Adipocyte SIRT1 Controls Systemic Insulin Sensitivity by Modulating Macrophages in Adipose Tissue

Transcription

1 Manuscript EMBO Adipocyte SIRT1 Controls Systemic Insulin Sensitivity by Modulating Macrophages in Adipose Tissue Xiaoyan Hui, Mingliang Zhang, Ping Gu, Kuai Li, Yuan Gao, Donghai Wu, Yu Wang, Aimin Xu Corresponding author: Aimin Xu, The University of Hong Kong Review timeline: Submission date: 10 August 2016 Editorial Decision: 15 September 2016 Revision received: 17 December 2016 Editorial Decision: 23 January 2017 Revision received: 26 January 2017 Editorial Decision: 31 January 2017 Revision received: 05 February 2017 Accepted: 07 February 2017 Editor: Achim Breiling Transaction Report: (Note: With the exception of the correction of typographical or spelling errors that could be a source of ambiguity, letters and reports are not edited. The original formatting of letters and referee reports may not be reflected in this compilation.) 1st Editorial Decision 15 September 2016 Thank you for the submission of your research manuscript to EMBO reports. We have now received reports from the three referees that were asked to evaluate your study, which can be found at the end of this . As you will see, all three referees acknowledge the potential interest of the findings. However, all three referees have raised a number of concerns and have come up with suggestions to improve the manuscript, or to strengthen the data and the conclusions drawn, which we feel need all to be addressed during the revision. As the reports are below, I will not detail them here, but in particular, the three points of referee #1 (experiments in Fig. 5D/E), point 2 of referee #2 and point 2 of referee #3 should be addressed experimentally. Given these constructive comments, we would like to invite you to revise your manuscript with the understanding that all referee concerns must be fully addressed in the revised manuscript and in a complete point-by-point response. Acceptance of your manuscript will depend on a positive outcome of a second round of review. It is EMBO reports policy to allow a single round of revision only and acceptance or rejection of the manuscript will therefore depend on the completeness of your responses included in the next, final version of the manuscript. REFEREE REPORTS European Molecular Biology Organization 1

2 Referee #1: The authors propose a novel function for adipocyte specific SIRT1 in protecting against onset of obesity and obesity driven metabolic dysregulations by blocking M1 macrophages and promoting anti-inflammatory M2 macrophages and reducing inflammation through deacetylating NFATc1 and promoting IL-4 expression. This will be an interesting addition to the field of metabolism and SIRT1 biology as it offers to delineate the function of adipocyte SIRT1 in adipose tissue from adipocyte precursors or myeloid tissue. It is very well written paper; however, my enthusiasm was diminished in the experiments performed in 5D and E, which are critical to make the claim that SIRT1 modulated IL-transcription through DEACETYLATING NFATc1 because of the following: 1. The band for SIRT1 in fig 5D, which is precipitating with NFATc1 is a very faint smear; suggesting the interaction is very weak if present at all. 2. Non-immune IgG as a negative control should have been used in 5E. 3. It is NOT made clear by the authors that acetylated-lys band in AKO panel is the same band as NFATc1. Based on this, the authors should provide stronger evidence by additional experiments on the proposed direct physical and functional interaction between adipocyte SIRT1 and NFATc1. Co-IP with SIRT1 in primary WT adipocytes or 3t3-L1 adipocytes and blot with NFATc1 could be included, and/or sequencing of the ac-lys could be done to identify the lysine residue is indeed in NFATc1 and is a direct target of SIRT1 deacetylation. Mapping the interaction domain between SIRT1 and NFATc1 to suggest few experiments that could strengthen the argument within the proposed mechanism Referee #2: The authors demonstrate the requirement of SIRT1 in mature adipocytes for protection against systemic insulin resistance in mice fed a high-fat diet. They observe increased accumulation of M1- polarized macrophages in SIRT1-deficient adipose tissue. They find that loss of SIRT1 in adipocytes disrupts the expression of key cytokines, including IL-4, MCP1, and adiponectin, which contributes to the altered macrophage phenotype. Finally, they provide evidence to suggest that SIRT1 regulates IL-4 expression in adipocytes by deacetylating the transcription factor NFATc1. Reviewer's Comments: This manuscript presents an interesting and novel set of observations concerning the crosstalk between adipocytes and macrophages in the setting of obesity-associated metabolic disease. The article should be suitable for publication in EMBO Reports after some minor revisions. The following concerns are provided to aid the authors in their revision of the manuscript prior to publication: 1. The authors need to show their flow cytometry data (plots) for Figure 2 to indicate how they identified the populations. 2. The authors evaluated the quantities of macrophages and T cells, but did not measure the quantities of other resident or recruited immune cells that play important roles in insulin resistance. It would be very informative to at least measure numbers of eosinophils and neutrophils in the adipose tissue of high-fat diet mice by flow cytometry to provide a more complete understanding. While the authors show that IL-4 quantity in the SVF is not reduced, suggesting that eosinophilderived IL-4 is not an altered variable, it would be helpful to show this directly. This is an important point because the expression of IL-4 and IL-13 by non-immune cells is very controversial. 3. The authors should show the same sets of M1 and M2 marker genes throughout their figures. In the current manuscript, some genes are shown only in some figures but not in others, without a clear reason for the discrepancy. The gene sets should be consistent or a reason should be provided for the inconsistency. European Molecular Biology Organization 2

