Efficient Reprogramming of the Epiblast Enables the Generation of Cloned Mice

Efficient Reprogramming of the Epiblast Enables the Generation of1 Cloned Mice2 Guomeng Li1,#, Wandong Bao2,#, Mei Hu1,3,#, Heng Xie1, Lianwei Li1, Jianlin Xu1, Shu Wei1,3 Yue Teng1, Yanyan Zhang4, Yunan Chen5, Lin Ran4, Jiawen Liu6, Juan Du6, Muhammad4 Ameen Jamal7, Zhanshan (Sam) Ma1, Chikai Zhou8,2, Jiangwei Lin1,2,*5 1. State Key Laboratory of Genetic Evolution & Animal Models, Kunming Institute of6 Zoology, Chinese Academy of Sciences, Kunming, Yunnan, 650223, China.7 2. Department of Laboratory Animal Science, Kunming Medical University, Kunming8 650500, China.9 3. NHC Key Laboratory of Study on Abnormal Gametes and Reproductive Tract, Anhui10 Medical University, Hefei, Anhui 230032, China.11 4. College of Animal Science and Technology, Yangzhou University, Yangzhou, 225009,12 Jiangsu, China.13 5. State Key Laboratory of Reproductive Medicine and Offspring Health, Nanjing Medical14 University, Nanjing 211166, China.15 6. College of Bioscience and Biotechnology, Yangzhou University, Yangzhou, 225009,16 Jiangsu, China.17 7. Institute for Engineering Medicine, Kunming Medical University, Kunming, 650500,18 China.19 20 8. Shenzhen Branch, Guangdong Laboratory for Lingnan Modern Agriculture, Key21 Laboratory of Synthetic Biology, Ministry of Agriculture and Rural Affairs, Agricultural22 Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen,23 518000, China.24 # These authors contributed equally.25 * Correspondence: [email protected] (J.L.)26 27 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Abstract28 Somatic cell nuclear transfer (SCNT) can generate viable mammals despite pervasive29 epigenetic abnormalities in cloned embryos, yet the mechanism underlying this paradox30 remains unclear. Here we show, using single-cell transcriptomics, that efficient31 reprogramming occurs exclusively in the epiblast (EPI), but the not primitive endoderm (PrE)32 of late mouse SCNT blastocysts. Integrating this with previous findings of trophectoderm (TE)33 aberration, we propose the EPI as the sole effectively reprogrammed lineage. By innovatively34 employing a Lego-like multi-lineage embryonic aggregation approach, where the35 extra-embryonic lineages were replaced with fertilization-derived counterparts, we36 demonstrated that SCNT-derived EPI inherently possesses full-term developmental potential37 nearly identical to that of fertilized EPI (20.5% vs. 21.8% birth rate), functionally confirming38 its effective reprogramming. Our study uncovers a lineage-specific asymmetric39 reprogramming mode where the EPI specifically achieves effective reprogramming, thereby40 constituting the deterministic basis for cloned animal generation. This work also provides a41 versatile strategy for investigating lineage potency and function.42 Keywords43 Somatic cell nuclear transfer, Generation of cloned animals/mice, Lineage-specific44 asymmetric reprogramming, Epiblast, Extra-embryonic lineages, Multi-lineage embryonic45 aggregation46 47 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Introduction48 Somatic cell nuclear transfer (SCNT) represents a transformative biotechnology with49 profound implications for animal reproduction, endangered species preservation, and50 regenerative medicine1-4. Despite its successful application across various species, the51 efficiency of generating cloned animals remains notoriously low, largely attributed to the52 incomplete epigenetic reprogramming of the donor somatic nucleus5,6. Extensive research has53 established that SCNT embryos encounter formidable barriers immediately following nuclear54 transfer, including the resistance of somatic H3K9me3 to erasure, aberrant DNA methylation55 patterns, and failures in zygotic genome activation (ZGA)7-10. Consequently, it is widely56 accepted that cloned embryos harbor severe, genome-wide epigenetic aberrations that persist57 throughout pre-implantation development11,12. However, this presents a fundamental paradox:58 despite these pervasive epigenetic defects that theoretically compromise developmental59 competence, healthy cloned animals can still be successfully born13-15. This raises a60 fundamental question: is the birth of cloned animals merely a stochastic “game of chance”61 where rare cells accidentally escape epigenetic errors, or does an underlying deterministic62 mechanism exist that lays the foundation for cloned animal generation?63 To decode this paradox, it is essential to examine the distinct cell lineages within the64 blastocyst, each of which possesses a specific developmental fate: the trophectoderm (TE),65 which forms the placenta; the primitive endoderm (PrE), which gives rise to the yolk sac; and66 the epiblast (EPI), which differentiates into the fetus proper16,17. Current research on67 reprogramming failure in SCNT blastocysts has predominantly focused on the68 extra-embryonic lineages, particularly the TE18. Indeed, our previous studies19, along with69 others, such as those highlighting the loss of H3K27me3-dependent non-canonical imprinting70 that leads to placental hyperplasia and early post-implantation lethality12, have demonstrated71 that the TE lineage in cloned embryos exhibits profound dysregulation. Given that the TE72 constitutes the majority of the blastocyst and is severely compromised, the key to73 understanding the “deterministic mechanism” of cloned animal generation likely lies within74 the inner cell mass (ICM). However, while studies have characterized global ICM defects20,75 current understanding of the precise molecular state, and thus the degree of reprogramming,76 of its two derived lineages, the PrE (extra-embryonic) and the EPI (embryonic), remains77 largely limited in SCNT blastocysts.78 Here, we performed precise isolation of the ICM from late-stage SCNT blastocysts followed79 by single-cell RNA sequencing (scRNA-seq) to dissect the lineage-specific reprogramming80 status. Surprisingly, our analysis revealed a striking divergence in reprogramming fates: the81 PrE lineage displayed severe transcriptional aberrations, whereas the EPI lineage maintained82 a global gene expression profile that closely mirrored that of fertilization-derived (FD)83 controls. To functionally interrogate whether this molecularly “normal” EPI is indeed the84 driver of cloned animal birth, we developed an innovative “Lego-like” embryonic85 reconstruction strategy. By systematically replacing the potentially compromised SCNT86 extra-embryonic lineages (both PrE and TE) with their FD counterparts, we demonstrated that87 the EPI lineage from late-stage SCNT blastocysts possesses a developmental potential nearly88 identical to that of its natural counterpart, achieving a comparable full-term birth rate (20.5%89 vs. 21.8%). Collectively, our findings support the hypothesis that effective reprogramming90 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint occurs specifically in the EPI lineage of SCNT embryos. This specific reprogramming91 outcome, distinct from the global failures in extra-embryonic tissues, serves as the92 deterministic cellular foundation enabling the generation of cloned animals.93 Results94 ScRNA-seq reveals highly efficient reprogramming in the EPI lineage of late-stage95 mouse SCNT blastocysts96 To investigate the reprogramming status of the ICM and its derived lineages, the PrE and EPI,97 in late-stage mouse SCNT blastocysts, we first profiled their gene expression landscapes.98 Using E4.5 blastocysts derived from the SCNT (NT) using cumulus cells and the99 fertilization-derived (FD) controls, we isolated ICMs by immunosurgery and performed bulk100 RNA sequencing (RNA-seq). Analysis of differentially expressed genes (DEGs) revealed that,101 overall, NT-ICMs exhibited distinct gene expression profiles compared to FD-ICMs,102 characterized by 515 up-regulated and 282 down-regulated DEGs (p.adjust < 0.05, Figure 1A).103 This finding is consistent with previous reports indicating global abnormalities in the ICM of104 cloned blastocysts20.105 For further dissecting the precise reprogramming states of the two constituent lineages, PrE106 and EPI, within late-stage cloned blastocysts, we dissociated E4.5 NT-ICMs (n = 3) and their107 FD counterparts (n = 2) into single cells (n = 84 and 53, respectively; Movie S1) and108 performed Smart-seq2 sequencing. Following quality control and clustering based on the109 expression of known lineage-specific marker genes, we obtained a total of 74 cells110 (comprising 63 PrE and 11 EPI cells) from NT blastocysts, and 49 cells (38 PrE and 11 EPI111 cells) from FD blastocysts (Figure S1A–D). We then performed DEG analyses in NT vs. FD112 (p.adjust < 0.05) separately for each lineage (Figure 1B, C). Strikingly, a massive number of113 DEGs were identified between NT-PrE and FD-PrE (n = 2,746, with 345 up-regulated and114 predominantly 2,401 down-regulated; Figure 1D), which showed significant associations with115 the biological processes (BPs) of kinase activity regulation, embryonic development, and116 RNA processing (Figure S2). In sharp contrast, only a minimal number of DEGs were117 detected between NT-EPI and FD-EPI (n = 42, with 4 up-regulated and also predominantly 38118 down-regulated; Figure 1D). These results indicate that in late-stage cloned blastocysts, the119 PrE lineage suffers from severe reprogramming abnormalities, whereas the EPI lineage120 achieves highly efficient reprogramming at the transcriptome level.121 Integrating these results with previous reports of severe TE dysregulation in cloned122 embryos18,19, we conclude that reprogramming failure in SCNT blastocysts is predominantly123 restricted to the extra-embryonic lineages (PrE and TE), while the EPI lineage, which124 develops the fetus proper, possesses the unique ability to achieve seemingly effective125 reprogramming. We thus hypothesize that this lineage-specific effective reprogramming126 endows the NT-EPI with