CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS

CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS CHAPTER 2 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS

37 MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS ANDRÁS VÁRADI, GÁBOR E TUSNÁDY AND BALÁZS SARKADI INTRODUCTION The most recent annotation* of the human genome sequence revealed 48 genes for ABC proteins, which were grouped into seven subclasses, from ABCA to ABCG (see: http:// nutrigene.4t.com/humanabc.htm; see also Chapter 3) As detailed elsewhere in this book, ABC proteins in all organisms can be recognized by their conserved motifs within the ATPbinding domains The majority of these proteins are membrane embedded and fulfill various membrane transport or regulatory functions (hence the designation ABC transporters), and our present review deals with such human proteins In contrast, the human ABCE and ABCF subfamilies contain proteins with no known transmembrane domains; therefore these are outside the scope of this chapter It is generally accepted that the minimum functional unit requirement for an ABC transporter is the presence of two transmembrane domains (TMD; IM in Chapter 1) and two ATPbinding cassette (ABC) units These may be present within one polypeptide chain (‘full transporters’), or within a membrane-bound homo- or heterodimer of ‘half transporters’ At present there are no high-resolution structural data available for any mammalian ABC transporter; therefore computer modeling and laborious biochemical experiments are CHAPTER necessary to elucidate membrane topology, i.e the position and orientation of membrane-spanning segments within the polypeptide chain The generally applied experimental methods include epitope insertion, localization of glycosylation sites, limited proteolysis and immunochemical techniques Several computer-assisted empirical prediction methods are available to generate the hydrophobicity profile for a putative transmembrane protein, but such an analysis may provide only a basis for developing experimental strategies for the elucidation of the actual membrane topology In the present review we summarize the available data on the membrane topology for various ABC transporters by examining the distinct subfamilies We discuss predicted topology models and their experimental reinforcement or negation, and assess the relationship between phylogenetic linkages and the arrangement of membrane topology patterns We call the reader’s attention to a recent thematic review, also analyzing the problems of membrane topology models within the ABCprotein kingdom (Dean et al., 2001) METHODOLOGY In the present study, for subfamily classification, we used the Human ABC Proteins Database * This chapter reflects the information available in June, 2001 ABC Proteins: From Bacteria to Man ISBN 0-12-352551-9 Copyright 2003 Elsevier Science Ltd All rights of reproduction in any form reserved 38 ABC PROTEINS: FROM BACTERIA TO MAN (http://nutrigene.4t.com/humanabc.htm, last update May 20, 2001) We included in the analysis only those members of a given subfamily whose full sequences were known We chose the sequences of the longest and/or the most common splicing variants, in cases where alternatively spliced cDNAs were described We considered a membrane topology to be ‘established’ if it was supported by independent experimental data Protein sequences were aligned by the ClustalW server (http://www.ebi.ac.uk/ clustalw), and hydrophobicity plots were generated according to a described method (von Heijne, 1992) For computer-assisted predictions the HMMTOP (Hidden Markov Model for Topology Prediction) transmembrane topology prediction server was applied (Tusnády and Simon, 1998, 2001), which is freely available for noncommercial users (http://www enzim.hu/ hmmtop) This method is based on the principle that the topology of transmembrane proteins is determined by considering the maximal divergence in the amino acid composition of defined sequence segments The results of the above analyses are presented as hydrophobicity plots of the aligned sequences within a given subfamily, and the phylogenetic trees