter is referred to as the "5' single-tailed hairpin adapter" or simply "5' hairpin adapter".

Following ligation of this adapter, the sample is then treated with one or more exonucleases (e.g., bacteriophage lambda exonuclease, E. coli Exol, PlasmidSafe™ ATP -dependent exonuclease) that degrade any DNA molecules that have not had the 5' single-tailed hairpin adapter ligated to both of their ends. The collection of molecules is then treated to induce blunt-end cuts, e.g., with a Cas9 nuclease complexed with a specific guide RNA (gRNA). The resulting nuclease-induced blunt ends are A-tailed and then a second hairpin adapter bearing a 3 ' end next-generation sequencing primer site and that also contains a single deoxyuridine nucleotide at a specific position is ligated to these ends; this adapter is also referred to herein as the "3' single-tailed hairpin adapter" or simply ' ^" 3 ^" hairpin adapter".

As a result of these treatments, the only DN A fragments that should have a 5' and a 3' single-tailed hairpin adapter ligated to their ends are those that were cleaved by the nuclease. Following treatment (e.g., by the USER enzyme mixture) to nick DNA wherever a deoxyuridine is present, only those fragments bearing the two types of adapters can be sequenced.

The present methods can be used to prepare libraries of fragments for next generation sequencing, e.g., to identify guideRNA/Cas9 combinations that induce the most specific DSBs (e.g., that have the fewest off-target effects), e.g., for therapeutic or research purposes.

Single -Tailed Hairpin A dapters

The present methods include the use of non-naturally occurring 3' and 5' single-tailed hairpin adapters. The hairpin adapters include (from 5' to 3') a first region, e.g., of about 10-20, e.g., 12 or 15, nucleotides; a second region, e.g., of about 45-65, e.g., 58 or 48, nucleotides that fonns one or more hairpi n loops and includes a sequence compatible for use in PGR priming and/or sequencing, e.g., next generation sequencing (NGS); and a third region, e.g., of about 10-20, e.g., 13 or 16, nucleotides that is complementary to the first region. The lengths of the first, second and third regions can vary' depending on the NGS method selected, as they are dependent on the sequences that are necessary for priming for use with the selected NGS platform.

The hairpin adapters include at least one, preferably only one, uracil that allows the adaptor to be opened by Uracil DNA glycosylase (UDG) and Endonuciease VIII, a DNA glycosylase-lyase, e.g., the USER (Uracil-Specific Excision Reagent) Enzyme mixture (New England BioLabs). The UDG catalyzes the excision of uracil bases to form an abasic site but leave the phosphodiester backbone intact (see, e.g., Lindhal et al., J. Biol. Chem.. 252:3286-3294 (1977); Lindhal, Annu. Rev. Biochem. 51 :61-64 (1982)). The Endonuciease VIII breaks the phosphodiester backbone at the 3 ^' and 5 ^' sides of the abasic site (see, e.g., Melamede et al., Biochemistry 33: 1255- 1264 (1994); Jiang et a!., J. Biol. Chem. 272:32230-32239 (1997)). This combination generates a single nucleotide gap at the location of a uracil. In some embodiments, in the 5' single-tailed hairpin adapters the uracil is placed at or within 1, 2, 3, or 4 nucleotides of the end of the first region and the beginning of the second region, and in the 3' single-tailed hairpin adapters the uracil is placed at or within 1, 2, 3, or 4 nucleotides of the end of the first region and the beginning of the second region.

In the present methods, ail parts of the hairpin adapters are preferably orthologous to the genome of the cell (i.e., are not present in or complementary to a sequence present in, i.e., have no more than 10%, 20%, 30%, 40%, or 50% identity to a sequence present in, the genome of the cell). The hairpin adapters can preferably be between 65 and 95 nts Song, e.g., 70-90 nts or 75-85 nts long.

