CRISPR-associated transposases and methods of use thereof

Abstract

Described herein are improved CRISPR-associated transposases (CASTs), including homing endonuclease-assisted large-array integration CRISPR-associated transposase (CAST) complexes, and methods for their use, as well as other strategies for improving the activity of native and engineered CASTs.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Application Publication No. 63 / 285,857, filed December 3, 2021, U.S. Application No. 63 / 291,264, filed December 17, 2021, and U.S. Application No. 63 / 411,735, filed September 30, 2022, the contents of which are incorporated by reference in their entireties herein.

[0002] Described herein is an improved homing endonuclease-assisted large-scale sequence integration CRISPR-associated transposase (CAST), including CAST and methods of using the same. [Background technology]

[0003] Programmable insertion of multi-kilobase DNA sequences into genomes, independent of homologous recombination and double-strand breaks (DSBs), offers new possibilities for precise genome editing. Methods for genome integration are usually viral vectors. 1、2 Or transposon 3~7 (both of which lack programmability and therefore insert stochastically throughout the genome) or nucleases coupled to a DNA donor that rely on cytotoxic DSBs and host homologous recombination factors. 8~10 In addition, recombination systems in bacteria rely on the cointegration of selectable markers. 12 or CRISPR-Cas counter selection 13 Without it, efficiency is low 11 CRISPR-associated transposase (CAST) is a promising new approach for programmable recombination-independent DNA insertion through the interaction between a transposase protein and CRISPR-Cas effector(s) to direct RNA-guided transposition. 14~16 . Summary of the Invention

[0004] CRISPR-associated transposases (CASTs) allow recombination-independent multi-kilobase DNA insertion at RNA-programmed genomic locations. VK-type CASTs offer clear technical advantages over I-type CASTs, given their smaller code size, fewer components, and unidirectional insertion. However, the utility of VK-type CASTs is hindered by high off-target integration, as well as a replicative transposition mechanism that results in a mixture of desirable simple cargo insertions and undesired plasmid cointegrate products. Here, we overcome both limitations by engineering a new CAST with improved integration product purity and genome-wide specificity. To do so, we compensate for the absence of the TnsA subunit in VK-type CASTs by engineering a homing endonuclease-assisted large sequence integration CAST complex (HELIX) that utilizes a nicking homing endonuclease (nHE) fused to TnsB to restore the 5' nicking ability required for cargo excision on the DNA donor. HELIX allows DNA insertion to be cut and pasted with up to 99.4% simple insertion product purity while retaining robust integration efficiency to genome target. We generate and characterize functional fusions between CAST subunits and demonstrate that HELIX has substantially higher on-target specificity compared to standard CAST. Furthermore, we identify fusion proteins and host factors that enhance the on-target specificity of HELIX, reducing the off-target integration profile to a level comparable to that of type I system. We also demonstrate the extendibility of HELIX to other VK-type orthologs, and the feasibility of CAST- and HELIX-mediated DNA insertion in human cell lysates and human cells. By leveraging the distinct features of both VK-type and type I systems, HELIX streamlines and improves the application of CRISPR-based transposition technology, eliminating the barriers for efficient and specific RNA-guided DNA insertion.

[0005] Thus, provided herein is a fusion protein comprising a translocation protein B (TnsB) protein, e.g., Tn7, Tn7-like, or Tn5053-like translocation protein B (TnsB), fused (optionally via an intervening linker) to a protein (e.g., an endonuclease, e.g., a nickase, a cleavase, or a catalytically inactive endonuclease, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)). In some embodiments, the endonuclease is a nickase, e.g., a homing endonuclease (HE), a nicking restriction endonuclease, a nicking Cas variant, or a phage HNH endonuclease, or a TnsA or Tn7 transposon from type I CAST, or a catalytic portion thereof. In some embodiments, the HE is a LAGLIDADG, HNH, His-Cys box, or a GIY-YIG HE. In some embodiments, the HE is I-AniI, e.g., I-AniI from Aspergillus nidulans (I-AniI) or a variant thereof, optionally including a K227M mutation (nAniI), a hyperactive variant (e.g., Y2 I-AniI (F13Y, SI11Y)), or both (K227M, F13Y, SI11Y). Also provided in some embodiments is a nucleic acid comprising a sequence encoding a fusion protein as described. Also provided is an expression construct comprising a nucleic acid as described and a regulatory sequence, e.g., a promoter, for expressing the protein.

[0006] In some embodiments, an expression construct is provided that includes a sequence encoding CRISPR-associated transposase (CAST), the sequence comprising a nucleic acid encoding a fusion protein as described, Cas12k; TnsC; TniQ; optionally one or more host proteins; and a guide RNA (gRNA) that interacts with Cas12k and directs Cas12k / gRNA complex to a target sequence, and a regulatory sequence for expressing the sequence, for example, one or more promoter sequences. In some embodiments, Cas12k is fused to at least one other protein, optionally TniQ and / or TnsC, optionally with a linker between each protein (e.g., Cas12k-TniQ, Cas12k-TniQ-TniQ, Cas12k-TnsC, Cas12k-TniQ-TnsC, or Cas12k-TnsC-TniQ). In some embodiments, the expression construct is a plasmid or a viral vector.

[0007] Also provided in some embodiments are host cells comprising and optionally expressing a nucleic acid as described comprising a nucleic acid sequence encoding a Tn-endonuclease fusion protein, e.g., a TnsB-endonuclease fusion protein; and optionally one or more, e.g., all, of Cas12k; TnsC; TniQ; optionally one or more host proteins; and a guide RNA that binds to Cas12k and directs the TnsB-endonuclease fusion protein to a selected target sequence, or a host cell comprising a CRISPR-associated transposase (CAST) comprising a fusion protein as described; Cas12k; TnsC; TniQ; optionally one or more host proteins; and a gRNA that interacts with Cas12k and directs the fusion protein to a selected target sequence. In some embodiments, Cas12k is fused to at least one other protein, optionally TniQ (e.g., Cas12k-TniQ, TniQ-Cas12k, TniQ-TniQ-Cas12k, TniQ-Cas12k-TniQ, or Cas12k-TniQ-TniQ) and / or at least one TnsC, optionally with a linker between each protein.

[0008] Also provided is a method for inserting a desired sequence into DNA, for example genomic DNA of a living cell, the method comprising expressing in the cell the nucleic acid of claim 5; Cas12k; TnsC; TniQ; optionally one or more host proteins; and a guide RNA that binds to cas12k and directs the endonuclease to a selected target sequence, and a donor DNA molecule (e.g., a plasmid) comprising the desired sequence to be inserted, wherein the desired sequence is flanked at the 5' and 3' ends, respectively, by LE and RE flanking sequences, and a target site for the endonuclease (e.g., I-AniI), preferably oriented to impart a nick on the donor plasmid 5' of the desired sequence to be inserted. In some embodiments, the donor DNA molecule has modified LE / RE flanking sequences, e.g., flanking sequences as shown in Table A, that are derived from a source organism other than the source organism of at least one of the CAST components, i.e., TnsB; cas12k; TnsC; or TniQ, and / or contain modifications or insertions at various distances from the LE and RE sequences (e.g., endonuclease recognition sequence or host factor binding sequence(s)). In some embodiments, the modified LE / RE flanking sequences are derived from Scytonema hofmannii (e.g., from ShCAST), and at least one of the Tn proteins; cas12k; TnsC; or TniQ are derived from a CAST or HELIX ortholog (e.g., AcCAST and AcHELIX), are modified ShCAST LE / RE flanking sequences, or are de novo LE / RE flanking sequences. In some embodiments, Cas12k is expressed as a fusion protein with at least one TniQ and / or at least one TnsC (e.g., Cas12k-TniQ, Cas12k-TniQ-TniQ, Cas12k-TnsC, Cas12k-TniQ-TnsC, or Cas12k-TnsC-TniQ), optionally with a linker between each protein.

[0009] Also provided are fusion proteins comprising Cas12k; optionally one or more host proteins; and at least one TniQ (e.g., Cas12k-TniQ or Cas12k-TniQ-TniQ) and / or at least one TnsC, optionally with linkers between each segment.

[0010] Also provided are fusion proteins comprising a host protein and one or more of Cas12k, TnsC, or TniQ, optionally with linkers between each segment.

[0011] Also provided are compositions comprising, or nucleic acids encoding, (i) fusion proteins comprising a transposon (Tn) protein, e.g., Tn7, Tn7-like, or Tn5053-like, e.g., transposition protein B (TnsB), fused to a protein (e.g., an endonuclease, e.g., a nickase, cleavase, or catalytically inactive endonuclease, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)), optionally via an intervening linker; and (ii) fusion proteins comprising a host protein and one or more of Cas12k, TnsC, or TniQ, optionally with a linker between each segment.

[0012] Also provided are compositions comprising, or nucleic acids encoding, (i) fusion proteins comprising a transposon (Tn) protein, e.g., Tn7, Tn7-like, or Tn5053-like, e.g., transposition protein B (TnsB), fused to a protein (e.g., an endonuclease, e.g., a nickase, cleavase, or catalytically inactive endonuclease, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)), optionally via an intervening linker; and (ii) fusion proteins comprising a host protein and one or more of Cas12k, TnsC, or TniQ, optionally with a linker between each segment.

[0013] In some embodiments, the host factor is a ribosomal protein S15 that alters DNA topology (e.g., a pi protein or a nucleoid-associated protein (NAP), such as HU, Fis, H-NS, IHF, or TF1), or the host factor is involved in DNA or cellular metabolism, protein degradation or protein folding, regulation, or transport (e.g., acyl carrier protein (ACP), Sigma S, DnaN, DnaA, DNA topoisomerase I, La protease, Dam methylase, or proteins expressed from the genes dcd, dinD, radA, recQ, clpX, fkpA, hflX, crl, rseB, rsxE, araJ, melB, mgtA, aspA, treC, proY, serA, yhbC, yidA, ykfA).

[0014] Also provided is a host cell containing or expressing a composition according to any one of claims 18 to 20 and a donor DNA molecule (e.g. a plasmid) comprising a desired sequence to be inserted, wherein the desired sequence is flanked at its 5' and 3' ends, respectively, by LE and RE flanking sequences and a target site for an endonuclease (e.g. I-AniI), preferably oriented so as to nick the donor plasmid 5' of the desired sequence to be inserted.

[0015] Unless otherwise stated, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.Methods and materials are described herein for use in the present invention, and other suitable methods and materials known in the art can also be used.Materials, methods, and examples are merely illustrative and are not intended to be limiting.All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety.In case of conflict, the present specification, including definitions, will control.

[0016] Other features and advantages of the invention will be apparent from the following detailed description and drawings, and from the claims.

[0017] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with the color drawing(s) will be provided by the Office upon request and payment of the necessary fee. [Brief description of the drawings]

