Fig 2: Effect of microglia depletion with PLX5622 on ZIKV infection of young adult mouse brain. a In a first experiment, TRIF-/- × IPS-1-/- mice received control or PLX5622 diet ad libitum for 10 days. Mice were sacrificed, and the brain was taken for immunofluorescence analysis. Representative micrographs illustrating the depletion of Iba1+/TMEM119+ microglia after immunofluorescence labeling in the stratum radiatum of untreated (upper images) and PLX5622-treated (lower images) mice. Brain sections were counterstained with DAPI. Scale bars on the picture are equivalent to 250 µm. b–d In a second experiment, TRIF-/- × IPS-1-/- mice received control or PLX5622 diet ad libitum for 10 days. Mice were then infected intravenously with 1 × 105 PFUs of ZIKV strain PRVABC59 (in 100 µL volume), and the respective diets were continued until sacrifice. b Body weight changes of untreated (?) and PLX5622-treated (¦) mice after ZIKV challenge. Results are the mean ± SEM of 3 mice per group and per time point. c–d Subsets of mice were sacrificed on day 0 (noninfected), 3, 7, and 10 post-infection, and viral RNA load was determined by reverse transcriptase ddPCR in serum (c) and brain homogenates (d). The dotted lines represent the limits of detection of the reverse transcriptase ddPCR assay. Results are the mean ± SEM of 3 mice per group and per time point. Statistical analyses were performed using a one-way analysis of variance with Tukey’s multiple comparison test. Results that are statistically different between indicated groups are shown as follows: *** P < 0.001

Fig 3: Confocal analysis of the time course of D-Cy5 in the cells of the lamina and optic nerve in rNAION and control eyes.A-I: sections of the retina:optic nerve junction region in naïve (A,D,G), 1 week post-rNAION induction (B,E,H), and 4 weeks post-induction (C,F,I) eyes. J-L: D-Cy5 localization in the distal optic nerve of an animal 1week post-rNAION induction. A,B,C. DAPI/D-Cy5 colocalization. Labeled D-Cy5 is external to the lamina in the naïve/non-induced eye (A) but accumulates at the primary rNAION lesion site (B). D-Cy5 is still present at 4 weeks post-induction (C), but the overall signal, although present, is less intense and less extensive. No D-Cy5 signal is present within the distal ON 1 week post-rNAION induction (J). D,E,F. D-Cy5/IBA1 co-localization within resident and invading inflammatory cells. There is no dendrimer signal co-localization with microglia in the lamina or the ON of the naïve eye 1 week post-injection (D). D-Cy5 colocalizes with IBA1(+) cells in the lamina and anterior ON of rNAION-induced eyes at both 1 week (E; red arrows) and 4 weeks post- induction (F; red arrows). G,H,I. Analysis of D-Cy5 and Aldh1L1 (astrocyte cytoplasm). G. There is no co-localization in the non-induced eye. H. D-Cy5 /Aldh1L1 co-localization is present 1 week post-induction (inset: higher magnification of same region). I. D-Cy5/Aldh1L1 continues to co-localize in the lamina region/anterior ON 4 weeks post-induction, but at a reduced level (inset:higher magnification of same region. There is minimal D-Cy5 present in the distal ON 1 week post-induction (J) and no co-localization of D-Cy5 with either IBA1(+) microglia (K) or Aldh1L1(+) astrocytes (L) at this time. Scale bar in B: 100µm.

Fig 4: D-Cy5 persists in astrocytes and macrophages in the lamina after rNAION.Higher magnification (20X) confocal analysis of the same regions seen in Fig 3 at 1 (A-C) and 4 (D-F) weeks. A. Three-channel (astrocyte, inflammatory cell, DAPI nuclear stain) confocal photo of D-Cy5 localization in the rat lamina 1week post-induction. Strong D-Cy5 signal is scattered within the lamina. B. Dual-channel photo showing D-Cy5 localization at high levels in astrocyte cytoplasm (arrows) in the lesioned area, revealed by Aldh1L1 immunostaining. C. Dual-channel photo showing D-Cy5 localization in inflammatory cells (arrows). D. Three-channel confocal photo of D-Cy5 localization in the rat lamina 4 weeks post-induction. D-Cy5 is reduced overall in the lesion area. E. Identification of astrocyte-specific D-Cy5 co-localization. D-Cy5 is still concentrated in astrocytes (arrows) but at considerably lower signal levels than at one week. F. Dual-channel (inflammatory cell-specific) D-Cy5 co-localization. D-Cy5 is still present in IBA1(+) cells (arrows) but at lower signal levels than at 1 week. Scale bars: 100µm.

Fig 5: PLX treatment for 28 days leads to microglia depletion. (A) The confocal images show perfect colocalization of tamoxifen-induced tdTomato (magenta) signal in Cx3Cr1-expressing cells and immunohistochemical staining for microglial marker Iba1 (cyan) in the CA1 region of the hippocampus. Insets illustrate the typical microglial morphology of labeled cells. EGFP labels a small subset of pyramidal neurons (green), which were used for dendritic spine analysis. (B) The bar graphs display the efficacy of the oral PLX3397 administration. After 4 weeks of treatment, virtually all microglia are depleted from the hippocampus. **** p <0.0001, t-test. Scale bars are 100 µm for main images and 25 µm for inserts.