Sensitization of cholangiocarcinoma cells to chemotherapy through BCRP inhibition with ß-caryophyllene oxide.

Cholangiocarcinomas (CCAs) are cancers originated in the biliary tree, which are characterized by their high mortality and marked chemoresistance, partly due to the activity of ATP-binding cassette (ABC) export pumps, whose inhibition has been proposed as a strategy for enhancing the response to chemotherapy. We have previously shown that ß-caryophyllene oxide (CRYO) acts as a chemosensitizer in hepatocellular carcinoma by inhibiting ABCB1, MRP1, and MRP2. Here, we have evaluated the usefulness of CRYO in inhibiting BCRP and improving the response of CCA to antitumor drugs. The TCGA-CHOL cohort (n = 36) was used for in silico analysis. BCRP expression (mRNA and protein) was assayed in samples from intrahepatic (iCCA) and extrahepatic (eCCA) tumors (n = 50) and CCA-derived cells (EGI-1 and TFK-1). In these cells, BCRP-dependent mitoxantrone transport was determined by flow cytometry. At non-toxic concentrations, CRYO inhibited BCRP function, which enhanced the cytostatic effect of drugs used in the treatment of CCA. The BCRP ability to confer resistance to a panel of antitumor drugs was determined in Chinese hamster ovary (CHO) cells with stable BCRP expression. At non-toxic concentrations, CRYO markedly reduced BCRP-induced resistance to known substrate drugs (mitoxantrone and SN-38) and cisplatin, gemcitabine, sorafenib, and 5-FU but not oxaliplatin. Neither CRYO nor cisplatin alone significantly affected the growth of BCRP-expressing tumors subcutaneously implanted in immunodeficient mice. In contrast, intratumor drug content was enhanced when administered together, and tumor growth was inhibited. In sum, the combined treatment of drugs exported by BCRP with CRYO can improve the response to chemotherapy in CCA patients.

Flucloxacillin Is a Weak Inducer of CYP3A4 in Healthy Adults and 3D Spheroid of Primary Human Hepatocytes.

Flucloxacillin is a widely used antibiotic. It is an agonist to the nuclear receptor PXR that regulates the expression of cytochrome P450 (CYP) enzymes. Treatment with flucloxacillin reduces warfarin efficacy and plasma concentrations of tacrolimus, voriconazole, and repaglinide. We conducted a translational study to investigate if flucloxacillin induces CYP enzymes. We also investigated if flucloxacillin induces its own metabolism as an autoinducer. We performed a randomized, unblinded, two-period, cross-over, clinical pharmacokinetic cocktail study. Twelve healthy adults completed the study. They ingested 1 g flucloxacillin 3 times daily for 31 days, and we assessed the full pharmacokinetics of the Basel cocktail drugs on days 0, 10, and 28, and plasma concentrations of flucloxacillin on days 0, 9, and 27. The 3D spheroid of primary human hepatocytes (PHHs) were exposed to flucloxacillin (concentration range: 0.15-250 µM) for 96 hours. Induction of mRNA expression, protein abundance, and enzyme activity of CYP enzymes were assessed. Flucloxacillin treatment reduced the metabolic ratio of midazolam (CYP3A4), (geometric mean ratio (GMR) 10 days (95% confidence interval (CI)): 0.75 (0.64-0.89)) and (GMR 28 days (95% CI): 0.72 (0.62-0.85)). Plasma concentrations of flucloxacillin did not change during 27 days of treatment. Flucloxacillin caused concentration-dependent induction of CYP3A4 and CYP2B6 (mRNA, protein, and activity), CYP2C9 (mRNA and protein), CYP2C19 (mRNA and activity), and CYP2D6 (activity) in 3D spheroid PHH. In conclusion, flucloxacillin is a weak inducer of CYP3A4, which may lead to clinically relevant drug-drug interactions for some narrow therapeutic range drugs that are substrates of CYP3A4.

Dicloxacillin-warfarin drug-drug interaction-A register-based study and in vitro investigations in 3D spheroid primary human hepatocytes.

