* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download The Malaria Parasite`s Chloroquine Resistance Transporter is a
Silencer (genetics) wikipedia , lookup
Gene expression wikipedia , lookup
Paracrine signalling wikipedia , lookup
Genetic code wikipedia , lookup
Ancestral sequence reconstruction wikipedia , lookup
Expression vector wikipedia , lookup
Biochemistry wikipedia , lookup
Signal transduction wikipedia , lookup
Bimolecular fluorescence complementation wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
Metalloprotein wikipedia , lookup
Homology modeling wikipedia , lookup
Point mutation wikipedia , lookup
Interactome wikipedia , lookup
Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup
Protein purification wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein structure prediction wikipedia , lookup
Protein–protein interaction wikipedia , lookup
Two-hybrid screening wikipedia , lookup
The Malaria Parasite’s Chloroquine Resistance Transporter is a Member of the Drug/Metabolite Transporter Superfamily Rowena E. Martin and Kiaran Kirk School of Biochemistry and Molecular Biology, Faculty of Science, The Australian National University, Canberra, Australia The malaria parasite’s chloroquine resistance transporter (CRT) is an integral membrane protein localized to the parasite’s acidic digestive vacuole. The function of CRT is not known and the protein was originally described as a transporter simply because it possesses 10 transmembrane domains. In wild-type (chloroquine-sensitive) parasites, chloroquine accumulates to high concentrations within the digestive vacuole and it is through interactions in this compartment that it exerts its antimalarial effect. Mutations in CRT can cause a decreased intravacuolar concentration of chloroquine and thereby confer chloroquine resistance. However, the mechanism by which they do so is not understood. In this paper we present the results of a detailed bioinformatic analysis that reveals that CRT is a member of a previously undefined family of proteins, falling within the drug/metabolite transporter superfamily. Comparisons between CRT and other members of the superfamily provide insight into the possible role of the protein and into the significance of the mutations associated with the chloroquine resistance phenotype. The protein is predicted to function as a dimer and to be oriented with its termini in the parasite cytosol. The key chloroquine-resistance-conferring mutation (K76T) is localized in a region of the protein implicated in substrate selectivity. The mutation is predicted to alter the selectivity of the protein such that it is able to transport the cationic (protonated) form of chloroquine down its steep concentration gradient, out of the acidic vacuole, and therefore away from its site of action. Introduction The emergence and spread of malaria parasites that are resistant to the widely-used antimalarial drug chloroquine (CQ) has been a disaster for world health. CQ is a weak base that accumulates in the parasite’s digestive vacuole, a lysosomal compartment in which haemoglobin, taken up from the host cell cytosol via an endocytotic feeding mechanism, is degraded to its component peptides and haem. Within the vacuole CQ interferes with the mechanism by which the potentially toxic haem monomers are converted to the inert crystalline substance haemozoin, causing monomeric haem to accumulate to levels that kill the parasite. Two proteins—Pgh1 (P-glycoprotein homolog 1; Reed et al. 2000) and CRT (CQ resistance transporter; Fidock et al. 2000)—have been implicated as playing a role in CQ resistance. Both are integral membrane proteins, localized to the parasite’s digestive vacuole membrane. In Plasmodium falciparum, the most virulent of the human malaria parasites, mutations in the CRT protein (PfCRT) confer CQ resistance on otherwise sensitive parasite strains (Sidhu, Verdier-Pinard, and Fidock 2002). CQ-resistant parasites have a markedly reduced concentration of CQ in their digestive vacuole (Fitch 1970; Saliba, Folb, and Smith 1998); however, neither the mechanism by which PfCRT influences the intravacuolar concentration of the drug nor the normal physiological role of this protein are understood. In their original description of the protein, Fidock et al. (2000) described PfCRT, together with orthologs from other Plasmodium species and a more distant homolog from the slime mould Dictyostelium discoideum, as being putative channels or transporters containing 10 transmembrane domains (TMDs). Very recently, two preliminary reports Key words: malaria, Plasmodium falciparum, chloroquine, drug resistance, PfCRT, drug/metabolite transporter. E-mail: [email protected]. Mol. Biol. Evol. 21(10):1938–1949. 2004 doi:10.1093/molbev/msh205 Advance Access publication July 7, 2004 have assigned the PfCRT protein to the drug/metabolite transporter superfamily (Martin, Trueman, and Kirk 2003; Tran and Saier 2004). Here we present a detailed bioinformatic analysis of the protein and of the family and superfamily to which it belongs. Comparisons between PfCRT and members of the superfamily provide insight into the possible role of the protein and into the significance of the mutations associated with the CQ resistance phenotype. Materials and Methods Identifying Homologs of PfCRT The amino acid sequence of PfCRT was queried against the NCBI nonredundant protein database using a BlastP (Altschul et al. 1990) search in which the query sequence was not masked in areas of low compositional complexity. This option was chosen because the filter sometimes excludes from the analysis stretches of hydrophobic residues that correspond to the membrane-spanning regions of the protein and which are therefore of biological interest. Proteins retrieved by this search but excluded from the subsequent phylogenetic study included duplicate sequences and truncated proteins. Several proteins displayed areas of reasonable alignment with PfCRT (as appraised by eye), even though the Blast output suggested that the probability of homology was low (P . 1024). A detailed bioinformatic analysis was performed on such proteins; for each, the number and position of putative TMDs was used to assess whether the regions of sequence similarity corresponded to regions of alignment between the predicted secondary structures of the retrieved protein and PfCRT. Proteins that satisfied this criteria were then queried against the NCBI nonredundant protein database (using BlastP) and the Entrez Conserved Domain Database (using Reverse Position-Specific Blast; Marchler-Bauer et al. 2002) to determine their relationships to other proteins. PfCRT was also queried against the Entrez Conserved Domain Database as well as the Conserved Domain Architecture Retrieval Tool database (Geer et al. 2002) and was the subject of a Position-Specific Iterated Blast Molecular Biology and Evolution vol. 21 no. 10 Ó Society for Molecular Biology and Evolution 2004; all rights reserved. The Chloroquine Resistance Transporter 1939 Phylogenetic Analyses Regions of the alignment that could not be aligned unambiguously were excluded prior to analysis. A phylogenetic tree was estimated using the NeighborJoining method (Saitou and Nei 1987) and uncorrected (‘‘p’’) amino acid distances in MacVectorÔ 7.1. Ties in the tree were resolved randomly and a bootstrap analysis (Felsenstein 1985) was performed with 1,000 replicates. FIG. 1.