Shaping Tomorrow's Liver Organoids: A Journey Toward Integrating Bile Ducts

Liver tissue engineering has undergone remarkable developments since the late 20th century, transitioning from simple two‐dimensional cultures to sophisticated three‐dimensional organoid models for drug toxicity assessments. Stem cell innovations have enabled the creation of liver organoids for disease modelling and tissue engineering. However, a key limitation is the absence of functional bile ducts in these organoids, crucial for replicating bile‐duct related diseases. Bile, synthesized by hepatocytes, plays a vital role in digesting fats and expelling lipid‐soluble wastes, including drug byproducts. Diseases impeding bile flow are responsible for many liver transplants and can cause severe conditions such as liver cirrhosis, causing over 50,000 annual deaths in the US. Current liver organoids, while bile‐producing, are devoid of bile ducts, limiting their efficacy in mimicking diseases related to bile flow. This article underscores the pressing need to incorporate bile ducts in engineered liver tissues, delves into the challenges faced in this effort, and highlights potential solutions through biomaterial and bioengineering techniques. Such advancements will offer researchers enhanced insights into bile duct disorders and pave the way for exploring innovative therapeutic strategies.


Introduction
Liver tissue engineering has evolved significantly since its inception in the late 20th century.Early in vitro systems, focusing on 2D cultures of primary hepatocytes and cell lines, paved the way for more complex, 3D organoid models to evaluate drug toxicity.Advances in stem cell biology facilitated the development of stem-cell derived liver organoids, composed of hepatocytes, hepatic stellate cells, and other liver constituents, with applications spanning disease modeling, toxicology studies, and drug development.Thus far, most efforts have been dedicated to improving the maturation of stem-cell derived hepatocytes and integrating vasculature in liver organoids.However, a key missing piece of the puzzle is the creation of bile ducts, which is critical to modeling liver bile-duct diseases.Hepatocytes produce bile, which is secreted into bile canaliculi -tiny channels between the apical DOI: 10.1002/adbi.202300450membranes of adjacent hepatocytes.Bile canaliculi merge with the liver bile duct network, enabling bile transport from hepatocytes to bile ducts.This bile is eventually stored in the gallbladder.Bile fluid aids in the digestion of fats and serves as a primary route for the excretion of lipid-soluble waste metabolites including bilirubin (a breakdown product of hemoglobin, the iron-rich pigment in red blood cells).In the case of drugs, certain compounds are metabolized in the liver and readily excreted in bile.Thus, the bile flow system is essential for both digestion and detoxification.Diseases that disrupt bile flow cause nearly 33% of adult and 70% of pediatric liver transplantations, and often result in liver cirrhosis, affecting millions of Americans, and causing > 50 000 deaths in the US annually. [1]nfortunately, although current liver organoids may produce bile, they lack bile ducts.Thus, current systems are inefficient at modeling liver diseases involving bile flow complications.In this perspective, I argue the importance of integrating bile ducts, discuss challenges in achieving this unmet need, and outline the bioengineering strategies that can potentially help us achieve this goal.

Importance of Integrating Bile Ducts in Liver Organoids
Bile-duct diseases such as primary biliary cirrhosis, primary sclerosing cholangitis, biliary atresia, and Alagille syndrome, cause damage to bile ducts, which result in liver damage due to the accumulation of bile acids in the liver, commonly referred to as liver cholestasis. [2]To model these diseases in vitro, we must create liver organoids with three key elements: 1) Functional hepatocytes that contain metabolism enzymes (e.g., cytochrome P450), 2) Bile ducts that are connected with bile canaliculi to enable biletransport from hepatocytes to bile duct channels, and 3) an outlet allowing us to collect and study the flow and composition of bile and disease markers therein.In current liver organoids, adjacent hepatocytes create a network of tiny channels that function as bile canaliculi.But these canaliculi are merely the luminal spaces between neighboring hepatocytes where the bile acids accumulate because there are no channels in which it can flow.The excessive accumulation of bile within the organoid may cause hepatocyte injury, and reduce hepatocyte functions. [2]Moreover, it is extremely challenging to collect bile acids from within an organoid in order to evaluate bile properties.Thus, the current hepatic 3D culture models, consisting of random, inaccessible bile reservoirs are inefficient to model bile duct diseases.Conversely, 3D culture of cholangiocytes (bile-duct forming cells) may contain bile-duct mimicking structures, but they lack hepatocytes to perform metabolism.Thus, the key role of bile ducts in liver function highlights a critical need to develop complex liver organoids with built-in bile-ducts in order to evaluate bile duct diseases.

