Effects of ATP depletion and phosphate analogues on P-glycoprotein conformation in live cells Katalin Goda 1 , Henrietta Nagy 1 , Eugene Mechetner 2 , Maurizio Cianfriglia 3 and Ga ´ bor Szabo ´ Jr 1 1 Department of Biophysics and Cell Biology, University of Debrecen, Hungary; 2 Chemicon International, Inc., Temecula, CA, USA; 3 Laboratorio di Immunologia, Istituto Superiore di Sanita ` , Rome, Italy P-glycoprotein (Pgp), a membrane pump often responsible for the multidrug resistance of cancer cells, undergoes con- formational changes in the presence of substrates/modula- tors, or upon ATP depletion, reflected by its enhanced reactivity with the UIC2 monoclonal antibody. When the UIC2-shift was elicited by certain modulators (e.g. cyclo- sporin A or vinblastine, but not with verapamil or Tween 80), the subsequent binding of other monoclonal anti-Pgp Ig sharing epitopes with UIC2 (e.g. MM12.10) was abolished [Nagy, H., Goda, K., Arceci, R., Cianfriglia, M., Mechetner, E. & Szabo ´ Jr, G. (2001) Eur. J. Biochem. 268, 2416–2420]. To further study the relationship between UIC2-shift and the suppression of MM12.10 binding, we compared, on live cells, how ATP depletion and treatment of cells with phosphate analogues (sodium orthovanadate, beryllium fluoride and fluoro-aluminate) that trap nucleo- tides at the catalytic site, affect the two phenomena. Similarly to modulators or ATP depleting agents, all the phosphate analogues increased daunorubicin accumulation in Pgp- expressing cells. Prelabeling of ATP depleted cells with UIC2 completely abolished the subsequent binding of MM12.10, in accordance with the enhanced binding of the first mAb. Vanadate and beryllium fluoride, but not fluoro-aluminate, reversed the effect of cyclosporin A, preventing UIC2 binding and allowing for labeling of cells with MM12.10. Thus, changes in UIC2 reactivity are accompanied by complementary changes in MM12.10 binding also in re- sponse to direct modulation of the ATP-binding site, con- firming that conformational changes intrinsic to the catalytic cycle are reflected by both UIC2-related phenomena. These data also fit a model where the UIC2 epitope is available for antibody binding throughout the catalytic cycle including the step of ATP binding, to become unavailable only in the catalytic transition state. Keywords: P-glycoprotein; multidrug resistance; UIC2; MM12.10; conformation. P-glycoprotein (Pgp) is an integral plasma membrane protein that functions as an ATP-dependent efflux pump for a broad range of lipophylic or amphiphylic compounds [1–3]. Expression of this, and related pumps on cancer cells, renders them resistant to a wide range of cytotoxic compounds, causing multidrug resistance (mdr) [1–3]. Based on the structure of its two ATP binding sites and the mechanism of ATP hydrolysis, Pgp belongs to the ATP Binding Cassette (ABC) family of transport ATPases [4,5]. The molecule is comprised of two homologous halves, each containing an ATP binding site characterized by Walker A and B sequence motifs, and six transmembrane segments. The two halves of Pgp are connected by a linker peptide ( 75 amino acids) composed of charged amino-acid residues with several phosphorylation sites [1,6]. As no difference was found in resistance between cells transfected with wild-type and phosphorylation defective forms of the human Pgp, the role of phosphorylation sites is not clearly understood [7]. Pgp interacts directly with its substrates, probably within the cell membrane, and transports them out reducing their intracellular concentration [1,8]. A number of compounds, often referred to as modulators (reversing agents, chemo- sensitizers), are capable of decreasing or eliminating mdr by preventing Pgp-mediated substrate export [1,9]. The protein exhibits a substrate-stimulated ATPase activity, suggesting that ATP hydrolysis and drug transport are intimately linked [10]. Phosphate (P i ) analogues, e.g. vanadate (V i ), beryllium fluoride (BeF x ; the exact compo- sition of the complex is unknown) and fluoro-aluminate (AlF 4 ) are potent inhibitors of Pgp ATPase activity [10–15]. The characteristics of their inhibitory effect are similar in general, as it is due to trapping of nucleotides at the catalytic site in a noncovalent but tenaciously bound form [10–15]. However, BeF x and V i trap only ADP, while AlF 4 traps both ATP and ADP, when Pgp molecules are preincubated with MgATP [11,12]. PP i protects effectively against BeF x inhibition of Pgp by competing with BeF x ,whereasPP i had no effect on inhibition by V i [12]. These data were interpreted to suggest that P i analogues may trap Pgp at different steps of the catalytic process [13]. The V i -trapped intermediate of Pgp (i.e. PgpÆMg– ADPÆV i ) is thought to be equivalent to the PgpÆMg–ADPÆP i complex, which represents the catalytic transition-state in the normal reaction pathway [14,15]. The formation of Correspondence to G. Szabo ´ Jr, Department of Biophysics and Cell Biology, University of Debrecen, PO Box 39, H-4012 Debrecen, Hungary. Fax/Tel.: + 36 52 412 623, E-mail: szabog@jaguar.dote.hu Abbreviations: Pgp, P-glycoprotein; mdr, multidrug resistance; ABC, ATP Binding Cassette; V i , vanadate; BeF x , beryllium fluoride; AlF 4 , fluoro-aluminate; CsA, cyclosporin A, FITC, fluorescein-5-isothiocyanate. (Received 6 December 2001, revised 18 March 2002, accepted 12 April 2002) Eur. J. Biochem. 269, 2672–2677 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02929.x Pgp Mg–ADP V i intermediate occurs randomly at one of the two nucleotide-binding sites of Pgp, but not simulta- neously [14–16]. It has been demonstrated in photoaffinity labeling experiments that the affinity of Pgp to different substrates is decreased after V i trapping, supporting the idea that during ATP hydrolysis Pgp undergoes conformational changes also affecting the drug binding site [17]. Conformational changes of Pgp upon its catalytic cycle can be detected by measuring the binding of certain monoclonal antibodies (mAbs; UIC2 [18], MC57 [19]). It was previously shown that reactivity of the UIC2 mAb with Pgp is increased in the presence of substrates/modulators or ATP depleting agents, or when both nucleotide-binding sites are inactivated by mutations [18]. The upshift of UIC2 binding in the presence of substrates/modulators is applied as an indicator of the expression of functional Pgp molecules in clinical tissue samples [20]. Changes in the proteolysis profile of Pgp were also detected in the presence of nucleotides alone or ATP with V i , indicative of conformational changes propagating to the extracellular domains of the pump upon its interaction with these nucleotides [21]. We have recently shown [22] that drug transport-related conformational changes of Pgp can also be detected via mAb competition involving UIC2. Using the UIC2- MM12.10 mAb pair, we have described that cyclosporin A (CsA), vinblastine, and valinomycin (and several other drugs; N. Nagy, K. Goda, F. Fenyvesi & G. Szabo ´ Jr, unpublished data) 1 interactwithPgpinsuchamannerthat preincubation of the cells with UIC2 completely suppresses the subsequent binding of MM12.10. In contrast, UIC2 mAb added at saturating concentration decreased the extent of MM12.10 labeling only mildly (up to  40%) without drug treatments, or when the cells were preincubated with verapamil or Tween-80 (and other drugs; N. Nagy, K. Goda, F. Fenyvesi & G. Szabo ´ Jr, unpublished data) 2 .The conformational state characterized by enhanced UIC2/ MM12.10 mAb competition (i.e. complete suppression of MM12.10 labeling after UIC2 binding) may be intrinsic to the normal catalytic cycle, or, alternatively, the CsA-type drugs may induce a special conformation adopted by Pgp only in the presence of these drugs. The first possibility would be confirmed if the same conformational state could be induced by nonsubstrate agents interfering with the ATPase cycle. To study this question, we compared the effects of P i analogues (V i ,BeF x ,AlF 4 ), trapping nucleotides at the catalytic site of Pgp, and of ATP depletion, on UIC2- shift and UIC2/MM12.10 competition. Accurate molecular interpretation of the changes in mAb binding are expected to help to visualize