The most abundant protein in microsomal membrane preparations from mammalian cells has been identified as a 100 × 103Mr concanavalin A-binding glycoprotein. The glycosyl moiety of the glycoprotein is completely sensitive to endoglycosidase H, suggesting a predominantly endoplasmic reticulum localization in the cell. Using a monospecific antibody it was shown by binding and immunofluorescence studies that the glycoprotein is intracellular. Immunoelectron microscopy showed that the glycoprotein was at least 100 times more concentrated in the endoplasmic reticulum than in any other cellular organelle. It was found to be substantially overexpressed in cells and tissues rich in endoplasmic reticulum. Since it is the major common protein component associated with the endoplasmic reticulum we refer to it as endoplasmin. Calcium-binding studies show that endoplasmin is a major calcium-binding protein in cells, suggesting that at least one of its roles might be in the calcium-storage function of the endoplasmic reticulum. The amino-terminal sequence of endoplasmin is identical to that of a 100 × 103Mr stress-related protein.

The endoplasmic reticulum (ER) consists of a complex system of intracellular membranes, which can account for as much as 90% of the total membrane in some cells (Alberts et al. 1983; Palade, 1975). It contains at least two distinct regions, the smooth and rough ER. It is generally accepted that the ER is one of the major synthetic organelles involved in the synthesis of secretory products, plasma membrane, lysosomes and secretory granules. In addition, the ER is the site of synthesis and assembly of the lipid bilayer that forms the basic unit of membrane structure.

Recently, another major function of the ER has gained prominence, i.e. its role as a calcium store that serves as a source of calcium for intracellular signalling during a variety of physiologically important processes (Berridge & Irvine, 1984; Streb et al. 1983). Thus in the Pl cycle, hormone-receptor interactions result in the release of inositol triphosphate, which passes to the ER and causes release of calcium, which acts on a variety of cellular processes. In a general sense, therefore, the ER appears to simulate the sarcoplasmic reticulum, which mediates calcium signalling during muscle contraction (MacLennan & Holland, 1975).

In spite of extensive investigations, the molecular basis of many ER functions remains obscure. One reason for this is that, unlike other membrane organelles such as mitochondria, ER cannot be isolated without considerable disruption and disorganization. Thus the definition of ER structure and function in molecular terms remains a major challenge in cell biology.

In this study we approached the analysis of ER structure and function by using a cultured plasmacytoma cell as the starting material for endoplasmic reticulum. Plasma cells are almost completely dedicated to the secretion of immunoglobulin and therefore contain large amounts of rough ER to effect this function (Alberts et al. 1983).

Using ER-rich membranes we have demonstrated, for the first time, that ER contains at least one common glycoprotein, which is actually one of the most abundant cellular proteins. By analogy with other organelle-associated common abundant proteins, such as nucleoplasmin, we refer to this ER-associated glycoprotein as endoplasmin.

Cells and cell culture

All cells were grown in RPM1 1640 medium supplemented with 10 % foetal calf serum, 4 mM-L-glutamine and 100 units ml−1 of penicillin and streptomycin. Cells were washed with phosphate-buffered saline (PBS) before use.

Preparation of tissue extracts

Organs from Balb/c mice were dissected out by standard procedures. Dissection of pancreas was kindly carried out by Dr N. Kamada. The tissues were weighed, coarsely minced with scissors and suspended in PBS with 1 % Nomdet P40 (NP40) and the protease inhibitors (PI) described previously (Koch et al. 1985). The suspension was homogenized in a Polytron homogenizer and the debris removed by centrifugation at 10 000 g for 30 min. The protein content of each extract was measured (see below) and all extracts adjusted to lmgml-1 protein for gel electrophoresis and immunoblotting.

Purification of endoplasmin for immunization and sequencing

The preparation of concanavalin A (ConA)-Sepharose, and its use in the purification of glycoproteins was as described (Koch & Smith, 1982). When these procedures are applied to the MOPC-315 line endoplasmin is the major protein in the purified preparation (Fig. 1). Therefore eluates from the ConA-Sepharose were used directly for immunization (see below). For sequencing studies endoplasmin was purified further by preparative polyacrylamide gel electrophoresis. The ConA-Sepharose eluate was electrophoresed on a 10% polyacrylamide gel, the protein located by lightly staining with PAGE Blue 83, the band cut out and the protein eluted by homogenization in 10 mM-Tris-HC1, pH 7’5. The eluate was dialysed extensively against deionized water and lyophilized. Sodium dodecyl sulphate (SDS)-polyacrylamide gel analysis showed that the preparation was >99% pure endoplasmin.

Fig. 1.

