WO2025224107A1 - Method and compositions for detecting off-target editing - Google Patents

Abstract

The present disclosure provides compositions and methods for discovering potential cryptic recombination sites and profiling the recombination landscapes of integrases in programmable genomic integration technologies.

Description

METHOD AND COMPOSITIONS FOR DETECTING OFF-TARGET EDITING

1. CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application Nos. 63/637,351, filed on April 22, 2024, and 63/643,820, filed on May 7, 2024, each of which is hereby incorporated in its entirety by reference herein.

2. BACKGROUND

[0002] The large serine integrase (LSI) family constitutes a diverse group of site-specific recombinases that play pivotal roles in mediating DNA rearrangements. The precise and efficient DNA manipulation capabilities of the large serine integrase family have positioned it as a valuable tool in gene therapy and gene editing applications. Therapeutic genome editing with site-specific integrases requires a genotoxic safety evaluation that includes an unbiased and genome-wide characterization of recombination specificity because unintended recombination will result in large genomic rearrangements like off-target transgene insertion and chromosomal translocations.

[0003] Whole genome sequencing (WGS) alone is not sensitive enough to characterize potential off-target editing in a population of cells. There is therefore a need for a new tool to discover the potential recombination landscape for large serine integrases in any genome and in a biochemical or cell-based context.

3. SUMMARY

[0004] The present disclosure provides new technologies for discovering potential off-target editing and profiling the recombination landscapes of integrases. This technology can be used to support and validate new drug development involving phage integrases, such as for site-specific genetic engineering using Programmable Addition via Site-Specific Targeting Elements (PASTE) (see lonnidi et al. doi: 10.1101/2021.11.01.466786; U.S. Patent No’s. 11,572,556; 11,834,658; 11,827881; PCT Publication No. WO 2022/087235 each of which is herein incorporated by reference in its entirety) or other suitable gene editing or gene incorporation technology.

[0005] The Cas9 off-target discovery technologies Digenome-Seq (see Kim et al. (2015). Digenome-seq: genome-wide profiling of CRISPR-Cas9 off-target effects in human cells. Nat. Methods 12, 237-243) and CHANGE-Seq (see Lazzarotto et al. (2020). CHANGE-seq reveals genetic and epigenetic effects on CRISPR-Cas9 genome-wide activity. Nat. BiotechnoL 38, 1317- 1327) were adapted for the use in the discovery of loci that are potential DNA substrates for phage integrase-mediated recombination and evaluated for sensitivity in comparison to Cryptic-Seq. Digenome-Seq discovered about 40 cryptic Bxbl attB sites in the human genome and CHANGE-Seq discovered about 250 cryptic Bxbl attB sites in the human genome, while Cryptic-Seq discovered about 40,000 cryptic Bxbl attB sites in the human genome. The application of the presently disclosed Cryptic-Seq technology begins with the creation of DNA recombination substrates that contain phage integrase attachment sequences and Illumina next generation sequencing (NGS) compatible PCR amplification priming sequences. The Illumina NGS compatible sequencing adapters can be used for PCR enrichment and priming for the Illumina sequencing by synthesis chemistry, which improves NGS library preparation efficiency. The optional UMls improve data analysis by reducing PCR biases through the deduplication NGS reads to unique molecules at the start of the reaction.

[0006] Plasmid DNA is a convenient format for the creation, propagation and transfection of Cryptic-Seq recombination substrates. Linear synthetic DNA substrates also work well. The Cryptic-Seq recombination substrates enable the enrichment, identification, and relative quantification of recombined genomic sequences from any species in either a cell-based or biochemical recombination reaction with a phage integrase.

[0007] The Cryptic-Seq recombination substrates are utilized in a phage integrase-mediated recombination reaction with genomic DNA from any species, which results in the integration of Cryptic-Seq recombination substrates with target genomic DNA at potential cryptic recombination sites. The integration of unique DNA sequences from the recombination reaction with Cryptic-Seq substrates allows targeted PCR enrichment and NGS of recombined locations in the genome of interest. Through bioinformatic analysis, the NGS reads are filtered for attL (left) and attR (right) sequences that are associated with the completion of an integrase-mediated recombination reaction. The filtered reads are then aligned to the correct reference genome to identify the attL and attR genomic junctions and report the coordinates ranked by reads deduplicated from PCR amplification through the UMIs. The list of genomic coordinates completes the discovery phase of the potential recombination landscape for any integrase in any genome. Loci discovered by Cryptic-Seq can then be validated for recombination in the cell type of interest after genome editing with targeted methods.

[0008] In some embodiments, the integration enzyme is a large serine integrase. In certain embodiments, the large serine integrase is an integration enzyme disclose in PCT Publication No. W02023/070031, the disclosure of which is incorporated by reference in its entirety.

4. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] These and other features, aspects, and advantages of the present invention will become better understood with regard to the following description, and accompanying drawings, where:

[0010] FIGs. 1A-1E shows analysis of AttP variants. FIG. 1A shows a non-limiting schematic of AttP mutations tested for improving integration efficiency (SEQ ID NOS: 394 and 540-542, respectively, in order of appearance). FIG. IB shows integration efficiencies of wildtype and mutant AttP sites across a panel of AttB lengths. FIG. 1C shows a non-limiting schematic of multiplexed integration of different cargo sets at specific genomic loci. Three fluorescent cargos (GFP, mCherry, and YFP) are inserted orthogonally at three different loci (ACTB, LMNB1, NOLC1) for in-frame gene tagging. FIG. ID shows orthogonality of top 4 AttB/AttP dinucleotide pairs evaluated for GFP integration with PASTE at the ACTB locus. FIG. IE shows efficiency of multiplexed PASTE insertion of combinations of fluorophores at ACTB, LMNB1, and NOLC1 loci. Data are mean (n= 3) ± s.e.m.

[0011] FIG. 2 illustrates a schematic of single atgRNA and dual atgRNA approaches for beacon placement (“integration recognition site”).

[0012] FIG. 3 shows a schematic of Cryptic-Seq. Cryptic-Seq begins with the construction of DNA sequences that contains an attP or attB attachment site, unique molecular identifiers (UMIs) and PCR primer targeting sites (P7 and N7) that serve dual purposes, PCR amplification and Illumina sequencing by synthesis (SBS). The DNA sequences with attachment sequences are then combined with genomic DNA (gDNA) from any species in a biochemical recombination reaction with any large serine integrase (LSI). After the biochemical integrase recombination reaction is complete, the genomic DNA is simultaneously sequence-tagged and fragmented (tagmented) with the transposase Tn5 to impart a custom DNA sequence tag that includes a partial Illumina P5 sequencing adapter. After tagmentation, the library is amplified with anchor PCR to enrich recombined DNA sequences and impart the full-length sequencing adapters for sequencing on the Illumina HTS platform. The next generation sequencing (NGS) data is converted to FASTQs for each sample that contains the sequences corresponding quality scores associated with each sample. These data are then analyzed by a custom algorithm that filters reads containing attachment sequences and then aligns them to any genome to generate the genome attachment site sequence junctions and curate a list of genomic coordinates optionally ranked by UMI read counts for subsequent validation of recombination at the cryptic loci.

[0013] FIG. 4 shows a schematic of HIDE-Seq. When genomic DNA is exposed to an LSI protein and linear DNA attachment substrates, a double strand break is produced with the attachment sequences appended at the break points. In addition to the discovery of sites with potential for off-target integration, HIDE-Seq can also detect double-strand DNA breaks and gross chromosomal rearrangements between two endogenous loci that were recombined by an LSI.

[0014] FIG. 5 shows a schematic of Cryptic-seq analysis workflow with tbChaSin.

[0015] FIG. 6 shows a schematic of HIDE-seq analysis workflow with tbDigln.

[0016] FIG. 7 shows a schematic of hybridization capture followed by next generation sequencing (“Hybrid Capture”).

[0017] FIG. 8 shows a schematic of CHASER-Seq.

[0018] FIG. 9 shows a schematic of Att-Seq.

[0019] FIG. 10 shows the potential cryptic attB sites discovered with Cryptic-Seq. The attB sequence logo that represents the specificity profile of Bxbl integrase was derived from recombinogenic human sites that match some described priority base pairs in attB from previous functional assessments of attB in isolation.

[0020] FIG. 11 shows the comparison of Bxbl Integrase-dependent recurrent DSBs as discovered by HIDE-Seq with the recurrent DSBs of CRISPR/Cas9.

[0021] FIG. 12 shows the potential cryptic attB sites discovered with HIDE-Seq. The attB sequence logo that represents the specificity profile of Bxbl integrase was derived from recombinogenic human sites that match some described priority base pairs in attB from previous functional assessments of attB in isolation. Figure discloses SEQ ID NO: 410.

[0022] FIG. 13 shows the highest ranked cryptic attB sites discovered by HIDE-Seq and Cryptic-Seq and evaluated by hybrid capture.

[0023] FIG. 14 shows the heatmap of the cryptic sites validated by hybrid capture.

[0024] FIGs. 15A-15D Cryptic-seq off-target discovery for the LSI Bxbl across multiple central dinucleotides. Fig. 15A is a sequence logo that shows genome-wide search with the PWM generated by HIDE-seq with HOMER identified 4,598,283 potential off-target loci. Fig. 15B is a Data matrix created with UMI count bins from >=1 UMIs up to >= 5000 UMIs of the number of cryptic attB sites detected in the human genome with DNA donor substrates containing an attP containing either AA, TC, GG, CT, CA, and GT central dinucleotides. Fig. 15C is a Manhattan-style scatter plot showing Genomic distribution of cryptic attB sites plotted against the UMI signal discovered by Cryptic-seq. Fig. 15D is a DNA sequence motif from the 410,776 cyptic attB sites discovered by Cryptic-seq. Natural Bxbl attB sequence is displayed on the bottom. Figure discloses SEQ ID NO: 426.

[0025] FIGs. 16A-16B. Target vs predicted cross-validation plots from IntQuery and a linear regression model. Fig. 16A is a density scatter plot showing five-fold cross-validation between the target values and predictions of IntQuery revealed a Spearman correlation of p = 0.42. Fig. 16B is a density scatter plot showing a linear regression model trained with an identical protocol achieved a Spearman correlation of p = 0.34 across all predictions.

[0026] FIGs. 17A-17C. Large Serine Integrase (LSI) Mechanism of Action. Fig. 17A is a schematic diagram showing serine integrases catalyzing recombination between attachment (all) sites on linear or circular DNA substrates. Integrase dimers bind to specific all sequences in the phage attP and bacterial host (attB) DNA. Integrase bound to attP and integrase bound to attB associate to form a synaptic complex that connects the paired homologous sequences. The integrase subunits cleave all four DNA strands at the central dinucleotide, forming 5 '-phosphoserine linkages between integrase subunits and DNA halfsites and generating 3 '-dinucleotide overhangs. The P' and B'-linked subunits exchange places by rotating 180° about a horizontal axis relative to the P and B-linked subunits. Basepairing between the central dinucleotides promotes ligation of the DNA strands, resulting in formation of two new attachment sites, attachment left (attL) and attachment right (attR). Figure discloses SEQ ID NOS 394, 590, 410, and 591, respectively, in order of appearance. Fig. 17B Loci in the human genome with homology to LSI attB or attP sites are often called cryptic attachment sites because they are not obvious but have sufficient homology to catalyze recombination. Figure discloses SEQ ID NOS 592 and 410, respectively, in order of appearance. Fig. 17C There are 2 classes of off-target editing from an LSI; DNA mutagenesis in the form of indels from FDEs that may occur if recombination is disrupted in between strand cleavage and re-ligation. The second class is DNA structural variants, and these can come in two forms, the first is from off-target cargo insertion at a site in the human genome with homology to the attachment sequences. The second appears in the form of gross chromosomal rearrangements that may involve the attachment sequence introduced by genome editing or the interaction between two sites with homology to the attachment sequences.

[0027] FIGs. 18A-18E. High-throughput Integrase-mediated DNA Event Sequencing (HIDE-seq). Fig. 18A is a graphic of HIDE-seq experimental workflow. Genomic DNA is recombined with linear attB or attP substrates by an LSI and then subjected to WGS.

Fig. 18B are Venn diagrams of the loci from FDEs detected by HIDE-seq for Bxbl and Digenome-seq for Cas9. Each circle represents a biochemical replicate reaction. Zero recurrent and background levels of FDEs were detected in samples treated with Bxbl alone, Bxbl and attP, or Bxbl and attB. In contrast, 12,597 recurrent FDEs were detected in our inline control using Digenome-seq and Cas9 with the off-target standard sgRNA VEGFA site 2. Fig. 18C is an integrative genomics viewer (IGV) of HIDE-seq reads supporting recombination display distinct opposing read alignments. The top panel corresponds to on- target attB integration reads supporting recombination with attL (P’) and attR (P) sequences appended to the ends. The reads align with lentiviral genomic sequence in the HEK293attB genome and then appear soft-clipped (multicolored) where the linear attP substrates were recombined because these sequences do not exist in the human genome hg38 alignment reference file. The bottom panel corresponds to off-target integration where soft clipped reads that correspond to attL (P’) and attR (P) sequences are detected amongst the WGS coverage. In contrast to the on- target locus where all reads covering the locus were recombined, only a fraction of reads appear recombined at potential off-target loci. Fig. 18D HIDE-seq discovered 36 cryptic attB and 2 cryptic attP loci in the human genome. Fig. 18E is a DNA sequence motif logo of cryptic attB sites created by aligning the 46 bp sequence from the 36 cryptic attB sites. [0028] FIGs. 19A-19E. Cryptic-seq | Fig. 19A is a graphic of Cryptic-seq experimental workflow. First, a library of genomic DNA is created with Tn5 tagmentation that imparts partial Illumina sequencing adapters. This library of genomic DNA is then recombined with an LSI and attP or attB DNA substrates with complementary Illumina sequencing adapters. Potential off-target loci in the human genome that are recombined by the LSI are subject to one-step PCR enrichment and deep sequencing. Fig. 19B is a representative IGV browser of Cryptic-Seq reads supporting recombination at the on-target attB site and a cryptic attB. The top panel corresponds to on-target alignment of the left library (N7) with reads supporting the right junction (attR). The reads that align with lentiviral genomic sequence in the HEK293< //7> genome appear soft-clipped where the linear attP substrates were recombined because these sequences do not exist in the human genome hg38 alignment reference file.

The bottom panel corresponds to off-target alignment of the N7 library with reads supporting attR (in this view the cryptic site is in the antiparallel orientation such that the GT dinucleotide is on the bottom DNA strand so the attR reads appear on the left). The reads align with genomic sequence and then appear soft-clipped where the linear attP substrates were recombined because these sequences do not exist in the human genome hg38 alignment reference file. Fig. 19C is a summary table with the number of cryptic attB sites detected in the human genome. The "Sites per sample" section at the top indicates the total number of sites detected in each replicate, at increasing UMI detection level, up to >= 5000 unique recombination events. The "Recurrence" section summarizes the overlap analysis across replicates. The number of sites shared between the corresponding set of replicates is indicated, e.g., the top row "Repl,Rep2,Rep3" corresponds to sites observed in all three Cryptic-seq replicates, at increasing UMI levels. For sites shared between replicates, the lowest of the UMI values is used. As shown here, all sites > 50 UMIs were observed in all three replicates. The bottom row "Unique sites" section indicates the total number of unique sites across all three replicates, at increasing UMI levels. A total of 44,311 unique cryptic attB sites were detected in the hg38 reference human genome with 1 UMI or more. Fig. 19D is a Venn diagram of cryptic attB discovery overlap from HIDE-Seq and Cryptic-seq.

Fig. 19E is a DNA sequence motif logos, unweighted by UMI count (top) and weighted by UMI count (bottom), created by aligning the 46 bp sequence from the 44,311 cryptic attB sites.

[0029] FIGs. 20A-20G. Hybrid Capture NGS Validation of Off-Target Editing Fig. 20A is a graphic showing that K562attB cells were edited with Bxbl integrase and the purified gDNA was used for hybrid capture followed by NGS. Hybrid capture probes were designed on the left and right sides of the cryptic attachment loci to enable unbiased detection of indels, cargo integration and genomic rearrangements. Fig. 20B consists of graphs showing the sensitivity of off-target insertion detection with hybrid capture NGS was 0.1% when using a synthetic DNA standard spike-in from one of the off-target loci discovered, CAS031. r²= Pearson coefficient of determination Fig. 20C is a graph of K562attB cells edited with Bxbl had 45- 48% levels of insertion measured by ddPCR. Fig. 20D is a scatter plot showing hybrid capture NGS with single-sided capture probes detected 52 sites with off-target integration in K562attB cells and most validated off-target frequencies were at or below our lower limit of detection. Fig. 20E is a graph showing quantification of off-target insertion at CAS421 in K562attB cells using ddPCR confirms the frequency of off-target integration detected by hybrid capture NGS. Fig. 20F is a graph of 5 cryptic attB sites validated by WGS. Cryptic- seq discovered all 5 WGS-validated off-target sites but hybrid capture captured only 4 of the 5 sites detected by WGS. Cryptic-seq discovered all cryptic attB sites validated by WGS, HIDE-seq discovered only 1 of the WGS-validated sites. Fig. 20G is a graph Discovery assay classification errors for HIDE-seq and Cryptic-Seq reveal high level of false positive rate for Cryptic-seq but achieve the target goal of 0% false negative rate.

5. DETAILED DESCRIPTION

[0030] In one aspect, this disclosure features a method for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) in a target genome, the method comprising:

(a) fragmenting genomic DNA isolated from target cells;

(b) tagging the 5’ and 3’ ends of the DNA fragments with a first oligonucleotide adapter;

(c) contacting the tagged genomic DNA fragments with

(i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, and a second oligonucleotide adapter;

(ii) optionally at least a second species of synthetic DNA molecule, wherein the second species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence in the central dinucleotide from the first attachment site of the first species of synthetic DNA molecule, and a second oligonucleotide adapter; and

(iii) the selected LSI, under conditions suitable for the LSI to effect the recombination of the synthetic DNA molecule into the tagged genomic DNA fragments at second attachment sites that are capable of functioning as cognates of the first attachment site;

(d) generating a sequencing library comprising synthetic DNA molecules recombined with the tagged genomic DNA fragments;

(e) sequencing the sequencing library; and

(f) identifying cryptic attachment sites based on the sequencing data.

[0031] In another aspect, the present disclosure features a method for identifying DNA recombination events of a selected integrase in a target genome, the method comprising:

(a) contacting genomic DNA isolated from target cells with

(i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule is linear and comprises a first attachment site known to be recognized by the selected integrase,

(ii) the selected integrase, under conditions suitable for the integrase to effect the recombination of the genomic DNA;

(b) generating a sequencing library comprising the recombined genomic DNA;

(c) sequencing the sequencing library; and

(d) detecting DNA recombination events based on the sequencing data.

5.1. Terminology

[0032] Unless defined otherwise, all technical and scientific terms used herein have the meaning commonly understood by a person skilled in the art to which this invention belongs. As used herein, the following terms have the meanings ascribed to them below. [0033] “Cryptic attachment site,” “pseudo attachment site,” “cryptic integration recognition site,” “cryptic site,” and “pseudo site” are used interchangeably herein to refer to a site in a genome that resembles, with some level of DNA similarity, an integration or recombination site that can be recognized/recombined by an integration or recombination enzyme. In some embodiments, the integration site is the canonical phage attachment site attP or attB and is functionally capable of supporting recombination, resulting in integration of attP or attB containing DNA cargo into the genome. In mammalian cells, the phage integrase can mediate the integration of plasmids bearing the canonical phage attachment site into native sequences that have partial sequence identity with attP or attB.

[0034] “Gene editor” as used herein, is a protein that that can be used to perform gene editing, gene modification, gene insertion, gene deletion, or gene inversion. As used herein, the terms “gene editor polynucleotide” refers to polynucleotide sequence encoding the gene editor protein. In various embodiments, the gene editor comprises DNA- or RNA-targetable nuclease protein (i.e., Cas protein) wherein target specificity is mediated by a complexed nucleic acid (i.e., guide RNA). In various embodiments, the gene editor is a DNA-targetable or RNA-targetable protein in which target specificity is mediated by internal, conjugated, fused, or linked amino acids, such as within TALENs, ZFNs, or meganucleases. The skilled person in the art would appreciate that the gene editor can demonstrate targeted nuclease activity, targeted binding with no nuclease activity, targeted nickase activity (or cleavase activity). A gene editor comprising a targetable protein may be fused, linked, complexed, operate in cis or trans to one or more proteins or protein fragment motifs. Gene editors may be fused or linked to one or more integrase, recombinase, polymerase, telomerase, reverse transcriptase, or invertase. A gene editor can be an alpha editor fusion protein or a gene writer fusion protein.

[0035] Base editors such as ADAR or AD AT may be delivered as cargo in various embodiments, as described herein.

[0036] “Alpha editor protein or fusion protein” as used herein, is a gene editor protein. “Alpha editor system” as used herein describes the components used in alpha editing and alpha editing and alpha editor are used interchangeably herein with the terms “prime editing or prime editor (PE). Alpha editing uses a CRISPR protein, which has enzymatic activity that nicks or cuts only single strand of double stranded DNA, i.e., a nickase; the nickase can occur either naturally or by mutation or modification of a nuclease that makes double stranded cuts. Prime editing (PE) and prime editing components (pegRNA) can be utilized as well. The nickase is programmed (targeted) with an alpha-editing guide RNA (aeg RNA or a pegRNA). The skilled person in the art would appreciate that the pegRNA both specifies the target site and encodes the desired edit. Attachment site containing guide RNA (atgRNA) that both specifies the target and encodes for the desired integrase target recognition site are provided. The nickase may be programmed (targeted) with an atgRNA. Advantageously the nickase is a catalytically impaired Cas9 endonuclease, i.e., a Cas9 nickase, that is fused to a reverse transcriptase. The reverser transcriptase can also be provided in the alpha editor system split from the Cas9 nickase. During gene editing, the Cas9 nickase part of the protein is guided to the DNA target site by the atgRNA (or pegRNA), whereby a nick or single stranded cut occurs. The reverse transcriptase domain then uses the atgRNA (or pegRNA) to template reverse transcription of the desired edit, directly polymerizing DNA onto the nicked target DNA strand. The edited DNA strand replaces the original DNA strand, creating a heteroduplex containing one edited strand and one unedited strand. Afterward, optionally, the alpha editor (AE) guides resolution of the heteroduplex to favor copying the edit onto the unedited strand, completing the process (typically achieved with a nickase gRNA). Other enzymes that can be used to nick or cut only a single strand of double stranded DNA includes a cleavase (e.g., cleavase I enzyme).

[0037] In some embodiments, an additional agent or agents may be added that improve the efficiency and outcome purity of the alpha or prime edit. In some embodiments, the agent may be chemical or biological and disrupt DNA mismatch repair (MMR) processes at or near the edit site (i.e., PE4 and PE5 and PEmax architecture by Chen et al. Cell, 184, 1-18, October 28, 2021; Chen et al. is incorporated herein by reference). In typical embodiments, the agent is a MMR-inhibiting protein. In certain embodiments, the MMR-inhibiting protein is dominant negative MMR protein. In certain embodiments, the dominant negative MMR protein is MLHldn. In particular embodiments, the MMR-inhibiting agent is incorporated into the multicomponent delivery method described herein. In some embodiments, the MMR- inhibiting agent is linked or fused to the alpha editor protein fusion, which may or may not have a linked or fused integrase. In some embodiments, the MMR-inhibiting agent is linked or fused to the Gene Writer™ protein, which may or may not have a linked or fused integrase. [0038] The alpha editor or gene editor system can be used to achieve DNA deletion and replacement. In some embodiments, the DNA deletion replacement is induced using a pair of atgRNAs or pegRNA that target opposite DNA strands, programming not only the sites that are nicked but also the outcome of the repair (i.e., PrimeDel by Choi et al. Nat.

Biotechnology, October 14, 2021; Choi et al. is incorporated herein by reference and TwinPE by Anzalone et al. BioRxiv, November 2, 2021; Anzalone et al. is incorporated herein by reference). In some embodiments described herein, the DNA deletion is induced using a single atgRNA. In some embodiments, the DNA deletion and replacement is induced using a wild type Cas9 alpha or prime editor (AE-Cas9 or PE-Cas9) system (i.e., PED AR by Jiang et al. Nat. Biotechnology, October 14, 2021; Jiang et al. is incorporated herein by reference in its entirety). In some embodiments, the DNA replacement is an integrase target recognition site or recombinase target recognition site. In certain embodiments, the constructs and methods described herein may be utilized to incorporate the pair of pegRNAs (or atgRNAs) used in PrimeDel, TwinPE (WO2021226558 incorporated by reference herein in its entirety), or PED AR, the alpha editor fusion protein or Gene Writer protein, optionally a nickase guide RNA (ngRNA), an integrase, a nucleic acid cargo, and optionally a recombinase into a LNP delivery system or vector delivery system (e.g., AAV or Adenovirus). The integrase may be directly linked, for example by a peptide linker, to the prime editor fusion or gene writer protein.

[0039] In some embodiments, the alpha editors can refer to a retrovirus or lentivirus reverse transcriptase such as a Moloney Murine Leukemia Virus (M-MLV) reverse transcriptase (RT) fused to a CRISPR enzyme nickase such as a Cas9 H840A nickase, a Cas9nickase. In some embodiments, the alpha editors can refer to a retrovirus or lentivirus reverse transcriptase such as a Moloney Murine Leukemia Virus (M-MLV) reverse transcriptase (RT) fused to a cleavase. In some embodiments the RT can be fused at, near or to the C- terminus of a Cas9nickase, e.g., Cas9 H840A. Fusing the RT to the C-terminus region, e.g., to the C-terminus, of the Cas9 nickase may result in higher editing efficiency. Such a complex is called PEI. In some embodiments, the CRISPR enzyme nickase, e.g., Cas9(H840A), i.e., a Cas9nickase, can be linked to a non-M-MLV reverse transcriptase such as an AMV-RT or XRT (Cas9(H840A)-AMV-RT or XRT). In some embodiments, instead of the CRISPR enzyme nickase being a Cas9 (H840A), i.e., instead of being a Cas9 nickase, the CRISPR enzyme nickase instead can be a CRISPR enzyme that naturally is a nickase or cuts a single strand of double stranded DNA; for instance, the CRISPR enzyme nickase can be Casl2a/b. Alternatively, the CRISPR enzyme nickase can be another mutation of Cas9, such as Cas9(D10A). A CRISPR enzyme, such as a CRISPR enzyme nickase, such as Cas9 (wild type), Cas9(H840A), Cas9(D10A) or Cas 12a/b nickase can be fused in some embodiments to a pentamutant of M-MLV RT (D200N/ L603W/ T330P/ T306K/ W313F), whereby there can be up to about 45-fold higher efficiency, and this is called PE2. In some embodiments, the M- MLV RT comprise one or more of the mutations Y8H, P51L, S56A, S67R, E69K, V129P, L139P, T197A, H204R, V223H, T246E, N249D, E286R, Q2911, E302K, E302R, F309N, M320L, P330E, L435G, L435R, N454K, D524A, D524G, D524N, E562Q, D583N, H594Q, E607K, D653N, and L671P. Specific M-MLV RT mutations are shown in Table 1.

[0040] In some embodiments, the reverse transcriptase can also be a wild-type or modified transcription xenopolymerase (RTX), avian myeloblastosis virus reverse transcriptase (AMV RT), Feline Immunodeficiency Virus reverse transcriptase (FIV-RT), FeLV-RT (Feline leukemia virus reverse transcriptase), HIV-RT (Human Immunodeficiency Virus reverse transcriptase). In some embodiments, the reverse transcriptase can be a fusion of MMuLV to the Sto7d DNA binding domain (see lonnidi et al.; doi.org/10.1101/2021.11.01.466786). The fusion of MMuLV to the Sto7d DNA binding domain sequence is given in Table 2. [0041] PE3, PE3b, PE4, PE5, and/or PEmax, which a skilled person can incorporate into the co-delivery system described herein, involves nicking the non-edited strand, potentially causing the cell to remake that strand using the edited strand as the template to induce HR. The nicking of the non-edited strand can involve the use of a nicking guide RNA (ngRNA).

[0042] The skilled person can readily incorporate into the co-delivery system described herein described herein a prime editing or CRISPR system. Examples of prime editors can be found in the following: W02020/191153, W02020/191171, WO2020/191233, WO2020/191234, WO2020/191239, W02020/191241, WO2020/191242, WO2020/191243, WO2020/191245, WO2020/191246, WO2020/191248, WO2020/191249, each of which is incorporated by reference herein in its entirety. In addition, mention is made, and can be used herein, of CRISPR Patent Applications and Patents of the Zhang laboratory and/or Broad Institute, Inc. and Massachusetts Institute of Technology and/or Broad Institute, Inc., Massachusetts Institute of Technology and President and Fellows of Harvard College and/or Editas Medicine, Inc. Broad Institute, Inc., The University of Iowa Research Foundation and Massachusetts Institute of Technology, including those claiming priority to US Application 61/736,527, filed December 12, 2012, including US Patents 11,104,937, 11,091,798, 11,060,115, 11,041,173, 11,021,740, 11,008,588, 11,001,829, 10,968,257, 10,954,514, 10,946,108, 10,930,367, 10,876,100, 10,851,357, 10,781,444, 10,711,285, 10,689,691, 10,648,020, 10,640,788, 10,577,630, 10,550,372, 10,494,621, 10,377,998, 10,266,887, 10,266,886, 10,190,137, 9,840,713, 9,822,372, 9,790,490, 8,999,641, 8,993,233, 8,945,839, 8,932,814, 8,906,616, 8,895,308, 8,889,418, 8,889,356, 8,871,445, 8,865,406, 8,795,965, 8,771,945, and 8,697,359; CRISPR Patent Applications and Patents of the Doudna laboratory and/or of Regents of the University of California, the University of Vienna and Emmanuelle Charpentier, including those claiming priority to US application 61/652,086, filed May 25, 2012, and/or 61/716,256, filed October 19, 2012, and/or 61/757,640, filed January 28, 2013, and/or 61/765,576, filed February 15, 2013 and/or 13/842,859, including US Patents 11,028,412, 11,008,590, 11,008,589, 11,001,863, 10,988,782, 10,988,780, 10,982,231, 10,982,230, 10,900,054, 10,793,878, 10,774,344, 10,752,920, 10,676,759, 10,669,560, 10,640,791, 10,626,419, 10,612,045, 10,597,680, 10,577,631, 10,570,419, 10,563,227, 10,550,407, 10,533,190, 10,526,619, 10,519,467, 10,513,712, 10,487,341, 10,443,076, 10,428,352, 10,421,980, 10,415,061, 10,407,697, 10,400,253, 10,385,360, 10,358,659, 10,358,658, 10,351,878, 10,337,029, 10,308,961, 10,301,651, 10,266,850, 10,227,611, 10,113,167, and 10,000,772; CRISPR Patent Applications and Patents of Vilnius University and/or the Siksnys laboratory, including those claiming priority to US application 62/046384 and/or 61/625,420 and/or 61/613,373 and/or PCT/IB2015/056756, including US Patent 10,385,336; CRISPR Patent Applications and Patents of the President and Fellows of Harvard College, including those of George Church’s laboratory and/or claiming priority to US application 61/738,355, filed December 17, 2012, including 11,111,521, 11,085,072, 11,064,684, 10,959,413, 10,925,263, 10,851,369, 10,787,684, 10,767,194, 10,717,990, 10,683,490, 10,640,789, 10,563,225, 10,435,708, 10,435,679, 10,375,938, 10,329,587, 10,273,501, 10,100,291, 9,970,024, 9,914,939, 9,777,262, 9,587,252, 9,267,135, 9,260,723, 9,074,199, 9,023,649; CRISPR Patent Applications and Patents of the President and Fellows of Harvard College, including those of David Liu’s laboratory, including 11,111,472, 11,104,967, 11,078,469, 11,071,790, 11,053,481, 11,046,948, 10,954,548, 10,947,530, 10,912,833, 10,858,639, 10,745,677, 10,704,062, 10,682,410, 10,612,011, 10,597,679, 10,508,298, 10,465,176, 10,323,236, 10,227,581, 10,167,457, 10,113,163, 10,077,453, 9,999,671, 9,840,699, 9,737,604, 9,526,784, 9,388,430, 9,359,599, 9,340,800, 9,340,799, 9,322,037, 9,322,006, 9,228,207, 9,163,284, and 9,068,179; and CRISPR Patent Applications and Patents of Toolgen Incorporated and/or the Kim laboratory and/or claiming priority to US application 61/717,324, filed October 23, 2012 and/or 61/803,599, filed March 20, 2013 and/or 61/837,481, filed une 20, 2013 and/or 62/033,852, filed August 6, 2014 and/or PCT/KR2013/009488 and/or PCT/KR2015/008269, including US Patent 10,851,380, and 10,519,454; and CRISPR Patent Applications and Patents of Sigma and/or Millipore and/or the Chen laboratory and/or claiming priority to US application 61/734,256, filed December 6, 2012 and/or 61/758,624, filed lanuary 30, 2013 and/or 61/761,046, filed February 5, 2013 and/or 61/794,422, filed March 15, 2013, including US Patent 10,731,181, each of which is hereby incorporated herein by reference, and from the disclosures of the foregoing, the skilled person can readily make and use a prime editing or CRISPR system, and can especially appreciate impaired endonucleases, such as a mutated Cas9 that only nicks a single strand of DNA and is hence a nickase, or a CRISPR enzyme that only makes a single-stranded cut that can be employed in a PASTE system of the invention. Further, from the disclosures of the foregoing, the skilled person can incorporate the selected CRISPR enzyme, as part of the prime editor fusion or gene editor fusion, into the codelivery method described herein.

[0043] Prior to RT -mediated edit incorporation, the prime editor protein (or system) (1) site-specifically targets a genomic locus and (2) performs a catalytic cut or nick. These steps are typically performed by a CRISPR-Cas. However, in some embodiments the Cas protein may be substituted by other nucleic acid programmable DNA binding proteins (napDNAbp) such as zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), or meganucleases. In addition, to the extent the “targeting rules” of other napDNAbp are known or are newly determined, it becomes possible to use new napDNAbp, beyond Cas9, to site specifically target and modify genomic sites of interest.

[0044] Similar to a prime editor protein, a Gene Writer can introduce novel DNA elements, such as an integration target site, into a DNA locus. A Gene Writer protein comprises: (A) a polypeptide or a nucleic acid encoding a polypeptide, wherein the polypeptide comprises (i) a reverse transcriptase domain, and either (x) an endonuclease domain that contains DNA binding functionality or (y) an endonuclease domain and separate DNA binding domain; and (B) a template RNA comprising (i) a sequence that binds the polypeptide and (ii) a heterologous insert sequence. Examples of such Gene Writer™ proteins and related systems can be found in US20200109398, which is incorporated by reference herein in its entirety.

[0045] In some embodiments, the prime editor or Gene Writer protein fusion or prime editor protein linked or fused to an integrase is expressed as a split construct. In typical embodiments, the split construct in reconstituted in a cell. In some embodiments, the split construct can be fused or ligated via intein protein splicing. In some embodiments, the split construct can be reconstituted via protein-protein inter-molecular bonding and/or interactions. In some embodiments, the split construct can be reconstituted via chemical, biological, or environmental induced oligomerization. In certain embodiments, the split construct can be adapted into one or more delivery vectors described herein.

[0046] In some embodiments, an integrase or recombinase is directly linked or fused, for example by a peptide linker, which may be cleavable or non-cleavable, to the prime editor fusion protein (i.e., fused Cas9 nickase-reverse transcriptase) or Gene Writer protein. Suitable linkers, for example between the Cas9, RT, and integrase, may be selected from Table 3:

[0047] In some embodiments, the prime editor or Gene Writer protein fusion or prime editor protein linked or fused to an integrase is expressed as a split construct. In typical embodiments, the split construct in reconstituted in a cell. In some embodiments, the split construct can be fused or ligated via intein protein splicing. In some embodiments, the split construct can be reconstituted via protein-protein inter-molecular bonding and/or interactions. In some embodiments, the split construct can be reconstituted via chemical, biological, or environmental induced oligomerization. In certain embodiments, the split construct can be adapted into one or more nucleic acid constructs described herein.

