Effect of sera from AChR-antibody negative myasthenia gravis patients on AChR and MuSK in cell cultures

Journal of Neuroimmunology 185 (2007) 136 – 144 www.elsevier.com/locate/jneuroim

Effect of sera from AChR-antibody negative myasthenia gravis patients on AChR and MuSK in cell cultures Maria Elena Farrugia, Domenico Marco Bonifati, Linda Clover, Judy Cossins, David Beeson, Angela Vincent ⁎ Neurosciences Group, Weatherall Institute of Molecular Medicine, Department of Clinical Neurology, University of Oxford, UK Received 11 December 2006; received in revised form 10 January 2007; accepted 17 January 2007

Abstract A proportion of patients with myasthenia gravis (MG) do not have antibodies to the acetylcholine receptor (AChR). Some of these patients have antibodies to muscle specific kinase (MuSK), whereas others have neither antibody (seronegative MG, SNMG). Both MuSK antibody positive MG (MuSK-MG) and SNMG are antibody-mediated diseases but how they cause neuromuscular junction failure is not clear. One possibility is that they reduce the clustering and expression of AChRs. We looked at the effects of MuSK-MG and SNMG sera/IgG on surface AChR distribution and expression, and AChR subunit and MuSK mRNA by quantitative RT-PCR, in TE671 and C2C12 myotubes. In TE671 cells MuSK-MG sera reduced AChR expression by about 20%, but had no effect on AChR subunit or MuSK mRNA expression. In C2C12 myotubes, MuSK-MG sera caused a reduction in the number of agrin-induced clusters, but the clusters became larger and there was no significant effect on total surface AChR numbers or AChR subunit or MuSK mRNA. By contrast, SNMG sera not only reduced AChR numbers by about 20% in TE671 cells, but modestly upregulated AChR γ subunit expression in TE671 cells and both AChR γ subunit and MuSK expression in C2C12 myotubes. Thus, although these results have, disappointingly, demonstrated little effect of MuSK antibodies on AChR expression, they do imply that SNMG antibodies act on AChR-associated pathways. © 2007 Elsevier B.V. All rights reserved. Keywords: Muscle specific tyrosine kinase (MuSK); Myasthenia gravis; Seronegative myasthenia gravis; Acetylcholine receptor

1. Introduction Myasthenia gravis (MG) is an autoimmune disease characterised by failure of transmission at the neuromuscular junction (NMJ). In 80–85% of myasthenia gravis patients, the disease is mediated by antibodies to the nicotinic acetylcholine receptor (AChR; (Lindstrom et al., 1976). Many investigations, both in vitro and in passive transfer models, have shown that these antibodies reduce the number of functional AChRs at the postsynaptic membrane by increasing AChR degradation (Drachman et al., 1978), inducing complement-mediated damage to the postsynaptic membrane (Engel et al., 1977),

⁎ Corresponding author. Neurosciences Group, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Oxford OX3 9DS, UK. Tel.: +44 1865 222321; fax: +44 1865 222402. E-mail address: [email protected] (A. Vincent). 0165-5728/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jneuroim.2007.01.010

and sometimes by blocking AChR function (Burges et al., 1990). In 2001, IgG antibodies to muscle specific kinase, MuSK, were identified in 70% of the patients without AChR antibodies (Hoch et al., 2001), and have since been detected in varying proportions of patients from Europe, the USA, Japan and Taiwan (reviewed by Vincent and Leite, 2005). MuSK plays an essential role in the agrin-induced clustering of AChRs at the NMJ during development, and preliminary studies showed that MuSK-MG IgG preparations inhibited the agrin-induced clustering of AChRs in the mouse myoblast cell line, C2C12 (Hoch et al., 2001). Moreover, immunoglobulin preparations (IgG) from AChR-antibody negative MG patients, since identified as positive for MuSK antibodies, transferred electrophysiological defects to mice (Mossman et al., 1986; Burges et al., 1994). But when motor endplates were studied in biopsied muscles, Shiraishi et al. (2005) reported no loss of AChR numbers or evidence of immune-complex deposition in

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

MuSK-MG patients, suggesting that the MuSK antibodies do not act by fixing complement and do not reduce AChR numbers overall. It is unclear, therefore, how MuSK antibodies cause a transmission defect in mature muscle, and some authors have questioned their pathogenicity (Selcen et al., 2004). Moreover, the target(s) and mechanism(s) of action of those antibodies in AChR/MuSK antibody negative MG patients (SNMG) are also unexplained. Here we tried to throw light on these questions by quantifying the effects of MuSK-MG and SNMG sera and IgG preparation on AChR and MuSK expression in the TE671 cell line, that expresses human AChR, or in the mouse C2C12 myotubes that provide a well-established model for studying processes involved in formation and turnover of the NMJ. 2. Materials and methods Sera or plasma samples were obtained from our archives. MG had been diagnosed on the basis of clinical features, neurophysiological investigations and the presence of AChR or MuSK antibodies when present. Sera were retested to confirm their antibody status. We initially used sera from a total of 13 AChR-MG sera, 12 MuSK-MG, 9 SNMG sera and 8 healthy control sera but for the main experiments, only sera from MuSK-MG 1, 2 and 3 and SNMG 1 and 3 were used (see Table 2). Table 1 Sequences of human (a) and mouse (b) probes for GAPDH, and the AChR alpha and gamma subunits and of MuSK (a) Human α subunit forward primer 5′ TGG AAG CAC TCC GTG ACC TAT 3′ α subunit reverse primer 5′ TCT GTG GGC AGG TAG AAT ACC A 3′ α subunit of AChR probe 6-FAM-5′ CCT GCT GCC CCG ACA CCC C 3′TAMRA γ subunit forward primer 5′ GTA CCG TCG CCA TCA ACG T 3′ γ subunit reverse primer 5′ AGC ATT CAC GAC AAT GAG GAT 3′ γ subunit probe 6-FAM-5′ ACT CAT CAG CAA GTA CCT GAC CTT 3′-TAMRA MuSK forward primer 5′ TCG GGA CAG CAT ATT CCA AAG T 3′ MuSK reverse primer 5′ GGA GCC CGC AGG ATC CT 3′ MuSK probe 6-FAM-5′AAG CTG GAA GTT GAG GTT TTT GCC AGG AT 3′-TAMRA GAPDH forward primer 5′ ACC ACA GTC CAT GCC ATC AC 3′ GAPDH reverse primer 5′ TCC ACC ACC CTG TTG CTG TA 3′ (b) Mouse α subunit forward primer α subunit reverse primer α subunit probe γ subunit forward primer γ subunit reverse primer γ subunit probe MuSK forward primer MuSK reverse primer MuSK probe