A major 100 K protein (endoplasmin) in microsomal membrane preparations from a murine plasmacytoma cell. Microsomal membranes were prepared from the MOPC-315 cell line as described in Materials and Methods. The soluble (lane a) and membrane fractions (lane b) were run on SDS-polyacrylamide gels and stained with Coomassie Blue; 1× 106 cell equivalents of each fraction were analysed. The two proteins enriched in the membrane fraction (large arrows) and the molecular weight markers (95, 65, 45, 30, 20 and 15 K; small arrows) are shown.

Fig. 1.

A major 100 K protein (endoplasmin) in microsomal membrane preparations from a murine plasmacytoma cell. Microsomal membranes were prepared from the MOPC-315 cell line as described in Materials and Methods. The soluble (lane a) and membrane fractions (lane b) were run on SDS-polyacrylamide gels and stained with Coomassie Blue; 1× 106 cell equivalents of each fraction were analysed. The two proteins enriched in the membrane fraction (large arrows) and the molecular weight markers (95, 65, 45, 30, 20 and 15 K; small arrows) are shown.

Preparation of monospecific antibody to endoplasmin

Rabbits were immunized with the enriched endoplasmin preparation from MOPC-315 cells prepared as described above. For the first injection 1 mg of protein in PBS was homogenized with Freund’s complete adjuvant, 1:1 (v/v), in a total volume of 3 ml and injected subcutaneously into Dutch white rabbits. After 1 month the rabbits were re-injected subcutaneously with 1 mg of protein in 3 ml PBS without adjuvant. After a further 2 weeks all animals treated showed antibodies to endoplasmin as measured by immunoblotting analysis.

The antibodies were rendered monospecific for endoplasmin by affinity’ purification. A preparative gel was prepared as described above and electroblotted onto nitrocellulose paper (see below). The endoplasmin was localized by staining with Ponceau Red and the relevant strip cut out and blocked with neat foetal calf serum for 5 h. The strip was rinsed with undiluted antiserum on a rolling mixer for at least 2 days at 4°C. The strip was then washed and eluted with 0-2 M-glycine-HCl buffer, pH 2-8, followed by neutralization to pH 7·0 with 1 M-Tris.

Immunoblotting

The standard procedure described by Towbin et al. (1979) was used throughout these studies. l25I-labelled protein A (purchased from Amersham, UK) was used to develop the antibody bound to the nitrocellulose. Unless otherwise stated, the antibody to endoplasmin was the affinity-purified reagent described above. For quantitative comparisons of endoplasmin expression in cells and tissues, a standard curve was prepared using various amounts of MOPC-315 endoplasmin run on a standard gel and immunoblotted by the standard procedure. The resultant curve was linear up to 5 μg endoplasmin per sample. Therefore all the analyses were carried out at endoplasmin levels below this.

Immunofluorescence

Samples were treated with affinity-purified antibody to endoplasmin (see above) for 15 min at room temperature, washed with PBS/10% foetal calf serum (washing buffer) and developed with fluorescein conjugated (fl-) goat anti-rabbit immunoglobulin (Garig, Miles-Yeda) for 15 min. After washing in washing buffer the samples were mounted in 50% glycerol in PBS and examined for epifluorescence on a Zeiss microscope. Live cells (usually MOPC-315) were stained and examined as above. 3T3 fibroblasts in culture medium were allowed to settle onto polylysine-coated glass coverslips and incubated for varying times at 37°C to enable spreading to proceed. Cells were rapidly fixed by immersing the covershps in 3·5 % formaldehyde in PBS (pH adjusted to 7·0 with NaOH) for 30 min at 37°C. The coverslips were washed and the cells permeabilized by immersion in 0T % saponin in PBS for 15 min. Excess aldehyde was neutralized with neat foetal calf serum and staining carried out as described above. When cells were analysed for both endoplasmin and tubulin, they were first stained for endoplasmin, treated with neat rabbit serum for 30 min to neutralize any excess fl-Garig and treated with monoclonal rat antibody to tubulin (YOL1/2; Kilmartin et al. 1982) followed by rhodamine conjugated (rh) rabbit anti-rat immunoglobulin (Rartig, Miles-Yeda). Photographs were taken on Kodak Tri-X film.

Immunoelectron microscopy

The methods used were essentially those described by Webster et al. (1985), based on the work of Griffiths et al. (1984a,b) and Tokuyasu (1973, 1980) ; the only modifications being that the cells were fixed in 1 % glutaraldehyde in PBS for 15 min, and that after gelatin embedding, prior to freezing, the cells were infused with 2M-sucrose in 100 mM-phosphate buffer containing 8% formaldehyde (from paraformaldehyde).

