Effect of Tunicamycin , an Inhibitor of Protein Glycosylation , on the Biological Properties of Acetylcholine Receptor in Cultured Muscle Cells

We have studied the effect of tunicamycin (TM), an antibiotic which inhibits the glycosylation of nascent proteins, on the properties of the acetylcholine receptor (AChR) at the surface of embryonic chick skeletal muscle cells. The use of two separate assays, specific binding of '2SI-~-bungarotoxin and carbamylcholine-activated "Na+ uptake, has allowed us to monitor the effects of impaired glycosylation on the metabolic and functional properties of AChR. A significant decrease in the amounts of surface AChR elaborated in the presence of TM is detected by both measurements. This decrease has been found to reflect an enhanced proteolytic degradation of the underglycosylated AChR. The underglycosylated AChR, expressed on the cell surface in the presence of TM, retains the capability of mediating agonist-activated ionic permeability changes, but displays quantitatively altered interactions with receptor ligands. We conclude that the carbohydrate moiety on AChR may play a role in determining the folding of newly synthesized polypeptides to form a conformation compatible with the metabolic properties and ligand interactions characteristic of glycosylated AChR.

significance of the carbohydrate components of glycoproteins. Studies utilizing TM have indicated a role for carbohydrate moieties in regulating the processing and turnover of specific glycoproteins (10-21).
The carbohydrate moieties of AChR represent approximately 5% of its total weight (2, 22, 23). The impairment of protein glycosylation by TM has recently been shown to diminish the accumulation of AChR in cultured muscle cells In the present study, we have examined the consequences of impaired protein glycosylation on the functional properties of AChR. We provide evidence that the inhibition of protein glycosylation by T M treatment results in the expression of functionally altered AChR on the surface of cultured muscle cells.

Effect of Tunicamycin on Acetylcholine Receptor Properties
Tris base), 5.5 mM glucose, 0.8 mM MgSO,, and 0.1 mM ouabain for 30 min at 37 "C. At the end of this period, medium was removed and cells were rinsed twice within 10 s with 2.5 ml of medium consisting of 5.4 mM KC1, 130 mM choline chloride, 50 mM Hepes (pH 7.4), 5.5 mM glucose, and 0.8 mM MgSO, a t room temperature. Uptake of "Na+ was then assayed a t room temperature in 1 ml of medium containing carbamylcholine chloride at the concentrations specified, 5.4 mM KCl, 125 mM choline chloride, 10 mM NaCI, 50 mM Hepes (pH 7.4), 5.5 mM glucose, 0.8 mM MgS04, 0.1 mM ouabain, and "'NaCI (5 pCi/ml). Under these conditions, "NaC uptake is linear with time for 20 s (see Fig. 4A). Linear "Na' uptake was assayed for 15 s and terminated by washing 5 times within 25 s with 2 ml of ice-cold medium consisting of 163 mM choline chloride, 50 mM Hepes (pH 7.4), 5.5 mM glucose, 0.8 mM MgSO,, 1.8 mM CaCI2, and 1 mM d-tubocurarine chloride. The cells were suspended in 1 N NaOH containing 1% Triton X-100 and radioactivity was determined by y spectroscopy. "2Na+ uptake measured in replicate cultures as described above but, in the absence of carbamylcholine, was subtracted from total uptake. Results are expressed as nmol of n.Na+ taken up/min/culture plate. Concanavalin A-Sepharose Column-Cultures were treated with TM (0.05 pg/ml) or with T M and leupeptin (100 pM) for 24 h and labeled with '"I-a-Bgt as described above. Unbound toxin was removed by 5 washes with DME. After an additional rinse with phosphate-buffered saline, cells were scraped from culture plates with a solution of 50 mM NaCI, 50 mM Tris.HC1, pH 7.4, 1 mM PMSF, and 100 PM leupeptin a t 4 "C and Triton X-100 was added to a final concentration of 1%. After incubation at 4 "C for 1 h, insoluble material was removed by centrifugation for 30 min at 100,000 X g . The supernatant (Triton extract) was adsorbed batchwise for 3 h at 4 "C to 0.2 ml of Con A-Sepharose 4B that had been prewashed twice with 2.5-ml volumes of 1% Triton X-100 in 50 mM NaCI, 50 mM Tris. HC1, pH 7.4. The beads were then washed successively in batch with 8-ml volumes of 50 mM NaC1, 50 mM Tris.HC1, pH 7.4, containing 1 mM PMSF, and 1% Triton X-100, until no radioactivity was detected in the washes. Elution was carried out in batch with 0.4 M a-methyl-D-mannoside in 50 mM NaC1, 50 mM Tris.HC1, pH 7.4, 14 Triton X-100 at 4 "C for 40 h.

