DGCR8haploinsufficiency leads to primate-specific RNA dysregulation and pluripotency defects

The 22q11.2 deletion syndrome (22qDS) is caused by a microdeletion in chromosome 22, 1 including DGCR8, an essential gene for miRNA production. The contribution of human 2 DGCR8 hemizygosity to the disease is still unclear . In this study, we generated two human 3 pluripotent cell models containing a single functional DGCR8 allele to elucidate its role on 4 22qDS. DGCR8+/- cells show increased apoptosis as well as self -renewal and differentiation 5 defects in both the naïve and primed state s. The expression of primate -specific miRNAs was 6 largely affected, due to impaired miRNA processing and chromatin accessibility. DGCR8+/- 7 cells also displayed a pronounced reduction in human endogenous retrovirus class H (HERVH) 8 expression, a primate -specific retroelement essential for pluripotency maintenance. 9 Importantly, the reintroduction of primate -specific miRNAs as well as the miR-371-3 cluster 10 rescued the cellular and molecular phenotypes of DGCR8+/- cells. Our results suggest that 11 DGCR8 is haploinsufficient in humans and that miRNAs and transposable elements may have 12 co-evolved in primates as part of an essential regulatory network to maintain stem cell identity. 13 14

DGCR8/22q11.2DS/endogenous retroviruses/miRNAs/pluripotency/ 15 16 17 18 19 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 3

20 The 22q11.2 deletion syndrome (22qDS) is a human genetic disorder caused by a heterozygous 21 microdeletion at chromosome 22. It is the most common human chromosom al deletion, with 22 an incidence of 1/3,000 to 1/6,000 in live births (McDonald-McGinn et al, 2015). Although the 23 deletion is variable in size (1.5Mb to 3Mb ), the largest and most frequent deletion (85% 24 patients) affects around 40 protein-coding genes (McDonald-McGinn et al., 2015). The major 25 clinical manifestations include developmental disabilities, congenital heart disease, palatal 26 abnormalities, immune deficiency and increased risk of autoimmune diseases and psychiatric 27 illnesses, such as autism and schizophrenia (McDonald-McGinn et al., 2015). 28 The precise relationship between the deletion of specific genes and the subsequent clinical 29 symptoms remains to be fully elucidated. Amongst the 40 genes affected by the microdeletion, 30 the DGCR8 gene has received much attention, due to its essential role in miRNA biogenesis. 31 MiRNAs are small non-coding RNAs that negatively regulate mRNA stability and translation 32 by imperfect base -pairing to complementary sequences (Gebert & Macrae, 2019) . Most 33 miRNAs are transcribed as long primary transcripts (pri -miRNAs) that fold into hairpin 34 structures. These are recognised and cleaved in the nucleus by the Microprocessor complex, 35 which is composed of the dsRNA-binding protein DGCR8 and the RNAse III endonuclease 36 Drosha (Denli et al, 2004; Gregory et al, 2004; Han et al, 2004; Landthaler et al, 2004). Next, 37 the excised hairpin is exported to the cytoplasm and further processed by the RNAse III 38 endonuclease Dicer. Finally, one of the strands of the mature miRNA duplex is incorporated 39 into the RNA-induced silencing complex ( RISC) to guide repression of the target mRNA 40 (Treiber et al, 2019). In addition to miRNA production, the Microprocessor can directly control 41 the levels of mRNAs by cleaving stem-loop structures which resemble pri-miRNAs, including 42 the DGCR8 transcript itself and retrotransposon -derived RNAs amongst others (Han et al , 43 2009; Heras et al, 2013; Knuckles et al, 2012; Macias et al, 2012; Triboulet et al, 2009). 44 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 4 Mouse models of Dgcr8 heterozygosity have led to contrasting results. While losing one copy 45 of Dgcr8 in mouse embryonic stem cells (mESCs) was not sufficient to observe significant 46 alterations in the expression of miRNAs by microarrays (Wang et al , 2007) , analyses of 47 miRNA expression by RT-qPCR or deep sequencing from brain tissue of Dgcr8 heterozygous 48 mice showed dysregulation of a subgroup of miRNAs (Earls et al, 2012; Fenelon et al, 2011; 49 Marinaro et al , 2017; Schofield et al , 2011; Stark et al , 2008) . These m ice also displayed 50 behavioural changes and cognitive defects, which have been attributed to changes in the 51 structure of neuronal dendrites and their synaptic properties (Earls et al., 2012; Schofield et al., 52 2011; Stark et al., 2008). These studies suggest that the alteration of the structure and function 53 of neuronal circuits in Dgcr8 heterozygous mice could provide a genetic explanation to the 54 neuropsychiatric manifestations in 22qDS. Although the complete deficiency of Dgcr8 is 55 embryonically lethal in mice (Wang et al., 2007), cell-specific Dgcr8 ablation has revealed that 56 this gene is also necessary for optimal function of m ouse immune cells, including helper T 57 cells, B cells, NK cells and thymic architecture as well as reproductive function, female fertility 58 and spermatogenesis (Bezman et al, 2010; Brandl et al, 2016; David et al, 2011; Khan et al, 59 2014; Kim et al, 2016; Zimmermann et al, 2014). 60 Here, we have developed two different human cell models of DGCR8 heterozygosity, in the 61 embryonic stem cell line H9 and teratocarcinoma cell line PA-1, to investigate if DGCR8 is 62 haploinsufficient in humans . Our results show that inactivating one DGCR8 allele results in 63 haploinsufficiency as manifested by dysregulation of miRNA biogenesis but also changes in 64 chromatin accessibility. Despite the functional conservation of DGCR8 throughout evolution, 65 we found that a high proportion of the affected mature miRNAs are primate specific. As a 66 result, DGCR8 heterozygote cells display alterations in the gene expression profile associated 67 with pluripotency maintenance and embryonic development, including the human -specific 68 endogenous retrovirus type-H (HERVH) family, a crucial Transposable Element for stem cell 69 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 5 identity (Gerdes et al, 2016; Lu et al, 2014; Römer et al, 2017; Wang et al, 2014). Interestingly, 70 the alterations caused by DGCR8+/- are conserved in both the naïve and primed stages of 71 pluripotency. Consistent with these findings, DGCR8 heterozygote cells display defects in self-72 renewal and impaired differentiation into the three primary major germ layers. Altogether, 73 these data indicate that DGCR8 has a significant role in the aetiology of 22qDS that is more 74 relevant than previously was revealed using mouse models of this human genetic disorder. 75 76

77 Characterization of DGCR8 heterozygosity in human pluripotent cellular models 78 22qDS is caused by a microdeletion in one chromosome 22, resulting in the hemizygosity of 79 around 40 protein coding genes (McDonald-McGinn et al. , 2015) . It is still unclear if the 80 disease originates from the haploinsufficiency of a small subset of these genes or from the 81 absence of the entire region . To investigate this, we assessed if the genes affected by the 82 microdeletion were predicted to be haploinsufficient by comparing their natural variation in 83 the human population using the Genome Aggregation Database (gnomAD) (Karczewski et al, 84 2020). Essential genes are predicted to have a very low frequency of loss-of-function mutations 85 in the general population, as these may be incompatible with life . When the frequency of 86 observed loss-of-function (LoF) mutations is lower than expected (obs/exp ≤ 0.089) the gene 87 is considered haploinsufficient (Karczewski et al., 2020). Obs/exp LoF ratios were plotted for 88 each of the genes affected by the most common microdeletion in 22qDS, and only 5 genes were 89 predicted to be haploinsufficient, including DGCR8 (Fig 1A ). Similar conclusions were 90 previously reported by (Karbarz, 2020) . To validate this prediction , we generated human 91 embryonic stem cells (H9 hESCs) and human embryonic carcinoma cells (PA-1 hECCs), where 92 a single copy of the DGCR8 gene was inactivated using the CRISPR/Cas9 nickase system. PA-93 1 cells are a diploid human embryonic teratocarcinoma cell line, which has retained limited 94 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 6 pluripotent capacity (Garcia-Perez et al , 2010; Zeuthen et al , 1980) . After targeting, two 95 different DGCR8 heterozygote clones (HET(1) and HET(2)) for each cell line were selected 96 for further studies ( Fig EV1A-EV1C). Inactivation of a single DGCR8 allele resulted in 97 reduced DGCR8 protein expression in both H9 and PA -1 cells, but also of DROSHA, as it 98 requires DGCR8 interaction for stabilisation (Han et al., 2009) (Fig 1B and EV1D-EV1F). 99 Despite the reduction in DGCR8 expression, H9 HET hESCs did not display obvious changes 100 in colony morphology ( Fig 1C) or in the expression of the typical pluripotency markers, 101 NANOG and TRA-1-60 (Fig 1D). Consistently, no differences were observed in the expression 102 of the other pluripotency markers, OCT4 and SOX2 (Fig 1E). Only KLF4 was slightly less 103 abundant in HET H9-HESCs, at both protein and RNA levels (Fig 1E and EV1F-EV1G). The 104 proportion of alkaline phosphatase expressing colonies was also similar between WT and HET 105 H9-hESCs cells (Fig EV2A). 106 In contrast, DGCR8 HET hESCs displayed a large reduction in their clonal expansion ability 107 when plated at low cell density or as single cells ( Fig 1F and EV2B-EV2C) suggesting poor 108 maintenance of self -renewal capacity. Consistent with this finding, both H9 and PA -1 HET 109 showed a decreased doubling time, confirming some proliferation defects (Fig EV2D). 110 Defective proliferation can result from a defect in cell cycle progression and/or increased 111 apoptosis. DGCR8 HET hESCs displayed delayed cell cycle progression, with a significant 112 accumulation in G0/G1 ( Fig EV2E), in addition to a significant increase in early (7 -AAD 113 negative and PE Annexin V positive) and late (7 -AAD positive and PE Annexin V positive) 114 apoptosis ( Fig 1G and EV2F). Importantly, these cellular phenotypes, including defective 115 proliferation, clonal expansion ability , increased apoptosis , and KLF4 misexpression were 116 reverted when DGCR8 expression was rescued in HET cells by lentiviral transduction (Fig 1F-117 1G, EV1F-EV1G and EV2C-EV2D, EV2F). All these together suggest that inactivati on of 118 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 7 one DGCR8 allele in human pluripotent cells results in defective self-renewal capacity, which 119 is characterized by increased apoptosis and cell cycle and proliferation defects. 120 121 DGCR8 heterozygous hESCs display differentiation defects 122 To evaluate if DGCR8 heterozygosity results in defects during human embryonic development, 123 WT and HET DGCR8 hESCs were differentiated in vitro using spontaneous embryoid body 124 (EB) formation. Remarkably, EBs formed by the HET clones were of smaller size in 125 comparison with WT cells, indicating differentiation and proliferation defects (Fig 2A). To 126 study the defects in differentiation, pluripotency and differentiation markers’ expression from 127 ectoderm, mesoderm and endoderm , was compared by RT-qPCR at day 0, 7, 14 and 21 of 128 differentiation. WT and HET EB differentiation resulted in a similar repression of the 129 pluripotency markers, NANOG and POU5F1. The expression of SOX2 was also similar 130 between WT and HET EBs (Fig 2B). Tested ectodermal markers ( OTX2, PAX6 and SOX1) 131 were also similarly increased during differentiation in both WT and HET hESCs. Only PAX6 132 displayed a subtle reduction at day 21 of differentiation (Fig 2C). In contrast, the expression 133 of most of the tested mesodermal ( CD34, FOXA2, TBXT and HAND1) and endodermal 134 (GATA6, HNF3, SOX7 and SOX17) markers was reduced upon differentiation of both clones 135 of HET hESCs, in comparison with WT ( Fig 2D-2E). These results suggest that DGCR8 136 heterozygosity in pluripotent human cells resulted in differentiation defects most markedly for 137 mesodermal and endodermal lineages. 138 To support these findings, directed differentiation protocols into ectoderm, mesoderm and 139 endoderm were performed followed by immunofluorescence of well -stablished markers for 140 these embryonic layers. This revealed that the average expression of the mesodermal marker 141 BRACHYURY (encoded by the TBXT) and the endodermal marker SOX17 was significantly 142 decreased, and nearly absent in a fraction of HET cells. A more homogeneous subtler reduction 143 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 8 of the ectodermal marker PAX6 was observed ( Fig 2F-2H), thus confirming the findings 144 observed upon EB differentiation. All these results suggest that DGCR8 heterozygosity leads 145 to decreased pluripotency, affecting the establishment of the three major embryonic lineages, 146 with a more pronounced defect in the mesodermal and endodermal germ layers. 147 148 DGCR8 heterozygous hESCs maintain cellular defects in a naïve-like state 149 Our findings demonstrate that the loss of a functional copy of DGCR8 results in cellular defects 150 in the biology of hESCs. Conversely, ablation of a single Dgcr8 allele in mESCs did not lead 151 to significant or clear phenotypes (Wang et al., 2007), raising the possibility that the defects 152 associated with human DGCR8 heterozygosity could be species-specific. Alternatively, these 153 findings could be attributed to the different pluripotency cellular states of hESCs, considered 154 to be in a primed state compared to mESCs, which represent a naïve state (Nichols & Smith, 155 2009). To rule out this possibility, WT and HET hESCs we re induced into a naïve-like state. 156 As a result of this transition, colonies acquired the typical domed morphology of naïve hESCs 157 (Fig 3A), preserving the reduced protein levels of both DGCR8 and DROSHA ( Fig 3B). To 158 confirm successful transition, upregulation of the na ïve pluripotency markers KLF17, 159 DNMT3L, DPPA3 and DPPA5, and silencing of the primed pluripotency marker, DUSP6 was 160 confirmed by qRT-PCR (Guo et al, 2017; Messmer et al, 2019; Theunissen et al, 2014) (Fig 161 3C). Despite the major differences in cellular phenotypes and gene expression profiles of naïve 162 versus primed pluripotent states, the main cellular defects , including increased apoptosis and 163 decreased colony formation capacity, were retained during the naïve stage ( Fig 3D-3G). All 164 these together suggest that the defects deriving from DGCR8 heterozygosity may be species-165 specific rather than cell -state-specific. These results prompted us to further characterise the 166 impact of DGCR8 heterozygosity at the molecular level. 