Organoids as powerful models of endometrium epithelium in transcriptomic, cellular and functional mimicry

The inner lining of the uterus, referred to as the endometrium, is a highly specialized tissue that plays a pivotal role in reproduction. As a critical interface between the maternal environment and the developing embryo, the endometrium supports embryo implantation, development and eventual successful pregnancy. Each month, the endometrium undergoes cyclical remodeling as part of the menstrual cycle, preparing for the potential nesting of an embryo. The endometrium’s dynamic nature arises from its responses to the oscillating ovarian hormones estradiol (E2), a form of estrogen, and progesterone (P4)—which are regulated by the hypothalamic-pituitary axis. Structurally, the human endometrium comprises two main layers: the lamina basalis (basal layer facing the myometrium) and the lamina functionalis (upper layer facing the uterine lumen). While the lamina basalis remains relatively unchanged during the menstrual cycle, the lamina functionalis undergoes significant alterations in architecture and cellular composition. During the proliferative phase, E2 stimulates tissue growth and thickening to restore the lamina functionalis after menstruation. In the subsequent secretory phase, elevated levels of P4 promote cellular differentiation, transforming the endometrium into a receptive state, thereby inducing the secretion of pro-implantation molecules and proteins. If implantation occurs, the endometrium undergoes further transformation to form the decidua, which provides nutrients for the developing embryo and supports early placental development. In the absence of implantation, hormonal decline triggers the breakdown of the lamina functionalis , while the basal layer remains to ensure the continuity of the cycle by regenerating the upper layer. At the cellular level, the endometrium comprises four major cell populations: epithelial, stromal, endothelial and immune cells, which collectively form a complex network integral to its structural integrity and functional roles [ 1 ]. The epithelial cell compartment consists of luminal and glandular cell types, each with distinct functions. The luminal epithelium, characterized by plasma membrane projections of cilia and pinopodes, plays a crucial role in the initial stages of embryo-endometrium interaction, facilitating oocyte movement and embryo implantation [ 2 ]. Of note, the necessary interaction between the blastocyst and the endometrium can only occur in a limited period of 5–10 days per cycle, described as the window of implantation (WOI). During this critical phase, the endometrium undergoes specific morphological and molecular changes, a process known as decidualization, to create a conducive environment for blastocyst adhesion and implantation [ 3 ]. The glandular epithelium, organized in a pseudostratified structure, is specialized to secrete critical factors such as leukemia inhibitory factor (LIF), progestogen-associated endometrial protein (PAEP) and heparin-binding EGF like growth factor (HB-EGF) to provide the necessary nourishment for the embryo. Equally important is the decidualized stroma, which lies under the luminal epithelium and is composed of specialized fibroblasts, tubular glands, spiral arteries and immune cells. P4 plays a pivotal role in triggering polyploidization and the differentiation of the elongated fibroblastic stromal cells into rounded secreting decidual cells, thereby initiating the secretion of stromal factors such as prolactin (PRL) and insulin-like growth factor binding protein 1 (IGFBP1) to establish an immunotolerant and nutrient-rich environment essential for successful embryo implantation and pregnancy maintenance [ 4 , 5 ]. In addition, the endometrium comprises immune cells, i.e. myeloid and lymphoid cells, which in turn give rise to mast cells, macrophages, neutrophils, dendritic cells, T cells, B cells and uterine Natural Killer cells (uNK) [ 6 ]. On top of this, the endometrium also encompasses atrial and venous endothelial cells, contributing to endometrial angiogenesis [ 7 , 8 ]. During menstruation, the endometrial epithelium is significantly impacted as the lamina functionalis is shed into the uterine cavity to be subsequently restored through stages of proliferation and differentiation, all processes prompted by the required transcriptomic alterations [ 9 ]. In this section, the transcriptomic dynamics of the endometrial epithelium during the menstrual cycle will be described by diligently integrating recent insights from publicly available single-cell RNA-sequencing (scRNA-seq) datasets of human endometrium (Table  1 ). Histological definitions of menstrual cycle phases (i.e. menstrual, proliferative, and secretory phase) have served as the golden standard for determining endometrial state since the 1950s [ 10 ]. However, a recent scRNA-seq study redefined four (instead of three) major phases in the endometrium based on (sc) transcriptomic patterns, described as phase 1 to 4 [ 11 ]. Based on in silico predictions, these phases show high correlation with the pre-defined cycling state of the analyzed endometrial biopsies (i.e. menstruation, early and late proliferative phases and early, mid and late secretory phases) (Fig.  1 ). Fig. 1 Transcriptomic landscape of the human endometrial epithelium across the menstrual cycle. Hormonal levels of E2 (blue) and P4 (purple) are shown, as well as the associated histologically and transcriptomically defined endometrial cycle phases. Transcriptomically, the cycle can be divided in four phases (phase 1-4), while the histological classification considers three phases, i.e. menstrual, proliferative and secretory phase. The proliferative and secretory phases exhibit further subdivisions into early and late, and early, mid, and late stages, respectively. Variations in cellular activity and differentiation observed during phase transitions are mainly characterized by cell mitosis, cilium assembly, cell secretion, and abrupt opening and gradual closing of the WOI. Key gene expression profiles are depicted per phase. Figure created with BioRender.com Table 1 Overview of performed scRNA-seq analyses of human endometrium Study Data public accession number Tissue Amount of healthy samples sequenced Cycle phase Main findings related to healthy endometrium Ref Fitzgerald et al. (2019) GSE136795 Org 1 NA First to perform scRNA-seq analysis on endometrial organoids [ 62 ] Wang et al. (2020) GSE111976 Prim 10 Proliferative Secretory Characterization of the transcriptomic transformation of human endometrium at sc resolution across the menstrual cycle [ 11 ] Garcia-Alonso et al. (2021) E-MTAB-10,287 Prim 15 Proliferative Secretory Identification of SOX9 + cell populations Characterization of ciliated and secretory epithelial lineage: WNT-NOTCH interplay [ 1 ] E-MTAB-10,283 Org NA KRT17 + population in organoids Tan et al. (2022) GSE179640 Prim 3 Proliferative Hormone-treated Identification of MUC5B + population [ 16 ] Org 2 NA Lai et al. (2022) GSE183837 Prim 3 Proliferative Secretory Characterization of recurrent implantation failure (RIF)-associated cellular abnormalities [ 113 ] Shih et al. (2022) GSE203191 Prim 9 Menstrual effluent First to perform scRNA-seq analysis on