Decision letter: Regionally distinct trophoblast regulate barrier function and invasion in the human placenta

Abstract

Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract The human placenta contains two specialized regions: the villous chorion where gases and nutrients are exchanged between maternal and fetal blood, and the smooth chorion (SC) which surrounds more than 70% of the developing fetus but whose cellular composition and function is poorly understood. Here, we use single cell RNA-sequencing to compare the cell types and molecular programs between these two regions in the second trimester human placenta. Each region consists of progenitor cytotrophoblasts (CTBs) and extravillous trophoblasts (EVTs) with similar gene expression programs. While CTBs in the villous chorion differentiate into syncytiotrophoblasts, they take an alternative trajectory in the SC producing a previously unknown CTB population which we term SC-specific CTBs (SC-CTBs). Marked by expression of region-specific cytokeratins, the SC-CTBs form a stratified epithelium above a basal layer of progenitor CTBs. They express epidermal and metabolic transcriptional programs consistent with a primary role in defense against physical stress and pathogens. Additionally, we show that SC-CTBs closely associate with EVTs and secrete factors that inhibit the migration of the EVTs. This restriction of EVT migration is in striking contrast to the villous region where EVTs migrate away from the chorion and invade deeply into the decidua. Together, these findings greatly expand our understanding of CTB differentiation in these distinct regions of the human placenta. This knowledge has broad implications for studies of the development, functions, and diseases of the human placenta. Editor's evaluation By using single-cell RNA sequencing, elegant computational approaches, protein validation, and in vitro functional assays, this study characterizes the cellular composition and gene expression profiles of the human placenta in mid-gestation. In addition, this work gives new insights into our understanding of trophoblast differentiation in distinct regions of the human placenta. The findings and dataset provided by the authors represent an important resource for readers interested in human development and placenta biology. https://doi.org/10.7554/eLife.78829.sa0 Decision letter Reviews on Sciety eLife's review process Introduction The human placenta is the first organ to develop and forms the essential bridge between maternal and fetal tissues beginning at implantation (Knöfler et al., 2019; Turco et al., 2018). The placenta must begin development rapidly upon conception in order to perform the roles of future organ systems that have not yet developed and matured in the fetus, including nutrient and oxygen transport and protection from mechanical and pathogenic insults. The placenta also performs unique functions such as modulation of maternal tolerance and hormone production (Maltepe and Fisher, 2015; Knöfler et al., 2019; Turco et al., 2018). Placental development begins with the generation of stem villi surrounding the entire embryo and then proceeds asymmetrically to produce two distinct regions. At the human implantation site, the embryonic pole, the villi grow and branch to give rise to the region essential for the exchange of gases and nutrients, the chorion frondosum (also known as the chorionic villi or the villous chorion [VC]). This region includes the placental villi and the invasive EVTs. The villi located on the opposite side, the abembryonic pole, degenerate resulting in a smooth surface lacking villi termed the chorion leave (also known as the smooth chorion [SC]). The SC fuses with the amnion forming the chorioamniotic membranes (also known as the fetal membranes) (Hamilton and Boyd, 1960; Boyd and Hamilton, 1967; Benirschke et al., 2006). Both the VC and SC are comprised of fetal derived cytotrophoblasts (CTBs), with CTBs in the VC differentiating to either multinucleate syncytiotrophoblast (STB) or to invasive extravillous trophoblast (EVT) (Knöfler et al., 2019). Compared to the VC, little effort has been made to analyze types and functions of the CTBs that comprise the SC (Benirschke et al., 