Enhanced separation of antidepressant drugs using a polymerized nonionic surfactant as a transient capillary coating

PURPOSE: To investigate the role of the P-glycoprotein (P-gp) drug efflux pump in the intracellular disposition of colchicine and vinblastine. METHODS: Uptake and efflux kinetics were studied in vitro in human lymphocytes and in HL-60 cells with or without the P-gp modulator, verapamil. RESULTS: In human lymphocytes, colchicine was slowly taken up (uptake half-life was 18.9+/-1.1 hr.) and verapamil increased colchicine uptake by 37%, whereas it did not modify colchicine efflux from cells. In HL-60 cells, colchicine uptake was non-linear and slower than that of vinblastine, the colchicine uptake half-life (11.1+/-0.5 hr.) being 25-fold longer than that of vinblastine at 25 nM. Verapamil did not significantly modify colchicine uptake half-life, but increased its intracellular accumulation by 23% and that of vinblastine by 81%. Immuno-flow cytometry showed that P-gp expression in HL-60 cells increased significantly from 24 hr. following colchicine or vinblastine exposure. The significant increase in colchicine uptake induced by verapamil at 24 hr. was correlated with this enhanced P-gp expression. The drug efflux half-life was 11.5-fold higher for colchicine (23+/-0.9 hr) than vinblastine, indicating a much slower elimination of colchicine from cells that could be related to its longer dissociation half-life from the tubulin receptor. Verapamil treatment did not modulate either colchicine or vinblastine efflux kinetics, suggesting that the intracellular drugs are not available to the transmembrane P-gp binding sites. CONCLUSIONS: P-gp may not be the main reason for the slowness of colchicine uptake. It may be more efficient at controlling entry of colchicine and vinblastine through the plasma membrane than at mediating their efflux from HL-60 cells.     

Role of P-glycoprotein in the blood-brain transport of colchicine and vinblastine

Classically, drug penetration through the blood-brain barrier depends on the lipid solubility of the substance, except for some highly lipophilic drugs, like colchicine and vinblastine, both substrates of P-glycoprotein, a drug efflux pump present at the luminal surface of the brain capillary endothelial cells. Colchicine and vinblastine uptake into the brain was studied in the rat using the in situ brain perfusion technique and two inhibitors of P-glycoprotein, verapamil and SDZ PSC-833. When rats were pretreated with PSC-833 (10 mg/kg, intravenous bolus), colchicine and vinblastine uptake was enhanced 8.42- and 9.08-fold, respectively, in all the gray areas of the rat brain studied. The mean colchicine distribution volume was increased from 0.67 +/- 0.41 to 5.64 +/- 0.70 microliters/g and vinblastine distribution volume from 2.74 +/- 1.15 to 24.88 +/- 4.03 microliters/g. When rats were pretreated with verapamil (1 mg/kg, intravenous bolus), colchicine distribution volume was increased 3.70-fold. The increase in colchicine and vinblastine did not differ between the eight brain gray areas. PSC-833 and verapamil pretreatment had no influence on the distribution volume of either drug in the choroid plexus. Nevertheless, distribution volumes remained small, considering the highly lipophilic nature of the substances. We suggest that P-glycoprotein is either only partially inhibited (difficulty of fully saturating P-glycoprotein, especially under in vivo conditions) or not the only barrier to these two drugs.     

Simultaneous microdialysis in brain and blood of the mouse: extracellular and intracellular brain colchicine disposition

A simultaneous brain and blood microdialysis system was developed to study the passage of colchicine through the blood-brain barrier in the mouse. Colchicine was administered as a bolus in the jugular vein (1.5 mg kg-1) and its hippocampal extracellular fluid (ECF) and blood kinetics were determined over a 4 h period using two microdialysis probes, one in the dorsal hippocampus, the other in the inferior vena cava. Colchicine rapidly diffused into the hippocampus (maximum concentration in the first dialysate sample) and brain and blood concentrations declined in parallel, suggesting rapid equilibration between these two compartments. However, only 6. 7% of total blood colchicine, 14% of unbound colchicine was present in the hippocampus suggesting that the P-glycoprotein efflux pump limits colchicine uptake by the brain. We also found, using conventional tissue homogenate analysis in parallel, that the concentration of colchicine in the hippocampal ECF was 10 times less than that in the intracellular space and that the hippocampus colchicine concentration was 2.8 times higher than that of the rest of the brain. This study shows that the simultaneous brain and blood microdialysis can be used to measure the passage of colchicine through the blood-brain barrier and to estimate the brain extra- and intracellular distribution of colchicine.     

