Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

ERp57 is involved in the oxidative folding of the low-density lipoprotein receptor in the endoplasmic reticulum

ERp57 is involved in the oxidative folding of the low-density lipoprotein receptor in the... Volume 2 † Number 1 † March 2009 10.1093/biohorizons/hzp003 ......................................................................................................................................................................................................................................... Research article ERp57 is involved in the oxidative folding of the low-density lipoprotein receptor in the endoplasmic reticulum Jamie-Lee Berry and Neil J. Bulleid* Faculty of Life Sciences, The University of Manchester, Manchester M13 9PT, UK. * Corresponding author: The Michael Smith Building, The Faculty of Life Sciences, The University of Manchester, Oxford Road, Manchester M13 9PT, UK. Tel: þ44 1612755103. Email: neil.bulleid@manchester.ac.uk Supervisor: Professor Neil J. Bulleid, Faculty of Life Sciences, The University of Manchester, Manchester M13 9PT, UK. ........................................................................................................................................................................................................................................ This work explores the role of the thiol-oxidoreductase ERp57 in the post-translational oxidative folding of the low-density lipoprotein receptor (LDL-R), a cell-surface glycoprotein responsible for the uptake of cholesterol from plasma. The LDL-R provides a general model to analyse oxidative folding of multi-domain proteins in the endoplasmic reticulum; yet its folding pathway is also of specific interest as a high proportion of mutations in disulphide-rich domains of the protein are evident in familial hypercholesterolemia. Previous studies have suggested that the LDL-R forms a set of distinct non-native disulphide intermediates during folding, which are extensively isomer- ized prior to secretion of the native conformer. In addition, ERp57 has been suggested to be predominantly reduced in vivo and to form a mixed disulphide with the LDL-R. In this study, the LDL-R was expressed in both wild-type cells and those lacking the thiol-oxidoreduc- tase ERp57 under conditions that prevent disulphide formation. The protein was then allowed to fold under oxidizing conditions, and samples taken at various timepoints. The electrophoretic mobility of folding intermediates from knock-out cells was compared with that of wild-type cells. The results show that dissimilar disulphide intermediates form between the two cell types, particularly during early stages of folding. A mutant form of ERp57, able to form but unable to resolve mixed disulphides, was also found to form mixed dis- ulphides with the LDL-R. The results signify the requirement for ERp57 in oxidative folding of the LDL-R and also suggest that non- native disulphide intermediates may be central to the process of multi-domain protein folding. Key words: ERp57, disulphides, oxidoreductase, protein folding. ........................................................................................................................................................................................................................................ generally buffered by a balanced ratio of reduced to oxidized Introduction glutathione (GSH:GSSG), among other sources of oxidizing The formation of disulphides in eukaryotic cells occurs in the equivalents from parallel oxidation pathways that have been endoplasmic reticulum (ER). Conditions in the ER are highly proposed to involve Ero1, Erv2 and flavin-containing optimized for the formation of disulphides, owing first to its mono-oxidase. There is also much conflicting evidence for oxidizing environment relative to the cytosol, but also to the the distinct roles of oxidoreductases as oxidases or iso- presence of a number of thiol-oxidoreductases—a broad merases in vivo. One protein, the low-density lipoprotein group of folding enzymes possessing an active site analogous receptor (LDL-R), provides a good model on which to to that in the cytosolic reductase thioredoxin. The enzymes study the action of oxidoreductases in cells, due to the dis- can catalyse disulphide bond formation; however, these covery of an unusual set of disulphide intermediates during enzymes can also reduce substrates allowing for the isomer- its folding pathway. ization of incorrectly paired cysteines during protein The LDL-R is a multi-domain, cell-surface membrane folding. Figure 1 depicts some of the main players involved glycoprotein. Various classes of mutations in the gene in maintenance of an oxidizing ER environment, based on (LDL-R) are evident in familial hypercholesterolemia (FH), their known functions in vivo. The ER environment is several of which instigate the misfolding of disulphide-rich ......................................................................................................................................................................................................................................... 2009 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. 13 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 1. Possible mechanisms by which the endoplasmic reticulum (ER) may maintain an environment optimized for disulphide bond formation in pro- teins. Disulphides are introduced into nascent chains soon upon entry into the ER, most likely by a thiol-oxidoreductase, such as protein disulphide iso- merase (PDI). PDI may be maintained in its oxidized form by the sulfhydryl oxidase Ero1, with oxygen as the final electron acceptor. If the substrate forms native disulphides, it may be released to further stages of secretion. If not, the non-native protein may be reduced again by a member of the PDI family, or alternatively passed on to the calnexin/calreticulin cycle for isomerization to native disulphides in the case of N-glycosylated substrates. As ERp57 is associ- ated with calnexin/calreticulin, it functions to unscramble non-native disulphides. domains in the receptor structure. For example, the struc- heavily disulphide-bonded clients of the calnexin/calreticulin ture includes three domains with epidermal growth factor cycle. It appears also to exhibit preference for a subset of (EGF) homology, each forming three intra-domain disul- substrates that share common structural domains, such as phides, and within which 54% of mutations causing FH the EGF-like domains in the LDL-R. ERp57 has also 8, 9 13 are located. Characterized mutations of LDL-R affecting been shown to be reduced at steady state in vivo, further the correct formation of disulphides have been implicated supporting its role as an isomerase. in FH, and these illustrate the importance of this process The aim of this study was to define a role for ERp57 in the for structural and functional integrity. post-translational folding of the LDL-R, by comparing the It has been suggested that gene-fusion events in evolution set of folding intermediates observed in wild-type and 2/2 have led to a situation where modules in multi-domain pro- ERp57 cells according to their electrophoretic mobility. teins become immediately foldable, independent of their It was predicted that in the absence of ERp57, the protein neighbours. Indeed, vectorial folding is a logical scenario would form a non-native disulphide intermediate from which one would expect to make the process more efficient. which it would be unable to recover. The LDL-R coding For the LDL-R at least, findings to the contrary raise import- sequence was ligated into the pSPUTK vector downstream ant questions about the other factors involved in the folding of the b-globin enhancer to improve the expression levels. pathway, which may be responsible for the extensive isomer- It was then allowed to translate fully in both cell types ization of disulphides. One enzyme, ERp57, has been under reducing conditions, before folding was followed in shown previously to form a mixed disulphide with the an oxidative environment. Marked differences were observed LDL-R. ERp57 is a thiol-oxidoreductase for N-glycosylated, during the early stages of folding, and the protein ultimately ......................................................................................................................................................................................................................................... 14 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... reached a conformation in each cell type with a distinct manufacturer’s protocol. The purified DNA and 5 mgof hydrodynamic volume. These results build on previous pSPUTK were then digested with XbaI and KpnI. work to provide insight into multi-domain protein folding Restriction digests were performed in a total aqueous and evidence for the role of ERp57 as an isomerase. volume of 50 ml, containing 20 units of each enzyme and provided ‘Buffer L’ to 2 (Promega, UK). Digestion pro- ceeded for 4 h at 378C. The cut PCR product and plasmid Materials and Methods were isolated by electrophoresis on a 1% agarose gel, and DNA was extracted using the QIAGEN II Gel Extraction Culture and Preparation of Semi-intact Cells Kit (Qiagen) protocol. Concentrations of eluted DNA were Cells were cultured in Dulbecco’s Modification of Eagle’s measured by nano-spectrophotometry. Ligation reactions Medium with a 10% (v/v) foetal calf serum and 1% (v/v) were each performed in a total aqueous volume of 20 ml, L-glutamine supplement. They were incubated at 378C and with 18 units of T4 DNA ligase and the provided buffer 5% CO to 70–90% confluence. Semi-intact cells were pre- (1:10 dilution) (Promega). The amount of DNA was varied pared from cells grown in culture so that the folding of pro- in each reaction mixture, between 100 and 160 ng. Control teins translated in vitro could be followed in an environment experiments were conducted omitting either DNA insert, equivalent to the intact cell. Cells were made permeable, and or T4 DNA ligase, from the reaction. endogenous