167 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 9 Expression of primate-specific miRNAs is altered in DGCR8 heterozygous cells 168 Considering the defects in proliferation and differentiation of HET cells and the reduced levels 169 of DGCR8 and DROSHA, we next investigated if these phenotypes were linked to abnormal 170 expression of miRNAs. For this purpose, we performed miRNA expression analyses using 171 small RNA-seq of two clones of DGCR8 HET cells in H9 hESCs and one clone of PA-1 HET 172 cells vs their WT counterpart s. DGCR8 HET hESCs displayed a remarkable reduction in 173 mature miRNAs in a primed stage, with only a small proportion of miRNAs showing a modest 174 upregulation (Fig 4A, for complete list of significant differentially expressed miRNAs see 175 Table EV1). We also noted that a significant proportion of the common differentially 176 expressed miRNAs in both hESC HET clones were primate -specific (~30%) (Fig 4B and 177 EV3A). Dysregulated primate -specific miRNAs mostly belonged to the big miRNA cluster 178 C19MC (Fromm et al, 2022) (Fig 4C). The expression of the miR-371-3 cluster, homologous 179 to the miR-291-295 cluster in mouse, was also markedly reduced (Fig 4C). Interestingly, most 180 members of both clusters share the seed sequence ‘AAGUGC’, which have been previously 181 involved in self-renewal, proliferation, and apoptosis of hESCs (Kobayashi et al, 2022; Mong 182 et al, 2020; Teijeiro et al, 2018). The decreased expression of miRNAs belonging to these 183 clusters was validated by RT-qPCR in both naïve and primed H9 hESC clones and was rescued 184 upon reintroduction of DGCR8 (Fig 4D-4E). 185 To investigate the relevance of dysregulated primate -specific miRNA expression, we 186 performed pathway enrichment analyses with predicted mRNA targets of the primate-specific 187 miRNAs. Obtained pathways included ‘signalling pathways regulating pluripotency of stem 188 cells’ and pathways involved in pluripotency maintenance and self -renewal, as well as 189 embryonic development (e.g., ‘TGF -β signalling’, ‘Hippo signalling’ and ‘ErbB signalling’) 190 (Chan et al, 2002; James et al, 2005; Ramos & Camargo, 2012; Xu et al, 2016) (Fig 4F). 191 Enriched pathways for primate -specific miRNAs were similar to those obtained with all the 192 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 10 dysregulated miRNAs (10 out of the top 20 predicted pathways), (compare Fig 4F and EV3B, 193 see common pathways in black ). Despite the ontogenic differences between H9 -hESCs and 194 PA-1 cells and differences in their miRNA profile, we found that PA -1 HET cells also 195 displayed a similar proportion of miRNAs being differentially expressed, and a similar 196 proportion of those being primate -specific ( Fig EV3A and EV3C). Dysregulated miRNAs 197 were also predicted to regulate pathways involved in pluripotency maintenance and self -198 renewal (Fig EV3D and EV3E). These results highlight the potential importance of primate -199 specific miRNAs, as a subgroup of dysregulated miRNAs during DGCR8 haploinsufficiency. 200 201 Alterations in the transcriptome of DGCR8 HET cells 202 To investigate the impact of miRNA dysregulation on the gene expression programme of 203 DGCR8 HET cells, we performed total RNA high-throughput sequencing of HET H9 and PA-204 1 cells ( Fig EV4A and EV4B and Table EV2). Functional enrichment analyses of 205 differentially expressed genes (p-adj £ 0.05) revealed that affected pathways common to both 206 cell lines were related to development, including ‘embryonic organ development’ (Fig 4G and 207 EV4C). To understand if changes in gene expression were caused by defective miRNA levels, 208 we investigated the expression of the miRNA targets using the RNA-seq datasets. To this end, 209 we compared the expression of predicted targets for all dysregulated miRNAs (ALL), targets 210 for the subset of primate -specific dysregulated miRNAs (PS), versus non -predicted targets 211 (remaining genes). Changes in the expression of miRNA-predicted targets were significantly 212 different from non-target controls, both for targets of all dysregulated miRNAs and primate -213 specific miRNAs in PA -1 and H9 HET cells (p -val < 2.22e-16, Fig EV4D and EV4E). This 214 approach could not differentiate between genes predicted to be regulated by the up - or the 215 down-regulated miRNAs, as many contain predicted binding sites for both of the subgroups. 216 All these findings suggest that, in part, alterations in the gene expression profile of HET cells 217 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 11 can be attributed to altered post -transcriptional gene silencing. To better define the molecular 218 mechanisms contributing to the defects of HET cells, we next investigated if the 219 haploinsufficiency resulted in alterations in both well-defined canonical and non -canonical 220 functions of DGCR8. 221 222 Heterozygous cells display defects in both the canonical and non -canonical functions of 223 DGCR8 224 To investigate the impact of DGCR8 heterozygosity on its canonical function, the biogenesis 225 of miRNAs, we quantified Microprocessor cleavage efficiency both in vitro and in cells. For 226 in vitro purposes, total cell extracts from the three PA -1 cell lines, WT, HET and KO for 227 DGCR8, were prepared. Extracts were incubated with radiolabelled pri -miRNAs to visualise 228 precursor miRNA cleavage products as an indirect measurement of processing efficiency. We 229 observed that extracts derived from HET cells only retained partial processing activity when 230 compared to WTs, while KO extracts were not capable of processing pri -miRNAs (Fig 5A). 231 Next, to quantify pri-miRNA processing in cells, we measured the Microprocessor Processing 232 Index (MPI) using high -throughput sequencing data of chromatin -associated RNA for WT, 233 HET and KO PA-1 cells, as described in (Conrad et al, 2014; Witteveldt et al, 2018). For this 234 purpose, cells were fractionated in cytoplasmic, nucleoplasmic and chromatin fractions and 235 confirmed that chromatin was enriched for pri -miRNA transcripts (EV5A and EV5B). After 236 sequencing, the MPI for each pri-miRNA was calculated as the negative log2 fraction of reads 237 mapping to the hairpin versus reads mapping to the flanks of the pr imary miRNA. The higher 238 the value , the better processed the pri -miRNA, while values around 0 indicate absence of 239 processing. Both Microprocessor -dependent and independent pri -miRNAs behaved as 240 expected, with accumulation of reads over the hairpin of the Microprocessor-dependent pri-241 miRNA pri-miR-374b, in HET and KO cells, while no changes were observed for the DGCR8-242 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 12 independent miRNA pri-miR-1234 (Fig 5B, for more examples see EV5C). We next assessed 243 how the MPI is affected by DGCR8 heterozygosity and observed that, globally, PA -1 HET 244 cells displayed an intermediate processing efficiency when compared to WT and KO cells (Fig 245 5C, for a full list see Table EV3), suggesting that the biogenesis of miRNAs is affected in 246 these cells. 247 To explore the relationship between changes in pri -miRNA processing efficiency and mature 248 miRNAs, the levels of several pri-miRNA transcripts and mature miRNAs were compared by 249 RT-qPCR in WT and HET cells . Despite pri -miRNAs accumulation in HET cells, no 250 significant decrease in the mature miRNA levels was observed for some miRNAs, except for 251 miR-135b and miR -767 (Fig 5D and 5E). All t hese data together suggest that DGCR8 252 heterozygosity results in defects in miRNA biogenesis, both in vitro and in cells. However, 253 additional mechanisms may contribute to control the final mature miRNA levels. 254 Besides their canonical role in miRNA biogenesis , DGCR8 and Drosha have also been 255 suggested to regulate gene expression at the transcriptional level, independently of miRNAs. 256 DGCR8 and Drosha have been shown to interact with promoter -proximal regions of human 257 genes enhancing their transcription (Gromak et al, 2013). Furthermore, in an indirect manner, 258 DGCR8 has been shown to alter gene transcription by regulating heterochromatin formation 259 through physical association with KAP1 and HP1gamma (Deng et al, 2019). Both functions of 260 DGCR8 seem to be independent of the catalytical activity of Drosha. Thus, we next explored 261 whether changes in chromatin structures or accessibility could be associated with the 262 perturbation of gene expression observed in HET cells. For this purpose, we performed ATAC-263 seq (Assay for Transposase-Accessible Chromatin coupled to high-throughput sequencing) in 264 WT and HET PA -1 cells. As expe cted, an enrichment of ATA C-seq reads around the 265 transcription star t sites (TSS) was observed for both cell lines (see EV5D). Genome-wide 266 differential peak analysis identified 52,347 high-confidence peaks and revealed that there was 267 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 13 only a small proportion of peaks that were gained (0.36%, n=190) and approximately twice as 268 many were lost (0.6%, n=317) in HET PA -1 cells (Fig 5F). Most differential peaks were 269 located in introns and distal intergenic regions , followed by peaks annotated in proximal (≤ 1 270 kb) promoters (Fig 5G). In agreement with a role of chromatin accessibility in gene expression, 271 8% of genes containing a differential ATAC peak (+/-10Kb in distance) were also differentially 272 expressed, according to the RNA -seq analysis. Although small, this enrichment was 273 statistically significant (p-val=1.509e-9), suggesting that changes in the chromatin accessibility 274 of HET cells could also be influencing the gene expression profile. To further characterize the 275 functional impact of changes in the accessibility of regulatory regions, we used the rGREAT 276 package which implements the Genomic Regions Enrichment of Annotations Tool (GREAT) 277 (Gu & Hubschmann, 2023) . This analysis revealed that some of the most significant terms 278 associated with regions that lost accessibility in HET cells were linked to ‘development’, 279 ‘differentiation’ and ´morphogenesis” ( Fig 5H). All these findings together suggest that the 280 gene expression profile resulting from the loss of a single copy of DGCR8 could be a 281 combinatorial effect of both miRNA dysregulation and changes in chromatin accessibility. 282 283 DGCR8 haploinsufficiency reduces expression of primate -restricted endogenous 284 retrovirus HERVH and derived RNAs 285 Many transposable elements (TEs) are transcribed during early human embryogenesis in a 286 stage-specific manner and their expression is associated with stemness and pluripotency 287 maintenance (Gerdes et al., 2016; Torres-Padilla, 2020). For instance, knocking down the RNA 288 derived from the human endogenous retrovirus H (HERVH) or specific HERVH -derived 289 RNAs (e.g., chimeric transcripts driven by the ir promoter activity) result s in the loss of 290 pluripotency and self-renewal capacity of hESCs (Gerdes et al., 2016; Lu et al., 2014; Römer 291 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 14 et al. , 2017; Wang et al. , 2014) . To investigate if the stemness defects upon DGCR8 292 haploinsufficiency originated from aberrant TE expression, we used a pipeline that allows to 293 analyse TE expression at a locus specific level using the RNA-seq datasets from WT and HET 294 H9 hESCs [e.g., Software for Quantifying Interspersed Repeat Eleme nts (SQuIRE) (Yang et 295 al, 2019) . This analysis revealed that a high number of genomic locations annotated by 296 RepeatMasker as endogenous retroviruses type 1 (ERV1) were significantly downregulated in 297 both H9 hESC HET clones compared to WT cells (log2FC < -1; p-adj ≤ 0.05; Fig 6A). ERV1 298 retrotransposons have a structure resembling simple retroviruses , as they encode for gag and 299 pol genes, and are flanked by ~450bp Long Terminal Repeats (LTRs). However, genomic 300 analysis indicate that these retrotransposons are not currently active in the human genome 301 (Gerdes et al., 2016). Downregulated ERV1 loci from HET H9 hESCs mostly belonged to the 302 family members of the primate -specific endogenous retrovirus class H, HERVH, with reads 303 mapping to both their internal region (HERVH -int) and LTRs (known as LTR7) (89.2% and 304 70.49% of the ERV1 mapped reads in HET1 and HET2, respectively) ( Fig 6B, for locus -305 specific examples see Fig 6C, upper panels). 306 HERVH elements are typically expressed in pluripotent human cells. Indeed, nearly half of all 307 HERVH genomic copies (550 out of the 1225 full -length HERVH copies) are transcribed in 308 hESC, although only a relatively small subset of loci (n~117) is highly expressed (Wang et al., 309 2014). High expression of HERVHs in hESCs is mostly driven by LTR7 rather than its 310 counterparts, LTR7b, LTR7c or LTR7y (Wang et al. , 2014) . Remarkably, a significant 311 proportion of the downregulated HERVH elements in both HET H9 hESCs clones belonged to 312 the subgroup previously shown to be highly expressed in hESCs and mostly associated with 313 LTR7 promoter activity (formerly known as Type I subfamily) ( Table EV4). To validate if 314 HERVH transcripts were reduced in H9 hESC HET cells, we measured the expression of both, 315 total HERVH RNA levels and specific HERVH -derived transcripts. Using RT -qPCR, we 316 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 15 observed a global reduction in LTR7 and HERVH -int mRNAs in both naïve and primed 317 DGCR8 hESCs HET, which were rescued upon reintroduction of DGCR8 (Fig 6D and 6E). 318 Additionally, we observed a downregulation of specific HERVH -derived transcripts , as 319 defined by (Wang et al. , 2014) , including the hESCs-specific lncRNAs and chimeric 320 transcripts ( Fig 6C, lower panels , and 6F). Downregulated HERVH -derived transcripts , 321 including linc-ROR, ESRG and LINC00458, have also been previously associated with 322 pluripotency maintenance (Römer et al., 2017; Sexton et al, 2022). 323 Thus, these data strongly suggest that DGCR8 haploinsufficiency in pluripotent human cells 324 lead to misregulation of HERVH expression . Furthermore, these data further support that the 325 phenotype associated with DGCR8 heterozygosity is species-specific, and that in human cells 326 impacts expression of primate-specific miRNAs and HERVH retroelements. 327 328 C19MC and miR -371-373 miRNAs restore the molecular and cellular phenotype of 329 DGCR8 heterozygous cells. 330 Our results suggest that DGCR8 HET pluripotent human cells display two different primate-331 specific molecular phenotypes . First, the downregulation of primate -specific miRNAs and 332 second, HERVH derived transcripts , both of which are necessary for embryogenesis and 333 pluripotency maintenance. We next wanted to test if HERVH downregulation was a 334 consequence of the depletion of certain miRNAs, or if miRNAs and TEs were operating on 335 separate pathways. To this end, HET cells were transfected with two different pools of miRNAs 336 to test their ability to rescue HERVH expression. Full restoration of HERVH RNA levels was 337 observed after reintroduction of miRNAs belonging to the miR-371-3 cluster (hsa-miR-372-3p 338 and hsa-miR-373-3p), and a partial rescue was observed after transfection of four miRNAs 339 from C19MC cluster (hsa-miR-520g-3p, hsa-miR-520d-3p, hsa-miR-519c-3p, hsa-miR-515-340 5p) (Fig 7A). Notably, the protein levels of KLF4 were also restored upon overexpression of 341 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 16 miRNAs from both clusters (Fig 7B). Next, we assessed whether overexpression of those 342 miRNAs was also sufficient to restore some of the cellular phenotypes of HET cells. Indeed, 343 transient transfection with miRNAs from both clusters, either as single miRNAs or as pools, 344 re-established the clonal expansion capacity of DGCR8 HET cells (Fig 7C and 7D). 345 These findings indicate that the primate-specific miRNA cluster C19MC, along with the miR-346 371-373 cluster, play crucial role s in human embryonic stem cell maintenance. Also, we 347 showed that decreased KLF4 and HERVH RNA levels are a consequence of miRNA 348 dysregulation. However, we predict that all these factors are acting in concert and contribute 349 to the observed pluripotency defects of DGCR8 HET cells. 350 351

