HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 39, 36075-36078, September 28, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/39/36075    most recent C100345200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Taylor, J. C. Articles by Begley, G. S. Search for Related Content PubMed PubMed Citation Articles by Taylor, J. C. Articles by Begley, G. S.

ACCELERATED PUBLICATION
The Multidrug Resistance P-glycoprotein

OLIGOMERIC STATE AND INTRAMOLECULAR INTERACTIONS^*

Jenny C. Taylor, Andrea R. Horvath^§, Christopher F. Higgins^¶, and Gail S. Begley

From the ICRF Laboratories, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, United Kingdom

Received for publication, June 21, 2001, and in revised form, July 23, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The human multidrug resistance P-glycoprotein (P-gp), a member of the ATP-binding cassette (ABC) superfamily of transporters, is frequently responsible for the failure of chemotherapy by virtue of its ability to export hydrophobic cytotoxic drugs from cells. Elucidating the inter- and intramolecular interactions of this protein is critical to understanding its cellular function and mechanism of action. Toward this end, we have used both biochemical and genetic techniques to probe potential oligomerization interactions of P-gp. Differentially epitope-tagged P-gp molecules did not co-immunoprecipitate when co-expressed in HEK293 cells or when co-translated in vitro, demonstrating that P-gp is monomeric in both the presence and absence of detergents. The two cytoplasmic domains of P-gp did not interact with each other in vivo when co-expressed as gene fusions in yeast. In contrast, the homologous domains of the transporter associated with antigen processing (TAP), which reside on separate polypeptides and must form a heterodimeric transporter (TAP1/TAP2), did interact in this system, suggesting a role for these domains in TAP dimerization. Implications for understanding the subunit organization of ABC transporters are discussed.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The product of the human MDR1 gene, P-glycoprotein (P-gp),¹ is an integral membrane protein that can confer multidrug resistance on cells and tumors (reviewed in Ref. 1). P-gp is an ATP-dependent efflux pump that transports an extraordinary range of hydrophobic substrates out of the cell and has also been implicated in the regulation of a heterologous cell swelling-activated chloride channel (reviewed in Ref. 2). P-gp is a member of the ATP-binding cassette (ABC) superfamily of transporters and has the characteristic architecture of this protein family: two large cytoplasmic nucleotide-binding domains (NBDs) and two hydrophobic transmembrane domains. The first cytoplasmic domain includes a linker region that can be phosphorylated by protein kinase C. Phosphorylation of the linker has little or no effect on drug transport (3, 4) but plays a role in modulation of heterologous ion channel activity by P-gp (2, 5).

Biochemical and biophysical techniques have previously been employed to investigate the oligomeric structure of P-gp, with ambiguous results. Freeze-fracture electron micrographs of membranes containing overexpressed P-gp showed aggregates in the plasma membrane corresponding to an oligomeric state (6). Radiation inactivation (7) yielded an apparent molecular mass of 250 kDa for P-gp, corresponding to approximately a dimer, and chemical cross-linking indicated a small proportion of a 340-kDa form (8). Sedimentation velocity centrifugation through sucrose gradients also indicated an oligomeric state (9), although this depended upon the detergent used and could not distinguish whether P-gp might be interacting with heterologous proteins rather than sedimenting as an oligomer. In brain capillaries and renal brush border membranes, radiation inactivation and mobility of Triton X-solubilized protein on native gels also suggested a possible dimer (10). In contrast, a molecular complementation approach suggested that the functional unit of P-gp is a monomer (11). Single-particle electron microscopy of purified reconstituted P-gp also showed a size for the functional protein particle most consistent with a monomeric form (12), and labeling with lectin-gold particles indicated a single site of glycosylation on each P-gp particle, consistent with a monomer (12).

As the strategies used to date are subject to a number of caveats or may reflect the fortuitous interaction of the target with other proteins and aggregation due to overexpression, we have used a combination of in vitro and in vivo techniques to assess the oligomeric state of P-glycoprotein, as well as potential intramolecular interactions between the cytoplasmic nucleotide-binding domains of the protein.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Manipulations and Bacterial Culture-- Restriction digests, polymerase chain reaction (PCR), and other manipulations of DNA were performed according to standard protocols. DNA was introduced into Escherichia coli Sure^TM cells (Stratagene) by electroporation. Transformed cells were grown in Luria-Bertani broth supplemented with 100 µg/ml ampicillin. Plasmid DNA was prepared by alkaline lysis (Qiagen). Double-stranded plasmid DNA was sequenced by the automated dye-terminator method (ABI PRISM^TM 377 DNA Sequencer, PerkinElmer Life Sciences).

