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

This Article Full Text (PDF) All Versions of this Article: 17/15/2287 03-0107fjev1    most recent Alert me when this article is cited Alert me if a correction is posted 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 STENHAM, D. R. Articles by LINTON, K. J. Search for Related Content PubMed PubMed Citation Articles by STENHAM, D. R. Articles by LINTON, K. J.

An atomic detail model for the human ATP binding cassette transporter P-glycoprotein derived from disulfide cross-linking and homology modeling¹

DANIELLA R. STENHAM², JEFF D. CAMPBELL^*,2, MARK S. P. SANSOM^*, CHRISTOPHER F. HIGGINS, IAN D. KERR^,3 and KENNETH J. LINTON³

MRC Clinical Sciences Centre, Imperial College School of Medicine, Hammersmith Hospital Campus, London, W12 0NN, UK;
^* Laboratory of Molecular Biophysics, University of Oxford, Oxford, OX1 3QU, UK; and
School of Biomedical Sciences, University of Nottingham, Queen’s Medical Centre, Nottingham, NG7 2UH, UK

³Correspondence: K.J.L., MRC Clinical Sciences Centre, Imperial College School of Medicine, Hammersmith Hospital Campus, Du Cane Road, London, W12 0NN, UK; E-mail: kenneth.linton{at}csc.mrc.ac.uk and I.D.K., School of Biomedical Sciences, University of Nottingham, Queen’s Medical Centre, Nottingham, NG7 2UH, UK; E-mail: ian.kerr{at}nottingham.ac.uk

SPECIFIC AIMS

The crystal structures of two prokaryotic ATP binding cassette (ABC) transporters, MsbA and BtuCD, have recently been solved but they differ in their tertiary structure and domain interfaces. We have produced an atomic scale model of the paradigm human ABC transporter, the multidrug resistance P-glycoprotein (P-gp), based on a combination of the available structural data and novel chemical cross-linking data to determine the interface between the two transmembrane domains.

PRINCIPAL FINDINGS

P-gp is comprised of four domains: two transmembrane domains (TMDs), which contain the drug binding sites and define the translocation pathway across the membrane; and two cytoplasmic nucleotide binding domains (NBDs), which couple the energy associated with ATP binding and hydrolysis to drug transport. P-gp is the best-characterized human ABC transporter in terms of catalytic and transport function, and a detailed structure for P-gp would be a significant advance in elucidating a molecular mechanism for substrate binding and allosteric communication, necessary for ABC transporters to function.

1. The tertiary structure of E. coli MsbA is incorrect
MsbA, the lipid A transporter from Escherichia coli, has the highest end-to-end sequence similarity to P-gp of any prokaryotic ABC transporter whereas BtuCD, the vitamin B12 transporter from E. coli, has sequence similarity to P-gp only in the NBDs. The MsbA polypeptide contains a single TMD and NBD, and dimerizes to form a functional transporter. The E. coli MsbA crystal contains a dimer in which the TMDs are in contact at the periplasmic ends of transmembrane (TM) -helices. However, it contains no TMD:TMD interface at the intracellular ends of the TM -helices, and the two NBDs are separated by 50 Å. This tertiary structure is incompatible with disulfide cross-linking studies of P-gp that demonstrate both the close proximity of the TM -helices at their intracellular ends and an NBD:NBD interface. It is incompatible with electron microscopy (EM) data for P-gp, which show that the TMDs form a chamber in the membrane open at the extracellular face and closed at the intracellular face. In contrast, the tertiary structure of BtuCD contains a parallel TMD:TMD interface (i.e., the TMDs are in contact at both the extracellular and intracellular ends of the membrane) and an NBD:NBD interface comparable to that obtained in two independent structural studies of isolated prokaryotic NBDs: Rad50 and MJ0796.

2. An initial model based on the secondary structure of MsbA and the tertiary structure of the BtuD dimer fails to satisfy biochemical and EM data
Each half of P-gp (TMD-NBD) was modeled onto the structure of the homologous MsbA. Next, the two halves of P-gp were associated to reflect the consensus NBD:NBD interface observed in BtuCD, Rad50, and MJ0796, thus generating P-gp-model A. This model was inconsistent with EM data available for P-gp, only partially compatible with previously published cysteine cross-linking data, and irreconcilable with our novel cysteine cross-linking data (point 3).