—Neighbour-Joining tree of CRT family based on uncorrected distances. Numbers indicate the percentage of 1,000 bootstrap replicates which support the topology shown. The scale bar represents the number of substitutions per site for a unit branch length. (Altschul et al. 1997) search of the NCBI nonredundant protein database. Alignment of Hydropathy Profiles Hydropathy profile alignments were generated at http://bioinformatics.weizmann.ac.il/hydroph/ using KyteDoolittle (x 2 1) values, a window size of 17, and an algorithm that introduces gaps into the alignment to find the best match between two profiles. The final alignment of the hydropathy plots was compiled and edited in AdobeÒ PhotoshopÒ 6.0.1. Secondary Structure Predictions TMpred (www.ch.embnet.org/software/TMPRED_ form.html) and TMMHM (www.cbs.dtu.dk/services/ TMHMM-2.0/) were used to detect putative membranespanning domains in the sequences of the proteins of interest. Predictions of protein orientation were made on the basis of the ‘positive inside’ rule (von Heijne 1986; van Klompenburg et al. 1997) as well as by TMMHM (which incorporates the positive inside rule in its prediction of membrane protein topology [Sonnhammer, von Heijne, and Krogh 1998]). The predicted secondary structure of loop 7 of the CRT family proteins was obtained using the PredictProtein server (http://cubic. bioc.columbia.edu/predictprotein/). Construction of Alignments The ClustalW program (Thompson, Higgins, and Gibson 1994) in MacVectorÔ 7.1 was used to generate and edit alignments. Sequences for various DMT proteins have been described (e.g., Jack, Yang, and Saier 2001), and these were used to retrieve many DMT members from the NCBI database. Proteins of different DMT families have diverged considerably at the amino acid level and this can make a one-step alignment method error-prone. We therefore first aligned proteins within a family and then used the ClustalW profile-alignment tool to assemble the families into one large alignment. This alignment corresponded very well with the alignment of the predicted TMDs in these proteins. The first half of the DMT superfamily alignment was aligned to the second using the ClustalW profile alignment tool. Results and Discussion PfCRT Belongs to a Family of Proteins The proteins with the greatest similarity to PfCRT are homologs from other Plasmodia species (P. vivax, P. knowlesi, P. yoelii yoelii, P. chabaudi, and P. berghei) and these are retrieved by a BlastP search with P values in the range of 102162 to 102137. The P value is the probability that the sequence similarity shared by the query protein and the retrieved protein arose by chance. A P value ,1024 is considered to indicate a significant sequence similarity between the two proteins, consistent with their having a related biological function. Following the Plasmodia proteins, the next best hits against PfCRT include a protein from Cryptosporidium parvum (1 3 10225), the D. discoideum protein (4 3 10220), and several proteins from Arabidopsis thaliana (2 3 1025). A phylogenetic analysis performed on the sequence alignment of the Plasmodia, C. parvum, D. discoideum, and A. thaliana proteins (fig. 1) provides good evidence in support of the hypothesis that these proteins form a family. The CRT Proteins Are Related to Known Transporters The search for relatives of PfCRT in the NCBI database retrieved many proteins that had P values . 1024, indicating low, if any, homology to PfCRT. Nevertheless, several of these proteins showed a reasonable similarity in sequence to PfCRT over specific regions of the alignment. All such proteins are members of the same group of transport proteins, the drug/metabolite transporter (DMT) superfamily. When the D. discoideum, C. parvum, and A. thaliana proteins were queried against the NCBI database, they also retrieved many proven or putative transporters of the DMT superfamily. Three iterations of a Position-Specific Iterated Blast search of the NCBI database using PfCRT as the query sequence retrieved, with good significance, several characterized transporters of the DMT superfamily. Proteins of the same family typically share distinct modules or domains that have a common evolutionary origin and function. The PfCRT protein was queried against the Entrez Conserved Domain Database using Reverse Position-Specific Blast and was found to have weak, but significant, hits to DMT superfamily conserved domains. These observations are consistent with the hypothesis that the proteins of the CRT family form part of, or are related to, the DMT superfamily. Members of the CRT and DMT Families Have Similar Hydropathy Plots A simple and commonly used predictor of the structure of a membrane protein is the hydropathy plot, 1940 Martin and Kirk FIG. 2.—An alignment of the hydropathy plots for the P. falciparum (yellow), D. discoideum (red) and A. thaliana (green) members of the CRT family and for the E. coli YdeD amino acid effluxer (blue) of the DMT superfamily. The positions of the predicted 10 TMDs are indicated by white bars. Gaps in the alignment are indicated below the plot by appropriately colored lines. in which putative TMDs and the connecting hydrophilic, extra-membrane loops are detected as ‘peaks’ and ‘troughs,’ respectively, in a plot of the hydrophobicity index of the polypeptide. Transporters of the same family usually share a very similar hydropathy plot, even when the underlying polypeptide sequences have diverged considerably (Lolkema and Slotboom 1998). Members of the DMT superfamily, like those of the CRT family, are predicted to contain 10 TMDs. Figure 2 shows an alignment of the hydropathy plots for PfCRT (yellow line), the D. discoideum protein (red line), one of the A. thaliana proteins (green line), and a protein from the DMT superfamily (the E. coli YdeD amino acid effluxer; blue line), consistent with these proteins having similar structure. In pairwise sequence alignments between members of the CRT family and the DMT protein, the regions of sequence similarity were found to correspond to the regions of alignment of predicted TMDs in the two proteins. et al. 2003). The NST and TPT families clustered together on a branch separate from the DME proteins and the DME family was comprised of many nodes, one of the outermost being the plant DME proteins. The RhaT, GRP, and CEO families clustered together and are distantly related to the DME proteins. The CRT family placed within the DMT superfamily, in which it branched from the DME proteins, after the RhaT, GRP, and CEO cluster had diverged. Bootstrap analysis was performed on a reduced data set of 53 proteins that included representatives