Purification and Partial Molecular Characterization ... - Semantic Scholar

PROTEIN EXPRESSION AND PURIFICATION

7, 114–121 (1996)

Article No. 0015

Purification and Partial Molecular Characterization of GRP94, an ER Resident Chaperone Pamela A. Wearsch and Christopher V. Nicchitta1 Department of Cell Biology, Duke University Medical Center, Durham, North Carolina 27710

Received August 18, 1995, and in revised form October 26, 1995

GRP94 is a resident glycoprotein of the endoplasmic reticulum (ER) and a member of the hsp90 family of molecular chaperones. Current experimental evidence indicates that GRP94 functions in an as yet undefined manner in protein folding and assembly in the ER. We report a rapid, high-yield GRP94 purification procedure that yields milligram quantities of homogeneous protein suitable for structural and biochemical analyses. Beginning with a rough microsome fraction derived from porcine pancreas, GRP94 was isolated by partial detergent extraction, anion exchange, and gel filtration chromatography. With this procedure, approximately 3 mg of homogeneous GRP94 can be prepared from a 25-g pancreas. Heterogeneity in the migration of purified GRP94 on native and denaturing PAGE was observed and demonstrated to reflect variability in the N-linked glycosylation state of the protein. Analysis by native and two-dimensional nonreducing/reducing gels indicated that the protein exists as a dimer of noncovalently associated subunits. The membrane localization of GRP94 in isolated pancreatic microsomes was assessed by alkali and detergent extraction. By comparison with the resident ER integral membrane protein TRAPa, GRP94 exists as a soluble, lumenal protein. q 1996 Academic Press, Inc.

GRP94,2 also known as endoplasmin, gp96, ERp99, hsp108, and hsp100, is an abundant resident protein of the endoplasmic reticulum (1–5). It is a highly acidic glycoprotein with a relative mobility of 94–108 kDa for which homologues have been identified in mammalian, avian, and plant species (1,4,6–8). By sequence homol1

To whom correspondence should be addressed. Abbreviations used: ER, endoplasmic reticulum; RM, rough microsomes; RMek, EDTA/high-salt washed rough microsomes; GRP, glucose-regulated protein; BiP, binding protein; PDI, protein disulfide isomerase; Con A, concanavalin A; HRP, horseradish peroxidase; EndoH, endoglycosidase H; KDEL, Lys-Asp-Glu-Leu. 2

ogy, GRP94 is included as a member of the hsp90 family of stress proteins (1,3,4,9). Analysis of the cDNAderived sequence indicates that the protein is synthesized with an amino-terminal signal sequence and a C-terminal KDEL motif, characteristics which support localization of the protein to the lumen of the ER (10,11). The cDNA sequence also predicts, however, a potential transmembrane domain, and experimental evidence suggesting that GRP94 exists in both a transmembrane and a lumenal form has been provided (12–14). The expression of ER chaperones, such as GRP94 and BiP (GRP78), is coordinately regulated and increased under a variety of stress conditions, including glucose starvation, inhibition of glycosylation, heat shock, and treatment with reducing agents (6,15,16). In general, conditions that result in the accumulation of unfolded or mutant proteins in the ER activate transcription of the GRP94 and BiP genes (17,18). Based on its association with unassembled oligomeric protein substrates, such as immunoglobulin chains, MHC Class II molecules, and a mutant form of the herpes simplex virus-1 glycoprotein B, it has been proposed that GRP94 acts as a molecular chaperone (19–22). In pulse-chase studies of immunoglobulin assembly, Melnick et al. (21) identified sequential interactions of GRP94 and BiP with immunoglobulin assembly intermediates. Because GRP94 was found only in association with fully oxidized molecules, it was proposed that GRP94 preferentially interacts with protein substrates at late stages of folding and assembly (21). This proposal is consistent with current views on the cellular role of the cytosolic homologue of GRP94, hsp90. In vivo, hsp90 associates with newly synthesized, yet structurally mature proteins such as steroid receptors and protein kinases (23). It may be that all members of the hsp90 protein family recognize distinct structural or conformational motifs, rather than discrete peptide sequences. In this communication, we report a rapid, high-yield

114

1046-5928/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

/ m4941f0555

01-04-96 08:15:13

pepa

AP-PEP

RAPID PURIFICATION OF GRP94

purification protocol for GRP94. Analysis of the purified native protein has allowed several conclusions on its structure and cellular localization. On native and SDS–PAGE gels, multiple species of GRP94 have been identified and demonstrated to reflect heterogeneity in N-linked glycosylation states. The observed mobility on native gels is consistent with reports that GRP94 exists as a dimeric protein (24–26). The behavior of GRP94 on two-dimensional nonreducing/reducing PAGE indicates that dimerization occurs through noncovalent protein–protein interactions. An analysis of the membrane localization of GRP94 in rough microsomes was performed with alkali and detergent extraction procedures. By comparison with the resident ER integral membrane protein TRAPa, it was observed that GRP94 is a soluble component of the ER lumen. MATERIALS AND METHODS

