Implementation of point-of-care testing and a temporary influenza ward in a Dutch hospital.

BACKGROUND: The seasonal influenza epidemic poses a significant burden on hospitals, both in terms of capacity and costs. Beds that are occupied by isolated influenza patients result in hospitals temporary being closed to admissions and elective operations being cancelled. Improving hospital and emergency department (ED) patient flow during the influenza season could solve these problems. Microbiological point-of-care-testing (POCT) could reduce unnecessary patient isolation by providing a positive/negative result before admission, but has not yet broadly been implemented. METHODS: A clinical pathway for patients with acute respiratory tract infection presenting at the ED was implemented, including a PCR-based POCT for influenza, operated by nurses and receptionists. In parallel, a temporary ward equipped with 15 beds for influenza-positive patients was established. In this retrospective observational study, we describe the results of implementing this pathway by comparison with the previous epidemic. RESULTS: Clinical performance of the POCT within the clinical pathway was good with strongly decreased time from ED presentation to sample collection (194 vs 47 min) and time from sample collection to result (1094 vs 62 min). Hospital patient flow was improved by a decreased percentage of admitted influenza-positive patients (91% vs 73%) and shorter length of subsequent stay (median 5.86 vs 4.61 days) compared to the previous influenza epidemic. In addition, 430 patient-days of unnecessary isolation have been prevented within a time span of 18 weeks. Roughly estimated savings were almost 400,000 euros. CONCLUSION: We recommend that hospitals explore possibilities for improving patient flow during an influenza epidemic.

The gut microbiota in internal medicine: implications for health and disease.

The human gut microbiota may be viewed as an organ, executing numerous functions in metabolism, development of the immune system and host defence against pathogens. It may therefore be involved in the development of a range of diseases such as gastrointestinal infections, inflammatory bowel disease, allergy and diabetes mellitus. Reversely, certain therapies that are often used, such as antibiotics and chemotherapy, may negatively affect the composition and function of the gut microbiota and thereby the wellbeing of patients. As the microbiota research field is currently moving from association studies to intervention studies and even clinical trials, implementation of this new knowledge into clinical practice is coming near. Several therapeutic interventions that target the gut microbiota are being evaluated, ranging from supplementation of food components to transplantation of faecal microbiota. In this review we provide an overview of current literature on the gut microbiota in both a healthy state and a range of diseases that are relevant for internal medicine. In anticipation of gut microbiota-targeted therapies, it is important to realise the key function of the gut microbiota in physiological processes and the collateral damage that may be caused when disrupting this ecosystem within us.

Pharmacokinetics of bolus 5-fluorouracil: relationship between dose, plasma concentrations, area-under-the-curve and toxicity.

The pharmacokinetics of 5-fluorouracil (5FU) have been related to toxicity and antitumor activity, in particular for continuous infusion schedules, but to a lesser extent for frequently used bolus injections. The use of intensive sampling schedules limits the application of pharmacokinetics to optimize individual dosing or to define the ideal combination with other drugs. We therefore reanalyzed a pharmacokinetic study in order to develop a limited sampling schedule. Patients received escalating doses of 5FU at 500, 600 and 720 mg/m2 as a bolus until toxicity developed. Blood samples were analyzed until 24 h after administration. The area under the concentration time curve from 0-90 min (AUC(0-90)) was strongly correlated with dose and also with toxicity (p = 0.0009). The 5FU concentrations at 30 and 60 min were correlated to the AUC(30-240) and to that of the AUC(0-90) (r2 = 0.970). The use of limited sampling (30, 60, 90 min) in a patient given 353 mg/m2 5FU with severe toxicity at initial dosing at 500 mg/m2 revealed that the AUC(0-90) at 353 mg/m2 was higher than the normal AUC(0-90) for 500 mg/m2. This patient appeared to have an 8-fold lower activity of the 5FU degradation enzyme dihydropyrimidine dehydrogenase. Limited sampling will allow us to define potential aberrant kinetics of pharmacokinetic interaction of 5FU with other drugs being developed for treatment of colorectal cancer.

Analysing gene expressions with GRANK.

UNLABELLED: Unravelling overlapping clusters of functionally linked genes is a major challenge to current clustering programs. Algorithm GRANK permits systematic study of overlap between clusters of genes of a similar expression profile and large variation across a series of microarrays. AVAILABILITY: GRANK is freely available from http://www.bio.vu.nl/microb/research/tumor.html provided one adheres to the end-user license agreement in the associated manual. SUPPLEMENTARY INFORMATION: Above website also supplies the in- and output files used in the example run.

