Basal astrocyte and microglia activation in the central nervous system of Familial Hemiplegic Migraine Type I mice

Abstract

Background

Gain-of-function missense mutations in the α_1A subunit of neuronal Ca_V2.1 channels, which define Familial Hemiplegic Migraine Type 1 (FHM1), result in enhanced cortical glutamatergic transmission and a higher susceptibility to cortical spreading depolarization. It is now well established that neurons signal to surrounding glial cells, namely astrocytes and microglia, in the central nervous system, which in turn become activated and in pathological conditions can sustain neuroinflammation. We and others previously demonstrated an increased activation of pro-algogenic pathways, paralleled by augmented macrophage infiltration, in both isolated trigeminal ganglia and mixed trigeminal ganglion neuron-satellite glial cell cultures of FHM1 mutant mice. Hence, we hypothesize that astrocyte and microglia activation may occur in parallel in the central nervous system.

Methods

We have evaluated signs of reactive glia in brains from naïve FHM1 mutant mice in comparison with wild type animals by immunohistochemistry and Western blotting.

Results

Here we show for the first time signs of reactive astrogliosis and microglia activation in the naïve FHM1 mutant mouse brain.

Conclusions

Our data reinforce the involvement of glial cells in migraine, and suggest that modulating such activation may represent an innovative approach to reduce pathology.

Keywords

Reactive astrocytes activated microglia neuroinflammation migraine

Introduction

Familial hemiplegic migraine (FHM) is a monogenic subtype of migraine with aura, characterized by recurrent attacks of headache that are accompanied by auras consisting of transient neurological symptoms that include visual, sensory, and motor disturbances (1). Transgenic FHM1 mutant mice that express gain-of-function missense mutations in CACNA1A, which encodes the α_1A subunit of voltage-gated Ca_V2.1 calcium channels, revealed neuronal involvement with increased Ca_V2.1 channel function and enhanced cortical glutamatergic neurotransmission that can explain the increased susceptibility to experimentally induced cortical spreading depolarization (CSD) (2,3). The consequences of the mutation may, however, extend to cell types that surround neurons in the brain. For instance, there is evidence suggesting that neuronal hyperexcitability can trigger so-called “neurogenic neuroinflammation”, which involves the vascular and glial cell components of brain tissue, and can recruit immune cells from the bloodstream (4). Although a homeostatic role for neuroinflammation has been postulated, it might also lead to further release of pro-inflammatory mediators by surrounding activated glial cells, namely reactive astrocytes and microglia, which in turn may trigger or aggravate an underlying pathological condition (4). Notably, in primary cultures from postnatal day 11 trigeminal ganglia tissue (containing a mix of trigeminal ganglion neurons and satellite glial cells) of FHM1 mutant mice, satellite glial cells exhibit an increased release of inflammatory mediators and an up-regulation of specific pro-algogenic receptors (5), but it is questionable whether findings in cultured cells can be translated to the intact trigeminal ganglion as no such increases were observed in isolated trigeminal ganglia from mice that were 11–14 weeks old (6). In addition, trigeminal infiltration of macrophages was observed, pointing to the development of neuroinflammation in the peripheral nervous system (7). Hence, both abnormal neuronal and glial cell function, and likely their interaction, may contribute to FHM pathophysiology. To further examine the involvement of glial cell activation in FHM, we here investigated whether the FHM1 R192Q missense mutation promotes the development of a reactive glia phenotype in the central nervous system of naïve mutant mice.

Methods

Animals

Three- to 4-month-old male homozygous FHM1 R192Q knock-in (KI) (“R192Q”) and wild-type (“WT”) mice were used. The KI mice were generated by introducing the human FHM1 pathogenic R192Q missense mutation in the orthologous Cacna1a gene using a gene targeting approach (8). Genotyping was performed as described before (8). Mice were kept under standard conditions (temperature: 22 ± 2℃; relative humidity: 50 ± 10%; light regimen: 12-hour light/12-hour dark cycle) with food and water ad libitum. All experimental procedures were in strict accordance with the Italian and EU regulation on animal welfare and were previously approved by the local ethics committee, and by the Italian Ministry of Health (authorization #736/2015-PR). All results are reported according to Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. Researchers were blinded for genotype information, and animals were randomised. Seven animals for both the WT and the FHM1 mutant group were analyzed per type of staining.

