WO2018067447A1

WO2018067447A1 - Improved methods for identifying double strand break sites

Info

Publication number: WO2018067447A1
Application number: PCT/US2017/054734
Authority: WO
Inventors: Hon-Ren HUANG; Christian Dombrowski
Original assignee: Itellia Therapeutics, Inc.
Priority date: 2016-10-03
Filing date: 2017-10-02
Publication date: 2018-04-12

Abstract

This application relates to compositions and methods to analyze double-strand break events. The methods may include steps of linear amplification of the target DNA with a primer comprising an internal cleavage site, circularization of the linear amplified product, cleaving the circular product at the cleavage site, and sequencing the linearized product.

Description

IMPROVED METHODS FOR IDENTIFYING DOUBLE STRAND

BREAK SITES

SEQUENCE LISTING

[0000] The instant application contains a Sequence Listing which has been submitted electronically in ASCII format and is hereby incorporated by reference in its entirety. Said ASCII copy, created on September 26, 2017, is named 01155-0003-00PCT_SL.txt and is 7,937 bytes in size.

[0001] This application claims the benefit of priority to United States Provisional Application No. 62/403,232, which was filed on October 3, 2016, and which is incorporated by reference in its entirety.

BACKGROUND

[0002] Over the last several decades, researchers have developed a variety of methods for editing genes in cells, providing tremendous potential for treating genetic, viral, and bacterial diseases. Many of these editing technologies take advantage of cellular mechanisms for repairing double-stranded breaks (DSB) created in DNA by enzymes such as meganucleases, zinc finger nucleases (ZFN), transcription activator-like effector nucleases (TALEN), and clustered regularly interspaced short palindromic repeats (CRISPR) associated nucleases (Cas). The CRISPR/Cas9 system, in particular, has spurred a tremendous amount of research in gene editing because it can be readily engineered to target and create DSBs at specific DNA sequences.

[0003] When a DSB occurs, a cell can use a number of different mechanisms to repair the break. Most commonly in animal cells, non-homologous end joining ("NHEJ") processes can correctly repair the DNA, although in some cases insertions or deletions ("indels") may result. In other circumstances, cells can repair DSBs using a template sequence, which may rely upon homology between the ends of the template and the broken DNA. When a cell uses NHEJ for repair, the loose ends of the DSB can rejoin on the same chromosome, which is termed intra-chromosomal translocation, or on two different chromosomes, which is called inter-chromosomal translocations. Identification of translocation junctions provides location information and can identify intra- and inter- chromosomal events. If a DSB enzyme cleaves only its intended target, then intra- chromosomal translocations are expected because only the loose ends of the on-target DSB site are available for rejoining. However, if a DSB enzyme cleaves any off-target sites, inter- and intra-chromosomal translocations may occur because the loose ends of the on-target cut site may rejoin with the loose ends of the off-target cut site and vice versa. Thus, analyzing translocation junctions can provide important information as to whether a DSB enzyme has off-target activity.

[0004] Therapeutically, on-target cleavage events are desired, and off-target cleavage events must be minimized for at least the reason that off-target events may lead to non-intended therapeutic consequences. In the context of regulatory approval, for example, demonstration that therapeutic DSB enzymes have minimal off-target cleavage events will be important. Methods are thus needed that can identify on- and off-target cleavage events, especially when off-target cleavage sites are rare. These methods must be amenable to using low quantities of starting DNA so that the assay can be applied to non-dividing cells such as primary cells or cells isolated from a treated patient, and must be sensitive enough to identify very rare off-target events. Described herein are methods for identifying on- and off -target DSB events that use very low amounts of starting DNA and that are able to identify very rare DSB events that are not identified by currently available methods.

SUMMARY

[0005] Provided herein are methods for identifying genome-wide on- and off-target double-strand break (DSB) cleavage sites. These methods may be referred to as multiplex amplification of genomic inverted circles (MAGIC), which generally involves formation of multiple single-stranded DNA (ssDNA) circles and re-linearization to generate libraries to analyze for DSB events.

[0006] In one embodiment, the MAGIC method comprises isolating genomic DNA from cells or animal tissues previously treated with an agent (e.g., enzyme) that makes DSBs, optionally shearing the DNA to create fragments of about 500 base pairs, conducting linear amplification PCR (LAM-PCR) using an oligonucleotide primer containing a 5' phosphate and an internal l ',2'-dideoxyribose modification to linearly amplify a genomic region around an on-target cleavage site or around a bait cleavage site, circularizing the amplified single- stranded DNA (ssDNA), and cleaving the circle with an APE 1 enzyme at the ,2'- dideoxyribose site to linearize the ssDNA molecules. The linearized ssDNA is analyzed to identify DSB sites.

[0007] In some embodiments, the analysis of the linearized ssDNA comprises performing one or more PCR reactions on the linearized ssDNA. In some embodiments, a first PCR reaction is done with two primers (e.g., a second and third primer) that are different from the primer used in the LAM-PCR step. The second primer may be a sense primer extending to and annealing to the region downstream of the LAM-PCR primer. The third primer may be an antisense primer that anneals to a region in the first LAM-PCR primer upstream of the l ',2'-dideoxyribose modification site (now cleaved. This "nested" PCR design provides a high degree of amplification specificity because the sequence of the sense primer either does not anneal to a part of the original LAM-PCR primer (e.g., it anneals to the known on-target sequence downstream of the LAM-PCR primer sequence) or it overlaps with and anneals to a part of the LAM-PCR primer sequence and a part of the on-target sequence immediately downstream. A second round of PCR may be conducted with different primers to add adapter sequences, such as, for example, next generation sequencing (NGS) adapters, to the ends of amplified DNA fragments. Resulting adapter-tagged libraries may be processed to obtain sequence information of both ends of the DNA fragments in the libraries.

Translocation fragments should contain one end mapped to the region where the sense PCR primer of the first PCR is located and the other end mapped to a genomic region either on other chromosomes or distant from the LAM-PCR primer site.

[0008] In some embodiments, the MAGIC method comprises cleaving the DNA at a cleavage site that is within 10 and 1000 base pairs and 3' to an induced double strand break. This additional cleavage is typically initiated after the LAM-PCR, and between the first and second round of PCR amplification. The cleavage may be with an enzyme that cuts less than 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 5, 4, 3, 2, or 1 times in the DNA. In some

embodiments, the DNA is cleaved by an enzyme such as a Cas9, Cpfl, zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), meganuclease, and group one intron encoded endocunclease (GIIEE). This additional cleavage removes or reduces DNA fragments that do not contain translocation events.