3 4. Figure 2 F and G should be labelled to indicate that F4/80 is being detected/measured (for clarity). 5. There is a mistake in the text: "Furthermore, the acetylation level of SIRT1 in freshly isolated SIRT1-deficient adipocytes". The word in bold should be NFATc Referee #3: In this study by Hui et al, the authors compared the effects of loss of SIRT1 in adipose tissue, to those in macrophage. For this, the CreLoxP method was used to generate MKO and AKO mice using the LysM or AP2 Cre promoters, respectively. The most obvious changes related to insulin sensitivity and inflammation appeared to occur in the AKO mouse. The authors concluded that adipocyte SIRT1 controls systemic glucose homeostasis and insulin sensitivity via crosstalk with macrophages. 1. In the AKO mouse, SIRT1 KO also occurred in BAT. Was this expected? Some discussion on the implications of this are warranted. It seems interesting that loss of SIRT1 in BAT did not affect energy expenditure, and this point should be included as part of the broader discussion. 2. Do the authors think that adipocyte SIRT1 is having effects on glucose metabolism etc by regulating adipocyte insulin sensitivity? There is evidence in the literature suggesting that adipocyte insulin sensitivity can regulate macrophage infiltration and inflammation (McCurdy et al. 2012;61: ), which may fit with the authors model. This may provide insight into their findings. As part of this, it would certainly be helpful to measure adipose tissue insulin sensitivity. 3. The authors suggest that they study the effects of aging, yet only study mice up to 45 weeks old. The reviewer appreciates that they studies them every month or so, but to study aging, you need to study mice up to 20 to 24 months of age. The aging angle should be minimised. 4. The overt jump in fasting insulin from week 30 to week 45 seems 'strange'. Do the authors have insight into why there was such a large jump in this time frame? 5. Are the WT mice littermate controls? Or, were the mice bred as 3 distinct lines? Please expand the Methods appropriately. If the WT mice are not littermates of the respective MKO or AKO mice lines, this needs to be justified, and is also a major confounding variable that needs to be taken into consideration. Also, were only floxed mice used as WT mice, or were Cre mice (that are negative for the floxed allele) also used as WT mice? 6. It seems important to verify that SIRT1 KO was maintained across "age" in MO and AKO mice, particularly given the turnover of these tissues. Please provide a WB or PCR of these tissues at the measured time point, particularly 30 and 45 weeks. 7. The energy expenditure data appears problematic. Isn't the VO2 data much lower than what would be expected in mice (usually in the ~4,000 ml/kg/h). And, the RER data is very low (well below 0.7). Are the values/units for VO2 (and VCO2) correct? Insight into this data is needed. 8. The authors used 2 deoxyglucose to measure tissue glucose uptake during the clamp. The reviewer did not see that data presented. Did they calculate insulin-stimulated glucose uptake from the 3H during the clamp? Please include the 2DOG data. 9. Given the focus of the paper is on the adipocyte, it seems very important to have a marker of adipose insulin sensitivity. Was the suppression of plasma fatty acids by insulin (during the clamp) different between groups? Or, pakt in adipose different? Adipose tissue insulin resistance can have an important impact on peripheral insulin sensitivity, so such analysis is important. 10. For much of the WT vs. AKO comparisons, it looks as though HFD increased those parameters (e.g. TNFalpha), it is just that the increase was less. Do the authors think there is a 'magic' (so to speak) threshold above which things need to be in order for things (such as insulin action, European Molecular Biology Organization 3

4 inflammation, mo infiltration) to get worse? Some discussion of this would be helpful to the reader. 11. It is not clear how CM was collected from MKO and WT adipose tissue. Please provide a more detailed description of this experiment. Was it from HFD mice? If so, what are the implications of this? f not, same thing, what does this mean? 12. The authors state that their results in the MKO mouse are novel. Is that really the case given that a number of papers have shown similar findings? The authors need to do a better job of convincing the reader as to why their work is novel, as compared to other works in the same area. 13. Overall the Methods are insufficient in detail. It should be expanded, where appropriate. 14. Were male or female mice studied? 15. How many days were the mice in the CLAMS unit? Typically, when mice are first placed the CLAMS it takes them at least 24 h to get accustomed to their new surroundings, and to adjust to the access to the food. Was there an acclimation period for the mice in the CLAMS? 16. In the Figure Legends, or Statistics section of the Methods, it would be helpful to identify which results were analysed using a 1 way or 2 way ANOVA. Was a repeated measures ANOVA used for aging data? That is, are the same mice used at each of the aging timepoints? Minor comments: - Please provide catalogue number for antibodies. - A brief description of the analysis for EV1B would be helpful. What primers were used, and where were they targeted? - Please provide quantification of data in EV1C. What % change in Ac-p53 was found? - What does fed insulin mean? Were the mice fasted, then fed? Or is it a random sampling? Please describe in Methods. 1st Revision - authors' response 17 December 2016 First of all, we would like to express our sincere thanks to the reviewers invaluable comments. We have now revised our manuscript by performing additional experiments. In particular, we have addressed the three points of referee #1 (experiments in Fig. 5D/E), point 2 of referee #2 and point 2 of referee #3, as requested by Dr. Breiling. Firstly, we performed co-ip in both primary WT adipocytes and in HEK293T cells co-transfected plasmids encoding human SIRT1 and NFATc1. By co-ip of SIRT1, the interactions between these two proteins were detected in both these two systems (Figure 5D-E), whereas the interaction was abolished when the C terminus of SIRT1 ( aa) was deleted (Figure 5E). Moreover, the experiment in Figure 5E was re-performed by including mouse non-immune IgG as controls. Secondly, we quantified the numbers of eosinophils and neutrophils in epididymal adipose tissue of obese WT and AKO mice by flow cytometry, which confirmed that that there was no significant difference in number/percentage