Effects of Castanospermine and 1-Deoxynojirimycin on Insulin Receptor Biogenesis EVIDENCE FOR A ROLE OF GLUCOSE REMOVAL FROM CORE OLIGOSACCHARIDES*

The insulin proreceptor is a 190-kDa glycoprotein that is processed to mature (Y (135-kDa) and /3 (95-kDa) subunits. In order to determine the role of carbohydrate chain processing in insulin receptor biogenesis, we investigated the effect of inhibiting glucose removal from core oligosaccharides of the insulin proreceptor with glucosidase inhibitors, castanospermine and 1-deoxynojirimycin. Cultured IM-9 lymphocytes treated with inhibitors had 50% reduction in surface insulin receptors as demonstrated by ligand binding, affinity cross-linking with ‘2sI-insulin, and lactoperoxidase/Na’2SI labeling studies. Degradation rates of surface labeled receptors were similar in both control and inhibitor-treated cells (tlk = 5 h); thus, accelerated receptor degradation could not account for this reduction. Biosynthetic labeling experiments with t3H]leucine and t3H]rnannose identi- fied an apparently higher molecular size proreceptor (-205 kDa) that failed to show the characteristic decline with time as seen in the normal 190-kDa prore- ceptor. Along with this finding, the biosynthetic label appearing in the mature subunits was reduced in these inhibitor-treated cells. Endoglycosidase H treatment of both precursors produced identical 170-kDa bands. Carbohydrate chains released from the 205-kDa precursor by endoglycosidase H migrated in the same position

The insulin proreceptor is a 190-kDa glycoprotein that is processed to mature (Y (135-kDa) and / 3 (95-kDa) subunits. In order to determine the role of carbohydrate chain processing in insulin receptor biogenesis, we investigated the effect of inhibiting glucose removal from core oligosaccharides of the insulin proreceptor with glucosidase inhibitors, castanospermine and 1deoxynojirimycin.
Cultured IM-9 lymphocytes treated with inhibitors had 50% reduction in surface insulin receptors as demonstrated by ligand binding, affinity cross-linking with '2sI-insulin, and lactoperoxidase/Na'2SI labeling studies. Degradation rates of surface labeled receptors were similar in both control and inhibitor-treated cells (tlk = 5 h); thus, accelerated receptor degradation could not account for this reduction. Biosynthetic labeling experiments with t3H]leucine and t3H]rnannose identified an apparently higher molecular size proreceptor (-205 kDa) that failed to show the characteristic decline with time as seen in the normal 190-kDa proreceptor. Along with this finding, the biosynthetic label appearing in the mature subunits was reduced in these inhibitor-treated cells. Endoglycosidase H treatment of both precursors produced identical 170-kDa bands. Carbohydrate chains released from the 205-kDa precursor by endoglycosidase H migrated in the same position as the Glc2-3ManeGlcNAc standards when separated by high performance liquid chromatography, whereas the 190-kDa proreceptor oligosaccharides migrated similar to the Man7-9GlcNAc chains. Although the mature subunits of control and inhibitor-treated cells demonstrated equal electrophoretic mobility, the endoglycosidase H-sensitive oligosaccharides of the mature subunits in treated cells also contained residues that migrated similar to the Gl~~-~Man~GlcNAc standards. Thus, glucose removal from core oligosaccharides is apparently not necessary for the cleavage of the insulin proreceptor, but does delay processing of this precursor, which probably accounts for the reduction in cell-surface receptors.
The insulin receptor is an integral membrane glycoprotein * Portions of this paper have been presented in abstract form (1).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-$ To whom correspondence should be addressed Diabetes Branch, National Institute of Diabetes and Digestive and Kidney Diseases, Bldg. 10, Rm. 88-243, National Institutes of Health, Bethesda, MD 20892. which is synthesized in the endoplasmic reticulum as a singlechain polypeptide precursor of 190 kDa (2, 3). This proreceptor contains oligosaccharides exclusively of the N-linked highmannose-type. After translocation to the Golgi complex, it is proteolytically cleaved to yield pre-a (120-kDa) and pre-P (80-kDa) subunits (4). Additional post-translational events are the processing of a number of the high-mannose carbohydrate chains to complex-type chains (2) and fatty acylation (5). The mature a (135-kDa) and P (95-kDa) subunits are inserted into the plasma membrane as disulfide-linked heterodimers (6,7).
