TNFα levels and macrophages expression reflect an inflammatory potential of trigeminal ganglia in a mouse model of familial hemiplegic migraine

Abstract

Latent changes in trigeminal ganglion structure and function resembling inflammatory conditions may predispose to acute attacks of migraine pain. Here, we investigated whether, in trigeminal sensory ganglia, cytokines such as TNFα might contribute to a local inflammatory phenotype of a transgenic knock-in (KI) mouse model of familial hemiplegic migraine type-1 (FHM-1). To this end, macrophage occurrence and cytokine expression in trigeminal ganglia were compared between wild type (WT) and R192Q mutant Ca(V)2.1 Ca(2+) channel (R192Q KI) mice, a genetic model of FHM-1. Cellular and molecular characterization was performed using a combination of confocal immunohistochemistry and cytokine assays. With respect to WT, R192Q KI trigeminal ganglia were enriched in activated macrophages as suggested by their morphology and immunoreactivity to the markers Iba1, CD11b, and ED1. R192Q KI trigeminal ganglia constitutively expressed higher mRNA levels of IL1β, IL6, IL10 and TNFα cytokines and the MCP-1 chemokine. Consistent with the report that TNFα is a major factor to sensitize trigeminal ganglia, we observed that, following an inflammatory reaction evoked by LPS injection, TNFα expression and macrophage occurrence were significantly higher in R192Q KI ganglia with respect to WT ganglia. Our data suggest that, in KI trigeminal ganglia, the complex cellular and molecular environment could support a new tissue phenotype compatible with a neuroinflammatory profile. We propose that, in FHM patients, this condition might contribute to trigeminal pain pathophysiology through release of soluble mediators, including TNFα, that may modulate the crosstalk between sensory neurons and resident glia, underlying the process of neuronal sensitisation.

Figure 1. Iba1 immunoreactivity in trigeminal ganglia from in WT and R192Q KI mice.

A, Representative confocal microscopy image of a longitudinal section of mouse trigeminal ganglion immunostained for neuronal β-tubulin III (green), and labeled with DAPI (blue). Discrete distribution of neuronal somata in three histological subdivisions of the trigeminal ganglion (i.e. V1, V2 and V3) is indicated by ellipsoids. Scale bar: 300 µm. B, Representative confocal microscopy images of WT (top row) or R192Q KI (bottom row) trigeminal ganglion sections from different V1, V2 and V3 regions of WT and KI ganglia and immunostained with Iba1 (red). Nuclear signal is also shown (blue). Scale bar: 15 µm. C, Histograms quantify Iba1-positive cells in V1, V2 and V3 regions (ROI: 370 µm×370 µm) of WT and KI trigeminal ganglia. Data were collected in parallel from at least 3 WT and 3 KI mice; * p<0.01. D, Panels show confocal images of WT (top row) and R192Q KI (bottom row) trigeminal ganglion sections immunostained with anti-Iba1 (red, macrophages) and the anti-glutamine synthetase antibodies (GS, green, satellite glial cells). Nuclei were labeled with DAPI (blue). Note the peculiar GS-positive immunostaining of satellite glial cells typically surrounding neurons (green). No Iba1 co-localisation was found in GS-immunolabeled cells. Scale bar: 20 µm. E, Histograms quantify the Iba1-immunoreactivity (IR) in whole ganglion slices from R192Q KI with respect to WT samples. Data are from an average of 30 ROIs per whole ganglion, respectively from n = 6 WT and KI mice. ROI: 640×480 µm. n = 6 mice, * p<0.001. F, Iba1 immunoreactivity was counted in neuronal- (β-tubulin III-positive) or fiber- (MAP-2-positive, not shown) enriched ganglia regions; n = 3 mice; * p<0.05; ** p<0.01; *** p<0.001.

Figure 2. Different morphology of Iba1 immunoreactive cells in WT and R192Q KI trigeminal ganglia.

A, Examples of 3D reconstruction of Iba1-positive cells (from confocal Z-stacks, each 0.5 µm-tick) from WT or R192Q KI ganglia. Note large branching of WT cell vs compact, process-free KI cell morphology. Scale bar: 5 µm. B, Histograms quantify average volume (µm³) of Iba1-positive cells in WT and R192Q KI ganglia, obtained from voxel analysis of 3D images. Data were collected from three independent experiments with a total of 83 cells for WT and 70 cells for R192Q KI; * p<0.001. C, Histograms quantify the average volume of Iba1-positive cells from different neurons or fibers enriched areas of WT and R192Q KI ganglia; n = 20–50 cells (3 WT and 3 KI mice). * p<0.05; ** p<0.01.

