Journal of the Peripheral Nervous System 7:54–58 (2002)

© 2002 Peripheral Nerve Society, Inc.

54

Anti-LM1 antibodies in the sera of patients withGuillain-Barré syndrome, Miller Fisher syndrome,

and Motor Neuron Disease

Hajime Harukawa,

1

Hiroya Utsumi,

1

Atsuko Asano,

2

and Hiide Yoshino

2

The Third Department of Internal Medicine

1

, Tokyo Medical University, Tokyo and Department of Neurology

2

, Kohnodai Hospital, National Center of Neurology and Psychiatry, Ichikawa City, Chiba, Japan

Abstract

This study is designed to establish whether sialosylneolactotetraosylceram-ide (LM1), a major component of human peripheral nerve ganglioside, is a potential targetantigen for the development of peripheral autoimmune neuropathies such as Guillain-Barré syndrome (GBS) and Miller Fisher syndrome (MFS). Serum antibodies against LM1in 116 patients with GBS, 56 patients with MFS, 88 patients with motor neuron disease(MND) and 60 normal control subjects were quantified using enzyme-linked immunosor-bent assay (ELISA). The presence of anti-LM1 antibodies were confirmed using an immu-nostaining method on high-performance thin-layer chromatographic plates (HPTLC). Anti-LM1 IgG antibodies were detected in 22% (25/116) of patients with GBS. The ratio of thedemyelination type to the axonal type of GBS was approximately 3:1. Among the 25 anti-LM1-positive GBS patients, additional anti-GM1 IgG antibodies were detected in 7 patients,4 of whom possessed the axonal form of GBS. Anti-LM1 antibodies were also detected in asignificant portion of patients with MFS (20%, 11/56). In contrast, anti-LM1 antibodies weredetected in only 2% (2/88) of patients with MND, and 7% (4/60) of normal control subjects.The results of this study suggest that serum antibodies against LM1 may have a patho-genic role in the development of GBS and MFS.

Key words:

LM1, Guillain-Barré syndrome, Miller Fisher syndrome, MND

Introduction

Although the etiology and pathogenesis of the Guil-lain-Barré syndrome (GBS) and Miller Fisher syndrome(MFS) remain poorly defined, the immunological mech-anisms may play important roles in these diseases.Gangliosides are important constituents of the plasmamembrane of cells and are particularly abundant in thenervous system. Recently, studies have detected se-rum antibodies against gangliosides such as GM1 andGQ1b in patients with GBS and MFS. Therefore, theseanti-ganglioside antibodies are important diagnosticmarkers of patients with GBS and MFS

(Chiba et al.,

1992; 1993; Ilyas et al., 1988; Kusunoki et al., 1994; Yukiet al., 1990)

. These gangliosides are also expressed inthe peripheral nervous system (PNS), but are enriched inthe central nervous system (CNS)

(Yu and Saito, 1989)

.On the other hand, serum anti-ganglioside antibody

against LM1 (sialosylneolactotetraosylceramide), thepredominant ganglioside in the human peripheral nervemyelin

(Ogawa-Goto et al., 1992)

, is an important mole-cule in the pathogenesis of immunologic diseases ofthe peripheral nerves. However, until now, there havebeen few studies investigating anti-LM1 antibodies thatincluded adequate numbers of normal control and dis-ease control subjects. Therefore, the significance ofanti-LM1 antibodies in GBS or MFS has not been estab-lished. In this study, we investigated the presence ofthe anti-LM1 antibodies in the sera of patients withGBS, MFS, motor neuron disease (MND), and normalsubjects using an enzyme-linked immunosorbent assay

Address correspondence to:

Dr. Hiide Yoshino, Department of Neu-

rology, Kohnodai Hospital, National Center of Neurology and Psy-

chiatry, 1-7-1 Kohnodai, Ichikawa-city, Chiba-prefecture, 272-8516,

Japan. Tel.: +81-47-372-3501; Fax: +81-47-372-1858; E-mail: yoshinos@lily.ocn.ne.jp

Harukawa et al. Journal of the Peripheral Nervous System 7:54–58 (2002)

55

(ELISA). The presence of anti-LM1 antibodies wereconfirmed using an immunostaining method on high-performance thin-layer chromatographic plates (HPTLC).The possible role of anti-LM1 antibodies in the patho-genesis of GBS and MFS is discussed.

