In order to define the basic properties of microglia within the basal ganglia (BG) of the adult CNS, we used CX3CR1-EGFP transgenic mice to visualize microglia within the ventral tegmental area (VTA), nucleus accumbens (NAc), substantia nigra pars compacta (SNc), and substantia nigra pars reticulata (SNr). Previously we found that microglia populate the VTA at significantly lower density and exhibit sparse branching compared to other BG regions. In contrast, microglia within the SNr are present at a dramatically high density and both SNr and NAc microglia display highly-ramified morphologies. Using electrophysiological recording, we found that microglia in the VTA and the adjacent SNr differed significantly in their membrane capacitance, resting potential, and the expression of voltage-gated potassium channels. We have expanded upon these findings by examining gene expression of microglial cells isolated from distinct BG regions. Following microdissection of the VTA, SNr, NAc, dorsal striatum, and cortex from CX3CR1-EGFP mice, microglial cells were isolated from dissociated tissue using fluorescence activated cell sorting (FACS). Messenger RNA was isolated from sorted cells and RT-PCR was used to examine expression of genes of interest. The microglial-specific gene CD11b was highly expressed in FACS-isolated EGFP+ cells, while expression of the neuron-specific gene NeuN, oligodendrocyte precursor cell specific gene NG2, and astrocyte-specific gene GLT-1 was absent, indicating successful enrichment of pure populations of microglial cells. Preliminary experiments examining purinergic receptors and inflammatory signaling factors indicate that microglial expression of these genes differs across distinct BG regions. These findings suggest that there are functional differences among microglia in different regions of the BG and raise questions about whether BG microglia will exhibit variable responses to pathological insults. We have also used CX3CR1-EGFP mice to visualize microglia within the BG of mice trained to self-administer (SA) cocaine. Density, morphology, and proliferation of microglia was examined in mice that self-administered cocaine for 3 weeks as well as in mice that self-administered cocaine followed by 2 weeks of withdrawal (WD). Quantification of the % of CNS parenchyma covered by microglial processes indicated that microglial process coverage was significantly reduced in the NAc following 3 weeks SA + 2 weeks WD. Similar trends were observed for microglial cells in the VTA and SNr. These observations suggest that microglial cell morphology is altered following chronic exposure to cocaine and that these effects may be more dramatic in some BG regions than others. Changes in microglial cell morphology are correlated with alterations in the release of trophic and inflammatory signaling factors and suggest that microglial interaction with surrounding neurons may be perturbed. We are currently using more rigorous methods to quantify microglial morphology in SA and SA+WD mice. Furthermore, we are using FACS isolation of BG microglia to carry out analysis of the whole microglial transcriptome using RNAseq, a type of next generation sequencing. In future experiments, this technique will be used for unbiased examination of the microglial transcriptome in mice that have self-administered cocaine or saline. Such experiments will clarify the degree to which microglial cell biology may be altered in the absence of overt morphological reactivity and could identify novel signaling pathways that regulate interactions between microglia and neurons. Previously we used CX3CR1-EGFP mice to quantify microglial cell density and morphology during early postnatal periods and found that the regional heterogeneity observed in the adult is established early in development. To build upon this analysis, we used immunostaining for markers of presynaptic terminals and confocal imaging to determine whether microglia in the developing reward circuitry participate in synaptic pruning through phagocytosis of presynaptic terminals. Punctate structures immunolabeled for vesicular glutamate transporter 1 (VGlut1) and 2 (VGlut2) and vesicular gaba transporter (VGAT) were found colocalized with the EGFP+ processes of microglia in postnatal day 12-20 mice within the VTA and NAc. This observation suggests that microglia in the developing BG are engulfing presynaptic terminals as has been observed within the hippocampus and visual system and raises important questions about the developmental consequences of perturbations of microglial cell function, such as childhood infections that penetrate the CNS. Future experiments in this project will be aimed at determining the peak developmental period during which microglial engulfment of synapses takes place. Electron microscopic analysis will then be used to confirm observations made with immunostaining and light microscopy. The data summarized above expand our understanding of microglial cells within the developing and adult BG. To begin to define whether these cells influence the membrane properties and synaptic transmission of neurons within the BG, we are using transgenic strategies to ablate microglia within the CNS. CX3CR1-CreER mice, which express inducible cre recombinase within microglia have been crossed to both rosa-flox-stop-diphtheria toxin (rfs-DTA) mice and rosa-flox-stop-diphtheria toxin receptor (rfs-DTR) mice. In double transgenic mice, injection of 4-hydroxytamoxifen activates Cre within microglia and those cells begin expressing DTA or DTR. In CX3CR1-CreER;rfs-DTA mice, this direct expression of diphtheria toxin within microglia causes the cells to undergo programmed cell death. In CX3CR1-CreER;rfs-DTR mice, subsequent injections with diphtheria toxin will only affect those cells expressing DTR, again pushing microglia into programmed cell death. Using immunohistochemistry, we have identified a dose of 4-hydroxytamoxifen that can effect a temporary ablation of 99% of the microglia within the CNS of CX3CR1-CreER heterozygous;rfs-DTA homozygous mice. By 5 days post-ablation, microglia have repopulated 30-60% of the CNS. Preliminary electrophysiological experiments in acute brain slices from microglial-ablated mice indicate that neurons in the SNr of ablated mice have significantly reduced hyperpolarization-activated potassium current (Ih), a current which is important for regulating the firing patterns of these neurons. Ongoing experiments are aimed at determining whether similar changes in Ih occur in dopaminergic neurons of the VTA. Preliminary behavioral experiments indicate that female microglial-ablated mice have reduced anxiety-like behavior in the elevated plus maze paradigm. In addition, female microglial-ablated mice have more pronounced cocaine conditioned place preference and cocaine locomoter sensitization. Together these observations suggest that microglia interact with neurons of the basal ganglia to regulate neuronal membrane properties and BG-dependent behaviors. Future experiments will be aimed at confirming these preliminary observations and identifying possible mechanisms underlying neuron-microglial interactions in the BG. In addition, we hope to obtain a small molecule inhibitor of colony stimulating factor receptor 1 (CSFR1), manufactured by Plexikkon, that can be used to ablate microglia for long periods. Prolonged microglial ablation using this antagonist will allow investigation of mouse behavior in longer, drug self-administration paradigms.
| De Biase, Lindsay M; Schuebel, Kornel E; Fusfeld, Zachary H et al. (2017) Local Cues Establish and Maintain Region-Specific Phenotypes of Basal Ganglia Microglia. Neuron 95:341-356.e6 |