AIMS: Dicloxacillin is used to treat staphylococcal infections and we have previously shown that dicloxacillin is an inducer of cytochrome P450 enzymes (CYPs). Here, we employed a translational approach to investigate the effect of a treatment with dicloxacillin on warfarin efficacy in Danish registries. Furthermore, we assessed dicloxacillin as an inducer of CYPs in vitro. METHODS: We conducted a register-based study and analysed international normalized ratio (INR) levels in chronic warfarin users before and after short- and long-term use of dicloxacillin (n = 1023) and flucloxacillin (n = 123). Induction of CYPs were investigated in a novel liver model of 3D spheroid primary human hepatocytes at the level of mRNA, and protein and enzyme activity. RESULTS: Short- and long-term dicloxacillin treatments decreased INR levels by -0.65 (95% confidence interval [CI]: -0.57 to -0.74) and -0.76 (95% CI: -0.50 to -1.02), respectively. More than 90% of individuals experienced subtherapeutic INR levels (below 2) after long-term dicloxacillin treatment. Flucloxacillin decreased INR levels by -0.37 (95% CI: -0.14 to -0.60). In 3D spheroid primary human hepatocytes, the maximal induction of CYP3A4 mRNA, protein and enzyme activity by dicloxacillin were 4.9-, 2.9- and 2.4-fold, respectively. Dicloxacillin also induced CYP2C9 mRNA by 1.7-fold. CONCLUSION: Dicloxacillin induces CYPs and reduces the clinical efficacy of warfarin in patients. This effect is substantially exacerbated during long-term treatment with dicloxacillin. The in vitro results corroborated this drug-drug interaction and correlated to the clinical findings. Caution is warranted for warfarin patients that initiate dicloxacillin or flucloxacillin, especially for a long-term treatment of endocarditis.

3D Spheroid Primary Human Hepatocytes for Prediction of Cytochrome P450 and Drug Transporter Induction.

Primary human hepatocytes (PHHs) have been the gold standard in vitro model for the human liver and are crucial to predict hepatic drug-drug interactions. The aim of this work was to assess the utility of 3D spheroid PHHs to study induction of important cytochrome P450 (CYP) enzymes and drug transporters. The 3D spheroid PHHs from three different donors were treated for 4 days with rifampicin, dicloxacillin, flucloxacillin, phenobarbital, carbamazepine, efavirenz, omeprazole, or ß-naphthoflavone. Induction of CYP1A1, CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A4, and transporters P-glycoprotein (P-gp)/ABCB1, multidrug resistance-associated protein 2 (MRP2)/ABCC2, ABCG2, organic cation transporter 1 (OCT1)/SLC22A1, SLC22A7, SLCO1B1, and SLCO1B3 were evaluated at mRNA and protein levels. Enzyme activity of CYP3A4, CYP2B6, CYP2C19, and CYP2D6 were also assessed. Induction of CYP3A4 protein and mRNA correlated well for all donors and compounds and had a maximal induction of five- to sixfold for rifampicin, which closely correlates to induction observed in clinical studies. Rifampicin induced the mRNA of CYP2B6 and CYP2C8 by 9- and 12-fold, whereas the protein levels of these CYPs reached 2- and 3-fold induction, respectively. Rifampicin induced CYP2C9 protein by 1.4-fold, whereas the induction of CYP2C9 mRNA was over 2-fold in all donors. Rifampicin induced ABCB1, ABCC2, and ABCG2 by 2-fold. In conclusion, 3D spheroid PHHs is a valid model to investigate mRNA and protein induction of hepatic drug-metabolizing enzymes and transporters, and this model provides a solid basis to study induction of CYPs and transporters, which translates to clinical relevance.

New Approach Methods for Hazard Identification: A Case Study with Azole Fungicides Affecting Molecular Targets Associated with the Adverse Outcome Pathway for Cholestasis.

Triazole fungicides such as propiconazole (Pi) or tebuconazole (Te) show hepatotoxicity in vivo, e.g., hypertrophy and vacuolization of liver cells following interaction with nuclear receptors such as PXR (pregnane-X-receptor) and CAR (constitutive androstane receptor). Accordingly, azoles affect gene expression associated with these adverse outcomes in vivo but also in human liver cells in vitro. Additionally, genes indicative of liver cholestasis are affected in vivo and in vitro. We therefore analyzed the capability of Pi and Te to cause cholestasis in an adverse outcome pathway (AOP)-driven approach in hepatic cells of human origin in vitro, considering also previous in vivo studies. Bile salt export pump (BSEP) activity assays confirmed that both azoles are weak inhibitors of BSEP. They alternate the expression of various cholestasis-associated target genes and proteins as well as the mitochondrial membrane function. Published in vivo data, however, demonstrate that neither Pi nor Te cause cholestasis in rodent bioassays. This discrepancy can be explained by the in vivo concentrations of both azoles being well below their EC50 for BSEP inhibition. From a regulatory perspective, this illustrates that toxicogenomics and human in vitro models are valuable tools to detect the potential of a substance to cause a specific type of toxicity. To come to a sound regulatory conclusion on the in vivo relevance of such a finding, results will have to be considered in a broader context also including toxicokinetics in a weight-of-evidence approach.

Okadaic acid influences xenobiotic metabolism in HepaRG cells.