from the major groups within the DMT superfamily. As shown in figure 3, the bootstrap values support the placement of the CRT proteins within the DMT superfamily, where they cluster as a distinct family that branches between the DME and NST families. The analysis was repeated on another two subsets of ;50 proteins drawn from the full data set of 368 proteins and both yielded results similar to the tree presented in figure 3. The CRT Family Falls Within the DMT Superfamily The CRT Proteins Arose from a Gene Duplication Event The DMT superfamily consists of a number of discrete membrane protein families including the drug/ metabolite effluxer (DME) family, the L-Rhamnose symporter (RhaT) family, the glucose/ribose permease (GRP) family, the C. elegans ORF (CEO) family (no members of which have been characterized as yet), the nucleotide-sugar transporter (NST) families, the triosephosphate transporter (TPT) family and the plant organocation permease (POP) family (Jack, Yang, and Saier 2001). An alignment of 368 proteins belonging to the DMT superfamily and CRT family was constructed from 10 CRT, 206 DME, 6 RhaT, 12 GRP, 7 CEO, 76 NST, 41 TPT, and 10 POP proteins. Four transporters from each of three unrelated 10-TMD transporter families—the Ca21:cation antiporters (CaCA), the C4-dicarboxylate importers (Dcu), and the glutamate:Na1 symporters (ESS)—were included as outgroups. The ‘best tree’ was estimated using the NeighborJoining method and the relationships within the superfamily were found to be similar to those described previously (Jack, Yang, and Saier 2001; Ward 2001; Knappe, Flugge, and Fischer 2003; Livshits et al. 2003; Martinez-Duncker Within members of the DMT superfamily there is significant homology between the first half of the protein (encompassing TMDs 1–5) and the second half (encompassing TMDs 6–10), consistent with these transporters having arisen from an ancestral gene that underwent an internal duplication event (Jack, Yang, and Saier 2001). Figure 4 shows alignments of the two halves of a representative selection of DMT proteins together with three members of the CRT family (including PfCRT). The first TMD is aligned with the sixth TMD, the second TMD with the seventh, and so on. As well as there being significant similarities along the whole lengths of the different proteins, there are similarities between the two halves of the proteins, including in the CRT proteins. The Membrane Orientation of the CRT Proteins As is evident from figure 4, there is a general trend within the DMT superfamily (including the CRT proteins) for the even-numbered extra-membrane loops to have a greater proportion of positively charged residues than the odd-numbered loops. The N- and C-termini also have The Chloroquine Resistance Transporter 1941 FIG. 3.—Neighbour-Joining tree of the DMT superfamily based on uncorrected distances. The analysis included three members of the CRT family and 38 DMT sequences. Four of the DMT proteins are characterized transporters of the nucleotide-sugar transporter (NST) family and 34 are known or putative members of the drug/metabolite efflux (DME) family. Included for comparison were four sequences from each of three unrelated 10 TMD transporter families: the Ca21:cation antiporters (CaCA), the C4-dicarboxylate importers (Dcu), and the Glutamate:Na1 symporters (ESS). Numbers indicate the percentage of 1,000 bootstrap replicates which support the topology shown. The scale bar represents the number of substitutions per site for a unit branch length. ! FIG. 4.—The alignment of the first half (encompassing TMDs 1–5) with the second half (encompassing TMDs 6–10) of a representative selection of DMT proteins, including three CRT family members. Residues are shaded as follows: positively charged, blue; negatively charged, red; tryptophan and tyrosine, white; polar, proline, and glycine, grey; remaining nonpolar, yellow. The conserved proline in TMDs 4 and 9 is highlighted in green. The variant N- and C-termini have been omitted. In some proteins loop 2 and/or 7 has been truncated and this is indicated by a solid black line. ‘(p)’ indicates that the assignment of function is only putative. The proteins are as follows: 1, M. musculus gi 2499227 (Eckhardt et al. 1996); 2, D. melanogster gi 21355345 (Martinez-Duncker et al. 2003); 3, K. lactis gi 6016590 (Abeijon, Robbins, and Hirschberg 1996); 4, C. elegans gi 20140026 (Berninsone et al. 2001); 5, C. albicans gi 14971021 (Nishikawa et al. 2002); 6, H. sapiens gi 14009667 (Lubke et al. 2001; Luhn et al. 2001); 7, Z. mays gi 1352200 (Fischer et al. 1994); 8, A. thaliana gi 15219121; 9, E. coli gi 26107188 (Livshits et al. 2003); 10, E. coli gi 25367911 (Santiviago et al. 2002); 11, B. subtilis gi 1175623 (Jack, Yang, and Saier 2001); 12, P. agglomerans gi 4098977; 13, E. coli gi 26250715 (Jack, Yang, and Saier 2001); 14, F. nucleatum gi 19705391; 15, M. loti gi 13473318; 16, M. rubra gi 2062658 (Berg, Hilbi, and Dimroth 1997); 17, E. coli gi 13361606 (Dassler et al. 2000); 18, T. erythraeum gi 23043464; 19, B. fungorum gi 22988226; 20, P. marinus gi 33862898; 21, T. fusca gi 23018393; 22, B. subtilis gi 16081133 (Jack, Yang, and Saier 2001); 23, C. aurantiacus gi 22972802; 24, A. thaliana gi 25408555; 25, A. thaliana gi 18415262; 26, A. thaliana gi 21536591; 27, P. falciparum gi 23612473 (Fidock et al. 2000); 28, D. discoideum gi 11139714; 29, S. xylosus gi 2226001 (Fiegler et al. 1999); 30, C. elegans gi 13384556. 1942 Martin and Kirk The Chloroquine Resistance Transporter 1943 FIG. 5.—Alignment of CRT proteins over TMD 1 and loop 7. Residues are shaded as follows: positively charged, black; negatively charged, dark grey; tryptophan and tyrosine, white; polar, proline and glycine, mid-grey; remaining nonpolar, light grey. The position of the K76T mutation in TMD 1 is indicated (star) and CQ-resistant parasite strains (numbered 5 and 6) are highlighted. a preponderance of positive charge. This makes it likely that the even-numbered loops, together with the N- and Ctermini, are located at the cytoplasmic face of the membrane (predicted by the positive inside rule [von Heijne 1986; van Klompenburg et al. 1997] and by TMHMM [Sonnhammer, von Heijne, and Krogh 1998]). Such an orientation (i.e., termini in the cytosol) has been proven experimentally for two DMT proteins, the mouse CMP-sialic acid transporter (NST family; Eckhardt, Gotza, and Gerardy-Schahn 1999) and the PecM protein of Erwinia chrysanthemi (DME family Rouanet and Nasser 2001). For the RhaT, GRP, and CEO proteins it is the oddnumbered loops that have the higher proportion of positive charge, suggestive of the opposite topology, and this has been demonstrated experimentally for the S. typhimurium RhaT protein (for which the termini were found to be noncytosolic [Tate and Henderson 1993]). 