Materials and Methods

Patients’ sera

Serum samples were collected from 116 patientswith GBS, 56 with MFS, 88 with MND, and 60 normalcontrol subjects. These samples were sent to KohnodaiHospital to test for the presence of serum gangliosideantibodies. The diagnosis of GBS was defined clinicallyaccording to the criteria established by

Asbury and Corn-blath (1990)

.

Purification of LM1 and brain gangliosides

LM1 was prepared from the plasma membrane ofhuman red blood cells

(Li et al., 1973)

, purified on aDEAE-Sephadex A25 column (Sigma, St. Louis, MO,USA), and applied to an Iatrobeads column (Iatron, To-kyo, Japan). A standard ganglioside mixture was ob-tained from bovine brain (Shibaura Animal Organ Sup-ply, Tokyo, Japan) gangliosides and Tay-Sachs diseasedbrain gangliosides.

Detection of anti-LM1 antibodies by ELISA method

ELISA was performed as described previously(

Yoshino et al., 1992

). Briefly, 50 nanograms of purifiedLM1 were coated to wells of microtiter plates (Nunc,Roskilde, Denmark). After the non-specific binding siteswere blocked with 1% bovine serum albumin (BSA)-phosphate-buffered saline (PBS), each well was over-laid with 1:100 dilution of sera overnight at 4

�

C. Theplate was washed with PBS and counterstained withperoxidase-conjugated goat antibodies to either humanIgG (gamma-chain specific) or human IgM (gamma-chain specific) (Dako, Tokyo, Japan) for 1 hour at roomtemperature. Bound peroxidase-conjugated antibodieswere visualized with 0.05% 4-chloro-1-naphtol contain-ing 0.01% H

2

O

2

, and the absorbance values at 492 nmwere determined by subtracting the optical density(OD) of wells treated identically, but without bound LM1.

Detection of anti-LM1 antibodies by the immunostaining method on HPTLC plate

Serum samples positive for anti-LM1 antibody byELISA (the OD value

�

0.2) were subject to the detec-tion of anti-LM1 antibody with the TLC-immunostainingmethod. We determined the cutoff to be 0.2 absor-bance unit, because we can usually demonstrate anti-ganglioside antibodies on TLC-immunostaining methodwhen the absorbance shows more than 0.2. Standardgangliosides, human peripheral nerve gangliosides, and

LM1 were developed on HPTLC plates (Merck, Darm-stadt, Germany) using a solvent system of chloroform/methanol/water containing 0.2%CaCl

2

•2H

2

O (50/45/10,by volume, solvent A)

(Ando and Yu, 1977)

. Immun-ostaining on the HPTLC plate was performed as de-scribed previously

(Matsumoto et al., 1995)

, except thatperoxidase-conjugated gamma-chain-specific anti-humanIgG (1:2000; Dako, Tokyo, Japan) was used as the sec-ondary antibody. Anti-LM1 antibodies were detected onx-ray film using ECL Western blotting detection re-agents (Amersham, Buckinghamshire, UK).

Statistical analyses were performed using the Chi-square test. A p value

�

0.05 was considered to be sta-tistically significant.

