Reconstitution of Drug Transport by Purified P-glycoprotein

Abstract

P-glycoprotein confers multidrug resistance upon cells in which it is highly expressed, reducing the effectiveness of numerous cytotoxic drugs, including many of those used for chemotherapy of cancer. Although P-glycoprotein is widely believed to function as an ATP-dependent drug efflux pump, the unusually broad substrate specificity of P-glycoprotein has engendered the proposal of other, less direct mechanisms. None of the hypothetical mechanisms has been definitively tested, however, in a purified system where other cellular components and processes are absent. We have used a fluorescent substrate of P-glycoprotein, Hoechst 33342, to measure transport activity in real-time of highly purified P-glycoprotein in a reconstituted liposome system in which the P-glycoprotein has a uniformly inside-out orientation. Using this system, we demonstrated MgATPdependent, chemosensitizer-inhibitable transport of Hoechst 33342. Transport was prevented by omission of Mg2+, by substitution of nonhydrolyzable adenylyl-β,γ-imidodiphosphate for ATP, by inhibition of the ATPase activity of P-glycoprotein with vanadate and N-ethylmaleimide, and by the chemosensitizers verapamil and amiodarone. Measurements of intraliposomal pH during Hoechst 33342 transport detected no large pH changes in P-glycoprotein-containing liposomes. These results are inconsistent with a mechanism in which P-glycoprotein affects drug accumulation by directly altering intracellular pH. The Hoechst 33342 transport assay results are consistent with mechanisms in which P-glycoprotein alone is sufficient to transport drugs out of the membrane bilayer. P-glycoprotein confers multidrug resistance upon cells in which it is highly expressed, reducing the effectiveness of numerous cytotoxic drugs, including many of those used for chemotherapy of cancer. Although P-glycoprotein is widely believed to function as an ATP-dependent drug efflux pump, the unusually broad substrate specificity of P-glycoprotein has engendered the proposal of other, less direct mechanisms. None of the hypothetical mechanisms has been definitively tested, however, in a purified system where other cellular components and processes are absent. We have used a fluorescent substrate of P-glycoprotein, Hoechst 33342, to measure transport activity in real-time of highly purified P-glycoprotein in a reconstituted liposome system in which the P-glycoprotein has a uniformly inside-out orientation. Using this system, we demonstrated MgATPdependent, chemosensitizer-inhibitable transport of Hoechst 33342. Transport was prevented by omission of Mg2+, by substitution of nonhydrolyzable adenylyl-β,γ-imidodiphosphate for ATP, by inhibition of the ATPase activity of P-glycoprotein with vanadate and N-ethylmaleimide, and by the chemosensitizers verapamil and amiodarone. Measurements of intraliposomal pH during Hoechst 33342 transport detected no large pH changes in P-glycoprotein-containing liposomes. These results are inconsistent with a mechanism in which P-glycoprotein affects drug accumulation by directly altering intracellular pH. The Hoechst 33342 transport assay results are consistent with mechanisms in which P-glycoprotein alone is sufficient to transport drugs out of the membrane bilayer. P-glycoprotein is a 170-kDa mammalian plasma membrane ATPase belonging to a large superfamily of integral membrane transport proteins, the ATP Binding Cassette (ABC) superfamily or traffic ATPases (16Higgins C.F. Annu. Rev. Cell Biol. 1992; 8: 67-113Crossref PubMed Scopus (3375) Google Scholar; 8Doige C.A. Ames G.F.-L. Annu. Rev. Microbiol. 1993; 47: 291-319Crossref PubMed Scopus (266) Google Scholar). P-glycoprotein is unusual, however, in having a very broad specificity for dissimilar lipophilic or amphiphilic substrates (7Chin K.-V. Pastan I. Gottesman M.M. Adv. Cancer Res. 1993; 60: 157-180Crossref PubMed Scopus (175) Google Scholar). There are three P-glycoprotein genes in rodents and two (classes I and III) in humans (6Childs S. Ling V. DeVita V.T. Hellman S. Rosenberg S.A. Important Advances in Oncology. J. B. Lippincott Company, Philadelphia, PA1994: 21-36Google Scholar). Cells which express class I or class II, but not class III, P-glycoprotein in substantial amounts exhibit the multidrug resistance phenotype, in which they are resistant to a wide variety of structurally unrelated cytotoxic compounds (15Gottesman M.M. Pastan I. Annu. Rev. Biochem. 