Immunohistochemistry

Mice were anesthetized with 1.5% isoflurane and perfused with 4% paraformaldehyde, as described (9). Brains were removed, postfixed in 4% formalin for 60 minutes, cryoprotected in 30% sucrose for 24 hours, embedded in mounting medium (OCT; Tissue Tek, Sakura Finetek, Zoeterwoude, The Netherlands), and kept at −80℃ until use. Twenty µm-thick coronal brain sections were collected for immunohistochemistry. Rabbit antibodies directed against: (i) astrocyte marker glial fibrillary acidic protein (anti-GFAP, 1:600; Dako, Milan, Italy), (ii) microglia marker ionized calcium binding adaptor molecule 1 (anti-Iba1, 1:500; Wako, Richmond, VA, USA), (iii) GPR17 receptor (home-made, 1:20,000); or mouse antibodies directed against mature oligodendrocyte marker glutathione S-transferase pi (anti-GSTπ, 1:500; MBL International, Woburn, MA, USA) were used. As secondary antibody, goat anti-rabbit or goat anti-mouse (1:600; Thermo Fisher Scientific, Monza, Italy) conjugated to AlexaFluor®488 or AlexaFluor®555, respectively, were used. Cell nuclei were counterstained with the Hoechst33258 dye (1:20,000; Sigma-Aldrich, Milan, Italy). Cell staining was evaluated by a Zeiss Axioskop fluorescent microscope (Carl Zeiss, Milan, Italy), with the aid of the NIH ImageJ software, as described below.

Image analysis

Cell counting

For Iba1, GPR17, and GSTπ staining, the number of immunopositive cells was counted in whole sections and in selected brain areas (i.e. the cortex and corpus striatum), acquired at 20× magnification, and expressed as the number of cells/area in µm².

Densitometric analysis

For GFAP staining, a digital image of the immunostained brain sections was acquired at low (10×) magnification, and the mean values of pixel intensity were automatically evaluated using the NIH Image-J software (10), and expressed as integrated density compared to values obtained in WT mice set to 100%.

Evaluation of microglia morphology and branch complexity

Fluorescent images of Iba1-positive microglia in the cortex and the corpus striatum were acquired at higher (40×) magnification, and converted to binary grayscale to better analyze cell morphology by the “Simple neurite tracer” tool of the Fiji-ImageJ software (11). In each brain area, three randomly chosen cells in each of 10 randomly chosen optical fields (for a total of 30 cells/animal) were manually traced (see Figure 2(a) for representative images) and i) the area covered by cell processes (which is proportional to the total cell ramification), ii) the number of junctions and branches and the mean and maximal length of the latter (which reflects the complexity and ramification of cells), and iii) the number of triple and quadruple points (an additional indicator of cell complexity, which indicates junctions where a single branch ramifies in three or four) were evaluated with the Skeleton analysis tool of the same software (11).

Figure 1.

Reactive astrogliosis in the naïve FHM1 mutant mouse brain. (a) Coronal sections from the brains of wild-type (WT) and R192Q mutant mice stained with anti-GFAP antibody. Representative pictures are shown with magnified details. Scale bars: 1.5 mm. (b) Quantification of GFAP immunoreactivity by densitometric analysis. Scatter plots show the results obtained in seven animals for each group. The pixel intensity values are expressed as mean ± SD compared to WT animals set to 100%. p = 0.007 with respect to WT, two-tailed non-parametric Mann-Whitney test. (c) Evaluation of GFAP expression in cortical and striatal tissue by Western blotting. β-actin was utilized as internal loading control (see representative filters at the bottom). Scatter plots show the optical density of protein bands from seven animals for each group normalized on corresponding β-actin and expressed as mean ± SD with respect to WT animals set to 1. p = 0.028 (cortex) and p = 0.32 (c. striatum) with respect to WT, two-tailed non-parametric Mann-Whitney test.

