Acetylcholine Receptor Synthesis from Membrane Polysomes*

We have established conditions for the fractionation of cytoplasmic and membrane-bound polyribosomes from the clonal mouse cell line BCSH-1. Polyribosome fractions are obtained in goad yield and purity. They are active in protein synthesis when incubated with nuclease-treated rabbit reticulocyte lysates, and we have demonstrated that the cytoplasmic and mem-brane-bound fractions direct the synthesis of distinctly different sets of proteins. Using immunoprecipitation and sodium dodecyl sulfate gel analysis, we have shown that the membrane-bound but not the cytoplasmic polyribosomes direct the synthesis of two protein species (Mr = 39,000 and 42,000) which are homologous to the native a subunit of acetylcholine receptor. Peptide maps suggest that the two species synthesized in vitro may correspond to the nonglycosylated and glycosyl- ated forms, respectively, of the a subunit. The importance of the acetylcholine to synapse at the regulated ACh’

We have established conditions for the fractionation of cytoplasmic and membrane-bound polyribosomes from the clonal mouse cell line BCSH-1. Polyribosome fractions are obtained in goad yield and purity. They are active in protein synthesis when incubated with nuclease-treated rabbit reticulocyte lysates, and we have demonstrated that the cytoplasmic and membrane-bound fractions direct the synthesis of distinctly different sets of proteins. Using immunoprecipitation and sodium dodecyl sulfate gel analysis, we have shown that the membrane-bound but not the cytoplasmic polyribosomes direct the synthesis of two protein species (Mr = 39,000 and 42,000) which are homologous to the native a subunit of acetylcholine receptor. Peptide maps suggest that the two species synthesized in vitro may correspond to the nonglycosylated and glycosylated forms, respectively, of the a subunit.
The importance of the acetylcholine receptor to synapse formation at the vertebrate neuromuscular junction gives special significance to the understanding of mechanisms by which the synthesis of this membrane glycoprotein is regulated (1,2). Several lines of evidence have led to the view that synthesis of ACh' receptor is regulated throughout the life of the muscle fiber, and that regulation is coupled in some way directly to the electrical or mechanical activity of the fiber (3). In most studies effects at the level of synthesis have been inferred from measurements of ACh receptor content (3,4) and the rate constant for receptor degradation (5). Synthesis in vivo has been measured directly in a few cases (6, 7), confirming the earlier studies. Information concerning the mechanisms of synthesis and its regulation has proven difficult to obtain because the most. interesting regulat,ory phenomena occur in intact tissue in living animals or organ culture and because the ACh receptor represents a very small proportion (less than 0.01%) of the total synthetic activity of such tissue.
Embryonic muscle ceUs in tissue culture provide some unique advantages for studies of ACh receptor synthesis. Even primary cultures have been shown to be as active as intact tissue for ACh receptor synthesis (8,9). In addition, at least *This work was supported by a Biomedical Research Support Grant to the University of Pittsburgh from the National Institutes of Health and by grants from the Muscular Dystrophy Associations of America. Inc., and the National Institutes of Health to J. P. M. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 so1eLy to indicate this fact. one clonal mouse cell line has been established which is more active than any known mammalian tissue for ACh receptor synthesis (10). Using such cell culture systems, Devreotes et al. (11) have shown that newly synthesized ACh receptor possessing a-bungarotoxin binding activity first appears in an intracellular precursor pool, and is transferred to the plasma membrane only after 2 h. We have shown recently (12) that newly synthesized ACh receptor subunits require approximately 15-30 min before they acquire a-bungarotoxin binding activity, indicating the existence of a pool of inactive precursors at a stage which precedes the pool defined by the experiments of Devreotes et al. (21).