Mammalian Cell Culture-- HEK293 cells (Imperial Cancer Research Fund Cell Culture Facility, Clare Hall, UK) were maintained in monolayers in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 10% fetal calf serum.

Construction of Epitope-tagged P-glycoproteins-- Myc (EEQKLISEEDL) and FLAG (DYKDDDDK) epitope tags were added to the C terminus of P-gp by oligonucleotide-directed mutagenesis of the human MDR1 gene. A PstI-SalI fragment of MDR1 encoding the C terminus of the protein was cloned into the pAlter vector and mutagenesis performed according to the Altered Sites procedure (Promega). The mutagenic oligonucleotide was designed to replace the stop codon of MDR1 with the epitope tag followed by an in-frame stop codon and a novel AatII restriction site to facilitate identification of the tagged MDR1 genes. Sequencing of the mutant plasmids confirmed the presence of the tags and demonstrated that no other mutations had been introduced. The mutated 480-base pair PstI-SalI fragment was then used to replace the corresponding fragment of the wild type MDR1 sequence in plasmid pMDR7 (13) to regenerate the full-length P-gp coding sequence with the appropriate epitope tags (pMDR7.M and pMDR7.F, respectively). The pMDR7 plasmid has a T7 promoter for in vitro expression. For in vivo studies, the tagged MDR1 fragments were cloned into pcDNA3 (Invitrogen), which contains a cytomegalovirus promoter, generating pcMDR.M (myc tagged) and pcMDR.F (FLAG-tagged). Plasmids were verified by restriction analysis and DNA sequencing.

In Vitro Expression-- For in vitro studies, proteins encoded by the pMDR7.M and pMDR7.F plasmids were expressed using the rabbit reticulocyte lysate-coupled transcription/translation system (Promega). Reactions were incubated at 30 °C for 1.5 h in the presence of [³⁵S]methionine. Labeled proteins were separated by SDS-PAGE (14) and detected by autoradiography.

Immunoprecipitation-- Dynabeads cross-linked with the anti-myc or anti-FLAG antibodies were used for immunoprecipitations. The beads were first coated with either the myc or FLAG antibodies (1.5 µg/10⁷ beads) by cross-linking using 20 mM dimethylpimelimidate in 0.2 M triethanolamine at pH 9 for 45 min at room temperature. The beads were then washed in triethanolamine for 2 h, washed three times in phosphate-buffered saline (PBS), and stored at 4 °C. For in vitro immunoprecipitation, Myc and FLAG-tagged P-gps were co-expressed and ³⁵S-labeled in the complete transcription-translation system. Dynabeads cross-linked with the myc antibody were added to the labeled protein and incubated for 2 h at 4 °C with continuous slow rotation. The Dynabeads were then harvested magnetically and washed extensively with PBS before eluting bound proteins in 50 µl of 1% SDS. The eluate was diluted to 500 µl with PBS. For the second immunoprecipitation, Dynabeads cross-linked with the FLAG antibody were added to the eluate and immunoprecipitation performed as described above. After harvesting and washing, bound proteins were eluted with 50 µl of Laemmli sample buffer and analyzed by SDS-PAGE and autoradiography.

For in vivo immunoprecipitations, epitope-tagged proteins encoded by plasmids pcMDR.M and pcMDR.F were expressed in HEK293 cells. Plasmid DNA (5 µg) was complexed with 20 µl of Lipofectin (Life Technologies, Inc.) in 2 ml of Opti-MEM (Life Technologies, Inc.) for 10 min at room temperature. The mixture was added to cells grown to ~80% confluence in 35-mm wells and incubated for 5 h. Cells were returned to Dulbecco's modified Eagles' medium for a further 48-h incubation and then lysed in solubilization buffer (150 mM NaCl, 50 mM Tris-HCl, 0.1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 20 µg/ml aprotinin, 1 mM EDTA, pH 7.4) containing, as appropriate, between 0.5 and 4% detergent (either Nonidet P-40, dodecyl maltoside, or Triton X-100). Lysates were incubated on ice for 1 h and then centrifuged at 100,000 × g for 1 h to sediment insoluble material. The supernatants were incubated overnight at 4 °C with continuous slow rotation with 150 µl of Dynabeads cross-linked with the anti-myc antibody. Dynabeads were harvested magnetically and washed in solubilization buffer. Bound proteins were eluted in 50 µl of Laemmli sample buffer. Immunoprecipitated proteins were analyzed by SDS-PAGE and detected by immunoblotting with the anti-FLAG or anti-myc antibody.

