Brain Advance Access originally published online on May 16, 2008
Brain 2008 131(7):1926-1939; doi:10.1093/brain/awn074
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Structural requirements of anti-GD1a antibodies determine their target specificity
1Department of Neurology, Johns Hopkins University, Baltimore, MD, USA, 2Department of Neurology, The Second Teaching Hospital, Hebei Medical University, Shijiazhuang, People's Republic of China, 3Department of Biophysics and Biophysical Chemistry and 4Department of Pharmacology, Johns Hopkins University, Baltimore, MD, USA
Correspondence to: Dr Kazim Sheikh, Department of Neurology, Johns Hopkins Hospital, 600 N. Wolfe St., 509 Pathology Building, Baltimore, MD, 21287, USA E-mail: ksheik{at}jhmi.edu
 |
Summary |
|---|
 Top Summary IntroductionMaterials and MethodsResultsReferences |
|---|
The acute motor axonal neuropathy (AMAN) variant of Guillain–Barré syndrome (GBS) is associated with anti-GD1a and anti-GM1 IgG antibodies. The basis of preferential motor nerve injury in this disease is not clear, however, because biochemical studies demonstrate that sensory and motor nerves express similar quantities of GD1a and GM1 gangliosides. To elucidate the pathophysiology of AMAN, we have developed several monoclonal antibodies (mAbs) with GD1a reactivity and reported that one mAb, GD1a-1, preferentially stained motor axons in human and rodent nerves. To understand the basis of this preferential motor axon staining, several derivatives of GD1a were generated by various chemical modifications of N-acetylneuraminic (sialic) acid residues (GD1a NeuAc 1-amide, GD1a NeuAc ethyl ester, GD1a NeuAc 1-alcohol, GD1a NeuAc 1-methyl ester, GD1a NeuAc 7-alcohol, GD1a NeuAc 7-aldehyde) on this ganglioside. Binding of anti-GD1a mAbs and AMAN sera with anti-GD1a Abs to these derivatives was examined. Our results indicate that mAbs with selective motor axon staining had a distinct pattern of reactivity with GD1a-derivatives compared to mAbs that stain both motor and sensory axons. The fine specificity of the anti-GD1a antibodies determines their motor selectivity, which was validated by cloning a new mAb (GD1a-E6) with a chemical and immunocytochemical binding pattern similar to that of GD1a-1 but with two orders of magnitude higher affinity. Control studies indicate that selective binding of mAbs to motor nerves is not due to differences in antibody affinity or ceramide structural specificity. Since GD1a-reactive mAb with preferential motor axon staining showed similar binding to sensory- and motor nerve-derived GD1a in a solid phase assay, we generated computer models of GD1a based on binding patterns of different GD1a-reactive mAbs to different GD1a-derivatives. These modelling studies suggest that critical GD1a epitopes recognized by mAbs are differentially expressed in motor and sensory nerves. The GD1a-derivative binding patterns of AMAN sera resembled those with motor-specific mAbs. On the basis of these findings we postulate that both the fine specificity and ganglioside orientation/exposure in the tissues contribute to target recognition by anti-ganglioside antibodies and this observation provides one explanation for preferential motor axon injury in AMAN.
Key Words: acute motor axonal neuropathy; anti-ganglioside antibodies; gangliosides; Guillain–Barré syndrome; immune neuropathies
Abbreviations: Ab, antibody; Abs, antibodies; AMAN, acute motor axonal neuropathy; DRG, dorsal root ganglion; ELISA, enzyme linked immunosorbent assay; GBS, Guillain–Barré syndrome; IgG, immunoglobulin G; mAb, monoclonal Ab; MoS, motor-specific; NeuAc, N-acetylneuraminic acid; NoS, nonselective; PBS, phosphate-buffered saline; TLC, thin layer chromatography; 3D, three-dimensional
Received April 20, 2007. Revised February 14, 2008. Accepted March 26, 2008.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsReferences |
|---|
Gangliosides are sialic-acid containing glycosphingolipids with ubiquitous expression but particular enrichment in the vertebrate nervous system (Yu and Saito, 1989
; Svennerholm, 1994). Several lines of evidence support diverse roles of these glycolipids in biology (Sheikh et al., 1999b; Kawai et al., 2001; Vyas et al., 2002) and pathobiology (Kusunoki et al., 1996; Goodyear et al., 1999; Yuki et al., 2001; Bullens et al., 2002; Sheikh et al., 2004; Lehmann et al., 2007). Despite their ubiquitous expression, how these moieties mediate specific functions, often selectively in certain cell types, remains a challenging biologic puzzle. The acute motor axonal neuropathy (AMAN) variant of Guillain–Barré syndrome (GBS) provides such a pathobiological example in which antibodies (Abs) directed against GM1 and GD1a gangliosides are associated with selective dysfunction and injury of motor axons, despite the similar expression of these gangliosides in human motor and sensory nerve fibers (Svennerholm et al., 1992, 1994; Gong et al., 2002; Willison and Yuki, 2002; Hughes and Cornblath, 2005). Notably, the role of anti-ganglioside Abs in the pathogenesis of AMAN is strongly supported by an animal model in which induction of these Abs is associated with reproduction of clinicopathological features of human disease (Yuki et al., 2001, 2004), and it has been demonstrated experimentally that Ab-mediated injury to nerve cells or nerve fibres is due to targeting of gangliosides (Zhang et al., 2004; Sheikh et al., 2004; Goodfellow et al., 2005; Lehmann et al., 2007).
To provide a better understanding of preferential motor axon involvement in AMAN, we generated several monoclonal antibodies (mAbs) with GD1a reactivity (Lunn et al., 2000; Schnaar et al., 2002) and previously reported that one mAb, GD1a-1, preferentially stains motor axons in human and rodent nerves (motor-selective, MoS) compared to two GD1a-reactive mAbs that bind both motor and sensory nerves (nonselective, NoS), despite similar concentrations of GD1a in human motor and sensory nerves (Gong et al., 2002). The fine specificity of MoS and NoS anti-GD1a mAbs is different, as suggested by varied levels of crossreactivity of these mAbs to synthetic GT1a
, a ganglioside that has not been reported to be expressed in the peripheral nervous system (Gong et al., 2002). Our previous studies show that MoS anti-GD1a mAb binds to GD1a derived from both motor and sensory human nerves in solid phase assays (Gong et al., 2002). From these observations we hypothesize that GD1a epitopes recognized by MoS and NoS mAbs in motor and sensory nerves are expressed differentially.