[0018] [Figure 1-1]Development and characterization of HELIX. a-c, Schematic diagram of type I and VK CAST and HELIX (panels a-c, respectively) and their transposition mechanisms resulting in simple insertion or cointegrate gene products. d, Workflow of transposition experiments targeting plasmid substrates. e, Transposition assessed by junction PCR across the LE / RE at TS1 in pTarget. Experiments were performed with nAniI fused to the N- or C-terminus of TnsB when using pDonor without an I-AniI site. f, Quantification of DNA integration efficiency for plasmid and donor plasmids using ShHELIX with various distances (d) between the I-AniI site and the LE / RE assessed by ddPCR using miniprep DNA. g, Coverage of expected insertion products into pTarget from long-read sequencing using exemplary simple insertion reads for ShHELIX and a subset of cointegrate reads for ShCAST (coverage from ShHELIX cointegrate reads and ShCAST simple insertion reads omitted for simplicity). h, Read length distribution using ShCAST and ShHELIX with sgRNA targeting TS1 on pTarget from long-read sequencing data. The top right panel is a zoomed-in view of the read length peak at approximately 8,000 bp. i, Comparison of the proportion of simple insertion and cointegrate products of transposition products for ShCAST and ShHELIX constructs using pDonor oriented to impart a 5' nick, 14 bp from the LE / RE, as assessed by long-read sequencing. j, k, Transposition product purity (panel j) and CFU (panel k) using the Lib4 I-AniI site on pDonor (with a 14 bp distance between the Lib4 site and the LE / RE), which has previously been shown to increase affinity for wild-type I-AniI by 5-fold. For panels f and k, mean values, SD, and individual data points are shown for n=3. TSD, target site overlap; LE and RE, left and right transposon ends, respectively; sgRNA, single guide RNA; ddPCR, droplet digital PCR. [Figure 1-2] (As stated above.) [Figure 1-3] (As stated above.) [Figure 1-4] (As stated above.) [Figure 1-5] (As stated above.) [Figure 1-6] (As stated above.) [Figure 1-7] (As stated above.) [Figure 1-8] (As stated above.) [Figure 2-1]Characterization of DNA insertion into genome targets using HELIX. a, Workflow of genome-targeted transposition experiments. b, Integration efficiency using two different amino acid linkers between nAniI and TnsB, sgRNA against genome target site 2 (TS2), and a set of eight donor plasmids with various distances between the I-AniI site and the LE / RE as determined by ddPCR. c, Insertion orientation percentage using ShCAST or ShHELIX targeting TS2 and using pDonor with 14 bp spacing between the I-AniI site and the LE / RE. d, Integration efficiency across six genome target sites for ShCAST and ShHELIX (left panel) and relative integration with ShHELIX normalized to ShCAST (right panel) assessed by ddPCR. e, Coverage of expected insertion products into the genome (TS2) from long-read sequencing using an exemplary simple insertion read for ShHELIX and a subset of cointegrate reads for ShCAST (coverage from ShHELIX cointegrate reads and ShCAST simple insertion reads omitted for simplicity). Transposition products were enriched prior to sequencing by Cas9 target enrichment. f, Read length distribution of transposition products using ShCAST and ShHELIX on genome target site 2 (TS2) from long-read sequencing data. The top right panel is a zoomed-in view of the read length peak at approximately 8,200 bp. g, Comparison of the ratio of simple insertion and cointegrate products at TS2 for ShCAST and ShHELIX as assessed by long-read sequencing. h, Integration efficiency with sgRNAs targeting ShHELIX and TS5 using pDonors encoding cargos of various sizes. Integration was assessed by ddPCR. For panels b, d, and h, mean values, SD, and individual data points are shown for n = 3. LE and RE, left and right transposon ends, respectively; sgRNA, single guide RNA; ddPCR, droplet digital PCR. [Figure 2-2] (As stated above.) [Figure 2-3](As stated above.) [Figure 2-4] (As stated above.) [Figure 2-5] (As stated above.) [Figure 3-1]Extension of HELIX to VK-type CAST orthologues. a, Phylogenetic tree showing the diversity of TnsB sequences from the recently identified VK-type CAST21, the CAST used in this study, as well as Tn5053. b, sgRNA designs for AcCAST. c, Integration efficiency with AcCAST using two sgRNA designs (from panel b) and donor plasmids with either natural flanking sequences (as previously reported14) or ShCAST flanking sequences, assessed by ddPCR. d, Schematic of AcHELIX with 14 bp ShCAST flanking sequences on pDonor. e, Coverage of insertion products into the genome (TS2) from long-read sequencing showing a selection of exemplary simple insertion reads for AcHELIX and cointegrate reads for AcCAST (coverage from AcHELIX cointegrate reads and AcCAST simple insertion reads omitted for simplicity). Transposition products were enriched prior to sequencing by Cas9 target enrichment. f, Read length distribution of transposition products using AcCAST and AcHELIX on TS2 from long-read sequencing data. The top right panel is a zoomed-in view of the approximately 8.3 kb read-length peak. g, Comparison of the ratio of simple insertion and cointegrate products for AcCAST and AcHELIX assessed by long-read sequencing. h, i, Integration efficiency of T-LR and T-RL orientations across six genomic target sites for AcCAST and AcHELIX assessed by ddPCR (panels h and i, respectively). In panel h, the integration efficiency of AcHELIX T-LR relative to AcCAST is shown in the right panel. All transformations contain a pDonor variant with ShCAST flanks and nAniI sites and 14 bp spacing between the LE / RE. j, Integration efficiency using AcHELIX with sgRNA targeting TS6 and pDonors encoding cargos of various sizes assessed by ddPCR. k, Schematic diagram of ShoHELIX with 14-bp ShCAST flanking sequences on pDonor.l, Coverage of expected insertion products into the genome (TS2) from long-read sequencing showing a selection of exemplary simple insertion reads for ShoHELIX and cointegrate reads for ShoCAST (coverage from ShoHELIX cointegrate reads and ShoCAST simple insertion reads omitted for simplicity). Transposition products were enriched prior to sequencing by Cas9 target enrichment. m, Read length distribution using ShoCAST and ShoHELIX on genomic targets (TS2) from long-read sequencing data. n, Comparison of the ratio of simple insertion and cointegrate products for ShoCAST and ShoHELIX assessed by long-read sequencing. o, p, Integration efficiency of T-LR and T-RL orientations across six genomic target sites for ShoCAST and ShoHELIX assessed by ddPCR (panels o and p, respectively). q, Integration efficiency using ShoHELIX with TS3-targeting sgRNA and pDonors encoding cargos of various sizes assessed by ddPCR. All ShoCAST and ShoHELIX transformations contain pDonor variants with ShCAST flanks. For panels c, h–j, and o–q, mean values, SD, and individual data points are shown for n=3. LE and RE, left and right transposon ends, respectively; sgRNA, single guide RNA. [Figure 3-2] (As stated above.) [Figure 3-3] (As stated above.) [Diagram 3-4] (As stated above.) [Figure 3-5] (As stated above.) [Diagram 3-6] (As stated above.) [Diagram 3-7] (As stated above.) [Diagram 3-8] (As stated above.) [Diagram 3-9] (As stated above.) [Figure 4-1]Specificity profiling of ShCAST and ShHELIX systems. a, Schematic of two-component and three-component ShCAST systems containing Cas12k fusions, b, Relative integration efficiency using three-component and two-component ShCAST systems using TnsC and / or TniQ fusions to Cas12k. c, Schematic of three-component ShHELIX system containing Cas12k fusions. d, Relative integration efficiency for three-component ShHELIX system. e, Integration efficiency of ShCAST and ShHELIX systems with or without Cas12k-TnsC fusions using target plasmids with pre-inserted transposons. f, On-target specificity of ShCAST and ShHELIX systems in Endura cells (pir') and PIR2 cells (pir+) with genome-targeting TS2 sgRNA measured by an unbiased specificity profiling approach (see Methods). g, Schematic of transformation protocol using pi protein co-expression in Endura (pir') cells. h, On-target specificity of ShCAST and ShHELIX with or without pi protein co-expression with genome-targeting TS2 sgRNA. i-l, Visualization of genome-wide integration events in Endura cells using ShCAST (6.67 M reads; panel i), ShHELIX with Cas12k-TniQ fusion (4.44 M reads; panel j), ShHELIX with Cas12k-TnsC fusion (3.29 M reads; panel k), or ShHELIX with pi protein co-expression (7.31 M reads; panel l) when programmed with TS2 sgRNA. Black triangles below the x-axis indicate on-target sites, and the y-axis represents the percentage of reads mapping to any given genomic site. For panels b, d, and e, mean values, SD, and individual data points are shown for n=3. LE and RE, left and right transposon ends, respectively; PAM, protospacer adjacent motif. [Figure 4-2] (As stated above.) [Figure 4-3] (As stated above.) [Figure 4-4] (As stated above.) [Figure 4-5] (As stated above.) [Figure 4-6] (As stated above.) [Figure 5-1]HELIX-mediated DNA insertion in human cell lysates and human cells. a, Schematic of N7HELIX with 14 bp ShCAST flanking sequences on pDonor. b, Workflow of plasmid targeting transposition experiments in human cell lysates. c, Qualitative assessment of integration by junction PCR across LE and RE using purified pTarget from lysate assay. d, Representative Sanger sequencing reaction of PCR reaction of insertion product (from panel c). e, PAM-to-LE insertion distance profile of N7HELIX with TS1 sgRNA from plasmid targeting experiment in HEK293T lysates (assessed by NGS; see FIG. 12A). f, Comparison of the ratio of simple insertion and cointegrate products for N7CAST and N7HELIX assessed by PCR enrichment of total and cointegrate insertions, followed by long-read sequencing (Example 11). g, Schematic of workflow for plasmid targeting experiments in HEK293T cells using five separate plasmids. All N7CAST or N7HELIX proteins were expressed from a single all-in-one plasmid. Two different sgRNA architectures using different promoters (sgRNA1 scaffold sequence is wild type, whereas sgRNA2 scaffold contains substitutions within the poly-T extension relative to sgRNA1 to allow U6 promoter compatibility) were tested, both targeting TS1. h, Junction PCR and Sanger sequencing across LE using insert products from HEK293T cell-based plasmid targeting assay. i, Quantification of integration efficiency assessed by ddPCR when transfecting various amounts of pTarget from HEK293T cell-based plasmid targeting assay. j, Quantification of integration efficiency assessed by ddPCR when co-expressing HU protein (in addition to S15) from HEK293T cell-based plasmid targeting assay. k, Integration efficiency of N7CAST and N7HELIX when targeting endogenous genomic target sites in HEK293T cells assessed by ddPCR.l, Schematic of areas of potential optimization to increase integration efficiency of the CAST and HELIX systems in human cells. For panels i–k, mean, SD, and individual data points are shown for n=3. LE and RE, left and right transposon ends, respectively; PAM, protospacer adjacent motif; sgRNA, single guide RNA; NT, non-targeting; HH, hammerhead ribozyme; HDV, hepatitis delta virus ribozyme. [Figure 5-2] (As stated above.) [Figure 5-3] (As stated above.) [Figure 5-4] (As stated above.) [Figure 6-1] Characterization of TnsA fusions to ShTnsB. a, Structures of various TnsA enzymes, either experimentally solved (E. coli TnsA; PDB 1F1Z) or computationally predicted by AlphaFold. b, Integration efficiency when targeting genomic site TS2 using either ShCAST (no fusion) or variants containing fusions of TnsA and ShTnsB linked by either short GSG or XTEN linkers. Integration was measured by ddPCR; mean, SD, and individual data points are shown for n=3. c, Characterization of on-target cointegrates, as measured by long-read sequencing following a Cas9-based target enrichment protocol. d, Percentage of total insertions occurring in pEffector plasmids using either no fusions (ShCAST), nAniI fusions (ShHELIX), or TnsA fusions. [Figure 6-2] (As stated above.) [Figure 6-3] (As stated above.) [Figure 7-1]Optimization and characterization of plasmid targeting experiments. a, Schematic of Donor carrying modified flanking sequences with I-AniI sites placed at various distances from the left and right transposon ends (LE / RE, respectively). b, Colony forming units (CFU) from transformation with ShCAST and ShHELIX plasmids targeting TS1 using a series of pDonor plasmids carrying I-AniI sites and various spacing between LE / RE. c, Integration efficiency assessed by ddPCR using ShCAST targeting TS1 and a series of pDonor with different LE / RE flanking sequences (corresponding to ShHELIX pDonor carrying I-AniI sites and different spacing between LE / RE, see panel a). d, Alignment of 10 exemplary reads carrying ShHELIX-mediated cargo integration 62 bp downstream of the PAM on pTarget. For panels b and c, mean values, SD, and individual data points are shown for n=3. LE and RE, left and right transposon ends, respectively. [Figure 7-2] (As stated above.) [Figure 8] Workflow for plasmid enrichment prior to long-read sequencing. Schematic diagram of the protocol to enrich transposition plasmid products and improve the read depth of the intended products by long-read sequencing. sgRNA, single guide RNA; LE and RE, left and right transposon ends, respectively. [Figure 9-1]Characterization of Y2 ShHELIX. a, Colony forming units (CFU) from transformation with Y2 ShHELIX plasmid targeting TS1 using a series of pDonor plasmids carrying I-AniI sites and various spacing between LE / RE. Means, SD, and individual data points are shown for n=3. b, Coverage of expected insertion products into pTarget from long-read sequencing showing exemplary subset simple insertion or cointegrate reads for Y2 ShHELIX. c, Read length distribution using ShCAST and Y2 ShHELIX with sgRNA targeting TS1 on pTarget. d, Comparison of the ratio of simple insertion and cointegrate products by long-read sequencing for various conditions using Y2-ShHELIX targeting TS1. LE and RE, left and right transposon ends, respectively. [Figure 9-2] (As stated above.) [Figure 9-3] (As stated above.) [Figure 10-1] ShHELIX control experiments, a, Comparison of the ratio of simple insertion and cointegrate products by long-read sequencing for a HELIX variant with catalytically attenuated nAniI (dShHELIX) and when using HELIX with pDonor without an I-AniI site. b, Comparison of the ratio of simple insertion and cointegrate products by long-read sequencing for ShCAST and ShHELIX when using pDonor with a flipped I-AniI site that places the nAniI nicking site on the same strand as the TnsB-derived nick. c, Potential alternative mechanism allowing simple insertion products when using pDonor containing a flipped I-AniI site. TSD, target site overlap. [Figure 10-2] (As stated above.) [Figure 11]Integration efficiency based on long-read sequencing. a, Comparison of integration efficiency for each system as measured by ddPCR or by Cas9 enriched long-read sequencing. The grey dashed line represents the diagonal (concordance between the two types of measurements). b, Integration efficiency at TS2 using the CAST and HELIX systems as assessed by long-read sequencing. The stacked bar graphs represent the percentage of Cas9 enriched targeted reads that lack or contain a cargo insertion. Integration (colored portion of each bar) represents the number of reads containing a cargo insertion divided by the total number of targeted reads. [Figure 12-1] Cargo insertion distance from PAM. a, Schematic of the workflow for characterizing PAM to LE insertion distance by next-generation targeted sequencing. PAM to LE insertion distance profiles for various CAST and HELIX constructs shown in panels: b, ShCAST (4 components); c, ShHELIX (4 components); d, AcCAST (4 components); e, AcHELIX (4 components); f, ShoCAST (4 components); g, ShoHELIX (4 components). h, ShCAST with Cas12k-TniQ (3 components); i, ShCAST with Cas12k-TniQ-TniQ (3 components); j, ShCAST with Cas12k-TnsC (3 components); k, ShHELIX with Cas12k-TniQ (3 components); l, ShHELIX with Cas12k-TniQ-TniQ (3 components); m, ShHELIX with Cas12k-TnsC (3 components); sgRNA, single guide RNA; PAM, protospacer adjacent motif; LE and RE, left and right transposon ends, respectively; NGS, next generation sequencing. [Figure 12-2] (As stated above.) [Figure 12-3] (As stated above.) [Figure 12-4] (As stated above.) [Figure 12-5] (As stated above.) [Figure 13-1]Comparison of type I INTEGRATE and type VK CAST and HELIX systems. a, Schematic of conditions and constructs tested, regulating growth time (24 h), donor cargo size (2.1 kb), approximate donor copy number (high copy), bacterial strain (PIR1), general target location (closest compatible PAM near TS2, TS5, and TS6 of genomic target sites), and efficiency measurement method (ddPCR). b, c, Integration efficiency of INTEGRATE, CAST, and HELIX in the intended forward orientation (panel b) or unintended reverse orientation (panel c). For panels b and c, mean values, SD, and individual data points are shown for n=3. [Figure 13-2] (As stated above.) [Figure 14] Integration efficiencies for the smaller CAST and HELIX systems. a, b, Absolute integration efficiencies when the genome was targeted at TS2 for the 2-, 3-, or 4-component ShCAST (panel a), and when TS2 or TS5 was targeted for the 3- and 4-component ShHELIX systems (panel b). For both panels, integration efficiencies were assessed by ddPCR and used to calculate relative integration as shown in Figure 3; mean values, SD, and individual data points are shown for n=3. [Figure 15-1] Genome-wide integration profiles of ShCAST and ShHELIX systems. a-d, Integration site profiles from unbiased genome-wide insertion analysis of various CAST and HELIX constructs. Experiments were performed in Endura cells (panels a and b) or PIR2 cells (panels c and d) using various ShCAST (panels a and c) or ShHELIX (panels b and d) configurations, including various donor architectures, fusions to Cas12k, pi co-expression, or I-AniI variants. [Figure 15-2] (As stated above.) [Figure 15-3] (As stated above.) [Figure 15-4] (As stated above.) [Figure 15-5] (As stated above.) [Figure 16] Effect of pDonor copy number and pi protein type on integration efficiency. Integration efficiency using ShCAST and ShHELIX and sgRNA targeting genomic site TS2 in two different bacterial strains expressing either wild type pi protein (pir) or a mutant copy number mutant (pir116), where PIR1 and PIR2 cells maintain approximately 250 and 15 copies of pDonor, respectively. Integration efficiency assessed via ddPCR; mean, SD, and individual data points are shown for n=3. R6Kg, origin of replication that requires gene pir to replicate. [Figure 17] Comparison of coding sequences and component numbers of CAST and HELIX systems. Approximate sizes of coding sequences and numbers of protein subunits for the prototype type I and VK-type CAST, the HELIX system developed in this study, and the recently described mini-CAST9 from metagenomic mining. nAniI, nicking I-AniI (K227M). [Figure 18-1]Further characterization of N7CAST and N7HELIX. a, Schematic of the genomic architecture of N7CAST as found in Nostoc sp. PCC7107 (identified by Strecker et al. 7, not drawn to scale). b, Profile of PAM to LE insertion distance using N7CAST and IVT sgRNA targeting TS1 on pTarget in lysate experiments assessed by NGS. c, Schematic of the all-in-one N7CAST and N7HELIX expression plasmids and two versions of an sgRNA encoding the canonical N7 scaffold expressed from a U6 promoter (sgRNA1), or a derivative in which the poly-T stretch in the scaffold has been replaced to be more compatible with transcription from a U6 promoter (sgRNA2). d, Junction PCR using N7CAST or N7HELIX with either IVT sgRNA1 or sgRNA2 targeting TS1 on pTarget in HEK293T lysate experiments. e, Junction PCR from HEK293T cell-based plasmid targeting experiments with or without N7 or Escherichia coli (E. coli) (Ec) S15 and pi proteins. [Figure 18-2] (As stated above.) [Figure 18-3] (As stated above.) [Figure 19-1] Exemplary pDonor sequences. The I-AniI site is shown in bold. The LE and RE sequences for ShCAST, AcCAST, ShoCAST, and N7CAST have been abbreviated for brevity in the pDonor sequence, but the sequences are also shown in the table. [Figure 19-2] (As stated above.) DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] CRISPR-associated transposases (CASTs) are an emerging class of genome editing technologies that enable programmable DNA insertion without relying on recombination, sequence-specific recombinases, or DSBs. However, currently discovered and characterized systems have limitations that limit their ease of use, including size (Figure 17), stoichiometry and component complexity, and / or purity of the insertion product. The two major types of CASTs, type I and type VK, have distinct and complementary properties. Characterized type I CASTs exhibit high on-target specificity and generally result in only the intended simple inserted gene product. 17 (However, there are exceptions. 18 On the other hand, the stoichiometric complexity and large code size of the larger number of Cas genes may limit downstream tool development in other organisms, such as eukaryotic cells. Furthermore, the tendency of some type I systems to result in bidirectional insertions leads to undesirable editing impurities. 15 In comparison, VK-type CAST is more compact in terms of code size, contains only four core components, and results in complete or nearly complete unidirectional insertion. 14、16 However, VK-type CAST results in a problematic mixture of simple insertion and cointegrate gene products, the latter consisting of cargo duplications and complete plasmid backbone insertions. 4、6、19 (affecting the "purity" of the desired product) (Figure 1b). Furthermore, compared to the type I system, VK-type CAST exhibits substantially lower integration specificity 14、16、17、20 .

