Animal model of axonal Guillain-Barr syndrome induced by sensitization with GM1 ganglioside.код для вставкиСкачать
Animal Model of Axonal Guillain-Barré Syndrome Induced by Sensitization with GM1 Ganglioside Nobuhiro Yuki, MD, PhD,1 Mitsunori Yamada, MD, PhD,2 Michiaki Koga, MD, PhD,1 Masaaki Odaka, MD, PhD,1 Keiichiro Susuki, MD,1 Yumi Tagawa, MD, PhD,1 Shuichi Ueda, MD, PhD,3 Takeshi Kasama,4 Akio Ohnishi, MD, PhD,5 Shintaro Hayashi, MD, PhD,2 Hitoshi Takahashi, MD, PhD,2 Mikiko Kamijo, MD, PhD,6 and Koichi Hirata, MD, PhD1 Some humans develop the axonal form of Guillain-Barré syndrome after receiving bovine brain ganglioside. On sensitization with the ganglioside mixture, all of a group of rabbits injected developed high anti-GM1 IgG antibody titers, flaccid limb weakness of acute onset, and a monophasic illness course. Pathological findings for the peripheral nerves showed predominant Wallerian-like degeneration, with neither lymphocytic infiltration nor demyelination. IgG was deposited on the axons of the anterior roots, and GM1 was proved to be present on the axons of peripheral nerves. Sensitization with purified GM1 also induced axonal neuropathy, indicating that GM1 was the immunogen in the mixture. A model of human axonal Guillain-Barré syndrome has been established that uses inoculation with a bovine brain ganglioside mixture or isolated GM1. This model may help to clarify the molecular pathogenesis of the syndrome and to develop new treatments for it. Ann Neurol 2001;49:712–720 The presence of axonal Guillain-Barré syndrome (GBS) has been established.1– 4 Cumulative evidence supports the speculation that gangliosides, highly expressed molecules in nerve tissues, are the target antigens of the autoantibodies in axonal GBS. First, patients frequently have IgG antibodies to several gangliosides during the acute phase of axonal GBS.5–7 Second, some patients develop axonal GBS subsequent to Campylobacter jejuni enteritis, and the C. jejuni strains isolated from these patients have lipopolysaccharides bearing ganglioside-like structure.8,9 Third, gangliosides extracted from bovine brain have been used to treat various neurological disorders, and reports of patients developing axonal GBS after ganglioside administration have been recorded.10,11 There are only two reports describing neurological dysfunction in GM1-immunized animals. In the first study, GM1-immunized rabbits developed a spastic paralysis, whereas GD1a-immunized rabbits had a flaccid paralysis.12 Histologically, phagocytic cells containing myelin debris could be observed. In the other study, rabbits developed a subclinical neuropathy.13 There was mild axonal degeneration in the sciatic nerve and IgM deposits at the nodes of Ranvier. These findings have failed confirmation in rodents. In contrast, the sensitization of rabbits with GD1b did induce sensory neuropathy associated with anti-GD1b antibody.14 These suggested that failure to induce neuropathy by sensitization with gangliosides might depend on species susceptibility and on the immunization procedure used. In this study, therefore, rabbits were inoculated with a bovine brain ganglioside (BBG) mixture according to the procedure of Kusunoki et al.14 The rabbits developed acute motor axonal neuropathy associated with anti-GM1 IgG antibody. Experimental motor axonal neuropathy also was induced by sensitization with GM1. From the Departments of 1Neurology, 3Histology and Neurobiology, Dokkyo University School of Medicine, Tochigi; 2Department of Pathology, Brain Research Institute, Niigata University, Niigata; 4 Instrumental Analysis Research Center for Life Science, Tokyo Medical and Dental University, Tokyo; 5Department of Neurology, School of Medicine, University of Occupational and Environmental Health, Fukuoka; and 6Department of Anatomy, Aichi Medical University, Aichi, Japan. Received Sep 19, 2000, and in revised form Jan 22, 2001. Accepted for publication Jan 22, 2001. 712 © 2001 Wiley-Liss, Inc. Materials and Methods Immunization Procedure Male Japanese white rabbits ( JW/CSK), weighing 2.0 –2.5 kg, were obtained from SLC (Hamamatsu Japan). A 2.5 mg portion of a BBG mixture (GM1 21%, GD1a 40%, GD1b 16%, GT1b 19%; CronassialTM; Fidia, Padova, Italy) or 0.5 Published online 23 March 2001. Address correspondence to Dr Yuki, Department of Neurology, Dokkyo University School of Medicine, Kitakobayashi 880, Mibu, Shimotsuga, Tochigi 321-0293, Japan. E-mail: firstname.lastname@example.org mg of GM1 isolated from bovine brain (SygenTM; Fidia) was dissolved in 0.5 ml of keyhole limpet hemocyanin (KLH; 2 mg/ml; Sigma, St. Louis, MO) in phosphate-buffered saline (PBS). A 0.5 ml portion of Freund’s complete