Materials. Chemicals were of analytical grade and were obtained from Fluka (Ronkonkoma, NY) and Sigma Chemicals (St. Louis, MO). Endoglycosidase H was from Boehringer Mannheim (Indianapolis, IN). Chromatography columns were obtained from Pharmacia Biotech (Piscataway, NJ). The GRP94 monoclonal antibody 9G10 was obtained from Stressgen (Vancouver, B.C.) and the ECL reagents from Amersham (Arlington Heights, IL). Membrane preparation and extraction of lumenal contents. Porcine pancreas rough microsomes (RM) were prepared by the method of Walter and Blobel (27). Peripheral membrane proteins were removed by a high-salt washing procedure. RM at a concentration of 50 A280 U/ml were diluted 1:10 into a buffer containing 0.55 M KOAc, 2.2 mM EDTA, 1 mM DTT, and 0.5 mM PMSF and incubated for 30 min on ice. Washed RM (RMek) were subsequently collected by centrifugation (47C, 1 h) through a cushion of 0.5 M sucrose, 25 mM K-Hepes, pH 7.2, 1 mM DTT at 45,000 rpm in a Beckman Ti50.2 rotor. The RMek were resuspended to the original volume in a buffer consisting of 50 mM KOAc, 50 mM K-Hepes, pH 7.2, 5 mM Mg(OAc)2 , and 0.5 mM PMSF by Dounce homogenization. The membrane suspension was then permeabilized by dropwise addition, with stirring, of the nonionic detergent Nikkol (C12E8) to a final concentration of 0.25%. After a 30-min incubation, the permeabilized membranes were separated from the released lumenal contents by centrifugation through a 0.5 M sucrose cushion as described above. The supernatant plus the top 2–3 ml of the cushion was collected and filtered through a 0.2-mm cellulose acetate filter. Chromatography. The filtered lumenal protein fraction was injected onto a Pharmacia MonoQ HR 10/ 10 column equilibrated in 50 mM NaCl, 25 mM Tris– Cl, pH 7.8, 0.01% Nikkol at a flow rate of 1 ml/min

/ m4941f0555

01-04-96 08:15:13

pepa

115

and washed with the same buffer until the absorbance returned to baseline. The lumenal proteins were eluted with a 100-ml linear gradient of 50 to 750 mM NaCl, in 25 mM Tris–Cl, pH 7.8, 0.01% Nikkol at a flow rate of 4 ml/min. Fractions were analyzed on 10% gels and the GRP94 peak fractions were identified and pooled. Further purification of GRP94 was accomplished by gel filtration chromatography on a Superdex 26/60 column equilibrated in 110 mM KOAc, 20 mM NaCl, 25 mM K-Hepes, pH 7.2, 2 mM Mg(OAc)2 , and 0.1 mM CaCl2 . The pooled MonoQ fraction was injected and run at a flow rate of 2 ml/min, collecting 4-ml fractions. The GRP94 fractions were pooled and concentrated in a Centiprep-30 (Amicon Inc., Beverly, MA) according to the manufacturer’s instructions. The concentrated protein was divided into aliquots and stored at 0807C. Alkali and detergent extraction of GRP94. Fifteen microliters of RM was diluted to 150 ml in one of the following buffers and incubated on ice for 30 min: 100 mM K-Bicine, pH 7.5; 100 mM Na2CO3 , pH 11.5; 6 mM Big CHAPS, 0.25 M sucrose, 50 mM K-Hepes, pH 7.2, 50 mM KOAc, 0.5 mM DTT; or 12 mM cholate, 0.25 M sucrose, 50 mM K-Hepes, pH 7.2, 50 mM KOAc, 0.5 mM DTT. The suspensions were overlaid onto a 0.5 M sucrose, 50 mM K-Hepes, pH 7.2, cushion at a volume ratio of 3:1 and centrifuged for 10 min at 60,000 rpm, 47C, in a Beckman TLA100 rotor. The supernatant and pellet fractions were collected and processed for SDS– PAGE as described in Nicchitta and Blobel (28). The samples were run on 10% gels and transferred to nitrocellulose for subsequent immunoblot analysis. EndoH digestion of GRP94. Purified GRP94 (1.5 mg), in a final volume of 40 ml, was incubated at 377C for 2.5 h in a buffer consisting of 110 mM KOAc, 25 mM MES pH 6.0, in the presence or absence of 5 munits of EndoH. The samples were processed for SDS–PAGE and analyzed in triplicate on 7.5% gels. One set of samples was stained with Coomassie blue and the remaining samples were transferred to nitrocellulose for immunoblot analysis. Immunoblotting. Nitrocellulose membranes were blocked for 2 h at RT in 3% BSA, PBS, 1% Triton X100, 0.2% SDS. Subsequent incubations were done in 1% BSA, PBS, 0.1% Triton X-100, 0.02% SDS. For GRP94 detection, a rat monoclonal antibody (9G10) was used at a dilution of 1:4000 for 1 h at RT. A 5-min wash was followed by a 30-min incubation with a 1:500 dilution of rabbit anti-rat IgG (Bio-Rad, Hercules, CA). After three additional 5-min washes, the membrane was incubated with goat anti-rabbit HRP (Bio-Rad, Hercules, CA) at a dilution of 1:1000. After 3 1 20 minute washes, detection was performed by ECL. Detection of glycoproteins by Con A–HRP. GRP94 samples were prepared for SDS–PAGE, separated on 7.5% gels, and transferred to nitrocellulose mem-

