PhoenixBio is heading to Denver for The Liver Meeting® 2026, AASLD’s 77th annual meeting and its first in the city. It runs 5–9 November 2026 at the Colorado Convention Center, and you’ll find our team at Booth #766 throughout the exhibition.
This year, four presentations at the meeting feature PXB-cells. They showcase our high quality, primary human hepatocytes adding value in areas from MASH to viral hepatitis (HDV) to cholestatic diseases such as Primary Biliary Cholangitis. Check out the presentations and then stop by our booth to speak to our scientists about how humanized liver models can support your research.
Humanized Liver Related Abstracts
0003 Shared and Isoform-specific Roles of PPAR-α, -δ, and -γ Signaling in mitigating Cholestatic Hepatocyte Injury in Primary Biliary Cholangitis
Garnik Arustamyan1 Masashi Okawa1 Yein Woo1 Daniel Pinski1 Meng Li1 Takeshi Saito2
1University of Southern California, Keck School of Medicine, 2University of Southern California
Sat, 11/7 2:45 - 3:00 pm, General Session Hall A
Background: PBC is a chronic cholestatic liver disease in which autoimmune-mediated bile duct injury impairs bile flow, leading to intrahepatic accumulation of toxic bile acids and consequent hepatocyte stress and injury. UDCA remains the first-line therapy; however, a substantial proportion of patients have an incomplete response and remain at risk for disease progression. PPAR agonists have emerged as effective second-line PBC therapies; however, the isoform-specific mechanisms by which PPAR activation protects hepatocytes from cholestatic bile acid injury remain poorly defined. We therefore aimed to define the shared and distinct roles of PPAR-α,-δ, and-γ signaling in regulating bile acid homeostasis and mitigating cholestatic injury in primary human hepatocytes (HHs). Methods: HHs were benchmarked against normal human liver tissue to confirm PPAR isoform expression and intracellular localization. HHs were then treated with PPAR isoformselective and multi-isoform agonists, including fenofibrate, seladelpar, elafibranor, and lanifibranor. Effects on bile acid homeostasis and canalicular bile secretory machinery were assessed by RNA-seq profiling, fluorescence microscopy with CDFDA and fluorescent bile acids, and LC-MS quantification of intracellular and extracellular bile acid species. PPAR isoform dependence was tested using solid-phase reverse-transfection–based siRNA knockdown of PPAR isoforms, individually and in combination. Parallel studies with UDCA and obeticholic acid assessed additive or synergistic interactions between PPAR pathways and FXR signaling. Therapeutic efficacy of PPAR isoform specific agonists alone or in combination with UDCA was evaluated in a CDCA-induced in vitro model of cholestatic HH injury by assessing reductions in bile acid–induced cellular stress and injury and modulation of canalicular bile secretory function. Results: Gain and loss-of-function studies using PPAR agonists and siRNA-mediated silencing revealed isoform-specific regulation of bile acid homeostasis. Dual PPAR-α/δ agonism most effectively reduced the intracellular toxic bile acid pool in HHs, whereas PPAR-γ signaling had minimal effect. The anticholestatic effect of PPAR-α/δ agonist was mediated primarily through enhanced canalicular bile acid export. Dual PPAR-α/δ agonism, particularly when combined with UDCA, exhibited the strongest anticholestatic and cytoprotective effects in the CDCA-induced cholestatic hepatocyte injury model. Conclusion: Dual PPAR-α/δ agonism, exemplified by elafibranor, enhanced canalicular bile acid export in primary HHs, thereby reducing the intracellular toxic bile acid pool that drives CDCA-induced cholestatic stress and injury. The additional protection observed with UDCA co-treatment further supports enhanced hepatocellular bile acid disposal as a central mechanism for mitigating cholestatic injury in PBC.
