Adipose-derived Mesenchymal Stromal Cells Modulate Lipid Metabolism and Lipid Droplet Biogenesis via AKT/mTOR –PPARγ Signalling in Macrophages

Mesenchymal stromal cells (MSCs) are a potential therapy for many chronic inflammatory diseases due to their regenerative, immunologic and anti-inflammatory properties. The two-way dialogue between MSCs and macrophages is crucial to tissue regeneration and repair. Previous research demonstrated that murine adipose-derived MSC conditioned medium (ASCcm) reprograms macrophages to an M2-like phenotype which protects from experimental colitis and sepsis. Here, our focus was to determine the molecular mechanism of lipid droplet biogenesis in macrophages re-educated using ASCcm. Adipose-derived MSC conditioned medium promotes phosphorylation of AKT/mTOR pathway proteins in macrophages. Furthermore, increased expression of PPARγ, lipid droplet biogenesis and PGE2 synthesis were observed in M2-like phenotype macrophages (high expression of arginase 1 and elevated IL-10). Treatment with mTOR inhibitor rapamycin or PPARγ inhibitor GW9662 suppressed lipid droplets and PGE2 secretion. However, these inhibitors had no effect on arginase-1 expression. Rapamycin, but not GW9662, inhibit IL-10 secretion. In conclusion, we demonstrate major effects of ASCcm to reprogram macrophage immunometabolism through mTOR and PPARγ dependent and independent pathways.

functional states 15 . In macrophages activated with LPS (M1), aerobic glycolysis generates energy and is mediated by mTOR/HIF-1α. In contrast, M2 macrophages exhibit mitochondrial metabolism and fatty acid oxidation mediated by mTOR/PPARγ [16][17][18] . These results indicate that the transition between either glycolytic metabolism or oxidative phosphorylation plays a role in macrophage polarization. Interesting, Vasandan et al. showed that MSC-secreted PGE 2 induce macrophage polarization by modulating the metabolic program 19 .
Nonetheless, few studies have addressed the modulation of lipid droplets in reprogrammed macrophages. Lipids are an essential energy source, critical structural components of cellular membranes and important to cellular signaling. Lipid droplets are lipid-enriched, multifunctional organelles that have multiple roles in physiology and pathological conditions including regulating lipid metabolism and energy homeostasis 20,21 . Indeed, lipid droplets act as essential platforms for immunometabolic regulation, including lipid storage and metabolism, inflammatory lipid mediator production, and signaling pathway compartmentalization. In addition to their function in fat tissues, lipid droplets increase in number and size in cells involved in inflammatory processes. Under different stimulatory conditions, e.g. sterile-or infection-driven inflammation, immune cells like macrophages, neutrophil, and eosinophils exhibit increased lipid droplet biogenesis [22][23][24] . This process has been associated with increased synthesis and the secretion of inflammatory lipid mediators, such as prostanoids and leukotrienes 25 .
Given the relevance of metabolic, morphological and biochemical alterations of macrophage during the pathophysiology of inflammation response, our hypothesis is that macrophage lipid droplet formation is tightly regulated by ASC; and lipid droplets have key roles in immunometabolic regulations in macrophages. The overall goal of our study was to determine the molecular mechanism of lipid droplet biogenesis and the release of inflammatory lipid mediator in re-educated macrophages by ASC. Here, the mechanisms of lipid droplet biogenesis in M2-like macrophage was demonstrated for the first time.

