A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs

Guanghui Tan; Xiaoru Yan; Xuesha Cao; Ao Zhang; Tao Shi; Yaqi Zhou; Xi Guo; Haoping Wang; Hengkuan Li; Jiezhang Li; Yingran Liu; Xihong Wang; Yu Jiang; Yi Zheng

doi:10.1186/s13059-026-04173-0

. 2026 Jun 27;27:293. doi: 10.1186/s13059-026-04173-0

A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs

Guanghui Tan ^1,^2,^3,^#, Xiaoru Yan ^2,^4,^#, Xuesha Cao ^1,^2,^#, Ao Zhang ^1,^2,^#, Tao Shi ^1,², Yaqi Zhou ^1,², Xi Guo ^1,⁵, Haoping Wang ⁶, Hengkuan Li ^1,², Jiezhang Li ⁷, Yingran Liu ^1,², Xihong Wang ^1,², Yu Jiang ^1,^2,^✉, Yi Zheng ^1,^✉

PMCID: PMC13587516 PMID: 42365320

Abstract

Background

Elucidating cellular hierarchies and regulatory mechanisms of placental development across gestation is critical for understanding pregnancy maintenance and improving reproductive outcomes in mammals, including cattle. However, a comprehensive, temporally resolved single-cell characterization of the bovine placenta has remained lacking.

Results

We construct a longitudinal single-nucleus transcriptomic atlas comprising 311,299 placental cells and nuclei across 13 developmental timepoints (E12, E14, E16, E18, E24, E30, E50, E60, E85, E110, E180, E240, and E280). We identify 13 major cell types and 14 trophoblast subtypes, revealing pronounced cellular heterogeneity and stage-specific transcriptional programs. Regulatory analyses highlight HAND1 and DLX5 as candidate key regulators of maternal recognition of pregnancy. Trajectory inference demonstrates that binucleate cells arise from specific uninucleate cell subpopulations around E24, with differentiation governed by genomic imprinting and metabolic reprogramming. Integration with genome-wide association study data identifies eight early trophoblast subtypes significantly associated with gestation length, along with candidate pathways and risk genes, including CYCS, HMGA1, and VDAC1, under strong evolutionary constraint. Additionally, we find that placental macrophages emerge from E30 in cattle and show significant associations with pregnancy loss in both cattle and humans, sharing conserved risk pathways.

Conclusions

This study provides a comprehensive spatiotemporal single-cell atlas of bovine placental development, defining cellular hierarchies, lineage dynamics, and regulatory networks at the maternal–fetal interface. These findings offer a valuable resource and conceptual framework for understanding pregnancy maintenance and for improving reproductive traits in ruminants.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13059-026-04173-0.

Background

Reproductive efficiency is a key determinant of profitability and sustainability in both beef and dairy cow operations [1–3]. Pregnancy loss remains a chronic constraint on productivity during gestation, even in well-managed herds [1, 4], with reported losses varying by stages (e.g., 27% on 19–32 d, 13% on 30–45 d, 7% on 45–60 d, and 2% on 60–90 d in dairy cows) [5]. Pregnancy establishment and maintenance in cattle involve coordinated events from fertilization to the formation of a functional placenta [6, 7]. The placenta is essential for nutrient transport and fetal growth, secreting hormones that support fetal development and maternal adaptation, influencing long-term offspring health [1, 4, 8, 9]. In ruminants, the placenta exhibits a typical cotyledonary structure and contains distinctive binucleate cells (BNCs) that play pivotal roles in hormone secretion and immunomodulation [9, 10]. Nonetheless, due to cellular heterogeneity and dynamic remodeling, placental lineages, differentiation trajectories, and regulatory mechanisms across gestation remain poorly defined.

This gap limits our understanding of major reproductive disorders, such as gestation length (GL) and retained placenta (RP) that can substantially affect maternal and offspring health as well as productivity [10]. GL is highly heritable (≈30–50% in humans and cattle) [11–13], but its cellular foundations and molecular mechanisms remain largely elusive. RP is a leading cause of infertility and culling in dairy cows and is also implicated in severe postpartum complications in humans [14]. In humans, while postpartum hemorrhage remains the leading cause of maternal mortality worldwide, RP accounts for nearly 20% of severe cases and contributes substantially to the global maternal morbidity [14, 15]. Nonetheless, the etiology of RP is intricate, often associated with placental inflammation, immune dysregulation, and impaired tissue remodeling at the maternal–fetal interface [16]. Although genome-wide association studies (GWAS) have revealed several genetic loci linked to GL and RP [13, 14, 16], the absence of a high-resolution placental cell atlas precludes the precise mapping of these genetic signals to specific cell types and developmental stages, further hindering the elucidation of their functional mechanisms.

Recently developed single-cell transcriptomics enable the cellular-resolution dissection of organ development and diseases. Despite the extensive application to human and mouse placentas [17–21], a high-resolution transcriptomic atlas of the ruminant placenta throughout gestation has so far not been reported. Consequently, some fundamental questions remain to be answered, including the full cellular diversity of the ruminant placenta, the developmental origin of BNCs, the transcriptional programs underlying key events (e.g., maternal recognition of pregnancy), and the cell types and gestational windows through which genetic risks for GL and RP operate.

Cows are a valuable model organism, and studies in cows can reveal fundamental mechanisms for placental development and dysfunction with potential relevance to human reproductive health. Here, we generated a single-cell/-nucleus transcriptomic atlas of the bovine placenta spanning 13 gestational timepoints (311,299 cells/nuclei). Using this resource, we mapped the placental cellular architecture, delineated dynamic trophoblast differentiation trajectories, and inferred key gene regulatory networks. We further integrated cross-species comparisons, GWAS summary statistics, and gene regulatory network analyses to define conserved and species-specific features of placental development and to link genetic signals for GL and RP to specific placental cell types and gestational windows, highlighting candidate regulators of pregnancy maintenance. Together, this study provides a comprehensive resource for ruminant reproductive biology and a framework for understanding of pregnancy loss and improvement of reproductive efficiency in livestock.

Results

Construction of a single-cell transcriptomic atlas of bovine placentas throughout gestation

To decode the temporal transcriptional dynamics of the cotyledonary placenta, we first collected fresh placental tissue from Holstein cow embryos at multiple gestational timepoints (E50, E60, E85, E110, E180, E240, and E280). These timepoints represent critical developmental windows characterized by dramatic morphological restructuring of the placenta and coincide with major phases of rapid increases in placental chorionic surface area and weight. Based on the 10 × Genomics Chromium platform, we generated single-nucleus transcriptomic data from 224,039 nuclei. After stringent quality control, 176,731 high-quality nuclei were retained for downstream analyses (Additional file 2: Table S1-S2).

To systematically investigate the transcriptomic features of placentas at more developmental timepoints, we incorporated publicly available cow single-cell transcriptomic datasets that cover the peri-implantation (E12-E18; initiation of maternal signaling) [22] and implantation stages (E24-E50; placental morphogenesis) [23] (Fig. 1a). While these public datasets were derived from cow breeds different from Holstein used in this study, substantial evidence has indicated that early key events of placental development, particularly initiation of implantation and morphogenesis, are highly conserved among ruminants [24–26]. To empirically evaluate the cross-breed comparability, we first performed hematoxylin & eosin (HE) staining on placental samples from both Holstein cows and the breeds adopted in the public datasets. The results demonstrated high histological concordance across breeds at the matched developmental stages, particularly in key indices such as trophoblast column architecture, stromal vascular density, and primary villous formation (Additional file 1: Fig. S1a). Next, to assess the comparability at the transcriptomic level, we generated placental single-nucleus RNA-sequencing (snRNA-seq) data from Holstein and crossbred cattle at the same developmental stage and compared them with a published Angus placental single-cell RNA-sequencing (scRNA-seq) dataset [23] at the corresponding stage (Additional file 2: Table S1-S2). These analyses revealed consistent major cell type composition, strong cross-breed correspondence of matched cell populations, and similar marker expression patterns in trophoblast subpopulations (Additional file 1: Fig. S1b-h), thereby providing additional transcriptomic support for the integrated developmental analyses across breeds. Overall, we integrated single-cell/-nucleus transcriptomic datasets from 41 libraries, including 21 newly generated datasets in this study (with 2–4 biological replicates per timepoint). Using scVI [27], a variational inference framework benchmarked as the optimal integration method (Additional file 2: Table S3), we successfully harmonized 311,299 high-quality cells/nuclei across 13 gestational timepoints (E12, E14, E16, E18, E24, E30, E50, E60, E85, E110, E180, E240, and E280, Fig. 1a), and the subsequent unsupervised clustering with Seurat showed 28 distinct cell clusters (Fig. S2a-d). Based on canonical markers, we systematically annotated 13 major cell types spanning the entire bovine gestation: hypoblasts, epiblasts, trophoblast cells (Trophs), epithelial cells (Epi, including allantoic, luminal, and superficial glandular epithelial subtypes), vascular endothelial cells (Endo), vascular smooth muscle cells (VSMCs), fibroblasts (Fibs), mesenchymal stromal cells (Mes), monocytes, macrophages (Macs) and erythrocyte (Fig. 1b, c and Additional file 1: Fig. S2c-d).

Fig. 1 — Construction of a single-cell transcriptomic atlas of bovine placentas throughout gestation. a A schematic diagram of experimental design and the analysis framework. Gray and black fonts indicate public single-cell RNA-sequencing (scRNA-seq) datasets and the single-nucleus RNA-sequencing (snRNA-seq) datasets generated in this study, respectively. b UMAP projection of cells from the scRNA-/snRNA-seq data of bovine placental tissue and the cell type identities of cell clusters. c UMAP projection of cells from the scRNA-/snRNA-seq data of bovine placental tissue, colored by gestational timepoints. d The dot plot showing the expression of representative markers in the 13 major cell types. LEs: luminal epithelial cells; Endo: vascular endothelial cells; VSMCs: vascular smooth muscle cells; Fibs: fibroblasts; Mes: mesenchymal stromal cells; GECs: superficial glandular epithelial cells

Specifically, the hypoblasts were characterized by high expression of SOX17, FST, and GATA6 [22, 28], while the epiblasts were marked by prominent expression of POU5F1, TDGF1, and NANOG [22, 29]. The trophoblast cells were confirmed by elevated expression of GATA2, GATA3, PPARG, and AKR1B1 [30–32]. The subtype analysis revealed that epithelial cells expressed WFDC2 and EPCAM [3, 33], with allantoic epithelial cells (Allantoic Epi) additionally expressing UPK3B and UPK1B [34, 35]. Luminal epithelial cells (LEs) highly expressed SLC26A4 and SLC4A9 [36], and superficial glandular epithelial cells (GECs) showed high expression of GRHL2 and PAX8 [37–40]. Within the stromal and vascular compartments, vascular endothelial cells were marked by PECAM1, EGFL7, and KDR [41, 42], while VSMCs expressed ACTA2, MMP2, and COL6A3 [43]. Fibroblasts were defined by elevated expression of LUM and DCN, and mesenchymal stromal cells expressed COL1A1 and COL1A2 [3]. Among immune cell types, monocytes showed high transcription levels of S100A8, S100A9, and CD14 [44, 45], whereas resident placental macrophages (Hofbauer cells) were marked by CD163 and C1QA [3]. Erythroid cells were identified by specific expression of HBE1, HBA, and HBM (Fig. 1d and Additional file 1: Fig. S2e). Further, the results from the comparison with a published bovine placental cell atlas [3] supported the consistency of our cell type annotation, as corresponding major cell populations showed high cross-dataset similarity, conserved marker expression patterns, and significant overlap of cluster markers (Additional file 1: Fig. S2f-h). The analysis of placental cellular composition across gestation revealed trophoblast cells as the predominant cell population (constituting ~ 49% of all cells), with dynamic shifts observed in all lineages (Additional file 1: Fig. S3a-c). Precisely, hypoblast and epiblast populations underwent rapid diminution following E18, and became undetectable at later timepoints. VSMCs and endothelial cells emerged on E50, indicating completion of implantation and commencement of placental vasculogenesis. Macrophages were first detected on E30 and exhibited marked enrichment at late gestation (E280), suggesting delicate immune regulation during terminal placental maturation. Collectively, we presented a comprehensive single-cell/-nucleus transcriptomic atlas of bovine placentas spanning the entire gestational period. This high-resolution resource lays the robust foundation for advancing reproductive biology in ruminants and establishment of a critical translational model for studies on placental development.

Spatiotemporal mapping of bovine trophoblasts uncovered species-specific lineage programs and stage-dependent shifts in cellular composition

Trophoblast cells are exclusive to the placental tissue and comprise multiple subtypes. Yet, their diversity remains poorly understood in the bovine placenta. To elucidate this complexity, we performed a re-clustering analysis for the trophoblast cells and discovered a multitude of cell subtypes yet to be characterized. Overall, we found 14 distinct trophoblast subtypes across all gestational timepoints in cows, based on their expression of specific markers and the conserved transcriptomic patterns (Fig. 2a-c and Additional file 2: Table S4). Specifically, we identified two trophectoderm (TE) subpopulations in cows, i.e., mural and polar TE, which were distinguished by KRT18, CDX2, and CCND2 expression [46]. Notably, polar TE exhibited pronounced stem-like features. This subpopulation showed elevated expression of FURIN (Additional file 1: Fig. S4a), a key convertase that restricts NODAL activity to the epiblast-trophoblast boundary [47], and concurrently upregulated proliferation-associated genes, including BIRC5, PCNA, CCNB1, and AURKB (Additional file 1: Fig. S4a). Importantly, polar TE also displayed enhanced expression of established bovine trophoblast stem cell (bTSC) markers [48], such as KRT8, GATA2, SFN, ASCL2, and RAB25, while the overall expression of differentiated/functional trophoblast markers was relatively low (Additional file 1: Fig. S4a), suggesting that polar TE may be enriched for a bTSC-like population. Consistently, we identified a bTSC-like subcluster within polar TE (PL0) that was predominantly derived from E12-E14 (Additional file 1: Fig. S4b-f), providing a basis for in vivo characterization of bTSC-like cells. We also identified two trophoblast subtypes with high expression of interferon tau (IFNT, a key factor initiating implantation [29, 49]), designated as IFNT⁺ Troph1 and IFNT⁺ Troph2. IFNT⁺ Troph1 uniquely co-expressed AHSG and FN1 that are involved in inflammatory response and cell adhesion [50], respectively, with both processes essential for trophoblast migration, attachment, and invasion during embryo implantation. Notably, IFNT⁺ Troph1 represented the most abundant subtype in all trophoblast populations, indicating a critical role in maternal recognition of pregnancy. In contrast, IFNT⁺ Troph2 specifically expressed KRT7, RAC2, and TMSB4. Given the known roles of RAC2 in regulation of cell polarity, inflammatory signaling, and reactive oxygen species (ROS) production [49], this subtype might be involved in local immune modulation and maintenance of maternal–fetal interface homeostasis during implantation. Besides, we delineated four uninucleate cell (UNC) subtypes (designated as Troph_U1-U4), based on the distinct PAG2 and PAG8 expression patterns [3]. The expression levels of PAG2 and PAG8 were higher in Troph_U1 and Troph_U2, and Troph_U1 and Troph_U2 also highly co-expressed PAG12, PAG11, and PAGE4. CDKN1C was broadly expressed in the UNC subtypes, whereas CDK14 was preferentially expressed in Troph_U1/U3/U4. Troph_U1 specifically expressed CTSL and SPINT1, the known markers for areolar epithelial cells (a subtype of trophoblast cells) in the porcine diffuse placenta [51], whilst Troph_U2 was defined as an epithelial-mesenchymal trophoblast population (UNC-EMT), characterized by upregulation of mesenchymal/extracellular matrix (ECM) programs, including VIM, COL1A1, and COL3A1. For Troph_U3 and Troph_U4, they both exhibited elevated expression of TEAD1, a key transcription factor (TF) implicated in trophoblast lineage differentiation. Troph_U3 was further characterized by specific expression of DIAPH3 and marked upregulation of multiple mitosis-associated genes (i.e., CENPF, DEPDC1B, and TPX2), as well as prominent expression of proliferation/cell cycle markers including MKI67, CDK1, and TOP2A, suggesting that Troph_U3 represents a proliferative trophoblast precursor-like population (Additional file 1: Fig. S4a). Notably, an akin subtype has also been identified in the porcine diffuse placenta [51], indicating potential evolutionary conservation. In contrast, Troph_U4 displayed high expression of SULT1E1 (which regulates the estrogen activity [52]) and SLC1A3 (which transports amino acids [53]) and represented the second most abundant trophoblast subtype after IFNT⁺ Troph1, suggesting that this subpopulation might modulate estrogen metabolism in bovine placentas.

We further identified six subtypes of BNCs, designated as Troph_B1-B6 (Fig. 2a-c). All subtypes specifically expressed a battery of pregnancy-associated glycoproteins (PAGs) and prolactin-related proteins (PRPs). Notably, Troph_B1 uniquely expressed SYCN (required for efficient compound exocytosis [54]), POU3F1 (a known inhibitor of trophoblast giant cell (TGC) differentiation in mice [55]), and SERPINE2 (a marker of human placental bed giant cells (GCs), interstitial extravillous trophoblasts (iEVTs), and cynomolgus macaque (Macaca fascicularis, hereafter referred to as macaca) EVTs [17, 42]), and additionally expressed E2F8. Also, Troph_B1 showed elevated expression of BIRC5, PCNA, CCNB1, and AURKB, predominantly derived from E24-E50 (around the onset of BNC emergence), and exhibited high CDKN1B expression together with relatively incipient expression of BNC markers (e.g., PAG6 and PRP1) and low expression of UNC markers (Additional file 1: Fig. S4a), collectively suggesting an early/intermediate transitional state from a proliferative progenitor-like program towards BNC differentiation. Compared to other BNC subtypes, Troph_B2 exhibited higher expression of PAG6, PAG7, PAG9, PAG19, PAG20, PAG21, PRP2, PRP4, PRP6, PRP8, and PRP12, along with high expression of chorionic somatomammotropin hormone 2 (CSH2, i.e., placental lactogen). Troph_B3 and Troph_B4 both showed significant upregulation of TFs PPARG and TCF7L2. Troph_B3 additionally expressed GCM1 and the higher level of CDK6/CDK18, whereas Troph_B4 uniquely expressed E2F3 and also showed high expression of E2F7, representing the most abundant BNC subtype (Additional file 1: Fig. S4a). The remaining Troph_B5 and Troph_B6 highly expressed TFRC, a marker shared by human syncytiotrophoblast (STB) [17], mouse syncytiotrophoblast layer I (SynT-I) [21], and macaca fusion-competent cytotrophoblasts (fcCTBs) [42]. Notably, Troph_B5 additionally upregulated GLIS1, a characteristic marker of mouse SynT-I [20], and expressed E2F5 and CDK8. In contrast, Troph_B6 specifically expressed CLDN3, a tight-junction component and a known marker of smooth chorion (SC) trophoblast-like cells (XTBs) in macacas [42], as well as CDKN2A, a canonical marker of cell cycle arrest, suggesting that Troph_B6 may represent a distinct homeostatic or trans-differentiation stage. Together, these findings highlight the extensive cellular heterogeneity within the developing bovine placenta and underscore the intricate temporal dynamics in cell populations and their transcriptional programs. To further validate the trophoblast subtypes and their spatial heterogeneity revealed by single-cell transcriptomics within the native tissue context, we performed multi-omics in situ pairwise sequencing (MiP-seq) [56] on E60 bovine placental sections (Fig. 2d, Additional file 1: Fig. S5a, and Additional file 2: Table S5). As expected, the expression patterns of canonical markers PAG8 and PAG6 exactly confirmed the spatial anatomical localization of UNCs and BNCs within the tissue architecture, respectively. Further, VIM was found to significantly co-localize with PAG8 in the Troph_U1 sub-cluster, while SULT1E1 exhibited distinctly localized signals across both UNC and BNC populations. Importantly, markers for specific BNC subtypes, including SYCN (Troph_B1), THSD4 (Troph_B3), E2F3 (Troph_B4), GLIS1 and TFRC (Troph_B5), and CLDN3 (Troph_B6), were accurately mapped to their respective anatomical niches within the E60 bovine placenta (Additional file 1: Fig. S5a). These high-throughput in situ data provide direct and definitive morphological evidence for the identified trophoblast heterogeneity and their spatial topology, thereby confirming the existence of these cellular sub-clusters in vivo.