mRNA was degraded according to a previously described method. Cell lines used were HT1080 human Transformation of Ligation Products into fibroblasts (ATCC, Maryland, USA), wild-type murine Escherichia coli cells 2/2 fibroblasts (MF), ERp57 MF and V5-ERp57 (C2,7A) Escherichia coli XL1 Blue cells were inoculated into 200 ml HT1080. In the latter cell type, ERp57 has a V5 tag and Luria Bertani (LB) broth and grown up to OD 0.5–0.6. its active site cysteine is mutated to alanine, disabling its Cells were centrifuged at 4000 g for 10 min at 48C and ability to resolve the mixed disulphides it forms with re-suspended to 100 ml in 10% glycerol (ice-cold). Fifty substrate. microlitres of cell culture were combined with 2 mlof each ligation reaction in E. coli Pulser cuvettes and chilled on Analysis of Samples by Electrophoresis ice for 5 min. Cells were electroporated, added to 250 mlof Protein samples in this study were analysed by SDS–PAGE, at LB broth and shaken for 1 h at 378C. Cells from each ligation 22 mA per gel. Proteins were resolved in 7.5% polyacrylamide reaction were plated onto separate LB and ampicillin gel. Gels were fixed in 10% (v/v) acetic acid and 10% (v/v) (100 mgml ) agar plates and left overnight. Single colonies methanol and dried onto filter paper in a vacuum for 1 h at from the non-control ligations were picked and inoculated 808C. Gels were exposed to Kodak Biomax MR Film (GRI, into 30 ml fresh LB broth with ampicillin (100 mgml ). UK) for a period of 1 week for imaging. Cell pellets from centrifugation were re-suspended in 100 ml GTE (50 mM glucose, 1 mM EDTA, 25 mM Tris– Cloning and Modification of LDL-R HCl, pH 8.0) containing RNase (100 mg ml ). Cultures The LDL coding sequence was amplified from pcDNA 3.1 were treated with 200 ml NaOH–SDS [0.2 M NaOH, 1% (Invitrogen, UK) by several cycles of the polymerase chain (w/v) SDS] and 150 ml potassium acetate [29% (v/v) reaction (PCR) and modified by the following primers. acetic acid and KOH to pH 4.8]. Supernatant was retained Forward: 5 GAACTCTCTAGAGCCACCATGGGGCCC after centrifugation, and plasmids were purified by washing 0 0 TGGGGC3 ; reverse: 5 GCACGCGGTACCTCACGCCAC once with phenol and twice with chloroform. GTCATC3 . The primers incorporated the appropriate TNT Lysate (Promega, UK) Translation XbaI and KpnI restriction enzyme recognition sites to enable ligation into the pSPUTK vector. Each PCR pro- A coupled transcription and translation of LDL-R in each of ceeded in a total aqueous volume of 100 ml containing 5 the circular vectors (pcDNA 3.1 and pSPUTK) was set up units of Taq polymerase with the provided buffer (1) using the TNT Lysate Translation System (Promega) accord- (Bioline, UK). Deoxy-nucleotides (dATP, dCTP, dTTP, ing to the manufacturer’s protocol. This minimizes any dGTP; Bioline) were each added to a final concentration experimental error that may be encountered when perform- of 0.25 mM, and primers to a concentration of 1 mM. ing separate reactions. Products were analysed by SDS– Reactions were performed containing both 5 and 10 mgof PAGE electrophoresis, and pixel density of the image was LDL-R in pcDNA 3.1. These were denatured for 1 min at measured using AIDA Image Analyzer software. 948C, annealed at 558C for 1 min and allowed to extend at Transcriptions in vitro 728C for 30 cycles. Controls were performed both in the absence of polymerase and in the absence of DNA. The pSPUTK vector was cut with HpaI, and linear DNA was PCR products were washed using DNA Clean and transcribed according to a previous method, using 30 units Concentrator Kit (Qiagen, Germany) according to the of SP6 polymerase per 50 ml of reaction mixture. Reagents ......................................................................................................................................................................................................................................... 15 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... were from Promega. RNA transcripts were purified by 25 mM NEM. Samples were pre-cleared to prevent non- washing once with phenol and twice with chloroform and specific binding by adding Protein A Sepharose (PAS) beads RNA was precipitated in ethanol for 60 min with 0.3 M to 1%, and incubating at 48C for 30 min with constant agi- sodium acetate (pH 5.2) at 2208C. The mRNA was tation to keep the beads suspended. After 30 min, samples re-suspended in DEPC-treated water. were centrifuged for 5 min, the supernatant was removed and PAS beads were added again to 1%, this time coupled Translations in vitro with polyclonal anti-LDL-R antibody 121 (1:300 dilution). The mRNA was translated with 16.5 ml of Rabbit Reticulocyte Samples were incubated at 48C overnight to allow specific Lysate (Flexilysate, Promega) per 25 ml translation, containing binding of the LDL-R to the antibody. The beads were iso- 1 ml of mRNA, 20 mM amino acids minus methionine lated by centrifugation and washed three times in lysis (Bioline), 20 mM potassium chloride and 10 mCi EasyTag buffer, before re-suspension in 15 ml5 SDS loading EXPRESS-35 Protein labelling mix [ S] (NEN/PerkinElmer). buffer. Samples were boiled for 2–3 min at 1008C and agi- Translations were performed either in the presence or absence tated to release the LDL-R into solution. Beads were of 10 selectively permeabilized (SP) cells and incubated settled by centrifugation, and the supernatants were analysed for 1 h at 308C. Samples not containing SP cells were placed by SDS–PAGE under non-reducing conditions. Antibody on ice, reduced with 5 mM dithiothreitol (DTT) and was provided by Ineke Braakman, University of Utrecht, boiled with equal volume of 5 SDS sample buffer for 3 min Netherlands. at 1008C before analysis by SDS–PAGE. Otherwise, translation mixtures were centrifuged to terminate translation and Analysis of Mixed Disulphides isolate the SP cells, which were then washed in KHM Four translations were conducted under non-reducing con- (110 mM potassium acetate, 2 mM magnesium acetate, ditions, two of which were in the presence of SP wild-type 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) HT1080 cells and two in the presence of SP V5-ERp57 to remove the lysate. Cells were then re-suspended in 10 ml, (C2,7A) cells. After 1 h, NEM was added to each sample 5 SDS buffer (156 mM Tris pH 6.8, 7.5% SDS, 10% glycerol, to 25 mM to prevent further thiol exchange. One of the 0.02% bromophenol blue), boiled and analysed by SDS– samples from each of the cell types was immunoisolated, PAGE. The LDL-R products from the post-translational using anti-V5 antibody (1:300 dilution) (Invitrogen) conju- folding assay were immunoisolated before SDS–PAGE. gated to Protein G Sepharose beads. The other two samples were washed with KHM and boiled in 5 SDS Post-translational Folding Assay sample buffer without immunoisolation. Samples were ana- Translation mixtures were set up to a 25 ml volume per time- lysed under non-reducing conditions by SDS–PAGE. point, including SP cells. Seven timepoints were taken at 0, 1, 5, 15, 30, 60 and 90 min from one total translation mixture. Prior to the 0 timepoint, the total mixture was allowed to Results and Discussion translate for 30 min under reducing conditions in the pre- LDL-R Expression is Optimized in pSPUTK sence of 5 mM DTT. This should enable full translation of adequate levels of the polypeptide in a completely reduced The LDL-R coding sequence was modified by oligonucleo- form. After 30 min, oxidized glutathione was added to a tide primers to enable its ligation into the mammalian final concentration of 15 mM, in order to recreate the oxidiz- pSPUTK vector, downstream of the b-globin enhancer. In ing environment of the ER and to allow folding of the poly- addition, downstream of the XbaI consensus site, a Kozak peptide under conditions reflecting those in vivo. To cease ribosomal binding region was included in the primer before the reaction at each timepoint, a 25 ml aliquot was taken the translation initiation codon. and placed on ice, then N-ethylmaleimide (NEM) was Figure 2A shows analysis of translation products by SDS– added immediately to a concentration of 25 mM to prevent PAGE. Coupling transcription and translation reactions further disulphide exchange by the transfer of alkyl groups minimize experimental error and loss of product during puri- to cysteine sulphydryls. This experiment was conducted fication. The LDL-R formed a single major translation twice: once with wild-type MF cells and once with product when expressed from pcDNA 3.1 and pSPUTK. 2/2 ERp57 MF cells. The resulting image was analysed by 2D Densitometry using AIDA Image Analyzer software (Fig. 2B) and trans- Immunoisolation of the LDL-R lation was shown to be improved by 23% after cloning Cells from each translation were isolated by centrifugation, and ligation of LDL-R downstream of the pSPUTK washed in an excess of KHM buffer and then isolated for a b-globin enhancer. Translation of LDL-R from pcDNA 3.1 second time. Supernatant containing any remaining trans- minus the enhancer appeared to yield less protein product. lation mixture was extracted and discarded. Cell pellets This increase can also be attributed to the 5 sequence were then re-suspended in 0.5 ml lysis buffer containing included in the modified gene. ......................................................................................................................................................................................................................................... 16 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... than the conformers observed between 1 and 10 min. The product at 90 min does not appear completely reduced, yet migrates at a rate closer to the reduced form than the earlier folding intermediates. These results indicate that the protein folds to a compact form early in the time course, but then becomes less compact, probably due to a rearrange- ment in the disulphide bonds. From previous published results, and those in this study, it is clear that the LDL-R folds via a non-native disulphide intermediate. It should be noted, however, that the shift in electrophoretic mobility towards that of the ‘native’ confor- mer in the MF cells is less pronounced than previously observed in HeLa cells. Although this may be purely due to differences in activity between the cell types, it should be kept in mind when drawing conclusions from these results. 