Positive-inside Rule Research Articles

Sec61p Contributes to Signal Sequence Orientation According to the Positive-Inside Rule

Protein targeting to the endoplasmic reticulum is mediated by signal or signal-anchor sequences. They also play an important role in protein topogenesis, because their orientation in the translocon determines whether their N- or C-terminal sequence is translocated. Signal orientation is primarily determined by charged residues flanking the hydrophobic core, whereby the more positive end is predominantly positioned to the cytoplasmic side of the membrane, a phenomenon known as the "positive-inside rule." We tested the role of conserved charged residues of Sec61p, the major component of the translocon in Saccharomyces cerevisiae, in orienting signals according to their flanking charges by site-directed mutagenesis by using diagnostic model proteins. Mutation of R67, R74, or E382 in Sec61p reduced C-terminal translocation of a signal-anchor protein with a positive N-terminal flanking sequence and increased it for signal-anchor proteins with positive C-terminal sequences. These mutations produced a stronger effect on substrates with greater charge difference across the hydrophobic core of the signal. For some of the substrates, a charge mutation in Sec61p had a similar effect as one in the substrate polypeptides. Although these three residues do not account for the entire charge effect in signal orientation, the results show that Sec61p contributes to the positive-inside rule.

Predicting the topology of transmembrane helical proteins using mean burial propensity and a hidden-Markov-model-based method.

Helices in membrane spanning regions are more tightly packed than the helices in soluble proteins. Thus, we introduce a method that uses a simple scale of burial propensity and a new algorithm to predict transmembrane helical (TMH) segments and a positive-inside rule to predict amino-terminal orientation. The method (the topology predictor of transmembrane helical proteins using mean burial propensity [THUMBUP]) correctly predicted the topology of 55 of 73 proteins (or 75%) with known three-dimensional structures (the 3D helix database). This level of accuracy can be reached by MEMSAT 1.8 (a 200-parameter model-recognition method) and a new HMM-based method (a 111-parameter hidden Markov model, UMDHMM(TMHP)) if they were retrained with the 73-protein database. Thus, a method based on a physiochemical property can provide topology prediction as accurate as those methods based on more complicated statistical models and learning algorithms for the proteins with accurately known structures. Commonly used HMM-based methods and MEMSAT 1.8 were trained with a combination of the partial 3D helix database and a 1D helix database of TMH proteins in which topology information were obtained by gene fusion and other experimental techniques. These methods provide a significantly poorer prediction for the topology of TMH proteins in the 3D helix database. This suggests that the 1D helix database, because of its inaccuracy, should be avoided as either a training or testing database. A Web server of THUMBUP and UMDHMM(TMHP) is established for academic users at http://www.smbs.buffalo.edu/phys_bio/service.htm. The 3D helix database is also available from the same Web site.

Membrane insertion of the chloroplast outer envelope protein, Toc34:constrains for insertion and topology

The insertion of the outer envelope protein Toc34 from chloroplasts was studied. Toc34 was chosen as a model protein because it contains one predicted transmembrane helix at the C-terminus and a large hydrophilic N-terminal located GTPase domain, which is exposed to the cytosol. Unlike proteins located in internal chloroplast compartments, Toc34 neither contains a cleavable presequence nor uses the general import pathway. The protein can insert into the outer envelope of chloroplasts but not into the outer membrane of mitochondria. Using protein-free liposomes we showed that Toc34 is able to insert directly into the lipid bilayer. This insertion is stimulated by GTP and the presence of nonbilayer lipids, but is independent of the presence or absence of charged lipids. The topology of the protein inserted into protein-free liposomes was not exclusively directed by the positive-inside rule but by the size of the hydrophilic domain.

Domain organization of the MscS mechanosensitive channel of Escherichia coli.

The major structural features of the Escherichia coli MscS mechanosensitive channel protein have been explored using alkaline phosphatase (PhoA) fusions, precise deletions and site-directed mutations. PhoA protein fusion data, combined with the positive-inside rule, strongly support a model in which MscS crosses the membrane three times, adopting an N(out)-C(in) configuration. Deletion data suggest that the C-terminal domain of the protein is essential for the stability of the MscS channel, whereas the protein will tolerate small deletions at the N-terminus. Four mutants that exhibit either gain-of-function (GOF) or loss-of-function have been identified: a double mutation I48D/S49P inactivates MscS, whereas the MscS mutants T93R, A102P and L109S cause a strong GOF phenotype. The similarity of MscS to the last two domains of MscK (formerly KefA) is reinforced by the demonstration that expression of a truncated MscK protein can substitute for MscL and MscS in downshock survival assays. The data derived from studies of the organization, conservation and the influence of mutations provide significant insights into the structure of the MscS channel.