Each of the 5' and 3' single-tailed hairpin adapters should include a primer site that is a randomized DNA barcode (e.g., SHAPE-SEQ, Lucks et al., Proc Natl Acad Sci U S A 108: 11063-11068), unique molecular identifier (UMI) (see, e.g., Kivioja et al. Nature Methods 9, 72-74 (2012); Islam, et al.. Nature Methods 11, 163-166 (2014); Karlsson et al., Genomics. 2015 Mar; 105(3): 150-8), or unique PCR priming sequence and/or unique sequence compatible for use in sequencing (e.g., NGS). The sequence compatible for use in sequencing can be selected for use with a desired sequencing method, e.g., a next generation sequencing method, e.g., lilumma, Ion Torrent or library preparation method like Roche/454, lllumina Solexa Genome Analyzer, the Applied Biosystems SOLiD™ system, Ion Torrent™ semiconductor sequence analyzer, PacBio® real-time sequencing and Helicos™ Single Molecule Sequencing (SMS). See, e.g., WO2014020137, Voelkerdmg et al., Clinical Chemistry 55:4 641-658 (2009) and Metzker, Nature Reviews Genetics 11 :31-46 (2010)). A number of kits are commercially available for preparing DNA for NGS, including the ThruPLEX DNA-seq Kit (Rubicon; see U.S. Patents 7,803,550; 8,071,312; 8,399,199; 8,728,737) and NEBNext® (New England BioLabs; see e.g., U.S. patent 8,420,319)). Exemplary sequences are shown in Table A, below.

In some embodiments, the hairpin adapters include a restriction enzyme recognition site, preferably a site that is relatively uncommon in the genome of the cell.

The hairpin adapters are preferably modified; in some embodiments, the 5 ' ends of the hairpin adapters are phosphorylated. In some embodiments, the hairpin adapters are blunt ended. In some embodiments, the hairpin adapters include a random variety of 1, 2, 3, 4 or more nucleotide overhangs on the 5' or 3 ' ends, or include a single T at the 5' or 3' end.

The hairpin adapters can also include one or more additional modifications, e.g., as known in the art or described in PCT/US2011/060493. For example, in some embodiments, the hairpin adapters is biotinylated. The biotin can be anywhere internal to the hairpin adapters (e.g., a modified thymidine residue (Biotin-dT) or using biotin azide), but not on the 5' or 3' ends. This provides an alternate method of recovering fragments that contain the FIND-seq hairpin adapters. Whereas in some embodiments, these sequences are retrieved and identified by PGR, in this approach they are physically pulled down and enriched by using the biotin, e.g., by binding to streptavidin-coated magnetic beads, or using solution hybrid capture; see, e.g., Gnirke et al., Nature Biotechnology 27, 182 - 189 (2009).

Although the present working examples include ligating the 5' adapter and then the 3' adapter, the adapters can be added in either order, i.e., 5' adapter then 3' adapter, or 3 ' adapter then 5 ' adapter. The order may be optimized depending on the exonuclease treatment used, e.g., for Lambda exonuclease, which is a highly processive 5' -> 3' exonuclease, w here adding the 5' adapter second may be advantageous.

Engineered Nucleases

There are presently four main classes of engineered nucleases: 1)

meganucleases, 2) zinc -finger nucleases, 3) transcription activator effector-like nucleases (T ^' ALEN), and 4) Clustered Regularly Interspaced Short Palindromic

Repeats (CRISPR) Gas RNA-guided nucleases (RGN). See, e.g., Gaj et al., Trends Biotechnol. 2013 Jul;31(7):397-405. Any of these, or variants thereof, can be used in the present methods. The nuclease can be transiently or stably expressed in the cell, using methods known in the art; typically, to obtain expression, a sequence encoding a protein is subcloned into an expression vector that contains a promoter to direct transcription. Suitable eukaryotic expression systems are well known in the art and described, e.g., in Sambrook et al., Molecular Cloning, A. Laboratory Manual (4th ed. 2013); Kriegler, Gene Trans fer and Expression: A Laboratory Manual (2006); and Current Protocols in Molecular Biology (Ausubel et al., eds., 2010). Transformation of eukaryotic and prokaryotic cells are performed according to standard techniques (see, e.g., the reference above and Morrison, 1977, J. Bacteriol. 132:349-351: Clark- Curtiss & Curtiss, Methods in Enzymology 101 :347-362 (Wu et al, eds, 1983), Homing Meganucleases

