Group X secretory phospholipase A₂ negatively regulates adipogenesis in murine models

Abstract

Studies in vitro indicate that group X secretory phospholipase A₂ (GX sPLA₂) potently releases arachidonic acid (AA) and lysophosphatidylcholine from mammalian cell membranes. To define the function of GX sPLA₂ in vivo, our laboratory recently generated C57BL/6 mice with targeted deletion of GX sPLA₂ (GX^−/− mice). When fed a normal rodent diet, GX^−/− mice gained significantly more weight and had increased adiposity compared to GX^+/+ mice, which was not attributable to alterations in food consumption or energy expenditure. When treated with adipogenic stimuli ex vivo, stromal vascular cells isolated from adipose tissue of GX^−/− mice accumulated significantly more (20%) triglyceride compared to cells from GX^+/+ mice. Conversely, overexpression of GX sPLA₂, but not catalytically inactive GX sPLA₂, resulted in a significant 50% reduction in triglyceride accumulation in OP9 adipocytes. The induction of genes encoding adipogenic proteins (PPARγ, SREBP-1c, SCD1, and FAS) was also significantly blunted by 50–80% in OP9 cells overexpressing GX sPLA₂. Activation of the liver X receptor (LXR), a nuclear receptor known to up-regulate adipogenic gene expression, was suppressed in 3T3-L1 and OP9 cells when GX sPLA₂ was overexpressed. Thus, hydrolytic products generated by GX sPLA₂ negatively regulate adipogenesis, possibly by suppressing LXR activation.—Li, X., Shridas, P., Forrest, K., Bailey, W., Webb, N. R. Group X secretory phospholipase A₂ negatively regulates adipogenesis in murine models.

the incidence of obesity is rapidly growing globally, with its prevalence reaching epidemic proportions, particularly in Western countries. The excess expansion of adipose tissue markedly increases the risk of type II diabetes and cardiovascular complications (1). However, lipodystrophy, the lack of adipose tissue, is also associated with insulin resistance and abnormal lipid metabolism (2, 3). In addition to its role as a central modulator of lipid homeostasis and energy balance, adipose tissue is also an important endocrine organ, producing a variety of cytokines and adipokines that regulate whole-body metabolic and inflammatory status (4, 5). Thus, the factors that regulate the formation and function of adipose tissue are critically important to human health.

The expansion of white adipose tissue (WAT) results from two processes, the conversion of precursor cells into differentiated adipocytes (hyperplasia), and the growth of individual fat cells due to increased storage of triglycerides (TGs) (hypertrophy). In humans, total body fat mass is determined by both adipocyte number and adipocyte size, although the total number of adipocytes increases only in childhood and adolescence (6). Despite an apparently fixed number of adipocytes, the adipocyte population nevertheless appears to undergo a remarkable rate of turnover in adults. According to one estimate, ∼10% of fat cells are renewed every year in adults, regardless of the size of the fat depot (6). Thus, in order to maintain a constant number of adipocytes, feedback mechanisms must be in place to regulate adipogenesis and adipocyte hypertrophy when energy balance is perturbed. In the setting of obesity, the failure to appropriately respond to the need to store excess energy in the form of TGs may lead to metabolic disorders (7, 8). Insights into the regulation of the adipogenic program at the molecular level may provide novel targets for manipulating the number of adipocytes and adipocyte expansion, thus ameliorating metabolic consequences of obesity.

The process of adipogenesis has been widely studied in vitro using well-characterized cell lines, including murine 3T3-L1 (9) and OP9 (10) preadipocyte cells. The differentiation of fibroblast-like preadipocytes into mature adipocytes is carried out by a tightly regulated transcription cascade that coordinates the induction of lipogenic and adipocyte-specific genes. The transcription factors CCAAT/enhancer binding proteins (C/EBPs) and peroxisome proliferator-activated receptor γ (PPARγ) are considered the master regulators of adipocyte differentiation (11–13). PPARγ, a member of the nuclear receptor family of ligand-activated receptors, has been shown to be indispensable for adipocyte differentiation. PPARγ forms a heterodimer with the retinoid X receptor (RXR) to interact with PPARγ response elements (PPREs) on adipocyte-specific genes (14, 15). Although PPARγ activation is critical for adipogenesis, accumulating evidence indicates that a number of other transcription factors can profoundly affect adipocyte differentiation, adipocyte-specific gene expression, and adipocyte lipid accumulation. For example, SREBP-1c mRNA is induced in 3T3-L1 cells 24 h after treatment with differentiation stimuli (16). Overexpression of SREBP-1c leads to enhanced adipogenesis, an effect that is PPARγ-dependent. Interestingly, overexpression of SREBP-1c in 3T3-L1 cells also induces the expression of PPARγ, indicating crosstalk between these two factors (17). More recently, liver X receptor (LXR)-responsive elements (LXREs) have been identified in the promoter region of PPARγ, linking the nuclear receptor LXR to adipogenesis (18). Treatment with the LXR agonist T0901317 increases TG accumulation and enhances the induction of adipogenic/lipogenic proteins PPARγ, genes encoding fatty acid-binding protein (FABP/aP2), sterol regulatory element binding protein-1c (SREBP-1c), and fatty acid synthase (FAS) in 3T3-L1 cells (18). Notably, mice deficient in LXRβ are protected from diet or age-induced obesity (19).