menstrual effluent Identification of IGFBP1 + endometrial stromal cells [ 114 ] Lv et al. (2022) PRJNA730360 Prim 3 Proliferative Characterization of thin endometrium-associated cellular abnormalities [ 115 ] Fang et al. (2022) HRA002555 Prim 3 Proliferative Cell-cell interactions in proliferative endometrium [ 59 ] Huang et al. (2023) GSE214411 Prim 7 Proliferative Secretory Characterization of endometriosis-associated cellular abnormalities [ 116 ] Fonseca et al. (2023) GSE213216 Prim 10 Proliferative Secretory Hormone-treated Characterization of endometriosis-associated cellular abnormalities [ 117 ] Yan et al. (2024) PRJCA009009 Prim 11 Proliferative Secretory Characterization of ovarian endometriosis-associated cellular abnormalities [ 118 ] Mareckova et al. (2024) E-MTAB-14,039 Prim 52 Proliferative Secretory Menstrual Hormone-treated Establishment of inclusive sc endometrial atlas “HECA” [ 25 ] Abbreviations: Prim, primary tissue; Org, organoids. Hormone-treated: endometrial biopsy obtained from women actively taking oral contraceptives at time of sample collection Transcriptomic landscape of the human endometrial epithelium across the menstrual cycle. Hormonal levels of E2 (blue) and P4 (purple) are shown, as well as the associated histologically and transcriptomically defined endometrial cycle phases. Transcriptomically, the cycle can be divided in four phases (phase 1-4), while the histological classification considers three phases, i.e. menstrual, proliferative and secretory phase. The proliferative and secretory phases exhibit further subdivisions into early and late, and early, mid, and late stages, respectively. Variations in cellular activity and differentiation observed during phase transitions are mainly characterized by cell mitosis, cilium assembly, cell secretion, and abrupt opening and gradual closing of the WOI. Key gene expression profiles are depicted per phase. Figure created with BioRender.com Overview of performed scRNA-seq analyses of human endometrium Proliferative Secretory Proliferative Secretory Identification of SOX9 + cell populations Characterization of ciliated and secretory epithelial lineage: WNT-NOTCH interplay Proliferative Hormone-treated Proliferative Secretory First to perform scRNA-seq analysis on menstrual effluent Identification of IGFBP1 + endometrial stromal cells Proliferative Secretory Proliferative Secretory Hormone-treated Proliferative Secretory Proliferative Secretory Menstrual Hormone-treated Abbreviations: Prim, primary tissue; Org, organoids. Hormone-treated: endometrial biopsy obtained from women actively taking oral contraceptives at time of sample collection Phase 1 encompasses menstruation and early proliferative phase. Menstruation is triggered by the decrease in P4, inducing an immune and inflammatory response that contributes to the shedding of endometrial tissue [ 9 , 11 ]. During this phase, matrix metalloproteinases (MMPs), primarily MMP7 , 10 and 11 , play a critical role in the degradation of the extracellular matrix (ECM) resulting in loss of structural integrity and tissue breakdown [ 11 , 12 ] (Fig.  1 ). The resulting menstrual effluent contains tissue fragments comprising immune, stromal and epithelial cells that have recently been shown to be a valuable source for modeling and investigating the human endometrium [ 11 , 13 ]. Concurrently with the shedding process, re-epithelialization of the lamina functionalis is initiated by increasing levels of E2, along with the expression of its target genes estrogen receptor 1 ( ESR1 ) and progesterone receptor ( PGR ) (Fig.  1 ). Interestingly, residual stromal fibroblasts were shown to undergo mesenchymal-to-epithelial transition and become incorporated into the re-epithelialized luminal surface of the repaired tissue [ 14 ]. This process was supported by scRNA-seq analysis and lineage tracing studies using a mouse model of endometrial repair, which identified platelet-derived growth factor alpha (PDGFRa) as a marker for the stromal cells involved [ 15 ]. The E2 peak further promotes glandular epithelial cell proliferation, inducing upregulation of tissue inhibitor of metalloproteinases 1 ( TIMP1 ) and cell adhesion molecule 1 ( CADM1 ). Thrombospondin-1 ( THBS1 ), which encodes for an adhesive protein, is also highly expressed (Fig.  1 ). THBS1 is involved in tissue remodeling by binding to ECM components and facilitates angiogenesis to provide oxygen and nutrients for tissue re-growth [ 11 , 16 , 17 ]. Moreover, THBS1 can mitigate the menstruation-driven inflammatory response by binding to cell-surface receptors (such as a4b1 integrin) of inflammatory cells, thereby restoring the overall immune homeostasis when proceeding to the next stages of the menstrual cycle [ 18 ]. Importantly, cilium assembly is enhanced at this stage, driven by the synergistic effect of E2 signaling activation and NOTCH signaling suppression, culminating in the upregulation of forkhead box J1 ( FOXJ1 ), a key transcription factor (TF) essential for ciliogenesis [ 19 ]. This tissue expansion process continues in phase 2, characterized by hormone responsiveness ( ESR1 , PGR ) and increased energy consumption (CKB , SCGB1D2 , HGD , IDO1 , ATP1A1 , PSAT1) [ 11 ] (Fig.  1 ). Cell mitosis is one of the most distinct histological features that differentiates the proliferative phase from the secretory phase [ 20 , 21 ]. Interestingly, endometrial G1/S and G2/M signatures are elevated in phase 1 and 2, but cease in the later phases, indicating that transition from proliferative to secretory phase occurs between phases 2 and 3 [ 11 ]. Phase 3 represents the early secretory phase and is marked by upregulation of MTF1 , a transcriptional activator of the metallothionein I promoter regulating the expression of multiple metallothionein I genes ( MT1E, MT1F, MT1G, MT1X ) [ 11 , 22 ] (Fig.  1 ). These heavy-metal-binding proteins are involved in protecting the body against oxidative stress and inflammation and have been advanced as potential key regulators in preparing the endometrium for upcoming pregnancy [ 11 , 23 ]. During this phase, epithelial cells also undergo morphological and functional changes, thereby increasing secretory activity [ 1 , 24 ]. Interestingly, a recent study proposed novel early secretory phase-specific markers showing that epithelial cell differentiation is not yet completed during this phase, with luminal and glandular cells remaining in a pre-luminal and pre-glandular state, respectively [ 25 ]. Sulfotransferase family 1E member 1 ( SULT1E1 ) has been proposed as a pre-luminal marker (Fig.  1 ), which is tightly regulated by P4. It encodes the enzyme estrogen sulfotransferase which plays a crucial role in the metabolism of estrogens by catalyzing their sulfation processes. OPRK1 , SUFU , CBR3 and HPRT1 were proposed as markers for the pre-glandular stage. Although their expression is not yet characterized in the human endometrium, both cell populations were recently successfully mapped to early, but not mid secretory samples using spatial transcriptomics [ 25 ]. The most noteworthy transcriptomic changes in the endometrial epithelium occur during the transition from phase 3 to 4 (Fig.  1 ), with an abrupt activation of a specific gene module upon entering phase 4. Genes within this module