Insulin-induced lipid body accumulation is accompanied by lipid remodelling in model mast cells

ABSTRACT Mast cell lipid bodies are key to initiation, maintenance and resolution of inflammatory responses in tissue. Mast cell lines, primary bone marrow-derived mast cells and peripheral blood basophils present a ‘steatotic’ phenotype in response to chronic insulin exposure, where cells become loaded with lipid bodies. Here we show this state is associated with reduced histamine release, but increased capacity to release bioactive lipids. We describe the overall lipid phenotype of mast cells in this insulin-induced steatotic state and the consequences for critical cellular lipid classes involved in stages of inflammation. We show significant insulin-induced shifts in specific lipid classes, especially arachidonic acid derivatives, MUFA and PUFA, the EPA/DHA ratio, and in cardiolipins, especially those conjugated to certain DHA and EPAs. Functionally, insulin exposure markedly alters the FcϵRI-induced release of Series 4 leukotriene LTC4, Series 2 prostaglandin PGD2, Resolvin-D1, Resolvin-D2 and Resolvin-1, reflecting the expanded precursor pools and impact on both the pro-inflammation and pro-resolution bioactive lipids that are released during mast cell activation. Chronic hyperinsulinemia is a feature of obesity and progression to Type 2 Diabetes, these data suggest that mast cell release of key lipid mediators is altered in patients with metabolic syndrome.


Introduction
Metabolic syndrome is associated with a chronic state of hyperinsulinemia, prior to the onset of Type 2 Diabetes mellitus [1,2]. Chronic insulin elevation has functional consequences for numerous cells, tissues and organs in the body, including those of the immune system [3][4][5]. We have previously shown that chronic insulin elevation alters mast cell functional phenotypes in vitro, and there is in vivo and clinical evidence that altered levels of insulin affect the outcomes of allergic and anaphylactic inflammatory responses [6][7][8][9]. Chronic insulin exposure quantitatively alters the lipid content of mast cells, causes a steatosis-like accumulation of lipid bodies similar to that observed in neutrophils and macrophages under conditions of infection. However, qualitative changes in the cellular profiles of bioactive lipids and their precursors have not been extensively explored. This is an important question arising from the marked effects that lipid remodelling has upon the net pro-and anti-inflammatory capacity of cells such as mast cells [10][11][12][13][14].
Several studies have evaluated the location of the bioactive lipid precursor AA in mast cells, evaluating distribution between membrane phospholipid (phosphatidylcholine (PC), phosphatidylethanolamine (PE) and phosphatidylserine (PS)) and free fatty acid (FFA) forms of AA. The distribution within these pools changes in response to mast cell activation in response to both calcium ionophore and FcεRI stimulation [19,21,22]. These changes are functionally important as the location of the AA changes its proximity and availability to phospholipases that are concomitantly activated and catalyse pathways leading to synthesis of prostaglandins, leukotrienes, thromboxanes, HETEs, resolvins and endocannabinoids [13,23,24]. Similarly, functional importance is ascribed to the ratios between omega-3 and omega-6-fatty acids and the abundance of DHA and EPA pools. These are precursors for inflammation resolving factors (protectins, resolvins, and maresins). Since n-6 and n-3 fatty acids are generally regarded as pro-inflammatory, and anti-inflammatory, respectively, the cellular abundance of these forms in mast cells has consequences for tissue inflammatory responses [25,26].
In the context of metabolic syndrome and obesity, lipid remodelling has been studied in adipose tissue and adipocytes [20]. In addition to the formation and storage of large lipid droplets, adipose depot expansion and increased infiltration of pro-inflammatory cells such as macrophages, obesity is also linked to qualitative changes in adipocyte lipid content. In particular, switching towards a more proinflammatory lipid profile (increased ratio of polyunsaturated fatty acids (PUFA) to mono-unsaturated fatty acids (MUFA), altered ratio of DHA and EPA, altered AA levels. In addition, infiltrating macrophages remodel their lipid precursor pools when in obese adipose tissue contexts. However, the effect of an altered metabolic/glycemic environment upon mast cell lipid profiles has not been studied.
In prior studies we have shown that chronic insulin exposure is associated with elevated lipid body numbers, overall increases in cellular lipid content and elevated leukotriene C4 (LTC4) production in model mast cells [6][7][8][9]27]. In the current study, we use a similar model system in which a mucosal mast cell-like line is exposed to elevated insulin over a chronic time period. Our data suggest remodelling of the mast cell lipidome and altered functional release of bioactive lipids in response to chronic insulin elevations.

