Gene-fusion techniques for determining membrane-protein topology

Topological and transcriptional analysis of pssL gene product: a putative Wzx-like exopolysaccharide translocase in Rhizobium leguminosarum bv. trifolii TA1

An identified pssL gene is yet another one, besides the pssT, pssN and pssP genes, encoding for a protein engaged in polysaccharide polymerization and export in Rhizobium leguminosarum bv. trifolii strain TA1 (RtTA1). Amino acid sequence similarity and hypothetical protein secondary structure placed the PssL protein within Wzx (RfbX) translocases with putative flippase function that belong to the polysaccharide specific transport (PST) family. The predicted secondary structure of the PssL membrane protein was examined with a series of PssL-PhoA and PssL-LacZ translational fusions. The results support the hypothesis of PssL being a member of PST protein family comprising transporters with 12 membrane spanning segments and amino and carboxyl termini located in the cytoplasm. Results of semi-quantitative RT-PCR showed that the initial abundance of mRNA encoding PssL protein was relatively lower when compared to the quantity of the previously identified PssT membrane protein. PssL might be a good candidate for Wzx-like protein that together with PssT (Wzy protein) could be responsible for Wzx/Wzy-like-dependent EPS polymerization and translocation in RtTA1.

Topological and deletion analysis of CorS, a Pseudomonas syringae sensor kinase

A modified two-component regulatory system consisting of two response regulators, CorR and CorP, and the histidine protein kinase CorS, regulates the thermoresponsive production of the phytotoxin coronatine (COR) in Pseudomonas syringae PG4180. COR is produced at the virulence-promoting temperature of 18 °C, but not at 28 °C, the optimal growth temperature of PG4180. Assuming that the highly hydrophobic N-terminus of CorS might be involved in temperature-signal perception, the membrane topology of CorS was determined using translational phoA and lacZ fusions, leading to a topological model for CorS with six transmembrane domains (TMDs). Interestingly, three PhoA fusions located downstream of the sixth TMD showed a thermoresponsive phenotype. Enzymic activity, immunoblot, and protease-sensitivity assays were performed to localize the CorS derivatives, to analyse the expression level of hybrid proteins and to examine the model. In-frame deletions of the last four, or all six TMDs gave rise to non-functional CorS. The results indicated that the transmembrane region is important for CorS to function as a temperature sensor, and that the membrane topology of CorS might be involved in signal perception.

Membrane topology of PssT, the transmembrane protein component of the type I exopolysaccharide transport system in Rhizobium leguminosarum bv. trifolii strain TA1

The pssT gene was identified as the fourth gene located upstream of the pssNOP gene cluster possibly involved in the biosynthesis, polymerization, and transport of exopolysaccharide (EPS) in Rhizobium leguminosarum bv. trifolii strain TA1. The hydropathy profile and homology searches indicated that PssT belongs to the polysaccharide-specific transport family of proteins, a component of the type I system of the polysaccharide transport. The predicted membrane topology of the PssT protein was examined with a series of PssT-PhoA fusion proteins and a complementary set of PssT-LacZ fusions. The results generally support a predicted topological model for PssT consisting of 12 transmembrane segments, with amino and carboxyl termini located in the cytoplasm. A mutant lacking the C-terminal part of PssT produced increased amounts of total EPS with an altered distribution of high- and low-molecular-weight forms in comparison to the wild-type RtTA1 strain. The PssT mutant produced an increased number of nitrogen fixing nodules on clover.

Identification by phage display of the linear continuous MRPr1 epitope in the multidrug resistance-associated protein (MRP1)

In order to study the structure of the multidrug resistance-associated protein (MRP1), which is one of the most important members of the ATP-binding cassette (ABC) protein family acting as drug-efflux systems, we have developed an epitope mapping-based strategy. By means of the mAb MRPr1, we have immunoselected clones from two distinct random peptide libraries displayed on phages and have identified several peptide sequences mimicking the internal conformation of this 190 kDa multidrug transporter protein. Phage clones able to block the immunolabeling of the MRPr1 antibody to MRP1-overexpressing multidrug resistance (MDR) H69/AR cells were isolated and, after sequencing the corresponding inserts, their amino acid sequence was compared to that of MRP1. This analysis led to the identification of the consensus sequence L.SLNWED, corresponding to the MRP1 segment LWSLNKED (residues 241-248). This MRP1 sequence is partially overlapping with the MRPr1 epitope GSDLWSLNKE (residues 238-247) previously mapped using peptide scanning techniques. These results demonstrate the high reliability of phage display technology to study not only the topography of complex integral membrane proteins such as MRP1, but also to help identify critical residues participating in the formation of the epitope structure.