Functional expression of the P-glycoprotein mdr in primary cultures of bovine cerebral capillary endothelial cells

The P-glycoprotein mdr is expressed not only in tumoral cells, but also in nontransformed cells, including the specialized endothelial cells of brain capillaries which build up the blood-brain barrier. Since all previously identified blood-brain barrier markers are rapidly lost when cerebral capillary endothelial cells are maintained in primary culture, we have investigated whether P-glycoprotein (P-gp) would follow the same rule, in order to address the influence of the cerebral environment on the specific P-gp expression in the brain endothelium. As compared to freshly isolated purified cerebral capillaries, P-glycoprotein was detected by immunochemistry at a high level in 5-7 day primary cultures. In our culture conditions, P-glycoprotein was immunodetected at a lower molecular weight than that found in freshly isolated capillaries. Enzymatic deglycosylation led to the same 130 kDa protein for both fresh and cultured samples, suggesting that P-gp post-translational modifications were altered in primary cultures. However, studies on the uptake and efflux of the P-gp substrate [3H]vinblastine, and on the effect of various mdr reversing agents on the uptake and efflux, clearly indicated that the efflux pump function of the P-glycoprotein was maintained in primary cultures of bovine cerebral capillary endothelial cells. P-Glycoprotein may thus represent the first blood-brain barrier marker which is maintained in cerebral endothelial cells cultured in the absence of factors originating from the brain parenchyma.     

Distribution of mRNA encoding the inwardly rectifying K+ channel, BIR1 in rat tissues

Chinese hamster P-glycoprotein ("multidrug-resistance protein") was purified and reconstituted in proteoliposomes by the procedure of I. L. Urbatsch, M. K. al-Shawi, and A. E. Senior (1994, Biochemistry 33, 7069-7076). The presence of lipid during the octylglucoside solubilization and Reactive Red 120 chromatography steps was found to be mandatory for retention of ATPase activity. Sheep brain or bovine liver lipid extracts could be substituted for the Escherichia coli lipids used previously. Stimulation of ATPase activity of purified, reconstituted P-glycoprotein by vinblastine, colchicine, and daunomycin was seen with sheep brain and bovine liver lipids, but not with E. coli lipids. Basal (i.e., not drug-stimulated) ATPase activity was different in the three lipids. Azidopine labeling of the drug binding sites in purified, reconstituted P-glycoprotein was carried out; vinblastine, colchicine, and daunomycin competed for labeling in all three lipids. It is therefore evident that the lipid environment can significantly influence the characteristics of purified, reconstituted P-glycoprotein ATPase activity and the apparent coupling between drug-binding and catalytic sites.     

