Review
Recapitulating human embryo implantation in vitro using stem cells and organoids
2024 年 6 巻 3 号 p. 68-71
詳細
2024 年 6 巻 3 号 p. 68-71
Implantation is the process in which an embryo invades the endometrium. Therefore, reliable models are required to understand the complex molecular mechanisms underlying human embryo implantation. We recently developed an advanced embryo implantation assembloid by co-culturing organoids that simulate the structure of the human endometrium with blastocyst-like entities derived from naïve pluripotent stem cells. This paper reviews and juxtaposes various studies on in vitro human embryo mimicry, centering on the implantation model we developed, and highlights the methodologies and findings of our investigation, along with other critical advances in the field.
• Using a collagen-based substrate, we engineered endometrial organoids featuring exposed apical surfaces and optimized them for embryo implantation studies.
• We developed an implantation assembloid by co-culturing a human blastoid with our endometrial model, accurately mimicking the human embryo implantation process.
• Analyses of the implantation assembloids confirmed the fusion of syncytial cells with maternal cells, advancing our understanding of the cellular dynamics during early pregnancy.
In recent years, a decline in birth rates has been observed in developed countries driven by societal trends such as delayed marriages and increased female participation in the workforce. Consequently, ARTs, such as in vitro fertilization and intracytoplasmic sperm injection have become essential measures for counteracting declining birth rates. However, the implantation rate of embryos via ART remains low and a significant number of cases involve refractory implantation failure, in which repeated transfers of high-quality embryos fail to result in pregnancy, leading to treatment abandonment. Therefore, addressing this issue, which is a major national challenge, is crucial. Understanding the interactions between fetal and maternal environments during the implantation process is key to improving implantation outcomes.
Implantation is a crucial process in pregnancy, involving initial direct interactions between the embryo and maternal body. The fertilized egg, resulting from the union of an egg and sperm, undergoes repeated cleavage and forms a structure known as the blastocyst by approximately Day 5. The blastocyst comprises an inner cell mass (ICM) that develops into the embryonic body and the surrounding trophectoderm, which forms the placenta. After hatching from the zona pellucida, the blastocysts adhere to the endometrium via the exposed trophectoderm. Although studies using animal models, such as mice, have provided significant knowledge about the molecular mechanisms of implantation, there are substantial interspecies differences. For instance, in humans, the ICM adheres to the apical surface of the endometrial epithelium and then deeply invades the endometrium, resulting in complete embedding of the embryo [1]. This direction is opposite to that observed in mice. In addition, the timing of decidualization differs between humans and mice. In humans, the endometrium undergoes cyclic changes in structure and properties owing to hormonal fluctuations during the menstrual cycle, with the basal layer remaining unchanged and the endometrium regenerating under the influence of estrogen. Progesterone produced by the corpus luteum prepares the endometrium for implantation, creating a “window of implantation”. In contrast, in mice, decidualization occurs after embryo implantation. Decidualized human endometrial tissue exhibits edematous thickening, and the luminal epithelium that separates the uterine cavity from the underlying stromal layer facilitates embryo invasion through changes in cellular polarity and a reduction in cell-cell and cell-matrix adhesion by the “window of implantation” [2].
For successful implantation, synchronizing the implantation ability of the embryo and receptive state of the endometrium is crucial. However, the phenomena occurring at the embryo–endometrial interface and the detailed molecular mechanisms of implantation are poorly understood because obtaining human samples is almost ethically and technically impossible. Thus, utilizing in vitro models to simulate implantation is a morally and practically viable option. However, implantation models developed using immortalized cell lines or cancer cells may not accurately represent normal physiological conditions [3, 4]. To address these limitations, it is essential to utilize in vitro models that mimic the implantation niche using cell sources that accurately reflect the physiological environment.
Endometrial organoids (EMO) capable of long-term propagation and demonstrating hormonal responsiveness have been developed by culturing endometrial epithelial cells in Matrigel [5, 6]. However, these organoids exhibit a simple spherical structure in which the luminal (apical) surface, which is critical for embryo attachment, is not exposed, rendering them unsuitable for implantation studies. Recently, we developed a new type of endometrial organoid called apical-out EMO (AO-EMO) [7]. We noticed that the collagen content was high in human endometrial tissues [8] and examined the addition of collagen to the culture substrates. As a result, in three-dimensional (3D) cultures using type I collagen, epithelial cells autonomously migrated to the surface layer of the extracellular matrix. In addition, immunostaining (laminin) confirmed that the cilium localized on the luminal surface was located outside, and the basement membrane was located inside. The AO-EMO demonstrated robust hormone responsiveness by sequentially undergoing treatment with hormones, including estradiol, medroxyprogesterone acetate, prolactin, and 8-Br-cAMP (termed EPCP treatment). Consequently, AO-EMOs exhibited significant glandular thickening characteristic of the secretory phase (Fig. 1A). This hormone-induced maturation in AO-EMO surpassed that observed in EMOs cultured in conventional Matrigel [7]. Furthermore, incorporation of human umbilical vein endothelial cells enabled the formation of an internal vascular network (Fig. 1B). By adding stromal cells, we created a composite human endometrial model that mirrored the spatial arrangement and cellular composition of endometrial tissues in vivo (Fig. 1C, 1D). Consequently, we successfully developed an endometrial model using a collagen-based 3D culture system that fosters self-formation of an apical-out configuration in the epithelium [7]. One notable method involves the creation of 3D endometrial constructs by embedding endometrial stromal cells in collagen followed by their overlay with epithelial cells [9]. Compared to approaches using cell culture inserts, our composite endometrial model utilizing AO-EMO represents an advanced model. This enables vascularization and is characterized by a dense environment of stromal cells autonomously formed through condensation.
