Ly49E separates liver ILC1s into embryo-derived and postnatal subsets with different functions

Abstract

Type 1 innate lymphoid cells (ILC1s) represent the predominant population of liver ILCs and function as important effectors and regulators of immune responses, but the cellular heterogeneity of ILC1s is not fully understood. Here, single-cell RNA sequencing and flow cytometric analysis demonstrated that liver ILC1s could be dissected into Ly49E+ and Ly49E- populations with unique transcriptional and phenotypic features. Genetic fate-mapping analysis revealed that liver Ly49E+ ILC1s with strong cytotoxicity originated from embryonic non-bone marrow hematopoietic progenitor cells (HPCs), persisted locally during postnatal life, and mediated protective immunity against cytomegalovirus infection in newborn mice. However, Ly49E- ILC1s developed from BM and extramedullary HPCs after birth, gradually replaced Ly49E+ ILC1s in the livers with age, and contained the memory subset in recall response to hapten challenge. Thus, our study shows that Ly49E dissects liver ILC1s into two unique subpopulations, with distinct origins and a bias toward neonatal innate or adult immune memory responses.

Graphical abstract

Figure 1.

Dissection and clustering of murine liver group 1 ILCs at steady state. (A) Overview of the study design. Group 1 ILCs (CD45⁺NK1.1⁺NKp46⁺CD3⁻CD19⁻) from the livers and spleens of naive WT C57BL/6 (B6) mice (6–8 wk old) were sorted for scRNA-seq. (B and C) t-SNE clustering plots of each hepatic and splenic group 1 ILC sample (B) and an overlay (C). Each point represents an individual cell. Hepatic and splenic group 1 ILCs could be grouped into five distinct clusters (C1–C5). Data were obtained from two independent experiments (liver) or one experiment (spleen). (D) Heatmap showing the expression levels of the top 200 genes (left) and top 10 genes (right) expressed in each liver group 1 ILC cluster (C1–C5). (E) Distribution of signature gene expression by the liver (top) and spleen (bottom) group 1 ILC clusters in t-SNE plots. (F–I) Distribution of Cd27, Itgam, Klrg1 (F), and Klra5 (H) expression in the t-SNE plot. Representative plots of CD27, CD11b, and KLRG1 expression on cNK cells (CD45⁺NK1.1⁺CD49b⁺CD49a⁻CD3⁻CD19⁻; G) and Ly49E/F expression on liver ILC1s (CD45⁺NK1.1⁺CD49a⁺CD49b⁻CD3⁻CD19⁻; I) of WT B6 mice (6–8 wk old). Fluorescence minus one (FMO) controls were gated on total cNK cells. Flow cytometric plots are representative of at least three independent experiments with n = 3–9 mice per experiment.

Figure 2.

Ly49E expression distinguishes two distinct clusters of liver ILC1s. (A) Representative plots of Ly49E expression on liver ILC1s (CD45⁺NK1.1⁺CD49a⁺CD49b⁻CD3⁻CD19⁻) and cNK cells (CD45⁺NK1.1⁺CD49b⁺CD49a⁻CD3⁻CD19⁻) of WT B6 mice (5–8 wk old). (B) Representative plots showing the expression of ILC1-lineage markers on Ly49E⁺CD45⁺ leukocytes from the livers of WT B6 mice (5–8 wk old). Data are representative of at least five independent experiments with n = 3–5 mice per experiment (A and B). (C) Representative plots of CD49a and Ly49E/F expression on CD45⁺NK1.1⁺NKp46⁺CD3⁻Ly49F⁻ cells from the indicated tissues of WT B6 mice (5–8 wk old). Data are representative of at least three independent experiments with n = 1–4 mice per experiment. (D and E) Percentages of cells with higher expression levels of the indicated molecular signatures in liver Ly49E⁻ ILC1s (D) or Ly49E⁺ ILC1s (E) of WT B6 mice (5–8 wk old). Data are pooled from one or two independent experiments with n = 3–4 mice. (F and G) Comparative transcriptome analysis between Ly49E⁺ or Ly49E⁻ ILC1s from WT B6 mice (5–8 wk old) using bulk cell RNA-seq. Volcano plot (F) and GO Biological Processes (G) associated with the genes with >2.0-fold changes in expression are highlighted in red (higher in Ly49E⁺ ILC1s) or blue (higher in Ly49E⁻ ILC1s). (H–K) 1–2 × 10⁵ Ly49E⁺ or Ly49E⁻ ILC1s were sorted from CD45.2⁺ WT mice (3–4 wk old) and adoptively transferred into sublethally irradiated CD45.1⁺ mice (6–8 wk old). 2 wk after transfer, CD45.2⁺ liver ILC1s in recipients were analyzed for Ly49E expression. (H) Schematic of the experimental design. (I) Representative plots showing Ly49E expression on donor-derived (CD45.2⁺) ILC1s in recipient liver. (J) Percentages of Ly49E⁺ or Ly49E⁻ cells among host- or donor-derived liver ILC1s. Data are pooled from two (Ly49E⁺ ILC1s) or three (Ly49E⁻ ILC1s) independent experiments with n = 7–9 mice (I and J). (K) Percentages of cells expressing the indicated molecular signatures among liver ILC1 subsets after 2 wk of adoptive transfer. Data are representative of at least two independent experiments with three to four recipient mice per group. Bar graphs show the mean ± SEM; an unpaired Student’s t test was used to compare experimental groups in D, E, J, and K; two-way ANOVA was used in I; *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