Protein Separation and Autoradiography-- Proteins were separated by SDS-polyacrylamide gel electrophoresis. Gels of radiolabeled proteins were fixed in 25% isopropanol, 10% acetic acid for 30 min, incubated in Amplify^TM (Amersham Pharmacia Biotech) for 30 min, dried, and exposed for 1-24 h.

Unlabeled proteins were detected by immunoblotting, performed as described previously (15) using enhanced chemiluminescence detection (Amersham Pharmacia Biotech) of horseradish peroxidase-conjugated secondary antibodies. Primary antibodies were M2 anti-FLAG (IBI Kodak) and 9E10 anti-myc (Imperial Cancer Research Fund Hybridoma Unit).

Construction of Gene Fusions for Two-hybrid Analysis-- The PCR was used to amplify, from plasmid DNA, sequences encoding the cytoplasmic domains of wild type P-gp and the P-gp-8E mutant in which phosphorylatable serine and threonine residues are replaced by glutamates (5). As controls, the cytoplasmic domains of related ABC transporter proteins, TAP1, TAP2, and CFTR (Table I), were also amplified. PCR primers were designed to add restriction endonuclease recognition sequences for EcoRI and XhoI to facilitate directional cloning. The PCR products were cloned into the yeast two-hybrid vectors pEG202 and pJG4-5 to generate in-frame gene fusions with the LexA DNA binding domain and the "acid blob" transcriptional activation domain, respectively (16). Gene fusions were verified by restriction analysis and DNA sequencing.

Yeast Two-hybrid Analysis-- Protein-protein interactions were studied using the LexA two-hybrid system as described previously (17). The Saccharomyces cerevisiae strain EGY48 (LEU2::pLexAop1-LEU2) containing pSH18-34, a 20-µm plasmid carrying a galactose-inducible lacZ gene (LexAop-Gal1-lacZ), was used as a reporter strain (17). Plasmids were transformed into this strain using the lithium acetate method (18). Transcriptional activation was assessed by plate assay (19). Fusion proteins were tested for nuclear localization and DNA binding by repression assay (17).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Generation of Epitope-tagged P-gps-- Oligonucleotide-directed mutagenesis was used to introduce short epitope tags (myc and FLAG) to the C terminus of P-gp. Cytotoxicity tests for ability to confer drug resistance demonstrated that addition of these tags did not affect the function of P-gp (data not shown).

Co-immunoprecipitation of Epitope-tagged P-gps Expressed in Vitro-- [³⁵S]Methionine-radiolabeled myc- and FLAG-tagged P-gps were co-expressed in vitro in a rabbit reticulocyte system and sequentially immunoprecipitated with antibody cross-linked Dynabeads. Sequential immunoprecipitation with anti-myc and anti-FLAG antibodies (myc/FLAG) gave no recovery of P-gp-FLAG in the final immunoprecipitate (Fig. 1). In contrast, P-gp was recovered in the second precipitation of positive control reactions in which anti-myc antibody was used in both precipitation steps (myc/myc). This suggests that the two differentially tagged populations of P-gp do not associate with each other and that P-gp is monomeric in the absence of detergent.

View larger version (74K): [in this window] [in a new window]   Fig. 1.   Co-immunoprecipitation of myc and FLAG-tagged P-gps in vitro. Autoradiogram of sequential immunoprecipitations of 35S-labeled, epitope-tagged P-gps transcribed and translated in vitro. S1, supernatant from first (myc) immunoprecipitation; P1, pellet from first immunoprecipitation; S2, supernatant from second (myc or FLAG) immunoprecipitation; P2, pellet from second immunoprecipitation. Arrow indicates approximate mobility of P-gp.

Co-immunoprecipitation of Epitope-tagged P-gps Co-expressed in Vivo-- To ascertain whether P-gp is also a monomer in vivo, Myc- and FLAG-tagged P-gps were co-expressed in HEK293 cells by transient transfection. Epitope-tagged P-gps were recovered in immunoprecipitates from cell lysates prepared in the presence of varied concentrations of several different detergents: dodecyl--maltoside, Nonidet P-40, or Triton X-100 (Fig. 2). In controls, when the anti-myc antibody was used for both immunoprecipitation and immunoblotting (myc/myc), P-gp was recovered with each detergent at all the concentrations used, demonstrating the efficacy of the procedure. No P-gp was detected on immunoblots with the anti-FLAG antibody following immunoprecipitation with the anti-myc antibody under any conditions (myc/FLAG). Experiments carried out in reverse using the anti-FLAG antibody for immunoprecipitation and anti-myc antibody for immunoblotting yielded similar results (data not shown). The inability to co-immunoprecipitate FLAG-tagged P-gp with the myc-tagged P-gp in these experiments indicated that the two forms of P-gp did not associate, suggesting that P-gp in the membranes is a monomer.