1993; 62: 385-427Crossref PubMed Scopus (3564) Google Scholar; 10Endicott J.A. Ling V. Annu. Rev. Biochem. 1989; 58: 137-171Crossref PubMed Scopus (2018) Google Scholar). Multidrug resistance, including that mediated by P-glycoprotein, is a serious problem limiting the effectiveness of cancer chemotherapy. It is of clinical as well as biochemical interest, therefore, to develop an understanding of the mechanism by which P-glycoprotein confers multidrug resistance upon cells. Reduced drug accumulation in P-glycoprotein-overexpressing cells is the result of an increased rate of drug efflux (21Nielsen D. Skovsgaard T. Biochim. Biophys. Acta. 1992; 1139: 169-183Crossref PubMed Scopus (156) Google Scholar; 4Altenberg G.A. Vanoye C.G. Horton J.K. Reuss L. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 4654-4657Crossref PubMed Scopus (114) Google Scholar) and in some cases a decreased rate of drug influx (32Stein W.D. Cardarelli C. Pastan I. Gottesman M.M. Mol. Pharmacol. 1994; 45: 763-772PubMed Google Scholar; 18Homolya L. Holló A. Germann U.A. Pastan I. Gottesman M.M. Sarkadi B. J. Biol. Chem. 1993; 268: 21493-21496Abstract Full Text PDF PubMed Google Scholar; 11Frézard F. Garnier-Suillerot A. Eur. J. Biochem. 1991; 196: 483-491Crossref PubMed Scopus (103) Google Scholar; 21Nielsen D. Skovsgaard T. Biochim. Biophys. Acta. 1992; 1139: 169-183Crossref PubMed Scopus (156) Google Scholar). Several mechanisms have been hypothesized to explain how P-glycoprotein performs this function. In the conventional model for a transporter, P-glycoprotein binds substrates in the cytoplasm and expels them directly into the extracellular medium, using ATP hydrolysis as an energy source. In the “hydrophobic vacuum cleaner” model (Gottesman and Pastan, 1993), P-glycoprotein interacts directly with substrates in the plasma membrane and pumps them out of the cell, accounting for both increased drug efflux and decreased drug influx rates (the substrates are expelled prior to entering the cytoplasm), as well as for the hydrophobic or amphiphilic nature of the known substrates. According to the “flippase” model (17Higgins C.F. Gottesman M.M. Trends Biochem. Sci. 1992; 17: 18-21Abstract Full Text PDF PubMed Scopus (709) Google Scholar), P-glycoprotein encounters drugs in the inner leaflet of the plasma membrane and flips them to the outer leaflet from which they diffuse into the extracellular medium. In a completely different model, P-glycoprotein raises the intracellular pH ( 26Roepe P.D. Biochemistry. 1992; 31: 12555-12564Crossref PubMed Scopus (106) Google Scholar) and/or depolarizes the plasma membrane electrical potential (27Roepe P.D. Wei L.Y. Cruz J. Carlson D. Biochemistry. 1993; 32: 11042-11056Crossref PubMed Scopus (145) Google Scholar) of the cell by acting as a proton pump, a chloride channel, or by a less direct means, thereby reducing intracellular accumulation of weakly basic, cationic, lipophilic compounds or reducing pH-dependent binding of the compounds to their intracellular targets. This model avoids the necessity of P-glycoprotein having to recognize numerous, structurally diverse compounds. For the most part, P-glycoprotein function has been studied with whole cell or plasma membrane systems in which numerous proteins and other cellular components and processes may complicate interpretation of the data. With studies of P-glycoprotein function, these confounding variables include differences between multidrug resistant and sensitive cells, such as differences in intracellular pH and plasma membrane electrical potential, membrane lipid composition, and energy metabolism. In addition, it has been proposed that drug transport by P-glycoprotein involves other cellular components such as glutathione S-transferase ( 35West I.C. Trends Biochem. Sci. 1990; 15: 42-46Abstract Full Text PDF PubMed Scopus (50) Google Scholar) and cytochrome P450 (reviewed in 12Gatmaitan Z.C. Arias I.M. Adv. Pharmacol. 1993; 24: 77-97Crossref PubMed Scopus (110) Google Scholar). Distinguishing definitively between hypotheses would be greatly facilitated by the availability of a means of measuring drug transport by purified P-glycoprotein reconstituted into unilamellar lipid vesicles where the pH and membrane potential can be controlled or monitored, the lipid composition is the same for control and P-glycoprotein-containing liposomes, the only source of energy is ATP hydrolysis by P-glycoprotein, and systems for chemically modifying xenobiotic substances are not present. 