include PAEP , glutathione peroxidase 3 ( GPX3 ), C-X-C motif chemokine ligand 14 ( CXCL14 ) and dipeptidyl peptidase 4 ( DPP4 ), all previously pinpointed as pivotal genes during the WOI [ 26 , 27 ]. Interestingly, the expression of monoamine oxidase A ( MAOA ) was also significantly increased at the onset of phase 4 [ 11 , 21 ]. Although primarily recognized for its role in the nervous system, MAOA is proposed as a critical marker of endometrial receptivity, as it inhibits endometrial proliferation through the activation of forkhead box O1 ( FOXO1 ), thereby forging the required environment for the nesting embryo [ 28 ]. Moreover, the TF assembly reaching peak expression during this phase is enriched with early developmental regulators of differentiation ( IRX3 , PAX8 ), underscoring the progression of the endometrium towards a mature state. In contrast to its abrupt opening, closure of the WOI occurs much more gradually. Here, three major gene groups with distinct transcriptomic characteristics were identified (Fig.  1 ): group 1 genes ( PAEP , GPX3 ) show sustained expression throughout phase 4 and uphold noticeable expression in phase 1 of the subsequent cycle. Group 2 genes ( CXCL14 , MAOA , DPP4 ) notably decline, whereas group 3 genes ( MMP7 , THBS1 ) arise and continue their expression while entering phase 1 [ 11 ]. Of note, Marečková et al. identified a glandular secretory population in the late secretory phase with upregulated expression of fibroblast growth factor 7 ( FGF7 ), a gene already known to be hormonally regulated and expressed in the endometrial tissue in multiple species [ 25 ]. Taken together, the detailed expression profile has the potential to expose specific markers to distinguish the WOI from the other phases. Intracellular ions are known to play an important regulatory role as secondary messengers in many cellular signaling pathways of the female reproductive system [ 29 ]. Specifically, intracellular calcium plays a pivotal signaling role in key physiological processes of the endometrium such as decidualization and early embryo implantation [ 30 ]. Calcium can be released from the intracellular stores in endometrial cells primarily via activation of the Inositol triphosphate (IP3) and ryanodine receptors on the endoplasmic reticulum, in response to hormonal signals. In the endometrium, changes in intracellular calcium are regulated by specific ion channels and transporters. These include members of the transient receptor potential (TRP) channel superfamily, which mediate extracellular calcium influx in response to hormonal and environmental cues, and store-operated calcium entry (SOCE) channels, such as the STIM1-ORAI1 complex, which replenish intracellular calcium stores following depletion [ 31 ]. Additionally, these ion channels, including calcium, sodium, potassium and chloride channels, exhibit distinct expression patterns and activity throughout the menstrual cycle. Several ion channels have been identified as key players in endometrial function [ 32 ]. The cystic fibrosis transmembrane conductance regulator (CFTR) in the endometrial epithelial cells transports chloride ions from the cytosol to the uterine lumen, which is crucial for maintaining uterine fluid homeostasis and regulating apoptotic activity of endometrial epithelial cells, all critical for embryo implantation. Indeed, CFTR upregulation was shown to lead to increased apoptosis in endometrial epithelial cells, as well as to abnormal fluid accumulation within the uterine lumen, hence disrupting the supportive environment needed for successful embryo implantation [ 7 , 33 – 35 ]. CFTR expression plays a prominent role in the receptivity of the endometrium and reproductive outcomes [ 34 ]. The epithelial sodium channel (ENaC) is highly expressed in endometrial epithelial cells. Activation of ENaC participates in uterine fluid absorption, by decreasing uterine fluid during the phases preceding implantation. Moreover, activation of ENaC is crucial to initiate the decidualization process by triggering the release and production of prostaglandins [ 36 ]. Of note, activation of ENaC causes depolarization of the membrane potential and induces activation of voltage-dependent calcium channels (VDCC) [ 37 , 38 ]. VDCC, particularly L-type, and potassium channels, including the calcium-activated potassium channels IKCa and BKCa, have also been implicated in embryo implantation [ 34 , 39 ]. However, the functional expression of VDCC in mouse and human endometrial epithelial cells could not be verified in other studies [ 40 ]. Mechanosensitive ion channels may also be involved in the early embryo-uterus crosstalk, as suggested by the high functional expression of the mechanosensitive PIEZO1 channels in epithelial cells of both mouse and human endometrium [ 40 ]. However, the physiological impact of these mechanosensors in the endometrium is still unclear. TRP channels represent another important superfamily of ion channels expressed in the endometrium [ 41 ]. This group of ion channels play a crucial role in the homeostasis of cations (like sodium, calcium and magnesium) and the regulation of calcium entry into endometrial cells, which is essential for numerous cellular functions such as gene expression, secretory activity and cellular motility. Additionally, TRP channels respond to hormonal fluctuations and mechanical stimuli in the endometrium, influencing key reproductive events such as tissue remodeling, inflammation, and cell migration, adhesion and proliferation [ 29 ]. These processes are critical for the repetitive monthly rebuilding of the endometrium. In addition, activation of several G-protein coupled receptors (GPCRs) cause the phospholipase C/IP3-induced depletion of the calcium stores of the endoplasmic reticulum and provoke intracellular calcium oscillations which regulate endometrial receptivity [ 42 ]. One example of such GPCR is the protease-activated receptor 2 (PAR2), which is specifically expressed in the plasma membrane of endometrium epithelial cells and modulates intracellular calcium levels upon activation by serine proteases like trypsin. This PAR2-induced calcium release plays a key role in processes such as inflammation and tissue remodeling, and is suggested to act as a biosensor for embryonic signals [ 42 , 43 ]. To examine the expression of these channels and receptors across different stages of the menstrual cycle, a publicly available endometrial scRNA-seq dataset was analyzed [ 1 ] (Fig.  2 a). In this dot plot, the expression of ion channels in human endometrial epithelial cells is visualized across the menstrual cycle phases, with temporal progression reflected along the y-axis proliferative, early secretory, mid secretory, and late secretory. Dot size represents the proportion of cells expressing a given gene, while color intensity indicates the average gene expression level. Notably, the secretory phase is subdivided into three transcriptionally defined stages based on clustering analysis of epithelial single-cell transcriptomes, to ensure consistency with the original publicly available scRNA-seq dataset. However, it is important to note that scRNA-seq has limited sensitivity for detecting lowly expressed genes, which may lead to an underestimation of the presence or variability