In vivo expression of the lacY gene in two segments leads to functional lac permease.

The lacY gene of Escherichia coli was cut into two approximately equal-size fragments with Afl II and subcloned individually or together under separate lac operator/promoters in plasmid pT7-5. Under these conditions, lac permease is expressed in two portions: (i) the N-terminal portion (the N terminus, the first six putative transmembrane helices, and most of putative loop 7) and (ii) the C-terminal portion (the last six putative transmembrane helices and the C terminus). Cells harboring pT7-5 encoding both fragments transport lactose at about 30% the rate of cells expressing intact permease to a comparable steady-state level of accumulation. In contrast, cells expressing either half of the permease independently do not transport lactose. As judged by [35S]methionine labeling and immunoblotting, intact permease is completely absent from the membrane of cells expressing lacY fragments either individually or together. Thus, transport activity must result from an association between independently synthesized pieces of lac permease. When the gene fragments are expressed individually, the N-terminal portion of the permease is observed inconsistently, and the C-terminal portion is not observed. When the gene fragments are expressed together, polypeptides identified as the N- and C-terminal moieties of the permease are found in the membrane. It is concluded that the N- or C-terminal halves of lac permease are proteolyzed when synthesized independently and that association between the two complementing polypeptides leads to a more stable, catalytically active complex.     

Insertional mutagenesis of hydrophilic domains in the lactose permease of Escherichia coli.

The lactose permease of Escherichia coli is a membrane transport protein postulated to contain a hydrophilic N terminus (hydrophilic domain 1), 12 hydrophobic transmembrane alpha-helices that traverse the membrane in zigzag fashion connected by hydrophilic domains, and a hydrophilic C terminus (hydrophilic domain 13). To test whether the hydrophilic domains are important for function, each domain was independently disrupted by insertion of two or six contiguous histidine residues, and the mutants were characterized with respect to initial rate of lactose transport and steady-state level of accumulation. Remarkably, histidine insertions into 10 out of 13 hydrophilic domains result in molecules that catalyze lactose accumulation effectively, although the initial rate of transport is compromised in certain cases. In contrast, insertions into hydrophilic domain 3, 9, or 10 cause a marked decrease in transport activity. As judged by immunoblots and [35S]methionine pulse-chase experiments, diminished activity is not due to decreased expression of the mutated permeases, defective insertion into the membrane, or increased rates of proteolysis after insertion. The results (i) suggest that most of the hydrophilic domains in the permease do not play an essential role in the transport mechanism and (ii) focus on the region of the permease containing putative helices IX and X as being particularly important for activity.     

The N-terminal 22 amino acid residues in the lactose permease of Escherichia coli are not obligatory for membrane insertion or transport activity.

When the lactose (lac) permease of Escherichia coli is expressed from the lac promoter at relatively low rates, deletion of amino acid residues 2-8 (delta 7) or 2-9 (delta 8) from the hydrophilic N terminus has a relatively minor effect on the ability of the permease to catalyze active lactose transport. Activity is essentially abolished, however, and the permease is hardly detected in the membrane when two additional amino acid residues are deleted (delta 10), and mutants deleted of residues 2-23 (delta 22) or 2-39 (delta 38) also exhibit no activity and are not inserted into the membrane. Dramatically, when the defective deletion mutants are overexpressed at high rates via the T7 promoter, delta 10 and delta 22 are inserted into the membrane in a stable form and catalyze active lactose transport in a highly significant manner, whereas delta 38 is hardly detected in the membrane and exhibits no activity. Interestingly, a fusion protein consisting of delta 38 and the ompA leader peptide is inserted into the membrane but exhibits no transport activity. The results indicate that the N-terminal hydrophilic domain of lac permease and the N-terminal half of the first putative transmembrane alpha-helix are not mandatory for either membrane insertion or transport activity.     

A five-residue sequence near the carboxyl terminus of the polytopic membrane protein lac permease is required for stability within the membrane.

