Adipose triglyceride lipase acts on neutrophil lipid droplets to regulate substrate availability for lipid mediator synthesis

In humans, mutations in ATGL lead to TG accumulation in LDs of most tissues and cells, including peripheral blood leukocytes. This pathologic condition is called Jordans’ anomaly, in which functional consequences have not been investigated. In the present study, we tested the hypothesis that ATGL plays a role in leukocyte LD metabolism and immune cell function. Similar to humans with loss‐of‐function mutations in ATGL, we found that global and myeloid‐specific Atgl−/− mice exhibit Jordans’ anomaly with increased abundance of intracellular TG‐rich LDs in neutrophil granulocytes. In a model of inflammatory peritonitis, lipid accumulation was also observed in monocytes and macrophages but not in eosinophils or lymphocytes. Neutrophils from Atgl−/− mice showed enhanced immune responses in vitro, which were more prominent in cells from global compared with myeloid‐specific Atgl−/− mice. Mechanistically, ATGL−/− as well as pharmacological inhibition of ATGL led to an impaired release of lipid mediators from neutrophils. These findings demonstrate that the release of lipid mediators is dependent on the liberation of precursor molecules from the TG‐rich pool of LDs by ATGL. Our data provide mechanistic insights into Jordans’ anomaly in neutrophils and suggest that ATGL is a potent regulator of immune cell function and inflammatory diseases.


Introduction
NLSD describes a group of rare, autosomal-recessive disorders of lipid metabolism with systemic deposition of neutral lipids in multiple tissues. The disease was discovered initially in 1953 by Jordans [1], who reported a case of progressive muscular dystrophy accompanied by fat-containing vacuoles in leukocytes, which, since then, has been termed Jordans' anomaly. Lipidladen leukocytes were confirmed in further clinical reports of NLSD [2][3][4]. Given the consistent presence of Jordans' anomaly on an otherwise heterogeneous background of clinical phenotypes (e.g., variable degrees of ichthyosis, liver steatosis, hepatosplenomegaly, skeletal and cardiac myopathy, hearing loss, growth, and mental retardation) [5], Jordans' anomaly was proposed as hallmark criterion for the diagnosis of NLSD [6]. In 2001 and 2007, mutations in CGI-58 (also termed a/b hydrolase domain-containing 5) and ATGL (also termed desnutrin or PNPLA2) were linked to NLSD with ichthyosis and NLSDM, respectively [7,8]. ATGL and CGI-58 are key regulators of lipolysis. The catabolism of TG-rich intracellular LDs takes place predominantly in adipose tissue. It is initiated by ATGL, followed by the action of HSL and MGL, and it largely depends on the presence of the ATGL coactivator CGI-58 [9]. As LDs have been recognized as functional and dynamic organelles beyond lipid storage depots for metabolic fuel, novel implications of LDs and their associated proteins in various cell types are steadily emerging. LDs have been linked to a variety of biologic functions, such as cellular regulation of lipid homeostasis and signaling, protein folding, storage and degradation, infection, and immunity (reviewed in refs. [10][11][12]). Several human diseases (e.g., inflammatory arthritis, respiratory distress syndrome, and atherosclerosis) are associated with LD accumulation in various types of leukocytes, implicating LDs as critical regulators of inflammation. Significant inflammation-relevant functions that have been associated with LDs (often termed lipid bodies in the context of inflammation) include the regulation of host/ pathogen interactions [13], virus assembly [14], nutrient source for intracellular pathogens [15], and functions in antigen crosspresentation [16]. Numerous studies have provided substantial evidence that LDs serve as major organelles for the generation of AA-derived lipid mediators (reviewed in ref. [17]).
In cells of the myeloblastic lineage (e.g., neutrophils, eosinophils, basophils, mast cells), TGs are the major lipid class of the LD core [18][19][20]. However, the function of lipolytic enzymes that use this TG pool in cells of the innate or adaptive immune system is largely unexplored. Here, we investigated the role of ATGL in leukocyte LD biology and immune cell function by examining neutrophils lacking ATGL. Our study shows that similar to humans with defective ATGL, Atgl 2/2 mice exhibit Jordans' anomaly. In addition, ATGL 2/2 leads to alterations in immune cell function and a marked reduction in lipid mediator generation. To our knowledge, this is the first report providing detailed information about the functional consequences of Jordans' anomaly caused by ATGL 2/2 .

