Liver group 2 innate lymphoid cells regulate blood glucose levels through IL-13 signaling and suppression of gluconeogenesis

Fujimoto, Masanori; Yokoyama, Masataka; Kiuchi, Masahiro; Hosokawa, Hiroyuki; Nakayama, Akitoshi; Hashimoto, Naoko; Sakuma, Ikki; Nagano, Hidekazu; Yamagata, Kazuyuki; Kudo, Fujimi; Manabe, Ichiro; Lee, Eunyoung; Hatano, Ryo; Onodera, Atsushi; Hirahara, Kiyoshi; Yokote, Koutaro; Miki, Takashi; Nakayama, Toshinori; Tanaka, Tomoaki

doi:10.1038/s41467-022-33171-6

Download PDF

Article
Open access
Published: 15 September 2022

Liver group 2 innate lymphoid cells regulate blood glucose levels through IL-13 signaling and suppression of gluconeogenesis

Nature Communications volume 13, Article number: 5408 (2022) Cite this article

4992 Accesses
9 Citations
19 Altmetric
Metrics details

Subjects

Abstract

The liver stores glycogen and releases glucose into the blood upon increased energy demand. Group 2 innate lymphoid cells (ILC2) in adipose and pancreatic tissues are known for their involvement in glucose homeostasis, but the metabolic contribution of liver ILC2s has not been studied in detail. Here we show that liver ILC2s are directly involved in the regulation of blood glucose levels. Mechanistically, interleukin (IL)-33 treatment induces IL-13 production in liver ILC2s, while directly suppressing gluconeogenesis in a specific Hnf4a/G6pc-high primary hepatocyte cluster via Stat3. These hepatocytes significantly interact with liver ILC2s via IL-13/IL-13 receptor signaling. The results of transcriptional complex analysis and GATA3-ChIP-seq, ATAC-seq, and scRNA-seq trajectory analyses establish a positive regulatory role for the transcription factor GATA3 in IL-13 production by liver ILC2s, while AP-1 family members are shown to suppress IL-13 release. Thus, we identify a regulatory role and molecular mechanism by which liver ILC2s contribute to glucose homeostasis.

A MAFG-lncRNA axis links systemic nutrient abundance to hepatic glucose metabolism

Article Open access 31 January 2020

Multiomics reveals persistence of obesity-associated immune cell phenotypes in adipose tissue during weight loss and weight regain in mice

Article Open access 26 May 2022

The transcriptional corepressor CtBP2 serves as a metabolite sensor orchestrating hepatic glucose and lipid homeostasis

Article Open access 02 November 2021

Introduction

The risk factors for cardiovascular diseases and stroke are closely linked to pathological conditions that are collectively referred to as metabolic syndrome. Over time, excessive glucose levels cause organ complications in patients with diabetes, and fat accumulates in patients who are obese or overweight. The tissue deposition of lipids coupled with chronic inflammation causes decreased insulin sensitivity, termed insulin resistance, which can lead to hyperglycemia, hyperinsulinemia, and tissue damage, including damage to blood vessels. Hepatic gluconeogenesis is critical in the process of glucose homeostasis because it promotes the production of glucose from stored glycogen, amino acids, and lactate in response to hypoglycemia. Glucose-6-phosphatase (G6PC) and PCK1 are key enzymes in this process, and G6PC is especially indispensable because it converts glucose-6-phosphate into glucose in the final pathway step. Insulin suppresses excessive gluconeogenesis, but insulin resistance in the liver interferes with the proper regulation of glucose production. This means that lipid deposition in the liver under obese conditions weakens insulin signaling in hepatocytes, leading to excessive gluconeogenesis and blood glucose elevation. Hepatic insulin resistance is a pathological hallmark of nonalcoholic fatty liver disease (NAFLD)^1,2,3, and its prevalence continues to increase. The drug metformin is administered to patients with diabetes to improve insulin resistance and mechanistically exerts its effects by suppressing hepatic gluconeogenesis. In addition to the current approaches, improvement of gluconeogenesis has attracted the attention of researchers as a medical treatment option.

Group 2 innate lymphoid cells (ILC2) are innate cells of lymphoid origin that develop from a common lymphoid progenitor and play pivotal roles in immune protection against helminth infection by secreting type 2 cytokines such as interleukin (IL)-4, IL-5, IL-9, and IL-13. Inflammation associated with type 2 immunity mediates responses to tissue repair, goblet cell hyperplasia, and autoimmune diseases such as arthritis. ILC2s are regulated by soluble factors, including cytokines and hormones, and IL-33 is known as a major activator of ILC2s. In addition to known roles in immunity^4,5,6, ILC2s have beneficial roles in metabolic disorders, and they reportedly limit adiposity by increasing caloric expenditure of beige adipocytes in white adipose tissue^7,8. Another report has suggested that IL-33-activated islet ILC2s produce IL-13 and accelerate insulin secretion via the production of retinoic acid in myeloid cells⁹. Mallat and colleagues reported that ILC2-derived IL-13 protects against atherosclerosis in aortic tissues¹⁰. The functions of ILC2s in the liver were previously reported to be related to intrahepatic and extrahepatic fibrosis^11,12, and ILC2s in the liver appear to respond to physiological conditions. However, the metabolic functions of ILC2s in the liver and whether they differ from the functions of ILC2s found in adipose or pancreatic tissues remain unknown.

Here, we report that IL-33 treatment increases liver ILC2 numbers, decreases blood glucose levels, and downregulates the gluconeogenic enzyme G6pc in the livers of both normal and obese mice. However, these effects are absent in ILC2-null NSG mice, indicating that IL-33-mediated suppression of gluconeogenesis is dependent on ILC2s. The results of single-cell RNA sequencing (scRNA-seq) analysis of liver tissue show that liver ILC2s suppress gluconeogenesis in the G6pc-high hepatocyte cluster, possibly through downregulation of G6pc expression via signal transducer and activation of transcription 3 (STAT3) signaling. In addition, we identify AP-1 family as GATA3-binding proteins and that JunB suppresses GATA3 function, resulting in reduced IL-13 expression. These observations uncover the additional function of liver ILC2s and the role of the AP-1 family in GATA3 signaling for IL-13 regulation. Our findings provide insights into the potential functions of liver ILC2s in the modulation of gluconeogenesis and blood glucose levels and may present a new strategy for diabetes treatment.

Results

IL-33 lowers blood glucose levels by limiting gluconeogenesis via liver ILC2s

As increases in the levels of gluconeogenic enzymes are significant contributing factors to blood glucose elevation in obese humans and mice, we first investigated whether ILC2 activation could reduce blood glucose and gluconeogenic enzyme levels in the livers of obese mice. We intraperitoneally (i.p.) injected leptin receptor-deficient (db/db) obesity model mice with phosphate-buffered saline (PBS) or 0.5 µg of recombinant IL-33 (rIL-33) to activate ILC2s for 5 consecutive days. rIL-33 injection decreased fasting blood glucose levels and reduced the gene expression levels of liver G6pc, a rate-limiting enzyme of gluconeogenesis (Fig. 1a, b), which is consistent with a previous report¹³. T cells are major immune cells found in liver tissue that also respond to rIL-33; therefore, we compared the parameters of glucose metabolism among specific immune cell-deficient models to clarify which cell types contributed to the rIL-33-induced glucose-lowering effect. Nude (Foxn^nu/nu) mice and NOD/Scid/Il2Rγ^null (NSG) mice were used as deficiency models lacking mature T cells and both T and ILC2s cells, respectively. In wild-type (wt) and nude mice, rIL-33 injection decreased fasting blood glucose levels (Fig. 1c), significantly reduced fasting insulin levels (Fig. 1d), and increased liver glycogen contents (Fig. 1e). However, these responses were not observed in NSG mice, indicating that the rIL-33-induced glucose-lowering effect was mediated by ILC2s in addition to glycogen storage in the liver.

**Fig. 1: Recombinant IL-33 (rIL-33) injection exerts a blood glucose-lowering effect by limiting gluconeogenesis, which may be mediated by liver group 2 innate lymphoid cells (ILC2).**

We then subjected mice to a pyruvate tolerance test (PTT) to determine whether rIL-33 could suppress gluconeogenesis and glucose elevation in wt mice. As expected, rIL-33 injection suppressed blood glucose elevation during PTT in wt and nude mice but not in NSG mice (Fig. 1f and Supplementary Fig. 1a). Similar to the test with pyruvate, we also compared the effect of rIL-33 in a challenge test with glycerol, another substrate of gluconeogenesis, in rIL-33-responsive wt and nude mice. We confirmed that rIL-33 suppressed glucose elevation in both types of mice subjected to the glycerol tolerance test (GlycerolTT) (Supplementary Fig. 1b). For de novo glucose production, pyruvate required both G6PC and PCK1, whereas glycerol required G6PC but bypassed PCK1. These results suggested that the G6PC-dependent process of gluconeogenesis was one of the sites of action of rIL-33. Furthermore, in an insulin tolerance test (ITT), rIL-33 injection suppressed blood glucose recovery during the late phase (90 − 240 min) in wt mice but not in NSG mice (Fig. 1g), consistent with a previous report describing glucose recovery during the late phase of ITT as being dependent on gluconeogenesis¹⁴. We also examined the gene expression of gluconeogenic enzymes in the liver tissues of PBS- or IL-33-treated wt nude and NSG mice. Consistent with the in vivo results, the expression of several gluconeogenic enzymes (G6pc, Pck1, and Hnf4a) was significantly suppressed by IL-33 treatment in both wt and nude mice (Supplementary Fig. 1d). Fbp1 and Pcx were inhibited somewhat, although not significantly, by IL-33 treatment in wt mice. We also found that none of the genes were significantly changed by IL-33 in NSG mice. These data suggested that the effect on gluconeogenic enzymes was mediated by ILC2s. Taken together, these findings indicated that rIL-33 administration suppressed gluconeogenesis via an ILC2-dependent mechanism.

To address whether ILC2s mediated the blood glucose-lowering effect of rIL-33, we intravenously transferred liver ILC2s (defined as Lin^-Thy1⁺IL-7Rα⁺ST2⁺) into ILC2-deficient NSG mice and then measured the blood glucose levels of the mice after 5 days of daily rIL-33 or PBS administration (Fig. 1h). We confirmed the presence of ILC2s in the livers of ILC2-transferred NSG mice (Fig. 1i). While rIL-33 had no effect on fasting blood glucose levels in NSG mice that did not receive transfers (left two groups in Fig. 1j), rIL-33 reduced the fasting blood glucose levels of NSG mice that received ILC2 transfers (right group in Fig. 1j). Thus, the results of the transfer experiments indicated that rIL-33 reduced blood glucose levels via liver ILC2s. Collectively, these data showed that liver ILC2s limited gluconeogenesis in response to rIL-33 injection.