352 In this study, we have used two independent human pluripotent cellular models containing a 353 single functional DGCR8 allele to understand its relevance in the context of 22qDS. Our results 354 indicate that DGCR8 is haploinsufficient and that the most prominent defects are primate -355 specific. Previous attempts to study the consequences of DGCR8 haploinsufficiency were 356 performed in mice and led to conflicting conclusions. Mouse ESCs harbouring a single Dgcr8 357 gene do not show significant defects in miRNA levels or differentiation (Han et al., 2009; 358 Wang et al. , 2007) . Despite the apparently negligible consequences, Dgcr8 HET mice 359 displayed behavioural and neuronal def ects, which were attributed to altered miRNA 360 expression (Stark et al., 2008). In contrast to some of these findings in mice, our human models 361 showed defects in pluripotency and dysregulat ion of miRNA expression. Inconsistencies 362 between human and mouse models were also previously highlighted when comparing the 363 transcriptome of 22qDS-derived neurons and those derived from the mouse model (Df16A+/-), 364 where no overlap was found (Khan et al, 2020; Sun et al, 2018). We hypothesise that this 365 discrepancy is due to intrinsic differences between species. For instance, our results showed 366 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 17 that DGCR8 HET hESC s display increased apoptosis and defects in self-renewal, which 367 demonstrates the importance of DGCR8 for hESC survival . In agreement with this finding, 368 DGCR8 has also been identified as an essential gene for the survival of haploid hESCs (Yilmaz 369 et al , 2018) . In contrast, the total absence of Dgcr8 in mESCs results in a reduction in 370 proliferation but without changes in cell death or self -renewal ability (Wang et al. , 2007) . 371 Similarly, species -driven differences have been observed for DICER deficiency, another 372 essential factor for miRNA biogenesis. While hESCs require DICER1 for self -renewal, it 373 seems to be dispensable for mESC survival (Teijeiro et al., 2018). These discrepancies could 374 arise from differential sensitivities to unbalanced miRNA levels within the different 375 developmental stages that mouse and human ESCs represent (naïve versus primed states, 376 respectively). However, o ur results showed that hESCs maintain the main molecular and 377 cellular defects after induction into a naïve-like stage, arguing against a cell -state-specific 378 phenomena and supporting species-specific differences. In agreement with these findings, our 379

suggest that primate-specific miRNA dysregulation could be largely responsible for the 380 stemness defects in DGCR8 HET hESCs. We observed that a third of the miRNAs that are 381 affected by DGCR8 haploinsufficiency are primate-specific and not present in rodents , some 382 of which have been shown to be associated with pluripotency maintenance and self -renewal, 383 including the C19MC cluster (Lin et al, 2010; Mong et al., 2020). Ree et al has also reported 384 defects in C19MC miRNA processing, despite an inconsistent efficiency in targeting DGCR8 385 expression in human cells (Reé et al, 2022). The clonal expansion defects observed in DGCR8 386 HET hESCs could be rescued by reintroducing four independent miRNAs belonging to this 387 cluster, indicating it s contribution to proliferation. Interestingly, only two of the rescued 388 miRNAs, miR-520d-3p and miR -519c-3p, share the seed sequence with miR-371-373 389 suggesting that the functions of these clusters in human pluripotent cells are partially 390 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 18 redundant. Similar to our findings, Teijero et al. showed that reintroduction of miR-372-3p and 391 miR-373-3p rescues the increased apoptosis of DICER knockout hESCs (Teijeiro et al., 2018). 392 Importantly, human and mouse early development display remarkable differences, some of 393 which seem to be driven by species -specific expression of transposable elements (Gerdes et 394 al., 2016) . For instance, the murine endogenous retrovirus -L (MERVL) is transiently 395 upregulated at the two-cell stage and is essential for mouse preimplantation development 396 (Sakashita et al, 2023). The endogenous retrovirus that colonised the common ancestor of apes, 397 HERVH, is highly expressed in human pluripotent stem cells (hESCs and iPSCs) and epiblast, 398 where it appears to play a role in promoting self-renewal and pluripotency (Goke et al, 2015; 399 Grow et al , 2015; Wang et al. , 2014) . Notably, almost all HERVH elements expressed in 400 hESCs belong to a subfamily of elements transcribed from LTR7, which contains binding sites 401 for specific transcription factor s, including K LF4 (Carter et al , 2022; Ohnuki et al , 2014) . 402 Haploinsufficiency of DGCR8 led to a significant reduction of the HERVH/LTR7 subfamily 403 transcripts and KLF4 levels, and these were both restored upon reintroduction of miRNAs. Our 404 findings suggest that HERVH/LTR7 silencing is a consequence of miRNA downregulation and 405 could be potentially mediated by KLF4 knockdown. The reduction of this primate -specific 406 endogenous retrovirus, in both naïve and primed stages , may also contribute to some of the 407 cellular phenotypes of DGCR8 HET hESCs. These findings lead us to hypothesis e that 408 important primate -specific non -coding RNAs, including those derived from miRNAs and 409 transposable elements , may have co -evolved to orchestrate crucial aspects of pluripotency 410 maintenance in primates. 411 Our previous results showed that RNAs derived from other types of TEs (LINEs and SINEs) 412 are bound and processed by the Microprocessor (Heras et al., 2013). However, HET hESCs 413 showed no significant differences in the expression levels of these active retrotransposons , 414 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 19 suggesting that inactivating a single allele of DGCR8 is not sufficient to abolish the post-415 transcriptional control of LINE/SINE in pluripotent cells. 416 To identify which other functions of DGCR8 were altered in HET cells, we studied both the 417 canonical and less known functions. Similar to Stark et al. findings in mice (Stark et al., 2008), 418 human DGCR8 HET cells showed differential expression of a small proportion of miRNAs . 419 However, miRNA biogenesis defects did not seem to be directly correlated with mature 420 miRNA levels, indicating that additional factors may influence mature miRNA abundance, 421 including differences in the transcription of precursor pri -miRNAs. In addition to miRNA 422 biogenesis, DGCR8 has been implicated in stimulating RNA -pol II transcription, but also 423 promoting heterochromatin formation through interaction with KAP1 (Deng et al. , 2019; 424 Gromak et al., 2013). ATAC-seq analysis of DGCR8 HET cells only showed subtle changes 425 in chromatin accessibility. Interestingly, affected regions were associated with development 426 pathways, suggesting that changes at chromatin level could also be involved in some of the 427 cellular phenotypes characterised in HET cells. 428 Collectively, we show that DGCR8 results in haploinsufficiency by altering the expression of 429 primate-specific miRNAs, as well as of primate -specific transposable elements in human 430 pluripotent cells independently of the developmental cellular stage. These findings stress the 431 potential limitations of studying the function of human genes in evolutionary distant animal 432 models (e.g., rodents, zebrafish, etc) where a proportion of the genome, especially the non -433 coding genome, is only partially conserved. Particularly, our results suggest that DGCR8 could 434 have a more profound role in the developmental issues present in 22qDS to what it had been 435 previously suggested using mouse models and open new avenues to design novel targets for 436 rescuing some of the developmental defects in 22qDS patients. 437 438

439 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 20 Cell lines 440 All cell lines were grown at 37℃ and 5% CO 2. H9 hESCs were obtained from WiCell and 441 cultured in mTeSR1 media (STEMCELL Technolog ies) in plates coated with Matrigel 442 (Corning). Human PA-1 cells were cultured in MEM (Gibco) supplemented with GlutaMAX, 443 20% heat-inactivated FBS (Gibco), 100 U/mL penicillin -streptomycin (P/S, Invitrogen) and 444 0.1 mM Non-Essential Amino Acids (Gibco). iROCK (10 uM, Y-27632, Sigma) was added to 445 medium during the first 24 hours after splitting, followed by replacement with fresh media. 446 HEK293T cells were obtained from ATCC and culture d in high -glucose DMEM (Gibco) 447 supplemented with GlutaMAX, 10% FBS, (Hyclone) and 100 U/mL P/S. STR (Short tandem 448 repeat) analysis was carried out at the Genomic Unit (Genyo, Granada). 449 450 Generation of CRISPR-edited clonal cell lines 451 Guide RNAs (gRNA) A (GCACCACTGGACGTTTGCAG) and B 452 (GAGGTAATGGACGTTGGCTC) were designed to target exon 2 of DGCR8, after the start 453 codon, using the double nickase design from CRISPR Design Tool (http://tools.genome -454 engineering.org). gRNAs were cloned into pX461 (pSpCas9n(BB) -2A-GFP, Addgene ID # 455 48140) as in (Ran et al, 2013). For CRISPR targeting, H9 hESCs were maintained in E8 media 456 (DMEM/F12, L-ascorbic acid-2-phosphate magnesium (64 mg/l), sodium selenium (14 µg/l), 457 FGF2 (100 µg/l), insulin (19.4 mg/l), sodium bicarbonate (1.74 g/l), NaCl (5 mM) and holo -458 transferrin (10.6 mg/l) and TGFβ1 (1.8 µg/l)). Approximately, 2x106 hESCs were nucleofected 459 with 2 µg each of pX461 -sgRNA(A) and (B) using the V -Kit solution (Amaxa) and the A -23 460 program and seeded at a density of 2x103 in a Matrigel-coated plate, as in (Macia et al, 2017). 461 Control transfection with plasmid pMAX -EGFP (Amaxa) revealed that ~50% of cells were 462 GFP+ by fluorescence microscopy. Five days after nucleofection, cells were dissociated with 463 TrypLE and single cell clonal cell lines generated by limited dilution in 96-well coated plates. 464 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 21 iROCK inhibitor was added to the media during passaging to increase cell survival. After two 465 passages, a fraction of the cells was used for genomic DNA extraction using QuickExtract 466 DNA Solution (Lucigen). PCR from genomic DNA was performed using KAPA2G Fast 467 Hotstart Ready Mix PCR Kit (Kapa Biosystems). PCR products were cloned in pGEM-T Easy 468 Vector (Promega), and at least 10 clones were sent for Sanger sequencing for each cell line. 469 Two DGCR8+/- (HET) clones with a frameshift mutation in one allele were selected for further 470 studies. These clones were named H9 hESCs HET(1) and HET(2). For CRISPR targeting of 471 PA-1 cells, 1.25 µg of each pX461 -sgRNA(A) and (B) were co -transfected using 472 Lipofectamine 2000. GFP + cells were sorted 48h post -transfection using a FACSAria Cell 473 Sorter (BD) and seeded in 96-well plates. iROCK was added during passaging. Genomic DNA 474 sequencing of PA -1 clones was performed as described for H9 hESC clones. As no 475 heterozygote clones were obtained during the first round of targeting, one clone containing 476 frameshift mutations in both DGCR8 alleles (DGCR8-/- KO) was used to generate DGCR8+/- 477 (HET) cells by repairing one of the mutated alleles. To this end, 245 pmols of crRNA 478 (AGGTAATGGACGTTGGACGT), complementary to only one of the mutated alleles in the 479 KO cells, and 245 pmols of tracrRNA (IDT) were incubated in 25 µl nuclease free buffer for 5 480 min at 95℃ and allowed to anneal at room temperature (RT). The resulting gRNA was 481 incubated with 25 µg of Cas9 protein (IDT) for 15 -25 min at 37℃ prior to transfection. 1.2 x 482 106 PA-1 DGCR8-/- (KO) cells were resuspended in 80 µl of buffer T (Neon Transfection 483 system), and the Cas9/gRNA mixture and 300 pmol of the repair template were added to be 484 electroporated using 3 pulses of 1600V and 10 milliseconds. Single cell clones were obtained 485 by limiting dilution in 96 -well plates. To test successful gene editing, genomic DNA was 486 extracted by incubating cells in lysis buffer (30 mM Tris -HCl pH 8.0, 10 mM EDTA, 0.1% 487 SDS, 0.5% Tween-20 and 10 ug/ml Proteinase K) for 15 min at RT. Next, lysate was transferred 488 to 57℃ for 10 min followed by 98℃ for 10 min. From this mix, amplification of the sequence 489 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 22 of interest was performed by PCR , and products were cloned in pGEM -T Easy Vector . 490 Successful gene editing was confirmed by Sanger sequencing. Two clones were selected for 491 further studies and named PA-1 HET(1) and HET(2). All primers used are listed in Table 1. 492 493 Lentiviral transduction of hESCs 494 The lentiviral particles overexpressing human DGCR8, pLV-EF1α:hDGCR8 were purchased 495 from Vector Builder. As a lentiviral empty control vector , the promoter EF1 a and human 496 DGCR8 sequences were removed using the restriction enzymes FseI and BsBtI (N EB). After 497 ligation, the right sequence was confirmed by Sanger sequencing. Lentiviral control particles 498 were generated as in (Tristan-Manzano et al, 2023). Viral titres (transduction units [TU]/mL) 499 were calculated using qPCR. To this end, 1 x 10 5 HEK293T cells were transduced with 500 different volumes of the viral supernatant (1, 5 and 10 µl). 72 h post -transduction, genomic 501 DNA was isolated (QiAamp DNA miniKit, Qiagen) and the lentiviral copy number integrated 502 per cell was calculated using a standard curve method. Primers used are listed in Table 1. For 503 transduction, hESCs were dissociated and mixed with lentiviruses at a multiplicity of infection 504 (MOI) of 5 and seeded on 24 -well plates coated with Matrigel in mTeSR1 medium 505 supplemented with iROCK. After 24 hours, medium was replaced. Three days later, 2 μg/ml 506 puromycin selection was initiated , replacing media every 2 days. After 5 days, transduced 507 hESCs were grown under normal conditions. 