Evaluating limited specificity of drug pumps reduced relative resistance in human MDR phenotypes.

In the parallel paper, we developed a property to characterize drug efflux pumps, i.e. the reduced relative resistance (RRR). Using this RRR, we here investigate whether the observed diversity in human multidrug resistance (MDR) phenotypes might be due to variable levels of P-glycoprotein encoded by MDR1. We analyzed resistance phenotypes of various human cell lines in which either one, or both, classical human multidrug resistance genes, MDR1 and MDR3, are overexpressed. In addition, RRR values were calculated for MDR phenotypes presented in the literature. The results suggest that more than a single mechanism is required to account for the observed phenotypic diversity of classical multidrug resistance. This diversity is only partly due to differences in plasma membrane permeabilities between cell line families. It is discussed whether the alternative MDR phenotypes might be MDR1 phenotypes modified by other factors that do not themselves cause MDR. The method we here apply may also be useful for other nonspecific enzymes or pumps.

Vinblastine and sulfinpyrazone export by the multidrug resistance protein MRP2 is associated with glutathione export.

The multidrug resistance proteins MRP1 and MRP2 are members of the same subfamily of ATP-binding cassette transporters. Besides organic molecules conjugated to negatively charged ligands, these proteins also transport cytotoxic drugs for which no negatively charged conjugates are known to exist. In polarized MDCKII cells, MRP1 routes to the lateral plasma membrane, and MRP2 to the apical plasma membrane. In these cells MRP1 transports daunorubicin, and MRP2 vinblastine; both transporters export reduced glutathione (GSH) into the medium. We demonstrate that glutathione transport in MDCKII-MRP1 cells is inhibited by the inhibitors of organic anion transporters sulfinpyrazone, indomethacin, probenecid and benzbromarone. In MDCKII-MRP2 cells, GSH export is stimulated by low concentrations of sulfinpyrazone or indomethacin, whereas export is inhibited down to control levels at high concentrations. We find that unmodified sulfinpyrazone is a substrate for MRP2, also at concentrations where GSH export is inhibited. We also show that GSH export in MDCKII-MRP2 cells increases in the presence of vinblastine, and that the stoichiometry between drug and GSH exported is between two and three. Our data indicate that transport of sulfinpyrazone and vinblastine is associated with GSH export. However, at high sulfinpyrazone concentrations this compound is transported without GSH. Models of MRP action are discussed that could explain these results.

P-glycoprotein in primary acute myeloid leukemia and treatment outcome of idarubicin/cytosine arabinoside-based induction therapy.

The expression of the drug transport protein, P-glycoprotein (Pgp/MDR1) has been found to be of prognostic significance for the achievement of complete remission (CR) or the duration of survival after daunorubicin (DNR)-containing induction therapy in acute myeloid leukemia (AML). This would suggest that the expression of Pgp in AML is high enough to have significant impact on intracellular DNR concentrations and on clinical therapy failure in AML. Recently, DNR has been replaced in many centers by idarubicin (IDA) as the first choice anthracycline in AML treatment. We have, therefore, performed a study in a group of 98 primary AML patients, who all received IDA, but not DNR during induction therapy in order to determine if the response to IDA-containing induction therapy might be related to the biologic characteristic of Pgp expression in AML. The AML samples were studied for Pgp expression by MRK16 antibody staining and for Pgp activity measured as the modulation of rhodamine 123 uptake by 2 microM PSC 833. No correlation of Pgp with complete response rate, event-free survival or overall survival was found. In addition to Pgp, the expression of another protein that has been implicated by some studies in response failure to DNR-containing therapy, the major vault protein (Mvp/LRP), was studied. This marker did not correlate with CR or survival after IDA-containing therapy. The results of this patient study are consistent with model studies showing that the steady-state cellular accumulation of lipophilic anthracyclines such as IDA are little affected by Pgp. Therefore, putative beneficial effects of the inclusion of PSC 833 in IDA-containing therapy might rather be related to alternative mechanisms than to inhibition of Pgp-mediated IDA efflux.

The effect of glutathione on the ATPase activity of MRP1 in its natural membranes.

