WO2015066205A1

WO2015066205A1 - Fast genetically modified organism and systems for precise genome editing and targeted genetic manipulation through enhanced homologous recombination

Info

Publication number: WO2015066205A1
Application number: PCT/US2014/062950
Authority: WO
Inventors: William J. BELDEN; Tammy M. JOSKA
Original assignee: Rutgers, The State University Of New Jersey
Priority date: 2013-10-29
Filing date: 2014-10-29
Publication date: 2015-05-07

Abstract

Methods and compositions for increasing homologous recombination disclosed.

Description

FAST GENETICALLY MODIFIED ORGANISM AND SYSTEMS FOR PRECISE GENOME EDITING AND TARGETED GENETIC MANIPULATION THROUGH ENHANCED HOMOLOGOUS

RECOMBINATION

This application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No. 61/896,825, filed October 29, 2013. The foregoing application is incorporated by reference herein.

FIELD OF THE INVENTION

The present invention relates to the field of gene recombination. More specifically, the invention provides compositions, systems, and methods for enhanced homologous recombination, e.g., for the purpose of precise gene conversion or genome editing in broad organisms. This invention includes the systems, compositions and methods for altering target gene sequences with precision and accuracy for use in mutagenesis and/or altering expression of a target gene sequence. The invention relates to the development of genetic models of human disease, gene therapy and/or the development of model research organisms.

BACKGROUND OF THE INVENTION

Several publications and patent documents are cited throughout the specification in order to describe the state of the art to which this invention pertains. Each of these citations is incorporated herein by reference as though set forth in full.

Homologous recombination (HR) allows targeted genetic manipulation of organisms and is an invaluable tool in biotechnology laboratories. Unfortunately, targeted HR is very inefficient or non-existent in most organisms because nonhomologous end joining (NHEJ) is the preferred method of genome repair. This represents an obstacle in more complex species (especially vertebrates), leaving many researchers unable to perform genetic manipulations common in E. coli and

Saccharomyces cerevisiae. To circumvent the deficiencies in higher eukaryotes and metazoans, enhanced methods of homologous recombination are desired. SUMMARY OF THE INVENTION

In accordance with the present invention, compositions and methods for increasing homologous recombination, particularly non-homologous end joining, within a cell are provided. In a particular embodiment, the method comprises delivering to the cell Rad51 and/or Rad52 (particularly yeast Rad51 and/or Rad52). The method may further comprise the delivery of at least one other protein (or nucleic acid encoding the protein) that supports homologous recombination. Examples of proteins that support homologous recombination include, without limitation, meiotic recombination- 11 (MRE11), PARP, Rad50, Rad54, nijmegen breakage syndrome-1 (NBS-1), RPA, BRCA1 and BRCA2 RNA. The method may further comprise delivering at least one inhibitor of a component of the NHEJ pathway. Examples of components of NHEJ include, without limitation, LiglV, XRCC4, XRCC7, Ku70, Ku80, DNA-dependent protein kinase (DNA-PKC), and NBS1. In a particular embodiment, the inhibitor is selected from the group consisting of a LiglV inhibitor, a XRCC4 inhibitor, a Ku70 inhibitor, and a Ku80 inhibitor. In a particular

embodiment, the method comprises delivering to the cell Rad51, Rad52, a LiglV inhibitor, a XRCC4 inhibitor, and a Ku70 inhibitor or a Ku80 inhibitor. In a particular embodiment, the cell is a zebrafish cell, such as a fertilized zebrafish egg. Rad51 and Rad52 may be delivered to the cell as proteins or as nucleic acid molecules. In a particular embodiment, the inhibitors are inhibitory nucleic acid molecules, particularly antisense oligonucleotides such as morpholino

oligonucleotides. The methods may further comprise delivering a nucleic acid molecule to the cell for homologous recombination. In a particular embodiment, the nucleic acid molecule comprises 5' and 3' flank regions (e.g., about 100 bp to 10,000 bases, particulary about 500 bp to about 5 kb or about lkb to about 3 kb) that match the target sequence within the genome of the cell and a selectable marker and/or fluorescence marker and/or kill gene or other detectable marker. In a particular embodiment, the nucleic acid molecule comprises a 5' flank, a nucleic acid encoding a first detectable protein operably linked to a promoter, a selectable marker, an origin or replication, a 3 ' flank, and a nucleic acid encoding a second detectable protein operably linked to a promoter, wherein the 5' flank and the 3' flank specifically hybridize with target sequences within the genome of the cell. The nucleic acid molecule may further comprise a multiple cloning site or a nucleic acid sequence of interest. BRIEF DESCRIPTIONS OF THE DRAWING

Figure 1 provides a schematic of HR cassette construction.

Figure 2 provides images of the fluorescence prescreening. GFP+RFP- (left) and GFP+RFP+ (right) zebrafish are shown to demonstrate HR versus ectopic insertions.

Figure 3 provides an image of a Southern blot showing HR. Total genomic DNA was isolated from 14 GFP+RFP- fish and probed with a g#?-specific probe.

Figure 4 provides a schematic of a split marker strategy. Two individual fragments can be co-injected with the MO/yRad51/52 cocktail. GFP positive lines will only be produced if HR has occurred.

DETAILED DESCRIPTION OF THE INVENTION

The instant invention provides methods to enhance homologous recombination by inhibiting the NHEJ DNA repair pathway while simultaneously adding factors designed to promote HR. NHEJ enzymes repair double-strand breaks in DNA by using single-stranded overhangs at the break point. Only disrupting NHEJ in organisms such as Neurospora crassa increase HR by 10-fold. Disrupting a single NHEJ enzyme in human cells increases HR by up to 20-fold while disrupting two enzymes from within the same pathway increased HR 164-fold, but the levels are still low and require time consuming screening and many starting samples. Herein, an alternative method is provided in zebrafish (D nio reri ) whereby multiple NHEJ enzymes are inhibited while simultaneously adding factors that promote HR.

Unexpectedly, this combination strategy has led to reproducible targeted integration at around 30-50% with an overall targeted integration rates of >90%. This increased rate of integration was found in a system where random integration occurs at a minuscule rate of one in 10² to 10⁴ and homologous integration occurs at an even lower rate of 10⁵ to 10⁷ (Sargent et al. (1998) Curr. Res. Mol. Ther., 1, 584-692).