Basic charge clusters and predictions of membrane protein topology.

The topology predictor SPLIT 4.0 (http://pref.etfos.hr) predicts the sequence location of transmembrane helices by performing an automatic selection of optimal amino acid attribute and corresponding preference functions. The best topological model is selected by choosing the highest absolute bias parameter that combines the bias in basic charge motifs and the bias in positive residues (the "positive inside rule") with the charge difference across the first transmembrane segment. Basic charge motifs, such as the BBB, BXXBB, and BBXXB motifs in alpha-helical integral membrane proteins, are significantly more frequent near cytoplasmic membrane surface than expected from the Arg/Lys (B) frequency. The predictor's accuracy is 99% for predicting 178 transmembrane helices in all membrane proteins or subunits of known 3D structure.

Analysis of Asymmetry of Membrane Proteins by Evolutionary Trace

The investigation of membrane proteins has revealed the difference in amino acid composition between the cytoplasmic and the extracellular segments. Von Heijine showed that positively charged residues are relatively abundant in the cytoplasmic segments, and the tendency in the amino acid composition is called ‘positive inside rule’ [6]. Nakashima and Nishikawa reported that alanine and arginine residues tend to be more prevalent in the cytoplasmic segment, while threonine and cysteine residues are preferentially located in the extracellular segment [4]. That is, there is asymmetry between the extracellular region and the cytosolic region in membrane proteins. Recently, the tertiary structures of two proteins belonging to the aquaporin family, transmembrane water-conduction channel (PDB code 1J4N) and glycerol-conducting channels (PDB code 1LDF), have been determined. The structures of two eubacterial ClC chloride channels have been also determined (PDB codes 1KPL and 1KPK). The difference in sequence and structure, suggests that the two membrane protein families have derived from different evolutionary origins. However, the two protein families share a common structural pattern. Both of them have a two-fold internal duplication in the primary structure. In addition, the X-ray crystallographic studies revealed that the two repetitive units are arranged in the anti-parallel manner in each tertiary structure. That is, the orientation of the N-terminal domain is opposite to that of the C-terminal domain against the membrane. Considering the asymmetric features of membrane proteins, it is expected that the N-terminal and the C-terminal domains in the same membrane protein have been subjected to different structural and/or functional pressures due to the opposite orientation. We applied several different methods of evolutionary trace (ET) to the multiple alignments of the repetitive units, in order to extract novel asymmetric features of the membrane proteins.

Anionic lipids stimulate Sec-independent insertion of a membrane protein lacking charged amino acid side chains.

We have investigated the influence of the different lipid classes of Escherichia coli on Sec-independent membrane protein insertion, using an assay in which a mutant of the single-spanning Pf3 coat protein is biosynthetically inserted into liposomes. It was found that phosphatidylethanolamine and other non-bilayer lipids do not have a significant effect on insertion. Surprisingly, the anionic lipids phosphatidylglycerol and cardiolipin stimulate N-terminal translocation of the protein, even though it has no charged amino acid side chains. This novel effect is general for anionic lipids and depends on the amount of charge on the lipid headgroup. Since the N-terminus of the protein is at least partially positively charged due to a helix dipole moment, apparently negatively charged lipids can stimulate translocation of slightly positively charged protein segments in a direction opposite to the positive-inside rule. A mechanism is proposed to explain these results.

Spontaneous insertion of gene 9 minor coat protein of bacteriophage M13 in model membranes

Gene 9 minor coat protein from bacteriophage M13 is known to be located in the inner membrane after phage infection of Escherichia coli. The way of insertion of this small protein (32 amino acids) into membranes is still unknown. Here we show that the protein is able to insert in monolayers. The limiting surface pressure of 35 mN/m for 1,2-dioleoyl- sn-glycero-3-phosphocholine and 1,2-dioleoyl- sn-glycero-3-phosphoglycerol lipid systems indicates that this spontaneous insertion can also occur in vivo. By carboxyfluorescein leakage experiments of vesicles it is demonstrated that protein monomers, or at least small aggregates, are more effective in releasing carboxyfluorescein than highly aggregated protein. The final orientation of the protein in the bilayer after insertion was addressed by proteinase K digestion, thereby making use of the unique C-terminal location of the antigenic binding site. After insertion the C-terminus is still available for the enzymatic digestion, while the N-terminus is not. This leads to the overall conclusion that the protein is able to insert spontaneously into membranes without the need of any machinery or transmembrane gradient, with the positively charged C-terminus remaining on the outside. The orientation after insertion of gene 9 protein is in agreement with the ‘positive inside rule’.

