LINE1 promotes nuclear compartmentalization to repress the 8-cell state in embryonic stem cells

50 Understanding early human development is of fundamental interest and may provide new 51 applications in Reproductive Health and Regenerative Medicine. However, work in this field is 52 hindered by ethical questions and the low availability of early human embryos for research. Human 53 embryonic stem cells (hESCs) provide a powerful platform to model human development. 54 Depending on culture conditions, hESCs can capture the naïve state of the pre -implantation 55 blastocyst or the primed state of the post -implantation embryo. These two states can be 56 interconverted as well as used to differentiate cells towards specific post-gastrulation lineages. 57 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Reversion of naïve hESCs to earlier developmental stages has been more difficult to achieve. Or 58 particular interest is the 8 -cell (8C) state, the stage at which zygotic genome activation occurs 1. 59 Recently, several groups reported that 8C -like cells (8CLCs), which express a distinct 60 transcriptional program similar to the 8C embryo, exist at low frequency in cultures of naïve 61 hESCs2–5. These modified hESC culture conditions represent an unprecedented opportunity to 62 model and study human pre-implantation development. In support of this notion, the transcription 63 factor TPRX1 was shown to be essential both for development of human 8-cell embryos ex vivo6 64 and for generation of 8CLCs from hESCs 3. However, it remains unclear what molecular factors 65 prevent hESCs from reverting to the 8CLC state. 66 Studies in mouse have revealed several repressors of the 2-cell (2C) state7–10, the equivalent to 67 the human 8C state. We and others have shown that the expression of the largest family of 68 mammalian transposable elements (TEs) , long interspersed nuclear element 1 (LINE1), is 69 essential for mouse embryonic stem cell (mESC) self-renewal and pre-implantation 70 development7,11. We found that mouse LINE1 acts as a nuclear non -coding RNA that interacts 71 with Nucleolin (NCL) at chromatin to repress the mouse 2C program7,12 (reviewed in13). This study 72 left several important questions unanswered, notably how LINE1 RNA might regulate chromatin 73 organization to maintain the mESC state. Moreover, these findings raised the question of whether 74 and how LINE1 could play a conserved role in between mouse and human. Unlike unique protein-75 coding genes, TEs are typically not conserved in different mammalian genomes, but rather 76 different TE sub-families are lost and gained independently in different lineages over evolution14–77 16. This mobility and replacement of TEs makes them ideal drivers of genome evolution, as has 78 been well documented in mammals 15,17–19, but makes it less straightforward to envision TEs 79 regulating conserved processes such as development or physiology across species. 80 In this study, we set out to determine the function of LINE1 RNA in hESCs. Our findings support 81 a model whereby LINE1 RNA is essential for self -renewal and identity of hESCs, by preventing 82 reversion to the 8CLC state in a manner dependent on TPRX1. In parallel, our data uncover key 83 novel regulatory features of the human 8CLC state, including a role for nucleolar architecture and 84 a unique localization of chromosome 19 and a repressive function for PRC2. 85 86

87 LINE1 RNA regulates self-renewal and the transcriptional program of hESCs 88 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint We began by analyzing the expression of LINE1 subfamilies20 during human preimplantation 89 development21–23. The expression of recently evolved primate LINE1 subfamilies is detected in 90 oocytes and rises further after fertilization, with a peak at the inner cell mass (ICM) of the 91 blastocyst (Figures 1A, S1A-S1B and S1D). In contrast, older mammalian LINE1 subfamilies do 92 not display a clear pattern of expression during early human development (Figures S1C and S1E). 93 We then determined the expression of primate LINE1 in cultures of hESCs that recapitulate pre- 94 and post-implantation human development (Figures S1F-S1H)3,24–26. We found that LINE1 RNA 95 is expressed in hESC s in all culture conditions tested and is primarily localized to the nucleus , 96 with some cytoplasmic foci also observed (Figure S1H). 97 To investigate the function of LINE1 in hESCs, we designed an antisense oligo (ASO) targeting a 98 sequence conserved in the interORF region of primate and human-specific LINE1 families (ASO-99 L1) (Figure 1B; Table S1 ). We found that ASO-L1 induces efficient knockdown of LINE1 RNA 100 (L1KD) relative to a control non-targeting ASO (ASO-Ctr) or a sense version of the ASO-L1 (SO-101 L1) across different hESC culture conditions (Figures 1C-1D and S2A-S2C). Regardless of culture 102 conditions, k nockdown of LINE1 leads to loss of self -renewal of hESCs ( Figures S3A -S3F), 103 consistent with results observed in mESCs7. 104 To determine the impact of LINE1 KD on the transcriptome of hESCs, we carried out bulk RNA-105 sequencing (RNA-seq). Three different culture conditions of karyotypically normal (Figures S4A-106 S4B) H9 hESCs were tested, which capture the pluripotency continuum between naïve and 107 primed states (Figure S1F): mTeSR medium, to maintain a primed state 26, RSeT medium, to 108 maintain an intermediate naïve-like state24 (a RT-PCR validation in Figure S1G), and RSeT+DT, 109 where RSeT medium is supplemented for 48 hours with 10nM DZNep (3-Deazaneplanocin A, 110 “D”), an S-adenosyl-L-homocysteine hydrolase , and 5nM TSA (Trichostatin A , “T” ), a histone 111 deacetylase inhibitor. These low doses of DT were recently shown to promote a more naïve state 112 permissive for the emergence of 8CLCs at low frequencies3 (Figure S1F). These conditions have 113 the advantage s of relying on standardized commercial base media (mTeSR and RSeT) and 114 avoiding prolonged cultures in DT, which we found can lead to increased expression of 8C 115 markers and cell death over time (data not shown). Across all three hESC culture conditions 116 tested, LINE1 KD induces reproducible changes in gene expression relative to controls, as 117 determined by principal component analysis and hierarchical clustering (Figures 1E, 1G, S5A -118 S5E and S5G-S5H). 119 We next performed Gene Set Enrichment Analysis (GSEA) to compare the transcriptional 120 changes induced by LINE1 KD to gene sets that define the stages of human preimplantation 121 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint embryos from published data 4,22 (Figure S5M; Table S2). This analysis revealed that LINE1 KD 122 induces transcriptional signatures of the 8C and morulae stages , which are not normally 123 expressed in hESCs (Figures 1F and S5J -S5K). Hierarchical clustering confirmed that the 8C 124 gene expression program4 is induced upon LINE1 KD in hESCs (Figures 1G and S5G-S5H). 125 While derepression of the 8C signature is detected upon LINE1 KD across all hESC culture 126 conditions tested, this effect is more prominent in cells of a more naïve state (RSeT+DT > RSeT > 127 primed) (Figures 1F and S5J -S5L). Moreover, a short diagnostic set of 8C markers 2 is robustly 128 induced by LINE1 KD in RSeT+DT ( Figures 1H-1I, S5C and S5F ) but much less so in RSeT 129 (Figures S5D and S5I ), suggesting that a permissive chromatin environment induced by DT 130 facilitates 8C gene derepression upon LINE1 KD (Figures S1F and S5L ). Importantly, the 131 induction of diagnostic 8C markers upon LINE1 KD was validated using an independent ASO 132 targeting a conserved region in ORF2 (ASO-L1_ORF2), or using a different, XY H1 hESC line 133 (Figures 1I-1J). Moreover, we corroborated our findings using a CRISPR interference (CRISPRi) 134 system in hESCs generated via sleeping beauty transposition27. Transfection of gRNAs to target 135 a stably expressed dCas9-KRAB repressor protein to the LINE1 promoter region28,29 (Figure S2D; 136 Table S1) leads to knockdown of LINE1 RNA (Figure S2E) and induction of 8C marker gene s 137 (Figure 1K). Taken together, these results indicate that knockdown of LINE1 RNA via independent 138

consistently leads to the derepression of the 8C program in hESCs. 139 140 LINE1 KD induces the generation of 8CLCs 141 8CLCs are expected to be a rare subset of cells in naïve hESC cultures2–5. To determine whether 142 the upregulation of the 8C program upon LINE1 KD arises from the induction of 8CLCs, we carried 143 out 10X Genomics single-cell RNA-seq (scRNA-seq) on hESCs transfected with SO-L1 (control) 144 or ASO-L1 (LINE1 KD). A total of 5471 cells in control and 7629 cells in LINE1 KD were sequenced 145 and passed quality control , respectively ( Figures 2A and S6A -S6B). Uniform manifold 146 approximation and projection (UMAP) shows that a population of cells expresses 8C gene 147 signatures22,3, including key 8C markers such as TPRX1, ZSCAN4, DPPA3, ZNF280A and 148 SLC34A2 (encircled in Figure 2A and cluster 15 in Figure S6B). This cluster of cells co-expresses 149 TEs previously reported as markers of the 8C stage in vivo and 8CLCs in vitro, such as MLT2A1/2 150 and LTR7B2,3,5 (Figure 2B). As expected from 8CLCs, this cluster also expresses lower levels of 151 pluripotency markers such as POU5F1, SOX2 and NANOG (Figures 2C and S6C -S6D). Some 152 8C markers (such as TPRX1, DPPA3, and CCNA1) are consistently detected across all cells of 153 this cluster, while others (DUXA, KLF17, and ZSCAN4) are detected in only some of them (Figures 154 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 2A, and 2C). These differences may reflect cellular heterogeneity with in the 8CLC population or 155

of sequencing depth . Nevertheless, it is clear that the 8CLC cluster of cells is 156 essentially only observed upon LINE1 KD: while only 0.07% (4/5471) of control hESCs belong to 157 the 8CLC cluster, LINE1 KD induces a ~65-fold enrichment, to 4.95% (378/7629) of the population. 158 To directly assess the similarity of LINE1 KD hESCs to embryonic cells in vivo, we mapped our 159 scRNA-seq data onto published scRNA -seq of early human embryos 30. This analysis revealed 160 that coherent groups of cells in the LINE1 KD dataset, but not the control dataset, are projected 161 onto the cell clusters of embryonic day (E) 3 (8C) and E4 (morula) embryos (Figure 2D). 162 In addition to the induction of a distinct 8CLC cluster , there are other transcriptional changes 163 induced by LINE1 KD that potentially reflec t a more general transcription al deregulation. We 164 analyzed the bulk RNA -seq dataset of LINE1 KD hESCs for enriched molecular signatures 165 (Figures 1E -1H, S5C and S5F) and projected them onto the scRNA -seq data. This analysis 166 revealed that cell clusters induced by LINE1 KD are enriched for genes upregulated upon 167 silencing of the pluripotency factors POU5F1 or SOX2 (Figures S6E-S6F), signatures of s tress 168 such as the p53 pathway (Figures 2E-2F and S6G ), and targets of the Polycomb Repressive 169 Complex 2 (PRC2), which deposits H3K27me3 (Figures 2E, 2G and S6H-S6I, see below). 170 To independently assess the emergence of 8CLCs in hESC cultures, we used 171 immunofluorescence (IF) to assess the expression of TPRX1 protein, a diagnostic marker of 172 8CLCs2,3. In H9 hE SCs, LINE1 KD induces an increase in TPRX1+ cells from ~0.8% to ~3.6% 173 (Figures 3A-3B); in H1 hES Cs, LINE1 KD induces an increase from ~0.2% to ~4.2% ( Figures 174 S7A-S7B). These percentages roughly parallel those derived from the scRNA -seq data. Taken 175 together, these results indicate LINE1 KD induces the emergence of 8CLCs in hESCs. 176 177 The induction of the 8C program upon LINE1 KD is mediated by TPRX1 178 We next sought to identify genes that may mediate the induction of the 8C program upon LINE1 179 KD. Several studies have suggested that the transcription factor DUX4 is a master regulator of 180 the human 8C program 21,31,32, similar to the role of its homolog Dux in mouse21,33,34. In mESCs, 181 the induction of the 2C program upon LINE1 KD is Dux-dependent7. However, we found that 182 DUX4 and DUX4-like genes are essentially undetectable in our RNA -seq data and are not 183 induced by LINE1 KD (Figure S7C). This is in contrast to TPRX1 (see above, Figures 1, 2, and 184 S7D), a known regulator of human 8CLCs in vitro 3 and 8C development in vivo 6. In addition to 185 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint TPRX1, we considered a potential role for H3.XY, a histone variant that is a marker of the 8C 186 state2 induced by LINE1 KD (Figures 1I-1K) which mediates the induction of DUX4 target genes 187 in human muscle cells 35. Finally, we considered TP53, as it has been linked to stress -induced 188 activation of 2C/8C genes in mouse and human cells36,37 and p53 pathway genes are induced by 189 LINE1 KD (Figures 2E-2F). 190 To test the potential regulatory roles of these candidate genes, we carried out ASO -mediated 191 LINE1 KD with or without simultaneous KD of TPRX1, H3.XY or TP53 using siRNAs (Figure 3C). 192 We validated that TPRX1, H3.XY and TP53 levels are knocked down by their corresponding 193 siRNAs ( Figures S8A-S8D). RNA-seq revealed that KD of TPRX1, but not H3.XY or TP53, 194 partially suppresses the induction of the 8C program upon LINE1 KD (Figures 3D-3G and S8E-195 S8G). In particular, the expression of key diagnostic markers of 8CLCs2, including KDM4E, 196 SLC34A2, ZSCAN4, RFPL2, DPPA3 and KLF17 (apart from TPRX1 itself), is reverted towards 197 control levels in LINE1 KD cells only in the case of TPRX1 RNAi (Figures 3E-3F). Note that there 198 are transcriptional changes induced by LINE1 KD not related to 8C program that are not rescued 199 by TPRX1 RNAi (Figures 3D-3E). We cannot exclude the possibility that RNAi did not lead to 200 sufficient KD of H3.XY or TP53. Moreover, there are likely other mediators of the transition to 201 8CLCs, in addition to TPRX1. Nevertheless, our findings indicate that the induction of the 8C 202 program upon LINE1 KD is partially mediated by derepression of TPRX1. 203 204 Lamina- and nucleolus-associated domains are enriched for LINE1 and reveal a dynamic 205 localization of 8C loci 206 In mouse 2CLCs, the activation of Dux coincides with its relocalization from the nucleolar 207 periphery to the nucleoplasm 7,38,39, although the underlying mechanisms remain poorly 208 understood. We therefore sought to understand the dynamics of nuclear compartmentalization, 209 including the localization of LINE1 and 8C genes, in hESCs. We carried out sequencing of NADs 210 (using nucleolar fractionation-seq40) and LADs (using LAMB1 -DamID-seq41,42) in hESCs (Figure 211 4A). Both of these compartments are associated with heterochromatin -mediated gene 212 repression43,44. While NADs have not previously been sequenced in hESCs, our LAD -seq data 213 are highly consistent with a previous report 45. We found that human LINE1 loci are enriched in 214 NADs and LADs in hESCs (Figures S9A-S9B), in agreement with data in mESCs12,46. Interestingly, 215 we found that LINE1 RNA accumulates in nuclear foci that are frequently located at the vicinity of 216 the nucleolus and lamina, in contrast to a lower density in the intervening nucleoplasm (Figures 217 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 4B and S9C -S9D). Moreover, cross -linking and RNA immunoprecipitation (CLIP) -qRT-PCR 218 revealed an interaction between Nucleolin (NCL) and Laminin B1 (LMNB1) proteins with LINE1 219 RNA (Figure S9E). In addition, we found that LADs are essentially constant between naïve -like 220 and primed hESCs, while NADs are less abundant in naïve-like hESCs and expand in signal and 221 size in primed hESCs ( e.g., see Figures S1 2A). These results suggest that the nucleolar 222 compartment may be particularly dynamic during early human development (see below). 