Meganucleases are sequence-specific endonucleases originating from a variety of organisms such as bacteria, yeast, algae and plant organelles. Endogenous meganucleases have recognition sites of 12 to 30 base pairs; customized DNA binding sites with 18bp and 24bp-long meganuclease recognition sites have been described, and either can be used in the present methods and constructs. See, e.g., Silva, G., et al., Current Gene Therapy, 1 1 : 11-27, (2011); Arnould et al., Journal of Molecular Biology, 355:443-58 (2006); Arnould et al.. Protein Engineering Design & Selection, 24: 27-31 (2011); and Stoddard, Q. Rev, Biophys. 38, 49 (2005); Gnzot et al., Nucleic Acids Research, 38:2006-18 (2010).

CRISPR-Cas Nucleases

Recent work has demonstrated that clustered, regularly interspaced, short palindromic repeats (CRISPR)/CRISPR-associated (Cas) systems (Wiedenheft et al.. Nature 482, 331-338 (2012); Horvath et al, Science 327, 167-170 (2010); Terns et al, Curr Opin Microbiol 14, 321-327 (2011)) can serve as the basis of a simple and highly efficient method for performing genome editing in bacteria, yeast and human cells, as well as in vivo in whole organisms such as fruit flies, zebrafish and mice (Wang et al. Cell 153, 910-918 (2013); Shen et al. Cell Res (2013); Dicarlo et al. Nucleic Acids Res (2013); Jiang et al, Nat Biotechnoi 31, 233-239 (2013); Jinek et al, Elife 2, e00471 (2013); Hwang et al, Nat Biotechnoi 31 , 227-229 (2013); Cong et al. Science 339, 819-823 (2013); Mali et al . Science 339, 823-826 (2013c); Cho et al, Nat Biotechnoi 31, 230-232 (2013); Gratz et al. Genetics 194(4): 1029-35 (2013)). The Cas9 nuclease from. S. pyogenes can be guided via simple base pair

complementarity between 17-20 nucleotides of an engineered guide RNA (gRNA), e.g., a single guide RN A or crRNA/tracrRNA pair, and the complementary strand of a target genomic DNA sequence of interest that lies next to a protospacer adjacent motif (PAM), e.g., a PAM matching the sequence NGG or NAG (Shen et al . Cell Res (2013); Dicarlo et al. Nucleic Acids Res (2013); Jiang et al, Nat Biotechnoi 31, 233- 239 (2013); Jinek et al, Elife 2, e00471 (2013); Hwang et al, Nat Biotechnoi 31, 227- 229 (2013); Cong et al. Science 339, 819-823 (2013); Mali et al. Science 339, 823- 826 (2013c); Cho et al, Nat Biotechnoi 31 , 230-232 (2013); Jinek et al. Science 337, 816-821 (2012)).

In some embodiments, the present system utilizes a wild type or variant Cas9 protein from S. pyogenes or Staphylococcus aureus, either as encoded in bacteria or codon-optimized for expression in mammalian cells. The guide RNA is expressed in the cell togetiier with the Cas9. Either the guide RNA or the nuclease, or both, can be expressed transiently or stably in the cell. TAL Effector Repeat Arrays

TAL effectors of plant pathogenic bacteria in the genus Xanthomonas play important roles in disease, or trigger defense, by binding host DNA and activating effector-specific host genes. Specificity depends on an effector-variable number of imperfect, typically -33-35 amino acid repeats. Polymorphisms are present primarily at repeat positions 12 and 13, which are referred to herein as the repeat variable- diresidue (RVD). The RVDs of TAL effectors correspond to the nucleotides in their target sites in a direct, linear fashion, one RVD to one nucleotide, with some degeneracy and no apparent context dependence. In some embodiments, the polymorphic region that grants nucleotide specificity may be expressed as a triresidue or triplet.

