Reprogramming site-specific retrotransposon activity to new DNA sites

a) Schematic of computational pipeline used to discover and classify site-specific nLTR retrotransposon systems. b) Size distribution of the ORFs from the first methionine for each of the 5 families of RLE containing nLTR retrotransposons. c) Distribution of distances from candidate retrotransposons to detected Rfam annotation or tandem repeat targets for each of the 5 families of RLE containing nLTR retrotransposons. d) Distribution of the predicted 5′ and 3′ UTR sizes for all nLTR RLE-containing retrotransposons. UTR sizes are predicted based on the distance from the ORF and nearest predicted target site. Box plots are shown with the median, 25th percentile, 75th percentile, and whiskers that are 1.5x the interquartile range. All outliers are shown as individual points. n(5’UTR) = 10,033; n(3’UTR) = 7642. e) Distribution of the lengths of observed non-coding conservation regions flanking the 5′ and 3′ ends of the retrotransposon ORF. Box plots are shown with the median, 25th percentile, 75th percentile, and whiskers that are 1.5x the interquartile range. All outliers are shown as individual points. n(5’UTR) = 3307; n(3’ UTR) = 6472. f) Schematic of typical nLTR retrotransposon insertion sites with target sites consistent on both sides of the retrotransposon. g) Phylogenetic tree of all 5 families of RLE-containing nLTR systems showing majority of detected Rfam targets in the vicinity of the nLTR ORF.

a) DNA sequence alignments of nLTR families with divergent target preferences in the non-coding areas surrounding the nLTR ORFs. Identified Rfam annotations in the surrounding locus are highlighted. b) Multiple sequence alignment of different nLTR retrotransposons using MUSCLE, with Pfam ___domain schematic above as determined by HHpred. c) Analysis of sequence identity similarity of chosen nLTR retrotransposon family members using the MUSCLE protein alignment from Extended Data Fig. 2b.

a) Analysis of the 5′ end of the nLTR1Mbr locus with the microsatellite repeat region and alignment to the human 28S rDNA region highlighted. b) Schematic of payload homology and target sites used to evaluate nLTR1Mbr insertion. c) Gluc payload insertion by nLTR1Mbr into a panel of luciferase reporters, as quantified by luciferase production, with R2Tg targeting the R2 28S sequence as control. Reporters with either similarity to the R2 28S region, or with similarity to the 28S homology region in the nLTR1Mbr locus are used for evaluation of alternative insertion sites. d) Phylogenetic tree of nLTR retrotransposons zoomed in on the R2Tocc system and surrounding orthologs. Tree branches corresponding to avian genomes are highlighted in blue and orthologs used in this study are labeled. e) Heatmap of 28S luciferase reporter assay, testing integration by R2Bm, R2Tg, R2Mes and R2Tg^RTmut (x axis) using RNA payloads containing UTRs from different retrotransposon ortholog systems (y axis). f) Validation of 28S NGS assessment of editing efficiencies. Synthetic eblocks containing editing and unedited DNA sequences were mixed at defined ratios (x-axis) and measured by NGS (y-axis). Agreement between the known editing percentage (x-axis) and measured editing percentage was calculated by linear regression and is shown as an inset. Schematic above shows the relationship of the three NGS primers to the inserted sequence where one forward primer is in the genomic sequence upstream and there is one reverse primer in the insert and one reverse primer in the downstream genomic sequence. g) Timecourse of biochemical TPRT by R2Tg into 28S DNA with or without RNA payloads with different incubation times, as indicated. h) NGS insertion quantification of TPRT shown in Extended Data Fig. 3g. i) Electrophoretic Mobility Shift Assay gel showing the shift of the 28S DNA target due to binding of the R2Tg protein alone or R2Tg-RNA complex. Schematics to the right of the gel show the identity of the different DNA complex products on the gel. The bottom strand is 5′ labeled. j) Biochemical insertion of Gluc sequence into the 28S target with a payload containing only homology arms to the 28S locus and no UTRs with +/–payload RNA, +/– 28S DNA, +/– R2 protein, +/– Mg²⁺ and +/– dNTPs, as indicated. Above, NGS quantitation of insertion efficiency and schematic of the used RNA payloads. Schematics to the right of the gels indicate the specific TPRT and cleavage products. k) Biochemical TPRT by R2Tg into 28S DNA using RNA payloads with 100 bp, 60 bp, 30 bp and 0 bp 28S homology and no RNA payload control. Insertion frequency is quantified by NGS. l) Biochemical TPRT by R2Tg into 28S DNA using RNA payloads with or without 5′ cap and/or 3′ poly-A tail modifications as well as no RNA payload control. m) NGS insertion quantification of TPRT shown in Extended Data Fig. 3l. n) Primer extension assay by WT R2Tg, RLE inactivated R2Tg^D1275A, and no protein, where 28S RNA payload and complementary primer were hybridized and extended by reverse transcription activity of the R2Tg protein. Error bars represent mean +/− (c, e) s.e.m. or (j) s.d. n = 3 (c, e, f, j) or n = 1 (h, k, m) where n represents biological replicates.