To explore whether liver ILC2s played roles in physiological models, we evaluated liver ILC2s in a 3-month normal diet (ND) or high-fat diet (HFD)-fed model. HFD-fed mice showed clear increases in fasting blood glucose, body weight, and blood glucose in the PTT and ITT (Supplementary Fig. 2a–c). These data confirmed that gluconeogenesis was enhanced in this model. We next performed flow cytometry analysis and found clear expansion of ILC2s in livers from the HFD-fed mice compared with those of the ND-fed mice (Supplementary Fig. 2d). These data might suggest that ILC2s play a compensatory role in this model.

Il13 is highly expressed in liver ILC2s, as revealed by scRNA-seq analysis

To characterize the molecular features of activated liver ILC2s following rIL-33 administration, we performed scRNA-seq analyses of sorted ILC2-enriched immune cell populations. We compared liver ILC2s and lung ILC2s to document the heterogeneity among organs, including differences in the presence and absence of rIL-33. Lung ILC2s were used as well-characterized nonliver ILC2 controls. We examined the liver and lung immune cell types sorted from wt mice injected with PBS (control) or rIL-33 for 5 days. The cells were classified based on differentially expressed genes (DEGs) and the expression of known markers (Fig. 2a–c). Most of the sorted cells were defined as ILC2s, but other cell types were also identified according to marker profiles. We excluded cells expressing non-ILC2 markers such as Cd3, Cd11b, Fceri and Cd45r and reclustered the remaining 14,026 ILC2s into 10 clusters (Fig. 3a). ILC2s from rIL-33-treated lungs were distributed to all clusters, whereas rIL-33-treated liver ILC2s were enriched in specific clusters (1, 2, 4, 6, and 8) (Fig. 3b, c). GATA3 is a crucial transcription factor involved in cell survival and cytokine production¹⁵ in ILC2s, and Il-1 receptor-like 1 (Il1rl1) induces type 2 cytokine production in ILC2s¹⁶. Both Gata3 expression and Il1rl1 expression were detected in most clusters. Interestingly, the expression levels of these genes displayed unique patterns (Fig. 3d), indicating heterogeneous characteristics for ILC2s derived from different organs and treatment conditions. To characterize the unique features of each cluster, unbiased marker selection was performed (Fig. 3e), which identified GATA3-regulating genes (Areg, Il1rl1 and Gzma). Therefore, we focused on the GATA3 downstream signaling pathway to identify whether effector molecules were specialized in each cluster. We identified significant differences in representative cytokines that characterized ILC2 function: high Il13 expression in clusters 1, 2, 4, and 7; high Il5 expression in clusters 7 and 9; high Areg expression in clusters 3, 7, and 8; and high Arg1 expression in clusters 2, 6, and 9 (Fig. 2d). We confirmed that ILC2s expressed Cd3g or Tcrg-C1 at very low levels compared with those in other cells, such as gamma delta T cells (Supplementary Fig. 3a). We also confirmed that the expression of TCRg in ILC2s was undetectable by flow cytometry analysis (Supplementary Fig. 3b).

**Fig. 2: scRNA-seq reveals that liver group 2 innate lymphoid cells (ILC2) highly express *Il13*, which may contribute to the blood glucose-lowering effect of IL-33.**

**Fig. 3: scRNA-seq reveals that liver group 2 innate lymphoid cells (ILC2) highly express Il13, which may contribute to the blood glucose-lowering effect of IL-33.**

Furthermore, we compared the overall gene expression levels across different treatment conditions in the liver and between liver and lung ILC2s. In addition to the gene profiles that were altered by rIL-33 stimulation (Fig. 3f, left), liver ILC2s presented a distinct gene profile that differed from that of lung ILC2s (Fig. 3f, right). In the context of GATA-dependent genes, Il13 expression was significantly upregulated in liver ILC2s compared with lung ILC2s (Fig. 3f, right), but this pattern was not observed for other genes downstream of GATA3, such as Il1rl1, Il5, and Areg (Fig. 3g), and Gata3 expression did not correlate with Il13 expression (Fig. 2e). We confirmed that IL-13 expression was significantly increased in liver ILC2s compared to lung ILC2s using flow cytometry (Fig. 3h, i). These data suggested that the regulation of GATA3-dependent cytokines differed under different conditions and across distinct organs even though GATA3 was a key transcription factor in ILC2s. In particular, the expression of IL-13 was more strongly induced in liver ILC2s than in lung ILC2s in response to IL-33.

Liver ILC2-derived IL-13 is involved in the blood glucose-lowering effect of IL-33

The IL-13 and STAT3 signaling pathway has been reported to directly suppress hepatic glucose production¹⁷; we therefore hypothesized that ILC2-derived IL-13 mediates a suppressive effect on gluconeogenesis. First, we investigated whether IL-13 limited gluconeogenesis in wt mice. The PTT after injection of either PBS or 1 µg of recombinant IL-13 (rIL-13) confirmed that rIL-13 injection significantly suppressed blood glucose elevation in response to pyruvate (Fig. 4a). As shown in Fig. 1c, NSG mice that lacked ILC2s did not exhibit an IL-33-induced glucose-lowering effect. To determine the role of IL-13 in this process, we measured IL-13 levels in specific immune cell-deficient models (nude and NSG mice) after rIL-33 treatment. The rIL-33 injection increased IL-13 levels in both liver tissues and blood serum of wt and nude mice. In contrast, IL-13 levels remained unchanged in NSG mice (Fig. 4b, c). Next, we analyzed glucose metabolism following IL-33 stimulation in Il13-knockout mice. Although rIL-33 injection decreased fasting blood glucose levels and suppressed gluconeogenesis in Il13^+/+ or Il13^+/− mice, Il13^−/− mice did not exhibit either glucose-lowering effects or gluconeogenesis suppression following rIL-33 injection (Fig. 4d, e). Finally, we transferred liver ILC2s derived from Il13^+/+, Il13^+/- or Il13^-/- mice into ILC2-deficient NSG mice to confirm the contributions of the IL-33-IL-13 axis to the effects observed following rIL-33 injection (Fig. 4f). Transferring Il13^+/- ILC2s into NSG mice decreased fasting blood glucose levels in response to rIL-33 injection, but Il13^-/- ILC2s did not demonstrate significant glucose-lowering effects (Fig. 4g). Collectively, these data suggested that ILC2-derived IL-13 mediated the blood glucose-lowering effects of rIL-33. In particular, the glucose-lowering effects mediated by ILC2s were able to influence gluconeogenesis in hepatocytes. We also examined whether ILC2s played a similar role in a 3-month HFD-fed model. In 3-month HFD-fed mice, the numbers of IL-13(+) ILC2s clearly expanded in liver tissue (Supplementary Fig. 2e, f). Il13 expression in the whole liver was also increased, as confirmed by real-time quantitative PCR (RT‒qPCR) (Supplementary Fig. 2g). In addition, neutralization of IL-13 significantly suppressed blood glucose elevation in the PTT in mice fed a HFD for 3 months (Supplementary Fig. 2h). Collectively, these data suggested that IL-13 suppressed gluconeogenesis in the HFD-fed obesity model.

**Fig. 4: Liver group 2 innate lymphoid cell (ILC2)-derived IL-13 directly suppresses gluconeogenesis in hepatocytes.**

To examine the direct effect of hepatic ILC2s on the gluconeogenesis of hepatocytes, we cocultured hepatic ILC2s and primary hepatocytes (Fig. 4h). To confirm the effect of ILC2-derived IL-13, we also tested the effect of additional IL-13 neutralization. As expected, several gluconeogenic enzymes (G6pc, Pck1, Hnf4a, Fbp1, Pcx and Ppargc1a) were downregulated by coculture with ILC2s, and most of the effect was abolished by IL-13 neutralization (Fig. 4i). These data support the concept that ILC2s directly suppress gluconeogenesis in hepatocytes via IL-13. Furthermore, to evaluate the effect of ILC2s on metabolites in hepatocytes, we next cocultured ILC2s and hepatocytes and then assessed the metabolome in hepatocytes (Fig. 4h). Coculture of ILC2s significantly altered the amino acids in hepatocytes, especially those in the pyruvic acid pathway: Ala, Cys and Gly (Fig. 4j). These three amino acids are classified as glucogenic amino acids, which can be substrates for gluconeogenesis. Additionally, we examined the metabolome in the gluconeogenesis-related pathway (Fig. 4k). Coculture with ILC2s significantly increased fructose-1,6-bisphosphate (F1,6P) and pyruvic acid levels and decreased 3-phosphoglyceric acid (3-PG), 2-phosphoglyceric acid (2-PG), and phosphoenolpyruvic acid (PEP) levels. The increases in F1,6P and pyruvic acid levels may have been related to decreased expression of Fbp1 and Pcx (Fig. 4i, k). In addition, the decrease in PEP might have been partly attributable to decreased expression of Pck1. With regard to the tricarboxylic acid (TCA) cycle, intermediate metabolites tended to accumulate after coculture with ILC2s, and citric acid and malic acid levels were especially significantly increased (Fig. 4l). These results suggested that liver ILC2-derived IL-13 directly suppressed the gene expression of gluconeogenic enzymes and significantly altered the metabolite content in hepatocytes.

To further investigate the effects of IL-13 on hepatocytes, we also performed RNA-seq analyses on primary hepatocytes treated with rIL-13. Primary hepatocytes were incubated with rIL-13 for 2 h and then stimulated with cAMP/glucagon because glucagon promotes gluconeogenesis through the cAMP signaling pathway (Supplementary Fig. 4a). To identify gluconeogenesis-related genes affected by rIL-13, DEGs were measured after cAMP stimulation and rIL-13 treatment. Pathway analysis showed that glucose metabolism was affected by the presence of rIL-13 (Supplementary Fig. 4b). In particular, forkhead box O (FOXO) signaling is known to induce liver gluconeogenesis by activating genes such as G6pc and Pck1 and was mostly downregulated by rIL-13 treatment in the presence of cAMP or glucagon (Supplementary Fig. 4c). In addition, we confirmed the gene expression of G6pc and Pck1 by RT–qPCR analyses (Supplementary Fig. 4d). Similar to the RNA-seq results, the RT–qPCR analysis results showed that these genes were upregulated by cAMP or glucagon treatment and suppressed by rIL-13 treatment, indicating that IL-13 directly suppressed gluconeogenesis in hepatocytes. Thus, these results suggested that liver ILC2-derived IL-13 inhibited gluconeogenesis in hepatocytes.

