Collateral sensitivity of resistant MRP1-overexpressing cells to flavonoids and derivatives through GSH efflux

The multidrug resistance protein 1 (MRP1) is involved in multidrug resistance of cancer cells by mediating drug efflux out of cells, often in co-transport with glutathione (GSH). GSH efflux mediated by MRP1 can be stimulated by verapamil. In cells overexpressing MRP1, we have previously shown that verapamil induced a huge intracellular GSH depletion which triggered apoptosis of the cells. That phenomenon takes place in the more global anticancer strategy called “collateral sensitivity” and could be exploited to eradicate some chemoresistant cancer cells. Seeking alternative compounds to verapamil, we screened a library of natural flavonoids and synthetic derivatives. A large number of these compounds stimulate MRP1-mediated GSH efflux and the most active ones have been evaluated for their cytotoxic effect on MRP1-overexpressing cells versus parental cells. Interestingly, some are highly and selectively cytotoxic for MRP1-cells, leading them to apoptosis. However, some others do not exhibit any cytotoxicity while promoting a strong GSH efflux, indicating that GSH efflux is necessary but not sufficient for MRP1-cells apoptosis. In support to this hypothesis, structure activity relationships show that the absence of a hydroxyl group at position 3 of the flavonoid C ring is an absolute requirement for induction of MRP1-cells death, but is not for GSH efflux stimulation. Chrysin (compound 8) and its derivatives, compounds 11 and 22, exhibit a high selectivity toward MRP1-cells with a IC₅₀ value of 4.1 μM for compound 11 and 4.9 μM for chrysin and compound 22, making them among the best described selective killer compounds of multidrug ABC transporter-overexpressing cells.     

Non-destructive micromethod for MRP1 functional assay in human lung tumor cells

Defense against toxic endo- and xenobiotics is a major concern of all living species and ABC transporters play a vital role in this defense system. Multidrug resistance associated proteins 1 (MRP1) is a cellular detoxifying factor supposed to transport a wide range of compounds across cell membranes either as GSH conjugates or as co-transport accompanying glutathione transposition. The cellular localization of MRP1 is a determining factor whether the transport function can take place. In this study we have undertaken experiments on the transport activity of MRP1 in cultured human lung tumor cells in order to check whether MRP1 is expressed as a functionally active protein. For this purpose we have adapted a quantitative fluorescence imaging assay to conditions where a small number of attached cells should be repeatedly measured by a non-destructive method. In cultured A549, H358 and H322 cells MRP1 is located in the cell membrane as observed by immunocytochemistry. Efflux of 5,6-carboxy-2′-7′-dichloro-fluorescein (CDF) from lung cells was sensitive toward the MRP1 inhibitor MK571 while verapamil had no effect. On the other hand, efflux of Rhodamin 123, a Pgp-glycoprotein substrate, from lung cells reacted to inhibition by verapamil, while MK571 had no effect. Modulation of glutathion content of lung cells by N-acetyl cystein and buthionine sulfoximine shifted CDF efflux toward higher or lower rates, respectively. These experiments confirm that MRP1 function can be followed in the attached cells in vitro under non-toxic concentrations of the substrates without the need to harvest and destroy the cells. Abdelrahman Torky and Ekkehard Stehfest have contributed equally to this study     

Resistance of human multidrug resistance-associated protein 1-overexpressing lung tumor cells to the anticancer drug arsenic trioxide

The human multidrug-resistance protein (MRP1) confers resistance to some heavy metals such as arsenic and antimony, mainly through mediating an increased cellular efflux of metal. However, it was recently suggested that arsenic, used under its trioxide derivative form as anticancer drug, is not handled by MRP1. The aim of the present study was to test this hypothesis in MRP1-overexpressing human lung tumor GLC4/Sb30 cells. Using the cytotoxicity MTT assay, GLC4/Sb30 cells were found to be 10.8-fold more resistant to arsenic trioxide (As2O3) than parental GLC4 cells. MK571, a potent inhibitor of MRP1 activity, almost totally reversed resistance of GLC4/Sb30 cells, but did not alter the sensitivity of GLC4 cells. Moreover, As2O3-loaded GLC4/Sb30 cells poorly accumulated arsenic through an increased MK571-sensitive efflux of metal. Finally, depletion of cellular glutathione levels in buthionine sulfoximine-treated GLC4/Sb30 cells was found to result in increased accumulation and reduced efflux of arsenic in cells exposed to As2O3, outlining the glutathione-dependence of MRP1-mediated transport of the metal. These results indicate that MRP1 overexpression in human tumor cells can confer resistance to As2O3, which may limit the clinical use of this anticancer drug for treatment of MRP1-positive tumors.     