a) Gluc payload insertion by R2Tg reverse transcriptase ___domain deletions, RLE inactivation mutants (D1275A), and reverse transcriptase mutations (R2Tg^{F876A/A877L/D878A/D879A/L880A/V881A/L882A}, RTmut), at the 28S locus luciferase reporter target, as quantified by luciferase activity. Luciferase activity was assayed in HEK293FT cells. b) Gluc payload insertion by R2Tg RT ___domain mutations, including R2Tg^{F876A/A877L/D878A/D879A/L880A/V881A/L882A} (RTmut), R2Tg^D878R/D879R, and R2Tg^D878H/D879H, and the RLE inactivation mutant (D1275A) at the 28S locus luciferase reporter, as quantified by luciferase. Luciferase activity was assayed in HEK293FT cells. c) Uncropped version of the gels shown in Fig. 2g; Above, RNA payload insertion into a 28S plasmid reporter by wild type R2Tg, RLE inactivated, RT inactivated, and complemented RT and RLE inactivated proteins +/– RNA payload, as indicated. RNA templates used were in vitro transcribed with a 5′ cap and a poly A tail. Expected band size = 294 bp. NT, non-targeting RNA templates that have homology to the NOLC1 target instead of the 28S locus. Below, R2Tg insertion into human 28S endogenous locus with payloads containing 100, 50, 30 or 0 homology to the 28S target site. RNA templates used were in vitro transcribed with a 5′ cap and a poly A tail. Expected band size = 374 bp. d) Luciferase assay of Gluc insertion of an IVT RNA payload with variable 3′ tail length into a 28S reporter target by WT R2Tg and RLE-inactivated R2Tg^D1275A. Luciferase activity was assayed in Huh-7 cells. e) Sanger sequencing of 5′ and 3′ insertion junctions at the 28S target for additional selected payload designs after R2Tg integration. Payload numbers correspond to those in Fig. 2h. f) Example indels at the 5′ junction for R2Tg insertion at the 28S target for selected payloads. Non-templated Cs from reverse transcription in the bottom strand (G in the top strand) are highlighted with red boxes. g) Example indels at the WT 28S locus target for selected payloads. Non-templated Cs from reverse transcription in the bottom strand (G in the top strand) are highlighted with red boxes. Error bars represent mean +/− (a,b,d) s.e.m. n = 3 where n represents biological replicates.

a) Gaussia luciferase exon 2 (Gluc) payload insertion by wild type and ___domain inactivated mutants of R2Tg into a 28S plasmid reporter, with editing outcomes profiled by NGS at the upstream (left) junction. Mutants tested are WT R2Tg and R2Tg^D1275A (RLE mutant) and outcomes are classified as perfect insertions, insertions with indels, or WT locus indels. b) Schematic of additional payload variant with internal homology arms against the 28S target. c) Gaussia luciferase exon 2 (Gluc) payload insertion by wild type R2Tg into a 28S plasmid reporter with payload variants shown in part B, with editing outcomes profiled by NGS at the upstream (left) junction. Outcomes are classified as perfect insertions, insertions with indels, or WT locus indels. d) Size analysis by gel electrophoresis of 5′ and 3′ insertion junctions at the 28S target reporter for payload designs from part (b) and (c) after R2Tg integration. Payload numbers correspond to those in B. e) Gluc exon 2 payload insertion by WT R2Tg, R2Tg^D1275A, or the RT ___domain deletion R2Tg^Δ(875-885) into a 28S plasmid reporter with payloads containing 28S or AAVS1 targeting homology arms, profiled by NGS. Statistics were calculated using unpaired t-test. f) Biochemical retrotransposition of different RNA payloads into the AAVS1 DNA target with the R2Tg protein, dNTPs, and MgCl₂, as indicated. Either no payload was used or the following two payloads were used: 1) payload with a 5′ UTR targeting AAVS1 and containing a Gluc insert, or 2) a payload with 5′ and 3′ UTRs targeting NOLC1 and containing an EGFP insert. Labels on the gel indicate the specific TPRT product, DNA target band, and R2Tg produced nicked fragments. g) Validation of AAVS1 NGS method. Synthetic eblocks containing editing and unedited DNA sequences were mixed at defined ratios (x-axis) and measured by NGS (y-axis). Agreement between the known editing percentage (x-axis) and measured editing percentage was calculated by linear regression and is shown inset. h) Validation of the NOLC1 3-primer NGS assay using mixes of genomic DNA from unedited or heterozygously inserted cells at NOLC1, as measured by ddPCR. Shown is the known pre-mixed ratio of edited and unedited gDNA (x-axis) vs the measured editing rate by NGS (y-axis). Inset, coefficient of determination between values on x- and