Gene Editing of Pluripotent Stem Cell-Derived Hepatic Cells for Liver Disease Modeling and Therapeutic Development

Abstract

The growing demand for physiologically relevant human liver models has driven significant progress in generating hepatic cells and organoids derived from pluripotent stem cells. These regenerative cell sources serve as powerful platforms for elucidating the mechanisms underlying liver diseases and for evaluating drug responses under human-relevant conditions. Moreover, they hold tremendous promise as cell-based therapeutics for various hepatic disorders. The utility of these regenerative cell technologies is further expanded when combined with gene-editing techniques, which enable precise modeling of pathogenic variants and targeted correction of disease-associated mutations. Gene editing can also be leveraged to enhance the functionality and therapeutic potential of regenerative hepatocyte products. In this review, we summarize recent advances at the interface of gene editing and hepatic cell regeneration, emphasizing their applications in genetic disease modeling, therapeutic gene correction, drug testing, and cell-based therapies for liver disorders. We also provide an overview of major gene-editing tools and practical guidance for implementing them in pluripotent stem cells-based regenerative workflows, concluding with future perspectives on the integration of gene editing and regenerative hepatocyte technologies.

INTRODUCTION

Genetic background profoundly shapes the onset and progression of many diseases. Specific gene variants can directly cause certain diseases, whereas others modulate susceptibility. In practice, genome-wide association studies (GWASs) have identified numerous loci that are directly and indirectly related to numerous diseases (Chen et al., 2021, 2023; Heyne et al., 2023). In parallel, the integration of genetic evidence into therapeutic decision-making is increasingly recognized as essential for effective disease management (King et al., 2019; Minikel and Nelson, 2025; Minikel et al., 2024). A prerequisite for genotype-guided treatment is determining the functional role of individual variants in defined disease contexts. To this end, gene-editing methods, most importantly clustered regularly interspaced short palindromic repeat (CRISPR)-Cas systems, can be employed to enable precise genetic and epigenetic perturbations in living cells.

Another essential requirement for translational progress is access to in vitro model systems that faithfully recapitulate in vivo tissue functions. In this context, stem cell-derived tissues and organoids have emerged as robust platforms for studying human biology under physiologically relevant conditions. For liver research, hepatic cells and organoids can be generated from primary liver samples, adult stem cells (ASCs), or pluripotent stem cells (PSCs). Among these, PSC-derived models are particularly versatile because of the unlimited proliferative capacity of PSCs and refined differentiation protocols that yield hepatic cells. Although the liver has regenerative capacity, primary hepatocytes are difficult to culture and maintain in vitro, which constrains their utility for gene editing and clonal isolation (Kaur et al., 2023; Luce et al., 2021; Wu et al., 2025a). By contrast, gene editing has been extensively optimized in PSCs due to their therapeutic potential, enabling efficient genetic modifications and clonal selection to obtain genetically well-defined cell lines (Blassberg, 2022; Hockemeyer and Jaenisch, 2016).

Gene-edited, PSC-derived hepatic cells provide powerful platforms to study the genetic basis of liver diseases either by introducing disease-relevant mutations into embryonic stem cells (ESCs) or induced pluripotent stem cells (iPSCs) from health donors. Alternatively, gene editing can be employed to correct pathogenic mutations in patient-derived iPSCs. Because these approaches preserve an isogenic background, they enable precise dissection of variant-specific disease mechanisms. Insights gained from such in vitro studies can inform in vivo gene-editing strategies, especially given that the liver is a key target organ amenable to therapeutic gene editing. In addition, PSC-derived hepatocytes can serve as versatile models for evaluating the safety and efficacy of various drug candidates, spanning from small molecules to biologics.

Beyond disease modeling and drug screening, gene-corrected, patient-derived iPSCs and their hepatic derivatives offer avenues for autologous cell therapies. Standardized and well-characterized PSC lines may also enable consistent manufacturing of allogenic cell therapies. In contrast, hepatic tissues derived from primary cells often suffer from batch-to-batch variability that complicates standardized production and quality control (Chang et al., 2021). Given the rising global burden of liver diseases (Gan et al., 2025; Wu et al., 2025b), PSC-derived and gene-edited cell therapies constitute compelling modalities for treatment and, in some cases, potential cure.

In this review, we provide an overview of recent advances in gene-editing technologies combined with PSC-derived hepatic cells. We begin by summarizing the current methods for generating hepatic cells and organoids. As the primary objective of this review is to guide researchers in choosing the most appropriate gene-editing tools for their work, we offer a detailed discussion of existing gene-editing technologies. We then explore how gene editing has deepened our understanding of disease genetics and facilitated the development of novel therapeutic modalities. Finally, we address key considerations for performing gene editing in PSCs and highlight future directions in this rapidly evolving field.

GENERATION OF PLURIPOTENT STEM CELL-DERIVED HEPATIC CELLS AND ORGANOIDS

The liver performs diverse functions, including the control of glucose and lipid homeostasis, detoxification of blood, secretion of essential plasma proteins, bile production, and innate immune response. These diverse functions are carried out primarily by hepatocytes, the dominant parenchymal cell type of the liver (Gong et al., 2022; Nagarajan et al., 2019; Schulze et al., 2019), which account for approximately 80% of liver volume (Liu et al., 2024a; Schulze et al., 2019). Accordingly, much of the efforts in liver cell generation has focused on producing functional hepatocytes.

Multiple strategies have been developed to derive hepatocytes from various cell sources, including ASCs, PSCs, mesenchymal stem cells, and fetal hepatocytes. Among these, PSCs provide a virtually unlimited source owing to their robust proliferative capacity and potential to differentiate toward the hepatic lineage, making them highly valuable for both basic research and regenerative therapies. High proliferative potential also makes PSCs amenable to gene editing, especially precise gene knockin followed by clonal isolation, which is impossible in non-proliferative cells. Consequently, significant attention has focused on generating of PSC-derived hepatic cells.

Protocols for directed differentiation of PSC toward hepatic cells recapitulate key stages of liver development using timed exposure to cytokines, growth factors, and small molecules. Although specific conditions vary across laboratories, most common protocols comprise three phases: definitive endoderm induction, hepatic progenitor specification, and hepatocyte differentiation (Fig. 1A). Definitive endoderm is typically induced with activin A. Activation of WNT3 signaling by a small molecule CHIR99021 and inhibition of phosphatidylinositol 3-kinase (PI3K) signaling by a small molecule LY294002 can enhance endodermal differentiation. Subsequently, hepatic progenitors are induced from the endoderm using bone morphogenetic protein 4 (BMP4) and fibroblast growth factor 2 (FGF2), often under hypoxic conditions. Hepatocyte growth factor (HGF) can be added to promote proliferation of hepatic progenitors while suppressing genes involved in pancreatic lineage development. As hepatic progenitors can generate either hepatocytes or bile ducts, a combination of specific protein factors and small molecules (e.g., HGF, oncostatin M, and glucocorticoid dexamethasone) is employed to complete differentiation into hepatic cells (Fig. 1A) (Blaszkiewicz and Duncan, 2024; Graffmann et al., 2022; Ju et al., 2024; Luo et al., 2023). Numerous similar, yet slightly different, protocols have been employed for directed differentiation, and the resulting cells are variably referred to as PSC-derived hepatocytes or PSC-derived hepatocyte-like cells. To avoid ambiguity, this review consistently uses the term PSC-derived hepatic cells throughout.

Fig. 1.

Generation of pluripotent stem cell (PSC)-derived hepatic cells and hepatic organoids. (A) Schematic of the protocol for generating PSC-derived hepatic cells. (B) Exposure of PSC-derived hepatic cells to additional specification factors and environmental cues promotes the formation of hepatic organoids. phosphatidylinositol 3-kinase (PI3K), fibroblast growth factor 2 (FGF2), hepatocyte growth factor (HGF), oncostatin M (OSM), dexamethasone (DEX).

The majority of reports have generated PSC-derived hepatic cells using 2-dimensional (2D) differentiation methods and demonstrated their utility. More recently, 3-dimensional (3D) culture platforms have been developed to yield hepatic organoids that better recapitulate liver architecture and physiological functions. Furthermore, advanced liver organoids incorporating multiple parenchymal liver cell types and displaying liver-like structural features can also be derived from PSCs, providing more physiologically relevant models for disease research (Ju et al., 2024; Luo et al., 2023). Such organoids are often generated by further culturing PSC-derived hepatic cells under defined environmental cues such as extracellular matrices, micropatterned substrates, and additional small molecules and protein factors (Fig. 1B) (Ju et al., 2024; Reza et al., 2025; Weng et al., 2023; Yuan et al., 2023). Despite progress, standardized and widely reproducible protocols for liver organoid generation remain to be established, underscoring the need for continued development and optimization.

When combined with gene-editing technologies, PSC-based hepatic cell generation holds substantial translational potential. Gene editing is performed at the undifferentiated PSC stages, followed by clonal expansion and directed differentiation into hepatic cells (Fig. 1A) (Kim et al., 2025b). The highly proliferative nature of PSCs makes them well suited for gene-editing approaches, including precise gene knockin via homologous recombination. In contrast, growth-arrested cells are refractory to such precise gene editing. Moreover, clonal isolation is feasible only in dividing cells, further supporting the advantage of editing at the PSC state. Overall, coupling gene editing with clonal selection in PSCs, followed by their differentiation into hepatic cells, represents a practical and powerful strategy to interrogate the genetic basis of liver disease and develop new regenerative therapies. In the following sections, we present gene-editing strategies applicable to the generation of gene-modified PSCs and their hepatic derivatives, highlighting both their applications and associated technical considerations.