3. Cross-linking of introduced triple cysteines defines the TMD:TMD interface
The TMD:TMD interface in E. coli MsbA predicts that the extracellular ends of the 2nd and 11th TM -helices (TM2 and TM11) of P-gp will be in close proximity. Given the twofold rotational symmetry of the MsbA dimer, P-gp TM5 and TM8 are predicted to be close at their extracellular ends. To test the veracity of this interface, three consecutive residues toward the extracellular ends of TM2, TM5, TM8, and TM11 were replaced with cysteine in P-gp-cys^- (a variant of P-gp in which the endogenous cysteines have been replaced by serine). Triple cysteines were introduced to maximize the chance that a sulfhydryl group on one TM would be suitably positioned to form an intramolecular cross-link with a sulfhydryl from an adjacent TM. All mutant isoforms were shown to adopt a near-native protein fold in a two-color flow cytometry assay by virtue of their ability to transport rhodamine 123 and their sensitivity to the inhibitor cyclosporin A.

The proximity of the introduced cysteines was tested using a series of sulfhydryl-reactive, di-maleimide cross-linkers. Cross-linked proteins were identified by altered mobility on SDS-PAGE. As expected, P-gp-cys^- and mutant isoforms with triple cysteine residues in a single TM -helix showed the same electrophoretic mobility irrespective of the presence of chemical cross-linking reagents (Fig. 1 , lanes b, d, f, h, and j). However, when cysteines were introduced into TM2 and TM11 of the same protein, a shift in mobility was observed after treatment with the cross-linking reagents (Fig. 1 , lanes k, l). Similar results were obtained when TM5 and TM8 contained triple cysteines (Fig. 1 , lanes m, n). The results were specific to these helix pairs: mutant P-gp-cys^- isoforms with triple cysteines in TM2 and TM8 or in TM5 and TM11 showed no evidence of cross-linking (Fig. 1 , lanes p, r). Cotransfection experiments (Fig. 1 , lanes t, v) demonstrated that the cross-linking between TMs 2 and 11 and between TMs 5 and 8 was intramolecular rather than intermolecular. These data afford confidence that the extracellular ends of TM2 and TM5 are in close proximity to those of TM11 and TM8, respectively, in agreement with the TMD:TMD interface observed at the periplasmic face of MsbA.

View larger version (15K): [in this window] [in a new window]   Figure 1. Cross-linking analysis of P-gp mutant isoforms. PAGE and Western blot analysis of the indicated P-gp mutant isoforms treated (+) or not (–) with the chemical cross-linking agent p-PDM (similar results were observed with both o-PDM and BMH; data not shown). Membrane fractions in lanes s, t, u, and v were prepared from cells in which two mutant isoforms were coexpressed. The 170 kDa protein represents mature glycosylated P-gp and the 140 kDa protein represents immature P-gp. Electrophoretically retarded protein was observed only with P-gp-TM2+11 and P-gp-TM5+8 and only after treatment with p-PDM (lanes l and n, respectively).

4. Biochemical data and EM data are compatible and can be incorporated into an atomic scale model
Our cross-linking data, which provide the first experimental support for contact between the TMDs at their extracellular ends, augment earlier data demonstrating that the intracellular ends of the TMDs are also in contact. This led us to speculate that the TMDs of P-gp may be approximately parallel (in agreement with the conclusion from a recent EM study of P-gp). A simple closure of the TMDs of MsbA to form a parallel TMD:TMD interface would, however, render impossible the consensus association between NBDs that has been observed in the crystal structures of BtuCD, MJ0796, and Rad50. We posited three arguments to reconcile this discrepancy. The first is that the helical packing within the TMDs of MsbA is different from P-gp, which seems highly unlikely given the high degree of sequence similarity between the two proteins. The second is that the crystallized E. coli MsbA is misfolded. This again seems unlikely as the partially resolved MsbA NBD suggests a typical NBD protein fold and, if the TMD fold had been disrupted during purification (by detergents, for instance), the protein would be less likely to crystallize. Our third hypothesis is that the orientation of the E. coli MsbA TMD with respect to its cognate NBD observed in the crystal structure does not reflect the physiological association. This assumption is embodied in P-gp-model B and is achieved by rotation of the NBDs 150° with respect to their cognate TMDs (Fig. 2 ). This model contains a consensus NBD:NBD interface and a parallel TMD:TMD interface. The model is entirely consistent with our chemical cross-linking data as well as data provided in extensive studies by Loo and Clarke and from EM studies of P-gp.

