Human liver organoids: From generation to applications

Abstract

In the last decade, research into human hepatology has been revolutionized by the development of mini human livers in a dish. These liver organoids are formed by self-organizing stem cells and resemble their native counterparts in cellular content, multicellular architecture, and functional features. Liver organoids can be derived from the liver tissue or pluripotent stem cells generated from a skin biopsy, blood cells, or renal epithelial cells present in urine. With the development of liver organoids, a large part of previous hurdles in modeling the human liver is likely to be solved, enabling possibilities to better model liver disease, improve (personalized) drug testing, and advance bioengineering options. In this review, we address strategies to generate and use organoids in human liver disease modeling, followed by a discussion of their potential application in drug development and therapeutics, as well as their strengths and limitations.

Generation and application of liver organoids in human disease modeling, drug development, and therapeutics. BME, basement membrane-like matrix; CLC, cholangiocyte-like cell; HLC, hepatocyte-like cell; HOs, hepatic organoids; iCOs, intrahepatic cholangiocyte organoids; iPSC; induced pluripotent stem cell; KCL, Kupffer-like cell; ULA, ultra-low attachment. This figure was created using BioRender.

INTRODUCTION

Extrapolating results from inappropriate liver model systems to humans have hampered disease modeling, drug development, and artificial liver engineering. Most in vitro models have been relying on the widely available and extensively characterized human hepatic cell lines derived from cancer or by the genetic engineering of primary liver cells (Figure 1). They display high proliferation capacity and stable metabolism under standardized experimental settings, which are accompanied by deficits in differentiated functional performance. Conversely, although primary human liver cells mirror the functionality of the human hepatic cells in vivo,¹ their low availability, short lifespan, and difficulties in standardization due to challenging procedures for isolation, cryopreservation, and cultivation, limit their broad use in biomedical research. Another methodological parameter to consider is the dimensionality of the cell culture; most studies have been relying on2 dimension (2D) models generated from dissociated cell cultures that are expanded on plastic dishes in the presence or absence of an extracellular matrix (ECM) scaffold due to their relatively low cost and ease to manipulate and maintain. Nevertheless, 2D cultures do not properly mimic cell polarity, morphology, and motility and cell-cell and cell-ECM interactions in 3 dimensions (3D). In addition, 2D cultures fail to recapitulate the gradients of soluble factors, nutrients, oxygen, and waste bioproducts the liver cells are exposed to in vivo. These shortcomings may be overcome by 3D cultures that better resemble the intricate architectural and functional properties of in vivo tissues, such as cell density, organization, and communication. However, 3D cultures are often laborious and limit the measurement of certain readouts in high-throughput screening assays.^2,3

FIGURE 1.

Comparison of organoids with other systems to model liver disease. Abbreviations: 2D, two-dimensions; 3D, three dimensions; PHH, primary human hepatocytes. This figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.

The liver has a heterogeneous cellular composition, comprising ~65% hepatocytes, ~10% tissue-resident macrophages Kupffer cells, ~5% HSC, ~15% liver sinusoidal endothelial cells, ~5% cholangiocytes, and less abundant cells, such as other immune cells and adult stem cells, which spatiotemporally cooperate to shape and maintain liver functions.⁴ Further, multidirectional interactions between the liver and other organs play pivotal roles in the coordination of cellular processes under homeostasis and localized or systemic stress. Ideally, such cell types should be co-cultured together in an in vitro model system. Although animal models are more likely to capture this complex communication between different liver cells and liver-organ interactions, interspecies variability often hinders the extrapolation of data to humans, particularly biological processes associated with development, metabolism, toxicology, and pharmacology.⁵ In addition, ethical precepts urge the technological development of manifold refined alternatives to reduce or replace the use of animals in biomedical research, eventually with even better outcomes.

One option in which most of the above-described hurdles are overcome is by using human precision-cut liver slices. In these thick liver slices, the complex and multicellular/ECM histoarchitecture of the hepatic environment is retained, and this makes for a unique ex vivo system located in between experimental and human studies. Samples for precision-cut liver slice preparation are obtained from freshly resected tissue, usually in the framework of surgical waste, and explanted or non-transplantable tissue, which requires tuned coordination between the clinical and laboratory teams and limits access to truly “healthy” liver tissue. In addition, human precision-cut liver slices retain optimal function and metabolic capacity only up to 5 days in culture.^6,7

In the last decade, liver organoids have emerged as a powerful tool to fill this gap in model systems more closely mimicking human liver function. In this review, we discuss this innovation in hepatology research and the breath of applications to model liver disease, aid in drug development pipelines, and open up the path towards liver bioengineering (Figure 2).

FIGURE 2.

Applications of human liver organoids. Liver organoids can be isolated from healthy donors and liver disease patients for disease modeling, including response to external stimuli and stress signals, gene editing or cell communication in microfluidic devices. Likewise, liver disease patient-derived organoids can be used to predict the drug response supporting the therapeutic decision-making process (precision medicine). Organoids can also feature as a platform for drug discovery by providing a robust assessment of human drug sensitivity and hepatoxicity. Finally, liver organoids may themselves constitute a source for regenerative medicine-based therapeutics alone or coupled with tissue and/or genetic engineering. This figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.

ORGANOIDS: A NEW DIMENSION IN LIVER MODELING

Liver organoids are defined as a 3D structure derived from pluripotent stem cells, progenitor, and/or differentiated cells, as well as primary and metastatic tumors, which self-organize through cell-cell and cell-matrix interactions to recapitulate architecture and function features of the tissue of origin.⁸ This definition and classification system was agreed upon by over 60 experts in hepatic, pancreatic, and biliary organoid culture through an adopted Delphi method. Since, in recent years, immense progress has been made in the field of liver and pancreatic organoid technology, revisiting the terminology was necessary to facilitate communication, interpretation, and reproducibility of liver organoid data. In this review, we will use the classification and nomenclature proposed accordingly. Liver organoids can be subclassified as tumor-organoids (tumoroids) and epithelial organoids, namely hepatocyte, cholangiocyte, and non-parenchymal organoids, or multi-tissue/lineage organoids comprising both parenchymal and supporting cells that mimic the intra-organ self-organization.⁸

The groundwork for the current state-of-the-art liver organoid technology lies in our improved understanding of the knowledge on molecular mechanisms of liver regeneration, stem cell maintenance, and differentiation, together with the knowledge of ECM requirements for the 3D culture of cells, which will be discussed in the following paragraphs.

Organoids derived from primary liver and biliary tree tissue

Adult tissue stem cells from various epithelial structures were shown to be a great source for organoid culture (Figure 3). The groundwork for this innovation came from our knowledge of liver development. During liver organogenesis, embryonic progenitor cells, or hepatoblasts, arise from the posterior foregut endoderm. In response to factors secreted by the surrounding tissue, such as FGF, bone morphogenic proteins, HGF, and Wnt, these hepatoblasts undergo morphology changes, start to proliferate, and migrate into the adjacent mesoderm to form the liver bud.⁹ The liver bud undergoes outgrowth, and the liver lobes form subsequently, after which differentiation of the hepatoblasts into hepatocytes and cholangiocytes takes place.

FIGURE 3.

Generation and classification of human iPSC-derived, adult stem cell-derived and liver cancer-derived liver organoids. iPSC-derived liver organoids are established upon epithelial progenitor commitment, followed by hepatocyte and/or cholangiocyte maturation by culturing them with specific growth factors. The co-culture or concomitant differentiation of different iPSC germ layers can result in muli-tissue liver organoids, composed of parenchymal and non-parenchymal cells. Adult stem cells–derived liver organoids require their isolation, followed by culture in extracellular matrix in the presence of specific growth factors. Those organoids lack mesenchymal or immune components, and depending on the origin of adult stem cells and differentiation protocol are designated ICO, ECO, GCO, and HO. Finally, liver cancer organoids can be generated from cancer stem cells isolated from primary cancers and their classification follows the nomenclature of the parental tumor. Abbreviations: ECO, extrahepatic cholangiocyte organoid; GCO, gallbladder cholangiocyte organoids; HO, hepatocyte organoids; ICO, intrahepatic cholangiocyte organoid; iPSC, induced pluripotent stem cell. This figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.