5′ GCT CTG TGG TGG CCA TTA 3′ 5′ CCG CTC TCC ATG AAG TTA CTC A 3′ 6-FAM-5′ CCC GGA AAG TGA CCA GCC CGA 3′-TAMRA 5′ TTG GCA GAA CTG TTC CCT CAT3′ 5′ TGG CTC AGC TGC AAG TTG AT3′ 6-FAM-5′ TTC CAA TCC CAG ACT TAC AGC ACC AGT GA 3′-TAMRA 5′ GGG CAC AGC TTA CTC CAA ACT G 3′ 5′ TTC AGG AGC ACG CAG GAT TC 3′ 6-FAM-5′ TTC CCA AAA CCT CCA CTT CCA GCT TCA 3′-TAMRA

The human GAPDH probe and the primers and probes for rodent GAPDH were supplied as standard by Applied Biosystems and the details of the sequences are not known.

137

2.1. Cell cultures TE671 cells (rhabdomyosarcoma, a human-derived cell line that expresses native human AChR (α12, β, δ, γ)) were grown in Dulbecco's Modified Eagle's Medium (DMEM) from SIGMA supplemented with 10% foetal calf serum and 100 U/ml penicillin and 100 U/ml streptomycin. In some experiments, cycloheximide (SIGMA) was applied to culture media at a concentration of 40 μg/ml to inhibit the synthesis of new AChRs. C2C12 myoblasts are a mouse-derived cell line (Silberstein et al., 1986). They were grown to 60–70% confluence in 175 cm2 flasks in DMEM supplemented with 15% foetal calf serum and 100 U/ml penicillin and 100 U/ml streptomycin. They were trypsinised and replated into 35 mm petri-dishes. For clustering studies, at 80% cell confluence the culture medium was replaced by differentiating medium consisting of DMEM supplemented with 10% horse serum and 100 U/ml penicillin and 100 U/ml streptomycin. Differentiation was allowed to take place over at least 5 days. 2.2. Agrin production For agrin expression, COS-7 cells (African green monkey renal epithelial cells, SV40-transformed) were grown in 175 cm2 flasks in DMEM supplemented with 10% foetal calf serum, 100 U/ml penicillin and 100 U/ml streptomycin. When approximately 70% confluent, COS-7 cells were transiently transfected with a plasmid construct for expression of rat agrin, which was kindly donated by the late Dr. Werner Hoch. This construct is referred to as N1 in the literature and denotes the C-terminal splice form of neural agrin, C-Ag12,4,8 (for details see Ferns et al., 1993). Plasmid DNA was prepared using a Qiagen plasmid Maxi kit, according to the manufacturer's instructions. The COS cells were transfected using polyethylenimine as described by Boussif et al. (1995). Culture medium containing agrin protein was collected on days 2 to 4 after transfection and concentrated approximately 5 times using Centriprep YM-30 filter. 2.3. Serum incubations The sera were first dialysed against phosphate buffered saline, then diluted in DMEM supplemented by antibiotics, complement-inactivated (heated in a water bath at 56 °C for 30 min) and filtered before use. Two to four days after plating TE671 cells, the medium was replaced by serum dilutions as above. For C2C12 myotubes, serum incubations were generally carried out on day 6 or 7 of differentiation, after changing to horse serum. 2.4. Surface AChRs measurements TE671 or C2C12 cells were incubated in 15 nM 125I-αbungarotoxin (Amersham Biosciences) and incubated for 30 min at room temperature in Hepes-buffered solution, pH 7.4. Non-specific binding of 125I-α-bungarotoxin was estab-