2/2 In the ERp57 cells, LDL-R appears to fold similarly to the wild-type. It was predicted that the receptor would be unable to recover from a non-native disulphide intermediate without the oxidoreductase present. This seems not strictly to be the case, as the results suggest that isomerization of dis- ulphides is taking place in the knock-outs (Fig. 3A, right- hand panel). In fact, the results shown in Figure 3A do not suggest any obvious defect in recovery from the non-native intermediate with ERp57 absent. However, the products do appear to form a slightly different pattern on each gel, suggesting differences in disulphide intermediates between the two cell types. This was confirmed by the results shown in Figure 3B. Here the folding intermediates from 2/2 the wild-type and ERp57 cells were compared directly Figure 2. Translation of LDL-R in vitro and quantification of protein levels. as the samples from each timepoint were run adjacent to (A) Coupled transcription and translations were conducted from LDL-R in pcDNA 3.1 (lane 1) and pSPUTK (lane 2). (B) Levels of protein from one another on the same gel. There is no difference in mobi- coupled reactions were compared with 2D Densitometry and the results lity between the reduced forms of the LDL-R in each cell are shown in the bar chart. Expression of the coding sequence from type at 0 min; however, intermediates formed between 1 pSPUTK increased protein levels by 23%. and 5 min in wild-type cells decrease more dramatically in hydrodynamic volume, compared with samples from The LDL-R Folds via a Non-native Disulphide 2/2 ERp57 cells. Interestingly, the intermediate reached at Intermediate 5 min in the wild-type cells has formed a different, more Figure 3A shows the post-translational folding of the LDL-R compact, disulphide intermediate to the species formed 2/2 2/2 in wild-type and ERp57 MF cells. In the wild-type, at with the ERp57 cells. Therefore, it may be proposed 0 min the protein appears completely reduced, forming a that alternative non-native disulphides are formed during concise band on the gel. Upon the addition of oxidized glu- early stages of folding without ERp57. This suggests a role tathione, the protein migrates a greater distance through the for ERp57 in the formation of non-native disulphides as gel and the product becomes much less distinct, appearing as well as their isomerization. 2/2 a smear. Between 15 and 30 min, the samples decrease in In the ERp57 cells, the LDL-R reduces slightly in electrophoretic mobility. This decrease in mobility is due to mobility after 15 min like the wild-type; however, the ulti- an increased hydrodynamic volume, i.e. the molecules are mate conformers at 60 and 90 min have a greater hydro- less compact and take up more space in solution, so dynamic volume than those from the wild-type, insinuating migrate more slowly through the gel network. Here there is that a different conformation has been reached. The disul- a small, yet observable shift in mobility towards that of the phides that form, therefore, may be central to the folding reduced form. After 90 min in an oxidizing environment, process as opposed to forming haphazardly. Indeed, it is the protein molecules occupy less hydrodynamic volume possible that non-native disulphide intermediates are an than the reduced form, yet failed to migrate as far through important step enabling the protein to eventually reach the the gel, therefore occupy a greater hydrodynamic volume, native conformation. Their formation may give rise to ......................................................................................................................................................................................................................................... 17 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 3. Post-translational folding of the LDL-R in semi-intact MF cells. (A) LDL-R was first translated with a rabbit reticulocyte lysate under reducing 2/2 conditions (5 mM DTT) for 30 min in the presence of either semi-intact wild-type (left panel), or ERp57 (right panel), MF cells. At 0 min, the LDL-R was allowed to fold by the addition of oxidized glutathione (15 mM) to permit disulphide exchange. Samples were taken at the timepoints shown, lysed in the presence of 25 mM NEM and the LDL-R was immunoisolated with a polyclonal anti-LDL-R antibody conjugated to Protein A Sepharose beads. Samples were analysed by SDS–PAGE. (B) The experiment was repeated, and samples taken from each cell type at each timepoint were analysed beside each other by SDS–PAGE under non-reducing conditions. large conformational loops, keeping parts of the protein analysis by SDS–PAGE run under non-reducing conditions apart in space that may otherwise form favourable inter- (Fig. 4, lanes 4 and 5). When the untransfected cells were actions during early stages of folding, thereby allowing the lysed and treated with the anti-V5 antibody, no protein individual domains to fold correctly. A disulphide intermedi- was precipitated (lane 2). This was expected because ate was previously observed in the tailspike endorhamnosi- HT1080 cells did not express the V5-tagged ERp57. dase from bacteriophage p22, whereas the native protein The presence of immunoisolated products with a higher contains no disulphides. Such evidence supports an evol- molecular weight than the LDL-R confirms that ERp57 utionary role for non-native disulphide intermediates forms mixed disulphides with the protein. Degradation of during protein folding. endogenous mRNA during semi-intact cell preparation, before the addition of LDL-R mRNA to the system, The LDL-R Forms a Mixed Disulphide with ERp57 ensures that the most likely interacting partner causing an in HT1080 Cells increase in molecular weight of the precipitated V5-tagged To confirm an interaction between ERp57 (57 kDa) and the ERp57 is the LDL-R. However, two distinct higher molecu- LDL-R, the mRNA was translated in either untransfected or lar weight species were immunoisolated from the V5-ERp57 a stable transfected cell line expressing a mutant form of (C2,7A) cells. The 160 kDa product corresponds well to con- V5-tagged ERp57, in which the second active site cysteine joined LDL-R and ERp57, yet the band over 220 kDa is of was replaced with alanine to trap mixed disulphides unknown origin. We do not, however, know the stoichi- (C2,7A). Without immunoisolation, a band corresponding ometry of the interaction in these mutant cells. Mixed disul- to LDL-R was observed in both cell types with an approxi- phides are too transient to detect normally, so the ratio may mate molecular weight of 97 kDa (Fig. 4, lanes 1 and 3). be 1:1 in the wild-type but it is plausible that more than one Upon immunoisolation of the V5-tagged ERp57 from cell ERp57 molecule becomes trapped with each LDL-R mol- lysates with an anti-V5 antibody, two proteins with a mol- ecule in the mutant. In the system used here, although it ecular weight greater than LDL-R were observed after can be assumed with a degree of conviction that an ......................................................................................................................................................................................................................................... 18 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... Figure 4. Mixed disulphides form between ERp57 and LDL-R. LDL-R was translated in both wild-type (lane 1) and V5-ERp57 (C2,7A) cells (lane 3) and cell lysates were analysed by SDS–PAGE. Next the LDL-R was translated in the wild-type (lane 2) and V5-ERp57 (C2,7A) (lane 4) cells and lysates were immu- noisolated with an anti-V5 antibody to isolate the tagged ERp57. Products were analysed under non-reducing conditions. Two higher molecular weight bands appear in lane 4 (clearer after 7 days image exposure in lane 5) corresponding to ERp57 and the LDL-R as a co-precipitant, trapped in a mixed disulphide state. interaction takes place, it cannot be inferred whether single has a specialized role. Therefore, there is indication that or multiple interactions will occur in vivo. they may share substrates, along with evidence to the con- trary; however, it is feasible that upon ERp57 deletion ERp72 is a good candidate replacement. 2/2 A Possible Redundancy Mechanism in ERp57 Cells Although thought to be tightly regulated in an oxidizing When the translation products from the different cell types state by Ero1 (Fig. 1), protein disulphide isomerase (PDI) were analysed adjacent to one another (Fig. 3B), it became has shown to exhibit distinct catalytic activity between evident that the wild-type samples form more compact struc- active sites of its a and a domains. Mutational studies tures throughout the folding pathway. This may be a have shown that although the a domain is an efficient knock-on effect of the alternative non-native disulphides 23 oxidase, the a domain acts as an isomerase. PDI itself that form at earlier folding stages in the presence of may be implicated in the situation without ERp57, perhaps ERp57. Alternatively, this may be because ERp57 is required in complex with BiP. It may be the case that a compensatory to isomerize the folding intermediates, a process defective in mechanism is less efficient than the calnexin/calreticulin and 2/2 2/2 the ERp57 cells. The samples from ERp57 cells, ERp57 system, hence the difference in folding intermediates; however, are ultimately forming structures with greater yet it may provide a means to an end, perhaps resulting in a hydrodynamic volume than those from earlier stages of the receptor with reduced function. folding pathway. This indicates that the protein in the knock- out cells does undergo some form of isomerization to try to recover from the non-native disulphide intermediate. Limitations and Further Direction However, the native structure