Using this knowledge, the seminal work of the Clevers group showed clonal expansion of mouse single leucine-rich repeat-containing G-protein coupled receptor 5 positive (Lgr5⁺) liver hepatoblast cells into cystic organoids.¹⁰ These mouse liver organoids turned out to originate from adult cholangiocytes that, by epigenetic reprogramming, adopt a progenitor cell state^11,12 and are therefore classified as intrahepatic cholangiocyte organoids (ICOs). In their described growth or expansion medium, these organoids resemble hepatoblast progenitor cells that express markers of both cholangiocyte and hepatocyte lineages, indicating a biopotential nature. Upon changing medium conditions promoting hepatocyte differentiation, organoid cultures indeed showed an increase in hepatocyte markers and albumin production.¹⁰ To allow for the culture and expansion of human ICOs, the medium needed to be adapted to contain inhibitors of TGF beta, ALK receptor tyrosine kinases 4/5/7, and the adenylate cyclase agonist forskolin.⁸ Also, these hICOs are bipotent, as a medium that lacks R-spondin and forskolin, resulting in organoids maturing into cholangiocyte-like cells. However, the addition of dexamethasone, bone morphogenetic protein 7, FGF19, and the Notch inhibitor N-[N-(3,5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester drove the acquisition of mature hepatocyte functions, such as albumin and bile acid secretion, glycogen storage, phase I and II drug metabolism, and ammonia detoxification.⁸

The fast progress in this area was led by the findings that upon injury, cholangiocyte-like oval cells represent activated liver stem cells that can produce both cholangiocytes and hepatocytes.^13–15 However, it was also shown that adult mature hepatocytes are reprogrammed into proliferative bipotent progenitor cells upon liver injury.^16–19 These dedifferentiated hepatocyte progenitor cells were also argued to be the main source of the liver’s regenerative capacity in vivo.^19–21 Based on these findings, attempts were also made to generate protocols to grow organoids from primary hepatocytes.

A protocol to culture organoids from primary mouse hepatocytes was developed.²² Using knowledge from the field of liver regeneration, it was inferred that a proinflammatory stimulus, which is known to drive liver regeneration, could aid in organoid expansion. Indeed, the addition of TNF alpha and the Wnt agonist CHIR99201 resulted in the outgrowth of hepatocyte organoids (HOs) from primary mouse hepatocytes that can be expanded for at least 8 months. Culturing them in a differentiation medium resulted in acquiring mature hepatocyte functions, that is, uptake of lipid particles, glycogen storage, and canalicular transport.²² Different expansion media, using different growth factors and R-respondin (Wnt potentiator) were developed to allow for mHO organoid outgrowth with similar results.²³ Also, HO outgrowth is feasible from primary hepatocytes from human fetal and adult livers, although the adult cultures had very low plating efficiency and could only be expanded for 2.5 months, after which the cultures deteriorated. Although very promising, this calls for further improvements in media conditions to support the long-term expansion of adult HOs. In a head-to-head comparison of human ICOs and adult HOs to primary hepatocytes, morphology and gene expression analyses point towards better differentiation towards hepatocytes of the HOs compared to ICOs.²³

Upon the success of culturing ICOs, successful attempts were also made to culture mouse and human gallbladder cholangiocyte organoids (GCOs) and extrahepatic cholangiocyte organoids (ECOs).^24–27 In contrast to primary tissue, ICOs, ECOs, and GCOs are very similar under the same Wnt-stimulated culture conditions, and have only a very limited number of markers that correspond to the regional primary tissue due to the lack of niche stimuli.^27,28 However, only ICOs could be differentiated in vitro into a hepatocyte-like phenotype, which is likely explained by their origin: hepatoblasts (ICOs) versus the caudal side of the hepatic diverticulum (ECOs). The great promise for clinical applications of culturing ECOs and GCOs is illustrated by the implantation of scaffolds on which ECOs were grown in mice suffering from common bile duct or gallbladder injury, resulting in the repair of the biliary epithelium. In addition, it was shown that hGCOs infused in ex vivo normothermic perfused livers resulted in the engraftment of these hGCOs, repair of the intrahepatic biliary tree, and a concomitant functional improvement of choleresis.²⁶

Lastly, it has been demonstrated that self-organizing 3D structures could be cultured from primary and metastatic tumors and even tumor needle biopsies of the liver and extrahepatic bile ducts.^29,30 HCC-derived organoids (HCCOs) have the histological organization of the native tumor, and the tumor mutation landscape and transcriptome resemble the original tumor.²⁹ This was also the case for the culture of intrahepatic cholangiocarcinoma organoids (iCCAOs).^29,31 Of note, iCCAOs retained the drug-resistance phenotype, paving the way for detailed mechanistic and personalized drug interaction studies.³¹

Organoids derived from pluripotent stem cells

Next to organoids from primary tissue, also embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) are nowadays used as starting material to grow organoids for different liver cell types (Figure 3). The advantage of pluripotent stem cells as the origin for organoid culture is that they can be generated from skin biopsies,³² hair,³³ kidney epithelial cells present in urine³⁴ or blood cells,³⁵ all relatively easily accessible cell types, and can be steered to differentiate into any cell type using the appropriate culture conditions. Since generating ESCs requires manipulation of the preimplantation stage embryo, the use of embryonic pluripotent stem cells is limited because of ethical objections. By introducing (analogues of the) the so-called “Yamanaka factors” (cMyc, Oct3/4, Sox2, and Klf4), the mature somatic cells are reprogrammed into iPSCs.³⁶

Self-organizing liver spheroids were developed from iPSCs from healthy persons and patients with Alagille syndrome (ALGS).³⁷ The resulting organoids were composed of both hepatocyte-like as well as cholangiocyte-like cells that were able to accumulate glycogen and produce albumin and bile acids. Also, by combining mesodermal and endodermal parallel differentiation, organoids containing hepatocyte-like cells and primitive vascular epithelial networks were produced and engrafted upon transplantation into immunocompromised mice, connected to the host vasculature.³⁸

Others developed co-differentiation protocols of endodermal and mesodermal lineages, resulting in multi-tissue organoids containing hepatocyte-like, cholangiocyte-like, Kupffer cell-like, and stellate cell-like cells in the organoids.^39,40 Very interestingly, a recent study reported the generation of hiPSC-derived organoids, containing functionally hepatic and biliary interconnected compartments.⁴¹ Also, protocols are available to generate branched tubular structures of cholangiocyte-like cells to study biliary disease mechanisms.⁴²

In conclusion, a lot of progress has already been made, and still, new protocols are being developed to further optimize in vitro “mini-liver” culture regimes. It seems imperative that all these protocols be evaluated similarly to compare the differentiation status of the single cell types for each protocol and cell of origin to decide on what is now the best way to proceed toward clinical applications. However, depending on the research question, several possibilities exist to investigate the basic mechanisms of human liver physiology and regeneration, drug interactions, and disease states in human liver organoids. These innovations will largely overcome the limitations of the use of cancer and immortalized cell lines and will add to current in vivo models, which, together with the difficulties of obtaining human liver samples, previously hampered research into basic hepatology mechanisms and disease interactions.

Extracellular matrix requirements for 3D culture of organoids

To allow for organoid growth in 3D, some form of extracellular support matrix needs to be provided. Alternatively, ultra-low-attachment plates or “hanging drop” methods are used to avoid adherence to the plastic and concurrent 2D growth but promote the clumping together of cells into a spheroid, which produces ECM proteins by itself. Irrespective of the way of ECM support, although often resulting in different phenotypes (cystic vs. cell clump phenotype,), both resulting 3D structures are termed “organoids”.⁴³ The ULA plates and hanging drop methods are useful for the analyses of the ultimate differentiation status of the organoid, however not so much for organoid expansion.

