Organoids are three-dimensional cellular models that can be used for fundamental research and clinical regenerative medicine purposes. By using organoids as an experimental model, the need for animal models can be reduced (Kim et al., 2020). In addition, clinical applications of organoids can reduce the need for donor organs, for example in metabolic liver diseases (Schneemann et al., 2020; Tsuchida et al., 2020).To reproducibly expand organoids for both experimental as well as clinical purposes, defined cell culture media are necessary. For most organoid culture media, Dulbecco’s Modified Eagle’s Medium (DMEM) is used as a basis. DMEM contains amino acids, vitamins, glucose, sodium pyruvate and a buffer system to maintain a physiological pH. The growth factor Wnt3A, most often added in the form of Wnt3A-conditioned medium, is a necessary supplement for expansion media to induce proliferation and organoid outgrowth. Wnt3A-conditioned medium is obtained by culturing a cell line that recombinantly expresses Wnt3A. To facilitate release and stabilization of recombinantly produced Wnt3A, fetal calf serum (FCS) is added to the cell culture medium (Pleguezuelos‐Manzano et al., 2020; VanDussen et al., 2019; Willert et al., 2003).

The use of FCS in cell culture protocols however goes accompanied by experimental, clinical and ethical concerns. The production of FCS is characterized by large inter-batch variations, and the undefined composition of the complex biological fluid negatively impacts the reproducibility of experimental results. Furthermore, exposure of cells in culture to FCS induces a proliferative phenotype which may be non-physiological (Timmerman et al., 2022). In addition, FCS has been suggested to induce cellular senescence in pancreatic organoids, thereby inhibiting long term culture (Mazzella et al., 2023; Seino et al., 2018; Rauch et al., 2011). Growth obstruction is especially detrimental for protocols that require organoid outgrowth from single cells (e.g. CRISPR-Cas9-mediated genome engineering) (Drost et al., 2015).

The use of FCS also poses risks for the clinical translatability (Schneemann et al., 2020; Tsuchida et al., 2020). Regulatory authorities promote switching to FCS-free media for organoids grown for clinical purposes, because of the potential presence of infectious agents and of the risk of inducing xeno-immunization in patients (Chimenti et al., 2014; Mackensen et al., 2000; Selvaggi et al., 1997; Bauman et al., 2018). Untreated FCS therefore does not comply with European Good Manufacturing Practices (GMP), and even gamma-irradiation does not eliminate all clinical risks (Chimenti et al., 2014; Giannitti et al., 2022). Xeno-immunization against FCS proteins has been demonstrated in clinical trials, causing anaphylaxis and fever reactions in patients (Mackensen et al., 2000; Selvaggi et al., 1997). Therefore xeno-free chemically defined media are encouraged in research fields focusing on developing clinical cellular therapies, like the multipotent mesenchymal stromal cells (MSCs) and the human embryonic stem cells (hESC) fields, and more basal organoid research will likely follow suit (Lee et al., 2022; Palamà et al., 2023).

Finally, the production of FCS also raises ethical concerns (Barosova et al., 2021). FCS is collected via cardiac puncture at a stage when the calf fetus can already feel pain and distress (Bauman et al., 2018). Approximately 3 calves are needed to obtain 1 liter of FCS, which results in the killing of around 2 million fetal calves yearly (Bauman et al., 2018).

The reasons to stop using FCS are thus numerous, but replacing FCS to produce Wnt3A-conditioned medium remains challenging as FCS is necessary for the solubilization of Wnt3A protein, and greatly enhances protein stability. Although FCS-free Wnt activation approaches have been studied for organoid expansion, a comparative review on the different strategies has thus far been lacking. This review provides such an overview, taking different experimental goals into account.