of eosinophils and neutrophils in epididymal WAT between WT and AKO mice (Figure EV5). Thirdly, we compared the insulin-stimulated glucose uptake and phosphorylation on S473 of Akt in SVF-derived adipocytes from WT and AKO adipose tissue (Figure EV4), showing that there was no significant difference on insulin sensitivity between WT and SIRT1-null adipocytes. In addition, we presented data from our hyperinsulinemic /euglycemic clamp which showed that glucose uptake in edididymal adipose tissue of AKO mice was not significantly altered (Figure 1L). Our point-by-point responses to the comments of the three referees are as follows. Responses to Referee #1 Comments from the referee: The authors propose a novel function for adipocyte specific SIRT1 in protecting against onset of obesity and obesity driven metabolic dysregulations by blocking M1 macrophages and promoting anti-inflammatory M2 macrophages and reducing inflammation European Molecular Biology Organization 4

5 through deacetylating NFATc1 and promoting IL-4 expression. This will be an interesting addition to the field of metabolism and SIRT1 biology as it offers to delineate the function of adipocyte SIRT1 in adipose tissue from adipocyte precursors or myeloid tissue. It is very well written paper; however, my enthusiasm was diminished in the experiments performed in 5D and E, which are critical to make the claim that SIRT1 modulated IL-transcription through DEACETYLATING NFATc1 because of the following: 1. The band for SIRT1 in fig 5D, which is precipitating with NFATc1 is a very faint smear; suggesting the interaction is very weak if present at all. Answer: We have now performed additional experiments to support the presence of physical interaction between SIRT1 and NFATc1. Firstly, we performed co-ip with SIRT1 in primary WT adipocytes and blotted with NFATc1. The results showed that NFATc1 co-immunoprecipitates with SIRT1. The data has now been presented as new Figure 5D. In addition, we also detected the interaction between SIRT1 and NFATc1 in 293T cells co-transfected with plasmids encoding human NFATc1 and FLAG-tagged human SIRT1, which showed that these two proteins readily interact with each other. This part of the study is now shown as Figure 5E. 2. Non-immune IgG as a negative control should have been used in 5E. Answer: According to your suggestion, we have re-performed the experiments by including nonimmune mouse IgG as control. The data are presented as Figure 5F. 3. It is NOT made clear by the authors that acetylated-lys band in AKO panel is the same band as NFATc1. Answer: Sorry that we did not make it clear. The samples were immunoprecipated eluate of NFATc1. The samples were blotted with acetylated-lysine, stripped and re-probed with NFATc1. The acetylated-lysine band shown in Figure 5F is at the same size as blotted with NFATc1 (around 100 kd). Therefore, we have reason to believe the acetylated-lysine band shown in Figure 5F is the same band as NFATc1. To avoid confusion, we have provided more detailed information in the method section and in the main text (p9, line 26 and p18, line 8, in marked version). 4. Based on this, the authors should provide stronger evidence by additional experiments on the proposed direct physical and functional interaction between adipocyte SIRT1 and NFATc1. Co-IP with SIRT1 in primary WT adipocytes or 3t3-L1 adipocytes and blot with NFATc1 could be included, and/or sequencing of the ac-lys could be done to identify the lysine residue is indeed in NFATc1 and is a direct target of SIRT1 deacetylation. Mapping the interaction domain between SIRT1 and NFATc1 to suggest few experiments that could strengthen the argument within the proposed mechanism. Answer: To provide more insight into the interactive domain between SIRT1 and NFATc1, we generated truncated forms of SIRT1 (1-243 aa, aa) which lack either the C terminus or C terminus plus catalytic domain (Figure 5E). Co-IP experiment showed that the interaction was almost completely abrogated when C terminus of SIRT1( aa) was truncated (Figure 5E), suggesting that this domain is essential for interaction with NFATc1. This part of data has now been added as Figure 5E. Responses to Referee #2 Comments from the referee: The authors demonstrate the requirement of SIRT1 in mature adipocytes for protection against systemic insulin resistance in mice fed a high-fat diet. They observe increased accumulation of M1-polarized macrophages in SIRT1-deficient adipose tissue. They find that loss of SIRT1 in adipocytes disrupts the expression of key cytokines, including IL-4, MCP1, and adiponectin, which contributes to the altered macrophage phenotype. Finally, they provide evidence to suggest that SIRT1 regulates IL-4 expression in adipocytes by deacetylating the transcription factor NFATc1. This manuscript presents an interesting and novel set of observations concerning the crosstalk between adipocytes and macrophages in the setting of obesity-associated metabolic disease. The article should be suitable for publication in EMBO Reports after some minor revisions. The following concerns are provided to aid the authors in their revision of the manuscript prior to publication: 1. The authors need to show their flow cytometry data (plots) for Figure 2 to indicate how they identified the populations. European Molecular Biology Organization 5

6 Answer: The plots of flow cytometry are now included in the manuscript as Appendix Figure S1. 2. The authors evaluated the quantities of macrophages and T cells, but did not measure the quantities of other resident or recruited immune cells that play important roles in insulin resistance. It would be very informative to at least measure numbers of eosinophils and neutrophils in the adipose tissue of high-fat diet mice by flow cytometry to provide a more complete understanding. While the authors show that IL-4 quantity in the SVF is not reduced, suggesting that eosinophilderived IL-4 is not an altered variable, it would be helpful to show this directly. This is an important point because the expression of IL-4 and IL-13 by non-immune cells is very controversial. Answer: Thank you for the constructive comment. We have now measured the numbers of eosinophils and neutrophils in edididymal WAT of obese WT and AKO mice. The flow cytometry results demonstrate that there is no significant difference in number/percentage of eosinophils and neutrophils in edididymal WAT between WT and AKO mice. The results are now resented as figure EV5 and described on p8, line 20, in marked version. 