Figure S1.

Differentially expressed genes between two liver ILC1 subsets revealed by scRNA-seq and bulk RNA-seq. (A) Distribution of selected genes (upregulated in C4 or C5 cluster) within C1–C5 clusters from scRNA-seq analysis. (B) Heatmap showing the expression of selected ILC-featured transcription factors from bulk RNA-seq data of liver Ly49E⁺ ILC1s and Ly49E⁻ ILC1s from WT B6 mice (5–8 wk old). (C) Representative plots of GATA3 and RORγt expression in ILC subsets of WT B6 mice (5–8 wk old). Flow cytometric plots are representative of two independent experiments with four to five mice per group. (D) Gene expression signature of the indicated cell populations revealed by bulk RNA-seq. Left heatmap shows total differential expressed genes with >2.0 fold-changes. Right heatmap shows genes that were manually selected from the left with different categories.

Figure S2.

Adult Ly49E⁻ ILC1s have the capacity to confer memory responses in the CHS model. (A) Representative dot plots showing Ly49E/F and IL-7Rα expression on liver ILC1 subsets from WT B6 mice (6–8 wk old). Data are representative of at least two independent experiments with three to four mice per experiment. (B) Percentages and GMFI of IL-7Rα expression among liver Ly49E⁺ or Ly49E⁻ ILC1s from B6 mice (6–8 wk old). Data are pooled from two independent experiments with seven mice per experiment. (C and D) WT B6 mice (6–8 wk old) were abdominally sensitized with 5% OXA and challenged with 1% OXA on the ears 5 d later (sensitized and challenged). The control mice were only challenged with 1% OXA on the ears without prior abdominal sensitization (challenged only). 48 h after 1% OXA challenge, the ear cell suspensions of two groups were prepared for flow cytometric analysis. Statistical analysis showing the percentages (C) and absolute numbers (D) of Ly49E⁺ or Ly49E⁻IL-7Rα⁺ ILC1s in mouse ears. Ear ILC1s were gated as CD3⁻CD19⁻NK1.1⁺NKp46⁺CD49a⁺. Data were pooled from two independent experiments with n = 6 mice per group. (E and F) Sublethally irradiated WT B6 mice (6–8 wk old) received 5 × 10⁴ Ly49E⁺ ILC1s, Ly49E⁻ ILC1s, or liver cNK cells obtained from 5% OXA-sensitized WT B6 mice (4 wk old) and were challenged 1 mo later with 1% OXA on the right ears and solvent control on the left ears. Schematic showing the experimental design (E). Ear swelling of recipient mice was measured 24–96 h after challenge (F). Data were pooled from three independent experiments with n = 5–7 mice per group. (G) Sublethally irradiated WT B6 mice (6–8 wk old) received OXA-sensitized Ly49E⁻ ILC1s and were challenged with 1% OXA on the right ears or 0.5% FITC on the left ears 1 mo later. Ear swelling of recipient mice was measured 48 h after challenge. Data were pooled from two independent experiments with n = 7 mice per group. (H and I) Sublethally irradiated WT B6 mice (6–8 wk old) received 5 × 10⁴ Ly49E⁺ ILC1s, Ly49E⁻IL-7Rα⁻ ILC1s, Ly49E⁻IL-7Rα⁺ ILC1s, or liver cNK cells from OXA-sensitized WT B6 mice (4 wk old) and were challenged 1 mo later with 1% OXA on the right ears or solvent control on the left ears. Schematic showing the experimental design (H). Ear swelling was measured 24–96 h after challenge of recipient mice (I). Data were pooled from four independent experiments with n = 6–14 mice per group. Bar graphs show the mean ± SEM; unpaired, two-tailed Student’s t test were used in B and G, two-way ANOVA in C, D, F, and I; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S3.

