Sulforaphane and its analogues inhibit CYP1A1 and CYP1A2 activity induced by benzo[a]pyrene

CYP1A1 and CYP1A2 enzymes metabolize polycyclic aromatic hydrocarbons (PAHs) to the reactive oxyderivatives. PAHs can induce the activity of both enzymes, which increases its conversion and enhances risk of carcinogenesis. Thus, the inhibition of CYP enzymes is recognized as a cancer chemoprevention strategy. A well‐known group of chemopreventive agents is isothiocyanates, which occur naturally in Brassica vegetables. In this paper, a naturally occurring sulforaphane and its two synthetic analogues isothiocyanate‐2‐oxohexyl and alyssin were investigated. The aim of the study was to determine whether the differences in the isothiocyanate structure change its ability to inhibit CYP1A1 and CYP1A2 activity induced by benzo[a]pyrene in HepG2 and Mcf7 cells. Also a mechanistic study was performed including isothiocyanates' influence on CYP1A1 and CYP1A2 catalytic activity, enzymatic protein level, and AhR translocation. It was shown that both enzymes were significantly induced by benzo[a]pyrene, and isothiocyanates were capable of decreasing the induced activity. The inhibitory properties depend on the types of isothiocyanate and enzyme. In general, CYP1A2 was altered in the more meaningful way than CYP1A1 by isothiocyanates. Sulforaphane exhibited weak inhibitory properties, whereas both analogues were capable of inhibiting BaP‐induced activity with the similar efficacy. The mechanistic study revealed that analogues decreased the CYP1A2 activity via the protein‐level reduction and CYP1A1 directly. The results indicate that isothiocyanates can be considered as potent chemopreventive substances and the change in the sulforaphane structure increases its chemopreventive potency. © 2009 Wiley Periodicals, Inc. J Biochem Mol Toxicol 23:18–28, 2009; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/jbt.20259     

Silybin and dehydrosilybin decrease glucose uptake by inhibiting GLUT proteins

Silybin, the major flavonoid of Silybum marianum, is widely used to treat liver diseases such as hepatocellular carcinoma and cirrhosis-associated insulin resistance. Research so far has focused on its anti-oxidant properties. Here, we demonstrate that silybin and its derivative dehydrosilybin inhibit glucose uptake in several model systems. Both flavonoids dose-dependently reduce basal and insulin-dependent glucose uptake of 3T3-L1 adipocytes, with dehydrosilybin showing significantly stronger inhibition. However, insulin signaling was not impaired, and immunofluorescence and subcellular fractionation showed that insulin-induced translocation of GLUT4 to the plasma membrane is also unchanged. Likewise, hexokinase activity was not affected suggesting that silybin and dehydrosilybin interfere directly with glucose transport across the PM. Expression of GLUT4 in CHO cells counteracted the inhibition of glucose uptake by both flavonoids. Moreover, treatment of CHO cells with silybin and dehydrosilybin reduced cell viability which was partially rescued by GLUT4 expression. Kinetic analysis revealed that silybin and dehydrosilybin inhibit GLUT4-mediated glucose transport in a competitive manner with K(i)=60 and 116 μM, respectively. We conclude that silybin and dehydrosilybin inhibit cellular glucose uptake by directly interacting with GLUT transporters. Glucose starvation offers a novel explanation for the anti-cancer effects of silybin.     

Wild-type CYP102A1 as a biocatalyst: turnover of drugs usually metabolised by human liver enzymes