y-axes. i) Schematic of payload engineering for R2Tg reprogramming to the NOLC1 locus. j) EGFP payload insertion at human endogenous NOLC1 locus by natural reprogrammed wild-type R2Tg as well as R2Tg^D1275A and R2Tg^RTmut. Insertion is quantified by ddPCR. Statistics calculated with unpaired t-test. k) Payload insertion by SpCas9^H840A-R2Tg^Δ1-183 or SpCas9^H840A-R2Tg^{Δ1-183,D1275A} into the endogenous NOLC1 locus, mediated by dual guides or non-targeting guides and quantified by ddPCR. Inset shows payload design and locus schematic with homology arms colored and top guide in red and bottom guide in blue. Statistics calculated with unpaired t-test. l) Secondary structure analysis of the 5′ UTR of R2Tg, including the full length, 15 nt truncated variant, and the 15 nt truncated variant with the 50 nt 28S homology sequence upstream. m) Validation of the 3-primer NGS assay for analysis of AAVS1 integration via the left insertion junction. Standards consist of edited and WT amplicons that are mixed in the listed ratios (x-axis) and the measured editing is determined by the 3-primer NGS assay (y-axis). n) Gluc integration at the endogenous AAVS1 locus via the SpCas9^H840A-R2Tg^Δ1-183 fusion using payloads with the full length or 15-nt truncated 5′ UTR, an upstream 28S 50 nt sequence, and internal AAVS1 homology arms. Integration is quantified by next-generation sequencing (left) and ddPCR (right). Error bars represent mean +/− s.e.m. n = 3 where n represents three biological replicates.

a) Biochemical retrotransposition of an RNA payload into the NOLC1 DNA target with and without withdrawal of R2Tg protein, RNA, dNTPs, SpCas9/guides, or MgCl₂, as indicated. Above, NGS quantification of insertion efficiency and a schematic of the RNA payload used. Gel is stained with SYBR gold for visualization of nucleic acid. b) Biochemical retrotransposition of an RNA payload into the NOLC1 DNA target with and without withdrawal of R2Tg protein, RNA, dNTPs, SpCas9/guides, or MgCl₂, as indicated. The DNA top strand is Cy5 labeled (red) and bottom strand is FAM labeled (green), allowing for visualization by fluorescence. Labels on the gel indicate the specific TPRT product, DNA target band, and R2Tg produced nicked fragments. c) Reprogrammed biochemical retrotransposition by R2Tg into the NOLC1 DNA target, using a homologous IVT NOLC1 payload (N) with +/– 5′ cap and 3′ tail modifications compared to EMX1 (E)- or 28-homologous (28S) payloads (i.e. non-homologous to NOLC1). Integration is quantified by NGS. d) Reprogrammed biochemical retrotransposition of an IVT RNA payload containing the optimized 5′ and 3′ UTR and homology regions into the AAVS1 DNA target by R2Tg +/– DNA target, +/– RNA, +/– Cas9-assisted nicking, and +/– R2Tg, as indicated. Black arrow on the gel indicates the specific TPRT product. The blue arrow denotes the cleaved DNA band generated by R2Tg protein alone reprogrammed by its payload RNA. e) NGS quantification of insertion data shown in Extended Data Fig. 6d. f) Integration efficiencies, quantified by NGS, of reprogrammed biochemical TPRT of an RNA payload by R2Tg into varying amounts of NOLC1 DNA target compared to no RNA controls. g) Integration efficiencies, quantified by NGS, of reprogrammed biochemical TPRT by R2Tg using NOLC1 RNA payloads incorporating either different single-base mismatches or insertions into the NOLC1 DNA, as indicated. Either in vitro transcribed mRNA or synthetic RNA templates were used as the payloads. h) Biochemical retrotransposition of an RNA payload into the AAVS1 DNA target with and without withdrawal of R2Tg protein, RNA, dNTPs, or MgCl₂, as indicated. Above, NGS quantification of insertion efficiency and a schematic of the RNA payload used. Labels on the gel indicate the specific TPRT product, DNA target band, and R2Tg produced nicked fragments. i) Schematic of DNA cleavage end detection using ligation and NGS. Ligation adaptor primers (shown in black) are used in combination with anchored primers on either the left (red) or right end (blue) are used to read out the variable R2Tg cleavage sites. j) Cleavage end detection by next-generation sequencing of the R2Tg generated nicks on the AAVS1 target from Extended Data Fig. 6h in the condition without dNTPs. The color of the reads for 5′ or 3′ ends match the anchored primers shown in the schematic in Extended Data Fig. 6i. Below the plot is a schematic of the AAVS1 target (black) and the homology arms of the payload template (beige and gray). k) Cleavage end detection by next-generation sequencing of the R2Tg generated nicks