[0009] In some embodiments, the invention comprises a method for detecting the location of double-strand breaks (DSBs) in DNA. In some aspects, the steps comprise: a. contacting DNA with a first primer and performing one or more cycles of linear amplification, wherein the primer comprises an internal cleavage site, and wherein there is no reverse primer, thereby producing single stranded DNA;

b. circularizing the amplified single stranded DNA to produce a single stranded DNA circle;

c. cleaving the circularized DNA at the first primer's internal cleavage site to linearize the DNA from step (b);

d. optionally amplifying the linearized DNA from step (c) with a second and third primer; and

e. analyzing the amplified DNA or the linearized DNA to determine the location of the DSB.

[00010] In some embodiments, the analyzing step comprises sequencing. In some aspects, the analyzing step comprises cloning the amplified or linearized DNA into a plasmid and sequencing the plasmid or a portion of the plasmid. In some embodiments, the analyzing step comprises an additional PCR with a fourth and fifth primer followed by sequencing. In some embodiments, the analyzing step comprises contacting the amplified or linearized DNA to known nucleic acids and detecting any hybridization of the amplified or linear DNA to the known nucleic acids.

[00011] In some embodiments, the DNA is sheared prior to step a). The DNA may be sheared to a size of about 1000, 750, 500, 250, 150, or 100 base pairs.

[00012] In some embodiments, the double-strand break is at a site that is different from an intended target site (off-target).

[00013] In some embodiments, the linearized DNA from step (c) is amplified.

Amplification may be via polymerase chain reaction (PCR) with a second and third primer. In some aspects, the second primer hybridizes at or near an end of the linearized DNA, the end optionally comprising part of the first primer, and the third primer hybridizes at or near the other end of the linearized DNA, the other end comprising part of the first primer.

[00014] In some embodiments, the method further comprises a second amplification via PCR with a fourth and a fifth primer, wherein the fourth and the fifth primer optionally comprise an adapter for sequencing.

[00015] In some aspects, the first LAM-PCR primer comprises a 5' phosphate. In some embodiments, the internal cleavage site is enzymatically or chemically cleaved. In some aspects, the internal cleavage site is a l ',2'-dideoxyribose modification site, such as an apurinic/apyrimidinic (AP) site. In some embodiments, the modification site is inducible such that it comprises a l ',2'-dideoxyribose modification site that is blocked by a protective group that when induced unblocks the site.

[00016] In some embodiments, the first LAM-PCR primer binds upstream or downstream of a potential double strand break and is oriented to allow extension of its 3' end toward the potential double strand break.

[00017] In some instances, the methods further comprise cleaving the DNA at a site that is within 10-1000 base pairs and 3' to the potential double strand break.

[00018] In some embodiments, the further cleavage is with an enzyme that makes DSBs in DNA. In some embodiments, the enzyme cuts less than 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 5, 4, 3, 2, or 1 times in the DNA. In some aspects, the enzyme is a nuclease such asCas9, Cpfl, zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), meganucleases, or group one intron encoded endonuclease (GIIEE).

[00019] In some embodiments, the meganuclease or GIIEE is I-Scel, I-Cre, I- Anil, I- Ceul, I-Chul, I-Cpal, I-CpaII, I-Dmol, H-Drel, I-Hmul, I-HmuII, I-Llal, I-Msol, PI-PfuI, PI- PkoII, I-Porl, I-Ppol, PI-PspI, I-Scal, PI-SceI, I-SceII, I-SecIII, I-SceIV, I-SceV, I-SceVI, I- SceVII, I-Ssp6803I, I-Tevl, I-TevII, I-TevIII, PI-Tlil, PI-THII, I-Tsp061I, and I-Vdil41I.

[00020] In some embodiments, the DNA is fragmented (e.g., sheared) prior to the first step of the method.

[00021] In some embodiments, any one or more of the first through fifth primers further comprises a tag that allows for purification or isolation of amplified single stranded DNA.

[00022] In some aspects, the tag comprises biotin, streptavidin, digoxigenin, a DNA sequence, or fluorescein isothiocyanate (FITC). In some aspects, prior to step b), the amplified DNA is contacted with a capture reagent that interacts with the tag to isolate DNA comprising the tag, and non-isolated DNA is discarded.

[00023] In some embodiments, the second and third primer comprise sequencing adaptors. In some embodiments, the fourth and fifth primer comprise sequencing adaptors.

[00024] In some instances, the DNA in step a) comprises about 50, 40, 30, 20, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, or 0.5 micrograms of DNA as input material. The DNA may be genomic or mitochondrial. In some instances, the DNA is mammalian. In some embodiments, the DNA is from a non-replicating cell. In some instances, the DNA is from a primary cell. In some instances, the DNA is from a subject treated with a DNA cutting enzyme. In some instances, the DNA is from a cell, tissue, or bodily fluid. The bodily fluid may be blood, serum, cerebral spinal fluid, sputum, lung secretions, or urine.

BRIEF DESCRIPTION OF DRAWINGS

[00025] Fig. 1 shows a schematic of endonuclease cleavage and DNA repair via nonhomologous end joining. Translocations (dark bars joined with lighter bars) can be detected in the methods of the invention to detect off-target cleavage events.

[00026] Fig. 2A shows a schematic of direct translocation capture, similar to Figure

1.

[00027] Fig. 2B shows a schematic of bait-based translocation capture, where a "bait" endonuclease is purposefully introduced into the test system to provide known landing sites for primers to detect on- and off -target cleavage events, e.g., for applications that would be universal for any "test" nuclease.

[00028] Fig. 3 shows a schematic of one embodiment of the invention wherein double-strand break sites (DSBs) are identified by linearly amplifying DNA with a primer comprising an internal apurinic/apyrimidinic (AP) site (small dot near the middle of the primer) and a 5' phosphate, circularizing the DNA, cleaving the circularized DNA with an enzyme that cleaves the AP site (i.e., APE 1), further amplifying the linearized DNA, and sequencing the amplified DNA.

[00029] Fig. 4 shows one embodiment of the invention wherein another sequence specific DNA endonuclease (e.g., I-Scel or Cas9 with a specific gRNA; cleavage site represented by triangles) is used to cleave DNA downstream of the original DNA

endonuclease to remove or reduce DNA fragments that do not contain translocation events. The DNA fragment containing a translocation is represented by the fragment having solid and dashed lines.

[00030] Fig. 5A shows a schematic of the first steps of the disclosed MAGIC method, where sheared genomic DNA linearly amplified with LAM-PCR primers designed to extend in the 3' direction toward a known endonuclease cleavage site. The RAGIB primers and cleavage site are exemplary only. The lighter colored chromosome joined to the darker colored chromosome indicates off-target DSB and translocation.