GENE-EDITING TECHNOLOGIES

Gene-editing technologies are now indispensable for modeling genetic disease and correcting pathogenic variants. Early targeted gene-editing platforms, such as zinc finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs), demonstrated significant value due to their programmability. However, their applications have been limited by the technical challenges associated with designing and producing custom nucleases for each new genomic target, primarily owing to the repetitive sequences within these proteins (Gaj et al., 2013). The advent of CRISPR-based technologies revolutionized the field by making gene editing more accessible, even to researchers with limited molecular biology expertise. A major advantage of CRISPR lies in its simplicity. Targeting a new genomic site required only the design and generation of a new guide RNA (gRNA), a process that is straightforward and readily scalable. For plasmid-based editing machinery delivery, generating a new gRNA involves a simple one-step cloning procedure. When delivering purified Cas protein-gRNA complexes, new gRNAs can be synthesized by in vitro transcription or purchased directly from commercial vendors.

In this section, we review contemporary gene-editing modalities, focusing on CRISPR-Cas nucleases, base editors, prime editors, and epigenome editors. These tools are broadly applicable to hepatocyte research due to their versatility and ease of implementation. We also provide practical guidelines for researchers aiming to perform PSC-based disease modeling and gene correction.

Cas9 nucleases

Cas9 generates a double-strand break (DSB) at a specified genomic locus with the aid of gRNAs that base-pairs with the target sequence. Once a DSB is introduced, endogenous DNA-repair pathways resolve the break via non-homologous end-joining (NHEJ), microhomology-mediated end joining (MMEJ), or homology-directed repair (HDR) pathways.

Under NHEJ, small insertions or deletions (indels) are introduced around the DSB, frequently disrupting gene function by introducing frameshifts, premature stop codons, or aberrant splicing sites (Fig. 2A). MMEJ can be engaged when short microhomology sequences flank the DSB. Annealing of these homologous sequence produces small-sized deletions that likewise yield loss-of-function alleles (Fig. 2A). In contrast, HDR uses an exogenous donor DNA template containing left and right homology arms that match the genomic sequence flanking the cut site, with the desired edit positioned between these arms. The homology arms direct the precise integration during the DSB repair, enabling accurate knockin of defined sequences (Fig. 2A). Because HDR installs specified nucleotide changes, it allows predictable gene knockout (e.g., targeted insertion of a premature stop codon) unlike the NHEJ. More importantly, HDR can be employed to introduce single nucleotide polymorphisms (SNPs). Because many disease-associated variants are SNPs, HDR is the preferable pathway for gene editing-based disease modeling.

Fig. 2.

Overview of gene-editing approaches applicable to PSC-derived hepatic cells. (A) Cas9-induced double-strand breaks repaired by NHEJ, MMEJ, or HDR result in knockout or knockin cells. (B) Cytosine base editors mediate C-to-T conversions. (C) Adenine base editors mediate A-to-G conversions. (D) Prime editors enable small base substitutions, insertions, or deletions. (E) CRISPR activation and (F) CRISPR interference are employed to upregulate or repress endogenous gene expression. non-homologous end-joining (NHEJ), microhomology-mediated end joining (MMEJ), homology-directed repair (HDR), cytosine base editor (CBE), adenine base editors (ABE), prime editor (PE).

Despite these advantages, HDR is far less efficient than the NHEJ (Lim et al., 2020). Accordingly, NHEJ remains the most effective and straightforward route for generating knockout models. Since NHEJ-mediated repair is imprecise and uncontrollable, investigators commonly derive multiple clonal PSC lines and sequence them to finally isolate knockout alleles. When precise knockin is required, HDR must be used even though its efficiency is often low. To address this problem, a variety of HDR-enhancing strategies have been developed, including transient use of NHEJ-blocking small molecules, Cas9-donor DNA conjugation, temporary p53 inactivation, and design of optimized donor DNA constructs (Jin et al., 2025; Liao et al., 2024). Because HDR efficiency in PSCs is even lower than in many transformed cell lines, empirically testing several HDR-enhancing tools is recommended to expedite recovery of edited clones.

Selecting appropriate type of donor DNA is important for maximizing HDR efficiency while maintaining cellular fitness. Options include linear double-stranded DNA (dsDNA), plasmid DNA, single-stranded oligodeoxynucleotide (ssODN), and adeno-associated virus serotype 6 (AAV6). For small edits such as SNP corrections, ssODNs are widely used due to facile chemical synthesis and comparatively low cytotoxicity (Liao et al., 2024). For larger insertions, dsDNA or plasmid donors are commonly employed because of their availability through standard molecular cloning techniques. AAV6 donors have achieved high knockin efficiency in therapeutically relevant cell types such as iPSCs and hematopoietic stem cells, although their production is more complicated compared to other donor types (Liao et al., 2024).

Cas9 from Streptococcus pyogenes (SpCas9) is the most widely used nuclease owing to its robust activity across diverse cell types including PSCs. Its deployment can be constrained, however, by the requirement for a 5′-NGG-3′ protospacer adjacent motif (PAM) and by potential off-target activity. If a suitable PAM is absent, Cas12 nucleases recognizing T-rich PAMs offer an alternative (Badon et al., 2024; Wu et al., 2024). When off-target propensity is a concern, higher-specificity SpCas9 variants (e.g., eSpCas9, Correct-Cas9) can be used (Slaymaker et al., 2016; Sung et al., 2025). Cas9 from Staphylococcus aureus (SaCas9) or Cas12 can may also mitigate off-target effects by virtue of distinct PAM requirements and consequently distinct target sequences (Zhou and Yao, 2023). Specificity can be further improved with a double-nicking strategy, in which a Cas9 nickase (nCas9) and paired gRNAs create staggered DSBs. Because single nicks are typically repaired with high fidelity back to the original sequence, off-target nicks would not produce mutations. Simultaneously, the requirement for two correctly spaced guides effectively doubles the length of recognition sequence and greatly enhances the editing specificity (Ran et al., 2013).

Whereas Cas9 produces blunt end DSBs, Cas12 generates DSBs with staggered ends, which in some contexts can increase HDR efficiency (Khan and Sallard, 2023). Similarly, double-nicking strategy yields staggered ends and may further facilitate HDR (Ran et al., 2013). When precise knockin is desirable, the applicability of these methods should be evaluated in combination with complementary HDR-enhancement methods to identify the most effective conditions for a given locus and cell line.

Base editors and prime editors

Despite its versatility, HDR is intrinsically inefficient, which can limit its broad application. Moreover, the Cas9-induced DSBs required for HDR can be toxic in PSCs and raise concerns about off-target gene editing (Ihry et al., 2018; Lim et al., 2022). These limitations have motivated the development of DSB-free precision editors such as base editors and prime editors that install defined sequence changes without creating DSBs.

Cytosine base editors (CBEs) consist of an nCas9 fused to a cytidine deaminase and uracil DNA glycosylase inhibitors (UGIs). The deaminase domain converts cytosine to uracil, which is ultimately resolved to thymine during DNA repair. The UGI domain inhibits the excision of DNA-bound uracil, thereby enhancing editing efficiency (Fig. 2B) (Anzalone et al., 2020). Adenine base editors (ABEs) comprise an nCas9 fused to a deoxyadenosine deaminase that converts deoxyadenosine to inosine, which is interpreted as guanosine during DNA repair (Fig. 2C) (Anzalone et al., 2020). In these base editors, nCas9 that nicks the gRNA-bound strand (e.g., SpCas9 D10A variant) is frequently used to direct subsequence DNA repair to proceed with the base-exchange strand, thereby increasing editing efficiency. Because CBEs and ABEs catalyze precise base substitutions, they are valuable tools for modeling SNPs found in many diseases. However, they cannot perform transversions between purines and pyrimidines (e.g., A to C or G to T), nor can they introduce insertions or deletions.

To broaden the editable mutation space, prime editors have been developed that enable the introduction of diverse genetic alterations, including all types of base transversions as well as small-sized deletions and insertions. Prime editors consist of an nCas9 (e.g., SpCas9 H840A variant) fused to a reverse transcriptase and use a prime editing gRNA (pegRNA) in place of standard gRNA. The pegRNA features a 3′ extension containing a primer binding site (PBS) and a reverse transcription template (RTT) that encodes the desired edit (Fig. 2D). Following nicking of pegRNA-unbound non-target strand by nCas9, the exposed single-stranded DNA region anneals to the PBS on the pegRNA, which primes reverse transcription. The reverse transcriptase then synthesizes a complementary DNA (cDNA) strand from the RTT, incorporating the precise edit into the cDNA. Subsequent DNA repair from the cDNA strand can be facilitated by nicking the opposite DNA strand, ultimately generating precisely edited cells (Fig. 2D) (Getachew et al., 2025).

Although base editors and prime editors avoid DSB-associated toxicity and can afford highly precise edits, their activity varies substantially by locus, necessitating through optimization efforts. For disease modeling, the target variant is predetermined and viable gene-editing strategies are limited. Therefore, we recommend empirically testing all precise editing routes available (i.e., HDR-mediated knockin, base editors, and prime editors) to identify the most effective approach for the locus and PSC line of interest. These gene-editing modalities are now widely used to generate genetically defined PSCs that can be differentiated into stem cell-derived tissues for mechanistic studies and translational applications.

CRISPR-based epigenetic modifiers

CRISPR-Cas epigenetic modification technologies, including CRISPR activation (CRISPRa) and CRISPR interference (CRISPRi), enable regulation of endogenous gene expression without altering underlying DNA sequence, thereby providing tools for reversible and temporally controllable gene modulation (Fig. 2E, 2F) (Bendixen et al., 2023). These approaches are particularly valuable for basic biological research since they enable highly specific control over gene expression compared to conventional approaches such as RNA interference (Smith et al., 2017; Stojic et al., 2018). In addition, RNA-targeting Cas proteins (e.g., Cas13a) extend the gene-editing toolkit to enable transcript-level modulation (Shi and Wu, 2024).