Figure 2.

Increased number of microglia in the naïve FHM1 mutant brain. (a) Representative pictures of coronal sections from the brains of wild-type (WT) and R192Q mutant mice stained with primary antibody against Iba1 (40× magnification; scale bars: 25 µm). Magnified details show representative cells traced with the «Simple neurite tracer» tool of the Fiji-ImageJ software. Junctions and branches are indicated. (b) The total number of Iba1-positive microglial cells was counted in brain sections (20× magnification; not shown). Scatter plots show the mean ± SD of Iba1-positive cell number/area from seven animals for each group. p = 0.002 with respect to WT, two-tailed non-parametric Mann-Whitney test.

Western blotting

Western blotting analysis of brain tissues was performed, as described before (12). Mouse antibodies directed against: i) GFAP (1:1,000; Cell Signaling, Danvers, MA, USA), or ii) the marker of mature oligodendrocyte 2',3'-cyclic-nucleotide 3'-phosphodiesterase (CNPase, 1:250; Millipore, Vimodrone, MI, Italy) and rat antibody against myelin basic protein (MBP, 1:500; Millipore) as markers of myelinating oligodendrocytes, were used. Beta-actin (rabbit anti-β-actin, 1:1,500; Sigma-Aldrich) or α-tubulin (mouse anti α-tubulin, 1:1,000; Sigma-Aldrich) expression was analyzed as internal loading controls. Next, filters were incubated with species-specific secondary antibodies conjugated to horseradish peroxidase (goat anti-rabbit, 1:4,000 and goat anti-mouse, 1:2,000; both from Sigma-Aldrich; goat anti-rat, 1:2,000; Thermo Fisher Scientific). Protein detection was performed by ECL (BioRad, Milan, Italy). After autoradiography, the relative amount of protein was evaluated by the NIH Image-J software, normalized for the corresponding β-actin or α-tubulin values, and expressed with respect to values obtained in WT mice set to 1.

Statistical analysis

Scatter plots show single data and their mean ± standard deviation (SD). Continuous variables were compared by use of the Mann-Whitney U test. A two-tailed p-value < 0.05 was considered statistically significant. Testing of hypotheses was performed using the free software R (13).

Results

Activated astrocytes in the naïve FHM1 R192Q mutant brain

We first examined the expression of GFAP, a typical astrocyte marker whose expression is increased during reactive astrogliosis (14). Counting cell numbers was not feasible, due to the complex interconnections among cells; thus, we measured the fluorescence intensity of immunostaining (see Methods). In the WT cortex the highest abundance of stained cells is seen in the subventricular zone lining lateral ventricles (Figure 1(a)), similar to what was reported before for that brain region (15). In the naïve FHM1 mutant mouse brain, however, GFAP immunoreactivity was typically present in higher cortical layers (Figure 1(a); see magnification in inset), with a significant 19.3 ± 4.9% increase in fluorescence when compared to the WT brain (Figure 1(b)), suggesting reactive astrogliosis. Western blotting analysis of cortical tissue confirmed the results by showing a significant 25.4 ± 8.7% increase in GFAP expression. In the striatum, GFAP expression was not significantly increased (+26.3 ± 15.3%, p = 0.32; Figure 1(c)), although the results were quite variable among samples. Moreover, Western blotting analysis of TG tissue also did not show a genotypic difference (normalized GFAP expression 1.00 ± 0.12 vs. 1.19 ± 0.10 in the TG from WT and R192Q mice, respectively; p = 0.144 Student’s t test).