View larger version (52K): [in this window] [in a new window]   Figure 2. Reconciliation of cross-linking data with a P-gp model. P-gp-Model-B (seen from two orthogonal directions, A and B, was generated as described in the text by rotation of each NBD with respect to its cognate, TMD. The final model contains a parallel TMD:TMD interface (blue and gold subunits) and a consensus NBD:NBD interface (green and purple subunits). B) TM -helices (numbered H1 to H12) define a central lumen consistent with EM data, and the relative inter-helix distances are consistent with available chemical cross-linking data. The extracellular loops connecting TM1-2 and TM7-8 have been removed for clarity.

CONCLUSIONS AND SIGNIFICANCE

We have used complementary research techniques in the form of experimental cysteine-directed chemical cross-linking and computational homology modeling to produce a testable, atomic scale model for the human multidrug transporter P-gp (Fig. 3 ). Essentially, the crystal structure of E. coli MsbA was used to independently model the structures of the four domains of P-gp. The tertiary structure of the P-gp–model B was based on the independently observed, consensus NBD:NBD interface, which would coordinate nucleotide between the two domains. The TMD:TMD interface, empirically tested by chemical cross-linking, is derived from the twofold rotational symmetry of the equivalent interface of MsbA but with each TMD perpendicular to the plane of the membrane. So arranged, the TMDs form a compact helix bundle surrounding a chamber that is open at the extracellular surface, entirely consistent with the low-resolution EM data for P-gp. The chamber is closed at the intracellular surface largely by the convergence of TM6 and TM12, similar to the arrangement described by helix 5 (and 5') from the homo-dimeric BtuC and consistent with the published finding that the NBDs of P-gp are not responsible for closing the chamber.

View larger version (21K): [in this window] [in a new window]   Figure 3. Schematic representation of the protocol used to generate alternative models of P-gp. The color scheme of domains is comparable to Fig. 2 . The consensus NBD:NBD interaction is denoted by the connecting arcs. By rotation of the NBD with respect to its cognate TMD, denoted by circular arrows, we are able to produce Model-B, which agrees with EM data and cross-linking data for P-gp.

An atomic scale model for P-gp will allow the position of amino acids, substitutions of which affect substrate specificity, to be mapped in three dimensions and should aid identification of the drug binding sites in the TMDs. The model will be important to the design of future fluorescence resonance energy transfer, electron paramagnetic resonance, and cross-linking experiments to address the question of interdomain communication within P-gp that is fundamental to the control of the drug transport cycle. As a paradigm for ABC transporters, a model for P-gp has direct implications for understanding the mechanisms of diverse physiological processes including prokaryotic nutrient uptake, mammalian antigen presentation, cystic fibrosis, as well as resistance to chemotherapy.

FOOTNOTES

¹ To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.03-0107fje; doi: 10.1096/fj.03-0107fje

² D.R.S. and J.D.C. contributed equally to this manuscript.

This article has been cited by other articles:

				J. K. Zolnerciks, C. Wooding, and K. J. Linton Evidence for a Sav1866-like architecture for the human multidrug transporter P-glycoprotein FASEB J, December 1, 2007; 21(14): 3937 - 3948. [Abstract] [Full Text] [PDF]

				K. J. Linton Structure and Function of ABC Transporters Physiology, April 1, 2007; 22(2): 122 - 130. [Abstract] [Full Text] [PDF]

				D. Y. Haubertin, H. Madaoui, A. Sanson, R. Guerois, and S. Orlowski Molecular Dynamics Simulations of E. coli MsbA Transmembrane Domain: Formation of a Semipore Structure Biophys. J., October 1, 2006; 91(7): 2517 - 2531. [Abstract] [Full Text] [PDF]

				R. G. Deeley, C. Westlake, and S. P. C. Cole Transmembrane Transport of Endo- and Xenobiotics by Mammalian ATP-Binding Cassette Multidrug Resistance Proteins. Physiol Rev, July 1, 2006; 86(3): 849 - 899. [Abstract] [Full Text] [PDF]

				H. Omote and M. K. Al-Shawi Interaction of Transported Drugs with the Lipid Bilayer and P-Glycoprotein through a Solvation Exchange Mechanism Biophys. J., June 1, 2006; 90(11): 4046 - 4059. [Abstract] [Full Text] [PDF]

				N. Maki, K. Moitra, P. Ghosh, and S. Dey Allosteric Modulation Bypasses the Requirement for ATP Hydrolysis in Regenerating Low Affinity Transition State Conformation of Human P-glycoprotein J. Biol. Chem., April 21, 2006; 281(16): 10769 - 10777. [Abstract] [Full Text] [PDF]

