Male C57BL/6J wild-type mice were purchased from Jackson Laboratories (Bar Harbor, ME). Mice were bred and maintained under specific pathogen-free condition at the Animal Facility of the University of Pittsburgh School of Medicine, VA Pittsburgh Healthcare System, and the Children’s Hospital, Zhejiang University School of Medicine. All mice used in the experiments were 8 weeks old. All mice were performed in compliance with the guidelines of the Institutional Animal Care and Use Committee of the University of Pittsburgh, VA Pittsburgh Healthcare System, and the Children’s Hospital, Zhejiang University School of Medicine, respectively.

Sepsis was induced by CLP procedure as described previously (23). In short, mice were deeply anesthetized with an intraperitoneal injection of xylazine (5 mg/kg) and ketamine (50 mg/kg). A midline incision (1 cm) on the abdomen was performed to allow exteriorization of the cecum. To obtain a moderate CLP, the cecum was ligated 0.8 cm from the apex with 4-0 silk suture and punctured once with a 22-gauge needle in the ligated segment. To induce a severe CLP, the cecum was ligated 1.2 cm from the apex and punctured twice with an 18-gauge needle. A droplet of cecal contents was then slowly squeezed out of the puncture holes. Then the cecum was placed back into the abdomen. The incision was then sutured in two layers. Sham surgery was identical to CLP without puncture and ligation. Mice were fluid resuscitated immediately after surgery (1 ml/mouse sterile saline, subcutaneously). At the specified time point, mice were sacrificed, and lung tissue samples were obtained under sterile condition. Whole blood was collected by cardiac puncture and spun down, and samples were stored at -80°C for further analysis.

In the survival analysis, the survival of animals (5 mice per group) was monitored each 3 hours for 72 consecutive hours after CLP surgery. To relieve pain, mice with signs of imminent death were overdosed with xylazine/ketamine. The survival rate was evaluated, followed by plotting the survival curve.

In some in vivo experiments, mice were injected intraperitoneally (i.p.) with 0.2 µg/g B.W. of NMU-23 peptide (Phoenix Pharmaceuticals, USA) at 6h before and after CLP or with a single dose of NMU-23 peptide (1 µg/g B.W.) at 6h before CLP. At 24h after CLP, lungs were harvested for further analysis. Control mice were treated with PBS.

Cell isolation from lung tissue was performed as described previously (24). Briefly, lungs were perfused with 5 ml cold PBS with 2% heparin through right ventricle of the heart, and then filled with 1 ml HBSS with Liberase ™ (100 μg/ml final concentration) (Roche, USA) and digested in 4 ml HBSS digestion medium for 45 min at 37°C with vortexing every 15 min. The resultant samples were mashed by 70-μm cell strainers, washed with Dulbecco’s modified Eagle media [DMEM; supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Thermo Fisher Scientific, Pittsburgh, PA, USA)], and treated with RBC Lysis Buffer (eBioscience ™) to lyse red blood cells. Cell suspensions were used for subsequent flow cytometry staining.

Western blot was performed using standard methods. In short, protein (30 μg) was electropharesed through 12% SDS polyacrylamide gels and transferred to PVDF membranes (Bio-Rad, USA). The membranes were incubated with primary antibodies at 4°C overnight, followed by secondary antibodies tagged with HRP (Thermo Fisher Scientific, USA) at room temperature for 1 hour. The signals were detected using ECL Kit (Pierce Biotech, Rockford, Illinois, USA). A GAPDH antibody was used as a control for whole-cell lysates.

For flow cytometry analysis, anti-mouse CD16/CD32 antibody (eBioscience, USA) was added to samples at a 1:200 dilution for 20 minutes at 4°C to block nonspecific binding to Fc receptors before cell staining.

