Prolonged stimulation of a brainstem raphe region attenuates experimental autoimmune encephalomyelitis

Abstract

Multiple sclerosis (MS), a neuroinflammatory disease, has few treatment options, none entirely adequate. We studied whether prolonged electrical microstimulation of a hindbrain region (the nucleus raphe magnus) can attenuate experimental autoimmune encephalomyelitis, a murine model of MS induced by MOG_35-55 injection. Eight days after symptoms emerged, a wireless electrical stimulator with an attached microelectrode was implanted cranially, and daily intermittent stimulation was begun in awake, unrestrained mice. The thoracic spinal cord was analyzed for changes in histology (on day 29) and gene expression (on day 37), with a focus on myelination and cytokine production. Controls, with inactive implants, showed a phase of disease exacerbation on days 19-25 that stimulation for >16days eliminated. Prolonged stimulation also reduced numbers of infiltrating immune cells and increased numbers of myelinated axons. It additionally lowered genetic expression of some pro-inflammatory cytokines (interferon gamma and tumor necrosis factor) and platelet-derived growth factor receptor alpha, a marker of oligodendrocyte precursors, while raising expression of myelin basic protein. Studies of restorative treatments for MS might profitably consider ways to stimulate the raphe magnus, directly or via its inputs, or to emulate its serotonergic and peptidergic output.

Keywords: cytokines; deep brain stimulation; multiple sclerosis; myelination; nucleus raphe magnus.

Fig. 1

Stimulators and histology. (A) Encapsulated stimulators just before implantation, showing protruding electrodes (wire anode and microelectrode cathodes). (B) An awake mouse with an implanted working stimulator. (C) Immunostained sagittal section through the brainstem, including gliosis around the electrode task. Red: NeuN; green, GFAP; blue, DAPI. Scale bar = 0.5 mm. (D) A sagittal section through the midline of the brainstem, stained with hematoxylin-eosin, showing the approximate trajectory of the microelectrode through the cerebellum and dorsal medulla. The abbreviation CC marks the start of the spinal central canal. The dotted line outlines the NRM.

Fig. 2

Electrical stimulation of NRM is protective in EAE. (A) Experimental timeline, as planned. (B) Time course of EAE, starting with the day of its first appearance, in the control (‘sham’) group and in mice receiving more than 16 days of stimulation (‘stim’). Stimulation was initiated on day 8. Results are expressed as daily mean scores ± SEM; n=12 for control group and n=9 for the stimulated group. The broken line indicates the percentage of functioning stimulators on a given day. The bar marked by an asterisk shows the 7 days of peak EAE subjected to repeated measures ANOVA (F_1,19=6.71, p=0.020, partial eta-squared=0.30). Enlarged symbols indicate significant differences from the control group found in t-tests on individual days (p≤0.05); day 19: t₁₉=|2.24|, p=0.040; day 21: t₁₉=|2.28|, p=0.037; day 22: t₁₉=|2.23|, p=0.041; day 23: t₁₉=|2.67|, p=0.017; day 24: t₁₉=|2.80|, p=0.013; day 25: t₁₉=|2.55|, p=0.022. (C) EAE scores for mice receiving stimulation for 4–15 days (n=7); the curve for control mice (n=12) appears also in graph B. (D) EAE score plotted against the numbers of days after cessation of stimulation. The daily score for an individual mouse was subtracted from the mean control score for the same post-disease day, and scores were averaged within each day counting from the end of stimulation (designated as day 1). Enlarged symbols represent significant within-subject contrasts for differences between the data point and all previous times (p≤0.05), as determined by repeated measures analysis. After stimulation for 4–7 days (n=4), days showing significant differences were day 10 (F_1,3=10.4, p=0.045, partial eta-squared=0.78) and day 11 (F_1,3=10.1, p=0.048, partial eta-squared=0.78). After stimulation for 15–21 days (n=6), days showing significant differences were day 4 (F_1,5=10.6 p=0.022, partial eta-squared=0.68) and day 5 (F_1,5=7.0, p=0.046, partial eta-squared =0.53).

Fig. 3

Histological analysis of spinal cord. (A): Toluidine blue staining of semi-thin thoracic spinal cord sections. Images of representative areas were acquired in the white matter of control (‘sham’) mice and stimulated (‘stim’) mice. Solid black arrows show collapsed axons and smaller arrows with white interiors show infiltrating cells. The scale bar in the lower left of the panel is 10 μm. Stereological quantification of myelinated axons (B), collapsed axons (C) and infiltrating cells (D) in the thoracic white matter of control mice (n=5) and stimulated mice (n=5). The vertical scale reflects the total counted axons within the sampled tissue (B, C) or infiltrating cells per area of the thin sections (D). Horizontal bars represent the mean. Asterisks above the column of the stimulated group indicate significant contrasts. The stimulated group received 19.4 ±1.5 (mean ±SEM) days of stimulation. Differences were significant (p≤0.05) in unpaired t-tests for myelinated axons (t₈=3.584, p=0.007, partial eta-squared=0.62) and infiltrating cells (t₈=2.591, p=0.032, partial eta-squared=0.46), but not for collapsed axons.

Fig. 4

Gene expression profiling in the thoracic spinal cord after EAE. Gene expression levels were evaluated 37 days after EAE symptoms emerged in samples of control (‘sham’) mice (filled squares) and stimulated (‘stim’) mice (filled circles). Expression of the gene of interest was normalized to GAPDH expression and results in the treated group are graphed as percent of the mean of the corresponding control group. Horizontal bars indicate medians for each group. Note the cytokines (lower three graphs) are displayed on logarithmic vertical axes, which are better to show the large range of values in stimulated mice. Mean days of stimulation ranged over groups from 18.4 to 22.8. Groups sizes were MBP stimulated n=7, control n=10; Notch1 stimulated n=8, control n=10; PDGFRα stimulated n=6, control n=9; TNF stimulated n=7, control n=10; IFNγ stimulated n=6, control n=10; IL-1β stimulated n=5, control n=11. Group sizes varied due to occasional problems with the chemical assay. Significant values in the Mann-Whitney test were for MBP U=10.00, 7, 10, p=0.015; for PDGFRα U=6.00, 6, 9, p=0.012; for TNF U=11.00, 7, 10, p=0.018; for IFNγ U=9.00, 6, 10, p=0.023. Significance is indicated on graphs by an asterisk above the column of the stimulated group.