To further examine whether trophoblast subtypes in cows are conserved with those from other mammalian species, we incorporated the human [57–59], macaca [42], and mouse [21] placental single-cell/-nucleus RNA-seq datasets harboring well-defined trophoblast subtypes and performed cross-species analyses. We found that, among the top 50 markers, except for certain genes associated with cell cycle and stemness that were shared by progenitor and stem cell populations, only a limited number (no more than eight) of genes were shared among almost all bovine trophoblast subtypes and their human, macaca, and mouse counterparts (Additional file 2: Table S6). Despite the overall limited conservation, bovine trophoblast subtypes, compared with their porcine counterparts [51], shared more markers with trophoblast subtypes from these species, suggesting that cows retain more evolutionarily conserved molecular features with primates and rodents.

Temporally, the proportions of IFNT⁺ Troph1 and IFNT⁺ Troph2 peaked on E16-E18, with a subset of cells beginning to transit towards the Troph_U1 subtype (Fig. 2e). This period corresponds to conceptus elongation to form a filamentous type and the onset of trophectoderm contact with the endometrial luminal epithelium, representing the first step towards implantation in cattle [22]. Notably, BNCs were found to emerge on E18-E24, consistent with previously reported timing for their first appearance [6, 10, 60]. To more precisely characterize the temporal distribution of trophoblast subtypes, we developed a customed metric termed the Stage Preference Index (SPI), which was inspired by previously published methods [61]. This index quantifies the temporal distribution bias of each cell type by calculating the ratio of observed to expected cell counts (R_o/e) across different gestational timepoints. As shown in Fig. 2f, trophoblast populations could clearly be categorized into four temporally distinct groups: 1, on E12-E14, mural and polar TE were significantly enriched and defined as UNC_First; 2, on E16-E18, IFNT⁺ Troph1 and IFNT⁺ Troph2 showed prominent enrichment, defined as UNC_Second; 3, on E24-E50, Troph_U1 and Troph_U2 were highly enriched and defined as UNC_Third, while the concurrent BNC subtypes (Troph_B1 and Troph_B2) were designated as BNC_First; 4, on E60-E280, Troph_U3 and Troph_U4 were predominantly enriched (defined as UNC_Four), and the concurrent BNC subtypes (Troph_B3, Troph_B4, Troph_B5, and Troph_B6) were defined as BNC_Second. These findings reveal the temporal transitions and functional shifts of trophoblast subtypes throughout bovine gestation.

Stage-resolved TF modules delineated the regulatory landscape of trophoblast development

To further elucidate the transcriptional regulatory mechanisms underlying the temporal transitions and functional evolution of trophoblast cells during bovine gestation, we subsequently constructed a customized cisTarget database based on the cow genome. We integrated single-nucleus transcriptomic data from 137,954 trophoblast cells (Fig. 3a) and applied a method called SCENIC (single-cell regulatory network inference and clustering) [62] to evaluate the regulatory activity of GRNs in each cell. Through the AUCell algorithm, we selected 190 bovine TF regulons with high regulatory activity that were defined as active TFs (Additional file 3: Table S1).

Fig. 3 — Stage-resolved TF modules delineated the regulatory landscape of trophoblast development in cows. a UMAP projection of trophoblast subtypes during gestation. b The identified regulon modules based on the regulon CSI matrix, as well as the representative TFs. c The dot plot showing the enriched GO terms in each regulon module. d The TF network in cows (CSI ≥ 0.8). Based on the regulon matrix of bovine trophoblast cells, eight TF modules were identified, illustrating the relationships between modules and the representative TFs. e The average module activity scores illustrated by UMAP. f The average expression level of each module in each cell type. g The dynamic changes in the activity of eight regulatory modules across gestational timepoints on E12-E280. Box plots represent the distribution of module activity at each timepoint, and red lines indicate the smooth trend curves

TFs often work synergistically to modulate gene expression. To systematically characterize the synergistic patterns, we compared the similarity of RAS scores for every regulon pair across cow trophoblast cells using the Connection Specificity Index (CSI) [63]. Strikingly, these 190 regulons were categorized into eight major modules (Fig. 3b). We then performed a Gene Ontology (GO) analysis to characterize the functions of module TF regulons (Fig. 3c and Additional file 3: Table S2-S3). Among the cell category-specific modules, modules 1 and 4, harboring the module TFs E2F1/8 and GATA2/3, respectively, were enriched with function terms such as regulation of transcription by RNA polymerase II and positive regulation of DNA-templated transcription, indicating that these two modules are mainly involved in cell cycle and might be related to stemness. Module 2 included TFs DLX3, HAND1, and TFAP2C, and the GO analysis revealed enrichment terms such as intracellular receptor signaling pathway and response to peptide hormone. Module 3, containing TFs DBP, NANOG, and SPI1, was related to myeloid cell differentiation and cell fate commitment. Module 5, consisting of a small number of TFs (e.g., E2F3, E2F7, and YY1), showed no salient enrichment patterns. Module 6, harboring TFs THRB and TBX5, was involved in response to hormone and cellular response to hormone stimulus. Module 7 contained BHLHE40 (a factor related to preeclampsia [64]) and TBX3 (a key regulator for the differentiation of CTBs into STB [65]). Module 8, containing ETS2, TEAD3, and SREBF2, was enriched with regulation of RNA biosynthetic and metabolic processes (Fig. 3d). Next, we confirmed cell type-specific modules with regulon activity. Mapping the regulons within each module across different cell types and gestational timepoints, as shown in Fig. 3e-g, revealed that these TF regulon modules also displayed cell type-specific expression patterns and stage-specific activity profiles. For example, Modules 1 and 2 exhibited strong regulatory activity on E12-E18 that declined rapidly thereafter, suggesting primary roles in pregnancy recognition and pre-implantation lineage specification. Module 3 was predominantly active in Troph_U2 cells, peaking on E24 before the rapid decrease. Modules 4 and 5 were highly activated in Troph_U3 and Troph_U4 cells. Module 6 showed enhanced activity during late gestation, particularly near term, potentially involved in immune response, hormonal signal sensing, and parturition-associated immune modulation. Modules 7 and 8 were prominently activated in BNCs and, like Modules 4–6, culminated during late gestation. Collectively, these temporal and cell type-specific activity patterns highlight coordinated transcriptional programs underlying trophoblast differentiation and functional maturation.

Later, we performed the conversion of orthologous TFs between cows and other species (humans [57–59], macacas [42], and mice [21]) (Additional file 1: Fig. S6a-c), to interrogate whether the TF regulon modules in bovine trophoblast cells are also conserved in trophoblasts from other mammalian species. We found that Modules 1, 3, and 6 displayed consistently low activity throughout gestation in all three species. In contrast, the activity of Module 4 showed a clear upward trend during late gestation across species. In the human placenta, this module was most active in late EVTs and STB, and in macacas, its peak activity occurred in late intrinsic CTBs (iCTBs), fcCTBs, and EVTs, whilst in mice, it was active in glycogen trophoblast cells (Gly-T) and primary parietal TGCs (P-TGCs). In addition, similar to that in bovine trophoblasts, the activity of Module 2 also exhibited manifest fluctuation during early gestation in humans, macacas, and mice, with its activity declining sharply on E7 in humans, on E26 in macacas, and on E8.5 in mice. These results mirror the species-specific and evolutionarily diverse transcriptional regulatory network in trophoblast development.

The developmental atlas of UNCs revealed key TFs during implantation

To delineate the temporal molecular features of trophoblast cells during gestation, we subsequently performed stage-specific differential expression analyses for 109,809 UNCs, using a replicate-aware pseudobulk + limma-voom framework [66] as the primary approach and Seurat FindAllMarkers + Wilcoxon as a complementary validation approach (Fig. 4a, b, Additional file 1: Fig. S7a-c, and Additional file 4). At the UNC_Second stage (E16-E18), members of the IFNT family and the ruminant-specific trophoblast Kunitz domain proteins (TKDP) were found to be markedly upregulated, constituting the core signaling axis for maternal recognition of pregnancy. Concomitantly, genes associated with conceptus elongation, such as HAND1 and FADS1, as well as trophoblast markers PAG2 and PAG12, were also activated. At the subsequent UNC_Third stage (E24-E50), expression of CSH2, SULT1E1, and PHLDA2 was markedly increased. The gene set enrichment analysis (GSEA) further corroborated this functional transition (Fig. 4c and Additional file 1: Fig. S8): compared with those in UNC_First, cells at the UNC_Second stage were significantly enriched with IL2_STAT5_SIGNALING and HALLMARK_IL6_JAK_STAT3_SIGNALING (adjusted P < 0.01), both closely linked to immune evasion mechanisms characteristic of human EVTs [67, 68]. Moreover, immune-related pathways including HALLMARK_COMPLEMENT, HALLMARK_TNFA_SIGNALING_VIA_NFKB, and Inflammatory Response remained enriched throughout the UNC_Third stage. Notably, MYC_TARGETS and HALLMARK_DNA_REPAIR were significantly downregulated in UNC_Second, consistent with the transition from a highly proliferative state to functional differentiation. Finally, at the UNC_Four stage (E60-E280), EMT pathways were significantly downregulated compared with those in UNC_Third, which was accompanied by concomitant repression of the apoptosis-associated HALLMARK_P53_PATHWAY, suggesting that trophoblast cells progressively exit the migratory and plastic state, favoring structural stabilization of the placenta.

We then systematically examined the temporal expression dynamics of stage-specific markers to pinpoint when their expression initiates, reaches the peak, and attenuates. We identified that trophoblast cells displayed clear sequential activation of transcriptional programs across developmental timepoints (Fig. 4d). On E12-E14, genes involved in mitochondrial energy metabolism, ribosomal activity, and maintenance of stemness (ATP5F1E, DDX21, NDUFC1, ATP5PD, ASCL2, and CDX2) were transiently upregulated, followed by rapid downregulation. In the subsequent E14-E30 interval, core genes associated with pregnancy recognition and conceptus elongation (IFNT, IFNT2, IFNT3, TKDP1, TKDP4, FADS1, and PTGS2) were transcriptionally activated on as early as E14, displaying maximal expression on E16-E18, and their expression then sharply declined to an undetectable level by E30, marking this period as a critical window for maternal recognition of pregnancy and implantation in cows. Notably, PAG2, PAG8, PAG11, and PAG12 were concurrently upregulated at this stage, with their expression exhibiting high overlap with that of IFNT, despite differences in the peak time, consistent with previous observations [69]. Besides, the expression of several TFs implicated in placental maturation (TEAD1, TCF7L1, TP63, and PPARG) displayed a consistent increase, peaking at late gestation (post-E60) before declining on E240.

Upon construction of the regulatory network, we next identified TFs with stage-specific activation patterns across gestation (Fig. 4e, Additional file 1: Fig. S9, and Additional file 4) that closely mirrored the temporal trends observed in TF module activity. Specifically, CDX2, SOX15, and OTX1 in Module 1, as well as SP6, HAND1, and DLX5 in Module 2, were selectively activated on E12-E18. We particularly focused on the E14-E30 window, in that it encompasses both conceptus elongation and the classical period of maternal recognition of pregnancy, and also represents a high-risk interval for early pregnancy loss (~ 30% average loss) in cows [1]. During this interval, the key signaling molecules IFNT and TKDP were dynamically expressed, with their expression initiating on E14 and terminating by E30. Along the temporal axis, TFs specifically expressed on E14 included OTX2, MYBL2, and ETS2, of which ETS2 is well established as a direct transcriptional activator for both IFNT and TKDP [70]. On E16-E18, the predominant TFs were SP6, HAND1, DLX5, and IRF1, of which HAND1 is a conserved regulator of trophoblast differentiation and has previously been shown to modulate multiple lineage-defining genes in both mouse and human trophoblasts, while IRF1 is a classical interferon-stimulated gene (ISG) induced by type I interferons including IFNT [71]. On E24-E30, the key TFs expressed were HOXB2, FLI1, and IRF9, with IRF9 also being a canonical type I interferon-induced ISG [72]. The sequential TF activation highlights the dynamic coupling between interferon signaling and the transcriptional regulatory network during the maternal recognition window. To further pinpoint the core regulators in this process, we extracted the top 30 predicted target genes from the E14-E30 stage-specific regulons (Fig. 4f, g, Additional file 1: Fig. S10, and Additional file 4), and identified that HAND1 and DLX5 co-regulated multiple genes critical for pregnancy recognition and conceptus elongation, such as IFNT, IFNT2, IFNT3, TKDP1, and TKDP4. Notably, HAND1 and DLX5 not only targeted each other reciprocally but also exhibited activation time (E16-E18) coinciding with the peak expression of their target genes, further supporting their synergistic roles during the maternal recognition window (Fig. 4h). Collectively, these findings delineate the stage-specific transcriptional dynamics and key TF networks in bovine trophoblast cells, providing novel mechanistic insights and potential targets for understanding of maternal–fetal interface establishment and pregnancy maintenance.

Cell trajectory analysis revealed the developmental origin and key regulatory networks of BNCs

BNCs were first observed on E24, suggesting that this timepoint represents a critical developmental window for the initiation of BNC differentiation. To further elucidate the origin and differentiation trajectory of BNCs, we probed the lineage relationships and pseudotime dynamics of trophoblast subpopulations on E24. High-resolution clustering (Fig. 5a) revealed that the UNC population could be further subdivided into seven subtypes (C1-C7), predominantly composed of Troph_U1 and Troph_U2 cells, with a minor fraction of IFNT⁺ Troph cells. For the BNC population, they could be divided into three subtypes (C8-C10): C8 corresponded to BNC2 and mainly derived from Troph_B3 and Troph_B4, whereas C9 and C10 (BNC1) were composed of Troph_B1, Troph_B2, and Troph_B6 (Additional file 1: Fig. S11a). Notably, both BNC subtypes and UNCs exhibited a continuous transitional pattern in the UMAP embedding (transcriptional state space) and the transcriptomic similarity tree topology (Fig. 5a, b), supporting the notion that BNCs arise from UNCs through gradual transcriptional reprogramming [10, 60].

We also found that all UNC subpopulations exhibited high expression of canonical trophoblast markers such as PAG2, PAG8, and PTGS2, while displaying distinct transcriptional profiles. Specifically, C1 showed marked upregulation of CITED1, CDKN1C, and FBP1; C2 expressed elevated levels of GRIA4, NCKAP5, and SIPA1L1; C3 upregulated MITF, DNM3, EPS8, SOX6, and SOX5; C4 expressed higher levels of GPC3, SLIT3, COL4A2, and COL4A1; C5 upregulated PAGE4, PINLYP, CXCL16, and MME; C6 expressed more transcripts of EPHB6; C7 specifically expressed SLC16A9 and MTUS2. Within the BNC compartment, BNC1 (C9 and C10) and BNC2 (C8) exhibited clear functional and status distinctions. Among these, BNC1 highly expressed PAG (e.g., PAG3, PAG10, PAG14-16, and PAG19-21) and PRP family members (e.g., PRP1, PRP6, and PRP14). Notably, BNC1 (C9 and C10) showed high expression of RUM1, GCM1, CSH2, PGF, and ADM. Although RUM1-mediated fusion is pH-dependent in vitro [73], its co-expression with the master regulator GCM1 in BNC1 suggests that BNC1 represents a mature secretory state primed for syncytialization, the physiological triggers of which warrant further functional investigation (Fig. 5c-e). In contrast to BNC1, BNC2 (C8) showed higher expression of TFs TP63, STAB2, JUN, PPARG, TCF7L2, TCF7L1, as well as MSI2, indicating that BNC2 represents a molecularly distinct BNC-associated state with a transcriptional program different from that of BNC1. To further assess whether BNC1 and BNC2 represent two persistent BNC-associated states or a transitional sequence within a single differentiation path, we explored the expression patterns of their representative markers/signatures at later developmental timepoints. The results showed that these two molecular features were already distinguishable on E24 and remained detectable in the corresponding trophoblast populations on E50, E180, and E240 (Additional file 1: Fig. S11b). In addition, in the published E195 dataset [3], BNC1 was more similar to the cotyledonary-derived population, whereas BNC2 was more similar to the intercotyledonary BNC population (Additional file 1: Fig. S11c). TF regulatory activity patterns further showed greater similarity between BNC1 and UNC1, and between BNC2 and UNC2 (Additional file 1: Fig. S11d). Together, these results support BNC1 and BNC2 as two BNC-associated states with distinct transcriptional and regulatory features and suggest that they may arise from different UNC subpopulations.