5.2. Type II CRISPR proteins

[0048] The skilled person can incorporate a selected CRISPR enzyme, described below, as part of the prime editor fusion, into the co-delivery method described herein. Streptococcus pyogenes Cas9 (SpCas9), the most common enzyme used in genome-editing applications, is a large nuclease of 1368 amino acid residues. Advantages of SpCas9 include its short, 5'-NGG-3' PAM and very high average editing efficiency. SpCas9 consists of two lobes: a recognition (REC) lobe and a nuclease (NUC) lobe. The REC lobe can be divided into three regions, a long a helix referred to as the bridge helix (residues 60-93), the RECI (residues 94-179 and 308-713) domain, and the REC2 (residues 180-307) domain. The NUC lobe consists of the RuvC (residues 1-59, 718-769, and 909-1098), HNH (residues 775- 908), and PAM-interacting (PI) (residues 1099-1368) domains. The negatively charged sgRNA:target DNA heteroduplex is accommodated in a positively charged groove at the interface between the REC and NUC lobes. In the NUC lobe, the RuvC domain is assembled from the three split RuvC motifs (RuvC I— III) and interfaces with the PI domain to form a positively charged surface that interacts with the 30 tail of the sgRNA. The HNH domain lies between the RuvC II— III motifs and forms only a few contacts with the rest of the protein. Structural aspects of SpCas9 are described by Nishimasu et al., Crystal Structure of Cas9 in Complex with Guide RNA and Target DNA, Cell 156, 935-949, February 27, 2014.

[0049] REC lobe'. The REC lobe includes the RECI and REC2 domains. The REC2 domain does not contact the bound guide Target heteroduplex, indicating that truncation of REC lobe may be tolerated by SpCas9. Further, SpCas9 mutant lacking the REC2 domain (D175-307) retained -50% of the wild-type Cas9 activity, indicating that the REC2 domain is not critical for DNA cleavage. In striking contrast, the deletion of either the repeat-interacting region (D97-150) or the anti-repeat-interacting region (D312-409) of the RECI domain abolished the DNA cleavage activity, indicating that the recognition of the repeat: anti -repeat duplex by the RECI domain is critical for the Cas9 function.

[0050] PAM-Inter acting domain'. The NUC lobe contains the PAM-interacting (PI) domain that is positioned to recognize the PAM sequence on the noncomplementary DNA strand. The PI domain of SpCas9 is required for the recognition of 5’-NGG-3’ PAM, and deletion of the PI domain (A1099-1368) abolished the cleavage activity, indicating that the PI domain is critical for SpCas9 function and a major determinant for the PAM specificity.

[0051] RuvC domain'. The RuvC nucleases of SpCas9 have an RNase H fold and four catalytic residues, AsplO (Ala), Glu762, His983, and Asp986, that are critical for the two- metal cleavage of the noncomplementary strand of the target DNA. In addition to the conserved RNase H fold, the Cas9 RuvC domain has other structural elements involved in interactions with the guide:target heteroduplex (an end-capping loop between a42 and a43) and the PI domain/stem loop 3 (P hairpin formed by P3 and [34).

[0052] HNH domain'. SpCas9 HNH nucleases have three catalytic residues, Asp839, His840, and Asn863 and cleave the complementary strand of the target DNA through a single-metal mechanism.

[0053] sgRNA:DNA recognition'. The sgRNA guide region is primarily recognized by the REC lobe. The backbone phosphate groups of the guide region (nucleotides 2, 4-6, and 13- 20) interact with the RECI domain (Argl65, Glyl66, Arg403, Asn407, Lys510, Tyr515, and Arg661) and the bridge helix (Arg63, Arg66, Arg70, Arg71, Arg74, and Arg78). The 20- hydroxyl groups of Gl, C15, U16, and G19 hydrogen bond with Vai 1009, Tyr450, Arg447/Ile448, and Thr404, respectively. [0054] A mutational analysis demonstrated that the R66A, R70A, and R74A mutations on the bridge helix markedly reduced the DNA cleavage activities, highlighting the functional significance of the recognition of the sgRNA “seed” region by the bridge helix. Although Arg78 and Argl65 also interact with the “seed” region, the R78A and R165A mutants showed only moderately decreased activities. These results are consistent with the fact that Arg66, Arg70, and Arg74 form multiple salt bridges with the sgRNA backbone, whereas Arg78 and Argl65 form a single salt bridge with the sgRNA backbone. Moreover, the alanine mutations of the repeat: anti -repeat duplex-interacting residues (Arg75 and Lysl63) and the stemloop- 1 -interacting residue (Arg69) resulted in decreased DNA cleavage activity, confirming the functional importance of the recognition of the repeat: anti -repeat duplex and stem loop 1 by Cas9.

[0055] RNA-guided DNA targeting'. SpCas9 recognizes the guide:target heteroduplex in a sequence-independent manner. The backbone phosphate groups of the target DNA (nucleotides 1, 9-11, 13, and 20) interact with the RECI (Asn497, Trp659, Arg661, and Gln695), RuvC (Gln926), and PI (Glul 108) domains. The C2’ atoms of the target DNA (nucleotides 5, 7, 8, 11, 19, and 20) form van der Waals interactions with the RECI domain (Leul69, Tyr450, Met495, Met694, and His698) and the RuvC domain (Ala728). The terminal base pair of the guide:target heteroduplex (G1 :C20’) is recognized by the RuvC domain via end-capping interactions; the sgRNA G1 and target DNA C20’ nucleobases interact with the Tyrl013 and Vai 1015 side chains, respectively, whereas the 20-hydroxyl and phosphate groups of sgRNA G1 interact with Vall009 and Gln926, respectively.

[0056] Repeat: Anti-Repeat duplex recognition'. The nucleobases of U23/A49 and A42/G43 hydrogen bond with the side chain of Argl 122 and the main-chain carbonyl group of Phe351, respectively. The nucleobase of the flipped U44 is sandwiched between Tyr325 and His328, with its N3 atom hydrogen bonded with Tyr325, whereas the nucleobase of the unpaired G43 stacks with Tyr359 and hydrogen bonds with Asp364.

[0057] The nucleobases of G21 and U50 in the G21 :U50 wobble pair stack with the terminal C20:G10 pair in the guide:target heteroduplex and Tyr72 on the bridge helix, respectively, with the U50 04 atom hydrogen bonded with Arg75. Notably, A51 adopts the syn conformation and is oriented in the direction opposite to U50. The nucleobase of A51 is sandwiched between Phel 105 and U63, with its Nl, N6, and N7 atoms hydrogen bonded with G62, Glyl 103, and Phel 105, respectively. [0058] Stem-loop recognition'. Stem loop 1 is primarily recognized by the REC lobe, together with the PI domain. The backbone phosphate groups of stem loop 1 (nucleotides 52, 53, and 59-61) interact with the RECI domain (Leu455, Ser460, Arg467, Thr472, and Ile473), the PI domain (Lysl 123 and Lysl 124), and the bridge helix (Arg70 and Arg74), with the 20- hydroxyl group of G58 hydrogen bonded with Leu455. A52 interacts with Phel 105 through a face-to-edge p-p stacking interaction, and the flipped U59 nucleobase hydrogen bonds with Asn77.

[0059] The single-stranded linker and stem loops 2 and 3 are primarily recognized by the NUC lobe. The backbone phosphate groups of the linker (nucleotides 63-65 and 67) interact with the RuvC domain (Glu57, Lys742, and Lysl097), the PI domain (Thrl 102), and the bridge helix (Arg69), with the 20-hydroxyl groups of U64 and A65 hydrogen bonded with Glu57 and His721, respectively. The C67 nucleobase forms two hydrogen bonds with Vail 100.

[0060] Stem loop 2 is recognized by Cas9 via the interactions between the NUC lobe and the non-Watson-Crick A68:G81 pair, which is formed by direct (between the A68 N6 and G81 06 atoms) and water-mediated (between the A68 N1 and G81 N1 atoms) hydrogen-bonding interactions. The A68 and G81 nucleobases contact Serl351 and Tyrl356, respectively, whereas the A68:G81 pair interacts with Thrl358 via a water-mediated hydrogen bond. The 20-hydroxyl group of A68 hydrogen bonds with Hisl349, whereas the G81 nucleobase hydrogen bonds with Lys33.

[0061] Stem loop 3 interacts with the NUC lobe more extensively, as compared to stem loop 2. The backbone phosphate group of G92 interacts with the RuvC domain (Arg40 and Lys44), whereas the G89 and U90 nucleobases hydrogen bond with Gin 1272 and Glul225/Alal227, respectively. The A88 and C91 nucleobases are recognized by Asn46 via multiple hydrogen-bonding interactions.

[0062] Cas9 proteins smaller than SpCas9 allow more efficient packaging of nucleic acids encoding CRISPR systems, e.g., Cas9 and sgRNA into one rAAV (“all-in-one-AAV”) particle. In addition, efficient packaging of CRISPR systems can be achieved in other viral vector systems (i.e., lentiviral, integration deficient lentiviral, hd-AAV, etc.) and non-viral vector systems (i.e., lipid nanoparticle). Small Cas9 proteins can be advantageous for multidomain-Cas-nuclease-based systems for prime editing. Well characterized smaller Cas9 proteins include Staphylococcus aureus (SauCas9, 1053 amino acid residues) and Campylobacter jejuni (CjCas9, 984 amino residues). However, both recognize longer PAMs, 5 '-NNGRRT-3 ' for SauCas9 (R = A or G) and 5 '-NNNNRYAC-3 ' for Cj Cas9 (Y = C or T), which reduces the number of uniquely addressable target sites in the genome, in comparison to the NGG SpCas9 PAM. Among smaller Cas9s, Schmidt et al. identified Staphylococcus lugdunensis (Siu) Cas9 as having genome-editing activity and provided homology mapping to SpCas9 and SauCas9 to facilitate generation of nickases and inactive (“dead”) enzymes (Schmidt et al., 2021, Improved CRISPR genome editing using small highly active and specific engineered RNA-guided nucleases. Nat Commun 12, 4219. doi.org/10.1038/s41467- 021-24454-5) and engineered nucleases with higher cleavage activity by fragmenting and shuffling Cas9 DNAs. The small Cas9s and nickases are useful in the instant invention.

[0063] Besides dead Cas9 and Cas9 nickase variants, the Cas9 proteins used herein may also include other “Cas9 variants” having at least about 70% identical, at least about 80% identical, at least about 90% identical, at least about 95% identical, at least about 96% identical, at least about 97% identical, at least about 98% identical, at least about 99% identical, at least about 99.5% identical, or at least about 99.9% identical to any reference Cas9 protein, including any wild type Cas9, or mutant Cas9 (e.g., a dead Cas9 or Cas9 nickase), or fragment Cas9, or circular permutant Cas9, or other variant of Cas9 disclosed herein or known in the art. In some embodiments, a Cas9 variant may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 21, 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, 50 or more amino acid changes compared to a reference Cas9. In some embodiments, the Cas9 variant comprises a fragment of a reference Cas9 (e.g., a gRNA binding domain or a DNA-cleavage domain), such that the fragment is at least about 70% identical, at least about 80% identical, at least about 90% identical, at least about 95% identical, at least about 96% identical, at least about 97% identical, at least about 98% identical, at least about 99% identical, at least about 99.5% identical, or at least about 99.9% identical to the corresponding fragment of wild type Cas9. In some embodiments, the fragment is at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% identical, at least 96%, at least 97%, at least 98%, at least 99%, or at least 99.5% of the amino acid length of a corresponding wild type Cas9 (e.g., SEQ ID NO: 18). [0064] In some embodiments, the disclosure also may utilize Cas9 fragments that retain their functionality and that are fragments of any herein disclosed Cas9 protein. In some embodiments, the Cas9 fragment is at least 100 amino acids in length. In some embodiments, the fragment is at least 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1050, 1100, 1150, 1200, 1250, or at least 1300 amino acids in length.

[0065] In various embodiments, the prime editors disclosed herein may comprise one of the Cas9 variants described as follows, or a Cas9 variant thereof having at least about 70% identical, at least about 80% identical, at least about 90% identical, at least about 95% identical, at least about 96% identical, at least about 97% identical, at least about 98% identical, at least about 99% identical, at least about 99.5% identical, or at least about 99.9% identical to any reference Cas9 variants.

[0066] In some embodiments, prime editors utilized herein comprise CRISPR-Cas system enzymes other than type II enzymes. In certain embodiments, prime editors comprise type V or type VI CRISPR-Cas system enzymes. It will be appreciated that certain CRISPR enzymes exhibit promiscuous ssDNA cleavage activity and appropriate precautions should be considered. In certain embodiments, prime editors comprise a nickase or a dead CRISPR with nuclease function comprised in a different component. [0067] In various embodiments, the nucleic acid programmable DNA binding proteins utilized herein include, without limitation, Cas9 (e.g., dCas9 and nCas9), Casl2a (Cpfl), Casl2bl (C2cl), Casl2b2, Casl2c (C2c3), Casl2d (CasY), Casl2e (CasX), C2c4, C2c5, C2c8, C2c9, C2cl0, Cast 3a (C2c2), Cast 3b (C2c6), Cast 3c (C2c7), Cast 3d, and Argonaute. Cas-equivalents further include those described in Makarova et al., “C2c2 is a singlecomponent programmable RNA-guided RNA-targeting CRISPR effector,” Science 2016; 353(6299) and Makarova et al., “Classification and Nomenclature of CRISPR-Cas Systems: Where from Here?,” The CRISPR Journal, Vol.l. No.5, 2018, the contents of which are incorporated herein by reference. One example of a nucleic acid programmable DNA-binding protein that has different PAM specificity than Cas9 is Clustered Regularly Interspaced Short Palindromic Repeats from Prevotella and Francisella 1 (i.e, Casl2a (Cpfl)). Similar to Cas9, Casl2a (Cpfl) is also a Class 2 CRISPR effector, but it is a member of type V subgroup of enzymes, rather than the type II subgroup. It has been shown that Casl2a (Cpfl) mediates robust DNA interference with features distinct from Cas9. Casl2a (Cpfl) is a single RNA- guided endonuclease lacking tracrRNA, and it utilizes a T-rich protospacer-adjacent motif (TTN, TTTN, or YTN). Moreover, Cpfl cleaves DNA via a staggered DNA double-stranded break. Out of 16 Cpfl -family proteins, two enzymes from Acidaminococcus and Lachnospiraceae are shown to have efficient genome-editing activity in human cells. Cpfl proteins are known in the art and have been described previously, for example Yamano et al., “Crystal structure of Cpfl in complex with guide RNA and target DNA.” Cell (165) 2016, p.949-962; the entire contents of which is hereby incorporated by reference.

5.3. Type V CRISPR proteins

[0068] In some embodiments, prime editors used herein comprise the type V CRISPR family includes Francisella novicida U112 Cpfl (FnCpfl) also known as FnCasl2a. FnCpfl adopts a bilobed architecture with the two lobes connected by the wedge (WED) domain. The N- terminal REC lobe consists of two a-helical domains (RECI and REC2) that have been shown to coordinate the crRNA-target DNA heteroduplex. The C-terminal NUC lobe consists of the C-terminal RuvC and Nuc domains involved in target cleavage, the arginine-rich bridge helix (BH), and the PAM-interacting (PI) domain. The repeat-derived segment of the crRNA forms a pseudoknot stabilized by intra-molecular base-pairing and hydrogen-bonding interactions. The pseudoknot is coordinated by residues from the WED, RuvC, and REC2 domains, as well as by two hydrated magnesium cations. Notably, nucleotides 1-5 of the crRNA are ordered in the central cavity of FnCasl2a and adopt an A-form-like helical conformation. Conformational ordering of the seed sequence is facilitated by multiple interactions between the ribose and phosphate moieties of the crRNA backbone and FnCpfl residues in the WED and RECI domains. These include residues Thrl6, Lys595, His804, and His881 from the WED domain and residues Tyr47, Lys51, Phel82, and Argl86 from the RECI domain. The structure of the FnCasl2a-crRNA complex further reveals that the bases of the seed sequence are solvent exposed and poised for hybridization with target DNA. Structural aspects of FnCpfl are described by Swarts et al., Structural Basis for Guide RNA Processing and Seed-Dependent DNA Targeting by CRISPR-Casl2a, Molecular Cell 66, 221-233, April 20, 2017.

[0069] Pre-crRNA processing'. Essential residues for crRNA processing include His843, Lys852, and Lys869. Structural observations are consistent with an acid-base catalytic mechanism in which Lys869 acts as the general base catalyst to deprotonate the attacking 2’- hydroxyl group of U(-19), while His843 acts as a general acid to protonate the 5’-oxygen leaving group of A(-18). In turn, the side chain of Lys852 is involved in charge stabilization of the transition state. Collectively, these interactions facilitate the intra-molecular attack of the 20-hydroxyl group of U(-19) on the scissile phosphate and promote the formation of the 2’, 3 ’-cyclic phosphate product.

[0070] R-loop formation'. The crRNA-target DNA strand heteroduplex is enclosed in the central cavity formed by the REC and NUC lobes and interacts extensively with the RECI and REC2 domains. The PAM-containing DNA duplex comprises target strand nucleotides dT0-dT8 and non-target strand nucleotides dA(8)*-dA0* and is contacted by the PI, WED, and RECI domains. The 5’-TTN-3’ PAM is recognized in FnCasl2a by a mechanism combining the shape-specific recognition of a narrowed minor groove, with base-specific recognition of the PAM bases by two invariant residues, Lys671 and Lys613. Directly downstream of the PAM, the duplex of the target DNA is disrupted by the side chain of residue Lys667, which is inserted between the DNA strands and forms a cation-7t stacking interaction with the dA0-dT0* base pair. The phosphate group linking target strand residues dT(-l) and dTO is coordinated by hydrogen-bonding interactions with the side chain of Lys823 and the backbone amide of Gly826. Target strand residue dT(-l) bends away from residue TO, allowing the target strand to interact with the seed sequence of the crRNA. The non-target strand nucleotides dTl*-dT5* interact with the Arg692-Ser702 loop in FnCasl2a through hydrogen-bonding and ionic interactions between backbone phosphate groups and side chains of Arg692, Asn700, Ser702, and Gln704, as well as main-chain amide groups of Lys699, Asn700, and Ser702. Alanine substitution of Q704 or replacement of residues Thr698-Ser702 in FnCasl2a with the sequence Ala-Gly3 (SEQ ID NO: 115) substantially reduced DNA cleavage activity, suggesting that these residues contribute to R-loop formation by stabilizing the displaced conformation of the nontarget DNA strand.

[0071] In the FnCasl2a R-loop complex, the crRNA-target strand heteroduplex is terminated by a stacking interaction with a conserved aromatic residue (Tyr410). This prevents base pairing between the crRNA and the target strand beyond nucleotides U20 and dA(-20), respectively. Beyond this point, the target DNA strand nucleotides re-engage the non-target DNA strand, forming a PAM-distal DNA duplex comprising nucleotides dC(-21)-dA(-27) and dG21*-dT27*, respectively. The duplex is confined between the REC2 and Nuc domains at the end of the central channel formed by the REC and NUC lobes.

[0072] Target DNA cleavage'. FnCpfl can independently accommodate both the target and non-target DNA strands in the catalytic pocket of the RuvC domain. The RuvC active site contains three catalytic residues (D917, E1006, and D1255). Structural observations suggest that both the target and non-target DNA strands are cleaved by the same catalytic mechanism in a single active site in Cpfl/Casl2a enzymes.

[0073] Another type V CRISPR is AsCpfl from Acidaminococcus sp BV3L6 (Yamano et al., Crystal structure of Cpfl in complex with guide RNA and target DNA, Cell 165, 949-962, May 5, 2016)

[0074] In certain embodiments, the nuclease comprises a Casl2f effector. Small CRISPR- associated effector proteins belonging to the type V-F subtype have been identified through the mining of sequence databases and members classified into Casl2fl (Casl4a and type V- U3), Casl2f2 (Casl4b) and Casl2f3 (Casl4c, type V-U2 and U4). (See, e.g., Karvelis et al., PAM recognition by miniature CRISPR-Casl2f nucleases triggers programmable doublestranded DNA target cleavage. Nucleic Acids Research, 21 May 2020, 48(9), 5016-23 doi.org/10.1093/nar/gkaa208). Xu et al. described development of a 529 amino acid Casl2f- based system for mammalian genome engineering through multiple rounds of iterative protein engineering and screening. (Xu, X. et al., Engineered Miniature CRISPR-Cas System for Mammalian Genome Regulation and Editing. Molecular Cell, October 21, 2021, 81(20): 4333-45, doi.org/10.1016/j.molcel.2021.08.008). [0075] Exemplary CRISPR-Cas proteins and enzymes used in the prime editors herein include the following without limitation.

5.4. Protospacer Adjacent Motif

[0076] As used herein, the term “protospacer adjacent sequence” or “protospacer adjacent motif’ or “PAM” refers to an approximately 2-6 base pair DNA sequence (or a 2-, 3-, 4-, 5-, 6-, 7-, 8-, 9-, 10-, 11-, 12-long nucleotide sequence) that is an important targeting component of a Cas9 nuclease. Typically, the PAM sequence is on either strand, and is downstream in the 5' to 3' direction of Cas9 cut site. The canonical PAM sequence (i.e., the PAM sequence that is associated with the Cas9 nuclease of Streptococcus pyogenes or SpCas9) is 5'-NGG-3' wherein “N” is any nucleobase followed by two guanine (“G”) nucleobases. Different PAM sequences can be associated with different Cas9 nucleases or equivalent proteins from different organisms. In addition, any given Cas9 nuclease may be modified to alter the PAM specificity of the nuclease such that the nuclease recognizes alternative PAM sequence.

[0077] For example, with reference to the canonical SpCas9 amino acid sequence, the PAM specificity can be modified by introducing one or more mutations, including (a) DI 135V, R1335Q, and T1337R “the VQR variant”, which alters the PAM specificity to NGAN or NGNG, (b) DI 135E, R1335Q, and T1337R “the EQR variant”, which alters the PAM specificity to NGAG, and (c) DI 135V, G1218R, R1335E, and T1337R “the VRER variant”, which alters the PAM specificity to NGCG. In addition, the DI 135E variant of canonical SpCas9 still recognizes NGG, but it is more selective compared to the wild type SpCas9 protein.

[0078] It will also be appreciated that Cas9 enzymes from different bacterial species (i.e., Cas9 orthologs) can have varying PAM specificities and some embodiments are therefore chosen based on the desired PAM recognition. For example, Cas9 from Staphylococcus aureus (SaCas9) recognizes NGRRT or NGRRN. In addition, Cas9 from Neisseria meningitis (NmCas) recognizes NNNNGATT. In another example, Cas9 from Streptococcus thermophilis (StCas9) recognizes NNAGAAW. In still another example, Cas9 from Treponema denticola (TdCas) recognizes NAAAAC. These examples are not meant to be limiting. It will be further appreciated that non-SpCas9s bind a variety of PAM sequences, which makes them useful to expand the range of sequences that can be targeted according to the invention. Furthermore, non-SpCas9s may have other characteristics that make them more useful than SpCas9. For example, Cas9 from Staphylococcus aureus (SaCas9) is about 1 kilobase smaller than SpCas9, so it can be packaged into adeno-associated virus (AAV). Further reference may be made to Shah et al., “Protospacer recognition motifs: mixed identities and functional diversity,” RNA Biology, 10(5): 891-899 (which is incorporated herein by reference). Gasiunas used cell-free biochemical screens to identify protospacer adjacent motif (PAM) and guide RNA requirements of 79 Cas9 proteins. (Gasiunas et al., A catalogue of biochemically diverse CRISPR-Cas9 orthologs, Nature Communications 11 :5512 doi.org/10.1038/s41467-020-19344-l) The authors described 7 classes of gRNA and 50 different PAM requirement.

[0079] Oh, Y. et al. describe linking reverse transcriptase to a Francisella novicida Cas9 [FnCas9(H969A)] nickase module. (Oh, Y. et al., Expansion of the prime editing modality with Cas9 from Francisella novicida, bioRxiv 2021.05.25.445577; doi.org/10.1101/2021.05.25.445577). By increasing the distance to the PAM, the FnCas9(H969A) nickase module expands the region of a reverse transcription template (RTT) following the primer binding site.

5.5. Prime Editors

[0080] “Prime editor fusion protein” describes a protein that is used in prime editing. Prime editing uses CRISPR enzyme that nicks or cuts only single strand of double stranded DNA, i.e., a nickase; and a nickase can occur either naturally or by mutation or modification of a nuclease that makes double stranded cuts. Such an enzyme can be a catalytically-impaired Cas9 endonuclease (a nickase). Such an enzyme can be a Casl2a/b, MAD7, or variant thereof. The nickase is fused to an engineered reverse transcriptase (RT). The nickase is programmed (directed) with a prime-editing guide RNA (pegRNA). The skilled person in the art would appreciate that the pegRNA both specifies the target site and encodes the desired edit. Advantageously the nickase is a catalytically-impaired Cas9 endonuclease, a Cas9 nickase, that is fused to the reverse transcriptase. During genetic editing, the Cas9 nickase part of the protein is guided to the DNA target site by the pegRNA, whereby a nick or single stranded cut occurs. The reverse transcriptase domain then uses the pegRNA to template reverse transcription of the desired edit, directly polymerizing DNA onto the nicked target DNA strand. The edited DNA strand replaces the original DNA strand, creating a heteroduplex containing one edited strand and one unedited strand. Afterward, optionally, the prime editor (PE) guides resolution of the heteroduplex to favor copying the edit onto the unedited strand, completing the process (typically achieved with a nickase gRNA).

[0081] As used herein, “PEI” refers to a PE complex comprising a fusion protein comprising Cas9(H840A) and a wild type MMLV RT having the following N-terminus to C-terminus structure: [NLS]-[Cas9(H840A)]- [linker]-[MMLV_RT(wt)] + a desired atgRNA (or PEgRNA). In various embodiments, the prime editors disclosed herein is comprised of PEI.

[0082] As used herein, “PE2” refers to a PE complex comprising a fusion protein comprising Cas9(H840A) and a variant MMLV RT having the following N-terminus to C-terminus structure:

[NLS]-[Cas9(H840A)]- [linker]-[MMLV_RT(D200N)(T330P)(L603W)(T306K)(W313F)] + a desired atgRNA (or PEgRNA). In various embodiments, the prime editors disclosed herein are comprised of PE2. In various embodiments, the prime editors disclosed herein is comprised of PE2 and co-expression of MMR protein MLHldn, that is PE4.

[0083] As used herein, “PE3” refers to PE2 plus a second-strand nicking guide RNA that complexes with the PE2 and introduces a nick in the non-edited DNA strand. The induction of the second nick increases the chances of the unedited strand, rather than the edited strand, to be repaired. In various embodiments, the prime editors disclosed herein are comprised of PE3. In various embodiments, the prime editors disclosed herein are comprised of PE3 and co-expression of MMR protein MLHldn, that is PE5. [0084] As used herein, “PE3b” refers to PE3 but wherein the second-strand nicking guide RNA is designed for temporal control such that the second strand nick is not introduced until after the installation of the desired edit. This is achieved by designing a gRNA with a spacer sequence with mismatches to the unedited original allele that matches only the edited strand. Using this strategy, mismatches between the protospacer and the unedited allele should disfavor nicking by the sgRNA until after the editing event on the PAM strand takes place.

5.6. Guides for Prime Editing

[0085] Anzalone et al., 2019 (Nature 576: 149) describes prime editing and a prime editing complex using a type II CRISPR and can be used herein. A prime editing complex consists of a type II CRISPR PE protein containing an RNA-guided DNA-nicking domain fused to a reverse transcriptase (RT) domain and complexed with a pegRNA. The pegRNA comprises (5’ to 3’) a spacer that is complementary to the target sequence of a genomic DNA, a nickase (e.g. Cas9) binding site, a reverse transcriptase template including editing positions, and primer binding site (PBS). The PE-pegRNA complex binds the target DNA and the CRISPR protein nicks the PAM-containing strand. The resulting 3' end of the nicked target hybridizes to the primer-binding site (PBS) of the pegRNA, then primes reverse transcription of new DNA containing the desired edit using the RT template of the pegRNA. The overall structure of the pegRNA is like that of a typical type II sgRNA with a reverse transcriptase template/primer binding site appended to the 3’ end. The structure leaves the PBS at the 3’ end of the pegRNA free to bind to the nicked strand complementary to the target which forms the primer for reverse transcription.

[0086] Guide RNAs of CRISPRs differ in overall structure. For example, while the spacer of a type II gRNA is located at the 5’ end, the spacer of a type V gRNA is located towards the 3’ end, with the CRISPR protein (e.g. Casl2a) binding region located toward the 5’ end. Accordingly, the regions of a type V pegRNA are rearranged compared to a type II pegRNA. The overall structure of the pegRNA is like that of a typical type II sgRNA with a reverse transcriptase template/primer binding site appended to the 3’ end. The pegRNA comprises (5’ to 3’) a CRISPR protein-binding region, a spacer which is complementary to the target sequence of a genomic DNA, a reverse transcriptase template including editing positions, and primer binding site (PBS).

[0087] In typical embodiments, an atgRNA comprises a reverse transcriptase template that encodes, partially or in its entirety, an integration recognition site (also referred to as an integration target recognition site) or a recombinase recognition site (also referred to as a recombinase target recognition site). The integration target recognition site, which is to be placed at a desired location in the genome or intracellular nucleic acid, is referred to as a “beacon” site or an “attachment site” or a “landing pad” or “landing site.” An integration target recognition site or recombinase target recognition site incorporated into the pegRNA is referred to as an attachment site containing guide RNA (atgRNA).

5.7. Attachment Site-Containing Guide RNA (atgRNA)

[0088] As used herein, the term “attachment site-containing guide RNA” (atgRNA) and the like refer to an extended single guide RNA (sgRNA) comprising a primer binding site (PBS), a reverse transcriptase (RT) template sequence, and wherein the RT template encodes for an integration recognition site or a recombinase recognition site that can be recognized by a recombinase, integrase, or transposase. In some embodiments, the RT template comprises a clamp sequence and an integration recognition site. As referred to herein an atgRNA may be referred to as a guide RNA. An integration recognition site or recombinase target recognition site incorporated into the pegRNA is referred to as an attachment site containing guide RNA (atgRNA).

[0089] Recombination between a first integration recognition site (e.g., any of the integration recognition sites described herein) and a second recognition site (e.g., any of the integration recognition sites described herein) is mediated by functional symmetry between the two integration recognition sites and the central dinucleotide of each of the two integration recognition sites. As used herein, the term “cognate pair” refers to a first integration recognition site (e.g., any of the integration recognition sites described herein) and a functionally symmetric second integration recognition site (e.g., any of the integration recognition sites described herein) that can be recombined. Each of the first and second integration recognition sites is a “cognate”, or “integration cognate”, or “integration cognate site”, of the other member of the cognate pair. A non-limiting example of a cognate pair are an attB site and an attP site, which are capable of recombination by the large serine integrase BxBl.

[0090] In typical embodiments, an atgRNA comprises a reverse transcriptase template that encodes, partially or in its entirety, an integration recognition site (also referred to as an integration target recognition site) or a recombinase recognition site (also referred to as a recombinase target recognition site). The integration target recognition site, which is to be placed at a desired location in the genome or intracellular nucleic acid, is referred to as a “beacon,” a “beacon” site or an “attachment site” or a “landing pad” or “landing site.” An integration target recognition site or recombinase target recognition site incorporated into the pegRNA is referred to as an attachment site containing guide RNA (atgRNA).

[0091] During genome editing, the primer binding site allows the 3’ end of the nicked DNA strand to hybridize to the atgRNA, while the RT template serves as a template for the synthesis of edited genetic information. The atgRNA is capable for instance, without limitation, of (i) identifying the target nucleotide sequence to be edited and (ii) encoding new genetic information that replaces (or in some cases adds) the targeted sequence. In some embodiments, the atgRNA is capable of (i) identifying the target nucleotide sequence to be edited and (ii) encoding an integration site that replaces (or inserts/deletes within) the targeted sequences.

[0092] In some embodiments, the co-delivery system described herein includes a polynucleotide sequence encoding an attachment site-containing guide RNA (atgRNA) packaged in an LNP. In some embodiments, the co-delivery system described herein includes a vector comprising a polynucleotide sequence encoding an atgRNA. In some embodiments, the atgRNA comprises a domain that is capable of guiding the prime editor fusion protein to a target sequence, thereby identifying the target nucleotide sequence to be edited; and a reverse transcriptase (RT) template that comprises a first integration recognition site. In some embodiments, the atgRNA comprises a domain that is capable of guiding the prime editor fusion protein (or prime editor system) to a target sequence, thereby identifying the target nucleotide sequence to be edited; and a reverse transcriptase (RT) template that comprises at least a portion first integration recognition site.

[0093] In some embodiments, the co-delivery system described herein includes a polynucleotide sequence encoding a first attachment site-containing guide RNA (atgRNA) and a polynucleotide nucleotide sequence encoding a second attachment site-containing guide RNA (atgRNA) packaged into the same LNP. In some embodiments, the co-delivery system described herein includes a polynucleotide sequence encoding a first attachment sitecontaining guide RNA (atgRNA) packaged into a first LNP and a polynucleotide nucleotide sequence encoding a second attachment site-containing guide RNA (atgRNA) packaged into a second LNP. [0094] In some embodiments, the co-delivery system described herein includes a vector comprising a polynucleotide sequence encoding a first attachment site-containing guide RNA (atgRNA), a polynucleotide sequence encoding a second atgRNA, or both.

[0095] In some embodiments, the co-delivery system described herein includes a polynucleotide sequence encoding a first attachment site-containing guide RNA (atgRNA) packaged into a first LNP and a vector comprising a polynucleotide sequence encoding a second atgRNA.

[0096] In some embodiments, where the co-delivery system contains a first atgRNA and a second atgRNA, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, where the at least first pair of atgRNAs have domains that are capable of guiding the gene editor protein or prime editor fusion protein to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site.

[0097] In some embodiments, the first atgRNA’ s reverse transcriptase template encodes for a first single-stranded DNA sequence (i.e., a first DNA flap) that contains a complementary region to a second single-stranded DNA sequence (i.e., a second DNA flap) encoded by a second atgRNA comprising a second reverse transcriptase template. In certain embodiments, the complementary region between the first and second single-stranded DNA sequences is comprised of more than 5 consecutive bases of an integrase target recognition site. In certain embodiments, the complementary region between the first and second single-stranded DNA sequences is comprised of more than 10 consecutive bases of an integrase target recognition site. In certain embodiments, the complementary region between the first and second singlestranded DNA sequences is comprised of more than 20 consecutive bases of an integrase target recognition site. In certain embodiments, the complementary region between the first and second single-stranded DNA sequences is comprised of more than 30 consecutive bases of an integrase target recognition site. Use of two guide RNAs that are (or encode DNA that is) partially complementarity to each other and comprised of consecutive bases of an integrase target recognition site are referred to as dual, paired, annealing, complementary, or twin attachment site-containing guide RNAs (atgRNAs). In certain embodiments, use of two guide RNAs that are (or encode DNA that is) full complementarity to each other and comprised of consecutive bases of an integrase target recognition site are referred to as dual, paired, annealing, complementary, or twin attachment site-containing guide RNAs (atgRNAs).

[0098] In some embodiments, upon introducing the nucleic acid construct into a cell, the first atgRNA incorporates the first integration recognition site into the cell’s genome at the target sequence.

[0099] Table 9 includes atgRNAs, sgRNAs and nicking guides that can be used herein. Spacers are labeled in capital font (SPACER), RT regions in bold capital (RT REGION), AttB sites in bold lower case (attB site), and PBS in capital italics (PBS). Unless otherwise denoted, the AttB is for Bxbl. 5.8. Integrases/Recombinases and Integration/Recombination Sites

[0100] In typical embodiments, the co-delivery system described herein contains an integrase and/or a recombinase. In some embodiments, the co-delivery system includes an integrase and/or a recombinase packaged in a LNP. In one embodiment, the co-delivery system includes a polynucleotide encoding an integrase and/or a recombinase. In some embodiments, the co-delivery system includes an integrase or a recombinase packaged in a vector (e.g., a viral vector). In some embodiments, the co-delivery system includes at least a first integrase (e.g., a first integrase and a second integrase) and/or at least a first recombinase (e.g., a first recombinase and a second recombinase).