138

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

lished by preincubation with excess unlabelled α-bungarotoxin (SIGMA) at 0.02 mg/ml before addition of 125I-α-bungaratoxin. After the incubation, the radioactive toxin was aspirated and the adherent cells were washed three times with physiological solution. The cells were then extracted using 100 mM Tris pH 7.4 with 1% Triton X-100. Cell extracts were counted in a gamma liquid scintillation counter. Tests were carried out in triplicate. Results from incubation of cells with unlabelled α-bungaratoxin, were considered as background and subtracted from each result with radiolabelled bungaratoxin. Bound 125 I-αbungaratoxin was considered to correspond to surface AChR numbers. 2.5. Immunofluorescence C2C12 myoblasts were seeded and grown onto sterilised coverslips (coverglass /borosilicate glass number 1 thickness, 13 mm in diameter). Cells were grown and differentiated as above. Incubation conditions included preincubation of myotubes with agrin prior to the addition of serum dilutions (with or without further agrin addition), and preincubation of myotubes with serum dilutions followed by agrin later. At the end of the incubations, the cells were stained using a 1:1000 dilution of tetramethylrhodamine α-Bungarotoxin (Molecular Probes) in DMEM supplemented with antibiotics. Cells were incubated at room temperature for 20 min, protected from daylight. Cells were washed and fixed in 3% formaldehyde/PBS. The coverslips were mounted onto slides (Superfrost BDH 1 mm thick) after having added a drop of DAKO® Faramount Aqueous Mounting Medium. OpenLab 3.1 program was used to visualise the myotubes and count the clusters in each field (158 μm × 120 μm; 19,000 μm2) at a magnification of × 40 (Olympus 40×/0.75 ∞ 0.17). For each experiment, 20 fields were examined. Clusters were defined as larger than 3 μm in length (Fuhrer et al., 1999; Marangi et al., 2002; Moransard et al., 2003) and results were averaged. 2.6. Quantification of messenger RNA by RT-PCR Total RNA was extracted using RNA-Bee (Tel-Test, INC) from cells grown in 35 mm petri-dishes and treated with serum in triplicate samples. RNA was subject to DNase I digestion and reverse transcribed using the Retroscript™ kit (Ambion). The level of respective first strand cDNAs was analysed using real time RT-PCR. Primers and probes were designed using Primer Express® software. The GeneAmp 5700 SDS software was used for these reactions. The thermal cycle included a 2 minute 50 °C step, a 10 minute 95 °C step and 40 cycles, each consisting of a “melt” phase at 95 °C for 15 s and an annealing/extension phase at 60 °C for 1 min. Details of sequences for primers (Invitrogen) and probes (Applied Biosystems) for the human and mouse α, γ and ϵ subunits of the AChR and of MuSK are given in Table 1. The internal standards applied for these experiments were Human GAPDH (VIC-TAMRA; Applied Biosystems which includes

Table 2 Clinical details of selected patients and results of mRNA expression in TE671 cells Antibody Healthy (n = 5) AChR-MG1 AChR-MG2 AChR-MG3 Mean ± SD MuSK-MG1 MuSK-MG2 MuSK-MG3 MuSK-MG4 MuSK-MG5 Mean ± SD SNMG1 SNMG2 SNMG3 SNMG4 Mean ± SD

Age MGFA grade ΔCt onset maximum AChR α 18 32 31

3a 3a 2b

6 25 3 27 24

3b 2b 4b 4b 3b

42 17 11 12

2b 3a 3a 4a

3.42 ± 1.0 3.4 2.4 2.7 2.81 ± 0.51 3.2 3.9 3.6 4.0 4.0 3.75 ± 0.34 3.7 2.1 2.8 2.3 2.75 ± 0.68

ΔCt AChR γ

ΔCt MuSK

11.33 ± 3.12 9.49 ± 0.97 8.7 8.4 10.8 10.0 10.7 9.6 10.07 ± 1.18 9.33 ± 0.83 10.6 7.8 10.3 9.1 13.0 10.8 9.6 8.3 11.8 9.9 11.06 ± 1.34 9.18 ± 1.21 9.3 9.1 8.4 6.6 9.0 – 8.4 7.5 8.78 ± 0.45 a 7.73 ± 1.27

SNMG1 and MuSK1 were chosen because they had been shown previously to inhibit (SNMG1) or not inhibit (MuSK1) AChR function in TE671cells (Spreadbury et al., 2005). a One way ANOVA, Bonferroni post-test, p = b0.05.

reagents for primers and probe) and Taqman® Rodent GAPDH Control Reagents (Applied Biosystems) consisting of VIC™ Probe 20 μM, Rodent GAPDH reverse primer 10 μM and forward primer 10 μM). Standard curves were set up by testing undiluted cDNA, and dilutions of cDNA of 1:5, 1:25 and 1:125. The reaction was considered efficient if the gradient of the standard curve was − 3.5 and if the correlation was 0.99. If the reaction was deemed to be inefficient, based on these criteria, then it was repeated until better results were achieved. Ct values were obtained by reading off the cycle value (Ct) at which fluorescence crosses the threshold value (Rn) at the maximal gradient of the slope on a logarithmic scale. The change in cycle threshold or (ΔCt), which was utilised for subsequent data analysis, represented the difference between the Ct value for undiluted cDNA for a specific gene and that for GAPDH. Reactions were always performed in triplicate. 2.7. Statistical analysis Results were analysed using PRISM 3.0 software, applying ANOVA and Mann Whitney nonparametric tests. 3. Results 3.1. Effects of AChR expression in TE671 cells Sera from a number of MG and healthy patients were first tested at 1:10 dilution to see if they reduced AChR expression in the TE671 cell line. The AChRs were labelled after 24 h of serum incubations, and the results were expressed relative to those in the healthy control sera (Fig. 1a). As expected from previous studies AChR-MG sera caused a marked reduction in surface 125 I-α-bungaratoxin binding to AChRs in TE671 cells, likely due to increased internalisation and degradation (Tzartos