To test this hypothesis of differential expression of GD1a epitopes, N-acetylneuraminic acid (NeuAc) residues on GD1a ganglioside were chemically modified and binding of MoS and NoS GD1a-reactive mAbs to these GD1a-derivatives was examined. Our results indicate that MoS and NoS mAbs had different structural requirements. Using this MoS mAb-binding template for screening, we cloned another GD1a-reactive mAb, which, as predicted, also preferentially stained motor nerves. With these Ab-binding profiles and published torsion angle data on GD1a conformers (Acquotti et al., 1994), we generated computer maps identifying differential exposure of critical residues recognized by MoS and NoS mAbs in different GD1a conformers. Reactivity of AMAN sera with high titers of anti-GD1a Abs to GD1a-derivatives resembled the binding pattern of MoS anti-GD1a mAbs. On the basis of these observations we postulate that motor nerves express GD1a conformer(s) that are different from those in sensory nerves: this provides one explanation for preferential Ab-binding and motor nerve injury in AMAN.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsReferences |
|---|
Specificity of motor-selective (MoS) and nonselective (NoS) anti-GD1a mAbs
The generation, specificity and purification of these mAbs were described in previous publications (Lunn et al., 2000
; Schnaar et al., 2002). Three anti-GD1a IgG monoclonal Abs (mAbs), GD1a-1, GD1a-2a and GD1a-2b, were used initially. These mAbs are designated according to their ganglioside specificity and IgG isotype (1, 2a or 2b); for example, GD1a-2b refers to a mAb with GD1a specificity and IgG2b isotype. These mAbs were purified according to previously described methods (Lunn et al., 2000; Schnaar et al., 2002) and were used in all assays; the specificity (Schnaar et al., 2002) and tissue binding patterns (Gong et al., 2002) are presented in Table 1.
|
Generation of GD1a derivatives
Figure 1 shows the structures of the sialic-acid derivatives used in this study. The glycerol chains of GD1a sialic acids were oxidized to the 7- and 8-aldehyde forms and then reduced to the corresponding truncated primary alcohols as described (Spiegel et al., 1979
). Briefly, GD1a (1 mg) was treated with 1 ml of ice-cold 150 mM NaCl, 50 mM sodium phosphate, pH 7.4, containing 2.5 mM sodium periodate for 90 min. The aldehyde products were purified by reverse phase chromatography (with a Sep-Pak C18 cartridge; Millipore Corp., Milford, MA) as described (Schnaar, 1994). A portion of the resulting GD1a-aldehyde was dissolved in 1 ml of 50 mM sodium bicarbonate containing 10 mM sodium borohydride and incubated for 2 h at 37°C. The resulting products were purified by reverse phase chromatography (Schnaar, 1994). The sialic acids on 1 µmol of GD1a were converted to the corresponding ethyl esters by incubation for 1 h at ambient temperature in 0.6 ml of Me2SO/iodoethane (5 : 1) (Handa and Nakamura, 1984). The resulting diester was purified by DEAE-Sepharose (acetate form) chromatography in methanol. The product in the unretained eluent was further purified by reverse phase chromatography (Schnaar, 1994). One portion of the GD1a diethyl ester was reduced to the corresponding dialcohol in 1% sodium borohydride in methanol for 2 h. The product was purified by DEAE-Sepharose and reverse phase chromatography. Another portion of the GD1a ethyl ester was converted to the corresponding diamide by treatment with 4 M ammonium hydroxide in methanol/water (5 : 2) for 12 h at ambient temperature. The product was purified by reverse phase and DEAE-Sepharose chromatography. The methyl ester derivative of GD1a was obtained by incubation of GD1a for 1 h at ambient temperature in 0.6 ml of Me2SO/iodomethane (5 : 1) (Handa and Nakamura, 1984). GD1a methyl ester was further purified as described above. Products were analysed by thin-layer chromatography (TLC) and fast atom bombardment-mass spectrometry (Dell et al., 1994) at the Middle Atlantic Mass Spectrometry Laboratory.
|
Binding of mAbs to GD1a-derivatives
mAb binding to GD1a and GD1a-derivatives was determined by TLC-immunooverlay as described (Lopez et al., 2000
). GD1a and GD1a-derivatives (100 pmol) were spotted on aluminum silica gel HPTLC plates and resolved using chloroform:methanol:aqueous 0.02% CaCl2 (60 : 35 : 8). For selected studies with NoS mAbs up to 300 pmol of GD1a—derivatives (3-fold higher concentration) were spotted to detect low affinity binding. Developed chromatograms were dipped in 2 mg/ml of polyisobutylmethacrylate in hexane and air-dried. Then the plates were incubated overnight with mAbs (1–10 µg/ml diluted in PBS, pH 7.4, containing 0.1% bovine serum albumin and 0.05% Tween 20) and developed with alkaline phosphatase-conjugated goat anti-mouse IgG (80 ng/ml for 90 min; 
chain-specific; Jackson Immunoresearch, USA) and nitroblue tetrazolium/bromochloroindolyl phosphate. The resulting bands were scanned and the signal was quantified with software ScionImage. These studies were repeated at least once with all mAbs.
Binding of GBS sera to GD1a-derivatives
AMAN sera with high titers of anti-GD1a reactivity were analysed for these proof of concept studies. Anti-GD1a Abs were affinity-purified from five sera according to methods described by Hirabayashi et al. (1983). Briefly, 1.5 ml of methanol:aqueous 0.2 M KCl (1 : 1) containing 0.25 µmol of GD1a were added to 1 ml of octyl-Sepharose (Sigma). The mixture was rotated end-over-end for 1 h at room temperature. After removal of the methanolic solution, the gel was washed thoroughly with PBS. All the remaining steps were performed at 4°C. Portions of gel (200 µl) were loaded on columns and patient sera were recycled through the column for 2 h with a peristaltic pump. The columns were washed with PBS, and the retained Abs were eluted with a solution of 100 mM glycine–HCl, pH 3.0, and immediately neutralized in a 1/10 volume of 1 M Tris–HCl, pH 8.0. Then, the antibodies were desalted in Sephadex G-25 in 0.1% BSA-PBS and frozen at –20°C until used. Affinity-purified anti-GD1a Abs from five patients and serum from one patient were examined for binding to GD1a derivatives by TLC-immunooverlay as described in the preceding section.
Five additional sera from AMAN cases were analysed by ELISA. GD1a ganglioside or its derivatives (100 pmol/well) were dissolved in methanol and adsorbed to butanol-washed 96-well plates. In control wells ganglioside was omitted. Ganglioside-adsorbed plates were washed with PBS, and then plates were blocked for 60 min at 4°C with 200 µl/well of PBS containing 1 mg/ml of bovine serum albumin (PBS/BSA; dilution buffer). Sera were incubated overnight at 4°C, the plates were washed with PBS, and developed with alkaline-phosphatase-conjugated secondary antibody (goat anti-human IgG, Fc-specific; Jackson Immunoresearch, West Grove, PA), diluted 1 : 500 (90 min) and substrate buffer (2 mg/ml p-nitrophenyl phosphate in 100 mM Tris, 100 mM NaCl, 5 mM MgCl2, pH 9.5, 100 µl/well) for 30 min. Absorbance was determined at 405 nm with microplate reader (Benchmark; BioRad, Hercules, CA).