the conformational steps during the catalytic cycle. MATERIALS AND METHODS Cell lines The human mdr1-transfected NIH 3T3 (NIH 3T3 MDR1 G185) mouse fibroblast cells [23] and the drug sensitive human epidermoid carcinoma cell line KB-3-1 and its multidrug resistant relative KB-V1 [24], obtained by vinblastine selection, were used. The cell lines (obtained from M. Gottesman, NIH, Bethesda, USA) were grown as monolayer cultures at 37 °C in an incubator containing 5% CO 2 and maintained by regular passage in Dulbecco’s minimal essential medium (supplemented with 10% heat- inactivated fetal bovine serum, 2 m ML -glutamine and 25 lgÆmL )1 gentamycin). KB-V1 and NIH 3T3 MDR1 cells were cultured in the presence of 180 n M vinblastine or 670 n M doxorubicin, respectively. Cells were trypsinized 2–3 days prior to the experiments and maintained without vinblastine or doxorubicin until use. The cells were occa- sionally checked for mycoplasma 3 by the mycoplasma tissue culture rapid detection system with a 3 H-labeled DNA probe from General-Probe Inc. and were found to be negative. Chemicals The cells were treated with the modulators or other agents at the following concentrations: 100 l M verapamil, 100– 500 l M Na-orthovanadate (from a 13.5-m M translucent stock solution freshly prepared in distilled water [25]), fluoro-aluminate (AlF 4 – ; prepared from 1 m M AlCl 3 and 5m M NaF), beryllium fluoride (BeF x ; prepared from 2 m M BeSO 4 plus 10 m M NaF) [11], 10 l M cyclosporin A, 50– 100 l M vinblastine or 10 l M valinomycin. Intracellular ATP was depleted by 5 l M oligomycin or 5 m M sodium azide applied together with 5 m M 2-deoxy-D-glucose. The optimal concentration of P i analogues (V i ,BeF x and AlF 4 ) and ATP depleting agents were pretitrated in daunorubicin accumulation experiments. The concentrations of the chemicals applied increased steady-state daunorubicin accumulation to the approximate level reached in Pgp – parental cells and did not significantly increase the ratio of dead cells, as assessed by propidium iodide exclusion. All of the above agents were from Sigma–Aldrich (Budapest). Cell culture media and supplements were also from Sigma. Fluorescein isothiocyanate (FITC) was purchased from Molecular Probes (Eugene, OR, USA). All the other chemicals used in the experiments were of analytical grade, from Sigma. The MRK16 (Cancer Chemotherapy Center, Tokyo, Japan), UIC2, 4E3, MC57, MM12.10 and MM8.15 anti-Pgp mAb preparations were > 97% pure by SDS/ PAGE. The FITC conjugates of MM12.10 and UIC2 were prepared as described previously [26,27]. Flow cytometric assays Nearly confluent monolayers of cells were harvested by 2–3 min trypsin treatment [0.05% trypsin and 0.02% EDTA in NaCl/P i (pH 7.4)] and washed twice with NaCl/P i before antibody labeling. The mAbs MRK16, MC57, MM8.15, MM12.10 (8 lgÆmL )1 )and10lgÆmL )1 of UIC2 and 4E3 were applied. The antibody competi- tion test was performed as follows: 10 6 cells in 1 mL NaCl/P i supplemented with 8 m M glucose were preincu- bated in the absence or presence of different drugs/ modulators at 37 °C for 20 min, then the first mAb, UIC2 was added, without washing the cells. After further 30 min incubation at 37 °C, the FITC-conjugated secon- dary mAb MM12.10 was added (again without washing the cells) and incubation followed at 37 °C for another 30 min. The extent of competition between mAbs UIC2 and FITC-MM12.10 were expressed as R competition ,the difference of mean fluorescence intensities of cell-bound FITC-MM12.10, in the absence and in the presence of Ó FEBS 2002 Catalytic cycle of P-glycoprotein (Eur. J. Biochem. 