An internal signal sequence: the asialoglycoprotein receptor membrane anchor

The human asialoglycoprotein receptor H1 is anchored in the membrane by a single stretch of 20 hydrophobic amino acids; the hydrophilic amino terminus faces the cytoplasm, and the carboxyl terminus is exoplasmic. We show here that glycosylation and insertion of the asialoglycoprotein receptor into the endoplasmic reticulum membrane is cotranslational and SRP-dependent and occurs without proteolytic cleavage. The membrane-anchor domain is necessary for membrane insertion, since a receptor with the segment deleted is neither inserted nor glycosylated. The segment is also sufficient for membrane insertion, since it will initiate translocation of a carboxy-terminal domain of rat alpha-tubulin across the membrane. We propose that a helical hairpin mechanism of membrane insertion is used both by cleaved amino-terminal and uncleaved internal signal sequences.     

Sec-dependent membrane protein biogenesis: SecYEG, preprotein hydrophobicity and translocation kinetics control the stop-transfer function.

Preprotein translocase catalyzes membrane protein integration as well as complete translocation. Membrane proteins must interrupt their translocation and be laterally released from the translocase into the lipid bilayer. We have analyzed the translocation arrest and lateral release activities of Escherichia coli preprotein translocase with an in vitro reaction and the preprotein proOmpA carrying a synthetic stop-transfer sequence. Membrane protein integration is catalytic, occurs with kinetics similar to those of proOmpA itself and only requires the functions of SecYEG and SecA. Though a strongly hydrophobic segment will direct the protein to leave the translocase and enter the lipid bilayer, a protein with a segment of intermediate hydrophobicity partitions equally between the translocated and membrane-integrated states. Analysis of the effects of PMF, varied ATP concentrations or synthetic translocation arrest show that the stop-translocation efficiency of a mildly hydrophobic segment depends on the translocation kinetics. In contrast, the lateral partitioning from translocase to lipids depends solely on temperature and does not require SecA ATP hydrolysis or SecA membrane cycling. Thus translocation arrest is controlled by the SecYEG translocase activity while lateral release and membrane integration are directed by the hydrophobicity of the segment itself. Our results suggest that a greater hydrophobicity is required for efficient translocation arrest than for lateral release into the membrane.     

Insertion of a multispanning membrane protein occurs sequentially and requires only one signal sequence

To study the insertion of multispanning membrane proteins into the endoplasmic reticulum, we constructed novel proteins on the cDNA level by repeating, up to four times, the internal signal-anchor domain of the asialoglycoprotein receptor H1. Upon in vitro translation in the presence of microsomes, these polypeptides are indeed inserted as polytopic membrane proteins. The first hydrophobic domain functions as a signal and the second as a stop-transfer sequence, while the third initiates a second translocation process, halted again by the fourth. We were able to demonstrate that insertion occurs sequentially, starting with the first apolar segment from the amino terminus. By replacing the original signal-anchor domains by a mutant sequence not recognized by signal recognition particle (SRP), it was shown that only the first hydrophobic domain needs to be a signal sequence and that the second translocation event does not require SRP.     

Negatively charged residues in the IgM stop-transfer effector sequence regulate transmembrane polypeptide integration

A non-hydrophobic sequence that contributes to the biogenesis of a transmembrane protein is termed a stop-transfer effector (STE). To examine the mechanism of STE-mediated stop-transfer, a series of fusion proteins were constructed containing variants of a putative STE from murine IgM fused to an otherwise translocated hydrophobic sequence. Unexpectedly, the fraction of molecules adopting transmembrane topology was insensitive to many amino acid substitutions within the STE sequence but varied directly with the number of negative charges. Furthermore, when present at the amino terminus of a reporter, mutants were observed that adopted type I (amino terminus lumenal) and type II (amino terminus cytoplasmic) transmembrane topologies, demonstrating that the STE sequence can be located at either side of the endoplasmic reticulum membrane. Our results suggest that recognition of a broad structural feature formed primarily by negatively charged residues within the STE halts translocation and triggers membrane integration, even when the negative charges end up on the cytoplasmic side of the membrane. Since functional STE sequences photocross-link to two membrane proteins not previously identified at the translocon, these unique proteins are presumably involved in recognizing STE sequences and/or facilitating STE function.     