In this study, we report the novel finding that group X secretory phospholipase A₂ (GX sPLA₂) plays a regulatory role in adipogenesis. GX sPLA₂ is a member of the sPLA₂ family of enzymes that has been implicated in a number of biological processes based on its potent ability to hydrolyze mammalian cell membranes (20, 21). On the basis of loss-of-function studies in vivo and gain-of-function studies in vitro, we conclude that hydrolytic products generated by GX sPLA₂ impede the adipogenic program. We also provide evidence that this effect on adipogenesis is due, at least in part, to the suppression of LXR activation.

MATERIALS AND METHODS

Animals

Male C57BL/6 mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). GX sPLA₂ deficient (GX^−/−) mice were generated by InGenious Targeting Laboratory, Inc. (Stony Brook, NY, USA) using embryonic stem cells derived from C57BL/6 mice (22). Unless otherwise indicated, age-matched male GX^+/+ and GX^−/− mice produced from heterozygous breeding were used for all experiments. Mice were fed ad libitum normal rodent diet (Teklad Global 18% Protein Rodent Diet 2018S, 6.2% of calories from fat; Teklad Diets, Madison WI, USA) and housed in rooms with 14 h light and 10 h dark. For body weight and food consumption measurements, GX^+/+ and GX^−/− mice were singly caged at weaning. Daily food intake was monitored for 8 consecutive weeks beginning at 3 wk of age. The fat and lean mass of 1-yr-old mice were determined by dual energy X-ray absorptiometry (DEXA) (GE Lunar PIXImus, Madison, WI, USA), as described previously (23). For measurement of O₂ consumption and respiratory exchange ratio (RER), 8-mo-old male mice were individually caged in the LabMaster Metabolism Research Platform (TSE, Chesterfield, MO, USA); metabolic parameters were measured every 30 min for a total of 72 h. RER was calculated as Vco₂/Vo₂. Animal procedures were performed with the approval of the Lexington Veterans Affairs Medical Center Institutional Animal Care and Use Committee.

Mouse plasma measurements

Plasma leptin concentrations were determined using the Mouse Leptin Quantikine kit (R&D System, Minneapolis, MN, USA); plasma adiponectin was measured using the Mouse Adiponectin/Acrp30 Quantikine kit (R&D System). Total cholesterol (TC) concentrations in plasma were determined using the Cholesterol E kit (Wako, Richmond, VA, USA). Plasma free fatty acid (FFA) was estimated using HR Series NEFA-HR Color Reagents (Wako, Osaka, Japan). Plasma TG concentrations were determined using the Infinity Triglyceride kit (Thermo Scientific, Middletown, VA, USA).

Histology

Epididymal fat was collected from 6-mo-old GX^+/+ and GX^−/− mice. Tissues were fixed in 10% formaldehyde, paraffin embedded, cut into 5-μm sections, and stained with hematoxylin (Vector Laboratories, Burlingame, CA, USA). Adipocyte area was quantified for 3 randomly selected sections from 2 GX^+/+ and GX^−/− mice using Image-Pro computer software (Media Cybernetics, Bethesda, MD, USA).

Primary stromal vascular fraction (SVF) isolation and differentiation

The SVF was isolated from adipose tissues of 1-mo-old male GX^+/+ and GX^−/− mice, as described previously (24, 25). Briefly, tissue (1 g tissue/5 ml solution) was incubated for 1 h in Krebs buffer containing 1 mg/ml collagenase type I (Worthington, Lakewood, NJ, USA) and 1% fatty acid free BSA (Sigma, St. Louis, MO, USA). The digested tissue was spun at 500 g, and the SVF pellet was collected. SVF cells were differentiated into adipocytes ex vivo by a modification of a published protocol (25). Cells were plated (50,000 cells/cm²) and maintained in culture medium (DMEM/F12 1:1, 10% FBS, 8 mg/L biotin, 4 mg/L pantothenate, 2 mM l-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin) until 90% confluent, and then incubated for 2 d in culture medium supplemented with 0.1 μM insulin, 1 μM dexamethasone, and 0.5 mM IBMX. Cells were subsequently incubated for an additional 2 d in culture medium containing 0.1 μM insulin, and then maintained in culture medium without supplementation for another 3 d.

3T3-L1 and OP9 cultures

3T3-L1 cells were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and maintained in DMEM supplemented with 10% FBS, 2 mM l-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin. OP9 cells were generously provided by Dr. Jianhua Shao (University of California, San Diego, CA, USA). OP9 preadipocytes were maintained and differentiated into adipocytes, as described previously (10). Briefly, cells were grown in MEMα (Invitrogen, Grand Island, NY, USA) supplemented with 20% FBS, 2 mM l-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin, at a density of 7000 cells/cm². For differentiation, OP9 cells were cultured for 2 d postconfluence, and then they were incubated for 2 d in MEMα supplemented with 15% KnockOut SR (Invitrogen), 50 U/ml penicillin and 50 μg/ml streptomycin. Where indicated, cells were incubated with 5 μm T0901317 (Cayman, Ann Arbor, MI, USA), a specific agonist for LXR (26).

