Reproductive fluids, commercial media, and organoids: bridging the gap in IVF culture systems.

EVs have been isolated from a wide range of reproductive sources (e.g. human, murine, livestock), including follicular fluid, OF, UF, embryo-conditioned media, and oviduct epithelial cell cultures, underscoring their central role in maternal–embryo communication ( Ma et al. , 2022 ; Li et al. , 2023 ; Hamdi et al. , 2024 ; Muraoka et al. , 2024 ; Pakniyat et al. , 2025 ). These vesicles transport bioactive cargo, such as miRNAs, proteins, and membrane receptors, that can be internalized by gametes and embryos, influencing preimplantation development, developmental competence, cell allocation, oxidative balance, and implantation-related pathways ( Ma et al. , 2022 ; Li et al. , 2023 ). Evidence from animal models further supports their function as dynamic mediators of maternal signaling, modulating embryonic development in a stage, dose, and tissue-specific manner. For instance, murine embryos co-cultured with varying doses of human fallopian tube (FT)-derived EVs (from patients with uterine fibroids) showed improved developmental outcomes, although the effects were not strictly dose-dependent nor consistently linked to a particular donor ( Li et al. , 2023 ). Consistent with the concept of temporally regulated EV signaling, murine oviduct-derived EVs isolated across estrous stages enhance blastocyst yield, cell number, and embryo quality, with the strongest effects in EVs isolated from the diestrus group (representing the highest levels of estrogen receptor, a factor known to support early embryogenesis ( Xue et al. , 2025 ). Importantly, when sequential and EV replenishment supplementation was done (particularly on days 3 and 4 post-fertilization), an increase in hatching rates was observed. This suggests that EV bioactivity declines in static culture, and temporal renewal more closely mimics the continuous in vivo exposure within the reproductive tract ( Xue et al. , 2025 ). While isolated studies in bovines have reported that specific EV-associated components, such as miR-146b, present in conditioned culture medium, may exert detrimental effects on embryos ( Pavani et al. , 2024 ), the broader body of evidence supports a predominantly beneficial role of reproductive EVs in early development. Studies using medium supplemented with microRNA (miR-378a-3p from embryonic sources) and follicular oviductal- and UF-derived EVs support an overall increase in blastocyst yield and hatching rates ( Hamdi et al. , 2018 , 2024 ; Pavani et al. , 2022 ; Benedetti et al. , 2024 ) and, furthermore, enhanced the expression of embryo development and implantation-relevant genes such as IFNT , the key cytokine required for maternal recognition of pregnancy ( Leal et al. , 2022 ; Benedetti et al. , 2024 ). Bovine and murine findings provide a compelling proof of concept, which is further supported by human data in which microRNAs were dysregulated in women with recurrent implantation failure, and were associated with pathways governing endometrial receptivity and implantation competence ( Liu et al. , 2021a ). In this case, endometrial EVs from fertile women improved implantation competence compared with EVs from women experiencing recurrent implantation failure ( Liu et al. , 2020 ). These findings support the concept that fluid-derived microRNAs function not only as biomarkers of reproductive outcome but also as integral regulatory signals within the implantation microenvironment.

Embryo culture medium performance is commonly evaluated based on early developmental parameters such as cleavage rate, cell division kinetics, and blastocyst formation. However, implantation and gestational success, typically measured by live birth rate, remain the most meaningful clinical endpoint ( Gnoth et al. , 2011 ). The period of in vitro embryo culture coincides with the global epigenetic reprogramming of the preimplantation embryo ( Zhu et al. , 2018 ). This developmental window is therefore considered particularly sensitive to environmental influences. Despite this biological susceptibility, relatively few studies have directly assessed the specific impact of culture media on the epigenome of embryos and extraembryonic tissues. So far, the direct effect of IVC conditions on placental DNA methylation (DNAm) is modest and inconsistent across cohorts, with only a limited number of recurrent candidate loci. Analysis of umbilical cord blood (UCB) from neonates conceived under different culture media (G5, Vitrolife; human tubal fluid (HTF) medium, HTF, Lonza) shows no differences in DNAm between media groups ( Koeck et al. , 2022 ). However, when UCB from ART-conceived newborns is compared with that of naturally conceived (NC) newborns, global DNA hypomethylation is observed ( Håberg et al. , 2022 ). Similarly, ART-derived term placentas exhibit less DNAm at imprinted loci such as the PEG1/MEST region ( Nelissen et al. , 2013 ; Barberet et al. , 2022 ), alongside downregulation of TRIM28 , a key stabilizer of genomic imprinting ( Auvinen et al. , 2024 ). Beyond methylation patterns, transcriptomic analyses indicate that ART-associated placental changes are enriched in pathways related to hormonal regulation, insulin secretion, and vascular development, particularly through downregulation of NOTCH3 and DLK1 genes ( Auvinen et al. , 2024 ). Whether these molecular signatures directly translate into apparent structural placental abnormalities, such as accelerated villous maturation, placenta previa, placental abruption, or placenta accreta spectrum, remains ambiguous ( Vermey et al. , 2019 ; Londero et al. , 2021 ; Matsuzaki et al. , 2021 ). Clinically, although children conceived through ART are generally healthy, the aforementioned molecular findings, together with epidemiological evidence, offer a plausible mechanistic link of ART with an increased risk of cardiometabolic and vascular disorders, imprinting disorders, preeclampsia, hypertensive complications of pregnancy, and low birth weight ( Qin et al. , 2017 ; Cui et al. , 2020 ; Pinborg et al. , 2023 ; Auvinen et al. , 2024 ). Birth weight, as a representative outcome, reflects a complex interplay of parental, embryological, and genetic determinants. Within this multifactorial framework, growing evidence suggests that embryo culture media formulation and culture strategy may contribute measurably to perinatal growth patterns, even if their molecular effects are subtle and context-dependent ( Ahlström et al. , 2023 ; Sonigo et al. , 2024 ; Li et al. , 2025 ). For instance, independent and randomized studies comparing human embryo culture mediums, including GI.3, G5, G5-plus (Vitrolife), and HTF versus K-SICM (Cook) medium, showed that singletons cultured using Vitrolife and HTF mediums had higher birth weights ( Dumoulin et al. , 2010 ; Li et al. , 2025 ). In fact, Vitrolife versus Cook culture medium growth differences persisted through the first 2 years of life, highlighting how subtle differences in media composition can have lasting postnatal effects ( Kleijkers et al. , 2014 ). Further examples include mitochondrial DNA (mtDNA) variants, which in NC individuals are associated with body weight ( Flaquer et al. , 2014 ). This relationship appears only partially preserved in ART cohorts, as an enrichment of de novo mtDNA mutations (particularly non-synonymous and rRNA variants, also present in NC groups) seems to be magnified in ART populations, with additional influence from maternal age, ovarian stimulation, and IVC-related factors ( Mertens et al. , 2024 ). In this case, exposure to Cook or UZB (an in-house medium made in the UZ Brussel) embryo culture medium appears to override the effect of mtDNA variation on fetal growth, whereas embryos cultured in Vitrolife medium show similar patterns to those of the NC groups ( Mertens et al. , 2024 ). Importantly, current findings underscore that ART-associated molecular signatures are not yet universal but vary across the populations, study designs, and developmental stages assessed. The lack of standardization in procedural variables (e.g. media type, oxygen conditions, IVF/ICSI embryo transfer protocols) across clinics and even within the same institution creates inconsistencies that can mask subtle biological effects and challenge the interpretation of comparative analyses ( Racowsky et al. , 2026 ). These limitations also reinforce the broader conceptual shift required in the field: moving beyond merely “supportive” culture environments toward culture systems that are actively biomimetic, capable of recapitulating the biochemical, biophysical, and temporal complexity of the reproductive tract. This focus is increasingly driving interest in the use of reproductive-fluid supplementation, dynamic co-culture platforms, and organoid or assembloids-derived secretomes, all of which aim to reintroduce aspects of the physiological niche into embryo culture. Their integration with 3D-cell culture models and embryo co-culture models represents a key frontier in constructing more faithful reproductive microenvironments, and insights from these platforms will likely inform the next generation of biomimetic media.