This work provides functional data showing that the bacterial CYP102A1 recognises compounds metabolised by human CYP3A4, CYP2E1 and CYP1A2 and is able to catalyse different reactions. Wild-type cytochrome CYP102A1 from Bacillus megaterium is a catalytically self-sufficient enzyme, containing an NADPH-dependent reductase and a P450 haem domain fused in a single polypeptidie chain. An NADPH-dependent method (Tsotsou et al. in Biosens. Bioelectron. 17:119–131, 2002) together with spectroscopic assays were applied to investigate the catalytic activity of CYP102A1 towards 19 xenobiotics, including 17 commercial drugs. These molecules were chosen to represent typical substrates of the five main families of drug-metabolising human cytochromes P450. Liquid chromatography–mass spectrometry analysis showed that CYP102A1 catalyses the hydroxylation of chlorzoxazone, aniline and p-nitrophenol, as well as the N-dealkylation of propranolol and the dehydrogenation of nifedipine. These drugs are typical substrates of human CYP2E1 and CYP3A4. The K _M values calculated for these compounds were in the millimolar range: 1.21 ± 0.07 mM for chlorzoxazone, 2.52 ± 0.08 mM for aniline, 0.81 ± 0.04 mM for propranolol. The values of v _max for chlorzoxazone and propranolol were 46.0 ± 9.0 and 7.6 ± 3.4 nmol min⁻¹ nmol⁻¹, respectively. These values are higher then those measured for the human enzymes. The v _max value for aniline was 9.4 ± 1.3 nmol min⁻¹ nmol⁻¹, comparable to that calculated for human cytochromes P450. The functional data were found to be in line with the sequence alignments, showing that the identity percentage of CYP102A1 with CYP3A4 and CYP2E1 is higher than that found for CYP1A2, CYP2C9 and CYP2D6 families.     

Lead and catechol hematotoxicity in vitro using human and murine hematopoietic progenitor cells

In vitro cloning assays for hematopoietic myeloid and erythroid precursor cells have been used as screening systems to investigate the hematotoxic potential of environmental chemicals in humans and mice. Granulocyte-monocyte progenitors (CFU-GM) from human umbilical cord blood and from mouse bone marrow (Balb/c and B6C3F1) were cultured in the presence of lead and the benzene metabolite catechol. Erythroid precursors (BFU-E) from human umbilical cord blood were cultured in the presence of lead. The in vitro exposure of the human and murine cells resulted in a dose-dependent depression of the colony numbers. The concentration–effect relationship was studied. Results showed that: (1) Based on calculated IC50 values, human progenitors are more sensitive to lead and catechol than are murine progenitors. The dose that caused a 50% decrease in colony formation after catechol exposure was 6 times higher for murine cells (IC50 = 24 μmol/L) than for human cord blood cells (IC50 = 4 μmol/L). Lead was 10–15 times more toxic to human hematopoietic cells (IC50 = 61 μmol/L) than to murine bone marrow cells from both mice strains tested (Balb/c, IC50 = 1060 μmol/L; B6C3F1, IC50 = 536 μmol/L). (2) A lineage specificity was observed after exposure to lead. Human erythroid progenitors (hBFU-E) (IC50 = 3.31 μmol/L) were found to be 20 times more sensitive to the inhibitory effect of lead than were myeloid precursors (hCFU-GM) (IC50 = 63.58 μmol/L). (3) Individual differences in the susceptibility to the harmful effect of lead were seen among cord blood samples. (4) Toxicity of lead to progenitor cells occurred at environmentally relevant concentrations. This revised version was published online in July 2006 with corrections to the Cover Date.     

Cell-specific differences in the susceptibility of potential cellular targets of human origin derived from blood and lung following treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)

The induction of cytochrome P4501A (CYP1A1) enzyme activity is one of the best-studied direct effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related compounds and has been shown to be a sensitive biomarker of exposure to polycyclic aromatic hydrocarbons (PAH) in different experimental animal species as well as in humans. TCDD has also been shown to modulate cytokine gene expression in human keratinocytes, including IL-1, TGF- and TFG-₂. In the present studies, the aim was to determine whether different cellular targets of human origin differed in susceptibility to TCDD as measured by CYP1A1 activity and mRNA expression, and whether cytokine gene induction/suppression correlated with TCDD susceptibility. Human airway epithelial cells, alveolar macrophages (AM), peripheral blood monocytes and lymphocytes (PBL) were exposed to 10^-10–10^-7 mol/L TCDD. CYP1A1 enzyme activity was determined by ethoxyresorufin-O-deethylase (EROD) activity, mRNA expression of CYP1A1 was measured by semiquantitative PCR assay. The secretion and/or gene expression of specific cytokines, including IL-6, IL-8, and IL-1 were also examined. Overall, there was a clear correlation between TCDD-induced enzyme activity and CYP1A1 mRNA levels, which were dose-dependently increased in the bronchoepithelial cells and PBL. The human airway epithelial cells (BEAS-S6 cell line and primary cells) appeared to be the most inducible cellular target, with up to 50-fold increases at 10^-8 mol/L TCDD with an EC₅₀ of 3×10^-11 mol/L TCDD. The pokeweed mitogen-activated peripheral blood lymphocytes revealed approximately 5-fold less capacity in CYP1A1 activity, with high interindividual variabilities (EC₅₀ 3×10^-9 mol/L TCDD). In contrast, CYP1A1 enzyme activity in both AM and purified peripheral blood monocytes, which were costimulated with LPS and/or GM-CSF, could not be detected. CYP1A1 mRNA levels, however, were detectable and only marginally enhanced in response to TCDD. The ability of all these cells to express and produce the proinflammatory cytokines IL-6 and IL-8 was neither enhanced nor impaired by TCDD. These results indicate that cell types found in human lung and peripheral blood vary in susceptibility to TCDD, with the lung epithelium being highly susceptible and the alveolar macrophage being nonsusceptible. However, expression and production of specific cytokines such as IL-6 and IL-8, which may potentiate inflammatory processes and/or work as mitogens, does not appear to be influenced by TCDD.     

