Isolation and characterization of nicotinic acetylcholine receptor-like protein from fetal calf thymus

Journal of the Neurological Sciences, 1988, 87:195-209

195

Elsevier JNS 03057

Isolation and characterization of nicotinic acetylcholine receptor-like protein from fetal calf thymus Sachiko Kawanami ~'*, Bianca Conti-Tronconi 1,**, John Racs 1 and Michael A. Raftery 1"2 lDivision of Biology and Chemical Engineering, California Institute of Technology, Pasadena, CA 91125 (U.S.A.), and 2Department of Chemistry, University of Minnesota, Minneapolis, MN 55455 (U.S.A.)

(Received22 March, 1988) (Revised, received 8 June, 1988) (Accepted 9 June, 1988)

SUMMARY A nicotinic acetylcholine receptor-like protein (AChR-LP) was isolated from fetal calf thymus by affinity chromatography using cobrotoxin-Sepharose after alkaline extraction and solubilization with Triton X-100. The AChR-LP had a specificity of 1.61 + 1.12 nmol of ~-bungarotoxin binding sites per mg of protein. The isoelectric point, sedimentation coefficient and amino acid composition of the purified AChR-LP were very similar to those of muscle and electric organ AChRs. Upon SDS-polyacrylamide gel electrophoresis purified thymus AChR-LP preparations contained up to 6 polypeptide bands of molecular weights of 40000, 43 000, 51000, 56000, 58 000, and 66000, respectively. The peptides of 40000, 51000, 56000, and 66000 dalton crossreacted with the four subunits of Torpedo californica and fetal calf muscle AChR.

Key words: Acetylcholine receptor; Thymus; Myasthenia gravis

* Correspondence andpresent address: First Department of Internal Medicine,Fukuoka University Hospital, Jonan-ku, Fukuoka 814-01, Japan. ** Present address: Department of Biochemistry,CBS,Universityof Minnesota, St. Paul, MN 55108, U.S.A.

0022-510X/88/$03.50 © 1988Elsevier SciencePublishers B.V.(BiomedicalDivision)

196 INTRODUCTION

Myasthenia gravis (MG) is a human paralysis due to an autoimmune response against the nicotinic cholinergic receptor (AChR) at the neuromuscular junction (Engel 1984; Lindstrom 1985). Anti-AChR antibodies cause accelerated destruction and functional impairment of the AChR, and failure of neuromuscular transmission (Patrick and Lindstrom 1973; Richman et al. 1980; Engel 1984; Lindstrom 1985). Myasthenic symptoms can be induced in a variety of animals by immunization against purified AChR (Patrick and Lindstrom 1973), or injection of anti-AChR antibodies (Toyka et al. 1975; Richman et al. 1980). The AChR from peripheral tissues like fish electric organ and mammalian muscle is a complex transmembrane glycoprotein formed by four homologous subunits, in a stoichiometry 0t2fl),b(Raftery et al. 1980; Conti-Tronconi et al. 1982; Numa et al. 1983; McCarthy et al. 1986). Corresponding subunits from different species are highly homologous proteins (Raftery et al. 1980; Conti-Tronconi et al. 1982; Numa et al. 1983; McCarthy et al. 1986). The source of antigenic stimulation which initiates autoantibody production in MG is unknown. A thymus dysfunction could play a role in MG pathogenesis, as suggested by the frequent thymic abnormalities found in myasthenic patients: hyperplasia or thymoma (Castleman 1966). Thymus lymphocytes from myasthenic patients can synthesize anti-AChR antibodies (Vincent et al. 1978; Fujii et al. 1986). The thymus could therefore be the site where the primary sensitization against AChR takes place. In support of this possibility it has been shown that thymus contains ~-bungarotoxin (BuTx) binding sites (Kao and Drachman 1977; Engel et al. 1977) and cultured cells from adult human thymus can differentiate into striated muscle fibers (Wekerle et al. 1975). Immunological crossreactivity between a thymus component and muscle AChR has been suggested (Van der Geld and Oosterhuis 1963; Ahronov et al. 1975; Ueno et al. 1980). It is therefore important to determine if the thymus does indeed contain an AChR-like protein, able to bind ~-BuTx, and how such a component is related to skeletal AChR. Purification of ~-BuTx-binding protein(s) from the thymus has been difficult because of the agonizingly low amounts of these components in the thymus (Lindstrom et al. 1976; Raimond et al. 1984). We report here the isolation and partial characterization of an AChR-like protein (AChR-LP) from fetal calf thymus, by the use of a combined method of alkaline extraction of peripheral membrane proteins and affinity chromatography with cobrotoxin-Sepharose.