Next, we conducted an integrative analysis for RNA velocity and the Monocle3 pseudotime trajectory (Fig. 5f), and found pronounced heterogeneity in the differentiation trajectories of BNC1 and BNC2. Specifically, BNC1 originated directly from C1 (Troph_U1/U2) via the differentiation pathway DF1, whereas BNC2 arose from C3 (comprising Troph_U3/U4), first transiting to C4 (Troph_U1) and subsequently differentiating into mature BNCs via the pathway DF2. Additionally, pathways DF3 and DF4 uncovered potential phenotypic remodeling and state transitions within the UNC population (Fig. 5f). To elucidate the regulatory mechanisms driving lineage specification of BNC subtypes, we then traced key feature genes that exhibited dynamic changes along the differentiation trajectories. The results demonstrated that BNC differentiation was closely associated with the parental genomic imprinting state (Fig. 5g and Additional file 1: Fig. S12a). It was found that multiple maternally and paternally imprinted genes, as well as epigenetic regulators (e.g., ATP2C2, INPP5F, PEG3, CDKN1C, PHLDA2, and PPP1R9A), displayed trajectory-specific dynamic expression patterns. Among these, the paternally imprinted genes ATP2C2, INPP5F, and PEG3, as well as the maternally imprinted gene PPP1R9A, showed evidently dynamic expression along the BNC2 trajectory (DF2), whereas the maternally imprinted genes CDKN1C and PHLDA2 exhibited markedly dynamic expression along the BNC1 trajectory (DF1). These findings suggest that genomic imprinting and epigenetic regulation may jointly orchestrate subtype-specific differentiation and functional specialization of BNCs.

Based on histological and cell biological evidence from ruminant placentation, BNCs are thought to arise from progenitor UNCs through endoreplication-related processes, potentially involving acytokinetic mitosis-/endomitosis-like events and/or endoreduplication-/endocycle-like programs [10, 74], and it has been reported that canonical (E2F1/2/3) and atypical (E2F7/8) E2Fs jointly regulate endocycles in vivo [75]. Our single-cell transcriptomic data showed increased expression of both arms during BNC formation, consistent with E2F-linked cell cycle rewiring during polyploidization. Also, the mitosis-associated core kinase CDK1 was upregulated, along with increased expression of additional cyclin-dependent kinases (e.g., CDK6, CDK8, CDK14, and CDK18) during BNC formation. Notably, GCM1, a key factor implicated in mouse TGC differentiation [76], was also upregulated during BNC formation (Additional file 1: Fig. S12b).

Further, by integrating the preceding TF module data (Fig. 5h and Additional file 1: Fig. S12c), we found that a host of members in Modules 4, 5, 7, and 8 were involved in BNC differentiation. In addition to well-established regulators of trophoblast fusion and differentiation (e.g., TBX3, TEAD1, TP63, and PPARG), we also identified novel candidate factors such as JARID2, SATB2, TCF7L2, BHLHE40, MTA3, and NCOR2. Notably, JARID2 and BHLHE40 have been demonstrated to be implicated in the pathogenesis of human preeclampsia [64, 77], whilst loss of NCOR2 led to abnormal trophoblast morphology and reduced proliferation [78] (Fig. 5h). Here, we found that TCF7L2 showed gradual upregulation concomitant with BNC1 and BNC2 formation and that it was present at the terminal differentiation branches, in sharp contrast with previous reports that TCF7L2 was exclusively expressed in undifferentiated UNCs [3]. The metabolic pathway analysis further disclosed that glycolysis/gluconeogenesis and oxidative phosphorylation were downregulated during BNC differentiation, whereas propanoate metabolism was significantly upregulated during BNC formation (Fig. 5i and Additional file 1: Fig. S12d). The TCA cycle, the pentose phosphate pathway, and fructose-mannose metabolism were specifically downregulated along the BNC1 trajectory, highlighting the tight correlation between metabolic reprogramming and BNC fate determination. Collectively, these findings delineate the lineage origins, differentiation trajectories, core regulatory factor networks, and metabolic remodeling of BNCs on E24, providing a new mechanistic framework and candidate targets for understanding of BNC genesis in ruminant placentas.

Trophoblast development in the bovine placenta involved extensive transcriptional and metabolic reprogramming

Subsequently, we performed a systematic clustering analysis for transcriptional and metabolic profiles in bovine trophoblasts throughout the entire gestational timeline, and identified stage-specific molecular programs. At the transcription level, clustering based on gene expression patterns partitioned trophoblasts into 16 transcriptional modules (C1-C16), with each exhibiting dynamic changes closely linked to specific biological functions (Fig. 6a and Additional file 5): Early-activated modules (C3 and C8; E12-E18) were characterized by IFNT and KRT18 and enriched with “embryo development” and “cation transport”, indicative of roles in early maternal–fetal signaling and cytoskeletal remodeling; Mid-peak modules (C1, C2, C4, C5, and C10; E24-E30) were characterized by PAG8, PAG4, and MMP2 and enriched with “mitotic cell cycle” and “ECM organization”, suggestive of active proliferation and matrix remodeling after implantation; Mid-late maintenance modules (C6, C9, C11, C12, C13, C14, and C15; E50-E180) were characterized by TFRC, GLG5, CDKN1B, and TAC1 and enriched with “multi-organism reproductive process”, “Hippo signaling”, and “lipid biosynthetic process”, suggesting roles in maintenance of placental architecture and functions; Late-specific modules (C7 and C16; E240-E280) were characterized by FLT1, CXCL8, and IL19 and enriched with “response to bacterium”, “inflammatory response”, and “response to cytokine”, related to pre-parturition immune regulation.

At the metabolic level, clustering of 85 metabolic pathways uncovered eight metabolic modules with distinct temporal patterns (Fig. 6b and Additional file 5): Early-activated modules (Mp7 and Mp6) were enriched with “galactose metabolism”, “fatty acid degradation”, and “butanoate metabolism”, suggesting frequent utilization of energy substrates and fatty acid catabolism during early gestation; Mid-/late-activated modules (Mp1, Mp3, and Mp8) were enriched with “propanoate metabolism”, “steroid hormone biosynthesis”, and “glycosphingolipid biosynthesis”, linked to placental hormone production and membrane remodeling; The constitutively active module (Mp5) was enriched with “D-amino acid metabolism”, “inositol phosphate metabolism”, and “lysine degradation”, potentially providing stable metabolic support throughout gestation; The continuously downregulated module (Mp4) was enriched with “glycolysis/gluconeogenesis”, “citrate cycle (TCA cycle)”, and “oxidative phosphorylation”, indicating a shift away from high glucose oxidation with placental maturation. This integrated transcriptomic and metabolic framework delineates the functional transitions and energy demand adaptation in bovine trophoblasts throughout gestation.

Early trophoblasts were associated with GL and exhibited a strong evolutionary constraint

To elucidate the role of distinct placental cell types in the genetic regulation of GL, we integrated large-scale GWAS summary statistics for GL (N = 3,683) with single-cell transcriptomic data from bovine placentas spanning the entire gestational period. Through the integrative analysis, we identified that eight trophoblast subtypes (Fig. 7a and Additional file 6: Table S1) (IFNT⁺ Troph_2, mural TE, polar TE, Troph_B1, Troph_U2, Troph_B2, IFNT⁺ Troph_1, and Troph_U1), two epiblast-derived cell types (hypoblasts and epiblasts), epithelial cells (urothelial epithelial cells and LEs), and mesenchymal cells (Mes) showed significant genetic associations with GL after multiple-testing correction (FDR < 0.05). Strikingly, all eight GL-associated trophoblast subtypes were present at early gestation (E12-E50) and exhibited gestational stage-dependent heterogeneity in their association patterns (Additional file 1: Fig. S13a-b and Additional file 6: Table S2), suggesting that early trophoblasts are likely important cellular contributors to GL regulation and pregnancy maintenance.

The further pathway analysis revealed the enriched key biological processes related to these trophoblast populations (Fig. 7b), such as innate immunity (e.g., NOD‑like receptor signaling), protein degradation and homeostasis (the proteasome pathway), cell migration and adhesion (phospholipase D signaling, regulation of the actin cytoskeleton, Ras signaling, and Rap1 signaling), metabolic regulation (glycolysis/gluconeogenesis, growth hormone signaling, oxidative phosphorylation), programmed cell death (apoptosis, p53 signaling), and reproductive hormone regulation (GnRH signaling). In addition, by correlating gene expression levels with the summed genetically associated pathway activity scores (gPAS) of placental cells, we identified top ten genes most strongly associated with GL (Fig. 7c and Additional file 6: Table S3): CYCS, CALML5, HMGA1, PERP, RAB25, VDAC1, CLDN6, ATP5MC1, LMO1, and C1QBP. Notably, the majority of these genes exhibited high expression in early trophoblast populations (Fig. 7d). Overall, these findings provide additional insights into the genetic and cellular basis of GL variations.

Based on these findings, we next performed cross-species analyses for the tolerance to functional mutations of genes [79] (Additional file 6: Table S4) in different placental cell types throughout gestation in cows, humans, mice, and macacas, and for the percentage of lethality-associated genes expressed in these species, on the basis of a set of neutrally ascertained mouse knockouts [80] (Additional file 6: Table S5). We found that early trophoblast populations, which were identified as the core cell types for GL determination, tended to exhibit stronger functional constraints than other placental cell types or stage counterparts, evidenced by higher mean pLI scores (i.e., lower loss-of-function tolerance), and a higher proportion of lethality-associated genes expressed across species (Fig. 7e-h and Additional file 1: Fig. S13c-h). Taken together, these findings suggest that during the critical window of pregnancy establishment, early trophoblasts may combine cellular plasticity and turnover with relatively strong gene-level constraints, a pattern potentially conducive to the execution of essential placental functions.

Dynamic polarization of placental macrophages and its correlation with genetic susceptibility and signaling networks underlying RP

In contrast to the trophoblast-centered pattern observed for GL, RP was associated with a macrophage-related program. Expression of canonical macrophage markers was not evident on E18 or E24, but became detectable since E30 and remained at later timepoints, indicating that placental macrophage-like populations become clearly discernable from E30 onwards and undergo continuous transcriptional reprogramming throughout gestation (Additional file 1: Fig.S14a). In early pregnancy (E30-E60), they were enriched with interferon/inflammatory and stress-response pathways (e.g., IFI27, ISG15, CCL8/CCL2, S100A family, FOS, and EGR1). During mid-gestation (E85-E110), genes associated with phagocytic clearance, iron export, and vascular/matrix remodeling (MERTK, MRC1, VCAM1, ITGAM, VSIG4, and SLC40A1) were markedly upregulated. In late gestation (E180-E280), programs related to chemotaxis, coagulation/hypoxia, and angiogenesis (CXCL family, SERPINE1, and FLT1) were enhanced, suggesting a stage-specific functional transition (immune surveillance to repair and remodeling to prepartum inflammatory reactivation) (Fig. 8a, Additional file 1: Fig. S15, and Additional file 7: Table S1-S3). The polarization analysis further revealed the pronounced plasticity and heterogeneity of placental macrophages. From E30 onwards, M2 scores were consistently higher than M1 scores, indicating an M2-dominant state at the maternal–fetal interface, in line with the report on human placentas [81]. On E30, macrophages exhibited an intermediate state with high M1 and M2 scores, both of which subsequently declined and bottomed out on E58-E84. Thereafter, M1 scores stabilized during E105-E180, and steadily increased from E180 to parturition, whereas M2 scores again went up from E85, culminated at mid-to-late gestation (~ E180), and then gradually decreased. Collectively, these findings delineate a dynamic polarization trajectory characterized by M2 predominance and prepartum M1 rebound, supporting their critical roles in immune tolerance and initiation of parturition (Fig. 8b and Additional file 7: Table S4).

To probe whether the aforementioned temporal programs carry genetic susceptibility to RP, we integrated bovine RP GWAS data with single-cell transcriptomic data, and calculated RP-relevant scores for each placental cell type (Fig. 8c and Additional file 7: Table S5-S7). The results showed that macrophages ranked among the top three significantly associated cell types (P = 4.37 × 10^–9), indicating that genetic signals preferentially converge on this immune-vascular remodeling hub. Vascular endothelial cells, VSMCs, fibroblasts, GECs, and specific trophoblast subtypes (Troph_U3 and Troph_U4) also reached statistical significance. The cross-species analysis further demonstrated that macrophages were also significantly enriched with RP GWAS signals in independent human placental datasets, supporting macrophages as a cross-species convergent cell type implicated in RP-associated genetic signals (Fig. 8d and Additional file 7: Table S8-S12). Later, by pathway integration in macrophages enriched with GWAS signals (Fig. 8e and Additional file 1: Fig. S16a-c), we identified seven conserved RP-associated pathways: GnRH secretion, fat digestion and absorption, the apelin signaling pathway, the p53 signaling pathway, pantothenate and CoA biosynthesis, glycerolipid metabolism, and the B cell receptor signaling pathway. These pathways aligned well with the aforementioned temporal polarization programs, i.e., metabolic/endocrine modules (lipid metabolism, GnRH, and apelin) and inflammatory/stress modules (p53 and BCR) converged in macrophages, suggesting that RP-associated genetic signals may converge on stage-specific polarization and communication programs in placental macrophages.

Finally, we performed a cell–cell communication analysis for RP-associated cell types in bovine placentas (Fig. 8f and Additional file 1: Fig. S17). We identified that since their emergence on E30, macrophages exhibited extensive interactions with endothelial cells, VSMCs, fibroblasts, and multiple trophoblast subtypes, with the TNF-TNFR, TGFβ-TGFβR, and SPP1 axes persistent throughout gestation but gradually decreasing in intensity. In contrast, the activity of inflammatory chemotactic pathways (IL1-IL1R and CCL/CXCL-CCR/CXCR) peaked on E280, suggesting the activation of parturition-associated inflammation. Collectively, these findings illuminate placental macrophages at three scales, i.e., genetic signal enrichment, core pathway involvement, and intercellular communication, and support a model in which placental macrophages act as a hub linking RP-associated genetic signals to microenvironmental remodeling during pregnancy.

Discussion

Due to the lack of documentation of the entire dynamic placentation, the mechanisms underlying ruminant placental development remain poorly understood. To acquire knowledge in this regard, novel techniques such as scRNA-seq and spatial transcriptomics have been applied to ruminant placentas. Nonetheless, typically only placental cells at a single developmental timepoint were analyzed [3, 23, 82], and recent snRNA-seq studies, albeit with multiple developmental timepoints covered, mainly focused on peri-implantation embryo development and embryo-extraembryonic interactions [22, 23, 29]. Here, we have for the first time, to the best of our knowledge, provided a single-cell-resolution atlas of trophoblast differentiation spanning the entire course of bovine gestation, and systematically characterized the cellular and molecular landscape of the bovine placenta. We identified 13 major cell types and 14 trophoblast subtypes, and precisely delineated their spatiotemporal dynamics and functional features. Furthermore, we fully annotated TFs closely associated with trophoblast differentiation trajectories and pregnancy progress, thereby illustrating the regulatory network and providing a valuable resource for future studies on cotyledonary placental development.

Pregnancy failure can occur at multiple stages and markedly undermines both reproductive performance and economic returns in dairy herds. Of overriding importance are E8-E27 that encompass embryo elongation and the classical "maternal recognition of pregnancy" period, when the average loss can arrive at ∼30% [1]. During this window, the embryo must secrete sufficient IFNT to prevent luteolysis by inhibiting the endometrial release of luteolytic prostaglandin F2α (PGF2α) pulses [83], thereby ensuring the survival of the corpus luteum (CL) and its vital progesterone production. Hence, failures or delays in trophoblast elongation and IFNT signaling are the primary causes for pregnancy loss in this period [1]. Here, our single-cell/-nucleus transcriptomic data revealed that IFNT and TKDP1/4 were synchronously activated (initiating on E14, peaking on E16-E18, and sharply declining by E30), whereas TKDP2 expression was clearly delayed and more closely aligned with the upregulation of PAG8/PAG2. While previous studies showed that only TKDP1 was synchronous with IFNT [70], our findings extend this paradigm across the TKDP family, a ruminant-specific gene family uniquely expressed in trophoblasts, revealing a distinct phase-partition among its members. This temporal divergence indicates that the IFNT-TKDP axis is not a single-channel signal but rather a “time-gated ensemble” composed of asynchronous members: early members (IFNT and TKDP1/4) that establish the threshold for maternal recognition of pregnancy, as well as the late member (TKDP2) that acts in concert with the PAG program to stabilize the maternal–fetal interface and to promote tissue remodeling, thereby interpreting the vulnerability during early pregnancy. Besides, we not only pinpointed the roles of canonical factors such as ETS2 [69], IRF1 [71], and IRF9 [72], but also identified HAND1 and DLX5 as newly activated regulators that peaked in synchrony with IFNT and that co-targeted key markers. Previous studies showed that conditional deletion of HAND1 disrupted placental vascular development, caused global dysregulation of trophoblast gene expression, and led to both early and late fetal loss [84, 85], while DLX5 overexpression reduced trophoblast proliferation, perturbed metabolism, and induced endoplasmic reticulum stress, yielding transcriptional profiles that closely resembled preeclamptic placentas [86]. Notably, DLX5 is expressed in human but not mouse trophoblasts, and it has been identified as a component of the human-specific regulatory network for trophoblast differentiation [86, 87]. Thus, the co-activation of HAND1 and DLX5 in cows not only provides new regulatory leverage for maternal recognition of pregnancy, but also highlights the evolutionary convergence of key regulatory networks between humans and ruminants but not mice, providing critical insights into the species-specific vulnerability during early pregnancy.

Subsequently, by performing a high-resolution cell trajectory analysis, we delineated the developmental origin of BNCs, demonstrating that they arise from specific UNC subpopulations on E24 and are coordinately modulated by genomic imprinting and metabolic reprogramming. The trophoblast cells spanning the entire course of gestation were further partitioned into 16 transcriptional modules and eight metabolic modules. This integrative analysis of transcriptional and metabolic dynamics reveals the functional transitions and energy demand adaptation in bovine trophoblasts at different stages of pregnancy, providing important insights into placental nutrient regulation and a valuable reference catalog for optimization of nutritional strategies during bovine gestation. Besides, our findings elicit an interesting topic of discussion in the context of the classic Wooding/Klisch model [10, 74] of bovine BNC differentiation, in which BNCs are proposed to arise from progenitor UNCs through endoreplication-related processes. Although our single-cell transcriptomic data did not directly resolve the exact cytological mode of polyploidization, the pseudotime analysis revealed dynamic remodeling of cell cycle- and polyploidization-associated gene programs during the UNC-to-BNC transition. In particular, the coordinated dynamics of E2F family members, CDK-related factors, and CDKN1C were consistent with cell cycle rewiring during BNC formation, while the concomitant upregulation of differentiation-associated regulators such as GCM1 suggests coupling between endoreplication-related programs and terminal trophoblast specialization. Thus, our data provide transcriptomic support for the hypothesis that bovine BNCs arise from UNCs through an endoreplication-associated differentiation process.