2006; Garrido-Gomez et al., 2017). The CTBs of the SC exist in an epithelial-like structure and lack STBs and proximity to the fetal vasculature, and thus cannot function in a manner comparable to villi in VC (Benirschke et al., 2006). Furthermore, in contrast to the VC where EVTs invade and remodel the maternal arteries, the cells of the SC do not invade the adjacent decidua and the maternal blood vessels it contains (Genbačev et al., 2015). Thus, the function of these SC CTBs remains unclear. Several pieces of evidence suggest that the SC is not simply a vestigial structure. First, the intact CTB layer contains proliferating cells and is maintained until term (Yeh et al., 1989; Benirschke et al., 2006; Garrido-Gomez et al., 2017). Second, the histological heterogeneity among SC CTBs suggests functional distinctions. Yeh et al., 1989, characterize two distinct populations of vacuolated and eosinophilic CTBs. Vacuolated CTBs were positive for placental lactogen and placental alkaline phosphatase, while the eosinophilic subpopulations was not. Both populations were rich in keratin and neither had the known characteristics of villous CTBs. Bou-Resli et al., 1981 also note high levels of variation among CTBs in the SC and the existence of a vacuolated population. A more molecular characterization was carried out by Garrido-Gomez et al., 2017, which demonstrated heterogeneity of ITGA4 and HLA-G expression, markers previously associated with stemness and invasion, respectively (Genbacev et al., 2016; McMaster et al., 1995). This study also uncovered an expansion of the SC in cases of severe pre-eclampsia, along with a disease-specific gene expression pattern. In sum, these results suggest that the SC CTBs are a heterogeneous and dynamic collection of cells with important functions in development and disease. Single cell RNA-sequencing (scRNA-seq) has emerged as the standard for transcriptional characterization of complex organs. This methodology was previously applied to the placenta, but with a focus on the maternal-fetal interface, specifically the chorionic villi and basal plate (Liu et al., 2018; Suryawanshi et al., 2018; Vento-Tormo et al., 2018). Recently, scRNA-seq was used to profile the smooch chorion at term (Pique-Regi et al., 2019; Pique-Regi et al., 2020; Garcia-Flores et al., 2022). However, in Pique-Regi et al., 2019, only 132 CTBs were identified in the SC out of 29,921 cells (0.44%) collected from this region. A comparable number of CTBs in the SC were recovered in Garcia-Flores et al., 2022, potentially reflecting a difficulty in capturing these cells at term. To better understand the differences in the cell types and functions of the two sides of the developing placenta, we applied scRNA-seq to matching samples of cells isolated from the VC and SC regions of human samples from mid to late in the second trimester. We used scRNA-seq to compare the composition and developmental trajectories of CTBs in the VC and SC. The data were validated and extended with functional studies to gain initial insights into the basis of differential migration of trophoblasts in each region. These results identified a novel SC-specific CTB population important for establishment of a protective barrier and the suppression of trophoblast invasion. In addition, these data represent a resource of CTB types, proportions, and gene expression at mid-gestation against which age-related and pathogenic alterations can be measured. Results The transcriptional landscape of the VC and SC at mid-gestation To understand the cellular composition of the SC, we isolated and profiled cells from both the VC and the SC regions of four second trimester human placentas spanning gestational weeks 18–24 (GW18–24) using scRNA-seq (Figure 1a). We chose to analyze second trimester samples because the maturation of the SC is complete but the inflammation and apoptosis associated with membrane rupture and parturition is absent (Benirschke et al., 2006; Yuan et al., 2006; Yuan et al., 2008; Yuan et al., 2009; Figure 1a). SC and VC cells were isolated from each human placental sample allowing within and across patient comparisons. VC samples included cells isolated from floating and anchoring villi and areas