223 The NAD and LAD data further predicted an enrichment of 8C gene loci like TPRX1 and H3.XY 224 in the nucleolus or the lamina (Figure S10A and S11A). By DNA-FISH, we found that TPRX1 loci 225 are localized in the nucleoplasm in 8CLCs, but display increasing association with the nuclear 226 lamina and the nucleolar periphery with developmental progression of hESCs ( Figures 4C and 227 S10B-S10D; Videos S3-S6). In contrast, H3.XY (Chr. 5, H3.Y1 and H3.Y2, see Figure S11A) is 228 located at the nuclear lamina across all hESC states, with only ~20% of H3.XY loci relocated to 229 the nucleoplasm in 8CLCs (Figures 4D and S11B-S11C). We find DUX4 to be mostly localized to 230 the nuclear lamina in naïve hESCs (Figures S11D-S11E), which stands in contrast to the nucleolar 231 association of its homolog Dux in mESCs7,39. Taken together, the LAD-seq and NAD-seq data, as 232 well as the imaging data for key loci, reveal a progressive association of 8C genes with both the 233 nucleolar periphery and the nuclear lamina, from 8CLCs to more developmentally advanced 234 stages of hESCs. 235 We next explored whether LINE1 KD impacts the 3D nuclear loca lization of specific 8C loci . 236 Interestingly, we found that in LINE1 KD hESCs a lower proportion of H3.XY loci are located at 237 the lamina (~72% vs ~85% in control cells), and their overall distance from the lamina increases 238 (Figure 4E-4G). These results suggest that LINE1 RNA promotes repression of H3.XY genes in 239 part by contributing to tethering of their loci to the nuclear lamina. 240 In contrast, we detected no significant change in the localization of TPRX1 loci upon LINE1 KD 241 (Figure S10E-S10F). The observation that only a minor percentage of TPRX1 loci are located 242 near the lamina or the nucleolus in naïve and primed hESCs (Figure 4C) complicates a 243 quantification of the impact of LINE1 KD. Moreover, the fact that TPRX1 is repressed in naïve and 244 primed hESCs despite its predominant nucleolplasmic localization indicates that mechanisms 245 other than 3D nuclear positioning may contribute to TPRX1 silencing in hESCs. 246 247 Chromosome 19 is enriched for genes induced upon LINE1 KD and preferentially 248 associates with the nucleolus 249 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint We noticed that TPRX1, as well as several other key markers of the 8C state (e.g., TPRX2, 250 ZSCAN4, DUXA or LEUTX) are all located on Chr. 19 ( Figure 5A). We therefore analyzed the 251 chromosomal distribution of genes differentially expressed in LINE1 KD hESCs. Intriguingly, Chr. 252 19 displays an enrichment for genes upregulated upon LINE1 KD ( Figure 5B), across all hESC 253 states tested (RSeT+DT, RSeT and Primed). Moreover, analysis of the LAD/NAD -seq data 254 revealed that, while most chromosomes display enrichments for both compartments, Chrs. 19 255 and 22 stand out as being enriched for NADs but not LADs in both naïve and primed hESC s 256 (Figures 5C and S12B). This observation is supported by data in differentiated cells showing that 257 Chr.19 occupies an internal nuclear territory closely associated with the nucleolus47,48. 258 The results above prompted us to explore the 3D nuclear positioning of Chr. 19 in 8CLCs vs 259 hESCs. We carried out chromosome painting in enhanced 4CL medium with high levels of DT 260 (“e4CL”), which was recently shown to enrich for 8CLCs, compared to naïve cells in basal 4CL 261 medium3. The results reveal a remarkably distinct association between Chr. 19 and the nucleolus: 262 while in naïve cells Chr. 19 is intimately associated with the nucleolus, as would be predicted from 263 the NAD data (Figures 5C and S12), in 8CLCs there is a significantly higher separation between 264 the two (Figures 5D-5F; Videos S1 and S2). This differential co-localization seems to be in part 265 related to a much rounder morphology of the nucleolus in 8CLCs. Thus, 8CLCs appear to be the 266 exception to the rule of an intimate association between Chr. 19 and the nucleolus across cell 267 types tested. The molecular mechanisms that underlie the peri -nucleolar localization of Chr. 19, 268 and that presumably are not yet in place in 8CLCs, remain unknown (see Discussion) . 269 Nevertheless, these results led us to focus our attention on the nucleolus for the remainder of this 270 study. 271 272 LINE1 RNA is required for maintenance of nucleolar architecture, which is critical to 273 suppress the 8C program 274 As mentioned above, 8CLCs tend to display a rounder nucleolar morphology that is remarkably 275 distinct from the irregular shape of the nucleolus typical of other hESC states and non-pluripotent 276 cells (Figures 5D and S10B; Videos S1-S6). This nucleolar structure in 8CLCs resembles what is 277 observed in vivo49. Early human pre-implantation embryos have round nucleoli up to 4-8 cell stage; 278 from then onwards , the nucleolus becomes more irregular in shape and accumulates 279 heterochromatin at the periphery 49. We found that i nhibition of RNA Pol I with 0.25 µM BMH in 280 RSeT+DT hESCs induces a rounder nucleolar morphology by 8 hours ( Figure S13A) and a 281 striking up-regulation of 8C marker genes by 24 hours (Figure 6A). These results are in agreement 282 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint with data showing that a functional nucleolus is required for repression of the 2C state in 283 mESCs38,39. Thus, the nucleolus plays a conserved role in the repression of totipotency-284 associated gene expression programs in mouse and human ESCs. 285 We next assessed the role of LINE1 in nucleolar architecture. Normal nucleoli are organized with 286 an interior dense fibrillar component marked by Fibrillarin (FBL), encased by an exterior granular 287 component marked by NCL 50. We found that LINE1 KD hESCs display a disruption of this 288 organization and other signs of nucleolar stress 51,52. In LINE1 KD hESCs, the FBL domain 289 frequently extrudes onto the outside of the NCL domain , forming what are known as nucleolar 290 caps (Figures 6B and S13B). In addition, LINE1 KD hESCs display abnormally high levels of FBL 291 (Figure 6B-6C and S13B), which have been reported for nucleolar stress in human cells52. LINE1 292 KD leads to a disruption of nucleolar morphology than is overall distinct from the “circularization” 293 observed upon inhibition of RNA Pol I (Figure S13A). These results indicate that LINE1 RNA is 294 essential to globally maintaining a mature nucleolar architecture in hESC cultures, which in turn 295 prevents a fraction of the cells from reverting to the 8CLC state. 296 To gain further insight into factors that might regulate exit from the 8C state, we examined the 297 expression of various heterochromatin repressors and nucleolar proteins during early human 298 development. Interestingly, this analysis revealed that core components of the PRC2 complex 299 (EZH2, EED, SUZ12), as well as key nucleolar proteins (NCL and FBL ), are highly induced at 300 cleavage stages (Figures 6D and S14A-S14C), consistent with the induction of primate LINE1 301 subfamilies (Figures 1A, S1A-S1B and S1D). This contrasts with to PRC1 complex members or 302 factors involved in H3K9 methylation, which are either not detected or show the oppositive pattern 303 of expression during early human development ( Figures 6D, S14A and S14D-S14G). Together 304 with the enrichment of genes induced upon LINE1 KD for targets of PRC2 (Figures 2E, 2G and 305 S6H-S6I), these results led us to explore a potential role for PRC2 in 8C gene regulation. 306 Remarkably, inhibition of PRC2 using UNC1999 (but not its inactive analog UNC2400) for just 307 48h induces a strong up -regulation of 8C genes, even in the absence of DZNep in the culture 308 medium ( Figure 6E). In parallel, CRISPRi-mediated repression of NCL also up -regulates the 309 expression of key 8C marker genes (Figures 6F and S13C-S13D). Taken together, these findings 310 indicate that nucleolar maturation promoted by LINE1 RNA and NCL, possibly acting in 311 combination with PRC2-mediated gene repression, prevents reversion of hESCs to the 8C state. 312 313

314 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint The regulation of the human 8C state remains very poorly understood. The transcription factor 315 TPRX1, part of a larger family of fast -evolving homeobox transcription factors 53,54, does not 316 appear to play a role in mouse but is essential for the generation of 8CLCs 3 in hESCs and 317 development of human 8C embryos6. Here we show that LINE1 is an essential regulator of hESC 318 self-renewal and identity by preventing developmental reversion to the 8CLC state. We found that 319 LINE1 RNA is associated with the nuclear lamina and the nucleolus and is required for the 320 maintenance of nucleolar architecture, acting as a barrier to the expression of TPRX1 and the 321 activation of the 8C program in hESCs ( Figure 6G). In addition, our study provides new insights 322 into chromosome 19 territory dynamics and a key organizing role for the nucleolus within the 323 totipotency-to-pluripotency continuum of hESCs. Our analyses further point to a key role for PRC2 324 in repression of the 8C program. Taken together, our results indicate that LINE1, rather than being 325 irrelevant junk DNA55 or a genomic parasite detrimental to cell survival56, coordinates 3D nuclear 326 organization and chromatin-mediated gene repression in hESCs. 327 Our findings indicate that a fundamental transition at the onset of development is regulated in 328 both mouse7 and human (this study) by the expression of LINE1 elements. We speculate that the 329 massive expansion of LINE1 in therian mammals 57 may have contributed to its role in nuclear 330 compartmentalization and early development. Notably, the partial similarities between the roles 331 of LINE1 in mouse and human ESCs are observed despite overall replacement of LINE1 sub -332 families and lack of syntheny in genomic locations of their elements across mammals 14–16,58,59. 333 LINE1 elements in both the mouse and human genome are located in AT -rich, gene poor areas 334 that tend to be silenced and enriched in LAD and NAD compartments 12,46,58,59. The AT -rich 335 consensus sequence of the LINE1 endonuclease (5’ -TTTT/AA-3’ and derivatives) 60,61 and the 336 selection for maintenance of LINE1 insertions in heterochromatin areas and/or against their 337 retention in gene -rich euchromatic areas 62,63 may contribute to preserving a role of LINE in 338 genome compartmentalization. It will be of interest to study potential developmental roles of 339 LINE1 and its distribution across nuclear compartments in other mammalian species. 340 This study points to a remarkable remodeling of the nucleolar compartment in the transition 341 between 8CLCs and naïve hESCs. We speculate that the strong induction of key nucleolar factors 342 at the 4-8C stage (Figures 6D and S14) may support the maturation of the nucleolus that occurs 343 at these stages 49. The association of LINE1 RNA with newly synthesized NCL protein may 344 facilitate liquid -liquid phase separation 64,65 and/or maturation of the nucleolus. In turn, the 345 nucleolar periphery may then become a hub for recruitment of chromosomal domains to be 346 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint repressed, including large parts of Chr. 19, and heterochromatin -mediated silencing of 8CLC 347 regulators such as TPRX1 (Figure 6G). 348 Our findings raise several challenging questions for future inquiry: i) what mediates the 349 preferential association of Chr. 19 with the nucleolus in hESCs and other human cells 47,48, and 350 what makes this association weaker in 8CLCs? The link to LINE1 could well be indirect via a role 351 of LINE1 in nucleolar maturation, as Chr. 19 has an overall lower frequency of LINE1 elements 352 relative to the rest of the genome 66; ii) How is the 3D nuclear compartmentalization of hESCs 353 coordinated with chromatin-level repression of the 8C program? Our results indicate that PRC2 354 plays a key role in silencing the 8C program, and PRC2 has been shown to establish long-range 355 chromatin interactions (e.g.67–69). Whether PRC2 directly represses genes like TPRX1 or acts via 356 some other mechanism remains to be determined; iii) is the coordinate expression of LINE1 from 357 its various loci across most/all chromosomes required for nuclear compartmentalization and 8C 358 program repression? The repeated nature of LINE1 in the genome and the association of its RNA 359 at chromatin with repressive factors, such as KAP17, YTHDC1/SETDB170 or the HUSH complex71 360 makes it an ideal genomic feature to coordinately silence gene expression programs. Alternatively, 361 it is possible that specific LINE1 elements in unique chromosomal locations are critical for 362 nucleolar organization and suppression of the 8C program. Moreover, it is worth keeping in mind 363 that LINE1 RNA as well as some 8C loci such as H3.XY are also enriched at the nuclear lamina. 364 In this regard, recent findings point to a remarkably dynamic nuclear lamina in mouse embryos at 365 the 2C stage that may regulate chromatin organization at ZGA 72. A systematic analysis of the 366 unique landscape of nuclear compartmentalization of 8CLCs may enable the dissection of the 367 role of activators (such as TPRX1) and repressors (such as LINE1) of the 8C state. 368 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 369 370 371 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 1 LINE1 RNA is a barrier to the 8C program in hESCs. 372 (A) Averaged normalized counts of n=2 independent RNA-seq samples21 showing induction of 373 human- and primate-specific LINE1 subfamilies (L1HS and L1PA2 to L1PA8) during human 374 cleavage embryonic stages. GV, germinal vesicle; MI, metaphase I; MII, metaphase II; PN, 375 pronuclear stage; CL, cleavage; MOR, morula; ICM, inner cell mass; TROPH, trophectoderm. 376 (B) Schematic of LINE1 RNA knockdown experiments in hESCs (see Figure S1F for hESC 377 culture conditions used in this study). ASOs targets (green arrows) at the inter-ORF and ORF2 378 sites, and CRISPRi gRNA targets (purple arrows) at the 5’-UTR of full-length LINE1 elements 379 are indicated. Cells were collected after 48 hours of transfection. 380 (C) LINE1 RNA FISH in RSeT hESCs showing predominantly nuclear localization of LINE1 RNA 381 and efficient knockdown in ASO-L1 transfected cells compared to ASO-Ctr and SO-L1 controls 382 after 48 hours of transfection. Representative of two independent experiments. Scale bar, 10 383 µm. 384 (D) Quantification of LINE1 RNA-FISH signal intensity in (C). LINE1 intensity in the ASO-L1 385 transfected cells is significantly reduced compared to the controls. Representative of three 386 independent experiments. See Extended Data 2a-b for validation of LINE1 KD in RSeT+DT and 387 primed hESCs. Mann-Whitney test. ns = p > 0.05, ** = p < 0.01. 388 (E) Multidimensional scaling (MDS) plot of genes across all samples, showing that ASO-L1 KD 389 RSeT+DT hESCs have distinct gene expression profiles from SO-L1 and ASO-Ctr samples. 390 (F) Gene Set Enrichment Analysis of the transcriptional profile of ASO-L1 KD RSeT+DT hESCs 391 compared to different stages of peri-implantation development and 8C/pre-8C gene sets4,22. 392 LINE1 KD hESCs display a significant enrichment for the 8-cell (plotted in the lower panel) and 393 morulae programs. NES, Normalized Enrichment Score. 394 (G and H) Heatmap showing induction of diagnostic gene sets of the 8C stage in vivo from X. 395 Yu et al.4 (G) and Taubenschmid-Stowers et al.2 (H) in RSeT+DT hESCs upon ASO-L1KD 396 compared to ASO-Ctr and SO-L1 controls. 