Abstract
During 1 year, we used immunocytochemical staining of human cerebellum to screen 1,488 serums for IgG autoantibodies to Hu and Yo antigens.Three serums had none of the classically described patterns of IgG binding but, instead, selectively stained the cerebellar molecular layer. Evaluation of clinical data showed that the patients had either typical Miller Fisher syndrome (MFS) or Guillain-Barre syndrome with ophthalmoplegia. Further analysis by ELISA assay showed that all three serums had high titers of IgG anti-GQ1b autoantibodies. IgG autoantibody staining of human cerebellum, which is used for the diagnosis of paraneoplastic disorders, may have additional specificity for other, presumably autoimmune, syndromes such as MFS. The specificity of the serum IgG autoantibody binding to the cerebellar molecular layer may be related to the ataxia that often occurs in these patients.
NEUROLOGY 1996;47: 1317-1320
Between May 1993 and May 1994, 1,488 serums from different patients referred to the Neuromuscular Clinical Laboratory at Washington University, St. Louis, for measurement of autoantibody titers to a variety of antigens were tested for their ability to immunostain normal postmortem human cerebellum. Three serums had an unusual pattern of staining, with selective binding of IgG to the cerebellar molecular layer. We questioned whether this pattern of staining correlated with a particular clinical syndrome. Review of the clinical records showed that all three patients had acute neuropathy syndromes, either Miller Fisher syndrome (MFS) or Guillain-Barre syndrome (GBS) with ophthalmoplegia and ataxia.
Methods.
Serums.
Serums from 1,488 different patients were frozen at -70 degrees C until evaluation or tested within 1 week of collection. Follow-up specimens from two of the three patients with serum IgG staining of the cerebellar molecular layer were also studied. Diagnoses from patients at our own institutions included Guillain-Barre syndrome without extraocular movement disorders or ataxia (19), polyneuropathy (93), myopathies (24), and amyotrophic lateral sclerosis (23). Other than the three positive serums, no other patients with Miller-Fisher-like syndromes were studied from our institutions.
Clinical data.
All three patients with serum IgG that stained the cerebellar molecular layer were personally examined by one of the authors. Follow-up clinical data were available for all three patients. "Early" serums were obtained within 2 weeks of the onset of symptoms.
ELISA assays.
Serum was assayed for antibodies to GQ1b, GT1b, GM1, GD1a, and sulfatide using ELISA methodology. GQ1b was obtained from Bachem Fine Chemicals (Torrance, CA) and GT1b, GM1, GD1a, and sulfatide from Sigma Chemical Company (St. Louis, MO). Substrates were attached to wells of microtiter plates by adding 150 ng of GQ1b, GT1b, GM1, and GD1a, or 400 ng of sulfatide in 50 micro liter of methanol, and evaporated to dryness. Any remaining binding sites were blocked with 100 micro liter of 1% normal goat serum (NGS) in phosphate-buffered saline (PBS). Subsequent steps were performed as previously described. [1] Titers of antibodies were calculated from optical density (OD) data by extrapolating readings to the OD that might be expected at a standard dilution of 1:100 and multiplied by 1,000. A serum antibody with a titer of x was detectable (>0.05 OD units over negative controls) up to a dilution of at least 1/x. On the basis of prior testing, titers of serum IgG binding to GQ1b or GT1b ganglioside that were greater than 1:600 were considered to be high. This binding was considered to be selective when the titer of IgG anti-sulfatide antibodies was less than 20% of the anti-GQ1b or anti-GT1b antibodies. [1] All positive results were repeated, separately and in batches.
Cerebellar immunostaining.