Gold particles (5 nm) made according to Slot & Geuze (1985). Staphylococcal protein A (Pharmacia Fine Chemicals AB., Uppsala, Sweden) was absorbed onto the gold particles and the reagents purified by centrifugation (Geuze et al. 1981).

Gel electrophoresis

One-dimensional SDS-polyacrylamide gel electrophoresis was carried out according to the method of Laemmli (1970) using 10% polyacrylamide gels or in the microgradient gel system of Matsudaira & Burgess (1978). Gels were stained with PAGE Blue 83.

Preparation of microsomal membranes

The plasmacytoma line MOPC-315 was used. Cells were suspended in 20 vol. of phosphate-buffered saline (PBS) and saponin added to 0·01 %. After all cells had been permeabilized, as evident from light microscopy (15 min), the soluble components were separated from the ‘ghosts’ by centrifugation at 1000g for 10 min. The ghosts were then disrupted by syringing through an 18 gauge hypodermic needle into a cushion of 45% sucrose/PBS at 100000×gf for 30 min. All operations were carried out in the presence of a protease inhibitor mix (Koch et al. 1985).

Calcium binding assays

Calcium-binding proteins were detected after SDS-polyacrylamide gel electrophoresis and electroblotting into nitrocellulose by overlay with 45Ca (Koch et al. 1986). For quantitative tests endoplasmin was purifed from the MOPC-315 cell line by affinity with ConA-agarose as described above. The preparation was >90% pure and endoplasmin was the only calcium-binding protein detectable (see Fig. 1C). Samples (10µg) were spotted onto nitrocellulose paper (Schleicher and Schüll, BA 85/20) in triplicate and incubated with varying concentrations of 45CaC12 (4×106cts min−1 ml−1, adjusted to the appropriate concentration with cold CaCh) in 25 μM-Hepes, pH 7·2, 100μM-KCI and 10 μ M-MgCh. After 15 min incubation at room temperature, the spots were rinsed in deionized water for 5 s, dried out on a vacuum filter and subjected to scintillation counting. The absorption of protein onto filter was monitored by staining the spots with Ponceau Red, followed by densitometry in the reflectance mode. Background controls were carried out in exact parallel, using identical sample spots (10 μg each) of cytochrome c which shows no calcium binding.

Composition of membranes from a murine plasmacytoma line

Fig. 1 shows the protein composition of the soluble and membrane fractions from the MOPC 315 cell line. Most of the major proteins are absent from the membrane fraction with the exception of an abundant lOOx 103Afr protein and a less-abundant 75×103Mr species, which are selectively retained by the membranes. Similar results are obtained with other cell lines (data unpublished). These studies revealed the existence of a previously undescribed abundant protein in microsomal membranes.

The major membrane-associated 100×103 Mr protein is a glycoprotein

Fig. 2 shows that the 100 K (100×103Mr) protein, selectively retained in microsomal membrane preparations, binds to ConA and is eluted with a-methylman-noside. The binding to ConA is eliminated by glycosidase treatment (see below) and the protein can be metabolically labelled with radioactive sugars such as glucosamine and mannose (data not shown). Thus it was concluded that the 100 K protein is a glycoprotein (GPIQQ), which is the same as the major mammalian cell glycoprotein described previously (Koche et al. 1985).

Fig. 2.

Isolation of endoplasmin with ConA. A sample of MOPC-315 microsomal membranes was solubilized in 1 % NP40 in PBS, mixed with ConA-Sepharose and eluted with α-methyl mannoside (see Materials and Methods). The eluate (10, 5, 2 and 1 Mg protein, respectively) was analysed by SDS-polyacrylamide gel electrophoresis and stained with Coomassie Blue. The major 100 K protein species (see Fig. 1) is quantitatively extracted by the ConA. M, protein marker standards (see Fig. 1).

Fig. 2.

Isolation of endoplasmin with ConA. A sample of MOPC-315 microsomal membranes was solubilized in 1 % NP40 in PBS, mixed with ConA-Sepharose and eluted with α-methyl mannoside (see Materials and Methods). The eluate (10, 5, 2 and 1 Mg protein, respectively) was analysed by SDS-polyacrylamide gel electrophoresis and stained with Coomassie Blue. The major 100 K protein species (see Fig. 1) is quantitatively extracted by the ConA. M, protein marker standards (see Fig. 1).

