Search

Membrane-protein integration, folding and assembly processes in vivo depend on complex targeting, translocation, chaperoning, and sorting machineries that somehow read the ‘molecular code’ built into the nascent polypeptide, ultimately producing a properly folded protein integrated into the correct target membrane. Although the main molecular constituents and the basic mechanistic principles of many of these machines are known in outline, the codes remain poorly defined and there is little quantitative information on how protein sequence affects the final structure of membrane proteins. By carefully designing model protein constructs, we have derived the first true biological hydrophobicity scale and have been able to get a first impression of how the position of a given type of residue within a transmembrane segment affects its ability to promote membrane insertion.

Using R-β-[1-11C]hydroxybutyrate and positron emission tomography, we studied the effect of acute hyperketonemia (range 0.7–1.7 μmol/ml) on cerebral ketone body utilization in six nondiabetic subjects and six insulin-dependent diabetes mellitus (IDDM) patients with average metabolic control (HbA1c = 8.1 ± 1.7%). An infusion of unlabeled R-β-hydroxybutyrate was started 1 h before the bolus injection of R-β-[1-11C]hydroxybutyrate. The time course of the radioactivity in the brain was measured during 10 min. For both groups, the utilization rate of ketone bodies was found to increase nearly proportionally with the plasma concentration of ketone bodies (1.0 ± 0.3 μmol/ml for nondiabetic subjects and 1.3 ± 0.3 μmol/ml for IDDM patients). No transport of ketone bodies from the brain could be detected. This result, together with a recent study of the tissue concentration of R-β-hydroxybutyrate in the brain by magnetic resonance spectroscopy, indicate that, also at acute hyperketonemia, the rate-limiting step for ketone body utilization is the transport into the brain. No significant difference in transport and utilization of ketone bodies could be detected between the nondiabetic subjects and the IDDM patients.

We describe and validate a new membrane protein topology prediction method, TMHMM, based on a hidden Markov model. We present a detailed analysis of TMHMM's performance, and show that it correctly predicts 97-98 % of the transmembrane helices. Additionally, TMHMM can discriminate between soluble and membrane proteins with both specificity and sensitivity better than 99 %, although the accuracy drops when signal peptides are present. This high degree of accuracy allowed us to predict reliably integral membrane proteins in a large collection of genomes. Based on these predictions, we estimate that 20-30 % of all genes in most genomes encode membrane proteins, which is in agreement with previous estimates. We further discovered that proteins with N(in)-C(in) topologies are strongly preferred in all examined organisms, except Caenorhabditis elegans, where the large number of 7TM receptors increases the counts for N(out)-C(in) topologies. We discuss the possible relevance of this finding for our understanding of membrane protein assembly mechanisms. A TMHMM prediction service is available at http://www.cbs.dtu.dk/services/TMHMM/.

Recently, a new protein translocation pathway, the twin-arginine translocation (TAT) pathway, has been identified in both bacteria and chloroplasts. To study the possible competition between the TAT- and the well-characterized Sec translocon-dependent pathways in Escherichia coli, we have fused the TorA TAT-targeting signal peptide to the Sec-dependent inner membrane protein leader peptidase (Lep). We find that the soluble, periplasmic P2 domain from Lep is re-routed by the TorA signal peptide into the TAT pathway. In contrast, the full-length TorA-Lep fusion protein is not re-routed into the TAT pathway, suggesting that Sec-targeting signals in Lep can override TAT-targeting information in the TorA signal peptide. We also show that the TorA signal peptide can be converted into a Sec-targeting signal peptide by increasing the hydrophobicity of its h-region. Thus, beyond the twin-arginine motif, the overall hydrophobicity of the signal peptide plays an important role in TAT versus Sec targeting. This is consistent with statistical data showing that TAT-targeting signal peptides in general have less hydrophobic h-regions than Sec-targeting signal peptides.

We have studied the effects of single charged residues on the position of a model transmembrane helix in the endoplasmic reticulum membrane using the glycosylation mapping technique. Asp and Glu residues cause a re-positioning of the C-terminal end of the transmembrane helix when placed in the one to two C-terminal turns but not when placed more centrally. Arg and Lys residues, in contrast, have little effect when placed in the two C-terminal turn but give rise to a more substantial shift in position when placed 9-11 residues from the helix end. We suggest that this difference between the effects of positively and negatively charged residues can be explained by the so-called snorkel effect, i.e. that the very long side-chains of Arg and Lys can reach up along the transmembrane helix to allow the terminal, charged moiety to reside in the lipid headgroup region while the Calpha of the residue is positioned well below the membrane/water interface.