Methods
Animals, cell culture and treatment. C57BL/6 and Balb/c mice (10-15 weeks) were obtained from the Institute of Science and Technology in Biomodels/FIOCRUZ. All animal procedures were approved and performed following the guidelines of the Brazilian Council for care and use of experimental animals (CONCEA). Adipose-derived MSCs (ASCs) were obtained, as previously described 10 . Briefly, perigonadal and inguinal adipose tissues from female Balb/c mice were mechanically and chemically dissociated using a type 1 collagenase 2 mg/mL for 1 h at 37 °C. The cell suspension was filtered using a 100 μm strainer followed by a second filtration step using a 40 μm cell strainer. Cells were centrifuged and resuspended in Mesencult medium (StemCell Technologic). ASCs were maintained at 37 °C in a humidified hypoxic atmosphere containing 5% O 2 , and the conditioned medium was collected between passage 2 and 4. ASC were characterized by morphology and positive expression of CD29 and negative expression of CD45 flow cytometry at passage 3 (Fig. S1). After reaching 80% of confluency, cells were split into 3 T75 flasks and the conditioned medium (ASCcm) was collected after three days in the culture conditions described above. To obtain bone marrow murine macrophages, cells isolated from femur and tibia of male C57Bl/6 mice were cultured using RPMI1640 medium containing 20% heat-inactivated FBS and 30% L929 cell-conditioned medium. Differentiation was performed at 37 °C in a humidified 5% CO 2 incubator. After seven days, adherent macrophages were harvested and seeded for assays. These macrophages were treated with 50% of ASCcm with the presence or absence of Rapamycin (20 nM) or GW9662 (10 nM) as indicated. Control macrophage (CTRL-MΦ) were treated with 50% of Mesencult medium. The ASC and macrophage culture medium contained 100 U/mL penicillin and 100 µg/mL streptomycin. enzymatic assay. To perform the arginase enzymatic activity assay, after treatment with ASCcm, macrophage cell lysate was collected by using lysis buffer (0.5% Triton-X100, 10 mM Tris, 0.1 mM EDTA, 150 mM NaCl, and 10% glycerol) containing protease inhibitor cocktail (Roche). MnCl 2 , at a final concentration of 1 mM, was added to the lysate; and this solution was heated for 10 min at 56 °C to activate the enzyme. Arginine hydrolysis was conducted by incubating this suspension with 100 μl of 500 mM of L-arginine (pH 9.7) at 37 °C for 40 min. The reaction was stopped using 900 μl of H 2 SO 4 (96%)/H 3 PO 4 (85%)/H 2 O (1:3:7, v/v/v) followed by the addition of 40 µL of 9% α-isonitrosopropiophenone (Sigma Aldrich -I3502) that had been dissolved in 100% ethanol. The final solution was incubated for 20 min at 100 °C. Serial dilutions of urea solution were used as the standard. The absorbance was measured at 540 nM to determine the urea concentration in the cell lysate. One unit of arginase activity (U) is defined as the amount of enzyme that catalyzes the formation of 1 µM of urea per min. Protein dosage was determined by Bradford assay and the result was normalized for total protein (mU/µg protein).
Lipid droplets and immunofluorescence staining. Macrophages were seeded in coverslips and allowed to adhere for 24 h. After treatments, cells were fixed using 3.7% formaldehyde and the lipid droplets were stained with 0.3% Oil Red O (diluted in isopropanol 60%) for 2 min at room temperature. For immunofluorescence staining, after a fixing step, cells were rinsed three times with PBS containing 0.1 M CaCl 2 and 1 M MgCl 2 (PBS/CM) and then permeabilized with 0.1% Triton X-100 plus 0.2% BSA in PBS/CM for 10 min (PBS/ CM/TB). Cells were stained using rabbit polyclonal anti-PLIN-2 antibody (Proteintech, #15294-1-AP) at 1:1000 dilution for 2 h, followed by a rabbit anti-IgG-Dylight 550 at 1:1000 dilution for 1 h. Lipid droplets were stained with BODIPY493/503 dye (dilution 1:5000 in water) for 5 min. Macrophages were stained with Alexa Fluor ™ 594 Phalloidin for 30 min at room temperature (Invitrogen ™ ) for cell morphology analysis. The coverslips were mounted in slides using an antifade mounting medium (VECTASHIELD ® ). Nuclear recognition was based on DAPI staining (1 µg/mL) for 10 min. Fluorescence was analyzed by fluorescence microscopy with an x100 objective lens (Olympus, Tokyo, Japan) or Confocal Microscopy (Laser scanning microscopy 510 Meta, Zeiss). The numbers of lipid droplets were automatically quantified by ImageJ software analysis or manually counted by light microscopy in 50 consecutive cells.