Expression of MRP1 and related transporters in human lung cells in culture

Multidrug resistance type 1 P-glycoproteins (P-gp) and multidrug resistance associated proteins (MRP) were studied in differentiated primary human lung cells in culture, in comparison with permanent human lung cell lines and primary alveolar type II cells from rat lung. AII cells exhibited low basal levels of mdr1b mRNA, that increased over time and after oxygen radical production induced by paraquat. mRNAs coding for antioxidative enzymes catalase (CAT), maganese superoxide dismutase (Mn-SOD) and copper/zinc superoxide dismutase (Cu/Zn-SOD) were not changed. H358, A549, H322 cells expressed low levels of MDR1 mRNA, but the mdr1 substrate rhodamine 123 (Rh 123) was transported out of H358 and H322 cells in a non-invasive, single cell fluorescence assay. The dye efflux could be inhibited by the chemosensitizer, verapamil. Normal human bronchial epithelial cells (NHBEC) expressed immuno-reactive MDR1 P-gp and the MPR protein that was active in the fluorescence assay using the MRP substrate carboxy-dichlorofluorescein (CDF) and MK-571 as an inhibitor. We did observe inter-individual variation of MRP in both the mRNA and the immunoreactive protein in NHBEC culture. Over time (12 weeks) the protein was relatively stable in NHBEC and epithelial cells from peripheral lung (PLC), but the mRNA level was drastically increased when explant cultures were continued (18 weeks).     

Differential cell-specific cytotoxic responses of oral cavity cells to tobacco preparations

To examine the effects of standardized (reference) tobacco preparations on human oral cavity cells, two oral squamous cell carcinoma cell lines (101A, 101B) and normal human gingival epithelial cells (HGEC) were treated with cigarette smoke total particulate matter (TPM), smokeless tobacco extracted with complete artificial saliva (ST/CAS), or whole-smoke conditioned media (WS-CM). EC-50 values, as determined by sulforhodamine B assays, varied among the cell types and agents. When normalized to nicotine content, cytotoxicity for WS-CM and TPM was higher compared to that observed with ST/CAS. Nicotine alone had no or only minimal cytotoxicity for all cell types in the applied range.Activation of pro-apoptotic caspase-3 was examined in all cell types at their respective EC-50 doses for the three agents. TPM, but not ST/CAS or WS-CM significantly activated caspase-3 in all three cell types. Fluorescence-activated cell sorting (FACS) for expression of the early apoptosis marker Annexin V and for nuclear staining by 7-aminoactinomycin (7-AAD) revealed different extents of apoptosis versus non-apoptotic cell death for the three agents.These data characterize differential responses of normal and malignant oral cells after exposure to TPM, ST/CAS, or WS-CM. They assist in understanding differential effects of combustible versus non-combustible tobacco products, and in identifying novel biomarkers for tobacco smoke exposure and effect in the oral cavity.     

Differential sensitivities of the MRP gene family and gamma-glutamylcysteine synthetase to prooxidants in human colorectal carcinoma cell lines with different p53 status

1Recent molecular cloning studies have identified six members in the multidrug-resistance protein (MRP) gene family. However, the regulation of expression of these genes is largely unknown. We previously reported that expression of MRP1, encoding multidrug-resistance associated protein, and gamma-GCSh, which encodes the heavy subunit of gamma-glutamylcysteine synthetase (gamma-GCS), could be up-regulated by prooxidants [Yamane et al., J Biol Chem 1998;273:31075-85]. In the present study, we investigated whether different members of the MRP family exhibit different responses to induction by prooxidants, and whether p53 status influences the levels of induction. A panel of colorectal cancer cell lines with different p53 status, i.e. HCT116 containing wild-type p53, and HT29, SW480, and Caco2 containing mutant p53, was treated with tert-butylhydroquinone (t-BHQ) and pyrrolidinedithiocarbamate (PDTC). MRP1 and gamma-GCSh mRNA levels were determined by the RNase protection assay, using gene-specific probes. We report here that induction of MRP1 and gamma-GCSh expression by these prooxidants varied among the different cell lines, and p53 mutations were not always associated with elevated levels of induction. These results suggest that the effects of p53 on the induced expression of MRP1 and gamma-GCSh depend on the environment of the cell and/or nature of p53 mutations. In an isogenic HCT116 cell line containing p53(-/-) alleles, we demonstrated that, as for MRP1, expression of MRP2 and MRP3 was induced by the prooxidants, whereas expression of MRP4 and MRP5 was not. MRP6 mRNA was not detectable. Induction of MRP2 expression by prooxidants seemed to be independent of p53 status. Our results demonstrated the differential regulation of the MRP gene family by p53 mutation under oxidative stress.     