We have carried out detailed statistical analyses of integral membrane proteins of the helix-bundle class from eubacterial, archaean, and eukaryotic organisms for which genome-wide sequence data are available. Twenty to 30% of all ORFs are predicted to encode membrane proteins, with the larger genomes containing a higher fraction than the smaller ones. Although there is a general tendency that proteins with a smaller number of transmembrane segments are more prevalent than those with many, uni-cellular organisms appear to prefer proteins with 6 and 12 transmembrane segments, whereas Caenorhabditis elegans and Homo sapiens have a slight preference for proteins with seven transmembrane segments. In all organisms, there is a tendency that membrane proteins either have many transmembrane segments with short connecting loops or few transmembrane segments with large extra-membraneous domains. Membrane proteins from all organisms studied, except possibly the archaeon Methanococcus jannaschii, follow the so-called "positive-inside" rule; i.e., they tend to have a higher frequency of positively charged residues in cytoplasmic than in extra-cytoplasmic segments.

Membrane proteins perform many of the essential functions required for life. They are often the targets of medicinal drugs and have many potential uses in biotechnological processes. Therefore our ability to understand them and manipulate their functions is both important and necessary to enable protein engineers to create 'designer' membrane proteins (that is, proteins designed to have desired properties).

We have analyzed the structure of mitochondrial cytochrome c oxidase in terms of general characteristics thought to be important for describing the architecture of helix bundle membrane proteins. Many aspects of the structure are similar to what has previously been found for the photosynthetic reaction center and bacteriorhodopsin. Our results lead to a considerably more precise general picture of membrane protein architecture than has hitherto been possible to obtain.

The chloroplast outer envelope protein OEP86 functions as a receptor in precursor protein translocation into chloroplasts. Sequence analysis suggests that the precursor of OEP86 is directed to the chloroplast outer envelope by a cleavable, negatively charged, and unusually long amino-terminal peptide. This presequence is unlike other potential targeting signals and suggests the existence of another membrane insertion pathway. Insertion of precursor OEP86 required the hydrolysis of adenosine triphosphate and the existence of surface exposed chloroplast membrane components, and it was not competed by another precursor protein destined for the internal plastid compartments.

We have studied the membrane insertion of two proteins from the inner membrane of Escherichia coli, both with two transmembrane segments connected by a short periplasmic loop: the M13 procoat protein and a mutant "inverted" leader peptidase. Neither molecule depends on the Sec machinery for insertion. We show that the introduction of a charged residue in the second transmembrane segment completely blocks insertion of both proteins. In contrast, a Sec-dependent procoat mutant, where the periplasmic region has been lengthened, inserts into the membrane even in the presence of a charged residue in the second hydrophobic domain. In addition, a large deletion within the second transmembrane domain of the leader peptidase mutant allows membrane translocation, but only under conditions where the SecA protein is functional. Furthermore, we show that the first hydrophobic domain is required for insertion of the short periplasmic loop of the "inverted" leader peptidase. These results suggest that Sec-independent insertion occurs by a synergistic entry of the two neighboring hydrophobic domains into the lipid bilayer.

Signal peptides (SPs) target proteins to the secretory pathway and are cleaved from the nascent chain once the translocase in the ER has been engaged. Signal-anchor (SA) sequences also interact transiently with the ER translocase, but are not cleaved and move laterally out of the translocase to become permanent membrane anchors. One obvious difference between SP and SA sequences is the considerably longer hydrophobic regions (h regions) of the latter. To study the interaction between SP/SA sequences and the ER translocase, we have constructed signal sequences with poly-Leu h regions ranging in length from 8 to 29 residues and have characterized their locations within the translocase using both a new assay that measures the minimum number of amino acids needed to span the distance between the COOH-terminal end of the h region and the active site of the oligosaccharyl transferase enzyme and an assay where the efficiency of signal peptidase catalyzed cleavage is measured. Our results suggest that SP and SA sequences are positioned differently in the ER translocase.

The membrane electrochemical potential is critical for the export of most periplasmic proteins in Escherichia coli. Its exact role during insertion of integral inner membrane proteins, however, remains obscure. Using derivatives of the inner membrane protein leader peptidase (Lep), we now show that the membrane potential appears to stimulate the membrane translocation of chain segments containing negatively charged residues, that positively charged regions appear to be more easily translocated in the absence of a potential, and that certain Lep constructs insert with different topologies in the presence and absence of a membrane potential, suggesting that the electrochemical potential introduces an asymmetry between the topological effects of positively and negatively charged amino acids during the process of membrane protein insertion in E. coli.