The transport mechanism by which the multidrug resistance protein 1 (MRP1) effluxes cytotoxic agents out of cells is still not completely understood. However, the cellular antioxidant glutathione (GSH) has been shown to have an important role in MRP1-mediated drug transport. In this study we show that GSH stimulates the ATPase activity of MRP1 in a natural plasma membrane environment. This stimulation was dose-dependent up to 5 mM. The MRP1 substrates vincristine and daunorubicin do not induce MRP1 ATPase activity. In addition, the effect of GSH on the MRP1 ATPase activity is not increased by daunorubicin or by vincristine. In contrast, a GSH conjugate of daunorubicin (WP811) does induce the ATPase activity of MRP1. In the presence of GSH the effect of WP811 was not significantly increased. Finally, (iso)flavonoid-induced MRP1 ATPase activity is not synergistically increased by the presence of GSH. In conclusion, we show that GSH has no apparent influence on the ATPase reaction induced by several MRP1 substrates and/or modulators. The subclasses of molecules had different effects on the MRP1 ATPase activity, which supports the existence of different drug binding sites.

Simvastatin improves disturbed endothelial barrier function.

BACKGROUND: Recent clinical trials have established that inhibitors of the enzyme 3-hydroxy-3-methylglutaryl coenzyme A reductase (statins) reduce the risk of acute coronary events. These effects of statins cannot be fully explained by their lipid-lowering potential. Improved endothelial function may contribute to the positive effects of statin treatment. METHODS AND RESULTS: In the present study, we report that simvastatin reduces endothelial barrier dysfunction, which is associated with the development of atherosclerosis. Treatment of human umbilical vein endothelial cells for 24 hours with 5 micromol/L simvastatin reduced the thrombin-induced endothelial barrier dysfunction in vitro by 55+/-3%, as assessed by the passage of peroxidase through human umbilical vein endothelial cell monolayers. Similar effects were found on the thrombin-induced passage of (125)I-LDL through human aortic endothelial cell monolayers. This reduction in barrier dysfunction by simvastatin was both dose and time dependent and was accompanied by a reduction in the thrombin-induced formation of stress fibers and focal adhesions and membrane association of RhoA. Simvastatin treatment had no effect on intracellular cAMP levels. In Watanabe heritable hyperlipidemic rabbits, treatment for 1 month with 15 mg/kg simvastatin reduced vascular leakage in both the thoracic and abdominal part of the aorta, as evidenced by the Evans blue dye exclusion test. The decreased permeability was not accompanied by a reduction of oil red O-stainable atherosclerotic lesions. CONCLUSIONS: These data show that simvastatin, in a relatively high concentration, improves disturbed endothelial barrier function both in vitro and in vivo. The data also support the beneficial effects of simvastatin in acute coronary events by mechanisms other than its lipid-lowering effect.

A mathematical model of drug transport in human breast cancer.

A mathematical model of drug transport in tissue has been developed on the basis of a clinical study of patients with breast cancer, treated with the drug doxorubicin and of drug transport experiments using cultured human breast cancer cells. The clinical study revealed doxorubicin gradients in tumor islets of densely packed cancer cells. The mathematical model allows simultaneous drug transport through the cellular network (transcellular pathway), through the intercellular interstitium (paracellular pathway), and across the boundary between the two networks. The effective diffusion coefficient of the interstitial network is found to be much higher than that of the cellular network, in spite of the fact that the interstitium thickness is only 20-40 nm. The model simulations can be made to fit the results of the clinical study. A long-continued simulation (40 days) of drug transport into a spherical islet with a radius of 150 microm, after a bolus injection of doxorubicin, reveals that the maximum average drug concentration at the islet centre is only reached after 224 h, while it decreases by a factor 15 from the boundary to the centre of the islet. The area under the curve in a plot of the average drug concentration versus time only decreases by 10% from the boundary to the centre of the islet.

The relative importance of passive and P-glycoprotein mediated anthracycline efflux from multidrug-resistant cells.