The gluconeogenesis pathway is suppressed in certain hepatocyte clusters via the IL-33-IL-13 axis in vivo, as indicated by scRNA-seq with cell–cell interaction analysis

To clarify the underlying mechanism of action through which the IL-33-IL-13 axis affected hepatocytes, we performed scRNA-seq analyses of primary hepatocytes from fasted mice treated with rIL-33 or PBS for 5 days (Fig. 5a). A total of 11,327 cells were individually analyzed, and the cell types were classified by markers (Fig. 5b). A total of 10,186 cells were defined as hepatocytes (e.g., expressing Alb, Pck1, G6pc, and Hnf4a), which were categorized into five groups based on their heterogeneity (Fig. 5b). Cells from both the PBS-treated and rIL-33-treated groups were distributed in all five clusters, but the frequencies were somewhat altered by IL-33 stimulation (Supplementary Fig. 5a, b). The graph of the cell cycle phase also shows that there were some differences among the hepatocyte clusters (Supplementary Fig. 5c). Next, to explore the traits of each hepatocyte cluster, we observed the representative zonation markers and DEGs in each cluster. Glul is highly expressed near the central vein (CV), and Cdh1 is highly expressed in the portal node (PN)¹⁸. We found that Glul expression was higher in the Hepatocyte1 and Hepatocyte2 clusters than in the other clusters (Fig. 5c left panel). In contrast, Cdh1 expression was higher in the Hepatocyte4 and Hepatocyte5 clusters. These data suggest that the hepatocyte cluster reflects the location in which the hepatocytes reside. DEG analysis for each hepatocyte cluster revealed high expression of gluconeogenic enzymes, including G6pc and Pck1, especially in the Hepatocyte4 and Hepatocyte5 clusters (Fig. 5c right panel, Supplementary Fig. 5d). In addition, Gene Ontology (GO) analysis of the DEGs in each cluster suggested that Hepatocyte4 genes played roles in metabolic processes (Supplementary Fig. 5e).

**Fig. 5: IL-33 suppresses gluconeogenesis in hepatocytes, especially in the cluster that most significantly interacts with liver group 2 innate lymphoid cells (ILC2).**

We next investigated which hepatocyte cluster interacts with ILC2s. To answer this question, we combined hepatocytes and ILC2s, as shown in Fig. 3b, and performed interaction analysis. The liver ILC2-enriched clusters (1, 2, 4, 6, and 8) displayed increased interactions with the Hepatocyte4 cluster, followed by the Hepatocyte5 cluster (Fig. 5d and Supplementary Fig. 5f). Moreover, the interaction between IL-13 and IL-13R was strongly observed in the Hepatocyte4 cluster (Fig. 5e). We also performed immunofluorescence and found that some ILC2s existed near Pck1-high hepatocytes (Fig. 5f). To predict the regulatory factors affected by IL-33 stimulation, we performed an upstream gene analysis of all identified DEGs between PBS-treated and rIL-33-treated hepatocytes. The most enriched regulator was Hnf4a in this analysis, and the expression of genes downstream of Hnf4a was significantly decreased in the IL-33-treated cells (Fig. 5g, h). Hnf4a is known to play roles in glucose metabolic processes, including gluconeogenesis, in the liver. In fact, several reports, including one on the results of an HNF4A chromatin IP (ChIP) assay, have shown that HNF4A binds to the promoter regions of G6pc and Pck1, leading to enhanced expression¹⁹. The enrichment of G6pc and Pck1 expression was confirmed by uniform manifold approximation and projection (UMAP, Fig. 5i upper panel), and rIL-33 injection significantly downregulated G6pc expression, especially in the Hepatocyte4 cluster (Fig. 5i lower panel). In support of this finding, analysis of the overall gene distribution revealed that rIL-33 treatment significantly downregulated G6pc and Pck1 levels (Fig. 5j).

A previous report has suggested that the IL-13-STAT3 pathway promotes the downregulation of gluconeogenic genes. IL-13 binds to the IL-13 receptor α1 subunit (encoded by Il13ra1) in a receptor complex expressed on hepatocytes that also contains IL-4Rα (encoded by Il4ra). We confirmed that the population of clusters that interacted with ILC2s (Hepatocyte4 and Hepatocyte5) highly expressed genes associated with the IL-13-STAT3 pathway, including Il3ra1, Il4ra, Stat3, and Stat6 (Fig. 5k). We also confirmed direct binding of STAT3 upstream of the G6pc and Pck1 transcription start sites (TSSs) using a ChIP assay method reported by Lee et al. (Fig. 5l). To confirm that gluconeogenic enzymes were suppressed by IL-13/STAT signaling, we performed a knockdown experiment. In the presence of IL-13, suppression of G6pc or Pck1 expression on hepatocytes was significantly blocked by knockdown of Il13ra1, Il4ra, and Stat3 but not by knockdown of Stat6 (Fig. 5m). These results suggest that IL-13 directly suppresses gluconeogenic enzymes on hepatocytes via IL-13 receptor/Stat3 signaling. We also found that Hnf4a signaling was a common enriched pathway in both DEGs between hepatocyte clusters and DEGs between the PBS- and IL-33-treated groups (Fig. 6a). In addition, Hnf4a expression itself was elevated in Hepatocyte4 cells but was significantly suppressed by IL-33 treatment (Fig. 6b). Therefore, we hypothesized that the downregulation of Hnf4a might be mediated by Stat3 activation. To confirm this possibility, we searched the STAT3 binding motif using the JASPAR database and found a STAT3 binding motif approximately 1.5 kb upstream from the Hnf4a TSS (Fig. 6c). In addition, we found that rIL-13 treatment promoted STAT3 binding upstream of Hnf4a in primary hepatocytes (Fig. 6d). As the hepatocyte isolation method includes a digestion process, we verified that at least most hepatocytes in the UMAP analysis were viable cells. Notably, cell death markers (Susd6, Fas, Bax, Bcl2) were rarely expressed on hepatocytes (Fig. 6e). In addition, the GO enrichment list for DEGs in Hepatocyte4 scarcely included terms related to cell death or stress; most GOs were related to metabolic processes (e.g., steroid metabolic process, monocarboxylic acid metabolic process) (Fig. 6f).

**Fig. 6: IL-13 suppresses the expression of gluconeogenic enzymes via Stat3 signaling.**

The Gata3-mediated complex involves AP-1 family members to achieve transcriptional regulation in ILC2s

GATA3 serves as a central mediator of cytokine expression associated with type 2 immunity in ILC2s and T helper (Th)2 cells. However, Il13 expression levels in ILC2s were not uniformly altered in response to IL-33 stimulation despite GATA3 being activated to a similar extent (Figs. 2d and 3d). This finding indicated that the transregulation of Il13 did not simply depend on the presence of GATA3. We previously showed that chromodomain helicase DNA-binding protein 4 (Chd4) plays a role in Il13 induction in Th2 cells as a component of the GATA3 functional complex^20,21. Therefore, we focused on GATA3-interacting proteins that are functionally specific in ILC2s. To identify potential GATA3-interacting proteins under conditions of IL-33 stimulation in ILC2s, we used GATA3 IP and subsequent LC–MS/MS (IP-LC–MS/MS) analysis^20,21 of liver and lung ILC2s (Fig. 7a, upper panel). In total, 459 and 352 GATA3-binding proteins were detected in liver and lung ILC2s, respectively, which were then categorized into nuclear proteins and epigenetic factors (Fig. 7a, lower panel). We selected 26 and 20 proteins from liver and lung ILC2s as transcriptional regulators and GATA3 complex components, respectively (Fig. 7b). The top-ranked proteins were similar between liver ILC2s and lung ILC2s, especially members of the activating protein 1 (AP-1) family of transcription factors, such as JunB and JunD.

**Fig. 7: AP-1 family members bind to GATA3 in liver and lung group 2 innate lymphoid cells (ILC2) and suppress GATA3 function to induce *Il13* expression.**

Next, we performed GATA3 ChIP sequencing (ChIP-seq) and assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq) to determine whether the active transcriptional targets of GATA3 were altered in liver ILC2s and lung ILC2s. IL-33-stimulated open chromatin loci were defined by comparison between PBS- and rIL-33-treated cells in ATAC-seq. Consequently, we identified 7,031 differentially accessible regions (DAR), 1371 of which were bound by GATA3 in rIL-33-treated ILC2s in the liver (Fig. 7c and Supplementary Fig. 6c–e). Indeed, the upstream region of Il13 became more open in rIL-33-treated ILC2s than in PBS-treated ILC2s (Fig. 7d). To predict the interacting partners of the GATA3 transcriptional complex, we searched for enriched enhancer motifs among the active GATA3-binding targets. A motif analysis revealed that AP-1, Etv2, and Runx1 binding motifs were enriched at these GATA3-bound DARs in liver ILC2s (Fig. 7c, right). GATA3 actively bound the Il-13 genomic locus under IL-33 stimulation conditions in liver ILC2s, and the Il13 enhancer locus included the motif for AP-1 (Fig. 7d). In particular, the JunB (also known as AP-1)-binding motif was significantly enriched in both cell types, suggesting that AP-1 family transcription factors, including JunB and JunD, bound GATA3 in ILC2s (Supplementary Fig. 6c). Based on this possibility, we searched for JunB binding motifs using the JASPAR database and found the motif in GATA3-binding peaks located 40 bases upstream (motif1) and approximately 2.9 kb upstream (motif2) from the Il13 TSS (Fig. 7d). ChIP analysis confirmed that JunB bound to this region (Fig. 7e). Furthermore, to confirm the direct binding of GATA3-JunB on ILC2s, we performed IP western blotting. The ILC2 cell line ILC2/b6 was infected with a retroviral vector overexpressing FLAG-GATA3 or JunB. FLAG-IP and subsequent JunB western blotting clearly showed the binding of GATA3-JunB (Fig. 7f and Supplementary Fig. 7).

To further investigate the functional roles of GATA3-interacting molecules, we knocked down potential regulators using small interfering RNA (siRNA). The target molecules were selected from the liver ILC2 data according to the Mascot score derived from the GATA3-IP-LC–MS/MS results and the expression levels obtained from scRNA-seq (Fig. 7g). As Runx1 and Runx3 are known to function as crucial transcriptional regulators of IL-13 on ILC2s, we used siRunx1 and siRunx3 as experimental controls along with siGata3 (Fig. 7h). Sorted liver ILC2s were cultured with rIL-33 supplementation for 4 days, and siRNA oligonucleotides (targeting Gata3, Junb, Jund, Pa2g4, Calr, Irf8, Runx1, and Runx3) were transfected using electrophoresis. The cells were harvested 24 h later, and Il13 mRNA levels were analyzed by RT–qPCR (Fig. 7i). As expected, Il13 mRNA levels were suppressed by the knockdown of Gata3, Runx1, or Runx3. On the other hand, siRNA targeting Junb significantly upregulated Il13 mRNA levels, suggesting that AP-1 family members, including Junb, had a suppressive effect on Il13 transcription. We also performed RNA-seq analyses using the same sample groups analyzed by RT‒qPCR (Fig. 7h). The data were classified based on the responses to siGata3, siJunb, and siRunx1, so we focused on the cluster that was downregulated by siGata3 and upregulated by siJunb (Fig. 7j). Taken together, the data suggested that AP-1 family members were involved in the GATA3 complex that formed in both liver and lung ILC2s. Moreover, AP-1 family members may have played suppressive roles in GATA3-dependent cytokine production in ILC2s.