As gene-editing methodologies continue to advance, studies leveraging gene-edited, stem cell-derived tissues are poised to become increasingly widespread. In the following sections, we examine applications of gene-editing tools for investigating the genetic basis of hepatopathies, and explore the regenerative medicine potential of gene-edited hepatic cells for treating diverse hepatic disorders. In particular, we outline technical aspects in each context to assist readers in selecting the appropriate gene-editing tools for their research.

GENE EDITING FOR MODELING AND CORRECTING GENETIC LIVER DISEASES

Here, we provide a comprehensive overview of liver disease models established through the integration of gene editing and hepatocyte-generation technologies (Table 1, Fig. 3). These studies have facilitated a detailed understanding of the molecular mechanisms underlying diverse hepatopathies and have informed the developments of next-generation gene-editing therapies aimed at restoring liver function. Notably, the liver is amenable to in vivo delivery of gene-editing machineries (Simoni et al., 2024), thereby enabling direct translation of preclinical findings into clinical applications. Indeed, recent clinical trials have demonstrated both the feasibility and therapeutic potential of in vivo gene editing for the treatment of liver diseases (Musunuru et al., 2025; Wang et al., 2021).

Table 1.

Applications of gene editing in pluripotent stem cell-derived hepatic cells

Disease modeling and disease correction
Diseases or targets	Tissue types investigated	Gene-editing strategies	Notes	Ref.
GWAS-identified SNPs linked to lipid metabolism	ESC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced SNPs identified from GWAS.	Gene editing enabled investigation into the impact of SNPs on lipid metabolism in hepatic cells.	(Pashos et al., 2017)
Familial hypercholesterolemia	iPSC-derived hepatic cells	HDR with a paired nCas9 and ssODN introduced 3-bp insertion into the LDLR gene.	Gene-edited hepatic cells displayed normal LDLR function.	(Omer et al., 2017)
Familial hypercholesterolemia	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor introduced hepatocyte-specific LDLR expression cassette at the AAVS1 safer harbor locus.	Gene-edited hepatic cells displayed normal LDLR function.	(Caron et al., 2019)
Non-alcoholic fatty liver disease	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced a variant (I148M) in the PNPLA3 gene or generated PNPLA3 knockout cells.	Gene-edited cells displayed the disease phenotype such as fat accumulation.	(Tilson et al., 2021)
Non-alcoholic fatty liver disease	iPSC-derived hepatic cells and organoids	HDR with Cas9 nuclease and ssODN introduced precise gene knockout at the TRIB1 gene.	Gene-edited 3D hepatic organoids could be used to study the biology of TRIB1 related to non-alcoholic fatty liver disease.	(Abbey et al., 2020)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	HDR with TALEN and a plasmid donor corrected the Z allele in the SERPINA1 gene while introducing a selection marker. The marker was then removed by piggyBac transposase.	Gene-corrected cells exhibited functional restoration of the disease phenotype.	(Choi et al., 2013)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN corrected the Z allele in the SERPINA1 gene.	Gene-corrected cells displayed normal AAT production.	(Kaserman et al., 2020)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced or corrected the Z allele in the SERPINA1 gene.	Gene editing enabled accurate modeling and correction of the disease phenotype.	(Gil et al., 2025)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN corrected the Z allele in the SERPINA1 gene. Use of dual ssODNs allowed the generation of heterozygous allele.	Precise gene editing enabled investigation into the function of both heterozygous and homozygous mutants.	(Kaserman et al., 2022)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	Adenine base editing corrected the Z allele in the SERPINA1 gene.	Base editing restored the normal alpha-1 antitrypsin production.	(Werder et al., 2021)
Urea cycle disorder	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced mutations in the OTC gene (D175V) or the ASS1 gene (G390R).	Urea secretion was decreased by the OTC mutation, but not by the ASS1 mutation.	(Łukasiak et al., 2024)
Urea cycle disorder	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN corrected the pathogenic mutation (R129H) in the OTC gene.	Gene-corrected cells displayed enhanced urea production.	(Zabulica et al., 2021)
Urea cycle disorder	iPSC-derived hepatic cells	HDR with a paired nCas9 and plasmid donor knocked in an arginase expression cassette in the middle of the HPRT locus.	Knock-in at the HPRT locus facilitated the efficient isolation of edited cells, while successfully restoring arginase function and urea production.	(Lee et al., 2016)
Glycogen storage disease IV	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced a variant of uncertain significance (I694N) in the GBE1 gene.	The gene-edited cells displayed the disease phenotype, revealing the effect of variant of uncertain significance on hepatocytes.	(Naito et al., 2024)
Cholestasis	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced a pathogenic variant (D482G) in the ABCB11 gene.	The mutant hepatic cells displayed impaired bile acid efflux.	(Łukasiak et al., 2024)
Alagille syndrome	iPSC-derived hepatic organoids	HDR with a paired nCas9 and plasmid donor introduced a pathogenic JAG1 mutation (C829X) and selection markers. The markers were then removed by piggyBac transposase.	Gene editing was employed to precisely model and correct the disease, demonstrating the feasibility of gene editing-based therapies.	(Guan et al., 2017)
Hepatocyte carcinomas	ESC-derived hepatic cells	Cas9 nuclease-mediated gene knockout and lentiviral oncogene delivery were performed.	Tumorigenic events previously observed in hepatocellular carcinoma cohorts were recapitulated in engineered hepatic cells.	(Zhang et al., 2024)
Hepatitis B virus infection	ESC-derived hepatic cells	Adenine base editing introduced a variant (S267F) in the NTCP gene.	Homozygous variants were resistant to infection, while heterozygous variants were fully susceptible to viral infection.	(Uchida et al., 2021)
Hepatitis C virus infection	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor introduced a hepatocyte-specific LDLR expression cassette at the AAVS1 safer harbor locus.	Gene editing allowed the investigation into the role of LDLR in viral entry, replication, and infection.	(Caron et al., 2019)
Drug screening, drug action study, and drug metabolism study
Diseases or targets	Tissue types investigated	Gene-editing strategies	Notes	Ref.
Mitochondrial DNA depletion syndromes	iPSC-derived hepatic cells	Cas9 nuclease-mediated knockout of the DGUOK and the RRM2B gene was performed.	The gene-edited cells recapitulated the disease phenotype and were successfully applied to drug screening.	(Jing et al., 2018)
Alpha-1 antitrypsin deficiency	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced Z allele in the SERPINA1 gene.	The effect of chemical drugs and siRNA drugs was investigated using the disease model.	(Gil et al., 2025)
Wilson’s disease	ESC-derived hepatic cells	HDR with Cas9 nuclease and ssODN introduced a disease-causing variant (R778L) in the ATP7B gene.	The model correctly recapitulated the disease phenotype and could be used to test drug candidates.	(Kim et al., 2020a)
Drug metabolism testing	iPSC-derived hepatic cells	Cas9 nuclease-mediated knockout of the CYP2C19 gene was performed.	The knockout cells served as a model of poor drug metabolism.	(Deguchi et al., 2019)
Drug metabolism testing	iPSC-derived hepatic cells	Cas9 nuclease-mediated knockout of the CYP3A4 gene was performed.	The knockout cells served as a model of poor drug metabolism.	(Deguchi et al., 2021)
Drug metabolism testing	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor introduced neoR and eGFP genes at the CYP3A4 locus.	The reporter knockin enabled the enrichment of cells with higher CYP3A4 activity and enhanced drug metabolism.	(Takayama et al., 2018)
Drug metabolism testing	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor introduced an mCherry gene at the CYP1A1 locus.	Image-based live-cell drug screening was performed using the reporter cells to identify CYP1A1-modulating compounds.	(Kim et al., 2020b)
Regenerative cell therapies for liver diseases
Diseases or targets	Tissue types investigated	Gene-editing strategies	Notes	Ref.
Wilson’s disease	iPSC-derived hepatic cells	HDR with Cas9 nuclease and ssODN corrected a disease-causing variant (R778L) in the ATP7B gene.	Correction of a single allele was sufficient to achieve therapeutic effects. The transplanted cells were effective in the diseased mouse model.	(Wei et al., 2022)
Urea cycle disorder	Mouse iPSC-derived hepatic cells	HDR with TALEN and donor DNA restored exon 7 and exon 8 in the Arg1 gene.	The gene-edited hepatic cells showed partial therapeutic effects, which is ascribed to suboptimal transplantation protocol.	(Sin et al., 2018)
Hemophilia B	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor knocked in F9 expression cassette along with a selection marker into the AAVS1 safer harbor.	The gene-edited hepatic cells secreted factor IX during a short-term transplantation.	(Lyu et al., 2018)
Hemophilia B	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor knocked in F9 expression cassette at the liver-specific APOC3 locus.	The gene-edited hepatic cells secrete factor IX and were successfully implanted into mice model as cell sheets.	(Bayarsaikhan et al., 2025)
Further applications
Diseases or targets	Tissue types investigated	Gene-editing strategies	Notes	Ref.
Liver injury	iPSC-derived liver organoids	CRISPR activation of CYP3A4 was employed along with lentiviral overexpression of ATF5 and PROX1.	Combined CRISPR activation and lentiviral overexpression enabled the formation of highly functional liver organoids, as demonstrated in a mouse model of liver injury.	(Velazquez et al., 2021)
Refining hepatic differentiation processes	iPSC-derived hepatic cells	HDR with Cas9 nuclease and plasmid donor knocked in a Venus report at the ALB locus. Genome-wide CRISPR knockout screen was performed.	The albumin-Venus cell line served as a reporter for hepatic differentiation. The CRISPR screen identified genetic factors improving the differentiation.	(Li et al., 2018)
Refining organoid generation processes	iPSC-derived hepatic cells and liver organoids	HDR with Cas9 nuclease and donor DNA knocked in fluorescent reporter genes at the AAVS1 safer harbor.	Fluorescent labeling of different liver cell types in liver organoids allowed facile monitoring of the organoids in vitro and in vivo.	(Reza et al., 2025)

Fig. 3.