30Sharom F.J. Yu X. Doige C.A. J. Biol. Chem. 1993; 268: 24197-24202Abstract Full Text PDF PubMed Google Scholar reported transport of radiolabeled colchicine by partially purified and reconstituted P-glycoprotein. We described previously (Shapiro and Ling, 1994) a method for purifying class I hamster P-glycoprotein to a high degree and reconstituting it into a lipid milieu with retention of ATPase activity. In this paper, we describe fluorescence-based methods for continuous monitoring of transport of Hoechst 33342 by purified, reconstituted P-glycoprotein. This assay demonstrates for the first time that highly purified P-glycoprotein is capable of multidrug transport and has been used to distinguish between hypothetical mechanisms of P-glycoprotein function. P-glycoprotein was purified as described (Shapiro and Ling, 1994) with a few modifications. The pre-elution wash of the immunoaffinity column with 10 ml of 25 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, and 2 mM CHAPS1( 1The abbreviations used are: CHAPS3-[(3-cholamidopropyl)dimethylammonio]-1-propane-sulfonateAMP-PNPadenylyl-β,γ-imidodiphosphateDPXp-xylene-bispyridinium bromideIAEDANS5-((((iodoacetyl)amino)ethyl)amino)naphthalene-1-sulfonic acidoctyl glucosiden-octyl-β-D-glucopyranosidepyranine8-hydroxypyrene-1,3,6-trisulfonic acidSDM4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid. )(Calbiochem) was pooled with the subsequent two fractions. These latter fractions were doubled in volume to 10 ml each. The 30-ml pool was then concentrated to 600-800 μl. These modifications doubled the yield of P-glycoprotein from the immunoaffinity column without loss of purity. 3-[(3-cholamidopropyl)dimethylammonio]-1-propane-sulfonate adenylyl-β,γ-imidodiphosphate p-xylene-bispyridinium bromide 5-((((iodoacetyl)amino)ethyl)amino)naphthalene-1-sulfonic acid n-octyl-β-D-glucopyranoside 8-hydroxypyrene-1,3,6-trisulfonic acid 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid. For most experiments, about 10 μg of purified P-glycoprotein in 50 μl of 25 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, and 2 mM CHAPS was mixed with 0.5 mg of sonicated, crude soybean L-α-phosphatidylcholine (Sigma), 17.5 mM octyl glucoside (Sigma), 50 mM Tricine-NaOH (pH 7.4), 125 mM NaCl, and 1 mM EDTA in 150 μl. The phosphatidylcholine pellets were suspended in water with a and for in a in a The was then to a column of with 50 mM Tricine-NaOH (pH 7.4), 125 mM NaCl, and 1 mM The column was with the same were from the volume between about and ml of were with the same composition but P-glycoprotein. The of control and P-glycoprotein-containing were by their between and a of and The for the two of were for a about in the P-glycoprotein-containing to the of the For in which was in the liposomes, for the in 1 mM was in the In addition, μl of 1 mM in was the prior to of the to loss of the from the by the to 1 for the of the for the The volume of the liposome was by or with a of into the liposome using the described that 10 mM or mM was used of 1 mM The and of the liposome were by using of for and for with for to the liposomes. The volume of 10 mM or mM to yield the in the volume of liposome was The volume of as described was μl of the was by 10 mM of a of that the was and that the were of control or P-glycoprotein-containing having the same of was in a with a for and a for was with an 2 using and The and were 0.5 and The was used to by of μl of mM Hoechst 33342 in water was Hoechst 33342 binding to the was as by the of the Hoechst 33342 to the This was by the by and measuring the Hoechst 33342 in the a μl of 50 mM Tricine-NaOH (pH 7.4), mM mM and mM was to The not the chloride was it ATP hydrolysis by P-glycoprotein (Shapiro and Ling, were as described in the was which during which was Transport was for For transport time were to an of 1 prior to the of the of P-glycoprotein, the for control was from the for P-glycoprotein-containing liposomes. The in many was very large to the high of the of the used to this time were a the of some were for from the are The was used to measure the intraliposomal pH Biochim. Biophys. Acta. PubMed