Human amniotic epithelial cells (hAECs) are derived from the amniotic ectoderm which differentiates from the epiblast around day 8 after fertilization. The cell morphology and transcriptome of amniotic ectoderm are distinctly different from epiblast. As the differentiation and expansion proceed, amniotic ectoderm gradually gets apart from epiblast, forming amniotic cavity filled up with amniotic fluids, and this structure is called amniotic sac. Amniogenesis occurs prior to the formation of primitive streak, the hallmark of the initiation of gastrulation, making amniotic ectoderm cells one of the primordial extraembryonic cells. hAECs are tightly packed as a monolayer, constituting the innermost layer of amnion, directly in contact with amniotic fluid (1, 2). It has been reported that pluripotency markers such as TRA1-60, TRA1-81, SSEA3, and SSEA4 are expressed in hAECs isolated from early gestational amnion, then are gradually lost over the pregnancy period (3). Although many clinical studies have shown that hAECs are of promising therapeutic potential in various diseases, the specific cell biological features and mechanisms for the treatments still remain elusive. In addition, it has been shown that hAECs have limited proliferation capability due to the scarce telomerase activity, impeding their development as a cellular product in the large-scale clinical applications and their further industrialization. To address these problems, it is critical to have a comprehensive understanding of hAECs, including cellular properties, the intrinsic proliferative capacity, and the therapeutic potentials.

Amniotic epithelial cells are derived from epiblast around day 8 post fertilization. During the early embryogenesis, a zygote is a totipotent cell having the potentials to develop into both embryonic and extra-embryonic tissues. Once it forms, it rapidly undergoes successive divisions including the 2-cell, 4-cell, 8-cell, 16-cell (morula), and blastocyst stages. At the morula stage, the cells exhibit differential division rates that faster-dividing cells forming a non-polarized inner cell mass (ICM) in the interior of embryo while the slow-dividing cells tightly aligning in the exterior to encompass the ICM and eventually differentiating into trophectoderm (TE), forming the blastocyst. As the blastocoel expands, the ICM predominantly aggregates on one side of the embryo, establishing an overall polarized embryo (4, 5). The ICM further differentiates into epiblast and hypoblast (primitive endoderm), forming the bilaminar disk. Amniotic ectoderm, is believed as the primordial amniotic epithelial cells, emerges during peri-implantation from the differentiation of epiblast. Subsequently, the primitive streak will arise from the derivation of the non-amniotic origin epiblast, which represents the initiation of the three embryonic germ-layers and organogenesis ( Figure 1A ).

The amnion, a crucial extraembryonic tissue, is an important milestone in animals’ evolutionary transitions from aquatic to terrestrial environment. Amnion provides mechanical protection to the fetus and secretes cytokines and hormones, contributing to the embryo development. Although the underlying mechanisms of amniogenesis and amniotic epithelial lineage development are not fully understood yet, the specifications of amniogenesis have been reported with two different patterns: folding and cavitation (6). In species such as bats, monkeys, and higher primates, the amnion emerges by delamination from pluripotent epiblast around the peri-implantation, followed by epithelialization and cavitation, forming the amniotic sac (7, 8). In contrast, in mice, rabbits, Pteropodid bats, dogs, pigs, cows, and lower primates, the amnion is formed by the folding of embryonic tissues during or shortly after gastrulation, whereby the amniotic folds extend and merge to form a closed sac (9–11). It is noteworthy that despite these different amniogenesis patterns, the anatomical structures and functions of amnion are similar across various species. Because of the ethical concerns, the investigations of human amniotic epithelial cells lineage development were restricted in the stage of formation of amniotic ectoderm. The entire process remains unclear; therefore, the substitute models using pluripotent stem cells in vitro have been established to learn its lineage development and to comprehensively understand the amniotic epithelial cells biological features and therapeutic potentials as a promising cellular drug candidate.

Previous studies have employed single-cell transcriptome sequencing (scRNA-seq) to analyze three stages of primate embryonic development, including the cultured human pre-gastrulation embryos, in vitro cultured Cynomolgus monkey gastrulating embryos, and a human gastrulating embryo implanted in utero. The integrated scRNA-seq analysis revealed that amniogenesis occurred in two distinct waves of epiblast differentiation during early embryonic development. The two waves occur independently and develop in different routes, amniotic epithelial cells-early (AME-E) and amniotic epithelial cells-late (AME-L), respectively. AME-E follows a trophectoderm-like route and the amniotic cavity is formed during the early wave, whereas AME-L follows a nonneural ectoderm-like transcriptional program (9). Similar results were also observed in the differentiation of human pluripotent stem cells (hPSCs) from different states. The naïve and primed hPSCs could model the two waves of amniogenesis (9).