IL-10 and PGE 2 measure. After 24 h of treatment with ASCcm, M2-like macrophages were rinsed twice with PBS and incubated in fresh medium containing 500 ng/mL of LPS (O127: B8, Sigma) plus 10 ng/mL of IFNy (PeproTech) for 24 h. The IL-10 was quantified by ELISA (murine IL-10 Duo Set, R&D systems) and PGE 2 by EIA kit (Cayman Chemicals), following manufacturers' recommended protocols.
Statistical analysis. Data are expressed as mean ± standard error of the mean (SEM) of three to six independent experiments. The normal distribution was performed using the Shapiro-Wilk test. The paired two-tailed t-test was used to evaluate the significance of the two groups. Multiple comparisons among three or more groups were performed by one-way ANOVA followed by Tukey's multiple comparison test. p values < 0.05 were considered statistically significant when compared to control non-stimulated group or ASCcm treated group.

Results
ASccm polarize and promotes lipid droplet biogenesis in macrophage. Studies from the past suggest that modulation of monocyte/macrophage is the hallmark of MSC therapeutic effects in vivo experimental model. Our previous research demonstrated that ASCcm promotes macrophage differentiation towards an M2-like phenotype, which shows high therapeutic capacity in colitis and sepsis experimental models 10 . In order to explore the modulation of lipid mediator on this reprogrammed macrophage, we first confirmed the classical aspects of alternative macrophage phenotype. ASCcm induces high expression and activity of arginase-1 enzyme, a well-known marker for M2/M2-like macrophages (Fig. 1A,B). After stimulation, enhancement in IL-10 production was observed in ASCcm-reprogrammed macrophage (ASC-MΦ) unlike the control macrophage (CTRL-MΦ, non-polarized) (Fig. 1C). In addition, ASC-MΦ showed a distinguished fusiform morphology of cytoplasmic shape compared to CTRL-MΦ, (Fig. 1D). Previously, it was described that cell elongation is a characteristic alteration in an alternative macrophage 26 . These results confirmed the ASCcm programs macrophage toward an alternative M2-like phenotype.
It has been well established that LPS induces M1 macrophages and promotes lipid droplet biogenesis. Here, we described for the first time a remarkable induction of lipid droplets in M2-like macrophages. Those organelles were properly identified by the expression of Plin2 (Fig. 1E). Our findings indicate that ASCcm leads to an increased content of lipid droplets within macrophage's cytoplasm that appeared as red fluorescent dots in the image (Fig. 1F,G). Lipid droplets are cytoplasmic organelles that compartmentalize PGE 2 synthesis machinery and final products. In accordance, ASCcm treatment results in the elevated secretion of PGE 2 by ASC-MΦ (Fig. 1G). Therefore, we evaluated the capability of ASCcm to modulate the expression of crucial enzymes for eicosanoids production. Macrophage treated with ASCcm showed a significant increase in COX-2 and cPLA2-α expression when compared to CTRL-MΦ (Fig. 1H,I). In summary, ASCcm seems to modulate lipid metabolism in macrophages. mtoR/ppARγ pathway did not affect macrophage polarization induced by ASCcm. Next, we investigated the mechanisms underlying macrophage polarization in macrophages treated with ASCcm. The mTOR pathway modulates different cells functions, such as survival, proliferation, protein, and lipid synthesis. Further, the regulation of mTOR was showed to be crucial to reprogramming macrophages 27 . The activation of the AKT/mTOR pathway was investigated in different time points after ASCcm treatment. It was concluded that detection of phosphorylated forms of AKT and mTOR was time-dependent. This change was apparent within 15 minutes after the addition of ASCcm and had decreased at 24 hours ( Fig. 2A,B). On the other hand, an increase in the arginase-1 expression was only detected at later time points. In addition, our results showed an ASCcm induce expression of PPARγ in comparison to CTRL-MΦ (Fig. 2C). PPARγ is a transcription factor regulated by the mTOR pathway, and it regulates lipid metabolism. Therefore, we used rapamycin, an mTOR inhibitor, to evaluate downstream proteins of this pathway.
