1. Introduction
Genome editing technology has revolutionized cell therapy and basic biomedical research, offering powerful tools to precisely manipulate genomic DNA and alter cellular characteristics and functions [
1]. In addition to the compelling therapeutic applications, CRISPR/Cas (clustered regularly interspaced short palindromic repeats/CRISPR-associated proteins)-mediated genome editing has significantly facilitated the generation of more accurate genetically altered disease models, which enable the precise recapitulation of pathogenesis and disease developmental processes.
Genetically modified cellular models hold great promise for providing deeper insights into disease mechanisms and the impact of genetic variants. For example, tumorigenesis has been recapitulated using CRISPR-mediated cellular models for many types of cancer, including brain [
2], breast [
3], colon [
4,
5], pancreas [
6], stomach [
7], etc. These models have been widely used to elucidate the roles of oncogenic genes, mutational signatures, and drug response. Beyond cancer, genetically modified cellular models have been created for studying congenital diarrheal disorder [
8], cystic fibrosis [
9,
10], and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [
11,
12]. These advances highlight the incredible value of modeling human disease by genetically manipulating cells resembling their tissue of origin.
The liver is a major organ responsible for detoxification, metabolism, and various biochemical processes in the human body [
13]. Dysregulation of liver functions leads to complex chronic diseases, such as hepatocellular carcinoma and nonalcoholic fatty liver disease. The first CRISPR-mediated genome editing in mouse liver used hydrodynamic injection to deliver DNA plasmids expressing Cas9 and single guide RNAs (sgRNAs) to the liver [
14]. Subsequently, other delivery methods, including adeno-associated virus (AAV) and nanoparticles, have also been employed for direct in vivo genome editing in animal models [
15].
Despite substantial progress in genome editing of hepatocytes in vivo, efficient genome editing methods for primary hepatocytes in vitro remain underdeveloped, due to challenges in maintaining long-term cultures of functional hepatocytes [
16,
17,
18]. To date, all efforts to extend the culture of primary hepatocytes in monolayer have faced limitations, such as species-specific differences hindering relevance to mouse hepatocytes or dedifferentiation processes impairing hepatocyte functions. In recent years, methods for culturing primary hepatocyte organoids derived from either adult or fetal liver have been described [
19,
20], showing promise for applications in disease modeling, gene and cell therapy, and drug screening. However, expanding adult primary hepatocyte organoids, unlike their fetal liver-derived counterparts, remains challenging and may limit their potential utility [
21]. Therefore, benchmarking the efficiency and robustness of current strategies for the construction of genetically modified cellular models of liver diseases is crucial. This allows assessing the applicability of these models and benefits further research in developing better technologies.
In this study, we systematically compared methods for the culturing and genome editing of murine primary hepatocytes in vitro. The genome editing efficiency in primary adult hepatocytes reached up to 80% in monolayer and 20% in organoids, respectively. In comparison, we performed genome editing in human pluripotent stem cell-derived hepatocytes, which enabled long-term culture with high editing efficiency. By introducing alterations in tumor-associated genes in hepatocytes, we established an in vitro model system to recapitulate liver tumorigenesis. Furthermore, we demonstrated that various oncogenic genes drive the acquisition of various malignant features, providing a platform for investigating gene functions during the onset of liver cancers. This model system offers valuable insights into liver cancer development and can be used to test potential therapeutic interventions.
2. Materials and Methods
2.1. Materials and Reagents
William’s E Medium, B27 (serum free), B27 (minus vitamin A), N2, Advanced DMEM/F12, insulin-transferrin-selenium (ITS), and fetal bovine serum were purchased from Gibco, Thermo Fisher Scientific (Waltham, MA, USA). Bovine Serum Albumin (BSA), LiberaseTM, N-acetylcysteine, Human [Leu15]-gastrin I, CHIR99021, and Nicotinamide were obtained from Sigma Aldrich, Merck KGaA (Darmstadt, Germany). Forskolin, SB431542, IWP2, DAPT, LDN-193189, Y-27632, and A83-01 were purchased from Selleck (Shanghai, China). Human recombinant epidermal growth factor (EGF), recombinant human FGF7, recombinant human FGF10, recombinant human HGF, and Activin a were obtained from Peprotech (Rocky Hill, NJ, USA). Recombinant Human Wnt-3a Protein was obtained from R&D systems (Minneapolis, MN, USA). The 96-well Ultra-Low-Attachment plate, growth factors-reduced Matrigel for organoids culture (#354230), and Matrigel for human pluripotent stem cell-coated (#354277) were purchased from Corning (Corning, NY, USA). Pluripotency Growth Master 1 (PGM1) human pluripotent stem cell culture medium was purchased from Cellapy (Beijing, China).