Hyper-ramified resident microglia in the naïve FHM1 R192Q mutant brain

Next, we evaluated the number and morphology of microglial cells in the brains of FHM1 mutant and WT mice, to assess whether cells are in a resting, activated, or intermediate state (16). The total number of Iba1-positive microglia cells in coronal brain sections was found to be significantly increased in FHM1 mutant mice (Figure 2(a), 2(b)) suggesting cell proliferation. In both the cortex and corpus striatum, the increased cell number was paralleled by an increase in the total area covered by microglia ramifications (Figure 3(a)), as further evidenced by a higher number of long branches (Figure 3(b)–(d)) and of junctions and triple points, which are representative of more complex ramification of cell processes (Figure 3(e), (f); see images in Figure 2(a)). The number of quadruple points was not different between the two groups (p = 0.165; Figure 3(g)). Thus, our data show morphological changes of microglia towards a hyper-ramified intermediate state (16), which likely reflects cell adaptation in response to changes in the (pathological) environment (16), and suggest an intensification of surveillance of this cell population.

Figure 3.

Increased branching and complexity of microglia in the cortex and corpus striatum of FHM1 mutant mice. The complexity and ramification of Iba1-positive cells were evaluated by the Skeleton analysis tool of the Fiji-ImageJ software (see Methods section). In both cortex and c. striatum from wild-type (WT) and R192Q mutant mice, scatter plots show: (a) the area covered by the cell projection tree, (b) the total number, (c) the average and (d) the maximal length of branches, (e) the total number of junctions and of (f) triple and (g) quadruple points. Data are the mean ± SD of 30 randomly chosen cells/animal from seven animals for each group (see Methods for details). (a) p = 0.0006 (cortex and c. striatum), (b) p = 0.004 (cortex) and p = 0.038 (c. striatum), (c) p = 0.053 (cortex) and p = 0.026 (c. striatum); (d) p = 0.007 (cortex) and p = 0.004 (c. striatum); (e) p = 0.011 (cortex and c. striatum); (f) p = 0.011 (cortex) and p = 0.038 (c. striatum); (g) p = 0.165 (cortex and c. striatum), two-tailed non-parametric Mann-Whitney test.

No basal changes in markers of mature oligodendrocytes in the FHM1 mutant brain

Based on the known reciprocal influence of neurons and oligodendrocytes, we have also evaluated possible changes in the expression of markers of myelination in the brains of FHM1 mutant mice. Counting the number of CNPase-positive mature and MBP-positive myelinating oligodendrocytes after immunohistochemistry (17) was not possible due to the high number of cells with a very complex morphology (not shown). We, therefore, performed semi-quantitative Western blotting analysis of portions of the cortex and of the corpus striatum from naïve FHM1 R192Q and WT mice. Results show no alteration in the expression of either protein (CNPase: Cortex, p = 0.295; c. striatum, p = 0.836; MBP: Cortex, p = 0.945; c. striatum, p = 0.445; Figure 4). Nevertheless, a reduction in the number of immature oligodendrocytes and pre-oligodendrocytes expressing the GPR17 receptor (18) was detected in FHM1 mutant brains (p = 0.017; Figure 5(a), (b)). When analyzing specific brain areas, differences in the number of GPR17-expressing cells were observed in the corpus striatum (p = 0.038), with similar values (p = 0.465) in the cortex (Figure 5(b)), whereas the expression of GSTπ, a marker of intermediate oligodendrocyte maturation was similar between WT and FHM1 mice (65.6 ± 25.6 vs. 59.4 ± 17.6 positive cells/counted field, respectively; p = 0.390). Overall, our results suggest that the activity of mutated neuronal Ca_V2.1 channels does not alter the basal maturation of oligodendrocytes but reduces the fraction of immature and pre-oligodendrocytes expressing GPR17.

Figure 4.

