Folates provoke cellular efflux and drug resistance of substrates of the multidrug resistance protein 1 (MRP1).

Cellular folate concentration was earlier reported to be a critical factor in the activity and expression of the multidrug resistance protein MRP1 (ABCC1). Since MRP1 mediates resistance to a variety of therapeutic drugs, we investigated whether the cellular folate concentration influences the MRP1-mediated cellular resistance against drugs. As a model system, we used the human ovarian carcinoma cell line 2008wt, and its stably MRP1/ABCC1-transfected subline 2008/MRP1. These cell types have a moderate and high expression of MRP1, respectively. In folate-deprived 2008/MRP1 cells, the MRP1-mediated efflux of its model substrate calcein decreased to ~55 % of the initial efflux rate under folate-rich conditions. In 2008wt cells, only a small decrease in efflux was observed. Folate depletion for 5-10 days markedly increased (~500 %) cellular steady-state accumulation of calcein in 2008/MRP1 cells and moderately in 2008wt cells. A subsequent short (24 h) exposure to 2.3 µM L-leucovorin decreased calcein levels again in MRP1-overexpressing cells. Folate deprivation markedly increased growth inhibitory effects of the established MRP1 substrates daunorubicin (~twofold), doxorubicin (~fivefold), and methotrexate (~83-fold) in MRP1-overexpressing cells, proportional to MRP1 expression. In conclusion, this study demonstrates that increased cellular folate concentrations induce MRP1/ABCC1-related drug efflux and drug resistance. These results have important implications in the understanding of the role of MRP1 and its homologs in clinical drug resistance.

Cellular pharmacology of multi- and duplex drugs consisting of ethynylcytidine and 5-fluoro-2'-deoxyuridine.

Prodrugs can have the advantage over parent drugs in increased activation and cellular uptake. The multidrug ETC-L-FdUrd and the duplex drug ETC-FdUrd are composed of two different monophosphate-nucleosides, 5-fluoro-2'deoxyuridine (FdUrd) and ethynylcytidine (ETC), coupled via a glycerolipid or phosphodiester, respectively. The aim of the study was to determine cytotoxicity levels and mode of drug cleavage. Moreover, we determined whether a liposomal formulation of ETC-L-FdUrd would improve cytotoxic activity and/or cleavage. Drug effects/cleavage were studied with standard radioactivity assays, HPLC and LC-MS/MS in FM3A/0 mammary cancer cells and their FdUrd resistant variants FM3A/TK(-). ETC-FdUrd was active (IC(50) of 2.2 and 79 nM) in FM3A/0 and TK(-) cells, respectively. ETC-L-FdUrd was less active (IC(50): 7 nM in FM3A/0 vs 4500 nM in FM3A/TK(-)). Although the liposomal formulation was less active than ETC-L-FdUrd in FM3A/0 cells (IC(50):19.3 nM), resistance due to thymidine kinase (TK) deficiency was greatly reduced. The prodrugs inhibited thymidylate synthase (TS) in FM3A/0 cells (80-90%), but to a lower extent in FM3A/TK(-) (10-50%). FdUMP was hardly detected in FM3A/TK(-) cells. Inhibition of the transporters and nucleotidases/phosphatases resulted in a reduction of cytotoxicity of ETC-FdUrd, indicating that this drug was cleaved outside the cells to the monophosphates, which was verified by the presence of FdUrd and ETC in the medium. ETC-L-FdUrd and the liposomal formulation were neither affected by transporter nor nucleotidase/phosphatase inhibition, indicating circumvention of active transporters. In vivo, ETC-FdUrd and ETC-L-FdURd were orally active. ETC nucleotides accumulated in both tumor and liver tissues. These formulations seem to be effective when a lipophilic linker is used combined with a liposomal formulation.

Bortezomib induces schedule-dependent modulation of gemcitabine pharmacokinetics and pharmacodynamics in non-small cell lung cancer and blood mononuclear cells.