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

139

Fig. 1. Surface AChR numbers in TE671 cells, which were radiolabelled with 125I-α-bungaratoxin following incubation with sera. Numbers were expressed as a percentage of values in cells treated with healthy control sera. (a) TE671 cells treated with sera for 24 h. (b) Time-course of AChR numbers in TE671 cells at 8, 24 and 48 h. (c) TE671 cells treated with sera and cycloheximide for 14 h and surface AChRs expressed as a percentage of cells treated with healthy control sera and cycloheximide. “N” denotes the number of sera tested for each experiment.

et al., 1986) (Fig. 1a). MuSK-MG and SNMG sera reduced surface AChR numbers by about 20%. One way ANOVA was highly significant (p b 0.0001), and Bonferroni post-tests for comparison of the results of each test group with the healthy group were significant (Fig. 1a). When we performed a time-course incubating the cells for between 8 and 48 h with 3–4 sera from each group, the results were similar (Fig. 1b) at 8 and 24 h. However, after a 48 hour incubation, the AChR numbers in each case tended to increase (Fig. 1b), perhaps because of a compensatory increase in AChR synthesis. To examine this further, we used cycloheximide to arrest protein synthesis. Preliminary studies showed that 14 hour incubation in 40 μg/ml cycloheximide substantially reduced the number of surface AChRs in untreated cells overall (to about 25%) but did not reduce cell viability. Fig. 1c shows the results of incubations with 2–6 sera from each group for 14 h with cycloheximide. Under these conditions, AChR-MG sera abolished AChR expression N 95%, but the results in the MuSK-MG and SNMG sera were not significantly different from those in healthy sera. These results suggest that the relative lack of effect of MuSK-MG and SNMG sera is not due to a compensatory rise in AChR synthesis. 3.2. mRNA expression of AChR subunits and of MuSK in TE671 cells To see, however, if there were any changes in AChR mRNA, we used quantitative RT-PCR to measure mRNA for AChR α and γ subunits and MuSK, after incubation of the cells in the

diluted sera (1:10) for 48 h. The results were expressed as the change in Ct (cycle threshold) value (ΔCt) comparing the test values with that of GAPDH (as described above). We tested only selected MuSK-MG and SNMG sera (see Table 2). Delta Ct values were somewhat lower for α AChR subunit mRNA after incubation in AChR-MG sera and for α and γ AChR subunit mRNAs after incubation in SNMG sera (suggesting upregulation of mRNA expression), but only the results for the γ AChR subunit in SNMG sera were significant. MuSKMG sera had no apparent effect on AChR subunit expression (Table 2). 3.3. Effects of AChR expression and clustering in C2C12 myotubes Although the TE671 cells express MuSK, they may not have the necessary intracellular machinery to demonstrate a MuSK antibody-induced effect. We, therefore, turned to the C2C12 myotubes which, although a mouse cell line, bind MuSK antibodies and show inhibition of agrin-induced AChR clustering by MuSK antibody positive sera (Hoch et al., 2001). Agrin-induced cluster numbers were measured on immunofluorescent images (as in Fig. 2a). Agrin alone caused a substantial rise in the number of AChR clusters (Fig. 2b). In the absence of agrin, MuSK-MG1 serum alone caused a small increase in the number of AChR clusters, which was slightly higher than the spontaneous clustering that occurs in the absence of agrin (Fig. 2b; as previously shown in Hoch et al., 2001). Incubation with MuSK-MG1 serum added simulta-

140

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

Fig. 2. Effect of sera on agrin-induced clusters in C2C12 myotubes. (a) AChRs were visualised using immunofluorescence with tetramethylrhodamine αbungarotoxin. The asterisks denote the location of each cluster that was counted. (b) The number of clusters/field plotted with time of incubation with agrin, MuSKMG serum (MuSK-MG 1) with agrin, MuSK-MG serum (MuSK-MG 1) alone, or with no agrin or serum. (c) The decline in cluster numbers with time after agrin is removed from the medium. This is accelerated by the presence of MuSK-MG serum. The numbers in brackets denote the number of sera tested. (d) Time-course of the number of clusters/field in myotubes incubated with healthy control serum or MuSK-MG serum (MuSK-MG 1) and agrin for 12 h after a 12 hour preincubation with agrin.

neously with agrin reduced the number of clusters compared with agrin alone (Fig. 2b). MuSK-MG1 serum applied for 6–12 h before 6–18 h of agrin application significantly reduced the numbers of clusters compared with those in healthy serum alone (Table 3). To see if the sera affected the number of preformed clusters, and to test the effects of SNMG sera, different dilutions of three MuSK-MG and two SNMG sera were incubated with myotubes for 3 h after removal of agrin (preincubated for six hours). MuSK-MG sera but not SNMG sera reduced the clusters at 1:20 dilutions (Fig. 2c). Fig. 2d shows the effects of two dilutions of MuSK-MG1 serum on cluster numbers in C2C12 myotubes preincubated with agrin for 12 h. MuSK-MG serum, even at

1/100 dilution, substantially reduced the number of clusters, whereas healthy serum showed no difference from agrin alone. We noticed that the size of the clusters appeared to be increased by MuSK-MG sera, an effect that had not been reported previously (Hoch et al., 2001). The lengths of the clusters were measured (Fig. 3a) and the results for the experiment in Fig. 2d are shown in Fig. 3b. With longer durations of agrin treatment, there was a slight increase in the cluster length in the presence of agrin alone or with healthy serum, but a greater increase in the presence of MuSK-MG1 serum (Fig. 3b and Table 3). The results in 1/100 dilution of MuSKMG1 serum overlapped with those of the 1/20 dilution and are therefore not shown.