No changes in the expression of markers of mature oligodendrocytes in the brains of FHM1 mutant mice. (a) Western blotting evaluation of CNPase and MBP expression in cortical and striatal tissues from wild-type (WT) and R192Q mutant mice. Alpha-tubulin was utilized as internal control for protein loading (see representative filters of cortex samples in (b)). The three MBP bands at different molecular weights (some 15-16-18 kDa) were analyzed together. Scatter plots show the optical density of protein bands from seven animals for each group expressed as mean ± SD with respect to WT animals set to 1. In (a) CNPase: p = 0.295 (cortex) and p = 0.836 (c. striatum); MBP: p = 0.945 (cortex) and p = 0.445 (c. striatum), two-tailed non-parametric Mann-Whitney test.

Figure 5.

Reduced number of GPR17-positive cells in the FHM1 mutant brain. (a) Representative pictures of coronal sections from the brains of wild-type (WT) and R192Q mutant mice stained with a home-made antibody against the membrane receptor GPR17. Cell nuclei were counterstained with the Hoechst33258 dye. The number of GPR17-positive cells was counted in the entire coronal section and in selected brain areas; that is, the cortex and the corpus striatum. Scale bars: 1.5 mm. (b) Scatter plots show the mean ± SD of GPR17-positive cell number/counted field in the different brain areas. Data were obtained by counting 10 sections from seven independent animals/group.

Discussion

Here we present evidence for increased glial cell reactivity in the naïve FHM1 R192Q mouse brain, as evidenced by morphological signs of astrogliosis, and of reactive microglia. Because of the previously reported increased glutamate neurotransmitter release from excitatory cortical neurons in the FHM1 mutant brain (3), we set out to investigate whether the observed enhanced neurotransmission could also modify neuron-to-glia cell communication, directly or indirectly, and promote the basal activation of resident glial cells, thus contributing to the FHM1 phenotype.

Given that non-neuronal cells, in particular glia cells, contribute to the initiation and maintenance of chronic pain (19), which is now referred to as a “gliopathy” (20), one possible role of permanently activated astrocytes and microglia is to contribute to the pathological brain environment relevant to FHM1 pathology.

Both astrocytes and microglia can respond to external stimuli and to changes in the environment by exerting either detrimental or protective functions (14,16). Although we do not have functional evidence of the polarization of these cells towards an overall pro- or anti-inflammatory phenotype in the FHM1 brain, it is relevant that here we detected changes in cell morphology, which for glial cells represent an indicative parameter of altered cell function, in the absence of a CSD trigger. Our data show that the presence of mutant neuronal Ca_V2.1 channels leads to what we would interpret as an adaptation of glial cells that have to cope, from time to time, with increased levels of glutamate (and likely ATP) and K⁺ ions, and likely prime cells to be ready to react to possible subsequent triggers.

Concerning astrocytes, consistent with the known role of calcium waves in cell-to-cell communication (21), here we have detected signs of reactive astrogliosis and we can speculate that this could be a factor that modulates headache pain. After all, astrocytes play a key role in the blood-brain barrier and control of the blood-brain exchanges of chemicals and other potential migraine triggers, such as certain nutrients (14). Additionally, glial cell dysfunction has been implicated in FHM2 mutant mice (22,23), in which loss-of-function mutations in the α₂ subunit of Na⁺/K⁺ ATPases expressed in astrocytes resulted in increased cortical glutamatergic neurotransmission, but in this case as the result of inadequate K⁺ and glutamate buffering ability due to abnormal glial function (24). Similar to FHM1 mice, FHM2 mice were shown to be more susceptible to experimentally induced CSD (22,25). Thus, data from a different model of FHM suggests that astrocytes can contribute to migraine pathology.