In accordance with the activation of mTOR, the phosphorylation of p70S6k and 4EBP1 was detected after ASCcm treatment. Thus, the rapamycin blocked these effects (Fig. 3A). In addition, the PPARγ expression induced by ASCcm was also significantly inhibited by rapamycin (Fig. 3B). It opened up a possibility of AKT/ mTOR/PPARγ pathway to be involved in both lipid droplet biogenesis and macrophage plasticity induced by ASCcm. However, rapamycin and GW9662 (a PPARγ inhibitor) did not affect the macrophage arginase-1 expression and fusiform morphology induced by ASCcm (Fig. 3C-E). Interestingly, rapamycin, but not GW9662, blocked the IL-10 secreted by ASC-MΦ (Fig. 4). www.nature.com/scientificreports www.nature.com/scientificreports/ Lipid droplet biogenesis induced by ASccm are regulated by mtoR/ppARγ pathway. Lipid droplets are dynamic organelles controlled by molecular mechanisms that are cell-and stimulus-dependent. Rapamycin and GW9662 treatment were able to reduce lipid droplets induced by ASCcm in macrophage. These effects were showed by the analysis of lipid droplet numbers and the stained area (Fig. 5A,B). Therefore, the www.nature.com/scientificreports www.nature.com/scientificreports/ production of PGE 2 was evaluated as a functional marker of these lipid droplets. As expected, rapamycin and GW9662 had a significant impact on PGE 2 production. This lipid mediator was significantly reduced after inhibition of mTOR and PPARγ protein (Fig. 5C). However, rapamycin treatment, but not GW9662, inhibited COX-2 expression in macrophage treated with ASCcm (Fig. 6A,B). The above observations support the mTOR/PPARγ pathway as a molecular mechanism for lipid droplet biogenesis induced by ASCcm.

Discussion
Recent advances in understanding the immunomodulatory properties of MSC have provided new insights into their benefits effects on modulating monocytes/macrophage 28 . Thus, it is well-known macrophage plasticity is a crucial component to tissues repair and regeneration. The present study was designed to focus on macrophage polarization and lipid metabolism regulated by ASCcm in vitro. We demonstrated that ASCcm induces macrophage polarization to M2-like phenotype accompanied by an increased amount of lipid droplets (Fig. 1). Previously, our research showed that ASCcm converted macrophages into a distinctive phenotype that shares markers with alternatively activated macrophages (M2/M2-like) and regulatory macrophages (Mreg). It includes high expression of arginase I and II, HO-1 and LIGHT and production of nitric oxide and IL-10 upon stimulation, but lack of classical markers of M2 macrophages (ie, Ym1, CCL22 or Fizz1) 10 . Here we demonstrate major effects of ASCcm to reprogram macrophage immunometabolism, leading to increased lipid droplet biogenesis and PGE 2 production through mTOR and PPARγ dependent pathways (Fig. 7).
It was demonstrated that Balb/c ASC modulates macrophages derived from Balb/c and C57Bl6 and these macrophages have therapeutic effects in two models of inflammatory bowel diseases: 1) TNBS model was performed in Balb/c and 2) DSS model in C57BL/6 10 . In our present study, all the experiments were performed using macrophages derived from bone marrow isolated from male C57Bl/6. Different studies have demonstrated that mesenchymal stem cells from different sources (i.e: tissues, strain) induces macrophage polarization. To better characterize and validate our findings, we performed an experiment using ASCcm isolated from Balb/c and C57BL/6 to compare the induction of lipid droplet biogenesis. ASCcm obtained from Balb/c or C57BL/6 similarly induce increase lipid droplet formation in macrophages (Fig. S2).