Each DNA binding repeat can include a RVD that determines recognition of a base pair in the target DNA sequence, wherein each DNA binding repeat is responsible for recognizing one base pair in the target DNA sequence. In some embodiments, the RVD can comprise one or more of: HA for recognizing C; ND for recognizing C; HI for recognizing C: HN for recognizing G; NA for recognizing G; SN for recognizing G or A; YG for recognizing T; and NK for recognizing G, and one or more of: HD for recognizing C; NG for recognizing T; NI for recognizing A; NN for recognizing G or A; N S for recognizing A or C or G or T; N* for recognizing C or T, wherein * represents a gap in the second position of the RVD; HG for recognizing T; H* for recognizing T, wherein * represents a gap in the second position of the RVD; and IG for recognizing T.

TALE proteins may be useful in research and biotechnology as targeted chimeric nucleases that can facilitate homologous recombination in genome engineering (e.g., to add or enhance traits useful for biofuels or biorenewables in plants). These proteins also may be useful as, for example, transcription factors, and especially for therapeutic applications requiring a very high le vel of specificity such as therapeutics against pathogens (e.g., viruses) as non-limiting examples.

Methods for generating engineered TALE arrays are known in the art, see, e.g., the fast ligation-based automatable solid-phase high-throughput (FLASH) system described in USSN 61/610,212, and Reyon et al, Nature Biotechnology 30,460-465 (2012); as well as the methods described in Bogdanove & Voytas, Science 333, 1843- 1846 (2011); Bogdanove et al., Curr Opm Plant Biol 13, 394-401 (2010); Scholze & Boch, J. Curr Opin Microbiol (2011); Boch et al. Science 326, 1509-1512 (2009); Moscou & Bogdanove, Science 326, 1501 (2009); Miller et al, Nat Biotechnol 29, 143-148 (2011); Morbitzer et al, T. Proc Natl Acad Sci U S A 107, 21617-21622

(2010) ; Morbitzer et al . Nucleic Acids Res 39, 5790-5799 (2011); Zhang et al, Nat Biotechnol 29, 149-153 (2011); Geissler et al, PLoS ONE 6, e!9509 (2011); Weber et al, PLoS ONE 6, e l9722 (2011); Christian et al. Genetics 186, 757-761 (2010); Li et al. Nucleic Acids Res 39, 359-372 (2011); Mahfouz et al, Proc Natl Acad Sci U S A 108, 2623-2628 (2011); Mussolino et al. Nucleic Acids Res (2011 ); Li et al. Nucleic Acids Res 39, 6315-6325 (2011); Cermak et al. Nucleic Acids Res 39, e82 (2011); Wood et al. Science 333, 307 (2011); Hockemeye et al. Nat Biotechnol 29, 731-734

(201 1) ; Tesson et al, Nat Biotechnol 29, 695-696 (2011 ); Sander et al, Nat

Biotechnol 29, 697-698 (2011); Huang et al, Nat Biotechnol 29, 699-700 (2011); and Zhang et al, Nat Biotechnol 29, 149-153 (2011); all of which are incorporated herein by reference in their entirety.

Also suitable for use in the present methods are MegaTALs, which are a fusion of a meganuclease with a TAL effector; see, e.g., Boissel et al, Nucl. Acids Res. 42(4):2591-2601 (2014); Boissel and Scharenberg, Methods Mol Biol.

2015;1239: 171 -96.