[0101] In some embodiments, the integration enzyme (e.g., the integrase or recombinase) is selected from the group consisting of Dre, Vika, Bxbl, <pC31, RDF, cpBTl, Rl, R2, R3, R4, R5, TP901-1, Al 18, cpFCl, cpCl, MR11, TGI, cp370.1, W , BL3, SPBc, K38, Peaches, Veracruz, Rebeuca, Theia, Benedict, KSSJEB, PattyP, Doom, Scowl, Lockley, Switzer, Bob3, Troube, Abrogate, Anglerfish, Sarfire, SkiPole, Conceptll, Museum, Severus, Airmid, Benedict, Hinder, ICleared, Sheen, Mundrea, BxZ2, cpRV, retrotransposases encoded by a Tcl/mariner family member including but not limited to retrotransposases encoded by LI, Tol2, Tel, Tc3, Himar 1 (isolated from the horn fly, Haematobia irrilans). Mosl (Mosaic element of Drosophila maiiriliana). and Minos, and any mutants thereof. As can be used herein, Xu et al describes methods for evaluating integrase activity in E. coli and mammalian cells and confirmed at least R4, cpC31, (pBTl, Bxbl, SPBc, TP901-1 and WP integrases to be active on substrates integrated into the genome of HT1080 cells (Xu et al., 2013, Accuracy and efficiency define Bxbl integrase as the best of fifteen candidate serine recombinases for the integration of DNA into the human genome. BMC Biotechnol. 2013 Oct 20; 13:87. doi: 10.1186/1472-6750-13-87). Durrant describes new large serine recombinases (LSRs) divided into three classes distinguished from one another by efficiency and specificity, including landing pad LSRs which outperform wild-type Bxbl in episomal and chromosomal integration efficiency, LSRs that achieve both efficient and site-specific integration without a landing pad, and multi-targeting LSRs with minimal site-specificity. Additionally, embodiments can include any serine recombinase such as BceINT, SSCINT, SACINT, and INT10 (see lonnidi et al, 2021; Drag-and-drop genome insertion without DNA cleavage with CRISPR directed integrases. bioRxiv 2021.11.01.466786, doi. org/10.1101/2021.11.01.466786). In some embodiments, the integration site can be selected from an attB site, an attP site, an attL site, an attR site, a lox71 site a Vox site, or a FRT site. In instances in this disclosure that refer to a Cre-lox system, the Cre-lox system is referred to either as a control for programmable gene insertion or as a tool for a recombinase- mediated event separate and distinct from insertion of the donor polynucleotide template (or exogenous nucleic acid) into the integrated recognition site.

[0102] It will be appreciated that desired activity of integrases, transposases and the like can depend on nuclear localization. In certain embodiments, prokaryotic enzymes are adapted to modulate nuclear localization. In certain embodiments, eukaryotic or vertebrate enzymes are adapted to modulate nuclear localization. In certain embodiments, the invention provides fusion or hybrid proteins. Such modulation can comprise addition or removal of one or more nuclear localization signal (NLS) and/or addition or removal of one or more nuclear export signal (NES). Xu et al compared derivatives of fourteen serine integrases that either possess or lack a nuclear localization signal (NLS) to conclude that certain integrases benefit from addition of an NLS whereas others are transported efficiently without addition, and a major determinant of activity in yeast and vertebrate cells is avoidance of toxicity. (Xu et al., 2016, Comparison and optimization of ten phage encoded serine integrases for genome engineering in Saccharomyces cerevisiae. BMC Biotechnol. 2016 Feb 9; 16: 13. doi: 10.1186/sl2896-016- 0241-5). Ramakrishnan et al. systematically studied the effect of different NES mutants developed from mariner-like elements (MLEs) on transposase localization and activity and concluded that nuclear export provides a means of controlling transposition activity and maintaining genome integrity. (Ramakrishnan et al. Nuclear export signal (NES) of transposases affects the transposition activity of mariner-like elements Ppmarl and Ppmar2 ofmoso bamboo. Mob DNA. 2019 Aug 19;10:35. doi: 10.1186/sl3100-019-0179-y). The methods and constructs are used to modulate nuclear localization of system components of the invention.

[0103] In typical embodiments, the integrase used herein is selected from below (Table 10).

[0104] Sequences of insertion sites (i.e., recognition target sites) suitable for use in embodiments of the disclosure are presented below (Table 11). FIGs. 14A-14E shows analysis of effect of variant AttP sites on integration efficiency.

5.9. Co-delivery of gene editor and donor DNA template

[0105] This disclosure features methods of delivering (e.g., co-delivery or dual delivery) a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, where the methods includes delivering to a (i) gene editor construct and a (ii) template polynucleotide, and (iii) at least a first attachment site-containing guide (atgRNA).

[0106] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes: delivering a lipid nanoparticle (LNP) comprising a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and a vector comprising a template polynucleotide and at least a first attachment site-containing guide RNA (atgRNA). In some embodiments, the first atgRNA comprises (i) a domain that is capable of guiding the prime editor system to a target sequence; and (ii) a reverse transcriptase (RT) template that comprises at least a portion of a first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the RT template comprises the entirety of the first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the vector also includes a sequence encoding a nicking guide RNA (ngRNA).

[0107] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes: delivering a lipid nanoparticle (LNP) comprising a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and a vector comprising a template polynucleotide and a first attachment site-containing guide RNA (atgRNA) and a second attachment sitecontaining guide RNA (atgRNA). In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the at least first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap (e.g., 6bp of complementarity).

[0108] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes: delivering into a cell a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct), and (ii) a first attachment sitecontaining guide RNA (atgRNA); and a vector comprising: (i) a template polynucleotide, and (ii) a second atgRNA. In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the a first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap (e.g., 6bp of complementarity).

[0109] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes delivering: a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct), (ii) a first attachment site-containing guide RNA (atgRNA), and (iii) a second atgRNA; and a vector comprising (i) a template polynucleotide. In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the at least first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the at least first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap (e.g., 6bp of complementarity).

[0110] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes delivering: a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and (ii) a first attachment site-containing guide RNA (atgRNA); and a vector comprising: (i) a template polynucleotide, and (ii) a nicking atgRNA. In some embodiments, the first atgRNA comprises (i) a domain that is capable of guiding the prime editor system to a target sequence; and (ii) a reverse transcriptase (RT) template that comprises at least a portion of a first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the RT template comprises the entirety of the first integration recognition site.

[OHl] In some embodiments, where the method includes delivering an LNP and a first vector, the LNP and the first vector are delivered at least 1 day, at least 2 days, at least 3 days, at least 4 days, at least 5 days, at least 6 days, at least 7 days, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, or at least 8 weeks apart. In some embodiments, where the method includes delivering an LNP and a second vector, the LNP and the second vector are delivered a different times on the same day, at least 1 day, at least 2 days, at least 3 days, at least 4 days, at least 5 days, at least 6 days, at least 7 days, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, or 8 weeks apart. In some embodiments, the LNP and the first vector are delivered about 6 weeks apart.

[0112] This disclosure also features a method for delivering a system capable of site- specifically integrating a template polynucleotide into the genome of a cell, where the method includes delivering the system in vivo. In a non-limiting example, the system is delivered to a fetus or a neonate to site-specifically integrate in vivo a template polynucleotide into the genome of a cell. Delivering the system to a fetus or a neonate provides advantages over delivering the system later in life (e.g., after the neonate phase ends), including: (i) fewer number of cells that need to be treated (e.g., in the adult, there are trillions of cells, but in a fetus, there are significantly fewer cells); (ii) developmental benefits: the early stage of development of a fetus or a neonate means that if a genetic disease is treated successfully, the individual could potentially develop normally, with significant reduction or even complete removal of any of the disease manifestations; (iii) preventing disease progression: in certain genetic conditions the physiological damage is irreversible damage and in some instances is exacerbated as the disease progresses, therefore, intervening at the fetal (or neonate) stage, it is possible to prevent or reduce the progression of the disease and potentially prevent any irreversible damage from occurring; (iv) higher cell turnover and cell division rate: in a developing fetus, cells are dividing rapidly as the fetus (or neonate) grows, which means that if programmable gene insertion is achieved in the fetus (or neonate) is introduced, it could be propagated more rapidly throughout the body than in an older child or adult; and (v) immune tolerance: for example, there is evidence to suggest that performing gene therapy (e.g., programmable gene insertion) early in development might result in immune tolerance to the vector, thereby reducing the risk of an immune response against the system.

[0113] In some embodiments, the method includes delivering an LNP and a first vector, the LNP and the first vector are delivered to a cell in vivo. In some embodiments, the in vivo cells are present in a fetus or a neonate. In some embodiments, the LNP is delivered between age 0 (day of birth) and age 7 days and the vector is delivered between age 5 weeks and age 7 weeks. In some embodiments, the LNP is delivered at about at 2 days and the vector is delivered at about age 6 weeks. [0114] In some embodiments, where the method includes delivering an LNP and a first vector, the LNP and the first vector are delivered to a cell in vivo, the LNP can be delivered to a fetus at a first time point and the vector is delivered to the fetus after the fetus is born (referred to after birth as a neonate). In some embodiments, the LNP is delivered to a fetus and the vector is delivered to the fetus after birth (i.e., at the neonate stage) at any point between birth and up to age 8 weeks. In some embodiments, the LNP is delivered to the fetus and the vector is delivered at about age 1 week, 2 weeks, 3 weeks, 4 weeks, 5 weeks, 6 weeks, 7 weeks or 8 weeks.

[0115] In some embodiments, where the method includes delivering an LNP and a first vector, the LNP and the first vector are delivered to a cell in vivo, the LNP can be delivered to a fetus at a first time point and the vector is delivered to the fetus (child) after the fetus (child) is bom, for example, when the child is age 90 days or older (e.g., age 6 months, age 9 months, age 1 year, age 2 years, age 3 years, age 4 years, age 5 years, age 6 years, or older).

[0116] As used herein, the term “fetus” refers to an unborn offspring. As used herein, the term “neonate” refers to a newborn infant, which includes the first 90 days of life.

[0117] This disclosure also features a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, the system comprising: a lipid nanoparticle (LNP) comprising a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and a vector comprising a template polynucleotide and at least a first attachment sitecontaining guide RNA (atgRNA). In some embodiments, the first atgRNA comprises (i) a domain that is capable of guiding the prime editor system to a target sequence; and (ii) a reverse transcriptase (RT) template that comprises at least a portion of a first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the RT template comprises the entirety of the first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the vector also includes a sequence encoding a nicking guide RNA (ngRNA).

[0118] This disclosure also features a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, the system comprising: a lipid nanoparticle (LNP) comprising a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and a vector comprising a template polynucleotide and a first attachment site-containing guide RNA (atgRNA) and a second attachment site-containing guide RNA (atgRNA). In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the a first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap.

[0119] This disclosure also features a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, the system comprising: a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct), and (ii) a first attachment site-containing guide RNA (atgRNA); and a vector comprising: (i) a template polynucleotide, and (ii) a second atgRNA. In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the a first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap.

[0120] This disclosure also features a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, the system comprising: co-delivering: a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct), (ii) a first attachment site-containing guide RNA (atgRNA), and (iii) a second atgRNA; and a vector comprising (i) a template polynucleotide. In some embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the a first integration recognition site; and the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap.

[0121] This disclosure also features a system capable of site-specifically integrating a template polynucleotide into the genome of a cell, the system comprising: a lipid nanoparticle (LNP) comprising: (i) a gene editor polynucleotide (e.g., a gene editor polynucleotide construct) and (ii) a first attachment site-containing guide RNA (atgRNA); and a vector comprising: (i) a template polynucleotide, and (ii) a nicking atgRNA. In some embodiments, the first atgRNA comprises (i) a domain that is capable of guiding the prime editor system to a target sequence; and (ii) a reverse transcriptase (RT) template that comprises at least a portion of a first integration recognition site. In some embodiments, where the vector comprises a polynucleotide encoding a first atgRNA, the RT template comprises the entirety of the first integration recognition site.

[0122] In typical embodiments, the LNP comprising a gene editor polynucleotide construct is capable delivering to a cell cytoplasm the gene editor polynucleotide construct. In some embodiments, the LNP comprising a gene editor polynucleotide construct is capable delivering to a cell nucleus the gene editor polynucleotide construct. In some embodiments, the LNP comprises a gene editor protein and associated guide nucleic acids. In some embodiments, the LNP comprises a gene editor protein and associated guide nucleic acids that are capable of localizing to cell nucleus.

[0123] In some embodiments, a gene editor polynucleotide construct is delivered to a cell by a fusosome. In some embodiments, a gene editor polynucleotide construct is delivered to a cell cytoplasm by a fusosome. In some embodiments, the fusosome comprises a gene editor protein and associated guide nucleic acids.

[0124] In some embodiments, a gene editor polynucleotide construct is delivered to a cell by an exosome. In some embodiments, a gene editor polynucleotide construct is delivered to a cell cytoplasm by an exosome. In some embodiments, the exosome comprises a gene editor protein and associated guide nucleic acids.

[0125] In some embodiments, the prime editor or Gene Writer protein fusion, either of which may have a fused/linked integrase, is incorporated (i.e., packaged) into LNP as protein. Further, associated atgRNA and optional ngRNAs may be co-packaged with gene editor proteins in LNP.

[0126] In some embodiments, the gene editor polynucleotide construct comprises (a) a polynucleotide sequence encoding a prime editor fusion protein or a Gene Writer™ protein,

(b) a polynucleotide sequence encoding an attachment site-containing guide RNA (atgRNA),

(c) optionally, a polynucleotide sequence encoding a nickase guide RNA (ngRNA), (d) a polynucleotide sequence encoding an integrase, (e) and optionally, a polynucleotide sequence encoding a recombinase.

[0127] In some embodiments, the prime editor or Gene Writer protein fusion, either of which may have a fused/linked integrase, is expressed as a split construct. In typical embodiments, the split construct in reconstituted in a cell. In some embodiments, the split construct can be fused or ligated via intein protein splicing. In some embodiments, the split construct can be reconstituted via protein-protein inter-molecular bonding and/or interactions. In some embodiments, the split construct can be reconstituted via chemical, biological, or environmental induced oligomerization. In certain embodiments, the split construct can be adapted into one or more nucleic acid constructs described herein.

5.9.1. Gene Editor Polynucleotide

[0128] In some embodiments, the systems described include a gene editor polynucleotide that is delivered to a cell using the methods described herein. In some embodiments, the gene editor polynucleotide is delivered as a polynucleotide (e.g., an mRNA). In some embodiments, the gene editor polynucleotide is delivered as a protein. In some embodiments, the gene editor polynucleotide or protein is packaged, and thereby vectorized, within a lipid nanoparticle (LNP). In some embodiments, the gene editor polynucleotide or protein is packaged in a LNP and is co-delivered with a template polynucleotide (i.e., nucleic acid “cargo” or nucleic acid “payload”) packaged into a separate vector (e.g., a viral vector (e.g., an AAV or adenovirus)) or a second lipid nanoparticle (LNP).

[0129] In some embodiments, the gene editor polynucleotide is delivered to the cells as a polynucleotide. For example, the gene editor polynucleotide is delivered to the cells as an mRNA encoding the gene editor polynucleotide (e.g., the gene editor protein or the prime editor system). In some embodiments, the mRNA comprises one or more modified uridines. In some embodiments, the mRNA comprises a sequence where each of the uridines is a modified uridine. In some embodiments, the mRNA is uridine depleted. In some embodiments, the mRNA encoding the nickase comprises one or more modified uridines. In some embodiments, the mRNA encoding the reverse transcriptase comprises one or more modified uridines. In some embodiments, the mRNA encoding the nickase comprises one or more modified uridines, and the mRNA encoding the reverse transcriptase comprises one or more modified uridines. In some embodiments, where the integrase is encoded in an mRNA, the mRNA comprises modified uridines. In some embodiments, a modified uridine is a Nl- Methylpseudouridine-5’ -Triphosphate. In some embodiments, a modified uridine is a pseudouridine. In some embodiments, the mRNA comprises a 5’ cap. In some embodiments, the 5’ cap comprises a molecular formula of C32H43N15O24P4 (free acid).

[0130] In some embodiments, the gene editor polynucleotide (e.g., a gene editor polynucleotide construct) comprises a polynucleotide sequence encoding a primer editor system (e.g., any of the prime editor systems described herein). In some embodiments, the prime editor system comprises a nucleotide sequence encoding a nickase (e.g., any of the Cas proteins or variants thereof (e.g., nickases) and nickases described herein, see Tables 4-8) and a nucleotide sequence encoding a reverse transcriptase (e.g., any of the reverse transcriptases described herein). In some embodiments, the nucleotide sequence encoding the nickase and the nucleotide sequence encoding the reverse transcriptase are positioned in the construct such that when expressed the nickase is linked to the reverse transcriptase. In some embodiments, the nickase is linked to the reverse transcriptase by in-frame fusion. In some embodiments, the nickase is linked to the reverse transcriptase by a linker. In some embodiments, the linker is a peptide fused in-frame between the nickase and reverse transcriptase.

[0131] In some embodiments, the gene editor polynucleotide (e.g., a gene editor polynucleotide construct) further comprises a polynucleotide sequence encoding at least a first integrase (e.g., any of the integrases described herein, e.g., as described in Table 10 and also in Yarnall et al., Nat. Biotechnol., 2022, doi.org/10.1038/s41587-022-01527-4 and Durrant et al., Nat. Biotechnol., 2022, doi.org/10.1038/s41587-022-01494-w, each of which are herein incorporated by reference in their entireties). In some embodiments, the linked nickase-reverse transcriptase are further linked to the first integrase. [0132] In some embodiments, the gene editor polynucleotide construct further comprises a polynucleotide sequence encoding at least a first recombinase (e.g., any of the recombinases described herein).

5.9.2. Vector

[0133] In some embodiments, the systems and methods described herein include a vector that is capable of co-delivering a template polynucleotide, one or more attachment site-containing gRNA, one or more integrases, one or more recombinases, a gene editor polynucleotide, one or more integration recognition sites, one or more recombinase recognition sites, or a combination thereof.

[0134] Non-limiting examples of vectors that can be used in the methods or systems described herein include the vectors described in FIGs. 3-6.

5.9.2.1 AtgRNA and/or ngRNA

[0135] In some embodiments, the vector includes a polynucleotide sequence encoding an attachment site-containing guide RNA (atgRNA). In such embodiments, the polynucleotide sequence encoding the attachment site-containing guide RNA (atgRNA) is operably linked to a regulatory element (e.g., a U6 promoter) that is capable of driving expression of the atgRNA. In such embodiments, the atgRNA comprises (i) a domain that is capable of guiding the prime editor system to a target sequence; and (ii) a reverse transcriptase (RT) template that comprises at least a portion of a first integration recognition site. In some embodiments, where the system, and thereby the vector, include a polynucleotide encoding only a first atgRNA, the RT template comprises the entirety of the first integration recognition site. In such embodiments, the vector or the LNP includes a polynucleotide sequence encoding a nicking gRNA.

[0136] In some embodiments, the vector includes a polynucleotide sequence encoding a first attachment site-containing guide RNA (atgRNA) and a polynucleotide sequence encoding a second attachment site-containing guide RNA (atgRNA). In such embodiments, the first atgRNA and the second atgRNA are an at least first pair of atgRNAs, wherein the at least first pair of atgRNAs have domains that are capable of guiding the prime editor system to a target sequence, the first atgRNA further includes a first RT template that comprises at least a portion of the a first integration recognition site; the second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site, and the first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap.

5.9.2.1 Template Polynucleotide

[0137] In typical embodiments, the vector includes a template polynucleotide and a sequence that is an integration cognate of an integration recognition site site-specifically incorporated into the genome of a cell. For example, the vector includes a template polynucleotide and a second integration recognition site that is a cognate pair with the first integration recognition site site-specifically incorporated into the genome of the cell. In such embodiments, the sequence that is an integration cognate (e.g., a second integration recognition site) enables integration of the template polynucleotide or portion thereof when contacted with an integrase and the site-specifically incorporated first integration recognition site.

[0138] In typical embodiments, the vector comprising a template polynucleotide is a recombinant adenovirus, a helper dependent adenovirus, an AAV, a lentivirus, an HSV, an annelovirus, a retrovirus, a Doggybone™ DNA (dbDNA), a minicircle, a plasmid, a miniDNA, an exosome, a fusosome, or an nanoplasmid. In preferred embodiments, the vector is capable of localizing to the nucleus.

[0139] In certain embodiments, the template polynucleotide is delivered to the cytoplasm and localizes to the nucleus. In certain embodiments, the template polynucleotide is delivered to the cytoplasm by LNP. In certain embodiments, the donor template polynucleotide construct comprises a recognition sequence that is recognized by a DNA binding protein (DNA binding domain) or a transcription factor binding domain. In certain embodiments, the donor template polynucleotide construct is delivered to the nucleus by an integrase or recombinase.

[0140] In certain embodiments, the template polynucleotide is delivered to the mitochondria. In certain embodiments, the donor template polynucleotide construct comprises a mitochondria targeting sequence.

[0141] In certain embodiments, the vector comprising a template polynucleotide is AAV. In some embodiments, the AAV contains a 5’ inverted terminal repeat (ITR). In some embodiments, the AAV contains a 3’ inverted terminal repeat (ITR). In some embodiments, the AAV contains a 5’ and a 3’ ITR. In some embodiments, the 5’ and 3’ ITR are not derived from the same serotype of virus. In some embodiments, the ITRs are derived from adenovirus, AAV2, and/or AAV5.

[0142] In certain embodiments, the vector comprising a template polynucleotide is single stranded AAV (ssAAV). In certain embodiments, the vector comprising a donor template polynucleotide construct is self-complementary AAV (scAAV).

[0143] In some embodiments, a vector comprises an attachment site-containing guideRNA (atgRNA), a nicking-guideRNA (ngRNA), and template polynucleotide. In typical embodiments, the vector comprising an attachment site-containing guideRNA (atgRNA), a nicking-guideRNA (ngRNA), and template polynucleotide is recombinant adenovirus, helper dependent adenovirus, AAV, lentivirus, HSV, annelovirus, retrovirus, Doggybone™ DNA (dbDNA), minicircle, plasmid, miniDNA, exosome, fusosome, or nanoplasmid. In preferred embodiments, the vector is capable of localizing to the nucleus. In typical embodiments, the attachment site-containing guideRNA (atgRNA) sequence and the nicking-guideRNA (ngRNA) sequence contain a terminal poly dT.

[0144] In some embodiments, a vector comprises an attachment site-containing guideRNA (atgRNA), and donor template. In typical embodiments, the vector comprising an attachment site-containing guideRNA (atgRNA) and donor template is recombinant adenovirus, helper dependent adenovirus, AAV, lentivirus, HSV, annelovirus, retrovirus, Doggybone™ DNA (dbDNA), minicircle, plasmid, miniDNA, exosome, fusosome, or nanoplasmid. In preferred embodiments, the vector is capable of localizing to the nucleus. In typical embodiments, the attachment site-containing guideRNA (atgRNA) sequence contain a terminal poly dT.

[0145] In typical embodiments, the template polynucleotide is capable of being integrated into a genomic locus that contains an integrase target recognition site or a recombinase target recognition site.

[0146] In certain embodiments, the template polynucleotide comprises at least one of the following: a gene, a gene fragment, an expression cassette, a logic gate system, or any combination thereof. In some embodiments, the template polynucleotide comprises at least one intron or exon.

[0147] In typical embodiments, the template polynucleotide further comprises at least one integrase target recognition site or a recombinase target integrase site. In certain embodiments, at least one integrase target recognition site or a recombinase target integrase site is placed within the donor template vector inverted terminal repeat.

5.9.1.3 Integrase- or recombinase-mediated self-circularization of a subsequence of a vector delivered as part of the co-delivery system

[0148] In some embodiments, the delivery system (e.g., co-delivery system) includes a vector having a sub-sequence that is capable of self-circularizing to form a self-circular nucleic acid. In some embodiments, the vector comprises a physical portion or region of the vector that is capable of self-circularizing to form a circular construct. As used herein, the term “subsequence” refers to a portion of the vector that is capable of self-circularizing, where the subsequence is flanked by integration recognition sites or recombinase recognition sites positioned to enable self-circularization. As used herein, the term “self-circular nucleic acid” refers to a double-stranded, circular nucleic acid construct produced as a result of recombination of a cognate pair of integrase or recombinase recognition sites present on the vector. Recombination occurs when the vector is contacted with an integrase or a recombinase under conditions that allow for recombination of the cognate pair of integrase or recombinase recognition sites.

[0149] In some embodiments, the sub-sequence of the vector includes a first recombinase recognition site and a second recombinase recognition site, wherein the first and second recombinase recognition sites are capable of being recombined by a recombinase. In some embodiments, the sub-sequence of the vector includes a first recombinase recognition site, a second recombinase recognition site, and a second integration recognition site (e.g., the second integration recognition site is a cognate pair of the first integration recognition site), where the first and second recombinase recognition sites flank the integration recognition site. In such cases, the first recombinase recognition site, the second recombinase recognition, and a recombinase enable the self-circularizing and formation of the circular construct.

[0150] In some embodiments, the sub-sequence of the vector includes a third integration recognition site and a fourth integration recognition site, wherein the third and fourth integration recognition sites are a cognate pair. In some embodiments, the subsequence of the vector includes the second integration recognition site, the third integration recognition site, the fourth integration recognition site, where the third and fourth integration recognition sites flank the second integration recognition site (where the second integration recognition site is a cognate pair of the first integration recognition site). In such cases, the third integration recognition site, the fourth integration recognition site, and an integrase enable self - circularization and formation of the circular construct. In such cases, the third integration recognition site and/or the fourth integration recognition sites cannot recombine with the first integration recognition site and/or the second integration recognition site due, in part, to having different central dinucleotides than the first and second integration recognition sites.

[0151] In some embodiments where the subsequence includes three or more integration recognition sites, each integration recognition site or each pair of integration recognition is capable of being recognized by a different integrase. In some embodiments where the subsequence includes three or more integration recognition sites, each integration recognition site or each pair of integration recognition comprises a different central dinucleotide.

[0152] In some embodiments, self-circularizing is mediated at the integration recognition sites or recombinase recognition sites. In some embodiments, the self-circularizing is mediated by an integrase or a recombinase.

[0153] In some embodiments, upon introducing the vector into a cell and after selfcircularizing to form the self-circular nucleic acid, the self-circular nucleic acid comprising the second integration recognition site is capable of being integrated into the cell’s genome at the target sequence that contains the first integration recognition site.

[0154] In some embodiments, following self-circularization, the self-circular nucleic acid comprises one or more additional integration recognition sites that enable integration of an additional nucleic acid cargo. In such cases, the additional nucleic acid cargo includes a sequence that is a cognate pair with one or more of the additional integration recognition sites in the self-circular nucleic acid. For example, integration of the self-circular nucleic acid into the genome of a cell results in integration of the one or more additional integration recognition sites into the genome along with the nucleic acid cargo. The integrated one or more additional integration recognition sites serve as an integration recognition site (beacon) for placing the additional nucleic acid cargo. Upon contacting the cell harboring the integrated nucleic acid cargo and the one or more additional integration recognition sites with an integrase and the second additional nucleic acid cargo that includes a sequence that is an integration cognate to the one or more additional integration recognition sites the additional nucleic acid cargo is integrated into the cell’s genome. [0155] In typical embodiments, the self-circularized nucleic acid comprises a DNA cargo, embodiments, the DNA cargo is a gene or gene fragment. In some embodiments the DNA cargo is an expression cassette. In some embodiments, the DNA cargo is a logic gate or logic gate system. The logic gate or logic gate system may be DNA based, RNA based, protein based, or a mix of DNA, RNA, and protein. In some embodiments, the nucleic acid cargo is a genetic, protein, or peptide tag and/or barcode.

5.9.1.4 A second vector

[0156] In some embodiments, the system or methods described herein include a second vector. In some embodiments, where the gene editor polynucleotide encodes a prime editor system comprising a nickase (e.g., any of the Cas proteins or variants thereof (e.g., nickases) and nickases described herein, see Tables 4-8) and a reverse transcriptase (e.g., any of the reverse transcriptase described herein), the second vector comprises a polynucleotide sequence encoding an integrase (e.g., any of the integrases described herein, e.g., as described in Table 10 and also in Yarnall et al., Nat. Biotechnol., 2022, doi.org/10.1038/s41587-022- 01527-4 and Durrant et al., Nat. Biotechnol., 2022, doi.org/10.1038/s41587-022-01494-w, each of which are herein incorporated by reference in their entireties).

[0157] In some embodiments, where the gene editor polynucleotide encodes a prime editor system comprising a nickase and a reverse transcriptase, the second vector comprises a polynucleotide sequence encoding at least a first recombinase. In some embodiments, where the gene editor polynucleotide encodes a prime editor system comprising a nickase, a reverse transcriptase, and an integrase, the second vector comprises a polynucleotide sequence encoding at least a first recombinase. In some embodiments, where the gene editor polynucleotide encodes a prime editor system comprising a nickase, a reverse transcriptase, and an integrase, the second vector comprises a polynucleotide sequence encoding at least a second integrase.

[0158] In some embodiments, the second vector includes a template polynucleotide and a sequence that is an integration cognate of an integration recognition site site-specifically incorporated into the genome of a cell. For example, the second vector includes a template polynucleotide and a second integration recognition site that is a cognate pair with the first integration recognition site site-specifically incorporated into the genome of the cell. In such embodiments, the sequence that is an integration cognate (e.g., a second integration recognition site) enables integration of the template polynucleotide or portion thereof when contacted with an integrase and the site-specifically incorporated first integration recognition site.

[0159] In some embodiments, the second vector is a vector selected from: adenovirus, AAV, lentivirus, HSV, annelovirus, retrovirus, Doggybone™ DNA (dbDNA), minicircle, plasmid, miniDNA, exosome, fusosome, or nanoplasmid.

[0160] In some embodiments, the polynucleotide sequence encoding the prime editor system is encoded on at least two different vectors. In one embodiment, a first vector comprises a polynucleotide sequence encoding a nickase and a second vector comprises a polynucleotide sequence encoding a reverse transcriptase. In such cases, the first vector and second are delivered concurrently.

[0161] In some embodiments, the polynucleotide sequence(s) encoding the prime editor system is encoded on at least two (non-contiguous) polynucleotide sequences. In one embodiment, a first polynucleotide sequence encodes a nickase and a second polynucleotide sequence encodes a reverse transcriptase. In such cases, the first vector and second are delivered concurrently (e.g., in a first LNP).

5.9.3. Split Lipid Nanoparticles (LNPs)

[0162] Also provided herein are methods of co-delivering a system capable of site- specifically integrating at least a first integration recognition site into the genome of a cell, where the method includes delivering to a cell a mixture of a first LNP and a second LNP (“split LNPs”). In one embodiment, the method includes co-delivering to a cell a first gene editor polynucleotide construct and a first attachment site-containing guide RNA (atgRNA) are packaged, and thereby vectorized, within the first LNP, and a second gene editor polynucleotide construct and a second attachment site containing guide RNR (atgRNA) are packaged, and thereby vectorized, within the second LNP, where the first atgRNA and the second atgRNA are an at least first pair of atgRNA. The at least first pair of atgRNAs comprise domains that are capable of guiding the prime editor system to a target sequence. The first atgRNA further includes a first RT template that comprises at least a portion of a first integration recognition site. The second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site. The first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap. [0163] In some embodiments, where the method includes delivering a first LNP (e.g., a first LNP comprising a first gene editor polynucleotide construct and a first atgRNA) and a second LNP (e.g., a second LNP comprising a second gene editor polynucleotide construct and a second atgRNA), the first LNP and the second LNP are mixed prior to delivering to a cell. In some embodiments, the first LNP and the second LNP are mixed at a ratio of first LNP to second LNP of 1:10, 1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1,2:1,3:1,4:1,5:1, 6:1, 7:1, 8:1, 9:1, or 10:1. In some embodiments, the first LNP and the second LNP are mixed at a ratio of 1 : 1.

[0164] In some embodiments, a first LNP comprising a first gene editor polynucleotide construct and a first attachment site-containing guide RNA (atgRNAl) comprises a ratio of ratio of gene editor polynucleotide construct (e.g., mRNA) to atgRNAl of 1 : 10, 1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1, 9:1, or 10:1. In some embodiments, the first LNP comprises a ratio of mRNA to atgRNAl of 2: 1.

[0165] In some embodiments, a second LNP comprising a second gene editor polynucleotide construct and a second attachment site-containing guide RNA (atgRNA2) comprises a ratio of gene editor polynucleotide construct (e.g., mRNA) to atgRNA2 of 1 : 10, 1:9, 1:8, 1 :7, 1 :6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1, 9:1, or 10:1. In some embodiments, the second LNP comprises a ratio of mRNA to atgRNA2 of 2: 1.

[0166] In some embodiments, where the method includes delivering a first LNP (e.g., a first LNP comprising a first gene editor polynucleotide construct and a first atgRNA) and a second LNP (e.g., a second LNP comprising a second gene editor polynucleotide construct and a second atgRNA), the first LNP and the second LNP are mixed such that the ratio of gene editor polynucleotide construct (e.g., mRNA) to first atgRNA (atgRNAl) to second atgRNA (atgRNA2) is 1:0.25:0.25, l:0.5:0.5, 1:0.75:0.75, or 1:1:1.

[0167] In some embodiments, the method of co-delivering to a cell a mixture of LNPs includes co-delivering three or more LNPs, four or more LNPs, five or more LNPs, six or more LNPs, seven or more LNPs, eight or more LNPs, nine or more LNPs, or ten or more LNPs.

[0168] Also provided herein is a system capable of site-specifically integrating at least a first integration recognition site into the genome of a cell, the system comprising: a first gene editor polynucleotide construct and a first attachment site-containing guide RNA (atgRNA) are packaged, and thereby vectorized, within the first LNP, and a second gene editor polynucleotide construct and a second attachment site containing guide RNR (atgRNA) are packaged, and thereby vectorized, within the second LNP, where the first atgRNA and the second atgRNA are an at least first pair of atgRNA. The at least first pair of atgRNAs comprise domains that are capable of guiding the prime editor system to a target sequence. The first atgRNA further includes a first RT template that comprises at least a portion of a first integration recognition site. The second atgRNA further includes a second RT template that comprises at least a portion of the first integration recognition site. The first atgRNA and the second atgRNAs collectively encode the entirety of the first integration recognition site. In such embodiments, the first atgRNA and second atgRNA include at least a 6bp overlap.

[0169] In some embodiments, the system comprises a first LNP (e.g., any of the first LNPs described herein) and a second LNP (e.g., any of the second LNPs described herein) at a ratio of first LNP to second LNP of 1:10, 1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1, 9:1, or 10:1. In some embodiments, the system comprise the first LNP and the second LNP at a ratio of 1 : 1.

[0170] In some embodiments, the system comprises a first LNP having a ratio of a first gene editor polynucleotide construct to a first attachment site-containing guide RNA (atgRNAl) of 1:10, 1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1, 2:1, 3:1, 4:1, 5:1, 6:1, 7:1, 8:1, 9: 1, or 10: 1. In some embodiments, the system includes a first LNP having a ratio of mRNA (i.e., mRNA encoding the gene editor protein) to atgRNAl of 2: 1.

[0171] In some embodiments, the system comprise a second LNP having a ratio of a second gene editor polynucleotide construct to a second attachment site-containing guide RNA (atgRNA2) of 1:10, 1:9, 1:8, 1:7, 1:6, 1:5, 1:4, 1:3, 1:2, 1:1, 1:0.75, 0.75:1,2:1,3:1,4:1,5:1, 6:1, 7:1, 8:1, 9:1, or 10:1. In some embodiments, the system includes a second LNP having a ratio of mRNA (i.e., mRNA encoding the gene editor protein) to atgRNA2 of 2: 1.

[0172] In some embodiments, the system comprises a ratio of gene editor polynucleotide construct (e.g., mRNA encoding the gene editor protein) to first atgRNA (atgRNAl) to second atgRNA (atgRNA2) of 1:0.25:0.25, l:0.5:0.5, 1:0.75:0.75, or 1:1:1.

[0173] In some embodiments, the system comprises a mixture of LNPs comprising three or more LNPs, four or more LNPs, five or more LNPs, six or more LNPs, seven or more LNPs, eight or more LNPs, nine or more LNPs, or ten or more LNPs. [0174] In some embodiments, where a split LNP (e.g., a mixture of two LNPs packaged with different cargo) is being used to site-specifically integrate the at least first integration recognition site into the genome, a vector comprising a template polynucleotide and a sequence that is an integration cognate (i.e., cognate to an integration recognition site site- specifically incorporated into the genome of a cell) can be delivered to the cell concurrently with the split LNPs or after delivery of the split LNPs. For example, after delivering the split LNPs to the cell, a vector that includes a template polynucleotide and a second integration recognition site that is a cognate pair with the first integration recognition site is delivered to the cell. In such embodiments, the sequence that is an integration cognate (e.g., a second integration recognition site) enables integration of the template polynucleotide or portion thereof when contacted with an integrase and the site-specifically incorporated first integration recognition site.