508 509 Immunofluorescence 510 hESCs were seeded on Matrigel-coated coverslips and fixed with 4% paraformaldehyde (PFA) 511 during 5 min followed by permeabilization with 0.1% Triton X -100 in PBS for 5 min at RT. 512 Blocking was performed with 10% donkey serum (Merck Life Science) in PBS containing 513 0.5% Triton X -100, for 1h at RT. Cells were incubated overnight at 4°C with primary 514 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 23 antibodies diluted in PBS containing 1% donkey serum and 0.5% Triton X -100, followed by 515 three washes with 1% donkey serum in PBS. Incubation with secondary antibodies was 516 performed for 30 min at RT, followed by three additional washes with 1% donkey serum in 517 PBS. Cells were counterstained with DAPI (ProLong Gold antifade, Invitrogen) and Zeiss 518 LSM 710 Confocal Microscopy with a Plan-Apochromat 63x/1.40 Oil DIC M27 was used for 519 imaging. Image J was used for quantification. Primary antibodies against Nanog (1:1000, 500-520 P236, Prepotech), Tra -1-60 (1:100, 09 -0068, Stemgent), Brachyury (1:1600, #81694, CST), 521 Sox17 (1:200, AF1924, R&D Systems) and Pax6 (1:200, #60433, CST) were used. Secondary 522 antibodies included an anti -mouse IgG Alexa flour 488 (1:1000, A21202, Invitrogen), anti -523 rabbit IgG Alexa Fluor 555 (1:1000, A31572, Invitrogen) and anti -goat IgG Alexa Fluor 555 524 (1:1000, A21432, Invitrogen). 525 526 Western Blot 527 Cells were lysed in RIPA buffer supplemented with 1x EDTA-free Protease Inhibitor cocktail 528 (Roche), 1mM PMSF and 35 nM β-mercaptoethanol (Sigma). Protein was quantified using the 529 Micro BCA Protein Assay Kit (Thermo Fisher Scientific). Lysates were subjected to SDS -530 PAGE electrophoresis and transferred to PVDF membranes using Trans -Blot Turbo Transfer 531 System (Bio-Rad). Membranes were blocked using 5% non -fat dried milk in TBS -T (0.1% 532 Tween-20 in 1x TBS) or PBS-T (0.1% Tween-20 in 1x PBS) and incubated with primary and 533 secondary antibodies. Images were acquired using an ImageQuant LAS4000 or Chemidoc 534 Imaging System (Bio -Rad), fluorescent signal was quantified using ImageJ software. 535 Antibodies used include anti -DGCR8 (1:1000, ab90579, Abcam), anti -DROSHA (1:1000, 536 NBP1-03349, NovusBio). anti -SOX2 (1:2000, AB5603, Merck), anti -NANOG (1:500, 500 -537 P236, Peprotech) , anti-OCT4 (1:300, sc -8628, Santa Cruz Biotechnology) and anti -KLF4 538 (1:1000, 4038, Cell Signaling Technology). a-tubulin (1:1000, sc -23948, Santa Cruz 539 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 24 Biotechnology) and b-actin (1:15000, A1978, Sigma) antibodies were used as loading controls. 540 As secondary antibodies, anti-rabbit HRP (1:1000, 7074S, Cell Signaling Technology), anti-541 mouse HRP (1:1000, 7076S, Cell Signaling Technology) and anti -goat HRP (1:10000, 305 -542 035-003, Jackson ImmunoResearch). 543 544 Naïve-like hESCs induction 545 To revert the primed -state of HET and WT H9 hESCs to a naïve -like state, RSeT™ Feeder-546 Free Medium (STEMCELL Technologies) was used. Briefly, hESCs were cultured in a T 25 547 flask to 90% confluency. Next, cells were dissociated and detached into aggregates of 548 approximately 100-200 µm in diameter. A dilution (1:20) of the hESCs aggregates was seeded 549 in mTeSR1 medium supplemented with iROCK on Matrigel-coated 6-well plates. One day 550 after seeding, medium was replaced with RSeT™ Feeder-Free Medium and hESCs were grown 551 under hypoxic conditions (37ºC, 5% CO2 and 5% O2). A full-medium change was done every 552 2 days. After 5 days, hESCs were reverted into naïve-like hESCs. 553 554 Clonal expansion and cell proliferation assays 555 For clonal expansion assays, 13 x 103 WT, HET(1) and HET(2) H9 hESCs were seeded in 12-556 well plates coated with Matrigel in mTeSR1 medium. For naïve-like hESCs, 8 x 103 WT, HET 557 (1) and HET (2) were plated in 6 -well plates coated with Matrigel in RSeT Feeder -Free 558 Medium. After ten days for hESCs and five days for naïve -like hESCs, cells were fixed (37% 559 formaldehyde, 50% glutaraldehyde in 10× PBS) for 30 min and stained with 0.5 % crystal 560 violet for 40 min at room temperature. Stained areas and number of colonies were quantified 561 using ImageJ software. The area of naïve -like hESCs colonies was measure d using cellSens 562 Entry software. For cell proliferation assays, hESCs were seeded at a density of 36,500 cells 563 per well in 12-well plates coated with Matrigel and maintained for 10 days. Cells were counted 564 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 25 at days 3, 5, 7 and 10 using a Neubauer chamber . For PA-1 cells, growth rates of WT, HETs 565 and KO cells were compared by seeding 1x105 cells in 6-well plates. Cells were harvested and 566 counted using a hemocytometer before reseeding at day 2, 4 and 7. 567 568 Alkaline Phosphatase staining analysis 569 H9 hESCs were seeded at a density of 13,000 cells per well in 12 -well plate coated with 570 Matrigel for 6 days. Cells were fixed in 4% PFA for 2 min and stained with Alkaline 571 Phosphatase Detection Kit (Sigma -Aldrich). Colonies were manually counted distinguishing 572 between differentiated, mixed and undifferentiated colonies depending on the staining grade 573 and morphology. For single-cells assay, cells were seeded at a density of 0.5 cells/well in a 96-574 well plate coated with Matrigel and supplemented with cloneR (STEMCELL Technolog ies) 575 during the first 96 hours after plating. After 10 days, colonies were stained with alkaline 576 phosphatase detection kit. Next, colonies were stained with crystal violet to visualise negative 577 alkaline phosphatase cells. 578 579 Cell-cycle and apoptosis analyses 580 hESCs were detached using TrypLE and fixed in 70% cold ethanol, washed 3 times with ice 581 cold PBS and centrifuged at 450g for 5 min at 4ºC. Cell pellets were resuspended in propidium 582 iodide staining buffer (1x PBS, 0.05% NP -40, 3mM EDTA pH 8, 1 mg/ml RNaseA, 0.05 583 mg/ml propidium iodide) for 10 min at RT followed by 20 min on ice. Cells were analysed by 584 flow cytometry using FlowJo software. For apoptosis, cells were labelled with the PE Annexin 585 V Apoptosis kit (BD Biosciences) to distinguish between early and late apoptosis by flow -586 cytometry. Results were represented using BD FACSDiva™ Software. 587 588 H9 hESC differentiation 589 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 26 For EB differentiation, WT and HET H9 hESC s were cultured to 60% confluency. Next, 590 mTeSR1 medium containing Matrigel (1:6 ratio) was added to increase the thickness of the 591 colonies. At 80% confluency, cells were gently detached and cultured in suspension in ultra -592 low-attachment plates, allowing for EB formation during 21 days in medium, consisting of 593 DMEM Knockout (Gibco) supplemented with 20% FBS (Hyclone), 0.1 mM Non -Essential 594 Amino Acids (Gibco), 1 mM L -glutamine, and 0.1 mM β -mercaptoethanol. Media was 595 replaced every 2 days. For guided differentiations to ectoderm, mesoderm and endoderm 596 lineages STEMdiff Trilineage differentiation kit was used (STEMCELL Technologies). 597 Briefly, H9 hESCs were seeded in mTeSR1 medium supplemented with iROCK on Matrigel-598 coated coverslips in 24 -well plates. 400,000 cells per well were used for ectoderm and 599 endoderm differentiation, and 100,000 for mesoderm. One day after seeding, medium was 600 replaced with lineage -specific medium for 5 days (mesoderm and endoderm), or 7 days 601 (ectoderm). Medium was replaced every day. Cells were next fixed and processed for 602 immunofluorescence analyses as described above. 603 604 RT-qPCR 605 For RT-qPCR, total RNA was extracted from cells using Trizol, followed by RQ1 DNAse 606 treatment and phenol/chloroform purification. Next, 1µg of total RNA was further treated with 607 DNase I (Invitrogen), and cDNA was synthesised using High -Capacity cDNA Reverse 608 Transcription Kit (Applied Biosystems) and used for qPCR (GoTaq qPCR Mix, Promega). 609 Alternatively, cDNA was synthesised using Transcriptor Universal cDNA Master (Roche) and 610 qPCR was carried out with LightCycler 480 SYBR Green I Master mix (Roche). GAPDH or 611 ACTB were used as normalisers. Gene expression levels were quantified using the second 612 derivative method. Primers used are listed in Table 1. 613 614 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 27 miRNA mimics transfection 615 H9 WT and HET s (7x105) hESCs were seeded in 6 -well plates. After 24 hours, cells were 616 transfected with 60 nM of control mimic (4464058, ThermoFisher) or 30 nM of each miR-372-617 3p and 373-3p mimics (MC10165 and MC11024) or 15 nM of each miR-520g-3p, miR-520d-618 3p, miR-519c-3p and miR-515-5p mimics (MC10365, MC12807, MH10575 and MC10387, 619 ThermoFisher) using Lipofectamine 2000 (Life Technologies). 48 hours post -transfection, 620 total RNA was extracted or, alternatively, cells were fixed and stained for clonal expansion 621 assays. 622 623 Chromatin RNA-sequencing and Microprocessor Processing Index 624 PA-1 cells fractionated similar to (Conrad et al., 2014; Witteveldt et al., 2018). In brief, 8x106 625 cells were lysed in mild buffer (10 mM Tris pH 7.4; 150 mM NaCl;0.075% NP -40) and the 626 nuclei and cytoplasm were separated by sucrose gradient. The nucleic fraction was 627 subsequently separated into nucleoplasmic and chromatin -associated fractions as described 628 before, and the chromatin -associated fraction was sonicated on a Bioruptor (5 times 20 sec 629 on/off intervals) prior to DNase treatment using RQ1 DNase (Promega) and RNA extraction 630 using Trizol. Four biological replicate samples for each of the three PA-1 cell lines (WT, HET, 631 KO) were prepared and sent to BGI Genomics for library preparation and high -throughput 632 sequencing after rRNA depletion (2x 100 nt). Reads from chromatin-associated RNA samples 633 were aligned to the human genome (GRCh38.p13) using HISAT2 (v2.1.0) with the options --634 no-discordant --no-mixed --no-unal (Kim et al , 2019) . Human pre -miRNA locations were 635 determined by aligning human precursor sequences, obtained using the mature miRNA and 636 hairpin sequences from miRbase (v22.1) against the same genome using Bowtie2 with options 637 --very-sensitive --no-unal (Langmead & Salzberg, 2012) . The genomic locations of the pre -638 miRNAs plus 100 nts on each side were determined using bedtools getfasta -s. The read depth 639 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 28 for each nucleotide in the alignment of chromatin -associated reads on the appropriate strand 640 for each pre-miRNA was extracted using R. 641 A 10 nt gap between each pre -miRNA and its flanking regions was created by excluding the 642 outer 5 nts from the pre-miRNA and flanking regions, creating leeway for potential alternative 643 Drosha cleaving. For each region, the average read depth of all included nucleotides was used 644 for Microprocessor Processing Index (MPI) calculation. MPI was defined as the negative log2-645 transformed ratio between the mean read depth (RD) 646 of the pre -miRNA region (hairpin) to the mean read 647 depth of the regions flanking the pre -miRNA. In this 648 manner, a high MPI and > 0 indicates efficient 649 processing, and a MPI close to 0 or negative, indicates 650 inefficient or absence of processing. 651 To filter out Microprocessor-independent pri-miRNAs, reduce noise and exclude artifacts, pri-652 miRNAs were only included if they: (1) were no mirtrons or tailed mirtrons; (2) produce 653 miRNAs which were detected and included in statistical analysis for small RNA sequencing; 654 (3) on average, the flank depth in WT samples is ≥ 2; (4) the ratio between the two flanks did 655 not differ more than 4-fold. Differences in processing for individual pri -miRNA between cell 656 lines (ΔMPI) were computed by subtracting MPI values in DGCR8(WT) vs (HET) or (KO) 657 cells (ΔMPI= MPI(HET/KO) – MPI(WT)). A negative ΔMPI indicates that the pri-miRNA is 658 less processed in HET/KO compared to WT cells. 659 660 MiRNA quantification and small RNA high-throughput sequencing 661 MiRNA quantification in H9 hESCs was performed as described (Tristán-Ramos et al, 2020). 662 For PA-1 cells, 200 ng of total RNA was used for the retrotranscription (RT) reaction using the 663 miRCURY LNA miRNA PCR system (Qiagen). After dilution of the resulting cDNA, qPCR 664 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 29 was performed using the miRCURY LNA SYBR Green PCR Kit (Qiagen) with corresponding 665 miRNA primers on a LightCycler 480 Instrument (Roche). Both H9 and PA -1 miRNA 666 quantification data were normalised to the DGCR8 -independent miRNA, hsa-miR-320a. For 667 miRNA-specific primers see Table 1. 668 For small RNA high -throughput sequencing, RNA from three different biological replicates 669 for WT, HET(1) and HET(2) hESCs was extracted using mirVana microRNA isolation kit 670 (ThermoFisher). Small RNA libraries were generated using NEXTFlex Small RNA Library 671 Prep Kit v3 (Cat #NOVA-5132–06) and sequenced on the NextSeq 500 system (Illumina, CA, 672 USA) by the Genomic Unit at Genyo. For PA -1 cells, RNA from four different biological 673 replicates from WT and HET cells was extracted using the miRNeasy Mini Kit (Qiagen) 674 followed by on -column DNAse digestion. Small RNA sequencing libraries using unique 675 molecular identifiers (UMIs) were sequenced using DNB-seq (Li et al, 2019) by BGI. For each 676 sample, identical small RNA-seq reads were collapsed and only reads that were present more 677 than once were included in further analysis. MiRNAs were identified and counted using 678 miRDeep2, using the quantification function with human hairpin and mature miRNA 679 sequences files from miRbase ( v22.1) and allowing no mismatches (Friedlander et al, 2012; 680 Kozomara et al, 2019; Kozomara & Griffiths -Jones, 2014). DESeq2 was used for statistical 681 analysis, comparing WT vs HET in H9 and PA -1 separately, and using apeglm as log fold -682 change shrinkage model (Love et al, 2014; Zhu et al, 2019). 683 684 In silico miRNA target prediction and pathway analysis 685 Differentially expressed miRNAs ( p-adj ≤ 0.05) that are common to both H9 hESCs HET(1) 686 and HET(2) (85 miRNAs) and differentially expressed miRNAs in PA -1 HET (155miRNAs) 687 underwent functional enrichment analysis with DIANA-mirPATH v3 software (Vlachos et al, 688 2015) (for PA-1 HET, only the top 100 miRNAs with the highest absolute log2FC were used). 689 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 30 MicroT-CDS prediction algorithm was used to identify putative mRNA targets of miRNAs and 690 associated significantly enriched (FDR ≤ 0.05) KEGG pathways (Kanehisa et al , 2014) 691 identified. Dot plots were generated using ggplot2 (v3.3.5) and DOSE v3.14.0 (Yu et al, 2015) 692 R packages. 