LIVE/DEAD Fixable Aqua Dead Cell Stain Kit (eBioscience), Fixable Viability Dye eFluor™ 780 (eBioscience) or 7AAD viability dye (eBioscience) were used to exclude dead cells. Lung cell suspensions were stained with anti-CD45 (30-F11), anti-CD3ϵ (17A2), anti-CD4 (RM4-5), anti-CD5 (53-7.3), anti-CD8α (53-6.7), anti-CD11b (M1/70), anti-CD11c (N418), anti-CD19 (eBio1D3), anti-NK1.1 (PK136), anti-TER119 (TER-119), anti-FceR1α (MAR-1), anti-TCRβ (H57-597), anti-TCRγδ (GL-3), anti-Gr1 (RB6-8C5), anti-Thy1.2 (CD90.2; 53-2.1), anti-CD25 (eBio7D4), anti-CD127 (IL-7Rα; A7R34), anti-IL-17A (eBio17B7) from eBioscience; anti-KLRG1 (2F1), anti-Sca1 (D7) from BD Biosciences, anti-T1/ST2 (DJ8) from MD Biosciences. Lineage was composed by CD3ϵ, CD4, CD5, CD8α, CD11b, CD11c, CD19, NK1.1, TER119, FceR1α, TCRβ, TCR γδ and Gr1; Cell populations were defined as: ILC2s, CD45⁺Lineage⁻Thy1.2⁺T1/ST2⁺CD127⁺CD25⁺KLRG1⁺Sca1⁺; γδ T cells, CD45⁺CD3ϵ⁺TCR γδ⁺CD4^-TCR β^-.

For intracellular cytokine protein analysis ex vivo, cells were stimulated using the Cell Stimulation Cocktail (eBioscience), containing PMA/Ionomycin/Brefeldin‐A/monensin, for 4 hours at 37°C before staining. Intracellular staining was performed using IC fixation/permeabilization kit (eBioscience).

Flow cytometry analysis and cell sorting were performed using LSR Fortessa, FACS Aria flow cytometers (BD Biosciences) and Cytek Aurora (Cytek Biosciences). The percentage of ILC2s is gated in live CD45⁺Lineage⁻ cells. Data analysis was done using FlowJo software (Tristar).

For flow cytometric sorting, Lin^-CD45⁺CD90.2⁺ST2⁺ ILC2s were sorted from the lungs of naive mice by FACSAria (BD Biosciences) or Beckman MoFlo Astrios EQ (Beckman Coulter Life Sciences, Indianapolis, IN, USA). The average purity of ILC2s is > 98%. In vitro culture of ILC2s was conducted as previously described (25). Sorted ILC2s were routinely grown in DMEM glutaMAX (Gibco) supplemented with 10% FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 1% hepes, sodium pyruvate, glutamine at 37°C. Lung ILC2s were plated in 96-well round-bottom plates with two densities (1.5 × 10⁴ or 3.0 × 10⁴ cells/well) in 10 ng/ml rmIL-2 (Biolegend, San Diego, CA, USA), 20 ng/ml rmIL-7 (Biolegend) and 20 ng/ml rmIL-33 (Biolegend) for 6 days. Before use, ILC2s were gently washed twice to remove residual rmIL-2, rmIL-7, and rmIL-33.

Lung γδ T cells were collected after flushing the lungs with 5 ml cold PBS through right ventricle to remove circulating cells. Fresh γδ T cells were enriched from lung by negative and positive selection using the TCRγ/δ⁺ T cell isolation Kit (Miltenyi Biotec, Gladbach Bergische, Germany), then CD45⁺CD3ϵ⁺TCR γδ⁺ γδ T cells were sorted by FACSAria (BD Biosciences) or Beckman MoFlo Astrios EQ (Beckman Coulter Life Sciences, Indianapolis, IN, USA). The average purity of γδ T cells is > 98%. In vitro culture of γδ T cells was conducted as previously described (26). Sorted γδ T cells were routinely grown in DMEM glutaMAX (Gibco) supplemented with 10% FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 1% hepes, sodium pyruvate, glutamine at 37°C. Lung γδ T cells were plated in 96-well round-bottom plates with a density of 5.0 × 10³ cells/well, in 100 ng/ml rmIL-1β (Biolegend, San Diego, CA, USA) and 100 ng/ml rmIL-23 (Biolegend) to polarize IL-17A-producing γδ T cells (26–28).