of the subfamilies are also shown together with the plots MEMBRANE TOPOLOGY OF THE ABC TRANSPORTERS ABCA SUBFAMILY The Human ABC Database lists 14 members for the ABCA subfamily, from which currently seven are represented by full cDNA sequences We have analyzed the membrane topology of the following seven proteins: A1 ϭ 2261 aa; A2 ϭ 2436 aa; A3 ϭ 1704 aa; A4 ϭ 2273 aa; A7 ϭ 2146 aa; A8 ϭ 1581 aa; and A12 ϭ 2277 aa The ABCA subfamily, as examined here, contains ‘full transporters’, and it is noteworthy that these proteins have the largest molecular masses within the entire human ABC protein family Within this subfamily the function and membrane topology of only two members have been studied in detail These are ABCA1, the gene associated with Tangier disease, and ABCA4, the retina-specific ABC transporter whose mutations cause severe retinopathies Both of these proteins (see Chapters 23 and 28) are located in the plasma membrane, and they share 50% identical amino acids All the suggested membrane topology models agree that both halves of these proteins contain one TMD and one ABC domain, and the TMDs consist of six transmembrane helices However, there are several different models for the actual distribution of the helices In the case of ABCA1, the first model, elaborated by Luciani et al (1994), predicted a large cytoplasmic domain at the extreme N-terminal part, and another large cytoplasmic domain (‘regulatory domain’) in the central portion of the protein A hydrophobic segment with a hairpin orientation was predicted to be located within this cytoplasmic regulatory domain, thus anchoring it to the plasma membrane After the cloning of ABCA4, a topology model smilar to ABCA1 was first suggested (Allikmets et al., 1997) However, Illing et al (1997) described a different model for ABCA4, with two large extracellular domains (ECDs) between the first and second transmembrane helices in each predicted TMD In this model the first TM helix within the N-terminal half was placed close to the N-terminus, while the first predicted TM helix in the C-terminal half was the very same hydrophobic segment suggested to form a hairpin with a membraneassociated structure in the previous models This latter model thus predicted no large intracellular regulatory domain for ABCA4 A combination of the two different models was also published (Azarian and Travis, 1997; Sun et al., 2000), in which the N-terminal half of the ABCA4 protein was represented by the Illing model, while the C-terminal half contained a large cytoplasmic (regulatory) domain, as suggested by the model of Luciani et al (1994) In the case of ABCA4, an elegant experimental proof for the Illing model has been recently published (Bungert et al., 2001) Eight functional N-glycosylation sites were mapped by mutagenesis within the bovine ABCA4 sequence, four in the N-terminal half, and four within the C-terminal half These results support the presence of a 600 amino acid extracellular domain within the N-terminal half, and a 275 amino acid extracellular domain within the C-terminal half of the protein The authors also presented experimental data suggesting that the two extracellular domains are linked by disulfide bridge(s) MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS Regarding ABCA1, after the initial description of its coding region, an in-frame, upstream methionine was discovered, and translation from this codon results in a 60 amino acid extension of the originally expected polypeptide chain (Costet et al., 2000; Pullinger et al., 2000; Santamarina-Fojo et al., 2000) Within this N-terminal 60 amino acid chain, sequence analysis predicted the presence of a transmembrane helix (between residues 22 and 44), and a potentially cleavable signal sequence (between amino acids 45 and 46) These predictions triggered experiments to test the membrane topology dictated by the presence of a TM helix (TMH) in the proximity of the N-terminus, which may serve as an anchor signal and orient a large loop in the N-terminal half of the protein extracellularly Fitzgerald et al (2001) expressed various truncated and tagged versions of human ABCA1, and