693 Total RNA high-throughput sequencing and gene ontology analyses 694 Total RNA from three biological replicates of H9 (WT, HET(1) and HET(2)) and four 695 biological replicates of PA -1 (WT and HET) cells was extracted using miRNeasy Mini Kit 696 (Qiagen) followed by on-column DNAse digestion. Purified RNA was rRNA-depleted prior to 697 sequencing (DNB-seq) by BGI. For PA-1 RNA-seq analyses, paired-end reads were aligned to 698 the human genome (GRCh38.p13) using HISAT2 with options --no-discordant --no-mixed --699 no-unal (Kim et al., 2019). Transcript counts for each sample were created with featureCounts 700 (Rsubread v2.2.6), using reverse counts and excluding reads that aligned to the genome 701 multiple times (Liao et al, 2014). For H9 RNA-seq analyses, quality of individual H9 WT and 702 HET sequences were evaluated using FastQC v0.11.5 software (Andrews, 2010). Paired-end 703 reads were aligned to GRCh38.p13 human genome assembly with STAR v2.7.6a (Dobin et al, 704 2013) and quantified with featureCounts v2.0.1 (Liao et al., 2014) using NCBI Annotation 705 Release 109. For both H9 and PA -1, differential expression analysis and count normalization 706 was performed with the R package DESeq2 v1.28 (Love et al. , 2014) . After differential 707 analysis, DESeq2's apeglm lfcShrink was applied to shrink log2 fold changes (Love et al., 2014; 708 Zhu et al., 2019). 709 Functional enrichment analysis was carried out for differentially expressed genes (p-adj ≤ 0.05) 710 using the enrichGO function in R package clusterProfiler v3.16.1 (Yu et al, 2012), and using 711 the Gene Ontology Biological Processes (GO-BP) gene sets. 712 713 Analysis of primate-specific C19MC cluster 714 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 31 A list of primate-specific miRNAs was obtained from MirGeneDB 2.1(Fromm et al., 2022) . 715 R package miRBaseConverter v1.14.0 (Haunsberger et al, 2017) was used to adapt this list to 716 mirBase v22.1 used for differential expression analysis. All miRNAs with padj ≤ 0.05 were 717 considered as dysregulated. Common primate-specific dysregulated miRNAs in H9 HET (34 718 miRNAs) and PA -1 HET cells (47 miRNAs) were analysed using DIANA -mirPATH v3 719 software. Statistical significance for the enrichment of primate-specific dysregulated miRNAs 720 was calculated using Fisher’s exact test (phyper function of R package stats, v4.2.1). 721 722 Transposable element expression analyses 723 The SQuIRE v0.9.9.92 pipeline (Yang et al., 2019) was used to measure transposable elements 724 expression changes using default parameters. SQuIRE downloads TE annotations from 725 RepeatMasker and uses an expectation-maximization algorithm to assign multimapping reads. 726 Next, it performs TE differential expression using DESeq2, either by grouped TE subfamilies 727 or by analysing individual loci. Expression of individual TEs was represented using ggplot2 728 (v3.3.5) and gghalves (v0.1.1) (Tiedemann, 2020) R packages. hESCs-specific chimeric 729 transcripts and lncRNAs derived from HERVH were extracted from (Wang et al., 2014) and 730 all gene symbols were updated using the R package HGNChelper v0.8.1 (Oh et al, 2020). This 731 gene list was compared with differentially expressed genes (p-adj ≤ 0.05) from hESCs HET(1) 732 and HET(2). Heatmap was generated using the R package ComplexHeatmap v2.4.3 (Gu et al, 733 2016) applied to Z-score for each gene and volcano graphs using ggplot2 (v3.3.5). Normalized 734 bigwig files were generated using Draw tool from SQuIRE and visualized using IGV. 735 736 Assay for Transposase Accessible Chromatin with sequencing (ATAC-seq) 737 ATAC-seq libraries were prepared as previously described (Buenrostro et al, 2015). Around 738 50,000 cells were used for each biological replicate (n=3). Cells were lysed in 50 µL cold lysis 739 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 32 buffer (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 3 mM MgCl2 and 0.1% IGEPAL CA-360) and 740 pelleted at 500g for 10 mins at 4°C. Pellets were resuspended in 50 µL transposition reaction 741 mix as follows: 2X Tagment DNA buffer (Illumina 15027866), 20X Tagment DNA enzyme 742 (Illumina 15027865) and incubated at 37°C for 30 mins. Transposed samples were purified 743 using the MinElute PCR purification kit (Qiagen 28204), and eluted in 10 µL. Transposed 744 DNA samples were amplified by PCR by setting up a 50 µL reaction as follows: 10 µL 745 transposed DNA, 9.7 µL ddH2O, 2.5 µL 25 µM customised Nextera PCR primer FW, 2.5 µL 746 25 µM customised Nextera PCR primer RV, 0.3 µL 100X SYBR Green I (Invitrogen S-7563) 747 and 25 µL 2X NEBNext high-fidelity PCR master mix (NEB, M0541). To reduce GC/size bias 748 and over amplification of libraries, the reaction was monitored by qPCR to stop amplification 749 prior to saturation. A qPCR 15 µL side -reaction was set up as follows: 5 µL 5 cycles PCR 750 amplified DNA, 4.44 µL ddH2O, 0.25 µL 25 µM customised Nextera PCR primer FW, 0.25 751 µL 25 µM customised Nextera PCR primer RV, 0.06 µL 100X SYBR Green I and 5 µL 2X 752 NEBNext high-fidelity PCR master mix. The additional number of cycles required for each 753 sample was determined by plotting a linear run v s. the cycle number. The number of cycles 754 that correspond to ¼ maximum fluorescent intensity was calculated for each sample. The 755 remaining 45 µL 5 cycles PCR amplified DNA was run as before (omitting 72°C initial step 756 and modifying the number of cycles to the calculated amount). Amplified DNA samples were 757 subjected to double size selection to remove DNA fragments 1000 758 bp that would hinder sequencing reactions. Samples were diluted up to 90 µL with ddH2O and 759 purified using the SPRIselect beads (Beckman Coulter B23317) . For removal of large DNA 760 fragments, a ratio of 0.55 DNA/beads slurry was used, and for removal of small DNA 761 fragments, a ratio of 1.8 was used. Purified DNA samples were quantified using the Qubit 762 dsDNA high sensitivity assay (ThermoFisher Q32854) with the Qubit 4 fluorometer. The 763 quality of DNA samples was analysed using the high sensitivity DNA kit (Agilent 5067-4626) 764 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 33 in Bioanalyzer. Samples with the nucleosomal fragment distribution profile expected for 765 ATAC-seq libraries (mono-, di- and trisomal fragments) were sent for sequencing. 766 Samples were sequenced on an Illumina HiSeq 4000 platform to obtain 50 bp paired-end reads 767 at the Wellcome Trust Clinical Research Facility (University of Edinburgh). Reads were 768 trimmed using cutadapt v3.5 paired-end trimming, and aligned to the hg38 human genome 769 using bowtie2 v2.4.4 (Langmead & Salzberg, 2012) paired-end alignment with options --very-770 sensitive --no-mixed --no-discordant -X 2000. Unmapped reads and those mapping to the 771 mitochondrial genome were removed and duplicate reads were filtered out using Picard 2.27.5 772 MarkDuplicates (http://broadinstitute.github.io/picard/). Reads were shifted by +4 bp for those 773 mapping to the positive strand and −5 bp for those mapping to the negative strand using 774 alignmentSieve tool from deepTools package. Broad peaks were called using MACS2 2.2.7.1 775 callpeak (Zhang et al, 2008) with options -g hs -f BED --keep-dup all -q 0.01 --nomodel --shift 776 -75 --extsize 150 --broad. Peaks overlapping blacklisted regions were removed using bedtools. 777 A union peak set across all samples was obtained following the iterative overlap peak merging 778 procedure described on (Grandi et al , 2022) (code provided in 779 https://github.com/corceslab/ATAC_IterativeOverlapPeakMerging). A count matrix over the 780 union peak set was computed using featureCounts 2.0.1 (Liao et al. , 2014) and differently 781 expressed peaks were obtained using DESeq2 1.28 R package (Love et al., 2014) and selecting 782 dysregulated peaks using p -adj ≤ 0.05 and abs( log2FC) ≥ 1. Peaks were annotated using 783 annotatePeaks.pl from Homer v4.11 package (Heinz et al , 2010) . Functional analysis of 784 differential ATAC-seq regions was carried out with rGREAT package (Gu & Hubschmann, 785 2023). Motif analysis was performed with findMotifsGenome.pl tool from Homer v4.11 786 package using default parameters and random background selection. Metagene plots were 787 generated using computeMatrix and plotProfile tools from deepTools. Genes contained in an 788 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 34 interval of ± 10kb from differentially expressed PA1 HET ATAC -seq peaks were obtained 789 using annotatePeakInBatch function in ChIPpeakAnno R package(Zhu et al, 2010). 790 791 Data and code availability 792 All PA -1 sequencing data are deposited in GEO database , accession number GSE197474 793 (token: qbulgmakldgrbap). ATAC -seq datasets are deposited in GEO , accession number 794 GSE205798 (token: ivaduicidhwnlor). H9 data are deposited in GEO , accession number 795 GSE209843 (token: obsnscqwbfshfuz). Bioinformatic and software packages are described in 796 the STAR Methods sections. 797 798

799 The work in the lab of S.R.H was supported by Grant PID2020 -115033RB-I00 funded by 800 MCIN/AEI/10.13039/501100011033, grants PEJ2018 -003280-A and RYC -2016-21395 801 funded both by AEI /10.13039/501100011033 and by “ESF investing in your future ”, Career 802 Integration Grant -Marie Curie (FP7-PEOPLE-2011-CIG-303812), FEDER/Junta de 803 Andalucía-Consejería de Transformación Económica, Industria, Conocimiento y 804 Universidades (PY20_00619 y A-CTS-28_UGR20 grants) and donation to “Aula de estudios 805 22qDS”. The work in S.M laboratory was funded by the Wellcome Trust grants 806 (221737/Z/20/Z and 107665/Z/15/Z), the Royal Society grant (RGS \R1\191368) and the 807 Wellcome Trust iTPA (PIII021). L.K. was funded by an MRC-Precision Medicine fellowship 808 and P.C was funded by a Darwin Trust fellowship. A.G.-G was supported by grant PRE2021-809 098878 funded by MCIN/AEI/10.13039/501100011033. J.L.G.-P. acknowledges funding from 810 ERC ( ERC-Consolidator ERC -STG-2012-309433), the Government of Spain (MINECO -811 FEDER SAF2017-89745-R and PID2021 -128934NB-I00), the Andalusian regional 812 Government ( PAIDI P12 -CTS-2256 and P18 -RT-5067) and a private donation from Ms 813 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 35 Francisca Serrano (Trading y Bolsa para Torpes, Granada, Spain). We are grateful to Javier F. 814 Caceres for comments and critical reading of the manuscript. We thank Gregorz Kudla for 815 initial bioinformatic analysis. Paul de Sousa and Rosa Montes for advice on hESC biology and 816 techniques; Marta García -Canadas, Jennifer Parra, Esther Prada , Meriem Benkaddour -817 Boumzaouad and the Genyo’s microscopy unit for technical support. We thank the Francisco 818 Martin’s group (Genyo, Granada) for help with the lentiviral transduction experiments. We 819 would like to specially thank the Association s of 22q11.2 patients in Andalucia and Levante 820 for their support and trust. 821 822 Author contributions 823 S.M. and S.R.H. conceived and supervised this study. A.C.B., L.K., S.M. and S.R.H designed 824 and interpreted the experiments. A.C.B . performed most experiments with hESC. L.K. 825 performed most experiments with PA -1 cells and performed data analyses for transcriptomic 826 approaches, under the supervision of A.I. and S.M. G.P. provided additional data analysis for 827 hESC and PA -1 datasets. P.C . and K.G. performed experimental validation with PA -1 cells. 828 L.S. provided technical support and performed experiments. S.P. performed and analysed 829 ATAC-seq experiment in the group of R.H. P.T.R generated DGCR8 knockout PA1 cells and 830 DGCR8 heterozygous hESCs. A.G.G . contributed to HERVH analysis. J.L.G.P contributed 831 with resources to generate DGCR8 heterozygous hESCs. G.B. performed preliminary 832 bioinformatic analysis. S.M. and S.R.H wrote the original draft and all the authors contributed 833 to the final version. 834 835 Competing Interest Statement 836 The authors declare no competing interests. 837 838 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 36

839 Andrews S (2010) FASTQC. A quality control tool for high throughput sequence data 840 https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ 841 842 Bezman NA, Cedars E, Steiner DF, Blelloch R, Hesslein DG, Lanier LL (2010) Distinct 843 requirements of microRNAs in NK cell activation, survival, and function. J Immunol 185: 844 3835-3846 845 Brandl A, Daum P, Brenner S, Schulz SR, Yap DY-H, Bösl MR, Wittmann J, Schuh W, Jäck 846 H-M (2016) The microprocessor component, DGCR8, is essential for early B-cell 847 development in mice. European Journal of Immunology 46: 2710-2718 848 Buenrostro JD, Wu B, Chang HY, Greenleaf WJ (2015) ATAC‐seq: A Method for Assaying 849 Chromatin Accessibility Genome‐Wide. Current Protocols in Molecular Biology 109 850 Carter TA, Singh M, Dumbović G, Chobirko JD, Rinn JL, Feschotte C (2022) Mosaic cis-851 regulatory evolution drives transcriptional partitioning of HERVH endogenous retrovirus in 852 the human embryo. eLife 11 853 Chan R, Hardy WR, Laing MA, Hardy SE, Muller WJ (2002) The Catalytic Activity of the 854 ErbB-2 Receptor Tyrosine Kinase Is Essential for Embryonic Development. Molecular and 855 Cellular Biology 22: 1073-1078 856 Conrad T, Marsico A, Gehre M, Orom UA (2014) Microprocessor activity controls 857 differential miRNA biogenesis In Vivo. Cell Rep 9: 542-554 858 David, Molly, Joyce, Yang Z, Joshua, Christopher, Matloubian M, Blelloch R, K (2011) 859 MicroRNA-29 Regulates T-Box Transcription Factors and Interferon-γ Production in Helper 860 T Cells. Immunity 35: 169-181 861 Deng L, Ren R, Liu Z, Song M, Li J, Wu Z, Ren X, Fu L, Li W, Zhang W et al (2019) 862 Stabilizing heterochromatin by DGCR8 alleviates senescence and osteoarthritis. Nature 863 Communications 10 864 Denli AM, Tops BBJ, Plasterk RHA, Ketting RF, Hannon GJ (2004) Processing of primary 865 microRNAs by the Microprocessor complex. Nature 432: 231-235 866 Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, 867 Gingeras TR (2013) STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29: 15-21 868 Earls LR, Fricke RG, Yu J, Berry RB, Baldwin LT, Zakharenko SS (2012) Age-Dependent 869 MicroRNA Control of Synaptic Plasticity in 22q11 Deletion Syndrome and Schizophrenia. 870 Journal of Neuroscience 32: 14132-14144 871 Fenelon K, Mukai J, Xu B, Hsu PK, Drew LJ, Karayiorgou M, Fischbach GD, Macdermott 872 AB, Gogos JA (2011) Deficiency of Dgcr8, a gene disrupted by the 22q11.2 microdeletion, 873