View larger version (54K): [in this window] [in a new window]   Fig. 2.   Immunoprecipitation of epitope-tagged P-gps from detergent-solubilized cells. Immunoblots of anti-myc precipitated epitope-tagged P-gps expressed in HEK293 cells solubilized with dodecyl--maltoside, Nonidet P-40 (NP40), or Triton X-100 (A, B, and C, respectively) at the detergent concentrations indicated. Blots were probed with either anti-myc (myc/myc) or anti-FLAG (myc/FLAG) antibodies. Arrow indicates the approximate mobility of P-gp.

Intramolecular Interactions between the Nucleotide-binding Domains of P-gp-- To determine whether P-gp oligomerizes in the absence of detergents and to assess the possibility of very weak or transient interactions, we tested the soluble domains of P-gp in the yeast two-hybrid system. The large cytoplasmic domains of P-gp were expressed as fusion proteins (Table I) as were the cytoplasmic domains of two other ABC transporters, CFTR and TAP, which served as controls. Fusion proteins were expressed singly in yeast and tested for expression, nuclear localization, and DNA binding activity (for binding domain fusions) to exclude autoactivation of reporter genes, as described previously (17). Fusion proteins were then co-expressed in pairwise combinations and tested for interaction by assaying -galactosidase and leucine reporters. All fusions were shown to be capable of positive interaction with other fusion proteins in related studies (data not shown).

No combination of wild-type P-glycoprotein cytoplasmic domains produced a detectable interaction with either reporter gene (Table II). Interestingly, the P-gp-8E mutant, derived from a form of P-gp generated by site-directed mutagenesis to study the effects of phosphorylation of the linker region, interacted weakly with the C-terminal cytoplasmic domain of P-gp in the more sensitive leucine assay. These results suggest that direct interaction between the cytoplasmic domains does not normally occur, but may be induced upon phosphorylation of the linker.

View this table: [in this window] [in a new window]   Table II Yeast two-hybrid analysis of interaction of P-gp cytoplasmic domains Reporter gene activities were determined by plate assay (lacZ, blue color on plates containing 5-bromo-4-chloro-3-indoyl -D-galactopyranoside; LEU2, growth on minimal medium plates lacking leucine). Positive interactions were dependent on galactose induction of expression of the activation domain fusion. Relative activities were scored qualitatively from no detectable activity (), to high activity (+++).

Control experiments with the cytoplasmic domains of CFTR were likewise negative (data not shown). In contrast, the cytoplasmic domains of the transporter associated with antigen processing (TAP) did interact in this system (Table III). This observation is also consistent with a recent study showing interaction of TAP1 and TAP2 cytoplasmic domains by co-immunoprecipitation (20). It is important to note that, unlike the homologous domains in P-gp and CFTR, the TAP cytoplasmic domains reside on separate polypeptides that must dimerize to form a complete ABC transporter. Furthermore, the TAP transporter does not include a regulatory domain homologous to either the P-gp linker or the CFTR R domain.

View this table: [in this window] [in a new window]   Table III Yeast two-hybrid analysis of interaction of TAP1 and TAP2 Reporter gene activities (lacZ, LEU2) were determined by plate assay (lacZ, blue color on plates containing 5-bromo-4-chloro-3-indoyl -D-galactopyranoside; LEU2, growth on minimal medium plates lacking leucine). Relative activities were scored qualitatively from no detectable activity () to high activity (+++).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Efforts to examine the oligomerization of P-gp have been hampered by the technical difficulties of studying protein-protein interactions of integral membrane proteins. The detergents used to solubilize membrane proteins may give rise to protein aggregates or disrupt normal associations. For these reasons, we chose to conduct oligomerization experiments both in the presence and absence of a variety of detergents. Expression of tagged P-gps in vitro in the absence of microsomal membranes obviates the need for detergent solubilization. To complement this approach, we investigated P-gp oligomerization in vivo, using a number of detergents in an attempt to control for detergent-specific effects. Detergents with differing properties in terms of head groups, hydrocarbon chain length, and critical micellar concentration were used for solubilization. Concentrations of 0.5-4% were used for each detergent, enabling efficient solubilization of P-gp and a wide enough range of concentrations to examine possible artifactual aggregation of the protein. We consistently isolated P-gp as a monomer in vitro and in vivo under all conditions tested.