demonstrated that the loop from amino acids 44 to 640 indeed has an extracellular orientation Tanaka et al (2001) obtained similar results, and they also predicted a large extracellular loop within the C-terminal half of the protein (between amino acids 1368 and 1655) However, the two publications contain contradicting data concerning whether cleavage occurs at position 45 According to the results of Tanaka et al (2001), the polypeptide chain is cleaved at this position during maturation, and they suggest that ABCA1 is present in the plasma membrane with a large N-terminal segment extending to the extracellular space Fitzgerald et al (2001), using a bulkier N-terminal GFP-tag, found no cleavage of ABCA1 at this position In summary, the experimental data obtained for ABCA1 and ABCA4 support a similar membrane topology for the two proteins, with a domain arrangement of TMH1-ECD1TMH(2–6)-ABC1-TMH7-ECD2-TMH(8–12)ABC2, where H indicates helix and ECD, an extracellular domain The primary sequence alignment and hydropathy analysis of the aligned ABCA sequences (as presented in Figure 2.1) are in line with this assumption In the projection of Figure 2.1, the location of TM helices within ABCA1, as predicted by Tanaka et al., is used as a guide to label similarities within the hydrophobicity patterns of the members in this subfamily Indeed, a similar domain arrangement and membrane topology can be predicted for all the currently known members of the ABCA subfamily However, the alignment also reveals that ABCA3 and ABCA8 have relatively short ECDs within their N-terminal and C-terminal halves ABCB SUBFAMILY Three members of this subfamily are ‘full transporters’, with ABCB1 (MDR1 or Pgp), ABCB4 (MDR3) and ABCB11 (sisterPgp or BSEP), all localized in the plasma membrane, in polarized cells in the apical membrane All the other family members are ‘half transporters’, including ABCB2 and ABCB3 (TAP1 and TAP2) residing in the endoplasmic reticulum ABCB6, ABCB7, ABCB8 and ABCB10 are mitochondrial membrane proteins, while ABCB9 has a putative lysosomal localization It is well established that ABCB2/TAP1 and ABCB3/TAP2 form a noncovalent heterodimer, and actively translocate peptides into the endoplasmic reticulum (see Chapter 26) Dimer formation for the mitochondrial half transporters has not been experimentally shown, but the presence of mitochondrial targeting signals within the N-terminal regions of both ABCB7 (Csere et al., 1998), and ABCB8 (Hogue et al., 1999) has been demonstrated For the first recognized mammalian ABC transporter, the original topology model of ABCB1 (MDR1-Pgp) predicted six TM helices in both TMDs of the protein, each followed by an ABC domain (Chen et al., 1986) This membrane topology has been fully supported by later epitope insertion experiments (Kast et al., 1995, 1996) ABCB4 and ABCB11 are close relatives of ABCB1, and similar membrane topology arrangements can be predicted for both of these proteins (Figure 2.2) Hydrophobicity plots of the aligned sequences reveal that the positioning of predicted helices in the ABCB family of half transporters is more closely related to the C-terminal halves (TMD2) of the full transporters than to the N-terminal halves, as presented in Figure 2.2 The hydrophobicity plots of the half transporters support the six TM helix model in their TMDs, although the N-terminal regions are clearly extended, and contain hydrophobic regions which may correspond to TM helices Indeed, in the case of TAP1 and TAP2, additional four and three TM helices, respectively, were predicted (Abele and Tampe, 1999) It is worth mentioning that currently no reliable algorithms are available for such predictions, and visual inspection and heuristic adjustments 39 40 ABC PROTEINS: FROM BACTERIA TO MAN A1 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 A7 200 100 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 A4/ABCR 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 A3 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 A12 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 A8 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 A2 