The N-glycosylation of the insulin receptor appears to follow the general pathway described for many secretory and membrane proteins in animal cells (8,9). After the initial transfer of a high-mannose oligosaccharide (Glc3Man9Glc-NAc2) from a lipid carrier to the nascent polypeptide chain, a number of processing steps take place. Glucosidases I and I1 remove the 3 glucose residues and mannosidases I and I1 may remove all but 3 mannose residues. Following the trimming of glucoses and mannoses, complex-type chains are generated by the addition of N-acetylglucosamine, galactose, fucose, and sialic acid.
The recent identification of the cDNA of the insulin receptor has revealed 15 potentia1 N-linked glycosylation sites in the a subunit and 4 on the presumed extracellular portion of the subunit (10,11). However, the biological significance of N-glycosylation of the insulin receptor is unclear. Complete inhibition of glycosylation by tunicamycin, which prevents the transfer of the core oligosaccharide, blocks any further processing of the proreceptor (12). On the other hand, inhibition of Golgi mannosidase I1 by swainsonine does not appear to have any effect on the translocation and function of the insulin receptor (13). Thus, whereas complete lack of N-linked carbohydrate chains is incompatible with normal receptor processing, abnormalities of late carbohydrate processing may not be sufficient to alter further receptor maturation and function. Castanospermine, a plant alkaloid, and l-deoxynojirimycin, a glucose analogue antibiotic, inhibit glucosidases I and 1/11, respectively. The oligosaccharide chains of glycopeptides produced in the presence of either inhibitor retain 1-3 glucose residues and a variable number of mannoses (14,15). The preservation of the glucosylated high-mannose oligosaccharide on the insulin receptor precursor allowed us to investigate the significance of oligosaccharide trimming for insulin receptor structure and function. We have found that in the presence of these inhibitors, cultured lymphocytes produce an abnormal proreceptor of approximately 205 kDa. The processing of this precursor was delayed and probably accounted for the observed 50% reduction in cell-surface insulin receptors of inhibitor-treated cells. All reagents used for SDS1-polyacrylamide gel electrophoresis were purchased from Bio-Rad. All other materials were reagent-grade.
Cell Culture and Treatment with Glucosidase Inhibitors-Human IM-9 lymphocytes were grown in RPMI 1640 with 25 mM HEPES and 10% fetal bovine serum at 37 "C. Cells at stationary phase of growth were used for all studies. Either castanospermine or l-deoxynojirimycin was added to the culture medium to a final concentration of 100 pg/ml and 7.5 mM, respectively. The cells were preincubated in the presence of the inhibitors at 37 'C for 24 h prior to binding assays, affinity labeling, or cell-surface iodination. In the case of biosynthetic labeling, the cells were preincubated with the appropriate inhibitor for 4 h prior to labeling. The inhibitors were present throughout the labeling procedures.
Affinity LabeZing-1251-Insulin was cross-linked to IM-9 lymphocytes as previously described (17). Control and treated lymphocytes (3-5 X lo7 cells/ml) were incubated with '251-insulin (5-10 ng/ml) with and without labeled insulin M) for 30 min at 15 "C in lymphocyte buffer. The cells were then washed and resuspended in lymphocyte buffer without bovine serum albumin, and bound insulin was cross-linked to the receptor with disuccinimidyl suberate (1 mM) for 10 min at 4 "C. The reaction was quenched with 100 mM Tris-HCl buffer.
Cell-surface Labeling-Control and treated lymphocytes (2 X 10' cells/ml) were suspended in 100 ml of PBS with glucose (20 mM) and iodinated with Na"' 1 (2 mCi), lactoperoxidase (2 mg), and glucose oxidase (200 units) (18 with phenylmethylsulfonyl fluoride (2 mM) and aprotinin (1.5 trypsin inhibitor units/ml). Solubilization was performed at 4 "C for 30 min, and nondissolved material was sedimented by ultracentrifugation at 200,000 X g for 1 h. The insulin receptors were immunoprecipitated with human autoantibodies (serum B-7) as previously described (19). The antireceptor or the normal nonimmune sera were added directly to the detergent extract (1:lOO dilution). In the case of biosynthetic labeling, the extracts were treated with S. aureus cell suspension or protein A-Sepharose (200 pl/l-ml sample) for 1 h prior to the addition of antibodies. After 12-14 h at 4 "C, the receptor-antibody complexes were immunoadsorbed with S. aureus cells or protein A-Sepharose bent by boiling in sample buffer (2% SDS, 0.1 M dithiothreitol, 0.002% (100 pl) for 2 h. The receptors were released from the immunoadsorbromphenol blue, 10% glycerol, and 10 mM phosphate) for 5 min. The receptor components were then separated by SDS-polyacrylamide gel electrophoresis (7.5%) according to the method of Laemmli (20). Gels with tritium-labeled receptors were pretreated with EN3HANCE prior to drying. Autoradiography or fluorography was performed by exposure at -70 "C of Kodak X-Omat AR film to the dried gels (19). Quantitative measurements of radioactivity in receptor bands were obtained by excising and counting in a y-counter for '''1 label or in a scintillation counter for tritium label after eluting with 3% Protosol in Econofluor for 18 h at 37 "C.