Generation of stably transfected preadipocyte cell lines

All of the results from studies of stably transfected OP9 cells were confirmed using two independently transfected and selected cell lines. The construction of expression plasmids was previously described (22). Briefly, a C-terminal Flag-tagged GX sPLA₂ cDNA (GX-Flag) was inserted into the mammalian expression vector pcDNA 3.1 (Invitrogen). An expression plasmid encoding Flag-tagged inactive GX sPLA₂ (H46Q-Flag) (27) was generated using the QuickChange kit (Stratagene, La Jolla, CA, USA). All expression plasmids were confirmed by DNA sequencing (Davis Sequencing, Davis, CA, USA). OP9 and 3T3-L1 preadipocytes were plated in 6-well plates at a density of 7000 cells/cm². At ∼50% confluence, cells were transfected with 4 μg of expression plasmids using the Optifect reagent (Invitrogen) according to the manufacturer's protocol. Transfected cells were maintained in culture medium containing 500 μg/ml G418 (Invivogen, San Diego, CA, USA) to select for control cells (OP9-C, 3T3-C cells) and cells expressing active (OP9-GX, 3T3-GX cells) or inactive (OP9-H46Q cells) GX sPLA₂. Phospholipase A₂ activity in 48 h-conditioned medium from the transfected cells was determined using the sPLA₂ Assay Kit (Cayman) and normalized to cellular protein, which was determined using BCA protein assay reagent (Pierce, Rockford, IL, USA).

Cellular TG measurements

Total lipid was extracted from cells using hexane:isopropanol (3:2, v/v), dried under a stream of nitrogen gas, and solubilized in chloroform containing 1% Triton-X100. Samples were dried again and resuspended in water. TG was quantified using the Infinity kit and normalized to cellular protein.

Oil red O staining

Cells were washed twice with PBS and fixed in 10% formaldehyde for 15 min, rinsed with 60% isopropyl alcohol, and stained with 0.3% oil red O in 60% isopropanol for 30 min. Cells were counterstained in hematoxylin for 30 s and mounted in 30% glycerol.

RNA isolation and real-time PCR

Total RNA was isolated from cells and tissues using TriZol reagent (Molecular Research Center, Cincinnati, OH, USA) as recommended by the manufacturer. The reverse transcriptase reaction was carried out using 0.2 μg of RNA and a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems, Warrington, UK) and the following parameters: 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C for 40 cycles, and a final extension step at 72°C for 10 min. The sequences of primers used for RT-PCR are presented in Supplemental Table S1.

TG lipolysis and synthesis measurements

The rate of TG turnover was measured by a modification of a previously described procedure (28). Briefly, confluent monolayers of OP9-C, OP9-H46Q, and OP9-GX preadipocytes were cultured overnight in delipidated medium [DMEM/low glucose, 10% delipidated serum (SeraCare, Milford, MA, USA), penicillin and streptomycin], and then incubated in the same medium supplemented with 0.37 μCi/ml ³H-oleic acid (PerkinElmer, Boston, MA, USA) for 16 h. The labeling medium was then replaced with fresh delipidated medium containing 6 μM triacsin C (Biomol Research Laboratories, Plymouth Meeting, PA, USA) to block re-esterification of fatty acids. Cells were collected at selected intervals up to 32 h in homogenization buffer (250 mM sucrose, 1 mM EDTA, and 20 mM Tris-HCl, pH 7.4). Lipid was extracted in chloroform:methanol (2:1, v/v) with 0.6 mg/ml triolein added as a carrier. Samples were applied to Polygram silicon membranes (Macherey-Nagel, Easton, PA, USA) and separated by thin-layer chromatography (TLC) in hexane:ethyl ether:acetic acid (80:20:1, v/v/v). Lipid bands were visualized by exposing the TLC plates to I₂ vapor. The amount of ³H-oleate remaining in the TG pool was measured in a γ counter (Cobra II; Packard Instruments, Minneapolis, MN, USA), and normalized to cellular protein.

The rate of incorporation of ³H-oleate into TG was measured in OP9 cells treated with diethyl-p-nitrophenyl phosphate (E600) (Sigma), which had been shown previously to block lipolysis in 3T3-L1 preadipocytes (28). To confirm the effectiveness of E600 in OP9 cells, we performed control experiments to measure TG turnover, as described above, except that in addition to triascin C, 0.6 mM E600 was present during a 24-h chase period. For measuring TG synthesis, confluent OP9-C, OP9-H46Q, and OP9-GX cells were incubated for 4 h with delipidated medium containing 0.37 μCi/ml ³H-oleate, 400 μM oleate complexed to BSA, and 0.6 mM E600. The amount of ³H-oleate incorporated in TG was determined by TLC as described above. For measurements of oleate incorporation into diacylglycerol (DAG), 0.6 mg/ml DAG was added as a carrier, and the lipid extracts were analyzed in the same TLC solvent system as for TG. For phosphatidylcholine (PC) measurements, PC was added as carrier, and the lipid extracts were separated in chloroform:methanol:water (65:25:4, v/v/v).