Results

Serum antibodies against LM1

Elevated levels of IgG antibodies against LM1 weredetected in 25 of 116 (22%) GBS patients, 11 of 56(20%) MFS patients, 2 of 88 (2%) MND patients, and 4of 60 (7%) normal control subjects (Table 1). The occur-rence of IgG anti-LM1 in GBS and MFS is significantlymore frequent than that in the normal population (GBS:p = 0.01; MFS: p = 0.04) or MND patients (GBS: p

�

0.01; MFS: p

�

0.01). Elevated levels of IgM antibodiesto LM1 were detected in only 2 patients with GBS.These 2 patients also possessed IgG antibodies to LM1.

Further analysis of the anti-LM1 antibody-positive GBS patients

IgG anti-GM1 antibodies were detected in 7 of the25 patients with GBS that possessed detectable LM1antibodies (Fig. 1, Panel B). In the 25 patients with GBSwho had anti-LM1 antibodies, the electrophysiologicalfindings suggested the demyelinating type of GBS in 16cases, and the axonal type of GBS in 5 cases. The re-maining 4 cases were either unclassified or the perti-nent data were not available. Four of the 5 patients withthe axonal type of GBS had additional anti-GM1 anti-bodies.

Table 1.

Anti-LM1 antibodies in the sera of patients with GBS, MFS, MND and normal control

Anti-LM1 antibody p value*

PatientPositive (n = 42)

Negative (n = 278)

versus MND

versus NC

GBS 25 91

�

0.01 =0.01MFS 11 45

�

0.01 =0.04MND 2 86 — —NC 4 56 — —

GBS = Guillain-Barré syndrome; MFS = Miller-Fisher syndrome; MND = motor neuron disease; NC = normal control*Chi-square test

Harukawa et al. Journal of the Peripheral Nervous System 7:54–58 (2002)

56

Further analysis of the anti-LM1 antibody-positiveMFS patients

Of the 11 MFS patients who had anti-LM1 antibod-ies, anti-GQ1b antibodies were detected in 9 (Fig. 1,Panels C and D). One of the 2 patients without detect-able anti-GQ1b antibody activity demonstrated a motornerve conduction delay in the electrophysiological study.

Further analysis of the anti-LM1-positiveMND patients

Anti-LM1 antibodies were detected in only 2 of theMND patients. One of 2 anti-LM1-positive MND pa-tients revealed an unusually slow course of the illness.Eight years after onset, this patient received a tracheot-omy. From 2 years later until the present, the patientdid not need a respirator in the daytime. In this case,the nerve conduction velocity was found to be normal,and the protein in the CSF remained elevated (108 mg/dLin March 2000).

Discussion

In this study, anti-LM1 antibodies were more fre-quently detected among the patients with GBS andMFS than in normal control subjects or MND patients.Of the various gangliosides, LM1 is specifically distrib-uted in the myelin of peripheral nerves

(Ogawa-Goto etal., 1992)

and cranial nerves

(Chiba et al., 1997)

. There-fore, LM1 may be a specific target antigen in the patho-genesis of GBS and MFS. In our study, anti-LM1 anti-bodies were detected in 22% of patients with GBS.

However,

Yuki et al. (1996)

and

Yako et al. (1999)

re-ported that anti-LM1 antibodies were detected in only8% (8/96) and 5% (7/140) of the patients with GBS, re-spectively. On the other hand,

Fredman et al. (1991)

re-ported that antibodies to LM1 were found in 43% (10/23)of patients with GBS, 67% (10/15) of patients withCIDP, and 20% (8/40) of normal subjects.

Ilyas et al.(1992)

reported that 23% (11/53) of GBS patients and2.3% (2/88) of the combined disease controls and nor-mal subjects had IgG anti-LM1 antibodies, but no signif-icant elevation of IgM anti-LM1 antibody titers was de-tected. Moreover,

Svennerholm and Fredman (1990)

detected anti-LM1 antibodies in 58% (29/50) of pa-tients with GBS, 30% (6/20) of healthy controls, and14%(28/197) of patients diagnosed with other neuro-logical diseases. These differences may be due to dif-ferences in the cutoff points for the determination ofdetectable antibody titers. In this study, we were usu-ally able to demonstrate anti-ganglioside antibodies bythe TLC-immunostaining method when the OD valuewas above 0.2 in the ELISA.