Endoglycosidase H Treatment-Enzymatic digestion with endoglycosidase H was performed as previously described (2). After immunoprecipitation, labeled receptors were recovered from immunoadsorbent by boiling in 1% SDS, 0.01 M dithiothreitol, and 10 mM phosphate buffer, pH 7.0. Samples were then diluted 1:4 with 0.3 M citrate buffer, pH 5.5, and endoglycosidase H was added to a final concentration of 0.2-0.5 unit/ml. Digestion was performed for 6 h at 37 "C in the presence of phenylmethylsulfonyl fluoride (1 mM) and pepstatin (10 p~) and terminated with ice-cold 10% trichloroacetic acid. Precipitates were washed twice with ethyl ether/ethanol (l:l, v/ v), and analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions as described above.
HPLC of Endoglycosidase H-sensitive Oligosaccharides-Analysis of labeled high-mannose oligosaccharides present in insulin receptors was performed as described previously (21). Dried gel bands containing [3H]mannose-labeled receptors were excised, washed in 20% methanol, and incubated in 0.15 M citrate buffer, pH 5.5. Endoglycosidase H was added to a final concentration of 0.2-0.5 unit/ml, and the digestion was performed for 18 h at 37 "C. After endoglycosidase H treatment, the supernatant was diluted in acetonitrile (HPLCgrade). Chromatography was performed using a Model 680 automated gradient controller, an M6000A pump, an M45 pump, and a Model U6K injection valve (Waters Associates, Millipore Coy.). A pBondapak-NH2 column, equilibrated with 65% acetonitrile, 35% water (v/v), was used to separate the oligosaccharides; the samples were fractionated at a flow rate of 1 ml/min at ambient temperature. Fractions (1 ml) were collected and mixed with scintillation fluid, and their radioactive content was measured. Labeled oligosaccharide standards were run adjacent to all samples.

RESULTS
Effect of Glucosidase Inhibitors on '251-Insulin Binding-Insulin binding to IM-9 lymphocytes was determined after a 24-h preincubation at 37 "C in the presence or absence of inhibitors. Cells treated with 100 pg/ml castanospermine had approximately a 50% reduction in specific insulin binding (Fig. 1). Competition binding with unlabeled insulin in treated and untreated cells revealed similar one-half maximal displacement (-5-10 ng/ml insulin). Thus, reduction in binding was attributed to a decrease in receptor number, not affinity. The effect of castanospermine was dose-dependent, with maximal reduction in binding observed at a concentration of 100 pglml. The earliest detectable decrease in binding occurred at 12 h of preincubation; the effect was maximal a t 18-24 h and persisted for up to 36 h at 37 "C (data not shown).
Treatment of cells with 1-deoxynojirimycin (7.5 mM) also reduced specific insulin binding t o 50% of maximal (Fig. 2). Lower concentrations had no effect on IM-9 cells, and higher doses were not examined due to the limited availability of this inhibitor. Both castanospermine and 1-deoxynojirimycin did not affect cell viability as determined by trypan blue dye  Affinity and Cell-surface Labeling of the Insulin Receptor-To investigate further the nature of cell-surface insulin receptors after treatment with castanospermine, affinity crosslinking with '251-insulin was performed. After a 24-h preincubation in the presence or absence of the inhibitor, cells were incubated with 1251-insulin and cross-linked with disuccinimidyl suberate. The labeled receptors were analyzed by SDS-polyacrylamide gel electrophoresis and autoradiography. Affinity labeling confirmed the reduction of specific insulinbinding sites in castanospermine-treated cells (Fig. 3). Quantitation of the ligand-bound a subunit showed an approximate 50% decrease in the labeled band from treated cells. The electrophoretic mobility of the 135-kDa band, however, was similar to control (Fig. 3). Because of the possibility that a population of receptors might go undetected by equilibrium binding or cross-linking studies, surface labeling experiments were undertaken. Following a 24-h preincubation in the presence or absence of castanospermine, IM-9 cells were labeled by the lactoperoxidase/Na'*'I method, and the insulin receptor was immunoprecipitated with antireceptor antibodies. kDa) and @ (95-kDa) receptor subunits in control and treated cells (Fig. 3). The labeling of both receptor subunits in treated cells was clearly reduced. Quantitation of the radioactivity present in the receptor bands showed that this reduction was approximately 50%, in agreement with the estimates obtained in the '251-insulin binding and affinity labeling studies.
