Interaction of Newly Synthesized Apolipoprotein ... - Bioscience Reports

Bioscience Reports, Vol. 19, No. 3, 1999

Interaction of Newly Synthesized Apolipoprotein B with Calnexin and Calreticulin Requires Glucose Trimming in the Endoplasmic Reticulum Utpal Tatu1,3 and Ari Helenius2 Apolipoprotein B (ApoB) is the only protein component of the low density lipoproteins (LDL) in plasma. It is a glycoprotein with a molecular mass of about 550 kDa (4536 amino acids) containing 16 N-glycans. We have studied the interaction of ApoB with two lectinlike chaperones of the Endoplasmic Reticulum (ER)—Calnexin (CN) and Calreticulin (CR). Using a co-immunoprecipitation approach we observed that newly synthesized ApoB associates with CN and CR. The interaction was transient; within 30-60 min after synthesis bulk of newly formed ApoB dissociated. Using McA Rh7777 cells expressing an N-terminal fragment of ApoB we found that inhibition of glucosidases in the ER prevented the association of CN and CR to newly synthesized ApoB. The results showed that like for association with other glycoprotein substrates, trimming of glucose residues was essential for ApoB binding to CN and CR. KEY WORDS: Apolipoprotein B; Assembly; Calnexin, Calreticulin; Chaperones; Endoplasmic Reticulum; Glycosylation. ABBREVIATIONS: ApoB, Apolipoprotein B; CN, Calnexin; CR, Calreticulin; CST, Castanospermine; ER, Endoplasmic Reticulum; LDL, Low Density Lipoprotein; MTP, Microsomal Triglyceride Transfer Protein; NEM, N-ethylmaleimide.

INTRODUCTION Biogenesis of secretory and membrane proteins takes place in the Endoplasmic Reticulum of animal cells. By virtue of the presence of various chaperone proteins, folding enzymes and oxidizing potential suited for disulfide oxidation, the lumen of the ER provides a specialized microenvironment for the folding and assembly reactions of a variety of structurally diverse proteins (Gething and Sambrook, 1992; Helenius et al., 1992). The folding of newly synthesized polypeptide chains begins co-translationally in association with the chaperone proteins and through repeated cycles of binding and dissociation the chaperones promote productive maturation of proteins. The newly formed full length proteins remain associated with the chaperones until they acquire a mature, folded conformation—competent for exit out of the ER (Hammond and Helenius, 1995). 1

Department of Biochemistry, Indian Institute of Science, Bangalore 560 012, India. Instituteof Biochemistry, Universitaetstr. 16, ETH-Zurich, CH-8092 Switzerland. 3 To whom correspondence should be addressed. 189 2

0144-8463/99/0600-OI89SI6.00/0 © 1999 Plenum Publishing Corporation

190

Tatu and Helenius

Among the chaperone proteins of the ER, CN and CR are unique in binding specifically to glycoproteins with N-linked oligosaccharides (Bergeron et al., 1994; Helenius et al., 1997). These two chaperones recognize monoglucosylated N-glycan (Glul Man8/9 GlcNAc2) as their ligand and bind therefore to a specific subset of trimming intermediate (Hammond et al., 1994). Driven by the de- and re-glucosylation cycle catalyzed by glucosidase II and glucosyl transferase activities in the ER, glycoprotein substrates are known to undergo a cycle of CN and CR binding that promotes proper folding, prevents premature oligomerization, and prevents degradation (Hammond et al., 1994; Zapun et al., 1998). ApoB is the only protein component of LDL circulating in the plasma. It is a glycoprotein with a molecular mass of about 550 kDa (4536 amino acids) containing 25 cysteines (8 disulfides) and 16 N-glycans. Like many secretory proteins, ApoB is synthesized in the ER of liver cells and is co-translationally assembled with the lipid moiety to give rise to lipoprotein particle (Boren et al., 1992). In this way the early maturation events of ApoB such as disulfide oxidation, N-glycosylation are coupled to its assembly into lipoprotein particles. The mechanism of this complex assembly process involving protein, lipid and carbohydrate moiety are only partly understood. The biogenesis of lipoproteins containing ApoB is dependent on the presence of at least one resident ER protein called as Microsomal Triglyceride transfer Protein (MTP, Gordon et al., 1995). This ER protein interacts with newly synthesized ApoB and aids its acquisition of lipids (Wu et al., 1996). Furthermore, heterologous expression studies with a carboxy terminal truncated fragment of ApoB (ApoB41) in COS cells have shown its association with CN in the ER (Patel and Grundy, 1996). Apart from this, there is very little information about the interaction of ApoB with other, more general, ER chaperones. To determine whether this unusual protein interacts with ER chaperones in the same way as more conventional glycoproteins, we studied the association of newly synthesized ApoB with both CN and CR. MATERIALS AND METHODS Chemicals and Reagents N-ethyl maleiamide, protein A sepharose beads and CST were purchased from Sigma Chemical Co., St. Louis, MO, USA. [35S]-Promix containing labeled Methionine and Cysteine was purchased from Amersham, Arlington Heights, IL, USA. Staph A powder was obtained from Zymed, So. San Francisco, CA, USA. G418 was from Media Tech., Virginia, USA. All tissue culture reagents were from Gibco BRL, Grass Island, New York, USA. Antibodies and Immunoprecipitations