Current IVF culture systems support acceptable early embryo development but fail to reproduce the dynamic biochemical, mechanical, and temporal cues of the reproductive tract ( Sciorio and Rinaudo, 2023 ). In vivo , preimplantation embryos are exposed to a continuously changing environment shaped by epithelial secretions, mucins, glycoproteins, extracellular vesicles (EVs), fluid flow, and hormonal oscillations ( Kölle et al. , 2009 ; Godakumara et al. , 2023 ). These factors regulate metabolic activity, epigenetic remodeling, developmental timing, and maternal–embryo communication ( Pavličev et al. , 2017 ). In contrast, in vitro culture (IVC) media still rely on static media that lack many of these features, offering only a partial approximation of physiological conditions ( Cairo Consensus Group, 2020 ). These discrepancies between in vivo and in vitro conditions contribute to measurable differences in metabolic programming, transcriptional profiles, and fetal growth trajectories between in vivo - and in vitro -derived conceptuses ( Aksu et al. , 2012 ; Driver et al. , 2012 ). Such limitations underscore the need for more biomimetic culture systems capable of integrating essential biochemical and biophysical properties of reproductive tract fluids. Advances in reproductive tract organoids and assembloids, as well as their secretome, now provide new opportunities to generate defined, reproducible, and physiologically relevant supplements that may overcome the constraints associated with native reproductive fluids. This review synthesizes the current evidence on reproductive fluids, commercial media, and three-dimensional (3D) cell culture-derived factors, highlighting conceptual advances, methodological limitations, and the translational challenges of implementing physiologically inspired culture systems.

Early studies showed that the co-culture of human gametes or embryos with reproductive tract epithelial cells enhanced developmental outcomes ( Bongso et al. , 1989 ; Yeung et al. , 1992 ). These systems aimed to restore the dynamic, bidirectional communication between the embryo and its maternal environment: interactions that are otherwise absent in conventional static culture. Although traditional co-culture systems are no longer used in routine ART due to advances in next-generation media, they have re-emerged as physiological bioreactors to investigate the maternal–embryo crosstalk and uncover molecular signals lost in simplified media. This renewed appreciation for the importance of dynamic maternal signaling prompted the development of advanced microfluidic reproductive platforms. Oviduct or uterus-on-a-chip models represent the next conceptual step, offering dynamic control over fluid flow, spatiotemporal gradients, compartmentalization, and luminal shear forces ( Ferraz et al. , 2018 ; Ahn et al. , 2021 ; Wang et al. , 2022 ), all of which are features impossible to recreate in conventional IVC drops. Dual chamber “endometrium-on-a-chip” devices, for instance, maintain long-term co-culture of stromal and endothelial cells, enabling decidualization patterns while simulating a full 28-day hormonal cycle ( Gnecco et al. , 2017 ). More recent iterations incorporate epithelial, stromal, endothelial, or myometrial compartments in configurations that reproduce the receptivity marker expression and transcriptomic fidelity of native tissue ( Ahn et al. , 2021 ; Busch et al. , 2024 ). Collectively, these platforms begin to approximate patient-specific, multi-cellular uterine microenvironments for functional interrogation of embryo–maternal crosstalk ( Lee et al. , 2025 ). Although the application of microfluidic systems to clinical ART remains largely exploratory, these technologies outline what the next generation of IVF platforms could become: dynamic, automated, and biomimetic. The parallel outset of robotic micromanipulation further illustrates this trajectory. The recent successful births following automated ICSI (ICSIA) demonstrate that automated systems can achieve clinical-grade precision, even when operated by individuals without extensive embryology training ( Costa-Borges et al. , 2023 ; Mendizabal-Ruiz et al. , 2025 ). The convergence of such robotic technologies and microfluidic culture devices suggests a future in which embryo handling, gamete injection, and early development occur within tightly controlled integrated platforms that function simultaneously as culture systems, physiological bioreactors, and precision instrumentation. Together, they offer a path toward standardizing embryo culture conditions, mitigating procedural heterogeneity (e.g. human-handling, lab-specific practices), one of the most significant sources of inconsistency in ART, and enabling the development of universally applicable protocols across laboratories and clinics ( Racowsky et al. , 2026 ).