Microglia are the immunocompetent cells in the central nervous system, which constantly scan the environment and contact surrounding neurons and astrocytes. Under pathological conditions, microglia acquire an activated phenotype, accompanied by the release of cytokines, increased proliferation, migration and phagocytic activity (19). Increased neuronal activity, as seen in epilepsy or during CSD propagation (26,27), is another trigger for microglia reaction. Additionally, higher microglia dynamic movements are detected hours after induction of CSD, and the level of microglia activation is proportional to the number of CSD waves (28). The transition between resting and fully activated (i.e. amoeboid phagocytic microglia) is characterized by the presence of an intermediate phenotype, which can be recognized by increased branching as in the naïve FHM1 brain, representing a “warning” state of cells to be fully able to respond to subsequent harms (16). Very recent data show that elongation of processes is mandatory for microglia to shift from a pro- to an anti-inflammatory phenotype, which can more efficiently patrol the surrounding brain tissue (29). Thus, simply based on changes in cell morphology we cannot definitively assume whether hyper-ramified microglia have an overall detrimental or beneficial role. Further studies are needed to analyze the expression of specific markers of microglia polarization towards an anti-inflammatory M1 or pro-inflammatory M2 phenotype (16). Nevertheless, we can conclude that in FHM1 mutant brains under basal conditions, already after minor disturbances in brain activity, microglia enlarge their area of patrolling by extending cell processes and are therefore able to react more efficiently, if needed, and to release various signaling molecules involved in the cross-talk with the other cell populations.

Oligodendrocytes and their precursor cells (OPCs) are also involved in cell-to-cell communication in the brain by reacting to neuronal signals and to activated astrocytes and surrounding microglial cells (30,31). To confirm their relevance to migraine pathology, altered patterns of myelination have been observed in the brains of migraineurs (32), and transient disruption of myelin structure is induced by CSD (33). Our data show no basal changes in the expression of mature oligodendrocytes in FHM1 brains, but a significant reduction of GPR17 expression.

As GPR17 receptor expression must be down-regulated to allow for the terminal maturation of pre-oligodendrocytes (34), its reduced expression could accelerate oligodendrocyte maturation. Nevertheless, since GPR17-expressing OPCs show a higher susceptibility to toxic signals, such as pathologically elevated concentrations of ATP (18) that are likely to be found in the FHM1 mutant brain as an additional consequence of the enhanced glutamatergic transmission (35), at present we cannot exclude that the observed reduction in their number is due to cell death.

A recent genetic study, which used genotype data from thousands of migraine patients and controls tested in genome-wide association studies (GWAS), evaluated gene sets containing astrocyte- and oligodendrocyte-related genes and found an association with the common forms of migraine (36). This supports the concept that genetic factors underlie glial cell dysfunction in migraine, so beyond the correlation presented here in FHM1 mutant mice. Glia cell activation, which is suggestive of neuroinflammation, may therefore be relevant to various types of migraine, although this has not firmly been established for common forms of migraine. It might be that glial cell activation is a relevant reflection of the basal reactive state of a FHM1 mutant brain that might influence how migraine triggers affect the brain. In that respect it is of interest that experimentally induced CSDs generate a molecular signature of activated inflammatory pathways specifically linked to interferon-γ signaling that is seen in the FHM1 mutant mice, but not the wild-type control mice (37). Together with data in FHM2 mice (see above), all in all the information supports the involvement of glial dysfuctions eventually leading to neuroinflammation in a migraine-relevant context.

As mentioned above, one limitation of our study is that the claim of glial cell activation in FHM1 mutant brains is only based on changes in cell morphology; that is, without investigating its functional consequences. Still, the observed features are regarded as highly representative of the functional state of the cells, so can be considered sufficient proof that our findings are genuine.

Taken together, we provided evidence that FHM1 mutations – at least the R192Q missense mutation, in a transgenic mouse model – not only impacts neurons, but also glia cells. This study therefore reinforces the concept that altered neuron-to-glia communication in naïve FHM1 mutant mice might contribute to the disease phenotype, and that normalizing glia cell function could have potential to treat migraine. We feel that our observations are relevant to the understanding of (hemiplegic) migraine pathophysiology, as one could ask the question whether the same occurs in patients.

