982 Journal ofNeurology, Neurosurgery, and Psychiatry 1993;56:982-987

Motor conduction block and high titres of anti- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from GM1 ganglioside antibodies: pathological evidence of a motor neuropathy in a patient with lower motor syndrome

David Adams, Thierry Kuntzer, Andreas J Steck, Alexander Lobrinus, Robert C Janzer, Franco Regli Abstract and temporal dispersion included a more A patient with a progressive lower motor than 20% reduction in the negative CMAP neuron syndrome had neurophysiologi- amplitude and an increase in the negative cal evidence of motor loss, multi- duration of no more than 15% of the maxi- focal proximal conduction mum CMAP elicited by stimulation at the block, and high titres of anti-ganglioside proximal site compared with that evoked by GM1 antibodies. Neuropathological find- distal stimulation.12 In cases where stimula- ings included a predominantly proximal tion could not be applied proximally, a con- motor radiculoneuropathy with multi- duction block was admitted when the number focal IgG and IgM deposits on nerve of motor unit potentials evoked by graded fibres associated with a loss of spinal stimulation was much greater than could be motor . These findings support obtained during maximal voluntary contrac- an autoimmune origin of this lower tion or when the total area of the CMAP was syndrome with retrograde greater than the sum of the areas of the indi- degeneration of spinal motor neurons vidual motor unit potentials."3 Motor nerve and severe neurogenic muscular atrophy. conduction studies were measured over long segments at conventional sites (Erb's point, (J Neurol Neurosurg Psychiatry 1993;56:982-987) axilla, elbow, wrist) and also over short seg- ments for the ulnar nerve. F wave latencies from at least 10 responses were recorded fol- Interest in the pathological role of anti-gan- lowing stimulation at the wrist. Using surface glioside antibodies has been stimulated by the electrodes, action potentials identification of monoclonal IgM antibodies were recorded orthodromically. Needle reacting with specific gangliosides in patients examination was performed using standard with motor neuron diseases and monoclonal Dantec concentric electrodes. gammopathy.l1 Polyclonal anti-GMl anti- bodies have also been found in the absence of ENZYME LINKED IMMUNOSORBENT ASSAY a gammopathy in patients with motor neuron The presence of anti-ganglioside GMl anti- syndromes, particularly in a subgroup of bodies in serum was measured by the enzyme patients with multiple motor neuropathy and linked immunosorbent assay (ELISA) using a lower motor neuron syndrome.5-9 The pres- technique previously described.8 The activity ence of these antibodies raises the possibility against other gangliosides GM2, GM3, http://jnnp.bmj.com/ of an ongoing autoimmune process in GD1a, GDlb, asialo-GMl was also evaluat- patients with . In a few ed. The purified gangliosides were a kind gift of these patients improvement in strength has of Dr A Martelli (Fidia Research Lab- occurred after reduction of antibody titres by oratories, Abano Terme, Italy). treatment with immunosuppressive drugs.5910 Anti-GM1 antibodies may bind to the nodes PATHOLOGICAL STUDY of Ranvier," but the of this anti- Brain, spinal cord and roots, several peripher-

significance on October 2, 2021 by guest. Protected copyright. body binding pattern and its role in nerve al nerves, and muscles were fixed for three weeks in a Centre Hospitalier conduction block remain to be determined.9 10% buffered paraformaldehyde Universitaire Vaudois, We report the clinical, immunological, physi- solution. Forty samples from relevant regions Lausanne, ological and pathological findings in a patient were routinely processed. Paraffin embedded Switzerland with lower motor neuron syndrome and a and frozen sections were stained with haema- Service de Neurologie toxylin and eosin, van Gieson and Luxol and D Adams high titre of anti-ganglioside GM1 antibodies. T Kuntzer Nissl and Luxol stains. Additionally, muscles A J Steck were stained with periodic acid Schiff, retic- F Regli Methods ulin, Congo red, muscle enzyme, and lipid Institut Universitaire NEUROPHYSIOLOGY stains. Immunochemical stains were per- de Pathologie studies were formed with the A Lobrinus Electrophysiological performed following commercial avail- R C Janzer using a Mystro system (Vickers Medical able antibodies against phosphorylated Correspondence to: Corp). Multisegmental motor nerve conduc- , ubiquitin, glial fibrillary acidic Professor A Steck, Service tion studies of both arms were performed protein (GFAP), lysozyme, basic pro- de Neurologie, CHUV, 1011 Lausanne, with percutaneous supramaximal stimulation tein (MBP) and myelin associated glycopro- Switzerland. while recording the compound muscle action tein (MAG), and anti-IgG, IgM, and IgA Received 12 February 1992 potential (CMAP) with TECA 4 mm surface antibodies. Samples of spinal roots were also and in final revised form 8 December 1992 disc electrodes and needle electrodes (radial postfixed in glutaraldehyde, epon embedded, Accepted 11 December 1992 nerve). Determination of conduction block and lpum sections stained with toluidine blue. Motor neuropathy and anti-GM1 ganglioside antibodies 983

