Glycoproteins form mixed disulphides with oxidoreductases during folding in living cells

The formation of intra- and interchain disulphide bonds constitutes an integral part of the maturation of most secretory and membrane-bound proteins in the endoplasmic reticulum. Evidence indicates that members of the protein disulphide isomerase (PDI) superfamily are part of the machinery needed for proper oxidation and isomerization of disulphide bonds. Models based on in vitro studies predict that the formation of mixed disulphide bonds between oxidoreductase and substrate is intermediate in the generation of the native intrachain disulphide bond in the substrate polypeptide. Whether this is how thiol oxidoreductases work inside the endoplasmic reticulum is not clear. Nor has it been established which of the many members of the PDI superfamily interacts directly with newly synthesized substrate proteins, because transient mixed disulphides have never been observed in the mammalian endoplasmic reticulum during oxidative protein folding. Here we describe the mechanisms involved in co- and post-translational protein oxidation in vivo. We show that the endoplasmic-reticulum-resident oxidoreductases PDI and ERp57 are directly involved in disulphide oxidation and isomerization, and, together with the lectins calnexin and calreticulin, are central in glycoprotein folding in the endoplasmic reticulum of mammalian cells.     

Defective co-translational formation of disulphide bonds in protein disulphide-isomerase-deficient microsomes

The formation of disulphide bonds in mammalian secretory and cell-surface proteins occurs in the lumen of the endoplasmic reticulum and is believed to be catalysed by the enzyme protein disulphide-isomerase (PDI). The evidence for this physiological role for PDI is circumstantial and relates to the cell and tissue distribution of the enzyme, its developmental behaviour and its catalytic properties in vitro. A clear requirement for PDI in the correct folding or assembly of disulphide-bonded proteins during biosynthesis has not been demonstrated. We have prepared dog pancreas microsomes which are deficient in soluble lumenal proteins, including PDI, but which are still able to translocate and process proteins synthesized in vitro. Using the formation of intramolecular disulphide bonds during the in vitro synthesis of gamma-gliadin, a wheat storage protein, as a model, we have demonstrated that these microsomes are defective in co-translational formation of disulphide bonds. Reconstitution of these microsomes with purified PDI reverses this defect.     

S-nitrosylated protein-disulphide isomerase links protein misfolding to neurodegeneration

Stress proteins located in the cytosol or endoplasmic reticulum (ER) maintain cell homeostasis and afford tolerance to severe insults. In neurodegenerative diseases, several chaperones ameliorate the accumulation of misfolded proteins triggered by oxidative or nitrosative stress, or of mutated gene products. Although severe ER stress can induce apoptosis, the ER withstands relatively mild insults through the expression of stress proteins or chaperones such as glucose-regulated protein (GRP) and protein-disulphide isomerase (PDI), which assist in the maturation and transport of unfolded secretory proteins. PDI catalyses thiol-disulphide exchange, thus facilitating disulphide bond formation and rearrangement reactions. PDI has two domains that function as independent active sites with homology to the small, redox-active protein thioredoxin. During neurodegenerative disorders and cerebral ischaemia, the accumulation of immature and denatured proteins results in ER dysfunction, but the upregulation of PDI represents an adaptive response to protect neuronal cells. Here we show, in brains manifesting sporadic Parkinson's or Alzheimer's disease, that PDI is S-nitrosylated, a reaction transferring a nitric oxide (NO) group to a critical cysteine thiol to affect protein function. NO-induced S-nitrosylation of PDI inhibits its enzymatic activity, leads to the accumulation of polyubiquitinated proteins, and activates the unfolded protein response. S-nitrosylation also abrogates PDI-mediated attenuation of neuronal cell death triggered by ER stress, misfolded proteins or proteasome inhibition. Thus, PDI prevents neurotoxicity associated with ER stress and protein misfolding, but NO blocks this protective effect in neurodegenerative disorders through the S-nitrosylation of PDI.     

Role of ATP and disulphide bonds during protein folding in the endoplasmic reticulum.