Among various anti-ganglioside antibodies ob-served in GBS, GM1 and GalNAc-GD1a antibodies arefrequently observed in GBS with predominantly motoraxonopathy

(Yuki et al., 1990; Kusunoki et al., 1994)

,whereas anti-GQ1b is observed in MFS, or GBS withophthalmoplegia

(Chiba et al., 1992; 1993)

. In addition,GT1a and GM1b might be associated with GBS withbulbar palsy

(Yoshino et al., 2000)

. Therefore, each anti-body species may be involved in the clinical manifesta-tion of GBS. So far, there is only limited informationconcerning the clinical manifestation in GBS patientswho have anti-LM1 antibodies.

Yako et al.

(1999)

re-ported that 5 of 7 patients with anti-LM1 antibodies intheir study showed demyelination electrophysiologi-cally, and only 1 patient showed axonopathy mixed withdemyelination. In our study, the clinical manifestationof patients with GBS who possessed detectable levelsof anti-LM1 antibodies is not uniform, in that both thedemyelinating and axonal forms of GBS were present.The ratio of the demyelination form versus the axonalform was approximately 3:1. Among the 5 GBS pa-tients with the axonal form who possessed anti-LM1antibodies, 4 also had anti-GM1 antibodies. Axonaldamage, therefore, may be more closely associatedwith the presence of anti-GM1 antibodies rather thanthe presence of anti-LM1 antibodies. This result sup-ports previous findings that anti-GM1 antibodies are animportant marker for the development of the axonalform of GBS

(Ho et al., 1995; Kornberg et al., 1994; Vis-ser et al., 1995; Yuki et al., 1990)

.Current studies suggest that anti-GQ1b antibodies

are detected very frequently in patients with MFS, andthat the presence of anti-GQ1b antibodies is a usefulmarker in the diagnosis of MFS

(Chiba et al., 1992; Will-

Figure 1. Results of immunostaining HPTLC plates with thesera of patients with GBS, MFS and MND.Standard ganglioside mixture (lane 1) and purified LM1 (lane2) were developed on a thin-layer chromatographic plate.The plate was visualized by spraying with the resorcinol-HClreagent (A) or immunostained with serum from a patientwith GBS (Panel B), and with sera from 2 patients with MFS(Panels C and D). The serum from the patient with GBS re-acted with GM1 and GalNAc-GD1a additionally (Panel B,lane 1). The serum of 1 patient with MFS reacted with GQ1band GT1a additionally (Panel C, lane 1), whereas the otherpatient with MFS did not react (Panel D) with GQ1b.

Harukawa et al. Journal of the Peripheral Nervous System 7:54–58 (2002)

57

ison et al., 1993; Willison and Veitch, 1994; Yuki et al.,1993a)

. Anti-GQ1b antibodies, however, were not nec-essarily detected in all of MFS cases. In this study, anti-LM1 antibodies were detected in 11 of 56 patients withMFS, and 2 of the 11 patients did not possess detect-able anti-GQ1b antibodies. Whether the central or pe-ripheral nervous system is originally damaged in MFSpatients with ataxia or sensory disturbances remainsunclear. However, because this study found that anti-LM1 antibodies were detected in 20% of patients withMFS, peripheral nerve lesions may account for the de-velopment of ataxia and sensory disturbances in somepatients with MFS.

Recently, several studies reported that anti-ganglio-side antibodies are detected in some MND or motorneuropathies

(Pestronk, 1991)

such as multifocal motorneuropathy (MMN)

(Pestronk et al., 1988)

. Yuki et al.(

1996)

reported that anti-LM1 antibody is detected in10% (5/50) of ALS patients.

