IL-6 signaling drives self-renewal and alternative activation of adipose tissue macrophages

Introduction Obesity is associated with chronic low-grade inflammation of adipose tissue (AT) and an increase of AT macrophages (ATMs) that is linked to the onset of type 2 diabetes. We have recently shown that neutralization of interleukin (IL)-6 in obese AT organ cultures inhibits proliferation of ATMs, which occurs preferentially in alternatively activated macrophage phenotype. Methods In this study, we investigated AT biology and the metabolic phenotype of mice with myeloid cell-specific IL-6Rα deficiency (Il6ra Δmyel) after normal chow and 20 weeks of high-fat diet focusing on AT inflammation, ATM polarization and proliferation. Using organotypical AT culture and bone marrow derived macrophages (BMDMs) of IL-4Rα knockout mice (Il4ra -/-) we studied IL-6 signaling. Results Obese Il6ra Δmyel mice exhibited no differences in insulin sensitivity or histological markers of AT inflammation. Notably, we found a reduction of ATMs expressing the mannose receptor 1 (CD206), as well as a decrease of the proliferation marker Ki67 in ATMs of Il6ra Δmyel mice. Importantly, organotypical AT culture and BMDM data of Il4ra -/- mice revealed that IL-6 mediates a shift towards the M2 phenotype independent from the IL-6/IL-4Rα axis. Discussion Our results demonstrate IL-4Rα-independent anti-inflammatory effects of IL-6 on macrophages and the ability of IL-6 to maintain proliferation rates in obese AT.


Introduction
Obesity is a worldwide growing epidemic and represents a threat to global human health by shortening life expectancy due to associated diseases as type 2 diabetes, stroke, cardiovascular disease, and cancer (1,2).Until now, the only effective long-term treatment for obesity is bariatric surgery and there is a strong need to develop new therapeutic strategies by decoding the pathophysiology and molecular mechanisms of AT dysfunction.
In obesity, abnormal lipid accumulation leads to adipocyte hyperplasia and hypertrophy, which provokes abnormal dimensions of adipocyte death.Thereupon, the number of immune cells in AT rises, resulting in a chronic low-grade inflammation associated with type 2 diabetes and other comorbidities.Especially the number of adipose tissue macrophages (ATMs) increases as a hallmark of AT inflammation and is associated with insulin resistance (IR) (3).Additionally, the ATM immune phenotype switches from an antiinflammatory M2-like to a pro-inflammatory M1-like state in obese mice and humans.Histologically, ATMs accumulate in so-called crown-like structures (CLS) surrounding dying adipocytes under obese conditions (4,5).Increasing ATM number results from either CCR2-dependent recruitment of monocytes from the bloodstream or by local proliferation (6)(7)(8)(9).Interestingly, local ATM proliferation can be stimulated by anti-inflammatory cytokines, such as IL-4 or IL-13 (9).Moreover, IL-4 treatment of obese mice affects glucose and lipid metabolism as well as insulin sensitivity in a positive manner (10).Controversially, myeloid-specific deficiency of the IL-4Ra subunit in obese mice also resulted in an improvement of metabolic parameters and less M1-like macrophages in AT and, therefore, partially protects from AT inflammation (11).The receptor-receptor-and cytokine-receptorinteractions as well as downstream signaling cascades are complex and can differ between auto-and paracrine secretion modes or between different cell types.Therefore, cytokines often have pleiotropic effects, which are poorly understood in the pathophysiology of obesity and AT inflammation.For instance, Han and colleagues showed in 2019 that pleiotropic effects of IL-6 in AT depend on the cellular source of the cytokine (12).They found that IL-6 released from adipocytes stimulates macrophage infiltration, whereas IL-6 secreted from myeloid and muscle cells decreased the infiltration process into AT (12).Moreover, cell-type specific IL-6 release results in switches between non-canonical (classical) signaling (IL-6 binds to membrane bound IL-6Ra) and canonical (trans-signaling) using a soluble receptor variant (sIL-6Ra) (12).Dimerization of the IL-6/IL-6Ra subunit with ubiquitously expressed gp130 leads to STAT3 or STAT6 phosphorylation.Furthermore, ciliary neurotrophic factor (CNTF) and IL-30 have been described as additional low-affinity ligands of the so-called IL-6R/gp130/LIFR complex in divergent pathologies (13).In obesity, circulating IL-6 is elevated in line with TNF-a serum levels (14-16).Therefore, IL-6 was seen as a main driver of AT inflammation over years.This perception changed with evidence of anti-inflammatory effects of IL-6, like improved insulin sensitivity, alternative activation of macrophages, and an augmented Il4ra gene expression (17).In addition, neutralization of IL-6 in murine obese AT explants showed an inhibition of IL-4 mediated ATM proliferation combined with a decrease in Il4ra (9).
To further address the role of IL-6Ra signaling on ATM proliferation in obese individuals in vivo, we analyzed lean and obese mice with a LysM-Cre driven myeloid cell-specific deficiency of the IL-6Ra (Il6ra Dmyel ) for markers of proliferation as well as ATM polarization, general markers of AT inflammation and insulin sensitivity.Interestingly, our investigation revealed a lower percentage of CD206+ macrophages within AT and a decrease of ATM proliferation in obese Il6ra Dmyel mice compared to control mice.Moreover, IL-6 stimulation of organotypical AT explants and bone marrow derived macrophages (BMDMs) lacking the IL-4Ra subunit indicates that IL-6 shifts towards an M2 phenotype independent of IL-4Ra.