The experiments were performed using mesenchymal stem cell obtained from female because they accumulate more adipose tissue than male mice. As such, it allowed us to obtain a reasonable amount of ASC using fewer mice. Although it is clear that there are many genes expressed in a sexually dimorphic manner, the impact www.nature.com/scientificreports www.nature.com/scientificreports/ of sex in lipid metabolism is explained by multiple factors, including hormones (catecholamine, estrogens, and progesterone), nutritional requirements lipid metabolism in the liver, and systemic metabolism. Therefore, to limit the impact of sex differences on our studies we focused on an in vitro model to study the lipid droplet and macrophage polarization. In addition, females were always used to obtain ASC while male mice were used to obtain bone marrow cells.
There is strong evidence that metabolic changes in distinct immune cells play a crucial role during the host immune response against invaders. The paradigm activation of M1 and M2 macrophages have been correlated with a noticeable difference in their metabolic status, by which classically macrophage activation (M1, i.e. activated by LPS plus IFNγ) undergo glycolytic status. In contrast, alternative activation phenotype (M2/Mreg, i.e. activated by IL4 or IL-10) requires mitochondrial β-oxidation of fatty acids 29 . Nonetheless, in vivo, it is not clear whether the metabolic and cellular switch is due to repolarization of M1 to M2 (and vice versa) or dependent Lipid-enriched foam cell macrophages are a hallmark of atherosclerosis. Recently, it was shown that MpOx-LDLs induced lipid droplet accumulation and enhanced anti-inflammatory phenotype in macrophage. MpOX-treated macrophages are more efficient removing tissue debris and apoptotic bodies than M1 and M2 macrophages. Thus, contributes to preventing uncontrolled tissue damage in the atherosclerotic milieu 30 . In this context, MSC inhibits foam cell formation by reducing the uptake of ox-LDL. Furthermore, it decreases dyslipidemia and atherosclerotic lesions size in mice experimental model 31,32 . Supporting this idea, Erpicum et al. suggested that administration of MSC prior to renal ischemia/reperfusion attenuates the injury by downregulating fatty acid biosynthesis in kidneys 33 . In contrast, our data showed the unmistakable M2-like polarization induced by adipose-derived MSC is accompanied by increased expression of lipid droplet biogenesis (Fig. 1). Distinct source of MSC could explain the discrepancies between the studies, as bone marrow-derived MSC were used in renal injury and atherosclerosis model. Moreover, we speculate that different sources of MSC (i.e bone marrow, umbilical cord, placenta, etc) as well the source of macrophage (i.e: Kupfer cell, microglia, peritoneal resident macrophage, tumor-associated macrophages, etc) could also impact the modulation of lipid droplets.
Recently, Flaherty et al. described in an elegant way that adipose lipid-filled extracellular vesicles (AdEXos) deliver large amounts of triacylglyceride to macrophages. Further, AdExos modulates macrophage differentiation in bone marrow 34 . However, the lipid accumulation is distinct from lipid droplets seen in foam cells. Here we demonstrate the ASCcm induces lipid droplet (confirmed by Plin2 staining; Fig. 1E) and that this event is mediated by mTOR/PPARγ pathway. mTOR pathway is a nutrient sensor mechanism that regulates cell growth, survival, and metabolic homeostasis. In our experiment, we used 50% of ASCcm and 50% of fresh medium containing serum, to assure macrophage are not under starvation due to possible ASCcm nutrient depletion during conditioning. The control cell (CTRL-MΦ) received 50% of Mesencult media instead of the ASCcm and 50% of fresh medium containing serum. Moreover, conditioned media from L-929 cells (mouse fibroblast-derived from subcutaneous adipose tissue) did not show the same effects as ASCcm (data not shown), suggesting that the effects observed were due to soluble factor secreted by ASC.
