Microglia modulate hippocampal synaptic transmission and sleep duration along the light/dark cycle

Abstract

Microglia, the brain's resident macrophages, actively contribute to the homeostasis of cerebral parenchyma by sensing neuronal activity and supporting synaptic remodeling and plasticity. While several studies demonstrated different roles for astrocytes in sleep, the contribution of microglia in the regulation of sleep/wake cycle and in the modulation of synaptic activity in the different day phases has not been deeply investigated. Using light as a zeitgeber cue, we studied the effects of microglial depletion with the colony stimulating factor-1 receptor antagonist PLX5622 on the sleep/wake cycle and on hippocampal synaptic transmission in male mice. Our data demonstrate that almost complete microglial depletion increases the duration of NREM sleep and reduces the hippocampal excitatory neurotransmission. The fractalkine receptor CX3CR1 plays a relevant role in these effects, because cx3cr1^GFP/GFP mice recapitulate what found in PLX5622-treated mice. Furthermore, during the light phase, microglia express lower levels of cx3cr1 and a reduction of cx3cr1 expression is also observed when cultured microglial cells are stimulated by ATP, a purinergic molecule released during sleep. Our findings suggest that microglia participate in the regulation of sleep, adapting their cx3cr1 expression in response to the light/dark phase, and modulating synaptic activity in a phase-dependent manner.

Keywords: cx3cr1; electroencephalography; long-term potentiation; microglial depletion; miniature excitatory post-synaptic currents; sleep; spontaneous excitatory post-synaptic currents.

FIGURE 1

Effect of microglial depletion on sleep and movement duration in the dark and light phases. Mean ± SEM and individual distribution of non‐rapid eye movement (NREM) sleep (a), rapid eye movement (REM) sleep (b) and movement (c) duration in 7 h of light versus 7 h of dark in C57BL/6 (control, C. n = 10) and PLX5622‐treated (n = 9) mice. The within‐group statistical analysis was performed by Wilcoxon test for paired measures and showed statistical differences for NREM sleep (C ## p = .0025 corrected, z = 2.80, t = 0.00; PLX5622 # p = .0054 corrected, z = 2.54, t = 0.00), REM sleep (C # p = .014 uncorrected, z = 2.19, t = 0.00) and movement duration (C ## p = .0025 corrected, z = 2.80, t = 0.00; PLX5622 ## p = .0038 corrected, z = 2.66, t = 0.00), in light versus dark conditions. The between‐group statistical analysis was performed by Mann–Whitney U test for unpaired measures and showed increased NREM sleep in PLX5622‐treated mice compared to C (**p = 0.0031 corrected, z = −2.74, U = 11.00), (d–f) time‐courses of NREM (d), REM (e) and movement (f) duration at the time‐points analyzed. Data are expressed as a percentage of the total time analyzed and are shown as mean ± SEM

FIGURE 2

The depletion of microglia alters synaptic transmission across the light–dark cycle. (a) Top: Representative traces of mEPSCs recorded at −70 mV from hippocampal CA1 neurons at ZT4 and ZT16 in control (C) and PLX5622 conditions; scale bars: 2 s (horizontal), 10 pA (vertical). bottom: histograms of the mean amplitude (left) and frequency (right) for mEPSCs at ZT4 and ZT16 in C and in PLX5622 conditions. Miniature EPSC amplitude is increased in C during dark condition (ZT16: 10/4 cells/mice; ZT4: 16/4, t = 2.332, p = .025) and PLX5622 treatment reduces mEPSC amplitude at ZT16 (ZT16 PLX5622: 13/4 cells/mice, t = 2.066, p = .045) compared to C, abolishing the difference at the two times considered (ZT4 PLX5622: 13/5, post hoc 13/4, t = 0.521, p = .605). Frequencies of mEPSC were similar in all groups. (b) Top: representative traces of sEPSC at ZT4 and ZT16 in control (C) and PLX5622 conditions; bottom: histogram of the mean amplitude (left) and frequency (right) for sEPSCs in control conditions (C) and after PLX5622 treatment. Spontaneous EPSC amplitude is increased in C during dark condition (ZT16: 10/4 cells/mice; ZT4: 10/4, t = 2.855, p = .007) and reduced upon PLX5622 treatment compared to control (ZT16 PLX5622: 10/4 cells/mice, t = 3.310, p = .002). The frequency of sEPSC is increased in C at ZT16 (t = 2.167, p = .037) but reduced after PLX5622 treatment only at dark (t = 2.292, p = .028). (c) Left. The cumulative probability curve for sEPSC amplitude is shifted rightward at ZT16 (n = 1200 events) compared to ZT4 (n = 900 events) in C (KS, p < .001), reflecting increased sEPSC amplitude. Right. The cumulative function for the inter‐event intervals (right) is shifted to the left in control (C) at ZT16 (KS, p < .0001), in line with the increased sEPSC frequency. Upon PLX5622 treatment the cumulative functions for both sEPSC amplitude and IEI were similar at the two time points (KS, p = .28 and p = .39). (d) Representative fEPSP traces (left) and mean values (right) for PPR experiments performed (ISI = 50 ms) at ZT4 and ZT16 in control (17/9 and 31/16 slices/mice) and PLX5622 treated mice (16/7 and 17/7 slices/mice respectively. PPR is increased in PLX5622‐treated mice at ZT16 (t = 3.012, p = .004). Scale bars: 0.3 mV (vertical), 10 ms (horizontal). Data are shown as mean ± SEM. Statistical analysis was performed with Two‐way ANOVAs, Holm‐Sidak post hoc comparison. *, # p < .05, **, ## p < .01. Cumulative probability functions were compared with Kolmogorov–Smirnov test

