Sterol Derivatives Specifically Increase Anti-Inflammatory Oxylipin Formation in M2-like Macrophages by LXR-Mediated Induction of 15-LOX

The understanding of the role of LXR in the regulation of macrophages during inflammation is emerging. Here, we show that LXR agonist T09 specifically increases 15-LOX abundance in primary human M2 macrophages. In time- and dose-dependent incubations with T09, an increase of 3-fold for ALOX15 and up to 15-fold for 15-LOX-derived oxylipins was observed. In addition, LXR activation has no or moderate effects on the abundance of macrophage marker proteins such as TLR2, TLR4, PPARγ, and IL-1RII, as well as surface markers (CD14, CD86, and CD163). Stimulation of M2-like macrophages with FXR and RXR agonists leads to moderate ALOX15 induction, probably due to side activity on LXR. Finally, desmosterol, 24(S),25-Ep cholesterol and 22(R)-OH cholesterol were identified as potent endogenous LXR ligands leading to an ALOX15 induction. LXR-mediated ALOX15 regulation is a new link between the two lipid mediator classes sterols, and oxylipins, possibly being an important tool in inflammatory regulation through anti-inflammatory oxylipins.


Introduction
Inflammation is a protective mechanism against infection or tissue injury.During the course of inflammation, immune cells, such as neutrophils and monocytes, are recruited in several phases and stimulate the formation and release of different cytokines, chemokines, growth factors, and lipid mediators [1][2][3].In the inflammatory process, monocytes differentiate into macrophages, which have three main tasks: the production of immunomodulators, phagocytosis, and antigen presentation [4,5].
Primary cell culture of monocyte-derived macrophages allows us to investigate the signaling and regulatory pathways of the innate immune response.Under cell culture conditions, macrophage differentiation is mimicked using different stimuli to investigate the regulation of the arachidonic acid cascade.The (simplified) classification of macrophages is characterized by the expression pattern of characteristic markers, including surface markers as well as pro/anti-inflammatory proteins [9].Stimulation with granulocyte/macrophage colony-stimulating factor (GM-CSF) leads to the development of a pro-inflammatory state with increased TNF expression and release.After priming with macrophage colonystimulating factor (M-CSF), the cells exert an anti-inflammatory character with increased IL-10 expression and release.In addition, the polarization of macrophages is induced by stimuli such as cytokines or ligands of the innate immune response, such as bacterial lipopolysaccharide (LPS) [10].Commonly, stimulation of macrophages with IFNγ and LPS is used to generate classically activated macrophages, i.e., M1-like macrophages [9,11].The polarization of alternatively activated macrophages is induced by IL-4, IL-10, or IL-13, i.e., M2-like macrophages [9,12].
During the inflammatory process, the polarization of macrophages can switch from pro-to anti-inflammatory depending on the inflammatory environment [13][14][15].A switch of the macrophage phenotype can be initiated by efferocytosis, i.e., the uptake of apoptotic cells [14], which is associated with nuclear receptors such as retinoid X receptors (RXR), PPAR, or liver X receptors (LXR) [16].Moreover, PPAR is known to be involved in macrophage polarization and inflammatory processes [17].While inactivation of PPAR or RXR impairs efferocytosis of apoptotic cells, LXR-deficient macrophages are unable to clear apoptotic cells altogether [18,19].In addition, LXR also drives efferocytosis indirectly by increasing its own expression in an autoregulatory manner [20].
LXR belongs to the family of nuclear receptors involved in the regulation of metabolic homeostasis and inflammation [21].There are two LXR isoforms that are expressed differently in tissues, despite their high structural similarity (77%).While LXRα is mainly expressed in the liver, kidney, intestine, adipose tissue, or macrophages, LXRβ can be found in all human cells [22,23].LXR target genes include sterol response element binding protein 1c (SREBP1c), apolipoprotein (apo) E, ATP-binding cassette transporters (ABC)A1, ABCG1, ABCG5, cytochrome P-450 7A1 (CYP7A1), and fatty acid synthase (FAS) [24,25].As LXR is involved in both cholesterol and lipid metabolism, it could be an important link in the regulation of inflammation by impacting the formation of anti-inflammatory lipid mediators and thus driving the inflammation towards the resolution phase.