Proprotein convertase PC3 is not a transmembrane protein

Proprotein convertase PC3 (also known as PC1) is an endopeptidase involved in proteolytic processing of peptide hormone precursors in granules of the regulated secretory pathway of endocrine cells. Lacking any extended hydrophobic segments, PC3 was considered to be a secretory protein only peripherally attached to the granule membrane. Recently, evidence has been presented that PC3 is a transmembrane protein with a 115-residue cytoplasmic domain and a membrane-spanning segment containing eight charged amino acids [Arnaoutova, I., et al. (2003) Biochemistry 42, 10445-10455]. Here, we analyzed the membrane topology of PC3 and of a PC3 construct containing a conventional transmembrane segment of 19 leucines. Alkaline extraction was performed to assess membrane integration. Exposure to the cytosol or to the ER lumen was tested by addition of C-terminal tags for phosphorylation or glycosylation, respectively. Protease sensitivity was assayed in permeabilized cells. The results show that the C-terminus of PC3 is translocated across the endoplasmic reticulum membrane. Furthermore, the proposed transmembrane segment of PC3 and a similar one of carboxypeptidase E did not stop polypeptide translocation when inserted into a stop-transfer tester construct. PC3 is thus not a transmembrane protein. These results have implications for the mechanism of granule sorting of PC3 as well as for the topology of PC2 and carboxypeptidase E, which have been reported to span the lipid membrane by homologous charged sequences.     

Structural requirements for interruption of protein translocation across rough endoplasmic reticulum membrane

Co-translational translocation of proteins across the membrane of rough endoplasmic reticulum (ER) is interrupted by particular amino acid sequences, which are functionally termed "stop-transfer sequence." We analyzed the structural requirements for the interruption of the peptide translocation. By the manipulation of the cDNA of interleukin 2 (IL2), which passes through ER membrane co-translationally, the middle portion of the IL2 molecule was replaced with systematically altered hydrophobic segments, leucine, alanine, or leucine/alanine mixed clusters. Furthermore, charged amino acid residues were introduced just downstream of the hydrophobic segments. These modified IL2 peptides were synthesized with wheat germ cell-free system in the presence of rough microsomes and the topology of the peptides in the microsomes was assessed by post-translational digestion with proteinase K. We obtained the following results. (i) Each modified protein was processed to the mature form but the extent of stop-translocation varied widely. The ratio of the stopped to the translocated products increased as the length and hydrophobicity of the inserted segment increased. (ii) Shorter hydrophobic segments than naturally occurring native transmembrane segment promoted stop-translocation. (iii) Proteins with hydrophobic segments followed by positive charges were more efficiently stop-translocated than those having negative charges. (iv) If the hydrophobicity of the segment was sufficiently high, the positive charges after the segment were not essential for stop-translocation. We also suggest that the stop-transfer process includes protein-protein interaction between the hydrophobic segment and translocation channel.     

Stop-transfer activity of hydrophobic sequences depends on the translation system

Signal and stop-transfer sequences are the known determinants involved in topogenesis of integral membrane proteins. To study the characteristics of stop-transfer sequences, artificial proteins have been created on the DNA level based on the cDNA of the asialoglycoprotein receptor H1. Its internal signal/anchor domain initiates translocation of the downstream sequence across the endoplasmic reticulum membrane. The ability of several hydrophobic sequences inserted into the translocating polypeptide to stop further transfer was analyzed by translation of the fusion proteins using the wheat germ extract and rabbit reticulocyte lysate systems with dog pancreas microsomes. We discovered that some of the sequences behave differently with respect to translocation across the membrane depending on the translation system. Expression of one of the fusion proteins in fibroblasts showed that the reticulocyte lysate system reflects more closely the in vivo situation than the wheat germ system. Our results suggest that in a homologous system the translating ribosomes interact with the translocation machinery and influence the termination of polypeptide transfer by hydrophobic sequences.     

