Cell Stem Cell
Available online 24 September 2024
2024 年 9 月 24 日上线
Journal home page for Cell Stem Cell

Article 文章
Marmoset and human trophoblast stem cells differ in signaling requirements and recapitulate divergent modes of trophoblast invasion
狨猴和人类滋养层干细胞的信号传导需求不同,并概括了滋养层入侵的不同模式

医学TOPSCI升级版 医学1区SCI基础版 生物1区IF 19.8 如果19.8
https://doi.org/10.1016/j.stem.2024.09.004 Get rights and content 获取权利和内容

Highlights 亮点

  • Marmoset naive PSCs form extraembryonic mesoderm in human TSC conditions
    狨猴幼稚 PSC 在人类 TSC 条件下形成胚胎外中胚层
  • TGF-β/NODAL, FGF, and WNT control marmoset peri/postimplantation trophoblast identity
    TGF-β/NODAL、FGF 和 WNT 控制狨猴植入周围/植入后滋养层身份
  • Marmoset TSCs recapitulate hallmarks of early embryo attachment and invasion
    狨猴 TSC 重现了早期胚胎附着和侵袭的特征
  • Human TSCs require WNT activation to repress extravillous trophoblast differentiation
    人类 TSC 需要 WNT 激活来抑制绒毛外滋养层分化

Summary 概括

Early human trophoblast development has remained elusive due to the inaccessibility of the early conceptus. Non-human primate models recapitulate many features of human development and allow access to early postimplantation stages. Here, we tracked the pre- to postimplantation transition of the trophoblast lineage in superficially implanting marmoset embryos in vivo. We differentiated marmoset naive pluripotent stem cells into trophoblast stem cells (TSCs), which exhibited trophoblast-specific transcriptome, methylome, differentiation potential, and long-term self-renewal. Notably, human TSC culture conditions failed to support marmoset TSC derivation, instead inducing an extraembryonic mesoderm-like fate in marmoset cells. We show that combined MEK, TGF-β/NODAL, and histone deacetylase inhibition stabilizes a periimplantation trophoblast-like identity in marmoset TSCs. By contrast, these conditions differentiated human TSCs toward extravillous trophoblasts. Our work presents a paradigm to harness the evolutionary divergence in implantation strategies to elucidate human trophoblast development and invasion.
由于早期概念难以接近,人类早期滋养层的发育仍然难以捉摸。非人类灵长类动物模型概括了人类发育的许多特征,并允许进入早期植入后阶段。在这里,我们追踪了体内浅表植入狨猴胚胎中滋养层谱系的植入前到植入后的转变。我们将狨猴幼稚多能干细胞分化为滋养层干细胞(TSC),其表现出滋养层特异性转录组、甲基化组、分化潜力和长期自我更新。值得注意的是,人类 TSC 培养条件无法支持狨猴 TSC 衍生,而是在狨猴细胞中诱导胚胎外中胚层样命运。我们发现,联合 MEK、TGF-β/NODAL 和组蛋白脱乙酰酶抑制可稳定狨猴 TSC 中的植入周围滋养层样特性。相比之下,这些条件使人类 TSC 分化为绒毛外滋养细胞。我们的工作提出了一个范例,利用植入策略的进化差异来阐明人类滋养层的发育和侵袭。

Keywords 关键词

human development
trophoblast
non-human primate trophoblast
trophoblast stem cells
primate trophoblast stem cells
superficial implantation
interstitial implantation
marmoset embryo
marmoset trophoblast stem cells
extraembryonic mesoderm

人类发展
滋养层
非人灵长类滋养层
滋养层干细胞
灵长类滋养层干细胞
浅表植入
间质植入
狨猴胚胎
狨猴滋养层干细胞
胚外中胚层

Introduction 介绍

Embryo implantation and placentation are defining hallmarks of eutherian development. The trophoblast lineage mediates these critical tasks and is essential to establishing the link between the embryo and the maternal tissues.1
胚胎植入和胎盘形成是真兽发育的标志。滋养层谱系介导这些关键任务,对于建立胚胎和母体组织之间的联系至关重要。 1
All trophoblast lineages originate from the trophectoderm, the outer cell layer of the mammalian preimplantation embryo.2,3,4,5 At implantation, the trophectoderm undergoes primary syncytialization to form invasive cells that attach and penetrate the luminal epithelium of the uterus.6,7 In the early stages of human postimplantation development, the trophectoderm differentiates into three major lineages: cytotrophoblast, syncytiotrophoblast, and extravillous trophoblast (EVT). Cytotrophoblast is an undifferentiated and proliferative cell population that, in humans, surrounds the entire conceptus and gives rise to multinucleated syncytiotrophoblast and highly migratory EVT cells.4,8 Primary syncytiotrophoblast are multinucleated cells that arise upon embryo attachment, form lacunae for nutrient exchange, and secrete chorionic gonadotropin (CGA and CGB) to sustain the pregnancy.6,8,9,10,11,12,13,14,15 EVT cells deeply invade into the uterus, where they play a pivotal role in spiral artery remodeling and immune modulation.4
所有滋养层谱系均源自滋养外胚层,即哺乳动物植入前胚胎的外细胞层。 2 3 4 5植入时,滋养外胚层经历初级合胞化,形成附着并穿透子宫腔上皮的侵袭性细胞。 6 7在人类着床后发育的早期阶段,滋养外胚层分化为三个主要谱系:细胞滋养层、合体滋养层和绒毛外滋养层 (EVT)。细胞滋养层是一种未分化和增殖的细胞群,在人类中,它包围整个概念并产生多核合体滋养层和高度迁移的 EVT 细胞。 4 8初级合体滋养层是多核细胞,在胚胎附着时产生,形成用于营养交换的腔隙,并分泌绒毛膜促性腺激素(CGA 和 CGB)以维持妊娠。6 8 9 10 11 12 13 14 15 EVT细胞深入侵入子宫,在螺旋动脉重塑和免疫调节中发挥关键作用。4
Abnormalities in trophoblast development and invasion have profound consequences for maternal and fetal health.16 Insufficient trophoblast invasion can lead to fetal growth restriction and pre-eclampsia,17 while excessive trophoblast invasion causes placenta accreta spectrum disorders.16,18,19 Despite the importance of the placenta for healthy pregnancy outcomes, the molecular mechanisms controlling human trophoblast invasion depth and placental development remain ill-understood due to the inaccessibility of early human postimplantation samples.
滋养层发育和侵袭的异常对母体和胎儿的健康具有深远的影响。 16滋养细胞侵入不足可导致胎儿生长受限和先兆子痫, 17而滋养细胞侵入过度则会导致植入性胎盘谱系疾病。 16 18 19尽管胎盘对于健康妊娠结局非常重要,但由于无法获得早期人类植入后样本,控制人类滋养层侵袭深度和胎盘发育的分子机制仍然不清楚。
The derivation of human trophoblast stem cells (TSCs)20,21,22 has provided an essential tool to delineate the mechanisms of cytotrophoblast self-renewal23 and differentiation.24,25,26 Trophectoderm-like identities can also be induced from naive (preimplantation) pluripotent stem cells (PSCs) by transforming growth factor (TGF)-β/NODAL and fibroblast growth factor (FGF)/ERK inhibition.25,27,28,29,30 Interestingly, recent studies have demonstrated that human trophoblast culture conditions20 promoted not only trophoblast-like identities from naive PSCs but also extraembryonic mesoderm (ExMes)-like cell types, suggesting some degree of overlap in extraembryonic lineage specification.31
人类滋养层干细胞 (TSC) 20 21 22的衍生为描述细胞滋养层自我更新23和分化机制提供了重要工具。 24 25 26还可以通过转化生长因子 (TGF)-β/NODAL 和成纤维细胞生长因子 (FGF)/ERK 抑制从幼稚(植入前)多能干细胞 (PSC) 诱导滋养外胚层样特征。 25 27 28 29 30有趣的是,最近的研究表明,人类滋养层培养条件20不仅促进了幼稚 PSC 的滋养层样特性,而且促进了胚外中胚层 (ExMes) 样细胞类型,表明胚外谱系规范存在一定程度的重叠。 31
Rare in vivo datasets from early pregnancy terminations provide essential insights into postimplantation development,32,33 but trophoblast samples from periimplantation and early gastrulation stages remain elusive. Therefore, the paucity of in vivo-developed early postimplantation transcriptional signatures in human has hampered the validation of in vitro-generated TSCs from naive PSCs.
早期妊娠终止的罕见体内数据集为植入后发育提供了重要的见解, 32 33但来自植入周围和早期原肠胚阶段的滋养层样本仍然难以捉摸。因此,人类体内开发的早期植入后转录特征的缺乏阻碍了体外从初始 PSC 产生的 TSC 的验证。
Non-human primate models are a promising platform to study early postimplantation primate development. In vivo studies in the common marmoset, cynomolgus macaque, and rhesus macaque have illuminated the molecular framework governing early postimplantation embryonic development.34,35,36,37 However, trophoblast formation and placentation are subject to evolutionary pressures and therefore vary across mammalian species.7,11,13,15,38 New World monkeys, Old World monkeys, apes, and humans exhibit increasing degrees of trophoblast invasion, respectively. Embryos of the common marmoset (Callithrix jacchus), a New World monkey, undergo superficial implantation in the central lumen of the uterus. This implantation mode is more shallow and less invasive compared with interstitial implantation in Great apes (orangutan, gorilla, chimpanzee, and human), where the embryo penetrates deeply into the uterus and becomes fully engulfed by the endometrium.4,7,39,40,41 We previously proposed that studying embryo implantation in different primate species may provide a model to tackle trophoblast invasion-related disorders.13 However, TSCs from a New World monkey have remained elusive.
非人类灵长类动物模型是研究早期植入后灵长类动物发育的有前景的平台。对普通狨猴、食蟹猴和恒河猴的体内研究阐明了控制早期植入后胚胎发育的分子框架。 34 35 36 37然而,滋养层的形成和胎盘的形成受到进化压力的影响,因此在不同的哺乳动物物种中存在差异。 7 11 13 15 38新世界猴、旧世界猴、猿和人类分别表现出越来越严重的滋养层侵袭。普通狨猴( Callithrix jacchus )(一种新世界猴)的胚胎在子宫中央腔进行浅层植入。与类人猿(猩猩、大猩猩、黑猩猩和人类)的间质着床相比,这种着床模式更浅、侵入性更小,胚胎深入子宫并被子宫内膜完全吞没。4 7 39 40 41我们之前提出,研究不同灵长类动物的胚胎植入可能提供一个模型来解决滋养层侵袭相关疾病。 13然而,来自新世界猴的 TSC 仍然难以捉摸。
Here, we investigated early marmoset trophoblast development in vivo to delineate the impact of implantation and early invasion on trophoblast identity and to establish a molecular framework for the derivation of marmoset TSCs in vitro. We determined culture conditions to stabilize periimplantation and early postimplantation trophoblast identities and uncovered a human-specific role for WNT signaling. Marmoset TSCs recapitulated primary syncytium formation and invasive EVT-like differentiation, providing insights into the regulators of trophoblast lineage progression. Our results demonstrate that marmoset TSCs are a powerful resource to capture the evolutionary divergence of primate trophoblast development.
在这里,我们研究了狨猴体内早期滋养层发育,以描述植入和早期侵袭对滋养层特性的影响,并建立了狨猴 TSC体外衍生的分子框架。我们确定了稳定植入周围和植入后早期滋养层特性的培养条件,并发现了 WNT 信号传导的人类特异性作用。狨猴 TSC 再现了初级合胞体形成和侵袭性 EVT 样分化,为滋养层谱系进展的调节因子提供了见解。我们的结果表明,狨猴 TSC 是捕捉灵长类滋养层发育进化差异的强大资源。

Results 结果

Marmoset trophoblast development in vivo
狨猴滋养层体内发育

We recently illuminated early marmoset implantation stages by spatial embryo profiling.34 To investigate marmoset trophoblast attachment, invasion, and early placentation, we performed further analysis of specimens from Carnegie stages (CSs)5–734 with a focus on trophoblast development. At CS5, the polar trophectoderm (proximal to the inner cell mass [ICM]) has adhered to the endometrial lining (primary implantation site) and started to break down the luminal epithelium (Figures S1A and S1B). We identified a discontinuous KRT7+ cytotrophoblast layer near the primary implantation site, which was interspersed with CGB-secreting multinucleated syncytiotrophoblast (Figures 1A–1C and S1B). Both cyto- and syncytiotrophoblast were positive for AP2γ (TFAP2C), with syncytiotrophoblast expressing the highest AP2γ levels (Figures 1B, 1C, and S1B). The thickness and density of the AP2γ+ trophoblast layer only marginally increased from CS5 (6–100 μm wide) to CS7 (6–135 μm wide) (Figures 1B, 1C, S1B, and S1C), demonstrating minimal trophoblast proliferation or invasion immediately after implantation. By contrast, human early postimplantation trophoblast proliferates rapidly, increasing in thickness from 3–90 μm at CS5 to 160–830 μm at CS7.42
我们最近通过空间胚胎分析阐明了狨猴的早期植入阶段。 34为了研究狨猴滋养层附着、侵袭和早期胎盘形成,我们对卡内基阶段 (CSs)5-7 34的标本进行了进一步分析,重点关注滋养层发育。在 CS5,极滋养外胚层(靠近内细胞团 [ICM])已粘附到子宫内膜衬里(主要植入部位)并开始分解管腔上皮(图 S1 A 和 S1B)。我们在主要植入部位附近发现了一个不连续的 KRT7+ 细胞滋养层,其中散布着分泌 CGB 的多核合体滋养层(图 1 A-1C 和S1 B)。细胞滋养层和合体滋养层均呈 AP2γ ( TFAP2C ) 阳性,其中合体滋养层表达的 AP2γ 水平最高(图 1 B、1C 和S1 B)。 AP2γ+滋养层的厚度和密度仅从CS5(6-100μm宽)到CS7(6-135μm宽)略有增加(图1B 、1C、 S1B和S1C),表明滋养层增殖最小或植入后立即发生侵袭。相比之下,人类早期着床后滋养层增殖迅速,厚度从CS5时的3-90μm增加到CS7时的160-830μm。42
  1. Download: Download high-res image (3MB)
    下载:下载高分辨率图像 (3MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 1. Postimplantation marmoset trophoblast exhibits shallow invasion
图1 。着床后狨猴滋养层表现出浅层侵袭

(A) Immunofluorescence (IF) staining of CS7 marmoset primary (1°) and secondary (2°) implantation sites. Tile-scanned images were automatically merged by the acquisition software.
(A) CS7 狨猴初级 (1°) 和次级 (2°) 植入位点的免疫荧光 (IF) 染色。平铺扫描图像由采集软件自动合并。
(B) Expanded view of (A). Scale bars are 200 μm.
(B) (A) 的扩展视图。比例尺为 200 μm。
(C) IF of CS5 implantation site. Scale bars are 100 μm.
(C) CS5 植入部位的 IF。比例尺为 100 μm。
(D) IF of contact point between luminal trophoblast of twin embryos.
(D) 双胚胎管腔滋养层之间接触点的 IF。
(E) Expanded view of (D). Scale bars are 50 μm.
(E) (D) 的扩展视图。比例尺为 50 μm。
(F) UMAP plot of trophoblast lineages. Tb_CS3: trophectoderm; Tb_CS5, Tb_CS6, and Tb_CS7: postimplantation embryonic trophoblast; Tb_abembryonic: postimplantation abembryonic trophoblast.
(F) 滋养层谱系的 UMAP 图。 Tb_CS3:滋养外胚层; Tb_CS5、Tb_CS6 和 Tb_CS7:植入后胚胎滋养层; Tb_abembryonic:植入后胚胎滋养层。
(G) UMAP plot from (F) showing normalized log expression.
(G) (F) 中的 UMAP 图显示归一化的对数表达。
(H) MA plot of differentially expressed genes (DEGs) between trophectoderm (Tb_CS3) and postimplantation trophoblast (Tb_Post).
(H) 滋养外胚层 (Tb_CS3) 和植入后滋养层 (Tb_Post) 之间差异表达基因 (DEG) 的 MA 图。
(I) Violin plots of normalized mRNA counts in trophoblast lineages.
(I) 滋养层谱系中标准化 mRNA 计数的小提琴图。
See also Figure S1.
另请参见图 S1
At CS5, the mural trophectoderm (abembryonic trophoblast, distal from the ICM) expanded up to 2 mm to fill the uterine cavity and made loose contact with the opposing uterine wall (Figures S1A and S1B). By CS6, attached mural trophectoderm breached the luminal epithelium and established a secondary implantation site on the uterine wall opposite to the embryo (Figures 1A and 1B). AP2γ+ syncytiotrophoblast appeared at the secondary implantation site (14 μm wide) by CS7 but was thinner compared with the primary implantation site (35 μm wide) (Figures S1C and S1D).11,34 In most Old World and New World monkeys, including marmosets, the endometrial epithelium is remodeled to form epithelial plaques at the implantation site.7,11,34,43,44 This process, which is not observed in Great apes, has been compared with decidualization in human.10,43 Interestingly, both primary and secondary embryo implantation induced widespread epithelial plaque reaction, transforming SOX17+ endometrial glands into rounded, densely packed clusters of KRT7+, SOX17− epithelial plaque cells (Figures 1A–1C).
在 CS5 时,壁滋养外胚层(胚胎滋养层,远离 ICM)扩张至 2 毫米以填充子宫腔,并与相对的子宫壁形成松散接触(图 S1 A 和 S1B)。到CS6时,附着的壁滋养外胚层突破管腔上皮,并在与胚胎相对的子宫壁上建立了二次植入位点(图1A和1B)。 AP2γ+合体滋养层出现在CS7的二次植入部位(14μm宽),但比初次植入部位(35μm宽)更薄(图S1C和S1D)。 11 34在大多数旧世界和新世界猴中,包括狨猴,子宫内膜上皮会被重塑,在植入部位形成上皮斑块。 7 11 34 43 44这一过程在类人猿中没有观察到,但已与人类的蜕膜化进行了比较。10 43有趣的是,初级和次级胚胎植入均诱导了广泛的上皮斑块反应,将 SOX17+ 子宫内膜腺体转化为圆形、密集的 KRT7+、SOX17− 上皮斑细胞簇(图 1 A-1C)。
Outside the primary and secondary implantation sites, the mural trophectoderm remained unattached throughout CS5–7. Luminal epithelium adjacent to unattached trophoblast maintained its columnar morphology (Figure 1B). We refer to this unattached postimplantation trophoblast as luminal trophoblast. In one specimen with a twin pregnancy, we were able to obtain cross-sections at the border between two implanting blastocysts (Figures 1D and 1E). The luminal trophoblast consisted of AP2γ+/KRT7+ epithelial cells, which were clearly distinguishable from PDGFRα+/AP2γ−/KRT7− ExMes lining the luminal trophoblast (Figure 1E).
在初次和二次植入位点之外,在整个 CS5-7 期间,壁滋养外胚层保持不附着。与未附着的滋养层相邻的管腔上皮保持其柱状形态(图1B )。我们将这种独立的植入后滋养层称为管腔滋养层。在一个双胎妊娠的标本中,我们能够获得两个植入囊胚之间边界的横截面(图 1D和 1E)。管腔滋养层由 AP2γ+/KRT7+ 上皮细胞组成,这些细胞与管腔滋养层内衬的 PDGFRα+/AP2γ−/KRT7− ExMes 明显区分(图 1 E)。
Transcriptome analysis of the pre- to postimplantation transition showed that trophectoderm samples separated from postimplantation trophoblast (Figure 1F). Examination of the postimplantation trophoblast cluster demonstrated no obvious sub-clustering along developmental time (Figure 1F). We identified NOTO, JAM2, and FABP3 as enriched in the trophectoderm, while postimplantation trophoblast samples predominantly upregulated MLLT1, PRR9, CGB3, and CGA (Figures 1G and 1H). Gene Ontology (GO)-term analysis of the pre- to postimplantation transition showed differences in WNT signaling and basement membrane organization (Figure S1E).
植入前到植入后转变的转录组分析表明,滋养外胚层样品与植入后滋养层分离(图1F )。对植入后滋养层簇的检查表明,随着发育时间的推移,没有明显的亚簇(图 1 F)。我们发现NOTOJAM2FABP3在滋养外胚层中富集,而植入后滋养层样本主要上调MLLT1PRR9CGB3CGA1G 和 1H)。对植入前到植入后过渡的基因本体 (GO) 术语分析显示 WNT 信号传导和基底膜组织存在差异(图 S1 E)。
To identify postimplantation trophoblast lineages, we annotated trophoblast samples according to their location and transcriptomic signature (see STAR Methods). We analyzed spatially distinct trophoblast lineages: trophectoderm (preimplantation trophoblast of the blastocyst), embryonic trophoblast (postimplantation, attached trophoblast at the primary implantation site), abembryonic trophoblast (postimplantation, attached trophoblast at the secondary implantation site), and luminal trophoblast (postimplantation, unattached trophoblast) (Figures 1I and S1F–S1H). Core trophoblast factors GATA2, GATA3, and KRT7 were expressed throughout all trophoblast lineages (Figure 1I). The marmoset trophectoderm expressed low levels of pluripotency markers POU5F1, SOX2, and NANOG, which were sharply downregulated in all postimplantation trophoblast samples (Figure 1I). Conversely, syncytiotrophoblast-associated genes such as CGA and CGB3 were upregulated upon implantation (Figures 1H and 1I).23 Human cytotrophoblast-associated transcripts MSX2 and OVOL1 were highest in luminal trophoblast (Figures 1I and S1H). Abembryonic trophoblast samples expressed nearly all postimplantation markers, including MLLT1, NR2F2, CGA, and CGB3, albeit at lower levels than embryonic trophoblast (Figures 1I and S1F). This observation is in line with the delay of trophoblast invasion at the secondary implantation site (Figures 1A and 1B). The interspersed nature of cytotrophoblast and syncytiotrophoblast made it difficult to discriminate between the two lineages. To overcome this challenge, we performed further hierarchical clustering of embryonic trophoblast, which revealed two closely related cell populations (Figure S1I). Differential expression analysis between the two clusters identified them as cytotrophoblast (KRT7+, CDH1+) and syncytiotrophoblast (CGB3+, MLLT1+) (Figure S1J).
为了识别植入后滋养层谱系,我们根据滋养层样本的位置和转录组特征对滋养层样本进行了注释(参见STAR 方法)。我们分析了空间上不同的滋养层谱系:滋养外胚层(囊胚的植入前滋养层)、胚胎滋养层(植入后,在初次植入部位附着的滋养层)、胚胎滋养层(植入后,在二次植入部位附着的滋养层)和管腔滋养层(植入后,未附着的滋养层)(图 1 I 和S1 F–S1H)。核心滋养层因子GATA2GATA3KRT7在所有滋养层谱系中均有表达(图 1 I)。狨猴滋养外胚层表达低水平的多能性标记物POU5F1SOX2NANOG ,这些标记物在所有植入后滋养层样本中均急剧下调(图 1 I)。相反,合体滋养层相关基因(例如CGACGB3)在植入时上调(图 1H和 1I)。 23人细胞滋养层相关转录物MSX2OVOL1在管腔滋养层中最高(图 1 I 和S1 H)。 胚胎滋养层样本表达几乎所有植入后标记,包括MLLT1NR2F2CGACGB3 ,尽管水平低于胚胎滋养层(图 1 I 和S1 F)。这一观察结果与二次植入部位滋养层侵袭的延迟一致(图 1A和 1B)。细胞滋养层和合体滋养层的分散性质使得很难区分这两个谱系。为了克服这一挑战,我们对胚胎滋养层进行了进一步的层次聚类,揭示了两个密切相关的细胞群(图S1 I)。两个簇之间的差异表达分析将其鉴定为细胞滋养层(KRT7+、CDH1+)和合体滋养层(CGB3+、MLLT1+)(图S1 J)。
We conclude that marmoset trophectoderm expresses residual levels of core pluripotency factors that are extinguished upon implantation. Postimplantation trophoblast gives rise to a discontinuous layer of AP2γ (TFAP2C)+/KRT7+ cytotrophoblast and syncytium at the implantation sites, which can be readily demarcated by transcription factors MLLT1, MSX2, and OVOL1, as well as pregnancy hormones CGA and CGB3.
我们得出的结论是,狨猴滋养外胚层表达残留水平的核心多能性因子,这些因子在植入后消失。植入后滋养层在植入位点产生不连续的 AP2γ ( TFAP2C )+/KRT7+ 细胞滋养层和合胞体层,可以通过转录因子MLLT1MSX2OVOL1以及妊娠激素CGACGB3轻松划分。