Cloning of new anti-GD1a mAb
We used a preexisting library of anti-GD1a clones derived from our previous immunization studies with GD1a (Lunn et al., 2000; Schnaar et al., 2002). These clones were preselected on the basis of their binding to GD1a among four major gangliosides (GM1, GD1a, GD1b and GT1b). GD1a-derivatives were used for screening studies as described in the preceding sections. One mAb, designated GD1a-E6, was selected because the pattern of reactivity with GD1a-derivatives was similar to that of mAb GD1a-1 (MoS). The specificity of GD1a-E6 against 14 gangliosides (Table 1) was also examined by ELISA. GD1a-E6 was included in parallel with other GD1a mAbs in all the assays used in this study.
Immunostaining
These studies were done by previously described methods (Sheikh et al., 1999a; Gong et al., 2002). Cauda equina collected from adult 8–12-week-old Sprague–Dawley rats were snap-frozen in isopentane at –70°C and cryosectioned. These sections were stained with mAbs GD1a-2a, GD1a-2b, GD1a-1 and GD1a-E6 (2–20 µg/ml; overnight incubations at 4°C). These sections were developed with specific secondary antibodies conjugated to FITC or Cy3 (1 : 200, 1 h at ambient temperature) and examined by an epifluorescence microscope, digitized and processed in Adobe® Photoshop® (San Jose, CA, USA).
In a subset of studies tissue binding of GD1a-E6 was competed with soluble GD1a methyl ester-oligosaccharide (see below). GD1-E6 (2 µg/ml) was preincubated with GD1a methyl ester-oligosaccharide (1 x 10–4 M) for 1 h at 4°C and then this mixture was used for staining the cauda equina as described above. The inhibitory studies were preferentially done with GD1a methyl ester-oligosaccharide in order to avoid non-specific adsorption to tissues of GD1a methyl ester ganglioside. GD1a-2b was used as negative control to show that GD1a methyl ester-oligosaccharide does not cause non-specific inhibition of antibody binding to tissues. GD1a-1 was not tested because the affinity of this mAb for this derivative is >1 x 10–4 M (see ‘Results’) and enough oligosaccharide was not available to perform these studies.
An AMAN serum with high titers of anti-GD1a Abs (98-2) was affinity-purified according to methods described by Hirabayashi et al. (1983). Fresh-frozen Cauda equina sections were immunostained with affinity-purified anti-GD1a Abs as described above (1 : 30; overnight incubations at 4°C). These sections were developed with biotin-conjugated anti-human chain-specific Abs (1 : 200; overnight incubation at 4°C; Jackson Immunoresearch Lab) and Alexa fluor 546-conjugated streptavidin (1 : 200; 90 min incubation at 4°C) and examined by an epifluorescence microscope, digitized and processed in Adobe® Photoshop®.
Determination of affinity
A soluble binding inhibition assay was used for this purpose, as described (Lopez et al., 2002). We used GD1a-oligosaccharides because the competing antigen is soluble and the value thus obtained can be considered as monovalent affinity. GD1a-oligosaccharides from GD1a or GD1a methyl ester from respective gangliosides were generated by a specific ceramide glycanase (1 U/mg of ganglioside) as suggested by the manufacturer's instructions (Calbiochem) that catalyses removal of ceramide moieties from gangliosides, and their identity was confirmed by mass spectrometry through a core facility at Johns Hopkins University. Four anti-GD1a mAbs in this study were compared by using soluble GD1a-oligosaccharide in a soluble binding inhibition assay (ELISA). Two MoS anti-GD1a mAbs were similarly compared by using soluble GD1a methyl ester-oligosaccharide. Because NoS mAbs did not bind to GD1a methyl ester, they were not included in affinity studies with respective oligosaccharide. Different concentrations of soluble sugars (10–4 to 10–10 M) were used to determine the amount of oligosaccharide required for 50% inhibition of Ab binding to the corresponding ganglioside.
Fatty acid content of GD1a ganglioside and binding of mAbs
Generation of GD1a-derivatives with varied fatty acid length [GD1a-derivatives (FA)]
C22:0 and C24:0 fatty acids were added to GD1a as described previously (Tagawa et al., 2002). Briefly, bovine brain GD1a (500 nmol), sphingolipid Cer N-deacylase (SCDase; 25 mU) and sodium cholate (4 mg) were added to 5 ml of 50 mM sodium acetate, pH 6.0, in a screw-capped test tube. n-Decane (5 ml) was added and the solution was thoroughly mixed. The two-phase system was stirred for 24 h at 37°C, after which the phases were separated by brief centrifugation and the decane was discarded. The aqueous phase was extracted twice with 5 ml of n-decane and then vacuum-evaporated. The resulting lyso-GD1a was purified by sequential reverse-phase and normal-phase chromatographies. Re-N-acylation was via carbodiimide coupling. The same protocol was used to prepare GD1a-C22:0 and GD1a-C24:0. The respective fatty acid (6 mmol) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (6.6 µmol) were dissolved in 600 µl of tetrahydrofuran. After 2 h at 60°C lyso-GD1a (300 nmol in 420 µl of water) was added. The reaction was stirred for 10 h at 60°C, cooled to ambient temperature, then 50 µl of 6N NaOH were added and the reaction was incubated at ambient temperature for 16 h. Similarly, C16:0 (palmitic) and C18:0 (stearic) acids (2 µmol) and EDAC (2.2 µmol) were incubated in 100 µl of water at 50°C for 2 h, then lyso-GD1a (200 nmol in 100 µl of water) was added. The reaction mixture was stirred at 50°C for 10 h, cooled to ambient temperature, 33 µl of 6N NaOH were added and the mixture was incubated at ambient temperature for 16 h. The products were purified by reverse phase chromatographies.
Cell ELISA
NGCR72, a mutant cell line derived from NG108-15, which does not express complex a-series gangliosides including GD1a (Wu et al., 2001a, b), was used for reconstitution assays with GD1a-derivatives (FA) to avoid Ab-binding to endogenous GD1a. These cells were cultured according to previously described methods (Zhang et al., 2004) and cell phenotype was reconfirmed by the absence of cholera toxin subunit B staining before all experiments.
For reconstitution with GD1a-derivatives (FA) these cells were suspended in serum-free medium containing a concentration of soluble GD1a derivatives able to produce an equal loading at the plasma membrane (determined empirically): 2 µM solution for GD1a-derivatives containing C-22 or C-24 fatty acid and 0.33 and 0.67 µM, respectively, for derivatives carrying C-16 or C-18 fatty acids for 2 h in a microrotator at 37°C. Cells without reconstitution were used as controls.
After reconstitution, cells were centrifuged, washed with serum-free medium, plated onto 96-well plates in serum-free medium at 20 000 cells/well, and allowed to attach for 2 h. The cells were washed with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature. Non-specific binding was blocked with 2% normal goat serum (NGS) in PBS and these wells were incubated overnight with mAbs (5–10 µg/ml in PBS buffer containing 0.1% bovine serum albumin and 2% NGS). Antibody binding was detected with alkaline phosphatase-conjugated goat anti-mouse IgG (1 µg/ml for 90 min). The reaction was developed with 2 mg/ml of p-nitrophenyl phosphate (in 100 mM Tris, 100 mM NaCl, 5 mM MgCl2, pH 9.5) for 60 min and absorbance was measured at 405 nm with a microplate reader (Benchmark, Bio-Rad, Hercules, CA).