Isolation of neutrophils
Neutrophils were isolated from peritoneal exudate or bone marrow. Cells of peritoneal exudate were collected 1 d after intraperitoneal injection of 2.5 ml 3% thioglycolate by rinsing the peritoneum with 10 ml HBSS containing 1 mM EDTA. For bone marrow isolation, tibia and femur from both hind legs were collected, flushed with HBSS-EDTA, and pushed through a cell strainer to disperse cell clumps. The cell suspension was centrifuged (400 g, 10 min, 4°C), and neutrophils were purified by use of a Percoll gradient, as described [23]. In brief, cell suspensions were layered on top of a 3-layer Percoll gradient of 78, 69, and 52% Percoll and centrifuged (1500 g, 30 min, room temperature) with brakes off. Neutrophils from the 69-78% interphase were harvested and counted. After washing with HBSS-EDTA containing 1% BSA, cells were suspended in HEPES buffer plus 9 mM glucose and used immediately for further experiments.

Transmission electron microscopy
Bone marrow cells were fixed in 2.5% glutaraldehyde (w/v) and 5% methanolfree formalin in 0.1 M phosphate buffer (pH 7.4) for 2 h at room temperature, postfixed in 2% osmium tetroxide (w/v) for 2 h, dehydrated in a graded series of ethanol, and embedded in a TAAB epoxy resin. Ultrathin sections (75 nm) were cut with a Leica UC6 Ultramicrotome (Leica Microsystems, Wetzlar, Germany) and stained with lead citrate for 5 min and with uranyl acetate for 15 min. Images were taken by use of a FEI Tecnai G2 20 transmission electron microscope (FEI, Eindhoven, Netherlands) with a Gatan Ultrascan 1000 chargecoupled device camera. Acceleration voltage was 120 kV.

Lipid analysis of peritoneal lavage
Peritoneal lavages were harvested from mice 1 d after intraperitoneal thioglycolate injection. Lipids from peritoneal exudates were extracted by use of the Folch method. After evaporation under constant nitrogen flow, 100 ml 1% Triton X-100 (dissolved in chloroform) was added, and lipids were dried again under a stream of nitrogen. Thereafter, samples were dissolved in 100 ml ddH 2 O for 15 min at 37°C. Aliquots were used for enzymatic measurements of TG, TC, and FC by use of kits, according to the manufacturer's protocols (DiaSys, Holzheim, Germany). The readings were normalized to protein concentrations. To determine the composition of TG, PC, and phosphatidylethanolamine species, extracted lipids were dissolved in 2-propanol:chloroform: methanol (7:2:1 v:v:v) and analyzed by LC/ESI-MS, as described [24].
Calcium flux and chemotaxis assays acetoxymethylester (Life Technologies) and 0.02% pluronic F-127 (Sigma-Aldrich) in the absence of Ca 2+ and Mg 2+ for 1 h at room temperature. Cells were washed with assay buffer and resuspended in assay buffer containing Ca 2+ and Mg 2+ to a final concentration of 0.3-0.5 3 10 6 neutrophils/ml. Aliquots of 500 ml were used for each condition. Maximal Ca 2+ responses were determined after adding various concentrations of fMLP and KC. Kinetics of Ca 2+ flux were analyzed by use of FlowJo software (Tree Star).
For chemotaxis assays, bone marrow-derived neutrophils were suspended in assay buffer containing Ca 2+ and Mg 2+ (2 3 10 6 cells/ml). Cell suspension (50 ml/well) were placed onto the top plate of an AP48 micro-Boyden chemotaxis chamber with a 5 mm pore-size polycarbonate filter (Neuro Probe, Gaithersburg, MD, USA). Cells were allowed to migrate toward 30 ml fMLP and KC in the bottom wells of the plate for 1 h at 37°C. The membrane was removed carefully, and migrated cells were enumerated by flow cytometric counting.