While the instant invention is exemplified herein in zebrafish, it can be used in other organisms such as higher eukaryotes including mammalian (e.g., human) cells (including stem cells), plants (e.g., crops), and single cell organisms (e.g., those used in the biofuels industry). The instant invention allows for the construction of targeted knockouts by gene replacement, thereby facilitating the creation of loss-of-function deletions, site-specific mutants, and conditional expression constructs. The instant invention also allows for the generation of knock-ins. In a particular embodiment, the instant invention can be applied to human stem cells to promote gene therapy. As explained herein, the instant invention also encompasses vectors which allow for epitope tagging of endogenous genes. The epitope tagging will be useful in detecting/tracking the endogenous protein in a variety of assays including

immunodetection, immunoprecipitation (including ChIP), immunoaffinity

purification, and fluorescence microscopy (including FRET), among other applications, in a number of model systems. The instant invention provides many advantages including, without limitation: ease and specificity of mutant generation; multiple model organisms; alternative, novel vector system; compliance with multiple tagging methods; and applications for human gene therapy. Additionally, this methodology has led to the development of a new ligation independent DNA cloning method that allows easy insertion of multiple DNA fragments into any DNA shuttle vector without the use of DNA ligase or restriction enzymes (Joska et al. (2014) J. Microb. Methods 100:46-51). Moreover, this cloning technology allows assembly of multiple fragments at once and is 100% efficient, supporting the rapid assembly of gene targeting cassettes.

Zebrafish are widely-used as an animal model for biomedical research. It is a vertebrate that shares many genetic and physiological similarities to mammals, yet it is significantly more cost effective to maintain in a laboratory compared to rodent models (Vascotto et al. (1997) Biochem. Cell Biol, 75:479-485; Parng et al. (2002) Assay Drug Dev. Technol., 1 :41-48). The biological and mechanistic pathways in zebrafish are highly conserved with humans making it an ideal system to model complex diseases. The zebrafish embryo is small and translucent making it well- suited for high-throughput analysis, drug discovery, and toxicological screening (Parng et al. (2002) Assay Drug Dev. Technol., 1 :41-48). Moreover, organ development and disease progression can be visualized in real-time. These attributes, and the ability to block proteins with morpholino oligonucleotides (MO), have propelled it to become the premier vertebrate system to study development. Zebrafish also has a sequenced and annotated genome including a transcriptional profile of developing embryos and the associated noncoding RNAs (Pauli et al. (2012) Genome Res., 22:577-591), libraries of full-length cDNAs, and an assortment of genetic techniques. These techniques allow the construction of transgenic lines and targeted mutations creating an attractive model. Despite these advantages, the zebrafish model lacks a simple and cost- effective method for targeted genetics that would further enhance its use as a model organism. A prime example is the inability to make targeted gene replacements by HR.

There is a method for HR in murine models that has allowed the construction of knockout mice for studying the functional requirements of individual genes. This has been the basis for the development of numerous mouse models of human diseases. HR in mammals is low, but the murine models have a method for selection and the ability to transform large numbers of blastocyst-derived stem cells via electroporation. However, mice are expensive to maintain, relatively slow to develop, and from a practical standpoint, the tractability is limited due to the high cost and expertise needed to generate a mouse line. Moreover, mice don't lend themselves to high throughput chemical screens, making it a resource-intensive system. Despite the high cost, mouse researchers are able to generate whole animal knockouts, generate conditional alleles via knockins, and are able to epitope tag endogenous genes (Hardy et al. (2010) Biol. Cell., 102:561-580).

Even though zebrafish lack efficient HR, there are a variety of methods to generate loss-of-function (LOF) zebrafish lines, but all of these current methods come with limitations. Currently, targeted knockout/knockdown technology in zebrafish is reliant on a handful of techniques: MO knockdown, TILLING (Targeting Induced

Local Lesions IN Genomes), zinc-finger nuclease (ZFN), RNAi knockdown, chemical mutagenesis, and transposable elements, all of which have some significant drawbacks (Bill (2009) Zebrafish 6:69-77; Eisen et al. (2008) Development 135: 1735- 1743; Moens et al. (2008) Brief Funct. Genomic Proteomic 7:454-459; Doyon et al. (2008) Nat. Biotechnol., 26:702-708; Zhao et al. (2001) Dev. Biol., 229:215-223; Bedell et al. (2011) Brief Funct. Genomics 10:181-188; Meng et al. (2008) Nat.

Biotechnol., 26:695-701 ; McCammon et al. (2010) Methods Mol. Biol., 649:281-298; Urnov et al. (2010) Nat. Rev. Genet., 11 :636-646; Kim et al. (1996) Proc. Natl. Acad. Sci., 93:1156-1160; Ekker, S.C. (2008) Zebrafish 5: 121-123; Wargelius et al. (1999) Biochem. Biophys. Res. Commun., 263: 156-161 ; De Rienzo et al. (2012) Zebrafish; Dong et al. (2009) PLoS One 4:e6125; Zhaoet al. (2008) FEBS J., 275:2177- 2184; Amsterdam et al. (2006) Trends Genet., 22:473-478; Jao et al. (2008) Brief Funct. Genomic Proteomic 7:427-443; Ni et al. (2008) Brief Funct. Genomic

Proteomic 7:444-453; Kawakami, K. (2007) Genome Biol., 8 Suppl 1 :S7; Koga, A. (2004) Adv. Biophys., 38: 161-180; Asakawa et al. (2009) Methods 49:275-281;

Kawakami et al. (2004) Dev. Cell 7: 133-144; Balciunas et al. (2006) PLoS Genet., 2:el69).

A solution to the genetic limitations in zebrafish is the development of efficient HR. HR allows targeted gene replacement, replacement of endogenous genes with specific mutated genes, epitope tagging, and construction of human disease models. Like mice, HR in zebrafish is possible, but it is remarkably inefficient. HR in zebrafish has been demonstrated in ES cells and in injected fertilized eggs (Wu et al. (2006) Mar. Biotechnol., 8:304-311; Fan et al. (2006) Transgenic Res., 15:21-30). In the case with the injected eggs, HR occurs, at best, once out of 2.4 x 10⁵ injections and germline transmission is only 1.7 x 10^'6 (Wu et al. (2006) Mar. Biotechnol., 8:304-31 1). The infrequency of HR in zebrafish using current methods creates a barrier for reverse genetic approach and genetic

manipulation. Mouse researchers overcome this problem with selection (neomycin gene) and are able to transform large numbers of stem cells by electroporation. If the efficiency of HR in zebrafish is improved, then making targeted knockout strains would emerge as workable method. Herein, it is shown that HR efficiency in zebrafish is increased by inhibiting at least one subunit/component of NHEJ (e.g., by using a mixture of MO) and adding the yeast proteins yRad51 and yRad52. The increased HR efficiency will allow for the characterization of pathways involved in development, organogenesis, regeneration, aging, braod genetic disease etc. and, overall, increase the power of zebrafish as a research model organism.

As demonstrated herein, inhibiting NHEJ components such as Ku80, LiglV and XRCC4 and adding yRad51 and yRad52 results in targeted, non-random integration. The injection cocktail described below contains a mixture of morpholino oligonucleotides (MO) that targets Ku80, LiglV and XRCC4 in the NHEJ pathway with each at 1 ΟμΜ. This concentration was determined by performing dose response curves varying the concentration between ΙμΜ and ΙΟΟμΜ compared to a scrambled- MO control (sMOC). At high concentrations (ΙΟΟμΜ) there is a significant effect on viability, and minor phenotypes (e.g., whirling with LiglV MO) were observed at 25μΜ. At 10μΜ, no visual defects were apparent relative to the sMOC. It is desirable to use a concentration of MO that gives maximum HR while minimizing the potential adverse effects of prolonged inhibition ofthe DNA repair machinery.