Workflow for the generation of patient-derived iPSCs, gene correction, and subsequent differentiation into PSC-derived hepatic cells and organoids, along with representative applications of these cells in studying genetic disease mechanisms, testing drug actions, and advancing regenerative medicine. alpha-1 antitrypsin (AAT), cytochrome P450 (CYP).

Hypercholesterolemia

The liver is the central organ responsible for lipid metabolism. Thus, blood lipid levels directly reflect the lipid-metabolizing capacity of hepatocytes. To investigate the functional link between genetic variants and hepatic lipid metabolism, Pashos et al. employed Cas9 nuclease- and ssODN-mediated HDR to model genetic alterations identified in GWAS analyses (Pashos et al., 2017). For instance, they introduced an SNP variant (rs2277862) into ESCs, which were subsequently differentiated into hepatic cells. This study revealed that the SNP reduces the expression of the nearby CPNE1 gene in hepatic cells. When the corresponding mutation was introduced into a mouse model, Cpne1 expression in the liver was similarly reduced compared to wild-type controls. Based on these key findings, they were able to establish CPNE1 as a lipid-associated functional gene that decreases high-density lipoprotein cholesterol levels in blood. This work highlights the utility of gene-edited, PSC-derived hepatic cells for dissecting the genetic basis of hepatocyte biology (Pashos et al., 2017).

Familial hypercholesterolemia (FH), characterized by elevated blood cholesterol and an increased risk of cardiovascular diseases, can be caused by mutations in the gene encoding low-density lipoprotein receptor (LDLR). To evaluate the feasibility of gene-editing for FH, Omer et al. corrected patient-derived iPSC using nCas9-based double-nicking strategy with ssODN donors to harness HDR pathway (Omer et al., 2017). This approach repaired a 3-bp deletion in the LDLR gene, restoring the wild-type sequence. The corrected iPSCs were differentiated into hepatic cells that exhibited normalized LDLR-mediated low-density lipoprotein (LDL) uptake. This proof-of-concept demonstrates the therapeutic potential of either in vivo gene editing or autologous cell therapies for FH (Omer et al., 2017).

Similarly, Caron et al. generated iPSCs from FH patients harboring a homozygous premature stop codon (Q12X) in the LDLR gene (Caron et al., 2019). To restore LDLR function, they knocked in an LDLR expression cassette driven by hepatocyte-specific apolipoprotein A2 promoter into the AAVS1 safe harbor locus along with a constitutive puromycin resistance gene, which facilitated efficient isolation of correctly edited iPSC clones. This safe harbor-based strategy offers an alternative approach for gene editing when precise base exchange at the endogenous locus is challenging. Notably, the edited iPSC-derived hepatic cells displayed restored LDLR expression and LDL uptake, supporting the use of safe harbor integration as a viable therapeutic strategy for developing cell therapies for FH (Caron et al., 2019).

Non-alcoholic fatty liver disease

Non-alcoholic fatty liver disease (NAFLD), characterized by excessive lipid accumulation in hepatocytes, is one of the leading causes of the chronic liver disease worldwide. Despite its substantial health burden, the molecular mechanisms underlying NAFLD remain incompletely understood. Building on recent GWAS findings that have identified potential genetic risk factors, Tilson et al. investigated the impact of the I148M mutation in the patatin-like phospholipase domain-containing protein 3 (PNPLA3) gene (Tilson et al., 2021). Using Cas9 nuclease- and ssODN-mediated HDR, they generated iPSCs carrying this variant. As HDR is always accompanied by NHEJ, PNPLA3 knockout iPSCs were also obtained during the editing process. Differentiation of these iPSC lines into hepatic cells enabled functional interrogation of the mutation under isogenic backgrounds. The study demonstrated that PNPLA3 knockout cells accumulated more lipids than wild type controls. In addition, the knockout caused diminished xenobiotic metabolism of hepatic cells, rendering them more susceptible to xenobiotic-induced toxicity. The I148M variant conferred a loss-of-function phenotype to predispose variant carriers to steatosis, producing intermediate phenotypes between the wild type and knockout cells (Tilson et al., 2021).

Abbey et al. explored the role of Tribbles homolog 1 (TRIB1), a gene identified through GWAS as being associated with NAFLD and plasma lipid levels (Abbey et al., 2020). In the absence of patient-derived iPSCs with homozygous disruption, they introduced a premature stop codon into the TRIB1 gene in normal iPSCs using Cas9 nuclease- and ssODN-mediated HDR. TRIB1 is known to post-translationally regulate CCAAT/enhancer-binding protein α (C/EBPα), and liver-specific Trib1 knockout in mice causes increased C/EBPα expression and elevated hepatic triglycerides (Bauer et al., 2015). Consistently, TRIB1 knockout iPSC-derived 3D hepatic organoids recapitulated the lipid-related phenotypes observed in vivo. In contrast, 2D iPSC-derived hepatic cells showed reduced expression of late-stage hepatic and lipogenesis markers, indicating the limitations of 2D cultures. This study not only highlights the value of gene editing for functional interrogation of poorly characterized genes but also emphasizes the importance of appropriate cell culture systems for faithfully modeling human liver diseases.

Alpha-1 antitrypsin deficiency

Alpha-1 antitrypsin deficiency is primarily caused by mutations in the SERPINA1 gene, which encodes the alpha-1 antitrypsin (AAT) protein. The most prevalent pathogenic variant, the Z allele (E342K), leads to protein misfolding and impaired function, thereby compromising its tissue-protective roles. Both the liver and lungs are highly susceptible organs critically affected by the mutation. As the accumulation of misfolded mutant proteins leads to toxic gain-of-function in hepatocytes, simple delivery of a normal copy (M allele) of SERPINA1 has shown suboptimal efficacy. Instead, in vivo gene-editing strategies that directly correct the causative mutation represent a promising curative modality for managing this disorder (Werder et al., 2021).

To validate the feasibility of gene correction, multiple gene-editing platforms, including TALEN, Cas9 nuclease and base editors, have been explored. In an early proof-of-concept study, Choi et al. (2013) employed a gene knockin strategy to introduce the wild-type sequence, a selection marker (puromycin resistance gene), and piggyBac transposon elements into patient-derived iPSCs using TALEN and plasmid donor. Following puromycin selection, the selection marker was excised by transposase to achieve seamless introduction of the genetic variant. Hepatic cells differentiated from these iPSCs demonstrated functional restoration of the disease phenotype (Choi et al., 2013). More recently, Kaserman et al. applied (2020) HDR-mediated gene editing with Cas9 nuclease and ssODN to correct the Z allele in patient-derived iPSCs. The resulting gene-corrected, iPSC-derived hepatic cells showed restored AAT function as evidenced by increased AAT secretion and decreased intracellular retention (Kaserman et al., 2020). Similarly, Gil et al. (2025) employed Cas9 nuclease and ssODN to correct the Z allele in patient-derived iPSCs as well as to introduce the pathogenic mutation into wild-type iPSCs, thereby generating isogenic pairs that enabled direct comparison of disease and corrected states. These approaches highlight the value of PSC-based isogenic models for mechanistic dissection and therapeutic evaluation (Gil et al., 2025).

A notable feature of precise gene-editing lies in its ability to manipulate zygosity. Delivering only M allele-encoding ssODNs enabled the correction of patient-derived iPSCs into SERPINA1 homozygous wild-type cells. In contrast, using dual ssODNs (each encoding M allele and Z allele) enabled the generation of heterozygous (MZ) mutant. Functional assessment of iPSC-derived hepatic cells revealed that even MZ heterozygotes exhibited significant impairment in AAT function, albeit to a lesser extent than ZZ homozygotes (Kaserman et al., 2022). This finding underscores the value of gene editing for elucidating the impact of variant zygosity in the context of clinical development.

To circumvent the risks associated with Cas9-induced DSB, Werder et al. (2021) employed adenine base editors to corrected the Z allele in patient-derived iPSCs. The resultant iPSC-derived hepatic cells exhibited restored AAT folding and function, alleviating endoplasmic reticulum stress (Werder et al., 2021). These results highlight the translational potential of base editing for clinical applications. Base editing offers distinct advantages over HDR, as it avoids the need for donor DNA templates and eliminates DSB-induced genotoxicity. This simplification facilitates delivery and provides a favorable off-target safety profile, making base editing particularly attractive for in vivo applications. Indeed, recent advances demonstrated successful in vivo correction of alpha-1 antitrypsin deficiency in mouse models using base editors delivered by liver-directed lipid nanoparticles (LNPs) (Kim et al., 2025a).

Urea cycle disorders

Urea cycle disorders are a group of inherited metabolic diseases characterized by the impaired ammonia elimination, leading to the accumulation of highly toxic ammonia in the body. Because hepatocytes are responsible for ammonia detoxification via the urea cycle, defects in this pathway can result in liver failure. These disorders commonly arise from mutations in genes encoding urea cycle enzymes, with pathogenic variants in the OTC gene (encoding ornithine transcarbamylase) being the most frequent.

A recent study by Łukasiak et al. exemplified precision disease modeling by introducing the D175V pathogenic variant into the OTC gene of iPSCs using Cas9 nuclease and ssODN to harness the HDR pathway (Łukasiak et al., 2024). The same study also generated an iPSC line carrying the G390R mutation in the ASS1 gene that encodes argininosuccinate synthase 1, another essential urea cycle enzyme. Functional assays with iPSC-derived hepatic cells revealed distinct outcomes. While the OTC-mutant hepatic cells exhibited a 40% reduction in urea secretion, the ASS1 mutation, despite its prevalence, did not measurably affect urea secretion. These results suggest a context-dependent functional impact of specific variants and highlight the need for further mechanistic investigation.