Binding of ConA to the 100 Kglycoprotein is completely eliminated by digestion with endoglycosidase H

When the partially purified protein is treated with endoglycosidase H (endo H), there is a slight increase in mobility on SDS-polyacrylamide gels, similar to that observed with ovalbumin but not with a non-glycosylated protein such as serum albumin (Fig. 3). Analysis of ConA binding shows a corresponding loss of binding to the lectin. In the fully digested protein no residual ConA binding is detectable. Thus the glycosyl moiety of the glycoprotein is completely sensitive to endoglycosidase H digestion. Since glycoproteins that have proceeded into the Golgi complex acquire resistance to endo H, the sensitivity of the 100 K glycoprotein is reminiscent of glycoproteins such as HMG Co A reductase (Liscum et al. 1983), cytochrome P450 and glucosidase 1 (Brands et al. 1985), which are confined to the endoplasmic reticulum, and suggests, but does not prove, that the 100 K glycoprotein may be predominantly an ER glycoprotein.

Fig. 3.

Endoglycosidase H sensitivity of endoplasmin. Samples (5 μg) of ConA-purified glycoprotein were mixed with 5 μg BSA in 30μl 0·1M-sodium citrate, pH 5·5, and incubated at 37 °C with 0, 0·25, 0·5 and 1·0μ1 units endoglycosidase H (Brands et al. 1985). A parallel analysis was carried out with 5 ug chicken ovalbumin (Ovalb. ; Sigma) as positive control. The samples (ovalbumin set on the right half of the same gel) were analysed by SDS-polyacrylamide gel electrophoresis, stained for protein with Coomassie Blue (left panel) and the same gel stained for ConA binding (right panel) with [125]ConA (Koch & Smith, 1982).

Fig. 3.

Endoglycosidase H sensitivity of endoplasmin. Samples (5 μg) of ConA-purified glycoprotein were mixed with 5 μg BSA in 30μl 0·1M-sodium citrate, pH 5·5, and incubated at 37 °C with 0, 0·25, 0·5 and 1·0μ1 units endoglycosidase H (Brands et al. 1985). A parallel analysis was carried out with 5 ug chicken ovalbumin (Ovalb. ; Sigma) as positive control. The samples (ovalbumin set on the right half of the same gel) were analysed by SDS-polyacrylamide gel electrophoresis, stained for protein with Coomassie Blue (left panel) and the same gel stained for ConA binding (right panel) with [125]ConA (Koch & Smith, 1982).

Monospecific antibody to the 100Kglycoprotein

To examine the cellular localization of the glycoprotein, a monospecific affinity-purified antibody was prepared. In whole cell lysates the antibody recognizes the 100 K glycoprotein only (Fig. 4). Binding studies with the antibody, to intact MOPC-315 cells, failed to reveal any antigen at the cell surface, although large amounts were detected in permeabilized cells (unpublished observation) suggesting that the glycoprotein is localized intracellularly. This was confirmed by immunofluorescence, using attached fibroblasts after permeabilization (Fig. 5). The antigen was associated with large structures dispersed throughout the cytoplasm. Staining was not observed at the cell surface or within the nucleus, although the nuclear membrane was often clearly stained. At early stages of cell spreading the glycoprotein was associated with perinuclear structures that completely surrounded the nucleus, but at later stages of spreading these structures dispersed throughout the entire cytoplasm. These studies clearly confirmed the intracellular location of the glycoprotein.

Fig. 4.

Monospecific antibody to endoplasmin. Immunoblotting analysis of whole cell lysates (1) and purified glycoprotein (2) with antibodies to the 100 K glycoprotein (see Materials and Methods). Left panel: whole anti-serum; right panel: affinity-purified antibodies. The molecular weight markers (arrowed) are 95, 65, 45, 30 and 25 K, respectively.

Fig. 4.

Monospecific antibody to endoplasmin. Immunoblotting analysis of whole cell lysates (1) and purified glycoprotein (2) with antibodies to the 100 K glycoprotein (see Materials and Methods). Left panel: whole anti-serum; right panel: affinity-purified antibodies. The molecular weight markers (arrowed) are 95, 65, 45, 30 and 25 K, respectively.

Fig. 5.

Immunofluorescence analysis of 3T3 fibroblasts with monospecific antibody to endoplasrnin. Cells were prepared as described in Materials and Methods. A,B. Cells after 2h spreading. C-F. Cells after 16h spreading. A-E were stained with antibody to the glycoprotein and F is the same field as E, double-labelled for tubulin with monoclonal anti-tubulin antibody.

Fig. 5.

Immunofluorescence analysis of 3T3 fibroblasts with monospecific antibody to endoplasrnin. Cells were prepared as described in Materials and Methods. A,B. Cells after 2h spreading. C-F. Cells after 16h spreading. A-E were stained with antibody to the glycoprotein and F is the same field as E, double-labelled for tubulin with monoclonal anti-tubulin antibody.

