Activation of group 2 innate lymphoid cells alleviates aging-associated cognitive decline

Abstract

Increasing evidence has challenged the traditional view about the immune privilege of the brain, but the precise roles of immune cells in regulating brain physiology and function remain poorly understood. Here, we report that tissue-resident group 2 innate lymphoid cells (ILC2) accumulate in the choroid plexus of aged brains. ILC2 in the aged brain are long-lived, are relatively resistant to cellular senescence and exhaustion, and are capable of switching between cell cycle dormancy and proliferation. They are functionally quiescent at homeostasis but can be activated by IL-33 to produce large amounts of type 2 cytokines and other effector molecules in vitro and in vivo. Intracerebroventricular transfer of activated ILC2 revitalized the aged brain and enhanced the cognitive function of aged mice. Administration of IL-5, a major ILC2 product, was sufficient to repress aging-associated neuroinflammation and alleviate aging-associated cognitive decline. Targeting ILC2 in the aged brain may provide new avenues to combat aging-associated neurodegenerative disorders.

Figure S1.

The expression of many ILC2 characteristic genes is upregulated in the CP of aged mice. Heatmap depicts the expression of the indicated genes in the CP of young and aged mice by genome-wide microarray analysis.

Figure 1.

ILC2 accumulate in the aged brain. (A) Representative flow cytometry profiles of hematopoietic cells in the brains of young (2–3-mo-old) and aged (18–22-mo-old) mice. (B) Quantification of ILC2 numbers in the brains of young and aged mice. Data are from seven mice per group, pooled from two independent experiments. (C) Representative flow cytometry profiles of ILC2 and T cells from the brains of young and aged mice. FSC-A, forward scatter area. (D) Representative flow cytometry profiles of brain ILC2 and T cells from young and aged mice injected i.v. with anti-CD45.2 PE antibody and euthanized 3 min after injection. PB, pacific blue. (E) Percentages of circulating cells labeled with i.v. injected anti-CD45.2 PE antibody. Data are from three mice per group and are representative of two independent experiments. (F) The number of various ILC subsets in the brain of young (Y) and aged (A) mice. Data are from four mice per group and are representative of two independent experiments. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01.

Figure 2.

ILC2 accumulate in the CP of the aged brain. (A) Quantification of ILC2 numbers in different regions of young and aged mice. Data are from five mice per group, pooled from two independent experiments. CTX, cortex; PFC, prefrontal cortex; STR, striatum; TH, thalamus; SVZ, subventricular zone. CP lining the lateral ventricle (LV), leptomeninges attached to the brain parenchyma of dorsal cerebrum (DC), and dura/arachnoid meninges attached to the skull (S) were isolated. (B) Representative flow cytometry profiles of ILC2 in the CP of aged mice. (C) The number of various ILC subsets in the CP of aged mice. Data are four mice per group and are representative of two independent experiments. (D) Representative flow cytometry profiles of ILC2 in the CP of young and aged mice. (E) Quantification of ILC2 numbers in the CP of young (Y) and aged (A) mice. Data are from six mice per group, pooled from two independent experiments. (F) Representative whole-mount CP immunofluorescence staining profiles of ILC2 and T cells from young and aged mice. Scale bar length, 10 µM. White arrowheads denote ILC2; blue arrowheads denote CD3⁺ T cells. Data represent six mice per group from two independent experiments. (G) Quantification of ILC2 numbers in the CP of mice at different ages. Data are from three or four mice per group and are representative of two independent experiments. (H) Representative flow cytometry profiles of ILC2 in the CP of deceased elderly people >65 yr old. Data represent four autopsy samples. (I) Representative flow cytometry profiles depicting the expression of CRTH2 and cKit in human CP ILC2 from deceased elderly people. Data represent four autopsy samples. (J) mRNA levels of the indicated genes in sorted ILC2, CD3⁺ T cells, and CD11b⁺ myeloid cells in the CP tissue of deceased elderly people. Data are from four autopsy samples. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01. Lin1 antibodies included anti-B220, anti-NK1.1, and anti-CD11b. Mouse total Lin antibodies included anti-B220, anti-NK1.1, anti-CD11b, anti-CD3, and anti-CD5. Human Lin1 antibodies included anti-CD56, anti-CD19, anti-CD11c, anti-CD11b, anti-CD14, and anti-FcγR1α.