in altered short-term plasticity in the prefrontal cortex. Proc Natl Acad Sci U S A 108: 874 4447-4452 875 Friedlander MR, Mackowiak SD, Li N, Chen W, Rajewsky N (2012) miRDeep2 accurately 876 identifies known and hundreds of novel microRNA genes in seven animal clades. Nucleic 877 Acids Res 40: 37-52 878 Fromm B, Hoye E, Domanska D, Zhong X, Aparicio-Puerta E, Ovchinnikov V, Umu SU, 879 Chabot PJ, Kang W, Aslanzadeh M et al (2022) MirGeneDB 2.1: toward a complete 880 sampling of all major animal phyla. Nucleic Acids Res 50: D204-D210 881 Garcia-Perez JL, Morell M, Scheys JO, Kulpa DA, Morell S, Carter CC, Hammer GD, 882 Collins KL, O'Shea KS, Menendez P et al (2010) Epigenetic silencing of engineered L1 883 retrotransposition events in human embryonic carcinoma cells. Nature 466: 769-773 884 Gebert LFR, Macrae IJ (2019) Regulation of microRNA function in animals. Nature Reviews 885 Molecular Cell Biology 20: 21-37 886 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 37 Gerdes P, Richardson SR, Mager DL, Faulkner GJ (2016) Transposable elements in the 887 mammalian embryo: pioneers surviving through stealth and service. Genome Biol 17: 100 888 Goke J, Lu X, Chan YS, Ng HH, Ly LH, Sachs F, Szczerbinska I (2015) Dynamic 889 transcription of distinct classes of endogenous retroviral elements marks specific populations 890 of early human embryonic cells. Cell Stem Cell 16: 135-141 891 Grandi FC, Modi H, Kampman L, Corces MR (2022) Chromatin accessibility profiling by 892 ATAC-seq. Nat Protoc 17: 1518-1552 893 Gregory RI, Yan K-P, Amuthan G, Chendrimada T, Doratotaj B, Cooch N, Shiekhattar R 894 (2004) The Microprocessor complex mediates the genesis of microRNAs. Nature 432: 235-895 240 896 Gromak N, Dienstbier M, Macias S, Plass M, Eyras E, Caceres JF, Proudfoot NJ (2013) 897 Drosha regulates gene expression independently of RNA cleavage function. Cell Rep 5: 898 1499-1510 899 Grow EJ, Flynn RA, Chavez SL, Bayless NL, Wossidlo M, Wesche DJ, Martin L, Ware CB, 900 Blish CA, Chang HY et al (2015) Intrinsic retroviral reactivation in human preimplantation 901 embryos and pluripotent cells. Nature 522: 221-225 902 Gu Z, Eils R, Schlesner M (2016) Complex heatmaps reveal patterns and correlations in 903 multidimensional genomic data. Bioinformatics 32: 2847-2849 904 Gu Z, Hubschmann D (2023) rGREAT: an R/bioconductor package for functional enrichment 905 on genomic regions. Bioinformatics 39 906 Guo G, von Meyenn F, Rostovskaya M, Clarke J, Dietmann S, Baker D, Sahakyan A, Myers 907 S, Bertone P, Reik W et al (2017) Epigenetic resetting of human pluripotency. Development 908 144: 2748-2763 909 Han J, Lee Y, Yeom KH, Kim YK, Jin H, Kim VN (2004) The Drosha-DGCR8 complex in 910 primary microRNA processing. Genes Dev 18: 3016-3027 911 Han J, Pedersen JS, Kwon SC, Belair CD, Kim Y-K, Yeom K-H, Yang W-Y, Haussler D, 912 Blelloch R, Kim VN (2009) Posttranscriptional Crossregulation between Drosha and 913 DGCR8. Cell 136: 75-84 914 Haunsberger SJ, Connolly NM, Prehn JH (2017) miRNAmeConverter: an R/bioconductor 915 package for translating mature miRNA names to different miRBase versions. Bioinformatics 916 33: 592-593 917 Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng JX, Murre C, Singh H, 918 Glass CK (2010) Simple combinations of lineage-determining transcription factors prime cis-919 regulatory elements required for macrophage and B cell identities. Mol Cell 38: 576-589 920 Heras SR, Macias S, Plass M, Fernandez N, Cano D, Eyras E, Garcia-Perez JL, Caceres JF 921 (2013) The Microprocessor controls the activity of mammalian retrotransposons. Nat Struct 922 Mol Biol 20: 1173-1181 923 James D, Levine AJ, Besser D, Hemmati-Brivanlou A (2005) TGFβ/activin/nodal signaling 924 is necessary for the maintenance of pluripotency in human embryonic stem cells. 925 Development 132: 1273-1282 926 Kanehisa M, Goto S, Sato Y, Kawashima M, Furumichi M, Tanabe M (2014) Data, 927 information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res 42: 928 D199-205 929 Karbarz M (2020) Consequences of 22q11.2 Microdeletion on the Genome, Individual and 930 Population Levels. Genes 11: 977 931 Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, Collins RL, 932 Laricchia KM, Ganna A, Birnbaum DP et al (2020) The mutational constraint spectrum 933 quantified from variation in 141,456 humans. Nature 581: 434-443 934 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 38 Khan IS, Taniguchi RT, Fasano KJ, Anderson MS, Jeker LT (2014) Canonical micro RNA s 935 in thymic epithelial cells promote central tolerance. European Journal of Immunology 44: 936 1313-1319 937 Khan TA, Revah O, Gordon A, Yoon S-J, Krawisz AK, Goold C, Sun Y, Kim CH, Tian Y, 938 Li M-Y et al (2020) Neuronal defects in a human cellular model of 22q11.2 deletion 939 syndrome. Nature Medicine 26: 1888-1898 940 Kim D, Paggi JM, Park C, Bennett C, Salzberg SL (2019) Graph-based genome alignment 941 and genotyping with HISAT2 and HISAT-genotype. Nature Biotechnology 37: 907-915 942 Kim YS, Kim H-R, Kim H, Yang SC, Park M, Yoon JA, Lim HJ, Hong S-H, Demayo FJ, 943 Lydon JP et al (2016) Deficiency in DGCR8-dependent canonical microRNAs causes 944 infertility due to multiple abnormalities during uterine development in mice. Scientific 945 Reports 6: 20242 946 Knuckles P, Vogt MA, Lugert S, Milo M, Chong MMW, Hautbergue GM, Wilson SA, 947 Littman DR, Taylor V (2012) Drosha regulates neurogenesis by controlling Neurogenin 2 948 expression independent of microRNAs. Nature Neuroscience 15: 962-969 949 Kobayashi N, Okae H, Hiura H, Kubota N, Kobayashi EH, Shibata S, Oike A, Hori T, 950 Kikutake C, Hamada H et al (2022) The microRNA cluster C19MC confers differentiation 951 potential into trophoblast lineages upon human pluripotent stem cells. Nature 952 Communications 13 953 Kozomara A, Birgaoanu M, Griffiths-Jones S (2019) miRBase: from microRNA sequences to 954 function. Nucleic Acids Res 47: D155-D162 955 Kozomara A, Griffiths-Jones S (2014) miRBase: annotating high confidence microRNAs 956 using deep sequencing data. Nucleic Acids Res 42: D68-73 957 Landthaler M, Yalcin A, Tuschl T (2004) The Human DiGeorge Syndrome Critical Region 958 Gene 8 and Its D. melanogaster Homolog Are Required for miRNA Biogenesis. Current 959 Biology 14: 2162-2167 960 Langmead B, Salzberg SL (2012) Fast gapped-read alignment with Bowtie 2. Nature 961

9: 357-359 962 Li Q, Zhao X, Zhang W, Wang L, Wang J, Xu D, Mei Z, Liu Q, Du S, Li Z et al (2019) 963 Reliable multiplex sequencing with rare index mis-assignment on DNB-based NGS platform. 964 BMC Genomics 20 965 Liao Y, Smyth GK, Shi W (2014) featureCounts: an efficient general purpose program for 966 assigning sequence reads to genomic features. Bioinformatics 30: 923-930 967 Lin S, Cheung WK, Chen S, Lu G, Wang Z, Xie D, Li K, Lin MC, Kung HF (2010) 968 Computational identification and characterization of primate-specific microRNAs in human 969 genome. Comput Biol Chem 34: 232-241 970 Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for 971 RNA-seq data with DESeq2. Genome Biology 15 972 Lu X, Sachs F, Ramsay L, Jacques PE, Goke J, Bourque G, Ng HH (2014) The retrovirus 973 HERVH is a long noncoding RNA required for human embryonic stem cell identity. Nat 974 Struct Mol Biol 21: 423-425 975 Macia A, Widmann TJ, Heras SR, Ayllon V, Sanchez L, Benkaddour-Boumzaouad M, 976 Muñoz-Lopez M, Rubio A, Amador-Cubero S, Blanco-Jimenez E et al (2017) Engineered 977 LINE-1 retrotransposition in nondividing human neurons. Genome Research 27: 335-348 978 Macias S, Plass M, Stajuda A, Michlewski G, Eyras E, Caceres JF (2012) DGCR8 HITS-979 CLIP reveals novel functions for the Microprocessor. Nat Struct Mol Biol 19: 760-766 980 Marinaro F, Marzi MJ, Hoffmann N, Amin H, Pelizzoli R, Niola F, Nicassio F, De Pietri 981 Tonelli D (2017) MicroRNA-independent functions of DGCR8 are essential for neocortical 982 development and TBR1 expression. EMBO Rep 18: 603-618 983 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 39 McDonald-McGinn DM, Sullivan KE, Marino B, Philip N, Swillen A, Vorstman JA, Zackai 984 EH, Emanuel BS, Vermeesch JR, Morrow BE et al (2015) 22q11.2 deletion syndrome. Nat 985 Rev Dis Primers 1: 15071 986 Messmer T, von Meyenn F, Savino A, Santos F, Mohammed H, Lun ATL, Marioni JC, Reik 987 W (2019) Transcriptional Heterogeneity in Naive and Primed Human Pluripotent Stem Cells 988 at Single-Cell Resolution. Cell Rep 26: 815-824 e814 989 Mong EF, Yang Y, Akat KM, Canfield J, VanWye J, Lockhart J, Tsibris JCM, Schatz F, 990 Lockwood CJ, Tuschl T et al (2020) Chromosome 19 microRNA cluster enhances cell 991 reprogramming by inhibiting epithelial-to-mesenchymal transition. Sci Rep 10: 3029 992 Nichols J, Smith A (2009) Naive and primed pluripotent states. Cell Stem Cell 4: 487-492 993 Oh S, Abdelnabi J, Al-Dulaimi R, Aggarwal A, Ramos M, Davis S, Riester M, Waldron L 994 (2020) HGNChelper: identification and correction of invalid gene symbols for human and 995 mouse. F1000Res 9: 1493 996 Ohnuki M, Tanabe K, Sutou K, Teramoto I, Sawamura Y, Narita M, Nakamura M, Tokunaga 997 Y, Nakamura M, Watanabe A et al (2014) Dynamic regulation of human endogenous 998 retroviruses mediates factor-induced reprogramming and differentiation potential. 999 Proceedings of the National Academy of Sciences 111: 12426-12431 1000 Ramos A, Camargo FD (2012) The Hippo signaling pathway and stem cell biology. Trends in 1001 Cell Biology 22: 339-346 1002 Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F (2013) Genome engineering 1003 using the CRISPR-Cas9 system. Nature Protocols 8: 2281-2308 1004 Reé D, Fóthi Á, Varga N, Kolacsek O, Orbán TI, Apáti Á (2022) Partial Disturbance of 1005 Microprocessor Function in Human Stem Cells Carrying a Heterozygous Mutation in the 1006 DGCR8 Gene. Genes 13: 1925 1007 Römer C, Singh M, Hurst LD, Izsvák Z (2017) How to tame an endogenous retrovirus: 1008 HERVH and the evolution of human pluripotency. Current Opinion in Virology 25: 49-58 1009 Sakashita A, Kitano T, Ishizu H, Guo Y, Masuda H, Ariura M, Murano K, Siomi H (2023) 1010 Transcription of MERVL retrotransposons is required for preimplantation embryo 1011 development. Nat Genet 55: 484-495 1012 Schofield CM, Hsu R, Barker AJ, Gertz CC, Blelloch R, Ullian EM (2011) Monoallelic 1013 deletion of the microRNA biogenesis gene Dgcr8 produces deficits in the development of 1014 excitatory synaptic transmission in the prefrontal cortex. Neural Development 6: 11 1015 Sexton CE, Tillett RL, Han MV (2022) The essential but enigmatic regulatory role of 1016 HERVH in pluripotency. Trends Genet 38: 12-21 1017 Stark KL, Xu B, Bagchi A, Lai W-S, Liu H, Hsu R, Wan X, Pavlidis P, Mills AA, 1018 Karayiorgou M et al (2008) Altered brain microRNA biogenesis contributes to phenotypic 1019 deficits in a 22q11-deletion mouse model. Nature Genetics 40: 751-760 1020 Sun Z, Williams DJ, Xu B, Gogos JA (2018) Altered function and maturation of primary 1021 cortical neurons from a 22q11.2 deletion mouse model of schizophrenia. Translational 1022 Psychiatry 8 1023 Teijeiro V, Yang D, Majumdar S, González F, Rickert RW, Xu C, Koche R, Verma N, Lai 1024 EC, Huangfu D (2018) DICER1 Is Essential for Self-Renewal of Human Embryonic Stem 1025 Cells. Stem Cell Reports 11: 616-625 1026 Theunissen TW, Powell BE, Wang H, Mitalipova M, Faddah DA, Reddy J, Fan ZP, Maetzel 1027 D, Ganz K, Shi L et al (2014) Systematic Identification of Culture Conditions for Induction 1028 and Maintenance of Naive Human Pluripotency. Cell Stem Cell 15: 524-526 1029 Tiedemann F (2020) gghalves: Compose Half-Half Plots Using Your Favourite Geoms 1030 https://cran.rproject.org/web/packages/gghalves/index.html 1031 Torres-Padilla ME (2020) On transposons and totipotency. Philos Trans R Soc Lond B Biol 1032 Sci 375: 20190339 1033 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 40 Treiber T, Treiber N, Meister G (2019) Regulation of microRNA biogenesis and its crosstalk 1034 with other cellular pathways. Nature Reviews Molecular Cell Biology 20: 5-20 1035 Triboulet R, Chang H-M, Lapierre RJ, Gregory RI (2009) Post-transcriptional control of 1036 DGCR8 expression by the Microprocessor. RNA 15: 1005-1011 1037 Tristan-Manzano M, Maldonado-Perez N, Justicia-Lirio P, Cortijo-Gutierrez M, Tristan-1038 Ramos P, Blanco-Benitez C, Pavlovic K, Aguilar-Gonzalez A, Munoz P, Molina-Estevez FJ 1039 et al (2023) Lentiviral vectors for inducible, transactivator-free advanced therapy medicinal 1040 products: Application to CAR-T cells. Mol Ther Nucleic Acids 32: 322-339 1041 Tristán-Ramos P, Rubio-Roldan A, Peris G, Sánchez L, Amador-Cubero S, Viollet S, 1042 Cristofari G, Heras SR (2020) The tumor suppressor microRNA let-7 inhibits human LINE-1 1043 retrotransposition. Nature Communications 11 1044 Vlachos IS, Zagganas K, Paraskevopoulou MD, Georgakilas G, Karagkouni D, Vergoulis T, 1045 Dalamagas T, Hatzigeorgiou AG (2015) DIANA-miRPath v3.0: deciphering microRNA 1046 function with experimental support. Nucleic Acids Res 43: W460-466 1047 Wang J, Xie G, Singh M, Ghanbarian AT, Raskó T, Szvetnik A, Cai H, Besser D, Prigione A, 1048 Fuchs NV et al (2014) Primate-specific endogenous retrovirus-driven transcription defines 1049 naive-like stem cells. Nature 516: 405-409 1050 Wang Y, Medvid R, Melton C, Jaenisch R, Blelloch R (2007) DGCR8 is essential for 1051 microRNA biogenesis and silencing of embryonic stem cell self-renewal. Nat Genet 39: 380-1052 385 1053 Witteveldt J, Ivens A, Macias S (2018) Inhibition of Microprocessor Function during the 1054 Activation of the Type I Interferon Response. Cell Rep 23: 3275-3285 1055 Xu Z, Robitaille AM, Berndt JD, Davidson KC, Fischer KA, Mathieu J, Potter JC, Ruohola-1056 Baker H, Moon RT (2016) Wnt/β-catenin signaling promotes self-renewal and inhibits the 1057 primed state transition in naïve human embryonic stem cells. Proceedings of the National 1058 Academy of Sciences 113: E6382-E6390 1059 Yang WR, Ardeljan D, Pacyna CN, Payer LM, Burns KH (2019) SQuIRE reveals locus-1060 specific regulation of interspersed repeat expression. Nucleic Acids Res 47: e27 1061 Yilmaz A, Peretz M, Aharony A, Sagi I, Benvenisty N (2018) Defining essential genes for 1062 human pluripotent stem cells by CRISPR-Cas9 screening in haploid cells. Nat Cell Biol 20: 1063 610-619 1064 Yu G, Wang L-G, Han Y, He Q-Y (2012) clusterProfiler: an R Package for Comparing 1065 Biological Themes Among Gene Clusters. OMICS: A Journal of Integrative Biology 16: 284-1066 287 1067 Yu G, Wang LG, Yan GR, He QY (2015) DOSE: an R/Bioconductor package for disease 1068 ontology semantic and enrichment analysis. Bioinformatics 31: 608-609 1069 Zeuthen J, Norgaard JO, Avner P, Fellous M, Wartiovaara J, Vaheri A, Rosen A, Giovanella 1070 BC (1980) Characterization of a human ovarian teratocarcinoma-derived cell line. Int J 1071 Cancer 25: 19-32 1072 Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, Nusbaum C, Myers 1073 RM, Brown M, Li W et al (2008) Model-based analysis of ChIP-Seq (MACS). Genome Biol 1074 9: R137 1075 Zhu A, Ibrahim JG, Love MI (2019) Heavy-tailed prior distributions for sequence count data: 1076 removing the noise and preserving large differences. Bioinformatics 35: 2084-2092 1077 Zhu LJ, Gazin C, Lawson ND, Pages H, Lin SM, Lapointe DS, Green MR (2010) 1078 ChIPpeakAnno: a Bioconductor package to annotate ChIP-seq and ChIP-chip data. BMC 1079 Bioinformatics 11: 237 1080 Zimmermann C, Romero Y, Warnefors M, Bilican A, Borel C, Smith LB, Kotaja N, 1081 Kaessmann H, Nef S (2014) Germ Cell-Specific Targeting of DICER or DGCR8 Reveals a 1082 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 41 Novel Role for Endo-siRNAs in the Progression of Mammalian Spermatogenesis and Male 1083 Fertility. PLoS ONE 9: e107023 1084 1085 1086 Figure legends 1087 Figure 1. Generation and characterization of DGCR8 heterozygote cell lines. 1088 (A) Genes commonly deleted in 22qDS were searched for loss-of-function (LoF) mutations in 1089 the human population using the gnomAD. The observed/expected (obs/exp) LoF for each gene 1090 was plotted. (*) obs/exp LoF ≤ 0.089, predicted as haploinsufficient. Only genes with at least 1091 10 predicted LoF mutations were considered haploinsufficient (in bold) (B) DGCR8 and 1092 DROSHA western blot analyses of WT and two different HET DGCR8 H9 hESC clones. 