The resolution phase macrophages produce a variety of anti-inflammatory cytokines and chemokines.In addition, the formation of lipid mediators-oxylipins-plays a decisive role in the course of inflammation.High levels of cyclooxygenase (COX)-and 5-lipoxygenase (LOX)-derived products from arachidonic acid (ARA) are well investigated pro-inflammatory oxylipins, e.g., prostaglandin (PG) E2 and D2 or leukotriene (LT) B4, which are predominantly formed and released at the beginning of the inflammation.In contrast, multi-hydroxylated metabolites of different polyunsaturated fatty acids (PUFA) such as ARA, eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) are described to be formed during the resolution phase and have anti-inflammatory properties [26,27].The formation of multi-hydroxylated oxylipins such as 5,15-diHETE and 5,12-diHETE predominantly involves different lipoxygenases (LOX) interacting with each other (Figure 1) and 15-LOX being a key enzyme [28].Although there is doubt about the formation, detectability, and receptors of multi-hydroxylated LOX products, so-called specialized pro-resolving mediators (SPM) and SPM bearing three hydroxy groups are not formed in macrophages [29], the anti-inflammatory effect of 15-LOX is undisputed [30].
Recently, we have shown that stimulation of macrophages with the synthetic LXR agonist T0901317 (T09) increases the expression of anti-inflammatory genes, such as ALOX15 (15-LOX gene) [31].Here, we aim to investigate the LXR-induced influence on the ARA cascade and formation of oxylipins in macrophages, as well as different markers involved in differentiation and polarization.For this, human peripheral blood mononuclear cells (PBMC) left untreated or differentiated into M1-and M2-like macrophages were stimulated with T09, and the formation of oxylipins, protein levels, and presence of surface markers were compared to non-stimulated cells.The specificity of nuclear receptor activation was investigated using FXR agonists and RXR agonists.Finally, sterols were identified as endogenous LXR ligands, elevating 15-LOX abundance and upregulating oxylipin formation.In summary, with the sterol-mediated induction of oxylipins, we demonstrated a new link between these lipid mediator classes in the regulation of M2-like macrophages.Recently, we have shown that stimulation of macrophages with the synthetic agonist T0901317 (T09) increases the expression of anti-inflammatory genes, su ALOX15 (15-LOX gene) [31].Here, we aim to investigate the LXR-induced influen the ARA cascade and formation of oxylipins in macrophages, as well as different ma involved in differentiation and polarization.For this, human peripheral blood mon clear cells (PBMC) left untreated or differentiated into M1-and M2-like macrophages stimulated with T09, and the formation of oxylipins, protein levels, and presence o face markers were compared to non-stimulated cells.The specificity of nuclear rec activation was investigated using FXR agonists and RXR agonists.Finally, sterols identified as endogenous LXR ligands, elevating 15-LOX abundance and upregulatin ylipin formation.In summary, with the sterol-mediated induction of oxylipins demonstrated a new link between these lipid mediator classes in the regulation of M2 macrophages.

Results
In the present study, we investigated the effect of LXR activation by different nu receptor agonists and endogenous lipids on the induction of ARA cascade enzyme oxylipin formation, as well as the overall effects on proteins and markers that are invo in the polarization of human primary macrophages.
PBMC were left untreated or differentiated into M1-or M2-like macrophages stimulated with the synthetic LXR agonist T09 (1 µM) for 3 h (Figure 2A).Oxylipins analyzed using a LC-MS/MS targeted oxylipin metabolomics method.Figure 2B s representative oxylipins derived from ARA.The analysis of the protein levels was ca

Results
In the present study, we investigated the effect of LXR activation by different nuclear receptor agonists and endogenous lipids on the induction of ARA cascade enzymes and oxylipin formation, as well as the overall effects on proteins and markers that are involved in the polarization of human primary macrophages.