Marmoset cells differentiate to ExMes in OK conditions
狨猴细胞在正常条件下分化为 ExMes

To investigate developmental trajectories of marmoset trophoblast upon implantation, we sought to establish in vitro models of peri- and postimplantation trophoblast. Human naive PSCs can be differentiated into extraembryonic lineages including trophoblast, hypoblast, and ExMes.25,27,28,29,30,31,45 The activation of WNT/EGF and inhibition of TGF-β/HDAC/ROCK in human TSC (OK) medium20 promotes a trophoblast-like state in naive human PSCs.25,28 We recently reported the establishment of preimplantation epiblast-like marmoset naive PSCs.34 To examine whether human trophoblast conditions can generate marmoset TSCs from naive PSCs, we switched culture conditions to human trophoblast (OK) medium20 (Figure 2A). The majority of dome-shaped naive PSC colonies flattened into epithelial colonies with varying morphologies at day 5 (Figures 2B and S2A). By passage 2, the dome-shaped colonies disappeared, resulting in a homogeneous culture of flat epithelial colonies (marmoset OK cells) that expressed trophoblast markers, including CDX2 and TFAP2C, and downregulated the core pluripotency factor OCT4 (Figures 2C, 2D, and S2B). However, marmoset OK cells lacked expression of core trophoblast markers, including GATA2, GATA3, and CGA. Moreover, we observed upregulation of hypoblast and ExMes lineage markers GATA6 and PDGFRα (Figure S2B). Immunofluorescence of marmoset embryos in vivo demonstrated exclusive expression of PDGFRα and CDH11 within the ExMes (Figures 1E and 2E).34 Marmoset OK cells expressed both PDGFRα and CDH11 and lacked expression of KRT7, in contrast to human OK TSCs20 (Figures 2F and 2G). These findings suggest that marmoset OK cells exhibit features of ExMes.
为了研究狨猴滋养层在植入后的发育轨迹,我们试图建立植入前后滋养层的体外模型。人类初始 PSC 可分化为胚胎外谱系,包括滋养层、下胚层和 ExMes。 25 27 28 29 30 31 45人 TSC (OK) 培养基中 WNT/EGF 的激活和 TGF-β/HDAC/ROCK 的抑制20可促进初始人 PSC 的滋养层样状态。 25 28我们最近报道了植入前外胚层样狨猴幼稚 PSC 的建立。 34为了检查人滋养层条件是否可以从初始 PSC 生成狨猴 TSC,我们将培养条件切换为人滋养层 (OK) 培养基20图 2 A)。 第 5 天时,大多数圆顶形初始 PSC 集落变平为具有不同形态的上皮集落(图 2 B 和S2 A)。到第 2 代时,圆顶形集落消失,形成扁平上皮集落(狨猴 OK 细胞)的均质培养物,表达滋养层标记物,包括 CDX2 和TFAP2C ,并下调核心多能性因子 OCT4(2C、2D 和S2 B)。然而,狨猴 OK 细胞缺乏核心滋养层标志物的表达,包括GATA2GATA3CGA 。此外,我们观察到下胚层和 ExMes 谱系标记GATA6PDGFRα的上调(图 S2 B)。体内狨猴胚胎的免疫荧光显示 ExMes 中 PDGFRα 和 CDH11 的专有表达(图 1 E 和2 E)。 34狨猴 OK 细胞表达 PDGFRα 和 CDH11,但缺乏 KRT7 表达,与人 OK TSC 20不同(2F 和 2G)。这些发现表明狨猴 OK 细胞表现出 ExMes 的特征。
  1. Download: Download high-res image (3MB)
    下载:下载高分辨率图像 (3MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 2. Human TSC culture conditions induce ExMes in marmoset naive PSCs
图2 .人类 TSC 培养条件在狨猴幼稚 PSC 中诱导 ExMes

(A) Schematic of naive marmoset PSC differentiation in OK conditions.
(A) 幼稚狨猴 PSC 在正常条件下分化的示意图。
(B) Phase contrast image of naive marmoset PSCs.
(B) 幼稚狨猴 PSC 的相差图像。
(C) Phase contrast image of passage 2 OK-derived cells.
(C) 第 2 代 OK 衍生细胞的相差图像。
(D) IF of PSCs and passage 2 OK-derived cells. Scale bars are 50 μm.
(D) PSC 和第 2 代 OK 衍生细胞的 IF。比例尺为 50 μm。
(E–G) IF of (E) in vivo CS5 implanting marmoset embryo and (F and G) OK differentiated marmoset and human cells. Scale bars are 100 μm (E and F) and 200 μm (G).
(E–G) (E)体内CS5 植入狨猴胚胎的 IF 和 (F 和 G) OK 分化的狨猴和人类细胞。比例尺为 100 μm(E 和 F)和 200 μm(G)。
(H) PCA of in vivo marmoset dataset and in vitro cells. EPI, preimplantation epiblast; HYPO, hypoblast; EmDisc, embryonic disc; VE, visceral endoderm.
(H)体内狨猴数据集和体外细胞的 PCA。 EPI,植入前外胚层; HYPO,下胚层; EmDisc,胚胎盘; VE,内脏内胚层。
(I and J) Violin plot of normalized expression of in vivo and in vitro lineages.
(I 和 J)体内体外谱系标准化表达的小提琴图。
(K) Spatial identity mapping of marmoset naive PSCs and OK cells to the embryo.
(K) 狨猴幼稚 PSC 和 OK 细胞与胚胎的空间同一性图谱。
See also Figure S2.
另请参见图 S2
To further interrogate the developmental identity of marmoset OK cells, we performed full-length single-cell transcriptome profiling and integrated OK cells into our updated marmoset embryo datasets (Figure 1).34,46 Global dimensionality reduction methods revealed that OK cells clustered closest with postimplantation ExMes cells (Figures 2H and S2C). Global correlation analysis showed that marmoset OK cells had the highest similarity with ExMes at CS6 and expressed known human ExMes markers (Figures S2D and S2E).31 Consistent with RT-qPCR data, OK cells expressed low levels of trophoblast markers TFAP2C and KRT7 in comparison to pre- and postimplantation trophoblast lineages in vivo (Figure 2I). Notably, CDX2 is highly expressed in both trophectoderm and ExMes. OK cells upregulated CDX2 and various other ExMes markers, including CDH11, GATA6, and VIM (Figures 2I and 2J). To establish the regional correlation of OK single-cell transcriptomes to the embryo, we performed spatial identity mapping to the embryo (Figure 2K). Naive PSCs correlated to the preimplantation epiblast, as previously reported.34 By contrast, OK cells showed the highest spatial correlation scores with postimplantation ExMes at CS5 and CS6 (Figure 2K). These results are consistent with the recent finding that human naive PSCs can generate both trophoblast and ExMes in OK conditions.31 Interestingly, a small number of OK cells clustered closely with postimplantation trophoblast (Figure 2H). To determine if OK conditions promoted a trophoblast phenotype during early differentiation of naive marmoset cells, we monitored KRT7 expression during the first 18 days of OK differentiation (Figures S2F and S2G). We observed that small KRT7+ subpopulations arose by day 12, which were lost in mature OK cultures (day 50+) (Figures S2F and S2G). Thus, human OK TSC medium does not allow direct TSC derivation from naive PSCs in marmoset and promotes ExMes lineage entry instead.
为了进一步探究狨猴 OK 细胞的发育特性,我们进行了全长单细胞转录组分析,并将 OK 细胞整合到我们更新的狨猴胚胎数据集中(图 1 )。 34 46全局降维方法显示 OK 细胞与植入后 ExMes 细胞聚类最接近(图 2 H 和S2 C)。全局相关性分析表明,狨猴 OK 细胞与 CS6 的 ExMes 具有最高的相似性,并表达已知的人类 ExMes 标记(图 S2D和 S2E)。 31与 RT-qPCR 数据一致,与体内植入前和植入后滋养层谱系相比,OK 细胞表达低水平的滋养层标记物TFAP2CKRT7图 2 I)。值得注意的是, CDX2在滋养外胚层和 ExMe 中均高度表达。 OK 细胞上调CDX2和各种其他 ExMes 标记物,包括CDH11GATA6VIM图 2I和 2J)。为了建立 OK 单细胞转录组与胚胎的区域相关性,我们对胚胎进行了空间同一性映射(图 2 K)。 如先前报道,幼稚 PSC 与植入前外胚层相关。 34相比之下,OK 细胞在 CS5 和 CS6 处与植入后 ExMe 表现出最高的空间相关性得分(图 2 K)。这些结果与最近的发现一致,即人类初始 PSC 在正常条件下可以产生滋养层和 ExMe。 31有趣的是,少数 OK 细胞与植入后滋养层紧密聚集(图 2 H)。为了确定 OK 条件是否在幼稚狨猴细胞的早期分化过程中促进滋养层表型,我们在 OK 分化的前 18 天监测了 KRT7 的表达(图 S2 F 和 S2G)。我们观察到,在第 12 天时出现了小型 KRT7+ 亚群,这些亚群在成熟的 OK 培养物(第 50 天以上)中消失了(图 S2 F 和 S2G)。因此,人类 OK TSC 培养基不允许从狨猴中的幼稚 PSC 直接衍生 TSC,而是促进 ExMes 谱系进入。

Combined WNT, NODAL, and FGF/ERK inhibition is required for marmoset TSCs
狨猴 TSC 需要结合 WNT、NODAL 和 FGF/ERK 抑制

Human OK TSC medium stimulates WNT and EGF-signaling and inhibits TGF-β/NODAL-signaling.20,22 Moreover, trophectoderm fate can be readily induced in human naive PSCs by inhibition of FGF/ERK and TGF-β/NODAL signaling in the absence of WNT stimulation.27,29 In the embryo, we observed dynamic changes of WNT signaling in the pre- to postimplantation transition of marmoset trophoblast development (Figure S1E). To identify the signaling requirements for trophoblast identity, we modulated WNT and FGF/ERK in OK medium during marmoset TSC derivation from naive PSCs on inactivated mouse embryonic fibroblast (MEFs) (Figure 3A). OK with the WNT inhibitor XAV939 (OK XAV), OK with the FGF/ERK inhibitor PD0325901 (OK PD), and OK with WNT and FGF/ERK inhibition (OK XAV PD) each induced flattening and epithelialization of naive PSCs after 5 days (Figure 3B). Immunocytochemistry revealed that OK XAV retained a proportion of OCT4+ cells and lacked KRT7 expression (Figures 3C, S3A, and S3B). OK PD efficiently downregulated OCT4 and mildly upregulated KRT7 (Figures 3C, S3A, and S3B), suggesting that mitogen-activated protein kinase kinase (MEK) inhibition partially suppresses ExMes differentiation. OK XAV PD exhibited high levels of KRT7, expressed CDX2, and was negative for OCT4 (Figures 3C, S3A, and S3B). Considering that KRT7 was strongly expressed in trophoblast in vivo, we decided to further investigate OK XAV PD cells.
人 OK TSC 培养基刺激 WNT 和 EGF 信号传导并抑制 TGF-β/NODAL 信号传导。 20 22此外,在缺乏 WNT 刺激的情况下,通过抑制 FGF/ERK 和 TGF-β/NODAL 信号传导,可以很容易地在人初始 PSC 中诱导滋养外胚层命运。 27 29在胚胎中,我们观察到狨猴滋养层发育的植入前到植入后转变过程中 WNT 信号传导的动态变化(图 S1 E)。为了确定滋养层身份的信号传导要求,我们在从灭活的小鼠胚胎成纤维细胞 (MEF) 上的幼稚 PSC 衍生狨猴 TSC 的过程中,在 OK 培养基中调节 WNT 和 FGF/ERK(图 3 A)。使用 WNT 抑制剂 XAV939(OK XAV)、使用 FGF/ERK 抑制剂 PD0325901(OK PD)以及使用 WNT 和 FGF/ERK 抑制(OK XAV PD)均能在 5 天后诱导初始 PSC 的扁平化和上皮化(图 3 B)。免疫细胞化学显示 OK XAV 保留了一部分 OCT4+ 细胞,并且缺乏 KRT7 表达(图 3CS3A和 S3B)。 OK PD 有效下调 OCT4 并轻度上调 KRT7(图 3CS3A和 S3B),表明丝裂原激活蛋白激酶激酶 (MEK) 抑制部分抑制 ExMes 分化。 OK XAV PD 表现出高水平的 KRT7,表达 CDX2,并且 OCT4 呈阴性(图 3CS3A和 S3B)。考虑到KRT7在体内的滋养层中强烈表达,我们决定进一步研究OK XAV PD细胞。
  1. Download: Download high-res image (3MB)
    下载:下载高分辨率图像 (3MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 3. Additional WNT and FGF/ERK inhibition stabilizes postimplantation trophoblast identity in marmoset
图3 .额外的 WNT 和 FGF/ERK 抑制可稳定狨猴植入后滋养层的特性

(A) Schematic of naive marmoset PSC differentiation in variations of OK conditions.
(A) 不同 OK 条件下幼稚狨猴 PSC 分化的示意图。
(B) Phase contrast images of naive PSCs and differentiated cells at passage 2 in indicated conditions. Scale bars are 100 μm.
(B) 在指定条件下第 2 代的初始 PSC 和分化细胞的相差图像。比例尺为 100 μm。
(C) IF of marmoset cells in indicated conditions.
(C) 狨猴细胞在指定条件下的 IF。
(D) IF of passage 2 OK and OK XAV PD-derived cells cultured on MEFs. MEFs constitutively express PDGFRα.
(D) 在 MEF 上培养的第 2 代 OK 和 OK XAV PD 衍生细胞的 IF。 MEF 组成型表达 PDGFRα。
(E) Spatial identity mapping of marmoset postTSCs to the embryo.
(E) 狨猴后 TSC 到胚胎的空间同一性映射。
(F) PCA of in vivo marmoset dataset and in vitro cells. EPI, preimplantation epiblast; HYPO, hypoblast; EmDisc, embryonic disc; VE, visceral endoderm.
(F)体内狨猴数据集和体外细胞的 PCA。 EPI,植入前外胚层; HYPO,下胚层; EmDisc,胚胎盘; VE,内脏内胚层。
(G) Violin plot of normalized expression in in vivo and in vitro lineages.
(G)体内体外谱系标准化表达的小提琴图。
(H) IF of postTSC differentiation with forskolin.
(H) 使用毛喉素进行 TSC 分化后的 IF。
(I) Quantification of multinucleated cell frequency. Each point represents a single frame where the proportion of multinucleated to single-nucleated cells was counted. Significance was calculated using a two-tailed Mann-Whitney test. n = 3. Error bars represent mean + SD (p ≤ 0.05). FK, forskolin.
(I) 多核细胞频率的量化。每个点代表一个帧,其中计算了多核细胞与单核细胞的比例。使用双尾曼-惠特尼检验计算显着性。 n = 3。误差线代表平均值 + SD ( * p ≤ 0.05)。 FK,毛喉素。
(J) Live imaging of tagged marmoset TSCswith nuclear GFP (red) and F-actin binding LifeACT (green). Images were taken every 30 min.
(J) 带有核 GFP(红色)和结合 F-肌动蛋白的 LifeACT(绿色)的标记狨猴 TSC 的实时成像。每 30 分钟拍摄一次图像。
See also Figure S3, Table S2, and Video S1.
另请参见图 S3表 S2视频 S1
Marmoset OK XAV PD cells could be stably maintained for at least 14 passages and were readily established in a second, independent marmoset PSC line (Figure S3C). In contrast to OK cells, OK XAV PD cells did not express the ExMes marker PDGFRα and upregulated the trophoblast marker AP2γ (TFAP2C) (Figure 3D). Hence, OK XAV PD cells recapitulated the hallmarks of TSCs, exhibiting long-term self-renewing capability and lineage markers expression specific to human TSCs and marmoset cytotrophoblast.
狨猴 OK XAV PD 细胞可以稳定维持至少 14 代,并且很容易在第二个独立的狨猴 PSC 系中建立(图 S3 C)。与 OK 细胞相反,OK XAV PD 细胞不表达 ExMes 标记物 PDGFRα,并上调滋养层标记物 AP2γ ( TFAP2C )(图 3D )。因此,OK XAV PD 细胞重现了 TSC 的特征,表现出长期自我更新能力和人类 TSC 和狨猴细胞滋养层特有的谱系标记表达。
To assess the developmental authenticity of OK XAV PD cells, we performed single-cell transcriptome profiling. Spatial identity mapping and integrated analysis showed that OK XAV PD cells did not correspond to ExMes but were most similar to postimplantation trophoblast in the marmoset embryo (Figures 3E, 3F, and S3D). Therefore, we denoted OK XAV PD cells as postimplantation TSCs (postTSCs). Notably, ExMes-associated genes, including CDH11, GATA6, and VIM, were extinguished in postTSCs, while trophoblast factors KRT7 and TFAP2C were expressed at the same levels as in the postimplantation trophoblast of the marmoset conceptus (Figure 3G). CDX2 levels were reduced compared with OK cells, ExMes, and preimplantation trophoblast (trophectoderm), but similar to postimplantation trophoblast. Correlation analysis showed that marmoset postTSCs corresponded to postimplantation trophoblast at CS6 (Figure S3E), similar to human TSCs.23,47
为了评估 OK XAV PD 细胞发育的真实性,我们进行了单细胞转录组分析。空间同一性图谱和整合分析表明,OK XAV PD 细胞与 ExMe 不对应,但与狨猴胚胎中的植入后滋养层最相似(3E、3F 和S3D )。因此,我们将 OK XAV PD 细胞称为植入后 TSC(postTSC)。值得注意的是,ExMes相关基因,包括CDH11GATA6VIM ,在后TSC中消失,而滋养层因子KRT7TFAP2C的表达水平与狨猴植入后滋养层相同(图3G )。与 OK 细胞、ExMes 和植入前滋养层(滋养外胚层)相比, CDX2水平降低,但与植入后滋养层相似。相关分析表明,狨猴后 TSC 对应于 CS6 处的植入后滋养层(图 S3 E),与人类 TSC 相似。 23 47
A defining hallmark of trophoblast cells is their capacity to form syncytiotrophoblast.3,38,48 Human TSCs are capable of differentiation into a multinucleated syncytium upon cyclic AMP (cAMP) stimulation with forskolin.20 Equally, marmoset postTSCs formed multinucleated cysts in the presence of forskolin (Figures 3H and 3I). Low cellular densities also promoted syncytium formation (Figures 3I and S3F). Moreover, marmoset syncytiotrophoblast genes were enriched in a subset of sequenced postTSCs, suggesting a propensity for syncytiotrophoblast differentiation in postTSCs (Figure S3G). To determine the mechanism by which multinucleated syncytium is formed in vitro, we tracked syncytium formation by live-cell imaging. We generated nuclear GFP and F-actin mCherry-labeled PSCs (nuclear localization sequence (NLS)-GFP LIFEACT) and derived marmoset postTSCs. Time lapse imaging showed that marmoset postTSCs predominantly formed syncytium via endoreduplication (Figures 3J and S3H; Video S1), where mitosis occurred without cytokinesis. On rare occasions, cells would surround the nucleus of another cell, suggesting that cellular fusion may also occur during marmoset syncytium formation (Figure S3I).
滋养层细胞的一个明确标志是它们形成合体滋养层的能力。 3 38 48人类 TSC 在毛喉素的环 AMP (cAMP) 刺激下能够分化为多核合胞体。 20同样,狨猴 TSC 后在毛喉素存在下形成多核囊肿(图 3 H 和 3I)。低细胞密度也促进合胞体形成(图 3 I 和S3 F)。此外,狨猴合体滋养层基因在已测序的 postTSC 子集中富集,表明 postTSC 中有合体滋养层分化的倾向(图 S3 G)。为了确定多核合胞体在体外形成的机制,我们通过活细胞成像追踪合胞体的形成。我们生成了核 GFP 和 F-肌动蛋白 mCherry 标记的 PSC(核定位序列 (NLS)-GFP LIFEACT)和衍生的狨猴 postTSC。延时成像显示狨猴后 TSC 主要通过核内复制形成合胞体(图 3 J 和S3 H;视频 S1 ),其中发生有丝分裂而没有胞质分裂。 在极少数情况下,细胞会包围另一个细胞的细胞核,这表明在狨猴合胞体形成过程中也可能发生细胞融合(图S3 I)。
Download: Download video (572KB)

Video S1. postTSCs form syncytium by endoreduplication, related to Figures 3J, S3H, and S3I. Live IF imaging of marmoset postTSCs tagged with GFP with a nuclear localisation tag (red) and stained with F-actin binding LifeACT (green).