1320 Human Hepatocyte-Based Functional Annotation of MASH-associated Genetic Risk Variants defines Distinct Stress States and Therapeutic Response
Takeshi Saito1 Yein Woo2 Jinyoung Ahn2 Masashi Okawa2 Meng Li1 Garnik Arustamyan2
1University of Southern California, 2Keck School of Medicine, University of Southern California
Thu, 11/5 12:00 - 1:00 pm, Hall DE
Background: Genetic variants in TM6SF2, PNPLA3, MBOAT7, TRIB1, GPAM, ADH1B, FTO, APOB, MTTP, and other MASH-associated loci are thought to confer variable risks of steatosis, hepatocellular stress, fibrosis, and cirrhosis. However, this concept derives largely from clinical association studies; these variants have not been systematically compared at the cellular and molecular levels for direct effects on metabolic stress pathways or responses to hepatocyte-targeted therapeutics. This study establishes a physiologically relevant platform using our state-of-the-art primary human hepatocyte (HH) culture system to define how MASH-associated genetic risk signatures shape hepatocyte pathophysiology and drug responsiveness. Methods: HHs were perturbed using solid-phase siRNA transfection to model MASH-associated risk genes and exposed to a carbohydrate–fatty acid cocktail as the primary MASH-relevant metabolic challenge. Ethanol alone and combined carbohydrate–fatty acid/ethanol exposure were included as comparator and extension arms to test MASH-specific versus shared effects across ALD- and MetALD-like stress contexts. Effects on lipid accumulation, lipotoxicity, oxidative and ER stress, hepatocyte-intrinsic inflammatory signaling, and cell injury were assessed by high-content imaging, biochemical assays, and transcriptomic profiling. Therapeutic responsiveness was evaluated with resmetirom (THR-β agonist) and lanifibranor (pan-PPAR agonist). Results: Despite comparable lipid droplet accumulation, individual gene silencing produced distinct metabolic stress and injury signatures under MASH-like challenge and in MetALD/ALD extension arms. Silencing of PNPLA3, MBOAT7, TM6SF2, APOB, and MTTP was strongly associated with metabolic-oxidative stress and cytotoxicity, whereas silencing of GPAM, TRIB1, ADH1B, and FTO caused minimal stress or injury. Therapeutic responses were risk gene–dependent: resmetirom reduced steatosis and stress signaling in PNPLA3-, APOB-, and MTTP-silenced HHs, but the magnitude of lipid reduction, injury mitigation, and stress-gene suppression varied by genetic background. Lanifibranor modestly reduced lipid droplet burden across models but did not significantly reduce cell stress or injury. Conclusion: This study establishes a HH-based framework to functionally annotate MASH associated genetic variants and link genetic risk to molecular pathophysiology and therapeutic response. These findings indicate that MASH risk genes define separable hepatocyte stress states with distinct drug response profiles across metabolic and alcohol-associated stress contexts. This platform provides a functional approach to stratify MASH risk variants by mechanism and test therapeutic responsiveness in a bona fide human hepatocyte model system.
3219 HDV Strain-specific Fitness and HBsAg Pairing refine the Genotype-based Paradigm of HDV Pathobiology
Takeshi Saito1 Daniel Pinski2 Masashi Okawa2 ShiuanChi Wang2 Meng Li1 Garnik Arustamyan2
1University of Southern California, 2Keck School of Medicine, University of Southern California
Sat, 11/7 1:00 - 2:00 pm, Hall DE
Background: Hepatitis delta virus (HDV) accelerates HBV-associated liver disease and exhibits marked genetic diversity across eight genotypes. Certain HDV genotypes have been clinically linked to aggressive necroinflammation, fibrosis progression, and hepatocellular carcinoma; however, these associations are confounded by strain diversity, co-infecting HBV, host genetics, and regional comorbidities. Building on our prior cross-genotype HDV platform, this study tested whether clinically recognized genotype differences reflect intrinsic HDV genotype properties, strain specific hepatocyte fitness, or functional compatibility between HDV genomes and HBV surface antigen (HBsAg). Methods: We analyzed 25 full-length HDV strains spanning all eight genotypes, including strains from genotypes clinically associated with aggressive disease. Using our established cross-genotype HDV HBsAg platform, each HDV construct was evaluated with HBsAg expression systems representing distinct HBV envelope backgrounds. Producer-cell readouts included intracellular HDV RNA replication, S-HDAg/ L-HDAg production, extracellular HDV RNA release, and particle-associated HDV output. The resulting HDV preparations were then assessed in optimized PHH cultures under HDV-alone and HBV-HDV coinfection conditions. PHH-based analyses quantified HDV entry/propagation, intracellular and extracellular HDV RNA, HBV-HDV interactions, and HBV envelope–supported HDV particle production. Results: Across 25 HDV strains, we observed marked strain-to-strain variability in intracellular HDV RNA replication and S-HDAg/L-HDAg production. These phenotypes were not consistently predicted by HDV genotype as strains within the same genotype differed in replication efficiency and antigen production. HBsAg pairing further modified HDV propagation in a bidirectional and combination-specific manner. For a given HDV strain, selected HBsAg backgrounds enhanced intracellular HDV RNA replication and increased extracellular virion production, whereas others suppressed both intracellular and extracellular HDV readouts. Conversely, the same HBsAg background supported different levels of intracellular replication and virion production depending on the paired HDV strain. In most combinations, intracellular HDV replication and virion output changed in parallel, although select pairings showed partial dissociation between genome replication and envelope-supported particle production. Conclusion: This cross-sectional analysis of HDV strains and HBsAg pairings refines the genotype-based view of HDV pathobiology. HDV replication, HDAg production, and virion formation were determined by strain-specific fitness and HBsAg compatibility rather than genotype alone, suggesting that aggressive disease may reflect high-fitness HDV HBsAg pairings.