Analysis of the role of interfacial tryptophan residues in controlling the topology of membrane proteins.

Tryptophans have a high affinity for the membrane-water interface and have been suggested to play a role in determining the topology of membrane proteins. We investigated this potential role experimentally, using mutants of the single-spanning Pf3 coat protein, whose transmembrane topologies are sensitive to small changes in amino acid sequence. Mutants were constructed with varying numbers of tryptophans flanking the transmembrane region and translocation was assessed by an in vitro translation/translocation system. Translocation into Escherichia coli inner membrane vesicles could take place under a variety of experimental conditions, with co- or posttranslational assays and proton motive force-dependent or -independent mutants. It was found that translocation can even occur in pure lipid vesicles, under which conditions the tryptophans must directly interact with the lipids. However, under all these conditions tryptophans neither inhibited nor stimulated translocation, demonstrating that they do not affect topology and suggesting that this may be universal for tryptophans in membrane proteins. In contrast, we could demonstrate that lysines clearly prefer to stay on the cis-side of the membrane, in agreement with the positive-inside rule. A statistical analysis focusing on interfacially localized residues showed that in single-spanning membrane proteins lysines are indeed located on the inside, while tryptophans are preferentially localized at the outer interface. Since our experimental results show that the latter is not due to a topology-determining role, we propose instead that tryptophans fulfill a functional role as interfacially anchoring residues on the trans-side of the membrane.

Sequence Analysis of Less Hydrophobic Transmembrane Segments in Polytopic Membrane Proteins

To understand the function of membrane proteins, the knowledge of their structure, especially the knowledge of the position of transmembrane segments (TMSs) and of their membrane topology (i.e., orientation of TMSs across the membrane), is crucial. Since most TMSs of α-type polytopic (i.e., having multiple TMSs) membrane proteins are composed of hydrophobic amino acids, they are usually detected as peaks in the hydropathy plot [1]. However, there are a certain amount of TMSs that are relatively hydrophobic and so it is hard to distinguish them with relatively hydrophobic loop segments. Although how such less hydrophobic segments span the membrane has been a mystery, Ota et al. found an important clue on its mechanism [4]; according their result, each TMS has its own topogenic tendency like membrane proteins with a single TMS. Their topogenic tendency is determined by the distribution of neighboring positively-charged residues like the positive-inside rule [3]. Namely, if a TMS has more positive charges on its nearby N-terminal side, it tends to have the Nin/Cout orientation; otherwise, a TMS tends to be in the Nout/Cin orientation. Thus, if two neighboring TMSs have the same topogenic tendency, a loop segment between them may become a TMS to resolve this contradiction. To confirm this idea, we systematically analyzed the sequence features of both less-hydrophobic TMSs and relatively-hydrophobic loops. Our result not only supports Ota et al.’s conclusion but also implies a more general model of membrane topogenesis.

Recent advances in the understanding of membrane protein assembly and structure.

