PPARɣ drives IL-33-dependent ILC2 pro-tumoral functions

Group 2 innate lymphoid cells (ILC2s) play a critical role in protection against helminths and in diverse inflammatory diseases by responding to soluble factors such as the alarmin IL-33, that is often overexpressed in cancer. Nonetheless, regulatory factors that dictate ILC2 functions remain poorly studied. Here, we show that peroxisome proliferator-activated receptor gamma (PPARγ) is selectively expressed in ILC2s in humans and in mice, acting as a central functional regulator. Pharmacologic inhibition or genetic deletion of PPARγ in ILC2s significantly impair IL-33-induced Type-2 cytokine production and mitochondrial fitness. Further, PPARγ blockade in ILC2s disrupts their pro-tumoral effect induced by IL-33-secreting cancer cells. Lastly, genetic ablation of PPARγ in ILC2s significantly suppresses tumor growth in vivo. Our findings highlight a crucial role for PPARγ in supporting the IL-33 dependent pro-tumorigenic role of ILC2s and suggest that PPARγ can be considered as a druggable pathway in ILC2s to inhibit their effector functions. Hence, PPARγ targeting might be exploited in cancer immunotherapy and in other ILC2-driven mediated disorders, such as asthma and allergy.

It would be helpful to provide information on how human T helper cell subsets shown in Fig. S1 are identified and purified? : Mitotracker dyes and TMRM allow to assess the mitochondrial mass and membrane potential. However, they do not allow assessing mitochondrial respiration. More should be done to quantify more standard parameters of cellular metabolism (such as extracellular flux analysis or similar metabolic assay) in response to PPARg inhibition or deletion.
In Fig. 4 ILC2 numbers should be displayed as numbers of total lymphocytes and ideally as total cell counts. The heading of Fig. 4 implies that a direct connection between IL-33 and ILC2 is investigated although the data supplied is a mere correlation. Increase of IL-33 plasma levels do not appear to result in increased PPARg levels in ILC2. More should be done to assess the connection between tumor derived IL-33 and increased PPARg expression in vivo.
In this regard the authors could use blocking antibodies for IL33R or IL33 to confirm that the ILC2 tumor cross-talk is driven by IL-33 in vitro (Fig. 5a) and in vivo (Fig. 6).
What were the culture conditions for mouse ILC2 displayed in Fig 6? For how long were ILC2 from WT and ID2Cre PPARg fl/fl mice incubated? Fig. 7: Do ID2Cre PPARg fl/fl mice also display a difference in survival?
Is PPARg upregulated in TIL ILC2 and how is PPARg activated in TIL ILC2? The authors should link IL-33 production to the expression of PPARg in ILC2. Do other ILC subtypes expressing ID2, such as ILC1/NK cells also express PPARg in the tumor microenvironment?
Could the PPARg inhibitor also be used therapeutically as suggested by the authors? Maybe the authors could include experiments to test the efficacy of this treatment.
The authors suggest that PPARg may be a relevant target for asthma and allergies. This work has been done and could be included in the discussion of the study (Chen et al. Science Immunol. 2017, Karagiannis et al. 2020 Immunity. Reviewer #2 (Remarks to the Author): with expertise in colon cancer and IL33 signaling The manuscript builds up on the previous discovery of ILC2s that play a critical role in protection against helminths and in diverse inflammatory diseases by responding to soluble factors such as the alarmin IL-33 through the receptor ST2. Here, the authors show that peroxisome proliferatoractivated receptor gamma (PPARγ) but not PPARα and PPARβ is selectively expressed in human expanded ILC2s but not ILC1s and ILCPs. Selective PPARγ antagonist T0070907 inhibition impaired IL-13 production and mitochondrial fitness. Using a COX-2 inhibitor, they showed a slight decrease in IL-13 secretion. Although supp fig 3 show no difference in PPARγ expression in HDs vs. CRC patients, the authors indicated that PPARγ blockade in ILC2s disrupted their pro-tumoral effect induced by IL-33-secreting cancer cells in human samples. They further confirmed this ILC2-driven protumoral effect by using Ppargfl/flId2CreERT2 mice and MC38 CRC model, and showed significantly suppressed tumor growth in vivo. The authors concluded that PPARγ targeting might be exploited in cancer immunotherapy and in other ILC2-driven mediated disorders, such as asthma and allergy.
Overall, the data on PPARɣ in ILC2s are novel, and the data are well presented. However, we have major criticisms shown below: Major: 1) The major concern is that Supp Figure 3 indicated no significant difference of the PPARɣ expression between freshly sorted ILC2s from healthy donor as compared to CRC patients, which is concerning for their hypothesis on tumor microenvironment. The point by point following comments derive from this major concern. 2) Figure 1 to Figure 3, we just see the ILCs data from healthy donors but the paper is about tumor microenvironment and data from tumor-infiltrated ILCs in patients or in murine models are needed.