of certain channels and receptors in the dataset [ 44 ]. In addition, RNA-levels do not always represent the functional expression of the ion channels, as these proteins are subject to post-translational modifications. Nevertheless, the observable expression changes underscore the physiological importance of these molecules in endometrial function and cyclical remodeling. Along this line, altered expression or function of these ion channels and receptors has been correlated with impaired receptivity, implantation failure and linked to several endometrial pathologies, making them interesting targets for diagnostic and therapeutic interventions [ 39 , 45 , 46 ]. While intracellular Ca 2+ regulation and ion channel activity are proposed to be important in endometrial function, their precise roles remain largely unexplored, at least partly due to the lack of relevant in vitro models. Further ion channel research using recently developed, relevant models (such as organoids) could provide critical molecular and functional insights into gynecological pathologies and implantation-related infertility problems. Fig. 2 Plasma membrane receptor transcriptomics in human endometrial epithelial cells and organoids.  (a) Dot plot displaying the log₂-transformed expression levels of genes encoding selected ion channels and receptors potentially implicated in embryo implantation and endometrial receptivity in endometrial epithelial cells across the different phases of the menstrual cycle [1]). This panel represents members of the Transient Receptor Potential (TRP) superfamily.  (b) Expression of the same genes in endometrial organoids under different hormone exposure conditions, modeling cycle phases in vitro. Dot size indicates the proportion of expressing cells; color shows mean expression (log₂).  CACNA1 genes represents proteins of the voltage-gated calcium channel family,  F2RL1 encodes the Protease Activated Receptor 2 (PAR2), KCNMA1 encodes the beta subunit of the calcium activated BK channel,  KCNN4 encodes the Potassium Calcium-Activated Channel Subfamily N Member 4, SCNN1A encodes the epithelial sodium channel (ENAC),  PIEZO1 the mechanosensitive ion channel PIEZO1, and CFTR (Cystic Fibrosis Transmembrane Conductance Regulator) Plasma membrane receptor transcriptomics in human endometrial epithelial cells and organoids.  (a) Dot plot displaying the log₂-transformed expression levels of genes encoding selected ion channels and receptors potentially implicated in embryo implantation and endometrial receptivity in endometrial epithelial cells across the different phases of the menstrual cycle [1]). This panel represents members of the Transient Receptor Potential (TRP) superfamily.  (b) Expression of the same genes in endometrial organoids under different hormone exposure conditions, modeling cycle phases in vitro. Dot size indicates the proportion of expressing cells; color shows mean expression (log₂).  CACNA1 genes represents proteins of the voltage-gated calcium channel family,  F2RL1 encodes the Protease Activated Receptor 2 (PAR2), KCNMA1 encodes the beta subunit of the calcium activated BK channel,  KCNN4 encodes the Potassium Calcium-Activated Channel Subfamily N Member 4, SCNN1A encodes the epithelial sodium channel (ENAC),  PIEZO1 the mechanosensitive ion channel PIEZO1, and CFTR (Cystic Fibrosis Transmembrane Conductance Regulator)

Most reproduction studies in vivo rely on rodent models (mainly mice), which exhibit significant physiological differences compared to humans, in terms of cycle length (e.g. 4-5-day estrous cycle in mice versus 28-day menstrual cycle in humans), and absence of cyclic spontaneous decidualization and menstruation in rodents. Moreover, mouse and human present dissimilar hormonal oscillations, and differ in uterine structure and embryo orientation during implantation. Collectively, animal models are limited in their ability to accurately replicate human reproductive physiology, alongside inherent ethical and technical constraints [ 47 ]. In vitro studies mainly use adenocarcinoma-derived (Ishikawa, HEC-1A, KLE) and immortalized (T-HESC) cell lines to study endometrial activity and characteristics. However, these models fail to accurately replicate physiological conditions due to their simplified (2D) microenvironment, phenotypic and genetic instability and functional alterations in culture [ 48 ]. For instance, Ishikawa cells are prone to losing cellular phenotype (such as loss of hormone receptor expression), exhibiting genetic modifications during prolonged culture, and showing instability at microsatellite loci. Moreover, most popular cell lines used are cancer-derived or immortalized, and do not represent the primary genomics of healthy cells [ 49 ]. On the other hand, primary cell cultures of human endometrial samples show only limited proliferative capacity and lose their phenotype rapidly during cell culturing. In addition, the cell lines and primary cells are mostly cultured in non-physiological 2D conditions. Alternative models, such as epithelial endometrial cells derived from bovine or non-human primate sources, have been explored yet failed to optimally replicate the human condition [ 50 , 51 ]. Human in vitro research has advanced significantly with the development of organoids, 3D cellular models originating from tissue-specific stem cells or pluripotent stem cells. These cells exhibit self-organizing, self-renewing and proliferative properties when cultured in an extracellular matrix (ECM) scaffold (typically Matrigel) in the presence of a specified cocktail of growth factors, in particular encompassing WNT-activating signals. Importantly, organoids are vastly expandable in culture while robustly maintaining their phenotype. Since their first establishment in 2009 from mouse intestine [ 52 ], organoids have been successfully established from various organs including, among others, brain, lung, kidney and placenta. Turco et al. and Boretto et al. established the first organoid models from human and mouse endometrium [ 53 , 54 ]. These organoids were found to accurately reproduce morphological and functional hallmarks of the endometrial epithelium, including glandular-like organization and presence of apical-faced microvilli and cilia. Moreover, exposure to the endometrium-regulating ovarian hormones E2 and P4 recapitulated aspects of the menstrual cycle, such as the proliferative phase with increased cell proliferation and associated gene expression. In addition, in conditions inducing the secretory phase (high levels of E2 and P4), the organoids showed a pseudostratified, folded glandular epithelial appearance with cilia, and upregulation of decidualization markers such as PAEP [ 53 – 55 ]. Hormone withdrawal induced organoids to shed dead cells mimicking the menstrual phase of the cycle [ 54 ]. To date, endometrial organoids have emerged as invaluable tools for investigating human endometrium (patho)physiology [ 54 , 56 ]. The organoids overcome several limitations associated with 2D in vitro models, such as the absence of biological 3D conditions and physiological phenotype, the limited availability and expandability of primary human endometrium epithelial cells, and the disappearance of the endometrial phenotype in culture [ 57 ]. Despite the