Bortezomib combination with gemcitabine/cisplatin in patients with advanced tumors, predominantly non-small cell lung cancer (NSCLC), showed an unexpected transient drop in the deoxycytidine plasma levels, a marker for gemcitabine activity. This study investigates the pharmacokinetic/pharmacodynamic effect of bortezomib on gemcitabine in NSCLC and peripheral blood mononuclear cells (PBMC). Gemcitabine metabolites, including difluoro-dCTP (dFdCTP), were studied in PBMCs from bortezomib/gemcitabine/cisplatin-treated patients and from volunteers and NSCLC cells (H460 and SW1573) exposed to 4 h simultaneous or sequential treatments of gemcitabine (50 µmol/L, 4 h) and bortezomib (100 nmol/L, 2 h). Gemcitabine total phosphate levels measured by liquid chromatography-tandem mass spectrometry in PBMCs from bortezomib/gemcitabine/cisplatin-treated patients were strongly reduced after 90 min (-82.2%) up to 4 h post-gemcitabine infusion compared with gemcitabine/cisplatin-treated patients. Accordingly, bortezomib/gemcitabine combinations reduced dFdCTP in PBMCs treated ex vivo. Surprisingly, differential effects were observed in NSCLC cells. dFdCTP decreased after 4 h following gemcitabine removal in H460 but continued to increase for 24 h in SW1573. However, dFdCTP significantly increased (2-fold) in both cell lines in the bortezomib → gemcitabine exposure, coinciding with a major reduction in cell growth compared with single drugs, and the highest increase of deoxycytidine kinase expression, possibly mediated via E2F-1. Bortezomib affects differently gemcitabine pharmacokinetics/pharmacodynamics in PBMCs and NSCLC cells, suggesting that PBMCs are not adequate to evaluate the anticancer activity of bortezomib/gemcitabine combinations. The bortezomib → gemcitabine/cisplatin schedule appeared a safe and active combination for the treatment of advanced NSCLC and the bortezomib → gemcitabine was the most cytotoxic combination in NSCLC cells. The increase of deoxycytidine kinase and dFdCTP might contribute to this synergistic interaction and supports its further clinical investigation.

Correlation of CDA, ERCC1, and XPD polymorphisms with response and survival in gemcitabine/cisplatin-treated advanced non-small cell lung cancer patients.

PURPOSE: Selecting patients according to key genetic characteristics may help to tailor chemotherapy and optimize the treatment in non-small cell lung cancer (NSCLC). Polymorphisms at the xeroderma pigmentosum group D (XPD), excision repair cross-complementing 1 (ERCC1), and cytidine deaminase (CDA) genes have been associated with alterations in enzymatic activity and may change sensitivity to the widely used cisplatin-gemcitabine regimen. EXPERIMENTAL DESIGN: Analyses of CDA, XPD, and ERCC1 polymorphisms were done on blood samples of 65 chemotherapy-naïve, advanced NSCLC patients treated with cisplatin-gemcitabine. Furthermore, CDA enzymatic activity was evaluated by high-performance liquid chromatography analysis. Association between XPD Asp(312)Asn and Lys(751)Gln, ERCC1 C118T, and CDA Lys(27)Gln polymorphisms and response, clinical benefit, toxicity, time to progression (TTP), and overall survival (OS) was estimated using Pearson's chi(2) tests, the Kaplan-Meier method, the log-rank test, and the Cox proportional hazards model. RESULTS: The CDA Lys(27)Lys polymorphism significantly correlated with better clinical benefit (P = 0.04) and grade > or =3 neutropenia and thrombocytopenia, as well as with longer TTP and OS (P = 0.006 and P = 0.002, respectively), whereas no significant associations were found among ERCC1 and XPD polymorphisms and both response and clinical outcome. Finally, the enzymatic activity assay showed a significant lower mean in subjects harboring the CDA Lys(27)Lys polymorphism. CONCLUSIONS: Our data suggested the role of CDA Lys(27)Lys polymorphism as a possible predictive marker of activity, toxicity, TTP, and OS in advanced NSCLC patients treated with cisplatin and gemcitabine. These results may be explained by the lower enzymatic activity associated with the Lys(27)Lys CDA and offer a potential new tool for treatment optimization.

Interaction between celecoxib and docetaxel or cisplatin in human cell lines of ovarian cancer and colon cancer is independent of COX-2 expression levels.