We conclude that marmoset postTSCs closely resemble early postimplantation trophoblast in vivo, are capable of long-term self-renewal, and give rise to syncytiotrophoblast.
我们得出的结论是,狨猴后TSCs与体内早期植入后滋养层非常相似,能够长期自我更新,并产生合体滋养层。

PAVS stabilizes a periimplantation trophectoderm-like state
PAVS 稳定植入周围滋养外胚层样状态

Human naive PSCs are capable of differentiation into trophectoderm-like cells upon FGF/ERK and TGF-β/NODAL inhibition.27,29 To test whether a developmentally earlier, preimplantation trophectoderm-like state could be stabilized, we cultured marmoset naive PSCs in media containing the MEK inhibitor PD0325901, two TGF-β/NODAL inhibitors (A83-01 and SB431542), and valproic acid (VPA), termed PAVS (Figure 4A). Naive PSCs in PAVS medium differentiated into flat, epithelial colonies, which could be sustained in culture for more than 15 passages on MEFs (Figure 4B). Immunofluorescence showed that cells cultured in PAVS medium expressed high levels of trophoblast markers KRT7, CDX2, and GATA3 in the absence of OCT4 (Figures 4C and 4D). Consistent with this trophoblast lineage marker profile, PAVS cultures occasionally formed multinucleated syncytium that expressed the syncytiotrophoblast marker CGB (Figure S4A). We confirmed trophoblast induction in PAVS medium in a second independent marmoset PSC line.
人类初始 PSC 在 FGF/ERK 和 TGF-β/NODAL 抑制后能够分化为滋养外胚层样细胞。 27 29为了测试发育早期、植入前滋养外胚层样状态是否可以稳定,我们在含有 MEK 抑制剂 PD0325901、两种 TGF-β/NODAL 抑制剂(A83-01 和 SB431542)和丙戊酸( VPA),称为 PAVS(图 4 A)。 PAVS 培养基中的初始 PSC 分化为扁平的上皮集落,可以在 MEF 上持续培养超过 15 代(图 4 B)。免疫荧光显示,在不存在 OCT4 的情况下,在 PAVS 培养基中培养的细胞表达高水平的滋养层标记物 KRT7、CDX2 和 GATA3(图 4C和 4D)。与该滋养层谱系标记谱一致,PAVS 培养物偶尔会形成表达合体滋养层标记 CGB 的多核合胞体(图 S4 A)。我们在第二个独立的狨猴 PSC 系中证实了 PAVS 培养基中滋养层的诱导。
  1. Download: Download high-res image (2MB)
    下载:下载高分辨率图像 (2MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 4. PAVS promotes a periimplantation trophoblast phenotype
图4 . PAVS 促进植入周围滋养层表型

(A) Schematic of naive marmoset PSC differentiation in PAVS.
(A) 幼稚狨猴 PSC 在 PAVS 中分化的示意图。
(B) Phase contrast images of naive PSCs and differentiated cells at day 5 and passage 15 in PAVS. Scale bars are 100 μm.
(B) PAVS 中第 5 天和第 15 代时幼稚 PSC 和分化细胞的相差图像。比例尺为 100 μm。
(C) IF of marmoset cells in indicated conditions. Scale bars are 100 μm.
(C) 狨猴细胞在指定条件下的 IF。比例尺为 100 μm。
(D) IF of multinucleated cells in PAVS. Scale bars are 100 μm.
(D) PAVS 中多核细胞的 IF。比例尺为 100 μm。
(E) PCA of in vivo marmoset dataset and in vitro cells. EPI, preimplantation epiblast; HYPO, hypoblast; EmDisc, embryonic disc; VE, visceral endoderm.
(E)体内狨猴数据集和体外细胞的 PCA。 EPI,植入前外胚层; HYPO,下胚层; EmDisc,胚胎盘; VE,内脏内胚层。
(F) Spatial identity mapping of marmoset periTSCs to the embryo.
(F) 狨猴 periTSC 到胚胎的空间同一性映射。
(G) Violin plot of normalized expression in in vivo and in vitro lineages.
(G)体内体外谱系标准化表达的小提琴图。
(H) Schematic of cross-species chimeras with mouse embryos and marmoset periTSCs.
(H) 小鼠胚胎和狨猴 periTSC 的跨物种嵌合体示意图。
(I) IF of E4.5 chimeric mouse embryos aggregated with tagged periTSCs. Scale bars are 50 μm.
(I) E4.5 嵌合小鼠胚胎与标记的 periTSC 聚集的 IF。比例尺为 50 μm。
(J) Quantification of lineage contribution of periTSCs in chimeric mouse embryos (n = 44). Embryos were collected from 7 different mice.
(J) 嵌合小鼠胚胎中 periTSC 谱系贡献的量化 ( n = 44)。从 7 只不同的小鼠身上收集胚胎。
(K) Relative methylation of the ELF5 locus in periTSCs and primed PSCs in comparison to in vivo postimplantation trophoblast and embryonic disc.
(K) 与体内植入后滋养层和胚胎盘相比,periTSC 和引发的 PSC 中 ELF5 基因座的相对甲基化。
(L) Phase contrast image of periTSCs spontaneously forming spheroids in 2D culture.
(L) 在 2D 培养中自发形成球体的 periTSC 的相差图像。
(M and N) IF of periTSC-spheroids. F-actin stain: phalloidin. LAM, laminin. Scale bars are 50 μm (M) and 100 μm (N).
(M 和 N) periTSC 球体的 IF。 F-肌动蛋白染色:鬼笔环肽。 LAM,层粘连蛋白。比例尺为 50 μm (M) 和 100 μm (N)。
See also Figure S4 and Table S2.
另请参见图 S4表 S2
Microvilli are characteristic for marmoset preimplantation trophoblast and thought to play an important role in primate embryo implantation.6,7 Scanning electron microscopy of PAVS cells revealed extensive microvilli formation, while naive PSCs did not form microvilli (Figure S4B). Examination of cell colony edges showed that microvilli were confined to the apical surface, as observed in implanting primate embryos (Figure S4B).6,7
微绒毛是狨猴植入前滋养层的特征,被认为在灵长类动物胚胎植入中发挥重要作用。 6 7 PAVS 细胞的扫描电子显微镜显示广泛的微绒毛形成,而初始 PSC 未形成微绒毛(图 S4 B)。对细胞集落边缘的检查表明,微绒毛仅限于顶端表面,正如在植入灵长类动物胚胎时观察到的那样(图 S4 B)。 6 7
To assess the developmental identity of PAVS-derived TSCs at the transcriptome level, we performed single-cell profiling and mapped PAVS cells to the spatial transcriptomic embryo dataset. Principal-component analysis (PCA) placed PAVS cells in between trophectoderm and early postimplantation trophoblast samples (Figure 4E). Spatial identity mapping revealed greatest correlation scores to trophectoderm and lower correlation to postimplantation trophoblast (Figure 4F). Uniform manifold approximation and projection (UMAP) showed similar results, with some PAVS cells clustering with trophectoderm at CS3 (Figure S4C). Global correlation analysis placed PAVS cells with trophectoderm, in contrast to postTSCs, which more closely correlated with postimplantation trophoblast (Figure S4D). Therefore, we annotated PAVS TSCs as periimplantation trophoblast-like TSCs (periTSCs). periTSCs lacked expression of ExMes-associated transcripts and showed robust trophoblast gene expression (Figure S4E). Importantly, periTSCs upregulated preimplantation-specific mRNAs, including NOTO, MIOX, and FABP3, in contrast to postTSCs and OK cells (Figure 4G). To examine whether periTSCs are able to contribute to trophectoderm, we generated cross-species aggregation chimeras with mouse embryos and marmoset periTSCs (Figure 4H). periTSCs contributed exclusively to the trophectoderm with an overall efficiency of 27% (13.5% and 13.5% for complete and partial trophectoderm contribution, respectively) (Figures 4I, 4J, S4F, and S4G).
为了在转录组水平评估 PAVS 衍生的 TSC 的发育特性,我们进行了单细胞分析并将 PAVS 细胞映射到空间转录组胚胎数据集。主成分分析 (PCA) 将 PAVS 细胞置于滋养外胚层和早期植入后滋养层样本之间(图 4 E)。空间同一性图谱显示与滋养外胚层的相关性最高,与植入后滋养层的相关性较低(图4F )。均匀流形近似和投影 (UMAP) 显示了类似的结果,一些 PAVS 细胞与滋养外胚层在 CS3 处聚集(图 S4 C)。全局相关性分析将 PAVS 细胞置于滋养外胚层,而后 TSC 与植入后滋养层的相关性更密切(图 S4 D)。因此,我们将 PAVS TSC 注释为植入周围滋养层样 TSC(periTSC)。 periTSC 缺乏 ExMes 相关转录本的表达,并显示出强大的滋养层基因表达(图 S4 E)。重要的是,与 postTSC 和 OK 细胞相比,periTSC 上调了植入前特异性 mRNA,包括NOTOMIOXFABP3图 4 G)。为了检查 periTSC 是否能够促进滋养外胚层,我们用小鼠胚胎和狨猴 periTSC 生成了跨物种聚集嵌合体(图 4 H)。 periTSC 专门对滋养外胚层做出贡献,总体效率为 27%(完全和部分滋养外胚层贡献分别为 13.5% 和 13.5%)(图 4 I、4J、 S4 F 和 S4G)。
Human TSCs exhibit lower global DNA-methylation levels compared with primed PSCs and somatic lineages.20 Little is known about the methylation dynamics in marmoset pre- and postimplantation development. Consequently, we generated single-cell bisulfite sequencing samples of marmoset in vivo postimplantation trophoblast and embryonic disc. Furthermore, we performed bisulfite sequencing on bulk populations of in vitro marmoset naive PSCs, primed PSCs, and periTSCs. Global CpG methylation levels were lowest in naive PSCs and highest in primed PSCs (Figure S4H), consistent with reported differences between naive and primed pluripotent states in human.49,50 periTSCs exhibited slightly lower DNA methylation levels compared with primed PSCs and characteristic hypomethylation of the ELF5 promoter region, consistent with the dynamics observed in vivo between the postimplantation trophoblast and the embryonic disc (Figures 4K, S4H, and S4I).51 Interestingly, ELF5 hypomethylation was also observed in naive PSCs, which suggests that some trophoblast factors remain unmethylated in the preimplantation epiblast (Figure S4J). In vivo trophoblast and periTSCs exhibited similar hypomethylation patterns for characteristic trophoblast genes such as KRT7 (Figure S4K). To investigate trophoblast-specific methylation features, we determined differentially methylated genes in naive PSCs versus periTSCs. periTSCs exhibited higher methylation levels of pluripotency factors POU5F1, NANOG, and KLF17 (Figure S4I). SMAD2 and the amnion-specific gene VTCN1 were also more highly methylated in periTSCs (Figure S4I),52,53,54,55 suggesting trophoblast-specific DNA methylation of epiblast- and amnion-related transcripts.
与引发的 PSC 和体细胞谱系相比,人类 TSC 表现出较低的整体 DNA 甲基化水平。 20人们对狨猴植入前和植入后发育中的甲基化动态知之甚少。因此,我们生成了狨猴体内植入后滋养层和胚胎盘的单细胞亚硫酸氢盐测序样本。此外,我们对体外狨猴幼稚 PSC、引发的 PSC 和 periTSC 的大量群体进行了亚硫酸氢盐测序。总体 CpG 甲基化水平在初始 PSC 中最低,在引发 PSC 中最高(图 S4 H),与报道的人类初始和引发多能状态之间的差异一致。与引发的 PSC 相比, 49 50 个periTSC 表现出略低的 DNA 甲基化水平和ELF5启动子区域的特征性低甲基化,与体内观察到的植入后滋养层和胚盘之间的动态一致(图 4 K、 S4 H 和 S4I)。 51有趣的是,在幼稚 PSC 中也观察到ELF5低甲基化,这表明一些滋养层因子在植入前外胚层中仍未甲基化(图 S4 J)。 体内滋养层和 periTSC 的特征性滋养层基因(例如KRT7 )表现出相似的低甲基化模式(图 S4 K)。为了研究滋养层特异性甲基化特征,我们确定了幼稚 PSC 与 periTSC 中的差异甲基化基因。 periTSC 表现出较高的多能性因子POU5F1NANOGKLF17甲基化水平(图 S4 I)。 SMAD2和羊膜特异性基因VTCN1在 periTSC 中的甲基化程度也更高(图 S4 I), 52 53 54 55表明外胚层和羊膜相关转录物的滋养层特异性 DNA 甲基化。
Human and non-human primate PSCs harbor the potential to differentiate into amnion.34,53,54,56,57,58 Marmoset amnion in vivo exhibits a flat squamous epithelial morphology and expresses several trophoblast-associated genes, including TFAP2C and GATA2.34 To determine whether periTSCs or postTSCs expressed molecular features of amnion, we extracted a comprehensive panel of markers that discriminate between amnion and trophoblast in vivo (Figure S4M). Both periTSCs and postTSCs expressed the majority of trophoblast-associated genes and lacked amnion-related transcripts, including VTCN1 and POU5F1 (Figure S4L), consistent with previous global dimensionality reduction methods (Figures 3F, 4E, S3D, and S4C). Immunofluorescence confirmed that periTSCs do not express VTCN1 at a protein level (Figure S4M). The addition of bone morphogenic protein 4 (BMP4) induced VTCN1 expression (Figure S4M), in line with the reported role of BMP signaling for amnion formation.34,53,54,56,57
人类和非人类灵长类动物 PSC 具有分化为羊膜的潜力。 34 53 54 56 57 58狨猴羊膜在体内表现出平坦的鳞状上皮形态,并表达多种滋养层相关基因,包括TFAP2CGATA234为了确定 periTSC 或 postTSC 是否表达羊膜的分子特征,我们提取了一组可在体内区分羊膜和滋养层的全面标记物(图 S4 M)。 periTSC 和 postTSC 均表达大部分滋养层相关基因,缺乏羊膜相关转录本,包括VTCN1POU5F1图 S4 L),与之前的全局降维方法一致(图 3 F、 4 E、 S3 D 和S4 C) )。免疫荧光证实 periTSC 在蛋白质水平上不表达 VTCN1(图 S4 M)。 添加骨形态发生蛋白 4 (BMP4) 可诱导 VTCN1 表达(图 S4 M),这与报道的 BMP 信号传导对羊膜形成的作用一致。34 53 54 56 57
We observed that periTSCs, when passaged or cultured at higher densities, formed free-floating vesicles that could be maintained in hanging drops for at least 5 days (Figures 4L and S4O). Immunocytochemistry of periTSC vesicles showed robust expression of the trophectoderm marker GATA3 (Figure 4M). EZRIN and F-actin localized to the outer membrane of periTSC vesicles, which demonstrated that periTSC-derived trophoblast spheroids (Tb-spheroids) exhibit an outward-facing apical domain, similar to the trophectoderm of the blastocyst (Figure 4M). Consistent with this result, Laminin was found on the inner membrane of the POU5F1−/AP2γ+/ZO1+ Tb-spheroids (Figure 4N).
我们观察到,当以较高密度传代或培养时,periTSC 形成自由漂浮的囊泡,可以在悬滴中维持至少 5 天(图 4 L 和S4 O)。 periTSC 囊泡的免疫细胞化学显示滋养外胚层标记物 GATA3 的强劲表达(图 4 M)。 EZRIN和F-肌动蛋白定位于periTSC囊泡的外膜,这表明periTSC衍生的滋养层球体(Tb球体)表现出面向外的顶端结构域,类似于囊胚的滋养外胚层(图4M )。与此结果一致,在 POU5F1−/AP2γ+/ZO1+ Tb 球体的内膜上发现了层粘连蛋白(图 4 N)。
We conclude that PAVS induces a periimplantation trophectoderm-like TSC state, as determined by transcriptome and methylome profiling, chimeric contribution to the trophectoderm of mouse blastocysts and spontaneous syncytium and trophectoderm-like spheroid formation.
我们得出的结论是,PAVS 诱导植入周围滋养外胚层样 TSC 状态,根据转录组和甲基化组分析、对小鼠囊胚滋养外胚层的嵌合贡献以及自发合胞体和滋养外胚层样球体形成确定。

CDX2 overrides naive and primed self-renewing culture conditions
CDX2超越幼稚和启动的自我更新培养条件

In mouse, CDX2 induces trophoblast lineage entry via the repression of pluripotency factors.59,60,61 In particular, CDX2-mediated downregulation of Pou5f1 (OCT4) is a key juncture during mouse trophectoderm specification.59 To examine the effect of the trophectoderm transcriptional program on pluripotency in the marmoset, we generated doxycycline-inducible CDX2 overexpression PSCs lines by PiggyBac transposition (Figures 5A, S5A, and S5B). CDX2 overexpression for 2 passages in naive PSCs promoted trophoblast identity, even in self-renewing naive PSCs culture conditions (Figures 5A and 5B). CDX2-overexpressing PSCs in naive culture conditions exhibited similar morphology to periTSCs and were maintained beyond passage 3 with CDX2 overexpression (Figures 5A and 5B). Notably, OCT4 was downregulated in CDX2 overexpressing cells in naive PSCs culture conditions (Figure S5C). This suggests that CDX2 induction extinguishes OCT4 expression in marmoset PSCs.
在小鼠中,CDX2 通过抑制多能因子诱导滋养层谱系进入。 59 60 61特别是,CDX2 介导的Pou5f1 (OCT4) 下调是小鼠滋养外胚层分化过程中的关键节点。 59为了检查滋养外胚层转录程序对狨猴多能性的影响,我们通过 PiggyBac 转座生成了强力霉素诱导的CDX2过表达 PSC 系(图 5 A、 S5 A 和 S5B)。即使在自我更新的幼稚 PSC 培养条件下, CDX2在幼稚 PSC 中过表达 2 代也能促进滋养层身份(图 5 A 和 5B)。在初始培养条件下过表达CDX2 的PSC 表现出与 periTSC 相似的形态,并且在第 3 代之后仍保持 CDX2 过表达(图 5A和 5B)。值得注意的是,在初始 PSC 培养条件下,OCT4 在CDX2过表达细胞中下调(图 S5 C)。这表明 CDX2 诱导消除了狨猴 PSC 中的 OCT4 表达。
  1. Download: Download high-res image (3MB)
    下载:下载高分辨率图像 (3MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 5. OCT4 does not inhibit trophoblast formation in the marmoset
图5 . OCT4 不抑制狨猴滋养层形成