Different expression levels of GATA3-interacting AP-1 family members differentially activate ILC2s in an organ-specific manner

Il13 induction in liver ILC2s was higher than that in lung ILC2s under IL-33 stimulation conditions; therefore, we investigated whether the suppressive effects against GATA3 activity mediated by AP-1 family members were associated with differential levels of Il13 expression. First, we measured the expression levels of AP-1 family members in ILC2s using scRNA-seq data. Most of the genes in the AP-1 family were suppressed by IL-33 stimulation, and their expression levels in liver ILC2s were significantly lower than those in lung ILC2s (Fig. 8a). It has been reported that AP-1 family genes are rapidly upregulated with digestion stress; however, there was a clear difference between the control and IL-33-treated liver ILC2s even though the samples were digested with the same protocol and under the same conditions²². To confirm the function of the AP-1 family directly, we knocked down other AP-1 family members (Jun, Fos, Fosb, and Fosl2) and performed RT‒qPCR for Il13. The results suggested that Fos and Fosb suppress Il13 expression, while Fosl2 increases it, on liver ILC2s (Fig. 8b). We also tested the effect of JunB overexpression on ILC2s and found that JunB overexpression suppressed Il13 expression in both primary liver ILC2s and the cell line ILC2s (Fig. 8c).

**Fig. 8: Trajectory analysis reveals that AP-1 family member expression negatively correlates with *Il13* expression, especially in liver group 2 innate lymphoid cells (ILC2s).**

Pseudotime analysis was used to clarify whether changes in AP-1 family gene expression correlated with changes in Il13 expression in liver ILC2s. The cells obtained from PBS-treated ILC2s were defined as the roots, and the response of the cells to IL-33 stimulation was analyzed. To characterize the end cluster containing activated liver ILC2s (cluster 3), we grouped similarly regulated genes into modules and then calculated the correlation between gene modules and clusters. This analysis showed that the end cluster containing activated ILC2s (cluster 3) was positively correlated with one specific module (Liver Module 6, Fig. 8d). Interestingly, the top GO term of the Liver Module 6 was TNF signaling, which was associated with Il13 and AP-1 family member-encoding genes. In contrast, data obtained from lung ILC2s showed that the AP-1 genes were not contained in the module (Lung Module 7) that was highly correlated with activated ILC2s (Fig. 8e). In Fig. 8f, the matrix shows the correlations of average gene expression in each cluster, revealing significant negative correlations between the expression of cytokines downstream of GATA3, such as Il13, and AP-1 gene expression. However, these correlations were weaker in lung ILC2s than in liver ILC2s (Fig. 8f). In detail, the course of Il13 gene activation from the pseudotime analysis of liver ILC2s was negatively associated with AP-1 gene activation, especially Fos and Fosb activation. However, this relationship was not observed in lung ILC2s (Fig. 8g). Taken together, the findings indicated that the expression of AP-1 genes was suppressed during the course of ILC2 activation in the liver compared with the lungs and that the elevated expression of Il13 observed in liver ILC2s could be explained, at least in part, by downregulation of AP-1 genes that reduced the AP-1 family-mediated suppression of GATA3 transcriptional activity (Fig. 8h).

In summary, our observations by physiological and comprehensive transcriptomic approaches combined with LC–MS/MS analysis reveal the additional function of liver ILC2s and the role of the AP-1 family in GATA3 signaling in metabolism.

Discussion

Innate lymphoid cells are known to play roles in protecting organs from infectious diseases and exacerbating allergies. For example, ILC2s in the lungs are involved in responses to viral infection starting in early life and are mediated by IL-33 stimulation, and a stronger cytokine induction in ILC2s corresponds with exacerbation of asthma severity⁵. In the gut, ILC2s are major sources of IL-10, which protects against parasitic helminths⁶. However, a limited number of studies have examined the characteristics and roles of ILC2s in the liver, especially their involvement in glucose metabolism. In this study, we found that IL-33 stimulation suppressed gluconeogenesis in the liver through ILC2-derived IL-13 expression. scRNA-seq revealed that Il13 expression was more strongly induced by IL-33 stimulation in liver ILC2s than in lung ILC2s. Finally, we found that differences in the expression patterns of Gata3-binding partners resulted in differential cytokine production in the lungs, even though GATA3 is a key transcription factor for ILC2s both in the liver and in the lungs.

Insulin resistance in the liver results in excessive gluconeogenesis, and gluconeogenesis dysregulation is associated with worsened regulation of glucose metabolism, especially under obese conditions. We demonstrated that IL-13 derived from ILC2s suppressed gluconeogenesis by downregulating the expression of enzymatic genes, including G6pc and Pck1, in response to rIL-33 administration. Inoue et al. reported that Kupffer cell-derived IL-6 suppresses gluconeogenic enzymes in hepatocytes through the STAT3 signaling pathway²³. In addition, a recent study reported that IL-13 directly suppresses gluconeogenic enzyme activity in hepatocytes via STAT3 signaling¹⁷. Consistent with these reports, our scRNA-seq data indicated that Stat3 was highly expressed in hepatocytes, demonstrating a significant interaction with ILC2s and correlating with the downregulation of G6pc and Pck1 and their upstream transcription factor gene Hnf4a. Our findings suggested that ILC2-derived IL-13 improves glucose metabolism via cellular interactions in the liver.

Numerous studies have been performed on heterogeneous populations of immune cells, especially macrophages, and the results have suggested that these populations have tissue- and niche-specific functions²⁴. Transcriptional profiling of ILCs across human and mouse tissues has revealed unique features that are dependent on the tissue microenvironment^{25,26,27,28,29,30}. We elucidated the features specific to liver ILC2s via comparison with lung ILC2s. IL-13 is a prevalent cytokine found in ILC2s; however, our data showed that Il13 transcription in response to IL-33 stimulation was significantly enhanced in the liver compared to the lungs. This finding supports the existence of interorgan heterogeneity among ILC2s, and the transcription complex analysis indicated that the effector cytokines in ILC2s may be regulated in an organ-specific manner.

The transcription factor GATA3 is a well-known regulator of cytokine production, and we found that the expression patterns of cytokines downstream differed in ILC2s from the liver compared with ILC2s from the lungs. This diversity of downstream molecule expression was not simply associated with concomitant differences in Gata3 expression; therefore, we examined the expression patterns of GATA3-binding partners. Our omics data, including GATA3-IP-LC–MS/MS data, suggested that GATA3 and JunB interacted in liver ILC2s. In the context of cancer immunity, tumor ILC2s express Junb more highly than lymph node ILC2s³¹. During helminth infections, another AP-1 family molecule, basic leucine zipper ATF-like transcription factor (BATF), induces IL-4 and IL-13 in ILC2s³². Our scRNA-seq data showed that AP-1 family members (Junb, Fos, and Fosb) were expressed at lower levels in rIL-33-treated liver ILC2s than in rIL-33-treated lung ILC2s, suggesting that AP-1 family members suppress Il13 expression by associating with the promoter or enhancer regions of Il13 gene loci via interaction with GATA3 in ILC2s. Although AP-1 family members are known as activators, several reports indicated that they can have suppressive functions depending on the context^33,34,35. Our functional assays also suggested that some AP-1 family members can behave as suppressors in liver ILC2s in some contexts (Figs. 7i and 8b, c).

Our findings demonstrate that IL-33 stimulation in vivo drastically modulates metabolic signaling in hepatocytes through liver ILC2-derived IL-13 expression, providing insights into additional therapeutic targets for managing excessive gluconeogenesis, especially under obese diabetic conditions. We believe the mechanism uncovered in our study could introduce new avenues for the treatment of obese patients with diabetes.

Methods

Our research complied with relevant ethical regulations and the research protocols were approved by the ethics committee for animals at Chiba University.

Mice

BALB/c, C57BL/6, and nude mice were obtained from Japan SLC. NSG mice (#: 005557) with the NOD-SCID genetic background and carrying a null allele of the Il2rg gene were obtained from Charles River Laboratories (Japan). Il13^-/- BALB/c mice were provided by The Medical Research Council (Cambridge, UK)³⁶. A 3-month HFD model was achieved by feeding mice a dietary chow consisting of 60% kcal fat (Research Diets D12492) beginning at the age of 8 weeks and extending for a period of 5 months. All mice were allowed ad libitum access to food and water and were maintained in specific pathogen-free conditions in a 22 °C temperature-controlled room with a 12 h light-12 h dark cycle. Humidity was maintained in the range of 40–70%. All mice were male. Unless otherwise noted, all mice were BALB/c and used for experiments at 6–8 weeks of age. The research protocols were approved by the ethics committee for animals at Chiba University (registration number: 3-161, 1-68).

In vivo cytokine treatment

Mice were administered 200 µL of sterile PBS as a control or 0.5 µg of carrier-free murine rIL-33 (BioLegend, 580506) in 200 µL of sterile PBS via daily intraperitoneal injection for 5 days, unless otherwise noted. Each mouse was used for blood glucose measurements and euthanized the day after the final injection. An independent series of mice were administered 1 µg of carrier-free murine rIL-13 (BioLegend, 575906) in a single injection and then subjected to PTT analysis the next day. Fifty micrograms of control IgG (BioLegend, 400192) or anti-mouse IL-13 neutralizing antibody (InvivoGen, mabg-mil13-5) was administered via daily intraperitoneal injection for 4 days.

Tolerance tests and evaluation of blood glucose and insulin levels

Before each tolerance test, the mice were fasted for 16 h with free access to drinking water. Then, 2 g/kg body weight pyruvate was administered for the PTT³⁷, 2 g/kg body weight glycerol was administered for the GlycerolTT³⁸, and 0.075–0.1 U/kg body weight insulin was administered for the ITT³⁷. In each test, fasting blood glucose levels of mice were measured, after which pyruvate, glycerol, and insulin were diluted in PBS and injected at the appropriate volume i.p. One series of loading tests was performed on day 1. Mice were subjected to PTT, and then the GlycerolTT or ITT was performed 1 or 3 days after PTT, with daily rIL-33 injections. Changes in blood glucose levels were measured during 0−60 min (PTT, GlycerolTT) or 0−240 min (ITT). All reagents utilized for these tests were purchased from Sigma–Aldrich.

Insulin measurements

Blood samples from mice (wt BALB/c, nude, and NSG mice) fasted for 16 h were collected from the retro-orbital vein. Insulin concentrations were measured using an ultrasensitive mouse insulin enzyme-linked immunosorbent assay (ELISA) (Morinaga, M1108).

IL-13 ELISA

Serum IL-13 levels were quantified using an ELISA kit (R&D Systems, DY413-05) according to the manufacturer’s instructions.

Glycogen content assay

The glycogen content in liver tissues was quantified using a glycogen content assay kit (Abcam) according to the manufacturer’s instructions. Liver tissue (<100 mg) was washed with cold PBS and ddH₂O, homogenized, and boiled for 10 min to inactivate enzymes. After centrifugation for 10 min at 4 °C at 18,000 × g, the supernatant was collected and treated with reagents, the optical density was measured at 570 nm, and the glycogen content was calculated.