Scopus Google Scholar; Transport the 1992; Scholar). of 10 in 50 mM 125 mM NaCl, and 1 mM EDTA of pH were to a pH for The was The and were both The of the to the the was the pH of the the were The was the pH of For of intraliposomal pH during transport of Hoechst 33342, 1 mM were was the between and was and the pH were from the was it was to 10 mM to the liposome from a concentrated in 50 mM Tricine-NaOH (pH 7.4), 125 mM NaCl, and 1 mM The ATPase activity of reconstituted P-glycoprotein was as described in and soybean L-α-phosphatidylcholine was as the source of lipid for reconstituting P-glycoprotein for transport The of this lipid are of and of unilamellar For 0.5 mg of lipid and 10 μg of P-glycoprotein in μl of the the of the octyl glucoside was in to the for retention of ATPase that ATP hydrolysis is to and intraliposomal that be for the substrate to The by the of was 17.5 mM octyl glucoside was used 0.5 mM which a of in the of octyl glucoside (Shapiro and Ling, The ATPase activity of the P-glycoprotein-containing liposomes, however, was by the octyl glucoside but no was into the liposomes. the ATPase activity was 17.5 mM octyl glucoside was This to a in the of P-glycoprotein into the In the of P-glycoprotein but in of ATPase that the was by high octyl glucoside or that the were the liposomes. For transport experiments, 17.5 mM octyl glucoside was used in the to the volume of the ATPase activity. ATP was only We to the of P-glycoprotein in the liposome using the CHAPS method of 30Sharom F.J. Yu X. Doige C.A. J. Biol. Chem. 1993; 268: 24197-24202Abstract Full Text PDF PubMed Google that a in the ATPase activity of partially purified P-glycoprotein in the of 2 mM CHAPS to the to The interpretation of this result was that of the P-glycoprotein was with the liposomes. were by 2 mM of the of 10 mM was by the of 2 mM CHAPS in the of 10 mM an of not In to the results of 30Sharom F.J. Yu X. Doige C.A. J. Biol. Chem. 1993; 268: 24197-24202Abstract Full Text PDF PubMed Google however, the ATPase activity of reconstituted P-glycoprotein was not increased by the of 2 mM the ATPase activity was decreased This result is consistent with the P-glycoprotein having a uniformly inside-out orientation. CHAPS increased the ATPase with a mM CHAPS This result was of CHAPS in of 2 mM were previously to the ATPase activity of P-glycoprotein in (Shapiro and Ling, The ATPase activity of reconstituted P-glycoprotein in the of mM CHAPS is a of μg of P-glycoprotein in the the activity of the purified P-glycoprotein prior to and Ling, CHAPS may the ATPase activity of reconstituted P-glycoprotein, as chemosensitizers verapamil (Shapiro and Ling, The ATPase activity of P-glycoprotein is by (Shapiro and Ling, 1994) and other that with C.A. Yu X. F.J. Biochim. Biophys. Acta. 1992; PubMed Scopus Google Scholar). by can be prevented by of J. Biol. Chem. 1994; Full Text PDF PubMed Google Scholar), which that in or the are the for the the of reconstituted P-glycoprotein, we it with 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid a to not of the liposomes, to the assay of the the were with mM were the of CHAPS have activity from was no ATPase activity The results are completely consistent with the that the reconstituted P-glycoprotein was uniformly in the inside-out orientation. these results that the increased ATPase activity in the of CHAPS is to of the ATPase activity of P-glycoprotein of The lipophilic nature of the substrates is a serious problem in an assay for transport by P-glycoprotein reconstituted into of the high by as to about for plasma The substrates in the lipid a very high of the substrate into the is to the of this problem we used to many we were to changes in to transport a high In addition, continuous monitoring an time to be from a Hoechst 33342 was previously to be from cells P-glycoprotein Ling V. Proc. Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar; Yu C. A. Cancer Res. 1993; Google Scholar) and is a substrate for P-glycoprotein. Hoechst 33342 may have potential in chemotherapy it I Yu C. A. Cancer Res. 1993; Google Scholar). Hoechst 33342 with the and P-glycoprotein substrate for binding to P-glycoprotein that Hoechst 33342 interacts directly with P-glycoprotein. the inhibition of was not to of the by Hoechst 33342 is demonstrated by the of a the of which is by Hoechst 33342. Hoechst 33342 25 has no the ATPase activity of P-glycoprotein-containing not known P-glycoprotein substrates such as and colchicine have or no ATPase activity (Shapiro and Ling, chemosensitizers such as and and some such as substantial of the ATPase activity. Hoechst 33342 is widely used for of increased upon binding to We that Hoechst 33342 was greatly increased by binding to We of this in an assay for transport by purified, reconstituted P-glycoprotein. the degree of of Hoechst 33342 as a function of the of or crude soybean phosphatidylcholine liposomes. The about 0.5 lipid and a of about about the with of Hoechst 33342 a upon from the lipid to the P-glycoprotein Hoechst 33342 from the liposome membrane to the of liposomes, the Hoechst 33342 The in be to the of Hoechst 33342 can be in of ATP the Hoechst 33342 of P-glycoprotein-containing decreased with time with to the control as the Hoechst 33342 was from the membrane by P-glycoprotein. The was to the of the was by loss of may have been by binding of Hoechst 33342 to ATP, thereby reducing the of Hoechst 33342. a to and of the concentrated to the in Hoechst 33342 The of this was about but was between between about and The for this is There was no consistent between control and P-glycoprotein-containing in the of the The of the to Hoechst 33342 transport by P-glycoprotein of was about in between and This was to in the of P-glycoprotein reconstituted as well as in the activity of the P-glycoprotein from different In some it was to the with μg of P-glycoprotein 10 μg to the This a rate of transport of about or about of P-glycoprotein, a of P-glycoprotein in the The was 1 of Hoechst 33342 of P-glycoprotein. This rate the between the rate of Hoechst 33342 transport out of the membrane and the rate of of Hoechst 33342 to the The rate of transport is to be the rate the rate is about of 25 Hoechst 33342 binds to the and The rate of ATP hydrolysis is about 50 the rate of into the of the liposomes, an to transport be B. and V. Ling, that the Hoechst 33342 was into the liposome 25 Hoechst 33342 can liposome and of the liposome There not to be in the rate of between control and P-glycoprotein-containing liposomes. and not the transport which of Hoechst 33342 from the lipid not of the substrate in the The purified transport system to transport is to the ATPase activity of P-glycoprotein. studies of the ATPase activity of purified, reconstituted P-glycoprotein, as well as studies of the ATPase activity of P-glycoprotein by (Shapiro and Ling, and demonstrated that the ATPase activity of P-glycoprotein such as Mg2+, and is by of and that transport of Hoechst 33342 by P-glycoprotein was by omission of and by 50 a which completely P-glycoprotein ATPase activity. that Hoechst 33342 transport was by of the reconstituted P-glycoprotein with 50 was partially prevented by of mM during the with J. Biol. Chem. 1993; 268: Full Text PDF PubMed Google Scholar that P-glycoprotein ATPase activity inhibition by transport of Hoechst 33342 by P-glycoprotein was not by the nonhydrolyzable ATP adenylyl-β,γ-imidodiphosphate These results that Hoechst 33342 transport by P-glycoprotein ATP compounds as multidrug resistance or drug to cells. The purified transport system to is to a direct P-glycoprotein. In verapamil to substantial inhibition of Hoechst 33342 transport 25 Hoechst 33342 not by verapamil and was however, the Hoechst 33342 was to 2 It may be that the chemosensitizers not with Hoechst 33342 for binding to P-glycoprotein the of high lipid in the transport of the that P-glycoprotein may transport drugs by intracellular pH we the intraliposomal pH using the in the liposomes. In the of Hoechst 33342, the pH the control was the of which a of the the pH of the control and P-glycoprotein-containing decreased as the proton as into the liposomes. was a but pH P-glycoprotein-containing with no Hoechst 33342, which was not in control liposomes. The for this is It not have been an it prior to the of There were no large differences in intraliposomal pH between control and P-glycoprotein-containing in the of Hoechst 33342 ATP addition, which demonstrates that P-glycoprotein not as an ATP-dependent proton the of intraliposomal pH in the of Hoechst 33342, we as a function of Hoechst 33342 the to and to liposome to from the that the pH not only the pH. the to the of the liposomes, we the pH time to those which