Apical-Out Endometrial Model (modified from [7]). A. Thickened apical-out endometrial organoids (AO-EMO) in response to hormones, stained for EpCAM (green), F-actin (red), and PGR (white); nuclei stained with Hoechst (blue). B. Fluorescence image of AO-EMO, including RFP-HUVECs. Network of endothelial cells were shown in red. C. Schematic overview of the co-culture of the endometrial organoid model. EMO, endometrial organoids; eSC, endometrial stromal cells; HUVECs, human umbilical vein endothelial cells. D. Human endometrial tissue in vivo (left) and the endometrial organoid model (right) stained for EpCAM (cyan), CD31 (red) or RFP, and VIM (yellow); nuclei are stained with Hoechst (blue).
Methods for culturing human embryos in vitro have been established, demonstrating that embryonic development can proceed for up to 14 days post-fertilization in both two-dimensional and 3D cultures without maternal factors [10,11,12]. However, the use of human embryos poses ethical concerns and is limited to a small number of laboratories. Recently, blastoids, blastocyst-like structures derived from pluripotent stem cells, have emerged as promising alternatives to human embryos [13,14,15]. In a recent study, we explored the interactions between these blastoids and a composite endometrial model, as described in this paper. To facilitate the tracing of endometrial epithelial cells, we used a lentivirus to stably express green fluorescent protein in the epithelial cells of the model. Furthermore, we generated blastoids from naïve human ES cells treated with Kusabira Orange and co-cultured these labeled blastoids with a composite EMO, which incorporated both endometrial stromal cells and human umbilical vein endothelial cells in suspension culture (Fig. 2A). Consequently, the blastoids adhered to the EMO via an ICM-like structure, flattening and penetrating the EMO (Fig. 2B). At the contact surface, disruption of the endometrial epithelium and infiltration of syncytiotrophoblast cells were observed beneath the ICM-like structures. These syncytiotrophoblasts, characterized by large nuclei, were in direct contact with the underlying stromal cells, resembling the primitive syncytium seen at the embryo–endometrial interface post-implantation [7]. Although the 3D culture of blastoids to promote development [16] and blastoids adhered to two-dimensionally cultured endometrial epithelium [13] have been reported, our model offers a powerful tool that enables the visualization of the implantation process in three dimensions and analysis of primitive syncytiotrophoblast formation.
Creation of an Implantation Assembloid (modified from [7]). A. Schematic of the co-culture procedure. B. Bright-field and fluorescence image of human blastoids (KuO+) and composite endometrial organoids (epi: eGFP+) on Days 0, 1, 2, and 3 of co-culture. Arrows and arrowheads indicate blastoids. GFP, green fluorescent protein; KuO, Kusabira Orange
In our recent study, we confirmed that cell fusion occurs between the blastoid-derived syncytium and endometrial stromal cells. Robust verification of cell fusion was achieved through DNA-FISH analysis, which identified both XY and XX chromosomes within the same cell, and a split-green fluorescent protein system that confirmed fusion events. The reproducibility of these fusion events was confirmed in human blastocysts [7]. These findings provide strong evidence for direct fusion between embryo-derived syncytiotrophoblast cells and maternal-derived endometrial stromal cells, underlining the utility of our assembloid model in simulating early embryonic implantation processes. We hypothesized that the successful incorporation of a dense layer of stromal cells beneath the trophoblast via cell fusion could enhance embryo implantation and subsequent penetration into the maternal uterus.
Human embryo implantation involves profound penetration into the uterus [17], with the level of syncytial and maternal integration likely influencing the penetration depth. Therefore, exploring the dynamics of cell fusion in vivo is essential for future studies.
We successfully developed an embryo-endometrium assembloid model using blastoids and an endometrial model as an embryo substitute. This model enables three-dimensional visualization of the embryo–endometrial interface, allowing us to observe the fusion of embryo-derived cells and maternal endometrial cells during early human embryonic development. Furthermore, by utilizing transfection and genome editing technologies with this model, we anticipate gaining a comprehensive understanding of the molecular mechanisms involved in implantation by examining the interactions between these cells. Additionally, a deeper understanding of the mechanisms underlying embryo implantation could improve the success rates of ART and potentially offer new solutions for refractory implantation failure.
The authors declare no conflicts of interest regarding this study.
This study was conducted in collaboration with research groups from Tohoku University Graduate School of Medicine, Trophoblast Research at the Institute of Molecular Embryology and Genetics at Kumamoto University, the Institute for Quantitative Biosciences at the University of Tokyo, and the Institute of Biomaterials and Bioengineering at Tokyo Medical and Dental University. This work was supported by the Japan Agency for Medical Research and Development [grant numbers: JP19gm1310001, JP21bm0704068, JP23gn0110072] and the Japan Society for the Promotion of Science KAKENHI [grant numbers: 21H04834, 21K15098, 24H01389].