5.9.4. Vector Delivery of a template polynucleotide

[0175] In certain aspects the invention involves vectors, e.g. for delivering or introducing in a cell, but also for propagating these components (e.g. in prokaryotic cells). A used herein, a "vector" is a tool that allows or facilitates the transfer of an entity from one environment to another. It is a replicon, such as a plasmid, phage, or cosmid, into which another DNA segment may be inserted so as to bring about the replication of the inserted segment. Generally, a vector is capable of replication when associated with the proper control elements. In general, the term "vector" refers to a nucleic acid molecule capable of transporting another nucleic acid to which it has been linked. Vectors include, but are not limited to, nucleic acid molecules that are single-stranded, double-stranded, or partially double-stranded; nucleic acid molecules that comprise one or more free ends, no free ends (e.g. circular); nucleic acid molecules that comprise DNA, RNA, or both; and other varieties of polynucleotides known in the art. One type of vector is a ’’plasmid,” which refers to a circular double stranded DNA loop into which additional DNA segments can be inserted, such as by standard molecular cloning techniques. Another type of vector is a viral vector, wherein virally-derived DNA or RNA sequences are present in the vector for packaging into a virus (e.g. retroviruses, replication defective retroviruses, adenoviruses, replication defective adenoviruses, and adeno-associated viruses (AAVs)). Viral vectors also include polynucleotides carried by a virus for transfection into a host cell. Certain vectors are capable of autonomous replication in a host cell into which they are introduced (e.g. bacterial vectors having a bacterial origin of replication and episomal mammalian vectors). Other vectors (e.g., non-episomal mammalian vectors) are integrated into the genome of a host cell upon introduction into the host cell, and thereby are replicated along with the host genome. Moreover, certain vectors are capable of directing the expression of genes to which they are operatively-linked. Such vectors are referred to herein as "expression vectors." Vectors for and that result in expression in a eukaryotic cell can be referred to herein as "eukaryotic expression vectors." Common expression vectors of utility in recombinant DNA techniques are often in the form of plasmids.

[0176] Recombinant expression vectors can comprise a nucleic acid of the invention in a form suitable for expression of the nucleic acid in a host cell, which means that the recombinant expression vectors include one or more regulatory elements, which may be selected on the basis of the host cells to be used for expression, that is operatively-linked to the nucleic acid sequence to be expressed. Within a recombinant expression vector, "operably linked" is intended to mean that the nucleotide sequence of interest is linked to the regulatory element(s) in a manner that allows for expression of the nucleotide sequence (e.g. in an in vitro transcription/translation system or in a host cell when the vector is introduced into the host cell). With regards to recombination and cloning methods, mention is made of U.S. patent application Ser. No. 10/815,730, published Sep. 2, 2004 as US 2004-0171156 Al, the contents of which are herein incorporated by reference in their entirety.

[0177] Vector delivery, e.g., plasmid, viral delivery: The CRISPR enzyme, for instance a Type V protein such as C2cl or C2c3, and/or any of the present RNAs, for instance a guide RNA, can be delivered using any suitable vector, e.g., plasmid or viral vectors, such as adeno associated virus (AAV), lentivirus, adenovirus or other viral vector types, or combinations thereof. Effector proteins and one or more guide RNAs can be packaged into one or more vectors, e.g., plasmid or viral vectors. In some embodiments, the vector, e.g., plasmid or viral vector is delivered to the tissue of interest by, for example, an intramuscular injection, while other times the delivery is via intravenous, transdermal, intranasal, oral, mucosal, or other delivery methods. Such delivery may be either via a single dose, or multiple doses. One skilled in the art understands that the actual dosage to be delivered herein may vary greatly depending upon a variety of factors, such as the vector choice, the target cell, organism, or tissue, the general condition of the subject to be treated, the degree of transformation/modification sought, the administration route, the administration mode, the type of transformation/modification sought, etc. [0178] Such a dosage may further contain, for example, a carrier (water, saline, ethanol, glycerol, lactose, sucrose, calcium phosphate, gelatin, dextran, agar, pectin, peanut oil, sesame oil, etc.), a diluent, a pharmaceutically-acceptable carrier (e.g., phosphate-buffered saline), a pharmaceutically-acceptable excipient, and/or other compounds known in the art. The dosage may further contain one or more pharmaceutically acceptable salts such as, for example, a mineral acid salt such as a hydrochloride, a hydrobromide, a phosphate, a sulfate, etc.; and the salts of organic acids such as acetates, propionates, mal onates, benzoates, etc. Additionally, auxiliary substances, such as wetting or emulsifying agents, pH buffering substances, gels or gelling materials, flavorings, colorants, microspheres, polymers, suspension agents, etc. may also be present herein. In addition, one or more other conventional pharmaceutical ingredients, such as preservatives, humectants, suspending agents, surfactants, antioxidants, anticaking agents, fillers, chelating agents, coating agents, chemical stabilizers, etc. may also be present, especially if the dosage form is a reconstitutable form. Suitable exemplary ingredients include microcrystalline cellulose, carboxymethylcellulose sodium, polysorbate 80, phenylethyl alcohol, chlorobutanol, potassium sorbate, sorbic acid, sulfur dioxide, propyl gallate, the parabens, ethyl vanillin, glycerin, phenol, parachlorophenol, gelatin, albumin and a combination thereof. A thorough discussion of pharmaceutically acceptable excipients is available in REMINGTON'S PHARMACEUTICAL SCIENCES (Mack Pub. Co., N.J. 1991) which is incorporated by reference herein.

[0179] In an embodiment herein the delivery is via an adenovirus, which may be at a single booster dose containing at least 1 x 10⁵ particles (also referred to as particle units, pu) of adenoviral vector. In an embodiment herein, the dose preferably is at least about 1 x 10⁶ particles (for example, about 1 x 10⁶- 1 x 10¹¹ particles), more preferably at least about 1 x 10⁷ particles, more preferably at least about 1 x 10⁸ particles (e.g., about 1 x 10⁸-l x 10¹¹ particles or about 1 x 10⁹-l x 10¹² particles), and most preferably at least about 1 x IO¹⁰ particles (e.g., about 1 x 10⁹-l x IO¹⁰ particles or about 1 x 10⁹-l x 10¹² particles), or even at least about 1 x IO¹⁰ particles (e.g., about 1 x 10¹⁰-l x 10¹² particles) of the adenoviral vector. Alternatively, the dose comprises no more than about 1 x 10¹⁴ particles, preferably no more than about 1 x 10¹³ particles, even more preferably no more than about 1 x 10¹² particles, even more preferably no more than about 1 x 10¹¹ particles, and most preferably no more than about 1 x IO¹⁰ particles (e.g., no more than about 1 x 10⁹ particles). Thus, the dose may contain a single dose of adenoviral vector with, for example, about 1 x 10⁶ particle units (pu), about 2 x 10⁶ pu, about 4 x 10⁶ pu, about 1 x 10⁷ pu, about 2 x 10⁷ pu, about 4 x 10⁷ pu, about 1 x 10⁸ pu, about 2 x 10⁸ pu, about 4 x 10⁸ pu, about 1 x 10⁹ pu, about 2 x 10⁹ pu, about 4 x 10⁹ pu, about 1 x IO¹⁰ pu, about 2 x IO¹⁰ pu, about 4 x IO¹⁰ pu, about 1 x 10¹¹ pu, about 2 x 10¹¹ pu, about 4 x 10¹¹ pu, about 1 x 10¹² pu, about 2 x 10¹² pu, or about 4 x 10¹² pu of adenoviral vector. See, for example, the adenoviral vectors in U.S. Pat. No. 8,454,972 B2 to Nabel, et. al., granted on Jun. 4, 2013; incorporated by reference herein, and the dosages at col 29, lines 36-58 thereof. In an embodiment herein, the adenovirus is delivered via multiple doses.

[0180] In an embodiment herein, the delivery is via an AAV. A therapeutically effective dosage for in vivo delivery of the AAV to a human is believed to be in the range of from about 20 to about 50 ml of saline solution containing from about 1 x 10¹⁰ to about 1 x 10⁵⁰ functional AAV/ml solution. The dosage may be adjusted to balance the therapeutic benefit against any side effects. In an embodiment herein, the AAV dose is generally in the range of concentrations of from about 1 x 10⁵ to 1 x 10⁵⁰ genomes AAV (sometimes referred to herein as “vector genomes” or “vg”), from about 1 x 10⁸ to 1 x IO²⁰ genomes AAV, from about 1 x 10¹⁰ to about 1 x 10¹⁶ genomes, or about 1 x 10¹¹ to about 1 x 10¹⁶ genomes AAV. A human dosage may be about 1 x 10¹³ genomes AAV. Such concentrations may be delivered in from about 0.001 ml to about 100 ml, about 0.05 to about 50 ml, or about 10 to about 25 ml of a carrier solution. Other effective dosages can be readily established by one of ordinary skill in the art through routine trials establishing dose response curves. See, for example, U.S. Pat. No. 8,404,658 B2 to Hajjar, et al., granted on Mar. 26, 2013, at col. 27, lines 45-60.

[0181] The promoter used to drive nucleic acid-targeting effector protein coding nucleic acid molecule expression can include: AAV ITR can serve as a promoter: this is advantageous for eliminating the need for an additional promoter element (which can take up space in the vector). The additional space freed up can be used to drive the expression of additional elements (gRNA, etc.). Also, ITR activity is relatively weaker, so can be used to reduce potential toxicity due to over expression of nucleic acid-targeting effector protein. For ubiquitous expression, can use promoters: CMV, CAG, CBh, PGK, SV40, Ferritin heavy or light chains, etc. For brain or other CNS expression, can use promoters: SynapsinI for all neurons, CaMKIIalpha for excitatory neurons, GAD67 or GAD65 or VGAT for GABAergic neurons, etc. For liver expression, can use Albumin promoter. For lung expression, can use SP-B. For endothelial cells, can use ICAM. For hematopoietic cells can use IFNbeta or CD45. For Osteoblasts can use OG-2. [0182] The promoter used to drive guide RNA can include: Pol III promoters such as U6 or Hl Use of Pol II promoter and intronic cassettes to express guide RNA Adeno Associated Virus (AAV).

[0183] Nucleic acid-targeting effector protein and one or more guide RNA can be delivered using adeno associated virus (AAV), lentivirus, adenovirus or other plasmid or viral vector types, in particular, using formulations and doses from, for example, U.S. Pat. No. 8,454,972 (formulations, doses for adenovirus), U.S. Pat. No. 8,404,658 (formulations, doses for AAV) and U.S. Pat. No. 5,846,946 (formulations, doses for DNA plasmids) and from clinical trials and publications regarding the clinical trials involving lentivirus, AAV and adenovirus. For examples, for AAV, the route of administration, formulation and dose can be as in U.S. Pat. No. 8,454,972 and as in clinical trials involving AAV. For Adenovirus, the route of administration, formulation and dose can be as in U.S. Pat. No. 8,404,658 and as in clinical trials involving adenovirus. For plasmid delivery, the route of administration, formulation and dose can be as in U.S. Pat. No. 5,846,946 and as in clinical studies involving plasmids. Doses may be based on or extrapolated to an average 70 kg individual (e.g., a male adult human), and can be adjusted for patients, subjects, mammals of different weight and species.

Frequency of administration is within the ambit of the medical or veterinary practitioner (e.g., physician, veterinarian), depending on usual factors including the age, sex, general health, other conditions of the patient or subject and the particular condition or symptoms being addressed. The viral vectors can be injected into the tissue of interest. For cell-type specific genome modification, the expression of nucleic acid-targeting effector can be driven by a cell-type specific promoter. For example, liver-specific expression might use the Albumin promoter and neuron-specific expression (e.g., for targeting CNS disorders) might use the Synapsin I promoter.

[0184] In terms of in vivo delivery, AAV is advantageous over other viral vectors for a couple of reasons: Low toxicity (this may be due to the purification method not requiring ultra centrifugation of cell particles that can activate the immune response) and Low probability of causing insertional mutagenesis because it doesn't integrate into the host genome.

[0185] AAV has a packaging limit of 4.5 or 4.75 Kb. This means that nucleic acid-targeting effector protein (such as a Type V protein such as C2cl or C2c3) as well as a promoter and transcription terminator have to be all fit into the same viral vector. Therefore embodiments of the invention include utilizing homologs of nucleic acid-targeting effector protein (such as a Type V protein such as C2cl or C2c3) that are shorter.

[0186] As to AAV, the AAV can be AAV1, AAV2, AAV5 or any combination thereof. One can select the AAV of the AAV with regard to the cells to be targeted; e.g., one can select AAV serotypes 1, 2, 5 or a hybrid capsid AAV1, AAV2, AAV5 or any combination thereof for targeting brain or neuronal cells; and one can select AAV4 for targeting cardiac tissue. AAV8 is useful for delivery to the liver. The herein promoters and vectors are preferred individually.

[0187] Packaging cells are typically used to form virus particles that are capable of infecting a host cell. Such cells include 293 cells, which package adenovirus, and psi2 cells or PA317 cells, which package retrovirus. Viral vectors used in gene therapy are usually generated by producing a cell line that packages a nucleic acid vector into a viral particle. The vectors typically contain the minimal viral sequences required for packaging and subsequent integration into a host, other viral sequences being replaced by an expression cassette for the polynucleotide(s) to be expressed. The missing viral functions are typically supplied in trans by the packaging cell line. For example, AAV vectors used in gene therapy typically only possess ITR sequences from the AAV genome which are required for packaging and integration into the host genome. Viral DNA is packaged in a cell line, which contains a helper plasmid encoding the other AAV genes, namely rep and cap, but lacking ITR sequences. The cell line may also be infected with adenovirus as a helper. The helper virus promotes replication of the AAV vector and expression of AAV genes from the helper plasmid. The helper plasmid is not packaged in significant amounts due to a lack of ITR sequences. Contamination with adenovirus can be reduced by, e.g., heat treatment to which adenovirus is more sensitive than AAV. Additional methods for the delivery of nucleic acids to cells are known to those skilled in the art. See, for example, US20030087817, incorporated herein by reference.

[0188] Millington-Ward et al. (Molecular Therapy, vol. 19 no. 4, 642-649 April 2011) describes adeno-associated virus (AAV) vectors to deliver an RNA interference (RNAi)- based rhodopsin suppressor and a codon-modified rhodopsin replacement gene resistant to suppression due to nucleotide alterations at degenerate positions over the RNAi target site. An injection of either 6.0 x 10⁸ vp or 1.8 x 10¹⁰ vp AAV were subretinally injected into the eyes by Millington-Ward et al. The AAV vectors of Millington-Ward et al. may be applied to the system of the present invention, contemplating a dose of about 2 x 10¹¹ to about 6 x 10¹¹ vp administered to a human.

[0189] Dalkara et al. (Sci Transl Med 5, 189ra76 (2013)) also relates to in vivo directed evolution to fashion an AAV vector that delivers wild-type versions of defective genes throughout the retina after noninjurious injection into the eyes' vitreous humor. Dalkara describes a 7 mer peptide display library and an AAV library constructed by DNA shuffling of cap genes from AAV1, 2, 4, 5, 6, 8, and 9. The rcAAV libraries and rAAV vectors expressing GFP under a CAG or Rho promoter were packaged and deoxyribonucleaseresistant genomic titers were obtained through quantitative PCR. The libraries were pooled, and two rounds of evolution were performed, each consisting of initial library diversification followed by three in vivo selection steps. In each such step, P30 rho-GFP mice were intravitreally injected with 2 ml of iodixanol-purified, phosphate-buffered saline (PBS)- dialyzed library with a genomic titer of about 1. times.10. sup.12 vg/ml. The AAV vectors of Dalkara et al. may be applied to the nucleic acid-targeting system of the present invention, contemplating a dose of about 1 x 10¹⁵ to about 1 x 10¹⁶ vg/ml administered to a human.

[0190] The tropism of a retrovirus can be altered by incorporating foreign envelope proteins, expanding the potential target population of target cells. Lentiviral vectors are retroviral vectors that are able to transduce or infect non-dividing cells and typically produce high viral titers. Selection of a retroviral gene transfer system would therefore depend on the target tissue. Retroviral vectors are comprised of cis-acting long terminal repeats with packaging capacity for up to 6-10 kb of foreign sequence. The minimum cis-acting LTRs are sufficient for replication and packaging of the vectors, which are then used to integrate the therapeutic gene into the target cell to provide permanent transgene expression. Widely used retroviral vectors include those based upon murine leukemia virus (MuLV), gibbon ape leukemia virus (GaLV), Simian Immuno deficiency virus (SW), human immuno deficiency virus (HIV), and combinations thereof (see, e.g., Buchscher et al., J. Virol. 66:2731-2739 (1992); Johann et al., J. Virol. 66: 1635-1640 (1992); Sommnerfelt et al., Virol. 176:58-59 (1990); Wilson et al., J. Virol. 63:2374-2378 (1989); Miller et al., J. Virol. 65:2220-2224 (1991); PCT/US94/05700). In applications where transient expression is preferred, adenoviral based systems may be used. Adenoviral based vectors are capable of very high transduction efficiency in many cell types and do not require cell division. With such vectors, high titer and levels of expression have been obtained. This vector can be produced in large quantities in a relatively simple system. Adeno-associated virus ("AAV") vectors may also be used to transduce cells with target nucleic acids, e.g., in the in vitro production of nucleic acids and peptides, and for in vivo and ex vivo gene therapy procedures (see, e.g., West et al., Virology 160:38-47 (1987); U.S. Pat. No. 4,797,368; WO 93/24641; Kotin, Human Gene Therapy 5:793-801 (1994); Muzyczka, J. Clin. Invest. 94: 1351 (1994). Construction of recombinant AAV vectors are described in a number of publications, including U.S. Pat. No. 5,173,414; Tratschin et al., Mol. Cell. Biol. 5:3251-3260 (1985); Tratschin, et al., Mol. Cell. Biol. 4:2072-2081 (1984); Hermonat & Muzyczka, PNAS 81 :6466-6470 (1984); and Samulski et al., J. Virol. 63:03822- 3828 (1989).

[0191] Packaging cells are typically used to form virus particles that are capable of infecting a host cell. Such cells include 293 cells, which package adenovirus, and yr2 cells or PA317 cells, which package retrovirus. Viral vectors used in gene therapy are usually generated by producing a cell line that packages a nucleic acid vector into a viral particle. The vectors typically contain the minimal viral sequences required for packaging and subsequent integration into a host, other viral sequences being replaced by an expression cassette for the polynucleotide(s) to be expressed. The missing viral functions are typically supplied in trans by the packaging cell line. For example, AAV vectors used in gene therapy typically only possess ITR sequences from the AAV genome which are required for packaging and integration into the host genome. Viral DNA is packaged in a cell line, which contains a helper plasmid encoding the other AAV genes, namely rep and cap, but lacking ITR sequences. The cell line may also be infected with adenovirus as a helper. The helper virus promotes replication of the AAV vector and expression of AAV genes from the helper plasmid. The helper plasmid is not packaged in significant amounts due to a lack of ITR sequences. Contamination with adenovirus can be reduced by, e.g., heat treatment to which adenovirus is more sensitive than AAV. Additional methods for the delivery of nucleic acids to cells are known to those skilled in the art. See, for example, US20030087817, incorporated herein by reference.

[0192] In some embodiments, a host cell is transiently or non-transiently transfected with one or more vectors described herein. In some embodiments, a cell is transfected as it naturally occurs in a subject. In some embodiments, a cell that is transfected is taken from a subject. Cells taken from a subject include, but are not limited to, hepatocytes or cells isolated from muscle, the CNS, eye or lung. Immunological cells are also contemplated, such as but not limited to T cells, HSCs, B-cells and NK cells.

[0193] Another useful method to deliver proteins, enzymes, and guides comprises transfection of messenger RNA (mRNA). Examples of mRNA delivery methods and compositions that may be utilized in the present disclosure including, for example, PCT/US2014/028330, US8822663B2, NZ700688A, ES2740248T3, EP2755693A4, EP2755986A4, WO2014152940A1, EP3450553B1, BRI 12016030852A2, and EP3362461A1. Expression of CRISPR systems in particular is described by W02020014577. Each of these publications are incorporated herein by reference in their entireties. Additional disclosure hereby incorporated by reference can be found in Kowalski et al., “Delivering the Messenger: Advances in Technologies for Therapeutic mRNA Delivery,” Mol Therap., 2019; 27(4): 710-728.

[0194] In some embodiments, the cell is derived from cells taken from a subject, such as a cell line. A wide variety of cell lines for tissue culture are known in the art. Examples of cell lines include, but are not limited to, C8161, CCRF-CEM, MOLT, mIMCD-3, NHDF, HeLa- S3, Huhl, Huh4, Huh7, HUVEC, HASMC, HEKn, HEKa, MiaPaCell, Panel, PC-3, TF1, CTLL-2, CIR, Rat6, CV1, RPTE, A10, T24, J82, A375, ARH-77, Calul, SW480, SW620, SKOV3, SK-UT, CaCo2, P388D1, SEM-K2, WEHL231, HB56, TIB55, Jurkat, J45.01, LRMB, Bcl-1, BC-3, IC21, DLD2, Raw264.7, NRK, NRK-52E, MRC5, MEF, Hep G2, HeLa B, HeLa T4, COS, COS-1, COS-6, COS-M6A, BS-C-1 monkey kidney epithelial, BALB/3T3 mouse embryo fibroblast, 3T3 Swiss, 3T3-L1, 132-d5 human fetal fibroblasts; 10.1 mouse fibroblasts, 293-T, 3T3, 721, 9L, A2780, A2780ADR, A2780cis, A172, A20, A253, A431, A-549, ALC, B16, B35, BCP-1 cells, BEAS-2B, bEnd.3, BHK-21, BR 293, BxPC3, C3H-10T1/2, C6/36, Cal-27, CHO, CHO-7, CHO-IR, CHO-K1, CHO-K2, CHO-T, CHO Dhfr-/-, COR-L23, COR-L23/CPR, COR-L23/5010, COR-L23/R23, COS-7, COV-434, CML Tl, CMT, CT26, D17, DH82, DU145, DuCaP, EL4, EM2, EM3, EMT6/AR1, EMT6/AR10.0, FM3, H1299, H69, HB54, HB55, HCA2, HEK-293, HeLa, Hepalclc7, HL- 60, HMEC, HT-29, Jurkat, JY cells, K562 cells, Ku812, KCL22, KG1, KYO1, LNCap, Ma- Mel 1-48, MC-38, MCF-7, MCF-10A, MDA-MB-231, MDA-MB-468, MDA-MB-435, MDCK II, MDCK II, MOR/0.2R, MONO-MAC 6, MTD-1 A, MyEnd, NCLH69/CPR, NCI- H69/LX10, NCI-H69/LX20, NCI-H69/LX4, NIH-3T3, NALM-1, NW-145, OPCN/OPCT cell lines, Peer, PNT-1A/PNT 2, RenCa, RIN-5F, RMA/RMAS, Saos-2 cells, Sf-9, SkBr3, T2, T-47D, T84, THP1 cell line, U373, U87, U937, VCaP, Vero cells, WM39, WT-49, X63, YAC-1, YAR, and transgenic varieties thereof. Cell lines are available from a variety of sources known to those with skill in the art (see, e.g., the American Type Culture Collection (ATCC) (Manassas, Va.)). In some embodiments, a cell transfected with one or more vectors described herein is used to establish a new cell line comprising one or more vector-derived sequences.

[0195] In some embodiments, one or more vectors described herein are used to produce a non-human transgenic animal or transgenic plant. In some embodiments, the transgenic animal is a mammal, such as a mouse, rat, or rabbit. In certain embodiments, the organism or subject is a plant. In certain embodiments, the organism or subject or plant is algae. Methods for producing transgenic plants and animals are known in the art, and generally begin with a method of cell transfection, such as described herein.

[0196] In one aspect, the invention provides for methods of modifying a target polynucleotide in a prokaryotic or eukaryotic cell, which may be in vivo, ex vivo or in vitro. In some embodiments, the method comprises sampling a cell or population of cells from a human or non-human animal or plant (including micro-algae) and modifying the cell or cells. Culturing may occur at any stage ex vivo. The cell or cells may even be re-introduced into the non-human animal or plant (including micro-algae).

[0197] In plants, pathogens are often host-specific. For example, Fusariumn oxysporum f. sp. lycopersici causes tomato wilt but attacks only tomato, and F. oxysporum f. dianthii Puccinia graminis f. sp. tritici attacks only wheat. Plants have existing and induced defenses to resist most pathogens. Mutations and recombination events across plant generations lead to genetic variability that gives rise to susceptibility, especially as pathogens reproduce with more frequency than plants. In plants there can be non-host resistance, e.g., the host and pathogen are incompatible. There can also be Horizontal Resistance, e.g., partial resistance against all races of a pathogen, typically controlled by many genes and Vertical Resistance, e.g., complete resistance to some races of a pathogen but not to other races, typically controlled by a few genes. In a Gene-for-Gene level, plants and pathogens evolve together, and the genetic changes in one balance changes in other. Accordingly, using Natural Variability, breeders combine most useful genes for Yield. Quality, Uniformity, Hardiness, Resistance. The sources of resistance genes include native or foreign Varieties, Heirloom Varieties, Wild Plant Relatives, and Induced Mutations, e.g., treating plant material with mutagenic agents. Using the present invention, plant breeders are provided with a new tool to induce mutations. Accordingly, one skilled in the art can analyze the genome of sources of resistance genes, and in Varieties having desired characteristics or traits employ the present invention to induce the rise of resistance genes, with more precision than previous mutagenic agents and hence accelerate and improve plant breeding programs.

[0198] Examples of target polynucleotides include a sequence associated with a signaling biochemical pathway, e.g., a signaling biochemical pathway-associated gene or polynucleotide. Examples of target polynucleotides include a disease associated gene or polynucleotide. A “disease-associated” gene or polynucleotide refers to any gene or polynucleotide which is yielding transcription or translation products at an abnormal level or in an abnormal form in cells derived from a disease-affected tissues compared with tissues or cells of a non disease control. It may be a gene that becomes expressed at an abnormally high level; it may be a gene that becomes expressed at an abnormally low level, where the altered expression correlates with the occurrence and/or progression of the disease. A disease- associated gene also refers to a gene possessing mutation(s) or genetic variation that is directly responsible or is in linkage disequilibrium with a gene(s) that is responsible for the etiology of a disease. The transcribed or translated products may be known or unknown, and may be at a normal or abnormal level.

5.9.5. Lipid Nanoparticle Delivery

[0199] In some embodiments, the delivery system is packaged in one or more LNPs and administered intravenously. In some embodiments, the co-delivery system is packaged in one or more LNPs and administered intrathecally. In some embodiments, the co-delivery system is packaged in one or more LNPs and administered by intracerebral ventricular injection. In some embodiments, the co-delivery system is packaged in one or more LNPs and administered by intracistemal magna administration. In some embodiments, the co-delivery system is packaged in one or more LNPs and administered by intravitreal injection.

[0200] The preparation of lipidmucleic acid complexes, including targeted liposomes such as immunolipid complexes, is well known to one of skill in the art (see, e.g., Crystal, Science 270:404-410 (1995); Blaese et al., Cancer Gene Ther. 2:291-297 (1995); Behr et al., Bioconjugate Chem. 5:382-389 (1994); Remy et al., Bioconjugate Chem. 5:647-654 (1994); Gao et al., Gene Therapy 2:710-722 (1995); Ahmad et al., Cancer Res. 52:4817-4820 (1992); U.S. Pat. Nos. 4,186,183, 4,217,344, 4,235,871, 4,261,975, 4,485,054, 4,501,728, 4,774,085, 4,837,028, and 4,946,787). In some embodiments, the LNP formulations are selected from LP01 (Cas No. 1799316-64-5), ALC-0315 (Cas No. 2036272-55-4), and cKK-E12 (Cas No. 1432494-65-9). In some embodiments, the LNP formulation is LP01. In some embodiments, the LNP formulation is ALC-0315. In some embodiment, the LNP formulation is cKK-E12.

[0201] In some embodiments, LNP doses range from about 0.1 mg/kg to about 100 mg/kg (or any of the values or subranges therein). In some embodiments, LNP doses is about 0.1 mg/kg, about 0.2 mg/kg, about 0.3 mg/kg, about 0.4 mg/kg, about 0.5 mg/kg, about 0.6 mg/kg, about 0.7 mg/kg, about 0.8 mg/kg, about 0.9 mg/kg, about 1.0 mg/kg, 1.5 mg/kg, about 2 mg/kg, about 2.5 mg/kg, about 3 mg/kg, about 3.5 mg/kg, about 4 mg/kg, about 4.5 mg/kg, about 5 mg/kg, about 6 mg/kg, about7 mg/kg, about 7 mg/kg, about 8 mg/kg, about 9 mg/kg, about 10 mg/kg, about 15 mg/kg, about 20 mg/kg, about 25 mg/kg, about 30 mg/kg, about 35 mg/kg, about 40 mg/kg, about 45 mg/kg, or about 50 mg/kg or more.

[0202] In another embodiment, LNP doses of about 0.01 to about 1 mg per kg of body weight administered intravenously are contemplated. Medications to reduce the risk of infusion- related reactions are contemplated, such as dexamethasone, acetaminophen, diphenhydramine or cetirizine, and ranitidine are contemplated. Multiple doses of about 0.3 mg per kilogram every 4 weeks for five doses are also contemplated.

[0203] The charge of the LNP must be taken into consideration. As cationic lipids combined with negatively charged lipids to induce nonbilayer structures that facilitate intracellular delivery. Because charged LNPs are rapidly cleared from circulation following intravenous injection, ionizable cationic lipids with pKa values below 7 were developed (see, e.g., Rosin et al, Molecular Therapy, vol. 19, no. 12, pages 1286-2200, December 2011). Negatively charged polymers such as RNA may be loaded into LNPs at low pH values (e.g., pH 4) where the ionizable lipids display a positive charge. However, at physiological pH values, the LNPs exhibit a low surface charge compatible with longer circulation times. Four species of ionizable cationic lipids have been focused upon, namely l,2-dilineoyl-3- dimethylammonium-propane (DLinDAP), l,2-dilinoleyloxy-3-N,N-dimethylaminopropane (DLinDMA), l,2-dilinoleyloxy-keto-N,N-dimethyl-3 -aminopropane (DLinKDMA), and 1,2- dilinoleyl-4-(2-dimethylaminoethyl)-[l,3]-dioxolane (DLinKC2-DMA). It has been shown that LNP siRNA systems containing these lipids exhibit remarkably different gene silencing properties in hepatocytes in vivo, with potencies varying according to the series DLinKC2- DMA>DLinKDMA>DLinDMA»DLinDAP employing a Factor VII gene silencing model (see, e.g., Rosin et al, Molecular Therapy, vol. 19, no. 12, pages 1286-2200, December 2011).

A dosage of 1 pg/ml of LNP in or associated with the LNP may be contemplated, especially for a formulation containing DLinKC2-DMA.

[0204] In some embodiments, the LNP composition comprises one or more one or more ionizable lipids. As used herein, the term "ionizable lipid" has its ordinary meaning in the art and may refer to a lipid comprising one or more charged moieties. In some embodiments, an ionizable lipid may be positively charged or negatively charged. In principle, there are no specific limitations concerning the ionizable lipids of the LNP compositions disclosed herein. In some embodiments, the one or more ionizable lipids are selected from the group consisting of 3-(didodecylamino)-Nl,Nl,4-tridodecyl-l-piperazineethanamine (KL10), Nl-[2- (didodecylamino)ethyl]-Nl,N4,N4-tridodecyl-l,4-piperazinediethanami- ne (KL22), 14,25- ditridecyl- 15, 18,21 ,24-tetraaza-octatriacontane (KL25), 1 ,2-dilinoleyloxy-N,N- dimethylaminopropane (DLin-DMA), 2, 2-dilinoleyl-4-dimethylaminomethyl-[l,3]-di oxolane (DLin-K-DMA), heptatriaconta-6,9,28,31-tetraen-19-yl 4-(dimethylamino)butanoate (DLin- MC3-DMA), 2, 2-dilinoleyl-4-(2-dimethylaminoethyl)-[l,3]-di oxolane (DLin-KC2-DMA), l,2-dioleyloxy-N,N-dimethylaminopropane (DODMA), 2-({8-[(3)-cholest-5-en-3- yloxy]octyl}oxy)-N,N-dimethyl-3-[(9Z,12Z)-octad- eca-9,12-dien-l-yloxy]propan-l-amine (Octyl-CLinDMA), (2R)-2-({8-[(3)-cholest-5-en-3-yloxy]octyl}oxy)-N,N-dimethyl-3- [(9Z,12Z)— octadeca-9,12-dien-l-yloxy]propan-l -amine (Octyl-CLinDMA (2R)), and (2S)- 2-({8-[(3)-cholest-5-en-3-yloxy]octyl}oxy)-N,N-dimethyl-3-[(9Z,12Z)— octadeca-9,12-dien- 1-y loxy]propan-l -amine (Octyl-CLinDMA (2S)). In one embodiment, the ionizable lipid may be selected from, but not limited to, an ionizable lipid described in International Publication Nos. WO2013086354 and WO2013116126.

[0205] In some embodiments, the lipid nanoparticle may include one or more (e.g., 1, 2, 3, 4, 5, 6, 7, or 8) cationic and/or ionizable lipids. Such cationic and/or ionizable lipids include, but are not limited to, 3-(didodecylamino)-Nl,Nl,4-tridodecyl-l-piperazineethanamine (KL10), Nl-[2-(didodecylamino)ethyl]-Nl,N4,N4-tridodecyl-l,4-piperazinediethanami- ne (KL22), 14,25-ditridecyl-15,18,21,24-tetraaza-octatriacontane (KL25), l,2-dilinoleyloxy-N,N- dimethylaminopropane (DLin-DMA), 2, 2-dilinoleyl-4-dimethylaminomethyl-[l,3]-di oxolane (DLin-K-DMA), heptatriaconta-6,9,28,31-tetraen-19-yl 4-(dimethylamino)butanoate (DLin- MC3-DMA), 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[l,3]-dioxolane (DLin-KC2-DMA), 2- ({8-[(3.beta.)-cholest-5-en-3-yloxy]octyl}oxy)-N,N-dimethyl-3-[(9Z,12Z)- -octadeca-9,12- dien-l-yloxy]propan-l -amine (Octyl-CLinDMA), (2R)-2-({8-[(3.beta.)-cholest-5-en-3- yloxy]octyl}oxy)-N,N-dimethyl-3-[(9Z- ,12Z)-octadeca-9,12-dien-l-yl oxy]propan-l -amine (Octyl-CLinDMA (2R)), (2S)-2-({8-[(3Pcholest-5-en-3-yloxy]octyl}oxy)-N,N-dimethyl-3- [(9Z- ,12Z)-octadeca-9,12-dien-l-yl oxy]propan-l -amine (Octyl-CLinDMA (2S)).N,N- dioleyl-N,N-dimethylammonium chloride ("DODAC"); N-(2,3-dioleyloxy)propyl-N,N— N- triethylammonium chloride ("DOTMA"); N,N-distearyl-N,N-dimethylammonium bromide ("DDAB"); N-(2,3-dioleoyloxy)propyl)-N,N,N-trimethylammonium chloride ("DOTAP"); l,2-Dioleyloxy-3 -trimethylaminopropane chloride salt ("DOTAP. Cl"); 3-.beta.-(N--(N',N'- dimethylaminoethane)-carbamoyl)cholesterol ("DC-Chol"), N-(l-(2,3-dioleyloxy)propyl)-N- 2-(sperminecarboxamido)ethyl)-N,N-dimethyl- -ammonium trifluoracetate ("DOSPA"), dioctadecylamidoglycyl carboxyspermine ("DOGS"), l,2-dioleoyl-3 -dimethylammonium propane ("DODAP"), N,N-dimethyl-2,3-dioleyloxy)propylamine ("DODMA"), and N-(l,2- dimyristyloxyprop-3-yl)-N,N-dimethyl-N-hydroxyethyl ammonium bromide ("DMRIE"). Additionally, a number of commercial preparations of cationic and/or ionizable lipids can be used, such as, e.g., LIPOFECTIN.RTM. (including DOTMA and DOPE, available from GIBCO/BRL), and LIPOFECTAMINE.RTM. (including DOSPA and DOPE, available from GIBCO/BRL). KL10, KL22, and KL25 are described, for example, in U.S. Pat. No. 8,691,750.

[0206] In some embodiments, the LNP composition comprises one or more amino lipids. The terms "amino lipid" and "cationic lipid" are used interchangeably herein to include those lipids and salts thereof having one, two, three, or more fatty acid or fatty alkyl chains and a pH-titratable amino head group (e.g., an alkylamino or dialkylamino head group). In principle, there are no specific limitations concerning the amino lipids of the LNP compositions disclosed herein. The cationic lipid is typically protonated (i.e., positively charged) at a pH below the pKa of the cationic lipid and is substantially neutral at a pH above the pKa. The cationic lipids can also be termed titratable cationic lipids. In some embodiments, the one or more cationic lipids include: a protonatable tertiary amine (e.g., pH- titratable) head group; alkyl chains, wherein each alkyl chain independently has 0 to 3 (e.g., 0, 1, 2, or 3) double bonds; and ether, ester, or ketal linkages between the head group and alkyl chains. Such cationic lipids include, but are not limited to, DSDMA, DODMA, DOTMA, DLinDMA, DLenDMA, .gamma. -DLenDMA, DLin-K-DMA, DLin-K-C2-DMA (also known as DLin-C2K-DMA, XTC2, and C2K), DLin-K-C3-DMA, DLin-K-C4-DMA, DLen-C2K-DMA, y-DLen-C2-DMA, C12-200, cKK-E12, cKK-A12, cKK-012, DLin-MC2- DMA (also known as MC2), and DLin-MC3-DMA (also known as MC3).