surrounding the cell column, while most of the decidua (including spiral arteries) were dissected away. SC samples included the chorion and underlying stroma (mesenchymal and endothelial cells), but not the amnion and little of the neighboring decidua, which were also removed during dissection. CTBs were further enriched over stromal and immune cells during cell preparation as previously described (Garrido-Gomez et al., 2017 and in Materials and methods). The transcriptomes of the resulting cells were captured using the 10× Genomics scRNA-seq platform. Figure 1 with 8 supplements see all Download asset Open asset The transcriptional landscape of the villous (VC) and smooth chorion (SC) at mid-gestation. (a) Left: Schematic of the placenta at mid-gestation, highlighting the regions sampled, together with the methods used for cell isolation and characterization. Right: Schematic of the cell types and their organization in each region. (b) UMAPs of integrated samples, shown by region of origin (left – VC, right – SC), and colored according to broad cell type clusters. (c) Heatmap of the transcript expression of select cell identity markers across broad cell type clusters and regions. Values are scaled expression across the clusters of each region independently. Figure 1—source data 1 Marker genes for the clusters of the Immune subset. https://cdn.elifesciences.org/articles/78829/elife-78829-fig1-data1-v2.csv Download elife-78829-fig1-data1-v2.csv Figure 1—source data 2 Marker genes for the clusters of the Stroma subset. https://cdn.elifesciences.org/articles/78829/elife-78829-fig1-data2-v2.csv Download elife-78829-fig1-data2-v2.csv Each of the eight datasets (GW17.6, 18.2, 23.0, 24.0; VC and SC) was captured independently, then integrated computationally (Figure 1b; Stuart et al., 2019). We classified the cells of the integrated dataset into broad cell type clusters according to functional identities and annotated each by expression of canonical markers. CTBs were annotated as KRT7+, HLA-G-; EVTs as KRT7+, HLA-G+; immune cells as CD45+, VIM-; stromal cells as VIM+, CD45-; and the uterine epithelium by expression of EPCAM+, MUC1+, and MUC16+ (Figure 1c; Lee et al., 2016; McMaster et al., 1995; Vento-Tormo et al., 2018). Cells expressing exclusive markers of disparate cell types (co-expression of KRT7, HLA-G, HLA-A, VIM, ACTA2) were labelled as doublets and excluded from further analysis. The complete dataset used for further analysis contained 50,496 cells that passed quality control (between 500 and 6500 unique genes, fewer than 15% mitochondrial reads, doublets removed) (Figure 1—figure supplement 1a; McGinnis et al., 2019). Cells originating from each region (VC – 25,367 and SC – 25,129) and each sample (7181–17,705 cells per sample) were well represented (Figure 1b; Figure 1—figure supplement 1b and c). The number of cells in each broad cell type cluster demonstrated enrichment for CTBs, which represented more than 60% of the cells in the integrated dataset, as expected given the enrichment protocol (Figure 1—figure supplement 1d). The sex of each fetus was inferred by assaying expression of XIST in trophoblast cells isolated from each sample (Figure 1—figure supplement 2a), which allowed the assignment of cell types to either fetal or maternal origin (Figure 1—figure supplement 2b). Even though there was enrichment for CTBs, we still identified 14,805 immune cells and 3883 stromal cells. We reclustered these immune and stromal cell subsets individually and compared them to previously published single cell analyses, allowing for the annotation of subtypes within each group (Figure 1—figure supplement 3a-d; Figure 1—figure supplement 4a-d; Figure 1—figure supplement 5a-c; Figure 1—figure supplement 7a-c; Vento-Tormo et al., 2018; Pique-Regi et al., 2019). Independent clustering analysis of VC and SC cells identified the similar populations as the integrated immune and stromal subsets and confirmed the relative proportions of cell identities across each region (Figure 1—figure supplement 6a-d; Figure 1—figure supplement 8a-d). All immune clusters robustly expressed Xist in each sample, identifying them to be of maternal origin (Figure 1—figure supplement 2c). Almost twofold more