Sheep antiserum to human ApoB was obtained from Boehringer Ingelheim Co., Indianapolis, INA, USA. Rabbit polyclonal antiserum to calreticulin was obtained from Affinity Bioreagents, Golden, CO, USA. The antiserum to calnexin was raised against a synthetic peptide corresponding to the C-terminal 20 amino acids of calnexin in rabbits (Hammond et al., 1994). Anti-FLAG monoclonal antibody (M2)

Role of ER-Chaperones in Apolipoprotein B Assembly

191

Fig. 1. Co-precipitation of newly synthesized ApoB100 with CN and CR in HepG2 cells. Two confluent dishes of HepG2 cells were pulsed for 15 min with radiolabeled methionine and cysteine. One dish was chased for 60 min with medium containing excess of unlabeled methionine and cysteine. After the chase the medium from the 60 min chased dish was recovered and cells were lysed with 2% CHAPS containing buffer. The lysates were spun at 1000g for 10 min to obtain the PNS. The PNS from both the dishes was divided into three aliquots and incubated at 4°C for 16hr with Protein A beads and antisera to CN, CR, or with PA beads alone as a negative control.

was purchased from Eastman Kodak Scientific Imaging Systems, New Haven, CT, USA. For immunoprecipitations the post nuclear supernatants were precleared with 10% staph A for 30 min at 4°C and then rotated for 15 hr at 4°C with 10 ul antisera against ApoB, Calnexin (50% in glycerol), calreticulin and protein A beads. Protein A control was included by adding just protein A beads without antibody to the cell lysate. The washes (two for 10 min each) were done in the cold with 0.5% CHAPS in 50 mM Hepes, 20 mM NaCl, pH 7.5 (HBS). The immunoprecipitates were analyzed by 5% SDS-PAGE (Fig. 1) or 4% SDS-PAGE (Fig. 2) and fluorography. Cell Lines and Metabolic Labeling

Human hepatoma HepG2 cell line was obtained from ATCC and were maintained in DMEM with 10% FCS, penicillin, streptomycin, Glutamine and 10% nonessential amino acids. McA Rh7777 cells expressing ApoB28F were obtained from

192

Tatu and Helenius

Dr. Greg Shelness at Bowman Gray School of Medicine, North Carolina, USA. For routine propagation, transfected McA R h 7 7 7 cells were maintained in DMEM with 10% FCS, Penicillin, Streptomycin, Glutamine and 350 mg/ml G418. For metabolic labeling, cells were washed in cold PBS and incubated for 30 min with methionine and cysteine free medium to starve the cells of these amino acids. The cells were pulsed in the same medium reconstituted with radiolabeled methionine and cysteine. The chases were done in complete media containing 5 mM unlabeled methionine and cysteine. For chases, the pulse medium was removed and transferred to a separate tube at cold. The labeled cells were washed once with complete medium containing cold amino acids and chased in the same medium for various time intervals. At the end of this time, the medium was transferred to a separate tube and used for immunoprecipitating ApoB secreted during the chase. Cells were finally washed with cold PBS containing 20 mM NEM and lysed with 2% CHAPS in HBS.