Two translocating hydrophilic segments of a nascent chain span the ER membrane during multispanning protein topogenesis

During protein integration into the endoplasmic reticulum, the N-terminal domain preceding the type I signal-anchor sequence is translocated through a translocon. By fusing a streptavidin-binding peptide tag to the N terminus, we created integration intermediates of multispanning membrane proteins. In a cell-free system, N-terminal domain (N-domain) translocation was arrested by streptavidin and resumed by biotin. Even when N-domain translocation was arrested, the second hydrophobic segment mediated translocation of the downstream hydrophilic segment. In one of the defined intermediates, two hydrophilic segments and two hydrophobic segments formed a transmembrane disposition in a productive state. Both of the translocating hydrophilic segments were crosslinked with a translocon subunit, Sec61α. We conclude that two translocating hydrophilic segment in a single membrane protein can span the membrane during multispanning topogenesis flanking the translocon. Furthermore, even after six successive hydrophobic segments entered the translocon, N-domain translocation could be induced to restart from an arrested state. These observations indicate the remarkably flexible nature of the translocon.     

Protein sorting between mitochondrial membranes specified by position of the stop-transfer domain

Recently, we fused a matrix-targeting signal to a large fragment of vesicular stomatitis virus G protein, which contains near its COOH-terminus a well-characterized endoplasmic reticulum (ER) stop-transfer sequence; the hybrid G protein was sorted to the inner mitochondrial membrane (Nguyen, M., and G. C. Shore. 1987. J. Biol. Chem. 262:3929-3931). Here, we show that the 19 amino acid G stop-transfer domain functions in an identical fashion when inserted toward the COOH-terminus of an otherwise normal matrix precursor protein, pre-ornithine carbamyl transferase; after import, the mutant protein was found anchored in the inner membrane via the stop-transfer sequence, with its NH2 terminus facing the matrix and its short COOH-terminal tail located in the intermembrane space. However, when the G stop-transfer sequence was placed near the NH2 terminus, the protein was inserted into the outer membrane, in the reverse orientation (NH2 terminus facing out, with a large COOH-terminal fragment located in the intermembrane space). These observations for mitochondrial topogenesis can be explained by a simple extension of existing models for ER sorting.     

Signal and membrane anchor functions overlap in the type II membrane protein I gamma CAT

I gamma CAT is a hybrid protein that inserts into the membrane of the endoplasmic reticulum as a type II membrane protein. These proteins span the membrane once and expose the NH2-terminal end on the cytoplasmic side and the COOH terminus on the exoplasmic side. I gamma CAT has a single hydrophobic segment of 30 amino acid residues that functions as a signal for membrane insertion and anchoring. The signal-anchor region in I gamma CAT was analyzed by deletion mutagenesis from its COOH-terminal end (delta C mutants). The results show that the 13 amino acid residues on the amino-terminal side of the hydrophobic segment are not sufficient for membrane insertion and translocation. Mutant proteins with at least 16 of the hydrophobic residues are inserted into the membrane, glycosylated, and partially proteolytically processed by a microsomal protease (signal peptidase). The degree of processing varies between different delta C mutants. Mutant proteins retaining 20 or more of the hydrophobic amino acid residues can span the membrane like the parent I gamma CAT protein and are not proteolytically processed. Our data suggest that in the type II membrane protein I gamma CAT, the signals for membrane insertion and anchoring are overlapping and that hydrophilic amino acid residues at the COOH-terminal end of the hydrophobic segment can influence cleavage by signal peptidase. From this and previous work, we conclude that the function of the signal-anchor sequence in I gamma CAT is determined by three segments: a positively charged NH2 terminus, a hydrophobic core of at least 16 amino acid residues, and the COOH-terminal flanking hydrophilic segment.     

Internally located cleavable signal sequences direct the formation of Semliki Forest virus membrane proteins from a polyprotein precursor.