MATERIALS AND METHODS

Purification of cobrotoxin and preparation of affinity resin The long ~-neurotoxin (cobrotoxin) from Naja naja siamensis venom (Sigma) was purified as described by Ong and Brady (1974). The purity of the cobrotoxin preparations was assessed by SDS-polyacrylamide gel electrophoresis (Laemmli 1970) using

197 linear or exponential polyacrylamide gradients from 8 to 15% and by aminoterminal gas-phase sequencing, using an Applied Biosystems sequencer. Only one protein band of the expected molecular weight (8 000) was present upon gel electrophoresis, and only the expected sequence was found, with 3 % contaminating sequences. The cobrotoxin was covalently attached to Sepharose 2B (Pharmacia) following CNBr activation of the agarose matrix as described by Porath et al. (1973). The protocol for weak resin activation was followed in order to achieve a low degree of toxin substitution, which has been demonstrated to be most effective for efficient adsorption and desorption of AChR (Ong and Brady 1974; Gotti et al. 1982). The ability of the cobrotoxin-Sepharose resin to specifically bind AChR, and to specifically release it after incubation with 1 M-carbamylcholine was measured using Triton X-100 extract of Torpedo californica electric organ, as described by Gotti et al. (1982). Typical preparations of cobrotoxinSepharose resin yielded a maximum of 4 nmol of AChR (measured as g-BuTx binding sites according to Schmidt and Raftery (1973)) per g of wet resin.

Isolation of A ChR-like protein(s)from bovine thymus Freshly dissected fetal calf thymus was homogenized in 4 volumes of 10 mM sodium phosphate buffer, pH 7.4, containing 50 mM sodium chloride, 0.02% sodium azide and the following protease inhibitors: 10mM phenylmethylsulfonyl fluoride (PMSF), 3 mM EDTA, 1 mM EGTA, 5 mM iodoacetamide (buffer A) for two 30-sec cycles at high speed in a Waring blender and three 30-sec cycles with a Polytron homogenizer at maximum speed. Tissue debris were pelleted by centrifugation at 5000 rpm for 10 min in a Sorvall GSA rotor (Elliot et al. 1980). The supernatant was filtered through two layers of cheese cloth and centrifuged at 100000 x g for 1 h at 4 °C. The pellet was resuspended using a Polytron homogenizer with an equal volume of buffer B (buffer A with 2 mM PMSF and without iodoacetamide). The resuspended pellet was diluted with 20 volumes of distilled water and the pH was adjusted to 11 with 1 M NaOH. The preparation was centrifuged at 100000 × g for 1 h. The pellet was resuspended in buffer B, diluted 10-fold with distilled water and the alkaline treatment was repeated. The final pellet was solubilized in buffer B containing 2~o Triton X-100 by stirring at 4 °C for 90 min. The extract was centrifuged at 100000 × g for 1 h and resulting supernatant recovered. Binding activity with [125I]-~-BuTx of this thymus extract was carried out as described later. Cobrotoxin-Sepharose 2B was added to the extract and shaken for 12 h at 4 °C. The gel was recovered by filtration and washed batch-wise with the following buffers: 200 ml of buffer B with 2 ~ Triton X-100, 200 ml of buffer B with lYo Triton X-100, 200 ml of buffer B with 1% Triton X-100 and 1 M NaCI and fmally 200 ml of buffer B with 0.1% Triton X-100. The washed gel was incubated with 2 bed volumes of 1 M carbamylcholine in buffer B containing 0.1% Triton X-100 for 12 h at 4 °C and the proteins present in the resulting eluate were concentrated using an Amicon Diaflo membrane (molecular weight cutoff 50 000). To investigate the effect ofpH 11 treatment, the same purification protocol was carried out omitting the two alkali-treatment steps.

198

~-BuTx binding assay The presence of ~-BuTx binding components in thymus extract was measured by a modification of the DEAE-cellulose disk assay of Schmidt and Raftery (1973), using [12sI]-0t-BuTx (New England Nuclear), calibrated by the method of Blanchard et al. (1979). The specific activity of the calibrated toxin ranged between 0.128 and 0.163 Ci/mmol. In the at-BuTx binding assay, 0.1 ml aliquots of thymus extract were incubated with 25 #1 of [ 125I]-~-BuTx solution (final concentration 10 n M [ 12sI]-~BuTx) for 10 h at 4 °C. 0.1-ml aliquots of the incubation mixture were pipetted onto the filter disks, washed and radioactivity was counted in a Beckman Biogamma II counter.

Protein assay Protein concentration was determined either by the method of Lowry et al. (1951) or by fluorescamine assay (Udenfriend et al. 1972), using bovine serum albumin as standard.

Preparation of AChR-rich membranes from Torpedo electric organ AChR-rich postsynaptic membrane fragments were prepared from the electric tissue of Torpedo californica as described by Elliot et al. (1980). The membrane preparations were subjected to alkaline extraction at pH 11 (Elliott et al. 1979, 1980; Neubig et al. 1979).

Density gradient centrifugation Linear gradients from 5 to 20~o sucrose (w/v) were made in 10 mM sodium phosphate, 50 mM NaCI, and 0.01~ sodium azide, pH 7.4 with 0.1~o Triton X-100, using a Beckman gradient maker. Torpedo AChR or bovine thymus AChR-LP solutions were incubated with 10 nM[ 12sI]-~-BuTx for 10 h at 4 ° C, loaded onto 12-ml gradients and centrifuged at 40000 rpm for 16 h at 4 °C in a Beckman SW41 rotor. Fractions of 0.6 ml were collected and the radioactivity was counted in a Beckman Biogamma II counter.