Human cerebellar tissue was obtained at autopsy after postmortem intervals of between 6 and 24 hours from patients without any history of neurologic disorders. The tissue was stored frozen at -70 degrees C. Eight-micro meter thick cryostat sections of frozen autopsied human cerebellum, spinal cord, or peripheral nerve were placed on gelatin-covered slides and allowed to air dry for 30 to 60 minutes. The slides were used immediately or frozen at -70 degrees C. Just prior to use, the sections were fixed in acetone for 10 minutes at -20 degrees C and then airdried for 2 to 3 minutes. Nonspecific binding was blocked with 100% NGS for 2 hours at room temperature. Patient serum was added at a dilution of 1:100 in 10% NGS in PBS, and incubated in a humidity chamber in a cold room at 4 degrees C for 24 hours. The slide was then washed three times with PBS, with each wash for 5 minutes. Endogenous peroxidase activity was blocked by incubating the sections in methanol with 0.3% hydrogen peroxide (H2 O2) for 10 minutes. Specific goat anti-human IgG linked to horseradish peroxidase (Organon Teknika-Cappel, West Chester, PA) at a dilution of 1:200 in 10% NGS in PBS was then added to the tissue sections and incubated for a minimum of 3 hours at room temperature in the humidity chamber. The sections were again washed three times in PBS and then developed in 3,3-diaminobenzidine tetrahydrochloride (DAB) solution (100 ml of PBS, 50 mg of DAB, and 33 micro liter of hydrogen peroxide [0.01%]) for 2 to 3 minutes in the dark. After development, the reaction was stopped by washing in deionized water. The sections were then dehydrated in increasing concentrations of alcohol, cleared with at least two washes of xylene, and mounted with Permount (Fisher Chemical). The positive staining results on human cerebellum were repeated on at least three occasions using different tissue specimens.
Case report.
A 68-year-old man (patient 1) presented with a 3-day history of double vision, blurred vision, and numbness in his palms, legs, and face. Over the next 2 days he noted increasing unsteadiness of gait associated with nasality of speech and nasal regurgitation. Approximately 2 weeks prior to the onset of symptoms, he had an illness with fever, chills, and diarrhea. General physical examination was normal. He was dysarthric with prominent nasality to his voice. Cranial nerve examination revealed a partial right abducens palsy, bilateral facial diplegia, poor palatal movement with phonation or on the gag reflex, and mild neck flexor weakness. Strength testing was normal in all four limbs. Tendon reflexes were present and normal in all limbs. Sensation was intact except for diminished vibratory sensation in the hands and feet. Finger/nose and heel/shin testing was dysmetric. The gait was ataxic.
Laboratory investigations, including electrolytes, complete blood count, and MRI of the head, were normal. The initial CSF examination showed 20 lymphocytes and normal protein. Subsequent CSF examinations showed an increased protein content. IgG anti-GQ1b antibody titers were elevated Table 1. Serology was consistent with recent Campylobacter jejuni infection.
Table 1. Clinical, ELISA and immunocytochemical data
Nerve conduction studies, 5 days after the onset of symptoms, showed borderline slowing of motor and sensory nerve conduction velocities and low-normal amplitudes of motor and sensory nerve action potentials. Electromyography was normal. Blink reflex testing showed an absence of the R2 response bilaterally. A follow-up electrodiagnostic examination approximately 1 month after the onset of the illness was unchanged.
The patient underwent five courses of plasma exchange. At discharge, the patient's ataxia had improved but his bulbar symptoms persisted, necessitating nasogastric tube placement for feeding. On follow-up 9 months after his presentation, the ataxia and cranial nerve dysfunction had resolved.
Results.
Cerebellar immunostaining Figure 1, Table 1.
(Figure 1, Table 1) IgG in serum from all three patients selectively stained the molecular layer of human cerebellum. This pattern was only found when we tested serum obtained from these patients early in the disease course. The molecular-layer staining was not present in the convalescent serum from two patients obtained 1 to 2 months after the onset of disease. Neither the early nor convalescent serums stained human spinal cord or peripheral nerve.