The Engelbreth-Holm-Swarm (EHS) matrix, also known as Matrigel, is widely used for organoid expansion. It contains a mixture of several ECM components, such as collagen type IV and laminin, which supplies a matrix needed for cell growth.⁴⁴ As the Engelbreth-Holm-Swarm matrix is produced from the murine Engelbreth-Holm-Swarm sarcoma tumor,⁴⁵ the production is costly, and batch-to-batch variability limits the reproducibility of experiments. In addition, the murine origin of the matrigel hampers the clinical application of human organoids. It is therefore justified that alternatives such as hydrogels and recombinant protein matrices are being sought after (reviewed in⁴³). For the mass expansion of organoids without the use of matrigel, the use of spinner flasks has recently been shown to increase the growth rate and also the differentiation status of hICOs to hepatocytes, probably due to improved oxygenation and nutrient availability.⁴⁶

In that respect, mentioning the developments toward liver-on-chip technology is also important. Liver-on-a-chip manufacturing involves the generation of microchambers through biomaterial-based microcontact printing or micropatterning techniques, thereby creating microchambers connected by channels to allow for the co-culture of different liver cell types and culture medium perfusion. These microfluidic devices were originally developed to overcome the main challenges of static 2D liver modeling. By controlling the cell culture medium flow, oxygenation, and biomolecule gradient-generation conditions, researchers aimed to better mimic chemotaxis and shear stress and zonation-dependent activity and differentiation artificially in vitro. For example, cell printing was used to develop a 3D liver-on-a-chip with human umbilical vein endothelial cells (HUVEC) to mimic the vascular endothelium or liver sinusoidal endothelial cells and HepaRG.⁴⁷ HepaRG cells were encapsulated in liver-decellularized ECM bio-ink and the HUVEC cells in gelatin to create a 3D-microenvironment prior to placement in the microfluidic system. They were placed amidst vascular and biliary fluidic channels for creating vascular and biliary flow systems.⁴⁷ In a comparative analysis, albumin and urea secretion and acetaminophen response were increased compared to conventional 2D and 3D culture conditions.

Based on these developments, the combination of liver-on-chip technology with organoid technology seems to allow for the best resemblance of native liver tissue in vitro. This would also allow for personalized medicine, as the effects of, for example, genetic mutations and drug response can be studied in this way. As an example of the combination of the 2 technologies, human iPSC-derived 3D organoids were cultured on a liver-on-a-chip, and decreased hepatic function was demonstrated with increased ethanol concentration, which could be reversed by providing fresh culture medium to the damaged 3D liver tissue for few days.⁴⁸

Taken together, innovations in ECM requirements and microfluidics for 3D organoid maturation are paving the way for optimal mechanistic studies of liver physiology and the clinical application of organoids.

LIVER DISEASE MODELING

Disease modeling using liver organoid models derived from patients opens exciting opportunities for the study of liver diseases.^49–51 Liver organoids can be directly generated from the biopsies of patients. The liver organoids retain the genetic background of the respective individual, including specific disease-causing mutations in monogenic diseases and cancer. Alternatively, the CRISPR/Cas9-mediated insertion or corrections of disease-causing mutations can be engineered in liver organoids generated from healthy donors. This allows us to study the effects on protein function, pathogenicity, and response to targeted therapies. In Table 1, we summarize the current progress on modeling diverse liver diseases in organoids, including monogenic liver diseases, metabolic liver diseases, liver cancer, and viral hepatitis, which will be further discussed below.

TABLE 1.

