Fumarates modulate microglia activation through a novel HCAR2 signaling pathway and rescue synaptic dysregulation in inflamed CNS

Abstract

Dimethyl fumarate (DMF), recently approved as an oral immunomodulatory treatment for relapsing-remitting multiple sclerosis (MS), metabolizes to monomethyl fumarate (MMF) which crosses the blood-brain barrier and has demonstrated neuroprotective effects in experimental studies. We postulated that MMF exerts neuroprotective effects through modulation of microglia activation, a critical component of the neuroinflammatory cascade that occurs in neurodegenerative diseases such as MS. To ascertain our hypothesis and define the mechanistic pathways involved in the modulating effect of fumarates, we used real-time PCR and biochemical assays to assess changes in the molecular and functional phenotype of microglia, quantitative Western blotting to monitor activation of postulated pathway components, and ex vivo whole-cell patch clamp recording of excitatory post-synaptic currents in corticostriatal slices from mice with experimental autoimmune encephalomyelitis (EAE), a model for MS, to study synaptic transmission. We show that exposure to MMF switches the molecular and functional phenotype of activated microglia from classically activated, pro-inflammatory type to alternatively activated, neuroprotective one, through activation of the hydroxycarboxylic acid receptor 2 (HCAR2). We validate a downstream pathway mediated through the AMPK-Sirt1 axis resulting in deacetylation, and thereby inhibition, of NF-κB and, consequently, of secretion of pro-inflammatory molecules. We demonstrate through ex vivo monitoring of spontaneous glutamate-mediated excitatory post-synaptic currents of single neurons in corticostriatal slices from EAE mice that the neuroprotective effect of DMF was exerted on neurons at pre-synaptic terminals by modulating glutamate release. By exposing control slices to untreated and MMF-treated activated microglia, we confirm the modulating effect of MMF on microglia function and, thereby, its indirect neuroprotective effect at post-synaptic level. These findings, whereby DMF-induced activation of a new HCAR2-dependent pathway on microglia leads to the modulation of neuroinflammation and restores synaptic alterations occurring in EAE, represent a possible novel mechanism of action for DMF in MS.

Figures

Fig. 1

MMF induces a molecular switch in activated microglia from a pro-inflammatory to an alternatively activated phenotype. Real-time PCR analysis of mRNA expression of relevant genes in LPS-activated N9 cells treated with MMF shows that a MMF decreases the expression of genes associated with a pro-inflammatory phenotype in activated microglia; and b MMF upregulates the expression of neuroprotective (alternatively activated) phenotype markers in activated microglia. Data are presented as fold induction over the expression of the genes in untreated N9 cells, and results are shown as mean ± SEM of at least three independent experiments. *P < 0.05 and **P < 0.01

Fig. 2

MMF prompts functional changes associated with an alternatively activated phenotype of microglia. MMF enhances the phagocytic activity of LPS-activated N9 cells in vitro as shown by confocal microscopy images of LPS-activated N9 cells treated or not with MMF in the presence of green fluorescent microspheres (a) and quantified as the cell area occupied by the microspheres (b); the average cell section area was 34 ± 1 and 180 ± 7 μm2 for non-activated and activated cells, respectively. c MMF upregulates mRNA expression of Trem2 in LPS-activated N9 cells. Data are presented as fold induction over the expression of Trem2 in untreated N9 cells. d MMF increases [Ca2+]i in LPS-activated N9 cells, as measured by fluorometric analysis of Fura 2AM. Results are shown as mean ± SEM of at least three independent experiments. *P < 0.05 and ***P < 0.001. Bar = 15 μm

Fig. 3

MMF signals through binding to HCAR2. a N9 cells express HCAR2. Confocal microscopy images of N9 cells stained with anti-HCAR2 (green) and anti-CD11b (red) antibodies, and with the nuclear marker DAPI (blue). Bars = 20 and 5 μm for left and rightpanels, respectively. Magnification = 40× and 100× for left and right panels, respectively. b Blocking HCAR2 reverts the effect of MMF on the expression of genes dependent on NF-κB, but not on that of Cx3cr1 and Cd200r. N9 cells were incubated with or without anti-HCAR2 antibody (aHCAR2) for 1 h prior to LPS activation for 6 h in the presence or absence of MMF, and mRNA expression of the indicated genes was assessed by real-time PCR. Data are presented as fold induction over the expression of the genes in untreated N9 cells and results are shown as mean ± SEM of at least three independent experiments. *P < 0.05, **P < 0.01