The lac permease (lacY gene product) of Escherichia coli contains 417 amino acid residues and is predicted to have a short hydrophilic amino terminus on the inner surface of the cytoplasmic membrane, multiple transmembrane hydrophobic segments in alpha-helical conformation, and a 17-amino acid residue hydrophilic carboxyl-terminal tail on the inner surface of the membrane. To assess the importance of the carboxyl terminus, the properties of several truncation mutants were studied. The mutants were constructed by site-directed mutagenesis such that stop codons were placed at specified positions, and the altered lacY genes were expressed at a relatively low rate from plasmid pACYC184. Permease truncated at position 407 or 401 retains full activity, and a normal complement of molecules is present in the membrane, as judged by immunoblot analyses. Thus, it is apparent that the carboxyl-terminal tail plays no direct role in membrane insertion of the permease, its stability, or in the mechanism of lactose/H+ symport. In marked contrast, when truncations are made at residues 396 (i.e., 4 amino acid residues from the carboxyl terminus of putative helix XII), 389, 372, or 346, the permease is no longer found in the membrane. Remarkably, however, when each of the mutated lacY genes is expressed at a high rate by means of the T7 RNA polymerase system [Tabor, S. & Richardson, C. C. (1985) Proc. Natl. Acad. Sci. USA 82, 1074-1079], all of the truncated permeases are present in the membrane, as indicated by [35S]methionine incorporation studies; however, permease truncated at residue 396, 389, 372, or 346 is defective with respect to lactose/H+ symport. Finally, pulse-chase experiments indicate that wild-type permease or permease truncated at residue 401 is stable, whereas permease truncated at or prior to residue 396 is degraded at a significant rate. The results are consistent with the notion that residues 396-401 in putative helix XII are important for protection against proteolytic degradation and suggest that this region of the permease may be necessary for proper folding.     

Evidence for the loop model of signal-sequence insertion into the endoplasmic reticulum.

The insertion of proteins into the endoplasmic reticulum is mediated by short hydrophobic domains called signal sequences, which are usually cleaved during insertion. We previously constructed DNAs encoding vesicular stomatitis virus glycoproteins with N-terminal extensions preceding the signal sequence and showed that these extensions allowed normal signal-sequence function and cleavage in vivo. To analyze signal sequence topology during membrane insertion, we generated a point mutation that blocks cleavage of these signal sequences. After expressing these proteins in HeLa cells, we used proteolysis of microsomal membranes to determine that the N terminus of the signal sequence and the C terminus of each protein remain on the cytoplasmic side of the endoplasmic reticulum after insertion. This result indicates that the proteins were inserted in a looped configuration. Extending this finding, we were able to reverse the orientation of such a mutant protein by deleting its normal C-terminal transmembrane and cytoplasmic domains. In addition to demonstrating that a signal sequence can function as a membrane anchor, these findings show that except for the presence of a cleavage site, the cleaved signal sequence of a type I transmembrane protein is structurally and functionally equivalent to the noncleaved signal sequences of type II transmembrane proteins.     

The central cytoplasmic loop of the major facilitator superfamily of transport proteins governs efficient membrane insertion

Deletion of 5 residues (Delta5) from the central cytoplasmic loop of the lactose permease of Escherichia coli has no significant effect on expression or activity, whereas Delta12 leads to increased rates of permease turnover after membrane insertion and decreased transport activity, and Delta20 abolishes insertion and activity. By expressing Delta12 or Delta20 in two halves, both expression and activity are restored to levels approximating wild type. Replacing deleted residues with random hydrophilic amino acids also leads to full recovery. However, introduction of hydrophobic residues decreases expression and activity in a context-dependent manner. Thus, a minimum length of the central cytoplasmic loop is vital for proper insertion, stability, and efficient transport activity, because of constraints at the cytoplasmic ends of helices VI and VII. Furthermore, the results are consistent with the idea that the middle cytoplasmic loop provides a temporal delay between insertion of the first six helices into the membrane before insertion of the second six helices.     

Distance determination in proteins using designed metal ion binding sites and site-directed spin labeling: application to the lactose permease of Escherichia coli.