(A) Schematic of CDX2 overexpression in naive marmoset PSCs for 2 passages.
(A) 幼稚狨猴 PSC 中 CDX2 过表达 2 代的示意图。
(B) IF of CDX2 overexpression in naive PSCs. Scale bars are 50 μm.
(B) 幼稚 PSC 中 CDX2 过表达的 IF。比例尺为 50 μm。
(C) Schematic of CDX2 overexpression in primed marmoset PSCs for 2 passages.
(C) 已引发的狨猴 PSC 中 CDX2 过表达 2 代的示意图。
(D) IF of CDX2 overexpression in primed PSCs. Scale bars are 50 μm.
(D) 引发的 PSC 中 CDX2 过表达的 IF。比例尺为 50 μm。
(E) Violin plot of normalized expression in preimplantation marmoset lineages.
(E) 植入前狨猴谱系中标准化表达的小提琴图。
(F) IF of PAVS-differentiated cells at day 3.
(F) 第 3 天 PAVS 分化细胞的 IF。
(G) Schematic of OCT4 overexpression in naive PSCs cultured in naive medium (PLAXA). Overexpression was maintained for 2 passages unless stated otherwise.
(G) 在初始培养基 (PLAXA) 中培养的初始 PSC 中 OCT4 过表达的示意图。除非另有说明,过表达维持2代。
(H) qPCR of overexpression-specific OCT4 expression.
(H) 过表达特异性 OCT4 表达的 qPCR。
(I) Phase contrast image of PAVS-differentiated cells from OCT4 overexpressing naive PSCs at passage 3. Overexpression was maintained for 3 passages.
(I) 来自第 3 代过表达初始 PSC 的 OCT4 的 PAVS 分化细胞的相差图像。过表达维持了 3 代。
(J) IF of OCT4 overexpression in PSCs and periTSCs in indicated conditions.
(J) 在指定条件下 PSC 和 periTSC 中 OCT4 过表达的 IF。
(K) Quantification of CDX2+ cells in (J). Threshold was determined by 2 times the standard deviation of CDX2 expression in PSCs. Significance was calculated using a Kruskal-Wallis followed by Dunn's multiple comparison test. n = 2. (p < 0.05.).
(K) (J) 中 CDX2+ 细胞的定量。阈值由 PSC 中 CDX2 表达标准差的 2 倍确定。使用 Kruskal-Wallis 和 Dunn 多重比较检验计算显着性。 n = 2。( * p < 0.05。)。
(L) IF PAVS-derived cells from OCT4 overexpressing naive PSCs at passage 2.
(L) 来自 OCT4 的 IF PAVS 衍生细胞在第 2 代过表达初始 PSC。
All scale bars (except for B and D) are 100 μm.
所有比例尺(B 和 D 除外)均为 100 μm。
See also Figure S5.
另请参见图 S5
To test if CDX2 can promote a trophoblast-like phenotype in primed (postimplantation-like) pluripotency, we assessed CDX2 overexpression in self-renewing primed PSC culture conditions (knockout serum replacement [KSR]/basic FGF [bFGF]) (Figures 5C and S5D). CDX2 overexpression in primed cells led to a decrease in OCT4 expression and differentiation (Figure 5D). However, DOX+ cells remained KRT7−, suggesting differentiation into other lineages (Figure 5D). A subset of CDX2 overexpressing PSCs in KSR/bFGF expressed TBXT+, which may indicate acquisition of a primitive streak-like mesodermal cell identity (Figure S5E).34,36,62 We conclude that CDX2 overrules pluripotency maintenance conditions and promotes trophoblast-like identity in naive, but not primed, marmoset PSCs.
为了测试 CDX2 是否可以促进启动(植入后样)多能性中的滋养层样表型,我们评估了自我更新启动 PSC 培养条件(敲除血清替代物 [KSR]/碱性 FGF [bFGF])中的CDX2过表达(图 5 C)和S5 D)。致敏细胞中 CDX2 过度表达导致 OCT4 表达和分化减少(图 5D )。然而,DOX+ 细胞仍然是 KRT7−,表明分化为其他谱系(图 5D )。 KSR/bFGF 中CDX2过表达 PSC 的子集表达 TBXT+,这可能表明获得了原条状中胚层细胞身份(图 S5 E)。 34 36 62我们得出的结论是,CDX2 会否决多能性维持条件,并促进幼稚但未启动的狨猴 PSC 中的滋养层样身份。

POU5F1 does not restrict marmoset trophoblast differentiation in vitro
POU5F1不限制狨猴滋养细胞体外分化

Analysis of pre- and postimplantation-specific trophoblast signatures (Figure 1) identified the presence of pluripotency factors in marmoset trophectoderm in vivo, including POU5F1 (Figures 5E and S5F). This is in contrast to mouse, where Pou5f1 (OCT4) is rapidly downregulated in the trophectoderm.63 Therefore, we set out to examine the role of POU5F1 in marmoset trophoblast differentiation. To elucidate POU5F1 dynamics during TSC derivation, we performed lineage marker analysis at day 3 during periTSCs differentiation from naive PSCs. SOX2 was downregulated in the majority of cells; however, OCT4 was present in all cells, including early GATA3+ and KRT7+ trophoblast cells (Figures 5F and S5H–S5J). Image quantification showed no anti-correlation (R = 0.1066) between OCT4 and GATA3 expression (Figure S5J). Equally, CDX2 started to become upregulated in a subset of OCT4+ cells (Figure S5G). To test whether POU5F1 upregulation and sustained expression would prevent differentiation into trophoblast-like cells, we generated marmoset doxycycline-inducible POU5F1 overexpression PSCs (Figure 5G). We confirmed tight control over the POU5F1-transgene (Figure 5H) and derived periTSCs in the presence and absence of doxycycline (DOX+/−) (Figures 5I and 5J). In the absence of DOX (DOX−), POU5F1-inducible naive PSCs converted into KRT7+/CDX2+/POU5F1− periTSCs (Figure 5J), similar to wild-type cells (Figure 4C). In the presence of DOX (DOX+) in PAVS for 2 passages, POU5F1-inducible naive PSCs equally flattened out, acquired a trophoblast-like morphology, and could be readily propagated (Figure 5I). Sustained POU5F1 expression did not interfere with the robust expression of trophoblast markers KRT7, GATA3, and CDX2 (Figures 5J–5L). We conclude that POU5F1 does not inhibit trophoblast-specific gene expression in vitro.
对植入前和植入后特异性滋养层特征的分析(图1 )确定狨猴体内滋养外胚层中存在多能因子,包括POU5F1图5ES5F )。这与小鼠相反,小鼠的滋养外胚层中的Pou5f1 (OCT4) 快速下调。 63因此,我们着手研究POU5F1在狨猴滋养层分化中的作用。为了阐明 TSC 衍生过程中的POU5F1动态,我们在 periTSC 与初始 PSC 分化期间的第 3 天进行了谱系标记分析。 SOX2 在大多数细胞中下调;然而,OCT4 存在于所有细胞中,包括早期 GATA3+ 和 KRT7+ 滋养层细胞(5F 和S5H -S5J)。图像量化显示 OCT4 和 GATA3 表达之间没有反相关性 (R = 0.1066)(图 S5 J)。同样,CDX2 在 OCT4+ 细胞的子集中开始上调(图 S5 G)。为了测试POU5F1上调和持续表达是否会阻止分化为滋养层样细胞,我们生成了狨猴强力霉素诱导的POU5F1过表达 PSC(图 5 G)。 我们确认了在存在和不存在强力霉素(DOX+/-)的情况下对POU5F1转基因(5H)和衍生的 periTSC 的严格控制(5I 和 5J)。在缺乏 DOX (DOX−) 的情况下, POU5F1诱导型幼稚 PSC 转化为 KRT7+/CDX2+/POU5F1− periTSC(图 5 J),与野生型细胞类似(图 4 C)。在 PAVS 中存在 DOX (DOX+) 的情况下传代 2 代, POU5F1诱导的幼稚 PSC 同样变平,获得滋养层样形态,并且可以轻松繁殖(图 5 I)。持续的POU5F1表达不会干扰滋养层标记物 KRT7、GATA3 和 CDX2 的稳健表达(图 5 J-5L)。我们得出结论, POU5F1在体外不会抑制滋养层特异性基因表达。

Marmoset TSCs recapitulate features of superficial attachment
狨猴 TSC 再现了表面附着的特征

A central function of trophectoderm is to form a vesicle that is capable of implantation. We previously established a microfluidic workflow to encapsulate PSCs into monodisperse agarose microgels.54,64,65 To set up a platform for trophoblast invasion of Tb-spheroids, we encapsulated either marmoset periTSCs or human OK TSCs in agarose microgels (Figures 6A, 6B, and S6A). Both marmoset periTSCs and human OK TSCs20 formed Tb-spheroids in agarose gels from day 3 onward that continued to expand, resulting in many structures escaping from the microgels by day 5 (Figure 6B). Marmoset and human Tb-spheroids robustly expressed trophoblast markers CDX2 and AP2γ and lacked the pluripotency factor SOX2 (Figures 6C and S6B).
滋养外胚层的核心功能是形成能够植入的囊泡。我们之前建立了一个微流体工作流程,将 PSC 封装到单分散琼脂糖微凝胶中。 54 64 65为了建立滋养层侵入 Tb 球体的平台,我们将狨猴 periTSC 或人类 OK TSC 封装在琼脂糖微凝胶中(图 6 A、6B 和S6 A)。从第 3 天开始,狨猴 periTSC 和人类 OK TSC 20在琼脂糖凝胶中形成 Tb 球体,并继续扩大,导致许多结构在第 5 天从微凝胶中逸出(图 6 B)。狨猴和人类 Tb 球体强烈表达滋养层标记 CDX2 和 AP2γ,但缺乏多能因子 SOX2(图 6 C 和S6 B)。
  1. Download: Download high-res image (3MB)
    下载:下载高分辨率图像 (3MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 6. Marmoset TSCs recapitulate characteristics of superficial implantation and early invasion
图6 .狨猴 TSC 再现了浅表植入和早期侵袭的特征

(A) Schematic of Tb-spheroid formation by encapsulation.
(A) 通过封装形成 Tb 球体的示意图。
(B) Phase contrast images of encapsulated marmoset Tb-spheroids at days 3 and 5. Escaped: structures that have broken out of agarose microgels and are growing in suspension. Scale bars are 50 μm.
(B) 第 3 天和第 5 天时封装的狨猴 Tb 球体的相差图像。逃逸:已从琼脂糖微凝胶中脱离并在悬浮液中生长的结构。比例尺为 50 μm。
(C) IF of marmoset Tb-spheroids at day 3. F-actin stain: phalloidin. Scale bars are 50 μm.
(C) 第 3 天狨猴 Tb 球体的 IF。F-肌动蛋白染色:鬼笔环肽。比例尺为 50 μm。
(D) Schematic of Tb-spheroid attachment assay.
(D) Tb 球体附着测定示意图。
(E) IF of attached marmoset Tb-spheroids on Matrigel beds. Scale bars are 50 μm.
(E) 基质胶床上附着的狨猴 Tb 球体的 IF。比例尺为 50 μm。
(F) IF of NLS-GFP and LifeACT (actin stain) overexpressing Tb-spheroids on a collagen-coated surface. Scale bars are 50 μm.
(F) NLS-GFP 和 LifeACT(肌动蛋白染色)在胶原蛋白包被表面上过表达 Tb 球体的 IF。比例尺为 50 μm。
(G) IF of attaching primed marmoset PSC and postTSC spheroids on a Matrigel-coated surface. Scale bars are 50 μm.
(G) 将已涂底狨猴 PSC 和 postTSC 球体附着在基质胶涂层表面上的 IF。比例尺为 50 μm。
(H) Schematic of Tb-spheroids released onto ECM of varying thickness.
(H) 释放到不同厚度的 ECM 上的 Tb 球体示意图。
(I) IF of marmoset Tb-spheroids attaching on gels of varying thickness. Scale bars are 50 μm.
(I) 狨猴 Tb 球体附着在不同厚度的凝胶上的 IF。比例尺为 50 μm。
(J) Schematic of TSC differentiation in human EVT conditions. LAM, laminin.
(J) 人类 EVT 条件下 TSC 分化的示意图。 LAM,层粘连蛋白。
(K) Phase contrast images of periTSCs (PAVS) and postTSCs (OK XAV PD) differentiated in human EVT conditions for 6 days. Scale bars are 100 μm.
(K) periTSC (PAVS) 和 postTSC (OK XAV PD) 在人类 EVT 条件下分化 6 天的相差图像。比例尺为 100 μm。
(L) IF of periTSCs (PAVS) before and after human EVT differentiation. Scale bars are 100 μm.
(L) 人类 EVT 分化前后 periTSC (PAVS) 的 IF。比例尺为 100 μm。
See also Figure S6.
另请参见图 S6
To assess marmoset Tb-spheroid implantation in vitro, we allowed spheroids to attach on a Matrigel bed after release from the microgels (Figure 6D). Marmoset Tb-spheroids maintained regular spherical shapes and expanded lumina (Figures 6E, S6D, and S6E), consistent with the superficial implantation mode of marmoset embryos (Figure 1A), where the blastocyst expands to fill the uterine cavity. By contrast, human Tb-spheroids exhibited an irregular and deflated morphology (Figures S6C–S6E). These findings suggest that marmoset TSCs recapitulate some morphological aspects of superficial attachment (Figure S6F).
为了评估狨猴 Tb 球体的体外植入,我们允许球体从微凝胶释放后附着在基质胶床上(图 6D )。狨猴 Tb 球体保持规则的球形形状和扩张的管腔(图 6 E、 S6 D 和 S6E),与狨猴胚胎的浅表植入模式一致(图 1 A),囊胚扩张以填充子宫腔。相比之下,人类 Tb 球体表现出不规则且扁平的形态(图 S6 C-S6E)。这些发现表明狨猴 TSC 概括了浅表附着的一些形态学方面(图 S6 F)。
Primate trophectoderm differentiates into primary syncytium upon contact with the uterine epithelium in vivo.13 To evaluate the ability of Tb-spheroids to recapitulate primary syncytium formation upon attachment, Tb-spheroids were transferred to adherent culture conditions. Tb-spheroids formed a multinucleated syncytium upon contact with thinly coated Matrigel or collagen IV (Figures 6F, 6G, and S6G). We confirmed that syncytium formation was specific to TSCs by plating marmoset primed PSC-derived spheroids, which did not give rise to multinucleated cells (Figure 6G). Notably, attachment of Tb-spheroids on thicker beds of Matrigel or collagen IV did not form syncytium (Figures 6H and 6I), suggesting that substrate stiffness may play a role in the regulation of primary syncytium formation.
灵长类滋养外胚层在体内与子宫上皮接触后分化为初级合胞体。 13为了评估 Tb 球体在附着后重现初级合胞体形成的能力,将 Tb 球体转移到贴壁培养条件下。 Tb 球体在与薄层基质胶或胶原 IV 接触后形成多核合胞体(图 6F 、6G 和S6G )。我们通过铺板狨猴引发的 PSC 衍生球体来确认合胞体形成对 TSC 具有特异性,该球体不会产生多核细胞(图 6 G)。值得注意的是,Tb 球体附着在较厚的基质胶或胶原 IV 床上并没有形成合胞体(图 6H和 6I),这表明基质硬度可能在调节初级合胞体形成中发挥作用。
Migratory EVT differentiation is essential for immune, endometrial, and vessel remodeling during placentation. Despite significantly delayed trophoblast invasion in the marmoset compared with human, migratory trophoblast cells have been observed in the marmoset endometrium and uteroplacental arteries in vivo.41 To examine if marmoset TSCs are able to differentiate into invasive trophoblast, we cultured periTSCs and postTSCs in human EVT conditions20 (Figure 6J). Supplementation of NRG1 and ECM together with TGF-β/NODAL inhibition robustly induced epithelial-mesenchymal transition and expression of the EVT marker MMP2 by day 6, similar to human EVT differentiation (Figures 6K, 6L, and S6I–S6K). The human EVT marker HLA-G showed very limited conservation in the marmoset (Figure S6H).66,67 To determine the migratory potential of marmoset EVT-like cells, we performed transwell migration assays. Marmoset EVT differentiation increased migratory behavior, similar to EVT differentiation in human (Figure S6L). This result suggests at least partial conservation of EVT specification in human and marmoset.
迁移性 EVT 分化对于胎盘期间的免疫、子宫内膜和血管重塑至关重要。尽管与人类相比,狨猴的滋养层侵袭显着延迟,但在狨猴体内的子宫内膜和子宫胎盘动脉中观察到了迁移的滋养层细胞。 41为了检查狨猴 TSC 是否能够分化为侵袭性滋养层,我们在人类 EVT 条件下培养 periTSC 和 postTSC 20图 6 J)。补充 NRG1 和 ECM 以及抑制 TGF-β/NODAL 可在第 6 天强烈诱导上皮间质转化和 EVT 标记物 MMP2 的表达,类似于人类 EVT 分化(图 6K 、6L 和S6 I-S6K)。人类 EVT 标记 HLA-G 在狨猴中表现出非常有限的保守性(图 S6 H)。 66 67为了确定狨猴 EVT 样细胞的迁移潜力,我们进行了 Transwell 迁移测定。狨猴 EVT 分化增加了迁徙行为,类似于人类的 EVT 分化(图 S6 L)。这一结果表明,人类和狨猴中 EVT 规范至少部分保留。
Together, these experiments show that marmoset periTSCs undergo superficial attachment, form primary syncytia, and are capable of EVT differentiation, thus recapitulating important aspects of New World monkey trophoblast development.
总之,这些实验表明,狨猴 periTSC 经历表面附着,形成初级合胞体,并且能够进行 EVT 分化,从而概括了新世界猴滋养层发育的重要方面。

WNT stimulation is required to suppress spontaneous EVT differentiation in human
抑制人类自发 EVT 分化需要 WNT 刺激

FGF/ERK and TGF-β/NODAL inhibition are integral to human TSC and marmoset periTSC culture conditions (Figures 3A–3C).27,29 To test whether the marmoset periTSC culture regime can induce trophoblast identity in human, we cultured naive human PSCs49,68 in PAVS conditions. PAVS induced flat, epithelial-like cells within 3 days that could readily be passaged (Figure 7A). Human PAVS colonies expressed KRT7 and AP2γ (TFAP2C) (Figure 7B), indicating that PAVS conditions promote a trophoblast-like state in human naive PSCs at early passages. However, after three to five passages, small cell populations gradually acquired a mesenchymal, spindle-shaped morphology (Figure 7C). Lineage marker analysis revealed that the spindle-shaped cells in passage 5 human PAVS cultures expressed the EVT marker HLA-G (Figure 7D). This result suggests that human trophoblast-like PAVS cells gradually differentiate into the EVT lineage.
FGF/ERK 和 TGF-β/NODAL 抑制是人类 TSC 和狨猴 periTSC 培养条件不可或缺的一部分(图 3 A-3C)。 27 29为了测试狨猴 periTSC 培养方案是否可以诱导人类滋养层身份,我们在 PAVS 条件下培养了初始人类 PSC 49 68 。 PAVS 在 3 天内诱导出易于传代的扁平上皮样细胞(图 7 A)。人 PAVS 集落表达 KRT7 和 AP2γ ( TFAP2C )(图 7 B),表明 PAVS 条件促进人幼稚 PSC 在早期传代时形成滋养层样状态。然而,经过三到五次传代后,小细胞群逐渐获得间质、纺锤形形态(图7C )。谱系标记分析显示,第 5 代人 PAVS 培养物中的纺锤形细胞表达 EVT 标记 HLA-G(图 7D )。这一结果表明人滋养层样 PAVS 细胞逐渐分化为 EVT 谱系。
  1. Download: Download high-res image (2MB)
    下载:下载高分辨率图像 (2MB)
  2. Download: Download full-size image
    下载:下载全尺寸图像

Figure 7. WNT activation prevents EVT differentiation in human PAVS
图 7 . WNT 激活可防止人类 PAVS 中的 EVT 分化