Yuki et al. (1993b)

also re-ported a single case of proximal lower MND with de-tectable anti-LM1 antibodies. In our study, 1 of 2 MNDpatients with anti-LM1 antibodies revealed an unusualcourse of the illness. However, we could not confirman association between the presence of anti-LM1 anti-bodies and MND because anti-LM1 antibodies werefound only in 2% of MND and 7% of normal controlsubjects.

From the results reported here, we conclude thatanti-LM1 antibodies are more frequently associatedwith patients with GBS and MFS than normal controlsubjects or patients with MND. Therefore, the pres-ence of anti-LM1 antibodies may have a pathogenicrole in the development of GBS and MFS, and testingfor the presence of anti-LM1 antibody would help diag-nose these diseases.

Acknowledgements

This work was supported in part by the Imai KimiMemorial Foundation for Neuromuscular Diseases, theFoundation of Japan ALS Association, a research granton neuroimmunological diseases from the Ministry ofHealth and Welfare, and a research grant on ALS fromthe Ministry of Health and Welfare, Japan.

References

Ando S, Yu RK (1977). Isolation and characterization of a noveltrisialoganglioside, GT1a, from human brain. J Biol Chem252:6247–6250.

Asbury AK, Cornblath DK (1990). Assessment of current diag-nostic criteria for Guillain-Barré syndrome. Ann Neurol 27(Suppl):21–24.

Chiba A, Kusunoki S, Shimizu T, Kanazawa I (1992). Serum IgGantibody to ganglioside GQ1b is a possible marker of MillerFisher syndrome. Ann Neurol 31:677–679.

Chiba A, Kusunoki S, Obata H, Machinami R, Kanazawa I (1993).Serum anti-GQ1b IgG antibody is associated with ophthal-moplegia in Miller Fisher syndrome and Guillain-Barré syn-drome: clinical and immunohistochemical study. Neurology43:1911–1917.

Chiba A, Kusunoki S, Obata H, Machinami R, Kanazawa I (1997).Ganglioside composition of the human cranial nerves, withspecial reference to pathophysiology of Miller Fisher syn-drome. Brain Res 745:32–36.

Fredman P, Vedeler CA, Nyland H, Aarli JA, Svennerholm L(1991). Antibodies in sera from patients with inflammatorydemyelinating polyneuropathy react with ganglioside LM1and sulphatide of peripheral nerve myelin. J Neurol 238:75–79.

Ho TW, Mishu B, Li CY, Gao CY, Cornblath DR, Griffin JW, As-bury AK, Blaser MK, McKhann GM (1995). Guillain-Barrésyndrome in northern China. Relationship to Campylobacterjejuni infection and anti-glycolipid antibodies. Brain 118:597–605.

Ilyas AA, Willison HJ, Quarles RH, Jungalwala FB, CornblathDR, Trapp BD, Griffin DE, Griffin JW, McKhann GM (1988).Serum antibodies to gangliosides in Guillain-Barré syn-drome. Ann Neurol 23:440–447.

Ilyas AA, Mithen FA, Dalakas MC, Chen ZW, Cook SD (1992).Antibodies to acidic glycolipids in Guillain-Barré syndrome andchronic inflammatory demyelinating polyneuropathy. J NeurolSci 107:111–121.

Kornberg AJ, Pestronk A, Bieser K, Ho TW, McKhann GM, WuHS, Jiang Z (1994). The clinical correlates of high-titer IgGanti-GM1 antibodies. Ann Neurol 35:234–237.

Kusunoki S, Chiba A, Kon K, Ando S, Arisawa K, Tate A, Kana-zawa I (1994). N-acetylgalactosaminyl GD1a is a target mole-cule for serum antibody in Guillain-Barré syndrome. Ann Neurol35:570–576.

Li Y-T, Månsson J-E, Vanier M-T, Svennerholm L (1973). Struc-ture of the major glucosamine containing ganglioside of hu-man tissues. J Biol Chem 248:2634–2636.