Isoelectric focusing Isoelectric focusing of AChR-LP-[ 125I]-~-BuTx complexes was performed according to Brockes and Hall (1975) with the following modifications. AChR-[ 125I]-ctBuTx complexes were obtained after a 10-h incubation at 4 °C. Focusing was carried out for 12 h at 400 V and 1 h at 800 V. Five-mm gel slices were counted in a Beckman Biogamma counter.

Amino acid analysis AChR-LP samples (13.1-13.3 #g) were hydrolyzed with 6 N HCI at 110 °C for 24 h, and then vacuum-dried. Amino acid composition was determined using a Beckman 120 amino acid analyzer.

199

Polyacrylamide gel electrophoresis Polyacrylamide gel electrophoresis in sodium dodecyl sulfate (SDS-PAGE) was performed by the method of Laemmli (1970) using 8.75 or 10~o acrylamide in the separating gel. Gels were stained for proteins with silver using a Bio-Rad silver staining kit.

Immunoblotting Polypeptides present in Torpedo AChR or in purified thymus AChR-LP were transferred from SDS-polyacrylamide slab gels onto nitrocellulose sheets and probed with anti-AChR antisera, according to Granger and Lazarides (1984). Anti-AChR antisera had been raised in rabbits by repeated immunizations with 50-100 #g of affinity purified Torpedoor fetal bovine muscle AChR, in 0.5 ml of 10 mM sodium phosphate buffer, 0.1 ~o Triton X-100, emulsified with 0.5 ml of Freund's complete adjuvant. After transfer, the nitrocellulose sheets were treated with Tris-buffered saline (TBS, 15 mM Tris-HC1, 150 mM NaC1, 1 mM EDTA, pH 7.5) and part of the sheet was stained with TBS-GT (TBS containing 0.1~o gelatin and 0.1~ Tween 20) containing 0.1~o India ink for 30 min. The nitrocellulose was rinsed in TBS-GT and washed overnight in TBS-GT buffer. Antisera against AChR diluted 4000-fold with TBS-GT were applied for 3 h, then the sheet was washed with 6 changes of TB S-GT buffer, and stained with horseradish peroxidase conjugated anti-rabbit IgG using an Immunoblot kit from Bio-Rad. All incubations were performed in rocking trays. The purified myofibrils from the pectoralis muscle of chicken were a generous gift from Dr. E. Lazarides.

1500

_jO

•

o_

h-

m 1000 rn X.__._.____~X_____.____X

i

[S J 5 0 0

~ ~, 4

rn

~1-~ ' °

1

g ×o

m I

o

~o

20

~o

" 0.2 0.4 0,6 0.8 r r i t o n l e x t r a c t (rag prof.)

~o

Triton e x t r a c t

Fig. 1. Binding of [125I]-~-BuTx to crude thymus extract. (a) Triton extract without alkaline treatment. (b) Same AChR preparation from fetal calf thymus was incubated with 5 #M unlabeled ~-BuTx before binding with [125I]-~-BuTx. [125I]-~-BuTx-AChR complex was detected with DEAE disk assay.

200 TABLE 1 EFFECT OF ALKALINE TREATMENT ON PURIFICATION OF ACETYLCHOLINE RECEPTOR PROTEIN FROM FETAL CALF THYMUS Specific binding activity of nicotinic AChR protein isolated from fetal calf thymus with [lasI]-~-BuTx according to the stage of isolation. The activity was measured by DEAE disk assay (Schmidt and Raftery 1973) as described in Methods and shown in pmol/mg protein. - = without, and + = with alkaline treatment. Alkaline treatment (pmol/mg protein) -

Triton extract

+

0.0069 +_ 0.0013 (n = 2)

Eluate of cobrotoxin-Sepharose

301 + 39.8

(n = 3)

13.8 _ 5.0

(n = 4)

1608 _+ 1123 (n = 4)

RESULTS

Purification of AChR-LP from fetal bovine thymus The presence of AChR-LP in crude thymus extracts was detected by binding assay with [ 125I]-ct-BuTx. The results of a typical experiment, carried out with non-alkali extracted thymus extract is reported in Fig. 1. Parallel control samples were preincubated with an excess (5 #M) of unlabelled 0t-BuTx to inhibit specific binding of the radiolabelled species. The unlabelled toxin clearly inhibited [~25I]-~t-BuTx binding activity. The concentration of [~25I]-~-BuTx binding sites in fetal calf thymuses was 0.04-0.3 pmol/g tissue, a value which agrees with those reported previously (0.05-0.29 pmol/g tissue; Raimond et al. 1984). Alkaline treatment carried out before solubilization with detergents was very effective in removing peripheral, extrinsic membrane proteins and proteins non-covalently attached to the AChR-LP molecules and it greatly improved AChR-LP purification. The efficacy of this treatment is quanti-

TABLE 2 ISOLATION OF ACETYLCHOLINE RECEPTOR PROTEIN FROM FETAL CALF THYMUS Purification process of AChR from fetal calf thymus. Solubilization of AChR protein in thymus tissue was started after two alkaline treatments of thymus homogenates. I and II means different lots of isolation and the age of fetal calf thymus is different.