7 is a relatively long hydrophilic domain, changes in which have been shown to influence transporter activity. The activity of the mouse CMP-sialic acid transporter is reduced when an epitope is inserted into loop 7 (but not loop 2), and increasing the length of the tag causes a further reduction in transporter activity (Eckhardt, Gotza, and Gerardy-Schahn 1999). As illustrated in figure 5, loop 7 is especially long and well conserved in composition in the proteins of the CRT family, and the predicted structure of this region (given by the PredictProtein sever) was similar for each protein: a ‘compact globular domain’ that is formed by a nine amino acid alpha helix followed by two short beta sheets (although in some proteins the first beta sheet may instead be an alpha-helix). The predicted orientation of PfCRT in the membrane (i.e., with the protein termini in the cytosol) places loop 7 in the digestive vacuole, where it may have a role in modulating the activity of the transporter. The Significance of the Extra-Membrane Loops Having established that PfCRT is a member of the DMT superfamily it is possible to assign putative functions to different regions of the PfCRT protein on the basis of previous studies of other members of the superfamily (fig. 4). A striking feature of figure 4 is the conservation of structural elements throughout the superfamily, despite the considerable divergence in sequence, mode of transport, and substrate specificity. For instance, the length and composition of the loops are conserved between proteins from different families. Furthermore, the loop regions in the second half of the transporter bear significant similarity to the corresponding loops in the first half. Loops 3 and 8 and 4 and 9, are particularly well conserved. Studies with proteins from both the NST and DME families have revealed that the insertion of an epitope tag or reporter molecule in loops 3, 4, 8, and 9 can inactivate the transporter and in some instances cause the protein to be localized incorrectly within the cell and/or degraded (Tate and Henderson 1993; Eckhardt, Gotza, and GerardySchahn 1999; Rouanet and Nasser 2001). By contrast, the same sequences introduced into loops 1, 2, 5, and 6 have no effect on transporter activity or localization. Compared to the other loops, 2 and 7 show less sequence conservation and show significant variation in length between DMT proteins. In most DMT proteins loop The Presence of Helix Packing Motifs in Transmembrane Domains 5 and 10 Indicates that PfCRT May Function as a Dimer While the structure of the DMT transporters has been retained over time and evolution, the underlying amino acid sequence has proven more plastic. Nevertheless, some regions of sequence have been strongly conserved. In TMDs 5 and 10 there are two conserved glycines that are separated by six hydrophobic residues (fig. 4; Knappe, Flugge, and Fischer 2003). This motif (GxxxxxxG) is a common feature of membrane proteins, in which it is thought to facilitate the packing of membrane-spanning helices, leading to the association of TMDs to form oligomers (Liu, Engelman, and Gerstein 2002). Other small residues (alanine, serine, and threonine) can replace one of the glycines in a glycine-packing motif (Russ and Engelman 2000; Eilers et al. 2002) and such a substitution has occurred in a number of the DMT proteins, including PfCRT (fig. 4). TMDs 5 and 10 are known to have a role in mediating the formation of homo-dimers by the NST and TPT transporters (Abe, Hashimoto, and Yoda 1999; Ishida et al. 1999; Streatfield et al. 1999; Gao and Dean 2000) and nonconservative mutations within the putative packing motifs in TMDs 5 and 10 abrogate transporter activity as well as stability (Ishida et al. 1999; Streatfield et al. 1999). Although it is not known whether members of the DME 1944 Martin and Kirk family function as dimers, the conservation of the GxxxxxxG motif in DME (as well as CRT proteins) are consistent with their doing so. Different Transmembrane Domains Are Implicated in Substrate Recognition, Binding, and Translocation Many studies with NST or TPT proteins have shown that substitutions at conserved positions in TMDs 3, 4, 8, and 9 interfere with the binding and translocation of the substrate (a divalent anion). The mutations either impair the rate of transport or abolish transport altogether, while not affecting either the localization of the protein within the cell or its ability to form homo-dimers (Abe, Hashimoto, and Yoda 1999; Berninsone et al. 2001; Gao, Nishikawa, and Dean 2001; Lubke et al. 2001; Luhn et al. 2001; Oelmann, Stanley, and Gerardy-Schahn 2001; Etzioni et al. 2002; Knappe, Flugge, and Fischer 2003). A large proportion of these functionally important residues are located in TMDs 4 or 9; in both cases they are concentrated in the center of the helix where they form a strongly conserved region of polar and positively charged residues that is thought to constitute a substrate binding motif (Fischer et al. 1994; Gao, Nishikawa, and Dean 2001; Lubke et al. 2001). The binding motif regions in TMDs 4 and 9 are also well conserved in DME and CRT proteins, but they do not contain positively charged residues. Instead, there is a conserved proline, usually in both TMDs 4 and 9, although a few proteins (including the Arabidopsis CRT protein) have a proline in only one of the binding motifs (fig. 4). The role of these conserved prolines in the function of DME transporters has not been studied, but proline residues located in the central part of a transmembrane helix are known to be essential for the biological activity of a number of membrane transport proteins (Webb, Rosenberg, and Cox 1992; Lin, Itokawa, and Uhl 2000; Shelden et al. 2001; Koike et al. 2004). The proline ring distorts the normal structure of a membrane-spanning helix via two mechanisms: a kink is introduced in the helix backbone to avoid a steric clash with the proline ring, and the hydrogen bonds that would normally stabilize this region of the helix are unable to form (Woolfson and Williams 1990; Visiers, Braunheim, and Weinstein 2000; Cordes, Bright, and Sansom 2002). This results in a flexible ‘hinge’ point in the helix that has the capacity to form hydrogen bonds (perhaps with a substrate). The presence of a proline hinge in the putative binding motif of DME and CRT proteins suggests a role for this residue in the binding and translocation of substrates. The codons encoding both proline and glycine are GC-rich and, as such, these codons are the least likely to be retained by chance in the AT-rich genomes of Plasmodium and Dictyostelium (Stevens and Arkin 2000). This is consistent with there being a critical role for the putative helix-packing glycine motifs and the intraTMD proline residues in the function of CRT proteins. While discrete regions of the DMT proteins are implicated in the binding and translocation of the substrate, other elements of the transporters have been implicated in the recognition of, and discrimination between, substrates. The participation of residues from a number of TMDs in substrate recognition is not uncommon among transporters. For example, in members of the major-facilitator superfamily, eight out of the 12 TMDs are predicted to be involved in determining