Densitometric Scanning

Quantitation of CN and CR associated ApoB was done from the band intensities using a Fuji Image Gauge program on a Fuji BAS 1600 system.

RESULTS Association of Newly Synthesized ApoB with CN and CR in HepG2 Cells

To examine whether newly synthesized ApoB interacts with chaperone proteins CN and CR during their maturation in the ER we labeled two confluent dishes of HepG2 cells with [35S]-methionine and -cysteine for 15 min. One of the two dishes was further chased for 1 hr before lysing the cells with 2% CHAPS containing buffer. After the chase the medium was recovered and kept at 4°C. The cell lysates from both the dishes and the chase medium were divided into four aliquots. One aliquot from each sample was precipitated with anti-ApoB antiserum, the second with antiCN antiserum, the third with anti-CR antiserum and the fourth aliquot was incubated with Protein A beads without any serum to see non-specifically precipitating labeled proteins. As shown in Fig. 1, in samples labeled for 15 min the anti-ApoB immunoprecipitates (lane 1) showed the presence of a band of about 500 kDa corresponding to full length ApoB 100. A number of other lower molecular weight bands were also seen below the ApoB 100 band. Most of these may correspond to the incompleted nascent chains that were initiated during the radioactive pulse but we do not show any evidence for this. Anti-CN and anti-CR precipitates of this sample (lane 2 and 3 respectively) also showed bands corresponding to ApoB 100 along with a number of other labeled bands corresponding to other glycoprotein substrates of CN in HepG2 cells (see Ou et al., 1993). Quantitation by densitometric scanning showed that about 20% of labeled ApoB 100 was precipitated with anti-CN or anti-CR antisera in HepG2 cells.

Role of ER-Chaperones in Apolipoprotein B Assembly

193

In samples chased for 1 hr, the anti-ApoB precipitates (lane 4) once again showed a band corresponding to full length ApoB100 while the bands corresponding to partially completed nascent chains were no longer visible after the chase. The intensity of ApoB100 was diminished after the chase, partly accounted for by the secretion of a small amount of the newly synthesized ApoB100. As shown in lane 7 a small amount of labeled ApoB100 was found in the medium of cells chased for an hour. The anti-CN and anti-CR precipitates of cell lysates from samples chased for 1 hr (lanes 5 and 6 respectively) did not show any band corresponding to ApoB100 indicating that it had dissociated from these chaperone during the chase period. Anti-CN and anti-CR precipitates from the medium of these samples also did not show any band corresponding to ApoB100 (lanes 8 and 9 respectively). A background band of about 200 kDa was found in all the immunoprecipitates from the medium of cells chased for 1 hr, including in the Protein A controls. It was clear that newly synthesized ApoB100 transiently associates with the ER-chaperones CN and CR in HepG2 cells.

Association of Newly Synthesized ApoB with CN and CR Requires Glucose Trimming in the ER

Studies performed by Ingram and Shelness (1997) have shown that amino terminal fragments of ApoB containing about 21% of the sequence (ApoB28F) can be incorporated into lipoprotein particles thus serving as convenient models to study many aspects of the biosynthesis of lipoproteins containing ApoB. We have used MacArdle RH7777 cells permanently transfected with ApoB28F to investigate the binding of ApoB to CN/CR and its dependence on glucose residues present on Nglycans. The presence of C-terminal FLAG tag on the ApoB28 fragment allowed efficient immunoprecipitation of this fragment using anti-FLAG antibodies. MacArdle RH7777 cells grown to confluency were labeled for 15 min with [35S]methionine and -cysteine in the presence or absence of 1 mM CST. CST is an inhibitor of ER-glucosidases (Elbein, 1991). The cells were lysed in 2% CHAPS containing buffer and the lysates were immunoprecipitated with anti-FLAG, anti-CN or antiCR. An aliquot was used as a Protein A control without added antibody. As shown in Fig. 2, anti-Flag precipitated a band of about 140 kDa corresponding to ApoB28 fragment. It was precipitated both in the absence (lane 1) and the presence (lane 5) of CST. ApoB28F synthesized in the presence of CST gave diffuse band that migrated slightly slower than the control indicating the presence of additional glucose residues. Anti-CN and anti-CR antisera also precipitated ApoB28F from the untreated dish (lanes 2 and 3) but failed to precipitate the apo-protein in cells treated with CST (lanes 6 and 7). Lanes 4 and 8 represent Protein A controls in which the lysates were incubated in the absence of any antibody. The results showed that like the full length ApoB, the amino terminal fragment Apo28F also associates with CN and CR in the ER and their binding requires trimming of glucose residues by glucosidases in the ER.