immune cells were recovered from the VC than the SC, although it is possible that this change in proportion is an artifact of dissection and/or the CTB enrichment protocol (Figure 1—figure supplement 5d and e). Comparing the immune cell types identified in each region revealed a higher proportion of macrophages in the VC as compared to the SC, which contained a greater proportion of NK/T cells (Figure 1—figure supplement 5e; Figure 1—source data 1). While few stromal cells were isolated in the preparations, subclustering still revealed a differential composition of fetal stromal cells between the VC and SC (Figure 1—figure supplement 7a-c; Figure 1—source data 2). The majority of stromal cells recovered originated from the SC (2941 compared to 942 from VC). These cells included lymphatic endothelium (Pique-Regi et al., 2019) and two largely SC-specific mesenchymal cell populations of fetal origin, Mesenchyme 1 and Mesenchyme 3 (Figure 1—figure supplement 7c-e; Figure 1—figure supplement 2b). These two clusters are marked by elevated expression of EGFL6, DLK1, and uniquely by expression of COL11A1, which is observed only in the SC (Figure 1—figure supplement 7b and f ). Interestingly, several canonical CTB support factors including HGF, WNT2, and RSPO3 were expressed in fetal stromal populations in both regions, suggesting shared requirements for WNT and MET signaling (Figure 1—figure supplement 7g). Taken together these data demonstrate the identification of broad classes of CTBs, immune, and support cells from both the VC and SC regions . Identification of an SC-specific CTB population CTBs are the fetal cells that perform the specialized functions of the VC, and are required for normal fetal growth and development (Maltepe and Fisher, 2015; Turco et al., 2018; Knöfler et al., 2019). To better understand the composition of CTBs in the SC versus the VC, we subclustered this population (KRT7+, VIM-, CD45-, MUC1-). The CTB subset is comprised of 29,668 cells with similar representation and cell quality control metrics across all eight samples (Figure 2—figure supplement 1a, b, and c). This analysis identified 13 clusters including several CTB, EVT, and STB subtypes (Figure 2a; Figure 2—source data 1). Figure 2 with 4 supplements see all Download asset Open asset Identification of a smooth chorion-specific cytotrophoblast. (a) UMAP of subclustered trophoblasts (n=29,668). Colors correspond to the clusters at the right. (b) Dot plot showing average expression and percent of cells in each cluster as identified by the marker genes listed on the x-axis. The clusters are listed on the y-axis. (c) UMAP of subclustered trophoblasts from the villous chorion (VC) (left) or the SC (right). Clusters and colors are the same as in panel a. (d) Quantification of the number of cells in each trophoblast cluster from each region. Cells from the VC are shown in black. Cells from the SC are shown in blue. (e) Violin plot of PAGE4 transcript expression across all trophoblast clusters. (f) Immunofluorescence co-localization of PAGE4 with pan-cytokeratin (marker of all trophoblast) in the VC (left) or SC (right). (g) Violin plot of KRT6A transcript expression across all trophoblast clusters. (h) Immunofluorescence co-localization of KRT6 with pan-cytokeratin (marker of all trophoblast) in the VC (left) or SC (right). (i) Immunofluorescence co-localization of CDH1 and KRT6 in the VC (left) or SC (middle). High magnification inset is denoted by the white box (right). For all images, nuclei were visualized by DAPI stain; scale bar = 100 μm. Abbreviations: AV = anchoring villi; FV = floating villi; SC = smooth chorion epithelium; Amn. = amnion; Dec. = decidua. Figure 2—source data 1 Marker genes for each cluster in the trophoblast subset. https://cdn.elifesciences.org/articles/78829/elife-78829-fig2-data1-v2.csv Download elife-78829-fig2-data1-v2.csv Broad classes of trophoblast were annotated by established markers (STBs – CGA, CYP19A1, CSH1, CSH2; EVTs – HLA-G, DIO2; CTBs – PAGE4, PEG10, and no expression of EVT and STB markers) (Figure 2b and Figure 2—figure supplement 2a; McMaster et al., 1995; Lee et al., 2016; Suryawanshi et al., 2018; Liu et al., 2018). Comparison to previously published cell types