Studies using organoids to model liver disease

Disease	Model	Main findings	Year	Reference
Alpha-1 antitrypsin deficiency (AATD)	hiPSCs	Hepatocytes differentiated from hiPSCs from skin biopsies of AATD patients presented the accumulation of misfolded alpha-1 antitrypsin in the ER.	2010	Rashid et al52
	hiPSCs	Genetic correction of point mutation in SERPINA1 in hiPSCs using zinc finger nucleases and piggyBac technology restored the structure and function of alpha-1 antitrypsin.	2011	Yusa et al53
	hiPSCs	Establishment of the drug screening platform using patient-derived hiPSCs; Five clinical drugs found to reduce alpha-1 antitrypsin accumulation in patient hiPSCs–derived hepatocytes; Mutation correction by transcription activator-like effector nuclease technology in patient hiPSCs–derived hepatocytes resolved alpha-1 antitrypsin accumulation.	2013	Choi et al54
	Human liver tissue	AATD patient liver biopsy–derived organoids presented reduced ability to block elastase, ER stress, and increased apoptosis.	2015	Huch et al55
	hiPSCs	Patient-specific hiPSCs differentiated into hepatocytes modeling individual disease phenotypes of AATD.	2015	Tafaleng et al56
	hiPSCs	Hepatocytes differentiated from control and AATD patients’ hiPSCs used to identify the differential expression of 135 genes; hiPSC-derived hepatocytes used for toxicity prediction upon exposure to hepatotoxic drugs.	2015	Wilson et al57
	hiPSCs	Patient-derived hiPSC hepatocytes used to explore role of JNK signaling in AATD pathophysiology.	2017	Pastore et al58
	hiPSCs	Patient-specific hiPSC hepatocyte-like cells used to study inflammatory intermediates and unfolded protein response in AATD pathophysiology.	2018	Segeritz et al59
	Human liver tissue	Establishment of organoids from ductal cells of human liver biopsies from controls and AATD patients; Organoids presented intracellular aggregation and lower secretion of alpha-1 antitrypsin, and lower ALB and APOB expression; Organoids treated with oncostatin M (SERPINA1 inducer) showed increased expression of alpha-1 antitrypsin transcripts.	2020	Gomez-Mariano et al60
	hiPSCs	Establishment and characterization of a repository of AATD hiPSCs; Differentiated hepatocytes presented increased retention of alpha-1 antitrypsin; Mutation correction by CRISPR/Cas9 decreased the intracellular accumulation of alpha-1 antitrypsin in the differentiated hepatocytes.	2020	Kaserman et al61
	hiPSCs	Using intra-splenic injections, hepatocytes differentiated from hiPSCs were transplanted into the livers of AATD transgenic mice; Transplantation induced a progressive repopulation of the mice livers without evidence of carcinogenicity.	2021	Chen et al62
	hiPSCs	Patient-derived hiPSC hepatocytes used to evaluate alpha-1 antitrypsin-targeting compound; Tested compound blocked alpha-1 antitrypsin polymerization and increased its secretion.	2021	Lomas et al63
	hiPSCs	Mutation correction by adenine base editing in patient-derived hiPSCs reduced ER stress and alpha-1 antitrypsin accumulation.	2021	Werder et al64
Alagille syndrome (ALGS)	hiPSCs	Differentiation of hiPSCs into cholangiocyte-like cells with Notch signaling deregulation.	2015	Sampaziotis et al42
	Human liver tissue	Long-term expansion of adult bile duct–derived bipotent progenitor cells from human liver; Bipotent cell-derived organoids presented duct-like phenotype that could be differentiated into hepatocyte-like in vitro and upon in vivo transplantation in mice; ALGS patient liver biopsy–derived organoids reflected ALGS phenotype with scarce biliary cells unable to integrate epithelium and undergoing apoptosis.	2015	Huch et al55
	hiPSCs	hiPSC-hepatic organoids derived from ALGS patients presented both reduced bile duct formation and regenerative ability, as well as downregulation of Notch signaling and biliary markers; hiPSC-derived hepatic organoids genetically engineered with CRISPR/Cas9 to introduce or reverse causing mutations.	2017	Guan et al37
	GEMM	Establishment of bile duct–derived organoids from Jag1Ndr/Ndr mice.	2018	Andersson et al65
	Human fetal liver	Establishment of liver organoids from human fetal liver progenitor cells, recapitulating hepatobiliary organogenesis; Medium supplementation with Notch signaling inhibitor hampered bile duct maturation.	2018	Vyas et al66
	Mouse liver	Adult mouse intrahepatic and extrahepatic duct organoids used to study Jagged/Notch signaling in the extrahepatic stem cell niche.	2022	Zhao et al67
Cystic fibrosis (CF) liver disease	hiPSCs	CF patient hiPSCs differentiated into functionally impaired cholangiocytes; VX-809 treatment corrected protein misfolding.	2015	Ogawa et al68
	hiPSCs	CF patient hiPSCs differentiated into cholangiocyte-like cells; VX-809 treatment increased CFTR function in vitro.	2015	Sampaziotis et al42
	hiPSCs	Cholangiocytes differentiated from hiPSCs from healthy donors and CF patients; CF patient-derived cholangiocytes presented impaired protein kinase A/cAMP-mediated fluid secretion, increased Src-TLR4 and proinflammatory changes; Src inhibition and VX-809 treatment improved fluid secretion and cytoskeletal abnormalities.	2018	Fiorotto et al69
	Human liver tissue	Establishment and characterization of extrahepatic and intrahepatic cholangiocyte organoids derived from human common bile duct and liver tissue; ECOs from a CF patient presented impaired CFTR channel activity.	2020	Verstegen et al28
	Human liver tissue	Establishment of intrahepatic cholangiocytes from human tissue; Derived cholangiocytes used to assess hypoxia effect on ion secretion.	2021	Roos et al70
Wilson’s disease	hiPSCs	Differentiation of pluripotent hiPSCs carrying R788L mutation into hepatocyte-like cells presenting altered copper transport; In vitro phenotype rescue using self-inactivating lentiviral vector or with curcumin treatment.	2011	Zhang et al71
	Canine liver tissue	Generation of canine hepatic organoids with increased copper accumulation; Copper excretion restored upon lentiviral expression of COMMD1.	2015	Nantasanti et al72
	Canine liver tissue	Canine hepatic COMMD1-deficient organoids with restored COMMD1 expression used for autologous transplantations through the portal vein.	2020	Kruitwagen et al73
Wolman disease	hiPSCs	Establishment of multicellular liver organoids derived from hiPSCs of Wolman disease patient cell lines presented increased lipid accumulation and stiffness; Organoid exposure to FGF19, simulating FXR agonism, suppressed lipid accumulation and improved cell survival.	2019	Ouchi et al40
Glycogen storage disease type 1	hiPSCs	Patient-derived hiPSCs differentiated into hepatocytes presented increased intracellular glycogen accumulation.	2010	Rashid et al52
	hiPSCs	Patient-derived hiPSCs differentiated into hepatocytes recapitulating glycogen, lactate, pyruvate, and lipid accumulation.	2013	Satoh et al74
Citrullinemia type 1	hiPSCs	Establishment of patient-derived hiPSCs presenting ammonia accumulation; ASS1 overexpression rescued ammonia detoxification.	2019	Akbari et al75
Cholangiopathies	Primary human cell lines	Cholangioids established from human primary cholangiocyte cell lines from healthy and primary sclerosing cholangitis patients; Cholangioids recapitulated cellular senescence, senescence-associated secretory phenotype, and macrophage recruitment.	2017	Loarca et al76
	Bile	Bile-derived organoids established from bile of primary sclerosing cholangitis patients; Organoids expressed cholangiocyte markers and showed distinct gene expression profile when compared with bile-derived organoids from control individuals; Patient-derived organoids reacted to inflammatory stimuli from IL-17A.	2018	Soroka et al77
	Bile	Bile-derived cholangioids established from bile of primary sclerosing cholangitis patients showed decreased TRG5 expression predisposing for more severe biliary injury; Biliary epithelial cells incubation with norUDCA restored TRG5 expression levels.	2021	Reich et al78
	Human liver tissue	Biliary organoids derived from liver biopsies of biliary atresia patients; Biliary atresia organoids presented polarity changes, cilia misorientation, and expressed less developmental and functional markers; Biliary atresia organoids phenotype restored upon treatment with EGF and FGF2.	2021	Amarachintha et al79
	Human liver tissue	ICOs derived from liver biopsies of primary sclerosing cholangitis patients; Necroptosis induced in patient-derived ICOs and used for necroptosis inhibitors drug screening.	2022	Shi et al80
	hiPSCs	iPSC-derived cholangiocytes and cholangioids established from skin fibroblasts of healthy individuals and PCS patients; Patient-derived cholangioids presented disease-specific features and predisposition to cellular senescence; RNA sequencing revealed enrichment of cell cycle, senescence and fibrosis-related pathways.	2022	Jalan-Sakrikar et al81
Metabolic liver disease	Liver tissue	Liver organoids established from feline, mouse, dog, and human liver tissue; All organoids presented lipid accumulation upon FFA treatment.	2017	Kruitwagen et al82
	hiPSCs	Establishment of multicellular liver organoids derived from hiPSCs from healthy cell lines; Functional profiles of hepatocyte-like, stellate-like and Kupffer-like cells present in the liver organoids were evaluated; FFA-treated liver organoids presented lipid accumulation, increased lipid droplet size, hepatocyte ballooning, and increased expression of inflammatory markers, changes compatible with steatohepatitis.	2019	Ouchi et al40
	Liver tissue	Liver organoids derived from cat liver tissue used for compound screening.	2020	Haaker et al83
	hiPSCs	Healthy persons and patients with NASH–derived hiPSCs differentiated into hepatocytes; NASH patient–derived hepatocytes presented spontaneous lipid accumulation in the absence of FA.	2020	Gurevich et al84
	hiPSCs hESCs	Hepatic organoids derived from hiPSCs and hESCs presented functional hepatocyte-like cells and cholangiocyte-like cells; Organoid incubation with FFA induced lipid accumulation, lipid droplet increase, and increased ROS and lipid peroxidation, representative of NAFLD; Organoid incubation with troglitazone induced bile canaliculi decay modelling cholestasis.	2020	Ramli et al41
	hiPSCs	Generation of liver organoids from hiPSCs using an organ-on-chip approach; Liver organoids generated in a perfusable PDMS chip and exposed to FFA; FFA exposed organoids presented lipid droplet formation, TG accumulation, increased ROS and expression of fibrogenic and proinflammatory markers.	2020	Wang et al85
	Human liver tissue	Bipotent ductal organoids differentiated from liver tissue of patients with NASH; NASH liver organoids presented upregulated proinflammatory pathways and fibrosis markers, lipid accumulation, and increased apoptosis sensitivity.	2021	McCarron et al86
Alcohol-associated liver disease	Fetal liver tissue	Co-culture of hepatic organoids with human fetal liver mesenchymal cells; Upon ethanol treatment, co-cultured organoids present steatosis, fibrosis, release of inflammatory cytokines, and oxidative stress.	2019	Wang et al87
Primary liver cancer	Human liver tissue	Liver cancer organoids derived from human tumor resection samples of HCC, CCA and CHC patients, presenting preserved histological and genetic features of original tumors; Tumor-organoids presented tumorigenic potential in xenograft models: Tumor-organoids used for drug screening of primary liver cancer-targeting compounds.	2017	Broutier et al29
	Human liver tissue	Organoids derived from human tumor needle biopsies of HCC and CCA patients, presenting morphological and expression patterns as original tumors; Sensitivity to sorafenib was evaluated in both HCC-derived and CC-derived organoids.	2018	Nuciforo et al30
	Human liver tissue	Liver cancer organoids derived from diverse regions of surgical specimens of HCC and CCA patients, and cell lines established; Cancer organoid sensitivity tested against 129 FDA-approved drugs.	2019	Li et al88
	Human liver tissue	Generation of CRISPR/Cas9 engineered cholangiocyte organoids to study the role of BAP1 tumor suppressor.	2019	Artegiani et al89
	hiPSCs	HOs derived from hiPSCs genetically engineered to model initial features of human liver cancers.	2019	Sun et al90
	Murine liver tissue	Generation of liver organoids from liver tissue of wild-type-, Kras- and p53- mutant mice; Organoids presented tumorigenic potential in xenograft mouse model, and developed tumors presented CC-compatible features; Sensitivity to gemcitabine was evaluated.	2019	Saborowski et al91
Viral hepatitis	hiPSCs	Functional liver organoids derived from hiPSCs and infected with HBV; HBV-infected organoids recapitulated virus life cycle and altered hepatic features.	2018	Nie et al92
	Human liver tissue	Liver organoids derived from liver specimens of healthy donors and HBV-infected patients; HBV-infected liver organoids used for drug screening; HVB-infected liver organoids presented an HCC-compatible gene signature.	2021	De Crignis et al93
	Human liver tissue	Co-culture of CD8+T cell and liver organoids in a microfluidic chip to monitor the response to HCV.	2022	Natarajan et al94
SARS-CoV-2 Infection	Liver progenitor cells	Human liver ductal organoids established at infected with SARS-CoV-2 virus; Infected organoids presented downregulation of tight junctions expression and increased expression of cell death- and cellular response to external stimulus–associated genes.	2020	Zhao et al95
	Human liver tissue	Liver organoids derived from human liver biopsies and infected with SARS-Cov-2; Cholangiocyte-like cells involved in viral replication efficiency.	2022	Lui et al96
	Human liver tissue	Liver organoids and biliary organoids derived from normal liver tissue from liver cancer patients; Organoids subjected to SARS-CoV-2 infection and liver infection route explored; Infected organoids presented the upregulation of proinflammatory cytokines and pathways.	2022	Zhao et al97