Figure 1. Immunostaining of human cerebellum by serum IgG from patient 1. (A) Early in the disease course when IgG versus GQ1b was present in high titer. (B) Late in the disease course after specific titers of IgG vs GQ1b had disappeared. Note the selective staining of the cerebellar molecular layer by IgG from the early sample that did not occur using the serum obtained during the convalescent period. IgG from the late serum sample stained nuclei in all layers of the cerebellum, a nonspecific finding. Bar = 215 micro meter (A), 245 micro meter (B).
Antibody assays Table 1.
(Table 1) High titer, specific IgG reactivity to GQ1b and GT1b gangliosides was found in all three patients. There was no specific serum IgM or IgG binding to GM1, asialo-GM1, GD1a ganglioside, or sulfatide, and no IgM binding to GD1b ganglioside or myelin-associated glycoprotein. In the two patients with follow-up serology 1 to 2 months later, specific IgG binding to these gangliosides had disappeared.
Clinical data Table 1.
(Table 1) All three patients presented with a clinical syndrome that included ataxia and ophthalmoplegia and were subsequently diagnosed as having MFS, or a GBS variant with prominent ataxia and ophthalmoplegia. Laboratory investigations in all patients were considered to be suggestive of an acute immune-mediated radiculoneuropathy. All had elevated CSF protein without cells (albumino-cytologic dissociation). Electrophysiological testing in two patients was normal except for abnormalities in blink reflexes. In the third patient electrical testing was consistent with a demyelinative polyneuropathy. All patients underwent plasma exchange. On follow-up all recovered to normal neurologic function.
Discussion.
MFS is manifested clinically by the triad of gait ataxia, external ophthalmoplegia, and areflexia. [2,3] It is uncommon, occurring in about 5 to 10% of cases in GBS series. [3] The occurrence of high titers of serum IgG reactivity to GQ1b ganglioside in most cases of MFS [4-6] is evidence for an immune pathogenesis in this disorder. Clinical evidence, including high titers of serum anti-GQ1b antibodies early in the course of the disease, falling titers with clinical improvement, and the high specificity of these antibodies for MFS, suggests that the antibodies could be pathogenic. [4-8] More direct experimental evidence includes the neurotransmitter release being inhibited in vitro by serums containing IgG anti-GQ1b antibodies from patients with MFS syndromes. [8] The abundance of gangliosides, such as GQ1b, in neuronal membranes, and the extracellular location of their carbohydrate epitopes, [7,9] probably make them available as targets of autoantibodies. Pathologic studies have shown binding of monoclonal anti-GQ1b antibodies to the paranodal region of oculomotor nerves in patients with MFS. [6] The binding of the GQ1b antibodies to the oculomotor nerve, where small amounts of GQ1b are present, [10] may explain some of the clinical features but not other symptoms such as the ataxia.
Debate continues regarding whether the ataxia in MFS is of central or peripheral origin. [11,12] Our results showing IgG binding to the cerebellar molecular layer provide further support for the possibility of a central origin of the ataxia in MFS and similar syndromes. Other evidence supporting a central origin includes CNS abnormalities on MRI, [13] electrodiagnostic testing, [12,13] and the presence of anti-GQ1b antibodies in Bickerstaff encephalitis, [14] a syndrome that has clinical similarities to MFS. [15]
One study correlated the presence of anti-GQ1b antibodies with the ataxia in MFS. [4] However, histochemical studies of rat cerebellum have shown that the predominant localization of GQ1b is in the granular layer and glomeruli, rather than in the molecular layer. [16] This apparent discordance may be explained by the cross-reaction of anti-GQ1b antibodies with structurally related gangliosides such as GT1b, [6] which is abundant in the molecular layer [16] of the cerebellum.
Whether or not the anti-cerebellar autoantibodies are pathogenic, the pattern of binding seems specific enough that it may have diagnostic utility. Study of additional patients is necessary to determine the strength of the correlation between the pattern of cerebellar molecular layer staining observed in our patients, binding to gangliosides such as GQ1b and GT1b, and specific clinical signs.
- Copyright 1996 by Advanstar Communications Inc.
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.