We hope to be able to define the mechanisms by which ACh receptor synthesis is regulated. As a fist step toward this goal, we have established a cell-free system for the synthesis of ACh receptor polypeptides. Using nuclease-treated rabbit reticulocyte lysates and polyribosome fractions prepared from the clonal mouse cell line BC3H-1 (IO), we have shown that ACh receptor a subunit is synthesized on membrane-bound polyribosomes. mM MgC12,l mM phenylmethylsulfonyl fluoride, and 10 mM vanadylribonucleoside complex as an RNase inhibitor (14). The first homogenate was made by using 10 strokes of a tight-fitting Dounce homogenizer. It was centrifuged for 15 min at 131,000 X g, and the pellet was rehomogenized and centrifuged again. The second supernatant was added to the first and this fraction is referred to as the cytoplasmic extract (Table I). The pellet of the second centrifugation was extracted with homogenization buffer, 4% (w/w) Triton X-100, and nuclei were removed by centrifugation at 12,000 X g for 5 min. The resulting supernatant is referred to as the membrane extract. The cytoplasmic and membrane extracts were layered over discontinuous sucrose gradients consisting of 1.5 ml of 2.0 M and 1.5 mi of 1.3 M sucrose in homogenization buffer without RNase inhibitor or cycloheximide. The gradients were centrifuged for 20 h at 178,000 X g . The upper sucrose layers were removed, t,he tubes were rinsed with buffer containing 10 mM Tris HC1, pH 7.4, 10 mM KC1, 1.5 mM MgCL, and the pellets were resuspended in the same buffer.
Polyribosomes were analyzed by velocity sedimentation in 10-409; (w/v) sucrose gradients prepared in homogenization buffer. Gradients were centrifuged for 110 min at 28,000 rpm in an SW-28 rotor.
Cell-free Protein Synthesis-Protein synthesis was performed with ribonuclease-treated rabbit reticulocyte lysates prepared as described by Pelham and Jackson (15). Resuspended polysomes were incubated with lysate and a reaction mixture containing 150 mM potassium acetate, 2 mM magnesium acetate, 25 mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid buffer, pH 7.4, and an amino acid mixture 6995 Cell-free Synthesis of ACh Receptor modified to that of the a subunit of ACh receptor of Torpedo marmarota (16). 0.64 ~-[~'S]methionine (35 Ci/mmol, New England Nuclear) was included for labeling. The Kt, Mg", and methionine concentrations were determined to be optima for incorporation stimulated by polyribosomes.

~mmunoprec~ztation-[35S]methionine-labeled
ACh receptor synthesized in vivo was immunoprecipitated by toxin-antitoxin and formalin-fixed, heat-treated Staphylococcus aureus as previously described (12). This method employs an anti-a-bungarotoxin antiserum and precipitates only native ACh receptor molecules to which abungarotoxin is bound. For immunoprecipitation of in vitro products an antiserum prepared against sodium dodecyl sulfate-denatured purified fetal bovine muscle ACh receptor (anti-SDS-ACh receptor) was used (12).
Treutment with Endo-P-N-acetylglucosum~idase H-Immunoprecipitates of "S-labeled ACh receptor labeled in vivo were treated with 0.2% sodium dodecyl sulfate in distilled Ha0 at 100 "C to elute labeled proteins. The S. aureus ceUs were removed by centrifugation and the supernatant was treated with or without 1 milliunit of endo- Poiyucrylamide Gel Electrophoresis and Peptide Mapping-Onedimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed as described previously (18, 12). High resolution two-dimensional gel electrophoresis was performed as described by Garrels (19). Partial proteolytic digests were performed using S. uureus V8 protease (Worthington) (20, 12).

RESULTS AND DISCUSSION
Our fwst objective was to develop a fractionation procedure for cytoplasmic and membrane-bound polyribosomes from the clonal mouse muscle cell line, BC3H-1. We have employed a modification of the method developed by Ramsey and Steele (13) for rat liver polysomes. Since our intention was to use the purified fractions for the cell-free synthesis of ACh receptor, and thereby to determine whether ACh receptor synthesis occurred on cytoplasmic or membrane-bound polysomes, it was essential to obtain the highest quality polyribosomes with respect to integrity, relative and absolute yield, and purity.