Degradation Rate of Cell-surface Insulin Receptor-Having demonstrated a decrease in cell-surface insulin receptors in the presence of glucosidase inhibitors, we wished to determine whether this could be due to accelerated receptor degradation. IM-9 cells were incubated for 24 h with and without castanospermine and then labeled by the lactopero~idase/Na'~~I method. After washing, the cells were returned to control or inhibitor-supplemented culture media and incubated at 37 "C for up to 9 h. At various time points, aliquots of cells were solubilized, and the insulin receptors were analyzed by SDS electrophoresis and autoradiography. Degradation rates of the receptor subunits were estimated by quantitation of the decrease in their radioactive content. The degradation rates of both subunits from control and castanospermine-treated cells were essentially identical (tlh -5 h (Fig. 4)). Thus, degradation of surface receptors was unaffected by the presence of inhibitor. The tH = 5 h observed differs from that reported by Kasuga et al. were identified by electrophoresis and fluorography. By 1 h of chase, the insulin proreceptor was prominently seen in both control and castanospermine-treated cells; and by 4 h of chase, both proreceptor and mature a and @ subunit bands were seen (Fig. 5). Several points are of note. First, the proreceptor, normally a 190-kDa band, as seen in control, had somewhat less mobility and migrated as a 205-kDa band in castanospermine-treated cells. At the 4-h chase point, the proreceptor band was more prominent in the castanospermine-treated cells than in control, and the a and p subunit bands were reciprocally less intense (Fig. 5). An essentially identical situation was found when [3H]mannose was used as the label in the same pulse-chase design (Fig. 6) except that the labeling of the 205-kDa proreceptor was more intense with this monosaccharide than with [3H]leucine. As in the cross-linking and surface labeling experiments, the a and p subunits from control and castanospermine-treated cells migrated to a similar position in the gel (compare Fig. 3 with Figs. 5 and 6). A continuous labeling study with [3H]mannose was performed in 1-deoxynojirimycin-treated cells (7.5 mM) for up to 18 h. In the treated cells, a larger molecular size precursor was also observed, similar to that seen with castanospermine, i.e. -205 kDa. The intensity of the label in this band was greater than the control proreceptor, and the label in the a and B subunit bands was reciprocally decreased (data not shown).

Endoglycosidase H Treatment and HPLC Analysis of Oli-
gosaccharide Chains-To demonstrate the role of the oligosaccharide chains in the generation of the abnormal 205-kDa insulin proreceptor, endoglycosidase H digestion was performed on the precursors of control and castanosperminetreated cells. This enzyme cleaves specifically high-mannosetype oligosaccharides, which account for all the carbohydrate chains of the insulin proreceptor. After [3H]leucine pulse labeling and a 1-h chase, cells were solubilized; the immunoprecipitated receptors were then treated with endoglycosidase H, and receptor subunits were separated by SDS electrophoresis. As shown in Fig. 7 the abnormal 205-kDa precursor was endoglycosidase H-sensitive, similar to the control 190-kDa proreceptor. A 170-kDa band, representing the nascent polypeptide (24), resulted from digestion of both precursors. Thus, the difference in the molecular size of the abnormal prorecep- and chased for 1 h; and the insulin proreceptor was isolated as described for Fig. 6. The receptor precursor bands were excised from the gel and digested with endoglycosidase H; the released high-mannose oligosaccharides were analyzed by HPLC as described under "Experimental Procedures." 3H radioactivity is plotted as a function of fraction number. The proreceptor oligosaccharides from control cells are shown (upper), as are the proreceptor oligosaccharides from castanospermine-treated cells (lower). The elution positions of oligosaccharide standards, profiled before and after each sample, are shown by arrows; M, represents Man,GlcNAcl, Gy represents Glc,Man9GlcNAcl. FRACTION NUMBER tor can be solely attributed to carbohydrate chain alterations induced by the inhibitor.