on the AAVS1 target from Extended Data Fig. 6h in the condition without the RNA template. The color of the reads for 5′ or 3′ ends match the anchored primers shown in the schematic in Extended Data Fig. 6i. Below the plot is a schematic of the AAVS1 target (black) and the homology arms of the payload template (beige and gray). l) Biochemical retrotransposition of an RNA payload into the NOLC DNA target with and without withdrawal of R2Tg protein, RNA, dNTPs, or MgCl₂, as indicated. Labels on the gel indicate the specific TPRT product, DNA target band, and R2Tg produced nicked fragments. m) Cleavage end detection by next-generation sequencing of the R2Tg generated nicks on the NOLC1 target from Extended Data Fig. 6l in the condition without dNTPs but with RNA template. The color of the reads for 5′ or 3′ ends match the anchored primers shown in the schematic inset. Below each plot is a schematic of the NOLC1 target (black) and the homology arms of the payload template (beige and gray). n) Cleavage end detection by next-generation sequencing of the R2Tg generated nicks on the NOLC1 target from Extended Data Fig. 6l in the condition without RNA template. The color of the reads for 5′ or 3′ ends match the anchored primers shown in the schematic inset. Below each plot is a schematic of the NOLC1 target (black) and the homology arms of the payload template (beige and gray). Error bars represent mean +/− s.e.m. n = 3 where n represents three biological replicates.

a) Schematic of SpCas9^H840A fused to N- and C-terminal truncations of R2Tg at different amino acid positions. Not all tested constructs are shown. b) Gluc payload insertion by different SpCas9^H840A-R2Tg^Δ1-183 fusions, according to the schematic in (a), into the endogenous AAVS1 locus quantified by NGS. N-term and C-term denote either N-terminal or C-terminal fusions of the full length R2Tg protein. Denoted residue positions indicate the starting amino acid position of N-terminal R2Tg truncations that are fused to the C-terminal of SpCas9^H840A. c) Gluc integration at the endogenous AAVS1 target by SpCas9^H840A-R2Tg^Δ1-183, SpCas9^H840A-R2Tg^{Δ1-183,F876A/A877L/D878A/D879A/L880A/V881A/L882A} (RTmut), and SpCas9^H840A-R2Tg^{Δ1-183,Δ(875-885)}, and SpCas9^H840A alone. Editing rates were quantified by NGS (left) and ddPCR (right). d) TPRT activity in HEK293FT cells with SpCas9^H840A alone or fused to R2Tg, R2Tg^{Δ1-183,F876A/A877L/D878A/D879A/L880A/V881A/L882A} (RTmut), or R2Tg^{Δ1-183,Δ875-885} into the NOLC1 genomic target with dual guides. EGFP payload contains the full 5′ and 3′ UTRs for R2Tg. e) Gluc payload insertion into a 28S plasmid reporter in HEK293FT cells by selected nLTR retrotransposons fused to SpCas9^H840A, with either targeting or non-targeting guides, quantified by Gluc production normalized to a control Cluc. Data is shown as ratio of targeting signal to non-targeting signal. f) Gluc payload insertion into the endogenous AAVS1 locus in HEK293FT cells by selected nLTR retrotransposons fused to SpCas9^H840A, with either targeting or non-targeting guides, profiled by next generation sequencing. Editing outcomes are quantified as perfect insertions, insertions with indels, and indels at the unmodified WT target site. g) Gluc payload insertion into the endogenous AAVS1 locus in HEK293FT cells by selected nLTR retrotransposons fused to so SpCas9^H840A and an AAVS1-targeting or non-targeting sgRNA control, quantified by ddPCR h) Validation of the AAVS1 3-primer NGS assay using mixes of genomic DNA from unedited or heterozygously inserted cells at AAVS1, as measured by ddPCR. Shown is the known pre-mixed ratio of edited and unedited gDNA (x-axis) vs the measured editing rate by NGS (y-axis). Inset, coefficient of determination between values on x- and y-axes. i) R2Tocc retrotransposition of a synthetic RNA payload into top- and bottom-strand labeled 28S DNA. The top strand is FAM labeled (red); the bottom strand is Cy5 labeled (green). Schematics on the side of the gel indicate the expected size of each band, including the TPRT product and cleaved target fragments. j) Indels or substitutions found in the sequencing reads of AAVS1 in vitro TPRT shown in Fig. 5b, analyzed by CRISPResso. Above, a reference sequence consisting of correct insertion of the Gluc payload into AAVS1 DNA. Below, a schematic of the different insertion outcomes found by sequencing, the raw number of reads and % of total reads which these correspond to. Error bars represent mean +/− (b,g) s.d. or (d, e, f, h) s.d. n = 3 where n represents 3 biological replicates.