Transmembrane helix predictions revisited

Methods that predict membrane helices have become increasingly useful in the context of analyzing entire proteomes, as well as in everyday sequence analysis. Here, we analyzed 27 advanced and simple methods in detail. To resolve contradictions in previous works and to reevaluate transmembrane helix prediction algorithms, we introduced an analysis that distinguished between performance on redundancy-reduced high- and low-resolution data sets, established thresholds for significant differences in performance, and implemented both per-segment and per-residue analysis of membrane helix predictions. Although some of the advanced methods performed better than others, we showed in a thorough bootstrapping experiment based on various measures of accuracy that no method performed consistently best. In contrast, most simple hydrophobicity scale-based methods were significantly less accurate than any advanced method as they overpredicted membrane helices and confused membrane helices with hydrophobic regions outside of membranes. In contrast, the advanced methods usually distinguished correctly between membrane-helical and other proteins. Nonetheless, few methods reliably distinguished between signal peptides and membrane helices. We could not verify a significant difference in performance between eukaryotic and prokaryotic proteins. Surprisingly, we found that proteins with more than five helices were predicted at a significantly lower accuracy than proteins with five or fewer. The important implication is that structurally unsolved multispanning membrane proteins, which are often important drug targets, will remain problematic for transmembrane helix prediction algorithms. Overall, by establishing a standardized methodology for transmembrane helix prediction evaluation, we have resolved differences among previous works and presented novel trends that may impact the analysis of entire proteomes.

Membrane topology of the Mep/Amt family of ammonium transporters

The Mep/Amt proteins constitute a new family of transport proteins that are ubiquitous in nature. Members from bacteria, yeast and plants have been identified experimentally as high-affinity ammonium transporters. We have determined the topology of AmtB, a Mep/Amt protein from Escherichia coli, as a representative protein for the complete family. This was established using a minimal set of AmtB-PhoA fusion proteins with a complementary set of AmtB-LacZ fusions. These data, accompanied by an in silico analysis, indicate that the majority of the Mep/Amt proteins contain 11 membrane-spanning helices, with the N-terminus on the exterior face of the membrane and the C-terminus on the interior. A small subset, including E. coli AmtB, probably have an additional twelfth membrane-spanning region at the N-terminus. Addition of PhoA or LacZ alpha-peptide to the C-terminus of E. coli AmtB resulted in complete loss of transport activity, as judged by measurements of [14C]-methylammonium uptake. This C-terminal region, along with four membrane-spanning helices, contains multiple residues that are conserved within the Mep/Amt protein family. Structural modelling of the E. coli AmtB protein suggests a number of secondary structural features that might contribute to function, including a putative ammonium binding site on the periplasmic face of the membrane at residue Asp-182. The implications of these results are discussed in relation to the structure and function of the related human Rhesus proteins.

Insertional gene fusion technology

The classical 'end to end' gene fusion technique has widely been used for monitoring gene expression, biological screening and purification of recombinant proteins. Recent progress with the 'insertional' gene fusion approach, on the other hand, has demonstrated that this technique can be utilized for membrane protein topology analysis, display of randomized protein libraries and design of biosensor proteins. In this review, we describe examples of insertional gene fusion and compare the old and new gene fusion techniques.

Membrane topology of MntB, the transmembrane protein component of an ABC transporter system for manganese in the cyanobacterium Synechocystis sp. strain PCC 6803

The structure of the membrane protein MntB, a component of a manganese transporter system in Synechocystis sp. strain PCC 6803, was examined with a series of fusions to the reporter proteins alkaline phosphatase and beta-galactosidase. The results support a topological model for MntB consisting of nine transmembrane segments, with the amino terminus of the protein being in the periplasm and the carboxyl terminus being in the cytoplasm.