Effect of silybin and its glycosides on the expression of cytochromes P450 1A2 and 3A4 in primary cultures of human hepatocytes

Four beta-glycosides of flavonoligan silybin, i.e. silybin beta-galactoside, silybin beta-glucoside, silybin beta-maltoside, silybin beta-lactoside were synthesized in order to improve silybin water solubility and bioavailability (Kren et al., J Chem Soc, Perkin Trans 1, 2467-2474, 1997). The presented paper deals with the effect of silybin and its synthetic beta-glycosides on the expression of two major cytochrome P450 isoforms, CYP1A2 and CYP3A4. Primary cultures of human hepatocytes were the model of choice. mRNAs were analyzed using Northern blot and P-radiolabelled probes. CYP protein content was determined by immunoblotting using specific antibodies. Silybin and its beta-glycosides do not induce expression of CYP1A2 and CYP3A4. Tested compounds did not affect inducible expression of CYP1A2 and CYP3A4 by dioxin and rifampicin, respectively, as evaluated at the level of mRNAs and proteins. Silybin and its beta-glycosides do not interfere with the expression of CYP1A2 and CYP3A4, are not likely to produce drug-drug interactions in terms of the inducibility of two important cytochromes P450.     

Characterization of human CYP1A1/1A2 induction by DNA microarray and alpha-naphthoflavone

DNA microarrays and real time PCR were used to analyze the mechanism of gene induction by CYP1A1 inducers, beta-naphthoflavone, and omeprazole, in the human hepatocellular carcinoma HepG2 cells. Reproducible and significant inductions were observed in a limited number of genes including CYP1A1 and CYP1A2. Genes induced by omeprazole included several protein tyrosine kinase targets. This result confirmed that omeprazole could modulate gene expressions through protein tyrosine kinase-mediated pathway. Induction ratios were considerably different from CYP1A1 and CYP1A2 (>10-fold) to other induced genes (<5-fold). alpha-Naphthoflavone, which is known as an antagonist to 2,3,7,8-tetrachlorodibenzo-p-dioxin, inhibited the inductions of heme oxygenase 1, glutamate-cysteine ligase (modifier unit), and thioredoxin reductase by beta-naphthoflavone but not those of CYP1A1 and CYP1A2. It unexpectedly enhanced the beta-naphthoflavone-mediated CYP1A1 and CYP1A2 induction. These results suggest that the CYP1A1 and CYP1A2 genes, which share their 5(') enhancer regions, are regulated differently from the other genes.     

Comparison of tetrahydrofuran,fetal calf serum,and Tween 40 for the delivery of astaxanthin and canthaxanthin to HepG2 cells

The present investigation aimed to compare fetal calf serum (FCS) and Tween 40 with the commonly employed tetrahydrofuran (THF) with respect to cytotoxicity, stability of the solubilized carotenoids, and uptake and accumulation of the xanthophylls astaxanthin (AX) and canthaxanthin (CX) in cultured human liver cells (HepG2). Incubation of HepG2 cells for 24 h with THF (≥1.25%) or FCS (≥11.25%) with or without AX (≥25 μmol/L) or CX (≥25 μmol/L) did not affect cell viability. Tween 40 (0.25–1.25% in medium) reduced cell viability by 75–99%. The stabilities of AX and CX in cell-free RPMI 1640 medium for ≤24 h were higher when delivered with THF instead of FCS. The dose- and time-dependent accumulations of AX and CX (1–10 μmol/L) in HepG2 cells were higher when carotenoids were delivered with FCS compared to THF. In conclusion, FCS and THF, but not Tween 40, were suitable solvent systems for the delivery of AX and CX to HepG2 cells. In our experiments FCS was superior with regard to the uptake and accumulation of both carotenoids.     