Figure 3.

ILC2 in the aged brain are long-lived, capable of switching between cell cycle quiescence and proliferation, and relatively resistant to cellular senescence and exhaustion. (A) mRNA expression of Mki67 (encoding Ki67) in ILC2 sorted from the brains of young and aged mice. Data are from four mice per group and are representative of two independent experiments. (B) Mean fluorescence intensity (MFI) of Ki67 in ILC2 from young and aged mice, assessed by intracellular staining and flow cytometry analysis. Data are from three mice per group and are representative of two independent experiments. (C) Schematic for IL-33 injections in D–G. Young or aged mice were administrated with IL-33 daily for 7 d. Brain ILC2 in young and aged mice were examined at day 0 (naive mice), day 8 (24 h after the last IL-33 administration), and day 36 (4 wk after the last IL-33 administration). (D) Quantification of brain ILC2 numbers at day 0, day 8, and day 36. Data are from three to seven mice per group and are representative of two independent experiments. (E) Fold change of brain ILC2 numbers at day 0, day 8, and day 36. The average ILC2 number at day 0 was arbitrarily set as 1. Data are from three to seven mice per group and are representative of two independent experiments. (F) EdU was administrated simultaneously with IL-33 treatment. EdU concentrations in ILC2 were measured at day 0 (naive mice), day 8 (24 h after the last EdU/IL-33 administration), and day 36 (4 wk after the last EdU/IL-33 administration). Percentages of EdU⁺ brain ILC2 at day 0, day 8, and day 36. Data are from three to seven mice per group and are representative of two independent experiments. (G) MFI of Ki67 in brain ILC2 at day 0, day 8, and day 36. Data are from three to seven mice per group and are representative of two independent experiments. (H) Percentage of ILC2 that expressed activated caspases at day 29. Data are from six mice per group, pooled from two independent experiments. (I) Aged mice were treated with IL-33 for 7 d together with EdU. The mice were then rested for 4 wk and administrated with IL-33 or PBS again for 2 d. EdU concentrations were measured at 7 d after secondary injections with IL-33 or PBS. Ki67 expression was measured at 24 h after secondary injections with IL-33 or PBS. (J) Percentages of Edu⁺ ILC2 at 7 d after secondary injection of IL-33 or PBS. Data are from three mice per group and are representative of two independent experiments. (K) MFI of Ki67 in aged brain ILC2 at 24 h after secondary injections with IL-33 or PBS. Data are from three mice per group and are representative of two independent experiments. (L) ILC2 from the brains of young and aged mice were sorted and cultured with IL-7, IL-2, IL-33, and SCF for up to 4 wk. Fold change of cell numbers over time is shown. Data are from three mice per group and are representative of three independent experiments. (M) Cytokine concentrations in the culture supernatant of ILC2 isolated from aged brains and cultured in vitro for 4 wk. Data are from three mice per group and are representative of two independent experiments. (N) Heatmap depicts the expression of representative genes by RNA-Seq comparing ILC2 isolated from the brains of young versus aged mice and cultured for 2 wk in vitro. (O) qPCR showing the expression of the indicated genes in ILC2 isolated from the brains of young and aged mice and cultured for 2 or 4 wk in vitro. Data are from three mice per group and are representative of two independent experiments. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01. n.s., not significant; U.D., undetectable.

Figure 4.

