Supplementary Figure 12: Genetic deletion of Eed in adult microglia leads to global changes in the adult brain.

(a) The number of neurons in the striatum with Eed-deficient microglia remains unchanged. Quantification of the DAPI+ cells (control: mean=223.1, SEM=14.90; mutant: mean=242.9, SEM=9.861; p=0.4079; F=3.425, t₃=0.9599) and NeuN+ neurons (control: mean=67.04, SEM=12.72; mutant: mean=107.4, SEM=7.274; p=0.1020; F=4.585, t₃=2.331). n=3 and 2 for control and mutant, respectively. (b) The survival of neurons in the striatum with Eed-deficient microglia remains unchanged. Quantification of the percentage of cCASP3+ neurons in control (mean=8.063, SEM=1.174) and mutant mice (mean=10.11, SEM=1.455). p=0.2997, F=1.537, t₁₀=1.094. (c) Cx3cr1^CreErt2/+;Eed^loxP/loxP mice display normal behavior in the open field paradigm at 3 months (n=5/genotype) and 15 months of age (n=12/genotype). Bar graphs with individual data points show total distance moved in cm (left: Eed^loxP/loxP control: mean=2402, SEM=174.7; mutant: mean=2181, SEM=372.0; p=0.6052, F=4.534, t₈=0.5380; right: control: mean=2144, SEM=235.9; mutant: mean=2003, SEM=192.1; p=0.6479, F=1.508, t₂₂=0.4630). (d) Cx3cr1^CreErt2/+;Eed^loxP/loxP mice show reduced response to cocaine. Total distance traveled is shown in 5 minute bins (control: black; mutant:grey), during 30-minute habituation period and the 60 minutes following cocaine injection on day 1 (control: mean=533.0, SEM=34.97; mutant: mean=458.4, SEM=41.79; p=0.5791, F=0.3170) and day 7 (control: mean=1688, SEM=217.6; mutant: mean=997.0, SEM=191.7; p=0.0453, F=4.504). Two-way ANOVA with repeated measures (n=12 mice/genotype). (e) Eed-deficiency in cbMg does not cause a change in Purkinje cell spine density. Control: mean=2.783, SEM=0.1092; mutant: mean=2.457, SEM=0.1672; p=0.1905, F=6.247, t₄=1.574 (n=3/genotype). (f) Cx3cr1^CreErt2/+;Eed^loxP/loxP mice display normal motor learning or balance by rotarod test at 3 month (control: mean=233.2, SEM=13.00; mutant: mean=242.0, SEM=11.86; p=0.6655, F=0.19) and 15 month of age (control: mean=177.0, SEM=8.765; mutant: mean=162.2, SEM=9.176; p=0.6655, F=0.19). Two-way ANOVA with repeated measures (young: n=5/genotype; old: n=11 control, 7 mutant mice). (g) Loss of H3K27me3 in cxMg leads to a progressive gain of cbMg signature in Cx3cr1^{CreErt2/+(Litt)}; Eef1a1^{LSL.eGFPL10a/+}; Eed^loxP/loxP mice. Left: Box-and-whisker plots show mean relative expression of the 27 cbMg signature genes that are up-regulated in Eed-deficient cxMg at 9 months. Control: ctrl: min=-2.657, 25%=-0.833, median=-0.68, 75%=-0.45, max=-0.30; 3mo: min=-2.46, 25%=-0.68, median=-0.439, 75%=0.055, max=0.854 ; 6mo: min=-1.68, 25%=-0.039, median=0.231, 75%=0.515, max=1.62; 9mo: min=0.588, 25%=1.014, median=1.19, 75%=2.022, max=2.788; p < 0.0001, F=68.56, KWS=75.96. Right: Heatmap shows the variation in the expression levels (z-scored log2 RPKM) of 27 cbMg-enriched genes in 3,6 & 9mo Eed-deficient cxMg and control cxMg. box-and-whisker plots (Tukey) show mean and variance for normalized expression values in z-scored log2-rpkm. Shapiro-Wilk normality p < 0.0001 for all samples. One-way ANOVA (Kruskal-Wallis test) with Dunn’s Multiple Comparison post-hoc analysis to compare all pairs of columns. **p ≤ 0.01, ***p ≤ 0.001. (h) Morphology of Eed-deficient cxMg reflect a gain of clearance phenotype. CD68+ lysosome content (red) in YFP+ microglia (green) was determined using immunofluorescence analysis from cortex of brain sections from 12-month-old control Cx3cr1^CreErt2/+;Eed^loxP/+ and mutant Cx3cr1^CreErt2/+;Eed^loxP/loxP mice (DAPI: blue). Scale: 10 μm. Representative image (left) and quantification of the lysosomal area / microglia area (right) is shown. Control: mean=0.04767, SEM=0.01230); mutant: mean=0.09713, SEM=0.007464; p=0.0263, F=2.716, t₄=3.438 (12 images from n=3/genotype). (i) Microglia-specific Eed deficiency is associated with a reduction in spine density of cortical pyramidal cells. Representative images of Golgi-stained neuronal processes of the layer IV-V pyramidal neurons from 15-month-old control, Eed^loxP/loxP, and Cx3cr1^CreErt2/+;Eed^loxP/loxP mice. Scale: 2 μm. Bar graph with individual data points show total spine densities (left) of layer IV-V cortical pyramidal neurons of control mice: mean=1.098, SEM=0.03408; and mutant mice: mean=0.8713, SEM=0.06293; p= 0.0337, F=3.411, t₄=3.174); and mushroom spine densities of layer IV-V cortical pyramidal neurons of control mice: mean=0.5369, SEM=0.01514; and mutant mice: mean=0.4271, SEM=0.01488; p=0.0066, F=1.035, t₄=5.177. 28 dendrites from n=3 mice / genotype. (j) The number of neurons in the cortex with Eed-deficient microglia remains unchanged. Quantification of the DAPI+ cells (control: mean=2773, SEM=193.8; mutant: mean=2844, SEM=116.5; p=0.7700; F=2.764, t₄=0.3129) and NeuN+ neurons (control: mean=1133, SEM=42.70; mutant: mean=1209, SEM=44.83; p=0.2838; F=1.102, t₄=1.237). n=3/genotype. Bar graphs with individual data points show mean ± SEM, line graphs show mean ± SEM, t-tests are two-tailed unpaired unless otherwise specified.