the developmental competence to support full-term development,127 serving as the deterministic cellular foundation for the generation of cloned animals.128 Devising an induced lineage-based multi-lineage aggregation strategy to verify the129 hypothesis of SCNT-EPI developmental potential130 To test this hypothesis, we designed a strategy to substitute the extra-embryonic lineages of131 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint NT blastocysts with their FD counterparts, retaining only the EPI lineage derived from SCNT.132 This approach aims to evaluate the developmental potential of the NT-EPI lineage for133 generating cloned mice by excluding the confounding effects of reprogramming abnormalities134 in the NT-TE and NT-PrE lineages. Initially, we planned to isolate NT-EPI and FD-PrE cells135 from E4.5 blastocysts – a stage when PrE and EPI lineages are considered fully differentiated136 and no longer interconvertible21 – and aggregate them with FD tetraploid (4N) embryos,137 which provide a functional TE lineage via tetraploid complementation19,22. The resulting138 reconstructed embryos (NT-EPI/FD-PrE/FD-4N), along with their strict controls139 (FD-EPI/FD-PrE/FD-4N), were intended to be transferred into surrogate females so that the140 developmental potential of the NT-EPI lineage could be measured by the birth rate of cloned141 mice. However, in practice, the PrE and EPI cells within the late-stage ICM are closely142 connected and thus difficult to precisely distinguish and separate, as confirmed in our143 experimental operations on the E4.5 blastocysts.144 Nevertheless, it has been reported that treating early embryos with FGF4 and heparin (F4H)145 induces all E4.5 ICM cells to become Gata6-positive PrE cells23, while treatment with146 PD0325901 and CHIR99021 (two inhibitors of MEK and GSK3, respectively24; 2i) induces147 all E4.5 ICM cells to become Nanog-positive EPI cells25. Here, we refer to these cells as148 induced PrE (iPrE) and induced EPI (iEPI), respectively. Inspired by these findings, we149 sought to isolate iEPI or iPrE cells from E4.5 NT and FD blastocysts via induction followed150 by immunosurgery, and then aggregate them with FD-4N embryos. This alternative strategy151 allowed us to successfully obtain the desired “Lego-like” reconstructed embryos –152 specifically, NT-iEPI/FD-iPrE/FD-4N and their control FD-iEPI/FD-iPrE/FD-4N – thereby153 serving as a robust model to evaluate the capacity of the NT-EPI lineage to develop into154 cloned mice.155 Before formally implementing this innovative reconstruction strategy, prudence dictated that156 we should comprehensively characterize these induced lineage cells to verify their identity157 and, importantly, confirm their equivalence to their naturally differentiated counterparts in158 both FD and NT contexts. To this end, we utilized FD and NT embryos carrying a compound159 transgenic background of Nanog-GFP26 and Rosa26-CAG-tdTomato (R26-tdTomato), which160 allows for the specific visualization of the pluripotent EPI lineage (GFP-positive) and the161 tracking of all donor-derived cells (tdTomato-positive), respectively. These embryos were162 treated with 2i or F4H from the 2-cell stage until the E4.5 blastocyst stage. The resulting163 induced ICMs (iICMs) – specifically, iEPI (2i-treated) or iPrE (F4H-treated) – were then164 isolated via immunosurgery and respectively aggregated with two 4-cell FD-4N embryos to165 monitor their chimeric behavior in the late-stage reconstructed blastocysts (Figure 2A).166 First, we assessed the molecular identity of the induced lineages. Fluorescence microscopy of167 the isolated iICMs confirmed that Nanog-GFP expression was significantly repressed in both168 FD- and NT-iPrE, but maintained at high levels throughout the iEPIs (Figure 2B). Bulk169 RNA-seq analysis of these iICMs demonstrated that the iEPIs and iPrEs derived from both170 FD and NT embryos specifically enriched the marker gene expression of their corresponding171 lineages (Figure S3), validating the correctness of their induced lineage identities.172 Furthermore, based on our transcriptome data, no significant DEGs were identified between173 NT-iEPI and NT-EPI, or between FD-iEPI and FD-EPI (p.adjust < 0.05, Figure S4). This174 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint finding provides robust molecular evidence that the induced EPI lineages are equivalent to175 their naturally differentiated counterparts, thereby ensuring that using NT-iEPI vs. FD-iEPI as176 a surrogate for NT-EPI vs. FD-EPI as the sole experimental variable in our final aggregation177 strategy allows us to rigorously achieve our research objectives.178 Next, to assess their integration capability, we aggregated each type of iICMs individually179 with two FD-4N embryos and cultured them in vitro for 48 hours to obtain late reconstructed180 blastocysts for monitoring the chimeric distribution. As expected, both FD-iPrE and NT-iPrE181 cells (tdTomato-positive only) were predominantly localized to the cavity side of the ICM in182 the reconstructed blastocysts – where the differentiated PrE lineage typically resides16,27 – and183 showed no Nanog-GFP expression (Figure 2C), indicating that the iPrEs maintained their184 lineage identity well within the reconstructed embryos. The chimeric behavior of both FD-185 and NT-iEPIs appeared slightly more complex: while the majority of cells (tdTomato-positive,186 GFP-positive) were identified as EPI cells embedded within the ICM region, a portion of cells187 (tdTomato-positive, GFP-negative) were found distributed in positions typical of PrE and TE188 (Figure 2C). This suggests that the iEPIs (isolated from E4.5 late blastocysts) exhibit some189 plasticity towards the two extra-embryonic lineages in this context. We speculate that this190 unusual lineage transdifferentiation may be attributed to the unique microenvironment formed191 within the reconstructed embryos generated by this specific lineage/embryo aggregation192 method, which might differ significantly from that of naturally developing embryos.193 Nonetheless, our data demonstrate that both iPrEs and iEPIs can efficiently integrate into the194 reconstructed embryos generated by lineage/embryo aggregation strategies and constitute195 their respective lineage components.196 The SCNT-derived EPI lineage achieves a low cloning efficiency via tetraploid197 complementation198 To gain a preliminary understanding of the in vivo developmental potential of these induced199 lineages, particularly the NT-iEPI which is of our primary interest, we assessed their200 outcomes under tetraploid complementation following transfer into surrogate females. Given201 our observation that NT-iEPI cells can partially contribute to PrE and TE components in202 NT-iEPI/FD-4N reconstructed blastocysts, we initially hypothesized that these reconstructed203 embryos might possess a high potential for full-term development. To test this, NT embryos204 with a CAG-GFP transgenic background were processed to yield iEPIs (as described above),205 which were subsequently aggregated with FD-4N embryos. After 24 hours of in vitro culture,206 the aggregates formed blastocysts (Figure 3A, B) and were then transferred into surrogate207 females to evaluate their full-term developmental capacity.208 Transgenic fluorescence imaging at E16.5 revealed that the fetuses generated from209 NT-iEPI/FD-4N were ubiquitously GFP-positive (Figure 3C), confirming their origin entirely210 from the NT-iEPI lineage. Meanwhile, GFP signal was also observed in the extra-embryonic211 tissues, such as the umbilical cord, visceral yolk sac, and the labyrinth zone of the placenta212 (Figure 3C), consistent with the reported involvement of EPI lineage in the development of213 these structures28,29. However, no detectable GFP signal was found in any other area of the214 placenta (Figure 3C), standing in contrast to the phenomenon observed in NT-iEPI/FD-4N215 blastocysts, where the NT-iEPI cells showed a tendency to transdifferentiate into the216 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint extra-embryonic lineages including the TE (Figure 2C) that is responsible for forming the217 placenta30. Ultimately, we transferred a total of 101 NT-iEPI/FD-4N reconstructed blastocysts218 and merely obtained 6 live cloned pups, representing a birth rate of 5.9%. This efficiency was219 only marginally higher than that of the baseline NT control group (4.8%, 6/126; Figure 3D,220 Table 1), in which NT embryos were cultured to the blastocyst stage in vitro and then221 transferred. The body weight of the cloned pups generated from NT-iEPI/FD-4N showed no222 significant difference compared to the NT group (Figure 3E), whereas their placental weight223 was significantly lower (Figure 3F), which is likely attributed to the rescue of224 SCNT-associated placental hyperplasia by the FD-TE lineage provided via tetraploid225 complementation19,20.226 Here, we speculate that the de novo NT-TE cells, as well as NT-PrE cells, derived from the227 NT-iEPI might still harbor severe reprogramming abnormalities due to the lack of maternal228 epigenetic inheritance (e.g., H3K27me310,11,31), and consequently, act as functionally229 defective cells that are gradually outcompeted and replaced by the non-fluorescent,230 FD-4N-derived extra-embryonic lineages during subsequent development. If this is indeed the231 case, such futile lineage transdifferentiation by the NT-iEPI likely represents a detrimental232 drain on the cell pool destined for the fetus proper, which might even be exacerbated by a233 persistent lack of effective negative feedback due to the failure of transdifferentiation into234 functional extra-embryonic lineages, especially its critical neighbour, the PrE lineage. In235 support of this, immunofluorescence staining revealed that within the ICM regions of236 numerous NT-iEPI/FD-4N reconstructed blastocysts, the