				O. Dalmas, C. Orelle, A.-E. Foucher, C. Geourjon, S. Crouzy, A. Di Pietro, and J.-M. Jault The Q-loop Disengages from the First Intracellular Loop during the Catalytic Cycle of the Multidrug ABC Transporter BmrA J. Biol. Chem., November 4, 2005; 280(44): 36857 - 36864. [Abstract] [Full Text] [PDF]

				J. Dong, G. Yang, and H. S. Mchaourab Structural Basis of Energy Transduction in the Transport Cycle of MsbA Science, May 13, 2005; 308(5724): 1023 - 1028. [Abstract] [Full Text] [PDF]

				K. Pleban, S. Kopp, E. Csaszar, M. Peer, T. Hrebicek, A. Rizzi, G. F. Ecker, and P. Chiba P-Glycoprotein Substrate Binding Domains Are Located at the Transmembrane Domain/Transmembrane Domain Interfaces: A Combined Photoaffinity Labeling-Protein Homology Modeling Approach Mol. Pharmacol., February 1, 2005; 67(2): 365 - 374. [Abstract] [Full Text] [PDF]

				M. F. Rosenberg, R. Callaghan, S. Modok, C. F. Higgins, and R. C. Ford Three-dimensional Structure of P-glycoprotein: THE TRANSMEMBRANE REGIONS ADOPT AN ASYMMETRIC CONFIGURATION IN THE NUCLEOTIDE-BOUND STATE J. Biol. Chem., January 28, 2005; 280(4): 2857 - 2862. [Abstract] [Full Text] [PDF]

				T. T. Tran, A. Mittal, T. Aldinger, J. W. Polli, A. Ayrton, H. Ellens, and J. Bentz The Elementary Mass Action Rate Constants of P-gp Transport for a Confluent Monolayer of MDCKII-hMDR1 Cells Biophys. J., January 1, 2005; 88(1): 715 - 738. [Abstract] [Full Text] [PDF]

				D. Situ, A. Haimeur, G. Conseil, K. E. Sparks, D. Zhang, R. G. Deeley, and S. P. C. Cole Mutational Analysis of Ionizable Residues Proximal to the Cytoplasmic Interface of Membrane Spanning Domain 3 of the Multidrug Resistance Protein, MRP1 (ABCC1): GLUTAMATE 1204 IS IMPORTANT FOR BOTH THE EXPRESSION AND CATALYTIC ACTIVITY OF THE TRANSPORTER J. Biol. Chem., September 10, 2004; 279(37): 38871 - 38880. [Abstract] [Full Text] [PDF]

				A. Rothnie, J. Storm, J. Campbell, K. J. Linton, I. D. Kerr, and R. Callaghan The Topography of Transmembrane Segment Six Is Altered during the Catalytic Cycle of P-glycoprotein J. Biol. Chem., August 13, 2004; 279(33): 34913 - 34921. [Abstract] [Full Text] [PDF]

				G. Tombline, L. A. Bartholomew, I. L. Urbatsch, and A. E. Senior Combined Mutation of Catalytic Glutamate Residues in the Two Nucleotide Binding Domains of P-glycoprotein Generates a Conformation That Binds ATP and ADP Tightly J. Biol. Chem., July 23, 2004; 279(30): 31212 - 31220. [Abstract] [Full Text] [PDF]

				A. Haimeur, G. Conseil, R. G. Deeley, and S. P.C. Cole Mutations of Charged Amino Acids in or near the Transmembrane Helices of the Second Membrane Spanning Domain Differentially Affect the Substrate Specificity and Transport Activity of the Multidrug Resistance Protein MRP1 (ABCC1) Mol. Pharmacol., June 1, 2004; 65(6): 1375 - 1385. [Abstract] [Full Text] [PDF]

				T. W. Loo, M. C. Bartlett, and D. M. Clarke Val133 and Cys137 in Transmembrane Segment 2 Are Close to Arg935 and Gly939 in Transmembrane Segment 11 of Human P-glycoprotein J. Biol. Chem., April 30, 2004; 279(18): 18232 - 18238. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) All Versions of this Article: 17/15/2287 03-0107fjev1    most recent Alert me when this article is cited Alert me if a correction is posted 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 STENHAM, D. R. Articles by LINTON, K. J. Search for Related Content PubMed PubMed Citation Articles by STENHAM, D. R. Articles by LINTON, K. J.