When it comes to GL, a compelling finding was obtained through our integrative analysis. Specifically, we for the first time, to our knowledge, found GL-associated genetic signals to be significantly enriched in eight early trophoblast subtypes (E12-E50). This finding is noticeable because it suggests that a trait measured in late pregnancy might have already be influenced by trophoblast-associated genetic programs established during early gestation, raising the possibility that some aspects of the placental developmental trajectory relevant to GL are established early through trophoblast proliferation, differentiation, and functional state transitions. These early trophoblast populations showed signatures consistent with relatively strong evolutionary constraints, characterized by both a high proportion of lethality-associated genes expressed and an elevated mean pLI score of expressed genes, in line with the lower tolerance to loss-of-function variations at the gene level. Because pLI is derived from human population genetic data, its relevance to bovine placentas should be interpreted with appropriate caution. This pattern aligns with the idea that early placental development may be subjected to relatively stringent functional constraints, which could contribute to the stability of GL-related developmental programs. Further, we identified CYCS, HMGA1, and VDAC1 as top risk genes, providing candidates for future functional validation and evaluation of their potential relevance to GL-related breeding strategies. Of these, CYCS and VDAC1 are central regulators of the mitochondrial function, orchestrating energy metabolism and cell fate [88, 89]. Notably, CYCS, which is highly evolutionarily conserved, serves as a key factor in the mitochondrial apoptosis pathway and plays an essential role in spermatogenesis [89], and it is also closely linked to human trophoblast cell apoptosis [90]. HMGA1 is known to promote the cancer cell growth and invasion [91] and perceived as a potential driver of preeclampsia by perturbing EVT invasion [92]. In sheep, HMGA1 is also associated with excessive trophoblast proliferation and conceptus elongation during early pregnancy [93]. Yet, given the modest sample size of the GL GWAS, these integrative results should be interpreted primarily as prioritization evidence for candidate cell types and pathways rather than the definitive causal proof.

Our study also provides a new cellular framework for understanding the pathogenesis of RP, a major reproductive disorder. Specifically, we identified placental macrophages as a major cell type prioritized by RP-associated genetic enrichment analyses. Similar enrichment patterns were observed in both bovine and human datasets, suggesting some extents of cross-species convergence at the cellular level. It was discerned that macrophages emerged around E30 and underwent dynamic M1/M2 polarization trajectories, with enrichment of pathways related to lipid metabolism, inflammation, and apoptosis at late gestation. The overlap between genetic signals and these functional programs suggests that the susceptibility to RP may arise when macrophage polarization is disrupted by genetic perturbation, further leading to immune imbalance, compromised tissue remodeling and, ultimately, abnormal placental detachment and expulsion failure. Besides, we identified the RP-associated pathways that were repeatedly enriched in both bovine and human analyses, such as GnRH secretion, lipid metabolism, p53 signaling, B cell receptor signaling, and the Apelin pathway, all of which were enriched in macrophages. Also, endothelial cells, fibroblasts, and specific trophoblast subtypes (Troph_U3 and Troph_U4) displayed significant RP associations, outlining a multi-cellular interaction network involving immune, vascular, stromal, and trophoblast compartments. Within this network, crossroads of metabolic, endocrine, and inflammatory pathways, such as the interaction of lipid metabolism with GnRH/Apelin signaling, and coupling between p53 and apoptosis/repair, are likely to represent key points responsible for the vulnerability. Together, these findings support a multi-layer convergence model in which RP-associated genetic signals may preferentially converge on macrophage polarization programs and the related metabolic, endocrine, and inflammatory pathways, with broader effects on multiple placental cell types. Notably, because bovine and human GWAS were conducted in different populations with distinct genetic architectures, these cross-species analyses were intended to highlight convergent cellular patterns rather than to imply direct equivalence of genetic effects.

To sum up, this longitudinal study provides an invaluable omics resource that goes beyond traditional histological studies to illuminate placental development in ruminants. Technically, we unravel the integrated landscape of cell types, developmental timepoints, and genetic traits within the molecular dynamics of bovine gestation, providing a spatiotemporal atlas that delineates the dynamic gene expression and regulatory networks in specific cell types during key developmental windows, as well as a trait mechanism map that denotes precise cellular origins (e.g., trophoblasts) and pathways (e.g., growth hormone signaling) for important genetic traits such as GL (Fig. 9). Our study also has the evolutionary and translational value, as it reveals conserved mechanisms and underscores the bovine model as a vital resource to understand mammalian reproduction. We acknowledge that technical noise remains a limitation of this integrated atlas. The diverse assay modalities, breeds, and library chemistries inevitably introduce batch effects that cannot be fully eliminated. While our cross-dataset comparisons confirmed the stability of the major cell type framework, these underlying variations may still mask subtle biological signals or rare cell states. Hence, further validation using more uniform platforms and single-breed cohorts would help refine the resolution of these placental networks.

Fig. 9 — A schematic overview of the primary findings in this study. The x-axis denotes the E12-E280 timeline. Trophoblasts differentiate continuously from polar trophectoderm (TE) through the early placenta to the fetal-stage placenta. *IFNT*⁺ trophoblasts emerge on E16, and BNCs arise on E20-E24 via two differentiation routes, followed by formation of multiple trophoblast subpopulations (U1/U2 and B1-B6). Stage-specific key TFs are: early: *OTX2*/*CDX2*/*ETS2* and *SP6*/*HAND1*/*DLX5* (with *DLX5* and *HAND1* involved in maternal–fetal recognition); mid: *ETS1*/*IRF9*/*HOXB2* and *MAF*/*MAFG*/*EGR1*; mid-late: *GCM1*/*TCF7L2*/*TEAD1*; late: *MEIS2*/*NFIA*/*CREB5* and *SOX9*/*BACH2*. The lower panels show the temporally shifting trophoblast gene expression patterns (C1-C16), pathways (Hippo, BMP, inflammation, etc.), and TF modules (Modules 1–8). The upper red/blue curves depict macrophage polarization from M1 to M2 across gestation, with inflammation-related signals going up again before parturition. The upper-left inset summarizes trophoblast molecular states and metabolic/stress pathways associated with GL, whereas the upper-right network illustrates the multicellular interactions at the maternal–fetal interface and the associations between dysregulated ligand-receptor axes (*TNF/TGF*, *IL1/IL6/IL7*, and *CXCL*) and RP during late pregnancy. Abbreviations: TE, trophectoderm; Troph, trophoblast; BNCs, binucleate cells; GL, gestation length; RP, retained placenta; Endo, endothelium; VSMCs, vascular smooth muscle cells; GECs, glandular epithelial cells

Conclusions

This study provides a spatiotemporal single‑cell transcriptomic atlas of bovine placental development across gestation, revealing the cellular dynamics and regulatory programs underlying trophoblast differentiation, pregnancy recognition, and genetic susceptibility to gestation length and retained placenta. These findings offer a valuable resource and conceptual framework for understanding of pregnancy maintenance and improvement of reproductive traits in ruminants.

Methods

Animal experiments and sample collection

Estrus synchronization was achieved via hormonal treatment, and embryo transfer was performed on day 7 post-estrus, as previously described [7]. Pregnant heifers were euthanized on gestational days E50 (n = 2), E60 (n = 3), E85 (n = 3), E110 (n = 2), E180 (n = 4), E240 (n = 4), and E280 (n = 3) to collect placental tissues, following our established protocol [82], which resulted in a total of 21 samples. All tissues were immediately snap-frozen in liquid nitrogen and stored at –80 °C for further use.

Sample preparation for snRNA-seq

Frozen tissue was minced into approximately 2 mm fragments and lysed using a Dounce homogenizer (885,302–0002, Kimble Chase) in 2 ml of chilled EZ lysis buffer (NUC-101, Sigma-Aldrich) supplemented with a protease inhibitor cocktail (5,892,791,001, Roche) and RNase inhibitors (Promega N2615 and AM2696, Life Technologies). An additional 2 ml of cold lysis buffer was added, and the suspension was kept on ice for 5 min. The homogenate was filtered through a 40 μm mesh (43–50,040–51, pluriSelect) and centrifuged at 500 × g for 5 min at 4 °C. The nuclear pellet was resuspended in 4 ml of cold lysis buffer, incubated again on ice for 5 min, and centrifuged. After discarding the supernatant, the nuclei were resuspended in Nuclei Suspension Buffer (1 × PBS containing 0.07% BSA and 0.1% RNase inhibitor) and filtered through a 20 μm mesh (43–50,020–50, pluriSelect). The nuclei were stained with 7-AAD (AP104, MultiSciences) for 5 min and sorted using a BD FACSAria II to remove cellular debris. The final nuclear concentration was adjusted to 700–1200 nuclei/μl prior to loading. The nuclear suspension was processed using the Chromium Single Cell 3′ Reagent Kit v3 (1,000,075, 10 × Genomics) according to the manufacturer's instructions. Libraries were assessed for quality and sequenced on an Illumina NovaSeq 6000 system with 150 bp paired-end reads.

snRNA-seq data analysis

Sample demultiplexing and barcode assignment were performed using Cell Ranger (v7.1.0), and the snRNA-seq reads were aligned to the Bos taurus reference genome ARS-UCD2.0. Quality control of nuclear data was conducted using the Seurat R package (v4.3.0.1), with the following filtering criteria: < 500 genes detected per nucleus, > 20% mitochondrial gene content, and removal of genes expressed in fewer than five nuclei. DoubletFinder (v2.0.3) [94] was subsequently used to identify and remove potential doublets (Additional file 2: Table S1-S2).

For each developmental timepoint, filtered and unscaled Seurat objects from individual placental samples were merged and processed separately for cell type annotation. The merged objects were subjected to normalization, variable feature selection, data scaling, and the principal component analysis using the “NormalizeData”, “FindVariableFeatures”, “ScaleData”, and “RunPCA” functions in Seurat, respectively. Batch effects across samples were corrected using the “RunHarmony” function from the Harmony package (version 1.0.1). Cell clustering was performed with the “FindNeighbors” and “FindClusters” functions, and dimensionality reduction for visualization was achieved using “RunUMAP”. Markers for each cluster were identified using the “FindAllMarkers” function, with significance thresholds set at log₂ fold change > 0.25 and adjusted P value < 0.05, to guide the manual cell type annotation.

To minimize batch effects, we retrieved raw sequencing data at other developmental timepoints from public datasets. These public datasets were generated as scRNA-seq data, whereas the newly generated Holstein data originated from snRNA-seq. For reanalysis, the public datasets were processed using the same reference genome and software versions and standardized analytical workflow as used for the newly generated data to generate comparable Seurat objects. Briefly, reads were aligned to the same bovine reference genome, and downstream processing was performed using the same software pipeline. Quality control for each public dataset was conducted according to the criteria reported in the corresponding original publication. These Seurat objects from all developmental timepoints were subsequently integrated. To select the most effective integration strategy, we applied the widely recognized scIB framework [95] to benchmark the performance of several commonly used integration methods, i.e., CCA, scVI, Harmony, BBKNN, fastMNN, and Scanorama. Each method was evaluated based on multiple biological conservation metrics (Isolated labels, Silhouette label, Clisi, KMeans NMI, and KMeans ARI) and batch correction metrics (Silhouette batch, iLISI, Graph connectivity, and KBET). These metrics were aggregated into a composite score to rank the overall performance. In our dataset, scVI demonstrated the consistent and robust performance across key metrics and achieved the highest overall score (Additional file 3: Table S3). Therefore, scVI was selected as the primary integration method in this study. Following integration, markers were re-identified using the FindAllMarkers function in Seurat (log₂ fold change > 0.25 and adjusted P value < 0.05) to guide cell type annotation.

To determine the genetic origin of each barcode-labelled cell in the absence of prior individual genotype information, we employed the Souporcell Python tool [96]. This approach involves the integration of the ARS-UCD2.0 reference genome, variant VCF files from the 1,000 Bull Genomes Project, and Cell Ranger-filtered barcodes to analyze BAM files generated by Cell Ranger for post-E24 placental samples, thereby enabling to infer the genetic identity of each cell. Upon completion of the Souporcell analysis, each cell was assigned a status (singlet, doublet, or unassigned) and a genetic origin (maternal, fetal, or ambiguous). These assignments were then integrated with transcriptome-based dimensionality reduction results to confirm the maternal or fetal origin of each annotated cell type. Only cells/nuclei classified as singlets with a confident and unambiguous origin by Souporcell were retained for downstream analyses.

MiP-seq

In situlibrary construction and high-plex imaging

To provide orthogonal spatial validation for the 14 annotated trophoblast sub-clusters, MiP-seq [56] was performed on E60 bovine placental cryosections (8–10 μm). Sections were fixed with 4% paraformaldehyde (PFA) and permeabilized using 2 mg/ml pepsin (Sigma) in 0.1 M HCl at 37 °C for 90 s to ensure optimal probe accessibility. Custom-designed padlock probes targeting 12 marker genes (e.g., PAG8, PAG6, and CSH2) were hybridized overnight at 37 °C in Secure-Seal chambers (Additional file 2: Table S5). To ensure high-fidelity detection, circularization was mediated by SplintR ligase (NEB) at 25 °C for 3 h. This enzyme specifically seals the padlock probe only when it is perfectly hybridized to the complementary target RNA, effectively eliminating non-specific background signals at the molecular level. Subsequently, Rolling Circle Amplification (RCA) was performed using Phi29 DNA polymerase (Vazyme) at 30 °C for 14 h to generate high-intensity rolling circle products (RCPs) for signal amplification. Subcellular-resolution images were captured using a Leica THUNDER Imager 3D Assay with a 25 × objective following fluorescent probe hybridization.

Spatial image processing and interactive single-cell mapping

The computational pipeline was designed to transform raw fluorescence signals into a single-cell-resolution spatial expression matrix. First, multi-round images were aligned using the Bigwarp plugin (ImageJ) via a thin-plate spline (TPS) elastic registration algorithm to correct for physical deviations. The mRNA signal spots were subsequently identified and decoded using RS-FISH and the deep learning-based U-FISH (U-Net) method to extract precise XY coordinates. For cell segmentation, the pre-trained Cellpose (v4.0) cyto3 model was employed to delineate cell boundaries by predicting pixel-level gradient vector fields, effectively resolving cell-dense regions. Signal spot assignment was conducted using the ImageFlow software integrated with a KD-tree (k-dimensional tree) data structure for radius-adaptive searching and probabilistic mapping. Finally, the integrated dataset, comprising multiplexed gene expression, cell segmentation masks, and tissue architecture, was visualized using the TissUUmaps (v3.1.1) framework. This interactive interface enabled precise spatial co-localization analysis of the 13 trophoblast sub-cluster markers within the complex placental microenvironment and facilitated the systematic exploration of large-scale spatial datasets.

Comparisons with published bovine placental single-cell transcriptomic atlases and cross-breed datasets

To assess the reliability of our cell type annotation and cross-dataset integration, we compared our dataset with previously published bovine placental single-cell transcriptomic atlases [3, 23]. To evaluate the transcriptomic consistency across breeds at the uniform developmental timepoint, we generated snRNA-seq data from Holstein and crossbred (Bos taurus x Bos indicus) placentas on E50 (Additional file 2: Table S1-S2) and compared them with a published Angus placental dataset from the same developmental timepoint [23]. For each major annotated cell population, average expression profiles were calculated, and the Spearman correlation analysis was used to evaluate the cross-dataset cell type similarity. Representative markers reported in the published studies were further visualized in both the reference datasets and our dataset using dot plots to assess the consistency of cell type-/subpopulation-specific expression patterns. In addition, the top 100 cluster markers from the published atlases were compared with those identified in our study, and the significance of marker overlap was evaluated using a one-sided Fisher’s exact test for overrepresentation. For the matched-stage cross-breed comparisons, UMAP distribution, major cell type composition, and the marker expression consistency of trophoblast subpopulations were further examined to systematically assess the transcriptomic correspondence across data sources.

Differentially expressed gene (DEG), GO, and GSEA analyses

We used the “FindAllMarkers” function in the Seurat R package to identify DEGs in each cell cluster or at each developmental timepoint. Based on adjusted P values and log₂ fold changes, the top-ranked upregulated genes in each cluster were selected and visualized using heatmaps or dot plots.

For developmental stage-wise differential expression analyses, we explored two complementary strategies. First, Seurat FindAllMarkers with the Wilcoxon rank-sum test was used to capture single-cell-level expression differences, including changes in expression magnitude and the proportion of cells expressing the genes. Second, following recent recommendations favoring the replicate-aware differential expression analysis in single-cell transcriptomic studies, we applied a pseudobulk + limma-voom framework to account for biological replications [66]. Briefly, cells from the same biological replicate within each timepoint were aggregated into pseudobulk count matrices, followed by differential expression testing using the limma-voom pipeline. DEGs between consecutive developmental timepoints were identified based on adjusted P values and log₂ fold changes. To obtain a more robust set of stage-specific DEGs, DEGs identified by both methods were prioritized as the main candidates for downstream interpretation and presentation in the main text.

These cluster-specific DEGs were then subjected to GO and KEGG pathway enrichment analyses using the “compareCluster” function in the clusterProfiler R package (v 4.6.2) [97]. Representative enriched terms were visualized using the “dotplot” function. For GSEA, the human Hallmark gene sets were downloaded from the MSigDB database. Only gene sets containing 1:1 orthologous genes between humans and cows were retained. Further filtering was adopted to keep gene sets harboring more than ten genes. The enrichment analysis was performed with the following parameters: nPermSimple = 10,000, pvalueCutoff = 0.05, minGSSize = 20, maxGSSize = 1000, and pAdjustMethod = "BH". Visualization of enrichment results was carried out using the enrichplot R package (v1.18.4).

Analysis of stage-specific cell type distribution

To systematically evaluate the enrichment or depletion patterns of different cell types across gestational timepoints, we developed a customized metric (SPI), inspired by previously published methods [61]. This index quantifies the temporal distribution bias of each cell type by calculating the ratio of observed to expected cell counts (R_o/e) across different gestational timepoints. Specifically, we first constructed a contingency table of cell types by gestational timepoints and adopted a chi-squared test to assess whether the observed distribution significantly deviates from random expectation. Then, we calculated the SPI for each cell type-stage combination using the following formula:

{SPI}_{Stage}^{cell} = \frac{O b s e r v e d}{E x p e c t e d} = R_{o / e}

Here, R_o/e represents the ratio of the observed number of cells to the expected number for a given cell type at a given timepoint, with the expected counts derived from the chi-squared model. Unlike the chi-squared statistic that only indicates the extent of deviation from randomness, the SPI also captures the direction of the deviation: R_o/e > 1: strong enrichment (higher than expected; strong preference); 0.8 < R_o/e ≤ 1: mild enrichment (weak preference); 0.2 ≤ R_o/e ≤ 0.8: neutral or mild depletion (low preference); 0 < R_o/e < 0.2: rare presence (minimal preference); R_o/e = 0: complete absence (strong exclusion). This index enables robust quantification of the temporal distribution preference of cell types during placental development and provides insights into their potential stage-specific functions.