The neonatal liver is enriched with Ly49E⁺ ILC1s that are rarely replenished after birth. (A) Representative plots showing the gating strategy and characteristics of neonatal liver group 1 ILCs (0–3 d old). (B) Percentages of Ly49E⁺ ILC1s, Ly49E⁻ ILC1s, and cNK cells among liver group 1 ILCs from WT mice at different ages. Data for each time point are representative of three independent experiments with n = 3–6 mice. (C) Flow cytometric plots showing Ly49E expression among CD45⁺ leukocytes from neonatal liver (0–3 d old). Data are representative of at least three independent experiments. (D and E) Lethally irradiated recipient CD45.1⁺ mice received donor CD45.2⁺ BM cells (6–8 wk old) or CD45.2⁺ fetal liver cells (E13.5–15.5). 1 mo later, the livers of recipient mice were harvested for flow cytometric analysis of Ly49E expression on ILC1s (D). Percentages of BM- or fetal liver–derived Ly49E⁺ cells among donor-derived ILC1s (E). Data are representative of two independent experiments with n = 3 or 4 mice. (F) Absolute cell numbers of Ly49E⁻ ILC1s in the liver of WT and Klra5^DTR mice at 1 or 4 wk after intrahepatic (i.h.) treatment with DT once during the neonatal period (0–3 d old). Data were pooled from two or three independent experiments with n = 5–16 mice. Bar graphs show the mean ± SEM; unpaired Student’s t test was used in E, and one-way ANOVA followed by Dunnett’s multiple comparison test in F; ****, P < 0.0001.

Figure 3.

BM hematopoiesis contributes little to the Ly49E⁺ ILC1 pool. (A) Representative plots showing Ly49E expression on liver ILC1s from WT B6 mice at different ages. Data are representative of three independent experiments with n = 3–9 mice per condition. (B) Percentages of Ly49E⁺ or Ly49E⁻ cells among liver ILC1s at different ages. Data for each time point are representative of three independent experiments with n = 3–9 mice. (C) Absolute cell numbers of ILC1 subsets per liver from WT B6 mice at ages 0–3 d or 23–30 wk. Data for 0–3 d are pooled from two independent experiments with n = 8 mice. Data for 23–30 wk are pooled from three independent experiments with n = 10 mice. (D and E) Fate-mapping analysis of neonatal (1 d and 1 wk) and adult (5 wk) Fgd5⁺ LSK cell–derived group 1 ILCs. Representative plots show tdTomato expression on BM LSK cells and liver group 1 ILC subsets from Fgd5^CreERT2Rosa26^tdTomato mice exposed to tamoxifen at different ages (D). Labeling efficiency of Ly49E⁻ ILC1s, Ly49E⁺ ILC1s, and cNK cells normalized against LSK cells (E). Data are pooled from at least three independent experiments (1 d: n = 5; 1 wk: n = 4; 5 wk: n = 4). (F–H) Representative plots showing Ly49E expression on liver ILC1s from lethally irradiated recipient CD45.1⁺ mice that received the adoptive transfer of donor CD45.2⁺ BM (6–8 wk old) and CD45.1⁺CD45.2⁺ fetal liver (E13.5 fetal liver [FL]) MNCs (mixed at 1:1; F). Absolute cell numbers of donor BM- or FL-derived ILC1s in recipient livers (G). Percentages of donor BM- or FL-derived Ly49E⁺ cells among donor-derived ILC1s (H). Data are representative of at least two independent experiments with n = 3 or 7 mice per group. Bar graphs show the mean ± SEM; two-way ANOVA was used in B, paired Student’s t test in C, one-way ANOVA followed by Dunnett’s multiple comparison test in E and H, and unpaired Student’s t test in G; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 4.