We have mapped the interface between the two transmembrane alpha-helices in the membrane domain of the Escherichia coli enzyme leader peptidase by analyzing disulfides formed between pairs of cysteine residues introduced near their respective periplasmic ends. The interface is formed primarily from aliphatic amino acids, and the two helices appear to pack against each other in a left-handed supercoil. We suggest that disulfide mapping may be a generally applicable approach for the construction of models of helix-helix interfaces in membrane proteins.

By in vitro transcription/translation of model proteins in the presence of dog pancreas microsomes, we have measured the minimum distance of an acceptor site from the lumenal end of a transmembrane segment required for N-linked glycosylation, both when the acceptor site is placed N- and C-terminally to the membrane anchor. We observe a sharp threshold at a distance of 12-14 residues, suggesting that the oligosaccharyltransferase active site is 30-40 A above the membrane and is oriented roughly parallel to the membrane surface.

Plasminogen-activator inhibitor type 2 (PAI-2) is a specific inhibitor of plasminogen activators that belongs to the serine protease inhibitor superfamily (SERPINS). PAI-2 exists in two molecular forms: an intracellular, non-glycosylated form and a secreted, glycosylated form. Like ovalbumin, PAI-2 contains an uncleaved internal secretion signal. By deletion analysis, we have mapped the secretion signal to two mildly hydrophobic regions near the NH2 terminus. We also show that both of these regions become more efficient translocation signals when their hydrophobicities are increased. The PAI-2 secretion signal provides a unique example of a signal that, by virtue of its poor efficiency, allows the synthesis of both an extracellular and an intracellular form of the protein.

To study the sequence requirements for addition of O-linked N-acetylgalactosamine to proteins, amino acid distributions around 174 O-glycosylation sites were compared with distributions around non-glycosylated sites. In comparison with non-glycosylated serine and threonine residues, the most prominent feature in the vicinity of O-glycosylated sites is a significantly increased frequency of proline residues, especially at positions -1 and +3 relative to the glycosylated residues. Alanine, serine and threonine are also significantly increased. The high serine and threonine content of O-glycosylated regions is due to the presence of clusters of several closely spaced glycosylated hydroxy amino acids in many O-glycosylated proteins. Such clusters can be predicted from the primary sequence in some cases, but there is no apparent possibility of predicting isolated O-glycosylation sites from primary sequence data.

1. Why membrane proteins? 2 Many functions 3 Bioinformatics and membrane proteins: is it feasible to predict the 3D structure of a membrane protein? 4 Predicting the topology of membrane proteins 5 How many methods to predict membrane protein topology? 5.1 From theory to practice 6 Benchmarking the predictors of transmembrane topology 6.1 Testing on membrane proteins of known structure and topology 6.2 Topological experimental data 6.3 Validation towards experimental data 7 How many membrane proteins in the Human genome? 8 Membrane proteins and genetic diseases: PhD-SNP at work 9 Last but not least: 3D MODELLING of membrane proteins 10 What can currently be done in practice? 11 Can we improve?

Although mitochondrial targeting peptides lack a common consensus sequence, a certain bias in the positional distribution of amino acids has recently been found. These patterns seem to be associated with cleavage of the precursor proteins by matrix processing proteases. We have extended the previous studies and found new sequence motifs that are conserved within subgroups of mitochondrial targeting peptides. These motifs have certain common themes, indicating that they are associated with cleavage by one single protease. Two of the conserved patterns have a high predictive value, but even for sequences that do not possess these patterns, a fairly accurate prediction of the cleavage site is shown to be possible. We also suggest that a well-conserved RXY decreases (S/A) pattern may be used to engineer efficiently recognized cleavage sites into uncleaved or artificial mitochondrial targeting peptides.

The prediction of protein structure from sequence has been a long-standing goal of molecular biology. Integral membrane proteins, once abhorred by protein chemists and crystallographers because of their insolubility and stubborn refusal to yield good crystals, now appear to hold great promises for efficient structure prediction and engineering. This is mainly due to the constraints on permissible structures imposed by the lipid environment, and to the apparent uncoupling between an initial membrane targeting and insertion process which determines the overall topological arrangement of the transmembrane segments and a subsequent 'condensation' of these segments into a unique folded state. Recent work suggests that the membrane insertion process is controlled by simple sequence elements composed of different combinations of long hydrophobic regions and flanking charged residues. In this review we sketch the most important structural rules relating amino acid sequence to membrane insertion to fully folded molecule, and their use for prediction and protein-engineering purposes.

Chances are that you have come across membrane proteins many times in your professional life: ion channels, aquaporins, G-protein-coupled receptors, drug resistance proteins. But it is also quite likely that you have never bothered to think about what the implications are of being a membrane protein, as opposed to a soluble protein. What is special about membrane proteins in terms of structure and function, how many membrane proteins are out there, how are they made in the cell? Welcome to the membrane protein universe!