Phagocytosis and ROS measurement
Neutrophils were isolated from peritoneal lavages, 1 d after thioglycolate injection. After 1 h of preincubation in culture medium (DMEM, 4 mM glutamine, 1 mM pyruvate, 10% lipoprotein-deficient serum, 1% penicillin/ streptomycin) in the absence of glucose for 1 h, neutrophils (1 3 10 6 ) were incubated with 100 ml fluorescein-labeled Escherichia coli particles (Vybrant phagocytosis assay; Molecular Probes, Life Technologies) for 1 h at 37°C in the dark under gentle rotation. The suspension was removed, and 100 ml trypan blue was added for 1 min to quench extracellular fluorescein-labeled bacteria. Cells were washed with cold HBSS-EDTA and stained with a neutrophil-specific marker (Ly6G-PE; BD Biosciences), and the amount of phagocytosed E. coli was measured by flow cytometry. Data are expressed as geometric means of fluorescence intensity of the intracellular fluorescence detected. Intracellular ROS generation, during phagocytosis and after a 20 min exposure to KC (20 ng/ml) and fMLP (1 mM), was measured according to the manufacturer's protocol (Life Technologies).

Inhibition of ATGL in vitro after FA loading
Peritoneal neutrophils were cultivated in culture medium containing 25 mM glucose, supplemented with 300 mM OA or AA (Sigma-Aldrich), complexed to BSA (OA-BSA, AA-BSA). FA:BSA complexes were prepared as described [25]. After an incubation period of 6 h in the absence or presence of the ATGL inhibitor Atglistatin (40 mM) and Hi (25 mM; kindly provided by Dr. Martina Schweiger, University of Graz), the medium was replaced by serum-free DMEM containing Atglistatin or Atglistatin/Hi. After 4 and 6 h of incubation, LDs were quantified by flow cytometry, as described above.

Quantification of PUFAs and lipid mediators
Peritoneal neutrophils were preincubated in culture medium (containing 25 mM glucose), with or without 40 mM Atglistatin, for 6 h. Thereafter, the medium was replaced by HBSS containing Ca 2+ and Mg 2+ in the absence or presence of the Ca 2+ ionophore A23187 (0.5 mM; Sigma-Aldrich) for 1 h. Then, cells were placed on ice, transferred into tubes, and centrifuged at 400 g for 10 min at 4°C. Supernatants were immediately frozen in liquid N 2 and stored at 280°C (maximum 1 wk) until analysis. To determine PUFA and lipid mediator concentrations in peritoneal lavage fluids, the peritoneum was rinsed with 3 ml HBSS-EDTA. The cell-free exudate was snap frozen in liquid N 2 and stored at 280°C. A QTRAP 5500 mass spectrometer (AB Sciex, Darmstadt, Germany), operating in negative ESI mode, was applied for LC-MS/MS analysis by use of 200 ml cell supernatant or peritoneal lavage fluid, as described previously [26].

Statistical analysis
Statistical analyses were performed by use of GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA, USA). Significance was calculated by Student's t test or ANOVA, followed by Bonferroni correction. Data are shown as means + SEM. For statistical analysis of PUFAs and lipid mediators, nonparametric Wilcoxon signed-rank tests and Mann-Whitney U tests were used. Data are expressed as median with minimum and maximum range. Statistical significance is indicated in the figure legends.

RESULTS
Atgl 2/2 mice exhibit Jordans' anomaly in neutrophil granulocytes Immunophenotyping of peripheral blood by flow cytometry revealed comparable blood counts between WT and Atgl 2/2 mice (Fig. 1A). By use of BODIPY 493/503 as a neutral lipid dye, we observed increased fluorescence intensity, exclusively in neutrophils of Atgl 2/2 mice (Fig. 1B). To gain insight into the ultrastructural morphology of the cells, we performed transmission electron microscopy of WT and Atgl 2/2 bone marrow cells. As shown in Fig. 1C, intracellular LDs were present in the cytoplasm of Atgl 2/2 bone marrow cells as electron-lucent, round-shaped structures, whereas no LDs were visible in WT cells. Next, we evaluated by flow cytometry whether hematopoietic precursor cells of Atgl 2/2 mice accumulate LDs. Relative populations of bone marrow cells were comparable between WT and Atgl 2/2 mice (Fig. 1D). We observed an increased amount of LDs exclusively in mature Gr-1 + neutrophils of Atgl 2/2 mice but not in hematopoietic stem cells, such as LSK, or any of the investigated precursor cells, such as CLPs, MDPs, and GMPs (Fig. 1E).