[0207] Anionic lipids suitable for use in lipid nanoparticles include, but are not limited to, phosphatidylglycerol, cardiolipin, diacylphosphatidylserine, diacylphosphatidic acid, N- dodecanoyl phosphatidylethanoloamine, N-succinyl phosphatidylethanolamine, N-glutaryl phosphatidylethanolamine, lysylphosphatidylglycerol, and other anionic modifying groups joined to neutral lipids.

[0208] Neutral lipids (including both uncharged and zwitterionic lipids) suitable for use in lipid nanoparticles include, but are not limited to, diacylphosphatidylcholine, diacylphosphatidylethanolamine, ceramide, sphingomyelin, dihydrosphingomyelin, cephalin, sterols (e.g., cholesterol) and cerebrosides. In some embodiments, the lipid nanoparticle comprises cholesterol. Lipids having a variety of acyl chain groups of varying chain length and degree of saturation are available or may be isolated or synthesized by well-known techniques. Additionally, lipids having mixtures of saturated and unsaturated fatty acid chains and cyclic regions can be used. In some embodiments, the neutral lipids used in the disclosure are DOPE, DSPC, DPPC, POPC, or any related phosphatidylcholine. In some embodiments, the neutral lipid may be composed of sphingomyelin, dihydrosphingomyeline, or phospholipids with other head groups, such as serine and inositol.

[0209] In some embodiments, amphipathic lipids are included in nanoparticles. Exemplary amphipathic lipids suitable for use in nanoparticles include, but are not limited to, sphingolipids, phospholipids, fatty acids, and amino lipids.

[0210] The lipid composition of the pharmaceutical composition may comprise one or more phospholipids, for example, one or more saturated or (poly)unsaturated phospholipids or a combination thereof. In general, phospholipids comprise a phospholipid moiety and one or more fatty acid moieties.

[0211] A phospholipid moiety can be selected, for example, from the non-limiting group consisting of phosphatidyl choline, phosphatidyl ethanolamine, phosphatidyl glycerol, phosphatidyl serine, phosphatidic acid, 2-lysophosphatidyl choline, and a sphingomyelin.

[0212] A fatty acid moiety can be selected, for example, from the non-limiting group consisting of lauric acid, myristic acid, myristoleic acid, palmitic acid, palmitoleic acid, stearic acid, oleic acid, linoleic acid, alpha-linolenic acid, erucic acid, phytanoic acid, arachidic acid, arachidonic acid, eicosapentaenoic acid, behenic acid, docosapentaenoic acid, and docosahexaenoic acid.

[0213] Particular amphipathic lipids can facilitate fusion to a membrane. For example, a cationic phospholipid can interact with one or more negatively charged phospholipids of a membrane (e.g., a cellular or intracellular membrane). Fusion of a phospholipid to a membrane can allow one or more elements (e.g., a therapeutic agent) of a lipid-containing composition (e.g., LNPs) to pass through the membrane permitting, e.g., delivery of the one or more elements to a target tissue.

[0214] Non-natural amphipathic lipid species including natural species with modifications and substitutions including branching, oxidation, cyclization, and alkynes are also contemplated. For example, a phospholipid can be functionalized with or cross-linked to one or more alkynes (e.g., an alkenyl group in which one or more double bonds is replaced with a triple bond). Under appropriate reaction conditions, an alkyne group can undergo a copper- catalyzed cycloaddition upon exposure to an azide. Such reactions can be useful in functionalizing a lipid bilayer of a nanoparticle composition to facilitate membrane permeation or cellular recognition or in conjugating a nanoparticle composition to a useful component such as a targeting or imaging moiety (e.g., a dye).

[0215] Phospholipids include, but are not limited to, glycerophospholipids such as phosphatidylcholines, phosphatidylethanolamines, phosphatidylserines, phosphatidylinositols, phosphatidy glycerols, and phosphatidic acids. Phospholipids also include phosphosphingolipid, such as sphingomyelin.

[0216] In some embodiments, the LNP composition comprises one or more phospholipids. In some embodiments, the phospholipid is selected from the group consisting of 1,2-dilinoleoyl- sn-glycero-3 -phosphocholine (DLPC), 1,2-dimyristoyl-sn-glycero-phosphocholine (DMPC), l,2-dioleoyl-sn-glycero-3 -phosphocholine (DOPC), l,2-dipalmitoyl-sn-glycero-3- phosphocholine (DPPC), l,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2- diundecanoyl-sn-glycero-phosphocholine (DUPC), l-palmitoyl-2-oleoyl-sn-glycero-3- phosphocholine (POPC), l,2-di-O-octadecenyl-sn-glycero-3 -phosphocholine (18:0 Diether PC), l-oleoyl-2-cholesterylhemisuccinoyl-sn-glycero-3-phosphocholine (OChemsPC), 1- hexadecyl-sn-glycero-3 -phosphocholine (Cl 6 Lyso PC), l,2-dilinolenoyl-sn-glycero-3- phosphocholine, 1 ,2-diarachidonoyl-sn-glycero-3 -phosphocholine, 1 ,2-didocosahexaenoyl- sn-glycero-3 -phosphocholine, l,2-dioleoyl-sn-glycero-3 -phosphoethanolamine (DOPE), 1,2- diphytanoyl-sn-glycero-3-phosphoethanolamine (ME 16:0 PE), l,2-distearoyl-sn-glycero-3- phosphoethanolamine, l,2-dilinoleoyl-sn-glycero-3 -phosphoethanolamine, 1,2-dilinolenoyl- sn-glycero-3 -phosphoethanolamine, l,2-diarachidonoyl-sn-glycero-3- phosphoethanolaminel,2-didocosahexaenoyl— sn-glycero-3-phosphoethanolamine, 1,2- dioleoyl-sn-glycero-3-phospho-rac-(l -glycerol) sodium salt (DOPG), sphingomyelin, and any mixtures thereof.

[0217] Other phosphorus-lacking compounds, such as sphingolipids, glycosphingolipid families, diacylglycerols, and .beta.-acyloxyacids, may also be used. Additionally, such amphipathic lipids can be readily mixed with other lipids, such as triglycerides and sterols.

[0218] In some embodiments, the LNP composition comprises one or more helper lipids. The term "helper lipid" as used herein refers to lipids that enhance transfection (e.g., transfection of an LNP comprising an mRNA that encodes a site-directed endonuclease, such as a SpCas9 polypeptide). In principle, there are no specific limitations concerning the helper lipids of the LNP compositions disclosed herein. Without being bound to any particular theory, it is believed that the mechanism by which the helper lipid enhances transfection includes enhancing particle stability. In some embodiments, the helper lipid enhances membrane fusogenicity. Generally, the helper lipid of the LNP compositions disclosure herein can be any helper lipid known in the art. Non-limiting examples of helper lipids suitable for the compositions and methods include steroids, sterols, and alkyl resorcinols. Particularly helper lipids suitable for use in the present disclosure include, but are not limited to, saturated phosphatidylcholine (PC) such as distearoyl-PC (DSPC) and dipalymitoyl-PC (DPPC), di oleoylphosphatidylethanolamine (DOPE), 1 ,2-dilinoleoyl-sn-glycero-3 -phosphocholine (DLPC), cholesterol, 5-heptadecylresorcinol, and cholesterol hemisuccinate. In some embodiments, the helper lipid of the LNP composition includes cholesterol.

[0219] In some embodiments, the LNP composition comprises one or more structural lipids. As used herein, the term "structural lipid" refers to sterols and also to lipids containing sterol moieties. Without being bound to any particular theory, it is believed that the incorporation of structural lipids into the LNPs mitigates aggregation of other lipids in the particle. Structural lipids can be selected from the group including but not limited to, cholesterol, fecosterol, sitosterol, ergosterol, campesterol, stigmasterol, brassicasterol, tomatidine, tomatine, ursolic acid, alpha-tocopherol, hopanoids, phytosterols, steroids, and mixtures thereof. In some embodiments, the structural lipid is a sterol. As defined herein, "sterols" are a subgroup of steroids consisting of steroid alcohols. In certain embodiments, the structural lipid is a steroid. In some embodiments, the structural lipid is cholesterol. In certain embodiments, the structural lipid is an analog of cholesterol.

[0220] The lipid component of a lipid nanoparticle composition may include one or more molecules comprising polyethylene glycol, such as PEG or PEG-modified lipids. In some embodiments, the LNP composition disclosed herein comprise one or more polyethylene glycol (PEG) lipid. The term "PEG-lipid" refers to polyethylene glycol (PEG)-modified lipids. Such lipids are also referred to as PEGylated lipids. Non-limiting examples of PEG- lipids include PEG-modified phosphatidylethanolamine and phosphatidic acid, PEG- ceramide conjugates (e.g., PEG-CerC14 or PEG-CerC20), PEG-modified dialkylamines and PEG-modified l,2-diacyloxypropan-3 -amines For example, a PEG lipid can be PEG-c- DOMG, PEG-DMG, PEG-DLPE, PEG-DMPE, PEG-DPPC, or a PEG-DSPE lipid. In some embodiments, the PEG-lipid includes, but not limited to 1,2-dimyristoyl-sn-glycerol methoxypolyethylene glycol (PEG-DMG), l,2-distearoyl-sn-glycero-3- phosphoethanolamine-N-[amino(polyethylene glycol)] (PEG-DSPE), PEG-disteryl glycerol (PEG-DSG), PEG-dipalmetoleyl, PEG-dioleyl, PEG-distearyl, PEG-diacylglycamide (PEGDAG), PEG-dipalmitoyl phosphatidylethanolamine (PEG-DPPE), or PEG-1, 2- dimyristyloxlpropyl-3-amine (PEG-c-DMA). In some embodiments, the PEG-lipid is selected from the group consisting of a PEG-modified phosphatidylethanolamine, a PEG-modified phosphatidic acid, a PEG-modified ceramide, a PEG-modified dialkylamine, a PEG-modified diacylglycerol, a PEG-modified dialkylglycerol, and mixtures thereof. In some embodiments, the lipid moiety of the PEG-lipids includes those having lengths of from about C. sub.14 to about C. sub.22, preferably from about C. sub.14 to about C. sub.16. In some embodiments, a PEG moiety, for example a mPEG-NH.sub.2, has a size of about 1000, 2000, 5000, 10,000, 15,000 or 20,000 daltons. In some embodiment, the PEG-lipid is PEG2k-DMG. In some embodiments, the one or more PEG lipids of the LNP composition comprises PEG-DMPE. In some embodiments, the one or more PEG lipids of the LNP composition comprises PEG- DMG.

[0221] In some embodiments, the ratio between the lipid components and the nucleic acid molecules of the LNP composition, e.g., the weight ratio, is sufficient for (i) formation of LNPs with desired characteristics, e.g., size, charge, and (ii) delivery of a sufficient dose of nucleic acid at a dose of the lipid component(s) that is tolerable for in vivo administration as readily ascertained by one of skill in the art.

[0222] In certain embodiments, it is desirable to target a nanoparticle, e.g., a lipid nanoparticle, using a targeting moiety that is specific to a cell type and/or tissue type. In some embodiments, a nanoparticle may be targeted to a particular cell, tissue, and/or organ using a targeting moiety. In particular embodiments, a nanoparticle comprises a targeting moiety. Exemplary non-limiting targeting moieties include ligands, cell surface receptors, glycoproteins, vitamins (e.g., riboflavin) and antibodies (e.g., full-length antibodies, antibody fragments (e.g., Fv fragments, single chain Fv (scFv) fragments, Fab' fragments, or F(ab')2 fragments), single domain antibodies, camelid antibodies and fragments thereof, human antibodies and fragments thereof, monoclonal antibodies, and multispecific antibodies (e.g., bispecific antibodies)). In some embodiments, the targeting moiety may be a polypeptide. The targeting moiety may include the entire polypeptide (e.g., peptide or protein) or fragments thereof. A targeting moiety is typically positioned on the outer surface of the nanoparticle in such a manner that the targeting moiety is available for interaction with the target, for example, a cell surface receptor. A variety of different targeting moieties and methods are known and available in the art, including those described, e.g., in Sapra et al., Prog. Lipid Res. 42(5):439-62, 2003 and Abra et al., J. Liposome Res. 12: 1-3, 2002.

[0223] In some embodiments, a lipid nanoparticle (e.g., a liposome) may include a surface coating of hydrophilic polymer chains, such as polyethylene glycol (PEG) chains (see, e.g., Allen et al., Biochimica et Biophysica Acta 1237: 99-108, 1995; DeFrees et al., Journal of the American Chemistry Society 118: 6101-6104, 1996; Blume et al., Biochimica et Biophysica Acta 1149: 180-184,1993; Klibanov et al., Journal of Liposome Research 2: 321-334, 1992; U.S. Pat. No. 5,013,556; Zalipsky, Bioconjugate Chemistry 4: 296-299, 1993; Zalipsky, FEBS Letters 353: 71-74, 1994; Zalipsky, in Stealth Liposomes Chapter 9 (Lasic and Martin, Eds) CRC Press, Boca Raton Fla., 1995). In one approach, a targeting moiety for targeting the lipid nanoparticle is linked to the polar head group of lipids forming the nanoparticle. In another approach, the targeting moiety is attached to the distal ends of the PEG chains forming the hydrophilic polymer coating (see, e.g., Klibanov et al., Journal of Liposome Research 2: 321-334, 1992; Kirpotin et al., FEBS Letters 388: 115-118, 1996). [0224] Standard methods for coupling the targeting moiety or moi eties may be used. For example, phosphatidylethanolamine, which can be activated for attachment of targeting moieties, or derivatized lipophilic compounds, such as lipid-derivatized bleomycin, can be used. Antibody -targeted liposomes can be constructed using, for instance, liposomes that incorporate protein A (see, e.g., Renneisen et al., J. Bio. Chem., 265: 16337-16342, 1990 and Leonetti et al., Proc. Natl. Acad. Sci. (USA), 87:2448-2451, 1990). Other examples of antibody conjugation are disclosed in U.S. Pat. No. 6,027,726. Examples of targeting moieties can also include other polypeptides that are specific to cellular components, including antigens associated with neoplasms or tumors. Polypeptides used as targeting moieties can be attached to the liposomes via covalent bonds (see, for example Heath, Covalent Attachment of Proteins to Liposomes, 149 Methods in Enzymology 111-119 (Academic Press, Inc. 1987)). Other targeting methods include the biotin-avidin system.

[0225] In some embodiments, a lipid nanoparticle includes a targeting moiety that targets the lipid nanoparticle to a cell including, but not limited to, hepatocytes, colon cells, epithelial cells, hematopoietic cells, epithelial cells, endothelial cells, lung cells, bone cells, stem cells, mesenchymal cells, neural cells, cardiac cells, adipocytes, vascular smooth muscle cells, cardiomyocytes, skeletal muscle cells, beta cells, pituitary cells, synovial lining cells, ovarian cells, testicular cells, fibroblasts, B cells, T cells, reticulocytes, leukocytes, granulocytes, and tumor cells (including primary tumor cells and metastatic tumor cells). In particular embodiments, the targeting moiety targets the lipid nanoparticle to a hepatocyte.

[0226] The lipid nanoparticles described herein may be lipidoid-based. The synthesis of lipidoids has been extensively described and formulations containing these compounds are particularly suited for delivery of polynucleotides (see Mahon et al., Bioconjug Chem. 2010 21 : 1448-1454; Schroeder et al., J Intern Med. 2010 267:9-21; Akinc et al., Nat. Biotechnol. 2008 26:561-569; Love et al., Proc Natl Acad Sci USA. 2010 107: 1864-1869; Siegwart et al., Proc Natl Acad Sci USA. 2011 108: 12996-3001).

[0227] The characteristics of optimized lipidoid formulations for intramuscular or subcutaneous routes may vary significantly depending on the target cell type and the ability of formulations to diffuse through the extracellular matrix into the blood stream. While a particle size of less than 150 nm may be desired for effective hepatocyte delivery due to the size of the endothelial fenestrae (see e.g., Akinc et al., Mol Ther. 2009 17:872-879), use of lipidoid oligonucleotides to deliver the formulation to other cells types including, but not limited to, endothelial cells, myeloid cells, and muscle cells may not be similarly size-limited.

[0228] In one aspect, effective delivery to myeloid cells, such as monocytes, lipidoid formulations may have a similar component molar ratio. Different ratios of lipidoids and other components including, but not limited to, a neutral lipid (e.g., diacylphosphatidylcholine), cholesterol, a PEGylated lipid (e.g., PEG-DMPE), and a fatty acid (e.g., an omega-3 fatty acid) may be used to optimize the formulation of the mRNA or system for delivery to different cell types including, but not limited to, hepatocytes, myeloid cells, muscle cells, etc. Exemplary lipidoids include, but are not limited to, DLin-DMA, DLin-K-DMA, DLin-KC2-DMA, 98N12-5, C 12-200 (including variants and derivatives), DLin-MC3-DMA and analogs thereof. The use of lipidoid formulations for the localized delivery of nucleic acids to cells (such as, but not limited to, adipose cells and muscle cells) via either subcutaneous or intramuscular delivery, may also not require all of the formulation components which may be required for systemic delivery, and as such may comprise the lipidoid and the mRNA or system.

[0229] According to the present disclosure, a system described herein may be formulated by mixing the mRNA or system, or individual components of the system, with the lipidoid at a set ratio prior to addition to cells. In vivo formulations may require the addition of extra ingredients to facilitate circulation throughout the body. After formation of the particle, a system or individual components of a system is added and allowed to integrate with the complex. The encapsulation efficiency is determined using a standard dye exclusion assays.

[0230] In vivo delivery of systems may be affected by many parameters, including, but not limited to, the formulation composition, nature of particle PEGylation, degree of loading, oligonucleotide to lipid ratio, and biophysical parameters such as particle size (Akinc et al., Mol Ther. 2009 17:872-879; herein incorporated by reference in its entirety). As an example, small changes in the anchor chain length of poly(ethylene glycol) (PEG) lipids may result in significant effects on in vivo efficacy. Formulations with the different lipidoids, including, but not limited to penta[3-(l-laurylaminopropionyl)]-triethylenetetramine hydrochloride (TETA-5LAP; aka 98N12-5, see Murugaiah et al., Analytical Biochemistry, 401 :61 (2010)), Cl 2-200 (including derivatives and variants), MD1, DLin-DMA, DLin-K-DMA, DLin-KC2- DMA and DLin-MC3-DMA can be tested for in vivo activity. The lipidoid referred to herein as "98N12-5" is disclosed by Akinc et al., Mol Ther. 2009 17:872-879). The lipidoid referred to herein as "C12-200" is disclosed by Love et al., Proc Natl Acad Sci USA. 2010 107: 1864- 1869 and Liu and Huang, Molecular Therapy. 2010 669-670.

[0231] The LNPs of the present disclosure, in which a nucleic acid is entrapped within the lipid portion of the particle and is protected from degradation, can be formed by any method known in the art including, but not limited to, a continuous mixing method, a direct dilution process, and an in-line dilution process. Additional techniques and methods suitable for the preparation of the LNPs described herein include coacervation, microemulsions, supercritical fluid technologies, phase-inversion temperature (PIT) techniques.

[0232] In some embodiments, the LNPs used herein are produced via a continuous mixing method, e.g., a process that includes providing an aqueous solution a nucleic acid described herein in a first reservoir, providing an organic lipid solution in a second reservoir (wherein the lipids present in the organic lipid solution are solubilized in an organic solvent, e.g., a lower alkanol such as ethanol), and mixing the aqueous solution with the organic lipid solution such that the organic lipid solution mixes with the aqueous solution so as to substantially instantaneously produce a lipid vesicle (e.g., liposome) encapsulating the nucleic acid molecule within the lipid vesicle. This process and the apparatus for carrying out this process are known in the art. More information in this regard can be found in, for example, U.S. Patent Publication No. 20040142025. The action of continuously introducing lipid and buffer solutions into a mixing environment, such as in a mixing chamber, causes a continuous dilution of the lipid solution with the buffer solution, thereby producing a lipid vesicle substantially instantaneously upon mixing. By mixing the aqueous solution comprising a nucleic acid molecule with the organic lipid solution, the organic lipid solution undergoes a continuous stepwise dilution in the presence of the buffer solution (e.g., aqueous solution) to produce a nucleic acid-lipid particle.

[0233] In some embodiments, the LNPs used herein are produced via a direct dilution process that includes forming a lipid vesicle (e.g., liposome) solution and immediately and directly introducing the lipid vesicle solution into a collection vessel containing a controlled amount of dilution buffer. In some embodiments, the collection vessel includes one or more elements configured to stir the contents of the collection vessel to facilitate dilution. In some embodiments, the amount of dilution buffer present in the collection vessel is substantially equal to the volume of lipid vesicle solution introduced thereto. [0234] In some embodiments, the LNPs are produced via an in-line dilution process in which a third reservoir containing dilution buffer is fluidly coupled to a second mixing region. In these embodiments, the lipid vesicle (e.g., liposome) solution formed in a first mixing region is immediately and directly mixed with dilution buffer in the second mixing region. These processes and the apparatuses for carrying out direct dilution and in-line dilution processes are known in the art. More information in this regard can be found in, for example, U.S. Patent Publication No. 20070042031.

5.10. Additional delivery modalities

[0235] This disclosure is not limited to systems and methods described herein. Any delivery method that is capable of delivering the systems described herein can be used as long as it is capable of site-specifically integrating a template polynucleotide into the genome of a cell.

5.11. Genes and Targets

[0236] This disclosure provides compositions and co-delivery methods for correcting or replacing genes or gene fragments (including introns or exons) or inserting genes in new locations. In certain embodiments, such a method comprises recombination or integration into a safe harbor site (SHS). A frequently used human SHS is the AAVS1 site on chromosome 19q, initially identified as a site for recurrent adeno-associated virus insertion. Another locus comprises the human homolog of the murine Rosa26 locus. Yet another SHS comprises the human Hl 1 locus on chromosome 22. In some cases, a complete gene may be prohibitively large and replacement of an entire gene impractical. In certain embodiments, a method of the invention comprises recombining corrective gene fragments into a defective locus.

[0237] The methods and compositions can be used to target, without limitation, stem cells for example induced pluripotent stem cells (iPSCs), HSCs, HSPCs, mesenchymal stem cells, or neuronal stem cells and cells at various stages of differentiation. In certain embodiments, methods and compositions of the invention are adapted to target organoids, including patient derived organoids.

[0238] In certain embodiments, methods and compositions of the invention are adapted to treat muscle cells, not limited to cardiomyocytes for Duchene Muscular Dystrophy (DMD). The dystrophin gene is the largest gene in the human genome, spanning ~2.3 Mb of DNA. DMD is composed of 79 exons resulting in a 14-kb full-length mRNA. Common mutations include mutations that disrupt the reading frame of generate a premature stop codon. An aspect of DMD that lends it to gene editing as a therapeutic approach is the modular structure of the dystrophin protein. Redundancy in the central rod domain permits the deletion of internal segments of the gene that may harbor loss-of-function mutations, thereby restoring the open reading frame (ORFs). In some embodiments, the methods and systems described herein are used to treat DMD by site-specifically integrating in the genome a polynucleotide template that repairs or replaces all or a portion of the defective DMD gene.

[0239] The following are non-limiting diseases that may be treated utilizing the methods and compositions of the present disclosure:

Inherited Retinal Diseases:

• Stargardt Disease (ABCA4)

• Leber congenital amaurosis 10 (CEP290)

• X linked Retinitis Pigmentosa (RPGR)

• Autosomal Dominant Retinitis Pigmentosa (RHO)

Liver Diseases:

• Wilson’s disease (ATP7B)

• Alpha-1 antitrypsin (SERPINA1)

Intellectual Disabilities:

• Rett Syndrome (MECP2)

• S YNGAP 1 -ID (S YNGAP 1 )

• CDKL5 deficiency disorder (CDKL5)

Peripheral Neuropathies:

• Charcot-Mari e-Tooth 2 A (MFN2)

Lung Diseases:

• Cystic Fibrosis (CFTR)

• Alpha-1 Antitrypsin (SERPINA1)

Autoimmune diseases:

IgA Nephropathy (Berger’s disease)

Anti-Neutrophil Cytoplasmic Antibody (ANCA) Vasculitis

Systemic Lupus Erythematosus (SLE) / Lupus Nephritis (LN) Membranous Nephropathy (MN)

C3 glomerulonephritis (C3GN)

Blood disorders:

• Sickle Cell

• Hemophilia

• Factor VIII or

• Factor IX

• Ornithine transcarbamylase deficiency (OTCD)

• Homocystinuria (HCU)

• Phenylketonuria (PKU)

Cancer

• Prostate cancer

• Renal cell cancer

• Thyroid cancer

[0240] CFTR (cystic fibrosis transmembrane conductance regulator). The most common cystic fibrosis (CF) mutation F508del removes a single amino acid. In some embodiments, recombining human CFTR into an SHS of a cell that expresses CFTR F508del is a corrective treatment path. In some embodiments, the methods and systems described herein are used to CF by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing CF. Proposed validation is detection of persistent CFTR mRNA and protein expression in transduced cells.

[0241] Sickle cell disease (SCD) is caused by mutation of a specific amino acid — valine to glutamic acid at amino acid position 6. In some embodiments, SCD is corrected by recombination of the HBB gene into a safe harbor site (SHS) and by demonstrating correction in a proportion of target cells that is high enough to produce a substantial benefit. In some embodiments, the methods and systems described herein are used to sickle cell disease by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the disease. In some embodiments, validation is detection of persistent HBB mRNA and protein expression in transduced cells. [0242] DMD — Duchenne Muscular Dystrophy. The dystrophin gene is the largest gene in the human genome, spanning ~2.3 Mb of DNA. DMD is composed of 79 exons resulting in a 14-kb full-length mRNA. Common mutations include mutations that disrupt the reading frame of generate a premature stop codon. An aspect of DMD that lends it to gene editing as a therapeutic approach is the modular structure of the dystrophin protein. Redundancy in the central rod domain permits the deletion of internal segments of the gene that may harbor loss- of-function mutations, thereby restoring the open reading frame (ORFs).

[0243] In some embodiments, recombination will be into safe harbor sites (SHS). A frequently used human SHS is the A4P57 site on chromosome 19q, initially identified as a site for recurrent adeno-associated virus insertion. In some embodiments, the site is the human homolog of the e murine Rosa26 locus (pubmed.ncbi.nlm.nih.gov/18037879). In some embodiments, the site is the human Hl 1 locus on chromosome 22. Proposed target cells for recombination include stem cells for example induced pluripotent stem cells (iPSCs) and cells at various stages of differentiation. In some cases, a complete gene may be prohibitively large and replacement of an entire gene impractical. In such instances, rescuing mutants by recombining in corrected gene fragments with the methods and systems described herein is a corrective option.

[0244] In some embodiments, correcting mutations in exon 44 (or 51) by recombining in a corrective coding sequence downstream of exon 43 (or 50), using the methods and systems described herein is a corrective option. Proposed validation is detection of persistent DMD mRNA and protein expression in transduced cells.

[0245] F8 (Factor VIII). A large proportion of severe hemophilia A patients harbor one of two types of chromosomal inversions in the FVIII gene. The recombinase technology and methods described herein are well suited to correcting such inversions (and other mutations) by recombining of the FVIII gene into a SHS.

[0246] In some embodiments, correcting factor VIII deficiency by recombining the FVIII gene into an SHS is a corrective path. In some embodiments, the methods and systems described herein are used to correct factor VIII deficiency by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the FIX deficiency. Proposed validation is detection of persistent FVIII mRNA and protein expression in transduced cells. [0247] Factor 9 (Factor IX) Hemophilia B, also called factor IX (FIX) deficiency is a genetic disorder caused by missing or defective factor IX, a clotting protein.

[0248] In some embodiments, the methods and systems described herein are used to correct factor IX deficiency by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the FIX deficiency. Proposed validation is detection of persistent FiX mRNA and protein expression in transduced cells.

[0249] Ornithine transcarbamylase deficiency (OTCD). Ornithine transcarbamylase deficiency is a rare genetic condition that causes ammonia to build up in the blood. The condition - more commonly called OTC deficiency - is more common in boys than girls and tends to be more severe when symptoms emerge shortly after birth.

[0250] In some embodiments, the methods and systems described herein are used to correct OTC deficiency by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the OTC deficiency or integrates a polynucleotide encoding a functional ornithine transcarbamylase enzyme. Proposed validation is detection of persistent OTC mRNA and protein expression in transduced cells.

[0251] Phenylketonuria, also called PKU, is a rare inherited disorder that causes an amino acid called phenylalanine to build up in the body. PKU is caused by a change in the phenylalanine hydroxylase (PAH) gene. This gene helps create the enzyme needed to break down phenylalanine.

[0252] In some embodiments, the methods and systems described herein are used to correct PKU by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the PKU deficiency or integrates a polynucleotide encoding a functional phenylalanine hydroxylase (PAH) gene. Proposed validation is detection of persistent PAH mRNA and protein expression in transduced cells.

[0253] Homocystinuria (HCU). Homocystinuria is elevation of the amino acid, homocysteine (protein building block coming from our diet) in the urine or blood. Common causes of HCU include: problems with the enzyme cystathionine beta synthase (CBS), which converts homocysteine to the amino acid cystathionine (which then becomes cysteine) and needs the vitamin B6 (pyridoxine); and problems with converting homocysteine to the amino acid methionine. [0254] In some embodiments, the methods and systems described herein are used to correct HCU by site-specifically integrating in the genome a polynucleotide template that corrects the mutation causing the HCU or integrates a polynucleotide encoding a functional copy of a gene (e.g., CBS) able to reduce or prevent buildup of homocysteine in the urine. Proposed validation is detection of persistent CBS mRNA and protein expression in transduced cells.

[0255] IgA Nephropathy (Berger’s disease). IgA nephropathy, also known as Berger's disease, is a kidney/autoimmune disease that occurs when an antibody called immunoglobulin A (IgA) builds up in the kidneys.

[0256] In some embodiments, the methods and systems described herein are used to treat Berger’s disease by administering to a patient an iPSC-derived Natural Killer cell that includes a polynucleotide site-specifically integrated in the genome of the cell using the methods described herein. Upon administering the iPSC-NK cell to the patient, the iPSC-NK cell is capable of removing native cells (e.g., B cells) that are responsible, at least in part, for the symptoms of Berger’s disease.

[0257] Anti-Neutrophil Cytoplasmic Antibody (ANCA) Vasculitis. ANCA vasculitis is an autoimmune disease affecting small blood vessels in the body. It is caused by autoantibodies called ANCAs, or Anti-Neutrophilic Cytoplasmic Autoantibodies. ANCAs target and attack a certain kind of white blood cells called neutrophils.

[0258] In some embodiments, the methods and systems described herein are used to treat ANCA vasculitis by administering to a patient an iPSC-derived Natural Killer cell that includes a polynucleotide site-specifically integrated in the genome of the cell using the methods described herein. Upon administering the iPSC-NK cell to the patient, the iPSC-NK cell is capable of removing native cells (e.g., B cells) that are responsible, at least in part, for the symptoms of ANCA vasculitis.

[0259] Systemic Lupus Erythematosus (SLE) / Lupus Nephritis (LN). Lupus is an autoimmune — a disorder in which the body’s immune system attacks the body’s own cells and organs.

[0260] In some embodiments, the methods and systems described herein are used to treat SLE/LN by administering to a patient an iPSC-derived Natural Killer cell that includes a polynucleotide site-specifically integrated in the genome of the cell using the methods described herein. Upon administering the iPSC-NK cell to the patient, the iPSC-NK cell is capable of removing native cells (e.g., B cells) that are responsible, at least in part, for the symptoms of SLE/LN.

[0261] Membranous Nephropathy (MN). MN is a kidney disease that affects the filters (glomeruli) of the kidney and can cause protein in the urine, as well as decreased kidney function and swelling. It can sometimes be called membranous glomerulopathy as well (these terms can be used interchangeably and mean the same thing).

[0262] In some embodiments, the methods and systems described herein are used to treat MN by administering to a patient an iPSC-derived Natural Killer cell that includes a polynucleotide site-specifically integrated in the genome of the cell using the methods described herein. Upon administering the iPSC-NK cell to the patient, the iPSC-NK cell is capable of removing native cells (e.g., B cells) that are responsible, at least in part, for the symptoms of MN.

[0263] C3 glomerulonephritis (C3GN). C3 glomerulopathy is a group of related conditions that cause the kidneys to malfunction. The major features of C3 glomerulopathy include high levels of protein in the urine (proteinuria), blood in the urine (hematuria), reduced amounts of urine, low levels of protein in the blood, and swelling in many areas of the body. Affected individuals may have particularly low levels of a protein called complement component 3 (or C3) in the blood.

[0264] In some embodiments, the methods and systems described herein are used to treat C3 glomerulopathy by administering to a patient an iPSC-derived Natural Killer cell that includes a polynucleotide site-specifically integrated in the genome of the cell using the methods described herein. Upon administering the iPSC-NK cell to the patient, the iPSC-NK cell is capable of removing native cells (e.g., B cells) that are responsible, at least in part, for the symptoms of C3 glomerulopathy.

[0265] Glucagon-like peptide 1 (GLP-1). Glucagon-like peptide 1 (GLP-1) is a small peptide component of the prohormone, proglucagon, that is produced in the gut. In some embodiments, the methods and systems described herein include administering, and thereby site-specifically integrating, a polynucleotide encoding GLP-1 or a GLP-1 agonist. 5.12. Methods of treatment

[0266] In another aspect, methods of treatment are presented. The method comprises administering an effective amount of the pharmaceutical composition comprising the nucleic acid construct or vectorized nucleic acid construct described above to a patient in need thereof. In some embodiments, the system (e.g., any of the systems described herein) are delivered to a cell ex vivo and the cell is then administered to the subject. In some embodiments, the systems (e.g., any of the systems described herein) are delivered to a patient, thereby delivering to a cell in vivo.

[0267] DNA or RNA viral vectors can be administered directly to patients (in vivo) or they can be used to treat cells in vitro, and the modified cells may optionally be administered to patients ex vivo). Conventional viral based systems to be used herein could include retroviral, lentivirus, adenoviral, adeno-associated and herpes simplex virus vectors for gene transfer. Integration in the host genome is possible with the retrovirus, lentivirus, and adeno- associated virus gene transfer methods, often resulting in long term expression of the inserted transgene. Additionally, high transduction efficiencies have been observed in many different cell types and target tissues.

[0268] In some embodiments, the co-delivery system described herein (e.g., a gene editor construct packaged in a LNP and a donor template packaged in a vector) is administered intravenously. In some embodiments, the co-delivery system described herein (e.g., a gene editor construct packaged in a LNP and a donor template packaged in a vector) is administered intrathecally. In some embodiments, the co-delivery system described herein (e.g., a gene editor construct packaged in a LNP and a donor template packaged in a vector) is administered by intracerebral ventricular injection. In some embodiments, the co-delivery system described herein (e.g., a gene editor construct packaged in a LNP and a donor template packaged in a vector) is administered by intracistemal magna administration. In some embodiments, the co-delivery system described herein (e.g., a gene editor construct packaged in a LNP and a donor template packaged in a vector) is administered by intravitreal injection.

[0269] Methods of non-viral delivery of the donor DNA template described herein include lipofection, nucleofection, microinjection, biolistics, virosomes, liposomes, immunoliposomes, polycation or lipidmucleic acid conjugates, naked DNA, artificial virions, and agent-enhanced uptake of DNA. Lipofection is described in e.g., U.S. Pat. Nos. 5,049,386, 4,946,787; and 4,897,355) and lipofection reagents are sold commercially (e.g., Transfectam™ and Lipofectin™). Cationic and neutral lipids that are suitable for efficient receptor-recognition lipofection of polynucleotides include those of Feigner, WO 91/17424; WO 91/16024. Delivery can be to cells (e.g. in vitro or ex vivo administration) or target tissues (e.g. in vivo administration).

5.12.1. mRNA delivery

[0270] Another useful method to deliver proteins, enzymes, and guides comprises transfection of messenger RNA (mRNA). Examples of mRNA delivery methods and compositions that may be utilized in the present disclosure including, for example, PCT/US2014/028330, US8822663B2, NZ700688A, ES2740248T3, EP2755693A4, EP2755986A4, WO2014152940A1, EP3450553B1, BRI 12016030852A2, and EP3362461A1. Expression of CRISPR systems in particular is described by W02020014577. Each of these publications are incorporated herein by reference in their entireties. Additional disclosure hereby incorporated by reference can be found in Kowalski et al., “Delivering the Messenger: Advances in Technologies for Therapeutic mRNA Delivery,” Mol Therap., 2019; 27(4): 710-728.