269) 2673 UIC2, divided by the fluorescence intensity obtained in the absence of UIC2. In the case of the ÔUIC2 shift assayÕ [18], the cells were pretreated with drugs, labeled first with UIC2 then with the secondary antibody [FITC-conjugated goat anti-(mouse IgG2a) Ig, F/P ¼ 4.3, from Sigma) on ice in 100 lLof NaCl/P i for 45 min. The samples were washed twice after staining and resuspended in 500 lLofNaCl/P i for flow cytometric analysis. Daunorubicin and calcein accumulation was measured as described previously [28,29]. Briefly, the cells were preincu- bated with modulators for 10 min or with P i analogues (V i , BeF x or AlF 4 ) for 20 min and then with daunorubicin or calcein-AM for further 40 min, at 1 and 0.5 l M final concentration, respectively. All the incubations were carried out at 37 °C. The cells were washed and the samples stored on ice until their measurement. The mean cellular fluores- cence in each sample was determined using a modified Becton Dickinson FACStar Plus flow cytometer (Mountain View, CA, USA) equipped with an argon ion laser (Spectra- Physics Inc. Mountain View, CA, USA). Dead cells stained with propidium iodide were excluded from the analysis. Fluorescence signals were collected in logarithmic mode and the cytofluorimetric data were analyzed by the FLOWIN software (written by M. Emri & L. Balkay, University of Debrecen, Positron Emission Tomography Center, Hungary). RESULTS As ATP depletion is known to increase UIC2 reactivity (UIC2-shift assay, [18]), we investigated if treatment of cells with oligomycin, or sodium azide together with 2-deoxy- D -glucose, elicit a similar effect on UIC2/ MM12.10 competition. R competition (i.e. the relative decrease of FITC-MM12.10 labeling after UIC2 pretreatment, see Materials and methods) was used to characterize the effect of the treatments. In ATP-depleted cells, UIC2 completely abolished MM12.10 labeling similarly to CsA treated cells, as shown by the R competition  1valuesinFig. 1.Thus,ATP depletion affects UIC2 binding and UIC2/MM12.10 com- petition in a parallel manner. P i analogues are often used for blocking Pgp ATPase activity in plasma membrane vesicles [10], and it was recently shown that V i also affects UIC2 binding in living cells [30]. We compared the effects of CsA and P i analogues (V i ,AlF 4 ,BeF x ) on Pgp function, measuring the cellular accumulation of different Pgp substrates. As demonstrated in Fig. 2, treatments with P i analogues restored steady-state daunorubicin levels in NIH 3T3 MDR1 cells, similarly to the effect of CsA applied at a concentration that completely inhibits the pump. V i treatment also increased calcein accumulation of Pgp + cells, albeit to a lesser degree ( twofold, as opposed to the 30-fold increase in intracel- lular calcein levels observed after CsA treatment; data not shown). As the addition of P i analogues did not influence significantly either the daunorubicin or calcein uptake of the parental cells (NIH 3T3 cells; Fig. 2B), their effect must be Pgp-specific. The consequences of V i ,BeF x and AlF 4 treatments on UIC2 binding were also examined. It was shown previously by Druley et al. [30] that V i prevents UIC2 binding even on vinblastine-treated cells. We have found that BeF x pretreat- ment also decreased the reactivity of UIC2 with cell surface Pgp molecules (data not shown), in contrast with CsA treatment or ATP depletion that increased UIC2 binding in accordance with what was previously shown by Mechetner et al.[18].V i and BeF x also suppressed the effect of substrates/modulators (verapamil, CsA, vinblastine) on UIC2 binding when the cells were incubated in the simultaneous presence of V i or BeF x , and any of the above agents. The order of treatments with substrates/modulators vs. V i or BeF x seemed to be indifferent as it is shown in the case of CsA and P i analogues in Fig. 3. In contrast to V i and BeF x ,AlF 4 did not affect the binding of UIC2 despite its inhibitory effect on Pgp function (compare Figs 2 and 3). Fig. 1. Effect of ATP depletion on UIC2-MM12.10 mAb competition in NIH 3T3 MDR1 cells. Cellular ATP production was inhibited by 30 min pretreatment of cells with 5 l M oligomycin or 5 m M sodium azide applied together with 5 m M 2-deoxy- D -glucose. Labeling with mAbs was carried out and R competition was calculated as described in Materials and methods. Means of three independent experiments are shown (± SEM). Fig. 2. Effect of phosphate analogues and cyclosporin A (CsA) on the accumulation of daunorubicin into Pgp + (NIH 3T3 MDRl) and Pgp – (NIH 3T3) cells. The cells were preincubated in the presence of phos- phate analogues (V i ,BeF x ,AlF 4 ), or 10 lmCsAfor20 min,then1 lm daunorubicin was added for 40 min. Solid line: daunorubicin only; bold line: 10 l M CsA; empty triangles: 500 l M V i ; black triangles: BeF x ; empty squares: AlF 4 treatment. Preparation of phosphate ana- logues is described in Materials and methods. 