In order to verify the effects of glucosidase inhibition induced by castanospermine as well as to characterize the structure of the oligosaccharides produced under these conditions, the proreceptor carbohydrate chains were isolated and subjected to HPLC analysis. IM-9 lymphocytes were pulse-labeled with [3H]mannose; and at 1 h of chase, the labeled receptors were isolated. After SDS-polyacrylamide gel separation, the proreceptor bands were excised and treated with endoglycosidase H; the released oligosaccharides were separated by HPLC on an NH,-derivatized column. The major components of the proreceptor oligosaccharides from control cells eluted in the same position as the Man8_gGlcNAc standards (Fig. 8, upper), as previously reported (21). It should be noted that endoglycosidase H cleaves high-mannose chains between the two inner, pl-4-linked N-acetylglucosamine residues; and therefore, the released oligosaccharides contain only 1 N-acetylglucosamine residue. A very different elution pattern was observed with the proreceptor oligosaccharides from castanospermine-treated cells (Fig. 8, lower); in this case, the major components migrated in the same position as the G l~~~~M a n~G l c N A c standards. The HPLC procedure we have employed separates highmannose chains essentially by the number of hexoses. Thus, the species which migrated with the GlcsMansGlcNAc standard has very likely this hexose composition since this is the longest oligosaccharide (12 hexoses) known to be transferred in animal cells (8). On the other hand, the peak eluting with the GlczMan9GlcNAc marker corresponds to a HexllGlcNAc chain which could be composed of either Glc,Man9GlcNAc or Glc3Man,GlcNAc. The latter possibility is quite likely since removal of 1 or 2 mannose residues by mannosidase I can occur on fully glucosylated chains as reported for other glycoproteins synthesized under the influence of castanospermine (25). In either case, it is clear that castanospermine treatment had a dramatic effect on the structure of the proreceptor oligosaccharides.
The mature a and p subunits of the insulin receptor are known to contain a mixture of high-mannose and complextype oligosaccharides (18). Although the electrophoretic mobility of the receptor subunits synthesized in the presence of castanospermine was not altered, endoglycosidase H-sensitive oligosaccharides were subjected to a similar HPLC analysis as that performed with the proreceptor chains. The chromatographic pattern of the oligosaccharides of the a and subunits from control cells was similar to that reported previously (21) and showed the presence of chains with a generally lower number of mannose residues (Mans-gGlcNAc) than those of the proreceptor (Fig. 9, A and C). In contrast, analysis of the oligosaccharides from castanospermine-treated cells demonstrated the presence of two larger components which migrated with the Glc3Man9GlcNAc and Glc2MangGlcNAc markers ( Fig. 9, B and D) in a similar fashion to that observed in the proreceptor chains. The presence of these glucosylated chains was observed in both the a and p receptor subunits. Thus, the abnormal glucosylation of the proreceptor oligosaccharides induced by castanospermine persists in some high-mannose chains of the mature LY and @ subunits.

DISCUSSION
Our data show that glucosidase inhibition in IM-9 lymphocytes with two chemically dissimilar compounds, castanospermine and 1-deoxynojirimycin, induces a reduction (~5 0 % ) in cell-surface insulin receptors. This reduction was demonstrated with '251-insulin binding, affinity labeling of the a subunit, and cell-surface labeling of the a and p receptor subunits. The usefulness of inhibitors is frequently limited by their toxicity and adverse effects on cell metabolism. In this regard, both castanospermine and 1-deoxynojirimycin were excellent tools to probe the role of glucosidases since their use was not accompanied by any alteration of cell viability or protein synthesis.
Given that the insulin receptor is subject to constant turnover and its concentration at the cell surface is the result of a dynamic equilibrium (22), a reduction in the number of receptors could be due to either an enhanced degradation rate or a decreased synthesis rate. The fact that the degradation rate of the receptor subunits was not modified in the presence of the glucosidase inhibitors suggested an effect of these compounds on the biosynthetic pathway. Indeed, biosynthetic labeling studies with [3H]leucine and [3H]mannose demonstrated the production of an abnormal, higher molecular size insulin proreceptor (205 uersus 190 kDa). This large proreceptor failed to decline with time at the same rate as the proreceptor from control cells, and the generation of the mature CY and p subunits was proportionally reduced. Therefore, a delayed processing of the abnormal proreceptor appeared to account for the reduction of insulin receptors on the plasma membrane in inhibitor-treated cells.