Key findings

Reactive astrocytes and activated microglia are detected under basal conditions in the brains of FHM1 mutant mice expressing neuronal Ca_V2.1 calcium channels with R192Q-mutated α_1A subunits.

Normalization of glia cell reactivity maybe a promising avenue for drug treatment.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work has been partially supported by the Fondazione Cariplo (Grant #2011-0505 to MPA), by the Department of Excellence grant program from the Italian Ministry of Research (MIUR), the Center of Medical System Biology (CMSB) established by the Netherlands Genomics Initiative/Netherlands Organisation for Scientific Research (NGI/NWO) (to AMJMvdM), and the European Community (EC) FP7-EUROHEADPAIN (no. 602633 to AMJMvdM).

References

Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia 2013; 33: 629–808.

Tottene

Conti

Fabbro

, et al. Enhanced excitatory transmission at cortical synapses as the basis for facilitated spreading depression in Ca(v)2.1 knockin migraine mice. Neuron 2009; 61: 762–773.

Ferrari

Klever

Terwindt

, et al. Migraine pathophysiology: Lessons from mouse models and human genetics. Lancet Neurol 2015; 14: 65–80.

Xanthos

Sandkühler

. Neurogenic neuroinflammation: Inflammatory CNS reactions in response to neuronal activity. Nat Rev Neurosci 2014; 15: 43–53.

Ceruti

Villa

Fumagalli

, et al. Calcitonin gene-related peptide-mediated enhancement of purinergic neuron/glia communication by the algogenic factor bradykinin in mouse trigeminal ganglia from wild-type and R192Q Cav2.1 knock-in mice: Implications for basic mechanisms of migraine pain. J Neurosci 2011; 31: 3638–3649.

Chan

Labastida-Ramírez

Ramírez-Rosas

, et al. Trigeminovascular calcitonin gene-related peptide function in Cacna1a R192Q-mutated knock-in mice. J Cereb Blood Flow Metab 2019; 39: 718–729.

Franceschini

Vilotti

Ferrari

, et al. TNFα levels and macrophages expression reflect an inflammatory potential of trigeminal ganglia in a mouse model of familial hemiplegic migraine. PLoS One 2013; 8: e52394–e52394.

van den Maagdenberg

Pietrobon

Pizzorusso

, et al. A Cacna1a knockin migraine mouse model with increased susceptibility to cortical spreading depression. Neuron 2004; 41: 701–710.

Villa

Ceruti

Zanardelli

, et al. Temporomandibular joint inflammation activates glial and immune cells in both the trigeminal ganglia and in the spinal trigeminal nucleus. Mol Pain 2010; 6: 89–89.

10.

Schneider

Rasband

Eliceiri

. NIH Image to ImageJ: 25 years of image analysis. Nature Meth 2012; 9: 671–675.

11.

Schindelin

Arganda-Carreras

Frise

, et al. Fiji: An open-source platform for biological-image analysis. Nat Meth 2012; 9: 676–682.

12.

Magni

Merli

Verderio

, et al. P2Y₂ receptor antagonists as anti-allodynic agents in acute and sub-chronic trigeminal sensitization: Role of satellite glial cells. Glia 2015; 63: 1256–1269.

13.

R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org/ (2016, accessed 15 February 2019).

14.

Buffo

Rolando

Ceruti

. Astrocytes in the damaged brain: Molecular and cellular insights into their reactive response and healing potential. Biochem Pharmacol 2010; 79: 77–89.

15.

Boccazzi

Rolando

Abbracchio

, et al. Purines regulate adult brain subventricular zone cell functions: Contribution of reactive astrocytes. Glia 2014; 62: 428–439.

16.

Walker

Beynon

Jones

, et al. Dynamic structural remodelling of microglia in health and disease: A review of the models, the signals and the mechanisms. Brain Behav Immun 2014; 37: 1–14.

17.