5.13. Discovery and validation of off-target editing

[0271] In another aspect, methods for identifying and validating off-target editing are provided.

5.13.1. Cryptic-Seq

[0272] In some embodiments, the method for identifying off-target editing is Cryptic site sequencing (Cryptic-Seq). Cryptic-Seq can be used to identify off-target editing of integration/recombination enzymes that recognize an attachment site, such as sequence specific integrase, transposase, insertion element, resolvase, invertase, or recombinase. In some embodiments, the integrase is an integrase of the serine or tyrosine family. In some embodiments, the integrase is a large serine integrase (LSI). In some embodiments, the transposase is a Mu transposase. In some embodiments, the recombinase is a recombinase of the serine or tyrosine family. In some embodiments, the recombinase is a RAG family recombinase. [0273] In some embodiments, Cryptic-Seq is used to identify the off-target editing of a large serine integrase. In certain embodiments, the large serine integrase is an integration enzyme disclosed in PCT Publication No. W02023/070031, the disclosure of which is incorporated by reference in its entirety.

[0274] In some embodiments, Cryptic-Seq is used with one large serine integrase (LSI) multiplexed with attachment sites that differ only in the central dinucleotide sequence. In some embodiments, Cryptic-Seq is used with multiple large serine integrases (LSIs) simultaneously with each of the LSI’s specific attachment site DNA substrates.

[0275] In some embodiments, provided herein is a method for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) in a target genome, the method comprising: (a) fragmenting genomic DNA isolated from target cells; (b) tagging the 5’ and 3’ ends of the DNA fragments with a first oligonucleotide adapter; (c) contacting the tagged genomic DNA fragments with (i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, and a second oligonucleotide adapter; (ii) optionally at least a second species of synthetic DNA molecule, wherein the second species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence in the central dinucleotide from the first attachment site of the first species of synthetic DNA molecule, and a second oligonucleotide adapter; and (iii) the selected LSI, under conditions suitable for the LSI to effect the recombination of the synthetic DNA molecule into the tagged genomic DNA fragments at second attachment sites that are capable of functioning as cognates of the first attachment site; (d) generating a sequencing library comprising synthetic DNA molecules recombined with the tagged genomic DNA fragments; (e) sequencing the sequencing library; and (f) identifying cryptic attachment sites based on the sequencing data.

[0276] In some embodiments, the first oligonucleotide adapter comprises a primer binding site. In some embodiments, the second oligonucleotide adapter comprises a primer binding site. In some embodiments, the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating PCR amplification. In some embodiments, the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating sequencing-by-synthesis. In some embodiments, the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating PCR amplification and sequencing-by-synthesis.

[0277] In some embodiments, the synthetic DNA molecule further comprises a unique molecular identifier (UMI). In some embodiments, the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI. In some embodiments, the UMI is positioned in the synthetic DNA molecule between the first attachment site and the second oligonucleotide adapter.

[0278] In some embodiments, the synthetic DNA molecule further comprises a barcode. In some embodiments, the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual-indexed barcode.

[0279] In some embodiments, the synthetic DNA molecule is circular. In some embodiments, the synthetic DNA molecule is linear.

[0280] In some embodiments, the large serine integrase is Bxbl. In some embodiments, the first attachment site is an attP or a modified attP site. In some embodiments, the first attachment site is an attB or a modified attB site.

[0281] In some embodiments, generating a sequencing library comprises an initial step of amplifying the tagged genomic DNA fragment after recombination with the synthetic DNA molecule.

[0282] In some embodiments, identifying a cryptic attachment site comprises determining the sequence of the genomic site into which the DNA molecule has been recombined or the sequence of DNA flanking the cryptic attachment site. In some embodiments, identifying a cryptic attachment site comprises detecting an attL-genomic DNA junction, an attR-genomic DNA junction, or both. In some embodiments, identifying a cryptic attachment site comprises: aligning the sequencing data to a reference genome; detecting an attachment site- genomic DNA junction; and reporting the coordinates and using the coordinates to identify cryptic attachment site. In some embodiments, detecting an attachment site-genomic DNA junction comprises detecting an attL-genomic DNA junction, an attR-genomic DNA junction, or both. In some embodiments, reporting the coordinates comprises ranking the coordinates based on sequencing reads de-duplicated from PCR amplification using the UMIs. In some embodiments, the UMI count represents the recombination efficacy of the first attachment site.

[0283] In some embodiments, the identified cryptic attachment site was not previously known to be an attachment site of the selected LSI.

[0284] In some embodiments, provided herein is a multiplexed method for identifying cryptic attachment sites that are separately recognizable by each of a plurality of large serine integrases (LSIs) in a target genome, the method comprising: (a) fragmenting genomic DNA isolated from target cells; (b) tagging the 5’ and 3’ ends of the DNA fragments with a first oligonucleotide adapter; (c) contacting the tagged genomic DNA fragments with (i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule comprises a first attachment site known to be recognized by a first LSI, and a second oligonucleotide adapter; (ii) at least a second species of synthetic DNA molecule, wherein the second species of synthetic DNA molecule comprises a first attachment site known to be recognized by a second LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence from the first attachment site of the first species of synthetic DNA molecule, and a second oligonucleotide adapter; and (iii) the first LSI and the at least second LSI, under conditions suitable for the LSIs to effect the recombination of the synthetic DNA molecule into the tagged genomic DNA fragments at second attachment sites that are capable of functioning respectively as cognates of the respective first attachment sites; (d) generating a sequencing library comprising synthetic DNA molecules recombined with the tagged genomic DNA fragments; (e) sequencing the sequencing library; and (f) identifying cryptic attachment sites for each of the plurality of LSIs based on the sequencing data.

[0285] The present disclosure also provides a synthetic DNA molecule for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) in a target genome, comprising: a first attachment site and an oligonucleotide adapter, wherein the first attachment site is known to be recognized by the selected LSI.

[0286] In some embodiments, the oligonucleotide adapter comprises a primer binding site. In some embodiments, the primer binding site is capable of mediating PCR amplification. In some embodiments, the primer binding site is capable of mediating sequencing-by-synthesis. In some embodiments, the primer binding site is capable of mediating PCR amplification and sequencing-by-synthesis.

[0287] In some embodiments, the synthetic DNA molecule further comprises a unique molecular identifier (UMI). In some embodiments, the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI. In some embodiments, the UMI is positioned in the synthetic DNA molecule between the first attachment site and the oligonucleotide adapter.

[0288] In some embodiments, the synthetic DNA molecule further comprises a barcode. In some embodiments, the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual-indexed barcode.

[0289] In some embodiments, the synthetic DNA molecule is circular. In some embodiments, the synthetic DNA molecule is linear.

[0290] In some embodiments, the large serine integrase is Bxbl. In some embodiments, the first integration recognition site is an attP or a modified attP site. In some embodiments, the first integration recognition site is an attB or a modified attB site.

[0291] Also provided herein is a set of synthetic DNA molecules for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI), comprising: a first species of synthetic DNA molecule, comprising a first attachment site known to be recognized by the selected LSI, an oligonucleotide adapter, and a first DNA barcode; and at least a second species of synthetic DNA molecule, comprising a first attachment site known to be recognized by the selected LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence in the central dinucleotide from the first attachment site of the first species of synthetic DNA molecule, an oligonucleotide adapter, and a second DNA barcode; wherein the first DNA barcode and the second DNA barcode allow for multiplexing of multiple first attachment sites of different sequences in one reaction.

[0292] In some embodiments, the oligonucleotide adapter comprises a primer binding site.

In some embodiments, the primer binding site is capable of mediating PCR amplification. In some embodiments, the primer binding site is capable of mediating sequencing-by-synthesis. In some embodiments, the primer binding site is capable of mediating PCR amplification and sequencing-by-synthesis.

[0293] In some embodiments, the synthetic DNA molecule further comprises a unique molecular identifier (UMI). In some embodiments, the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI. In some embodiments, the UMI is positioned in the synthetic DNA molecule between the first attachment site and the oligonucleotide adapter.

[0294] In some embodiments, the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual-indexed barcode.

[0295] In some embodiments, the synthetic DNA molecule is circular. In some embodiments, the synthetic DNA molecule is linear.

[0296] In some embodiments, the large serine integrase is Bxbl. In some embodiments, the first integration recognition site is an attP or a modified attP site. In some embodiments, the first integration recognition site is an attB or a modified attB site.

[0297] The present disclosure also provides a set of synthetic DNA molecules for identifying cryptic attachment sites that are separately recognizable by each of a plurality of large serine integrases (LSIs), comprising: a first species of synthetic DNA molecule, comprising a first attachment site known to be recognized by a first LSI, an oligonucleotide adapter, and a first DNA barcode; and at least a second species of synthetic DNA molecule, comprising a first attachment site known to be recognized by a second LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence from the first attachment site of the first species of synthetic DNA molecule, an oligonucleotide adapter, and a second DNA barcode; wherein the first DNA barcode and the second DNA barcode allow for identifying the respective cryptic attachment sites of each of the plurality of LSIs in one reaction.

[0298] Also provided herein are kits for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) or by each of a plurality of large serine integrases (LSIs). The kit comprises the synthetic DNA molecule of the present disclosure and instructions for performing the method of the present disclosure. 5.13.2. HIDE-Seq

[0299] In some embodiments, the method for identifying off-target editing is High- throughput Integrase-mediated DNA Event Sequencing (HIDE-Seq). HIDE-Seq can identify loci of recombination with DNA substrates, loci of recombination without substrates, loci with double-strand break formation from abortive recombination events, and loci with intra- genomic recombination independent of the synthetic DNA substrates.

[0300] In some embodiments, provided herein is a method for identifying DNA recombination events of a selected integrase in a target genome, the method comprising: (a) contacting genomic DNA isolated from target cells with (i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule is linear and comprises a first attachment site known to be recognized by the selected integrase, and (ii) the selected integrase, under conditions suitable for the integrase to effect the recombination of the genomic DNA; (b) generating a sequencing library comprising the recombined genomic DNA; (c) sequencing the sequencing library; and (d) detecting DNA recombination events based on the sequencing data.

[0301] In some embodiments, the DNA recombination events are selected from: a DNA double strand break, a recombination of the attachment site with a cryptic site in the genomic DNA, and a recombination between two cryptic sites in the genomic DNA.

[0302] In some embodiments, the integrase is a large serine integrase (LSI). In some embodiments, the large serine integrase is Bxbl. In some embodiments, the first integration recognition site is an attP or a modified attP site. In some embodiments, the first integration recognition site is an attB or a modified attB site.

[0303] In some embodiments, generating a sequencing library does not comprise amplifying the genomic DNA after recombination of the synthetic DNA molecule. In some embodiments, generating a sequencing library comprises generating a whole genome sequencing library. In some embodiments, sequencing the sequencing library comprises whole genome sequencing.

[0304] In some embodiments, detecting an DNA recombination event comprising: aligning the sequencing data to a reference genome; and detecting a DNA double strand break, a recombination of the integration recognition site with a cryptic site in the genomic DNA, or a recombination between two cryptic sites in the genomic DNA.

[0305] Also provided herein is a synthetic DNA molecule for identifying DNA recombination events of a selected integrase in a target genome, comprising: a first attachment site, wherein the first attachment site is known to be recognized by the selected integrase, wherein the synthetic DNA molecule is linear.

[0306] In some embodiments, the DNA recombination events are selected from: a DNA double strand break, a recombination of the attachment site with a cryptic site in the genomic DNA, and a recombination between two cryptic sites in the genomic DNA.

[0307] In some embodiments, the integrase is a large serine integrase (LSI). In some embodiments, the large serine integrase is Bxbl. In some embodiments, the first integration recognition site is an attP or a modified attP site. In some embodiments, the first integration recognition site is an attB or a modified attB site.

[0308] Also provided herein is a kit for identifying DNA recombination events of a selected integrase, comprising: (a) the synthetic DNA molecule of the present disclosure; and (b) instructions for performing the method of the present disclosure.

5.13.3. Hybrid Capture

[0309] In some embodiments, the method for validating off-target editing is hybridization capture followed deep sequencing of capture targets by with NGS (hybrid capture NGS). Hybrid capture NGS can be used to determine the risk of off-target editing across the potential ‘landscape’ of cryptic attachment sequences in the human genome (FIG. 7). Hybridization capture followed by next generation sequencing (hybrid capture NGS) is a target enrichment approach for comprehensive and quantitative validation of off-target sites in edited genomic DNA isolated from edited cells. For assessment of potential off-target sites due to LSI editing, such as discovered sites in the human genome from the discovery assays such as HIDE-seq, and Cryptic-seq described herein, in some embodiments, probes can be designed and ordered from vendors such as IDT or Twist Biosciences by submitting hg38 genome coordinates for the registered dinucleotide breakpoint and allow for probe design directly 5’ or 3’ adjacent to the breakpoint and distances up to 180 nucleotides upstream (5’) or downstream (3’) of the LSI site breakpoint. In some embodiments, the probe is designed to be adjacent, but not overlapping, with the LSI dinucleotide breakpoint in the off-target site to be queried. This adjacent placement of probes allows direct relative quantification of unedited and edited read frequencies at a given locus, with accuracy near on par with ddPCR (due to UMI deduplication) and qualifiable to a limit of detection of 0.1% and observed detection of edits down to 0.001%. In some embodiments, two sets of hybrid capture validation probes can be designed to enable comprehensive validation of potential off-target editing; probe set 1 hybridizes 5’ (i.e., upstream) of the potential off-target site and probe set 2 hybridizes 3’ (i.e., downstream) of the potential off-target site which allows the detection of off-target insertion, intra- or inter-chromosomal translocations and indels. Derisking of any validated off-target involves assessing gene function and cancer relevance, guiding decisions on therapeutic suitability.

5.13.4. CHASER-Seq

[0310] In some embodiments, the method for identifying off-target editing is Circularization for High-throughput Analysis of Large SErine Recombinase by Sequencing (CHASER-Seq) (FIG. 8). This technology is an amplification-based method that takes advantage of the sequencing of double stranded DNA circles generated from ligated fragments of genomic DNA that linearize upon recombination with the unique attachment sequences in combination with PCR amplification of linearized circles followed by next genome sequencing to discover potential LSI off-target sites. When genomic circularized DNA is exposed to LSI protein and linear DNA attachment substrates, a double strand break is produced with the attachment sequences appended at the break points, while un-recombined DNA circles do not proceed forward in the library prep and are thus not sequenced. In addition to the discovery of sites with potential for off-target integration, CHASER-Seq can also detect double-strand DNA breaks.

5.13.5. Att-Seq

[0311] In some embodiments, the method for identifying off-target editing is Attachment site sequencing (Att-Seq) (FIG. 9). Att-Seq is a targeted but amplification-free sequencing method that takes advantage of direct DNA sequencing of small attP and attB containing double stranded oligonucleotides post-recombination by Bxb 1 integrase in a biochemical reaction. In some embodiments, Att-Seq is performed with known attB and attP oligonucleotide substrates. In some embodiments, Att-Seq provides quantitative information on site recombination preferences. In some embodiments, Att-Seq detects rare events, such as DSBs or antiparallel recombination that can occur with mismatched attP and attB substrates. In some embodiments, multiple attP and or attB substrates of known quantity and sequence are inputted to survey the recombination potential in the presence of LSI or mutated variants of an LSI (e.g., high fidelity LSI protein are compared to WT LSI proteins in the ability to recombine multiple off-target attB substrates with a wildtype attP substrate).

6. EXAMPLES

6.1. Example 1: Cryptic-Seq

[0312] Genomic DNA from internally developed cell line HEK Clone 12 was isolated using the Qiagen Blood and Tissue Kit. The gDNA was incubated with Tn5 transposase (custom recombinant protein purification by Genscript, pre-annealed with MES_Rev_3InvdR and MES_AmpSeq_P5 oligos) for 7 minutes at 55°C to tagment the gDNA and shear it to an average size of around 700bp.

[0313] Six cargo donor constructs were designed to include integrase attachment sites with 6 different central dinucleotides (GT (PL2312), AC (PL2328), CT (PL2334), AG (PL2329), CA (PL2331), TG (PL2340) and Illumina sequencing primer binding sites N7 and P7. The 6 donor constructs were normalized to the same concentration and pooled. The DNA sequences used in the experiment are shown in Table 12 below. Table 12 [0314] The biochemical reaction was performed in a buffer consisting of lOmM Tris HCl, lOOmM KC1, and 5% glycerol. 1 pg tagmented gDNA, InM of the pooled donor constructs, and I M integrase were combined and incubated at 37°C for 4h. The reaction was stopped by adding SDS to a final concentration of 0.1% and the products were cleaned using the Zymo clean and concentrator kit (Zymo research).

[0315] PCR was used to amplify the regions of integration using Q5 polymerase and primers for P5 containing distinct barcodes for each sample (LM P5 F1-F6) and a single N7 primer (GN037_CrypticSeq_N7_i7_N701). Following PCR, a 1.5x bead clean-up was performed and products were run on tape station using DI 000 tape to confirm amplification. Library concentration was determined using the NEB Next Library Quant Kit for Illumina. Libraries were normalized to 2nM, loaded at a concentration of 750pM and sequenced via Illumina Next Seq according to the manufacturer's instructions.

[0316] For bioinformatic analysis of cryptic-seq data, FASTQ files from Illumina sequencing were loaded into a custom cryptic-seq bioinformatic pipeline to discover and quantify cryptic recombination sites from Cryptic-seq data. The bioinformatic workflow for Cryptic-seq is outlined in FIG. 5 which starts with the trimming of reads that contain leading attP or attB sequences (P,P’ or B,B’) followed by a search and trim of any Tn5 mosaic end (ME) sequence (CTGTCTCTTATACACATCT (SEQ ID NO: 580)) in the reads. Reads trimmed for att and ME sequences were aligned against the hg38 human reference genome with appended attP and attB sequences by BWA aligner to create mapped BAM files for each sample. The resulting BAM files were deduplicated by Picard and queried for integration sites with to generate output .csv files containing sites per sample along with collation of sites across samples to generate output .csv files containing collated sites.

[0317] As shown in FIG. 10, 920 potential cryptic attB sites were discovered for Bxbl integrase with Cryptic-Seq.

6.2. Example 2: HIDE-seq

[0318] Recombination reactions with 8 micrograms of purified human genomic DNA (gDNA) from PBMCs (Qiagen Blood and Tissue kit) or, in separate reactions, internal HEK293 cell line C12 that contains an Bxbl integrase attB site in an integrated lentiviral construct on chromosome 5 was incubated with 10 nM attB_66bp annealed oligonucleotide (for cryptic attP site or DSB discovery) or 10 nM attP_72bp annealed oligonucleotide (for cryptic attB site or DSB discovery) and 400 nM of recombinant Bxbl integrase (Genscript) in recombination buffer (10 nM Tris-HCl, pH 8.0, 100 mM KC1, 5% glycerol from) for 8 hours at 37°C and then held at 4°C. To anneal the attB_66 and attP_72 oligonucleotide substrates for HIDE-seq, 25 pl of attB_66bp_P_Top (lOOpM) and 25 pl of attB_66bp_P_Bot (lOOpM) or 25 pl of attP_72bp_P_Top (lOOpM) and 25 pl of attP_72bp_P_Bot (lOOpM), respectively were mixed together in wells of a 20 pl 96 well plate. Oligonucleotides were resuspended in IDT duplex buffer (IDT). The plate with oligonucleotide samples were placed into a thermocycler, and the following anneal program was performed with a 105°C heated lid:

95°C, 2 minutes

750 cycles:

-0.1 °C per cycle, Is

4°C hold

[0319] Reactions were inactivated with RNAase and then Proteinase K and genomic DNA was purified with 0.7x Ampure bead cleanup. Isolated gDNA was quantified by qubit and submitted for PCR free whole genome Illumina short read sequencing at 3 Ox coverage (Azenta). FASTQ files from Illumina sequencing were loaded into a custom bioinformatic pipeline called tbDigin which quantified the number of unclipped reads (for DNA double strand break detection) or soft-clipped reads (recombined sites). The bioinformatic workflow for HIDE-seq is outlined in FIG. 6 and starts with the alignment of sequencing reads from FASTQ files against the hg38 human reference genome with appended attP and attB sequences by BWA aligner to create mapped BAM files for each sample. The resulting BAM files are deduplicated by Picard and inputted into HIDE pipeline (github.com/editasmedicine/digenomitas) to output .csv files containing sites per sample as clipped counts for integration events and unclipped counts for DSBs.

[0320] The results from HIDE-Seq showed that Bxbl integrase generated 0 recurrent DSBs across samples with total levels of singleton DSBs the same as untreated DNA (FIG. 11). In contrast, in-line control CRISPR/Cas9 Digenome-seq reactions with a promiscuous VEGFA s2 sgRNA yielded more than 11,000 recurrent DSBs. See Kim, D., et al., (2015) Nat Methods 12, 237-243 and Tsai, S.Q., et al., (2015) Nat Biotechnol 33, 187-197, each of which is incorporated by reference in its entirety.

[0321] As shown in FIG. 12, 35 potential cryptic attB sites were discovered for Bxbl integrase with HIDE-Seq. High on-target integration frequencies (100%) with lower cryptic integration frequencies (<10%) was observed.

6.3. Example 3: Hybrid Capture

[0322] For hybrid capture NGS library preparation, cell pellets from iPSCs (induced pluripotent stem cells harboring an on-target Bxbl integrase attB sequence at the B2M gene (Tome iPSC clone 52) that were edited by electroporation of Bxbl integrase mRNA and plasmid DNA cargo containing an Bxbl integrase attP sequence (EXP23001648) were treated with the Qiagen Blood and Tissue DNA extraction kit to isolate gDNA. 500 ng of gDNA from edited and unedited control samples were sheared with a Covaris ME220 sonicator to average fragment lengths of 550 bp for a longer fragment experiment or 350 bp for a separate shorter experiment to assess the effect of DNA fragment length on LSI off- target detection in hybrid capture. The Xgen library prep and hybridization protocol was followed according to the manufacturer’s instructions (IDT). Library concentration was determined by Qubit Fluorometry (ThermoFisher) and sequenced on an Illumina Next Seq 2000 with a P3 600 cycle kit according to the manufacturer's instructions.

[0323] For bioinformatic analysis of hybrid capture data, FASTQ files from Illumina sequencing were loaded into a custom bioinformatic pipeline called tbHCA (Tome Bio Hybrid Capture Analysis, github.com/tomebio/tbHCA) developed to quantify edited reads corresponding to unedited reads, reads containing indels, reads containing cargo DNA sequence and thus represent recombined reads, or structural variants such as translocations. Briefly, input data in FASTQ or BAM format underwent quality control using FASTP, which included adapter trimming, quality filtering, and extraction of unique molecular identifiers (UMIs). Additionally, paired-end reads were merged into a single file. The processed reads were aligned to the hg38 reference genome using BWA aligner. UMIs were deduplicated from the aligned BAM files to remove PCR duplicates. Target information was collected and used to generate reference amplicons. Target reads were extracted from the deduplicated BAM files and aligned to the reference amplicons using BWA aligner. Editing events were quantified using a custom Python script designed to analyze the aligned target reads. Visualization of editing events was performed using a custom script to generate graphical representations. Editing sites were collated across samples are outputted into an excel spreadsheet using a custom script. An HTML report was generated using papermill (papermill. readthedocs.io/en/latest/index.html) to summarize the results of the analysis.

[0324] Two separate upstream and downstream hybrid capture probes for each of 465 highest ranked cryptic attB sites discovered by HIDE-Seq and Cryptic-Seq were synthesized (FIG. 13). The 465 sites were surveyed with average read count of 65,000 reads on target with chimeric reads containing full P or P’ half-sites called as cryptic integration events with frequencies ranging from 0.008 to 0.617%, with no indels observed at off-target sites. 90 cryptic sites with less than 1% frequency in cells were validated with hybrid capture. WGS at 60x coverage only detected 2 out of 90 of the discovered cryptic sites, indicating at least an about 45 times increase in sensitivity of hybrid capture for detection rare or low frequency cryptic sites (FIG. 14).

6.4. Example 4: Large Serine Integrase Off-Target Discovery with Deep Learning for Genome Wide Prediction

[0325] Large Serine Integrases (LSIs) hold significant therapeutic promise due to their ability to efficiently incorporate gene-sized DNA into the human genome, offering a method to integrate healthy genes in patients with monogenic disorders or to insert gene circuits for the development of advanced cell therapies. To advance the application of LSIs for human therapeutic applications, new technologies and analytical methods for predicting and characterizing off-target recombination by LSIs are required. It is not experimentally tractable to validate off-target editing at all potential off-target sites in therapeutically relevant cell types because of sample limitations and genetic variation in the human population. To address this gap, we constructed a deep learning model named IntQuery that can predict LSI activity genome-wide. For Bxbl integrase, IntQuery was trained on quantitative off-target data from 410,776 cryptic attB sequences discovered by Cryptic-seq, an unbiased in vitro discovery technology for LSI off-target recombination. We show that IntQuery can accurately predict in vitro LSI activity, providing a tool for in silico off-target prediction of large serine integrases to advance therapeutic applications.

[0326] The large serine integrase (LSI) family constitutes a diverse group of site-specific recombinases that play pivotal roles in mediating DNA rearrangements¹'³. Serine integrases, in contrast to their tyrosine recombinase counterparts, utilize a serine residue for catalysis, leading to distinct mechanistic features4. This large family encompasses integrases with varying sizes and functionalities, with notable members including PhiC31 integrase from Streptomyces bacteriophage PhiC31⁵ and Bxbl integrase discovered in my cobacteriophage Bxbl⁶. Both PhiC31 and Bxbl integrases are well-recognized for their utility in site-specific recombination applications by virtue of direct recombination between phage attachment site attP and bacterial attachment site attB with the requirement of no co-factors or DNA supercoiling⁷'⁹. The precise and efficient DNA manipulation capabilities of the large serine integrase family have positioned it as an attractive tool for synthetic biology and genome editing applicationslO-14.

[0327] Large serine integrases facilitate recombination between attachment sites on linear or circular DNA substrates^{5, 15}. Recently CRISPR-directed integrase editing strategies have been described^{11, 14} that enable programmable genomic insertion of DNA cargo to facilitate gene replacement strategies¹⁶. Unlike CRISPR/Cas9 knock-in approaches, which rely on the host cell’s DNA repair mechanisms¹⁷'¹⁹, LSI integration operates independently of host cell factors and does not require double-stranded breaks (DSBs). Instead, it directly integrates the template DNA, resulting in significantly higher fidelity compared to other error-prone gene writing methods^{5, 15}. This approach to gene insertion has the advantage of minimizing unintended editing at the on-target locus, but it still presents the risk of potential off-target insertion and gross chromosomal rearrangements related to LSI-mediated recombination at ‘cryptic’ or ‘pseudo’ attachment sequences that may be present in the human genome²⁰'²².

[0328] The rapid development of new medicines driven by the expansion of genome editing technologies²³ has resulted in approved cures for sickle cell disease using ex vivo products^{24, 25}. Additionally, promising clinical data has emerged for in vivo genome editing, where lipid nanoparticles encapsulating Cas9 mRNA and guide RNA are administered ^{svslcmcallv26}- ²⁷. [_n response to these exciting advances in genetic medicines, the FDA has released non-binding recommendations for the assessment of safety, including potential genotoxicity from off- target editing events²⁸. Nonclinical safety studies designed to discover potential risks should use multiple methods (e.g., in silico, biochemical and cellular-based assays) that include a genome-wide analysis to reduce bias in identification of potential off-target sites²⁸.

[0329] To help safely advance CRISPR-directed integrases for clinical trials, we previously described two empirical genome-wide discovery technologies, HIDE-seq and Cryptic-seq²⁰, to enable unbiased off-target discovery in isolated human genomic DNA (gDNA) samples. However, an effective rank-ordering in silico prediction tool for LSI off-target recombination has not been described. To address this gap, we constructed a deep learning model named IntQuery, using the empirical off-target discovery data from Cryptic-seq for the LSI Bxbl integrase across multiple central dinucleotide targets. The quantitative nature of Cryptic-seq data from 410,776 potential off-target sites in the human genome allowed us to leverage a simple deep learning model to effectively predict and rank LSI off-target recombination potential genome-wide.

6.4.1. Results

[0330] To determine if new methods for the computational prediction of potential off-target sites for an LSI were required to advance therapeutic CRISPR-directed integrases, we first evaluated HOMER²⁹, a well-established DNA-protein interaction model based on position weight matrices (PWM). Utilizing the position weight matrix identified by HIDE-Seq²⁰, HOMER’s default parameters predicted 4,598,283 potential off-target sites within the human genome (Figure 15 A). This number of potential off-target sites is too high for effective verification of potential off-target editing in cells with the current state of technology. Therefore, we sought to leverage the wealth of empirical and quantitative off-target discovery data from the highly sensitive Cryptic-seq technology²⁰ to develop IntQuery, a deep learning model for quantitative prediction of LSI off-target integration at any DNA sequence.

[0331] LSI attachment sequences are demarcated by a canonical central dinucleotide² which promotes annealing and ligation after 180-degree rotation to complete the recombination reaction. While the central dinucleotide does not directly interact with the LSI protein, it plays a critical role in determining recombination specificity because the landscape of cryptic attachment sites aligns with the central dinucleotide of the complementary attachment sequence. When applied to CRISPR-directed integrases, this feature allows simultaneous and specific multiplex gene insertion of unique cargos by utilizing orthogonal central dinucleotidesl4. Although there are 16 possible central dinucleotide combinations, 4 of the combinations are palindromic and permit bidirectional insertion, and 6 of the combinations represent complementary counter parts that allow directional control over the orientation of insertion. Therefore, we performed Cryptic-seq with substrates spanning the 12 central dinucleotides with directional insertion (AA/TT, TC/GA, GG/CC, CT/AG, CA/TG, GT/AC) to determine the off-target editing landscape of Bxbl. Cryptic-seq discovered a total of 410,776 unique potential off-target sites (Figure 15B) distributed throughout the human genome (Figure 15C). The DNA motif created from these discovery data was similar to the motif discovered by HIDE-seq20 and revealed the expected features of sequence conservation and palindromicity related to the dimerization of Bxbl on attachment sequences (Figure 15D).

[0332] Cryptic-seq is a quantitative biochemical off-target discovery assay because each recombination event imparts a unique molecular identifier (UMI) to the NGS reads²⁰. We reasoned that our empirical database of 410,776 cryptic attachment sites in the human genome could be used to train a machine learning model to predict integrase activity from a DNA sequence alone. To demonstrate this approach, we trained a simple multi-layer perceptron model with two hidden layers to perform a regression task, which we call IntQuery³⁰. We trained IntQuery to predict log-transformed UMIs from a one-hot encoding of the cryptic attachment site sequence, replacing the central dinucleotide of each sequence with NN. To prevent oversampling low UMI sites (17% of sites have UMI = 1), we used UML weighted sampling of the training data for each epoch. We conducted five-fold cross- validation on the training data, achieving a Spearman correlation of p = 0.42 between the target values and predictions (Figurel6A). A linear regression model trained with an identical protocol achieved a Spearman correlation of p = 0.34 across all predictions. Plotting target vs predicted values for both models revealed that while the MLP model maintains correlation across the entire range of UMI values, the linear regression model is unable to discriminate between low UMI and high UMI sites (MLP p = .51, linear regression p = 0.15 for sites with target values > 4). Therefore, IntQuery provides a simple method for predicting LSI cryptic attachment site identity and activity directly from DNA sequence. We anticipate that IntQuery will be valuable as an in-silico method for discovery of off-target sites and prioritization of potential high-activity sites for verification in LSI-edited cells of interest. 6.4.2. Discussion

[0333] IntQuery is a machine learning approach to predict the potential for LSI recombination at any site in the human genome. The approach taken by IntQuery is applicable to empirical LSI off-target discovery data from any genome. Computational off- target prediction with IntQuery complements empirical discovery data generated in the laboratory to help scientists prioritize potential off-target sites for experimental verification in edited cells. This two-step strategy was inspired by the approaches pioneered for the first Cas9 genome editing therapies^{24, 26, 31}, because the risk of off-target editing from Cas9 and LSIs are both dependent on DNA sequence homology^{22, 32}.

[0334] IntQuery can also be used for variant-aware off-target prediction³³ for any large serine integrase, including naturally discovered^{12, 14} and engineered ^{21, 34, 35} versions, by generating a large quantitative discovery dataset with technologies like Cryptic-Seq²⁰. Experimental verification of IntQuery off-target predictions in edited cells will increase its relevance in supporting the non-clinical studies required for a new drug application. We hope our communication of a deep learning off-target prediction tool will help enable the safe development of LSI-based therapeutics.

6.4.3. Methods

6.4.3.1 Position Weight Matrix Predication of Bxbl Cryptic Sites with HOMER

[0335] The position weight matrix discovered by HIDE-Seq²⁰ was used to predict locations in the hg38 human reference genome³⁶ that might be loci for off-target recombination by Bxbl. Briefly, we aligned the sequences flanking the integration sites discovered by HIDE- seq20 and generated a custom motif file based on the position frequency matrix and ran HOMER v4.11 motif analysis (scanMotifGenomeWide.pl)²⁹ with default parameters to predict binding at all sites in the hg38 human reference genome.

6.4.3.1 Cryptic-seq

[0336] Cryptic-seq was performed as previously described²⁰ with the following modifications. HEK293attB20 gDNA sheared in a ME220 focused-ultrasonicator (Covaris) to an average fragment length of approximately 350-500 bp. End prep and dA tailing of the gDNA was performed by using NEBNext® Ultra™ II DNA Library Prep Kit for Illumina® (E7645L) as protocol instructed, after which Illumina TruSeq sequencing annealed Y adaptors containing 8 bp UMIs were ligated to gDNA listed below.

[0337] Two independent cryptic-seq reactions with the following pools of cryptic-seq donor plasmids were performed. Reaction 1 contains the following plasmids at equimolar ratios: PL2327 AA, PL2341 TT, PL2337 GG, PL2332 CC, PL2339 TC, PL2335 GA, with a final total plasmid concentration of 30 nM. The plasmid pool was then reacted with 1 pg sheared/adapter-ligated gDNA and 1 pM Bxbl integrase at 37°C for 4h in a recombination buffer⁷. Reaction 2 contains the following plasmids at equimolar ratios: PL2312 GT, PL2328 AC, PL2331 CA, PL2340 TG, PL2334 CT, PL2329 AG and was reacted under the same conditions as reaction described above. A representative sequence map of the cryptic-seq plasmid PL2312 GT can be found in our previous publication²⁰ with all plasmids in this study being identical to PL2312 except for the indicated central attP dinucleotide. The reaction was stopped by adding sodium dodecyl sulfate to a final concentration of 0.1% and the products were cleaned using the Zymo clean and concentrator kit (Zymo research). PCR was used to amplify the regions of integration using Q5 polymerase (NEB) and common primer for P5 (LM_P5) and N7 primer (GN037_CrypticSeq_N7_i7_N701). Following PCR, a 1.5x AMPure bead clean-up was performed, and products were run on a Tapestation 4200 (Agilent Technologies) using DI 000 tape to confirm amplification. Library concentration was determined using the Next Library Quant Kit (NEB) for Illumina. Libraries were sequenced on an Illumina NovaSeq (Fulgent Genetics).

[0338] For bioinformatic analysis of Cryptic-seq data, FASTQ files from Illumina sequencing were loaded into a custom Cryptic-seq bioinformatic pipeline tbChaSIn developed by Tome Biosciences and Fulcrum Genomics to discover and quantify cryptic recombination sites from Cryptic-seq data (github.com/didacs/tbChaSIn). The bioinformatic workflow starts with the trimming of reads that contain leading attP or attB sequences (P, P’ or B, B’) at the 5’ end of R2. Untrimmed reads were discarded. Included reads were further trimmed to remove Illumina TrueSeq adapter sequence (AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT (SEQ ID NO: 584)) from the 3’ end of R2. Trimmed read sequences were aligned against the hg38 human reference genome³⁶ with appended attP and attB sequences by BWA aligner³⁷ to create mapped BAM files for each sample. The resulting BAM files were deduplicated by Picard³⁸. The integration sites are identified by the first base of R2 and quantified by the number of deduplicated reads that pass mapping quality threshold (MAPQ > 20). Output .csv files containing sites per sample along with collation of sites across samples were generated.