LXRE reporter assays

LXRE reporter and mLXRα plasmids were generously provided by Dr. Peter Tontonoz (University of California, Los Angeles). OP9-C, OP9-GX, 3T3-C, and 3T3-GX cells were transfected with an Amaxa Cell Line Nucleofector Kit V (Lonza, Cologne, Germany) following the manufacturer's protocol using 0.5 μg pTK-3xLXRE-Luc reporter construct, 0.5 μg mLXRα and 0.01 μg Renilla luciferase (pRL-TK; Promega, Madison, WI, USA) per 1 × 10⁶ cells. Additional reporter assays were performed in OP9 and 3T3-L1 preadipocytes transiently expressing GX sPLA₂, in which case cells were transfected with 0.5 μg GX-Flag or control expression plasmid along with pTK-3xLXRE-Luc, mLXRα, and Renilla luciferase, as described above. At 24 h after transfection, fresh medium containing 5 μM T0901317 or an equal volume of vehicle (DMSO) was added to the cells and incubated for 8 h. Luciferase activities were analyzed using the Dual-Luciferase Reporter Assay System (Promega). Transfection efficiency was normalized using Renilla luciferase activity.

Statistical analysis

Mean ± se values were calculated for each parameter. Statistical significance in experiments comparing only 2 groups was determined by a 2-tailed Student's t test. Unless otherwise indicated, the significance of the difference in mean values among >2 groups was evaluated by 2-way ANOVA using a Tukey post hoc test (SigmaStat; Systat Software, San Jose, CA, USA). For assessment of weight gain, data was analyzed by 2-way ANOVA with repeated measures. Statistical comparison between 2 genotypes within the same treatment group and between different treatment groups within the same genotype was carried out for each experiment, and all significant differences (P<0.05) are given in the figures and/or figure legends.

RESULTS

GX sPLA₂-deficient mice have increased adiposity

We recently developed C57BL/6 mice with targeted deletion of GX sPLA₂ (GX^−/− mice). Unexpectedly, GX^−/− mice fed a normal rodent diet for 12 mo were heavier compared to age-matched C57BL/6 mice (Fig. 1A). This finding prompted us to monitor body weights of GX^+/+ and GX^−/− littermates that were singly housed at weaning until 10 mo of age. Results from this study confirmed that deficiency in GX sPLA₂ leads to significantly increased weight gain in mice (Fig. 1B). On the basis of results from DEXA, the difference in body weight appears to be primarily due to an increase in adiposity in GX^−/− mice; lean mass was not significantly different between the two strains (Fig. 1C). Histological staining of epididymal fat sections from 6-mo-old mice showed that GX^−/− mice have enlarged adipocytes compared to GX^+/+ mice (GX^+/+ mice: 3630±95 μm²; GX ^−/− mice: 6290±144 μm², P<0.0001) (Fig. 1D, E). Unlike some genetic models of obesity, the increase in adiposity did not appear to be the result of altered food consumption (Fig. 1F). No difference in oxygen consumption or RER was detected in GX^−/− mice compared to GX^+/+ mice (Supplemental Fig. S1), indicating that differences in adiposity are not due to major alterations in energy balance. GX sPLA₂ deficiency did not result in any apparent differences in plasma TG, free fatty acids (FFAs), adiponectin, or leptin concentrations (Supplemental Table S2), although there was a trend for increased leptin in GX^−/− mice (P=0.06). There was a modest, but significant decrease in plasma TC in 3-mo-old GX^−/− mice (Supplemental Table S2). However, lipoprotein cholesterol distributions were similar in GX^−/− and C57BL/6 mice (Supplemental Figure S2). Differences in plasma TC were not observed in other cohorts of mice we analyzed that were older than 3-mo (data not shown).

Figure 1.

Increased adiposity in GX^−/− mice. A) Male C57BL/6 and GX^−/− mice in C57BL/6 background were fed a normal rodent diet for 12 mo. B) Male GX^+/+ and GX^−/− littermates were singly housed at weaning, and body weights were determined weekly (n=4). C) Body composition of 12-mo-old male mice was determined by DEXA (n=3). D) Histological staining of paraffin-embedded epididymal fat sections from 6-mo-old mice. E) Mean ± se adipocyte area was determined for 3 randomly chosen sections, 2 mice/strain. F) Daily food consumption by male GX^+/+ and GX^−/− mice was measured for 8 wk, starting at 3 wk of age (n=4). *P < 0.05; ***P < 0.0001. n.s., not significant.