Signaling Pathways Involved in Creating Bile Ducts During Development, and Challenges in Applying Them In Vitro
It is critical to understand how bile-ducts form during liver development.Fortunately, liver development studies have determined the signaling pathways that are key to the differentiation of cholangiocytes and their morphogenesis into bile ducts.Both hepatocytes and cholangiocytes are differentiated from bipotent liver progenitor cells known as hepatoblasts.Differentiation media containing various biochemical signals such as epidermal growth factor (EGF), interleukin 6 (IL-6), sodium taurocholate, Jagged1, and transforming growth factor beta (TGF-) have been shown to push the hepatoblasts towards cholangiocyte fate, expressing mature markers (e.g., SOX9, OPN, CK7, CK19, and CFTR).Among these biochemical signals, Jagged1-induced Notch signaling is widely recognized to be the key signaling pathway to induce cholangiocyte fate (Figure 1).During liver development, the portal vein smooth muscle cells express Jagged1 (Notch signaling ligand), which interacts with the Notch2 receptors in adjacent hepatoblasts to activate Notch signaling, eventually differentiating them to cholangiocytes. [3]Thus, Notch signaling determines where in the liver cholangiocytes will differentiate, i.e., in regions around the portal veins.In vitro studies indicate that inhibition of Notch signaling prevents cholangiocyte growth, further confirming the critical role of Notch in bile duct formation. [4]Conversely, activation of Notch signaling in hepatocytes induces fibrosis in a mouse model of nonalcoholic steatohepatitis (NASH), [5] and converted hepatocytes into malignant cholangiocarcinoma cells. [6]These studies emphasize the need to provide targeted, cell-type-specific Notch signaling in 3D cocultures to maintain maturation of both hepatocytes and cholangiocytes, something lacking in current 3D co-culture systems.
One way to achieve spatial control of Notch signaling is by embedding Notch ligands in specific locations within the cell cul- a) Jagged1 (Jag1) were caged using photo-cleavable linker to inactivate ligand-receptor interaction.Upon UV exposure, the cages can be removed to activate the Jag1 ligands in selected regions of the cell culture matrix using photomask.b) Spatially controlled Notch activation in selected region of the cell culture substrate as verified using fluorescence of Notch2 reporter (green) reporter cells.Adapted with permission. [9]Copyright 2021, Wiley-VCH.ture microenvironment.For example, embedding of DLL4 ligands in specific locations of a matrix lead to activated Notch signaling in those selected regions. [7]Another method is to use light-responsive chemistry.For example, Jagged1 can be modified with light-responsive functional groups (e.g., photocaged thiols) so that it reacts with the cell-culture matrix in the presence of light.In this way, controlled light exposure can allow immobilization of Jagged1 in the desired regions of the matrix.Photobased patterning of ligands in a cell culture matrix is a common approach.Yet, the photo-patterning process has a limitation in 3D culture because it may briefly expose all encapsulated cells to soluble Notch ligands during the immobilization step, [7] which can inhibit Notch signaling in some cells, instead of activating it, due to soluble Notch ligands acting as decoys. [8]To circumvent this limitation, we directly photo-caged the Jagged1 ligands, thus making them inactive, unless the cages are photo-chemically removed later during cell-culture (Figure 2). [9]Upon photo-based removal of the cages, Notch signaling is activated and induces differentiation of hepatoblasts to cholangiocytes in a specific time and location of our choosing.While promising, these methods employ native ligand-receptor interactions to activate cells, which cannot discriminate between cell types if mixed populations are dispersed throughout a cell-culture, such as a co-culture of hepatocytes and cholangiocytes.Thus, targeting a specific population of cells in a heterogenous mixture during 3D co-culture is an unmet need.It should be noted that the Notch signaling is required not just for the differentiation of cholangiocytes, but also to promote their self-assembly into bile ducts. [10]Thus, Notch pathway still needs to be activated spatially even when differentiated cholangiocytes and hepatocytes are co-cultured.
Cholangiocytes and hepatocytes seem to require different biomechanical cues for their optimum growth and morphogenesis.For example, mono-culture studies show that a matrix stiffness of ≈1500 Pa is optimum for human hepatocytes functions. [11]onversely, the promotion of bile-duct growth only requires a stiffness of ≈500 Pa. [10]Hepatocytes and cholangiocyte can also sense the stress-relaxation rate of the matrix.In cholangiocytes, stress-relaxing matrix increased YAP activation (Yes-associated protein) -a widely known readout of mechano-sensing -which was critical for the formation of tubular bile ducts. [10]By contrast, decreased YAP activation is critical for maintaining the maturation of hepatocytes. [11]These studies highlight that there may be no "one-size-fits-all" matrix for co-cultured liver cells because of the diverse effect of bio-mechanical cues on liver cell functions.Yet, the effect of matrix stiffness and stress-relaxation on the functions of both hepatocytes and cholangiocytes in a co-culture system is unclear.Thus, there is an obvious need to further evaluate and identify the most conducive cues for each cell type and develop a matrix that provides both cells with their optimum cues.