Immunoelectron microscopic localization of the glycoprotein to the endoplasmic reticulum

The identity of the intracellular structures in which the glycoprotein was localized was determined by immunoelectron microscopy on thin frozen sections with a colloid gold marker. This is probably the most reliable approach since it provides equal access to all intracellular compartments (Tokuyasu, 1973, 1980). The results show (Fig. 6) that most of the gold particles are located within the elongated membrane structures characteristic of endoplasmic reticulum. In the example shown, which is accompanied by a diagram to facilitate identification of gold particles and membranes, 118 particles were clearly within the ER and 12 clearly outside, with about 28 being of doubtful location. When the density of gold label associated with various organelles was examined (Table 1), the ER was found to have about 100 times that in any other compartment. It was notable that labelling of the Golgi region was not above the background level, consistent with the results of the endo H sensitivity and cell fractionation experiments. Because of the localization of the glycoprotein to the ER, it is now referred to as endoplasmin.

Table 1.

Distribution of endoplasmin in the different cellular compartments

Distribution of endoplasmin in the different cellular compartments
Distribution of endoplasmin in the different cellular compartments
Fig. 6.

Immunoelectron microscopy on thin, frozen sections of MOPC-315 cells using antibody to endoplasmin. Samples were prepared as described in Materials and Methods and developed with 5 nm protein A-gold particles. The drawing on the right illustrates the regions identified as trilamellar membrane (lines) and gold particles (spots). Broken lines define regions where the membrane appears discontinuous or is not unambiguously identifiable. Bar, 0·5 μm.

Fig. 6.

Immunoelectron microscopy on thin, frozen sections of MOPC-315 cells using antibody to endoplasmin. Samples were prepared as described in Materials and Methods and developed with 5 nm protein A-gold particles. The drawing on the right illustrates the regions identified as trilamellar membrane (lines) and gold particles (spots). Broken lines define regions where the membrane appears discontinuous or is not unambiguously identifiable. Bar, 0·5 μm.

Hyperexpression of the major glycoprotein in secretory cells and tissues

Secretory cells such as plasma and pancreatic cells contain large amounts of rough endoplasmic reticulum (Alberts et al. 1983). A survey of the expression of the endoplasmin showed that it was most abundant in such cells (Fig. 7). Hyperexpression in plasma cells permits the detection of such cells by direct immunofluorescence on spleen cell smears (Fig. 8). Comparison of the cultured cell lines derived from B lymphocytes (WEH1-231) with plasma cells also shows the significant increase in the glycoprotein on development of the secretory phenotype (Fig-7).

Fig. 7.

Analysis of endoplasmin expression in secretory tissues and cells. Samples (1 mg) of protein from tissue or cell extracts were analysed by SDS-polyacrylamide gel electrophoresis and immunoblotting with antibody to endoplasmin. The immuno-positive band from each sample is shown. *, plasmacytoma cell lines.

Fig. 7.

Analysis of endoplasmin expression in secretory tissues and cells. Samples (1 mg) of protein from tissue or cell extracts were analysed by SDS-polyacrylamide gel electrophoresis and immunoblotting with antibody to endoplasmin. The immuno-positive band from each sample is shown. *, plasmacytoma cell lines.

Fig. 8.

Identification of endoplasmin-rich plasma cells by immunofluorescence on spleen cell smears. Spleen cell smears were prepared as described in Materials and Methods and developed for immunofluorescence with anti-endoplasmin antibody. Note the sub-population of large plasma cells that stam strongly compared with the weakly staining lymphocytes.

Fig. 8.

Identification of endoplasmin-rich plasma cells by immunofluorescence on spleen cell smears. Spleen cell smears were prepared as described in Materials and Methods and developed for immunofluorescence with anti-endoplasmin antibody. Note the sub-population of large plasma cells that stam strongly compared with the weakly staining lymphocytes.

The 100 K glycoprotein is the major calcium-binding protein of the ER

When total cell lysates from the MOPC-315 cell line were probed for calcium-binding proteins after SDS-polyacrylamide gel electrophoresis and electroblotting, one of the major calcium-binding proteins detected was a 100K protein (Fig. 9A). This protein was retained in a crude membrane fraction (Fig. 9B) and was shown to be endoplasmin by purification with ConA (Fig. 9C). Using the purified undenatured glycoprotein in a calcium-binding assay (Fig. 10) revealed that the binding kinetics were complex, with evidence of two classes of binding site, one involving about four calcium sites with half-maximal binding at 0·4 mM and a second involving about 8-10 calcium sites with half-maximal binding at 6 mM. Thus endoplasmin has the properties of a relatively low-affinity high-capacity calcium-binding protein localized in the endoplasmic reticulum.