2674 K. Goda et al. (Eur. J. Biochem. 269) Ó FEBS 2002 The reactivity of several other anti-Pgp mAbs (MRK16, 4E3, MM.12.10, MC57, MM.8.17) was not affected by either V i or substrate/modulator treatment (data not shown). The effects of P i analogue treatments on UIC2/MM12.10 competition are shown in Fig. 4. The degree of the UIC2/ MM12.10 competition (i.e. the suppression of MM12.10 binding after UIC2) correlates with the level of UIC2 binding upon the same treatment (compare Figs 3 and 4). DISCUSSION The UIC2-shift [18] and the mAb competition phenomenon involving UIC2 and MM12.10 mAbs [22] were affected by pharmacological modulation of the ATPase cycle in a parallel manner. Near the catalytic transition state stabilized by V i or BeF x , Pgp apparently undergoes a global conformational change involving the extracellular loops, manifest in the drastic decrease of its affinity to UIC2. ATP depletion increased UIC2 reactivity as expected [18], preventing subsequent MM12.10 binding. AlF 4 that prob- ably freezes the catalytic cycle at an earlier phase, trapping unhydrolysed ATP [11], inhibited transport function with- out preventing UIC2 binding. In our experiments, different P i analogues affected Pgp function and conformation in live cells. As the nucleotide- binding domains of Pgp do not seem to be accessible from the extracellular surface of the cell [31], our results suggest that P i analogues penetrate the cell membrane, despite of their negative charge. V i can increase phosphorylation of Pgp at its linker peptide region [32] similar to phorbol esters or phosphatase inhibitors (e.g. ocadaic acid). However, the latter agents do not affect or decrease drug accumulation [33,34], while in our experiments, V i significantly inhibited drug transport. In addition, other P i analogues, e.g. BeF x and AlF 4 also inhibited Pgp mediated drug transport. Thus, the V i effect seems to be independent of the phosporylation state of the pump and is best interpreted assuming that V i trapping of Pgp also occurs in live cells. Coupling of Pgp-mediated drug transport and ATP hydrolysis are generally interpreted in terms of ligand- induced ATPase activation and concomitant transitions between substrate-binding and substrate-releasing conform- ational states. The consecutive steps of the catalytic cycle appear to be discriminated by the changing affinity of Pgp to the UIC2 mAb [18,30]. Reactivity of Pgp with UIC2 is Fig. 4. Effect of phosphate analogues on UIC2/MM12.10 mAb competition in NIH 3T3 MDR1 cells. Cells were preincubated for 20 min in the presence or absence of phosphate analogues (V i ,BeF x , AlF 4 ), followed by the addition or omission of 10 l M CsA for an additional 30 min and finally labeled with UIC2 and MM12.10 mAbs (upper panel). Incubations with CsA and phosphate analogues were also carried out in the reverse order (lower panel). Preparation of phosphate analogues, labeling with mAbs and calculation of R competition values are as described in Materials and methods. Means of three independent experiments are shown (± SEM). Fig. 3. Flow-cytometric fluorescence intensity distribution histograms demonstrating UIC2-shift assays performed on NIH 3T3 MDR1 cells. Cells were preincubated for 20 min in the presence or absence of phosphate analogues, then 10 l M CsA was added and the cells were further incubated for 30 min and finally labeled with UIC2 mAb. In parallel experiments, incubations with CsA and phosphate analogues were carried out in the reverse order. Phosphate analogues and CsA were continuously present till the end of labeling with UIC2. UIC2 binding was visualized by indirect immunofluorescence. Solid line: no CsA and phosphate analogue treatment; bold line: 10 l M CsA; empty triangles: 10 l M CsA + 500 l M V i; black triangles: 10 l M CsA + BeF x ; empty squares: 10 l M CsA + AlF 4 treatment. Ó FEBS 2002 Catalytic cycle of P-glycoprotein (Eur. J. Biochem. 