the substrate specificity of the transporter (Hirai et al. 2003). The functional analyses of chimeras constructed from two human NSTs, the UDPgalactose transporter, and the CMP-sialic acid transporter, revealed the participation of TMDs 1, 2, 3, 7, and 8 in determining substrate specificity (Aoki, Ishida, and Kawakita 2001, 2003). The involvement of TMDs 3 and 8 is not surprising, as these domains are also thought to influence the binding and translocation of the substrate; residues in these helices are likely to face the translocation pore where they may interact with the substrate. Amino acid residues are considered to be potentially ‘substrate-specific’ when they are conserved among transporters with identical substrates, but are different between transporters of differing substrate specificity. In DMT proteins there are many such residues present in TMDs 2 and 7 (fig. 4). Indeed, TMD 7 is one of the helices that varies the most in sequence between DMT proteins of different subgroups. Consistent with this observation, the substrate specificity of the UDP-galactose transporter is broadened to include CMP-sialic acid simply by replacing TMD 7 with the corresponding helix from the CMP-sialic acid transporter. TMD 1 is also crucial for determining the substrate specificity of hybrids between the UDP-galactose and CMP-sialic acid transporters, and it is thought to have a similar involvement in the substrate specificity of the UDP-galactose transporter from fruit fly (Segawa, Kawakita, and Ishida 2002). Similarly, in TPT proteins, residues in TMD 1 are proposed to line the translocation pore of the transporter (Knappe, Flugge, and Fischer 2003). TMDs 7, 8, 9, and 10 all fulfill the same role in transporter function as the corresponding domains in the first half of the protein (fig. 6). It might therefore be expected that TMD 6, as the counterpart of TMD 1 in the second half of the protein, should also play a role in determining substrate specificity. Experimental support for this comes from a study with the hamster CMP-sialic acid transporter (Eckhardt, Gotza, and Gerardy-Schahn 1998). Mutation of a glycine residue in TMD 6 to glutamate, glutamine, or isoleucine severely reduces transporter activity without affecting the expression or trafficking of the protein. Overexpression of the mutant proteins restores a low level of CMP-sialic acid transport, consistent with the mutations having affected the affinity of the transporter for its substrate. The Chloroquine Resistance-Conferring K76 Mutation Lies in a Region Implicated in Substrate Selectivity The substitution of the lysine at position 76 for threonine (K76T) has been identified as a crucial determinant of PfCRT-mediated CQ resistance (Sidhu, Verdier-Pinard, and Fidock 2002). As depicted in figure 6, this mutation lies towards the C-terminal end of TMD 1, a region of the transporter that we predict to be involved in substrate recognition. Field isolates from both Old and New World strains of CQ-resistant parasites all have the critical The Chloroquine Resistance Transporter 1945 FIG. 6.—Charge distribution, topology, and putative roles of the TMDs of PfCRT. Positively charged residues (lysine or arginine) predicted to lie in the extra-membrane segments or in the termini regions are shown in black. The putative TMDs are numbered 1–10. The position of the CQresistance-associated K76T (TMD 1) and A220S (TMD 6) mutations (black), the conserved prolines (TMDs 4 and 9; dark grey) and the helix packing motifs (GxxxxxxT; TMDs 5 and 10; mid-grey) are highlighted. TMDs 4 and 9 (boxed in dark grey) are implicated in the binding and translocation of substrates by DMT proteins. The conserved helix packing motifs in TMDs 5 and 10 (boxed in mid-grey) play a role in the formation of homo-dimers by TPT and NST proteins. Certain residues in TMDs 3 and 8 (boxed in light grey) in DMT transporters assist in the binding and translocation of the substrate and both domains also influence the substrate specificity of the transporter. Several elements of the DMT transporter seem to be involved in recognizing and discriminating between substrates. Along with TMDs 3 and 8, TMDs 1, 2, 6, and 7 (boxed in black) also participate in determining substrate specificity. The topology cartoon was generated using the TOPO2 server (www.sacs.ucsf.edu/TOPO/topo.html). K76T mutation, accompanied by a number of what are thought to be ‘compensatory’ PfCRT mutations that enable the transporter to maintain a semblance of its normal physiological role (Carlton et al. 2001). The A. thaliana CRT proteins, like the PfCRT proteins of CQ-sensitive strains, have a positively charged residue in this position (fig. 5). By contrast, the D. discoideum and C. parvum proteins are more similar to the PfCRT proteins from resistant strains of P. falciparum in that they have a serine (serine and threonine are hydroxy amino acids and a S ! T mutation is considered to be a conservative substitution). A CQ-sensitive isolate from Sudan (106/1) contains all of the PfCRT mutations found in CQ-resistant strains from the Old World, with the exception of the K76T mutation which has reverted to the wild-type form (Fidock et al. 2000). When this strain was subjected to CQ-selection pressure in vitro, two CQ-resistant clones arose, both with novel mutations in the K76 position (K76I and K76N; Fidock et al. 2000; Cooper et al. 2002). Characterization of these mutants revealed that the K76I mutation had the unusual effect of increasing the parasite’s sensitivity to quinine while decreasing its sensitivity to the diastereomer quinidine. This adds further support to the hypothesis that TMD 1 of PfCRT, and in particular position 76, is involved in determining the substrate specificity of the transporter. The absence of the K76I and K76N mutations in CQ-resistant field populations (and the prevalence of the K76T mutation) may reflect a less fit phenotype in vivo for these artificially-derived mutants (Cooper et al. 2002). This is consistent with the observation that only two types of amino acid residues are found in this position among other proteins of the CRT family: positively charged or hydroxy. Warhurst and colleagues (2002) have suggested previously that PfCRT is related to proteins of the chloride channel (ClC) family, members of which have been determined to possess 18 alpha helices (Dutzler et al. 2002). In their analysis, the region of PfCRT containing the K76T mutation was predicted to correspond to helix C of the Salmonella typhimurium ClC protein and the loss of the positive charge in this position was postulated to alter the specificity of the parasite chloride channel, such that it permitted the transport of chloroquine. However, residues in helix C of the S. typhimurium ClC protein are not known to influence substrate specificity, nor do they line either the channel’s selectivity filter or translocation pore (Dutzler et al. 2002). The mechanism by which the K76T mutation would influence the selectivity of a channel of this type, and thereby confer chloroquine resistance, is therefore unclear. Apart from the K76T mutation, there is another PfCRT mutation (A220S) conserved in most CQ-resistant isolates and absent from CQ-sensitive strains (with the exception of the 106/1 strain). The A220S mutation is located in TMD 6 and does not confer resistance in the absence of the K76T mutation (e.g., 106/1). However, its presence in most CQ-resistant strains analyzed to date suggests that it acts in synergy with K76T, perhaps by 1946 Martin and Kirk FIG. 7.