Assembloids provide a major step forward for studying reproductive tract secretions because their multicellular organization restores epithelial-stromal communication (with unprecedented in vitro fidelity, Figs 2 and 3 ): a key driver of physiologically relevant secretome profiles, including during the window of implantation ( Zhang et al. , 2024 ). Unlike epithelial-only organoids, human endometrial assembloids generate hormone-responsive secretions that more accurately represent the transition to a receptive state. Stromal cells further contribute with decidual products such as PRL, IGFBP1, and senescence-associated signals that shape the composition and timing of the essential molecules for implantation ( Gnecco et al. , 2023 ; Tian et al. , 2023 ). Beyond physiological modeling, endometrial assembloids retain donor-specific signatures and capture clinically relevant alterations, such as progesterone-resistant endometriosis, adenomyosis, and recurrent implantation failure ( Rawlings et al. , 2021 ; Stratopoulou et al. , 2025 ), making them highly relevant for understanding implantation failure and ART outcomes. Current multicompartment assembloids (MA) are innovative constructs that remain in earlier stages of development. FT MA incorporates a basement membrane-like Matrigel core containing the epithelial cells, surrounded by a collagen I-rich stromal compartment ( Fig. 3B ), recreating the epithelial-stromal organization and mucosal folding characteristic of the in vivo FT. This architecture enhances epithelial differentiation, motile ciliation, and associated proteomic diversity, resulting in luminal secretions and fluid dynamics that resemble native tubal physiology. Moreover, cilia developed by the FT assembloids are not only present but functionally active, producing directional ciliary beating sufficient to transport oocyte-sized cargo along the lumen-facing epithelial surface, recapitulating a fundamental mechanical function of the FT ( Crawford et al ., 2024 ). A parallel effort has focused on building endometrial MAs. This platform recapitulates the proliferative, secretory, and menstrual regression phases within a single culture system, supports decidualization, and reveals compartment-specific paracrine signaling. By uniting tissue-informed design with full-cycle hormonal responsiveness, these endometrial MA establish a powerful foundation for studying implantation, modeling gynecological disease, and advancing precision reproductive diagnostics ( Ren et al. , 2025 ). Receptive endometrial scaffold platforms like the cell-engineered receptive endometrial scaffold technology ( Fig. 4A and B ), the 3D endometrioid on a chip ( Fig. 4C ), and the 3D post-implantation co-culture system ( Fig. 4D ) are even able to model the early human embryo and blastoid implantation process in vitro , recapitulating key post-implantation hallmarks such as yolk sac formation, primordial germ cell specification, and early placental development ( Dong et al. , 2025 ; Molè et al. , 2026 ; Song et al. , 2026 ). Crucially, these models not only advance our understanding of maternal–embryo communication but also faithfully reproduce the luminal, glandular, and stromal compartments of the receptive endometrium ( Fig. 4B ), providing a physiologically relevant context for the study and collection of in vitro secretions, including those involved in oviductal, endometrial, and embryonic interactions ( Dong et al. , 2025 ; Molè et al. , 2026 ; Song et al. , 2026 ). Moreover, these overcome the limitations of accessibility and ethical constraints in early pregnancy research. All these features position assembloids as an intermediate system between current organoid cultures, which lack complex tissue interactions, and whole-uterus models, which remain experimentally constrained. Three-dimensional models for studying embryo implantation and early post-implantation development. ( A ) Schematic representation of a reconstructed endometrial scaffold generated by layering a stromal hydrogel within a compartmentalized culture system and overlaying it with organoid-derived epithelial fragments to form a luminal epithelial layer. Separate culture media are used for stromal and epithelial compartments to preserve cell-specific differentiation. This engineered platform recreates stromal-epithelial architecture and supports a hormonally responsive endometrial environment. ( B ) Hormone-driven differentiation and embryo interaction within the reconstructed endometrial model. Upper panels show epithelial remodeling in response to hormonal stimulation. Under estrogen exposure, the epithelium displays a proliferative phenotype with shallow gland-like structures. Subsequent treatment with progesterone and cyclic AMP induces a secretory phenotype, characterized by increased glandular complexity and expression of differentiation markers. Lower panels illustrate embryo or blastoid attachment and invasion, including trophoblast expansion and differentiation into distinct lineages surrounding embryonic compartments. ( C ) Three-dimensional endometrial implantation model within a microfluidic platform. A hormonally primed endometrial tissue is assembled through sequential layering of extracellular matrix, stromal cells, glandular epithelial organoids, and a luminal epithelial layer. This compartmentalized system enables controlled embryo attachment and invasion under dynamic culture conditions. ( D ) Three-dimensional post-implantation co-culture model incorporating epithelial, stromal, immune, and supporting-cell populations. Organoid-derived epithelial structures are hormonally primed to mimic the receptive phase of the endometrium. Embryos are introduced into a defined niche within the structure, allowing modelling of key implantation processes, including attachment, epithelial penetration, and trophoblast differentiation, within a stabilized in vitro microenvironment. cAMP, cyclic adenosine monophosphate; E2, estradiol; GE, glandular epithelial organoids; LE, luminal epithelial cells; mP4, medroxyprogesterone acetate; ST, stromal cells.