RESULTS
We have monitored the expression of AChR on the surface of muscle cells by two distinct measurements i.e. the specific binding of "'I-a-Bgt and the linear uptake of ""af induced by the AChR activator carbamylcholine. Fig. 1 shows the time course of expression of AChR during muscle differentiation as monitored by these two assays. Highly similar kinetics of developmental appearance are observed. These results confirm that the linear uptake of 'lNaf induced by carbamylcholine reflects the amount of cell surface AChR.
With the aim of achieving maximal expression of underglycosylated AChR, cultures were treated with T M (0.05 pg/ml) at 48 h after plating for a 24-h interval. As can be seen (Fig. I), during this period a major increase in AChR elaboration occurs. Under these experimental conditions, TM treatment inhibits incorporation of ["Hlmannose into trichloroacetic acid precipitable material by over 80%, while causing less than 20% inhibition of [''C]leucine incorporation into protein (data not shown). To investigate the consequences of impaired protein glycosylation, the effects of TM and protease inhibitors on carbamylcholine-activated 92Na+ uptake have been studied. Table I shows that the linear uptake of "Na+ mediated by AChR elaborated in the presence of T M is inhibited by approximately 80% as compared to untreated cultures. This inhibitory effect of TM is attenuated in the presence of the protease inhibitors leupeptin and chloroquine. In addition, Table I shows that the effects of TM and protease inhibitors on ""a' uptake and '"I-a-Bgt binding, measured in replicate cultures, are quantitatively similar. These results suggest that the TM-induced decrease in linear uptake of 'lNa+ primarily reflects the reduced amounts of surface AChR. Furthermore, the partial reversal of the effect of T M by protease inhibitors is consistent with the possibility that the TM-induced loss of AChR activity is associated with enhanced proteolytic degradation. and by the linear uptake of "Na+ activated by carbamylcholine (A).
For binding measurements, cultures were labeled with '"I-a-Bgt ( M) for 60 min at 37 "C. Nonspecific labeling, determined in the presence of the competitive ligand decamethonium (10 p~) , was subtracted from total labeling and did not account for more than 10% of total labeling. z2Na+ linear uptake was measured in replicate cultures by incubation for 15 s at room temperature in the presence of carbamylcholine (10 mM). Uptake of "Na+, measured under the same conditions but in the absence of carbamylcholine, was subtracted. Points represent the averages of 2 experiments each consisting of 3 determinations.

TABLE I
Effect of tunicamycin on the expression of AChR measured by "'2a-Bgt specific binding a n d carbamylcholine-activated 22Na+ linear uptake AChR was monitored by the specific binding of '"I-a-Bgt and by the linear uptake of T'Na+ activated by carbamylcholine as described in Fig. 1 and under "Experimental Procedures." Where specified, cultures were pretreated with T M (0.05 pg/ml) and the protease inhibitors leupeptin (100 p~) and chloroquine (10 PM) for a 24-h period. Values shown are averages of three determination carried out on replicate culture plates. Values for per cent effect obtained in repeat experiments did not vary by more than 3% from the average shown.