[00031] Fig. 5B shows a picture of a gel run to visualize DNA libraries produced from RAGlB.LAMpl and RAGlB.LAMp2 PCR2 primers as described in Example 1 and purified with Ampure XP beads. [00032] Fig. 6 shows that the indel frequency at the RAG IB on-target site was 49.3%, and the off-sites reported previously (OT1 and OT3) had 21.5% and 5.3% indel formation, respectively. The novel RAG1B off-target site (OT2) identified by the MAGIC method but not the HTGTS method showed an indel frequency of 12.5%. Cells that were not treated with RAG1B gRNA had less than 1% indel formation at any of OT1-OT3 (bars labelled "No treatment control). These results demonstrate that OT2 is a genuine RAG1B off- target site which was cleaved by Cas9/RAG1B gRNA in cells.

[00033] Fig. 7 shows another schematic of the MAGIC methods of the invention, where a DNA endonuclease (here shown as I-Scel) is introduced after completion of a first PCR reaction to amplify re-linearized ssDNA and before a second PCR reaction to add sequencing adaptors.

[00034] Fig. 8 shows a picture of a gel showing that I-Scel was able to cut fragments on each of three noncanonical sites on chromosomes 1, 11 and 16, as discussed in Example 2.

[00035] Fig. 9 shows a picture of a gel showing the results of MAGIC on genomic DNA isolated from HEK293/Cas9 cells transfected with G16-1 or G16-2 gRNA in the absence or presence of the RAG1B gRNA as described in Example 2.

[00036] Fig. 10 shows the translocation reads from the experiment described in Example 2, showing the improved sensitivity of the use of a rare cutting endonuclease I-Scel in the MAGIC method. I-Scel is exemplary, and the use of any rare cutting endonuclease, or CRISPR/cas9 is contemplated.

DESCRIPTION OF THE SEQUENCES

[00037] Table 1 provides a listing of certain sequences referenced herein.

OT1-R 9 GGAGTTCAGACGTGTGCTCTTCCGATCTTAAGGGCTTATCATCCAACTA

ACAGA

OT2-F 10 CACTCTTTCCCTACACGACGCTCTTCCGATCTTCAACTATGGCCATATT

CCTGG

OT2-R 11 GGAGTTCAGACGTGTGCTCTTCCGATCTGCTTCTAAATTAATGGGTTGT

TACTCC

OT3-F 12 CACTCTTTCCCTACACGACGCTCTTCCGATCTTAAATGATAGTACTTCC

AAGAAGAG

OT3-R 13 GGAGTTCAGACGTGTGCTCTTCCGATCTCATAATATGAAAGTGATTACG

CAG

G16-1 gRNA 14 UGCCAAGACUUGGACAGCACGUUUUAGAGCUAUGCUGUUUUG

(crRNA)

G16-2 gRNA 15 UAGCCAGAAGUUUGACCACAGUUUUAGAGCUAUGCUGUUUUG

(crRNA)

G16.LAMp 16 /5Phos/GCCTTTGTTCTCTCGACAGA/AP

site/CCAGGGCTGTTCCTCAAGAC

G16.LAMp-F 17 TACACGACGCTCTTCCGATCTGCTGTTCCTCAAGACCATTG

G16.LAMp-R 18 AGACGTGTGCTCTTCCGATCTTCTGTCGAGAGAACAAAGG mTTR gRNA 19 CCCAUACUCCUACAGCACCAGUUUUAGAGCUAGAAAUAGCAAGUUAAAA

UAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCUU UU

mTTR. LAMp 20 /5Phos/TTCACAGCCAACGACTCTGG/AP

site/CATCGCCACTACACCATCGC

mTTR. LAMp-F 21 TACACGACGCTCTTCCGATCTACTACACCATCGCAGCCCTG mTTR. LAMp-R 22 AGACGTGTGCTCTTCCGATCTCCAGAGTCGTTGGCTGTGA i501 23 AATGATACGGCGACCACCGAGATCTACACNNNNNNNNACACTCTTTCCC

TACACGACGCTCTTCCGATC*T

is phosphorothioate (PS) bond modification. i701 24 CAAGCAGAAGACGGCATACGAGATNNNNNNNNGTGACTGGAGTTCAGAC

GTGTGCTCTTCCGATC*T

is phosphorothioate (PS) bond modification.

On-target 25 CCCATACTCCTACAGCACCA ( CGG)

translocation

site

Off-target 26 CCCATGCTACTACAGCACCA (AGG)

translocation

site

tracrRNA 27 AACAGCAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAAC

UUGAAAAAGUGGCACCGAGUCGGUGCUUUUUUU

DETAILED DESCRIPTION

[00038] The methods described herein provide for the detection of double-strand breaks (DSBs) in DNA. The methods generally involve: linearly amplifying DNA with a primer comprising an internal apurinic/apyrimidinic (AP) site and a 5' phosphate,

circularizing the DNA, cleaving the circularized DNA with an enzyme that cleaves the AP site (i.e., APE 1), further amplifying the linearized DNA, and sequencing the amplified DNA. The target DNA to be analyzed may be genomic or mitochondrial DNA of a prokaryotic or eukaryotic cell, including humans. The method may be used to detect on- and off-target DSBs in samples from human and animal subjects, including subject receiving therapeutic agents capable of inducing DSBs.

Linear amplification

[00039] As used herein, linear amplification (sometimes referred to as "LAM" or "LA") refers to methods where the number of copies of amplification product increases linearly according to the number of amplification cycles. Each amplification cycle may involve a denaturation step to separate DNA duplexes, an annealing step to allow primer binding, and an elongation step. The denaturing and annealing steps may be carried out by using standard heating and cooling conditions, which are known to those of skill in the art. An example protocol of denaturing and annealing steps is denaturing at 98°C for 10s;

annealing at 67°C for 15s; and elongation at 72°C for 5s.

[00040] The annealing temperature may vary depending on the primer design and sequence. In addition, conditions for elongation may vary according to the desired sequence to be analyzed and the relative location of the linear amplification primer. The number of cycles performed can vary from 1 to 50 or more. For example, the number of cycles may be 1-50 cycles, 1-60 cycles, 1-70 cycles, 1-80 cycles, 1-90 cycles, or 1-100 cycles. The number of cycles may be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50, and so on increasing via multiples of 1 to about 100.

[00041] In certain cases, isothermal linear amplification methods (i.e., linear amplification approaches that do not require changing the reaction temperature) can be used. Example methods of isothermal linear amplification include loop-mediated isothermal amplification (LAMP), strand displacement amplification (SDA), helicase-dependent amplification, nicking enzyme amplification reaction, or other methods of isothermal linear amplification known to those skilled in the art. In some embodiments, strand displacement polymerase(s) are used for SDA. In some embodiments, the strand displacement polymerases are Bst DNA Polymerase, Large Fragment or Klenow Fragment (3 '-5' exo-).