3) Figure1 display the expression of PPARγ on different human ILCs subsets at mRNA level. Verification at the protein level of PPARγ by flow cytometry will strengthen the data. Also, Supp. Figure1 just show the expression of PPARγ at mRNA level in CD4 T helper subsets, if the protein level of PPARγ could be further verified by flow cytometry. 4) The data of binding motif in the promoter region of IL-5 and IL-13 was generated by the HOMER software (Figure2d and 6f). It will be preferred to verify this with cell lines by co-IP and any other biochemistry pathway. 5) MitoTracker Green and Deep Red dye uptake were significantly decreased in ILC2s after treatment with PPARγ inhibitor, suggesting reduced mitochondrial mass and impaired respiration, respectively ( Figure 3). However, if the authors would like to focus on the ILC2s-mediated immunotherapy, they should use tumor-infiltrated ILC2s from patients or murine models. 6) The data displayed in Figure 4a, 4b and 4c have previously been published by the Krebs' group. 7) It would be important to show the effect of IL-13 in the in vivo data with for example with IL13 knockout or anti-IL-13 tumor model. 8) In Figure 6, could the authors explain why they sorted the ILC2s from lung tissues instead of the intestinal tissues, while the authors used a CRC model? Minor: 1) The authors should provide high resolution graph of Figure 2d and 6f. 2) As a represent flow data (Figure 2a, 4d and 6b), the percentage should be labeled in the graph.
3) The tumor volume in Figure 7c should be monitored for at least 28 days.
Reviewer #3 (Remarks to the Author): with expertise in ILC Ercolano et al. examined the role of PPARg for the pro-tumor activity of group 2 innate lymphoid cells (ILC2) in both human and mouse. The authors showed the inhibitor of PPARg, T0070907 suppressed IL-5 and IL-13 secretion from both human and mouse ILC2. PPARg inhibition was shown to suppress mitochondrial function as well. Importance of PPARg in IL-5 and IL-13 secretion was also confirmed by deletion of PPARg in mouse. The authors then examined the pro-tumor effect of ILC2 using colorectal cancer (CRC) in both human and mouse. The authors showed that ILC2-derived IL-13 stimulated migration of SW1116 human CRC cell line associated with the induction of EMT markers MMP9 and N-cadherin. The authors also showed the presence of IL-33 and CD3-GATA3+ ILC2 in CRC tissues. In a mouse model, IL-33 expressing MC38 mouse CRC cell line was better controlled in PPARg-deficient mice. Mouse deficient for ILC2 survived longer with MC38-IL-33 compared with control mice. From these results, the authors propose that PPARg is a potential target for cancer immunotherapy.
This study was well designed and preformed in a logical manner to provide readers with some new findings. As for the role of PPARg in ILC2 functions, Karagiannis et al. have already reported that PPARg together with Dgat1 and mTOR controls lipid metabolism in ILC2 (PMID: 32268121). However, this paper was not even mentioned in the manuscript, which is not appropriate. One of major potential advance of Ercolano et al. paper over Karagiannis et al. papers is that Ercolano et al. propose that PPARg directly activates IL-5 and IL-13 gene transcription with RARA but no evidence is provided, which dampens the priority of this paper.
Although this paper provides the readers with some new findings, there are several points that need to be clarified. 1) To confirm the pro-tumor effect of ILC2 in Figures 7a and 7b, the authors should adoptively transfer wild type ILC2 into RORa<fl/sg>IL7rCre and Pparg<fl/fl>Id2Cre+ mice and examine whether the transfer of ILC2 promote tumor progression and mortality.
2) Figures 2d and 6f: the authors must examine whether the motives are functional or not using either reporter assay in vitro or CRSPR-mediated gene deletion assay in vivo.
3) In figure 5g, the authors should also examine the effect of anti-IL-13 antibody as well. 4) Bottom of page8: because there is no significant difference in IL-5 and IL-13 secretion between T0070907 treatment and T0070907 plus CXB (Supplementary Figure 2b), the authors cannot conclude the interaction of COX-2 and PPARg. 5) As mentioned by the authors, the role of ILC2 on tumor immunity is multiphasic. Some papers reported anti-tumor activity of ILC2 against melanoma (PMID: 22174445, 32101749), suggesting that the benefit of suppressing ILC2 activities depends on the tumor type. The authors should discuss these points.
6) The authors should provide readers with detailed methods to obtain Th1, Th2, and Th17 cells ex vivo and in vitro expansion. 7) Similarly, methods for pre-treatment of ILC2 with T0070907 should be provided. 8) Genetic background of mice should be provided. 9) Stock number should be provided for mouse line obtained from the Jackson Laboratory.
10) The first line of page 2: "Type 2 innate lymphoid cells" should be "Group 2 innate lymphoid cells".