Zinc Fingers

Zinc finger proteins are DNA-binding proteins that contain one or more zinc fingers, independently folded zinc-containing mini-domains, the structure of which is well known in the art and defined in, for example, Miller et al, 1985, EMBO J, 4: 1609; Berg, 1988, Proc. Natl. Acad. Sci. USA, 85:99; Lee et al, 1989, Science. 245:635; and King, 1993, Gene, 135:83. Crystal structures of the zinc finger protein Zif268 and its variants bound to DNA show a semi-conserved pattern of interactions, in which typically three amino acids from the alpha-helix of the zinc finger contact three adjacent base pairs or a ' ^" subsite" in the DNA (Pavletich et al., 1991, Science, 252:809; Elrod-Erickson et al, 1998, Structure, 6:451). Thus, the crystal structure of Zif268 suggested that zinc finger DNA-binding domains might function in a modular manner with a one-to-one interaction between a zinc finger and a three-base-pair "subsite" m the DNA sequence. In naturally occurring zinc finger transcription factors, multiple zinc fingers are typically linked together in a tandem array to achieve sequence-specific recognition of a contiguous DNA sequence (Klug, 1993, Gene 135:83).

Multiple studies have shown that it is possible to artificially engineer the DNA binding characteristics of individual zinc fingers by randomizing the amino acids at the alpha-helical positions involved in DNA binding and using selection

methodologies such as phage display to identify desired variants capable of binding to DNA target sites of interest (Rebar et ai . 1994, Science, 263:671: Choo et al, 1994 Proc. Natl. Acad. Sci. USA, 91 : 11163; Jamieson et al, 1994, Biochemistry 33:5689; Wu et al, 1995 Proc. Natl. Acad. Sci. USA, 92: 344). Such recombinant zinc finger proteins can be fused to functional domains, such as transcriptional activators, transcriptional repressors, methylation domains, and nucleases to regulate gene expression, alter DNA metliylation, and introduce targeted alterations into genomes of model organisms, plants, and human cells (Carroll, 2008, Gene Ther., 15: 1463-68; Cathomen, 2008, Mol. Ther., 16: 1200-07; Wu et al., 2007, Cell. Mol. Life Sci , 64:2933-44).

One existing method for engineering zinc finger arrays, known as "modular assembly," advocates the simple joining together of pre-selected zinc finger modules into arrays (Segal et al, 2003, Biochemistry, 42:2137-48; Beerli et al, 2002, Nat. Biotechnol., 20: 135-141; Mandell et al, 2006, Nucleic Acids Res, 34:W516-523; Carroll et al., 2006, Nat. Protoc. 1 : 1329-41; Liu et al., 2002, J. Biol. Chem.,

277:3850-56; Bae et al, 2003, Nat. Biotechnol., 21 :275-280; Wright et al, 2006, Nat. Protoc, 1 : 1637-52). Although straightforward enough to be practiced by any researcher, recent reports have demonstrated a high failure rate for this method, particularly in the context of zinc finger nucleases (Ramirez et al, 2008, Nat.

Methods, 5:374-375; Kim et ai, 2009, Genome Res. 19: 1279-88), a limitation that typically necessitates the construction and cell-based testing of very large numbers of zinc finger proteins for any given target gene (Kim et al., 2009, Genome Res.

19: 1279-88).

Combinatorial selection-based methods that identify zinc finger arrays from randomized libraries have been shown to have higher success rates than modular assembly (Maeder et al, 2008, Mol. Cell, 31 :294-301; Joung et al, 2010, Nat.

Methods, 7:91-92; Isalan et al, 2001, Nat. Biotechnol , 19:656-660). In preferred embodiments, the zinc finger arrays are described in, or are generated as described in, WO 2011/017293 and WO 2004/099366. Additional suitable zinc finger DBDs are described in U.S. Pat. Nos. 6,511,808, 6,013,453, 6,007,988, and 6,503,717 and U.S. patent application 2002/0160940.

DNA

The methods described herein can be applied to any double stranded DNA, e.g., genomic DNA isolated from any cell, artificially created populations of DNAs, or any other DNA pools, as it is performed in vitro.