Immunofluorescence staining

Liver samples were harvested and fixed in 4% paraformaldehyde (PFA) overnight at 4 °C. The cryosections of tissues were permeabilized in PBS with 0.1% Triton X-100 and blocked for 30 min at room temperature (RT) in 5% normal goat serum. After eliminating the background signal with Endogenous Biotin-Blocking Kit (Invitrogen, E21390), immunofluorescence staining was performed with the following antibodies: anti-KLRG1 (1:25, FITC conjugated, Biolegend, 138409), rabbit anti-PCK1 (1:50, Abcam, ab70358) and anti-CD3e (1:25, biotin conjugated, eBioscience, 13-0033-82). The secondary antibodies were an Alexa Fluor 647 goat anti-rabbit antibody (1:50, Molecular Probes, A-21244) and Alexa Fluor 594 streptavidin (1:50, Jackson ImmunoResearch, 016-580-084). Primary antibody staining was performed at 4 °C overnight, and secondary antibody staining was performed at RT for 2 h. DAPI was used for nuclear staining. The immunofluorescence images were captured with Zeiss LSM 980 confocal microscope (microscope: Axio Observer7, objective lens: ×20 Plan Apochromat) and analyzed with Zeiss ZEN Blue software. Pseudocolored magenta was used for Alexa Fluor 647 dye, and orthogonal projection was performed to represent a 12 μm-section in two dimensions.

Primary hepatocyte isolation and culture

Primary hepatocytes were isolated from 8- to 10-week-old mice via a two-step collagenase digestion method³⁹. Specifically, mice were euthanized and sequentially perfused with ethylene glycol-bis(β-aminoethyl ether)-N,N,N,N-tetraacetic acid (EGTA) and collagenase (Gibco) starting from the inferior vena cava and draining through the portal vein. Single-cell suspensions were washed with Dulbecco’s modified Eagle’s medium (DMEM), and 1 × 10⁵/cm² cells were plated in William’s E medium (Thermo Fisher Scientific, 12551032) supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin.

Treatment of primary hepatocytes

Primary hepatocytes were incubated with or without 10 ng/mL rIL-13 for 2 h at 37 °C and then stimulated with 100 µM cAMP or 100 nM glucagon and with or without 10 ng/mL rIL-13 for 3 h at 37 °C. Cells were assigned to one of the following five groups: the control (vehicle), cAMP-treated (cAMP), cAMP- and rIL-13-treated (cAMP+rIL-13), glucagon-treated (Gluca), and glucagon- and rIL-13-treated (Gluca+rIL-13) groups.

Sampling of organ ILC2s and cell sorting

Male Il13^+/+, Il13^+/- and Il13^-/- BALB/c mice received daily injections of rIL-33 (0.5 µg, i.p.) for 5 days. The mice were euthanized and perfused with PBS, and then liver and lung ILC2s were harvested. Briefly, the mouse livers and lungs were digested using a gentle MACS Octo Dissociator with Heaters (Miltenyi Biotec) using a liver dissociation kit (Miltenyi, 130-105-807, program: 37C_m_LIDK_1) and a lung dissociation kit (Miltenyi, 130-095-927, m_lung_01, incubation at 37 °C for 30 min, and then m_lung_02 program), respectively. Debris was removed using the protocol published by Itoh, with some modifications⁴⁰. After filtration through a 70-µm cell strainer, the cells were collected by centrifugation at 500 × g for 5 min at 4 °C and resuspended in 40% Percoll Plus (GE Healthcare). Then, 75% Percoll was gently added below the 40% Percoll layer. After centrifugation at 700 × g for 20 min at RT, the cells at the interface were collected in PBS containing 1% FBS and 2 mM EDTA, and lineage-positive cells were depleted using MACS (Miltenyi) with an anti-mouse lineage panel (BioLegend, 133307) and streptavidin nanobeads (BioLegend, 480016). Then, single-cell suspensions were stained using blocking and staining antibodies in the dark at 4 °C. Liver Lin^-Th1.2⁺IL-7Rα⁺ST2⁺ ILC2s were sorted using a FACSAria cell sorter (BD Biosciences) and cultured or directly used for experiments.

ILC2 in vitro cultures

ILC2 culture was performed as before⁴¹. ILC2s isolated from each organ were cultured with complete medium (RPMI-1640 medium [Sigma–Aldrich], 10 mM HEPES [Life Technologies], 1 mM sodium pyruvate [Life Technologies], 1% MEM nonessential amino acids [Life Technologies], 0.1% 2-mercaptoethanol and L-glutamine) with 25 U/mL rIL-7Rα (BioLegend), 10 ng/mL rIL-25 (BioLegend), and 10 ng/mL rIL-33. The ILC2s were maintained at 37 °C with 5% CO₂ and passaged after 2−3 days into the same volume of medium containing cytokines. The ILC2 cell line ILC2/b6 was generously donated by Dr. Q Yang. The ILC2/b6 cells were cultured in OP9 medium (α-MEM, 20% FBS, 50 μM β-mercaptoethanol, and Pen-Step-Glutamine) supplemented with 10 ng/mL IL-2, IL-7, and IL-33 (PeproTech Inc.)⁴².

Coculture of primary hepatocytes and hepatic ILC2s

Primary hepatic ILC2s were cultured for several days. Then, the ILC2s were washed with PBS, and 2 × 10⁴/cm² ILC2s were plated on top of a monolayer of primary hepatocytes in the presence or absence of 5 µg/mL anti-mouse IL-13 neutralizing antibody (InvivoGen, mabg-mil13-5). After 6 h of culture, floating ILC2s were aspirated and washed with PBS, and hepatocytes were immediately subjected to RNA extraction or metabolomic analysis.

Metabolomic analysis

Primary hepatocytes were washed with 5% (w/w) mannitol and immersed in methanol, and extracts were subjected to ultrafiltration and frozen immediately. CE-TOFMS and CE-QqQMS (C-Scope) were performed to evaluate central energy metabolism in hepatocytes (Human Metabolome Technologies, Japan).

ILC2 transfer into NSG mice

Liver ILC2s were purified from wt BALB/c mice that were treated with rIL-33 for 5 consecutive days and then cultured for 3−5 days. After culturing, 5 × 10⁵ to 1 × 10⁶ liver ILC2s were suspended in 300 µL of PBS and then intravenously injected into NSG mice. Control NSG mice were intravenously injected with 300 µL of PBS. The control and recipient mice received rIL-33 injections every day, and fasting blood glucose was measured after 3 and 5 days.

Flow cytometric analysis

Cells were incubated with an anti-CD16/CD32 monoclonal antibody (mAb, eBioscience, 14-0161-82, 93, 1:33) and then stained with the following dye or anti-mouse antibodies: viability dye APC-Cy7 (eBioscience, 65-0865-14), anti-lineage cocktail-FITC (BioLegend, 133302, 145-2C11; RB6-8C5; RA3-6B2; Ter-119; M1/70, 1:40), anti-TCRβ-FITC (BioLegend, 109206, H57-597, 1:200), anti-CD90.2-PE/Cy7 (BioLegend, 105326, 30-H12, 1:200), anti-CD127-APC (BioLegend, 135012, A7R34, 1:100), anti-ST2-BV421 (BioLegend, 145309, DIH9, 1:100), anti-CD3E-FITC (BioLegend, 100203, 17A2, 1:200), anti-TCRg-PE (BioLegend, 118107, GL3, 1:100). Gating was performed as follows: ILC2s were defined as live Lin⁻ (CD3e⁻Ly6G⁻Ly6c⁻CD11b⁻CD45R⁻TER119⁻TCRb⁻) CD127⁺Thy1⁺ST2⁺ cells. Stained cells were examined with a FACSCanto II flow cytometer (BD Biosciences), and the results were analyzed using FlowJo software version 10.4.2 (TreeStar, FlowJo).

Intracellular cytokine staining

Freshly isolated lymphocytes from livers and lungs were restimulated with phorbol 12-myristate 13-acetate (50 ng/mL) and ionomycin (500 ng/mL) for 4 h in the presence of monensin (2 µM) and then stained for ILC2 surface markers. The cells were fixed with 4% PFA for 10 min at RT and permeabilized in permeabilization buffer (50 mM NaCl, 5 mM EDTA, 0.02% NaN₃, pH 7.5) containing 0.5% Triton X-100 for 10 min on ice. After blocking with 3% bovine serum albumin in PBS for 15 min, the cells were stained on ice for 30 min with anti-IL-13 (BioLegend, 159403, W17010B, 1:100). Flow cytometry was performed with a FACSCanto II (BD Biosciences) or FACSymphony A3 (BD Biosciences) flow cytometer, and the results were analyzed with FlowJo.

Real-time PCR

Real-time PCR was performed as previously described⁴³. Liver tissues were collected from mice soon after euthanasia and then homogenized using a TissueRuptor II system (Qiagen). RNA was extracted using an RNeasy mini kit (Qiagen, 74104), and its quantity and quality were evaluated with a NanoDrop spectrophotometer (Thermo Fisher Scientific). The RNA was then reverse-transcribed into cDNA using a ReverTra Ace qPCR RT kit (Toyobo), and the resulting cDNA was amplified by RT–PCR using Fast SYBR Green (Thermo Fisher Scientific, 4385614). Ct values were quantified using a StepOnePlus Real-Time PCR system (Applied Biosystems) with the default fast mode setting (stage 1; 95.0 °C for 20 s, stage 2; 40 cycles at 95.0 °C for 1 s, and 60.0 °C for 20 s). Relative gene expression was quantified using the 2^−ΔΔCT method and normalized to Rpl32 expression. The genes and primer pairs are listed in Supplementary Table 5.

siRNA or plasmid transfection

Cultured ILC2s were electroporated with a NEON transfection kit and device (Thermo Fisher Scientific) using three pulses of 1800 V for 10 ms to transfect select siRNAs or pCMV-myc-JunB vector. After 24 h of transfection, cells were collected and used for RT–qPCR or RNA-seq. Primary hepatocytes were transfected with RNAi using Lipofectamine RNAi Max (Thermo Fisher, 13778075) according to the manufacturer’s instructions. The siRNAs were as follows: Ct si (Negative Control #1, Cat# AM4635) and Target si (Cat# 4390771, 4390815). The IDs were as follows: Gata3 si: s61810, Junb si: s68567, Pa2g4 si: s201806, Calr si: s63270, Jund si: s201553, Runx1 si: s201126, Runx3 si: s63447, Il4ra si: s68281, Il13ra1 si: s68211, Stat3 si: s74452, Stat6 si: s74463. The pCMV-myc-JunB vectors used were described previously⁴⁴.

GATA3-IP and subsequent LC–MS/MS analysis

The protocol was performed as before^20,21,45,46. Liver ILC2s (1 × 10⁷) and lung ILC2s (1.8 × 10⁶) were solubilized with the following protease inhibitor-containing IP buffer: 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 0.1% Tween, 1 mM EDTA, 10 mM NaF, 1 mM DTT, and a protease inhibitor cocktail (Roche Applied Science). The cells were lysed on ice for 30 min with gentle shaking and sonicated on a Misonix S-4000 sonicator (Qsonica) for three cycles at an amplitude of 20 for 30 s followed by 30 s of rest. Protein complexes were subjected to IP with or without an anti-GATA3 antibody (Santa Cruz Biotechnology, sc-268; R&D Systems, MAB26051) and then separated by SDS–PAGE. The bands were excised from the gel, and the gel pieces were washed twice with 100 mM bicarbonate in acetonitrile. The proteins were then digested with trypsin for 8 h at 37 °C. After adding 0.1% formic acid to the supernatant, the peptides were analyzed by LC–MS/MS with an Advance UHPLC (Bruker) and an Orbitrap Velos Pro Mass Spectrometer (Thermo Fisher Scientific). The resulting LC–MS/MS dataset was analyzed using the Mascot software program (version 2.5.1, Matrix Science). The Mascot score indicated the probability that the observed match was a random event (a Mascot score > 100 means the absolute probability is <1e⁻¹⁰). Proteins with significant Mascot scores with a high signal/noise ratio (anti-GATA3/IgG control > 1) were considered specific GATA3-interacting proteins.