Reviewer #1
The paper by Ercolano et al. investigates the importance of PPARg in driving pro-tumorigenic ILC2 responses. The authors show that human and mouse ILC2 express PPARg and that blockade or genetic deletion of PPARg in ILC2 impairs cytokine production of ILC2 and thus their pro-tumoral function. This suggests that PPARg could be a potential target in cancer immunotherapy. While the analysis performed is interesting some of the findings are based primarily on correlations, which allows alternative interpretations. In particular it remains unclear how PPARg drives cytokine production in ILC2. Although the authors mention that PPARg controls lipids metabolism, no experiments are included to address the involvement of this important metabolic pathway. This is in particular important since published data suggest an essential role of PPARg by promoting lipid metabolism in ILC2 in the context of airway inflammation (Karagiannis et al. 2020 Immunity). Hence, it would be important to provide new mechanistic insights of how PPARg drives the pro-tumoral function of ILC2 in vivo.
Specific points:

Answer: We thank the reviewer for pointing out this important aspect. In the work by Angela et al., the authors showed that PPARγ expression is increased in TCR/CD28 activated CD4 T cells. In contrast, in our work human Th2 cells express PPARγ and that absence of PPARγ in Th2 cells, but not in Th1 or Th17 cells affects their polarization. In the current manuscript, we analysed PPARγ expression in freshly-sorted and in vitro expanded Th subsets (cultured in the presence of IL-2), without any TCR triggering. This differences in the experimental setups might explain the discrepant results with other reports.
It would be helpful to provide information on how human T helper cell subsets shown in Fig. S1 are identified and purified?  Figure 1 below, the PPARγ antagonist did not impair human ILC2 proliferation.

Figure 1: Proliferation rate was assessed by MTT assay in human ILC2s, after incubation with T0070907 for 72 hours.
Is PPARg directly controlling PD1 expression in ILC2 as suggested by (Batyruva et al. 2020 Immun Inflamm Dis.)?