PBMC were left untreated or differentiated into M1-or M2-like macrophages and stimulated with the synthetic LXR agonist T09 (1 µM) for 3 h (Figure 2A).Oxylipins were analyzed using a LC-MS/MS targeted oxylipin metabolomics method.Figure 2B shows representative oxylipins derived from ARA.The analysis of the protein levels was carried out by LC-MS/MS based targeted proteomics (Section 1 + Tables S1-S3, Supplementary Material).The effect of T09-mediated LXR activation on ARA cascade enzymes, receptors, and other protein levels specific for macrophages is depicted in Figure 2C.The presence of macrophage surface markers reflecting their polarization was analyzed by immunofluorescence staining (Figure 2D).T09 had no effect on the morphology or confluency of untreated M1-and M2-like macrophages.
The T09-induced 15-LOX abundance and activity were further evaluated in a timeand dose-dependent manner (Figure 3 and Table 1).
Following the evaluation of the optimal incubation conditions, the induction of 15-LOX by other LXR-, FXR-, and RXR-specific compounds as well as cholesterol derivatives (Figure 4) was investigated.The results are shown in Figure 5 and Table S4.Effect of the LXR agonist T09 on the ARA cascade and differentiation of primary macrophages.(A) Primary human blood monocytic cells were treated with 10 ng/mL GM-CSF (M1 type) for 8 days or with M-CSF (M2 type) for 8 days, as well as 10 ng/mL IFNγ (M1 type) or IL-4 (M2 type) for the final 2 days.Additionally, M1 cells were challenged with 100 ng/mL LPS for 6 h.For the untreated macrophages, the adhered monocytes were left untreated for 8 days.The three different macrophage phenotypes were incubated with (blue) or without the synthetic LXR agonist T09 (1 µM; green) for the final three hours.Shown are (B) oxylipin concentrations and (C) protein levels (mean ± SEM; cells from 3-5 donors).(D) The three macrophage phenotypes with or without T09 (1 µM) were immunostained for specific macrophage markers CD14 (untreated), CD86 (M1), or CD163 (M2) using specific mouse anti-human antibodies (green) (magnification 63×).Nuclei were counterstained with Hoechst (blue).The immuno-positivity of the antibodies towards the respective macrophage types is shown in Figure S1.Differences were considered significant at p-values < 0.05 (**) or < 0.01 (***) using a two-tailed, unpaired student's t-test.
The T09-induced 15-LOX abundance and activity were further evaluated in a timeand dose-dependent manner (Figure 3 and Table 1).Following the evaluation of the optimal incubation conditions, the induction of 15-LOX by other LXR-, FXR-, and RXR-specific compounds as well as cholesterol derivatives (Figure 4) was investigated.The results are shown in Figure 5 and Table S4.Following the evaluation of the optimal incubation conditions, the induction of 15-LOX by other LXR-, FXR-, and RXR-specific compounds as well as cholesterol derivatives (Figure 4) was investigated.The results are shown in Figure 5 and Table S4.

Discussion
The nuclear receptor LXR is associated with different immune cell functions [32].We have recently shown that stimulation of M2-like macrophages with LXR agonist T09 leads to overexpression of ALOX15 as well as increased formation of 15-LOX-derived oxylipins Table 1.Oxylipin concentrations and 15-LOX levels in M2-like macrophages.Primary blood monocytic cells were differentiated into M2-like macrophages with 10 ng/mL M-CSF for 8 days and incubated with IL-4 for the final 48 h.M2-like macrophages were incubated with 1 µM T09 for different periods of time to investigate the time-dependent correlation of 15-LOX abundance and activity (top) and with different T09 concentrations for 24 h to determine the dose-dependent correlation (bottom).The results are shown as mean ± SEM (n = 3-5).

Discussion
The nuclear receptor LXR is associated with different immune cell functions [32].We have recently shown that stimulation of M2-like macrophages with LXR agonist T09 leads to overexpression of ALOX15 as well as increased formation of 15-LOX-derived oxylipins [31].15-LOX as well as its products, such as 15-HETE or 15-HEPE, are discussed as having anti-inflammatory properties [30,[33][34][35].
