In the past 2 decades, organoids have greatly expanded our ability to experimentally model gastrointestinal (GI) physiology and pathology. Early studies of two-dimensional cell cultures provided a simple and cost-effective way to study fundamental biology of cells in the GI system and to test phenotypic responses to altered genetics or extrinsic stimuli. However, many GI cells can quickly lose function when cultured in 2D (Elaut et al., 2025; Houbracken et al., 2011). In addition, 2D cultures do not recapitulate the structure and physiology of the native tissues or their microenvironment, leading to an incomplete understanding of the cells and organs in this system. Mouse models, by nature, incorporate this complexity and have been instrumental in advancing understanding of GI organ development, normal homeostasis, and disease. The ability to drive spontaneous pathologies through genetic engineering or inducing injury or inflammation has provided in vivo platforms for understanding disease mechanisms and testing therapeutics (Baydi et al., 2021; Klauss et al., 2018; Hayakawa et al., 2013). Mouse models, however, can take a long time to generate, are low throughput, and fail to recapitulate the complexity and heterogeneity found in the patient population at large.

Organoids begin to bridge the gap between cell culture and animal models by mimicking key functional, structural, and biological complexities of organs. Organoid cultures are derived from adult stem cells (ASC) or pluripotent stem cells (PSC) that are grown in a 3D matrix with specific factors that provide mechano-chemical cues. These signals support self-organization of complex structures (Sato et al., 2009; Clevers, 2016; McCracken et al., 2011). The development of organoid cultures has enabled long-term growth of cells from many tissues of the GI system including the intestines, colon, stomach, pancreas, liver, esophagus, and salivary glands (Sato et al., 2009; DeWard et al., 2014; Yoon et al., 2022; Bartfeld et al., 2015; Crespo et al., 2017; Guan et al., 2017; Huang et al., 2021). The function of these models has been demonstrated in normal and disease conditions. For example, heterogeneous salivary gland organoids were shown to display swelling morphology and calcium influx following treatment with neurotransmitters (Yoon et al., 2022) and colon organoids respond to spatiotemporal induction of tumor-associated genetic alterations with increased proliferation and plasticity as they undergo tumorigenesis (Lorenzo-Martin et al., 2024). Moreover, multiple studies comparing global gene expression have now shown organoids to mimic transcriptional profiles of native tissues (Yoon et al., 2022; Paul et al., 2025; Raghavan et al., 2021; Peng et al., 2018; Fujii et al., 2018; Cherubini et al., 2024; Mead et al., 2018).

By mimicking the critical components and function of complex gastrointestinal organs in an in vitro system, organoids provide a robust experimental platform to investigate mechanisms of development and disease. Organoids allow for rapid manipulation to mimic genetic alterations or extrinsic stress (Huang et al., 2021; Lorenzo-Martin et al., 2024; Li et al., 2014; Mitrofanova et al., 2024; Farin et al., 2014). In addition, they are amenable to high throughput applications such as drug screening (Du et al., 2020; Mead et al., 2022; Hirt et al., 2022). Finally, they enable modeling patient-to patient heterogeneity and can maintain patient specificity in terms of treatment response, allowing them to serve as models for personalized medicine approaches (Papargyriou et al., 2024; Ooft et al., 2019; Zhao et al., 2024; Demyan et al., 2022; Vlachogiannis et al., 2018). The current field of organoids is quickly evolving, with organoids being incorporated into more complex platforms including co-cultures with stromal and immune components (Saheli et al., 2018; Min et al., 2020; Recaldin et al., 2024), organ-on-a-chip devices (Ballerini et al., 2025; Skardal et al., 2015; Zhang et al., 2017) and bioprinted models (Bernal et al., 2022; Wang et al., 2024). As a new generation of models is further developed and widely adopted, a continued consideration of the variety of mechano-chemical signals provided by culture conditions and how they impact phenotype and function is important.