Figure 3. Characterization of Iba1-positive cells in trigeminal ganglia from WT and R192Q KI mice.

A, C, E, Representative confocal images of WT or R192Q KI trigeminal ganglion sections (from V3 region) immunostained for Iba1 (red) and CD11b (A), ED1 (C), or F4/80 (E) in green. Nuclei were labeled with DAPI (blue). Scale bar: 40 µm. Insets represent larger magnification of immunoreactive Iba1 cells. Scale bar: 7 µm. B, D, F, Histograms quantify the percentage of occurrence of Iba1 signal with CD11b (B), ED1 (D) or F4/80 (F) in different WT and KI ganglion regions (V1, V2 or V3). Expression of different markers was quantified in Iba1 positive cells only. n = 3 WT and 3 KI mice; * p<0.05; ** p<0.01. Data are expressed as mean ± S.D.

Figure 4. Expression of inflammatory mediators in WT and R192Q KI trigeminal ganglia under basal conditions.

A, Histograms quantify IL1β, IL6, IL10 and TNFα cytokine protein levels from WT or R192Q KI ganglia; n = 4 WT and 4 KI mice; data were normalized on total protein content, and represented as fraction of WT. * p<0.001. B, Real-time RT-PCR experiments quantify IL1β, IL6, IL10 and TNFα mRNA levels in WT and R192Q KI trigeminal ganglia. PCR data were normalized with respect to corresponding GAPDH and β-Tubulin housekeeping gene expression and expressed as fraction of WT; n = 4 WT and 4 KI mice; * p<0.05. C, Real-time RT-PCR experiments quantify MCP-1 mRNA levels in WT and R192Q KI trigeminal ganglia (expressed as in B); n = 4 WT and 4 KI mice; * p<0.05. D, Representative western blot experiment of WT or R192Q KI trigeminal ganglia extracts immuno-probed with anti-MCP-1 antibodies. Actin levels were used as loading control. Histograms quantify the differences. n = 3; * p<0.05.

Figure 5. TNFα expression in WT and KI ganglia.

A, Representative confocal microscopy images of WT (top row) or R192Q KI (bottom row) trigeminal ganglion sections immunostained for Iba1 (red) or TNFα(green) in basal condition. Pseudocolor images showing areas of high (yellow) and low (blue) Iba1-TNFα expressing cell co-localization. Color scale was also included. Note TNFα immunostaining detected as spots along perimembrane regions. The larger magnification insets show immunostaining of Iba1-TNFα signal (yellow) in KI rather than WT. Scale bar: 30 µm, for large images; Scale bar: 10 µm for larger magnification insets. B, Histograms quantify the percentage of TNFα immunoreactivity over the total of Iba1 expressing cells in different V1, V2 or V3 trigeminal regions (ROI: 370×370 µm). n = 4 WT and 4 R192Q KI mice; * p<0.05. Data are expressed as mean ± S.D.

Figure 6. LPS evoked acute TNFα expression in R192Q KI ganglia.

A, Representative confocal microscopy images of WT (top row) or KI (bottom row) trigeminal ganglion sections immunostained for Iba1 (red) or TNFα(green) after saline (left) or LPS injection (i.p., 5 h; right). LPS evokes TNFα expression in WT and KI after injection. Scale bar: 20 µm. B, C, Histograms quantify the occurrence of Iba1 signal (B) and Iba1-TNFα co-localisation (C) in different V1, V2 or V3 trigeminal ganglion regions from WT or KI mice, after saline or LPS injection (i.p., 5 h). ROI: 370×370 µm. n = 3 WT and 3 R192Q KI mice; * p<0.05; ** p<0.01; *** p<0.001. D, Histograms quantify changes in TNFα mRNA fold increase in WT or R192Q KI ganglia following LPS-injection (i.p., 1 h). Data are expressed as fold increase with respect to saline-injected mice samples; n = 3 WT and 3 KI mice; * p<0.05. E, TNFα protein levels in whole ganglia (pg cytokine/mg protein content) from LPS-injected (i.p. 5 h) WT or R192Q KI mice, expressed as fold increase with respect to saline-injected mice. n = 3 WT and 3 KI mice; * p<0.05.