Case report nal and respiratory muscles, but the cranial J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from A 71 year old man first experienced painful nerves remained normal. There was no senso- cramps in the thighs and calves on exertion at ry deficit or symptoms. Deep tendon reflexes the end of 1983. Past medical history was not were all absent, plantar responses were neu- relevant and there was no family history of tral, and no sphincter dysfunction or fascicu- similar illness. Over the ensuing three years, lations were seen. On general examination the patient-developed progressive paraparesis, diffuse oedema of the four limbs was noted. at first characterised by asymmetrical (left Biochemical studies were within the normal more than right) difficulties in climbing stairs; range, including serum electrophoresis and then a bilateral foot drop appeared. In 1985, immunoelectrophoresis, triiodothyronine, thy- he needed a stick to walk; in summer 1986 he roxine thyroid stimulating hormone, and could only walk with the assistance of two immunoelectrophoresis. Serological tests for canes. Over the next six months, the weakness syphilis, anti-DNA antibodies, CSF, and extended to the hands and he fell frequently. blood Lyme serology were normal or negative. In December 1986 on examination the cranial The patient was treated with chlorambucil, nerves were normal but there was a moderate 9 mg daily but no improvement was observed atrophy of shoulders and intrinsic hand mus- over the next four months. Death occurred six cles and diffuse wasting of the lower limbs. years after onset. The necropsy was per- were observed in proximal formed 48 hours after death. muscles. The muscle strength was weak and asymmetrical, left side more than the right side, MRC Grade 3 + to 4/5 distally, 4 + /5 Results proximally in the upper limbs, and 2 to 4/5 NEUROPHYSIOLOGY distally, 2 to 3/5 proximally in the lower The results of the nerve conduction studies limbs. Sensory responses to light touch, pin- are summarised in the table. There were prick, vibration, and joint position sense were motor conduction blocks in four nerves out of normal. Deep tendon reflexes were absent in the eight tested on both sides, all of them the left leg and present but decreased in the located at the right Erb's point. The degree of other limbs. Plantar responses were neutral on the block was more than 50% for the radial the right side and questionable on the left. A and ulnar nerves. For the axillary and muscu- general examination revealed oedema of the locutaneous nerves the number of motor unit legs up to the knees. There was physiological potentials evoked by graded stimulation was evidence of diffuse denervation reinnervation greater than could be obtained during maxi- by needle EMG. The distal motor and senso- mal voluntary contraction. The blocks were ry nerve conduction velocities and amplitudes associated with a low distal amplitude of the of lower and upper limb sensory nerve action CMAP, expressing axonal loss. Despite the potentials were normal. A diagnosis of proba- low distal amplitude which leads to difficulty ble motor neuron disease was made. in measuring the block, the proximal and dis- The weakness progressed in the arms and tal responses were of the same shape and he was found to be tetraplegic in October could be superimposed: the difference in 1988. In November 1989 he was found to duration of CMAPs did not exceed 10% have high anti-ganglioside GM1 antibodies between distal and proximal responses. The and was reinvestigated. Examination revealed differences in amplitude were thus not simply

a severe hypotonic tetraplegia: in his arms and due to increased dispersion of CMAPs. The http://jnnp.bmj.com/ legs, only slight fingers and toes flexion could conduction velocity for the remaining nerve be seen. He also had weakness of the abdomi- fibres was nearly normal or diminished by less

Table Results ofconduction studies in motor and sensoryfibres Amplitude (i1W)