Sequencing

As used herein, "'sequencing"' includes any method of determining the sequence of a nucleic acid. Any method of sequencing can be used in the present methods, including chain terminator (Sanger) sequencing and dye terminator sequencing. In preferred embodiments, Next Generation Sequencing (NGS), a high- throughput sequencing technology that performs thousands or millions of sequencing reactions in parallel, is used. Although the different NGS platforms use varying assay chemistries, they all generate sequence data from a large number of sequencing reactions run simultaneously on a large number of templates. Typically, the sequence data is collected using a scanner, and then assembled and analyzed bioinformatically. Thus, the sequencing reactions are performed, read, assembled, and analyzed in parallel; see, e.g., US 20140162897, as well as Voelkerding et al, Clinical Chem., 55: 641 -658, 2009; and MacLean et a!., Nature Rev. Microbiol, 7: 287-296 (2009). Some NGS methods require template amplification and some that do not.

Amplification-requiring methods include pyrosequencing (see, e.g., U.S. Pat. Nos. 6,210,89 and 6,258,568; commercialized by Roche); the Solexa/lllumina platform (see, e.g., U.S. Pat. Nos. 6,833,246, 7,115,400, and 6,969,488); and the Supported Oligonucleotide Ligation and Detection (SOLID) platform (Applied Biosvstems; see, e.g., U.S. Pat. Nos. 5,912,148 and 6, 130,073). Methods that do not require amplification, e.g., single -molecule sequencing methods, include nanopore sequencing, HeliScope (U.S. Pat. Nos. 7,169,560; 7,282,337; 7,482,120; 7,501,245; 6,818,395; 6,911,345; and 7,501,245); real-time sequencing by synthesis (see, e.g., U.S. Pat. No. 7,329,492); single molecule real time (SMRT) DNA sequencing methods using zero-mode waveguides (ZMWs); and other methods, including those described in U.S. Pat. Nos. 7,170,050; 7,302,146; 7,313,308; and 7,476,503). See, e.g., US 20130274147; US20140038831 ; Metzker, Nat Rev Genet 11 (1): 31-46 (2010).

Alternatively, hybridization-based sequence methods or other high-throughput methods can also be used, e.g., microarray analysis, NANOSTRING, ILLUMINA, or other sequencing platforms.

Kits

Also provided herein are kits for use in the methods described herein. The kits can include one or more of the following: 5' single-tailed hairpin adapters; 3' single- tailed hairpin adapters; reagents and/or enzymes for end repair and A tailing (e.g., T4 polymerase, Klenow fragment, T4 Polynucleotide Kinase (PNK), and/or Taq DNA Polymerase): exonuclease; uracil DNA glycosylase (UDG) and/or endonuclease VIII, a DNA glycosylase-iyase, e.g., the USER (Uracil-Specific Excision Reagent) Enzyme mixture (New England BioLabs); purified cas9 protein; guideRNA (e.g., control gRNA); gDNA template (e.g., control gDNA template); and instructions for use in a method described herein.

EXAMPLES

The invention is further described in the following examples, which do not limit the scope of the invention described in the claims.

Materials

The following materials were used in Examples 1-2.

2.0 The sequences in CAPS are dual-index barcodes (for demultiplexing samples) in the Illumina Truseq library preparation system. The * represents a phosphorothioate linkage.

Example 1. Optimization of Exemplary FIND-Seq Methodology

Described herein is the development of an in vitro method that enables comprehensive determination of DSBs induced by nucleases on any genomic DNA of interest. This method enables enrichment of nuclease-induced DSBs over random DSBs induced by shearing of genomic DN A. An overview of how the method works can be found in Figures 1A and IB. In brief, genomic DNA from a cell type or organism of interest is randomly sheared to a defined average length. In the present experiments, an average length of 500 bps works well for subsequent next-generation sequencing steps. The random broken ends of this genomic DNA are end repaired and then A-tailed. This then enables the ligation of a hairpin adapter bearing a 5' end next- generation sequencing primer site that also contains a single deoxyuridine nucleotide at a specific position (Figure IB); this adapter is also referred to herein as the "5' single-tailed hairpin adapter". Following ligation of this adapter, the sample is then treated with one or more exonucleases (e.g., bacteriophage lambda exonuclease, E. coli Exol, PlasmidSafe ATP-dependent exonuclease), which degrades any DNA molecules that have not had the 5' single-tailed hairpin adapter ligation to both of their ends. The collection of molecules is then treated with Cas9 nuclease complexed with a specific guide RNA (gRNA). The resulting nuclease-induced blunt ends are A- tailed and then a second hairpin adapter bearing a 3' end next-generation sequencing primer site and that also contains a single deoxyuridine nucleotide at a specific position is ligated to these ends; this adapter is also referred to herein as the "3' single-tailed hairpin adapter". As a result of these treatments, the only DNA fragments that should have a 5 ^" and a 3' single-tailed hairpin adapter ligated to their ends are those that were cleaved by the nuclease. The adapters can be added in any order. Following treatment by the USER enzyme mixture, which nicks DNA wherever a deoxyuridine is present, only those fragments bearing the two types of adapters can be sequenced.