Functional assays
Resolvin, leukotriene and prostaglandin assays were purchased from Cayman Chemical (Ann Arbor, MI). Assays were developed according to the manufacturer's instructions, using cell pellets or supernatants derived from 2.5 million RBL2H3 per assay point, performed in triplicate. Mast cell degranulation was assayed as follows: RBL2H3 were plated in cluster plates at 5 × 10 [4] cells/well. Monolayers were washed and incubated in 200 µl Tyrode's buffer before stimulating as described. After 45 min at 37°C, 25 µl supernatant was removed, clarified by microcentrifugation, and transferred to a 96 well plate containing 100 µl per well 1 mM p-N-acetyl glucosamine (Sigma) in 0.05 M citrate buffer pH 4.5. After 1 h at 37°C reactions were quenched by the addition of 100 µl per well 0.2 M glycine, pH 9.0. Beta-hexosaminidase levels were read as OD at 405 nm. Results are shown as the mean ± standard deviation.

Imaging
Bright field and fluorescence imaging of cells in MatTek dishes (50,000 cells per cm 2 ) were performed on a Nikon Ti Eclipse C1 epi-fluorescence and confocal microscopy system, equipped with heated stage. Available laser lines in FITC, TxRed and Cy5 were supplied by a 488 nm 10 mW solid-state laser, a 561 nm 10 mW diode pump solid state laser and a 638 nm 10 mW modulated diode laser. Z stack size was 15 microns. Each z disc (optical section) was 150 nm. Pinhole size for all images was 60 microns. Images were analysed in NIS Elements (Nikon, Melville, NY).

Lipidomic analysis
The lipids were extracted by the method of Folch et al. using chloroform:methanol (2:1 v/v). For the separation of neutral lipid classes (FFA, triacyl glycerols (TAG), diacylglycerol (DAG), free cholesterol (FC), cholesterol ethers (CE)), a solvent system consisting of petroleum ether/diethyl ether/acetic acid (80:20:1, by vol) was employed. Individual phospholipid classes within each extract are separated by liquid chromatography (Agilent Technologies model 1100 Series). Each lipid class is transesterified in 1% sulfuric acid in methanol in a sealed vial under a nitrogen atmosphere at 100°C for 45 min. The resulting fatty acid methyl esters are extracted from the mixture with hexane containing 0.05% butylated hydroxytoluene and prepared for gas chromatography by sealing the hexane extracts under nitrogen. Fatty acid methyl esters are separated and quantified by capillary gas chromatography (Agilent Technologies model 6890) equipped with a 30 m DB 88 capillary column (Agilent Technologies) and a flame ionization detector. Whole cell lipidomic analysis was performed in collaboration with Metabolon, Inc.

Lysis and western blot
Cells were pelleted (2000 x g, 2 min) and washed once in ice-cold PBS. For total lysates, 1 × 10 7 cells were lysed (ice, 30 min) in 350 μl lysis buffer (50 mM HEPES [pH 7.4], 250 mM NaCl, 20 mM NaF, 10 mM iodoacetamide, 0.5% (w/v) Triton X-100, 1 mM phenylmethyl-sulphonylfluoride, 500 mg aprotinin/ml, 1.0 mg leupeptin/ml and 2.0 mg chymostatin/ml). Lysates were clarified (17,000 x g, 20 min). For preparation of total protein, lysates were acetone precipitated (1.4 vol acetone for 1 h at −20°C, followed by 10,000 x g for 5 min). Protein samples were resolved by 10% SDS-PAGE under reducing conditions. Resolved proteins were electrotransferred to PVDF membranes in 192 mM glycine/25 mM Tris (pH 8.8) solutions. For Western blotting, membranes were blocked using 5% non-fat milk in PBS for 1 h at RT. Primary antibodies were dissolved in PBS/0.05% Tween-20/0.05% NaN 3 and incubated with the membranes for 16 h at 4°C. After washing the membranes four times with PBS/ 0.1% Tween-20 (5 min/wash), developing antibodies (anti-rabbit or anti-mouse IgGs conjugated to horseradish peroxidase diluted to 0.1 μg/ml in PBS/0.05% Tween-20; Amersham) were placed with the membranes for 45 min at RT. Washes were then performed again, and the adherent secondary antibody then visualized using enhanced chemiluminescence solution (Amersham) and exposure of the membrane to BioMax film (Kodak, Rochester, NY). Films were then scanned at >600 dpi, and quantification of the band intensities performed using Image J (NIH, Bethesda, MD).