AP-PEP

116

WEARSCH AND NICCHITTA

branes. A buffer of 0.5 M NaCl, 25 mM Tris–Cl, pH 7.5, 2.5 mM Mg(OAc)2 , 1 mM CaCl2 , 0.5% Tween-20 was used for all incubations. The membrane was blocked for 2 h at RT and then incubated with a 1:4000 dilution of Con A–HRP for 3 h (Sigma Chemicals). After three 20-min washes, glycosylated bands were detected using ECL. Native gel electrophoresis. A 5–10% nondenaturing gradient gel was made as described in the Electrophoresis Applications Guide from Hoefer Scientific Instruments (San Francisco, CA). Sample volume was adjusted to 40 ml with 5% sucrose, 0.01% bromophenol blue and samples were run at 100 V for 20 h at 47C. Ten micrograms of purified GRP94 was analyzed by comparison to standards. The native molecular weight standards included thyroglobulin (660 kDa), urease trimer (270 kDa), catalase (230 kDa), b-amylase (200 kDa), yeast alcohol dehydrogenase (140 kDa), and BSA (67 kDa). Proteins were visualized by Coomassie blue staining. Two-dimensional electrophoresis. For the first (nonreducing) dimension, samples were placed in 0.5 M Tris, 5% SDS, 0.01% bromophenol blue and heated for 10 min at 557C. Two 1-mg samples of purified GRP94 were prepared, one of which included 50 mM DTT in the sample buffer. Five micrograms of high-molecular-weight standards and IgG were included as controls. Samples were loaded on 7.5% gels cast in capillary tubes and run at 200 V in the Mini-Protean II 2-D Cell (Bio-Rad). The tube gels were extracted, incubated in 0.5 M Tris, 5% SDS, 0.1 M DTT for 15 min at 557C, overlaid on 7.5% preparative slab gels, and run in the second dimension. Proteins were visualized by Coomassie blue staining. Miscellaneous. Protein quantitation was done with the BCA assay reagents (Pierce, Rockford, IL) using BSA as a standard. With the exception of the native gel previously described, all gels were prepared according to the method of Laemmli (29). RESULTS AND DISCUSSION

Purification The starting material for the described purification of GRP94 is a highly enriched rough endoplasmic reticulum (RER) fraction derived from porcine or canine pancreas. GRP94 is estimated to be present in the RER at concentrations exceeding 10 mg/ml (30). In addition, in secretory cells, such as the pancreatic acinar cell, the RER can constitute as much as 70–80% of the total membrane surface area (31). The high specific enrichment of GRP94 in the pancreatic RM fraction, combined with the ease of preparation and substantial yield of RM from isolated pancreas, greatly simplifies the purification. This conclusion is supported by the

/ m4941f0555

01-04-96 08:15:13

pepa

FIG. 1. Purification of GRP94 from porcine RM. Samples from each step of the purification from RM were analyzed by SDS–PAGE and Coomassie blue staining. Lane 1, 10 equivalents (eq) of rough microsomes (RM); lane 2, 10 eq of EDTA/KOAc-washed microsomes (RMek); lane 3, 10 eq of depleted membranes (DM); lane 4, 10 eq of lumenal proteins; lane 5, 20 eq of the MonoQ fraction; and lane 6, 16 eq of the gel filtration fraction. The diamonds mark the positions of the major lumenal protein bands, which are, in descending order, GRP94,100 kDa; BiP, 78 kDa; ERp72, 72 kDa; calreticulin, 55 kDa; and PDI, 50 kDa.

