(A) Schematic of the protocol schedule for cocaine induced CPP training, testing and microglia depletion during withdrawal. (B) Representative images of Iba1+ immunofluorescence labeling in the NAc shell at WD21 under the indicated conditions. Nuclei were counterstained with DAPI (not shown). (C) The density of Iba1+ cells was significantly reduced in both saline- and cocaine- treated mice after 20 days of PLX5622 supplemented diet (Diet: F(1, 45) = 465.0, p<0.05; N=12 fields/4 mice per group). (D) Place preference score was higher in cocaine-treated mice than saline-treated mice both at WD1 and at WD21 (Drug: F(1, 77) = 56.95, p<0.05) independently of PLX5622 treatment. (E) Place preference score at WD21 time point (Drug: F(1, 77) = 20.07, p<0.05). (F) The distance traveled in the apparatus during tests increases from WD1 to WD21 in cocaine control mice (interaction Drug x Diet x WD: F(1, 77) = 4.960, p<0.05). (G) Distance traveled during test at WD21 time point (interaction Drug x Diet: F(1, 77) = 5,456, p<0.05). (H) The frequency of crossings to side-chambers increases from WD1 to WD21 in cocaine control mice (interaction Drug x Diet x WD: F(1, 77) = 7.785, p<0.05). (I) Crossing frequency during the test at WD21 time point (interaction Drug x Diet: F(1, 77) = 4.297, p<0.05). For convenience, behavioral data from mice in the first behavioral experiment and subsequent experiments (Fig.2) are pooled together (saline control N=16 mice, saline PLX N=17 mice, cocaine control N=23 mice, cocaine PLX N=25 mice). Data are presented as mean ± SEM. Scale bars = 50 μm. The symbols indicate, & p<0.05 compared to control diet, # p<0.05 compared to saline, ‡ p<0.05 cocaine control WD1 vs cocaine control WD21, * p<0.05 compared to saline control, saline PLX and cocaine PLX, $ p<0.05 compared to cocaine PLX, by using Tukey’s test for multiple comparisons).