NT-iEPI(-derived) cells, despite237 exhibiting Nanog expression, almost universally displayed significant Gata6 expression238 (Figure S5), suggesting an aberrantly high degree of ongoing transdifferentiation toward the239 PrE lineage. This would therefore drastically diminish the NT-iEPI cell pool available for240 embryonic proper development, which helps to explain the severely compromised success241 rate of generating full-term cloned pups under this aggregation strategy.242 In parallel, utilizing the same methodology, we investigated the birth efficiency of FD-iEPI,243 NT-ICM, and FD-ICM aggregated with FD-4N embryos (Table 1). Possibly owing to its244 purely FD cellular composition and the consequent potential to generate a relatively normal245 PrE lineage, the FD-iEPI/FD-4N group achieved a remarkably high birth rate of 29.4%246 (15/51), which was only marginally lower than the 31.3% (5/16) observed in the247 FD-ICM/FD-4N group. In contrast, the NT-ICM/FD-4N group yielded a relatively lower birth248 rate of 16.0% (8/50), theoretically constrained by the overall SCNT origin of its ICM.249 However, this rate was still notably higher than the 5.9% observed in the NT-iEPI/FD-4N250 group, suggesting that although the NT-PrE lineage harbors severe reprogramming251 abnormalities, its presence (as in NT-ICM/FD-4N) is still more advantageous than its initial252 absence (as in NT-iEPI/FD-4N) for facilitating the successful birth of cloned pups to some253 extent.254 Additionally, we utilized R26-tdTomato FD embryos to derive iPrE cells and tested their255 developmental potential in vivo via tetraploid complementation (Figure S6A). As expected,256 following the transfer of blastocysts developed from FD-iPrE/FD-4N aggregates (Figure S6B),257 no live pups were obtained (0%, 0/64; Table 1), with only occasional resorption sites258 observed upon dissection (data not shown). This failure should be attributed to the lack of an259 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint embryonic (EPI) lineage competent to form the fetus proper in FD-iPrE/FD-4N embryos,260 where the FD-4N component can only effectively support extra-embryonic development19,20261 and the FD-iPrE was observed to strictly maintain its lineage identity (Figure 2C) as also262 further confirmed by immunofluorescence assays (Figure S6C). Given these results, the263 NT-iPrE appeared even less likely to support full-term development via tetraploid264 complementation, further due to its potential SCNT-associated reprogramming defects; thus,265 in vivo developmental testing was omitted for this group.266 The SCNT-derived EPI lineage possesses full-term developmental potential equivalent267 to that of the fertilization-derived EPI lineage268 Having comprehensively characterized the induced lineages in terms of their molecular269 signatures, chimeric integration capabilities, and in vivo developmental potentials under270 tetraploid complementation, we proceeded to the definitive phase of our study by271 implementing the “Lego-like” multi-lineage embryonic aggregation strategy, aiming to272 rigorously explore the developmental potential of the SCNT-derived EPI lineage for273 generating full-term cloned mice under conditions where the extra-embryonic lineages are274 fully functional and normal. Using CAG-GFP NT embryos and R26-tdTomato FD embryos,275 we induced and isolated iEPI and iPrE cells, respectively, as described above, and276 subsequently aggregated them simultaneously with two FD-4N embryos. After 24 hours of in277 vitro culture, the aggregates developed into NT-iEPI/FD-iPrE/FD-4N reconstructed278 blastocysts, which were then transferred into surrogate females to generate cloned pups279 (Figure 4A).280 Transgenic fluorescence imaging revealed that approximately 75% of the reconstructed281 blastocysts developed from the aggregates retained significant amounts of both NT-iEPI and282 FD-iPrE components (GFP-positive, tdTomato-positive; Figure 4B), and these were selected283 for subsequent transfer. Immunofluorescence staining showed that within the ICM regions of284 reconstructed blastocysts, the FD-iPrE(-derived) cells (tdTomato-positive) exhibited Gata6285 expression, whereas the NT-iEPI(-derived) cells (GFP-positive) were uniformly negative286 (Figure S7), indicating no transdifferentiation from the NT-iEPI to PrE lineage in this context.287 Subsequently, we examined the fluorescence distribution at E9.5 and found that the fetus288 proper was almost exclusively GFP-positive (derived from NT-iEPI), with only a few289 tdTomato-positive cells (derived from FD-iPrE) observed in the gut region, particularly the290 hindgut (Figure 4C). This distribution aligns perfectly with current knowledge regarding EPI291 and PrE lineage contributions during embryonic development, where the EPI forms the fetus292 proper32-34, while a minority of PrE cells transiently contribute to parts of the gut293 endoderm35,36. The VYS enveloping the E9.5 fetus exhibited double positivity for GFP and294 tdTomato (Figure 4D), consistent with the fact that the VYS is composed of an inner layer of295 extra-embryonic mesoderm derived from EPI and an outer layer of visceral endoderm derived296 from PrE37,38. Furthermore, this dual-fluorescence pattern was also observed in the VYS297 obtained from cesarean sections at E18.5 (Figure 4E). At this stage, prominent GFP signals298 were accurately detected in other structures that EPI developmentally involves, including the299 fetus proper, amnion, umbilical cord, and the labyrinth zone of the placenta28,29,39 (Figure 4E),300 with the majority of the placental tissue being non-fluorescent, presumably originating from301 the FD-4N embryos. These data robustly demonstrate that in our multi-lineage aggregation302 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint strategy, both NT-iEPI and FD-iPrE strictly adhere to their respective lineage identities and303 correctly fulfill their developmental fates.304 Ultimately, we transferred a total of 88 NT-iEPI/FD-iPrE/FD-4N reconstructed blastocysts305 and successfully obtained 18 live cloned pups (designated as Tri-NT mice due to their306 generation via tri-lineage/embryo aggregation), achieving a birth rate of 20.5% (Figure 4F,307 Table 1). This rate was significantly higher than the 4.8% (6/126) of the baseline NT control308 group, indicating that the holistic replacement of extra-embryonic lineages with their normal309 counterparts can drastically improve cloning efficiency. Crucially, compared to the strict310 control group FD-iEPI/FD-iPrE/FD-4N prepared using the same methodology, the birth rate311 of Tri-NT mice was nearly identical to that of the control pups (designated as Tri-FD mice),312 which stood at 21.8% (12/55; Figure 4F, Table 1). This provides direct evidence that NT-iEPI313 (equivalent to NT-EPI) possesses a full-term developmental potential comparable to that of314 FD-iEPI (equivalent to FD-EPI). Moreover, Tri-NT pups showed no significant differences in315 body weight (Figure 4G) or placental weight (Figure 4H) compared to Tri-FD pups.316 Attributable to tetraploid complementation, the placental weight of Tri-NT pups was317 significantly ameliorated compared to the NT group (Figure 4H), so was the morphology of318 the labyrinth and spongiotrophoblast layers within their placentae, as indicated by histologic319 sectioning (Figure S8). The Tri-NT mice could grow to adulthood (Figure 4I) and were fertile320 to produce healthy offspring (Figure 4J). Furthermore, to confirm the reproducibility of these321 findings, we performed an independent replicate of the experiments, which yielded statistical322 outcomes that recapitulated the trends presented above (Figure S9A–C).323 Compared to the NT-iEPI/FD-4N aggregation, the NT-iEPI/FD-iPrE/FD-4N strategy324 achieved a substantial increase in cloning efficiency (20.5% vs. 5.9%), attributed to the325 inclusion of FD-iPrE. The underlying mechanism likely involves two factors: first, the326 presence of FD-iPrE at the onset of aggregation allows the NT-iEPI to maintain a more327 abundant cell pool for fetal development by eliminating its need for compensatory328 transdifferentiation into the PrE lineage; second, as a representative of the normal PrE lineage,329 FD-iPrE theoretically supports NT-iEPI embryonic development more effectively than the330 potentially dysfunctional de novo NT-PrE cells generated in the NT-iEPI/FD-4N strategy.331 Therefore, only by incorporating FD-iPrE into the aggregation can the developmental332 potential of NT-iEPI to form the embryo proper and achieve full-term development be333 realized to the maximum extent, confirming the feasibility and necessity of implementing the334 NT-iEPI/FD-iPrE/FD-4N strategy to achieve the objectives of this study.335 In summary, by implementing this innovative multi-lineage embryonic aggregation strategy336 to holistically replace the reprogramming-compromised extra-embryonic lineages with their337 FD counterparts, we successfully unlocked the latent full-term developmental potential of the338 NT-(i)EPI lineage. Our data demonstrate that the NT-(i)EPI lineage inherently possesses339 developmental competence equivalent to that of the FD-(i)EPI lineage, thereby serving as the340 core driver for the generation of cloned mice. Finally, based on our findings and existing341 studies, we propose a theoretical model of “lineage-specific asymmetric reprogramming”342 governing the developmental potential of cloned embryos: while SCNT-induced343 reprogramming defects are pervasively retained in the extra-embryonic lineages (TE and PrE),344 severely compromising the cloning efficiency, the embryonic lineage (EPI) specifically345 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint achieves effective reprogramming and possesses intact pluripotency, thereby constituting the346 biological foundation that enables the generation of cloned animals (Figure 4K).347 Discussion348 Historically, the