6.4.3.3 IntQuery

[0339] Cryptic attachment sequences were one-hot encoded and used as input to predict their corresponding UMI values. UMI values were scaled to log (1 + UMI) before training. The MLP model was written in Pytorch and consisted of an input layer, two hidden layers with dropout, and an output layer with a single dimension, with mean-squared-error between prediction and target used as an error function39. To prevent oversampling of low UMI sites, we randomly sampled the training data with a weight directly proportional to each sites log (UMI) value. For demonstration purposes in this paper, we trained a model with hidden dimensions of 200 and 5% dropout. In preliminary experiments we found that model performance is relatively insensitive to model architecture but don’t rule out the possibility that alternative models might outperform the simple MLP described here.

6.4.3.4 Data and model availability

[0340] All code and training data have been made publicly available on GitHub: github.com/mhbakalar/intquery. The Cryptic-seq next-generation sequencing data is available under NCBI SRA bioproject PRJNA1169517. The code used for Bxbl Cryptic-seq is available on GitHub github.com/didacs/tbChaSIn.

6.4.4. References

1. Stark, W.M. The Serine Recombinases. Microbiol Spectr 2 (2014).

2. Smith, M.C.M. Phage-encoded Serine Integrases and Other Large Serine Recombinases. Microbiol Spectr 3 (2015).

3. Grindley, N.D., Whiteson, K.L. & Rice, P.A. Mechanisms of site-specific recombination. Annu Rev Biochem 75, 567-605 (2006). 4. Groth, A.C. & Calos, M.P. Phage integrases: biology and applications. J Mol Biol 335, 667-678 (2004).

5. Thorpe, H.M. & Smith, M.C. In vitro site-specific integration of bacteriophage DNA catalyzed by a recombinase of the resolvase/invertase family. Proc Natl Acad Sci U S A 95, 5505-5510 (1998).

6. Kim, A.I. et al. Mycobacteriophage Bxbl integrates into the Mycobacterium smegmatis groELl gene. Mol Microbiol 50, 463-473 (2003).

7. Wang, X. et al. Bxbl integrase serves as a highly efficient DNA recombinase in rapid metabolite pathway assembly. Acta Biochim Biophys Sin (Shanghai) 49, 44-50 (2017).

8. Xu, Z. et al. Accuracy and efficiency define Bxbl integrase as the best of fifteen candidate serine recombinases for the integration of DNA into the human genome. BMC Biotechnol 13, 87 (2013).

9. Groth, A.C., Olivares, E.C., Thyagarajan, B. & Calos, M.P. A phage integrase directs efficient site-specific integration in human cells. Proc Natl Acad Sci U S A 91, 5995-6000 (2000).

10. Farruggio, A.P., Bhakta, M.S., du Bois, H., Ma, J. & M, P.C. Genomic integration of the full-length dystrophin coding sequence in Duchenne muscular dystrophy induced pluripotent stem cells. Biotechnol J 12 (2017).

11. Anzalone, A.V. et al. Programmable deletion, replacement, integration and inversion of large DNA sequences with twin prime editing. Nat Biotechnol 40, 731-740 (2022).

12. Durrant, M.G. et al. Systematic discovery of recombinases for efficient integration of large DNA sequences into the human genome. Nat Biotechnol 41, 488-499 (2023).

13. Blanch- Asensio, A. et al. STRAIGHT-IN enables high-throughput targeting of large DNA payloads in human pluripotent stem cells. Cell Rep Methods 2, 100300 (2022).

14. Yarnall, M.T.N. et al. Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases. Nat Biotechnol 41, 500-512 (2023).

15. Ghosh, P., Kim, A.I. & Hatfull, G.F. The orientation of mycobacteriophage Bxbl integration is solely dependent on the central dinucleotide of attP and attB. Mol Cell 12, 1101-1111 (2003). 16. Tou, C.J. & Kleinstiver, B.P. Programmable RNA-guided enzymes for nextgeneration genome editing. Nature 630, 827-828 (2024).

17. Mali, P. et al. RNA-guided human genome engineering via Cas9. Science 339, 823- 826 (2013).

18. Nambiar, T.S., Baudrier, L., Billon, P. & Ciccia, A. CRISPR-based genome editing through the lens of DNA repair. Mol Cell 82, 348-388 (2022).

19. Allen, D. et al. CRISPR-Cas9 engineering of the RAG2 locus via complete coding sequence replacement for therapeutic applications. Nat Commun 14, 6771 (2023).

20. Hazelbaker, D.Z. et al. Large Serine Integrase Off-target Discovery and Validation for Therapeutic Genome Editing. bioRxiv, 2024.2008.2023.609471 (2024).

21. Pandey, S. et al. Efficient site-specific integration of large genes in mammalian cells via continuously evolved recombinases and prime editing. Nat Biomed Eng (2024).

22. Chalberg, T.W. et al. Integration specificity of phage phiC31 integrase in the human genome. J Mol Biol 357, 28-48 (2006).

23. Wang, J.Y. & Doudna, J. A. CRISPR technology: A decade of genome editing is only the beginning. Science 379, eadd8643 (2023).

24. Frangoul, H. et al. CRISPR-Cas9 Gene Editing for Sickle Cell Disease and betaThalassemia. N Engl J Med 384, 252-260 (2021).

25. Kanter, J. et al. Biologic and Clinical Efficacy of LentiGlobin for Sickle Cell Disease. N Engl J Med 6, 617-628 (2022).

26. Gillmore, J.D. et al. CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis. N Engl J Med 3135, 493-502 (2021).

27. Longhurst, H.J. et al. CRISPR-Cas9 In Vivo Gene Editing of KLKB1 for Hereditary Angioedema. N Engl J Med 390, 432-441 (2024).

28. fda.gov/regulatory-information/search-fda-guidance-documents/human-gene-therapy- products-incorporating-human-genome-editing (2022).

29. Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 38, 576-589 (2010). 30. Rumelhart, D.E., Hinton, G.E. & Williams, R.J. Learning representations by back- propagating errors. Nature 323, 533-536 (1986).

31. Tsai, S.Q. et al. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases. Nat Biotechnol 33, 187-197 (2015).

32. Hsu, P.D. et al. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat Biotechnol 31, 827-832 (2013).

33. Cancellieri, S. et al. Human genetic diversity alters off-target outcomes of therapeutic gene editing. Nature Genetics 55, 34-43 (2023).

34. Fauser, F. et al. Systematic Development of Reprogrammed Modular Integrases Enables Precise Genomic Integration of Large DNA Sequences. bioRxiv, 2024.2005.2009.593242 (2024).

35. Hew, B.E. et al. Directed evolution of hyperactive integrases for site specific insertion of transgenes. bioRxiv, 2024.2006.2010.598370 (2024).

36. Schneider, V.A. et al. Evaluation of GRCh38 and de novo haploid genome assemblies demonstrates the enduring quality of the reference assembly. Genome Res 27, 849-864 (2017).

37. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754-1760 (2009).

38. broadinstitute. github.io/picard/ (2019).

39. Paszke, A. et al., Vol. 32. (eds. H. Wallach et al.) (2019).

6.5. Example 5: Large Serine Integrase Off-target Discovery and Validation for Therapeutic Genome Editing

[0341] While numerous technologies for the characterization of potential off-target editing by CRISPR/Cas9 have been described, the development of new technologies and analytical methods for off-target recombination by Large Serine Integrases (LSIs) are required to advance the application of LSIs for therapeutic gene integration. Here we describe a suite of off-target recombination discovery technologies and a hybrid capture validation approach as a comprehensive framework for off-target characterization of LSIs. HIDE-Seq (High- throughput Integrase-mediated DNA Event Sequencing) is a PCR-free unbiased genome- wide biochemical assay capable of discovering sites with LSI-mediated free DNA ends (FDEs) and off-target recombination events. Cryptic-Seq is a PCR-based unbiased genomewide biochemical or cellular-based assay that is more sensitive than HIDE-Seq but is limited to the discovery of sites with off-target recombination. HIDE-Seq and Cryptic-Seq discovered 38 and 44,311 potential off-target sites respectively. 2,455 sites were prioritized for validation by hybrid capture NGS in LSI-edited K562 cells and off-target integration was detected at 52 of the sites. We benchmarked the sensitivity of our LSI off-target characterization framework against unbiased whole genome sequencing (WGS) on LSI- edited samples, and off-target integration was detected at 5 sites with an average genome coverage of 40x. This reflects a greater than 10-fold increase in sensitivity for off-target detection compared to WGS, however only 4 of the 5 sites detected by WGS were also validated by hybrid capture NGS. The dissemination of these technologies will help advance the application of LSIs in therapeutic genome editing by establishing methods and benchmarks for the sensitivity of off-target detection.

[0342] The large serine integrase (LSI) family constitutes a diverse group of site-specific recombinases that play pivotal roles in mediating DNA rearrangements¹'³. Serine integrases, in contrast to their tyrosine recombinase counterparts, utilize a serine residue for catalysis, leading to distinct mechanistic features⁴. This large family encompasses integrases with varying sizes and functionalities, with notable members including PhiC31 integrase from Streptomyces bacteriophage PhiC31⁵ and Bxbl integrase discovered in my cobacteriophage Bxbl6. Both PhiC31 and Bxbl integrases are well-recognized for their utility in site-specific recombination applications by virtue of direct recombination between phage attachment site attP and bacterial attachment site attB with the requirement of no co-factors or DNA supercoiling⁷'⁹. The precise and efficient DNA manipulation capabilities of the large serine integrase family have positioned it as an attractive tool for synthetic biology and genome editing applications¹⁰'¹⁴.

[0343] Large serine integrases facilitate recombination between attachment sites on linear or circular DNA substrates^{5, 15} and the mechanism of action is outlined in Figurel7A. Dimerized integrases bind to specific sequences in phage (attP) and bacterial host (attB) DNA2 The bound integrases associate to form a synaptic complex, connecting paired homologous sequences. Integrases cleave all four DNA strands at the central dinucleotide, forming 5'- phosphoserine linkages and generating 3 '-dinucleotide overhangs³. This cleavage mechanism covalently bonds DNA strands to integrase, avoiding free DNA ends and the opportunity for mis-repair by host DNA machinery. Subunits exchange places, promoting irreversible ligation through the formation of new attachment sites, attL and attR. Unlike CRISPR/Cas9 knock-in approaches that depend on cellular DNA repair machineries5, 15, LSI integration is deterministic and independent of host cell factors ⁵‘ ¹⁵.

[0344] A comprehensive understanding of the specificity of integrase-mediated insertion for gene editing strategies, such as integrase-mediated programmable genomic integration (I- PGI), is crucial for safe and effective therapeutic use. While LSIs attachment sites of attP and attB are typically between 40-50 bp¹⁶ and recombine efficiently with a high degree of sitespecificity², sites with degenerate homology to attB or attP, known as cryptic sites or pseudo sites can exist in the human genome and can support recombination at lower frequencies¹², ^{17, 18} (Figure 17B). In terms of off-target editing events potentially mediated by cryptic sites in I- PGI, and other integrase-based editing approaches^{12, 13}, we broadly classify these events into three genetic classes: indels, DNA insertions, and large chromosomal rearrangements (Figure 17Cc). In the case of indels, because site-specific recombination by LSIs proceeds through a cleaved DNA intermediate, disruption of this intermediate could result in de-protection of the cleaved free DNA ends (FDEs) and repair of the exposed DNA ends by the cellular DNA repair 19. Biochemical evidence for FDE formation via disruption of LSI recombination by chemical treatment or mismatched central dinucleotides exists in biochemical systemsl5. In the case of off-target integration events, Bxbl integrase has been reported to integrate DNA cargoes into a small number of sites in the human genome at low levels^{12, 18}, though the extent of potential Bxbl integrase cryptic sites in the human genome for Bxbl is not clear. In addition to off-target recombination of DNA cargo into cryptic sites, there exists the possibility of recombination between two cryptic sites in the human genome with previous publications describing multiple chromosomal rearrangements by PhiC31 integrases in human cells, potentially mediated by cryptic sites in the human genome²⁰ (Figure 17C). To address these potential outcomes, we set out to develop more sensitive methods to determine the off-target profile of Bxbl integrase in human cells.

[0345] Here we demonstrate an empirical off-target discovery and validation framework for the characterization of LSI specificity for therapeutic genome editing that is sensitive and comprehensive of unintended DNA variants. We describe the generation of a suite of sensitive biochemical-based discovery and characterization approaches in addition to a targeted hybrid capture NGS approach for the validation of off-target LSI editing in cells at sites nominated by our discovery assays. For discovery of genomic sites that support LSI off- target activity, we present HIDE-seq and Cryptic-seq. HIDE-Seq (High-throughput Integrase- mediated DNA Event Sequencing) is a PCR-free unbiased genome-wide biochemical assay capable of discovering sites with LSI-mediated FDEs and off-target recombination events. Cryptic-Seq is a PCR-based unbiased genome-wide biochemical or cellular-based assay that is more sensitive than HIDE-Seq but is limited to the discovery of sites with off-target recombination. The application of these biochemical characterization technologies in combination with cell-based hybrid capture NGS validation, help advance the implementation of LSIs in therapeutic genome editing approaches by establishing benchmark criteria for sensitive off-target discovery and validation in therapeutically relevant cell-types.

6.5.1. Results

6.5.1.1 HIDE-Seq is an unbiased and genome-wide LSI off- target discovery technology

[0346] For unbiased and genome-wide assessment of potential LSI off-target editing events, we combined a biochemical LSI recombination reaction containing recombinant integrase, substrates and purified and deproteinized human genomic DNA with whole genome sequencing. We named this approach HIDE-seq (High-throughput Integrase-mediated DNA Event Sequencing), a PCR-free genome-wide biochemical assay for discovering LSI- mediated off-target events, including FDEs, cryptic integration events, or structural variants via WGS (Figure 18 A). For the development of HIDE-seq, we drew inspiration from Digenome-seq21, a biochemical assay for the genome-wide detection of CRISPR-Cas9 off- target cleavage via WGS of DNA samples digested with Cas9 to identify read pairs reads that correspond to FDEs.

[0347] To demonstrate the utility of HIDE-seq, we leveraged the well-characterized LSI Bxbl integrase^{6, 22} and isolated human genomic DNA (gDNA) from two sources (1) a HEK293 cell line that contains two copies of the Bxbl attB site in a lentiviral vector integrated into a single locus on chromosome 5 (referred to from here on as HEK293attB) and (2) commercially available peripheral blood mononuclear cells (PBMCs).

[0348] The cell-free nature of HIDE-seq allows for tuning and supraphy siological saturation of LSI and DNA substrates in the reaction to enable sensitive discovery of potential off-target events to deeply map the landscape of potential off-target events across any given DNA sample. For HIDE-seq reactions, 8 pg of PBMC or HEK293< //7> gDNA was incubated with 10 nM attP or attB linear substrate and 400 nM of recombinant Bxbl integrase in various reactions in triplicate for 8 hours at 37C° as outlined in Table 14. As a positive control reaction for free DNA end formation, we performed Digenome-seq21 reactions with Cas9 complexed with a previously described low specificity sgRNA23 (Table 14). After completion of the reactions, the gDNA was purified and processed via amplification-free library preparation and subjected to Illumina WGS with a target of 30x coverage (recovered mean coverage of 18.5x, with a minimum of lOx and maximum of 28x across samples) followed by bioinformatics analysis by Digenomitas24 (for FDE discovery) or a custom HIDE-seq pipeline (for LSI off-target discovery).

Table 14. HIDE-seq sample summary. Overview of HIDE-seq samples performed in triplicate (3x) and the experimental purpose.

[0349] No LSI associated FDE’s were identified, as Bxbl treated samples had a similar number of FDEs (69-212 total counts) to background levels of untreated DNA samples (104- 206 total counts) (Figure 18B). Furthermore, no recurrent FDEs were detected by HIDE-seq, where recurrence is defined as the appearance of a FDE at the same genomic location in the triplicate samples (Figure 18B). In contrast, 12,597 recurrent FDEs were detected in samples treated with spCas9 complexed with VEGFA sgRNA (Figure 18B). These results indicate that Bxbl integrase does not generate significant levels of FDEs during its canonical recombination mechanism consistent with previous results shown for Bxbl in biochemical contextsl5.

[0350] HIDE-seq identified high levels of on-target integration between the attP donor and the endogenous attB sites in the HEK293< //7> gDNA samples, with integration read count frequencies of 100%, 100%, and 94% for HEK293attS Bxbl + attP donor replicates 1, 2, and 3 respectively (quantified as percentage of alii. and attR reads versus attB total read counts, data not shown). An example visualization image of HEK293attB Bxbl + attP donor replicate 1 displaying 100% integration of the attP donor in integrative genomics viewer²⁵ (IGV) is shown in Figure 18C. These high levels of on-target integration indicate our biochemical reaction conditions and concentrations of Bxbl integrase and DNA donor substrates are robust enough to detect potential off-target events, as shown by the integration reads at an example cryptic attB site (Figure 18C). Across the aggregate of PBMC and HEK293attB gDNA samples treated with Bxbl and attP donor, we discovered 36 unique cryptic attB sites, which we designate as CAS sites (cryptic attachment sites) (Figure 18D). For the PBMC gDNA samples treated with Bxbl and attB donor, we discovered 2 unique cryptic attP sites. Our detection of fewer numbers of cryptic attP sites compared to cryptic attB sites is consistent with recently published resultsl8. With the 36 cryptic attB sites uncovered by HIDE-seq, we generated detailed sequence logos that represent the specificity profile of Bxbl across the human genome (Figure 18E). Lastly, to examine if any split reads in the HIDE-seq dataset support potential recombination between any of the 38 discovered cryptic sites to another genomic location, we utilized fgsv caller by Fulcrum Genomics26. With fgsv caller analysis, we found no split reads that map at both the cryptic site and another genomic location, suggesting that these 38 discovered cryptic sites did not engage in detectable cryptic recombination in the HIDE-seq reactions (data not shown). Taken together, these results demonstrate HIDE-seq is a sensitive and unbiased approach to identify any potential FDEs or off-target integration events by LSIs.

6.5.1.2 Cryptic-Seq is a sensitive discovery technology for LSI off-target recombination

[0351] While HIDE-seq can identify potential LSI cryptic sites via sequencing whole genomes, we sought to develop a genome-wide approach that preferentially enriches recombined species to enable more sensitive discovery of cryptic sites. We named this approach Cryptic-seq, and it relies on the application of a specialized DNA donor substrate to enable one-step PCR enrichment and sequencing of recombined cryptic sites (Figure 19A). To execute Cryptic-seq, gDNA is first tagmented with Tn5 transposase21 loaded DNA oligonucleotides containing Illumina P5 sequencing adaptors, i5 indexes, and a unique molecule identifier (UMI)²⁷'²⁹ (Figure 19A). This pre-tagmentation of the gDNA is a crucial step to increase the sensitivity of the assay as it prevents sequencing of any unrecombined donor DNA substrate. After tagmentation, gDNA is incubated with recombinant Bxbl integrase and a Cryptic-seq plasmid-based vector that contains either attP (GT dinucleotide) flanked by a unique barcode sequence (BC1) and an Illumina N7 primer binding site 5’ of the P half-site of attP, and on 3’ side of the P’ site of the attP a second unique barcode sequence (BC2) and an Illumina P7 primer binding site (Figure 19A). After the reaction is complete, the gDNA is purified and PCR amplified with i7 indexed primers that prime off either the N7 or P7 primer binding site in the cryptic-seq vector for one-step amplification of recombined DNA fragments (Figure 19A), followed by NGS and bioinformatic identification of genomic reads sequence-tagged with either the P or P’ half-sites from the attP donor. Like HIDE-seq, Cryptic-seq is a biochemical-based assay and the reactions are performed with supraphy siological concentrations of Bxbl integrase and DNA donor, conditions far higher than we can achieve in a cell, for maximum sensitivity to enable deep discovery of any potential off-target events.

[0352] To apply Cryptic-seq to the human genome, we isolated gDNA from HEK293attB cells and used Tn5 transposase to tagment the library to an average fragment size of 700 bp. Three independent replicate reactions were performed at 37C° for 4 hours with each reaction containing 1 pg tagmented HEK293attS gDNA, 10 nM of the attP-GT vector, and 1 pM Bxbl integrase. After quenching of the reaction with sodium dodecyl sulfate (SDS) the gDNA was purified and PCR enrichment was performed with N7 primer. In this report, Cryptic-seq was performed with one-sided enrichment, with only indexed N7 primers used for detection of cryptic attR fragments because the creation of attL from a cryptic attB is inextricably linked to the formation of attR by Bxbl integrase. After PCR-based enrichment and indexing, NGS libraries were quantified and sequenced by Illumina next generation sequencing on a NextSeq 2000. Examples of representative IGV²⁵ plots showing the structure of both on-target and cryptic site genomic reads tagged with the P half-site are shown in Figure 19B.

[0353] In total across the three replicates (Cryptic-seq Repl, Rep2, and Rep3), 44,311 unique cryptic attB sites with > 1 UMI support were discovered by Cryptic-seq, with each replicate containing > 22,000 discovered cryptic sites in the human genome that can recombine in a Bxbl integrase-dependent manner (Figure 19C). The recombination signal produced by Cryptic-Seq is reflected by the number of UMIs detected for each site and the distribution is negatively skewed with a long-tail of single UMI events detected. In each replicate, over 200 sites show greater than 500 UMIs counted, reflecting that these sites have recombined individual and unique DNA fragments containing these cryptic sites over 500 times in the Cryptic-seq reaction (Figure 19C). Of note, 4 unique cryptic sites displayed UMI counts greater than 5000 UMIs in all three replicates (Figure 19C). For reference, the endogenous attB site in the HEK293attB genome displayed integration read UMI counts of 6033 UMIs in Repl, 5622 UMIs in Rep2, and 4808 UMIs in Rep3, indicative of high levels of on-target recombination (data not shown). The recurrence of recombination at these sites varies also across replicates, with sites with high UMI counts (> 50 UMIs) displaying higher levels of recurrence than sites with low UMI counts (< 5 UMIs) (Figure 19C).

[0354] Cryptic-Seq re-discovered 33 of the 36 cryptic attB sites that were discovered by HIDE-seq (Figure 19D), which reflected a 91% concordance in that comparison. However, Cryptic-seq also discovered an additional 44,278 cryptic attB sites in the human genome. Despite this large increase in the number of loci detected, the Bxbl sequence motif (Figure 19E) did not change significantly compared to the HIDE-seq data, which indicates that these data represent a 1,265-fold increase in sensitivity and not the result of spurious biochemical signals.

6.5.1.3 Validation of Hybrid Capture NGS as a scalable approach for LSI editing quantification

[0355] With a >1200-fold increase in sensitivity for LSI cryptic site discovery by Cryptic-seq over HIDE-seq and thousands of potential off-target sites discovered, we required a scalable validation approach to characterize off-target editing frequencies in a relevant cellular context. Droplet digital PCR³⁰ (ddPCR), while quantitatively accurate, cannot be readily scaled to thousands of unique sites. Multiplex targeted NGS amplicon sequencing assays, such as rhAmpSeq can scale to thousands of sites for approaches like CRISPR/Cas9 editing detection³¹, however the unique sequence of recombined reads generated by LSIs does not allow for quantitative detection with forward and reverse genomic primers surrounding the target site. To overcome these limitations, we turned to multiplex hybrid capture NGS³² which has been applied previously to assess off-target editing frequencies of thousands of potential off-target sites of the first FDA approved CRISPR-based therapeutic, Exagamglogene autotemcel (marketed as CASGEVY)^{33, 34}

[0356] To customize hybrid capture NGS for the detection of LSI-editing events (indels, cryptic integration events, and structural variants) in DNA from edited cells, we first designed 120-mer DNA capture probes to specifically bind left (5’) and right (3’) of the discovered cryptic site central dinucleotide (Figure 20A). By designing probes flanking both sides of the central dinucleotide of the cryptic site, we could compare the capture and quantification efficiency of the left and right probes. Importantly, by designing probes that flanking the central dinucleotide of the genomic site, we ensure the quantitative nature of the assay is preserved as the capture probe will enrich both edited and unedited genomic sites with equal efficiency, thus allowing accurate quantification of edited and unedited DNA species after bioinformatic UMI deduplication of NGS reads.

[0357] To qualify hybrid capture NGS for the quantitative detection of cryptic integration events we performed a standard curve experiment using commercially available purified human gDNA spiked with a custom synthetic DNA fragment at varying relative copy ratios. The DNA sequence of the fragments was designed to simulate integration at a HIDE-seq and Cryptic-seq discovered cryptic attB site CAS031 (chr6: 142247516-142247561; GT dinucleotide located in minus strand starting at 142247538; Figure 18C). Notably, CAS031 was also discovered in a recent report of Bxbl integrase off-target sitesl8. The synthetic DNA fragment was titrated into human gDNA at specific copy number concentrations (95%, 50%, 10%, 1%, 0.1%, 0.05%, 0.001%, and 0%) to generate an 8-point standard curve. These DNA standards were processed in triplicate by NGS library prep and target enrichment with hybrid capture panels containing probes hybridizing either left or right of CAS031. Illumina short-read sequencing and bioinformatic analysis was performed via our custom hybrid capture analysis pipeline. As shown in Figure 20B, by comparing observed versus expected integration frequencies, hybrid capture NGS affords accurate quantitation across standard curve and reliable detection of recombined reads (as defined as detection of the edit in > 2/3 replicates) down to 0.1%and single-point detection with both left and right CAS031 probes and single point detection as low as 0.05% with the left probe at the level of sequencing performed in this experiment. For positive calling of integration events in hybrid capture NGS, we set the threshold that a read (UMI count) must be informative, i.e. the read spans the complete cryptic attB target sequence (unedited read) or the complete cryptic attL or attR sequence with a full P or P’ half-site sequence (integrated read). In this experiment, we recovered an average of 440 informative UMI counts for the left CAS031 probe samples and an average of 797 informative UMI counts for the right CAS031 probe samples (data not shown). We also observed strong correlation between observed and expected integration frequencies with Pearson r-squared values of 0.9897 and 0.9954 for the left and right CAS031 probes respectively, highlighting the quantitative ability of hybrid capture NGS for quantitative detection of cryptic integration events in edited gDNA (Figure 20D).

6.5.1.4 Hybrid Capture NGS Validation of LSI off-target editing at cryptic attB sites in K562 cells

[0358] To examine the potential of HIDE-seq and Cryptic-seq to discover bona fide recombinogenic cryptic sites in cells, we leveraged an engineered K562 cell line that contains two atlB-Gl sequences in different genomic locations. This line was generated by transducing K562 cells with a lentivirus carrying insert in a PGK-attB-EFla-PuroR cassette that integrated at unique genomic sites in both chromosome 6 and 17 (K562attB cells). Biological triplicate samples (Rep 1, Rep 2, and Rep 3) of K562attB cells were co-transfected with mRNA expressing Bxbl integrase and an attP-GT plasmid donor, along with control samples in which K562attB cells were transfected with only the attP-GT plasmid donor (Control 1, Control 2, and Control 3), followed by harvesting and lysis 3 days posttransfection to isolate gDNA. Assessment of on-target editing by ddPCR showed robust on- target average integration frequencies of 45.5%, 48.4%, and 48.3% across edited Rep 1, Rep 2, and Rep 3 samples, respectively (Figure 20C).

[0359] To create a hybrid capture panel composed of discovered sites to test for off-target editing in K562s, we prioritized inclusion of cryptic sites located in Tier I (gene coding regions), Tier II (non-coding regions (exonic UTR or intronic) of coding genes, and Tier III (exonic or intronic regions of non-coding genes) as defined by the off-target threat matrix from The Broad Institutes 5 in combination with a prioritization of cryptic sites with UMI counts > 10 from Cryptic-seq (Figure 19C). Of note, the probe panel consists of discovered cryptic sites and does not contain an on-target probe targeting the lentiviral cassette that contains the canonical attB sequence and, given the concordance in detection by left and right probes shown in Figure 20B, we designed the panel to contain only a single capture probe (either left or right) for each target (i.e., single sided panels). This inclusion criteria resulted in a single-sided probe panel size of 2,455 genomic sites. Of the 2,455 included sites, only 16 were discovered in HIDE-seq (CAS002, CAS005, CAS006, CAS008, CAS011, CAS013, CAS017, CAS018, CAS019, CAS020, CAS023, CAS024, CAS026, CAS030, CAS031, CAS032 in Figure 18D). For hybrid capture NGS, 2 of the edited K562attB biological replicates (Rep 1 and Rep 2) and 2 of the cargo alone samples (Control 1 and Control 2) were chosen for captured enrichment and analysis. Samples were sequenced on an Illumina NextSeq 2000 with 2x300 paired end sequencing, yielding an average of 3.7 million reads near probe across samples and an average target capture percentage of 37% (derived from quantifying reads near probes divided by total reads in the sample). FASTQ files were analyzed through a custom hybrid capture bioinformatic pipeline for indel and cryptic integration event detection.

[0360] For quantification of indels, we filtered targets containing >1,000 informative UMI average read counts across edited Rep 1 and Rep 2 and unedited Control 1 and Control 2 samples (to reduce any miscalls due to sampling bias). For a positive indel call, the indel must overlap within a 6 bp window surrounding the central GT dinucleotide with the minimal mapping criteria that the read that contains minimum of 50 bp of flanking each side of the dinucleotide. Using these inclusion criteria, 621 unique targets were analyzed for indel detection by comparing edited and unedited samples³³. Statistical testing by Welch’s onesided T test, revealed only 16 sites with p-values <0.05. However, across all 621 sites assayed, including the 16 sites with p-values <0.05, we identified only one potential off-target site, CAS 14693, which was not significant but did display an Aindel frequency between the edited and control samples greater than 0.2% (data not shown). CAS 14693 displayed averaged indel frequencies of 7.5% in edited samples and 6.7% in unedited samples with a Aindel frequency of -0.7% (data not shown) . Visual inspection of the indel sequences indicates they are likely somatic indels present in the K562attB cell line as they represent precise and single CAG repeat deletions present in both edited and unedited samples in a CAG-rich repetitive region of the potential cryptic site (data not shown). In conclusion, of the 621 sites that passed our indel analysis criteria, we identified no off-target indels scored that associated with Bxbl integrase activity.

[0361] For quantification of cryptic integration, positive calls were made on any informative genomic reads tagged at the GT dinucleotide with a full P or P’ half-site sequence from the DNA cargo in the UMI deduplicated read. In total, 52 cryptic sites were validated for off- target insertion in K562attB cells (12 sites in Repl and 48 sites in Rep2) with 8 sites showing recurrent off-target insertion in both Repl and Rep2 samples (CAS421, CAS023, CAS6083, CAS4310, CAS024, CAS985, CAS4668, CAS008) with editing frequencies ranging from 4.26% (CAS421, Rep 2) to 0.049% (CAS3972, Rep 2) (Figure 20D). Of the 52 unique validated sites across Rep 1 and Rep 2, 5 of the 52 sites were discovered in HIDE-seq (CAS002, CAS008, CAS023, CAS024, CAS031) while all 52 validated sites were discovered in Cryptic-seq (Figure 20D). It is important to note that K562 cells display inherent chromosomal copy number variation³⁶, so integration frequencies of off-target edits may not be directly comparable between targets. Across the 52 validated sites, the minimum number of informed reads was 694 UMI counts for CAS761 with 1 UMI count supporting integration (Rep 2) and the maximum number of informed reads was 2,177 UMI counts for CAS3785 with 2 UMI counts supporting integration (Rep 2) (data not shown). To corroborate our findings with an orthogonal approach, we assessed off-target integration at CAS421 by ddPCR in Rep 1 and Rep 2 gDNA samples. Analysis by ddPCR reveal the integration frequencies of 2.96% and 5.91% in Rep 1 and Rep 2 respectively (Figure 20E), closely matches the integration frequencies at CAS421 via hybrid capture NGS (1.29% in Rep 1 and 4.26% in Rep 2) (Figure 20D). These confirmatory results at a single site by ddPCR, along with the validation data in Figure 20D, demonstrate the sensitivity and scalability of hybrid capture NGS for LSI off-target frequency determination.

6.5.1.5 The LSI off-target characterization framework consisting of HIDE-Seq, Cyptic-Seq and Hybrid capture NGS, is more sensitive than WGS alone

[0362] To benchmark our discovery and validation framework against unbiased whole genome sequencing, we isolated gDNA from Rep 1 and Rep 2 edited K562attB cells (Figure 20C) and performed WGS to an average genomic coverage of 40x. To identify any Bxbl integrase-mediated integration events from WGS, fgsv26 was used to call and filter any split reads that contain genomic sequences and either P or P’ sequences of the cargo tagged at the GT dinucleotide. Across the two WGS samples, we identified a total of 5 unique off-target insertions at sites CAS031 (Rep 1), CAS071 (Rep 1), CAS421 (Rep 1, Rep 2), and CAS985 (Repl, Rep2) and CAS1991 (Rep 1) (Figure 20F). In terms of detection of these sites with our framework, 1 WGS validated site (CAS031) was discovered with HIDE-seq while all 5 WGS validated sites were discovered with Cryptic-seq (Figure 20F). In terms of validation, hybrid capture validated 4 of the 5 WGS validated sites (CAS031, CAS421, CAS985, CAS1991) with only CAS071 not called in the hybrid capture dataset. In contrast, WGS at 40x coverage did not detect the remaining 48 sites that were validated in the hybrid capture datasets (Figure 20D). Taken together, our results from both hybrid capture NGS and WGS validate a composite total of 53 off-target sites across the edited K562attB samples.

[0363] Examining the classification error rates described in Gillmore et al., ¹ for CRISPR-

Cas9 off-target discovery in the context of off-target characterization and potential off-target editing of a genome editing therapeutic, a high rate of false positives is tolerated in order to mitigate the potential for false negatives to occur. HIDE-seq discovered 36 potential cryptic attB sites (Figure 18E) with 16 of these sites present in the hybrid capture panel and 5 of the 16 tested sites displaying off-target editing in K562attB cells. As a result, HIDE-seq displays a false positive rate of 68.8% (l-(5 validated sites 16 tested discovered sites) and false negative rate of 90.6% (l-(5 validated sites^-53 validated sites) (Figure 20G). Cryptic-seq discovered 44,311 unique cryptic attBs with 2,455 of these sites present in the hybrid capture panel and 53 of these sites displaying off-target editing in K562attB cells. Accordingly, Cryptic-seq displays a false positive rate of 97.8% ( 1 -(53 validated sites 2,455 tested discovered sites) and false negative rate of 0% ( 1 -(53 validated sites 53 validated sites) (Figure 20G).

6.5.2. Discussion

[0364] Large serine integrases are a unique class of compact enzymes that have evolved to insert large DNA sequences (>50 kb) in specific genomic locations, and can catalyze all steps of the integration reaction, independent of host cell DNA repair pathways2. Thus, integrases have immense therapeutic potential genome editingl 1-14, 18. In particular, LSIs can be used in combination with Cas9 nickases and writing enzymes (e.g. reverse transcriptase¹¹, ¹⁴ or ligases³⁸) to enable therapeutic applications including endogenous gene replacement and the development of highly engineered cell-based medicines. LSIs allow for the directional, seamless integration of large DNA sequences without depending on FDEs or cellular DNA repair pathways. To date, comprehensive methods for detecting and validating LSI specificity have not been developed but will be required to develop integrase-based therapeutics.

[0365] Standard genotoxicity evaluation approaches, like the Ames test or Comet assays³⁹, are not suitable for homology dependent off-target editing from an LSI or an RNA-guided nuclease because they are designed to detect the effects of homology independent sources DNA damage, such as chemical or physical agents. Homology dependent off-target editing poises the risk of recurrent gene disruption of a tumor suppressor that would be undetectable with standard approaches. Therefore, we have sought to contribute fit for purpose and sensitive methods to evaluate the specificity any LSI in the human genome to help advance the therapeutic application of endogenous gene replacement.

[0366] In contrast to iPSC-based cell therapy applications, where a clonal cell line is genome engineered and WGS alone is often sufficient to characterize off-target editing⁴⁰, ex-vivo approaches that employ gene editing of bulk T cells or CD34+ hematopoietic stem and progenitor cells, as well as in vivo editing approaches that target the liver, require a two-step approach to first discover potential off-target sites and then validate sites with off-target edits through the detection off-target activity in edited cells^{34, 37}

[0367] The approaches we advance here comprise a comprehensive and sensitive suite of assays to discover and validate LSI off-target events to directly address regulatory and genotoxicity considerations for LSI-based strategies in gene and cell therapy approaches.