in the VC and SC confirmed the identities of most clusters (Figure 2—figure supplement 1d-g; Vento-Tormo et al., 2018; Pique-Regi et al., 2019). Two cell clusters showed high expression of canonical phasic transcripts, including MKI67, with the S-phase cluster denoted by expression of PCNA and the G2/M-phase cluster by expression of TOP2A (Figure 2—figure supplement 2b; Tirosh et al., 2016). Both populations share gene expression with all clusters of CTBs, and therefore, were identified as actively cycling CTBs (Figure 2—figure supplement 2a). No STB or EVT markers were identified in the cycling clusters as was expected due to the requirement for cell cycle exit upon terminal differentiation to these lineages (Lu et al., 2017; Genbacev et al., 1997). CTBs separated into four clusters, CTB 1–4. CTB 1 cells highly expressed PAGE4, PEG10, and CDH1 (Figure 2b, Figure 2—figure supplement 2a, Figure 2—figure supplement 4a), which have been shown to be canonical markers of villous CTB (Lee et al., 2016; Suryawanshi et al., 2018). Overall, CTB 2–4 were more transcriptionally similar to each other than to CTB 1, indicating a transcriptional program that was distinct from canonical villous CTBs (Figure 2—figure supplement 2a and c). CTB 2–4 existed along a gradient of gene expression changes suggestive of various stages of a common differentiation pathway. However, each population expressed distinct transcripts corresponding to important proposed functions of the SC. CTB 2 cells highly expressed CLU, CFD, and IFIT3, suggesting roles in responding to bacterial or viral infection through the innate arm of the immune system (Thurman and Holers, 2006; Liu et al., 2011). CTB 3 cells upregulated EGLN3 and SLC2A3, indicating an HIF-mediated response to hypoxia and a switch toward glucose metabolism, likely as an adaptation to the decreased oxygen levels in the largely avascular SC region (del Peso et al., 2003; Maxwell et al., 1997). Finally, CTB 4 specifically expressed several cytokeratins — KRT6A, KRT17, and KRT14, found in many epithelial barrier tissues and important for maintenance of integrity in response to mechanical stressors (Karantza, 2011; Figure 2b, Figure 2—figure supplement 2a). Previous analysis of the SC region at term using scRNA-seq identified fewer than 0.5% of trophoblasts as being CTB and did not identify a separate KRT6 expressing CTB population (Pique-Regi et al., 2019). Computational integration of the data from Pique-Regi et al., 2019, with the trophoblast subset confirmed CTB 3 and CTB 4 to be unique to this study (Figure 2—figure supplement 1f and g ). These results establish a previously underappreciated transcriptional diversity of CTB subpopulations. Quantification of cells showed a strong regional bias in the number and proportion contributing to each CTB cluster (Figure 2c and d). In the VC samples, 53.8% of CTBs clustered in CTB 1 compared to 11.1% in the SC (Supplementary file 1). In contrast, the SC had a much larger proportion of the CTB 2–4 clusters. In this region, 71.3% of CTBs were nearly equally distributed among CTB 2 (25.6%), CTB 3 (24.9%), and CTB 4 (20.8%). In the VC, only 24.1% of CTBs were found in the same clusters, with the majority in CTB 2 (16.1%). The contribution to cycling clusters was consistent across regions: 22.1% and 17.5% of CTBs in the VC and SC, respectively. The relative proportions of CTB 1–4 in the VC and SC were consistent across individual samples indicating that this difference was not driven by sample variability (Figure 2—figure supplement 1b-c). Furthermore, independent clustering analysis of VC and SC cells identified the same populations as the integrated trophoblast subset. Importantly, independent clustering of each region did not identify CTB 4 in the VC or STB in the SC, suggesting CTB 4 and STB to be specific to the SC and VC, respectively. The small number of cells in the integrated dataset identified as VC CTB 4 (173 cells) or SC STB (14 cells) may be in part an artifact of computational integration (Figure 2—figure supplement 3a-f). Next, we immunolocalized the protein products of genes that distinguished the subpopulations. At the mRNA level, PAGE4 expression was highest in CTB 1 and decreased across CTB 2–4 (Figure 2b and e). In the VC, the CTB monolayer between the fetal stromal villous core and the overlying STB layer showed strong PAGE4 immunoreactivity, which diminished upon differentiation to EVT (Figure 2f - left). PAGE4 mRNA and protein expression matched that of known villous CTB marker CDH1 (Figure 2—figure supplement 4a and b; Genbacev et al., 1997). In the SC, the PAGE4 signal was strong in the epithelial layer directly adjacent to the fetal stroma and then decreased in cells distant from the basal layer, again matching CDH1 (Figure 2f – right, Figure 2—figure supplement 4b – right). Both RNA expression and protein localization were consistent with CTB 1 cells existing in both chorionic regions and occupying a similar niche. Staining for KRT6, a marker highly enriched in CTB 4 cells (Figure 2g) showed a strikingly different result. Cells occupying the upper layers of SC epithelium showed a strong KRT6 signal, a pattern opposite to CDH1 (Figure 2h and i – right). KRT6 was absent from either the floating or anchoring villi of the VC (Figure 2h and i – left), although rare decidual resident KRT6 positive cells were identified in the VC region (Figure 2—figure supplement 4c, Figure 6—figure supplement 1 – top). KRT6 isoforms, KRT6B and KRT6C, were not expressed, confirming KRT6A transcript and protein as highly specific markers of a CTB population found only in the SC (Figure 2—figure supplement 4d). These data describe a novel subpopulation of CTBs unique to the SC, which going forward we term CTB 4 or SC-CTBs for SC-specific CTBs. A common CTB progenitor gives rise to STBs in the VC and SC-CTBs in the SC Next, we investigated the developmental origin of the SC-CTBs. We performed RNA velocity analysis to predict the relationships between cells based on the proportion of exonic and intronic reads. These predictions are shown as vectors representing both the magnitude (predicted rate) and the direction of differentiation (Bergen et al., 2020). We first asked whether RNA velocity could recapitulate the well-established differentiation trajectories of trophoblasts in the VC (Knöfler et al., 2019; Turco et al., 2018; Vento-Tormo et al., 2018). In accordance with previous results, RNA velocity projections identified CTB 1 as the root for three differentiation trajectories: self-renewal, differentiation to STBs, and differentiation to EVTs (Figure 3a). Cells at the boundary of the CTB 1 cluster showed differentiation vectors of high magnitude toward STB Precursors and upregulated canonical drivers of STB differentiation and fusion (ERVW-1 and ERVFRD-1). These cells also expressed transcription factors (GCM1 and HOPX) and hormones (CSH1) necessary for STB function (Figure 3—figure supplement 1a; Baczyk et al., 2009; Mi, 2000; Blaise et al., 2003; Yabe et al., 2016). Figure 3 with 2 supplements see all Download asset Open asset A common cytotrophoblast (CTB) progenitor gives rise to syncytiotrophoblasts (STBs) in the villous chorion (VC) and smooth chorion (SC)-CTBs in the SC. RNA velocity vector projections overlaid on to UMAPs for trophoblast cells isolated from the (a) VC and (b) SC. Arrows denote direction and magnitude is represented by line thickness. (c) Pseudotime reconstruction of SC derived CTB 1–4 clusters from the scVelo dynamical model of latent time. Each column represents one cell. Cells at the left are clustered in CTB 1 and progress through CTB 2, 3, and 4 along the x-axis. Select genes that were the major drivers of the pseudotime alignment are shown on the y-axis. Expression ranged from dark blue (lowest) to yellow (highest). (d) Violin plots of select factors from (c) demonstrated shared or region-specific expression for genes associated with the CTB 4 differentiation trajectory. In the SC, CTB 1 cells once again were identified as the root for differentiation. However, CTB 1 cells showed strong directionality and magnitude toward CTB 2–4 (Figure 3b). All cells in CTB 1–4 clusters displayed uniform directionality indicating a robust differentiation trajectory ending at CTB 4. High levels of transcriptional similarity between CTB 2 and 4 compared to CTB 1 suggested CTB 2 and CTB 3 are intermediate states between CTB 1 and CTB 4 (Figure 2—figure supplement 