Wnt3A is such a popular organoid medium supplement because it can bind to all FZD receptors to induce canonical Wnt signaling, as opposed to other -more selective- Wnt ligands (Voloshanenko et al., 2017). Active canonical Wnt signaling is characterized by inhibition of the β-catenin destruction complex. This results in more stable cytosolic β-catenin translocating to the nucleus (Niehrs, 2012). Figure 1 illustrates the Wnt3A-induced activation of the canonical Wnt pathway (Figure 1A), as well as the further below-described FCS-free strategies to achieve a similar effect (Figure 1B).

There are several possibilities to obtain serum-free Wnt3A. The most straightforward one is to purify recombinant Wnt3A protein by using Cu2+-charged Sepharose beads (Witkowski et al., 2015). This is however expensive, difficult to scale up, and requires maintaining a recombinant cell line culture (Witkowski et al., 2015; Wilson et al., 2020). Alternatively, commercial purified recombinant Wnt3A can be used. As per product information available via the manufacturers’ website or via the product spec sheet, some manufacturers grow these cell lines using FCS, while others disclose using serum-free defined media, resulting in GMP-grade proteins that are suitable for clinical applications as well.

However, the addition of recombinant, serum-free, Wnt3A to expansion medium is not a viable option because of problems with protein solubility and stability. The hydrophobic nature of Wnt3A protein causes aggregation in aqueous solutions in the absence of serum or detergent (Willert et al., 2003; Willert, 2008). In addition, macro biomolecules that are present in FCS, like heparan sulfate proteoglycans (HPSGs), stabilize and preserve Wnt3A activity (Fuerer et al., 2010). Supplementation of Wnt3A to FCS-free DMEM medium resulted in an instant decrease of Wnt3A signaling by 3-fold, and further incubation resulted in a complete loss of signaling (Fuerer et al., 2010). By contrast, Wnt3A signaling activity was only marginally reduced after 6 h of incubation in medium supplemented with 10% FCS, due to HPSGs stabilization (Fuerer et al., 2010).

To expand organoids, Wnt3A is an essential growth factor which is often produced in media that contain FCS. FCS helps to solubilize and stabilize Wnt3A, but the use of FCS goes accompanied by ethical and scientific concerns. It has however proven difficult to replace the use of FCS in the production process of Wnt3A, since supportive cell lines need FCS for survival and for the release of Wnt3A. In this review we explored the use of Wnt mimetics, Wnt carriers and small molecules as animal-free replacements for FCS, taking into account effectiveness, specificity, cost and the type(s) of organoids used for testing (Table 1).

Wnt mimetics can support organoid outgrowth with similar or better efficacy than conventional Wnt3A-conditioned medium. The Wnt mimetic platform allows for effective and precise activation of Wnt pathways, because multivalency and simultaneous binding of multiple receptors can result in more potent pathway activation than is achieved by natural Wnt3A (Chen et al., 2020; Tao et al., 2019; Post et al., 2023). The potency of Wnt mimetics to support organoid outgrowth has been confirmed for many different organoid types (Janda et al., 2017; Chen et al., 2020; Post et al., 2023; Miao et al., 2020), although differential activity for some organoid types was observed (Janda et al., 2017). This may be due to differential requirements for FZD triggering between different organoids or to unknown factors that may be present in the Wnt3A-conditioned medium (Janda et al., 2017). The broad application combined with the potency to support organoid outgrowth for singe cells make Wnt mimetics a good choice for high-throughput organoid workflows, such as CRISPR-Cas9-mediated genome editing (Drost et al., 2015) and for the establishment of organoid biobanks (van de Wetering et al., 2015). However, out of the three FCS replacement approaches described here, Wnt mimetics are the most expensive (Table 1). When acquired commercially they are expensive, and also when produced in-house the initial investments in time and money are high because of the need to establish and maintain a recombinant cell line combined with the use of expensive antibody purification resins (Post et al., 2023).