Analysis of TF regulation

TF regulatory activity was inferred using the command-line workflow of pySCENIC (v0.12.1), as previously described [62]. Since pySCENIC by default supports only a limited number of species, we customized the cisTarget database for the species involved in this study. Specifically, we downloaded the reference genome FASTA file and corresponding annotation files from Ensembl, extracted the upstream 1 kb regulatory sequence for each gene (named by gene symbols), and used the create_cistarget_motif_databases (https://github.com/aertslab/create_cisTarget_databases) script to generate the three core cisTarget database files: motifs_vs_regions.scores.feather, regions_vs_motifs.rankings.feather, and regions_vs_motifs.scores.feather. The list of TFs was obtained from the AnimalTFDB4 database. To compare the cell type specificity of TF regulons, we applied the Regulon Specificity Score (RSS) as previously defined [62]. Based on this, TF modules in bovine trophoblast cells were identified using the previously reported method [98]. Briefly, it contains four steps: (1) Regulon activity inference: SCENIC was used to infer the activity of each TF in individual cells using AUCell scores; (2) Correlation analysis: Pearson correlation coefficients (PCCs) were computed between the activity scores of all TF pairs; (3) CSI calculation: For a given TF pair (a and b), CSI was defined as the proportion of TFs connected to either a or b whose PCCs with a or b were lower than the PCC between a and b; (4) Module identification and network construction: A CSI matrix was used for hierarchical clustering (Ward’s method) to identify TF modules. A co-regulatory TF network was then constructed using TF pairs with CSI > 0.7. TF-TF interactions between modules were visualized using Cytoscape, displaying connections only with CSI ≥ 0.8. Finally, AUCell was used to assess the activity of each TF module across different gestational timepoints. For cross-species transfer of bovine trophoblast regulons, target genes were converted to one-to-one orthologs in the destination species. The low-coverage regulons were filtered (≥ ten mapped targets), and AUC scores were computed and summarized by timepoints. To identify key TFs regulating the IFNT gene during maternal–fetal recognition, we extracted the regulatory target gene sets of the top 30 stage-specific activated TFs from the adjacent file generated by the pySCENIC analysis. Candidate TFs potentially regulating IFNT were identified based on the presence of IFNT within these target sets. Subsequently, we ranked the regulatory importance of these candidate TFs using the GENIE3 (v1.20.0) algorithm to determine the most likely upstream regulators of IFNT.

RNA velocity and pseudotime analyses

Read annotation of the sequencing samples was performed using the command-line tool velocyto run10x, with input files including BAM, genome annotation, and repeat annotation files [99]. The BAM files were generated using default parameters of Cell Ranger (10 × Genomics). Transcripts were classified into three categories, i.e., “spliced”, “unspliced”, and “ambiguous”, based on the bovine genome annotation. Repeat annotation files were downloaded from the UCSC Genome Browser. The RNA velocity analysis was carried out using the Python package scVelo v0.2.2 [100], based on the UMAP embedding coordinates obtained from the Seurat pipeline. Specifically, after completion of the Seurat analysis, loom files containing the three transcript categories were imported into the R environment using the ReadVelocity function from the SeuratWrappers R package (v0.3.0), and spliced and unspliced data were added to the Seurat object. The object was then converted to the H5ad format using the SaveH5Seurat and Convert functions from the SeuratDisk R package (v0.0.0.9013) [101], and subsequently loaded into Python using scv.read from scVelo. After filtering and normalization of the Scanpy object, RNA velocities were computed using scv.pp.moments and scv.tl.velocity, and the velocity vectors were embedded into the UMAP plot generated in Seurat. The final RNA velocity stream plot was produced with the scv.pl.velocity_embedding_stream function. The developmental pseudotime analysis was performed using the Monocle3 R package (v1.3.4) [102]. After completion of the Seurat pipeline, the Seurat object was converted into a Monocle3 object using the as.cell_data_set function from the SeuratWrappers package. The developmental trajectory graph was constructed using the learn_graph function, and the root node was selected based on the RNA velocity direction inferred from velocyto. The Pseudotime value was calculated using the order_cells function, and visualized with plot_cells. To assess transcriptomic relationships among E24 trophoblast subclusters, we computed cluster-level average expression profiles and constructed a hierarchical clustering tree (Ward’s D2 linkage on Euclidean distances of averaged log-normalized expression) that was visualized in a phylogeny-like format using ggtree (v3.14.0).

Developmental trajectory and regulatory characterization of BNCs emerging on E24

To assess whether BNC1 and BNC2 on E24 represent two persistent BNC-associated states, representative markers were selected based on their differential expression and compared across the corresponding trophoblast populations on E24, E50, E180, and E240 using dot plots. To evaluate their relations to different anatomical regions, a published E195 placental single-cell transcriptomic dataset [3] was incorporated into our E24 trophoblast cell dataset using Seurat CCA. Average expression profiles from the integrated data were then used to construct a hierarchical clustering tree with Ward’s D2 linkage. In addition, based on the highly active regulons inferred by pySCENIC from all E24 trophoblast cells, regulon activity was calculated for BNC1, BNC2, UNC1, and UNC2, and their regulatory similarity was compared by heatmap clustering.

Identification of gene expression and metabolic activity modules

To identify trophoblast-specific gene expression modules across different developmental timepoints, we first extracted the raw UMI count matrix of trophoblast cells at all timepoints from the Seurat object. Genes with non-zero expression across all cells were retained for downstream analyses. Cells were then grouped by developmental timepoints, and the total UMI counts for each gene at each timepoint were calculated to generate a stage-level pseudobulk expression matrix. We applied soft clustering to the pseudobulk matrix using the clusterData function (with cluster.method = "mfuzz") from the ClusterGVis package (v0.1.4). The clustering results were visualized with the visCluster function, and heatmaps were produced to display the dynamic expression patterns of gene modules across developmental timepoints.

To explore the metabolic activity trajectories, we retrieved bovine metabolic pathway information using the KEGGREST package (v1.38.0). The AUCell package (v1.20.2) was used to assess the enrichment of each metabolic pathway in individual cells, generating a cell × pathway activity matrix. These activity scores were then aggregated by developmental timepoints to construct a pseudo metabolic activity matrix. Finally, the identical soft clustering approach was employed to identify and visualize metabolic activity modules, revealing stage-specific metabolic dynamics.

Integration of GWAS and snRNA-seq data to identify trait-associated cell types

In this study, we harnessed scPagwas [103] (v1.3.1) to integrate GWAS summary statistics of pregnancy-related traits with placental single-cell transcriptomic data. scPagwas leverages the cumulative effect of biological pathways to combine single-cell gene expression with genome-wide association signals, enabling the assessment of the relevance between the cell and the target trait at single-cell resolution, and therefore the identification of key trait-associated genes and pathways. Specifically, for cows, we prepared GWAS summary statistics according to the software documentation and used the reference panel from the 1,000 Bull Genomes Project to estimate linkage disequilibrium (LD) between SNPs. KEGG pathway information was retrieved using the R package KEGGREST. For human data, the built-in preprocessing pipeline of scPagwas was directly used.

Expression patterns of functionally constrained and lethality-associated genes

We obtained pLI scores for human genes from reference [104] and retrieved 1:1 orthologous gene sets across humans, cows, mice, and macacas using Ensembl BioMart. To enable cross-species comparisons, we only retained those 1:1 orthologs that were expressed (UMI ≥ 1) in all three species. For each nucleus, we computed the median pLI score of the expressed orthologous genes. To assess the expression patterns of genes essential for organismal survival, we used a neutrally ascertained gene knockout dataset consisting of 4,742 protein-coding genes, of which 1,139 were classified as lethality-associated. After intersection with the shared 1:1 orthologs expressed across species, we calculated the proportion of expressed genes known to produce a lethal phenotype in each cell. The denominator was the number of expressed genes with available lethality data, with the numerator being the subset number of those classified as lethality-associated. Tested genes for viability and associated phenotype information were downloaded from the International Mouse Phenotyping Consortium [80].

Macrophage polarization activity fraction

We curated markers for M1 and M2 macrophage polarization from the literature (see Additional file 7: Table S4). For each macrophage cell across gestational timepoints, signature activity was quantified with AUCell that computes the area under the recovery curve (AUC) of the ranked expression profile for a given gene set. Before scoring, gene symbols were intersected with the expression matrix, and only the markers present in the dataset were used. AUC was computed with aucMaxRank = 0.05 × nOfRankedGenes (default, 5%). A cell was considered “active” for a signature if its AUC exceeded a data-driven threshold returned by AUCell_exploreThresholds (per signature, global across cells). The activity fraction at each gestational timepoint was defined as the proportion of macrophages classified as active for the corresponding signature among all macrophages at that timepoint. For robustness, results were confirmed using a fixed percentile threshold (90th percentile of AUC) that showed the uniform qualitative trends.

Cell–cell communication analysis

We analyzed intercellular communication using CellPhoneDB v3.1.0 [19]. Following the recommended workflow, the raw count matrix and cell type annotation were exported from the Seurat object using provided scripts. These files were then used as input for the statistical analysis module with default parameters, and executed via the Linux command-line interface. To account for species differences, bovine genes were mapped to their human orthologs as outright as possible, enabling compatibility with the CellPhoneDB human-based receptor-ligand database. Significant cell–cell interactions between key cell types were visualized using the “dot_plot” function.

Supplementary Information

Additional file 1: Fig. S1-S17.^{(5.8MB, pdf)}

13059_2026_4173_MOESM2_ESM.xlsx^{(637.2KB, xlsx)}

Additional file 2: Tables S1-S6. Supplementary information on bovine placental development. Description: This Excel file contains six supplementary tables. Tables S1 and S2 provide detailed dataset information and quality metrics. Table S3 summarizes integration evaluation results. Table S4 lists DEGs among trophoblast subtypes. Table S5 presents probe design targets for spatial transcriptomics. Table S6 shows the top 50 markers shared between bovine trophoblasts and their human, mouse, or macaca counterparts.

13059_2026_4173_MOESM3_ESM.xlsx^{(25.4KB, xlsx)}

Additional file 3: Tables S1-S3. TF regulon activity and module organization in bovine trophoblasts. Description: This Excel file contains three supplementary tables. Table S1 lists highly active TF regulons in bovine trophoblasts across gestation. Table S2 presents GO enrichment results for TF regulatory modules composed of these regulons. Table S3 specifies which module each regulon belongs to.

13059_2026_4173_MOESM4_ESM.xlsx^{(6.9MB, xlsx)}

Additional file 4: DEGs of trophoblasts across gestation.

13059_2026_4173_MOESM5_ESM.xlsx^{(543.4KB, xlsx)}

Additional file 5: Clustering of gene expression and metabolic pathways in trophoblasts across gestational stages.

13059_2026_4173_MOESM6_ESM.xlsx^{(1.2MB, xlsx)}

Additional file 6: Tables S1-S5. GWAS and cross‑species constraint analyses for cow GL. Description: Table S1: Cell type association P values. Table S2: Gestational stage enrichment P values for bovine GL GWAS. Table S3: Pearson correlation coefficientsfor genes correlated with GL. Table S4: One‑to‑one mapping of human pLI scores to cows, mice, and macacas. Table S5: One‑to‑one mapping of mouse lethal genes to humans, cows, and macacas.

13059_2026_4173_MOESM7_ESM.xlsx^{(2.6MB, xlsx)}

Additional file 7: Tables S1-S12. Hofbauer celldynamics and RP associations in cows and humans. Description: This Excel file contains twelve supplementary tables. Tables S1-S3: Bovine HBC stage markers and their KEGG/GO enrichment results. Table S4: M1/M2 macrophage signatures. Tables S5-S6: Bovine RP cell type enrichment P values and gene PCC. Table S7: The top 50 KEGG pathways per bovine cell type and trait. Tables S8-S9: Human RP cell type P values and gene PCC. Tables S10-S11: The top 50 KEGG pathways for cow and human macrophages/monocytes with RP. Table S12: Human RP GWAS summary data sources.

Acknowledgements

We thank the High-Performance Computing Platform of Northwest A&F University and the Computing Center in Xi'an for providing computing resources.

Peer review information

Wenjing She was the primary editor of this article and managed its editorial process and peer review in collaboration with the rest of the editorial team. The peer-review history is available in the online version of this article.

Authors’ contributions

Conceptualization, G.T., Y.Zheng, and Y.J.; Methodology, G.T., X.Y., A.Z., X.C., and T.S.; Investigation, G.T. (major), X.Y. (major), A.Z. (major), X.C. (major), T.S., Y.Zhou, X.G., H.W., H.L., Y.L., and X.W.; Writing—Original Draft, G.T., X.Y., and Y.Zheng; Writing—review & editing, all authors; Funding acquisition, Y.Zheng and Y.J.; Resources, Y.J. and Y.Zheng; Supervision, Y.Zheng, Y.J., and T.S.

Funding

This work was supported by the National Key R&D Program of China (Grant No. 2022YFF1000100 and 2023YFD1300402), and the Science and Technology Development Program of Shaanxi Province (Grant No. 2025SYS-SZSYS-23).

Data availability

The snRNA-seq data generated in this study have been deposited at NCBI under accession BioProject PRJNA1302120 [105], and the processed data are available at GEO under accession GSE306749 [106]. The bovine placental scRNA-seq datasets used in this study included those from gestational stages E12-E18 (GSE234335) [107] and E24-E50 (GSE234524) [108]. The human placental scRNA-seq datasets used in this study were obtained from the following resources: first trimester data from ArrayExpress (https://www.ebi.ac.uk/biostudies/arrayexpress; accession E-MTAB-6701) [109], second trimester data from GEO (GSE198373) [110], and third trimester data from the European Genome-phenome Archive (EGAS00001002449) [111]. The mouse placental scRNA-seq data were obtained from GEO (GSE156125) [112], and the macaca placental datasets were also retrieved from GEO (accessions GSE180637 [113] and GSE193007 [114]). The GWAS summary statistics for bovine GL and RP, as well as the raw and processed MiP-seq data for E60 bovine placental sections, are available at Figshare (https://doi.org/10.6084/m9.figshare.31937505) [115]. For human RP, we retrieved the summary data from the GWAS Catalog under accession GCST90044489 [116]. No custom code was developed in this study. All analyses were performed using publicly available software and packages as described in the Methods section. All key parameters and command lines are provided in the main text.

Declarations

Ethics approval and consent to participate

All experiments were conducted using sexually mature, cycling Holstein heifers (Bos taurus), with approval from the Animal Care and Use Committee of Northwest A&F University (Approval No. XN2024-0416).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Guanghui Tan, Xiaoru Yan, Xuesha Cao and Ao Zhang contributed equally to this work.

Contributor Information

Yu Jiang, Email: yu.jiang@nwafu.edu.cn.

Yi Zheng, Email: y.zheng@nwafu.edu.cn.