a) Expression of wild type and mutant R2Tg orthologs (x-axis), quantified by luciferase signal. b) Schematic of STITCHR insertion using intron-containing templates in the following subpanels. An EGFP STITCHR payload containing an interrupting intron is expressed by a CAG promoter. After RNA splicing and TPRT, it is inserted into the genome as an uninterrupted EGFP ORF. Shown are the landing sites of the NGS primers used in the subsequent panels. Shown are the GFP cargo (green bar, approximately 500 bp), interrupting intron (USF1, 245 bp, tetrahymena self-splicing intron, 399 bp), homology sequences (yellow bar, 50 bp each), poly-A tail, genomic sequence (grey bar), external F and R NGS primers (black) and internal reverse primer (blue). c) NGS evaluation of insertion at AAVS1 (left) and EMX1 (right) loci after delivering a plasmid template containing GFP with an interrupting self-splicing tetrahymena intron. Shown is the % insertion of GFP lacking the interrupting intron (i.e. spliced insertion) by SpCas9^H840A-R2Tocc^Δ1-169 or SpCas9^H840A. Insertions are quantified as perfect insertions or insertions with indels. d) ddPCR evaluation of AAVS1 insertion after delivering a plasmid template containing an interrupting USF1 intron, which interrupts in two locations in the payload. The ddPCR assay used detects spliced insertion only. Shown is the % spliced insertion by SpCas9^H840A-R2Tocc^Δ1-169, SpCas9^H840A-R2Tg^{Δ1-183,RTmut}, SpCas9^H840A-R2Tg^{Δ1-183,RLEmut}, SpCas9^H840A-R2Tg^{Δ1-183,RTmut} and SpCas9^H840A control. e) Gluc reconstitution by correction of a 20 bp deletion by delivering plasmid or synthetic RNA payloads, quantified by Gluc expression normalized to control Cluc. The synthetic RNA template is an extension of the Cas9 sgRNA. f-g) Gluc reconstitution by R2Tg mutants with synthetic RNA payloads extended off the guide RNA as quantified by NGS (f) and Gluc (g) expression normalized to control Cluc. h) STITCHR 20 bp payload insertion on a luciferase reporter plasmid from a synthetic RNA lacking the 5′ UTR and containing a Cas9 guide scaffold (fused) or a synthetic RNA delivered in trans containing the 5′ UTR and the correction sequence (trans). Editing is with or without a Cas9 nicking guide that allows for initiation of TPRT for the trans template. Integration is quantified by NGS and is represented as perfect insertions or insertions with indels. WT indels are also shown which are defined as indels at the unintegrated Gluc locus. i) STITCHR 22 bp payload insertion on an EGFP reporter plasmid from a synthetic RNA lacking the 5′ UTR and containing a Cas9 guide scaffold (fused) or a synthetic RNA delivered in trans containing the 5′ UTR and the correction sequence (trans). Editing is with or without a Cas9 nicking guide that allows for initiation of TPRT for the trans template. Integration is quantified by NGS and is represented as perfect insertions or insertions with indels. WT indels are also shown which are defined as indels at the unintegrated plasmid reporter. j) STITCHR 20 bp payload insertion on a luciferase reporter plasmid from a synthetic RNA lacking the 5′ UTR and containing a Cas9 guide scaffold (fused), a synthetic RNA delivered in trans containing the 5′ UTR and the correction sequence (trans with UTR), and a synthetic RNA delivered in trans containing the correction sequence without a UTR (trans without UTR). SpCas9^H840A-R2Tg^Δ1-183 and SpCas9^H840A-R2Tg^{Δ1-183,RTmut} are compared to each other and editing is performed +/− a Cas9 nicking guide that allows for initiation of TPRT for the trans template. Integration is quantified by NGS. k) STITCHR 38 bp payload insertion at the endogenous LMNB1 locus from a synthetic RNA lacking the 5′ UTR and containing a Cas9 guide scaffold. Integration is quantified by NGS. l) STITCHR 700 bp EGFP payload insertion in Huh-7 cells at the endogenous NOLC1 locus from an in vitro transcribed mRNA, insertion is quantified by ddPCR. m) Indels or substitutions found in the sequencing reads of NOLC1 insertion experiment shown in Extended Data Fig. 8i, analyzed by CRISPResso. Above, a reference sequence consisting of correct insertion of the Gluc payload into AAVS1 DNA. Below, a schematic of the different insertion outcomes found by sequencing, the raw number of reads and % of total reads which these correspond