Celecoxib, an inhibitor of cyclooxygenase-2 (COX-2), is being investigated for enhancement of chemotherapy efficacy in cancer clinical trials. We determined whether continuous exposure to celecoxib would increase the antiproliferative effects of a 1-h treatment with docetaxel or cisplatin in four human ovarian cancer cell lines. COX-2 protein could not be detected in these cell lines, because of which three COX-2 positive human colon cancer cell lines were included. Multiple drug effect analysis demonstrated additive to borderline antagonistic effects of celecoxib combined with docetaxel. Combination indices with values of 1.4-2.5 in all cancer cell lines indicated antagonism between celecoxib and cisplatin regardless whether celecoxib preceded cisplatin for 3h, was added simultaneously or immediately after cisplatin. Apoptotic features measured in COX-2-negative H134 ovarian cancer cells and COX-2-positive WiDr colon cancer cells, such as the activation of caspase-3 and the number of cells in sub-G0 of the cell cycle, induced by docetaxel were increased in the presence of celecoxib, but were abrogated upon addition of celecoxib to cisplatin. Moreover, the G2/M accumulation in cisplatin-treated cells was less pronounced when celecoxib was present. Drugs did not affect p-Akt. Celecoxib upregulated p-ERK1/2 in H134 cells, but not in WiDr cells. Platinum-DNA adduct formation measured in WiDr cells, however, was reduced when celecoxib was combined with cisplatin. Taken together, our data demonstrate clear antagonistic effects when celecoxib is given concurrently with cisplatin, which is independent of COX-2 expression levels.

In vivo and in vitro activity and mechanism of action of the multidrug cytarabine-L-glycerylyl-fluorodeoxyuridine.

Multidrugs have the potential to bypass resistance. We investigated the in vitro activity and resistance circumvention of the multidrug cytarabine-L-fluorodeoxyuridine (AraC-L-5FdU), linked via a glycerophospholipid linkage. Cytotoxicity was determined using sensitive (A2780, FM3A/0) and resistant (AG6000, AraC resistant, deoxycytidine kinase deficient; FM3A/TK-, 5FdU resistant, thymidine kinase deficient) cell lines. Circumvention of nucleoside transporter and activating enzymes was determined using specific inhibitors, HPLC analysis and standard radioactivity assays. AraC-L-5FdU was active (IC50: 0.03 microM in both A2780 and FM3A/0), had some activity in AG6000 (IC50: 0.28 microM), but no activity in FM3A/TK(-) (IC50: 18.3 microM). AraC-nucleotides were not detected in AG6000. 5FdU-nucleotides were detected in all cell lines. AraC-L-5FdU did not inhibit TS in FM3A/TK(-) (5%). Since phosphatase/nucleotidase-inhibition reduced cytotoxicity 7-70-fold, cleavage seems to be outside the cell, presumably to nucleotides, and then to nucleosides. The multidrug was orally active in the HT-29 colon carcinoma xenografts which are resistant toward the single drugs.

Activity and substrate specificity of pyrimidine phosphorylases and their role in fluoropyrimidine sensitivity in colon cancer cell lines.