Ly49E⁺ ILC1s are primarily derived from embryonic hematopoiesis. (A and B) tdTomato expression on fetal liver LSK cells (E14.5), ILC1 subsets (E16.5), adult BM LSK cells, and adult liver ILC1 subsets from Fgd5^CreERT2Rosa26^tdTomato mice pulsed with 4-OHT at E8.5 (A). Labeling efficiency of Ly49E⁻ ILC1s and Ly49E⁺ ILC1s in A were normalized to LSK cells (B). Data are pooled from at least three experiments (fetus: n = 7; adult: n = 4). (C and D) tdTomato expression on liver Ly49E⁻ ILC1s, Ly49E⁺ ILC1s, and cNK cells from E19.5, 1-wk-old, and 4-wk-old Ncr1^CreERT2Rosa26^td-Tomato mice pulsed with 4-OHT at E17.5, Rosa26^td-Tomato mice were used as controls (C). Contribution of fetus-derived Ly49E⁻ ILC1s, Ly49E⁺ ILC1s, and cNK cells were calculated as labeling efficiency of the cells from Ncr1^CreERT2Rosa26^tdTomato mice at 1 or 4 wk relative to that at E19.5, and the unlabeled cells were considered postnatally derived (D). Data are pooled from at least three independent experiments (E19.5: n = 4; 1 wk: n = 5; 4 wk: n = 5). (E and F) Representative plots (E) and percentages (F) of Ly49E⁺ cells among liver ILC1s from WT and Klra5^DTR mice 1 or 4 wk after intrahepatic (i.h.) treatment with DT once during the neonatal period. Data are pooled from two or three independent experiments with n = 5 or 8 (Klra5^DTR mice), n = 8 or 15 (DT-treated WT mice), or n = 7 or 12 (DT-treated Klra5^DTR mice). (G–J) Representative plots (G and I) and percentages (H and J) of Ki67⁺ or BrdU⁺ cells among liver ILC1 subsets in neonatal (0–3 d old) and adult (6–8 wk old) B6 mice. Data are pooled from two or three independent experiments (neonate: n = 7 or 4; adult: n = 12 or 8). Bar graphs show the mean ± SEM; unpaired Student’s t test was used in B, one-way ANOVA in D and F, paired Student’s t test in H and J; *, P < 0.05; ****, P < 0.0001.

Figure S4.

Comparison of effector functions between Ly49E⁺ and Ly49E⁻ ILC1s. (A and B) Representative histograms (A) and percentages (B) of cells expressing the indicated intracellular functional molecules (GzmA, GzmB, and GzmC) in liver group 1 ILC subsets of WT B6 mice (5–8 wk old) at steady state. Data are representative of two or three independent experiments with n = 3–5 mice per experiment. (C and D) Percentages of cells expressing perforin and CD107a (C) or TNF-α and IFN-γ (D) among liver group 1 ILC subsets of WT B6 mice (5–8 wk old) after stimulation for 18 h with 10 ng/ml IL-12, 10 ng/ml IL-15, and 50 ng/ml IL-18, 4 h with PMA plus ionomycin (PMA + Ion), or with Yac-1 cells. Data are representative of two or three independent experiments with n = 3 mice per experiment. (E and F) Representative plots (E) and percentages (F) of cells expressing IL-2 in liver T cells and group 1 ILC subsets of WT B6 mice (5–8 wk old) after stimulation for 4 h with PMA and Ion. Data are representative of two independent experiments with n = 5 mice per experiment. (G) GMFI of cells expressing FasL and TRAIL among Ly49E⁺ or Ly49E⁻ ILC1s derived from WT B6 mice (5–8 wk old). Data are representative of two independent experiments with n = 3 mice per experiment. (H and I) Real-time cytotoxicity assays were performed using liver group 1 ILC subsets from WT B6 mice (3–6 wk old) as effector cells and MC38 tumor cells as the target cells, with an effector-target (E:T) ratio of 10:1. The cell index during 4 h (H) and percentages of killing efficiency at the 4-h time point (I) after incubation with effector cell. Data are representative of two independent experiments with n = 3 samples per group. Bar graphs show the mean ± SEM; one-way ANOVA followed by Dunnett’s multiple comparison test was used in B–D, F and I, and unpaired Student’s t test in G; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 5.

Ly49E⁺ ILC1s are effective in cytotoxicity. (A and B) Representative histograms (A) and percentages (B) of cells expressing the indicated intracellular functional molecules (GzmA, GzmB, and GzmC) in liver group 1 ILC subsets of WT B6 mice (5–8 wk old) after stimulation for 18 h with 10 ng/ml IL-12, 10 ng/ml IL-15, and 50 ng/ml IL-18. Fluorescence minus one (FMO) control was gated on total group 1 ILCs (A). (C and D) Percentages of cells expressing the indicated molecules among the liver group 1 ILC subsets for 4 h with PMA plus ionomycin (PMA + Ion; C) or with Yac-1 cells (D). Data are representative of two or three independent experiments with n = 3–6 mice per experiment (A–D). (E and F) Flow-based killing assays were performed by using liver group 1 ILC subsets of WT B6 mice (3–6 wk old) that were preactivated with IL-12, IL-15, and IL-18 for 18 h as effector cells and Yac-1 tumor cells as the target cells, with an effector–target (E:T) ratio of 10:1. Representative plots (E) and percentages (F) of 7-AAD–positive target cells after incubation with effector cells. Data are pooled from two independent experiments with n = 14–30 samples per group. Bar graphs show the mean ± SEM; one-way ANOVA, followed by Dunnett’s multiple comparison test, was used to compare the experimental groups in B–D and F; **, P < 0.01; ****, P < 0.0001.