1093 ACTIN serves a loading control. Quantification of DGCR8 and DROSHA protein levels in 1094 HET H9 hESCs. Data represents the average of three independent experiments +/ - st. dev (C) 1095 Colony morphology in normal culturing conditions for WT, HET hESCs and HET hESCs with 1096 rescued DGCR8 expression by lentiviral transduction (Scale bar = 100 µ m) (D) 1097 Immunofluorescence images for pluripotency markers Tra-1-60 and NANOG (Scale bar = 20 1098 µm) (E) NANOG, OCT4, SOX2 and KLF4 western blot analyses of both WT and two different 1099 HET DGCR8 hESC lines. Actin serves as a loading control ( F) Relative clonal expansion 1100 capacity, expressed as the fraction of stained area of HET hESCs lines vs WT transduced with 1101 a control -lentiviral vector or with lentivirus expressing human DGCR8. Data represent the 1102 average +/- st. dev. of 4 biological replicates. (****) p -val ≤ 0.0001, by one -way ANOVA 1103 followed by Tukey’s multiple comparison test ( G) Percentage of cells in early (EA) and late 1104 apoptosis (LA). Data are the average of (n=3) +/ - st. dev. (*) p -val ≤ 0.05, (**) p -val ≤ 0.01, 1105 by two-way ANOVA followed by Tukey’s multiple comparison test. 1106 1107 Figure 2. DGCR8 heterozygote hESCs display differentiation defects. 1108 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 42 (A) Quantification of the area of embryoid bodies at day 21. Data are the average +/ - s.e.m; 1109 (****) p -val ≤ 0.0001 by generalised linear models with negative binomial distribution, 1110 adjusted by Bonferroni post hoc test (top) . Representative images of EBs after differentiation 1111 at day 21 (bottom) (Scale bar = 100 µm) (B) RT -qPCR analyses of pluripotency, ( C) 1112 ectodermal, (D) mesodermal and (E) endodermal markers after EB differentiation of DGCR8 1113 HET H9 hESCs compared to the parental cell line (WT). Data are the average (n=3) +/- s.e.m. 1114 (*) p-val ≤ 0.5, (**) p-val ≤ 0.01, (***) p-val ≤ 0.001 (****), p-val ≤ 0.0001, by two-tailed 1115 Student t-test (F-H) Representative immunofluorescence images from WT and HET cells after 1116 differentiation to ( F) ectoderm (PAX6), ( G) mesoderm (BRACHYURY) and ( H) endoderm 1117 (SOX17). Scale bar = 10 µm. Data represent the log10 integrated density for staining in WT and 1118 HET cells (n=150), (***) p-val ≤ 0.001, (****) p-val ≤ 0.0001 by generalised linear models 1119 with negative binomial distribution, adjusted by Bonferroni post hoc test. 1120 1121 Figure 3. Defects in DGCR8 heterozygous cells are conserved in a naïve pluripotent state. 1122 (A) Colony morphology for WT and HET hESCs in naïve culturing conditions (Scale bar = 1123 100 µm) (B) DGCR8 and DROSHA western blot analyses of WT and HET DGCR8 hESC 1124 lines. Actin serves as a loading control ( C) RT-qPCR analyses of naïve ( KLF17, DNMT3L, 1125 DPP3A, DPP5A) and primed ( DUSP6) markers in both DGCR8 HET and WT hESCs. Data 1126 are normalized to GAPDH relative to WT naïve or WT primed, respectively, and represent the 1127 average of 3 biological replicates +/- st. dev. (D) Relative clonal expansion capacity, expressed 1128 as the number of colonies of HET hESCs lines in comparison to WT. Data represents the 1129 average (n=3) +/- st. dev. (***) p -val ≤ 0.001, by one-way ANOVA followed by Dunnett’s 1130 multiple comparison test (E) Quantification of the area in naïve colonies at day 5. Data are the 1131 average +/ - (n=10) (****) p -val ≤ 0.0001, by one -way ANOVA followed by Dunnett’s 1132 multiple comparison test. Colonies are visualised by crystal violet staining (F) Percentage of 1133 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 43 cells in early (EA) and late apoptosis (LA). Data represents the average +/ - st. dev of 3 1134 biological replicates. (*) p -val ≤ 0.05, (**) p -val ≤ 0.01 by two -way ANOVA followed by 1135 Tukey’s multiple comparison test. 1136 1137 Figure 4. miRNA expression is affected in DGCR8 heterozygote hESCs. 1138 (A) Differential expression (log2FC) of miRNAs from two different clones (HET 1 and 2) of 1139 DGCR8 HET hESCs compared to the parental cell line (WT). 69 and 93 miRNAs are 1140 downregulated whereas 59 and 76 miRNAs are upregulated in HET(1) and HET(2), 1141 respectively. MiRNAs in pink, indicate primate-specific miRNAs (B) Commonly dysregulated 1142 miRNAs in both HET H9 hESCs are significantly enriched in primate -specific miRNAs (PS) 1143 (hypergeometric p-val 1.112e-5) (C) Log2FC differential expression of miRNAs derived from 1144 the primate-specific C19MC cluster and the miR-317-3 cluster in the two HET clones vs WT 1145 H9 hESCs (D) RT-qPCR of some mature C19MC and 371-3 miRNAs expression in H9 naïve 1146 hESCs and (E) in H9 hESCs and HET hESCs transduced with a control or a lentiviral vector 1147 expressing human DGCR8. Data represent the average (n=3) +/ - s.e.m. Expression levels for 1148 each miRNAs are normalised to the levels of the DGCR8-independent miRNA, hsa-miR-320-1149 3p, and expressed relative to WT levels, (*) p-val ≤ 0.05, (**) p-val ≤ 0.01, (***) p-val ≤ 0.001, 1150 (****) p -val ≤ 0.0001, by one -way ANOVA followed by Dunnett’s or Tukey’s multiple 1151 comparison test, respectively ( F) KEGG pathway analyses for the predicted targets (microT -1152 CDS) of the commonly dysregulated primate-specific miRNAs in both HET H9 clones (G) GO 1153 pathway enrichment for differentially expressed genes in DGCR8 HET hESCs clones. 1154 1155 Figure 5. DGCR8 heterozygosity in PA-1 cells results in impaired miRNA processing and 1156 subtle changes on chromatin accessibility. 1157 (A) In vitro processing assays of radiolabelled pri-miRNAs, pri-let-7f-2 and pri-miR-23b, with 1158 WT, HET and KO PA -1 derived extracts. Arrow is the resulting cleavage product, the pre -1159 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 44 miRNA (B) Read depth coverage across a DGCR8 -dependent miRNA (hsa-miR-374b) and a 1160 DGCR8-independent mirtron ( has-miR-1234) from chromatin -associated RNA high -1161 throughput sequencing in WT, HET and KO PA-1 cells. Grey boxes indicate mature miRNAs, 1162 black line represents surrounding genomic regions ( C) Violin plots for MPI (Microprocessor 1163 processing index) values in WT, HET and KO. Only pri-miRNAs with a minimum MPI > 0.3 1164 in WT samples (processed in WT cells) were included in these analyses (***) p-val ≤ 0.001 by 1165 one-way ANOVA followed by Tukey’s multiple comparison test ( D) Quantification of 1166 unprocessed pri-miRNA levels by RT-qPCR in WT and HET PA-1 cells. Data are normalised 1167 to GAPDH and represented relative to WT. Data are the average (n=3) +/ - s.e.m. (*) p-val ≤ 1168 0.5, (**) p-val ≤ 0.01, (***) p-val ≤ 0.001 by one-way ANOVA followed by Dunnett’s multiple 1169 comparison test (y-axis, logarithmic) (E) Quantification of mature miRNA levels by RT-qPCR 1170 in WT and HET PA -1 cells. Data are normalised to the DGCR8 -independent miRNA, hsa-1171 miR-320-3p and relative to WT levels. Data are the average (n=3) +/ - s.e.m. (*) p-val ≤ 0.5, 1172 (***) p-val ≤ 0.001, (****) p-val ≤ 0.001 by one-way ANOVA followed by Dunnett’s multiple 1173 comparison test. (F) Differential ATAC accessibility analysis using DEseq2 of HET vs WT 1174 PA-1 cells (selected up/down hits with abs. log2FC ≥1, p-adj ≤ 0.05, hits outside plot as open 1175 triangles) (G) Proportion and genomic distribution of common (grey), less accessible (down, 1176 blue), or more accessible ATAC peaks (red, up) in HET vs WT PA - 1 (UTR; untranslated 1177 region) (H) Top twenty more significant GO terms associated with genomic regions that are 1178 significantly less accessible in HET cells obtained by rGREAT package. 1179 1180 Figure 6. DGCR8-mediated control of HERVH expression. 1181 (A) Violin plots for locus -specific expression of retrotransposons in the two DGCR8 hESC 1182 clones compared to WT cells. Significantly downregulated (log2FC 100) and upregulated (log2FC > 1; p -val ≤ 0.05; baseMean > 100) elements are 1184 represented as blue and red dots, respectively ( B) Distribution of downregulated ERV1 1185 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 45 subfamilies in both HET hESC clones (C) Genome browser view of RNA-seq data from WT 1186 and HET DGCR8 hESCs clones. Two representative regions containing differentially 1187 expressed LTR7 -HERVH elements are shown, XR_942793.1 and LMNB1 (top) and two 1188 chimeric transcripts driven by LTR7 promoter activity are shown as representative examples, 1189 LINC-ROR and ESRG (bottom). Sense (+) and antisense (-) strands are represented. Genes are 1190 depicted in blue and LTR7-HERVH in green (D) RT-qPCR analyses for HERVH-int and LTR7 1191 in WT and HET naïve hESCs ( E) the same as ( D), but in WT and HET primed hESCs 1192 transduced with lentiviral control vector or with lentivirus expressing DGCR8. Data are 1193 normalised to GAPDH and relative to WT levels. Data are the average (n=3) +/- st. dev. (*) p-1194 val ≤ 0.05, (**) p-val ≤ 0.01, (***) p-val ≤ 0.001, (****) p-val ≤ 0.0001, by one-way ANOVA, 1195 followed by Tukey’s or Dunnett’s multiple comparison test, respectively ( F) Heatmap for 1196 differentially expressed (p-adj ≤ 0.05) chimeric transcripts and lncRNAs driven by LTR7 1197 promoter activity in HET hESCs. 1198 1199 Figure 7. C19MC and miR -371-373 miRNAs rescue molecular and cellular defects in 1200 HET hESCs. 1201 (A) RT-qPCR for HERVH-int in WT and HET hESCs transfected with mimic control or with 1202 two miRNA mimics belonging to 371 -3 cluster (372-3p and 373-3p) or five miRNA mimics 1203 from C19MC cluster (520g-3p, 520d-3p, 519c-3p and 515-5p). Data are normalised to GAPDH 1204 and relative to WT levels. Data are the average (n=3) +/- st. dev. (*) p-val ≤ 0.05, (***) p-val 1205 ≤ 0.001, (****) p -val ≤ 0.0001, by one -way ANOVA, followed by Sidak’s multiple 1206 comparison test (B) KLF4 western -blot analyses for WT and HET hESCs transfected with 1207 mimic control or with miRNA mimics belonging to 371 -3 cluster or miRNA mimics from 1208 C19MC cluster. Actin serves as a loading control (top). Quantification of KLF4 protein levels 1209 in WT and HETs cells transfected with miRNAs mimics. Data are the average (n=3) +/- st. dev 1210 and relative to WT levels (down) (C) Relative clonal expansion capacity, expressed as fraction 1211 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 46 of stained area of HET hESCs transfected with miRNA mimics relative to HET hESCs lines 1212 transfected with a control miRNA mimic. Data represent the average +/- st. dev. of 3 biological 1213 replicates. (***) p -val ≤ 0.001, (****) p -val ≤ 0.0001 by one -way ANOVA followed by 1214 Dunnett’s multiple comparison test (D) Relative clonal expansion capacity, expressed as 1215 fraction of stained area of WT and HET hESCs transfected with a control miRNA mimic or a 1216 pool of miRNA mimics from miR-371-3 cluster (372-3p and 373-3p) or with C19MC miRNA 1217 mimics (520d -3p, 520g -3p, 519c -3p and 515 -5p) and relative to WT transfected with the 1218 control mimic. Data represent the average +/ - st. dev. of 3 biological replicates. (*) p-val ≤ 1219 0.05, (****) p-val ≤ 0.0001 by one-way ANOVA, followed by Sidak’s, multiple comparison 1220 test. Colonies are visualised by crystal violet staining. 1221 1222 EVF1. Characterization of DGCR8 heterozygous human pluripotent cell models. 1223 (A) Sanger sequencing of H9 hESC DGCR8 HET clone 1 (left) and clone 2 (right). HET(1) 1224 contains a 35-nt insertion in one allele of DGCR8 that contains a stop codon. HET(2) contains 1225 a frameshift 8-nt deletion in one allele of DGCR8 (B) Sanger sequencing of DGCR8 KO PA-1226 1 cells, allele 1 contains a 60 nt insertion which includes a stop codon, and allele 2 contains a 1227 frameshift 7-nt insertion in exon 2 (C) Sanger sequencing of DGCR8 PA-1 HET clones (both 1228 HET1 and HET2 are genetically identical). HET cells were generated by correcting the 1229 frameshift 7-nt insertion in exon 2 of DGCR8 in KO cells (D) DGCR8 and DROSHA western 1230 blot analyses of WT, two different HET DGCR8 PA -1 and KO clones. α -Tubulin serves as a 1231 loading control ( E) Quantification of DGCR8 and DROSHA protein levels in PA -1 WT, 1232 HET and KO cells. Data represent the average of three independent experiments +/- st.dev. α-1233 Tubulin serves as a loading control (F) DGCR8, DROSHA and KLF4 western blot analyses of 1234 H9 WT and HET DGCR8 hESCs transduced with a control or a lentiviral vector expressing 1235 human DGCR8. Actin serve as loading control ( G) RT-qPCR analyses of KLF4 for WT and 1236 HET hESCs transduced with control or lentiviral vector expressing human DGCR8. Data are 1237 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 47 normalised to GAPDH and relative to WT levels. Data are the average (n=3) +/- st. dev. (***) 1238 p-val ≤ 0.001, (****) p -val ≤ 0.0001, by one -way ANOVA followed by Tukey’s multiple 1239 comparison test. 1240 1241 EVF 2. DGCR8 heterozygous human pluripotent cell models display proliferation and 1242 pluripotent defects. 1243 (A) Alkaline phosphatase staining of WT and HET H9 hESCs, with high (left) or low (right) 1244 magnification. Quantification of alkaline positive colonies, distinguishing between positive or 1245 mixed, regarding morphology and staining intensity ( B) Quantification of alkaline positive, 1246 mixed or negative (differentiated) colonies, obtained after single-cell dilution of WT and HET 1247 cells. Data are the average of three independent experiments +/ - st. dev., (***) p -val ≤ 0.001 1248 by two-way ANOVA, followed by Dunnett’s multiple comparison test ( C) Clonal expansion 1249 assay for WT and HET cells transduced with a control or a lentiviral vector expressing human 1250 DGCR8. Colonies are visualised by crystal violet staining ( D) Cell proliferation essay of WT 1251 and HET hESCs transduced with control and DGCR8-expressing lentiviruses (left) and PA-1 1252 WT, HET and KO cells (right) ( E) Cell cycle analyses by flow cytometry. Data represent the 1253 average of 3 biological replicates +/ - st. dev. (*) p -val ≤ 0.05, (**) p -val ≤ 0.01 by two -way 1254 ANOVA followed by Dunnett’s multiple comparison test ( F) Representative flow cytometry 1255 scatter plots for WT and HET hESCs transduced with a control lentiviral vector or DGCR8 -1256 expressing vector, stained with 7-AAD (y-axis) and PE Annexin V (x-axis). Cells positive for 1257 PE Annexin V are designated as ‘early apoptosis’, double labelled are considered ‘late 1258 apoptosis’. 1259 EVF3. Functional analysis of primate -specific dysregulated miRNAs in both human 1260 pluripotent cell models. 1261 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 48 (A) Fraction of expressed and significantly dysregulated primate -specific miRNAs (pink). 1262 Dysregulated miRNAs were significantly enriched in primate-specific miRNAs in all HET cell 1263 lines: HET1 H9 hESCs (p -val =6.835985e-05), HET2 H9 hESCs (p -val =2.488492e-04) and 1264 HET PA -1 (p -val =3.65392e -07) ( B) KEGG pathway analyses for the predicted targets 1265 (microT- CDS) of all the common significantly dysregulated miRNAs in HET DGCR8 hESCs 1266 (C) Volcano plot for miRNA expression in HET PA -1 cells vs WT. MiRNAs in pink are 1267 primate-specific miRNAs (D) KEGG pathway analyses for predicted targets (microT-CDS) of 1268 dysregulated miRNAs (TOP100) in HET PA -1 cells ( E) the same as ( D), but only using 1269 primate-specific miRNA predicted targets. Common pathways in ( E) and ( D) are in black, 1270 unique in grey. 1271 EVF 4. Analysis of gene expression defects in both human pluripotent cell models. 