In an independent study, Zabulica et al. examined patient-derived iPSCs harboring the R129H mutation in the OTC gene, a well-known pathogenic allele (Zabulica et al., 2021). Targeted correction of this mutation was achieved using Cas9 nuclease and ssODN-mediated HDR. Upon hepatic differentiation, corrected hepatocytes demonstrated significantly enhanced urea production compared with unedited controls, underscoring the therapeutic potential of precise gene-editing for restoring urea cycle function.

Arginase deficiency caused by mutations in the ARG1 gene also represents a urea cycle disorder amenable to gene-editing intervention. In one approach, the enzyme deficiency was corrected by targeted knock-in of an arginase expression cassette into exon 1 of the endogenous HPRT locus using a paired nCas9 and plasmid donor (Lee et al., 2016). The HPRT locus was selected as the insertion site to enable enrichment of correctly edited cells, as unedited HPRT-positive cells could be selectively eliminated with 6-thioguanine treatment. This strategy is particularly advantageous given the limited efficiency of HDR in PSCs. Upon hepatic differentiation, corrected cells displayed restored arginase function and competent urea production, demonstrating the feasibility of this safe harbor-based therapeutic strategy (Lee et al., 2016).

Glycogen storage diseases

Glycogen storage diseases (GSD) are a group of inherited metabolic disorders characterized by excessive accumulation of glycogen in the liver and skeletal muscles, the primary sites of glycogen storage in the human body. Many GSDs arise from mutations in genes encoding enzymes essential for glycogen synthesis or degradation. Because multiple enzymes orchestrate the consecutive steps of glycogen breakdown, deficiency in any one enzyme can result in pathological glycogen accumulation, frequently within the lysosomes of hepatic cells and muscle cells. Pompe disease, for example, is caused by mutations in the GAA gene that encodes α-glucosidase. GSD type I (GSD I) arises from mutations in either the G6PC gene (encoding glucose-6-phosphatase-α) or the SLC37A4 gene (encoding glucose-6-phosphate exchanger), both of which are pivotal for the final steps of glycogenolysis in hepatocytes. GSD type III (GSD III) is attributed to mutations in the AGL gene that encodes the glycogen debranching enzyme, while GSD type IV (GSD IV) arises from defects in glycogen branching enzyme 1 encoded by the GBE1 gene (Walsh and Jin, 2024).

To functionally characterize the impact of a variant of uncertain significance (VUS) in GBE1 associated with GSD IV, Naito et al. employed Cas9 nuclease and ssODN-medicated HDR to introduce the I694N mutation into the GBE1 gene in iPSCs derived from healthy donors (Naito et al., 2024). Hepatic cells differentiated from these edited iPSCs displayed the disease phenotype, revealing the effect of the VUS on hepatocytes. This proof-of-concept study underscores the utility of iPSC-based gene-edited disease models for clarifying the clinical significance of VUSs and highlights their potential to accelerate the development of models for other forms of glycogen storage diseases.

Cholestasis

Cholestasis is a pathological condition characterized by impaired or obstructed bile flow, which can lead to inflammation and progressive injury of the liver and bile ducts. Progressive familial intrahepatic cholestasis type 2 (PFIC2) is associated with mutations in the ABCB11 gene which encodes ATP-binding cassette sub-family B member 11 (also known as the bile salt export pump) (Łukasiak et al., 2024). To model PFIC2, Cas9 nuclease and ssODN-mediated HDR were employed to introduce a common missense mutation (D482G) in the ABCB11 locus of iPSCs. Hepatic cells differentiated from the mutant iPSCs exhibited impaired ABCB11 activity and defective bile acid efflux. These findings underscore the value of gene-edited iPSC-derived hepatocytes for modeling PFIC2, particularly in the absence of approved therapies due to the lack of proper disease models (Łukasiak et al., 2024).

Alagille syndrome

Alagille syndrome is a rare inherited disorder that affects multiple organs. In the liver, it is characterized by bile duct paucity, cholestasis, and progression to liver failure. The disease most frequently arises from loss-of-function mutations in the JAG1 gene, leading to defects in Notch signaling (Masek and Andersson, 2024). To model the pathogenic C829X mutation in JAG1, HDR-mediated gene editing was performed using a paired nCas9 and donor DNA to introduce the mutation together with a selection marker and piggyBac transposon elements (Guan et al., 2017). Following puromycin selection, the marker was excised by transient expression of piggyBac transposase. Because Alagille syndrome is inherited in an autosomal dominant manner, heterozygous iPSC clones were selected for downstream analysis. Hepatic organoids derived from these mutant iPSCs recapitulated hallmark disease features, including markedly reduced formation of duct-like structures compared with JAG1 wild-type controls. In a complementary approach, patient-derived iPSCs carrying the mutation were corrected by gene editing to restore wild-type JAG1. Gene-corrected hepatic organoids were then compared with the uncorrected counterparts. Both approaches revealed that the JAG1 mutation impaired ductal morphogenesis, with markedly reduced levels of CK7+ cholangiocytes. Subsequent analyses demonstrated that the JAG1 mutation exerts a dominant-negative effect, a conclusion enabled by the generation of precise heterozygous mutant models through gene editing (Guan et al., 2017).

From a therapeutic perspective, gene editing holds promise for directly correcting pathogenic JAG1 mutations. This strategy is particularly attractive because conventional Notch activators cannot be employed due to the high risk of tumorigenesis, given the pivotal role of Notch signaling in several cancers. Likewise, viral delivery of JAG1 is not feasible, as forced overexpression may itself promote carcinogenesis. Instead, restoration of endogenous, basal-level expression of JAG1 is required. To achieve this, variant correction through precise gene editing is the most rational approach (Sanchez et al., 2021). In rare cases, mutations in NOTCH2 can also cause Alagille syndrome. Because the functional impact of many NOTCH2 variants on Alagille syndrome remains poorly understood, PSC-derived hepatic cells and organoids will serve as valuable platforms for future mechanistic studies and therapeutic evaluation.

Hepatocyte carcinomas

Hepatocytes, the predominant cell type in the liver, represent a major source of tumorigenesis owing to their abundance. Although introducing precise genetic modifications into primary hepatocytes remains technically challenging, iPSC-derived hepatic cells can provide a tractable platform for generating genetically defined tumor models through gene-editing technologies. To demonstrate the feasibility of modeling cancer-associated genetic alterations, Zhang et al. employed Cas9 nuclease to establish a panel of ESC lines carrying either single- or dual-gene knockouts (Zhang et al., 2024). In parallel, lentiviral gene delivery was employed to generate knockin cell lines overexpressing specific oncogenes. Upon hepatic differentiation, cells harboring dual knockouts of APC and TP53, or engineered to overexpress both the PI3KCA E542K mutant and MYC, exhibited markedly enhanced cell proliferation, recapitulating tumorigenic events observed in hepatocellular carcinoma cohorts (Zhang et al., 2024). These findings highlight the potential of integrating gene-editing technologies with hepatocyte regeneration approaches to create robust platforms for dissecting the genetic basis of hepatocellular carcinoma in future studies.

Hepatitis virus infection

Hepatitis B virus (HBV) is a major cause of chronic liver diseases. The sodium-dependent taurocholate co-transporting polypeptide (NTCP), encoded by the SLC10A1 gene, serves as a cellular receptor for HBV through its interaction with the large surface protein of the virus. A naturally occurring variant, NTCP S267F, predominantly found in East Asian populations, has been associated with reduced susceptibility to HBV infection (Uchida et al., 2021). To examine the functional role of this variant in a physiologically relevant setting, Uchida et al. employed adenine base editors to generate ESCs carrying either homozygous or heterozygous mutations (Uchida et al., 2021). Hepatic cells derived from homozygous NTCP S267F ESCs were resistant to HBV infection, whereas heterozygous cells carrying only one copy of the variant remained susceptible. These results demonstrated that the variant does not exert a dominant-negative effect (Uchida et al., 2021). This work represents another key demonstration of how precise gene editing can clarify the functional consequences of zygosity.

LDLR has been implicated in hepatitis C virus (HCV) entry into hepatocytes, although its role in the viral life cycle remains controversial. To clarify its contribution, Caron et al. generated iPSC-derived hepatic cells carrying the LDLR Q12X mutation, alongside gene-corrected cells in which an LDLR expression cassette was inserted into the AAVS1 safe harbor locus (Caron et al., 2019). Control iPSC-derived hepatic cells with intact LDLR sequence were also established. Comparative analyses revealed that viral entry and genome replication were unaffected by LDLR deficiency. However, viral load and infectious titer were decreased in LDLR-null hepatic cells compared with gene-corrected or wild-type controls. These findings suggest that LDLR is dispensable for HCV entry and replication but contributes to later stages of the viral life cycle, such as virus assembly or secretion (Caron et al., 2019).

Improving protocols for generating PSC-derived hepatic cells and liver organoids

Current protocols for generating PSC-derived hepatic cells remain suboptimal, often yielding immature cells. To overcome this limitation, Li et al. conducted a CRISPR-Cas9 screen to identify key genetic regulators of hepatic differentiation (Li et al., 2018). They first generated an albumin-Venus reporter iPSC line by knocking in a Venus fluorescence reporter protein at the C-terminal locus of the ALB gene. Because mature hepatocytes secrete abundant albumin, this strategy enabled selective enrichment of fully differentiated, functional iPSC-derived hepatic cells. Using these reporter cells, they performed a genome-wide CRISPR-Cas9 knockout screen and found that cells exhibiting high Venus fluorescence (top 5%) were enriched for gRNAs targeting ATG7, RPS6KA2, and HDAC3. Targeted knockout of these genes significantly enhanced hepatic differentiation compared with control cells. Complementing the genetic screen, a parallel chemical screen identified CI-994, a selective class I HDAC inhibitor, as a compound that promotes hepatic differentiation. Mechanistic analysis revealed that loss of HDAC3 increased histone acetylation marks H3K9ac and H3K27ac, thereby augmenting gene expression driven by the hepatic transcription factor HNF4 (Li et al., 2018). This study underscores the value of gene-editing approaches for refining cell regeneration protocols.