Table 3 Results of different incubations with MuSK-MG1 or control sera (1:20 dilution) on agrin-induced cluster numbers and cluster lengths in C2C12 cells Length (hours) of serum incubation

Length (hours) of further incubation with agrin in absence of serum

Cluster numbers per field

Cluster numbers per field

Cluster length (μm)

Cluster length (μm)

Healthy

MuSK-MG1

Healthy

MuSK-MG1

6 6 12 6

6 12 12 18

17.8 ± 4.8 16.3 ± 3.0 15.1 ± 3.3 14.0 ± 3.3

10.0 ± 4.0, p b 0.001 a 10.7 ± 1.3, p b 0.001 a 9.3 ± 4.0, p b 0.05 a 9.8 ± 2.0, p b 0.001 a

9.0 ± 1.2 11.1 ± 1.0 11.3 ± 2.1 10.6 ± 2.2

12.8 ± 4.4, p b 0.001 a 12.3 ± 1.7, NS 14.2 ± 4.2, p b 0.05 a 16.7 ± 3.0, p b 0.001 a

a

One way ANOVA on Dunnett's multiple tests comparing with results from C2C12 myotubes incubated with agrin and healthy control serum.

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

141

Fig. 3. Effects on length of clusters and removal of agrin. (a) Within each field studied, clusters were identified and counted and the length of each cluster recorded. (b) The mean length of each cluster for the same conditions as in the experiment in Fig. 2d. (c) AChR measurements in C2C12 myotubes, preincubated with agrin for 36 h, then with serum for 24 h in the absence of agrin. (d) The rate of decline in AChR numbers following addition of serum to 125I-bungarotoxin-labelled C2C12 myotubes. For these experiments, MuSK-MG 1 and 2 and SNMG 1 and 3 were used.

3.4. Surface AChR numbers in C2C12 myotubes These results suggested that despite the reduction in numbers of AChR clusters, the total number of AChRs might not be substantially reduced. C2C12 myotubes were preincubated with agrin for 36 h, in an attempt to simulate physiological conditions, then incubated with serum for 12, 24 or 48 h, before the myotubes were incubated in 125I-α-BgTx. Incubation with three MuSK-MG and three SNMG sera had no significant effect on subsequent surface AChR numbers (Fig. 3c). To see whether there might be an effect on the turnover of the AChRs, the cells were pre-labelled with 125I-α-bungarotoxin before the addition of agrin and sera (MuSK-MG 1 and 2; SNMG 1 and 3). The AChR numbers fell over time, as expected due to their normal rate of internalisation and degradation, but there was no additional effect of MuSK-MG or SNMG sera (Fig. 3d). 3.5. mRNA expression of AChR subunits and MuSK in C2C12 myotubes Finally, we used quantitative RT-PCR to measure mRNA levels of mouse AChR and MuSK. We used purified IgG from a pool of three healthy sera, MuSK-MG1 and SNMG1 sera. The C2C12 myotubes were pretreated with agrin for 36 h and then dilutions of the IgG samples (equivalent to 1:20 of serum levels) were added for a further 12 h. MuSK-MG IgG had no

significant effect on AChR or MuSK expression, although there was a trend towards downregulation of the AChR α mRNA expression (increased ΔCt values) and upregulation of the AChR γ mRNA. Although none of the results were significant by ANOVA, there was a trend towards upregulation of the AChR α gene (lower ΔCt values) in the presence of the SNMG sera and a significant upregulation of both the AChR γ and MuSK mRNAs (Fig. 4). 4. Discussion MuSK antibodies have, to date, only been found in patients with myasthenia gravis and only in those patients without AChR antibodies, except in one report (Ohta et al., 2004), which was subsequently corrected. The manner by which these antibodies reduce neuromuscular transmission is not known, but a reasonable hypothesis is that they reduce the number or stability of AChRs. We used two in vitro cell lines to investigate the effects of MuSK antibodies on AChR numbers, distribution and mRNA expression. We found a small reduction in AChR numbers in TE671 cells but unexpectedly no change in C2C12 myotubes when incubated in the presence of MuSK-MG sera over varying time periods and at different concentrations. Moreover, the effect on agrin-induced AChR clusters in C2C12 myotubes, previously reported (Hoch et al., 2001), appears to relate as much to coalescence of small clusters to form larger ones, as to dispersion of clusters. Overall there was little effect

142

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

Fig. 4. mRNA expression of AChR subunits and MuSK in C2C12 cells after addition of sera. RT-PCR on C2C12 myotubes preincubated with agrin for 36 h, then IgG for MuSK-MG1 and SNMG 1 added for 12 h. Results were expressed in relation to the housekeeping gene, GAPDH, as ΔCt for alpha (a), gamma (b) and MuSK (c).