2.2. Primary Mouse Hepatocytes Isolation
All animal experiments were performed according to the institutional animal care ethics guidelines and approved by the Institutional Animal Care and Use Committee (IACUC) of Shanghai Institute of Materia Medica, Chinese Academy of Sciences. Mouse hepatocytes were isolated from transgenic Cas9-EGFP (Jackson Laboratory, #026175, Bar Harbor, ME, USA) mice using a two-step collagenase perfusion method, as previously described [
22]. Briefly, after successfully anesthetizing the mice, a “U” incision was made in the lower abdomen to expose the liver, portal vein, and inferior vena cava. A needle was then inserted into the vena cava, the portal vein was cut, and the liver was perfused at 3 mL/min with pre-warmed perfusion buffer for 10 min to clear blood. Following this, perfusion was continued with pre-warmed digestion buffer containing 25 μg/mL Liberase and Ca
2+ at 3 mL/min for 3 min. After dissociation, the cells were filtered through a 70 μm filter. Hepatocytes were further separated and purified by Percoll gradient centrifugation at low speed. The hepatocytes were then incubated in DMEM (low glucose) with 5% FBS for at least 3 h. The plating medium was then aspirated, and each well was gently rinsed twice with PBS to remove unadhered cells. The maintenance medium was changed every 2 days.
The primary hepatocyte maintenance media were as follows: Williams’ Medium E with supplements (WEM), Williams’ medium E containing 1× B27, 1× Glutamax, and 1× penicillin-streptomycin; 5C condition (5C): WEM supplemented with Forskolin (20 μM), SB431542 (10 μM), IWP2 (0.5 μM), DAPT (5 μM), and LDN193189 (0.1 μM); Hepatocyte Culture Medium (HCM): manufactured by Lonza (CC3198); Proliferation human hepatocytes condition (ProliHH): Advanced DMEM/F12 supplemented with 1 mM N-acetylcysteine, 1× B27 (minus vitamin A), 1× N2 supplement, 10 nM gastrin I, 25 μg/mL recombinant Human Wnt-3a Protein, 50 ng/mL EGF, 2 ng/mL FGF10, 25 ng/mL recombinant human HGF, 25 ng/mL human recombinant bFGF, 5 μM A83-01, 10 μM Y-27632, 1% penicillin-streptomycin, and 1% fetal bovine serum.
2.3. Cell Viability Assay
As mentioned above, primary mouse hepatocytes were plated in Collagen I-coated 96-well white/transparent bottom plates at a density of 1.5 × 105 cells/cm2.
Cell viability was quantified on days 1, 4, 7, and 10 using a CellTiter-Meiluncell Luminescent Cell Viability Assay Kit (Meilunbio, Dalian, China). Briefly, 100 μL of CellTiter-Meiluncell reagent was added to each well, incubated with cells for 10 min at room temperature, and then measured using a PE EnVision® Multi-Label Plate Reader (PerkinElmer, Waltham, MA, USA). All experiments were performed independently in triplicate.