Anti-GD1a mAbs and ganglioside complexes
To address the issue whether MoS and NoS mAbs recognize ganglioside complexes ELISA were done on mixtures of GD1a and other major neural gangliosides. For these studies 100 pmol GD1a ganglioside alone or a combination of GD1a with gangliosides GM1 or GD1b or GT1b at ratio 75:25, 50:50 or 25:75 in methanol were adsorbed into each well. Then the reactivity of mAbs was tested as described above. Antibody binding was detected with alkaline phosphatase-conjugated goat anti-mouse IgG (Jackson Immunoresearch; 1 µg/ml for 90 min).
Computer modelling of GD1a conformers
The software Quanta (Accelrys Software Inc.) was used to model the different conformers of GD1a oligosaccharide portion as described (Acquotti et al., 1994). Figures of these computerized models were reproduced by using software PyMOL (DeLano Scientific LLC, CA, USA).
Analysis of anti-GD1a mAbs reactivity with sialidase-treated cauda equina sections
Rat cauda equina cryosections were pretreated with 100 mU/ml of sialidase at 37°C for 20, 40, 60, 90 min, and overnight before staining with GT1b-1 [specificity and immunostaining pattern as described (Gong et al., 2002)], GD1a-2a, GD1a-2b, GD1a-1 and GD1a-E6 (2–20 µg/ml; overnight incubations at 4°C) mAbs, as described above. These sections were developed with specific secondary antibodies conjugated to Cy3 (1 : 200, 1 h at ambient temperature) and examined by epifluorescence microscopy. FITC-conjugated isolectin B4 (IB4) (5 µg/ml; Sigma, St Louis, MO, USA) was used to mark Remak bundles in sensory roots in cauda equina, as described (Gong et al., 2002). Sialidase (V. cholerae) was obtained by overexpression in E. coli by using an expression plasmid (pET30 b(+)/VCNA) kindly provided by Dr G. Taylor, University of St Andrews, Fife, Scotland, and was purified as described (Moustafa et al., 2004). Quantification of area of immunoreactivity (
500 nerve fibre/condition) pre- and post-sialidase treatment was determined by ImageJ 1.34s software (NIH, USA).
|
Results |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsReferences |
|---|
Motor-selective (MoS) and non-selective (NoS) GD1a mAbs have distinct patterns of binding to GD1a-derivatives
TLC-immunooverlay showed that none of the mAbs recognized the amide form of GD1a. Likewise elimination of the glycerol tail of NeuAc abolished binding of all mAbs. GD1a-2a and GD1a-2b did not recognize any of derivatives carrying modifications in the carboxylic group despite use of 3-fold higher concentration of the derivatives. In contrast, GD1a-1 showed various degrees of reactivity with derivatives containing methyl and ethyl esters of the sialic acid carboxyl groups or replacement of the carboxylate with an alcohol (Fig. 2). These findings suggest that the carboxyl group is critical for differentiating between MoS and NoS mAbs as the glycerol chain is recognized by both group of mAbs.
|
Binding of AMAN sera to GD1a-derivatives
Figure 3 shows that all patient-derived anti-GD1a Abs showed some degree of binding to different chemical derivatives of GD1a. None of the sera matched the binding pattern of GD1a-2a or GD1a-2b mAbs. Similar to MoS mAbs, most of the binding was directed to the derivatives with substitutions of the carboxyl group (GD1a NeuAc 1-amide, GD1a NeuAc ethyl ester, GD1a NeuAc 1-alcohol and GD1a NeuAc 1-methyl ester) of NeuAc and 6 out of 10 patients had >30% reactivity with two or more of these derivatives compared to parent GD1a ganglioside. Most sera retained reactivity to GD1a-1-alcohol, 8 out of 10 sera had 20% or more (range 20–75%) binding to this derivative compared to parent GD1a. One patient's serum (98-10) had only minor reactivity with GD1a-1-alcohol and GD1a NeuAc 7-alcohol but not with other derivatives. The modifications in the glycerol chain of NeuAc variably affected the Ab-binding in AMAN sera; for example three sera had no binding to these derivatives, two sera had <20%, and five sera had 20% or more binding to GD1a NeuAc 7-alcohol and/or GD1a NeuAc 7-aldehyde. In contrast to MoS and NoS mAbs, anti-GD1a antibodies in some patient sera recognize the altered glycerol chain of NeuAc. These findings suggest that the binding patterns of sera contain a broader degree of reactivity because of the polyclonal/oligoclonal nature of the anti-GD1a antibody repertoire in patients. Overall, these results demonstrate that, similar to the MoS mAbs, anti-GD1a Abs in most AMAN sera did not require an intact underivatized carboxyl group in the NeuAc.
|
Characterization of a new motor-selective anti-GD1a mAb
GD1a-derivatives were used to screen a preexisting library of anti-GD1a clones. We selected a mAb (GD1a-E6) whose binding profile to GD1a derivatives was similar to that of GD1a-1 (minimal reactivity to GT1a
). Binding of this mAb to different gangliosides is summarized in Table 1. Immunostaining studies with this mAb showed that it preferentially stained ventral roots without significant binding to sensory fibres (Fig. 4). These results again emphasize that retention of binding to GD1a-derivatives with modifications in carboxyl groups correlates with motor-selective immunostaining.
|
Affinity of anti-GD1a mAbs
To examine whether mAb affinity could be a factor determining the differential binding to dorsal and ventral roots, affinity of anti-GD1a mAbs was analysed by soluble binding inhibition assays. Figure 5 shows the inhibition curves obtained by co-incubating individual mAbs with different concentrations of GD1a-oligosaccharide. mAbs GD1a-2a, GD1a-2b and GD1a-1 showed less than 50% inhibition with GD1a-oligosaccharide at the highest concentration tested (1 x 10–4 M). In contrast, GD1a-E6 was inhibited in a concentration-dependent manner with GD1a-oligosaccharide concentrations as low as 1 x 10–6 M. Affinity values are expressed as the concentration of GD1a-oligosaccharide that produced 50% inhibition of mAb binding (IC50) (Fig. 5). These studies indicate that the IC50 of both MoS and NoS mAbs (GD1a-1, GD1a-2a and GD1a-2b) was >1 x 10–4 M, but that these mAbs have diverse tissue-binding patterns. Furthermore, the higher affinity of a GD1a-selective mAb (GD1a-E6, IC50
2 x 10–6 M) did not diminish its motor selectivity.
|
Affinity of GD1a methyl ester-oligosaccharide was compared to GD1a-oligosaccharide because MoS mAbs have highest binding for this derivative on TLC analysis (shown above). The IC50 of GD1a-1 was >1 x 10–4 M and GD1a-E6 was
2 x 10–5.