Figure 6.

Ly49E⁺ ILC1s play a role in the host defense against MCMV infection during early life. (A and B) Schematic showing the experimental design used for MCMV infections in neonatal mice within 3 d of age (A). Percentages and GMFI of GzmB, perforin, or IFN-γ expressed by Ly49E⁺ or Ly49E⁻ ILC1s derived from neonatal mice at different time points after infection with MCMV-eGFP (B). Data presented for each time point are representative of at least three independent experiments with n = 3–5 mice. (C) Representative images of the intact liver of neonatal WT and Tbx21^−/− mice 0.5 dpi with MCMV-eGFP; scale bar = 100 μm. (D and E) egfp expression (D) and viral copies (E) in the livers of neonatal Tbx21^−/− and WT mice were measured 0.5, 1.5, 2.5, and 3.5 dpi with MCMV-eGFP. Data are pooled from two independent experiments with n = 4–12 mice per group. (F and G) 1–1.5 × 10⁵ neonatal Ly48E⁺ or Ly49E⁻ ILC1s were transferred into neonatal Tbx21^−/− mice 1 d before MCMV-eGFP infection. Liver viral copies were measured 3.5 dpi (F). Data were pooled from three independent experiments with n = 9–12 mice per group. Survival of neonatal Tbx21^−/− mice that received Ly49E⁺ or Ly49E⁻ ILC1s followed by MCMV infection (G). Data are pooled from three independent experiments with n = 7–14 mice per group. (H and I) Neonatal Klra5^DTR mice were administered DT 10 h before MCMV-eGFP infection. Liver viral copies were measured 3.5 dpi. MCMV-infected Klra5^DTR littermates and DT-treated WT mice were as controls (H). Data were pooled from three independent experiments with n = 6–26 mice per group. Survival of neonatal Klra5^DTR mice were shown (I). Data are pooled from two independent experiments with n = 9 mice per group. (J) Viral titers in the livers of neonatal Ncr1^Cre/+Eomes ^fl/fl and Eomes ^fl/fl mice were measured 3.5 dpi with MCMV-eGFP. Data are pooled from three independent experiments with n = 11–16 mice per group. Bar graphs show the mean ± SEM; two-way ANOVA was used in B, D, and E, one-way ANOVA followed by Dunnett’s multiple comparison test in F and H, unpaired two-tailed Student’s t test in J, and log-rank (Mantel–Cox) test in G and I; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S5.

Ly49E⁺ ILC1s mediate protective responses against MCMV infection in neonatal mice. (A) Percentages of Ly49E⁺ ILC1s, Ly49E⁻ ILC1s, and cNK cells among liver group 1 ILCs from neonatal WT mice within 3 d of age after infection with MCMV-eGFP. Data for each time point are representative of three independent experiments with n = 4–5 mice. (B) Representative plots showing Ly49H expression on neonatal liver group 1 ILCs after MCMV-eGFP infection. (C) Flow cytometric analysis of GzmB expression in CD45⁺ leukocytes from neonatal livers before and after MCMV-eGFP infection. (D) Viral copies in the spleen of neonatal Tbx21^−/− and WT mice were measured 0.5, 1.5, 2.5, and 3.5 dpi with MCMV-eGFP. Data are representative of at least three independent experiments with n = 3 mice per group. (E) 1–1.5 × 10⁵ neonatal Ly48E⁺ or Ly49E⁻ ILC1s were transferred into neonatal Tbx21^−/− mice 1 d before MCMV-eGFP infection. Spleen viral copies were measured 3.5 dpi. Data were pooled from two independent experiments with n = 7–12 mice per group. (F) The depletion effect of Klra5^DTR mice after DT treatment. Plots are representative of at least three independent experiments. (G) Neonatal Klra5^DTR mice were administered DT 10 h before MCMV-eGFP infection. Spleen viral copies were measured 3.5 dpi. Data were pooled from three independent experiments with 21–23 mice. (H and I) 0–3-d-old neonatal Ncr1^Cre/+Eomes^fl/+ and Ncr1^Cre/+Eomes^fl/fl mice were infected with MCMV-eGFP. The absolute cell number of liver ILC1s (H) and viral titers in the spleen (I) were measured 3.5 d later. **, P < 0.01. Data are representative of three independent experiments with five to nine mice per experiment. Bar graphs show the mean ± SEM; two-way ANOVA was used in D and E, unpaired two-tailed Student’s t test in G–I.