269) 2675 increased in the presence of substrates/modulators or ATP depleting agents, and in double Walker A and B mutants incapable of ATP binding and hydrolysis [18,35]. These findings may be interpreted suggesting that changes in the UIC2 reactivity of Pgp are brought about by nucleotide binding and release [18,30], or assuming that the hydrolysis of ATP is the key element in driving Pgp from UIC2- reactive conformation to a nonreactive state. We favor the latter interpretation considering the following data. Previ- ous studies demonstrated that linker region mutants capable of binding ATP but unable to hydrolyze it, are recognized by UIC2 similarly to the functional molecules that have been preincubated by Pgp substrates [36], suggesting that changes in the availability of the UIC2 epitope cannot be explained by the step of ATP binding itself [35]. This conclusion was further confirmed by the following obser- vations. Mutation restricted to the N-terminal Walker B motif abolished ATP binding at both ATP sites, while mutation of the C-terminal motif did not affect ATP binding at the N-terminal site [35]. Interestingly, both of the above mutants were recognized by UIC2, implying that ATP binding per se does not affect the availability of the UIC2 epitope. In our experiments, V i (in agreement with [30]) and BeF x treatment induced a significant decrease of UIC2 binding, despite the presence of substrates or modulators, while AlF 4 ,knowntotrapbothATPand ADP produced upon hydrolysis [11], inhibited drug pumping similarly to V i and BeF x but did not affect UIC2 binding (Figs 3 and 4). These data also support a model where the UIC2 epitope is available for antibody binding throughout the catalytic cycle including the step of ATP binding, to become unavailable only in the catalytic transition state. ATP hydrolysis could cover the energy expenses of a global conformational change effecting a decreased affinity for substrates, as shown in vanadate- trapping experiments [17]. As both substrate/modulator category identified on the basis of UIC2/MM1210 mAb binding include agents that stimulate, and others that inhibit, ATPase activity [37–39], the two classes of substrates/modulators may not be distinguished based on the overall catalytic rate achieved by Pgp in their presence. The investigated conformational effect of CsA (or vinblastine) manifest at the externally located UIC2-binding epitopes is apparently reproduced by ATP depletion and can be thwarted 4 by P i analogues, confirming that the state elicited by CsA-type substrates/ modulators is part of the normal catalytic cycle. This latter conformational state must be present for periods long enough to be recognized by UIC2 in a way that leads to mAb competition. In the presence of verapamil-like agents, this state may be by-passed or its duration may be diminished. As UIC2 binding and UIC2/MM12.10 competition could be influenced by nonsubstrate agents in a parallel manner, the two phenomena are, in the case of the CsA-type modulators, negative replicas of each other. Thus, the epitopes opening up upon UIC2-shift are eventually titrated back by MM12.10, without modulator treatment. A practical implication of this finding is that the often variable functional modulation of Pgp by modulators (e.g. CsA) in the UIC2-shift assay can be substituted by the consecutive application of UIC2 and MM12.10 for the labeling of Pgp + cells, in the absence of modulators. ACKNOWLEDGEMENTS This work was financially supported by OTKA funding T 032563 and the research grant of the Hungarian Academy of Sciences AKP 98-83 3,3. This publication was also sponsored by the research grant of the Ministry of Public Health ETT T01/103 and the OMFB grant 02692/ 2000. M. C. is in part supported by grant 502 from Istituto Superiore di Sanita ` , Rome, Italy. The technical assistance of Eniko ¨ Pa ´ sztor is gratefully acknowledged. REFERENCES 1. Gottesman, M.M. & Pastan, I. (1993) Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu. Rev. Biochem. 