Additions to control ""I-a-Bgt hound To compare the depletion rates of AChR-mediated "Na+ uptake in TM-treated and untreated cultures, the linear uptake of 22Na+ was measured at various intervals after the addition of the protein synthesis inhibitor, cycloheximide (10 pg/ml). Since the net accumulation of AChR represents the balance between receptor synthesis and degradation, the observed loss of surface AChR activity, which follows first order kinetics (Fig. 2, inset), can provide an estimate of AChR degradation rate. In cultures not treated with TM, after an initial 4-h period of AChR accumulation which presumably reflects the continuing appearance of surface AChR synthesized prior to the interruption of protein synthesis ( 3 , 3 2 ) , the AChR-mediated uptake of "Na' decreases ( Fig. 2 ) with a half-time of approximately 14 h (Fig. 2,  significantly enhanced, with a half-time of approximately 4 h ( Fig. 2 and inset). These values are in close agreement with the degradation rates of '"I-a-Bgt-AChR complexes in TMtreated and untreated cultures reported previously (24). ConA, a lectin with high affinity for glucose and mannose residues on glycoproteins, has been shown to interact with surface AChR in muscle cells (33). T o monitor the effect of TM on the glycosylation of surface AChR, intact TM-treated and untreated cells were labeled with I2'I-a-Bgt. The 1251-a-Bgt-AChR complexes were then detergent-solubilized and analyzed by ConA-Sepharose affinity chromatography. A comparison of the capacity for specific adsorption to ConA-Sepharose of "'I-a-Bgt-AChR complexes from TM-treated and untreated cultures is shown in Table 11. As can be seen, the fraction of surface AChR from TM-treated cells that is adsorbed to ConA-Sepharose is significantly lower than that from untreated cells. In addition, the fraction of the adsorbed '""I-a-Bgt-AChR complexes that is specifically eluted by -amethyl-D-mannoside is considerably smaller in extracts from TM-treated cells. Similar results are obtained with extracts from cultures exposed to TM in the presence of the protease inhibitor leupeptin (Table 11), a treatment that results in a partial reversal of the TM-induced depletion of surface AChR (Table I) glycosylated. Since TM blocks the synthesis and transfer of core oligosaccharides to asparaginyl residues (7-9), this underglycosylated state of AChR apparently corresponds to the presence of fewer asparagine-linked core oligosaccharides.
To further investigate if the carbamylcholine-induced Na' uptake in TM-treated cells is mediated by AChR deficient in glycosylation, we have compared the effects of ConA on AChR activity in TM-treated and untreated cultures. As shown in Fig. 3 (inset), ConA inhibits carbamylcholine-activated z'Na+ uptake in a dose-dependent manner, to a maximum of approximately 70% a t 0.5 ,UM ConA. This inhibitory effect of ConA is almost abolished in cultures previously exposed to T M (Fig. 3 ) .
To examine if the exposure of muscle cells to TM produces changes in the functional properties of AChR, we have measured '%a+ linear uptake at increasing concentrations of carbamylcholine in TM-treated and -untreated cultures. As indicated by the dose-response relationships shown in Fig. 4B, the apparent V,,, is significantly reduced in TM-treated cells, reflecting the reduced number of functional receptors. Moreover, the apparent affinity of AChR for carbamylcholine is reduced in TM-treated cultures, as shown by the 3-4-fold increase in the apparent K,, (Fig. 4B, inset). To rule out the possibility that the observed difference in apparent K,,, reflects the large difference in AChR levels in TM-treated and untreated cultures, these measurements were repeated under   Effects of TM on the functional properties of AChR. A, relationship between Na' uptake and the duration of exposure to carbamylcholine in TM-treated and untreated muscle cells?' Na+ uptake was measured, as described under "Experimental Procedures," in the presence of carbamylcholine (IO mM) for the specified intervals. B, effect of increasing concentrations of carbamylcholine on linear uptake of Z2Na+ in TM-treated and untreated muscle cells. Linear uptake of "Na+ was measured, as described under "Experimental Procedures," in the presence of the specified carbamylcholine concentrations. Inset, double reciprocal plot of the data. C, effect of increasing concentrations of d-tubocurarine on carbamylcholine-activated 'lNa+ uptake in TM-treated and untreated muscle cells. Cultures were preincubated in the presence of d-tubocurarine at the concentrations specified for 30 min. Linear "Na+ uptake activated by carbamylcholine (10 mM) was then measured, as described under "Experimental Procedures," in the presence of &tubocurarine at the concentrations indicated. In all experiments (A-Q, values shown were obtained by subtracting "Na+ uptake in the absence of carbamylcholine from the total uptake. Measurements in TM-treated (0) and untreated (0) muscle cells were carried out in cultures from the same plating. Points represent averages from 2 experiments, each consisting of 3 determinations. to accumulate new AChR for a further 4-h period. After this interval, the amounts of new surface AChR, as measured by both '"I-a-Bgt binding and "Na+ uptake, were equivalent to AChR levels in cultures exposed to TM for 24 h. However, the apparent K,,, under these conditions (1.6 mM, results not shown) is significantly lower than the K , in TM-treated cells (4.5 111~, Fig. 4B), and resembles that of untreated cultures (1.2 mM, Fig. 4B).