Thymidine phosphorylase (TP) and uridine phosphorylase (UP) are often upregulated in solid tumors and catalyze the phosphorolysis of natural (deoxy)nucleosides and a wide variety of fluorinated pyrimidine nucleosides. Because the relative contribution of each of the two enzymes to these reactions is still largely unknown, we investigated the substrate specificity of TP and UP in colon cancer cells for the (fluoro)pyrimidine nucleosides thymidine (TdR), uridine (Urd), 5'-deoxy-5-fluorouridine (5'DFUR), and 5FU. Specific inhibitors of TP (TPI) and UP (BAU) were used to determine the contribution of each enzyme in relation to their cytotoxic effect. The high TP expressing Colo320TP1 cells were most sensitive to 5'DFUR and 5FU, with IC50 values of 1.4 and 0.2 microM, respectively, while SW948 and SW1398 were insensitive to 5'DFUR (IC50>150 microM for 5'DFUR). TPI and BAU only moderately affected sensitivity of Colo320, SW948, and SW1398, whereas TPI significantly increased IC(50) for 5'DFUR (50-fold) and 5FU (11-fold) in Colo320TP1 and BAU that in C26A (9-fold for 5'DFUR; p<0.01). In the epithelial skin cell line HaCaT both inhibitors were able to decrease sensitivity to 5'DFUR and 5FU separately. HaCaT might be a model for 5'DFUR toxicity. In the colon cancer cells 5'DFUR degradation varied from 0.4 to 50 nmol 5FU/h/10(6)cells, that of TdR from 0.3 to 103 nmol thymine/h/10(6)cells, that of Urd from 0.8 to 79 nmol uracil/h/10(6)cells, while conversion of 5FU to FUrd was from 0.3 to 46 nmol/h/10(6)cells. SW948 and SW1398 were about equally sensitive to 5'DFUR and 5FU, but SW1398 had higher phosphorylase activity (>65-fold) compared to SW948. In SW948 and HaCaT TPI and BAU inhibited TdR and Urd phosphorolysis (>80%), respectively. Both TP and UP contributed to the phosphorolysis of 5'DFUR and 5FU. In the presence of both inhibitors, still phosphorolysis of 5FU (>40%) was detected in the tumor and HaCaT cell lines, and remarkably, that of all four substrates in SW1398 cells. 5'DFUR phosphorolysis was also measured in situ, where Colo320TP1, SW1398, and HaCaT cells produced significant amounts 5FU from 5'DFUR (>10 nmol/24h/10(6)cells). In Colo320TP1 and in HaCaT cells TPI completely prevented 5FU production, but not in SW1398 cells, where BAU decreased this by 67% (p<0.01). High uracil and dUrd levels were detected in the medium. Uracil accumulation was heavily reduced in the presence of TPI for Colo320TP1 and HaCaT cells, whereas 5FU-induced dUrd production by these cell lines increased (p<0.01). In contrast, for SW1398 cells only BAU was able to reduce uracil levels, and dUrd production remained unchanged. In conclusion, overlapping substrate specificity was found for TP and UP in the cell lines, in which both enzymes were responsible for converting TdR and Urd, and 5'DFUR. 5'DFUR and 5FU were converted to their products in both the colon cancer cells and keratinocytes.

Folate concentration dependent transport activity of the Multidrug Resistance Protein 1 (ABCC1).

The Multidrug Resistance Protein MRP1 (ABCC1) can confer resistance to a variety of therapeutic drugs. In addition, MRP1/ABCC1 mediates cellular export of natural folates, such as folic acid and l-leucovorin. In this study we determined whether cellular folate status affected the functional activity of MRP1/ABCC1 mediated efflux of an established substrate, the anthracycline daunorubicin (DNR). As a model system we used the human ovarian carcinoma cell line 2008wt, and its MRP1/ABCC1 transfected subline 2008/MRP1. Both types of these moderate- and high-MRP1/ABCC1 expressing cells displayed efflux of DNR when maintained in standard culture media (2.3microM folic acid). The initial total cellular DNR efflux rate in 2008/MRP1 cells was approximately 2-fold higher compared to 2008wt cells. This efflux consisted of MRP1/ABCC1 mediated transport, possibly non-MRP1 mediated transport, as well as passive diffusion. Benzbromarone, a specific MRP1 inhibitor, decreased the initial efflux rate in 2008/MRP1 cells (4-fold) and in 2008wt cells (2-fold). When 2008/MRP1 cells were challenged for 2 days in folate-free medium, total cellular DNR efflux was decreased to 43% of the initial efflux rate under folate-rich conditions. In 2008wt cells DNR efflux was decreased to 84% of the folate-rich conditions. Benzbromarone did not inhibit DNR efflux after the folate-free period in both cell lines. Repletion of folate by a 2-24hr exposure to 2.5microM l-leucovorin or folic acid resulted in a complete restoration of DNR efflux. In contrast, expression of MRP1/ABCC1 protein was not changed significantly during the folate-free period or the repletion-period, nor were cellular ATP or ADP pools. In conclusion, this study demonstrates that the cellular folate status can influence the transport activity of MRP1/ABCC1. These results have potentially important implications in the understanding of the (patho-)physiological roles of MRP1/ABCC1, and possibly other ABC transporter proteins in cellular folate homeostasis and drug resistance.