397 (I) qRT-PCR validation of upregulation of 8C genes in LINE1 KD RSeT+DT H9 hESCs with 398 ASOs targeting the inter-ORF and ORF2 sites, respectively. Data are mean ± SEM, n = 6 399 biological replicates. Ratio paired Student’s t-tests. ns = p > 0.05, * = p < 0.05, ** = p < 0.01, *** 400 = p< 0.001, **** = p < 0.0001. 401 (J) qRT-PCR validation of upregulation of 8C genes in LINE1 KD RSeT+DT H1 hESCs with 402 ASOs targeting the inter-ORF site. Data are mean ± SEM, n = 5 biological replicates. Ratio 403 paired Student’s t-tests. ns = p > 0.05, * = p < 0.05, ** = p < 0.01, *** = p< 0.001. 404 (K) qRT-PCR validation of upregulation of 8C genes in LINE1 KD RSeT+DT H9 hESCs with 405 CRISPRi targeting LINE1. Data are mean ± SEM, n = 5 biological replicates. Ratio paired 406 Student’s t-tests. ** = p < 0.01, *** = p< 0.001, **** = p < 0.0001. 407 408 409 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 410 411 412 413 414 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 2 LINE1 KD induces the generation of 8CLCs. 415 (A) UMAP visualization of scRNA-seq data from LINE1 KD and control RSeT+DT hESCs. The 416 top left panel displays ASO-L1 KD and SO-L1 control cells as red and blue dots, respectively. 417 The circled population (cluster 15 in Figure S6B) represents LINE1 KD-induced 8CLCs 418 (ASO_8CLCs) expressing 8C gene sets annotated from 8CLCs3 and 8C embryos22; individual 419 8C marker genes are also indicated. 420 (B) Violin plot showing the log-normalized expression of representative 8C-enriched TEs 421 (MLT2A1, MLT2A2 and LTR7B) in human E3 to E7 stages, ASO-L1 KD-induced 8CLCs 422 (ASO_8CLCs, circled in (A), upper left panel), and total hESCs transfected with ASO-L1 or SO-423 L1 (red and blue dots in (A), upper left panel). 424 (C) Bubble plot representing the frequency of expression and average expression of 425 representative pluripotency and totipotency genes in early human embryonic stages30 , ASO-L1 426 KD-induced 8CLCs (ASO_8CLCs), e4CL induced 8CLCs (e4CL_8CLCs)3, RSeT+DT control 427 hESCs (SO-L1 transfected), 4CL hESCs and primed hESCs. 428 (D) UMAP visualization of scRNA-seq data from early human embryos (E3 to E7)30. To the right, 429 data from ASO-L1 or SO-L1 transfected hESCs are projected onto the in vivo data. Purple and 430 orange arrows point to corresponding E3 and E4 cell clusters, respectively. 431 (E) Top chromatin-bound factors enriched at genes upregulated upon L1KD in RSeT+DT 432 hESCs (see Methods). 433 (F) UMAP view of expression levels of genes of the p53 pathway. 434 (G) UMAP view of expression levels of H3K27me3-marked genes (left) and targets of EZH2 435 (right) in hESCs. 436 437 438 439 440 441 442 443 444 445 446 447 448 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 449 450 451 452 453 454 455 456 457 458 459 460 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 3 The induction of the 8C program upon LINE1 KD is mediated by TPRX1 461 (A) Immunostaining showing an increase in TPRX1-positive cells in RSeT+DT H9 hESCs upon 462 ASO-L1KD compared to controls. Representative of three independent experiments. Scale, 100 463 µm. 464 (B) Quantification of (A) for percentage of TPRX1+ cells in each indicated sample. 1000~1500 465 cells per sample from six randomly scanned images of two independent experiments are 466 quantified. Data are mean ± SEM. Welch’s t-test. * = p < 0.05, ** = p < 0.01. Representative of 467 three independent experiments. 468 (C) Schematic of LINE1 RNA knockdown, co-transfected with siRNAs targeting TPRX1, TP53, 469 and H3.XY, compared with a control non-targeting (NT) siRNA. Cells were collected after 48 470 hours of transfection. 471 (D and E) MA plot showing log2 fold change in gene expression following (E) L1 KD+NT-siRNA 472 and (F) L1 KD+TPRX1-siRNA compared to control transfected cells (ASO-Ctr+NT-siRNA). Red 473 or blue highlight genes of adj.P-value 1.5 or < -1.5, respectively. 8C marker 474 genes from Taubenschmid-Stowers et al.2 are labeled in dark blue. 475 (F) Heatmap showing a suppression in the induction of 8C marker genes when LINE1 KD is 476 combined with TPRX1 RNAi but not with TP53 or H3.XY RNAi. 477 (G) Violin plot of the same data as in (F). TPRX1 RNAi, but not TP53 or H3.XY RNAi, 478 significantly suppresses the induction of 8C marker genes that is observed upon LINE1 KD, 479 relative to an random gene set of equal size. Wilcoxon test, * = p < 0.05, **** = p < 0.0001. 480 481 482 483 484 485 486 487 488 489 490 491 492 493 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 494 495 496 497 498 499 500 501 502 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 4 Lamina- and nucleolus-associated domains are enriched for LINE1 and reveal a 503 dynamic localization of 8C loci. 504 (A) Schematic of LAD-seq and NAD-seq in hESCs. See Methods for details. 505 (B) Representative image of LINE1 and 45s-rRNA RNA-FISH and co-IF staining for LBR 506 antibody in RSeT+DT hESCs, showing that LINE1 RNA are enriched in the vicinity of nucleolar 507 and laminar domains. The max projected image (two top left panels) displays high expression of 508 LINE1 in the nucleus. A representative single-plane is shown in the two top right panels, with 509 boxed areas (* and **) enlarged in the middle and bottom panels, with color channels separated. 510 Yellow dotted line encircles the nucleolus area. Yellow and red arrows point to LINE1 RNA foci-511 enriched spots at the nucleolar and laminar domains, respectively. Noted the nucleoplasm (np) 512 areas are sparse in LINE1 foci. Representative of at least three independent experiments. Scale 513 bar, 10 µm. 514 (C) Distribution of the location of TPRX1 loci across the nuclear compartments listed. See also 515 Figures S10. n = cells /TPRX1 loci quantified. 516 (D) Distribution of the location of H3.XY loci across the nuclear compartments listed. See also 517 Figures S11. n = cells /H3.XY loci quantified. 518 (E) Representative image of H3.XY DNA-FISH and and co-IF staining for B23 and LMNB1 in 519 control and LINE1 KD RSeT+DT hESCs. H3.XY loci localize further away from the lamina in 520 LINE1 KD cells. The boxed area in the max projected images (top panels) is enlarged and 521 displayed as representative z-stack (Z) images (bottom panels). The yellow arrows point to 522 H3.XY loci in the nucleoplasm. Representative of three independent experiments. Scale bar, 10 523 µm. 524 (F) Quantification of data from H3.XY DNA-FISH and co-IF staining (shown in E), plotting the 525 distance of H3.XY loci to the lamina domain. If the distance is < 0.5 µm, it is defined as within 526 LADs. Number of cells quantified in each group is indicated in (G). Mann-Whitney test. * = p < 527 0.05. Data are from three independent experiments. 528 (G) Distribution of the location of H3.XY loci across the nuclear compartments listed in control 529 and LINE1 KD RSeT+DT hESCs based on the distance quantification in F. n=cells /H3.XY loci 530 quantified. Data are from three independent experiments. 531 532 533 534 535 536 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 537 538 539 540 541 542 543 544 545 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 5 Chromosome 19 is enriched for genes induced upon LINE1 KD and 546 preferentially associates with the nucleolus. 547 (A) Ideogram of chromosome 19 with the location of several 8C genes indicated. 548 (B) Distribution of protein-coding genes upregulated (red bars) or downregulated (blue bars) 549 upon LINE1 KD as percentage of total number of genes in each chromosome. Chr. 19 shows a 550 bias towards genes upregulated vs downregulated in comparison to other chromosomes, in all 551 hESC culture condition used. 552 (C) Representative genome browser views of LAD and NAD enrichment signatures of Chr. 19, 553 Chr. 1, Chr. 13, and Chr. X. Chr. 19 displays a distinct enrichment of NAD but not LAD 554 throughout the chromosome. 555 (D) Example still images from videos (Videos S1 and S2) of Chr. 19 DNA-FISH and co-IF 556 staining for NCL antibody, showing that Chr.19 is less associated with the nucleolus in H3,XY+ 557 8CLCs than in naïve hESCs, cultured in e4CL and 4CL3, respectively. 3D reconstructed 558 fluorescence and surface data are shown, see Methods for details. Representative of two 559 independent experiments. Scale bar, 20 µm. 560 (E and F) Quantification of Chr.19 DNA-FISH and NCL IF co-staining (as shown in D) indicates 561 a significantly higher overlap between these two territories in naïve hESCs than in 8CLCs 562 (whether H3.XY+ or H3.XY-). Data are from two replicated independent experiments. Number of 563 cells quantified in each group is indicated. Mann-Whitney test. * = p < 0.05, ** = p < 0.01. 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 580 581 582 583 584 585 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure 6 LINE1 RNA promotes maintenance of nucleolar architecture, which suppresses 586 the 8C program. 587 (A) qRT-PCR data showing strong upregulation of 8C genes following inhibition of RNA Pol I in 588 hESCs using 0.25 µM BMH for 24 hours. Data are mean ± SEM, n = 7-13 biological replicates. 589 Ratio paired Student’s t-tests. * = p < 0.05, *** = p< 0.001, **** = p < 0.0001. 590 (B) NCL and FBL IF in control and LINE1 KD hESCs. In the LINE1 KD cells, FBL extrudes from 591 the NCL territory. Representative of two independent experiments. Scale bar, 10 µm. 592 (C) Quantification of the FBL IF images (as shown in B). LINE1 KD cells have significantly 593 higher levels of FBL than control hESCs Data are from two independent experiments. Mann-594 Whitney test. **** = p < 0.0001. 595 (D) Heatmap representation of the expression of select heterochromatin and nucleolus factors 596 during early human embryonic development, profiled using bulk RNA-seq21. See Figure S14 for 597 similar scRNA-seq data. Nucleolar proteins and PRC2 subunits are among the genes 598 upregulated (bolded) at cleavage stages (CL). 599 (E) qRT-PCR data showing upregulation of 8C genes following inhibition of EZH2 in RSeT+DT 600 or RSeT+TSA hESCs using 2.5 µM UNC1999 for 48 hours, compared to RSeT+DT alone or 601 RSeT+DT+UNC2400 controls. Data are mean ± SEM, n = 4 biological replicates. Ratio paired 602 Student’s t-tests. * = p < 0.05, ** = p < 0.01, *** = p< 0.001. 603 (F) qRT-PCR data showing upregulation of 8C genes upon knockdown of NCL by CRISPRi. 604 Data are mean ± SEM, n = 5-7 biological replicates. Ratio paired Student’s t-tests. * = p < 0.05, 605 ** = p < 0.01. 606 (G) Model for the maturation of NADs and LADs, and associated 3D nuclear redistribution of 607 Chr. 19 and key 8C loci between 8CLCs and naïve hESCs. Human LINE1 RNA promotes 608 maintenance of nucleolar architecture and repression of the 8C program. See text for details. 609 610 611 612 613 614 615 616 617 618 619 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint

620 Primed hESC culture 621 H9 and H1 primed hESC lines73 were cultured in mTeSR Plus Medium (STEMCELL # 05825) on 622 hESC-Qualified Matrigel -(VWR # CA89050 -192) coated plates at 37 °C, 5% CO2, 20% O2 623 conditions26. Cells were passaged every 4 -5 days by performing a PBS wash, adding then 624 removing ReLeSR (STEMCELL # 05873) and incubating at 37 °C for 5 min. mTeSR medium was 625 then added and the cell plate was tapped gently to release colonies. Without pipetting up and 626 down, the cell suspension was dropped into new pre -Matrigel-coated wells. When single -cell 627 seeding was required, the cell suspension was pipetted up and down, and cells were counted and 628 seeded at the desired numbers with the addition of 20 µM ROCK inhibitor Y -27632 (Stemgent # 629 S1049). 630 RSeT hESC culture 631 RSeT hESCs were cultured in RSeT Medium (STEMCELL # 05975) at 37°C, 5% CO2, and 5% 632 O2 conditions24. Primed hESCs were converted to RSeT conditions and passaged when colony 633 sizes reached ~100 µm, following the RSeT Medium manufacturer’s instructions. RSeT hESCs 634 were used for experiments within ten passages of conversion. 635 For RSeT+DT conditions, RSeT hESCs were supplemented with 10 nM DZNep (Selleck # S7120) 636 and 5nM TSA (Vetec # V900931) for 48 hours, with the medium changed at 24h. Similarly, for 637 RSeT+eDT conditions, RSeT hESCs were supplemented with 50 nM DZNep and 20 nM TSA as 638 described3 for 48 hours. 639 MEF culture 640 Primary MEFs (passage 3) were obtained from the Lunenfeld-Tanenbaum Research Institute 641 (LTRI) ESC Facility . About 2 x 10^5 MEFs were thawed on 0.1% gelatin -coated plates and 642 cultured in MEF medium (DMEM, high glucose, GlutaMAX ™ Supplement, pyruvate (Life 643 Technologies # 10569), 15% FBS (BioTechne # S12450), 0.1 mM MEM nonessential amino acids 644 (Thermo Fisher Scientific # 11140050) and 100 U/mL penicillin -streptomycin (Thermo Fisher 645 Scientific # 15140122)). MEF feeders were used within 4-5 days of thawing. 646 4CL and e4CL hESC culture 647 4CL hESCs were converted from primed hESCs according to Mazid et al. 3. Briefly, 1-1.5 x 10^5 648 primed hESCs were seeded as single cells per well of a 6-well plate pre-seeded with MEF feeder 649 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint cells in mTeSR medium supplemented with 10 µM ROCK inhibitor Y-27632 at 37 °C in 20% O2. 650 After 24 hours, medium was switched to 4CL [1:1 mix of Neurobasal medium (Gibco #21103049) 651 and Advanced DMEM/F12 (Gibco #12634028) supplemented with N2 (Gibco # 17502048) and 652 B27 (Gibco # 17504044) supplements, sodium pyruvate (Life Technologies # 11360070), non -653 essential amino acids (Life Technologies # 11140050L), GlutaMAX (Gibco # 35050061), penicillin-654 streptomycin (Life Technologies # 15140122), 10 nM DZNep (Selleck # S7120), 5 nM TSA (Vetec 655 # V900931), 1 µM PD0325901 (Stemolecule # 04-0006), 5 µM IWR-1 (Sigma # I0161), 20 ng/ml 656 human LIF (STEMCELL # 78055), 20 ng/ml ACTIVIN A (Peprotech # 120 -14E), 50 µg/ml L -657 ascorbic acid (Sigma # A8960) and 0.2% (v/v) Geltrex (Life Technologies # A1413301)] and kept 658 at 37°C in 5% O2. The medium was changed daily. When colony size s reached ~100 m, cells 659 were incubated in a 1:1 mix of TypleE:0.5 mM EDTA for 5 min treatment. Without tapping plates, 660 the TypleE:EDTA mixture was replaced with 1 ml 4CL media. After tapping the plate and pipetting 661 up and down, single cells were reseeded on Geltrex -(Life Technologies # A1413301) pre-coated 662 or MEF-pre-seeded 6-well plates. 10 µM Y-27632 was added for 24 h the day after each passage. 663 Cells were used for experiments within 3-5 passages. 664 To convert 4CL cells to e4CL cells, 2-3 x 10^5 per 6-well 4CL cells were seeded in 4CL medium, 665 supplemented 24 h later with 50 nM DZNep and 20 nM TSA 3 (e4CL). After 5 days of culturing in 666 e4CL medium, cells were processed for analysis. 667 dCas9-KRAB hESC line 668 Plasmids were built based on the sleeping -beauty transposon system27, as indicated in Figure 669 S2D using enzymes from NEB. In a first step, the amaxaGFP sequence of pT2-CAG-amaxaGFP 670 (a gift of the Izsvák lab) was removed using the AgeI and BglII sites and replaced with an artificial 671 oligo sequence containing multiple enzyme digestion sites (FW - 672 ccggtGGCGCGCCCGGGTGATCAa; RV- tGGCGCGCCCGGGTGATCAagatc). Next, AgeI and 673 NotI sites were used to clone in the hUbC -dCas9-KRAB-T2a-Puro sequence from the pLV-hU6-674 sgRNA-hUbC-dCas9-KRAB-T2a-Puro plasmid 29 (Addgene # 71236), generating pT2 -hUbC-675 dCas9-KRAB-T2a-Puro such that the insert is now flanked by sleeping beauty inverted/direct 676 repeats. 677 Primed H9 hESCs were used to generate the dCas9-KRAB transgenic line. For one well of a 6 -678 well plate containing 5 x 10^6 single cells, 1.5 µg pT2 -hUbC-dCas9-KRAB-T2a-Puro and 0.5 µg 679 pCMV(CAT)T7-SB100X (a gift of the Izsvák lab, expressing an SB100X hyperactive transposase), 680 were mixed with 4 µl Lipofectamine Stem (Life Technologies # STEM00008), in 200 µl Opti-MEM 681 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint I (Thermo Fisher # 31985070). 