62, 385–427. 2. Nooter, K. & Sonneveld, P. (1994) Clinical relevance of P-glyco- proteinexpressioninhaematologicalmalignancies.Leukemia Res. 18, 233–243. 3. Bosch, I. & Croop, J. (1996) P-glycoprotein multidrug resistance and cancer. Biochim. Biophys. Acta 1288, 37–54. 4. Gottesman, M.M., Hrycyna, C.A., Schoenlein, P.V., Germann, U.A. & Pastan, I. (1995) Genetic analysis of the multidrug transporter. Annu. Rev. Genet. 29, 607–649. 5. Walker, J.E., Saraste, M., Runswick, M.J. & Gay, N.J. (1982) Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP requiring enzymes and a common nucleotide binding fold. EMBO J. 1, 945–951. 6. Gottesman, M.M. & Pastan, I. (1988) The multidrug transporter, a double-edged sword. J. Biol. Chem. 263, 12163–12166. 7. Germann, U.A., Chambers, T.C., Ambudkar, S.V., Licht, T., Cardarelli, C.O., Pastan, I. & Gottesman, M.M. (1996) Char- acterization of phosphorylation-defective mutants of human P-glycoprotein expressed in mammalian cells. J. Biol. Chem. 271, 1708–1716. 8. Homolya, L., Hollo ´ , Zs, Germann, U.A., Pastan, I., Gottesman, M.M. & Sarkadi, B. (1993) Fluorescent cellular indicators are extruded by the multidrug resistance protein. J. Biol. Chem. 268, 21493–21496. 9. Seelig, A. (1998) A general pattern for substrate recognition by P-glycoprotein. Eur. J. Biochem. 251, 252–261. 10. Sarkadi, B., Price, E.M., Boucher, R.C., Germann, U.A. & Scarborough, G.A. (1992) Expression of the human multidrug resistance cDNA in insect cells generates a high activity drug sti- mulated membrane ATPase. J. Biol. Chem. 267, 4854–4858. 11. Sankaran, B., Bhagat, S. & Senior, A.E. (1997) Inhibition of P-glycoprotein ATPase activity by procedures involving trapping of nucleotide in catalytic sites. Arch. Biochem. Biophys. 341, 160–169. 12. Sankaran, B., Bhagat, S. & Senior, A.E. (1997) Inhibition of P-glycoprotein ATPase activity by beryllium fluoride. Biochem- istry 36, 6487–6853. 13. Szaka ´ cs, G., O ¨ zvegy, Cs, Bakos, E ´ ., Sarkadi, B. & Va ´ radi, A. (2000) Transition-state formation in ATPase-negative mutants of human MDR1 protein. Biochem. Biophys. Res. Commun. 276, 1314–1319. 14. Urbatsch, I.L., Sankaran, B., Weber, J. & Senior, A.E. (1995) P-glycoprptein is stably inhibited by vanadate-induced trapping of nucleotide at a single catalytic site. J. Biol. Chem. 270, 19383– 19390. 15. Senior, A.E. (1998) Catalytic mechanism of P-glycoprotein. Acta Physiol. Scand. 163, 213–218. 16. Hrycyna, C.A., Ramachandra, M., Ambudkar, S.V., Ko, Y.H., Pedersen, P.L., Pastan, I. & Gottesman, M.M. (1998) Mechanism of action of human P-glycoprotein ATPase activity. J. Biol. Chem. 273, 16631–16634. 17. Ramachandra, M., Ambudkar, S.V., Chen, D., Hrycyna, C.A., Dey, S., Gottesman, M.M. & Pastan, I. (1998) Human P-glyco- 2676 K. Goda et al. (Eur. J. Biochem. 269) Ó FEBS 2002 protein exhibits reduced affinity for substrates during a catalytic transition state. Biochemistry 37, 5010–5019. 18. Mechetner, E.B., Schott, B., Morse, S.B., Stein, W., Druley, T., Dvis, K.A., Tsuruo, T. & Roninson, I. (1997) P-glycoprotein function involves conformational transitions detectable by dif- frential immmunoreactivity. Proc.NatlAcad.Sci.USA94, 12908– 12913. 19. Jachez, B., Cianfriglia, M. & Loor, F. (1994) Modulation of human P-glycoprotein epitope expression by temperature and/or resistance modulating agents. Anti-Cancer Drugs 5, 655– 665. 20. Mechetner,E.,Kyshtoobayeva,A.,Zonis,S.,Kim,H.,Stroup, R., Garcia, R., Parker, R.J. & Fruehauf, J.P. (1998) Levels of multidrug resistance (MDR1) P-glycoprotein expression by human breast cancer correlate with in vitro resistance to taxol and doxorubicin. Clin. Cancer Res. 4, 389–398. 21. Julien, M. & Gros, P. (2000) Nucleotide-induced conformational changes in P-glycoprotein and in nucleotide binding site mutants monitored by trypsin sensitivity. Biochemistry 39, 4559–4568. 22. Nagy,H.,Goda,K.,Arceci,R.,Cianfriglia,M.,Mechetner,E.& Szabo ´ Jr, G. (2001) P-glycoprotein conformational changes detected by antibody competition. Eur. J. Biochem. 