Further, it is desirable for the DNA repair machinery to still function, but significantly favor HR over NHEJ when a DNA fragment is introduced. Complete abrogation of NHEJ may not be needed for this to occur. Therefore, the optimal concentrations of MO for Ku80, LiglV and XRCC4 that remove the NHEJ enzymes may be

accomplished by performing a dose response curve varying the concentration of each MO individually, then measuring the amount of corresponding transcript by qRT- PCR. Once the concentration of each individual MO is determined, the dose response may be repeated using all three MOs simultaneously to ensure there are no

compensatory effects. This will allow for the determination of the minimal concentration of MO that cause loss of the NHEJ. This concentration can then be used as the maximum concentration in an additional set of dose response curves designed to measure integration rates. These experiments would be designed to determine the minimal effective concentrations of each MO that allows a high rate of HR.

While MOs are described, other inhibitory nucleic acid molecules such as siRNA (or shRNA) or standard antisense oligonucleotides may be used instead.

Indeed, the stability of MO may block NHEJ longer than necessary. For example, NHEJ may only need to be inhibited for a short period of time (one to two hours instead of days) and the full 72 hours post-fertilization (hpf) inhibition that is typical for high concentrations of MO may not be needed. Accordingly, inhibitory nucleic acid molecules such as antisense oligonucleotides that are rapidly metabolized may be used.

The zebrafish genome has a significant amount of duplications and adding recombinant yRad52 and yRad51 protein in the injection cocktail commits the cells to HR rather than NHEJ. Excess yRad51 and yRad52 may produce unwanted side- effects of enhancing genomic translocations between homologous sequences on different chromosomes (hybrid chromosomes). These translocations may cause the fish to become non- viable for making homozygous knockouts, or prevent germline transmission. Therefore, the levels of yRad51 and yRad52 proteins in the injection cocktail may be titrated to determine a minimal effective concentration. Additionally, instead of injecting protein, nucleic acid molecules encoding yRad51 and yRad52

(e.g., in vitro transcribed mRNA of rad51 and rad52 generated using the mMESSAGE mMACHINE® Ultra Kit (Applied Biosystems) or other available kits) may be used. This kit produces an mRNA with a 5' 7-methylguanosine cap and a poly A tail and is commonly used for mRNA rescue experiments in developing zebrafish. The use of nucleic acids may be desirable to avoid potential solubility issues with the proteins. Regardless, the addition of yRad51 and yRad52 leads to yRad51/52 being bound DNA and commits the DNA repair machinery to HR, whereas Ku70/80 bound DNA block HR and ensures NHEJ (Sonoda et al. (2006) DNA Repair (Amst) 5:1021-1029).

With regard to the targeting construct, the construct may comprise a multiple cloning site (MCS), a selection marker (e.g., URA3, an antibiotic resistance marker such as the Kan^R/G418 gene), and/or bacterial and yeast origin of replication (e.g., fl ori and 2u ori, respectively) within the cassette sequence for cassette assembly and to create a convenient method to determine the integration site. In a particular embodiment, the construct may comprise a 5' flank (which specifically hybridizes a desired target sequence in the genome), a nucleic acid molecule encoding a first detectable protein (e.g., fluorescent or luminescent protein) operably linked to a promoter, a multiple cloning site (MCS) (or a nucleic acid of interest to be inserted), an optional selection marker (e.g., an antibiotic resistance marker such as the

Kan^R/G418 gene) for bacterial and eukaryotic selection, an optional origin of replication (e.g., pUC19 ori), a 3' flank (which specifically hybridizes a desired target sequence), and a nucleic acid molecule encoding a second detectable protein (e.g., fluorescent or luminescent protein; distinguishable/different than the first detectable protein; e.g., a kill gene (e.g., a toxin)) operably linked to a promoter. The integration site can be recovered by isolating the plasmid from genomic DNA. Genomic DNA can first be digested with an enzyme that cuts within the MCS and outside the 3' or 5' flank, and the vector can then be ligated back on itself in a proximity-based reaction before being transformed into E. coli. The integration site can then be determined by sequencing of the integration site. Genomic isolation, digestion, ligation,

transformation and sequencing the integration site can be performed in 96-well plates. Such a method can be a superior means to confirm knockouts compared to Southern blot hybridization or PCR, which can produce both false positive and false negatives. This method requires only a minimal amount of genomic DNA and commercial of DNA sequencing means are available.

Adding additional MO that targets other subunits of NHEJ can also be performed. Additionally, a zebrafish line that expresses high levels of Parp can be used. Parp covalently attaches Poly[ADP-ribosylation] onto proteins and is one of the earliest cellular responses to DNA damage. Parp appears to commit the cell to HR and loss of Parp increases the rate of NHEJ (Ame et al. (2004) Bioessays 26:882-893; Hochegger et al. (2006) EMBO J., 25: 1305-1314).

In addition to the above, a split-marker approach may be used (see, e.g., Figure 4). The GFP/RFP prescreening method allows for the screening of positive HR events versus ectopic insertions or random insertions into the genome by NHEJ. The GFP/RFP prescreening also allows for the screening of germline vs mosaic integrations whereby HR occurs at the single cell stage vs the multi-cell stage thereby transmitting gene conversion to every cell in the organism vs a subpopulation of cells rendering the mosaic phenotype (a portion of cells that are GFP+).

As an alternative, the split-marker strategy may be used. Under these conditions, two fragments that, by themselves, will not produce a GFP+ signal are injected. The integrated DNA sequence will only produce a GFP+ signal if HR has occurred. The two separate pieces of targeting DNA will be injected simultaneously with the MO/yRad51/52 cocktail. GFP+ signal in zebrafish will indicate a HR event.

Another alternative method is the kill gene strategy that may be used to replace RFP. In this case, the kill gene may be the diphtheria toxin open reading frame or any other toxin or lethal protein designed to universally kill if and when expressed. As shown in Figure 1, the transgene includes GFP between the 5' and 3' flanks and the kill gene outside of the flanks, or to either side of the flanks.

Recombinant fish survive if and only if the kill gene is homologously recombined of the targeting transgene by homologous recombination. The embodiment includes any counter selection in place of the diphtheria toxin rendering the organism non- viable. In the event of NHEJ, the toxin will be incorporated into the genome in a random location and expressed, resulting in cell death. Only targeted homologous recombinant fish will survive.