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 Figure 2.1 Relative similarity dendrogram and the hydrophobicity plots of the aligned sequences of the ABCA subfamily transporters The shaded areas show transmembrane helices with locations supported by experimental results from studies of ABCA1 and ABCA4 (see text) of the sequences can be most helpful for predicting the probable location of some TM helices, such as TM2 in ABCB6, or TM6 in ABCB8 ABCC SUBFAMILY The ABCC subfamily consists of 11 members in the human genome and most of these (ABCC1–6 or MRP1–6) have been identified as active membrane transporters for various organic anions, several of which are discussed in other chapters in this volume Although the majority of the characterized ABC proteins are active pumps, the cystic fibrosis transmembrane conductance regulator, ABCC7 (CFTR), is a chloride channel which may also regulate other channel proteins The sulfonylurea receptors, ABCC8 (SUR1) and ABCC9 (SUR2), are best described as intracellular ATP sensors, regulating the permeability of specific Kϩ channels (with which they form transmembrane complexes) Nothing is currently known about the function of ABCC10 and ABCC11 The membrane topology of human CFTR/ ABCC7 was originally predicted based on the MDR1 model, and supported experimentally by glycosylation site insertion mutagenesis (Chang et al., 1994) This study strongly supported the original model, with a TMD1-ABC1R-TMD2-ABC2 domain arrangement Each TMD is predicted to consist of six TM helices, and a regulatory domain (R) is present between MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 B1 MDR1 100 200 300 400 500 600 700 800 900 1000 1100 1200 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 B4 MDR3 100 200 300 400 500 600 700 800 900 1000 1100 1200 B11 sPgp 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 B6 100 200 300 400 500 600 700 800 B7 100 200 300 400 500 600 700 B2/TAP1 100 200 300 400 500 600 700 800 100 200 300 400 500 600 B9 100 200 300 400 500 600 700 B8 100 200 300 400 500 600 700 B10 100 200 300 400 500 600 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 B3/TAP2 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 700 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 Figure 2.2 Relative similarity dendrogram and the hydrophobicity plots of the aligned sequences of the ABCB subfamily transporters The shaded areas show transmembrane helices, predicted with locations as projected from experimental results based on ABCB1 (see text) the two halves of the protein, which seems to be unique for CFTR The membrane topology models for human ABCC1 have been independently formulated by two research groups (Bakos et al., 1996; Stride et al., 1996) They found that when the CFTR/ABCC7 and MRP1/ABCC1 sequences were aligned, the hydrophobicity analysis of the aligned sequences yielded a close matching of putative transmembrane segments, thus also suggesting a six plus six transmembrane helix topology for MRP1/ABCC1 However, ABCC1 contains an additional N-terminal segment of about 230 amino acids, which has no counterpart in CFTR/ABCC7 On the basis of the hydropathy profiles and limited proteolysis experiments, the hydrophobic N-terminal segment of ABCC1 was suggested to be membrane embedded, with four to six transmembrane helices (Bakos et al., 1996; Stride et al., 1996) Subsequent investigations of the membrane topology of human ABCC1 by epitope insertion (Kast and Gros, 1997, 1998) and by mutation of glycosylation sites (Hipfner et al., 1997) fully supported the above topology and 41 42 ABC PROTEINS: FROM BACTERIA TO MAN revealed the presence of five TM helices in the N-terminal segment These results indicated a domain arrangement of TMD0-L0-TMD1ABC1-L-TMD2-ABC2, in which TMD0 represents the N-terminal five TM helix extension, while L0 and L represent intracellular linker sequences In fact, by aligning the linear sequences and the hydrophobicity plots for all full-size ABC transporters present in the sequence database in 1997, we concluded that a common membrane topology and domain arrangement distinguishes a subfamily (MRP or ABCC subfamily) within the ABC kingdom (Tusnády et al., 1997) In the present review the amino acid sequences of 11 members of the ABCC subfamily (those with full cDNA sequences) were aligned, and the hydrophobicity plots of the aligned protein sequences were determined (Figure 2.3) This comparison indicates that seven proteins in the family (ABCC1–3, ABCC6, and ABCC8–10) form a subcluster within which each member possesses the N-terminal TMD0 domain, first described for ABCC1 Thus this subgroup is characterized by the TMD0-L0TMD1-ABC1-L-TMD2-ABC2 arrangement The N-terminal TMD0 domain is absent from ABCC4, ABCC5 and CFTR/ABCC7 Recent studies revealed that the TMD0 domain of ABCC1 does not play a crucial role either in the transport activity of the protein, or in the proper routing of the protein into the basolateral membrane compartment However, the presence of the L0 region (together with the TMD1-ABC1-L-TMD2-ABC2 core) is necessary for the ABCC1 GS-conjugate transport activity, and for the proper intracellular routing of the protein (Bakos et al., 1998) Thus, the ABCC1 L0 polypeptide was found to be membrane associated, and a 10 amino acid deletion within this region, encompassing a putative amphipathic helix, abolished the L0–membrane interaction and eliminated transport function, while not affecting membrane routing (Bakos et al., 2000) We have concluded from these studies that the L0 region forms a distinct structural and functional domain, which interacts with both the membrane and the core region of the transporter In harmony with the above conclusions, the cytoplasmic amino-terminal of CFTR/ABCC7 (which corresponds to the L0 domain) was found to have a major role in the control of CFTR channel gating, via physical interaction with the regulatory (R) domain (Naren et al., 1999) Of course, the actual sites of interactions still need to be explored in these different proteins ABCD SUBFAMILY Four half transporters with TMD-ABC arrangements, ABCD1/ALDP, ABCD2/ALDR, ABCD3/PMP70 and ABCD4/PMP70R, are the members of this family They are localized to the peroxisomal membrane and their mutant forms are involved in different inherited peroxisomal disorders It has been proposed that peroxisomal transporters need to dimerize in order to exert their function Co-immunoprecipitation experiments demonstrated the homodimerization of ABCD1, while a heterodimerization of ABCD1 with ABCD3 or ABCD2, and heterodimerization of ABCD2 with ABCD3 were found (Liu et al., 1999) The presence of six TM helices in the TMDs of the ABCD half transporters is generally predicted, but the experimental verification of this prediction has yet to be approached ABCG SUBFAMILY The members of this subfamily are also half transporters, with a unique domain arrangement of ABC-TMD, i.e the ABC domain located at the N-terminus Four proteins have been described with full sequences in this subgroup The best-characterized transporter is ABCG2 (MXR/BCRP), whose overexpression confers multidrug resistance Interestingly, this protein performs an active drug extrusion from the cells and is N-glycosylated in its mature form These data suggest that ABCG2 is localized in the plasma membrane, while several other half ABC transporters in the ABCB and ABCD subfamilies have been suggested to localize in membranes of various intracellular organelles, e.g the TAP proteins are located in the ER There is genetic evidence that ABCG5 and ABCG8 form heterodimers, as it was found that mutations of either of these genes cause the recessive genetic disease sitosterolemia (Berge et al., 2000; Lee et al., 2001) On the other hand, overexpression of the human ABCG2 multidrug resistance protein in a heterologous insect cell system generated an active, drugstimulated ATPase activity, strongly suggesting that this protein can act as a homodimer (Özvegy et al., 2001) No experimental data are available as yet on the exact membrane topology of the ABCG transporters However, here we combined some experimental data with empirical predictions to localize the transmembrane helices We used the MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C1/MRP1 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C3 