Fatty acid content of GD1a ganglioside and mAb binding
Previous studies indicate that the length of the fatty acid could modulate exposure of the carbohydrate moiety of glycosphingolipids and modulate the anti-glycan antibody binding (Itonori et al., 1989; Tagawa et al., 2002; Stewart and Boggs, 1993a, b). Whether differences in ceramide fatty acid length could be the mechanism underlying the motor-selective binding was examined in studies of cell membrane reconstitution with GD1a-derivatives carrying fatty acids of different lengths. All anti-GD1a mAbs had higher binding to GD1a-derivatives carrying long chain fatty acids (Fig. 6). It is known that long chain fatty acids (C:22 and C:24) are predominant ceramide constituents in GD1a of dorsal root origin and that short chain fatty acids (C:16 and C:18) are predominant ceramide constituents in GD1a of ventral root origin. This suggests that fatty acid length is not the basis of preferential ventral root binding by MoS mAbs.
|
Ganglioside complexes and mAb binding
Antibodies against ganglioside complexes have been associated with different forms of GBS (Kaida et al., 2004
, 2007). To examine whether antibody binding to ganglioside complexes could provide a potential explanation for MoS mAb binding to ventral roots the following studies were conducted. GD1a ganglioside alone or a combination of GD1a with gangliosides GM1 or GD1b or GT1b at ratio 75 : 25, 50 : 50 or 25 : 75 were used as antigen. Figure 7 shows that both MoS (upper panel) and NoS (bottom panel) mAbs showed reduced reactivity to GD1a in the presence of other gangliosides. This reduced reactivity was correlated with the size of the carbohydrate head group of the ganglioside added to GD1a as well as the ratio GD1a/other ganglioside in the mixture. Thus, GT1b ganglioside with the largest carbohydrate head group was most efficient in inhibiting anti-GD1a mAb binding to the mixture of GD1a/GT1b. In contrast, GM1 ganglioside with the smallest carbohydrate head group showed almost no effect at ratio 75 : 25 or 50 : 50 and only a very modest negative effect on the reactivity when tested at a ratio of 25 : 75. This solid phase assay suggests that anti-GD1a mAbs used in this study do not recognize neo-epitopes in ganglioside mixtures, which could explain preferential ventral root staining by anti-GD1a MoS mAbs.
|
Modelling of GD1a conformers
To understand how the chemical modifications in the sialic acid on GD1a affect the binding of different GD1a-mAbs, we analysed the three-dimensional (3D) structures of different GD1a-derivatives, an approach that has been previously published (Lopez et al., 2002
). Previous 3D NMR studies showed that GD1a can adopt different configurations when incorporated in micelles (Acquotti et al., 1994). We modelled four different GD1a conformers in agreement with Acquotti et al. (1994) as described under Materials and Methods. The set of glycosidic torsion angles described by Acquotti et al. (1994) were preferentially used because micelles replicate the biophysical properties of the lipid environment in plasma membranes. We found no variation in the inner core GalNAc-(NeuAc)Gal-Glc in the four GD1a conformers. In contrast, they differed in the orientation of the terminal NeuAc-Gal residues. The differences in the orientation of the terminal NeuAc-Gal in these four conformers of GD1a can be explained by a gradual torsion of these residues toward the inner core of the ganglioside. We only used conformers 1 and 4 that represent the two extreme rotations of NeuAc-Gal residues for comparative modelling (Fig. 8A). Figure 8B depicts the structure of conformers 1 and 4 as seen from top view, which is opposite to the ceramide portion that anchors the ganglioside to the cell membrane, and represents the face of GD1a accessible to antibodies. Figure 8C illustrates the solvent-accessible surface of the top views of conformers 1 and 4. Since our solid phase studies with GD1a-derivatives established that the substitutions of the carboxyl group of the terminal NeuAc is key to discriminating MoS versus NoS mAbs, we analysed the orientation of this group in conformers 1 and 4. We observed that the carboxyl group remained above the ring core of NeuAc in conformer 1, whereas this group remained below the plane of the NeuAc ring in conformer 4 (Fig. 8C) and thus, the carboxyl group appears fully accessible to the Abs in conformer 1, whereas this group is tucked under the glycerol tail in conformer 4 and is not accessible to Abs. These modelling data suggest that the carboxyl group is not part of the epitope recognized by MoS mAbs, which thus preferentially recognize conformers 3 and 4 in which this group is not accessible to the mAbs. In contrast, the carboxyl group is a critical part of the epitope recognized by NoS mAbs and thus preferentially recognizes conformers 1 and 2 in which this group is easily accessible.
|
Sialidase treatment and binding pattern of MoS and NoS mAbs
The binding of MoS and NoS anti-GD1a mAbs to cauda equina sections was compared with mAb GT1b-1 at various intervals of sialidase treatment. Different durations of sialidase treatment were initially determined empirically with GD1a-1 and GT1b-1 mAbs. Our results indicate that cauda equina treatment with sialidase for 40 min at 37°C significantly decreased the binding of GT1b-1 mAb to GT1b (Fig. 9A and B). Notably, the binding of MoS mAb GD1a-1 to both motor and sensory nerves increased after pretreatment with sialidase for 40 min (Fig. 9C and D) but that sialidase treatment for 90 min (data not shown) reduced the GD1a-1 binding to both motor and sensory nerves compared to untreated sections or sections treated with sialidase for 40 min. GD1a-1 binding was completely abolished after overnight sialidase treatment (data not shown). The binding of both NoS mAbs increased to a variable degree in sensory and motor nerve fibers after sialidase treatment for 40 min. In comparison GD1a-E6 binding to motor but not to sensory nerve roots increased after sialidase treatment for 40 min. Table 2 summarizes the quantitative results for the four anti-GD1a mAbs.
|
|
Binding inhibition of mAb reactivity to GD1a ganglioside in situ by GD1a methyl ester oligosaccharide
These studies were done to determine whether GD1a derivatives compete with anti-GD1a mAb binding to GD1a in nerve fibers. Our results show that GD1a-E6 mAb binding to motor nerve fibres was almost completely abolished by pre- and co-incubation with GD1a methyl ester oligosaccharide (Fig. 10A and B). In contrast, GD1a-2b binding to motor or sensory nerve fibres was not affected by pre- and co-incubation with GD1a methyl ester oligosaccharide (Fig. 10C and D).
|
Binding of affinity purified anti-GD1a Abs from an AMAN serum to cauda equina
We found that affinity-purified anti-GD1a Abs from GBS patient (98-2) preferentially stained motor nerve fibres in rat cauda equina (Fig. 11). In contrast, to NoS mAbs the patient-derived Abs did not bind to unmyelinated sensory nerve fibers and there was faint staining of some myelinated sensory nerve fibres. Prolonged incubation with secondary antibodies was necessary to detect the Ab-binding probably due to the low amount of affinity-purified anti-GD1a Abs. The reactivity of this particular serum (98-2) with GD1a derivatives is quite similar to those of MoS mAbs (see above).
