Pathogenesis

Molecular mimicry is currently the major theme in the pathogenesis of GBS, particularly in cases that follow Campylobacter jejuni infection. This hypothesis is supported by the following key observations: (1) C. jejuni enteritis is the most commonly recognized antecedent infection in GBS; (2) different variants of GBS, particularly FS and AMAN, are strongly associated with specific anti-ganglioside antibodies; (3) the lipooligosaccharides (LOSs) of C. jejuni isolates from patients with GBS carry relevant ganglioside-like moieties; (4) gangliosides, the purported target antigens, are enriched in the nerve fibers; and (5) pathological and immunopathological studies in the AMAN variant of GBS indicate antibody-mediated axonal injury. The current discussion on molecular mimicry is restricted to C. jejuni because most of the evidence has been obtained from the post- Campylobacter cases, although other bacteria such as Haemophilus influenzae have also been reported to carry ganglioside-like moieties (Mori et al., 2001). Further, anti-ganglioside antibodies in patients with GBS can also be detected in cases preceded by viral infections (Jacobs et al., 1998). Another relatively novel feature of molecular mimicry in GBS is the nature of antigens on the infecting organism and the relevant target antigens in peripheral nerves. These purported antigens on the infectious organism and in the peripheral nerves are carbohydrates that are carried on lipid backbones. In this context, relevant pathogenetic properties/issues relating to C. jejuni and LOSs containing ganglioside-like moieties, anti-ganglioside antibodies, and gangliosides in peripheral nerves are discussed further.

1. Campylobacter jejuni

C. jejuni is a gram-negative non-spore-forming entero-pathogen that is one of the most common causes of bacterial gastroenteritis worldwide, especially in children (Hughes and Rees, 1997; Friedman et al., 2000; Oberhelman and Taylor, 2000). Infection with C. jejuni is the most frequently recognized event preceding AMAN and other variants of GBS (reviewed in Hughes and Rees, 1997). Most Campylobacter infections are sporadic and are associated with ingestion of improperly handled or cooked food; poultry products are a major source of human infection. In GBS cases, both stool culture and serologic methods are needed to diagnose Campylobacter infection, because by the time neurological symptoms develop, the yield of C. jejuni from stool culture is relatively low (Nachamkin, 1997). Rhodes and Tattersfield described the first case of GBS following C. jejuni gastroenteritis in 1982 (Rhodes and Tattersfield, 1982). Subsequent studies have confirmed this association; however, the incidence of preceding C. jejuni infection varies widely, ranging from 4% in North America to 74% in northern China (Hughes and Rees, 1997; Ho et al., 1999), with an overall prevalence estimated at approximately 30% (reviewed in Moran et al., 2002). Based on the known incidences of GBS and C. jejuni enteritis, it is estimated that 1 in 1,000 cases of Campylobacter infection is complicated by GBS. Because GBS is a rare complication after C. jejuni infection, attention has focused on host and Campylobacter properties that may lead to this sequela. The host properties that could confer susceptibility to GBS after C. jejuni infection are not well established. Although some reports indicate post-Campylobacter GBS cases preferentially associate with certain HLA alleles, the significance of these findings remains unclear because of lack of confirmatory studies (Yuki et al., 1991; Rees et al., 1995c).

The issue of Campylobacter-related factors has been examined mainly by serotyping and characterizing the gan-glioside-like mimicry in the LOS of the GBS and enteritis isolates. C. jejuni is genetically heterogeneous and a large number of serotypes are recognized. Penner serotyping, based on the heat-stable capsular polysaccharide antigens, is most commonly used for classification of clinical isolates, including those from patients with GBS. Numerous C. jejuni serotypes have been associated with GBS and of particular interest is the observation that certain GBS-associated strains are uncommon in patients with uncomplicated gastroenteritis; for example, Penner serotype HS:19 is overrep-resented in GBS patients compared to diarrhea isolates in some but not all populations (Fujimoto et al., 1992; Yuki et al., 1992b; Kuroki et al., 1993; Sheikh et al., 1998; Rees et al., 1995b; Jacobs et al., 1997). Another striking example of the association of uncommon serostrains with GBS is that of serostrain HS:41 in GBS cases from Cape Town, South Africa (Goddard et al., 1997). The overrepre-sentation of serostrains HS:19 and HS:41 in patients with GBS has supported the notion that as yet undefined properties related to the organism are critical in determining whether or not GBS follows C. jejuni infection. GBS-asso-ciated isolates are not restricted to these two uncommon serotypes and other C. jejuni serotypes isolated from GBS include a large number of other Penner serotypes (Prendergast and Moran, 2000).