Preferential inducibility of CYP1A1 and CYP1A2 by TCDD: differential regulation in primary human hepatocytes versus transformed human cells

Cytochrome P4501A1 (CYP1A1) induction, a marker of aryl hydrocarbon (Ah) receptor activation, has been associated with carcinogenicity of the environmental agent 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Consistently, we show that TCDD treatment led to induction of CYP1A1 in responsive human cancer cell lines including HepG2, LS174T, and MCF-7, as determined by Western blotting and CYP1A form-selective R-warfarin 6- and 8-hydroxylation. TCDD, however, preferably induced CYP1A2, not CYP1A1, in primary human hepatocytes. Such CYP1A form-preferred induction at the protein level was apparently uncorrelated with non-preferred mRNA induction in any cells studied. Moreover, while both genes were up-regulated by TCDD in primary hepatocytes and HepG2 cells, the induction of CYP1A1 and CYP1A2 at the mRNA level was distinguishable, indicated by the marked differences in activation kinetics and the response to the protein synthesis inhibitors, anisomycin and cycloheximide. Furthermore, formation of total benzo(a)pyrene (BaP)-DNA adducts was not altered following BaP exposure in TCDD-treated primary hepatocytes, whereas significantly elevated, in a CYP1A1-dependent manner, in the treated HepG2 cells. Taken together, our findings, demonstrating the complexities of TCDD-associated human Ah receptor function and differential regulations of CYP 1A enzymes, suggest clearly the need for caution when extrapolating data obtained in cell-based models.     

Induction of Human UDP-glucuronosyltransferase 1A1 by Cortisol-GR

During the course of the study of UGT1A1 induction by bilirubin, we could not detect the induction of the reporter gene (−3174/+14) of human UGT1A1 in HepG2 by bilirubin (Mol. Biol. Rep. 31: 151–158 (2004)). In this report, we show the finding of the induction of the reporter gene of UGT1A1 by cortisol at 1 μM, a major natural cortico-steroid, with human glucocorticoid receptor (GR). RU486 of a typical GR antagonist at 10 μM inhibited the induction by cortisol from 5.9- to 1.8-fold. This result indicates that the induction by cortisol-GR is dependence on ligand-binding. This induction is caused by the UGT reporter gene itself, from the results of noinduction with control vector pGL2 (equal to pGV-C) in the presence of cortisol-GR. We confirmed that the induction of the reporter gene by cortisol is dependent on the position of proximal element (−97/−53) of UGT1A1. From this result, we concluded that the increase of corticosteroid in neonates must induce the elevation of UGT1A1 after birth and prevent jaundice. With the study of induction by corisol, we studied the influence of co-expression of PXR (pregnenolone xenobiotic receptor) with the UGT1A1 reporter gene and we could not find the induction of UGT1A1 expression in the presence of dexamethasone, rifampicin, or pregnenolone 16α-carbonitrile of the PXR ligands. These results suggest that the induction of UGT1A1 expression by GR is not mediated by PXR, unlike the induction of CYP3A4 through PXR.     

Inhibition of CYP3A4 and CYP2C9 by podophyllotoxin: Implication for clinical drug–drug interactions

Podophyllotoxin (PPT) and its derivatives exert significant anti-cancer activities, and one derivative etoposide is often utilized to treat various cancers in the clinic. The aim of the present study is to investigate the inhibitory effects of PPT on major cytochrome P450 (CYP) isoforms in human livers. Inhibition of CYP3A4, CYP2C9, CYP2C8, CYP2D6, CYP2E1 and CYP2A6 by PPT was investigated in the human liver microsomal system. Time-dependent inhibition of CYP3A4 by PPT was also evaluated. The results showed that PPT strongly exhibited inhibitory effects on CYP3A4 and CYP2C9 in a concentration-dependent manner. Half inhibition concentration (IC₅₀) was 1.1 ± 0.3 and 4.6 ± 0.3 μM for CYP3A4 and CYP2C9, respectively. Inhibition kinetic analysis showed that PPT exhibited competitive inhibition towards CYP3A4 and CYP2C9 with K_i of 1.6 and 2.0 μM, respectively. Additionally, PPT exerted time-dependent inhibition towards CYP3A4 and the kinetic parameters were 4.4 ± 2.1 μM and 0.06 ± 0.01 min^–1 for K_I and k_inact, respectively. Our experimental data indicate that potential drug–drug interaction (DDI) might exist when PPT is co-administered with the substrates which mainly undergo CYP3A4- or CYP2C9-mediated metabolism.     