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C2 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C6 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C8/SURI 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C9/SUR2 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 C10 100 200 C5 100 300 200 400 500 300 400 600 500 700 800 600 700 900 1000 1100 1200 1300 1400 800 900 1000 1100 1200 1300 1400 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 C4 C7/CFTR 1.0 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 1.0 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C11 1.0 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 100 200 300 100 200 400 300 400 500 600 500 600 700 700 800 1.0 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 800 900 1000 1100 1200 1300 900 1000 1100 1200 1300 1400 1.0 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 Ϫ0.8 Figure 2.3 Relative similarity dendrogram and the hydrophobicity plots of the aligned sequences of the ABCC subfamily transporters The shaded areas show transmembrane helices, predicted with locations as projected from experimental results based on ABCC1 (see text) HMMTOP transmembrane topology prediction server for this purpose in ABCG2, with the following restrictions: the N-terminal 290 amino acid chain, which harbors the ABC domain, should be intracellular, and – as the protein was shown to be N-glycosylated (Özvegy et al., 2001) – the predicted extracellular loops should contain consensus N-glycosylation site(s) The model predicted six TM helices with the following localization in the linear sequence of 43 44 ABC PROTEINS: FROM BACTERIA TO MAN 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 G5 100 200 300 400 500 600 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 G8 100 200 300 400 500 600 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 G2/BCRP 100 200 300 400 500 600 0.8 0.6 0.4 0.2 0.0 Ϫ0.2 Ϫ0.4 Ϫ0.6 G1 100 200 300 400 500 600 Figure 2.4 Relative similarity dendrogram and the hydrophobicity plots of the aligned sequences of the ABCG subfamily transporters The shaded areas show transmembrane helices, predicted with locations as projected from the HMMTOP server data and experimental results, based on ABCG2 (see text) ABCG2: 394–416, 427–449, 474–497, 506–530, 539–563 and 632–651 Two of the potential N-glycosylation sites (at positions 418 and at 596) were predicted within extracellular loops and 3, respectively Figure 2.4 shows the hydrophobicity plots of the aligned ABCG sequences, with the HMMTOP predicted TMDs superimposed This analysis suggests a similar location of the six TM helices in the TMDs for all ABCG proteins CONCLUSIONS Membrane topology models without experimental studies are just ‘educated hallucinations’ (to quote J Riordan), but even at this stage they have an important stimulatory role in searching for structure–function relationships As demonstrated in the case of CFTR or Pgp-MDR1, biochemical experiments may efficiently validate such models, while in other cases, e.g in the ABCA subfamily, experimentbased models still remain contradictory The various membrane topology predictions in the ABCC-MRP family have led to numerous experimental and theoretical studies, yielding important information regarding functional domains and those involved in membrane routing/ targeting A membrane topology is still difficult to define for any given ABC transporter, but a correct prediction for domain arrangements may provide a major help in devising useful antibodies, site-directed mutants and even specific functional modulators or inhibitors We can hardly wait for the detailed crystal-based structures of these large human membrane proteins, but until then there is still much fun to be had in working out more and more sophisticated and accurate models After completing this chapter, the structure of a bacterial ABC transporter, MsbA of Escherichia coli, determined by X-ray crystallography to a resolution of 4.5 Å (Chang and Roth, 2001, see Chapter 7) was published MsbA is a half transporter with a TMD-ABC domain arrangement, and the functional protein is a homodimer The published