henhaobio
Sign in

bioRxivPharmacology & ToxicologyPreprint

Chemical augmentation of the validated HepaRGTM CYP induction test method Part 2: Additional laboratory study supported by mRNA analysis

Emma Quartermain, Jinkang Zhang, Tim H. MarczyloiD, Timothy W. GantiD, Miriam Naomi JacobsiD

PDFDOI

Abstract

Cytochrome P450 (CYP)-mediated biotransformation of endogenous and xenobiotic substances can lead to altered exposure, toxicological impact, or adverse drug reactions. CYP induction data are fundamental to regulatory chemical toxicity hazard assessment because they directly affect the in vivo fate of xenobiotics, potentially influencing their safety and efficacy of pharmaceuticals, and impacting the safety assessment of industrial chemicals, and environmental contaminants. Here we report on the third laboratory supplementary validation of an established and previously validated human HepaRGTM in vitro method able to detect CYP1A2, CYP2B6, and CYP3A4 induction, to support the expansion of the chemical applicability domain beyond pharmaceuticals. This study was conducted to support the part 1 study with additional robust data. We established the test method in-house using the 10 previously validated pharmaceutical proficiency chemicals, then tested a further 6 proposed augmentation chemicals, tebuconazole, benfuracarb, chlorpyrifos, N, N-Diethyl-meta-toluamide, fipronil, permethrin, as tested in part 1, and then four additional chemicals: prochloraz, atrazine, pyrimethanil, and chlorpyrifos-methyl. LC-MS/MS was utilised to measure the conversion of a cocktail mixture of prototypical selective CYP probe substrates to their metabolites, in parallel with mRNA measurements. We achieved high concordance with expected classifications for proficiency and additional chemicals. Comparisons with mRNA-based measurements suggested gene expression may serve as a cost-effective pre-screening tool for CYP1A2 and CYP3A4, though with greater uncertainty for CYP2B6. The data support the robustness of the HepaRG method for CYP induction testing and the adoption of the test method in 2026 as an Organisation for Economic Cooperation and Development Test Guideline. Plain language summaryCytochrome P450 (CYP) enzymes metabolize drugs, pesticides, and other chemicals. Chemicals that increase or decrease CYP enzyme activity can change internal exposure levels, potentially leading to unexpected toxicity or impact drug effectiveness. Reliable in vitro methods to assess CYP induction are needed for regulatory chemical safety assessment. This study describes results from a third laboratory applying a previously validated human HepaRG cell-based method to assess induction of CYP1A2, CYP2B6, and CYP3A4. After successful in-house implementation using ten reference pharmaceutical compounds, the method was extended to ten more industrial chemicals. CYP induction was evaluated by measuring enzyme activity and changes in gene expression. The test method showed a high level of agreement with expected induction outcomes. Gene expression data supported enzyme activity results, particularly for CYP1A2 and CYP3A4. These results strengthen confidence in the robustness and wider applicability of the method for Organisation for Economic Cooperation and Development Test Guideline adoption.

In-depth summary

Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (20 Jun 2026) · CC BY 4.0

A third laboratory transferred the HepaRG CYP induction assay to six augmentation and four additional non-pharmaceutical chemicals; parallel mRNA data supported enzyme activity calls, with CYP3A4 most concordant.

Earlier validation used only pharmaceuticals, so some OECD national coordinators doubted the method covered industrial chemicals and adoption of the draft guideline was halted. Data from one of two laboratories in the part 1 study were not robust enough, so this third-laboratory dataset adds supporting evidence.

Key results

  1. ShownThe laboratory established the method with pharmaceutical proficiency chemicals, obtaining expected classifications for all three CYPs. Batch-level agreement with the earlier GOLIATH results was much lower for CYP2B6 than for CYP1A2 and CYP3A4.

    Reference chemicals gave at least 2-fold induction on every plate. Batch agreement vs GOLIATH was about 70% for CYP1A2/3A4 and 40% for CYP2B6. Several chemicals were equivocal for CYP1A2/2B6 under the stricter DIP. · 3.1. Positive control CYP induction; 3.2. Establishment with pharmaceutical…

  2. ShownMost of the six augmentation chemicals matched predictions for CYP1A2 and CYP3A4, with more variability for CYP2B6. Permethrin and chlorpyrifos deviated from the predicted pattern only in the equivocal category.

    Across batches, tebuconazole and benfuracarb scored positive for each of the three CYPs. Chlorpyrifos: CYP1A2 positive, CYP2B6 negative, CYP3A4 equivocal. Permethrin, predicted to be a non-inducer, was equivocal for CYP1A2 and CYP2B6, with no significance. · 3.4 Augmentation chemicals

  3. ShownThe four additional chemicals were meant to broaden the applicability domain. Prochloraz matched earlier reports for CYP2B6 and CYP3A4 and looked like a plausible CYP1A2 inducer. Atrazine was equivocal rather than the expected inducer, and the authors advise more batch testing.