Bioengineering Approaches to Overcome Challenges in Integrating Bile Ducts in Liver Organoids
Broadly, there are two ways to create bile ducts within liver organoids.One is to use the bottom-up approach, i.e., the ability of the cells to differentiate and self-assemble into bile-duct structures in co-culture with hepatocytes, when provided suitable growth factors and conducive 3D culture matrix. [13]This can be achieved using stem cells or hepatoblasts as the starting point, or by co-culturing cholangiocytes and hepatocytes.For example, stem cells have been co-differentiated into a mixture of hepatocytes and cholangiocytes where the cholangiocytes selfassembled into cysts with luminal space. [14]Although promising, the organoids lacked tubular bile ducts and luminal connections between the bile duct regions and hepatocyte regions.I argue that this is due to the lack of targeted Notch signaling, which is critical to the morphogenesis of bile ducts as discussed earlier.
A recent study co-cultured mouse hepatocytes and cholangiocytes and demonstrated that metabolites could flow from bile canaliculi of hepatocytes toward bile duct regions (Figure 3a). [12]nfortunately, the use of animal-derived cells and a poorly defined substrate confounded results, limiting its impact and reproducibility.Moreover, the cells were cultured in 2D conditions, rather than 3D conditions, which is needed to mimic the structure of liver.Notwithstanding, this study, for the first time, recapitulated the bile flow dynamic of the liver in vitro.In a more recent study, we showed that cholangiocytes self-assemble into 3D bile duct like structures when grown in a viscoelastic hyaluronanlaminin matrix with immobilized Jagged1 ligands to activate Notch signaling (Figure 3b).The bile duct growth was abrogated when the Notch signaling was inhibited using small molecule inhibitor (Compound E).Together, these studies highlight the potential of creating liver constructs containing integrated hepatocytes and bile ducts.Our research group is developing welldefined and highly tunable biomaterials for 3D culture of liver cells, which will allow us to precisely evaluate the role of Notch signaling and biomechanical cues in the context of co-cultured liver cells.
Another approach that we are developing to create bile ducts within liver constructs is to utilize top-down methods which involve design-based bio-fabrication processes.A key benefit for using bio-fabrication is its ability to create customized microenvironments for both hepatocytes and bile-duct cells within the same construct.For example, instead of using one type of biomaterial to embed both cell types (hepatocytes and cholangiocytes), we are developing two different bioink materials that will but not with mature hepatocytes, when co-cultured in 2D on a collagen substrate.Adapted according to the terms of the CC BY license. [12]Copyright 2021, The authors, published by Springer Nature.b) The cholangiocytes grew into tubular, bile duct structures when cultured in a 3D hyaluronan-laminin matrix with immobilized Jagged1 (left and middle).When the cells were cultured in the presence of compound E (Notch inhibitor) in the same matrix, the cells mostly grew as cysts instead of forming bile-duct structures (right).Adapted with permission. [10]Copyright 2022, Wiley-VCH GmBH.
be used to create compartmentalized, niche regions for each cell type.One bioink material can be used to encapsulate cholangiocytes, while the other bioink can be used to encapsulate hepatocytes.Given that each cell type is embedded in a different material, the physio-chemical properties and ligands therein can be tuned independently for each cell type.Thus, this strategy will potentially allow us to independently tune both bio-mechanical and bio-chemical cues, which is not possible when using one matrix for both cells.Finally, the biofabrication approach can allow the creation of a bile fluid outlet, thereby providing a possibility of creating an organoid-on-a-chip system.Even though liver-ona-chip techniques have benefits for producing functioning liver organoids, they continue to be a difficult and costly method that needs further research.Biofabrication approaches make it easier to create and scale up 3D architecture, particularly to examine liver disorders and conduct drug screening.
In a bio-fabricated liver construct, where the cholangiocyte and hepatocytes will be present in distinct regions (and not mixed together), a key challenge will be to ensure that the bile-canaliculi of the hepatocytes are seamlessly connected with the cholangiocyte ducts.This will be important for a continuous flow of bile fluid from regions of mostly hepatocyte to the cholangiocyte-laden regions.This will require: 1) conducive matrix that allows cellular movement and self-assembly at the hepatocytes/cholangiocytes interface so that two cell types can integrate as they do in the liver, and 2) identifying bio-chemical signaling that improves the cellular connections between cholangiocyte and hepatocytes.
Lastly, we are developing a combination of cell and biomaterial engineering that can be used synergistically to achieve targeted Notch signaling in co-cultured cholangiocytes.Morsut et.al. developed synthetic genetically-engineered Notch receptors that, when expressed in cells, make them responsive to synthetic ligands of choice such as GFP. [15]In this way, user-defined responses were activated specifically in the cell population of interest, such as altering the T-cell secretion profile. [15]Yet, the current synthetic Notch receptors lack Notch intracellular domain (NICD) which is needed to activate the Notch target genes.Thus, whether the synthetic Notch receptors with NICD can activate Notch signaling in a targeted cell population is untested.In theory, cholangiocytes can be engineered to express synthetic Notch receptors containing NICD, that can then recognize corresponding ligands immobilized in the engineered cell-culture matrix.In this way, synthetic Notch receptor-ligand interaction can be used to activate Notch signaling in cholangiocytes, while ensuring that the hepatocytes are not activated.We hope to develop this approach and combine this with biofabrication to create a rationally designed bile-duct system in liver constructs.
In conclusion, bile duct integration in liver organoids is critical to effectively model bile duct diseases.Recent developments have made it possible to grow and differentiate cholangiocytes and hepatocytes from both primary and stem cells.Significant progress has been made in elucidating the signaling pathways and developing synthetic biomaterials to provide a conducive microenvironment for the 3D in vitro growth of bile ducts.Continuing work in developing methods to spatially activate cell signaling at cellular-level resolution, and advances in biofabrication will eventually pave the way to form liver organoids with integrated bile ducts.This will enable researchers to gain a deeper understanding of bile duct disorders and evaluate novel treatment options.