References

1.Wiltbank MC, Baez GM, Garcia-Guerra A, Toledo MZ, Monteiro PL, Melo LF, et al. Pivotal periods for pregnancy loss during the first trimester of gestation in lactating dairy cows. Theriogenology. 2016;86:239–53. [DOI] [PubMed] [Google Scholar]
2.Sánchez JM, Mathew DJ, Passaro C, Fair T, Lonergan P. Embryonic maternal interaction in cattle and its relationship with fertility. Reprod Domest Anim. 2018;53(Suppl 2):20–7. [DOI] [PubMed] [Google Scholar]
3.Davenport KM, Ortega MS, Liu H, O’Neil EV, Kelleher AM, Warren WC, et al. Single-nuclei RNA sequencing (snRNA-seq) uncovers trophoblast cell types and lineages in the mature bovine placenta. Proc Natl Acad Sci U S A. 2023;120:e2221526120. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Diskin MG, Morris DG. Embryonic and early foetal losses in cattle and other ruminants. Reprod Domest Anim. 2008;43(Suppl 2):260–7. [DOI] [PubMed] [Google Scholar]
5.Albaaj A, Durocher J, LeBlanc SJ, Dufour S. Meta-analysis of the incidence of pregnancy losses in dairy cows at different stages to 90 days of gestation. JDS Commun. 2023;4:144–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Green JA, Geisert RD, Johnson GA, Spencer TE. Implantation and placentation in ruminants. Adv Anat Embryol Cell Biol. 2021;234:129–54. [DOI] [PubMed] [Google Scholar]
7.Moraes JGN, Behura SK, Geary TW, Hansen PJ, Neibergs HL, Spencer TE. Uterine influences on conceptus development in fertility-classified animals. Proc Natl Acad Sci U S A. 2018;115:E1749-e1758. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Eley RM, Thatcher WW, Bazer FW, Wilcox CJ, Becker RB, Head HH, et al. Development of the conceptus in the bovine. J Dairy Sci. 1978;61:467–73. [DOI] [PubMed] [Google Scholar]
9.Davenport KM, Ortega MS, Johnson GA, Seo H, Spencer TE. Review: implantation and placentation in ruminants. Animal. 2023;17(Suppl 1):100796. [DOI] [PubMed] [Google Scholar]
10.Wooding FBP. The ruminant placental trophoblast binucleate cell: an evolutionary breakthrough. Biol Reprod. 2022;107:705–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Davies Morel MC, Newcombe JR, Holland SJ. Factors affecting gestation length in the Thoroughbred mare. Anim Reprod Sci. 2002;74:175–85. [DOI] [PubMed] [Google Scholar]
12.Clausson B, Lichtenstein P, Cnattingius S. Genetic influence on birthweight and gestational length determined by studies in offspring of twins. BJOG. 2000;107:375–81. [DOI] [PubMed] [Google Scholar]
13.Fang L, Jiang J, Li B, Zhou Y, Freebern E, Vanraden PM, et al. Genetic and epigenetic architecture of paternal origin contribute to gestation length in cattle. Commun Biol. 2019;2:100. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Moradi M, Zhandi M, Sharafi M, Akbari A, Atrabi MJ, Totonchi M. Gene expression profile of placentomes and clinical parameters in the cows with retained placenta. BMC Genomics. 2022;23:760. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Weeks AD. The retained placenta. Best Pract Res Clin Obstet Gynaecol. 2008;22:1103–17. [DOI] [PubMed] [Google Scholar]
16.Rojas de Oliveira H, Chud TCS, Oliveira GA Jr., Hermisdorff IC, Narayana SG, Rochus CM, et al. Genome-wide association analyses reveal copy number variant regions associated with reproduction and disease traits in Canadian Holstein cattle. J Dairy Sci. 2024;107:7052–63. [DOI] [PubMed]
17.Arutyunyan A, Roberts K, Troulé K, Wong FCK, Sheridan MA, Kats I, et al. Spatial multiomics map of trophoblast development in early pregnancy. Nature. 2023;616:143–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Pavličev M, Wagner GP, Chavan AR, Owens K, Maziarz J, Dunn-Fletcher C, et al. Single-cell transcriptomics of the human placenta: inferring the cell communication network of the maternal-fetal interface. Genome Res. 2017;27:349–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Suryawanshi H, Morozov P, Straus A, Sahasrabudhe N, Max KEA, Garzia A, et al. A single-cell survey of the human first-trimester placenta and decidua. Sci Adv. 2018;4:eaau4788. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Marsh B, Blelloch R. Single nuclei RNA-seq of mouse placental labyrinth development. Elife. 2020. 10.7554/eLife.60266. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Jiang X, Wang Y, Xiao Z, Yan L, Guo S, Wang Y, et al. A differentiation roadmap of murine placentation at single-cell resolution. Cell Discov. 2023;9:30. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Scatolin GN, Ming H, Wang Y, Iyyappan R, Gutierrez-Castillo E, Zhu L, et al. Single-cell transcriptional landscapes of bovine peri-implantation development. iScience. 2024. 10.1016/j.isci.2024.109605. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Davenport KM, O’Neil EV, Ortega MS, Patterson A, Kelleher AM, Warren WC, et al. Single-cell insights into development of the bovine placenta†. Biol Reprod. 2024;110:169–84. [DOI] [PubMed] [Google Scholar]
24.Hradecký P, Mossman HW, Stott GG. Comparative development of ruminant placentomes. Theriogenology. 1988;29:715–29. [DOI] [PubMed] [Google Scholar]
25.Bauersachs S, Ulbrich SE, Gross K, Schmidt SEM, Meyer HHD, Wenigerkind H, et al. Embryo-induced transcriptome changes in bovine endometrium reveal species-specific and common molecular markers of uterine receptivity. Reproduction. 2006;132:319–31. [DOI] [PubMed] [Google Scholar]
26.Bairagi S, Quinn KE, Crane AR, Ashley RL, Borowicz PP, Caton JS, et al. Maternal environment and placental vascularization in small ruminants. Theriogenology. 2016;86:288–305. [DOI] [PubMed] [Google Scholar]
27.Gayoso A, Lopez R, Xing G, Boyeau P, Valiollah Pour Amiri V, Hong J, et al. A Python library for probabilistic analysis of single-cell omics data. Nat Biotechnol. 2022;40:163–6. [DOI] [PubMed] [Google Scholar]
28.Hue I. Determinant molecular markers for peri-gastrulating bovine embryo development. Reprod Fertil Dev. 2016;28:51–65. [DOI] [PubMed] [Google Scholar]
29.Jia G-X, Ma W-J, Wu Z-B, Li S, Zhang X-Q, He Z, et al. Single-cell transcriptomic characterization of sheep conceptus elongation and implantation. Cell Rep. 2023. 10.1016/j.celrep.2023.112860. [DOI] [PubMed] [Google Scholar]
30.Liu T, Li J, Yu L, Sun HX, Li J, Dong G, et al. Cross-species single-cell transcriptomic analysis reveals pre-gastrulation developmental differences among pigs, monkeys, and humans. Cell Discov. 2021;7:8. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Steinhauser CB, Landers M, Myatt L, Burghardt RC, Vallet JL, Bazer FW, et al. Fructose synthesis and transport at the uterine-placental interface of pigs: cell-specific localization of SLC2A5, SLC2A8, and components of the polyol pathway. Biol Reprod. 2016;95:108. [DOI] [PubMed] [Google Scholar]
32.Degrelle SA, Murthi P, Evain-Brion D, Fournier T, Hue I. Expression and localization of DLX3, PPARG and SP1 in bovine trophoblast during binucleated cell differentiation. Placenta. 2011;32:917–20. [DOI] [PubMed] [Google Scholar]
33.Yahyazadeh Mashhadi SM, Kazemimanesh M, Arashkia A, Azadmanesh K, Meshkat Z, Golichenari B, et al. Shedding light on the EpCAM: an overview. J Cell Physiol. 2019;234:12569–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Kuriyama S, Tamiya Y, Tanaka M. Spatiotemporal expression of UPK3B and its promoter activity during embryogenesis and spermatogenesis. Histochem Cell Biol. 2017;147:17–26. [DOI] [PubMed] [Google Scholar]
35.Deng FM, Liang FX, Tu L, Resing KA, Hu P, Supino M, et al. Uroplakin IIIb, a urothelial differentiation marker, dimerizes with uroplakin Ib as an early step of urothelial plaque assembly. J Cell Biol. 2002;159:685–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Shin JH, Son EJ, Lee HS, Kim SJ, Kim K, Choi JY, et al. Molecular and functional expression of anion exchangers in cultured normal human nasal epithelial cells. Acta Physiol (Oxf). 2007;191:99–110. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Chen W, Kang KL, Alshaikh A, Varma S, Lin YL, Shin KH, et al. Grainyhead-like 2 (GRHL2) knockout abolishes oral cancer development through reciprocal regulation of the MAP kinase and TGF-β signaling pathways. Oncogenesis. 2018;7:38. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.MacFawn I, Wilson H, Selth LA, Leighton I, Serebriiskii I, Bleackley RC, et al. Grainyhead-like-2 confers NK-sensitivity through interactions with epigenetic modifiers. Mol Immunol. 2019;105:137–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Di Palma T, Filippone MG, Pierantoni GM, Fusco A, Soddu S, Zannini M. Pax8 has a critical role in epithelial cell survival and proliferation. Cell Death Dis. 2013;4:e729. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Fu DJ, De Micheli AJ, Bidarimath M, Ellenson LH, Cosgrove BD, Flesken-Nikitin A, et al. Cells expressing PAX8 are the main source of homeostatic regeneration of adult mouse endometrial epithelium and give rise to serous endometrial carcinoma. Dis Model Mech. 2020. 10.1242/dmm.047035. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Wang F, Ding P, Liang X, Ding X, Brandt CB, Sjöstedt E, et al. Endothelial cell heterogeneity and microglia regulons revealed by a pig cell landscape at single-cell level. Nat Commun. 2022;13:3620. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Jiang X, Zhai J, Xiao Z, Wu X, Zhang D, Wan H, et al. Identifying a dynamic transcriptomic landscape of the cynomolgus macaque placenta during pregnancy at single-cell resolution. Dev Cell. 2023;58:806-821.e807. [DOI] [PubMed] [Google Scholar]
43.Wu JJ, Zhu S, Gu F, Valencak TG, Liu JX, Sun HZ. Cross-tissue single-cell transcriptomic landscape reveals the key cell subtypes and their potential roles in the nutrient absorption and metabolism in dairy cattle. J Adv Res. 2022;37:1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Ziegler-Heitbrock L. The CD14+ CD16+ blood monocytes: their role in infection and inflammation. J Leukoc Biol. 2007;81:584–92. [DOI] [PubMed] [Google Scholar]
45.Müller I, Vogl T, Kühl U, Krannich A, Banks A, Trippel T, et al. Serum alarmin S100A8/S100A9 levels and its potential role as biomarker in myocarditis. ESC Heart Fail. 2020;7:1442–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Seong J, Frias-Aldeguer J, Holzmann V, Kagawa H, Sestini G, Heidari Khoei H, et al. Epiblast inducers capture mouse trophectoderm stem cells in vitro and pattern blastoids for implantation in utero. Cell Stem Cell. 2022;29:1102-1118.e1108. [DOI] [PubMed] [Google Scholar]
47.Pfeffer PL. Alternative mammalian strategies leading towards gastrulation: losing polar trophoblast (Rauber’s layer) or gaining an epiblast cavity. Philos Trans R Soc Lond B Biol Sci. 2022;377:20210254. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Wang Y, Ming H, Yu L, Li J, Zhu L, Sun H-X, et al. Establishment of bovine trophoblast stem cells. Cell Rep. 2023. 10.1016/j.celrep.2023.112439. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Zhao G, Jiang K, Zhang T, Wu H, Qiu C, Deng G. Specific interferon tau gene-regulation networks in bovine endometrial luminal epithelial cells. Theriogenology. 2018;105:51–60. [DOI] [PubMed] [Google Scholar]
50.Strieder-Barboza C, de Souza J, Raphael W, Lock AL, Contreras GA. Fetuin-A: a negative acute-phase protein linked to adipose tissue function in periparturient dairy cows. J Dairy Sci. 2018;101:2602–16. [DOI] [PubMed] [Google Scholar]
51.Wu J-J, Zheng E, Liu L, Quan J, Ruan D, Yao Z, et al. Cell-cell communication-mediated cell-type-specific parent-of-origin effects in mammals. Nat Commun. 2025;16:5106. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Khatri P, Frenette G, Sullivan R, Hoffmann B, Schuler G. Expression of SULT1E1 protein in bovine placentomes: evidence for localization in uninucleated trophoblast cells. Placenta. 2011;32:431–40. [DOI] [PubMed] [Google Scholar]
53.Simner C, Novakovic B, Lillycrop KA, Bell CG, Harvey NC, Cooper C, et al. DNA methylation of amino acid transporter genes in the human placenta. Placenta. 2017;60:64–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Hecker N, Sharma V, Hiller M. Convergent gene losses illuminate metabolic and physiological changes in herbivores and carnivores. Proc Natl Acad Sci U S A. 2019;116:3036–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Lee B-K, Jang Y, Kim M, LeBlanc L, Rhee C, Lee J, et al. Super-enhancer-guided mapping of regulatory networks controlling mouse trophoblast stem cells. Nat Commun. 2019;10:4749. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Wu X, Xu W, Deng L, Li Y, Wang Z, Sun L, et al. Spatial multi-omics at subcellular resolution via high-throughput in situ pairwise sequencing. Nat Biomed Eng. 2024;8:872–89. [DOI] [PubMed] [Google Scholar]
57.Vento-Tormo R, Efremova M, Botting RA, Turco MY, Vento-Tormo M, Meyer KB, et al. Single-cell reconstruction of the early maternal-fetal interface in humans. Nature. 2018;563:347–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Marsh B, Zhou Y, Kapidzic M, Fisher S, Blelloch R. Regionally distinct trophoblast regulate barrier function and invasion in the human placenta. Elife. 2022. 10.7554/eLife.78829. [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Tsang JCH, Vong JSL, Ji L, Poon LCY, Jiang P, Lui KO, et al. Integrative single-cell and cell-free plasma RNA transcriptomics elucidates placental cellular dynamics. Proc Natl Acad Sci U S A. 2017;114:E7786-e7795. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Spencer TE, Hansen TR. Implantation and establishment of pregnancy in ruminants. Adv Anat Embryol Cell Biol. 2015;216:105–35. [DOI] [PubMed] [Google Scholar]
61.Zhang L, Yu X, Zheng L, Zhang Y, Li Y, Fang Q, et al. Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. Nature. 2018;564:268–72. [DOI] [PubMed] [Google Scholar]
62.Van de Sande B, Flerin C, Davie K, De Waegeneer M, Hulselmans G, Aibar S, et al. A scalable SCENIC workflow for single-cell gene regulatory network analysis. Nat Protoc. 2020;15:2247–76. [DOI] [PubMed] [Google Scholar]
63.Fuxman Bass JI, Diallo A, Nelson J, Soto JM, Myers CL, Walhout AJ. Using networks to measure similarity between genes: association index selection. Nat Methods. 2013;10:1169–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Mi C, Ye B, Gao Z, Du J, Li R, Huang D. BHLHE40 plays a pathological role in pre-eclampsia through upregulating SNX16 by transcriptional inhibition of miR-196a-5p. Mol Hum Reprod. 2020;26:532–48. [DOI] [PubMed] [Google Scholar]
65.Lv B, An Q, Zeng Q, Zhang X, Lu P, Wang Y, et al. Single-cell RNA sequencing reveals regulatory mechanism for trophoblast cell-fate divergence in human peri-implantation conceptuses. PLoS Biol. 2019;17:e3000187. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Sharifi O, Haghani V, Neier KE, Fraga KJ, Korf I, Hakam SM, et al. Sex-specific single cell-level transcriptomic signatures of Rett syndrome disease progression. Commun Biol. 2024;7:1292. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Fitzgerald JS, Poehlmann TG, Schleussner E, Markert UR. Trophoblast invasion: the role of intracellular cytokine signalling via signal transducer and activator of transcription 3 (STAT3). Hum Reprod Update. 2008;14:335–44. [DOI] [PubMed] [Google Scholar]
68.Mor G, Cardenas I. The immune system in pregnancy: a unique complexity. Am J Reprod Immunol. 2010;63:425–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Ezashi T, Imakawa K. Transcriptional control of IFNT expression. Reproduction. 2017;154:F21-f31. [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Chakrabarty A, Roberts RM. Ets-2 and C/EBP-beta are important mediators of ovine trophoblast Kunitz domain protein-1 gene expression in trophoblast. BMC Mol Biol. 2007;8:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Ma B, Cui H, Wang X, Feng W, Zhang J, Chen N, et al. IFNT-induced IRF1 enhances bovine endometrial receptivity by transactivating LIFR. J Reprod Immunol. 2024;163:104212. [DOI] [PubMed] [Google Scholar]
72.Basavaraja R, Przygrodzka E, Pawlinski B, Gajewski Z, Kaczmarek MM, Meidan R. Interferon-tau promotes luteal endothelial cell survival and inhibits specific luteolytic genes in bovine corpus luteum. Reproduction. 2017;154:559–68. [DOI] [PubMed] [Google Scholar]
73.Cornelis G, Heidmann O, Degrelle SA, Vernochet C, Lavialle C, Letzelter C, et al. Captured retroviral envelope syncytin gene associated with the unique placental structure of higher ruminants. Proc Natl Acad Sci U S A. 2013;110:E828-837. [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Klisch K, Pfarrer C, Schuler G, Hoffmann B, Leiser R. Tripolar acytokinetic mitosis and formation of feto-maternal syncytia in the bovine placentome: different modes of the generation of multinuclear cells. Anat Embryol (Berl). 1999;200:229–37. [DOI] [PubMed] [Google Scholar]
75.Chen HZ, Ouseph MM, Li J, Pécot T, Chokshi V, Kent L, et al. Canonical and atypical E2Fs regulate the mammalian endocycle. Nat Cell Biol. 2012;14:1192–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Hughes M, Dobric N, Scott IC, Su L, Starovic M, St-Pierre B, et al. The Hand1, Stra13 and Gcm1 transcription factors override FGF signaling to promote terminal differentiation of trophoblast stem cells. Dev Biol. 2004;271:26–37. [DOI] [PubMed] [Google Scholar]
77.Jiang Y, Chen Y, Chen Y. Knockdown of JARID2 inhibits the viability and migration of placenta trophoblast cells in preeclampsia. Mol Med Rep. 2017;16:3594–9. [DOI] [PubMed] [Google Scholar]
78.Kim M, Adu-Gyamfi EA, Kim J, Lee BK. Super-enhancer-associated transcription factors collaboratively regulate trophoblast-active gene expression programs in human trophoblast stem cells. Nucleic Acids Res. 2023;51:3806–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Dickinson ME, Flenniken AM, Ji X, Teboul L, Wong MD, White JK, et al. High-throughput discovery of novel developmental phenotypes. Nature. 2016;537:508–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Koscielny G, Yaikhom G, Iyer V, Meehan TF, Morgan H, Atienza-Herrero J, et al. The International Mouse Phenotyping Consortium web portal, a unified point of access for knockout mice and related phenotyping data. Nucleic Acids Res. 2014;42:D802-809. [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Zhang YH, He M, Wang Y, Liao AH. Modulators of the balance between M1 and M2 macrophages during pregnancy. Front Immunol. 2017;8:120. [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Tan G-H, Liu S-J, Dou M-L, Zhao D-F, Zhang A, Li H-K, et al. Spatially resolved transcriptomic profiling of placental development in dairy cow. Zool Res. 2024;45:586–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Brooks K, Burns G, Spencer TE. Conceptus elongation in ruminants: roles of progesterone, prostaglandin, interferon tau and cortisol. J Anim Sci Biotechnol. 2014;5:53. [DOI] [PMC free article] [PubMed] [Google Scholar]
84.Courtney JA, Wilson RL, Cnota J, Jones HN. Conditional mutation of Hand1 in the mouse placenta disrupts placental vascular development resulting in fetal loss in both early and late pregnancy. Int J Mol Sci. 2021. 10.3390/ijms22179532. [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Fresch R, Courtney J, Brockway H, Wilson RL, Jones H. HAND1 knockdown disrupts trophoblast global gene expression. Physiol Rep. 2023;11:e15553. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.Zadora J, Singh M, Herse F, Przybyl L, Haase N, Golic M, et al. Disturbed placental imprinting in preeclampsia leads to altered expression of DLX5, a human-specific early trophoblast marker. Circulation. 2017;136:1824–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
87.Novakovic B, Fournier T, Harris LK, James J, Roberts CT, Yong HEJ, et al. Increased methylation and decreased expression of homeobox genes TLX1, HOXA10 and DLX5 in human placenta are associated with trophoblast differentiation. Sci Rep. 2017;7:4523. [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Shoshan-Barmatz V, Shteinfer-Kuzmine A, Verma A. VDAC1 at the intersection of cell metabolism, apoptosis, and diseases. Biomolecules. 2020. 10.3390/biom10111485. [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Liang YD, Zhao K, Chen Y, Zhang SQ. Role of CYCS in the cytogenesis and apoptosis of male germ cells and its clinical application. Zhonghua Nan Ke Xue. 2020;26:265–70. [PubMed] [Google Scholar]
90.Liao Y, Peng S, He L, Wang Y, Li Y, Ma D, et al. Methylmercury cytotoxicity and possible mechanisms in human trophoblastic HTR-8/SVneo cells. Ecotoxicol Environ Saf. 2021;207:111520. [DOI] [PubMed] [Google Scholar]
91.Shi M, Lv X, Zhu M, Dong Y, Hu L, Qian Y, et al. HMGA1 promotes hepatocellular carcinoma proliferation, migration, and regulates cell cycle via miR-195-5p. Anticancer Drugs. 2022;33:e273–85. [DOI] [PubMed] [Google Scholar]
92.Matsubara K, Matsubara Y, Uchikura Y, Takagi K, Yano A, Sugiyama T. HMGA1 is a potential driver of preeclampsia pathogenesis by interference with extravillous trophoblasts invasion. Biomolecules. 2021. 10.3390/biom11060822. [DOI] [PMC free article] [PubMed] [Google Scholar]
93.Ali A, Stenglein MD, Spencer TE, Bouma GJ, Anthony RV, Winger QA. Trophectoderm-specific knockdown of LIN28 decreases expression of genes necessary for cell proliferation and reduces elongation of sheep conceptus. Int J Mol Sci. 2020. 10.3390/ijms21072549. [DOI] [PMC free article] [PubMed] [Google Scholar]
94.McGinnis CS, Murrow LM, Gartner ZJ. DoubletFinder: doublet detection in single-cell RNA sequencing data using artificial nearest neighbors. Cell Syst. 2019;8:329-337.e324. [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Luecken MD, Büttner M, Chaichoompu K, Danese A, Interlandi M, Mueller MF, et al. Benchmarking atlas-level data integration in single-cell genomics. Nat Methods. 2022;19:41–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
96.Heaton H, Talman AM, Knights A, Imaz M, Gaffney DJ, Durbin R, et al. Souporcell: robust clustering of single-cell RNA-seq data by genotype without reference genotypes. Nat Methods. 2020;17:615–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Yu G, Wang LG, Han Y, He QY. ClusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
98.Suo S, Zhu Q, Saadatpour A, Fei L, Guo G, Yuan GC. Revealing the critical regulators of cell identity in the mouse cell atlas. Cell Rep. 2018;25:1436-1445.e1433. [DOI] [PMC free article] [PubMed] [Google Scholar]
99.La Manno G, Soldatov R, Zeisel A, Braun E, Hochgerner H, Petukhov V, et al. RNA velocity of single cells. Nature. 2018;560:494–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
100.Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ. Generalizing RNA velocity to transient cell states through dynamical modeling. Nat Biotechnol. 2020;38:1408–14. [DOI] [PubMed] [Google Scholar]
101.Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM 3rd, et al. Comprehensive integration of single-cell data. Cell. 2019;177:1888-1902.e1821. [DOI] [PMC free article] [PubMed]
102.Qiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, et al. Reversed graph embedding resolves complex single-cell trajectories. Nat Methods. 2017;14:979–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
103.Ma Y, Deng C, Zhou Y, Zhang Y, Qiu F, Jiang D, et al. Polygenic regression uncovers trait-relevant cellular contexts through pathway activation transformation of single-cell RNA sequencing data. Cell Genom. 2023;3:100383. [DOI] [PMC free article] [PubMed] [Google Scholar]
104.Lek M, Karczewski KJ, Minikel EV, Samocha KE, Banks E, Fennell T, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536:285–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
105.Tan GH, et al. Single-cell RNA-sequencing reveals the placental molecular dynamics during bovine gestation. NCBI SRA BioProject PRJNA1302120. 2026. https://www.ncbi.nlm.nih.gov/sra/PRJNA1302120.
106.Tan G, et al. A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs. Gene Expression Omnibus. 2026. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE306749. [DOI] [PMC free article] [PubMed]
107.Scatolin GN, Ming H, Wang YJ, Iyyappan R, Gutierrez-Castillo E, Zhu LK, Sagheer M, Song C, Bondioli K, Jiang ZL. Single-cell transcriptional landscapes of bovine peri-implantation development. Datasets. Gene Expression Omnibus. 2023.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234335. [DOI] [PMC free article] [PubMed]
108.Davenport KM, O'Neil EV, Ortega MS, Patterson A, Kelleher AM, Warren WC, Spencer TE. Single cell insights into development of the bovine placenta. Datasets. Gene Expression Omnibus. 2023.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234524. [DOI] [PubMed]
109.Vento-Tormo R, Efremova M, Botting RA, Turco MY, Vento-Tormo M, Meyer KB, Park JE, Stephenson E, Polanski K, Goncalves A, et al. Reconstructing the human first trimester fetal-maternal interface using single cell transcriptomics -10x data. Datasets. ArrayExpress. 2018. https://www.ebi.ac.uk/biostudies/arrayexpress/studies/E-MTAB-6701.
110.Marsh B, Zhou Y, Kapidzic M, Fisher S, Blelloch R. Regionally distinct trophoblast regulate barrier function and invasion in the human placenta. Datasets. Gene Expression Omnibus. 2022. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE198373. [DOI] [PMC free article] [PubMed]
111.Tsang JCH, Vong JSL, Ji L, Poon LCY, Jiang P, Lui KO, Ni Y, To KF, Cheng YKY, Chiu RWK, et al. Integrative single-cell and cell-free plasma RNA transcriptomics elucidates placental cellular dynamics. Datasets. European Genome-Phenome Archive. 2017. https://ega-archive.org/datasets/EGAD00001003705. [DOI] [PMC free article] [PubMed]
112.Jiang X, Wang Y, Xiao Z, Yan L, Guo S, Wang Y, Wu H, Zhao X, Lu X, Wang H. A differentiation roadmap of murine placentation at single-cell resolution. Datasets. Gene Expression Omnibus. 2023. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE156125. [DOI] [PMC free article] [PubMed]
113.Jiang X, Zhai J, Xiao Z, Wu X, Wang H, Wan H, Xu Y, Zhang D, et al. A single-cell transcriptome landscape of the non-human primate placenta across pregnancy. Datasets. Gene Expression Omnibus. 2023 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE180637.
114.Zhai J, Guo J, Wan H, Qi L, Liu L, Yan L, Xiao Z, Xu Y, Yu D, Wu X, et al. Primate gastrulation and early organogenesis at single-cell resolution. Datasets. Gene Expression Omnibus. 2022. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE193007. [DOI] [PMC free article] [PubMed]
115.Tan GH, et al. A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs. 2026. Figshare. 10.6084/m9.figshare.31937505. [DOI] [PMC free article] [PubMed]
116.Jiang L, Zheng Z, Fang H, Yang J. A generalized linear mixed model association tool for biobank-scale data. Datasets. GWAS Catalog. 2021.https://www.ebi.ac.uk/gwas/studies/GCST90044489. [DOI] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional file 1: Fig. S1-S17.^{(5.8MB, pdf)}