268, 2416– 2420. 23. Brugemann, E.P., Currier, S.J., Gottesman, M.M. & Pastan, I. (1992) Characterization of the azidopine and vinblastine binding site of P-glycoprotein. J. Biol. Chem. 267, 21020–21026. 24. Shen, D.W., Cardarelli, C., Hwang, J., Cornwell, M.M., Richert, N., Ishii, S., Pastan, I. & Gottesman, M.M. (1986) Multiple drug- resistant human KB carcinoma cells independently selected for high-level resistance to colchicine, adriamycin, or vinblastine show changes in expression of specific proteins. J. Biol. Chem. 261, 7762–7770. 25. Gordon, J.A. (1991) Use of vanadate as protein-phosphotyrosine phosphatase inhibitor. Methods Enzymol. 201, 477–482. 26. Spack, E.G. Jr, Packard, B., Wier, M.L. & Edidin, M. (1986) Hydrophobic adsorption chromatography to reduce nonspecific staining by rhodamine–labeled antibodies. Anal. Biochem. 158, 233–237. 27. Szo ¨ llo ¨ si, J., Damjanovich, S., Goldman, C.K., Fulwyler, M.J., Aszalo ´ s,A.,Goldstein,G.,Rao,P.,Talle,M.&Waldmann, T.A. (1987) Flow cytometric resonance energy transfer measure- ments support the association of a 95-kDa peptide termed T27 with the 55-kDa Tac peptide. Proc.NatlAcad.Sci.USA84, 7246–7250. 28. Hollo ´ ,Zs 5 , Homolya, L., Davis, C.W. & Sarkadi, B. (1994) Calcein accumulation as a fluorometric functional assay of the multidrug transporter. Biochim. Biophys. Acta 1191, 384–388. 29. Goda, K., Balkay, L., Maria ´ n, T., Tro ´ n, L., Aszalo ´ s, A. & Szabo ´ Jr, G. (1996) Intracellular pH does not affect drug extrusion by P-glycoprotein. J. Photochem. Photobiol. 34, 177–182. 30. Druley, T.E., Stein, W.D. & Roninson, I.B. (2001) Analysis of MDR1 P-glycoprotein conformational changes in permeabilized cells using differential immunoreactivity. Biochemistry 40, 4312– 4322. 31. Blott, E.J., Higgins, C.F. & Linton, K.J. (1999) Cystein-scanning mutagenesis provides no evidence for the extracellular accessibility of the nucleotide-binding domains of the multidrug resistance transporter P-glycoprotein. EMBO J. 18, 6800–6808. 32. Lelong, I.H., Cardarelli, C.O., Gottesman, M.M. & Pastan, I. (1994) GTP-stimulated phosphorylation of P-glycoprotein in transporting vesicles from KB-V1 multidrug resistant cells. Biochemistry 33, 8921–8929. 33. Wielinga, P.R., Heijn, M., Broxterman, H.J. & Lankelma, J. (1997) P-glycoprotein-independent decrease in drug accumulation by phorbol ester treatment of tumor cells. Biochem. Pharm. 54, 791–799. 34. Chambers, T.C., Pohl, J., Raynor, R.L. & Kuo, J.F. (1993) Identification of specific sites in human P-glycoprotein phos- phorylated by protein kinase C. J. Biol. Chem. 268, 4592–4595. 35. Hrycyna, C.A., Ramachandra, M., Germann, U.A., Cheng, P.W., Pastan, I. & Gottesman, M.M. (1999) Both ATP sites of P-glycoprotein are essential but not symmetric. Biochemistry 38, 13887–13899. 36. Hrycyna, C.A., Airan, L.E., Germann, U.A., Ambudkar, S.V., Pastan, I. & Gottesman, M.M. (1998) Structural flexibility of the linker region of human P-glycoprotein permits ATP hydrolysis and drug transport. Biochemistry 37, 13660–13673. 37. Borginia, M.J., Eytan, G.D. & Assaraf, Y.G. (1996) Competition of hydrophobic peptides, cytotoxic drugs, and chemosensitizers on a common P-glycoprotein pharmacophore as revealed by its ATPase activity. J. Biol. Chem. 271, 3163–3171. 38. Regev, R., Assaraf, Y.G. & Eytan, G.D. (1999) Membrane flui- dization by ether, other anesthetics, and certain agents abolishes P-glycoprotein ATPase activity and modulates efflux from multi- drug-resistant cells. Eur. J. Biochem. 259, 18–24. 39. Rebbeor, J.F. & Senior, A.E. (1998) Effects of cardiovascular drugs on ATPase activity of P-glycoprotein in plasma membranes and in purified reconstituted form. Biochim. Biophys. Acta 1369, 85–93. Ó FEBS 2002 Catalytic cycle of P-glycoprotein (Eur. J. Biochem. 269) 2677 . Effects of ATP depletion and phosphate analogues on P-glycoprotein conformation in live cells Katalin Goda 1 , Henrietta Nagy 1 , Eugene Mechetner 2 ,. antibody binding throughout the catalytic cycle including the step of ATP binding, to become unavailable only in the catalytic transition state. ATP hydrolysis