Unlike monolayer hepatocyte cultures, liver organoids comprise multiple cell types, including hepatocytes and non-parenchymal cells, organized in a 3D architecture. To more faithfully recapitulate in vivo liver physiology, Reza et al. used gene editing to integrate fluorescent reporter proteins into the AAVS1 safe harbor locus of human iPSCs (Reza et al., 2025). Upon differentiation into hepatic lineages and subsequent treatment with zone-specification factors, distinct liver zones were induced, each labeled with different fluorescent markers. This enabled real-time imaging of organoid assembly and morphogenesis, allowing the generation of physiologically structured liver organoids. Moreover, the gene editing facilitated in vivo tracking of transplanted organoids, enabling longitudinal assessment of their engraftment and fate (Reza et al., 2025). Collectively, these studies illustrate how gene-editing technologies can be leveraged not only to enhance differentiation efficiency but also to monitor regenerative cell therapies post-implantation.

GENE EDITING IN DRUG SCREENING, DRUG ACTION STUDY, AND DRUG METABOLISM STUDY

Recently, the U.S. Food and Drug Administration (FDA) announced that preclinical testing of drug candidates in animals is no longer mandatory prior to initiating human clinical trials. This regulatory shift underscores the growing importance of using human stem cell-derived tissues and organoids as critical platforms for evaluating the safety and efficacy of drug candidates, including both small molecules and biopharmaceuticals (Berreur et al., 2025; Park et al., 2024; Zhou et al., 2025). Among these, stem cell-derived hepatic tissues are particularly well positioned to serve as key tools in this evolving landscape. They not only enable the assessment of drugs targeting liver diseases but also facilitate hepatotoxicity evaluation for drug candidates across diverse therapeutic areas, given that liver toxicity testing is an essential prerequisite in drug development to prevent drug-induced liver injury. Increasing evidences indicate that the use of inappropriate models to predict human hepatotoxicity is a major factor behind numerous drug withdrawals (Serras et al., 2021). Therefore, establishing standardized protocols for generating stem cell-derived liver tissues and conducting drug toxicity assay is crucial to meet the demand for more predictive and reliable preclinical models. In this section, we review early efforts and recent advances in employing gene-edited, PSC-derived hepatic tissues for drug testing.

Mitochondrial diseases

Mitochondrial DNA (mtDNA) depletion syndromes (MTDPSs) are genetic disorders characterized by a depletion of mtDNA copy number and consequent impairment of ATP synthesis, leading to multi-tissue dysfunction. In particular, liver malfunction associated with MTDPSs can be fatal, underscoring the importance of modeling the disease using liver tissues (Viscomi and Zeviani, 2017). mtDNA depletion can arise from mutations in genes involved in nucleotide metabolism or mtDNA replication. Among these, mutations in the DGUOK gene, which encodes deoxyguanosine kinase, are frequently observed in hepatic MTDPSs, wherein impaired production of purine deoxyribonucleosides disrupts nucleotide metabolism (Jing et al., 2018). Owing to the scarcity of the patient-derived liver samples, Jing et al. (2018) generated DGUOK knockout iPSCs using Cas9 nucleases and differentiated them into hepatic cells. These cells recapitulated hallmark disease phenotype, including mitochondrial dysfunction, and were subsequently applied in drug screening campaign that identified compounds capable of enhancing ATP production and restoring mitochondrial function. Furthermore, the gene-edited cells provided a platform to investigate the mechanism of action on the hit compound nicotinamide adenine dinucleotide (NAD⁺). Elevated NAD⁺ level enhanced ATP production in both DGUOK-deficient cells and Dguok-deficient rat models, demonstrating the translational relevance of PSC-derived hepatocyte models in drug discovery. In a parallel approach, Cas9 nuclease was employed to knock out RRM2B, a gene required for the conversion of ribonucleoside diphosphates (NDPs) into deoxyribonucleoside diphosphates (dNDPs). Similar to the DGUOK knockout model, impaired ATP production in RRM2B-deficient hepatic cells was rescued by NAD⁺ treatment (Jing et al., 2018). Collectively, these findings highlight the critical role of gene editing in modeling mitochondrial diseases and in facilitating the discovery of novel therapeutic interventions.

Alpha-1 antitrypsin deficiency

A study by Gil et al. (2025) employed Cas9 nuclease and ssODN-mediated gene knockin to generate iPSC-derived hepatic cells carrying the Z allele (E342K) of the SERPINA1 gene. The model cells were subsequently used to evaluate the efficacy of various drugs in the search for therapies for alpha-1 antitrypsin deficiency, a condition currently lacking approved treatments (Gil et al., 2025). Given that autophagy is one of the key mechanisms for the clearance of misfolded protein aggregates, the autophagy-enhancing drug carbamazepine (CBZ) was tested in the iPSC-derived disease model. Consistent with findings from cell line and animal studies, CBZ treatment reduced intracellular levels of polymeric alpha-1 antitrypsin in a dose-dependent manner (Gil et al., 2025; Hidvegi et al., 2010). In addition, the therapeutic potential of short interfering RNA (siRNA) was demonstrated in this model, further supporting the utility of iPSC-derived hepatic cells for preclinical evaluation of novel therapeutic strategies (Gil et al., 2025).

Wilson’s Disease

It is often difficult to obtain patients’ somatic cells for rare genetic variants. Even when such cells are accessible, their reprogramming into iPSCs with robust characteristics, including multi-lineage pluripotency, remains difficult. Wilson’s disease is a rare disease caused by mutations in the ATP7B gene, which encodes a copper-transporting P-type ATPase. Pathogenic ATP7B mutations result in excessive copper accumulation in multiple tissues including the liver, and can ultimately cause irreversible liver failure. These mutations range from SNP to frameshifts, leading to either partial impairment or complete loss of ATP7B activity. To model Wilson’s Disease, Kim et al. employed Cas9 nuclease and ssODN-mediated HDR pathway to introduce the R778L variant of the ATP7B gene into ESCs (Kim et al., 2020a). The resulting ESC-derived hepatic cells carrying the mutation exhibited heightened susceptibility to copper-induced toxicity compared to wild-type controls, faithfully recapitulating the disease phenotype. This model was subsequently used to evaluate candidate compounds for hepatocyte-protective effects. Notably, clinically used copper chelators such as trientine effectively alleviated copper-induced toxicity, demonstrating the value of gene-edited hepatic cells as a reliable platform for drug screening in Wilson’s Disease (Kim et al., 2020a).

Studying drug metabolism and drug toxicity

PSC-derived hepatic cells are increasingly used for precise evaluation of hepatotoxicity in drug development. Gene editing further enhances this utility by enabling the generation of cells carrying defined genetic variants, particularly in cytochrome P450 (CYP) enzymes, thereby allowing personalized assessment of drug responses based on individual genetic backgrounds. For example, Deguchi et al. (2019) applied Cas9 nuclease in iPSCs to knockout cytochrome P450 family 2 subfamily C member 19 (CYP2C19), a key phase 1 drug-metabolizing enzyme. Poor metabolizers with impaired CYP2C19 enzyme activity may experience altered pharmacokinetics and increased toxicity when exposed to CYP2C19 substrates (Deguchi et al., 2019). Although CYP2C19-knockout iPSC-derived hepatic cells expressed hepatocyte markers comparable to wild-type controls, CYP2C19 enzymatic activity was abolished as expected by the knockout. Indeed, these mutant cells successfully modeled hepatotoxicity by clopidogrel, a CYP2C19 substrate, thereby demonstrating the feasibility of performing personalized drug testing with gene-edited tissues (Deguchi et al., 2019). In a similar approach, the same group generated Cytochrome P450 3A4 (CYP3A4) knockout iPSC-derived hepatic cells, given the central role of CYP3A4 in hepatic drug metabolism. These cells reliably predicted CYP3A4-mediated drug metabolism and toxicity. For example, desipramine-induced hepatotoxicity was evident in wild-type cells but absent in CYP3A4-deficient cells, whereas antiviral efficacy of asunaprevir was strengthened due to reduced drug metabolism in knockout cells (Deguchi et al., 2021).

Takayama et al. also focused on CYP3A4. Although CYP3A4 is one of the most representative drug-metabolizing enzymes in hepatocytes, its expression level is often low in PSC-derived hepatic cells (Takayama et al., 2018). To address this, they used Cas9 nuclease-mediated HDR to knock in two reporters, neomycin resistance protein and enhanced green fluorescent protein (eGFP), at the CYP3A4 locus. This gene editing allowed neomycin-based enrichment of CYP3A4-expressing cells, while the eGFP reporter facilitated real-time tracking of the enriched cells. Notably, the enriched cells displayed elevated expression of drug-metabolizing enzymes and hepatic transcription factors, along with enhanced biliary excretion. Consistent with these improvements, the cells enabled sensitive detection of drug-induced hepatotoxicity comparable to the primary human hepatocytes, highlighting their suitability for drug safety assessment (Takayama et al., 2018).

Similarly, Kim et al. (2020b) employed Cas9 nuclease-mediated HDR to generate reporter cells wherein the cytochrome P450 1A1 (CYP1A1) is fused with mCherry protein. CYP1A1 was selected as a target because it is transcriptionally regulated by aryl hydrocarbon receptor (AHR), a critical mediator of xenobiotic toxicity. The resulting CYP1A1-mCherry reporter iPSC-derived hepatic cells were applied in image-based drug screening to identify small molecules that modulate the AHR- and CYP1A1-associated processes (Kim et al., 2020b).

Given the critical role of hepatic metabolism in the manifestation of drug toxicity, loss of drug efficacy, and prodrug activation, robust in vitro models for drug metabolism testing are essential for successful drug development. However, the use of primary hepatocytes is limited by donor-to-donor variability and short lifespan in culture, leading to inconsistent test results. Human cell lines exhibit altered genomic and metabolic profiles, while animal models are constrained by species-specific differences that limit translational relevance. Therefore, PSC-derived hepatic cells and organoids integrated with gene-editing technologies represent promising platforms for physiologically relevant and reliable drug metabolism and toxicity testing, advancing both precision medicine and drug discovery.