SVF cells from GX^−/− mice accumulate more TG when differentiated into adipocytes ex vivo compared to GX^+/+ mice

To our knowledge, the expression of GX sPLA₂ in adipose tissue has not been previously reported. Analysis by qRT-PCR demonstrated GX sPLA₂ mRNA expression in adipose tissue, at levels comparable to other tissues reported to express it, such as small intestine, brain, thymus, and spleen (29) (Supplemental Fig. S3). Our data indicated that GX sPLA₂ mRNA was most abundant in testis and large intestine, consistent with earlier reports (29). Adipose tissue is heterogeneous, consisting not only of mature adipocytes, but also other cell types (collectively known as the SVF), including fibroblasts, vascular cells, mesenchymal cells, and inflammatory cells. Therefore, adipose tissue from 1-yr-old male C57BL/6 mice was separated into SVF and adipocytes to analyze the expression and distribution of GX sPLA₂ mRNA. RNAs isolated from the SVF and adipocytes were enriched more than 100-fold in F4/80 and aP2, respectively, confirming the effectiveness of our separation procedure (Supplemental Fig. S4). GX sPLA₂ mRNA was detected in both adipocytes and SVF (Fig. 2A). When treated with adipogenic stimuli ex vivo, SVF cells were efficiently differentiated into mature adipocytes, as evidenced by the presence of abundant intracellular lipid droplets (Supplemental Fig. S5). GX sPLA₂ mRNA could be detected at comparable levels throughout all stages of SVF differentiation into adipocytes (Fig. 2B). Notably, SVF from GX^−/− mice accumulated significantly more TG during differentiation compared to GX^+/+ mice (Fig. 2C). These data suggest that the increased adiposity of GX^−/− mice may be due at least, in part, to a direct effect in adipose tissue, rather than a systemic consequence of GX sPLA₂ deficiency.

Figure 2.

Deficiency of GX sPLA₂ leads to increased TG accumulation during adipocyte differentiation. A) Adipocytes and SVF cells were isolated from collagenase-digested epididymal fat from 1-yr-old C57BL/6 mice. Data are means ± se of values from 4 mice. B) Cultured SVF cells from 5-wk-old male C57BL/6 mice were treated with adipogenic stimuli, as described in Materials and Methods. RNA was prepared from cells at the indicated time of differentiation for real-time RT-PCR. C) SVF cells from 5-wk-old GX^+/+ and GX^−/− mice were plated and differentiated into mature adipocytes. At the indicated time of differentiation, cellular lipid was extracted, and TG was quantified. Data are means ± se of triplicate samples and are representative of 4 independent experiments. *P < 0.05. TG content in both cell types was significantly higher at d 7 of differentiation compared to d 4; P < 0.001. n.s., not significant.

Overexpression of GX sPLA₂ reduces TG accumulation in OP9 cells

The data from GX^−/− mice suggested the interesting possibility that GX sPLA₂ may regulate adipogenesis. We next turned to an established in vitro model, OP9 cells (10), to investigate the molecular mechanism of this regulation. When incubated with adipogenic stimuli, OP9 preadipocytes were effectively differentiated into mature adipocytes, as evidenced by the accumulation of lipid droplets and the up-regulation of PPAR-γ and aP2, key transcription factors that drive adipocyte differentiation (Supplemental Fig. S6A, B). GX sPLA₂ mRNA was detected in OP9 preadipocytes, and its expression was not significantly altered during differentiation (Supplemental Fig. S6C). GX sPLA₂ mRNA was similarly detected in 3T3-L1 cells before, during and after adipocyte differentiation (data not shown).

To study the role of GX sPLA₂ in adipogenesis, we established cell lines stably transfected with either control expression vector (OP9-C) or vector expressing wild-type (OP9-GX) or catalytically inactive (OP9-H46Q) GX sPLA₂. OP9-GX preadipocytes secreted ∼80% more phospholipase activity into the medium compared to OP9-H46Q and OP9-C cells (Supplemental Fig. S7). Interestingly, when incubated under conditions that induce adipocyte differentiation (10), OP9-GX cells, but not OP9-H46Q cells, accumulated significantly less TG compared to OP9-C cells, as evidenced by oil red O staining (Fig. 3A) and biochemical determinations of cellular TG (Fig. 3B). Thus, the ability of GX sPLA₂ to decrease TG accumulation in OP9 adipocytes was consistent with our observation that GX sPLA₂ deficiency in vivo leads to adipocyte hypertrophy.

Figure 3.

Overexpression of GX sPLA₂ in OP9 cells leads to decreased TG accumulation during differentiation. OP9 preadipocytes stably transfected with control expression plasmid (OP9-C), GX sPLA₂ (OP9-GX) or catalytically inactive GX sPLA₂ (OP9-H46Q) were differentiated into adipocytes as described in Materials and Methods. A) After 4 d of differentiation, cells were stained with oil red O. B) At the indicated day of differentiation, lipid was extracted, and TG was quantified (n=3). ***P < 0.001 vs. OP9-C. C) Incorporation of ³H-oleate into TG in 4 h was measured in OP9-C, OP9-H46Q, and OP9-GX cells as described in Materials and Methods. Data are means ± se of triplicate samples and are representative of 5 independent experiments. **P < 0.01 vs. OP9-C (1-way ANOVA).