ILC2 in the aged brain are functionally dormant at homeostasis but can be activated by IL-33 to produce large amounts of type 2 cytokines in vivo. (A) Representative flow cytometry profiles showing the expression of IL-5 and IL-13 by brain ILC2 in young versus aged mice at homeostasis. (B) MFI of IL-5 and IL-13 by total brain ILC2 or cytokine-expressing ILC2 in young versus aged mice at homeostasis. Data are from three mice per group and are representative of three independent experiments. (C) Representative flow cytometry profiles of the expression of IL-5 and IL-13 by brain ILC2 in young or aged mice treated with IL-33 or PBS daily for 7 d. (D) MFI of IL-5 and IL-13 expressed by brain ILC2 in young or aged mice treated with IL-33 or PBS daily for 7 d. Data are from three to six mice per group and are representative of two independent experiments. (E) Experimental scheme for the Object Placement Test. (F) Percentage of time spent with familiar or displaced objects for young and aged mice. Data are from 9 or 10 mice per group and are representative of two independent experiments. (G) Distance traveled by aged mice treated with IL-33 or PBS in the Open Field Test. Data are from 9–14 mice per group and are representative of two independent experiments. (H) Percentage of time spent with familiar or displaced objects for aged mice treated with IL-33 or PBS daily for 2 d. Data are from 9–14 mice per group and are representative of two independent experiments. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01. n.s., not significant.

Figure 5.

Activated ILC2 improve cognitive function in aged mice. (A) ILC2 were isolated from brains of aged mice and cultured in vitro with IL-33, IL-2, IL-7, and SCF for 4 wk. The Object Placement Test was performed with aged mice that received i.c.v. injection of 2 × 10⁵ in vitro expanded and activated brain ILC2 or PBS. Mice were rested for 1 wk after surgery before behavior testing. Distance traveled in the Open Field Test was quantified. Data are from 8–11 mice per group, pooled from two independent experiments. (B) Percentage of time spent with familiar or displaced objects in the Object Placement Test. Data are from 8–11 mice per group, pooled from two independent experiments. (C) The Morris Water Maze was performed with mice that were i.c.v. injected with in vitro expanded and activated brain ILC2 or PBS. Percentage of time spent in the target quadrant and the number of entries to the target zone were quantified. Data are from 9–11 mice per group, pooled from two independent experiments. (D) RNA-Seq was performed with sorted microglia, ILC2, or T cells from the brains of aged mice treated with IL-33 daily for 7 d. Principal component analysis was performed. (E) MA (log-ratio versus mean log-expression) plots comparing the gene expression profiles. (F) Venn diagrams depicting the numbers of differentially expressed genes. (G) Pathway analysis of differentially expressed genes. (H) Pathway analysis of genes that are highly expressed in brain ILC2 but not in microglia or brain T cells of IL-33–treated aged mice. (I) List of effector molecules, receptors, and transcription factors that are highly expressed in brain ILC2 of IL-33–treated aged mice. (J) Heatmap depicts the expression of effector molecules expressed by different immune cell subsets in the brains of IL-33–treated aged mice. (K) Representative flow cytometry profiles of IL-5 and IL-13 expression by brain T cells or ILC2 in PBS- or IL-33–treated aged mice. (L) Quantification of the number of IL-5– or IL-13–producing brain T cells and ILC2 in IL-33–treated aged mice. Data are from three mice per group and are representative of two independent experiments. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01. n.s., not significant.

Figure S2.

Behavior test results in young and aged mice treated with PBS, IL-5, or IL-13. (A) The Morris Water Maze test was performed with young or aged mice. Percentage of time spent in the target quadrant was quantified. Data are from six to eight mice per group and are representative of two independent experiments. (B) Distance traveled in the Open Field Test for aged mice treated with PBS, IL-5, or IL-13. Data are from 9 or 10 mice per group and are representative of two independent experiments. (C) The Object Placement Test was performed with young mice treated with PBS, IL-5, or IL-13. Percentage of time with the displaced object was quantified. Data are from eight mice per group and are representative of two independent experiments. (D) The Morris Water Maze test was performed with young mice treated with PBS, IL-5, or IL-13. Percentage of time spent in the target quadrant was quantified. Data are from 11–15 mice per group, pooled from two independent experiments. Error bars are mean ± SEM. *, P < 0.05.

Figure S3.