Notably, when embryos are used, maternally inherited NHEJ enzymes may be present. If the proteins are present at high concentrations in the egg, then MO knockdown may be ineffective. One option to deal with this circumstance would be to treat the zebrafish embryos with low levels of reagents that induce double-strand breaks (e.g., bleomycin) either before or immediately after the co-injection process to sequester existing NHEJ enzymes. A risk with this modification is treating eggs with DNA-damaging agents may introduce mutations. However, if this improves targeted HR, the resulting transgenic fish can be backcrossed to remove any potential deleterious mutations. Alternatively, restriction enzymes can be co-injected that would function like ZFN. This would require that the construct pair effectively with the cut site and the minimal amount of restriction enzyme would need to be determined empirically. In another embodiment, an shR A zebrafish line may be generated that inhibits NHEJ enzymes in the female germline to minimize the amount of these enzymes present in the fertilized eggs.

In a particular embodiment, the targeting construct may be co-injected with antibodies immunologically specific for zebrafish NHEJ enzymes. Co-injection of the construct with antibodies (with or without the MO) would inhibit NHEJ activity and improve HR rates. In order to facilitate entry into the nucleus, small antibody fragments may be used. For example, Ku80 and/or Ku70 anti-Fab fragments may be generated by treating antibodies with papain that will reduce the size to approximately 50 kDa. Monoclonal single domain antibody (sdAb) may be used. The sdAb are small antibody-like molecules ranging in size from 12-15 kDa, allowing for passive diffusion through the nucleus as proteins smaller than 30 kDa are capable of passive diffusion through the NPC. In a particular embodiment, nuclear localization signals may be added to the antibodies to facilitate their movement into the nucleus.

As an alternative to adding yRad51 and yRad52 protein or nucleic acids, bacteria-derived RecA protein or RecA nucleic acids may be used. The RecA could be added with the MO and split markers and antibodies.

The instant invention also encompasses vectors for use in the described HR methods. The ability to detect, quantify and characterize proteins is vital to molecular biology research. Generating fusion proteins for use in fluorescent and luminescent experiments, in addition to adding peptide epitopes to analyze proteins either by immunoblot analysis, immunoprecipitation (including ChIP), or

immunohistochemistry are all extremely valuable research tools. These techniques are currently available in zebrafish, but are often reliant on ectopic expression that coexist alongside the endogenous WT copies making it a limited method to confirm functionality. Accordingly, the instant invention encompasses nucleic acid molecules or vectors for the convenient insertion of DNA sequencing encoding peptide epitopes on the 3' end of genes of interest. For example, fluorescent or luminescent proteins (e.g., GFP, RFP, YFP, or luciferase) and/or peptide epitopes (e.g., FLAG, MYC, HA, or V5) can be added to the 3' of a gene of interest. The vectors may have a combination of epitopes for tandem-affinity purification purposes (e.g., TAP-Tag). These vectors can be created in yeast and construction can follow the methodologies similar to the vector assembly outlined in Figure 1. The 5' and 3' flanks of the gene of interest will be amplified and inserted into the vector.

In a particular embodiment, the vector is a DamID construct. The DamID method is an alternative to ChIP and allows one to identify the locations where a protein associates with DNA based on adenine methylation (van Steensel et al. (2001) Nat. Genet., 27:304-308). This powerful technique offers many advantages over the ChIP reaction. It can be used in instances where the antibody has yet to be made, or if the antibody is unsuitable for ChIP. It will also work if there is only a transient interaction with chromatin that is normally difficult to detect in formaldehyde crosslinking, and it also reduces the labor-intensive task of optimizing conditions for the ChIP reaction. In a typical DamID experiment, the restriction endonuclease Dpnl is used to detect if a specific enzyme localizes to a specific region because this enzyme will only digest DNA if the adenine is methylated. Dpnl is a 4 bp cutter and cuts on average (assuming random sequence) once every 256 bp so it appears quite regularly throughout the genome.

All the epitopes discussed can be designed to work with a given set of flanking DNA allowing them to be interchangeable as desired. Similar systems have been developed in Saccharomyces and Neurospora and have greatly facilitated biochemical and molecular manipulations in both these organisms (Longtine et al. (1998) Yeast 14:953-961 ; Honda et al. (2009) Genetics 182: 11-23).

Notably, not every gene knockout will allow germline transmission and this creates barriers for research. Moreover, the zebrafish genome contains duplications, so deleting a single gene may have little effect. Instead of post-transcriptional gene silencing (PTGS) using shRNA, RNAi-mediated transcriptional gene silencing (TGS) could be induced by inserting an inducible promoter at the 3' region of a gene so that it drives expression of an antisense transcript. RNAi-mediated TGS could work in trans and produce localized heterochromatic regions at multiple different locations on the genome corresponding to the gene of interest. This would silence duplicated or essential genes and it would allow silencing only at desired times. The method may be performed by choosing multicopy and essential genes, inducing expression of the antisense (w.g., with the tet-on/tet-off system) and then examining endogenous transcripts levels by qRT-PCR. ChIP experiments may also be performed using H3K9me3, H3K27me3 and HP1 (or ChlP-seq) to ensure that heterochromatin has formed in trans and silenced all copies of the gene. Genome-wide association studies (GWAS) have unveiled numerous advances in causative genetic mutations of disease and the advent of high-throughput sequencing will likely produce significantly more information. Therefore, a convenient method to rapidly produce disease models is needed. The instant HR technology can be used to insert mutations (e.g., mutations identified via GWAS) to generate zebrafish disease models for research. These genetic models can then be used in high throughput screens of chemical compounds that may have therapeutic potential or they can be used to research the etiology of diseases.

The worldwide obesity epidemic is one of the largest global-health problems. Obesity is associated with numerous diseases including type 2 diabetes,

cardiovascular disease and cancer. There is a zebrafish model of type 1 diabetes that involves intraperitoneal injection of streptozocin resulting in a sustained

hyperglycemic state, but other good genetic models of obesity appear to be lacking. Recent GWAS studies have identified several SNPs in the fat mass and obesity associated (Fto) gene (Frayling et al. (2007) Science 316:889-894; Dina et al. (2007) Nat. Genet. 39:724-726; Fall et al. (2012) Mol. Cell Endocrinol.). Fto encodes a protein of unknown function with homology to 2-oxoglutarate oxygenase

superfamily. Fto-deficient mice have a significant decrease in body adipose tissue (Fischer et al. (2009) Nature 458:894-898) whereas mice carrying additional copies of Fto have higher levels of food intake causing obesity (Church et al. (2010) Nat.

Genet., 42:1086-1092). These reports have a positive correlation with humans carrying the rs9939609 SNP that also display high Fto transcript levels and increase body mass index (BMI).

Another gene implicated in obesity is stearoyl-CoA desaturase (SCD). SCD1 converts stearoyl-CoA to oleate or palmitoleate by a desaturation reaction at 9 position of CI 8 or CI 6 saturated hydrocarbons respectively. SCD 1 -deficient mice are protected from obesity and insulin resistance when fed a high-fat diet (Ntambi et al. (2002) Proc. Natl. Acad. Sci., 99: 1 1482-11486). Mice expressing additional copies of Scdl have not been tested. There are four Scd isoforms in mice and one Fto that all appear to be circadian regulated, linking circadian gene expression and obesity (Panda et al. (2002) Cell 109:307-320).