Mitochondrial staining
Mitotracker (Molecular Probes) was used at 1μM for 15 min at 37°C.

Statistical analysis
Results are shown as mean ± SD or standard error. Statistical significance was determined based on analysis of variance (ANOVA; GraphPad Prism 6 v6.02; San Diego, CA). Adjacent to datapoints in the respective graphs, significant differences were recorded as *p < 0.05, **p < 0.01, or ***p < 0.001. Data shown are all based on an n of at least three experiments.

Chronic insulin exposure results in lipid body accumulation in RBL2H3
Hyper-insulinemia was mimicked in vitro using chronic exposure to elevated insulin levels. Hyper-insulinemia in either escalated dose schedules (not shown) or singlesustained doses resulted in an altered phenotype in RBL2H3. As we have previously described, this phenotype includes the steatotic accumulation of lipid bodies [18,29,30] (Figure 1(a-d), associated with overall gains in lipid content. In the populations of cells used here, at least 80% of the cells were confirmed (by Oil Red O staining and microscopy) to have gained significant lipid body accumulation and exhibit a 'steatotic' staining pattern by the 6-d insulin exposure timepoint (Figure 1(e)). These data are consistent with both our prior studies and observations from adipocyte and immunocytes systems that see lipid body accumulation and overall gains in lipid content in response to insulin, nutrient excess and infection, respectively [6][7][8][9][31][32][33].

Comparison of fatty acid profiles in major lipid classes between control and insulin-treated cells
We used a lipidomic approach to obtain an overview of the fatty acid profile of nine major lipid classes in insulintreated and control RBL2H3. Figure 2 c. e.

Remodelling of AA, DHA and EPA pools in insulin-treated cells
The location of fatty acids within specific lipid pools is of functional significance for the cell. Whether fatty acids such as Arachidonic Acid (AA, 20:4n6), eicosapentenoic acid (EPA, 20:5n3) and Docosahexaenoic acid (DHA, 22:6n3) are in a free fatty acid pool, are esterified onto membrane phospholipids, or are associated with neutral lipid pools changes their ability to be accessed as precursors for the production of bioactive messengers. Organellar locations, for example defined by association with a primarily mitochondrial lipid such as cardiolipin, are also important determinants of lipid availability for signalling and metabolic purposes. We assessed the degree to which insulininduced remodelling between lipid pools and locations in RBL2H3. Figure 4(a-e) shows the effect on insulin on the distribution of AA, DHA, and EPA between various pools. We assessed the percentages of the total lipid-associated esterified onto phospholipid, associated with neutral lipid pools, present as free fatty acids or associated with c.
d. cardiolipin (shown in Figure 4(a)). Figure 4(b-d) shows in absolute quantities, the changes in the total DHA, EPA and AA and the changes in the amount associated with phospholipid (PL), neutral lipid (NL) free fatty acid (FFA) and (CL) pools. Figure 4(e) provides a representation normalized to the totals in each category, allowing us to visualize the proportional remodelling that occurs in each pool. We note that DHA remodels towards CL and away from PL. For EPA and AA the NL pool decreases and PL esterification increases.

Altered functional responses including bioactive lipid release in steatotic mast cells
We extended our previous data by evaluating a range of functional responses in control, 2-d and 6-d insulin-exposed cells. Figure 5(a) tabulates the effects of 2 and 6 d insulin exposure on release of histamine, the Series 4 leukotriene LTC4 and the Series 2 prostaglandin PGD2, Resolvin D1 and Resolvin E1 [17]. Figure 5(b) shows the insulin-induced upregulation of key biosynthetic enzymes in the LTC4 and Resolvin synthesis pathways. Figure 5(c) shows a dose response of the impact of insulin on Resolvin D1 and E1 levels.
Cardiolipin levels and mitochondrial/lipid body complexes are altered in insulin-exposed RBL2H3 Cardiolipin emerged from our analysis as a highly regulated lipid species following chronic insulin exposure. CL was overall increased in abundance by >3 folds and it follows that AA, DHA, EPA and LA all increased a. b. c.
d. their proportional representation in the CL lipid class. Given the mitochondrial function of CL, we would not have expected this lipid to be highly regulated in cells that were simply enriched in lipid bodies. However, there have been several studies showing that LB and mitochondria are physically interacting [34][35][36] and that there may be functional interactions between the two. In addition, omega-3 and omega-6 FA-induced remodelling in CL [37]. We asked if this correlated with increased mitochondrial load. Figure 6(a) shows that the mitochondrial marker Cytochrome Oxidase IV is upregulated in insulin-treated cells. Mitochondrial abundance is positively correlated with ORO staining intensity ( Figure 6  Mitotracker ( Figure 6(c), PCC 0.91) and we observe higher levels of Mitotracker fluorescence in the latter at the single cell level (Figure 6(d,e)).