Abbreviations: AATD, Alpha-1 antitrypsin deficiency; ALGS, Alagille syndrome; CCA, cholangiocarcinoma; CF, cystic fibrosis; CHC, combined HCC/CC tumors; ECOs, extrahepatic cholangiocyte organoids; ER, endoplasmic reticulum; FA, fatty acid; FFA, free fatty acid; GEMM, genetically engineered mouse model; hESCs, human embryonic stem cells; hiPSCs, human induced pluripotent stem cell; HO, hepatocyte organoids; ICOs, intrahepatic cholangiocyte organoids; PDMS, poly dimethylsiloxane; ROS, reactive oxygen species; TG, triglycerides.

Monogenic liver diseases

Monogenic liver diseases comprise a heterogeneous group of disorders caused by single gene mutations associated to phenotypes with variable degrees of liver parenchyma damage, extrahepatic complications, or liver involvement. Some, if not all, are associated with a high-risk factor of HCC development. In most cases, liver transplantation is currently the only curative option, and replacing the entire dysfunctional organ corrects the genetic defect in the liver.^49,98 Although monogenic diseases originate from mutations in a single gene, there are hundreds of monogenic diseases in which the liver is the primary organ affected. In addition, hundreds of causal mutations per monogenetic disorder may be involved. The ability to culture liver organoids that recapitulate the specific genetic background for each disease and/or patient opens new avenues to the understanding of genotype-phenotype correlations and to the development of new therapeutic approaches.^49,50

Alpha-1 antitrypsin deficiency

Alpha-1 antitrypsin deficiency (AATD) is a common autosomal codominant disorder caused by mutations in the SERPINA1 gene, encoding alpha-1 antitrypsin, a serine protease inhibitor mainly produced in the liver. Mutations in SERPINA1 cause the accumulation of misfolded proteins, primarily in the liver and/or lungs, leading to liver cirrhosis and emphysema.⁹⁹ Several disease models of AATD have been described using organoids, iPSCs, or patient-specific hiPSCs that recapitulate key phenotypic features of AATD such as increased alpha-1 antitrypsin retention or endoplasmic reticulum stress.^{52,55,56,60,61} AATD patient-derived hiPSCs have been used to deepen the knowledge of AATD molecular mechanisms, exploring the role of Jun N-terminal kinase signaling, inflammatory response, unfolded protein response, or the establishment of AATD-associated gene expression profiles.^57–59 Importantly, organoids and hiPSCs have been pivotal for drug discovery and therapeutic approaches. hiPSCs were used to correct and insert AATD mutations by adenine and CRISPR/Cas9 base editing, piggyBack, transcription activator-like effector, and zinc finger nuclease technology, with the successful rescue of cellular phenotypes.^53,54,61,64 Importantly, hIPSC-derived hepatocyte-like cells were successfully transplanted into AATD transgenic mice in which the “Z” patient mutation was introduced, which resulted in a progressive repopulation of the mouse livers.⁶² Moreover, AATD patient-derived hiPSCs and organoids have been used to create a high-throughput screening platform, and several compounds have been tested as treatment options for AATD.^54,60,63

Alagille syndrome

ALGS is an autosomal dominant disorder caused by mutations in JAG1 and NOTCH2 genes, encoding important members of the Notch signaling pathway. The main hepatic feature of ALGS is the lack of development of functional bile ducts leading to chronic cholestasis.¹⁰⁰ Several authors have described differentiation protocols to generate ALGS-recapitulating liver organoids from iPSCs, genetically engineered mouse models, or human liver tissue.^37,42,55,65 In addition, using CRISPR/Cas9 gene editing, both the introduction and reversion of ALGS-causing mutations have been successfully engineered into iPSC-derived hepatic organoids.³⁷ Other approaches have focused on using primary tissue-derived organoids to explore liver development disease models. Notch signaling was inhibited in liver organoids derived from human fetal liver progenitor cells in an attempt to create an ALGS liver developmental disease model.⁶⁶ In turn, by combining a zebrafish model, mouse-derived organoids, and multiple lineage tracing, it has been suggested that the modulation of Jagged/Notch signaling might improve the regeneration of the intrahepatic duct in ALGS.⁶⁷

Cystic fibrosis

Cystic fibrosis (CF) is possibly the most common autosomal recessive genetic disease in Caucasians. Caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, CF mainly affects the lungs, but as a multisystemic disease, it often causes liver involvement featuring decreased bile fluidity, which concomitantly leads to biliary cirrhosis.¹⁰¹ Several authors have generated hiPSC-derived and primary tissue-derived organoids to study extrahepatic bile duct diseases, including CF. The resulting cholangiocyte-like cells have successfully recapitulated features of CF phenotype, such as impaired CFTR channel activity.^{28,42,68–70} Notably, the treatment of these cholangiocyte-like cells with corrector drug VX-809 increased CFTR function in vitro, as evidenced by the improvement of fluid secretion.^42,68,69

Wilson disease

Wilson disease is an autosomal recessive disorder caused by mutations in the ATP7B gene, which encodes an ATP-dependent transporter that exports copper from cells (eg, hepatocytes). Mutations in ATP7B lead to copper accumulation in the liver and brain.¹⁰² The phenotype of patient-derived hiPSCs–derived organoids was rescued by the lentiviral overexpression of both wild-type ATP7B and curcumin, showing the potential of using iPSCs for personalized therapy and opening doors to future autologous transplantation in Wilson disease.⁷¹ Further advances have been made by modeling copper-overload diseases in canine organoid models. Hepatic organoids were established from CommD1-deficient dogs. These organoids indeed accumulate high concentrations of copper, which were rescued by wild-type gene supplementation.⁷² Autologous portal vein injections with liver organoids supplemented with wild-type CommD1 engrafted in the dog liver partly corrected the excessive copper accumulation in these dogs.⁷³

Other diseases

iPSCs-derived and primary tissue-derived organoids are also being applied to other monogenic liver diseases such as Wolman disease,⁴⁰ glycogen storage disease type 1,^52,74 or citrullinemia type I.⁷⁵ Overall, these models recapitulate cellular phenotypic traits of each disease and present a platform for therapy testing.