An indication of the integrity of the fractionated polyribosomes can be gotten from the distribution of absorbance at 260 nm and 3H-labeled RNA in analytical sucrose gradients shown in Fig. 1. This experiment was done using cells which had been labeled with [3H]uridine so that we could determine quantitatively the distribution of RNA in the different fractions. Table I represents a summary of another such experiment. Table I shows that: 1) approximately 50% of the polysomes present in the initial extracts are recovered in the purified fractions for both cytoplasmic and membrane fractions.
2) The distribution of polysomes between the cytoplasmic and membrane fractions is 9:L 3) Approximately 85% of the ribosomes in the purified polysome fractions were distributed equally between small and large polysomes. These data indicate that our fractionation procedure does not lead to any selective loses from either cytoplasmic or membrane polyribosome fractions, Purified polysome fractions contain less monosomes than the initial extracts because monosomes are lost during the discontinuous sucrose gradient step. The relatjvely high value for monosomes in the cytoplasmic extract is not unusual for cells in stationary phase (21), it is approximately twice that reported for primary cultures of quail myotubes (22). The occurrence of monosomes in the purified membrane polysome fraction was variable (compare data from separate experiments in Fig. 1 and Table I) and probably due to polysome runoff or breakage during fractionation.
A further indication of the integrity of the isolated polyribosome fractions was obtained from experiments which assayed their capacity to stimulate incorporation of ~-["~S]methionine by nuclease-treated rabbit reticulocyte lysates (15). Fig. 2 shows that incorporation of [35S]methionine by nuclease-treated lysates was absolutely dependent upon addition of polyribosomes. Incorporation stimulated by polyribosome: continued for longer times at greater rates than the incorpo ration due to endogenous mRNA in non-nuclease-treatec lysates. Cytoplasmic and membrane-bound polyribosome: were equally active when activity was expressed in terms o radioactivity incorporated per unit of absorbance at 260 nn in a 90-min incubation. We have determined that the nuclease treated rabbit reticulocyte lysates used for these experiment, were completely inactive when purified functional BC3Hpoly(A) + RNA or pure rabbit globin mRNA was added

Cell-free
Therefore, the incorporation of ['"S]methionine stimulated by BC3H-1 polyribosomes shown in Fig. 2, as well as in those figures to follow, was most likely due to "runoff' or elongation, and not to re-initiation in vitro.
The purity or degree of fractionation of the two polyribosome classes was estimated by comparing the protein products synthesized in reticulocyte lysates. The products were analyzed by high resolution two-dimensional polyacrylamide gel electrophoresis (19). Fig. 3 shows portions of two such gels, with A and B containing cytoplasmic and membrane polyribosome products, respectively. The actual identities of the  subjected to two-dimensional electrophoresis as described by Garrels (19). The first dimension gel was prepared with pH 5-7 ampholines (basic ends to the right), the second dimension was a sodium dodecyl sulfate-10% polyacrylamide gel. The gels were processed for fluorography (23) and exposed to x-ray film for 3 weeks. Only a portion of each gel is presented. Open and filled arrowheads indicate spots which are enriched in cytoplasmic and membrane-bound fractions, respectively. As a reference, the 3 intense spots in the middle of A are actins (19). proteins are not important, although it is possible to identify tentatively many of the abundant structural proteins. The important aspect of the comparison of the two gels is that there is a different set of proteins enriched in the two fractions.
In the case of the cytoplasmic polyribosome fraction, we have indicated 4 (of many) abundant species which are synthesized in greatly reduced quantities by the membrane-bound fraction; we estimate an enrichment of about 5-fold. There are fewer abundant proteins which appear to be unique to the membrane-bound fraction: we have indicated 3. On the basis of these 7 proteins we conclude that our fractionation procedure does yield polyribosome fractions which are enriched for different mRNAs. Finally, the fact that both fractions produce discrete products is a good indication of polyribosome integrity.