prevailing consensus regarding reprogramming in SCNT embryos posits that349 the failure to fully surmount early epigenetic barriers, such as the resistance of H3K9me3350 erasure7,40,41 that can lead to defective ZGA5,42, leaves cloned embryos burdened with351 widespread and severe reprogramming defects across the entire genome4,7. However, this352 perception of “global failure” has failed to coherently reconcile a core paradox: how can353 SCNT embryos, theoretically laden with such a profound and pervasive epigenetic burden,354 still possess the probability to break through these barriers and develop into completely355 healthy cloned animals? This cognitive gap largely stems from the fact that prior research has356 predominantly focused on the global embryonic state following nuclear transfer11,43, while357 overlooking the potential divergence in reprogramming fates among distinct cell lineages as358 development progresses to the blastocyst stage. In this study, by precisely dissecting the fine359 architecture within ICM of late-stage blastocysts and coupling scRNA-seq analysis with360 functional reconstruction strategies, we have uncovered and confirmed a mode of361 “lineage-specific asymmetric reprogramming”. It is crucial to emphasize that, for those362 cloned embryos capable of successfully developing to the late blastocyst stage,363 reprogramming failure is not persistently diffuse throughout the embryo. Instead, our findings364 reveal that defects are primarily restricted within the extra-embryonic lineages (TE and PrE);365 in sharp contrast, the EPI lineage, destined to form the fetus, specifically exhibits a366 remarkably effective reprogrammed state at this stage. This discovery not only theoretically367 helps to resolve the aforementioned “survivor paradox” but also offers a fundamental368 revelation: the birth of cloned animals is not the result of absolutely stochastic chance, but369 rather is founded upon a deterministic biological mechanism – the successful reconstruction370 of intact pluripotency specifically within the EPI lineage during the critical window of371 blastocyst formation.372 The lineage-specific effective reprogramming observed in SCNT blastocysts might be373 intrinsically linked to the developmental properties of the EPI lineage. Accumulating374 evidence indicates that maternal H3K27me3-dependent non-canonical imprinting375 predominantly resides and functions within the extra-embryonic lineages, as exemplified by376 its critical role in ensuring the activity of the maternal X chromosome by repressing Xist44 as377 well as in regulating the expression of key developmental genes12, whereas these imprints are378 generally transient or dispensable in the EPI lineage45. Owing to the inherent lack of these379 oocyte-specific H3K27me3 modifications in SCNT donor nuclei, the extra-embryonic380 lineages of cloned embryos are thus foreseeable to suffer from severe and potentially381 irreparable gene expression dysregulation. By contrast, the EPI lineage, due to its low382 dependency on these imprints, appears relatively insensitive to this specific epigenetic defect,383 thereby exhibiting a markedly superior reprogramming state. Nevertheless, the EPI lineage of384 cloned embryos might be more than merely a “passive survivor” as described above; based on385 current knowledge, we boldly speculate that it may also act as an “active seeker” capable of386 dynamically correcting reprogramming errors. In contrast with the extra-embryonic lineages,387 EPI should undergo a profound genome-wide remodeling associated with establishment of388 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint pluripotency, a process considered dependent on its robust pluripotency regulatory network389 (e.g., Oct4, Sox2, and Nanog) and characterized by hallmark events such as global DNA390 demethylation46,47 and X-chromosome reactivation (XCR) achieved through the potent391 repression of Xist48,49. A compelling question arises: could this EPI-specific, powerful392 epigenetic remodeling mechanism, while establishing the naïve pluripotency, also393 concurrently purge the residual epigenetic aberrations derived from the somatic nucleus? If so,394 this would constitute a “secondary reprogramming” event specifically in the NT-EPI lineage395 to successfully reconstruct intact developmental pluripotency. This represents a highly396 intriguing hypothesis that warrants future investigation.397 While our functional assays robustly demonstrate the full developmental potential of the398 NT-EPI lineage, a small number of DEGs departing from its FD compartment are still399 detected, consistent with the previous reports of EPI transcriptional defects in cloned400 blastocysts50. The NT-EPI has been further shown to be incapable of executing a successful401 naïve-to-primed transition due to the persistently aberrant activation of the Wnt signaling402 pathway, leading to the failure to correctly form the egg cylinder peri-implantation50. Rather403 than attributing these seemingly contradictory findings solely to incomplete reprogramming404 within the NT-EPI itself, we propose a reasonable and nonnegligible explanation: these405 aberrations may be extrinsically induced by its compromised extra-embryonic lineage406 neighbours. Given the well-established understanding of the precise and complex signaling407 crosstalk between embryonic and extra-embryonic lineages that crucially modulates EPI408 pluripotency and development51, it is highly probable that the extra-embryonic lineages with409 severe reprogramming disorder in cloned embryos fail to secrete signaling molecules or410 provide essential supportive cues normally, thereby inflicting potential “secondary damage”411 upon the molecular state and developmental trajectory of the EPI. For example, our412 scRNA-seq data reveal that the expression of Dkk1, encoding a key secreted inhibitor of the413 Wnt signaling pathway52, is significantly downregulated in NT-PrE cells (Figure S10),414 strongly suggesting a reduction of Wnt antagonist received by NT-EPI cells from surrounding415 microenvironment. This plausibly explains the previously reported persistent aberrant416 activation of Wnt signaling in NT-EPI, which results in defective peri-implantation417 development due to an inability to execute a successful naïve-to-primed transition50.418 Therefore, we posit that the detected molecular and developmental defects in NT-EPI should,419 at least partially, be attributed to suffering from the dysfunctional extra-embryonic420 components. This rationale underscores the necessity of holistically replacing the421 extra-embryonic lineages to establish an optimal “undisturbed” microenvironment for fully422 unleashing the true developmental potential of the NT-EPI, which ultimately proved to be423 equivalent to that of its FD counterpart.424 In the present study, we adopted the induced lineages from E4.5 blastocysts as a surrogate,425 avoiding the technical challenge of physically dissecting late-stage blastocyst lineages and the426 concomitant risk of lineage cross-contamination, which ensures the accuracy of our427 reconstruction experiments. It is generally accepted that the lineage fates of EPI and PrE in428 E4.5 late-stage blastocysts are stable and irreversible21,23,25. Consistent with this, we observed429 that the iPrEs tenaciously maintained their lineage identity within reconstructed embryos, and430 both iEPIs and iPrEs correctly contributed to their respective embryonic and extra-embryonic431 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint tissues during the in vivo development of the tri-lineage aggregates. Surprisingly, a fraction of432 iEPIs underwent transdifferentiation towards extra-embryonic lineages, when solely433 aggregated with FD-4N embryos. This unconventional, seemingly “totipotent-like” behavior434 is unlikely to be an intrinsic property of E4.5 EPI, but is instead plausibly induced by the435 unique intercellular microenvironment formed within the iEPI/FD-4N reconstructed embryos.436 In this system, developmentally advanced iEPI cells are forced to aggregate with blastomeres437 from a much earlier developmental stage – 4-cell stage FD-4N embryos, which are speculated438 capable of creating a relatively undifferentiated signaling microenvironment that potentially439 induces fate reprogramming in the adjacent iEPI cells. In contrast, this special situation is not440 present in a naturally developing late-stage blastocyst, where the EPI lineage identity is stable441 post-specification. Nonetheless, our data unexpectedly reveal that the E4.5 EPI lineage still442 retains a latent degree of cellular plasticity, which can be unveiled under certain extreme443 non-physiological embryonic conditions.444 Beyond the primary focus on validating the inherent full-term developmental potential of445 NT-EPI via tri-lineage aggregation, the present study, based on various lineage/embryo446 aggregation strategies, also demonstrates or reflects many aspects of lineage developmental447 characteristics and effects. Some comparisons of cloning birth rates among different448 aggregation groups highlight the significant impact of the PrE lineage on their full-term449 development (Table 1). For example, the NT-ICM/FD-4N group, despite possessing a450 presumably defective NT-PrE lineage, achieved a much higher birth rate than the451 NT-iEPI/FD-4N group (16.0% vs. 5.9%); conversely, when the latter was supplemented with452 a normal PrE lineage (NT-iEPI/FD-iPrE/FD-4N), the birth rate increased substantially (up to453 20.5%). Both observations indicate that the presence of a PrE lineage at the onset of454 aggregation has a significant promoting effect on the full-term development of reconstructed455 cloned embryos. However, an excess of PrE cells may inversely have a detrimental effect on456 cloning efficiency, as indicated by the fact that the NT-ICM/FD-iPrE/FD-4N group, despite457 