2.4. Primary Mouse Hepatocytes Organoids Culture
The dome-cultured organoids were established from freshly isolated primary mouse hepatocytes, as previously described [
20]. Briefly, approximately 20,000 hepatocytes were mixed with 30 μL Matrigel and then seeded into a well of a 24-well plate. After the Matrigel had solidified, organoid culture medium was added. The organoid culture medium consisted of AdDMEM/F12 (with HEPES, GlutaMax and Penicillin-Streptomycin) plus 15% RSPO1 conditioned medium (home-made), B27 (minus vitamin A), 50 ng/mL EGF, 1.25 mM N-acetylcysteine, 10 nM Human [Leu15]-gastrin I, 3 mM CHIR99021, 25 ng/mL HGF, 50 ng/mL FGF7, 50 ng/mL FGF10, 1 mM A83-01, 10 mM Nicotinamide, and 10 μM Y-27632.
For floating culture, approximately 10,000 freshly isolated hepatocytes were suspended in 100 μL organoid culture medium supplemented with 5% Matrigel, and then cultured in a 96-well ultra-low-attachment plate. The medium was refreshed every 4 days.
2.5. Immunofluorescence Staining
Immunofluorescence staining was performed according to a previously established protocol [
23]. Briefly, the cell culture medium was discarded, and the cells were fixed with 4% paraformaldehyde (pH 7.4) for 15 min at room temperature, followed by washing with PBS. The cells were then incubated in blocking buffer (PBS containing 5% normal donkey serum and 0.1% Triton X-100) at room temperature for 60 min. After washing with PBS, the cells were incubated overnight at 4 °C with primary antibodies diluted in blocking buffer at the recommended concentrations (CYP3A4, Proteintech, Chicago, IL, USA, 1:600; Albumin, Bethyl, Montgomery, TX, USA, 1:2000). The following day, the primary antibody solution was discarded, and the cells were washed with PBS, then incubated with fluorescent dye-labeled secondary antibodies (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA, 1:2000) diluted in PBS containing 1% BSA for 60 min at room temperature, protected from light. After discarding the secondary antibody solution and washing with PBS, the cell nuclei were stained with PBS containing 1 μg/mL DAPI for 5 min at room temperature, also protected from light. Finally, after discarding the DAPI solution and washing with PBS, the cells were visualized using an Olympus IX73 fluorescence microscope. For whole-mount immunofluorescence of 3D organoids, the organoids were collected and incubated in cold PBS containing 0.1% BSA to dissolve the Matrigel. All subsequent procedures were performed as described for monolayer cells.
2.6. Electroporation
Electroporation was performed using a P3 Primary Cell 4D-Nucleofector™ X Kit (Lonza, Basel, Switzerland). According to the manufacturer’s instructions, 600,000 freshly isolated mouse hepatocytes were resuspended in 100 μL of Lonza electroporation buffer P3 (82 μL Nucleofector Solution and 18 μL Supplement) containing 5 μg of DNA. The cell suspension was then transferred into NucelocuvetteTM vessels and electroporated using a Lonza 4D-NucleofectorTM core unit with the DS-150 program. After electroporation, the cells were resuspended in plating medium and seeded onto collagen I-coated plates for at least 3 h. The plating medium was then replaced with pre-warmed HCM. After 72 h, images were captured using an Olympus IX73 fluorescence microscope and the transfection efficiency was quantified using Image J software (version 1.53e).
2.7. Lipofection
Lipofectamine™ 3000 (Invitrogen, Thermo Fisher Scientific) was used for lipofection, according to the manufacturer’s instructions. For each well of a 24-well plate, 1 μL of Lipofectamine 3000 reagent was diluted in 25 μL of Opti-MEM (Gibco, Thermo Fisher Scientific). Separately, 0.5 μg of DNA and 1 μL of P3000 reagent were added to 25 μL of Opti-MEM in another tube. The DNA solution and the Lipofectamine 3000 reagent solution were then combined and incubated at room temperature for 15 min. After incubation, a 50 μL mixture of DNA-Lipofectamine complex was added to the cells, ensuring even distribution across the well by gently swirling the plate. Twelve hours later, the medium was replaced with fresh HCM. Then, 72 h after lipofection, images were captured using an Olympus IX73 fluorescence microscope, and transfection efficiency was quantified using ImageJ software (version 1.53e).