    Prochloraz: CYP2B6 non-inducer, CYP3A4 equivocal. Atrazine: equivocal for CYP1A2/3A4 with no significance, weak CYP2B6 inducer. Pyrimethanil and chlorpyrifos-methyl: CYP1A2 induction supported across batches, CYP2B6 equivocal, CYP3A4 weak. · 3.5 Prochloraz, Atrazine, Pyrimethanil and…

  4. SuggestedmRNA fold change (2-fold threshold) agreed with expected classifications at about the 80% benchmark for CYP2B6 and CYP3A4 but not for CYP1A2. Authors describe it only as an exploratory pre-screen run in one laboratory.

    Concordance: CYP2B6 13/16 (81.25%), CYP3A4 15/17 (88.2%), CYP1A2 14/19 (73.7%). Prochloraz raised CYP2B6 mRNA ~10x. CYP1A2 discordances: DEET, fipronil, chlorpyrifos-methyl. Batch variability was higher for mRNA than for activity. · 3.4 Gene expression measurement performance; 4.2 Gene expression endpoint role

  5. SuggestedThe two-classifier DIP gave perfect concordance for pharmaceuticals but over-called inducers in the wider GOLIATH data. The multi-classifier DIP (positive, borderline, equivocal, negative) flags uncertain results instead.

    Two-classifier DIP: 100% concordance for CYP1A2 and CYP3A4 and 9/9 for CYP2B6 (pharmaceuticals). Between-laboratory concordance for pharmaceuticals was 80% for CYP1A2 and CYP3A4. For augmentation chemicals (INRAE vs UKHSA) it was above 80% for CYP2B6 and CYP3A4. · 4.1.1 Two classifier DIP; 4.1.2 Multi-classifier DIP: Positive…

Methods, data and reproducibility

Differentiated, cryopreserved HepaRG cells were exposed for 48 h to six concentrations per chemical in triplicate. CYP1A2, CYP2B6 and CYP3A4 activity was measured by LC-MS/MS probe metabolites normalized to protein and classified with a fold-change DIP. RT-qPCR (ΔΔCt, TBP reference) ran in parallel.

  • Cryopreserved differentiated HepaRG HPR116-TA08 cells, six batches (#308, #332, #338, #344, #345, #358), 96-well induction assays · human (HepaRG cell line) · n = three batches per chemical set; three replicates (n = 3) per batch
  • Pharmaceutical proficiency chemicals (omeprazole, carbamazepine, phenytoin, rifampicin, etc.) · human (HepaRG cell line) · n = 10 chemicals
  • Augmentation chemicals (tebuconazole, benfuracarb, chlorpyrifos, DEET, fipronil, permethrin) · human (HepaRG cell line) · n = 6 chemicals
  • Additional chemicals (prochloraz, atrazine, pyrimethanil, chlorpyrifos-methyl) · human (HepaRG cell line) · n = 4 chemicals
Methods & tools
HepaRG HPR116-TA08 cells (Biopredic), Alamar Blue cytotoxicity assay, Thermo Ultimate 3000 UPLC with Q Exactive Plus Orbitrap, Xcalibur 4.2 Quan Browser, Micro BCA protein assay, RNeasy Mini Kit, High-Capacity RNA-to-cDNA kit, QuantStudio 7 Pro RT-qPCR, PrimerQuest (IDT), GraphPad Prism v11, One-way ANOVA with Dunnett's post-hoc test, EURL ECVAM TM2009-14 SOP
Reagents & environment
Xcalibur 4.2 for LC-MS/MS data processing; GraphPad Prism v11 for plots and statistics.
Reproducibility gaps
No code or raw data repository is named; results are in supplementary tables and in companion papers listed as submitted. The revised SOP is hosted on the JRC TSAR website (test method TM2009-14).
Limitations (6)
  • mRNA measurements came from a single laboratory, were not formally validated across laboratories, and are not part of the adopted OECD guideline; they cannot support definitive classification at this stage. · stated by the authors
  • CYP2B6 and CYP1A2 show notable cell-batch variability; atrazine and some other calls need further batch testing to confirm, and mRNA batch variability exceeded that of enzyme activity. · stated by the authors
  • The design addresses induction only. Inhibition (bosentan, chlorpyrifos, artemisinin, omeprazole) was observed incidentally and would need dedicated work with known inhibitors and IC50 determination. · stated by the authors
  • The method does not currently identify test chemical metabolites, which limits some intended regulatory applications such as metabolism-pathway databases. · stated by the authors
  • Reported statistics are internally inconsistent: Dunnett's significance is written as p > 0.05, and CYP1A2 mRNA concordance appears as 73.7%, 72.2% and 'at least 80%' in different places, so the true figure is unclear. · noted in this summary
  • Each chemical was tested in three cell batches with three replicates per batch, and the mRNA concordance figures rest on small counts of chemical calls (13/16, 15/17, 14/19). · noted in this summary

Read this if you work on in vitro hepatic metabolism, OECD test guideline validation, or CYP induction classification rules and want third-laboratory performance data for non-pharmaceutical chemicals.

Citations
2

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.