Being topologically equivalent to the extracellular space, the lumen of the endoplasmic reticulum (ER) provides a unique folding environment for newly synthesized proteins. Unlike other compartments in the cell where folding occurs, the ER is oxidizing and therefore can promote the formation of disulphide bonds. The reducing agent dithiothreitol, when added to living cells, inhibits disulphide formation with profound effects on folding. Taking advantage of this effect, we demonstrate here that folding of influenza haemagglutinin is energy dependent. Metabolic energy is required to support the correct folding and disulphide bond formation in this well characterized viral glycoprotein, to rescue misfolded proteins from disulphide-linked aggregates, and to maintain the oxidized protein in its folded and oligomerization-competent state.     

The molten globule protein conformation probed by disulphide bonds

The molten globule is a compact protein conformation that has a secondary structure content like that of the native protein, but poorly defined tertiary structure. It is a stable state for a few proteins under particular conditions and could be a ubiquitous kinetic intermediate in protein folding. The extent to which native interactions, above the level of the secondary structure, are preserved in this conformation is not so far known. Here we report that alpha-lactalbumin can adopt a molten globule conformation when one of its four disulphide bonds is reduced. In this state, the three other disulphide bonds rearrange spontaneously, at the same rate as when the protein is fully unfolded, to a number of different disulphide bond isomers that tend to maintain the molten globule conformation. That the molten globule state is compatible with a variety of disulphide bond pairings suggests that it is unlikely to be stabilized by many specific tertiary interactions.     

Probing the chemistry of thioredoxin catalysis with force

Thioredoxins are enzymes that catalyse disulphide bond reduction in all living organisms. Although catalysis is thought to proceed through a substitution nucleophilic bimolecular (S(N)2) reaction, the role of the enzyme in modulating this chemical reaction is unknown. Here, using single-molecule force-clamp spectroscopy, we investigate the catalytic mechanism of Escherichia coli thioredoxin (Trx). We applied mechanical force in the range of 25-600 pN to a disulphide bond substrate and monitored the reduction of these bonds by individual enzymes. We detected two alternative forms of the catalytic reaction, the first requiring a reorientation of the substrate disulphide bond, causing a shortening of the substrate polypeptide by 0.79 +/- 0.09 A (+/- s.e.m.), and the second elongating the substrate disulphide bond by 0.17 +/- 0.02 A (+/- s.e.m.). These results support the view that the Trx active site regulates the geometry of the participating sulphur atoms with sub-?ngstr?m precision to achieve efficient catalysis. Our results indicate that substrate conformational changes may be important in the regulation of Trx activity under conditions of oxidative stress and mechanical injury, such as those experienced in cardiovascular disease. Furthermore, single-molecule atomic force microscopy techniques, as shown here, can probe dynamic rearrangements within an enzyme's active site during catalysis that cannot be resolved with any other current structural biological technique.     

Disulphide-isomerase-enabled shedding of tumour-associated NKG2D ligands

Tumour-associated ligands of the activating NKG2D (natural killer group 2, member D; also called KLRK1) receptor-which are induced by genotoxic or cellular stress-trigger activation of natural killer cells and co-stimulation of effector T cells, and may thus promote resistance to cancer. However, many progressing tumours in humans counter this anti-tumour activity by shedding the soluble major histocompatibility complex class-I-related ligand MICA, which induces internalization and degradation of NKG2D and stimulates population expansions of normally rare NKG2D+CD4+ T cells with negative regulatory functions. Here we show that on the surface of tumour cells, MICA associates with endoplasmic reticulum protein 5 (ERp5; also called PDIA6 or P5), which, similar to protein disulphide isomerase, usually assists in the folding of nascent proteins inside cells. Pharmacological inhibition of thioreductase activity and ERp5 gene silencing revealed that cell-surface ERp5 function is required for MICA shedding. ERp5 and membrane-anchored MICA form transitory mixed disulphide complexes from which soluble MICA is released after proteolytic cleavage near the cell membrane. Reduction of the seemingly inaccessible disulphide bond in the membrane-proximal alpha3 domain of MICA must involve a large conformational change that enables proteolytic cleavage. These results uncover a molecular mechanism whereby domain-specific deconstruction regulates MICA protein shedding, thereby promoting tumour immune evasion, and identify surface ERp5 as a strategic target for therapeutic intervention.     