Figure 2: A, Frequency (left) and MFI (right) of PD1 expression in mouse expanded ILC2s from the lung of Pparg fl/fl Id2Cre + mice, treated or not with 4-Hydroxytamoxifen. B, Frequency of PD1 in human expanded ILC2s after 48h silencing with a PPARG SiRNA.
Fig. 3: Mitotracker dyes and TMRM allow to assess the mitochondrial mass and membrane potential. However, they do not allow assessing mitochondrial respiration. More should be done to quantify more standard parameters of cellular metabolism (such as extracellular flux analysis or similar metabolic assay) in response to PPARg inhibition or deletion.  Figure 3 here

Figure 3: Quantification of Mitotracker Deep Red in human expanded ILC2s pre-treated or not with T0070907 and stimulated with Oligomycin for 2.5 hours. The result is represented as fold change MDR post-vs pre-Oligomycin.
In Fig. 4 ILC2 numbers should be displayed as numbers of total lymphocytes and ideally as total cell counts. Figure 4d, the absolute numbers are comparable to ILC frequencies and no significant difference is observed in the different conditions.

Answer: as suggested, we quantified ILC numbers as total cell counts. As shown in the Revised
The heading of Fig. 4 implies that a direct connection between IL-33 and ILC2 is investigated although the data supplied is a mere correlation. Increase of IL-33 plasma levels do not appear to result in increased PPARg levels in ILC2. More should be done to assess the connection between tumor derived IL-33 and increased PPARg expression in vivo. In this regard the authors could use blocking antibodies for IL33R or IL33 to confirm that the ILC2 tumor cross-talk is driven by IL-33 in vitro (Fig. 5a) and in vivo (Fig. 6). Figure 4a (pages 16, 18 and 30). Is PPARg upregulated in TIL ILC2 and how is PPARg activated in TIL ILC2? The authors should link IL-33 production to the expression of PPARg in ILC2. Do other ILC subtypes expressing ID2, such as ILC1/NK cells also express PPARg in the tumor microenvironment? Supplementary Figure 3, panels a-b (pages 13, 15, 32).

Figure 5: A, Expression of PPARγ assessed by qPCR in human freshly sorted ILC subsets and NK cells from PBMCs of HDs, and PBMCs and TILs of CRC patients. B, Expression of CPT1A assessed by qPCR in freshly sorted ILC2s from HDs and CRC patients.
Could the PPARg inhibitor also be used therapeutically as suggested by the authors? Maybe the authors could include experiments to test the efficacy of this treatment.

Answer: We thank the reviewer for raising this important translational aspect. To test the in vivo efficacy of T0070907, we implanted MC38-IL33 CRC cells in C57BL6 mice and treated the animals daily with T0070907 i.p. 7,5 mg/kg, as previously reported (Burton et al., PPAR Res. 2008). As shown in the revised supplementary Figure 6b we observed a significant reduction of both tumor volume and weight in T0070907 treated mice compared to controls.
The authors suggest that PPARg may be a relevant target for asthma and allergies. This work has been done and could be included in the discussion of the study (Chen et al. Science Immunol. 2017, Karagiannis et al. 2020 Immunity.