Cholangiopathies

Cholangiopathies, also designated as biliary diseases, comprise an array of human disorders and syndromes that may arise from several inflicting insults, including idiopathic (eg, biliary atresia), genetic defects (eg, ALGS and CF), toxin or drug toxicity (eg, carbamazepine), dysregulation of the immune system (eg, primary biliary cholangitis and primary sclerosing cholangitis [PSC]), hepatobiliary malignancies, or obstruction of the biliary tract. Despite their heterogeneity, those diseases share basic hepatocyte and cholangiocyte-targeting processes that contribute to their pathogenesis. In response to injury, cholangiocytes and hepatocytes proliferate, which may lead to progressive periductular and biliary fibrosis, and eventually cirrhosis. In fact, cholangiopathies are progressive disorders that may lead to end-stage liver disease owing to a lack of effective medical therapies.¹⁰³ Each cholangiopathy has a differential etiology and a unique presentation and course; however, necroptosis is emerging as a common pathway of programmed cell death in all cholangiopathies. Evidence of increased markers of necroptosis effector proteins has been found in cholangiocytes of patients with primary biliary cholangitis and during human end-stage liver diseases, while this cell death routine contributes to liver damage in animal models of cholestatic liver disease.^80,104–106 Further, it has been shown that the necroptosis-associated cytokine milieu may control the lineage commitment of liver carcinogenesis in mice, favoring CCA development.¹⁰⁷ Importantly, hICOs can recapitulate cholangiopathy-associated necroptosis. Specifically, death receptor stimulation and human bile-induced necroptosis in both healthy donor–derived hlCOs and PSC-derived hICOs. Those organoid cultures further uncovered variable inter-individual drug responses.⁸⁰ Noteworthy, biliary organoids can be cultured from stem cells isolated from human bile collected upon endoscopic retrograde cholangiopancreatography. Bile-derived healthy donors and PSC organoids presented similar growth rates and cholangiocyte markers but differentially expressed genes involved in immune regulation and secreted chemokines upon inflammatory stimuli, suggesting that PSC-derived organoids retain the characteristic immuno-reactive phenotype.⁷⁷ Moreover, a similar approach was used to study bile acid toxicity in PSC, showing that TGR5 levels were decreased in PSC-derived bile ducts.⁷⁸ Thus, bile-derived organoids can be an additional source of biobanked organoids that could be obtained longitudinally over the course of cholangiopathies for the study of their pathogenesis and drug sensitivity. By using human primary cholangiocyte cell lines, other studies have also established and characterized normal-bile duct and PSC-bile duct organoids (cholangioids) that were functionally active. Curiously, senescent biliary organoids derived from normal cholangiocytes mimicked several features of PSC, including senescence-associated secretory phenotype and macrophage attraction, showing utility in exploring disease pathogenesis and therapeutic target identification.⁷⁶ Likewise, biliary epithelial organoids grown from the skin fibroblasts of patients with PSC reprogrammed into hiPSCs and subsequently differentiated to cholangiocytes were intrinsically predisposed to cellular senescence and fibrosis, as opposed to healthy donors hiPSCs–derived biliary organoids.⁸¹

Biliary atresia is a rare, congenital, severe obstructive cholangiopathy of unknown etiology present in neonates. To explore the mechanisms underlying the abnormal cholangiocyte development, cholangiocyte organoids were obtained from normal, biliary atresia, and non-biliary atresia cholestatic livers. Biliary atresia-derived organoids showed abnormal polarity and increased permeability, which were reversed by cholangiocyte differentiation through EGF and FGF signaling, supporting a role of defects in epithelial maturation in disease pathogenesis and hints toward potential treatment approaches for biliary atresia.⁷⁹

Metabolic liver disease

NAFLD affects ~one-quarter of the global adult population, causing a substantial disease burden with wide-ranging social and economic implications.¹⁰⁸ NAFLD is the hepatic manifestation of metabolic syndrome and encompasses a range of conditions with varying degrees of parenchymal liver damage ranging from hepatic steatosis or NAFL, NASH, cirrhosis, and eventually HCC.¹⁰⁹ Mechanistically, lipid overload from increased fatty acid influx due to insulin resistance in the adipose tissue leads to metabolically stressed hepatocytes with the activation of cell death and proinflammatory signaling pathways. Human in vitro models of NAFLD have involved either primary human hepatocytes or human cell lines,¹¹⁰ whose limitations were discussed above. In addition, most studies have addressed underlying NAFLD disease mechanisms individually rather than multiple parallel cellular events and how they impact NAFLD progression. Interestingly, using several different healthy and diseased pluripotent stem cell lines, a reproducible method was developed to derive multicellular human liver organoids comprising hepatocytes, stellate and Kupffer-like cells that, when exposed to free fatty acids, phenocopied the progressive features of the steatohepatitis, from steatosis to inflammation and fibrosis.⁴⁰ Furthermore, organoids from patients with genetic dysfunction of lysosomal acid lipase were rescued by farnesoid X receptor (FXR) agonism-mediated FGF19, leading to reactive oxygen species suppression. Several other studies have reported the use of liver organoids for NAFLD modeling.^{41,84–86,111} Of note, the generation yield and growth capacity of adult liver tissue-derived liver organoids of patients with NASH are far lower than those of normal donor-derived liver organoids, although successfully maintaining the hallmarks of NASH phenotypes, including lower albumin production, increased lipid accumulation, and sensitivity to apoptotic stimuli.⁸⁶ In an attempt to study how liver disease induces HCC, iCOs were generated from patients with (N)ASH. Surprisingly, these precancerous organoids did not harbor an increased mutational load, and therefore, HCC rather seems to derive from disease-associated changes in the microenvironment favoring the outgrowth of precancerous cells.¹¹²

Other studies confirmed that treatment of liver organoids of different origins with free fatty acids resulted in the accumulation of intracellular lipids and may be used to identify drug candidates for further clinical evaluation.^82,83 Thus, liver organoid models can provide robust model systems for studies of NAFLD pathophysiology and progression, as well as testing of new drug candidates.

Alcohol-associated liver disease (ALD) is a highly prevalent chronic liver disease worldwide caused by chronic excessive alcohol consumption. The disease course typically involves the development of alcohol-driven hepatic steatosis, its progression to steatohepatitis, and finally cirrhosis, which in a minority of patients culminates in the development of HCC.¹¹³ Cessation of alcohol consumption is the most important factor for successful therapy, which may involve anti-inflammatory treatments or liver transplantation in the most severe cases. Recently, an in vitro organoid model system was developed that recapitulates typical features of ALD pathophysiology. Human HOs co-cultured with mesenchymal cells, both derived from fetal liver tissue, were exposed to alcohol, which resulted in alcohol-induced hepatocyte-like cell injury, including increased activity of the cytochrome P450 family members CYP2E1 and CYP3A4, as well as ALD phenotypes such as oxidative stress, steatosis, and inflammation.⁸⁷ Interestingly, the mesenchymal cells promoted hepatocyte maturation and showed fibrogenic responses to alcohol, recapitulating the pathophysiology of ALD.

Primary liver cancer

Primary liver cancer is a major cause of cancer-related deaths worldwide, with HCC accounting for 75%–85% of cases and intrahepatic cholangiocarcinoma (ICC) for 10%–15%.^114,115 In most countries, the incidence of HCC has more than doubled in the past 25 years and is predicted to rise significantly even further in the coming years. The high mortality rate of HCC is associated with its predominant development in the context of liver cirrhosis, resulting from HBV and HCV, alcohol abuse, hereditary causes, and severe metabolic syndrome. The differences in prognosis reflect the diversity in the genomic landscape of cancer patients and tumors and the large degree of intratumor and interpatient heterogeneity. For many years, in vitro models of hepatobiliary cancers represented by cancer cell lines (eg, HepG2, Huh7, and Hep3B) were considered of great value, despite their monoclonal nature and unclear representation of the tumor biology. Then, the field moved to the generation of challenging and costly patient-derived xenograft (PDX) models following the transplantation of HCC tissue into immunodeficient mice.¹¹⁶ However, since HCC development in humans as a result of the cirrhotic disease involves a highly active inflammatory environment, which is absent in these PDX models, the question remains on how effectively PDX mimics human liver tumorigenesis.

