Research ArticleIMMUNOGENOMICS

Transcription factor ID2 prevents E proteins from enforcing a naïve T lymphocyte gene program during NK cell development

See allHide authors and affiliations

Science Immunology  27 Apr 2018:
Vol. 3, Issue 22, eaao2139
DOI: 10.1126/sciimmunol.aao2139
  • Fig. 1 ID2 was required for NK cell differentiation.

    (A) The frequency of NKP46+CD49b+ NK cells among Lin (Ter119Gr1) lymphoid cells was determined in the BM, spleen, and liver of Ctrl or RId2−/− mice by flow cytometry. Representative plots are shown with NK cell frequencies indicated. (B) Bar graphs showing the number of LinNKP46+CD49b+ NK cells/3 × 107 cells in the BM or spleen of Ctrl or RId2−/− mice ± SD. n > 7. (C) EOMES and TBET in BM LinNKP46+CD49b+ NK cells from Ctrl and RId2−/− mice. Representative of five experiments. (D) Semiquantitative PCR for the deleted and floxed alleles of Id2 in genomic DNA isolated from LinNKP46+CD49b+ BM NK cells of two RId2−/− mice. Two concentrations of DNA were tested as indicated. DNA from RId2f/f mice is also shown. (E) LinNKP46+CD49b+ NK cells from the BM, spleen, and liver of Ctrl and RId2−/− mice were examined for CD27 and CD11b by flow cytometry. (F) Bar graphs showing the average number of CD27+CD11b, CD27+CD11b+, and CD27CD11b+ NK cells/3 × 107 cells in the BM and spleen of Ctrl and RId2−/− mice. n = 7. (G and H) Chimeric mice created with RId2−/− (CD45.2+) and WT CD45.1+ BM and analyzed 8 weeks after reconstitution. The frequency of CD45.1 cells among the indicated NK cell subsets in the BM and spleen is shown. Representative of three experiments. **P < 0.01, ***P < 0.001.

  • Fig. 2 RId2−/− NK cells proliferate in response to IL-2 or IL-15.

    (A) Heat map of probe sets that were differentially expressed by greater than 1.5-fold (P < 0.05) in Ctrl and RId2−/− CD27+CD11b NK cells. Microarray analysis was performed on RNA isolated in three independent experiments. (B) Fold change in expression of a set of differentiation-associated probe sets, identified by Chiossone et al. (5), in RId2−/− as compared with Ctrl CD27+CD11b NK cells. (C) Ctrl or RId2−/− mice were injected with phosphate-buffered saline (PBS), IL-2, or polyI:C, and 3 days later, the frequency of KI67+ CD27+CD11b NK cells was determined in the BM (C) or spleen (D) by flow cytometry. Representative of three independent experiments. The average frequency of Ctrl (red) or RId2−/− (blue) CD27+CD11b NK cells expressing KI67 + SD. Three days after injection of PBS, IL-2 or polyI:C was determined for the (E) BM and (F) spleen. n = 3, ***P < 0.005. (G) qRT-PCR for Mcl1 and Socs3 mRNA shown relative to Hprt mRNA in Ctrl (red) or RId2−/− (blue) CD27+CD11b NK cells from the BM. (H) BM NK cells from Ctrl (red) or RId2−/− (blue) mice were treated with increasing concentrations of IL-15 for 3 days before addition of BrdU for 30 min. The cells were stained for NK cell surface markers, intracellular BrdU, and KI67. LinNKP46+CD49b+CD27+CD11b cells are shown. Representative of four independent experiments. (I) Average ratio ± SD of the percent of NK cells that were BrdU+ in cultures of RId2−/− as compared with Ctrl CD27+CD11b NK cells isolated from the BM (blue) or spleen (black), identified as in (H). n = 4.

  • Fig. 3 ID2 promotes the cytotoxic effector program of NK cells.