2.8. AAV Production and Infection
We constructed an AAV vector expressing the EGFP reporter under the control of the EF1a promoter, and verified the accuracy of its DNA sequences by Sanger sequencing. For AAV production, the AAV vector, rep/cap packaging plasmids, and adenoviral helper plasmids were co-transfected into HEK293T cells. After 72 h of incubation, both the cell pellet and supernatant were collected. The cell pellet was lysed through 3 freeze–thaw cycles, and the combined cell lysate and supernatant were concentrated by precipitation in 10% PEG 8000 and 1.0 M NaCl at 4 °C overnight. The precipitate was then resuspended in PBS buffer with a final concentration of 200U/mL Benzonase (Millipore, Merck KGaA, Darmstadt, Germany). Four discontinuous iodixanol gradient solutions (15%, 25%, 40%, and 60%) were prepared in a Beckman tube. The virus-containing supernatant was layered onto the top of the gradient and centrifuged at 58,000 rpm in a Beckman Type 70Ti rotor for 2 h at 18 °C. The 40% iodixanol fraction containing AAVs was collected and the purified virus was further concentrated via buffer exchange into PBS with 0.001% F68 (Gibco, Thermo Fisher Scientific).
The titers of AAVs were measured by qPCR using SYBR Green, with the primers targeting the AAV2 ITR regions: forward ITR primer, 5′-GGAACCCCTAGTGATGGAGTT; reverse ITR primer, 5′-CGGCCTCAGTGAGCGA. The quality of the AAV was confirmed through viral infection assays in cell lines such as HEK293T and Hepa1-6.
For AAV infection, freshly isolated mouse hepatocytes were resuspended in plating medium and seeded onto collagen I-coated plates for at least 3 h. Following this, the plating medium was replaced with pre-warmed HCM. High-titer AAVs (approximately MOI of 30) were added to the HCM, along with polybrene at a final concentration of 8 μg/mL. 12 h later, the medium was replaced with fresh HCM. After 72 h, images were captured using an Olympus IX73 fluorescence microscope, and transfection efficiency was quantified using ImageJ software (version 1.53e).
2.9. Lentivirus Production and Infection
For lentivirus production, 6 μg pMD2.G, 24 μg psPAX2, and 18 μg of lentiviral vector were co-transfected in one 15 cm dish of HEK293T cells using 120 μL PEI. After 12 h, the media was changed for fresh culture media. Viral particles were collected at 48 h and 72 h after transfection, respectively, then concentrated with Millipore 100,000 NMWL centrifugal ultrafilters, divided into aliquots and frozen at −80 °C.
The infection process and the multiplicity of infection of lentivirus was the same as for AAV.
2.10. Cloning of Exogenous Oncogenes Plasmids
The PIK3CA E542K and MYC open reading frame (ORF) sequences were cloned into a lentiviral vector derived from EF1a_mCherry_P2A_Hygro (Addgene, #135003, Cambridge, MA, USA). In this modified vector, the mCherry was replaced with a puromycin resistance gene, and the hygromycin resistance gene was replaced with the ORF. The accuracy of the DNA sequences of each plasmid was confirmed using Sanger sequencing.
2.11. sgRNA Selection and Cloning
We initially selected two sgRNA sequences per gene from the genome-wide CRISPR screen library (Brunello, Addgene, #73179), which was designed using optimized metrics that combine improved on-target activity predictions (Rule Set 2) with an off-target score, the cutting frequency determination [
24]; we then validated the high knockout efficiency of each sgRNA in HEK293T cells during preliminary experiments, and the sgRNA sequences we eventually used are shown in
Table 1.
The selected sgRNAs were cloned into the lentiviral vector CROP-seq-EGFP, a modified version of CROP-seq-opti (Addgene, #106280), where the puromycin resistance gene was replaced with an EGFP reporter. The empty vector was initially digested with the restriction enzyme BsmBI and then recombined with a dsDNA fragment containing the sgRNA sequences using Gibson Assembly. The accuracy of the DNA sequences of each plasmid was confirmed using Sanger sequencing.