Figure 1 .
Figure 1.Notch signaling pathway controls differentiation of bipotent liver progenitor cells.Upon Notch activation, the hepatoblasts differentiate to cholangiocyte which eventually form bile ducts in the liver.Conversely lack of Notch activation induced hepatoblasts differentiation to hepatocytes.

Figure 2 .
Figure 2. Photo-based spatial control of Notch activation using light.a)Jagged1 (Jag1) were caged using photo-cleavable linker to inactivate ligand-receptor interaction.Upon UV exposure, the cages can be removed to activate the Jag1 ligands in selected regions of the cell culture matrix using photomask.b) Spatially controlled Notch activation in selected region of the cell culture substrate as verified using fluorescence of Notch2 reporter (green) reporter cells.Adapted with permission.[9]Copyright 2021, Wiley-VCH.

Figure 3 .
Figure 3. Creation of bile ducts during in-vitro culture.a) Cholangiocytes established luminal connections with bile-canaliculi of small hepatocytes,but not with mature hepatocytes, when co-cultured in 2D on a collagen substrate.Adapted according to the terms of the CC BY license.[12]Copyright 2021, The authors, published by Springer Nature.b) The cholangiocytes grew into tubular, bile duct structures when cultured in a 3D hyaluronan-laminin matrix with immobilized Jagged1 (left and middle).When the cells were cultured in the presence of compound E (Notch inhibitor) in the same matrix, the cells mostly grew as cysts instead of forming bile-duct structures (right).Adapted with permission.[10]Copyright 2022, Wiley-VCH GmBH.