Cell line identity rather than medium composition determines transcriptomic profiles of HepaRG and HuH7 cells cultured in chemically defined or serum-based media: comparison with primary human hepatocytes

This study demonstrates that while cell line identity is the primary determinant of transcriptomic profiles in HepaRG and HuH7 cells, switching from serum-based to chemically defined media differentially enhances xenobiotic metabolism in HepaRG cells to better resemble primary human hepatocytes, whereas HuH7 cells primarily exhibit stress and lipid homeostasis responses.

The Gist

The Big Question: What's in the Lunchbox?

Imagine you are trying to train two different types of athletes: a Marathon Runner (representing the HepaRG cells) and a Power Lifter (representing the HuH7 cells).

For decades, scientists have fed these athletes a "mystery stew" called Fetal Bovine Serum (FBS). This stew is made from cow blood. It's packed with nutrients, but it's a total "black box." You don't know exactly what's in every spoonful, and one batch might taste different from the next. This makes it hard to know if the athlete's performance is due to their natural talent or just the specific batch of stew they ate that day.

Recently, scientists started using Chemically Defined Media (CDM). Think of this as a precise, custom-made meal plan. Every ingredient is measured down to the gram. No mystery, no cow blood, just pure, controlled nutrition.

The Big Question: If we switch these athletes from the "mystery stew" to the "precise meal plan," does it change who they are? Does it make them run faster or lift heavier? Or does their natural talent (their DNA) matter more than what they eat?

The Experiment: A Taste Test

The researchers took two famous liver cell lines (HepaRG and HuH7) and grew them side-by-side.

• Group A: Ate the old "mystery stew" (FBS).
• Group B: Ate the new "precise meal plan" (CDM).

They also included a group of Primary Human Hepatocytes (PHH). These are the "Gold Standard" athletes—actual liver cells taken directly from a human donor. They represent what a perfect liver cell should look like.

Then, they took a "snapshot" of every single gene in these cells (like taking a photo of every muscle fiber and instruction manual) to see how the diet changed them.

The Results: Identity Wins, Diet Tweaks

Here is what they found, broken down simply:

1. The "Personality" Matters More Than the "Menu"

The biggest surprise was that who the cell is matters way more than what it eats.

• Analogy: Imagine a Chihuahua and a Great Dane. If you feed them both the exact same premium dog food, the Chihuahua will still be small and yappy, and the Great Dane will still be huge and calm. The food changes them slightly, but it doesn't turn the Chihuahua into a Great Dane.
• The Science: The biggest difference in the data was between the two cell lines (HepaRG vs. HuH7), not between the two diets. The cells kept their core "personality" regardless of the medium.

2. The Marathon Runner (HepaRG) Got a Boost

The HepaRG cells are the "good students" of the liver world. They are already pretty close to real human liver cells.

• The Change: When these cells switched to the precise meal plan (CDM), they actually got better. They started acting more like the Gold Standard (PHH).
• The Metaphor: It's like giving a talented musician a high-quality instrument. They were already good, but the new tool helped them play even closer to the master's level. Their genes for "detoxifying" (cleaning up toxins) and "metabolizing" (processing drugs) turned up louder and clearer.
• The Takeaway: If you want to test liver drugs, using HepaRG cells in this new diet is a great idea. It makes them more reliable and human-like.

3. The Power Lifter (HuH7) Got Stressed

The HuH7 cells are different. They are tougher but less specialized.

• The Change: When these cells switched to the precise meal plan, they didn't get "more liver-like." Instead, they got confused. They started obsessing over cholesterol and stress.
• The Metaphor: Imagine a Power Lifter who is used to a messy cafeteria diet. Suddenly, you put them on a strict, low-fat diet. Instead of lifting better, they start panicking about their weight and obsessing over their body fat. They didn't become a better liver cell; they just started reacting to the lack of "fatty" cow blood in their food.
• The Takeaway: For HuH7 cells, the new diet didn't make them better models for human liver function. They remained quite different from real human liver cells.

The Conclusion: Why This Matters

For a long time, scientists worried that switching away from the "mystery stew" (FBS) would ruin their data, making it impossible to compare new results with old studies.

This paper says: "Don't worry!"

1. The Cell is King: The cell's own identity is the main driver of its behavior. Changing the food doesn't turn a cat into a dog.
2. CDM is a Refinement, Not a Reset: Switching to the precise meal plan (CDM) doesn't break the science; it actually clarifies it. It removes the "noise" of the cow blood.
3. Better Tools for Specific Jobs:
   • If you are studying drug safety, use HepaRG cells in the new diet. They are now even closer to real human livers.
   • If you are using HuH7 cells, be aware that the diet changes their stress levels, but it won't magically turn them into perfect human liver cells.

In a nutshell: The researchers proved that while the "menu" (the medium) matters, the "chef" (the cell type) is the one who decides the final dish. Switching to a clean, defined menu helps the best chefs (HepaRG) cook even better, while the others (HuH7) just react to the change in ingredients. This gives scientists a clearer, more reliable way to test new medicines.

Technical Summary

Title

Cell line identity rather than medium composition determines transcriptomic profiles of HepaRG and HuH7 cells cultured in chemically defined or serum-based media: comparison with primary human hepatocytes.