Barateiro

Fernandes

. Temporal oligodendrocyte lineage progression: In vitro models of proliferation, differentiation and myelination. Biochim Biophys Acta 2014; 1843: 1917–1929.

18.

Ceruti

Viganò

Boda

, et al. Expression of the new P2Y-like receptor GPR17 during oligodendrocyte precursor cell maturation regulates sensitivity to ATP-induced death. Glia 2011; 59: 363–378.

19.

Salter

Stevens

. Microglia emerge as central players in brain disease. Nat Med 2017; 23: 1018–1027.

20.

Berta

Nedergaard

. Glia and pain: Is chronic pain a gliopathy? Pain 2013; 154: S10–S28.

21.

de Bock

Decrock

Wang

, et al. The dual face of connexin-based astroglial Ca(2+) communication: A key player in brain physiology and a prime target in pathology. Biochem Biophys Acta 2014; 1843: 2211–2232.

22.

Leo

Gherardini

Barone

, et al. Increased susceptibility to cortical spreading depression in the mouse model of familial hemiplegic migraine type 2. PLoS Genet 2011; 7: e1002129–e1002129.

23.

Bøttger

Glerup

Gesslein

, et al. Glutamate-system defects behind psychiatric manifestations in a familial hemiplegic migraine type 2 disease-mutation mouse model. Sci Rep 2016; 6: 22047–22047.

24.

Capuani

Melone

Tottene

, et al. Defective glutamate and K+ clearance by cortical astrocytes in familial hemiplegic migraine type 2. EMBO Mol Med 2016; 8: 967–986.

25.

Kros

Lykke-Hartmann

Khodakhah

. Increased susceptibility to cortical spreading depression and epileptiform activity in a mouse model for FHM2. Sci Rep 2018; 8: 16959–16959.

26.

Wendt

Wogram

Korvers

, et al. Experimental cortical spreading depression induces NMDA receptor dependent potassium currents in microglia. J Neurosci 2016; 36: 6165–6174.

27.

Grinberg

Milton

Kraig

. Spreading depression sends microglia on Lévy flights. PloS One 2011; 6: e19294–e19294.

28.

Shibata

Suzuki

. Exploring the role of microglia in cortical spreading depression in neurological disease. J Cereb Blood Flow Metab 2017; 37: 1182–1191.

29.

Yang

Wang

Wen

, et al. Regulation of microglial process elongation, a featured characteristic of microglial plasticity. Pharmacol Res 2019; 139: 286–297.

30.

Maldonado

Angulo

. Multiple modes of communication between neurons and oligodendrocyte precursor cells. Neuroscientist 2015; 21: 266–276.

31.

Lundgaard

Osório

Kress

, et al. White matter astrocytes in health and disease. Neuroscience 2014; 276: 161–173.

32.

Granziera

Daducci

Romascano

, et al. Structural abnormalities in the thalamus of migraineurs with aura: A multiparametric study at 3 T. Hum Brain Mapp 2014; 35: 1461–1468.

33.

Pusic

Mitchell

Kunkler

, et al. Spreading depression transiently disrupts myelin via interferon-gamma signaling. Exp Neurol 2015; 264: 43–54.

34.

Fumagalli

Daniele

Lecca

, et al. Phenotypic changes, signaling pathway, and functional correlates of GPR17-expressing neural precursor cells during oligodendrocyte differentiation. J Biol Chem 2011; 286: 10593–10604.

35.

Burnstock

. Purinergic cotransmission. Exp Physiol 2009; 94: 20–24.

36.

Eising

de Leeuw

Min

, et al. Involvement of astrocyte and oligodendrocyte gene sets in migraine. Cephalalgia 2016; 36: 640–647.

37.

Eising

Shyti

t Hoen

PAC

, et al. Cortical spreading depression causes unique dysregulation of inflammatory pathways in a transgenic mouse model of migraine. Mol Neurobiol 2017; 54: 2986–2996.