Linear amplification primer

[00042] Only one linear amplification (LA) primer is used in the initial step of the methods described herein. The LA primer may be designed to amplify the desired DNA site to be analyzed (see Figures 2 and 7, for example). In certain embodiments, the primer is designed to extend its 3' end toward the on-target endonuclease cut site. Importantly, the LA primer has an internal cleavage site and a 5' modification to allow circularization of LAM- PCR products.

[00043] The internal cleavage site may comprise a series of nucleic acids that encode a recognition site for an enzyme that can cleave single- or double-stranded DNA. Such recognition sites are known to those of skill in the art. In some embodiments, the cleavage site is a l ',2'-dideoxyribose site, such as an apurinic/apyrimidinic (AP) site. In some embodiments, the cleavage site is inducible such that it is blocked by a protective group that when induced unblocks the site. The internal cleavage site is unique within the primer, i.e., there is only one recognition site for a particular enzyme within one primer so that application of the enzyme will result in only one cut.

[00044] The 5' modification to allow circularization may be a phosphate, although any modification known to those of skill in the art as allowing circularization of single- stranded DNA is contemplated.

[00045] Generally, the primer may be as far upstream or downstream as permitted by the maximum processivity of the polymerase being employed, e.g., over 10,000 nucleotides. In some embodiments, the primer may be 50 to 300 nucleotides away from the expected DSB site. In some embodiments, the primer may be at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210, at least 220, at least 230, at least 240, at least 250, at least 260, at least 270, at least 280, at least 290, or at least 300 nucleotides away from the expected DSB site. In each instance, however, the primer is designed to anneal to the DNA at a site that is upstream or downstream of the expected DSB site so that it can extend its 3' end past the repaired DSB site junction.

[00046] Primers may be optimized for GC content, melting temperature, and amplification efficiency.

[00047] In certain circumstances, the primer (e.g., the pool of primers) can include a molecular barcode(s), e.g., a unique molecular identifier (UMI). The molecular barcodes may comprise one or more target specific regions, label regions, sample index regions, universal PCR regions, adaptors, linkers, or any combination of these elements. In some embodiments, the molecular barcode may be a sample tag or a molecular identifier label. In some embodiments, multiple molecular barcodes may be used. In some embodiments, the molecular barcode may comprise one or more oligonucleotides. In some embodiments, the molecular barcode may be at least about 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more nucleotides in length.

[00048] In some embodiments, the molecular barcode may comprise a random nucleotide sequence. In some embodiments, the random nucleotide sequence may be computer generated. In some embodiments, the random nucleotide sequence may have no pattern associated with it.

[00049] In some embodiments, the molecular barcode may comprise a non-random nucleotide sequence, wherein the nucleotides comprise a pattern. In some embodiments, the molecular barcode may be a commercially available sequence.

[00050] The molecular barcode may comprise one or more secondary structures. The molecular barcode may comprise a hairpin structure.

[00051] In certain embodiments, the linear amplification primers include a tag. In some embodiments the tag is a purification tag that may be associated with a capture reagent such that the tag-capture reagent interaction is stronger than standard DNA-DNA interactions to allow for proper purification of the linear amplification products resulting from such tagged primers. The tag should be of a size and structure such that it will not interfere with binding or priming of an amplification reaction. In certain embodiments, the tag is at the 5' end of the primer. In some embodiments, the tag is at the 3' end of the primer. In some embodiments, the tag is at an internal nucleotide site. The tag may be attached, conjugated, or otherwise associated with the primer by methods known to those of skill in the art.

Exemplary tags include biotin, streptavidin, digoxigenin, a DNA sequence, or fluorescein or a derivative of fluorescein such as fluorescein isothiocyanate (FITC).

[00052] Capture reagents are chosen according to the tag. For example, a biotin tag may utilize a streptavidin capture. A streptavidin tag may utilize a biotin capture. A digoxigenin tag may utilize an anti-digoxigenin antibody or Fab fragment capture(s). A DNA sequence tag may utilize a complementary DNA sequence capture. A fluorescein or a derivative of fluorescein tag, such as FITC tag, may utilize an anti-FITC or anti -fluorescein antibody or Fab fragment capture. In this application, "FITC" is used as an example fluorescein derivative. As such, the term "FITC" would be interchangeable with any fluorescein derivative, including 6-FAM (6-carboxyfluorescein), carboxylates, and succinimidyl esters. [00053] The tags and capture reagents are interchangeable such that in some instances the capture reagent may be called a capture reagent and vice versa - the tag-capture pairing is key.

[00054] A purification or isolation step may be utilized to allow for removal of any non-tagged DNA or RNA. Only amplification products primed from the desired tagged LA primer will be captured after purification/isolation. Generally, purification or isolation processes can be designed to separate any untagged nucleic acids (e.g., free template DNA, genomic DNA, RNA) from the tagged linear amplification products. In certain examples, the capture reagent can be on beads, a solid surface, or a column. Wash steps may be designed to reduce nonspecific binding. For example, when using a biotin-tagged linear amplification primer, the linear amplification products can be purified using streptavidin beads and high salt (e.g., 1 M NaCl) and/or high pH buffer washes. In some embodiments, the purification or isolation step comprises contacting the linear amplification product or products with a capture reagent that is specific for the tag used on the primer. In some embodiments, the purification or isolation step comprises discarding any DNA or RNA that is not captured. In some embodiments, the purification or isolation step comprises washing the capture-tag so as to remove essentially all of the non-tagged DNA and RNA.

[00055] In some embodiments, the tag-capture pair is biotin and streptavidin. In some embodiments, biotin is the tag and streptavidin is the capture reagent. In some embodiments, more than one biotin may be used per primer as a tag. In some embodiments, the biotin tag is added using photoreactive biotinylation reagents. In some embodiments, streptavidin is the tag and biotin is the capture reagent. In some embodiments, more than one streptavidin may be used per primer as a tag.

[00056] In one embodiment, the tag-capture is digoxigenin and an anti-digoxigenin antibody or Fab fragment. In some embodiments, digoxigenin is the tag and an anti- digoxigenin antibody or Fab fragment is the capture. In some embodiments, more than one digoxigenin may be used per primer as a tag. In some embodiments, digoxigenin is attached to a 5 '-amino substituted primer. In some embodiments, an anti-digoxigenin antibody or Fab fragment is the tag and digoxigenin is the capture. In some embodiments, more than one anti- digoxigenin antibody or Fab fragment may be used per primer as a tag.