Nowadays, liver organoids can provide advanced models to study cancer biology and drug response. Patient-derived organoids obtained from the tumor tissue of patients with liver cancer^{29,30,88,115,117–120} are particularly useful as they resemble the histological architecture and genomic features of the original tumor in vitro and form tumors after subcutaneous xenotransplantation into immunodeficient mice. Indeed, using surgical resection tissue, organoid cultures were successfully established from individuals encompassing 3 of the most common subtypes of primary liver cancer, namely HCC, CCA and combined hepato-cholangiocarcinoma.²⁹ Liver cancer-derived organoids were shown to identify genes that have prognostic value. In addition, using a drug screening approach, an ERK inhibitor was identified as a potential therapeutic agent for primary liver cancer. The generation of long-term organoid cultures from tumor needle biopsies of patients with HCC, with various etiologies and tumor stages has also been reported,³⁰ which greatly expands our ability to model liver tumor tissue, as the surgically resected specimens are usually confined to the minority of HCC patients with early tumor stages and/or non-cirrhotic liver. Cancer organoids are also used to test sensitivity to conventional chemotherapies and standard-of-care targeted therapies such as sorafenib.^29,30,88 Nevertheless, drug screening is usually associated with high intratumor drug-response heterogeneity, indicating that tumor-organoid production from several tumor sites, including metastatic foci, and co-culture with mesenchymal cells and immune cells, might help to overcome the influence of tumor heterogeneity and microenvironment.⁸⁸

Much like patient-derived liver cancer organoids, genome-edited liver organoid models may also be useful tools for mechanistic studies of liver cancer initiation and progression and the discovery of new therapeutics against liver cancer. Cancer organoid models can be generated from normal donor cells,^{89,90,121,122} which are then genome-edited by CRISPR/Cas9 technology, introducing cancer driver gene mutations.¹²³ Remarkably, transplantation of such engineered organoids into immunodeficient mice gave rise to xenografts with carcinoma features, with characteristics of either HCC or CCA, depending on the inserted oncogenic driver. These approaches provide ideal models for probing the tumorigenic potential of individual oncogenes and tumor suppressors in vitro and in vivo.^89,91 Thus, patient-derived liver cancer organoids and genome-edited liver organoids have great potential as tools to facilitate personalized medicine.

Viral hepatitis

Despite the outstanding progress in prevention and treatment, viral hepatitis caused by infections with HBV and HCV remains a significant global health problem.¹²⁴ HBV and HCV are both hepatotropic viruses, and liver organoids may be applied in modeling HBV infection and exploring virus-host interactions in an individualized manner. Indeed, cultured hiPSC-derived endodermal, mesenchymal, and endothelial cells were used to generate functional liver organoids in a 3D microwell culture system and were shown to be susceptible to HBV infection. Organoids maintained HBV propagation and produced infectious viruses for a prolonged duration.⁹² Hepatic dysfunction of HBV-infected hiPSC liver organoids was also evident, with impaired hepatic gene expression (albumin, glucose-6-phosphatase catalytic-subunit, and hepatocyte nuclear factor 4 alpha), the release of early acute liver failure markers (aminotransferase and lactate dehydrogenase), and altered hepatic ultrastructure (vacuoles and reduced membrane microvilli). Unlike hepatoma cell lines, iPSC-derived liver organoids endogenously expressed high levels of the sodium-taurocholate co-transporting polypeptide, the bile acid import pump that is also used by HBV to enter the cell. In addition, the authors also infected 2D hepatocyte-like cells derived from the same donors and found a lower susceptibility to HBV compared to the organoids, suggesting that 3D growth might result in a more mature phenotype necessary for HBV infection. More recently, HOs from the livers of healthy donors and patients infected with HBV were shown to model HBV infection and related HCC.⁹³ In this study, exposure to the recombinant virus or serum from HBV-infected patients led to organoid infection and active virus replication. HBV-infected organoids have further been used in drug screening campaigns for both anti-HBV activity and drug-induced toxicity. Of note, transcriptomic analysis of organoids from non-tumor cirrhotic liver tissue of patients with HBV showed the presence of an early cancer gene signature. This implies that these organoids can provide interesting biomarkers for the development of HCC as well as surveillance in HBV-infected patients.

Liver organoids have also been used to dissect the mechanisms of liver injury caused by other viral infections such as SARS-CoV-2. Firstly, it was reported how the human liver is attacked by SARS-CoV-2 through the use of human liver ductal organoids derived from progenitor cells.⁹⁵ Robust replication of SARS-CoV-2 occurred in bile duct cells and downregulated the expression of genes involved in cellular tight junctions and bile acid transport while inducing the production of proinflammatory cytokines. These results were later corroborated by other groups using organoid technology.^96,97 Overall, human organoids are a promising tool to investigate virus-induced liver tissue damage ex vivo at the cellular and molecular levels and eventually test antivirals.

LIVER ORGANOIDS FOR DRUG DISCOVERY AND THERAPEUTICS

The development of new drugs is typically a long and costly process comprising a preclinical stage composed of in vitro and in vivo studies to develop a drug that can safely and effectively be administered for clinical trials. The simplicity of traditional 2D cultures makes them particularly suitable for cell-based high-throughput drug screening assays of relevance for liver disease pathogenesis.^{105,125–127} However, although significant advancements have been made in drug discovery, including those based on tailored precision medicine, the large majority of the promising drug candidates picked from conventional drug screens using 2D cultured cell lines and in vivo xenograft models still failed in clinical trials. Lack of efficacy accounts for ~50% of clinical trial failures, highlighting the limitations of disease models currently used in drug discovery.^128,129

The particular poor clinical translation observed in oncology¹²⁸ can largely be associated with the inability of traditional models to accurately mirror important pathophysiological aspects of the human disease, such as complex intratumor and intertumor heterogeneity, interaction with tumor microenvironment, and drug penetration in the tumor. Because of their ability to model liver diseases, liver organoids are emerging as a promising alternative platform for pharmacological testing. A proof-of-principle screening campaign showed a good correlation between drug sensitivities of 29 anti-cancer compounds, including drugs in clinical use or development, and mutational profiles of patient-derived primary liver cancer organoids while also identifying novel players in the therapeutic arena of primary liver cancer.²⁹ Importantly, this platform is amenable to high-throughput drug testing compatible with personalized medicine approaches. The reported discovery of novel potential prognostic biomarkers and treatment drugs for primary liver cancer was assisted by the integration of multiple-level data from primary liver cancer-derived organoids, including genomic and transcriptomic profiling.²⁹ This supports a tremendous potential for the coupling of organoids with multiomics technologies for accelerating personalized medicine and drug discovery. Indeed, healthy and diseased tissues might be amplified using organoid technology, followed by phenotypic profiling using single-cell analysis and omics platforms to study causal mutations, disease pathogenesis, and disease prognosis, as well as to discover and track treatment regimens.

Liver organoids of CF patients have also been used to quantitate individual drug responses in vitro and guide tailored treatments. Of note, in 2015, a patient carrying a rare genetic defect was the first person to receive effective treatment for CF on the basis of drug screens in intestinal organoids or “mini-guts”.¹³⁰ Human iPSC-derived mature cholangiocyte organoids display high levels of CFTR and primary cilia capable of sensing flow. Those cholangiocytes derived from different patients with CF have been grown in 96-well plates and shown different rescue responses to different combinations of CF drugs.¹³¹ Finally, non-parenchymal liver organoids comprising HSC, co-cultured or not with hepatocytes, have been generated with potential application in hepatotoxicity tests and screening of anti-fibrotic agents since they respond to toxins, infections, or hepatotoxic drugs by producing ECM.^132,133 All of these examples support that biobanks of organoids derived from manifold liver diseases might be used as platforms for drug screening campaigns, eventually with more robust outcomes than the classical in vitro models.

