Locus-resolution analysis of L1 regulation and retrotransposition potential in mouse embryonic development

Locus-resolution analysis of L1 regulation and retrotransposition potential in mouse embryonic development Patricia Gerdes, 1 Dorothy Chan, 1 Mischa Lundberg, 1,2,3 Francisco J. Sanchez-Luque, 1,4,5 Gabriela O. Bodea, 1,6 Adam D. Ewing, 1 Geoffrey J. Faulkner, 1,6 and Sandra R. Richardson 1 1 Mater Research Institute - University of Queensland, TRI Building, Woolloongabba, Queensland 4102, Australia; 2 The University of Queensland Diamantina Institute, The University of Queensland, Woolloongabba, Queensland 4102, Australia; 3 Translational Bioinformatics, Commonwealth Scientific and Industrial Research Organisation, Sydney, New South Wales 2113, Australia; 4 GENYO. Centre for Genomics and Oncological Research (Pfizer-University of Granada-Andalusian Regional Government), PTS Granada, 18016, Spain; 5 MRC Human Genetics Unit, Institute of Genetics and Cancer (IGC), University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, United Kingdom; 6 Queensland Brain Institute, University of Queensland, Brisbane, Queensland 4072, Australia Mice harbor ∼2800 intact copies of the retrotransposon Long Interspersed Element 1 (L1). The in vivo retrotransposition capacity of an L1 copy is defined by both its sequence integrity and epigenetic status, including DNA methylation of the monomeric units constituting young mouse L1 promoters. Locus-specific L1 methylation dynamics during development may therefore elucidate and explain spatiotemporal niches of endogenous retrotransposition but remain unresolved. Here, we interrogate the retrotransposition efficiency and epigenetic fate of source (donor) L1s, identified as mobile in vivo. We show that promoter monomer loss consistently attenuates the relative retrotransposition potential of their offspring (daughter) L1 insertions. We also observe that most donor/daughter L1 pairs are efficiently methylated upon differentiation in vivo and in vitro. We use Oxford Nanopore Technologies (ONT) long-read sequencing to resolve L1 methylation genome-wide and at individual L1 loci, revealing a distinctive “smile”pattern in methylation levels across the L1 promoter region. Using Pacific Biosciences (PacBio) SMRT sequencing of L1 5′RACE products, we then examine DNA methylation dynamics at the mouse L1 promoter in parallel with transcription start site (TSS) distribution at locus-specific resolution. Together, our results offer a novel perspective on the interplay between epigenetic repression, L1 evolution, and genome stability. [Supplemental material is available for this article.] Retrotransposons are major contributors to ongoing mutagenesis in mammalian genomes. The autonomous non-long terminal repeat (non-LTR) retrotransposon Long Interspersed Element 1 (LINE-1 or L1) is actively mobilizing in both humans and mice, and L1 sequences occupy ∼17% of human DNA and ∼18% of mouse DNA (International Human Genome Sequencing Consortium 2001; Waterston et al. 2002). While humans contain a single active L1 subfamily, the mouse genome harbors three active L1 subfamilies, termed T F ,G F , and A (Wincker et al. 1987; Schichman et al. 1993; Casavant and Hardies 1994; DeBerardinis et al. 1998; Naas et al. 1998; Saxton and Martin 1998; Hardies et al. 2000; Goodier et al. 2001; Mears and Hutchison 2001) which are each further divided into several sublineages, for example, T FI ,T FII , and T FIII (Sookdeo et al. 2013). Although T F and G F elements descended from the old and now inactive F subfamilies, A elements evolved independently. However, generating a single phylogenetic tree describing the relation of the subfamilies with each other is difficult because of frequent recombination among elements (Sookdeo et al. 2013). The vast majority of the ∼600,000 L1 copies in the mouse genome reference are 5′truncated and mutated (Voliva et al. 1983; Waterston et al. 2002), leaving approximately 2800 full-length L1s (Penzkofer et al. 2017). Ongoing L1 activity has generated substantial variation in L1 content among inbred strains, as well as interindividual variation in L1 content within strains (Akagi et al. 2008; Nellåker et al. 2012; Richardson et al. 2017; Schauer et al. 2018; Gerdes et al. 2022; Ferraj et al. 2023). L1 insertions also are responsible for several spontaneous mouse mutants, driven by L1 T F elements in all cases in which the L1 subfamily can be identified (Gagnier et al. 2019). A full-length mouse L1 is ∼6–7 kb long and begins with a 5′untranslated region (5′UTR) containing an internal RNA polymerase II promoter (Severynse et al. 1991; DeBerardinis and Kazazian 1999). The mouse L1 5′UTR has a distinctive structure, wherein a variable number of tandemly repeated ∼200 bp monomer units are situated upstream of a nonmonomeric region (Adey et al. 1994; Kong et al. 2022). Each monomer contributes Corresponding authors: [email protected]m, [email protected] Article published online before print. Article, supplemental material, and publication date are at https://www.genome.org/cgi/doi/10.1101/gr.278003.123. Freely available online through the Genome Research Open Access option. © 2023 Gerdes et al. This article, published in Genome Research, is available under a Creative Commons License (Attribution 4.0 International), as described at http://creativecommons.org/licenses/by/4.0/. Research 33:1465–1481 Published by Cold Spring Harbor Laboratory Press; ISSN 1088-9051/23; www.genome.org Genome Research 1465 www.genome.org additive promoter activity (DeBerardinis and Kazazian 1999). Individual monomers of young L1 subfamilies generally comprise sufficient CpG dinucleotides to qualify as CpG islands (Lee et al. 2010), and also contain several transcription factor binding sites, including for YY1 transcription factor (YY1) (DeBerardinis and Kazazian 1999; Lee et al. 2010). The YY1 binding site is required for accurate L1 transcription initiation (Athanikar et al. 2004; Lee et al. 2010), and an intact YY1 binding site also is important for methylation of the human L1 promoter during cellular differentiation (Sanchez-Luque et al. 2019). The L1 5′UTR is followed by two open reading frames encoding the proteins ORF1p and ORF2p, and a 3′UTR incorporating a polyadenylation signal (Scott et al. 1987; Skowronski et al. 1988; Dombroski et al. 1991). ORF1p is ∼40 kD and harbors RNA binding and chaperone activities (Holmes et al. 1992; Hohjoh and Singer 1996; Martin and Bushman 2001; Khazina and Weichenrieder 2009, 2018) whereas the ∼150 kD ORF2p has showed endonuclease (EN) and reverse transcriptase (RT) activities (Mathias et al. 1991; Feng et al. 1996; Ergün et al. 2004; Doucet et al. 2010; Taylor et al. 2013). Both proteins are required for L1 mobilization through reverse transcription of an RNA intermediate in a process termed target-site primed reverse transcription (TPRT) (Scott et al. 1987; Holmes et al. 1992; Luan et al. 1993; Feng et al. 1996; Moran et al. 1996). While L1 retrotransposition can occur in nondividing cells (Kubo et al. 2006; Macia et al. 2017), a growing body of evidence has emerged linking L1 retrotransposition to DNA replication during the S phase of the cell cycle (Mita et al. 2018, 2020; Flasch et al. 2019). Hallmarks of L1 integration by TPRT include flanking target site duplications (TSDs), and the incorporation of a3 ′poly(A) tract which reflects the necessity of L1 mRNA polyadenylation forefficient retrotransposition (Grimaldi et al. 1984; Doucet et al. 2015). Unchecked retrotransposition presents a threat to genome stability, and is countered by a variety of host defense mechanisms (Bourc’his and Bestor 2004; Goodier 2016; MacLennan et al. 2017; Liu et al. 2018; Deniz et al. 2019; Greenberg and Bourc’his 2019; Mita et al. 2020; Tristán-Ramos et al. 2020; Senft and Macfarlan 2021).While repressivehistonemarks suchasH3K9me3 playamajor role in silencing older mouse L1 subfamilies (Tan et al. 2013; Castro-Diazetal.2014;Jacobsetal.2014),youngerL1saretypically silencedbymethylationoftheCpG islandsintheirpromoters(Furanoetal.1988;HataandSakaki1997;Leeetal.2010;dela Ricaetal. 