is not attained, as samples at 90 min maintain distinct conformers to those from wild-type It has been reported previously that ERp57 is primarily cells. Incorrectly folded LDL-R polypeptides, as a result of reduced at steady state, so its apparent involvement in the incorrect cysteine pairing, would be retained for continuous formation of non-native disulphides may not ring true 18, 19 futile cycles of re-association with calnexin/calreticulin. in vivo. The large injection of oxidized glutathione into the Of course this would be in vain in the absence of ERp57, but system here may have caused an initial surge in levels of oxi- it is conceivable that another, less well-characterized oxido- dized ERp57, permitting thiol exchange with the LDL-R in a reductase acts in its place upon deletion. ERp72, for way that would not occur normally. The subsequent reba- example, has been shown to share 37% sequence homology lance of the system by an increase in reduced glutathione (Fig. 5) with ERp57. ERp72 has been shown to form would then increase levels of reduced ERp57, akin to the 20, 21 mixed disulphides with substrates of ERp57, yet its situation in cells, allowing it to isomerize the disulphides depletion has no effect on certain substrates where ERp57 (see Fig. 1). The system used in this study is perhaps not ......................................................................................................................................................................................................................................... 19 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 5. A schematic showing the similar domain organization observed for PDI-family proteins. Using the example of ERp72, it is unlikely that eukar- yotes have evolved without a possible compensatory mechanism in the event of ERp57 dysfunction. ERp72 shares 37% homology with ERp57, which 24 20 shares 29% identity and 56% similarity with PDI. All share striking similarity with respect to domain organization, except ERp72 has an extra catalytic a domain and ERp57 has a basic (þ) as opposed to acidic (2) tail. The b domains affiliate with substrate specificity and/or targeting to substrates, for example, ERp57 has been shown to interact with calnexin via its b domain. an accurate reflection of conditions in the ER. Indeed, it is in the wild-type cells, it would be necessary to carry out more likely that a primarily oxidized enzyme such as arche- immunoisolation using a conformational specific antibody typical PDI will have a key role in the formation of disul- which only binds to the native LDL-R conformer. The phides in LDL-R; yet it is not unreasonable to propose the results shown here, however, strongly suggest that a dis- involvement of a proportion of oxidized ERp57 in vivo.It tinct set of disulphide intermediates arise during folding should also be noted that the difference in electrophoretic of the LDL-R in the absence of ERp57. mobility of the LDL-R from two different cell types, 2/2 namely ERp57 MF and wild-type MF, may not be defini- Funding tively attributed to the presence or absence of ERp57. It may be useful to reproduce the results in a further study, using This work was funded by The Wellcome Trust (grant various cell types. #074081) and the University of Manchester. A substitute oxidoreductase may be acting in the absence of ERp57, yet may not be able to access the LDL-R in the calnexin cycle while ERp57 is present. Therefore, this may not be stearically possible in vivo in cases where ERp57 is References dysfunctional but not absent. A more suitable approach to 1. Sevier C, Kaiser C (2002) Formation and transfer of disulphide bonds in living study ERp57 function may be to introduce a loss of function cells. Nat Rev Mol Cell Biol 3: 836–847. mutation into the active site so that the protein is still phys- 2. Jessop C, Chakravarthi S, Watkins R et al. (2004) Oxidative protein folding in ically present and interacting with calnexin and calreticulin. the mammalian endoplasmic reticulum. Biochem Soc Trans 32: 655–658. With this approach, we may see the LDL-R reaching a non- 3. Chakravarthi S, Jessop C, Bulleid NJ (2006) The role of glutathione in disul- native intermediate that cannot be isomerized, confirming phide bond formation and endoplasmic reticulum-generated oxidative the requirement for ERp57 in vivo. stress. EMBO Rep 7: 271–275. To study a compensatory mechanism further, one 4. Sevier C, Cuozzo J, Vala A et al. (2001) A flavoprotein oxidase defines a new endoplasmic reticulum pathway for biosynthetic disulphide formation. Nat method would be to block entry into the calnexin cycle Cell Biol 3: 874–882. with the glucosidase inhibitor castanospermine, then look 5. Suh J, Poulsen L, Ziegler D et al. (1999) Yeast flavin-containing monooxygen- at LDL-R folding. Substrate binding to calnexin or calreti- ase generates oxidizing equivalents that control protein folding in the endo- culin is dependent upon the presence of a monoglucosy- plasmic reticulum. Proc Natl Acad Sci USA 96: 2687–2691. lated oligosaccharide side chain; castanospermine inhibits 6. Jansens A, Duijin E, Braakman I (2002) Coordinated non-vectorial folding in a the ER glucosidase activity and therefore prevents trim- newly synthesized multi-domain protein. Science 298: 2401–2403. ming of the glycan chain. If the set of folding intermedi- 7. Hobbs H, Brown M, Goldstein J (1992) Molecular genetics of the LDL recep- ates matches the wild-type, it is likely another tor gene in familial hypercholesterolemia. Hum Mutat 1: 445–466. oxidoreductase and ER retention system can act in place 8. Jeon H, Meng W, Takagi J et al. (2001) Implications for familial hypercholes- terolemia from the structure of the LDL receptor YWTD-EGF domain pair. of ERp57 and the calnexin cycle. In this study, the Nat Struct Biol 8: 499–504. LDL-R was able to enter the calnexin cycle, making a 9. Leren T, Solberg K, Rødningen OK et al. (2005) Two novel point mutations in compensatory mechanism unlikely. Also, the only oxido- the EGF precursor homology domain of the LDL receptor gene causing reductase known to conserve the residues interacting familial hypercholesterolemia. Hum Genet 96: 241–242. with calnexin is ERp72, which has been shown not to 10. Saha S, Boyd J, Werner J et al. (2001) Solution structure of the LDL receptor 2/2 21 interact with calnexin in ERp57 cells. Before we EGF-AB pair: a paradigm for the assembly of tandem calcium binding EGF can assume that the native conformer has been reached domains. Structure 9: 451–456. ......................................................................................................................................................................................................................................... 20 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... 11. Netzer W, Hartl F (1997) Recombination of protein domains facilitated by 19. Kleizen B, Braakman I (2004) Protein folding and quality control in the endo- co-translational folding in eukaryotes. Nature 388: 343–349. plasmic reticulum. Curr Opin Cell Biol 16: 343–349. 12. Jessop C, Chakravarthi S, Garbi N et al. (2007) ERp57 is essential for efficient 20. Maattanen P, Kozlov G, Gehring K et al. (2006) ERp57 and PDI: folding of glycoproteins sharing common structural domains. EMBO J 26: multifunctional protein disulfide isomerases with similar domain architec- 28–40. tures but differing substrate-partner associations. Biochem Cell Biol 84: 881–889. 13. Jessop C, Bulleid N (2004) Glutathione directly reduces an oxidoreductase in the endoplasmic reticulum of mammalian cells. J Biol Chem 279: 55341–55347. 21. Solda` T, Garbi N, Hammerling G et al. (2006) Consequences of ERp57 del- etion on oxidative folding of obligate and facultative clients of the calnexin 14. Wilson R, Allen A, Oliver J et al. (1995) The translocation, folding, assembly cycle. J Biol Chem 281: 6219–6226. and redox-dependent degradation of secretory and membrane proteins in semi-permeabilised mammalian cells. Biochemical J 307: 679–687. 22. Zhang Y, Balg E, Williams D (2006) Functions of ERp57 in the folding and assembly of major histocompatibility complex class I molecules. J Biol 15. Duffy M, Noormohammadi A, Bassegio N et al. (1998) Polyacrylamide gel- Chem 281: 14622–14631. electrophoresis separation of whole-cell proteins. In RJ Miles, RAJ Nicholas eds, Methods in Molecular Biology. Vol. 104. Totowa, NJ: Humana Press. 23. Kulp M, Frickel E, Ellgaard L et al. (2006) Domain architecture of protein- disulphide isomerase facilitates its dual role as an oxidase and an isomerase 16. Gurevich V, Pokrovskaya I, Obukhova T et al. (1991) Preparative in vitro mRNA in Ero1p-mediated disulphide formation. J Biol Chem 281: 876–884. synthesis using SP6 and T7 RNA polymerases. Anal Biochem 195: 207–213. 24. High S, Fabienne L, Russell S et al. (2000) Glycoprotein folding in the endo- 17. Robinson A, King J (1997) Disulphide-bonded intermediate on the folding plasmic reticulum: a tale of three chaperones? FEBS Lett 476: 38–41. and assembly pathway of a non-disulphide bonded protein. Nat Struct Biol 4: 450–455. 25. Russell S, Ruddock L, Salo K et al. (2004) The primary substrate binding site in the b domain of ERp57 is adapted for endoplasmic reticulum lectin 18. Brook D (1999) Introduction: molecular chaperones of the ER: their role in association. J Biol Chem 279: 18861–18869. protein folding and genetic disease. Semin Cell Dev Biol 10: 441–442. ........................................................................................................................................................................................................................................ Submitted on 30 September 2008; accepted on 20 January 2009; advance access publication 10 February 2009 ......................................................................................................................................................................................................................................... http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Bioscience Horizons Oxford University Press