Retrovirus infection and subsequent FLAG-IP-immunoblotting

ILC2/b6 cells were infected with either a mock control (pMxs-IRES-hNGFR) or a Myc-Flag-GATA3-containing retrovirus^47,48. Five days after infection, Myc-Flag-tagged GATA3-infected hNGFR⁺ cells were solubilized with the following protease inhibitor-containing IP buffer: 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 0.1% Tween, 1 mM EDTA, 10 mM NaF, 1 mM DTT and a protease inhibitor cocktail (Roche Applied Science). The cells were lysed on ice for 30 min with gentle shaking and sonicated on a Misonix S-4000 sonicator (Qsonica) for 3 cycles at an amplitude of 20 for 30 s followed by 30 s of rest. The insoluble materials were removed by centrifugation, and IP with 50 µL anti-Flag M2 agarose (Sigma–Aldrich, A2220) was performed overnight at 4 °C. The immune complexes were subjected to immunoblotting with anti-JunB (Sant Cruz, sc8051, 1 µg/mL), anti-Flag (Sigma, M2, A2220, 1 µg/mL) and anti-Tubulinα (Sigma, T6119, 1 µg/mL). Image data were obtained using Fusion Solo S (Vilber).

GATA3-ChIP-seq and data analysis

ChIP-seq analyses were performed as reported previously⁴⁹. Sorted liver ILC2s (1.7 × 10⁶) and liver ILCs (0.5 × 10⁶) were fixed with 1% formaldehyde (Thermo Fisher Scientific, 28906) in MEM alpha (Thermo Fisher Scientific, 12561-056) for 10 min at RT. The reaction was quenched by the addition of a 1/10 volume of 0.125 M glycine, incubated for 5 min at RT, washed with PBS, and centrifuged at 500 × g. The pelleted nuclei were resuspended in lysis buffer (0.5% SDS, 10 mM EDTA, 0.5 mM EGTA, 50 mM Tris-HCl [pH 8], and protein inhibitor cocktail) and sonicated on a Bioruptor (Diagenode) for 18 cycles of 30 s of sonication followed by 30 s of rest at maximum power. Per 10⁷ cells, 5 µg of antibody was hybridized to Dynabeads M-280 sheep anti-rabbit, Dynabeads M-280 sheep anti-mouse, or Dynabeads Protein A/G (Invitrogen) and incubated with diluted chromatin complexes overnight at 4 °C. The beads were then washed five times with IP buffer and eluted for 6 h at 65 °C in ChIP elution buffer (20 mM Tris-HCl, pH 7.5, 5 mM EDTA 50 mM NaCl, 1% SDS, and 50 mg proteinase K). The precipitated chromatin fragments were cleaned using ChIP DNA Clean & Concentrator (Zymo Research) and subjected to qPCR on a 7900HT Fast Real-Time PCR System (Applied Biosystems) with SYBR GreenER qPCR SuperMix (Invitrogen). Mouse anti-Gata3 mAbs (a mixture of sc-268 from Santa Cruz Biotechnology and MAB26051 from R&D Systems) were used. ChIP-seq libraries were constructed using an NEBNext ChIP-Seq Library Preparation Kit (Illumina, San Diego, CA, USA; NEB #E6240) according to the manufacturer’s instructions. Deep sequencing was performed on the Illumina NextSeq 500 platform using a TruSeq Rapid SBS Kit (Illumina) in 60-base single-end mode according to the manufacturer’s protocol. The annotation of enhancers for the nearest genes was performed with GREAT (http://bejerano.stanford.edu/great/public/html/index.php). A heatmap of the ChIP-seq peaks was produced using HOMER for enrichment calculations, and IGV was used for visualization. Reads were mapped to the mouse genome assembly version from December 2011 (mm10) by using the Bowtie2 mapping tool with a default setting. Peaks were called using MACS2⁵⁰ software, and those identified in both replicates were used for further analyses (Fig. 7c). Data quantification was conducted using the “annotatePeaks.pl” routine in HOMER⁵¹. Motif analysis was performed using the “findMotifsGenome.pl” routine in HOMER.

ChIP-PCR

Cultured liver ILC2s (5 × 10⁶) were incubated with phorbol 12-myristate 13-acetate (50 ng/mL) and ionomycin (500 ng/mL) for 3 h and then crosslinked. ChIP was performed as in the protocol for ChIP-seq with some modifications. Cells were sonicated on Q700 (QSONICA) for 20 to 25 cycles of 15 s of sonication followed by 45 s of rest at an amplitude of 60 (total: 6000J). Mouse anti-JunB mAbs (a mixture of 10 µL [1:50] of sc-8051 from Santa Cruz Biotechnology and 10 µL [1:50] of C37F9 [#3753] from Cell Signaling Technology) were used. Primary hepatocytes were isolated from wt BALB/c mice and cultured with serum-free DMEM in the presence of 100 µM cAMP with or without 10 ng/mL rIL-13 for 3 h at 37 °C. Next, 1 × 10⁷ harvested hepatocytes were processed using a SimpleChIP kit (Cell Signaling Technology, #9002). Anti-STAT3 mAbs (a mixture of 5 µL [1:100] of 124H6 [#9139] and 10 µL [1:50] of D3Z2G [#12640] from Cell Signaling Technology) were used. RT–PCR was performed using eluted DNA.

ATAC-seq

ATAC-seq analyses of sorted liver and lung ILC2s were performed as before, with minor modifications⁵². A total of 5,000 − 50,000 cells were pelleted and washed with 50 µL of 1× PBS before treatment with 50 µL of 2× lysis buffer. The nuclei were resuspended in 40 µL of Tagment DNA enzyme buffer supplemented with 2.5 µL of Tn5 transposase (Illumina, 15027865) to tag and fragment the accessible chromatin. The mixture was incubated at 37 °C for 30 min. The DNA fragments were purified with a MinElute kit (Qiagen, 28004) and amplified by 11−13 PCR cycles based on the amplification curve. The libraries were sequenced for 60 cycles (single-end reads) on an Illumina HiSeq 1500 instrument. The reads were mapped to the mouse genome assembly version from December 2011 (GRCm38/mm10) by using the Bowtie2 mapping tool with a default setting. Peaks were called using MACS2 software with the broad option, and those identified in both replicates were used for further analyses (Fig. 7c). Data quantification was conducted using the “annotatePeaks.pl” routine in HOMER. DARs were detected using the “getDifferentialPeaksReplicates.pl” routine in HOMER.

RNA-seq and data analysis

Total RNA was purified from 30,000−100,000 liver ILC2s and 100,000 primary hepatocytes using an RNeasy Plus Micro Kit (Qiagen). The RNA concentration and RNA integrity were measured using an Agilent 2100 bioanalyzer (Agilent). Library construction and sequencing were performed as previously described⁵³. Briefly, libraries for RNA-seq were prepared using a TruSeq RNA Sample Prep Kit (Illumina, 20020594) in accordance with the manufacturer’s protocol. Sequencing was performed on the Illumina HiSeq 1500 or NextSeq 500 platform using a TruSeq Rapid SBS Kit (Illumina, FC-402-4002) in the 50-base single-end mode, according to the manufacturer’s protocol. The TopHat program (version 1.3.2 with default parameters) was used to align reads to a mouse reference genome (mm10). Then, gene expression values were calculated as reads per kilobase of exon unit per million mapped reads using the Cufflinks program (version 2.0.2). DEGs with a false-discovery rate of less than 0.05 were identified using an edgeR software package (version 3.28.1). A heatmap and dendrogram were also drawn using R software. Upstream gene (transcriptional regulatory relationships unraveled by Sentence-based Text-mining, TRRUST), GO, and Kyoto Encyclopedia of Genes and Genomes (KEGG) terms were used to classify the DEG functions using Metascape.

scRNA-seq library preparation and sequencing

Single-cell suspensions of immune cells from the livers or lungs of wt 8-week-old BALB/c mice were prepared by digestion, Percoll enrichment, and the sorting method as stated above (gating: Lin⁻ [CD3e⁻Ly6G⁻Ly6c⁻CD11b⁻CD45R⁻TER119⁻TCRb⁻] CD127⁺Thy1⁺ST2⁺). Single-cell hepatocytes were also prepared using the collagenase perfusion method described above. Liberase (Gibco) was substituted for collagenase to perform milder digestion. The scRNA-seq libraries were prepared according to 10× Genomics specifications (10× Chromium Next GEM single-cell 3ʹ reagent kits and the library construction kit). Briefly, cell suspensions were loaded onto the 10× Genomics Chromium Controller to generate gel beads in emulsion (GEM). After GEM generation, the samples were incubated at 53 °C for 45 min in a thermal cycler (Veriti, Applied Biosystems) to generate polyA cDNA barcoded at the 5ʹ end by the addition of a template switch oligo linked to a cell barcode and unique molecular identifiers (UMI). The GEMs were broken, and the single-stranded cDNA was cleaned up with DynaBeads MyOne Silane Beads (Thermo Fisher Scientific). The cDNA was amplified (98 °C for 3 min, 11 cycles of 98 °C for 15 s and 63 °C for 20 s, and 72 °C for 1 min), and cDNA quality was assessed using an Agilent TapeStation. The cDNA was enzymatically fragmented, end-repaired, A-tailed, subjected to double-sided size selection with SPRIselect beads (Beckman Coulter), and ligated to adaptors provided in the kit. A unique sample index for each library was introduced through 14 cycles of PCR amplification using the indices provided in the kit (98 °C for 45 s; 14 cycles of 98 °C for 20 s, 54 °C for 30 s, and 72 °C for 20 s; 72 °C for 1 min; and a hold at 4 °C). The indexed libraries were subjected to a second round of double-sided size selection, and the libraries were then quantified and quality-assessed with an Agilent TapeStation. The libraries were submitted to GENEWIZ and clustered using a NovaSeq 6000 on a paired-end read flow cell. Sequencing was performed on R1 (10× barcode and UMI), I7 index (8 cycles; sample index) and R2 (89 cycles; transcript). The 10× Genomics Cell Ranger Single Cell Software was used for sample demultiplexing, alignment to the mouse reference genome (mm10), filtering, UMI counting, single-cell 3ʹ-end gene counting, and quality control according to the manufacturer’s parameters.