structure reveals that each MsbA subunit contains a TMD with six transmembrane helices, an ABC domain, and an ‘intracellular domain’ which is composed of three intracellular loops, connecting TMH2 to TMH3, TMH4 to TMH5, and TMH6 to the ABC domain A very important result of the published MsbA structure is the first demonstration that the membrane-spanning segments of this ABC transporter are indeed ␣-helices Helix organization and interactions similar to those in the MsbA protein most probably will be characteristic for many other ABC transporters REFERENCES Abele, R and Tampe, R (1999) Function of the transport complex TAP in cellular immune recognition Biochim Biophys Acta 461, 405–419 Allikmets, R., Singh, N., Sun, H., Shroyer, N.F., Hutchinson, A., Chidambaram, A., et al (1997) A photoreceptor cell-specific ATPbinding transporter gene (ABCR) is mutated in recessive Stargardt macular dystrophy Nat Genet 15, 236–246 Azarian, S.M and Travis, G.H (1997) The photoreceptor rim protein is an ABC transporter encoded by the gene for recessive Stargardt’s disease (ABCR) FEBS Lett 409, 247–252 Bakos, É., Hegedûs, T., Holló, Z., Welker, E., Tusnády, G.E., Zaman, G.J., Flens, M.J., Váradi, A and Sarkadi, B (1996) Membrane topology and glycosylation of the human MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS multidrug resistance-associated protein J Biol Chem 271, 12322–12326 Bakos, É., Evers, R., Szakács, G., Tusnády, G.E., Welker, E., Szabó, K., et al (1998) Functional multidrug resistance protein (MRP1) lacking the N-terminal transmembrane domain J Biol Chem 273, 32167–32173 Bakos, É., Evers, R., Calenda, G., Tusnády, G., Szakács, G., Váradi, A and Sarkadi, B (2000) Characterization of the amino-terminal regions in the human multidrug resistance protein (MRP1) J Cell Sci 113, 4451– 4461 Berge, K.E., Tian, H., Graf, G.A., Yu, L., Grishin, N.V., Schultz, J., Kwiterovich, P., Shan, B., Barnes, R and Hobbs, H.H (2000) Accumulation of dietary cholesterol in sitosterolemia caused by mutations in adjacent ABC transporters Science 290, 1771–1775 Bungert, S., Molday, L.L and Molday, R.S (2001) Membrane topology of the ATP binding cassette transporter ABCR and its relationship to ABC1 and related ABCA transporters Identification of N-linked glycosylation sites J Biol Chem 276, 23539–23546 Chang, G and Roth, C.B (2001) Structure of MsbA from E coli: a homolog of the multidrug resistance ATP binding cassette (ABC) transporters Science 293, 1793–1800 Chang, X.B., Hou, Y.X., Jensen, T.J and Riordan, J.R (1994) Mapping of cystic fibrosis transmembrane conductance regulator membrane topology by glycosylation site insertion J Biol Chem 269, 18572–18575 Chen, C.J., Chin, J.E., Ueda, K., Clark, D.P., Pastan, I., Gottesman, M.M and Roninson, I.B (1986) Internal duplication and homology with bacterial transport proteins in the mdr1 (P-glycoprotein) gene from multidrug-resistant human cells Cell 47, 381–389 Costet, P., Luo, Y., Wang, N and Tall, A.R (2000) Sterol-dependent transactivation of the ABC1 promoter by the liver X receptor/retinoid X receptor J Biol Chem 275, 28240–28245 Csere, P., Lill, R and Kispal, G (1998) Identification of a human mitochondrial ABC transporter, the functional orthologue of yeast Atm1p FEBS Lett 441, 266–270 Dean, M., Hamon, Y and Chimini (2001) The human ATP-binding cassette (ABC) transporter superfamily J Lipid Res 42, 1007–1017 Fitzgerald, M.L., Mendez, A.J., Moore, K.J., Andersson, L.P., Panjeton, H.A and Freeman, M.W (2001) ATP-binding cassette transporter A1 contains an NH2-terminal signal anchor sequence translocates the protein’s first hydrophilic domain to the exoplasmic space J Biol Chem 276, 15137–15145 Hipfner, D.R., Almquist, K.C., Leslie, E.M., Gerlach, J.H., Grant, C.E., Deeley, R.G and Cole, S.P (1997) Membrane topology of the multidrug resistance protein (MRP) J Biol Chem 272, 23623–23630 Hogue, D.L., Liu, L and Ling, V (1999) Identification and characterization of a mammalian mitochondrial ATP-binding cassette membrane protein J Mol Biol 285, 379–389 Illing, M., Molday, L.L and Molday, R.S (1997) The 220-kDa rim protein of retinal rod outer segments is a