Distal latency Ratio proximal/distal Conduction velocity on October 2, 2021 by guest. Protected copyright. Nerve Proximal Distal (ms) amplitude (%) (range, mls) Motor: Right ulnarl* 0 310 5 0 44-55 7 Right radial2* 20 110 6-4 18-1 39-45 Right musculocutaneous* 640 ND 6-2 ND ND Right axillary* 300 ND 7-4 ND ND Right median ND 0 Left median3 150 180 5-6 83 35-8-46 Left musculocutaneous 0 ND Left ulnar' 160 180 6-4 88 38-45 Sensory (right side): Median digit I to wrist 8-4 3 45 Median palm to wrist 28-5 1-5 53-3 Ulnar digit V to wrist 6 3 45 Radial digit I to wrist 7-9 3 41-6 *Conduction block at Erb's point (see discussion). ND = Not done or not measured. 0 = No response could be recorded despite different stimulators and position of the patient. 'For ulnar nerve, multiple sites of stimulation included wrist, below and above elbow, axilla and Erb's point while recording CMAPs over abductor digiti minimi muscle. Amplitude ratio was calculated between the most proximal and distal site of stimulation. Between axilla and wrist, no significant drop in amplitude was seen. The nerve conduction velocity within the segment above-below elbow was normal. 2For the radial nerve, both needle and disc electrodes were used in an attempt to record the same shape of evoked response. Stimulation sites included forearm, external bicipital fossa, axilla, and Erb's point while recording the CMAPs over extensor indicis proprius muscle. Between axilla and forearm, no significant changes in amplitude was seen. 3For median nerve, stimulation sites included wrist and elbow while recording CMAPs over abductor pollicis brevis muscle. 984 Adams, Kuntzer, Steck, Lobrinus, Janzer, Regli

Figure 1 Spinal cord at J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from C5 level showing cellular details oflower motor neurons. (A) Central 4i ~ ' chromatolysis (Nissl and Luxol, bar = 50,um). (B) Diffuse expression of 4 ~ - ¾ ..- ;i^ phosphorylated proteins Macroscoically theneuro-Psigniican (immunoperoxidase methodforphosphorylated p. neurofilament, bar = 100 :I : i. Aim). consectiv preomnatl rigtsie cardiacU

consecutive peoiatly right side card.i.ac

Macrscoial the sinfcn -n;uRioo than 30% of the normal value and no evi- Baisp.a(13 g) cod an peripheral dence of superimposed focal neuropathy was found (see legend of table). No F waves could be recorded. The sensory nerve action pneuronsa(fi B aior roots (fig 2) potentials were normal in both upper and Micatroscopicallyethelmostr leions lower limbs. By needle EMG, fibrillations mothor ipal nerves (figsdimportean3ando4)anderi- and positive sharp waves were recorded at wyeretconfinedslrtesr, doinasevereto thermianteriormnurogenichrornmotoatrophy. rest in proximal and distal muscles. Thenseveriy ofted inanlsos ainth spinlcorda Insertional activity was not increased, and andervtsh anterido rormtsal me mrkedhat myokymic or complex repetitive discharges muclsM bcroscopicallygeveraiedtheloswinortaicntlevson were not seen. In proximal muscles, asyn- sathowediroshowinglessiolvemntlofneurogenictheant eriatrophy.to- chronous and arrhythmic long duration fasci- anteario noralot aoemred culation potentials were recorded. During anderhes weak effort there were prominent signs of reinnervation with prolonged, enlarged, and neraccondThe iseae.t cervicAlT he lesionsofsegments,iothespinalsuggestfingcordicona) polyphasic motor unit potentials in widely slowly afiscending cors ofd distributed muscles. The number of motor heuronsacrinperoghralnressvanalyadlowerthorantlesiond unit potentials was reduced during maximum wshie in theupperments the uprn mof effort. etorl pcromratolyti ascefigs andi te patho ELISA Theosegera thelesionof thespinal cord n- Anti-ganglioside GM1 antibodies were found ritdae ants ino othupeiarm oftemotor in high titre in the serum (788 units, normal neuronscentrly chromatoyivegrefin gh1),teoderasedigcoreandrofauiqutci linlupto- <20). Anti-ganglioside GD1a and anti-gan- log isealeaccumlation ofpheosphorylatrd cne- glioside GM3 antibodies were also found in motoeroilamentfaosinvoelur.erinthuperioary0%rmentinof theofthsppiernmotorao the serum but in a lower titre (below 100 neurons (igthe motdevoidelyubiquitin inclu- units). The anti-ganglioside GM1 antibodies siomnts,suggstedanretrogradaccompniedebretrgadeleiona lesion glofstheglofsis, http://jnnp.bmj.com/ were of IgG class.