Here, we investigated the effect of LXR activation on the enzymes of the ARA cascade and oxylipin formation, as well as the effect on markers of human primary macrophages involved in polarization, in detail (Figure 2A).Our experiments demonstrate that in IL-4stimulated M2-like macrophages, 15-LOX abundance is increased by T09 (Figure 2B,C).The LXR agonist T09 has a specific effect on 15-LOX abundance and its metabolites (15-HETE, 5,15-diHETE).T09 showed no or a low effect on PPARγ, IL-1RII (Figure 2C), and the surface marker CD163 (Figure 2D), which are highly expressed in M2-like macrophages [36,37], and thus this indicates that polarization of M2 macrophages was not influenced by T09-stimulation.
In M1-like macrophages, T09 also has a neglectable effect on the investigated markers involved in the differentiation process (Figure 2C,D).The surface marker CD86, which is highly expressed in M1-like macrophages after IFNγ stimulation [38], remained unchanged.
Stimulation with T09 had no effect on the 15-LOX-derived 15-HETE and did not increase ALOX15 abundance in these cells.Stimulation with T09 hardly modulated M1-specific proteins, with a trend towards higher TLR2 and TLR4 amounts and a decrease of COX-2 and its products PGE2 and 12-HHT (Figure 2B,C).
The untreated macrophages have low levels of PPARγ and the surface marker CD14 remained unchanged after stimulation with T09.Interestingly, incubations with T09 reduced 12-HETE formation (Figure 2B) and 12-LOX protein levels (Figure 2C).However, concentrations of 12-HETE were low and originated from platelets present in cell preparations due to unavoidable platelet contaminations [39].Again, neither elevated 15-HETE nor changed 15-LOX abundance were observed after T09 incubation in untreated macrophages.15-LOX-2, which also catalyzes the formation of 15-HETE among other oxylipins, is found in low concentration in untreated cells, and its abundance is not stimulated by T09.While all macrophage phenotypes have comparable levels of LXRα (Figure 2C), only incubations of M2-like macrophages with the LXR agonist T09 specifically increase 15-LOX abundance and its oxylipins.No other major effect of T09 on macrophage markers, namely IL-1RII and PPARγ in M2-like and TLR2 and TLR4 in M1-like macrophages, was observed (Figure 2C).Surface markers CD14, CD86, and CD163 (Figure 2D) were not changed after T09 stimulation, suggesting T09 has no effect on specific surface receptors characteristic of macrophage polarization.
A strong LXR-induced dose-dependent ALOX15 abundance in M2-like macrophages was observed after 24 h of incubation (Figure 3A).At 1 µM T09, 15-LOX abundance can be maximally increased 2-3-fold; higher concentrations led to no further increase in the abundance of 15-LOX (Figure 3B).The mono-hydroxylated oxylipins show a 2-4-fold higher concentration (Figures 3B and S3B), whereas the concentration of the multi-hydroxylated metabolites is increased up to 15-fold (Figures 3B and S3B).No saturation could be achieved in the dose-dependent stimulation of M2-like cells; higher T09 concentrations were found to be cytotoxic (Figure S5).Nevertheless, incubations with 1 µM T09 reached maximum levels of 15-LOX and oxylipins, higher than all other reports about this enzyme and its products, even compared to IL-4-induced macrophages [34,40].
So far, the presence of the cytokines IL-4 and IL-13 has been required to induce ALOX15 at the mRNA and protein levels in human macrophages in cell culture experiments [34,41].Even with prolonged incubations with 100 ng/mL LPS for more than 16 h, the 15-LOX abundance at the mRNA as well as protein level remain unchanged, whereas the 15-LOXderived oxylipins increase by a factor of 1.5-2 [40].In addition, the concentrations of multihydroxylated oxylipins, such as 5,15-diHETE and 7,17-diHDHA, were at most increased by factor 2 [40].The LXR-mediated effect is stronger than the previously described results on 15-LOX activity in macrophages and thus could be of high biological relevance in the regulation of inflammatory pathways in macrophages.However, other pathways may also be involved in the regulation of 15-LOX abundance and activity.