Microsatellite instability is correlated with lymph node-positive breast cancer

1. P-glycoprotein, a 170-180 kDa membrane glycoprotein that mediates multidrug resistance, hydrolyses ATP to efflux a broad spectrum of hydrophobic agents. In this study, we analysed the effects of three MDR reversing agents, verapamil, cyclosporin A and [3'-keto-Bmt1]-[Val2]-cyclosporin (PSC 833), on the adenosine triphosphatase (ATPase) activity of human P-glycoprotein. 2. P-glycoprotein was immunoprecipitated with a monoclonal antibody (MRK-16) and the P-glycoprotein-MRK-16-Protein A-Sepharose complexes obtained were subjected to a coupled enzyme ATPase assay. 3. While verapamil activated the ATPase, the cyclosporin derivatives inhibited both the substrate-stimulated and the basal P-glycoprotein ATPase. No significant difference was observed between PSC 833 and cyclosporin A on the inhibition of basal P-glycoprotein ATPase activity. PSC 833 was more potent than cyclosporin A for the substrate-stimulated activity. 4. Kinetic analysis indicated a competitive inhibition of verapamil-stimulated ATPase by PSC 833. 5. The binding of 8-azido-[alpha-32P]-ATP to P-glycoprotein was not altered by the cyclosporin derivatives, verapamil, vinblastine and doxorubicin, suggesting that the modulation by these agents of P-glycoprotein ATPase cannot be attributed to an effect on ATP binding to P-glycoprotein. 6. The interaction of the cyclosporin derivatives with ATPase of P-glycoprotein might present an alternative and/or additional mechanism of action for the modulation of P-glycoprotein function.     

Multidrug resistance-reversing agents increase vinblastine distribution in normal tissues expressing the P-glycoprotein but do not enhance drug penetration in brain and testis

The aim of this study was to assess whether P-glycoprotein (Pgp) inhibitors altered the blood-brain barrier and enhanced vinblastine (VBL) distribution in brain, testis and other Pgp-expressing tissues. Trifluoperazine, cyclosporin A, amiodarone, quinidine, the nifedipine analog Bay K8644 and verapamil were selected among Pgp inhibitors and were administered intraperitoneally 1 hr before an intravenous dose of 10 mg/kg VBL. Trifluoperazine and cyclosporin A were also administered intraperitoneally for 7 days before VBL. VBL and its metabolite O4-deacetylvinblastine were measured in tissues by high-performance liquid chromatography assay. None of the reversing agents (RA) appreciably raised VBL concentrations in brain and testis, whereas all except quinidine significantly enhanced VBL distribution in liver and kidney; the most effective were trifluoroperazine and cyclosporin A. In mice treated with RA and VBL combined, O4-deacetylvinblastine levels in liver and kidney reached either the same or higher levels than in mice treated with VBL alone, indicating that the increase in VBL levels is not due to inhibition of its metabolism. The main conclusions are that (1) inhibitors of Pgp, even at high doses, do not increase the permeability of the blood-brain barrier in mice, suggesting caution in the clinical use of RA combined with antitumor agents for brain tumors; and (2) several RA achieve high enough concentrations to enhance the distribution of VBL in other normal tissues expressing Pgp, thus potentially increasing VBL toxicity.     

Interaction of combinations of drugs, chemosensitizers, and peptides with the P-glycoprotein multidrug transporter

P-Glycoprotein functions as an ATP-driven efflux pump for hydrophobic natural products and peptides, and gives rise to resistance to multiple chemotherapeutic drugs. The inhibition of colchicine transport via P-glycoprotein by various compounds was determined in a plasma membrane vesicle model system. A chemotherapeutic drug (vinblastine) and several chemosensitizers (verapamil, reserpine, cyclosporin A) and hydrophobic peptides (N-acetyl-leucyl-leucyl-methioninal, leupeptin, pepstatin A, valinomycin) were examined, both as individual species and as combinations of compounds. The median effect analysis was used to determine the concentration of each combination required to produce a median effect, Dm, as well as the sigmoidicity of the concentration-effect plot, m. The combination of cyclosporin A and verapamil was the only one established to be mutually nonexclusive, whereas several mutually exclusive pairs of compounds were identified. The combination index, CI, was calculated for several combinations of drugs, chemosensitizers, and peptides, and used to ascertain whether effects were synergistic, antagonistic, or additive. Some combinations (vinblastine/verapamil; verapamil/valinomycin) showed antagonism over the entire concentration range. Other combinations (valinomycin/N-acetyl-leucyl-leucyl-methioninal; cyclosporin A/verapamil) displayed both synergism and antagonism over different regions of the CI plot. Many combinations of compounds displayed additive interactions over most of the CI plot. The median effect analysis may be helpful in identifying potentially useful additive or synergistic combinations of compounds for reversal of Pgp-mediated drug resistance.     