In vivo two-hybrid analysis allowed us to test for possible weak dimerization interactions between cytoplasmic domains on individual molecules as well as intramolecular interactions between the two cytoplasmic domains of a single molecule. The latter type of interaction could have both structural and mechanistic implications. Although the yeast two-hybrid system provides an extremely sensitive assay for interaction, allowing detection of weak and/or transient protein-protein interactions, we were unable to detect interaction between the cytoplasmic domains of P-gp or of CFTR. These results suggest that dimerization does not occur via direct interaction of the cytoplasmic domains of these ABC transporters.

In contrast, interaction was observed between TAP1 and TAP2, which represent discrete subunits that upon heterodimerization constitute the four-domain structure present in a single monomer of P-gp or CFTR. While a monomer is apparently sufficient for P-gp activity (21), TAP activity requires both the TAP1 and TAP2 subunits, each consisting of one membrane domain and one cytoplasmic domain (NBD). The interaction of TAP cytoplasmic domains suggests that these regions may be directly involved in TAP1/TAP2 dimerization.

The absence of cytoplasmic domain interaction for P-gp is consistent with the low resolution structure of P-gp (12) in which two distinct lobes, predicted to be the two cytoplasmic domains, are observed. However, intramolecular interactions between cytoplasmic domains of P-gp may occur under certain conditions, as implied by the weak interaction that we observe when a pseudo-phosphorylated form of cytoplasmic domain 1 is used in interaction assays. A recent study (22) suggests that, at least during part of the reaction cycle, the two nucleotide-binding domains of P-gp are in close proximity to one another, as they could be chemically cross-linked. Interestingly, the cross-linking only occurred at 37 °C and was inhibited by ATP, suggesting a role for structural flexibility and substrate-induced conformational change.

Our data suggest a model in which four-domain ABC transporters, such as P-gp and CFTR, are normally monomeric, whereas two-domain ABC transporters, such as TAP, may dimerize through direct protein-protein interaction to form a paradigmatic, four-domain ABC protein. However, the data do not rule out the possibility that four-domain transporters may oligomerize through indirect interaction. A recent paper by Wang and colleagues (23) identifies a protein, CAP70, which interacts with CFTR, facilitating oligomerization and modulating CFTR channel activity. This interaction requires both ATP and phosphorylation of the R domain and yields a more active channel configuration. Likewise, P-gp may be regulated by oligomerization in response to effectors or via specific binding proteins. Such interactions may account for at least some of the cases in which apparent P-gp oligomers have been observed.

	ACKNOWLEDGEMENTS

The two-hybrid interaction reagents were the generous gift of Lauren Ha and Roger Brent.

	FOOTNOTES

^* This work was supported by the Cancer Research Campaign and the Imperial Cancer Research Fund.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Oxagen Limited, 91 Milton Park, Abingdon, Oxon OX14 4RY, UK.

^§ Present address: Dept. of Clinical Chemistry, Albert Szent-Gyorgyi Medical University, Somogyi Bela Ter 1., Szeged, H-6722, Hungary.

^¶ Howard Hughes International Research Scholar.

To whom correspondence should be addressed. Present address: The Marine Biological Laboratory, 7 MBL St., Woods Hole, MA 02543. E-mail: gbegley@mbl.edu.

Published, JBC Papers in Press, August 8, 2001, DOI 10.1074/jbc.C100345200

	ABBREVIATIONS

The abbreviations used are: P-gp, P-glycoprotein; CFTR, cystic fibrosis transmembrane regulator; MDR, multidrug resistance; NBD, nucleotide-binding domain; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; TAP, transporter associated with antigen presentation; PAGE, polyacrylamide gel electrophoresis; ABC, ATP-binding cassette.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This article has been cited by other articles:

				Y. Yang, Y. Liu, Z. Dong, J. Xu, H. Peng, Z. Liu, and J.-T. Zhang Regulation of Function by Dimerization through the Amino-terminal Membrane-spanning Domain of Human ABCC1/MRP1 J. Biol. Chem., March 23, 2007; 282(12): 8821 - 8830. [Abstract] [Full Text] [PDF]

				J. Xu, Y. Liu, Y. Yang, S. Bates, and J.-T. Zhang Characterization of Oligomeric Human Half-ABC Transporter ATP-binding Cassette G2 J. Biol. Chem., May 7, 2004; 279(19): 19781 - 19789. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/39/36075    most recent C100345200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Taylor, J. C. Articles by Begley, G. S. Search for Related Content PubMed PubMed Citation Articles by Taylor, J. C. Articles by Begley, G. S.