Pronounced accumulation of LDs under inflammatory conditions in the absence of ATGL
In naive lavages, resident peritoneal cells include predominantly macrophages, lymphocytes, and small numbers of granulocytes and monocytes ( Fig. 2A). Similar to our findings in peripheral blood, we observed a selective LD accumulation in neutrophils of naive lavages from Atgl 2/2 mice (Fig. 2B). To elucidate the effect of ATGL 2/2 under inflammation-driven conditions, we used a thioglycolate-induced peritonitis model [27]. The cellular infiltration into the peritoneum was unchanged between WT and Atgl 2/2 mice, 1 d (Fig. 2C) and 3 d (Fig. 2E) postinjection. An increased LD abundance in neutrophils and monocytes was observed on d 1 (Fig. 2D) and in neutrophils, monocytes, and macrophages on d 3 postinjection (Fig. 2F) in peritoneal lavages from Atgl 2/2 mice. Oil Red O staining of sedimented peritoneal cells confirmed this increased abundance of LDs in cells from Atgl 2/2 mice, whereas in WT cells, only few LDs were visible (data not shown). Lipid parameters of peritoneal cells isolated from mice 1 d after thioglycolate injection showed a 2.6-fold TG accumulation in Atgl 2/2 cells. TC, FC, and CE concentrations were unaffected (Fig. 3A). With the use of LC-ESI/MS analysis, we found that the relative FA composition of the most abundant TG species (abundance .3% of total TGs), ranging from C 50 to C 56 , differed significantly between WT and Atgl 2/2 cells (Fig. 3B). In neutrophils from Atgl 2/2 mice, TG species with .4 double bonds, including predominantly 18:1, 18:2, 20:3, and 20:4, showed an increased relative abundance, whereas TG species with up to 3 double bonds, containing mostly 16:0, 18:0, and 18:1, were decreased. Absolute concentrations of the most abundant phospholipid species, PC, were unchanged (Fig. 3C,  inset). The relative PC composition was changed to an increase in 36:2 and a reduction in 34:1 in Atgl 2/2 neutrophils (Fig. 3C). No alterations were observed in phosphatidylethanolamine between WT and Atgl 2/2 neutrophils, neither in absolute (Fig.  3D, inset) nor relative amounts (Fig. 3D).