GENE EDITING IN REGENERATIVE CELL THERAPIES FOR LIVER DISEASES

The liver is particularly amenable to in vivo gene editing because of its relatively higher gene delivery efficiency compared with other tissues, making it a prime candidate for translational applications of gene-editing therapies. However, pre-existing anti-Cas9 immune responses may compromise the effectiveness of in vivo gene editing. In contrast, transplantation of ex vivo-edited, cell-based therapeutics has not been associated with such immune responses (Essawi et al., 2023; Hakim et al., 2021). Moreover, current in vivo gene delivery technologies often lack sufficient efficiency for many diseases, especially when correction of a large fraction of cells is required. In these cases, ex vivo gene editing of PSC-derived hepatic cells followed by transplantation represents a promising therapeutic strategy. This regenerative cell-based approach also circumvents challenges associated with liver organ transplantation (Reza et al., 2021). For instance, autologous iPSC-derived hepatic cell transplantation could eliminate the need for long-term immunosuppression. Ex vivo editing further avoids concerns related to the long-term presence of AAV vectors in the primate liver, including risks of unexpected vector genome integration (Greig et al., 2024).

Prior to transplantation, cells can be engineered through various strategies to improve therapeutic outcomes, with gene editing representing the most powerful tool. Beyond correcting disease-causing variants in patient-derived cells, gene editing can also endow stem cell-derived hepatic cells with entirely new functions, thereby addressing current limitations of these regenerative cell sources. In this section, we summarize studies in which gene-edited, stem cell-derived hepatic cells have been experimentally applied as therapeutics. Although these examples remain limited due to the early stage of the field, the gene-correction strategies discussed in the previous sections hold strong potential for future development as cell replacement therapies.

Treatment of Wilson’s disease

Because the underlying genetic defect in Wilson’s disease is well defined (section 5.3), it is amenable to precise gene-editing approaches. For instance, Wei et al. corrected patient-derived iPSCs carrying the homozygous ATP7B R778L mutation using Cas9 nuclease- and ssODN-mediated HDR (Wei et al., 2022). Although the gene editing corrected only one allele yielding ATP7B^WT/– iPSC-derived hepatic cells, a single functional copy was sufficient to rescue copper export activity and normal hepatocyte properties in vitro. Upon transplantation into immunocompromised Wilson’s disease mice model, the gene-corrected hepatic cells rescued hepatic copper accumulation and attenuated copper-induced hepatotoxicity. This proof-of-concept study highlights the therapeutic potential of gene-edited, iPSC-derived hepatic cells in regenerative medicine (Wei et al., 2022).

Treatment of urea cycle disorders

To treat the urea cycle disorder caused by arginase-1 (Arg1) deficiency in a mouse model, Sin et al. employed a TALEN and donor DNA-mediated knockin strategy to restore exons 7 and 8 of the Arg1 gene, which were lost in mouse iPSCs derived from Arg1^{Δ/ Δ} mice (Sin et al., 2018). Transplantation of the resulting iPSC-derived hepatic cells into these mice led to partial restoration of Arg1 expression. However, the improvement in survival was modest. This limited phenotypic benefit was attributed to the low engraftment efficiency and the inability of transplanted cells to localize within the optimal metabolic zone of the liver (Sin et al., 2018). These findings suggest that optimized cell transplantation protocols will be essential to fully realize the therapeutic potential of such regenerative approaches.

Treatment of hemophilia B

Hemophilia B is caused by mutations in the X-linked coagulation factor IX (F9) gene. Defects in factor IX impairs proper blood clot formation, resulting in bleeding disorders. Currently, intravenous infusion of recombinant factor IX is the standard treatment. However, it requires lifelong administration and is costly while not providing a cure for the disease. As an alternative, transplantation of autologous hepatocytes engineered to secrete coagulation factors holds promise as a curative strategy.

In an early study, Lyu et al. performed gene editing in patient-derived iPSCs to restore factor IX expression (Lyu et al., 2018). Instead of correcting the missense mutation (R225W) in the endogenous F9 gene, they knocked in a functional F9 expression cassette into the AAVS1 safe harbor locus. This approach enabled puromycin-based selection of gene-edited cells, providing a practical strategy when precise correction of the endogenous F9 locus is challenging. The resulting iPSC-derived hepatic cells secreted functional factor IX, and transplantation into mice model led to detectable levels of human factor IX in the plasma, demonstrating the feasibility of cell-based therapies for hemophilia B (Lyu et al., 2018).

More recently, Bayarsaikhan et al. conducted gene editing to knock in an F9 expression cassette at the APOC3 locus in iPSCs, which enables factor IX secretion specifically after hepatic differentiation under the control of the liver-specific APOC3 promoter (Bayarsaikhan et al., 2025). To enhance in vivo functionality and durability, they generated the hepatocyte cell sheets supported by polymeric scaffolds, resulting in tight cell-to-cell junctions. Transplantation of these cell sheets into mice led to sustained secretion of factor IX, outperforming single-cell transplants that were rapidly lost post-engraftment (Bayarsaikhan et al., 2025). This study underscores the importance of integrating biomaterial-based cell engineering with gene editing to achieve effective regenerative therapies.

Similar gene-editing strategies may also be applicable to other hereditary coagulation disorders such as hemophilia A (factor VIII deficiency). Although hepatocytes do not naturally secrete factor VIII, several studies have demonstrated their utility as a protein secretion platform (Arruda, 2015; Zhang et al., 2023). Furthermore, hepatocytes engineered via gene editing could serve as a robust secretion platform for a wide range of therapeutic proteins, given their well-developed secretory machinery. Thus, gene-edited, PSC-derived hepatic cells hold substantial potential for regenerative medicine across diverse diseases.

Treatment of liver injury

Previous example demonstrated the use of gene-editing approaches to restore enzyme functions compromised by pathogenic variants. Beyond variant correction, gene editing can also be applied to enhance the performance of regenerative cell therapies, particularly since current cell differentiation protocols do not yet yield fully mature hepatocytes.

Cytochrome P450 3A4 (CYP3A4) is an important hepatic enzyme involved in the metabolism of endogenous substrates and a wide range of drugs. However, PSC-derived hepatic cells often express CYP3A4 at levels significantly lower than those in adult liver tissues. To improve the reliability of PSC-derived hepatic cells as a tool for drug testing, Velazquez et al. (2021) employed CRISPR-based transcriptional activation (CRISPRa) system to upregulate CYP3A4 expression. While this approach increased CYP3A4 expression, levels remained below those in adult hepatocytes. To address this, they analyzed adult human liver tissues and liver organoids, identifying activating transcription factor 5 (ATF5) and Prospero-related homeobox1 (PROX1) as key regulators of hepatocyte maturation. Indeed, overexpression of these factors in liver organoids induced CYP3A4 expression comparable to adult liver tissue. The resulting ‘designer liver organoids’ exhibited advanced transcriptional signatures and enhanced hepatic functionality. Moreover, transplantation of these organoids into mouse models of liver injury improved in vivo hepatic function and increased survival rates compared with controls, underscoring the potential of gene-editing approaches to enhance the therapeutic efficacy of regenerative cell therapies (Velazquez et al., 2021).

CONSIDERATIONS FOR GENE EDITING AND APPLICATIONS OF PSC-DERIVED HEPATIC CELLS AND ORGANOIDS

Considerations for safe and efficient gene editing

Conventionally, genetically modified cells, particularly knockin cells, have been generated through lentiviral transduction, which has proven useful for studying hepatocyte biology and developing novel therapeutics. However, such modifications exhibit limited physiological relevance because transgenes are often integrated randomly into unspecified genomic loci, thereby confounding the precise interrogation of genetic factors in liver diseases. In addition, permanent incorporation of viral components into the cell products raises significant safety concerns, especially when these cells are intended for regenerative therapies (Collin de l’Hortet et al., 2019). By contrast, CRISPR-based gene editing enables targeted modifications of endogenous, disease-relevant loci without leaving behind exogenous genetic footprints. Importantly, gene editing also allows control over the zygosity of target genes, which is critical since some diseases require homozygous biallelic mutations for phenotypic manifestation, whereas others can arise even from heterozygous monoallelic variants (Heyne et al., 2023). Thus, a deeper understanding of hepatocyte biology and the development of safer cell-based therapeutics can only be realized through precise gene-editing approaches.

Despite its versatility, several important considerations remain when applying gene editing to PSCs. A primary concern is the risk of off-target editing. In the context of correcting disease-causing variants, the editing site is predetermined and gRNA selection is restricted, as the DNA cleavage site must be located in close proximity to the target variant. This constraint often necessitates the use of gRNAs with higher off-target potential. Therefore, rigorous validation of genomic integrity is required, including sequencing of predicted off-target loci. For therapeutic applications, comprehensive analyses such as whole-genome sequencing and karyotyping are essential to exclude unintended edits and chromosomal abnormalities (Kitano et al., 2022). When gRNAs with low off-target risk are unavailable, alternative strategies, such as employing high-fidelity Cas9 variants or temporally regulated delivery of Cas9 inhibitors, can help mitigate off-target effects (Kim et al., 2023; Lim et al., 2022). Another critical consideration is the role of p53, as CRISPR-mediated DSB can trigger p53 activation, and surviving edited clones may be biased toward cells with compromised p53 function (Ihry et al., 2018). Accordingly, thorough assessment of p53 integrity is necessary before advancing PSC-derived gene-edited cells for clinical translation.