Particulate depleted uranium is cytotoxic and clastogenic to human lung cells

Depleted uranium (DU) is commonly used in military armor and munitions, and thus, exposure of soldiers and non-combatants is potentially frequent and widespread. DU is considered a suspected human carcinogen, affecting the bronchial cells of the lung. However, few investigations have studied DU in human bronchial cells. Accordingly, we determined the cytotoxicity and clastogenicity of both particulate (water-insoluble) and soluble DU in human bronchial fibroblasts (WTHBF-6 cells). We used uranium trioxide (UO3) and uranyl acetate (UA) as prototypical particulate and soluble DU salts, respectively. After a 24 h exposure, both UO3 and UA induced concentration-dependent cytotoxicity in WTHBF-6 cells. Specifically, 0.1, 0.5, 1, and 5 microg/cm2 UO3 induced 99, 57, 32, and 1% relative survival, respectively. Similarly, 100, 200, 400, and 800 microM UA induced 98, 92, 70, and 56% relative survival, respectively. When treated with chronic exposure, up to 72 h, of either UO3 or UA, there was an increased degree of cytotoxicity. We assessed the clastogenicity of these compounds and found that at concentrations of 0, 0.5, 1, and 5 microg/cm2 UO3, 5, 6, 10, and 15% of metaphase cells exhibit some form of chromosome damage. UA did not induce chromosome damage above background levels. There were slight increases in chromosome damage induced when we extended the UO3 treatment time to 48 or 72 h, but no meaningful increase in chromosome damage was observed with chronic exposure to UA.     

Muramyl peptides augment cytotoxic effect of tumor necrosis factor-alpha in combination with cytotoxic drugs on tumor cells

We have demonstrated that biologically active muramyl peptides, in particular, glucosaminylmuramyl dipeptide (GMDP), augmented in vitro cytotoxic activity of tumor necrosis factor-alpha (TNF-alpha) against murine fibrosarcoma L929 cells. The introduction of GMDP resulted in cytotoxic effect characteristic for substantially higher dose of cytokine. Even more potent was the combination of GMDP, TNF-alpha and Actinomycin D (ActD). According to clonogenic and MTT assays 100% L929 cells could be killed in culture with low doses of TNF-alpha and ActD if GMDP was present. When cisplatin was substituted for ActD similar results were obtained. GMDP also enhanced cytotoxicity of TNF-alpha and cisplatin against human breast carcinoma MCF7 and histiocytic lymphoma U937 cells. Normal cells, namely human peripheral blood leucocytes and murine peritoneal macrophages, were resistant to selected doses of TNF-alpha/cisplatin/GMDP.     

Differential cytotoxic responses of PC12 cells chronically exposed to psychostimulants or to hydrogen peroxide

Repeated abuse of stimulant drugs, cocaine and amphetamine, is associated with extraneuronal dopamine accumulation in specific brain areas. Dopamine may be cytotoxic through the generation of reactive oxygen species, namely hydrogen peroxide (H2O2), resulting from dopamine oxidative metabolism. In this work, we studied the cytotoxicity in PC12 cells (a dopaminergic neuronal model) chronically and/or acutely exposed to cocaine or amphetamine, as compared to H2O2 exposure. Chronic cocaine treatment induced sensitization to acute cocaine insult and increased cocaine-evoked accumulation of extracellular dopamine, although no changes in dihydroxyphenylacetic acid (DOPAC) levels were observed. Moreover, dopamine was depleted in cells chronically exposed to amphetamine and acute amphetamine toxicity persisted in these cells, indicating that dopamine was not involved in amphetamine cytotoxicity. PC12 cells chronically treated with H2O2 were totally resistant to acute H2O2, but not to acute cocaine or amphetamine exposure, suggesting that the toxicity induced by these stimulant drugs is unrelated to adaptation to oxidative stress. Interestingly, chronic cocaine treatment largely, but not completely, protected the cells against a H2O2 challenge, whilst a decrement in intracellular ATP was observed. This study shows that chronic treatment of PC12 cells with cocaine or H2O2 modifies the cytotoxic response to an acute exposure to these agents.     