13059_2026_4173_MOESM2_ESM.xlsx^{(637.2KB, xlsx)}

13059_2026_4173_MOESM3_ESM.xlsx^{(25.4KB, xlsx)}

13059_2026_4173_MOESM4_ESM.xlsx^{(6.9MB, xlsx)}

Additional file 4: DEGs of trophoblasts across gestation.

13059_2026_4173_MOESM5_ESM.xlsx^{(543.4KB, xlsx)}

Additional file 5: Clustering of gene expression and metabolic pathways in trophoblasts across gestational stages.

13059_2026_4173_MOESM6_ESM.xlsx^{(1.2MB, xlsx)}

13059_2026_4173_MOESM7_ESM.xlsx^{(2.6MB, xlsx)}

Data Availability Statement

[CR1] 1.Wiltbank MC, Baez GM, Garcia-Guerra A, Toledo MZ, Monteiro PL, Melo LF, et al. Pivotal periods for pregnancy loss during the first trimester of gestation in lactating dairy cows. Theriogenology. 2016;86:239–53. [DOI] [PubMed] [Google Scholar]

[CR2] 2.Sánchez JM, Mathew DJ, Passaro C, Fair T, Lonergan P. Embryonic maternal interaction in cattle and its relationship with fertility. Reprod Domest Anim. 2018;53(Suppl 2):20–7. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Davenport KM, Ortega MS, Liu H, O’Neil EV, Kelleher AM, Warren WC, et al. Single-nuclei RNA sequencing (snRNA-seq) uncovers trophoblast cell types and lineages in the mature bovine placenta. Proc Natl Acad Sci U S A. 2023;120:e2221526120. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Diskin MG, Morris DG. Embryonic and early foetal losses in cattle and other ruminants. Reprod Domest Anim. 2008;43(Suppl 2):260–7. [DOI] [PubMed] [Google Scholar]

[CR5] 5.Albaaj A, Durocher J, LeBlanc SJ, Dufour S. Meta-analysis of the incidence of pregnancy losses in dairy cows at different stages to 90 days of gestation. JDS Commun. 2023;4:144–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Green JA, Geisert RD, Johnson GA, Spencer TE. Implantation and placentation in ruminants. Adv Anat Embryol Cell Biol. 2021;234:129–54. [DOI] [PubMed] [Google Scholar]

[CR7] 7.Moraes JGN, Behura SK, Geary TW, Hansen PJ, Neibergs HL, Spencer TE. Uterine influences on conceptus development in fertility-classified animals. Proc Natl Acad Sci U S A. 2018;115:E1749-e1758. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Eley RM, Thatcher WW, Bazer FW, Wilcox CJ, Becker RB, Head HH, et al. Development of the conceptus in the bovine. J Dairy Sci. 1978;61:467–73. [DOI] [PubMed] [Google Scholar]

[CR9] 9.Davenport KM, Ortega MS, Johnson GA, Seo H, Spencer TE. Review: implantation and placentation in ruminants. Animal. 2023;17(Suppl 1):100796. [DOI] [PubMed] [Google Scholar]

[CR10] 10.Wooding FBP. The ruminant placental trophoblast binucleate cell: an evolutionary breakthrough. Biol Reprod. 2022;107:705–16. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Davies Morel MC, Newcombe JR, Holland SJ. Factors affecting gestation length in the Thoroughbred mare. Anim Reprod Sci. 2002;74:175–85. [DOI] [PubMed] [Google Scholar]

[CR12] 12.Clausson B, Lichtenstein P, Cnattingius S. Genetic influence on birthweight and gestational length determined by studies in offspring of twins. BJOG. 2000;107:375–81. [DOI] [PubMed] [Google Scholar]

[CR13] 13.Fang L, Jiang J, Li B, Zhou Y, Freebern E, Vanraden PM, et al. Genetic and epigenetic architecture of paternal origin contribute to gestation length in cattle. Commun Biol. 2019;2:100. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Moradi M, Zhandi M, Sharafi M, Akbari A, Atrabi MJ, Totonchi M. Gene expression profile of placentomes and clinical parameters in the cows with retained placenta. BMC Genomics. 2022;23:760. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Weeks AD. The retained placenta. Best Pract Res Clin Obstet Gynaecol. 2008;22:1103–17. [DOI] [PubMed] [Google Scholar]

[CR16] 16.Rojas de Oliveira H, Chud TCS, Oliveira GA Jr., Hermisdorff IC, Narayana SG, Rochus CM, et al. Genome-wide association analyses reveal copy number variant regions associated with reproduction and disease traits in Canadian Holstein cattle. J Dairy Sci. 2024;107:7052–63. [DOI] [PubMed]

[CR17] 17.Arutyunyan A, Roberts K, Troulé K, Wong FCK, Sheridan MA, Kats I, et al. Spatial multiomics map of trophoblast development in early pregnancy. Nature. 2023;616:143–51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Pavličev M, Wagner GP, Chavan AR, Owens K, Maziarz J, Dunn-Fletcher C, et al. Single-cell transcriptomics of the human placenta: inferring the cell communication network of the maternal-fetal interface. Genome Res. 2017;27:349–61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Suryawanshi H, Morozov P, Straus A, Sahasrabudhe N, Max KEA, Garzia A, et al. A single-cell survey of the human first-trimester placenta and decidua. Sci Adv. 2018;4:eaau4788. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Marsh B, Blelloch R. Single nuclei RNA-seq of mouse placental labyrinth development. Elife. 2020. 10.7554/eLife.60266. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Jiang X, Wang Y, Xiao Z, Yan L, Guo S, Wang Y, et al. A differentiation roadmap of murine placentation at single-cell resolution. Cell Discov. 2023;9:30. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Scatolin GN, Ming H, Wang Y, Iyyappan R, Gutierrez-Castillo E, Zhu L, et al. Single-cell transcriptional landscapes of bovine peri-implantation development. iScience. 2024. 10.1016/j.isci.2024.109605. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Davenport KM, O’Neil EV, Ortega MS, Patterson A, Kelleher AM, Warren WC, et al. Single-cell insights into development of the bovine placenta†. Biol Reprod. 2024;110:169–84. [DOI] [PubMed] [Google Scholar]

[CR24] 24.Hradecký P, Mossman HW, Stott GG. Comparative development of ruminant placentomes. Theriogenology. 1988;29:715–29. [DOI] [PubMed] [Google Scholar]

[CR25] 25.Bauersachs S, Ulbrich SE, Gross K, Schmidt SEM, Meyer HHD, Wenigerkind H, et al. Embryo-induced transcriptome changes in bovine endometrium reveal species-specific and common molecular markers of uterine receptivity. Reproduction. 2006;132:319–31. [DOI] [PubMed] [Google Scholar]

[CR26] 26.Bairagi S, Quinn KE, Crane AR, Ashley RL, Borowicz PP, Caton JS, et al. Maternal environment and placental vascularization in small ruminants. Theriogenology. 2016;86:288–305. [DOI] [PubMed] [Google Scholar]

[CR27] 27.Gayoso A, Lopez R, Xing G, Boyeau P, Valiollah Pour Amiri V, Hong J, et al. A Python library for probabilistic analysis of single-cell omics data. Nat Biotechnol. 2022;40:163–6. [DOI] [PubMed] [Google Scholar]

[CR28] 28.Hue I. Determinant molecular markers for peri-gastrulating bovine embryo development. Reprod Fertil Dev. 2016;28:51–65. [DOI] [PubMed] [Google Scholar]

[CR29] 29.Jia G-X, Ma W-J, Wu Z-B, Li S, Zhang X-Q, He Z, et al. Single-cell transcriptomic characterization of sheep conceptus elongation and implantation. Cell Rep. 2023. 10.1016/j.celrep.2023.112860. [DOI] [PubMed] [Google Scholar]

[CR30] 30.Liu T, Li J, Yu L, Sun HX, Li J, Dong G, et al. Cross-species single-cell transcriptomic analysis reveals pre-gastrulation developmental differences among pigs, monkeys, and humans. Cell Discov. 2021;7:8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Steinhauser CB, Landers M, Myatt L, Burghardt RC, Vallet JL, Bazer FW, et al. Fructose synthesis and transport at the uterine-placental interface of pigs: cell-specific localization of SLC2A5, SLC2A8, and components of the polyol pathway. Biol Reprod. 2016;95:108. [DOI] [PubMed] [Google Scholar]

[CR32] 32.Degrelle SA, Murthi P, Evain-Brion D, Fournier T, Hue I. Expression and localization of DLX3, PPARG and SP1 in bovine trophoblast during binucleated cell differentiation. Placenta. 2011;32:917–20. [DOI] [PubMed] [Google Scholar]

[CR33] 33.Yahyazadeh Mashhadi SM, Kazemimanesh M, Arashkia A, Azadmanesh K, Meshkat Z, Golichenari B, et al. Shedding light on the EpCAM: an overview. J Cell Physiol. 2019;234:12569–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Kuriyama S, Tamiya Y, Tanaka M. Spatiotemporal expression of UPK3B and its promoter activity during embryogenesis and spermatogenesis. Histochem Cell Biol. 2017;147:17–26. [DOI] [PubMed] [Google Scholar]

[CR35] 35.Deng FM, Liang FX, Tu L, Resing KA, Hu P, Supino M, et al. Uroplakin IIIb, a urothelial differentiation marker, dimerizes with uroplakin Ib as an early step of urothelial plaque assembly. J Cell Biol. 2002;159:685–94. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Shin JH, Son EJ, Lee HS, Kim SJ, Kim K, Choi JY, et al. Molecular and functional expression of anion exchangers in cultured normal human nasal epithelial cells. Acta Physiol (Oxf). 2007;191:99–110. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Chen W, Kang KL, Alshaikh A, Varma S, Lin YL, Shin KH, et al. Grainyhead-like 2 (GRHL2) knockout abolishes oral cancer development through reciprocal regulation of the MAP kinase and TGF-β signaling pathways. Oncogenesis. 2018;7:38. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.MacFawn I, Wilson H, Selth LA, Leighton I, Serebriiskii I, Bleackley RC, et al. Grainyhead-like-2 confers NK-sensitivity through interactions with epigenetic modifiers. Mol Immunol. 2019;105:137–49. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Di Palma T, Filippone MG, Pierantoni GM, Fusco A, Soddu S, Zannini M. Pax8 has a critical role in epithelial cell survival and proliferation. Cell Death Dis. 2013;4:e729. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Fu DJ, De Micheli AJ, Bidarimath M, Ellenson LH, Cosgrove BD, Flesken-Nikitin A, et al. Cells expressing PAX8 are the main source of homeostatic regeneration of adult mouse endometrial epithelium and give rise to serous endometrial carcinoma. Dis Model Mech. 2020. 10.1242/dmm.047035. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Wang F, Ding P, Liang X, Ding X, Brandt CB, Sjöstedt E, et al. Endothelial cell heterogeneity and microglia regulons revealed by a pig cell landscape at single-cell level. Nat Commun. 2022;13:3620. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Jiang X, Zhai J, Xiao Z, Wu X, Zhang D, Wan H, et al. Identifying a dynamic transcriptomic landscape of the cynomolgus macaque placenta during pregnancy at single-cell resolution. Dev Cell. 2023;58:806-821.e807. [DOI] [PubMed] [Google Scholar]

[CR43] 43.Wu JJ, Zhu S, Gu F, Valencak TG, Liu JX, Sun HZ. Cross-tissue single-cell transcriptomic landscape reveals the key cell subtypes and their potential roles in the nutrient absorption and metabolism in dairy cattle. J Adv Res. 2022;37:1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Ziegler-Heitbrock L. The CD14+ CD16+ blood monocytes: their role in infection and inflammation. J Leukoc Biol. 2007;81:584–92. [DOI] [PubMed] [Google Scholar]

[CR45] 45.Müller I, Vogl T, Kühl U, Krannich A, Banks A, Trippel T, et al. Serum alarmin S100A8/S100A9 levels and its potential role as biomarker in myocarditis. ESC Heart Fail. 2020;7:1442–51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Seong J, Frias-Aldeguer J, Holzmann V, Kagawa H, Sestini G, Heidari Khoei H, et al. Epiblast inducers capture mouse trophectoderm stem cells in vitro and pattern blastoids for implantation in utero. Cell Stem Cell. 2022;29:1102-1118.e1108. [DOI] [PubMed] [Google Scholar]

[CR47] 47.Pfeffer PL. Alternative mammalian strategies leading towards gastrulation: losing polar trophoblast (Rauber’s layer) or gaining an epiblast cavity. Philos Trans R Soc Lond B Biol Sci. 2022;377:20210254. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Wang Y, Ming H, Yu L, Li J, Zhu L, Sun H-X, et al. Establishment of bovine trophoblast stem cells. Cell Rep. 2023. 10.1016/j.celrep.2023.112439. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Zhao G, Jiang K, Zhang T, Wu H, Qiu C, Deng G. Specific interferon tau gene-regulation networks in bovine endometrial luminal epithelial cells. Theriogenology. 2018;105:51–60. [DOI] [PubMed] [Google Scholar]

[CR50] 50.Strieder-Barboza C, de Souza J, Raphael W, Lock AL, Contreras GA. Fetuin-A: a negative acute-phase protein linked to adipose tissue function in periparturient dairy cows. J Dairy Sci. 2018;101:2602–16. [DOI] [PubMed] [Google Scholar]