ERp57 is involved in the oxidative folding of the low-density lipoprotein receptor in the endoplasmic reticulum

Berry, Jamie-Lee; Bulleid, Neil J.
Bioscience Horizons , Volume 2 (1) – Mar 10, 2009
Download PDF
9 pages

Loading next page...
 
/lp/oxford-university-press/erp57-is-involved-in-the-oxidative-folding-of-the-low-density-Af0dxkH0BW

References (27)

Publisher
Oxford University Press
Copyright
© Published by Oxford University Press.
Subject
Research articles
eISSN
1754-7431
DOI
10.1093/biohorizons/hzp003
Publisher site
See Article on Publisher Site

Abstract

Volume 2 † Number 1 † March 2009 10.1093/biohorizons/hzp003 ......................................................................................................................................................................................................................................... Research article ERp57 is involved in the oxidative folding of the low-density lipoprotein receptor in the endoplasmic reticulum Jamie-Lee Berry and Neil J. Bulleid* Faculty of Life Sciences, The University of Manchester, Manchester M13 9PT, UK. * Corresponding author: The Michael Smith Building, The Faculty of Life Sciences, The University of Manchester, Oxford Road, Manchester M13 9PT, UK. Tel: þ44 1612755103. Email: neil.bulleid@manchester.ac.uk Supervisor: Professor Neil J. Bulleid, Faculty of Life Sciences, The University of Manchester, Manchester M13 9PT, UK. ........................................................................................................................................................................................................................................ This work explores the role of the thiol-oxidoreductase ERp57 in the post-translational oxidative folding of the low-density lipoprotein receptor (LDL-R), a cell-surface glycoprotein responsible for the uptake of cholesterol from plasma. The LDL-R provides a general model to analyse oxidative folding of multi-domain proteins in the endoplasmic reticulum; yet its folding pathway is also of specific interest as a high proportion of mutations in disulphide-rich domains of the protein are evident in familial hypercholesterolemia. Previous studies have suggested that the LDL-R forms a set of distinct non-native disulphide intermediates during folding, which are extensively isomer- ized prior to secretion of the native conformer. In addition, ERp57 has been suggested to be predominantly reduced in vivo and to form a mixed disulphide with the LDL-R. In this study, the LDL-R was expressed in both wild-type cells and those lacking the thiol-oxidoreduc- tase ERp57 under conditions that prevent disulphide formation. The protein was then allowed to fold under oxidizing conditions, and samples taken at various timepoints. The electrophoretic mobility of folding intermediates from knock-out cells was compared with that of wild-type cells. The results show that dissimilar disulphide intermediates form between the two cell types, particularly during early stages of folding. A mutant form of ERp57, able to form but unable to resolve mixed disulphides, was also found to form mixed dis- ulphides with the LDL-R. The results signify the requirement for ERp57 in oxidative folding of the LDL-R and also suggest that non- native disulphide intermediates may be central to the process of multi-domain protein folding. Key words: ERp57, disulphides, oxidoreductase, protein folding. ........................................................................................................................................................................................................................................ generally buffered by a balanced ratio of reduced to oxidized Introduction glutathione (GSH:GSSG), among other sources of oxidizing The formation of disulphides in eukaryotic cells occurs in the equivalents from parallel oxidation pathways that have been endoplasmic reticulum (ER). Conditions in the ER are highly proposed to involve Ero1, Erv2 and flavin-containing optimized for the formation of disulphides, owing first to its mono-oxidase. There is also much conflicting evidence for oxidizing environment relative to the cytosol, but also to the the distinct roles of oxidoreductases as oxidases or iso- presence of a number of thiol-oxidoreductases—a broad merases in vivo. One protein, the low-density lipoprotein group of folding enzymes possessing an active site analogous receptor (LDL-R), provides a good model on which to to that in the cytosolic reductase thioredoxin. The enzymes study the action of oxidoreductases in cells, due to the dis- can catalyse disulphide bond formation; however, these covery of an unusual set of disulphide intermediates during enzymes can also reduce substrates allowing for the isomer- its folding pathway. ization of incorrectly paired cysteines during protein The LDL-R is a multi-domain, cell-surface membrane folding. Figure 1 depicts some of the main players involved glycoprotein. Various classes of mutations in the gene in maintenance of an oxidizing ER environment, based on (LDL-R) are evident in familial hypercholesterolemia (FH), their known functions in vivo. The ER environment is several of which instigate the misfolding of disulphide-rich ......................................................................................................................................................................................................................................... 2009 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. 13 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 1. Possible mechanisms by which the endoplasmic reticulum (ER) may maintain an environment optimized for disulphide bond formation in pro- teins. Disulphides are introduced into nascent chains soon upon entry into the ER, most likely by a thiol-oxidoreductase, such as protein disulphide iso- merase (PDI). PDI may be maintained in its oxidized form by the sulfhydryl oxidase Ero1, with oxygen as the final electron acceptor. If the substrate forms native disulphides, it may be released to further stages of secretion. If not, the non-native protein may be reduced again by a member of the PDI family, or alternatively passed on to the calnexin/calreticulin cycle for isomerization to native disulphides in the case of N-glycosylated substrates. As ERp57 is associ- ated with calnexin/calreticulin, it functions to unscramble non-native disulphides. domains in the receptor structure. For example, the struc- heavily disulphide-bonded clients of the calnexin/calreticulin ture includes three domains with epidermal growth factor cycle. It appears also to exhibit preference for a subset of (EGF) homology, each forming three intra-domain disul- substrates that share common structural domains, such as phides, and within which 54% of mutations causing FH the EGF-like domains in the LDL-R. ERp57 has also 8, 9 13 are located. Characterized mutations of LDL-R affecting been shown to be reduced at steady state in vivo, further the correct formation of disulphides have been implicated supporting its role as an isomerase. in FH, and these illustrate the importance of this process The aim of this study was to define a role for ERp57 in the for structural and functional integrity. post-translational folding of the LDL-R, by comparing the It has been suggested that gene-fusion events in evolution set of folding intermediates observed in wild-type and 2/2 have led to a situation where modules in multi-domain pro- ERp57 cells according to their electrophoretic mobility. teins become immediately foldable, independent of their It was predicted that in the absence of ERp57, the protein neighbours. Indeed, vectorial folding is a logical scenario would form a non-native disulphide intermediate from which one would expect to make the process more efficient. which it would be unable to recover. The LDL-R coding For the LDL-R at least, findings to the contrary raise import- sequence was ligated into the pSPUTK vector downstream ant questions about the other factors involved in the folding of the b-globin enhancer to improve the expression levels. pathway, which may be responsible for the extensive isomer- It was then allowed to translate fully in both cell types ization of disulphides. One enzyme, ERp57, has been under reducing conditions, before folding was followed in shown previously to form a mixed disulphide with the an oxidative environment. Marked differences were observed LDL-R. ERp57 is a thiol-oxidoreductase for N-glycosylated, during the early stages of folding, and the protein ultimately ......................................................................................................................................................................................................................................... 