of the MuSK-MG sera on AChR or MuSK expression, whereas SNMG sera marginally increased AChR γ subunit and MuSK expression. Thus these experiments, which are the first to quantify AChRs in C2C12 cells treated with MuSK-MG antibodies, do not demonstrate a potentially pathogenic effect on AChR or MuSK expression. MuSK is crucial for neuromuscular development and MuSK deficient mice have profound NMJ defects as do agrin-deficient mice (DeChiara et al., 1996; Gautam et al., 1996). Myotubes derived from these mice failed to undergo agrin-induced AChR clustering (Glass et al., 1996). AChR cluster formation is dependent on rapsyn (Frail et al., 1988; Froehner, 1991) and mice that are null for rapsyn fail to form AChR clusters during embryological formation (Gautam et al., 1995). However, the importance of MuSK in the maintenance of the mature NMJ has not been clear until recently when Kong et al. (2004) showed that inhibition of MuSK synthesis in rodent muscle, by injection of siRNA, leads to a slow but dramatic dispersion of AChRs and disruption of the NMJ. The mechanisms by which MuSK antibodies alter neuromuscular transmission, however, are not yet clear and it has been suggested that these antibodies may not be pathogenic (Selcen et al., 2004). In order to investigate their effects in vitro, therefore, we first used TE671 cells, which have provided a useful source of human AChR for in vitro and radioimmunoassay studies. We found no marked effect of either MuSK-MG or SNMG sera on AChR numbers or on AChR subunit expression. This appears to contrast with the work of Guyon et al. (1994, 1998) and Poea et al. (2000) who reported that more clinically severe MG patients, including SNMG patients, showed an upregulation of some

AChR genes both in patient-derived muscle biopsies, obtained from sternocleidomastoid muscle samples, and in in vitro experiments on TE671 cells. However, these findings were more prominent in AChR-MG than AChR-antibody negative MG patients, and the MuSK antibody status of the SNMG patients was not available at that time. Our results also contrast with a recent study by Boneva et al. (2006) which demonstrated that some MuSK-MG sera can alter morphology, growth and cell proliferation of TE671 cells when incubated for up to 72 h. Their sera also reduced mRNA expression of RhoA and cdc42 and of AChR α and β and of rapsyn — all important players in the formation of AChR clusters and in the maintenance of the postsynaptic membrane at the NMJ. Interestingly, this effect was seen with only about 50% of MuSK-MG sera and results with the other MuSK-MG sera were not distinguishable from controls suggesting that the effect may not be related to the MuSK antibody. Even in incubations of up to 48 h, we did not find any evidence suggesting an effect of MuSK-MG sera on cell numbers of TE671 cells; the reason for this discrepancy is unclear. It is difficult in any case to compare our results, since there is no evidence that they heated and dialysed their sera as we did here, and the MuSK antibody titres were generally lower than those which we find (eg. McConville et al., 2004). On the other hand, the upregulation of atrogin-1 expression (Boneva et al., 2006) parallels our findings of upregulation of another atrophy-related gene product, MuRF1, by MuSK-MG sera in C2C12 myotubes (Benveniste et al., 2005) and supports our clinical findings that MuSK-MG is associated with muscle atrophy in MRI studies (Farrugia et al., 2006).

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

The TE671 cells, however, cannot be differentiated to form myotubes, and do not express rapsyn, a component that is crucial in the signalling cascade and clustering pathway. The C2C12 myotube system allowed us to study the effect of MuSK antibodies on agrin-induced AChR clusters. We showed that MuSK-MG sera, but not SNMG sera, reduced the number of AChR clusters, as previously reported (Hoch et al., 2001). This decline in cluster numbers was associated with an increase in cluster length and no overall effect on AChR numbers. Altering the duration of exposure to agrin or to the serum or both did not make any substantial difference to the results. Quantitative mRNA measurements showed a trend towards reduced AChR α subunit mRNA in MuSK-treated cells but no other changes. These results, therefore, suggest that either the MuSK antibodies do not affect AChR numbers, or that some key component is missing in these cells. The recently identified Dok-7 protein, that is a binding partner for MuSK, could be such a protein (Beeson et al., 2006). Therefore, from our in vitro model of synaptogenesis, it remains unclear how MuSK antibodies cause MG but our data are not in disagreement with other recent findings. There is evidence to suggest that MuSK antibodies do not activate complement and do not act via complement-mediated destruction (McConville et al., 2004; Shiraishi et al., 2005). In rat muscle, silencing of MuSK RNA, using RNAi demonstrated changes in the NMJs with severe fragmentation of AChR clusters and loss of the integrity of the postsynaptic membrane but only after 6 weeks (Kong et al., 2004). We know that there are limitations in the use of the C2C12 cells in that they express the foetal isoform of the mouse AChR, and the AChRs within the agrin-induced clusters have a higher turnover rate than at the adult NMJ (Fambrough et al., 1979). This suggests that the in vitro model is insufficient to demonstrate the effects of MuSK antibodies since there is a limit to the duration of treatment of cells in culture, and the density and stability of the AChRs are different. Active immunisation models in mice or rabbits immunised with recombinant MuSK have found evidence of clinical weakness and altered neuromuscular junction morphology (Jha et al., 2006; Shigemoto et al 2006). Sera from these animals inhibited agrin-induced AChR aggregation in C2C12 myotubes (Shigemoto et al., 2006). Further active as well as passive immunisation animal models are crucial to help us unravel further the pathogenicity of these antibodies. Finally, one can ask whether these results throw any light on the nature of the target in SNMG patients. Interestingly, the results of the in vitro incubations suggested that SNMG sera might increase AChR expression to a small extent, raising the possibility that SNMG patients, who have thymic changes not unlike those in AChR-MG patients (Leite et al., 2005), may contain low levels of antibodies that bind to the AChR and lead to down-stream effects on AChR and MuSK expression. Indeed, binding of some SNMG IgG antibodies to AChRs by immunofluorescence on AChR-transfected cell lines has recently been demonstrated (Leite et al., in preparation). Half of the SNMG sera used in the TE671 cell experiments, and the SNMG serum used in the C2C12 experiments, were positive by this method.