member of the ABC transporter superfamily J Biol Chem 272, 10303–10310 Kast, C and Gros, P (1997) Topology mapping of the amino-terminal half of multidrug resistance-associated protein by epitope insertion and immunofluorescence J Biol Chem 272, 26479–26487 Kast, C and Gros, P (1998) Epitope insertion favors a six transmembrane domain model for the carboxy-terminal portion of the multidrug resistance-associated protein Biochemistry 37, 2305–2313 Kast, C., Canfield, V., Levenson, R and Gros, P (1995) Membrane topology of P-glycoprotein as determined by epitope insertion Biochemistry 34, 4402–4411 Kast, C., Canfield, V., Levenson, R and Gros, P (1996) Transmembrane organization of mouse P-glycoprotein determined by epitope insertion and immunofluorescence J Biol Chem 271, 9240–9248 Lee, M.H., Lu, K., Hazard, S., Yu, H., Shulenin, S., Hidaka, H., et al (2001) Identification of a gene, ABCG5, important in the regulation of dietary cholesterol absorption Nat Genet 27, 79–83 Liu, L.X., Janvier, K., Berteaux-Lecellier, V., Cartier, N., Benarous, R and Aubourg, P (1999) Homo- and heterodimerization of peroxisomal ATP-binding cassette halftransporters J Biol Chem 274, 32738–32743 Luciani, M.F., Denizot, F., Savary, S., Mattei, M.G and Chimini, G (1994) Cloning of two novel ABC transporters mapping on human chromosome Genomics 21, 150–159 Naren, A.P., Cormet-Boyaka, E., Fu, J., Villain, M., Blalock, J.E., Quick, M.W and Kirk, K.L (1999) CFTR chloride channel regulation by an interdomain interaction Science 286, 544–548 45 46 ABC PROTEINS: FROM BACTERIA TO MAN Özvegy, C., Litman, T., Szakács, G., Nagy, Z., Bates, S., Váradi, A and Sarkadi, B (2001) Functional characterization of the human multidrug transporter, ABCG2, expressed in insect cells Biochem Biophys Res Commun 285, 111–117 Pullinger, C.R., Hakamata, H., Duchateau, P.N., Eng, C., Aouizerat, B.E., Cho, M.H., Fielding, C.J and Kane, J.P (2000) Analysis of hABC1 gene 5Ј end: additional peptide sequence, promoter region, and four polymorphisms Biochem Biophys Res Commun 271, 451–455 Santamarina-Fojo, S., Peterson, K., Knapper, C., Qiu, Y., Freeman, L., Cheng, J.F., et al (2000) Complete genomic sequence of the human ABCA1 gene: analysis of the human and mouse ATP-binding cassette A promoter Proc Natl Acad Sci USA 97, 7987–7992 Stride, B.D., Valdimarsson, G., Gerlach, J.H., Wilson, G.M., Cole, S.P and Deeley, R.G (1996) Structure and expression of the messenger RNA encoding the murine multidrug resistance protein, an ATP-binding cassette transporter Mol Pharmacol 49, 962–971 Sun, H., Smallwood, P.M and Nathans, J (2000) Biochemical defect in ABCR protein variants associated with human retinopathies Nat Genet 26, 242–246 Tanaka, A.R., Ikeda, Y., Abe-Dohmae, S., Arakawa, R., Sadanami, K., Kidera, A., et al (2001) Human ABCA1 contains a large amino-terminal extracellular domain homologous to an epitope of Sjogren’s Syndrome Biochem Biophys Res Commun 283, 1019–1025 Tusnády, G.E and Simon, I (1998) Principles governing amino acid composition of integral membrane proteins: application to topology prediction J Mol Biol 283, 489–506 Tusnády, G.E and Simon, I (2001) The HMMTOP transmembrane topology prediction server Bioinformatics 17, 849–850 Tusnády, G.E., Bakos, É., Váradi, A and Sarkadi, B (1997) Membrane topology distinguishes a subfamily of the ATP-binding cassette (ABC) transporters FEBS Lett 402, 1–3 von Heijne, G (1992) Membrane protein structure prediction Hydrophobicity analysis and the positive-inside rule J Mol Biol 225, 487–494  ... Sarkadi, B (1996) Membrane topology and glycosylation of the human MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS multidrug resistance-associated protein J Biol Chem 27 1, 123 22 123 26 Bakos,... localize the transmembrane helices We used the MEMBRANE TOPOLOGY OF THE HUMAN ABC TRANSPORTER PROTEINS 0.8 0.6 0.4 0 .2 0.0 Ϫ0 .2 Ϫ0.4 Ϫ0.6 Ϫ0.8 C1/MRP1 100 20 0 300 400 500 600 700 800 900 1000 1100 120 0... summary, the experimental data obtained for ABCA1 and ABCA4 support a similar membrane topology for the two proteins, with a domain arrangement of TMH1-ECD1TMH (2 6) -ABC1 -TMH7-ECD2-TMH(8– 12) ABC2 , where