    (A) Fold change in expression of probe sets associated with NK cell differentiation or the cytotoxic effector program as measured by microarray analysis. (B) qRT-PCR analysis of Gzma, Gzmb, Gzmk, Prf1, and Tbet mRNA relative to Hprt mRNA in BM CD27+CD11b NK cells from Ctrl (red) or RId2−/− (blue) mice. Representative of three experiments, n = 3 per experiment. Error bars are SD. (C) TBET, KLRG1, and IL-18R1 on CD27+CD11b NK cells from Ctrl (shaded, gray) or RId2−/− (open, black) mice by flow cytometry. The open gray histogram is the FMO (fluorescence minus one) control. The mean fluorescence intensity (MFI) or percent positive cells is indicated for Ctrl (red) and RId2−/− (blue). Representative of three to eight experiments. (D) Klr8−/− mice were injected with NK cells from the spleen of Ctrl or RId2−/− (CD45.2+) mice along with an equal number of WT CD45.1+ splenic NK cells and infected with MCMV. The percent of Ly49H+ NK cells in the spleen that were CD45.1+ (black gate) or CD45.2+ (red for Ctrl cells and blue for RId2−/− cells) was examined by flow cytometry on day 7. One representative experiment is shown. n = 3. (E) Summary of data for multiple NK cell chimeric mice, set up as in (D). Each dot is one mouse. The mean ± SD is shown. **P < 0.01, ***P < 0.005.

  • Fig. 4 ID2 represses a naïve or memory-like T lymphocyte gene program.

    (A) Gene set enrichment analysis of the microarray data revealed enrichment of multiple T cell–associated pathways in RId2−/− CD27+CD11b NK cells as compared with Ctrl. (B) Fold change in expression of gene-specific probes in RId2−/− NK cells as compared with Ctrl as determined by microarray analysis. Cytokine- and chemokine-associated genes (blue), signaling molecule genes (green), and T cell–associated genes (red) are shown. (C) qRT-PCR for CD3d and Cd3g mRNA in Ctrl (red) and RId2−/− (blue) CD27+CD11b BM NK cells is shown relative to Hprt mRNA. Error bars are SD. One of three experiments is shown. (D). Surface expression of CXCR5, CXCR3, IL-4Rα, and CD27 and (E) intracellular staining for TCF1 on CD27+CD11b NK cells from the BM of Ctrl (gray, shaded) or RId2−/− (black, open) was determined by flow cytometry. Gray or stippled line is FMO control. For (D) and (E), the MFI or percent positive is indicated for Ctrl (red) and RId2−/− (blue). *P < 0.05, **P < 0.01.

  • Fig. 5 ID2 represses chromatin accessibility at T cell–associated genes by preventing E protein binding.

    (A) ATAC-seq was used to identify regions of open chromatin in RId2−/− (red) and Ctrl CD27+CD27 (blue) NK cells and in Ctrl CD27CD11b+ (green, labeled as CD11b+) NK cells. The graph shows normalized reads at the indicated distance from the center of open chromatin domains that had increased (left graph) or decreased (right graph) accessibility in RId2−/− as compared with Ctrl CD27+CD11b NK cells. (B) Heat map and K-means clustering of the regions analyzed in (A) with increased or decreased accessibility in RId2−/− as compared with Ctrl CD27+CD11b NK cells. (C) Fraction of genomic regions that showed increased accessibility (red), decreased accessibility (blue), or no change in accessibility (gray) in RId2−/− CD27+CD11b NK cells and were associated with genes that increased (left plot) or decreased (right plot) expression by greater than twofold. Black represents the background/null distribution (all genes in the genome). (D) HOMER motif enrichment at regions with increased accessibility near genes that increased expression in RId2−/− compared with Ctrl NK cells. (E) Normalized reads for open chromatin (left) and nucleosome differential (right) centered over E box motifs associated with genes that increased expression in RId2−/− NK cells. (F) Accessible chromatin at the Cxcr3, Cxcr5, and Tgfbr1 genes in RId2−/− (orange) and Ctrl (blue) CD27+CD11b NK cells and Ctrl CD11b+ NK cells (green) viewed in IGV. Black arrows indicate regions that are more accessible in RId2−/− compared with Ctrl. (G) Same as (F) but for the Socs3 gene.

  • Fig. 6 RId2−/− NK cells acquire an accessible chromatin state similar to naïve rather than MP or TE CD8 T cells.

    Accessible chromatin tracks for Ctrl, RId2−/−, and Ctrl CD11b+ NK cells from this study compared with CD8 T naïve, TE, or MP from GSE95237 after normalization of the data sets. Accessible chromatin domains at the (A) Il4ra, (B) Cd3d, and CD3g are shown, with black arrows indicating regions of increased accessibility in RId2−/− as compared with Ctrl. The red arrow indicates a peak unique to NK cells. (C) Peaks that increased in RId2−/− as compared with Ctrl by greater than 400% (P < 0.05; 486 peaks) were compared with the accessible peaks in naïve as compared with MP, TE as compared with naïve, or TE as compared with MP. A change in the CD8 T cell populations was accepted as positive if it was greater than 200% (P < 0.05). (D) As in (C) for peaks that increased by 400% (P < 0.05; 1274 peaks) in Ctrl as compared with RId2−/−.