Assisted reproductive technologies have resulted in more than 10 million births since 1978 ( Edwards et al. , 1969 , 1970 ; Adamson et al. , 2025 ; Baker et al. , 2025 ). Contemporary culture media incorporate salts, carbohydrates, amino acids, buffers, and protein supplements but lack many critical components and physical properties characteristic of reproductive tract fluids ( Sunde et al. , 2016 ; Zagers et al. , 2025 ). Since the earliest mammalian media (e.g. Earle’s, Bavister’s, Waymouth’s, Whittingham’s), human embryo culture formulations have been progressively refined into modern sequential and single-step systems ( Edwards et al. , 1970 ; Morbeck et al. , 2017 ; Zagers et al. , 2025 ). Yet, no consensus exists on a superior medium, and reported differences in clinical or laboratory outcomes are often modest and inconsistent. Meaningful comparison is further hampered by proprietary compositions, divergent protein supplements, variable oxygen tension, and non-standardized laboratory practices, which together limit mechanistic interpretation ( Sciorio and Rinaudo, 2023 ). Contemporary media are broadly categorized as chemically defined or undefined, depending on their protein source. Defined formulations offer greater reproducibility and quality control, whereas undefined supplements, such as serum or reproductive fluids, introduce inter- and intra-donor variability that complicates standardization. In general, commercial media combine inorganic salts, carbohydrates, amino acids, buffers, and protein supplements, with optional additions such as antibiotics, vitamins, and chelators ( Table 1 ). However, the precise contribution of individual components to embryo competence remains difficult to disentangle, and authors have therefore called for full transparency in media composition and patient cohorts ( Sunde et al. , 2016 ; Tarahomi et al. , 2019 ; Zagers et al. , 2025 ). Even under optimized conditions, these formulations force embryos to adapt to an artificial environment and to activate metabolic and molecular stress responses to maintain development in vitro ( Lee et al. , 2022 ; Zhang et al. , 2025 ). Composition of selected commercial in vitro embryo culture media compared with reference basal media. × indicates the presence of the component in the formulation according to manufacturer descriptions. Commercial media compositions are based on publicly available information and may not reflect proprietary or undisclosed components. Culture protocols are classified as one-step (single medium throughout development) or sequential (stage-specific media). CaCl 2 , calcium chloride; KCl, potassium chloride; KH 2 PO 4 , potassium dihydrogen phosphate; MgSO 4 , magnesium sulfate; Na 2 HPO 4 , disodium hydrogen phosphate; NaCl, sodium chloride; NaH 2 PO 4 , sodium dihydrogen phosphate; NaHCO 3 , sodium bicarbonate; EDTA, ethylenediaminetetraacetic acid; GM-CSF, granulocyte–macrophage colony-stimulating factor.

At present, we lack standardized benchmarks to define what an “optimal” or “physiologically relevant” organoid secretome should look like. Moreover, it remains unclear whether all components of these secretions (e.g. mRNAs, signaling proteins, redox enzymes, and EV-associated factors) retain their biological activity throughout the entire IVC period or whether some undergo degradation or inactivation. Addressing these issues will require parallel phenotypic and molecular characterization of organoids (e.g. transcriptomic markers of differentiation state) alongside quantitative proteomic/EV profiling of their secretions. Knowing which molecules are necessary and sufficient to recapitulate the beneficial effects of native reproductive fluids requires a systematic fractionation and functional testing of individual groups of molecules (proteins, lipids, metabolites, and EV-associated cargo). Following this, establishing a standard protocol and sample origin/use among laboratories, and determining the optimal concentration ranges for secreted factors, are crucial parameters that should be evaluated hereafter. Studies must also be accompanied by new quality control pipelines, including functional assays (such as embryo developmental kinetics, zona pellucida remodeling, or metabolic readouts), which are essential for determining potency rather than simply relying on the presence or absence of factors. A practical intermediate step would be to evaluate these secretomes in well-established model systems, such as mouse or bovine, where developmental endpoints and embryo competence are routinely measured and where ethical and logistical barriers are comparatively lower. Moreover, the recent development of human blastoids and embryo-like structures ( Liu et al. , 2021b ; Kagawa et al. , 2022 ) provides an additional platform to assess how ODS influence early developmental trajectories in a controlled and ethically regulated framework ( Kagawa et al. , 2022 ; Shibata et al. , 2024 ). Addressing these questions will not only improve reproducibility but also clarify how close we are to achieving a physiologically meaningful, translationally relevant in vitro milieu.