3. The authors should show the same sets of M1 and M2 marker genes throughout their figures. In the current manuscript, some genes are shown only in some figures but not in others, without a clear reason for the discrepancy. The gene sets should be consistent or a reason should be provided for the inconsistency. Answer: Yes, we agree with the referee that the same sets of M1/M2 markers should be consistently shown throughout the figures. We thus have unified our data presentation in terms of M1 and M2 marker gene expressions in the whole manuscript, using Tnfα, Mcp1, inos as M1 markers, and Arg1, Mgl1, Mgl2, Mrc2 as M2 markers. 4. Figure 2 F and G should be labelled to indicate that F4/80 is being detected/measured (for clarity). Answer: Labels have been added to Figure 2F and 2G to improve the clarity. 5. There is a mistake in the text: "Furthermore, the acetylation level of SIRT1 in freshly isolated SIRT1-deficient adipocytes". The word in bold should be NFATc1. Answer: The typo has been corrected (p9, line 25). Thank you for the careful checkup. Responses to Referee #3 Comments from the referee: In this study by Hui et al, the authors compared the effects of loss of SIRT1 in adipose tissue, to those in macrophage. For this, the CreLoxP method was used to generate MKO and AKO mice using the LysM or AP2 Cre promoters, respectively. The most obvious changes related to insulin sensitivity and inflammation appeared to occur in the AKO mouse. The authors concluded that adipocyte SIRT1 controls systemic glucose homeostasis and insulin sensitivity via crosstalk with macrophages. 1. In the AKO mouse, SIRT1 KO also occurred in BAT. Was this expected? Some discussion on the implications of this are warranted. It seems interesting that loss of SIRT1 in BAT did not affect energy expenditure, and this point should be included as part of the broader discussion. Answer: (1) Our adipose tissue specific SIRT1 KO mice (AKO) is constructed by mating Sirt1flox/flox mice with ap2-cre transgenic mice. ap2 is pan-adipocyte marker and is expressed in both white and brown adipocytes. Therefore, deletion of SIRT1 in BAT is expected. (2) Liang et al. reported that whole body knockout of Sirt1 did not influence expression of brown markers in BAT, but only led to lower levels of brown markers in subcutaneous WAT after overnight cold exposure at 4 (Liang Q et al, Cell , 620). Our CLAMS study was performed under ambient temperature, and therefore it is possible that no significant difference on energy expenditure was detected between WT and AKO mice. We have discussed the above two points in the discussion section (p10, line 11 in marked version) as suggested by the reviewer. European Molecular Biology Organization 6

7 2. Do the authors think that adipocyte SIRT1 is having effects on glucose metabolism etc by regulating adipocyte insulin sensitivity? There is evidence in the literature suggesting that adipocyte insulin sensitivity can regulate macrophage infiltration and inflammation (McCurdy et al. 2012;61: ), which may fit with the authors model. This may provide insight into their findings. As part of this, it would certainly be helpful to measure adipose tissue insulin sensitivity. Answer: Thanks reviewer for the constructive comment. To investigate whether deletion of SIRT1 affects insulin sensitivity, insulin-stimulated glucose uptake and phosphorylation of Akt (S473) were compared using SVF-derived mature adipocytes from WT and AKO mice. No differences between WT and SIRT1-deficient adipocytes were observed in terms of both glucose uptake and Akt phosphorylation on S473 (Figure EV4), indicating that SIRT1 does not directly regulate adipocyte insulin sensitivity. In addition, data from our hyperinsulinemic /euglycemic clamp showed that while insulin-stimulated glucose uptake in skeletal muscle was strongly suppressed in AKO mice, glucose uptake in edididymal adipose tissue was not significantly altered (Figure 1L). 3. The authors suggest that they study the effects of aging, yet only study mice up to 45 weeks old. The reviewer appreciates that they studies them every month or so, but to study aging, you need to study mice up to 20 to 24 months of age. The aging angle should be minimised. Answer: The claims of aging has now been avoided throughout the manuscript (Figure EV2, descriptions on p5, line 10 and other places where necessary). 4. The overt jump in fasting insulin from week 30 to week 45 seems 'strange'. Do the authors have insight into why there was such a large jump in this time frame? Answer: The exact reason for the jump of fasting insulin from week 30 to 45 is not clear. However, as shown in Figure EV2D, a steady increase of AUC for GTT, from week 8 to 30, was observed, suggesting that mice developed glucose intolerance at week 30. However, AUC at week 45 was comparable to that at week 30, it is thus possible that at this stage the mice became hyperinsulinemic as a defense mechanism to combat against the glucose intolerance and insulin resistance of the mice. 5. Are the WT mice littermate controls? Or, were the mice bred as 3 distinct lines? Please expand the Methods appropriately. If the WT mice are not littermates of the respective MKO or AKO mice lines, this needs to be justified, and is also a major confounding variable that needs to be taken into consideration. Also, were only floxed mice used as WT mice, or were Cre mice (that are negative for the floxed allele) also used as WT mice? Answer: WT controls used in this study were littermates of AKO and MKO mice and only floxed mice without cre were used. Details on mice breeding have now been provided in the method section (p14, line 20). 6. It seems important to verify that SIRT1 KO was maintained across "age" in MKO and AKO mice, particularly given the turnover of these tissues. Please provide a WB or PCR of these tissues at the measured time point, particularly 30 and 45 weeks. Answer: As suggested by the editor and also due to time limitation for submitting the revised manuscript, we currently are unable to obtain samples from 30 and 45 weeks old mice. However, SIRT1fl/fl and cre transgenic mice, which are used to generate the AKO and MKO, are germlinetransmitted, meaning that the genome in every single cell (both germline and somatic) of AKO and MKO are SIRT1 floxed and cre-positive. Therefore, even if there are turnover of tissues or cells, newly arising cells have exactly the same genomic information as their early ages, and therefore SIRT1 will be consistently deleted as long as the cell expresses ap2 (as in mature adipocytes) or Lysozyme M (myeloid cells). 7. The energy expenditure data appears problematic. Isn't the VO2 data much lower than what would be expected in mice (usually in the ~4,000 ml/kg/h). And, the RER data is very low (well below 0.7). Are the values/units for VO2 (and VCO2) correct? Insight into this data is needed. Answer: Thanks for the comment. We have carefully checked our raw data and found that the abnormal value of VO2 and VCO2 were caused by our mistakes for calculation, which subsequently European Molecular Biology Organization 7

