Fig. 5

Monocyte-derived macrophages uniquely affect disease modification in PD-L1 blockade in DM-hTAU mice. a, b Flow cytometry of splenocytes, CD44⁺CD62L^low effector memory T (T_EM) cells, versus CD44⁺CD62L^high central memory T (T_CM) cells in DM-hTAU mice, treated with 0.5 mg of anti-PD-L1 (n = 10) or IgG (n = 11) (one-way ANOVA, Fisher’s exact test). c, d Flow cytometry of brains from anti-PD-L1-treated mice (n = 10), and IgG-treated mice (n = 16) analyzed for CD45^highCD11b^high, pooled from two experiments. e–g Repeated experiment as in a, b using GFP-BM-chimeric DM-hTAU mice. e Flow cytometry of GFP-labeled cells gated from CD45^highCD11b^high cells, expressing Ly6C. f Quantitation of the number of GFP⁺CD45^highCD11b^high cells in anti-PD-L1 (n = 4), relative to IgG-treated mice (n = 6). g Representative projections of confocal z-axis stacks, showing colocalization of GFP⁺ cells (green) with IBA-1 (blue), in the cortex of DM-hTAU^GFP/+ mice, treated with anti-PD-L1 antibody (arrowheads). Scale bar: 100 μm. h Representative confocal z-axis stacks, showing colocalization of GFP⁺ cells (green), IBA-1 (blue), and IL-10 (red) in the brains of anti-PD-L1-treated DM-hTAU^GFP/+ mice. Scale bar: 50 μm. i Sorted CD45^highCD11b^high from DM-hTAU mice treated with anti-PD-L1, analyzed by single-cell RNASeq (Supplementary Fig. 8). tSNE plot depicting 899 cells. Clusters indicated by color and number. j Average Unique Molecular Identifier counts for selected genes across the 12 clusters. k, l Representative projections of confocal z-axis stacks, showing colocalization of GFP⁺ cells (green) with MSR1 (red) and IBA-1 (blue) in the cortex (k), and of GFP⁺ cells (green) with MSR1 (red) in the hippocampus of DM-hTAU^GFP/+ mice treated with anti-PD-L1 antibody (l). Scale bars: 25 and 50 μm. m, n BM-chimeric DM-hTAU and WT mice (male and female) prepared using WT or MSR1^−/− mice as BM donors. m T-maze task, 2 weeks after BM transplant, of WT > WT (n = 4), MSR1^−/− > WT (n = 5), WT > DM-hTAU (n = 8) and MSR1^−/− > DM-hTAU (n = 8) chimeric mice. n The same mice were treated after the behavioral assessment in m with 1.5 mg of anti-PD-L1 antibody or IgG control antibody, and were tested again 1 month later for their performance in T-maze; nonchimeric IgG-treated DM-hTAU littermates were used as additional controls. Improved performance of WT > DM-hTAU treated with anti-PD-L1 (n = 5) versus IgG-treated WT > DM-hTAU (n = 3) and IgG-treated nonchimeric DM-hTAU mice (n = 6). MSR1^−/− > DM-hTAU mice failed to show beneficial effect following treatment with anti-PD-L1 (n = 5), performing similarly to MSR1^−/− > DM-hTAU treated with IgG (n = 3). In all panels, error bars represent mean ± s.e.m.; *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA and Fisherʼs exact test)