Model for signal sequence recognition from amino-acid sequence of 54K subunit of signal recognition particle

Protein targeting to the endoplasmic reticulum in mammalian cells is catalysed by signal recognition particle (SRP). Cross-linking experiments have shown that the subunit of relative molecular mass 54,000 (Mr 54K; SRP54) interacts directly with signal sequences as they emerge from the ribosome. Here we present the sequence of a complementary DNA clone of SRP54 which predicts a protein that contains a putative GTP-binding domain and an unusually methionine-rich domain. The properties of this latter domain suggest that it contains the signal sequence binding site. A previously uncharacterized Escherichia coli protein has strong homology to both domains. Closely homologous GTP-binding domains are also found in the alpha-subunit of the SRP receptor (SR alpha, docking protein) in the endoplasmic reticulum membrane and in a second E. coli protein, ftsY, which resembles SR alpha. Recent work has shown that SR alpha is a GTP-binding protein and that GTP is required for the release of SRP from the signal sequence and the ribosome on targeting to the endoplasmic reticulum membrane. We propose that SRP54 and SR alpha use GTP in sequential steps of the targeting reaction and that essential features of such a pathway are conserved from bacteria to mammals.     

Signal-sequence recognition by an Escherichia coli ribonucleoprotein complex.

Hydrophobic signal-sequences direct the transfer of secretory proteins across the inner membrane of prokaryotes and the endoplasmic reticulum membranes of eukaryotes. In mammalian cells, signal-sequences are recognized by the 54K protein (M(r) 54,000) of the signal recognition particle (SRP) which is believed to hold the nascent chain in a translocation-competent conformation until it contacts the endoplasmic reticulum membrane. The SRP consists of a 7S RNA and six different polypeptides. The 7S RNA and the 54K signal-sequence-binding protein (SRP54) of mammalian SRP exhibit strong sequence similarity to the 4.5S RNA and P48 protein (Ffh) of Escherichia coli which form a ribonucleoprotein particle. Depletion of 4.5S RNA or overproduction of P48 causes the accumulation of the beta-lactamase precursor, although not of other secretory proteins. Whether 4.5S RNA and P48 are part of an SRP-like complex with a role in protein export is controversial. Here we show that the P48/4.5S RNA ribonucleoprotein complex interacts specifically with the signal sequence of a nascent secretory protein and therefore is a signal recognition particle.     

Protein-disulphide isomerase and prolyl isomerase act differently and independently as catalysts of protein folding

Two enzymes are now known that catalyse slow steps in protein folding. Peptidyl-prolyl cis-trans isomerase catalyses the cis-trans isomerization of Xaa-Pro peptide bonds in oligopeptides and during the refolding of several proteins. The other enzyme, protein-disulphide isomerase, accelerates the reactivation of reduced proteins, presumably by catalysis of thiol-disulphide exchange reactions. Recent evidence indicates that the beta-subunit of prolyl 4-hydroxylase, an enzyme involved in collagen biosynthesis, is identical with disulphide isomerase. On the basis of this important finding, it was suggested that disulphide isomerase accelerates protein folding, not by 'reshuffling' incorrect disulphide bonds, but in the same way as prolyl isomerase by catalysing proline isomerization which is known to be important for the folding of collagen and other proteins. Here we show that the catalytic activities of these two enzymes are different. Disulphide isomerase accelerates the reformation of native disulphide bonds during protein reoxidation. We find no evidence that this enzyme can catalyse the isomerization of proline peptide bonds, a reaction efficiently accelerated by prolyl isomerase. When both enzymes are present simultaneously during protein folding, they act independently of one another.     

Substantial increase of protein stability by multiple disulphide bonds

Disulphide bonds can significantly stabilize the native structures of proteins. The effect is presumed to be due mainly to a decrease in the configurational chain entropy of the unfolded polypeptide. In phage T4 lysozyme, a disulphide-free enzyme, engineered disulphide mutants that crosslink residues 3-97, 9-164 and 21-142 are significantly more stable than the wild-type protein. To investigate the effect of multiple-disulphide bonds on protein stability, mutants were constructed in which two or three stabilizing disulphide bridges were combined in the same protein. Reversible thermal denaturation shows that the increase in melting temperature resulting from the individual disulphide bonds is approximately additive. The triple-disulphide variant unfolds at a temperature 23.4 degrees C higher than wild-type lysozyme. The results demonstrate that a combination of disulphide bonds, each of which contributes to stability, can achieve substantial overall improvement in the stability of a protein.     