Thymus tissue (g) Triton extract (pmol) Eluate of cobrotoxin-Sepharose (pmol) The values in parentheses mean total protein.

I

II

883 342.2 (19.0 mg) 123.5 (37.2 #g)

1059 23.2 (3.2 mg) 8.16 (11.34 #g)

201 3600

O A

lib 300C 2400

~

180C

I ~ t l

12 O0 600

o~

...... : , . , , ! ~ c . ( c ~ . ~ . ~ . . ~ , 10

20 30 40 Fraction n u m b e r

50

Fig. 2. Sucrose gradient sedimentation. Linear density gradients (5-20%) were run by using samples of AChR bound with [125I]-~t-BuTx from Torpedo californica(A), or from fetal calf thymus (B).

fled in Table 1. At the stage of the crude Triton extract, alkaline treatment caused a 2000-fold enrichment in AChR-LP, presumably by removing extrinsic membrane and membrane-attached proteins. AChR-LP purified from alkaline treated fetal calf thymus membranes had an average specific activity of 1608 + 1123 (mean + SD) pmol/mg protein, as opposed to 301 + 39.8 when the pH 11 treatment was not performed. A summary of the yields obtained in two different isolations of AChR-LP from fetal calf thymus is reported in Table 2. The preparations were treated at pH 11 twice before solubilization with Triton X-100. The yield after affinity chromatography was 36% and 35% of the ~-BuTx binding sites present in the crude Triton extract in lot I and II, respectively.

Density gradient centrifugation Samples of purified AChR-LP from fetal calf thymus and solubilized AChR-rich membrane fractions from Torpedocalifornica electric organ were centrifuged in a 5-20 % sucrose gradient containing 10 mM sodium phosphate, pH 7.4, and 0.1% Triton X- 100. The results of a typical experiment are reported in Fig. 2. Torpedo AChR sedimented

25O A

200

8 7 6

'

350

B

8

30C

~

7 6

~ 25C

150

5

5 20C

a

100

3 ~

15C

a

10C 50 O

5c 2

4

6 15 10 12 14 Slice n u m b e r

16

18

0

ii,it,tt

tK 1

3

5 7 9 11 13 Slice n u m b e r

2 1

O 15 17 19

Fig. 3. Isoelectric focusing of AChR bound with [125I]-~-BuTx. (A) Solubilized membrane fractions of electric organ from Torpedocalifornica with Triton X-100. (B) Purified AChR from fetal calf thymus.

202 as two components, with sedimentation coefficients of approx. 9.5 and 13 S, which are known to correspond to the monomeric and dimeric forms of AChR, respectively (Raftery et al. 1972). Purified AChR-LP from fetal calf thymus migrated as single species, having a sedimentation similar to, although slightly lighter than, Torpedo AChR monomers (Fig. 2).

Isoelectric focusing Isoelectric focusing revealed 5.75 for bovine thymus AChR-LP-[ t25I]-~-BuTx complexes (Fig. 3B). This is very similar to the value of 5.76 we found for TorpedoAChR (Fig. 3A), which agrees with the values reported in the literature (reviewed in ContiTronconi and Raftery 1982).

Amino acids analysis The amino acid composition of two different preparations of purified fetal calf thymus AChR is reported in Table 3. Comparison of our findings with those reported for Torpedo californica (Vandlen et al. 1979) and fetal calf muscle (Gotti et al. 1982) AChRs, reveals a remarkable similarity between AChR-LP from bovine thymus and fetal calf muscle AChR (Gotti et al. 1982). Four amino acids (Ser, Gly, lie, Tyr) had a molto ratio closer to AChR from fetal calf muscle than Torpedo electric organ. The values obtained for other amino acids (Asp, Thr, Pro, Val, Phe) was closer to those reported for Torpedo californica AChR.

TABLE 3 AMINO ACID COMPOSITION OF PURIFIED AChR (mol %) I and II means different lots of isolation from fetal calf thymus.

Asp Thr Ser Glu Pro Gly Ala Cys

I

II

10.8 6.2 7.0 11.8 5.7 9.2 7.8

10.3 5.3 7.4 11.0 5.3 10.8 7.2

Val

7.5

Met

trace

lle

Leu Tyr Phe His Lys Arg

4.5 9.1 2.9 4.8 2.0 5.6 5.5

7.0 4.2 8.4 2.9 4.2 2.0 4.8 4.3

203

A

B C D

| 5K

4K

Fig. 4. SDS-PAGEofpurifiedAChRon 10% acrylamidegelswith0.1% SDS. Gelswerestainedwithsilver. (A) and (B) PurifiedAChRfrom fetal calf thymusof differentlot. (C) Triton extract of membranefraction of electric organ obtained from T. californica. (D) Low molecular weight protein standards (Bio-Rad).