endometrium being a highly dynamic tissue with remarkable regenerative capabilities, a specific population of endometrial stem cells has yet to be pinpointed. Over time, several epithelial stem/progenitor cell candidates have been proposed, as comprehensively reviewed in [ 24 ]. Markers such as N-cadherin ( NCAD) , stage-specific embryonic antigen 1 ( SSEA-1 ) and leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) have been advanced as potential markers [ 58 ]. The ability to generate organoids from endometrial epithelium points to the existence of a genuine stem/progenitor cell population, and organoid technology will be a valuable tool for tracking down these cells and deciphering their exact role in the endometrial (patho)physiology. While morphological and functional assays have demonstrated their resemblance to primary endometrial epithelium, the precise cell-type composition of endometrial organoids remains to be fully elucidated. Here, we provide a comprehensive overview of what is known and how it compares to the in vivo situation. In recent years, several research groups have performed scRNA-seq analyses to map the cellular landscape of the endometrium. Recently, Marečková et al. compiled a Human Endometrial Cell Atlas (HECA), a comprehensive (sc) reference map that integrates both previously published and newly generated sc transcriptomic datasets [ 25 ]. The large-scale resource also includes validation through spatial transcriptomics, providing a detailed map of cell populations and gene expression patterns across the human endometrium and cycle phases [ 25 ]. In general, four main cell categories are distinguished: (1) epithelial, (2) mesenchymal, (3) immune (myeloid and lymphoid) and (4) endothelial cells [ 1 , 11 , 16 , 25 , 59 , 60 ]. Since organoids only represent the epithelial component of the endometrium, the transcriptomic portrait provided here focuses on the endometrium’s epithelial cell cluster. In the proliferative phase, a SOX9-expressing (SOX9 + ) cell population is identified [ 25 ] (Fig.  3 ). The group can be further subdivided in SOX9 + LGR5 + cells (expressing MMP7 , ESR1 , KRT17 , WNT7A ), proliferative SOX9 + cells (expressing MKI67) containing both LGR5 + and LGR5 − cells, and SOX9 + CDH2 + LGR5 − cells (expressing AXIN2 , ALDH1A1 , IHH ) [ 1 ]. Spatial transcriptomics mapped these populations to (i) the luminal epithelium (SOX9 + LGR5 + ); (ii) regenerative superficial glandular layer (proliferative SOX9 + ) and (iii) basal glands (SOX9 + CDH2 + LGR5 − ) (Fig.  3 ) [ 1 , 25 ]. Furthermore, Tan et al. discerned an epithelial subset expressing mucin 5B (MUC5B + ) ( RUNX3 , TFF3 , SAA1 ) [ 16 ] (Fig.  3 ). However, doubts have been raised regarding the source of this MUC5B + cell population, hypothesized to rather originate from cervical contamination in the endometrial samples. This statement was based solely on in silico analyses, involving a matched MUC5B + cell population after projection of a cervical reference map onto the HECA [ 25 ]. However, high similarity with cervical MUC5B + cells does not necessarily exclude their presence in the endometrial tissue [ 9 , 25 ]. Hence, further studies are needed to definitively determine the precise origin of the MUC5B + cell population. Fig. 3 Schematic overview of the spatio-temporal organization of the human endometrial epithelium in vivo and its corresponding 3D organoid model in vitro . In organoid models, the proliferative and secretory phases are recapitulated by exposure to estrogen receptor (ER) agonist for the proliferative phase, and a combination of ER agonist + progesterone receptor (PGR) agonist + cAMP for the secretory phase. Corresponding markers are depicted in the organoids. Figure created with BioRender.com Schematic overview of the spatio-temporal organization of the human endometrial epithelium in vivo and its corresponding 3D organoid model in vitro . In organoid models, the proliferative and secretory phases are recapitulated by exposure to estrogen receptor (ER) agonist for the proliferative phase, and a combination of ER agonist + progesterone receptor (PGR) agonist + cAMP for the secretory phase. Corresponding markers are depicted in the organoids. Figure created with BioRender.com To date, the spatio-temporal nature of ciliated cells in the human endometrium remains debated and requires further elucidation. Wang et al. visualized their spatial distribution in both luminal and glandular epithelium using RNA in situ hybridization (RNAscope) and protein immunostaining with antibodies against the newly identified cilia-associated markers ( C11orf88 , C20orf85 , FAM183A and CDHR3) , and further enriched the list of specific markers for the ciliated epithelial cells ( PIFO , FOXJ1 , TPPP3 , TP73 , C11orf88 , C20orf85 , FAM183A , CDHR3 ) (Fig.  3 ) [ 11 ]. The dynamics of ciliated cells during the transition from proliferative to secretory phase is also still under debate. It was initially proposed that P4 suppresses the number of ciliated cells during the secretory phase, to prioritize for secretory and glandular cell functions [ 25 ]. However, recent studies suggest that the number of ciliated cells further increases during the secretory stage, and that P4 inhibits ciliary beat frequency rather than cilium assembly to create a suitable environment for embryo implantation [ 1 , 61 ]. The secretory phase is characterized by a marked reduction in SOX9 + cell populations as the endometrium differentiates to prepare for blastocyst implantation. During this process, pre-luminal ( SULT1E1 ) and pre-glandular ( OPRK1 , SUFU , CBR3 , HPRT1 ) cells differentiate into mature luminal ( WNT7A , LGR5 , PTGS1 ) and glandular ( ITGA1 , HPGD , SLC47A1 ) cells, respectively [ 25 ]. Moreover, a new post-ovulatory population of secretory cells ( PAEP , SCGB1D2 , CCL20 , DEFB1 , CXCL8 ) emerges (Fig.  3 ). To study the menstrual cycle in vitro, endometrial organoids are subjected to in vivo mimicking hormonal regimens [ 53 , 54 , 62 ]. In particular, to recapitulate the proliferative phase, organoids are exposed to an estrogen receptor (ER) agonist (typically E2), while a combination of E2 with a progesterone receptor (PGR) agonist (typically P4 or medroxyprogesterone acetate (MPA)) and decidualization-promoting cyclic adenosine monophosphate (cAMP) is employed to replicate the transition to the secretory phase [ 53 , 54 , 63 ]. This hormonal priming induces the organoids to exhibit phase-specific morphological/cellular behavior and molecular characteristics and allows the study of endometrial cycling physiology. Sc profiling of endometrial organoids [ 1 , 16 , 62 ] revealed that in culture conditions without hormones, most cells are in a proliferative state ( TOP2A , PCNA ) and express ESR1 , indicating their existing responsiveness to E2 (Fig.  3 ). Of note, no ciliated cells were identified in this baseline organoid culture condition, while a so far unidentified cluster of keratin 17 ( KRT17 ) positive cells ( KRT17 , SOX9 , MMP7 , PLAU , IL32 ) could be distinguished, representing an estimated 20% of the cells present [ 1 ]. This KRT17 + population showed the highest similarity to the SOX9 + cell population, and more specifically the SOX9 + proliferative and SOX9 + LGR5 + populations as present in vivo [ 1 ]. When organoids were exposed to E2, three additional cell populations arose (Fig.  3 ). A PGR -expressing population emerged as a result of activated estrogen signaling, along with a preciliated population ( CCNO , CDC20B ) that closely aligns with the in vivo counterpart [ 1 ]. Moreover, a MUC5B  + cell population was also observed in the E2-treated endometrial organoids [ 16 ]. Interestingly, purified MUC5B + epithelial cells generated organoids that were significantly larger and more numerous than organoids derived from MUC5B − cells. Notably, Tan et al. cultured the organoids using organoid medium supplemented with E2. However, no data are available on MUC5B + -derived organoids cultured without hormones or under conditions supplemented with E2 + PGR agonists + cAMP, thus limiting the ability to draw conclusions on the phenotype and hormone-responsiveness of the MUC5B + epithelial cells [ 16 ]. Additional exposure of the organoids to P4 or MPA, together with activation of cAMP signaling, creates a decidualized endometrial phenotype comparable to the in vivo situation [ 53 , 54 , 64 ]. scRNA-seq analysis revealed high expression of advanced differentiation-stage markers, including those of secreted products ( PAEP , DEFB1 ), glands ( SCGB2A2 ) and ciliated cells ( FOXJ1 , PIFO ) (Fig.  3 ). Overall, the hormone-induced differentiated organoids markedly mirror the in vivo endometrial epithelium across cycle stages. These organoids exhibit the presence of key cell types, such as ciliated and secretory cells, and express genes associated with physiological processes such as decidualization, reflecting the functional characteristics of the endometrium during the menstrual cycle. However, it is important to acknowledge that our understanding of the full spectrum of cell types present within organoids remains incomplete. Some populations (indicated as “unknown” in Fig.  3 ), are yet to be fully characterized, highlighting the need for additional research. Furthermore, it should be mentioned that there is currently no standardized protocol for secretory-phase differentiation of endometrial organoids. Although exposure to E2, P4 and cAMP is widely used, the specific conditions, including the type of PGR agonist (e.g., P4 vs. MPA), the timing of E2 exposure and the concentrations used vary between most of the studies (overview of different protocols in Table  2 ). This variability complicates the ability to draw detailed conclusions about the cell types and behavior based on scRNA-seq data. Thus, while general conclusions can be made regarding hormone responses, potential differences because of distinct hormone exposure protocols should still be considered. Table 2 Overview of human endometrial organoid studies and used hormonal differentiation protocols References in bold include publications with publicly available scRNA-seq datasets of human endometrial organoids. Abbreviations: NH, no hormones; d, day; ExM, (organoid) expansion medium; PRL, prolactin; hPL, human placental lactogen; hCG, human chorionic gonadotropin. Colors: organoid expansion without hormones (grey), proliferative-like phase (typically with E2) (blue), secretory-like phase (typically with P4) (pink) Overview of human endometrial organoid studies and used hormonal differentiation protocols References in bold include publications with publicly available scRNA-seq datasets of human endometrial organoids. Abbreviations: NH, no hormones; d, day; ExM, (organoid) expansion medium; PRL, prolactin; hPL, human placental lactogen; hCG, human chorionic gonadotropin. Colors: organoid expansion without hormones (grey), proliferative-like phase (typically with E2) (blue), secretory-like phase (typically with P4) (pink) Currently, the main limitation of organoids in terms of cellular complexity is the apparent lack of luminal cells and associated spatial organization. García-Alonso et al. investigated the regulation of epithelial cell differentiation throughout the menstrual cycle by examining the activity of lineage-specific TFs. A close interplay between WNT and NOTCH signaling was observed, establishing luminal-glandular gradients across the endometrium (Fig.  3 ). WNT-activated TFs ( FOXJ1 ) were mostly expressed in the ciliated lineage at the luminal surface, while WNT-inhibited TFs ( FOXO1 , CSRNP1 ) and NOTCH-induced transcriptional regulators ( HEY1 , HES1 ) were activated in the secretory lineage in the glands of the lamina functionalis [ 1 ]. Concordantly, endometrial organoids cultured with WNT inhibitors show a low proportion of preciliated and ciliated cells and a high proportion of secretory cells which is the opposite with NOTCH inhibitors [ 1 ]. Recent studies have also explored whether endometrium-derived organoids can be used as a valid model to characterize the functional expression of endometrial ion channels and receptors. These studies illustrated highly comparable expression levels, specifically of TRPM4 , TRPM7 , TRPV2 and TRPV4 , in endometrial organoids (baseline culture condition) compared to short-term cultured endometrial epithelial cells [40]. In addition, the functional expression of the mechanosensitive PIEZO1 channels in endometrial organoids, evaluated by monitoring intracellular calcium changes, more closely aligned with its activity in primary endometrial epithelial cells compared to the Ishikawa cell line, thereby further underscoring the superior functional mimicry of endometrial organoids. Notably, analysis of scRNA-seq data (from [1]) for different subunits of the voltage-gated calcium channels (CACNA) showed only marginal expression levels in endometrial organoids in baseline culture conditions (Fig. 2b). However, expression of CACNA1C was upregulated following E2/P4/cAMP treatment which correlates with upregulation during the secretory phase in vivo (as observed in endometrial scRNA-seq dataset, Fig. 2a). The CACNA1F expression increases under E2 exposure corresponding to its increase during the proliferative phase in vivo . However, at the functional level, no increase in current density was detected after application of a voltage step protocol in endometrial epithelial cells in patch clamp experiments, suggesting the absence of functional voltage-dependent calcium channels. This observation could potentially be explained by the very low expression levels of other subunits of the channel, as needed for activity. This hypothesis is further supported by the current dataset indicating that CACNA1F is expressed at levels insufficient to support functional channel formation [ 65 ]. In general, expression levels of these ion channels appear lower in endometrial organoids compared to primary endometrial epithelial cells (Fig.  2 b vs. Fig.  2 a). These findings suggest that the specific microenvironment may be affected by the absence of non-epithelial cell types, although hormone-regulatory mechanisms are present in the organoids. In summary, the organoid model recapitulates most of the endometrial cell types and maintains the gene expression complexity. Ultimately, a more comprehensive understanding of the signaling pathways driving endometrial cellular differentiation will enable enrichment and fine-tuning of organoids towards specific cell fates, thereby allowing deeper research into endometrial physiology.