Reviewer #2
The manuscript builds up on the previous discovery of ILC2s that play a critical role in protection against helminths and in diverse inflammatory diseases by responding to soluble factors such as the alarmin IL-33 through the receptor ST2. Here, the authors show that peroxisome proliferator-activated receptor gamma (PPARγ) but not PPARα and PPARβ is selectively expressed in human expanded ILC2s but not ILC1s and ILCPs. Selective PPARγ antagonist T0070907 inhibition impaired IL-13 production and mitochondrial fitness. Using a COX-2 inhibitor, they showed a slight decrease in IL-13 secretion. Although supp fig 3 show no difference in PPARγ expression in HDs vs. CRC patients, the authors indicated that PPARγ blockade in ILC2s disrupted their pro-tumoral effect induced by IL-33-secreting cancer cells in human samples. They further confirmed this ILC2-driven protumoral effect by using Ppargfl/flId2CreERT2 mice and MC38 CRC model, and showed significantly suppressed tumor growth in vivo. The authors concluded that PPARγ targeting might be exploited in cancer immunotherapy and in other ILC2-driven mediated disorders, such as asthma and allergy.
Overall, the data on PPARɣ in ILC2s are novel, and the data are well presented. However, we have major criticisms shown below: Major: 1) The major concern is that Supp Figure 3 indicated no significant difference of the PPARɣ expression between freshly sorted ILC2s from healthy donor as compared to CRC patients, which is concerning for their hypothesis on tumor microenvironment. The point by point following comments derive from this major concern.
Answer: We thank the reviewer for this comment, also raised by reviewer 1 (see Figure 6 Supplementary Figure 3 a- Figure 3, panels a-b (pages 13,  15, 32).

b, we observed a trend for increased expression of PPARγ in ILC2s in patients compared to HDs, whilst the expression of PPARγ was low and comparable in ILC1s, ILC3s and NKs of HDs and patients (Figure 6 PBP reply, Panel A). Moreover, we also assessed the expression of CPT1A, one of the well-characterized, canonical PPARγ-direct target genes involved in fatty acid oxidation (Mascaró et al., J Biol Chem. 1998). In line with the increased expression of PPARγ, a trend for increased CPT1A expression was found in ILC2s from CRC patients compared to HDs which reflects an increase in PPARγ activity (Figure 6 PBP reply, Panel B). The results on the expression of PPARγ and CPT1A in ILC2s from CRC patients have been included in the revised Supplementary
2) Figure 1 to Figure 3, we just see the ILCs data from healthy donors but the paper is about tumor microenvironment and data from tumor-infiltrated ILCs in patients or in murine models are needed. Supplementary Figure 3 c-e (pages 13, 15).

Answer: We agree with the reviewer that patients' data were lacking in our original manuscript. As suggested, we performed in vitro experiments using ILC2s from CRC patients. The inhibitory effects of T0070907 on cytokine production, cytokine secretion and mitochondrial fitness were confirmed in patients' ILC2s, as shown in the revised
3) Figure1 display the expression of PPARγ on different human ILCs subsets at mRNA level. Verification at the protein level of PPARγ by flow cytometry will strengthen the data. Also, Supp. Figure1 just show the expression of PPARγ at mRNA level in CD4 T helper subsets, if the protein level of PPARγ could be further verified by flow cytometry.

Answer: We thank the reviewer for this comment. We tested several PPARγ antibodies in flow cytometry, but in our hands, we were unable to obtained reliable results. However, we have previously shown (Nobs et al, JEM, 2017) that activated human Th2 cells express PPARγ and that absence of PPARγ in Th2 cells, but not in Th1 or Th17 cells affects their polarization. In the current manuscript, we analysed PPARγ expression in freshly-sorted and in vitro expanded Th subsets (cultured in the presence of IL-2), without any TCR triggering.
4) The data of binding motif in the promoter region of IL-5 and IL-13 was generated by the HOMER software (Figure2d and 6f). It will be preferred to verify this with cell lines by co-IP and any other biochemistry pathway. (pages 8, 10, 34). Figure 6A and B here below). The results show the presence of several PPARγ binding sites in Th2 cells, suggesting that PPARγ might bind to these elements also in mouse ILC2s.

5) MitoTracker Green and Deep
Red dye uptake were significantly decreased in ILC2s after treatment with PPARγ inhibitor, suggesting reduced mitochondrial mass and impaired respiration, respectively ( Figure 3). However, if the authors would like to focus on the ILC2s-mediated immunotherapy, they should use tumorinfiltrated ILC2s from patients or murine models. ILC2s (pages 13, 15).