For the four anthracyclines idarubicin, daunorubicin, epirubicin and doxorubicin the passive and active efflux rates in intact multidrug resistant cells were compared. Although highly similar structurally, these anti-tumor agents differ in lipophilicity and membrane permeability (k). The method we used was based on the continuous measurement of the cellular efflux and determination of the ratio (RVp) of transport rates just before and just after inhibition of the active transport with verapamil (Vp). Hence, RVp - 1 should reflect the active transport rate relative to the passive transport rate. If cells were single, well-stirred compartments, RVp - 1 should equal Vmax/(k.Km), where Vmax is the maximal pumping rate and Km is the Michaelis constant. However, using the plasma membrane permeabilizing agent digitonin, we found an effective intracellular anthracycline store. Particularly, when the efflux was fast, e.g. with idarubicin or in intensively pumping cells, the intracellular transport began to control the cellular efflux. Under these conditions, k underestimated the true plasma membrane permeability (k0) and RVp - 1 underestimated Vmax/(k.Km). Based on the effects of digitonin on the efflux rates in pumping and nonpumping cells, we developed an index (RVp,corrected - 1) which should equal Vmax/(k0. Km). The term Vmax/(k0.Km) varied substantially between the drugs. It appears that differences in lipophilicity between the drugs do not affect passive efflux and pumping equally. This demonstrates that passive permeation plays a substantial and independent role in determining the drug resistance for these anthracyclines. The methods developed here enable dissection of this role from that of drug pumping and intracellular subcompartmentation.

In vitro transepithelial drug transport by on-line measurement: cellular control of paracellular and transcellular transport.

Studies on transcellular transport across epithelial cell layers are performed mostly by discontinuous sampling of the transported compound. This has several drawbacks, e.g., it gives disturbances in volume, it limits the time-resolution, and is often laborious. In this report we introduce a method to measure transepithelial transport of fluorescent compounds continuously. The time-resolution is at the (sub)minute scale, allowing the measurement of the change in transport rate before and after transport modulation. We will describe how we used the method to measure transcellular and paracellular transport. For highly membrane-impermeable compounds, the paracellular transport and the regulation of the tight junctions was studied in wild-type and MDR1 cDNA transfected epithelial canine kidney cells (MDCKII). The effect of the multidrug transporter P-glycoprotein (Pgp) on the transepithelial transport was studied. Addition of the Pgp inhibitor SDZ PSC 833 showed a modulation of the idarubicin (IDA) and daunorubicin (DNR) transport, which was larger during transport from the basolateral to the apical side than in the reverse direction. By modeling the transepithelial transport, we found that in these cells Pgp had more effect on the basolateral to apical transport than vice versa, which can be attributed to a relatively large passive permeation coefficient for the cellular basolateral plasma membrane.

Potent interaction of flavopiridol with MRP1.

The multidrug resistance protein 1 (MRP1) is an ATP-dependent transport protein for organic anions, as well as neutral or positively charged anticancer agents. In this study we show that flavopiridol, a synthetic flavonoid currently studied in phase 1 trials for its antiproliferative characteristics, interacts with MRP1 in a potent way. Flavopiridol, as well as other (iso)flavonoids stimulate the ATPase activity of MRP1 in a dose-dependent way at low micromolar concentrations. A new specific monoclonal antibody against MRP1 (MIB6) inhibits the (iso)flavonoid-induced ATPase activity of plasma membrane vesicles prepared from the MRP1 overexpressing cell line GLC4/ADR. The accumulation of daunorubicin in GLC4/ADR cells is increased by flavopiridol and by other non-glycosylated (iso)flavonoids that interact with MRP1 ATPase activity. However, flavopiridol is the only tested compound that affects the daunorubicin accumulation when present at concentrations below 1 microM. Glycosylated (iso)flavonoids do not affect MRP1-mediated transport or ATPase activity. Finally, MRP1 overexpressing and transfected cells are resistant to flavopiridol, but not to other (iso)flavonoids tested. These findings may be of relevance for the development of anticancer therapies with flavopiridol.

Doxorubicin gradients in human breast cancer.

Ten patients with locally advanced breast cancer were given doxorubicin i.v., and an incision biopsy was subsequently taken. Doxorubicin autofluorescence was examined using computerized laser scanning microscopy, and microvessels were immunostained in the same sections. Overlays of both pictures revealed doxorubicin gradients in tumor islets with high concentrations in the periphery and low concentrations in the center of the tumor islets. Gradients were most pronounced shortly after the injection, but they could still be detected 24 h later. No gradients were observed in connective tissue. This study demonstrates a serious risk of the drug not reaching all of the cancer cells in those cases in which the cancer cells are densely packed in islets. The efficacy of drug treatment will thus depend on the histology of the tumor tissue.

Do P-glycoprotein and major vault protein (MVP/LRP) expression correlate with in vitro daunorubicin resistance in acute myeloid leukemia?