Human endometrial organoids represent innovative 3D biomimetics that open a vast horizon of previously unattainable applications. In this section, we provide an overview of several applications based on the endometrial organoid technology. Building upon the success of organoid models, researchers are developing in vitro co-culture systems incorporating, among others, stromal, immune, and endothelial cells, referred to as endometrial assembloids, aiming at better replicating the multicellular architecture and functionality of the endometrium [ 27 , 82 – 84 ]. Rawlings et al. demonstrated the integration of human gland-like epithelial organoids within a stromal matrix, and extended their study to include embryos in the assembloid co-culture [ 85 ]. Another study developed air–liquid interface (ALI) endometrial assembloids combining stromal and epithelial cells within a matrix, mimicking the human endometrial anatomy with stromal, glandular and luminal epithelia, and showing menstrual cycle-related hormone responses, gene expression and dynamic ciliogenesis [ 86 ]. Since the endothelial compartment is fundamental to endometrial function, its integration into models has also been explored. Adding human umbilical vein endothelial cells (HUVECs) to apical-out (AO) hormone-responsive endometrial organoids (that replicate the natural orientation of endometrial epithelial cells), organoids self-organized to form an integrated endothelial network [ 87 ]. This approach addresses potential challenges such as apoptosis in the center of organoids and enhances the physiological relevance of the model for studying endometrial function. The first paracrine and physical interplay between the blastocyst and the endometrium is critical for successful pregnancy, with 30% of conceptions lost at the time of implantation [ 88 ]. However, our current understanding of the early implantation events is very limited, mainly because of inaccessibility in vivo as well as lack of appropriate in vitro study models. Recently, a novel human in vitro implantation model was developed by turning the closed endometrial organoid model into an ‘open-faced endometrial layer’ (OFEL), thus opening the apical epithelial surface of the organoid to allow the accessibility of the embryo. This adhesion model mimics the embryo-endometrium interface and simulates positioning as in the womb [ 27 ]. Interestingly, stem cell-derived human blastocyst models (blastoids) were shown to attach to the OFEL, yet only in physiologically permissive conditions, i.e. when the OFEL was hormonally primed (with a mixture of E2 + P4 + cAMP + XAV939) to reach the receptive WOI state, supporting the physiological relevance of this new embryo adhesion model. Upon attaching, the blastoids started to produce pregnancy hormone (human chorionic gonadotropin-β, hCGβ) by the trophectoderm [ 27 ]. In addition, an endometrial AO assembloid model was used in a floating, suspension co-culture system with blastoids and was found to mimic embryo implantation [ 87 ]. Disruption of endometrial epithelial cells by syncytial cells was observed, which invaded and fused with the endometrial stromal cells. Taken together, these advanced organoid-based models are expected to close the knowledge gap of human embryo implantation, enabling breakthroughs in the study of female infertility. Organoids have already proven to be powerful tools for drug screening and discovery, also with respect to personalized medicine. Indeed, conventional treatment of gynecologic malignancies often involves a standardized regimen, which may not be effective for every individual. The patient-derived organoids from endometrial pathologies will allow preclinical drug testing as well as toxicity screening, offering a promising solution for more personalized medicine. For instance, endometrial and ovarian cancer-derived organoids reflect individual responses to therapy underscoring their value as preclinical platform for personalized cancer treatment [ 89 ]. By integrating organoid technology into drug and toxicity testing [ 90 ], the accuracy of preclinical assessments will be enhanced, and more effective, personalized treatment strategies for gynecological cancers will be developed. Furthermore, the ability to create and introduce specific mutations in patient-derived endometrial epithelial organoids (see section on genetic and epigenetic studies) enhances the potential for customized therapeutic strategies. This approach will enable evaluation of tissue responses to various treatments, particularly for conditions like EC, where patient-specific organoids can help to identify the most effective chemo- or targeted therapy options [ 91 , 92 ]. Beyond oncology, advancements in reproductive medicine also leverage endometrial organoids to personalize hormonal therapies aimed at improving endometrial receptivity in infertile women by mimicking the WOI. Analysis of these organoids enables clinicians to determine the most effective timing and hormonal regimen, thereby improving the success rates of assisted reproductive technologies (ART) [ 93 ]. Organoids are amenable to genetic editing tools like CRISPR/Cas9, enabling detailed investigations into gene functions, disease mechanisms and epigenetic impacts. Innovative techniques such as nanoblades (virus-like particles) deliver CRISPR/Cas9 components or RNA/DNA directly into cells with high precision and minimal toxicity [ 94 ]. Such genetic interventions enable researchers to mimic diseases like EC and explore targeted personalized therapies [ 95 – 98 ]. However, a key limitation remains in achieving efficient and uniform CRISPR/Cas9-based transduction within intact 3D organoid structures. Current approaches, as shown for human liver ductal organoids, addresses this challenge by using electroporation to introduce genetic material, followed by dissociation into single-cells to isolate successfully edited cells based on fluorescence markers [ 95 ]. This additional step of dissociation is required because effective transduction across all cells within the 3D structure is not yet feasible, making it difficult to preserve the organoid’s spatial organization and cell-cell interactions. The transplantation of organoids into host tissues has emerged as a promising strategy in regenerative medicine, offering new possibilities for treating various tissue injuries and degenerative conditions [ 99 ]. In the field of endometrial research and infertility [ 100 ], organoids have been used to model conditions such as AS and adenomyosis, among others. As discussed in the previous paragraph, endometrial organoids, both mouse and human-derived, have been successfully employed as grafts to treat post-traumatic endometrial regeneration disorders in mouse models [ 80 , 81 ]. Additionally, comparing the regenerative potential of bone marrow mesenchymal stem cells (BMSCs) with endometrial organoids in rats demonstrated that regeneration of the injured endometrium was superior when endometrial organoids were used [ 100 , 101 ]. The integration of endometrial organoids into functional glandular epithelium also resulted in increased fertility outcomes, outperforming BMSCs in terms of regenerative capability. A major advantage of using organoids in regenerative medicine is the use of autologous tissue transplantation to avoid tissue rejection. When derived from patient-specific cells, organoids are immunologically compatible with the recipient, reducing the risk of immune rejection and enhancing the safety of the treatment. This property makes organoid transplantation an attractive option for personalized regenerative therapies, particularly in the treatment of diseases like AS, where the endometrial tissue requires careful restoration. Taken together, the application of endometrial organoids in regenerative medicine offers significant promise for advancing the treatment of endometrial disorders, paving the way for personalized, effective therapies aimed at restoring tissue function and fertility. Organoids can be integrated into microfluidic systems, commonly referred to as organ-on-chip (OoC) models, to create advanced platforms for studying the native tissue. These systems allow precise control over fluid flow, mimicking physiological processes in a dynamic microenvironment [ 102 ]. Incorporating endometrial organoids into OoC models provides a robust tool for replicating key processes, including cyclical hormonal fluctuations and controlled pharmaceutical interventions [ 103 ]. Interestingly, OoC models have been developed for various parts of the female reproductive system, including the oviduct [ 104 ] and placenta [ 105 ]. Notably, other OoC models incorporated epithelial, stromal and microvascular endometrial components within a dynamic flow system, allowing a more comprehensive understanding of endometrial physiology [ 106 ]. Furthermore, a microfluidic platform has been developed that combines ten distinct organ models, including the endometrium, along with the kidney, liver, heart and others. This platform successfully maintained the phenotypic functions of all ten mini-organs for up to four weeks, demonstrating its potential as a multifunctional system to explore interactions between the endometrium and other organs, as well as endocrine influences [ 107 ]. Future plans will increase the complexity of OoC systems by incorporating multiple cell populations and organs to further assess their interactions [ 108 ].