The lipopolysaccharide (LPS) of gram-negative bacteria, including C. jejuni, consists of three components: lipid A, the hydrophobic region inserted into the membrane; an oligosaccharide core divided into inner and outer parts; and capsular polysaccharides, also called O-chains. LOS is differentiated from LPS by lack of O-chains. In C. jejuni it is the oligosaccharide core region that carries ganglioside-like moieties (Fig.1A). Several studies have characterized the core regions of LPS/LOS of GBS- and diarrhea-associated C. jejuni strains. Yuki et al. first described the presence of a GM1-like structure in C. jejuni LPS isolated from a GBS patient and subsequent studies have shown the presence of GM1-, GD1a-, GalNAc-GD1a-, GM1b-, GT1a-, GD2-, GD3-, and GM2-like structures (Yuki et al., 1992a; Aspinall et al., 1993, 1994a, 1994b, 1996; Sheikh et al., 1998; Nachamkin et al., 2002). Although structural studies have failed to demonstrate a GQ1b-like structure, the purported target antigen for FS, antibody binding assays with human or murine monoclonal antibodies have shown the presence of GQ1b- and GT1a-cross-reactive moieties in C. jejuni LPS/LOSs (Jacobs et al., 1995, 1997; Yuki et al., 1994). Figure 1B shows the ganglioside-like moieties, as determined by mass spectrometry, in C. jejuni LOS implicated in AMAN and FS. It is clear that both GBS- and diarrhea-associated isolates carry ganglioside-like moieties but

A. Structure of LPS

Lipooligosaccharide (LOS)

Lipopolysaccharide (LPS)

B. Structure of ganglioside-mimics in the inner core of LPS

GM1-like

GD1a-like

GT1a-like

C. Structure of the gangliosides implicated as target antigens AMAN FS

GM1a

GalNAAc-GD1a

GM1b

GQ1b

GT1a

Key: □ Glc V 2,3-NeuAc ■ Gal ▼ 2,8-NeuAc OGalNAcO Kdo • Hep

Figure I (A) Diagrammatic representation of the structure of LPS and LOS in gram-negative bacteria. (B) Structure of ganglio-side-mimics in C.jejuni core oligosaccharide as determined by different mass spectroscopy studies. (C) Structure of gangliosides implicated as target antigens in acute motor axonal neuropathy and Fisher syndrome.

GBS-related organisms are more likely to do so (Nachamkin et al., 1999, 2002). These studies indicate that expression of ganglioside-structures in LPS by itself is not sufficient to impart GBS-inducing properties to the infecting organism.

2. Anti-Ganglioside Antibodies

The possibility that carbohydrate moieties on glycolipids could be autoantigens in neuropathic disorders was realized with the description of a case with chronic IgM parapro-

teinemic neuropathy in which the paraprotein recognized a structure shared by several gangliosides (Ilyas et al., 1985). Initial descriptions of anti-ganglioside antibodies in patients with GBS were based on small case series or case reports (Ilyas et al., 1988, 1992; Yuki et al., 1990; Walsh et al., 1991). Subsequently, several large studies examined the clinicoserological correlations in GBS, leading to identification of some putative ganglioside target antigens (Rees et al., 1995a; Jacobs et al., 1996; Ho et al., 1999;

Ogawara et al., 2000). Besides identifying the target antigens, antiglycolipid serological studies tend to correlate the specificity of antibody responses with specific clinical and pathological features, such as motor vs. sensory fiber involvement or axonal vs. demyelinating injury. There are several limitations inherent to such large population-based serological studies, but despite these caveats, certain patterns/associations have emerged.

