Western blots of brain homogenates showing protein antigens recognized by serum and cerebrospinal fluid samples from patients 1,2 and 3, affected by stiff-man syndrome and breast cancer (lane 1,2 and 3) and by serum samples from control patients and normal subjects. The sera from the three patients affected by affected by stiff-man syndrome and breast cancer did not react with the Gad band, which was recognized by a human serum positive for GAD antibodies (lane 4). Lane 5, serum from a control patient with the stiff-man syndrome who did not have GAD antibodies; and lane 6 serum of a normal subject. A band with an apparent molecular mass of 128 kd was recognized by the serum samples of patient 1, 2 and 3. Lane 7 and 8 were both loaded with 20 ug of protein from a homogenate of a rat brain and were labelled with serum and cerebrospinal fluid, respectively, from patient 3. (Reprinted with permission from F. Folli et al, Autoantibodies to a 128 kd protein in three women with the stiff-man syndrome and breast cancer. N Engl J Med 1993; 328: 546-551. Copyright 1993 Massachusetts Medical Society. All rights reserved).

Photomicrographs of Sections of Rat Cerebellum and Brain Stem Stained with Serum Samples (Diluted 1:20) from Patients 1, 2, and 3 affected by stiff-man syndrome and breast cancer, and a Control Patient affected by stiff-man syndrome with autoantibodies against glutamic acid decarboxilase and with Rabbit Antibodies to Synaptic-Vesicle Proteins.

In Panels A through I, immunoperoxidase staining was used (immunoreactive areas are black), and in Panel J and Panel K, immunofluorescence staining was used (immunoreactive areas are white). In Panel A, a sagittal section of cerebellum stained with serum from Patient 1 shows staining of all regions containing synapses in the cerebellar cortex and the deep cerebellar nuclei (DCN). The dense staining of the molecular layer (ML) and the patchy staining of the granular layer (GL) correspond to the known distribution of synapses in the two layers.

In Panel B, cerebellar cortex stained with serum from Patient 3 shows widespread immunoreactivity on synapses.

In Panel C, cerebellar cortex stained with serum from a patient with the stiff-man syndrome that contained glutamic acid decarboxylase antibodies shows only a few immunoreactive synapses, corresponding to aminobutyric acid-ergic synapses. These include synapses made by basket-cell axons around the axon hillocks of Purkinje cells (arrow).

In Panel D, Panel E, and Panel F, adjacent sections of cerebellar cortex stained with serum samples from Patients 1, 2, and 3, respectively, have identical patterns of immunoreactivity. This pattern is very similar to that produced by rabbit antibodies directed against a nerve-terminal marker, synapsin I, as shown in Panel G.

In Panel H and Panel I, adjacent sections of deep cerebellar nuclei stained with serum samples from Patients 1 and 3, respectively, show localized immunoreactivity in the cytoplasm of neuronal perikarya (P) and dendrites as well as in the nerve terminals that form synapses at their surfaces (arrows). Double-immunofluorescence labeling of a brain-stem section with serum from Patient 3 (Panel J) and antibodies directed against synaptophysin (Panel K), a synaptic-vesicle marker, shows nerve terminals filled with synaptic vesicles (white puncta). The profiles of neuronal perikarya and dendrites (arrows) are apparent. The scale bar measures 0.25 mm in Panel A, 25 microm in Panel B and Panel C, 50 microm in Panels D, E, F, and G, and 25 microm in Panels H, I, J, and K. (Reprinted with permission from F. Folli et al, Autoantibodies to a 128 kd protein in three women with the stiff-man syndrome and breast cancer. N Engl J Med 1993; 328: 546-551. Copyright 1993 Massachusetts Medical Society. All rights reserved).

Anti-amphiphysin antibody titers during two series of plasma exchanges (filled squares: anti-amphysin titer, arrows: days of plasma exchange) in a patient with anti-amphiphysin associated stiff-person syndrome. Clinical improvement closely correlated with drop in titer. ( Reprinted from Wessig et al, Neurology 2003; 61:195-198, with permission).