2016; Gerdes et al. 2022). During embryonic development, the epigenome undergoes reprogramming including phases of global DNA demethylation (Hajkova et al. 2002; Seki et al. 2005; Abe et al. 2011; Saitou et al. 2012; Seisenberger et al. 2012; Smith et al. 2012; Cantone and Fisher 2013). The developmental methylation dynamics of L1 promoters have been examined using subfamily-specific and whole genome bisulfite sequencing (WGBS) (Hajkova et al. 2002; Kuramochi-Miyagawa et al. 2008; Watanabe et al. 2008; Popp et al. 2010; Saitou et al. 2012; Seisenberger et al. 2012; Smith et al. 2012; Molaro et al. 2014; Schöpp et al. 2020; Zoch et al. 2020). However, because of the repetitive structure and variable length of the mouse L1 promoter, as well as the high sequence identity among young L1 copies in the genome, assignmentof shortinternalreadsto specificmouseL1 lociischallenging (Lanciano and Cristofari 2020), and assessment of L1 methylation status en masse may mask individual L1s whose methylation dynamics differ from those of their subfamily. Indeed, studies of human L1s suggest certain loci can “escape”methylation and thus contribute to somatic retrotransposition throughout development and in cancer (Pitkänen et al. 2014;Tubioet al. 2014; Paterson et al. 2015; Philippe et al. 2016; Scott et al. 2016; Gardner et al. 2017; Nguyen et al. 2018; Schauer et al. 2018; Salvador-Palomeque et al. 2019; Sanchez-Luque et al. 2019; Ewing et al. 2020). Locusspecific resolutionof murine developmental L1 methylation, however, remains largely unexplored. Heritability of locus-specific retrotransposon methylation has been found in the context of “metastable epialleles”mostly involving variably methylated young IAP elements (VM-IAPs) (Bertozzi and Ferguson-Smith 2020). However, genome-wide screens have not revealed evidence of this phenomenon for L1s (Kazachenka et al. 2018; Elmer et al. 2021). We previously characterized five de novo (daughter) L1 insertions arising in pedigrees of inbred mice (Richardson et al. 2017), and identified their source (donor) elements through unique L1 3′transductions (Holmes et al. 1994; Moran et al. 1999; Goodier et al. 2000; Pickeral et al. 2000; Xing et al. 2006; Beck et al. 2010). In this study, we use the mosaic tissues from the animals in which theseinsertionsaroseandtheirheterozygousinsertion-bearingoff- spring, to investigate the retrotransposition potential and epigenetic fate of de novo and retrotransposition-competent L1 copies in vivo. We also use in vitro differentiation of mouse embryonic stem cells (mESCs) as a model to explore developmental L1 methylation dynamics at locus-specific resolution and genomewide. We use cell culture-based L1 retrotransposition assays, locus-specific bisulfite sequencing, Oxford Nanopore Technologies (ONT) long-read DNA sequencing and methylation profiling, and Pacific Biosciences (PacBio) SMRT sequencing of L1 5′RACE cDNAs to explore the relationship between developmental DNA methylation and L1 expression and retrotransposition capacity at single-locus resolution. Results L1 retrotransposition efficiency is diminished by ongoing promoter shortening To evaluate the retrotransposition potential of de novo and polymorphic daughter elements relative to their donors, we amplified via polymerase chain reaction (PCR), cloned and capillary sequenced five donor/daughter pairs (Richardson et al. 2017) to derive the exact nucleotide sequence of each element (Sanchez- Luque et al. 2019) (see Methods). All 10 L1s contained at least one T F monomer unit and encoded intact open reading frames (ORFs), and each daughter L1 contained between 0.6 and 1.8 fewer monomer units than the corresponding donor L1 (Table 1; Fig. 1A; Supplemental Fig. S1A). The remaining sequence of each daughter L1 was identical to its donor, with the exception of Insertion 2 which had a single nonsynonymous substitution in ORF1 (V303A) (Fig. 1A; Supplemental Fig. S1A). This result, representing a single nucleotide substitution among 33,374 reverse transcribed bases, is consistent with a high fidelity for mouse L1 reverse transcriptase activity in vivo. The loss of daughter element 5′UTR sequence could be explained either by 5′truncation during retrotransposition (Ostertag and Kazazian 2001; Symeret al. 2002; Zingler et al. 2005), or by the use of TSSs within internal monomers of the donor element promoter(DeBerardinisand Kazazian1999).We analyzedthe putative initiator dinucleotide (−1,+1) (Carninci et al. 2006) for the 10 elements under study (Fig. 1A), as well as 10 additional L1 T F elements (Supplemental Fig. S1B) comprising five likely donor/daughter pairs (Table 1). Under the assumption that the most 5′position of each element represents the first transcribed nucleotide, Gerdes et al. 1466 Genome Research www.genome.org transcription of 14/20 elements initiated at the preferred mammalian PolII initiator pyrimidine/purine dinucleotide (Carninci et al. 2006) (Fig. 1B; Supplemental Fig. S1C; Table 1). 15 of the 20 elements analyzed here were 5′truncated within the first 108 nt of the 5′-most T F monomer. 5/20 elements truncated within, and an additional 7/20 truncated in close proximity (≤21 bp) to, the YY1 corebindingmotif(GCCATCTT)(Fig.1C;Table1) aspreviouslyobserved for mouse L1s (Shehee et al. 1987; DeBerardinis and Kazazian 1999; Zhou and Smith 2019). The observed clustering of 5′truncationpoints,and theircoincidencewith thePy/Pu initiator dinucleotide, are consistent with the daughter L1 promoters being shortened due to transcription initiation internal to the 5′UTR of the donor element. To quantify the impact of monomer loss on daughter element mobility, we evaluated the five donor/daughter pairs identified in our previous study (Richardson et al. 2017) in a cultured cell L1 retrotransposition assay (Moran et al. 1996; Wei et al. 2000). We tested each element driven either by a cytomegalovirus promoter and the native L1 promoter (CMVp+5′UTR), or by the native L1 promoter only (5′UTR only), and quantified their activity relative to L1 spa (Fig. 1D; Kingsmore et al. 1994; Naas et al. 1998). In HeLa cells, when driven by CMVp+5′UTR, all elements mobilized efficiently (∼160% of L1 spa ), with donor and daughter elements showing similar activity. Notably, Insertion 2 which has an amino acid change in ORF1p retrotransposed with the same efficiency as its donor (∼160%) when transcribed from the CMVp, indicating that the mutation does not influence retrotransposition efficiency. In contrast, when driven by the 5′UTR alone, each daughter element mobilized less efficiently than its donor (Fig. 1E,F). This trend was most pronounced for Insertion 2, which retrotransposed at 8% of L1 spa +CMVp, compared to 62% for Donor 2, and reached statistical significance for three of the five donor/daughter pairs (One-way ANOVA followed by Sidak’s post-hoc test, P=0.0060, 0.7632, <0.0001, 0.5652, 0.0066 for donor/daughter pairs 2, 5, 7, 3 and 4, respectively; Fig. 1E,F). A similar trend was observed when donor/daughter pairs 2 and 5 (5′UTR only) were tested in Table 1. Characterization of L1 donor/daughter pairs L1 donor/ daughter