194

Tatu and Helenius

Fig. 2. Binding of ApoB28F to CN and CR is CST sensitive. Two dishes of McA Rh7777 cells permanently transfected with an amino terminal fragment of ApoB100 (ApoB28F) were used. One dish was preincubated and labeled for 15 min in the presence of CST while the other was labeled without CST for 15 min. Both dishes were lysed with buffer containing 2% CHAPS and the lysates were divided into four parts. These were immunoprecipitated with antibodies to FLAG, CN and CR. The fourth aliquot was used as a protein A control.

DISCUSSION ApoB is a highly unusual glycoprotein not only through its huge size but also through its affinity for lipids which allows it to assemble into lipoprotein particles co-translationally in the ER. Thus its maturation not only involves the folding, formation of intrachain disulfides and glycosylation but specific association with nonmembrane lipids in the ER. It was of interest to determine whether any of the normal "household" chaperones of the ER, such as the lectin-like CN and CR, participate in the maturation of this protein. CN and CR chaperones are known to be involved in the folding of virtually all other glycoproteins synthesized in the ER of HepG2 and other cells (Hammond et al., 1994; Nauseef et al., 1995; Ou et al., 1993; Peterson et al., 1995; Vassilakos et al., 1996). We found that ApoB associates with both of these chaperones. Binding probably starts already on the nascent chains but is transient. Thirty minutes after synthesis, only a small amount of ApoB remained associated with CN (not shown). By 60 min no binding was observed with either chaperone. The binding is dependent on glucose trimming as previously observed for most other substrate proteins that bind

Role of ER-Chaperones in Apolipoprotein B Assembly

195

to these two lectin-like folding factors. The synthesis of a ApoB molecule takes about 14 min and its assembly with lipids is known to begin co-translationally (Bostrom et al., 1986). Association of CN and CR thus coincides the assembly of ApoB into lipoprotein particles. MTP, the other resident ER protein known to interact with ApoB, also binds at an early stage (maximal binding at 10 min after synthesis) and participates directly in its assembly with lipids (Wu et al., 1996). Together the results suggest an overlap in the time of association of newly synthesized ApoB with CN/CR and MTP. It remains to be seen if ApoB interacts simultaneously with these two functionally distinct groups of chaperones. It is possible, for example, that CN and CR binding may facilitate MTP dependent transfer of triglycerides onto newly synthesized ApoB. Of the 19 potential glycosylation sites on ApoB 100, 16 are known to be utilized (Schumaker et al., 1995) providing sufficient binding sites for both of the two homologous lectins. ApoB28F, the N-terminal fragment of ApoB used in this study, contains 28% of mass of ApoB (Ingram and Shelness, 1994) this includes two 7Vglycosylation sites. It is not clear as to how many glycans are actually added to ApoB28F, but it is known that more than a single glycan is needed to provide stable binding of calnexin to a glycoprotein (Cannon et al., 1996). The intensity of the ApoB 100 bands indicated that no more than 20% of labeled full length ApoB 100 was bound to CN and CR. ApoB28Fon the other hand showed somewhat greater binding. The reason for this difference is not clear but may have to do with the different mechanisms by which ApoB 100 and ApoB28F are assembled into lipoproteins in human hepatoma HepG2 cells and rat hepatoma McA Rh7777 cells respectively. While in rat hepatoma ApoB28F is assembled in a two step process similar to ApoB48, the second step does not occur in the biogenesis of ApoB100 containing lipoproteins in HepG2 (Boren et al., 1994). This initial study shows that, similar to conventional secretory glycoproteins, interaction of ApoB with CN and CR is dependent on glucose trimming in the ER. The significance of this interaction in LDL biogenesis is however unclear. A key question is—whether the lipoprotein assembled in the presence of CST is qualitatively and quantitatively similar to that in its absence. Clearly, understanding the details of association of ApoB with CN and CR is important not only to further understand the roles of these chaperones but also to understand the complex process of assembling a lipoprotein particle. ACKNOWLEDGEMENTS U.T. would like to thank Dr. Greg Shelness, Bowman Gray School of Medicine, North Carolina, USA and Dr. Michael Kashgarian, Department of Pathology, Yale University, New Haven, CT, USA for their encouragement and generous gifts of cell lines and reagents used in this study. REFERENCES 1. Bergeron, J. J. M., Brenner, M. B., Thomas, D. Y., and Williams, D. B. (1994) Trends Biochem. Sci. 19:124-128. 2. Boren, J., Graham, L., Wettesten, M., Scott, )., White, A., and Olofsson, S.-O. (1992) J. Biol. Chem. 267:9858-9867.