[0020] Another major difference between type I and VK-type CASTs is whether they encode or lack TnsA, respectively (although type I systems may also lack TnsA in rare cases). 21), this difference contributes to their different integration product purity (defined as the ratio between simple insertion and cointegrate products). In both Tn7 transposon and type I CAST, TnsA and TnsB perform 5' and 3' donor nicking, respectively, resulting in simple insertion via cut-and-paste transposition (Fig. 1a). In Tn5053 transposon and VK-type CAST, which lack TnsA, as well as in Tn7 transposon and catalytically inactive TnsA, 17、22 In the modified type I system with , only 3' donor nicking occurs via TnsB. Singly nicked donors result in a significant proportion of cointegrate insertions through replicative transposition instead of cut-and-paste. 23 (Fig. 1b). To overcome the lack of TnsA in the VK-type system, we hypothesized that an orthogonal DNA nickase could be exploited to restore 5' donor nicking. An ideal nickase would be small (so as to add minimal code size for the system), have predictable nicking sites and strand preference, and function in a variety of organisms for downstream tool development and application. Potential nickases to be considered include type I CAST or other transposon-derived orthogonal TnsA enzymes. 17、24 , nicking restriction endonuclease 25 , nicking Cas variants 9、26、27 , phage HNH endonuclease 28 , or nicking homing endonuclease (nHE) 29~32 Examples include:

[0021] For genome editing applications, the ideal DNA insertion technology would generate a programmable, highly specific, unidirectional, recombination-independent, pure, simple insertion product, all with few components and minimal coding sequences. Therefore, we sought to develop an engineered CAST that combines the simplicity and orientation predictability of the VK-type system with the product purity and specificity of the I-type system. Our results reveal that an optimized and engineered HE-assisted large-sequence integration CAST complex (HELIX), consisting of nHE fusion to TnsB with the remaining CAST components, can substantially improve the purity and specificity of CAST-mediated DNA insertion.

[0022] As shown herein, HELIX utilizes the technical advantages of VK-type CAST to achieve programmable and efficient cut-and-paste DNA insertion similar to type I CAST using nHE fusions and modified donor plasmids. HELIX dramatically increases the purity of simple insertion products on plasmids and genomic targets in E. coli and retains robust RNA-guided transposition at or near wild-type levels. Furthermore, a simplified CAST and HELIX system is shown herein that includes a three-component system via subunit fusions to Cas12k, which enhances integration efficiency.

[0023] CAST is a new class of genome editing technology that allows programmable DNA insertion without relying on recombination, sequence-specific recombinases, or DSBs. Here, we overcome some of the major limitations of CAST by utilizing the technical advantages of VK-type CAST to develop HELIX, which achieves programmable, specific, and efficient cut-and-paste DNA insertion. We demonstrate that HELIX enhances the purity of simple insertion products on plasmid and genomic targets in Escherichia coli (E. coli) and retains robust RNA-guided transposition at or near wild-type levels. HELIX is effective across several VK-type CAST orthologs, establishing the universality of this approach. We also demonstrate that HELIX is substantially more specific than the CAST from which it is derived, and that co-expression of Cas12k fusions and / or pi proteins can further reduce genome-wide off-target integration. Finally, we demonstrate that the advantages of HELIX can be translated to a human cell context on plasmid targets. Overall, our approach represents the first delineation of CAST engineering and highlights how other naturally occurring enzymes can be exploited to enhance CAST properties for use in a variety of systems.

[0024] Our results also provide insight into certain mechanistic aspects of HELIX. First, nAniI must be proximal to TnsB via fusion to potentially coordinate the nicking reactions of nAniI and TnsB to reduce cointegrates. Similarly, in Tn7 and type I CAST, physical proximity is mediated by protein-protein interactions between TnsA, TnsB, and TnsC. 33Second, fusions of the TnsA domain from Tn7 or type I CAST to ShTnsB were ineffective in reducing cointegrates, presumably because TnsA is only active in complex with its cognate TnsB and TnsC, physically and temporally coordinating strand-specific cleavage. These results suggest that generating 5' nicks in VK-type systems via fusion proteins to TnsB is optimal from a standalone nicking endonuclease (such as nHE in HELIX), a conclusion supported by our efficiency and target immunity data sets that reveal that nAniI-TnsB fusions do not substantially interfere with other CAST components (i.e., donor or target DNA, or TnsC).

[0025] The continued discovery and optimization of CAST will result in more robust integration techniques. We anticipate the identification of new systems with useful properties (e.g., via metagenomic mining for smaller VK-type systems). 21 ) contribute to the diversity of the enzyme that can be further engineered by HELIX or other methods to enhance various integration parameters. During our characterization, we discovered optimization of various regions to modulate CAST properties. For example, modification of the flanking sequences immediately adjacent to the LE / RE on pDonor likely has sequence-specific effects (as demonstrated for Mu transposase). 52 ) and / or changes in interactions with unknown host factors may affect integration. Furthermore, fusion proteins to various CAST components resulted in unexpected changes in properties. Our findings suggest that a better understanding of several parameters (enhancing donor flanking sequences, amino acid linkers, spacing between nHE sites and LE / RE, nHE selection, etc.) combined with efforts to generate hyperactive variants of VK-type CAST (potentially through TnsB and Cas12k directed evolution and structure-guided engineering) will result in more powerful next-generation CAST and HELIX systems.

[0026] While HELIX resolves many of the limitations of VK CAST, our work also leaves open questions that merit continued investigation: incomplete ablation of cointegrate products could result from non-congruent donor nicking by nAniI and TnsB, which may also be the case for the minimal cointegrate products observed in type I systems, potentially due to asynchronous TnsA and TnsB donor nicking. 17 Further studies are worthwhile to explore the mechanisms of various HELIX improvements, including how pi proteins or fusions (such as nAniI-TnsB, Cas12k-TnsC, Cas12k-TniQ) contribute to specificity regulation. We suggest that the change in CAST conformation via the nAniI-TnsB fusion and the change in donor topology via modified TnsB-donor interactions, as well as the alteration of the pi junction of the iteron and / or AT-rich sequences at the left and right transposon ends and / or parts of the donor backbone, may contribute to specificity regulation. 53 Furthermore, how the component fusions and / or pi proteins function in concert with HELIX to enhance specificity, but not generally with CAST, requires further investigation.

[0027] We have demonstrated that CAST and HELIX can function in human lysates and cells on plasmid targets, but integration efficiencies were low using the constructs and conditions described. Methods that can improve efficiency are therefore important for the conversion of these systems in a variety of contexts. The recent discovery that ribosomal protein S15 is a bacterial host factor required for efficient transposition. 43 Therefore, it is plausible that additional bacterial host protein(s) may be required for efficient human cell integration. Our results support the requirement for S15. Indeed, the nucleoid-associated proteins (NAPs) HU and IHF are required for efficient Mu transposition. 51The same and / or other NAPs and DNA bending proteins are also involved in other transposon families (e.g., Tn10, IS903, Tn552, Sleeping Beauty, etc.). 54~56 The pi proteins that we observed to enhance insertion specificity are also known to distort DNA. 53 , which can act as a competitive binder with IHF 57 Thus, protein-induced changes in donor topology may affect the properties of transposition, possibly in addition to specificity, paired complex formation and / or transposase activity. Furthermore, host-encoded acyl carrier protein (ACP) and ribosomal protein L29 are involved in TnsD-mediated Tn7 transposition. 58 and TnsE-mediated pathway 59 It has been shown that DnaN is involved in the transcription of endothelial cells. Along with the discovery of host factors, the donor, sgRNA, and the protein itself (e.g., a more active nHE 35 Engineering and optimization of the HELIX components through modifications to Cas12k variants with improved binding affinity (e.g., TnsB variants, Cas12k variants with improved binding affinity), should enable more efficient and specific targeting of the human genome, as has been done with other Cas orthologs, including some that initially showed minimal activity. 60~62 Component fusions may also prove useful in facilitating the localization of these multicomponent systems.

[0028] Besides CAST, other advances have been made in DSB-free large-scale sequence integration techniques. Recent studies have combined prime editing (PE) with a site-specific serine recombinase to integrate DNA into the human genome in an RNA-programmed manner. 63、64With successful discovery and engineering efforts to enable more efficient use in human cells, HELIX represents a complementary technology with advantages compared to PE-based methods: smaller code size, the need to design only a single sgRNA instead of multiple PEGRNAs, complete elimination of DSBs, more minimal dependency on host cell repair, and the vast versatility of CAST that may be naturally amenable to efficient eukaryotic function and therapeutic deliverability.

[0029] Transposon-nickase fusion protein Described herein is a fusion protein comprising a transposition protein B (TnsB) protein (e.g., a Tn7, Tn7-like, or Tn5053-like transposition protein B (TnsB) protein) fused to a protein (e.g., a nickase), optionally via an intervening linker. In some embodiments, a DNA cleavase fusion can be used in place of a nickase fusion for cut-and-paste DNA insertion. The methods and compositions can be applied in a number of transposon / CAST systems, for example, in:

[0030] Standard Tn7 transposon 42、43、44 Tn7 has four components, TnsABCD, which form a heterotrimeric complex (TnsA and TnsB create 5' and 3' nicks at the transposon ends, and TnsC is an ATPase that regulates transposition activity). Tn7 is mediated by (1) TnsD, a sequence-specific DNA-binding protein that recognizes the Tn7 binding site; 45、46 (2) mediated by TnsE, which facilitates transfer and DNA replication into the conjugative plasmid 47 , is targeted to DNA by two alternative pathways.

[0031] CRISPR-Cas systems associated with Tn7-like transposons (type I CAST): Type I CRISPR Cas system is associated with Tn7-like transposons containing TnsA, TnsB, TnsC and TniQ genes and CRISPR system. TnsD / TnsE in standard Tn7 transposon is replaced by these CRISPR-Cas systems. "Tn7-like" refers to the association with standard system (i.e., with Tn7 family of transposons) and includes components TnsABC. Such systems may include VchCAST (from Vibrio cholerae Tn6677), AsaCAST (from Aeromonas salmonicida S44), AvCAST (from Anabaena variabilis ATCC29413), PmcCAST (Peltigera membranacea cyanobiont 210A) and PtrCAST in BL21(DE3). 57 .

[0032] CRISPR-Cas systems associated with the Tn5053 family of transposons (VK-type CAST): VK-type CAST is most closely related to the Tn5053 family of transposons. 48、21 Such systems include shCAST (from Scytonema hofmannii), AcCAST (from Anabaena cylindrica), and ShoCAST (from Scytonema hofmannii PCC 7110), in which the Tn5053 transposon, although not fully characterized, is known to lack TnsA, which results in cointegrates that are resolved by the transposon-encoded recombinase TniR. 49For VK-type CAST, the transposon does not encode an identifiable resolvase / recombinase to do so. In some embodiments, the VK-type CAST is a CAST as described in Rybarski JR, Hu K, Hill AM, Wilke CO, Finkelstein IJ. Metagenomic discovery of CRISPR-associated transposons. Proc Natl Acad Sci US A. 2021 Dec 7;118(49):e2112279118. doi: 10.1073 / pnas.2112279118, or in Table 2 of US Pat. No. 1,384,344.

[0033] Nickase / Chrysanthemum vase The nickase can be fused to either the N- or C-terminus of the transposon. Preferably, the nickase is smaller than about 500 amino acids. Many suitable nickases are known in the art and can be used, with exemplary nickases being nicking restriction endonucleases such as 22 , nicking Cas variants 9、23、24 , or phage HNH endonuclease 25 , or the catalytic part of the TnsA enzyme from type I CAST or the Tn7 transposon 26 or catalytic portions thereof. In some embodiments, the nickase is a homing endonuclease (HE), such as a LAGLIDADG HE (LHE), for example, an LHE from Aspergillus nidulans (I-AniI) optionally containing a K227M mutation (nAniI) or a hyperactive variant thereof (e.g., Y2I-AniI) can be used. Further examples of homing endonucleases (classified based on sequence motifs / domains) include LAGLIDADG, such as I-SceI (which has been engineered to be a sequence-specific nickase), and I-SceI (which has been engineered to be a sequence-specific nickase). 49 ) and I-DmoI (also engineered to be a sequence-specific nickase 50); HNH, e.g., I-PfoP3I (which occurs naturally as a nickase) 51 and I-BasI (also occurs naturally as a nickase); GIY-YIG, e.g., I-BmoI 5 and I-TevI14; or His-Cys Box, e.g. I-PpoI 52 For a comprehensive review, see Stoddard et al., 2011 16 As noted above, in some embodiments, fusions of cleaved versions of these enzymes to transposon proteins, such as TnsB, are used, which can improve the purity of the integration products and reduce cointegrates.

[0034] Linker In some embodiments, the fusion protein comprises a linker between the transposon protein and the nickase. For example, linkers known in the art comprising 1-100 amino acids, such as flexible linkers (e.g., XTEN linkers (containing GEDSTAP amino acids)) or Gly-Ser or Gly-Ser-Ala rich linkers (e.g., GSAGSAAGSGEF, GGSGGGSGG, (GGGGS)3 or (Gly) n ), PAS repeats, GQAP-like repeats, or SOBI linkers, or rigid linkers, such as alpha helical linkers (e.g., (EAAAK)3) or (XP) n(X refers to any amino acid, preferably Ala, Lys, or Glu) can be used.See, for example, Chen et al., Advanced Drug Delivery Reviews, 15 October 2013, 65(10): 1357-1369; An Overview of Linkers for Recombinant Fusion Proteins, kbdna.com / publishinglab / lnkr (05 / 08 / 2021); Podust et al., Protein Engineering, Design & Selection (2013), 26 (11), 743-753; Kjeldsen et al., ACS Omega 2020, 5, 31, 19827-1983.

[0035] Flanking sequences As provided herein, the constructs include flanking sequences that can affect integration, e.g., nucleotides immediately adjacent to the LE and RE of the donor sequence to be inserted on a donor plasmid (one example of which is referred to herein as pDonor). The flanking sequences can be, e.g., about 10-100, 10-20, 10-50, 10-30, 12-100, 12-50, 12-30, or 25-50 nucleotides in length, and can be altered to affect integration efficiency (Figures 4c and 6b). As used herein, modified flanking sequences have at least one change with respect to the corresponding flanking sequences from the organism from which the transposon sequence was obtained. The flanking sequences can be altered to increase transposition efficiency. Exemplary flanking sequences and their source organisms are provided in Table A. Flanking sequences can also be modified to include endonuclease recognition sites, e.g., I-AniI sites, at the 5' and / or 3' ends, e.g., 4-50, 4-25, 10-20, 12-20, 4-15, 10-15, 12-15, 10-16, or 10-18 nt away from the end of the sequence to be inserted. See further exemplary sequences below and in FIG. 15.

[0036] [Table 1]

[0037] HE-assisted large-scale array integration CAST complex (HELIX) Described herein are compositions and systems that can be used for programmable insertion of up to multi-kilobase DNA sequences into DNA, e.g., into the genome of a cell. The HELIX system component(s) include a fusion protein as described herein, e.g., including a transposon, e.g., TnsB, fused to a protein (e.g., nickase) optionally via an intervening linker. In some embodiments, DNA cleavage fusions can be used in place of nickase fusions for cut-and-paste DNA insertion.

[0038] Other HELIX system component(s) include cas12k, TnsC, and TniQ. A functional system includes a TnsB-nickase fusion protein, cas12k, TnsC, TniQ, and a guide RNA (e.g., single guide RNA (sgRNA)) that binds to cas12k and guides the HELIX system to the intended insertion site, and a donor nucleic acid, e.g., a donor plasmid, preferably containing the sequence to be inserted, flanked by LE and RE sequences at the 5' and 3' ends, respectively, and a target site (e.g., I-AniI) for the nickase, preferably oriented to give a 5' nick on the donor plasmid. The Cas12k enzyme itself is catalytically inactive, and it is directed to bind the gRNA and bind the target site (but does not cut or nick). The bound Cas12k reinforces the downstream transposition machinery (such as TniQ, TnsC, and TnsB / nAniI-TnsB).

