Respective implications of glutamate decarboxylase antibodies in stiff person syndrome and cerebellar ataxia

Abstract

Background: To investigate whether Stiff-person syndrome (SPS) and cerebellar ataxia (CA) are associated with distinct GAD65-Ab epitope specificities and neuronal effects.

Methods: Purified GAD65-Ab from neurological patients and monoclonal GAD65-Ab with distinct epitope specificities (b78 and b96.11) were administered in vivo to rat cerebellum. Effects of intra-cerebellar administration of GAD65-Ab were determined using neurophysiological and neurochemical methods.

Results: Intra-cerebellar administration of GAD65-Ab from a SPS patient (Ab SPS) impaired the NMDA-mediated turnover of glutamate, but had no effect on NMDA-mediated turnover of glycerol. By contrast, GAD65-Ab from a patient with cerebellar ataxia (Ab CA) markedly decreased the NMDA-mediated turnover of glycerol. Both GAD65-Ab increased the excitability of the spinal cord, as assessed by the F wave/M wave ratios. The administration of BFA, an inhibitor of the recycling of vesicles, followed by high-frequency stimulation of the cerebellum, severely impaired the cerebello-cortical inhibition only when Ab CA was used. Moreover, administration of transcranial direct current stimulation (tDCS) of the motor cortex revealed a strong disinhibition of the motor cortex with Ab CA. Monoclonal antibodies b78 and b96.11 showed distinct effects, with greater effects of b78 in terms of increase of glutamate concentrations, impairment of the adaptation of the motor cortex to repetitive peripheral stimulation, disinhibition of the motor cortex following tDCS, and increase of the F/M ratios. Ab SPS shared antibody characteristics with b78, both in epitope recognition and ability to inhibit enzyme activity, while Ab CA had no effect on GAD65 enzyme activity.

Conclusions: These results suggest that, in vivo, neurological impairments caused by GAD65-Ab could vary according to epitope specificities. These results could explain the different neurological syndromes observed in patients with GAD65-Ab.

Figure 1

Illustration of bilateral implantation of cerebellar guides. A: Two holes (black arrows) in the skull to implant the guides. B: Probes are inserted in cerebellar nuclei (CN). C: Bilateral injection of Ab. Brain sections at the level of cerebellar nuclei in B and C.

Figure 2

Sigmoidal fitting of recruitment curves of motor evoked potentials (MEP). Upper panels: Ab Ctrl, middle panels: Ab CA, lower panels: Ab SPS. Left panels: Ringer's. Right panels: BFA. Amplitudes are expressed in mV. Dash: 95% prediction band, dotted: 95% confidence band, p < 0.01 for each side. R² value > 0.970 (p < 0.005).

Figure 3

Cerebello-cortical inhibition. Thin trace: no conditioning stimulus; thick trace: test stimulus with conditioning stimulus. Left panels: baseline, middle panels: high-frequency, right panels: trains of tDCS. Microinjection of Ab Ctrl in A and B; Ab CA in C and D; Ab SPS in E and F.

Figure 4

Cerebello-cortical inhibition (CCI). Condition 1: baseline, Condition 2: high-frequency stimulation of the cerebellar cortex, Condition 3: trains of tDCS. BFA and high-frequency stimulation of the cerebellar cortex in rats injected with Ab CA (triangles), Ab Ctrl (circles), and Ab SPS (squares). Values are mean +/- SD. *: p < 0.05.

Figure 5

Motor maps (contour plots and color code). X-axis: sagittal coordinates, Y-axis coronal coordinates, Z-axis: corticomotor responses recorded in the contralateral gastrocnemius muscle. So-called "hot spots" (red) are identified. Coordinates of stimulation are established using the stereotactic frame. Maximal responses set at 100% on each side. Left panels: basal, right panels: high-frequency and trains of tDCS.

Figure 6

Mean F/M ratios. Group 1: Ab Ctrl+Ringer's, left side; Group 2: Ab Ctrl+BFA, right side; Group 3: Ab CA+Ringer's, left side; Group 4: Ab CA+BFA, right side; Group 5: Ab SPS+Ringer's, left side; Group 6: Ab SPS+BFA, right side. White bars: baseline, dash: high-frequency stimulation, black: tDCS. Means (+/- SD) are illustrated. Ratios are in %. *: p < 0.05, **: p < 0.01.

Figure 7

GAD65 enzyme inhibition and epitope specificity of Ab SPS, Ab CA, and Ab Ctrl. A: Enzyme activity in the presence of Ab SPS, Ab CA, and Ab Ctrl is reported as percent activity with enzyme activity in the absence of Ab set as 100%. B: Binding of Ab SPS and Ab CA to radiolabeled GAD65 in the presence of rFab b78 and b96.11 is reported as percent binding with uncompeted binding set at 100%.

Figure 8

Concentrations of glutamate, glucose, and glycerol in cerebellar nuclei. A: Concentrations of glutamate in control rats, rats receiving b96.11, and b78. Values are mean +/- SD. **: p < 0.01. B: glucose concentrations in cerebellar nuclei during NMDA challenge following infusion of b96.11 and b78. C: glycerol concentrations following administration of b96.11 and b78. Values are mean +/- SEM for glucose, median +/- SEM for glycerol. *: p < 0.05.

Figure 9

Modulation of the excitability of the motor cortex. Control: no cerebellar intervention; Ringer: intra-cerebellar administration of Ringer's; b78: administration of b78; b96.11 administration of b96.11. Recordings in gastrocnemius muscle, before (black columns) and after ipsilateral intra-cerebellar microinjections (white columns). *: p < 0.05; **: p < 0.01.

Figure 10

Effects of tDCS on the amplitudes of motor evoked potentials (contour plots). X-axis: sagittal coordinates, Y-axis coronal coordinates, Z-axis: intensities of corticomotor responses. Coordinates of stimulation are established using the stereotactic frame. Responses are expressed as % of baseline. Upper left: control rat; upper right: Ringer's; lower left: b78; lower right: b96.11.

Figure 11

F/M ratios. Controls: (black column), Ringer's administration: (dark grey column), b78: (grey column), b96.11: (white column). Values obtained in left plantaris following stimulation of the plantar nerve. Means (+/- SD) are illustrated. Ratios are expressed in %. *: p < 0.05, **: p < 0.01.