The sulphonylurea glibenclamide inhibits multidrug resistance protein (MRP1) activity in human lung cancer cells

1. Glibenclamide, a sulphonylurea widely used for the treatment of non-insulin-dependent diabetes mellitus, has been shown to inhibit the activities of various ATP-binding cassette (ABC) transporters. In the present study, its effects towards multidrug resistance protein 1 (MRP1), an ABC efflux pump conferring multidrug resistance and handling organic anions, were investigated. 2. Intracellular accumulation of calcein, an anionic dye substrate for MRP1, was strongly increased by glibenclamide in a dose-dependent manner in MRP1-overexpressing lung tumour GLC4/Sb30 cells through inhibition of MRP1-related calcein efflux. By contrast, glibenclamide did not alter calcein levels in parental control GLC4 cells. Another sulphonylurea, tolbutamide, was however without effect on calcein accumulation in both GLC4/Sb30 and GLC4 cells. 3. Glibenclamide used at 12.5 microM was, moreover, found to strongly enhance the sensitivity of GLC4/Sb30 cells towards vincristine, an anticancer drug handled by MRP1. 4. Efflux of carboxy-2',7'-dichlorofluorescein, an anionic dye handled by the ABC transporter MRP2 sharing numerous substrates with MRP1 and expressed at high levels in liver, was also strongly inhibited by glibenclamide in isolated rat hepatocytes. 5. In summary, glibenclamide reversed MRP1-mediated drug resistance likely through inhibiting MRP1 activity and blocked organic anion efflux from MRP2-expressing hepatocytes. Such effects associated with the known inhibitory properties of glibenclamide towards various others ABC proteins suggest that this sulphonylurea is a general inhibitor of ABC transporters.     

Effect of flavonoids on MRP1-mediated transport in Panc-1 cells

The purpose of this study was to identify the effects of dietary flavonoids, which are present in fruits, vegetables, and plant-derived beverages, on the transport of daunomycin (DNM) and vinblastine (VBL) in Panc-1 cells. Panc-1 is a human pancreatic adenocarcinoma cell line, which expresses Multidrug Resistance-Associated Protein1 (MRP1). The 2-h accumulation of (3)H-DNM and (3)H-VBL was determined in the presence and absence of 22 flavonoids. Biochanin-A, genistein, quercetin, chalcone, silymarin, phloretin, morin, and kaempferol, at 100 microM concentrations, all significantly increased the accumulation of both DNM and VBL in Panc-1 cells, with morin increasing DNM and VBL accumulation by 546 +/- 50% (mean +/- SE, n = 9) and 553 +/- 37% (n = 9), respectively. Fisetin treatment significantly decreased the accumulation of both DNM and VBL. Concentration-dependent studies demonstrated significant effects on VBL accumulation at 50 microM, but not at 10 microM concentrations, except for chalcone that was effective at a 10 microM concentration. Following a 24-h incubation, there were no changes in MRP1 membrane expression or glutathione-S-transferase activity in cells. Cellular glutathione (GSH) concentrations were significantly decreased following a 2-h incubation with biochanin A, chalcone, genistein, phloretin, quercetin, and silymarin, and following a 24-h incubation with biochanin A, chalcone, genistein, and phloretin. These results therefore indicate that the flavonoids morin, chalcone, silymarin, phloretin, genistein, quercetin, biochanin A, and kaempferol can inhibit MRP1-mediated drug transport, effects that may involve binding interactions with MRP1, as well as modulation of GSH concentrations.     

Expression of lung resistance-related protein, LRP, and multidrug resistance-related protein, MRP1, in normal human lung cells in long-term cultures