[00057] In one embodiment, the tag-capture is a DNA sequence tag and a

complementary DNA sequence capture. In some embodiments, the complementary DNA sequence capture nucleotides are contained on microspheres. In some embodiments, the microspheres are oligonucleotide-coupled polystyrene microparticles, which may be termed "beads." In some embodiments, the tag is a "TAG" sequence and the capture is performed using MagPlex-TAG™ microspheres (Luminex) with an "anti-TAG" sequence.

[00058] In one embodiment, the tag-capture is fluorescein or a derivative of fluorescein and an anti-FITC or anti -fluorescein antibody or Fab fragment. In some embodiments, the fluorescein or derivative of fluorescein is FITC, 6-FAM (6- carboxyfluorescein), carboxylates, and succinimidyl esters. In some embodiments, the fluorescein or derivative of fluorescein is the tag and an anti-FITC and anti-fluorescein antibody or Fab fragment is the capture. In some embodiments, more than one fluorescein or derivative of fluorescein may be used per primer as a tag. In some embodiments, an anti- FITC or anti-fluorescein antibody or Fab fragment is the tag and fluorescein or derivative of fluorescein is the capture. In some embodiments, more than one anti-FITC or anti-fluorescein antibody or Fab fragment may be used per primer as a tag.

[00059] A number of other tag-capture combinations would be known to those skilled in the art. It should be appreciated that antibodies and aptamers against a wide range of specific targets are available to those skilled in the art and may be commercially available. In some embodiments, aptamer sequences for use as a tag may be designed to bind to a capture protein. Thus, the use of this invention is not limited by the particular tag-capture combinations recited herein.

[00060] The linear amplification products can be separated from the capture reagent or the tag prior to further amplification, or the further amplification may be performed in the presence of the capture reagent and tag.

Single-stranded DNA circularization and re-linearization

[00061] The ssDNA products of the LAM-PCR reaction are circularized at the LAM- PCR primer's 5' modification site. Methods to ligate single stranded DNA to form a circle are known to those of skill in the art and are contemplated herein. In some embodiments, the ligation reaction uses a ssDNA ligase, such as, for example, intra-molecular ligation by CircLigase™ ssDNA Ligase.

[00062] The LAM-primer used in the LAM-PCR reaction is incorporated into the single-stranded DNA amplification products allowing circularization of the linear products. The circularized products comprise a single unique cleavage site (derived from the primer) that can be utilized to re-linearize the ssDNA. See, Figures 3 and 7. Re-linearization at the LAM-primer' s internal cleavage site is key to the MAGIC method to provide known sequences for primer binding at one or both ends of the re-linearized ssDNA useful in further amplification processes.

Analysis of re-linearized ssDNA

[00063] Analysis of the linear amplification products may involve unbiased methods that provide information on the nature and location of each DSB and translocation event. In some embodiments, the re-linearized DNA can be analyzed for location of DSBs. In some embodiments, the re-linearized DNA is further amplified prior to analyzing for DSBs.

Analysis can be performed using methods such as next generation sequencing (NGS). In certain embodiments, the re-linearized ssDNA or further amplified products can be contacted with known DNA sequences and hybridization can be detected. Hybridization of the re- linearized or further amplified DNA to known sequences can provide information on location of DSBs. In further embodiments, the linear amplification or further amplified products can be cloned into a plasmid and sequenced by Sanger sequencing.

[00064] In some embodiments, the analysis of the re-linearized ssDNA comprises an amplification step comprising a first, non-linear, PCR reaction using a second and third primer that are distinct from the first LAM-PCR primer. The second primer hybridizes at or near an end of the linearized ssDNA, wherein the end comprises a portion of the first LAM- PCR primer. The third primer hybridizes at or near the other end of the linearized ssDNA, wherein the other end comprises a different portion of the first LAM-PCR primer.

[00065] In some embodiments, the second primer has a 3' end that hybridizes to portions of the first LAM-PCR primer that resides at the end of the re-linearized ssDNA. In some embodiments, the second primer does not hybridize to any portion of the first LAM- PCR primer. In some embodiments, the 5' end of the second primer does not hybridize to any portion of the first LAM-PCR primer. In some embodiments, the third primer hybridizes at or near the other end of the linearized ssDNA (i.e., the end opposite that to which the second primer binds). The third primer has a 3' end that hybridizes to portions of the first LAM-PCR primer that reside at the end of the linearized ssDNA. In some embodiments, the 5' end of the third primer does not hybridize to any portion of the first LAM-PCR primer. In some embodiments, the second and third primer comprise adapter sequences to assist in

sequencing. In some embodiments, the second and third primer comprise tags for

purifi cati on/i sol ati on . [00066] In some embodiments, a second, non-linear, PCR reaction using a fourth and a fifth primer is provided, wherein the fourth and the fifth primer optionally comprise an adapter for sequencing. In some embodiments, the fourth and the fifth primer are designed so that their 3' end binds to a portion of the second and third primer. In some embodiments, the fourth and fifth primer bind to the unique portions of the second and third primer, i.e., to the portions of the third and fourth primer that are not complementary to the LAM-PCR primer. See, Figures 3

and 7.

[00067] One or more of the second through fifth primers may comprise an adaptor, such as, for example, a next generation sequencing adapter. One or more of the second through fifth primers may comprise a tag to assist in isolation/purification.

[00068] In certain circumstances, the primer (e.g., the pool of primers) can include a molecular barcode(s). The molecular barcodes may comprise one or more target specific regions, label regions, sample index regions, universal PCR regions, adaptors, linkers, or any combination of these elements. In some embodiments, the molecular barcode may be a sample tag or a molecular identifier label. In some embodiments, multiple molecular barcodes may be used. In some embodiments, the molecular barcode may comprise one or more oligonucleotides. In some embodiments, the molecular barcode may be at least about 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more nucleotides in length.

[00069] In some embodiments, the molecular barcode may comprise a random nucleotide sequence. In some embodiments, the random nucleotide sequence may be computer generated. In some embodiments, the random nucleotide sequence may have no pattern associated with it.

[00070] The molecular barcode may comprise one or more secondary structures. The molecular barcode may comprise a hairpin structure.

[00071] The phrase "next generation sequencing" refers to non-S anger-based sequencing technologies having increased throughput, for example with the ability to generate hundreds of thousands of relatively small sequence reads at a time. Some examples of next generation sequencing techniques include, but are not limited to, sequencing by synthesis, sequencing by ligation, and sequencing by hybridization. Some relatively well- known next generations sequencing methods further include pyrosequencing developed by 454 Corporation, the Solexa system, and the SOLiD (Sequencing by Oligonucleotide Ligation and Detection) developed by Applied Biosystems (now Life Technologies, Inc.), and the Sequel System (PacBio^®).