Results
Myeloid-specific Il6ra knockout does not affect weight gain or metabolism during protracted diet-induced obesity Previous studies showed that IL-6 signaling affects glucose homeostasis in a positive manner and leads to alternative activation of BMDMs (17).Furthermore, the neutralization of IL-6 in obese murine AT explant cultures leads to the inhibition of ATM proliferation (9).During obesity, serum levels of IL-6 as well as Il6 gene expression are augmented in AT (9,18).To verify the effect of IL-6 signaling on obesity-driven ATM proliferation in vivo, we used a myeloid-specific knockout of the IL-6Ra (Il6ra Dmyel ).First, IL-6Ra knockout was confirmed by a significant reduction of Il6ra gene expression in BMDMs measured by qRT-PCR (Figure 1A).Moreover, we found no differences in body weight of mice lacking the IL-6Ra compared to wildtype controls in all groups (female NCD, male NCD and male HFD) (Figure 1B).This might be due to utilization of different regimen for dietinduced obesity as published previously.Interestingly, lean male Il6ra Dmyel mice exhibited more perigonadal (PWAT) and subcutaneous (SWAT) AT compared to control mice on a chow diet, while showing no differences in food intake (Supplementary Figures 1A-C).However, HFD-provoked body weight gain was still present in Il6ra Dmyel mice (Figure 1B).Weight of several organs related to obesity-caused pathologies, e.g.fat depots, liver, and pancreas showed no significant differences in Il6ra Dmyel mice after 20 weeks of HFD (Figure 1C).Additionally, ipITT and ipGTT data of obese Il6ra Dmyel and Il6ra fl/fl mice revealed no differences in insulin sensitivity or glucose tolerance after HFD (Figures 1D, E) as well as for male mice fed a NCD (Supplementary Figures 1D-G; Supplementary Table 1).As shown in previous studies, diet-induced obesity leads to changes in gene expression in visceral AT (9).Here, we found no differences in gene expression between obese Il6ra Dmyel and control mice using qRT-PCR (Supplementary Figure 2A).

Adipocyte expansion and ATM distribution are unaltered in obese mice lacking the IL-6Ra
Next, we analyzed paraffin-embedded sections of lean and HFD-fed Il6ra Dmyel and Il6ra fl/fl mice stained for Perilipin to visualize adipocytes and the macrophage marker Mac-2 (Figure 2A).We studied macrophage distribution, formation of crown-like structures (CLS) and adipocyte size, as hallmarks of obesity induced AT dysfunction.However, analyses revealed no differences for CLS density (Figure 2B), percentage of interstitial macrophages per adipocyte (Figure 2C), or mean adipocyte diameter (Figures 2D-F) in lean and obese Il6ra Dmyel compared to Il6ra fl/fl mice.Flow cytometry analysis verified normal leukocyte and ATM enhancement after HFD in both mouse lines (Supplementary Figures 2B, C).Diet-induced obesity is concomitant with altered levels of triglycerides, free fatty acids, or cholesterol.In this study, we found no significant changes in these parameters comparing HFD Il6ra Dmyel and their littermate controls (Supplementary Table 1).