[0039] Co-expression of certain bacterial proteins (i.e., host factors) with standard CAST components can alter activity in bacteria or rescue and improve activity in eukaryotic cells. Thus, in some embodiments, host factors known to alter DNA topology to enhance insertion efficiency or specificity in prokaryotic or eukaryotic cells are also included. For example, ribosomal protein S15 is required for VK-type CAST integration, ribosomal protein L29 (and host acyl carrier protein ACP) is required for efficient TnsD-mediated Tn7 transposition, and DnaN is required for efficient TnsE-mediated Tn7 transposition. DnaA, DNA topoisomerase I, La protease, and Dam methylase alter Tn5 transposition (Schmitz, M., Querques, I., Oberli, S., Chanez, C., & Jinek, M. (2022). Structural basis for RNA-mediated assembly of type V CRISPR-associated transposons. Biorxiv, Chandler, M., and Mahillon, J. (2002) Insertion sequences revisited. In Mobile DNA II, Vol. II. Craig, N. L., Craigie, R., Gellert, M., and Lambowitz, A. M. (eds). Washington, DC: American Society for Microbiology Press, pp. 305 - 366, Craig, N. L., Craigie, R., Gellert, M., and Lambowitz, A. M. (eds). , AM (2002) Mobile DNA II. Washington, DC: American Society for Microbiology, Nagy , Z. , and Chandler , M. (2004) Regulation of transposition in bacteria .Res Microbiol 155: 387 - 398; Sharpe, PL & Craig, NL Host proteins can stimulate Tn7 transposition: a novel role for the ribosomal protein L29 and the acyl carrier protein. EMBO J. 17, 5822-5831 (1998); Parks, AR et al. Transposition into replicating DNA occurs through interaction with the processivity factor. Cell 138, 685-695 (2009)). In addition, the nucleoid-associated proteins (NAPs) HU and IHF are required for efficient Mu transposition, and the same and / or other NAPs and DNA bending proteins are transposition requirements or enhancements for other transposition families (e.g., Tn10, IS903, Tn552, Sleeping Beauty, etc.). Other examples of NAPS are H-NS, Fis, and TF1. The pi protein also alters DNA topology. .

[0040] In other embodiments, host factors are involved in DNA or cellular metabolism, protein degradation or protein folding, regulation, transport, and unknown functions in prokaryotic or eukaryotic cells. Exemplary proteins are acyl carrier protein (ACP), Sigma S, or proteins expressed from genes dcd, dinD, radA, recQ, clpX, fkpA, hflX, crl, rseB, rsxE, araJ, melB, mgtA, aspA, treC, proY, serA, yhbC, yidA, ykfA.

[0041] Delivery and Expression Systems To use the HELIX system described herein, it may be desirable to express one or more of the components from the nucleic acid that encodes them. This can be done in various ways. For example, the nucleic acid encoding the HELIX system component(s) can be cloned into an intermediate vector for transformation into a prokaryotic or eukaryotic cell for replication and / or expression. The intermediate vector is usually a prokaryotic vector, such as a plasmid, or a shuttle vector, or an insect vector, for storage or manipulation of the nucleic acid encoding the HELIX system component(s) for production of the HELIX system component(s). The nucleic acid encoding the HELIX system component(s) can also be cloned into an expression vector for administration to a plant cell, an animal cell, preferably a mammalian cell or a human cell, a fungal cell, a bacterial cell, or a protozoan cell.

[0042] In some embodiments, a single expression vector is used that contains sequences encoding a TnsB-nickase fusion protein, cas12k, TnsC, TniQ, and a single guide RNA that binds cas12k. CAST and its component parts have been described in the art, see, for example, Streckeret et al., Science. 2019 Jul 5;365(6448): 48-53, Rybarski et al., PNAS December 7, 2021 118 (49) e2112279118, and US Patent Publication No. 20200190487.

[0043] To obtain expression, the sequence encoding the HELIX system component(s) is usually subcloned into an expression construct, such as a vector containing a promoter to direct transcription. Suitable bacterial and eukaryotic promoters are well known in the art and are described, for example, in Sambrook et al., Molecular Cloning, A Laboratory Manual (3d ed. 2001), Kriegler, Gene Transfer and Expression: A Laboratory Manual (1990), and Current Protocols in Molecular Biology (Ausubel et al., eds., 2010). Bacterial expression systems for expressing proteins are available, for example, in E. coli, Bacillus sp., and Salmonella (Palva et al., 1983, Gene 22: 229-235). Kits for such expression systems are commercially available. Eukaryotic expression systems for mammalian cells, yeast, and insect cells are well known in the art and are commercially available as well.

[0044] The promoter used to direct the expression of nucleic acid depends on the specific application. For example, strong constitutive promoters are usually used for the expression and purification of fusion proteins. In some embodiments, for example, when the HELIX system component(s) should be expressed in vivo, either constitutive or inducible promoters can be used depending on the specific use of the HELIX system component(s). Furthermore, the preferred promoter for administration of the HELIX system component(s) can be a weak promoter such as HSV TK or a promoter with similar activity. Promoters can also contain elements that respond to transactivation, e.g., hypoxia response elements, Gal4 response elements, lac repressor response elements, and small molecule control systems such as the tetracycline regulatory system and the RU-486 system (see, e.g., Gossen & Bujard, 1992, Proc. Natl. Acad. Sci. USA, 89: 5547; Oligino et al., 1998, Gene Ther., 5: 491-496; Wang et al., 1997, Gene Ther., 4: 432-441; Neering et al., 1996, Blood, 88: 1147-55; and Rendahl et al., 1998, Nat. Biotechnol., 16:757-761).

[0045] Besides the promoter, expression vectors usually contain other regulatory elements, such as transcription units or expression cassettes, which contain all the additional elements required for the expression of nucleic acid in either prokaryotic or eukaryotic host cells.Thus, a typical expression cassette contains, for example, a promoter operably linked to the nucleic acid sequence encoding the HELIX system component(s), and, for example, any signal required for efficient polyadenylation of the transcript, transcription termination, ribosome binding site, or translation termination.Additional elements of the cassette may include, for example, enhancers, and heterologous splicing intron signals.

[0046] The particular expression vector used to transport genetic information into cells is selected based on the intended use of the HELIX system component(s), for example, expression in plants, animals, bacteria, fungi, protozoa, etc. Standard bacterial expression vectors include plasmids such as pBR322-based plasmids, pSKF, pET23D, and commercially available tag fusion expression systems such as GST and LacZ. Naked DNA and viral vectors (e.g., AAV), preferably non-integrating vectors, can also be used.

[0047] Expression vectors containing regulatory elements from eukaryotic viruses are often used in eukaryotic expression vectors, such as SV40 vectors, papilloma virus vectors, and vectors derived from Epstein-Barr virus. Other exemplary eukaryotic vectors include pMSG, pAV009 / A+, pMTO10 / A+, pMAMneo-5, baculovirus pDSVE, and any other vector that allows expression of a protein under the induction of the SV40 early promoter, SV40 late promoter, metallothionein promoter, mouse mammary tumor virus promoter, Rous sarcoma virus promoter, polyhedrin promoter, or other promoters shown to be effective for expression in eukaryotic cells.

[0048] Some expression systems have markers for selection of stably transfected cell lines, such as thymidine kinase, hygromycin B phosphotransferase, and dihydrofolate reductase. Also suitable are high-yield expression systems, e.g., using baculovirus vectors in insect cells, e.g., with the gRNA coding sequence under the guidance of the polyhedrin promoter or other strong baculovirus promoters.

[0049] Other elements typically included in expression vectors include a replicon that functions in E. coli, a gene encoding antibiotic resistance to allow for selection of bacteria harboring the recombinant plasmid, and a unique restriction site in a non-essential region of the plasmid to allow for insertion of recombinant sequences.

[0050] Standard transfection methods are used to produce bacterial, mammalian, yeast or insect cell lines that express large amounts of protein, which are then purified using standard techniques (see, e.g., Colley et al., 1989, J. Biol. Chem., 264: 17619-22; Guide to Protein Purification, in Methods in Enzymology, vol. 182 (Deutscher, ed., 1990)). Transformation of eukaryotic and prokaryotic cells is carried out according to standard techniques (see, e.g., Morrison, 1977, J. Bacteriol. 132:349-351; Clark-Curtiss & Curtiss, Methods in Enzymology 101:347-362 (Wu et al., eds, 1983)).

[0051] Any known procedure can be used to introduce foreign nucleotide sequences into host cells. These include calcium phosphate transfection, polybrene, protoplast fusion, electroporation, nucleofection, liposome, microinjection, naked DNA, plasmid vector, viral vector (e.g., AAV) (both episomal and integrative), and any other known method for introducing cloned genomic DNA, cDNA, synthetic DNA or other foreign genetic material into host cells (see, for example, Sambrook et al., supra). The specific genetic engineering procedure used only needs to be able to successfully introduce at least one gene into host cells that can express HELIX system component(s).

[0052] Alternatively, the method may include delivering the HELIX system component(s) protein and guide RNA together, e.g., as a complex. For example, the HELIX system component(s) and gRNA may be overexpressed in a host cell, purified, and then complexed with a guide RNA (e.g., in a test tube) to form a ribonucleoprotein (RNP) and delivered to the cell. In some embodiments, the variant Cas9 may be expressed in bacteria and purified from bacteria through the use of a bacterial Cas9 expression plasmid. For example, a His-tagged variant Cas9 protein may be expressed in a bacterial cell and then purified using nickel affinity chromatography. The use of RNPs avoids the need to deliver plasmid DNA that encodes the nuclease or guide, or encodes the nuclease as mRNA. Delivery of RNPs may also improve specificity, likely due to the shorter half-life of RNPs and the absence of sustained expression of the nuclease and guide (as obtained from a plasmid). RNPs can be delivered to cells in vivo or in vitro, for example, using lipid-mediated transfection or electroporation.For example, Liang et al. "Rapid and highly mammalian cell engineering via Cas9 protein transfection." Journal of biotechnology 208 (2015): 44-53, Zuris, John A., et al. "Cationic lipid-mediated efficient delivery of proteins enables efficient protein-based genome editing in vitro and in vivo." Nature biotechnology 33.1 (2015): 73-80, Kim et al. "Highly RNA-guided genome editing in human cells via delivery of purified Cas9 ribonucleoproteins." Genome research 24.6 (2014): 1012-1019.

[0053] Thus, provided herein are HELIX system component(s) (proteins and nucleic acids), vectors, and cells comprising the vectors.

[0054] How to use the HELIX system Provided herein is a method for inserting a desired sequence into DNA, e.g., genomic DNA of a living cell, e.g., a mammalian cell, e.g., a cell derived from a human or non-human animal. The method includes expressing in the cell a donor DNA molecule (e.g., a plasmid or linear dsDNA) that includes a nucleic acid sequence encoding a TnsB-nickase fusion protein as described herein; a nucleic acid sequence encoding a TnsB-nickase fusion protein, cas12k, TnsC, TniQ, and a guide RNA that binds to cas12k; and the desired sequence to be inserted, where the desired sequence is flanked by LE and RE sequences at the 5' and 3' ends, respectively, and the target site for the nickase (e.g., I-AniI) is preferably oriented to impart a 5' nick on the donor plasmid. EXAMPLES

[0055] The invention is further described in the following examples, which do not limit the scope of the invention described in the claims.

[0056] method In the examples below, the following materials and methods were used.

[0057] Plasmids and oligonucleotides All plasmids and selected sequences used in this study are listed in Table 1. New plasmids were generated by isothermal assembly or Golden Gate assembly, some of which have been deposited at Addgene (Table 1). pHelper and pDonor plasmids for ShCAST and AcCAST, and pTarget were donated by Feng Zhang (Addgene plasmid numbers 127921, 127924, 127923, 127925, 127926). For plasmids encoding gRNAs, spacer sequences were cloned into pCAST and pHELIX plasmids by Golden Gate assembly using SapI (New England Biolabs, NEB). Target site characteristics for all gRNAs used in this study are found in Appendix 2. Oligonucleotides and probes used in this study were purchased from Integrated DNA Technologies (IDT) and are listed in Appendix 3. Gene fragments for construct cloning were ordered from Twist Biosciences, and synthetic SpCas9 sgRNA was ordered from Synthego (Appendix 2).

[0058] [Table 2-1]

[0059] [Table 2-2]

[0060] [Table 2-3]

[0061] [Table 2-4]

[0062] [Table 2-5]

[0063]

Table 2-6

[0064]

Table 2-7

[0065]

Table 2-8

[0066]

Table 2-9

[0067]

Table 2-10

[0068]

Table 2-11

[0069]

Table 2-12

[0070]

Table 2-13

[0071]

Table 2-14

[0072]

Table 2-15

[0073] [Table 3]

[0074] [Table 4-1]

[0075] [Table 4-2]

[0076] [Table 4-3]

[0077] [Table 4-4]

[0078] [Table 4-5]

[0079] Plasmid and genomic site-targeted transposition assays Transformations for plasmid targeting experiments were performed in chemically competent PIR1 cells (original PIR1 strain obtained from Invitrogen) containing pTarget using 25ng of pCAST or pHELIX and 25ng of pDonor. For target immunity experiments, 25ng of pTarget (containing a different cargo than pDonor) encoding a pre-inserted mini-transposon was co-transformed with pCAST or pHELIX and pDonor in PIR1 cells not carrying any plasmid. Transformed cells were allowed to recover in SOC for 1 hour at 37°C and then plated on LB agar plates containing 50 μg / mL kanamycin, 25 μg / mL chloramphenicol, and 100 μg / mL carbenicillin. Plates were incubated at 37°C for 18 hours. Colonies were counted, scraped, and plasmid DNA was extracted by miniprep (Qiagen). The resulting plasmid pools were used for downstream analysis by junction PCR and long-read sequencing. Junction PCR was analyzed by QIAxcel Capillary Electrophoresis (Qiagen) and visualized with QIAxcel ScreenGel Software (vl. 5.0.16; Qiagen).

[0080] Transformations for genome targeting experiments were performed using PIR1 cells (or PIR2 cells (Invitrogen) for FIG. 12) and 25 ng of pCAST or pHELIX and 25 ng of pDonor. Transformed cells were allowed to recover in SOC for 1 h at 37° C. and then plated on LB agar plates containing 50 μg / mL kanamycin and 100 μg / mL carbenicillin. For transformations containing ShCAST, ShHELIX, ShoCAST, or ShoHELIX plasmids, plates were incubated at 37° C. for 18 h, and for AcCAST and AcHELIX transformations, plates were incubated at 37° C. for 24 h due to the relatively smaller colonies (although the numbers were similar). Colonies were scraped and gDNA was collected using the Wizard Genomic DNA Purification Kit (Promega) for downstream analysis by ddPCR and long-read sequencing.

[0081] Assessment of integration efficiency via ddPCR Plasmid or genomic DNA from the E. coli transposition assay was normalized to 10ng / μL or 100ng / μL, respectively, and subsequently further diluted to a working stock of 0.2ng / μL or 2ng / μL. DNA (genomic / plasmid mixture) extracted from the plasmid targeting HEK293T transposition assay was used undiluted for insertion detection and diluted 100-fold to count total pTarget plasmid. Insertion events were measured using target-specific primers and donor-specific probes (Appendix 3). Specifically for the target immunity experiment, the reverse primer to detect insertion was bound just inside the LE on the cargo (different for pre-placed insertions and cargo to be inserted) rather than directly on the LE. ddPCR reactions contained 20 pg of plasmid DNA (E. coli, from a plasmid targeting assay), 2 ng of E. coli gDNA, or 4 μL of gDNA / plasmid mix (from a HEK293T plasmid targeting assay), 250 nM of each primer, 900 nM of probe, and ddPCR supermix for probe (no dUTP) (BioRad) in a 20 μL reaction, and droplets were generated using a QX200 Automated Droplet Generator (BioRad). Thermal cycling conditions were 1 cycle of (95°C for 10 min), 40 cycles of (94°C for 30 s, 58°C for 1 min), and 1 cycle of (98°C for 10 min) with a 4°C hold. PCR products were analyzed using a QX200 Droplet Reader (BioRad) and absolute quantification of inserts was determined using QuantaSoft (vl.7.4). Additionally, total template DNA was analyzed and integration efficiency was calculated by insert / template×100.

[0082] Long-read sequencing of plasmids and genomic integration The purity of the integration products was analyzed by long-read sequencing using plasmids obtained from plasmid-targeted transposition reactions in E. coli (where HELIX pDonor was used for all conditions). Transposition products were enriched by electroporating approximately 100 ng of the plasmid pool into Endura Electrocompetent cells (Lucigen), a non-PIR strain that limits recombination. Cells were allowed to recover in SOC for 1 h at 37°C and spread on LB agar plates containing 50 μg / mL kanamycin and 25 μg / mL chloramphenicol. Plates were incubated at 30°C (to limit recombination) for 24 h, scraped, and plasmid DNA was extracted by miniprep. The enriched plasmids were digested with EcoRV (NEB) for 8 h at 37°C. Amplification-free long-read sequencing library preparation (Oxford Nanopore Technologies, SQK-LSK109) was performed using a barcode expansion kit (Oxford Nanopore Technologies, NBD-104). The final pooled library was loaded onto an R9.4.1 flow cell and sequenced for 24 h.