pairs Coordinates a Location L1 subfamily Intact ORFs Monomer # 5′truncation point b TSS References Donor 2 Chr 19: 28,149, 495–28,156,131 Intergenic T FII Y 3.59 85 (0) CT Richardson et al. 2017 Insertion 2 Chr 2: 101,251, 997–101,252,012 Intergenic T FII Y 1.8 41 (−38) CC Richardson et al. 2017 Donor 5 Chr 1: 7,328, 682–7,335,740 Intergenic T FI Y 5.61 82 (0) CA Richardson et al. 2017 Insertion 5 Chr 10: 89,133, 398–89,133,411 Intronic (Ano4) T FI Y 5.01 208 (−83) CG Richardson et al. 2017 Donor 7 Chr 6: 95,658, 065–95,664,762 Intronic (Suclg2) T FII Y 3.82 37 (−41) CG Richardson et al. 2017 Insertion 7 Chr 18: 85,067, 455–85,067,469 Intergenic T FII Y 2.61 82 (0) CA Richardson et al. 2017 Donor 3 Chr 16: 65,055, 909–65,062,518 Intronic (Htr1f) T FII Y 3.49 108 (+21) TA Richardson et al. 2017 polyL1Tf_3 Chr 6: 108,844, 479–108,844,492 Intronic (Edem1) T FII Y 2.69 64 (−15) CA Richardson et al. 2017 Donor 4 Chr 7: 58,421, 728–58,428,359 Intergenic T FI Y 3.41 125 (+38) CA Richardson et al. 2017 polyL1Tf_4 Chr 17: 41,334, 114–41,334,129 Intergenic T FI Y 2.61 82 (0) CA Richardson et al. 2017 Donor L1_Lama2 Chr 13: 4,065, 522–4,076,041 Intronic (Akr1c14) T FII Y c 4.56 93 (+6) GG Gagnier et al. 2019 L1_Lama2 Chr 10: 27,124, 144–27,124,153 Intronic (Lama2) T FII N 3.22 166 (+79) CA Besse et al. 2003 Donor L1_HCC Chr 1: 86,584, 921–86,584,935 Intronic (Cops7b) T FI Y 3.57 92 (+5) CG Schauer et al. 2018 L1_HCC Chr 1: 12,998, 087–12,998,074 Intergenic T FI n/d 2.55 96 (+9) AC Schauer et al. 2018 Donor L1 spa Chr 12: 91,936, 221–91,943,749 Intergenic T FI Y 7.83 36 (−43) CC Not previously described L spa Chr 3: 80,861, 509–80,861,524 Intronic (G1rb) T FI Y 7.51 103 (+16) GA Kingsmore et al. 1994; Naas et al. 1998 Donor L1_Chr6 Chr 18: 10,875, 468–10,882,471 Intergenic T FI Y 5.35 137 (+50) TG Not previously described. L1_Chr6 Chr 6: 82,256, 067–82,262,701 Intergenic T FI N 3.61 82 (0) CA Not previously described Donor L1_Chr1 Chr 6: 24,673, 352–24,678,785 Intergenic T FII Y 2.65 75 (−4) CG Not previously described L1_Chr1 Chr 1: 32,542, 809–32,542,822 Intronic (Khdrbs2) T FII N d 1.33 142 (+55) CA Not previously described a Coordinates for nonreference insertions are the first and the last nucleotides of the TSD; coordinates for the reference insertions are the first and the last nucleotides delimiting the L1 in the mm10 reference genome. b 5′truncation position in the 5′-end monomer, numbers in brackets indicate how far upstream (−) or downstream (+) of YY1 core motif (GCCATCTT) truncation occurred, “0”indicates truncation within YY1 core motif. c IAP sequence inserted into ORF2 leading to a slightly truncated ORF2p with intact EN and RT domains. d Sequence obtained from ONT sequencing could be inaccurate. Locus-resolution L1 analysis in mouse development Genome Research 1467 www.genome.org A DE FGHI B C L1spa L1spa L1spa L1 spa +CMVp L1 spa -CMVp L1 spa +CMV L1 spa +CMVp L1 spa -CMVp Figure 1. L1 donor/daughter pairs retrotranspose efficiently in vitro. (A) Amino acid changes in ORF1 and ORF2 compared to the L1 T FI and L1 T FII consensus sequences (Sookdeo et al. 2013) are annotated in red. L1 spa refers to the published disease-causing insertion used in later experiments (Naas et al. 1998). Functional domains in ORF1 and ORF2 are shown: CC = coiled-coiled, RRM = RNA recognition motif, CTD = C-terminal domain, EN = endonuclease, RT = reverse transcriptase, C = cysteine-rich motif. Triangles within the 5′UTR represent monomer units. For nucleotide substitutions in promoters see Supplemental Figure S1.(B) Sequence logo (Crooks et al. 2004) of putative transcription initiation start sites for all ten donor and daughter L1 pairs. Sequence represents transcription initiator dinucleotide in the center ± 9 nucleotides upstream and downstream. −1,+1 indicates transcription initiator dinucleotide. The first nucleotide of L1 sequence corresponds to the second nucleotide in transcription initiator dinucleotide. The position of the first base of each daughter L1 relative to its donor element, and the first base of each donor L1 relative to the L1 T FI /T FII consensus sequences was analyzed. (C) Donor and daughter L1 5′truncation points. Sequences or locations of donors and daughters were previously published as indicated in Table 1 (Kingsmore et al. 1994; Naas et al. 1998; Besse et al. 2003; Richardson et al. 2017; Schauer et al. 2018; Gagnier et al. 2019). Lines indicate truncation points of elements in the 5′-most monomer. YY1 binding site (GCCATCTT) is shown in red. (D) Rationale of a cultured cell retrotransposition assay (Moran et al. 1996; Wei et al. 2000). Constructs used in this study were previously published (Moran et al. 1996; Goodier et al. 2001; Han and Boeke 2004) or generated by modifying the pTN201 construct [L1 spa (Naas et al. 1998)]. An antisense orientated neomycin-resistance (NEO r ) reporter cassette interrupted by a sense-oriented intron is inserted into a mouse L1 3′UTR. The mouse L1 is driven by its native 5′UTR promoter or a CMV promoter (CMVp). Cells harboring a retrotransposition event become neomycin (G418) resistant. The colony number reflects the relative activity of the L1 construct. (E) Comparison of L1 donor/daughter pair retrotransposition efficiency in HeLa cells. The retrotransposition assay timeline is shown in the top (S: seeding, T: transfection, M: change of media, G418: start of G418 selection, TE: measurement of transfection efficiency, F: Fixing and staining of colonies). Constructs: L1SM (positive control), L1SMmut2 (negative control), pTN201, L1 spa +CMVp/−CMVp, L1 donor/daughter pairs +CMVp/−CMVp. Colony counts were normalized to L1 spa + CMVp and are shown as mean ± SD of three independent biological replicates, each of which comprised three technical replicates. (∗)P≤0.0332, (∗∗)P≤0.0021, (∗∗∗)P≤0.0002, (∗∗∗∗)P< 0.0001, ns = not significant (One-way ANOVA followed by Sidak’s post-hoc test, P= 0.0060, 0.7632, <0.0001, 0.5652, 0.0066 for donor/daughter pairs –CMVp from left to right). Representative well pictures are shown below each construct. 5×10 3 cells were plated per well in a six-well plate. (F) Percentage change (ΔChange) in retrotransposition activity between L1 donor/daughter pairs. Shown is the decrease of retrotransposition efficiency per daughter L1 compared to its respective donor L1. Data is shown as mean ± SD of three independent biological replicates. (G) Comparison of L1 donor/daughter pair retrotransposition efficiency in mESCs. The retrotransposition assay timeline is shown at the top (S: seeding, T: transfection, M: change of media 8 h after transfection, P: passaging of cells into 10 cm plates, TE: measurement of transfection efficiency, G418: start of G418 selection, F: Fixing and staining of colonies). Constructs as described in (E). Colony counts were normalized to L1 spa + CMVp and are shown as mean ± SD of three independent biological replicates, each of which comprised two technical replicates. (∗)P≤0.0332, (∗∗)P≤0.0021, (∗∗∗)P≤0.0002, (∗∗∗∗)P< 0.0001, ns = not significant (One-way ANOVA followed by Sidak’s post-hoc test, P= 0.2851, 0.3305 for donor/daughter pairs – CMVp from left to right). Representative well pictures are shown below each construct. 