Myeloid Il6ra knockout impairs alternative activation and proliferation of ATMs in diet-induced obesity
Although we could not detect any differences in CLS formation, ATM distribution, and adipocyte size (Figures 2B-F), we hypothesized that IL-6 signaling in myeloid cells impacts on the activation state of ATMs and their proliferation as described before ex vivo (9).To test this hypothesis, we analyzed ATMs (CD45+F4/ 80+) using flow cytometry (see gating strategy Supplementary Figure 4).Obesity is associated with an augmentation in AT leukocytes, especially ATMs.We found diet-induced enhancement of these cell types independent from IL-6Ra depletion (Supplementary Figures 2B, C; Supplementary Table 1).Next, we tested the ATM expression of the pro-inflammatory marker CD11c (Integrin alpha-X; encoded by the Itgax gene) and the anti-inflammatory marker CD206 (Mannose receptor 1; encoded by the Mrc1 gene) to detect alterations in the activation state of ATMs between Il6ra knockout and wildtype mice (Figures 3A-D).Here, we defined CD11c+CD206-as pro- inflammatory M1 macrophages, whereas CD11c-CD206+ ATMs were defined as M2-like macrophages.Interestingly, we could not detect an impact of disrupted IL-6 signaling on CD11c+CD206expressing ATMs, neither in NCD nor HFD mice (Figures 3A, D; Supplementary Table 1).Hence, the increase of CD11c+CD206macrophages in AT due to obesity-related inflammation is still existent in mice lacking the IL-6Ra in myeloid cells (Figures 3A, D).Importantly, CD206 expression in ATMs of obese Il6ra Dmyel mice is reduced, whereas IL-6Ra knockout has no influence in lean individuals (Figures 3B, D).Of note, the impact of IL-6 signaling on CD206 expression is also reflected by increased CD11c-CD206and decreased CD11c+CD206+ ATM populations in obese Il6ra Dmyel mice (Supplementary Figures 2D, E).The overall activation state of ATMs reflected by the M1/M2 ratio is not significantly affected by disrupted IL-6 signaling (Figures 3C, D).
Ex vivo data suggest a role of IL-6 signaling in ATM proliferation.Hence, we measured the percentage of all ATMs expressing the proliferation marker Ki67 within obese AT comparing Il6ra Dmyel and their control littermates by flow cytometry (Figures 3E, F) or by microscopy (Figures 3G, H).In accordance with previous ex vivo data, proliferation of ATMs measured by Ki67 expression is significantly decreased in mice lacking IL-6Ra in myeloid cells during HFD as measured by two independent methods (Figures 3E-H).In lean mice, we found no difference in ATM Ki67 expression (Supplementary Table 1).Of note, with our protocol we were not able to detect changes in Bromodesoxyuridine (BrdU) incorporation in AT leukocytes of obese Il6ra Dmyel and control animals (Supplementary Figures 2F, G).However, also different T-cell populations can affect ATM polarization and proliferation, which in turn affects insulin sensitivity and the outcome of diet-induced obesity (19)(20)(21).Therefore, we analyzed T-cell subsets isolated from obese AT of either Il6ra Dmyel or Il6ra fl/fl mice fed a NCD or HFD (Supplementary Figure 3).However, by comparing male Il6ra Dmyel and Il6ra fl/fl mice on a HFD, we could not detect significant differences (Supplementary Figures 3A, D; Supplementary Table 1).We also measured the expression of ST2 and FoxP3 in CD4+ T-cell populations to quantify prevalence of Th2 and regulatory T-cells.Here, we also did not find differences in male HFD mice with or without an intact IL-6Ra subunit (Supplementary Table 1).