Okadaic acid (OA) is an algae-produced lipophilic marine biotoxin that accumulates in the fatty tissue of filter-feeding shellfish. Ingestion of contaminated shellfish leads to the diarrheic shellfish poisoning syndrome. Furthermore, several other effects of OA like genotoxicity, liver toxicity and tumor-promoting properties have been observed, probably linked to the phosphatase-inhibiting properties of the toxin. It has been shown that at high doses OA can disrupt the physical barrier of the intestinal epithelium. As the intestine and the liver do not only constitute a physical, but also a metabolic barrier against xenobiotic exposure, we here investigated the impact of OA on the expression of cytochrome P450 (CYP) enzymes and transporter proteins in human HepaRG cells liver cells in vitro at non-cytotoxic concentrations. The interplay of OA with known CYP inducers was also studied. Data show that the expression of various xenobiotic-metabolizing CYPs was downregulated after exposure to OA. Moreover, OA was able to counteract the activation of CYPs by their inducers. A number of transporters were also mainly downregulated. Overall, we demonstrate that OA has a significant effect on xenobiotic metabolism barrier in liver cells, highlighting the possibility for interactions of OA exposure with the metabolism of drugs and xenobiotics.

Characterization of hepatic zonation in mice by mass-spectrometric and antibody-based proteomics approaches.

Periportal and perivenous hepatocytes show zonal heterogeneity in metabolism and signaling. Here, hepatic zonation in mouse liver was analyzed by non-targeted mass spectrometry (MS) and by the antibody-based DigiWest technique, yielding a comprehensive overview of protein expression in periportal and perivenous hepatocytes. Targeted immunoaffinity-based proteomics were used to substantiate findings related to drug metabolism. 165 (MS) and 82 (DigiWest) zonated proteins were identified based on the selected criteria for statistical significance, including 7 (MS) and 43 (DigiWest) proteins not identified as zonated before. New zonated proteins especially comprised kinases and phosphatases related to growth factor-dependent signaling, with mainly periportal localization. Moreover, the mainly perivenous zonation of a large panel of cytochrome P450 enzymes was characterized. DigiWest data were shown to complement the MS results, substantially improving possibilities to bioinformatically identify zonated biological processes. Data mining revealed key regulators and pathways preferentially active in either periportal or perivenous hepatocytes, with ß-catenin signaling and nuclear xeno-sensing receptors as the most prominent perivenous regulators, and several kinase- and G-protein-dependent signaling cascades active mainly in periportal hepatocytes. In summary, the present data substantially broaden our knowledge of hepatic zonation in mouse liver at the protein level.

P-Glycoprotein Inhibition Exacerbates Paclitaxel Neurotoxicity in Neurons and Patients With Cancer.

Paclitaxel-induced peripheral neuropathy (PIPN) is a common and dose-limiting adverse event. The role of P-glycoprotein (P-gp) in the neuronal efflux of paclitaxel was assessed using a translational approach. SH-SY5Y cells were differentiated to neurons and paclitaxel toxicity in the absence and presence of a P-gp inhibitor was determined. Paclitaxel caused marked dose-dependent toxicity in SH-SY5Y-derived neurons. Paclitaxel neurotoxicity was exacerbated with concomitant P-gp inhibition by valspodar and verapamil, consistent with increased intracellular accumulation of paclitaxel. Patients with cancer treated with paclitaxel and P-gp inhibitors had a 2.4-fold (95% confidence interval (CI) 1.3-4.3) increased risk of peripheral neuropathy-induced dose modification while a 4.7-fold (95% CI 1.9-11.9) increased risk for patients treated with strong P-gp inhibitors was observed, and a 7.0-fold (95% CI 2.3-21.5) increased risk in patients treated with atorvastatin. Atorvastatin also increased neurotoxicity by paclitaxel in SH-SY5Y-derived neurons. Clinicians should be aware that comedication with P-gp inhibitors may lead to increased risk of PIPN.

Polycyclic Aromatic Hydrocarbons Activate the Aryl Hydrocarbon Receptor and the Constitutive Androstane Receptor to Regulate Xenobiotic Metabolism in Human Liver Cells.