1. Problem Statement

• The Issue: Mammalian cell culture, particularly in liver research, has historically relied on media supplemented with Fetal Bovine Serum (FBS). FBS is an undefined "black box" with lot-to-lot variability, potential contaminants (endotoxins, prions, viruses), and ethical concerns regarding xenogeneic sourcing.
• The Gap: While Chemically Defined Media (CDM) offer a solution for standardization and reproducibility, the specific impact of transitioning from FBS-supplemented media (FBS-SM) to CDM on the transcriptomic profiles of widely used hepatic cell lines (HepaRG and HuH7) remains insufficiently characterized.
• The Concern: Researchers worry that switching to CDM might create data incomparable to historical FBS-based literature or alter critical metabolic and detoxification pathways essential for hepatotoxicity testing.

2. Methodology

• Cell Models:
  • HepaRG: A differentiated human hepatic cell line known for metabolic competence.
  • HuH7: A human hepatocellular carcinoma cell line used for toxicology and viral research.
  • HeLa: Included as a non-hepatic outgroup.
  • Primary Human Hepatocytes (PHH): Pooled cryopreserved hepatocytes (10 donors) used as the physiological gold standard reference.
• Culture Conditions:
  • Cells were cultured in parallel under two conditions:
    1. FBS-SM: Standard media supplemented with 10% FBS.
    2. CDM: Custom, chemically defined, serum-free formulations (e.g., ITS, EGF, HGF, lipids) tailored specifically for each cell line.
  • HepaRG cells underwent a differentiation protocol (DMSO induction) in both media types.
  • Cells were passaged in parallel for five generations before RNA isolation to ensure adaptation.
• Omics Approach:
  • RNA-seq: High-throughput sequencing (Illumina, 2x150bp) with ~30 million reads per sample.
  • Bioinformatics:
    • Alignment to hg38 using STAR.
    • Differential Expression Analysis (DEA) using DESeq2 (FDR < 0.05, |log2FC| > 0.5).
    • Unsupervised analysis: Principal Component Analysis (PCA) and hierarchical clustering.
    • Functional Analysis: Gene Ontology (GO:BP) enrichment, Gene Set Variation Analysis (GSVA), and ssGSEA.
  • Benchmarking: Transcriptomic similarity to PHH was quantified using Root Mean Square Error (RMSE) on curated panels for Hepatic Identity, Phase I/II Detoxification Enzymes, and Transporters.

3. Key Contributions

• Systematic Benchmarking: Provides a comprehensive, side-by-side transcriptomic comparison of two major hepatic cell lines in FBS vs. CDM, directly benchmarked against primary human hepatocytes.
• Quantification of Variance: Demonstrates that cell line identity is the primary driver of transcriptomic variance, while medium composition acts as a secondary, model-dependent modulator.
• Functional Specificity: Reveals that the biological response to CDM is not uniform; it differs significantly between HepaRG (metabolic/transport focus) and HuH7 (lipid/sterol focus).
• Validation of CDM: Offers evidence that CDM does not disrupt biological relevance but rather refines it, particularly for HepaRG cells, by reducing serum-induced noise and enhancing hepatocyte-specific programs.

4. Key Results

• Global Transcriptomic Structure:
  • PCA and clustering showed that samples grouped primarily by cell type (HepaRG vs. HuH7 vs. PHH) rather than by media type.
  • HepaRG cells clustered closer to PHH than HuH7 cells did, regardless of media.
  • The distance between CDM and FBS-SM conditions within a cell line was minimal compared to the distance between different cell lines or between cell lines and PHH.
• Differential Gene Expression (DEG):
  • HepaRG: CDM induced 2,272 DEGs (1,104 up, 1,168 down). Upregulated genes were enriched for xenobiotic metabolism, organic anion transport, and lipid metabolism. Downregulated genes were associated with immune response and cell adhesion.
  • HuH7: CDM induced 1,017 DEGs (614 up, 403 down). Upregulated genes were dominated by sterol biosynthesis and metabolic processes. Downregulated genes were linked to metal ion stress and detoxification.
• Pathway Activity (ssGSEA):
  • HepaRG in CDM showed increased activity in xenobiotic detoxification and transport pathways.
  • HuH7 in CDM showed a shift toward sterol homeostasis and stress response pathways.
• PHH Benchmarking (The "Gold Standard" Test):
  • HepaRG: Culturing in CDM significantly improved alignment with PHH across identity markers and detoxification panels (Phase I, Phase II, Transporters).
  • HuH7: Remained transcriptionally distant from PHH in both media conditions, with only marginal improvement in CDM. This suggests HuH7's limitations are intrinsic to the cell line, not just the medium.

5. Significance and Conclusion

• Reassurance for Legacy Data: The study confirms that historical FBS-based datasets remain interpretable because cell identity is the dominant biological signal. Switching to CDM does not render past data obsolete but rather clarifies the specific contribution of serum to the transcriptome.
• Optimization of HepaRG for Toxicology: The findings strongly support the use of HepaRG cells in CDM for hepatotoxicity and drug metabolism studies. This condition enhances the expression of critical hepatic functions (metabolism/transport) and brings the cell line closer to the physiological state of primary human hepatocytes.
• Model-Specific Responses: Researchers must recognize that "serum-free" does not mean "uniform." The biological response to CDM is highly cell-line dependent; while it benefits HepaRG's metabolic profile, it primarily alters lipid handling in HuH7 without conferring full hepatocyte fidelity.
• Future Direction: The paper advocates for CDM as a "controlled refinement" rather than a disruptive change, promoting reproducibility and reducing confounding variables in preclinical liver research.