196

Tatu and Helenius

3. Boren, J., Rustaeus, S., and Olofsson, S.-O. (1994) J. Biol. Chem. 269:25879-25888. 4. Bostrom, K., Wettesten, M., Boren, J., Bondjers, G., Wiklund, O., and Olofsson, S.-O. (1986) J. Biol. Chem. 261:13800-13806. 5. Cannon, K. S., Hebert, D. N., and Helenius, A. (1996) J. Biol. Chem. 271:14280-14284. 6. Elbein, A. D. (1991) FASEB J. 5:3055-3063. 7. Gordon, D. A., Wettarau, J. R., and Gregg, R. E. (1995) Trends Cell Biol. 5:317-321. 8. Gething, M.-J., and Sambrook, J. (1992) Nature (Land.) 355:33-45. 9. Hammond, C. H., and Helenius, A. (1995) Curr. Opin. Cell Biol. 7:523-529. 10. Hammond, C. H., Braakman, I., and Helenius, A. (1994) Proc. Natl. Acad. Sci. (USA) 91:913-917. 11. Helenius, A., Marquardt, T., and Braakman, I. (1992) Trends Cell Biol. 2:227-231. 12. Helenius, A., Trombetta, E. S., Hebert, D. N., and Simons, J. F. (1997) Trends Cell Biol. 7:193-200. 13. Ingram, M. F., and Shelness, G. S. (1997) J. Biol. Chem. 272:10279-10286. 14. Nauseef, W. M., McCormick, S. J., and Clark, R. A. (1995) J. Biol. Chem. 270:4741-4747. 15. Ou, W.-J., Cameron, P. H., Thomas, D. Y., and Bergeron, J. J. M. (1993) Nature (Lond.) 364:771776. 16. Patel, S. B., and Grundy, S. M. (1996) J. Biol. Chem. 271:18686-18694. 17. Peterson, J. R., Ora, A., Nguyen, V. A., and Helenius, A. (1995) Mol. Biol. Cell. 6:1173-1184. 18. Schumaker, V. N., Phillips, M. L., and Chatterton, J. E. (1995) Adv. Protein Chem. 45:205-244. 19. Vassilakos, A., Cohen-Doyle, M. F., Peterson, P. A., Jackson, M. R., and Williams, D. B. (1996) EMBO J. 15:1495-1506. 20. Wu, X., Zhou, M., Huang, L.-S., Wetterau, J., and Ginsberg, H. N. (1996) J. Biol. Chem. 271:1027710278. 21. Zapun, A., Petrescu, S. M., Rudd, P. M., Dwek, R. A., Thomas, D. Y., and Bergeron, J. J. (1997) Cell 88:29-38.