Membrane-protein integration, folding and assembly processes in vivo depend on complex targeting, translocation, chaperoning, and sorting machineries that somehow read the 'molecular code' built into the nascent polypeptide, ultimately producing a properly folded protein integrated into the correct target membrane. Although the main molecular constituents and the basic mechanistic principles of many of these machines are known in outline, the codes remain poorly defined and there is little quantitative information on how protein sequence affects the final structure of membrane proteins. By carefully designing model protein constructs, we have derived the first true biological hydrophobicity scale and have been able to get a first impression of how the position of a given type of residue within a transmembrane segment affects its ability to promote membrane insertion.

The helical hairpin, two closely spaced transmembrane helices separated by a short turn, is a common structural element in integral membrane proteins. Previous studies on the sequence determinants of helical hairpin formation have focussed on the role of polar and charged residues placed centrally in a long stretch of hydrophobic residues, and have yielded a "propensity scale" for the relative efficiency with which different residues promote the formation of helical hairpins. In this study, we shift our attention to the role of charged residues flanking the hydrophobic stretch. Clusters of charged residues are known to hinder membrane translocation, and thus flanking charged residues may conceivably force a long hydrophobic segment to form a helical hairpin even if there are no or only weakly turn-promoting residues in the hydrophobic stretch. We indeed find that Lys and, more surprisingly, Asp residues strongly affect helical hairpin formation when placed next to a poly-Leu-based transmembrane segment. We also find that a cluster of four consecutive Lys residues can affect the efficiency of helical hairpin formation even when placed approximately 30 residues downstream of the hydrophobic stretch. These observations have interesting implications for the way we picture membrane protein topogenesis within the context of the endoplasmic reticulum (ER) translocon.

We describe and validate a new membrane protein topology prediction method, TMHMM, based on a hidden Markov model. We present a detailed analysis of TMHMM's performance, and show that it correctly predicts 97-98 % of the transmembrane helices. Additionally, TMHMM can discriminate between soluble and membrane proteins with both specificity and sensitivity better than 99 %, although the accuracy drops when signal peptides are present. This high degree of accuracy allowed us to predict reliably integral membrane proteins in a large collection of genomes. Based on these predictions, we estimate that 20-30 % of all genes in most genomes encode membrane proteins, which is in agreement with previous estimates. We further discovered that proteins with N(in)-C(in) topologies are strongly preferred in all examined organisms, except Caenorhabditis elegans, where the large number of 7TM receptors increases the counts for N(out)-C(in) topologies. We discuss the possible relevance of this finding for our understanding of membrane protein assembly mechanisms. A TMHMM prediction service is available at http://www.cbs.dtu.dk/services/TMHMM/.

Publisher Summary This chapter reviews the characteristics of mitochondrial targeting peptides (mTPs) defined by statistical and experimental studies and discusses their use for predicting the presence of mTPs in protein sequences. Most methods for mTP prediction focus on the amino terminus of possible mitochondrial proteins. However, other portions of proteins also contribute to the import process. It is determined that C-terminal features of the small subunit of Rubisco modulate import of its precursor into chloroplasts. Some experimental evidences are provided for the existence of a cotranslational pathway for protein import into mitochondria. It has been demonstrated that the mTP is involved in the regulation of ribosome binding to the mitochondrial outer membrane. This import pathway may require additional, hitherto unknown mTP features. Thus, improved mTP prediction tools may have to account for features of the mature protein, rather than focusing only on the N-terminal parts of the precursor.

The helical hairpin, two closely spaced transmembrane helices separated by a short turn, is a recurring structural element in integral membrane proteins, and may serve as a compact unit that inserts into the membrane en bloc. Previously, we have determined the propensities of the 20 natural amino acids, when present in the middle of a long hydrophobic stretch, to induce the formation of a helical hairpin with a lumenally exposed turn during membrane protein assembly into the endoplasmic reticulum membrane. Here, we present results from a similar set of measurements, but with the turn placed on the cytoplasmic side of the membrane. We find that a significantly higher number of turn-promoting residues need to be present to induce a cytoplasmic turn compared to a lumenal turn, and that, in contrast to the lumenal turn, the positively charged residues Arg and Lys are the strongest turn-promoters in cytoplasmic turns. These results suggest that the process of turn formation between transmembrane helices is different for lumenal and cytoplasmic turns.

