New three-dimensional preclinical models to understand and treat liver cancers activated for the β-catenin pathway

This study establishes a robust, dynamic suspension culture method using rotating bioreactors to generate uniform, viable 3D organoids and tumouroids from mouse models of beta-catenin-activated liver cancers, which successfully recapitulate native tissue features and demonstrate responsiveness to the beta-catenin antagonist WNTinib for future drug screening and personalized therapy development.

The Gist

Imagine you are trying to understand how a specific type of cancer grows and, more importantly, how to stop it. For a long time, scientists have tried to study liver cancer in two main ways:

1. Flat Petri Dishes (2D): Growing cancer cells on a flat plastic surface. This is like trying to understand a bustling city by looking at a single, flat map. You miss the skyscrapers, the traffic, and the people interacting in 3D space. The cells often get confused, lose their "identity," and die in the middle because they can't get enough air or food.
2. Live Mice (In Vivo): Growing tumors inside actual mice. This is very accurate but expensive, slow, and ethically complex. It's like trying to test a new car engine by building a whole new car for every single test.

The Problem:
Liver cancers driven by a specific "engine" called the β-catenin pathway (found in about 40% of adult liver cancers and 80% of childhood liver cancers) are particularly tricky. They are resistant to many current treatments, and we don't have a good, simple model to test new drugs on them.

The Solution: The "Zero-Gravity" Salad Spinner
This paper introduces a new, clever way to grow these cancers in a lab. The researchers used a special machine called a ClinoStar® bioreactor.

Think of this machine as a high-tech salad spinner or a space station simulator.

• Instead of letting cells sit flat on a dish, the machine gently rotates them in a liquid soup.
• This rotation creates a "micro-gravity" effect. The cells don't stick to the plastic; they float and clump together naturally, forming perfect little 3D spheres.
• Because they are spinning gently, nutrients and oxygen can reach the very center of the ball, just like air circulating in a room. This prevents the "dead zone" in the middle that happens in flat dishes.

The scientists call these 3D balls "Tumouroids."

What They Did:

1. The Ingredients: They took liver cells from mice that were genetically programmed to develop liver cancer (specifically the kind driven by the β-catenin pathway).
2. The Recipe: They put these cells into the "salad spinner" bioreactor.
3. The Result: Within days, the cells formed dozens of tiny, perfect, 3D tumor balls.

Why This is a Big Deal:

• They Look Real: Unlike the flat dish cells, these Tumouroids kept their original shape and structure. They even kept their "neighbors" (immune cells and blood vessel cells) alive inside the ball, just like a real tumor does.
• They Act Real: The scientists checked the DNA and found that the Tumouroids were speaking the exact same "genetic language" as the original mouse tumors. They weren't confused or changing their identity.
• They Respond to Medicine: To prove they were useful, the researchers fed these Tumouroids a new drug called WNTinib (a drug designed to shut down the β-catenin engine).
  • The Result: The drug worked! It stopped the cells from multiplying and made the cancer cells commit "suicide" (apoptosis).
  • The Analogy: It's like taking a car engine that won't turn off, putting it in a test rig, and successfully injecting a chemical that makes it shut down.

The Bottom Line:
This new method is like having a miniature, ethical, and fast "test drive" for liver cancer treatments.

• For Scientists: It's a reliable, repeatable way to study how these specific cancers work without needing hundreds of mice.
• For Patients: It speeds up the discovery of new drugs. Instead of waiting years to test a drug in a mouse, scientists can test it on these Tumouroids in days.
• For the Future: The authors hope this method can eventually be used with human tumor samples to create "personalized" test tubes. Imagine taking a tiny piece of a patient's tumor, growing a "Tumouroid" in the salad spinner, and testing ten different drugs to see which one kills it best before ever giving the drug to the patient.

In short, they built a 3D, rotating, living model of liver cancer that is cheaper, faster, and more accurate than the old ways, offering a new hope for finding cures for stubborn liver cancers.

Technical Summary

1. Problem Statement

Constitutive activation of the β-catenin pathway is a critical driver in two primary liver cancers:

• Hepatocellular Carcinoma (HCC): Found in 30–40% of cases.
• Hepatoblastoma (HB): Found in 80–90% of cases.

Current Challenges:

• Therapeutic Gaps: There are no specific treatments for advanced β-catenin-mutated HCC, and these tumors often show innate resistance to standard immunotherapy/anti-angiogenic combinations. HB patients with chemo-resistant or recurrent disease have poor prognoses.
• Model Limitations: Existing in vitro models (2D cultures) fail to maintain hepatocyte identity and rapidly dedifferentiate. Traditional 3D models (e.g., Ultra-Low Attachment [ULA] plates) often develop necrotic cores due to poor nutrient/oxygen diffusion, leading to cell death and loss of tumor heterogeneity. Patient-derived xenografts (PDXs) are expensive, time-consuming, and ethically challenging for high-throughput screening.
• Need: There is an urgent need for robust, reproducible, and physiologically relevant 3D models that recapitulate the cellular diversity, histology, and transcriptome of native β-catenin-driven liver tumors for drug screening and mechanistic studies.

2. Methodology

The authors developed a novel methodology using dynamic suspension culture in a rotating bioreactor system (ClinoStar® by Celvivo) to generate organoids and tumouroids from mouse models.