Inflammatory cells express the full cascade of metabolic lipases
Next, we investigated whether lipolytic enzymes are expressed in immune cells. We analyzed flow-sorted macrophages, eosinophils, neutrophils, monocytes (derived from thioglycolate-elicited lavage), as well as B and T lymphocytes (derived from blood) for mRNA expression of ATGL, HSL, and MGL (Fig. 4A) by quantitative realtime PCR. mRNA expression of ATGL was highest in neutrophils, followed by B lymphocytes, eosinophils, T lymphocytes, monocytes, and macrophages. HSL was highly expressed in B lymphocytes and in comparable amounts in eosinophils and neutrophils. T lymphocytes, monocytes, and macrophages showed the lowest levels of HSL mRNA. MGL was most abundant in eosinophils, whereas all other cell types showed little expression. We confirmed the presence of ATGL (54 kDa), HSL (84 kDa), and MGL (33 kDa) in lysates of purified WT neutrophils via immunoblotting (Fig. 4B). Together, our data demonstrate that metabolic lipases are expressed in inflammatory cells.
Agonist-induced Ca 2+ flux and chemotaxis are increased in Atgl 2/2 neutrophils Of particular interest was to examine the functionality of LD-rich Atgl 2/2 neutrophils. Therefore, we determined agonist-induced, [Ca 2+ ] i flux and chemotaxis. As agonists, we used the bacterial peptide fMLP and the neutrophil-specific chemoattractant KC (also known as growth-related oncogene a or CXCL1). Compared with WT cells, which reached a maximal Ca 2+ response of ;230% over baseline, we observed an enhanced mobilization of [Ca 2+ ] i by .300% in Atgl 2/2 neutrophils with fMLP concentrations ranging from 1 to 5 mM (Fig. 5A). KC (5 and 50 mM) induced an even higher response in Atgl 2/2 cells with peak values of 369% and 447% over baseline in WT and Atgl 2/2 neutrophils, respectively (Fig. 5A). We next examined the cellular phenotype of Atgl 2/2 cells by measuring in vitro migration. These studies revealed that neutrophils from Atgl 2/2 mice possessed increased chemotactic properties toward both tested agonists compared with WT neutrophils (Fig. 5B). As another marker for neutrophil activation, we determined agonist-induced CD11b expression by flow cytometry. fMLP-and KC-mediated stimulation showed a slight yet nonsignificant trend for an increased activation in Atgl 2/2 neutrophils (Supplemental Fig. 1). By use of PMA to achieve a maximal and robust activation of neutrophils, we found an increased responsiveness in Atgl 2/2 compared with WT cells.
The intracellular TG accumulation in Atgl 2/2 neutrophils might be caused by alterations in cellular uptake. After 10 min exposure to fluorescently labeled FAs, neutrophils from Atgl 2/2 mice showed 2.9-fold increased uptake of FAs compared with control cells (Fig. 5C). Next, we addressed whether the lack of ATGL affects phagocytosis and ROS production of neutrophils. In vitro phagocytosis assays revealed comparable phagocytosis rates between neutrophils from WT and Atgl 2/2 mice (Fig. 5D). Simultaneously, we measured intracellular ROS formation following synchronized phagocytosis, as well as ROS production, in response to KC and fMLP. All conditions tested showed comparable fluorescence intensities, indicating unchanged ROS concentrations (Fig. 5E). Moreover, we assessed apoptosis under various conditions, which was unaltered between WT and Atgl 2/2 neutrophils, as Annexin V/propidium iodide staining revealed the same amount of alive, early, and late apoptotic, as well as necrotic, cells (Supplemental Fig. 2).
Similar but less pronounced phenotype in myeloidspecific Atgl 2/2 mice To elucidate whether Jordans' anomaly is caused by the specific lack of ATGL in immune cells or by the systemic lack of ATGL, we included mice with a targeted deletion in myeloid cells. Similar to our observations from global Atgl 2/2 mice with a 3.7fold increase in LD abundance in peripheral blood neutrophils, myeloid-specific Atgl 2/2 mice showed a 1.7-fold increase in neutrophil LDs (Fig. 6A). Under inflammatory conditions, myeloid-specific Atgl 2/2 mice showed an increased LD abundance exclusively in neutrophils, both 1 d (Fig. 6B) and 3 d (Fig. 6C) post-thioglycolate injection, which was of lower magnitude compared with global Atgl 2/2 mice (2.2-fold in myeloid Atgl 2/2 vs. 3.1-fold in global Atgl 2/2 mice). Ca 2+ flux assays with neutrophils isolated from myeloid-specific Atgl 2/2 mice that uses fMLP and KC as agonists showed a comparable response as WT cells (Fig. 6D). Chemotaxis was increased in response to fMLP but compared with WT cells after KC stimulation (Fig. 6E). Comparable with neutrophils from global Atgl 2/2 , we found increased FA uptake (Fig. 6F) and no alterations in phagocytosis (Fig. 6G) and ROS formation (Fig.  6H). Together, these observations indicate a similar yet less pronounced phenotype in neutrophils from myeloid-specific versus global Atgl 2/2 mice.

Pharmacological inhibition of ATGL results in LD accumulation and reduced release of lipid mediators from neutrophils
To exclude potential systemic effects caused by ATGL 2/2 , we inhibited ATGL in neutrophils ex vivo by treatment with the specific inhibitor Atglistatin [28]. By use of thioglycolate-elicited neutrophils collected from WT mice 1 d after thioglycolate injection, we determined whether Atglistatin effectively inhibited ATGL in these cells. To induce LD formation, we loaded WT neutrophils with 300 mM OA (18:1), and thereafter, we cultured cells in serum-free medium in the absence and presence of Atglistatin. After 4 and 6 h of starvation, we quantified LDs and observed an increased LD content in the presence of Atglistatin (Fig. 7A). These results indicate that the inhibitor blocks FA mobilization, which consequently results in TG accumulation. The combined use of Atglistatin and Hi tended to promote accumulation further of LDs in neutrophils. Similar results were observed after loading neutrophils with 300 mM AA (20:4; Fig. 7B).