8 led to lowered RER value as well. We have recalculated and analyzed the data as shown in Figure EV3E-I. 8. The authors used 2 deoxyglucose to measure tissue glucose uptake during the clamp. The reviewer did not see that data presented. Did they calculate insulin-stimulated glucose uptake from the 3H during the clamp? Please include the 2DOG data. Answer: We measured 3H-labeld 2-deoxyglucose uptake during the clamp. The insulin-stimulated glucose uptake in skeletal muscle and edididymal adipose tissue are now shown as Figure 1K-1L. 9. Given the focus of the paper is on the adipocyte, it seems very important to have a marker of adipose insulin sensitivity. Was the suppression of plasma fatty acids by insulin (during the clamp) different between groups? Or, pakt in adipose different? Adipose tissue insulin resistance can have an important impact on peripheral insulin sensitivity, so such analysis is important. Answer: As shown in Figure 1L, the insulin-stimulated glucose uptake in edididymal adipose tissue were not significantly different between WT and AKO mice, while glucose uptake in muscle was marginally impaired in AKO mice, suggesting that the blunted peripheral insulin sensitivity was not mainly contributed by adipose tissue insulin resistance. 10. For much of the WT vs. AKO comparisons, it looks as though HFD increased those parameters (e.g. TNFalpha), it is just that the increase was less. Do the authors think there is a 'magic' (so to speak) threshold above which things need to be in order for things (such as insulin action, inflammation, mo infiltration) to get worse? Some discussion of this would be helpful to the reader. Answer: This is an interesting idea in terms of magic threshold which we do not have definitive evidences to support. However, as a strong inducing factor for a cluster of metabolic diseases, HFDinduced obesity is well-known to evoke a series of biological events, such as inflammation, insulin resistance etc. Studies by us and others have demonstrated that SIRT1 serves as a protective player to antagonize metabolic deterioration under various conditions. Our current study clearly demonstrated that deletion of SIRT1 further exacerbates obesity-induced parameters compared to the WT mice, suggesting SIRT1 contributes to safeguard metabolic homeostasis. We have now discussed this point in the discussion section (p10, line 18). 11. It is not clear how CM was collected from MKO and WT adipose tissue. Please provide a more detailed description of this experiment. Was it from HFD mice? If so, what are the implications of this? If not, same thing, what does this mean? Answer: The CM was collected from SVF-derived mature adipocytes from WT and AKO adipose tissue. We have provided a more detailed description of the experiment on p16, line 33. The CM were not from HFD mice but from SVF-derived mature adipocytes. Our results showed that bone marrow derived macrophages (from normal C57BL/6J mice) exhibited enhanced M1 and reduced M2 markers when they were cultured in CM from SIRT1-deficient adipocyte, compared to that from WT adipocyte, whereas supplementation of recombinant IL-4 to AKO adipocyte CM to levels comparable to that of WT adipocyte CM reversed the changes in M1 and M2 markers. These data thus provide in vitro evidence that (1) SIRT1-deficinet adipocytes induce macrophage toward more M1 and less M2 phenotypes by secreted factors; (2) and the reduced level of IL-4 secreted from SIRT1-deficient adipocyte is at least in part responsible for its altered actions on macrophages. 12. The authors state that their results in the MKO mouse are novel. Is that really the case given that a number of papers have shown similar findings? The authors need to do a better job of convincing the reader as to why their work is novel, as compared to other works in the same area. Answer: Sorry we did not make it clear. Actually, we did not claim that our findings on MKO are novel and we are fully aware that several papers on MKO mice have been reported. Compared to existing publications, what we provide addition insight in this field is that we compared in parallel the metabolic parameters among WT, AKO and MKO mice, which showed that AKO mice developed diet-induced insulin resistance much earlier than WT and MKO mice, suggesting that SIRT1 in adipocyte play a more important role in controlling adipose inflammation and systemic glucose homeostasis during the onset of obesity-associated adipose inflammation and systemic European Molecular Biology Organization 8

9 insulin resistance. This point is discussed in the first paragraph of the discussion section (10, line 24). 13. Overall the Methods are insufficient in detail. It should be expanded, where appropriate. Answer: As suggested by the reviewer, we have provided additional information for several experimental procedures in the method section, including mice model generation, serum collection, glucose tolerance test, CM collection and other details where appropriate. 14. Were male or female mice studied? Answer: Male mice were studied in our study, which is specified in method section (p14, line 22). 