Fig. 9.

Endoplasmin is a major calcium-binding protein. Calcium-binding was examined after SDS-polyacrylamide gel electrophoresis and electro-transfer to nitrocellulose paper as described in Materials and Methods. Lanes a, MOPC-315 cell lysate; b, MOPC-315 microsomal membrane fraction; c, purified endoplasmin; d, protein standards (95, 65, 45, 30, 20 and 15 K). Left panel: protein stain (Ponceau Red); right panel: 45Ca autoradiograph. The calcium-binding protein in the standards (lane d) is αr-lactalbumin.

Fig. 9.

Endoplasmin is a major calcium-binding protein. Calcium-binding was examined after SDS-polyacrylamide gel electrophoresis and electro-transfer to nitrocellulose paper as described in Materials and Methods. Lanes a, MOPC-315 cell lysate; b, MOPC-315 microsomal membrane fraction; c, purified endoplasmin; d, protein standards (95, 65, 45, 30, 20 and 15 K). Left panel: protein stain (Ponceau Red); right panel: 45Ca autoradiograph. The calcium-binding protein in the standards (lane d) is αr-lactalbumin.

Fig. 10.

Calcium-binding kinetics of endoplasmin. ConA affinity-purified endoplasmin was analysed for calcium-binding as described in Materials and Methods.

Fig. 10.

Calcium-binding kinetics of endoplasmin. ConA affinity-purified endoplasmin was analysed for calcium-binding as described in Materials and Methods.

Endoplasmin is a stress-related protein

Numerous studies have shown that cells subjected to stresses, such as treatment with calcium ionophores, tunicamycin or glucose depletion, overexpress a 100 K protein (Subjeck & Shyy, 1986; Lee, 1981). To determine whether endoplasmin was the same as this protein, it was subjected to amino-terminal sequence analysis. The sequence obtained was: Asp-Asp-Glu-Val-Asp-Val-Asp-Gly-Thr-Val-Glu-Glu-Asp-, which is identical to that obtained for the corresponding stress-related protein (Lee et al1984).

The main outcome of these studies has been the first clear demonstration that one of the major cellular proteins is a glycoprotein localized in the endoplasmic reticulum. Because this glycoprotein is the most abundant protein subunit in the ER of a wide variety of cells in all mammalian (and vertebrate) species examined to date, we refer to it as endoplasmin.

Numerous studies on microsomal membranes that are thought to be predominantly ER-derived have failed to demonstrate the presence of endoplasmin. However, we have found that endoplasmin is quite easily lost from such membrane preparations upon mechanical disruption (unpublished observations) unless special care is taken to prevent excessive vesiculation. Furthermore, endoplasmin is very sensitive to proteolysis and is rapidly degraded in the absence of a mixture of protease inhibitors. Finally, the use of a cultured cell line, a plasmacytoma, rich in ER, permits rapid and efficient isolation of ER membranes, overcoming some of the above-mentioned difficulties. The practical consequence of this ready loss of endoplasmin from ER membrane preparations is that many of those prepared by conventional procedures may not be of sufficient integrity for the analysis of ER function. We suggest that retention of endoplasmin should be used as a test for the structural integrity of ER membranes. It is interesting from this standpoint that dog pancreas microsomal membranes, which have proved particularly amenable to the analysis of protein translocation across ER membranes, do contain substantial amounts of endoplasmin (unpublished observation).

As mentioned in the Introduction, several cellular functions have been found to involve the endoplasmic reticulum and endoplasmin could participate in one or more of these. In this study we have obtained evidence that one such function might be as part of the calcium store of the ER. Current estimates indicate that the overall calcium pool of cells such as those used in this study is around 1 mM (Alberts et al. 1983). Endoplasmin exists at total concentrations of around l×103−5M but there are at least 10 calcium-binding sites in the millimolar range on the protein. Therefore, the protein could account for a significant proportion of the stored calcium within the ER. The corollary of this is that endoplasmin actually exists to increase the calcium storage capacity of the ER and thereby provide the calcium required for intracellular signalling. In fact, if the results shown in Fig. 9 reflect the situation in vivo, endoplasmin would indeed comprise the major calcium storage protein in ER and thereby perform the function ascribed to calsequestrin in sarcoplasmic reticulum (MacLennan & Holland, 1975).