to. n) Insertion of a GFP payload delivered as an IVT mRNA without UTRs into the human endogenous NOLC1 locus by SpCas9^H840A-R2Tocc^Δ1-169 in Huh-7 cells. Insertion is quantified by NGS and is represented as perfect insertions or insertions with indels. WT indels are also shown which is defined as indels at the unintegrated NOLC1 locus. o) Insertion of a GFP payload delivered as an IVT mRNA with UTRs and other variable modifications, as indicated, into the human endogenous NOLC1 locus by SpCas9^H840A-R2Tocc^Δ1-169 in HEK293FT cells. Insertion is quantified by NGS. p) EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous NOLC1 locus, with combinations of single and dual guides, compared to a non-targeting guide control and quantified by NGS. q) EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous LMNB1 locus, with combinations of single and dual guides, compared to a non-targeting guide control and SpCas9^H840A alone. Editing was quantified by digital droplet PCR (ddPCR). r) EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous NOLC1 locus, with combinations of single and dual guides, compared to a non-targeting guide control and profiled by ddPCR. s) EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous EMX1 locus, with combinations of single and dual guides, compared to a non-targeting guide control and SpCas9^H840A alone. Editing quantified by digital droplet PCR (ddPCR). t) EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous AAVS1 locus, with combinations of single and dual guides, compared to SpCas9^H840A alone, SpCas9^H840A-R2Tocc^{Δ1-169,RTmut}, and SpCas9. Insertion is quantified by ddPCR. u) Comparison of ddPCR and NGS quantification of EGFP payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous AAVS1 locus, with combinations of single and dual guides. Error bars represent mean +/− (a, f, g, h, i, l, n, o, p, q, s, t, u) s.e.m or (c, d, e, j, k) s.d. n = 3 where n represents 3 biological replicates.

a) Gluc payload insertion by SpCas9^H840A-R2Tocc^Δ1-169 (WT), SpCas9^H840A-R2Tocc^{Δ1-169,F811A,A812L,D813A,D814A,L815A,V816A,L817A} (RTmut), SpCas9^H840A-R2Tocc^{Δ1-169,Δ(811-814)}, SpCas9^H840A-R2Tocc^{Δ1-169,Δ(810-820)}, and SpCas9^H840A at AAVS1. Editing quantified by NGS. b) EGFP insertion by SpCas9^H840A-R2Tocc^Δ1-169 (WT), SpCas9^H840A-R2Tocc^{Δ1-169,F811A,A812L,D813A,D814A,L815A,V816A,L817A} (RTmut), SpCas9^H840A-R2Tocc^{Δ1-169,Δ(876-879)}, SpCas9^H840A-R2Tocc^{Δ1-169,Δ(875-885)}, and SpCas9^H840A at NOLC1. Editing is quantified by ddPCR. c) GFP insertion by SpCas9^H840A-R2Tocc^Δ1-169 (WT), SpCas9^H840A-R2Tocc^{Δ1-169,RLEmut}, and SpCas9^H840A at the endogenous NOLC1 target site. Editing quantified by ddPCR. d) EGFP payload insertion by SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous NOLC1 locus, using payloads with 50 nt homology arms targeting NOLC1 or AAVS1 targets, or without homology. Payloads are evaluated with single, dual, or non-targeting guides and are compared to SpCas9^H840A. Editing quantified by ddPCR. N = NOLC1 target. A = AAVS1 target. e) EGFP insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous NOLC1 locus, with payloads with varying homology arm lengths. Payloads are evaluated with dual or non-targeting guides and are compared to SpCas9^H840A. Editing quantified by ddPCR. f) GFP insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous NOLC1 locus in HepG2 cells, compared to SpCas9^H840A. Editing quantified by ddPCR. g) STITCHR EGFP insertion at endogenous EMX1, NOLC1 and two AAVS1 loci in Huh-7 cells by SpCas9^H840A-R2Tocc^Δ1-169 compared to SpCas9^H840A-R2Tocc^{Δ1-169,RTmut}. Insertion quantified by ddPCR. h) STITCHR EGFP insertion at endogenous EMX1 and NOLC1 loci in HepG2 cells by SpCas9^H840A-R2Tocc^Δ1-169 compared to SpCas9^H840A-R2Tocc^{Δ1-169,RTmut}. Insertion quantified by ddPCR. i) EGFP insertion at endogenous NOLC1 by STITCHR, delivered by different adenovirus amounts to HEK293FT cells. Shown is a comparison of insertion efficiency when delivering STITCHR machinery with one vector and guides and template with the other, compared to delivery of guides and template only as a control. Editing quantified by NGS. j) EGFP insertion by SpCas9^H840A-R2Tocc^Δ1-169 at NOLC1 in quiescent primary human hepatocyte cells compared to SpCas9^H840A control. 