Activation of mTOR, a serine/threonine kinase, induces the formation of two distinct protein complexes, known as mTOR Complex 1 (mTORC1) and 2 (mTORC2). Cellular processes downstream of mTORC1 include protein, nucleotide, fatty acid, and lipid synthesis 35 . We demonstrated that ASCcm direct affects the mTORC1 signaling pathway in macrophages by phosphorylation of mTOR and p70S6K (Fig. 2). It was confirmed by rapamycin inhibition of P706K and 4EBP1 phosphorylation, two protein downstream mTORC1 complex (Fig. 3A).
Although we did not have deeply investigated the mTORC2 pathway, our results suggest this pathway could be involved in ASCcm effects over macrophage polarization. First, ASCcm promotes phosphorylation of AKT at Ser473 residue (Fig. 2). The two phosphorylation site of AKT(T308 and S473) are described as functional divisor between mTORC1 and mTORC2 complex 36 . In addition, Rapamycin treatment did not affect arginase-1 expression induced by ASCcm. mTORC2 plays an essential role in M2 macrophages polarization induced by parasitic helminths 27 and is critical to the metabolic switch induced by IL-4 37 . Our results provide evidence that mTORC1 and mTORC2 cooperate to M2-like macrophage phenotype induced by ASCcm. IL-10 induced by ASCcm was www.nature.com/scientificreports www.nature.com/scientificreports/ significantly inhibited by rapamycin treatment. This is consistent with the previous report that impaired mTORC1 signals attenuate IL-10 in dendritic cells 38 .
Furthermore, our results showed that ASCcm induces PPARγ expression in macrophages. The activation of AKT/mTOR/PPARγ regulates lipid synthesis and lipid uptake in adipocytes 39,40 . Lipophosphoglycan, a major Leishmania surface glycoconjugate, triggers activation of PPARγ signaling, which leads a rising lipid droplet biogenesis, COX-2, and PGE 2 production in macrophages 41 . PPARγ is a nuclear receptor associated with glucose and fatty acid metabolism and participates in immune regulation. Rapamycin treatment inhibits partially PPARγ induced by ASCcm and affects the lipid droplet biogenesis (Fig. 5). In accordance, rapamycin treatment reduced fatty acid uptake of adipose tissue explants in vitro 39 . The mechanism by which mTOR regulates PPARγ is not   www.nature.com/scientificreports www.nature.com/scientificreports/ completely understood. However, the mTORC1 complex could play a critical role in this pathway, as mTORC2 is less sensitive to rapamycin 35 .
PPARγ regulates COX-2 expression and, consequently PGE 2 production by PPARγ response element (PPRE) present in the COX-2 gene promoter 42 . However, in our study, PPARγ inhibitor decreases lipid droplet biogenesis and PGE 2 secretion. But it did not affect COX-2 expression in macrophage re-educated by ASCcm (Fig. 6). This would suggest that PPARγ could affect the expression or the compartmentalization of different enzymes involved in the synthesis and release of inflammatory lipid mediator.
In summary, our data indicate that paracrine factors released by ASC could activate mTORC1 and mTORC2 to stimulate lipid droplet biogenesis and macrophage polarization. However, the specific soluble factors (growth factor, cytokine, lipid mediators) linked to AKT/mTOR activation, macrophage polarization, and lipid droplet biogenesis will be elucidated in future studies. Taken together, our results provide evidence that lipid droplet accumulation could play a dynamic role in macrophage polarization induced by ASCcm. Previous studies showed the effects of human ASC on macrophage polarization 11 . In addition, Zhang et al. showed that human and mouse ASC ameliorate the symptoms of acute lung injury in mice 43 . Therefore, we believe that the lipid droplet biogenesis on macrophage could be regulated by human ASC to modulate the inflammatory response and/or tissue repair. However, it needs further investigation to clarify this hypothesis. Nonetheless, further studies are needed to determine the relationship of lipid droplets during the anti-inflammatory/pro-resolution macrophage reprogramming.