Due to the “14-day rule”, referring to the International Society for Stem Cell Research (ISSCR) recommendation that human embryos created by in vitro fertilization, either frozen or unfrozen, cannot survive outside the body beyond the 14th day after fertilization without embryo transfer, and technological constraints, the studies of embryonic development remain quite limited, as well as a similarly scant comprehension to amniotic development. The amnion is composed of epithelium, basement membrane, the compact stromal layer, fibroblast layer, and the intermediate spongy layer ( Figure 1B ). The amniotic epithelial layer localizes at the innermost of amnion and directly contacts with the amniotic fluid. The protein secreted by amniotic epithelial cells, such as glycoproteins and collagens, constitute the underlying basement membrane (12). Amniotic epithelial cells and amniotic mesenchymal cells primarily reside in the epithelium and the fibroblast layer, respectively (13). These two cell types have distinct origins. Amniotic epithelial cells are derived from amniotic ectoderm (14), in contrast, amniotic mesenchymal cells originate from extra-embryonic mesoderm, which is developed posterior to amniotic epithelial cells (15). As the two major cell types derived from the amnion membrane, both cells have the advantages of low immunogenicity, no tumorigenicity, and limited ethical considerations. The characterized markers of hAECs and human amniotic mesenchymal stem cells (hAMSCs) are listed in Table 1 . It has been demonstrated that the epithelial markers, such as cytokeratins, E-cadherins, and CD9, are highly expressed in hAECs but not in hAMSCs. Multiple mesenchymal markers like CD90, CD29, and CD105 are also expressed in hAECs, indicating the spontaneous occurrence of epithelial-mesenchymal transition (EMT) during the cell cultivation in vitro. Both of them low express HLA-A, -B, -C, and are negative in the detection of hematopoietic stem cell markers, representing their low immunogenic potential. Although some pluripotency markers (SSEA4, OCT4, TRA1-60, and REX1) have been detected in hAECs and hAMSCs, the expression levels are extremely low compared with that of induced pluripotent stem cells (iPSCs).

Pluripotent stem cells, including embryonic stem cells (ESCs) and iPSCs, have been often employed in the studies of early embryonic development because of their differentiation potentials to the three-germ layers. It has been demonstrated that PSCs possess of the formative naive-to- primed transition in vitro. As widely recognized, naive PSCs are akin to the pre-implantation epiblast, having the differentiation potentials for both embryonic and some extra-embryonic tissues, while primed PSCs are more like the post-implantation epiblast and differentially develop into the embryonic tissues (16). Fu and his team established the first 3D model of amniotic sac embryoid using microfluidic devices and human PSCs, known as microfluidic platform for analysis of single embryos (μPASE), to explore the early development of amniotic sac, and found that the activation of BMP-SMAD signaling enabled the self-organization of hPSCs into amniotic sacs (17).

In another study, Wu et al. likewise induced the organization of hPSCs into structures resembling to the early embryos, termed Blastoids, which expressed the amniotic genes, exhibiting the morphological similarities to human blastocyst. Unlike μPASE, which collapsed within 48h, Blastoids were able to be maintained up to 4 days in vitro (18). Same results were also obtained using iPSCs, named as iBlastoids (19). On the basis of the previous research, Qin and his associates upgraded the microfluidics amniotic sac platform and exerted a perfuse-able microfluidic device to fabricate an advanced 3D amnion microchip. The constant-rate perfusion of culture medium prolonged the presence and integrity of the amniotic sac-like structure to 20 days. It offers a great strategy and model to investigate human amniotic development in mid-gestation (20).

Although the therapeutic effects of hAECs have been evidenced in many clinical studies, their applications are restricted by the current in vitro culture expansion systems. The yield and biological characteristics of primary hAECs are affected by various factors like the physical condition of parturients, placental size and quality, origins from different amnion regions, gestation stages, isolation and preservation procedures (3, 21, 22). The average quantity of hAECs isolated from an amnion is approximately 1 × 10⁸, whereas a single treatment typically requires the amount in the range of 0.5~1×10⁸, leading to the homogenous cells might not be sufficient for multiple treatments in one patient. Therefore, the requirement to obtain adequate hAECs complying with Good Clinical Laboratory Practice (GCLP) standards has become an absolute challenge. Developing a system that is sufficient to expand hAECs in vitro requires understanding the properties of hAECs, as well as their proliferation mechanisms. However, regardless of the extensive clinical studies of hAECs in various diseases, the lack of typical markers and standard production process for the quality control of hAECs still are the most urgent issues to be addressed.

Stem cell-based therapy has recently emerged as a key player in regenerative medicine. The commonly used cell types include pluripotent stem cells, extraembryonic tissue-derived cells and adult stem cells. Among those stem cell-based clinical applications, hematopoietic stem cells (HSCs) transplantation takes the leading position, and the second place goes to mesenchymal stem cell (MSC)-based studies. Autologous or allogeneic HSC transplantation is a revolutionary life-saving procedure after irradiation or chemotherapy. The first bone marrow transplantation was reported in 1957 and it currently has become the most effective treatment for leukemia. However, HSCs extraction requires the invasive procedures to human bodies. More severely, post-operative complications are often occurred to patients. As per the statistic study of a retrospective analysis over 11 years after HSCs transplantation, of all patients, 74.2% suffered of early or late compilations, including infections, graft-versus-host disease (GVHD), CNS disorders and cardio-vascular complications, etc (59).