The proteolytic processes involved in the cotranslational production of the Semliki Forest virus proteins p62, 6K, and E1 from a common precursor polypeptide were analyzed by an in vitro translation-translocation assay. By studying the behavior of wild-type and mutant variants of the polyprotein, we show that the signal sequences responsible for membrane translocation of the 6K and E1 proteins reside in the C-terminal regions of p62 and 6K, respectively. We present evidence suggesting that the polyprotein is processed on the luminal side by signal peptidase at consensus cleavage sites immediately following the signal sequences. Our results also lead us to conclude that the 6K protein is a transmembrane polypeptide with its N terminus on the luminal side of the membrane (type I). Thus, the production of all three membrane proteins is directed by alternating signal and stop-transfer (anchor) sequences that function in translocation and cleavage of the virus precursor polyprotein. This also shows conclusively that internally located signal sequences can be cleaved by signal peptidase.     

Nucleotide sequences of the mRNA''s encoding the vesicular stomatitis virus G and M proteins determined from cDNA clones containing the complete coding regions.

The complete nucleotide sequences of the vesicular stomatitis virus mRNA's encoding the glycoprotein (G) and the matrix protein (M) have been determined from cDNA clones that contain the complete coding sequences from each mRNA. The G protein mRNA is 1,665 nucleotides long, excluding polyadenylic acid, and encodes a protein of 511 amino acids including a signal peptide of 16 amino acids. G protein contains two large hydrophobic domains, one in the signal peptide and the other in the transmembrane segment near the COOH terminus. Two sites of glycosylation are predicted at amino acid residues 178 and 335. The close correspondence of the positions of these sites with the reported timing of the addition of the two oligosaccharides during synthesis of G suggests that glycosylation occurs as soon as the appropriate asparagine residues traverse the membrane of the rough endoplasmic reticulum. The mRNA encoding the vesicular stomatitis virus M protein is 831 nucleotides long, excluding polyadenylic acid, and encodes a protein of 229 amino acids. The predicted M protein sequence does not contain any long hydrophobic or nonpolar domains that might promote membrane association. The protein is rich in basic amino acids and contains a highly basic amino terminal domain. Details of construction of the nearly full-length cDNA clones are presented.     

Topogenesis of two transmembrane type K+ channels,Kir 2.1 and KcsA

Potassium channels, which control the passage of K+ across cell membranes, have two transmembrane segments, M1 and M2, separated by a hydrophobic P region containing a highly conserved signature sequence. Here we analyzed the membrane topogenesis characteristics of the M1, M2, and P regions in two animal and bacterial two-transmembrane segment-type K+ channels, Kir 2.1 and KcsA, using an in vitro translation and translocation system. In contrast to the equivalent transmembrane segment, S5, in the voltage-dependent K+ channel, KAT1, the M1 segment in KcsA, was found to have a strong type II signal-anchor function, which favors the Ncyt/Cexo topology. The N-terminal cytoplasmic region was required for efficient, correctly orientated integration of M1 in Kir 2.1. Analysis of N-terminal modification by in vitro metabolic labeling showed that the N terminus in Kir 2.1 was acetylated. The hydrophobic P region showed no topogenic function, allowing it to form a loop, but not a transmembrane structure in the membrane; this region was transiently exposed in the endoplasmic reticulum lumen during the membrane integration process. M2 was found to possess a stop-transfer function and a type I signal-anchor function, enabling it to span the membrane. The C-terminal cytoplasmic region in KcsA was found to affect the efficiency with which the M2 achieved their final structure. Comparative topogenesis studies of Kir 2.1 and KcsA allowed quantification of the relative contributions of each segment and the cytoplasmic regions to the membrane topology of these two proteins. The membrane topogenesis of the pore-forming structure is discussed using results for Kir 2.1, KcsA, and KAT1.     

Membrane insertion of uracil permease, a polytopic yeast plasma membrane protein.

