Pathogenic roles of antiganglioside antibodies in immune-mediated neuropathies

Authors


Correspondence

Kenichi Kaida, MD, PhD, Division of Neurology, Department of Internal Medicine 3, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan.

Tel: +81-42-2995-1617

Fax: +81-42-2996-5202

Email: adiak901@ndmc.ac.jp

Abstract

Recent studies have shown that antiganglioside antibody-mediated complement activation plays a key role in development of Guillain–Barré syndrome and related disorders. Additionally, complement-independent nerve dysfunction is suggested by in vitro studies, showing that antiganglioside antibodies directly inhibit voltage-gated Ca channel currents and change the integrity of lipid rafts. These pathogenic actions of antiganglioside antibodies might be governed by the avidity of the antibodies, which is influenced by specific localization of target gangliosides in the peripheral nervous system or glycolipid environment around the target antigens. The recent discovery of antibodies to ganglioside complexes has expedited the understanding of the mechanisms underlying antiganglioside antibody-mediated nerve dysfunction in Guillain–Barré syndrome and related disorders. In chronic immune-mediated neuropathy, some antiganglioside antibodies have also been identified as diagnostic markers, although their pathophysiological roles remain to be determined.

Introduction

Gangliosides are N-acetylneuraminic acid (sialic acid)-containing glycosphingolipids that aggregate and form clusters on the surfaces of neuronal membranes with display of oligosaccharides on the cell surface.[1] Gangliosides reside in membrane microdomains referred to as lipid rafts or detergent-resistant membranes, together with other sphingolipids, cholesterol and glycosylphosphatidylinositol (GPI)-anchored proteins.[2] These microdomains form platforms and facilitate a variety of membrane-mediated functions, including signal transduction.

Gangliosides localized in the peripheral nervous system (PNS) are targeted by serum antibodies in approximately 60% of patients with acute immune-mediated polyradiculoneuropathy, Guillain–Barré syndrome (GBS).[3] Antiganglioside antibodies might function in the pathogenesis of GBS through antibody-antigen interactions in the PNS.[3] There has been considerable progress on the pathophysiological mechanisms of antiganglioside antibody-associated nerve dysfunction in GBS. Studies in the past 10 years have shown that antiganglioside antibody-mediated nerve damage is likely to occur through complement activation, although some in vitro studies show that some of the nerve damage is independent of complement activation. The discovery of antibodies to ganglioside complexes (GSC) consisting of two different gangliosides has also allowed a better understanding of the pathophysiological action of antiganglioside antibodies in GBS. In the present article, we present an updated overview of the pathogenetic roles of antiganglioside antibodies in immune-mediated neuropathy, and especially in GBS.

Antiganglioside antibodies in Guillain–Barré syndrome and related disorders

Antibodies to glycolipids including gangliosides are present in up to 60% of cases of GBS and 90% of cases of Fisher syndrome (FS), a GBS variant characterized by ophthalmoplegia, ataxia and areflexia.[3, 4] In a recent collaborative study in Japan and Italy, 83% of patients with acute motor axonal neuropathy (AMAN) had immunoglobulin G (IgG) antibodies to GM1, GD1a, N-acetylgalactosaminyl GD1a (GalNAc-GD1a) or GM1b.[5] In acute inflammatory demyelinating polyneuropathy (AIDP), which is the prevalent form of GBS in Western countries, there is no definitive association with antiglycolipid antibodies, although antibodies to galactocerebroside (Gal-C), LM1 or GD1b are present in some AIDP patients.[6-8] A recent study showed that five of 40 patients (12.5%) with GBS had IgG anti-LM1 antibodies, including four patients classified as having AIDP.[9] Whether LM1 localized in peripheral nerve myelin is attacked by autoantibodies remains to be confirmed in vivo.