Next to a lack of reliable in vitro model systems, safety concerns are the second main cause that precipitates clinical trial failure^128,129 mainly due to DILI. Human primary hepatocytes are the ” gold standard” for drug metabolism and toxicity testing; however, their limitations call for the development of a better predictive human system to identify compounds at risk of causing DILI at the preclinical stage. Liver organoids derived from iPSCs containing polarized immature hepatocytes with bile canaliculi-like architecture were used in a miniaturized multiplexed DILI screening platform measuring viability and cholestatic and/or mitochondrial toxicity.¹³⁴ This screening test showed an 88% sensitivity and 89% specificity in identifying hepatotoxic drugs from a library of 238 marketed drugs, while also predicting some aspects of human genomic predisposition for drug-induced cholestasis, such as Bosentan-induced cholestasis. This type of miniaturized platform may also be coupled with electromembrane extraction based on electrophoresis for segregating drugs and selected metabolites from components that could interfere with subsequent mass spectrometry-based measurements, allowing more detailed metabolism and toxicological studies.¹³⁵ Overall, liver organoid-based screening platforms have a high potential for liver toxicology studies, facilitating drug screening applications, compound optimization, and personalized medicine.

LIVER BIOENGINEERING

Liver transplant remains the definitive treatment for a wide range of end-stage liver diseases.^136–138 Despite a high rate of patient survival following transplantation, a gap between supply and demand for donated organs, the costs, and potential complications associated with the long-term management of the patient's immune system response to the transplanted graft remain key constraining factors to implementing liver transplantation in “routine” therapy.^136,137 Importantly, demand for liver transplantation is projected to increase, paralleling the increased incidence of chronic liver disease and subsequent progression to advanced stages, whereby alternatives are urgently needed.¹³⁹ Moreover, besides simultaneously providing a platform for modeling liver disease and accelerating (personalized) drug development, liver organoids may themselves constitute a source for regenerative medicine-based therapeutics. Bipotential organoids derived from Lgr5⁺ liver stem cells engrafted in the mouse liver and differentiated to hepatocyte-like cells in vivo, prolonging the survival of fumarylacetoacetate hydrolase-deficient (Fah ^-/- ) mice, a model for tyrosinemia type I liver disease associated with severe liver damage.^10,55 In turn, primary hepatocyte-derived organoids display a higher engraftment efficiency in the injured liver and functionality, including regeneration potential, but are less versatile due to the commitment to the hepatocyte lineage.^22,23 Similarly, primary cholangiocyte organoids are able to repair the biliary tree following injury when in mice, assuming different regional identities according to the site of engraftment.^25,26 This plasticity might eventually allow the patient autologous transplantation of organoids derived from functional cholangiocytes (eg, gallbladder) to repair bile ducts affected by cholangiopathies, avoiding the need for immunosuppression. Noteworthy, ectopic transplantation of multi-tissue organoids may be important in certain contexts of liver damage. Mesenteric transplantation of “mini-livers” composed of parenchymal and supporting cells into immunocompromised mice resulted in functional hepatic cord-like structures connected with host vasculature that rescued the drug-induced lethal liver failure model.¹⁴⁰ The generation of those organoids could be adapted to a clinical-scale production platform using only human leukocyte antigen (HLA) homozygous iPSCs clones, which might facilitate both autologous transplant or generation of a liver organoid haplobank for allogenic transplantation from a matched donor.³⁸

Transplantable organ bioscaffolds might be necessary for surgical reconstruction in cases of extensive liver injury. ECOs have been seeded on biodegradable tubular scaffolds of densified collagen to generate bioengineered human bile ducts, which reconstructed the gallbladder wall and repaired the biliary epithelium following transplantation into a mouse model of injury.²⁵ Decellularized human or porcine livers could be attractive alternative scaffolds that might recapitulate the complex architecture and mechanical properties of the native tissue and retain intact vasculature and biliary network while allowing the repopulation with functional organoids,^66,141 but its applicability in transplantation remains to be established. Liver ECM hydrogels resulting from decellularization can also successfully replace the currently used mouse tumor-derived basement membrane extracts (eg. Matrigel) currently used in organoid culture in vitro, which pose risks for tumorigenesis and human immune reaction.¹⁴² Further applications and limitations of liver or organoids in regenerative medicine could be found in recent reviews subordinated to this topic.^4,143

CHALLENGES AND FUTURE DIRECTION

The use of liver organoids for clinical decision-making and optimizing patient therapy is tempting, particularly since organoids can be generated from liver biopsies or patient-derived iPSCs. However, although responses of liver disease organoids to a wide range of drugs have been successfully correlated with patient outcomes,^29,131 large cohort prospective studies are required before their implementation in clinical settings. In addition, its use in precision medicine raises specific challenges for governance. The creation of biobanks for the storage and distribution to researchers and commercial partners of organoids—living cell lines created from human stem cells or tissue—raises ethical and practical issues concerning identity, informed consent, privacy, disclosure policies, appropriate use, and commercial access and benefit-sharing.¹⁴⁴ Genetic engineering of organoids broadens their potential applications but could also pose additional ethical concerns. On the one hand, it can open the door for new human genetic studies, as it is possible to study the causal relationship between the mutation and disease phenotype in “mini-organs” resembling human physiology, which could be particularly promising to model primary liver cancer and monogenic liver diseases. On the other hand, it could allow the generation of autologous genome-edited organoids for transplantation, such as the precise correction of CFTR mutation by CRISPR/Cas9 gene editing, which in turn may invoke the ethical discussion about human genome editing.¹⁴⁵

Noteworthy, despite all the excitement about liver organoids, this technology should be largely considered a model system under development, continually evolving due to improved derivation protocols but still facing several challenges. The diversity of the system is one of the most urgent issues with organoid technology—widely accepted standardized methodologies for the generation and characterization of different liver organoid types are still missing. Moreover, intracultural liver organoid variability has also been reported, including the impact of passage number on gene expression profile and organoid-to-organoid variability within the culture.¹⁴⁶ Nevertheless, while technical variability could delay the translation of data related to organoid technology to clinical settings, variations in the organoid systems brought on by age and genetic background may offer an opportunity to assess the role of inter-individual variability in liver biology and disease pathogenesis.

Likewise, increasing the complexity of liver organoid systems also has its pros and cons. The traditional liver organoid derivation techniques are designed to maintain and expand epithelial cells, lacking other cell types that are generally present in the liver. However, it could be relevant to model cell-to-cell contact with stromal cell populations and the establishment of vasculature in organoid systems, which has been difficult to obtain.¹⁴⁰ Multi-organ organoids are complex systems composed of the dynamic morphogenesis of hepatic, biliary, and pancreatic structures, invaginating from 3D human iPSC cultures, which display a fascinating interconnectivity of hepatic, biliary, and pancreatic domains.¹⁴⁷ Combining organoid technology with other disciplines, such as microfluidic chips, can also represent a promising tool to study inter-cell and inter-organ communication, as well as drug/antibody discovery¹⁴⁸ in non-static conditions to better mimic the physiologic status. CD8⁺ T cell and adult stem cell-derived liver organoids were co-culture in a microfluidic chip to follow the T cell invasion and morphological changes in the liver organoids in response to a pulsed peptide specific for HCV exposure.⁹⁴ Of note, it should be recognized that organoid systems already have a high level of complexity, whereby the challenging integration of additional components into an intricate system could ultimately blur mechanistic findings and dampen the comprehensive understanding of the pathogenic mechanisms. Thus, it is crucial to select the model with the proper complexity degree to match the adequate study design for a given research question.¹⁴⁵

CONCLUSIONS

Organoid technology is emerging rapidly as a valuable tool in hepatology research, bridging the gap between animal models and humans in modeling liver disease, testing drug sensitivity and toxicology, and advancing novel cell-based therapies. Given the tremendous variability of liver organoid systems, a collective effort should move forward with the creation of consensus guidelines for the selection, generation, characterization, and validation of each system to fully exploit the tremendous potential of human organoids to model liver diseases.