Answer: we thank the reviewer for the comment. As reported above in a similar comment raised by reviewer 1, we have now used expanded ILC2s from CRC patients to perform analyses of cytokine production, cytokine secretion and the mitochondrial assays. As shown in the revised Supplementary Figure 3 c-e, the inhibitory effects of T0070907 have been confirmed also in patients'
6) The data displayed in Figure 4a, 4b and 4c have previously been published by the Krebs' group.

Answer: while we agree with the reviewer that previous publications by the Krebs' group and by others reported increased IL-33 levels in CRC patients, these observations were not linked to ILC2 infiltration. In the current Figure 4, we have performed staining of IL-33 and of markers for ILC2 visualization (GATA3, CD3) on independent tissue sections as compared to previously published reports.
7) It would be important to show the effect of IL-13 in the in vivo data with for example with IL13 knockout or anti-IL-13 tumor model.

Answer: We thank the reviewer for the suggestion. To evaluate the role of IL-13 in CRC progression, we treated C57BL6 tumor-bearing mice with an anti-IL-13 neutralizing antibody, daily i.p. 50 μg/mouse, for eighteen days after tumor implant. As shown in the revised Supplementary Figure 6c, the neutralization of IL-13 significantly reduced both tumor volume and weight (pages 22, 24, 37).
8) In Figure 6, could the authors explain why they sorted the ILC2s from lung tissues instead of the intestinal tissues, while the authors used a CRC model? (Figure 7 here below).

Figure 7: Expression of Pparg assessed by qPCR analysis in freshly-sorted ILC subsets from the intestine of C57BL6 mice.
Minor: 1) The authors should provide high resolution graph of Figure 2d and 6f. Figure 2d and 6f.

Answer: as requested by the reviewer, we increased the size of graphs in
2) As a represent flow data (Figure 2a, 4d and 6b), the percentage should be labelled in the graph.

Answer: as suggested by the reviewer, we included the percentage of positive cells in the representative flow graphs.
3) The tumor volume in Figure 7c should be monitored for at least 28 days.
Answer: we thank the reviewer for this comment. Unfortunately, we are unable to monitor tumor volumes 28 days after implant, since the volume would be higher than 1000 cc, the maximal limit set by the Swiss Veterinary Authorities.

Reviewer #3
Ercolano et al. examined the role of PPARg for the pro-tumor activity of group 2 innate lymphoid cells (ILC2) in both human and mouse. The authors showed the inhibitor of PPARg, T0070907 suppressed IL-5 and IL-13 secretion from both human and mouse ILC2. PPARg inhibition was shown to suppress mitochondrial function as well. Importance of PPARg in IL-5 and IL-13 secretion was also confirmed by deletion of PPARg in mouse. The authors then examined the pro-tumor effect of ILC2 using colorectal cancer (CRC) in both human and mouse. The authors showed that ILC2-derived IL-13 stimulated migration of SW1116 human CRC cell line associated with the induction of EMT markers MMP9 and N-cadherin. The authors also showed the presence of IL-33 and CD3-GATA3+ ILC2 in CRC tissues. In a mouse model, IL-33 expressing MC38 mouse CRC cell line was better controlled in PPARg-deficient mice. Mouse deficient for ILC2 survived longer with MC38-IL-33 compared with control mice. From these results, the authors propose that PPARg is a potential target for cancer immunotherapy. This study was well designed and preformed in a logical manner to provide readers with some new findings. As for the role of PPARg in ILC2 functions, Karagiannis et al. have already reported that PPARg together with Dgat1 and mTOR controls lipid metabolism in ILC2 (PMID: 32268121). However, this paper was not even mentioned in the manuscript, which is not appropriate. One of major potential advance of Ercolano et al. paper over Karagiannis et al. papers is that Ercolano et al. propose that PPARg directly activates IL-5 and IL-13 gene transcription with RARA but no evidence is provided, which dampens the priority of this paper.
Although this paper provides the readers with some new findings, there are several points that need to be clarified.
1) To confirm the pro-tumor effect of ILC2 in Figures 7a and 7b, the authors should adoptively transfer wild type ILC2 into RORa<fl/sg>IL7rCre and Pparg<fl/fl>Id2Cre+ mice and examine whether the transfer of ILC2 promote tumor progression and mortality. (pages 22, 24, 37).