Heterogeneity of ganglioside expression in the PNS is likely to influence the symptomatology of GBS and related disorders, and distribution of gangliosides is crucial for regulation of the clinical phenotype of GBS.[10, 11] IgG anti-GQ1b antibodies are pathophysiologically associated with ophthalmoplegia in FS and GBS,[4] and also with ataxia.[12] GD1b-specific IgG antibodies without cross-reaction to GM1 and other gangliosides induce ataxia and limb weakness in GBS,[8, 13] and an IgG antibody to a minor ganglioside, GalNAc-GD1a, is associated with AMAN.[14, 15] Immunohistochemical studies in the human PNS show a correlation between localization of gangliosides and the clinical phenotype. Thus, GQ1b is densely distributed in paranodal myelin of cranial nerves innervating extraocular muscles,[4] a subpopulation of large neurons in dorsal root ganglia,[12] and in nerve terminals inside muscle spindles and in the vicinity of intrafusal fibers;[16] GD1b is localized on dorsal root ganglion (DRG) neurons with a large diameter and paranodal myelin;[17, 18] and GalNAc-GD1a is distributed in the vicinity of the nodes of Ranvier, especially in the nodal and paranodal axolemma in motor nerves.[19]

IgG subclasses of antiganglioside antibodies might contribute to clinical features of GBS and FS. The IgG subclass of anti-GM1 antibodies in GBS and anti-GQ1b antibodies in FS are IgG1 or IgG3.[20, 21] IgG3 is the major subclass of anti-GQ1b antibodies in patients with FS after respiratory infection.[22] A study of the IgG subclass of anti-GM1 antibodies showed that the IgG1 subclass is associated with slow recovery and IgG3 with rapid recovery.[23] In GBS patients with IgG antibodies to GM1, GM1b, GD1a or GalNAc-GD1a, the presence of only IgG1 antibodies is related to diarrhea, positive Campylobacter serology, cross-reactive antibodies to C. jejuni lipo-oligosaccharides (LOS) and a poor outcome; whereas the presence of both IgG1 and IgG3 antibodies is related to upper respiratory infections, cross-reactive antibodies to Haemophilus influenzae LOS and a better outcome.[24] These findings suggest that the type of infectious agent precipitating GBS or FS governs variation of the IgG subclass through a mechanism that is unclear.

Antibodies to ganglioside complexes in GBS and its variants

Screening of antiganglioside antibodies has generally been carried out for purified single ganglioside antigens with enzyme-linked immunosorbent assay (ELISA) or thin-layer chromatogram-immunostaining. Recently, we detected antibodies to GSC that comprised of two different gangliosides in sera from patients with GBS or FS.[25] Some anti-GSC antibodies have hitherto been reported in GBS and related disorders (Table 1). When the ratio of two gangliosides is 1 : 1 (w/w), the activity of antibodies toward the GSC is optimal. Anti-GSC antibodies have little or no reactivity with each constituent ganglioside, indicating that novel glycoepitopes formed in the GSC are target molecules in antibody-mediated events. Since the discovery of anti-GSC antibodies in GBS and related disorders, ELISA for antiganglioside antibody-screening has been carried out in a grid manner with horizontal and vertical mixing lines.

Table 1. Immunoglobulin G anti-ganglioside complex antibodies and the associated clinical features
GSC antigensDisorders (frequency)aAssociated clinical features
  1. AIDP, acute inflammatory demyelinating polyneuropathy; AMCBN, acute motor conduction block neuropathy; BBE, Bickerstaff brainstem encephalitis; FS, Fisher syndrome; GBS, Guillain–Barré syndrome; GSC, ganglioside complex; OP, ophthalmoplegia.

  2. a

    “Frequency” indicates frequency of anti-ganglioside complex (GSC) antibodies in the disorder.

GM1/GD1aGBS (17%) 
GD1a/GD1bGBS (7%)Severe disability, artificial ventilation, impairment of lower cranial nerves
GD1b/GT1bGBS (6%)Severe disability, artificial ventilation, impairment of lower cranial nerves
GM1/GalNAc-GD1aGBS (3–5%)Pure motor, AMCBN
GM1/GalNAc-GD1aGBS (3–5%)Pure motor, AMCBN
GM1/GQ1b, GM1/GT1a, GD1b/GQ1b, GD1b/GT1aFS (41%), GBS with OP (28%)OP, infrequent sensory dysfunction
GD1a/GQ1b, GD1a/GT1a, GT1b/GQ1b, GT1b/GT1aFS (6%), GBS with OP (19%)OP
GA1/GQ1b, GA1/GT1aFS, GBS, BBEOP
GM1/LM1GBS (7.5%)AIDP?