24 hours later, the mTeSR medium was changed, and 0.3 µg/ml 682 Puromycin (Life Technologies # A1113803) was added. For details on transfection, see below. 683 Eight days after puromycin treatment, colonies were picked and transferred individually into wells 684 of a 24 -well plate. Colonies were expanded, confirmed to remain puromycin -resistant and 685 validated for dCas9 protein expression by IF. Primed dCas9-KRAB transgenic hESCs thus 686 generated were converted to RSeT conditions as above and used for CRISPRi as described 687 below. 688 ASO, siRNA and CRISPRi gRNA design 689 Locked nucleic acid (LNA) -gapmeR ASOs were designed targeting the L1HS full -length LINE1 690 L1RP (Genbank AF148856). Candidate target sites were first selected using CRISPR -RT 691 (CRISPR RNA -Targeting), a Cas13a -based RNA target screening tool to pre -filter against off -692 target effects 74. The candidate targets were then screened by BLASTN against the human 693 genomic and transcript database, sites of low or no “Query cover” and “Percent Identity” scores 694 were retained. Candidate target sites were further ranked for high conservation in full -length 695 LINE1 elements using L1base ( http://l1base.charite.de/l1base.php)75. Lastly, selected target 696 sequences were adjusted to LNA -GapmeR design recommendations with assistance from 697 QIAGEN customer support. QIAGEN Negative control A Antisense LNA GapmeR (LG00000002) 698 and a sense oligo (SO) sequence of the interORF LINE1 ASO were used as controls (Table S1). 699 Both naked and 6 -FAM (Fluorescein)-labeled versions of the ASOs were synthesized and used 700 (see below). 701 siRNAs targeting TPRX13, TP5376,77 and H3.XY35 were selected from the literature (Table S1). 702 The Dharmacon siGENOME Non -Targeting Control siRNAs (D -001210-03) were used as 703 negative control (Table S1). All siRNAs were 6 -FAM (Fluorescein) labeled and ordered from 704 Horizon Discovery. 705 CRISPRi gRNAs (Table S1) targeting the 5’ UTR of L1RP with low off-target scores were selected 706 by combining the output of CRISPOR ( http://crispor.tefor.net/) and CRISPick 707 (https://portals.broadinstitute.org/gppx/crispick/public)78,79. Individual sgRNA thus designed, as 708 well as non -targeting control gRNAs80 were synthesized with overhangs (sense: 5′ -ACCG, 709 antisense: 5′ -AAAC) subcloned into the BsaI sites of pGL3 -U6-sgRNA-EGFP (Addgene # 710 107721)81. 711 Oligos or plasmids transfection 712 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Transfection of either ASOs or CRISPRi gRNA plasmids was conducted using Lipofectamine 713 Stem, per manufacturer’s instructions. Typically , per 6 -well with 2 ml of medi um, 4 µl of 714 Lipofectamine Stem and 25 nM ASOs or 1 µg/ml of plasmid were mixed in 200 µl of Opti-MEM 715 Medium for one well of a 6 -well plate containing single -cell hESCs at a density of 2.5 -4 x 10^5 716 cells per well. The transfection mixture was added directly while the cells were in suspension. At 717 the same time, 0.7 µM Trans-ISRIB (Sigma # 5095840001), 5 µM Emricasan (Sigma # SML2227) 718 and 10 µM Y-27632 were added to improve transfection efficiency and cell survival82; these were 719 removed during media chang e the next day. Cells were collected for downstream analyses 48 720 hours after transfection. 721 For siRNAs transfection, or co-transfection of ASOs and siRNAs, DharmaFECT 1 (Dharmacon # 722 T-2001-02) was used. The siRNA transfection was performed following the manufacturer’s 723 instructions, similar to above, including the additions of Trans-ISRIB, Emricasan, and Y-27632. 40 724 nM siRNAs were transfected, alone or together with 25 nM ASOs, per well of a 6-well plate. 725 RNA polymerase I inhibition 726 RSeT hESCs were plated at 4 x 10^5 cells per well of a 6 -well plate and switched to RSeT+DT 727 medium the following day. After 24 hours in RSeT+DT conditions, cells were treated with 0.25 µM 728 of BMH-21 (Sigma # 5099110001), while DMSO was added to control wells. Cells were fixed after 729 8 hours for immunofluorescence or collected after 24 hours for RNA isolation (see below). 730 EZH2 inhibition 731 RSeT hESCs were plated at 8 x 10^4 cells per well of a 12-well plate. After 24 hours, the medium 732 was changed to RSeT+DT, RSeT+DT+UNC2400, RSeT+DT+UNC1999 or 733 RSeT+TSA+UNC1999. UNC1999 and UNC2400 were added at 2.5 µM final concentration. Cells 734 were collected after 48 hours for RNA isolation. 735 Self-renewal assays 736 Primed hESCs were plated on Matrigel -coated 12-well plates as single cells at a density of 8 x 737 10^4 per well in a mTeSR -based Cloning Medium, in which CloneR (STEMCELL # 05888) was 738 used per manufacturer’s instructions to improve single cell survival and colony formation. Y-27632 739 was added for the first day. Three hours after cell seeding, the transfection mixture with 10 nM 740 ASOs, 1 µl Lipofectamine Stem, and 50 µl Opti -MEM was added. Six days later, colonies were 741 stained with Alkaline Phosphatase (AP) Red Substrate Kit (Vector Laboratories # SK-5100). AP+ 742 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint colonies were manually counted under the microscope. Representative images showing colony 743 morphologies were acquired with a Zeiss Axio z1 inverted microscope. 744 RSeT and RSeT+DT hESCs do not survive at the low densities used for standard colony formation 745 assays (above). Thus, colony area was quantified by ImageJ analysis of microscop e images 48 746 hours after ASO transfection (as above). 747 RNA Extraction and qRT-PCR 748 In primed and RSeT hESCs, FAM+ ASO-transfected cells were isolated by fluorescence-activated 749 cell sorting (FACS) on Sony MA900. The ASO transfection in RSeT+DT hESCs was highly 750 efficient, with consistently >80% FAM+ cells (data no t shown). Accordingly, in RSeT+DT 751 conditions, cells were directly collected from the plates for RNA extraction, without sorting. In the 752 case of co-transfection of ASOs (non-FAM labeled) with FAM-labeled siRNAs, FAM+ cells were 753 isolated by FACS. 754 RNA was extracted using the RNeasy mini kit (QIAGEN # 74104) or Direct -zol RNA 755 Microprep/Miniprep Kits (Zymo # R2061, # R2051) with on-column DNAse digestion , following 756 the manufacturer’s instructions. cDNA was generated using the SuperScript IV VILO Master Mix 757 (Life Technologies # 11756050) and used for qPCR using PowerUp SYBR Green Master Mix (Life 758 Technologies # A25777) on a QuantStudio ™ 5 Real-Time PCR System (Thermo Fisher). Data 759 were normalized to housekeeping genes (UBB and/or GAPDH), and the results were further 760 plotted and analyzed using GraphPad Prism. qPCR primers are listed in Table S1. 761 RNA-sequencing and data analysis 762 30-200 ng of high-quality RNA (equal amounts of RNA per sample for each experiment) was used. 763 3-4 replicates were processed per sample, except where indicated. Sequencing libraries were 764 prepared using the NEBNext Poly(A) mRNA Magnetic Isolation Module (NEB # E7490S) and 765 NEBNext Ultra II Directional RNA Library Prep Kit (NEB # E7760L), following the instruction 766 manuals. The quality of RNA and resulting libraries was assessed on a Fragment Analyzer 767 (Agilent). Sequencing was performed on an Illumina NextSeq500 with 75bp single -end reads at 768 the LTRI Sequencing Facility. 769 The RNA-seq reads underwent quality control and trimming using Trim Galore! v0.4.0, and were 770 subsequently aligned to the human reference genome (GRCh19) with TopHat2 v2.0.13 83. Gene 771 or repetitive element counts (using an annotation file from UCSC RepeatMasker) were obtained 772 using the FeatureCounts function from the Subread package (v1.5.0) 84. TopHat and 773 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint FeatureCounts settings used were as previously described85. The raw counts table was analyzed 774 in R with edgeR (v3.40.2) for generating the log2 -normalized counts table and comparing gene 775 expression changes between the treated and control samples. Low-count genes - with normalized 776 counts per million (CPM) ≤0.2 for RNA -seq in RSeT and RSeT+DT hESCs and CPM ≤0.5 for 777 RNA-seq in primed hESCs - were filtered out. Differential gene expression was exported as 778 topTable, which was further analyzed using tidyverse v1.3.2 and plotted using ggplot2 v3.4.1. MA 779 plots and volcano plots were generated to display differential gene expression relative to controls: 780 genes with an adj.P value 1.5 were considered to be significantly 781 up- or down-regulated. The log2-normalized counts were batch -corrected by the ComBat_seq 782 function in sva v3.46.086, and then used for generating multidimensional scaling (MDS) plots with 783 limma v3.54.1, and heatmaps with gplots v3.1.3. Genes were clustered with the setting of hclust 784 (as.dist(1-cor(t(df), method="spearman")), method="complete"). Gene Set Enrichment Analysis 785 (GSEA)87 was performed with the fgsea v1.24.0 package using pre -ranked t -values from the 786 topTable. Cell or embryo signature gene sets used in the analysis above are listed in Table S2. 787 Sets of genes upregulated or downregulated (adj.P value log2(1) or log2FC 788 < -log2(1)) were intersected with gene lists of each chromosome in R with the dplyr function, and 789 percentages were calculated relative to the total number of genes in each chromosome. 790 The gene sets of embryonic stages were generated with hierarchial clustering of the averaged 791 Penalized Kernel Matrix Regression (avg.PKMR) values of genes in each embryonic stage from 792 the scRNA-seq Yan et al. study22 using gplots v3.1.3 (see code in GitHub entry). Clusters of genes 793 that define specific stages or lineages ( Figure S5M) were used for further analysis (Table S2). 794 8cell_genes and Morulae_genes were further defined from the 8cell_Morulae_1687 gene cluster, 795 by filtering avg.PKMR > 20 first, and then filtering as 8C genes if avg.PKRM (8cell / Morulae) >1.5 796 or as Morulae genes if avg.PKMR (Morulae / 8cell) >1.5. The other 8C -genes and Pre8C-gene 797 sets were reported by Yu et al. 4. The 8cell_genes_Stowers set are genes highlighted in 798 Taubenschmid-Stowers et al.2. 799 To analyze the expression of LINE1 sub -families, 8C-TEs88 and selected protein-coding genes 800 during embryonic stages, the raw counts table from bulk RNA -seq by Hendrickson et al. 21 was 801 analyzed with edgeR (v3.40.2) in R. The raw counts of all genes were used to calculate 802 normalization factors, which were applied to calculate read-depth normalized counts for repetitive 803 elements. The normalized counts table from this bulk RNA-Seq dataset 21 and the scRNA -Seq 804 datasets22,23 of early-stage human embryos were sub-selected for the repeats indicated (Figures 805 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1A and S1A-S1E) or select protein-coding genes (Figures 5E and S14) and plotted as heatmaps 806 with gplots v3.1.3. See in GitHub entry for details. 807 Single-cell RNA-sequencing and data analysis 808 100,000 FAM+ (SO-L1 or ASO-L1 transfected) RSeT+DT hESCs were sorted into HBSS+2%FBS 809 on ice. Cells were centrifuged at 500x g for 4 min, resuspended in 100 µl HBSS+2%FBS and 810 immediately processed on a 10x Genomics Chromium platform using the Single Cell 3’ v3 RNA-811 seq kit, targeting a yield of ~12,000 cells per sample. Final libraries were sequenced on an 812 Illumina NovaSeq6000 S1 flow cell (Illumina) to a depth of ~94,000 reads/cell for ASO-L1 and 813 ~50,000 reads/cell for SO-L1. 814 The raw sequencing data were aligned to the human reference genome (GRCh38) using Cell 815 Ranger (v7.1.0) with default parameters. Cells with less than 5000 genes, more than 100000 816 unique molecular identifiers (UMIs), or higher than 15% mitochondrial UMIs were removed from 817 the downstream analysis. All mitochondrial genes were filtered out from the gene expression 818 matrices, which were then normalized using the Seurat v4 SCTransform function89. Cell clustering 819 and visualization were performed by FindClusters and RunUMAP functions in the Seurat 820 package89, with top 20 Principal components analysis (PCA) dimensions. 821 To project the hESC dataset onto the UMAP of human embryo data, the embryonic dataset was 822 reanalyzed with Seurat at first. In brief, gene expression matrices were normalized using the 823 NormalizeData function, and the top 3000 most variable genes were identified. The data was 824 visualized using the RunUMAP with parameters: dims = 1:8, seed.use = 2023, n.neighbors = 30, 825 n.epochs = 100, min.dist = 0.5. The SO-L1 and ASO-L1 datasets were also normalized using the 826 NormalizeData function and filtered for the top 2000 most variable genes. The anchors between 827 the embryo dataset and our hESC dataset were identified using FindTransferAnchors, with 828 parameters: dims = 1:10, and k.anchor = 100. The projections were performed using MapQuery. 829 The set of genes up-regulated upon ASO-mediated L1 KD from bulk RNA -seq was assessed 830 using Enricher ( https://maayanlab.cloud/Enrichr/), which predict ed high enrichment for certain 831 gene modules, such as the p53 pathway or PRC2 targets. Select gene sets from this Enricher 832 analysis were plotted onto the UMAP. 833 LAD-seq 834 Three biological replicates per sample were prepared for LAD sequencing using pA-DamID, with 835 an LMNB1 antibody (Abcam # ab16048), as described 41. The Dam -treated Control and 836 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint LMNB1+pA-Dam treated samples were subjected to DNA extraction using the ISOLATE II 837 Genomic DNA Kit (Bioline # BIO-52066). 838 0.5 µg of genomic DNA was used for library preparation following the DamID-seq protocol90, with 839 minor modifications: we omitted the DpnII digestion step for the pA -Dam reaction, as described 840 in the pA-Dam Seq protocol41. Adaptors and primers used in this study are listed in Table S1. At 841 the last step, after ten cycles of PCR amplification, the PCR products (size 1.2-1.5 kb) were 842 purified using 1.6x volume AMPure XP beads (Beckman Coulter # A63880) and assessed by 843 Fragment Analyzer (Agilent), then pooled for 100 bp single -end sequencing on an Illumina 844 HiSeq2500 machine with 1%PhiX spike-in. 845 NAD-seq 846 Three biological replicates per sample were prepared for NAD -seq following the crosslinked 847

