The modeling of human implantation and early placentation: achievements and perspectives

Abstract BACKGROUND Successful implantation is a critical step for embryo survival. The major losses in natural and assisted human reproduction appeared to occur during the peri-implantation period. Because of ethical constraints, the fascinating maternal–fetal crosstalk during human implantation is difficult to study and thus, the possibility for clinical intervention is still limited. OBJECTIVE AND RATIONALE This review highlights some features of human implantation as a unique, ineffective and difficult-to-model process and summarizes the pros and cons of the most used in vivo, ex vivo and in vitro models. We point out the variety of cell line-derived models and how these data are corroborated by well-defined primary cells of the same nature. Important aspects related to the handling, standardization, validation, and modus operandi of the advanced 3D in vitro models are widely discussed. Special attention is paid to blastocyst-like models recapitulating the hybrid phenotype and HLA profile of extravillous trophoblasts, which are a unique yet poorly understood population with a major role in the successful implantation and immune mother-embryo recognition. Despite raising new ethical dilemmas, extended embryo cultures and synthetic embryo models are also in the scope of our review. SEARCH METHODS We searched the electronic database PubMed from inception until March 2024 by using a multi-stage search strategy of MeSH terms and keywords. In addition, we conducted a forward and backward reference search of authors mentioned in selected articles. OUTCOMES Primates and rodents are valuable in vivo models for human implantation research. However, the deep interstitial, glandular, and endovascular invasion accompanied by a range of human-specific factors responsible for the survival of the fetus determines the uniqueness of the human implantation and limits the cross-species extrapolation of the data. The ex vivo models are short-term cultures, not relevant to the period of implantation, and difficult to standardize. Moreover, the access to tissues from elective terminations of pregnancy raises ethical and legal concerns. Easy-to-culture cancer cell lines have many limitations such as being prone to spontaneous transformation and lacking decent tissue characteristics. The replacement of the original human explants, primary cells or cancer cell lines with cultures of immortalized cell lines with preserved stem cell characteristics appears to be superior for in vitro modeling of human implantation and early placentation. Remarkable advances in our understanding of the peri-implantation stages have also been made by advanced three dimensional (3D) models i.e. spheroids, organoids, and assembloids, as placental and endometrial surrogates. Much work remains to be done for the optimization and standardization of these integrated and complex models. The inclusion of immune components in these models would be an asset to delineate mechanisms of immune tolerance. Stem cell-based embryo-like models and surplus IVF embryos for research bring intriguing possibilities and are thought to be the trend for the next decade for in vitro modeling of human implantation and early embryogenesis. Along with this research, new ethical dilemmas such as the moral status of the human embryo and the potential exploitation of women consenting to donate their spare embryos have emerged. The careful appraisal and development of national legal and ethical frameworks are crucial for better regulation of studies using human embryos and embryoids to reach the potential benefits for human reproduction. WIDER IMPLICATIONS We believe that our data provide a systematization of the available information on the modeling of human implantation and early placentation and will facilitate further research in this field. A strict classification of the advanced 3D models with their pros, cons, applicability, and availability would help improve the research quality to provide reliable outputs.