As shown in the accompanying paper, the magnetic dipolar interaction between site-directed metal-nitroxide pairs can be exploited to measure distances in T4 lysozyme, a protein of known structure. To evaluate this potentially powerful method for general use, particularly with membrane proteins that are difficult to crystallize, both a paramagnetic metal ion binding site and a nitroxide side chain were introduced at selected positions in the lactose permease of Escherichia coli, a paradigm for polytopic membrane proteins. Thus, three individual cysteine residues were introduced into putative helix IV of a lactose permease mutant devoid of native cysteine residues containing a high-affinity divalent metal ion binding site in the form of six contiguous histidine residues in the periplasmic loop between helices III and IV. In addition, the construct contained a biotin acceptor domain in the middle cytoplasmic loop to facilitate purification. After purification and spin labeling, electron paramagnetic resonance spectra were obtained with the purified proteins in the absence and presence of Cu(II). The results demonstrate that positions 103, 111, and 121 are 8, 14, and > 23 A from the metal binding site. These data are consistent with an alpha-helical conformation of transmembrane domain IV of the permease. Application of the technique to determine helix packing in lactose permease is discussed.     

Functional complementation of internal deletion mutants in the lactose permease of Escherichia coli.

Using the lactose permease of Escherichia coli, a well-characterized membrane protein with 12 transmembrane domains, we demonstrated that certain paired in-frame deletion constructs complement each other functionally. Although cells expressing the deletion mutants individually are unable to catalyze active lactose accumulation, cells simultaneously expressing specific pairs of deletions catalyze transport up to 60% as do cells expressing wild-type permease. Moreover, complementation clearly does not occur at the level of DNA but probably occurs at the protein level. Remarkably, functional complementation is observed only with pairs of permease molecules containing large deletions and is not observed with missense mutations or point deletions. Although the mechanism of functional complementation is obscure, the findings indicate that certain pairs of permease molecules containing specific internal deletions can interact to form a functional complex in the same way phenomenologically as do independently expressed polypeptides corresponding to different N- and C-terminal portions of the permease.     

Ordered membrane insertion of an archaeal opsin in vivo

The prevailing model of polytopic membrane protein insertion is based largely on the in vitro analysis of polypeptide chains trapped during insertion by arresting translation. To test this model under conditions of active translation in vivo, we have used a kinetic assay to determine the order and timing with which transmembrane segments of bacterioopsin (BO) are inserted into the membrane of the archaeon Halobacterium salinarum. BO is the apoprotein of bacteriorhodopsin, a structurally well characterized protein containing seven transmembrane alpha-helices (A-G) with an N-out, C-in topology. H. salinarum strains were constructed that express mutant BO containing a C-terminal His-tag and a single cysteine in one of the four extracellular domains of the protein. Cysteine translocation during BO translation was monitored by pulse-chase radiolabeling and rapid derivatization with a membrane-impermeant, sulfhydryl-specific gel-shift reagent. The results show that the N-terminal domain, the BC loop, and the FG loop are translocated in order from the N terminus to the C terminus. Translocation of the DE loop could not be examined because cysteine mutants in this region did not yield a gel shift. The translocation order was confirmed by applying the assay to mutant proteins containing two cysteines in separate extracellular domains. Comparison of the translocation results with in vivo measurements of BO elongation indicated that the N-terminal domain and the BC loop are translocated cotranslationally, whereas the FG loop is translocated posttranslationally. Together, these results support a sequential, cotranslational model of archaeal polytopic membrane protein insertion in vivo.     

Atomic structure of a thermostable subdomain of HIV-1 gp41

Infection by HIV-1 involves the fusion of viral and cellular membranes with subsequent transfer of viral genetic material into the cell. The HIV-1 envelope glycoprotein that mediates fusion consists of the surface subunit gp120 and the transmembrane subunit gp41. gp120 directs virion attachment to the cell–surface receptors, and gp41 then promotes viral–cell membrane fusion. A soluble, α-helical, trimeric complex within gp41 composed of N-terminal and C-terminal extraviral segments has been proposed to represent the core of the fusion-active conformation of the HIV-1 envelope. A thermostable subdomain denoted N34(L6)C28 can be formed by the N-34 and C-28 peptides connected by a flexible linker in place of the disulfide-bonded loop region. Three-dimensional structure of N34(L6)C28 reveals that three molecules fold into a six-stranded helical bundle. Three N-terminal helices within the bundle form a central, parallel, trimeric coiled coil, whereas three C-terminal helices pack in the reverse direction into three hydrophobic grooves on the surface of the N-terminal trimer. This thermostable subdomain displays the salient features of the core structure of the isolated gp41 subunit and thus provides a possible target for therapeutics designed selectively to block HIV-1 entry.     