of nucleoli isolation described by Vertii et al.40.In brief, after serial sonication with a Misonix 848 XL-2000 ultrasonic cell disruptor (10s each burst at full power) and re-suspension in high then 849 low magnesium buffers, cell nucleoli were released and monitored under the microscope. The 850 nucleolar-fraction DNA (NF-DNA) and the control whole-cell DNA (WC-DNA) were extracted using 851 NucleoBond AGX-20 columns ( Takara # 740544) and NucleoBond® Buffer Set III ( Takara # 852 740603), following the manufacturer’s instructions. To verify nucleoli enrichment, 4.6 ng of WC -853 DNA or NF -DNA per reaction was used for qPCR using primers listed in Table S1. NADs 854 sequencing libraries were generated using the TruSeq DNA PCR -free Library Preparation kit 855 (Illumina # 20015962) ; 350 bp fragments were sonicated and AMpure bead size -selected 856 following the manual’s instructions. The libraries were subjected to 150 bp paired-end sequencing 857 on an Illumina NovaSeq6000 instrument. 858 LAD-seq and NAD-seq data analysis 859 For both the pA -DamID-LAD and NAD sequencing, raw sequencing data files were mapped to 860 the human reference genome (GRCh38) using bwa, the output of which was filtered using 861 samtools 1.10 (by “samtools view -b -F 4 -q 10”). LAD or NAD peaks were called from the bam 862 file of each replicate using epic2 91 with parameters bin_size=10000, gap=10 for the LADs and 863 bin_size=10000, gap=5 for the NADs. LAD or NAD peaks called from at least two replicates out 864 of the three samples sequenced were considered for further analysis. 865 Bam files from above analysis were uploaded to the UCSC genome browser to inspect LADs and 866 NADs across whole chromosomes. The final LAD or NAD peaks bed files were intersected with 867 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint bedtools92 2.27.1 for their co -LAD/NAD coordinates, and calculated in R with the dplyr function 868 for the coverage length ratio of LADs, NADs and co -LADs/NADs in each chromosome. To 869 calculate the enrichment of LINE1 and other repetitive elements in LAD and NAD datasets, the 870 final LAD or NAD peaks bed files were intersected with each indicated repetitive element’s bed 871 file subsetted from RepeatMasker using the Intersection function in bedtools; this was also applied 872 to a random DNA set of equal number and averaged size to LADs or NADs generated using the 873 Random function in bedtools. The enrichment score was calculated as the ratio of the percentage 874 of the length of LADs or NADs occupied by the repeat element divided by the corresponding 875 percentage in the respective random DNA set. 876 Cross-Linking Immunoprecipitation (CLIP) and qPCR 877 Plates of RSeT+DT hESCs were washed once with cold PBS, then were placed in cold PBS on 878 ice with the lid open and UV-irradiated once in UV stratalinker 2400 with 400 mJ/cm2. Cells were 879 harvested, and nuclear extraction was performed using a standard Abcam protocol 880 (https://www.abcam.com/protocols/nuclear-extraction-protocol-nuclear-fractionation-protocol). 881 Prior to extraction, Nucleolin, LaminB1 or control Rabbit-IgG antibodies (See Table S3) were pre-882 bound to 50 µl Protein A Dynabeads (Life Technologies # 10002D) and incubated rotating for 3-6 883 hours at 4 °C. Beads were collected on a DynaMag (Thermo Fisher Scientific # 12321D) and re-884 suspended in High Salt CLIP buffer [1 M NaCl, 50 mM Tris pH 7.4, 1 mM EDTA, 0.5% Sodium 885 deoxycholate, 1% NP40, protease and RNase inhibitors ] containing 500 ng/mL tRNA (Invitrogen 886 # AM7119 ) and 1 mg/ml RNase -free BSA (Sigma Aldrich # A9418) to block for 30 min, then 887 collected and used immediately in RNA immunoprecipitation. Nuclear extracts (10 x 10^6 cells 888 per IP) were pre -blocked for 30 min with 20 µl of Protein A Dynabeads at 4 °C, 30 min, then 889 incubated with antibody-bound blocked beads overnight at 4 °C. Beads were washed twice with 890 1 ml High Salt Buffer and then twice with 1 ml Proteinase K Buffer [100 mM Tris-HCl pH 7.4, 50 891 mM NaCl, 10 mM EDTA, DnaseI ( Life Technologies # 18068015) and RNase inhibitors (Life 892 Technologies #10777019)]. Proteinase K Buffer washes were done for 15 min each on a 37 °C 893 shaker at 800 rpm. After washes, beads were resuspended in 100uL of Proteinase K Buffer and 894 RNA was eluted via addition of 50 µg Proteinase K (Life Technologies # 25530049) and 1 hour 895 incubation at 55 °C. RNA was extracted from beads using Trizol and standard phenol-chloroform 896 extraction. The aqueous phase containing the RNA was loaded onto RNeasy mini columns 897 (QIAGEN) with 2x volume of 100% ethanol and RNA was purified according to kit’s manual, 898 including an on-column DNase I treatment. Purified CLIP RNA was then used to generate cDNA 899 for qRT-PCR. Primers are listed in Table S1. 900 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Immunofluorescence (IF) 901 Sterilized coverslips were inserted in wells of a 12 -well plate and pre-Matrigel coated if needed. 902 hESCs, whether transfected or not, were plated on the coverslips as described above. Two days 903 post seeding (except where otherwise indicated), they were washed once with PBS and then fixed 904 in 4% Paraformaldehyde (VWR # CAAAJ61899-AP) for 8 min, followed by two PBS washes. The 905 plate was sealed with parafilm and stored in PBS at 4 °C for later use. 906 Fixed cells were permeabilized with 0.2% Triton X -100 in PBS for 2 min on ice. After two PBS -T 907 (PBS + 0.1% Tween-20) washes, cells were incubated in blocking buffer [PBS-T + 1% BSA (Sigma 908 Aldrich # A9418) + 10% Donkey serum (Sigma Aldrich # D9663) ] for 30 minutes at room 909 temperature (rt). Subsequently, cells were incubated with the primary antibodies diluted in the 910 blocking buffer, washed 3 times in PBS -T for 5 min, followed by secondary antibody incubation 911 (Table S3) and another set of PBS-T washes. Blocking and antibody incubations was performed 912 on parafilm, with the coverslips facing down, and the wash steps were performed in the plates 913 with the coverslips facing up. Coverslips were mounted onto slides with the VECTASHIELD® 914 Antifade Mounting Medium with DAPI (Vector Laboratories # VECTH1200) and kept at rt for at 915 least 1 hour before sealing with nail polish. 916 Most IF Images were acquired with an Andor BC43 Benchtop Confocal Microscope, except for 917 the H3.XY and TPRX1 co -IF in e4CL -8CLCs ( Figure S1F ) and the Cas9 IF in dCAS9 -KRAB 918 hESCs (Figure S2D), which were acquired with on an Olympus Optigrid M Structured Illumination 919 Microscope. 920 RNA-FISH and RNA-FISH/IF 921 Cells were plated as above for the IF protocol, and fixed in 4% PFA for 15 minutes, followed by 922 one PBS wash and two washes with ice -cold 70% Ethanol in DEPC -treated water. Coverslips 923 were stored in 70% Ethanol in a parafilm-sealed plate at 4 °C for later use. 924 To perform RNA-FISH, cells were rehydrated in RNA wash buffer (WB) [10% Formamide + 2x 925 SSC in DEPC water, prepared freshly ] at rt for two 5 min incubations. Coverslips were then 926 transferred to a piece of parafilm in a 15cm dish and incubated face down cell on RNA-FISH probe 927 solution. Probes were designed and generated by Stellaris (see Table S4) to a stock concentration 928 of 12.5 µM and were diluted 1:150 in hybridization buffer. To prepare the hybridization buffer, 929 dextran sulfate (Sigma # D6001) was first dissolved in DEPC water at 10%, to which 1 mg/ml 930 E.coli tRNA (Sigma # 10109541001), 2X SSC, 0.02% BSA, and 2 mM VRCs (NEB # S1402S) 931 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint were added. After sterilizing with a 0.02 µM filter, the mixture was aliquoted into 475 µl per tube 932 and frozen at -20 °C. The hybridization buffer was ready to use after adding 75 µl of formamide 933 per aliquot freshly after thawing. Samples were incubated in the dark with probe solution, first for 934 45 min at 37 °C in a hybridization oven, then at 30 °C overnight without sealing the petri dish with 935 parafilm. The next day, the dried coverslips were rehydrated by adding 50 -100 µL of WB at the 936 edges and waiting for a few minutes. Coverslips were transferred to a 12-well plate with WB for 5 937 min, followed by a 1-hour incubation in WB with 2.5 µl/ml RNAseOUT (Life Technologies # 938 10777019) at 37 °C. Coverslips were mounted onto microscope slides with VECTASHIELD® 939 Antifade Mounting Medium with DAPI and then sealed with nail polish. 940 To perform RNA-FISH and IF co -staining, the RNA-FISH steps above were carried out with the 941 following modifications. Primary antibodies were added with the probe in the hybridization buffer 942 at 1:100 dilution and incubated in a humified container, in the dark first for 4 hours at 37 °C, then 943 at 30 °C overnight. The next day, all steps being carried out in the dark, coverslips were transferred 944 to a 12-well plate with WB for 5 min, then transferred to a new well containing WB with 2.5 µl/ml 945 RNAseOUT and 1:200 diluted secondary antibodies, followed by a 1-hour incubation at 37 °C. 946 After two 5 min washes in WB (with 0.5 µl/ml RNAseOUT) at rt, coverslips were mounted onto 947 microscope slides with VECTASHIELD® Antifade Mounting Medium with DAPI and kept at rt for 948 at least 1 hour before sealing with nail polish. RNA-FISH images were acquired on an Andor BC43 949 Benchtop Confocal Microscope, except for Figures 1C and S1H, which were acquired on a Leica 950 DMI 6000 Spinning Disc Confocal Microscope. 951 DNA-FISH/IF 952 DNA-FISH was performed essentially using the method described by Bolland et al. 93, with 953 modifications to perform DNA -FISH and IF co -staining by combining it with the method of 954 Chaumeil et al.94. 955 DNA-FISH probes were generated from BACs (listed in Table S4) or LINE1 DNA plasmids. To 956 enrich for signal from full-length LINE1 elements, a hL1-5’_3.3kb plasmid was generated by sub-957 cloning the 5’ end ~3.3kb LINE1 sequence from the EF06R plasmid (Addgene # 42940) 95 to the 958 backbone of the pU6 -sgRosa26-1Cbh-Cas9-T2A-BFP plasmid (Addgene # 64216) 96, using the 959 NotI and XbaI sites. BAC DNA or hL1 -5’_3.3kb plasmids were purified using the NucleoBond® 960 Xtra Midi Plus EF kit (Takara # 740422.50) following instructions for low -copy or high -copy 961 plasmids, respectively. 10 µg BAD or hL1 -5’_3.3kb plasmids were nick -translated with DNase I 962 (Sigma # 04716728001) and DNA Polymerase I (NEB # M0209S). 150bp-700bp DNA fragments 963 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint were generated and incorporated with Aminoallyl -dUTP (Invitrogen™ # AM8439), then purified 964 and subjected to fluorescent dye coupling as described93 (see Table S4). 965 Single cells were collected as above (ASO-transfected cells were collected for FAM positivity by 966 FACS), and plated in their respective medium on Poly-Prep Slides (Sigma # P0425-72EA) in an 967 area encircled with a hydrophobic pen and pre -coated with Matrigel. After three hours in the 968 incubator, the cells were fixed, permeabilized, washed and stored in 50% glycerol in PBS at -969 20 °C, as described93. 970 For DNA-FISH/IF, slides were thawed in 20% glycerol/PBS at rt and subjected to 3 cycles of 971 freeze/thawing in liquid nitrogen, followed by two PBS washes and permeabilization in 0.5% 972 saponin/0.5% Triton X-100/PBS for 30 min at rt. Next, 100 µl blocking solution (2.5% BSA/10% 973 Donkey serum/PBS-T) was added to the cell area and covered with a coverslip, followed by a 30 974 min incubation. The coverslip was then gently removed, and immediately 100 µl of a 1:100 dilution 975 of primary antibody in blocking buffer was added, followed by an incubation at rt for 2 hours. After 976 3 washes in 0.2% BSA/PBS -T solution, cells were incubated in 100 µl of a 1:200 dilution of 977 secondary antibody in the dark for 1 hour (see Table S3 for antibodies used). After the IF steps, 978 cells were washed 3 times in PBS -T and fixed in 2% PFA for 10 minutes at rt, followed by a 30 979 min incubation in 0.1 M HCl, a PBS wash, and 30 min permeabilization in 0.5% saponin, 0.5% 980 Triton X-100 in PBS. After two PBS washes and equilibration in 50% formamide in 2x SSC, cells 981 were ready for addition of DNA-FISH probes mixture. 982 The DNA-FISH probe mixtures were prepared as described93 during the cell permeabilization and 983 IF steps. The Chromosome 19 probe (Creative Bioarray # FWCP -19) mixture was treated with 7 984 min 80°C denaturation treatment followed by a 10 min 37°C incubation and then directly added 985 to the slides, per manufacturer’s instructions. 986 When cells and probes were both ready, the slides were briefly drained with Q -tips and 987 immediately covered with 10 µl of probe mixture that were pre -dropped on a coverslip. After 988 sealing the coverslips with rubber cement (Marabu Fixogum), slides were heated at 78 °C for 2 989 min on a hot plate. Slides were then incubated overnight at 37 °C in a dark and humidified 990 chamber. The next day, the slides were subjected to a series of wash steps following the protocol 991 by Bolland et al. 93., then mounted with VECTASHIELD® Antifade Mounting Medium with DAPI 992 and kept at room temperature for at least 1 hour before sealing with nail polish. The DNA -FISH 993 and IF co-staining images were acquired on an Andor BC43 Benchtop Confocal Microscope. 994 Image quantification 995 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint ImageJ was used for quantifying signal intensity, area sizes and distances, as detailed in figure 996 legends. The area of the nucleus or the nucleolus was manually traced based on DAPI or 997 NCL/B23 signal. Quantification of the number of LINE1 foci per cell was acquired at threshold of 998 5 using a published RNA-FISH quantification method97. 999 The DNA-FISH/IF images were analyzed using Imaris. Chr.19, NCL (nucleolus), DAPI (nucleus), 1000 and TPRX1 DNA loci were 3D reconstructed using the Create Surface function in Imaris to identify 1001 their territories and location, and quantify their overlapping volumes. Representative cells and 1002 their 3D reconstructed territories were exported in supplementary videos. 1003 Statistical analyses 1004 The image quantification data are presented as mean ± SEM and analyzed with GraphPad Prism 1005 v9.3.1 or R v4.0.3., using student t-test. Depending on the data distribution or comparisons being 1006 made, the t-test methods are selected differently, and noted in each figure legend. 1007 RT-qPCR and CLIP-qPCR are plotted for the averaged fold changes ±SEM of n times biological 1008 replicates with GraphPad Prism. Ratio-paired student’s t-test was used for statistical analysis. 1009 For RNA -seq data analysis, differential gene expression was calculated as described above. 1010 Comparison of expression changes between genes sets were displayed by violin plot and 1011 analyzed for statistical significance using the Wilcoxon test in R. 1012 Sample size, number of replicates of studies, and statistical tests were chosen based on 1013 experience and variability of in vitro studies and stated in each figure legend. Unless otherwise 1014 indicated, all experiments were repeated at least three independent times. 1015 1016 Data and code availability 1017 Sequencing data, including raw reads and processed data (raw counts, normalized counts and 1018 differential gene expression table) have been deposited on the NCBI Gene Expression Omnibus 1019 repository (GEO, http://ncbi.nlm.nih.gov/geo) and will be accessible upon publication. The data 1020