Tissue residency and gene expression of brain ILC2 in IL-33–treated aged mice. (A) Young or aged mice were treated with IL-33 daily for 7 d. At 24 h after the last injection, mice were injected with anti-CD45.2 PE antibody i.v. and euthanized 3 min later. Flow cytometry was performed to examine circulating cells that were labeled with i.v. injected anti-CD45.2 PE antibody. Representative flow cytometry profiles. Plots were pregated on brain T cells or ILC2. (B) Percentages of circulating cells in IL-33–treated mice. Data are from three mice per group and are representative of two independent experiments. (C) qPCR analysis of the indicated genes in ILC2 sorted from different anatomical sites at homeostasis. Data are from three mice per group, pooled from three independent experiments. SILP, small intestinal laminal propria; BM, bone marrow. (D) Expression of the indicated genes in CP and meninges of ILC2 in IL-33–treated mice by scRNA-Seq. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01.

Figure 6.

CP ILC2 possess more potent proliferative and cytokine-producing capability than meningeal ILC2 in aged mice. (A) Number of ILC2 at different regions of aged mice after daily IL-33 treatment for 7 d. Data are from four mice per group and are representative of two independent experiments. CTX, cortex. (B) ILC2 from the meninges of young mice and the meninges of aged mice and CP from aged mice were sorted and cultured with IL-7, IL-2, IL-33, and SCF for up to 4 wk. Fold change of cell numbers over time is shown. Data are from four mice per group and are representative of two independent experiments. (C) Numbers of meningeal and CP ILC2 in aged mice treated with PBS or IL-33 daily for 7 d. Data are from four mice per group and are representative of two independent experiments. (D) Fold change of ILC2 numbers in IL-33–treated mice over those in PBS-treated mice. The average ILC2 number in PBS-treated mice was set as 1. Data are from four mice per group and are representative of two independent experiments. (E) Representative flow cytometry profiles of IL-5 and IL-13 expression in meningeal and CP ILC2 in IL-33–treated aged mice. (F) MFI of IL-5 and IL-13 in meningeal and CP ILC2 in IL-33–treated aged mice. Data are from five mice per group and are representative of two independent experiments. (G) scRNA-Seq was performed with sorted meningeal and CP ILC2 from IL-33–treated aged mice. UMAP analysis of the combined ILC2 and meningeal ILC2 population. (H) Expression of the indicated genes by different subsets of ILC2 in the combined meningeal and CP ILC2 population. (I) Split analysis of meningeal and CP ILC2 by UMAP. (J) Percentage of each ILC2 subset in meningeal and CP ILC2. (K) Expression of the indicated genes by meningeal and CP ILC2. (L) Expression of the indicated genes in each ILC2 subset in meningeal and CP ILC2. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01.

Figure 7.