Subunits composition of bovine thymus A ChR-LP Upon SDS-PAGE analysis on 10~ acrylamide slab gels, stained by silver staining, purified AChR-LP preparations showed up to 6 polypeptide bands, of apparent molecular weights 40000, 43000, 51000, 56000, 58000, and 66000 (Fig. 4). The 43 kdalton protein band was present in purified AChR-LP preparations even after two alkaline treatment cycles (Fig. 5a). Immunoblot staining with anfi-Torpedo-AChR rabbit sera revealed a four subunits pattern (Fig. 5b, lane A), whose molecular weights (40 000, 51000, 56 000 and 66 000) were very similar to those of Torpedo AChR subunits (Fig. 5b, lane C). The 43 kdalton peptide band reacted only marginally with anti-Torpedo AChR antisera (Fig. 5b). The AChR preparation from fetal calf thymus also reacted with antibody against AChR from fetal calf muscle, indicating antigenic similarity between these two proteins (data not shown).

204

a

C

Fig. 5. (a) SDS-PAGE on 10% acrylamide slab gel stained with silver. A was purified AChR sample from fetal calf thymus. (B) Myofibrils from chicken pectoralis muscle, used as a source of actin. (C) Triton extract of membrane fraction of electric organ from T. cal~fornica.The ~t,8, ~, and 5 show the four subuults of AChR. (b) Immunoblot of SDS-PAGE slabgel run simultaneously as same as in (a). A, B, C, are the same samples used for (a). The polypeptides transferred onto nitrocellulose were stained with anti-Torpedo-AChR-rabbitIgG and horseradish peroxidase conjugated anti-rabbit IgG goat serum.

DISCUSSION W e report here the isolation from bovine thymus and partial characterization of a m e m b r a n e protein which, like muscle and Torpedo A C h R s , is extracted by the non-denaturing detergent Triton X-100, irreversibly binds ~t-BuTx and has strong structural similarities with muscle A C h R . It is a complex protein whose molecular weight, as determined by density gradient centrifugation, isoelectric point and amino

b

205 acids composition are very similar to those found for mammalian muscle and fish electroplax AChR. The purified ~-BuTx binding protein is formed by at least four subunits of molecular weights similar to muscle and elcctroplax AChR subunits. Four of these constituent peptides cross-reacted with antibodies raised against Torpedo and bovine muscle AChR. We can therefore conclude that the thymus does indeed contain a protein with strong structural and immunological similarities to muscle (AChR-LP). Preparations of purified thymus AChR-LP had an average specific activity of 1.61 + 1.12 nmol of ~-BuTx bound/rag of protein, i.e., approx. 22~ of the maximum theoretical activity of pure Torpedo AChR (270 000 dalton). This can be explained by (i) the presence of extra protein bands, in addition to the four subunits recognized by the anti-Torpedo AChR antibodies, which could be contaminants; and (ii) the use, during the purification procedure, of large amounts of iodoacetamide, which has been reported to partially destroy the ability of AChR to bind ~-BuTx (Sumikawa et al. 1982). The AChR-LP concentration in fetal bovine thymus reported here is in the same range as that reported for the thymus of normal children, i.e., 0.05-0.29 pmol/g tissue (Raimond et al. 1984). No ~-BuTx binding activity has been found in normal human adult thymus (Nicholson and Appel 1977). In normal adult human limb muscle, the AChR content is 0.18-2.55 pmol/g tissue (Momoi and Lennon 1982), in fetal bovine muscle it is 2.45-5 pmol/g tissue (Gotti et al. 1982). Young or fetal thymus therefore contains 10-100 times less AChR than muscle, and this scarcity well explains the difficulties so far encountered in its isolation. A key step for successful isolation of AChR-LP is the treatment at pH 11, which had been shown to extract peripheral membrane proteins without affecting the function and ~-BuTx binding properties of the AChR (Neubig et al. 1979; Elliott et al. 1979, 1980). The efficacy of the alkaline treatment in improving the yield and purity of AChR-LP preparations is proven by the data reported in Table 1, where AChR-LP prepared with and without pH 11 extraction are compared. Alkaline treatment yielded a preparation of AChR-LP 5 times purer than without this treatment. The low specific activity previously reported for AChR-LP isolated from fetal bovine thymus 0.41-2.04 pmol/mg protein (Kawanami et al. 1984) and 1.39-2.14 pmol/mg protein (Kawanami et al. 1987) are probably due to the lack of this crucial step, and consequent isolation of AChR-LP having other protein components still attached to it. In the case of Torpedo AChR, the main AChR-associated protein(s) has, upon SDS-gel electrophoresis, a molecular weight of 43 000 which can be extracted by alkaline treatment or with lithium salicylate (Neubig et al. 1979; Elliot et al. 1979, 1980). This 43 kdalton protein band contains three components, which can be separated by isoelectric focusing and have pI values of 7.0, 8.0 and 5.6, respectively. One of the 43 kdalton components has been shown to be muscle-type creatine kinase (Perryman et al. 1985; Froehner et al. 1981; Giraudat et al. 1984; Barrantes et al. 1985); another one, of acidic pI (5.6), has been shown to be actin (Strader et al. 1980; Porter and Froehner 1983). One or more other 43 kdalton component(s) are localized on the cytoplasmic surface of the postsynaptic membrane in both electric organ (Sealock et al. 1984; Porter and Froehner 1985) and muscle (Bloch and Resneck 1984; Burden 1985). Protein(s) of 43 kdalton and actin have been proposed to be involved in the immobilization of the AChR in clusters within