Since the development of the first organoids, several methods for culturing GI organoids have been tested. The nutrients, growth factors, matrix components, and physical cues, such as stiffness, that cells experience influence their signaling pathway usage and phenotype at baseline. They also impact how cells respond to additional perturbations or stress (see Figure 1). Optimizing culture conditions to recapitulate innate biology of tissues can help ensure organoid phenotypes mimic in vivo cellular response to perturbations while reductionist models can uncover mechanistic insights into cell and tissue function. In this mini review, we will outline many matrix and media compositions that have been used for culturing organoids of the gastrointestinal system. We will highlight the progress that has been made thus far in creating physiologically relevant environments for organoid culture and will discuss future directions in this field.

The extracellular matrix (ECM) is a crucial component of the tissue microenvironment. Its composition, organization, and mechanical properties affect cell behavior through chemical signaling and mechanotransduction. In experimental systems such as organoids, the matrix can be chosen to either recapitulate the normal environment or to recapitulate specific components. While mimicking the native matrix gives higher fidelity to the in vivo response of organoids (Lee et al., 2021), as of now, there are no perfect matrices that fully recapitulate the native environments of gastrointestinal organs. Reductionist models, therefore, serve as powerful tools for understanding the impact of individual elements, which can then be combined to develop an idealized model.

Several types of matrices have been produced and tested for gastrointestinal organoid culture, including basement membrane extract (BME)-based hydrogels, decellularized ECM hydrogels, defined natural protein hydrogels, hydrogels composed of recombinant proteins and peptides, and synthetic polymer hydrogels (Kozlowski et al., 2021). Each of these matrices has advantages and disadvantages, depending on the experimental design. The material source, composition, cost, availability, ease of use, and variability all need to be considered. Many types of matrices allow for some degree of control over elastic and viscoelastic properties, degradability, and composition (Table 1). These controllable matrix properties, through activation of cell surface receptors such as integrins, can influence downstream signaling pathways. Ultimately, these signals impact epigenetic and transcriptional programs and control organoid phenotypes and cellular behavior (Below et al., 2022; Saraswathibhatla et al., 2023). For example, matrix stiffness leads to integrin activation and focal adhesion assembly. This drives activation and nuclear translocation of YAP/TAZ, which mediate transcriptional responses to the mechanical cues (Dasgupta and McCollum, 2019; Jafarinia et al., 2024; Panciera et al., 2017). In vivo, gastrointestinal cells in physiological and pathological states experience a wide range of chemical and physical signals from their native ECM. The mechanisms by which distinct components or properties of the matrix impact cellular behavior are incompletely understood. In this section, we discuss recent developments in several types of matrices that impact gastrointestinal organoid culture efficiency or phenotype.

As discussed above, the matrix composition and mechanical properties play key roles in the formation and organization of organoids and can be altered to drive distinct cell states or to recapitulate dynamic environments. Historically, media was optimized to maintain and propagate tissue cultures, but media composition can mimic developmental signals and niche ligands to serve as a powerful tool to impact cell states and behaviors. In this section, we discuss how media components have been optimized to drive development and differentiation of heterogeneous cultures as well as how they contribute to selection and drug-responses of patient-derived organoids.

Basic nutrients including amino acids, vitamins, salts, metabolites, and proteins are provided to organoid cultures with a base media, often Advanced DMEM/F12. Supplementation with additional factors including antioxidants, such as N-acetylcysteine, vitamins, such as nicotinamide, hormones, such as Gastrin, and a variety of growth factors and/or signaling pathway inhibitors supports the growth of a range of organoid types. For reproducibility, as well as for potential clinical translation, there has been effort toward ensuring that components added to the media are well defined. Serum, for example, can be replaced with defined factors to promote organoid growth including the Wnt agonist R-spondin-1 (RSPO1), epidermal growth factor (EGF), and Noggin (Sato et al., 2009). Defined medias are considerably more expensive than some alternatives including conditioned media from growth factor producing cells. However, leftover serum and other undefined factors in conditioned media have the potential to introduce lot-to-lot variability affecting experimental outcomes. A recent assessment of conditioned media from the commonly used L-WRN producer cell line across five laboratories at three institutions found highly replicable Wnt3a, R-spondin3, and Noggin growth factor activity across all groups (VanDussen et al., 2019). In general, however, similar to control of the matrix components discussed above, reproducibility of organoid experiments can be enhanced with defined media components.