Answer: We thank the reviewer for the comment. As suggested, we performed ILC2 transfer from C57BL6 mice to ILC2 KO mice after implant of MC38-IL33 CRC cells. As shown in the revised Supplementary Figure 6a, ILC2 transfer in ILC2 KO mice significantly increased tumor growth compared to nontransferred ILC2 KO mice
2) Figures 2d and 6f: the authors must examine whether the motives are functional or not using either reporter assay in vitro or CRSPR-mediated gene deletion assay in vivo. Figure 2b- ILC2 activation (pages 8, 10, 33-34).

e). Moreover, we performed chromatin immunoprecipitation (ChIP) on expanded human ILC2s stimulated or not with the cytokine cocktail composed of IL-33 and IL-25, and treated or not with the PPARγ antagonist T0070907. As shown in the revised Supplementary Figure 2f, our results suggest that PPARγ directly regulates IL-13 production during
Lastly, we also used YRS reporter mice recently described by Ricardo-Gonzalez et al., Nat Immunol. 2018. ILC2s were sorted from the lungs of these mice and stimulated in vitro with IL-33 and IL-25, in the presence or absence of T0070907. The addition of T0070907 reduced the production of both IL-13 and IL-5, as assessed using the reporter system (Figure 8 here below).

Figure 8: (Left) Representative example of flow cytometry analysis of mouse ILC2-derived IL-13 upon in vitro T0070907 treatment in YRS reporter mice. (Right) Frequencies of IL-13 and IL-5 positive ILC2s after T0070907 treatment in YRS reporter mice.
3) In figure 5g, the authors should also examine the effect of anti-IL-13 antibody as well.

Answer: We thank the reviewer for the suggestion. As reported in the revised Figure 5g, we examined the expression of MMP9 and NCAD in SW1116 after incubation with ILC2 CM in the presence of an anti-IL-13 blocking antibody. We observed that the addition of the antibody reduced the EMT marker expression towards basal levels (pages 16-17).
4) Bottom of page8: because there is no significant difference in IL-5 and IL-13 secretion between T0070907 treatment and T0070907 plus CXB (Supplementary Figure 2b), the authors cannot conclude the interaction of COX-2 and PPARg.

Answer: we agree with the reviewer and have removed the speculation on a possible interaction of COX-2 and PPARγ (page 9).
5) As mentioned by the authors, the role of ILC2 on tumor immunity is multiphasic. Some papers reported anti-tumor activity of ILC2 against melanoma (PMID: 22174445, 32101749), suggesting that the benefit of suppressing ILC2 activities depends on the tumor type. The authors should discuss these points.

Answer: we agree with the reviewer and have included this point in the discussion (page 26).
6) The authors should provide readers with detailed methods to obtain Th1, Th2, and Th17 cells ex vivo and in vitro expansion.

Answer: as requested, details about Th subset sorting and culture have been added in the Materials and Methods section (pages 29,31).
7) Similarly, methods for pre-treatment of ILC2 with T0070907 should be provided.

Answer: as requested, details about T0070907 treatment have been added in the Materials and Methods section (page 30).
8) Genetic background of mice should be provided.

Answer: as requested, the genetic background of mice has been included in the Materials and Methods section (page 37).
9) Stock number should be provided for mouse line obtained from the Jackson Laboratory.

Answer: as requested, the stock number for the mouse obtained from the Jackson Laboratory has been included in the Materials and Methods section (page 37).
10) The first line of page 2: "Type 2 innate lymphoid cells" should be "Group 2 innate lymphoid cells".