additionally including even a normal PrE lineage, exhibited a lower birth rate than the458 NT-ICM/FD-4N group (13.5% vs. 16.0%). Furthermore, corresponding to the aberrant459 transcriptome profile of NT-PrE that implies its functional defects, the birth rates obtained460 from NT-ICM/FD-4N and NT-ICM/FD-iPrE/FD-4N (16.0% and 13.5%, respectively) were461 both lower than that of the NT-iEPI/FD-iPrE/FD-4N group (20.5%), which included only an462 FD-PrE and no NT-PrE lineage at the start of aggregation. This may indirectly reflect the463 potential detrimental effect of NT-PrE on the full-term development of reconstructed cloned464 embryos. Far beyond the experimental data already obtained in this study, the innovative465 “Lego-like” induced lineage/embryo aggregation strategy we employed, when combined with466 lineage-specific transgenic fluorescence labeling and gene editing (knockout or467 overexpression), is envisioned to have broad applications in customizable studies on the468 tracing, interaction, and molecular regulatory mechanisms of lineages in embryonic469 development.470 Despite the compelling functional evidence presented herein, our study still has several471 limitations. For instance, the omics assessment of the lineage reprogramming state is472 primarily based on the scRNA-seq data; while this can reliably indicate the reprogramming473 levels of lineages, the corresponding single-cell epigenomic profiles (e.g., whole-genome474 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint methylation or histone modifications) from the isolated NT-ICMs were unable to be obtained475 due to significant technical challenges. Second, although our “Lego-like” aggregation strategy476 allows for rigorous variable control, these artificially reconstructed embryos are speculated to477 possess an intercellular microenvironment different from that of natural development, which478 may exhibit a certain degree of negative impact on their development. This represents an479 intrinsic effect of the lineage/embryo aggregation strategies for both the experimental and480 control groups – an inevitable trade-off for achieving scientific validation. Finally, the deep481 molecular mechanisms driving the EPI-specific effective reprogramming remain to be482 elucidated in future studies.483 In conclusion, our study uncovers a “lineage-specific asymmetric reprogramming” mode in484 SCNT blastocysts, where only the EPI lineage achieves effective reprogramming, constituting485 the deterministic developmental foundation for the generation of cloned animals. This finding486 not only provides a new theoretical basis for understanding SCNT reprogramming487 mechanisms, but also further defines the direction for improving cloning efficiency in the488 future – rescuing or replacing the defective extra-embryonic lineages, ideally in a holistic489 manner. Furthermore, the advanced lineage/embryo aggregation strategy we developed is, in490 itself, a powerful new tool with broad prospects for studying embryonic lineage development491 and interaction in both normal and aberrant contexts. From a broader perspective, SCNT can492 be viewed as an extreme stress model for investigating embryonic robustness, which here493 allows us to marvel at the extraordinary resilience embryos exhibit in the face of such severe494 epigenetic aberrations – a resilience ultimately concentrated within the embryonic (EPI)495 lineage destined to form life itself, thereby conferring and safeguarding the very possibility of496 cloned animal birth.497 Methods498 Mouse strains499 The mouse strains used in this study include the wild-type C57BL/6J, DBA/2, ICR (CD-1)500 and FVB/N, and the C57BL/6J transgenic strains, Nanog-GFP, Rosa26-CAG-tdTomato501 (R26-tdTomato), and CAG-GFP. For the bulk RNA-seq and scRNA-seq of ICMs, FD502 blastocysts were obtained by mating C57BL/6J females with DBA/2 males, and NT503 blastocysts were generated using the cumulus cells and oocytes from B6D2F1 (produced by504 mating C57BL/6J females with DBA/2 males) females. For acquiring the (i)ICMs used for505 chimeric assessment in reconstructed blastocysts, FD 2-cell embryos were obtained by mating506 Nanog-GFP;R26-tdTomato homozygous females (produced by crossing Nanog-GFP and507 R26-tdTomato strains) with DBA/2 males, and NT 2-cell embryos were generated using the508 cumulus cells and oocytes from B6D2F1 Nanog-GFP;R26-tdTomato females (produced by509 mating Nanog-GFP;R26-tdTomato homozygous females with DBA/2 males). For acquiring510 the ICMs and iEPIs used for in vivo developmental potential assessment under511 lineage/embryo aggregation strategies, FD 2-cell embryos were obtained by mating512 CAG-GFP females with DBA/2 males, and NT 2-cell embryos were generated using the513 cumulus cells and oocytes from B6D2F1 CAG-GFP (produced by mating CAG-GFP females514 with DBA/2 males) females. For acquiring the FD-iPrEs used for in vivo developmental515 potential assessment under lineage/embryo aggregation strategies, FD 2-cell embryos were516 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint obtained by mating R26-tdTomato females with FVB/N males. For preparing tetraploid517 embryos, FD 2-cell embryos were obtained by intercrossing B6D2F1 (generating B6D2F2) or518 ICR. For embryo transfer, pseudo-pregnant ICR females were used as recipients.519 Fertilized embryo collection520 Females of C57BL/6J (3–6 weeks old), B6D2F1 (8–12 weeks old) or ICR (3–8 weeks old)521 background received an intraperitoneal injection of 5–7.5 IU PMSG, followed 48 hours later522 by an equivalent dose of hCG. The females were then mated with males of corresponding523 strains as required by the experiments. The presence of a vaginal plug the next morning was524 designated as E0.5. At E1.5, 2-cell embryos were collected in HEPES-buffered CZB (HCZB)525 medium, washed, and transferred to preheated KSOM medium supplemented with amino526 acids (KSOM-AA) overlaid with mineral oil. Embryos were cultured at 37°C in an527 atmosphere of 5% CO₂ in air.528 Somatic cell nuclear transfer529 B6D2F1 females (8–12 weeks old) used for SCNT experiments were superovulated by530 PMSG/hCG injection as described above, and MII oocytes were collected 14 hours post-hCG531 injection. The ampulla of the oviduct was punctured with a 30-gauge needle attached to a532 1-mL syringe to release the cumulus-oocyte complexes (COCs). The COCs were treated with533 0.1% hyaluronidase in HCZB medium for 3–5 min, and the denuded oocytes were then534 collected using a mouth-controlled pipette. Concurrently, the cumulus cells were collected in535 HCZB with 2.5% PVP and temporarily stored at 4℃, serving as nuclear donors. The residual536 hyaluronidase was removed from the oocytes by washing with HCZB, and the oocytes were537 then washed and cultured in CZB medium, overlaid with mineral oil at 37°C in 5% CO₂ in538 air.539 The oocytes were enucleated using an 8–10 μm glass needle and maintained in CZB at 37°C.540 Piezoelectric pulses were applied to penetrate the oocyte membrane, facilitating the injection541 of a donor cumulus cell into an enucleated oocyte. Subsequently, the reconstructed oocytes542 were returned to CZB medium for 1-hour recovery, followed by activation for 5.5–6 hours in543 Ca²⁺-free CZB medium containing 10 mM SrCl₂, 5 ng/mL trichostatin A (TSA), and 5544 µg/mL cytochalasin B (CB). All reconstructed embryos were then cultured in KSOM medium545 supplemented with 5 ng/mL TSA for another 3–4 hours, and finally maintained in KSOM546 medium with amino acids at 37°C under 5% CO₂ in air.547 Preparation of tetraploid embryos548 FD 2-cell embryos of B6D2F2 or ICR background were subjected to electrofusion to obtain549 tetraploid embryos. Upon completion of the program, the embryos were removed from the550 apparatus and placed in fresh KSOM culture droplets, which were then transferred to an551 incubator for 30–60 min recovery. Thereinto, the resulting fused 1-cell embryos were552 considered tetraploid embryos. After thorough washing with KSOM medium, the 1-cell553 tetraploid embryos were transferred to fresh KSOM droplets for further culture.554 Preparation of induced lineages555 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint For either the FD 2-cell embryos harvested from mated females or the NT 2-cell embryos556 developed in vitro after SCNT, 2i treatment was performed with KSOM medium557 supplemented with 1 µM PD0325901 and 3 µM CHIR99021 (termed 2i-KSOM) until the558 E4.5 late blastocyst stage to generate iEPI lineage (FD-iEPI and NT-iEPI, respectively), and559 F4H treatment was performed with KSOM medium supplemented with 1 µg/mL FGF4 and560 1 µg/mL heparin (termed F4H-KSOM) until the E4.5 late blastocyst stage to generate iPrE561 lineage (FD-iPrE and NT-iPrE, respectively). These blastocysts post-induction were then562 treated with acidic Tyrode’s solution to remove the zona pellucida, followed by 10 washes in563 KSOM. The blastocysts were then transferred to rabbit anti-mouse serum (diluted 1:1 with564 KSOM) and incubated for 30 min at 37°C in 5% CO₂. After 10 washes in KSOM, the565 blastocysts were exposed to guinea pig serum (diluted 1:1 with KSOM) and incubated for 15566 min at 37°C in 5% CO₂ until TE cell lysis was observed. The blastocysts were then567 recovered in KSOM for 30 min. Finally, lysed TE cells were mechanically removed by gentle568 pipetting with a 40–60 μm glass needle, and the isolated iICMs were then transferred to569 KSOM for temporary recovery and culture.570 Lineage/embryo aggregation and reconstruction571 Aggregation plates were prepared using an aggregation needle to create depressions in the572 bottom of a 35-mm easy grip petri dish, as described previously19,22. For573 NT-iEPI/FD-iPrE/FD-4N aggregation strategy, two 4-cell FD-4N embryos after zona574 pellucida removement by acidic Tyrode’s solution were placed into the aggregation well,575 followed by transferring one NT-iEPI and one FD-iPrE into the depression. Upon