data presented in Fig. 1, lane 1, in which a Coomassie blue-stained protein profile of the RM fraction is depicted. It is important to note that unlike a crude membrane preparation, the isolated rough microsome fraction is essentially free of other membranes and therefore represents a substantial enrichment of ER localized proteins. From this protein profile it is also evident that GRP94 is an abundant ER protein. Since an enzymatic activity for GRP94 has not yet been identified, the purification was monitored on a protein basis (Table 1). Once a microsome fraction has been prepared, two simple, effective steps were employed to obtain a highly enriched lumenal protein pool. First, the RM were washed in a high-salt/EDTA buffer (RMek) to remove peripheral membrane proteins (Fig. 1, lane 2). The lumenal contents were then selectively released from the RMek by partial detergent extraction. In the presence of low (50 mM) salt concentrations, nonionic detergents are quite inefficient at solublizing RM integral membrane proteins (32). Under these same conditions, however, the membranes are permeabilized and the lumenal contents released. As shown in Fig. 1, lane 4, the lumenal contents consist primarily of five proteins, GRP94, BiP, ERp72, calreticulin, and protein disulfide isomerase (PDI), which are absent in the depleted membrane fraction (Fig.1, lane 3). Densitometric analysis of a Coomassie blue-stained lumenal protein frac-

AP-PEP

117

RAPID PURIFICATION OF GRP94 TABLE 1

Purification of GRP94 from Porcine Rough Microsomes Sample

Volume (ml)

Concn (mg/ml)

Total protein (mg)

Percentage yield

Homogenate 12,000g supernatant Rough microsomes High-salt, EDTA-washed RM Depleted microsomes Lumenal contents Ion exchange: MonoQ Gel filtration: Superdex

110.0 91.4 23.0 23.0 23.0 42.5 7.8 5.4

25.87 9.48 8.26 7.10 3.18 1.04 0.59 0.65

2845.7 866.5 190.0 163.3 73.1 44.2 4.6 3.5

— — — 100 — 27.1 2.8 2.1

tion indicated that GRP94 accounted for approximately 19% of the total protein. The RER-derived lumenal protein pool was further fractionated by MonoQ anion exchange chromatography and elution with a 100-ml linear gradient of 50 to 750 mM NaCl (Fig. 2A). The protein composition of the MonoQ fractions was analyzed by SDS–PAGE and illustrates a very effective separation of lumenal proteins (Fig. 2B). Arrows indicate the peaks for GRP94 (fractions 32–34), BiP (fractions 20–24), PDI (fractions 23–24), ERp72 (fractions 28–30), and calreticulin (fractions 31–32). Being a highly acidic protein with a pI of 4.7, GRP94 elutes off the MonoQ column in the range 425 to 485 mM NaCl. The peak fractions from the MonoQ column were pooled and, as shown in Fig. 1, lane 5, are nearly 80% pure. Gel filtration was next performed to separate GRP94 from the predominant contaminant, calreticulin, which has an approximate native molecular weight of 55 kDa (33). The MonoQ pool was injected onto a Superdex 26/ 60 prep column equilibrated in 110 mM KOAc, 50 mM NaCl, 25 mM K-Hepes, pH 7.2, 2 mM Mg(OAc)2 , 0.1 mM CaCl2 . GRP94 eluted in a peak clearly separated from calreticulin (data not shown), the fractions of which were pooled and concentrated. As can be seen in Fig. 1, lane 6, the protein is essentially homogeneous following the gel filtration step. The final yield and summary of the purification are presented in Table 1. Over 3 mg of homogeneous protein was isolated from 25 g of starting material, representing approximately a 10- to 150-fold improvement relative to yields from protocols using cultured cells or liver as a tissue source (5,34). Although the purification described herein is devoted to the purification of GRP94, these procedures can be readily adapted for the purification of the other resident lumenal proteins, such as BiP, PDI, ERp72, and calreticulin. Further inspection of the Coomassie bluestained gels in Fig. 2B indicates that MonoQ chromatography provides a substantial purification for all of the lumenal proteins. In many instances, purification

/ m4941f0555

01-04-96 08:15:13

pepa

to homogeneity can be accomplished by a single additional chromatography step (data not shown). Molecular Characterization Reports concerning the membrane localization of GRP94 have concluded that the protein exists either

FIG. 2. Anion exchange chromatography of lumenal proteins. (A) MonoQ chromatography of ER lumenal proteins. The supernatant obtained from the detergent permeabilization step was loaded onto a MonoQ column and proteins were eluted using a 100-ml gradient of 50 to 750 mM NaCl. The solid line indicates the absorbance of the eluent at 280 nm and the dashed line represents the NaCl concentration. (B) Protein profile of Mono-Q fractions. The composition of fractions 13–37 was analyzed on 10% gels and visualized by staining with Coomassie blue. The arrows indicate the elution peaks of the major ER lumenal proteins: (1) BiP, (2) PDI, (3) ERp72, (4) calreticulin, and (5) GRP94.

