Microglia play beneficial roles in multiple experimental seizure models

Abstract

Seizure disorders are common, affecting both the young and the old. Currently available antiseizure drugs are ineffective in a third of patients and have been developed with a focus on known neurocentric mechanisms, raising the need for investigations into alternative and complementary mechanisms that contribute to seizure generation or its containment. Neuroinflammation, broadly defined as the activation of immune cells and molecules in the central nervous system (CNS), has been proposed to facilitate seizure generation, although the specific cells involved in these processes remain inadequately understood. The role of microglia, the primary inflammation-competent cells of the brain, is debated since previous studies were conducted using approaches that were less specific to microglia or had inherent confounds. Using a selective approach to target microglia without such side effects, we show a broadly beneficial role for microglia in limiting chemoconvulsive, electrical, and hyperthermic seizures and argue for a further understanding of microglial contributions to contain seizures.

FIGURE 1.

Pharmacological microglial elimination and repopulation. a-b, Representative images (a) and quantification of the percent of microglial cells (b) from the cortex and hippocampal dentate gyrus of CX3CR1^GFP/+ treated with PLX3397 chow (660mg/kg) at 0, 1, 4 and 7 days of treatment. n = 3 mice per group. c-d, Schematic representation of splenic (c) and bone marrow (d) isolation following a 7-day treatment with control of PLX3397 chow for processing by flow cytometry of various immune cells including T cells, B cells, neutrophils, inflammatory and patrolling monocytes. n = 3 mice per group (e-g), Representative images (e) and quantification (f-g) of the density of microglial cells from the cortex of CX3CR1^GFP/+ treated with control or PLX3397chow for 7 days and then withdrawn from the PLX3397 chow at 0, 1, 4 and 7 days. n = 2–4 mice per group. Data presented as mean ± s.e.m f and g. Statistics calculated by Student’s T-test, *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001.

FIGURE 2.

Endogenous and repopulated microglia are protective in chemoconvulsive seizures. (a) Diagram of experimental scheme for microglial depletion and chemoconvulsive kainic acid (KA) seizures. (b and c) Overall (b) and area under the curve (AUC, c) Racine seizure scores from control and PLX3397 treated mice monitored over 240 mins (4h). (d) Percent animal survival by 4h of KA treatment. n = 18–20 mice per group. (e) Diagram of experimental scheme for microglial depletion and subsequent repopulation followed by KA seizure induction after 7 days. (f and g) Overall (f) and AUC (g) Racine seizure scores from control and PLX3397 treated mice monitored over 120 mins (2h). n = 20 mice per group. (h) Diagram of experimental scheme for microglial depletion and subsequent repopulation followed by KA seizure induction after 28 days. (i and j) Overall (i) and AUC (j) Racine seizure scores from control and PLX3397 treated mice monitored over 120 mins (2h). n = 16–20 mice per group. Data presented as median in b, f and I and mean ± s.e.m in c, g and j. Statistics calculated by Student’s T-test and Chi-squared with Fisher’s exact test in d. ****p < 0.0001.

FIGURE 3.

Microglia are important in the recovery from chemoconvulsive seizures. (a) Diagram of experimental scheme for microglial depletion and behavioral assays. (b-d) Mice weights (b) distance travelled (c) and degree of mobility (d) exhibited by mice in an open field without KA treatment. n = 10–18 mice per group. (e) Diagram of experimental scheme for microglial depletion followed by KA seizure induction after 7 days and behavioral assays for 7 days. (f and g) Overall (f) and area under the curve (AUC, g) Racine seizure scores from control and PLX3397 treated mice monitored over 240 mins (4h). n = 8 mice per group. (h-j) Mice weights (h) distance travelled (i) and degree of mobility (j) exhibited by mice in an open field following KA treatment. n = 4–6 mice per group. Data presented as mean ± s.e.m except in f where data is presented as median. Statistics calculated by Student’s T-test for b, c, d and g and Two-way ANOVA for h-j. *p < 0.05; **p < 0.01.

FIGURE 4.

Microglia are beneficial in electrically induced seizures. (a) Diagram of experimental scheme for microglial depletion and chronic hippocampal stimulation (CHS) after 7 days using C57Bl/6J mice (n=16 per group). (b) Evolution of spikes during CHS stimulation (see Methods for details). (c) Time to CHS-evoked discrete seizures. (d) Time to CHS-induced death from start of CHS. (e) Diagram of experimental scheme for microglial depletion for 7 days followed by electrical kindling. (f) Standardized seizure durations elicited during kindling of VGAT-Cre mice. n = 9 – 11 mice per group. (g) Kaplan-Meyer plot of the number of stimulations required to elicit the full kindled state (5 evoked bilateral tonic-clonic seizures with loss of balance) in VGAT-Cre mice. (h) Quantification of Iba1⁺ cells in control and PLX3397-treated mice. n = 4 mice per group. Data presented as mean ± s.e.m. Statistics calculated by Student’s t-test (b, d, f), linear regression (c), Log-rank Mantel Cox test (g), or 2-way ANOVA followed by Sidak’s test (h). Significance is denoted by: *, p < 0.05; **, p < 0.01; ***, p < 0.001 and ****, p < 0.0001.

FIGURE 5.

Microglia are beneficial following secondary and spontaneously occurring seizures. (a) Diagram of experimental scheme for two KA exposures with a 7-day microglial depletion between the KA treatments. (b and c) Median Racine seizure scores from a single and following a double bout of KA treatment in control (b) and PLX3397 (c) treated mice n = 10–20 mice per group. (d and e) Overall (d) and average (e) Racine seizure scores from a second bout of KA-induced seizures following control and PLX3397 treatment monitored over 180 mins (3h). n = 8 mice per group. (f) Diagram of experimental scheme for hybrid kindling followed by monitoring of spontaneous recurrent seizures (SRS) with microglial depletion. (g) Quantification of the number SRS over a 2-week period with microglial depletion. (h and i) Time course of spontaneous seizures after chow was changed (day 0, binned in 2 day intervals). n = 7 mice per group Data were analyzed by area under curve (AUC) and the results are plotted in (i). Data presented as median in b – d and mean ± s.e.m in e, g and i. Statistics calculated by Student’s t-test *p < 0.05; ****p < 0.0001.

FIGURE 6.

Microglia are beneficial in hyperthermia-induced seizures in developing mice. (a) Image of hyperthermic (febrile) seizure setup with a mouse in a chamber made of plexiglass with heat provided by a heating plate below and a heat lamp from above. Temperature in chamber can be set and measured externally. (b) Schematic of a mouse head with implanted electrode in the left cortex (LC), right cortex (RC), left hippocampus (LH) and right hippocampus (RH). (c) Electrical activity is detected in all four regions upon hyperthermia exposure. (d) The percent of mice that show behavioral convulsions under hyperthermia at the different temperatures. (e) cFos expression in the cortex and dentate gyrus of normothermic and hyperthermic conditions as well as with microglial depletion. (f) Efficient microglial elimination with PLX3397 treatment in developing mice. (g-i) Microglial elimination and its effects of time to the first seizures (g) percent of mice that seize at 41°C (h) and mortality rate (i) during hyperthermia-induced seizures in developing mice. n = 13 mice per group. Statistics calculated by Student’s T-test and Chi-squared with Fisher’s exact test in i. **p < 0.01; ****p < 0.0001.