Signaling pathway activators and inhibitors that are critical for the development, differentiation, or maintenance of organ-specific cell types can be added to organoid cultures to drive PSCs or ASCs toward mature fates (Clevers, 2016; Lancaster and Knoblich, 2014). In early organoid work, Sato et al. showed that intestinal ASCs required RSPO1, EGF, and Noggin for proliferation and successful passaging as organoids (Sato et al., 2009). Spence et al. drove human PSCs toward hindgut cell fates with Activin A, Wnt, and fibroblast growth factor (FGF) treatments, and then cultured them in Matrigel with RSPO1, EGF, and Noggin to develop intestinal organoids (Spence et al., 2011). Huch et al. cultured liver stem cells with RSPO1, EGF, FGF10, hepatocyte growth factor (HGF), and nicotinamide to develop liver progenitor cultures. Additional inhibition of Notch and transforming growth factor beta (TGFβ) signaling drove expression of mature hepatocyte markers and some hepatocyte function (Huch et al., 2013a). In similar approaches, recapitulating developmental signaling pathways or niche ligands has allowed development of distinct cell fates or the balanced differentiation of heterogenous cultures. Huang et al. used multiple signaling ligands and inhibitors to induce pancreatic progenitors to differentiate to either acinar or ductal fates. They further built on this system, showing lineage-specific responses to the expression of distinct oncogenes GNAS^R201C and KRAS^G12D as well as differential transcriptional and morphological responses to TGFβ signaling in KRAS^G12D-expressing acinar versus ductal organoids (Huang et al., 2021). Yoon et al. used a variety of growth factors and inhibitors to refine culture conditions for salivary organoids to promote long-term growth as well as phenotypic and functional heterogeneity of salivary glands (Yoon et al., 2022). Yang et al. optimized niche signals to drive a balance in intestinal stem cell proliferation and differentiation in their organoid cultures and showed that manipulation of specific signaling pathways could shift that balance and drive specific intestinal fates (Yang et al., 2025). Finally, Mead et al. used high throughput screening on a miniaturized organoid platform to identify new regulators of intestinal stem cell differentiation (Mead et al., 2022).

As organoid research has expanded to model patient-derived tissues from specific pathologies, including cancer, there has been recognition that specific media components contribute to the efficiency of organoid generation from a heterogeneous patient population and to the selection of subpopulations of patient-derived cells. Seino et al. developed a library of 39 patient-derived organoids from pancreatic cancer and identified three functional subtypes of these organoids based on their dependence on Wnt and R-spondin in the media. They also altered the media composition with EGF depletion, Noggin depletion, or Nutlin3 addition to intentionally select for pancreatic cancer organoids expressing mutant KRAS, mutant SMAD4, or mutant p53, respectively (Seino et al., 2018). Hawkins et al. similarly observed different patient-derived pancreatic cancer organoid lines required distinct concentrations of Wnt. They further showed that this information may be useful in designing combination treatment strategies (Hawkins et al., 2024).