The structural and molecular determinants that govern the correct membrane insertion and folding of membrane proteins are still ill-defined. By following the addition of sugar chains to engineered glycosylation sites (glycosylation mapping) in Na,K-ATPase beta isoforms expressed in vitro and in Xenopus oocytes, in combination with biochemical techniques, we have defined the C-terminal end of the transmembrane domain of these type II proteins. N-terminal truncation and the removal of a single charged residue at the N-terminal start of the putative transmembrane domain influence the proper positioning of the transmembrane domain in the membrane as reflected by a repositioning of the transmembrane domain, the exposure of a putative cryptic signal peptidase cleavage site, and the production of protein species unable to insert into the membrane. Glycosylation mapping in vivo revealed that the degree of glycosylation at acceptor sites located close to the membrane increases with the time proteins spend in the endoplasmic reticulum. Furthermore, core sugars added to such acceptor sites cannot be processed to fully glycosylated species even when the protein is transported to the cell surface. Thus, the glycosylation mapping strategy applied in intact cells is a useful tool for the study of determinants for the correct membrane insertion of type II and probably other membrane proteins, as well as for the processing of sugar chains in glycoproteins.

Leader peptidase (Lep) is a membrane-bound enzyme of the Escherichia coli inner membrane that serves to remove signal peptides from exported proteins. Statistical and experimental studies of known signal peptides have defined a short C-terminal region that seems to provide the information for correct cleavage by Lep. Based on the patterns of conserved amino acids found in this region, we have designed a signal peptide "cleavage cassette." This cassette is processed at the expected site when introduced after an uncleaved signal peptide. Furthermore, processing is blocked in the predicted manner when the (-3, -1)-rule for signal peptide cleavage is violated. This suggests that current understanding of the sequence requirements for signal peptide cleavage is sufficiently advanced to be used in, e.g. protein engineering applications.

Structural analysis of chloroplast transit peptides (cTPs), including secondary structure prediction and analysis of ‘cTP-like’ peptides of known 3D structure, suggests that cTPs are essentially flexible peptides devoid of regular secondary or tertiary structure. It is proposed that cTPs may be designed to interact with a succession of different chaperones on the chloroplast protein-import pathway.

We have recently reported a first experimental turn propensity scale for transmembrane helices. This scale was derived from measurements of how efficiently a given residue placed in the middle of a 40 residue poly(Leu) stretch induces the formation of a "helical hairpin" with two rather than one transmembrane segment. We have now extended these studies, and have determined the minimum length of a poly(Leu) stretch compatible with the formation of a helical hairpin. We have also derived a more fine-grained turn propensity scale by (i) introducing each of the 20 amino acid residues into the middle of the shortest poly(Leu) stretch compatible with helical hairpin formation, and (ii) introducing pairs of residues in the middle of the 40 residue poly(Leu) stretch. The new turn propensities are consistent with the amino acid frequencies found in short hairpin loops in membrane proteins of known 3D structure.

alpha-Helix formation in globular proteins has been studied both theoretically and experimentally for decades, while a lack of both high-resolution structures and suitable experimental techniques has hampered the study of helices in membrane proteins. We have developed a new experimental approach, glycosylation mapping, where the active site of the lumenally exposed endoplasmic reticulum enzyme oligosaccharyl transferase is used as a point of reference against which the position of a transmembrane segment in the membrane can be measured. Here, we report an initial analysis of the helix-breaking properties of proline residues inserted in a transmembrane helix. We find that proline residues can break a transmembrane helix, but only when inserted near the end, and only when the helix is sufficiently long. The glycosylation mapping technique may be generally useful for determining the position of transmembrane helices in the membrane.

Integral membrane proteins have recently moved to center stage in structural biology, partly as a result of the realization that a large number of widely used drugs are targeted to membrane proteins. 3D structure determination of membrane proteins is, however, exceedingly difficult, and alternative means of obtaining structurally relevant information must be sought. In this short communication, a new way to study the conformation and membrane localization of a single transmembrane protein segment--glycosylation mapping--is introduced.

A novel method to model and predict the location and orientation of alpha helices in membrane-spanning proteins is presented. It is based on a hidden Markov model (HMM) with an architecture that corresponds closely to the biological system. The model is cyclic with 7 types of states for helix core, helix caps on either side, loop on the cytoplasmic side, two loops for the non-cytoplasmic side, and a globular domain state in the middle of each loop. The two loop paths on the non-cytoplasmic side are used to model short and long loops separately, which corresponds biologically to the two known different membrane insertions mechanisms. The close mapping between the biological and computational states allows us to infer which parts of the model architecture are important to capture the information that encodes the membrane topology, and to gain a better understanding of the mechanisms and constraints involved. Models were estimated both by maximum likelihood and a discriminative method, and a method for reassignment of the membrane helix boundaries were developed. In a cross validated test on single sequences, our transmembrane HMM, TMHMM, correctly predicts the entire topology for 77% of the sequences in a standard dataset of 83 proteins with known topology. The same accuracy was achieved on a larger dataset of 160 proteins. These results compare favourably with existing methods.