For direct co-culture assays, sorted lung ILC2s were cultured with rmIL-2, rmIL-7 and rmIL-33 for 6 days to obtain mature ILC2s, then sorted lung γδ T cells were added to the each ILC2 well, in DMEM glutaMAX (Gibco) supplemented with 10% FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 1% hepes, sodium pyruvate, glutamine at 37°C. To maintain ILC2 survival, rmIL-7 (20 ng/ml) was included in all assays with ILC2s as well as controls, including when co-culturing with γδ T cells. ILC2s and γδ T cells were co-cultured for 48 hours before the next tests if not otherwise specified.

The following substances were added to cultures as indicated: NMU (1 or 10 μg/ml, Phoenix Pharmaceuticals), rmIL-1β (100 ng/ml; Biolegend), rmIL-23 (100 ng/ml; Biolegend), LPS (1 μg/ml; Sigma-Aldrich) plus TNF-α (20 ng/ml; Biolegend) were added to mimic sepsis stimulation (13).

Total RNA from sorted cells or tissues were extracted by RNeasy Plus Mini Kit (Qiagen, Germantown, MD, USA) or Trizol method (Thermo Fisher Scientific, Pittsburgh, PA, USA) and stored at -80°C for further analysis. Total RNA concentration was measured using a Nanodrop One spectrophotometer (Thermo Fisher Scientific, Pittsburgh, PA, USA). Total RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription kit (Thermo Fisher Scientific, Pittsburgh, PA, USA) according to the protocol.

Quantitative PCR was conducted in triplicate on a CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) with TaqMan Gene Expression Master Mix (Applied Biosystems) using the following TaqMan Gene Expression Assays (Applied Biosystems): Nmu (Mm00479868_m1); Nmur1 (Mm00515885_m1); Il5 (Mm00439646_m1); Il9 (Mm00434305_m1); Il13 (Mm00434204_m1); Il17a (Mm00439618_m1).

Gene expression was normalized as n-fold difference to the gene Hprt1 (Mm00446968_m1) and S18 (Mm03928990_g1) for mouse according to the cycling threshold. Calculation of mRNA levels was performed with the CFX Manager Software version 3.1 (Bio-Rad).

For lung homogenate, the whole lung was snap frozen on dry ice homogenized in RIPA buffer (Sigma-Aldrich) containing 0.01% protease and phosphatase inhibitor cocktail (Thermo Scientific). The cell debris and tissue were removed by centrifugation at 19,000 g for 30 min at 4°C. The supernatant was collected for analysis of IL-17A (R&D System) by ELISA according to the manufacturer’s instructions.

Blood samples were collected, and plasma was obtained by centrifugation at 5,000 g for 20 min at 4°C. For determination of mouse IL-1β, IL-9, IL-17A and IL-23, Quantikine ELISA Kits from R&D Systems were used according to manufacturers’ instructions.

For CRISPR/Cas9-mediated gene knockdown, the following synthetic guide RNA (sgRNA) sequences were used: 5’-CGATATGCTGGTGCTCCTGG-3’ (targeting nmur1); 5’-GTGAGCGGACAGCTGTGTCA-3’ (targeting Il5); 5’-ATTGTACCACACCGTGCTAC-3’ (targeting Il9); 5’-CTTCGATTTTGGTATCGGGG-3’ (targeting Il13); 5’-AAUGUGAGAUCAGAGUAAU-3’ (non-target control) (ThermoFisher Scientific, USA). Ex vivo–expanded ILC2s were transfected with nmur1/Il5/Il9/Il13 CRISPR/Cas9 plasmid or its non-target control (NTC) in accordance with the manufacturer’s instructions. Transfected cells were cultured for 2 days before next step.

Statistical analyses were done using GraphPad Prism 7.00 software (GraphPad Software, Inc., La Jolla, CA, USA). Survival differences were assessed using the Kaplan-Meier analysis followed by a log-rank test. Student’s t test or ANOVA was used in all other experiments. Data were expressed as mean ± SEM. A P value < 0.05 was considered statistically significant, and significance is presented as * P < 0.05, ** P < 0.01, *** P < 0.001, or **** P < 0.0001.