Cell-free Synthesis of ACh Receptor
We have demonstrated that the membrane-bound polyribosome fraction directs the synthesis in vitro of ACh receptor a subunit. This has been done by immunoprecipitation of the in vitro products with rabbit anti-SDS-ACh receptor and sodium dodecyl sulfate-polyacrylamide gel analysis.  Fig. 4, lane A. The most abundant single band immunoprecipitated from the in vitro products synthesized by membrane-bound polyribosomes (Fig. 4, lane D ) co-migrates exactly with the a subunit.
This band is completely absent from the immunoprecipitates of the in vitro products of cytoplasmic polyribosomes (Fig. 4, lane 0. One-dimensional peptide maps of partial proteolytic digests of a subunit "S-labeled in vivo and the M , = 42,000 band from the in vitro reaction are very similar (Fig. 5, lanes A , B, and  E ) , confirming the identity of the a subunit synthesized in vitro. However, we expected that the in vitro product might differ from native a in its migration on gels and perhaps its peptide map due to incomplete processing of prepeptides and/ or oligosaccharide side chains (24). In particular, we have evidence that the a subunit synthesized in vivo is glycosylated with at least one "high mannose" or "simple" N-linked oligosaccharide side chain.' One characteristic of this type of '' J. P. Merlie, unpublished results.
glycoprotein is that the oligosaccharide chain is sensitive to endoglycosidase H (25). Treatment of the native a subunit with endoglycosidase H results in a reduction in the apparent molecular weight ( M r = 42,000 to M , = 39,000) (Fig. 4, lane  B ) . A polypeptide with very similar apparent molecular weight is immunoprecipitated from the in vitro products of the membrane-bound polyribosomes. The peptide map of the in vitro M , = 39,000 species identifies it as an a-related polypeptide. Furthermore, the peptide maps suggest that the M, = 39,000 in vitro product is the nonglycosylated form of the a subunit and the M, = 42,000 in vitro product is a glycosylated form.
There is a clear difference in a single peptide between the in vivo native a and the in vivo native a treated with endoglycosidase H (Fig. 5, lanes A and 0. We can detect this same difference in the M , = 39,000 and 42,000 in vitro products. These peptides are clearly visible on the x-ray films, but are poorly reproduced in prints. We have placed asterisks to indicate their position. More direct chemical analyses of the peptides are necessary to confirm this suggestion. The other bands specific to the membrane-bound fraction (Fig. 4, lane  D ) have not yet been analyzed.
If the major a subunit-like product of our in vitro reactions is glycosylated, it could be so only because the nascent chains of the a subunit polyribosomes are glycosylated in vivo before isolation. This is completely consistent with the accepted mechanism of N-linked glycosylation of glycoproteins (24). Furthermore, if the M , = 39,000 polypeptide represents the nonglycosylated form, the relative amounts of the M , = 42,000 and 39,000 polypeptides indicate that the site of glycosylation must be very close to the NH' terminus of the protein.
Finally, both the M, = 39,000 and 42,000 in vitro products appear to be contaminated with actin, since their peptide maps show a faint band corresponding to the major actin peptide (Fig. 5, lane F ) . This would seem to be an indication of: 1) the relative degree of contamination of membrane polyribosomes with cytoplasmic polyribosomes, and 2) of the degree of premature chain termination in the cell-free protein synthesis reaction. These experiments demonstrate that the ACh receptor a subunit is synthesized in vivo on membrane-bound polyribosomes, as was predicted by the current version of the signal hypothesis (26). That integral plasma membrane proteins of eukaryotes are synthesized on membrane-bound polyribosomes has been demonstrated for only a few other proteins (23, 27, 28). In addition to establishing this important aspect of ACh receptor synthesis, our results provide us with an effective step in the purification of BC3H-1 mRNA specific for ACh receptor polypeptides. The 5-fold enrichment of the membrane-bound polyribosomes obtained by fractionation should increase the abundance of ACh receptor mRNA to approximately 0.5% of the total membrane-associated mRNA. We think these experiments represent an important fmt step in our attempts to study mechanisms involved in the synthesis of ACh receptor.