We investigated the effect of 0.07% alcohol on regional brain activity at rest and during cognitive performance in order to elucidate the anatomical substrate for the effects of alcohol in man as well as to clarify the interaction between changes in cerebral activity induced by cognitive performance and alcohol inebriation.Regional cerebral blood flow (3D-PET, 15O Butanol) was measured in 13 male, nonalcoholic volunteers. Each subject was scanned 12 times, three scans in each of the following four situations: sober/rest, sober/test and inebriated/rest, inebriated/test. We used statistical parametric mapping and a computerized brain atlas for localization.Alcohol induced a sense of inebriation and elation as well as a relative increase of the cerebral blood flow in medial parts of the temporal lobes, in the anterobasal parts of the anterior cingulate cortex including the septal region. In addition, there was an increase of blood flow in midline parts of the lower brain stem. Relative decreases of flow were observed in the cerebellum and in the occipital cortex bilaterally. In the sober state, a computerized perceptual maze test and a (silent) serial seven test induced two distinct neocortical activation patterns that were specific to the tasks. Alcohol did not change these patterns and the test performance was also uninfluenced.A moderate dose of alcohol selectively activates target structures that pertain to the so-called cerebral reward system and the ascending reticular activating system. Alcohol (approximately 0.07%) appears to have only minor effects in the neocortical systems that are involved in on-line cognitive activity. This apparent independence between the subcortical alcohol target and the neocortical cognitive mechanisms is a new finding that appears to be of importance for an understanding of the effect of moderate doses of alcohol on the brain.

We have carried out detailed statistical analyses of integral membrane proteins of the helix-bundle class from eubacterial, archaean, and eukaryotic organisms for which genome-wide sequence data are available. Twenty to 30% of all ORFs are predicted to encode membrane proteins, with the larger genomes containing a higher fraction than the smaller ones. Although there is a general tendency that proteins with a smaller number of transmembrane segments are more prevalent than those with many, uni-cellular organisms appear to prefer proteins with 6 and 12 transmembrane segments, whereas Caenorhabditis elegans and Homo sapiens have a slight preference for proteins with seven transmembrane segments. In all organisms, there is a tendency that membrane proteins either have many transmembrane segments with short connecting loops or few transmembrane segments with large extra-membraneous domains. Membrane proteins from all organisms studied, except possibly the archaeon Methanococcus jannaschii, follow the so-called "positive-inside" rule; i.e., they tend to have a higher frequency of positively charged residues in cytoplasmic than in extra-cytoplasmic segments.

A new, simple method for predicting transmembrane segments in integral membrane proteins has been developed. It is based on low-stringency dot-plots of the query sequence against a collection of non-homologous membrane proteins using a previously derived scoring matrix [Cserzö et al., 1994, J. Mol. Biol., 243, 388-396]. This so-called dense alignment surface (DAS) method is shown to perform on par with earlier methods that require extra information in the form of multiple sequence alignments or the distribution of positively charged residues outside the transmembrane segments, and thus improves prediction abilities when only single-sequence information is available or for classes of membrane proteins that do not follow the 'positive inside' rule.

We have analyzed the structure of mitochondrial cytochrome c oxidase in terms of general characteristics thought to be important for describing the architecture of helix bundle membrane proteins. Many aspects of the structure are similar to what has previously been found for the photosynthetic reaction center and bacteriorhodopsin. Our results lead to a considerably more precise general picture of membrane protein architecture than has hitherto been possible to obtain.

Sorting signals route proteins to the correct subcellular compartment and define classes of evolutionarily conserved protein motifs. Their surprisingly low degree of sequence conservation suggests that partially functional sorting signals may arise continuously during evolution, and thus that existing proteins may be continually tested in new compartments. Intertaxonic combination events such as those involved in the establishment of mitochondria and chloroplasts have necessitated the evolution of new classes of organellar import signals, and have resulted in interesting hybrids between preexisting and newly created sorting signals.

We have developed a new method for the identification of signal peptides and their cleavage sites based on neural networks trained on separate sets of prokaryotic and eukaryotic sequence. The method performs significantly better than previous prediction schemes and can easily be applied on genome-wide data sets. Discrimination between cleaved signal peptides and uncleaved N-terminal signal-anchor sequences is also possible, though with lower precision. Predictions can be made on a publicly available WWW server.

