The lack of approved treatments for non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatosis (NASH) is largely explained by the shortcomings of cellular models used to unravel disease mechanisms, identify and validate new drug targets, and screen compounds for therapeutic potential. To address this problem, researchers at Hubrecht Institute in the Netherlands explored a novel approach to target discovery and compound screening for NAFLD, using prenatal human hepatocyte organoids as a model for mature human hepatocytes. Their work, published recently in Nature Biotechnology, identified compounds that could resolve a key NAFLD symptom in various genotypic backgrounds, revealed a prominent mechanism of action among effective compounds, and also led to a CRISPR-based screen for disease modulators and NAFLD risk genes.
Liver organoids – an attractive 3D model for the human liver
In Part 1 of this liver disease blog series, we summarized the cell types implicated in NASH and the limitations of currently used 2D cell models to study the disease, many of which also apply to therapeutic development for NAFLD. The solution to this problem might be edging closer with recent developments in liver organoid engineering.
In simple terms, liver organoids are miniature 3D cell models that are grown from various hepatic cells in vitro, typically liver sinusoidal endothelial cells, Kupffer cells, hepatic stellate cells, and hepatocytes. Liver organoids can either be cultured from isolated liver progenitor cells or from healthy donor- or patient-derived iPSCs whose self-renewal and differentiation properties make it possible to generate organoids that contain virtually any cell type. Unlike cell lines, organoid lines can be generated directly from wild-type/healthy donors of different genetic backgrounds, and the possibility to reproduce findings within the same organoid line allows for high reproducibility. With additional features such as greater proliferation capacity and longevity than 2D cell models, spatial and temporal liver-relevant gene expression, cell-cell communication and physical cell-cell interactions, and amenability to genetic manipulation, liver organoids allow for a much more complex and diverse experimental strategy than any 2D model.
Advances in 3D cell culture technology aid organoid organization into life-like 3D structures in a similar manner to in vivo organogenesis. When cultured in a certain ratio, with appropriate physical and biochemical cues, liver cells can self-organize into liver organoids that are similar to the whole liver that they are developed to mimic. Physical cues provide support for cell attachment and survival, while biochemical cues encompass molecules that modulate the signaling pathways governing proliferation, differentiation, and self-renewal, e.g., growth factors, enzymes and material that mimics the extracellular matrix (ECM). The most popular ECM material is Matrigel, a commercially-available reconstituted thin layer of ECM sheets known as the basement membrane, which consists mainly of collagen, entactin, perlecan, and laminin. As a side note, ethical issues about animal welfare, concerns about contamination from the animal source, and major supply challenges in recent years, mean that the use of animal-free alternatives will become necessary in the future.
Leveraging prenatal human hepatocyte organoids to model the mature human liver
Leveraging the potential of liver organoids to study NAFLD, a recent study led by Benedetta Artegiani (Hubrecht Institute at Royal Netherlands Academy of Arts and Sciences, and The Princess Maxima Center for Pediatric Oncology, The Netherlands) and Hans Clevers (Head of Pharma, Research and Early Development, Roche) deployed human-derived hepatocyte organoids to develop three in vivo-relevant human models that recapitulate different triggers for steatosis, which is characterized by an accumulation of lipids in the liver. They found that organoids developed from human hepatocytes (isolated from human fetal livers) could capture key features of mature human hepatocytes and early-stage NAFLD with relevance to the in vivo environment, including lipid metabolism profiles as evidenced by expression analysis of more than 400 genes known to be involved in lipid metabolism in humans.
Their findings, which were published in Nature Biotechnology earlier this year, describe a CRISPR-based target discovery approach that allowed them to identify disease-modifying genes and screen compounds for potential impact on key NAFLD symptoms, CRISPR-mediated gene knock-in and knock-out in human liver organoids (1).