(A) Phase contrast images of naive human PSCs differentiated in PAVS until passage 3.
(A) 原始人类 PSC 在 PAVS 中分化直至第 3 代的相差图像。
(B) IF of trophoblast markers in indicated conditions.
(B) 指定条件下滋养层标记物的 IF。
(C) Phase contrast image of passage 5 human PAVS.
(C) 第 5 代人类 PAVS 的相差图像。
(D) IF of markers of trophoblast differentiation in indicated conditions.
(D) 在所示条件下滋养层分化标记物的 IF。
(E) Cross-species diffusion map analysis in human, marmoset, and cynomolgus monkey. TE, trophectoderm; CTB, cytotrophoblast; STB, syncytiotrophoblast; EVT, extravillous trophoblast.
(E) 人类、狨猴和食蟹猴的跨物种扩散图分析。 TE,滋养外胚层; CTB,细胞滋养层; STB,合体滋养层; EVT,绒毛外滋养层。
(F) Violin plot of normalized expression in extraembryonic marmoset lineages and in remodeled and native endometrial tissue. reGland: remodeled glands; reStroma: remodeled stroma.
(F) 胚胎外狨猴谱系以及重塑和天然子宫内膜组织中标准化表达的小提琴图。 reGland:重塑腺体; reStroma:重塑的基质。
(G) Gene Ontology terms of significantly enriched signaling pathways.
(G) 显着丰富的信号通路的基因本体术语。
(H and I) Phase contrast images of naive human PSCs differentiated for (H) 5 days in indicated conditions or (I) at passage 5 of indicated conditions.
(H 和 I) 在指定条件下分化 (H) 5 天或 (I) 在指定条件下第 5 代分化的原始人类 PSC 的相差图像。
(J) qPCR of lineage markers in human PSCs and TSCs at different passages. Expression values are normalized by row using Z score. PXGL, naive PSCs; PAVS, TSCs; P2, passage 2; P5, passage 5; Epi, epiblast; Am, amnion; TE, trophectoderm; TB, general trophoblast markers; STB, syncytiotrophoblast; EVT, extravillous trophoblast.
(J) 不同传代的人类 PSC 和 TSC 中谱系标记的 qPCR。使用Z分数按行对表达值进行归一化。 PXGL,幼稚 PSC; PAVS、TSC; P2,第 2 段; P5,第 5 段; Epi,外胚层; Am,羊膜; TE,滋养外胚层; TB,一般滋养层标志物; STB,合体滋养层; EVT,绒毛外滋养层。
(K) IF of passage 6 trophoblast differentiation markers in indicated conditions.
(K) 在指定条件下第 6 代滋养层分化标记物的 IF。
All scale bars are 100 μm.
所有比例尺均为 100 μm。
See also Figure S7.
另请参见图 S7
To determine potential drivers of primate EVT and syncytiotrophoblast formation, we generated a single-cell embryo profiling compendium for the trophoblast lineages of human,9 cynomolgus monkey,69 and marmoset.34 Integrated analysis revealed the formation of two branches, representing syncytiotrophoblast and EVT differentiation (Figures 7E and S7A). Trophectoderm and early cytotrophoblast clustered in close proximity at the root of the syncytiotrophoblast and EVT trajectories in human and cynomolgus. Notably, marmoset samples did not form EVT at the examined stages (CS5–7), demonstrating a species-specific adaptation. This observation is consistent with the reported delayed formation of proliferative trophoblast projections in marmoset compared with human.70 Our previous experiments suggested that NRG1 is a conserved agonist of EVT differentiation in both marmoset and human (Figure 6). Notably, NRG1 is secreted by decidualized stromal cells in human.71 In accordance with delayed EVT formation, marmoset decidualized stromal cells did not express NRG1 at CS5–7 (Figure 7F). To further identify species-specific adaptations in cell signaling, we performed signaling pathway analysis between human and marmoset cytotrophoblasts. GO showed that WNT signaling was significantly enriched in the marmoset trophoblast (Figure 7G), suggesting that endogenous WNT may play a role in preventing premature EVT differentiation. Consistent with this notion, the soluble WNT inhibitor DKK1 was highly expressed in human decidualized endometrial stroma33 but scarcely detectable in marmoset (Figure 7F). These data highlight a potential role for WNT signaling in human EVT differentiation in vivo.
为了确定灵长类动物 EVT 和合体滋养层形成的潜在驱动因素,我们为人类、 9食蟹猴、 69和狨猴的滋养层谱系生成了单细胞胚胎分析纲要。 34综合分析揭示了两个分支的形成,代表合体滋养层和 EVT 分化(图 7 E 和S7 A)。在人和食蟹猴中,滋养外胚层和早期细胞滋养层紧密聚集在合体滋养层和 EVT 轨迹的根部。值得注意的是,狨猴样本在检查阶段(CS5-7)并未形成 EVT,这表明了物种特异性的适应。这一观察结果与报道的狨猴与人类相比增殖性滋养层投射的形成延迟一致。 70我们之前的实验表明,NRG1 在狨猴和人类中都是 EVT 分化的保守激动剂(图 6 )。值得注意的是,NRG1 由人类蜕膜基质细胞分泌。 71根据 EVT 形成延迟,狨猴蜕膜基质细胞在 CS5-7 处不表达 NRG1(图 7 F)。为了进一步确定细胞信号传导中的物种特异性适应性,我们对人类和狨猴细胞滋养层之间的信号传导途径进行了分析。 GO显示WNT信号在狨猴滋养层中显着富集(图7G ),这表明内源性WNT可能在防止EVT过早分化中发挥作用。与这一观点一致,可溶性 WNT 抑制剂DKK1在人类蜕膜化子宫内膜基质中高度表达33,但在狨猴中几乎检测不到(图 7 F)。这些数据强调了 WNT 信号在人类 EVT体内分化中的潜在作用。
WNT activation is a core component of OK medium for TSC self-renewal and is suggested to stabilize the cytotrophoblast state20,21 (Shannon et al.,72). We tested if WNT activation with the GSK-3β inhibitor CHIR99021 could prevent EVT differentiation in PAVS conditions (Figure 7H). Naive human PSCs were cultured in PAVS and PAVS plus CHIR99021 (PAVS+CHIR). In both media, naive PSCs flattened out and generated epithelial colonies within 3–5 days (Figure 7H). By passage 5, PAVS+CHIR samples sustained a TSC-like epithelial morphology, whereas PAVS TSCs differentiated into EVT-like cells (Figures 7C and 7I). Transcriptional profiling of lineage markers of naive PSCs, PAVS, and PAVS CHIR at passage 5 showed that human cells in PAVS downregulated cytotrophoblast markers, including GATA2, GATA3, and NR2F2, and showed strong upregulation of EVT markers HLA-G and ANXA4 (Figure 7J). However, PAVS+CHIR robustly expressed GATA3, GATA2, TFAP2C, NR2F2, and OVOL1, consistent with TSC identity (Figure 7J). At the protein level, human PAVS cells showed widespread upregulation of EVT-specific HLA-G and CGB at passage 6 (Figures 7K and S7B). By contrast, PAVS+CHIR cultured human cells did not express trophoblast differentiation markers, maintained epithelial morphology, and were positive for AP2γ (TFAP2C) (Figure 7K). These results demonstrate that in human, but not marmoset, TSCs require WNT signaling to block EVT differentiation in PAVS conditions.
WNT 激活是 TSC 自我更新 OK 培养基的核心成分,并被认为可以稳定细胞滋养层状态20 21 (Shannon 等人, 72 )。我们测试了 GSK-3β 抑制剂 CHIR99021 激活 WNT 是否可以阻止 PAVS 条件下的 EVT 分化(图 7 H)。初始人类 PSC 在 PAVS 和 PAVS 加 CHIR99021 (PAVS+CHIR) 中培养。在两种培养基中,初始 PSC 在 3-5 天内变平并产生上皮集落(图 7 H)。到第 5 代时,PAVS+CHIR 样品维持了 TSC 样上皮形态,而 PAVS TSC 分化为 EVT 样细胞(7C 和 7I)。第 5 代幼稚 PSC、PAVS 和 PAVS CHIR 谱系标记物的转录谱显示,PAVS 中的人类细胞下调了细胞滋养层标记物,包括GATA2GATA3NR2F2 ,并且 EVT 标记物HLA-GANXA4强烈上调(图 7) J)。然而,PAVS+CHIR 强烈表达GATA3GATA2TFAP2CNR2F2OVOL1 ,与 TSC 身份一致(图 7 J)。 在蛋白质水平上,人 PAVS 细胞在第 6 代时表现出 EVT 特异性 HLA-G 和 CGB 的广泛上调(7K 和S7B )。相比之下,PAVS+CHIR培养的人细胞不表达滋养层分化标记物,维持上皮形态,并且AP2γ( TFAP2C )呈阳性(图7K )。这些结果表明,在人类(而非狨猴)中,TSC 需要 WNT 信号传导来阻断 PAVS 条件下的 EVT 分化。

Discussion 讨论

Here, we report the derivation of marmoset postimplantation- and periimplantation-like TSCs from naive PSCs. Marmoset TSCs exhibited characteristic lineage marker expression, transcriptional correspondence to trophoblast in vivo, long-term self-renewal, and the ability to differentiate into multinucleated syncytia and EVT. Importantly, marmoset periTSCs readily formed Tb-spheroids that recapitulated trophectoderm-like cell polarity and underwent primary syncytium formation upon attachment in vitro. The establishment of marmoset TSCs from naive PSCs corroborates the concept of epiblast plasticity in human.25,27,29 Future studies will be required to delineate the timing of lineage restriction in human and non-human primate embryos.
在这里,我们报告了从原始 PSC 中衍生出狨猴植入后和植入周围类似 TSC。狨猴 TSC 表现出特征性谱系标记表达、体内与滋养层的转录对应、长期自我更新以及分化为多核合胞体和 EVT 的能力。重要的是,狨猴 periTSC 很容易形成 Tb 球体,再现滋养外胚层样细胞极性,并在体外附着后经历初级合胞体形成。从原始PSCs建立狨猴TSCs证实了人类外胚层可塑性的概念。 25 27 29未来的研究需要确定人类和非人类灵长类胚胎中谱系限制的时间。
Amnion exhibits a remarkable degree of similarity to trophoblast in the primate embryo: both form thin, squamous epithelia, give rise to a central lumen, and are controlled by a closely related transcriptional circuitry.34,53,56,73,74 Marmoset peri- and postTSCs demonstrated substantially reduced expression of all amnion markers. Methylome analysis of marmoset TSCs revealed hypermethylation of the amnion marker VTCN1, which may present a potential regulatory mechanism to prevent amnion trans-differentiation. Moreover, another key difference between amnion and trophoblast is cell polarity: trophectoderm has an outer apical side and a basal surface toward the inside of the cavity, while the amnion forms a rosette with an inner apical and outer basal orientation.11,75,76,77 We show that Tb-spheroids exhibit trophectoderm-like polarity in the absence of exogenous ECM, in contrast to amnion-like spheroids.54
羊膜与灵长类动物胚胎中的滋养层表现出显着的相似性:两者都形成薄的鳞状上皮,产生中央管腔,并受到密切相关的转录回路的控制。 34 53 56 73 74狨猴 TSC 周围和 TSC 后的所有羊膜标记物的表达均显着降低。狨猴 TSC 的甲基化分析揭示了羊膜标记 VTCN1 的高度甲基化,这可能是防止羊膜转分化的潜在调节机制。此外,羊膜和滋养层之间的另一个主要区别是细胞极性:滋养外胚层具有外部顶端侧和朝向腔内部的基部表面,而羊膜形成具有内部顶端和外部基部方向的玫瑰花结。 11 75 76 77我们发现,在没有外源 ECM 的情况下,Tb 球体表现出滋养外胚层样极性,与羊膜样球体相反。 54
New World monkeys, including the marmoset, exhibit the most shallow modes of trophoblast invasion among primate species.11,39,78 Thus, marmoset TSCs provide an avenue to capture features of superficial implantation and divergent dynamics of EVT differentiation compared with human. Elucidating the molecular mechanisms regulating trophoblast invasion depths will be important for our understanding of pathophysiological changes in placental development. Prominent examples include pre-eclampsia, where the trophoblast fails to invade the uterus sufficiently, or placenta accreta spectrum disorders, which are caused by excessive trophoblast invasion.13 Marmoset trophoblast expands within the uterine cavity without interstitial or endovascular trophoblast invasion during the first month of development in vivo.11,70,79 Cross-species transcriptome analysis of human, cynomolgus monkey, and marmoset embryo datasets indicated an absence of EVT-like cells in the marmoset at CS5–7, yet marmoset TSCs were capable of EVT lineage acquisition. It is tempting to speculate that the absence of NRG1 expression in the marmoset endometrium may contribute to a later onset of EVT formation in vivo. Alternatively, human and marmoset trophoblast cells may differ with regard to endogenous WNT signaling. Marmoset cytotrophoblast was enriched for WNT signaling compared with human, and WNT activation was required to prevent EVT differentiation in human, but not in marmoset. Future studies will be required to delineate the role of WNT in EVT formation and to determine if the marmoset maternal niche secretes signals to delay trophoblast invasion for superficial implantation.
新世界猴,包括狨猴,在灵长类动物中表现出最浅的滋养层入侵模式。 11 39 78因此,狨猴 TSC 提供了一种捕获浅表植入特征以及与人类相比 EVT 分化不同动态的途径。阐明调节滋养层侵袭深度的分子机制对于我们理解胎盘发育的病理生理变化非常重要。突出的例子包括先兆子痫(滋养细胞未能充分侵入子宫)或植入性胎盘谱系疾病(由于滋养细胞过度侵入而引起)。 13狨猴滋养层在体内发育的第一个月内在子宫腔内扩张,没有间质或血管内滋养层侵入。 11 70 79人类、食蟹猴和狨猴胚胎数据集的跨物种转录组分析表明,CS5-7 狨猴中不存在 EVT 样细胞,但狨猴 TSC 能够获得 EVT 谱系。人们很容易推测狨猴子宫内膜中 NRG1 表达的缺失可能导致体内EVT 形成的较晚发生。或者,人类和狨猴滋养层细胞在内源性 WNT 信号传导方面可能有所不同。 与人类相比,狨猴细胞滋养层的 WNT 信号转导富集,并且 WNT 激活是阻止人类 EVT 分化所必需的,但在狨猴中则不然。未来的研究需要阐明 WNT 在 EVT 形成中的作用,并确定狨猴母体生态位是否分泌信号来延迟滋养层入侵以进行浅表植入。
Collectively, our work presents a proof-of-concept for capturing the evolutionary divergence of primate trophoblast development. The comparative analysis of human and marmoset TSCs will be a powerful framework to elucidate primate placentation, regulatory mechanisms of invasion depth, and pathophysiological changes in human placental development.
总的来说,我们的工作为捕捉灵长类滋养层发育的进化差异提供了概念验证。人类和狨猴 TSC 的比较分析将成为阐明灵长类动物胎盘、侵袭深度的调节机制以及人类胎盘发育的病理生理变化的有力框架。

Limitations of the study 研究的局限性

Profiling of marmoset TSCs derived from marmoset blastocysts or first trimester placenta was not included due to technical limitations. We were not able to perform chimeric integration with marmoset embryos due to embryo scarcity. Matrigel alone is insufficient to faithfully recapitulate the maternal environment. Future studies must incorporate endometrial glands and stromal cells. While we have performed significant validation of in vivo marmoset sample identity, the inability to identify cell borders during tissue lysis makes samples with multiple cell types possible.
由于技术限制,未包括源自狨猴囊胚或妊娠早期胎盘的狨猴 TSC 分析。由于胚胎稀缺,我们无法与狨猴胚胎进行嵌合整合。单独的基质胶不足以忠实地再现母体环境。未来的研究必须纳入子宫内膜腺体和基质细胞。虽然我们对体内狨猴样品的身份进行了重要验证,但在组织裂解过程中无法识别细胞边界,使得具有多种细胞类型的样品成为可能。

Resource availability 资源可用性

Lead contact 铅接触

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Thorsten Boroviak (teb45@cam.ac.uk).
有关资源和试剂的更多信息和请求应直接发送给主要联系人 Thorsten Boroviak ( teb45@cam.ac.uk ) ,并由其完成)。

Materials availability 材料可用性

Plasmids and marmoset lines generated in this study are available from the lead contact upon request.
本研究中产生的质粒和狨猴系可根据要求从主要联系人处获得。

Data and code availability
数据和代码可用性

  • RNA sequencing data have been deposited at ArrayExpress and are publicly available as of the date of publication. Accession numbers are listed in the key resources table.
    RNA 测序数据已存入 ArrayExpress,并自发布之日起公开可用。关键资源表中列出了入藏号。
  • This paper analyzes existing, publicly available data. These accession numbers for the datasets are listed in the key resources table.
    本文分析了现有的公开数据。关键资源表中列出了这些数据集的入藏号。
  • Any additional data reported in this paper will be shared by the lead contact upon request.
    本文中报告的任何其他数据将根据要求由主要联系人共享。
  • All original code has been deposited at https://github.com/Boroviak-Lab/Trophoblast and is publicly available as of the date of publication. DOIs are listed in the key resources table.
    所有原始代码已存放在https://github.com/Boroviak-Lab/Trophoblast并自发布之日起公开发布。 DOI 列在关键资源表中。
  • Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
    重新分析本文报告的数据所需的任何其他信息可根据要求向主要联系人提供。

Acknowledgments 致谢

We are grateful to Prof. Takahiro Arima for the permission to use their human TSCs and all members of the Boroviak lab for their enthusiasm and critical discussion of the manuscript. This research was generously supported by the Wellcome Trust (WT RG89228 and WT G117980) and the Cambridge Stem Cell Institute. D.S. held a Centre for Trophoblast Graduate Studentship and was supported by the Cambridge Philosophical Society. The Gates Cambridge Trust (OPP1144) provided a PhD studentship for C.M. E. Slatery was supported by a Wellcome Trust PhD studentship (WT108438/C/15/Z). T.E.B. and C.P. were supported by the Wellcome Trust (WT RG89228 and WT G117980). Moreover, this research was funded in part, by the Wellcome Trust (203151/Z/16/Z, 203151/A/16/Z) and the UKRI Medical Research Council (MC_PC_17230). For the purpose of open access, the author has applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission. A.W. was supported by EU Horizon 2020 Marie Skłodowska-Curie Actions (ImageInLife, 721537). T.N.K, A.L.E., and F.H. were supported by the EPSRC (EP/Y032756/1) as part of the UKRI ERC guarantee scheme. M.L.-A. was supported by a Novo Nordisk Foundation grant number NNF21CC007329, an EMBO Scientific Exchange Grant, and a Company of Biologists Travelling Fellowship from Development.
我们感谢 Takahiro Arima 教授允许使用他们的人类 TSC,并感谢 Boroviak 实验室的所有成员对手稿的热情和批判性讨论。这项研究得到了Wellcome Trust (WT RG89228和 WT G117980 )和剑桥干细胞研究所的慷慨支持。 DS 设有滋养层研究生中心,并得到剑桥哲学会的支持。盖茨剑桥信托基金( OPP1144 ) 为 CME Slatery 提供了博士生奖学金,并得到了Wellcome Trust博士生奖学金 ( WT108438/C/15/Z ) 的支持。 TEB 和 CP 得到Wellcome TrustWT RG89228WT G117980 )的支持。此外,这项研究得到了Wellcome Trust ( 203151/Z/16/Z203151/A/16/Z ) 和 UKRI医学研究委员会( MC_PC_17230 ) 的部分资助。出于开放获取的目的,作者已对由此提交的任何作者接受的手稿版本应用了 CC BY 公共版权许可。 AW 得到了EU Horizo​​n 2020 Marie Skłodowska-Curie Actions (ImageInLife, 721537 )的支持。 TNK、ALE 和 FH 作为 UKRI ERC担保计划的一部分,得到了EPSRC ( EP/Y032756/1 ) 的支持。 ML-A。得到了诺和诺德基金会拨款号NNF21CC007329EMBO 科学交流拨款以及生物学家公司发展旅行奖学金的支持。

Author contributions 作者贡献

D.S. and T.E.B. conceptualized the project. D.S., C.M., T.N.K., A.L.E., S.B., A.W., and E.S. conducted experiments. A.W. carried out live imaging, T.N.K. and A.L.E. conducted encapsulations, and S.B., E. Slatery, D.S., and C.M. processed marmoset postimplantation embryos. M.L.-A., C.M., and T.N.K. generated cross-species TSC chimeras. C.M., D.S., T.N.K., T.M.R., M.L.-A., S.J.C., M.Z.-G., J.J.B., J.M.B., and W.R. performed experiments/supervised revision experiments. D.S. carried out all other experiments. R.B. provided marmoset postimplantation embryos. E. Sasaki provided marmoset PSCs and in vitro embryo culture facilities. C.P. and D.S. performed bioinformatics. T.E.B., D.S., and C.M. wrote the manuscript. T.E.B. and F.H. acquired funding and supervised the project.
DS 和 TEB 构思了该项目。 DS、CM、TNK、ALE、SB、AW 和 ES 进行了实验。 AW 进行实时成像,TNK 和 ALE 进行封装,SB、E. Slatery、DS 和 CM 处理狨猴植入后胚胎。 ML-A.、CM 和 TNK 生成跨物种 TSC 嵌合体。 CM、DS、TNK、TMR、ML-A.、SJC、MZ-G.、JJB、JMB 和 WR 进行了实验/监督修订实验。 DS 进行了所有其他实验。 RB 提供狨猴植入后胚胎。 E. Sasaki 提供狨猴 PSC 和体外胚胎培养设施。 CP 和 DS 进行生物信息学分析。 TEB、DS 和 CM 撰写了手稿。 TEB 和 FH 获得了资金并监督了该项目。