IL-6 elevates proliferation of antiinflammatory M2 macrophages partially independent of the IL-4Ra
In previous studies, alternative activation of ATMs as well as ATM proliferation was linked to IL-6 signaling via the IL-4Ra-Stat6axis (9,17).To verify IL-6-mediated impact on ATM polarization in inflammatory AT and to get more insights into the interrelation with IL-4Ra signaling, we investigated the effect of IL-6 stimulation in AT of Il4ra +/+ (wildtype) and Il4ra -/-(knockout) mice.To mimic obesityinduced AT inflammation in lean individuals, AT explants were cultured for 7d to induce adipocyte death, CLS formation and augmentation of the pro-inflammatory M1 phenotype as shown before (9).Treatment of AT organ culture with IL-6 (50 ng/ml) for 48h resulted in a significant decrease of CD11c+CD206-ATMs and a significant increase of CD11c-CD206+ ATMs as measured by flow cytometry (Figures 4A-D) confirming our results from myeloidspecific IL-6Ra examinations.Of note, our Il4ra knockout model showed a variation in the baseline of ATM polarization due to global IL-4Ra deficiency.Therefore, data were normalized to the respective PBS control.Stimulation of AT explants with IL-13, a potent ligand of IL-4Ra, generated conventionally M2-polarized ATMs for comparison in control AT and, on the other hand, confirmed efficient knockout in IL-4Ra-deficient AT (Figures 4A-D).Most importantly, upon IL-6 stimulation the ratio of M1 to M2 ATMs shifted towards M2 irrespective of IL-4Ra deficiency, indicating that IL-6 can stimulate M2 polarization without involvement of the IL-4Ra axis (Figures 4A-D).Moreover, we investigated the influence of IL-6 stimulation on ATM proliferation.In our explant model, we could only detect a trend towards higher Ki67 expressing ATMs after IL-6 stimulation, which can probably be explained by intrinsically high IL-6 levels ( Figures 4E, F) as discussed elsewhere (9).Of note, IL-6 treatment shows significant reductions in proliferative M1 macrophages (Figures 4G) and a trend towards a more proliferative M2 ATM phenotype, indicating preferential M2 polarization of proliferating ATMs (Figures 4G, H).

IL-6 signaling boosts CD206 expression partially independent of the IL-4Ra subunit
To get further insights into the IL-6-induced macrophage phenotype and to study the role of the IL-4Ra in IL-6 signaling in more detail, we performed gene expression analysis and bulk RNA sequencing of Il4ra +/+ and Il4ra -/-BMDMs stimulated with IL-13 and IL-6 (Figures 5A, B).Alternative activation and, therefore, an anti-inflammatory M2 phenotype of macrophages is closely linked to the expression of macrophage mannose receptor 1 (CD206) encoded by the Mrc1 gene.In contrast, the proinflammatory M1 phenotype is associated to CD11c expression on macrophages, which is encoded by the Itgax gene.By IL-13 stimulation of BMDMs from Il4ra +/+ and Il4ra -/-mice, we were able to generate conventionally M2-polarized BMDMs and to verify the Il4ra knockout by abrogated IL-13 signaling (Figures 5A, B).Interestingly, after stimulation with IL-6 a similar enhancement of Mrc1 gene expression was seen in both, control as well as Il4ra knockout mice (Figure 5A).qRT-PCR also revealed a decrease of Itgax expression in control and Il4ra -/-BMDMs after IL-6 stimulation (Figure 5B).IL-6 is known to increase IL-10 production and alternative activation of macrophages was described to be closely linked to signaling via IL-10 and IL-10Ra (22).Therefore, we tested whether the IL-6-dependent increase of Il10 relies on IL-4Ra signaling by stimulation of Il4ra -/-BMDMs with IL-6, which revealed a similar increase in Il4ra -/-BMDMs (Figures 5C, D).
To gain further insights into macrophage gene expression patterns after IL-13 or IL-6 stimulation and the dependence on IL-4Ra signaling, we performed bulk RNA sequencing analysis of BMDMs derived from Il4ra -/-and control mice (the 100 most differentially expressed genes are shown in Supplementary Figure 5).In Figure 5E, log 2 fold changes of significantly regulated genes of interest are shown and allocated to the M2 or M1 phenotype.IL-13 stimulation of BMDMs, derived from wildtype mice, induces M2-like as well as M1-like gene expression (Figure 5E, RNA sequencing data accession PRJNA971096), as shown in our previous study (11).Only very few IL-13-mediated effects are detectable in Il4ra -/-BMDMs and may be explained by other receptors (e.g.IL-13Ra2; Supplementary Figure 5).In line with our previous experiments, IL-6 treatment of BMDMs causes enhanced expression of Mrc1 in both wildtype and Il4ra knockout mice to a similar extent (Figure 5E, RNA sequencing data accession PRJNA971096).In contrast to OC-AT and qRT-PCR findings, we cannot confirm the decrease of Itgax expression after IL-6 stimulation of Il4ra -/-BMDMs, which may be due to higher discovery power of isotype forms using RNA sequencing.However, the increase of Il10 expression by IL-6 was even more pronounced in Il4ra -/-BMDMs (Figure 5E).

