Fig 2: Foxo3a interacts with Tet2 and modulates the expression of key genes involved in neurogenesis.(a) Merged Foxo3a ChIP-seq and 5hmC-seq peaks from WT and Tet2−/− aNSCs were used to calculate their overlap. 8,179 out of a total 21 395 Foxo3a peaks located 500 bp adjacent to aNSC 5hmC peaks (binomial test, P<0.0001). (b) Endogenous Foxo3a and Tet2 co-immunoprecipitation (co-IP) experiments were performed using WT aNSCs, followed by immunoblotting with either Tet2 or Foxo3a antibody. Endogenous Foxo3a interacted with Tet2 in WT aNSCs. (c) Mouse Myc-Tet2 and FLAG-Foxo3a expression plasmids were co-transfected into HEK293 cells, and co-IP experiments were performed by anti-FLAG and anti-Myc antibodies followed by immunoblotting. Exogenous mouse Foxo3a and Tet2 interacted in reciprocal IP. (d) Average normalized Foxo3a mapped reads per million from WT and Tet2−/− were calculated on promoters (1 kb±TSS) of 439 upregulated genes bearing higher intragenic 5hmC in Tet2−/− (Fig. 4d). Foxo3a became further enriched on 310 out of 439 (binomial test, P<0.0001) genes in the Tet2−/− aNSCs. (e) GO analyses on 310 upregulated genes bearing higher intragenic 5hmC and promoter Foxo3a in Tet2−/− were genuine in GO terms related to the cell cycle and nervous system development. Fold enrichment is demonstrated by heatmap view. Selected terms with fold enrichment >2 are indicated.

Fig 3: Foxo3a directly regulated genes involved in aNSC proliferation and signature by binding to their promoters.(a,b) Average normalized Foxo3a and 5hmC mapped reads from both WT and Tet2−/− aNSCs were calculated on promoters and gene bodies of the cell cycle (a, n=38) and nervous system development (b, n=54) related genes. Log2 of average reads per million on each gene were displayed by heatmap. Violin plots summarized the mean value of each column, and statistical significance was assessed by unpaired t-test. Foxo3a was substantially enriched on promoters and became significantly increased in the absence of Tet2. Genes involved in nervous system development showed general higher Foxo3a and 5hmC enrichment than genes involved in the cell cycle. P values were indicated. Genomic regions between 500 bp upstream and downstream of transcription start sites (TSS) were referred as promoter regions. Genomic regions between 500 bp downstream of TSS and transcription end sites (TES) were referred as gene bodies. (c–e) Real-time PCR to determine the expression dynamics of selected loci related to the cell cycle and nervous system development. The expression of these genes was upregulated in the absence of Tet2 in aNSCs and restored when Foxo3a was knocked down by siRNA in Tet2−/− aNSCs. n=3; unpaired t-test. *P<0.05; **P<0.01; ***P<0.001. (f) Quantification of BrdU+ aNSCs from WT, Tet2−/− and Foxo3a knockdown in Tet2−/− background. The relative percentage of BrdU+ aNSCs were normalized to total cells by DAPI staining. Tet2−/− aNSCs displayed significant higher BrdU+ aNSCs, indicating their higher proliferation rates. Depletion of Foxo3a in Tet2−/− aNSCs significantly reduced their proliferation. (n=3; unpaired t-test, *P<0.05; ***P<0.001).