1. Introduction 2862. Membrane protein assembly inE. coli2862.1. Role of the SRP 2872.2. YidC – a translocon component devoted to membrane proteins? 2872.3. The TAT pathway 2882.4. ‘Spontaneous’ membrane protein insertion 2883. Membrane protein assembly in the ER 2893.1. How TM segments exit the translocon 2893.2. Proteins with multiple topologies 2903.3. Stop-transfer effector sequences 2913.4. Non-hydrophobic TM segments? 2913.5. ‘Frustrated’ topologies 2913.6. N-tail translocation across the ER 2924. Membrane protein assembly in mitochondria 2924.1. The Oxa1p pathway 2924.2. The TIM22/54 pathway 2935. Evolution of membrane protein topology 2935.1. RnfA/RnfE – two homologous proteins with opposite topologies 2935.2. YrbG – duplicating an odd number of TMs 2946. Genome-wide analysis of membrane proteins 2956.1. Prediction methods 2956.2. How many membrane proteins are there? 2956.3. The positive-inside rule 2966.4. Dominant classes of membrane proteins 2967. The structure of transmembrane α-helices 2967.1. What TM helices look like 2977.2. The ‘helical hairpin’ 2977.3. Prolines in TM helices 2977.4. Charged residues in TM helices: the ‘snorkel’ effect 2987.5. The ‘aromatic belt’ 2988. Helix–helix packing in a membrane environment 2988.1. Lessons learnt from glycophorin A 2988.2. Genetic screens for helix–helix interactions 2998.3. Statistical studies 2998.4. Membrane protein folding 2999. Recent 3D structures 3009.1. KcsA – the first ion channel 3009.2. MscL – sensing lateral pressure changes 3009.3. The cytochrome bc 1 complex 3009.4. Fumarate reductase 3019.5. Bacteriorhodopsin – watching a membrane protein at work 30110. Concluding remarks 30111. Acknowledgements 30212. References 302For a variety of reasons – not the least biomedical importance – integral membrane proteins are now very much in focus in many areas of molecular biology, biochemistry, biophysics, and cell biology. Our understanding of the basic processes of membrane protein assembly, folding, and structure has grown significantly in recent times, both as a result of new methodological developments, more high-resolution structure data, and the possibility to analyze membrane proteins on a genome-wide scale.So what is new in the membrane protein field? Various aspects of membrane protein assembly and structure have been reviewed over the past few years (Cowan & Rosenbusch, 1994; Hegde & Lingappa, 1997; Lanyi, 1997; von Heijne, 1997; Bernstein, 1998); here, I will try to bring together a number of exciting recent developments. Particularly noteworthy are the discoveries related to the mechanisms of membrane protein assembly into the inner membrane of E. coli, the inner membrane of mitochondria, and the way transmembrane segments are handled by the ER translocon.Other advances include detailed studies of the interaction between transmembrane helices and the lipid bilayer, and of helix–helix packing interactions in the membrane environment. The availability of full genomic sequences have made it possible to study membrane proteins on a genome-wide scale. Finally, a handful of new high-resolution 3D structures have appeared.This review will deal only with helix bundle proteins, i.e. integral membrane proteins where the transmembrane segments form α-helices. For reviews on the other major class of integral membrane proteins – the β-barrel proteins – see Schirmer (1998) and Buchanan (1999). For readers who prefer a more ‘literary’ introduction to the membrane protein field, may I suggest von Heijne (1999).

Membrane topology of CadA homologous P-type ATPase of Helicobacter pylori as determined by expression of phoA fusions in Escherichia coli and the positive inside rule

The only experimental data available on the membrane topology of transition metal ATPases are from in vitro studies on two distinct P-type ATPases (CadA and CopA) of a gastric bacterium, Helicobacter pylori, both postulated to contain eight transmembrane domains (H1 to H8). In this study, H. pylori CadA ATPase was subjected to analysis of membrane topology in vivo by expression of ATPase-alkaline phosphatase (AP) hybrid proteins in Escherichia coli using a novel vector, pBAD phoA. This vector contains an inducible arabinose promoter and unique restriction sites for fusion of DNA fragments to phoA. The phoA gene lacking sequences encoding its N-terminal signal peptide was linked to the C-terminal regions of the postulated five cytoplasmic and four periplasmic segments of the H. pylori pump. The results obtained by heterologous expression of ATPase-AP hybrid proteins showed consistence with a model of eight transmembrane domains. They also demonstrated that the H. pylori ATPase sequences are well assembled in the cytoplasmic membrane of E. coli, a neutralophilic bacterium. Cloning and amino acid sequence analysis of the homologous ATPase of Helicobacter felis further verified the topological model for the H. pylori pump analyzed here, although the degree of amino acid sequence identity varied between the corresponding transmembrane segments, from 25% for H1 up to 100% for H6. It was found that the topology of ATPase follows the `positive inside rule'. With respect to the bioenergetic capacities of H. pylori, we discuss here the membrane potential as a possible factor directing insertion of ATPases in the cytoplasmic membrane of gastric bacteria.