206 or outside the postsynaptic membrane (Conti-Tronconi et al. 1982; Bloch and Resneck 1984). In addition, actin copurified with mammalian muscle AChR (Conti-Tronconi et al. 1982),//-cytoplasmic sonic actin is associated with AChR clusters in rat muscle cultures (Lubit 1984), and a 43 kdalton protein associated with the AChR is an actinbinding protein (Walker et al. 1984). Therefore, there are consistent indications that the AChR is anchored to cytoskeletal/extrinsic membrane proteins including actin. Actin is present in the thymus, and thymus actin has been purified by affinity chromatography (Lindberg and Skoog 1970) by virtue of its interaction with deoxyribonuclease 1, of which actin is a natural inhibitor (Lazarides and Lindberg 1974). It is therefore reasonable to assume that also in the thymus the AChR-LP is anchored to cytoskeletal elements, including actin, and only after release of the AChR-LP from these attached cellular structures AChR-LP can be isolated in a highly purified form. One interesting, unresolved question about the thymus AChR-LP is whether it is most similar to innervated (junctional) or denervated (extrajunctional) muscle AChR. It is known that these two forms of muscle AChR vary substantially in their antigenic structure, turnover rate, functional consequences of curare binding, channel open time and perhaps subunit composition (Schuetze 1976; Ziskind and Dennis 1978; Dwyer et al. 1981 ; Trautmann 1982; Hall et al. 1983, 1985). Although the presence of musclelike (myoid) cells in the thymus (Van de Velde and Friedman 1970) and of AChR on cultured thymic muscle cells (Kao and Drachman 1977) has been reported, there is no evidence on innervation. Therefore AChR-LP in the thymus is probably more similar to extrajunctional non-innervated AChR. The presence in the thymus of a protein with strong structural and antigenic similarities with muscle AChR supports the notion of a primary anti-AChR sensitization within the thymus as one of the first steps in the pathogenesis of myasthenia gravis.

ACKNOWLEDGEMENTS We thank Drs. E. Lazarides and C. Wood Lazarides for instruction on immunoblot methods, and Dr. S.M.J. Dunn and other colleagues for assistance with the experiments.

REFERENCES Ahranov, A., R. Tarrab-Hazdai, O. Abramsky and S. Fuchs (1975) Immunologicalrelationshipbetween acetylcholinereceptor and thymus. A possible significancein myastheniagravis,Proc. Natl. Acad. Sci. USA, 72: 1456-1459. Barrantes, F.J., A. Braceras, H.A. Caldironi, G. Mieskes, H. Moser, E.C. Torch, Jr., M.E. Roque, T. Wallimann and A. Zechel (1985)Isolation and characterizationof acetylcholinereceptor membraneassociated (nonreceptor-o2protein) and soluble electrocytecreatine kinases, J. Biol. Chem., 260: 3024-3034. Blanchard, S. G., U. Quast, K. Reed,T. Lee,M.I. Schimerlik,R. Vandlen, T. Claudio, C. D. Strader, H.-P. H. Moore and M.A. RaRery(1979) Interaction of [~25I]-~t-bungarotoxinwith acetylcholinereceptor from Torpedo californica. Biochemistry, 18: 1875-1883.