4 × 10 5 cells were plated per well in a six-well plate. (H) Schematic of an L1 monomer unit. The YY1 binding site is indicated as red rectangle. The extended YY1 binding motif sequence is shown below. The core YY1 binding motif sequence is underlined. A mutation in the extended YY1 binding motif sequence adjacent to the core motif in Insertion 2 is indicated in red. (I) Comparison of retrotransposition efficiency of Insertion 2 and Insertion 2 with intact YY1 binding sites (Insertion 2-YY1 fixed) in retrotransposition assay in HeLa cells. Constructs as described in (D). Colony counts were normalized to L1 spa in pCEP4-mneoI-G4 + CMVp and are shown as mean ± SD of three independent biological replicates, each of which comprised three technical replicates. (∗)P≤0.05, ns = not significant (two-tailed t-test, P= 0.1607). Representative well pictures are shown below each construct. 1 × 10 4 cells were plated per well in a six-well plate. Note: L1SM retrotransposed very efficiently, leading to cell colony crowding in wells, and a likely underestimate of retrotransposition. mESCs (MacLennan et al. 2017), but did not reach statistical significance for either pair (One-way ANOVA followed by Sidak’s posthoc test, P=0.2851, 0.3305 for pairs 2 and 5, respectively; Fig. 1F,G). We also noted that Insertion 2 had a single nucleotide mutation (1115C>T) in the extended YY1 binding motif (GGTCGCCATCTTGGT) in its second monomer (Fig. 1H) which could decrease YY1 binding affinity (Kim and Kim 2009). To determine whether the 1115C> T mutation impacts the retrotransposition efficiency of this element, we restored the YY1 binding site (Insertion 2-YY1 fixed), and observed a ∼15% increase in retrotransposition activity which was not statistically significant (two-tailed t-test, P=0.1607; Fig. 1I). Together, these results are consistent with previous luciferase reporter assays showing mouse L1 promoter strength is proportionate to monomer number (DeBerardinis and Kazazian 1999). Moreover, we find that monomer loss consistently diminishes the retrotransposition potential of de novo mouse L1 insertions relative to their donor elements. Donor and daughter L1 insertions are largely methylated in adult tissues Having established the retrotransposition competence of the donor and daughter elements in vitro, we next used mouse L1 locus-specific bisulfite sequencing (Schauer et al. 2018) to evaluate the methylation status of each L1 in the somatic tissues and gonads of the mosaic animal in which the daughter insertion arose, and in subsequent generations of heterozygous insertion-bearing animals (Fig. 2A–D). For comparison, we analyzed the genomewide methylation of the T FI and T FII subfamily monomers using primers internal to the monomer sequence (Schauer et al. 2018). Because of their high sequence similarity, it was not possible to design primers specific to the T FI or T FII subfamily. Overall, the T FI / T FII subfamily monomer sequence was >80% methylated in adult tissues, although a few demethylated reads were observed across animals and tissues (Fig. 2E; Supplemental Fig. S2A). The three de novo daughter L1s (Insertions 2, 5, and 7) (Table 1) were >80% methylated in the somatic tissues and gonads of adult mice, including mosaic animals E (Insertion 5), CD14 (Insertion 7), and their heterozygous F1, F2, and F3 descendants (Fig. 2C,F; Supplemental Fig. S2B). Notably, Insertion 2 originated as germline-restricted mosaic in mouse B and was transmitted only to F1 animal AB15, which was harvested as a postimplantation embryo. Insertion 2 was highly methylated in the adult testis of mouse B, but partially demethylated in embryo AB15 (Fig. 2C,F). As the genomic DNA of embryo AB15 wasanalyzed in bulk, the demethylated sequences may potentially correspond to primordial germ cells (PGCs) or multipotent stem cells. Consistently, the T FI /T FII subfamily monomer sequence and Donor 2 (see below) also showed partial demethylation in embryo AB15 (Fig. 2E). We also analyzed the methylation status of unfixed polymorphic insertions polyL1Tf_3 and polyL1Tf_4 (Table 1). Insertion polyL1Tf_3 was nearly 100% methylated across all adult tissues examined (Fig. 2G; Supplemental Fig. S2B). However, methylation of polyL1Tf_4 was more relaxed (<90%), with a tendency to be especially demethylated in liver (Fig. 2G; Supplemental Fig. S2B). This variability may reflect the influence of genomic location and physiological context on L1 element methylation (Salvador-Palomeque et al. 2019; Sanchez-Luque et al. 2019; Ewing et al. 2020). Together, these results indicate that de novo L1 insertions arising during embryonic development are likely silenced by DNA methylation during later embryogenesis. This methylation is maintained in subsequent generations, with an average methylation level of 93% in brain, 89% in heart and 86% in liver for daughter insertions and 91% in brain, 87% in heart and 88% in liver for the donor L1s. We viewed donor L1s active during embryonic development as candidate “escapee”loci in mice, potentially akin to specific human L1 loci that evade epigenetic repression in differentiated cells (Scott et al. 2016; Salvador-Palomeque et al. 2019; Sanchez-Luque et al. 2019; Ewing et al. 2020). We therefore assessed methylation of Donor 2, Donor 5, and Donor 7 in somatic tissues and germ cells of their mosaic founder animals (mouse B, mouse E, and mouse CD14, respectively). We also analyzed methylation of Donor 3 and Donor 4 in animals that carried the respective polymorphic daughter insertions polyL1Tf_3 and polyL1Tf_4 (Fig. 2B,D). Nearly all donor elements showed >80% methylation in somatic tissues and gonads (Fig.2H; Supplemental Fig. S2C). The exception was Donor 7 which, although it was completely methylated in the somatic tissues of founder mouse CD14, was hypomethylated in the germ cell fraction of mouse CD14 testis. This is a notable departure from the genome-wide state of T FI /T FII monomer sequences, which we found to be largely methylated in adult gonads (Fig. 2E,H; Supplemental Fig. S2A,C). Thus, Donor 7 may represent an L1 that is refractory to methylation during germline development and therefore privileged for heritable retrotransposition. Young L1s are rapidly methylated during in vitro mESC differentiation As adult tissues reveal only the end point of developmental L1 methylation dynamics, we next analyzed L1 methylation at genome-wide and locus-specific resolution during cellular differentiation. To model various states of pluripotency, we cultured feederfree E14 mESCs in three conditions: serum complemented with leukemia inhibitory factor (serum+ LIF), which generates a heterogeneous population of mESCs in a pluripotent state (Smith et al. 1988; Williams et al. 1988); two small kinase inhibitors+LIF (2i + LIF), which maintains the mESCs in a naive ground state similar to that of the inner cell mass (ICM) (Silva et al. 2008); and under 2i+serum conditions, which are shown to support engineered mouse L1 retrotransposition (MacLennan et al. 2017). To recapitulate specification and differentiation of cells into the three germ lineages (ectoderm, mesoderm, and endoderm), we collected genomic DNA overa time course,from differentiation induction of serum+LIF mESCs to embryoid bodies (EBs) for 6 d, and through subsequent differentiation over 15 d (Fig. 3A,B; Supplemental Fig. S3A,B; Behringer et al. 2016). L1 methylation state was assessed using genome-wide and locus-specific L1 bisulfite sequencing in the three mESC culture conditions and on days 3, 6, 9, 12, 15, 18 and 21 of differentiation. The mobile L1 subfamilies T F ,G F and A each showed their lowest methylation levels in 2i +LIF ground-state conditions, and reached maximal methylation by day 6 of differentiation (Fig. 3E–H; Supplemental Fig. S4C–F). Notably, the G F subfamily was less methylated (<80%) than the other subfamilies (>80%) both during differentiation and in fully differentiated EBs (Fig. 3G; Supplemental Fig. S4F). The 129/Ola genetic background from which E14 mESCs are derived