PATHOLOGICAL STUDY mtrneurons. Other portionsiuiitheof sinalu The general necropsy findings included a on October 2, 2021 by guest. Protected copyright.

fC"I:0 7-

AD 0 - O~~~~~~~op<-.z-

4 '~

0 B~~~~~~~' M '-44~''C Figur2Aterioroo pathlogyat loer toraci levl (A)Sevee nerefire los (imunopvoxiase mthodfor nuroflamen proein br 10 jim (B)and(C)jimepon ebedded sectins stained ith toluidin blue (bar 5 jim), shoing severe lss of myehnted nerve fires in anteior root an relativepreervedmyelintednervefiresin posteior root at he same leve (C) Distenson of myehn heaths is cased by postmrtem artefac Motor neuropathy and anti-GM1 ganglioside antibodies 985

ially showed no neuronal loss, neuronal J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from inclusions, or gliosis. The corticospinal tract showed no axonal loss or demyelination or any other significant lesion at all the levels examined, including corona radiata, , mesencephalic peduncle, pontine and bulbar portions, and along the spinal *. - -- 1-ev~- cord. Pure motor peripheral nerves were not examined, but the large mixed sensorimotor sciatic and peroneal nerves showed severe *-2 pathological changes clearly affecting some fascicles more than others (fig 3). There was severe axon depletion with loss of myelin and - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t-. a reactive proliferation of Schwann cells and fibrosis. Immunohistochemistry performed on unfixed frozen samples of the same nerves showed multifocal deposits of IgM (fig 4) and IgG, but not IgA, in relatively well preserved *~ ~~~~osiaTh\erordtract ntr oh fascicles on the nerve fibres.

Figure3 Pathologyofperipheralnerve Fascicular nerve fibres lossthesciaticnerve (immunoperoxidase method2forneurofilament protein, bar = 1s00pm). Discussion The slow progression, lack of brisk reflexes, and absence of bulbar dysfunction were clini- were intact as were the ascending cal features that helped to distinguish our and descending tracts, including the corti- case from amyotrophic lateral sclerosis. cospinal tract. The retrograde nature of the Antibodies to ganglioside GM1 are not usual- observed lesion was supported by the finding ly found in patients with presumably non- of an important loss of anterior root fibres (fig immune lower motor neuron syndromes such 2A, B) accompanied by a severe gliosis in the as .7-9 High titres of intraspinal portion and a proliferation of serum anti-ganglioside GM1 antibodies (>4 Schwann cells and fibrosis in the extraspinal SD above the mean value), are found in some portion. Myelin loss, as shown by myelin motor neuron disease and peripheral motor basic protein (MBP) and myelin associated neuropathy, whereas low titres (<100 in our glycoprotein (MAG) immunohistochemistry, laboratory) are less specific and may be found was not more pronounced than nerve fibre in sera from patients with a variety of neuro- loss. Axonal ballooning was not observed. logical and autoimmune disorders.89 Two The immunohistochemical results for IgG motor syndromes may be considered in our and IgM deposits in the spinal cord were patient: a multifocal motor neuropathy and a inconclusive. lower motor neuron syndrome.9 These syn- All the structures of the and the dromes may overlap with similar clinical and brain were intact; all cranial motor nerve immunological features, but in lower motor nuclei and the espec- neuron syndrome no motor conduction block may be seen on physiological studies,9 and http://jnnp.bmj.com/ these patients may develop widespread weak- Figure 4 (A) Proximal sciatic nervefibres with ness and atrophy suggestive of a spinal mus- multifocal IgM deposits cular atrophy and a progressive course (IgG deposits not shown) leading to death. in endoneurial Recent papers have that a compart,nent (arrows) and underscored around vessels drop in the recorded CMAP amplitude from

