Fig 2: Frequency of TReg populations is altered in individuals with chronic SCI. The frequency of CD3+ (a) is again lower in this group of individuals with chronic SCI as compared to a control group, while a trend toward a lower CD4+ cell population (b) is observed in individuals with chronic SCI. The frequency of CD25+ CD127lo expressing cells is equivalent between uninjured and individuals with chronic SCI (c). Within the CD25+ CD127lo population, the frequencies of CCR4+ or HLA-DR+ cells are significantly elevated in individuals with chronic SCI (d, e). Within the CD25+ CD127lo population, the frequency of cells that were positive for CCR4 and HLA-DR was also elevated (f). Significant P values are shown

Fig 3: Gating strategy of regulatory CD4+ T Regs. Flow cytometry dot plots illustrate the gating of TRegs. Cells expressed CD3 (a) and CD4 (b). CD4+ T cells were then gated for the expression of CD25 and low expression of CD127 (c). CD25+ CD127lo T cells were then gated for the expression of CCR4 (d) and HLA-DR (e)

Fig 4: CCRF-CEM and Treg migrate towards Raji in a CCR4-dependent manner.(A) Chemotaxis assays were performed on CCRF-CEM cells towards recombinant CCL22 or Raji cell supernatant with or without the CCR4 antagonist CCR4-351. Relative luciferase units (RLU) are shown as a proxy for migrated cell number. (B) Migration assays were performed with human iTreg. (C) Raji and Daudi xenograft tumors grown in NOD/SCID mice (n = 5) were assayed for chemokine levels by ELISA. (D) Human iTreg were transferred into Daudi or Raji tumor-bearing mice (n = 8) and allowed to migrate for 5 days, followed by harvesting of the tumors. Mice were treated with vehicle or CCR4-351. Fractions of human CD45+ cells which were iTreg (identified as CD45+ CD4+ CD19-) were quantitated by FACS.