[CR51] 51.Wu J-J, Zheng E, Liu L, Quan J, Ruan D, Yao Z, et al. Cell-cell communication-mediated cell-type-specific parent-of-origin effects in mammals. Nat Commun. 2025;16:5106. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Khatri P, Frenette G, Sullivan R, Hoffmann B, Schuler G. Expression of SULT1E1 protein in bovine placentomes: evidence for localization in uninucleated trophoblast cells. Placenta. 2011;32:431–40. [DOI] [PubMed] [Google Scholar]

[CR53] 53.Simner C, Novakovic B, Lillycrop KA, Bell CG, Harvey NC, Cooper C, et al. DNA methylation of amino acid transporter genes in the human placenta. Placenta. 2017;60:64–73. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Hecker N, Sharma V, Hiller M. Convergent gene losses illuminate metabolic and physiological changes in herbivores and carnivores. Proc Natl Acad Sci U S A. 2019;116:3036–41. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Lee B-K, Jang Y, Kim M, LeBlanc L, Rhee C, Lee J, et al. Super-enhancer-guided mapping of regulatory networks controlling mouse trophoblast stem cells. Nat Commun. 2019;10:4749. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Wu X, Xu W, Deng L, Li Y, Wang Z, Sun L, et al. Spatial multi-omics at subcellular resolution via high-throughput in situ pairwise sequencing. Nat Biomed Eng. 2024;8:872–89. [DOI] [PubMed] [Google Scholar]

[CR57] 57.Vento-Tormo R, Efremova M, Botting RA, Turco MY, Vento-Tormo M, Meyer KB, et al. Single-cell reconstruction of the early maternal-fetal interface in humans. Nature. 2018;563:347–53. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Marsh B, Zhou Y, Kapidzic M, Fisher S, Blelloch R. Regionally distinct trophoblast regulate barrier function and invasion in the human placenta. Elife. 2022. 10.7554/eLife.78829. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Tsang JCH, Vong JSL, Ji L, Poon LCY, Jiang P, Lui KO, et al. Integrative single-cell and cell-free plasma RNA transcriptomics elucidates placental cellular dynamics. Proc Natl Acad Sci U S A. 2017;114:E7786-e7795. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Spencer TE, Hansen TR. Implantation and establishment of pregnancy in ruminants. Adv Anat Embryol Cell Biol. 2015;216:105–35. [DOI] [PubMed] [Google Scholar]

[CR61] 61.Zhang L, Yu X, Zheng L, Zhang Y, Li Y, Fang Q, et al. Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. Nature. 2018;564:268–72. [DOI] [PubMed] [Google Scholar]

[CR62] 62.Van de Sande B, Flerin C, Davie K, De Waegeneer M, Hulselmans G, Aibar S, et al. A scalable SCENIC workflow for single-cell gene regulatory network analysis. Nat Protoc. 2020;15:2247–76. [DOI] [PubMed] [Google Scholar]

[CR63] 63.Fuxman Bass JI, Diallo A, Nelson J, Soto JM, Myers CL, Walhout AJ. Using networks to measure similarity between genes: association index selection. Nat Methods. 2013;10:1169–76. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Mi C, Ye B, Gao Z, Du J, Li R, Huang D. BHLHE40 plays a pathological role in pre-eclampsia through upregulating SNX16 by transcriptional inhibition of miR-196a-5p. Mol Hum Reprod. 2020;26:532–48. [DOI] [PubMed] [Google Scholar]

[CR65] 65.Lv B, An Q, Zeng Q, Zhang X, Lu P, Wang Y, et al. Single-cell RNA sequencing reveals regulatory mechanism for trophoblast cell-fate divergence in human peri-implantation conceptuses. PLoS Biol. 2019;17:e3000187. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR66] 66.Sharifi O, Haghani V, Neier KE, Fraga KJ, Korf I, Hakam SM, et al. Sex-specific single cell-level transcriptomic signatures of Rett syndrome disease progression. Commun Biol. 2024;7:1292. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Fitzgerald JS, Poehlmann TG, Schleussner E, Markert UR. Trophoblast invasion: the role of intracellular cytokine signalling via signal transducer and activator of transcription 3 (STAT3). Hum Reprod Update. 2008;14:335–44. [DOI] [PubMed] [Google Scholar]

[CR68] 68.Mor G, Cardenas I. The immune system in pregnancy: a unique complexity. Am J Reprod Immunol. 2010;63:425–33. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR69] 69.Ezashi T, Imakawa K. Transcriptional control of IFNT expression. Reproduction. 2017;154:F21-f31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Chakrabarty A, Roberts RM. Ets-2 and C/EBP-beta are important mediators of ovine trophoblast Kunitz domain protein-1 gene expression in trophoblast. BMC Mol Biol. 2007;8:14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Ma B, Cui H, Wang X, Feng W, Zhang J, Chen N, et al. IFNT-induced IRF1 enhances bovine endometrial receptivity by transactivating LIFR. J Reprod Immunol. 2024;163:104212. [DOI] [PubMed] [Google Scholar]

[CR72] 72.Basavaraja R, Przygrodzka E, Pawlinski B, Gajewski Z, Kaczmarek MM, Meidan R. Interferon-tau promotes luteal endothelial cell survival and inhibits specific luteolytic genes in bovine corpus luteum. Reproduction. 2017;154:559–68. [DOI] [PubMed] [Google Scholar]

[CR73] 73.Cornelis G, Heidmann O, Degrelle SA, Vernochet C, Lavialle C, Letzelter C, et al. Captured retroviral envelope syncytin gene associated with the unique placental structure of higher ruminants. Proc Natl Acad Sci U S A. 2013;110:E828-837. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Klisch K, Pfarrer C, Schuler G, Hoffmann B, Leiser R. Tripolar acytokinetic mitosis and formation of feto-maternal syncytia in the bovine placentome: different modes of the generation of multinuclear cells. Anat Embryol (Berl). 1999;200:229–37. [DOI] [PubMed] [Google Scholar]

[CR75] 75.Chen HZ, Ouseph MM, Li J, Pécot T, Chokshi V, Kent L, et al. Canonical and atypical E2Fs regulate the mammalian endocycle. Nat Cell Biol. 2012;14:1192–202. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR76] 76.Hughes M, Dobric N, Scott IC, Su L, Starovic M, St-Pierre B, et al. The Hand1, Stra13 and Gcm1 transcription factors override FGF signaling to promote terminal differentiation of trophoblast stem cells. Dev Biol. 2004;271:26–37. [DOI] [PubMed] [Google Scholar]

[CR77] 77.Jiang Y, Chen Y, Chen Y. Knockdown of JARID2 inhibits the viability and migration of placenta trophoblast cells in preeclampsia. Mol Med Rep. 2017;16:3594–9. [DOI] [PubMed] [Google Scholar]

[CR78] 78.Kim M, Adu-Gyamfi EA, Kim J, Lee BK. Super-enhancer-associated transcription factors collaboratively regulate trophoblast-active gene expression programs in human trophoblast stem cells. Nucleic Acids Res. 2023;51:3806–19. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR79] 79.Dickinson ME, Flenniken AM, Ji X, Teboul L, Wong MD, White JK, et al. High-throughput discovery of novel developmental phenotypes. Nature. 2016;537:508–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR80] 80.Koscielny G, Yaikhom G, Iyer V, Meehan TF, Morgan H, Atienza-Herrero J, et al. The International Mouse Phenotyping Consortium web portal, a unified point of access for knockout mice and related phenotyping data. Nucleic Acids Res. 2014;42:D802-809. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR81] 81.Zhang YH, He M, Wang Y, Liao AH. Modulators of the balance between M1 and M2 macrophages during pregnancy. Front Immunol. 2017;8:120. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR82] 82.Tan G-H, Liu S-J, Dou M-L, Zhao D-F, Zhang A, Li H-K, et al. Spatially resolved transcriptomic profiling of placental development in dairy cow. Zool Res. 2024;45:586–600. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR83] 83.Brooks K, Burns G, Spencer TE. Conceptus elongation in ruminants: roles of progesterone, prostaglandin, interferon tau and cortisol. J Anim Sci Biotechnol. 2014;5:53. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR84] 84.Courtney JA, Wilson RL, Cnota J, Jones HN. Conditional mutation of Hand1 in the mouse placenta disrupts placental vascular development resulting in fetal loss in both early and late pregnancy. Int J Mol Sci. 2021. 10.3390/ijms22179532. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR85] 85.Fresch R, Courtney J, Brockway H, Wilson RL, Jones H. HAND1 knockdown disrupts trophoblast global gene expression. Physiol Rep. 2023;11:e15553. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR86] 86.Zadora J, Singh M, Herse F, Przybyl L, Haase N, Golic M, et al. Disturbed placental imprinting in preeclampsia leads to altered expression of DLX5, a human-specific early trophoblast marker. Circulation. 2017;136:1824–39. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR87] 87.Novakovic B, Fournier T, Harris LK, James J, Roberts CT, Yong HEJ, et al. Increased methylation and decreased expression of homeobox genes TLX1, HOXA10 and DLX5 in human placenta are associated with trophoblast differentiation. Sci Rep. 2017;7:4523. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR88] 88.Shoshan-Barmatz V, Shteinfer-Kuzmine A, Verma A. VDAC1 at the intersection of cell metabolism, apoptosis, and diseases. Biomolecules. 2020. 10.3390/biom10111485. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR89] 89.Liang YD, Zhao K, Chen Y, Zhang SQ. Role of CYCS in the cytogenesis and apoptosis of male germ cells and its clinical application. Zhonghua Nan Ke Xue. 2020;26:265–70. [PubMed] [Google Scholar]

[CR90] 90.Liao Y, Peng S, He L, Wang Y, Li Y, Ma D, et al. Methylmercury cytotoxicity and possible mechanisms in human trophoblastic HTR-8/SVneo cells. Ecotoxicol Environ Saf. 2021;207:111520. [DOI] [PubMed] [Google Scholar]

[CR91] 91.Shi M, Lv X, Zhu M, Dong Y, Hu L, Qian Y, et al. HMGA1 promotes hepatocellular carcinoma proliferation, migration, and regulates cell cycle via miR-195-5p. Anticancer Drugs. 2022;33:e273–85. [DOI] [PubMed] [Google Scholar]

[CR92] 92.Matsubara K, Matsubara Y, Uchikura Y, Takagi K, Yano A, Sugiyama T. HMGA1 is a potential driver of preeclampsia pathogenesis by interference with extravillous trophoblasts invasion. Biomolecules. 2021. 10.3390/biom11060822. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR93] 93.Ali A, Stenglein MD, Spencer TE, Bouma GJ, Anthony RV, Winger QA. Trophectoderm-specific knockdown of LIN28 decreases expression of genes necessary for cell proliferation and reduces elongation of sheep conceptus. Int J Mol Sci. 2020. 10.3390/ijms21072549. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR94] 94.McGinnis CS, Murrow LM, Gartner ZJ. DoubletFinder: doublet detection in single-cell RNA sequencing data using artificial nearest neighbors. Cell Syst. 2019;8:329-337.e324. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR95] 95.Luecken MD, Büttner M, Chaichoompu K, Danese A, Interlandi M, Mueller MF, et al. Benchmarking atlas-level data integration in single-cell genomics. Nat Methods. 2022;19:41–50. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR96] 96.Heaton H, Talman AM, Knights A, Imaz M, Gaffney DJ, Durbin R, et al. Souporcell: robust clustering of single-cell RNA-seq data by genotype without reference genotypes. Nat Methods. 2020;17:615–20. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR97] 97.Yu G, Wang LG, Han Y, He QY. ClusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR98] 98.Suo S, Zhu Q, Saadatpour A, Fei L, Guo G, Yuan GC. Revealing the critical regulators of cell identity in the mouse cell atlas. Cell Rep. 2018;25:1436-1445.e1433. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR99] 99.La Manno G, Soldatov R, Zeisel A, Braun E, Hochgerner H, Petukhov V, et al. RNA velocity of single cells. Nature. 2018;560:494–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR100] 100.Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ. Generalizing RNA velocity to transient cell states through dynamical modeling. Nat Biotechnol. 2020;38:1408–14. [DOI] [PubMed] [Google Scholar]

[CR101] 101.Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM 3rd, et al. Comprehensive integration of single-cell data. Cell. 2019;177:1888-1902.e1821. [DOI] [PMC free article] [PubMed]

[CR102] 102.Qiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, et al. Reversed graph embedding resolves complex single-cell trajectories. Nat Methods. 2017;14:979–82. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR103] 103.Ma Y, Deng C, Zhou Y, Zhang Y, Qiu F, Jiang D, et al. Polygenic regression uncovers trait-relevant cellular contexts through pathway activation transformation of single-cell RNA sequencing data. Cell Genom. 2023;3:100383. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR104] 104.Lek M, Karczewski KJ, Minikel EV, Samocha KE, Banks E, Fennell T, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536:285–91. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR105] 105.Tan GH, et al. Single-cell RNA-sequencing reveals the placental molecular dynamics during bovine gestation. NCBI SRA BioProject PRJNA1302120. 2026. https://www.ncbi.nlm.nih.gov/sra/PRJNA1302120.

[CR106] 106.Tan G, et al. A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs. Gene Expression Omnibus. 2026. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE306749. [DOI] [PMC free article] [PubMed]

[CR107] 107.Scatolin GN, Ming H, Wang YJ, Iyyappan R, Gutierrez-Castillo E, Zhu LK, Sagheer M, Song C, Bondioli K, Jiang ZL. Single-cell transcriptional landscapes of bovine peri-implantation development. Datasets. Gene Expression Omnibus. 2023.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234335. [DOI] [PMC free article] [PubMed]

[CR108] 108.Davenport KM, O'Neil EV, Ortega MS, Patterson A, Kelleher AM, Warren WC, Spencer TE. Single cell insights into development of the bovine placenta. Datasets. Gene Expression Omnibus. 2023.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234524. [DOI] [PubMed]

[CR109] 109.Vento-Tormo R, Efremova M, Botting RA, Turco MY, Vento-Tormo M, Meyer KB, Park JE, Stephenson E, Polanski K, Goncalves A, et al. Reconstructing the human first trimester fetal-maternal interface using single cell transcriptomics -10x data. Datasets. ArrayExpress. 2018. https://www.ebi.ac.uk/biostudies/arrayexpress/studies/E-MTAB-6701.

[CR110] 110.Marsh B, Zhou Y, Kapidzic M, Fisher S, Blelloch R. Regionally distinct trophoblast regulate barrier function and invasion in the human placenta. Datasets. Gene Expression Omnibus. 2022. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE198373. [DOI] [PMC free article] [PubMed]

[CR111] 111.Tsang JCH, Vong JSL, Ji L, Poon LCY, Jiang P, Lui KO, Ni Y, To KF, Cheng YKY, Chiu RWK, et al. Integrative single-cell and cell-free plasma RNA transcriptomics elucidates placental cellular dynamics. Datasets. European Genome-Phenome Archive. 2017. https://ega-archive.org/datasets/EGAD00001003705. [DOI] [PMC free article] [PubMed]

[CR112] 112.Jiang X, Wang Y, Xiao Z, Yan L, Guo S, Wang Y, Wu H, Zhao X, Lu X, Wang H. A differentiation roadmap of murine placentation at single-cell resolution. Datasets. Gene Expression Omnibus. 2023. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE156125. [DOI] [PMC free article] [PubMed]

[CR113] 113.Jiang X, Zhai J, Xiao Z, Wu X, Wang H, Wan H, Xu Y, Zhang D, et al. A single-cell transcriptome landscape of the non-human primate placenta across pregnancy. Datasets. Gene Expression Omnibus. 2023 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE180637.

[CR114] 114.Zhai J, Guo J, Wan H, Qi L, Liu L, Yan L, Xiao Z, Xu Y, Yu D, Wu X, et al. Primate gastrulation and early organogenesis at single-cell resolution. Datasets. Gene Expression Omnibus. 2022. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE193007. [DOI] [PMC free article] [PubMed]

[CR115] 115.Tan GH, et al. A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs. 2026. Figshare. 10.6084/m9.figshare.31937505. [DOI] [PMC free article] [PubMed]

[CR116] 116.Jiang L, Zheng Z, Fang H, Yang J. A generalized linear mixed model association tool for biobank-scale data. Datasets. GWAS Catalog. 2021.https://www.ebi.ac.uk/gwas/studies/GCST90044489. [DOI] [PubMed]

PERMALINK

A longitudinal single-nucleus transcriptomic atlas of bovine placentation reveals dynamic cellular hierarchies and regulatory programs

Guanghui Tan

Xiaoru Yan

Xuesha Cao

Ao Zhang

Tao Shi

Yaqi Zhou

Xi Guo

Haoping Wang

Hengkuan Li

Jiezhang Li

Yingran Liu

Xihong Wang

Yu Jiang

Yi Zheng

Abstract

Background

Results

Conclusions

Supplementary Information

Background

Results

Construction of a single-cell transcriptomic atlas of bovine placentas throughout gestation

Fig. 1.

Spatiotemporal mapping of bovine trophoblasts uncovered species-specific lineage programs and stage-dependent shifts in cellular composition

Fig. 2.

Stage-resolved TF modules delineated the regulatory landscape of trophoblast development

Fig. 3.

The developmental atlas of UNCs revealed key TFs during implantation

Fig. 4.

Cell trajectory analysis revealed the developmental origin and key regulatory networks of BNCs

Fig. 5.

Trophoblast development in the bovine placenta involved extensive transcriptional and metabolic reprogramming

Fig. 6.

Early trophoblasts were associated with GL and exhibited a strong evolutionary constraint

Fig. 7.

Dynamic polarization of placental macrophages and its correlation with genetic susceptibility and signaling networks underlying RP

Fig. 8.

Discussion

Fig. 9.

Conclusions

Methods

Animal experiments and sample collection

Sample preparation for snRNA-seq

snRNA-seq data analysis

MiP-seq

In situlibrary construction and high-plex imaging

Spatial image processing and interactive single-cell mapping

Comparisons with published bovine placental single-cell transcriptomic atlases and cross-breed datasets

Differentially expressed gene (DEG), GO, and GSEA analyses

Analysis of stage-specific cell type distribution

Analysis of TF regulation

RNA velocity and pseudotime analyses

Developmental trajectory and regulatory characterization of BNCs emerging on E24

Identification of gene expression and metabolic activity modules

Integration of GWAS and snRNA-seq data to identify trait-associated cell types

Expression patterns of functionally constrained and lethality-associated genes

Macrophage polarization activity fraction

Cell–cell communication analysis

Supplementary Information

Acknowledgements

Peer review information

Authors’ contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles