MicroRNA-142 Is Critical for the Homeostasis and Function of Type 1 Innate Lymphoid Cells

Abstract

Natural killer (NK) cells are cytotoxic type 1 innate lymphoid cells (ILCs) that defend against viruses and mediate anti-tumor responses, yet mechanisms controlling their development and function remain incompletely understood. We hypothesized that the abundantly expressed microRNA-142 (miR-142) is a critical regulator of type 1 ILC biology. Interleukin-15 (IL-15) signaling induced miR-142 expression, whereas global and ILC-specific miR-142-deficient mice exhibited a cell-intrinsic loss of NK cells. Death of NK cells resulted from diminished IL-15 receptor signaling within miR-142-deficient mice, likely via reduced suppressor of cytokine signaling-1 (Socs1) regulation by miR-142-5p. ILCs persisting in Mir142^-/- mice demonstrated increased expression of the miR-142-3p target αV integrin, which supported their survival. Global miR-142-deficient mice exhibited an expansion of ILC1-like cells concurrent with increased transforming growth factor-β (TGF-β) signaling. Further, miR-142-deficient mice had reduced NK-cell-dependent function and increased susceptibility to murine cytomegalovirus (MCMV) infection. Thus, miR-142 critically integrates environmental cues for proper type 1 ILC homeostasis and defense against viral infection.

Keywords: IL-15; ILC1-like cells; NK cells; cytokine receptors; innate lymphoid cells; integrin; microRNA-142; murine cytomegalovirus; tissue resident.

Figure 1.. miR-142 deficiency reduces peripheral type-1 ILC numbers and alters trafficking.

BM, SP, and blood (BL) of Mir142^+/+ (control, open green) and Mir142^−/− (filled blue) were harvested and assessed for type-1 ILCs (CD45⁺CD3⁻NK1.1⁺NKp46⁺). (A) Representative flow plot of NK1.1⁺ NKp46⁺ cells (gated on CD3⁻) from the SP and BM. (B) Summary data showing the total NK1.1⁺ NKp46⁺ cells, see Fig. S1 (C-D) Schema of the BM chimera assay. (D) Summary of 2 independent experiments with >8 mice per group. (E) Non-competitive BM chimera experiment where control or Mir142^−/− BM were transferred into congenic, irradiated recipients. Summary data showing percent CD45.2⁺CD3⁻NK1.1⁺NKp46⁺ cells from 3 independent experiments with 6 mice per group. (F) Summary data showing the NK cell precursor population numbers: pre-NK precursor (pre-NKp) and restricted NK precursor (rNKp). Data are from >3 independent experiments with >12 mice per group. Data were compared using Student’s T test or Mann-Whitney test. (G) Experimental schema for trafficking experiments. Briefly, control and miR142^−/− BM cells were differently labeled with cell trace violet and cell trace yellow. The cells were mixed, injected iv. into WT recipients, and tissues examined for the presence of labeled NK1.1⁺NKp46⁺ cells. Flow plots of pre-infusion NK1.1⁺ cells present at 1:1 ratio. (H) Representative flow cytometry showing NK1.1⁺ cells. (I) Summary data from (H) showing the ratio of control and miR142^−/− NK1.1⁺ cells. N=10 mice from 2 independent experiments. Significance was determined by Wilcoxon signed rank test against the theoretical value of one (no trafficking alteration, dashed line). Please also see Figure S3.

Figure 2.. miR-142-deficient mice display altered, tissue-specific type-1 ILC compartment.