Discussion
This study characterizes alterations in lipid classes following chronic insulin exposure in a model mast cell line. i. j. k.
e. c. acids. Insulin treatment resulted in a significant upregulation of the polyunsaturated fatty acids (PUFAs), specifically the omega-3 and omega-6 PUFAs. PUFAs, in particular the 18-20 carbon long omega-6 PUFAs, constitute the precursor molecules required for AA synthesis. AA is a precursor for production of eicosanoids such as prostaglandins, prostacyclins, thromboxanes and leukotrienes.
We have previously demonstrated that chronic insulin exposure drives the formation of lipid bodies in mast cells, with important collateral effects on mast cell function [39] including a decrease in the intensity of responses such as histamine secretion. In this study, our data suggest that mast cells chronically exposed to insulin in vitro also have marked alterations in their lipidomic landscape. Mast cells have been linked to a variety of pathologies defined by an underlying inflammatory component, and the role of mast cell-derived lipid messengers (leukotrienes, prostaglandins, thromboxanes, etc.) in these pathologies is clear [40]. The physiological roles of mast cells in a. b. c.
d. e. wound healing, regulation of smooth muscle tone and vasoregulation are also dependent upon these lipid messengers [40,41]. Obesity and type 2 diabetes are also inflammatory diseases, and excess adiposity is associated with adiposopathy, where the adipose itself becomes dysfunctional and inflamed [42]. Mast cells have been implicated in adipose inflammation [43], and mast cell stabilization may change outcomes in metabolic syndrome as well as the status of the adipose [44][45][46][47]. The finding that insulin changes mast cell phenotype [6][7][8][9] has relevance to the status of adipose-relevant mast cells and the mast cell-derived mediators (cytokines, growth factors and bioactive lipids) that affect adipose health.
There are commonalities and differences between the insulin-induced lipid profile changes in mast cells and those previously observed in the obese liver or adipocyte systems that are steatotic [48][49][50]. Isolated adipocytes from obese tissue, with expanded lipid droplet numbers, show large increases in PUFA and TAG levels [48]. Similarly, adipose tissue as a whole, and the macrophages isolated from it show increased PUFA levels in the obese setting and elevated TAG [50]. TAG are stored in mast cell LB [11]. Our analysis suggests that under the hyperinsulinaemic conditions modelled in our experiments, mast cells show increases in PUFA but marginal increases overall in TAG. The increase in PUFA that we observe, if translated to the mast cells found in vivo, could provide a new source for the PUFA that are dramatically elevated in adipose tissue.
As described above, we find the largest lipid increases in response to chronic insulin to occur in the polyunsaturated fatty acid class, which are the precursor molecules responsible for synthesizing a variety of pro-inflammatory and pro-thrombotic agents in the mast cell. More specifically, we observe significant upregulations in the particular fatty acids involved in the biosynthetic pathway of AA, including large increases in the AA molecule (20:4n6) itself. We also found (data not shown) remodelling of AA to different membrane phospholipids (PC, PE, PS) that are consistent with prior studies [19,51,52], and which changes their ability to act as precursors for certain mediators. At least under these experimental conditions, these data indicate that in situations of ectopic lipid deposition caused by chronic exposure to insulin, mast cells may be biased to producing the required lipid-derived precursor molecules necessary and sufficient to drive localized and systemic inflammatory responses. These seem to be supported by the functional assays that we performed to look at leukotriene, prostaglandin and resolvins released after FcεRI stimulation of control and steatotic cells [53][54][55]. The concomitant regulation of both pro-inflammatory and proresolution mediators is not paradoxical when we consider the stages of inflammation in a situation such as wound healing [41]. The initial establishment of an inflammatory site is guided by pro-inflammatory mediators but the 'focus' of the inflammation shifts over time to resolution and tissue repair. The increase in precursors measured lipidomically here appears to translate to real gains in the amounts of fully formed mediators that are released from the cell with impact on both the pro-inflammation and pro-resolution bioactive phases of processes such as wound healing [41]. Moreover, these gains may change outcomes of inflammation reactions in hyperinsulinemic patients [14].