Declaration of interests 利益申报

The authors declare no competing interests.
作者声明没有竞争利益。

STAR★Methods STAR★方法

Key resources table 关键资源表

REAGENT or RESOURCESOURCEIDENTIFIER
Antibodies
Monoclonal rabbit IgG anti-AP2γAbcamab218107; RRID:AB_2891087
Polyclonal goat IgG anti-AP2γR&D SystemsAF5059; RRID:AB_2255891
Polyclonal rabbit IgG anti-KLF17Atlas AntibodiesHPA024629;RRID:AB_1668927
Monoclonal mouse IgG anti-SOX2R&D systemsMAB2018; RRID:AB_358009
Monoclonal mouse IgG anti-ZO-1Thermo Fischer33-9100; RRID:AB_87181
Polyclonal rabbit IgG anti-lamininAbcamab11575; RRID:AB_298179
Polyclonal goat IgG anti-SOX17R&D SystemsAF1924; RRID:AB_355060
Monoclonal mouse IgG anti-KRT7Abcamab9021; RRID:AB_306947
Polyclonal goat IgG anti-GATA6R&DAF1700; RRID:AB_2108901
Monoclonal rabbit IgG anti-β-cateninCell signalling8480; RRID:AB_11127855
Polyclonal rabbit IgG anti-CDH11Invitrogen71-7600; RRID:AB_2533995
Monoclonal rabbit IgG anti-PDGFRαAbcamab203491; RRID:AB_2892065
Monoclonal rabbit IgG anti-GATA3Cell signalling5852; RRID:AB_10835690
Polyclonal rabbit IgG anti-GATA2Abcamab173817
Monoclonal rabbit IgG anti-VTCN1Abcamab209242; RRID:AB_2801513
Monoclonal mouse IgG anti-OCT4Santa Cruz Biotech.SC/5279; RRID:AB_628051
Polyclonal goat IgG anti-TBXTR&D systemsAF2085; RRID:AB_2200235
Monoclonal mouse IgG anti-EzrinMerck/Sigma-AldrichE8897; RRID:AB_476955
Monoclonal mouse IgG anti-MMP2Thermo Fischer436000; RRID:AB_2532214
Monoclonal mouse IgG anti-HLA-GBio-RadMCA2043; RRID:AB_323365
Polyclonal rabbit IgG anti-human CGBAbcamab53087; RRID:AB_870731
Polyclonal rabbit IgG anti-human CGBAgilent DakoIR508
Polyclonal rabbit IgG anti-CDX2Cell Signalling3977; RRID:AB_2077043
Acti-Stain 670 PhalloidinUniversal BiologicalsPHDN1-A
Chemicals, peptides, and recombinant proteins
MAPK inhibitor PD032590Cambridge Stem Cell InstituteN/A
aPKC inhibitor Gö6983Bio-Techne2285
Basic fibroblast growth factor (bFGF)Cambridge Stem Cell InstituteN/A
Human leukaemia inhibitory factorCambridge Stem Cell InstituteN/A
Tankyrase inhibitor XAV939Cell Guidance SystemsSM38-200
Rock inhibitor Y-27632Tocris1254
L-ascorbic acidSigma-AldrichA4403
Activin ACambridge Stem Cell InstituteN/A
Epidermal growth factor (EGF)Sigma-AldrichE9644
Activin receptor inhibitor A83-01Stem cell technologies72022
TGF- β receptor inhibitor SB431542Tocris1614
Bone morphogenic protein 4 (BMP4)R&D systems314-BP
ForskolinCambridge BioscienceCAY11018
LipofectamineTM 2000 Transfection ReagentThermo Fisher Scientific11668019
Critical commercial assays
Polyethylene terephthalate membrane transwell insertsCorning353097
Total RNA Miniprep KitMonarchT2010S
High Sensitivity DNA Analysis KitAgilent5067-4626
KAPA HiFi HotStart ReadyMix 2xKAPA BiosystemsKK2601
Nextera XT DNA sample preparation kitIlluminaFC-131-1096
Nextera XT 24-index kit, 96 samplesIlluminaFC-131-1001
Deposited data
Bulk bisulfite sequencing of in vitro marmoset cellsThis paperArray Express: E-MTAB-12453
Single cell bisulfite sequencing of marmoset pre- and postimplantation embryosThis paperArray Express: E-MTAB-13893
Single cell RNA-sequencing of marmoset preimplantation embryoThis paperArray Express: E-MTAB-13895
Single cell-RNA sequencing of in vitro marmoset cellsThis paperArray Express: E-MTAB-12463
Single-cell RNAseq of cynomolgus in vitro cultured embryosMa et al.69GEO: GSE130114
Single-cell RNAseq of cynomolgus in vitro cultured embryosYang et al.74GEO: GSE148683
Single-cell RNAseq of marmoset in vivo embryosBergmann et al.34Array Express: E-MTAB-9367
Single-cell RNAseq of human in vitro cultured embryosDaSilva-Arnold et al.24GEO: GSE136447
Single-cell RNAseq of human 1st trimester trophoblastVento-Tormo et al.9Array Express: E-MTAB-6701
Single-cell RNAseq of human in vitro blastoid modelKagawa et al.80GEO: GSE177689
Single-cell RNAseq of human in vitro blastoid modelYu et al.81GEO: GSE150578
Single-cell RNAseq of human in vitro blastoid modelLiu et al.82GEO: GSE156596
Experimental models: Cell lines
Human: SHEF6Aflatoonian et al.83hPSC line
Marmoset: NEW4Kishimoto et al.84cmPSC line
Marmoset: NEW2Kishimoto et al.84cmPSC line
Marmoset: NEW2 (NLS-GFP and LifeACT-mCherry)This papercmPSC line NEW2
Marmoset: NEW4 (cerulean)This papercmPSC line NEW4
Marmoset: NEW2 (CDX2 overexpression)This papercmPSC line NEW4
Marmoset: NEW2 (OCT4 overexpression)This papercmPSC line NEW4
Experimental models: Organisms/strains
Common marmoset (Callithrix jacchus)Bergmann et al.34N/A
Mouse: CD-1Charles River LaboratoriesN/A
Oligonucleotides
Primers used in Figures 7J and S2B. See Table S1This paperN/A
Plasmid: OCT4-overexpressionThis paperN/A
Plasmid: CDX2-overexpressionThis paperN/A
Plasmid: CeruleanThis paperN/A
Plasmid: LifeACT-mCherryRiedl et al.85N/A
Plasmid: NLS-GFPThis paperN/A
Software and algorithms
Rhttps://www.R-project.org/v4.1.1
FijiSchindelin et al.86N/A
IGV V2.16.2https://data.broadinstitute.org/igv/projects/downloads/2.16/N/A
TrimGalore!https://github.com/FelixKrueger/TrimGaloreN/A
Bismark v 0.22.1https://www.bioinformatics.babraham.ac.uk/projects/bismark/N/A
Deep Tools v 3.1.3https://deeptools.readthedocs.io/en/develop/N/A
KentUtilshttps://github.com/ENCODE-DCC/kentUtilsN/A
DMRcaller v 1.4.2Catoni et al.87N/A
Seurat v3.2.0Butler et al.88; Stuart et al.89N/A
Original codeThis paperhttps://github.com/Boroviak-Lab/Trophoblast

Experimental model and study participant details
实验模型和研究参与者详细信息

Animals 动物

All animal work was carried out according to the Animals (Scientific Procedures) Act 1986 Amendment Regulations 2012 with ethical review by the University of Cambridge Animal Welfare and Ethical Review Body (AWERB).
所有动物工作均根据《1986 年动物(科学程序)法》2012 年修正案进行,并由剑桥大学动物福利和伦理审查机构 (AWERB) 进行伦理审查。
Marmoset embryo samples were originally obtained in Bergmann et al.34 with new specimens analysed in this study. Briefly, marmoset embryos were collected at the German Primate Center (Deutsches Primatenzentrum-Leibniz Institute for Primate Research) according to the German Animal Protection Law and approved by German Primate Center ethics committee. Animals were obtained from self-sustaining marmoset monkey (C.jacchus) breeding colony of the German Primate Center and housed according to the standard German Primate Center practice for common marmoset monkeys. Females were aged between 8-11 years old. Animal procedures to retrieve the marmoset embryos used in this study were approved by the Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, LAVES, under licence number 33.19-42502-04-16/2130 ‘Gewinnung früher Implantationsembryonen des Weißbüschelaffen zur molekularen Charakterisierung frühembryonaler Differenzierungsschritte bei Primaten’, which included a positive ethics statement. Uterus isolation was performed as described in Bergmann et al.34 and was carried out by specialised and experienced veterinarians.
狨猴胚胎样本最初由 Bergmann 等人获得。 34本研究中分析了新样本。简而言之,狨猴胚胎根据德国动物保护法在德国灵长类动物中心(Deutsches Primatezenzentrum-Leibniz Institute for Primate Research)采集,并经德国灵长类动物中心伦理委员会批准。动物获自德国灵长类动物中心的自我维持狨猴( C.jacchus )繁殖群,并根据德国灵长类动物中心普通狨猴的标准做法进行饲养。女性年龄在8-11岁之间。本研究中使用的狨猴胚胎取回动物程序已获得 Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, LAVES 的批准,许可证号为 33.19-42502-04-16/2130 'Gewinnung früher Implantationsembryonen des Weißbüschelaffen zur molecularkularen Charakterisierung frühembryon灵长类动物的区分,其中包括积极的道德声明。子宫隔离按照 Bergmann 等人的描述进行。 34由专业且经验丰富的兽医进行。
Mice used in this study were kept according to national and international guidelines in the animal facility of the University of Cambridge. Mouse embryos were obtained from Charles River through natural mating of CD1 wildtype mice and culled at the time of embryo collection by cervical dislocation. Mice used for wildtype embryo matings were 6 weeks old.
本研究中使用的小鼠根据国家和国际指南饲养在剑桥大学的动物设施中。小鼠胚胎通过CD1野生型小鼠的自然交配从Charles River获得,并在胚胎收集时通过颈脱位法进行剔除。用于野生型胚胎交配的小鼠是六周大的。

Cell lines 细胞系

Embryo-derived conventional marmoset PSCs lines New2 (female) and New4 (female) were provided by E. Sasaki. Marmoset PSCs were maintained in KSR/bFGF medium, which is comprised of Dulbecco’s modified Eagle medium (DMEM)/F12 (21331, Gibco) supplemented with 20% Knockout Serum Replacement (KSR) (10828028, Thermo Fisher Scientific), 1% GlutaMAX (35050061, Thermo Fisher Scientific), 1% MEM non-essential amino acids (11140035, Thermo Fisher Scientific), 100 μM β-mercaptoethanol (21985023, Thermo Fisher Scientific) and 10 ng/mL bFGF (Cambridge Stem Cell Institute). Cells were routinely cultured on mitomycin C (M4287, Sigma-Aldrich) inactivated mouse embryonic fibroblast (MEF) feeder cells (Cambridge Stem Cell Institute) under 5% O2 and 5% CO2 at 37 °C. The medium was changed daily and cells were passaged every 2–4 days by dissociation with Accutase (00-4555-56, Thermo Fisher Scientific) for 5 min.
胚胎来源的常规狨猴 PSC 系 New2(雌性)和 New4(雌性)由 E. Sasaki 提供。狨猴 PSC 维持在 KSR/bFGF 培养基中,该培养基由 Dulbecco 改良 Eagle 培养基 (DMEM)/F12(21331,Gibco)组成,辅以 20% 敲除血清替代品 (KSR)(10828028,Thermo Fisher Scientific)、1% GlutaMAX( 35050061,Thermo Fisher Scientific)、1% MEM 非必需氨基酸(11140035,Thermo Fisher Scientific)、100 μM β-巯基乙醇(21985023,Thermo Fisher Scientific)和 10 ng/mL bFGF(剑桥干细胞研究所)。将细胞常规培养在丝裂霉素C(M4287,Sigma-Aldrich)灭活的小鼠胚胎成纤维细胞(MEF)饲养细胞(剑桥干细胞研究所)上,在37℃、5% O 2和5% CO 2下。每天更换培养基,并每 2-4 天通过 Accutase(00-4555-56,Thermo Fisher Scientific)解离 5 分钟对细胞进行传代。
All human PSCs experiments were approved by the UK Stem Cell Bank Steering Committee and comply with the regulations of the UK Code of Practice for the use of Human Stem Cell Lines. The human embryonic stem cell line SHEF683 was provided by the laboratory of A. Smith with ethical approval from the UK Stem Cell Bank. Conventional SHEF690 were cultured on vitronectin-coated dishes (10 μg/ml; A14700, Thermo Fisher Scientific) in Essential 8 (E8) medium (A1517001, Thermo Fisher Scientific) under hypoxic conditions (37 °C, 5% CO2, 5% O2). Cells were routinely passaged in clumps using 0.5 mM EDTA (AM9261, Invitrogen). All stem cells were karyotyped, routinely tested for mycoplasma contamination and authenticated by RNA-seq.
所有人类 PSC 实验均获得英国干细胞库指导委员会的批准,并符合英国关于使用人类干细胞系的实践规范的规定。人类胚胎干细胞系SHEF6 83由A. Smith实验室提供,并获得英国干细胞银行伦理批准。常规 SHEF6 90在低氧条件(37 °C、5% CO 2 、5 % O 2 )。使用 0.5 mM EDTA(AM9261,Invitrogen)将细胞常规传代成团。所有干细胞均经过核型分析、常规支原体污染检测并通过 RNA-seq 进行验证。

Method details 方法详情

Marmoset and human PSC resetting
狨猴和人类 PSC 重置

Marmoset PSCs were reset chemically as previously described in Bergmann et al.34 In short, conventional marmoset PSCs were seeded as clumps of 2–5 cells one day before resetting at 50,000 cells per well of a 12-well plate (1.3 × 104 cells per cm2) on MEFs. After 24 h, the medium was changed to PLAXA, which is comprised of N2B27 (Ndiff; Y40002,Takara Bio) medium supplemented with 1% chemically defined lipids (11905031, Gibco), 1 μM PD0325901 (Cambridge Stem Cell Institute), 10 ng/ml recombinant human leukaemia inhibitory factor (LIF; Cambridge Stem Cell Institute), 50 μg/ml L-ascorbic acid (A4403, Sigma-Aldrich), 2 μM XAV939 (SM38-200, Cell Guidance Systems) and 20 ng/ml activin A (Cambridge Stem Cell Institute). Throughout conversion, cells were passaged with Accutase 1:1.5 every 3–4 days. Dome-shaped colonies first emerge at days 4–5 and naive conversion is complete by day 9.
如 Bergmann 等人先前所述,对狨猴 PSC 进行化学重置。 34简而言之,传统的狨猴 PSC 以 2-5 个细胞的团块形式接种,一天后以 12 孔板每孔 50,000 个细胞(每 cm 2 1.3 × 10 4 个细胞)的密度重置在 MEF 上。 24小时后,将培养基更换为PLAXA,其由N2B27(Ndiff;Y40002,Takara Bio)培养基组成,补充有1%化学限定脂质(11905031,Gibco)、1μM PD0325901(剑桥干细胞研究所)、10ng /ml 重组人白血病抑制因子(LIF;剑桥干细胞研究所)、50 μg/ml L-抗坏血酸(A4403,Sigma-Aldrich)、2 μM XAV939(SM38-200,Cell Guidance Systems)和 20 ng/ml 激活素A(剑桥干细胞研究所)。在整个转化过程中,细胞每 3-4 天用 Accutase 1:1.5 进行传代。圆顶形菌落首先在第 4-5 天出现,幼稚转化在第 9 天完成。
SHEF6 human PSCs were reset chemically using the protocol established by the Smith laboratory.68,91 Cells were treated with 1 μM PD0325901, 10 ng/ml human LIF and 1 mM valproic acid (VPA; P4543, Sigma-Aldrich) in N2B27 for 3 days on inactivated MEFs. Then, the medium was changed to PXGL: 1 μM PD0325901, 2 μM XAV939, 2 μM Gö6983 (2285, Bio-Techne) and 10 ng/mL human LIF in N2B27. 10 μM Y-27632 (1254, Tocris) was added during initial resetting in some cultures for the first 10 days.
使用史密斯实验室制定的方案对 SHEF6 人类 PSC 进行化学重置。 68 91在灭活的 MEF 上,用 N2B27 中的 1 μM PD0325901、10 ng/ml 人 LIF 和 1 mM 丙戊酸(VPA;P4543,Sigma-Aldrich)处理细胞 3 天。然后,将培养基更换为 PXGL:N2B27 中的 1 μM PD0325901、2 μM XAV939、2 μM Gö6983(2285,Bio-Techne)和 10 ng/mL 人 LIF。在某些培养物的前 10 天初始重置期间添加 10 μM Y-27632(1254,Tocris)。

Trophoblast stem cells derivation from PSCs
PSC 衍生的滋养层干细胞

Naive PSCs were prepared as previously described on MEFs and cultured to 60-70% confluency. Naive PSCs media was removed and PSCs were washed in PBS (D8537, Sigma-Aldrich). Media was replaced with indicated culture conditions and differentiated for 5 days.
如前所述在 MEF 上制备初始 PSC,并培养至 60-70% 汇合。除去初始 PSC 培养基,并在 PBS(D8537,Sigma-Aldrich)中洗涤 PSC。用指定的培养条件更换培养基并分化 5 天。
OK conditions were as previously described in Okae et al.20 Other conditions include OK XAV PD: advanced DMEM/F12 (12634010, Thermo Fischer Scientific) supplemented with 55 μM β-mercaptoethanol, 0.3% bovine serum albumin (BSA; A9418, Sigma-Aldrich), 1% ITS-X supplement (41400045, Gibco), 1.5 mg/ml L-ascorbic acid, 50 ng/ml EGF (E9644, Sigma-Aldrich), 2 μM XAV939, 1 μM PD0325901, 0.5 μM A83-01 (72022, Stem cell technologies), 1 μM SB431542 (1614, Tocris), 0.7 mM VPA and 1 μM Y27632 and PAVS: advanced DMEM/F12 supplemented with 0.3% BSA, 1% ITS-X supplement, 55 μM β-mercaptoethanol, 1% chemically defined lipids, 1x GlutaMAX, 1 μM PD0325901, 0.5 μM A83-01, 1μM SB431542, 0.7 mM VPA and 1 μM Y27632. Both conditions were cultured on inactive MEFs unless otherwise specified. 5 μg/ml collagen IV (C5533, Sigma-Aldrich) coating was generated by resuspending collagen in PBS and coating at 37°C for 2 hours. Additional molecules added include 2 μM XAV939, 20 ng/mL BMP4 (314-BP, R&D systems), 2 μM CHIR99021 (CHIR; Cambridge Stem Cell Institute), and 20 μM forskolin (CAY11018, Cambridge Bioscience). Media was changed every 24 hours.
OK 条件如之前 Okae 等人所述。 20其他条件包括 OK XAV PD:高级 DMEM/F12(12634010,Thermo Fischer Scientific)补充有 55 μM β-巯基乙醇、0.3% 牛血清白蛋白(BSA;A9418,Sigma-Aldrich)、1% ITS-X 补充剂(41400045) ,Gibco),1.5 mg/ml L-抗坏血酸,50 ng/ml EGF(E9644,Sigma-Aldrich),2 μM XAV939,1 μM PD0325901,0.5 μM A83-01(72022,干细胞技术),1 μM SB431542 (1614,Tocris),0.7 mM VPA 和 1 μM Y27632 和 PAVS:高级 DMEM/F12,补充有 0.3% BSA、1% ITS-X 补充剂、55 μM β-巯基乙醇、1% 化学成分确定的脂质、1x GlutaMAX、1 μM PD0325901、0.5 μM A83-01、1μM SB431542、0.7 mM VPA 和 1 μM Y27632。除非另有说明,这两种条件均在非活性 MEF 上培养。通过将胶原蛋白重悬于 PBS 中并在 37°C 下涂覆 2 小时,生成 5 μg/ml IV 型胶原蛋白(C5533,Sigma-Aldrich)涂层。添加的其他分子包括 2 μM XAV939、20 ng/mL BMP4(314-BP,R&D 系统)、2 μM C​​HIR99021(CHIR;剑桥干细胞研究所)和 20 μM 毛喉素(CAY11018,剑桥生物科学)。介质每 24 小时更换一次。
After 4-5 days or at 90% confluency, epithelial colonies were washed with PBS and dissociated with TrypLE (12605010, Thermo Fisher Scientific) for 6-8 minutes or until cells balled up and formed 2-3 cell clumps. TrypLE was carefully removed to not disturb cells and cells were resuspended and replated in culture media supplemented with 10 μM Y-27632 for 24 hours. Marmoset TSCs were passaged 1:2 every 7 days or until 80% confluence was reached.
4-5天后或达到90%汇合后,用PBS洗涤上皮集落,并用TrypLE(12605010,Thermo Fisher Scientific)解离6-8分钟,或直至细胞聚集并形成2-3个细胞团块。小心地取出 TrypLE,以免干扰细胞,将细胞重悬并重新铺在补充有 10 μM Y-27632 的培养基中 24 小时。狨猴 TSC 每 7 天以 1:2 传代,或直至达到 80% 汇合。