Many integral membrane proteins span the hydrophobic core of a membrane with one or more a-helical segments each consisting of about 20 hydrophobic amino acids. The orientation of the hydrophobic stretch in the membrane is determined by its flanking amino acids. In general there are more positively charged residues present in cytoplasmic loops than in extra-cytoplasmic ones (von Heijne, 1986). In both prokaryotic and eukaryotic cells, proteins are targeted for secretion by N-terminal signal sequences with a common basic design: a positively charged N-terminus, a central hydrophobic stretch, and a C-terminal cleavage region that serves as a recognition site for the signal peptidase enzyme (von Heijne, 1985). Signal sequences from prokaryotes and eukaryotes look very similar and are often functionally interchangeable. They are essential for the efficient and selective targeting of the nascent protein chains either to the endoplasmic reticulum (in eukaryotes) or to the cytoplasmic membrane (in prokaryotes) (Gierasch, 1989). Signal sequences also play a central role in the interaction with the translocation machinery of the cell and in the translocation of the polypeptide chains across the membrane. Proteins are co-translationally translocated across the endoplasmic reticulum (ER) membrane. In eukaryotes two kinds of topogenic sequences are important for assembly in ER membrane: Those that initiate translocation such as signal peptides (SPs) and signalanchor sequences (SAs) and those that halt translocation, stop-transfer signals (STs). Signal peptides and signal anchor sequences differ in that SA sequences tend to have longer hydrophobic cores and lack a signal peptidase cleavage site (von Heijne, 1988).

We have shown previously that the 100-residue-long periplasmic N-terminal tail of the Escherichia coli inner membrane protein ProW can be translocated across the inner membrane in a sec-independent manner and that its translocation is blocked by the introduction of three positively charged residues near its C-terminal end (Whitley, P., Zander, T., Ehrmann, M., Haardt, M., Bremer, E., and von Heijne, G. (1994) EMBO J. 13, 4653-4661). We have now further analyzed the requirements for translocation of the N-terminal tail and found that the introduction of even a single arginine can block translocation. Position-specific differences in the effects on translocation of arginine insertions suggest that the C-terminal end of the N-terminal tail is more critical for translocation than the central and N-terminal regions. We also show that the N-terminal tail is translocated in a truncation mutant where a stop codon is placed immediately after the first transmembrane segment, provided that the transmembrane segment is flanked on its C-terminal end by positively charged residues. Thus, sec-independent translocation of a relatively large domain can be induced by a translocation signal located at the extreme C terminus of a protein.

The extramembraneous segments in a large collection of G-protein coupled receptors (GPCRs) have been analysed in terms of amino acid composition and length. It is shown that this family of multi-spanning integral membrane proteins conforms well to the 'positive inside' rule. Further, the extracellular N-terminal tails of GPCRs lacking a cleavable signal peptide are shown to be considerably shorter and to have a reduced content of positively charged amnio acids compared with the N-terminal tails of GPCRs endowed with a signal peptide. This suggests that extracellular N-terminal tails of eukaryotic plasma membrane proteins may be translocated by different mechanisms depending on whether or not they are preceded by a signal peptide.

Protein sorting signals provide good examples of peptides that can be studied both from a chemical and a biochemical perspective. Their simple designs and low degree of sequence conservation suggest that they are involved in rather non-specific peptide-lipid interactions, yet their ability to discriminate efficiently between the import machineries of different subcellular compartments rather points to the importance of peptide-receptor interactions. The study of protein sorting signals thus invites a cross-disciplinary approach.

Integral membrane proteins have parts of their polypeptide backbones embedded in a phospholipid bilayer. These hydrophobic, membrane-spanning domains are separated by hydrophilic segments exposed to the aqueous environment and seem to fall into two basic modalities: a-helix bundle and anti-parallel /3-barrel membrane domains. For the helix bundle proteins, two easily identifiable features appear to be the major structural determinants: the long apolar stretches that form the transmembrane a-helices, and the biased distribution of charged residues in the polar regions, known as the 'positive-inside' rule (von Heijne, 1986). Recently, attempts to predict topologies of eukaryotic and prokaryotic integral membrane proteins that form a-helical structures, have been reported (von Heijne, 1992; Sipos and von Heijne, 1993). This work has indeed shown that the topology of such proteins can be effectively deduced from the amino acid sequence. Thus, the aim of developing TopPred II is to compile all existing knowledge about topology in order to permit easy access to prediction of membrane protein topologies. TopPred II is an improved version of the preceding freeware TOP-PRED (von Heijne, 1992). It is now a stand-alone application written in Symantec THINK Pascal to allow its operation on any Macintosh computer with a 6.0.2 (or beyond) system, respecting the standard Apple file- and window-handling procedures, and making extensive use of the graphic abilities of Macintosh computers. The compiled program is very compact (~90 kbytes) and all the default parameters, scales and texts have been built into resources to allow easy access and ability for permanent modifications by the user. Input sequence files are limited to 2000 amino acids. Sequences can be handled one by one or in groups of up to 20. Parameters and scales can be easily modified through standard dialogs, which also enable one to re-establish the TopPred II default values. Calculation of the hydrophobicity profile, transmembrane segments and topologies (see below) can be requested. All the displayed results (texts as well as graphics) can be printed out or saved in files that can be imported by other programs.