High-resolution transcriptomics, proteomics, and metabolomics analyses of human organoid-derived secretions (ODS) now enable systematic mapping of epithelial secretory pathways and direct comparisons to in vivo fluids across the proliferative, secretory and peri-implantation window ( Dong et al. , 2023 ). ODS secreted factors not only provide nutrients, but they also modulate embryo requirements, endometrial receptivity trophoblast invasion, and immunological responses ( Salamonsen et al. , 2013 ; Simintiras et al. , 2021 ). When secretions of human EMOs generated under mid-to-late luteal phase hormone stimulation are applied to macrophages, the cells shift their phenotype toward a decidual-like profile, characterized by reduced pro-inflammatory signaling and increased markers of tissue remodeling and tolerance (e.g. CD209, NRP1) ( Lin et al. , 2023 ). Extending this argument, metabolites from extraorganoid fluid (EOF) and intraorganoid fluid (IOF) of these EMOs are enriched in metabolic and immunomodulatory molecules, such as hypoxanthine, which is needed for nucleotide synthesis during rapid embryonic cell divisions, and spermine, a polyamine that modulates cytokine balance and facilitates decidualization ( Simintiras et al. , 2021 ). Currently, the application of ODS to embryo culture is no longer theoretical. Novel studies employing bovine organoid-derived, cargo-specific oviductal EVs, as well as diestrus-phase endometrial ODS (organoids hormonally programmed with estradiol and medroxyprogesterone acetate), have demonstrated their capacity to maintain/improve the quality of in vitro -produced embryos ( Devkota et al. , 2026 ; Menjivar et al. , 2026 ). Importantly, emerging evidence further indicates that ODS derived from hormonally responsive oviductal organoids can also promote sperm capacitation ( Navarro-Serna et al. , 2026 ), thereby highlighting their broader role in coordinating gamete–embryo interactions. Metabolomic profiling further supports the physiological relevance of these systems. For instance, the detection of metabolites such as isobutyrylcarnitine (normally present in native bovine UF) within bovine endometrial IOF suggests that ODS recapitulates key aspects of the luminal biochemical milieu. Functionally, although not yet equivalent to synthetic oviductal fluid controls, Day 7 blastocysts cultured solely in IOF, without the need for a standard commercial culture medium, exhibit prolonged survival, increased hatching rates, and enhanced trophectoderm specification by day 10 ( Devkota et al. , 2026 ). So far, the results indicate that the intrinsic secretory programs of ODS, particularly when synchronized to relevant hormonal states, can generate a microenvironment that supports embryogenesis more effectively than conventional media.

Despite the importance of the fallopian tube (FT), in-depth investigations have historically been constrained by limited access to primary tissue and the short lifespan of ex vivo epithelial cultures. Fallopian tube organoids (FTOs) overcome these constraints by allowing the formation of polarized epithelia composed of secretory (PAX8 + ) and ciliated acetylated (tubulin + ) cells, interconnected by tight junctions. FTOs are responsive to physiological concentrations of hormones (estradiol and progesterone), inducing the expression of differentiation markers reflective of reproductive phases and exhibiting minimal transcriptional divergence from native tissue ( Kessler et al. , 2015 ). FTOs are also versatile for practical utility, with the apical compartment outperforming commercial media in maintaining human sperm motility and viability for up to 96 h ( Gatimel et al. , 2025 ). This reinforces FTOs as powerful platforms for studying gamete–epithelium interactions, hormone signaling, early fertilization dynamics, and disease modeling in a patient-specific context ( Yucer et al. , 2021 ). Endometrial organoids (EMOs) emulate, at a certain level, the uterine complexity. EMOs express canonical markers of glandular epithelium ( MUC1 , E-CADHERIN , CK7 , and EPCAM ) ( Gnecco et al. , 2023 ) and produce luminal/apical secretions according to hormonal stimulations. Upon exposure to estradiol, EMOs undergo proliferative expansion, upregulate key genes, such as ESR1 , TRH , MCM2–4 , and OLFM4 , and exhibit pseudostratified glandular morphology characteristic of the in vivo proliferative endometrium ( Boretto et al. , 2017 ; Turco et al. , 2017 ). Subsequent progesterone treatment induces a secretory phenotype, marked by glandular folding, ciliation, and increased mucin production ( Fig. 3A ). This is paralleled by the upregulation of genes associated with secretory differentiation and early pregnancy, including PAEP , SPP1 , MUC1 , ALOX15 , AQP3 , and 17βHDS2 ( Boretto et al. , 2017 ; Fitzgerald et al. , 2023 ). Approaches for generating advanced endometrial organoids and multicellular assembloids. ( A ) Scaffold-free three-dimensional organoids generated using agarose micromolds, containing both epithelial and stromal cell populations. The inset illustrates the secretion of structural and mucosal components, including collagen, mucins, and glycoproteins. Immunostaining for cell-type-specific markers, such as vimentin (stromal cells) and E-cadherin (epithelial cells), highlights the spatial organization of the different compartments. ( B ) Multicellular assembloid model comprising a central epithelial compartment surrounded by a stromal cell layer, partially recapitulating epithelial-stromal tissue architecture. ( C ) Apical-out organoid configuration in which the epithelial apical (luminal) surface is oriented outward, allowing direct exposure to the external environment. Insets show ciliated epithelial structures and luminal organization. Cytoskeletal and polarity markers, including basal lamina components, filamentous actin, and acetylated α-tubulin, illustrate epithelial polarity and cellular specialization. ( D ) Apical-out organoid incorporating epithelial, stromal, and endothelial cell populations to model epithelial–stromal–vascular interactions. Cell-type-specific markers confirm the presence and organization of the three cellular compartments. AO, apical-out. It is worth mentioning that the use of 3D advanced cultures aligns with the 3Rs framework for animal studies. These models can reduce the number of animals required for validation and refine experimental design by enabling controlled models of hormone response and cell communication. Additionally, because organoids retain donor-specific transcriptional and epigenetic traits, they offer opportunities to investigate individual variability in receptivity and ART outcomes, something that cannot be captured in traditional animal models.