The antibody correlations for AIDP variant are not well defined. In contrast, two major gangliosides, GM1 and GD1a, and two minor gangliosides, GalNAc-GD1a and GM 1b, are implicated as target antigens in AMAN (Rees et al., 1995a; Jacobs et al., 1996; Ogawara et al., 2000; Hadden et al., 1998; Ho et al., 1999; Yuki et al., 1993b; Kusunoki et al., 1994, 1996a). Antibodies against GM1 and GD1a gangliosides can be detected in up to 50% to 60% of patients with AMAN in Japanese and northern Chinese populations (Ogawara et al., 2000; Ho et al., 1999). The frequency of anti-GalNAc-GD1a and GM 1b antibodies in motor-predominant syndromes is much lower, in the range of 10% to 15% (Ang et al., 1999; Yuki et al., 1999, 2000). Anti-GQ1b antibodies, frequently cross-reactive with other structurally related gangliosides containing disialosyl moieties, are present in up to 80% to 90% of patients with FS (Willison et al., 1993; Yuki et al., 1993a; Carpo et al., 1998), and this correlation provides the strongest association between antibodies to a specific ganglioside and a clinical phenotype.

In chronic neuropathies associated with anti-ganglioside reactivity, it is usually the pentameric IgM that binds to gly-colipids. In contrast, the anti-ganglioside antibodies in acute neuropathic conditions such as AMAN and FS are predominantly of IgG isotype, generally complement-fixing subtypes IgG1 and IgG3 (Willison and Veitch, 1994; Ogino et al., 1995). These anti-ganglioside antibody responses in GBS are almost always polyclonal and can have a broad range of cross-reactivity with structurally related ganglio-sides (Koga et al., 2001). It has been proposed that differences in geography and or genetic background may affect the specificity and isotype distribution of anti-ganglioside responses in different populations (Ogawara et al., 2000; Ang etal., 2001b).

The mechanism(s) of generation of anti-ganglioside responses in GBS remain elusive. The following observations support the hypothesis of molecular mimicry in post-Campylobacter GBS cases. First, studies that demonstrate that either human GBS sera or purified human anti-ganglio-side antibodies bind to ganglioside-like moieties contained in the core oligosaccharide region of the LPS/LOS (Wirguin et al., 1994; Oomes et al., 1995; Sheikh et al., 1998); one extension of this observation is that the cross-reactive sugar moieties in LOS triggered the production of these antibodies. Second, immunization with C. jejuni LPSs/LOSs can successfully induce anti-ganglioside antibodies (Wirguin et al., 1997; Goodyear et al., 1999; Ang et al., 2001a). These immune responses in experimental animals have generally induced low-affinity, non-T-cell-dependent antibodies of IgM and IgG3 type (non-complement-fixing isotype in mice), despite the use of adjuvants to recruit T-cell help (Wirguin et al., 1997; Goodyear et al., 1999), reflecting a high level of tolerance to self-gangliosides restricting antibody responses to C. jejuni LPSs (Bowes et al., 2002). In contrast, serological studies in AMAN and FS indicate class switching to IgG and subclass restriction to IgG1 and IgG3 (complement-fixing isotypes in humans), both usually features of T-cell (Th1) help and atypical of the human anticar-bohydrate antibody responses. The high level of tolerance to self-gangliosides in experimental animals can be repro-ducibly overcome, probably with successful recruitment of T-cell help and production of IgG subclasses, by immunization with gangliosides or C. jejuni LPSs in transgenic animals lacking complex gangliosides (Lunn et al., 2000; Bowes et al., 2002). The possibility that IgG anti-ganglioside antibody production does not require immunoglobulin class switching but reflects a primary B-cell response has not been examined in patients with GBS or experimental models.