not and the 10 mM of was completely the their or of In the of Hoechst 33342, of the not and the results those of The large in both of Hoechst 33342 most of the The method of the pH is to the In the in the pH of the P-glycoprotein-containing was that of the control This was not in the P-glycoprotein-containing a pH the control of pH in a of was no in the in pH between control and P-glycoprotein-containing P-glycoprotein were acting as an ATP-dependent proton pump, but were only in a of the liposomes, the be The pH is sensitive to a of about pH the is most pH Although we not of the or P-glycoprotein we that only of the P-glycoprotein, which an of pH from to then the pH would to be for the liposome such large changes to P-glycoprotein-containing of in the no the pH It is that P-glycoprotein Hoechst 33342 by a pH using purified, reconstituted P-glycoprotein, we have demonstrated for the first time that P-glycoprotein by is capable of direct drug The reconstituted system P-glycoprotein in with to the for ATP hydrolysis and inhibition by system it to definitively hypotheses about P-glycoprotein function as the “hydrophobic vacuum cleaner” (Gottesman and Pastan, method was to distinguish of the substrate between lipid and which is not with that measure the of substrates. The results of the Hoechst 33342 transport assay the that transport from the membrane the assay loss of Hoechst 33342 from the lipid for an is that it be to substrate a cells intracellular drug well the extracellular 30Sharom F.J. Yu X. Doige C.A. J. Biol. Chem. 1993; 268: 24197-24202Abstract Full Text PDF PubMed Google Scholar reported accumulation of colchicine a by partially purified P-glycoprotein. and reported accumulation of colchicine by P-glycoprotein-containing vesicles to J. Biol. Chem. 1994; Full Text PDF PubMed Google Scholar demonstrated of by P-glycoprotein-containing inside-out plasma membrane the of by Hoechst in system it to the by which P-glycoprotein Hoechst 33342 the liposomes. is that the of the Hoechst 33342 is to the lipid The of Hoechst 33342 in the is the 25 the in the lipid which has a very volume to the is very This of substrate where the of is as in In P-glycoprotein was Hoechst 33342 but this energy of the of the into the lipid The of of the liposome by P-glycoprotein to that P-glycoprotein is not directly for the increased intracellular pH of some cell and that this is not to for in intracellular drug to out of the membrane potential as the means by which P-glycoprotein affects drug In experiments, was no membrane potential the liposome was time a chloride by which P-glycoprotein a membrane potential by acting as a chloride channel, as has been proposed C.F. 1992; PubMed Scopus Google Scholar; S. C.F. 1992; Full Text PDF PubMed Scopus Google Scholar). The of of the of P-glycoprotein-containing to the of control out the of a membrane potential by proton by P-glycoprotein. that the mechanism of of P-glycoprotein is not to the plasma membrane electrical potential is the by and that P-glycoprotein-containing vesicles an membrane The rate of Hoechst 33342 transport by P-glycoprotein in was very ATP-dependent exhibit of ATP hydrolysis to substrate P-glycoprotein, however, has a high rate of ATP hydrolysis (Shapiro and Ling, and In transport P-glycoprotein about 50 ATP of Hoechst 33342 This ATP hydrolysis to both the rate of ATP hydrolysis and of the Hoechst 33342 between lipid and It be a for the to a method to of the substrate to the lipid in to the rate of Hoechst 33342 transport by P-glycoprotein. the function of P-glycoprotein. P-glycoprotein encounters substrates in the as results and those of (Gottesman and Pastan, 1993), then in which leaflet of the membrane is the substrate and how is the energy by ATP hydrolysis into substrate some chemosensitizers the rate of ATP hydrolysis by P-glycoprotein drug some substrates have no the ATPase activity of P-glycoprotein, and P-glycoprotein have ATPase activity in the of can P-glycoprotein recognize and transport numerous chemically P-glycoprotein function as a or an Ling V. Biochemistry. 1994; PubMed Scopus Google such as and in drug purified P-glycoprotein reconstituted into liposomes, we have to develop a system which is to many of these about P-glycoprotein. We are to for of and to for the to Hoechst 33342 in the transport We are to for many to and for and to for large cell