0051 Geometry-tuned Primary Human Hepatocyte Culture defines a Mechanotransduction Set Point for Hepatocyte Identity Loss
Takeshi Saito1 Masashi Okawa2 Yein Woo2 Garnik Arustamyan2 Daniel Pinski2 Meng Li1 Hans I-Chen Harn2
1University of Southern California, 2Keck School of Medicine, University of Southern California
Sun, 11/8 11:30 - 11:45 am, Bluebird Ballroom 2DEF
Background: Maintenance of mature human hepatocyte (HH) identity depends in part on preservation of native cellular architecture. In normal human liver, hepatocytes display constrained lateral dimensions (20–25 µm), whereas cirrhotic hepatocytes show marked geometric heterogeneity (18–40 µm), with enlargement associated with identity loss. Primary HH culture models this geometry–identity relationship, in which loss of tissue-imposed constraints renders hepatocytes susceptible to seeding density dependent distortion and lateral spreading. This study aimed to define the cellular geometry mechanotransduction set point that preserves hepatocyte identity by integrating liver tissue benchmarking with density-tuned primary HH culture. Methods: HHs were seeded as 2D monolayers at graded densities on substrates of varying stiffness to modulate lateral cell geometry and cell–matrix tension. Low-density culture permitted aberrant lateral spreading, whereas high density culture imposed spatial confinement approximating normal liver dimensions. Cell geometry was quantified by projected area, equivalent circular diameter, and major/minor lateral axes, while cellular mechanics were measured by atomic force microscopy. Hepatocyte architecture and polarity were assessed by confocal microscopy, including bile canaliculi formation. These readouts were integrated with single-cell RNA sequencing to link geometric and mechanical states with mature hepatocyte identity and YAP/TAZ-associated mechanotransduction. In vitro geometry–identity states were benchmarked against normal and cirrhotic human liver tissue. Results: Density- and stiffness tuned HH cultures generated distinct geometry states ranging from spatially confined cells within the normal liver size range to laterally enlarged cells resembling cirrhosis-associated geometric heterogeneity. Lateral enlargement >28 µm correlated with altered mechanics, disrupted canalicular architecture, YAP/TAZ-associated mechanotransduction, and loss of mature hepatocyte identity. Single-cell RNA sequencing identified a 16-gene identity-loss signature enriched in enlarged HHs and upregulated in cirrhotic hepatocytes. Signature-positive cells segregated into four subpopulations: residual mature hepatocyte, cholangiocyte-like, hepatic progenitor-like, and mesenchymal-like states. Conclusion: Expansion beyond an equivalent lateral diameter of 28 µm marked the emergence of hepatocyte identity-loss programs, with maladaptive mechanotransduction and canalicular disorganization mirroring cirrhosis. These findings refine cirrhosis as a hepatocyte identity-loss state, not solely hepatocyte loss, and support geometry-tuned primary HH culture as an in vitro platform to model cirrhosis-associated hepatocyte failure and identify therapeutic targets.