143

Acknowledgements We are very grateful to the Myasthenia Gravis Association and Muscular Dystrophy Campaign (MEF, LC) and the Medical Research Council (JC, DB) for support. References Beeson, D., Higuchi, O., Palace, J., Cossins, J., Spearman, H., Maxwell, S., Newsom-Davis, J., Burke, G., Fawcett, P., Motomura, M., Muller, J.S., Lochmuller, H., Slater, C., Vincent, A., Yamanashi, Y., 2006. Dok-7 mutations underlie a neuromuscular junction synaptopathy. Science 313, 1975–1978. Benveniste, O., Jacobson, L., Farrugia, M.E., Clover, L., Vincent, A., 2005. MuSK antibody positive myasthenia gravis plasma modifies MURF-1 expression in C2C12 cultures and mouse muscle in vivo. J. Neuroimmunol. 170, 41–48. Boneva, N., Frenkian-Cuvelier, M., Bidault, J., Brenner, T., Berrih-Aknin, S., 2006. Major pathogenic effects of anti-MuSK antibodies in myasthenia gravis. J. Neuroimmunol. 177, 119–131. Boussif, O., Lezoualc'h, F., Zanta, M.A., Mergny, M.D., Scherman, D., Demeneix, B., Behr, J.P., 1995. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. U. S. A. 92, 7297–7301. Burges, J., Wray, D.W., Pizzighella, S., Hall, Z., Vincent, A., 1990. A myasthenia gravis plasma immunoglobulin reduces miniature endplate potentials at human endplates in vitro. Muscle Nerve 13, 407–413. Burges, J., Vincent, A., Molenaar, P.C., Newsom-Davis, J., Peers, C., Wray, D., 1994. Passive transfer of seronegative myasthenia gravis to mice. Muscle Nerve 17, 1393–1400. DeChiara, T.M., Bowen, D.C., Valenzuela, D.M., Simmons, M.V., Poueymirou, W.T., Thomas, S., Kinetz, E., Compton, D.L., Rojas, E., Park, J.S., Smith, C., DiStefano, P.S., Glass, D.J., Burden, S.J., Yancopoulos, G.D., 1996. The receptor tyrosine kinase MuSK is required for neuromuscular junction formation in vivo. Cell 85, 501–512. Drachman, D.B., Angus, C.W., Adams, R.N., Michelson, J.D., Hoffman, G.J., 1978. Myasthenic antibodies cross-link acetylcholine receptors to accelerate degradation. N. Engl. J. Med. 298, 1116–1122. Engel, A.G., Lambert, E.H., Howard, F.M., 1977. Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis: ultrastructural and light microscopic localization and electrophysiologic correlations. Mayo Clin. Proc. 52, 267–280. Fambrough, D.M., Devreotes, P.N., Gardner, J.M., Card, D.J., 1979. The life history of acetylcholine receptors. Prog. Brain Res. 49, 325–334. Farrugia, M.E., Robson, M.D., Clover, L., Anslow, P., Newsom-Davis, J., Kennett, R., Hilton-Jones, D., Matthews, P.M., Vincent, A., 2006. MRI and clinical studies of facial and bulbar muscle involvement in MuSK antibodyassociated myasthenia gravis. Brain 129, 1481–1492. Ferns, M.J., Campanelli, J.T., Hoch, W., Scheller, R.H., Hall, Z., 1993. The ability of agrin to cluster AChRs depends on alternative splicing and on cell surface proteoglycans. Neuron 11, 491–502. Frail, D.E., McLaughlin, L.L., Mudd, J., Merlie, J.P., 1988. Identification of the mouse muscle 43,000-dalton acetylcholine receptor-associated protein (RAPsyn) by cDNA cloning. J. Biol. Chem. 263 (30), 15602–15607. Froehner, S.C., 1991. The submembrane machinery for nicotinic acetylcholine receptor clustering. J. Cell Biol. 114, 1–7. Fuhrer, C., Gaurtam, M., Sugiyama, J.E., Hall, Z.W., 1999. Roles of rapsyn and agrin in interaction of postsynaptic proteins with acetylcholine receptors. J. Neurosci. 19, 6405–6416. Gautam, M., Noakes, P.G., Mudd, J., Nichol, M., Chu, G.C., Sanes, J.R., Merlie, J.P., 1995. Failure of postsynaptic specialization to develop at neuromuscular junctions of rapsyn-deficient mice. Nature 377, 232–236. Gautam, M., Noakes, P.G., Moscoso, L., Rupp, F., Scheller, R.H., Merlie, J.P., Sanes, J.R., 1996. Defective neuromuscular synaptogenesis in agrindeficient mutant mice. Cell 85, 525–535. Glass, D.J., DeChiara, T.M., Stitt, T.N., DiStefano, P.S., Valenzuela, D.M., Yancopoulos, G.D., 1996. The receptor tyrosine kinase MuSK is required