The time-dependence of carbamylcholine-activated "Na+ uptake is shown in Fig. 4A. The uptake of "Na+ is linear during the initial 20 s of exposure to carbamylcholine (10 mM), and subsequently attenuates sharply. This decrease in AChRmediated Na+ uptake in the presence of activator has been attributed to the development of receptor inactivation (5), a process characteristic of AChR (34, 35). To examine the significance of protein glycosylation for the development of AChR inactivation, we compared the time dependence of carbamylcholine-induced "Na+ uptake in TM-treated and untreated muscle cells. Fig. 4A shows that, although the maximal uptake is reduced, the duration of the initial period of linear "Naf uptake and its subsequent attentuation are not significantly altered by T M treatment.
The effect of various concentrations of d-tubocurarine, a competitive inhibitor of AChR, on carbamylcholine-induced 22Na+ uptake in TM-treated and untreated muscle cells is shown in Fig. 4C. As can be seen, d-tubocurarine retains its inhibitory effect on carbamylcholine-induced AChR activation under conditions of impaired protein glycosylation. The altered dose-response relationship observed in cultures exposed to TM can reflect the change in apparent affinity towards carbamylcholine, and possibly an additional contribution of a modified interaction of underglycosylated AChR with d-tubocurarine.

DISCUSSION
In the present study, the consequences of impaired protein glycosylation on the functional and metabolic properties of AChR in cultured muscle cells have been investigated. We present evidence that AChR expressed on the surface of intact muscle cells treated with TM, an inhibitor of protein glycosylation, displays altered properties.
Under the culture conditions used, myogenesis is marked by a rapid burst of AChR appearance on the surface of newly fused muscle cells. The amounts of AChR on the cell surface were measured by the specific binding of '"SI-a-Bgt and AChR activity was monitored by receptor-mediated Na+ uptake. The time-course of AChR accumulation is essentially identical when measured by either '"I-a-Bgt binding or the initial rate of "Na+ uptake triggered by the AChR activator carbamylcholine (Fig. 1). This close similarity verifies that the initial rate of carbamylcholine-activated Na' influx reflects the amount of AChR. Furthermore, the Naf flux assay provides a quantitative estimate of functional AChR capable of transducing agonist binding into activation of ionic channels. The simultaneous appearance of toxin-binding and Na' flux activities under conditions of synchronous differentiation (Fig. 1) indicates, in addition, that AChR reaches the cell surface in a functional form.
The AChR elaborated during the 24-h exposure to TM is expressed on the cell surface in reduced amounts but can be readily detected by both '""I-a-Bgt specific binding and carbamylcholine-activated Na' uptake. In order to restrict our measurements to the AChR synthesized in the presence of TM, the contribution of pre-existing AChR was abolished by exposure of muscle cells to unlabeled a-Bgt during the initial 3 h of T M treatment.
In the context of the present study, it is important to assess by guest on October 16, 2017 http://www.jbc.org/

Effect of Tunicamycin on Acetylcholine Receptor Properties 1779
if the AChR detected on the surface of TM-treated cells is underglycosylated. The impaired glycosylation of surface AChR is suggested by our finding that exposure of muscle cells to TM results in decreased capacity of "'I-a-Bgt-AChR complexes to adsorb specifically to ConA-Sepharose (Table  11). Furthermore, we observe that ConA significantly inhibits Na+ uptake in control cultures (Fig. 3, and inset) while, in marked contrast, ConA treatment fails to inhibit receptormediated Na+ uptake in TM-treated cells (Fig. 3) even a t a higher ConA concentration (1 ,UM, results not shown). These findings indicate that AChR appearing on the surface of TMtreated muscle cells is underglycosylated. The phenomenon of inhibition of carbamylcholine-activated Na' uptake by ConA is itself of interest, indicating the proximity of receptorlinked oligosaccharides to either agonist binding sites or channel components.
The most pronounced effect of TM treatment is the extensive reduction in the amounts of AChR present on the cell surface. The net accumulation of AChR represents the balance between receptor synthesis and degradation. An enhanced degradation rate has been demonstrated for several glycoproteins synthesized in the presence of T M (10, 12, 15). The degradation rate of AChR, measured by the release of radioactivity from '""I-a-Bgt specifically bound to AChR, was recently shown to increase markedly upon treatment of muscle cells with T M (24). To estimate the extent to which the TM-induced reduction of Na+ flux (Table I) could be accounted for by enhanced degradation, carbamylcholine-activated Na+ uptake was monitored a t various intervals after the addition of cycloheximide, an inhibitor of protein synthesis. Under these conditions, a significantly enhanced degradation rate of the surface underglycosylated receptor is observed (Fig. 2). The notion that the reduced Nat uptake induced by TM treatment reflects depletion of surface AChR due to enhanced receptor degradation is further supported by the partial reversal of the TM effect by the protease inhibitors leupeptin and chloroquine (Table I). These findings are consistent with evidence that the absence of carbohydrate moieties on glycoproteins renders these proteins more susceptible to proteolytic degradation both in vitro and in intact cells (10, 15, 36).