|
Discussion
Our results demonstrate that IgG anti-GD1a mAbs with MoS and NoS binding have distinct structural requirements on GD1a ganglioside. Anti-GD1a Abs from AMAN patients had structural requirements similar to those of motor-selective mAbs and preferentially stained motor axons. Computer modelling studies suggest that the presence or absence of motor selectivity of the mAbs is based on recognition of distinct GD1a conformers with different 3D orientations of the critical terminal NeuAc-Gal residues. Sialidase studies indicate that GD1a can be retrieved in the sensory nerve fibres and becomes accessible to one MoS mAb. This raises the possibility that specific interactions between endogenous GD1a in sensory nerve fibres and other molecules in the plane of the membrane may result in the stable expression of a GD1a conformer(s) that is inaccessible/cryptic to MoS anti-GD1a mAbs. These findings provide one explanation for the specificity of neural targeting in peripheral neuropathies associated with anti-ganglioside Abs. Other, as yet unidentified, factors may also be relevant to selective motor nerve injury in immune neuropathies such as AMAN.
The precise basis of predominant motor nerve injury in the AMAN form of GBS remains unclear. De Angelis et al. first reported that serum from an AMAN patient or purified IgGs with reactivity to GD1a and GalNAc-GD1a selectively stained nodes of Ranvier of motor fibres (De Angelis et al., 2001). Subsequently, we reported that an anti-GD1a mAb preferentially stained motor nerve fibres in humans and rodents (Gong et al., 2002). These results emphasized that the clinical phenotype in AMAN could in part be attributed to selective binding of motor nerve fibres by anti-ganglioside Abs present in patient sera. Since crossreactivity with naturally occurring (structurally related) gangliosides did not show major differences in the specificity among four anti-GD1a mAbs used in this study, we investigated the substructural requirements of antibody binding to its target ganglioside (epitope) by introducing chemical modifications in terminal NeuAc residues. Similar approaches have been used previously for characterization of lectin and autoantibody binding to glycosphingolipids (Ilyas et al., 1990; Schengrund et al., 1991; Collins et al., 1997). The data obtained from this biochemical approach were used to model the epitopes recognized by our mAbs. An important extension of the modelling data is that different GD1a conformers can exist in motor and sensory spinal roots. The validity of this biochemical and modelling approach was substantiated by cloning a new mAb (GD1a-E6) with preferential motor binding.
Our studies with AMAN sera indicate that the structural requirements (binding patterns) of patient-derived anti-GD1a Abs to GD1a ganglioside did not match those of NoS GD1a mAbs. Rather, the binding patterns of anti-GD1a Abs in AMAN sera to GD1a derivatives resembled binding of motor-specific mAbs. These binding patterns do not match exactly with motor-specific mAbs or with each other, which is not surprising, because the patient sera contain oligoclonal or polyclonal responses and are expected to have a wider repertoire of binding compared to mAbs (Lopez et al., 2001, 2002). Furthermore, patient sera are collected at different time points after the onset of disease, which may allow subpopulations of anti-GD1a antibodies to be preferentially cleared from the circulation because of their higher affinity for target antigen or IgG isotypes/subclasses [which affects their plasma half-life (Koga et al., 2003)]. Overall, these studies show that anti-GD1a Abs in most AMAN sera retain reactivity to GD1a derivatives with modifications in carboxyl group of NeuAc, a finding that resembles those with MoS anti-GD1a mAbs, which could result in preferential binding of patient anti-GD1a Abs to motor nerves. This hypothesis is supported by immunostaining studies with affinity-purified anti-GD1a Abs from a patient with AMAN.
Our previous studies (Gong et al., 2002) indicated that human sensory and motor spinal roots contain comparable quantities of GD1a, but that GD1a-1 mAb preferentially stained both human and rodent motor spinal roots. This established that differential expression of GD1a gangliosides in sensory versus motor roots is not the basis of selective motor recognition by GD1a mAbs. Furthermore, these previous solid phase and immunostaining studies (Gong et al., 2002) showed apparent paradoxical findings in that motor-selective mAb (GD1a-1) binds to both sensory and motor nerve-derived GD1a in solid phase assays, but not in immunocytochemistry studies. GD1a ganglioside in the dorsal roots is not inaccessible to all antibodies since sensory fibers are immunostained by NoS GD1a mAbs.
Our results suggest that endogenous GD1a in sensory nerve fibres is cryptic for MoS anti-GD1a mAbs and the basis of this crypticity is complex and not completely elucidated by our results. For example, ‘neo-GD1a’ created after salidase treatment is accessible to GD1a-1 but not to GD1a-E6. Generation of ‘neo-GD1a’ that is accessible to GD1a-1 in sensory nerve fibres after sialidase treatment could be due to: (i) conversion of endogenous GT1b to GD1a; (ii) elimination of interaction(s) between endogenous GD1a and other molecules carrying sialic acids in the vicinity of this ganglioside; (iii) unmasking of GD1a. Conversion of GT1b to GD1a is unlikely to be the sole explanation because the related MoS mAb GD1a-E6 does not bind to sensory nerve fibres after sialidase treatment despite its increased binding to motor nerve fibres. Given the modelling data, we favour the hypothesis that oligosaccharide moieties on GD1a ganglioside assume different 3D conformations in motor and sensory nerve fibres. The paradoxical behaviour of GD1a-E6 (in sensory nerve fibres) after sialidase treatment is not explained. We speculate that despite similar structural requirements for terminal sialic acid between GD1a-1 and GD1a-E6, these mAbs may differ in their substructural requirements for other portions of the terminal sialyl-Gal epitope. High affinity interactions between experimentally induced anti-GM Abs and their target antigens require larger contact area compared to low affinity interactions (GD1a-1) (Lopez et al., 2002) and this could be a potential explanation for discrepancy between GD1a-E6 (high affinity) and GD1a-1 (low affinity).
Antibody affinity is not the basis of selective motor recognition by GD1a mAbs, as indicated by determinations of antibody affinity showing that low (GD1a-1) and high (GD1a-E6) affinity mAbs retained their motor specificity. Furthermore, our reconstitution assays with GD1a containing fatty acids of different length in the ceramide indicate that fatty acid length per se is not the major determinant of preferential ventral root binding by motor-selective mAbs. We are aware that exogenous insertion of gangliosides could be qualitatively different from that of endogenous insertion, but such an approach has been widely used to study the influence of the ceramide portion of glycosphingolipids on their localization and recognition in model membranes (Stewart and Boggs, 1993a, b; Singh et al., 1995; Jones et al., 1997). We examined the ganglioside complex hypothesis (Kaida et al., 2004, 2006, 2007) by analysing the reactivity of MoS and NoS mAbs in solid phase assays in the presence of different gangliosides at a wide range of ratios, and found that binding of the mAbs used in the current study was not enhanced in the presence of ganglioside mixtures.