2.12. Cell Differentiation and Genetic Modification
The H1 human embryonic stem cells were obtained from WiCell and cultured in Pluripotency Growth Master 1 (PGM1) human pluripotent stem cell culture medium (Cellapy). H1 cells were seeded at a density of 5 × 10
4 cells/cm
2 in Matrigel-coated 48-well plates and maintained in complete PGM1 culture medium for 1–2 days. To generate definitive endoderm, H1 cells were cultured in RPMI1640 medium supplemented with 0.5 mg/mL albumin fraction V and 100 ng/mL Activin a for 1 day. Over the next 2 days, 0.1% and 1% insulin-transferrin-selenium (ITS) were added to the DAY1 medium, respectively. Subsequent steps for the sequential differentiation of the definitive endoderm into posterior foregut, hepatic progenitors, and hepatocytes stages were carried out according to a previously established protocol [
25].
We conducted lentivirus-based genetic modification of hPSC-derived hepatocytes at the hepatic progenitor stage, when the cells exhibited high proliferative capacity and transfection efficiency. To ensure high modification efficiency, we enriched the sgRNA-expressing cells through fluorescence-activated cell sorting (FACS) and the ORF-expressing cells through puromycin selection. After eliminating uninfected cells, the remaining cells continued through differentiation to produce mature hepatocytes for subsequent functional analysis experiments.
2.13. Cell Proliferation Quantification Assay
All operations were carried out according to the recommendations of the BeyoClick™ EdU-594 kit. Briefly, 20 μM EdU solution was prepared by diluting it in the culture medium. When the genetic modified hepatocytes reached 80% confluence, 20 μM EdU solution was used to replace a half of the culture media in the plates, and the cells were incubated for 2 h. After labeling with EdU, the cells were digested and fixed with 4% paraformaldehyde at room temperature for 15 min. Following fixation, cells were permeabilized with 0.3% Triton X-100 in PBS for 10 min. The click additive solution was then prepared according to the manufacturer’s instructions, and cells were incubated with the click additive solution at room temperature, protected from light, for 30 min. EdU quantification was performed by flow cytometry.
2.14. RNA Sequencing
RNA-sequencing libraries were prepared based on the smart-seq3 and smart-seq3xpress method [
26,
27]. First, ASGPR (hepatic surface marker) and EGFP (sgRNA lentivirus infected) double-positive cells were enriched by fluorescence-activated cell sorting (FACS). The sorted cells were then resuspended in 15 µL of lysis buffer and incubated for 10 min at 72 °C in a thermal cycler, followed by immediate cooling on ice. The lysis buffer contained 0.5 U µL
−1 of recombinant RNase inhibitor (RRI) (Takara, Shiga, Japan), 0.15% Triton X-100 (Sigma Aldrich, Merck KGaA), 0.5 mM dNTP (Thermo Fisher Scientific), 1 µM Smart-seq3 oligo-dT primer (5′-biotin-ACGAGCATCAGCAGCATACGA T30VN-3′; GENEWIZ, Suzhou, China), and 5% PEG (Sigma Aldrich, Merck KGaA). Reverse transcription (RT) was performed by the addition of 5 µL RT mix, containing 25 mM Tris-HCl (pH 8.0, Sigma Aldrich, Merck KGaA), 30 mM NaCl (Invitrogen, Thermo Fisher Scientific), 2.5 mM MgCl
2 (Invitrogen, Thermo Fisher Scientific), 1 mM GTP (Invitrogen, Thermo Fisher Scientific), 8 mM DTT (Thermo Fisher Scientific), 0.5 U µL
−1 RRI, 2 µM of improved TSOs (improved TSOs; 5′-Biotin-AGAGACAGATTGCGCAATGNNNNNNNNWWrGrGrG-3′; GENEWIZ), and 2 U µL
−1 of Maxima H-minus reverse transcriptase enzyme (Thermo Fisher Scientific). The reaction was carried out at 42 °C for 90 min, followed by 10 cycles of 2 min at 50 °C and 2 min at 42 °C. The reaction was then terminated by incubating at 85 °C for 5 min. cDNA amplification was performed by adding 30 µL of PCR mix, containing 1x KAPA HiFi HotStart ReadyMix (Roche, Basel, Switzerland), 0.5 µM Smartseq3 forward PCR primer (5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGATTGCGCAATG-3′; GENEWIZ), and 0.5 µM Smartseq3 reverse PCR primer (5′-ACGAGCATCAGCAGCATACGA-3′; GENEWIZ).