Sequence and topology of a model intracellular membrane protein, E1 glycoprotein, from a coronavirus

In the eukaryotic cell, both secreted and plasma membrane proteins are synthesized at the endoplasmic reticulum, then transported, via the Golgi complex, to the cell surface. Each of the compartments of this transport pathway carries out particular metabolic functions, and therefore presumably contains a distinct complement of membrane proteins. Thus, mechanisms must exist for localizing such proteins to their respective destinations. However, a major obstacle to the study of such mechanisms is that the isolation and detailed analysis of such internal membrane proteins pose formidable technical problems. We have therefore used the E1 glycoprotein from coronavirus MHV-A59 as a viral model for this class of protein. Here we present the primary structure of the protein, determined by analysis of cDNA clones prepared from viral mRNA. In combination with a previous study of its assembly into the endoplasmic reticulum membrane, the sequence reveals several unusual features of the protein which may be related to its intracellular localization.     

Identification by anti-idiotype antibodies of an intracellular membrane protein that recognizes a mammalian endoplasmic reticulum retention signal

Monoclonal antibodies were raised against antibodies to distinct carboxy-terminal KDEL sequences of two soluble, resident endoplasmic reticulum proteins. These anti-idiotype reagents recognize an intrinsic membrane protein with characteristics expected of a receptor responsible for the recognition and return of resident proteins to the endoplasmic reticulum.     

X-ray structure of a bacterial oligosaccharyltransferase

Asparagine-linked glycosylation is a post-translational modification of proteins containing the conserved sequence motif Asn-X-Ser/Thr. The attachment of oligosaccharides is implicated in diverse processes such as protein folding and quality control, organism development or host-pathogen interactions. The reaction is catalysed by oligosaccharyltransferase (OST), a membrane protein complex located in the endoplasmic reticulum. The central, catalytic enzyme of OST is the STT3 subunit, which has homologues in bacteria and archaea. Here we report the X-ray structure of a bacterial OST, the PglB protein of Campylobacter lari, in complex with an acceptor peptide. The structure defines the fold of STT3 proteins and provides insight into glycosylation sequon recognition and amide nitrogen activation, both of which are prerequisites for the formation of the N-glycosidic linkage. We also identified and validated catalytically important, acidic amino acid residues. Our results provide the molecular basis for understanding the mechanism of N-linked glycosylation.     

Signal sequence directs localized secretion of bacterial surface proteins

All living cells require specific mechanisms that target proteins to the cell surface. In eukaryotes, the first part of this process involves recognition in the endoplasmic reticulum of amino-terminal signal sequences and translocation through Sec translocons, whereas subsequent targeting to different surface locations is promoted by internal sorting signals. In bacteria, N-terminal signal sequences promote translocation across the cytoplasmic membrane, which surrounds the entire cell, but some proteins are nevertheless secreted in one part of the cell by poorly understood mechanisms. Here we analyse localized secretion in the Gram-positive pathogen Streptococcus pyogenes, and show that the signal sequences of two surface proteins, M protein and protein F (PrtF), direct secretion to different subcellular regions. The signal sequence of M protein promotes secretion at the division septum, whereas that of PrtF preferentially promotes secretion at the old pole. Our work therefore shows that a signal sequence may contain information that directs the secretion of a protein to one subcellular region, in addition to its classical role in promoting secretion. This finding identifies a new level of complexity in protein translocation and emphasizes the potential of bacterial systems for the analysis of fundamental cell-biological problems.     

Live imaging of yeast Golgi cisternal maturation

There is a debate over how protein trafficking is performed through the Golgi apparatus. In the secretory pathway, secretory proteins that are synthesized in the endoplasmic reticulum enter the early compartment of the Golgi apparatus called cis cisternae, undergo various modifications and processing, and then leave for the plasma membrane from the late (trans) cisternae. The cargo proteins must traverse the Golgi apparatus in the cis-to-trans direction. Two typical models propose either vesicular transport or cisternal progression and maturation for this process. The vesicular transport model predicts that Golgi cisternae are distinct stable compartments connected by vesicular traffic, whereas the cisternal maturation model predicts that cisternae are transient structures that form de novo, mature from cis to trans, and then dissipate. Technical progress in live-cell imaging has long been awaited to address this problem. Here we show, by the use of high-speed three-dimensional confocal microscopy, that yeast Golgi cisternae do change the distribution of resident membrane proteins from the cis nature to the trans over time, as proposed by the maturation model, in a very dynamic way.     