Organoid technology has revolutionized the study of tissue-specific biology, offering unprecedented opportunities for modeling reproductive tract function in vitro . The field was established by seminal work from Hans Clevers’s laboratory, which demonstrated that adult stem cells can generate self-organizing 3D intestinal structures in vitro that recapitulate key features of the native tissue, ( Sato et al ., 2009 ). This foundational breakthrough paved the way for organoid generation from almost all tissues, from a variety of sources of cell types (embryonic, adult, or differentiated cells). Current advances enabled the derivation of oviductal and uterine organoids from a wide range of species, including humans, mice, cattle, pigs, and other domestic animals ( Bourdon et al. , 2021 ; Thompson et al. , 2023 ). Resembling native epithelia ( Fig. 2A ), these organoids comprise at least two cell types: ciliated and non-ciliated or secretory cells ( Boretto et al. , 2017 ; Turco et al. , 2017 ). In general, these fundamental traits are critically dependent on active Wnt and Notch signaling pathways ( Sato et al. , 2011 ). Accordingly, culture conditions are typically supplemented with growth factors such as epidermal growth factor (EGF), Noggin, and the Wnt agonist R-spondin 1 (RSPO1). Where Wnt signaling (through LGR4/5/6 receptors) maintains epithelial stemness, Notch signaling modulates cell fate, and Noggin prevents premature differentiation ( Sato et al. , 2011 ; Mahe et al. , 2013 ). Generation and structural features of endometrial organoids and assembloids derived from reproductive tissues. ( A ) Schematic representation of the female reproductive tract in human and bovine species, including the ovaries, fallopian tubes (oviducts), uterus, and cervix. The inset highlights the endometrium, showing the luminal epithelial layer and the underlying stromal compartment. ( B ) Overview of the generation of in vitro three-dimensional models. Epithelial cells are isolated and embedded within an extracellular matrix scaffold to form epithelial organoids. Stromal cells can be cultured separately as a monolayer or combined with epithelial cells to generate multicellular epithelial-stromal assembloids that partially recapitulate tissue organization. ( C ) Diagram of epithelial cell polarity within organoids, illustrating apical–basal organization. The apical (luminal-facing) surface is oriented toward the organoid lumen, while the basolateral surface interacts with the surrounding extracellular matrix. ( D ) Representative morphology of epithelial organoids and epithelial-stromal assembloid-like structures grown in extracellular matrix, showing three-dimensional architecture and cellular organization. Female reproductive tract organoids can be generated from fresh, cryopreserved biopsies and even menstrual flow samples, reflecting the donor’s physiological and genetic background ( Kessler et al. , 2015 ; Kopper et al. , 2019 ; Bui et al. , 2020 ; Cindrova-Davies et al. , 2021 ; Rizo et al. , 2025 ). These models can be established from tissue representing different cycle phases, including secretory, proliferative, decidual, and atrophic endometrium. Once established, organoids grow long-term through the expansion of individual cells ( Fig. 2B ) and acquire a ‘conserved’ epithelial polarity (the apical or luminal surface is enclosed (apical-in), whereas the basolateral surface faces the outside and interacts with the extracellular matrix (ECM) (basolateral-out), Fig. 2C and D ) ( Kessler et al. , 2015 ; Boretto et al. , 2017 ; Turco et al. , 2017 ). Importantly, organoids secrete a variety of proteins, lipids, metabolites, and EVs that reflect features of reproductive tract fluid composition, making them attractive candidates for developing next-generation biomimetic culture supplements ( Simintiras et al. , 2021 ; Dong et al. , 2023 ).

OF provides a dynamic milieu shaped by ciliary activity, epithelial secretion, and endocrine fluctuations. Before fertilization, OF undergoes key modifications that facilitate gamete interaction and fertilization, e.g. oviduct-specific glycoprotein 1 (OVGP1) promotes zona pellucida hardening in polyspermy-prone species, such as pigs and goats, thereby reducing fertilization anomalies ( Coy et al. , 2008 ; Bragança et al. , 2021 ). Because preimplantation development naturally spans two distinct environments (the oviduct during cleavage and the uterus during compaction and blastocyst formation), several studies have explored sequential supplementation strategies ( Hamdi et al. , 2018 ; Leal et al. , 2022 ; Pakniyat et al. , 2025 ). IVC of porcine embryos, first in OF (1% v/v) and then in UF (1% v/v), produced blastocysts that more closely resembled their in vivo counterparts in cell number and epigenetic signatures ( Canovas et al. , 2017 ; París-Oller et al. , 2021 ). Follow-up analyses in this porcine model revealed that supplementation with reproductive fluids also mitigated aberrant placental gene expression typically associated with in vitro embryo production. Placentas derived from embryos cultured with reproductive fluids did not show the increased expression of developmentally relevant genes such as PEG3 and LUM , which were otherwise dysregulated in conventional (IVC) conditions, suggesting that fluid-derived factors contribute to early placental programming with downstream consequences for offspring development ( París-Oller et al. , 2021 ). Human reproductive fluids collected under well-established quality control protocols demonstrated bioactivity in a bovine embryo assay, with 1% (v/v) supplementation of culture medium supporting high cleavage rates and blastocyst formation. As a subsequent proof-of-concept in this study, embryos cultured with autologous UF could develop into successful pregnancies in three women, with two of them resulting in healthy live births ( Canha-Gouveia et al. , 2021 ). Despite this evidence, responses to reproductive fluids are species-specific and dose-dependent, as excessive concentrations of reproductive fluids can impair development. This is relevant in cattle embryos, where the use of oviductal and UFs during IVC in more than 5% of cases was negative ( Hamdi et al. , 2018 ). Similarly, exposure of mouse oocytes/cumulus complexes to ovarian endometriotic fluid led to a significant reduction in the proportion of hatching/hatched blastocysts, even when initial fertilization and cleavage rates were not substantially altered ( Piromlertamorn et al. , 2013 ). These findings highlight that defining supplementation merely as a percentage (v/v) may be insufficient, as protein composition and concentration can vary substantially between fluid batches. Future strategies should therefore aim to standardize supplementation based on quantitative characterization of key protein and bioactive component concentrations, enabling tighter control of exposure levels and minimizing the risk of toxicity or developmental perturbations. Overall, reproductive fluids and their bioactive components provide a valuable baseline for designing physiologically relevant culture systems. When incorporated in a controlled and standardized way, these components have the potential to enhance embryo quality and developmental progression, supporting improved outcomes in ART ( Lopes et al. , 2020 ; Heras et al. , 2025a,b ; Serrano-Albal et al. , 2025 ). However, the direct use of biological fluids poses potential biosafety risks, including the transmission of infectious agents, inflammatory mediators, and immunogenic components. Translating approaches that incorporate reproductive fluids as standard supplements for IVC systems would require addressing stringent safety regulations and donor-screening requirements. Moreover, only a limited number of studies have demonstrated the feasibility of implementing such strategies in a controlled and reproducible manner.