Specific amino acid substitutions in the proline-rich motif of the Rhizobium meliloti ExoP protein result in enhanced production of low-molecular-weight succinoglycan at the expense of high-molecular-weight succinoglycan

The production of the acidic exopolysaccharide succinoglycan (EPS I) by Rhizobium meliloti exoP* mutants expressing an ExoP protein lacking its C-terminal cytoplasmic domain and by mutants characterized by specific amino acid substitutions in the proline-rich motif (RX4PX2PX4SPKX9IXGXMXGXG) located from positions 443 to 476 of the ExoP protein was analyzed. The absence of the C-terminal cytoplasmic ExoP domain (positions 484 to 786) and the substitution of both arginine443 by isoleucine443 and proline457 by serine457 within the proline-rich motif resulted in enhanced production of low-molecular-weight (LMW) EPS I at the expense of high-molecular-weight (HMW) EPS I. The ratios of HMW to LMW EPS I of the wild type and mutant strains increased with osmolarity.

Unusual structure of the tonB-exb DNA region of Xanthomonas campestris pv. campestris: tonB, exbB, and exbD1 are essential for ferric iron uptake, but exbD2 is not

The nucleotide sequence of a 3.6-kb HindIII-SmaI DNA fragment of Xanthomonas campestris pv. campestris revealed four open reading frames which, based on sequence homologies, were designated tonB, exbB, exbD1, and exbD2. Analysis of translational fusions to alkaline phosphatase and beta-galactosidase confirmed that the TonB, ExbB, ExbD1, and ExbD2 proteins are anchored in the cytoplasmic membrane. The TonB protein of X. campestris pv. campestris lacks the conserved (Glu-Pro)n and (Lys-Pro)m repeats but harbors a 13-fold repeat of proline residues. By mutational analysis, the tonB, exbB, and exbD1 genes were shown to be essential for ferric iron import in X. campestris pv. campestris. In contrast, the exbD2 gene is not involved in the uptake of ferric iron.

Analysis of the transmembrane topology and membrane assembly of the GAT-1 gamma-aminobutyric acid transporter

The transmembrane topology of the Na+- and Cl--dependent gamma-aminobutyric acid transporter GAT-1 has been studied using protein chimeras in Xenopus oocytes. A series of COOH-terminal truncations was generated to which a prolactin epitope was fused. Following expression of transporter-prolactin chimeras in Xenopus oocytes, the transmembrane orientation of each chimera was determined by testing for protease sensitivity in an oocyte membrane preparation. Data from protease protection assays with GAT-1-prolactin chimeras has shown that residues in the loops connecting hydrophobic domain (HD)3 and HD4 and HD7 and HD8 are accessible to protease in the cytoplasm and suggest the presence of pore loop structures which extend into the membrane from the extracellular face. Such pore loop structures may be involved in the formation of the substrate-binding pocket. Studies presented herein confirm that the NH2 and COOH termini are cytosolic and hydrophobic domains span the membrane in a manner consistent with the predicted hydropathy model for Na+- and Cl--dependent transporters. These data also provide insight into GAT-1 transmembrane assembly and suggest that a complex series of topogenic sequences directs this process. A potential pause-transfer sequence has been identified and may be responsible for the translocational pausing observed in this study.

Activity of protein MalE (maltose-binding protein) fused to cytoplasmic and periplasmic regions of an Escherichia coli inner membrane protein

We analysed the properties of mature MBP (maltose-binding protein or MalE protein) fused to an integral cytoplasmic membrane protein of Escherichia coli. Fusion of MalE to the first MalG periplasmic loop enabled a strain defective in the malE gene to utilize maltose. In contrast, fusion of MalE to a cytoplasmic loop did not complement the malE delta 444 deletion. We obtained results highly correlated with those obtained by using alkaline phosphatase as a reporter for the topology of MalG. We discuss the possibility of genetically determining the topology of cytoplasmic membrane proteins by a method based on engineered fusions to MBP.

Evidence for a six-transmembrane domain structure of presenilin 1

Mutations in genes encoding presenilin 1 and presenilin 2 account for the majority of cases of early-onset familial Alzheimer's disease. The presenilins have been localized to the endoplasmic reticulum and Golgi, but which of the multiple hydrophobic segments of the polypeptide chain span the lipid bilayer is unclear. To address this question, we have constructed a series of chimeric molecules in which a topologically neutral reporter protein (a C-terminal fragment of prolactin) containing three artificial glycosylation sites is fused to presenilin 1 following each of the 10 potential transmembrane domains identified in hydrophobicity plots. We have expressed these chimeras by translation in reticulocyte lysate containing canine pancreatic microsomes and by synthesis in transfected COS cells. Based on utilization of the glycosylation sites and sensitivity of the reporter to protease digestion, we provide evidence that presenilin 1 has six transmembrane segments with the N and C termini in the cytoplasm. This model provides important clues to the potential functions of different parts of the presenilin molecule and how these might relate to the pathogenesis of Alzheimer's disease.