The demonstration that endoplasmin is the same as one of the major stress-related proteins (GRP) in cells is interesting for at least two reasons. First, it has been claimed that the GRP is localized in the Golgi complex and also at the plasma membrane and even in the nucleus of heat-treated cells (Lin & Queally, 1982; Lin et al. 1982; Welch et al. 1983). It is clear that in the plasmacytoma cells, the only compartment showing significant amounts of the glycoprotein is the ER. Furthermore, in other cell lines such as fibroblasts, no evidence for endoplasmin in the nucleus or at the plasma membrane was obtained. Thus the variable results obtained in the previous study could reflect the anomalous behaviour of the monoclonal antibody used.

The apparent increase in the levels of endoplasmin (GRP) as a result of treatment of cells with calcium ionophores (Wu et al. 1981) is also interesting in view of the evidence that the protein might be part of the calcium store of the ER. Thus increases in endoplasmin could assist in the modulation of cellular calcium levels generally. One can speculate that the increases in endoplasmin upon glucose starvation and other stresses are also consistent with such a role since increased calcium uptake and ER hypertrophy appear to be amongst the earliest responses of cells to stress (Trump et al. 1981).

Finally, it should be emphasized that endoplasmin is one of the most abundant proteins in cells generally. Rough calculations, based on the assumption that the ER is 10 % of the total cell volume, indicate that the concentration in the ER could be as much as 10 mg ml1. Such high concentrations usually occur with proteins such as actin, which perform a structural role in cells. Thus the possibility that endoplasmin also performs some structural role in the ER should not be ignored.

In summary, these studies have shown that the endoplasmic reticulum contains at least one abundant glycoprotein, which could play a central role in the functions of this organelle, and which should provide a novel approach to the study of ER structure and function.