Membrane topology analysis of Escherichia coli mannitol permease by using a nested-deletion method to create mtlA-phoA fusions.

The Escherichia coli mannitol permease catalyzes the concomitant transport and phosphorylation of D-mannitol. This 68-kDa protein consists of a membrane-bound, N-terminal domain involved in mannitol binding and translocation and a C-terminal, cytoplasmic domain responsible for mannitol phosphorylation. Secondary-structure prediction methods suggest that the N-terminal half of the permease spans the membrane approximately seven times in alpha-helical segments, but these data cannot conclusively predict the structure. We have used gene fusions between mtlA (encoding the permease) and 'phoA (encoding alkaline phosphatase lacking its signal sequence) to further investigate the topology of the mannitol permease. Initially, fusions were constructed by using a lambda TnphoA vector and in vitro cloning of 'phoA into naturally occurring restriction sites in mtlA. However, the former method gave severe problems with insertion "hot-spots" in our vector systems, and the latter method was limited by the number of useful restriction sites. Therefore, we developed a nested-deletion method for creating mtlA-phoA fusions. 'phoA was first cloned downstream from the part of mtlA encoding the membrane-bound half of the permease. This construct was then treated with the appropriate restriction enzymes and with exonuclease III to create random fusions. An analysis of greater than 40 different fusion clones constructed by these methods provides strong evidence for six membrane-spanning regions in the mannitol permease with three relatively short periplasmic loops and two large cytoplasmic loops in the membrane-bound half of the protein.     

Properties and purification of an active biotinylated lactose permease from Escherichia coli.

A simplified approach for purification of functional lactose permease from Escherichia coli is described that is based on the construction of chimeras between the permease and a 100-amino acid residue polypeptide containing the biotin acceptor domain from the oxaloacetate decarboxylase of Klebsiella pneumoniae [Cronan, J. E., Jr. (1990) J. Biol. Chem. 265, 10327-10333]. Chimeras were constructed with a factor Xa protease site and the biotin acceptor domain in the middle cytoplasmic loop (loop 6) or at the C terminus of the permease. Each construct catalyzes active lactose transport in cells and right-side-out membrane vesicles. Moreover, the constructs are biotinylated in vivo, and in both chimeras, the factor Xa protease site is accessible from the cytoplasmic surface of the membrane. Both biotinylated permeases bind selectively to immobilized monomeric avidin and are eluted with free biotin in a high state of purity, and the loop 6 chimera catalyzes active transport after reconstitution into proteoliposomes. The methodology described should be applicable to other membrane proteins.     

Surface-exposed positions in the transmembrane helices of the lactose permease of Escherichia coli determined by intermolecular thiol cross-linking

Intermolecular thiol cross-linking was used to determine surface-exposed positions in 250 lactose permease mutants containing single-Cys replacements in each transmembrane helix. Significant cross-linking of monomers to produce homodimers is observed in nine mutants with a 5-A-long cross-linking agent containing bis-methane thiosulfonate reactive groups [position 78 (helix III); positions 185, 186, and 187 (helix VI); positions 263, 275, and 278 (helix VIII); and positions 308 (helix IX) and 398 (helix XII)]. The results are consistent with a current helix-packing model of the permease. Seven of the nine mutants that exhibit intermolecular cross-linking are located at or near the cytoplasmic ends of transmembrane helices; two are near periplasmic ends. The results suggest that only those Cys replacements accessible from the aqueous phase and not from the hydrophobic core of the membrane are susceptible to cross-linking because of the much higher reactivity of the thiolate anion relative to the thiol. Single-Cys mutants at positions 278 (helix VIII) and 398 (helix XII), which are located in opposite sides of the 12-helix bundle, exhibit similar rates of cross-linking with sigmoid kinetics. Furthermore, cross-linking is markedly decreased at 0 degrees C, suggesting that lateral diffusion of the permease within the plane of the membrane is important for intermolecular cross-linking. The findings confirm previous observations indicating that intermolecular cross-linking is a stochastic process resulting from random collisions and support a number of other lines of evidence that lactose permease is a monomer.     