These observations create a unique scenario where the epistatic and holistic relationship of two unrelated genes can be combined with diet and environment (chemical screen/xenobiotics) to better understand metabolic syndrome. Using the HR method, zebrafish lines can be generated where Fto and Scdl are both

overexpressed with the constitutive promoter and deleted with the knockout cassette. The collection of these fish can be set up in reciprocal crosses to determine if loss of Fto creates a protective phenotype in Sdcl overexpressing cells, or vice versa. These fish can then be further characterized by lipidomic profiling using gas

chromatography mass spectrometry (GS-MS) from liver tissue and by examine adipose tissue using magnetic resonance imaging (MRI).

Herein the functionality of zebrafish (Danio rerio) is improved as a research model by providing a new genetic research tools to enhance targeted homologous recombination (HR). The methods proposed here aid in characterization of genes, and detection of proteins allowing phenotypes and pathways to be elucidated. The tools will advance research in development, aging, organ formation, neural process, and human diseases by providing a method for convenient and efficient reverse genetics. Zebrafish is an exceptional model vertebrate system. Vectors that can be used with HR to insert DNA sequences encoding C-terminal epitopes to detect proteins are also encompassed by the instant invention. Further, zebrafish models of human diseases constructed by the HR methods described herein are also encompassed.

In accordance with the instant invention, methods of increasing homologous recombination in a cell (e.g., a eukaryotic cell) are provided. In a particular embodiment, the method comprises delivering Rad51 and/or Rad52. The method may further comprise the delivery of at least one other protein (or nucleic acid encoding the protein) that supports homologous recombination. Examples of proteins that support homologous recombination include, without limitation, meiotic recombination- 11 (MRE11), PARP, Rad50, Rad54, nijmegen breakage syndrome-1 (NBS- 1 ), RPA, BRCA1 and BRCA2 RNA. The method may further comprise delivering at least one inhibitor of a component of the NHEJ pathway. Examples of components of NHEJ include, without limitation, LiglV, XRCC4, XRCC7, Ku70, Ku80, DNA-dependent protein kinase (DNA-PKC), and NBS 1. In a particular embodiment, the method comprises delivering to the cell: 1) Rad51 ; 2) Rad52; 3) an inhibitor of Ku80 (and/or an inhibitor of Ku70); 4) an inhibitor of LiglV; and 5) an inhibitor of XRCC4. In a particular embodiment, the method comprises delivering Rad51 and Rad52 and one, two, three, or four inhibitors selected from the group consisting of an inhibitor of Ku80; an inhibitor of Ku70; an inhibitor of LiglV; and an inhibitor of XRCC4. The method may further comprise the delivery of at least one nucleic acid molecule of interest.

Rad51 and Rad52 can be from any species. In a particular embodiment, Rad51 and Rad52 are from yeast (e.g., GenelD: 856831 for yeast rad51 ; GenelD: 854976 for yeast rad52). In yeast (Saccharomyces cerivisiae), HR is the sole method of DNA repair, whereby NHEJ events are absent or rare. As mentioned, in higher vertebrates NHEJ is by far the favored method of double strand break (DSB) repair. Corresponding vertebrate enzymes of Rad51 and Rad52 orchestrate NHEJ and HR. Inasmuch as yeast Rad51 and Rad52 orchestrate only HR, they are superior to vertebrate enzymes which may not facilitate HR over NHEJ. GenBank Accession No. CAA45563.1 provides an amino acid sequence of yeast Rad51 and GenBank Accession No. CAA86623.1 provides an amino acid sequence of yeast Rad52. In a particular embodiment, Rad51 and Rad52 are human. Rad51 and Rad52 may be delivered to the cell as proteins or as nucleic acid molecules encoding the proteins. In a particular embodiment, Rad51 and/or Rad52 may be replaced in the instant methods with RecA or other bacterial derived recombinases. In a particular embodiment, Rad51 and/or Rad52 are delivered as a fusion protein with a cell penetrating peptide (e.g., the arginine-rich domain of the TAT protein of HIV (tatl 1) (48-60); penetratin (antennapedia (43-58)); pVEC (cadherin(615-632)); transportan; MPG; Pep-1;

polyarginines; MAP; R6W3; etc.) among other such molecules.

The inhibitors of the instant methods may be any compound that inhibits the activity of the protein such as a small molecule, inhibitory nucleic acid molecule (e.g., nucleic acid molecules which specifically hybridize (e.g., are complementary) with a target nucleic acid thereby inhibiting its expression; inhibitory nucleic acid molecules include antisense, siRNA, shRNA, etc.), antagonist, ligand, antibody or fragment thereof, or the like. In a particular embodiment, the inhibitor is an antisense molecule. In a particular embodiment, the inhibitor is a morpholino antisense oligonucleotide. In a particular embodiment, inhibitors of Ku70 may be included or replace the Ku80 inhibitors. GenBank Accession No. NM 001017360 provides a nucleotide and an amino acid sequence of zebrafish Ku80 (XRCC5); GenBank Accession No.

NM_199904.1 provides a nucleotide and an amino acid sequence of zebrafish Ku70 (XRCC6); GenBank Accession No. NM 001103123 provides a nucleotide and an amino acid sequence of zebrafish ligase IV (LiglV); and GenBank Accession No. NM_200786 provides a nucleotide and an amino acid sequence of zebrafish XRCC4. While the above are described as being from zebrafish, the inhibitors may directed against any species, particularly the species of the cell being treated. Any number and combination of NHEJ inhibitor may be used in the instant invention.

The instant invention also encompasses compositions comprising one or more of the elements described above in the methods. For example, the composition may comprise Rad51 and/Rad52 and, optionally, at least one inhibitor of a

component/subunit of the NHEJ pathway as identified above. The composition may further comprise at least one other protein (or nucleic acid encoding the protein) that supports homologous recombination as described above. In a particular embodiment, the composition comprises 1) Rad51; 2) Rad52; 3) an inhibitor of Ku80 (and/or an inhibitor of Ku70); 4) an inhibitor of LiglV; and 5) an inhibitor of XRCC4. In a particular embodiment, the composition comprises Rad51 and Rad52 and one, two, three, or four inhibitors selected from the group consisting of an inhibitor of Ku80; an inhibitor of Ku70; an inhibitor of LiglV; and an inhibitor of XRCC4. The

compositions may further comprise at least one carrier.

The instant invention also encompasses kits. The kits may comprise one or more of the elements described above in the methods and compositions. For example, the kit may comprise Rad51 and/Rad52 and, optionally, at least one inhibitor of a component/subunit of the NHEJ pathway as described above. The composition may further comprise at least one other protein (or nucleic acid encoding the protein) that supports homologous recombination as described above. In a particular embodiment, the kit comprises 1) Rad51; 2) Rad52; 3) an inhibitor of Ku80 (and/or an inhibitor of Ku70); 4) an inhibitor of LiglV; and 5) an inhibitor of XRCC4. In a particular embodiment, the kit comprises Rad51 and Rad52 and one, two, three, or four inhibitors selected from the group consisting of an inhibitor of Ku80; an inhibitor of Ku70; an inhibitor of LiglV; and an inhibitor of XRCC4. The components of the kits may be contained separately or combined together. For example, the components of the kits may be contained individually in compositions further comprising at least one carrier or may be combined into one or more compositions further comprising at least one carrier. In a particular embodiment, the kit comprises a composition comprising the proteins and a composition comprising the nucleic acid molecules. The kits may further comprise cells (e.g., eukaryotic cells). Definitions

The following definitions are provided to facilitate an understanding of the present invention:

The singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise.