adjuvant (FCA; Sigma) was added and the mixture emulsified. A 1 ml sample of the emulsion of the BBG mixture or GM1 was injected subcutaneously to the back and intraperitoneally at 3 week intervals until limb weakness developed or 6 months had passed after the first inoculation. Control rabbits were injected under the same protocol with the same inoculums but without gangliosides. The rabbits were checked daily for clinical signs and weighed twice per week. Weekly plasma samples were taken by ear vein puncture. All the animal procedures conformed to the Dokkyo University School of Medicine institutional guidelines. Pathological Studies Rabbits were deeply anesthetized by an intraperitoneal injection of sodium pentobarbital and the left sciatic nerves removed from some of them for teased fiber analysis. Nerve specimens were immersed directly in 3% glutaraldehyde for 10 minutes and stained with 1% osmium tetroxide, then immersed in 60% glycerol, stored in 100% glycerol, and examined after teasing. At least 100 teased myelinated fibers were prepared for each sciatic nerve. The teased fiber preparations were blindly assessed by one of the authors (A.O.). Differences in frequency distribution of Wallerian-like degeneration were statistically examined with the Mann-Whitney U test. After nerve removal, the rabbits were perfused transcardially with 0.4 liters/kg of heparinized PBS (10 IU/ml) followed by 1.2 liters/kg of 3% glutaraldehyde-1% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. After this perfusion, the right sciatic nerve, spinal cord with anterior and posterior nerve roots, and brain were removed and immersed in the same fixative. Multiple blocks of tissue from the brain and spinal cord were embedded in paraffin, after which 4 m thick sections were cut and stained with hematoxylin and eosin or Klüver-Barrera. The sciatic nerves and lumbar segment of the spinal cord with anterior and posterior nerve roots were all postfixed with 1% osmium tetroxide, dehydrated in a graded ethanol series, then embedded in Epon 812 (Polysciences, Warrington, PA). Sections 1 m thick were cut from each sample and stained with 0.5% toluidine blue and 1% safranine. The image analysis program MacscopeTM (Mitani, Fukui, Japan) was used for morphological examination of the sciatic nerves. Digital photo images were obtained with a light microscope with a final magnification of ⫻400 and an attached CCD camera interfaced with a computer. The computerassisted image analysis allowed for the determination of total fascicular area and density of degenerative or surviving axons. The number of these axons was counted manually from every three flame areas (single flame, 0.03/mm2) in the three largest fascicles. Degenerative axon profiles and macrophages were counted according to the criteria of Midroni and Bilbao.15 The profile of the surviving axons was characterized as round myelin profiles that enclosed a transcular space. This evaluation for the determination of surviving axons has been done by other investigators.16 The morphometric analysis was performed by one of the authors (M.Ka.), who was unaware of the clinical information. Data were analyzed by a one-way analysis of variance (ANOVA) with appropriate Bonferroni-corrected posthoc Student’s t-test comparisons. Immunohistochemical Studies The rabbits were deeply anesthetized, perfused transcardially with 1 liter/kg of PBS, pH 7.4, and then with 1 liter/kg of 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. After perfusion, the sciatic nerves and spinal nerve roots were removed. The samples were postfixed on ice for 30 minutes with the same fixative, then cryoprotected in 20% sucrose solution at 4°C overnight, after which they were frozen in isopentane at –70°C and then sectioned by cryostat. Sections 20 m thick were dried on gelatin-coated glass slides and then incubated at 20°C for 2 hours with peroxidaseconjugated protein G (2 g/ml; Sigma). Next, they were incubated for 5 minutes at 20°C in a solution of 50 mM Trishydrochloric acid buffer, pH 7.4, containing 0.02% 3,3⬘diaminobenzidine tetrahydrochloride (DAB) and 0.006% hydrogen peroxide. The reaction was terminated by washing the slides with PBS. Treatment with 0.05% osmium tetroxide in distilled water for 2 minutes was done to intensify the DAB reaction products. As described elsewhere,17 the sciatic nerves and spinal nerve roots of normal rabbits were incubated at 4°C overnight with peroxidase-conjugated cholera toxin B-subunit (2 g/ml; List Biological Laboratories, Campbell, CA), then developed with DAB, followed by osmium tetroxide. Sciatic nerves were removed from deeply anesthetized normal rabbits, quickly frozen in cold