Overall, these findings position organoid systems as ‘secretory bioreactors’, laying the conceptual groundwork for next-generation embryo culture strategies. For this transition to occur, researchers should bear in mind that while organoids successfully reproduce epithelial signaling, they lack the stromal, vascular, immune, and endocrine cell interactions present in vivo . As a result, their secretomes, though informative, do not fully capture the complexity or dynamic hormonal composition of natural reproductive fluids. The composition of ODS can also shift over time depending on passage number, media formulation, and the degree of cell differentiation. In addition, ECM scaffolds such as Matrigel (ECM derived from the Engelbreth-Holm-Swarm mouse sarcoma) introduce non-physiological components and limitations for mechanistic and translational studies. All these variables and phenotypic drifts challenge reproducibility and raise questions about whether ODS can be produced with the consistency required for clinical application. Despite the intrinsic variability of ODS, the modularity of organoids remains an attractive proof-of-concept for identifying and characterizing physiological factors under certain in vitro conditions. Looking ahead, ODS opens the possibility to adjust them to specific embryonic stages, physiological states (follicular versus luteal phases), or patient-specific biobanks. Achieving clinical applicability does not require eliminating biological variation; rather, it requires defining acceptable ranges and reproducible manufacturing steps, as is done for other complex biological products (e.g. human serum albumin, platelet lysates, and conditioned media used in cell therapy). In response to some of these limitations, a scaffold-free EMO model was generated using low-adhesion agarose micromoulds to produce apical-out (AO) spheroids with an epithelial perimeter and stromal core ( Fig. 3A ) ( Wiwatpanit et al. , 2020 ). Although originally designed to study polycystic ovary syndrome, its AO architecture provides an accessible luminal interface ( Fig. 3A and C ), offering a unique opportunity to directly expose gametes or embryos to native apical secretions. Similarly, AO-EMOs incorporating stromal and endothelial cells allow secretions from the internal gland-like epithelial cells to diffuse to the exterior ( Fig. 3C ), thereby modulating cell behavior in ways consistent with early pregnancy, as happens when they are co-cultured with human blastocysts or their in vitro counterparts: derived blastoids ( Shibata et al. , 2024 ).

Technical limitations, particularly the difficulty of collecting sufficient volumes of oviductal and UF from individual donors, often necessitate pooling samples, thereby introducing inter-individual variability that may mask specific biological effects. Even in controlled animal models, differences in donor physiology, collection methods, storage conditions, and processing pipelines yield fluids with markedly different biochemical profiles and biological effects. Such variability is incompatible with the strict standardization required for human ART. From a regulatory standpoint, reproductive fluids would likely be classified as complex human-derived biological products, necessitating good manufacturing practice (GMP)-compliant procurement, pathogen screening, sterility testing, endotoxin validation, and full traceability: requirements that currently lack standardized implementation pathways in ART laboratories. Another unresolved key issue is what the authors mean when they state that a certain percentage of “native” or “physiological” reproductive fluid is added as a supplement to the culture media. These additions are often described only in volumetric terms (e.g. 1–10% fluid supplementation), and the biochemical concentration and composition of the supplemented solution remain unknown. Without defining which molecular components are biologically active, at what effective concentrations, and their functional integrity (e.g. mRNA may degrade rapidly under the culture conditions), researchers cannot determine which molecules mediate the observed effect ( Chronopoulou and Harper, 2015 ). Hence, the field would benefit from the establishment of standardized reporting frameworks requiring biochemical characterization, batch qualification criteria, and absolute molecular quantification before clinical translation. A rational translational pathway may involve deconstructing reproductive fluids into defined molecular components and reconstructing synthetic, quality-controlled mimetics that retain biological functionality while meeting regulatory and safety standards. In this sense, the emergence of 3D tissue cultures and systems offers a captivating alternative. As 3D technology continues to evolve, it holds the potential to redefine IVF and IVC systems by providing this biomimetic microenvironment that closely mirrors the in vivo reproductive tract, ultimately improving embryo quality and implantation outcomes in a clinically applicable and ethically sustainable manner.

In vitro embryo development, while foundational to both human ART and livestock breeding programs, continues to fall short of replicating the dynamic and finely tuned environment of the female reproductive tract. Efforts to supplement culture media with reproductive fluids have demonstrated improved embryo quality, developmental competence, and epigenetic stability. However, practical limitations, including donor variability, biosafety concerns, and lack of standardization, have restricted their widespread implementation. ODS, 3D IVC (organoids and assembloids), scaffolds, chips, and co-culture systems now offer a scalable and ethically viable alternative. These models generate bioactive secretomes that closely mimic oviductal and UFs, providing a physiologically relevant and reproducible source of factors essential for early development. In human assisted reproductive technologies, 3D-based platforms hold the potential to personalize embryo culture conditions, reduce reliance on undefined supplements, and improve implantation and long-term developmental outcomes. As the field advances, the integration of organoid technologies, EV biology, and functional co-culture systems represents a transformative shift toward biomimetic embryo culture. These innovations not only bridge the gap between in vivo and in vitro environments but also offer translational benefits supporting safer, more effective, and physiologically informed reproductive strategies in clinical settings.