contained two of the donor L1s analyzed above, Donor 3 and Donor 7. Both loci were largely demethylated (<50%) across all mESC culture conditions (Fig. 3C,D; Supplemental Fig. S4A,B). Donor 3 showed 80% methylation at day 3 of EB differentiation, and >90% at day 6. Donor 7 showed >90% methylation at day 9 and day 21 (Fig. 3C; Supplemental Fig. S4A). These results were in line with our Genome Research 1469 www.genome.org Locus-resolution L1 analysis in mouse development A D F G H E BC Figure 2. L1 donor/daughter elements are methylated in somatic tissues of adult mice. (A) Schematic of CpG dinucleotides in a mouse L1 T F element. Triangles in 5′UTR represent monomer units. A magnification of the 5′UTR is shown below. Black boxes represent monomer units. Dark gray box represents unique (nonmonomeric) region within 5′UTR. Orange strokes represent CpG dinucleotides. Red boxes represent YY1 binding sites. (B) Experimental design of mouse L1 locus-specific bisulfite sequencing. Genomic DNA was extracted from tissues of C57BL6/J mice harboring previously identified donor and daughter L1 insertions (Richardson et al. 2017). The parental generation “P”(square = male, circle = female) is mosaic for the de novo daughter L1 insertion (represented by stripes). F1-F3 generations are heterozygous for L1 insertions (filled diamond = male or female). F2 and F3 generation animals were only available for Insertion 5 and polyL1Tf_4. Tissues for analysis of donor elements were only collected from the P generation. DNA was isolated from brain (green), heart (red), liver (orange), and gonads (gray) if available. After bisulfite conversion, the 5′monomeric region of each L1 was PCR amplified. Amplicons were pooled and sequenced as 2 × 300-mer Illumina reads. Circles represent methylated (black circles) and unmethylated (white circles) CpG dinucleotides in L1 5′UTR. (C) Methylation of a de novo L1 promoter sequence (Insertion 2) shown in the germline mosaic parental testis (animal B) and in the following F1 generation embryonic tissue (animal AB15). Displayed are 50 nonidentical sequences extracted at random from a much larger pool of available Illumina reads. Each cartoon panel corresponds to an amplicon (black circle, methylated CpG; white circle, unmethylated CpG; ×, mutated CpG). Colored line above the cartoon represents amplicon (gray = genomic sequence, colored = L1 sequence). The overall percentage of methylated CpG dinucleotides is indicated below each cartoon. Gray letters indicate methylation of CpG dinucleotides in genomic sequence. Colored letters indicate methylation of CpG dinucleotides in L1 sequence. (D) Locus-specific methylation analysis schematic representation for L1 T F monomer, 3 full-length de novo L1 insertions (Insertion 2, 5, 7), 2 polymorphic L1 insertions (polyL1Tf_3 and polyL1Tf_4) and their 5 respective donor elements (Donor 2, 5, 7, 3, 4). 5′monomeric sequences of each L1 were PCR amplified using primer pairs (green arrows) specific to that locus. Orange strokes indicate L1 CpG dinucleotides covered by the assay. Blue strokes represent covered genomic CpG dinucleotides. Gray strokes in the gray shaded area represent CpG dinucleotides not reached by Illumina sequencing. Red boxes indicate YY1 binding sites. Colored boxes represent L1 monomer units. (E) Genome-wide methylation of L1 T FI / T FII promoter sequence shown in all animals. Animals are labeled E, EF19, 137, 138, 235, B, AB5, AB15, CD14 and 55 in the bottom part of the x-axis label, with the generation (P, F1, F2 or F3) indicated in brackets. Each graph contains animals from the same family. The different tissues used for DNA extraction and bisulfite sequencing in each animal is indicated in the top part of the x-axis labeled as: brain, B; heart, H; liver, L; testis, T; ovaries, O; and embryonic tissues, E. Displayed are 1000 nonidentical sequences extracted at random from a much larger pool of available Illumina reads. The violin plots represent the methylation distribution as per Supplemental Figure S2. The black line and dashed lines show the distribution median and quartiles, respectively. The percentage of methylated CpGs per read is indicated on the y-axis. (F) As for (E) but for de novo L1 promoter sequences shown in the mosaic P generation in which each de novo L1 insertion was identified (animals E and CD14) and in following F1–F3 generations if available (animals EF19, 138, 235 and 55). Displayed are 1000 nonidentical sequences (if available) extracted at random from a much larger pool of available Illumina reads (exceptions: Insertion 5: 138 H, 235 B, H, L, O, EF19 T [368, 328, 337, 352, 958, 282 reads, respectively]; Insertion 7: 55 B, H, L, CD14 T [598, 885, 890, 662 reads, respectively]). (G) As for (E,F) but for polymorphic L1 insertions. (H) As for (E,F) but showing methylation of 5 donor L1 elements in the animal they mobilized in (P generation). 1470 Genome Research www.genome.org A DE FG HI BC Figure 3. Dynamic methylation of L1 elements during differentiation of mESCs. (A) Differentiation of mESCs to cells of all three germ layers using a standard differentiation protocol (Behringer et al. 2016). Undifferentiated E14 mESCs are grown on gelatin (day 0). Embryoid bodies (EBs) are generated by “hanging drop culture”(day 3) and are grown in suspension culture (day 6). After 6 d, EBs are plated and differentiated for 2 wk (day 21). Scale bar, 200 µm. (B) Immunofluorescence image of mesodermal (Actin, green), endodermal (AFP, red) and ectodermal (Tubulin, violet) lineage markers in differentiated E14 mESCs on day 21. Nuclei were stained with Hoechst (blue). Scale bar, 100 µm. (C) Methylation of Donor 7 promoter sequence shown in the mESCs cultured in three different conditions (2i + L = 2i + LIF, 2i + s = 2i + serum, s + L = serum + LIF) and on differentiation day 9 (d9) and day 15 (d15). Because of the technical challenge posed by PCR amplification of long bisulfite-treated fragments, sufficient material was generated to assess Donor 7 methylation only at day 9 and day 15 of differentiation. Displayed are 1000 nonidentical sequences (if available) extracted at random from a much larger pool of available Illumina reads (exceptions: Donor 7: 2i + LIF, 2i + serum, serum + LIF, d9, d15 [874, 837, 876, 110, 124 reads, respectively]). The violin plots represent the methylation distribution as per Supplemental Figure S4. The black line and dashed lines show the distribution median and quartiles, respectively. (D–I)Asfor(C)butforDonor3(D), L1 T FI /T FII family (E), L1 T FIII family (F), L1 G F family (G), L1 A family (H), and the imprinted gene Snrpn (I). Primers for L1 families are within the L1 promoter sequence. Shown is methylation in three different mESC culture conditions, during EB culture and during differentiation. Displayed are 1000 nonidentical sequences (if available) extracted at random from a much larger pool of available Illumina reads (exceptions: L1 G F family: 2i + LIF, 2i + serum, serum + LIF, d3, d6, d9, d12, d15, d18, d21 [86, 127, 119, 120, 168, 162, 161, 155, 155, 175 reads, respectively]). Locus-resolution L1 analysis in mouse development Genome Research 1471 www.genome.org observation that both donors were completely methylated in somatic tissues of adult mice (Fig. 2H) and is consistent with methylation occurring during differentiation in embryonic development in vivo. As an internal control, analysis of the maternally imprinted Snrpn gene revealed the expected bimodal distribution of methylation in pluripotent cells and across the differentiation time course (Fig. 3I; Supplemental Fig. S4G). Together, these results