As used herein, the term "small molecule" refers to a substance or compound that has a relatively low molecular weight (e.g., less than 4,000, less than 2,000, particularly less than 1 kDa or 800 Da). Typically, small molecules are organic, but are not proteins, polypeptides, or nucleic acids, though they may be amino acids or dipeptides.

A "carrier" refers to, for example, a diluent, adjuvant, preservative (e.g., Thimersol, benzyl alcohol), anti-oxidant (e.g., ascorbic acid, sodium metabisulfite), solubilizer (e.g., Tween 80, Polysorbate 80), emulsifier, buffer (e.g., Tris HCl, acetate, phosphate), bulking substance (e.g., lactose, mannitol), excipient, auxilliary agent or vehicle with which an active agent of the present invention is administered. Carriers can be sterile liquids, such as water and oils, including those of petroleum, animal, vegetable or synthetic origin, such as peanut oil, soybean oil, mineral oil, sesame oil and the like. Water or aqueous saline solutions and aqueous dextrose and glycerol solutions are preferably employed as carriers, particularly for injectable solutions. The compositions can be incorporated into particulate preparations of polymeric compounds such as polylactic acid, polyglycolic acid, etc., or into liposomes or micelles. Such compositions may influence the physical state, stability, rate of in vivo release, and rate of in vivo clearance of components of a pharmaceutical composition of the present invention. The pharmaceutical composition of the present invention can be prepared, for example, in liquid form, or can be in dried powder form (e.g., lyophilized). Suitable pharmaceutical carriers are described in "Remington's

Pharmaceutical Sciences" by E.W. Martin (Mack Publishing Co., Easton, PA);

Gennaro, A. R., Remington: The Science and Practice of Pharmacy, (Lippincott, Williams and Wilkins); Liberman, et al., Eds., Pharmaceutical Dosage Forms, Marcel Decker, New York, N.Y.; and Kibbe, et al., Eds., Handbook of Pharmaceutical Excipients, American Pharmaceutical Association, Washington.

The phrase "small, interfering RNA (siRNA)" refers to a short (typically less than 30 nucleotides long, more typically between about 21 to about 25 nucleotides in length) double stranded RNA molecule. In a particular embodiment, the siRNA is about 21 nucleotides in length. Typically, the siRNA modulates the expression of a gene to which the siRNA is targeted. The term "short hairpin RNA" or "shRNA" refers to an siRNA precursor that is a single RNA molecule folded into a hairpin structure comprising an siRNA and a single stranded loop portion of at least one, typically 1-10, nucleotide.

The term "antisense" refers to an oligonucleotide having a sequence that hybridizes to a target sequence in an RNA by Watson-Crick base pairing, to form an RNA:oligonucleotide heteroduplex with the target sequence, typically with an mRNA. The antisense oligonucleotide may have exact sequence complementarity to the target sequence or near complementarity. These antisense oligonucleotides may block or inhibit translation of the mRNA, and/or modify the processing of an mRNA to produce a splice variant of the mRNA. Antisense oligonucleotides are typically between about 5 to about 100 nucleotides in length, more typically, between about 7 and about 50 nucleotides in length, and even more typically between about 10 nucleotides and about 30 nucleotides in length.

As used herein, the term morpholino oligonucleotide refers to an antisense oligonucleotide comprising one or more (or only) morpholino subunits or morpholine bases (e.g., wherein a morpholine ring replaces the ribose or deoxyribose sugar moiety and a non-ionic phosphorodiamidate linkage replaces the anionic phosphates). Morpholino oligonucleotides are described in Bill et al. (Zebrafish (2009) 6:69-77).

An "antibody" or "antibody molecule" is any immunoglobulin, including antibodies and fragments thereof, that binds to a specific antigen. As used herein, antibody or antibody molecule contemplates intact immunoglobulin molecules, immunologically active portions of an immunoglobulin molecule, and fusions of immunologically active portions of an immunoglobulin molecule. Antibody fragments include, without limitation, immunoglobulin fragments including, without limitation: single domain (Dab; e.g., single variable light or heavy chain domain), Fab, Fab', F(ab')₂, and F(v); and fusions (e.g., via a linker) of these immunoglobulin fragments including, without limitation: scFv, scFv₂, scFv-Fc, minibody, diabody, triabody, and tetrabody.

As used herein, the term "immunologically specific" refers to

proteins/polypeptides, particularly antibodies, that bind to one or more epitopes of a protein or compound of interest, but which do not substantially recognize and bind other molecules in a sample containing a mixed population of antigenic biological molecules.

The term "vector" refers to a carrier nucleic acid molecule (e.g., DNA) into which a nucleic acid sequence can be inserted for introduction into a host cell where it will be replicated. An "expression vector" is a specialized vector that contains a gene or nucleic acid sequence with the necessary regulatory regions needed for expression in a host cell.

The term "operably linked" means that the regulatory sequences necessary for expression of a coding sequence are placed in the DNA molecule in the appropriate positions relative to the coding sequence so as to effect expression of the coding sequence. This same definition is sometimes applied to the arrangement of coding sequences and transcription control elements (e.g. promoters, enhancers, and termination elements) in an expression vector. This definition is also sometimes applied to the arrangement of nucleic acid sequences of a first and a second nucleic acid molecule wherein a hybrid nucleic acid molecule is generated.

The term "cell-penetrating peptide" as used herein refers to any peptide which translocates across a cell membrane, e.g., across the plasma and or nuclear membrane, particularly with an attached polypeptide.

As used herein, the terms "multiple cloning site" refer to a nucleotide sequence (typically artificially created) comprising at least one restriction site

(typically two or more) for the purpose of cloning nucleic acid fragments into another nucleic acid such as a vector.

The following example is provided to illustrate certain embodiments of the invention. It is not intended to limit the invention in any way.