Replicating the in vivo environment in in vitro embryo culture is by far more than a simple combination of salts, energy substrates, and amino acids ( Fig. 1 ). Modern approaches emphasize the need to adapt the composition of the culture media to the specific metabolic and developmental needs of the embryo at each stage. This includes not only adjusting inorganic elements but also incorporating a wide range of organic compounds (vitamins, lipids, growth factors, and antioxidants), many of which are naturally present in oviductal and uterine fluids (UFs) ( Table 2 ). Analyses of human oviductal fluid (OF) revealed substantial discrepancies in amino acids, energy substrates, and ionic composition when compared with conventional media, supporting the development of oviduct-inspired formulations such as OVIT ( Utsunomiya et al. , 2022 ). Comparison of fertilization and early embryo development in the female reproductive tract and in laboratory culture conditions. The upper panel illustrates fertilization and preimplantation embryo development in vivo within the female reproductive tract. In this physiological context, gametes and embryos are exposed to a dynamic and complex microenvironment composed of follicular, oviductal, and uterine fluids. These fluids contain a wide range of bioactive components, including proteins, metabolites, lipids, hormones, and extracellular vesicles, and are influenced by epithelial secretions, fluid flow, and hormonal changes. Together, these factors regulate fertilization, early embryonic development, and maternal–embryo communication. The lower panel depicts the conventional IVF workflow, in which sperm preparation, fertilization, and embryo culture are performed using static, chemically defined culture media. These systems lack many of the biochemical and physical properties of the in vivo environment. The figure highlights how supplementation with reproductive fluids, or their bioactive components, at specific stages of in vitro culture may improve fertilization outcomes and embryo quality by more closely mimicking physiological conditions. IVP, in vitro embryo production. Concentrations of major osmolytes in human uterine fluid, serum, and oviductal fluid across reproductive phases. Values are reported as ranges or mean ± standard deviation, depending on the original study. Uterine fluid values correspond to different phases of the menstrual cycle (proliferative, mid-cycle, and luteal). Oviductal fluid values correspond to preovulatory and postovulatory stages. Ca, calcium; Cl, chloride; K, potassium; Na, sodium; mOsm/kg, milliosmoles per kilogram. Data sources:  Casslén and Nilsson, (1984) ;  Borland et al . (1980) ;  Lippes et al . (1972) . Supplementation of media for in vitro oocyte maturation, fertilization, and embryo culture may serve as a foundation for next-generation approaches aimed at biomimicking natural reproductive fluids ( Canovas et al. , 2017 ; París-Oller et al. , 2021 ), although their ability to emulate the biochemical and physical conditions of the oviduct and uterus remains limited. In vivo , embryos experience a dynamic microenvironment characterized not only by temporally regulated nutrient concentrations but also by fluid flow, ciliary movement, mucus-rich secretions, and complex interactions with epithelial surfaces ( Coy et al. , 2012 ). Current culture media compensate only partially for these attributes, often lacking immunoregulatory glycoproteins, cytokines, EVs, and the dynamic fluctuations of ions and amino acid profiles that characterize the in vivo uterine environment ( Zagers et al. , 2025 ). Reproductive tract fluids contain thousands of components, including proteins, glycoproteins, lipids, metabolites, EVs, cytokines, miRNAs, and hormones, which contribute to the embryo’s natural microenvironment ( Li et al. , 2023 ; Piibor et al. , 2023 ; Gonella-Diaza et al. , 2024 ; Apostolov et al. , 2025 ). Unsurprisingly, numerous studies have explored the use of reproductive fluids or their derivatives as supplements to the medium as a strategy to improve biomimicry in vitro ( Canovas et al. , 2017 ; París-Oller et al. , 2021 ). Studies in livestock demonstrate that reproductive fluids can influence fertilization, cleavage, gene expression, metabolic activity, placental vascularization, and pregnancy rates, although outcomes vary markedly across species, experimental conditions, and fluid origin ( París-Oller et al. , 2022 ; Párraga-Ros et al. , 2023 ; Heras et al. , 2025b ). These inconsistencies highlight both the promise and the complexity of using native physiological secretions to support embryo development.

Future studies should integrate more physiologically relevant communication between epithelial, stromal, and vascular compartments into reproductive assembloids. In this way, next-generation secretomes would more faithfully reflect in vivo uterine and oviductal environments. Advances such as apical-out endometrial organoids incorporating stromal and endothelial components ( Fig. 3C and D ) provide a blueprint for engineering reproductive assembloids in which endothelial networks contribute angiocrine factors essential for receptivity, decidual remodelling, and early embryo support. Looking ahead, the incorporation of immune components may further enhance the physiological relevance of these models. Additionally, the incorporation of micropatterned hydrogel scaffolds or hybrid microfluidic platforms with cell culture inserts could further refine secretion dynamics by recreating tissue-specific architecture, such as glandular invaginations in the endometrium or the folded ciliated surface of the fallopian tube. These features are analogous to villus-crypt-like interfaces that enhance intestinal maturation ( Shin and Kim, 2022 ). Such patterned systems would enable spatially organized secretory activity, including hormone-mediated signaling and epithelial–stromal–vascular interactions, as well as embryo or trophoblast engagement under physiologically relevant mechanical and biochemical cues.