Transport processes form part of the bodys defense mechanism, and they determine the intracellular levels of many endogenous and exogenous compounds. The multidrug resistance-related protein MRP1 and the lung resistance-related protein LRP are associated with drug resistance against chemotherapeutics; they protect cells against toxic compounds. There is much experimental evidence to suggest that both of these transporter proteins serve important physiological functions. The expression of LRP and MRP1 was studied in normal human bronchial epithelial cells (NHBEC) and peripheral lung cells (PLC) obtained from explant cultures from morphologically-normal human lung tissue taken from patients with lung cancer. LRP (mRNA and protein) was detected in the cells of the bronchi as well as the peripheral lung with low (a factor of 2.6) inter-individual variation in the first generation. No significant alterations were noted for LRP within three-to-four generations in the same patient. LRP expression was not substantially different between cultures from different topographic regions of the human lung. MRP1 protein and MRP1 mRNA could also be detected in all of the NHBEC and PLC cultures studied, but with substantially higher (a factor of 7.7) intra-individual variation in the first generation than for LRP. MRP expression was the same for bronchial cells and PLC when the material was obtained from both sites. The level of mRNA for MRP1 was, in general, less stable than that for LRP. In multigeneration explant cultures, the levels of LRP mRNA and protein and MRP1 protein did not fluctuate greatly, but the level of MRP1 mRNA dropped to about 25% of the reference value within four generations (after about 8–10 weeks of culture). In one case, NHBEC subpassages were followed over a period of 20 weeks. In this system MRP mRNA levels increased by more than threefold, while levels of MRP1 protein and LRP mRNA and protein were expressed at almost constant rates.     

Differential sensitivities of the prostacyclin and nitric oxide biosynthetic pathways to cytosolic calcium in bovine aortic endothelial cells.

1. Bovine aortic endothelial cells were cultured in vitro, and shown to release both prostacyclin (PGI2; Kact = 24.1 nM) and endothelium-derived relaxing factor (EDRF, NO; Kact = 0.7 nM) in a concentration-dependent manner when exposed to bradykinin. 2. The bradykinin-dependent release of PGI2 (but not EDRF) was inhibited by 1 microM isoprenaline or 5 microM forskolin, and the inhibitory effect of isoprenaline could be reversed by the beta 2-adrenoceptor antagonist, ICI 118551. In contrast, isoprenaline had no capacity to inhibit PGI2 release stimulated by exogenous arachidonic acid. 3. Exposure of cells to bradykinin increased the cytosolic concentration of Ca2+ ions ([Ca2+]i; Kact = 4.8 nM), and the effect was inhibited by both 1 microM isoprenaline and 5 microM forskolin. 4. In similar experiments, exposure of cells to ionomycin also increased [Ca2+]i and the values of [Ca2+]i were calibrated in terms of the ionomycin concentration. In subsequent experiments involving exposure of endothelial cells to selected concentrations of ionomycin, it was possible to show that the biosynthesis of NO was triggered at ionomycin concentrations about one tenth of the required for PGI2 biosynthesis and that these corresponded to a [Ca2+]i threshold of 350 nM for PGI2 release while that for EDRF release was less than 200 nM. 5. These differences in Ca2+ ion sensitivity explain the selective inhibition of bradykinin-stimulated PGI2 biosynthesis (to the exclusion of NO biosynthesis) by isoprenaline or forskolin, both of which attenuate bradykinin-dependent increases in [Ca2+]i.     

Low-dose twice-daily fractionated X-irradiation of ovarian tumor cells in vitro generates drug-resistant cells overexpressing two multidrug resistance-associated proteins, P-glycoprotein and MRP1

Failure of chemotherapy is frequently observed in patients previously treated with radiotherapy. To establish a cellular model for examining this resistance phenotype a series of mammalian tumor cell lines were exposed in vitro to fractionated X-irradiation and were then shown to express resistance to multiple antitumor drugs, including vincristine, etoposide and cisplatin. In these experiments the radiation was delivered as 10 fractions of 5 Gy (dose resulting in 1 log cell kill) given intermittently over several months. We now report that a comparable multidrug-resistance profile is expressed by human SK-OV-3 human ovarian tumor cells exposed in vitro to low dose (2 Gy) twice-daily fractions of X-rays given for 5 days on two consecutive weeks, essentially mimicking clinical practice, involving an overexpression of two MDR-associated proteins, P-glycoprotein and the multidrug resistance protein 1 (MRP1), with the latter being readily detectable by immunocytochemistry.     