FIGURE 3

Microglial expression of cx3cr1 decreases during the light phase and upon ATP stimulation in vitro. (a) cx3cr1 mRNA expression analysis by RT‐PCR in CD11b + cells obtained at ZT4 and ZT16 from hippocampus (n = 12 and 13 mice respectively), PFC (n = 9 and 12 mice respectively) and hypothalamus (n = 15 and 13 mice respectively) and individual distribution of data. Data are expressed as cx3cr1 mRNA fold increase normalized to gapdh expression; values shown are normalized to ZT16. cx3cr1 mRNA levels are increased during dark in all regions analyzed. Data are expressed as mean ± SEM. * p = .034 (hippocampus) and p = .03 (PFC); ** p < .001 (one‐way ANOVA). (b) cx3cr1 mRNA expression analysis by RT‐PCR in primary wild type microglia obtained from pups treated for 4 h with LPS (100 ng/ml) and ATP at different concentrations. LPS treatment and ATP at 1 and 100 μM decrease cx3cr1 mRNA levels. Data are expressed as cx3cr1 mRNA fold increase normalized to gapdh expression; values shown are normalized to vehicle (c). Data are expressed as the mean ± SEM. ** p < .001 (one‐way ANOVA, followed by Dunn's post hoc test). (c) Median fluorescence intensity (MFI) of CX3CR1 staining is decreased in primary wild type microglia stimulated for 4 h with LPS (100 ng/ml) and ATP 1 mM and 100 μM. Data are expressed as mean ± SEM of the MFI. ** p < .001 C versus LPS, * p = 0.006 C versus ATP 1 mM, one‐way ANOVA followed by Holm‐Sidak post hoc test. (d) cx3cr1 mRNA expression by RT‐PCR is also decreased in primary wild type microglia obtained from adult mice following 4 h of ATP (100 μM) treatment. Data are expressed as mean ± SEM of cx3cr1 mRNA fold increase normalized to gapdh expression; values shown are normalized to vehicle (c). ** p < 0.001 (Student's t‐test)

FIGURE 4

c3cr1 deficiency effects on sEPSC and video‐EEG recordings. Mean ± SEM and individual distribution of non‐rapid eye movement (NREM) sleep (a), non‐rapid eye movement (REM) (b) sleep and movement (c) duration in 7 h of light versus 7 h of dark in cx3cr1GFP/GFP mice (n = 7) and control mice (C; n = 4). The within‐group statistical analysis was performed by Wilcoxon test for paired measures and showed statistical differences for NREM sleep (C # p = .03 uncorrected, z = 1.83, t = 0.00; cx3cr1GFP/GFP ## p = .0089 uncorrected, z = 2.36, t = .00), REM sleep (C # p = .04 uncorrected, z = 1.10, t = 2.00) and movement duration (cx3cr1GFP/GFP ## p = 0.0089 uncorrected, z = 2.36, t = 0.00), in light versus dark conditions. The between‐group statistical analysis was performed by Mann–Whitney test for unpaired measures and showed increased NREM sleep (* p = 0.04 uncorrected, z = 1.80, U = 4.00) in cx3cr1 ^GFP/GFP mice respect to (c,d) representative traces of spontaneous EPSC (top) and histograms for the mean amplitude and frequency (bottom) in controls (C) and cx3cr1 ^GFP/GFP mice at ZT4 and ZT16. The amplitude of sEPSC is increased in C during dark (ZT16 C: 12/3 cells/mice, ZT4 C: 15/3, t = 2.084, p = 0.043) and reduced in cx3cr1 ^GFP/GFP mice in dark (13/3, t = 2.084, p = .042). Scale bars: 1 s (horizontal), 20 pA (vertical). (e) Cumulative probability function for sEPSCs amplitudes compared at the two daily times in controls (ZT4:1121 and ZT16:1267 events) showed a rightward shift at ZT16 (KS, p < .0001, top left). The IEI cumulative function is shifted to the left in C at ZT16 indicating an increased sEPSC frequency (KS, p = .0036, top right). No shift is detected in cx3cr1 ^GFP/GFP mice for sEPSC amplitude and IEI at the two time points considered (KS, p = .748 and p = .503, bottom). (f) Representative fEPSP traces for PPR experiments performed (ISI = 50 ms) at ZT4 and ZT16 in controls and cx3cr1 ^GFP/GFP . (g) Histogram of the mean values of PPR at ZT4 and ZT16 in control (C, ZT4: 8/4 and ZT16: 13/5 slices/mice) and cx3cr1 ^GFP/GFP mice (ZT4: 8/4 and ZT16: 15/6 slices/mice). Paired‐pulse ratio values are decreased in control (t = 2.924, p = .005) and increased in cx3cr1 ^GFP/GFP mice (t = 2.144, p = .038) at ZT16. Scale bars: 0.3 mV (vertical), 10 ms (horizontal). Data are shown as mean ± SEM statistical analysis was performed with two‐way ANOVAs, holm‐Sidak post hoc comparison. *, # p < .05, **, ## p < .01. Cumulative probability functions were compared with Kolmogorov–Smirnov test