Fig 5: Gut dysbiosis and immune cell changes during disease progression in 5XFAD transgenic (Tg) mice. a Changes in the relative RNA expression levels of synaptophysin in the hippocampus of 5XFAD transgenic (Tg) mice at 2, 3, 5, 7 and 9 months and in wild-type (WT) mice at 2 months (n = 5–12). The data are presented as mean ± standard error of the mean (mean ± SEM) relative to the expression level of actin. *P < 0.05, **P < 0.01 by one-way ANOVA (F (5, 43) = 2.952). b Changes in the time out of 104 s taken to achieve 80% success (see “Materials and methods”) in a test to evaluate the discrimination learning abilities of 5XFAD transgenic (Tg) mice at 2, 3, 5, 7, and 9 months and wild-type (WT) mice at 2 months (n = 4–8). Data are presented as mean ± standard error of the mean (mean ± SEM). *P < 0.05 by Student’s t-test. s seconds. c Enterotype analysis at the genus level of the gut microbiomes of 5XFAD transgenic (Tg) mice and wild-type (WT) mice at 2-, 3-, 5-, 7- and 9-month old (n = 4–10). The separation of gut microbial taxa into a norank genus under the family of Muribaculaceae of Tg mice and the genus Lactobacillus enterotypes of WT mice is achieved by calculating Bray-Curtis distance based on the relative abundance at the genus level and clustered using PAM (Partitioning Around Medoids). The data are most naturally separated into two clusters, as determined by the Calinski-Harabasz (CH) index and represented using principal coordinate analysis (PCoA). The shapes and colours of the points indicate samples from each individual from various months. The coloured ellipses indicate the 0.95 confidence interval (CI) ranges within each enterotype group. M months. d Principal component analysis (PCA) of the gut microbiome composition of WT and 5XFAD transgenic (Tg) mice on the operational taxonomic unit (OTU) level at different time points (n = 4–10). The shapes and colours of the points indicate samples from each individual from various months. The coloured ellipses indicate 0.95 confidence interval (CI) ranges within each tested group. M months. e Abundance changes of operational taxonomic units (OTUs) in the overall population in the gut microbiome of 5XFAD transgenic (Tg) mice at various months, coloured at the phylum level on a stream graph (n = 4–10). The two most abundant phyla, Bacteroidetes and Firmicutes, are labelled on the graph. Colours indicate different phyla of the gut microbiota. f Changes in the positive densities of IBA1 immune-fluorescent staining, reflecting activation of microglial cells in the hippocampus of 5XFAD transgenic (Tg) mice at 2, 3, 5, 7, and 9 months relative to the values of 2-month-old wild-type (WT) mice (n = 2–7). The data are presented as mean ± standard error of the mean (mean ± SEM); lines are fitted with a cubic spline. g Changes in activated M1 and M2 type microglia detected in the whole-brain homogenates of 5XFAD transgenic (Tg) mice at 2, 3, 5, 7 and 9 months (n = 4–8). M1-type microglia (CD45lowCD11b+CX3CR1+Siglec-H+CD86+) and M2-type (CD45lowCD11b+CX3CR1+Siglec-H+ CD206+) microglia were detected by flow cytometry, and their cell counts are presented relative to the frequency of CD45low CD11b+ cells. Red points and lines: M1 microglia. Green points and lines: M2 microglia. The data are presented as mean ± standard error of the mean (mean ± SEM); lines are fitted with a cubic spline algorithm. h Changes in infiltrating cells (CD45high) detected in the whole-brain homogenates of 5XFAD transgenic (Tg) mice (red points and lines) and WT mice (black points and lines) at different time points as detected by flow cytometry (n = 4–8). Cell counts are presented relative to the frequency of CD45+ cells and formatted as mean ± standard error of the mean (mean ± SEM). Lines are fitted with a cubic spline algorithm. i Changes in CD45high cells in 5XFAD transgenic (Tg) mice at different time points (n = 4–8). On the barplot, cell counts are presented relative to the frequency of CD45high cells. Colours indicate different subtypes of CD45high cells: Neu, neutrophils; DC, dendritic cells; NK, natural killer cells; Mo/M? ? monocytes and macrophages; B, B cells; Others, unclassified cells. j Changes in infiltrating CD4 T cells (CD45highCD4+) detected in the whole-brain homogenates of 5XFAD transgenic (Tg) mice (red points and lines) at 2, 3, 5, 7 and 9 months as detected by flow cytometry (n = 4–8). Cell counts are presented relative to the frequency of CD45high cells and formatted as mean ± standard error of the mean (mean ± SEM). Lines are fitted with a cubic spline. k Changes in infiltrating peripheral Th1 and Th2 cells detected in the whole-brain homogenates of 5XFAD transgenic (Tg) mice at 2, 3, 5, 7 and 9 months (n = 4–8). Th1 cells (CD45highCD4+CXCR3+) and Th2 cells (CD45highCD4+CCR4+) were detected by flow cytometry, and presented relative to the frequency of CD45highCD4+ T cells. Red points and lines: Th1 cells. Green points and lines: Th2 cells. The data are presented as mean ± standard error of the mean (mean ± SEM). Lines are fitted with a cubic spline. l Correlation between the trend changes of the frequency of brain lymphocytes and the abundance of gut microbiota represented at genus level during the Th2/M2-related stage and Th1/M1-related stage in the early and late phase (2–3 months and 7–9 months), respectively (left panels, n = 4–8). Bacteria that are significantly correlated with Th2/M2 or Th1/M1 in 5XFAD mice are listed in the right-hand panels. Squares in red (positive correlation) or blue (negative correlation) with a yellow asterisk (*) indicate significant correlations with P-values < 0.05 measured by the Pearson parametric correlation test (see “Materials and methods”). f family, M months, T regulatory cells Treg, Mo monocytes, B B cells, DC dendritic cells, NK natural killer cells