[0083] To perform long-read sequencing of E. coli genomic target insertions, we implemented an amplification-free Cas9 target enrichment protocol to improve selective sequencing of the intended on-target sites (Oxford Nanopore Technologies, SQK-CS9109; sgRNAs listed in Appendix 2). As described in the SQK-CS9109 protocol, normalized aliquots of genomic DNA from genome-targeted transposition assays (where HELIX pDonor was used for all conditions) were dephosphorylated, and Cas9 and gRNA RNPs were targeted to cleave approximately + / - 1.5 kb of the target site on the dephosphorylated gDNA according to the SQK-CS9109 protocol. Adapters were selectively ligated to these segments, thereby enriching the target region and increasing the sensitivity of our sequencing of genomic targets. The resulting libraries were loaded onto an R9.4.1 flow cell and sequenced for 30 hours.

[0084] To analyze the purity of the integration products from the N7CAST and N7HELIX human lysate experiments (described below), a PCR-based enrichment strategy was used that minimizes size and template bias due to low efficiency of transposition (Example 11). Two sets of primers were used that amplify from upstream of TS1 to the RE of the insertion product (regardless of simple insertion or cointegrate) or from upstream of TS1 to the backbone of the cointegrate. These two reactions were performed in separate PCR reactions using Q5 high fidelity DNA polymerase (NEB) and containing the same volume of terminated lysate reaction as template (2 μL). The thermal cycling conditions for both PCRs were 98° C. for 2 minutes, followed by 20 cycles of (98° C. for 10 seconds, 64° C. for 15 seconds, 72° C. for 90 seconds), and a final extension of 72° C. for 3 minutes. The two reactions were combined and purified using 1× AmpureXP beads. Amplification-free long-read sequencing library preparation (Oxford Nanopore Technologies, SQK-LSK109) was performed using a barcode extension kit (Oxford Nanopore Technologies, NBD-104) and the final pooled library was sequenced for 20 h on an R9.4.1 flow cell.

[0085] Data processing of long-read sequencing results Fast5 files were base-called in real time using Miknow (v21.06.9) with the accelerated base-calling model, and the resulting FastQ files were filtered for Q-scores above 8. 65BBDuk from was used to filter reads containing 20 bp LE and RE and 30 bp target site sequences with a maximum Hamming distance of 2. Of these reads, those containing the 20 bp sequence (with a maximum Hamming distance of 2) found in the plasmid backbone (not expected to occur in simple insertion products) were classified as potential cointegrates, and those not containing this sequence were classified as potential simple insertions. Reads for plasmid targeting experiments were further filtered for appropriate read length. Reads containing products assigned as simple insertions or cointegrates were merged into a single FastQ file and analyzed using Minimap2 with the map-ont parameters specified. 66 The reads were aligned to either synthetic simple insertions or cointegrate products with the . Coverage plots were generated from an exemplary set of 100 reads using Geneious (v2021.2.2) and its built-in aligner (medium sensitivity and up to 5 iterations). Sam files containing the aligned reads were also generated and used to create length histograms.

[0086] For sequencing results from human lysate experiments, FastQ files were also filtered for Q-scores above 8, LEs and REs of 20 bp, and target site sequences of 30 bp with a maximum Hamming distance of 2. Reads containing 20 bp sequences found in the plasmid backbone were classified as cointegrates, whereas those that did not were classified as "total". Filtered reads were aligned to a synthetic reference using Geneious (v2021.2.2) and its built-in aligner (medium sensitivity and up to 5 iterations) and manually inspected. The percentage of cointegrates was calculated as the number of reads classified as cointegrates divided by the number of reads classified as "total".

[0087] Analysis of insertion distance using targeted sequencing The insertion distance from the PAM to the LE was assessed by next-generation sequencing using a two-step PCR-based library construction method. 50 ng of genomic DNA from genome targeting experiments was PCR amplified using Q5 high-fidelity DNA polymerase (NEB) and primers binding just outside TS2 or just inside the LE (Appendix 3). Thermal cycling conditions were 98°C for 2 min, followed by 25 cycles of (98°C for 10 s, 64°C for 15 s, 72°C for 20 s), and a final extension at 72°C for 3 min. PCR products were analyzed by QIAxcel capillary electrophoresis (Qiagen) and analyzed as previously described. 67、68 The PCR products were purified using prepared paramagnetic beads. 20 ng of the purified PCR products were used as template for a second PCR to add Illumina barcodes and adapter sequences (Appendix 3). Thermal cycling conditions were 98°C for 2 min, followed by 10 cycles of (98°C for 10 s, 65°C for 30 s, 72°C for 30 s) and a final extension of 72°C for 5 min. PCR products were analyzed and purified before quantification by QuantiFluor (Promega) and combined into equimolar pools. The final libraries were quantified by qPCR (KAPA Library Quantification Kit; Roche 7960140001) and sequenced on a MiSeq using the 300 cycle v2 kit (Illumina).

[0088] Data processing of targeted sequencing results Paired FastQ reads were first filtered for Q > 30 using BBDuk from the BBTool suite and merged via BBMerge. Then, reads containing 20bp TS2 and 20bp terminal LE, each with a maximum Hamming distance of 1, were extracted. Each read was then snipped with sequences upstream and including the PAM, and downstream and including the LE, resulting in only sequences between the PAM and the LE (i.e., the site of insertion). The length of the resulting reads was calculated and used to plot the insertion distance profile from the PAM to the LE.

[0089] Unbiased genome-wide specificity analysis Depending on the donor plasmid origin (R6K or SC101), two versions of specificity analysis library preparation were performed. When the R6K origin donor was used, transposition experiments were performed by heat shocking PIR2 cells with 25 ng each of pDonor and pCAST or pHELIX. After 18 h of growth on agar plates containing 50 μg / mL kanamycin and 25 μg / mL carbenicillin, colonies were scraped and gDNA was extracted using the Wizard Genomic Purification Kit (Promega).

[0090] When the temperature-sensitive SC101 origin donor was used, electroporation with 100 ng each of pDonor and pCAST or pHELIX was performed using electrocompetent Endura cells. Cells were allowed to recover in SOC for 1 hour at 30°C, after which 100 μL of the recovery was inoculated into 3 mL of LB medium containing kanamycin and carbenicillin. Cultures were shaken at 750 RPM for 8 hours at 30°C. 150 μL of culture was plated onto carbenicillin-containing agar plates and grown for 14 hours at 42°C. Resulting colonies were scraped and gDNA was extracted using the Wizard Genomic Purification Kit (Promega) with a final resuspension step in Buffer EB without EDTA (Qiagen).

[0091] 600ng of gDNA was used as input for library preparation using the HyperPlus Kit (Roche). Briefly, gDNA was subjected to enzymatic random fragmentation for 8 minutes, ligation was performed with the fragmented gDNA and Stubby Adaptors (IDT) for 90 minutes, and the adapter-ligated fragments were bead washed using 0.9x Ampure XP beads (Beckman Coulter) (all according to the manufacturer's protocol). When the R6K origin donor was utilized, the adapter-ligated fragments were subjected to double digestion with NruI and ScaI at 37°C for 6 hours to delete the fragment resulting from the uninserted donor (for the SC101 origin, the uninserted donor was heat hardened in the previous step) and bead washed with 0.9x Ampure XP beads. Genomic LE junctions were then enriched by PCR with Q5 high fidelity DNA polymerase (NEB) using i7-specific primers and transposon LE-specific primers containing i5 adapter sequences (Appendix 3). Thermal cycling conditions were 98°C for 2 min, followed by 25 cycles of (98°C for 10 s, 66°C for 15 s, 72°C for 30 s), and a final extension of 72°C for 2 min. 50 ng of purified PCR product was used as template for a second 10-cycle PCR to add Illumina barcodes and adapter sequences (Appendix 3). The final library was quantified by Quibit Fluorimeter and submitted to the Walk-Up Sequencing service at the Broad Institute of MIT and Harvard for sequencing on a high-throughput 75-cycle NextSeq sequencing kit.

[0092] Data processing of specificity analysis results Specificity Analysis Single-end adapter-trimmed demultiplexed reads from NGS were filtered for Q > 20 and used for downstream processing using BBDuk from the BBTool suite. Reads containing 20bp of ShCAST LE were extracted and resulting reads containing 20bp of donor backbone were removed. The remaining reads contained genome-LE junctions. Reads were then trimmed of LE sequences, leaving only LE-flanking genomic sequences, and mapped to the E. coli genome (GenBank: U00096.2). Mapped reads were filtered for uniquely aligned reads. The coordinates of uniquely aligned reads were used for specificity calculation and visualization, where on-target insertion events were defined as those that occurred within 55-75bp downstream of the PAM.

[0093] Human cell culture Human HEK293T cells (ATCC) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated FBS and 1% penicillin / streptomycin (ThermoFisher) at 37°C with 5% CO. Supernatant media obtained from cell cultures was analyzed monthly for the presence of mycoplasma using MycoAlert PLUS (Lonza).

[0094] Plasmid-targeted transposition assay in human cell lysates Approximately 150,000 HEK293T cells per well were seeded in 24-well plates approximately 20 hours prior to transfection. Transfections were performed using 600 ng of DNA and 1.8 μL of TransIT-X2 (Mirus) whether a single all-in-one plasmid was used or components were expressed from individual plasmids (for the latter, 150 ng of each plasmid encoding NLS-Cas12k, NLS-TniQ, TnsC, NLS-nAniI-TnsB or NLS-TnsB was used). Transfected cells were incubated at 37°C for 48 hours, followed by cell lysates collected by removing the culture medium and adding lysis buffer (20 mM Hepes pH 7.5, 100 mM KCl, 5 mM MgCl2, 5% (vol / vol) glycerol, 1 mM DTT, 0.1% (vol / vol) TritonX-100, and 1x SigmaFast Protease Inhibitor Cocktail (EDTA-free) (wherein 1x solution is 1 tablet per 100 mL)) to each well and placed on a rocker at 4°C for 20 minutes. Suspended cells were placed into a 96-well PCR plate, vortexed vigorously for 3-5 seconds, and spun down briefly in a centrifuge to remove cell debris. Lysates were then aliquoted into PCR strip tubes and flash frozen in liquid nitrogen for further use.

[0095] N7CAST sgRNA was transcribed in vitro using a PCR template that adds a T7 promoter and a TS1 spacer to the sgRNA scaffold (T7 RiboMax Express Large Scale RNA Production System; Promega) (Appendix 3). For the transposition reaction, 15 μL of cell lysate was combined with 20 ng of pTarget, 100 ng of N7HELIX pDonor, and 1 mg of TS1 targeting sgRNA. The reactions were mixed gently and incubated at 37 °C for 4 h. To stop the reaction, 0.8 U of proteinase K (NEB) was added to each reaction and the reactions were incubated at room temperature for 15 min, followed by a heat inactivation step at 95 °C for 10 min. 2 mL of terminated and heat inactivated product was used as input for junction PCR and long-read sequencing enrichment (as described above).

[0096] Plasmid-targeted transposition assay in human cells Approximately 20,000 HEK293T cells were seeded into 96-well plates approximately 20 hours prior to transfection.

[0097] Transfections were performed using 0.6 μL of TransIT-X2 (Mirus) with 0.5, 1, 2, or 10 ng of pTarget, 80 ng of all-in-one N7CAST or N7HELIX plasmid, 60 ng of N7HELIX pDonor, 20 ng of CMV-sgRNA1 or U6-sgRNA2 plasmid, and, where applicable, 20 ng of HU expression plasmid and / or 20 ng of N7S15 expression plasmid. Transfected cells were incubated at 37°C for 72 h, culture medium was removed, and cells were lysed by addition of 100 μL of lysis buffer (20 mM Hepes pH 7.5, 100 mM KCl, 5 mM MgCl2, 5% (vol / vol) glycerol, 1 mM DTT, 0.1% (vol / vol) Triton X-100). Lysis reactions were incubated at 65°C for 6 min, followed by 98°C for 2 min. DNA (gDNA / plasmid mixture) was extracted by performing a clean-up reaction on the lysates using 1× Ampure XP beads and then used as input for junction PCR and ddPCR (as described above).

[0098] [Example 1] HELIX Development and Optimization We first sought to engineer a cointegrate-free VK-type CAST capable of cut-and-paste transposition by restoring the missing function of TnsA. To do so, we first created a fusion of the TnsA enzyme (from where it occurs as a natural TnsA-B fusion in various Tn7 transposons or type I CAST) to TnsB of the canonical VK-type CAST (ShCAST) from Scytonema hofmannii. The N-terminal domain of E. coli Tn7 TnsA executes the 5' donor cleavage, whereas the C-terminal domain interacts with downstream transposition components. 33、24The predicted structures of additional TnsA enzymes that we sought to examine also revealed a distinction between the N- and C-terminal domains (Figure 6a). Because the C-terminal domain of TnsA is not predicted to play a functional role in transposition when combined with the orthogonal VK-type CAST, we chose to fuse the N-terminal domains of various TnsAs to ShTnsB. Evaluation of ShCAST integration with TnsA-TnsB fusions revealed a substantial reduction in integration efficiency compared to wild-type ShCAST (Figure 6b). Furthermore, for the three TnsA-TnsB fusions that showed detectable integration, we observed a gradual decrease in the proportion of insertion product cointegrates in only one case (Figure 6c), while also observing an increase in the proportion of insertions occurring in the pEffector plasmid (Figure 6d).

[0099] Next, we explored the use of LAGLIDADGHE (LHE) fusions to TnsB. LHE has been utilized for genome editing in bacteria and human cells and has moderate reprogrammability via protein engineering or chimeric assembly. 34 The LHE from Aspergillus nidulans (I-AniI) has a small coding sequence (254 amino acids), cleaves a 19-bp asymmetric DNA target sequence, and contains a single K227M 29 It has previously been engineered to be a sequence-specific nickase through mutation of the .sup.1AniI gene (nAniI). Furthermore, a hyperactive variant of I-AniI, designated Y2 I-AniI, has been shown to have a nine-fold higher affinity for its cognate target site. 35We hypothesized that fusion of either nAniI or Y2 nAniI to TnsB (creating a HELIX fusion protein) could enable double nicking on the donor plasmid required for cut-and-paste DNA insertion by VK-type CAST (Fig. lc). Importantly, the recognition sequence for nAniI can be encoded on the donor plasmid backbone without complicating or restricting RNA-programmed targeting. Furthermore, the length of the nAniI recognition sequence reduces the likelihood of undesired nAniI-mediated nicking at the Cas12k-binding target site due to TnsB localization.

[0100] Therefore, we determined whether nAniI could adequately replace the lack of TnsA in ShCAST. To do so, we constructed a series of ShCAST expression plasmids each containing nAniI fused to the N- or C-terminus of (1) a single guide RNA (sgRNA) targeting target site 1 (TS1) on a separate target plasmid (pTarget), (2) Cas12k, (3) TniQ, (4) TnsC, and (5) TnsB (Figure 1d). The ShCAST expression plasmids were transformed into the donor plasmid (pDonor) described previously. 14 (containing 2.1 kb cargo and the left and right transposon ends (LE and RE, respectively) of ShCAST) were co-transformed into an E. coli strain harboring pTarget (Fig. 1d). To determine whether ShCAST retained transposition activity with the TnsB fusion to nAniI, we assessed integration by performing junction PCR across both the LE and RE in pTarget on miniprep DNA from pooled colonies harboring the transposition products. Fusion of nAniI to the N-terminus of TnsB supported RNA-guided DNA insertion, whereas a C-terminal fusion did not (Fig. 1e), suggesting that the C-terminal TnsC-interacting domain of TnsB is less accommodating for fusion proteins. 36Recent structural studies of ShCAST TnsB support this view due to the observation that a 15-residue C-terminal "hook" in TnsB is the primary means of physical TnsB-TnsC binding. 37、38 Hereafter, the nAniI-TnsB fusion architecture, together with the remaining CAST components, is referred to as HELIX (Fig. lc).