14 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... reached a conformation in each cell type with a distinct manufacturer’s protocol. The purified DNA and 5 mgof hydrodynamic volume. These results build on previous pSPUTK were then digested with XbaI and KpnI. work to provide insight into multi-domain protein folding Restriction digests were performed in a total aqueous and evidence for the role of ERp57 as an isomerase. volume of 50 ml, containing 20 units of each enzyme and provided ‘Buffer L’ to 2 (Promega, UK). Digestion pro- ceeded for 4 h at 378C. The cut PCR product and plasmid Materials and Methods were isolated by electrophoresis on a 1% agarose gel, and DNA was extracted using the QIAGEN II Gel Extraction Culture and Preparation of Semi-intact Cells Kit (Qiagen) protocol. Concentrations of eluted DNA were Cells were cultured in Dulbecco’s Modification of Eagle’s measured by nano-spectrophotometry. Ligation reactions Medium with a 10% (v/v) foetal calf serum and 1% (v/v) were each performed in a total aqueous volume of 20 ml, L-glutamine supplement. They were incubated at 378C and with 18 units of T4 DNA ligase and the provided buffer 5% CO to 70–90% confluence. Semi-intact cells were pre- (1:10 dilution) (Promega). The amount of DNA was varied pared from cells grown in culture so that the folding of pro- in each reaction mixture, between 100 and 160 ng. Control teins translated in vitro could be followed in an environment experiments were conducted omitting either DNA insert, equivalent to the intact cell. Cells were made permeable, and or T4 DNA ligase, from the reaction. endogenous mRNA was degraded according to a previously described method. Cell lines used were HT1080 human Transformation of Ligation Products into fibroblasts (ATCC, Maryland, USA), wild-type murine Escherichia coli cells 2/2 fibroblasts (MF), ERp57 MF and V5-ERp57 (C2,7A) Escherichia coli XL1 Blue cells were inoculated into 200 ml HT1080. In the latter cell type, ERp57 has a V5 tag and Luria Bertani (LB) broth and grown up to OD 0.5–0.6. its active site cysteine is mutated to alanine, disabling its Cells were centrifuged at 4000 g for 10 min at 48C and ability to resolve the mixed disulphides it forms with re-suspended to 100 ml in 10% glycerol (ice-cold). Fifty substrate. microlitres of cell culture were combined with 2 mlof each ligation reaction in E. coli Pulser cuvettes and chilled on Analysis of Samples by Electrophoresis ice for 5 min. Cells were electroporated, added to 250 mlof Protein samples in this study were analysed by SDS–PAGE, at LB broth and shaken for 1 h at 378C. Cells from each ligation 22 mA per gel. Proteins were resolved in 7.5% polyacrylamide reaction were plated onto separate LB and ampicillin gel. Gels were fixed in 10% (v/v) acetic acid and 10% (v/v) (100 mgml ) agar plates and left overnight. Single colonies methanol and dried onto filter paper in a vacuum for 1 h at from the non-control ligations were picked and inoculated 808C. Gels were exposed to Kodak Biomax MR Film (GRI, into 30 ml fresh LB broth with ampicillin (100 mgml ). UK) for a period of 1 week for imaging. Cell pellets from centrifugation were re-suspended in 100 ml GTE (50 mM glucose, 1 mM EDTA, 25 mM Tris– Cloning and Modification of LDL-R HCl, pH 8.0) containing RNase (100 mg ml ). Cultures The LDL coding sequence was amplified from pcDNA 3.1 were treated with 200 ml NaOH–SDS [0.2 M NaOH, 1% (Invitrogen, UK) by several cycles of the polymerase chain (w/v) SDS] and 150 ml potassium acetate [29% (v/v) reaction (PCR) and modified by the following primers. acetic acid and KOH to pH 4.8]. Supernatant was retained Forward: 5 GAACTCTCTAGAGCCACCATGGGGCCC after centrifugation, and plasmids were purified by washing 0 0 TGGGGC3 ; reverse: 5 GCACGCGGTACCTCACGCCAC once with phenol and twice with chloroform. GTCATC3 . The primers incorporated the appropriate TNT Lysate (Promega, UK) Translation XbaI and KpnI restriction enzyme recognition sites to enable ligation into the pSPUTK vector. Each PCR pro- A coupled transcription and translation of LDL-R in each of ceeded in a total aqueous volume of 100 ml containing 5 the circular vectors (pcDNA 3.1 and pSPUTK) was set up units of Taq polymerase with the provided buffer (1) using the TNT Lysate Translation System (Promega) accord- (Bioline, UK). Deoxy-nucleotides (dATP, dCTP, dTTP, ing to the manufacturer’s protocol. This minimizes any dGTP; Bioline) were each added to a final concentration experimental error that may be encountered when perform- of 0.25 mM, and primers to a concentration of 1 mM. ing separate reactions. Products were analysed by SDS– Reactions were performed containing both 5 and 10 mgof PAGE electrophoresis, and pixel density of the image was LDL-R in pcDNA 3.1. These were denatured for 1 min at measured using AIDA Image Analyzer software. 948C, annealed at 558C for 1 min and allowed to extend at Transcriptions in vitro 728C for 30 cycles. Controls were performed both in the absence of polymerase and in the absence of DNA. The pSPUTK vector was cut with HpaI, and linear DNA was PCR products were washed using DNA Clean and transcribed according to a previous method, using 30 units Concentrator Kit (Qiagen, Germany) according to the of SP6 polymerase per 50 ml of reaction mixture. Reagents ......................................................................................................................................................................................................................................... 15 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... were from Promega. RNA transcripts were purified by 25 mM NEM. Samples were pre-cleared to prevent non- washing once with phenol and twice with chloroform and specific binding by adding Protein A Sepharose (PAS) beads RNA was precipitated in ethanol for 60 min with 0.3 M to 1%, and incubating at 48C for 30 min with constant agi- sodium acetate (pH 5.2) at 2208C. The mRNA was tation to keep the beads suspended. After 30 min, samples re-suspended in DEPC-treated water. were centrifuged for 5 min, the supernatant was removed and PAS beads were added again to 1%, this time coupled Translations in vitro with polyclonal anti-LDL-R antibody 121 (1:300 dilution). The mRNA was translated with 16.5 ml of Rabbit Reticulocyte Samples were incubated at 48C overnight to allow specific Lysate (Flexilysate, Promega) per 25 ml translation, containing binding of the LDL-R to the antibody. The beads were iso- 1 ml of mRNA, 20 mM amino acids minus methionine lated by centrifugation and washed three times in lysis (Bioline), 20 mM potassium chloride and 10 mCi EasyTag buffer, before re-suspension in 15 ml5 SDS loading EXPRESS-35 Protein labelling mix [ S] (NEN/PerkinElmer). buffer. Samples were boiled for 2–3 min at 1008C and agi- Translations were performed either in the presence or absence tated to release the LDL-R into solution. Beads were of 10 selectively permeabilized (SP) cells and incubated settled by centrifugation, and the supernatants were analysed for 1 h at 308C. Samples not containing SP cells were placed by SDS–PAGE under non-reducing conditions. Antibody on ice, reduced with 5 mM dithiothreitol (DTT) and was provided by Ineke Braakman, University of Utrecht, boiled with equal volume of 5 SDS sample buffer for 3 min Netherlands. at 1008C before analysis by SDS–PAGE. Otherwise, translation mixtures were centrifuged to terminate translation and Analysis of Mixed Disulphides isolate the SP cells, which were then washed in KHM Four translations were conducted under non-reducing con- (110 mM potassium acetate, 2 mM magnesium acetate, ditions, two of which were in the presence of SP wild-type 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) HT1080 cells and two in the presence of SP V5-ERp57 to remove the lysate. Cells were then re-suspended in 10 ml, (C2,7A) cells. After 1 h, NEM was added to each sample 5 SDS buffer (156 mM Tris pH 6.8, 7.5% SDS, 10% glycerol, to 25 mM to prevent further thiol exchange. One of the 0.02% bromophenol blue), boiled and analysed by SDS– samples from each of the cell types was immunoisolated, PAGE. The LDL-R products from the post-translational using anti-V5 antibody (1:300 dilution) (Invitrogen) conju- folding assay were immunoisolated before SDS–PAGE. gated to Protein G Sepharose beads. The other two samples were washed with KHM and boiled in 5 SDS Post-translational Folding Assay sample buffer without immunoisolation. Samples were ana- Translation mixtures were set up to a 25 ml volume per time- lysed under non-reducing conditions by SDS–PAGE. point, including SP cells. Seven timepoints were taken at 0, 1, 5, 15, 30, 60 and 90 min from one total translation mixture. Prior to the 0 timepoint, the total mixture was allowed to Results and Discussion translate for 30 min under reducing conditions in the pre- LDL-R Expression is Optimized in pSPUTK sence of 5 mM DTT. This should enable full translation of adequate levels of the polypeptide in a completely reduced The LDL-R coding sequence was modified by oligonucleo- form. After 30 min, oxidized glutathione was added to a tide primers to enable its ligation into the mammalian final concentration of 15 mM, in order to recreate the oxidiz- pSPUTK vector, downstream of the b-globin enhancer. In ing environment of the ER and to allow folding of the poly- addition, downstream of the XbaI consensus site, a Kozak peptide under conditions reflecting those in vivo. To cease ribosomal binding region was included in the primer before the reaction at each timepoint, a 25 ml aliquot was taken the translation initiation codon. and placed on ice, then N-ethylmaleimide (NEM) was Figure 2A shows analysis of translation products by SDS– added immediately to a concentration of 25 mM to prevent PAGE. Coupling transcription and translation reactions further disulphide exchange by the transfer of alkyl groups minimize experimental error and loss of product during puri- to cysteine sulphydryls. This experiment was conducted fication. The LDL-R formed a single major translation twice: once with wild-type MF cells and once with product when expressed from pcDNA 3.1 and pSPUTK. 2/2 ERp57 MF cells. The resulting image was analysed by 2D Densitometry using AIDA Image Analyzer software (Fig. 2B) and trans- Immunoisolation of the LDL-R lation was shown to be improved by 23% after cloning Cells from each translation were isolated by centrifugation, and ligation of LDL-R downstream of the pSPUTK washed in an excess of KHM buffer and then isolated for a b-globin enhancer. Translation of LDL-R from pcDNA 3.1 second time. Supernatant containing any remaining trans- minus the enhancer appeared to yield less protein product. lation mixture was extracted and discarded. Cell pellets This increase can also be attributed to the 5 sequence were then re-suspended in 0.5 ml lysis buffer containing included in the modified gene. ......................................................................................................................................................................................................................................... 16 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... than the conformers observed between 1 and 10 min. The product at 90 min does not appear completely reduced, yet migrates at a rate closer to the reduced form than the earlier folding intermediates. These results indicate that the protein folds to a compact form early in the time course, but then becomes less compact, probably due to a rearrange- ment in the disulphide bonds. From previous published results, and those in this study, it is clear that the LDL-R folds via a non-native disulphide intermediate. It should be noted, however, that the shift in electrophoretic mobility towards that of the ‘native’ confor- mer in the MF cells is less pronounced than previously observed in HeLa cells. Although this may be purely due to differences in activity between the cell types, it should be kept in mind when drawing conclusions from these results. 