Membrane topology of the ATP-binding cassette transporter associated with antigen presentation (Tap1) expressed in Escherichia coli

The ATP-binding cassette transporters associated with antigen presentation (Tap1 and Tap2) mediate the transport of peptide fragments across the endoplasmic reticulum membrane of mammalian cells. Tap1 and Tap2 are closely related to one another and are believed to function as a heterodimer. Each protein possesses a hydrophobic domain predicted to span the membrane multiple times and a highly conserved nucleotide-binding domain. We have assessed the transmembrane topology of Tap1 by expressing a series of fusions to a reporter protein, the mature form of beta-lactamase in Escherichia coli. From these data a topological model can be derived in which Tap1 spans the membrane eight times, with several large loops exposed in the lumen of the endoplasmic reticulum and with both the N and C termini (including the nucelotide-binding domain) residing in the cytoplasm.

Membrane topology analysis of the Bacillus subtilis BofA protein involved in pro-sigma K processing

The Bacillus subtilis BofA protein is involved in regulation of pro-sigma K processing in the mother cell during the late stages of sporulation. A computer analysis of the BofA amino acid sequence indicates that it is an integral membrane protein. To determine the membrane topology of the protein, a series of gene fusions of bofA with lacZ or phoA reporter genes in Escherichia coli were analysed. A BofA topological model with two membrane-spanning segments, and with the N- and the C-terminal domains located in the region between the inner and outer membranes surrounding the forespore is presented. The analysis of different modifications of the last five amino acid residues of the BofA protein, obtained by PCR site-directed mutagenesis, suggests a possible role of the C-terminal domain in the regulation of pro-sigma K processing.

Membrane topology of Kch, a putative K+ channel from Escherichia coli

We have mapped the topology of the C-terminal half of the putative potassium channel protein Kch in the inner membrane of Escherichia coli using PhoA fusions. Our results are consistent with the widely assumed six-transmembrane helix model for eukaryotic voltage-gated ion channels and place both ends of the proposed channel-lining P-segment on the periplasmic side of the inner membrane. The rather hydrophobic P-segment is found to translocate only slowly across the inner membrane and seems to be near a threshold for stop-transfer function.

Topology prediction for helical transmembrane proteins at 86% accuracy

Previously, we introduced a neural network system predicting locations of transmembrane helices (HTMs) based on evolutionary profiles (PHDhtm, Rost B, Casadio R, Fariselli P, Sander C, 1995, Protein Sci 4:521-533). Here, we describe an improvement and an extension of that system. The improvement is achieved by a dynamic programming-like algorithm that optimizes helices compatible with the neural network output. The extension is the prediction of topology (orientation of first loop region with respect to membrane) by applying to the refined prediction the observation that positively charged residues are more abundant in extra-cytoplasmic regions. Furthermore, we introduce a method to reduce the number of false positives, i.e., proteins falsely predicted with membrane helices. The evaluation of prediction accuracy is based on a cross-validation and a double-blind test set (in total 131 proteins). The final method appears to be more accurate than other methods published: (1) For almost 89% (+/-3%) of the test proteins, all HTMs are predicted correctly. (2) For more than 86% (+/-3%) of the proteins, topology is predicted correctly. (3) We define reliability indices that correlate with prediction accuracy: for one half of the proteins, segment accuracy raises to 98%; and for two-thirds, accuracy of topology prediction is 95%. (4) The rate of proteins for which HTMs are predicted falsely is below 2% (+/-1%). Finally, the method is applied to 1,616 sequences of Haemophilus influenzae. We predict 19% of the genome sequences to contain one or more HTMs. This appears to be lower than what we predicted previously for the yeast VIII chromosome (about 25%).