EXAMPLE

The differences between Saccharomyces and vertebrates are numerous, but in the case with HR, the disparity resides in the method of DNA repair. Yeast and vertebrate cells repair double strand breaks (DSB) either by NHEJ or HR. Yeast almost universally employs HR to repair DSB (NHEJ is extremely rare in yeast) and it occurs by a sequence-specific strand exchange reaction requiring yRad51 and yRad52. The yRad51 and yRad52 proteins also appear to be more suited for HR than their vertebrate counterparts. Contrary to yeast, vertebrates prefer NHEJ over HR because there is significantly more DNA, and the DNA is more compacted, making it extremely difficult to find homologous sequences. NHEJ interferes with, and inhibits, HR. Once a cell becomes committed to NHEJ, the HR enzymes are actively excluded. Given a choice, HR is a far better method to repair DSB and results in fewer errors, but because DNA compaction hinders homology searches, vertebrates are reliant on, and biased toward, NHEJ over HR (Sonoda et al. (2006) DNA Repair (Amst) 5 : 1021 - 1029). When foreign DNA is introduced in vertebrates, random integration occurs at DSB via NHEJ and it occurs at 1000-fold higher frequency over HR (Sonoda et al. (2006) DNA Repair (Amst) 5 : 1021 - 1029; Vasquez et al. (2001 ) Proc. Natl. Acad. Sci., 98:8403-8410). Repair of DSB by NHEJ involves capturing both DNA ends by a complex containing Ku70/Ku80 and re-ligating the DNA back together by a separate complex containing XRCC4 and the DNA ligase, LiglV (Weterings et al. (2008) Cell Res., 18: 114-124).

In view of the foregoing, it was determined whether blocking NHEJ in zebrafish (e.g., using MO) while simultaneously adding factors that promote HR in Saccharomyces (e.g., yRad51 and yRad52), would increase the rates of targeted integration to reasonable levels. Moreover, it is likely that homology searches will be less difficult when relatively large homologous flanking DNA sequences are added. Support for this methodology comes from numerous organisms from Neurospora crassa to mammalian cells lines. Knocking out the NHEJ enzymes Ku70 or Ku80 in Neurospora increased HR dramatically. Wild type (WT) Neurospora normally has HR rates of roughly 10% using 2 kilobases (kb) of homologous flanking DNA.

However, when ku70 (mus-5\) or ku80 (mus-52) are disrupted, targeted integration increases to 100% with the same 2 kb of flanking DNA (Ninomiya et al. (2004) Proc. Natl. Acad. Sci., 101 : 12248-12253). Recombination rates are greater than 90% with only 500 bp of flanking DNA in ku70 or ku80 mutants compared to WT where 500 bp is insufficient for HR. This increase in HR efficiency, combined with robot-assisted construct assembly, facilitated the systematic deletion of nearly all 10,000 predicted open reading frames in Neurospora (Colot et al. (2006) Proc. Natl. Acad. Sci.,

103:10352-10357; Collopy et al. (2010) Methods Mol. Biol., 638:33-40). This also led to the creation of epitope tagging and tunable expression strategies at endogenous loci similar to methods in yeast (Honda et al. (2009) Genetics 182:11-23; Larrondo et al. (2009) Eukaryot. Cell 8:800-804). In addition to removing Ku70 and Ku80, loss of the DNA ligase LiglV, resulted in significant increases in HR in both Neurospora and Aspergillus (Ishibashi et al. (2006) Proc. Natl. Acad. Sci., 103: 14871-14876; Takahashi et al. (2011) J. Biosci. Bioeng., 112:529-534).

Inhibiting a single subunit of NHEJ has little effect on HR rates in mammalian culture cells, but combinatorial silencing of both Ku70 and XRCC4 resulted in a 33- fold increase of targeted integration (Bertolini et al.. (2009) Mol. Biotechnol., 41 : 106- 114). This indicates that inhibiting multiple subunits further increases HR rates. Therefore, inhibiting LiglV along with Ku70/Ku80 and XRCC4 can reduce NHEJ to an even greater extent. LiglV is a key component because it ligates the DNA ends back together. The importance of this enzyme in DSB repair is highlighted by the observation that LiglV-deficient mice are not viable (Frank et al. (2000) Mol. Cell 5:993-1002). Other efforts to improve HR in mammalian culture cells showed that adding purified, membrane permeable yRad52-HIV(tatl 1) fusion protein increased HR by 37-fold (Kalvala et al. (2010) Nucleic Acids Res., 38:el49). It is clear that inhibiting NHEJ, or adding yRad52, leads to increases in HR efficiency. Herein, HR is improved by inhibiting multiple NHEJ components (Ku80, LiglV and XRCC4), while adding yRad51 and its cognate partner of yRad52. An enhanced selection step that enables a subset of positive integrations to be efficiently separated out with little effort is also provided herein. Indeed, proficient method to (1) make targeting constructs, (2) select likely HR replacements, and (3) simplify genetic confirmation of proper integration are provided.

The targeting construct was selected on the basis of a variety of criteria. First, the targeting construct would preferably include an extremely efficient prescreening method to filter out double HR events (needed for gene replacement) from ectopic insertions. This would significantly reduce the number of fish needed for molecular confirmation. The vector system should also allow convenient insertion of PCR- generated 5' and 3' DNA targeting sequences in a one-step cloning reaction so other constructs can be easily made. The vector/targeting sequence backbone that integrates into the zebrafish genome should also contain components that aid in confirmation of site-specific integration from a minimal amount of DNA acquired through tail snips (PCR from genomic DNA has the potential to produce false positives and false negatives). Lastly, the entire system from start to finish, excluding the injection into the fertilized egg, should be portable to a 96- well plate robot to aid screening or the eventual construction of an entire zebrafish knockout collection.

Construction of DNA knockout cassettes almost universally requires a three- point cloning reaction (at a minimum) where two homologous DNA flanking sequences straddle a selectable marker. To simplify this process, a method for cloning that takes advantage of the high rates of HR in Saccharomyces cerevisiae was utilized to assemble the parent vector (pZFHRl). The vector and the methodology for its use are outlined in Figure 1. The process starts by PCR amplification of 3kb flanking sequences of the 5' and 3' regions from zebrafish genomic DNA. The primers used to generate the flanks contain 20 bp of homology to the amplicon and 29 bp of homology to sequences in the vector pZFHRl for assembly via yeast cloning. To insert the PCR-generated flanks, pZHFRl is digested with the homing

meganuclease I-Ceul and then mixed in equal molar concentration with 5' and 3' targeting sequences (Step 1). The mixture of four DNA fragments is transformed into yeast and the vector is assembled by HR. The assembled construct is then recovered in E. coli (Step 2). The targeting construct is isolated by gel purification after digesting with l-Scel, before being injected into zebrafish fertilized eggs (note the yeast ura3 gene and 2μ ori are also removed). The final DNA fragment for injection consists of the following linear order of sequences: 3kb 5' flank, LoxP site, cmv-egfp, LoxP multiple cloning site (MCS), kanamycin resistance gene (Kan^R), pUC origin of replication (ori), site, 3kb 3' flank, and scmv-rfp. The 5' and 3' flanks are chosen so that both the start of transcription and translation are removed.