AP-PEP

118


as a soluble lumenal component (13) or in both a transmembrane and a soluble form (12,14). In protease protection studies of MOPC-315 cell homogenates, Lewis et al. (12) observed that, unlike the lumenal proteins ERp72 and calreticulin, GRP94 was susceptible to tryptic digestion in the absence of detergents and thus contained a substantial cytosolic domain. Additional protease protection experiments and surface iodination labeling experiments with microsomes isolated from HeLa cells have yielded similar results (14). As the cloning and sequencing of the murine GRP94 homologue revealed one putative transmembrane domain near the N-terminus of the protein, an orientation was proposed for GRP94 in which residues 1–169 of the mature protein are in the lumen of the ER, residues 170–192 function as a transmembrane domain, and the remaining 65 kDa of the protein extends into the cytoplasm (3). In contrast, Koch et al. (13) observed that GRP94 could be isolated from membrane fractions by mechanical disruption and, by immunogold analysis, was dispersed throughout the ER lumen, indicating that it is a soluble, lumenal protein. To explain both sets of data it was proposed that GRP94 exists in both soluble and integral membrane forms (14). The described protocol for the purification of GRP94 was designed to take advantage of a selective permeabilization step in which integral membrane proteins are retained in the microsomes and a soluble lumenal protein fraction is released. As seen in Fig. 1, GRP94 is present in the lumenal fraction (lane 4) and virtually absent in the depleted microsome fraction (lane 3). To provide further evidence that GRP94 behaves as a soluble, lumenal protein, rough microsomes were extracted and the membrane association of GRP94 was compared to both a resident ER integral membrane protein, TRAPa, and a soluble lumenal protein, PDI. Soluble and membrane-associated components were recovered by centrifugation over a sucrose cushion, separated on SDS–PAGE gels, transferred to nitrocellulose, and screened by immunoblotting with antibodies directed against GRP94, TRAPa, and PDI. As shown in Fig. 3A, when RM were diluted in an isotonic buffer at physiological pH, GRP94, PDI, and TRAPa were found predominantly in the pellet (or membrane) fraction. The small amount of GRP94 and PDI found in the supernatant fraction is the consequence of a relatively minor degree of membrane lysis that accompanies the thawing of frozen RM prior to the experiment (C. Nicchitta, unpublished observations). Alkali extraction in a pH 11.5 buffer converts the vesicles to membrane sheets and effectively separates peripheral and lumenal proteins from integral membrane proteins (35). Under these conditions, GRP94 and PDI are recovered in the supernatant whereas TRAPa is completely retained in the pellet (Fig. 3B). The distribution of GRP94 was also assayed following treatment of the RM with sublytic

/ m4941f0555

01-04-96 08:15:13

pepa

FIG. 3. Membrane topology of GRP94 in pancreatic rough microsomes. RM were diluted 1:10 in 100 mM K-Bicine, pH 7.5 (A), 100 mM Na2CO3 (B), 6 mM Big CHAPS (C), or 12 mM cholate (D). Detergents were in a buffer consisting of 0.25 M sucrose, 50 mM K-Hepes, pH 7.2, 50 mM KOAc, 0.5 mM DTT. After centrifugation over a 0.5 M sucrose, 50 mM K-Hepes, pH 7.2, cushion, the pellet and supernatant fractions were collected, processed for SDS–PAGE, and transferred to nitrocellulose. Immunoblots of these fractions were performed for GRP94, the lumenal marker, PDI, and the membrane protein marker, TRAPa.

concentrations of the detergents Big CHAPS and cholate (Figs. 3C and 3D). For both extraction conditions, GRP94 and PDI were observed to fractionate almost entirely in the supernatant fraction whereas TRAPa remained associated with the membranes. From these experiments, it can be concluded that GRP94 exists as a soluble component of pancreatic rough microsomes. It is uncertain why topology determinations of GRP94 have yielded such conflicting conclusions. It is noteworthy, however, that when the topology is determined in microsomes isolated from organ tissue, a lumenal disposition is reported (Fig. 3; ref. 13), whereas in isolations from pulse-labeled tissue culture cells, a transmembrane orientation appears possible (12,14). The membrane topology proposed by Mazzarella and Green (3) is problematic in that it places the C-terminal KDEL motif, known to function as an ER retention/ retrieval sequence for lumenal proteins, in the cytopalsm (11). Similarly, it has been reported that Asn196 is the acceptor site predominantly used by wildtype GRP94 (26). If the postulated transmembrane orientation for GRP94 is correct, Asn-196 would also reside in the cytosol and thus be inaccessible to the oligosaccharyl transferase. These objections notwithstanding, it is unclear why in tissue culture cell-derived homogenates or microsomes, GRP94, unlike other ER lumenal components, remains sensitive to impermeant probes such as proteases or lactoperoxidase. Upon close inspection of the purified protein on SDS– PAGE (Fig. 1, lane 6), several closely migrating bands were observed. To investigate whether these were different forms of GRP94, or if a comigrating contaminant existed, the protein composition of the purified fraction was more clearly resolved on 7.5% gels. By Coomassie