Uracil permease is a multispanning protein of the Saccharomyces cerevisiae plasma membrane which is encoded by the FUR4 gene and produced in limited amounts. It has a long N-terminal hydrophilic segment, which is followed by 10 to 12 putative transmembrane segments, and a hydrophilic C terminus. The protein carries seven potential N-linked glycosylation sites, three of which are in its N-terminal segment. Overexpression of this permease and specific antibodies were used to show that uracil permease undergoes neither N-linked glycosylation nor proteolytic processing. Uracil permease N-terminal segments of increasing lengths were fused to a reporter glycoprotein, acid phosphatase. The in vitro and in vivo fates of the resulting hybrid proteins were analyzed to identify the first signal anchor sequence of the permease and demonstrate the cytosolic orientation of its N-terminal hydrophilic sequence. In vivo insertion of the hybrid protein bearing the first signal anchor sequence of uracil permease into the endoplasmic reticulum membrane was severely blocked in sec61 and sec62 translocation mutants.     

Evidence that insertion of Tomato ringspot nepovirus NTB-VPg protein in endoplasmic reticulum membranes is directed by two domains: a C-terminal transmembrane helix and an N-terminal amphipathic helix

The NTB-VPg protein of Tomato ringspot nepovirus is an integral membrane protein found in association with endoplasmic reticulum (ER)-derived membranes active in virus replication. A transmembrane helix present in a hydrophobic region at the C terminus of the NTB domain was previously shown to traverse the membranes, resulting in the translocation of the VPg domain in the lumen. We have now conducted an in planta analysis of membrane-targeting domains within NTB-VPg using in-frame fusions to the green fluorescent protein (GFP). As expected, the entire NTB-VPg protein directed the GFP fluorescence to ER membranes. GFP fusion proteins containing the C-terminal 86 amino acids of NTB-VPg also associated with ER membranes, resulting in ER-specific glycosylation at a naturally occurring glycosylation site in the VPg domain. Deletion of the hydrophobic region prevented the membrane association. The N-terminal 80 amino acids of NTB were also sufficient to direct the GFP fluorescence to intracellular membranes. A putative amphipathic helix in this region was necessary and sufficient to promote membrane association of the fusion proteins. Using in vitro membrane association assays and glycosylation site mapping, we show that the N terminus of NTB can be translocated in the lumen at least in vitro. This translocation was dependent on the presence of the putative amphipathic helix, suggesting that oligomeric forms of this helix traverse the membrane. Taken together, our results suggest that at least two distinct elements play a key role in the insertion of NTB-VPg in the membranes: a C-terminal transmembrane helix and an N-terminal amphipathic helix. An updated model of the topology of the protein in the membrane is presented.     

Glycosylation can influence topogenesis of membrane proteins and reveals dynamic reorientation of nascent polypeptides within the translocon.

The topology of multispanning membrane proteins in the mammalian endoplasmic reticulum is thought to be dictated primarily by the first hydrophobic sequence. We analyzed the in vivo insertion of a series of chimeric model proteins containing two conflicting signal sequences, i.e., an NH(2)-terminal and an internal signal, each of which normally directs translocation of its COOH-terminal end. When the signals were separated by more than 60 residues, linear insertion with the second signal acting as a stop-transfer sequence was observed. With shorter spacers, an increasing fraction of proteins inserted with a translocated COOH terminus as dictated by the second signal. Whether this resulted from membrane targeting via the second signal was tested by measuring the targeting efficiency of NH(2)-terminal signals followed by polypeptides of different lengths. The results show that targeting is mediated predominantly by the first signal in a protein. Most importantly, we discovered that glycosylation within the spacer sequence affects protein orientation. This indicates that the nascent polypeptide can reorient within the translocation machinery, a process that is blocked by glycosylation. Thus, topogenesis of membrane proteins is a dynamic process in which topogenic information of closely spaced signal and transmembrane sequences is integrated.     