• Mouse Models:
  • APCΔHep Model: Apclox/lox/TTR-CreERT2 mice induced with tamoxifen (for preneoplastic organoids) or Adenovirus-Cre (for HCC and HB-like tumors).
  • βcatΔEx3 Model: Mice injected with AAV8-saCas9/sgRNA to delete Ctnnb1 exon 3 (stabilizing β-catenin), generating HB-like mesenchymal tumors.
• Culture System (ClinoStar®):
  • Mechanism: Uses a rotating bioreactor to counterbalance gravitational forces, creating near-microgravity conditions. This minimizes shear stress, eliminates plastic contact, and ensures uniform gas/nutrient exchange without requiring extracellular matrix (Matrigel) or scaffolds.
  • Protocol:
    • Organoids: Freshly isolated hepatocytes and non-parenchymal cells (NPCs) from APCΔHep livers were seeded at a 2:1 ratio.
    • Tumouroids: Tumors were excised, enzymatically digested, and dissociated into hepatocytes and NPCs, then seeded into the bioreactor.
  • Comparison: Results were compared against static 3D cultures in ULA plates and 2D collagen cultures.
• Characterization Techniques:
  • Histology/IHC: H&E staining, immunohistochemistry (GS, β-catenin, KI67, F4/80, CD146, cleaved caspase-3).
  • Molecular Analysis: RT-qPCR for gene expression; RNA-seq for transcriptomic profiling (PCA, differential expression, KEGG pathway enrichment).
  • Drug Screening: Treatment with WNTinib (a multi-kinase inhibitor targeting mutant β-catenin) at 1 µM (organoids) and 5 µM (tumouroids) for 5 days.

3. Key Contributions

• Novel Culture Platform: First demonstration of using the ClinoStar® rotating bioreactor to generate murine liver organoids (preneoplastic) and tumouroids (HCC and HB-like) without matrix additives.
• Preservation of Tumor Heterogeneity: The method successfully maintains the complex cellular architecture of the original tumors, including hepatocytes, macrophages (F4/80+), and endothelial cells (CD146+), which is often lost in other 3D models.
• Elimination of Necrotic Cores: Unlike ULA plate cultures, the dynamic culture prevents central necrosis, ensuring high viability and uniform size distribution.
• Translational Relevance: Validated the models as effective tools for drug screening by demonstrating sensitivity to WNTinib, a clinically relevant compound.

4. Key Results

A. Superiority of ClinoStar® over ULA Plates

• Morphology: ClinoStar® cultures produced uniform, compact organoids/tumouroids (250–300 µm) with no structural loosening after 16 days. In contrast, ULA plate cultures showed structural disintegration and smaller sizes.
• Viability: ClinoStar® organoids showed no central necrosis or cell death (negative for cleaved caspase-3 and TUNEL). ULA plate organoids exhibited significant apoptosis and necrotic cores.
• Gene Expression: Glul (glutamine synthetase), a direct β-catenin target, was maintained at high levels in ClinoStar® cultures (similar to fresh tissue) but significantly downregulated in ULA plate cultures.
• Cellular Diversity: ClinoStar® cultures retained macrophages and endothelial cells, whereas ULA cultures showed poor IHC quality due to background noise from dead cells.

B. Transcriptomic Fidelity

• RNA-seq Analysis: Principal Component Analysis (PCA) showed that tumouroids clustered perfectly with their native tumor of origin, distinct from non-tumoral liver tissue.
• Pathway Preservation:
  • HCC Tumouroids: Retained high expression of canonical β-catenin targets, glutamine/bile metabolism genes, and low amino acid catabolism genes.
  • HB-like Tumouroids: Retained embryonic liver gene signatures, mesenchymal markers, and Hippo pathway signatures.
  • Differential Expression: Only two genes (Fam177a2 and 2810047C21Rik) were significantly upregulated in HB tumouroids compared to native tumors, indicating high fidelity.

C. Drug Screening Efficacy (WNTinib)

• Response: Treatment with WNTinib induced significant structural disorganization and apoptosis in both preneoplastic organoids and HCC tumouroids.
• Molecular Markers:
  • Increased cleaved caspase-3 and TUNEL positivity.
  • Downregulation of β-catenin targets (Glul, Axin2) and anti-apoptotic gene (Bcl2l1).
  • Upregulation of pro-apoptotic gene (Noxa).
  • Reduced proliferation (KI67+ cells).
• Conclusion: The models accurately recapitulated the therapeutic response seen in human cells, validating their use for preclinical testing.

5. Significance and Impact

• Reduced Animal Use: These in vitro models can replace large cohorts of mice for pre-selection and toxicity testing, aligning with the "3Rs" (Replacement, Reduction, Refinement) in animal research.
• Precision Medicine: The models allow for "à la carte" therapy testing for specific β-catenin-driven subtypes of HCC and HB, which are currently difficult to treat.
• Scalability and Reproducibility: The method is simple, uses commercially available media, and generates large numbers of uniform structures, making it suitable for high-throughput drug screening.
• Future Potential: The authors are currently investigating the adaptation of this protocol for human liver tumor fragments, which would further enhance its translational value for personalized oncology.

In summary, this study provides a robust, matrix-free, dynamic 3D culture system that overcomes the limitations of existing models, offering a highly faithful representation of β-catenin-driven liver cancers for both basic research and therapeutic development.