15. How many days were the mice in the CLAMS unit? Typically, when mice are first placed the CLAMS it takes them at least 24 h to get accustomed to their new surroundings, and to adjust to the access to the food. Was there an acclimation period for the mice in the CLAMS? Answer: Yes, here we confirm that the mice were housed in CLAMS unit in a total of 72h. The mice were pre-acclimated in CLAMS for 24 h and the data collected during h were used for analysis. The details have now been specified in the method section on p15, line In the Figure Legends, or Statistics section of the Methods, it would be helpful to identify which results were analysed using a 1 way or 2 way ANOVA. Was a repeated measures ANOVA used for aging data? That is, are the same mice used at each of the aging timepoints? Answer: Most of the results were analyzed using two-way ANOVA, and the rest of the analysis were using two-sided student t test or one-way ANOVA. The statistics analysis methods are now specified in figure legends of each of the figure panels. And two-way repeated measures ANOVA was used for aging data. Minor comments: Please provide catalogue number for antibodies. Answer: Catalogue number for all the antibodies and clone number for all the monoclonal antibodies used in the study are now provided in the method section. A brief description of the analysis for EV1B would be helpful. What primers were used, and where were they targeted? Answer: The primers used in EV1B are now shown in Table 1. The primers and genotype strategy are provided by the Jackson Laboratory ( DE:13808,008041). Basically, the primer set spans exon 4 of mouse SIRT1 gene and amplifies a band of 900 bp in WT mice, while in KO mice, since regions between two flox sites are deleted by cre recombinase, the band is 450 bp in length. The positions of the primers used are now illustrated in Figure EV1A and a brief description is added in the figure legend of EV1 (p26, line 23). Please provide quantification of data in EV1D. What % change in Ac-p53 was found? Answer: EV1D is quantified using image J, and as shown in Figure EV1E, a ~4 fold and ~3 fold increase of Ac-p53 were observed in adipocyte of AKO and SVF of MKO mice. What does fed insulin mean? Were the mice fasted, then fed? Or is it a random sampling? Please describe in Methods. Answer: It was random sampling. Specifically, the serum sample of mice was collected from tail tip at 10:00 am. The details have been provided in the method section on p14, line 26. European Molecular Biology Organization 9

10 2nd Editorial Decision 23 January 2017 Thank you for the submission of your revised manuscript to our editorial offices. We have now received the report from the referees that were asked to re-evaluate your study (you will find enclosed below). As you will see, the referees now support the publication of your manuscript in EMBO reports. However, referee #2 has some further concerns that we ask you to address in a final revision (please clarify the gating strategy, add simple histograms as indicated, and discuss and tone down the claim of adipocytes secreting IL-4). Also referee #3 has some minor points, we ask you to address. Further, I have a few editorial requests that also need to be addressed in a final revised version of the manuscript. - Please format the references according to EMBO reports style. See: - The title of the paper is currently too long. Please provide a shorter title (max. 100 characters, including spaces). Also the abstract is slightly too long. Please shorten it to below 175 words. - I suggest to move the table with the primer information to the Appendix. Please add the table to the Appendix (including the TOC), name it Appendix Table S1, cut the empty rows, and add a title and a legend. Please update the callouts regarding the table in the manuscript text. - Further, there is a callout for Fig. EV6 in the manuscript text, but there is no Fig. EV6. The legend of Fig. 4 lists a panel "e", but there is no "e" in the figure. The legend of Fig. EV3 describes "A-C", but D is part of that section (I guess it should be "A-D"). There are no callouts in the text for Appendix Figures S1 and S2. Please correct this. - Finally, as also referee #3 notes, please have the paper reviewed and corrected by a native speaker. REFEREE REPORTS Referee #1: I recommend that Manuscript to be accepted for publication in EMBO reports. The authors have adequately responded to the points I raised in my review. Controls controls have been included in fig 5 to provide strong support for a functional and physical interaction between SIRT1 and NFATc Referee #2: The authors have addressed most of my concerns. However, the new data provided on gating strategy for M1 and M2 macrophages in adipose tissue of mice SCD or HFD is questionable (Appendix Figure S1). The separation of signal from noise is not clean. In fact, it is not clear whether the authors gated out debris and doublets, because the signal being seen is Figure S1 could easily come from debris and doublets (especially the low level expression of F4/80). Similarly, countour plots can be misleading for looking at expression of CD11c and CD206. For all these antigens, simple histograms should be shown to demonstrate the expression of these antigens. While it is unreasonable to ask the authors to repeat their entire experiment to address this concern, I think it might be more appropriate for the authors to temper their conclusions, as this raises concerns about data presented in Figures 2I, J, and K. In addition, I remain unconvinced that adipocytes express IL-4. Given the concerns about the authors gating and flow cytometry, low level contamination of their adipocyte fractions with immune cells could give a false positive result. While AKO mice are clearly insulin resistant, the mechanism might involve expression of chemotactic factors, which recruit the appropriate immune cells for secretion of IL-4 into the adipose tissue. I had suggested to the authors in the initial review that this should be toned down, which they have not done. I don't think the presented data is strong enough to claim that adipocytes secrete IL European Molecular Biology Organization 10