Matsumoto A, Yoshino H, Yuki N, Hara Y, Cashman NR, HandaS, Miyatake T (1995). Ganglioside characterization of a cellline displaying motor neuron-like phenotype: GM2 as a pos-sible major ganglioside in motor neurons. J Neurol Sci 131:111–118.

Ogawa-Goto K, Funamoto N, Ohta Y, Abe T, Nagashima K(1992). Myelin gangliosides of human peripheral nervoussystem; an enrichment of GM1 in the motor nerve myelinisolated from Cauda equina. J Neurochem 59:1844–1849.

Pestronk A, Cornblath DR, Ilyas AA, Baba H, Quarles RH, GriffinJW, Alderson K, Adams RN (1988). A treatable multifocalmotor neuropathy with antibodies to GM1 gangliosides. AnnNeurol 24:73–78.

Pestronk A (1991). Invited review: motor neuropathies, motorneuron disorders, and antiglycolipid antibodies. MuscleNerve 14:927–936.

Svennerholm L, Fredman P (1990). Antibody detection in Guil-lain-Barré syndrome. Ann Neurol 27(Suppl):36–40.

Visser LH, van der Meché FG, van Doorn PA, Meulstee J, Ja-cobs BC, Oomes PG, Kleyweg RP, Meulstee J (1995). Guil-lain-Barré syndrome without sensory loss (acute motor neu-ropathy). A subgroup with specific clinical, electrodiagnosticand laboratory features. Dutch Guillain-Barré Study Group.Brain 118:841–847.

Willison HJ, Veitch J, Paterson G, Kennedy PG (1993). Miller

Harukawa et al. Journal of the Peripheral Nervous System 7:54–58 (2002)

58

Fisher syndrome is associated with serum antibodies to GQ1bganglioside. J Neurol Neurosurg Psychiatry 56:204–206.

Willison HJ, Veitch J (1994). Immunoglobulin subclass distribu-tion and binding characteristics of anti-GQ1b antibodies inMiller Fisher syndrome. J Neuroimmunol 50:159–165.

Yako K, Kusunoki S, Kanazawa I (1999). Serum antibody againsta peripheral nerve myelin ganglioside, LM1, in Guillain-Barrésyndrome. J Neurol Sci 15:85–89.

Yoshino H, Inuzuka T, Miyatake T (1992). IgG antibody againstGM1, GD1b and asialo-GM1 in chronic polyneuropathy fol-lowing Mycoplasma pneumoniae infection. Eur Neurol 32:28–31.

Yoshino H, Harukawa H, Asano A (2000). IgG antiganglioside anti-bodies in Guillain-Barré syndrome with bulbar palsy. J Neuro-immunol 105:195–201.

Yu RK, Saito M (1989). Structure and Localization of Ganglio-

sides. In: Neurobiology of Glycoconjugates. Margolis RU,Margolis RK (Eds). Plenum Press, New York, pp 1–42.

Yuki N, Yoshino H, Sato S, Miyatake T (1990). Acute axonalpolyneuropathy associated with anti-GM1 antibodies follow-ing Campylobacter enteritis. Neurology 40:1900–1902.

Yuki N, Sato S, Tsuji S, Ohsawa T, Miyatake T (1993a). Frequentpresence of anti-GQ1b antibody in Fisher's syndrome. Neu-rology 43:414–417.

Yuki N, Yanagisawa K, Saito Y, Aikawa M, Miyatake T (1993b).Proximal lower motor neuron syndrome associated with se-rum antibodies to asialo-GM1, GM1 and LM1. Clin Neurol(Tokyo) 33:891–894.

Yuki N, Tagawa Y, Handa S (1996). Autoantibodies to peripheralnerve glycosphingolipids SPG, SLPG, and SGPG in Guillain-Barré syndrome and chronic inflammatory demyelinatingpolyneuropathy. J Neuroimmunol 70:1–6.