Series for this study is GSE232939. 1021 scRNA-Seq from this study was compared to the scRNAseq data from human embryos 30 and 1022 e4CL-8CLCs3, which are available under the accession numbers E -MTAB-3929 and 1023 CNP0001454, respectively. 1024 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Raw-counts tables or normalized -counts tables from other studies of human embryos using 1025 scRNA-seq or bulk RNA -seq studies 21–23 were also used in this work. Processed data were 1026 downloaded from the supplementary tables in these publications. 1027 The authors declare that all other data supporting the findings of this study are available within 1028 the paper and its supplementary information files. Code supporting this study is available at a 1029 dedicated Github repository: https://github.com/tzzhangjuan1/LINE1_hESC. 1030 1031 Acknowledgments 1032 We thank members of the Santos lab, M. Lupien and M. Percharde for critical reading of the 1033 manuscript. We thank all members of Santos lab for their input throughout the project, in particular 1034 E. Collignon and T. Macrae for guidance on bioinformatics analyses. We are grateful to the J. Ellis 1035 and B. Di Stefano for assistance with hESC cell culture, to Z. Izsvák and J. Wang for sharing 1036 plasmids, to C. Arrowsmith and the Structural Genomics Consortium for sharing PRC2 inhibitors 1037 and to P . Maass, M. Lupien, L. Pelletier, B. van Steensel, T. van Schaik, M. Percharde and Y. Yin 1038 for precious advice and technical assistance. We are also thankful to K. Chan at the LTRI 1039 Sequencing Core, A. Bang at the LTRI Flow Cytometry Facility, R. Bielecki, L. Brown at the LTRI 1040 Microscopy Facility and M. Kownacka at the LTRI ESC Facility for their core facility support. This 1041 work was supported by a 2022 Medicine by Design Research Application Support Initiative (RASI) 1042 Award to S.S., the Natural Sciences and Engineering Research Council of Canada (RGPIN-2022-1043 05134) to M.M.H., and a Canada 150 Research Chair in Developmental Epigenetics, the Great 1044 Gulf Homes Charitable Foundation, and a 2019 Medicine by Design New Ideas Award to M.R. -1045 S.. 1046 Author contributions 1047 M.R.-S. and J.Z. conceived of the project and designed the experiments. J.Z. performed majority 1048 of the experiments and interpreted the data. L.M. performed CLIP, and the CRISPRi experiments 1049 with a cell line generated by J.Z.. K.M. conducted Pol I inhibition experiments. D.P.C., D.T., and 1050 J.L.W. provided guidance with scRNA -seq and prepared the libraries. L.W. analyzed and 1051 interpreted the scRNA -seq data, with assistance from M.A.M. and M.A.E.. M.A.E. provided 1052 plasmids and relevant advice regarding 4CL and e4CL cultures. L.H. and M.M.H. processed the 1053 LADs-seq and NADs-seq data, and L.H and J.Z. performed downstream analysis. S.S. and K.T. 1054 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint assisted J.Z. in IF and DNA -FISH co-staining experiments. M.R.-S. supervised the project. J.Z. 1055 and M.R.-S. wrote the manuscript with input from all authors. 1056 Declaration of interests 1057 The authors declare no competing interests. 1058 1059 Supplemental information 1060 Table S1. Sequences of ASOs-, siRNAs-, CRISPRi-gRNAs, primers and adaptor oligos 1061 used in this study. 1062 1063 Table S2. Gene sets used in this study 1064 Notes: The gene sets from scRNA-seq22 are re-derived here by hierarchical clustering. See 1065 STAR Methods and Figure S5M for details. 1066 1067 Table S3. Antibodies used for Immunofluorescence (IF) and Cross-Linking 1068 Immunoprecipitation (CLIP) 1069 1070 Table S4. Sequences of RNA-FISH probes and list of plasmids for DNA-FISH probes 1071 1072 Videos S1-S2. Representative videos of chromosome 19 by DNA-FISH and co-IF staining with 1073 NCL antibody in e4CL-induced 8CLCs and 4CL-cultured naïve hESCs. Corresponds to 1074 representative cells in Figure 5D. The Chr.19 and NCL-marked nucleolus territories imaged by 1075 microscope are three-dimensionally reconstructed in Imaris using the Create surface function. 1076 Scale bar, 20 µm. 1077 1078 Video S3-S6. Representative videos of TPRX1 loci by DNA-FISH and co-IF staining with B23, 1079 and H3.XY (e4CL cells) or LMNB1 (4CL, RSeT and Primed hESCs). Corresponds to 1080 representative cells in Figure S10B, indicated by large yellow arrows. The B23-marked 1081 nucleolus territories and DAPI-marked nucleus regions are three-dimensionally reconstructed in 1082 Imaris using the Create surface function. Scale bar, 20 µm. 1083 1084 1085 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint

1086 1. Braude, P., Bolton, V., and Moore, S. (1988). Human gene expression first occurs between 1087 the four- and eight-cell stages of preimplantation development. Nature 332, 459–461. 1088 10.1038/332459a0. 1089 2. Taubenschmid-Stowers, J., Rostovskaya, M., Santos, F., Ljung, S., Argelaguet, R., Krueger, 1090 F., Nichols, J., and Reik, W. (2022). 8C-like cells capture the human zygotic genome 1091 activation program in vitro. Cell Stem Cell 29, 449-459.e6. 10.1016/j.stem.2022.01.014. 1092 3. Mazid, M.A., Ward, C., Luo, Z., Liu, C., Li, Y., Lai, Y., Wu, L., Li, J., Jia, W., Jiang, Y., et al. 1093 (2022). Rolling back human pluripotent stem cells to an eight-cell embryo-like stage. Nature 1094 605, 315–324. 10.1038/s41586-022-04625-0. 1095 4. Yu, X., Liang, S., Chen, M., Yu, H., Li, R., Qu, Y., Kong, X., Guo, R., Zheng, R., Izsvák, Z., et 1096 al. (2022). Recapitulating early human development with 8C-like cells. Cell Reports 39, 1097 110994. 10.1016/j.celrep.2022.110994. 1098 5. Moya-Jódar, M., Ullate-Agote, A., Barlabé, P., Rodríguez-Madoz, J.R., Abizanda, G., 1099 Barreda, C., Carvajal-Vergara, X., Vilas-Zornoza, A., Romero, J.P., Garate, L., et al. (2023). 1100 Revealing cell populations catching the early stages of human embryo development in naive 1101 pluripotent stem cell cultures. Stem Cell Reports 18, 64–80. 10.1016/j.stemcr.2022.11.015. 1102 6. Zou, Z., Zhang, C., Wang, Q., Hou, Z., Xiong, Z., Kong, F., Wang, Q., Song, J., Liu, B., Liu, 1103 B., et al. (2022). Translatome and transcriptome co-profiling reveals a role of TPRXs in 1104 human zygotic genome activation. Science 378, abo7923. 10.1126/science.abo7923. 1105 7. Percharde, M., Lin, C.-J., Yin, Y., Guan, J., Peixoto, G.A., Bulut-Karslioglu, A., Biechele, S., 1106 Huang, B., Shen, X., and Ramalho-Santos, M. (2018). A LINE1-Nucleolin Partnership 1107 Regulates Early Development and ESC Identity. Cell 174, 391-405.e19. 1108 10.1016/j.cell.2018.05.043. 1109 8. Macfarlan, T.S., Gifford, W.D., Driscoll, S., Lettieri, K., Rowe, H.M., Bonanomi, D., Firth, A., 1110 Singer, O., Trono, D., and Pfaff, S.L. (2012). Embryonic stem cell potency fluctuates with 1111 endogenous retrovirus activity. Nature 487, 57–63. 10.1038/nature11244. 1112 9. Maksakova, I.A., Thompson, P.J., Goyal, P., Jones, S.J., Singh, P.B., Karimi, M.M., and 1113 Lorincz, M.C. (2013). Distinct roles of KAP1, HP1 and G9a/GLP in silencing of the two-cell-1114 specific retrotransposon MERVL in mouse ES cells. Epigenetics Chromatin 6, 15. 1115 10.1186/1756-8935-6-15. 1116 10. Ishiuchi, T., Enriquez-Gasca, R., Mizutani, E., Bošković, A., Ziegler-Birling, C., Rodriguez-1117 Terrones, D., Wakayama, T., Vaquerizas, J.M., and Torres-Padilla, M.-E. (2015). Early 1118 embryonic-like cells are induced by downregulating replication-dependent chromatin 1119 assembly. Nat Struct Mol Biol 22, 662–671. 10.1038/nsmb.3066. 1120 11. Jachowicz, J.W., Bing, X., Pontabry, J., Bošković, A., Rando, O.J., and Torres-Padilla, M.-E. 1121 (2017). LINE-1 activation after fertilization regulates global chromatin accessibility in the 1122 early mouse embryo. Nat Genet 49, 1502–1510. 10.1038/ng.3945. 1123 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 12. Lu, J.Y., Shao, W., Chang, L., Yin, Y., Li, T., Zhang, H., Hong, Y., Percharde, M., Guo, L., 1124 Wu, Z., et al. (2020). Genomic Repeats Categorize Genes with Distinct Functions for 1125 Orchestrated Regulation. Cell Reports 30, 3296-3311.e5. 10.1016/j.celrep.2020.02.048. 1126 13. Percharde, M., Sultana, T., and Ramalho-Santos, M. (2020). What Doesn’t Kill You Makes 1127 You Stronger: Transposons as Dual Players in Chromatin Regulation and Genomic 1128 Variation. BioEssays 42, 1900232. 10.1002/bies.201900232. 1129 14. Furano, A.V., Duvernell, D.D., and Boissinot, S. (2004). L1 (LINE-1) retrotransposon 1130 diversity differs dramatically between mammals and fish. Trends in Genetics 20, 9–14. 1131 10.1016/j.tig.2003.11.006. 1132 15. Bourque, G., Burns, K.H., Gehring, M., Gorbunova, V., Seluanov, A., Hammell, M., Imbeault, 1133 M., Izsvák, Z., Levin, H.L., Macfarlan, T.S., et al. (2018). Ten things you should know about 1134 transposable elements. Genome Biology 19, 199. 10.1186/s13059-018-1577-z. 1135 16. Venuto, D., and Bourque, G. (2018). Identifying co-opted transposable elements using 1136 comparative epigenomics. Development, Growth & Differentiation 60, 53–62. 1137 10.1111/dgd.12423. 1138 17. Cordaux, R., and Batzer, M.A. (2009). The impact of retrotransposons on human genome 1139 evolution. Nat Rev Genet 10, 691–703. 10.1038/nrg2640. 1140 18. Fedoroff, N.V. (2012). Transposable Elements, Epigenetics, and Genome Evolution. 1141 Science 338, 758–767. 10.1126/science.338.6108.758. 1142 19. Chuong, E.B., Elde, N.C., and Feschotte, C. (2017). Regulatory activities of transposable 1143 elements: from conflicts to benefits. Nat Rev Genet 18, 71–86. 10.1038/nrg.2016.139. 1144 20. Khan, H., Smit, A., and Boissinot, S. (2006). Molecular evolution and tempo of amplification 1145 of human LINE-1 retrotransposons since the origin of primates. Genome Res 16, 78–87. 1146 10.1101/gr.4001406. 1147 21. Hendrickson, P .G., Doráis, J.A., Grow, E.J., Whiddon, J.L., Lim, J.-W., Wike, C.L., Weaver, 1148 B.D., Pflueger, C., Emery, B.R., Wilcox, A.L., et al. (2017). Conserved roles of mouse DUX 1149 and human DUX4 in activating cleavage-stage genes and MERVL/HERVL 1150 retrotransposons. Nat Genet 49, 925–934. 10.1038/ng.3844. 1151 22. Yan, L., Yang, M., Guo, H., Yang, L., Wu, J., Li, R., Liu, P., Lian, Y., Zheng, X., Yan, J., et al. 1152 (2013). Single-cell RNA-Seq profiling of human preimplantation embryos and embryonic 1153 stem cells. Nat Struct Mol Biol 20, 1131–1139. 10.1038/nsmb.2660. 1154 23. Yandım, C., and Karakülah, G. (2019). Expression dynamics of repetitive DNA in early 1155 human embryonic development. BMC Genomics 20, 439. 10.1186/s12864-019-5803-1. 1156 24. Gafni, O., Weinberger, L., Mansour, A.A., Manor, Y.S., Chomsky, E., Ben-Yosef, D., Kalma, 1157 Y., Viukov, S., Maza, I., Zviran, A., et al. (2013). Derivation of novel human ground state 1158 naive pluripotent stem cells. Nature 504, 282–286. 10.1038/nature12745. 1159 25. Bayerl, J., Ayyash, M., Shani, T., Manor, Y.S., Gafni, O., Massarwa, R., Kalma, Y., Aguilera-1160 Castrejon, A., Zerbib, M., Amir, H., et al. (2021). Principles of signaling pathway modulation 1161 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint for enhancing human naive pluripotency induction. Cell Stem Cell 28, 1549-1565.e12. 1162 10.1016/j.stem.2021.04.001. 1163 26. Ludwig, T.E., Bergendahl, V., Levenstein, M.E., Yu, J., Probasco, M.D., and Thomson, J.A. 1164 (2006). Feeder-independent culture of human embryonic stem cells. Nat Methods 3, 637–1165 646. 10.1038/nmeth902. 1166 27. Mátés, L., Chuah, M.K.L., Belay, E., Jerchow, B., Manoj, N., Acosta-Sanchez, A., Grzela, 1167 D.P., Schmitt, A., Becker, K., Matrai, J., et al. (2009). Molecular evolution of a novel 1168 hyperactive Sleeping Beauty transposase enables robust stable gene transfer in 1169 vertebrates. Nat Genet 41, 753–761. 10.1038/ng.343. 1170 28. Gilbert, L.A., Larson, M.H., Morsut, L., Liu, Z., Brar, G.A., Torres, S.E., Stern-Ginossar, N., 1171 Brandman, O., Whitehead, E.H., Doudna, J.A., et al. (2013). CRISPR-Mediated Modular 1172 RNA-Guided Regulation of Transcription in Eukaryotes. Cell 154, 442–451. 1173 10.1016/j.cell.2013.06.044. 1174 29. Thakore, P.I., D’Ippolito, A.M., Song, L., Safi, A., Shivakumar, N.K., Kabadi, A.M., Reddy, 1175 T.E., Crawford, G.E., and Gersbach, C.A. (2015). Highly Specific Epigenome Editing by 1176 CRISPR/Cas9 Repressors for Silencing of Distal Regulatory Elements. Nat Methods 12, 1177 1143–1149. 10.1038/nmeth.3630. 1178 30. Petropoulos, S., Edsgärd, D., Reinius, B., Deng, Q., Panula, S.P., Codeluppi, S., Plaza 1179 Reyes, A., Linnarsson, S., Sandberg, R., and Lanner, F. (2016). Single-Cell RNA-Seq 1180 Reveals Lineage and X Chromosome Dynamics in Human Preimplantation Embryos. Cell 1181 165, 1012–1026. 10.1016/j.cell.2016.03.023. 1182 31. Yoshihara, M., Kirjanov, I., Nykänen, S., Sokka, J., Weltner, J., Lundin, K., Gawriyski, L., 1183 Jouhilahti, E.-M., Varjosalo, M., Tervaniemi, M.H., et al. (2022). Transient DUX4 expression 1184 in human embryonic stem cells induces blastomere-like expression program that is marked 1185 by SLC34A2. Stem Cell Reports 17, 1743–1756. 10.1016/j.stemcr.2022.06.002. 1186 32. Vuoristo, S., Bhagat, S., Hydén-Granskog, C., Yoshihara, M., Gawriyski, L., Jouhilahti, E.-1187 M., Ranga, V., Tamirat, M., Huhtala, M., Kirjanov, I., et al. (2022). DUX4 is a multifunctional 1188 factor priming human embryonic genome activation. iScience 25, 104137. 1189 10.1016/j.isci.2022.104137. 1190 33. De Iaco, A., Planet, E., Coluccio, A., Verp, S., Duc, J., and Trono, D. (2017). DUX-family 1191 transcription factors regulate zygotic genome activation in placental mammals. Nat Genet 1192 49, 941–945. 10.1038/ng.3858. 1193 34. Yang, F., Huang, X., Zang, R., Chen, J., Fidalgo, M., Sanchez-Priego, C., Yang, J., Caichen, 1194 A., Ma, F., Macfarlan, T., et al. (2020). DUX-miR-344-ZMYM2-Mediated Activation of MERVL 1195 LTRs Induces a Totipotent 2C-like State. Cell Stem Cell 26, 234-250.e7. 1196 10.1016/j.stem.2020.01.004. 1197 35. Resnick, R., Wong, C.-J., Hamm, D.C., Bennett, S.R., Skene, P.J., Hake, S.B., Henikoff, S., 1198 van der Maarel, S.M., and Tapscott, S.J. (2019). DUX4-Induced Histone Variants H3.X and 1199 H3.Y Mark DUX4 Target Genes for Expression. Cell Reports 29, 1812-1820.e5. 1200 10.1016/j.celrep.2019.10.025. 1201 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 36. Grow, E.J., Weaver, B.D., Smith, C.M., Guo, J., Stein, P., Shadle, S.C., Hendrickson, P.G., 1202 Johnson, N.E., Butterfield, R.J., Menafra, R., et al. (2021). p53 convergently activates 1203 Dux/DUX4 in embryonic stem cells and in facioscapulohumeral muscular dystrophy cell 1204 models. Nat Genet 53, 1207–1220. 10.1038/s41588-021-00893-0. 1205 37. Maeda, T., Kimura, Y., Nakano, T., and Yamaguchi, S. (2021). Ribosomal stress induces 2-1206 cell embryo-like state transition of the mouse ESCs through p53 activation. Biochemical and 1207 Biophysical Research Communications 579, 175–180. 10.1016/j.bbrc.2021.09.068. 1208 38. Xie, S.Q., Leeke, B.J., Whilding, C., Wagner, R.T., Garcia-Llagostera, F., Low, Y., Chammas, 1209 P., Cheung, N.T.-F., Dormann, D., McManus, M.T., et al. (2022). Nucleolar-based Dux 1210 repression is essential for embryonic two-cell stage exit. Genes Dev. 36, 331–347. 1211 10.1101/gad.349172.121. 1212 39. Yu, H., Sun, Z., Tan, T., Pan, H., Zhao, J., Zhang, L., Chen, J., Lei, A., Zhu, Y., Chen, L., et 1213 al. (2021). rRNA biogenesis regulates mouse 2C-like state by 3D structure reorganization of 1214 peri-nucleolar heterochromatin. Nat Commun 12, 6365. 10.1038/s41467-021-26576-2. 1215 40. Vertii, A., Ou, J., Yu, J., Yan, A., Pagès, H., Liu, H., Zhu, L.J., and Kaufman, P.D. (2019). 1216 Two contrasting classes of nucleolus-associated domains in mouse fibroblast 1217 heterochromatin. Genome Res 29, 1235–1249. 10.1101/gr.247072.118. 1218 41. van Schaik, T., Vos, M., Peric-Hupkes, D., HN Celie, P., and van Steensel, B. (2020). Cell 1219 cycle dynamics of lamina-associated DNA. EMBO reports 21, e50636. 1220 10.15252/embr.202050636. 1221 42. Vogel, M.J., Peric-Hupkes, D., and van Steensel, B. (2007). Detection of in vivo protein–1222 DNA interactions using DamID in mammalian cells. Nat Protoc 2, 1467–1478. 1223 10.1038/nprot.2007.148. 1224 43. Guetg, C., and Santoro, R. (2012). Formation of nuclear heterochromatin. Epigenetics 7, 1225 811–814. 10.4161/epi.21072. 1226 44. van Steensel, B., and Belmont, A.S. (2017). Lamina-Associated Domains: Links with 1227 Chromosome Architecture, Heterochromatin, and Gene Repression. Cell 169, 780–791. 1228 10.1016/j.cell.2017.04.022. 1229 45. Shah, P.P., Keough, K.C., Gjoni, K., Santini, G.T., Abdill, R.J., Wickramasinghe, N.M., 1230 Dundes, C.E., Karnay, A., Chen, A., Salomon, R.E.A., et al. (2023). An atlas of lamina-1231 associated chromatin across twelve human cell types reveals an intermediate chromatin 1232 subtype. Genome Biology 24, 16. 10.1186/s13059-023-02849-5. 1233 46. Lu, J.Y., Chang, L., Li, T., Wang, T., Yin, Y., Zhan, G., Han, X., Zhang, K., Tao, Y., Percharde, 1234 M., et al. (2021). Homotypic clustering of L1 and B1/Alu repeats compartmentalizes the 3D 1235 genome. Cell Res 31, 613–630. 10.1038/s41422-020-00466-6. 1236 47. Croft, J.A., Bridger, J.M., Boyle, S., Perry, P., Teague, P., and Bickmore, W.A. (1999). 1237 Differences in the Localization and Morphology of Chromosomes in the Human Nucleus. 1238 Journal of Cell Biology 145, 1119–1131. 10.1083/jcb.145.6.1119. 1239 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 48. Németh, A., Conesa, A., Santoyo-Lopez, J., Medina, I., Montaner, D., Péterfia, B., Solovei, 1240 I., Cremer, T., Dopazo, J., and Längst, G. (2010). Initial Genomics of the Human Nucleolus. 1241 PLOS Genetics 6, e1000889. 10.1371/journal.pgen.1000889. 1242 49. Tesařík, J., Kopečný, V., Plachot, M., and Mandelbaum, J. (1987). Ultrastructural and 1243 autoradiographic observations on multinucleated blastomeres of human cleaving embryos 1244 obtained by in-vitro fertilization. Human Reproduction 2, 127–136. 