show rapid remethylation of L1 sequences during cellular differentiation, with subtle but notable variability between active L1 subfamilies and among individual loci. Methylation fluctuates within mouse L1 promoters Although locus-specific bisulfite sequencing allows base-pair resolution of individual L1 methylation, it is limited to the 5′-most portion of each L1 promoter. To attain complete methylation profiles of L1 loci without bisulfite conversion, we performed PCR-free ONT sequencing of mESCs in serum + LIF (day 0; d0), day 3 (d3) EBs, and day 21 (d21) differentiated cells. We achieved ∼15–20× genome-wide depth using an ONT PromethION platform, and surveyed CpG methylation via the methylartist package (Ewing et al. 2020; Cheetham et al. 2022). Evaluated by ONT sequencing, DNA methylation increased during differentiation (Fig. 4A). Examining the methylation profiles of each L1 subfamily averaged acrosstheir 5′UTRs, we found that in d0 mESCs the 5′UTRs of full-length T F , G F and A subfamily L1s were less methylated compared to the older L1 F subfamily (Fig. 4A). Average methylation of all L1 subfamilies rapidly increased to ∼80% by day 3 of differentiation; however, T FI and T FII elements were less methylated (≤80%) in d3 EBs compared to other L1 subfamilies (Fig. 4A). B1, B2, and BC1 SINEs were only slightly demethylated in d0 mESCs (∼80%), and fully methylated (>90%) in d3 EBs and d21 fully differentiated cells (Fig. 4B). While IAP retrotransposon LTRs were generally almost 100% methylated at all three differentiation time points, the IAP subfamily IAPEY4 LTRs showed an average methylation level of <40% in d0 mESCs and only ∼70% methylation in d3 EBs and d21 differentiated cells. MERVL/MT2 LTRs were >80% methylated at all three time points. Etn elements showed an average methylation level of only ∼60% in d0 mESCs but >80% in d3 EBs and d21 differentiated cells (Fig. 4C). Next, we quantified the methylation of active L1 subfamily promoters independently of the L1 body (unique 5′UTR region, ORFs, 3′UTR). We binned promotersgenome-wide based onmonomer count, with the majority of young L1 subfamily members containing between 2 and 5 monomer units (Fig. 4D), consistent with previous analyses of the mouse reference genome (Zhou and Smith 2019). On the whole, L1 promoter methylation was distributed between ∼0% and 90% in the d0 mESCs, indicating variability among loci even in pluripotent serum+LIF culture conditions. Methylation of L1 loci in d3 EBs was higher than in d0 mESCs but a substantial proportion of L1s were still <80% methylated. Notably, none of the L1 loci displayed here appeared to be completely demethylated in d3 EBs. Methylation of the majority of L1s was re-established (>80%) in d21 differentiated cells (Fig. 4D). However, some loci appeared to be <70% methylated even in d21 differentiated cells. We examined three of these methylation “escapees”and found for each a mixture of methylated and demethylated reads at d21 potentially belonging to specific cell types in our mixed population of differentiatedcells(SupplementalFig.S5)indicating again a likely lineageor cell type specificity for L1 methylation “escapees”(Salvador-Palo- meque et al. 2019; Sanchez-Luque et al. 2019; Ewing et al. 2020). Based on the mouse genome reference sequence all three L1s contain intact ORFs and multiple (4–7) monomers, indicating their potential retrotransposition competence. We next assessed composite methylation profiles covering the previously inaccessible interiors of full-length mouse L1s and in particular the entire mouse L1 promoter (Fig. 4E). We observed a consistent methylation trough in the 5′UTR promoter region in d0 mESCs and d3 EBs, whereas the L1 body was consistently hypermethylated at all three time points (Fig. 4F). Zooming in on the promoter region of the composite L1 methylation profiles revealed a consistent “smile”pattern across the 5′UTR, with the innermost monomers less methylated compared to the 5′-most and 3′-most monomers (Fig. 4G; Supplemental Fig. S6). This methylation pattern was most pronounced in T FI and T FII elements, and in elements containing three or more monomer units (Fig. 4G; Supplemental Fig. S6A–D). Elements with two monomers appeared highly variable in their methylation status in d0 mESCs (Fig. 4G; Supplemental Fig. S6A–D). Our analyses also revealed peaks and valleys of methylation along mouse L1 promoters, with a periodicity corresponding to the monomer units. Taking advantage of the locus-specific resolution offered by ONT sequencing, we examined methylation of the two donor L1s (Donor 3 and Donor 7) present in E14 mESCs (Supplemental Fig. S7). These elements recapitulated the methylation trough observed in the composite plots in d0 mESCs and d3 EBs. As an internal control, we readily identified the differentially methylated regions (DMRs) of two imprinted genes, Snrpn and Impact (Supplemental Fig. S8). To elucidate the pattern of DNA methylation within the L1 T FI 5′UTR at single nucleotide resolution, we determined the median percent methylation at each CpG dinucleotide, as well as the average percent methylation for all CpG dinucleotides (%mCpG) within each monomer unit, among at least 20 individual L1 T FI loci containing 2, 3, 4, or 5 monomers, (Fig. 5; Supplemental Fig. 9A–C). This analysis allowed us to quantify the “smile”pattern described above. Forexample, at day 3 among L1 loci containing five monomer units, the average %mCpG per monomer from the genome-proximal monomer inwards was 68.1%, 51.5%, 38.7%, 50.3%, and 73.5% (Fig. 5, middle panel); a similar pattern was observed for L1 loci containing 2, 3, and 4 monomers (Supplemental Fig. S9). Although methylation was highly variable at most CpG positions, the CpGs flanking the monomer-monomer borders were consistently hypomethylated relative to the CpGs internal to the monomer units. This trend was consistent across L1 T FI elements regardless of monomer count (Fig. 5; Supplemental Fig. S9). We also observed that methylation levels peak at the CpGs flanking the YY1 binding site within each monomer unit, with a slight dip in methylation around the core YY1 binding site. In sum, our ONT methylation analysis of full-length L1s provides unprecedented resolution of interior mouse promoter methylation dynamics. DNA methylation during differentiation impacts mouse L1 TSS distribution The fluctuating L1 promoter methylation patterns during mESC differentiation led us to investigate the influence of DNA methylation on L1 T FI TSS usage. We performed 5′rapid amplification of cDNA ends (5′RACE) on total RNA isolated from d0, d3, and d21 differentiated mESCs, followed by PCR with an L1 T FI ORF1-specif- ic primer in an approach similar to that used by Deininger et al. to study human L1 expression (Fig. 6A; Deininger et al. 2017). Purified 5′RACE products were sequenced on a PacBio SMRT flow cell. Reads were filtered to retain those aligned to only one Gerdes et al. 1472 Genome Research www.genome.org reference genome L1 T FI element, leveraging internal L1 sequence polymorphisms to identify unique TSSs supported by at least one read. At all three time points, TSSs within L1 T FI monomers clustered around the YY1 binding site as previously reported, consistent with our sequence analyses of donor/daughter L1 pairs (Figs. 1C, 6B; DeBerardinis and Kazazian 1999). The most prominent peak at all three time points was at position 65 of the T F monomer consensus sequence, 15 bp upstream of the YY1 binding site (positions 80–87) (Fig. 6B). This finding recalls the situation in the human L1 5′UTR, in which YY1 directs transcription initiation A C E F G D B Figure 4. ONT CpG methylation profiles of TEs. (A) Violin plots are showing methylated CpG fraction for the whole genome (6 kbp windows), L1 5′UTRs belonging to the active L1 T FI ,T FII ,T FIII ,G F and A subfamilies and the evolutionary older and inactive L1 F subfamily at three time points of differentiation: d0 (undifferentiated mESCs in serum + LIF), d3 (EBs on day 3 of differentiation) and on d21 (completely differentiated cells). (B) As for (A), but for B1, B2 and BC1 SINE subfamilies. (C) As for (A), but for IAP LTR1a_Mm and LTR1_Mm, IAPEY4 LTR, MERV-L/MT2 LTR and RLTR ETn_Mm copies. (D) As for (A), but for violin plots showing the methylated CpG fraction of active L1 subfamily (T F ,G F , and A together) promoters (monomers only) depending on the number of monomers (including partial monomers). Only elements with a minimum coverage of five reads across the whole 5′UTR were included in the plot. The number of loci represented in each bin is shown in the top.