Topological analysis of an RND family transporter, MexD of Pseudomonas aeruginosa

The membrane topology of a resistance-nodulation-division (RND) family transporter, MexD of Pseudomonas aeruginosa, was determined. Although it had been predicted previously that most RND proteins contain 12 transmembrane helices, three independent computer programs used in the present study predicted that MexD possessed 11, 14 or 17 transmembrane segments. To investigate the topology of MexD more thoroughly, 25 MexD-PhoA (alkaline phosphatase) and 18 MexD-Bla (β-lactamase) fusion plasmids were constructed and analyzed. The resulting topological model had just 12 transmembrane helices and two periplasmic loops of about 300 residues between helices 1 and 2 and helices 7 and 8. It is therefore proposed that the N- and C-termini are located in the cytoplasm and the predicted orientation is consistent with the ‘positive-inside rule’. This topological model can be applied to other RND proteins.

Insertion of leader peptidase into the thylakoid membrane during synthesis in a chloroplast translation system

The mechanisms of targeting and insertion of chloroplast-encoded thylakoid membrane proteins are poorly understood. In this study, we have used a translation system isolated from chloroplasts to begin to investigate these mechanisms. The bacterial membrane protein leader peptidase (Lep) was used as a model protein because its targeting and insertion mechanisms are well understood for Escherichia coli and for the endoplasmic reticulum. Lep could thus provide insight into the functional homologies between the different membrane systems. Lep was efficiently expressed in the chloroplast translation system, and the protein could be inserted into thylakoid membranes with the same topology as in E. coli cytoplasmic membranes, following the positive-inside rule. Insertion of Lep into the thylakoid membrane was stimulated by the trans-thylakoid proton gradient and was strongly inhibited by azide, suggesting a requirement for SecA activity. Insertion most likely occurred in a cotranslational manner, because insertion could only be observed if thylakoid membranes were present during translation reactions but not when thylakoid membranes were added after translation reactions were terminated. To halt the elongation process at different stages, we translated truncated Lep mRNAs without a stop codon, resulting in the formation of stable ribosome nascent chain complexes. These complexes showed a strong, salt-resistant affinity for the thylakoid membrane, implying a functional interaction of the ribosome with the membrane and supporting a cotranslational insertion mechanism for Lep. Our study supports a functional homology for the insertion of Lep into the thylakoid membrane and the E. coli cytoplasmic membrane.

An Artificial Transmembrane Segment Directs SecA, SecB, and Electrochemical Potential-dependent Translocation of a Long Amino-terminal Tail

Many integral membrane proteins contain an amino-terminal segment, often referred to as an N-tail, that is translocated across a membrane. In many cases, translocation of the N-tail is initiated by a cleavable, amino-terminal signal peptide. For N-tail proteins lacking a signal peptide, translocation is initiated by a transmembrane segment that is carboxyl to the translocated segment. The mechanism of membrane translocation of these segments, although poorly understood, has been reported to be independent of the protein secretion machinery. In contrast, here we describe alkaline phosphatase mutants containing artificial transmembrane segments that demonstrate that translocation of a long N-tail across the membrane is dependent upon SecA, SecB, and the electrochemical potential in the absence of a signal peptide. The corresponding mutants containing signal peptides also use the secretion machinery but are less sensitive to inhibition of its components. We present evidence that inhibition of SecA by sodium azide is incomplete even at high concentrations of inhibitor, which suggests why SecA-dependent translocation may not have been detected in other systems. Furthermore, by varying the charge around the transmembrane segment, we find that in the absence of a signal peptide, the orientation of the membrane-bound alkaline phosphatase is dictated by the positive inside rule. However, the presence of a signal peptide is an overriding factor in membrane orientation and renders all mutants in an Nout-Cin orientation.

Topological analysis of DcuA, an anaerobic C4-dicarboxylate transporter of Escherichia coli.

Escherichia coli possesses three independent anaerobic C4-dicarboxylate transport systems encoded by the dcuA, dcuB, and dcuC genes. The dcuA and dcuB genes encode related integral inner-membrane proteins, DcuA and DcuB (433 and 446 amino acid residues), which have 36% amino acid sequence identity. A previous amino acid sequence-based analysis predicted that DcuA and DcuB contain either 12 or 14 transmembrane helices, with the N and C termini located in the cytoplasm or periplasm (S. Six, S. C. Andrews, G. Unden, and J. R. Guest, J. Bacteriol. 176:6470-6478, 1994). These predictions were tested by constructing and analyzing 66 DcuA-BlaM fusions in which C terminally truncated forms of DcuA are fused to a beta-lactamase protein lacking the N-terminal signal peptide. The resulting topological model differs from those previously predicted. It has just 10 transmembrane helices and a central, 80-residue cytoplasmic loop between helices 5 and 6. The N and C termini are located in the periplasm and the predicted orientation is consistent with the "positive-inside rule." Two highly hydrophobic segments are not membrane spanning: one is in the cytoplasmic loop; the other is in the C-terminal periplasmic region. The topological model obtained for DcuA can be applied to DcuA homologues in other bacteria as well as to DcuB. Overproduction of DcuA to 15% of inner-membrane protein was obtained with the lacUV5-promoter-based plasmid, pYZ4.