the576 completion of aggregation operation, the aggregate was cultured in KSOM medium for 24577 hours to obtain reconstructed blastocyst. The reconstructed blastocysts with significant578 chimerism of iICM cells (NT-iEPI and FD-iPrE) were selected by fluorescence microscopy579 and subsequently transferred into pseudo-pregnant recipients. For other tri-lineage580 aggregation strategies, the experimental procedures remained the same except for using581 different (i)ICM combinations. For dual-lineage (i)ICM/FD-4N aggregation strategies, one582 (i)ICM and two 4-cell FD-4N embryos were combined, and all other procedures were583 performed as described above.584 Embryo transfer585 ICR females (8–16 weeks old) in good estrus were paired with vasectomized males. The day586 following plug detection was designated as E0.5. At E2.5 of these pseudo-pregnant recipients,587 the NT embryos or reconstructed embryos cultured in vitro to the blastocyst stage were588 transferred into the uterus of pseudo-pregnant recipients. Generally, fetuses were delivered by589 cesarean section at E18.5, and data related to fetuses and placentae were recorded and590 analyzed. Additionally, nursing ICR mice were prepared in advance for fostering the newborn591 pups.592 Single-cell dissociation of inner cell mass593 FD or NT embryos were cultured in vitro to the E4.5 late blastocyst stage. The zona pellucida594 was removed by brief exposure to acidic Tyrode’s solution. Immunosurgery was then595 performed to lyse the outermost TE cells using anti-mouse serum and guinea pig complement.596 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint The lysed TE cells surrounding the ICM were mechanically aspirated and discarded using597 glass needles of various diameters (100–120 μm and 40–50 μm). Subsequently, the ICM was598 digested with trypsin-EDTA in calcium-free medium for 5–10 min at 37°C. Single-cell599 dissociation was then performed using a 40–60 µm glass needle and a micromanipulator600 (Movie S1). The dissociated single cells were washed 3 times in 0.1% BSA solution, and601 individually collected for subsequent scRNA-seq.602 RNA library preparation and sequencing603 The single cells isolated from the ICMs of E4.5 FD or NT blastocysts were individually604 transferred into tubes containing 200 µL lysis buffer by a mouth-controlled pipette. After lysis605 at 72°C for 3 min, reverse transcription mixture was added into each tube, followed by606 incubation at 25°C for 5 min, 42°C for 1 hour, and 50°C for 30 min, with a final inactivation607 step at 70°C for 10 min. The resulting cDNA samples were subsequently amplified, and608 purified using AMPure XP beads. The DNA abundance was enhanced through PCR using the609 Nextera XT DNA sample preparation kit. According to the manual of the KAPA HyperPrep610 Kit, single-cell RNA-seq libraries were created, and then sequenced on the Illumina Hiseq611 platform. For the preparation and sequencing of bulk RNA-seq libraries, the isolated E4.5612 (i)ICMs were individually transferred into tubes containing 200 µL lysis buffer, all other613 procedures were the same as described above.614 RNA-seq data processing and analysis615 For both the scRNA-seq and bulk RNA-seq data, raw reads were trimmed using Trim Galore616 (v0.6.10) with default parameters to remove low-quality bases and adapter sequences, and617 then aligned to the mouse reference genome (mm10) using Salmon (v0.14.1). Gene618 expression levels were quantified by the normalization of transcripts per million mapped619 fragments (TPM). DEGs between groups were identified using the Wilcoxon rank-sum test620 with FDR correction (p < 0.05). Prior to DEG analysis, scRNA-seq data were processed to621 assign EPI or PrE lineage identity to each cell based on the expression of established622 lineage-specific marker genes33,50,53 (Figure S1A, B), and the classification was further623 validated by examining additional reported lineage markers33 (Figure S1C, D).624 Immunofluorescence staining625 Reconstructed blastocysts were fixed in 4% PFA at room temperature for 30 min, followed by626 washing 3 times with PBS containing 0.5% BSA. Subsequently, the embryos were627 permeabilized using 0.1% Triton X-100 in PBS (PBTX) for 30 min, and washed 3 times by628 PBS. Transferred to 3% BSA in PBS, the embryos were blocked at room temperature for 1629 hour, and then incubated with the primary antibody diluted in the 3% BSA (1:200) overnight630 at 4°C. The embryos were washed 3 times by PBS, followed by incubation with the secondary631 antibody diluted in the 3% BSA (1:500) for 1–2 hours at room temperature. The nuclei were632 then stained with DAPI for 5 min, and the embryos were washed 3 times with PBS before633 being observed and photographed under the LSM880 confocal microscope. The primary634 antibodies used in the present study include rabbit anti-Nanog (Abcam) and goat anti-Gata6635 (R&D Systems). The corresponding secondary antibodies used in the present study were636 purchased from Thermo Fisher Scientific.637 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Frozen sectioning638 Placental tissues collected from E18.5 mouse conceptuses were washed briefly in cold PBS to639 remove excess blood, and fixed in 4% paraformaldehyde (PFA) at 4°C overnight. Fixed640 tissues were cryoprotected in graded sucrose solutions (15% and 30% in PBS) at 4°C,641 embedded in optimal cutting temperature (OCT) compound, and frozen at −80°C. Frozen642 blocks were sectioned at 7 µm thickness using a cryostat (Leica CM1950), mounted onto643 glass slides, air-dried for 30 min at room temperature, and stored at −80°C until use.644 H&E staining645 Frozen sections were removed from −80°C storage, equilibrated to room temperature for646 10–15 min, and briefly rinsed in PBS to remove residual OCT compound. Sections were647 refixed in 4% PFA for 10 min, followed by 4 washes with deionized water. Then, sections648 were stained with hematoxylin for 3 min, rinsed in tap water, differentiated in 1% acid649 alcohol, and counterstained with eosin for 30 sec. After dehydration through graded ethanol650 and clearing in xylene, sections were mounted with neutral balsam and observed under a light651 microscope. Images were captured using a digital camera attached to the microscope.652 Statistical analysis and figure preparation653 Statistical analyses and visualization were performed using GraphPad Prism 9.0.654 Visualization of DEG analyses was generated using the ggplot2 package in R. Schematic655 diagrams were created and image assembly was completed using Adobe Photoshop and656 Illustrator.657 Declarations658 Ethics statement659 Mouse care and all the experimental procedures were conducted in compliance with the660 guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Kunming661 Institute of Zoology, Chinese Academy of Sciences. The approval number for all the contents662 of this research is IACUC-RE-2024-01-006.663 Data availability664 All data supporting the findings of this study are available within the paper and its665 Supplementary Information. Additional analysis data are available from the corresponding666 author upon reasonable request.667 Competing interests668 The authors declare that they have no competing interests.669 Authors' contributions670 J.L. conceived this project. G.L, J.L., M.H., W.B., H.X., S.W., J.X., Y.T., Y.Z., Y.C., L.R.,671 J.L., and J.D. performed the experiments. L.L. performed sequencing data analysis. Z.M.672 provided guidance. J.L., W.B., G.L., M.J., and C.Z. drafted the manuscript. All authors read673 and approved the final manuscript.674 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Acknowledgements675 We sincerely thank Prof. Jinsong Li, Maria Elena Torres-Padilla, Ping Zheng, and Shuhui676 Bian for critical reading of the manuscript. This work was supported by the National Natural677 Science Foundation of China (31970823 and 32270862).678 References679 1 Gurdon, J. B. The developmental capacity of nuclei taken from intestinal epithelium cells680 of feeding tadpoles. J Embryol Exp Morphol 10, 622-640 (1962).681 2 Wakayama, T., Perry, A. C., Zuccotti, M., Johnson, K. R. & Yanagimachi, R. Full-term682 development of mice from enucleated oocytes injected with cumulus cell nuclei. Nature683 394, 369-374 (1998). https://doi.org:10.1038/28615684 3 Wilmut, I., Schnieke, A. E., McWhir, J., Kind, A. J. & Campbell, K. H. Viable offspring685 derived from fetal and adult mammalian cells. Nature 385, 810-813 (1997).686 https://doi.org:10.1038/385810a0687 4 Rideout, W. M., 3rd, Eggan, K. & Jaenisch, R. Nuclear cloning and epigenetic688 reprogramming of the genome. Science 293, 1093-1098 (2001).689 https://doi.org:10.1126/science.1063206690 5 Matoba, S. & Zhang, Y. Somatic Cell Nuclear Transfer Reprogramming: Mechanisms691 and Applications. Cell Stem Cell 23, 471-485 (2018).692 https://doi.org:10.1016/j.stem.2018.06.018693 6 Yang, X. et al. Nuclear reprogramming of cloned embryos and its implications for694 therapeutic cloning. Nat Genet 39, 295-302 (2007). https://doi.org:10.1038/ng1973695 7 Matoba, S. et al. Embryonic development following somatic cell nuclear transfer696 impeded by persisting histone methylation. Cell 159, 884-895 (2014).697 https://doi.org:10.1016/j.cell.2014.09.055698 8 Chung, Y. G. et al. Histone Demethylase Expression Enhances Human Somatic Cell699 Nuclear Transfer Efficiency and Promotes Derivation of Pluripotent Stem Cells. Cell700 Stem Cell 17, 758-766 (2015). https://doi.org:10.1016/j.stem.2015.10.001701 9 Gao, R. et al. Inhibition of Aberrant DNA Re-methylation Improves Post-implantation702 Development of Somatic Cell Nuclear Transfer Embryos. Cell Stem Cell 23,703 426-435.e425 (2018). https://doi.org:10.1016/j.stem.2018.07.017704 10 Zhang, X. et al. Setd2 overexpression rescues bivalent gene expression during705 SCNT-mediated ZGA. Protein Cell 16, 439-457 (2025).706 https://doi.org:10.1093/procel/pwaf010707 11 Matoba, S. et