Topology of the membrane protein LamB by epitope tagging and a comparison with the X-ray model

We previously developed a genetic approach to study, with a single antibody, the topology of the outer membrane protein LamB, an Escherichia coli porin with specificity towards maltodextrins and a receptor for bacteriophage lambda. Our initial procedure consisted of inserting at random the same reporter epitope (the C3 neutralization epitope from poliovirus) into permissive sites of LamB (i.e., sites which tolerate insertions without deleterious effects on the protein activities or the cell). A specific monoclonal antibody was then used to examine the position of the inserted epitope with respect to the protein and the membrane. In the present work, we set up a site-directed procedure to insert the C3 epitope at new sites in order to distinguish between two-dimensional folding models. This allowed us to identify two new surface loops of LamB and to predict another periplasmic exposed region. The results obtained by random and directed epitope tagging are analyzed in light of the recently published X-ray structure of the LamB protein. Study of 23 hybrid LamB-C3 proteins led to the direct identification of five of the nine external loops (L4, L5, L6, L7, and L9) and led to the prediction of four periplasmic loops (I1, I4, I5, and I8) of LamB. Nine of the hybrid proteins did not lead to topological conclusions, and none led to the wrong predictions or conclusions. The comparison indicates that parts of models based on secondary structure predictions alone are not reliable and points to the importance of experimental data in the establishment of outer membrane protein topological models. The advantages and limitations of genetic foreign epitope insertion for the study of integral membrane proteins are discussed.

Topological analysis of the lysine-specific permease of Escherichia coli

Escherichia coli accumulates lysine via two systems, one specific for lysine (LysP) and a second inhibited by arginine or ornithine (LAO). The lysP gene encodes a polypeptide of 489 residues. A topological analysis of the LysP protein was performed using gene fusions. Random in-frame fusions of the lysP gene with the lacZ or blaM genes were generated. Site-directed mutagenesis was also used to generate additional blaM fusions at specific locations in the lysP gene. Two methods were used to alleviate the problem of lethal expression of some lysP::blaM fusions. First, ternary fusions were constructed in which the arsD gene was fused at the 5' end of the lysP gene and the blaM gene fused at specific sites within the lysP gene. In these plasmids lysP expression was controlled by the ars promoter. Secondly, an E. coli strain with a pcnB mutation was used with some fusions to maintain the plasmids at a reduced copy number. From analysis of 30 gene fusions, a topological model of the LysP protein is proposed in which the protein has 12 membrane-spanning regions, with the N- and C-terminal in the cytosol.

Low-molecular-weight succinoglycan is predominantly produced by Rhizobium meliloti strains carrying a mutated ExoP protein characterized by a periplasmic N-terminal domain and a missing C-terminal domain

The membrane topology of the Rhizobium meliloti 2011 ExoP protein involved in polymerization and export of succinoglycan was analysed by translational fusions of lacZ and phoA reporter genes to the exoP gene. Based on this analysis, the ExoP protein could be divided into an N-terminal domain mainly located in the periplasmic space and a C-terminal domain located in the cytoplasm. Whereas the C-terminal domain of ExoP is characterized by a potential nucleotide-binding motif, the N-terminal ExoP domain contains the sequence motif 'PX2PX4SPKX11GXMXG', which is also present in proteins involved in the determination of O-antigen chain length. R. meliloti strains carrying mutated exoP* genes, exclusively encoding the N-terminal ExoP domain, produced a reduced amount of succinoglycan. This reduction could be suppressed by a mutation in the regulatory gene exoR. The ratio of low-molecular-weight to high-molecular-weight succinoglycan was significantly increased in the exoP* mutant strain. In the exoP*/exoR mutant strain only low-molecular-weight succinoglycan could be detected. Based on sequence homologies and similar hydropathic profiles, the N-terminal domain of ExoP was proposed to be a member of a protein family thought to be involved in polysaccharide chain-length determination.

Topology profile for a glutamate receptor: three transmembrane domains and a channel-lining reentrant membrane loop

We investigated the transmembrane topology of the GluR3 subunit that was translated in rabbit reticulocytes supplemented with microsomal membranes. A prolactin reporter epitope was fused to GluR3 at six locations, bracketing each of the proposed transmembrane domains. The sidedness of the epitope in the microsomal membrane was then assessed by proteinase K sensitivity. The N terminus and the entire region between M3 and M4 was extracellular, and the C terminus was intracellular by this method. Four native N-linked glycosylation sites in the amino terminus and one introduced site between M3 and M4 were utilized, confirming the extracellular location of these regions. Epitopes inserted upstream and downstream of M2 were protease sensitive and thus intracellular. Our results support a topological model for glutamate receptor subunits that consists of three transmembrane domains, M1, M3, and M4, and another domain, the proposed channel-lining M2, which forms a reentrant membrane segment with both ends facing the cytoplasm.