Sepsis induced significant increases in plasma IL-17A levels, lung tissue Il17a mRNA expression, and IL-17A protein concentration ( Figures 1A–C ), which were consistent with the previous observations (29). Noteworthy, sepsis also induced a markedly increase in the percentage of lung IL-17A-producing cells ( Figures 1D, E ).

Our previous study has shown that sepsis induces ILC2 expansion in the lungs (13). However, it was unknown whether ILC2s secrete IL-17A during sepsis, although it has been reported that in allergic conditions ST2⁺ nILC2s secrete IL-17A (17). We found that IL-17A-producing ILC2s were significantly upregulated in the lungs following the CLP procedure ( Figures 1F, G ). However, importantly, IL-17A-producing ILC2s only occupy 1~2% of total IL-17A-producing cells ( Figure 1H ).

Since multiple cell types, including lineage^- ILCs, CD4⁺ T helper 17 (Th17) cells, CD8⁺ (Tc17) cells, and γδ T cells can produce and secret IL-17A (30), we then assessed the relative contribution of these cells to the elevated lung IL-17A in sepsis. Using the flow cytometry gating strategy, we found the γδ T cell lineage, but not lineage^- ILCs, CD4⁺ T cells, and CD8⁺ T cells, are the major source of IL-17A ( Figures 1D, H ). This finding underscores the important role of γδ T cells in producing IL-17A in the lung in sepsis. We also found that the percentage and numbers of lung γδ T cells were significantly higher in the CLP group than that in the sham group ( Figures 1I–K ). Collectively, these findings suggest an important role for lung γδ T cells in producing and secretion of IL-17A in sepsis.

NMU-NMUR1 signaling has been reported to play an important role in the regulation of inflammation in asthma and helminth infection models (19, 22). To determine the role of NMU-NMUR1 signaling in sepsis-induced inflammation, we measured the expression of NMU in the lungs and NMUR1 expression in ILC2s following CLP. The results show that sepsis markedly increased the expressions of nmu mRNA by ~2.4-fold and NMU protein by ~2.1-fold in the lungs ( Figures 2A, B ). We then collected ILC2s by flow sorting and treated the ILC2s with LPS + TNF-α to mimic a septic condition in vitro. We found that nmur1 mRNA expression in the ILC2s treated with LPS + TNF-α is significantly increased as compared to that in the PBS treated (control) group ( Figure 2C ).

Previous studies have shown that lung nmur1 is selectively expressed in ILC2s (22, 31). To determine this is also true in sepsis, we isolated lung cells from CLP mice, then categorized the cells into three populations, including ILC2s, γδ T cells, and other cells, using flow sorting, followed by measurement of nmur1 expression in these three populations using real-time qPCR. The results showed that nmur1 was specifically expressed by ILC2s, but not by γδ T cells and other cells ( Figure 2D ) (22, 31). These findings suggest that ILC2s are the major cell population to respond to the increased NMU expression in the lungs in sepsis.

Based on the data shown above, we hypothesized that NMU might act through ILC2s to upregulate lung IL-17A-producing γδ T cells. To test this hypothesis, we intraperitoneally injected (i.p.) recombinant NMU into the CLP mice at 6 h before and 6 h after the CLP procedure ( Figure 3A ). We found that NMU significantly reduced the mortality of septic mice ( Figure 3A ). Furthermore, exogenous NMU increased the percentage of ILC2s ( Figures 3B, C ), IL-17A-producing cells ( Figures 3D, E ), and γδ T cells ( Figures 3F, G ) in the lungs of septic mice; whereas NMU administration did not alter the percentage of IL-17A⁺ ILC2s ( Supplemental Figures A, B ). Given that nmur1 is specifically expressed by ILC2s but not by γδ T cells, the increased γδ T cells in response to NMU is mediated through ILC2s.