(arrowheads). (B) Serial proximal to distal stimulation is insufficient to on October 2, 2021 by guest. Protected copyright. step sectionfrom same recognise conduction block and alternative sample as illustrated in (A) showing no multifocal interpretations should be considered.13 In our staining with a control patient, we did not find abnormal desynchro- incubation. nisation that could result in a reduction of the CMAP. Insufficient stimulus intensity is unlikely because we stimulated Erb's point with different stimulators and differing the position of the patient without modifying the amplitude of the evoked responses. The inter- pretation of these findings is difficult when the usual physiological consequences of demyelination, such as slight conduction delay and prolonged F responses, are absent.'2 We postulate that the proximal con- duction blocks in our patient could be due to the severity of the axonal loss with an increased excitation threshold in the remain- ing motor nerve fibres. The onset of the dis- ease in the legs and the severity of the axonal 986 Adams, Kuntzer, Steck, Lobrinus, Janzer, Regli

loss, with "axonal" proximal conduction antigen might be more concentrated or more J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from blocks on physiological studies make the accessible in motor nerves than in sensory diagnosis of lower motor neuron syndrome nerves: a recent study of gangliosides GM1 in more likely than multifocal motor neuro- peripheral nerves suggested that there were pathy. different lengths of hydrophobic moieties of Conditions involving a primary motor neu- gangliosides GM1 between motor and sen- ropathy with secondary degeneration of sory .'9 Further evidence of interaction motor neurons or a primary motor neuron with the was provided by the disorder with secondary axonal degeneration demonstration that anti-gangliosides GM1 have to be considered in the light of the antibodies bind to nerve terminals.20 In our pathological study. We found a predominant- case, deposits of IgM could be demonstrated ly motor, proximally accentuated, radiculo- in peripheral nerves, but not on the surface of neuropathy with multifocal IgG and IgM motor neurons, as has been previously deposits on the nerve fibres and a slowly demonstrated." These findings additionally ascending retrograde degeneration of the favour the probable autoimmune origin of spinal motor neurons with a corresponding this lower motor neuron syndrome with sec- severe, partly endstage neurogenic muscular ondary neurogenic muscle atrophy and retro- atrophy. grade degeneration of spinal motor neurons. A primary spinal motor neuron disease Further pathological studies are necessary in seems to be unlikely in our case, since the patients with lower motor neuron syndromes severity of the observed changes were more with high titre of anti-gangliosides GM1 anti- accentuated in the anterior roots, followed by bodies to confirm our findings and to further the peripheral nerves and finally by the identify the targets and role of anti-GM1 changes in the anterior horn cells. In favour antibodies. of a primary axonal lesion are the presence of evenly distributed phosphorylated neurofila- Presented in part at the third meeting of the European ments within the perikarya of spinal motor Neurological Society, Lausanne, Switzerland, June 1992. neurons with central chromatolysis, suggest- We are grateful for the excellent technical assistance of Drs JP Baumann and B Bosshard, Institut Neuchatelois d'Anatomie ing impaired axonal transport,'4 and deposits Pathologique, and of Dr J Miklossy, Institut Universitaire de of immunoglobulins on the nerve fibres. We Pathologie, CHUV, Lausanne. cannot provide definitive evidence that the immunoglobulin deposits are at or near the 1 Freddo L, Yu RK, Latov N, et al. Gangliosides GM1 and GDlb are antigens for IgM M-Proteins in a patient with nodes of Ranvier, since myelin stained adja- motor neuron disease. Neurology 1986;36:454-8. cent sections or myelin stains performed on 2 Latov N, Hays AP, Donofrio PD, et al. Monoclonal IgM with unique specificity to gangliosides GM1 and GDlb sections previously stained by immunofluo- and to lactotetraose associated with human motor neu- rescence, or even double immunofluores- ron disease. Neurology 1988;38:763-8. 