Pathogenic roles of anti-GSC antibodies in GBS and related disorders might be similar to those of antibodies to single gangliosides, which are not well understood. In a survey of antibodies to GSC consisting of two of the four major gangliosides (GM1, GD1a, GD1b and GT1b) in 234 GBS patients, 39 (17%) had IgG anti-GSC antibodies, and the antibodies to GSC as GD1a/GD1b or GD1b/GT1b were closely associated with lower cranial nerve deficits and severe disability requiring artificial ventilation.[26] An immunoabsorption study of anti-GSC antibodies showed that the antibody activity to GD1a/GD1b was completely absorbed by GSC antigens, such as GM1/GD1a, GD1b/GT1b and GM1/GT1b, but only slightly by single constituent gangliosides, indicating that an epitope formed by a combination of [Galβ1-3GalNAc] and [NeuAcα2-3Galβ1-3GalNAc] in the terminal moieties of ganglio-N-tetraose structures is required for antibody binding (Fig. 1). Why antibodies to GSC, such as GD1a/GD1b, are associated with severe disability is unclear, but glycoepitopes in GSC are likely to be more multivalent than in single gangliosides, and interactions between anti-GSC antibodies and GSC antigens might be tighter, which might precipitate stronger antibody-mediated immunoreaction and more severe nerve damage.

Figure 1.

Pattern diagrams of glycolipid complexes GM1/GD1a, GD1a/GD1b, GM1/GT1b and GD1b/GT1b. Squares with dotted lines indicate putative antigenic epitopes for antiganglioside complex antibodies.

GM1 and GalNAc-GD1a are target gangliosides in AMAN, and are distributed on axolemma at the nodes of Ranvier in motor nerves.[19, 27, 28] Recently, an antibody to GM1/GalNAc-GD1a, a GSC consisting of GM1 and GalNAc-GD1a, was found in approximately 5% of GBS patients, and anti-GM1/GalNAc-GD1a antibody-positive patients had the pure motor variant of GBS.[29, 30] Electrophysiological findings in these patients showed early conduction block at intermediate nerve segments of motor nerves.[30] Serial electrophysiological studies showed rapid recovery of the conduction block and no findings indicative of remyelination and axonal degeneration, suggesting that the conduction block results from impairment of axonal membrane properties at the nodes of Ranvier, where voltage-gated sodium channels (Nav) are densely clustered.[31] A study of axonal excitability in AMAN with IgG antibodies to GM1, GM1b or GalNAc-GD1a showed rapid normalization of extreme refractoriness related to improvement of compound muscle action potentials,[32] suggesting that dysfunction of Nav at the nodes of Ranvier is a primary cause of reversible conduction failure in AMAN. Temporary blockade of Nav can provoke conduction block with rapid normalization within days, as often observed in poisoning by Nav-blocking toxins, saxitoxin and tetrodotoxin.[33, 34] GM1 and GalNAc-GD1a are located on nodal axolemma in motor nerves, where they can form a GSC, GM1/GalNAc-GD1a. The binding of anti-GM1/GalNAc-GD1a antibodies to target antigens on the nodal axolemma might provoke direct or indirect alteration of regulatory functions of Nav, leading to conduction block in anti-GM1/GalNAc-GD1a-positive GBS patients.

IgG antibodies to GSC containing GQ1b or GT1a are detected in half of patients with FS with ganglioside GQ1b as the prime antigen.[35] Anti-GM1/GQ1b-positive sera react with GD1b/GQ1b, GM1/GT1a or GD1b/GT1a, and anti-GD1a/GQ1b-positive sera react with GD1a/GT1a or GT1b/GQ1b. Such a pattern of anti-GSC antibodies shows that anti-GQ1b/GM1-reactive sera react with a combination of [Galβ1-3GalNAc] and [NeuAcα2-8 NeuAcα2-3Galβ1-3GalNAc] in the terminal residues of ganglio-N-tetraose structures, and that anti-GQ1b/GD1a-reactive sera react with a combination of [NeuAcα2-3Galβ1-3GalNAc] and [NeuAcα2-8 NeuAcα2-3Galβ1-3GalNAc] in the terminal residues (Fig. 2).[35, 36] Therefore, antiganglioside antibodies in FS are subdivided into GQ1b-specific, GQ1b/GM1-reactive or GQ1b/GD1a-reactive antibodies, regulated by the conformation of terminal residues containing sialic acids. IgG antibodies to GSC containing GQ1b or GT1a are also found in some GBS patients and are associated with development of ophthalmoplegia in GBS.[36] GQ1b-specific, GQ1b/GM1-reactive or GQ1b/GD1a-reactive antibodies are likely to react with the same glycoepitopes on the nerve membrane, especially in nerves innervating extraocular muscles.