isopentane, and stored at – 80°C. Cryostat sections (6 m thick) from these nerves were fixed with cold acetone at –20°C for 7 minutes. After a wash with PBS, the sections were incubated successively at 20°C for 15 minutes with 10% normal goat serum, at 4°C overnight with plasma from a BBG-immunized rabbit (1:100 dilutions), and at 20°C for 1 hour with peroxidaseconjugated goat anti-rabbit IgG (1:200; Southern Biotechnology, Birmingham, AL). After each step, the sections were washed thoroughly with PBS. Finally, the immunoreaction products were developed with DAB solution. Thin-Layer Chromatography (TLC) Immunostaining and TLC Blotting/Secondary Ion Mass Spectrometry Total gangliosides were extracted from the peripheral nerves of normal rabbits. The monosialosyl-, disialosyl-, and polysialosylganglioside fractions were prepared as described elsewhere.18 The BBG mixture; authentic GM1; or the monosialosyl-, disialosyl-, and polysialosylgangliosides from rabbit peripheral nerves were layered on TLC plates (Merk, Darmstadt, Germany). The plates were developed with chloroform/methanol/12 mM magnesium chloride in water (5: 4:1, by volume), dipped in n-hexane-containing 0.4% polyisobutylmethacrylate for 1 minute, then dried under an air stream. Each plate was overlaid with rabbit plasma (1:50 dilutions with PBS-0.5% casein) and kept at 4°C overnight. Next, the plates were washed and overlaid with peroxidaseconjugated anti-rabbit - or ␥-chain-specific antibodies (1: 500 dilutions; Nordic, Tilburg, The Netherlands), kept at 20°C for 2 hours, then washed. Binding activities were made Yuki et al: Animal Model of Axonal GBS 713 visible with 4-chloro-1-naphtol. After marking the immunostainings with a 6B pencil, the TLC plates were dipped in chloroform/methanol (1:1 by volume) to remove the visible product and polymer. TLC blotting and negative secondary ion mass spectrometry were done as described elsewhere.19 After being coated with the polymer, the plates were incubated at 20°C for 2 hours with peroxidase-conjugated cholera toxin B-subunit (0.2 g/ml). Binding was made visible with 4-chloro-1-naphtol. Enzyme-Linked Immunosorbent Assay The enzyme-linked immunosorbent assay (ELISA) was done as reported elsewhere.20 Five picomoles of GM2, GM1, GD1a, GD1b, GT1b, or GQ1b were placed in individual wells of microtiter plates. Plasma samples were diluted serially starting at 1:500. The diluted plasma was added to each well, and the plates were incubated at 4°C overnight. Peroxidase-conjugated anti-rabbit - or ␥-chain-specific antibody (1:2,000) was added, after which the plates were kept at 20°C for 2 hours, then developed with o-phenylenediamine. Results Clinical and Pathological Findings for BBGImmunized Rabbits None of the ten control rabbits inoculated with KLH and FCA showed limb weakness until 6 months after the first inoculation. In contrast, all of the 13 rabbits immunized with BBG, KLH, and FCA developed flaccid paresis of the hind limbs, the onset ranging from 35 to 57 days (median 43 days) after the initial inoculation (Table 1, Fig 1a,b). Eleven of the rabbits began to lose weight 3–20 days (median 6 days) before the onset of limb weakness. Rabbits Cr-2, Cr-3, and Cr-5 had body and limb tremors several days before the onset of limb weakness. Tetraparesis developed rapidly in rabbits Cr-3 and Cr-4, with respiration becoming labored and gasping. Limb weakness progressed for 4 –13 days (median 5 days) after onset in 9 rabbits (Cr-5 to Cr-13) then reached a plateau. One rabbit (Cr-6) could lift neither its head nor body but was able to walk 2 weeks after the onset of limb weakness. No significant changes were found in the brains or spinal cords of any of the rabbits immunized with BBG. In contrast, the sciatic nerves showed severe to mild Wallerian-like degeneration, with rare demyelination or remyelination (Fig 1c). Macrophage invasion was prominent in endoneurial-perivascular areas, but no lymphocytic infiltration was found in any of the sciatic nerve regions. Quantitative comparisons showed that the mean degenerative axon density was significantly increased in 11 BBG-immunized rabbits [Table 1; 1,289 ⫾ 476/mm2 (mean ⫾ standard deviation)] than in 4 adjuvant controls (498 ⫾ 152/mm2; p ⫽ Table 1. Rabbits Inoclulated with Bovine Brain Gangliosides or GM1 Quantitative Analysis of Sciatic Nerves Rabbit Immunogen Inoculation Times Cr-1 Cr-2 Cr-3 Cr-4 Cr-5 Cr-6 Cr-7 Cr-8 Cr-9 Cr-10 Cr-11 Cr-12 Cr-13 Sy-1 Sy-2 Sy-3 Sy-4 Sy-5 Sy-6 Sy-7 Sy-8 Sy-9 Sy-10 Sy-11 BBG BBG BBG BBG BBG BBG BBG BBG BBG BBG BBG BBG BBG GM1 GM1 GM1 GM1 GM1 GM1 GM1 GM1 GM1 GM1 GM1 2 3 3 2 2 2 3 3 3 3 3 3 3 2 9 3 9 2 5 5 9 4 4 4 Onset of Limb Weakness (Day) Days from Onset