IL-5 improves the cognitive function in aged mice. (A) The Object Placement Test was performed with aged mice after injection of IL-5, IL-13, or PBS daily for 2 d. Percentage of time spent with the displaced objects was quantified. Data are from 9 or 10 mice per group and are representative of two independent experiments. (B) The Morris Water Maze test was performed with aged mice after PBS, IL-5, or IL-13 treatment for 2 d. Percentage of time spent in the target quadrant was quantified. Data are from 9 or 10 mice per group and are representative of two independent experiments. (C) IL-5 was deleted in activated ILC2/b6 cells using the CRISPR-mediated gene knockout technique. The Object Placement Test was performed with aged mice 1 wk after i.c.v injection with control or IL-5–deficient ILC2/b6 cells. Percentage of time spent with displaced objects was quantified. Data are from 9 or 10 mice per group, pooled from two independent experiments. (D) Aged mice received i.c.v injection of 10 µg anti–IL-5 antibody or isotype control, followed by i.p. injection of IL-33 or PBS daily for 2 d. The Object Placement Test was performed after 1 wk of i.c.v injection. Percentage of time spent with the displaced objects was quantified. Data are from 8–11 mice per group, pooled from two independent experiments. (E) IL-5- or PBS-treated aged mice were injected with EdU in vivo. EdU⁺ cells in the HC DG regions was examined and quantified by immunofluorescence at 2 d after the last dose of IL-5 or PBS treatment. Data are from four mice per group. (F) Heatmap depicts the gene expression profiles of CD3⁺ T cells from the brains of PBS- or IL-5–treated aged mice by RNA-Seq. (G) Numbers of CD4⁺ and CD8⁺ T cells in the brains of young (Y) or aged (A) mice treated with PBS or IL-5. Data are from four mice per group and are representative of three independent experiments. (H) Representative flow cytometry profiles of TNFα expression by CD4⁺ or CD8⁺ T cells in the brains of aged mice treated with PBS or IL-5. (I) MFI of TNFα in brain CD4⁺ or CD8⁺ T cells of aged mice treated with PBS or IL-5. Data are from three to five mice per group and are representative of three independent experiments. (J) Representative flow cytometry profiles of CD8⁺ cells in the brains of aged mice that were injected i.v. with anti-CD45.2 PE antibody and euthanized 3 min later. PB, pacific blue. (K) The number of noncirculating brain CD8⁺ T cells that were not labeled with i.v. injected anti-CD45.2 antibody. Data are from three mice per group and are representative of two independent experiments. (L) The number of noncirculating CD8⁺ T cells in the CP of young or aged mice treated with PBS or IL-5. Data are from four mice per group and are representative of two independent experiments. (M) The number of microglia in IL-5– or PBS-treated aged mice. Data are from three mice per group and are representative of two independent experiments. (N) Representative flow cytometry profiles of CD68 expression by microglia in IL-5– or PBS-treated aged mice. (O) MFI of CD68 on microglia of IL-5– or PBS-treated aged mice. Data are from three mice per group and are representative of two independent experiments. (P) Heatmap depicts the gene expression profiles of microglia from PBS- or IL-5–treated aged mice by RNA-Seq. (Q) CD4⁺ or CD8⁺ T cells from the brains of aged mice were sort-purified and cultured in the presence or absence of IL-5 for 24 h. Tnf expression was examined by qPCR. Data are from six mice per group and are representative of two independent experiments. (R) Expression of the indicated genes in CD8⁺ T cells sorted from the aged brain and cultured in the presence or absence of IL-5 for 24 h. Data are from six mice per group and are representative of two independent experiments. (S) Annexin V staining was performed at 3 d of culture. Data are from four mice per group and are representative of two independent experiments. Error bars are mean ± SEM. *, P < 0.05; **, P < 0.01. n.s., not significant.

Figure S4.

Efficient deletion of IL-5 in ILC2/b6 cells. (A) Brain ILC2 from aged mice were transduced with lenti-CRISPRv2-GFP lentivirus containing an NTC sequence. GFP expression was examined at 5 d after transduction. Cells without lentiviral transduction, lung ILC2 sorted from young mice, and an immortalized ILC2 line ILC2/b6 were used as controls. Data represent four independent experiments. (B) ILC2/b6 line cells were transduced with lenti-CRISPRv2-GFP lentivirus containing a gRNA targeting IL-5 or NTC. Expression of IL-5 was examined at 10 d after transduction. Data represent three independent experiments.

Figure S5.

IL-5 did not induce proliferation of brain eosinophils or the apoptosis of spleen CD8⁺ T cells in aged mice. (A) Aged mice were treated with IL-5 daily for 2 d. Brains were harvested, and immune cells were examined by flow cytometry analysis at 24 h after last injection. Bronchoalveolar lavage (BAL) cells from young mice that inhaled 40 µg of Alternaria alternata extract daily for 3 d were used as positive controls for eosinophil staining. Data are from three mice per group and are representative of two independent experiments. (B) Spleen CD8⁺ T cells were sorted from aged mice and cultured with anti-CD3, anti-CD8, IL-7, and IL-2 in the presence or absence of IL-5. mRNA of Tnf was examined at 24 h of culture. (C) Annexin V staining was performed with aged spleen CD8⁺ T cells after 3 d of culture. Data are from three mice per group and are representative of two independent experiments. Error bars are mean ± SEM.