Figure 2.

Pattern diagrams of glycolipid complexes GM1/GQ1b, GD1a/GQ1b, GA1/GQ1b and GA1/GT1a. Squares with dotted lines indicate putative antigenic epitopes for antiganglioside complex antibodies.

We recently found that some FS or GBS patients with ophthalmoplegia had IgG antibodies to a glycolipid complex consisting of GA1 and GQ1b, or GA1 and GT1a.[37] Of the antibodies to GA1/GQ1b or GA1/GT1a, 70% did not react with GM1/GQ1b and GD1b/GQ1b, although the terminal residues of GA1/GQ1b are similar to those of GM1/GQ1b or GD1b/GQ1b (Fig. 2). Consequently, the specificity of antibodies to GSC containing GQ1b or GT1a might be governed by sialic acids attached to an internal galactose, as well as by those on terminal residues. Analyses of the conformation of glycoepitopes in the GSC are required for identification of the exact target antigens in GBS and its variants.

Pathophysiological action of antiganglioside antibodies in GBS and related disorders

Complement activation on the nerve cell membrane plays a key role in the process of nerve injury in GBS and related disorders, especially with antiganglioside antibodies,[38-43] as shown in in vitro and ex vivo studies.[11, 43] In experimental models of GBS or FS, formation of a membrane attack complex (MAC) has been shown at motor nerve terminals with bound antiganglioside antibodies.[44, 45] In in vitro and in vivo studies using C6-deficient mice and sera, MAC formation was not induced at nerve terminals by binding of antiganglioside antibodies. In CD59-deficient (CD59−/−) mice, which cannot inhibit MAC formation, deposits of MAC and damage to perisynaptic Schwann cells and neurofilaments at nerve terminals were greater than in CD59+/+ mice.[44] In an experiment in Ca+-free Ringers solution, antiganglioside antibody-mediated complement activation was shown to be induced mainly through the Ca2+-dependent, classical pathway.[44]

In an anti-GM1-positive rabbit model of AMAN immunized with a bovine brain ganglioside mixture including GM1, complement-mediated impairment of paranodal axoglial junctions, nodal cytoskeleton and Schwann cell microvilli are observed with destruction of clusters of Nav, with gradual improvement in the late recovery phase.[46] In an anti-GD1b antibody-positive rabbit model of acute sensory ataxic neuropathy (ASAN) immunized with bovine GD1b, immunohistochemical analyses showed deposition of IgG, C3 and MAC at the nodes of Ranvier in dorsal roots, where clusters of nodal and paranodal molecules such as Nav and contactin-associated protein (Caspr) were disrupted.[47] The axon diameter at C3-positive nodes was larger than at C3-negative nodes, indicating that larger axons in dorsal roots might be preferentially affected by IgG anti-GD1b antibodies, consistent with sensory ataxia (impaired deep sense).[47] Taken together, these findings show that complement-mediated nerve dysfunction is essential for development of neuropathy in anti-GD1b antibody- and anti-GM1 antibody-positive rabbit models. In a rabbit model of anti-GD1b-positive ataxic neuropathy, an apoptotic mechanism in dorsal root ganglion cells, especially large diameter cells, has also been proposed as a cause of development of ataxia.[48] Thus, activation of an apoptotic cascade might have a primary role in development of ataxia in anti-GD1b-positive GBS.

It is inferred from recent in vitro studies that a complement activation-independent mechanism can cause nerve dysfunction in GBS and related disorders. Rabbit IgG anti-GalNAc-GD1a antibody blocks neurotransmitter release by a presynaptic inhibitory effect on voltage-gated Ca channel currents through its binding to motor nerve terminals, independent of complement activation.[49, 50] IgG antibodies to GM1, GalNAc-GD1a or GD1a from AMAN patients also inhibit the Cav2.1 voltage-gated Ca channel current in cerebellar Purkinje cells.[51] Ex vivo, anti-GM1 or anti-GD1a mouse monoclonal antibodies decrease presynaptic-transmitter release in a complement-independent manner in the presynaptic membrane of motor nerves, probably because depolarization-induced calcium influx is inhibited.[52] Given that synaptic transmitter release is regulated by entry of Ca2+ through voltage-gated Ca channels at the presynaptic membrane, these observations indicate that antiganglioside antibodies might cause a complement-independent functional blockade of motor nerve terminals, leading to limb weakness in GBS. Neuromuscular transmission failure, however, has not been shown in clinical electrophysiological tests in GBS patients with antiganglioside antibodies.