11 Referee #3: The authors have made significant additions and changes to their manuscript, which strengthens their findings. Some minor issues need to be addressed. Although the point of the manuscript is clear, it would benefit by being reviewed for grammar and general writing clarity. For Figure 1K, please clarify in the manuscript that it is for skeletal muscle, and also, which muscle it being measured. The authors discuss a protective role of SIRT1 in a variety of tissues on metabolic disease and HFD. However, this protection does not occur in muscle, as shown by various studies, which should be acknowledged, particularly given its major role in glucose metabolsim (PMID: , , , ). 2nd Revision - authors' response 26 January 2017 Responses to Referee #2 1.The authors have addressed most of my concerns. However, the new data provided on gating strategy for M1 and M2 macrophages in adipose tissue of mice SCD or HFD is questionable (Appendix Figure S1). The separation of signal from noise is not clean. In fact, it is not clear whether the authors gated out debris and doublets, because the signal being seen is Figure S1 could easily come from debris and doublets (especially the low level expression of F4/80). Similarly, countour plots can be misleading for looking at expression of CD11c and CD206. For all these antigens, simple histograms should be shown to demonstrate the expression of these antigens. Answer: In the previous version, we only showed the representative contour plots of our flow cytometry analysis. The gating strategy has now been added as Appendix fig 1A, which shows how we exclude debris and doublets. In addition, simple histograms for all three antigens are presented as Appendix fig 1B. 2. While it is unreasonable to ask the authors to repeat their entire experiment to address this concern, I think it might be more appropriate for the authors to temper their conclusions, as this raises concerns about data presented in Figures 2I, J, and K. In addition, I remain unconvinced that adipocytes express IL-4. Given the concerns about the authors gating and flow cytometry, low level contamination of their adipocyte fractions with immune cells could give a false positive result. While AKO mice are clearly insulin resistant, the mechanism might involve expression of chemotactic factors, which recruit the appropriate immune cells for secretion of IL-4 into the adipose tissue. I had suggested to the authors in the initial review that this should be toned down, which they have not done. I don't think the presented data is strong enough to claim that adipocytes secrete IL-4. Answer: The conclusions on Fig 2I-K has been tempered as suggested (line 3, p7). In addition, we have toned down our claim on adipocyte-derived IL-4 and discussed other possible mechanisms in the discussion section (line 17, p12) Responses to Referee #3 The authors have made significant additions and changes to their manuscript, which strengthens their findings. Some minor issues need to be addressed. 1. Although the point of the manuscript is clear, it would benefit by being reviewed for grammar and general writing clarity. Answer: The manuscript has now been edited by a native speaker. 2. For Figure 1K, please clarify in the manuscript that it is for skeletal muscle, and also, which muscle it being measured. European Molecular Biology Organization 11

12 Answer: Soleus muscle was used in our study and the information is now provided in the main text (line 2, p6). 3. The authors discuss a protective role of SIRT1 in a variety of tissues on metabolic disease and HFD. However, this protection does not occur in muscle, as shown by various studies, which should be acknowledged, particularly given its major role in glucose metabolsim (PMID: , , , ). Answer: The role of SIRT1 in muscle and the related literatures have now been added in the discussion section (line 12, p10). 3rd Editorial Decision 31 January 2017 Thank you for the submission of your revised manuscript to our editorial offices. We think that now all referee concern are addressed. Before we can proceed with formal acceptance, I have a few editorial requests that need to be addressed: Please format the references according to EMBO reports style. See: The legend of Fig. 4 lists a panel "e", but there is no "e" in the figure. I guess that the "E" in Fig. 4 is missing (lower right panel). Please add this and upload the revised figure. Please submit your final revised manuscript in black fonts throughout and remove the highlighting. You submitted source data only for figure 5D. Is there a special reason for that? Or, would it be possible to submit source data for all the Western panels (that are substantially cropped). In case, please submit the source data (scans of entire gels or blots) of your key experiments together with the revised manuscript. Please include size markers for scans of entire gels, label the scans with figure and panel number, and send one PDF file per figure or per figure panel. I look forward to seeing a revised version of your manuscript when it is ready. Please let me know if you have questions or comments regarding the revision. 3rd Revision - authors' response 05 February 2017 Thank you for your swift response and we are glad to know that all the scientific questions raised have been satisfactorily addressed. Now we are submitting the updated version of the manuscript as suggested. Hopefully the manuscript in the current form is eligible for final acceptance by EMBO Reports. The detailed changes are listed below: 1. The references have been reformatted to EMBO Reports style. 2. The missing E in Figure 4 has been added. 3. Source data files for all the Western blot panels are uploaded. 4th Editorial Decision 07 February 2017 I am very pleased to accept your manuscript for publication in the next available issue of EMBO reports. Thank you for your contribution to our journal. European Molecular Biology Organization 12

13 EMBO PRESS YOU MUST COMPLETE ALL CELLS WITH A PINK BACKGROUND ê PLEASE NOTE THAT THIS CHECKLIST WILL BE PUBLISHED ALONGSIDE YOUR PAPER Corresponding Author Name: Aimin Xu Journal Submitted to: EMBO Reports Manuscript Number: EMBOR V1 Reporting Checklist For Life Sciences Articles (Rev. July 2015) This checklist is used to ensure good reporting standards and to improve the reproducibility of published results. These guidelines are consistent with the Principles and Guidelines for Reporting Preclinical Research issued by the NIH in Please follow the journal s authorship guidelines in preparing your manuscript. A- Figures 1. Data The data shown in figures should satisfy the following conditions: è the data were obtained and processed according to the field s best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. è figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way. è graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. è è if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation. USEFUL