scRNA-seq data analysis

The R package Seurat (v.3.1.2 and v4.0.4)⁵⁴ was used to cluster cells in a merged matrix. Several quality control steps were performed based on the plot pattern of the expression gene number and ratio of mitochondrial RNA in each sample. First, immune cells including doublets and dead cells were removed with flow cytometry, and those expressing fewer than 500 genes (low quality) were excluded from further analysis. Next, primary hepatocytes in which libraries were directly constructed without flow cytometry that expressed more than 3000 genes (potential doublets) and fewer than 200 genes or >50% mitochondrial genes (low quality) were first filtered out. The threshold for hepatocytes was determined by examining the validity of each mitochondrial cutoff (10–70%) because mitochondrial gene expression is constitutively active in hepatocytes due to the roles of mitochondria in energy metabolism. A strict cutoff (threshold of >10% mitochondria) eliminated enzymatically active cell populations, as reported by MacParland et al., who also used a threshold of >50% mitochondrial genes for human hepatocyte scRNA-seq⁵⁵. The gene counts for each cell were divided by the total gene counts for each cell and multiplied by a scale factor of 10,000, after which a natural log transformation was applied. The FindVariableFeatures function was used to select variable genes with default parameters. The ScaleData function was used to scale and center the counts in the dataset. We excluded several genes for clustering and listed them in Supplementary Table 7. To define the cell types of all cells, we performed principal component analysis on the variable genes, and 50 principal components were used for cell clustering (resolution = 5) and UMAP dimensional reduction⁵⁶. For further ILC2 subclustering, a resolution of 0.5 was used. The cluster markers were found using the FindAllMarkers function, and cell types were manually annotated based on the cluster markers. DEGs between two groups were defined as those with values of P < 0.05 as determined by the Welch t-test. Cluster markers defined by FindAllMarkers were used as DEGs of each cluster. These DEGs were used for functional analyses using the GO and KEGG pathway databases. The single-cell gene expression data were visualized with UMAP overlays, dot heatmaps, and violin plots. A volcano plot was constructed in R. We illustrated the cell cycle phase (G1, S, G2M) using the CellCycleScoring function based on cell cycle-related genes. To estimate the stage of maturation, a pseudotime trajectory analysis was performed using Monocle3⁵⁷. To calculate the correlations of clusters and gene sets, we grouped coregulated genes into modules using Louvain community analysis with Monocle3.

Cell–cell ligand–receptor interaction analysis

CellPhoneDB (v.2.0.0) was used to perform ligand–receptor analysis of cell populations in the liver, especially for assessment of the interactions between hepatocytes and ILC2s^58,59. The raw counts and cell-type annotation for each cell were inputted into CellPhoneDB to determine potential ligand–receptor pairs. Pairs with P > 0.05 were not included in further analyses. The numbers of interactions between each pair of clusters (hepatocyte clusters 1−5 and ILC2 clusters 0−9) were plotted using downsized cells (500 cells in total).

Analysis with g:GOSt

DEGs identified in siRNA-transfected ILC2s were included in a functional enrichment analysis query and then analyzed using g:GOSt functional profiling⁶⁰.

Statistics and reproducibility

The data were analyzed with GraphPad Prism 9 (GraphPad Software) and are expressed as the mean ± standard deviation (SD). Significant differences between two groups were determined by unpaired one-sided t-test. Comparisons of more than three groups were performed using one-way analysis of variance followed by Sidak’s or Tukey’s multiple comparisons test. Significance was set at P < 0.05. The experimental findings were reliably reproduced with at least two independent experiments in Fig. 1a–g, i, j; Fig. 3h, i; Fig. 4a–e, g, i; Fig. 5f, l, m; Fig. 6d; Fig. 7e, f, i; Fig. 8b, c; Supplementary Fig. 1a–d; Supplementary Fig. 2a–h; Supplementary Fig. 3b; Supplementary Fig. 4d. The micrograph of immunofluorescence in Fig. 5f is representative of 24 locations of 5 independent mice. Western blotting in Fig. 7f is representative figures of two independent experiments. We performed scRNA-seq of ILC2s using pooled tissues of 2–6 mice in each group (Fig. 3). ATAC-seq and ChIP-seq (Fig. 7c, d and Supplementary Fig. 6c–e) were performed with duplicates only except for GATA3-ChIP-seq from IL-33 treated lung (Supplementary Fig. 6c right, 6e). RNA-seq (Fig. 7j) of cultured ILC2s were performed with triplicates.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The raw and processed data generated in this study have been deposited in the Gene Expression Omnibus (GEO) database under accession code GSE189287. The source data are provided as a Source Data file. Source data are provided with this paper.

Code availability

The code will be made available upon reasonable request.