Functional characterization of the rat mdr1b encoded P-glycoprotein: not all inducing agents are substrates

The multidrug resistance (mdr) genes encode P-glycoproteins, integral membrane proteins which function as drug efflux transporters. Exposure of animals in vivo and cells in vitro to a variety of xenobiotics leads to increased mdr1 gene expression and higher levels of P-glycoprotein. This response may protect cells from the cytotoxic effects of these compounds. In this investigation we functionally expressed the rat mdr1b gene in NIH 3T3 cells and assessed the ability of the encoded P-glycoprotein to protect these cells from the cytotoxicity of xenobiotics known to induce mdr1b expression. In long-term colony survival assays, stably expressed mdr1b conferred resistance to cytotoxic drugs such as colchicine, vinblastine and doxorubicin, but not to 5-fluorouracil nor to the carcinogens aflatoxin B1 and N-hydroxy-acetylaminofluorene. The mdr reversal agent verapamil restored cytotoxicity of colchicine, doxorubicin, actinomycin D, vinblastine and taxol, but had no effect on the sensitivity of these cells to 5-fluorouracil, aflatoxin B1 or N-hydroxy-acetylaminofluorene. In a competitive transport assay, verapamil and, to a lesser extent, colchicine blocked the increased efflux of the fluorescent dye rhodamine 123 from mdr1b-transfected cells, whereas aflatoxin B1 did not compete for this export. These data demonstrate that expression of the rat mdr1b encoded P-glycoprotein can protect cells from a diverse group of compounds previously identified to be mdr substrates, however, other effective inducers of mdr expression, such as aflatoxin B1 and N-hydroxy-acetylaminofluorene, remain potent cytotoxins despite high levels of P-glycoprotein. The fact that compounds which are not themselves substrates can induce P-glycoprotein expression may have implications for pharmacokinetic interactions and chemotherapy.     

Functional expression of P-glycoprotein in an immortalised cell line of rat brain endothelial cells, RBE4

The presence of P-glycoprotein in the cell plasma membrane limits the penetration of many cytotoxic substances into cells that express the gene product. There is considerable evidence also to indicate that P-glycoprotein is expressed as part of the normal blood-brain barrier in the luminal membranes of the cerebral capillary endothelial cells, where it presumably performs a protective function for the brain. This report describes the functional expression of P-glycoprotein in an immortalised cell line, RBE4, derived from rat cerebral capillary endothelial cells. The expression of P-glycoprotein is demonstrated by western immunoblotting and by immunogold and fluorescent staining with monoclonal antibodies. The cellular accumulation of [3H]colchicine and [3H]-vinblastine is investigated and shown to be enhanced by the presence of azidothymidine, chlorpromazine, verapamil, cyclosporin A, and PSC 833 ([3'-keto-Bmt1]-[Val2]-cyclosporin) at 50 or 100 microM concentration. It is concluded that the RBE4 cell line is a valuable tool for investigating the mechanisms of P-glycoprotein activity both in the blood-brain barrier and in multidrug resistance in general.     

Handling of doxorubicin by the LLC-PK1 kidney epithelial cell line

The characteristics of doxorubicin handling have been studied in the cultured kidney epithelial cell line LLC-PK1, which has structure and function similar to those of renal tubular cells and expresses P-glycoprotein. The uptake of doxorubicin by LLC-PK1 cells was time dependent, reaching a steady state at about 4 hr, and reduced at low temperature; the initial uptake was saturable. The efflux of doxorubicin from LLC-PK1 cells was also temperature dependent but, even at 37 degrees C, a significant percentage of the drug remained associated with the cells after 180 min, which suggests a strong cellular binding, and the fluorescence microscopy revealed that the drug was concentrated in intracellular organelles. Substances that are substrates for P-glycoprotein, such as verapamil, vinblastine, vincristine and quinidine, significantly increased doxorubicin concentrations in LLC-PK1 cells. Similar results were obtained with the metabolic inhibitors sodium metavanadate and 2,4-dinitrophenol. On the other hand, the uptake was not affected by the classic organic cation transport drugs cimetidine, decynium 22 or decynium 24, nor by the organic anion drug probenecid. These results indicate that, in LLC-PK1 cells, doxorubicin enters by passive diffusion, is trapped in intracellular organelles and then is extruded from cells by a mechanism that probably involves P-glycoprotein. On the contrary, substances that interfere with the renal organic cation or anion secretory system have no effect on doxorubicin net accumulation in these cells.     