To determine the role of ILC2s in mediating NMU-induced upregulation of lung γδ T cells, we applied an in vitro ILC2s and γδ T cells coculture system. IL-17A was undetectable in the culture supernatant of the ILC2 alone group after NMU treatment ( Supplemental Figure C ). However, coculture of ILC2s and γδ T cells with the treatment of NMU resulted in significant increases in the number and percentage of γδ T cells ( Figures 4A–C ) and supernatant IL-17A concentrations ( Figure 4D ). NMU failed to induce γδ T cell expansion and IL-17A release in γδ T cell alone group ( Figures 4A–D ). More importantly, the supernatant IL-17A concentrations were further elevated in the ILC2-γδ T cell coculture group treated with NMU and LPS + TNF-α ( Figure 4D ).

To further establish the role of ILC2s in regulating γδ T cell expansion and IL-17A producing, we cocultured γδ T cells (5.0 × 10³ cells) with different numbers of ILC2s (1.5 × 10⁴ and 3.0 × 10⁴ cells/well). After 48-hour coculture, the final numbers of γδ T cells in the group co-cultured with 3.0 × 10⁴ ILC2s/well were ~3.2-fold higher than that in the group cocultured with 1.5 × 10⁴ ILC2s/well ( Figure 4E ); and the supernatant IL-17A concentrations of 3.0 × 10⁴ ILC2s/well group were also significantly higher than that of 1.5 × 10⁴ ILC2s/well group ( Figure 4F ). In addition, we treated the cocultures with different NMU concentrations (1 μg/ml and 10 μg/ml) but fixed the number of ILC2s (1.5 × 10⁴ cells/well). We found that higher NMU concentration (10 μg/ml) induced higher numbers of γδ T cells and higher IL-17A levels as compared to the group treated with lower NMU concentration (1 μg/ml) ( Figures 4G, H ). These results further suggested that ILC2s mediate NMU-induced γδ T cell expansion and IL-17A production.

To confirm the role of NMUR1 in ILC2s in transducing NMU signaling, we knocked down nmur1 in ILC2s using the CRISPR/Cas9 approach, which was confirmed by RT-PCR, as shown in Figure 4I . Knockdown of nmur1 significantly attenuated γδ T cell expansion in response to NMU in the co-culture system ( Figures 4J–L ), and decreased IL-17A production ( Figure 4M ).

Next, we wanted to identify the possible mediators that mediate the ILC2 regulation of γδ T cells in the co-culture system. Previous studies have shown that IL-1β and IL-23 can polarize γδ T cells to produce IL-17A (26–28, 32). Thus, we measured the supernatant levels of IL-1β and IL-23 and found that the concentrations of both cytokines were not changed in all groups of ILC2 cultures ( Supplemental Figures D, E ).

Our previous studies have shown that ILC2-derived IL-9 serves as an important mediator in the interaction between ILC2s and lung endothelial cells (13). Reports also showed that NMU can induce ILC2s to secrete IL-9 (18, 22), and IL-9/IL-9R signaling regulates γδ T-cell activation (33). In our current coculture experiments of ILC2 with γδ T cells, we observed a significant increase in supernatant IL-9 in response to treatment with NMU and LPS + TNF-α ( Figure 5A ). We further found that in cocultures of nmur1 knockdown ILC2s with γδ T cells, supernatant IL-9 levels were remarkably lower than that in cocultures of WT ILC2 with γδ T cells ( Figure 5B ). In order to further determine the role of IL-9 in mediating γδ T cell activation, we knocked down IL-9 in ILC2s using a CRISPR/Cas9 approach. The efficiency of Il9 knockdown was confirmed by detecting culture supernatant IL-9 using ELISA ( Figure 5C ). Coculture of Il9 knockdown ILC2s with γδ T cells resulted in lower γδ T cell expansion ( Figures 5D–F ) and lower levels of supernatant IL-17A ( Figure 5G ).

To determine if other ILC2-derived cytokines are also involved in mediating the interaction between ILC2s and γδ T cells, we knocked-down IL-5 and IL-13 in ILC2s using the CRISPR/Cas9 method, respectively. We found that the knockdown of IL-5 and IL-13 in ILC2s did not affect γδ T cell expansion and IL-17A production in the co-culture system (data not shown).

Collectively, the data demonstrate an important role for IL-9 in mediating ILC2 regulation of γδ T cell expansion, activation, and subsequent production of IL-17A.