Lidamycin shows highly potent cytotoxic to myeloma cells and inhibits tumor growth in mice

Aim： To investigate the effects of lidamycin （LDM） on a mouse myeloma cell line （SP2/0） and human multiple myeloma cell lines （U266 and SKO-007）, and provide the basis for the use of LDM in cancer therapy.
Methods： A 3-[4,5-dimethylthiazol-2-yl]5-[3-carboxymethoxyphenyl]-2-[4-sulfophenyl]2H-tetrazolium inner salt （MTS） assay was used to determine the degree of growth inhibition by the drugs analyzed in this study. Cell cycle distribution and analysis were measured by flow cytometry combined with propidium iodide （PI） staining. The effects on apoptosis were measured by Hoechst 33342 staining and by flow cytometry combined with fluorescein-isothiocyanate-Annexin V/propidium iodide （FITC-Annexin V/PI） staining. Protein expression was determined by Western blot analysis. In vivo antitumor activity was measured using a murine myeloma model in BALB/c mice.
Results： There was a significant reduction in cell proliferation after treatment with LDM. The overall growth inhibition correlated with increased apoptotic cell death. LDM-induced cell apoptosis was associated with the activation of c-Jun-N-terminal kinase （JNK）, and cleavage of caspase-3/7 and poly （ADP-ribose） polymerase （PARP）. LDM markedly suppressed tumor growth in a murine myeloma model.
Conclusion： LDM induces apoptosis in murine myeloma SP2/0 cells as well as in human myeloma U266 and SKO-007 cell lines. The sustained activation of JNK might play a critical role in LDM-induced apoptosis in the SP2/0 cell line. LDM demonstrates significant antitumor efficacy against myeloma SP2/0 cells in mice. Taken together, our data provide some clues for further research of the effects of LDM on human multiple myeloma.     

Differential cytotoxic effects of arsenic compounds in human acute promyelocytic leukemia cells

Arsenic trioxide, As₂O₃, has successfully been used to treat acute promyelocytic leukemia (APL). Induction of apoptosis in cancerous cells has been proposed to be the underlying mechanism for the therapeutic efficacy of arsenic. To further understand the cytotoxicity of arsenic compounds in APL cells, HL-60 cells were exposed to graded concentrations of the following arsenicals for up to 48 h: arsenic trioxide (As^III), sodium arsenate (As^V), phenylarsine oxide (PAO^III), monomethylarsonous acid (MMA^III), monomethylarsonic acid (MMA^V) and dimethylarsinic acid (DMA^V), and the viability and modes of cell death assessed. The arsenic-exposed cells were stained with annexin V-PE and 7-aminoactinomycin D (7-AAD) and analyzed by flow cytometry in order to detect apoptotic and viable cells while cell morphology was visualized using scanning and transmission electron microscopy. Acridine orange staining and microtubule-associated protein 1 light chain 3 (MAP-LC3) detection were used to recognize autophagic cell death. The results showed that the compounds reduced viable HL-60 cells by inducing apoptosis in a concentration-dependent manner. None of the compounds tested caused a significant change in binding of acridine orange or redistribution of MAP-LC3. Potencies of the six different arsenic compounds tested were ranked as PAO^III > MMA^III ≥ As^III > As^V > MMA^V > DMA^V. An increase in caspase-3 activity by PAO^III, MMA^III and DMA^V implied that these compounds induced apoptosis in HL-60 cells through a caspase-dependent mechanism, but the other arsenic compounds failed to activate caspase-3, suggesting that they induce apoptosis by an alternative pathway.     

Reversal effects of antifungal drugs on multidrug resistance in MDR1-overexpressing HeLa cells

In this study, the antiproliferative effects of vinblastine (VLB), paclitaxel (TXL), doxorubicin (DXR), daunorubicin (DNR) and 5-fluorouracil (5-FU) were assessed in the human cervical carcinoma cell line HeLa-Ohio (HeLa) and Hvr100-6 cells, established by growing the parental HeLa cells in the presence of progressively greater concentrations of VLB in the culture medium. Flow cytometric analysis indicated the induction of MDR1 (P-glycoprotein) in Hvr100-6 cells with no alterations in levels of multidrug resistance-associated protein (MRP). Resistance to VLB, TXL, DXR and DNR was found in Hvr100-6 cells with relative resistances of ca. 300, 4000, 50 and 200, respectively, whereas no resistance was found to 5-FU. The reversal effects of antifungal drugs, fluconazole, itraconazole, ketoconazole, miconazole and amphotericin B on multidrug resistance were also assessed using Hvr100-6 cells. Itraconazole was found to have potent reversal effect on the resistance to VLB and TXL, but the others had no such effect. This reversal effect of itraconazole was concentration-dependent, with dose modifying factors of 3.2, 10.1 and 435.7 at 0.1, 0.25 and 0.5 microM of itraconazole, respectively. In addition, this reversal effect of itraconazole was explained by the inhibition of accumulation of the anticancer drugs.