2c). Mitotic KRT6+ cells were identified, and while the interaction between the cell cycle and differentiation of SC-CTBs remains unclear, these data show that differentiation to SC-CTB does not require cell cycle exit, unlike STBs and EVTs (Figure 3—figure supplement 1b). In contrast to the VC, we observed no velocity vectors with directionality toward the STB lineage from the CTB clusters in the SC samples. Further, the smaller number of STB precursors (460 cells) and STBs (14 cells) exhibited reduced expression of STB canonical markers such as ERVFRD-1 and GCM1, and notably, a near absence of ERVW-1 (Figure 3—figure supplement 1a). These cells may be associated with ghost villi (Benirschke et al., 2006). In sum, these data show differential developmental trajectories for the CTB 1 cells in the SC and VC, with the former largely giving rise to SC-CTBs and the latter to STBs. To identify the genes that were correlated with progression from CTB 1–4 in the SC, we used the velocity vector predictions to construct a pseudotemporal model of differentiation. All the cells in these clusters were plotted in one dimension from the least to the most differentiated according the pseudotime model (Figure 3c). Genes that were highly expressed at the start of the pseudotemporal differentiation included pan-trophoblast factors such as EGFR, which was expressed throughout all four CTB populations in both the VC and SC. Progression along the pseudotime trajectory identified regulators of cell fate and function, including the transcription factor KLF4 and extracellular matrix (ECM) components COL5A1 and LAMA3 (Figure 3c–d); all demonstrated SC-specific expression. Elevated expression of ECM transcripts (COL4A2, FN1) and transcription factors responsive to cell contact and mechanical stress (HES1, YAP1) were coordinately upregulated, potentially highlighting the effects of the extracellular environment on fate specification (Figure 3—figure supplement 1c). To identify differential signaling events that might regulate alternative paths of differentiation in the VC and SC, we used CellPhoneDB to predict receptor-ligand interactions between CTB clusters within each region (Figure 3—figure supplement 2a-b; Efremova et al., 2020). This analysis identified BMP, Notch, and Ephrin signaling events specific to the SC region, which may help to determine cell fate and/or cell sorting within the SC trophoblast epithelium (Figure 3—figure supplement 2b). Together, these data demonstrated that SC-CTBs originate from CTB 1 progenitors common to both the VC and SC. In the SC, instead of upregulating syncytialization factors such as GCM1 and ERVFRD-1, CTB 1 progenitors upregulate transcription factors such as KLF4, YAP1, and HES1, which drive an epithelial cell fate in other contexts (Segre et al., 1999; Harvey et al., 2013; Rock et al., 2011). SC-CTBs express a distinct epidermal transcriptional program Next, we sought a better understanding of the physiological functions of the SC trophoblast clusters. We performed gene ontology analysis as a summary of functional processes (Figure 4a, Figure 4—figure supplement 1; Yu et al., 2012). We focused on the progenitor CTB 1 and terminally differentiated SC-CTBs as they showed enrichment for strikingly different functional categories. In CTB 1, we identified enrichment for WNT signaling, epithelial morphogenesis, and membrane transport, categories commonly associated with progenitors (Figure 4a, Figure 4—figure supplement 1). We validated the activity of WNT signaling and the location of these cells by immunolocalization of non-phosphorylated CTNNB1 (np-CTNNB1). Staining was localized to the most basal epithelial layer nearest to the stroma in both the VC and SC regions (Figure 4b), matching expression of the CTB 1 marker CDH1 (Figure 4—figure supplement 2a). WNT signaling has an important role in the maintenance of villous CTBs in vivo and in the derivation and culture of self-renewing human trophoblast stem cells (Knöfler et al., 2019; Haider et al., 2018; Okae et al., 2018). We investigated proliferation of np-CTNNB1 expressing cells in both regions using KI67 as a mitotic marker. This revealed a similar percentage of KI67+ CTB 1 cells, suggesting similar proliferat