As gangliosides cluster in biological membranes and are located on microdomains referred to as lipid rafts, antiganglioside antibodies might bind to gangliosides in the lipid rafts and exert a harmful effect on cell function, leading to complement-independent nerve damage. An intriguing study of biological effects of antiganglioside antibodies on PC12 cells showed that IgG anti-GM1-positive GBS sera inhibited nerve growth factor (NGF)-associated neurite outgrowth and reduced NGF-induced Trk autophosphorylation.[53] These sera caused displacement of Trk protein from the lipid raft fraction to the non-lipid raft fraction in cultured cell membranes; whereas a lipid raft marker, Ras, remained anchored in the lipid raft fraction. These observations indicate that anti-GM1 antibodies associated with GBS have a direct effect on the integrity of membrane lipid rafts and provoke complement-independent nerve cell damage.

Antibodies to GSC have a neurophysiological blocking effect at motor nerve terminals, as shown in a recent ex vivo study using a mouse phrenic/diaphragm muscle experimental model.[54] The pathophysiological potential of 31 GBS sera containing IgG anti-GM1/GD1a or anti-GM1/GQ1b antibodies was studied at mouse motor nerve terminals, and 17 of these sera induced increases in miniature end-plate potential frequency in the model. Immunohistochemical studies with confocal fluorescence microscopy revealed deposits of C3 and MAC at motor nerve terminals, accompanied by deposits of human IgG and loss of neurofilaments.[54] This shows that anti-GSC antibodies react with GSC antigens in living neural membranes and induce complement-mediated nerve damage. In view of the clustering of gangliosides in lipid rafts, another complement-independent mechanism of nerve damage can be proposed. Thus, binding of anti-GSC antibodies to GSC antigens that tend to form in lipid rafts might exert a direct effect on the integrity of the signaling platform, as shown in a recent in vitro study.[53]

Factors regulating the pathophysiological action of antiganglioside antibodies

The pathological effect of antiganglioside antibodies on the nerve cell membrane is regulated by antibody-antigen interactions, which are influenced by various factors. First, antibody specificity and the specific distribution of target gangliosides in the PNS are essential for antiganglioside antibody-mediated nerve dysfunction (Table 2).[10] Anti-GQ1b antibody-associated ophthalmoplegia in GBS and FS is principally based on the specific localization of GQ1b on paranodal myelin in human oculomotor, trochlear and abducens nerves.[4] Antibodies specific to GD1b precipitate ataxia in GBS,[13] and induce ataxic neuropathy in rabbits sensitized with GD1b, correlated with the location of GD1b in subsets of large dorsal root ganglion cells.[55] Second, complex glycolipid environments in the cell membrane affect accessibility and avidity of antiganglioside antibodies against target gangliosides.[13, 56] An examination of the specificity of IgG anti-GD1b antibodies using GSC antigens containing GD1b showed that activities of antibodies highly specific to GD1b were strongly inhibited by the addition of gangliosides with two or more sialic acids.[13] These GD1b-specific antibodies were closely associated with the development of ataxia.

Table 2. Target antigens of antiganglioside antibodies: localization in human peripheral nervous system and clinical features
AntigensClinical featuresLocalization in human peripheral nervous system
  1. AIDP, acute inflammatory demyelinating polyneuropathy; AMAN, acute motor axonal neuropathy; CIDP, chronic inflamatory demyelinating polyneuropathy; DRG, dorsal root ganglia; FS, Fisher syndrome; GBS, Guillain–Barré syndrome; PCB, pharyngeal-cervical-brachial.