LINKS FOR COMPLETING THIS FORM Antibodypedia 1DegreeBio ARRIVE Guidelines NIH Guidelines in animal use MRC Guidelines on animal use Clinical Trial registration CONSORT Flow Diagram CONSORT Check List Reporting Guidelines (marker prognostic studies) Dryad Figshare dbgap EGA 2. Captions Biomodels Database Each figure caption should contain the following information, for each panel where they are relevant: MIRIAM Guidelines è a specification of the experimental system investigated (eg cell line, species name). JWS Online è the assay(s) and method(s) used to carry out the reported observations and measurements Biosecurity Documents from NIH è an explicit mention of the biological and chemical entity(ies) that are being measured. List of Select Agents è an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner. è the exact sample size (n) for each experimental group/condition, given as a number, not a range; è a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.). è a statement of how many times the experiment shown was independently replicated in the laboratory. è definitions of statistical methods and measures: common tests, such as t-test (please specify whether paired vs. unpaired), simple χ2 tests, Wilcoxon and Mann-Whitney tests, can be unambiguously identified by name only, but more complex techniques should be described in the methods section; are tests one-sided or two-sided? are there adjustments for multiple comparisons? exact statistical test results, e.g., P values = x but not P values < x; definition of center values as median or average; definition of error bars as s.d. or s.e.m. Any descriptions too long for the figure legend should be included in the methods section and/or with the source data. Please ensure that the answers to the following questions are reported in the manuscript itself. We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects. In the pink boxes below, provide the page number(s) of the manuscript draft or figure legend(s) where the information can be located. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). B- Statistics and general methods 1.a. How was the sample size chosen to ensure adequate power to detect a pre-specified effect size? 1.b. For animal studies, include a statement about sample size estimate even if no statistical methods were used. Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) Sample sizes of animal studies was chosen on the basis of literature documentation of similar wellcharacterized experiments, and no statistical method was used to predetermine sample size. This is specified in the method section on p18. The statement about sample sie estimate is shown in method section on p Describe inclusion/exclusion criteria if samples or animals were excluded from the analysis. Were the criteria preestablished? No inclusion/exclusion criteria were adopted in the study. 3. Were any steps taken to minimize the effects of subjective bias when allocating animals/samples to treatment (e.g. randomization procedure)? If yes, please describe. For animal studies, include a statement about randomization even if no randomization was used. Animals with the same genotypes were randomized as neither standard chow or high fat diet (p14). Differentiated primary adipocytes, 3T3-L1 adipocytes and bone marrow derived macrophages were randomized for variours treatments. Animals with the same genotypes were randomized as neither standard chow or high fat diet (p14). 4.a. Were any steps taken to minimize the effects of subjective bias during group allocation or/and when assessing results Animals with the same genotypes were randomized as neither standard chow or high fat diet (e.g. blinding of the investigator)? If yes please describe. (p14). Differentiated primary adipocytes, 3T3-L1 adipocytes and bone marrow derived macrophages were randomized for variours treatments. 4.b. For animal studies, include a statement about blinding even if no blinding was done Animals with the same genotypes were randomized as neither standard chow or high fat diet (p14). 5. For every figure, are statistical tests justified as appropriate? Yes. Do the data meet the assumptions of the tests (e.g., normal distribution)? Describe any methods used to assess it. All the data meet the assumptions of the testes and the assessing statistical methods are specified in each panel of the figure legends (from p14). Is there an estimate of variation within each group of data? Yes. Is the variance similar between the groups that are being statistically compared? Yes. C- Reagents 6. To show that antibodies were profiled for use in the system under study (assay and species), provide a citation, catalog number and/or clone number, supplementary information or reference to an antibody validation profile. e.g., Antibodypedia (see link list at top right), 1DegreeBio (see link list at top right). 7. Identify the source of cell lines and report if they were recently authenticated (e.g., by STR profiling) and tested for mycoplasma contamination. NA D- Animal Models * for all hyperlinks, please see the table at the top right of the document 8. Report species, strain, gender, age of animals and genetic modification status where applicable. Please detail housing and husbandry conditions and the source of animals. These details can be found in method section on page For experiments involving live vertebrates, include a statement of compliance with ethical regulations and identify the committee(s) approving the experiments. 10. We recommend consulting the ARRIVE guidelines (see link list at top right) (PLoS Biol. 8(6), e , 2010) to ensure Other relevant aspects of animal studies comply with the guidelines. that other relevant aspects of animal studies are adequately reported. See author guidelines, under Reporting Guidelines. See also: NIH (see link list at top right) and MRC (see link list at top right) recommendations. Please confirm compliance. E- Human Subjects 11. Identify the committee(s) approving the study protocol. 12. Include a statement confirming that informed consent was obtained from all subjects and that the experiments conformed to the principles set out in the WMA Declaration of Helsinki and the Department of Health and Human Services Belmont Report. 13. For publication of patient photos, include a statement confirming that consent to publish was obtained.