Engineered liver organoids display steatosis with high phenotypic resolution
To model genetic predisposition to NAFLD, the researchers generated organoid models of two monogenic lipid disorders: familial hypobetalipoproteinemia (FHBL) and abetalipoprotenemia (ABL), and the well-characterized PNPLA3 I148M variant. They modeled these disorders using conventional CRISPR-Cas9 gene editing to disrupt the ApoB and MTP genes in hepatocytes, whose protein products play critical roles in packaging triglycerides into very low-density lipoprotein particles.
To generate PNPLA3 variant organoids, the team turned to prime editing, a search-and-replace genome-editing approach that can make small genomic alterations without introducing double-strand DNA breaks. Using two different donors, they generated a complete set of isogenic human PNPLA3 variant organoids, in which the I148M variant was introduced into one or both PNPLA3 alleles in hepatocytes to create a heterozygous or homozygous phenotype, respectively. They also created a homozygous PNPLA3 knockout organoid using conventional CRISPR-Cas9 gene editing. They found that the PNPLA3 genotype directly influenced lipid levels within hepatocytes, whereby all three engineered PNPLA3 variant organoids spontaneously accumulated lipids, the extent of which corresponded to the severity of the PNPLA3 mutation, with the knockout most severely affected. As expected, wild-type organoids did not accumulate lipids during homeostasis.
In parallel to generating the PNPLA3 variant and monogenic lipid disorder organoids, the researchers defined a free fatty acid (FFA)-induced steatosis model in wild-type hepatocyte organoids by ‘feeding’ them with a mixture of oleic acid and palmitic acid, two of the most abundant FFAs in the plasma pool. Other triggers of hepatic steatosis exist, but serum FFAs are the major contributor to the hepatic fatty acid pool and are increased in NAFLD patients. Thus, they were used in this study to mimic a Western diet. While wild-type organoids were able to process exogenous FFAs at low concentrations, they progressively accumulated lipids at higher concentrations. FFA exposure reproducibly induced similar levels of steatosis in organoids from different donors, and as steatosis progressed, organoid proliferation capacity became impaired.
By investigating the interplay between the FFA-induced steatosis and the NAFLD-predisposing genotypes of the engineered organoids, the researchers found that the organoid models recapitulated distinct and clinically-relevant biological mechanisms of steatosis with high phenotypic resolution. While the APOB- and MTTP- knockout organoids developed steatosis through accumulation of lipids derived from de novo lipogenesis – an endogenous pathway that turns excess glucose derived from the diet into specific fatty acids – steatosis in the FFA model was primarily driven by a surplus of FFAs. In contrast to studies in PNPLA3-knockout mice, where loss of PNPLA3 does not induce steatosis, the team found that PNPLA3 knockout in human organoids did lead to steatosis. This is important given that PNPLA3 depletion has been suggested as a potential therapeutic strategy for steatosis based on findings in mice, and illustrates the serious challenges posed by interspecies differences in currently-used disease models based on animals.
Drug screening reveals repression of de novo lipogenesis as common mechanism of action
Once the models had been characterized, the team tested them in screens for 17 candidate NAFLD drugs, to identify those that would directly impact steatosis and to probe whether the effect of a given drug depended on the steatosis trigger. The team focused on liver-targeted drugs that were either used in recent development programs or based on newly reported drug targets. To allow for comparison with the ApoB- and MTP-knockout organoids and to measure any effect of the drugs, steatosis was induced in the FFA model before exposure to the drugs. The team found that all liver organoid models responded to the drugs in a similar and mostly dose-dependent manner, and they could pinpoint a subset of drugs that were effective across all models.
Further analysis revealed repression of de novo lipogenesis (DNL) to be a common mechanism of action for the most effective drugs, which is noteworthy given that clinical studies have revealed increased rates of DNL as a prominent feature in NAFLD patients (2). Organoids with the PNPLA3 I148M variant genotype did not respond to all drugs in the same way as wild-type organoids, with only modest effects observed for certain drugs in clinical development for NAFLD. The authors suggest that these findings indicate that therapeutic effectiveness might be influenced by genetic risk factors for NAFLD, and highlight the usefulness of this model system for genotype-specific drug studies to target steatosis in early NAFLD.