Patient-derived organoids provide an opportunity not only to study the biology underlying tumorigenic phenotypes, but also to test patient-specific responses to therapeutics. Tiriac et al., for example, developed a biobank of 66 patient-derived pancreatic cancer organoids and demonstrated that the drug sensitivity profiles of the organoids reflected the patient response to therapy (Tiriac et al., 2018). Huang et al. leveraged patient-derived xenografts (PDX) of pancreatic cancer to compare the therapeutic sensitivity of the PDX models to that of PDX-derived organoids. They demonstrated that, when grown in Wnt-free media, the organoids retained the tumor differentiation status and histological heterogeneity and were concordant with the PDXs for drug sensitivity (Huang et al., 2020). Hogenson et al. evaluated patient-derived organoids from multiple gastrointestinal cancers for their ability to predict clinical response. They also found that Wnt in the organoid media impacted transcriptional signatures and drug sensitivity, with the media lacking Wnt giving better concordance with patient responses (Hogenson et al., 2022). These studies indicate that factors in the media can critically affect the interpretation of organoid response to specific stressors. Raghavan et al. analyzed matched pancreatic cancer patient samples and organoid models to investigate how microenvironmental signals present at distinct sites or associated with distinct molecular subtypes could control organoid state or drug response. They found that organoid cultures supplemented with TGFβ exhibited plasticity toward a basal-like phenotype and had therapeutic responses distinct from organoids with classical gene expression. In addition, treatment of the organoids with IFNγ to model signaling from CD8⁺ T cells, was found to increase IFN response gene signatures and drive an intermediate state expressing both basal and classical programs (Raghavan et al., 2021).

Together, the properties of the matrix and media in which organoids are cultured can alter the efficiency of organoid generation, affect signaling pathways leading to phenotypic changes, and alter the responses to perturbation. Many recent studies of patient-derived tumor organoids have focused on the presence of Wnt ligands in the media, finding Wnt signaling to be a major player in driving organoid selection, heterogeneity and drug sensitivity (Huang et al., 2020; Hogenson et al., 2022). TGFβ was found to alter organoid subtype, which also impacted drug sensitivity (Raghavan et al., 2021). Future work using high throughput screens to test responses to exogenous ligands, inhibitors, or drugs that impact specific signaling pathways (Mead et al., 2022; Hirt et al., 2022) can be leveraged to continue to uncover mechanisms driving distinct organoid phenotypes or vulnerabilities. As integrins and other receptors regulating mechanotransduction cooperate with many growth factor receptors, including those in the Wnt pathway (Crampton et al., 2009; Tejeda-Muñoz et al., 2022; Cooper and Giancotti, 2019; Sarker et al., 2020; Li et al., 2023), it will also be of interest in future work to study these signals in combination. Through adaptation of the media along with the matrix, organoid studies can address how specific signaling pathways converge to contribute to lineage commitment and progression, function, and response to perturbations.

The matrix and media can be chosen to recapitulate the native environment or can be designed in a reductionist approach to understand specific aspects of cellular responses to the environment. Matrices that support organoid growth, for example, can range from inert gels with a single adhesive site to the complex decellularized ECM from the tissue of interest. Proteomics analyses on ECM-derived matrices has been used to compare hydrogels to native matrices (Giobbe et al., 2019). Proteomics studies have also enabled the design of manipulable gels in which specific components of the native environment can be modulated to define which matrix proteins or properties control cell fates (Below et al., 2022). Extending these types of studies to additional tissues or disease states will be of interest. Similarly, large scale atlasing of transcriptional profiles in normal and tumor tissues is providing unprecedented information about the ligands, receptors, and signaling pathways that control cell-cell communication in niche environments (Yanai et al., 2024). As these studies suggest additional pathways to control through changes in the media or matrix, single cell transcriptomic approaches can be used to compare organoid cell composition and cell state to that found in normal tissues to validate these efforts (Yoon et al., 2022; Fujii et al., 2018; Cherubini et al., 2024; Mead et al., 2022).

Future work incorporating GI organoids into increasingly complex model platforms will increase understanding of organ development, homeostatic function, and disease. In addition, leveraging patient-derived GI organoids in precision medicine approaches may provide routes for clinical translation. Throughout these studies, systematic perturbation of the matrix and media components will allow for continued interrogation of mechanical and chemical signals regulating gastrointestinal physiology and pathology.