Helix packing of lactose permease in Escherichia coli studied by site-directed chemical cleavage.

Biotinylated lactose permease from Escherichia coli containing a single-cysteine residue at position 330 (helix X) or at position 147, 148, or 149 (helix V) was purified by avidin-affinity chromatography and derivatized with 5-(alpha-bromoacetamido)-1,10-phenanthroline-copper [OP(Cu)]. Studies with purified, OP(Cu)-labeled Leu-330 --> Cys permease in dodecyl-beta-D-maltopyranoside demonstrate that after incubation in the presence of ascorbate, cleavage products of approximately 19 and 6-8 kDa are observed on immunoblots with anti-C-terminal antibody. Remarkably, the same cleavage products are observed with permease embedded in the native membrane. Comparison with the C-terminal half of the permease expressed independently as a standard indicates that the 19-kDa product results from cleavage near the cytoplasmic end of helix VII, whereas the 6- to 8-kDa fragment probably results from fragmentation near the cytoplasmic end of helix XI. Results are entirely consistent with a tertiary-structure model of the C-terminal half of the permease derived from earlier site-directed fluorescence and site-directed mutagenesis studies. Similar studies with OP(Cu)-labeled Cys-148 permease exhibit cleavage products at approximately 19 kDa and at 15-16 kDa. The larger fragment probably reflects cleavage at a site near the cytoplasmic end of helix VII, whereas the 15- to 16-kDa fragment is consistent with cleavage near the cytoplasmic end of helix VIII. When OP(Cu) is moved 100 degrees to position 149 (Val-149 --> Cys permease), a single product is observed at 19 kDa, suggesting fragmentation at the cytoplasmic end of helix VII. However, when the reagent is moved 100 degrees in the other direction to position 147 (Gly-147 --> Cys permease), cleavage is not observed. The results suggest that helix V is in close proximity to helices VII and VIII with position 148 in the interface between the helices, position 149 facing helix VII, and position 147 facing the lipid bilayer.     

Sequential truncation of the lactose permease over a three-amino acid sequence near the carboxyl terminus leads to progressive loss of activity and stability.

Previous experiments are consistent with the notion that residues 396-401 (... SVFTLS ...) at the carboxyl terminus of the last putative transmembrane helix of the lactose (lac) permease of Escherichia coli are important for protection against proteolytic degradation and suggest that this region of the permease may be necessary for proper folding. Stop codons (TAA) have now been substituted sequentially for amino acid codons 396-401 in the lacY gene, and the termination mutants were expressed from the plasmid pT7-5. With respect to transport, permease truncated at residue 396 or 397 is completely defective, while molecules truncated at residues 398, 399, 400, and 401, respectively, exhibit 15-25%, 30-40%, 40-45%, and 70-100% of wild-type activity. As judged by pulse-chase experiments with [35S]methionine, wild-type permease or permease truncated at residue 401 is stable, while permease molecules truncated at position 400, 399, 398, 397, or 396 are degraded at increasingly rapid rates. The findings indicate that either the last turn of putative helix XII or the region immediately distal to helix XII is important for proper folding and protection against proteolytic degradation.     

Folding and activity of circularly permuted forms of a polytopic membrane protein

The transmembrane subunit of the Glc transporter (IICB(Glc)), which mediates uptake and concomitant phosphorylation of glucose, spans the membrane eight times. Variants of IICB(Glc) with the native N and C termini joined and new N and C termini in the periplasmic and cytoplasmic surface loops were expressed in Escherichia coli. In vivo transport/in vitro phosphotransferase activities of the circularly permuted variants with the termini in the periplasmic loops 1 to 4 were 35/58, 32/37, 0/3, and 0/0% of wild type, respectively. The activities of the variants with the termini in the cytoplasmic loops 1 to 3 were 0/25, 0/4 and 24/70, respectively. Fusion of alkaline phosphatase to the periplasmic C termini stabilized membrane integration and increased uptake and/or phosphorylation activities. These results suggest that internal signal anchor and stop transfer sequences can function as N-terminal signal sequences in a circularly permuted alpha-helical bundle protein and that the orientation of transmembrane segments is determined by the amino acid sequence and not by the sequential appearance during translation. Of the four IICB(Glc) variants with new termini in periplasmic loops, only the one with the discontinuity in loop 4 is inactive. The sequences of loop 4 and of the adjacent TM7 and TM8 are conserved in all phosphoenolpyruvate-dependent carbohydrate:phosphotransferase system transporters of the glucose family.     