  • Fig. 7 ID3 is required for the development of ID2-deficient NK cells.

    (A) Accessible chromatin tracks for the indicated populations surrounding the Id3 gene. Arrows indicate regions that are more accessible in RId2−/− as compared with Ctrl NK cells. The red arrows indicate peaks unique to NK cells. (B) CD45.2+ CD3εTCRβTCRδNK1.1+DX5+ NK cell numbers/104 BM cells in chimeric mice created by transplanting Id3+/+ (black) or Id3−/− (gray) BM into lethally irradiated CD45.1+ mice. Each circle represents one transplanted mouse with a unique donor. n = 4. (C) Flow cytometry of BM for CD45.2+ NK cells in CD45.1+ mice transplanted with CD45.2+ Ctrl, LId2−/−, and LId2−/−Id3−/− BM. Each row of plots is gated on the population indicated in the row to the left. Lineage = CD3ε, TCRβ, and TCRγ. (D) Average percent reconstitution, determined as the percent of LinCD122+NK1.1+DX5+ cells that were CD45.2+ ± SD, is shown for each population. n = 5. (E) and (F) are for the spleen as in (C) and (D). ***P < 0.001.

Supplementary Materials

  • immunology.sciencemag.org/cgi/content/full/3/22/eaao2139/DC1

    Fig. S1. ID2 is required for the development of NK cells, liver ILC1, and BM ILC2.

    Fig. S2. RId2−/− NK cells expand in vitro and phosphorylate S6 kinase similarly to Ctrl CD27+CD11b NK cells.

    Fig. S3. Activation of AP-1 restores IFN-γ production in RId2−/− NK cells.

    Fig. S4. Comparison of differential chromatin accessibility indicates that NK cell maturation has partial parallels to a CD8 MP to effector transition.

    Fig. S5. Cell-extrinsic deficiency of NK cells in Id3−/− mice caused by “γδNKT-like” cells.

    Table S1. Summary of gene set enrichment analysis for genes differentially expressed in Ctrl and RId2−/− CD27+CD11b NK cells.

    Table S2. Metascape pathway analysis for genes differentially expressed in Ctrl and RId2−/− CD27+CD11b NK cells.

    Table S3. Location of top five E box motifs in open chromatin with increased accessibility near genes with increased expression in RId2−/− NK cells.

    Table S4. Raw data for Figs. 1 to 4 and 7, and figs. S1 to S3 and S5.

  • Supplementary Materials

    Supplementary Material for:

    Transcription factor ID2 prevents E proteins from enforcing a na?ve T lymphocyte gene program during NK cell development

    Erin C. Zook, Zhong-Yin Li, Yiying Xu, Renée F. de Pooter, Mihalis Verykokakis, Aimee Beaulieu, Anna Lasorella, Mark Maienschein-Cline, Joseph C. Sun, Mikael Sigvardsson, Barbara L. Kee*

    *Corresponding author. Email: bkee{at}bsd.uchicago.edu (A.D.)

    Published 27 April 2018, Sci. Immunol. 3, eaao2139 (2018)
    DOI: 10.1126/sciimmunol.aao2139

    This PDF file includes:

    • Fig. S1. ID2 is required for the development of NK cells, liver ILC1, and BM ILC2.
    • Fig. S2. RId2-/- NK cells expand in vitro and phosphorylate S6 kinase similarly to Ctrl CD27+CD11b- NK cells.
    • Fig. S3. Activation of AP-1 restores IFN-γ production in RId2-/- NK cells.
    • Fig. S4. Comparison of differential chromatin accessibility indicates that NK cell maturation has partial parallels to a CD8 MP to effector transition.
    • Fig. S5. Cell-extrinsic deficiency of NK cells in Id3-/- mice caused by "γδNKTlike" cells.
    • Legends for tables S1 to S4

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Table S1 (Microsoft Excel format). Summary of gene set enrichment analysis for genes differentially expressed in Ctrl and RId2-/- CD27+CD11b- NK cells.
    • Table S2 (Microsoft Excel format). Metascape pathway analysis for genes differentially expressed in Ctrl and RId2-/- CD27+CD11b- NK cells.
    • Table S3 (Microsoft Excel format). Location of top five E box motifs in open chromatin with increased accessibility near genes with increased expression in RId2-/- NK cells.
    • Table S4 (Microsoft Excel format). Raw data for Figs. 1 to 4 and 7, and figs. S1 to S3 and S5.

    Files in this Data Supplement:

Navigate This Article