Sensing charges of the Ciona intestinalis voltage-sensing phosphatase

Voltage control over enzymatic activity in voltage-sensitive phosphatases (VSPs) is conferred by a voltage-sensing domain (VSD) located in the N terminus. These VSDs are constituted by four putative transmembrane segments (S1 to S4) resembling those found in voltage-gated ion channels. The putative fourth segment (S4) of the VSD contains positive residues that likely function as voltage-sensing elements. To study in detail how these residues sense the plasma membrane potential, we have focused on five arginines in the S4 segment of the Ciona intestinalis VSP (Ci-VSP). After implementing a histidine scan, here we show that four arginine-to-histidine mutants, namely R223H to R232H, mediate voltage-dependent proton translocation across the membrane, indicating that these residues transit through the hydrophobic core of Ci-VSP as a function of the membrane potential. These observations indicate that the charges carried by these residues are sensing charges. Furthermore, our results also show that the electrical field in VSPs is focused in a narrow hydrophobic region that separates the extracellular and intracellular space and constitutes the energy barrier for charge crossing.     

A tripartite structure of the signals that determine protein insertion into the endoplasmic reticulum membrane

Multilineage colony stimulating factor is a secretory protein with a cleavable signal sequence that is unusually long and hydrophobic. Using molecular cloning techniques we exchanged sequences NH2- or COOH-terminally flanking the hydrophobic signal sequence. Such modified fusion proteins still inserted into the membrane but their signal sequence was not cleaved. Instead the proteins were now anchored in the membrane by the formerly cleaved signal sequence (signal-anchor sequence). They exposed the NH2 terminus on the exoplasmic and the COOH terminus on the cytoplasmic side of the membrane. We conclude from our results that hydrophilic sequences flanking the hydrophobic core of a signal sequence can determine cleavage by signal peptidase and insertion into the membrane. It appears that negatively charged amino acid residues close to the NH2 terminal side of the hydrophobic segment are compatible with translocation of this segment across the membrane. A tripartite structure is proposed for signal-anchor sequences: a hydrophobic core region that mediates targeting to and insertion into the ER membrane and flanking hydrophilic segments that determine the orientation of the protein in the membrane.     

Residues flanking the COOH-terminal C-region of a model eukaryotic signal peptide influence the site of its cleavage by signal peptidase and the extent of coupling of its co-translational translocation and proteolytic processing in vitro

The polar, COOH-terminal c-region of signal peptides has been considered to be most important for influencing the efficiency and fidelity of signal peptidase cleavage while the hydrophobic core or h-region appears indispensable for initiating translocation. To identify structural features of residues flanking the c-region that influence the fidelity and efficiency of signal peptidase cleavage as well as co-translational translocation, we introduced six amino acid substitutions into the COOH terminus of the hydrophobic core and seven substitutions at the NH2 terminus of the mature region (the +1 position) of a model eukaryotic preprotein-human pre(delta pro)apoA-II. This preprotein contains several potential sites for signal peptidase cleavage. The functional consequences of these mutations were assayed using an in vitro co-translational translocation/processing system and by post-translational cleavage with purified, detergent-solubilized, hen oviduct signal peptidase. The efficiency of translocation could be correlated with the hydrophobic character of the residue introduced at the COOH terminus of the h-region. Some h/c boundary mutants underwent co-translational translocation across the microsomal membrane with only minimal cleavage yet they were cleaved post-translationally by hen oviduct signal peptidase more efficiently than other mutants which exhibited a high degree of coupling of co-translational translocation and cleavage. These data suggest that features at the COOH terminus of the h-domain can influence "presentation" of the cleavage site to signal peptidase. The +1 residue substitutions had minor effects on the extent of co-translational translocation and processing. However, these +1, as well as h/c boundary mutations, had dramatic effects on the site of cleavage chosen by signal peptidase, indicating that residues flanking the c-region of this prototypic eukaryotic signal peptide can affect the fidelity of its proteolytic processing. The site(s) selected by canine microsomal and purified hen oviduct signal peptidase were very similar, suggesting that "intrinsic" structural features of this prepeptide can influence the selectivity of eukaryotic signal peptidase cleavage, independent of the microsomal membrane and associated translocation apparatus.