144

M.E. Farrugia et al. / Journal of Neuroimmunology 185 (2007) 136–144

for neuromuscular junction formation and is a functional receptor for agrin. Cold Spring Harbor Symp. Quant. Biol. 61, 435–444. Guyon, T., Levasseur, P., Truffault, F., Cottin, C., Gaud, C., Berrih-Aknin, S., 1994. Regulation of acetylcholine receptor alpha subunit variants in human myasthenia gravis. Quantification of steady-state levels of messenger RNA in muscle biopsy using the polymerase chain reaction. J. Clin. Invest. 94, 16–24. Guyon, T., Wakkach, A., Poea, S., Mouly, V., Klingel-Schmitt, I., Levasseur, P., Beeson, D., Asher, O., Tzartos, S., Berrih-Aknin, S., 1998. Regulation of acetylcholine receptor gene expression in human myasthenia gravis muscles. Evidences for a compensatory mechanism triggered by receptor loss. J. Clin. Invest. 102, 249–263. Hoch, W., McConville, J., Helms, S., Newsom-Davis, J., Melms, A., Vincent, A., 2001. Auto-antibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies. Nat. Med. 7, 365–368. Jha, S., Xu, K., Maruta, T., Oshima, M., Mosier, D.R., Atassi, M.Z., Hoch, W., 2006. Myasthenia gravis induced in mice by immunization with the recombinant extracellular domain of rat muscle-specific kinase (MuSK). J. Neuroimmunol. 175, 107–117. Kong, X.C., Barzaghi, P., Ruegg, M.A., 2004. Inhibition of synapse assembly in mammalian muscle in vivo by RNA interference. EMBO Rep. 5, 183–188. Leite, M.I., Strobel, P., Jones, M., Micklem, K., Moritz, R., Gold, R., Niks, E.H., Berrih-Aknin, S., Scaravilli, F., Canelhas, A., Marx, A., Newsom-Davis, J., Willcox, N., Vincent, A., 2005. Fewer thymic changes in MuSK antibodypositive than in MuSK antibody-negative MG. Ann. Neurol. 57, 444–448. Lindstrom, J.M., Seybold, M.E., Lennon, V.A., Whittingham, S., Duane, D.D., 1976. Antibody to acetylcholine receptor in myasthenia gravis. Prevalence, clinical correlates, and diagnostic value. Neurology 26, 1054–1059. Marangi, P.A., Wieland, S.T., Fuhrer, C., 2002. Laminin-1 redistributes postsynaptic proteins and requires rapsyn, tyrosine phosphorylation, and Src and Fyn to stably cluster acetylcholine receptors. J. Cell Biol. 157, 883–895. McConville, J., Farrugia, M.E., Beeson, D., Kishore, U., Metcalfe, R., Newsom-Davis, J., Vincent, A., 2004. Detection and characterization of MuSK antibodies in seronegative myasthenia gravis. Ann. Neurol. 55, 580–584.

Moransard, M., Borges, L.S., Willmann, R., Marangi, P.A., Brenner, H.R., Ferns, M.J., Fuhrer, C., 2003. Agrin regulates rapsyn interaction with surface acetylcholine receptors, and this underlies cytoskeletal anchoring and clustering. J. Biol. Chem. 278, 7350–7359. Mossman, S., Vincent, A., Newsom-Davis, J., 1986. Myasthenia gravis without acetylcholine-receptor antibody: a distinct disease entity. Lancet 1, 116–119. Ohta, K., Shigemoto, K., Kubo, S., Maruyama, N., Abe, Y., Ueda, N., Ohta, M., 2004. MuSK antibodies in AChR Ab-seropositive MG vs AChR Abseronegative MG. Neurology 62, 2132–2133. Poea, S., Guyon, T., Bidault, J., Bruand, C., Mouly, V., Berrih-Aknin, S., 2000. Modulation of acetylcholine receptor expression in seronegative myasthenia gravis. Ann. Neurol. 48, 696–705. Selcen, D., Fukuda, T., Xin-Ming, S., Engel, A., 2004. Are MuSK antibodies the primary cause of myasthenic symptoms? Neurology 62, 1945–1950. Shigemoto, K., Kubo, S., Maruyama, N., Hato, N., Yamada, H., Jie, C., Kobayashi, N., Mominoki, K., Abe, Y., Ueda, N., Matsuda, S., 2006. Induction of myasthenia by immunization against muscle-specific kinase. J. Clin. Invest. 116, 1016–1024. Shiraishi, H., Motomura, M., Yoshimura, T., Fukudome, T., Fukuda, T., Nakao, Y., Tsujihata, M., Vincent, A., Eguchi, K., 2005. Acetylcholine receptors loss and postsynaptic damage in MuSK antibody-positive myasthenia gravis. Ann. Neurol. 57, 289–293. Silberstein, L., Webster, S.G., Travis, M., Blau, H.M., 1986. Developmental progression of myosin gene expression in cultured muscle cells. Cell 46, 1075–1081. Spreadbury, I., Kishore, U., Beeson, D., Vincent, A., 2005. Inhibition of acetylcholine receptor function by seronegative myasthenia gravis non-IgG factor correlates with desensitisation. J. Neuroimmunol. 162, 149–156. Tzartos, S.J., Sophianos, D., Zimmerman, K., Starzinski-Powitz, A., 1986. Antigenic modulation of human myotube acetylcholine receptor by myasthenic sera. Serum titer determines receptor internalization rate. J. Immunol. 136, 3231–3238. Vincent, A., Leite, M.I., 2005. Neuromuscular junction autoimmune disease: muscle specific kinase antibodies and treatments for myasthenia gravis. Curr. Opin. Neurol. 18, 519–525.