TG synthesis is reduced in OP9-GX cells

We investigated whether reduced TG accumulation in OP9-GX cells was due to decreased TG synthesis or increased lipolysis. To assess TG synthesis rate, the incorporation of ³H-oleate into TG was measured in OP9-C, OP9-H46Q, and OP9-GX cells in the presence of 0.6 mM E600, which had been previously shown to block lipolysis in 3T3-L1 cells (28). We confirmed that lipolysis is inhibited in E600-treated OP9 cells for at least 6 h (Supplemental Fig. S8). Compared to OP9-C and OP9-H46Q cells, OP9-GX cells incorporated significantly less ³H-oleate into TG during a 4-h labeling period, indicating that TG synthesis is reduced in OP9-GX cells (Fig. 3C). On the other hand, the incorporation of ³H-oleate into DAG and PC was not significantly different between OP9-GX and OP9-C cells (Supplemental Fig. S9A, B). There was also no evidence that the rate of turnover of the ³H-oleate-labeled TG pool was significantly different between OP9-C and OP9-GX cells (data not shown). Thus, GX sPLA₂ appears to reduce TG accumulation in OP9 cells by blunting lipogenesis rather than enhancing lipolysis.

Induction of adipogenic genes is reduced in OP9-GX cells

We next investigated whether reduced TG synthesis in OP9-GX cells was accompanied by alterations in lipogenic gene expression. Prior to adipocyte differentiation, the expression of SREBP-1c was significantly lower in OP9-GX cells compared to OP9-C cells (Fig. 4A). Expression of the remaining genes examined, FAS, stearoyl-coenzyme A desaturase 1 (SCD1), PPAR-γ, aP2, and diacylglycerol O-acyltransferase 1 (DGAT1) was similar in OP9-C preadipocytes and OP9-GX preadipocytes (Fig. 4B–F). As expected, the expression of each of these genes was significantly increased compared to baseline levels in OP9-C cells after treatment with adipogenic stimuli. Interestingly, the induction of all of the genes examined was significantly blunted in OP9-GX cells compared to OP9-C cells. The effect of GX sPLA₂ to lower adipogenic gene expression was most pronounced for SREBP-1c, FAS, and SCD1, where expression was reduced 70–80% compared to control values.

Figure 4.

Induction of adipogenic genes is reduced in OP9-GX cells. RNA was isolated from OP9-C and OP9-GX cells either before (d 0) or after (d 3) differentiation. Real-time RT-PCR was performed using primers specific for the genes SREBP-1c (A), FAS (B), SCD1 (C), PPARγ (D), aP2 (E), and DGAT1 (F). Data are means ± se of triplicate samples and are representative of 3 independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

The adipogenic effect of the LXR agonist T0901317 is reduced in OP9-GX cells compared to OP9-C cells

Published studies in our laboratory indicate that hydrolytic products generated by GX sPLA₂ suppress LXR target gene expression (22). Given that SREBP-1c, SCD1, and FAS are known LXR targets (30–32), and LXR activation promotes adipogenesis and increases TG accumulation in adipocytes (18, 33), we investigated the possibility that GX sPLA₂ reduces TG accumulation in OP9 cells by suppressing LXR activation. OP9-C and OP9-GX cells were differentiated in the presence or absence of 5 μM T0901317, and markers of adipogenesis were assessed. As published previously for 3T3-L1 cells and human primary preadipocytes (18, 33, 34), the addition of the LXR agonist during 2 d of differentiation significantly increased TG accumulation in OP9-C cells (Fig. 5A). This effect was totally abolished in cells overexpressing GX sPLA₂ (Fig. 5A). Furthermore, the ability of the LXR agonist to increase the expression of adipogenic genes was significantly blunted in OP9-GX cells (Fig. 5B–D). Interestingly, T0901317 treatment did not result in an enhanced induction of PPARγ (Fig. 5E) during the differentiation of either OP9-C or OP9-GX cells, in contrast to what has been shown in 3T3-L1 cells (18). This lack of effect of the LXR agonist may be due to the fact that OP9 cells are at a later stage of adipocyte differentiation compared to 3T3-L1 cells and express PPARγ even in the absence of adipogenic stimuli (10). In the case of 3T3-L1 cells, LXR agonists failed to have an effect on PPARγ expression at a later stage of differentiation (18). Taken together, we interpret our results to suggest that one mechanism by which GX sPLA₂ reduces TG accumulation in adipocytes is to suppress the adipogenic effect of LXR.

Figure 5.