Human pluripotent stem cells (hPSCs) are defined as self-renewable cell types conferring the ability to differentiate into various cellular phenotypes of the human body, including three germ layers, and have gained significant interest and attention in regenerative medicine field. Multiple clinical trials using hPSC-derived cells have been launched (clinicaltrials.gov). The therapeutic potential of hPSCs is tremendous, but there are still some challenges that need to be overcome. One of them is the teratoma formation because of their potential for infinite proliferation. Another challenge is the need for standardization. hPSCs, especially iPSCs, are greatly heterogeneous due to their origins and preparation methods. On that account, a set of critical quality procedures and evaluation systems have to be established.

MSCs are multipotent progenitor cells possessing self-renewal ability (limited in vitro) and differentiation potential into mesenchymal lineages, according to the International Society for Cell and Gene Therapy (ISCT). To date, a total of 12 MSCs products have received regulatory approval for commercial use worldwide, including autologous and allogenic, to be used for spinal cord injury, osteoarthritis, GVHD, acute myocardial infarction, and Crohn’s disease. MSCs are commonly derived from umbilical cord blood, bone marrow, and adipose tissue. The cell sources are relatively limited, and their cell isolation requests higher technical skills.

Compare to all the stem cells motioned above, hAECs are morphological epithelial cells, possessing low immunogenicity, immunomodulatory and anti-inflammatory effects. Moreover, hAECs lack the telomerase activity and the potential for sustainable cell divisions, making them have a low risk of tumorigenicity. Consequently, hAECs have been considered a promising candidate for cell therapy ( Figure 3 ). Despite of some uncertainty about the molecular mechanisms by which hAECs act effective on diseases, whether they are in vivo engrafted and differentiated or they modulate the biochemical reactions and cellular responses to injuries, the work of Wallace team has clear demonstrated that hAECs can exert a reparative effect without the need for engraftment or differentiation (60). They suggested that the primary mechanism of hAECs for lung injury repair was likely to be paracrine signaling to the surrounding tissues to reduce proinflammatory and profibrotic mediators. Similar results were also found in the studies in brain ischemia, Parkinson’s disease, spinal cord injury and wound healing (61–64). Furthermore, the transcriptome sequencing results have revealed that hAECs transplantation lead to the upregulation of several angiogenesis and inflammation molecules, such as interferon regulatory factor 7 (IRF7), Mx dynamin-like GTPase 1 (Mx1), vascular endothelial growth factor receptor 1 (VEGFR1) and VEGFR2 (65). Coculture of hAECs and freshly isolated human blood neutrophils significantly attenuated the level of oxidative burst of neutrophils (66), but directly inhibited the proliferation of naïve CD4 T cells and the production of Th1 and Th17 cytokines (67). All these cellular properties have led hAECs ideal candidate cells for various disease cell therapies.

In 1981. Akle, C.A., et al. firstly isolated primary hAECs and transplanted subcutaneously into the upper arms of 7 volunteers. Four weeks later, it was observed that none of the participants exhibited detectable HLA antibodies in their blood samples, and no lymphocyte reaction was shown in 2 of the participants, indicating the transplantation of hAECs would not cause acute immune rejection (34). Since then, cell therapy using hAECs has become an emerging treatment for diseases (68–70). Since then many studies have shown satisfactory effects of hAECs in the pre-clinical treatments, however, very little outcomes were revealed on the clinical research (71, 72). As of January 2024, there are 25 clinical trials listed in the database (ClinicalTrials.gov, Anzctr.org.au, etc.) using “human amniotic epithelial cells” as biological interventions for the treatments of various diseases ( Table 3 ), most promising include neurological disorders, immune diseases and tissue repairs.

Despite recent advances in stem cell-based therapies for various disease, including ESCs, iPSCs, and MSCs, there are many unclear questions and unsolved issues regarding to their potency, stability, oncogenicity, immune response, cell sources, and ethics. The major concerns for their therapeutic potentials are the high risk of tumorigenicity and invasive extraction procedures. hAECs are derived from the human amnion, which as a medical waste are readily available and under less ethical dispute, and have no tumorigenic and low immunogenic potential. The applications of hAECs on various diseases have been studied and relevant underlying mechanisms have also been explored, making them a better alternative cell source for diseases.

Although hAECs have exhibited a good efficacy in various clinical studies, their limited intrinsic capacity of proliferation lets the expansion in vitro and industrial manufacture challenging. Another common question about hAECs in clinical applications is the shortness of comprehensive knowledge of cell properties. Therefore, cell quality assessment for therapeutic effectiveness could not be controlled, which might lead to inconsistent outcomes in treatments.

These issues can be addressed by establishing optimized in vitro expansion strategies. Simultaneously, it is crucial to establish a complete and standardized evaluation system for cell properties and qualities. In the studies of hAECs lineage development and therapeutical mechanisms, by adopting new analytical methods such as single-cell transcriptome sequencing, machine learning, etc. Numerous research has looked into the deep insights of hAECs composition, as well as, cell fates and attributes during development, fundamentally solving the challenges in developing cell therapy products. On the other hand, by exploring new methods, such as combined therapy, and gene editing, the therapeutic potential of hAECs has been enhanced. In future research, the exploration of the functional features of hAECs will be the prime work to better target the disease treatment.