1245 10.1093/oxfordjournals.humrep.a136496. 1246 50. Boisvert, F.-M., van Koningsbruggen, S., Navascués, J., and Lamond, A.I. (2007). The 1247 multifunctional nucleolus. Nat Rev Mol Cell Biol 8, 574–585. 10.1038/nrm2184. 1248 51. Yang, K., Yang, J., and Yi, J. (2018). Nucleolar Stress: hallmarks, sensing mechanism and 1249 diseases. Cell Stress 2, 125–140. 10.15698/cst2018.06.139. 1250 52. Pinho, M., Macedo, J.C., Logarinho, E., and Pereira, P.S. (2019). NOL12 Repression 1251 Induces Nucleolar Stress-Driven Cellular Senescence and Is Associated with Normative 1252 Aging. Mol Cell Biol 39, e00099-19. 10.1128/MCB.00099-19. 1253 53. Leidenroth, A., and Hewitt, J.E. (2010). A family history of DUX4: phylogenetic analysis of 1254 DUXA, B, C and Duxbl reveals the ancestral DUXgene. BMC Evolutionary Biology 10, 364. 1255 10.1186/1471-2148-10-364. 1256 54. Lewin, T.D., Royall, A.H., and Holland, P.W.H. (2021). Dynamic Molecular Evolution of 1257 Mammalian Homeobox Genes: Duplication, Loss, Divergence and Gene Conversion Sculpt 1258 PRD Class Repertoires. J Mol Evol 89, 396–414. 10.1007/s00239-021-10012-6. 1259 55. Ohno S. (1972). So much “junk” DNA in our genome. Brookhaven Symp Biol. 1972;23:366-1260 70. PMID: 5065367. Brookhaven Symp Biol. 1261 56. Orgel, L.E., and Crick, F.H.C. (1980). Selfish DNA: the ultimate parasite. Nature 284, 604–1262 607. 10.1038/284604a0. 1263 57. Ivancevic, A.M., Kortschak, R.D., Bertozzi, T., and Adelson, D.L. (2018). Horizontal transfer 1264 of BovB and L1 retrotransposons in eukaryotes. Genome Biol 19, 85. 10.1186/s13059-018-1265 1456-7. 1266 58. Chinwalla, A.T., Cook, L.L., Delehaunty, K.D., Fewell, G.A., Fulton, L.A., Fulton, R.S., 1267 Graves, T.A., Hillier, L.W., Mardis, E.R., McPherson, J.D., et al. (2002). Initial sequencing 1268 and comparative analysis of the mouse genome. Nature 420, 520–562. 1269 10.1038/nature01262. 1270 59. Lander, E.S., Linton, L.M., Birren, B., Nusbaum, C., Zody, M.C., Baldwin, J., Devon, K., 1271 Dewar, K., Doyle, M., FitzHugh, W., et al. (2001). Initial sequencing and analysis of the 1272 human genome. Nature 409, 860–921. 10.1038/35057062. 1273 60. Feng, Q., Moran, J.V., Kazazian, H.H., and Boeke, J.D. (1996). Human L1 Retrotransposon 1274 Encodes a Conserved Endonuclease Required for Retrotransposition. Cell 87, 905–916. 1275 10.1016/S0092-8674(00)81997-2. 1276 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 61. Jurka, J. (1997). Sequence patterns indicate an enzymatic involvement in integration of 1277 mammalian retroposons. Proceedings of the National Academy of Sciences 94, 1872–1877. 1278 10.1073/pnas.94.5.1872. 1279 62. Sultana, T., van Essen, D., Siol, O., Bailly-Bechet, M., Philippe, C., Zine El Aabidine, A., 1280 Pioger, L., Nigumann, P., Saccani, S., Andrau, J.-C., et al. (2019). The Landscape of L1 1281 Retrotransposons in the Human Genome Is Shaped by Pre-insertion Sequence Biases and 1282 Post-insertion Selection. Molecular Cell 74, 555-570.e7. 10.1016/j.molcel.2019.02.036. 1283 63. Flasch, D.A., Macia, Á., Sánchez, L., Ljungman, M., Heras, S.R., García-Pérez, J.L., 1284 Wilson, T.E., and Moran, J.V. (2019). Genome-wide de novo L1 Retrotransposition 1285 Connects Endonuclease Activity with Replication. Cell 177, 837-851.e28. 1286 10.1016/j.cell.2019.02.050. 1287 64. Lafontaine, D.L.J., Riback, J.A., Bascetin, R., and Brangwynne, C.P . (2021). The nucleolus 1288 as a multiphase liquid condensate. Nat Rev Mol Cell Biol 22, 165–182. 10.1038/s41580-1289 020-0272-6. 1290 65. Doron-Mandel, E., Koppel, I., Abraham, O., Rishal, I., Smith, T.P., Buchanan, C.N., Sahoo, 1291 P.K., Kadlec, J., Oses-Prieto, J.A., Kawaguchi, R., et al. (2021). The glycine arginine-rich 1292 domain of the RNA-binding protein nucleolin regulates its subcellular localization. EMBO J 1293 40, e107158. 10.15252/embj.2020107158. 1294 66. Grimwood, J., Gordon, L.A., Olsen, A., Terry, A., Schmutz, J., Lamerdin, J., Hellsten, U., 1295 Goodstein, D., Couronne, O., Tran-Gyamfi, M., et al. (2004). The DNA sequence and 1296 biology of human chromosome 19. Nature 428, 529–535. 10.1038/nature02399. 1297 67. Long-range chromatin contacts in embryonic stem cells reveal a role for pluripotency factors 1298 and polycomb proteins in genome organization - PubMed 1299 https://pubmed.ncbi.nlm.nih.gov/24035354/. 1300 68. Ngan, C.Y., Wong, C.H., Tjong, H., Wang, W., Goldfeder, R.L., Choi, C., He, H., Gong, L., 1301 Lin, J., Urban, B., et al. (2020). Chromatin interaction analyses elucidate the roles of PRC2-1302 bound silencers in mouse development. Nat Genet 52, 264–272. 10.1038/s41588-020-1303 0581-x. 1304 69. Kraft, K., Yost, K.E., Murphy, S.E., Magg, A., Long, Y., Corces, M.R., Granja, J.M., Wittler, L., 1305 Mundlos, S., Cech, T.R., et al. (2022). Polycomb-mediated genome architecture enables 1306 long-range spreading of H3K27 methylation. Proc Natl Acad Sci U S A 119, e2201883119. 1307 10.1073/pnas.2201883119. 1308 70. Liu, J., Gao, M., He, J., Wu, K., Lin, S., Jin, L., Chen, Y., Liu, H., Shi, J., Wang, X., et al. 1309 (2021). The RNA m6A reader YTHDC1 silences retrotransposons and guards ES cell 1310 identity. Nature 591, 322–326. 10.1038/s41586-021-03313-9. 1311 71. Seczynska, M., Bloor, S., Cuesta, S.M., and Lehner, P.J. (2021). Genome surveillance by 1312 HUSH-mediated silencing of intronless mobile elements. Nature, 1–9. 10.1038/s41586-021-1313 04228-1. 1314 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 72. Masahito, T., Sakanoue, R., Takasu, A., Watanabe, N., Shimamoto, Y., and Miyamoto, K. 1315 (2023). Transition to the structurally vulnerable nuclear state is an integral part of mouse 1316 embryonic development. 2023.02.20.529332. 10.1101/2023.02.20.529332. 1317 73. Thomson, J.A., Itskovitz-Eldor, J., Shapiro, S.S., Waknitz, M.A., Swiergiel, J.J., Marshall, 1318 V.S., and Jones, J.M. (1998). Embryonic stem cell lines derived from human blastocysts. 1319 Science 282, 1145–1147. 10.1126/science.282.5391.1145. 1320 74. Zhu, H., Richmond, E., and Liang, C. (2018). CRISPR-RT: a web application for designing 1321 CRISPR-C2c2 crRNA with improved target specificity. Bioinformatics 34, 117–119. 1322 10.1093/bioinformatics/btx580. 1323 75. Penzkofer, T., Jäger, M., Figlerowicz, M., Badge, R., Mundlos, S., Robinson, P.N., and 1324 Zemojtel, T. (2017). L1Base 2: more retrotransposition-active LINE-1s, more mammalian 1325 genomes. Nucleic Acids Research 45, D68–D73. 10.1093/nar/gkw925. 1326 76. Kano, M., Matsushita, K., Rahmutulla, B., Yamada, S., Shimada, H., Kubo, S., Hiwasa, T., 1327 Matsubara, H., and Nomura, F. (2016). Adenovirus-mediated FIR demonstrated TP53-1328 independent cell-killing effect and enhanced antitumor activity of carbon-ion beams. Gene 1329 Ther 23, 50–56. 10.1038/gt.2015.84. 1330 77. Winsel, S., Sommer, A., Eschenbrenner, J., Mittelstaedt, K., Klar, U., Hammer, S., and 1331 Hoffmann, J. (2011). Molecular Mode of Action and Role of TP53 in the Sensitivity to the 1332 Novel Epothilone Sagopilone (ZK-EPO) in A549 Non-Small Cell Lung Cancer Cells. PLoS 1333 ONE 6, e19273. 10.1371/journal.pone.0019273. 1334 78. Concordet, J.-P., and Haeussler, M. (2018). CRISPOR: intuitive guide selection for 1335 CRISPR/Cas9 genome editing experiments and screens. Nucleic Acids Research 46, 1336 W242–W245. 10.1093/nar/gky354. 1337 79. Sanson, K.R., Hanna, R.E., Hegde, M., Donovan, K.F., Strand, C., Sullender, M.E., 1338 Vaimberg, E.W., Goodale, A., Root, D.E., Piccioni, F., et al. (2018). Optimized libraries for 1339 CRISPR-Cas9 genetic screens with multiple modalities. Nat Commun 9, 5416. 1340 10.1038/s41467-018-07901-8. 1341 80. Gilbert, L.A., Horlbeck, M.A., Adamson, B., Villalta, J.E., Chen, Y., Whitehead, E.H., 1342 Guimaraes, C., Panning, B., Ploegh, H.L., Bassik, M.C., et al. (2014). Genome-Scale 1343 CRISPR-Mediated Control of Gene Repression and Activation. Cell 159, 647–661. 1344 10.1016/j.cell.2014.09.029. 1345 81. Liu, Z., Lu, Z., Yang, G., Huang, S., Li, G., Feng, S., Liu, Y ., Li, J., Yu, W., Zhang, Y ., et al. 1346 (2018). Efficient generation of mouse models of human diseases via ABE- and BE-mediated 1347 base editing. Nat Commun 9, 2338. 10.1038/s41467-018-04768-7. 1348 82. Chen, Y., Tristan, C.A., Chen, L., Jovanovic, V.M., Malley, C., Chu, P.-H., Ryu, S., Deng, T., 1349 Ormanoglu, P., Tao, D., et al. (2021). A versatile polypharmacology platform promotes 1350 cytoprotection and viability of human pluripotent and differentiated cells. Nat Methods 18, 1351 528–541. 10.1038/s41592-021-01126-2. 1352 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 83. Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R., and Salzberg, S.L. (2013). 1353 TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and 1354 gene fusions. Genome Biology 14, R36. 10.1186/gb-2013-14-4-r36. 1355 84. Liao, Y., Smyth, G.K., and Shi, W. (2013). The Subread aligner: fast, accurate and scalable 1356 read mapping by seed-and-vote. Nucleic Acids Research 41, e108. 10.1093/nar/gkt214. 1357 85. Macrae, T.A., and Ramalho-Santos, M. (2021). The deubiquitinase Usp9x regulates PRC2-1358 mediated chromatin reprogramming during mouse development. Nat Commun 12, 1865. 1359 10.1038/s41467-021-21910-0. 1360 86. Zhang, Y., Parmigiani, G., and Johnson, W.E. (2020). ComBat-seq: batch effect adjustment 1361 for RNA-seq count data. NAR Genomics and Bioinformatics 2, lqaa078. 1362 10.1093/nargab/lqaa078. 1363 87. Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., 1364 Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., et al. (2005). Gene set enrichment 1365 analysis: A knowledge-based approach for interpreting genome-wide expression profiles. 1366 Proceedings of the National Academy of Sciences 102, 15545–15550. 1367 10.1073/pnas.0506580102. 1368 88. Grow, E.J., Flynn, R.A., Chavez, S.L., Bayless, N.L., Wossidlo, M., Wesche, D.J., Martin, L., 1369 Ware, C.B., Blish, C.A., Chang, H.Y., et al. (2015). Intrinsic retroviral reactivation in human 1370 preimplantation embryos and pluripotent cells. Nature 522, 221–225. 10.1038/nature14308. 1371 89. Stuart, T., Butler, A., Hoffman, P., Hafemeister, C., Papalexi, E., Mauck, W.M., Hao, Y., 1372 Stoeckius, M., Smibert, P., and Satija, R. (2019). Comprehensive Integration of Single-Cell 1373 Data. Cell 177, 1888-1902.e21. 10.1016/j.cell.2019.05.031. 1374 90. Leemans, C., van der Zwalm, M.C.H., Brueckner, L., Comoglio, F., van Schaik, T., Pagie, L., 1375 van Arensbergen, J., and van Steensel, B. (2019). Promoter-Intrinsic and Local Chromatin 1376 Features Determine Gene Repression in LADs. Cell 177, 852-864.e14. 1377 10.1016/j.cell.2019.03.009. 1378 91. Stovner, E.B., and Sætrom, P. (2019). epic2 efficiently finds diffuse domains in ChIP-seq 1379 data. Bioinformatics 35, 4392–4393. 10.1093/bioinformatics/btz232. 1380 92. Quinlan, A.R., and Hall, I.M. (2010). BEDTools: a flexible suite of utilities for comparing 1381 genomic features. Bioinformatics 26, 841–842. 10.1093/bioinformatics/btq033. 1382 93. Bolland, D.J., King, M.R., Reik, W., Corcoran, A.E., and Krueger, C. (2013). Robust 3D DNA 1383 FISH Using Directly Labeled Probes. J Vis Exp, 50587. 10.3791/50587. 1384 94. Chaumeil, J., Micsinai, M., and Skok, J.A. (2013). Combined Immunofluorescence and DNA 1385 FISH on 3D-preserved Interphase Nuclei to Study Changes in 3D Nuclear Organization. J 1386 Vis Exp, 50087. 10.3791/50087. 1387 95. Farkash, E.A., Kao, G.D., Horman, S.R., and Prak, E.T.L. (2006). Gamma radiation 1388 increases endonuclease-dependent L1 retrotransposition in a cultured cell assay. Nucleic 1389 Acids Research 34, 1196–1204. 10.1093/nar/gkj522. 1390 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 96. Chu, V.T., Weber, T., Wefers, B., Wurst, W., Sander, S., Rajewsky, K., and Kühn, R. (2015). 1391 Increasing the efficiency of homology-directed repair for CRISPR-Cas9-induced precise 1392 gene editing in mammalian cells. Nat Biotechnol 33, 543–548. 10.1038/nbt.3198. 1393 97. Raj, A., van den Bogaard, P., Rifkin, S.A., van Oudenaarden, A., and Tyagi, S. (2008). 1394 Imaging individual mRNA molecules using multiple singly labeled probes. Nat Methods 5, 1395 877–879. 10.1038/nmeth.1253. 1396 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1397 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S1. LINE1 RNA is highly expressed in early embryos and hESCs. 1398 (A) Heatmap showing induction of primate-specific LINE1 subfamilies and 8C-morula expressed 1399 TEs88 in RNA-seq datasets from developing human embryos21. 1400 (B and C) Average normalized counts of n=2 independent RNA-seq samples21 showing 1401 induction of primate-specific L1PA subfamilies (B), but not of older L1M subfamilies (C) during 1402 human embryonic cleavage stages. GV, germinal vesicle; MI, metaphase I; MII, metaphase II; 1403 PN, pronuclear stage; CL, cleavage; MOR, morula; ICM, inner cell mass; TROPH, 1404 trophectoderm. 1405 (D and E) Average normalized counts from scRNA-seq22 data showing induction of L1PAs (D) 1406 but not L1Ms (E) during early human embryonic development. MOR, morula; BLAST, 1407 blastocyst. 1408 (F) Schematic of hESCs culture conditions used in this study, and their corresponding embryo 1409 stages in vivo. DZNep and TSA, key components in the 4CL media3, were supplemented to the 1410 RSeT medium for 48 hours. IF staining of TPRX1 and H3.XY identifies 8CLCs induced by e4CL 1411 medium3. IF images, scale bar, 20 µm. Cell colony images, Scale bar, 100 µm. 1412 (G) qRT-PCR showing strong induction of markers of the naïve state in both RSeT and 4CL 1413 hESCs (3 batches each), relative to Primed hESCs. 1414 (H) RNA-FISH in 4CL, RSeT and Primed hESCs showing predominant nuclear localization of 1415 LINE1 RNA in hESCs. Representative of at least two independent experiments. Scale bar, 10 1416 µm. 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S2. ASOs and CRISPRi efficiently knockdown L1 RNA. 1440 (A) Quantification of LINE1 RNA-FISH signal intensity in RSeT+DT hESCs after 48 hours of 1441 ASO-L1, SO-L1, or ASO-Ctr transfection, showing efficient knockdown in ASO-L1 transfected 1442 cells compared to controls. Representative of two independent experiments. Welch’s t-test. ns = 1443 p > 0.05, ** = p < 0.01. 1444 (B) Quantification of LINE1 RNA-FISH foci per cell in primed hESCs after 48 hours of ASO-L1, 1445 SO-L1, or ASO-Ctr transfection, indicating significant reduction of LINE1 in ASO-L1 transfected 1446 cells compared to controls. Representative of at least four independent experiments. Mann-1447 Whitney test. ns = p > 0.05, **** = p < 0.0001. 1448 (C) Heatmap showing reduction of L1HS and L1PAs upon ASO-L1 KD in hESCs compared to 1449 SO-L1 and ASO-Ctr controls. 1450 (D) Schematic of generation of the CRISPRi system in hESCs using the Sleeping Beauty 1451 transposon system (see Methods for details). After puromycin selection and expansion of 1452 resistant colonies, Cas9 expression was validated by IF staining (bottom panel). Scale bar, 20 1453 µm. 1454 (E) Quantification of LINE1 RNA-FISH signal intensity in dCas9-KRAB hESCs 48 hours after 1455 transfection of L1-gRNA or Ctr-gRNA plasmids. The LINE1 RNA intensity in the L1-gRNA 1456 transfected EGFP+ cells is significantly reduced compared to control. Representative of two 1457 independent experiments. Welch’s t-test. * = p < 0.05. 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1472 1473 Figure S3. LINE1 KD impairs self-renewal of hESCs. 1474 (A-D) Colony size analysis in (A and B) RSeT+DT and (C and D) RSeT hESCs, respectively. 