Optional further cleavage and DNA fragmenting

[00072] In some embodiments, the DNA being tested via the MAGIC method is cleaved with an agent, e.g., an enzyme, that makes infrequent double-stranded breaks (DSBs) in genomic or mitochondrial DNA. In some embodiments, the agent cleaves the DNA within 10 to 1000 base pairs and 3' to an induced double-strand break. In some embodiments, the agent cleaves the DNA 10 to 500, 10 to 250, 10 to 200, 10 to 150, 10 to 100, 10 to 50, or 10- 25 base pairs away from and 3' to the induced double-strand break. In some embodiments, the agent cleaves 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 base pairs away from and 3' to the induced double-strand break site.

[00073] In some embodiments, the agent cuts less than 10,000, 5,000, 1,000, 500, 400, 300, 200, 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 5, 4, 3, 2, or 1 times in genomic DNA. In some embodiments, the agent cuts less than 5, 4, 3, 2, or 1 times in re-linearized ssDNA or in the further amplified DNA.

[00074] In some embodiments, the agent is an enzyme such as a DNase. In some embodiments, the enzyme is a Cas9, Cpfl, zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), meganuclease, or group one intron encoded endonuclease (GIIEE).

[00075] In some embodiments, the meganuclease or GIIEE is I-Scel, I-Cre, I-Anil, I- Ceul, I-Chul, I-Cpal, I-Cpall, I-Dmol, H-Drel, I-Hmul, I-HmuII, I-Llal, I-Msol, ΡΙ-PfuI, PI- PkoII, I-Porl, I-Ppol, PI-PspI, I-Scal, Pl-Scel, I-Scell, I-SecIII, 1-SceIV, I-SceV, I-SceVI, I- SceVII, I-Ssp6803I, I-Tevl, I-TevII, I-TevIII, PI-Tlil, PI-THII, I-Tsp061I, or I-Vdil41I.

[00076] In some embodiments, cleavage with the agent improves efficiency of the MAGIC method. See, e.g., Figure 10.

[00077] In some embodiments, the DNA is fragmented (e.g., sheared) prior to linear amplification. The shearing may produce DNA fragments having about 1000, 750, 500, 250, 150, or 100 base pairs.

Samples for Testing

[00078] In some embodiments, the DNA to be tested is from a replicating or non- replicating cell. In some instances, the DNA to be tested is genomic DNA from a subject treated with a DNA cutting enzyme. The subject may be mammal or non-mammal. The mammalian subject may be human, dog, cat, horse, cattle, goat, deer, or other livestock. The non-mammalian subject may be a reptile or bird, e.g., a chicken, hen, or rooster. In some embodiments, the genomic DNA is from a cell, tissue, or bodily fluid. In some instances, the bodily fluid is blood, serum, cerebral spinal fluid, sputum, lung secretions, or urine. In some instances, DNA is isolated from a tissue biopsy from a subject treated with a DNA cutting enzyme.

[00079] The words "a", "an" or "the" when used in conjunction with the term "comprising" in the claims and/or the specification may mean "one," but each is also consistent with the meaning of "one or more," "at least one," and "one or more than one." The use of "or" means "and/or" unless stated otherwise. The use of the term "including" and "containing," as well as other forms, such as "includes," "included," "contains," and

"contained" is not limiting. All ranges given in the application encompass the endpoints unless stated otherwise.

EXAMPLES

Example 1: Direct Translocation Capture

[00080] HEK293 cells stably expressing Cas9 (HEK293/Cas9) were transfected with a plasmid that expresses a sgRNA (SEQ ID NO: 1) targeting the RAGIB gene under the control of the human U6 promoter. RAGIB is the same gene previously evaluated in the context of potential Cas9 off-targets by Frock et al. (Nature Biotechnology 33, 179-186 (2015) PMTD: 25503383) using the high-throughput, genome-wide, translocation sequencing (HTGTS) method. Genomic DNA was purified 24 hours post transfection using PureLink® Genomic DNA Mini Kit (Therm oFisher) and concentration was quantified using a NanoDrop spectrophotometer (Thermo Scientific). Aliquots of 10 μg of genomic DNA were used in the liner amplification-mediated PCR (LAM-PCR) using 10 nM of RAGIB. LAMpl (SEQ ID NO:2) or RAGlB.LAMp2 (SEQ ID NO:3) with an annealing temperature of 68 °C for 80 cycles (Figure 5 A). Note that the method of Frock et al. using HTGTS requires at least 50 μg of genomic DNA.

[00081] Amplified single-stranded DNAs were purified with Ampure XP beads (Beckman Coulter) and eluted in 45 μΙ_, of water followed by intra-molecular ligation by CircLigaseTM ssDNA Ligase (Epicentre) according to manufacturer's condition at 60 °C for two hours. Circularized ssDNA were purified with Ampure XP beads and eluted in 50 μΙ_, of water. Circular ssDNA molecules were re-linearized by the APE 1 enzyme (20 units), and purified with Ampure XP beads. Purified linear DNA was used in the PCR1 reaction using Q5 enzyme. RAGlB.LAMpl-F (SEQ ID NO:4) and RAGlB.LAMpl -R (SEQ ID NO:5) primers were used for samples generated from the RAGlB.LAMpl linear amplification, and RAGlB.LAMp2-F (SEQ ID NO:6) and RAGlB.LAMp2-R (SEQ ID NO:7) primers were used for samples generated from the RAGlB.LAMp2 linear amplification. For PCR1, the reactions were cycled for 35 rounds at 98°C for 10 seconds, 62°C for 20 seconds and 72°C for 15 seconds. The resulting PCR products were purified with Ampure XP beads and used for a second round of PCR amplification (PCR2) with NGS index i5 (SEQ ID NO:23) and i7 (SEQ ID NO:24) PCR primers. For PCR2, the reactions were cycled for 25 rounds at 98°C for 10 seconds, 65°C for 20 seconds and 72°C for 15 seconds. The DNA libraries from PCR2 were purified with Ampure XP beads (Figure 5B) and quantified with KAPA Library Quantification Kit (Kapa Biosystems). The purified DNA libraries generated from

RAGlB.LAMpl or RAGlB.LAMp2 were then analyzed on the MiSeq sequencer (Illumina). Sequencing data were mapped to a reference human genome (e.g., hg38) to identify translocation reads.

[00082] In this set of experiments, three off-target sites (denoted RAGIB-OTI, RAG1B-OT2 and RAG1B-OT3) were identified in both libraries (Table 2).