Discussion
AT of obese individuals exhibits a chronic low-inflammatory state, linked to the onset of type 2 diabetes and other comorbidities.This chronic low-grade inflammation is characterized by an altered composition of ATM phenotypes, an augmented ATM number, and local proliferation of ATMs (9,23).Interestingly, local proliferation of macrophages is preferentially found in antiinflammatory macrophages and seems necessary for maintaining tissue homeostasis and refreshing the tissue microenvironment (23).Of note, local proliferation of ATMs can be linked to IL-6 signaling as shown before in ex vivo AT organ culture studies (9).In subjects with overweight and obesity, circulating levels of IL-6 are found to be augmented (24).Therefore, IL-6 was thought to be a driver of AT inflammation and associated with the activation of ATMs towards the pro-inflammatory M1 phenotype.Most importantly, Mauer and colleagues found evidence for an antiinflammatory role of IL-6 in AT of obese mice.They reported a decreased number of alternatively activated macrophages in obese AT and a deterioration of insulin sensitivity in mice lacking the IL-6Ra (17).These findings are in line with recent studies suggesting anti-inflammatory effects of IL-6 in other diseases, like cancer and neuro-inflammation (25,26).Here, our group investigated the correlation of ATM proliferation and polarization with IL-6 signaling in vivo using myeloid cell-specific IL-6Ra-deficient adult mice after 20 weeks of HFD.We found no significant differences in hallmarks of obesity related AT inflammation analyzing Il6ra Dmyel and Il6ra fl/fl mice using AT histology to examine CLS formation, macrophage distribution or numbers as well as adipocyte size, which is in contrast to previous studies (17).This may be explained by different study protocols, e.g.divergent ages of mice (23 weeks vs. 28 weeks), HFD protocols (starting week 3 of age vs. starting at week 8 of age), the circadian rhythm of mice or different animal facilities, which impacts on microbiota (27)(28)(29)(30)(31).
Analyzing ATM subsets from AT of obese Il6ra Dmyel and Il6ra fl/fl mice reveals a reduced number of CD206-positive ATMs in knockout mice, which matches our previous reports approving beneficial effects of IL-6 signaling on alternative ATM activation (9).Additionally, the subset of CD11c+CD206+ ATMs, which is a marker of ongoing AT inflammation and correlates to insulin resistance, was decreased in AT of Il6ra Dmyel mice (32).
Noteworthy, the M1/M2 classification is an oversimplification based on cell culture work and ATMs show quite diverse phenotypes and physiological characteristics (33).Beside the M1/ M2 paradigm, ATMs are also grouped to metabolically active ATMs, CD9+ ATMs, TREM2+ lipid associated macrophages (LAM) and sympathetic neuron-associated macrophages (SAM) (34-37).Whether occurrence and/or function of either of these ATM subsets is affected by IL-6 signaling needs to be studied in more detail in the future, e.g. by using single cell RNA sequencing.
Stimulation of AT explants with IL-13, a typical Th2 cytokine, leads to an augmentation of macrophages with CD11c and CD206 expression as shown before (11).In contrast, IL-6 treatment of BMDMs shows anti-inflammatory effects by enhancing Mrc1 and Il10 expression and lowering Itgax gene expression.In this study, we detected lower levels of Ki67 expression in obese Il6ra Dmyel mice, which confirms pro-proliferative effects of IL-6Ra signaling on macrophages.Additionally, IL-6 treatment of AT explants with ex vivo generated AT inflammation shows a trend to more proliferative macrophages, which seems to boost antiinflammatory M2 macrophage proliferation and significantly lowers M1 renewal.Notably, we were unable to support this result by our BrdU incorporation protocol, maybe due to limitation of BrdU incorporation to active DNA replication.Of note, BrdU incorporation preferentially labels proliferating cells in S phase of the cell cycle (38), whereas Ki67 expression depicts cells also during G2 and M phase (39).
In our previous study, we speculated that anti-inflammatory polarization and proliferation of ATMs stimulated by IL-6 is dependent on the IL-4Ra (9), since the activation of the IL-4Ra subunit by IL-4 and IL-13 is the most common pathway for alternative macrophage activation (40)(41)(42).Therefore, we tested IL-6 signaling in ATMs of an AT inflammation tissue inflammation model as well as in BMDMs of mice with a global knockout of this IL-4/IL-13 receptor subunit.Surprisingly, IL-6 stimulation lowers the percentage of the M1 subpopulation in ATMs and increases the number of M2 macrophages due to stimulation of M2 proliferation.These effects can also be observed in mice lacking the IL-4Ra, whereat it should be mentioned that IL-6-driven macrophage polarization strongly depends on the polarization state prior to IL-6 stimulation (43).Additionally, our ex vivo AT inflammation model cannot fully imitate obesity-associated AT inflammation e.g.due to missing recruitment of leukocytes from the bloodstream.But also IL-6 treatment of BMDMs leads to the enhancement of Mrc1 and Il10 expression independently of the IL-4Ra.Taken together this study describes a new mechanism for alternative activation of macrophages by IL-6 independently of the IL-4Ra-Stat6-axis.
RNA sequencing data revealed IL-4Ra-independent changes in RNA levels of pro-and anti-inflammatory macrophage genes after IL-6 stimulation of wildtype versus Il4ra -/-mice.Consistent with the finding of Mauer et al., IL-6 stimulation of BMDMs induces the expression of the Il4ra which was confirmed by qRT-PCR (17).Of note, IL-6-mediated effects on alternative activation of macrophages can be enhanced by treating BMDMs that were pre-stimulated with IL-4 and IL-13 (17,43).In our model of diet-induced obesity, a substantial proportion of ATMs is M1 polarized.However, since effects of IL-6 signaling might depend on M2 ATM occurrence, pathophysiological effects of IL-6Ra depletion, such as changes in glucose tolerance, might depend on the duration of pre-existing AT inflammation.Hence, a shorter HFD, as Mauer and colleagues utilized in their study, could imply a greater M2 population as mediator of disrupted IL-6 signaling.Further, we are also unable to exclude pro-inflammatory effects of a soluble mannose receptor variant (25).In addition, the source of IL-6 secretion seems to play an important role for its action, whereby adipocyte-derived IL-6 triggers macrophage recruitment.In contrast, IL-6 secretion by myeloid cells suppresses macrophage recruitment from the blood stream in obese AT (12).Thus, a mouse model with a conditional cell-specific overactivation of IL-6 signaling in macrophages could be a promising model to further investigate the role of IL-6 in obesity.Nevertheless, in this study myeloid cell-specific deficiency of IL-6Ra in obese mice leads to a decline in ATM proliferation and CD206 protein expression in ATMs indicating anti-inflammatory effects of IL-6 signaling on ATMs.The participation of other myeloid cells, such as neutrophils, in long-term HFD-induced AT inflammation seems unlikely due to low quantity (~1% of stroma cells) but cannot be excluded due to LysM-driven recombination in neutrophils (26).Since CD206 is also described to mediate phagocytosis (44, 45), augmentation of CD206 on ATMs could be helpful in the resolution of dying adipocytes and AT inflammation.Owing to IL-6-mediated effects on ATM proliferation, polarization and recruitment, IL-6 and mannose receptor 1 expression sustain potential targets to rescue insulin sensitivity in obesity and the role of IL-6 as a pro-inflammatory clinical parameter should be revised.