Characterization of adjacent E-box and nuclear factor 1-like DNA binding sequence in the human CYP1A2 promoter

To better understand the molecular mechanisms of cytochrome P450 1A2 (CYP1A2) regulation, we have characterized a region of the promoter (+3 to -176) that contains a single E-box and an adjacent nuclear factor 1 (NF1)-like DNA binding site. The E-box was shown to specifically bind nuclear proteins that were recognized by antibodies against upstream stimulatory factor (USF) 1 and 2. Comparison of NF1 binding proteins in HepG2 cells and primary cultures of rat hepatocytes revealed different patterns of DNA-protein complexes, all of which were recognized by a general NF1 antibody. Mutations of the E-box resulted in substantial reduction of promoter activity in either primary hepatocytes or HepG2 cells regardless of the presence in the reporter constructs of other CYP1A2 regulatory elements, such as the hepatic nuclear factor 1 (HNF-1) binding site. In contrast, reporter gene activity of the promoter construct harboring the mutated NF1-like binding site was affected by upstream sequences when transfected into HepG2 cells, but not in primary hepatocytes. We conclude that both USF proteins and different isoforms of NF1 contribute to the constitutive expression of CYP1A2.     

In vitro assessment of inhibition by natural polyphenols of metabolic activation of procarcinogens by human CYP1A1

Previously, inhibitors of CYP1A1 were rated as candidate chemopreventive agents against cancer mainly according to their effects on the 7-ethoxyresorufin O-deethylation (EROD) of diagnostic probe substrates. Surprisingly, several polyphenols including resveratrol, formerly identified as potent inhibitors by the EROD assay, exhibited no or weak inhibition of procarcinogen activation. We compared the effects of 11 representative natural polyphenols, which normally occur in food, on different activities of CYP1A1, namely epoxidation of 7,8-dihydrodiol-benzo[a]pyrene, the terminal step in the activation leading to the ultimate carcinogenic diolepoxides, hydroxylation of benzo[a]pyrene, and EROD. For the first time, a reconstituted system was used for the determination of IC(50) values, consisting of purified enzymes (human CYP1A1 and human NADPH-cytochrome P450 reductase) and dilaurylphosphatidylcholine. The results demonstrate that the inhibitory effects of dietary polyphenols on CYP1A1 activity depend on both the structure of the inhibitor and the type of the reaction and substrate used in the assay. Consequently, a potent EROD inhibition alone is insufficient to count a substance among the chemoprotective agents.     

Sigma1 binding in a human neuroblastoma cell line

Behaviorally, sigma₁ agents modulate opioid analgesia. To examine possible mechanisms responsible for these interactions, we have identified a cell line containing both sigma₁ and opioid receptors. [³H](+)-pentazocine binding in BE(2)-C human neuroblastoma cells is high affinity (K_D 3.4±0.7 nM) and high density (B_max 2.98±0.14 pmol/mg protein). Competition studies reveal a selectivity profile similar to that of sigma₁ sites in guinea pig brain. (+)-Pentazocine has no effect upon either basal or forskolin-stimulated cyclase in the BE(2)-C cells, but cAMP accumulation is inhibited by the morphine, DPDPE and naloxone benzoylhydrazone. (+)-Pentazocine at concentrations as high as 10 μM does not affect this opioid effect, implying that sigma₁/opioid interactions are not mediated at the level of the cell. This suggests that their behavioral interactions result from interacting neural circuits. Although (+)-pentazocine is without effect in the cyclase system, it does block carbachol-stimulated phosphoinositol turnover (IC₅₀ 6.5±1.14 μM). The specificity of the effect is confirmed by the ability of haloperidol (1 μM) to shift the IC₅₀ value of (+)-pentazocine 2-fold to the right. Special issue dedicated to Dr. Eric J. Simon.