to Peak 42 48 45 38 40 35 54 43 48 48 57 43 43 37 Not clarified Not clarified Undetermined Undetermined 4 6 13 4 5 4 5 5 5 Not clarified 62 Not clarified 41 90 90 177 70 70 72 7 2 Not clarified 6 27 10 2 Poorest Functional Grade Days from Onset to Sacrifice 1 1 4 4 4 4 2 4 4 3 4 4 4 1 0 1 0 2 2 1 2 3 5 5 52 53 2 4 29 31 32 9 8 11 13 20 16 29 11 9 119 81 43 35 Not done Not done Degenerative Axon Density (n/mm2) Surviving Axon Density (n/mm2) 1,446 795 8,266 7,462 952 1,127 1,455 1,163 730 10,355 8,692 7,673 7,264 15,916 1,428 1,399 2,481 1,271 336 394 1,238 536 1,122 1,115 1,091 800 802 10,306 10,084 9,091 10,456 12,340 10,988 10,712 10,152 11,646 11,923 14,446 9,528 13,717 Functional grade: 0 ⫽ normal; 1 ⫽ showing weakness of the hind limbs; 2 ⫽ showing mild weakness of the four limbs but able to walk; 3 ⫽ showing moderate weakness of the four limbs and unable to walk; 4 ⫽ showing severe weakness of the four limbs, which were spread out; 5 ⫽ dead. Because rabbits Cr-1, Cr-2, Sy-1, Sy-3, and Sy-7 had only weakness of the hind limbs, the days from onset to peak were not clarified. Sy-2 and Sy-4 were killed, respectively, 285 and 292 days after the first inoculation. Because rabbits Cr-3 and Cr-4 were killed, respectively, 2 and 4 days after the onset of limb weakness, the days from onset to peak could not be determined. Sciatic nerves of Cr-3 and Cr-9 were not available for the morphometric examination. 714 Annals of Neurology Vol 49 No 6 June 2001 Fig 1. (a) Rabbit Cr-6 (see Table 1) with limb weakness induced by sensitization with bovine brain ganglioside (BBG) mixture; 14 days after onset. It could not maintain a normal standing position nor lift its head or body. The muscles of the extremities and trunk were weak and slack, offering less than the usual resistance to passive movement. (b) Rabbit Cr-4 with limb weakness induced by sensitization with the BBG mixture; 4 days after onset. It lay splayed out, all its extremities extended and head on the floor, instead of sitting in the usual compact, hunched posture. This rabbit attempted to stand but could not. (c) Transverse section of the sciatic nerve from BBG-immunized rabbit Cr-6. Many myelin ovoids produced by Wallerian-like degeneration of the myelinated fibers are present. Toluidine blue-safranine stain. (d) Transverse section of the anterior root from BBG-immunized rabbit Cr-6. Clusters of small myelinated fibers indicative of regenerating sprouted fibers are present (arrowheads). Toluidine bluesafranine stain. (e) Myelin ovoids indicative of Wallerian-like degeneration of myelinated fibers from BBG-immunized rabbit Cr-11 are present in three sciatic nerve teased fiber preparations. (f ) Transverse section of the spinal anterior nerve root from a BBGimmunized rabbit with high anti-GM1 IgG antibody titer. Some axons are deeply stained by peroxidase-conjugated protein G (arrowheads). Scale bars ⫽ 10 m in c,d,f, 20 m in e. 0.0024). The surviving axon density was significantly decreased in the BBG-immunized rabbits (9,597 ⫾ 2,419/mm2) than in the adjuvant controls (14,971 ⫾ 3,131/mm2; p ⫽ 0.0006). The anterior spinal nerve roots from rabbit Cr-6 in the recovery phase of weakness had occasional clusters of small fibers showing various degrees of myelination, indicative of axonal sprout regeneration (Fig 1d). No pathological changes were seen in the posterior roots. Teased fiber studies showed a significantly higher percentage of fibers that had undergone Wallerian-like degeneration in the BBGimmunized rabbits (12% in Cr-11, 21% in Cr-12, and 24% in Cr-13; Fig 1e) than in 3 adjuvant controls (0%, 0%, and 3%; p ⫽ 0.046). In contrast, the paralyzed rabbits had low percentages of fibers that had undergone paranodal demyelination (1% in Cr-11, 0% in Cr-12, and 0% in Cr-13) as did the controls (0%, 1%, and 2%). Protein G positively stained some axons in Yuki et al: Animal Model of Axonal GBS 715 spinal anterior roots obtained from BBG-immunized rabbits (Fig 1f ), evidence of the presence of IgG in or around the axons. Similar findings were obtained when anterior roots were immunostained with anti-rabbit IgG antibody. No pathological changes or immunoglobulin deposits were detected in the sciatic nerves or anterior roots of the normal rabbits and adjuvant controls. Induction of Antiganglioside Antibody Plasma obtained within 1 week after the onset of limb weakness was used. TLC immunostaining showed that the rabbit IgGs strongly bound to GM1 (Fig 2a). An ELISA confirmed that the immune response was directed predominantly against GM1 (Table 2). Neither the IgMs nor the IgGs obtained from these rabbits before inoculation nor those from the control rabbits bound to GM1. Anti-GM1 IgM