(E) Annotated full-length L1 T F1 consensus showing the monomer units in green, unique region in light gray, ORF1 in dark gray, ORF2 in dark green, and 3′UTR in light gray. CpG dinucleotides throughout the whole element are displayed as orange strokes. The promoter CpG island (CGI) is indicated as an orange line. Number of bp are shown above the element. (F) Data is shown for full-length L1s (T FI , T FII ,T FIII ,G F and A) containing four monomers in the promoter at three time points of differentiation: d0 (undifferentiated mESCs in serum +LIF), d3 (EBs on day 3 of differentiation), and on d21 (completely differentiated cells). Each graph displays up to 50 methylation profiles for the specified L1 subfamily. Annotated consensus sequences as per (E) are shown at top including CpG positions. (G) As for (F), but for promoters of L1 T FI subfamily members containing between 2 and 5 monomers at three time points of differentiation (d0, d3, and d21). Locus-resolution L1 analysis in mouse development Genome Research 1473 www.genome.org et al. 2017. Mobilization of LINE-1 retrotransposons is restricted by Tex19.1 in mouse embryonic stem cells. eLife 6: e26152. doi:10.7554/ eLife.26152 Martin SL, Bushman FD. 2001. Nucleic acid chaperone activity of the ORF1 protein from the mouse LINE-1 retrotransposon. Mol Cell Biol 21: 467– 475. doi:10.1128/MCB.21.2.467-475.2001 Mathias S, Scott A, Kazazian H, Boeke J, Gabriel A. 1991. Reverse transcriptase encoded by a human transposable element. Science 254: 1808– 1810. doi:10.1126/science.1722352 Mears ML, Hutchison CA. 2001. The evolution of modern lineages of mouse L1 elements. J Mol Evol 52: 51–62. doi:10.1007/s002390010133 Mita P, Wudzinska A, Sun X, Andrade J, Nayak S, Kahler DJ, Badri S, LaCava J, Ueberheide B, Yun CY, et al. 2018. LINE-1 protein localization and functional dynamics during the cell cycle. eLife 7: e30058. doi:10 .7554/eLife.30058 Mita P, Sun X, Fenyö D, Kahler DJ, Li D, Agmon N, Wudzinska A, Keegan S, Bader JS, Yun C, et al. 2020. BRCA1 and S phase DNA repair pathways restrict LINE-1 retrotransposition in human cells. Nat Struct Mol Biol 27: 179–191. doi:10.1038/s41594-020-0374-z Molaro A, Falciatori I, Hodges E, Aravin AA, Marran K, Rafii S, McCombie WR, Smith AD, Hannon GJ. 2014. Two waves of de novo methylation during mouse germ cell development. Genes Dev 28: 1544–1549. doi:10.1101/gad.244350.114 Moran JV, Holmes SE, Naas TP, DeBerardinis RJ, Boeke JD, Kazazian HH. 1996. High frequency retrotransposition in cultured mammalian cells. Cell 87: 917–927. doi:10.1016/S0092-8674(00)81998-4 Moran JV, DeBerardinis RJ, Kazazian HH. 1999. Exon shuffling by L1 retrotransposition. Science 283: 1530–1534. doi:10.1126/science.283.5407 .1530 Naas TP, DeBerardinis RJ, Moran JV, Ostertag EM, Kingsmore SF, Seldin MF, Hayashizaki Y, Martin SL, Kazazian HH. 1998. An actively retrotransposing, novel subfamily of mouse L1 elements. EMBO J 17: 590–597. doi:10 .1093/emboj/17.2.590 Nam CH, Youk J, Kim JY, Lim J, Park JW, Oh SA, Lee HJ, Park JW, Won H, Lee Y, et al. 2023. Widespread somatic L1 retrotransposition in normal colorectal epithelium. Nature 617: 540–547. doi:10.1038/s41586-023- 06046-z Nellåker C, Keane TM, Yalcin B, WongK, Agam A, Belgard TG, Flint J, Adams DJ, Frankel WN, Ponting CP. 2012. The genomic landscape shaped by selection on transposable elements across 18 mouse strains. Genome Biol 13: R45. doi:10.1186/gb-2012-13-6-r45 Nguyen THM, Carreira PE, Sanchez-Luque FJ, Schauer SN, Fagg AC, Richardson SR, Davies CM, Jesuadian JS, Kempen MJHC, Troskie RL, et al. 2018. L1 retrotransposon heterogeneity in ovarian tumor cell evolution. Cell Rep 23: 3730–3740. doi:10.1016/j.celrep.2018.05.090 Okano M, Xie S, Li E. 1998. Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nat Genet 19: 219– 220. doi:10.1038/890 Ostertag EM, Kazazian HH. 2001. Twin priming: A proposed mechanism for the creation of inversions in L1 retrotransposition. Genome Res 11: 2059–2065. doi:10.1101/gr.205701 Paterson AL, Weaver JMJ, Eldridge MD, Tavaré S, Fitzgerald RC, Edwards PAW. 2015. Mobile element insertions are frequent in oesophageal adenocarcinomas and can mislead paired-end sequencing analysis. BMC Genomics 16: 473. doi:10.1186/s12864-015-1685-z Penzkofer T, Jäger M, Figlerowicz M, Badge R, Mundlos S, Robinson PN, Zemojtel T. 2017. L1Base 2: more retrotransposition-active LINE-1s, more mammalian genomes. Nucleic Acids Res 45: D68–D73. doi:10 .1093/nar/gkw925 Philippe C, Vargas-Landin DB, Doucet AJ, Van Essen D, Vera-Otarola J, Kuciak M, Corbin A, Nigumann P, Cristofari G. 2016. Activation of individual L1 retrotransposon instances is restricted to cell-type dependent permissive loci. eLife 5: e13926. doi:10.7554/eLife.13926 Pickeral OK, Makałowski W, Boguski MS, Boeke JD. 2000. Frequent human genomic DNA transduction driven by LINE-1 retrotransposition. Genome Res 10: 411–415. doi:10.1101/gr.10.4.411 Pitkänen E, Cajuso T, Katainen R, Kaasinen E, Välimäki N, Palin K, Taipale J, Aaltonen LA, Kilpivaara O. 2014. Frequent L1 retrotranspositions originating from TTC28 in colorectal cancer. Oncotarget 5: 853–859. doi:10 .18632/oncotarget.1781 Popp C, Dean W, Feng S, Cokus SJ, Andrews S, Pellegrini M, Jacobsen SE, Reik W. 2010. Genome-wide erasure of DNA methylation in mouse primordial germ cells is affected by AID deficiency. Nature 463: 1101–1105. doi:10.1038/nature08829 Richardson SR, Gerdes P, Gerhardt DJ, Sanchez-Luque FJ, Bodea GO, Muñoz-Lopez M, Jesuadian JS, Kempen MJHC, Carreira PE, Jeddeloh JA, et al. 2017. Heritable L1 retrotransposition in the mouse primordial germline and early embryo. Genome Res 27: 1395–1405. doi:10.1101/gr .219022.116 Richardson SR, Chan D, Gerdes P, Han JS, Boeke JD, Faulkner GJ. 2022. Revisiting the impact of synthetic ORF sequences on engineered LINE-1 retrotransposition. bioRxiv doi:10.1101/2022.08.29.505632 Saitou M, Kagiwada S, Kurimoto K. 2012. Epigenetic reprogramming in mouse pre-implantation development and primordial germ cells. Development 139: 15–31. doi:10.1242/dev.050849 Salvador-Palomeque C, Sanchez-Luque FJ, Fortuna PRJ, Ewing AD, Wolvetang EJ, Richardson SR, Faulkner GJ. 2019. Dynamic methylation of an L1 transduction family during reprogramming and neurodifferentiation. Mol Cell Biol 39: e00499-18. doi:10.1128/MCB.00499-18 Sanchez-Luque FJ, Kempen MJHC, Gerdes P, Vargas-Landin DB, Richardson SR, Troskie RL, Jesuadian JS, Cheetham SW, Carreira PE, Salvador- Palomeque C, et al. 2019. LINE-1 evasion of epigenetic repression in humans. Mol Cell 75: 590–604.e12. doi:10.1016/j.molcel.2019.05.024 Saxton JA, Martin SL. 1998. Recombination between subtypes creates a mosaic