207 Bloch, R.J. and W.G. Resneck (1984) Isolation of acetylcholine receptor clusters in substrate-associated material from cultured rat myotubes using saponin, J. Cell Biol., 99: 984-993. Brocks, J. P. and Z. W. Hall (1975) Acetylcholine receptors in normal and denervated rat diaphragm muscle. II. Comparison of junctional and extrajunctional receptors, Biochemistry, 14: 2100-2106. Burden, S.J. (1985) The subsynaptic 43-kDa protein is concentrated at developing nerve muscle synapses in vitro, Proc. Natl. Acad. Sci. USA, 82: 8270-8273. Castleman, B. (1966) The pathology of the thymus gland in myasthenia gravis, Ann. N. Y. Acad. Sci., 135: 496-503. Conti-Tronconi, B. and M.A. Raf~ery (1982) The nicotinic cholinergic receptor: correlation of molecular structure with functional properties, Annu. Rev. Biochem., 51 : 491-530. Conti-Tronconi, B., C. Gotti, M. Hankapiiler and M.A. Raftery (1982) Mammalian muscle acetylcholine receptor: a supramolecular structure formed by four related proteins, Science, 218: 1227-1229. Dwyer, D. S., R.J. Bradley, R. L. Furner and G. E. Kemp (1981) Immunological properties of junctional and extrajunctional acetylcholine receptor, Brain Res., 217: 23-40. Elliot, J., S. Dunn, S. Blanchard and M.A. Raftery (1979) Specific binding of perhydrohistrionicotoxin to Torpedo acetylcholine receptor, Proc. Natl. Acad. Sci. USA, 76: 2576-2579. Elliot, J., S. G. Blanchard, W. Wu, J. Miller, C. D. Strader, P. Hartig, H-P. Moore, J. Racs and M.A. Raftery (1980) Purification of Torpedo caflfornica postsynaptic membranes and fractionation of their constituent proteins, Biochem. J., 185: 667-677. Engel, A.G. (1984) Myasthenia gravis and myasthenic syndromes, Ann. Neurol., 16:519-534. Engel, W.K., J.L. Trotter, D.E. McFarlin and C.L. McIntosh (1977) Thymic epithelial cell contains acetylcholine receptor, Lancet, i: 1310-1311. Froehner, S.C. (1981) Identification of exposed and buried determinants of the membrane-bound acetylcboline receptor from Torpedo californica, Biochemistry, 120: 4905-4915. Froehner, S. C., V. Gulhrandsen, C. Hyman, A. Y. Jeng, R. R. Neubig and J. B. Cohen (1981) Immunofluorescence localization at the mammalian neuromuscular junction of the M r 43000 protein of Torpedo postsynaptic membranes, Proc. Natl. Acad. Sci. USA, 78: 5230-5234. Fujii, Y., J. Hashimoto, Y. Monden, T. Ito, K. Nakahara and Y. Kawashima (1986) Specific activation of lymphocytes against acetylcholine receptor in the thymus in myasthenia gravis, J. ImmunoL, 136: 887-891. Girandat, J., A. DeviUers-Thiery, J-C. Perriard and J-P. Changeux (1984) Complete nucleotide sequence of Torpedo marmorata mRNA coding for the 43,000-dalton ½ protein: muscle-specific creatine kinase, Proc. Natl. Acad. Sci. USA, 81: 7313-7317. Granger, B.L. and E. Lazar ides (1984) Membrane skeletal protein 4.1 avian erythroeytes is composed of multiple variants, that exhibit tissue specific expression, Cell, 37: 595-607. Gotti, C., B. M. Conti-Tronconi and M.A. Raftery (1982) Mammalian muscle acetylcholine receptor purification and characterization, Biochemistry, 21: 3148-3154. Hall, Z.W., M-P. Roisin, Y. Gu and P.D. Gorin (1983) A developmental change in the immunological properties of acetylcholine receptors at the rat neuromuscular junction, CoM Spring Harb. Syrup. Quant. Biol., 48: 101-108. Hall, Z.W., P.D. Gorin, L. Silberstein and C. Bennett (1985) A postnatal change in the immunological properties of the acetylcholine receptor at rat muscle endplates, J. Neurosci., 5: 730-734. Kao, I. and D.B. Drachman (1977) Thymic muscle cells bear acetylcholine receptors. Possible relation to myasthenia gravis, Science, 195: 74-75. Kawanami, S., M. Yuki, K. Oda, I. Goto and Y. Kuroiwa (1984) Isolation of acetylcholine receptor-like protein from fetal calf thymus. Biochem. Int., 8: 377-384. Kawanami, S., T. Kakuno and T. Horio (1987) Purification of acetylcholine receptor like protein from fetal calf thymus, Jap. J. Psychiat. NeuroL, 41: 97-104. Lazarides, E. and U. Lindberg (1974) Actin is the naturally occurring inhibitor of deoxyribonuclease 1, Proc. Natl. Acad. Sci. USA, 71: 4742-4746. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227: 680-685. Lindberg, M.U. and L. Skoog (1970) Purification from calf thymus of an inhibitor of deoxyribonuclease 1, Eur. J. Biochem., 13: 326-335. Lindstrom, J. M. (1985) Immunobiology ofmyasthenia gravis, Experimental autoimmune myasthania gravis, and Lambert-Eaton syndrome, Annu. Rev. Immunol., 3: 109-131. Lindstrom, J.M., V.A. Lennon, M.E. Seybold and S. Whittingham (1976) Experimental autoimmune myasthenia gravis and myasthenia gravis: biochemical and immunological aspects, Ann. N. Y. Acad. Sci., 274: 254-274.