3 Nardelli E, Steck AJ, Barkas T, et al. Motor neuron syn- cence staining, were not conclusive. What can drome and Monoclonal IgM with antibody activity reasonably be said is that there is multifocal against gangliosides GM1 and GDlb. Ann Neurol 1988;23:524-8. deposition of IgG and IgM on nerve fibres 4 Nobile Orazio E, Legname G, Daverio R, et al. Motor and that this pattern is compatible with the neuron disease in a patient with monoclonal IgM k directed against GM1, GDlb, and high-molecular- localisation at or near the nodes of Ranvier as weight neural specific glycoproteins. Ann Neurol 1990; described in experimental models. We did 28:190-8. 5 Pestronk A, Cornblath DR, Ilyas AA, et al. A treatable not find prominent demyelinative features, multifocal motor neuropathy with antibodies to GMI onion bulbs, or inflammatory cells in periph- ganglioside. Ann Neurol 1988;24:73-8. http://jnnp.bmj.com/ 6 Pestronk A, Chaudhry V, Feldman EL, et al. Lower motor eral nerves. The conclusions regarding the neuron syndromes defined by patterns of weakness, specificity of cytoplasmic inclusions and Ig nerve conduction abnormalities, and high titres of antiglycolipid antibodies. Ann Neurol 1990;27:316-26. deposits in our case must be regarded with 7 Sadiq SA, Thomas FP, Kilidireas K, et al. The spectrum caution in view of the small number of of neurologic disease associated with anti-GMl anti- bodies. Neurology 1990;40:1067-72. necropsies that have been reported in spinal 8 Adams D, Kuntzer T, Burger D, et al. Predictive value of muscular atrophy and pure motor neu- anti-GMl ganglioside antibodies in neuromuscular dis- eases: a study of 180 sera Jf Neuroimmunol 1991;32: ropathies,'4 15 and further investigations of the 223-30. on October 2, 2021 by guest. Protected copyright. role of anti-GM1 antibodies in peripheral 9 Pestronk A. Invited review: Motor neuropathies, motor neuron disorders, and antiglycolipid antibodies. Muscle nerve are needed to elucidate the exact struc- Nerve 1991;14:927-36. ture to which they react. 10 Shy ME, Heiman-Patterson T, Parry GJ, et al. Lower motor neuron disease in a patient with autoantibodies High anti-gangliosides GM1 titres have been against Gal(1-3)GalNAc in gangliosides GMI and shown to be relatively specific for patients with GDlb: improvement following immunotherapy. Neurology 1990;40:842-4. lower motor neuron disease,' 3 6 9 16 motor neu- 11 Santoro M, Thomas FP, Fink ME, et al. IgM deposits at ropathies,567896 or, more rarely, sensorimo- nodes of Ranvier in a patient with amyotrophic lateral sclerosis, anti-GMl antibodies, and multifocal motor tor neuropathy.7'6 Gangliosides are important conduction block. Ann Neurol 1990;28:373-7. components of nerve membrane in the 12 Feasby TE, Brown WF, Gilbert JJ, Hahn AF. The patho- logical basis of conduction block in human neuro- peripheral and central , con- pathies. JNeurol Neurosurg Psychiatry 1985;48:239-44. centrated at the nodes of Ranvier'7 and at 13 Cornblath DR, Summer AJ, Daube J, et al. Conduction block in clinical practice. Muscle Nerve 1991;14:869-71. synaptic terminals.'8 It is interesting to note 14 Sobue G, Hashizume Y, Yasuda T, et al. Phosphorylated that TgM deposits have been localised to the high molecular weight neurofilament protein in lower motor neurons in amyotrophic lateral sclerosis and nodes of Ranvier in a patient presenting with other neurodegenerative diseases involving ventral horn anti-GM1 antibodies and multifocal motor cells. Acta Neuropathol 1990;79:402-8. 15 Hays A. Separation of motor neuron diseases from pure conduction block." It is, however, not clear motor neuropathies. In: Rowland LP, ed. Amyotrophic under which conditions the antibodies gain lateral sclerosis and other motor neuron diseases. Advances in Neurology, 1991;56:385-98. access to their targets and why there is a 16 Lamb NL, Patten BM. Clinical correlations of anti- selective vulnerability of motor nerves.9 The GM1 antibodies in amyotrophic lateral sclerosis and Motor neuropathy and anti-GM1 ganglioside antibodies 987 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.9.982 on 1 September 1993. Downloaded from neuropathies. Muscle Nerve 1991;14:1021-7. 19 Ogawa-Goto K, Funamoto N, Abe T, Nagashima K. 17 Ganser AL, Kirschner DA, Willinger M. Ganglioside Different ceramide composition of gangliosides between localization on myelinated nerve fibers by cholera toxin human motor and sensory nerves. Neurochem binding. Neurocytol 1983;12:921-38. 1990;55: 1686-93. 18 Hansson HA, Holmgren J, Svennerholm L. Ultra- 20 Schluep M, Steck AJ. Immunostaining of motor nerve ter- structural localization of cell membrane GM1 ganglio- minals by IgM M protein with activity against ganglio- side by choleratoxin. Proc Natl Acad Sci USA 1977; sides GM1 and GDlb from a patient with motor 74:3782-6. neuron disease. Neurology 1988;38: 1890-2. http://jnnp.bmj.com/ on October 2, 2021 by guest. Protected copyright.