[00083] Table 2: Results of MAGIC assay to identify on- and off-target DSB events as compared to HTGTS method

[00084] The OT1 and OT3 sites were previously identified as the only RAG1B off- target sites by Frock et al. using the HTGTS method. However, the method provided herein identified an additional RAGIB off-target site (OT2).

[00085] To validate whether OT2 is a genuine RAGIB off -target site, genomic DNAs from HEK293/Cas9 cells or HEK293/Cas9 cells transfected with a plasmid expressing RAGIB gRNA were used to amplify a genomic region surrounding the OT2 site with PCR primers OT2-F (SEQ ID NO: 10) and OT2-R (SEQ ID NO: 11). The resulting DNA fragments were sequenced by NGS to determine insertion/deletion (indel) frequency. The indel frequency at the RAGIB on-target site was 49.3%, and the off-sites reported previously (OTl and OT3) had 21.5% and 5.3% indel formation, respectively. The novel RAGIB off-target site (OT2) showed an indel frequency of 12.5% (Figure 6). Cells that were not treated with RAGIB gRNA had less than 1% indel formation at any of OT1-OT3. These results demonstrated that OT2 is a genuine RAGIB off -target site which was cleaved by

Cas9/RAG1B gRNA in cells.

Example 2: Bait-based Translocation Capture with I-Scel Digestion: RAGIB Off- Target Sites

[00086] To enrich for NGS reads that contain chromosome translocations, sequence- specific DNA digestion can be introduced to cut DNA fragments that do not contain translocations (Figure 4). The number of DNase recognition sites in the genome should be limited and the site cannot be present in the region where translocation events occur. There are few noncanonical I-Scel cleavage sites present in the human genome (Petek et. al, Mol Ther 2010). To first show that I-Scel can cut at these noncanonical sites, PCR primers that anneal on either side of the sites (on chromosomes 1, 11 and 16) were used to amplify DNA fragments containing the sites. The amplified PCR products were purified and incubated with I-Scel, and I-Scel was able to cut all three fragments (Figure 8). Two gRNAs (e.g., crRNA sequences G16-1 (SEQ ID NO: 14) and G16-2 (SEQ ID NO: 15)) were designed to cleave around the noncanonical I-Scel site on chromosome 16. One linear amplification primer (G16.LAMp (SEQ ID NO: 16)) was designed according to the criteria described previously.

[00087] Genomic DNA isolated from HEK293/Cas9 cells transfected with gRNA (e.g., crRNAs G16-1 or G16-2) and tracrRNA (SEQ ID NO: 27) in the absence or presence of the RAGIB gRNA was purified using PureLink® Genomic DNA Mini Kit

(ThermoFisher). Concentration of genomic DNA was quantified using a NanoDrop spectrophotometer. Aliquots of 5 μg of genomic DNA were used in the liner amplification- mediated PCR (LAM-PCR) using 10 nM of G16.LAMp with an annealing temperature of 67°C for 50 cycles (Figure 5A). Amplified single-stranded DNAs were purified with Ampure XP beads and eluted in 45 μΙ_, of water followed by intra-molecular ligation by CircLigaseTM ssDNA Ligase (Epicentre) according to manufacturer's condition at 60°C for two hours. Circularized DNA were purified with Ampure XP beads and eluted in 50 μΙ_, of water. Circular ssDNA molecules were re-linearized by the APE 1 enzyme (20 units), and purified with Ampure XP beads. Purified linear DNA was used in the PCR1 reaction using Q5 enzyme with proper PCR primers. G16.LAMp-F (SEQ ID NO: 17) and G16.LAMp-R (SEQ ID NO: 18) primers were used on re-linearized DNA to amplify the intermediate libraries. PCR reactions were cycled 35 rounds at 98°C for 10 seconds, 63 °C for 20 seconds and 72°C for 15 seconds. PCR products were purified with Ampure XP beads. Eluted DNA was treated with I-Scel to cleave the DNA fragments that did not contain translocations. Cleaved products were purified with Ampure XP and used for the second round of PCR amplification (PCR2) with NGS index i5 (SEQ ID NO:23) and i7 PCR (SEQ ID NO:24) primers. PCR reactions were cycled 25 rounds at 98°C for 10 seconds, 65°C for 20 seconds and 72°C for 15 seconds. The DNA libraries from PCR2 were purified with Ampure XP beads (Figure 9) and quantified with KAPA Library Quantification Kit (Kapa Biosystems). DNA libraries were sequenced on MiSeq according to the manufacturer's instructions. The I- Scel treatment increased translocation reads from less than 0.5% to the range of 8% - 17%. In other words, this step increased the efficiency more than ten times which allows for the identification of endonuclease mediated events using ten times less sequencing reads. Both G16-1 and G16-2 cleavage sites captured all three RAG1B off-target sites identified in the previous Example, where RAGIB off -target sites were captured by the RAGIB on-target cleavage site (Table 3).

[00088] Table 3

Identified sites Chromosome G16-1 Method G16-2 Method

Identified Identified

1 RAGIB. OT1 14 Identified Identified

1 RAGIB. OT2 12 Identified Identified

1 RAGIB. OT3 4 Identified Identified

[00089] Figure 10 shows the translocation reads from this experiment, showing the improved sensitivity of the use of a rare cutting endonuclease in the MAGIC method. Example 3: Direct Translocation Capture: In Vivo Samples

[00090] Using the methods provided herein, it is also possible to identify DSBs in vivo, for example in animals that have been treated with an endonuclease such as Cas9. In this Example, the livers from mice that were treated with lipid nanoparticles (LNPs) comprising Cas9 mRNA and gRNA (SEQ ID: 19) targeting the TTR gene were utilized. Genomic DNA was purified from the liver tissues using the PureLink® Genomic DNA Mini Kit (ThermoFisher). Concentration of genomic DNA was quantified using a NanoDrop spectrophotometer. Aliquots of 2.5 μg of genomic DNA were used in the liner amplification- mediated PCR (LAM-PCR) using 10 nM of mTTR.LAMp (SEQ ID NO: 20) with an annealing temperature of 68°C for 80 cycles (Figure 5A). Amplified single-stranded DNAs were purified with Ampure XP beads and eluted in 45 μΐ. of water followed by intramolecular ligation by CircLigaseTM ssDNA Ligase (Epicentre) according to manufacturer's instructions at 60°C for two hours. Circularized DNA were purified with Ampure XP beads and eluted in 50 μΐ. of water. Circular ssDNA molecules were re-linearized by the APE 1 enzyme (20 units), and purified with Ampure XP beads. Purified linear DNA was used in the PCR1 reaction using Q5 enzyme with proper PCR primers. mTTR.LAMP-F (SEQ ID NO:21) and mTTR.LAMP-R (SEQ ID NO:22) primers were used for samples generated from the mTTR.LAMp linear amplification. The PCR1 reactions were cycled for 35 rounds at 98°C for 10 seconds, 66°C for 20 seconds and 72°C for 15 seconds. PCR1 products were purified with Ampure XP beads. Eluted DNA was used for second round of PCR

amplification (PCR2) with NGS index i5 (SEQ ID NO:23) and i7 PCR (SEQ ID NO:24) primers. The PCR2 reactions were cycled for 25 rounds at 98°C for 10 seconds, 65°C for 20 seconds and 72°C for 15 seconds. The DNA libraries from PCR2 were purified with Ampure XP beads (Figure 5B) and quantified with KAPA Library Quantification Kit (Kapa