Recombinant scrapie-like prion protein of 106 amino acids is soluble

The N terminus of the scrapie isoform of prion protein (PrP^Sc) can be truncated without loss of scrapie infectivity and, correspondingly, the truncation of the N terminus of the cellular isoform, PrP^C, still permits conversion into PrP^Sc. To assess whether additional segments of the PrP molecule can be deleted, we previously removed regions of putative secondary structure in PrP^C; in the present study we found that deletion of each of the four predicted helices prevented PrP^Sc formation, as did deletion of the stop transfer effector region and the C178A mutation. Removal of a 36-residue loop between helices 2 and 3 did not prevent formation of protease-resistant PrP; the resulting scrapie-like protein, designated PrP^Sc106, contained 106 residues after cleavage of an N-terminal signal peptide and a C-terminal sequence for glycolipid anchor addition. Addition of the detergent Sarkosyl to cell lysates solubilized PrP^Sc106, which retained resistance to digestion by proteinase K. These results suggest that all the regions of proposed secondary structure in PrP are required for PrP^Sc formation, as is the disulfide bond stabilizing helices 3 and 4. The discovery of PrP^Sc106 should facilitate structural studies of PrP^Sc, investigations of the mechanism of PrP^Sc formation, and the production of PrP^Sc-specific antibodies.     

Mutations in the lacY gene of Escherichia coli define functional organization of lactose permease.

Mutations in the lacY gene of Escherichia coli have been used to analyze the functional organization of lactose permease. Deletions suggest that the NH2 terminus of lactose permease is not essential and can be replaced by residues of the cytoplasmic enzyme beta-galactosidase. Negative dominant mutations in the lacY gene can be explained by the assumption that membrane-associated lactose permease is active as a dimer or oligomer. The map positions of these mutations and other point mutations that lower or alter the sugar specificity define regions of lactose permease involved in sugar or proton binding and transport.     

An approach to membrane protein structure without crystals

The lactose permease of Escherichia coli catalyzes coupled translocation of galactosides and H(+) across the cell membrane. It is the best-characterized member of the Major Facilitator Superfamily, a related group of membrane proteins with 12 transmembrane domains that mediate transport of various substrates across cell membranes. Despite decades of effort and their functional importance in all kingdoms of life, no high-resolution structures have been solved for any member of this family. However, extensive biochemical, genetic, and biophysical studies on lactose permease have established its transmembrane topology, secondary structure, and numerous interhelical contacts. Here we demonstrate that this information is sufficient to calculate a structural model at the level of helix packing or better.     

A general method for determining helix packing in membrane proteins in situ: Helices I and II are close to helix VII in the lactose permease of Escherichia coli

It was previously shown that coexpression of the lactose permease of Escherichia coli in two contiguous fragments leads to functional complementation. We demonstrate here that site-directed thiol crosslinking of coexpressed permease fragments can be used to determine helix proximity in situ without the necessity of purifying the permease. After coexpression of the six N-terminal (N₆) and six C-terminal (C₆) transmembrane helices, each with a single Cys residue, crosslinking was carried out in native membranes and assessed by the mobility of anti-C-terminal-reactive polypeptides on immunoblots. A Cys residue at position 242 or 245 (helix VII) forms a disulfide with a Cys residue at either position 28 or 29 (helix I), but not with a Cys residue at position 27, which is on the opposite face of helix I, thereby indicating that the face of helix I containing Pro-28 and Phe-29 is close to helix VII. Similarly, a Cys residue at position 242 or 245 (helix VII) forms a disulfide with a Cys residue at either position 52 or 53 (helix II), but not with a Cys residue at position 54. Furthermore, low-efficiency crosslinking is observed between a Cys residue at position 52 or 53 and a Cys residue at position 361 (helix XI). The results indicate that helix VII lies in close proximity to both helices I and II and that helix II is also close to helix XI. The method should be applicable to a number of different polytopic membrane proteins.