Characterization of a novel bisacridone and comparison with PSC 833 as a potent and poorly reversible modulator of P-glycoprotein

Novel compounds, composed of two acridone moieties connected by a propyl or butyl spacer, were synthesized and tested as potential modulators of P-glycoprotein (P-gp)-mediated multidrug resistance. The propyl derivative 1,3-bis(9-oxoacridin-10-yl)-propane (PBA) was extremely potent and, at a concentration of 1 microM, increased steady state accumulation of vinblastine (VLB) approximately 9-fold in the multidrug-resistant cell line KB8-5. In contrast to the readily reversible effects of VRP and cyclosporin A on VLB uptake and similar to the effects of the cyclosporin analog PSC 833, this modulation by PBA was not fully reversed 6-8 hr after transfer of cells to PBA-free medium. Continuous exposure to 3 microM PBA was nontoxic and could completely reverse VLB resistance in KB8-5 cells. Consistent with its effects on VLB transport, the drug resistance-modulating effect of PSC 833 was significantly more persistent than that of VRP. However, the effect of PBA was, like that of VRP, rapidly reversed once the modulator was removed from the extracellular environment. PBA was able to compete with radiolabeled azidopine for binding to P-gp and to stimulate P-gp ATPase activity. However, both the steady state accumulation of PBA and the rate of efflux of PBA were similar in drug-sensitive KB3-1 and drug-resistant KB8-5 cells, suggesting that this compound is not efficiently transported by P-gp. These results indicate that PBA represents a new class of potent and poorly reversible synthetic modulators of P-gp-mediated VLB transport.     

Stabilization of microtubule dynamics by estramustine by binding to a novel site in tubulin: a possible mechanistic basis for its antitumor action

The cellular targets for estramustine, an antitumor drug used in the treatment of hormone-refractory prostate cancer, are believed to be the spindle microtubules responsible for chromosome separation at mitosis. Estramustine only weakly inhibits polymerization of purified tubulin into microtubules by binding to tubulin (Kd, approximately 30 microM) at a site distinct from the colchicine or the vinblastine binding sites. However, by video microscopy, we find that estramustine strongly stabilizes growing and shortening dynamics at plus ends of bovine brain microtubules devoid of microtubule-associated proteins at concentrations substantially below those required to inhibit polymerization of the microtubules. Estramustine strongly reduced the rate and extent both of shortening and growing, increased the percentage of time the microtubules spent in an attenuated state, neither growing nor shortening detectably, and reduced the overall dynamicity of the microtubules. Significantly, the combined suppressive effects of vinblastine and estramustine on the rate and extent of shortening and dynamicity were additive. Thus, like the antimitotic mechanisms of action of the antitumor drugs vinblastine and taxol, the antimitotic mechanism of action of estramustine may be due to kinetic stabilization of spindle microtubule dynamics. The results may explain the mechanistic basis for the benefit derived from combined use of estramustine with vinblastine or taxol, two other drugs that target microtubules, in the treatment of hormone-refractory prostate cancer.     