References

Williams, C. D. et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle-aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology 140, 124–131 (2011).
Article PubMed Google Scholar
Eguchi, Y. et al. Prevalence and associated metabolic factors of nonalcoholic fatty liver disease in the general population from 2009 to 2010 in Japan: a multicenter large retrospective study. J. Gastroenterol. 47, 586–595 (2012).
Article CAS PubMed Google Scholar
Friedman, S. L., Neuschwander-Tetri, B. A., Rinella, M. & Sanyal, A. J. Mechanisms of NAFLD development and therapeutic strategies. Nat. Med. 24, 908–922 (2018).
Article CAS PubMed PubMed Central Google Scholar
Moro, K. et al. Innate production of T(H)2 cytokines by adipose tissue-associated c-Kit(+)Sca-1(+) lymphoid cells. Nature 463, 540–544 (2010).
Article ADS CAS PubMed Google Scholar
Wallrapp, A. et al. The neuropeptide NMU amplifies ILC2-driven allergic lung inflammation. Nature 549, 351–356 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Moriyama, S. et al. beta2-adrenergic receptor-mediated negative regulation of group 2 innate lymphoid cell responses. Science 359, 1056–1061 (2018).
Article ADS CAS PubMed Google Scholar
Brestoff, J. R. et al. Group 2 innate lymphoid cells promote beiging of white adipose tissue and limit obesity. Nature 519, 242–246 (2015).
Article ADS CAS PubMed Google Scholar
Lee, M. W. et al. Activated type 2 innate lymphoid cells regulate beige fat biogenesis. Cell 160, 74–87 (2015).
Article CAS PubMed Google Scholar
Dalmas, E. et al. Interleukin-33-activated islet-resident innate lymphoid cells promote insulin secretion through myeloid cell retinoic acid production. Immunity 47, 928–942.e927 (2017).
Article CAS PubMed Google Scholar
Newland, S. A. et al. Type-2 innate lymphoid cells control the development of atherosclerosis in mice. Nat. Commun. 8, 15781 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Jeffery, H. C. et al. Human intrahepatic ILC2 are IL-13positive amphiregulinpositive and their frequency correlates with model of end stage liver disease score. PLoS ONE 12, e0188649 (2017).
Article PubMed PubMed Central CAS Google Scholar
Peters, A. L. et al. Single cell RNA sequencing reveals regional heterogeneity of hepatobiliary innate lymphoid cells in a tissue-enriched fashion. PLoS ONE 14, e0215481 (2019).
Article CAS PubMed PubMed Central Google Scholar
Gao, Y. et al. IL-33 treatment attenuated diet-induced hepatic steatosis but aggravated hepatic fibrosis. Oncotarget 7, 33649–33661 (2016).
Article PubMed PubMed Central Google Scholar
Frizzell, R. T. et al. Role of gluconeogenesis in sustaining glucose production during hypoglycemia caused by continuous insulin infusion in conscious dogs. Diabetes 37, 749–759 (1988).
Article CAS PubMed Google Scholar
Mjösberg, J. et al. The transcription factor GATA3 is essential for the function of human type 2 innate lymphoid cells. Immunity 37, 649–659 (2012).
Article PubMed CAS Google Scholar
Schmitz, J. et al. IL-33, an interleukin-1-like cytokine that signals via the IL-1 receptor-related protein ST2 and induces T helper type 2-associated cytokines. Immunity 23, 479–490 (2005).
Article CAS PubMed Google Scholar
Stanya, K. J. et al. Direct control of hepatic glucose production by interleukin-13 in mice. J. Clin. Invest. 123, 261–271 (2013).
Article CAS PubMed Google Scholar
Halpern, K. B. et al. Single-cell spatial reconstruction reveals global division of labour in the mammalian liver. Nature 542, 352–356 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Sharabi, K. et al. Selective chemical inhibition of PGC-1α gluconeogenic activity ameliorates type 2 diabetes. Cell 169, 148–160.e115 (2017).
Article CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Gata3/Ruvbl2 complex regulates T helper 2 cell proliferation via repression of Cdkn2c expression. Proc. Natl Acad. Sci. USA 110, 18626–18631 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Functionally distinct Gata3/Chd4 complexes coordinately establish T helper 2 (Th2) cell identity. Proc. Natl Acad. Sci. USA 110, 4691–4696 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
van den Brink, S. C. et al. Single-cell sequencing reveals dissociation-induced gene expression in tissue subpopulations. Nat. Methods 14, 935–936 (2017).
Article PubMed CAS Google Scholar
Inoue, H. et al. Role of hepatic STAT3 in brain-insulin action on hepatic glucose production. Cell Metab. 3, 267–275 (2006).
Article CAS PubMed Google Scholar
Davies, L. C., Jenkins, S. J., Allen, J. E. & Taylor, P. R. Tissue-resident macrophages. Nat. Immunol. 14, 986–995 (2013).
Article CAS PubMed PubMed Central Google Scholar
Robinette, M. L. et al. Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets. Nat. Immunol. 16, 306–317 (2015).
Article CAS PubMed PubMed Central Google Scholar
Crinier, A. et al. High-dimensional single-cell analysis identifies organ-specific signatures and conserved NK cell subsets in humans and mice. Immunity 49, 971–986.e975 (2018).
Article CAS PubMed PubMed Central Google Scholar
Ricardo-Gonzalez, R. R. et al. Tissue signals imprint ILC2 identity with anticipatory function. Nat. Immunol. 19, 1093–1099 (2018).
Article CAS PubMed PubMed Central Google Scholar
Xu, H. et al. Transcriptional atlas of intestinal immune cells reveals that neuropeptide α-CGRP modulates group 2 innate lymphoid cell responses. Immunity 51, 696–708.e699 (2019).
Article CAS PubMed PubMed Central Google Scholar
Yudanin, N. A. et al. Spatial and temporal mapping of human innate lymphoid cells reveals elements of tissue specificity. Immunity 50, 505–519.e504 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mazzurana, L. et al. Tissue-specific transcriptional imprinting and heterogeneity in human innate lymphoid cells revealed by full-length single-cell RNA-sequencing. Cell Res. 31, 554–568 (2021).
Article CAS PubMed PubMed Central Google Scholar
Moral, J. A. et al. ILC2s amplify PD-1 blockade by activating tissue-specific cancer immunity. Nature 579, 130–135 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Miller, M. M. et al. BATF acts as an essential regulator of IL-25-responsive migratory ILC2 cell fate and function. Sci. Immunol. 5, eaay3994 (2020).
Mittelstadt, M. L. & Patel, R. C. AP-1 mediated transcriptional repression of matrix metalloproteinase-9 by recruitment of histone deacetylase 1 in response to interferon β. PLoS ONE 7, e42152 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Shaulian, E. & Karin, M. AP-1 in cell proliferation and survival. Oncogene 20, 2390–2400 (2001).
Article CAS PubMed Google Scholar
Schüle, R. et al. Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 62, 1217–1226 (1990).
Article PubMed Google Scholar
McKenzie, G. J. et al. Impaired development of Th2 cells in IL-13-deficient mice. Immunity 9, 423–432 (1998).
Article CAS PubMed Google Scholar
Gray, L. R. et al. Hepatic mitochondrial pyruvate carrier 1 is required for efficient regulation of gluconeogenesis and whole-body glucose homeostasis. Cell Metab. 22, 669–681 (2015).
Article CAS PubMed PubMed Central Google Scholar
Maeda, N. et al. Adaptation to fasting by glycerol transport through aquaporin 7 in adipose tissue. Proc. Natl Acad. Sci. USA 101, 17801–17806 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Charni-Natan, M. & Goldstein, I. Protocol for primary mouse hepatocyte isolation. STAR Protoc. 1, 100086 (2020).
Article PubMed PubMed Central Google Scholar
Itoh, M. et al. CD11c+ resident macrophages drive hepatocyte death-triggered liver fibrosis in a murine model of nonalcoholic steatohepatitis. JCI Insight 2, e92902 (2017).
Yamamoto, T. et al. DUSP10 constrains innate IL-33-mediated cytokine production in ST2(hi) memory-type pathogenic Th2 cells. Nat. Commun. 9, 4231 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Zhang, K. et al. Cutting edge: Notch signaling promotes the plasticity of group-2 innate lymphoid cells. J. Immunol. 198, 1798–1803 (2017).
Article CAS PubMed Google Scholar
Ochiiwa, H. et al. TAS4464, a NEDD8-activating enzyme inhibitor, activates both intrinsic and extrinsic apoptotic pathways via c-Myc-mediated regulation in acute myeloid leukemia. Oncogene 40, 1217–1230 (2021).
Article CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Methylation of Gata3 protein at Arg-261 regulates transactivation of the Il5 gene in T helper 2 cells. J. Biol. Chem. 290, 13095–13103 (2015).
Article CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Akt1-mediated Gata3 phosphorylation controls the repression of IFNgamma in memory-type Th2 cells. Nat. Commun. 7, 11289 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Bcl11b sets pro-T cell fate by site-specific cofactor recruitment and by repressing Id2 and Zbtb16. Nat. Immunol. 19, 1427–1440 (2018).
Article CAS PubMed PubMed Central Google Scholar
Hosokawa, H. et al. Cell type-specific actions of Bcl11b in early T-lineage and group 2 innate lymphoid cells. J. Exp. Med. 217, e20190972 (2020).
Hosokawa, H. et al. Stage-specific action of Runx1 and GATA3 controls silencing of PU.1 expression in mouse pro-T cells. J. Exp. Med. 218, e20202648 (2021).
Okabe, A. et al. Cross-species chromatin interactions drive transcriptional rewiring in Epstein-Barr virus-positive gastric adenocarcinoma. Nat. Genet. 52, 919–930 (2020).
Article CAS PubMed Google Scholar
Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol 9, R137 (2008).
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 38, 576-589 (2010).
Buenrostro, J. D., Wu, B., Chang, H. Y. & Greenleaf, W. J. ATAC-seq: A Method for assaying chromatin accessibility genome-wide. Curr. Protoc. Mol. Biol. 109, 21.29.21–21.29.29 (2015).
Article Google Scholar
Rizq, O. et al. Dual inhibition of EZH2 and EZH1 sensitizes PRC2-dependent tumors to proteasome inhibition. Clin. Cancer Res. 23, 4817–4830 (2017).
Article CAS PubMed PubMed Central Google Scholar
Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).
Article CAS PubMed PubMed Central Google Scholar
MacParland, S. A. et al. Single cell RNA sequencing of human liver reveals distinct intrahepatic macrophage populations. Nat. Commun. 9, 4383 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Shinoda, K. et al. Nematode ascarosides attenuate mammalian type 2 inflammatory responses. Proc. Natl Acad. Sci. USA 119, e2108686119 (2022).
Trapnell, C. et al. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014).
Article CAS PubMed PubMed Central Google Scholar
Efremova, M., Vento-Tormo, M., Teichmann, S. A. & Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nat. Protoc. 15, 1484–1506 (2020).
Article CAS PubMed Google Scholar
Hildreth, A. D. et al. Single-cell sequencing of human white adipose tissue identifies new cell states in health and obesity. Nat. Immunol. 22, 639–653 (2021).
Article CAS PubMed PubMed Central Google Scholar
Wong, K. L. et al. Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets. Blood 118, e16–e31 (2011).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We are grateful to Y. Endo for helpful suggestions and experimental support throughout this work, M. Mastumoto and K. Nakayama for LC‒MS/MS analysis, and A. Kaneda and A. Okabe for helpful suggestions on ChIP-seq analysis and data interpretation. The authors would also like to thank R. Yuki for helpful support for ATAC-seq analysis, K. Murata for supporting scRNA-seq library construction, M. Nakanishi for basic suggestion on scRNA-seq analysis, K. Sugaya for cell sorting, Y. Sano and H. Miya for sample preparation. This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology (Japan) (Grants-in-Aid: for Scientific Research [S] 26221305, JP19H05650, [B] #21H02974, #19H03708, #22300325; [C] #22K08644, #22K07205, #22K08619, #21K08524, #20K08397, #20K07561, #19K07635, 19K08972, #18K07439, #18K08464; Challenging Research [Exploratory] #21K19398; Early-Career Scientists #20K17527; Japan Agency for Medical Research and Development, AMED [No JP21ek0410060] AMED-CREST, AMED [No. JP21gm1210003]). T. Tanaka was supported by the Japan Society for the Promotion of Science KAKENHI grant JP19H03708. This work was partly supported by the Uehara Memorial Foundation, the Mochida Memorial Foundation for Medical and Pharmaceutical Research, the Naito Foundation, the Mitsui Life Social Welfare Foundation, the Princes Takamatsu Cancer Research Fund, the Takeda Science Foundation, the Senshin Medical Research Foundation, the Kose Cosmetology Research Foundation, the Japan Diabetes Foundation, the Yamaguchi Endocrine Research Foundation, the Cell Science Research Foundation, the Ichiro Kanehara Foundation for the Promotion of Medical Sciences and Medical Care, the Yasuda Memorial Medical Foundation, the MSD Life Science Foundation, the Hamaguchi Foundation for the Advancement of Biochemistry, the Novartis Foundation (Japan) for the Promotion of Science and the Medical Institute of Bioregulation Kyushu University Cooperative Research Project Program.

Author information

Authors and Affiliations

Department of Molecular Diagnosis, Graduate School of Medicine, Chiba University, Chiba, Japan
Masanori Fujimoto, Masataka Yokoyama, Akitoshi Nakayama, Naoko Hashimoto, Ikki Sakuma, Hidekazu Nagano, Kazuyuki Yamagata & Tomoaki Tanaka
Department of Endocrinology, Hematology and Gerontology, Graduate School of Medicine, Chiba University, Chiba, Japan
Masanori Fujimoto & Koutaro Yokote
Department of Immunology, Graduate School of Medicine, Chiba University, Chiba, Japan
Masahiro Kiuchi, Atsushi Onodera, Kiyoshi Hirahara & Toshinori Nakayama
Department of Immunology, Tokai University School of Medicine, Isehara, Kanagawa, Japan
Hiroyuki Hosokawa
Department of Systems Medicine, Graduate School of Medicine, Chiba University, Chiba, Japan
Fujimi Kudo & Ichiro Manabe
Department of Medical Physiology, Chiba University, Graduate School of Medicine, Chiba, Japan
Eunyoung Lee, Ryo Hatano & Takashi Miki
Institute for Advanced Academic Research, Chiba University, Chiba, Japan
Atsushi Onodera
Research Institute of Disaster Medicine, Chiba University, Chiba, Japan
Takashi Miki & Tomoaki Tanaka
AMED-CREST, AMED, Otemachi, Tokyo, Japan
Toshinori Nakayama

Authors

Masanori Fujimoto
View author publications
You can also search for this author in PubMed Google Scholar
Masataka Yokoyama
View author publications
You can also search for this author in PubMed Google Scholar
Masahiro Kiuchi
View author publications
You can also search for this author in PubMed Google Scholar
Hiroyuki Hosokawa
View author publications
You can also search for this author in PubMed Google Scholar
Akitoshi Nakayama
View author publications
You can also search for this author in PubMed Google Scholar
Naoko Hashimoto
View author publications
You can also search for this author in PubMed Google Scholar
Ikki Sakuma
View author publications
You can also search for this author in PubMed Google Scholar
Hidekazu Nagano
View author publications
You can also search for this author in PubMed Google Scholar
Kazuyuki Yamagata
View author publications
You can also search for this author in PubMed Google Scholar
Fujimi Kudo
View author publications
You can also search for this author in PubMed Google Scholar
Ichiro Manabe
View author publications
You can also search for this author in PubMed Google Scholar
Eunyoung Lee
View author publications
You can also search for this author in PubMed Google Scholar
Ryo Hatano
View author publications
You can also search for this author in PubMed Google Scholar
Atsushi Onodera
View author publications
You can also search for this author in PubMed Google Scholar
Kiyoshi Hirahara
View author publications
You can also search for this author in PubMed Google Scholar
Koutaro Yokote
View author publications
You can also search for this author in PubMed Google Scholar
Takashi Miki
View author publications
You can also search for this author in PubMed Google Scholar
Toshinori Nakayama
View author publications
You can also search for this author in PubMed Google Scholar
Tomoaki Tanaka
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

M.F., M.Y., and T.T. designed the experiments. M.F., M.Y., M.K., H.H., A.N., N.H., I.S., H.N., K. Yamagata, F.K., E.L., and R.H. performed the experiments. M.F., M.Y., M.K., I.M., A.O., K.H., K. Yokote, T.M., T.N., and T.T. analyzed the data and provided discussions. M.F., M.Y., H.H., A.O., T.M., and T.T. wrote manuscript. T.T. supervised the study.

Corresponding author

Correspondence to Tomoaki Tanaka.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Fujimoto, M., Yokoyama, M., Kiuchi, M. et al. Liver group 2 innate lymphoid cells regulate blood glucose levels through IL-13 signaling and suppression of gluconeogenesis. Nat Commun 13, 5408 (2022). https://doi.org/10.1038/s41467-022-33171-6

Download citation

Received: 11 November 2021
Accepted: 04 September 2022
Published: 15 September 2022
DOI: https://doi.org/10.1038/s41467-022-33171-6

This article is cited by

MicroRNA-874 targets phosphomevalonate kinase and inhibits cancer cell growth via the mevalonate pathway
- Alimasi Aersilan
- Naoko Hashimoto
- Tomoaki Tanaka
Scientific Reports (2022)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.