—Model for the mechanism of PfCRT-mediated resistance to CQ. The protein is shown as a dimer, functioning to export ‘metabolites’ (perhaps amino acids or peptides, and perhaps in symport with H1) from the parasite’s digestive vacuole (DV). The ‘positive-inside’ rule, and a presumed inwardly-positive electrical potential across the DV membrane, predicts the N- and C-termini to be cytosolic and the ‘compact globular domain’ of loop 7 to be located at the vacuolar face of the membrane. CQ is a diprotic weak base and therefore accumulates in the acidic DV in the protonated (positively charged, CQ21) form. In parasites expressing ‘wild type’ PfCRT, the positive charge of K76 prevents the interaction of CQ21 with the transporter. The CQ resistance–conferring K to T mutation removes the positive charge and alters the substrate selectivity, allowing CQ21 to interact with the transporter and to be effluxed from the DV, perhaps in symport with H1. This results in a reduction of the concentration of CQ within the vacuole. aiding the recognition of CQ as a substrate for PfCRT or as a compensatory mutation that stabilizes the interaction of the transporter with its physiological substrate(s). The location of the A220 mutation in a PfCRT domain predicted to participate in substrate recognition is consistent with both of these scenarios. Substrates Effluxed by DME Transporters The members of the DMT superfamily bearing the closest similarity to the CRT proteins in the region of the substrate binding motif fall within the DME transporter subfamily. Substrates for DME transporters include amino acids, weak bases, and organic cations. The YdeD protein of E. coli exports cysteine metabolites (Dassler et al. 2000), whereas the E. coli YbiF protein exports a broad range of amino acids including homoserine, threonine, lysine, and histidine (Livshits et al. 2003). In other species of bacteria, DME proteins are implicated in the efflux of methylamine (MttP; Ferguson and Krzycki 1997), the di-cationic herbicide methyl viologen (YddG; Santiviago et al. 2002) and the pigment indigoidine, which is, like chloroquine, a weak base (PecM; Rouanet and Nasser 2001). The fact that DME transporters are known to transport both weak bases and divalent organic cations lends support to the hypothesis that the CQ-resistant form of PfCRT transports the chloroquine in the di-cationic form. DME systems are postulated to be H1-coupled and this has been confirmed experimentally for at least one DME transporter (the E. coli YbiF protein) [Livshits et al. 1993]. The Role of PfCRT in Chloroquine Resistance Figure 7 shows a model for the mechanism of PfCRT-mediated CQ-resistance, based on the insights gained from the bioinformatic analysis presented here, and consistent with previous reports of enhanced CQ efflux from CQ-resistant parasites (Krogstad et al. 1987; Sanchez, Stein, and Lanzer 2003). The protein is shown as a dimer, functioning to export ‘metabolites’ from the parasite’s digestive vacuole. The facts that (1) a number of related DME proteins transport amino acids, and (2) that the only known metabolite transport function of the digestive vacuole is the efflux of peptides (Kolakovich et al. 1997) and perhaps amino acids, prompt the hypothesis that PfCRT is an amino acid/peptide transporter (perhaps H1-coupled), but this remains to be tested. The protein is predicted, on the basis of the ‘positiveinside’ rule as well as by TMHMM, to be oriented with the N- and C-termini in the parasite cytosol and the ‘compact globular domain’ of loop 7 located at the vacuolar face of the membrane. In this orientation, the crucial CQRconferring K76T mutation lies close to the surface of the vacuolar face of the membrane, within a region of the protein postulated to be involved in substrate selectivity. The positive charge is predicted to play a key role in repulsing the protonated (cationic) form of chloroquine (CQ21, the predominant species present within the acidic vacuole) and preventing it from interacting with the transporter. The CQR-conferring mutation of Lys to Thr (or Ile or Asn) removes the positive charge, allowing CQ21 to interact with and be transported by the protein, down the steep outward CQ21 concentration gradient (again perhaps in symport with H1). On exiting the vacuole and entering the parasite cytosol (pH 7.3; Saliba and Kirk 1999), the CQ is deprotonated and diffuses out of the parasite as the neutral (membrane-permeant) species. The net result is a decrease in the overall concentration of CQ at its site of action within the digestive vacuole, and hence a decreased CQ-sensitivity of the parasite. Experimental testing of the key aspects of the model presented in figure 7 await the expression of PfCRT in a heterologous system in a form in which the transport properties of the protein can be investigated directly. PfCRT has been successfully expressed in yeast (Zhang, Howard, and Roepe 2002); however, there has not, as yet, been any direct demonstration of its transport function. Efforts are The Chloroquine Resistance Transporter 1947 presently underway to express the protein in Xenopus laevis oocytes and to measure and compare the transport of radiolabeled chloroquine via PfCRT with and without the K76T mutation. It is predicted that oocytes expressing PfCRT from CQ-resistant strains (i.e., having the K76T mutation) in their plasma membrane will transport [3H]CQ, whereas those expressing wild-type PfCRT from CQsensitive strains (having K76) will not. The successful expression of PfCRT in Xenopus oocytes will also allow: (1) a direct test of the hypothesis that PfCRT proteins from both CQ-resistant and CQ-sensitive strains transport amino acids/peptides; (2) screening of other classes of substrate for their ability to be transported via PfCRT; (3) the determination of whether PfCRT-mediated transport is H1 coupled; and (4) an investigation of whether, as has been proposed (Warhurst 2003), the chloroquine resistance reversal agent verapamil interacts directly with PfCRT (in TMD 1) to inhibit the transport of CQ. Such experiments have the potential to yield important insights into the molecular mechanism underlying chloroquine resistance. Acknowledgments The authors are grateful to Stephen Allen and Jan Martin for technical assistance and to Rhys Hayward, Kevin Saliba, and John Trueman for helpful discussions. This work was supported by the Australian National Health and Medical Research Council (grant ID 179804). Literature Cited Abe, M., H. Hashimoto, and K. Yoda. 1999. Molecular characterization of Vig4/Vrg4 GDP-mannose transporter of the yeast Saccharomyces cerevisiae. FEBS Lett. 458:309–312. Abeijon, C., P. W. Robbins, and C. B. Hirschberg. 