In conclusion, our data support the hypothesis that fine specificity of the antibody and corresponding conformation of the target antigen is one factor that contributes to the pure motor phenotype observed in patients with AMAN and high titres of IgG anti-GD1a antibodies. Analysis of structural requirements of target recognition by anti-ganglioside antibodies could provide insight into the mechanisms of target selectivity and clinical phenotype in patients with GBS.
|
Acknowledgements |
|---|
This work was supported by grants from the National Institute of Health (NS42888, NS54962 and NS37096) and the GBS Foundation. We thank Dr Pamela Talalay for editorial discussion and Dr John Griffin for helpful suggestions and advice. ‘Under a licensing agreement between Seikagaku America and the Johns Hopkins University, Drs Sheikh and Schnaar are entitled to a share of royalty received by the University on sales of products described in this article. The terms of this arrangement are being managed by the Johns Hopkins University in accordance with its conflict of interest policies’.
|
References |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsReferences |
|---|
Acquotti D, Cantu L, Ragg E, Sonnino S. Geometrical and conformational properties of ganglioside Galnac-G(D1A), Iv(4)Galnaciv(3)Neu5Acii(3)Neu5Acggose(4)Cer. Eur J Biochem (1994) 225:271–88.[Web of Science][Medline]
Bullens RW, O’Hanlon GM, Wagner E, Molenaar PC, Furukawa K, Furukawa K, et al. Complex gangliosides at the neuromuscular junction are membrane receptors for autoantibodies and botulinum neurotoxin but redundant for normal synaptic function. J Neurosci (2002) 22:6876–84.
Collins BE, Kiso M, Hasegawa A, Tropak MB, Roder JC, Crocker PR, et al. Binding specificities of the sialoadhesin family of I-type lectins. Sialic acid linkage and substructure requirements for binding of myelin- associated glycoprotein, Schwann cell myelin protein, and sialoadhesin. J Biol Chem (1997) 272:16889–95.
De Angelis MV, Di Muzio A, Lupo S, Gambi D, Uncini A, Lugaresi A. Anti-GD1a antibodies from an acute motor axonal neuropathy patient selectively bind to motor nerve fiber nodes of Ranvier. J Neuroimmunol (2001) 121:79–82.[CrossRef][Web of Science][Medline]
Dell A, Reason AJ, Khoo KH, Panico M, Mcdowell RA, Morris HR. Mass-spectrometry of carbohydrate-containing biopolymers. Guide Tech Glycobiol (1994) 230:108–32.[CrossRef]
Gong Y, Tagawa Y, Lunn MP, Laroy W, Heffer-Lauc M, Li CY, et al. Localization of major gangliosides in the PNS: implications for immune neuropathies. Brain (2002) 125:2491–506.
Goodfellow JA, Bowes T, Sheikh K, Odaka M, Halstead SK, Humphreys PD, et al. Overexpression of GD1a ganglioside sensitizes motor nerve terminals to anti-GD1a antibody-mediated injury in a model of acute motor axonal neuropathy. J Neurosci (2005) 25:1620–8.
Goodyear CS, O’Hanlon GM, Plomp JJ, Wagner ER, Morrison I, Veitch J, et al. Monoclonal antibodies raised against Guillain-Barre syndrome- associated Campylobacter jejuni lipopolysaccharides react with neuronal gangliosides and paralyze muscle-nerve preparations. J Clin Invest (1999) 104:697–708. [published erratum appears in J Clin Invest 1999; 104: 1771].[Web of Science][Medline]
Handa S, Nakamura K. Modification of sialic-acid carboxyl group of ganglioside. J Biochem (1984) 95:1323–9.
Hirabayashi Y, Suzuki T, Suzuki Y, Taki T, Matsumoto M, Higashi H, et al. A new method for purification of anti-glycosphingolipid antibody – avian anti-hematoside (Neugc) antibody. J Biochem (1983) 94:327–30.
Hughes RA, Cornblath DR. Guillain-Barre syndrome. Lancet (2005) 366:1653–66.[CrossRef][Web of Science][Medline]
Ilyas AA, Chou DKH, Jungalwala FB, Costello C, Quarles RH. Variability in the structural requirements for binding of human monoclonal anti-myelin-associated glycoprotein immunoglobulin-M antibodies and Hnk-1 to sphingoglycolipid antigens. J Neurochem (1990) 55:594–601.[Web of Science][Medline]
Itonori S, Hidari K, Sanai Y, Taniguchi M, Nagai Y. Involvement of the acyl chain of ceramide in carbohydrate recognition by an anti-glycolipid monoclonal antibody: the case of an anti-melanoma antibody, M2590, to GM3-ganglioside. Glycoconj J (1989) 6:551–60.[CrossRef][Web of Science][Medline]
Jones DH, Lingwood CA, Barber KR, Grant CWM. Globoside as a membrane receptor: a consideration of oligosaccharide communication with the hydrophobic domain. Biochemistry (1997) 36:8539–47.[CrossRef][Web of Science][Medline]
Kaida K, Kanzaki M, Morita D, Kamakura K, Motoyoshi K, Hirakawa M, et al. Anti-ganglioside complex antibodies in Miller Fisher syndrome. J Neurol Neurosurg Psychiatry (2006) 77:1043–6.
Kaida K, Morita D, Kanzaki M, Kamakura K, Motoyoshi K, Hirakawa M, et al. Ganglioside complexes as new target antigens in Guillain-Barre syndrome. Ann Neurol (2004) 56:567–71.[CrossRef][Web of Science][Medline]
Kaida K, Morita D, Kanzaki M, Kamakura K, Motoyoshi K, Hirakawa M, et al. Anti-ganglioside complex antibodies associated with severe disability in GBS. J Neuroimmunol (2007) 182:212–8.[CrossRef][Web of Science][Medline]
Kawai H, Allende ML, Wada R, Kono M, Sango K, Deng C, et al. Mice expressing only monosialoganglioside GM3 exhibit lethal audiogenic seizures. J Biol Chem (2001) 276:6885–8.
Koga M, Yuki N, Hirata K, Morimatsu M, Mori M, Kuwabara S. Anti-GM1 antibody IgG subclass: a clinical recovery predictor in Guillain-Barre syndrome. Neurology (2003) 60:1514–8.
Kusunoki S, Shimizu J, Chiba A, Ugawa Y, Hitoshi S, Kanazawa I. Experimental sensory neuropathy induced by sensitization with ganglioside GD1b. Ann Neurol (1996) 39:424–31.[CrossRef][Web of Science][Medline]
Lehmann HC, Lopez PHH, Zhang G, Ngyuen T, Zhang JY, Kieseier BC, et al. Passive immunization with anti-ganglioside antibodies directly inhibits axon regeneration in an animal model. J Neurosci (2007) 27:27–34.