15-LOX is involved in the regulation of inflammation and is thought to play a protective role in arthritis, promote wound healing and host defense and counteract fibrosis [42][43][44].This is supported on the one hand by experiments of ALOX15 silencing in various experimental models, which was associated with inflammation and tissue damage [42].On the other hand, overexpression of human reticulocyte 15-LOX in experimental models protected transgenic animals against atherosclerosis [45].
Anti-inflammatory properties have been described for several of the 15-LOX products.ARA-derived 15-HETE can activate PPARγ and inhibit neutrophil migration, degranulation, and superoxide formation [17,[45][46][47][48][49]. 15-HETE can also reduce inflammatory signaling by regulating the TNFα mRNA half-life [50,51].EPA-derived 15-HEPE is also thought to have anti-inflammatory properties.In transgenic fat1 mice expressing an n3-PUFA generating desaturase, a protective effect was observed after DSS-induced colitis, which was attributed to the high concentrations of 12-HEPE, the major 15-LOX metabolite of EPA in mice.The results were corroborated by silencing of ALOX15 in fat1 transgenic mice, where 12-HEPE concentration was strongly reduced.Moreover, the 15-LOX metabolite 15-HEPE was also reduced in ALOX15-silenced fat1 transgenic mice, and the wild-type mice were protected against the development of DSS-induced colitis after intraperitoneal injections with 15-HEPE [35].
Upon interaction with several LOX enzymes, 5,15-diHETE or 8,15-diHETE can be formed, which are associated with neutrophil and eosinophil chemotactic activity [46,52,53].EPAderived 5,15-diHEPE, also known as resolvin E4, has been described to induce efferocytosis of apoptotic neutrophils in human macrophages and senescent red blood cells along with an upregulation of PPARγ gene expression, leading to a resolution of inflammation [54].In addition, the ALOX15 DHA-derived multi-hydroxylated metabolite 7,17-diHDHA, also called resolvin D5, increases phagocytic activity in neutrophils and macrophages and decreases the formation of pro-inflammatory mediators such as TNFα and NF-κB [55,56].However, these multi-hydroxylated oxylipins are synthesized in cells and tissues only at low concentrations, and many of the proposed effects and signaling pathways are a matter of discussion [30].
Here, we show that LXR agonist T09 specifically elevates ALOX15 abundance, while the effects on markers and proteins involved in the regulation of inflammation (PPARγ, TLR2, and TLR4 or COX) were minor.This suggests a trend towards a more anti-inflammatory biology in the macrophages due to the increased abundance and activity of 15-LOX.Thus, with the correlation between LXR activation and the formation of oxylipins, we conclude that part of the anti-inflammatory effects of LXR [16,20,31] are mediated by 15-LOX and its products.Given that the effects of 15-LOX modulation are more pronounced compared to the TLR4 ligand LPS, this argues that LXR seems to be a major regulatory pathway in macrophages.However, other pathways might also contribute to this effect, as T09 also has a side activity on other nuclear receptors.
We examined the specificity of the ALOX15 regulation by LXR activation.We analyzed the effects of the related nuclear receptors FXR and RXR (Figure 4) with regard to 15-LOX modulation in M2-like macrophages using specific agonists and antagonists.Incubations with the LXR antagonist GSK2033 did not lead to changes in 15-LOX abundance or the formation of its metabolites (Figure 5), indicating that LXR might not be the only factor involved in ALOX-15 regulation.