1996. Molecular cloning of the Golgi apparatus uridine diphosphateN-acetylglucosamine transporter from Kluyveromyces lactis. Proc. Natl. Acad. Sci. USA 93:5963–5968. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403–410. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped Blast and PSI-Blast: a new generation of protein database search programs. Nucleic Acids Res. 25:3389–3402. Aoki, K., N. Ishida, and M. Kawakita. 2001. Substrate recognition by UDP-galactose and CMP-sialic acid transporters. Different sets of transmembrane helices are utilized for the specific recognition of UDP-galactose and CMP-sialic acid. J. Biol. Chem. 276:21555–21561. ———. 2003. Substrate recognition by nucleotide sugar transporters: further characterization of substrate recognition regions by analyses of UDP-galactose/CMP-sialic acid transporter chimeras and biochemical analysis of the substrate specificity of parental and chimeric transporters. J. Biol. Chem. 278:22887–22893. Berg, M., H. Hilbi, and P. Dimroth. 1997. Sequence of a gene cluster from Malonomonas rubra encoding components of the malonate decarboxylase Na1 pump and evidence for their function. Eur. J. Biochem. 245:103–115. Berninsone, P., H. Y. Hwang, I. Zemtseva, H. R. Horvitz, and C. B. Hirschberg. 2001. SQV-7, a protein involved in Caenorhabditis elegans epithelial invagination and early embryogenesis, transports UDP-glucuronic acid, UDP-N- acetylgalactosamine, and UDP-galactose. Proc. Natl. Acad. Sci. USA 98:3738–3743. Carlton, J. M., D. A. Fidock, A. Djimde, C. V. Plowe, and T. E. Wellems. 2001. Conservation of a novel vacuolar transporter in Plasmodium species and its central role in chloroquine resistance of P. falciparum. Curr. Opin. Microbiol. 4:415–420. Cooper, R. A., M. T. Ferdig, X. Z. Su, L. M. Ursos, J. Mu, T. Nomura, H. Fujioka, D. A. Fidock, P. D. Roepe, and T. E. Wellems. 2002. Alternative mutations at position 76 of the vacuolar transmembrane protein PfCRT are associated with chloroquine resistance and unique stereospecific quinine and quinidine responses in Plasmodium falciparum. Mol. Pharmacol. 61:35–42. Cordes, F. S., J. N. Bright, and M. S. Sansom. 2002. Prolineinduced distortions of transmembrane helices. J. Mol. Biol. 323:951–960. Dassler, T., T. Maier, C. Winterhalter, and A. Bock. 2000. Identification of a major facilitator protein from Escherichia coli involved in efflux of metabolites of the cysteine pathway. Mol. Microbiol. 36:1101–1112. Dutzler, R., E. B. Campbell, M. Cadene, B. T. Chait, and R. MacKinnon. 2002. X-ray structure of a ClC chloride channel at 3.0 Å reveals the molecular basis of anion selectivity. Nature 415:287–294. Eckhardt, M., B. Gotza, and R. Gerardy-Schahn. 1998. Mutants of the CMP-sialic acid transporter causing the Lec2 phenotype. J. Biol. Chem. 273:20189–20195. ———. 1999. Membrane topology of the mammalian CMPsialic acid transporter. J. Biol. Chem. 274:8779–8787. Eckhardt, M., M. Muhlenhoff, A. Bethe, and R. Gerardy-Schahn. 1996. Expression cloning of the Golgi CMP-sialic acid transporter. Proc. Natl. Acad. Sci. USA 93:7572–7576. Eilers, M., A. B. Patel, W. Liu, and S. O. Smith. 2002. Comparison of helix interactions in membrane and soluble alphabundle proteins. Biophys. J. 82:2720–2736. Etzioni, A., L. Sturla, A. Antonellis, E. D. Green, R. GershoniBaruch, P. M. Berninsone, C. B. Hirschberg, and M. Tonetti. 2002. Leukocyte adhesion deficiency (LAD) type II/carbohydrate deficient glycoprotein (CDG) IIc founder effect and genotype/phenotype correlation. Am. J. Med. Genet. 110: 131–135. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783–791. Ferguson, D. J. Jr., and J. A. Krzycki. 1997. Reconstitution of trimethylamine-dependent coenzyme M methylation with the trimethylamine corrinoid protein and the isozymes of methyltransferase II from Methanosarcina barkeri. J. Bacteriol. 179:846–852. Fidock, D. A., T. Nomura, A. K. Talley et al. (14 co-authors). 2000. Mutations in the P. falciparum digestive vacuole transmembrane protein PfCRT and evidence for their role in chloroquine resistance. Mol. Cell 6:861–871. Fiegler, H., J. Bassias, I. Jankovic, and R. Bruckner. 1999. Identification of a gene in Staphylococcus xylosus encoding a novel glucose uptake protein. J. Bacteriol. 181:4929– 4936. Fischer, K., B. Arbinger, B. Kammerer, C. Busch, S. Brink, H. Wallmeier, N. Sauer, C. Eckerskorn, and U. I. Flugge. 1994. Cloning and in vivo expression of functional triose phosphate/ phosphate translocators from C3- and C4-plants: evidence for the putative participation of specific amino acid residues in the recognition of phosphoenolpyruvate. Plant J. 5:215–226. Fitch, C. D. 1970. Plasmodium falciparum in owl monkeys: drug resistance and chloroquine binding capacity. Science 169: 289–290. Gao, X. D., and N. Dean. 2000. Distinct protein domains of the yeast Golgi GDP-mannose transporter mediate oligomer 1948 Martin and Kirk assembly and export from the endoplasmic reticulum. J. Biol. Chem. 275:17718–17727. Gao, X. D., A. Nishikawa, and N. Dean. 2001. Identification of a conserved motif in the yeast golgi GDP-mannose transporter required for binding to nucleotide sugar. J. Biol. Chem. 276:4424–4432. Geer, L. Y., M. Domrachev, D. J. Lipman, and S. H. Bryant. 2002. CDART: protein homology by domain architecture. Genome Res. 12:1619–1623. Hirai, T., J. A. Heymann, P. C. Maloney, and S. Subramaniam. 2003. Structural model for 12-helix transporters belonging to the major facilitator superfamily. J. Bacteriol. 185:1712–1718. Ishida, N., S. Yoshioka, M. Iida, K. Sudo, N. Miura, K. Aoki, and M. Kawakita. 1999. Indispensability of transmembrane domains of Golgi UDP-galactose transporter as revealed by analysis of genetic defects in UDP-galactose transporterdeficient murine had-1 mutant cell lines and construction of deletion mutants. J. Biochem. 126:1107–1117. Jack, D. L., N. M. Yang, and M. H. Saier, Jr. 2001. The drug/ metabolite transporter superfamily. Eur. J. Biochem. 268: 3620–3639. Knappe, S., U. I. Flugge, and K. Fischer. 2003. Analysis of the plastidic phosphate translocator gene family in Arabidopsis and identification of new phosphate translocator-homologous transporters, classified by their putative substrate-binding site. Plant Physiol. 131:1178–1190. Koike, K., G. Conseil, H. E. Leslie, H. I. Deeley, and S. P. Cole. 2004. Identification of proline residues in the core cytoplasmic and transmembrane regions of multidrug resistance protein 1 (MRP1/ABCC1) important for transport function, substrate specificity, and nucleotide interactions. J. Biol. Chem. 279: 12325–12336. Kolakovich, K. A., I. Y. Gluzman, K. L. Duffin, and D. E. Goldberg. 1997. Generation of hemoglobin peptides in the acidic digestive vacuole of Plasmodium falciparum implicates peptide transport in amino acid production. Mol. Biochem. Parasitol. 87:123–135. Krogstad, D. J., I. Y. Gluzman, D. E. Kyle, A. M. Oduola, S. K. Martin, W. K. Milhous, and P. H. Schlesinger. 