After PCR amplification, the cDNA was purified using a 0.6:1 volume of VAHTS DNA Clean Beads (Vazyme, Nanjing, China). The purified cDNA was analyzed on an Agilent 4150 Bioanalyzer to assess cDNA size distribution, and its concentration was measured on a Qubit 3.0 Fluorometer (Invitrogen, Thermo Fisher Scientific) with a Qubit dsDNA High Sensitivity Assay kit (Invitrogen, Thermo Fisher Scientific). The cDNA was then diluted to 200–500 pg µL−1. For tagmentation using our home-made Tn5, 50 ng of cDNA in 10 µL water was mixed with 40 µL of tagmentation mix, which consisted of 5 µL of Tn5, 25 µL of 2× TD buffer (25 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 20% DMF), and 10 µL of water. The mixture was incubated at 55 °C for 5 min in a thermal cycler. Tagmented samples were then purified with 0.8:1 volume of VAHTS DNA Clean Beads.
Sequence library amplification was performed using 1.25 µL of custom-designed Nextera index primers (0.5 µM each) containing 8 bp indexes and 12.5 µL of NEBNext® High-Fidelity 2X PCR Master Mix (NEB, Ipswich, MA, USA). The samples were purified with a 0.7:1 volume of VAHTS DNA Clean Beads, analyzed on an Agilent 4150 Bioanalyser, and DNA concentration was measured using a Qubit 3.0 Fluorometer. The libraries were sequenced on a NovaSeq S4 flow cell with a 150 bp paired end.
2.15. RNA Sequencing Data Analysis
Sequencing reads from SMART-seq3 were processed using UMI-tools (version 1.1.4) to extract UMI-containing reads. The reads were then aligned to the GRCh38 human reference genome (Gencode v43) using STAR (version 2.7.11a). UMI reads were assigned to exon features using featureCounts (version 2.0.3). Normalization and differential expression analysis of the counts were performed using the Deseq2 R package (version 1.2.4).
Differences in gene set variation analysis (GSVA) scores, generated by the GSVA R package, were evaluated using the limma R package.
Weighted gene co-expression network analysis (WGCNA) was conducted using the WGCNA R package. Genes within each identified co-expression module were grouped based on their expression correlation with the entire module. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed using the clusterProfiler R package. The co-expression network was visualized using Cytoscape software (v3.10.2). The accession number for the raw and analyzed data reported in this paper is GSE275241.
2.16. Statistical Analysis
The data for Figure 1B and 1E were analyzed using two-way ANOVA and one-way ANOVA, respectively. The data for Figures 2E and 3G were analyzed using a t-test and two-way ANOVA, respectively. All statistical analyses were performed using GraphPad Prism 9 software. All experiments were performed in three replicates, except for the RNA sequencing experiments, which were performed in two replicates. All values are expressed as means ± SD. p values less than 0.05 were considered statistically significant. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. “ns” means not significant, p ≥ 0.05.
4. Discussion
In this study, we benchmarked two-dimensional (2D) and three-dimensional (3D) cell culture techniques along with various gene delivery methods, to optimize the genome editing of functional hepatocytes in vitro. Our findings indicated that up to 80% genome editing efficiency can be achieved in primary adult hepatocytes cultured in the monolayer format. Additionally, culturing hepatocytes as organoids yields moderate genome editing efficiency, while enabling the long-term maintenance of mature hepatocyte characteristics. Given the limited access to human primary hepatocytes due to ethical considerations, we established genome-engineered cellular models using hepatocytes derived from human pluripotent stem cells to recapitulate tumorigenesis.