Following treatment of M2-like cells with the FXR agonist hyodeoxycholic acid (HDCA) [57], the 15-LOX abundance remains unchanged.The FXR agonist fexaramine leads to a 2-fold increase in 15-LOX abundance (Figure 5 Proteomics), and the corresponding oxylipin concentrations are also increased by 2 to 5-fold (Figure 5 Oxylipins).However, due to the rather nonspecific binding pocket [58,59] of the nuclear receptors, the binding of ligands overlaps.Thus, the effect of fexaramine could be caused by a side activity on LXR, though it was previously described that fexaramine cannot activate LXR [60].Consistently, it has been shown that the RXR agonists bexarotene and 9-cis retinoic acid (9-RA) have agonistic activity toward LXR [61], and thus, bexarotene increased 15-LOX abundance 2-fold (Figure 5 Proteomics) and its products 2-to 5-fold (Figure 5 Oxylipins).These results also might indicate the involvement of other nuclear receptors or an off-target effect of the substances on LXR.9-RA showed lower effects on 15-LOX: the 15-LOX concentrations apparently remained unchanged while 12-HETE and 5,15-diHETE formation increased (Figure 5).Finally, we searched for endogenous ligands eliciting the LXR-mediated effects on the ARA cascade in M2-like macrophages (Figure 4D).Oxysterols and metabolites from cholesterol biosynthesis have been previously described as LXR activators [22,[62][63][64].
The activity of oxysterols on ALOX15 in M2-like macrophages is consistent with ligand binding affinity to the LXR [21,63].In structure/activity studies, it was shown that the position of the functional group is of great importance, the essential positions being C-22 and C-24 [63,65].In our experiments, we were also able to correlate the effect strength with the position of the functional group.Thus, when incubated with 22(R)-OH cholesterol and 24(S),25-Ep cholesterol, we observed the strongest effect on 15-LOX abundance and activity, whereas with 25-OH cholesterol, hardly any changes were observed.Our results are consistent with the effects described, where a significant LXR activation is observed for 22(R)-OH cholesterol > 24(S)-OH cholesterol > 25-OH cholesterol, whereas 25-OH cholesterol has little to no effect [21,63,64,66].Other significantly important ligands are described to be 24(S),25-Ep cholesterol, 20(S)-OH cholesterol, and 27-OH cholesterol [63,64,67].Of note, 24(S),25-Ep cholesterol is not formed from cholesterol like other oxysterols but is a by-product of cholesterol biosynthesis (mevalonate pathway) [63,67,68].In addition, the central intermediate from cholesterol biosynthesis, desmosterol, caused strong effects on 15-LOX.Following stimulation with desmosterol, 15-LOX is overexpressed almost 3-fold, which is comparable to the effect of T09 on 15-LOX abundance (Figure 5 Proteomics).15-LOX activity in incubations with desmosterol and T09 is also comparable, with the mono-hydroxylated oxylipins also being increased 4-fold and the multi-hydroxylated oxylipins up to 16-fold (Figure 5 Oxylipins).Thus, we have found active endogenous LXR ligands, namely desmosterol, 22(R)-OH cholesterol and 24(S),25-Ep cholesterol, that are able to massively increase 15-LOX abundance and activity through LXR.
Oxysterols occur in mammalian tissues in micromolar concentrations and circulate in only nanomolar concentrations [63,67,69].Elevated (circulating) oxysterol levels are associated with pathological structures, such as foam cells or atherosclerotic lesions [32,67,70,71].High 24(S)-OH cholesterol plasma levels are associated with Alzheimer's disease [67,72,73].Desmosterol can be found accumulated mainly in macrophage foam cells and atherosclerotic plaques, where it regulates via LXR the activation of genes involved in cholesterol efflux [74].Due to the induction of ALOX15 abundance and the elevated formation of anti-inflammatory oxylipins, desmosterol might also be involved in the reduction of atherosclerotic lesions via LXR activation.22(R)-OH cholesterol, 24(S)-OH cholesterol, and 24(S),25-Ep cholesterol have been shown to inhibit inflammatory signaling in cell culture models [75].As many oxysterols and metabolites of cholesterol biosynthesis are involved in the regulation of inflammatory and immune responses [16,[76][77][78], this study of the effect of cholesterol derivatives on macrophages is of great importance to better understand the interactions between sterol metabolism and LXR-mediated lipid mediators.The induction of ALOX15 in macrophages is associated with anti-inflammatory properties [45].Our results show that LXR upregulation may be related to the modulation of 15-LOX and the formation of anti-inflammatory oxylipins.Thus, the observed relationship between LXR and oxylipin formation activated by endogenous sterols may represent an important feedback regulation in macrophages, where sterols indirectly modulate 15-LOX through their binding to LXR, initiating inflammatory resolution.