Experimental animals
Mice strains were maintained in pathogen-free facilities at the University of Leipzig on a 12-h light/dark cycle at 22 ± 2°C with free access to food and water.For diet-induced obesity, Il6ra fl/fl x LysM-Cre -/+ (Il6ra Dmyel ) or wildtype (Il6ra fl/fl ) littermate controls on a 57BL/6N background were fed a high-fat diet (HFD) (60% kcal fat; Ssniff Spezialdiäten) for 20 weeks starting at the age of 8 weeks.Control littermates as well as Il4ra +/+ and Il4ra -/-mice were kept on a regular chow diet (9% kcal fat; Ssniff Spezialdiäten; Soest, Germany).

Induction of ex vivo AT inflammation
For mimicking AT inflammation as seen in obesity, organotypic organ-cultures of lean visceral AT were generated as described before (9).Explants were placed under a sterile cell culture insert (pore size 0.4 µm; Sarstedt, Nümbrecht, Germany) and cultured in RPMI 1640 medium (Sigma-Aldrich, Deisenhofen, Germany) supplemented with 1% insulin/transferrin/selenium mixture, antibiotics (100 U/ml penicillin and streptomycin; all reagents from Sigma-Aldrich) and 10% fetal bovine serum at 5% CO 2 , 21% O 2 and 37°C for 7 days without any intervention.Subsequently, AT explants were treated with PBS, IL-6 or IL-13 (50 ng/ml) for 48h.