FatTracer – A CRISPR-screening platform to identify new drug targets for NAFLD
To identify new putative therapeutic targets for steatosis, the team deployed the new ApoB and MTP-knockout organoid models to build Fat-Tracer, a scalable CRISPR-based screening platform that can screen loss-of-function mutations and evaluate their impact on steatosis. They screened a library of 35 lipid metabolism-related genes, as well as PNPLA3 and some of the effective drug targets identified in the drug screening experiments.
To set up FatTracer, single cells from ApoB and MTP-knockout organoid lines were transfected with plasmids encoding Cas9 and a target-specific guide RNA to create loss-of-function mutations as well as a selection marker to enrich for transfected cells. Using this setup, the team targeted every gene in the library independently and cultured the transfected cells into organoids, one for each target. Loss-of-function organoids were phenotypically compared to mock-transfected organoids in fat staining and quantitative experiments, and genotypically characterized to confirm genotype-phenotype correlation.
Overall, the targets identified during the drug screening experiments were also identified with FatTracer, and the screen also revealed the functional importance of PNPLA3 in preventing steatosis. The team notes that FatTracer evaluates gene-steatosis relationships in the context of an existing steatosis state, and can reveal reducing as well as aggravating effects, and that the screen can be applied genome-wide, and to study the healthy to disease state if applied to wild-type organoids in parallel.
Finally, the FatTracer studies revealed a previously unknown yet critical role for the FADS2 gene (fatty acid desaturase 2) in NAFLD, whereby its disruption exacerbated steatosis. Importantly, this role of FADS2 would not have been uncovered had the research relied on animal models. Subsequent overexpression of FADS2 in organoids resulted in almost complete clearance of existing steatosis, and protected wild-type organoids from developing steatosis when exposed to FFAs at the same doses used in the drug screening experiments. These findings suggest that by reducing total TAG content, FADS2 may be a potential new therapeutic target for NAFLD, despite the fact that mouse studies have yielded conflicting results regarding its role in steatosis.
This work represents a significant step forward in organoid technology that can eventually help to address many questions about the mechanisms of liver diseases such as NAFLD and NASH in a human context. It also illustrates the serious discrepancies that can arise between rodent and human model studies and highlights the importance of taking genotypic susceptibility to disease into account when screening potential new drugs.
- Hendriks D, Brouwers JF, Hamer K, et al. Engineered human hepatocyte organoids enable CRISPR-based target discovery and drug screening for steatosis. Nat Biotechnol. 2023 Feb 23. doi: 10.1038/s41587-023-01680-4. Epub ahead of print. PMID: 36823355.
- Lambert JE, Ramos-Roman MA, Browning JD, Parks EJ. Increased de novo lipogenesis is a distinct characteristic of individuals with nonalcoholic fatty liver disease. Gastroenterology. 2014 Mar;146(3):726-35. doi: 10.1053/j.gastro.2013.11.049.
- Commentary on the reference (Hendriks et al., Nat Biotechnol. 2023 Feb 23.) Wang Y, Xu H. CRISPR-edited hepatic organoids as drug screening platform for non-alcoholic fatty liver disease. Hepatobiliary Surg Nutr 2023;12(4):593-594. doi: 10.21037/hbsn-23-247
Karen O’Hanlon Cohrt is an independent Science Writer with a PhD in biotechnology from Maynooth University, Ireland (2011). After her PhD, Karen relocated to Denmark where she held postdoctoral positions in mycology and later in human cell cycle regulation, before moving to the world of drug discovery. Karen has been a full-time science writer since 2017, and has since then held numerous contract roles in science communication and editing spanning diverse topics including diagnostics, molecular biology, and gene therapy. Her broad research background provides the technical know-how to support scientists in diverse areas, and this in combination with her passion for learning helps her to keep abreast of exciting research developments as they unfold. Karen is currently based in Ireland, and you can follow her on Linkedin here.