[0101] Next, to generate a 5' nick on pDonor via nAniI, we encoded I-AniI target sequences on a series of donor plasmids with variable distances to the LE / RE (Fig. 1f and Fig. 7a). When ShCAST or ShHELIX plasmids were co-transformed with various pDonors into our pTarget strain, we observed similar numbers of transformant colonies, suggesting comparable cell viability (Fig. 7b). With ShHELIX, we observed variable integration efficiencies assessed by droplet digital PCR (ddPCR) across various I-AniI-LE / RE spacings on pDonor, with 14 bp spacing resulting in the highest integration (Fig. 1f). Surprisingly, ShCAST also showed variable integration efficiencies depending on the spacing between the I-AniI site and the LE / RE site (where, unlike ShHELIX, the I-AniI site has no direct role in transposition). For ShCAST, pDonor with 4-12 bp spacing resulted in substantially higher insertion efficiency than pDonor without an I-AniI site (Figure 7c). Varying the location of the I-AniI site altered the sequences directly adjacent to the LE / RE on pDonor, suggesting that the composition of the adjacent sequences, especially the first 12 bp, may be an important determinant of integration efficiency (Figures 7a and 7c). Separately, we also performed integration experiments using Y2 nAniI fused to TnsB (Y2 ShHELIX) and observed substantially fewer colonies with peak numbers when 14 bp spacing was used (Figure 9a and Example 7). For subsequent experiments, we used HELIX constructs with nAniI-TnsB fusions and pDonor with an I-AniI site and 14 bp between the LE / RE.

[0102] Next, we used long-read sequencing to evaluate whether restoration of the 5' nick on pDonor with ShHELIX could improve product purity compared to ShCAST. We enriched for transposition products from our miniprep plasmid pool by retransforming into non-pir cells (to eliminate uninserted donor plasmid) and selecting for insertion products (Figure 8), linearized the extracted plasmid DNA, and performed long-read sequencing to determine the ratio of simple insertions versus cointegrates (Figures 1g-i). With ShCAST, we found that the purity of the transposition products was improved compared to ShCAST. 6 We observed 18.06% cointegrates consistent with HELIX (Fig. 1i). Surprisingly, ShHELIX nearly eliminated cointegrates, reducing them to only 0.49% of the total products (a 37-fold reduction when compared to ShCAST; Fig. 1h and Fig. 1i). Expression of unfused nAniI together with ShCAST did not result in a reduction in cointegrates, demonstrating that the fusion of nAniI to TnsB is important for HELIX function (Fig. 1i). Furthermore, we did not observe I-AniI sites in the insertion product reads, suggesting that the 5' flap carrying these sequences is removed during HELIX-mediated transposition (Fig. lc and Fig. 7d). We also performed long-read sequencing of Y2 ShHELIX products and similarly observed improved purity of simple insertion products with only Y2-nAniI (Fig. 9b-d).

[0103] We also performed a series of control experiments to further characterize ShHELIX (Example 8). First, a catalytically attenuated variant of I-AniI (K227M, Q171K) reduced cointegrates by 1.7-fold compared to ShCAST (presumably due to incomplete inactivation of I-AniI nicking) (Fig. 10a). Second, pDonor lacking the I-AniI target site reduced cointegrates by 1.7-fold compared to ShCAST (Fig. 10a and Example 8). Next, an experiment using pDonor with a "flipped" I-AniI site that places the nick on the same strand as the TnsB nick reduced cointegrates by 9-fold (Fig. 10b). The resulting "gapped" Shapiro intermediates are the product of 5' flap and / or gap endonucleases. 39 (in addition to the potential for low-level DSB-mediated cargo excision) could be processed by Lib4 to yield a simple insertion product (Figure 10c). Finally, the "Lib4" variant target site for I-AniI (previously shown to increase affinity of wild-type I-AniI by 5-fold) 40 When ShHELIX was used on pDonor, we observed a further reduction in cointegrates to 0.18% of all transposition products (for a 100-fold reduction in cointegrates compared to ShCAST) (Fig. 1j). However, this product purity improvement was also accompanied by a reduction in CFUs (Example 7 and Fig. 1k), and was therefore not used in subsequent experiments. In summary, ShHELIX, coupled with an I-AniI site oriented on pDonor to impart a 5' nick, showed the most significant increase in simple insertions relative to cointegrate percentages, resulting in near perfect product purity on the plasmid target.

[0104] [Example 2] Characterization of HELIX in genomic targets Prompted by our transposition results on the plasmid target, we next investigated the efficacy of ShHELIX-mediated DNA integration at the genomic site. We used a construct similar to the plasmid targeting experiment, but instead performed transformation with the genome-targeting sgRNA and without pTarget (Figure 2a). First, we examined the effect of two different lengths of amino acid linkers between nAniI and TnsB on genome integration efficiency across our set of eight donor plasmids with various distances between the I-AniI site and the LE / RE. The experiment was carried out by using a previously characterized sgRNA against the genomic target site (TS2). 14 For both amino acid linkers, we observed the highest integration efficiency when using a spacing of 14 bp between the I-AniI site and the LE / RE (Fig. 2b), which was consistent with our plasmid targeting results. All detectable insertions were in the T-LR orientation (Fig. 2c).

[0105] Having identified optimal I-AniI site spacing relative to LE / RE spacing on pDonor for genome targeting, we then compared the integration efficiency and product purity of ShCAST and ShHELIX across various genome sites. ShHELIX retained robust RNA-programmed integration across six genome target sites at levels comparable to ShCAST (Figure 2d). To analyze the on-target product purity of HELIX integration when targeting the genome with TS2, we performed long-read sequencing (following an in vitro Cas9-based genome target enrichment strategy) to determine the on-target product purity of HELIX integration. 41) was utilized. Analysis of target enriched reads using ShCAST and ShHELIX containing or lacking cargo insertions showed that integration efficiencies calculated from our long-read sequencing data were similar to our ddPCR results at TS2 (Fig. 11a). With ShCAST, we observed that 46.31% of insertion reads were cointegrates (Fig. 2e-g), indicating that integration efficiencies were consistent across different target sites and when using alternative long-read sequencing methods. 17 Despite this, the results are generally lower than previously observed. With ShHELIX, we observed only 2.97% cointegrates, a 16-fold reduction compared to ShCAST (Figure 2e-g).

[0106] Next, we evaluated the ability of ShHELIX to integrate DNA cargoes of various sizes. We performed transposition experiments using donor plasmids carrying cargoes of either 5.2, 7.8, or 9.8 kb sequences (compared to pDonor with a 2.1 kb cargo used in previous experiments). When transposing each cargo, ShHELIX showed relatively high efficiency of targeted DNA integration, regardless of cargo size (Figure 2h). Overall, our results demonstrate that ShHELIX is capable of highly active, unidirectional cut-and-paste DNA insertion and is insensitive to cargo size, at least up to 10 kb.

[0107] [Example 3] Extensibility of HELIX to VK-type CAST orthologs All discovered VK-type CASTs lack TnsA. 21This observation supports the evolutionary hypothesis that Tn5053-like transposons, which contain TnsB, TnsC, and TniQ, but not TnsA, have co-opted and repurposed this CRISPR system. Thus, all VK-type CASTs are expected to act through replicative transposition, resulting in a significant proportion of undesired cointegrate products. We therefore investigated HELIX as a generalizable approach to enable cut-and-paste DNA insertion with other diverse VK-type CASTs (Figure 3a).

[0108] To study the applicability of HELIX to other CAST orthologues, we characterized and optimized two previously reported VK-type CASTs, either from Anabaena cylindrica (AcCAST) or from a different strain of Scytonema hofmannii (ShoCAST). First, for the standard AcCAST system, we designed two sgRNA scaffolds (Figure 3b) and two pDonor architectures, the latter varied by containing different 25 bp sequences flanking the LE and RE (as previously reported for AcCAST). 14, or using ShCAST flanking sequences). Using two sgRNA designs that differ based on their crRNA-tracrRNA fusion points, we observed only minor differences in integration efficiency (Fig. 3b and Fig. 3c). However, pDonor containing ShCAST flanking sequences resulted in an increase in absolute integration efficiency of 19.6% or 20.4% for sgRNAl and sgRNA2, respectively (1.28- and 1.31-fold increase over pDonor with natural AcCAST flanks; Fig. 3c). As we previously observed for ShCAST (Fig. 7c), these results suggest that the sequences directly adjacent to the LE and RE on pDonor are key determinants of VK-type CAST-mediated integration efficiency. Moreover, AcCAST showed a minimal but still detectable number of T-RL-oriented insertions, making it a nearly perfect unidirectional inserter (Fig. 3b).

[0109] We constructed AcHELIX containing the nAniI-TnsB fusion with sgRNA2 design and pDonor carrying an I-AniI site 14bp from the LE / RE separated by ShCAST flanking sequences (Fig. 3d). To determine the integration product purity with AcHELIX compared to AcCAST when targeting genomes, we performed Cas9 target enrichment followed by long-read sequencing (Fig. 3e). With AcCAST, we observed 37.99% cointegrate products, whereas for AcHELIX, we found only 0.60%, indicating a 63-fold improvement in product purity with AcHELIX (Fig. 3f and Fig. 3g). Across six genome targets, AcHELIX retained RNA-guided DNA integration and insertion directionality comparable to AcCAST (Fig. 3h, Fig. 3i, and Fig. 11a and Fig. 11b). Furthermore, similar to ShHELIX, AcHELIX showed no decline in efficiency when integrating cargo sequences of various sizes up to 9.8 kb, and maintained integration efficiencies above 83% for all four cargo sizes at TS6 (Figure 3j). Thus, similar to ShHELIX, AcHELIX is an effective engineered CAST with near-perfect simple insertion product purity for DNA inserts of various sizes.

[0110] Next, we characterized ShoCAST and ShoHELIX using pDonor with 14 bp spacing separating the I-AniI site and the LE / RE with the ShCAST flanking sequence (Figure 3k). 16We performed genome targeting experiments with ShoCAST and ShoHELIX using the ELISA kit. Characterization of the insertion products by long-read sequencing revealed 54.09% integrants for ShoCAST and 21.37% integrants for ShoHELIX, indicating a 2.5-fold reduction in cointegrates when ShoHELIX was used (Fig. 3l-3m). Across genome targets TS2-TS7, we observed a range of integration efficiencies, with ShoHELIX showing integration comparable to ShoCAST (Fig. 3o and Fig. 11a and 11b). Similar to AcCAST and AcHELIX, the orientation of ShoCAST and ShoHELIX insertions was predominantly T-LR oriented, despite detectable T-RL insertions (Fig. 3o and Fig. 3p). Furthermore, in contrast to ShHELIX and AcHELIX, ShoHELIX showed a decrease in integration efficiency with increasing cargo size on pDonor at TS3 (Fig. 3q). Finally, to test whether the nAniI fusion to TnsB alters the distance between the PAM and the insertion site, we performed amplicon sequencing across the genome-LE junction (Fig. S12a). ShHELIX, AcHELIX, and ShoHELIX did not alter the insertion distance profile of their canonical CASTs (Fig. S12b-g).

[0111] [Example 4] Comparison of types I, VK, and HELIX A streamlined type I CAST, termed INTEGRATE, has recently been described. 16Therefore, we sought to compare the efficiency and directionality of integration using ShHELIX and AcHELIX with Vibrio Cholerae INTEGRATE. We performed transposition assays that controlled for growth time (24 hours), donor cargo size (2.1 kb), approximate donor copy number (high copy), cell type (PIR1), common genomic target location (by the closest compatible PAM), and efficiency measurement method (ddPCR) (Figure 13a). We found that HELIX was more efficient or equally efficient to INTEGRATE, depending on the construct and growth temperature used (Figure 13b). Notably, for INTEGRATE-mediated insertions performed at 30°C, we observed substantial integration in the reverse orientation (Figure 13c).

[0112] [Example 5] Characterization and optimization of VK-type CAST and HELIX specificities In contrast to the highly specific insertion profile of type I CAST, VK-type CAST is prone to off-target integrations that are spread across the bacterial genome 14、16、17、20 Recent structural studies of ShCAST revealed Cas12k-independent TnsC filament formation on DNA in a sequence-independent manner. 36、42、43 (Similar to MuB in Mu transposase 44 ), potentially leading to off-target integration due to non-targeted assembly of the transpososome. TniQ has also been shown to play an important role in transposition events by capping and nucleating TnsC filaments. 42、43 Therefore, one potential approach to increase the specificity of VK-type CAST is to fuse TnsC and / or TniQ to Cas12k and localize the transposition event to the Cas12k-target-bound DNA.

[0113] To test this hypothesis, we constructed various three-component ShCAST systems in which Cas12k was fused to TniQ or TnsC in any orientation, as well as two-component systems that fused Cas12k, TniQ, and TnsC (Figure 4a). Transposition experiments demonstrated that Cas12k-TniQ, Cas12k-TniQ-TniQ, and Cas12k-TnsC fusions retained most of their activity compared to unfused standard CAST (Figure 4b and Figure 14a). HELIX versions of these three best-performing fusion constructs also maintained substantial integration at TS2 and TS5 (Figure 4c, Figure 4d, and Figure 14b). Furthermore, ShCAST and ShHELIX with Cas12k fusions did not change the distance between the PAM and the integration site (Figure 12h-7m). Both ShCAST and ShHELIX with or without the Cas12k-TnsC fusion preserved target immunity (Fig. 4e), such that sites that underwent integration events were resistant to subsequent integration. 14、45、46 Our observation that Cas12k-TniQ fusions retain functionality in combination with identical insertion distance profiles for all fusions supports the proposed model in which Cas12k and TniQ directly associate during transposition. 42、43 .

[0114] To compare the specificity of ShCAST, ShHELIX, and versions carrying Cas12k-TniQ or -TnsC fusions, we performed an unbiased analysis of genome-wide integration of previously described methods. 14、16、20 Analogously, we performed transformation in Endura cells and analyzed insertion specificity by random enzymatic fragmentation of genomic DNA followed by integration junction enrichment and sequencing. Our results revealed 54.4% on-target integration when targeting TS2 with ShCAST (Fig. 4f), which is the highest reported value for this target site to date. 14The specificity profile was consistent with that of ShCAST. Strikingly, ShHELIX showed 88.4% on-target integration with TS2 sgRNA, a 34% absolute increase in on-target specificity compared to ShCAST (Figure 4f and Figure 15a,b). Furthermore, the use of ShHELIX with a donor that does not contain an I-AniI site or dShHELIX (containing catalytically inactive I-AniI) also showed on-target specificity of over 88% (Figure 15b), indicating that neither I-AniI binding nor cleavage is the primary cause of this 1.6-fold enhanced specificity. Instead, these results potentially indicate how fusion of nAniI to TnsB structurally alters the CAST conformation and / or how TnsB distorts the donor topology to make translocation at sites where Cas12k is not bound energetically unfavorable. Similar experiments with ShHELIX containing Cas12k-TniQ and Cas12k-TnsC fusions further improved the specificity to 94.5% and 96.5% on-target integration, respectively (Fig. 4f). Comparable ShCAST specificity with Cas12k-TniQ and Cas12k-TnsC fusions was 65.3% and 51.7%, respectively (Fig. 4f and Fig. 15a). We also evaluated integration specificity in another E. coli strain by performing genome-wide insertion analysis in PIR2 cells (Fig. 15c and Fig. 15d). Interestingly, we observed enhanced on-target specificity for all conditions, with ShHELIX constructs achieving over 97% on-target integration (Fig. 4f and Fig. 15c). Moreover, this highly specific ShCAST- and ShHELIX-mediated transposition in PIR2 cells did not reduce transposition efficiency (FIG. 16).