The effect of drugs on the mortality of mice inoculated with Friend leukaemia virus or toxoplasma gondii

P-glycoprotein (P-gp), a product of the multidrug-resistant (mdr) genes, is expressed in the endothelial cells of the blood-brain barrier (BBB). Effects of glial factors and retinoic acid (RA) on P-gp activity and level were investigated in the immortalized rat brain endothelial cell line RBE4, which expressed immunodetectable P-gp associated with a decrease in accumulation of the P-gp substrates, vinblastine and colchicine. When RBE4 cells were cultured either in the presence of C6-conditioned medium or on C6- or astrocyte-extracellular matrix, intracellular vinblastine and colchicine concentrations were decreased. When the cells were treated with RA, increases in P-gp activities were correlated with increases in P-gp levels. Effects of simultaneous treatments with glial factors and RA were studied in RBE4 cells cultured on astrocyte-extracellular matrix and were shown to be additive on P-gp activity and level. RBE4 cells may serve as a useful in vitro model for basic research on P-gp regulation at the level of the BBB.     

Properties of tubulin in unfertilized sea urchin eggs. Quantitation and characterization by the colchicine-binding reaction

The colchicine-binding assay was used to quantitate the tubulin concentration in unfertilized Strongylocentrotus purpuratus eggs and to characterize pharmacological properties of this tubulin. Specificity of colchicine binding to tubulin was demonstrated by apparent first-order decay colchicine-binding activity with stabilization by vinblastine sulfate, time and temperature dependence of the reaction, competitive inhibition by podophyllotoxin, and lack of effect of lumicolchicine. The results demonstrate that the minimum tubulin concentration in the unfertilized egg is 2.71 mg per milliliter or 5.0% of the total soluble cell protein. Binding constants and decay rates were determined at six different temperatures between 8 degrees C and 37 degrees C, and the thermodynamic parameters of the reaction were calculated. delta H0=6.6 kcal/mol, delta S0=46.5 eu, and, at 13 degrees C, delta G=-6.7 kcal/mol. The association constants obtained were similar to those of isolated sea urchin egg vinblastine paracrystals (Bryan, J. 1972. Biochemistry. 11:2611-2616) but approximately 10 times lower than that obtained for purified chick embryo brain tubulin at 37 degrees C (Wilson, L.J.R. Bamburg, S.B. Mizel, L. Grisham, and K. Creswell. 1974. Fed Proc. 33:158-166). Therefore, the lower binding constants for colchicine in tubulin-vinblastine paracrystals are not due to the paracrystalline organization of the tubulin, but are properties of the sea urchin egg tubulin itself.     

pH-dependent growth retardation of Escherichia coli caused by overproduction of Na+/H+ antiporter

PSC 833, a nonimmunosuppressive cyclosporin, is a potent inhibitor of the efflux of antitumor drugs mediated by P-glycoprotein and thus has been introduced in clinical trials as an agent to overcome multidrug resistance. The purpose of this study was to evaluate the dose-dependent pharmacokinetics of PSC 833 and its effects on the biliary excretion of endogenous substrates in rats. The major elimination route for PSC 833 is metabolism, followed by excretion into bile. The biliary clearance of PSC 833 was reduced in a dose-dependent manner, whereas no urinary excretion of PSC 833 was detectable. The tissue/blood concentration ratios for PSC 833 in the liver, kidney, intestine, and spleen were reduced in a dose-dependent manner, suggesting the presence of a saturable uptake process and/or saturable binding in these tissues. The dose-dependent increase in the tissue/blood concentration ratio in the brain suggests the presence of efflux transporters on the blood-brain barrier. PSC 833 reduced the bile flow rate by decreasing the biliary excretion of bile acids and reduced and oxidized glutathione, in a dose-dependent manner. The mechanism for the dose-dependent disposition of PSC 833 and its effects on the biliary excretion of endogenous substrates could be related to interactions with transporters.     