Biosystems). DNA libraries were analyzed on the MiSeq sequencer (Illumina). Sequencing data were mapped to a mouse reference genome (e.g., mm 10) to identify translocation reads. One off-target site on chromosome 16 was identified for the TTR guide used. This off-target site contains two mismatched nucleotides as compared to the guide recognition region (compare SEQ ID NO:25 (guide recognition sequence) to SEQ ID NO:26 (off target site containing two mismatches)).

[00091] This off-target site was validated in a similar approach described in Example 1. Genomic DNA from the livers of mice that were treated with lipid nanoparticles (LNPs) comprising Cas9 mRNA and sgRNA (SEQ ID: 19) targeting the TTR gene was purified as described above and primers were used to amplify a genomic region surrounding the potential off-target site. NGS analysis showed an average indel rate of 1.8% from five treated animals (data not shown). This finding indicated the identified site is a genuine TTR off- target site.

Claims

CLAIMS What is claimed is:

1. A method for detecting the location of double-strand breaks in DNA comprising:

a. contacting DNA with a first primer and performing one or more cycles of linear amplification, wherein the primer comprises an internal cleavage site, and wherein there is no reverse primer, thereby producing single stranded DNA;

d. optionally amplifying the linearized DNA from step (c); and

e. analyzing the amplified DNA, or the linearized DNA, to determine the

location of double strand breaks.

2. The method of claim 1, wherein the analyzing step comprises:

a. sequencing the amplified or linearized DNA;

b. cloning the amplified or linearized DNA into a plasmid and sequencing the plasmid or a portion of the plasmid; or

c. contacting the amplified or linear DNA to known nucleic acids and detecting any hybridization of the amplified or linear DNA to the known nucleic acids.

3. The method of claim 1, wherein prior to step a) the DNA is sheared.

4. The method of claim 3, wherein the DNA is sheared to a size of about 1000, 750, 500, 250, 150, or 100 base pairs.

5. The method of claim 1, wherein the linearized DNA from step (c) is amplified.

6. The method of claim 5, wherein the amplification is via polymerase chain reaction (PCR) with a second and third primer, wherein the second primer hybridizes at or near an end of the linearized DNA, the end optionally comprising part of the first primer, and the third primer hybridizes at or near the other end of the linearized DNA, the other end comprising part of the first primer.

7. The method of claim 6, wherein the method further comprises a second amplification via PCR with a fourth and a fifth primer, wherein the fourth and the fifth primer optionally comprise an adapter for sequencing.

8. The method of claim 1, wherein the first primer further comprises a 5' phosphate.

9. The method of claim 1, wherein the internal cleavage site is enzymatically or chemically cleaved.

10. The method of claim 1, wherein the internal cleavage site is a l ',2'-dideoxyribose site.

11. The method of claim 10, wherein the l ',2'-dideoxyribose site is an

apurinic/apyrimidinic (AP) site.

12. The method of claim 1, wherein the internal cleavage site comprises an inducible cleavage site.

13. The method of claim 12, wherein the inducible cleavage site is an inducible AP site comprising an AP site that is blocked by a protective group that when induced unblocks the AP site.

14. The method of claim 1, wherein the first primer binds upstream or downstream of a potential double strand break and is oriented to allow extension of its 3' end toward the potential double strand break.

15. The method of claim 1, further comprising cleaving the DNA at a cleavage site that is within 10-1000 bp and 3' to an induced double strand break.

16. The method of claim 15, wherein cleavage is after the linear amplification step a).

17. The method of claim 16, wherein the cleavage is with an enzyme that cuts less than 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 5, 4, 3, 2, or 1 times in the DNA.

18. The method of claim 14, wherein the DNA is cleaved by an enzyme selected from Cas9, Cpfl, zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), meganuclease, and group one intron encoded endonuclease (GIIEE).

19. The method of claim 15, wherein the meganuclease or GIIEE is I-Scel, I-Cre, I- Anil, I-Ceul, I-Chul, I-Cpal, I-CpaII, I-Dmol, H-Drel, I-Hmul, I-HmuII, I-Llal, I-Msol, Pl-Pful, PI-PkoII, I-Porl, I-Ppol, PI-PspI, I-Scal, Pl-Scel, I-SceII, I-SecIII, I-SceIV, I-SceV, I-SceVI, I-SceVII, I-Ssp6803I, I-TevI, I-TevII, I-TevIII, PI-TliI, PI-TliII, I-Tsp061I, or I-Vdil41I.

20. The method of claim 1, wherein the first primer further comprises a tag that allows for purification of amplified single-stranded DNA.

21. The method of claim 20, wherein the tag comprises biotin, streptavidin, digoxigenin, a DNA sequence, or fluorescein isothiocyanate (FITC).

22. The method of claim 20, wherein prior to step b), the amplified DNA is contacted with a capture reagent that interacts with the tag to isolate DNA comprising the tag, and nonisolated DNA is discarded.

23. The method of claim 6, wherein the second and third primer comprise sequencing adaptors.

24. The method of claim 1, wherein the double strand break is at a site that is different from an intended target site (off-target).

25. The method of claim 1, wherein the DNA in step a) comprises about 50, 40, 30, 20, 10, 5, 4, 3, 2, 1, 0.5 micrograms of DNA.

26. The method of claim 1, wherein the DNA is genomic or mitochondrial.

27. The method of claim 1, wherein the DNA is mammalian.

28. The method of claim 1, wherein the DNA is from a non-replicating cell.

29. The method of claim 1, wherein the DNA is from a primary cell.

30. The method of claim 26, wherein the genomic DNA is from a subject treated with a

DNA cutting enzyme.

31. The method of claim 30, wherein the genomic DNA is from a cell, tissue, or bodily fluid.

32. The method of claim 31, wherein the bodily fluid is blood, serum, cerebral spinal fluid, sputum, lung secretions, or urine.