2/2 In the ERp57 cells, LDL-R appears to fold similarly to the wild-type. It was predicted that the receptor would be unable to recover from a non-native disulphide intermediate without the oxidoreductase present. This seems not strictly to be the case, as the results suggest that isomerization of dis- ulphides is taking place in the knock-outs (Fig. 3A, right- hand panel). In fact, the results shown in Figure 3A do not suggest any obvious defect in recovery from the non-native intermediate with ERp57 absent. However, the products do appear to form a slightly different pattern on each gel, suggesting differences in disulphide intermediates between the two cell types. This was confirmed by the results shown in Figure 3B. Here the folding intermediates from 2/2 the wild-type and ERp57 cells were compared directly Figure 2. Translation of LDL-R in vitro and quantification of protein levels. as the samples from each timepoint were run adjacent to (A) Coupled transcription and translations were conducted from LDL-R in pcDNA 3.1 (lane 1) and pSPUTK (lane 2). (B) Levels of protein from one another on the same gel. There is no difference in mobi- coupled reactions were compared with 2D Densitometry and the results lity between the reduced forms of the LDL-R in each cell are shown in the bar chart. Expression of the coding sequence from type at 0 min; however, intermediates formed between 1 pSPUTK increased protein levels by 23%. and 5 min in wild-type cells decrease more dramatically in hydrodynamic volume, compared with samples from The LDL-R Folds via a Non-native Disulphide 2/2 ERp57 cells. Interestingly, the intermediate reached at Intermediate 5 min in the wild-type cells has formed a different, more Figure 3A shows the post-translational folding of the LDL-R compact, disulphide intermediate to the species formed 2/2 2/2 in wild-type and ERp57 MF cells. In the wild-type, at with the ERp57 cells. Therefore, it may be proposed 0 min the protein appears completely reduced, forming a that alternative non-native disulphides are formed during concise band on the gel. Upon the addition of oxidized glu- early stages of folding without ERp57. This suggests a role tathione, the protein migrates a greater distance through the for ERp57 in the formation of non-native disulphides as gel and the product becomes much less distinct, appearing as well as their isomerization. 2/2 a smear. Between 15 and 30 min, the samples decrease in In the ERp57 cells, the LDL-R reduces slightly in electrophoretic mobility. This decrease in mobility is due to mobility after 15 min like the wild-type; however, the ulti- an increased hydrodynamic volume, i.e. the molecules are mate conformers at 60 and 90 min have a greater hydro- less compact and take up more space in solution, so dynamic volume than those from the wild-type, insinuating migrate more slowly through the gel network. Here there is that a different conformation has been reached. The disul- a small, yet observable shift in mobility towards that of the phides that form, therefore, may be central to the folding reduced form. After 90 min in an oxidizing environment, process as opposed to forming haphazardly. Indeed, it is the protein molecules occupy less hydrodynamic volume possible that non-native disulphide intermediates are an than the reduced form, yet failed to migrate as far through important step enabling the protein to eventually reach the the gel, therefore occupy a greater hydrodynamic volume, native conformation. Their formation may give rise to ......................................................................................................................................................................................................................................... 17 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 3. Post-translational folding of the LDL-R in semi-intact MF cells. (A) LDL-R was first translated with a rabbit reticulocyte lysate under reducing 2/2 conditions (5 mM DTT) for 30 min in the presence of either semi-intact wild-type (left panel), or ERp57 (right panel), MF cells. At 0 min, the LDL-R was allowed to fold by the addition of oxidized glutathione (15 mM) to permit disulphide exchange. Samples were taken at the timepoints shown, lysed in the presence of 25 mM NEM and the LDL-R was immunoisolated with a polyclonal anti-LDL-R antibody conjugated to Protein A Sepharose beads. Samples were analysed by SDS–PAGE. (B) The experiment was repeated, and samples taken from each cell type at each timepoint were analysed beside each other by SDS–PAGE under non-reducing conditions. large conformational loops, keeping parts of the protein analysis by SDS–PAGE run under non-reducing conditions apart in space that may otherwise form favourable inter- (Fig. 4, lanes 4 and 5). When the untransfected cells were actions during early stages of folding, thereby allowing the lysed and treated with the anti-V5 antibody, no protein individual domains to fold correctly. A disulphide intermedi- was precipitated (lane 2). This was expected because ate was previously observed in the tailspike endorhamnosi- HT1080 cells did not express the V5-tagged ERp57. dase from bacteriophage p22, whereas the native protein The presence of immunoisolated products with a higher contains no disulphides. Such evidence supports an evol- molecular weight than the LDL-R confirms that ERp57 utionary role for non-native disulphide intermediates forms mixed disulphides with the protein. Degradation of during protein folding. endogenous mRNA during semi-intact cell preparation, before the addition of LDL-R mRNA to the system, The LDL-R Forms a Mixed Disulphide with ERp57 ensures that the most likely interacting partner causing an in HT1080 Cells increase in molecular weight of the precipitated V5-tagged To confirm an interaction between ERp57 (57 kDa) and the ERp57 is the LDL-R. However, two distinct higher molecu- LDL-R, the mRNA was translated in either untransfected or lar weight species were immunoisolated from the V5-ERp57 a stable transfected cell line expressing a mutant form of (C2,7A) cells. The 160 kDa product corresponds well to con- V5-tagged ERp57, in which the second active site cysteine joined LDL-R and ERp57, yet the band over 220 kDa is of was replaced with alanine to trap mixed disulphides unknown origin. We do not, however, know the stoichi- (C2,7A). Without immunoisolation, a band corresponding ometry of the interaction in these mutant cells. Mixed disul- to LDL-R was observed in both cell types with an approxi- phides are too transient to detect normally, so the ratio may mate molecular weight of 97 kDa (Fig. 4, lanes 1 and 3). be 1:1 in the wild-type but it is plausible that more than one Upon immunoisolation of the V5-tagged ERp57 from cell ERp57 molecule becomes trapped with each LDL-R mol- lysates with an anti-V5 antibody, two proteins with a mol- ecule in the mutant. In the system used here, although it ecular weight greater than LDL-R were observed after can be assumed with a degree of conviction that an ......................................................................................................................................................................................................................................... 18 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... Figure 4. Mixed disulphides form between ERp57 and LDL-R. LDL-R was translated in both wild-type (lane 1) and V5-ERp57 (C2,7A) cells (lane 3) and cell lysates were analysed by SDS–PAGE. Next the LDL-R was translated in the wild-type (lane 2) and V5-ERp57 (C2,7A) (lane 4) cells and lysates were immu- noisolated with an anti-V5 antibody to isolate the tagged ERp57. Products were analysed under non-reducing conditions. Two higher molecular weight bands appear in lane 4 (clearer after 7 days image exposure in lane 5) corresponding to ERp57 and the LDL-R as a co-precipitant, trapped in a mixed disulphide state. interaction takes place, it cannot be inferred whether single has a specialized role. Therefore, there is indication that or multiple interactions will occur in vivo. they may share substrates, along with evidence to the con- trary; however, it is feasible that upon ERp57 deletion ERp72 is a good candidate replacement. 2/2 A Possible Redundancy Mechanism in ERp57 Cells Although thought to be tightly regulated in an oxidizing When the translation products from the different cell types state by Ero1 (Fig. 1), protein disulphide isomerase (PDI) were analysed adjacent to one another (Fig. 3B), it became has shown to exhibit distinct catalytic activity between evident that the wild-type samples form more compact struc- active sites of its a and a domains. Mutational studies tures throughout the folding pathway. This may be a have shown that although the a domain is an efficient knock-on effect of the alternative non-native disulphides 23 oxidase, the a domain acts as an isomerase. PDI itself that form at earlier folding stages in the presence of may be implicated in the situation without ERp57, perhaps ERp57. Alternatively, this may be because ERp57 is required in complex with BiP. It may be the case that a compensatory to isomerize the folding intermediates, a process defective in mechanism is less efficient than the calnexin/calreticulin and 2/2 2/2 the ERp57 cells. The samples from ERp57 cells, ERp57 system, hence the difference in folding intermediates; however, are ultimately forming structures with greater yet it may provide a means to an end, perhaps resulting in a hydrodynamic volume than those from earlier stages of the receptor with reduced function. folding pathway. This indicates that the protein in the knock- out cells does undergo some form of isomerization to try to recover from the non-native disulphide intermediate. Limitations and Further Direction However, the native structure is not attained, as samples at 90 min maintain distinct conformers to those from wild-type It has been reported previously that ERp57 is primarily cells. Incorrectly folded LDL-R polypeptides, as a result of reduced at steady state, so its apparent involvement in the incorrect cysteine pairing, would be retained for continuous formation of non-native disulphides may not ring true 18, 19 futile cycles of re-association with calnexin/calreticulin. in vivo. The large injection of oxidized glutathione into the Of course this would be in vain in the absence of ERp57, but system here may have caused an initial surge in levels of oxi- it is conceivable that another, less well-characterized oxido- dized ERp57, permitting thiol exchange with the LDL-R in a reductase acts in its place upon deletion. ERp72, for way that would not occur normally. The subsequent reba- example, has been shown to share 37% sequence homology lance of the system by an increase in reduced glutathione (Fig. 5) with ERp57. ERp72 has been shown to form would then increase levels of reduced ERp57, akin to the 20, 21 mixed disulphides with substrates of ERp57, yet its situation in cells, allowing it to isomerize the disulphides depletion has no effect on certain substrates where ERp57 (see Fig. 1). The system used in this study is perhaps not ......................................................................................................................................................................................................................................... 