Observations from these immunization studies raise the possibility that under some conditions, the high level of tolerance to self-gangliosides can be overcome after C. jejuni infection, leading to induction of T-cell-dependent IgG anti-ganglioside responses similar to those seen in human GBS cases. These experimental findings support the notion that breakdown of tolerance to self-gangliosides is critical in the pathogenesis of post-Campylobacter GBS. In such a model, it would not be far-fetched to propose that although generation of these antibodies reflects breakdown of strong tolerance to gangliosides, the immune system has another check to contain such an aberrant immune response, when it arises, either by turning off the anti-ganglioside antibody production by B-cells or more likely by their elimination/deletion. The existence of such a system of containment for autoreac-tive B-cells secreting anti-ganglioside antibodies is argued by the observations that post-Campylobacter GBS is monophasic, associated anti-ganglioside antibody titers reach a peak with the nadir of the disease and decay thereafter, and the extremely rare incidence of recurrence of clinical disease even in northern China, where C. jejuni is endemic and chances of reexposure are not trivial. An alternate explanation is that the development of these antibodies reflects thymus-independent B-cell responses, which typically do not induce long-term immunological memory. Further experimental work in this area may clarify the mechanisms involved in the generation and elimination of human anti-ganglioside antibodies after C. jejuni infection.

3. Peripheral Nerve Gangliosides

Gangliosides are sialic acid-containing glycosphin-golipids that are widely distributed in mammalian tissues but are particularly enriched in the nervous system. Gangliosides are classified on the basis of the number and linkages of the sugar backbone and attached sialic acids. All complex gangliosides have a tetraose (four-sugar) backbone/core, consisting of galactose, ^-acetylgalactosamine, galactose, and glucose, to which variable numbers of sialic acids are attached; the ceramide or lipid portion of the molecule is attached to the internal glucose. Although there are a large number of ganglioside species, GM1, GD1b, GD1a, and GT1b are the four most abundant complex gangliosides in the nervous system. Simpler gangliosides have a shorter core structure instead of the four-sugar backbone structure. Anti-ganglioside antibody binding to these carbohydrate moieties determines their specificities. The structures of gangliosides implicated as target antigens in AMAN and Fisher variants of GBS are depicted in Figure 1C.

One reason for defining the associations between anti-ganglioside antibody specificity and clinical variants of GBS is the simple hypothesis that differences in clinical manifestations may relate to differences in the distribution of target antigens (gangliosides) in different regions of the nervous system. A large number of biochemical studies indicate that there are no consistent differences in the ganglio-side content of motor and sensory fibers to explain preferential motor nerve fiber injury, except that carbohydrate-anchoring lipids in sensory and motor nerves of gan-gliosides are different (Ogawa-Goto et al., 1990; Gong et al., 2002). Immunolocalization studies have examined the distribution of GM1 and GD1a in peripheral nerves. These gan-gliosides are localized at nodes of Ranvier and on the nodal and internodal axolemma (Fig. 2A and B) (Sheikh et al., 1999; Gong et al., 2002). GM1 is also enriched in the para-nodal myelin (Fig. 2A) (Ganser et al., 1983; Sheikh et al., 1999). Most of the available data suggest that anti-ganglio-side antibodies bind to the Schwann cell surface but not to compact myelin. GM1 and GD1a are also concentrated in motor nerve terminals (Fig. 2C). We have recently generated several monoclonal anti-GD1a antibodies and found that anti-GD1a monoclonal antibodies preferentially stain myelinated motor axons but not sensory myelinated axons (Fig. 3) (Gong et al., 2002), but the basis of this differential antibody binding to motor fibers remains unclear. Our finding was supported by a recent study showing the preferential staining of nodes of Ranvier in motor nerve fibers by serum with GD1a reactivity obtained from a patient with AMAN (De Angelis et al., 2002). The distribution of minor gangliosides GalNAc-GD1a and GM1b is not well characterized.