Organoids are powerful tools for human disease modeling, with the potential to provide new insights into the molecular and cellular mechanisms underlying complex gynecological pathologies, including endometrial cancer, endometriosis, adenomyosis and Asherman’s syndrome. Here, we describe the different endometrial pathologies for which organoids have been developed and characterized. Endometrial cancer (EC) ranks among the most common gynecological malignancies in women worldwide, with rising incidence rates attributed to factors such as obesity, sedentary lifestyles and aging populations [ 66 ]. It primarily affects the endometrial lining of the uterus and is broadly categorized in two main types, i.e. type I (endometrioid) which is often E2-dependent and associated with better prognoses, and type II (non-endometrioid), which is more aggressive, less hormone-dependent and linked to poorer outcomes [ 67 ]. Despite recent advances in the EC research field, cellular and molecular mechanisms driving EC remain poorly understood, largely due to a lack of high-fidelity models that replicate the disease pathology. Conventional models, such as Ishikawa cells, are limited by genomic instability and loss of primary tumor phenotype, underscoring the need for more reliable models [ 68 ]. Organoids have been successfully developed from EC primary tumor samples of a spectrum of clinical types, ranging from hyperplasia to low- and high-grade cancer [ 64 , 69 ]. The organoids accurately recapitulate the original EC’s subtype, and show (and maintain) the tumor’s genotype and mutational profile. Of note, organoids are interesting disease models to harness in drug discovery. Indeed, the EC-derived organoids were found to replicate patient- and tumor-specific responses to drugs, an observation which was further expanded to discover patient-specific drug efficacy (personalized medicine) at higher throughput [ 70 , 71 ]. Taken together, EC-derived organoids provide powerful tools to optimize high-throughput screening approaches, thereby also paving the way towards personalized medicine. Endometriosis is a chronic and debilitating gynecological disorder affecting 10% of reproductive-age women and is characterized by the presence of endometrial-like tissue outside of the uterus (i.e. ectopic lesions). Although the exact cause remains uncertain, researchers consider the theory of retrograde menstruation as one of the most likely mechanisms contributing to the development of endometriosis. Ectopic lesions grow and wane under the influence of the E2/P4 oscillations resulting in a local chronic inflammatory state. Underlying mechanisms of endometriosis’ pathobiology remain poorly understood, partially due to the limitations of existing research models such as cell lines and non-menstruating animals, which fail to accurately recapitulate the disease [ 71 ]. Organoids derived from ectopic endometriosis lesions have shown promising utility in modeling disease-specific characteristics, including the altered expression of ECM-receptor interaction genes ( COL3A1 , FN1 , ITGA11 , LAMA2 ), WNT-pathway genes ( CTNNA2 , LEF1 , WNT11) and hormone-responsive genes ( HSD11B1 , LIFR , PGR , PRLR ) [ 40 , 55 , 64 ]. Additionally, these organoids display luminal invasion with related gene expression ( MMP2 , SNAI2 , TIMP4 ) reminiscent of the in vivo lesions that implant in the ectopic tissues [ 40 , 55 ]. Taken together, endometriosis organoids have facilitated deeper understanding of hormonal dysregulation and inflammatory pathways driving ectopic endometrial tissue implantation and proliferation. These organoid models from endometriotic lesions offer a novel platform for modeling parts of the endometriosis pathobiology and identifying novel treatment options [ 64 , 72 ]. Adenomyosis is a condition characterized by the abnormal presence of endometrial epithelial and stromal cells within the myometrium, the muscular layer of the uterus responsible for uterine contractions. This condition causes hyperplasia and hypertrophy of surrounding smooth muscle cells, triggering uterine cramps, pelvic pain and heavy bleeding. Organoids derived from adenomyosis patients’ endometrial biopsies demonstrated hormone-responsive behavior, showing cycle phase-specific markers such as Ki67, mucin-1 ( MUC-1 ) and glycogen and maintained cell apicobasal polarity and chromosomal stability [ 73 ]. However, adenomyosis-derived organoids demonstrated notable differences compared to healthy endometrial organoids, including overexpression of transforming growth factor beta 2 ( TGFβ2) and SMAD family member 3 ( SMAD3 ). Of note, the TGFβ/SMAD3 signaling pathway is involved in endometrial fibrosis and is thought to play an important role in adenomyosis pathobiology [ 74 , 75 ]. Moreover, the observed overexpression of secretory stage specific markers PAEP, SPP1 and LIF may contribute to impaired endometrial receptivity, potentially underlying adenomyosis-associated infertility [ 75 – 77 ]. These results demonstrate that organoids derived from endometrium of adenomyosis patients show disease-specific traits, making them promising preclinical models to study adenomyosis-related infertility. Asherman’s syndrome (AS), marked by intrauterine adhesions or intrauterine synechiae, is a condition affecting approximately 4 in 10,000 women, yet its pathophysiology remains incompletely understood. AS often arises after surgical procedures such as curettage or cesarean sections, or due to infections like schistosomiasis or genital tuberculosis, which lead to damage and abnormal endometrial repair. The related damage can result in implantation failure, recurrent pregnancy loss, preterm labor and/or placenta accreta [ 78 ]. Organoids derived from the endometrium of AS patients have been developed, providing valuable models for studying the disease. scRNA-seq analysis of AS-derived endometrial organoids has identified significant gene expression alterations, particularly in WNT and NOTCH signaling pathways, mirroring changes seen in the in vivo endometrium [ 79 ]. Furthermore, transplantation of both human and mouse healthy endometrial organoids into a mouse AS model of damaged endometrium demonstrated successful engraftment, repair of AS-induced damage, reduction of fibrotic lesions, and improved fertility outcomes [ 80 ]. In a similar study [ 81 ], it was demonstrated that dysfunctional mitochondria, a likely key feature of AS endometrium, were fully restored through the transfer of functional mitochondria from engrafted endometrial organoids. However, long-term effectiveness has yet to be demonstrated. These findings suggest that endometrial organoids derived from AS patients have great potential as tools for investigating the pathobiology of the disease and that the organoid approach can be used for developing regenerative therapies aimed at restoring endometrial function.

In conclusion, endometrium-derived organoids mark a significant advancement in the research toolbox to assess women’s reproductive health and disease. Importantly, organoids offer significant potential for personalized medicine and the exploration of endometrium-associated disease mechanisms. Despite significant progress, several challenges remain to be addressed to improve the endometrial identity of the endometrial organoids. Notably, the recent endometrial organoids represent only the epithelial fraction and lack the full complexity of endometrial stromal cells, vascular components (endothelial cells) and immune cells. This major limitation is the current research focus of multiple research groups, aiming at new co-culture models (e.g., assembloids and OoC models) to improve the mimicry of the endometrium. Scalability and high-throughput drug screening using endometrial organoids are other research challenges, primarily due to the lack of standardized culturing methods and the limited control over the inherent heterogeneity and genetic variability in organoid cultures [ 109 ]. This variability often causes inconsistencies in experimental outcomes and results in limited reproducibility of datasets among different research teams. A key factor in the experimental protocol is the use of Matrigel as an ECM scaffold, because of its heterogeneity and its poorly defined composition. Solutions can be found in the use of synthetic hydrogels, which offer more precise control over the composition, stiffness, and biochemical properties. Additionally, the use of decellularized tissues or tissue-specific matrices has been proposed as a way to more accurately replicate the in vivo environment of the endometrium [ 110 – 112 ].