Fig. 4

Postulated signaling pathway for MMF through binding to HCAR2

Fig. 5

Validation of the HCAR2-mediated pathway for MMF signaling. a Blocking HCAR2 suppresses MMF-induced increase in [Ca2+]i in activated microglia. N9 cells were incubated with or without aHCAR2 for 1 h prior to LPS activation for 6 h in the presence or absence of MMF and [Ca2+]i was determined by fluorimetric analysis. b MMF treatment activates AMPK in LPS-activated microglia. Representative Western blot (left panel) and quantification (right panel) of at least three Western blot analyses of N9 cell lysates are shown. N9 cells were incubated or not with aHCAR2 prior to activation with LPS for 30 min in the presence or absence of MMF. Western blot analysis of the lysates was performed with anti-p-AMPK, which recognizes phosphorylated Thr172, anti-AMPK, and anti-GRB2 (as loading control) antibodies. Quantification by densitometric analysis of Western blot films is presented as the proportion of p-AMPK over total AMPK (fold induction over N9), normalized to GRB2. Results are shown as mean ± SEM of at least three independent experiments; *P < 0.05. c MMF treatment increases NAD+ levels in LPS-activated microglia. Graph represents colorimetric quantification of intracellular NAD+ produced by N9 cells incubated with or without aHCAR2 prior to activation with LPS in the presence or absence of MMF for 30 min. d MMF treatment leads to NF-κB inhibition in activated microglia through deacetylation of NF-κB p65. Representative Western blot (left panel) and quantification (right panel) of at least three Western blot analyses of N9 cell lysates are shown. N9 cells were incubated with or without aHCAR2 and/or EX527 (10 μM) for 1 h prior to activation with LPS in the presence or absence of MMF or Resv (50 μM) for 6 h. Western blot analysis was performed with anti-Ac-NF-κB antibody, which recognizes acetylated lysine 310 of NF-κB p65, anti-NF-κB, and anti-GRB2 (as loading control) antibodies. Quantification by densitometric analysis of Western blot films is presented as the proportion of Ac-NF-κB over total NF-κB (fold induction over N9). Results are shown as mean ± SEM of at least three independent experiments; *P < 0.05

Fig. 6

Treatment with DMF ameliorates EAE and induces an increase in markers of alternatively activated microglia in vivo. a Administration of DMF at clinical onset reduces EAE severity at post-peak/early chronic phase. C57BL/6J mice (n = 20 per group, randomly allocated) were immunized with MOG 35–55 and treated daily by oral gavage with DMF (DMF) or with hydroxypropyl methylcellulose  alone (Vehicle) from the day of disease onset until day 30 post-immunization. Mice (n = 3 per group at each time point) were sampled at indicated time points (arrows) for PCR analysis of brain mRNA (see below). Data are presented as the mean ± SEM daily clinical score (left panel) and as the area under the curve (AUC) of EAE clinical course calculated for each mouse between days 15–30 post-immunization (right panel). *P < 0.05, #P < 0.01. b In vivo treatment with DMF downregulates the expression of pro-inflammatory genes and upregulates genes of the alternatively activated microglia phenotype in the CNS of EAE mice. RNA was isolated from brains of EAE mice treated or not with DMF at indicated time points corresponding to acute (day 15), post-peak (day 22) and chronic (day 30) phases. Expression of the indicated genes was quantified by real-time PCR. Results are shown as mean ± SEM. *P < 0.05, **P < 0.01

Fig. 7

DMF treatment normalizes pre-synaptic abnormalities of glutamatergic transmission in EAE mice. a–e DMF treatment normalizes the frequency, but not the increased decay time and half-width of glutamatergic sEPSCs of EAE mice. sEPSC recordings from single striatal neurons (n = 12 neurons for a total of four mice for each group) were performed on EAE mice sampled between day 20 and day 25 post-immunization. Corticostriatal slices obtained from naïve mouse brain (Control) were used for control recordings (n = 14 neurons recorded for a total of four mice). *P < 0.05. f Representative electrophysiological traces of voltage clamp recordings. g MMF applied directly on control corticostriatal slices induces significant reduction of sEPSC frequency. MMF (1 μM) was added to the bathing ACSF from the indicated time and continuously for 12 min. Recordings were performed on single striatal neurons (n = 10)

Fig. 8

DMF modulates microglia action at synapses. (a–c) The increase in sEPSC decay time and half-width induced upon exposure of control corticostriatal slices to LPS-activated microglia is reverted when activated microglia are treated with MMF at high concentration. Corticostriatal slices obtained from naïve mouse brain were exposed to microglia activated with LPS in the presence or absence of 1 or 100 μM MMF (N9 + LPS + MMF 1 μM or 100 μM) for 30 min prior to electrophysiological recordings from single striatal neurons (n = 11). Control recordings were obtained from untreated slices (Untreated) or slices exposed to N9 cells (N9) or N9 cells activated with LPS (N9 + LPS); ***P < 0.001. d Electrophysiological traces are example of mean peak obtained by group analysis in the presence of activated microglia untreated or treated with high-dose MMF