The larger molecular mass of the proreceptor in inhibitortreated cells was due solely to changes in the carbohydrate chains. Endoglycosidase H digestion of the [3H]leucine-labeled proreceptors from control and treated cells generated polypeptides which migrated electrophoretically in identical position (170 kDa). Furthermore, molecular sizing by HPLC of the carbohydrate chains of the insulin proreceptor demonstrated the presence of Hexlo-,,GlcNAc2 chains in treated cells, whereas the largest chain found in control cells was Hex9GlcNAcz. Although monosaccharide analysis was not performed, the largest oligosaccharide detected in treated cells, Hex,,GlcNAc,, must correspond to Glc3Man9GlcNAc2; the identification of this chain confirms that this is the initial form of high-mannose oligosaccharide in the insulin proreceptor, and the rapid removal of this oligosaccharide in the absence of glucosidase inhibitors accounts for the failure of its detection in previous studies (21). The other major oligosaccharide found in the proreceptor of treated cells was a Hexl1G1cNAc2 species; it is likely that its composition is Glc3MansGlcNAc2 rather than Glc2Man9GlcNAc, since removal of 1 or 2 mannose residues has been reported on fully glucosylated chains in other glycoproteins synthesized in the presence of castanospermine (25). Previous work has shown the presence of a mixture of complex-type and high-mannose chains in the mature CY and p subunits of the insulin receptor (19). Furthermore, the recent identification of the nucleotide sequence of a cDNA clone of the insulin receptor has predicted 15 potential Nlinked sites in the CY subunit and 4 on the extracellular portion of the p subunit (10, ll), although the number of actual glycosylation sites and the type of chains attached to each site are still unknown. Our data show that the mature CY and p subunits generated in the presence of glucosidase inhibitor had identical electrophoretic mobility as the mature subunits in control cells. However, the subunits in treated cells showed a somewhat greater sensitivity to endoglycosidase H (data not shown). Nevertheless, HPLC analysis of high-mannose oligosaccharides from treated cells also demonstrated the presence of glucosylated chains (Hex10-12GlcNAc2) in the CY and p Effects of Glucosidase Inhibitors on the Insulin Receptor subunits. Therefore, it appears that at least some glucosylated high-mannose chains persist in the a and p subunits after processing. The normal electrophoretic mobility of the intact subunits may be explained if the high-mannose chains make a smaller contribution to the electrophoretic behavior as compared to the complex-type chains.
Our results are consistent with the conclusion that prevention of glucose removal from core oligosaccharides retards processing of the insulin receptor and produces a marked decrease in cell-surface receptors. However, proteolytic cleavage of the proreceptor is not blocked, although it takes place a t a slower rate; and further processing of some of the carbohydrate chains is not completely inhibited. Furthermore, the processed receptors are inserted in the plasma membrane, and their insulin binding affinity is normal despite the presence of an undetermined number of glucosylated chains. The occurrence of complex-type oligosaccharide chains suggests that some glucosidase activity is resistant to the inhibitors as previously noted by other authors (14).
The present data agree with those reported for certain membrane and secretory glycoproteins, although other glycoproteins appear to be unaffected by glucosidase inhibitors (26). Secretion of IgD (27), a-antitrypsin, and a-antichymotrypsin (26) as well as the intracellular transport of the glycoprotein E, of mouse hepatitis virus (28) are greatly reduced in the presence of glucosidase inhibitors. Furthermore, the epidermal growth factor receptor of l-deoxynojirimycin-treated A431 cells show delayed acquisition of both ligand binding and endoglycosidase H resistance (29). In addition, cell-surface acetylcholine receptors of BC3H1 cells are reduced in the presence of 1-deoxynojirimycin because of an increase in receptor degradation rate (30).
The mechanism by which prevention of glucose removal retards the processing of the insulin proreceptor is not fully understood. The presence of glucose residues may reduce the affinity of the substrate for the subsequent enzyme(s) involved in the processing pathway. The presence of the glucose residues might impair binding to a putative transport receptor, retarding the movement from the endoplasmic reticulum to the Golgi complex. Alternatively, it could be speculated that the glucosylated chains may interfere with oligomerization or with the tertiary configuration of the polypeptide chain which may be required for its appropriate vesicular transfer (31).
In conclusion, we found that glucose removal from core oligosaccharides represents an important signal in the translocation and rate of processing of the insulin receptor to the plasma membrane.