Anionic phospholipids are determinants of membrane protein topology.

The orientation of many membrane proteins is determined by the asymmetric distribution of positively charged amino acid residues in cytoplasmic and translocated loops. The positive-inside rule states that loops with large amounts of these residues tend to have cytoplasmic locations. Orientations of constructs derived from the inner membrane protein leader peptidase from Escherichia coli were found to depend on the anionic phospholipid content of the membrane. Lowering the contents of anionic phospholipids facilitated membrane passage of positively charged loops. On the other hand, elevated contents of acidic phospholipids in the membrane rendered translocation more sensitive to positively charged residues. The results demonstrate that anionic lipids are determinants of membrane protein topology and suggest that interactions between negatively charged phospholipids and positively charged amino acid residues contribute to the orientation of membrane proteins.

The Membrane Anchor of Microsomal Epoxide Hydrolase from Human, Rat, and Rabbit Displays an Unexpected Membrane Topology

The microsomal epoxide hydrolase (mEH) and cytochrome P450s catalyze the sequential formation of carcinogenic metabolites. According to one algorithm for predicting the membrane topology of proteins, the human, the rabbit, and the rat mEH should adopt a type II topology. The type II topology is also predicted by a recently established neuronal network which is trained to recognize signal peptides with very high accuracy. In contrast to these predictions we find, based on N-glycosylation analysis in a cell-free and in a cellular system, that the membrane anchor of human, rat, and rabbit mEH displays a type I topology. This result is correctly predicted by the positive inside rule in which negatively charged residues, the distribution of which differs in the mEH membrane anchor of these species, have only a modulating role for the membrane topology of proteins. However, our results demonstrate that this role is not strong enough to force the mEHs into a type II topology, not even in the case of the rabbit mEH, in which the only positively charged residue in the C-terminal part of the topogenic sequence is flanked by five negatively charged residues.

Membrane topology of the di- and tripeptide transport protein of Lactococcus lactis.

Transport of hydrophilic di- and tripeptides into Lactococcus lactis is mediated by a proton motive force-driven peptide transport protein (DtpT) that shares similarity with eukaryotic peptide transporters, e.g., from kidney and small intestine of rabbit, man, and rat. Hydropathy profiling in combination with the "positive inside rule" predicts for most of the homologous proteins an alpha-helical bundle of 12 transmembrane segments, but the positions of these transmembrane segments and the location of the amino and carboxyl termini are by no means conclusive. The secondary structure of DtpT was investigated by analyzing 42 DtpT-alkaline phosphatase fusion proteins, generated by random or directed fusions of the corresponding genes. These studies confirm the presence of 12 transmembrane segments but refute several other predictions made of the secondary structure. Data obtained from the fusion proteins were substantiated by studying the accessibility of single cysteine mutants in putative cytoplasmic or extracellular loops by membrane (im)permeant sulfhydryl reagents. The deduced topology model of DtpT consists of a bundle of 12 alpha-helixes with a short amino and a large carboxyl terminus, both located at the cytoplasmic site of the membrane. On the basis of sequence comparisons with DtpT, it seems likely that the structure model of the amino-terminal half of DtpT also holds for the eukaryotic peptide transporters, whereas the carboxyl-terminal half is largely different.

Phylogenetic classification of the mitochondrial carrier family of Saccharomyces cerevisiae.

The screening of the open reading frames identified in the whole yeast genome has allowed us to discover 34 proteins belonging to the mitochondrial carrier family. By phylogenetic study, they can be divided into 27 subfamilies including ADP/ATP, phosphate and citrate carriers, putative oxoglutarate and GDC carriers and 22 new subfamilies. Topology predictions using the 'positive inside rule' approach have shown that the yeast carriers are similarly oriented with both extremities exposed to the cytosol. In each subfamily, a strict conservation of the charged residues in the six transmembrane alpha-helices is observed, suggesting a functional role for these residues and the existence of 27 functionally distinct carriers.