The use of carriers to solubilize and stabilize Wnt3A proteins can be considered as a more economical option, especially in combination with in-house produced recombinant Wnt3A (Table 1). Although there are initial investments in money and time associated with the purchase or with the establishment of a recombinant cell line and protein purification columns, further maintenance is relatively cheap. An additional advantage is that the use of carrier-stabilized recombinant Wnt3A closely mimics the in vivo situation Whereas the use of lipids as carriers for Wnt3A has been described to enable activation of canonical Wnt signaling and maintenance of duodenum organoids (Tüysüz et al., 2017), reports on the use of afamin are equivocal. Technical issues may underlie these differences, and more research should be performed to be able to conclusively decide on the suitability of either carrier for specific applications (Tüysüz et al., 2017; Mihara et al., 2016).

The small molecule inhibitor CHIR represents the least precise way to induce Wnt signaling in organoids. On the other hand, it is by far the least expensive alternative to the use of Wnt3A-conditioned medium (Table 1) since it does not require the purchase or maintenance of recombinant cell lines and associated purification steps. CHIR inhibits GSK3, which is also involved in signaling by a variety of other pathways. The use of CHIR has been proven effective for certain types of organoids only (Tüysüz et al., 2017; Konagaya and Iwata, 2019; Wang et al., 2020), and can be considered for liver and pancreatic organoids in particular. The molecular mechanism(s) behind this selectivity are currently unknown and whether this is e.g. correlated to a differential dependency of liver and pancreatic organoids on events downstream of GSK3 signaling as compared to other organoid types remains to be established. Even when CHIR is used for the expansion of liver and pancreatic organoids specific attention should be paid to potential effects on other GSK3-dependent signaling pathways (Jope and Johnson, 2004).

Studies that have directly compared the suitability of different Wnt activation methods for the expansion of different types of organoids are currently lacking, rendering conclusive statements impossible. Such studies could assess the potential to support organoid outgrowth from single cells, the efficacy in % organoid outgrowth as well as expansion speed related to Wnt3A-conditioned medium as a reference. Preferably, to ensure replicability results should be compared to the same batch of Wnt3A-conditoned medium. By summarizing the data available, we hope to stimulate the discussion and the development of efficient testing strategies for the replacement of FCS from organoid expansion media.

Although it will require an initial investment to find the best FCS replacement strategy for specific purposes, there are important benefits to be gained. The use of a defined FCS-free expansion medium will allow optimization of organoid yield and interlaboratory replicability. In addition, when gene-edited organoids are to be transplanted back into the patient, for example to ameliorate metabolic liver diseases, it is important that expansion and culture media are GMP-compliant and xeno-free to avoid complications (Schneemann et al., 2020; Tsuchida et al., 2020; Mackensen et al., 2000; Selvaggi et al., 1997; Giannitti et al., 2022; Macy et al., 1989). As mentioned before, complications like anaphylaxis and fever reactions after cell therapy treatment have been reported due to xeno-immune responses directed against FCS-derived proteins (Mackensen et al., 2000; Selvaggi et al., 1997). Currently three clinical trials have been registered that involve organoids for transplantation or cell based therapy. These studies investigate the regenerative potential of intestinal organoids for ulcerative colitis treatment (clinical trial UMIN000030117, Japanese Registry), islet organoids for diabetes treatment (clinical trial NCT06415643) and salivary gland organoids for the treatment of radiation-induced xerostomia (clinical trial NCT04593589). Although it is not disclosed which organoid expansion medium were used in these studies, the scientists of the ulcerative colitis study specifically mention the development of clinical-grade culture methods as one of their challenges for safe in-human studies (Takebe et al., 2018; Shimizu et al., 2019). Finally, more efficient organoid outgrowth will also help with culturing gene-edited organoids from single cells, which has a very low success rate when the conventional Wnt3A-conditioned medium is used. Optimization of the culture of gene-edited organoids in particular has great potential to reduce the need for animal experiments and animal-derived products in e.g. drug testing, drug discovery, personalized medicine and disease modeling.