Adipogenic effect of LXR agonist is reduced in OP9-GX cells. OP9-C and OP9-GX cells were incubated in differentiation medium supplemented with 5 μM T0901307 (T09) or vehicle. A) After 2 d, cellular lipid was extracted, and cellular TG content, normalized to cellular protein, was determined. B–E) RNA was collected at d 1 of differentiation, and the expression of adipogenic genes SREBP-1c (B), FAS (C), SCD1 (D), and PPARγ (E) was quantified by real-time RT-PCR. Data are means ± se of triplicate samples and are representative of 2 independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

GX sPLA₂ overexpression reduces LXR luciferase reporter activity in adipocytes

To examine directly the effect of GX sPLA₂ on LXR activity in adipocytes, 3T3-L1 and OP9 preadipocytes were cotransfected with mouse LXRα, a luciferase reporter construct containing a synthetic LXRE and Renilla luciferase along with GX sPLA₂ or control expression vector. One day after transfection, cells were treated with 5 μM T0901317 or vehicle control. In the absence of LXR agonist, 3T3-L1 and OP9 cells overexpressing GX sPLA₂ exhibited reduced LXRE promoter activity compared to their corresponding controls, although this effect was only significant in 3T3-L1 cells (Fig. 6A, B). The addition of the agonist failed to activate reporter activity in 3T3-L1 and OP9 cells overexpressing GX sPLA₂ to the same extent as their corresponding control cells. These findings were confirmed in additional experiments whereby GX sPLA₂ was stably expressed in 3T3-L1 and OP9 cells (Supplemental Fig. S10).

Figure 6.

Reporter activity in transiently transfected 3T3-L1 and OP9 cells: A) 3T3-L1 cells were transiently transfected with plasmids encoding pTK-3xLXRE-Luc, mLXRα, and Renilla luciferase, along with an empty vector expression plasmid (C, control) or an expression plasmid encoding GX sPLA₂ (GX). Luciferase activity was measured and presented as values after normalization with Renilla luciferase activity. Data are means ± se of triplicate samples and are representative of 3 independent experiments. B) OP9 cells were transfected with plasmids listed in A. Luciferase activity was measured and normalized to Renilla. Results were confirmed using stably transfected preadipocytes (Supplemental Fig. S7). Data are means ± se of triplicate samples. **P < 0.01; ***P < 0.001.

DISCUSSION

In the present study, we report for the first time that male GX^−/− mice fed a normal rodent diet had increased weight gain compared to GX^+/+ mice, which was attributable to increased adiposity and enlarged adipocytes. This finding contrasts to a previously published study, which reported no difference in body weight between GX^+/+ and GX^−/− mice (35). In our experience, the body weight phenotype is not clearly evident until mice have reached 4–5 mo of age or when mice are group housed, due to large variations in weights of individual mice within a cage. We do not know whether the authors of the earlier report monitored mice that were singly caged for 20–30 wk. Although we did not comprehensively study the gender specificity of this phenotype, results from DEXA analyses indicated a similar propensity for increased adiposity in female GX^−/− mice (data not shown). Adipocyte hypertrophy did not appear to arise from alterations in energy balance, since differences in daily food intake or oxygen consumption were not detected in GX^−/− mice. Rather, the expansion of adipocytes appeared to be related to an alteration in fat storage. When forced to undergo adipocyte differentiation ex vivo, SVF cells from adipose tissue of GX^−/− mice accumulated significantly more TG compared to cells from GX^+/+ mice. However, on the basis of oil red O staining of intracellular lipid droplets, there was no evidence that GX sPLA₂ deficiency altered the ability of SVF cells to acquire an adipocyte phenotype.

Our data indicate that GX sPLA₂ is expressed in both mature adipocytes and SVF of epididymal fat, although transcripts were relatively more abundant in the SVF. We did not investigate which specific cell types in the SVF express GX sPLA₂. In addition to fibroblasts, endothelial cells, and inflammatory cells, the SVF contains a population of cells capable of differentiating into several lineages (25). Interestingly, when SVF cells from 1-mo-old wild-type mice were incubated with adipogenic stimuli ex vivo, most of the cells acquired the morphology of adipocytes and continued to express relatively stable amounts of GX sPLA₂ mRNA, indicating that GX sPLA₂ is not transcriptionally regulated during adipocyte conversion. This finding was corroborated in 3T3-L1 and OP9 cells, where GX sPLA₂ mRNA was expressed at comparable levels in preadipocytes and throughout adipocyte differentiation. Taken together, our data provide strong evidence that GX sPLA₂ expressed by adipocytes regulates the adipogenic program. However, given published data that macrophages secrete GX sPLA₂ (29), it seems likely that macrophages may also be a source of GX sPLA₂ in adipose tissue. We speculate that macrophage-derived GX sPLA₂ may act in a paracrine manner to regulate adipocyte expansion. This regulation may be particularly important in the setting of obesity, when macrophage accumulation in adipose tissue is known to occur (36, 37). Further studies are necessary to determine whether GX sPLA₂ is post-transcriptionally regulated in adipose tissue. Unlike most members of the sPLA₂ family, GX sPLA₂ is expressed in an inactive form that requires removal of an N-terminal propeptide for catalytic activity (29, 38). The identity of the proteases responsible for the proteolytic processing and activation of GX sPLA₂ is currently unknown. Interestingly, recent data from transgenic mice suggest that cleavage of the precursor form of GX sPLA₂ is under tight regulation, and that GX sPLA₂ may be activated during inflammation (39).