1475 ASO-L1 transfected cells display significantly smaller colony sizes than control ASO-Ctr and 1476 SO-L1 transfected cells. Representative images on the left (A and C), quantified on the right (B 1477 and D). Representative of three independent experiments. Mann-Whitney test. ns = p > 0.05, * 1478 = p < 0.05, *** = p< 0.001, **** = p < 0.0001. Scale bar, 100 µm. 1479 (E and F) Colony formation assay in primed hESCs. ASO-L1 transfected cells generate 1480 significantly less colonies than control ASO-Ctr and SO-L1 transfected cells. AP-stained (upper 1481 panel in E) colony numbers per 12-well were quantified and plotted in (F). ASO-L1 transfected 1482 primed hESCs also display abnormal colony morphology (lower panel in E). Representative of 1483 three independent experiments. Mann Welch’s t-test. ns = p > 0.05, *** = p< 0.001. Scale bar, 1484 100 µm. 1485 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1486 1487 Figure S4. hESC lines used in this study display a normal karyotype. 1488 (A and B) Karyotyping of H9 hESCs maintained in (A) mTeSR primed and (B) RSeT naïve-like 1489 conditions. Cells are confirmed to have a normal karyotype. 1490 (C) Karyotyping of primed H1 hESCs. Cells are confirmed have a normal karyotype. 1491 (D) Karyotyping of the dCas9-KRAB transgenic H9 hESC line maintained in RSeT naïve-like 1492 conditions. Cells after dCas9-KRAB sequence insertion is confirmed to be normal karyotype. 1493 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1494 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S5. LINE1 KD disrupts the transcriptional profile of hESCs and induces 8C 1495 signatures. 1496 (A and B) MDS plot of genes across all samples, showing that ASO-L1 KD (A) RSeT and (B) 1497 primed hESCs have distinct gene expression profiles from their respective control SO-L1 and 1498 ASO-Ctr samples. 1499 (C-E) Volcano plot showing log2-fold change (FC) in gene expression following ASO-L1 KD in 1500 (C) RSeT+DT, (D) RSeT and (E) Primed hESCs compared to control SO-L1 transfected cells. 1501 Red or blue highlight genes of adj.P-value 1.5 or < -1.5, respectively. 8C 1502 marker genes from Taubenschmid-Stowers et al.2 are labeled. 1503 (F) Violin plot of data from ASO-L1KD in RSeT+DT hESCs showing significant upregulation of 1504 morulae and 8C gene sets from different studies22,2,4, compared to a random set of n=3000 1505 genes. Wilcoxon test. **** = p < 0.0001. 1506 (G and H) Heatmap showing induction of gene sets of the 8C stage from X. Yu et al.4 in (G) 1507 RSeT and (H) Primed hESCs upon ASO-L1KD compared to ASO-Ctr and SO-L1 controls. 1508 (I) qRT-PCR showing a lack of upregulation of key 8C marker genes in LINE1 KD RSeT H9 1509 hESCs (without DT) with ASOs targeting the inter-ORF and ORF2 sites, respectively. Data are 1510 mean ± SEM, n = 2 biological replicates. Ratio paired Student’s t-tests. ns = p > 0.05, 1511 (J and K) Gene Set Enrichment Analysis of the transcriptional profile of ASO-L1 KD (J) RSeT 1512 and (K) Primed hESCs for the enrichment of gene sets from different stages of pre-implantation 1513 development and 8C/pre-8C gene sets4,22. 1514 (L) Schematic illustrating the impact of LINE1 KD in hESCs of distinct naïve states. hESCs of a 1515 higher naïve nature (RSeT+DT > RSeT >Primed) are more permissive for derepression of the 1516 8C program upon KD of LINE1. See Figures 1F and S5J-S5K. 1517 (M) Heatmap analysis of scRNA-seq data from early human embryos22, see Methods for details. 1518 Gene sets of each human embryonic stage are hierarchy clustered and listed in Table S2. 1519 TROPH, trophectoderm; PRIMT, Primitive endoderm. 1520 1521 1522 1523 1524 1525 1526 1527 1528 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1529 1530 1531 1532 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S6. Additional analyses of scRNA-seq of LINE1 KD hESCs. 1533 (A) Metrics assessing the quality of scRNA-seq libraries, plotting transcripts per cell 1534 (nCount_RNA), genes detected per cell (nFeature_RNA) and mitochondria reads percentage 1535 (percet.mt) in the ASO-L1 KD and SO-L1 samples. 1536 (B) Visualization of 28 cell clusters with uniform manifold projection (UMAP). Cluster 15 is 1537 identified to as 8CLCs cluster. 1538 (C and D) UMAP view of cellular expression levels of (c) pluripotency markers; (d) naïve hESC 1539 markers. 1540 (E) Top signatures of genetic perturbations enriched at genes upregulated upon L1KD in 1541 RSeT+DT hESCs (see Methods). 1542 (F) UMAP view of expression levels of genes induced by POU5F1 KD or SOX2 KO. 1543 (G) Top chromatin-bound factors enriched at genes upregulated upon L1KD in RSeT+DT 1544 hESCs (complementary to analysis shown in Figure 2E). 1545 (H) UMAP view of expression levels of stress-related gene targets of ATF3 or JUND. 1546 (I) UMAP view of expression levels of targets of PRC2. 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1560 1561 1562 1563 1564 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S7. The expression of TPRX1, DUX4 and DUX4-like genes in LINE1 KD hESCs. 1565 (A) IF of TPRX1 in ASO-L1 transfected H1 hESCs compared to ASO-Ctr control. 1566 Representative of two independent experiments. Scale, 100 µm. 1567 (B) Quantification of TPRX1+ cells ratio in (A). Same as in Figure 3B, 380~400 cells per sample 1568 from five random views were quantified. Data are mean ± SEM. Welch’s t-test. Representative 1569 of two independent experiments. 1570 (C and D) Plot of expression of select genes from RNA-seq data of ASO-L1 KD RSeT+DT 1571 ESCs compared to SO-L1 control for (upper panel) fold change (F.C.) and (lower panel) 1572 normalized log2 counts per million (log2 CPM). These data indicate (C) undetectable expression 1573 of DUX4 and DUX4-like genes, in contrast to (D) TPRXs, ZSCANs, and LEUTX genes. 1574 1575 1576 1577 1578 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1579 1580 1581 1582 1583 1584 1585 1586 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S8. TPRX1 KD partially rescues LINE1 KD-mediated 8C gene upregulation, 1587 whereas TP53 and H3.XY KD has no detectable impact. 1588 (A) Schematic of siRNAs knockdown of TPRX1, TP53, and H3.XY in RSeT+eDT hESCs, 1589 compared with a control non-targeting (NT) siRNAs. RSeT+eDT conditions were used to induce 1590 a higher level of putative 8C regulators and better assess their siRNA-mediated KD (see STAR 1591 Methods). 1592 (B-D) qRT-PCR data of two biological replicates (Rep.) of siRNA KD in RSeT+eDT hESCs, as 1593 shown in a, validating efficient KD of corresponding targets: (B) TPRX1-siRNA (TPRX1si); (C) 1594 H3.XY-siRNA (H3.XYsi); and (D) TP53-siRNA (TP53si). Data are mean fold changes of n=3 1595 technical repeats, ± SD. 1596 (E and F) MA plot showing log2 fold changes in gene expression following (E) L1KD+TP53-1597 siRNA and (F) L1KD+H3.XY-siRNA compared to control transfected cells (ASO-Ctr+NT-siRNA). 1598 Red or blue highlight genes of adj.P-value 1.5 or < -1.5, respectively. 8C 1599 marker genes from Taubenschmid-Stowers et al.2 are labeled in dark blue. 1600 (G) Plot of the NES value from the GSEA analysis showing reduced enrichment of 8C marker 1601 genes2 in L1 KD+TPRX1-siRNA transfected hESCs compared to L1 KD+NTsi-RNA transfected 1602 cells, while L1 KD+TP53-siRNA and L1 KD+H3.XY-siRNA transfected hESCs shows similar 1603 induction level of 8C genes. adj.P values of each sample by GESA analysis are indicated on top 1604 of the bar. 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1619 1620 1621 1622 1623 1624 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S9. LINE1 elements are enriched at the nuclear lamina and the periphery of the 1625 nucleolus. 1626 (A) Enrichment of transposon families in LADs and NADs of RSeT (left panel) and Primed 1627 hESCs (right panel). Coverage ratio of each family in LADs or NADs was compared to the 1628 coverage ratio in a random set of genomic sequences of equal number and averaged size (set 1629 as 1). L1HSPA: L1HS and L1PA1-8 of length > 5.5 kb. 1630 (B) Representative image of LINE1 DNA-FISH and co-IF staining for NCL and LBR antibody in 1631 RSeT+DT hESCs, showing that LINE1 loci are enriched in the vicinity of nucleolar and laminar 1632 domains. Boxed area (*) in the max projected image (top left panel) is enlarged and displayed 1633 as a representative single z-stack image in separate channels. Yellow dotted line encircles the 1634 nucleolus area. Yellow and red arrows point to LINE1 loci-enriched spots at the nucleolar and 1635 laminar domains, respectively. Noted the nucleoplasm (np) areas are sparse in LINE1 loci. 1636 Representative of three independent experiments. Scale bar, 10 µm. 1637 (C and D) Representative image of LINE1 RNA-FISH and co-IF staining for NCL and LBR 1638 antibody in (C) RSeT+DT hESCs and (D) Primed hESCs, showing that LINE1 RNA are enriched 1639 in the vicinity of nucleolar and laminar domains. The max projected image (top left panels in C 1640 and two top left panels in D) displays high expression of LINE1 in the nucleus. A representative 1641 plane is shown at the bottom left panels in (C) and two top right panels in (D), with the boxed 1642 area (*) enlarged at the right panels in (C) and bottom panels in (D), with color channels 1643 separated. Yellow dotted line encircles the nucleolus area. Yellow and red arrows point to 1644 LINE1 RNA foci-enriched spots at the nucleolar and laminar domains, respectively. Noted the 1645 nucleoplasm (np) areas are sparse in LINE1 foci. Representative of at least three independent 1646 experiments. Scale bar, 10 µm. 1647 (E) Cross-linking immunoprecipitation (CLIP) qRT-PCR analysis for indicated RNAs pulled-1648 down with NCL or LMNB1 antibodies, compared to IgG pulldown. LINE1 RNA is associated with 1649 both NCL and LMNB1. Note the high enrichment for pre-rRNA is highly in the NCL pulldown. 1650 Data are mean ± SEM, n = 3 biological replicates. Multi Ratio paired t-test. * = p < 0.05. 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1661 1662 1663 1664 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S10. TPRX1 loci gain association with the nucleolus and the lamina in the 1665 progression from 8CLCs to more developmentally advanced hESC states. 1666 (A) Genome browser view of the TPRX1 region, displaying enrichment in NADs. 1667 (B) Representative image of TPRX1 DNA-FISH and co-IF staining for B23 and LMNB1 in 4CL, 1668 RSeT and primed hESCs. In e4CL, instead of LMNB1 antibody, co-IF was carried out for H3.XY 1669 to identify 8CLCs; thus, the distance of TPRX1 loci to the lamina in the images of e4CL cells 1670 was measured to the DAPI periphery. The boxed area in the max projected images (left panels) 1671 is enlarged and displayed as two representative z-stack (Z) images (right panels). The large 1672 yellow arrows point to representative cells exported in Videos S3-S6. Representative of two 1673 independent experiments. Scale bar, 10 µm. 1674 (C and D) Quantification of data from the TPRX1 DNA-FISH and co-IF staining (shown in A), 1675 plotting the distance of TPRX1 loci to (C) DAPI periphery (e4CL) or LMNB1-marked laminar 1676 domain (4CL, RSeT and primed hESCs); and (D) B23-marked nucleolus domain. If the distance 1677 is < 0.5 µm, it is defined as within LADs or NADs. Data are from two replicated independent 1678 experiments. Number of cells quantified in each group is indicated in Figure 4C. Brown-Forsythe 1679 and Welch Anova tests. ns = p > 0.05, ** = p < 0.01, *** = p< 0.001, **** = p < 0.0001. 1680 (E and F) Quantification of data from the TPRX1 DNA-FISH and co-IF staining for B23 and 1681 LMNB1 in control and LINE1 KD RSeT+DT hESCs, plotting the distance of TPRX1 loci to (E) 1682 nucleolar and (F) lamina domains. n=cells /TPRX1 loci quantified. Data are from three 1683 independent experiments. Mann-Whitney test. ns = p > 0.05. 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1696 1697 1698 1699 1700 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint Figure S11. H3.XY and DUX4 loci are primarily located at the lamina in hESCs. 1701 (A) Genome browser view of the H3.XY (annotated as H3.Y1 and H3.Y2) loci, displaying 1702 enrichment in LADs and NADs. H3.XY DNA-FISH probes are generated with the RP11-845I22 1703 BAC plasmid. See Table S4 for DNA-FISH probes for other loci. 1704 (B) Representative z-stack image of H3.XY DNA-FISH and co-IF staining for B23 and LMNB1 in 1705 4CL, RSeT and primed hESCs. In e4CL, instead of B23 antibody, co-IF was carried out for 1706 H3.XY to identify 8CLCs; thus, the distance of H3.XY loci to nucleolus in the images of e4CL 1707 cells was not measured. The yellow arrows point to H3.XY loci in the nucleoplasm. 1708 Representative of two independent experiments. Scale bar, 10 µm. 1709 (C) Quantification of data from the H3.XY DNA-FISH and co-IF staining (shown in B), plotting 1710 the distance of H3.XY loci to the laminar domain. If the distance is 0.05, * = p < 0.05. Representative of two independent 1713 experiments. 1714 (D) Representative image of DUX4 DNA-FISH and co-IF staining for NCL in 4CL hESCs. The 1715 distance of DUX4 loci to the lamina in the images was measured to the DAPI periphery. The 1716 max projected image (left panel) is displayed as three z-stack (Z) images (right panels). The 1717 yellow arrows point to DUX4 loci, locating frequently at the DAPI periphery. Representative of at 1718 least two independent experiments. Scale bar, 10 µm. 1719 (E) Plot of the DUX4 location percentile in each listed category based on the distance 1720 quantification (data not shown). The definition of its location in LADs or NADs is the same as 1721 above in (C and D). n=cells / H3.XY loci quantified. Representative of two independent 1722 experiments. 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1734 1735 Figure S12. Chromosome 19 contains NADs but not LADs. 1736 (A) Representative genome browser views of LADs and NADs enrichment signatures of Chr. 1737 19, Chr. 1, Chr. 13, and Chr. X., showing large-scale changes in NADs and very stable LADs in 1738 Naïve-like (N) and Primed (P) hESCs. 1739 (B) Plot of NADs, LADs, co-LADs/NADs, and non-LADs/non-NADs coverage as percentage to 1740 the total length in each chromosome. Chr. 19 lacks LADs in both RSeT and Primed hESCs. 1741 1742 1743 1744 1745 1746 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1747 1748 Figure S13. RNA Pol I inhibition and CRISPRi KD of NCL impair nucleolar architecture in 1749 hESCs. 1750 (A) NCL and B23 IF in hESCs treated with 0.25 µM BMH for 8 hours, showing disrupted 1751 nucleolar architecture, relative to DMSO controls. Representative of two independent 1752 experiments. Scale bar, 10 µm. 1753 (B) Representative NCL and FBL signal intensity profile plot across nucleolus (white-dotted line 1754 in Figure 6B), indicating an extrusion of FBL from the NCL territory in LINE1 KD cells. 1755 (C) NCL IF in dCas9-KRAB transgenic hESCs transfected with NCL-gRNAs, compared to Ctr-1756 gRNAs as control. GFP co-IF identifies cells with positive gRNA transfection. Representative 1757 images of two independent experiments. Scale bar, 10 µm. 1758 (D) Quantification of NCL and GFP IF images (examples in C), showing significant KD of NCL 1759 protein in GFP+ transfected with NCL-gRNAs, relative to controls. Data are from two 1760 independent experiments. Welch’s t-test. ns = p > 0.05, * = p < 0.05. 1761 1762 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint 1763 1764 Figure S14. Nucleolar factors and components of PRC2 are highly induced at the 8C 1765 stage in vivo. 1766 (A) Heatmap representation of the expression of select heterochromatin and nucleolus factors 1767 during early human embryonic development, profiled using scRNA-seq22. Nucleolar proteins 1768 and PRC2 subunits are among genes significantly upregulated (bolded) at the 8C stage. 1769 TROPH, trophectoderm; PRIMT, Primitive endoderm. 1770 (B-H) Plot of the average expression levels (Penalized Kernel Matrix Regression, avg.PKMR of 1771 indicated genes in cells of each embryonic stage sample, profiled using scRNA-seq22. TROPH, 1772 trophectoderm; PRIMT, Primitive endoderm. 1773 1774 1775 1776 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted September 23, 2024. ; https://doi.org/10.1101/2024.09.22.614332doi: bioRxiv preprint