al. Loss of H3K27me3 Imprinting in Somatic Cell Nuclear Transfer708 Embryos Disrupts Post-Implantation Development. Cell Stem Cell 23, 343-354.e345709 (2018). https://doi.org:10.1016/j.stem.2018.06.008710 12 Inoue, K. et al. Loss of H3K27me3 imprinting in the Sfmbt2 miRNA cluster causes711 enlargement of cloned mouse placentas. Nat Commun 11, 2150 (2020).712 https://doi.org:10.1038/s41467-020-16044-8713 13 Li, Y. & Sun, Q. Epigenetic manipulation to improve mouse SCNT embryonic714 development. Front Genet 13, 932867 (2022).715 https://doi.org:10.3389/fgene.2022.932867716 14 Corr, S. A. et al. Radiographic assessment of the skeletons of Dolly and other clones717 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint finds no abnormal osteoarthritis. Sci Rep 7, 15685 (2017).718 https://doi.org:10.1038/s41598-017-15902-8719 15 Sinclair, K. D. et al. Healthy ageing of cloned sheep. Nat Commun 7, 12359 (2016).720 https://doi.org:10.1038/ncomms12359721 16 Chazaud, C. & Yamanaka, Y. Lineage specification in the mouse preimplantation722 embryo. Development 143, 1063-1074 (2016). https://doi.org:10.1242/dev.128314723 17 Allègre, N. et al. NANOG initiates epiblast fate through the coordination of pluripotency724 genes expression. Nat Commun 13, 3550 (2022).725 https://doi.org:10.1038/s41467-022-30858-8726 18 Jiang, L. et al. Deciphering the placental abnormalities associated with somatic cell727 nuclear transfer at single-nucleus resolution. Protein Cell (2023).728 https://doi.org:10.1093/procel/pwad030729 19 Lin, J. et al. Defects in trophoblast cell lineage account for the impaired in vivo730 development of cloned embryos generated by somatic nuclear transfer. Cell Stem Cell 8,731 371-375 (2011). https://doi.org:10.1016/j.stem.2011.02.007732 20 Li, Y. et al. Efficient Somatic Cell Nuclear Transfer by Overcoming Both Pre- and733 Post-Implantation Epigenetic Barriers. Adv Sci (Weinh), e04669 (2025).734 https://doi.org:10.1002/advs.202504669735 21 Grabarek, J. B. et al. Differential plasticity of epiblast and primitive endoderm736 precursors within the ICM of the early mouse embryo. Development 139, 129-139737 (2012). https://doi.org:10.1242/dev.067702738 22 Nagy, A. et al. Embryonic stem cells alone are able to support fetal development in the739 mouse. Development 110, 815-821 (1990). https://doi.org:10.1242/dev.110.3.815740 23 Yamanaka, Y., Lanner, F. & Rossant, J. FGF signal-dependent segregation of primitive741 endoderm and epiblast in the mouse blastocyst. Development 137, 715-724 (2010).742 https://doi.org:10.1242/dev.043471743 24 Ying, Q. L. et al. The ground state of embryonic stem cell self-renewal. Nature 453,744 519-523 (2008). https://doi.org:10.1038/nature06968745 25 Nichols, J., Silva, J., Roode, M. & Smith, A. Suppression of Erk signalling promotes746 ground state pluripotency in the mouse embryo. Development 136, 3215-3222 (2009).747 https://doi.org:10.1242/dev.038893748 26 Xenopoulos, P., Kang, M., Puliafito, A., Di Talia, S. & Hadjantonakis, A. K.749 Heterogeneities in Nanog Expression Drive Stable Commitment to Pluripotency in the750 Mouse Blastocyst. Cell Rep 10, 1508-1520 (2015).751 https://doi.org:10.1016/j.celrep.2015.02.010752 27 Frum, T. & Ralston, A. Cell signaling and transcription factors regulating cell fate during753 formation of the mouse blastocyst. Trends Genet 31, 402-410 (2015).754 https://doi.org:10.1016/j.tig.2015.04.002755 28 Watson, E. D. & Cross, J. C. Development of structures and transport functions in the756 mouse placenta. Physiology (Bethesda, Md.) 20, 180-193 (2005).757 https://doi.org:10.1152/physiol.00001.2005758 29 Nehme, E., Panda, A., Migeotte, I. & Pasque, V. Extra-embryonic mesoderm during759 development and in in vitro models. Development 152 (2025).760 https://doi.org:10.1242/dev.204624761 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 30 Hemberger, M., Hanna, C. W. & Dean, W. Mechanisms of early placental development762 in mouse and humans. Nat Rev Genet 21, 27-43 (2020).763 https://doi.org:10.1038/s41576-019-0169-4764 31 Wang, L. Y. et al. Overcoming Intrinsic H3K27me3 Imprinting Barriers Improves765 Post-implantation Development after Somatic Cell Nuclear Transfer. Cell Stem Cell 27,766 315-325.e315 (2020). https://doi.org:10.1016/j.stem.2020.05.014767 32 Rossant, J. & Tam, P. P. Blastocyst lineage formation, early embryonic asymmetries and768 axis patterning in the mouse. Development 136, 701-713 (2009).769 https://doi.org:10.1242/dev.017178770 33 Mohammed, H. et al. Single-Cell Landscape of Transcriptional Heterogeneity and Cell771 Fate Decisions during Mouse Early Gastrulation. Cell Rep 20, 1215-1228 (2017).772 https://doi.org:10.1016/j.celrep.2017.07.009773 34 Nowotschin, S. et al. The emergent landscape of the mouse gut endoderm at single-cell774 resolution. Nature 569, 361-367 (2019). https://doi.org:10.1038/s41586-019-1127-1775 35 Kwon, G. S., Viotti, M. & Hadjantonakis, A. K. The endoderm of the mouse embryo776 arises by dynamic widespread intercalation of embryonic and extraembryonic lineages.777 Dev Cell 15, 509-520 (2008). https://doi.org:10.1016/j.devcel.2008.07.017778 36 Batki, J. et al. Extraembryonic gut endoderm cells undergo programmed cell death779 during development. Nat Cell Biol 26, 868-877 (2024).780 https://doi.org:10.1038/s41556-024-01431-w781 37 Jollie, W. P. Development, morphology, and function of the yolk-sac placenta of782 laboratory rodents. Teratology 41, 361-381 (1990).783 https://doi.org:10.1002/tera.1420410403784 38 Martinelli, L. M., Carucci, A., Payano, V. J. H., Connor, K. L. & Bloise, E. Translational785 Comparison of the Human and Mouse Yolk Sac Development and Function.786 Reproductive sciences (Thousand Oaks, Calif.) 30, 41-53 (2023).787 https://doi.org:10.1007/s43032-022-00872-8788 39 Lawson, K. A., Meneses, J. J. & Pedersen, R. A. Clonal analysis of epiblast fate during789 germ layer formation in the mouse embryo. Development 113, 891-911 (1991).790 https://doi.org:10.1242/dev.113.3.891791 40 Liu, W. et al. Identification of key factors conquering developmental arrest of somatic792 cell cloned embryos by combining embryo biopsy and single-cell sequencing. Cell793 Discov 2, 16010 (2016). https://doi.org:10.1038/celldisc.2016.10794 41 Chen, M. et al. Chromatin architecture reorganization in murine somatic cell nuclear795 transfer embryos. Nat Commun 11, 1813 (2020).796 https://doi.org:10.1038/s41467-020-15607-z797 42 Xu, R. et al. Unreprogrammed H3K9me3 prevents minor zygotic genome activation and798 lineage commitment in SCNT embryos. Nat Commun 14, 4807 (2023).799 https://doi.org:10.1038/s41467-023-40496-3800 43 Liu, Y. et al. Transcriptional defects and reprogramming barriers in somatic cell nuclear801 reprogramming as revealed by single-embryo RNA sequencing. BMC Genomics 19, 734802 (2018). https://doi.org:10.1186/s12864-018-5091-1803 44 Inoue, A., Chen, Z., Yin, Q. & Zhang, Y. Maternal Eed knockout causes loss of804 H3K27me3 imprinting and random X inactivation in the extraembryonic cells. Genes805 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Dev 32, 1525-1536 (2018). https://doi.org:10.1101/gad.318675.118806 45 Inoue, A., Jiang, L., Lu, F., Suzuki, T. & Zhang, Y. Maternal H3K27me3 controls DNA807 methylation-independent imprinting. Nature 547, 419-424 (2017).808 https://doi.org:10.1038/nature23262809 46 Leitch, H. G. et al. Naive pluripotency is associated with global DNA hypomethylation.810 Nature structural & molecular biology 20, 311-316 (2013).811 https://doi.org:10.1038/nsmb.2510812 47 Ficz, G. et al. FGF signaling inhibition in ESCs drives rapid genome-wide demethylation813 to the epigenetic ground state of pluripotency. Cell Stem Cell 13, 351-359 (2013).814 https://doi.org:10.1016/j.stem.2013.06.004815 48 Navarro, P. et al. Molecular coupling of Xist regulation and pluripotency. Science 321,816 1693-1695 (2008). https://doi.org:10.1126/science.1160952817 49 Ohhata, T. & Wutz, A. Reactivation of the inactive X chromosome in development and818 reprogramming. Cell Mol Life Sci 70, 2443-2461 (2013).819 https://doi.org:10.1007/s00018-012-1174-3820 50 Li, Y. et al. Inhibition of Wnt activity improves peri-implantation development of821 somatic cell nuclear transfer embryos. Natl Sci Rev 10, nwad173 (2023).822 https://doi.org:10.1093/nsr/nwad173823 51 Wei, Y. et al. Dissecting embryonic and extra-embryonic lineage crosstalk with stem cell824 co-culture. bioRxiv (2023). https://doi.org:10.1101/2023.03.07.531525825 52 Li, J. et al. Roles of Dickkopf-1 and its receptor Kremen1 during embryonic826 implantation in mice. Fertil Steril 90, 1470-1479 (2008).827 https://doi.org:10.1016/j.fertnstert.2007.09.003828 53 Ohinata, Y. et al. Establishment of mouse stem cells that can recapitulate the829 developmental potential of primitive endoderm. Science 375, 574-578 (2022).830 https://doi.org:10.1126/science.aay3325831 832 833 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 834 Figure 1. ScRNA-seq reveals highly efficient reprogramming in the EPI lineage of835 late-stage mouse SCNT blastocysts. (A) DEG analysis in E4.5 NT-ICM vs. FD-ICM based836 on their bulk RNA-seq (p.adjust < 0.05). (B, C) DEG analyses in NT-PrE vs. FD-PrE (B) and837 NT-EPI vs. FD-EPI (C) based on the scRNA-seq of E4.5 NT-ICMs and FD-ICMs (p.adjust <838 0.05). (D) Statistics of DEG counts in E4.5 NT-PrE vs. FD-PrE and NT-EPI vs. FD-EPI.839 840 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 841 Figure 2. Chimeric behavior characteristics of induced lineages in reconstructed842 blastocysts under tetraploid complementation. (A) Schematic illustration of strategies for843 investigating chimeric behavior of induced lineages in reconstructed blastocysts under844 tetraploid complementation. The FD and NT embryos with Nanog-GFP;R26-tdTomato845 transgenic background were treated with 2i or F4H from the 2-cell stage until the E4.5846 blastocyst stage; subsequently, the resulting iEPIs (2i-treated) or iPrEs (F4H-treated) were847 isolated