208 Lowry, O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall (1951) Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193: 265-275. Lubit, B.W. (1984) Association of fl-cytoplasmic actin with high concentrations of acetylcholine receptor (AChR) in normal and anti-AChR-treated primary rat muscle cultures, J. Histochem. Cytochem., 32: 973-981. McCarthy, M. P., J.P. Earnest, E. E. Young, S. Choe and R.M. Stroud (1986)The molecular neurobiology of the acetylcholine receptor, Annu. Rev. Neurosci., 9: 383-413. Momoi, M.Y. and V.A. Lennon (1982) Purification and biochemical characterization of nicotinic acetylcholine receptors of human muscle, J. Biol. Chem., 257: 12757-12764. Neubig, R.R., E.K. Krodel, N.D. Boyd and J.B. Cohen (1979) Acetylcholine and local anesthetic binding to Torpedo nicotinic postsynaptic membranes after removal of nonreceptor peptides, Proc. Natl. Acad. Sci. USA, 76: 690-694. Nicholson, G.A. and S.H. Appel (1977) Is there acetyleholine receptor in human thymus?, J. Neurol. Sci., 34: 101-108. Numa, S., M. Noda, H. Takahashi, T. Tanabe, M. Toyosato, Y. Furutani and K. Kikyotani (1983) Molecular structure of the nicotinic acetylcholine receptor, CoM Spring Harb. Syrup. Quant. Biol., 48: 57-69. Ong, D.E. and R.N. Brady (1974) Isolation of cholinergic receptor protein(s) from Torpedo nobiliana by affinity chromatography, Biochemistry, 13: 2822-2827. Patrick, J. and J.M. Lindstrom (1973) Autoimmune response to acetylcholine receptor, Science, 180: 871-872. Perryman, M.B., J.D. Knell, J. Ifegwu and R. Roberts (1985) Identification of a 43-kDa polypeptide associated with acetylcholine receptor-enriched membranes as MM creatine kinase, J. Biol. Chem., 260: 9399-9404. Porath, J., K. Aspberg, H. Drevin and R. Axen (1973) Preparation of cyanogen bromide-activated agarose gels, J. Chromatogr., 86: 53-56. Porter, S. and S.C. Froehner (1983) Characterization and localization of the M r 43,000 proteins associated with acetylcholine receptor-rich membranes, J. Biol. Chem., 258: 10034-10040. Porter, S. and S. C. Froehner (1985) Interaction of the 43k protein with components of Torpedo postsynaptic membranes, Biochemistry, 24: 425-432. Raimond, F., E. Morel and J.F. Bach (1984) Evidence for the presence of immunoreactive acetylcholine receptors on human thymus cells, J. Neuroimmunol., 6: 31-40. Raftery, M.A., J. Schmidt and D.C. Clark (1972) Specific activity of ct-bungarotoxin binding to Torpedo californica electroplax, Arch. Biochem. Biophys., 152: 882-886. Raftery, M.A., M.W. HunkapiUer, C.D. Strader and L.E. Hood (1980) Acetylcholine receptor: complex of homologous subunits, Science, 208: 1454-1457. Richman, D. P., C. Gomez, P. Berman, S. Burres and B. G. W. Arnason (1980) Monoclonal anti-acetylcholine receptor antibodies can cause experimental myasthenia, Nature, 286: 738-739. Schmidt, J. and M.A. RaRefy (1973) A simple assay for the study of solubilized acetylcholine receptors, Anal. Biochem., 52: 349-354. Schuetze, S. (1986) Embryonic and adult acetylcholine receptors: molecular basis of developmental changes in ion channel properties, Trends Neurosci., 9: 386-388. Sealock, R., B. E. Wray and S. C. Froehner (1984) Ultrastructural localization of the Mr 43,000 protein and the acetylcholine receptor in Torpedo postsynaptic membranes using monoclonal antibodies, J. Cell Biol., 98: 2239-2244. Strader, C., E. Lazarides and M.A. Raftery (1980) The characterization of actin associated with postsynaptic membranes from Torpedo californica, Biochem. Biophys. Res. Commun., 92: 365-370. Sumikawa, K., E.A. Barnard and J.O. Dolly (1982) Similarity of acetylcholine receptors of denervated, innervated embryonic chicken muscle. 2. Subunit compositions., Eur. J. Biochem., 126: 473-479. Toyka, K. V., D. B. Drachman, A. Pestronk and I. Kao (1975) Myasthenia gravis: passive transfer from man to mouse, Science, 190: 397-399. Trautmann, A. (1982) Curare can open and block ionic channels associated with cholinergic receptors, Nature, 298: 272-275. Udenfriend, S., S. Stein, P. Bohlen, W. Dairman, W. Leimgruber and M. Weigele (1972) Fluorescamine: a reagent for assay of amino acids, peptides, proteins, and primary amines in the picomole range, Science, 178: 871-872. Ueno, S., K. Wada, M. Takahashi and S. Tarui (1980) Acetylcholine receptor in rabbit thymus: antigenic similarity between acetylcholine receptors of muscle and thymus, Clin. Exp. Immunol., 42: 436-439. Van der Geld, H. and H.J.G.H. Oosterhuis (1963) Muscle and thymus antibodies in myasthenia gravis, Vox. Sanguinis, 8: 196-204.

209 Van de Velde, R.L. and N. B. Friedman (1970) Thymic myoid cells and myasthenia gravis, Am. J. Pathol., 59: 347-367. Vandlen, R. L., W. C.-S. Wu, J.C. Eisenach and M.A. Ral~ery (1979) Studies of the composition of purified Torpedo californica acetylcholine receptor and of its subunits, Biochemistry, 18: 1845-1854. Vincent, A., G. K. Scadding, H. C. Thomas and J. Newson-Davis (1978) In vitro synthesis of anti-acetylcholine receptor antibody by thymic lymphocytes in myasthenia gravis, Lancet, i: 305-307. Walker, J. H., C. M. Boustead and V. Witzsemann (1984) The 43-K protein, el, associated with acetylcholine receptor containing membrane fragments in an actin-binding protein, EMBO J., 3: 2287-2290. Wekerle, H.Y., B. Paterson, U.P. Ketelsen and M. Feldman (1975) Striated muscle fibers differentiate in monolayer cultures of adult thymus retieulum, Nature, 256: 493-494. Ziskind, L. and M.J. Dennis (1978) Depolarizing effect of curare on embryonic rat muscles, Nature, 276: 622-623.