Following paired end next-generation sequencing, the resulting reads can be mapped back to the genome. Sites of nuclease cleavage will have multiple bi- direction reads originating at the site of the nuclease-induced DSB, typically within a nucleotide of the cut site (Figures 2A and 2B). As nuclease-induced cleavage produces 'uniform' ends, in contrast to the staggered ends that results from random physical shearing of DNA, these sites can be simply bioinformatically identified by their signature uniform read alignments.

Using an initial non-optimized version of this approach, off-target sites induced by four different gRNAs and Cas9 nuclease were successfully identified. The target sites are listed in Table 1.

Table 1. List of sgRNA target sites tested with in vitro cleavage selection assay.

Target site name Cei!s Target Site (5' ~> 3') SEQ !D MO:

VEGFA sitel U 20S GGGTGGGGGGAGTTTGCTCCNGG 19

VEGFA site2 U20S GACCCCCTCCACCCCGCCTCNGG 20

VEGFA site3 U20S GGTGAGTGAGTGTGTGCGTGNGG 21

EMXl U20S GAGTCCGAGCAGAAGAAGAANGG 22

The results of table 2 were encouraging for this assay, as the majority of GUIDE- detected sites were also detected by this method. In the first trial, 58-77% of sitet overlapped between methods using 2.5-3M reads. ~20,000X enrichment was estimated for cleaved sites.

A critical parameter for optimization of the present method was the efficiency of degradation of linear DNA fragments following the ligation of the first 5' single- tailed hairpin adapter. Indeed, without wishing to be bound by theory, most of the ' ^" background" reads mapped throughout the genome may be due to incomplete digestion of unligated DNA at this step, thereby leaving ends to which the second 3' single tailed hairpin adapter can be ligated.

Therefore, the methods were further optimized by testing various kinds and numbers of exonucleases and lengths of incubation to find an optimal treatment method. As seen in Figures 3A-3D, three treatments consistently yielded the highest number of reads and sites and also found 100% of the off-target sites identified by matched GUIDE-seq experiments with the same gRNAs: 1 hour PlasmidSafe treatment, two serial 1 hour PlasmidSafe treatments, and 1 hour of Lambda exonuclease treatment followed by 1 hour of PlasmidSafe treatment- Using an optimized exonuclease treatment, the methods were again performed using CRISPR-Cas9 nuclease and a gRNA targeted to VEGFA site 1. This experiment yielded a larger number of off-target sites. Combining the various experiments performed with the VEGFA site 1 gRNA yields a larger number of off- target sites. GUIDE-seq had previously been performed using this same VEGFA site 1 gRNA and a range of genome-wide off-target sites identified. Importantly, the present in vitro method performed with the optimized exonuclease treatment found ALL of the off-target sites previously identified by GUIDE-seq for the gRNA tested and a very large number of additional, previously unknown off-target sites as well (Figure 4). This result demonstrates that these new methods are as sensitive as GUIDE-seq but offers the ability to identify sites that are not found by that earlier method. Reasons for this might include greater sensitivity of this in vitro method, negative biological selection against certain off-target mutations when practicing GUIDE-seq in cells, and/or negative effects of chromatin, DNA methylation, and gene expression on the activity of nucleases in cells.

OTHER EMBODIMENTS

It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.