GM1AMAN, pure motor GBSNot determined
GD1aAMANNot determined
GalNAc-GD1aAMAN, pure motor GBSPeriaxonal membrane of motor nerve at node and paranode, axolemma of small fibers in sural nerve
GM1bPure motor GBSNot determined
GD1bAtaxia in GBSLarge neurons in DRG, paranodal myelin
GD3AIDP, CIDPNot determined
GQ1bFS GBS with ophthalmoplegia

Paranodal myelin of oculomotor, trochlear, and abducens nerves

A part of DRG neurons

Nerve terminals inside muscle spindles and in the vicinity of intrafusal fibers

GT1aBulbar palsy in GBS PCB-GBSNot determined

It is inferred from these results that target epitopes of GD1b can be masked or modified by colocalization of gangliosides with two or more sialic acids, such as GD1a. Thus, colocalization of another ganglioside with GD1b might influence the accessibility of the anti-GD1b antibodies. Cis-interaction of the sugar chain of gangliosides in membranes might modify the conformations of glycoepitopes, and such complex glycolipid environments might regulate the accessibility and avidity of antiganglioside antibodies for target gangliosides. The hypothesis is supported by an in vitro and ex vivo study in GalNAc transferase-deficient (GalNAcT−/−) and GD3 synthase-deficient (GD3s−/−) mice.[56] The avidity of the pathogenic anti-GM1 antibody to GM1-like epitopes is governed by which glycolipid neighbors GM1 on the cell membrane and whether GM1 is unmasked. In another study, an epitope targeted by a monoclonal anti-GA1 antibody was masked in a GA1/GQ1b complex, whereas that targeted by a monoclonal anti-GQ1b antibody was preserved.[37] Finally, the anti-GQ1b antibody can access GQ1b epitopes in GA1/GQ1b, but the anti-GA1 antibody cannot access GA1 epitopes in the same complex. In addition to gangliosides, phospholipids might also exert an effect on the binding of antiganglioside antibodies to glycoepitopes. In GBS, anti-GM1 antibody activity against a mixture of GM1 and phospholipids, such as phosphatidic acid, phosphatidylinositol or phosphatidylserine, is higher than against GM1 alone, but anti-GM1 antibody activity against a mixture of GM1 and sphingomyelin is lower.[57] Thus, it should be borne in mind that the local glycolipid environment in the plasma membrane influences the avidity of antiganglioside antibodies.

Third, the large amount of targeted gangliosides in specific regions of peripheral nerves predisposes these regions to antiganglioside antibody-mediated damage. Anti-GD1a antibody-mediated nerve damage at motor terminals is observed in GD3-synthase knockout mice that overexpress GD1a, but not in normal mice,[58] indicating that specific loci with a high level of expression of GD1a might be predisposed to attack by anti-GD1a antibodies. Similarly, another study showed that anti-GM1/GD1a complex antibodies had a more harmful effect on nerve terminals in GD3-synthase knockout mice, which have an increased membrane density of a-series gangliosides GM1 and GD1a, compared with wild-type mice.[54] Fourth, the steric microstructure of gangliosides can influence avidity of antiganglioside antibodies. An immunohistochemical study using GD1a derivatives with chemically modified sialic-acid residues showed that anti-GD1a monoclonal antibodies that preferentially stained motor axons bound specifically to such GD1a derivatives as GD1a-1-ethylester, GD1a-1-alcohol and GD1a-1-methylester, in contrast to the reaction of another anti-GD1a monoclonal antibody that stained both motor and sensory axons.[59] Serum anti-GD1a antibodies from AMAN patients had a similar response to that of the motor-specific anti-GD1a monoclonal antibodies.

Fifth, antibody binding and development of nerve injury might be influenced by the conformation of glycoepitopes. Biochemical analysis of gangliosides in human motor and sensory nerves showed that the amount of GM1 and GD1a was almost equal in both nerves, but that the ceramide compositions differed between the two types of nerves.[60] The gangliosides from sensory nerves contain abundant long-chain fatty acids, in contrast to those from motor nerves. A binding assay using derivatives of GD1a bearing very long chain fatty acids showed that the difference in length of ceramide fatty acids decreased the binding of monoclonal anti-GD1a antibodies to the GD1a derivatives,[61] suggesting that the ceramide composition might regulate the structure of gangliosides. These results might partly explain the preferential binding of anti-GD1a antibodies from AMAN patients to GD1a in motor nerves.[62]