A schematic representation of a double HR event is shown at the bottom of Figure 1 (replacing the first 2 exons). A double HR event will yield a strong GFP signal and no RFP signal (GFP+RFP-) where, as an ectopic insertion that entered into a DSB, will express both GFP and RFP (GFP+RFP+). The interior of the

replacement sequence contains MCS, Kan^R gene, and pUC ori, arrayed in such a way to allow convenient recovery of the insertion site from zebrafish genomic DNA. There are also LoxP sites contained within the construct for removal of the cmv-egfp by injection of CRE recombinase mRNA for downstream applications. Alternative selectable markers besides GFP may be used. The above design is very efficiency and this protocol is easily modified for high throughput construct assembly.

In order to facilitate HR, 25 pg of the targeting construct is mixed with an injection cocktail composed of MO designed to knockdown the NHEJ enzymes XRCC4, LiglV and Ku80 (XRCC5), while adding purified yRad51 and yRad52 to promote HR. Both Ku70 and Ku80 have been successfully targeted using MO and embryos injected with either the Ku70 or Ku80 MO appear to develop normally, but are more susceptible to ionizing radiation-induced cell death (Bladen et al. (2007) Radiat. Res., 168:149-157; Bladen et al. (2005) Nucleic Acids Res., 33:3002-3010). Knockdown of Ku80, LiglV, and XRCC4 are non-lethal and display no phenotypic differences from WT at 10μΜ concentration. Approximately 8 nL of the cocktail is microinjected into fertilized eggs at the one-cell stage. 10 μΜ for each MO (XRCC4: GTGTCATTCATCGAGACTCACCGAG (SEQ ID NO: 1); LIG4:

GACACTTTCCATAATTGCAGAAGAC (SQE ID NO: 2); KU80:

TAGTGTAACAGGAAGGATACAGTCT (SEQ ID NO: 3)) was chosen because this did not yield any visible developmental defects. Microinjection efficiency is determined by assessing if fluorescein (contained in the MO) is contained in every cell at 24 hpf. If only a subpopulation of cells include fluorescein, or it is only in the yolk sac, the embryos are separated and discarded. After hatching, zebrafish are once again screened under a fluorescence microscope. This corresponds to the vector prescreen step. GFP+RFP- fish are indicative of a double HR event and these fish are segregated from the GFP+ RFP+ fish and allowed to mature to adults. A

representative example of this prescreening event is shown in Figure 2. The

GFP+RFP- fish shown in Figure 2 corresponds to Fish #2 in the southern blot shown below. Between 60-90% GFP+ RFP- fish were routinely obtained, highlighting that double HR events are occurring a much higher rate than anticipated.

Once the GFP+RFP- zebrafish have matured, insertions can be confirmed molecularly. To accomplish this, genomic DNA can be isolated from tail snips, digested with a unique restriction endonuclease in the MCS that is not present in the 3' flank, but is located within a couple kb downstream of the flank. The DNA is then ligated in a proximity-based reaction and plasmids containing the integration site are transformation into E. coli. Integration sites can be determined by single sequencing reaction. DNA isolation from tail snips, ligation, transformation into E. coli, plasmid isolation, and sequencing can all be done with the assistance of a robot in 96-well plate format for future high throughput genomic efforts.

For the preliminary studies thus far, integration was examined by Southern blot hybridization. For the Southern blot, a ^-specific probe was used. Figure 3 shows that 14 out of 14 GFP+ RFP- zebrafish are gfp positive on the Southern blot. Interestingly, reproducible patterns of integrations are observed that are highly indicative of HR. Had integrations occurred ectopically, one would have obtained a random pattern that was different in each individual fish. Here, the ku70 (xrcc6) gene was targeted for deletion (e.g., to generate a ku70 deletion zebrafish line that could serve as the background line for future injections). The ku70 locus is duplicated on chromosome 12. Therefore, the differences in the Southern hybridization pattern may be the result of targeting to either locus individually (fish 1, 6, and 8 or fish 9 and 14), or to both loci (fish 2, 3, 4, 5, 7, 10, 11, 12 and 13) simultaneously. Regardless, it is clear for all GFP+RFP- fish, a pattern that indicates reproducible, targeted integration is observed. Moreover, when it is considered that up to 90% of the injected fish are GFP+ RFP-, and nearly all of these appear to have targeted integrations, a convenient and efficient method for HR in zebrafish has clearly been established.

Claims

What is claimed is:

1. A method for increasing homologous recombination within a cell, said method comprising delivering to the cell Rad51 and Rad52 and at least one inhibitor of a component of the non-homologous end joining pathway.

2. The method of claim 1, wherein said inhibitor is selected from the group consisting of a LiglV inhibitor, a XRCC4 inhibitor, a Ku70 inhibitor, and a Ku80 inhibitor.

3. The method of claim 1, wherein said method comprises delivering to the cell

Rad51, Rad52, a LiglV inhibitor, a XRCC4 inhibitor, and a Ku70 inhibitor or a Ku80 inhibitor.

4. The method of claim 1 , wherein said cell is a zebrafish cell.

5. The method of claim 4, wherein said cell is a fertilized zebrafish egg.

6. The method of claim 1, comprising delivering to the cell Rad51 and Rad52 polypeptides.

7. The method of claim 1 , comprising delivering to the cell nucleic acid molecules encoding Rad51 and Rad52.

8. The method of claim 1, wherein said Rad51 and Rad52 are from yeast.

9. The method of claim 1, wherein said inhibitors are antisense oligonucleotides or siRNA.

10. The method of claim 9, wherein said inhibitors are morpholino oligonucleotides.

11. The method of claim 1 , further comprising delivering a nucleic acid molecule to the cell for homologous recombination.

12. The method of claim 11, wherein said nucleic acid molecule comprises a 5' flank, a nucleic acid encoding a first detectable protein operably linked to a promoter, a selectable marker, an origin or replication, a 3' flank, and a nucleic acid encoding a second detectable protein operably linked to a promoter, wherein said 5' flank and said 3' flank specifically hybridize with target sequences within the genome of the cell.

13. The method of claim 12, wherein nucleic acid molecule further comprises a multiple cloning site or a nucleic acid sequence of interest.

14. A composition comprising Rad51, Rad52, and at least one inhibitor of a component of the non-homologous end joining pathway.

15. The composition of claim 14, wherein said inhibitor is selected from the group consisting of a LiglV inhibitor, a XRCC4 inhibitor, a Ku70 inhibitor, and a Ku80 inhibitor.

16. The composition of claim 14, comprising Rad51, Rad52, a LiglV inhibitor, a XRCC4 inhibitor, and a Ku70 inhibitor or a Ku80 inhibitor.

17. The composition of claim 14, comprising Rad51 and Rad52 polypeptides.

18. The composition of claim 14, comprising nucleic acid molecules encoding Rad51 and Rad52.

19. The composition of claim 14, wherein said Rad51 and Rad52 are from yeast.

20. The composition of claim 14, wherein said inhibitors are antisense

oligonucleotides or siRNA.

21. The composition of claim 20, wherein said inhibitors are morpholino

oligonucleotides.

22. The composition of claim 14, further comprising a nucleic acid molecule for homologous recombination.

23. A kit comprising Rad51, Rad52, and at least one inhibitor of a component of the non-homologous end j oining pathway.