Two proteins that have been correlated with the occurrence of multidrug resistance in acute myeloid leukemia (AML) are P-glycoprotein (Pgp) and the major vault protein (Mvp/LRP). With the purpose of further quantifying the potential contributions of Pgp-mediated drug efflux and Mvp/LRP to drug resistance in AML we have investigated whether the transport function of Pgp and the expression of Mvp/LRP correlated with the accumulation of daunorubicin (DNR) and the in vitro resistance to DNR cytotoxicity (LC50 by MTT assay) in AML cells. In de novo adult AML, the steady-state DNR accumulation (in pmol/10(6) cells) correlated with Pgp activity or expression, whereas the LC50 for DNR did not correlate with Pgp activity (measured as the modulation of rhodamine 123 or DNR accumulation by the Pgp inhibitor PSC833) or Pgp expression (measured by flow cytometry with the MRK-16 antibody). The contribution of MRP1 expression to a reduced DNR accumulation seems minor compared to Pgp. In addition, the modulation of the DNR LC50 by PSC833 did not correlate with Pgp protein or activity. The steady-state DNR accumulation showed no correlation with the DNR LC50. The Mvp/LRP expression (immunocytochemical staining) did neither correlate with DNR accumulation nor with the DNR LC50. A significant negative correlation was seen between the Mvp/LRP immunocytochemical staining and Pgp activity, indicating that both markers define (partially) different populations. In conclusion, it is shown that Pgp function, but not Mvp/LRP or MRP1 expression correlate with a low steady-state DNR accumulation in de novo AML. The Pgp activity does, however, not predict the DNR sensitivity in AML measured as in vitro DNR LC50 with an MTT-based assay. The reason for that seems to be that a low DNR accumulation may not be the most important factor in determining the LC50. While the clinical usefulness of these drug resistance tests remains to be proven they do not seem to provide as yet a straightforward explanation for the major cause(s) of clinical chemotherapy failure.

A method for studying plasma membrane transport with intact cells using computerized fluorometry.

A new method is presented for measuring rapid efflux of fluorescent compounds from monolayer cells. Cells grown on a glass coverslip were loaded with a fluorescent substrate. Thereafter, the coverslip was installed outside the light path in a stirred and thermostated cuvette of a fluorometer. The efflux was recorded by measuring the changes of fluorescence in the extracellular medium. The method was used to study the kinetics of active and passive plasma membrane transport of the P-glycoprotein substrates rhodamine 123 and daunorubicin. The method has advantages over other methods: (1) no radioactively labeled substrate is needed, (2) fluorescence of the transported substrate is not compromised by the cells, (3) changes in the extracellular concentration of the substrate can be monitored continuously and therefore a substantial improvement of the kinetic resolution is obtained, and (4) the measurement setup is relatively simple and a standard fluorometer can be used. From the efflux data, cellular transport parameters could be calculated, such as passive permeation coefficients and active transport rates.

P-glycoprotein-independent decrease in drug accumulation by phorbol ester treatment of tumor cells.

The effect of a change in the phosphorylation state of the drug transporter P-glycoprotein (P-gp) on its drug transport activity was studied for the substrates daunorubicin (DNR), etoposide (VP-16), and calcein acetoxymethyl ester (Cal-AM). Phorbol ester (PMA), added to stimulate phosphorylation of P-gp by protein kinase C (PKC), caused a decrease in the cellular accumulation of DNR and VP-16, both in multidrug-resistant (MDR) P-gp-overexpressing cells and in wild-type cells. Since treatment of cells with kinase inhibitor staurosporine (ST) reversed this effect of PMA and the non-PKC-stimulating phorbol ester 4alpha-phorbol, 12,13-didecanoate (4alphaPDD) did not result in a decreased DNR accumulation, we conclude that this effect is the result of kinase activity. The concentration dependence of the inhibition of P-gp by verapamil (Vp) was not influenced by PMA. Accumulation of the P-gp substrate Cal-AM was not influenced by PMA in wild-type cells. Therefore, Cal-AM was used to study the effect of PMA-induced phosphorylation of P-gp on its transport activity. Activation of PKC with PMA or inhibition of protein phosphatase 1/2A (PP1/PP2A) with okadaic acid (OA) did not affect the accumulation of Cal-AM in the MDR cells or wild-type cells. The kinase inhibitor ST increased the Cal-AM accumulation only in the MDR cells. Neither stimulating PKC with PMA nor inhibiting PP1/PP2A with OA led to a decreased inhibition of P-gp by ST, indicating that ST inhibits P-gp directly. From these experiments, we conclude that PKC and PP1/PP2A activity do not regulate the drug transport activity of P-gp. However, these studies provide evidence that PMA-induced PKC activity decreases cellular drug accumulation in a P-gp-independent manner.