[0368] We demonstrate that HIDE-seq is a readily applicable and unbiased biochemicalbased WGS approach for identification of potential LSI off-target outcomes such as indels, cryptic integrations, and large structural variants. We increase the sensitivity of off-target discovery of cryptic integration events with Cryptic-seq, an ultra-sensitive biochemical approach that is greater than 1,200-fold more sensitive than HIDE-seq but is limited to detection of cryptic integrations. Cryptic-seq uses supraphy si ologi cal amounts of integrase and DNA template, significantly higher than a human cell would ever be exposed to in a therapeutic setting, followed by PCR enrichment, to identify a deep pool of off target sites for expansive analysis. We have demonstrated that this approach is extremely sensitive and has been specifically tuned to give a high false positive rate (>97%), increasing confidence that the total spectrum of potential off target sites with any integrase-dinucleotide combination will be discovered. Both HIDE-seq and Cryptic-seq can be adapted to accommodate multiple LSIs (i.e., multiplexing) or novelly designed LSIs. Importantly, both assays can be performed with genomic DNA isolated from therapeutically relevant cell types (e.g., human donor derived T cells, CD34+ hematopoietic stem and progenitor cells, primary human hepatocytes) or mixed populations of genomic DNA for more genetically relevant discovery of potential off-target sites. To readily perform scalable off-target validation of up to 1000s of potential LSI off-target sites in relevant edited cell types, we present a customized application of hybrid capture NGS that is tailored to accurately detect and quantify LSI off-target events, such as indels and cryptic integrations, which we benchmark to bulk WGS. In the context of the experimental data presented here, we demonstrate hybrid capture NGS has greater than 98% apparent validation hit rate (52/53 validated sites detected) compared to an apparent hit rate of WGS (at 40x coverage) of 9.4% (5/53 validated sites detected). By sharing our advances in molecular analytics for LSIs, which comprise powerful and promising tools for basic and clinical applications of programmable genomic insertion, we hope to help further the establishment of benchmarks for sensitivity of off-target detection in therapeutic paradigms that employ these fascinating enzymes.

6.5.3. Methods

6.5.3.1 HEK293aftB cell line generation

[0369] In-house produced lentivirus containing a transfer plasmid with an EFla-PuroR- WPRE backbone containing a 46 bp Bxbl attB insert (PL312) was transduced into HEK293 cells. Cells with Low MOI were plated in sterile 96 well plates under puromycin selection via serial dilutions for clone selection. Following clone selection, confirmation of lentiviral copy and identification of the lentivirus insertion site was performed using ligation mediated PCR with primers targeting the 5’ and 3’ LTRs along with Cergentis TLA mapping confirmed single lentiviral insertion on chromosome 5 (data not shown). Assessment of the insertion site by long read PCR with flanking genomic primers followed by Oxford Nanopore Technologies long read sequencing confirmed the lentiviral insert sequence, which had partially concatemerized to ultimately contain two Bxbl attB sequences, separated by 4,272 bp of intervening lentiviral sequence.

6.5.3.1 K562aftB cell line generation

[0370] Lentiviral plasmid (PL2811) with a PGK-EFla-PuroR backbone containing a 46 bp Bxbl attB inserts was acquired from GenScript and used for lentiviral production at Azenta. To generate engineered cell lines containing attB sequences, K562 cells (ATCC, Cat# CCL- 243) were transduced with lentivirus doses of 5 pL, 10 pL, 20 pL, 30 pL and 50 pL infected by spinfection at 1000g for 30 minutes at 33 °C to find appropriate dose of lentivirus with low multiplicity of infection (MOI). Cells with Low MOI were plated in sterile 96 well plates under puromycin selection via serial dilutions for clone selection. Following clone selection, confirmation of lentivirus insertion site was performed using Lenti-X™ Integration Site Analysis Kit (Takara) and WGS, which both confirmed the presence of single lentiviral integration events on chromosome 6 and 17.

6.5.3.3 HIDE-seq

[0371] HIDE-seq reactions with 8 pg of purified gDNA from PBMCs (Qiagen Blood and Tissue kit) or, in separate reactions, 8 pg of purified gDNA HEK293attB cells was incubated with 10 nM attB_66bp annealed oligonucleotide (for cryptic attP site or FDE discovery) or 10 nM attP_72bp annealed oligonucleotide (for cryptic attB site or FDE discovery) and 400 nM of recombinant Bxbl integrase (GenScript) in recombination buffer? (10 nM Tris-HCl, pH 8.0, 100 mM KC1, 5% glycerol) for 8 hours at 37°C and then held at 4°C. To anneal the attB_66 and attP_72 oligonucleotide substrates for HIDE-seq, 25 pl of attB_66bp_P_Top (100 mM) and 25 pl of attB_66bp_P_Bot (100 mM) or 25 pl of attP_72bp_P_Top (100 mM) and 25 pl of attP_72bp_P_Bot (100 mM), respectively were mixed together in wells of a 20 pl 96 well plate. Note, oligonucleotides were resuspended in duplex buffer (Integrated DNA technologies (IDT)). The plate with oligonucleotide samples were placed into a thermocycler, and the following anneal program was performed with a 105°C heated lid: 95°C/2 minutes, 750 cycles at -0.1°C per cycle/ls, 4°C hold. Reactions were inactivated with RNAase A (New England Biolabs (NEB) Monarch) and then Proteinase K (NEB) and gDNA was purified with 0.7x AMPure (Beckmann Coulter) bead cleanup. Control Digenome-seq reactions were performed with gDNA and spCas9 (IDT) with VEGFA S2 sgRNA (IDT) were performed according to the published protocol²¹. Isolated gDNA from both HIDE-seq and Digenome-seq reactions was quantified by Qubit fluorometry (Thermo Fisher Scientific) and submitted for PCR-free WGS at 30x coverage (Azenta). FASTQ files from Illumina sequencing were loaded into a custom bioinformatic pipeline for HIDE-seq called tbDigln (github.com/didacs/tbDigln) developed by Tome Biosciences and Fulcrum Genomics which quantified the number of unclipped reads (for FDE detection) or soft-clipped reads (recombined sites). The bioinformatic workflow for tbDigln starts with the alignment of sequencing reads from FASTQ files against the hg38 human reference genome41 with appended attP and attB sequences by BWA aligner42 to create mapped BAM files for each sample. The resulting BAM files are deduplicated by Picard⁴³ and inputted into HIDE, a modified Digenomitas²⁴ pipeline to output .csv files containing sites per sample as clipped counts for integration events and unclipped counts for FDEs.

6.5.3.4 Cryptic-seq

[0372] HEK293attB gDNA was incubated with Tn5 transposase²⁷ (custom purification by GenScript) pre-annealed with MES_Rev_3InvdR and MES_AmpSeq_P5 oligos) for 7 minutes at 55°C to tagment the gDNA and shear it to an average size of approximately 700bp. The Cryptic-seq donor plasmid (PL2312) includes an attP integrase attachment site with a GT dinucleotide flanked by a unique 12 bp barcode sequence (BC1) and Illumina N7 primer binding site (PBS) on the 5’ side of the attP and a unique barcode sequence (BC2) sequence on and an Illumina P7 sequencing primer binding site (PBS) on the 3’ side of the attP. The biochemical reaction was performed with 1 pg tagmented gDNA, 10 nM of the Cryptic-seq donor plasmid, and 1 pM integrase and incubating these components at 37°C for 4h in a recombination buffer?. The reaction was stopped by adding sodium dodecyl sulfate to a final concentration of 0.1% and the products were cleaned using the Zymo clean and concentrator kit (Zymo research).

[0373] PCR was used to amplify the regions of integration using Q5 polymerase (NEB) and primers for P5 containing distinct barcodes for each sample (LM P5 F1-F3) and a single N7 primer (GN037_CrypticSeq_N7_i7_N701). Following PCR, a 1.5x AMPure bead clean-up was performed, and products were run on a Tapestation 4200 (Agilent Technologies) using DI 000 tape to confirm amplification. Library concentration was determined using the Next Library Quant Kit (NEB) for Illumina. Libraries were normalized to 2nM, loaded at a concentration of 750pM and sequenced via Illumina Next Seq according to the manufacturer's instructions.

[0374] For bioinformatic analysis of cryptic-seq data, FASTQ files from Illumina sequencing were loaded into a custom cryptic-seq bioinformatic pipeline tbChaSIn (github.com/didacs/tbChaSIn) developed by Tome Biosciences and Fulcrum Genomics to discover and quantify cryptic recombination sites from Cryptic-seq data. The bioinformatic workflow starts with the trimming of reads that contain leading attP or attB sequences (P, P’ or B, B’) followed by a search and trim of any Tn5 mosaic end (ME) sequence (CTGTCTCTTATACACATCT (SEQ ID NO: 580), Illumina) in the reads. Reads trimmed for att and ME sequences are aligned against the hg38 human reference genome41 with appended attP and attB sequences by BWA aligner42 to create mapped BAM files for each sample. The resulting BAM files are deduplicated by Picard43 and queried for integration sites with to generate output .csv files containing sites per sample along with collation of sites across samples to generate output .csv files containing collated sites.

6.5.3.5 Hybrid Capture Quantitative Validation with DNA standards

[0375] Simulated LSR-mediated cargo integration for CAS031 DNA fragments were synthesized as gBlocks (IDT). gBlocks were designed to include approximately 200 bp left and right of the recombination junction (attL or attK) dinucleotide followed by approximately 1300 bp of random DNA stuff er sequence at each end to achieve a fragment length of approximately 3000 bp for optimal DNA fragmentation (CHR6_RC_Off_PL753_attL, CHR6_RC_Off_PL753_attR). Standard curve titrations were prepared by spiking in gBlocks into human gDNA (Promega catalog #G304A) at molar ratios representing 95%, 50%, 10%, 1%, 0.1%, 0.05%, 0.001%, and 0% I-PGI. Samples were fragmented to an average size of 550 bp via sonication (Covaris ME220) and NGS libraries prepared according to the IDT xGen DNA Library Prep Kit MC UNI (Version 2) protocol using xGen™ UDI-UMI Adapters (IDT catalog #10005903). Target enrichment was performed according to xGen™ hybridization capture of DNA libraries (Version 7) protocol using custom hybrid capture panels containing probes either left (5’) or right (3’) of CAS031. Hybrid capture libraries were sequenced by Illumina short-read sequencing via NextSeq 2000 paired-end (2 x 150 bp) sequencing using standard NextSeq 2000 P3 Reagents (300-Cycles) (Illumina catalog #20040561) and analyzed using a custom hybrid capture bioinformatics pipeline called tbREVEAL developed at Tome Biosciences (described below).

6.5.3.6 Editing of K562aftB cell lines

[0376] K562< //7> cells were maintained in RMPI1640 media (Gibco) +10% fetal bovine serum (Gibco) + 2ug/ml puromycin (Life Technologies) and passaged at ratio of 0.1 million cells per ml every 3-5 days. For editing, 200,000 cells were electroporated with 3 pg mRNA for Bxb 1 integrase and 3 pg of a donor construct containing attP integrase attachment sites and a GT central dinucleotide using the Lonza 4D-Nucleofector™ X Unit according to the manufacturer’s instructions for K562 cells. After 3 days, 80% of the cells were collected for gDNA extraction using the Qiagen blood and tissue kit and the remaining 20% were expanded then banked.

6.5.3.7 Hybrid capture NGS of edited K562a#B cells

[0377] Genomic DNA extracted from edited K562attB cells were used as input for hybridization capture NGS. A Covaris ME220 focused ultrasonicator was used to shear 400 ng of Rep 1, Rep 2, Control 1, and Control 2 gDNA samples to an average size of 400 bp. Fragmented DNA was end repaired and A-tailed followed by ligation of Illumina sequencing adapters using Twist library preparation kit (cat# 104177) to create Illumina paired end sequencing libraries. Hybridization, capture and post capture amplification of the prepared libraries to probes was performed using Twist target enrichment standard hybridization kit as per manufacturer’s instructions (Twist Library Preparation Kit 1, Mechanical Fragmentation, cat # 100876). Libraries were sequenced on Illumina NextSeq 2000 platform using a P2 600 cycle kit (Illumina) according to standard manufacturer’s protocol. [0378] For bioinformatic analysis of hybrid capture NGS data, FASTQ files from Illumina sequencing were loaded into a custom bioinformatic pipeline called tbREVEAL (github.com/jessie-wangjie/tbREVEAL) developed by Tome Biosciences to quantify edited reads corresponding to unedited reads, reads containing indels, reads containing cargo DNA sequence and thus represent recombined reads. Briefly, input data in FASTQ or BAM format undergo quality control using FASTP (github.com/OpenGene/fastp), which includes adapter trimming, quality filtering, and extraction of unique molecular identifiers (UMIs).

Additionally, paired-end reads are merged into a single file. The processed reads are aligned to the hg38 reference genome⁴¹ using BWA aligner⁴². UMIs are deduplicated from the aligned BAM files to remove PCR duplicates. Target information is collected and used to generate reference amplicons. Target reads are extracted from the deduplicated BAM files and aligned to the reference amplicons using BWA aligner⁴². Editing events are quantified using a custom Python script designed to analyze the aligned target reads. Visualization of editing events is performed using a custom script to generate graphical representations. Editing sites are collated across samples are outputted into an excel spreadsheet using a custom script. HTML reports by papermill (papermill. readthedocs.io/en/latest/index.html) are generated to summarize the results of the analysis.

6.5.3.8 Whole genome sequencing of edited K562aftB cells

[0379] For WGS, Bxbl integrase-edited Rep 1 and Rep 2 K562a//// gDNA genomic DNA was extracted using DNeasy blood and tissue kit (Qiagen) and samples library preparation and Illumina short read sequencing with a target of 60x genomic coverage. For library preparation, genomic DNA was sheared using a Covaris ME220 sonicator to average size of 400 bp, sheared fragments were end repaired, A-tailed and ligated with Illumina sequencing adapters and the resulting library was sequenced on an Illumina NovaSeq platform (Azenta). For bioinformatic analysis of WGS data, FASTQ files from Illumina sequencing were mapped to the hg38 reference genome and cargo reference using BWA⁴² followed by fgsv²⁶ from Fulcrum Genomics was used to predict potential chimeric reads. Chimeric reads with one breakpoint at the reference genome and with the second breakpoint at the attP dinucleotides in the DNA cargo reference were retained for further investigation.

6.5.3.9 Data and Software Availability

[0380] The HIDE-seq, Cryptic-seq, and Hybrid Capture analysis pipelines are accessible through GitHub at github.com/didacs/tbDigln, github.com/didacs/tbChaSIn, github.com/jessie-wangjie/tbREVEAL, respectively. The project number for the nextgeneration sequencing data reported in this paper is NCBI SRA: PRJNA1167015.

6.5.4. Materials

6.5.5. References

1. Stark, W.M. The Serine Recombinases. Microbiol Spectr 2 (2014).

2. Smith, M.C.M. Phage-encoded Serine Integrases and Other Large Serine Recombinases. Microbiol Spectr 3 (2015).

3. Grindley, N.D., Whiteson, K.L. & Rice, P.A. Mechanisms of site-specific recombination. Annu Rev Biochem 75, 567-605 (2006).

4. Groth, A.C. & Calos, M.P. Phage integrases: biology and applications. J Mol Biol 335, 667-678 (2004). 5. Thorpe, H.M. & Smith, M.C. In vitro site-specific integration of bacteriophage DNA catalyzed by a recombinase of the resolvase/invertase family. Proc Natl Acad Set U S A 95, 5505-5510 (1998).

6. Kim, A.I. et al. Mycobacteriophage Bxbl integrates into the Mycobacterium smegmatis groELl gene. Mol Microbiol 50, 463-473 (2003).

7. Wang, X. et al. Bxbl integrase serves as a highly efficient DNA recombinase in rapid metabolite pathway assembly. Acta Biochim Biophys Sin (Shanghai) 49, 44-50 (2017).

8. Xu, Z. et al. Accuracy and efficiency define Bxbl integrase as the best of fifteen candidate serine recombinases for the integration of DNA into the human genome. BMC Biotechnol 13, 87 (2013).

9. Groth, A.C., Olivares, E.C., Thyagarajan, B. & Calos, M.P. A phage integrase directs efficient site-specific integration in human cells. Proc Natl Acad Sci U S A 97, 5995-6000 (2000).

10. Farruggio, A.P., Bhakta, M.S., du Bois, H., Ma, J. & M, P.C. Genomic integration of the full-length dystrophin coding sequence in Duchenne muscular dystrophy induced pluripotent stem cells. Biotechnol J 12 (2017).

11. Anzalone, A.V. et al. Programmable deletion, replacement, integration and inversion of large DNA sequences with twin prime editing. Nat Biotechnol 40, 731-740 (2022).

12. Durrant, M.G. et al. Systematic discovery of recombinases for efficient integration of large DNA sequences into the human genome. Nat Biotechnol 41, 488-499 (2023).

13. Blanch- Asensio, A. et al. STRAIGHT-IN enables high-throughput targeting of large DNA payloads in human pluripotent stem cells. Cell Rep Methods 2, 100300 (2022).

14. Yarnall, M.T.N. et al. Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases. Nat Biotechnol 41, 500-512 (2023).

15. Ghosh, P., Kim, A.I. & Hatfull, G.F. The orientation of mycobacteriophage Bxbl integration is solely dependent on the central dinucleotide of attP and attB. Mol Cell 12, 1101-1111 (2003).

16. Olorunniji, F.J. et al. Control of serine integrase recombination directionality by fusion with the directionality factor. Nucleic Acids Res 45, 8635-8645 (2017). 17. Chalberg, T.W. et al. Integration specificity of phage phiC31 integrase in the human genome. J Mol Biol 357, 28-48 (2006).

18. Pandey, S. et al. Efficient site-specific integration of large genes in mammalian cells via continuously evolved recombinases and prime editing. Nat Biomed Eng (2024).

19. Nambiar, T.S., Baudrier, L., Billon, P. & Ciccia, A. CRISPR-based genome editing through the lens of DNA repair. Mol Cell 82, 348-388 (2022).

20. Ehrhardt, A., Engler, J. A., Xu, H., Cherry, A.M. & Kay, M.A. Molecular analysis of chromosomal rearrangements in mammalian cells after phiC31 -mediated integration. Hum Gene Ther 17, 1077-1094 (2006).

21. Kim, D. et al. Digenome-seq: genome-wide profiling of CRISPR-Cas9 off-target effects in human cells. Nat Methods 12, 237-243, 231 p following 243 (2015).

22. Mediavilla, J. et al. Genome organization and characterization of mycobacteriophage Bxbl. Mol Microbiol 38, 955-970 (2000).

23. Tsai, S.Q. et al. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases. Nat Biotechnol 33, 187-197 (2015).

24. Maeder, M.L. et al. Development of a gene-editing approach to restore vision loss in Leber congenital amaurosis type 10. Nat Med 25, 229-233 (2019).

25. Robinson, J.T. et al. Integrative genomics viewer. Nat Biotechnol 29, 24-26 (2011).

26. Nils Homer, P.R., Tim Fennell, Clint Valentine, John Didion, Matt Stone fulcrumgenomics/fgsv: Release 0.2.0 (0.2.0). i (2024).

27. Picelli, S. et al. Tn5 transposase and tagmentation procedures for massively scaled sequencing projects. Genome Res 24, 2033-2040 (2014).

28. Giannoukos, G. et al. UDiTaS, a genome editing detection method for indels and genome rearrangements. BMC Genomics 19, 212 (2018).

29. Hennig, B.P. et al. Large-Scale Low-Cost NGS Library Preparation Using a Robust Tn5 Purification and Tagmentation Protocol. G3 (Bethesda) 8, 79-89 (2018).

30. Findlay, S.D., Vincent, K.M., Berman, J.R. & Postovit, L.M. A Digital PCR-Based Method for Efficient and Highly Specific Screening of Genome Edited Cells. PLoS One 11, e0153901 (2016). 31. Amit, I. et al. CRISPECTOR provides accurate estimation of genome editing translocation and off-target activity from comparative NGS data. Nat Commun 12, 3042 (2021).

32. Gnirke, A. et al. Solution hybrid selection with ultra-long oligonucleotides for massively parallel targeted sequencing. Nat Biotechnol 27, 182-189 (2009).

33. Chaudhari, H.G. et al. Evaluation of Homology -Independent CRISPR-Cas9 Off- Target Assessment Methods. CRISPR J 3, 440-453 (2020).

34. Frangoul, H. et al. CRISPR-Cas9 Gene Editing for Sickle Cell Disease and betaThalassemia. N Engl J Med 384, 252-260 (2021).

35. Doench, J.G. et al. Optimized sgRNA design to maximize activity and minimize off- target effects of CRISPR-Cas9. Nat Biotechnol 34, 184-191 (2016).

36. Kasai, F., Mizukoshi, K. & Nakamura, Y. Variable characteristics overlooked in human K-562 leukemia cell lines with a common signature. Sci Rep 14, 9619 (2024).

37. Gillmore, J.D. et al. CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis. N Engl J MedP&5, 493-502 (2021).

38. Nan, A.X. et al. Ligase-mediated programmable genomic integration (L-PGI): an efficient site-specific gene editing system that overcomes the limitations of reverse transcriptase-based editing systems. bioRxiv, 2024.2009.2027.615478 (2024).

39. Menz, J. et al. Genotoxicity assessment: opportunities, challenges and perspectives for quantitative evaluations of dose-response data. Arch Toxicol 97, 2303-2328 (2023).

40. Yamanaka, S. Pluripotent Stem Cell-Based Cell Therapy -Promise and Challenges. Cell Stem Cell 21, 523-531 (2020).

41. Schneider, V.A. et al. Evaluation of GRCh38 and de novo haploid genome assemblies demonstrates the enduring quality of the reference assembly. Genome Res 27, 849-864 (2017).

42. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754-1760 (2009).

43. broadinstitute. github.io/picard/ (2019).

7. EQUIVALENTS AND INCORPORATION BY REFERENCE [0381] All references cited herein are incorporated by reference to the same extent as if each individual publication, database entry (e.g. Genbank sequences or GenelD entries), patent application, or patent, was specifically and individually indicated incorporated by reference in its entirety, for all purposes. This statement of incorporation by reference is intended by Applicants, pursuant to 37 C.F.R. §1.57(b)(1), to relate to each and every individual publication, database entry (e.g. Genbank sequences or GenelD entries), patent application, or patent, each of which is clearly identified in compliance with 37 C.F.R. § 1.57(b)(2), even if such citation is not immediately adjacent to a dedicated statement of incorporation by reference. The inclusion of dedicated statements of incorporation by reference, if any, within the specification does not in any way weaken this general statement of incorporation by reference. Citation of the references herein is not intended as an admission that the reference is pertinent prior art, nor does it constitute any admission as to the contents or date of these publications or documents.

[0382] It is an object of the invention not to encompass within the invention any previously known product, process of making the product, or method of using the product such that Applicant reserves the right and hereby disclose a disclaimer of any previously known product, process, or method. It is further noted that the invention does not intend to encompass within the scope of the invention any product, process, or making of the product or method of using the product, which does not meet the written description and enablement requirements of the USPTO (35 U.S.C. 112(a)) or the EPO (Article 83 of the EPC), such that Applicant reserves the right and hereby disclose a disclaimer of any previously described product, process of making the product, or method of using the product. It may be advantageous in the practice of the invention to be in compliance with Art. 53(c) EPC and Rule 28(b) and (c) EPC. Nothing herein is to be construed as a promise. It is noted that in this disclosure and particularly in the claims and/or paragraphs, terms such as "comprises", "comprised", "comprising" and the like can have the meaning attributed to it in U.S. Patent law; e.g., they can mean "includes", "included", "including", and the like; and that terms such as "consisting essentially of' and "consists essentially of' have the meaning ascribed to them in U.S. Patent law, e.g., they allow for elements not explicitly recited, but exclude elements that are found in the prior art or that affect a basic or novel characteristic of the invention.

[0383] While the invention has been particularly shown and described with reference to a preferred embodiment and various alternate embodiments, it is understood by persons skilled in the relevant art that various changes in form and details can be made therein without departing from the spirit and scope of the invention.

Claims

WHAT IS CLAIMED IS:

1. A method for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) in a target genome, the method comprising:

(a) fragmenting genomic DNA isolated from target cells;

(b) tagging the 5’ and 3’ ends of the DNA fragments with a first oligonucleotide adapter;

(c) contacting the tagged genomic DNA fragments with

(i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, and a second oligonucleotide adapter;

(ii) optionally at least a second species of synthetic DNA molecule, wherein the second species of synthetic DNA molecule comprises a first attachment site known to be recognized by the selected LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence in the central dinucleotide from the first attachment site of the first species of synthetic DNA molecule, and a second oligonucleotide adapter; and

(iii) the selected LSI, under conditions suitable for the LSI to effect the recombination of the synthetic DNA molecule into the tagged genomic DNA fragments at second attachment sites that are capable of functioning as cognates of the first attachment site;

(d) generating a sequencing library comprising synthetic DNA molecules recombined with the tagged genomic DNA fragments;

(e) sequencing the sequencing library; and

(f) identifying cryptic attachment sites based on the sequencing data.

2. The method of claim 1, wherein the first oligonucleotide adapter comprises a primer binding site.

3. The method of claim 1 or 2, wherein the second oligonucleotide adapter comprises a primer binding site.

4. The method of claim 2 or 3, wherein the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating PCR amplification.

5. The method of any one of claims 2 to 4, wherein the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating sequencing-by-synthesis.

6. The method of any one of claims 2 to 5, wherein the primer binding site of the first oligonucleotide adapter, the primer binding site of the second oligonucleotide adapter, or both, are capable of mediating PCR amplification and sequencing-by-synthesis.

7. The method of any one of claims 1 to 6, wherein the synthetic DNA molecule further comprises a unique molecular identifier (UMI).

8. The method of claim 7, wherein the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI.

9. The method of claim 7 or 8, wherein the UMI is positioned in the synthetic DNA molecule between the first attachment site and the second oligonucleotide adapter.

10. The method of any one of claims 1 to 9, wherein the synthetic DNA molecule further comprises a barcode.

11. The method of claim 10, wherein the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual-indexed barcode.

12. The method of any one of claims 1 to 11, wherein the synthetic DNA molecule is circular.

13. The method of any one of claims 1 to 11, wherein the synthetic DNA molecule is linear.

14. The method of any one of claims 1 to 13, wherein the large serine integrase is Bxbl.

15. The method of any one of claims 1 to 14, wherein the first attachment site is an attP or a modified attP site.

16. The method of any one of claims 1 to 14, wherein the first attachment site is an attB or a modified attB site.

17. The method of any one of claims 1 to 16, wherein generating a sequencing library comprises an initial step of amplifying the tagged genomic DNA fragment after recombination with the synthetic DNA molecule.

18. The method of any one of claims 1 to 17, wherein identifying a cryptic attachment site comprises determining the sequence of the genomic site into which the DNA molecule has been recombined or the sequence of DNA flanking the cryptic attachment site.

19. The method of any one of claims 1 to 18, wherein identifying a cryptic attachment site comprises detecting an attL-genomic DNA junction, an attR-genomic DNA junction, or both.

20. The method of any one of claims 1 to 19, wherein identifying a cryptic attachment site comprises: aligning the sequencing data to a reference genome; detecting an attachment site-genomic DNA junction; and reporting the coordinates and using the coordinates to identify cryptic attachment site.

21. The method of claim 20, wherein detecting an attachment site-genomic DNA junction comprises detecting an attL-genomic DNA junction, an attR-genomic DNA junction, or both.

22. The method of claim 20, wherein reporting the coordinates comprises ranking the coordinates based on sequencing reads de-duplicated from PCR amplification using the UMIs.

23. The method of claim 22, wherein the UMI count represents the recombination efficacy of the first attachment site.

24. The method of any one of claims 1 to 23, wherein the identified cryptic attachment site was not previously known to be an attachment site of the selected LSI.

25. A multiplexed method for identifying cryptic attachment sites that are separately recognizable by each of a plurality of large serine integrases (LSIs) in a target genome, the method comprising:

(a) fragmenting genomic DNA isolated from target cells; (b) tagging the 5’ and 3’ ends of the DNA fragments with a first oligonucleotide adapter;

(c) contacting the tagged genomic DNA fragments with

(i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule comprises a first attachment site known to be recognized by a first LSI, and a second oligonucleotide adapter;

(ii) at least a second species of synthetic DNA molecule, wherein the second species of synthetic DNA molecule comprises a first attachment site known to be recognized by a second LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence from the first attachment site of the first species of synthetic DNA molecule, and a second oligonucleotide adapter; and

(iii) the first LSI and the at least second LSI, under conditions suitable for the LSIs to effect the recombination of the synthetic DNA molecule into the tagged genomic DNA fragments at second attachment sites that are capable of functioning respectively as cognates of the respective first attachment sites;

(d) generating a sequencing library comprising synthetic DNA molecules recombined with the tagged genomic DNA fragments;

(e) sequencing the sequencing library; and

(f) identifying cryptic attachment sites for each of the plurality of LSIs based on the sequencing data.

26. A synthetic DNA molecule for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI) in a target genome, comprising: a first attachment site and an oligonucleotide adapter, wherein the first attachment site is known to be recognized by the selected LSI.

27. The synthetic DNA molecule of claim 26, wherein the oligonucleotide adapter comprises a primer binding site.

28. The synthetic DNA molecule of claim 27, wherein the primer binding site is capable of mediating PCR amplification.

29. The synthetic DNA molecule of claim 27, wherein the primer binding site is capable of mediating sequencing-by-synthesis.

30. The synthetic DNA molecule of claim 28 or 29, wherein the primer binding site is capable of mediating PCR amplification and sequencing-by-synthesis.

31. The synthetic DNA molecule of any one of claims 26 to 30, wherein the synthetic DNA molecule further comprises a unique molecular identifier (UMI).

32. The synthetic DNA molecule of claim 31, wherein the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI.

33. The synthetic DNA molecule of claim 31 or 32, wherein the UMI is positioned in the synthetic DNA molecule between the first attachment site and the oligonucleotide adapter.

34. The synthetic DNA molecule of any one of claims 26 to 33, wherein the synthetic DNA molecule further comprises a barcode.

35. The synthetic DNA molecule of claim 34, wherein the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual -indexed barcode.

36. The synthetic DNA molecule of any one of claims 26 to 35, wherein the synthetic DNA molecule is circular.

37. The synthetic DNA molecule of any one of claims 26 to 35, wherein the synthetic DNA molecule is linear.

38. The synthetic DNA molecule of any one of claims 26 to 37, wherein the large serine integrase is Bxbl.

39. The synthetic DNA molecule of any one of claims 26 to 38, wherein the first integration recognition site is an attP or a modified attP site.

40. The synthetic DNA molecule of any one of claims 26 to 38, wherein the first integration recognition site is an attB or a modified attB site.

41. A set of synthetic DNA molecules for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI), comprising: a first species of synthetic DNA molecule, comprising a first attachment site known to be recognized by the selected LSI, an oligonucleotide adapter, and a first DNA barcode; and at least a second species of synthetic DNA molecule, comprising a first attachment site known to be recognized by the selected LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence in the central dinucleotide from the first attachment site of the first species of synthetic DNA molecule, an oligonucleotide adapter, and a second DNA barcode; wherein the first DNA barcode and the second DNA barcode allow for multiplexing of multiple first attachment sites of different sequences in one reaction.

42. The set of synthetic DNA molecules of claim 41, wherein the oligonucleotide adapter comprises a primer binding site.

43. The set of synthetic DNA molecules of claim 42, wherein the primer binding site is capable of mediating PCR amplification.

44. The set of synthetic DNA molecules of claim 42, wherein the primer binding site is capable of mediating sequencing-by-synthesis.

45. The set of synthetic DNA molecules of claim 43 or 44, wherein the primer binding site is capable of mediating PCR amplification and sequencing-by-synthesis.

46. The set of synthetic DNA molecules of any one of claims 41 to 45, wherein the synthetic DNA molecule further comprises a unique molecular identifier (UMI).

47. The set of synthetic DNA molecules of claim 46, wherein the UMI is selected from: a random UMI, a variable length UMI, a variable length non-random UMI, a fixed length random UMI, a dual-indexed UMI, and a duplex UMI.

48. The set of synthetic DNA molecules of claim 46 or 47, wherein the UMI is positioned in the synthetic DNA molecule between the attachment site and the oligonucleotide adapter.

49. The set of synthetic DNA molecules of any one of claims 41 to 48, wherein the barcode is selected from: a predefined set of DNA barcodes, a variable length barcode, and a dual-indexed barcode.

50. The set of synthetic DNA molecules of any one of claims 41 to 49, wherein the synthetic DNA molecule is circular.

51. The set of synthetic DNA molecules of any one of claims 41 to 49, wherein the synthetic DNA molecule is linear.

52. The set of synthetic DNA molecules of any one of claims 41 to 51, wherein the large serine integrase is Bxbl.

53. The set of synthetic DNA molecules of any one of claims 41 to 52, wherein the first integration recognition site is an attP or a modified attP site.

54. The set of synthetic DNA molecules of any one of claims 41 to 52, wherein the first integration recognition site is an attB or a modified attB site.

55. A set of synthetic DNA molecules for identifying cryptic attachment sites that are separately recognizable by each of a plurality of large serine integrases (LSIs), comprising: a first species of synthetic DNA molecule, comprising a first attachment site known to be recognized by a first LSI, an oligonucleotide adapter, and a first DNA barcode; and at least a second species of synthetic DNA molecule, comprising a first attachment site known to be recognized by a second LSI, wherein the first attachment site of the at least second species of synthetic DNA molecule is different in sequence from the first attachment site of the first species of synthetic DNA molecule, an oligonucleotide adapter, and a second DNA barcode; wherein the first DNA barcode and the second DNA barcode allow for identifying the respective cryptic attachment sites of each of the plurality of LSIs in one reaction.

56. A kit for identifying cryptic attachment sites that are recognizable by a selected large serine integrase (LSI), comprising:

(a) the synthetic DNA molecule of any one of claims 26 to 40 or the set of synthetic DNA molecules of any one of claims 41 to 54; and

(b) instructions for performing any one of the methods of claims 1 to 24.

57. A kit for identifying cryptic attachment sites that are recognizable by each of a plurality of large serine integrases (LSIs), comprising:

(a) the synthetic DNA molecule of any one of claims 26 to 40 or the set of synthetic DNA molecules of claim 55; and

(b) instructions for performing the method of claim 25.

58. A method for identifying DNA recombination events of a selected integrase in a target genome, the method comprising:

(a) contacting genomic DNA isolated from target cells with

(i) a first species of synthetic DNA molecule, wherein the first species of synthetic DNA molecule is linear and comprises a first attachment site known to be recognized by the selected integrase, and

(ii) the selected integrase, under conditions suitable for the integrase to effect the recombination of the genomic DNA;

(b) generating a sequencing library comprising the recombined genomic DNA;

(c) sequencing the sequencing library; and

(d) detecting DNA recombination events based on the sequencing data.

59. The method of claim 58, wherein the DNA recombination events are selected from: a DNA double strand break, a recombination of the attachment site with a cryptic site in the genomic DNA, and a recombination between two cryptic sites in the genomic DNA.

60. The method of claim 58 or 59, wherein the integrase is a large serine integrase (LSI).

61. The method of claim 60, wherein the large serine integrase is Bxbl.

62. The method of any one of claims 58 to 61, wherein the first integration recognition site is an attP or a modified attP site.

63. The method of any one of claims 58 to 61, wherein the first integration recognition site is an attB or a modified attB site.

64. The method of any one of claims 58 to 63, wherein generating a sequencing library does not comprise amplifying the genomic DNA after recombination of the synthetic DNA molecule.

65. The method of any one of claims 58 to 64, wherein generating a sequencing library comprises generating a whole genome sequencing library.

66. The method of any one of claims 58 to 65, wherein sequencing the sequencing library comprises whole genome sequencing.

67. The method of any one of claims 58 to 66, wherein detecting an DNA recombination event comprising: aligning the sequencing data to a reference genome; and detecting a DNA double strand break, a recombination of the integration recognition site with a cryptic site in the genomic DNA, or a recombination between two cryptic sites in the genomic DNA.

68. A synthetic DNA molecule for identifying DNA recombination events of a selected integrase in a target genome, comprising: a first attachment site, wherein the first attachment site is known to be recognized by the selected integrase, wherein the synthetic DNA molecule is linear.

69. The synthetic DNA molecule of claim 68, wherein the DNA recombination events are selected from: a DNA double strand break, a recombination of the attachment site with a cryptic site in the genomic DNA, and a recombination between two cryptic sites in the genomic DNA.

70. The synthetic DNA molecule of claim 68 or 69, wherein the integrase is a large serine integrase (LSI).

71. The synthetic DNA molecule of claim 70, wherein the large serine integrase is Bxbl.

72. The synthetic DNA molecule of any one of claims 68 to 71, wherein the first integration recognition site is an attP or a modified attP site.

73. The synthetic DNA molecule of any one of claims 68 to 71, wherein the first integration recognition site is an attB or a modified attB site.

74. A kit for identifying DNA recombination events of a selected integrase, comprising:

(a) the synthetic DNA molecule of any one of claims 68 to 73; and

(b) instructions for performing any one of the methods of claims 58 to 67.