Mechanism of action cryptophycin. Interaction with the Vinca alkaloid domain of tubulin

Cryptophycin is a potent antitumor agent that depletes microtubules in intact cells, including cells with the multidrug resistance phenotype. To determine the mechanism of action of cryptophycin, its effects on tubulin function in vitro were analyzed. Cryptophycin reduced the in vitro polymerization of bovine brain microtubules by 50% at a drug:tubulin ratio of 0.1. Cryptophycin did not alter the critical concentration of tubulin required for polymerization, but instead caused substoichiometric reductions in the amount of tubulin that was competent for assembly. Consistent with its persistent effects on intact cells, cryptophycin-treated microtubule protein remained polymerization-defective even after cryptophycin was reduced to sub-inhibitory concentrations. The effects of cryptophycin were not due to denaturation of tubulin and were associated with the accumulation of rings of microtubule protein. The site of cryptophycin interaction with tubulin was examined using functional and competitive binding assays. Cryptophycin blocked the formation of vinblastine-tubulin paracrystals in intact cells and suppressed vinblastine-induced tubulin aggregation in vitro. Cryptophycin inhibited the binding of [3H]vinblastine and the hydrolysis of [gamma32P]GTP by isolated tubulin, but did not block the binding of colchicine. These results indicate that cryptophycin disrupts the Vinca alkaloid site of tubulin; however, the molecular details of this interaction are distinct from those of other antimitotic drugs.     

Flip-flop of doxorubicin across erythrocyte and lipid membranes

Doxorubicin, an anticancer drug, is extruded from multidrug resistant (MDR) cells and from the brain by P-glycoprotein located in the plasma membrane and the blood-brain barrier, respectively. MDR-type drugs are hydrophobic and, as such, enter cells by diffusion through the membrane without the requirement for a specific transporter. The apparent contradiction between the presumably free influx of MDR-type drugs into MDR cells and the efficient removal of the drugs by P-glycoprotein, an enzyme with a limited ATPase activity, prompted us to examine the mechanism of passive transport within the membrane. The kinetics of doxorubicin transport demonstrated the presence of two similar sized drug pools located in the two leaflets of the membrane. The transbilayer movement of doxorubicin occurred by a flip-flop mechanism of the drug between the two membrane leaflets. At 37 degrees, the flip-flop exhibited a half-life of 0.7 min, in both erythrocyte membranes and cholesterol-containing lipid membranes. The flip-flop was inhibited by cholesterol and accelerated by high temperatures and the fluidizer benzyl alcohol. The rate of doxorubicin flux across membranes is determined by both the massive binding to the membranes and the slow flip-flop across the membrane. The long residence-time of the drug in the inner leaflet of the plasma membrane allows P-glycoprotein a better opportunity to remove it from the cell.     

The ability of verapamil to restore intracellular accumulation of anthracyclines in multidrug resistant cells depends on the kinetics of their uptake

The basic distinguishing feature of all cells expressing functional P-glycoprotein-multidrug resistance is a decrease of steady state drug levels as compared to those in drug-sensitive controls. A variety of small molecules, such as verapamil and cyclosporin A, bind to P-glycoprotein and inhibit its ability to pump out antitumor drugs. The kinetics of P-glycoprotein-mediated efflux of various anthracycline derivatives was measured in multidrug-resistant (MDR) K562 cells in the presence of verapamil. Used for the purpose were daunorubicin, idarubicin and 8-S-fluoro-idarubicin which have the same pKa of deprotonation equal to 8.4, but different lipophilicity, 4'-epi-2'-bromo-daunorubicin which has a lipophilicity which is comparable to that of daunorubicin but a pKa equal to 6.3, pirarubicin with pKa equal to 7.7 and lipophilicity different from that of these derivatives were used. Our data show (1) that verapamil is unable to completely block the P-glycoprotein-mediated efflux of anthracyclines and that 10% of its functionality remains even with high verapamil concentrations, (2) that the ability of verapamil to restore intracellular accumulation of anthracyclines in MDR cells depends on the kinetics of their uptake. With fast kinetics uptake, as is the case for idarubicin, 8-S-fluoro-idarubicin, 4'-epi-2'-bromo-daunorubicin and pirarubicin (which have either a low pKa and/or high lipophilicity), verapamil can restore in multidrug resistant cells an intracellular drug level which is comparable to that observed in sensitive cells. This is not possible when the kinetics of uptake is low as is the case for daunorubicin. Cyclosporin A is a more potent modulator and is able to fully restore daunorubicin accumulation in multidrug resistant cells.