To rigorously assess off-target effects, whole-genome sequencing (WGS) of edited PSC clones can be performed, particularly prior to their use in therapeutic applications. This approach enables the selection of PSC clones that do not harbor detectable off-target modifications. If off-target events are unavoidable, the genomic locations of these alterations should be carefully examined, and clones carrying mutations with potential tumorigenic consequences should be excluded. In addition, potential off-target modifications that may disrupt hepatic differentiation process or metabolic functions of differentiated cells should be avoided. Given the rapidly decreasing cost of WGS, we recommend performing comprehensive genomic analysis before initiating differentiation procedures which require substantial time and resources.

Among the available editing modalities, HDR pathway has been most frequently employed to introduce precise genetic alterations. While HDR provides broad versatility, its efficiency in PSCs remains inherently low, thereby limiting the recovery of correctly edited clones and constraining opportunities to model genetic predispositions to liver disease. To overcome this challenge, multiple HDR-enhancing methods have been developed (Liao et al., 2024; Lim et al., 2020), and researchers should carefully select approaches best suited to their laboratory settings. Of note, use of prime editors in combination with site-specific recombinases has recently emerged as a promising alternative to HDR-mediated knock-in to introduce large gene fragments into the defined genomic loci. Because this method circumvents the requirement for DSB that may lead to genotoxicity in stem cells, it holds potential for widespread application in generating stem cell-derived tissues and engineered hepatic models (Pandey et al., 2025). Furthermore, base editors and prime editors continue to undergo iterative improvements that enhance their robustness and applicability across diverse cell types (Xu et al., 2024), making it essential to adopt the latest available versions to achieve optimal performance. Nevertheless, even with advanced platforms, the overall efficiency of PSC editing often remains suboptimal, which in turn hampers downstream clonal isolation. In such cases, specialized enrichment and selection strategies should be implemented to enable the isolation of precisely edited PSC clones, thereby ensuring reliable downstream applications (Reuven and Shaul, 2022).

Considerations for safety concerns of PSC-derived therapies

Safety remains a key consideration when using PSC-derived therapies. Risks include teratoma formation from residual undifferentiated stem cells as well as severe immune responses from transplanted tissues (Martin et al., 2020; Wiebking et al., 2020). To mitigate these hazards, strategies are being developed to incorporate inducible safety switches that allow selective elimination of transplanted cells in the event of adverse effects. Such switches, initially developed for CAR-T cell therapies (Di Stasi et al., 2011), are readily adaptable for PSC-based products and could facilitate safer and more rapid clinical translation (Martin et al., 2020). For example, drug-inducible caspase-9 (iCasp9) can be introduced into the safe harbor loci of PSCs via gene editing. In the event of teratoma formation, drug administration triggers selective elimination of transplanted cells, thereby functioning as a safeguard for cell-based therapies. Another widely used safety mechanism is the herpes simplex virus-derived thymidine kinase (TK) system. When cells expressing TK are treated with the prodrug ganciclovir, toxic metabolites are generated exclusively in the TK-positive cells, enabling their targeted removal. These safety switches have been shown to function effectively in vivo in PSC-based applications (Martin et al., 2020).

In addition to genetic safety switches, purification strategies to eliminate residual undifferentiated PSCs are critical for ensuring the safety of regenerative cell therapies, as even a small number of remaining undifferentiated cells can result in tumor formation. Representative purification methods include genetic manipulations to overexpression tumor suppressor genes or suppress genes highly enriched in PSCs and teratomas. Cell culture media can also be optimized, or PSC-specific small molecules can be applied to selectively remove undifferentiated PSCs. Cell-sorting techniques offer another option, although scalable purification methods will be required for practical applications (Movahed et al., 2025).

FUTURE PERSPECTIVES

The liver is uniquely amenable to relatively high-efficiency in vivo gene delivery compared to other tissues (Simoni et al., 2024). Accordingly, insights gained from gene editing of PSC-derived hepatic cells can be translated into in vivo gene-editing therapies. Recent advances in AAV- and LNP-based delivery systems further highlight this potential, positioning precise in vivo gene correction as a curative strategy for a wide range of liver-associated genetic disorders. A notable example is the successful delivery of a PCSK9-targeting base editor to hepatocytes using a novel LNP formulation, which demonstrated the feasibility of liver-directed precision gene editing in vivo (DeFrancesco, 2025).

While in vivo gene editing is a powerful tool for correcting pathogenic mutations, it does not restore lost liver mass, which is a limitation in the context of end-stage liver failure. Thus, ex vivo gene editing of regenerative liver-cell therapies and their transplantation can serve as complementary approaches to treat diverse end-stage liver diseases. The recent regulatory approval of Cas9-edited cell therapeutics (Casgevy for sickle cell disease) and the ongoing advanced clinical trials of PSC-derived cell products underscore the plausibility of clinical translation of gene-edited PSC-derived hepatic cells (Healey, 2025; Liu et al., 2024b; Philippidis, 2024). Although liver transplantation remains the current gold-standard treatment for end-stage liver failure, the procedure is limited by an insufficient donor pool, meeting less than 10% of global demand (Luo et al., 2023). Consequently, iPSC-derived liver tissues are being actively explored as alternative therapies for conditions such as acute liver injury and cirrhosis (Ortuno-Costela et al., 2025; Tadokoro et al., 2024).

Gene editing can substantially enhance the functional capacity of these cells, thereby improving therapeutic outcomes. For example, patient-derived iPSCs can be gene-corrected and differentiated into hepatic cells to provide autologous cell therapies. In parallel, the development of universal hypoimmune PSC-derived liver cells provides a foundation for cost-effective, off-the-shelf allogeneic therapies. In this context, gene editing can confer hypoimmunity by disrupting HLA class I and II molecules via knockout of HLA, B2M, and CIITA genes. Proof-of-concept has already been demonstrated in diabetes research, where transplantation of PSC-derived hypoimmune β cells obviated the need for toxic immunosuppressants (Carlsson et al., 2025; Hu et al., 2024a, 2024b; Kim et al., 2021; Kitano et al., 2022; Xu et al., 2019). Similar strategies are expected to be widely applicable to hepatocyte-based therapies.

Beyond genetic modification, advances in biomaterial technologies are emerging as critical enablers for hepatic tissue engineering. Novel biomaterials provide structural support that promotes differentiation into more mature hepatocytes and organoids. Embedding PSC-derived hepatic cells into biomaterial-based scaffolds can further enhance engraftment, function, and stability in vivo (Zhao et al., 2025). Additionally, biomaterial-based encapsulation technologies can protect transplanted cells from immune rejection, thereby prolonging survival without immunosuppression (Liu et al., 2024b; Samadi et al., 2023; Wang et al., 2024). When combined with the precision gene editing, these biomaterial platforms are expected to accelerate the clinical translation of next-generation hepatocyte-based cell therapies.

Although most current gene-editing research in PSC-derived liver models has focused on hepatic cells, fully understanding liver disease will require expanding these efforts to include diverse non-parenchymal cells. For example, severe liver disorders such as fibrosis and cirrhosis arise from the activation of hepatic stellate cells (Friedman, 2008). Recent advances in generating PSC-derived non-parenchymal cells, including hepatic stellate cells and Kupffer-like cells, have begun to enable the reconstruction of more physiologically relevant multicellular liver organoids (Li et al., 2025; Ouchi et al., 2019). However, reports demonstrating efficient gene editing in these models remain limited. As differentiation and co-culture strategies continue to improve, future studies should integrate gene-edited non-parenchymal cells and evaluate their interactions with hepatocytes within mature organoid systems. Such multicellular, gene-edited liver models will be essential for recapitulating complex human liver pathologies and advancing next-generation therapeutic strategies.

Large-scale manufacturing also requires attention, particularly for applications in large-scale drug screening and transplantation as cell-based therapeutics. Current differentiation protocols are continually being optimized, and bioprocess engineering efforts are advancing scalable systems for generating mature hepatocytes at clinically and industrially relevant scales (Harrison et al., 2023; Weng et al., 2023; Wu et al., 2025a).

While this review primarily discusses PSC-derived cells for treating liver diseases, the therapeutic utility of hepatocytes can extend beyond hepatic disorders. Owing to their inherent protein secretion capacity (Schulze et al., 2019), engineered hepatocytes can be programmed to produce therapeutic proteins for diverse systemic diseases. For example, hepatocytes engineered to secrete coagulation factor VIII, which is normally secreted by endothelial cells, present a promising strategy for hemophilia A therapy (Sharma et al., 2015). Thus, the combination of hepatocyte regenerative technologies with precise gene editing could establish entirely new therapeutic modalities with broad clinical applications.

Beyond PSC-derived hepatic cells, other cellular sources, including primary adult liver stem cells, fetal hepatocytes, and organoids derived from these populations, have also been employed in liver disease modeling and drug discovery (Artegiani et al., 2020; Hendriks et al., 2023; Schene et al., 2020). However, these primary cell-derived systems are constrained by low gene-editing efficiency and difficulty in long-term cultures, which makes obtaining genetically homogenous tissues impossible. By contrast, PSCs are highly amenable to precise gene editing coupled with clonal isolation, enabling reproducible studies with genetically uniform cell populations. As such, PSC-derived hepatic cells offer unique advantages, and research leveraging these systems is expected to expand substantially in the near future.

Many genetic alterations affect not only the liver but also a wide range of other organs. Consequently, gene editing-based disease modeling and therapeutic correction have been actively pursued in diverse iPSC-derived tissues beyond hepatocytes. For example, mucopolysaccharidosis type I has been modeled using gene-edited iPSC-derived fibroblasts (Miki et al., 2019), and glycogen storage disease type III has been investigated in gene-edited iPSC-derived skeletal muscle cells (Rossiaud et al., 2023). Since both of these disorders, among many others, also manifest in the liver, the reported gene-editing approaches can be readily adapted for investigating hepatopathies. Conversely, the liver-focused examples highlighted in this review could serve as valuable frameworks for extending gene-editing research into other tissue-specific contexts, thereby broadening our understanding of the systemic nature of genetic diseases and their potential treatments.