Overexpression marmoset PSC lines generation
过度表达狨猴 PSC 系的生成

Overexpression plasmids were designed in the Vectorbuilder vector design tool and were transfected with a pBase plasmid containing the piggybac transposase. LifeACT plasmid is as described in Riedl et al.85
Vectorbuilder载体设计工具中设计过表达质粒,并用含有 Piggybac 转座酶的 pBase 质粒转染。 LifeACT 质粒如 Riedl 等人所述。 85
OCT4 OE: pPB[TetOn]-TRE>cmPOU5F1:BGH pA:CMV>Puro-rev(CAG>tTS: T2A:rtTA)
OCT4 OE: pPB[TetOn]-TRE>cm POU5F1 :BGH pA:CMV>Puro-rev(CAG>tTS: T2A:rtTA)
CDX2 OE: pPB[TetOn]-TRE>cmCDX2:BGH pA:CMV>Neo-rev(CAG>tTS: T2A:rtTA)
NLS-GFP: pPB[Exp]-Puro-CAG>NLS-EGFP
LifeACT-mCherry: pPB[Exp]-Blast-CAG> LifeACT-mCherry
LifeACT-mCherry:pPB[Exp]-Blast-CAG> LifeACT-mCherry
Cerulean: pPB[Exp]-Puro-CAG>Cerulean
天蓝色:pPB[Exp]-Puro-CAG>Cerulean
Solution 1: 250μL of OptiMEM (31985062, Thermo Fisher Scientific) containing 5 μg of the gene plasmid and 2.5 μg of pBase. Solution 2: 250 μL of OptiMEM with 15 μL of LipofectamineTM 2000 Transfection Reagent (11668019, Thermo Fisher Scientific). Solutions were incubated separately at room temperature for 5 minutes before being combined and incubated at room temperature for 30 minutes. Marmoset PSCs were cultured in 6-well culture dishes to 75-80% confluency and 2mL of fresh media was placed on top of cells. Combined solution was then evenly distributed drop-wise over the marmoset PSCs. Cells and transfection solutions were incubated overnight. Medium was replaced with fresh culture medium the next day and cells were allowed to recover for 24-48 hours. Culture medium supplemented with either 2.5 μg/mL of puromycin (Cambridge Stem Cell Institute), 1 μg/mL of blasticidin (Cambridge Stem Cell Institute) or 200 μg/mL G418 (MIR 5920, Cambridge Bioscience) for selection of transfected colonies for 48 hours before cells were passaged. Plasmid-specific gene expression of the gene of interested was assayed via qPCR to confirm successful transfection. Dox-inducible gene expression was promoted by supplementing culture medium with 1 μg/mL doxycycline (CAY14422, Cambridge Bioscience) for a minimum of 2 passages.
溶液 1:250μL OptiMEM(31985062,Thermo Fisher Scientific),含有 5 μg 基因质粒和 2.5 μg pBase。溶液 2:250 μL OptiMEM 和 15 μL Lipofectamine TM 2000 转染试剂(11668019,Thermo Fisher Scientific)。将溶液在室温下单独孵育5分钟,然后合并并在室温下孵育30分钟。狨猴 PSC 在 6 孔培养皿中培养至 75-80% 汇合,并将 2mL 新鲜培养基置于细胞顶部。然后将合并的溶液逐滴均匀地分布在狨猴 PSC 上。将细胞和转染溶液孵育过夜。第二天用新鲜培养基更换培养基,并使细胞恢复24-48小时。培养基中添加 2.5 μg/mL 嘌呤霉素(剑桥干细胞研究所)、1 μg/mL 杀稻瘟菌素(剑桥干细胞研究所)或 200 μg/mL G418(MIR 5920,剑桥生物科学),用于选择 48 个转染菌落细胞传代前数小时。通过 qPCR 检测感兴趣基因的质粒特异性基因表达,以确认转染成功。通过在培养基中添加 1 μg/mL 多西环素(CAY14422,Cambridge Bioscience)至少 2 代来促进 Dox 诱导的基因表达。

Trophoblast spheroids generation and culture
滋养层球体的生成和培养

Cells were encapsulated as previously described.54,64,65,92 In short, microfluidic devices were designed with one inlet for aqueous low-melting agarose with cells and one for a continuous oil phase. Cells were resuspended at 1.5 × 106 cells/100 μL in PBS with 3% BSA. The cell suspension was mixed 1:1 with low-melting-point agarose (50302, Lonza) solution at 37°C. HFE-7500 (Fluorochem) supplemented with surfactant (0.3%; C022, Pico-Surf by Sphere Fluidics) was used as the continuous oil phase. Syringes (SGE Analytical Science) controlled by automated pumps (CETONI, neMYSIS) injected the agarose-cell suspension separately from the oil-surfactant solution into the microfluidic chips for emulsification. Agarose droplets left the microfluidic chip through the outlet and were collected on ice for polymerization. For demulsification, 200 μL medium and 45 μL 1H,1H,2H,2H-perfluoro-1-octanol (B20156.18, AlfaAesar) were used.
如前所述封装细胞。 54 64 65 92简而言之,微流体装置设计有一个入口用于含有细胞的水性低熔点琼脂糖,另一个入口用于连续油相。将细胞以 1.5 × 10 6 个细胞/100 μL 重悬于含有 3% BSA 的 PBS 中。将细胞悬浮液与低熔点琼脂糖(50302,Lonza)溶液在 37°C 下按 1:1 混合。使用补充有表面活性剂(0.3%;CO22,Sphere Fluidics 的 Pico-Surf)的 HFE-7500(Fluorochem)作为连续油相。由自动泵(CETONI、neMYSIS)控制的注射器(SGE Analytical Science)将琼脂糖细胞悬浮液与油表面活性剂溶液分开注入微流控芯片中进行乳化。琼脂糖液滴通过出口离开微流控芯片并收集在冰上用于聚合。为了破乳,使用 200 μL 培养基和 45 μL 1H,1H,2H,2H-全氟-1-辛醇(B20156.18,AlfaAesar)。
Encapsulated marmoset and human TSCs were cultured in indicated medium supplemented with 1% penicillin-streptomycin (15140122, ThermoFisher Scientific), 1% Matrigel for non-invasion assays (354230, Corning) and 10 μM Y-27632. Microgel-suspension cultures were cultured at 37°C under hypoxic conditions (5% CO2 and 5% O2). Marmoset TSC spheroids were cultured in the presence of MEFs. Media was topped up every other day with twice the amount of media. On alternate days, media was completely changed by spinning microgels down at 200g for 5 minutes. Supernatant was carefully removed, leaving at least 0.5 mL of media per well. Microgels were resuspended and transferred back to the same well. On day 4-5 (depending on structure growth), structures were released from microgels by incubation with 1 U/ml Agarase (EO0461, Thermo Fisher Scientific) in N2B27 medium for 5 min at 37°C, followed by gentle pipetting up and down to free structures from digested agarose.
封装的狨猴和人类 TSC 在补充有 1% 青霉素-链霉素(15140122,ThermoFisher Scientific)、1% 用于非侵入测定的基质胶(354230,Corning)和 10 μM Y-27632 的指定培养基中培养。微凝胶悬浮培养物在37°C、低氧条件(5% CO 2和5% O 2 )下培养。在 MEF 存在的情况下培养狨猴 TSC 球体。每隔一天用两倍的介质补充介质。每隔一天,通过以 200g 的速度旋转微凝胶 5 分钟来完全更换培养基。小心地去除上清液,每孔至少留下 0.5 mL 的介质。将微凝胶重新悬浮并转移回同一孔中。第 4-5 天(取决于结构生长),通过在 N2B27 培养基中于 37°C 下与 1 U/ml 琼脂酶(EO0461,Thermo Fisher Scientific)一起孵育 5 分钟,然后轻轻上下吹打,从微凝胶中释放结构从消化的琼脂糖中释放结构。
Hanging drops were generated by gently dissociating cells with TrypLE for 6-8 minutes, ensuring cell clumps remained. TrypLE was removed carefully to not disturb dissociated cells. Cells were gently resuspended in the appropriate TSC media and plated in 10-15 μL droplets on the lid of a 15 cm dish. The bottom of 15 cm dish was filled with distilled water and the lid was carefully flipped. Hanging droplets were cultured for 2-3 days. 10 μL droplets were given 10 μL of fresh media on day 2.
用 TrypLE 轻轻分离细胞 6-8 分钟,产生悬滴,确保细胞团块保留。小心去除 TrypLE,以免干扰解离的细胞。将细胞轻轻重悬于适当的 TSC 培养基中,并以 10-15 μL 液滴涂在 15 cm 培养皿的盖子上。 15厘米培养皿底部装满蒸馏水,小心地翻转盖子。悬滴培养2-3天。第 2 天,向 10 μL 液滴添加 10 μL 新鲜培养基。

Embryo cryosections 胚胎冷冻切片

Pregnant marmoset uteri were obtained as described in Bergmann et al. (2022) and were embedded unfixed into optimum cutting temperature (OCT) compound (4583, TissueTek) and snap-frozen.34 Each OCT block containing uteri with implanted embryos was sectioned fully at a thickness of 12 μm using a Leica cryostat microtome (CM3050) to obtain consecutive slices of the whole organ. Sections containing embryo tissue were collected either on Naphthalate (PEN) membrane slides (Zeiss, 1.0PEN) or histological slides (Superfrost Plus, Thermo Fisher Scientific) for LCM and immunostaining analysis, respectively, and immediately transferred to dry ice.
按照 Bergmann 等人的描述获得怀孕的狨猴子宫。 (2022) 并未固定地嵌入最佳切割温度 (OCT) 化合物 (4583, TissueTek) 中并速冻。 34使用徕卡低温切片机 (CM3050) 将每个包含植入胚胎的子宫的 OCT 块以 12 μm 的厚度完全切片,以获得整个器官的连续切片。将含有胚胎组织的切片收集在萘二甲酸盐(PEN)膜载玻片(Zeiss,1.0PEN)或组织学载玻片(Superfrost Plus,Thermo Fisher Scientific)上,分别用于 LCM 和免疫染色分析,并立即转移至干冰。

Cross-species marmoset TSCs and mouse chimera
跨物种狨猴 TSC 和小鼠嵌合体

Zygotes were collected at E0.5 by removing the cumulus cells with brief incubation in hyaluronidase (H4272, Sigma-Aldrich). Embryos were then cultured for 2 days in EmbryoMax Advanced KSOM medium (MMR106D, Sigma-Aldrich) until they reached 8-cell stage. For aggregation chimeras, zona pellucidae were removed from the embryos by brief incubation in acidic Tyrode’s solution (T1788, Sigma-Aldrich) for 30 sec - 1 min at RT.
通过去除卵丘细胞并在透明质酸酶(H4272,Sigma-Aldrich)中短暂孵育,在 E0.5 收集受精卵。然后将胚胎在 EmbryoMax Advanced KSOM 培养基(MMR106D,Sigma-Aldrich)中培养 2 天,直至达到 8 细胞阶段。对于聚集嵌合体,通过在酸性 Tyrode 溶液(T1788,Sigma-Aldrich)中在室温下短暂孵育 30 秒至 1 分钟,从胚胎中去除透明带。
Marmoset cerulean-tagged periTSCs were washed with PBS and dissociated with TrypLE for 10-15 min. TrypLE was carefully removed and cells were resuspended in N2B27. Clumps of 2-3 cells were rinsed in KSOM (MR-101, Sigma-Aldrich) and manually aggregated with zona pellucida-depleted 8-cell stage mouse embryos in microwells generated with a Hungarian darning needle. Aggregated embryos were then cultured in vitro in EmbryoMax Advanced KSOM overlayed with mineral oil for 48h, until E4.5 equivalent. All embryo culture steps were performed in 5% CO2 at 37°C.
用 PBS 清洗带有天蓝色标签的狨猴 periTSC,并用 TrypLE 解离 10-15 分钟。小心地除去 TrypLE,并将细胞重悬于 N2B27 中。将 2-3 个细胞团块在 KSOM(MR-101,Sigma-Aldrich)中冲洗,并与透明带耗尽的 8 细胞阶段小鼠胚胎手动聚集在用匈牙利织补针生成的微孔中。然后将聚集的胚胎在铺有矿物油的 EmbryoMax Advanced KSOM 中体外培养 48 小时,直至 E4.5 相当。所有胚胎培养步骤均在 37°C、5% CO 2中进行。

EVT differentiation and migration assay
EVT分化和迁移测定

Marmoset TSCs and human OK cells were differentiated to EVT according to the protocol described in Okae et al. (2018).20 Medium was changed every 24h for marmoset TSCs. After 6 days of differentiation, cells were further analysed or assayed for migratory properties.
根据 Okae 等人描述的方案将狨猴 TSC 和人类 OK 细胞分化为 EVT。 (2018)。 20对于狨猴 TSC,每 24 小时更换一次培养基。分化 6 天后,进一步分析或测定细胞的迁移特性。
Cells were cultured in 5% CO2 and 5% O2 at 37°C.
细胞在37°C、5% CO 2和5% O 2中培养。
Migration was assessed by the ability of TSCs to cross polyethylene terephthalate membrane transwell inserts (8 μm pore size; 353097, Corning). TSCs were dissociated using TrypLE for 10–15 min and resuspended in the appropriate medium. Media used for control periTSCs, control human TSCs and EVT cells are PAVS (described previously), OK (described in Okae et al.20) and EVT medium without Matrigel (advanced DMEM/F12 with 0.1 mM β-mercaptoethanol, 0.3% BSA, 1% ITS-X supplement, 7.5 μM A83-01, 2.5 μM Y27632, 4% KSR as described in Okae et al.,20 respectively. TSCs were placed in the upper chamber (50 000 cells per insert) and the lower compartment was filled with medium supplemented with 10% FBS. After 24 hours, media in the upper and bottom chambers were replaced with the respective fresh media without and with FBS respectively. The following day, non-migratory cells on the upper surface of the membrane were wiped away with a cotton swab and migratory cells at the bottom of the membrane were fixed with 4% paraformaldehyde (PFA; 15714S, Electron Microscopy Sciences/Thermo Fisher Scientific) for 15 min at room temperature.
通过 TSC 穿过聚对苯二甲酸乙二醇酯膜 Transwell 插入物(8 μm 孔径;353097,康宁)的能力来评估迁移。使用 TrypLE 解离 TSC 10-15 分钟,然后重悬于适当的培养基中。用于对照 periTSC、对照人 TSC 和 EVT 细胞的培养基是 PAVS(先前描述)、OK(Okae 等人20中描述)和不含基质胶的 EVT 培养基(高级 DMEM/F12,含有 0.1 mM β-巯基乙醇、0.3% BSA、 1% ITS-X 补充剂、7.5 μM A83-01、2.5 μM Y27632、4% KSR(如 Okae 等人所述),分别将20 个TSC 放置在上室(每个插入物 50 000 个细胞)中,下室放置在下室中。 24小时后,分别用不含和含FBS的新鲜培养基替换上室和下室中的培养基。第二天,擦拭膜上表面上的非迁移细胞。用棉签除去膜底部的迁移细胞,用 4% 多聚甲醛(PFA;15714S,Electron Microscopy Sciences/Thermo Fisher Scientific)在室温下固定 15 分钟。

Trophoblast attachment and invasion assay
滋养层附着和侵袭测定

Ibidi μ-Slide 8 wells (80806, Ibidi) were coated with indicated ECM. Thin collagen IV coating was generated by incubating 5 μg/ml human placental collagen IV in PBS for 2 hours at 37°C. Thin Matrigel coatings were generated by incubating wells with 1% Matrigel in DMEM for 2 hours at 37°C. Thick Matrigel beds were generated by pre-chilling pipette tips and μ-Slide 8 wells to -20°C. 50μL of Matrigel was pipetted into each chilled well, transferred back into their packaging and spun down in a plate spinner for 30s to create a flat bed. Matrigel-coated Ibidi μ-Slide 8 wells were incubated at 37°C for 1 hour.
Ibidi μ-Slide 8 孔(80806,Ibidi)涂有指定的 ECM。将 5 μg/ml 人胎盘 IV 型胶原蛋白在 PBS 中于 37°C 下孵育 2 小时,形成薄的 IV 型胶原蛋白涂层。通过在 DMEM 中将孔与 1% 基质胶在 37°C 下孵育 2 小时来生成薄基质胶涂层。通过将移液器吸头和 μ-Slide 8 孔预冷至 -20°C 来生成厚基质胶床。将 50μL 基质胶移入每个冷冻孔中,转移回包装中,并在平板旋转器中旋转 30 秒以形成平坦床。将基质胶包被的 Ibidi μ-Slide 8 孔在 37°C 下孵育 1 小时。
Structures were allowed to attach to the ECM coating/beds over 24–48 hours. After attachment, media was changed every 24 hours. Implanted structures were fixed in 4% PFA and treated for IF staining as described below for 2D culture, with special care taken to not disrupt attached structures.
允许结构在 24-48 小时内附着到 ECM 涂层/床上。连接后,每 24 小时更换一次介质。将植入的结构固定在 4% PFA 中,并按照下文所述的 2D 培养进行 IF 染色处理,特别注意不要破坏附着的结构。

Quantificative polymerase chain reaction
定量聚合酶链式反应

RNA extraction was performed using a Total RNA Miniprep Kit (T2010S, Monarch). Complementary DNA was obtained with GoScript Reverse Transcriptase (A5003, Promega). qPCR was performed with SYBR green PCR Master Mix (4309155, Thermo Fisher Scientific) in a StepOnePlus Real-Time PCR machine (Applied Biosystems). Results were normalized to the geometric mean of UBC and ACTB using the dCt method93 Primer sequences can be found in Table S1.
使用 Total RNA Miniprep Kit (T2010S, Monarch) 进行 RNA 提取。使用 GoScript 逆转录酶(A5003,Promega)获得互补 DNA。 qPCR 使用 SYBR green PCR Master Mix(4309155,Thermo Fisher Scientific)在 StepOnePlus 实时 PCR 机(Applied Biosystems)中进行。使用 dCt 方法将结果归一化为 UBC 和 ACTB 的几何平均值。 93引物序列可在表 S1中找到。

Immunocytochemistry 免疫细胞化学

Cryosection slides were thawed at room temperature and fixed for 8 min in 4% PFA/PBS solution. In vitro cells were cultured in Ibidi μ-Slide 8-wells and fixed with 4% PFA in PBS for 10 min at room temperature. Samples were washed three times with PBS and permeabilised with 0.25% Triton X-100 (13444259, Thermo Fisher Scientific) in PBS for 30 min. Slides and in vitro cells were blocked in 2% donkey serum (C06SB, Bio-Rad), 0.1% BSA, 0.01% Tween-20 (P9416-100, Thermo Fisher Scientific) in PBS for 30 min and incubated with primary antibody solution overnight 4°C. Slides were kept in a humidified chamber for incubation steps. Tile scanning was performed to image the uterine cavity with embryo in its entirety. Tile-scanned images were automatically merged by the acquisition software.
将冷冻切片载玻片在室温下解冻并在 4% PFA/PBS 溶液中固定 8 分钟。体外细胞在 Ibidi μ-Slide 8 孔中培养,并用 4% PFA 的 PBS 溶液在室温下固定 10 分钟。用 PBS 洗涤样品 3 次,并用 PBS 中的 0.25% Triton X-100(13444259,Thermo Fisher Scientific)透化 30 分钟。将载玻片和体外细胞在 PBS 中的 2% 驴血清(C06SB,Bio-Rad)、0.1% BSA、0.01% Tween-20(P9416-100,Thermo Fisher Scientific)中封闭 30 分钟,并与一抗溶液一起孵育过夜4°C。将载玻片保存在加湿室中进行孵育步骤。进行平铺扫描以对带有胚胎的子宫腔进行完整成像。平铺扫描图像由采集软件自动合并。
Secondary antibodies supplemented with nuclear-staining DAPI (4′,6-diamidino-2-phenylindole, Sigma-Aldrich) in blocking buffer were applied after washing steps (three times with PBS) and incubated for 60 min at room temperature. Slides were rinsed and mounted using Vectashield mounting medium (H-1200, Vector laboratories) and coverslips (12343138, Thermo Fisher Scientific).
洗涤步骤(用 PBS 三次)后应用在封闭缓冲液中补充有核染色 DAPI(4',6-二脒基-2-苯基吲哚,Sigma-Aldrich)的二抗,并在室温下孵育 60 分钟。使用 Vectashield 封固剂(H-1200,Vector 实验室)和盖玻片(12343138,Thermo Fisher Scientific)冲洗并封固载玻片。
Mouse embryos were fixed with 4% PFA for 15 min at room temperature. They were washed 3 times in PBS supplemented with 3 mg/mL polyvinylpyrolidone (PBS/PVP; PVP10, Sigma-Aldrich) and permeabilised in 0.25% Triton X-100 PBS/PVP for 30 min. Blocking was achieved with PBS containing 2% donkey serum, 0.1% BSA and 0.01% Tween-20 for 60 min. Embryos were incubated with primary antibodies in blocking buffer overnight at 4°C. Subsequently, embryos were washed 3 times in blocking buffer for a minimum of 15 min each time before incubation with the secondary antibodies for 2h at room temperature in the dark.
小鼠胚胎用 4% PFA 在室温下固定 15 分钟。将它们在补充有 3 mg/mL 聚乙烯吡咯烷酮(PBS/PVP;PVP10,Sigma-Aldrich)的 PBS 中洗涤 3 次,并在 0.25% Triton X-100 PBS/PVP 中透化 30 分钟。使用含有 2% 驴血清、0.1% BSA 和 0.01% Tween-20 的 PBS 封闭 60 分钟。将胚胎与一抗在封闭缓冲液中在 4°C 下孵育过夜。随后,将胚胎在封闭缓冲液中洗涤3次,每次至少15分钟,然后与二抗在室温下避光孵育2小时。
Fixed transwell membranes were washed 3 times with PBS and permeabilised with 0.25% Triton X-100 in PBS for 30 min. Following a 30 min incubation in blocking solution (2% donkey serum, 0.1% BSA and 0.01% Tween-20), membranes were transferred into blocking solution supplemented with DAPI for 2h at room temperature. Membranes were rinsed 3 times with blocking buffer and mounted using Vectashield mounting medium and coverslips.
固定的 Transwell 膜用 PBS 洗涤 3 次,并用 PBS 中的 0.25% Triton X-100 透化 30 分钟。在封闭液(2% 驴血清、0.1% BSA 和 0.01% Tween-20)中孵育 30 分钟后,将膜转移到补充有 DAPI 的封闭液中,在室温下孵育 2 小时。用封闭缓冲液冲洗膜 3 次,并使用 Vectashield 封固剂和盖玻片封固。
Antibody details are listed in key resources table. Anti-human CGB antibodies ab53087 (Abcam) and IR508 (Agilent, Dako) were used to detect CGB in marmoset and human samples, respectively.
抗体详细信息列于关键资源表中。抗人 CGB 抗体 ab53087 (Abcam) 和 IR508 (Agilent, Dako) 分别用于检测狨猴和人类样本中的 CGB。