via removal of zona pellucida and immunosurgery, and then aggregated with two848 4-cell FD-4N embryos for detecting their chimeric behavior in the blastocysts reconstructed849 after 48-hour culture. The ICMs isolated from the naturally developed blastocysts were used850 as controls. ZP, zona pellucida; IS, immunosurgery. (B) Fluorescence microscopy of the851 (i)ICMs isolated from FD and NT blastocysts. Scale bar = 25 μm. (C) Fluorescence852 microscopy of the reconstructed blastocysts after 24-hour post-aggregation with FD-4N853 embryos for each (i)ICM type in (B). Scale bar = 65 μm.854 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 855 Figure 3. The SCNT-derived EPI lineage achieves a low cloning efficiency via tetraploid856 complementation. (A) Schematic illustration of strategy for evaluating the full-term857 development potential of NT-iEPI lineage based on tetraploid complementation. The NT858 embryo with CAG-GFP transgenic background was treated with 2i from the 2-cell stage until859 the E4.5 blastocyst stage; then, the resulting iEPI was isolated via removal of zona pellucida860 and immunosurgery, and aggregated with two 4-cell FD-4N embryos; after 24-hour in vitro861 culture, the reconstructed blastocyst developed from the aggregate was subsequently862 transferred to obtain cloned pup. ZP, zona pellucida; IS, immunosurgery; EF, electro-fusion.863 (B) Fluorescence microscopy of the NT-iEPI/FD-4N reconstructed blastocysts after 24-hour864 in vitro culture post-aggregation. Scale bar = 100 μm. (C) Fluorescence microscopy of the865 E16.5 fetus and extra-embryonic tissues generated by NT-iEPI/FD-4N strategy. f, fetus; u,866 umbilical cord; v, visceral yolk sac; l, labyrinth zone; p, placenta. (D–F) Statistics of the birth867 rate (D), body weight (E), and placenta weight (F) of the cloned pups generated by868 NT-iEPI/FD-4N strategy. The NT group was used as a baseline control. *, p < 0.05; ns, not869 significant; Student’s t-test.870 871 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 872 Figure 4. Tri-lineage aggregation strategy validates the inherent robust full-term873 developmental potential of SCNT-derived EPI lineage. (A) Schematic illustration of874 strategy for evaluating the full-term development potential of NT-iEPI lineage based on875 tri-lineage embryonic aggregation. Specifically, the CAG-GFP NT embryo and R26-tdTomato876 FD embryo were treated respectively with 2i and F4H from the 2-cell stage until the E4.5877 blastocyst stage; then, the resulting NT-iEPI and FD-iPrE were isolated via removal of zona878 pellucida and immunosurgery, and aggregated with two 4-cell FD-4N embryos; after 24-hour879 in vitro culture, the reconstructed blastocyst developed from the aggregate was subsequently880 transferred to obtain cloned pup. ZP, zona pellucida; IS, immunosurgery; EF, electro-fusion.881 (B–E) Fluorescence microscopy of the reconstructed blastocysts after 24-hour in vitro culture882 post-aggregation (B), E9.5 embryo (C) and visceral yolk sac (D), and E18.5 fetus and883 extra-embryonic tissues (E) generated by NT-iEPI/FD-iPrE/FD-4N strategy. The arrowheads884 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint in (B) indicate the representative reconstructed blastocysts with significant dual fluorescence885 (blue arrowheads) that were selected for subsequent transfer operation, and the ones with886 residual single fluorescence (tdTomato only, yellow arrowhead; GFP only, purple arrowhead)887 that were thereafter discarded. The white dashed lines indicate the outline of the E9.5 embryo888 (C) and E18.5 fetus (E). The white arrowheads indicate the significant tdTomato-positive889 positions in E9.5 embryo (C). f, fetus; u, umbilical cord; v, visceral yolk sac; a, amnion; l,890 labyrinth zone; p, placenta (E). Scale bar = 100 μm (B), 2.5 mm (C, D). (F–H) Statistics of891 the birth rate (F), body weight (G), and placenta weight (H) of the neonatal pups generated by892 the tri-lineage aggregation strategies. The NT group was used as a baseline control. ****, p <893 0.0001; ns, not significant; Student’s t-test. (I) Adult Tri-NT mice. (J) Tri-NT mice were894 fertile to produce healthy offspring. The Tri-NT female parents are indicated (*). (K)895 Theoretical model of “lineage-specific asymmetric reprogramming” in SCNT-derived late896 blastocyst. As pre-implantation reprogramming of NT embryos progresses to the late897 blastocyst stage, the differentiated cell lineages exhibit an asymmetric reprogramming mode:898 the embryonic lineage (EPI) achieves effective reprogramming to exercise fully functional899 pluripotency, constituting the biological foundation of cloned pup generation, whereas the900 extra-embryonic lineages (TE and PrE) display seriously defective reprogramming, leading to901 extra-embryonic dysfunction and consequent low cloning efficiency.902 903 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 904 905 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Supplementary Information906 907 Figure S1. Identification of EPI and PrE lineage cells in the scRNA-seq of E4.5908 FD-ICMs and NT-ICMs. (A, B) Expression heatmaps of lineage-specific marker genes used909 to identify the EPI and PrE lineage cells in the scRNA-seq of E4.5 FD-ICMs (A) and910 NT-ICMs (B). (C, D) Expression heatmaps of additional lineage-specific marker genes used911 to further confirm the EPI and PrE lineage cells identified in the scRNA-seq of E4.5912 FD-ICMs (C) and NT-ICMs (D).913 914 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 915 Figure S2. GO enrichment analyses on the DEGs in E4.5 NT-PrE vs. FD-PrE and916 NT-EPI vs. FD-EPI. The up-regulated and down-regulated DEGs were analyzed separately.917 Representative BP terms are shown for each analysis.918 919 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 920 Figure S3. Expression pattern of lineage-specific marker genes in the bulk RNA-seq of921 each (i)ICM type.922 923 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 924 Figure S4. DEG analyses in E4.5 NT-iEPI vs. NT-EPI and FD-iEPI vs. FD-EPI. The bulk925 RNA-seq data of E4.5 NT-/FD-iEPIs and the transcriptional profiles of EPI lineage cells in926 the scRNA-seq of E4.5 NT-/FD-ICMs were correspondingly utilized for the DEG analyses927 (p.adjust < 0.05).928 929 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 930 Figure S5. Immunofluorescence staining of NT-iEPI/FD-4N reconstructed blastocyst for931 Nanog and Gata6 expression. The NT-iEPI(-derived) cells were of CAG-GFP transgenic932 background. The reconstructed blastocysts for immunofluorescence staining (n = 13) were933 collected after 48-hour in vitro culture post-aggregation.934 935 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 936 Figure S6. Assessment of the in vivo developmental potential of FD-iPrE under937 tetraploid complementation. (A) Schematic illustration of strategy for assessing the in vivo938 developmental potential of FD-iPrE under tetraploid complementation. The R26-tdTomato939 FD embryo was treated with F4H from the 2-cell stage until the E4.5 blastocyst stage; then,940 the resulting FD-iPrE was isolated via removal of zona pellucida and immunosurgery, and941 aggregated with two 4-cell FD-4N embryos; after 24-hour in vitro culture, the reconstructed942 blastocyst developed from the aggregate was subsequently transferred to assess its in vivo943 developmental potential. ZP, zona pellucida; IS, immunosurgery; EF, electro-fusion. (B)944 Fluorescence microscopy of the FD-iPrE/FD-4N reconstructed blastocysts after 24-hour in945 vitro culture post-aggregation. Scale bar = 50 μm. (C) Immunofluorescence staining of946 FD-iPrE/FD-4N reconstructed blastocyst for Nanog and Gata6 expression. The947 FD-iPrE(-derived) cells were of R26-tdTomato transgenic background. The reconstructed948 blastocysts for immunofluorescence staining (n = 17) were collected after 48-hour in vitro949 culture post-aggregation.950 951 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 952 Figure S7. Immunofluorescence staining of NT-iEPI/FD-iPrE/FD-4N reconstructed953 blastocyst for Gata6 expression. The NT-iEPI(-derived) and FD-iPrE(-derived) cells were of954 CAG-GFP and R26-tdTomato transgenic background, respectively. The reconstructed955 blastocysts for immunofluorescence staining (n = 9) were collected after 48-hour in vitro956 culture post-aggregation.957 958 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 959 Figure S8. Histological sections with H&E staining show the phenotypic rescue of E18.5960 placenta generated by NT-iEPI/FD-iPrE/FD-4N strategy. The sections of FD and NT961 placentae serve as the normal and pathological controls, respectively. For each sample, the962 labyrinth layer is outlined by the blue line, with the spongiotrophoblast layer located above.963 964 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 965 Figure S9. Statistical analysis of the neonatal data obtained from repeated tri-lineage966 embryonic aggregation strategies. Statistical aspects include the birth rate (A), body weight967 (B), and placenta weight (C) of the neonatal pups. The FD and NT groups serve as the normal968 and pathological controls, respectively. Statistical significance was determined by Student’s969 t-test. Groups with different letters (a, b) indicate significant differences (p 0.05).971 972 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint 973 Figure S10. Dkk1 expression levels in the FD-PrE and NT-PrE lineages. The gene974 expression data was acquired from the transcriptional profiles of PrE lineage cells in the975 scRNA-seq of E4.5 FD-ICMs and NT-ICMs. Statistical significance was determined by976 Wilcoxon rank-sum test.977 978 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint Movie S1. Demonstration of dissociating ICM micro-masses into single cells.979 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted March 14, 2026. ; https://doi.org/10.64898/2026.03.12.711217doi: bioRxiv preprint