AP-PEP

RAPID PURIFICATION OF GRP94

FIG. 4. Molecular heterogeneity of purified porcine GRP94. 1.5 mg of purified GRP94 was treated in the presence or absence of EndoH for 3 h at 377C. The reaction products were run in triplicate on 7.5% SDS–PAGE gels and either stained with Coomassie blue (A) or transferred to nitrocellulose for detection with either a monoclonal antibody directed against GRP94 (B) or horseradish peroxidase-conjugated Con A (C). The asterisks denote the positions of the three GRP94 bands identified by immunoblotting. The monoglycosylated species of GRP94 is indicated by the ‘‘Y,’’ and the multiglycosylated species by the ‘‘YY.’’

blue staining two bands were clearly visible (Fig. 4A). Both bands reacted with the 9G10 anti-GRP94 monoclonal antibody (Fig. 4B). Upon longer exposures of the immunoblot, a third band, found just slightly above the slowest migrating band of the pair, was also detected (data not shown). To determine if the heterogeneity might be the result of differences in glycosylation, purified GRP94 was incubated with endoglycosidase H (EndoH), an enzyme that cleaves high-mannose N-linked oligosaccharides. As shown by Coomassie blue staining and immunoblotting (Figs. 4A and 4B), EndoH treatment resulted in a mobility shift of the most predominant form to that of the faster migrating band, which itself did not appear to be EndoH sensitive. Since the difference in relative mobility of these bands corresponds approximately to that of one oligosaccharide chain (1.5 kDa), it was postulated that the different forms represented monoglycosylated and nonglycosylated species, respectively. To explore this possibility, GRP94 was incubated in the presence or absence of EndoH, separated on 7.5% SDS– PAGE gels, and transferred to nitrocellulose, and glycosylated species were detected by HRP-conjugated Con A. Of the three bands identified thus far, only the fastest migrating species did not bind Con A and is therefore a nonglycosylated form of GRP94 (Fig. 3, compare B and C). The two slower migrating bands were able to bind Con A and after EndoH treatment comigrated with the nonglycosylated GRP94. These bands likely represent mono- and multiglycosylated forms of GRP94. In the preparation depicted in Fig. 4, the multiglycosylated form of GRP94 constitutes a very small fraction of the total GRP94 pool. Although only trace amounts of this species were detectable by immunoblotting, it is reasonable to expect that the presence

/ m4941f0555

01-04-96 08:15:13

pepa

119

of additional oligosaccharides would increase the binding capacity of GRP94 for the lectin and thus the relative sensitivity to detection with Con A–HRP. Densitometric quantitation of the major species on Coomassie blue-stained gels indicates that the majority (95%) of the total native porcine GRP94 contains a single oligosaccharide and £5% is present in a nonglycosylated state. To further study the observed heterogeneity and assess the oligomerization state of the native protein, purified GRP94 was analyzed by native gel electrophoresis. As depicted in Fig. 5, three species of GRP94 can be distinguished on native gels and are expected to represent the same differences in glycosylation mentioned above. GRP94 has been reported to exist as a dimer of disulfide bonded subunits (3,24–26). The mobility of GRP94 relative to that of the catalase (Fig. 5, lane 3) and b-amylase (Fig. 5, lane 4) standards is consistent with a dimer. To determine whether quaternary structure was maintained by covalent or noncovalent interactions, two-dimensional electrophoresis was performed with native purified GRP94. Samples were run under nonreducing conditions in tube gels, incubated in a reducing buffer, and overlaid on slab gels for electrophoresis in the second dimension. Using this type of 2-D analysis, proteins that are either monomers or nondisulfide bonded oligomers will migrate on a diagonal, as demonstrated with molecular weight markers in Fig. 6A. In contrast, disulfide bonded oligomers will migrate off of the diagonal. The behavior of IgG heavy chains is illustrated as a control (Fig. 6B). IgG heavy chain migrated as a dimer of disulfide bonded

FIG. 5. Native gel electrophoresis of GRP94. 10 mg of purified GRP94 was run on a 5–10% nondenaturing gradient gel along with known standards and visualized by Coomassie blue staining. The standards and their native molecular weights are as follows: lane 1, thyroglobulin, 660 kDa; lane 2, urease trimer, 270 kDa; lane 3, catalase, 230 kDa; lane 4, b-amylase, 200 kDa; lane 5, yeast ADH, 140 kDa; and lane 6, BSA, 66 kDa. The bars indicate the positions of three GRP94 bands (lane 7), each of which represents a native species.