Interestingly, complement-mediated autoimmune nodal lesions were found in the dorsal root in one of 12 anti-GM1 antibody-positive rabbits with AMAN,[47] and some rabbits had morphological changes in dorsal roots or dorsal columns, although motor axons were mainly injured. These results suggest that IgG anti-GM1 antibodies can bind to GM1-like molecules in sensory nerves, as well as in motor nerves, and the greater damage to motor nerves compared with sensory nerves might be a result of biochemical and structural differences of glycoepitopes in gangliosides in the different nerves.[60-62]

Antibodies to gangliosides and ganglioside complexes in chronic immune-mediated neuropathies

In chronic immune-mediated neuropathies, such as multifocal motor neuropathy (MMN) and chronic inflammatory demyelinating polyneuropathy (CIDP), the pathogenetic role of antiganglioside antibodies remains to be elucidated. IgM anti-GM1 antibody is useful as a diagnostic marker of MMN, but the frequency is only approxiamtely 50%, and depends the clinical definition and detection techniques. Recent studies have reported frequencies of IgM amti-GM1 antibodies of 43%,[63] 46%,[64] and 58%.[65] In MMN, IgM anti-GM2 antibodies are found in 6%,[63] 10%[64] and 18%[65] of cases; and in CIDP, IgM antibodies to GM1 or GD1b are found in 3%.[64] A recent study by Kuwahara et al.[9] showed that IgG antibodies to peripheral nerve myelin, LM1, occur in 17.5% of CIDP patients. Whether the anti-LM1 antibody plays a pathogenetic role in development of CIDP requires further investigation.

IgM antibodies to GSC are also detected in sera from patients with MMN or CIDP, with antibodies to GM1/GT1b or GM2/GT1b found at rates of 3% in both diseases.[64] In a recent study using a newly developed combinatorial glycoarray with a polyvinylidene difluoride (PVDF) membrane in 33 patients with MMN, all patients had IgM anti-GM1/Gal-C antibodies and 94% had IgM anti-GM2/Gal-C antibodies.[65] A lipid mixture containing GM1 and Gal-C is a more sensitive antigen than GM1 alone,[66] but high frequencies of these antibodies are beyond expectation. Use of the combinatorial glycoarray might increase the diagnostic sensitivity of antiglycolipid antibodies, but these results require independent confirmation.

Future perspectives

Recent studies on the immunobiological mechanism in GBS have shed light on the pathophysiological effect of antiganglioside antibodies on nerve dysfunction. Many in vitro and ex vivo studies have suggested that antiganglioside antibody-mediated nerve dysfunction is mainly induced by complement activation and partly by a complement-independent mechanism, such as direct action on voltage-gated Ca channels or the integrity of lipid rafts in the cell membrane, which can be influenced by the avidity of antiganglioside antibodies. The concept of GSC provides new vistas for research on the antibody-antigen interactions in GBS. First, in addition to expansion of the spectrum of antiganglioside antibodies in GBS, investigation of anti-GSC antibodies will enhance their value as diagnostic markers. Second, the concept of GSC in microdomains can expedite elucidation of the mechanism underlying antiganglioside antibody-mediated nerve dysfunction. Microdomain function is governed by carbohydrate-binding proteins, such as selectins and Siglecs (sialic-acid-recognizing immunoglobulin-superfamily lectins), and is dependent on cis or trans carbohydrate-carbohydrate interactions.[67, 68] Complex glycoconjugates, such as GSC with clustered sialic acid epitopes, might form rigid rodlike structures with multivalency and strict binding specificity, and are likely to function in cell–cell recognition or immune-mediated events more effectively than a solo glycoepitope of an isolated ganglioside. This hypothesis is supported by an in vitro study showing that a GM2/GM3 complex more efficiently suppresses cell motility through blocking of cMet activation, compared with GM2 or GM3 alone.[69] Investigation of antibody-GSC antigen interactions in lipid rafts will unveil microdomain function mediated by carbohydrate–carbohydrate interactions in the cell membrane, and it will be intriguing and crucial to examine whether anti-GSC antibodies effectively interrupt the signaling pathway through binding to GSC antigens in the microdomains. A dimeric GM1-GD1a hybrid ganglioside derivative that contains two structurally different oligosaccharide chains has recently been prepared.[70] Such dimeric hybrid ganglioside derivatives that mimic GSC are extremely useful tools for research on anti-GSC antibody-mediated cell dysfunction and might open new perspectives for the understanding and treatment of immune-mediated neuropathies.

Disclosure

The author has no competing interests.

Ancillary