Aggregation and vesiculation of membrane proteins by curvature-mediated interactions

Membrane remodelling plays an important role in cellular tasks such as endocytosis, vesiculation and protein sorting, and in the biogenesis of organelles such as the endoplasmic reticulum or the Golgi apparatus. It is well established that the remodelling process is aided by specialized proteins that can sense as well as create membrane curvature, and trigger tubulation when added to synthetic liposomes. Because the energy needed for such large-scale changes in membrane geometry significantly exceeds the binding energy between individual proteins and between protein and membrane, cooperative action is essential. It has recently been suggested that curvature-mediated attractive interactions could aid cooperation and complement the effects of specific binding events on membrane remodelling. But it is difficult to experimentally isolate curvature-mediated interactions from direct attractions between proteins. Moreover, approximate theories predict repulsion between isotropically curving proteins. Here we use coarse-grained membrane simulations to show that curvature-inducing model proteins adsorbed on lipid bilayer membranes can experience attractive interactions that arise purely as a result of membrane curvature. We find that once a minimal local bending is realized, the effect robustly drives protein cluster formation and subsequent transformation into vesicles with radii that correlate with the local curvature imprint. Owing to its universal nature, curvature-mediated attraction can operate even between proteins lacking any specific interactions, such as newly synthesized and still immature membrane proteins in the endoplasmic reticulum.     

T gamma protein is expressed on murine fetal thymocytes as a disulphide-linked heterodimer

During the search for genes coding for the mouse alpha and beta subunits of the antigen-specific receptor of mouse T cells we encountered a third gene, subsequently designated gamma. This gene has many properties in common with the alpha and beta genes, somatic assembly from gene segments that resemble the gene segments for immunoglobulin variable (V), joining (J) and constant (C) regions; rearrangement and expression in T cells and not in B cells; low but distinct sequence homology to immunoglobulin V, J and C regions; other sequences that are reminiscent of the transmembrane and intracytoplasmic regions of integral membrane proteins; and a cysteine residue at the position expected for a disulphide bond linking two subunits of a dimeric membrane protein. Despite these similarities the gamma gene also shows some interesting unique features. These include a relatively limited repertoire of the germ-line gene segments, more pronounced expression at the RNA level in immature T cells such as fetal thymocytes and an apparent absence of in-frame RNA in some functional, alpha beta heterodimer-bearing T cells or cultured T clones and hybridomas. To understand the function of the putative gamma protein it is essential to define the cell population that expresses this protein. To this end we produced a fusion protein composed of Escherichia coli beta-galactosidase and the gamma-chain (hereafter referred to a beta-gal-gamma) using the phage expression vector lambda gt11 and raised rabbit antisera against the gamma determinants. Using the purified anti-gamma antibody we detected a polypeptide chain of relative molecular mass 35,000 (Mr 35K) on the surface of 16-day old fetal thymocytes. The gamma-chain is linked by a disulphide bridge to another component of 45K. No such heterodimer was detected on the surface of a cytotoxic T lymphocyte (CTL) clone 2C from which an in-phase gamma cDNA clone was originally isolated.     

Assembly of yeast Sec proteins involved in translocation into the endoplasmic reticulum into a membrane-bound multisubunit complex

Secretory-protein translocation into the endoplasmic reticulum (ER) is thought to be catalysed by integral membrane proteins. Genetic selections uncovered three Saccharomyces cerevisiae genes (SEC61, SEC62 and SEC63), mutations in which block import of precursor proteins into the ER lumen in vivo and in vitro. The DNA sequences of SEC62 and SEC63 predict multispanning membrane proteins, and biochemical characterization of the SEC62 protein (Sec62) confirms that it is an integral ER membrane protein. Here we show that Sec61, Sec62 and Sec63 are assembled with two additional proteins into a multisubunit membrane-associated complex. These results confirm previous predictions, based upon genetic interactions between the SEC genes, that Sec61, Sec62 and Sec63 act together to facilitate protein translocation into the ER.