AP-PEP

120


REFERENCES

FIG. 6. 2-D nonreducing/reducing electrophoresis of native GRP94. Molecular weight standards (5 mg) (A), IgG (5 mg) (B), or GRP94 (1 mg), either nonreduced (C) or reduced (D), were first run under nonreducing conditions in 7.5% tube gels. After incubation in 0.5 M Tris, 5% SDS, 0.1 M DTT for 15 min at 557C, the tube gels were run in the second (reducing) dimension on 7.5% preparative slab gels. Proteins were visualized by Coomassie blue staining. The arrow in B denotes the position of the disulfide bonded IgG heavy chains.

subunits at an apparent molecular weight of 110 kDa in the first dimension and, following reduction, as 55kDa monomers in the second dimension. When GRP94 was analyzed by this method, the native protein migrated on the diagonal (Fig. 6C). As an additional control, GRP94 was treated with DTT prior to the first dimension to demonstrate that GRP94 behaves identically in the presence or absence of reducing agents (Fig. 6D). Dimerization of GRP94 thus occurs through noncovalent protein–protein interactions. In summary, we have devised a rapid, efficient protocol for the purification of GRP94 from pancreatic rough microsomes. Milligram quantities of homogeneous protein that were suitable for biochemical studies were isolated. This procedure can be easily adapted for the purification of the major ER lumenal proteins BiP, calreticulin, PDI, and ERp72. By selective alkali and detergent extraction from rough microsomes, it was concluded that GRP94 behaves as a soluble lumenal protein of pancreatic rough microsomes. Analysis of the purified protein on SDS–PAGE and native gels revealed that heterogeneity exists in the endogenous GRP94 pools and is the result of differential glycosylation. GRP94 exists as a dimer with oligomerization conferred solely by protein–protein interactions. ACKNOWLEDGMENTS We thank Edwin C. Murphy III for helpful comments during the preparation of the manuscript, Waldo Stressman for technical assistance, and the Duke University Department of Surgery for donating the tissue used in the purification. This work was supported by NIH Grant DK47897.

/ m4941f0555

01-04-96 08:15:13

pepa

1. Smith, M. J., and Koch, G. L. E. (1987) Isolation and identification of partial cDNA clones for endoplasmin, the major glycoprotein of mammalian endoplasmic reticulum. J. Mol. Biol. 194, 345–347. 2. Srivastava, P. K., Chen, Y. T., and Old, L. J. (1987) Structural analysis of genes encoding polymorphic antigens of chemically induced tumors. Proc. Natl. Acad. Sci. USA 84, 3804–3807. 3. Mazzarella, R. A., and Green, M. (1987) ERp99, an abundant, conserved glycoprotein of the endoplasmic reticulum, is homologous to the 90-kDa heat shock protein (hsp90) and the 94-kDa glucose regulated protein (GRP94). J. Biol. Chem. 262, 8875– 8883. 4. Sargan, D. R., Tsai, M. J., and O’Malley, B. W. (1986) hsp108, a novel heat shock inducible protein of chicken. Biochemistry 25, 6252–6258. 5. Koyasu, S., Nishida, E., Miyata, Y., Sakai, H., and Yahara, I. (1989) HSP100, a 100-kDa heat shock protein, is a Ca2/-calmodulin-regulated actin-binding protein. J. Biol. Chem. 264, 15083– 15087. 6. Welch, W. J., Garrels, J. I., Thomas, G. P., and Feramisco, J. R. (1983) Biochemical characterization of the mammalian stress proteins and identification of two stress proteins as glucose- and Ca2/-ionophore-regulated proteins. J. Biol. Chem. 258, 7102– 7111. 7. Lewis, M. J., Mazzarella, R. A., and Green, M. (1985) Structure and assembly of the endoplasmic reticulum: The synthesis of three major ER proteins during lipopolysaccharide-induced differentiation of murine lymphocytes. J. Biol. Chem. 260, 3050– 3057. 8. Walther-Larsen, H., Brandt, J., Collinge, D. B., and ThordalChristensen, H. (1993) A pathogen-induced gene of barley encodes a HSP90 homologue showing striking similarity to vertebrate forms resident in the endoplasmic reticulum. Plant Mol. Biol. 21, 1097–1108. 9. Sorger, P. K., and Pelham, H. R. (1987) The glucose-regulated protein grp94 is related to heat shock protein hsp90. J. Mol. Biol. 194, 341–344. 10. Blobel, G. (1980) Intracellular protein topogenesis. Proc. Natl. Acad. Sci. USA 77, 1496–1500. 11. Munro, S., and Pelham, H. R. B. (1987) A C-terminal signal prevents secretion of luminal ER proteins. Cell 48, 899–907. 12. Lewis, M. J., Turco, S. J., and Green, M. (1985) Structure and assembly of the endoplasmic reticulum: Biosynthetic sorting of endoplasmic reticulum proteins. J. Biol. Chem. 260, 6926–6931. 13. Koch, G. L. E., Macer, D. R. J., and Wooding, F. B. P. (1988) Endoplasmin is a reticuloplasmin. J. Cell Sci. 90, 485–491. 14. Kang, H. S., and Welch, W. J. (1991) Characterization and purification of the 94-kDa glucose-regulated protein. J. Biol. Chem. 266, 5643–5649. 15. Shiu, R. P. C., Pouyssegur, J., and Pastan, I. (1977) Glucose depletion accounts for the induction of two transformation-sensitive membrane proteins in Rous sarcoma virus-transformed chick embryo fibroblasts. Proc. Natl. Acad. Sci. USA 74, 3840– 3844. 16. Lee, A. S. (1987) Coordinated regulation of a set of genes by glucose and calcium ionophores in mammalian cells. Trends Biochem. Sci. 12, 20–23. 17. Kim, Y. K., Kim, K. S., and Lee, A. S. (1986) Regulation of the glucose-regulated protein genes by b-mercaptoethanol requires de novo protein synthesis and correlates with inhibition of protein glycosylation. J. Cell. Physiol. 133, 533–559. 18. Kozutsumi, Y., Segal, M., Normington, K., Gething, M. J., and