We observed that the “5C” culture method, while effective in preserving the functionality of human hepatocytes, was not suitable for murine hepatocytes. This finding underscores the species differences between rodent and human liver, including variations in gene expression, metabolic capabilities, and disease progression [
38]. The ProliHH method demonstrated efficacy in both human and mouse hepatocytes. However, its potential utility is hindered by the dedifferentiation of hepatocytes during culture. On the other hand, the organoid culture system strikes a balance between maintaining hepatocyte characteristics and genome editing efficiency; however, it is less convenient compared to the traditional monolayer culture format.
Despite these limitations, human primary hepatocytes remain essential for studying liver physiology, signal transduction, metabolism, and hepatotoxicity in basic biomedical research and drug discovery, as mature hepatic functions cannot be fully reproduced by mouse models or hepatocytes derived from alternative sources. Therefore, successful genome editing of human primary hepatocytes in vitro is highly valuable for modeling liver diseases and conducting mechanistic studies.
In addition to primary hepatocytes, hPSC-derived hepatocytes provide an alternative cell source for modeling tumorigenesis, with serval advantages. Firstly, human cells avoid significant species-specific differences, thereby representing the molecular mechanisms of human cancer initiation and progression more accurately than animal models. Secondly, the use of hPSC-derived cells circumvents ethical concerns associated with the use of primary human tissues. Thirdly, hPSC-derived hepatocytes ensure an unlimited, consistent, and reproducible cell population for experiments, as they can be generated under controlled conditions. Lastly, hPSC-derived hepatocytes can be easily cultured and genetically manipulated, enhancing the feasibility of experiments compared to primary hepatocytes. Overall, genome engineered hPSC-derived hepatocytes offer a versatile and scalable platform for modeling liver-related diseases, complementing research conducted using animal model primary hepatocytes.
Isogenic disease models, which differ exclusively in individual disease-causing genetic alterations, allow researchers to isolate and study the effect of specific genetic alterations, without the confounding variations common in non-isogenic models. This enhances the reliability and reproducibility of the experimental results, enabling more confident attribution of precise mechanisms to the genetic alterations under investigation. In this research, we developed a series of isogenic cancer models by introducing single or combinations of tumorigenic alterations into wild-type hPSC-derived hepatocytes. This strategy allowed us to shed light on early neoplastic events, discover key pathways and gene modules that alter during tumorigenesis, and identify potential therapeutic targets for hepatocellular carcinoma.
We observed a significant increase in cell proliferation and transcriptomic changes in the PIK3CA E542K overexpression groups, underscoring the potent effect of the activated PI3K/AKT pathway in driving malignant transformation. Notably, two key hallmarks of this transformation, loss of mature hepatic function and metabolic reprogramming, are predominantly influenced by the PI3K/AKT pathway. Previous studies have shown that IL-6, a growth factor upstream of the PI3K/AKT pathway, plays a crucial role in liver regeneration following partial hepatectomy in vivo [
39], and in the expansion of cultured hepatocytes in vitro [
40], both processes also involving hepatocyte dedifferentiation. These findings suggest that physiological liver regeneration and neoplastic transformation may share common pathways that regulate hepatocyte proliferation and dedifferentiation.
It is intriguing that lncRNA
PURPL emerged as a central node among the top 30 interactions within the TP53-related gene module.
PURPL has been reported to be upregulated in liver cancer cell lines, although its exact mechanism remains unclear [
36,
37]. Additionally, the depletion of
PURPL has been shown to elevate basal p53 levels by inhibiting the formation of the p53-MYBBP1A complex in colorectal cancer [
41]. As the role of lncRNA
PURPL is still not well understood, further research into its functions could provide new insights into HCC pathogenesis and potentially lead to novel tools for early diagnosis and treatment.
Historically, generating a panel of isogenic cell models has been labor-intensive. Here, we overcome this challenge by developing an efficient and scalable platform using genome-engineered hPSC-derived hepatocytes. Our approach facilitates systematic analysis of single or combinations of genetic alterations in a single batch of experiments, thereby identifying both common and alteration-specific mechanisms of tumorigenesis.