antibodies were detected in the 13 paralyzed rabbits 2–3 weeks after their first sensitization with BBG (Table 3). Titers increased for 1–5 weeks (median 2 weeks) after detection (range 2,000 – 64,000; median 32,000). Anti-GM1 IgG antibodies were detected in these rabbits 3– 4 weeks after the first inoculation. The titers increased for 1– 6 weeks (median 3 weeks) after detection (range 4,000 –512,000; median 32,000). Ten of the thirteen developed flaccid paresis of the hind limbs within 1–3 weeks (median 1 week) after the peak anti-GM1 IgG titer was reached. Axonal Neuropathy Induced by Sensitization with GM1 Nine of eleven rabbits immunized with GM1, KLH, and FCA developed flaccid paresis of the hind limbs (Table 1). Quadriparesis and respiratory paresis rapidly progressed in rabbits Sy-10 and Sy-11, which died, respectively, 11 and 3 days after the onset of limb weakness. Limb weakness progressed for 2, 6, 7, and 27 days, respectively, after onset in rabbits Sy-6, Sy-8, Sy-5, and Sy-9, then reached a plateau. Weakness lessened in 2 rabbits (Sy-6 and Sy-7) 11 and 20 days after onset and disappeared in 2 (Sy-7 and Sy-8) 5 and 29 days after onset. TLC immunostaining and the ELISA showed that anti-GM1 IgM antibodies were induced and that a switch from IgM to IgG occurred in all 11 rabbits with or without development of paralysis. IgG from a paralyzed rabbit reacted with the GM1 of rabbit peripheral nerve (Fig 2b). Mild to moderate Wallerian-like degeneration was present in Eponembedded sciatic nerve sections obtained from rabbits with limb weakness (Table 1); the degenerative axon density was increased (929 ⫾ 310/mm2). There was mild to extensive fiber loss; the surviving axon density was decreased (12,044 ⫾ 1,680/mm2). Neither lymphocytic infiltration nor demyelination was found. Teased fiber preparations showed Wallerian-like degeneration in 23%, 21%, and 26% of the large myelinated fibers in Sy-6, Sy-8, and Sy-9, respectively, whereas paranodal demyelination was seen in 0%, 0%, and 2% of these fibers. Target Molecule for the IgG Autoantibody Immunoblotting showed that none of the IgGs from the paralyzed rabbits bound to proteins from rabbit peripheral nerves (data not shown). In contrast, TLC immunostaining showed that the IgG strongly reacted with a band in the monosialosylganglioside fraction from rabbit peripheral nerve (Fig 2b). The mobility of the band was similar to that of authentic GM1. Figure 2c shows the TLC blotting/negative secondary ion mass spectrum of the IgG-reactive band. On the TLC plates, cholera toxin B-subunit, which specifically recognizes GM1, reacted with a band in the monosialosylganglioside fraction (Fig 2b), which had mobility similar to that of the IgG-reactive band. These findings suggest that GM1 is present in rabbit peripheral nerves. The cholera toxin B-subunit bound to the axons of normal rabbit sciatic nerves (Fig 2d) and anterior and posterior roots, evidence that GM1 is expressed on those axons. Immunohistochemical experiments with plasma from a BBG-immunized rabbit that had high anti-GM1 activity clearly showed that its plasma IgG reacted positively with the axon surfaces of sciatic nerves from a normal rabbit (Fig 2e,f ). Discussion The clinical features of paresis with an acute, monophasic course in BBG-immunized rabbits were very similar to those of patients with GBS. Axonal GBS has predominant axonal involvement, characterized by Wallerian-like degeneration of nerve fibers, with only minimal demyelination and minimal lymphocytic infiltration.1– 4 Our paralyzed rabbits showed similar pathological changes with predominant axonal degeneration of the peripheral nerves. Results of teased fiber studies confirmed a higher percentage of Wallerian-like degeneration of the large myelinated fibers of the sciatic nerves. In addition, there was regenerative sprouting of anterior root fibers taken from a rabbit during the recovery phase. Anti-GM1 IgG antibody specifically is present in GBS during the acute phase and is significantly associated with axonal GBS.5,6 The paralyzed rabbits also had the anti-GM1 IgG antibody. As in rodents and humans,17 GM1 is expressed on the axons of rabbit peripheral nerves, as our study indicates. We showed that the IgG with antiGM1 activity from a paralyzed rabbit strongly bound to the axon. Moreover, as in Chinese patients with axonal GBS,21 IgG was deposited on the motor nerve axons in the paralyzed rabbits. The cholera toxin 716 Annals of Neurology Vol 49 No 6 June 2001 Yuki et al: Animal Model of Axonal GBS 717 Table 2. Antiganglioside Antibodies in Plasma Obtained within 1 Week after the Onset of Limb Weakness Rabbits Sensitized with Bovine Brain Ganglioside