We have studied the membrane insertion of two proteins from the inner membrane of Escherichia coli, both with two transmembrane segments connected by a short periplasmic loop: the M13 procoat protein and a mutant "inverted" leader peptidase. Neither molecule depends on the Sec machinery for insertion. We show that the introduction of a charged residue in the second transmembrane segment completely blocks insertion of both proteins. In contrast, a Sec-dependent procoat mutant, where the periplasmic region has been lengthened, inserts into the membrane even in the presence of a charged residue in the second hydrophobic domain. In addition, a large deletion within the second transmembrane domain of the leader peptidase mutant allows membrane translocation, but only under conditions where the SecA protein is functional. Furthermore, we show that the first hydrophobic domain is required for insertion of the short periplasmic loop of the "inverted" leader peptidase. These results suggest that Sec-independent insertion occurs by a synergistic entry of the two neighboring hydrophobic domains into the lipid bilayer.

The ProW protein, located in the inner membrane of Escherichia coli, has a very unusual topology with a 100-residue-long N-terminal tail protruding into the periplasmic space. We have studied the mechanism of membrane translocation of the periplasmic tail by analysing ProW-PhoA and ProW-Lep fusion proteins, both in wild-type cells and in cells with an impaired sec machinery. Our results show that the translocation efficiency is not affected by treatments that compromise the SecA and SecY functions, but that translocation is completely blocked by dissipation of the proton motive force or by the introduction of extra positively charged residues into the N-terminal tail. This suggests that the sec machinery can act properly only on domains located on the C-terminal side of a translocation signal, and that the N-terminal tail is driven through the membrane by a mechanism that involves the proton motive force.

Positively charged amino acids are major determinants of the topology of bacterial inner membrane proteins, whereas negatively charged residues by themselves have little or no influence on the transmembrane orientation. Further, positively charged amino acids can very efficiently block the function of signal sequences when placed immediately downstream, while negatively charged residues are much less potent also in this regard. Here, we show that a negatively charged aspartic acid situated close to a positively charged lysine can attenuate both of these effects in a position-specific manner, suggesting that intra- or intermolecular charge pairing can modulate the interactions between positively charged residues in the nascent chain and parts of the secretory machinery or membrane phospholipids. These observations further underscore the importance of charged amino acids during protein translocation and membrane protein assembly.

By in vitro transcription/translation of model proteins in the presence of dog pancreas microsomes, we have measured the minimum distance of an acceptor site from the lumenal end of a transmembrane segment required for N-linked glycosylation, both when the acceptor site is placed N- and C-terminally to the membrane anchor. We observe a sharp threshold at a distance of 12-14 residues, suggesting that the oligosaccharyltransferase active site is 30-40 A above the membrane and is oriented roughly parallel to the membrane surface.

Hydrophobic organization: Determination of the structure of the bacterial photosynthetic reaction center, bacterial porins, and bacteriorhodopsin allows a comparison of the basic structural features of integral membrane proteins. Structure parameters of membrane- and water-soluble proteins are surprisingly similar, given the different dielectric environments, except for the polarity of residues on the protein surface. Hydrophobic and electrostatic forces: 1) Intramembrane helix-helix interactions that are sensitive to small structure changes can dictate assembly of membrane proteins, as indicated by reconstitution of bacteriorhodopsin from proteolytic fragments and specific dimer formation of the human erythrocyte sialoglycoprotein glycophorin A. 2) Electrostatic interactions have an important role in determining the trans-membrane orientation of integral membrane proteins of the bacterial inner membrane, as expressed by the "positive-inside" rule for the distribution of basic residues on the cis relative to the trans side of the membrane-spanning alpha-helices. The use of this charge asymmetry rule, in conjunction with a hydrophobicity algorithm for prediction of membrane-spanning domains, allows accurate prediction of the folding patterns of such polypeptides across the membrane. A role of electrostatic interactions in assembly and maintenance of the structure of oligomeric integral membrane protein complexes is also implied by the separation and extrusion from the membrane, at high pH, of the major hydrophobic subunits of the cytochrome b6f complex from the chloroplast thylakoid membrane. It is inferred that the hydrophobic helix-helix interactions between the subunits of this complex, whose function is electron transfer and proton translocation, are relatively weak compared to those in bacteriorhodopsin.