The CD3⁻NK1.1⁺NKp46⁺ compartment of Mir142^+/+ (open green) and Mir142^−/− (filled blue) mice was assessed for CD49a and CD49b by flow cytometry. (A) Representative flow plots of CD3⁻NK 1.1 ⁺NKp46⁺ cells. (B) Total number of NK cells (NK, CD49a⁻CD49b⁺), ILC1-like (CD49a⁺CD49b⁺), and ILC1s (CD49a⁺CD49b⁻) within the indicated tissues. Summary from >3 independent experiments with > 10 mice per group. (C) Total numbers of NK, ILC1-like, and ILC1 cells within control (Ncr1-cre⁺ Mir142^+/+) and ILC-specific miR-142-deficient (Ncr1-cre⁺ Mir142^f/f) mice, cells gated on CD3⁻NK1.1⁺NKp46⁺ YFP⁺. (D) Experimental Schema. BM from control and Ncr1-cre⁺ Mir142^f/f mice were transferred into irradiated recipients and spleens assessed 6 weeks later. (E) Summary data showing the percent of ILC1-like cells from CD3-NK1.1⁺NKp46⁺ CD45.1⁺ (WT) or YFP⁺ (Mir142^f/f) within the indicated recipient mice. Data from 2-3 independent experiments, N=7-10 mice, compared using T-test or Mann-Whitney corrected for multiple comparisons, when appropriate. See also Figure S2.

Figure 3.. miR-142-deficient type-1 ILCs display ILC1-like phenotype in lymphoid tissues.

(A) Representative histograms depict CD43 expression on NK1.1⁺NKp46⁺ cells from Mir142^−/− (blue) mice compared to controls (green). Gray filled histograms represent CD43⁻ lymphocytes. (B) Summary from (A). (C) Summary CD73 and TIGIT median expression on NK1.1⁺NKp46⁺ cells. (D) Transcription factor median expression on Mir142^−/− and control NK1.1⁺NKp46⁺ cells. Data are from 2-4 independent experiments with N=5-10 mice per group. Data were compared using Student’s T or Mann Whitney tests.

Figure 4.. TGF-β signaling is increased in Mir142^−/− type-1 ILCs.

(A-B) CD3⁻NK1.1⁺NKp46⁺ cells were flow-sorted from Mir142^+/+ (WT) or Mir142^−/− (KO) BM or SP and microarray performed. (A) GSEA of the TGF-β Hallmark Signaling Pathway gene set in Mir142^−/− versus WT from SP and BM. NES, Normalized enrichment score. (B) Heat map of differentially regulated genes between Mir142^+/+, Smad4^Δ/Δ, and Mir142^−/− NK1.1⁺ cells. Genes associated with the ILC1-like signature (black), miR-142 targets (blue), and NK cell development and effector functions (green) are indicated. Data are from 3 biological replicates. (C) Predicted miR-142-3p binding site in the 3’UTR of Tgfbr1. (D) Luciferase reporter assay for miR-142 overexpression (OE) without the Tgfbr1 3’UTR (empty), with Tgfbr1 3’ UTR (WT) or with Tgfbr1-mutated at the miR-142-3p binding site in the 3’ UTR (Δ). Data summarize 3 independent experiments and were compared using an ANOVA. (E) Summary TGFBR1 expression on type-1 ILCs from SP of indicated mice (relative expression: median minus FMO); data from > 3 independent experiments were compared with T test (n=6-8 mice per group). (F-G) SP cells were stimulated with 10 ng/mL IL-15 plus 10 ng/mL TGF-β1 or left unstimulated (0 minutes) and assessed for phosphorylated (p) SMAD-2 and/or −3. (F) Representative flow plot showing pSMAD-2 and/or −3 in unstimulated (0m, top) or stimulated (60m, bottom) CD3⁻NK1.1⁺ cells. (G) Summary from (F). Data are from 2 independent experiments with 6-7 mice per group. Significance was determined by 2-Way ANOVA. Please also see Figure S4.

Figure 5.. miR-142-deficient type-1 ILCs are hyporesponsive to IL-15 signaling.

(A) WT CD3⁻NK1.1⁺ cells were sorted from the SP and incubated with 10 ng/mL IL-15. Immediately after sorting (baseline) and after stimulation, RNA was extracted from the cells and miR-142-3p/5p expression were assessed by qPCR. Summary data (normalized to baseline) from 2 independent experiments with 3 biological replicates. (B) Splenocytes were incubated with IL-15 for 48 hours in vitro. Summary data depict the percent of type-1 ILCs (CD45⁺CD3⁻NK1.1⁺) positive for 7-AAD and/or Annexin V. (C) Representative histograms of surface expression of CD122 (IL-2 and −15Rβ) on Mir142^+/+ and Mir142^−/− CD45⁺CD3⁻NK1.1⁺ cells (Global, Top) and Ncr1-cre⁺ Mir142^+/+ and Ncr1-cre⁺ Mir142^f/f CD45⁺CD3⁻NK1.1⁺YFP⁺ cells (Ncr1-cre⁺, Bottom). Gray filled histograms depict CD122-negative lymphocytes. (D) Summary data from (C) showing the MFI. (E) Representative histograms showing intracellular pSTAT-5 (Y694) staining in global (left) and Ncrl-cre⁺ (right) NK1.1⁺ and/or YFP⁺ cells after 15-minute stimulation with IL-15. (F) Summary data from (E). (G) Representative histograms showing pSTAT-4 in global (Top) and Ncr1-cre⁺ (Bottom) CD3⁻ NK1.1⁺ and/or YFP⁺ cells after stimulation with IL-12 for 15 minutes. (H) Summary data from (G). (I) Predicted miR-142-5p binding site in Socs1 3’ UTR (top). A luciferase reporter assay for Socs1 3’UTR. Data are compared using an ANOVA. Data summarize 3 independent experiments. (J-K) CD3⁻NK1.1⁺ SP cells from control and Mir142^−/− SP were sorted and assessed for Socs1 and β-actin by western blot. (J) Representative blot from 3 independent experiments showing Socs1 (top) and β-Actin (bottom). (K) Summary data depicting Socs1 band intensity (normalized using β-actin). Comparisons were made using RM-ANOVA (A, I), Student’s T test/Mann-Whitney (B-D, H), RM-ANOVA (F), paired T test (K). Please also see Figure S5.

Figure 6.. Integrin aV promotes type-1 ILC survival in miR-142-deficient mice.

(A) Summary data of integrin αVβ3 and CD49a on CD3⁻NK1.1⁺NKp46⁺CD49b⁺ cells from the SP of Il15^−/− or control mice. (B) Representative histogram showing integrin αV expression in the indicated tissues of Global and Ncr1-cre⁺ mice. (C) Summary showing percent integrin αV⁺ NK1.1⁺NKp46⁺ cells in the indicated tissues from Mir142^−/− (left), Ncr1-cre⁺ Mir142^f/f (YFP⁺, right), and controls. (D) Predicted miR-142-3p binding site in 3’UTR of Itgav (top). Luciferase reporter assay for Itgav 3’UTR (bottom). Data summarize 3 independent experiments and were compared using an ANOVA. (E) Mir142^−/− mice were treated for 2 weeks with an αV inhibitor or vehicle. Summary data of total type-1 ILCs in the SP from 4 independent experiments, n=11 mice per group. (C, E) Data were compared using Student’s T test. See also Figure S3.

Figure 7.. miR-142-deficient type-1 ILCs display altered effector responses in vitro and in vivo.

(A-B) Splenocytes from Mir142^+/+ and Mir142^−/− mice were stimulated for 6 hours. (A) Representative flow plots showing intracellular IFN-γ staining in CD45⁺CD3⁻NK1.1⁺ cells. (B) Summary data from (A). (C-D) Splenocytes were isolated from Ncr1-cre⁺ Mir142^+/+ (control, open green bars) and Ncr1-cre⁺ Mir142^f/f (filled blue bars) mice and stimulated for 6 hours. (C) Representative flow plots showing intracellular IFN-γ staining in CD45⁺CD3⁻NK1.1⁺YFP⁺ cells. (D) Summary data from (C). (E-F) 36 hours after MCMV infection, SP NK1.1⁺ cells were assessed immediately for intracellular IFN-γ protein by flow cytometry. (E) Representative flow plots showing IFN-γ in CD45⁺CD3⁻NK1.1⁺ cells. (F) Summary from (E). (G) Viral copy numbers in the SP of MCMV-naïve control, Mir142^+/+ (WT) and Mir142^−/− (KO) MCMV-infected mice, were assessed using qPCR 4 days after infection (units: IE1×1000/B-Actin). (H) Mice were infected with 1e⁵ PFU MCMV and survival assessed. Pooled data from 2-3 independent experiments with >4 mice per group per experiment. Comparisons were made using Student’s T test or Mann-Whitney test, and log-rank (H). See also Figure S7.