Live imaging 实时成像

Cells were plated in Ibidi μ-Slide 8 wells and given 400 μL of media to enable extended culture. Cells with larger nuclei or multiple nuclei were preferentially imaged. Cells were imaged at 30-minute intervals on a Leica multiphoton SP8 microscope with a 25x water objective using a z-step size of 1 μm. For excitation, 488 nm and 555 nm lasers were used.
将细胞铺板于 Ibidi μ-Slide 8 孔中,并加入 400 μL 培养基以进行扩展培养。优先对具有较大细胞核或多细胞核的细胞进行成像。使用 Leica 多光子 SP8 显微镜,使用 25 倍水物镜,使用 1 μm 的 z 步长,以 30 分钟的间隔对细胞进行成像。使用 488 nm 和 555 nm 激光进行激发。

SEM imaging 扫描电镜成像

SEM imaging was performed by the Cambridge advanced imaging centre. Cells were grown on glass coverslips of 13 mm diameter. Cells were fixed for at least 24h in fixative (2% glutaraldehyde/2% formaldehyde in 0.05 M sodium cacodylate buffer pH 7.4) at 4°C. Coverslips were briefly dipped twice in cold deionised water (DIW) and plunge-frozen by dipping into liquid nitrogen-cooled ethane. Samples were transferred to liquid nitrogen-cooled brass inserts and freeze-dried overnight in a liquid nitrogen-cooled turbo freeze-drier (Quorum K775X).
SEM 成像由剑桥高级成像中心进行。细胞在直径 13 mm 的玻璃盖玻片上生长。将细胞在固定液(0.05 M 二甲胂酸钠缓冲液 pH 7.4 中的 2% 戊二醛/2% 甲醛)中于 4°C 固定至少 24 小时。将盖玻片短暂浸入冷去离子水 (DIW) 中两次,然后浸入液氮冷却的乙烷中进行急速冷冻。将样品转移至液氮冷却的黄铜插入物中,并在液氮冷却的涡轮冷冻干燥机 (Quorum K775X) 中冷冻干燥过夜。
Sections were imaged in a Verios 460 SEM (FEI/Thermo Fisher Scientific) run at an accelerating voltage of 4 keV and 0.2 nA probe current using the concentric backscatter detector in field-free mode (low magnification) or immersion mode (high resolution).
使用 Verios 460 SEM(FEI/Thermo Fisher Scientific)在无场模式(低放大率)或浸没模式(高分辨率)下使用同心反向散射检测器在 4 keV 加速电压和 0.2 nA 探针电流下运行,对切片进行成像。

Single cell RNA sequencing
单细胞RNA测序

Transcriptomic characterization of pre- and postimplantation marmoset trophoblast utilized previously published marmoset in vivo single-cell RNA sequencing dataset.34
利用先前发表的狨猴体内单细胞 RNA 测序数据集对植入前和植入后狨猴滋养层进行转录组表征。 34
Naive derived OK, postTSCs and periTSCs were transferred using glass capillaries into individual tubes containing RLT buffer (1053393, Qiagen) and immediately frozen in dry ice.
使用玻璃毛细管将原始衍生的 OK、postTSC 和 periTSC 转移到含有 RLT 缓冲液(1053393,Qiagen)的单独管中,并立即冷冻在干冰中。
Smart-seq2 library preparation was carried out in 96-well format as previously described.94 Library quality was assessed using the High Sensitivity DNA Analysis Kit (5067-4626, Agilent) on the 2100 Bioanalyzer system (Agilent). Pooled libraries were sequenced on an Illumina NovaSeq platform with a read length of PE 150 bp. Reads were processed as described in Bergmann et al.34 Specifically; reads were trimmed of adapter sequences using TrimGalore! (https://github.com/FelixKrueger/TrimGalore) and mapped to the Common marmoset genome (Callithrix jacchus 3.2.1) using STAR95 aligner v2.5.4. Only samples with >100,000 mapped reads and mapping efficiency >40% were used for downstream analysis. Gene counts were quantified using FeatureCounts96 v1.6.0 using a modified Ensembl gene annotation file (release 91).34 Marmoset samples passing QC were analysed using Seurat95,96 v3.1.2. Feature counts were normalised and standardised using the NormalizeData and ScaleData function.
如前所述,Smart-seq2 文库制备以 96 孔格式进行。 94在 2100 生物分析仪系统 (Agilent) 上使用高灵敏度 DNA 分析试剂盒(5067-4626,Agilent)评估文库质量。合并的文库在 Illumina NovaSeq 平台上进行测序,读长为 PE 150 bp。按照 Bergmann 等人的描述处理读取。 34具体而言;使用 TrimGalore 修剪读数的接头序列! ( https://github.com/FelixKrueger/TrimGalore )并使用 STAR 95对准器 v2.5.4 映射到普通狨猴基因组 (Callithrix jacchus 3.2.1)。仅使用映射读数 >100,000 且映射效率 >40% 的样本进行下游分析。使用FeatureCounts 96 v1.6.0使用修改后的Ensembl基因注释文件(版本91)对基因计数进行量化。使用 Seurat 95 96 v3.1.2 分析了34 个通过 QC 的狨猴样本。使用 NormalizeData 和 ScaleData 函数对特征计数进行归一化和标准化。

DNA methylation DNA甲基化

Bulk whole genome bisulfite sequencing was conducted on three replicates of marmoset periTSCs, primed marmoset PSCs and naive marmoset PSCs. Genomic DNA was extracted using a Monarch Genomic DNA Purification Kit (NEB). Bulk bisulfite sequencing was conducted on an Illumina NovaSeq PE150 Sequencing with ∼30x coverage by CD Genomics. For bulk methylation data, reads were first trimmed of adapter sequences using Trim Galore! (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) and aligned to marmoset genome C. jacchus 3.2.1 using Bismark v0.163 (https://www.bioinformatics.babraham.ac.uk/projects/bismark/). The percentage DNA methylation in CpG context was calculated and saved as individual bigwig files. Percentage methylation was summarised over various genomic feature contexts, including promoter regions, the entire gene body, over exons and introns and at CpG islands and visualised as a line plot or heatmap using Deep Tools v 3.1.3 (https://deeptools.readthedocs.io/en/develop/). Note that CpG Islands were identified using the cpg_lh function of KentUtils (https://github.com/ENCODE-DCC/kentUtils). Differential methylation regions were calculated on incremental 1kb windows using DMRcaller v 1.4.2.87
对狨猴 periTSC、引发的狨猴 PSC 和幼稚狨猴 PSC 的三个重复进行批量全基因组亚硫酸氢盐测序。使用 Monarch 基因组 DNA 纯化试剂盒 (NEB) 提取基因组 DNA。 CD Genomics 在 Illumina NovaSeq PE150 测序仪上进行批量亚硫酸氢盐测序,覆盖率约为 30 倍。对于批量甲基化数据,首先使用 Trim Galore 修剪读数的接头序列! ( https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/ )并使用 Bismark v0.163 与狨猴基因组C. jacchus 3.2.1 进行比对( https://www.bioinformatics.babraham.ac.uk/projects/bismark/ )。计算 CpG 背景下的 DNA 甲基化百分比并保存为单独的大佬文件。总结了各种基因组特征背景的甲基化百分比,包括启动子区域、整个基因体、外显子和内含子以及 CpG 岛,并使用 Deep Tools v 3.1.3 将甲基化百分比可视化为线图或热图 ( https://deeptools.readthedocs .io/en/develop/ )。请注意,CpG 岛是使用 KentUtils 的 cpg_lh 函数识别的( https://github.com/ENCODE-DCC/kentUtils) )。使用 DMRcaller v 1.4.2 在增量 1kb 窗口上计算差异甲基化区域。 87

Single cell bisulfite-sequencing
单细胞亚硫酸氢盐测序

Laser capture microdissection was used to isolate tissues at near single-cell resolution in Carnegie Stage 5 (CS5) and CS7 marmoset embryos. RNA was extracted and used in a separate study34 while DNA was extracted for bisulfite sequencing (BS-seq) for use in this study, as outlined in Macaulay et al.97
使用激光捕获显微切割以接近单细胞的分辨率分离卡内基 5 期 (CS5) 和 CS7 狨猴胚胎中的组织。提取 RNA 并用于一项单独的研究34,同时提取 DNA 用于亚硫酸氢盐测序 (BS-seq),用于本研究,如Macaulay 等人所述。 97
Samples were analysed as outlined in Clark et al.98 Reads were first trimmed of adapter sequences using TrimGalore and subsequently aligned to the marmoset genome (Callithrix jacchus 3.2.1) with Bismark (v 0.22.1), using single-end and nondirectional mode. Duplicate reads were removed using bismark_deduplicate, and methylation extraction run on each sample to capture the percent methylation in CpG context. As the per cell coverage for single cell bisulfite sequencing is typically very low, individual bam files were combined to create pseudobulk representations for each tissue. Cell annotations for the individual samples were first assigned based on the transcriptomic annotation34 or by location within the embryo. Methylation extraction in CpG context was run on the pseudobulk tissue bam files. For visualisation, bedGraphs were converted to bigwig format and visualised in IGV (2.16.2).
按照 Clark 等人的概述对样品进行分析。首先使用 TrimGalore 修剪98 个读数的接头序列,然后使用 Bismark (v 0.22.1) 使用单端和非定向模式与狨猴基因组 (Callithrix jacchus 3.2.1) 进行比对。使用 bismark_deduplicate 删除重复读数,并对每个样本运行甲基化提取以捕获 CpG 背景中的甲基化百分比。由于单细胞亚硫酸氢盐测序的每个细胞覆盖率通常非常低,因此将各个 bam 文件组合起来为每个组织创建伪批量表示。首先根据转录组注释34或根据胚胎内的位置来分配各个样品的细胞注释。 CpG 上下文中的甲基化提取在伪大量组织 bam 文件上运行。为了可视化,bedGraphs 被转换为 bigwig 格式并在 IGV (2.16.2) 中可视化。

Quantification and statistical analysis
量化和统计分析

In vivo and in vitro sample sizing
体内和体外样品大小

In vivo marmoset single-cell RNA sequencing was performed in 2 biological replicates at each developmental stage (CS5, CS6, CS7). In vitro OK, postTSCs (OK XAV PD) and periTSCs (PAVS) were generated using New2 and New4 cell lines in two independent rounds. The “N” represented in Figures 3I, 5J, S2G, S3A, S3B, S5A, S5J, S6E, S6I, S6K, S6L, and S7B represents independent replicates of each experiment. Cross-species chimeras were performed with periTSCs (PAVS) derived from New4 cell line aggregated with 44 embryos collected from 7 different mice.
在每个发育阶段(CS5、CS6、CS7)进行 2 次生物重复体内狨猴单细胞 RNA 测序。在体外,使用 New2 和 New4 细胞系分两轮独立生成 OK、postTSC (OK XAV PD) 和 periTSC (PAVS)。3I、 5JS2GS3A 、S3B、 S5A 、S5J、 S6E 、S6I、S6K、S6L 和S7B中代表的“N”代表每个实验的独立重复。使用源自 New4 细胞系的 periTSC (PAVS) 进行跨物种嵌合,该细胞系与从 7 只不同小鼠收集的 44 个胚胎聚集在一起。

Dimensionality reduction, correlation and diffusion map analysis
降维、相关性和扩散图分析

The input dataset for dimensionality reduction consists of the 10000 most variably expressed genes, as determined by the Seurat function FindVariableFeatures.88 In vitro cultures were jointly analysed using Seurat based on the canonical correlation analysis and mutual nearest neighbour approaches. For visualisation, data from all in vitro models were jointly integrated within the marmoset in vivo reference datasets34 using Seurat, based on Canonical Correlation Analysis (CCA) and mutual nearest neighbour (MNN) approaches. Specifically, FindIntegrationAnchors was run using 10000 features and IntegrateData (with 20 dimensions) was used to calculate corrected gene expression matrices for the three datasets. Datasets were visualised using PCA on the corrected gene expression matrix, with in vivo and in vitro datasets split to aid interpretation.
用于降维的输入数据集由 10000 个表达差异最大的基因组成,由 Seurat 函数 FindVariableFeatures 确定。 88使用 Seurat 基于典型相关分析和相互最近邻方法对体外培养物进行联合分析。为了可视化,基于典型相关分析 (CCA) 和相互最近邻 (MNN) 方法,使用 Seurat 将所有体外模型的数据联合整合到狨猴体内参考数据集34中。具体来说,FindIntegrationAnchors 使用 10000 个特征运行,IntegrateData(具有 20 个维度)用于计算三个数据集的校正基因表达矩阵。使用 PCA 在校正的基因表达矩阵上可视化数据集,并分割体内体外数据集以帮助解释。
Conserved markers for lineages were identified using the FindMarkers function on the 10,000 integration genes with cutoffs based on an adjusted p value of < 0.05 and an average log foldchange >0.1 or <-0.1. Marker expression of key markers was visualized using a scatter plot on the first two principal components. Functional analysis was conducted using the Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Reactome databases via the R package.99 The Seurat function AddModuleScore evaluated the expression levels of genes within a particular signalling pathway, metabolic pathway or a cluster of genes.100 PCA and UMAP analysis were performed using built-in Seurat functions. Pearson correlation between subgroups of cells were calculated based on corrected gene expression values following integration with Seurat using the inbuilt R function, cor. Cross correlation matrices were visualised as heatmaps using pheatmap 1.0.12. Differential expression analysis was done using the cut-off of FSR < 0.01.
使用 FindMarkers 函数对 10,000 个整合基因鉴定谱系的保守标记,并根据调整后的 p 值 < 0.05 和平均对数倍数变化 >0.1 或 <-0.1 进行截断。使用前两个主成分的散点图可视化关键标记的标记表达。通过 R 包使用京都基因和基因组百科全书 (KEGG) 和 Reactome 数据库进行功能分析。 99 Seurat 函数 AddModuleScore 评估特定信号通路、代谢通路或基因簇内基因的表达水平。使用内置 Seurat 函数执行100 次PCA 和 UMAP 分析。使用内置 R 函数与 Seurat 整合后,根据校正后的基因表达值计算细胞亚组之间的 Pearson 相关性。使用 pheatmap 1.0.12 将互相关矩阵可视化为热图。使用 FSR < 0.01 的截止值进行差异表达分析。

Cross-species integrative analyses
跨物种综合分析

For cross-species analysis several embryonic and embryonic-model systems were integrated together. Each dataset for integration contained both an embryonic and extraembryonic component and included:
为了进行跨物种分析,将几个胚胎和胚胎模型系统集成在一起。用于整合的每个数据集都包含胚胎和胚胎外成分,包括:
  • 1.
    Cynomolgus samples from in vitro cultured embryos from day 9 to 20 using Smart-Seq2.69
    使用 Smart-Seq2 从体外培养胚胎第 9 天至第 20 天获取食蟹猴样本。 69
  • 2.
    Cynomolgus samples of in vitro cultured embryos from day 10 -14 using 10X.74
    使用 10X 从第 10 天 -14 天获取体外培养胚胎的食蟹猴样本。 74
  • 3.
    Marmoset samples of in vivo embryos from CS5-7 using SS2.34
    使用 SS2 从 CS5-7 采集狨猴体内胚胎样本。 34
  • 4.
    Human in vitro cultured embryos using SS2 from days 8 – 14.73
    使用 SS2 进行人类体外培养胚胎第 8 – 14 天。 73
  • 5.
    Human 10X samples of 1st trimester trophoblast.9
    人类一个三个月滋养层的 10X 样本。 9
  • 6.
    Two blastoid models80,81 using SS2 and 10X respectively and one 10X amnioid model.82
    两个胚层模型80 81分别使用 SS2 和 10X,以及一个 10X 羊膜模型。 82
Datasets were integrated in Seurat v3.2.088,89 using IntegrateData on 5000 marker genes (FindIntegrationMarkers) and 20 principal components (PCs). Once integrated, dimensionality reduction was run based on corrected gene expression matrices. For UMAP, dimensionality reduction was based on the first 20 PCs. Clustering was generated using the FindClusters function. For all annotated datasets, preliminary alignments separated out embryonic from extraembryonic cell lineages into separate clusters that separated well using PCA and UMAP. For trajectory inference, clusters associated with the TE (Cl 1 and 9), trophoblast (Clusters 0,2–7,10–11,13–16,18–20) and amnion (Clusters 6,8,9,12,17) were selected for datasets 1–5 (by subsetting clusters) and a refined clustering was generated. Any cell lineages with <10 cells were filtered out. Lineages for datasets with available annotations remained unchanged and lineages in unannotated or partially unannotated datasets were assigned lineages based on the dominant cell type in that cluster in other (annotated) datasets. Finally, diffusion maps were generated using destiny v 2.12.0.101 Three diffusion maps were generated to visualise dynamics:
使用 IntegrateData 在 Seurat v3.2.0 88 89中对 5000 个标记基因 (FindIntegrationMarkers) 和 20 个主成分 (PC) 进行数据集整合。整合后,基于校正的基因表达矩阵进行降维。对于 UMAP,降维是基于前 20 个 PC。聚类是使用 FindClusters 函数生成的。对于所有带注释的数据集,初步比对将胚胎细胞谱系与胚胎外细胞谱系分离成单独的簇,并使用 PCA 和 UMAP 很好地分离。对于轨迹推断,与 TE(Cl 1 和 9)、滋养层(簇 0,2–7,10–11,13–16,18–20)和羊膜(簇 6,8,9,12,17)相关的簇)被选择用于数据集 1-5(通过子集聚类)并生成精细的聚类。任何具有 <10 id=8> 的细胞谱系101生成三个扩散图以可视化动态:
  • 1.
    Divergence between syncytiotrophoblast and EVT from CTB dominated the first few diffusion components with amnion and TE sitting close to CTB.
    合体滋养层和 EVT 与 CTB 之间的分歧主导了前几个扩散成分,羊膜和 TE 靠近 CTB。
  • 2.
    A subsequent Diffusion map was generated for the TE, CTB and amnion was generated to better visualise the divergence between these lineages.
    随后生成了 TE、CTB 和羊膜的扩散图,以更好地可视化这些谱系之间的差异。
  • 3.
    A final DM focussed on the divergence between CTB and the syncytiotrophoblast and EVT lineages.
    最后的 DM 重点关注 CTB 与合体滋养层和 EVT 谱系之间的分歧。

Image analysis 图像分析

IF images were analysed using the open-source software Fiji86 to extract signal intensities, circularity and lumen size. DAPI was used to generate a nuclei segmentation mask. EZRIN and F-actin were used for structure and lumen segmentation. For syncytium quantifications, ZO1 and DAPI were used to define cell boundaries and nuclei count, respectively. Individual cells were segmented and quantified individually. In Figure 5K, CDX2+ cells were determined by 2 times the standard deviation of CDX2 expression in PSCs.
使用开源软件 Fiji 86分析 IF 图像,以提取信号强度、圆度和流明尺寸。 DAPI 用于生成细胞核分割掩模。 EZRIN 和 F-肌动蛋白用于结构和管腔分割。对于合胞体定量,ZO1 和 DAPI 分别用于定义细胞边界和细胞核计数。单个细胞被单独分割和量化。在图5K中,CDX2+细胞通过PSC中CDX2表达的标准差的2倍来确定。
Immunofluorescence analysis was performed in R. Two-tailed Mann–Whitney test was used to compare between two means for samples that were not normally distributed and Kruskal–Wallis followed by Dunn's multiple comparison test was used to compare between more than two means. All quantifications were performed on three independent biological replicates, except for Figures 5K and S2G with 2 biological replicates. Significance was determined by a p value < 0.05 as indicated by in the figures and figure captions.
在R中进行免疫荧光分析。双尾Mann-Whitney检验用于比较非正态分布样品的两种均值,Kruskal-Wallis和随后的Dunn多重比较检验用于比较两种以上均值。除图 5 K 和S2 G 具有 2 个生物重复外,所有定量均在三个独立的生物重复上进行。显着性由 p 值 < 0.05 确定,如图和图标题中的*所示。

Supplemental information 补充信息

What’s this? 这是什么?
Download: Download Acrobat PDF file (16MB)

Document S1. Figures S1–S7.

Download: Download spreadsheet (12KB)

Table S1. Primer sequences, related to Figures 7J and S2B.

Download: Download spreadsheet (363KB)

Table S2.. Image quantification data, related to Figures 3, 4, S1, S2, and S4–S7

Download: Download Acrobat PDF file (25MB)

Document S2. Article plus supplemental information.

References