The distribution of FS targets GQ1b and related cross-reactive gangliosides in peripheral and cranial nerves have also been studied by biochemical and immunohistochemical techniques. Anti-GQ1b antibodies bind to paranodal myelin and nodes of Ranvier, and it has been shown that the extraoc-ular cranial nerves are slightly enriched in GQ1b compared to other cranial and peripheral nerves (Chiba et al., 1993, 1997). That GQ1b is not restricted to cranial nerves is also supported by studies showing binding of disialosyl antibodies to the nodes of Ranvier in somatic nerves. Muscle spindles, motor nerve terminals in somatic and extraocular muscles, and intrafusal muscle fibers are also labeled by antibodies cross-reactive with GQ1b (Willison et al., 1996; Goodyear et al., 1999). The observation that anti-GQ1b reactive antibodies bind to motor nerve terminals has been exploited by investigators for experimental modeling (see later) and may be relevant to FS pathogenesis.

In summary, except for GD1a localization, ganglioside distribution studies do not provide a satisfactory explanation for preferential motor fiber injury in AMAN or regional-localization of symptoms in FS. These findings suggest that it is more than likely that other factors besides ganglioside density and distribution also play important roles in anti-ganglioside antibody-mediated injury. These factors include relative differences in the blood-nerve barrier in different regions of the nerve fibers and between sensory and motor spinal roots; the motor nerve terminal provides such an example. One favored explanation for the absence of CNS involvement in AMAN despite a high level of ganglioside expression is that the blood-brain barrier is much less permeable than is the blood-nerve barrier. The existence of such differences in barrier permeability between sensory and motor roots, however, has not been reported. Further, the conduction properties of motor fibers are significantly different from those of sensory fibers, and one could speculate that these fibers are more prone to antibody-mediated conduction failure; this may be particularly relevant early in the pathogenesis of AMAN. Moreover, susceptibility to injury could also differ between motor and sensory fibers or different regions of the same nerve fiber based on the differences in expression of target gangliosides or their accessibility to anti-ganglioside antibodies. In summary, the basis of preferential motor injury in AMAN remains incompletely understood, and ongoing work is likely to provide further clues to this interesting pathogenetic issue.

The following properties of gangliosides in nerves make them suitable to serve as target antigens in nerve fibers. The ceramide portion of gangliosides anchors them into the plasma membranes, whereas oligosaccharide moieties extend into the extracelluar space from the cell surface. This membrane organization allows anti-ganglioside antibodies to bind to sugar moieties extending into the extracellular space. Immunolocalization studies have confirmed that patient-derived or experimental anti-ganglioside antibodies bind these sugars on nerve fibers, thus confirming their accessibility to immune effectors. Recent work indicates that gan-gliosides are focally enriched into functional domains called rafts into which protein receptors or ion channels can be actively recruited or excluded. Experimental evidence indicates that signal transduction can occur through lipid rafts, and these ganglioside-enriched microdomains can also modulate receptor function. Further, sodium and postassium channels are clustered in the ganglioside-enriched regions of myelinated nerve fibers (i.e., nodes and paranodes) (Black et al., 1990). These properties of myelinated nerve fiber gangliosides provide clues to the potential mechanisms of

Figure 2 Localization of GM1 ganglloside in peripheral nerves. (A) Teased nerve fiber preparation showing GM1 staining at node of Ranvier (arrow) and paranodal Schwann cell. (B) Immunoelectron microscopy showing GM1 staining at the nodal axolemma and extension along paranodal axolemma. (C) Fresh-frozen muscle preparation showing GM1 staining (brown) of an intramuscular nerve (arrow) and a motor nerve terminal (arrowhead); blue staining shows motor end plates. (Adapted with permission from Sheikh et al., 1999.)

Figure 2 Localization of GM1 ganglloside in peripheral nerves. (A) Teased nerve fiber preparation showing GM1 staining at node of Ranvier (arrow) and paranodal Schwann cell. (B) Immunoelectron microscopy showing GM1 staining at the nodal axolemma and extension along paranodal axolemma. (C) Fresh-frozen muscle preparation showing GM1 staining (brown) of an intramuscular nerve (arrow) and a motor nerve terminal (arrowhead); blue staining shows motor end plates. (Adapted with permission from Sheikh et al., 1999.)

anti-ganglioside antibody-mediated nerve fiber dysfunction, as discussed in the next section.

Peripheral Neuropathy Natural Treatment Options

Peripheral Neuropathy Natural Treatment Options

This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.

Get My Free Ebook


Post a comment