Cr-1 Cr-2 Cr-3 Cr-4 Cr-5 Cr-6 Cr-7 Cr-8 Cr-9 Cr-10 Cr-11 Cr-12 Cr-13 IgM antibody to GM2 0 0 0 0 0 0 0 0 0 0 0 0 0 GM1 1,000 8,000 8,000 8,000 32,000 32,000 8,000 16,000 64,000 64,000 1,000 4,000 16,000 GD1a 0 500 500 2,000 2,000 16,000 500 0 4,000 4,000 0 500 500 GD1b 0 2,000 2,000 4,000 16,000 16,000 4,000 8,000 16,000 16,000 0 1,000 2,000 GT1b 0 0 0 0 0 0 0 0 0 0 0 0 0 GQ1b 0 0 0 0 0 0 0 0 0 0 0 0 0 IgG antibody to GM2 0 0 0 0 0 0 0 0 0 0 0 0 0 GM1 4,000 4,000 32,000 128,000 32,000 128,000 32,000 16,000 64,000 32,000 16,000 4,000 4,000 GD1a 0 0 1,000 0 1,000 0 500 0 8,000 500 0 0 0 GD1b 0 0 4,000 4,000 4,000 32,000 1,000 500 32,000 1,000 500 500 500 GT1b 0 0 0 0 0 0 0 0 0 0 0 0 0 GQ1b 0 0 0 0 0 0 0 0 0 0 0 0 0 bound with both anterior and posterior roots, indicative of the presence of GM1 on sensory fibers as well as motor fibers. Why neither the IgG deposition nor the Wallerian-like degeneration was seen in the posterior roots is unknown. Motor nerve terminals are proposed to be the target site of anti-GM1 antibody in acute motor axonal neuropathy.10,22 The pathology of the terminals should be investigated in this animal model. The pathophysiological role of anti-GM1 antibodies is still a matter of controversy. In some studies antiGM1 sera caused conduction block of the motor Š nerves,23,24 whereas in others it did not.25,26 Immunohistochemical findings in our study, however, provide evidence that IgG antibodies directed at GM1 on motor nerve axons could cause motor axonal neuropathy. We showed that sensitization with GM1 highly induced acute motor axonal neuropathy similar to that induced by the BBG mixture injection, strong support that GM1 is the immunogen in the BBG mixture and the target molecule for the autoantibodies. GM1 and galactocerebroside are expressed in both the peripheral nervous system (PNS) and central nervous system Fig 2. (a) Antiganglioside antibody from rabbits that developed limb weakness after sensitization with a bovine brain ganglioside (BBG) mixture. A: Thin-layer chromatogram stained with the orcinol reagent for hexose. B: Immunostained chromatogram that had been overlaid first with plasma from the rabbits, then with peroxidase-conjugated anti-rabbit ␥-chain specific antibodies. Lanes 1–10 show plasma from rabbits Cr-1 to Cr-10 in Table 1. Orcinol reagent stains GM1, GD1a, GD1b, and GT1b. Plasma IgGs from the rabbits strongly bind to GM1, and some weakly react with GD1b. (b) Target molecule for IgG autoantibodies among peripheral nerve gangliosides. A: Thin-layer chromatogram stained with the orcinol reagent. B: Immunostained chromatogram incubated first with plasma from rabbit Cr-6 inoculated with BBG then with peroxidase-conjugated anti-rabbit ␥-chain specific antibodies. C: Binding of the peroxidase-conjugated cholera toxin B-subunit. D: Immunostained chromatogram incubated first with plasma from rabbit Sy-5 inoculated with GM1, then with peroxidase-conjugated anti-rabbit ␥-chain specific antibodies. Lane 1: BBG mixture (CronassialTM). Lane 2: Authentic GM1 (SygenTM) from bovine brain. Lanes 3–5: Monosialosyl-, disialosyl-, and polysialosylganglioside fractions, respectively, from rabbit peripheral nerves. The mobility of the monosialosylganglioside, with which the cholera toxin B-subunit and IgGs from rabbits Cr-6 and Sy-5 react, is similar to that of authentic GM1. (c) Negative secondary ion mass spectrum of the monosialosylganglioside that reacted with IgG from rabbit Cr-6. Cer ⫽ ceramide; Hex ⫽ hexose; HexNAc ⫽ N-acetylhexosamine; NeuAc ⫽ N-acetylneuraminic acid. The ion at m/z 1,544 consisted of stearic acid (C18:0) and sphingenine (d18:1), and that at m/z 1,572 of C18:0 and icosasphingenine (d20:1). The 7 major fragment ions of the sugar sequence ions in the spectrum were representative ceramides (m/z 564, 592; out of spectrum), glucosylceramides (a; m/z 726, 754), lactosylceramides (b; m/z 888, 916), gangliotriaosylceramides (c; m/z 1,091, 1,119), gangliotetraosylceramides (d; m/z 1,253, 1,281), II 3-N-acetylneuraminosyllactosylceramides (e; m/z 1,179, 1,207), and II 3-N-acetylneuraminosylgangliotriaosylceramides (f; m/z 1,382, 1,410). The two fragment ions that corresponded to the nonreducing terminal side of the carbohydrate chain were m/z 833 (g; [(Hex-HexNAc-[NeuAc]Hex-OH)-H2-H] –) and m/z 995 (h; [(Hex-Hex-HexNAc-[NeuAc]Hex-OH)-H2-H] –). The m/z 290 and 308 ions, which corresponded to N-acetylneuraminic acid, also were present. Its spectrum and its mobility on the thinlayer chromatogram plate similar to that of authentic GM1 and its binding with the cholera toxin B-subunit suggest that the structure of this ganglioside is Gal ␤1–3 GalNAc ␤1– 4 (NeuAc ␣2–3) Gal ␤1– 4 Glc ␤1–1⬘ Cer. (d)Cross section of normal rabbit sciatic nerve stained with the peroxidase-conjugated cholera toxin B-subunit. Axons are stained. (e,f ) Cross (e) and longitudinal (f ) sections of normal rabbit sciatic nerve immunostained with plasma IgG that has anti-GM1 activity from the BBG-immunized rabbit Cr-4. Axon surfaces are positively stained. Scale bars ⫽ 10 m. 718 Annals of Neurology Vol 49 No 6 June 2001 Table 3. Anti-GM1 Antibody Titers in the Rabbits Sensitized with Bovine Brain Ganglioside Rabbit Isotype 1 Cr-1 IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG IgM IgG Cr-2 Cr-3 Cr-4 Cr-5 Cr-6 Cr-7 Cr-8 Cr-9 Cr-10 Cr-11 Cr-12 Cr-13 Week(s) After the First Inoculation 2 3 4 5 6 7 8 9 10 11 12 13 14 0 0 1,000 2,000 2,000 2,000 1,000 500 500 500 1,000 1,000 1,000 0 0 1,000 4,000 4,000 4,000 4,000 4,000 4,000 4,000 4,000 4,000 4,000 0 0 8,000 8,000 8,000 8,000 8,000 16,000 16,000 16,000 16,000 16,000 16,000 16,000 0 0 0 4,000 4,000 4,000 4,000 8,000 16,000 32,000 32,000 32,000 32,000 32,000 0 0 1,000 8,000 8,000 8,000 8,000 0 0 1,000 8,000 16,000 32,000 16,000 0 2,000 32,000 32,000 16,000 8,000 0 0 1,000 32,000 128,000 128,000 0 0 16,000 32,000 64,000 32,000 32,000 64,000 64,000 32,000 0 0 8,000 32,000 64,000 32,000 32,000 64,000 64,000 64,000 0 500 16,000 16,000 32,000 32,000 16,000 16,000 8,000 0 0 8,000 32,000 64,000 128,000 256,000 512,000 256,000 0 0 500 4,000 16,000 16,000 16,000 8,000 4,000 2,000 0 0 0 4,000 32,000 32,000 32,000 32,000 32,000 32,000 0 0 1,000 8,000 16,000 32,000 16,000 0 0 0 500 4,000 16,000 16,000 0 500 8,000 32,000 32,000 64,000 64,000 0 0 1,000 8,000 32,000 32,000 64,000 0 2,000 32,000 64,000 64,000 64,000 64,000 32,000 0 0 500 8,000 16,000 32,000 32,000 32,000 0 0 500 2,000 2,000 2,000 2,000 2,000 1,000 1,000 0 0 0 500 2,000 2,000 4,000 16,000 16,000 16,000 0 0 4,000 16,000 16,000 16,000 4,000 1,000 1,000 0 0 0 4,000 4,000 8,000 4,000 8,000 8,000 0 0 4,000 16,000 32,000 16,000 16,000 4,000 2,000 0 0 0 500 1,000 2,000 4,000 4,000 1,000 (CNS), but sensitization with these molecules produces only peripheral neuropathy.27 The blood–nerve barrier that guards the PNS is not as tight as the blood–brain barrier; therefore, small amounts of circulating IgG, which cannot enter the CNS, can penetrate the endoneurial space. This relative leakiness may make the PNS, especially the roots and nerve terminals, more vulnerable than constituents of the CNS to IgG antibody-mediated disorders. We assume that inoculation with the BBG mixture or the isolated GM1 induced high anti-GM1 IgG antibody titers, after which the autoantibodies bound to the GM1 expressed on the axons in the rabbit PNS, producing dysfunction of the motor nerves followed by Wallerian-like degeneration. Studies on BBG or purified GM1 as the agent for treating a variety of neurological disorders were initially reported with enthusiasm to be successful, but later double-blind controlled studies have failed to confirm these findings.28 CronassialTM (BBG mixture) and SygenTM (isolated GM1), which had been available on the Italian market, were used in our study. Although whether there is an epidemiological relationship between exogenous gangliosides and GBS is still not clear,29,30 our results indicate that in some cases a BBG or GM1 injection may trigger axonal GBS. As in the patients with axonal GBS after BBG therapy,10,31 in- terestingly, the immune response in the BBGinoculated rabbits predominantly was directed against GM1. The BBG mixture contains more GD1a than GM1, but it is unknown why the immune response against GD1a was poor in the rabbits. All the antiGD1b antibody titers were lower than the anti-GM1 antibody titers (Table 2). The anti-GD1b antibodies in the rabbits were absorbed with GM1 (data not shown), indicating that the antibodies cross-react with GM1. In conclusion, experimental inoculation of rabbits with BBG or GM1 provides a new and useful model for determination of the molecular pathogenesis of axonal GBS. This model should be useful for developing new treatments for patients with GBS. References 1. Feasby TE, Gilbert JJ, Brown WF, et al. An acute axonal form of Guillain-Barré polyneuropathy. Brain 1986;109:1115–1126. 2. McKhann GM, Cornblath DR, Griffin JW, et al. Acute motor axonal neuropathy: a frequent cause of acute flaccid paralysis in China. Ann Neurol 1993;33:333–342. 3. Griffin JW, Li CY, Ho TW, et al. Guillain-Barré syndrome in northern China: the spectrum of neuropathological changes in clinically defined cases. Brain 1995;118:577–595. 4. Griffin JW, Li CY, Ho TW, et al. Pathology of the motorsensory axonal Guillain-Barré syndrome. Ann Neurol 1996;39: 17–28. 