1272 (A) Volcano plots for differentially expressed genes in the two HET hESC clones vs WT 1273 hESCs. Blue are downregulated genes, and red, upregulated ( B) the same as in ( A) for HET 1274 PA-1 cells (C) GO pathway enrichment for differentially expressed genes in DGCR8 HET PA-1275 1 cells (only included TOP 30 categories) ( D-E) Kernel density estimation of log2FC 1276 distributions for predicted targets of all dysregulated miRNAs (blue), targets of primate -1277 specific dysregulated miRNAs (pink) and controls or non -target genes (black) for ( D) PA-1 1278 HET cells and (E) HET (1) and HET (2) hESCs. Distribution is significantly different for both 1279 targets of all dysregulated miRNAs and primate -specific miRNAs vs. non -targets (p -val < 1280 2.22e-16). 1281 EVF 5. In vitro pri-miRNA processing, chromatin -associated RNA enrichment and 1282 ATAC-seq controls 1283 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 49 (A) Top, Ponceau staining of cytoplasmic (cyto), nucleoplasmic (nucl) and chromatin (chr) 1284 fractions from WT, HET and KO PA -1 cells. Bottom, western blot analyses of the same 1285 fractions against TUBULIN, which serves as a cytoplasmic fraction marker, and histone H3, 1286 which is a chromatin marker ( B) RT-qPCR analysis of cytoplasmic, nuclear and chromatin 1287 fractions. GAPDH pre-mRNA serves as a positive control for chromatin fractions, while 1288 GAPDH mRNA is enriched in the cytoplasm. All three tested pri -miRNAs were enriched in 1289 the chromatin fractions from WT PA -1 cells. An equal proportional volume from the three 1290 fractions was used for cDNA preparation and qPCR analyses. Data for each primer pair are 1291 represented as a fold -change over the ‘cytoplasmic’ sample ( C) Read depth coverage across 1292 several DGCR8 -dependent miRNAs from chromatin -associated RNA high -throughput 1293 sequencing in WT, HET and KO PA-1 cells. Grey boxes indicate mature miRNAs, black line 1294 represents surrounding genomic regions (D) Read distribution around TSS (transcription start 1295 sites) of ATAC-seq peaks for WT (blue), and HET (orange) PA-1 cell lines. 1296 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Table 1. Oligonucleotides used in this study H9 cells Primer name Sequence (5’ to 3’) DGCR8 gRNA(A) Fw CACCGCACCACTGGACGTTTGCAG DGCR8 gRNA(A) Rv AAACCTG CAA ACG TCC AGT GGT GC DGCR8 gRNA(B) Fw CACCGAGGTAATGGACGTTGGCTC DGCR8 gRNA(B) Rv AAAC GAG CCA ACG TCC ATT ACC TC DGCR8 PCR Fw ACTCGCTTAGTCGCCAGTCA DGCR8 PCR Rv TTACTCCTGCAGCTCTCGGT DGCR8 Fw TGCAAAGATGAATCCGTTGA DGCR8 Rv AGTAACTTGCTCAAAGTCAAA NANOG Fw TGCAGTTCCAGCCAAATTCTC NANOG Rv CCTAGTGGTCTGCTGTATTACATTAAGG SOX2 Fw TCAGGAGTTGTCAAGGCAGAGAAG SOX2 Rv CTCAGTCCTAGTCTTAAAGAGGCAGC KLF4 Fw GCTGCCGAGGACCTTCTG KLF4 Rv GCGAACGTGGAGAAAGATGG OCT4 Fw AGTGAGAGGCAACCTGGAGA OCT4 Rv ACACTCGGACCACATCCTTC HERVH Fw ACGCTTTACAGCCCTAGACC HERVH Rv GTCGGGAGCAGATTGGGTA LTR7 Fw GGCCAGTCCTTGCCTTAACT LTR7 Rv GCTCAGTGGGGGTGCTTTT OTX2 Fw GACCCGGTACCCAGACATC OTX2 Rv GCTCTTCGATTCTTAAACCATACC SOX1 Fw CACAACTCGGAGATCAGCAA SOX1 Rv GGTACTTGTAATCCGGGTGC PAX6 Fw CCGGCAGAAGATTGTAGAGC PAX6 Rv CGTTGGACACGTTTTGATTG HAND1 Fw AACCTCAGCCCTATCTCC HAND1 Rv CTTTCATCTTCCTGCGTC TBXT Fw GATGATCGTGACCAAGAACG TBXT Rv CTTCCAGCGGTGGTTGTC CD34 Fw CCGCGCTTTGCTTGCTGAG CD34 Rv TCTGGGGTAGCAGTACCGT FOXA2 Fw CGCCCTACTCGTACATCTCG FOXA2 Rv AGCGTCAGCATCTTGTTGG GATA6 Fw AATACTTCCCCCACAACACAA .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint GATA6 Rv ACTCTCCCGCACCAGTCAT HNF3 Fw GTGGCTCCAGGATGTTAGGA HNF3 Rv GCCTGAGTTCATGTTGCTGA SOX7 Fw ACGCCGAGCTCAGCAAGAT SOX7 Rv TCCACGTACGGCCTCTTCTG SOX17 Fw CAGAATCCAGACCTGCACAACGC SOX17 Rv CTTCAGCCGCTTCACCTGCTTG KLF17 Fw GGGATGGTGCGATAGATTCA KLF17 Rv GCCTCACCCTCACCTAACAA DPPA3 Fw ATCGGAAGCTTTACTCCGTCGAG DPPA3 Rv CCCTTAGGCTCCTTGTTTGTTGG DPPA5 Fw ACATCGAGCAGGTGAGCAAGG DPPA5 Rv CATGGCTTCGGCAAGTTTGAG DNMT3L Fw GGACCCTTCGATCTTGTGTA DNMT3L Rv ACCAGATTGTCCACGAACAT DUSP6 Fw GCTGTGGCACCGACACAGT DUSP6 Rv ACTCGCCGCCCGTATTCT GAPDH Fw TGCACCACCAACTGCTTAGC GAPDH Rv GGCATGGACTGTGGTCATGAG Actin Fw AGAGCTATGAGCTGCCTGACG Actin Rv TGTGTTGGCATAGAGGTCTTTACG Human miR371a-5pAAA ACUCAAACUGUGGGGGCACUAAA Human miR373-3pAAA GAAGUGCUUCGAUUUUGGGGUGUAAA Human miR372-3pAAA AAAGUGCUGCGACAUUUGAGCGUAAA Human miR515-5pAAA UUCUCCAAAAGAAAGCACUUUCUGAAA Human miR519c-3pAAA AAAGUGCAUCUUUUUAGAGGAUAAA Human miR520c-3pAAA AAAGUGCUUCCUUUUAGAGGGUAAA Human miR520d-3pAAA AAAGUGCUUCUCUUUGGUGGGUAAA Human miR520g-3pAAA ACAAAGUGCUUCCCUUUAGAGUGUAAA Human miR520f-3pAAA AAGUGCUUCCUUUUAGAGGGUUAAA Viral Titration DU3_Fw GACGGTACAGGCCAGACAA Viral Titration PBS_Rv TGGTGCAAATGAGTTTTCCA PA1 cells Primer name Sequence (5’ to 3’) Repair template for one of the alleles in DGCR8-/- PA-1 cells TTAGAGAAGGATCCTTTGGAGAGAAGAGAAGCTCC GTAGAAGTTGAAGGGGTCCTCAGCAGGGAGTTCGG ACTGTCCATCACCACCAGAGCCAACGTCCATTACC TCTGCACCACTGGAC exon 2 human DGCR8 Fw AGGAGAAGCGGTGATGGAG exon 2 human DGCR8 Rv CATCCACTCTGTCTCTCTGAAC .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Human pri-miR-302a Fw CGTGGATGTACTTGCTTTGAAAC Human pri-miR-302a Rv GCTGCGGTCAATACAATAAAGT Human pri-miR-302b Fw CATGGAAGTGCTTTCTGTGACT Human pri-miR-302b Rv CCTTCAAATGAGGTTAGCGTGT Human pri-miR-21 Fw TCTCATGGCAACACCAGTCG Human pri-miR-21 Rv AAGTGCCACCAGACAGAAGG Human pri-miR-92a-2 Fw AGTATTGCACTTGTCCCGGC Human pri-miR-92a-2 Rv TGACTAAATATCAGAACTTACGGCT Human pri-miR-135b Fw TCGCTTCCCTATGAGATTCCT Human pri-miR-135b Rv TGGGACAGCAATCACATAGGA Human pri-miR-767 Fw TGCACCATGGTTGTCTGAG Human pri-miR-767 Rv GACAATGAAGGTTCCTGCTCA Human miR-302a-3p TAAGTGCTTCCATGTTTTGGTGA Human miR-302b-3p TAAGTGCTTCCATGTTTTAGTAG Human miR-21-5p TAGCTTATCAGACTGATGTTGA Human miR-92a-3p TATTGCACTTGTCCCGGCCTGT Human miR-135b-5p TATGGCTTTTCATTCCTATGTGA Human miR-767-5p TGCACCATGGTTGTCTGAGCATG .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Figure 1 B D F 0 0.5 1 1.5 2 2.5 obs/exp LoF DGCR6PRODHDGCR2ESS2 SLC25A1CLTCL1 HIRA MRPL40 UFD1CDC45SEPT5TBX1GNB1LTXNRD2ARVCFTANGO2DGCR8TRMT2ARANBP1ZDHHC8CCDC188 RTN4RDGCR6LUSP41ZNF74SCARF2KLHL22MED15PI4KA SERPIND1SNAP29AIFM3LZTR1THAP7SLC7A4LRRC74B * * * * * A C E HET(1) mergeDAPI TRA-1-60 NANOG Relative clonal expansion 0 0.5 1 1.5 **** **** **** **** WT + control HET(1) + control HET(1) + DGCR8 HET(2) + control HET(2) + DGCR8 +PE-A-7AAD (EA) +PE-A+7AAD (LA) 0 5 10 15 20 * ** ** ** *** % cells G DGCR8 DROSHA WT HET(1)HET(2) 0.5 1 1.5 WT HET(1)HET(2) 0.5 1 1.5 0 DGCR8 β-ACT DROSHA β-ACT WT HET(1)HET(2) 0 WTHET(2) WT HET(1)+DGCR8 HET(2)+DGCR8 HET(1) HET(2) WT HET(1)HET(2) NANOG β-ACT OCT4 β-ACT SOX2 β-ACT KLF4 β-ACT 1 2 3 1 2 3 H9 WT + control H9 HET(1) + control H9 HET(1) + DGCR8 H9 HET(2) + control H9 HET(2) + DGCR8 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Figure 2 GATA6 HNF3 SOX7 SOX17 CD34 FOXA2 TBXT HAND1 WT HET(1) HET(2) norm GAPDH B E D WT HET(1) HET(2) norm GAPDH WT HET(1) HET(2) A area ( m ) EBs day 21 2 **** ****8x105 6x10 5 4x10 5 2x105 0 C F PAX6 MERGE G H DAPI DAPI BRACHYURY MERGE DAPI SOX17 MERGE PAX6 BRACHYURY SOX17 WTHET(1)HET(2) 0 2 4 6 8 0 2 4 6 8 log10(integrated density) 0 2 4 6 8 log10(integrated density) **** **** **** **** *** **** log10(integrated density) * POU5F1 0 7 14 21 2 0 7 14 21 NANOG ** 0 7 14 21 SOX2 ****** 0 7 14 21 ******** 0 7 14 21 ******** 0 7 14 21 *** * 0 7 14 21 ** * 0 7 14 21 norm GAPDH ** **** *** 0 7 14 21 * 0 7 14 21 ** 0 7 14 21 WT HET(1) HET(2) HET(1)HET(2) WTHET(1)HET(2) WTHET(1)HET(2) WT HET(1) HET(2) WT HET(1) HET(2) WT HET(1) HET(2) WT 1.5 1 0.5 0 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 0 0 0.5 0.5 1 1 1.5 1.5 0 0.5 1 1.5 2 0 0.5 1 1.5 0 0.5 1 1.5 0 0.5 1 1.5 0 0.5 1 1.5 0 0.5 1 1.5 OTX2 PAX6 SOX1 WT HET(1) HET(2) * * 0 7 14 21 0 7 14 21 0 7 14 21 norm GAPDH 0 0.5 1 1.5 2 0 0.5 1 1.5 0 0.5 1 1.5 2 µ .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint WT HET(1) HET(2) A B C 0 0.5 1 1.5 2 rel to WT naïve (norm GAPDH) 0.5 1 1.5 2 0 0 0.5 1 1.5 2 2.5 0.5 1 0 0.5 0 0 2 4 6 8 % of cells ** ** * F 0 5000 10000 15000 20000 25000Area ( m ) colonies 0 0.5 1 1.5 Relative number of colonies WT HET(1) HET(2) DGCR8 DROSHA 1 1 0.25 0.51 0.5 0.3 Figure 3 rel to WT primed (norm GAPDH) KLF17 DNMT3L DPPA3 DPPA5 DUSP6 WT naïve HET(1) naïve HET(2) naïve WT primed HET(1) primed HET(2) primed +PE-A-7AAD (EA) +PE-A-7AAD (LA) WT naïve HET(1) naïve HET(2) naïve D WT naïve HET(1) naïve HET(2) naïve WT naïve HET(1) naïve HET(2) naïve E β-ACT β-ACT 2 1.5 1.5 1 µ 2 .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Figure 4 A B 373-3p 512-5p 512-3p 1323 498-5p 520e-3p 515-5p 520f-3p 519c-5p519c-3p 1283 520a-5p 520a-3p 526b-5p 519b-5p 519b-3p 525-5p 523-5p 523-3p 518f-5p 518f-3p520b-5p 518b 526a-5p 520c-5p 520c-3p 518c-3p 524-5p 517a-3p 519d-3p 520d-3p 517b-3p 520g-5p 520g-3p 516b-5p 518e-5p 518a-3p 518d-5p 517c-3p 520h 522-5p 522-3p 519a-3p 516a-5p 519a-5p 371a-5p 371a-3p 372-5p 372-3p 373-5p -1 0 -2 -3 -4 -5 -6 log2FC vs. WTHET(1) HET(2) C19MC miR-371-3 371a-5p 372-3p 373-3p 0 0.5 1 1.5rel to WT (norm to hsa-miR-320a) **** **** ** *** **** **** **** **** **** **** *** ** miR: D H9 WT + control H9 HET(1) + control H9 HET(2) + control H9 HET(1) + DGCR8 H9 HET(2) + DGCR8 0 0.5 1 1.5 2 *** **** **** ** *** **** ** * *** ** **** **** *** **** * ******* *** ** **** * * * * E 0 0.5 1.0 1.5 rel t o W T (norm to hsa-miR-320a) *** * * ** * ** ** ** * * * * * HET(2) naïve 371a-5p 372-3p 373-3p 520f-3p 520c-3p 520d-3p 520g-3p 519c-3p 515-5p miR: C 0.07 0.08 0.09 0.10 embryonic organ development forebrain development regulation of protein kinase B signaling protein kinase B signaling inner ear development ear development regulation of neuron apoptotic process neuron apoptotic process GeneRatio 0.02 p-adj 0.03 0.04 0.05 Count 7 8 9 10 Dysregulated genes (H9-HET) G miR−141−3p miR−302a−5p miR−34a−5p miR−520f−3p miR−520g−3p miR−200c−3p miR−371a−5p miR−372−3p miR−373−3p 0 40 80 120 −4 40 −log10(padj) log2(FC) HET(2) vs WT miR−141−3p miR−302a−5p miR−34a−5p miR−520f−3p miR−520g−3p miR−200c−3p miR−371a−5p miR−372−3p miR−373−3p 0 40 80 120 −2.5 0 2.5 −log10(padj) log2(FC) HET(1) vs WT ALL PS number of miRNAs0 50 100 150 200 common unique HET(1) unique HET(2) F Predicted targets of dysregulated primate-specific miRNAs (H9-HET) −7 −6 −5 −4 −3 1518 2124 # miRNAs 25 50 75 100 # genes log10(p−val) Adipocytokine signaling pathway MAPK signaling pathway FoxO signaling pathway Chronic myeloid leukemia Adrenergic signaling in cardiomyocytes Pancreatic cancer Arrhythmogenic right ventricular cardiomyopathy (ARVC) AMPK signaling pathway Viral carcinogenesis Glioma Prostate cancer Hepatitis B Signaling pathways regulating pluripotency of stem cells L ysine degradation Prion diseases Thyroid hormone signaling pathway TGF−beta signaling pathway ErbB signaling pathway Hippo signaling pathway Proteoglycans in cancer rel to WT (norm to hsa-miR-320a) * ** ** ** ** ***** *** *** **** *** WT naïve HET(1) naïve HET(2) naïve 371-5p372-3p373-3p520f-3p520c-3p520d-3p520g-3p519c-3p515-5p 515-5p 519c-3p 520c-3p 520d-3p 520f-3p 520g-3p .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint A HET KO microprocessor processing index (MPI) B C 1.5 1 0.5 0 302a-3p 302b-3p **** **** 135b-5p 767-5p 21-5p 92a-3pmiR: norm miR-320-3p (rel to WT) D 0.5 5 pri-miR: 302a 302b 21 92a 135b *** ** 767 WT HET(1) HET(2) norm to GAPDH (rel to WT) * *** *** * *** * * * *** * WT *** *** pri-miRNAs miRNAs pri-let-7f-2 KO WT HET(1) HET(2) KO WT HET (1) 1 2 3 4 HET(2) 1 2 3 4 pri-miR-23b E Figure 5 F G H 0 10 20 30 % peaks common down up promoter (2-3Kb)promoter (1-2Kb)promoter (≤1Kb) 5’UTRexonintron3’UTR distal intergenic negative regulation of cardiocyte differentiation negative regulation of striated muscle cell differentiation negative regulation of cardiac muscle differentiation regulation of cell proliferation regulation of cartilage development positive regulation of cartilage development mesenchyme development regulation of cardiocyte differentiation regulation of cardiac muscle tissue development macromolecule metabolic process heart morphogenesis in utero embryonic development regulation of cardiac muscle cell differentiation positive regulation of cell proliferation embryo development ending in birth or egg hatching chordate embryonic development regulation of osteoblast proliferation muscle tissue morphogenesis mesenchymal cell differentiation circulatory system development binom_adj p_BH 1e-04 2e-04 3e-04 4e-04 5e-04 0 50 100 150 200 binom_observed_region-hits log2(fold-change) -6 -3 0 3 6 -log10(adj p-val) 10 5 0 n=317 n=1900.6% 0.36% DGCR8-dep DGCR8-indep WTHETKO 200 100 0 200 100 0 200 100 0 hsa-miR-374b hsa-miR-1234 60 40 20 0 60 40 20 0 60 40 20 0 0 2 4 6 WT HET(1) HET(2) ATAC-seq .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint 2 0 -2 -4 HET(1) -6 2 -4 0 -2 HET(2) log2FC log2FC ERVK ERVL-MaLR ERVLERV1 L1 L2SVAAluMIR HERVH-int:67,30% LTR7:21,90% other:8,25% LTR12C:1,59%LTR7Y:0,95% HERVH-int:54,92% LTR7:15,57% other: 27,05% LTR49-int: 2,49% HET(2) HET(1) Figure 6 A B HERVH-int LTR7LTR7 CACNA2D3 ESRG WT + WT - HET(1) + HET(1) - HET(2) + HET(2) - C HERVH 0 0.5 1 1.5 norm GAPDH (rel to WT) LTR7 0 0.5 1 1.5**** **** **** ** *** **** **** **** H9 WT + control H9 HET(1) + control H9 HET(1) + DGCR8 H9 HET(2) + control H9 HET(2) + DGCR8 0.5 1 1.5 2 0 HERVH 0.5 1 1.5 0 LTR7 WT naïve HET(1) naïve HET(2) naïve norm GAPDH (rel to WT) E HERVH-int LTR7LTR7 LINC-ROR WT + WT - HET(1) + HET(1) - HET(2) + HET(2) - D F WT HET(1) HET(2) PRRG1 PCSK9 RTP1 GRPR LIN7A KLKB1 ESRG LRRTM4 HHLA1 LINC00458 TRDN LINC−ROR MDGA2 RPL39L GUCY2C OLMALINC −3 −2 −1 0 1 2 3 Z-score ERVK ERVL-MaLR ERVLERV1 L1 L2SVAAluMIR XR_942793.1 HERVH-int LTR7LTR7 [0 - 1.5] [0 - 1.5] [0 - 1.5] [0 - 1.5] [0 - 1.5] [0 - 1.5] LMNB1 HERVH-int [0 - 4] [0 - 4] [0 - 4] [0 - 4] [0 - 4] [0 - 4] WT + WT - HET(1) + HET(1) - HET(2) + HET(2) - WT + WT - HET(1) + HET(1) - HET(2) + HET(2) - ** ** * * [0 - 0.2] [0 - 0.2] [0 - 0.2] [0 - 0.2] [0 - 0.2] [0 - 0.2] [0 - 80] [0 - 80] [0 - 80] [0 - 80] [0 - 80] [0 - 80] .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint Figure 7 A B HET(1) HET(2)WT KLF4 0 .5 1 1 .5 2 0 1 2 3 C HET(1) HET(2) 0 0 .5 1 1 .5 WT 372-3p 373-3p 519c-3p 515-5p 520g-3p 520d-3p ctrl HET HET(2) 372-3p 373-3p 519c-3p 515-5p 520g-3p 520d-3p D HERVH norm GAPDH (rel to WT) HET(1) HET(2) *** **** ********* WT ctrl miR372/3C19MC ctrl miR372/3C19MC 1.5 1 0.5 0 norm actin (rel to WT) ctrl ctrl miR372/3C19MCctrl miR372/3C19MC 1 2 3 4 5 6 7 KLF4 HET(1) HET(2) WT ctrl miR372/3C19MC ctrl miR372/3C19MC 0 1 2 3 Relative clonal expansion **** **** **** **** **** **** **** **** **** ******* *** (1) (2) (1) (2) (1) (2) (1) (2) (1) (2) (1) (2) (1) (2) ctrl miR372-3pmiR373-3pmiR520g-3pmiR520d-3pmiR519c-3pmiR515-5p HET(1) ctrl Relative clonal expansion * **** **** **** **** ctrl miR372/3C19MC ctrl miR372/3C19MC ctrl miR372/3C19MC HET(1) HET(2)WT ctrlmiR372/3C19MC β-ACT .CC-BY-NC-ND 4.0 International licensemade available 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 The copyright holder for this preprintthis version posted May 3, 2024. ; https://doi.org/10.1101/2024.05.02.592145doi: bioRxiv preprint