AP-PEP

RAPID PURIFICATION OF GRP94 Sambrook, J. (1988) The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature (London) 332, 462–464. 19. Schaiff, W. T., Hruska, K. A., McCourt, D. W., Green, M., and Schwartz, B. D. (1992) HLA-DR associates with specific stress proteins and is retained in the endoplasmic reticulum in invariant chain negative cells. J. Exp. Med. 176, 657–666. 20. Melnick, J., Aviel, S., and Argon, Y. (1992) The endoplasmic reticulum stress protein GRP94, in addition to BiP, associates with unassembled immunoglobulin chains. J. Biol. Chem. 267, 21303–21306. 21. Melnick, J., Dul, J. L., and Argon, Y. (1994) Sequential interaction of the chaperones BiP and GRP94 with immunoglobulin chains in the endoplasmic reticulum. Nature (London) 370, 373– 375. 22. Ramakrishnan, M., Tugizov, S., Pereira, L., and Lee, A. S. (1995) Conformation-defective herpes simplex virus-1 glycoprotein B activates the promoter of the grp94 gene that codes for the 94kD stress protein in the endoplasmic reticulum. DNA Cell Biol. 14, 373–384.

27.

28.

29.

30.

31. 32.

23. Jakob, U., and Buchner, J. (1994) Assisting spontaneity: The role of Hsp90 and small Hsps as molecular chaperones. Trends Biochem. Sci. 19, 205–211.

33.

24. Lee, A. S., Bell, J., and Ting, J. (1984) Biochemical characterization of the 94- and 78-kilodalton glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem. 259, 4616–4621.

34.

25. Koyasu, S., Nishida, E., Kadowaki, T., Matsuzaki, F., Iida, K., Harada, F., Kasuga, M., Sakai, H., and Yahara, I. (1986) Two mammalian heat shock proteins, HSP90 and HSP100, are actinbinding proteins. Proc. Natl. Acad. Sci. USA 83, 8054–8058. 26. Qu, D., Mazzarella, R. A., and Green, M. (1994) Analysis of the

/ m4941f0555

01-04-96 08:15:13

pepa

35.

121

structure and synthesis of GRP94, an abundant stress protein of the endoplasmic reticulum. DNA Cell Biol. 13, 117–124. Walter, P., and Blobel, G. (1983) Preparation of microsomal membranes for cotranslational protein translocation. Meth. Enzymol. 96, 84–93. Nicchitta, C. V., and Blobel, G. (1993) Lumenal proteins of the mammalian endoplasmic reticulum are required to complete protein translocation. Cell 73, 989–998. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 259, 680–685. Koch, G., Smith, M., Macer, D., Webster, P., and Mortara, R. (1986) Endoplasmic reticulum contains a common, abundant calcium-binding glycoprotein, endoplasmin. J. Cell Sci. 86, 217– 232. Palade, G. (1975) Intracellular aspects of the process of protein synthesis. Science 189, 347–358. Kelleher, D. J., Kreibich, G., and Gilmore, R. (1992) Oligosaccharyltransferase activity is associated with a protein complex composed of ribophorins I and II and a 48 kD protein. Cell 69, 55–65. Waisman, D. M., Salimath, B. P., and Anderson, M. J. (1985) Isolation and characterization of CAB-63, a novel calcium-binding protein. J. Biol. Chem. 260, 1652–1660. Rowling, P. J. E., McLaughlin, S. H., Pollock, G. S., and Freedman, R. B. (1994) A single purification procedure for the major resident proteins of the ER lumen; endoplasmin, BiP, calreticulin and protein disulfide isomerase. Protein Expression Purif. 5, 331–336. Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1982) Isolation of intracellular membranes by means of sodium carbonate treatment: Application to endoplasmic reticulum. J. Cell Biol. 93, 97–102.

AP-PEP