DISCUSSION
Lipolysis is mainly catalyzed by ATGL, which specifically cleaves TG in cytosolic LDs. We show here that the systemic loss of ATGL in mice results in an increased abundance of TG-rich LDs specifically in neutrophil granulocytes from bone marrow, peripheral blood, and peritoneal lavage under basal conditions. Under inflammatory conditions, we observed that thioglycolateelicited inflammation triggers LD formation in all investigated immune cells from WT and Atgl 2/2 mice. This is evident by an increased amount of LDs in cells from thioglycolate-injected compared with naive lavages. Glycosylated proteins in the thioglycolate medium activate receptors for advanced glycation end products, which further induce proinflammatory responses through NF-kB-mediated secretion of cytokines and inflammatory factors [30]. In Atgl 2/2 cells, this proinflammatory stimulus boosts LD accumulation to increased amounts of LDs in neutrophils, monocytes, and macrophages, whereas eosinophils and B and T lymphocytes do not show any alterations in LD content between WT and Atgl 2/2 mice. This finding is somehow unexpected, as all immune cells investigated express ATGL. However, as inflammation-mediated induction of LDs in leukocytes is highly cell and stimulus dependent [31], it remains to be determined whether other stimuli are required to trigger ATGL-mediated functions in these cells. LDs have a dual function within cells by controlling lipid storage, thereby protecting cells from lipotoxicity and by serving as energy fuel upon starvation. In contrast to many other cell types, in which ATGL-mediated lipolysis represents an important biochemical mechanism to generate FAs as energy substrates, most lymphoid and myeloid cells (including neutrophils) derive their energy exclusively from glycolysis [32]. In accordance, we did not observe alterations in cell viability or any limitations in highly energy-demanding processes, such as migration and phagocytosis in Atgl 2/2 neutrophils. Of note, the absence of ATGL was associated with an elevated Ca 2+ mobilization and increased chemotaxis of neutrophils toward chemoattractants, indicating a proinflammatory phenotype of these cells. This phenotype was more pronounced in cells from global compared with myeloid-specific Atgl 2/2 mice, indicating that these alterations are likely caused by secondary effects as a result of systemic ATGL 2/2 rather than by the cell-specific lack of ATGL. A causative reason for this might be additional factors provoked by the severe phenotype of global ATGL 2/2 , including perivascular oxidative stress and endothelial dysfunction [33], ER stress, and mitochondrial dysfunction in various tissues [34], with the most dramatic phenotype in cardiac muscle resulting in lethal cardiomyopathy within a few months after birth [35].
cPLA 2a (also known as group IVA PLA 2 ) is assigned a central role in stimulus-dependent 20:4 mobilization from phospholipids during eicosanoid generation. Subcellular eicosanoid-generating compartments include membranes of the ER, phagosome, nuclear envelope, and LD but not the plasma membrane [17]. In a recent study, the TG-rich core of LDs in human mast cells has been identified as an additional substrate reservoir for eicosanoid precursors, with ATGL being critically involved in the release of these precursor FAs [36]. Our observations further support and extend this study showing that ATGL inhibition leads to an accumulation of TG-rich cytoplasmic LDs in neutrophils, monocytes, and macrophages and results in a reduced production of lipid mediators from neutrophils. A central question arising from the mast cell study was whether ATGL provides substrates for lipid mediator synthesis by its TG hydrolase activity [37] or its putative phospholipase activity [38]. Furthermore, it remained to be determined whether ATGL-mediated 20:4 release provides substrates directly for eicosanoid generation or indirectly by replenishing 20:4 in glycerophospholipids for the subsequent release by cPLA 2a . Our lipidomics data clearly show a pronounced increase of 18:1, 18:2, and 20:4 in the TG fraction in the absence of ATGL and no accumulation of linoleyl-or arachidonyl-containing PC or phosphatidylethanolamine, demonstrating that ATGL acts as a TG hydrolase and directly liberates FAs. FA mobilization and release are critical processes, and Atgl 2/2 neutrophils try to compensate for defective lipolysis by increased FA uptake from extracellular sources.