It is of interest to compare our present findings with a recent study by Merlie et al. (26). These authors utilized pulse-chase and immunoprecipitation procedures in detergent extracts to study the effects of T M on the expression of AChR subunits in a mouse muscle cell line. Their observations that T M treatment results in the elaboration of underglycosylated AChR that displays accelerated degradation kinetics are qualitatively consistent with previous observations (24) and our current findings. However, there are apparent quantitative discrepancies between our findings and the conclusions drawn by Merlie et al. As proposed by these authors, TM treatment results in a significant inhibition of AChR subunit assembly, and the small amounts of the nonglycosylated subunits that undergo assembly are degraded too rapidly to allow their accumulation on the cell surface. In marked contrast, our present findings, based on two independent assays for surface AChR, clearly indicate that underglycosylated receptors reach the surface of cultured muscle cells exposed to TM. How can our current findings be reconciled with the observations of Merlie and co-workers? Firstly, the different types of measurements utilized in each of the studies make meaningful quantitative comparisons difficult. In particular, the values of AChR catabolic half-life obtained by Merlie et al. (26) may have been influenced by several factors including the brief pulse intervals used, the short chase period, and the possible reduced detergent-solubility of the nonglycosylated subunits (18). In addition, the increased susceptibility of nonglycosylated proteins to proteolytic degradation (10, 12, 15, 36) may be amplified by enhanced protease activity in the detergent extract, leading to an overestimate of the degradation rate of underglycosylated AChR as compared to measurements performed in intact cells. Secondly, the AChR elaborated in the different cell type used by Merlie et al. displays different metabolic properties (3), and possibly different kinetics of synthesis, from those of AChR expressed on the surface of cultured embryonic chick skeletal muscle cells. Though it is conceivable that TM-induced inhibition of assembly of AChR subunits, as suggested by Merlie et al. (26), contributes to the reduced accumulation of AChR observed in the present study (Table I), treatment of muscle cells with TM during the period of sharply increased AChR elaboration (2 to 3 days postplating) allows the accumulation of significant amounts of underglycosylated AChR on the cell surface.
The oligomeric glycoprotein AChR transduces the binding of activators into changes in transmembrane ionic permeability (for review see Refs. 1 and 2). The measurement of surface AChR by carbamylcholine-activated Na' uptake provides a useful approach for investigating the functional significance of oligosaccharide chains on these receptors.
In the present study, we find that AChR activation, desensitization, and susceptibility to inhibition by d-tubocurarine are retained under conditions of impaired glycosylation (Fig.  4). However, the apparent affinity of surface AChR for carbamylcholine is reduced as a consequence of T M treatment (Fig. 43), independently of the TM-induced reduction in AChR levels (see "Results"). This finding is of particular interest when contrasted with the absence of detectable differences in the functional properties of several nonglycosylated glycoproteins synthesized in the presence of TM. Examples include the observation that nonglycosylated fibronectin is as effective as the glycosylated protein in promoting fibroblast attachment and erythrocyte agglutination (36), as well as the finding that nonglycosylated G protein is functional in vesicular stomatitis virus attachment and penetration (18,37).
Although retaining functional properties, the nonglycosylated G protein synthesized in the presence of T M displays modified physical properties consistent with an altered conformation (18, 38). Evidence for the significance of carbohydrate in maintaining glycoprotein tertiary structure has been advanced by studies utilizing T M (39) as well as enzymatic removal of oligosaccharide chains (40, 41). The notion that conformation is an important determinant of the degradative rates of proteins is well documented (for review see Ref. 42).
The findings of increased susceptibility of several nonglycosylated glycoproteins to degradation by intracellular proteases constitute indirect evidence for the influence of carbohydrate moieties on tertiary structure. The association between the conformational features and the function of AChR has been established in studies using covalent modification (43). By analogy, the change in apparent affinity of underglycosylated AChR toward carbamylcholine observed in the present study may reflect the significance of carbohydrate for the conformational features of functional AChR.
In conclusion, we suggest that impairment of protein glycosylation by T M results in the epxression of AChR altered in conformation and, as a consequence, changed in its metabolic and functional properties.