1987. Efflux of chloroquine from Plasmodium falciparum: mechanism of chloroquine resistance. Science 238:1283–1285. Lin, Z., M. Itokawa, and G. R. Uhl. 2000. Dopamine transporter proline mutations influence dopamine uptake, cocaine analog recognition, and expression. FASEB J. 14:715–728. Liu, Y., D. M. Engelman, and M. Gerstein. 2002. Genomic analysis of membrane protein families: abundance and conserved motifs. Genome Biol. 3:research0054. Livshits, V. A., N. P. Zakataeva, V. V. Aleshin, and M. V. Vitushkina. 2003. Identification and characterization of the new gene rhtA involved in threonine and homoserine efflux in Escherichia coli. Res. Microbiol. 154:123–135. Lolkema, J. S., and D. J. Slotboom. 1998. Hydropathy profile alignment: a tool to search for structural homologues of membrane proteins. FEMS Microbiol. Rev. 22:305–322. Lubke, T., T. Marquardt, A. Etzioni, E. Hartmann, K. von Figura, and C. Korner. 2001. Complementation cloning identifies CDG-IIc, a new type of congenital disorders of glycosylation, as a GDP-fucose transporter deficiency. Nat. Genet. 28:73–76. Luhn, K., M. K. Wild, M. Eckhardt, R. Gerardy-Schahn, and D. Vestweber. 2001. The gene defective in leukocyte adhesion deficiency II encodes a putative GDP-fucose transporter. Nat. Genet. 28:69–72. Marchler-Bauer, A., A. R. Panchenko, B. A. Shoemaker, P. A. Thiessen, L. Y. Geer, and S. H. Bryant. 2002. CDD: a database of conserved domain alignments with links to domain three-dimensional structure. Nucleic Acids Res. 30: 281–283. Martin, R. E., J. W. H. Trueman, and K. Kirk. 2003. Bioinformatic analysis of PfCRT places it in a known family of transport proteins. Exp. Parasitol. 105:56–57. Martinez-Duncker, I., R. Mollicone, P. Codogno, and R. Oriol. 2003. The nucleotide-sugar transporter family: a phylogenetic approach. Biochimie 85:245–260. Nishikawa, A., J. B. Poster, Y. Jigami, and N. Dean. 2002. Molecular and phenotypic analysis of CaVRG4, encoding an essential Golgi apparatus GDP-mannose transporter. J. Bacteriol. 184:29–42. Oelmann, S., P. Stanley, and R. Gerardy-Schahn. 2001. Point mutations identified in Lec8 Chinese hamster ovary glycosylation mutants that inactivate both the UDP-galactose and CMPsialic acid transporters. J. Biol. Chem. 276:26291–26300. Reed, M. B., K. J. Saliba, S. R. Caruana, K. Kirk, and A. F. Cowman. 2000. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Nature 403:906–909. Rouanet, C., and W. Nasser. 2001. The PecM protein of the phytopathogenic bacterium Erwinia chrysanthemi, membrane topology and possible involvement in the efflux of the blue pigment indigoidine. J. Mol. Microbiol. Biotechnol. 3: 309–318. Russ, W. P., and D. M. Engelman. 2000. The GxxxG motif: a framework for transmembrane helix-helix association. J. Mol. Biol. 296:911–919. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425. Saliba, K. J., P. I. Folb, and P. J. Smith. 1998. Role for the Plasmodium falciparum digestive vacuole in chloroquine resistance. Biochem. Pharmacol. 56:313–320. Saliba, K. J., and K. Kirk. 1999. pH regulation in the intracellular malaria parasite, Plasmodium falciparum: H1 extrusion via a V-type H1-ATPase. J. Biol. Chem. 274:33213–33219. Sanchez, C. P., W. Stein, and M. Lanzer. 2003. Trans stimulation provides evidence for a drug efflux carrier as the mechanism of chloroquine resistance in Plasmodium falciparum. Biochemistry 42:9383–9394. Santiviago, C. A., J. A. Fuentes, S. M. Bueno, A. N. Trombert, A. A. Hildago, L. T. Socias, P. Youderian, and G. C. Mora. 2002. The Salmonella enterica sv. Typhimurium smvA, yddG and ompD (porin) genes are required for the efficient efflux of methyl viologen. Mol. Microbiol. 46:687–698. Segawa, H., M. Kawakita, and N. Ishida. 2002. Human and Drosophila UDP-galactose transporters transport UDP-Nacetylgalactosamine in addition to UDP-galactose. Eur. J. Biochem. 269:128–138. Shelden, M. C., P. Loughlin, M. L. Tierney, and S. M. Howitt. 2001. Proline residues in two tightly coupled helices of the sulphate transporter, SHST1, are important for sulphate transport. Biochem. J. 356:589–594. Sidhu, A. B., D. Verdier-Pinard, and D. A. Fidock. 2002. Chloroquine resistance in Plasmodium falciparum malaria parasites conferred by pfcrt mutations. Science 298:210–213. Sonnhammer, E. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol. 6:175–182. Stevens, T. J., and I. T. Arkin. 2000. The effect of nucleotide bias upon the composition and prediction of transmembrane helices. Protein Sci. 9:505–511. Streatfield, S. J., A. Weber, E. A. Kinsman, R. E. Hausler, J. Li, D. Post-Beittenmiller, W. M. Kaiser, K. A. Pyke, U. I. Flugge, and J. Chory. 1999. The phosphoenolpyruvate/phosphate translocator is required for phenolic metabolism, palisade cell The Chloroquine Resistance Transporter 1949 development, and plastid-dependent nuclear gene expression. Plant Cell 11:1609–1622. Tate, C. G., and P. J. Henderson. 1993. Membrane topology of the L-rhamnose-H1 transport protein (RhaT) from enterobacteria. J. Biol. Chem. 268:26850–26857. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. ClustalW: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positionspecific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680. Tran, C. V., and M. H. Saier, Jr. 2004. The principal chloroquine resistance protein of Plasmodium falciparum is a member of the drug/metabolite transporter superfamily. Microbiology 150:1–3. van Klompenburg, W., I. Nilsson, G. von Heijne, and B. de Kruijff. 1997. Anionic phospholipids are determinants of membrane protein topology. EMBO J. 16:4261–4266. Visiers, I., B. B. Braunheim, and H. Weinstein. 2000. Prokink: a protocol for numerical evaluation of helix distortions by proline. Protein Eng. 13:603–606. von Heijne, G. 1986. The distribution of positively charged residues in bacterial inner membrane proteins correlates with the trans-membrane topology. EMBO J. 5:3021–3027. Ward, J. M. 2001. Identification of novel families of membrane proteins from the model plant Arabidopsis thaliana. Bioinformatics 17:560–563. Warhurst, D. C. 2003. Polymorphism in the Plasmodium falciparum chloroquine-resistance transporter protein links verapamil enhancement of chloroquine sensitivity with the clinical efficacy of amodiaquine. Malaria J. 2:31. Warhurst, D. C., J. C. Craig, and I. S. Adagu. 2002. Lysosomes and drug resistance in malaria. Lancet 360:1527–1529. Webb, D. C., H. Rosenberg, and G. B. Cox. 1992. Mutational analysis of the Escherichia coli phosphate-specific transport system, a member of the traffic ATPase (or ABC) family of membrane transporters. a role for proline residues in transmembrane helices. J. Biol. Chem. 267:24661–24668. Woolfson, D. N., and D. H. Williams. 1990. The influence of proline residues on alpha-helical structure. FEBS Lett. 277:185–188. Zhang, H., E. M. Howard, and P. D. Roepe. 2002. Analysis of the antimalarial drug resistance protein Pfcrt expressed in yeast. J. Biol. Chem. 277:49767–49775. Michele Vendruscolo, Associate Editor Accepted June 28, 2004