1.4e11 viral copies was used in the dual vector condition; half of that for the single vector payload only condition. Editing quantified by NGS. k) EGFP payload insertion by STITCHR at the NOLC1 endogenous locus in HEK293FT cells, comparing editing efficiencies with and without PAM elimination. Editing quantified by ddPCR. l) STITCHR EGFP insertion at the endogenous NOLC1 locus by SpCas9^H840A-R2Tocc^Δ1-169 and SpCas9^D10A-R2Tocc^Δ1-169. Editing quantified by ddPCR. m) PacBio sequencing of a 700 bp EGFP insertion at the endogenous NOLC1 locus by SpCas9^H840A-R2Tocc^Δ1-169. Reads are aligned to the expected reference sequences of scarless NOLC1 insertion. Gray bases indicate a match to the reference sequence; red or black indicate mismatched. n) PacBio sequencing of a 280 bp Gluc payload insertion at the endogenous AAVS1 locus by SpCas9^H840A-R2Tocc^Δ1-169. Reads are aligned to the corresponding expected reference sequences of scarless AAVS1 inseriton. Gray bases indicate a match to the reference sequence; red or black indicate mismatched. o) Additional analysis of the PacBio long read sequencing for complete, incomplete, and concatemeric insertions at the respective sites. p) Schematic of the cross-junction ddPCR assay. Primers amplify across the central junction of a hypothetical concatemeric GFP insertion. Above, a single GFP insert in which the primers are facing in opposite directions and will not amplify. Below, a hypothetical concatemeric insertion in which the primers are facing each other across the concatemer junction, producing amplification. q) Cross-junction ddPCR readout of concatemers generated by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 and a 700 bp GFP payload into the endogenous NOLC1 locus, benchmarked against synthetic standards r) Cross-junction ddPCR of genomic DNA standards generated from mixtures of genomic DNA from a heterozygous clone containing a 2x concatemeric GFP insertion at NOLC1 with gDNA from WT cells. gDNA was mixed at ratios corresponding to editing efficiencies ranging from 0.01% to 1% editing. Shown is the percentage of gDNA mixing (editing percentage) versus editing detected by cross-junctional ddPCR (measured editing). s) Schematic of ddPCR assay used to assess copy number of insertions. Above, 4 possible insertion outcomes of GFP insertion at AAVS1 by STITCHR: a single insert, tail-to-head and two tail-to-tail concatemeric insertions. Other outcomes are possible that are not depicted (e.g. head-to-head, partial concatemers, >2x concatemers) but would still be detected by the ddPCR design. Shown are primers (black) and the probe (pink box) used in the assay, plus the site of the restriction enzyme Xho1 which separates any concatemers (below), detected as increasing positive droplet concentration. t) CNV ddPCR assay depicted in s), of 10 HEK293FT clones containing a monoallelic, scarless STITCHR insertion of GFP at AAVS1 (indicated with a dotted line), a HEK293FT clone (22n115) containing a tail-to-head 2x GFP insertion at NOLC1 and a negative control (22n22) containing no insertion. Each sample was assayed +/− Xho1 digestion. * = p < 0.05, statistics calculated with unpaired t-test. u) Design of two Southern blots detecting STITCHR inserts at AAVS1 and their expected outcomes. Shown are two designs: an internal probe (left) which hybridizes to the GFP insertion and an external probe (right) which hybridizes outside the insert. For both, shown are 3 possible editing outcomes and their expected sizes: a 2x monoallelic insertion, a 1x monoallelic insertion and no insertion. Other outcomes are possible that are not depicted but will alter the expected band sizes (e.g. 3x insertion, insertion with unexpected insertions/deletions). Shown are the restriction enzyme cut sites, the site where the probe (pink box) hybridizes and, right, the expected banding pattern for each depicted editing outcome. v) Southern blots of 10 HEK293FT clones containing a scarless GFP insert by STITCHR at AAVS1 and a negative clone (WT), utilizing an internal (above) or external (below) probe. Expected band sizes are indicated with a red (inserted) or black (uninserted) asterisk. Error bars represent mean +/− s.d. (a, c, f, g, j, k, r, t) or s.e.m. (b, e, h, i, l, q, r) n = 3 for panels (a-i, k-l, q, r, t), n = 2 (j), n = 1 (o) where n represents biological replicates.