Metabolic characterization
Mice were weighed weekly, starting at 5 weeks of age until euthanization.At ages of 8 weeks (before starting HFD feeding) and 28 weeks (after 20 weeks of HFD) intraperitoneal insulin (ipITT) and glucose tolerance tests (ipGTT) were performed.For ipITT of lean mice, baseline glucose levels were measured before mice are injected intraperitoneally with insulin (Insuman Rapid, 100 IU/ml) in a concentration of 0,75 U per kg bodyweight.Blood glucose was measured again at 15, 30 and 60 min after injection.ipGTT of lean mice was performed 3 days after ipITT.Again baseline glucose levels were analyzed before mice were injected intraperitoneally with 2 g glucose per kg bodyweight (20%; B. Braun, Melsungen, Germany).Glucose levels were measured again at 15, 30, 60 and 120 min after injection.After HFD mice received for ipITT 1,5 U insulin per kg bodyweight because of a lower response to insulin.For ipGTT 1 g glucose per kg bodyweight was administered to avoid the possibility of prolonged hyperglycemia.Food intake was measured for 5 days in lean male and female Il6ra fl/fl and Il6ra Dmyel mice.

Analytical procedures
Free fatty acids (FFA), triglycerides, total cholesterol, lowdensity lipoprotein-(LDL) and high-density lipoprotein-(HDL) cholesterol in plasma were determined by an automatic chemical analyzer at the Institute of Laboratory Medicine and Clinical Chemistry at the University of Leipzig.

RNA isolation and quantitative real-time PCR analysis
RNA was extracted using TRI Reagent solution (Thermo Fischer Scientific; 15596018) and reverse transcribed into cDNA with RevertAid H Minus Reverse Transcriptase (Thermo Fischer Scientific; EP0451).mRNA expression of genes was measured on an Applied Biosystems StepOnePlus Real-Time PCR-Cycler (Applied Biosystems; Waltham, MA, USA) with Hot FirePol EvaGreen qPCR Mix Plus (ROX) (Biotium Iinc.; Hayward, CA, USA; 31077).Relative gene expression was adjusted to Ipo8 and calculated according to DDCt method by Pfaffl (33).Primers for relative gene expression analysis are given in Supplementary Table 2.

RNA sequencing
BMDMs were treated with IL-6, IL-13 or PBS for 48h.Afterwards 500.000-1.000.000 cells per condition were harvested and stored at −80 °C in TRIzol (Thermo Fischer Scientific) until

1 Myeloid
FIGURE 1 Myeloid Il6ra knockout does not affect weight gain and disturbed metabolism of protracted diet-induced obesity.(A) Relative gene expression analysis of Il6ra fl/fl and Il6ra Dmyel BMDMs to verify the Il6ra knockout (n=7-8).(B) Comparison of body weight in either lean female, lean male and obese male Il6ra fl/fl and Il6ra Dmyel mice.(C) Organ weights of different fat depots, liver and pancreas in obese male Il6ra fl/fl and Il6ra Dmyel after 20 weeks of HFD (n=9-16).Blood glucose measurements by ipITT (D) and ipGTT (E) of Il6ra fl/fl and Il6ra Dmyel male mice after 20 weeks of HFD (n=17).Data are presented as mean ± SEM. ***p < 0.001.

2
FIGURE 2 Adipocyte expansion and ATM distribution are unaltered in obese mice lacking IL-6Ra.(A) Representative images of immunohistochemistry of paraffin-embedded AT from obese (left side) and lean (right side) Il6ra fl/fl and Il6ra Dmyel mice stained with Perilipin (adipocytes, green), pan macrophage marker Mac-2 (red) and DAPI (nuclei, blue).(B) Abundance of CLS in % of adipocyte number, (C) interstitial macrophages in % of adipocyte number and (D) measurements of adipocyte size in µm from paraffin-embedded AT sections of obese Il6ra Dmyel and Il6ra fl/fl controls (n=6-7).(E, F) Distribution of adipocytes concerning adipocyte size in (E) lean and (F) obese male Il6ra Dmyel and Il6ra fl/fl mice.Data are presented as mean ± SEM. *p < 0.05.Scale bar represents 100 µm.