Limitations: As with all mechanistic cell culture studies using cell model systems and chemical probes, our study has limitations: The primary macrophages used, derived from different subjects, showed different (basal) levels of 15-LOX abundance and activity, which could be a confounder in the observed regulation by LXR agonists.Macrophages derived from healthy donors from a blood donation center, but no detailed information about their health status, sex, age, or medications were available.These factors might also contribute to the regulation of 15-LOX.Regarding the chemical probes, it should be noted that one agonist and one antagonist of LXR and no antagonists of FXR and RXR could be tested in the study, and they may elicit non-specific effects at the tested concentrations.

Cultivation of Macrophages from Human PBMC
Primary human macrophages were purified and differentiated as described [40].Buffy coats were obtained from the generation of erythrocyte concentrates from the blood donor services at the University Hospital of Düsseldorf, Germany, with the informed consent of healthy human subjects.The study was approved by the Ethical Committee of the University of Wuppertal.In brief, primary human blood monocytic cells (PBMC) were isolated from fresh buffy coats by dextran (5%) sedimentation for 45 min.The supernatant was layered on lymphocyte separation medium and centrifuged for 10 min at 1000× g without deceleration.PBMC sedimented on top of the lymphocyte separation medium was collected.Following washing with PBS twice, cell pellets were resuspended in RPMI 1640 medium supplemented with 1% L-glutamine and 1% penicillin/streptomycin (P/S).Cell suspensions were transferred to petri dishes and incubated for 1 h (37 • C; 5% CO 2 ; humidified atmosphere).Afterwards, dishes were washed twice with RPMI 1640 medium supplemented with 1% L-glutamine and 1% P/S to remove non-adherent cells and layered with RPMI 1640 growth medium, which contains 2 g/L glucose and is supplemented with 1% L-glutamine, 1% P/S, and 5% heat-inactivated human AB plasma.The monocytes were differentiated into different macrophage phenotypes for 8 days.For the M1-like phenotype, the growth medium was supplemented with 10 ng/mL GM-CSF, and for the M2-like phenotype, with 10 ng/mL M-CSF.The growth medium was refreshed every other day.In addition, for the last 48 h the cells were incubated with 10 ng/mL IFNγ (M1) or 10 ng/mL IL-4 (M2).Six hours before the harvest, the M1-like cells were additionally stimulated with 100 ng/mL LPS.For the untreated phenotype, cells were cultivated in growth medium without the addition of cytokines.
The differentiated macrophages were treated with test compounds at different time points and concentrations.
The cytotoxic effect of the test substances was investigated by means of the resazurin (alamar blue) assay [79], and only non-toxic concentrations were used (Figure S5).
The cells were harvested using the cold shock method by washing with PBS and incubating in ice cold PBS/EDTA (20 min/4 • C).The cells were collected by scraping, centrifugation, and pelleting followed by washing with PBS containing protease inhibitors.The harvested primary macrophage pellets were frozen at −80 • C until use.

Analysis of Oxylipins and Protein Levels by LC-MS/MS
The analysis of oxylipins and protein levels was carried out from the same cell pellet.Cell pellets were resuspended in PBS containing an antioxidant/inhibitor mixture (0.2 mg/mL BHT, 100 µM indomethacin, 100 µM trans-4-(-4-(3-adamantane-1-yl-ureido-9-cyclohexyloxy)-benzoic acid (t-AUCB) in MeOH) and sonicated [80,81].The protein content was determined by the bicinchoninic acid (BCA) assay.Following addition of 10 µL of the oxylipin internal standards (   C; 30 min), the supernatant was used for the oxylipin analysis whereas the protein pellet was frozen at −80 • C for a later protein level analysis [39].