5. Kuwabara S, Yuki N, Koga M, et al. IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degen- Yuki et al: Animal Model of Axonal GBS 719 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 720 eration in Guillain-Barré syndrome. Ann Neurol 1998;44:202– 208. Hadden RD, Cornblath DR, Hughes RAC, et al. Electrophysiological classification of Guillain-Barré syndrome: clinical associations and outcome. Ann Neurol 1998;44:780 –788. Ho TW, Willison HJ, Nachamkin I, et al. Anti-GD1a antibody is associated with axonal but not demyelinating forms of Guillain-Barré syndrome. Ann Neurol 1999;45:168 –173. Yuki N, Taki T, Inagaki F, et al. A bacterium lipopolysaccharide that elicits Guillain-Barré syndrome has a GM1 ganglioside-like structure. J Exp Med 1993;178:1771–1775. Aspinall GO, McDonald AG, Pang H, et al. Lipopolysaccharides of Campylobacter jejuni serotype O:19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barré syndrome. Biochemistry 1994;33:241–249. Illa I, Ortiz N, Gallard E, et al. Acute axonal Guillain-Barré syndrome with IgG antibodies against motor axons following parenteral gangliosides. Ann Neurol 1995;38:218 –224. Landi G, D’Alessandro R, Dossi BC, et al. Guillain-Barré syndrome after exogenous gangliosides in Italy. Br Med J 1993; 307:1463–1464. Nagai Y, Momoi T, Saito M, et al. Ganglioside syndrome, a new autoimmune neurologic disorder, experimentally induced with brain gangliosides. Neurosci Lett 1976;2:107–111. Thomas FP, Trojaborg W, Nagy C, et al. Experimental autoimmune neuropathy with anti-GM1 antibodies and immunoglobulin deposits at the nodes of Ranvier. Acta Neuropathol 1991;82:378 –383. Kusunoki S, Shimizu J, Chiba A, et al. Experimental sensory neuropathy induced by sensitization with ganglioside GD1b. Ann Neurol 1996;39:424 – 431. Midroni G, Bilbao JM. The axon: normal structure and pathological alterations. In: Midroni G, Bilbao JM, editors. Biopsy diagnosis of peripheral neuropathy. Boston: ButterworthHeinemann, 1995:45–74. Crawford TO, Hsieh S-T, Schryer BL, Glass JD. Prolonged axonal survival in transected nerves of C57BL/Ola mice is independent of age. J Neurocytol 1995;24:333–340. Sheikh KA, Deerinck TJ, Ellisman MH, Griffin JW. The distribution of ganglioside-like moieties in peripheral nerves. Brain 1999;122:449 – 460. Yoshino H, Ariga T, Latov N, et al. Fucosyl-GM1 in human sensory nervous tissue is a target antigen in patients with autoimmune neuropathies. J Neurochem 1993;61:658 – 663. Kasama T, Hisano Y, Nakajima M, et al. Microscale analysis of Annals of Neurology Vol 49 No 6 June 2001 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. glycosphingolipids by TLC blotting/secondary ion mass spectrometry: a novel blood group A-active glycosphingolipid and changes in glycosphingolipid expression in rat mammary tumour cells with different metastatic potentials. Glycoconj J 1996;13:461– 469. Yuki N, Tagawa Y, Irie F, et al. Close association of GuillainBarré syndrome with antibodies to minor monosialogangliosides GM1b and GM1␣. J Neuroimmunol 1997;74:30 –34. Hafer-Macko C, Hsieh ST, Li CY, et al. Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann Neurol 1996;40:635– 644. Ho TW, Hsieh S-T, Nachamkin I, et al. Motor nerve terminal degeneration provides a potential mechanism for rapid recovery in acute motor axonal neuropathy after Campylobacter infection. Neurology 1997;48:717–724. Santoro M, Uncini A, Corbo M, et al. Experimental conduction block induced by serum from a patient with anti-GM1 antibodies. Ann Neurol 1992;31:385–390. Takigawa T, Yasuda H, Kikkawa R, et al. Antibodies against GM1 ganglioside affect K⫹ and Na⫹ currents in isolated rat myelinated nerve fibers. Ann Neurol 1995;37:436 – 442. Hirota N, Kaji R, Bostock H, et al. The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 1997;120:2159 – 2169. Paparounas K, O’Hanlon GM, O’Leary CP, et al. Antiganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro. Brain 1999;122:807– 816. Saida T, Saida K, Dorfman SH, et al. Experimental allergic neuritis induced by sensitization with galactocerebroside. Science 1979;204:1103–1106. Bradley WG. Critical review of gangliosides and thyrotropinreleasing hormone in peripheral neuromuscular diseases. Muscle Nerve 1990;13:833– 842. Govoni V, Granieri E, Tola MR, et al. Exogenous gangliosides and Guillain-Barré syndrome: an observational study in the local health district of Ferrara, Italy. Brain 1997;120:1123–1130. Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology. Risk factors for Guillain-Barré syndrome: a population-based case-control study. J Neurol 1999;246:1004 –1009. Odaka M, Yuki N, Nobile-Orazio E, et al. Antibodies to GM1(NeuGc) in Guillain-Barré syndrome after ganglioside therapy. J Neurol Sci 2000;175:96 –106.