lineage of LINE-1 that is expressed and actively retrotransposing in the mouse genome. J Mol Biol 280: 611–622. doi:10.1006/jmbi.1998 .1899 Schauer SN, Carreira PE, Shukla R, Gerhardt DJ, Gerdes P, Sanchez-Luque FJ, Nicoli P, Kindlova M, Ghisletti S, Dos Santos AD, et al. 2018. L1 retrotransposition is a common feature of mammalian hepatocarcinogenesis. Genome Res 28: 639–653. doi:10.1101/gr.226993.117 Schichman SA, Adey NB, Edgell MH, Hutchison CA. 1993. L1 A-monomer tandem arrays have expanded during the course of mouse L1 evolution. Mol Biol Evol 10: 552–570. doi:10.1093/oxfordjournals.molbev .a040025 Schöpp T, Zoch A, Berrens RV, Auchynnikava T, Kabayama Y, Vasiliauskaite L, Rappsilber J, Allshire RC, O’Carroll D. 2020. TEX15 is an essential executor of MIWI2-directed transposon DNA methylation and silencing. Nat Commun 11: 3739. doi:10.1038/s41467-020-17372-5 Scott AF, Schmeckpeper B, Abdelrazik M, Comey C, O’Hara B, Rossiter J, Cooley T, Heath P, Smith K, Margolet L. 1987. Origin of the human L1 elements: Proposed progenitor genes deduced from a consensus DNA sequence. Genomics 1: 113–125. doi:10.1016/0888-7543(87) 90003-6 Scott EC, Gardner EJ, Masood A, Chuang NT, Vertino PM, Devine SE. 2016. A hot L1 retrotransposon evades somatic repression and initiates human colorectal cancer. Genome Res 26: 745–755. doi:10.1101/gr .201814.115 Seisenberger S, Andrews S, Krueger F, Arand J, Walter J, Santos F, Popp C, Thienpont B, Dean W, Reik W, et al. 2012. The dynamics of genomewide DNA methylation reprogramming in mouse primordial germ cells. Mol Cell 48: 849–862. doi:10.1016/j.molcel.2012.11.001 Seki Y, Hayashi K, Itoh K, Mizugaki M, Saitou M, Matsui Y. 2005. Extensive and orderly reprogramming of genome-wide chromatin modifications associated with specification and early development of germ cells in mice. Dev Biol 278: 440–458. doi:10.1016/j.ydbio.2004.11.025 Senft AD, Macfarlan TS. 2021. Transposable elements shape the evolution of mammalian development. Nat Rev Genet 22: 691–711. doi:10.1038/ s41576-021-00385-1 Severynse DM, Hutchison CA, Edgell MH. 1991. Identification of transcriptional regulatory activity within the 5′A-type monomer sequence of the mouse LINE-1 retroposon. Mamm Genome 2: 41–50. doi:10.1007/ BF00570439 Shehee WR, Chao SF, Loeb DD, Comer MB, Hutchison CA, Edgell MH. 1987. Determination of a functional ancestral sequence and definition of the 5′end of A-type mouse L1 elements. J Mol Biol 196: 757–767. doi:10.1016/0022-2836(87)90402-5 Silva J, Barrandon O, Nichols J, Kawaguchi J, Theunissen TW, Smith A. 2008. Promotion of reprogramming to ground state pluripotency by signal inhibition. PLoS Biol 6: e253. doi:10.1371/journal.pbio.0060253 Simpson JT, Workman RE, Zuzarte PC, David M, Dursi LJ, Timp W. 2017. Detecting DNA cytosine methylation using nanopore sequencing. Nat Methods 14: 407–410. doi:10.1038/nmeth.4184 Skowronski J, Fanning TG, Singer MF. 1988. Unit-length LINE-1 transcripts in human teratocarcinoma cells. Mol Cell Biol 8: 1385–1397. doi:10 .1128/mcb.8.4.1385-1397.1988 Slater GSC, Birney E. 2005. Automated generation of heuristics for biological sequence comparison. BMC Bioinformatics 6: 31. doi:10.1186/1471- 2105-6-31 Smith AG, Heath JK, Donaldson DD, Wong GG, Moreau J, Stahl M, Rogers D. 1988. Inhibition of pluripotential embryonic stem cell differentiation by purified polypeptides. Nature 336: 688–690. doi:10.1038/ 336688a0 Smith ZD, Chan MM, Mikkelsen TS, Gu H, Gnirke A, Regev A, Meissner A. 2012. A unique regulatory phase of DNA methylation in the early mammalian embryo. Nature 484: 339–344. doi:10.1038/nature10960 Sookdeo A, Hepp CM, McClure MA, Boissinot S. 2013. Revisiting the evolution of mouse LINE-1 in the genomic era. Mob DNA 4: 3. doi:10.1186/ 1759-8753-4-3 Gerdes et al. 1480 Genome Research www.genome.org Symer DE, Connelly C, Szak ST, Caputo EM, Cost GJ, Parmigiani G, Boeke JD. 2002. Human L1 retrotransposition is associated with genetic instability in vivo. Cell 110: 327–338. doi:10.1016/S0092-8674(02)00839-5 Tachibana M, Nozaki M, Takeda N, Shinkai Y. 2007. Functional dynamics of H3K9 methylation during meiotic prophase progression. EMBO J 26: 3346–3359. doi:10.1038/sj.emboj.7601767 Tan X, Xu X, Elkenani M, Smorag L, Zechner U, Nolte J, Engel W, Pantakani DVK. 2013. Zfp819, a novel KRAB-zinc finger protein, interacts with KAP1 and functions in genomic integrity maintenance of mouse embryonic stem cells. Stem Cell Res 11: 1045–1059. doi:10.1016/j.scr.2013.07 .006 Taylor MS, LaCava J, Mita P, Molloy KR, Huang CRL, Li D, Adney EM, Jiang H, Burns KH, Chait BT, et al. 2013. Affinity proteomics reveals human host factors implicated in discrete stages of LINE-1 retrotransposition. Cell 155: 1034–1048. doi:10.1016/j.cell.2013.10.021 Tristán-Ramos P, Rubio-Roldan A, Peris G, Sánchez L, Amador-Cubero S, Viollet S, Cristofari G, Heras SR. 2020. The tumor suppressor microRNA let-7 inhibits human LINE-1 retrotransposition. Nat Commun 11: 5712. doi:10.1038/s41467-020-19430-4 Tubio JMC, Li Y, Ju YS, Martincorena I, Cooke SL, Tojo M, Gundem G, Pipinikas CP, Zamora J, Raine K, et al. 2014. Extensive transduction of nonrepetitive DNA mediated by L1 retrotransposition in cancer genomes. Science 345: 1251343. doi:10.1126/science.1251343 Voliva CF, Jahn CL, Comer MB, Hutchison CA, Edgell MH. 1983. The L1Md long interspersed repeat family in the mouse: almost all examples are truncated at one end. Nucleic Acids Res 11: 8847–8859. doi:10.1093/ nar/11.24.8847 Watanabe T, Totoki Y, Toyoda A, Kaneda M, Kuramochi-Miyagawa S, Obata Y, Chiba H, Kohara Y, Kono T, Nakano T, et al. 2008. Endogenous siRNAs from naturally formed dsRNAs regulate transcripts in mouse oocytes. Nature 453: 539–543. doi:10.1038/nature06908 Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R, Ainscough R, Alexandersson M, An P, et al. 2002. Initial sequencing and comparative analysis of the mouse genome. Nature 420: 520–562. doi:10.1038/nature01262 Wei W,Morrish TA, Alisch RS, Moran JV. 2000. A transient assay revealsthat cultured human cells can accommodate multiple LINE-1 retrotransposition events. Anal Biochem 284: 435–438. doi:10.1006/abio.2000.4675 Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF, Metcalf D, Nicola NA, Gough NM. 1988. Myeloid leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 336: 684–687. doi:10.1038/336684a0 Wincker P, Jubier-Maurin V, Roizés G. 1987. Unrelated sequences at the 5′ end of mouse LINE-1 repeated elements define two distinct subfamilies. Nucleic Acids Res 15: 8593–8606. doi:10.1093/nar/15.21.8593 Xing J, Wang H, Belancio VP, Cordaux R, Deininger PL, Batzer MA. 2006. Emergence of primate genes by retrotransposon-mediated sequence transduction. Proc Natl Acad Sci 103: 17608–17613. doi:10.1073/pnas .0603224103 Zhou M, Smith AD. 2019. Subtype classification and functional annotation of L1Md retrotransposon promoters. Mob DNA 10: 14. doi:10.1186/ s13100-019-0156-5 Zingler N, Willhoeft U, Brose H-P, Schoder V, Jahns T, Hanschmann K-MO, Morrish TA, Löwer J, Schumann GG. 2005. Analysis of 5′junctions of human LINE-1 and Alu retrotransposons suggests an alternative model for 5′-end attachment requiring microhomology-mediated end-joining. Genome Res 15: 780–789. doi:10.1101/gr.3421505 Zoch A, Auchynnikava T, Berrens RV, Kabayama Y, Schöpp T, Heep M, VasiliauskaiteL, Pérez-Rico YA, Cook AG, Shkumatava A, et al. 2020. SPOCD1 is an essential executor of piRNA-directed de novo DNA methylation. Nature 584: 635–639. doi:10.1038/s41586-020-2557-5 Received April 25, 2023; accepted in revised form August 21, 2023. Locus-resolution L1 analysis in mouse development Genome Research 1481 www.genome.org