Polycyclic aromatic hydrocarbons (PAHs) are environmental pollutants produced by incomplete combustion of organic matter. They induce their own metabolism by upregulating xenobiotic-metabolizing enzymes such as cytochrome P450 monooxygenase 1A1 (CYP1A1) by activating the aryl hydrocarbon receptor (AHR). However, previous studies showed that individual PAHs may also interact with the constitutive androstane receptor (CAR). Here, we studied ten PAHs, different in carcinogenicity classification, for their potential to activate AHR- and CAR-dependent luciferase reporter genes in human liver cells. The majority of investigated PAHs activated AHR, while non-carcinogenic PAHs tended to activate CAR. We further characterized gene expression, protein abundancies and activities of the AHR targets CYP1A1 and 1A2, and the CAR target CYP2B6 in human HepaRG hepatoma cells. Enzyme induction patterns strongly resembled the profiles obtained at the receptor level, with AHR-activating PAHs inducing CYP1A1/1A2 and CAR-activating PAHs inducing CYP2B6. In summary, this study provides evidence that beside well-known activation of AHR, some PAHs also activate CAR, followed by subsequent expression of respective target genes. Furthermore, we found that an increased PAH ring number is associated with AHR activation as well as the induction of DNA double-strand breaks, whereas smaller PAHs activated CAR but showed no DNA-damaging potential.

Propiconazole is an activator of AHR and causes concentration additive effects with an established AHR ligand.

Consumers are exposed to pesticide residues and other food contaminants via the diet. Both can exert adverse effects on different target organs via the activation of nuclear receptor pathways. Hepatotoxic effects of the widely used triazole fungicide propiconazole (Pi) are generally attributed to the activation of the constitutive androstane receptor (CAR) or the pregnane X receptor (PXR). We now investigated the effects of Pi on the aryl hydrocarbon receptor (AHR) and possible mixture toxicity when Pi is present in combination with BbF, an AHR ligand. In silico docking simulations indicate that Pi can bind to human AHR. Subsequent dual luciferase reporter gene assays in human HepG2 cells showed that Pi activates the AHR in vitro. This concentration-dependent activation was confirmed by real-time RT-PCR analyses of the model AHR target genes CYP1A1 and CYP1A2 in human HepaRG and HepG2 cells. In addition, induction of CYP1A1 protein levels and enzyme activity were recorded. Similarly, increased mRNA expression and enzyme activity of Cyp1a1 and Cyp1a2 was observed in livers of rats treated with Pi for 28 days via the diet. Gene expression analysis in AHR-knockout HepaRG cells showed no induction of CYP1A1 and CYP1A2, whereas gene expression in CAR-, and PXR-knockout cells was induced. Finally, mixture effects of Pi and BbF were analyzed in human cell lines: modeling of concentration-response curves revealed concentration additivity. In conclusion, our results demonstrate that the triazole Pi is an activator of AHR in silico, in vitro and in vivo and causes additive effects with an established AHR ligand.

Direct Quantification of Cytochromes P450 and Drug Transporters-A Rapid, Targeted Mass Spectrometry-Based Immunoassay Panel for Tissues and Cell Culture Lysates.

The quantification of drug metabolizing enzymes and transporters has recently been revolutionized on the basis of targeted proteomic approaches. Isotope-labeled peptides are used as standards for the quantification of the corresponding proteins in enzymatically fragmented samples. However, hurdles in these approaches are low throughput and tedious sample prefractionation steps prior to mass spectrometry (MS) readout. We have developed an assay platform using sensitive and selective immunoprecipitation coupled with mass spectrometric readout allowing the quantification of proteins directly from whole cell lysates using less than 20,000 cells per analysis. Peptide group-specific antibodies (triple X proteomics antibodies) enable the enrichment of proteotypic peptides sharing a common terminus. These antibodies were employed to establish a MS-based immunoassay panel for the quantification of 14 cytochrome P450 (P450) enzymes and nine transporters. We analyzed the P450 enzyme and transporter levels in genotyped liver tissue homogenates and microsomes, and in samples from a time course induction experiment in human hepatocytes addressing different induction pathways. For the analysis of P450 enzymes and transporters only a minute amount of sample is required and no prefractionation is necessary, thus the assay platform bears the potential to bridge cell culture model experiments and results from whole organ tissue studies.

Variability in Mass Spectrometry-based Quantification of Clinically Relevant Drug Transporters and Drug Metabolizing Enzymes.

Many different methods are used for mass-spectrometry-based protein quantification in pharmacokinetics and systems pharmacology. It has not been established to what extent the results from these various methods are comparable. Here, we compared six different mass spectrometry-based proteomics methods by measuring the expression of clinically relevant drug transporters and metabolizing enzymes in human liver. Mean protein concentrations were in general quantified to similar levels by methods using whole tissue lysates. Methods using subcellular membrane fractionation gave incomplete enrichment of the proteins. When the enriched proteins were adjusted to levels in whole tissue lysates, they were on average 4-fold lower than those quantified directly in whole tissue lysates. The differences in protein levels were propagated into differences in predictions of hepatic clearance. In conclusion, caution is needed when comparing and applying quantitative proteomics data obtained with different methods, especially since membrane fractionation is common practice for protein quantification used in drug clearance predictions.