[0115] The major genotypic difference between the Endura and PIR2 strains is the pir gene of PIR cells, which encodes the pi protein required for conditional replication of the R6K-origin plasmid. 47、48Therefore, we sought to determine whether pir co-expression could increase the specificity of HELIX in non-pir cells, potentially obviating the need to alter the efficiency of Cas12k fusions. To do so, we co-expressed either the wild-type pir gene or the pir116 mutant, which has been shown to initiate higher copy replication of the R6K origin plasmid. 48 ) were cloned into separate plasmids carrying pDonor and ShCAST or ShHELIX plasmids containing the TS2 genome targeting sgRNA (Figure 4g). Specificity profiling revealed that wild-type pi, together with ShHELIX, provided an additional absolute 7.6% increase in specificity, with 96.0% of reads occurring at the on-target site (Figure 4h) (comparable to the specificity observed with ShHELIX and Cas12k-TniQ or Cas12k-TnsC fusions in PIR2 cells, Figure 4f). Co-expression of pi with ShCAST or mutant pi with either ShCAST or ShHELIX provided only minor changes in specificity (Figure 4h).

[0116] Comparative mapping of genome-wide integration sites of ShCAST (Figure 4i), ShHELIX with Cas1k-TniQ (Figure 4j), ShHELIX with Cas12k-TnsC (Figure 4k), and ShHELIX with pi co-expression (no fusion) (Figure 4l) from specificity experiments performed in Endura cells visualized a significant reduction in genome-wide off-target integration events when using the ShHELIX system. Furthermore, comparison of specificity profiles for ShCAST with or without pi protein co-expression reveals that pi protein generally reduces the distribution of off-target integrations but increases their occurrence at select sites (Figure 15a). A similar trend was observed when using ShHELIX and pi protein co-expression, but less dramatic due to higher on-target integration specificity (Figure 15b). Overall, ShHELIX, coupled with component fusions (albeit at the expense of some integration efficiency) and pi co-expression, substantially improves the genome-wide specificity of the VK-type system, while using fewer molecular components and a smaller code size, and provides a novel approach to the development of the VK-type system. 15~17、49 We can achieve levels of on-target integration comparable to those achieved by the ELISA kit (Figure 17).

[0117] Example 6: HELIX-mediated DNA integration in a human cell context The ability to accomplish targeted DNA insertion in human cells has significant implications for basic research and therapy. To determine whether the CAST or HELIX system can function in human cells, we first determined whether ShCAST or AcCAST can function in a human context by attempting a lysate-based insertion assay. Plasmids encoding human codon-optimized CAST components were transfected into HEK293T cells, incubated for 48 hours, and then lysed. Subsequently, HEK293T human cell lysates containing CAST proteins were incubated with pDonor, pTarget, and an in vitro transcribed sgRNA targeting TS1 on pTarget. However, for both ShCAST or AcCAST, we did not detect insertion into pTarget by junction PCR for the conditions tested. Next, given the generalizability of HELIX to various orthologs, we explored other CASTs and found that have previously been shown to function in human cell lysates. 50We identified VK-type CAST from Nostoc sp. PCC7101 (N7CAST; Figure 18a). After confirming that N7CAST could demonstrate detectable DNA insertion of sgRNA to TS1 on pTarget in HEK293T cell lysates (Figure 18b), we constructed an initial, unoptimized N7HELIX system (Figure 5a and Example 10). Transposition experiments with N7HELIX in lysates, followed by junction PCR on pTarget, yielded amplicons of the correct size (Figure 5b, Figure 5c), indicating productive insertion. Sanger sequencing of these amplicons revealed donor insertion downstream of TS1 with expected overlap of the target site at the insertion site (Figure 5d), and high-throughput sequencing revealed that insertion occurs primarily 57-62 bp downstream of the PAM (Figure 5e). To determine whether N7HELIX could improve the desired insert purity by reducing cointegrate products compared to N7CAST, we utilized a PCR enrichment strategy on our lysate reactions and used long-read sequencing (Example 11). We observed 41.9% cointegrates when using N7CAST, whereas a comparable experiment with N7HELIX yielded only 7.9% cointegrate products (a 5.3-fold reduction, FIG. 5f), demonstrating the scalability of HELIX to the human cell context.

[0118] We then attempted to streamline N7HELIX for experiments in human cells by constructing a single all-in-one expression plasmid while also varying the sequences of the sgRNA scaffold and promoter (Fig. 18c and Example 10). When human cell lysates containing N7HELIX expressed from the all-in-one plasmid were incubated with sgRNA2 (containing a poly-T stretch mutated in the wild-type sgRNA to allow U6 promoter compatibility), pDonor, and pTarget, we observed sgRNA-dependent DNA insertion at TS1, confirming that all components were active when expressed from a single plasmid (Fig. 18d). We then assessed whether N7HELIX could mediate targeted DNA integration in human cells. We co-transfected pTarget and pDonor with plasmids encoding N7CAST or N7HELIX and either U6-sgRNA2 or CMV-driven wild-type sgRNA flanked by hammerhead and HDV ribozymes (Fig. 5g). However, no DNA integration was detected by junction PCR (Fig. 18e). Recent studies reveal that ribosomal S15 may be a critical component of VK-type CAST by facilitating complex assembly. 43 Informed by (Example 10), we next attempted co-transfection of the same plasmids, but now also containing a plasmid encoding N7S15 (Fig. 5g). Junction PCR across the left transposon end of extracted plasmid DNA revealed N7CAST or N7HELIX-mediated donor integration on pTarget only with N7S15 and U6-sgRNA2 (Fig. 5h, Fig. 18e, and Example 10). Quantification of DNA insertion into pTarget revealed comparable integration between N7CAST and N7HELIX in the presence of N7S15, albeit at a lower efficiency (Fig. 5i). Structural and functional similarities between TnsB and TnsC in VK-type CAST for MuA and MuB, respectively, of Mu transposons 37、42, and the requirement for the host cofactor HU in Mu transposition 51 Considering this, we next attempted transposition with N7CAST or N7HELIX together with co-transfection of additional plasmids expressing N7S15 and N7HU. Integration quantification showed similar efficiency with or without HU co-expression (Fig. 5j). Next, experiments in HEK293T cells targeting endogenous genomic target sites with N7CAST or N7HELIX and co-expression of N7S15 (but not N7HU) showed minimal but detectable insertions in VEGFA and EMX1 (Fig. 5k). Overall, these results demonstrate the scalability of HELIX to the human cell context in the presence of S15, motivating the continued development of CAST and HELIX to achieve higher levels of integration in mammalian genomes (Fig. 5l).

[0119] [Example 7] Y2 Extending the discussion of ShHELIX results While developing and characterizing ShHELIX, we identified a Y2 nAniI variant that was previously shown to have 9-fold higher affinity for its cognate target site. 1 We assessed whether Y2 ShHELIX constructs could further increase the purity of the simple insertion product. With the Y2 ShHELIX construct, we observed a reduction in transformant colonies (Figure 8a) when compared to ShCAST or non-Y2 ShHELIX (Figure 6a). Furthermore, this reduction varied with the spacing between the I-AniI site on pDonor and the LE / RE, where a spacing of 14 bp showed the highest number of colony forming units (CFU) (also consistent with the spacing that confers the highest integration efficiency by ddPCR on plasmid and genomic targets). Combined with a similar observation (as shown in Figure 1k) when using the Lib4 I-AniI site (where the Lib4 I-AniI site was previously shown to increase wild-type I-AniI affinity sites by 5-fold), we found that the Lib4 I-AniI site increased the number of transformant colonies by 5-fold when compared to ShCAST or non-Y2 ShHELIX (Figure 6a). Furthermore, this reduction varied with the spacing between the I-AniI site on pDonor and the LE / RE, where a spacing of 14 bp showed the highest number of colony forming units (CFU) (also consistent with the spacing that confers the highest integration efficiency by ddPCR on plasmid and genomic targets). Combined with a similar observation (as shown in Figure 1k) when using the Lib4 I-AniI site (where the Lib4 I-AniI site was previously shown to increase wild-type I-AniI affinity sites by 5-fold). 2), we recognized a potential correlation between the affinity of I-AniI for its target sequence and the number of colonies present on plates selecting for pShHELIX or pShCAST, pDonor and / or the transposition products, and pTarget.

[0120] While further studies into the mechanism of HELIX will elucidate the basis for the decreased cell viability when using Y2-ShHELIX, we speculate that a combination of two phenomena may be occurring. First, the higher affinity of Y2 nAniI to its target, or the use of nAniI at the Lib4 site, causes an increased prevalence of DNA double-strand breaks (DSBs) on pDonor at early time points of recovery after transformation. In the absence of rapid and efficient cargo integration into pTarget, DSBs caused by AniI lead to loss of kanamycin resistance due to pDonor degradation before transposition. In this scenario, colony counts for different spacings on pDonor may correlate with higher or lower integration efficiency. For example, for spacings where transposition is most efficient and rapid, the loss of CFUs is less pronounced because integration into pTarget occurs more quickly than DSBs on pDonor. The second hypothesis is that the higher affinity of Y2 nAniI for its target, or the use of nAniI at the Lib4 site, causes increased occurrence of DSBs on pDonor. Given the high copy number of pDonor in PIR1 cells, this could lead to the induction of an SOS response and cell death.

[0121] [Example 8] ShHELIX control experiment While performing long-read sequencing of the transposition products obtained from the plasmid targeting experiments, we included several control conditions. First, we isolated a catalytically attenuated I-AniI variant (harboring the K227M and Q171K mutations) and isolated a catalytically attenuated I-AniI variant (harboring the K227M and Q171K mutations). 3) to create a "dead" ShHELIX (dShHELIX). With dShHELIX, we observed a 1.8-fold decrease in cointegrate products compared to wild-type ShCAST (Figures 9a and 1i, respectively). We hypothesize that this somewhat unexpected decrease in cointegrate products is the result of incomplete inactivation of the I-AniI catalyst, which may lead to low levels of 5' pDonor nicking (at a slower rate than nAniI-based ShHELIX). Indeed, the I-AniI Q171K variant has previously been shown to exhibit residual nicking activity on both DNA strands in vitro. 3 .

[0122] Second, we performed experiments using a pDonor variant that does not carry an I-AniI site. In transformation with this modified pDonor, which lacks ShHELIX and I-AniI sites, we observed a 1.7-fold reduction in cointegrates compared to ShCAST (Fig. 9a and Fig. 1i, respectively). We hypothesize that this may be due to low levels of I-AniI activity on sequences adjacent to the LE and RE (where tethering to TnsB induces energetically unfavorable interactions that do not occur in the absence of fusion). Previous studies in which each base in the I-AniI recognition sequence was mutated to every other base revealed that the specificity of nAniI was greatest over the half base pair positions ±3, 4, 5, and 6 of each site, and least specific over bases -2 to +1 and the bases at the outer edge of the recognition sequence. 3From this data, I-AniI recognition requires a minimal approximate core sequence of 5'-GAGGNNNCTG-3', with activity decreasing depending on the base substituted. While we are unable to identify the exact sequence match, we note that sequences similar to these core motifs occur on pDonor at 5'-GTGGNNNNGTCTA-3' (11 bp from LE) and 5'-GAGGNNNCATTG-3' (13 bp from RE), the latter in an orientation that nicks the same strand as TnsB (see next point). Low level nicking on these flanking sequences at these degenerate I-AniI core sequences may cause a slight increase in the purity of the simple insertion product (as observed).

[0123] Third, we performed experiments using a "flipped" I-AniI site on pDonor oriented to nick the same strand as TnsB. In experiments using pDonor with a flipped I-AniI site, we observed a 10-fold reduction in cointegrates with ShHELIX compared to ShCAST (Figure 9b). We hypothesize that this reduction in cointegrates may be the result of an alternative transposition mechanism involving 5' flap cleavage of the gapped Shapiro intermediate (Figure 9c).

[0124] [Example 9] Mechanistic implications of Cas12k-TnsC fusions Recent structural studies have provided insight into the mechanism of ShCAST-mediated DNA insertion. 4~6These studies suggest that TnsB recruitment to TniQ-nucleated TnsC filaments mimics filament disassembly, exposing the target site and directing insertion at a coordinated distance from the sgRNA-Cas12k-DNA complex. Our experiments with fusions of Cas12k to TnsC monomers in the context of ShCAST or ShHELIX (Figure 3) are intriguing given these proposed mechanisms, particularly with regard to the role of TnsC filamentation in recruiting downstream translocation machinery. Furthermore, because the extent of TnsC filament disassembly (or footprint of TniQ alone or bound to TnsC) may dictate the insertion distance from the bound DNA-bound Cas12k in the context of standard four-component ShCAST, it is intriguing that Cas12k-TnsC fusions (in the context of ShCAST and ShHELIX systems) enable target DNA insertion with the same insertion distance profile as standard four-component ShCAST and ShHELIX systems (Figure 12). We speculate that TnsC filamentation may still occur despite Cas12k fusion, or that only a single TnsC subunit fused to Cas12k is sufficient to allow translocation. In the latter case, TnsB-mediated depolymerization could collapse the TnsC filament into a single monomer, resulting in the fixed insertion distance profile observed for the native system, consistent with the same profile observed for our monomer fusion. Alternatively, TnsC may not be involved in determining the insertion distance, and a TniQ and TnsB-defined insertion distance model may be more plausible. However, the molecular ruler mechanism of CAST remains unclear. Furthermore, for ShCAST, our results reveal that Cas12k-TniQ-TnsC fusions are functional (albeit with reduced activity), whereas Cas12k-TnsC-TniQ fusions completely abolish activity (Figure 4b). This observation supports the current model, in which Cas12k and TniQ must be able to directly interact. 5Our results with Cas12k-TnsC and Cas12k-TniQ-TnsC fusions provide insight into the roles of TnsC and TniQ in ShCAST-mediated transposition and motivate further studies to elucidate the transposition mechanisms of both native CAST and engineered HELIX 2-, 3-, or 4-component systems.

[0125] [Example 10] Construction and characterization of N7HELIX in a human cellular context To construct N7HELIX, a human codon-optimized nicking variant of I-AniI was fused to N7TnsB via an 18 amino acid XTEN linker. The I-AniI site was placed 14 bp from the LE and RE on pDonor in the correct orientation to confer a 5' nick, and the flanking sequences immediately adjacent to the LE and RE were exchanged with those of ShCAST (Figure 5a). While this donor flank configuration was the most efficient for ShHELIX, it is possible that N7-specific optimization for N7HELIX may result in higher integration efficiency. To streamline N7HELIX expression, we used a 18 amino acid XTEN linker as previously described. 7 To this end, we constructed a single all-in-one plasmid in which all four HELIX components were driven by a single CMV promoter. Specifically, NLS-Cas12k and TnsC, as well as NLS-nAniI-TnsB and NLS-TniQ, were linked by a T2A sequence. The polypeptide pairs were separated by an EMCV internal ribosome entry site (IRES) (Figure 17c). We also eliminated several poly-T stretches within the scaffold of the wild-type sgRNA, which may act as a termination signal for the U6 promoter. 8 ) was generated (sgRNA2) with a substitution in (Figure 17c).

[0126] Recent studies have demonstrated that the bacterial host-encoded ribosomal protein S15 is a bona fide component of VK-type CAST and allosterically stimulates complex assembly at Cas12k-binding target sites.5 Surprisingly, the ShCAST sgRNA scaffold secondary structure to which S15 was found to bind is strikingly similar to the secondary structure of 16S rRNA, to which S15 binds in its primary role in promoting ribosomal complex assembly. Both E. coli S15 (EcS15) and S. Hofmannii S15 (ShS15) have previously been shown to substantially enhance translocation in vitro. 5 Due to these observations, we generated expression plasmids for both the N7 ribosomal protein S15 (N7S15) and EcS15 to determine whether they could promote N7CAST and N7HELIX (Fig. 5g, Fig. 5h, and Fig. 18e). We found that N7S15 coexpression was required for N7CAST and N7HELIX integration in human cells (Fig. 18e), confirming previous findings that S15 is likely required for optimal targeted integration and that S15 should be heterologously expressed when VK-type CAST or HELIX are used in human cells. 5 Under the conditions we examined, we did not observe N7CAST and N7HELIX integration in human cells when EcS15 was coexpressed (FIG. 18e).

[0127] Although CAST and HELIX-mediated transposition was detected in human cells when S15 was expressed, the overall insertion efficiency remained low for the constructs and conditions examined. Expanding on our paper, the discovery of additional essential host factors involved in VK-type CAST function, as well as the screening of VK-type CAST orthologues that may naturally fit into the human cell context, are necessary. Directed evolution and structure-guided engineering of the CAST system, particularly TnsB and Cas12k, may allow for more efficient integration on human genome targets. In addition, continued optimization of protein and sgRNA expression constructs and methods will prove important given the complexity of these systems and the need to localize all components to the nucleus. Optimized component fusions may prove useful to help promote nuclear localization.