MS data analysis was performed with the software MultiQuant TM 3.0.2(Sciex) using a Gaussian smooth width of 1 and the MQ4 integration algorithm.Concentrations were calculated via external calibration using internal standards based on the ratio of the analyte and internal standard areas.The quantified peptide/protein and oxylipin concentrations were normalized to the absolute protein content determined via the BCA assay.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism (GraphPad Software Inc. version 6.01, San Diego, CA, USA).Data are presented as mean ± standard error of mean (SEM).Statistical analyses were performed by a two-tailed, unpaired student's t-test.Differences were considered significant at p-values < 0.05 or 0.01.

Conclusions
We demonstrate that LXR activation by T09 exerts dramatic effects on the ARA cascade, specifically in M2-like macrophages, by increasing the abundance of 15-LOX and increasing the formation of its products.No or only a moderate effect of T09 was observed on untreated and M1-like macrophages, as well as surface markers and PPARγ, IL-1RII, TLR2, and TLR4 abundance.
The investigated agonists of FXR and RXR only had a moderate effect on 15-LOX abundance and activity compared to T09, presumably mediated by LXR through side activity.
We found that sterols are important endogenous LXR agonists regulating ALOX15.We could show that particularly the oxysterols 24(S),25-Ep cholesterol, and 22(R)-OH cholesterol and the cholesterol precursor desmosterol modulate the ARA cascade in macrophages, presumably by LXR.Thus, our results show a new cross-link between two lipid mediator classes: sterols/oxysterols and oxylipins.Sterols affect the LXR-mediated formation of 15-LOX-derived oxylipins.This regulatory mechanism should be further investigated, as the 15-LOX-derived oxylipins are suggested to play an important role in inflammatory resolution.

Figure 2 .
Figure 2. Effect of the LXR agonist T09 on the ARA cascade and differentiation of primary macrophages.(A) Primary human blood monocytic cells were treated with 10 ng/mL GM-CSF (M1 type) for 8 days or with M-CSF (M2 type) for 8 days, as well as 10 ng/mL IFNγ (M1 type) or IL-4 (M2 type) for the final 2 days.Additionally, M1 cells were challenged with 100 ng/mL LPS for 6 h.For the untreated macrophages, the adhered monocytes were left untreated for 8 days.The three different macrophage phenotypes were incubated with (blue) or without the synthetic LXR agonist T09 (1 µM; green) for the final three hours.Shown are (B) oxylipin concentrations and (C) protein levels (mean ± SEM; cells from 3-5 donors).(D) The three macrophage phenotypes with or without T09 (1 µM) were immunostained for specific macrophage markers CD14 (untreated), CD86 (M1), or CD163 (M2) using specific mouse anti-human antibodies (green) (magnification 63×).Nuclei were counterstained with Hoechst (blue).The immuno-positivity of the antibodies towards the respective macrophage types is shown in Figure S1.Differences were considered significant at p-values < 0.05 (**) or < 0.01 (***) using a two-tailed, unpaired student's t-test.

Figure 2 .
Figure 2. Effect of the LXR agonist T09 on the ARA cascade and differentiation of primary macrophages.(A) Primary human blood monocytic cells were treated with 10 ng/mL GM-CSF (M1 type) for 8 days or with M-CSF (M2 type) for 8 days, as well as 10 ng/mL IFNγ (M1 type) or IL-4 (M2 type) for the final 2 days.Additionally, M1 cells were challenged with 100 ng/mL LPS for 6 h.For the untreated macrophages, the adhered monocytes were left untreated for 8 days.The three different macrophage phenotypes were incubated with (blue) or without the synthetic LXR agonist T09 (1 µM; green) for the final three hours.Shown are (B) oxylipin concentrations and (C) protein levels (mean ± SEM; cells from 3-5 donors).(D) The three macrophage phenotypes with or without T09 (1 µM) were immunostained for specific macrophage markers CD14 (untreated), CD86 (M1), or CD163 (M2) using specific mouse anti-human antibodies (green) (magnification 63×).Nuclei were counterstained with Hoechst (blue).The immuno-positivity of the antibodies towards the respective macrophage types is shown in Figure S1.Differences were considered significant at p-values < 0.05 (**) or < 0.01 (***) using a two-tailed, unpaired student's t-test.