Modulation by (iso)flavonoids of the ATPase activity of the multidrug resistance protein.

The multidrug resistance protein (MRP) is an ATP-dependent transport protein for organic anions, as well as neutral or positively charged anticancer agents. In this study we report that dinitrophenyl-S-glutathione increases ATPase activity in plasma membrane vesicles prepared from the MRP-overexpressing cell line GLC4/ADR. This ATPase stimulation parallels the uptake of DNP-SG in these vesicles. We also show that the (iso)flavonoids genistein, kaempferol and flavopiridol stimulate the ATPase activity of GLC4/ADR membranes, whereas genistin has no effect. The present data are consistent with the hypothesis that certain (iso)flavonoids affect MRP-mediated transport of anticancer drugs by a direct interaction with MRP.

Theoretical and practical considerations for the measurement of P-glycoprotein function in acute myeloid leukemia.

This paper summarizes experimental data and theoretical considerations, that are important for the measurement of P-glycoprotein (Pgp) function in acute myeloid leukemia (AML). The data are presented in subdivisions based on the techniques used, which will facilitate finding specific information. Based on our extensive experience with Pgp analysis, which includes radioactive assays, flow cytometry and fluorescence microscopy, we recommend a flow cytometry-based assay, that measures the effect of 2 microM PSC 833 on rhodamine 123 (R123) accumulation as the most practical and sensitive functional Pgp test. In combination with the flow cytometric measurement of Pgp using an antibody against an extracellular epitope (eg MRK16), this offers a sensitive and reproducible method for Pgp detection in AML, which is also rapid and practical. Furthermore, an R123 accumulation assay is specific for Pgp, because R123 is transported much less efficiently by the multidrug resistance protein (MRP) than by Pgp. Another probe of similar sensitivity and specificity is 3,3'-diethyloxacarbocyanine iodide. Alternatively, especially for the analysis of small numbers of cells (for example sorted subpopulations of leukemic cells), convenient and sensitive procedures are being developed by using DNA-binding Pgp substrates which remain fixed in the nuclei of the cells upon formaldehyde exposure for quantitative fluorescence laser scanning microscopy with image analysis. Less experimental data have been published to establish the optimal conditions for dual parameter flow cytometry (Pgp function, in eg Pgp+ or CD34+ cells). However, laboratories with flow cytometry experience will be able to implement this useful option to analyze subpopulations of cells.

Anthracyclines modulate multidrug resistance protein (MRP) mediated organic anion transport.

We studied the ATP-dependent uptake of dinitrophenyl-glutathione (GS-DNP) into plasma membrane vesicles derived from parental GLC4 cells and from multidrug resistant GLC4/ADR cells. The latter have a high expression of the multidrug resistance protein (MRP). Uptake of GS-DNP into membrane vesicles from GLC4/ADR cells was highly stimulated by the addition of ATP, compared to the uptake into membrane vesicles from GLC4 cells. This ATP-dependent uptake into membrane vesicles from GLC4/ADR cells was saturable with a Km of 1.2 +/- 0.2 microM and a Vmax of 560 +/- 80 pmol/mg prot./min. ATP stimulated GS-DNP uptake with a Km of 187 +/- 4 microM. This uptake was specifically inhibited by a polyclonal serum raised against a fusion protein containing a segment of MRP. The ATP-dependent uptake of GS-DNP was not only inhibited by organic anions, such as oxidized glutathione (GSSG), methotrexate (MTX) and some bile acids, but also by non-anionic natural product drugs, such as anthracyclines, vinca alkaloids and etoposide (VP-16). Uptake of GSSG and MTX into membrane vesicles from GLC4/ADR cells could be stimulated by ATP. The ATP-dependent uptake of GSSG had a Km of 43 +/- 3 microM and a Vmax of 900 +/- 200 nmol/mg protein/min. The ATP-dependent uptake of GS-DNP seemed to be non-competitively inhibited by the anthracycline daunorubicin (DNR), whereas the ATP-dependent GSSG uptake seemed to be competitively inhibited by DNR. A substrate binding site on MRP is proposed that comprises a pocket in which both DNR and GS-DNP or GSSG bind in random order to different, only partly overlapping sites. In this pocket binding of a second compound is influenced by the compound which was bound first.