Genetic and pharmacological ATGL inhibition, both in vivo and in vitro, reduces the release of some but not all lipid mediators derived from all 3 enzymatic pathways. Lipid mediator generation is regulated in a highly time-dependent manner at various levels, e.g., through the availability of enzymes and substrates, as well as the site of mediator generation within different cellular compartments [17]. Our results indicate that lipid mediators, which remain unaffected by ATGL inhibition, are derived from non-LD compartments, whereas mediators with a reduced release upon ATGL inhibition are specifically generated from the TG-LD pool in an ATGL-dependent manner. Recently, a comprehensive work describing the cellular and molecular events, including lipid mediator generation in mice with thioglycolate-induced peritonitis, has been published [27]. This study revealed that in this model, the inflammatory response after 24 h is characterized by high levels of 20:4-derived proinflammatory eicosanoids, such as PGD 2 , LTB 4 , PGE 2 , 5-HETE, and TXB 2 . This time point correlates with a high infiltration of the peritoneal cavity with neutrophils, rendering 24 h as a suitable time point to investigate neutrophil-mediated lipid mediator release. In accordance, we detected a release of these eicosanoids from lavages collected from 1 d thioglycolatestimulated mice, which were reduced in Atgl 2/2 mice. As eicosanoids are critical determinants for the modulation of an inflammatory response, a dysregulation in any dimension might have pathologic consequences, such as chronic inflammation, failure of pathogen clearance, and both local and distal tissue damage [39,40]. The physiologic significance of ATGL 2/2 associated with reduced eicosanoid generation remains to be determined.
Clinical reports from patients with mutations in the ATGL gene describe an analogous phenotype as Atgl 2/2 mice: they suffer from massive TG accumulation in multiple tissues and develop severe myopathy, which often requires cardiac transplantation [41]. In humans, Jordans' anomaly affects various immune cells, predominantly granulocytes [42,43]. In some case reports of NLSDM, an increased prevalence of reoccurring infections was observed [44]. Unfortunately, it is not elusive whether these immune defects can be linked to Jordans' anomaly and whether eicosanoid production by blood leukocytes is altered in these patients. In general, a direct correlation between LD number and eicosanoid release was described in human and murine leukocytes [45]. Thus, the increased presence of LDs in neutrophils from humans and mice with Jordans' anomaly should result in an increased potential of these cells to generate lipid mediators. As a result of inhibited lipid mobilization from LDs in Atgl 2/2 conditions, however, the release of lipid mediators is partly inhibited, further underlining the important role of ATGL in the regulation of eicosanoid generation. Observations that mutations in the ATGL coactivator CGI-58 also result in Jordans' anomaly in humans raise the possibility of a similar phenotype in immune cells deficient in CGI-58. Figure 7. Neutrophil LD catabolism is blocked by the ATGL inhibitor Atglistatin . WT neutrophils collected from peritoneal lavages, 1 d postthioglycolate injection were incubated in culture medium supplemented with 300 mM OA (18:1) (A) or AA (20:4) (B) for 6 h, after which, the medium was replaced by serum-free DMEM containing ATGL or ATGL/HSL inhibitors. After 4 and 6 h of starvation, LDs were quantified by flow cytometry. Fluorescence intensities are expressed as percentages of untreated control cells + SEM (n = 6-7). *P , 0.05; **P # 0.01; ***P # 0.001.  Macrophages lacking ATGL or CGI-58, however, showed opposing phenotypes in terms of macrophage function and atherosclerosis susceptibility [46], rendering expectations in this context rather speculative.
In summary, our work strengthens the hypothesis that "lipolysis meets inflammation" with ATGL and probably other lipid hydrolases participating in inflammatory signaling processes [47]. Our data provide mechanistic insights that the release of lipid mediators from neutrophils is dependent on liberation of FAs as precursor molecules from the TG-rich pool of LDs by ATGL. As lipid mediator-synthesizing LDs are not only restricted to neutrophils but are also present in many other immunomodulatory, endothelial, epithelial, and neoplastic cells, it is reasonable to speculate that ATGL regulates FA availability and consequently, the release of lipid mediators also in these cells. Our work provides an important basis for future experiments exploring the role of ATGL-mediated lipid mediator release in the context of inflammatory and neoplastic diseases.