Alberts
,
B. A.
,
Bray
,
D.
,
Lewis
,
J.
,
Raff
,
M.
,
Roberts
,
K.
&
Watson
,
J. D.
(
1983
).
The Molecular Biology of the Cell
.
New York
:
Garland Publishing Inc
.
Berridge
,
M. J.
&
Irvine
,
R. F.
(
1984
).
Inositol triphosphate, a novel second messenger in cellular signal transduction
.
Nature, Land
.
312
,
315
321
.
Brands
,
R.
,
Snider
,
M. D.
,
Hino
,
Y.
,
Park
,
S. S.
,
Gelbain
,
H. V.
&
Rothman
,
J. E.
(
1985
).
Retention of membrane proteins by the endoplasmic reticulum
.
J. Cell Biol
.
101
,
1724
1732
.
Geuze
,
H.
,
Slot
,
J. W.
,
Van Der Ley
,
P. A.
,
Scheffer
,
R. C. T.
&
Griffiths
,
J. M.
(
1981
).
Use of colloidal gold particles in double-label immunoelectron microscopy of ultrathin frozen tissue sections..7
.
Cell Biol
.
89
,
653
665
.
Griffiths
,
G.
,
Mcdowell
,
A.
,
Back
,
R.
&
Dubochet
,
J.
(
1984a
).
On the preparation of cryosections for immunocytochemistry
.
Ultrastruct. Res
.
89
,
65
78
.
Griffiths
,
G.
,
Simons
,
K.
,
Warren
,
G.
&
Tokuyasu
,
K. T.
(
1984b
).
Immunoelectron microscopy using thin, frozen sections: Application to studies of the intracellular transport of Semliki Forest Virus spike glycoproteins
.
Meth. Enzym
.
96
,
466
485
.
Kilmartin
,
J. V.
,
Wright
,
B.
&
Milstein
,
C.
(
1982
).
Rat monoclonal antitubulin antibodies derived by using a new non-secreting rat cell line
.
J. Cell Biol
.
93
,
576
582
.
Koch
,
G. L. E.
&
Smith
,
M. J.
(
1982
).
Analysis of the glycoproteins of murine tumour cell lines with 125I-concanavalin A in two-dimensional electrophoresis gels
.
Eur. J. Biochem
.
128
,
107
112
.
Koch
,
G. L. E.
,
Smith
,
M. J.
&
Mortara
,
R. A.
(
1985
).
An abundant ubiquitous glycoprotein (GPioo) in nucleated mammalian cells
.
FEBS Lett
.
179
,
294
298
.
Koch
,
G. L. E.
,
Smith
,
M. J.
,
Twentyman
,
P.
&
Wright
,
K.
(
1986
).
Identification of a novel calcium-binding protein (CP22) in multidrug resistant murine and hamster cells
.
FEBS Lett
.
195
,
275
279
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
.
Nature, Land
.
227
,
680
685
.
Lee
,
A. S.
(
1981
).
The accumulation of three specific proteins related to glucose-regulated proteins in a temperature-sensitive hamster mutant cell line K12..7
.
cell. Physiol
.
106
,
119
125
.
Lee
,
A. S.
,
Bell
,
J.
&
Ting
,
J.
(
1984
).
Biochemical characterisation of the 94- and 78-kilodalton glucose-related protein in hamster fibroblasts
.
J. biol. Chem
.
259
,
4616
4621
.
Lin
,
J. J. C.
&
Queally
,
S. A.
(
1982
).
A monoclonal antibody that recognises Golgi-associated protein of cultured fibroblast cells
.
J. Cell Biol
.
92
,
108
112
.
Lin
,
J. J. C.
,
Welch
,
W. J.
,
Garells
,
J. I.
&
Feramisco
,
J. R.
(
1982
).
In Heat Shock: from Bacteria to Man
(ed.
M. J.
Schlessinger
,
M.
Ashburner
& A. Tissières
), pp.
267
273
.
New York
:
Cold Spring Harbor Laboratory Press
.
Liscum
,
L.
,
Cummings
,
R. D.
,
Anderson
,
R. G. W.
,
Demartino
,
G. N.
,
Goldstein
,
J. L.
&
Brown
,
M. S.
(
1983
).
3-hydroxy-3-methylglutaryl-CoA reductase: A transmembrane glycoprotein of the endoplasmic reticulum with N-linked “high-mannose” oligosaccharides
.
Proc, natn. Acad. Sci. U.SA
.
80
,
7165
7169
.
Maclennan
,
D. H.
&
Holland
,
P. C.
(
1975
).
Calcium transport in sarcoplasmic reticulum
.
A. Rev. Biophys. Bioeng
.
4
,
377
404
.
Matsudaira
,
P. T.
&
Burgess
,
D. R.
(
1978
).
SDS microslab linear gradient polyacrylamide gel electrophoresis
.
Analyt. Biochem
.
87
,
386
396
.
Palade
,
G.
(
1975
).
Intracellular aspects of the process of protein synthesis
.
Science
189
,
347
358
.
Slot
,
J. W.
&
Geuze
,
H. J.
(
1985
).
A new method of preparing gold-probes for multiple-labelling cytochemistry
.
Eur.J. Cell Biol
.
38
,
87
93
.
Streb
,
H.
,
Irvine
,
R. F.
,
Berridge
,
M. J.
&
Schulz
,
I.
(
1983
).
Release of calcium from a non- mitochondrial intracellular store in pancreatic acinar cells by inositol-l,4,5-triphosphate
.
Nature, Land
.
306
,
67
69
.
Subjeck
,
S. R.
&
Shyy
,
T. T.
(
1986
).
Stress protein systems of mammalian cells
.
Am. J. Physiol
.
250
,
C1
C17
.
Tokuyasu
,
K. T.
(
1973
).
A technique for ultrac?yotomy of cell suspensions and tissues
.
J. Cell Biol
.
57
,
551
565
.
Tokuyasu
,
K. T.
(
1980
).
Immunochemistry on ultrathin frozen sections
.
Histochem. J
.
12
,
381
403
.
Towbin
,
H.
,
Staehlin
,
T.
&
Gordon
,
J.
(
1979
).
Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications
.
Proc. natn. Acad. Sci. U.SA
.
76
,
4350
4354
.
Trump
,
B. F.
,
Beregesky
,
I. K.
&
Osornio-Vargas
,
A. R.
(
1981
).
In Cell Death tn Biology and Pathology
(ed.
I. D.
Bowen
&
R. A.
Lockshin
), pp.
209
242
.
London, New York
:
Chapman & Hall
.
Webster
,
P.
,
Dobbelaere
,
D. A. W.
&
Fawcett
,
D. W.
(
1985
).
The entry of sporozoites of Theileria parva into bovine lymphocytes in vitro. Immunoelectron microscopic observations
.
Eur.J. Cell Biol
.
36
,
157
162
.
Welch
,
W. J.
,
Garrells
,
J. I.
,
Thomas
,
G. P.
,
Lin
,
J. J. C.
&
Feramisco
,
J. R.
(
1983
).
Biochemical characterisation of the mammalian stress proteins and identification of two stress proteins as glucose- and calcium-ionophore regulated proteins
.
J. biol. Chern
.
258
,
7102
7111
.
Wu
,
F. S.
,
Park
,
V. C.
,
Rouja
,
D.
&
Martiniosi
,
A.
(
1981
).
Selective stimulation of the synthesis of an 80,000-dalton protein by calcium ionophores
.
J. biol. Chern
.
256
,
5309
5312
.