a-b) Circos plots depicting genome-wide insertion sites of payloads by SpCas9^H840A-R2Tocc^Δ1-169 using sgRNAs and payload homologies to a) AAVS1 (chr19) and b) NOLC1 (chr10). Counts are defined as the number of mapped reads occurring within a 5 kb window. c) Schematic of STITCHR using SpCas9^H840A-R2Tocc^Δ1-169 to insert EGFP as a scarless in-frame fusion at the N-terminus of the human NOLC1 gene. The EGFP template is transcribed in a reverse complement manner to minimize background expression in the absence of insertion with 50 nt homology arms. d) STITCHR-mediated EGFP tagging of NOLC1, visualized by confocal microscopy, and compared to immunofluorescence staining of NOLC1. White scale bar denotes 10 µm. e) Therapeutically relevant payload insertion by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous AAVS1 locus, with sizes and identities of payload panel members shown and 100 nt homology arms. Integration is quantified by ddPCR and compared to SpCas9^H840A. For the NGS, shown are the number of total left junction inserts, left junction inserts containing indels and the WT locus containing indels. f) Evaluation of different sized edits using STITCHR at the NOLC1 locus using either SpCas9^H840A-R2Tocc^Δ1-169 or SpCas9^H840A. Inset shows payload design and locus schematic with homology arms colored and top guide in red and bottom guide in blue. g) PacBio HiFi long-read sequencing of a 12.7 kb insert at the endogenous NOLC1 locus by SpCas9^H840A-R2Tocc^Δ1-169 using primers that land externally on the 5’ side and within the insert on the 3’ side. Reads are aligned to the corresponding expected reference sequences of scarless insertion at the NOLC1 locus. Black bases indicate a match to the reference sequence; red indicate mismatched. h) Additional analysis of the PacBio long read sequencing for the 12.7 kb insert at NOLC1 from Extended Data Fig. 10g showing complete, incomplete, and concatemeric insertions at the respective sites. i) Short read sequencing of the right junction of the same sample containing a 12.7 kb insert at NOLC1 used in Extended Data Fig. 10g, showing complete insertion of the right junction. j) Installation of small edits and insertions using STITCHR at the NOLC1 locus, using a U6 promoter for payload expression. k) SpCas9-mediated HDR editing of the EMX1 gene in cells treated with varying concentrations of aphidicolin. Genome editing is quantified by NGS. l) EGFP payload insertion efficiencies at endogenous NOLC1 locus by homology-directed repair (HDR), using SpCas9, at different concentrations of the cell cycle inhibitor aphidicolin or DMSO control. m) EGFP payload insertion (50 nt homology arms) by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 into the endogenous AAVS1 locus in cells treated with cell cycling inhibitor Mirin or double thymidine. Integration is quantified by NGS and compared to SpCas9^H840A. n) SpCas9-mediated HDR editing of the EMX1 gene in cells treated with cell cycling inhibitor Mirin or double thymidine. Genome editing is quantified by NGS. o) Schematic of STITCHR-replace methodology involving replacement of a region of the genome while inserting the STITCHR payload. Top guide is shown in red and the bottom guide in blue. p) Evaluation of STITCHR-replace at the NOLC1 locus using a single guide and homology arms spaced 50–150 bp apart on the genome. R2Tocc^RTmut corresponds to the RT inactivation mutant: F811A/A812L/D813A/D814A/L815A/V816A/L817A. q) Example sequencing reads of the EGFP insertion site at NOLC1 for STITCHR replace, showing the desired 50–150 bp deletions. r) ddPCR quantification of multiplexed gene integration by STITCHR with SpCas9^H840A-R2Tocc^Δ1-169 at NOLC1 and AAVS1 sites. EGFP payload insertion at NOLC1 is quantified by ddPCR, and Gluc insertion at AAVS1 is quantified by NGS. Targeting conditions are compared to non-targeting guide controls. Error bars represent mean +/− s.e.m (d, e, k, l, m, n) or s.d. (f, j, p, r). n = 3 where n represents 3 biological replicates.