[0128] It should also be noted that the HELIX architecture may require optimization for each CAST ortholog. These optimizations include the spacing between the I-AniI site and the LE / RE, the linker between nAniI and TnsB or other components (if applicable), the uniqueness of the LHE itself, and the adjacent sequences on the donor. We designed and constructed N7HELIX according to the optimal parameters from our ShHELIX / AcHELIX experiments, so we did not perform system-specific optimization for the other orthologs described in this study (AcCAST, ShoCAST, and N7CAST). Thus, ortholog-specific optimization may enable more efficient HELIX-mediated human genome targeting.

[0129] [Example 11] Characterization of cointegrates from experiments in HEK293T cell lysates We investigated the scalability of HELIX to reduce cointegrates relative to its standard CAST in the context of human cells. Due to the low efficiency of transposition in human lysates with the constructs and conditions we examined, the enrichment process we utilized for bacterial plasmid targeting experiments was not feasible or applicable to experiments performed in human lysates. Therefore, we chose to utilize a PCR-based enrichment strategy from lysate reactions to quantify the approximate ratio of simple insertions versus cointegrate products (see figure below). Two separate 20-cycle PCRs were performed, each using the same volume of terminated lysate reaction as template, differing only in the sequence of the downstream reverse primer. The PCRs attempted to amplify (A) from upstream of TS1 on pTarget to the edge of the RE on the inserted cargo (to approximate the "total" insertion) and (B) from upstream of TS1 on pTarget (same 5' primer as in the first PCR reaction) to the donor backbone near the edge of the RE. As described in the methods, both PCRs were performed for CAST and HELIX, and the PCRs were combined and analyzed by long-read sequencing. Reads from PCR-A represent the "total" insertions, whereas reads from PCR-B represent the "cointegrate" insertions. The ratio of "cointegrates" to "total insertions" was used to estimate the relative proportion of cointegrates from the total transposition products, but this was an approximate quantification and was only intended to compare the relative differences between CAST and HELIX.

[0130] Exemplary Sequences Note: The sequences vary for the different CAST systems in which HELIX is applied. For the ones used in this study, see:

[0131] ShCAST subunits ShCAST Cas12k

[0132] [ka]

[0133] ShCAST TnsB

[0134] [ka]

[0135] ShCAST TnsC

[0136] [ka]

[0137] ShCAST TniQ

[0138] [ka]

[0139] ShCAST sgRNA scaffold ribonucleotides

[0140] [ka]

[0141] AcCAST Cas12k amino acids

[0142] [ka]

[0143] AcCAST TnsB

[0144] [ka]

[0145] AcCAST TnsC

[0146] [ka]

[0147] AcCAST TniQ

[0148] [ka]

[0149] AcCAST sgRNA scaffold

[0150] [ka]

[0151] ShoCAST Cas12k

[0152] [ka]

[0153] ShoCAST TnsB

[0154] [ka]

[0155] ShoCAST TnsC

[0156] [ka]

[0157] ShoCAST TniQ

[0158] [ka]

[0159] ShoCAST sgRNA scaffold

[0160] [ka]

[0161] N7CAST Cas12k

[0162] [ka]

[0163] N7CAST TnsB

[0164] [ka]

[0165] N7CAST TnsC

[0166] [ka]

[0167] N7CAST TniQ

[0168] [ka]

[0169] N7CAST sgRNA scaffold (wild type sequence)

[0170] [ka]

[0171] N7CAST sgRNA scaffold (a poly-U stretch in the wild-type scaffold mutated to reduce or prevent premature transcription termination)

[0172] [ka]

[0173] I-AniI and variants: Wild-type I-AniI amino acid sequence

[0174] [ka]

[0175] I-AniI amino acid sequence containing two mutations (F80K, L232K) that confer increased solubility / solution behavior

[0176] [ka]

[0177] Nicking variants of I-AniI amino acid sequence (also contains solution behavior mutations F80K, L232K, and K227)

[0178] [ka]

[0179] Y2 I-AniI-amino acid sequence carrying two additional mutations (F80K, L232K, F13Y, S111Y) that have been shown to increase affinity by 9-fold

[0180] [ka]

[0181] Y2 I-AniI amino acid sequence nicking variants (F80K, L232K, K227M, F13Y, S111Y)

[0182] [ka]

[0183] TnsB fusions (expressed together with TnsC, TniQ, and Cas12k in the HELIX system) nAniI-XTEN18-ShTnsB: NickingI-AniI fused to ShCAST TnsB using an 18 amino acid XTEN linker

[0184] [ka]

[0185] Y2 nAniI-XTEN18-ShTnsB: Nicking I-AniI fused to ShCAST TnsB using an 18 amino acid XTEN linker

[0186] [ka]

[0187] nAniI-XTEN18-AcTnsB: Nicking I-AniI fused to AcCAST TnsB using an 18 amino acid XTEN linker (as seen in line 26)

[0188] [ka]

[0189] nAniI-XTEN18-ShoTnsB: NickingI-AniI fused to ShoCAST TnsB using an 18 amino acid XTEN linker

[0190] [ka]

[0191] nAniI-XTEN18-N7TnsB: Nicking NLS-I-AniI fused to N7CAST TnsB using an 18 amino acid XTEN linker

[0192] [ka]

[0193] Cas12k fusions to generate 3-component CAST (TnsB fused to nothing) or 3-component HELIX (nAniI-TnsB) Cas12k-XTEN18-TniQ: ShCAST Cas12k fused to ShCAST TniQ via an 18 amino acid XTEN linker; the other two components are TnsB (or nAniI-TnsB for HELIX) and TnsC

[0194] [ka]

[0195] Cas12k-XTEN18-TniQ-3×GGGS-TniQ: ShCAST Cas12k fused to ShCAST TniQ via an 18 amino acid XTEN linker. The two TniQs are fused via a 3×(GGGS) linker; the other two components are TnsB (or nAniI-TnsB for HELIX) and TnsC.

[0196] [ka]

[0197] Cas12k-XTEN18-TnsC: ShCAST Cas12k fused to ShCAST TnsC via an 18 amino acid XTEN linker; the other two components are TnsB (or nAniI-TnsB for HELIX) and TnsQ

[0198] [ka]

[0199] Cas12k-XTEN18-TniQ-3×GGGS-TnsC: ShCAST Cas12k fused to ShCAST TniQ via an 18 amino acid XTEN linker fused to ShCAST TnsC via a 3×(GGGS) linker

[0200] [ka]

[0201] Cas12k-XTEN18-TnsC-3×GGGS-TniQ: ShCAST Cas12k fused to ShCAST TnsC via an 18 amino acid XTEN linker fused to ShCAST TniQ via a 3×(GGGS) linker

[0202] [ka]

[0203] pDONOR sequence without I-AniI site (LE is underlined, RE is italicized) ShCAST pDonor with natural flanking sequences (no I-AniI site)

[0204] [ka]

[0205] AcCAST pDonor with natural flanking sequences (no I-AniI site)

[0206] [ka]

[0207] ShoCAST pDonor with ShCAST flanking sequences (no I-AniI site)

[0208] [ka]

[0209] AcCAST pDonor with ShCAST flanking sequences (no I-AniI site)

[0210] [ka]

[0211] N7CAST pDonor (no I-AniI site) with natural flanking sequences and 400bp of LE / RE (not minimized)

[0212] [ka]

[0213] [Table 5-1]

[0214] [Table 5-2]

[0215] [Table 5-3]

[0216] [Table 5-4]

[0217] References for Examples 1-6:

[0218] [Table 6-1]

[0219] [Table 6-2]

[0220] [Table 6-3]

[0221] [Table 6-4]

[0222] [Table 6-5]

[0223] [Table 6-6]

[0224] [Table 6-7]

[0225] References for Examples 7-11:

[0226] [Table 7]

[0227] Other embodiments Although the present invention has been described in conjunction with the detailed description thereof, it will be understood that the foregoing description is intended to be illustrative of the invention, as defined by the appended claims, and is not intended to limit its scope. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

1. A fusion protein comprising a transposition protein B (TnsB) protein, e.g., Tn7, Tn7-like, or Tn5053-like transposition protein B (TnsB), fused to a protein (e.g., an endonuclease, e.g., nickase, cribase, or an endonuclease without catalytic activity, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)).

2. The fusion protein according to claim 1, wherein the endonuclease is a nickasase, for example, a homing endonuclease (HE), a nicking restriction endonuclease, a nicking Cas variant, or a phage HNH endonuclease, or a TnsA or Tn7 transposon derived from type I CAST, or a catalytic portion thereof.

3. The fusion protein according to claim 2, wherein the HE is LAGLIDADG, H-N-H, His-Cys box, or GIY-YIGHE.

4. The fusion protein according to claim 3, wherein the HE is I-AniI, for example, I-AniI (I-AniI) or a variant derived from Aspergillus nidulans, which optionally includes the K227M mutation (nAniI), a highly active variant (e.g., Y2 I-AniI (F13Y, S111Y)), or both (K227M, F13Y, S111Y).

5. A nucleic acid comprising a sequence encoding the fusion protein described in Claim 1.

6. An expression construct comprising the nucleic acid according to claim 5, and a regulatory sequence for expressing the protein, such as a promoter.

7. An expression construct comprising a sequence encoding a CRISPR-related transposase (CAST), wherein the sequence comprises a nucleic acid encoding the fusion protein described in Claim 1, Cas12k; TnsC; TniQ; optionally one or more host proteins; and a guide RNA (gRNA) that interacts with Cas12k and directs the Cas12k / gRNA complex to a target sequence; and a regulatory sequence for expressing the sequence, for example, one or more promoter sequences.

8. The expression construct according to claim 7, wherein Cas12k is optionally fused to at least one other protein, optionally TniQ and / or TnsC, with a linker between each protein (e.g., Cas12k-TniQ, Cas12k-TniQ-TniQ, Cas12k-TnsC, Cas12k-TniQ-TnsC, or Cas12k-TnsC-TniQ).

9. The expression construct according to claim 8, which is a plasmid or a viral vector.

10. A nucleic acid according to claim 5 comprising a nucleic acid sequence encoding a Tn-endonuclease fusion protein, for example, a TnsB-endonuclease fusion protein; and optionally one or more, for example, all Cas12k; TnsC; TniQ; optionally one or more host proteins; and optionally a host cell expressing them, comprising a guide RNA that binds to cas12k and directs the TnsB-endonuclease fusion protein to a selected target sequence.

11. A host cell comprising a CRISPR-related transposase (CAST) containing the fusion protein described in Claim 1; Cas12k; TnsC; TniQ; optionally one or more host proteins; and a gRNA that interacts with Cas12k and directs the fusion protein to a selected target sequence.

12. The host cell according to claim 10, wherein Cas12k is optionally fused with at least one other protein, optionally TniQ (e.g., Cas12k-TniQ, TniQ-Cas12k, TniQ-TniQ-Cas12k, TniQ-Cas12k-TniQ, or Cas12k-TniQ-TniQ) and / or at least one TnsC, with linkers between each protein.

13. The host cell according to claim 11, wherein Cas12k is optionally fused with at least one other protein, optionally TniQ (e.g., Cas12k-TniQ, TniQ-Cas12k, TniQ-TniQ-Cas12k, TniQ-Cas12k-TniQ, or Cas12k-TniQ-TniQ) and / or at least one TnsC, with linkers between each protein.

14. A method for inserting a desired sequence into DNA, for example, into the genomic DNA of a living cell, the method comprising expressing in the cell a donor DNA molecule (e.g., plasmid) comprising the nucleic acid described in claim 5; Cas12k; TnsC; TniQ; optionally, one or more host proteins; and a guide RNA that binds to cas12k and directs the endonuclease to a selected target sequence, and the desired sequence to be inserted, wherein the desired sequence is flanked by LE and RE flanking sequences and a target site for the endonuclease (e.g., I-AniI) at its 5' and 3' ends, respectively, and preferably the target site is oriented to nick the donor plasmid 5' of the desired sequence to be inserted.

15. The method of claim 12, wherein the donor DNA molecule is derived from a source organism other than the source organism of at least one of the CAST components, i.e., TnsB;cas12k;TnsC; or TniQ, and / or has a fringe sequence as shown in Table A, which includes modifications or insertions at various distances from the LE and RE sequences (e.g., endonuclease recognition sequences or host factor binding sequences(s)).

16. The method according to claim 15, wherein the modified LE / RE facilitation sequence is derived from Scytonema hofmannii (e.g., from ShCAST), and at least one of the Tn proteins; cas12k; TnsC; or TniQ is derived from CAST or HELIX orthologues (e.g., AcCAST and AcHELIX), or is a modified ShCAST LE / RE facilitation sequence, or is a de novo LE / RE facilitation sequence.

17. The method according to claim 14, wherein Cas12k is expressed as a fusion protein (e.g., Cas12k-TniQ, Cas12k-TniQ-TniQ, Cas12k-TnsC, Cas12k-TniQ-TnsC, or Cas12k-TnsC-TniQ) with optionally linkers between each protein.

18. A fusion protein comprising Cas12k, optionally with linkers between each segment; optionally, one or more host proteins; and at least one TniQ (e.g., Cas12k-TniQ or Cas12k-TniQ-TniQ) and / or at least one TnsC.

19. A fusion protein comprising a host protein and one or more Cas12k, TnsC, or TniQ, with linkers optionally between each segment.

20. (i) a fusion protein comprising a transposon (Tn) protein, e.g., Tn7, Tn7-like, or Tn5053-like, e.g., transposition protein B (TnsB), fused to a protein (e.g., an endonuclease, e.g., nickase, cribase, or an endonuclease without catalytic activity, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)) via an intervening linker of any choice; and (ii) A fusion protein comprising the host protein and one or more Cas12k, TnsC, or TniQ, with linkers between each segment at the discretion of the user. A composition containing, or encoding, nucleic acid.

21. (i) a fusion protein comprising a transposon (Tn) protein, e.g., Tn7, Tn7-like, or Tn5053-like, e.g., transposition protein B (TnsB), fused to a protein (e.g., an endonuclease, e.g., nickase, cribase, or an endonuclease without catalytic activity, a fluorescent protein, or a peptide tag (e.g., NLS, His, Flag)) via an intervening linker of any choice; and (ii) A fusion protein comprising the host protein and one or more Cas12k, TnsC, or TniQ, with linkers between each segment at the discretion of the user. A composition containing, or encoding, nucleic acid.

22. The host factor is a ribosomal protein S15 that alters DNA topology (e.g., pi protein or nucleoid-associated protein (NAP), e.g., HU, Fis, H-NS, IHF, or TF1), or the host factor is involved in DNA or cellular metabolism, proteolysis or protein folding, regulation, or transport (e.g., acyl carrier protein (ACP), Sigma A protein expressed from the genes dcd, dinD, radA, recQ, clpX, fkpA, hflX, crl, rseB, rsxE, araJ, melB, mgtA, aspA, treC, proY, serA, yhbC, yidA, ykfA), an expression construct according to claim 7 or 8, a host cell according to any one of claims 9 to 11, the method according to any one of claims 14 to 17, a fusion protein according to claim 18 or 19, or the composition according to claim 20 or 21.

23. A host cell comprising or expressing a donor DNA molecule (e.g., a plasmid) containing the composition of claim 20 or 21, and the desired sequence to be inserted, wherein the desired sequence is flanked by LE and RE flanking sequences and a target site for the endonuclease (e.g., I-AniI) at its 5' and 3' ends, respectively, and preferably the target site is oriented to nick the donor plasmid 5' of the desired sequence to be inserted.

24. A host cell comprising the composition of claim 22 and a donor DNA molecule (e.g., plasmid) containing a desired sequence to be inserted, or expressing such a host cell, wherein the desired sequence is flanked by LE and RE flanking sequences and a target site for the endonuclease (e.g., I-AniI) at its 5' and 3' ends, respectively, and preferably the target site is oriented to nick the donor plasmid 5' of the desired sequence to be inserted.

Images