Lopez PH, Irazoqui FJ, Nores GA. Normal human plasma contains antibodies that specifically block neuropathy-associated human anti-GM1 IgG-antibodies. J Neuroimmunol (2000) 105:179–83.[CrossRef][Web of Science][Medline]
Lopez PH, Lardone RD, Irazoqui FJ, Villa AM, Di Egidio M, Saizar RD, et al. Variable patterns of anti-GM(1) IgM-antibody populations defined by affinity and fine specificity in patients with motor syndromes: evidence for their random origin. J Neuroimmunol (2001) 119:131–6.[CrossRef][Web of Science][Medline]
Lopez PH, Villa AM, Sica RE, Nores GA. High affinity as a disease determinant factor in anti-GM(1) antibodies: comparative characterization of experimentally induced vs. disease- associated antibodies. J Neuroimmunol (2002) 128:69–76.[CrossRef][Web of Science][Medline]
Lunn MP, Johnson LA, Fromholt SE, Itonori S, Huang J, Vyas AA, et al. High-affinity anti-ganglioside IgG antibodies raised in complex ganglioside knockout mice: reexamination of GD1a immunolocalization. J Neurochem (2000) 75:404–12.[CrossRef][Web of Science][Medline]
Moustafa I, Connaris H, Taylor M, Zaitsev V, Wilson JC, Kiefel MJ, et al. Sialic acid recognition by Vibrio cholerae neuraminidase. J Biol Chem (2004) 279:40819–26.
Schengrund CL, DasGupta BR, Ringler NJ. Binding of botulinum and tetanus neurotoxins to ganglioside GT1B and derivatives thereof. J Neurochem (1991) 57:1024–32.[CrossRef][Web of Science][Medline]
Schnaar RL. Isolation of glycosphingolipids. Methods Enzymol (1994) 230:348–70.[Web of Science][Medline]
Schnaar RL, Fromholt SE, Gong Y, Vyas AA, Laroy W, Wayman DM, et al. IgG-class mouse monoclonal antibodies to major brain gangliosides. Anal Biochem (2002) 302:276–84.[CrossRef][Web of Science][Medline]
Sheikh KA, Deerinck TJ, Ellisman MH, Griffin JW. The distribution of ganglioside-like moieties in peripheral nerves. Brain (1999a) 122:449–60.
Sheikh KA, Sun J, Lui Y, Kawai H, Crawford TO, Proia RL, et al. Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc Natl Acad Sci USA (1999b) 96:7532–7.
Sheikh KA, Zhang G, Gong Y, Schnaar RL, Griffin JW. An anti-ganglioside antibody-secreting hybridoma induces neuropathy in mice. Ann Neurol (2004) 56:228–39.[CrossRef][Web of Science][Medline]
Singh DM, Shan X, Davis JH, Jones DH, Grant CWM. Oligosaccharide behavior of complex natural glycosphingolipids in multicomponent model membranes. Biochemistry (1995) 34:451–63.[CrossRef][Web of Science][Medline]
Spiegel S, Ravid A, Wilchek M. Involvement of gangliosides in lymphocyte stimulation. Proc Natl Acad Sci USA (1979) 76:5277–81.
Stewart RJ, Boggs JM. A carbohydrate carbohydrate interaction between galactosylceramide-containing liposomes and cerebroside sulfate-containing liposomes: dependence on the glycolipid ceramide composition. Biochemistry (1993a) 32:10666–74.[CrossRef][Web of Science][Medline]
Stewart RJ, Boggs JM. Exposure of galactosylceramide to galactose-oxidase in liposomes:dependence on lipid environment and ceramide composition. Biochemistry (1993b) 32:5605–14.[CrossRef][Web of Science][Medline]
Svennerholm L. Designation and schematic structure of gangliosides and allied glycosphingolipids. Prog Brain Res (1994) 101:XI–XIV.[Medline]
Svennerholm L, Bosträm K, Fredman P, Jungbjer B, Lekman A, Månsson J-E, et al. Gangliosides and allied glycosphingolipids in human peripheral nerve and spinal cord. Biochim Biophys Acta (1994) 1214:115–123.[Medline]
Svennerholm L, Bostrom K, Fredman P, Jungbjer B, Mansson JE, Rynmark BM. Membrane lipids of human peripheral nerve and spinal cord. Biochim Biophys Acta (1992) 1128:1–7.[Medline]
Tagawa Y, Laroy W, Nimrichter L, Fromholt SE, Moser AB, Moser HW, et al. Anti-ganglioside antibodies bind with enhanced affinity to gangliosides containing very long chain fatty acids. Neurochem Res (2002) 27:847–55.[CrossRef][Web of Science][Medline]
Vyas AA, Patel HV, Fromholt SE, Heffer-Lauc M, Vyas KA, Dang J, et al. Gangliosides are functional nerve cell ligands for myelin-associated glycoprotein (MAG), an inhibitor of nerve regeneration. Proc Natl Acad Sci USA (2002) 99:8412–7.
Willison HJ, Yuki N. Peripheral neuropathies and anti-glycolipid antibodies. Brain (2002) 125:2591–625.
Wu G, Lu ZH, Xie X, Ledeen R. Comparison of ganglioside profiles in nuclei and whole cells of NG108-15 and NG-CR72 lines: changes in response to different neuritogenic stimuli. Brain Res Dev Brain Res (2001a) 126:183–90.[CrossRef][Medline]
Wu G, Lu ZH, Xie X, Li L, Ledeen RW. Mutant NG108-15 cells (NG-CR72) deficient in GM1 synthase respond aberrantly to axonogenic stimuli and are vulnerable to calcium-induced apoptosis: they are rescued with LIGA-20. J Neurochem (2001b) 76:690–702.[CrossRef][Web of Science][Medline]
Yu RK, Saito M. Structure and localization of gangliosides. In: Neurobiology of glycoconjugates—Margolis RU, Margolis RK, eds. (1989) New York, NY: Plenum Publishing Corporation. 1–42.
Yuki N, Susuki K, Koga M, Nishimoto Y, Odaka M, Hirata K, et al. Carbohydrate mimicry between human ganglioside GM1 and Campylobacter jejuni lipooligosaccharide causes Guillain-Barre syndrome. Proc Natl Acad Sci USA (2004) 101:11404–9.
Yuki N, Yamada M, Koga M, Odaka M, Susuki K, Tagawa Y, et al. Animal model of axonal Guillain-Barre syndrome induced by sensitization with GM1 ganglioside. Ann Neurol (2001) 49:712–20.[CrossRef][Web of Science][Medline]
Zhang G, Lopez PH, Li CY, Mehta NR, Griffin JW, Schnaar RL, et al. Anti-ganglioside antibody-mediated neuronal cytotoxicity and its protection by intravenous immunoglobulin: implications for immune neuropathies. Brain (2004) 127:1085–100.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Kaida, T. Ariga, and R. K Yu Antiganglioside antibodies and their pathophysiological effects on Guillain-Barre syndrome and related disorders--A review Glycobiology, July 1, 2009; 19(7): 676 - 692. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||