Deficiency of the microglial receptor CX3CR1 impairs postnatal functional development of thalamocortical synapses in the barrel cortex

Abstract

Accumulative evidence indicates that microglial cells influence the normal development of brain synapses. Yet, the mechanisms by which these immune cells target maturating synapses and influence their functional development at early postnatal stages remain poorly understood. Here, we analyzed the role of CX3CR1, a microglial receptor activated by the neuronal chemokine CX3CL1 (or fractalkine) which controls key functions of microglial cells. In the whisker-related barrel field of the mouse somatosensory cortex, we show that the recruitment of microglia to the sites where developing thalamocortical synapses are concentrated (i.e., the barrel centers) occurs only after postnatal day 5 and is controlled by the fractalkine/CX3CR1 signaling pathway. Indeed, at this developmental stage fractalkine is overexpressed within the barrels and CX3CR1 deficiency delays microglial cell recruitment into the barrel centers. Functional analysis of thalamocortical synapses shows that CX3CR1 deficiency also delays the functional maturation of postsynaptic glutamate receptors which normally occurs at these synapses between the first and second postnatal week. These results show that reciprocal interactions between neurons and microglial cells control the functional maturation of cortical synapses.

Figure 1.

CX3CL1/R1 signaling controls the entry of microglia into TCA clusters. A, B, Merge confocal images of tangential sections through layer 4 showing TCA (red, anti-5-HTT) and fluorescent microglia (green) of the barrel field of P5, P7, and P9 CX3CR1^+/eGFP (A) and CX3CR1^eGFP/eGFP (B) mice. C, Confocal images of a tangential section through layer 4 of a P7 CX3CR1^+/+ mouse showing microglia (green, anti-Iba1), TCAs (red, anti-VGlut2) and nuclei (blue, TO-PRO-3). D, Comparison of the ratio of microglia number inside and outside TCA clusters in CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice. *p < 0.02, **p < 0.01 between CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice; ^#p < 0.05, ^##p < 0.01 between P5 and other ages; Mann–Whitney U test. E, Coronal section of the barrel cortex of a P7 CX3CR1^eGFP/eGFP used to count microglial cells (green) in layer 4 (red, TCAs, anti-5-HTT). F, Comparable microglia density in layer 4 of CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice (p = 0.44, Mann–Whitney U test). G, Postnatal development of CX3CL1 (fractalkine) immunoreactivity in coronal sections of the mouse somatosensory cortex. Scale bars: A, E, 100 μm; C, 50 μm; G, 200 μm. Numbers above bar histograms refer to animal numbers.

Figure 2.

CX3CR1 deficiency impairs the functional maturation of thalamocortical synapses. A, DIC image of a P7 thalamocortical slice with the recording pipette in a barrel and the bipolar stimulating electrode in the internal capsule. Scale bar, 200 μm. B, Peak amplitude of individual (black dots) and mean (white circles) thalamocortical EPSCs plotted as a function of the stimulation intensity for the determination of the minimal stimulation (22 V in this example). C, Same cell as in B, thalamocortical EPSCs evoked by two stimulations for the determination of the paired-pulse ratio. D, Similar paired-pulse ratios of thalamocortical EPSCs in CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice at P5, P7, and P9. Statistical differences between P5 and other ages within each genotype are indicated. E, Effect of NBQX and d-AP-5 on thalamocortical EPSCs evoked in a P9 neuron. F, AMPAR- and NMDAR-mediated EPSCs in layer 4 neurons of P5 and P9 CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice. G, Comparison of the AMPAR/NMDAR ratio for P5, P7, and P9 CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice. Each trace is an average of 15–20 individual sweeps. *p < 0.05, **p < 0.01 (unpaired t test with Welch correction).

Figure 3.

CX3CR1 deficiency impairs the GluN2B to GluN2A developmental switch occurring at thalamocortical synapses between the first and second postnatal week. A, Effect of 300 nm Ro256981 on averaged (15–20 sweeps) NMDAR-mediated EPSCs in P10 neurons of CX3CR1^+/eGFP (upper traces) and CX3CR1^eGFP/eGFP (lower traces) mice. B, Summary of Ro256981 effects on NMDAR-mediated current charges for P9–P10 neurons. C, Normalized average traces of NMDAR-mediated currents recorded in layer 4 neurons of P5 (black) and P10 (red) of CX3CR1^+/eGFP (upper traces) and CX3CR1^eGFP/eGFP (lower traces) mice. D, Comparison of the weighted decay time constant in P5–P7, P9–P10, and adult CX3CR1^+/eGFP and CX3CR1^eGFP/eGFP mice. *p < 0.05, **p < 0.01,***p < 0.001 (unpaired t test with Welch correction).