19 Research article Bioscience Horizons † Volume 2 † Number 1 † March 2009 ......................................................................................................................................................................................................................................... Figure 5. A schematic showing the similar domain organization observed for PDI-family proteins. Using the example of ERp72, it is unlikely that eukar- yotes have evolved without a possible compensatory mechanism in the event of ERp57 dysfunction. ERp72 shares 37% homology with ERp57, which 24 20 shares 29% identity and 56% similarity with PDI. All share striking similarity with respect to domain organization, except ERp72 has an extra catalytic a domain and ERp57 has a basic (þ) as opposed to acidic (2) tail. The b domains affiliate with substrate specificity and/or targeting to substrates, for example, ERp57 has been shown to interact with calnexin via its b domain. an accurate reflection of conditions in the ER. Indeed, it is in the wild-type cells, it would be necessary to carry out more likely that a primarily oxidized enzyme such as arche- immunoisolation using a conformational specific antibody typical PDI will have a key role in the formation of disul- which only binds to the native LDL-R conformer. The phides in LDL-R; yet it is not unreasonable to propose the results shown here, however, strongly suggest that a dis- involvement of a proportion of oxidized ERp57 in vivo.It tinct set of disulphide intermediates arise during folding should also be noted that the difference in electrophoretic of the LDL-R in the absence of ERp57. mobility of the LDL-R from two different cell types, 2/2 namely ERp57 MF and wild-type MF, may not be defini- Funding tively attributed to the presence or absence of ERp57. It may be useful to reproduce the results in a further study, using This work was funded by The Wellcome Trust (grant various cell types. #074081) and the University of Manchester. A substitute oxidoreductase may be acting in the absence of ERp57, yet may not be able to access the LDL-R in the calnexin cycle while ERp57 is present. Therefore, this may not be stearically possible in vivo in cases where ERp57 is References dysfunctional but not absent. A more suitable approach to 1. Sevier C, Kaiser C (2002) Formation and transfer of disulphide bonds in living study ERp57 function may be to introduce a loss of function cells. Nat Rev Mol Cell Biol 3: 836–847. mutation into the active site so that the protein is still phys- 2. Jessop C, Chakravarthi S, Watkins R et al. (2004) Oxidative protein folding in ically present and interacting with calnexin and calreticulin. the mammalian endoplasmic reticulum. Biochem Soc Trans 32: 655–658. With this approach, we may see the LDL-R reaching a non- 3. Chakravarthi S, Jessop C, Bulleid NJ (2006) The role of glutathione in disul- native intermediate that cannot be isomerized, confirming phide bond formation and endoplasmic reticulum-generated oxidative the requirement for ERp57 in vivo. stress. EMBO Rep 7: 271–275. To study a compensatory mechanism further, one 4. Sevier C, Cuozzo J, Vala A et al. (2001) A flavoprotein oxidase defines a new endoplasmic reticulum pathway for biosynthetic disulphide formation. Nat method would be to block entry into the calnexin cycle Cell Biol 3: 874–882. with the glucosidase inhibitor castanospermine, then look 5. Suh J, Poulsen L, Ziegler D et al. (1999) Yeast flavin-containing monooxygen- at LDL-R folding. Substrate binding to calnexin or calreti- ase generates oxidizing equivalents that control protein folding in the endo- culin is dependent upon the presence of a monoglucosy- plasmic reticulum. Proc Natl Acad Sci USA 96: 2687–2691. lated oligosaccharide side chain; castanospermine inhibits 6. Jansens A, Duijin E, Braakman I (2002) Coordinated non-vectorial folding in a the ER glucosidase activity and therefore prevents trim- newly synthesized multi-domain protein. Science 298: 2401–2403. ming of the glycan chain. If the set of folding intermedi- 7. Hobbs H, Brown M, Goldstein J (1992) Molecular genetics of the LDL recep- ates matches the wild-type, it is likely another tor gene in familial hypercholesterolemia. Hum Mutat 1: 445–466. oxidoreductase and ER retention system can act in place 8. Jeon H, Meng W, Takagi J et al. (2001) Implications for familial hypercholes- terolemia from the structure of the LDL receptor YWTD-EGF domain pair. of ERp57 and the calnexin cycle. In this study, the Nat Struct Biol 8: 499–504. LDL-R was able to enter the calnexin cycle, making a 9. Leren T, Solberg K, Rødningen OK et al. (2005) Two novel point mutations in compensatory mechanism unlikely. Also, the only oxido- the EGF precursor homology domain of the LDL receptor gene causing reductase known to conserve the residues interacting familial hypercholesterolemia. Hum Genet 96: 241–242. with calnexin is ERp72, which has been shown not to 10. Saha S, Boyd J, Werner J et al. (2001) Solution structure of the LDL receptor 2/2 21 interact with calnexin in ERp57 cells. Before we EGF-AB pair: a paradigm for the assembly of tandem calcium binding EGF can assume that the native conformer has been reached domains. Structure 9: 451–456. ......................................................................................................................................................................................................................................... 20 Bioscience Horizons † Volume 2 † Number 1 † March 2009 Research article ......................................................................................................................................................................................................................................... 11. Netzer W, Hartl F (1997) Recombination of protein domains facilitated by 19. Kleizen B, Braakman I (2004) Protein folding and quality control in the endo- co-translational folding in eukaryotes. Nature 388: 343–349. plasmic reticulum. Curr Opin Cell Biol 16: 343–349. 12. Jessop C, Chakravarthi S, Garbi N et al. (2007) ERp57 is essential for efficient 20. Maattanen P, Kozlov G, Gehring K et al. (2006) ERp57 and PDI: folding of glycoproteins sharing common structural domains. EMBO J 26: multifunctional protein disulfide isomerases with similar domain architec- 28–40. tures but differing substrate-partner associations. Biochem Cell Biol 84: 881–889. 13. Jessop C, Bulleid N (2004) Glutathione directly reduces an oxidoreductase in the endoplasmic reticulum of mammalian cells. J Biol Chem 279: 55341–55347. 21. Solda` T, Garbi N, Hammerling G et al. (2006) Consequences of ERp57 del- etion on oxidative folding of obligate and facultative clients of the calnexin 14. Wilson R, Allen A, Oliver J et al. (1995) The translocation, folding, assembly cycle. J Biol Chem 281: 6219–6226. and redox-dependent degradation of secretory and membrane proteins in semi-permeabilised mammalian cells. Biochemical J 307: 679–687. 22. Zhang Y, Balg E, Williams D (2006) Functions of ERp57 in the folding and assembly of major histocompatibility complex class I molecules. J Biol 15. Duffy M, Noormohammadi A, Bassegio N et al. (1998) Polyacrylamide gel- Chem 281: 14622–14631. electrophoresis separation of whole-cell proteins. In RJ Miles, RAJ Nicholas eds, Methods in Molecular Biology. Vol. 104. Totowa, NJ: Humana Press. 23. Kulp M, Frickel E, Ellgaard L et al. (2006) Domain architecture of protein- disulphide isomerase facilitates its dual role as an oxidase and an isomerase 16. Gurevich V, Pokrovskaya I, Obukhova T et al. (1991) Preparative in vitro mRNA in Ero1p-mediated disulphide formation. J Biol Chem 281: 876–884. synthesis using SP6 and T7 RNA polymerases. Anal Biochem 195: 207–213. 24. High S, Fabienne L, Russell S et al. (2000) Glycoprotein folding in the endo- 17. Robinson A, King J (1997) Disulphide-bonded intermediate on the folding plasmic reticulum: a tale of three chaperones? FEBS Lett 476: 38–41. and assembly pathway of a non-disulphide bonded protein. Nat Struct Biol 4: 450–455. 25. Russell S, Ruddock L, Salo K et al. (2004) The primary substrate binding site in the b domain of ERp57 is adapted for endoplasmic reticulum lectin 18. Brook D (1999) Introduction: molecular chaperones of the ER: their role in association. J Biol Chem 279: 18861–18869. protein folding and genetic disease. Semin Cell Dev Biol 10: 441–442. ........................................................................................................................................................................................................................................ Submitted on 30 September 2008; accepted on 20 January 2009; advance access publication 10 February 2009 .........................................................................................................................................................................................................................................

Journal

Bioscience Horizons – Oxford University Press

Published: Mar 10, 2009

Keywords: Key words ERp57 disulphides oxidoreductase protein folding

There are no references for this article.