Although oil red O staining did not reveal a marked defect in the ability of GX^−/− SVF to undergo adipocyte conversion, these cells accumulated significantly more TG during differentiation compared to GX^+/+ SVF. In our experience, biochemical measurements are a more sensitive and specific method for determining cellular TG content than oil red O, which stains all neutral lipids. OP9 cells stably transfected with GX sPLA₂ cDNA accumulated significantly less TG during differentiation compared to control cells, consistent with the conclusion that GX sPLA₂ negatively regulates the capacity of adipocytes to store fat. The effect of GX sPLA₂ on TG accumulation was not observed in OP9 cells overexpressing H46Q, a mutated variant lacking catalytic activity. The incorporation of ³H-oleate into TG was significantly reduced in OP9-GX cells compared to control cells, indicating that GX sPLA₂ overexpression results in reduced lipogenesis. There was no evidence that the rate of lipolysis was altered by GX sPLA₂. Additional studies showed that GX sPLA₂ affects the incorporation of FFA into the TG pool but not overall utilization of FFA for PL synthesis or cholesterol esterification (data not shown).

We also showed that GX sPLA₂ overexpression leads to significantly reduced expression of adipogenic genes, including PPARγ, SREBP-1c, FAS, and SCD1 in OP9 adipocytes. Thus, GX sPLA₂ appears to inhibit TG accumulation by interfering with the adipogenic program. Extensive investigations have established the role of PPARγ in adipogenesis and adipocyte function (40, 41). Activation of PPARγ in preadipocytes initiates a transcriptional cascade that signals the cells to differentiate and acquire the full range of adipocyte functions, including the ability to produce TG from FFA derived either from circulating lipoproteins or de novo lipogenesis. In the case of OP9 preadipocytes, PPARγ expression is readily detected in cells grown to confluence, indicating that the adipogenic differentiation program is already under way (10). When confluent OP9 cells are exposed to the appropriate stimuli, PPARγ is rapidly induced, and the cells acquire the morphology and adipocyte marker protein expression of mature adipocytes (10) (See also Supplemental Fig. S6A, B). A role for LXRs in adipogenesis has also been investigated in a number of studies. In the liver, LXRs modify the expression of lipogenic genes such as FAS either through direct interactions with the target gene promoter (32) or by regulating SREBP-1c expression (30, 42). Both LXRα and LXRβ are highly expressed in adipose tissue, where they appear to regulate distinct sets of adipocyte genes (43). LXRα expression is induced by PPARγ during adipocyte differentiation, whereas LXRβ expression remains relatively constant (18, 33, 44). According to some reports, PPARγ is a target of LXR, and activation of LXR stimulates adipocyte differentiation by up-regulating PPARγ expression (18). Thus, LXR has been proposed to function in adipocyte differentiation by promoting the expression of both adipocyte-specific genes (through PPARγ) and lipogenic genes (directly, or through SREBP-1c). However, a subsequent investigation in human adipocytes indicated that LXR activation promotes TG accumulation by strongly stimulating SREBP-1c and lipogenic gene expression without activating the differentiation process through PPARγ (34); other published studies have reported no role for LXR in mediating adipogenesis (44, 45). In our studies, the LXR agonist T0901317 resulted in significantly increased TG accumulation in differentiating mouse OP9 cells (Fig. 5A). This effect was accompanied by a significant up-regulation of the lipogenic genes SREBP-1c, FAS, and SCD1, but no change in PPARγ expression in cells after 1 d of differentiation. Notably, GX sPLA₂ overexpression markedly reduced the effect of LXR agonist to enhance TG accumulation and adipogenic gene expression during OP9 adipocyte differentiation (Fig. 5).

Our results from synthetic LXR promoter luciferase constructs provide evidence that GX sPLA₂ may negatively regulate adipogenesis by inhibiting the ability of LXR to transactivate its target promoters. Consistent with our conclusion that GX sPLA₂ suppresses LXR, the phenotype of GX^−/− mice (adipocyte hypertrophy) is opposite to LXRβ-deficient mice, which are resistant to the increase in adipocyte size associated with aging or high-fat diet feeding (19). We speculate from our data that hydrolytic products generated by GX sPLA₂, most likely AA, suppress the expression of LXR target genes, including SREBP-1c in adipocytes by negatively regulating LXR activation. This conclusion is in keeping with other studies showing that polyunsaturated fatty acids, especially AA, block LXR-mediated gene induction (46, 47). Gel shift mobility and ligand binding domain activation assays indicate that AA competes with LXR agonists to block LXR activation (46). Nevertheless, we cannot rule out the possibility that GX sPLA₂ modulates adiposity through additional mechanisms, including the production of lipid mediators such as prostaglandins and lysophosphatidylcholine, which are known to have effects on adipose tissue (48, 49). The release of hydrolytic products from lipoprotein particles should be considered, given the potent ability of GX sPLA₂ to act on HDL and LDL (50, 51).

In recent studies in our laboratory, we determined that GX sPLA₂ also acts in mouse adrenals to negatively regulate the expression of the LXR target gene, steroidogenic acute regulatory protein (StAR) (22). StAR mediates the rate-limiting step of corticosteroid synthesis by delivering cholesterol to the inner mitochondrial membrane, where steroidogenic enzymes are localized. Interestingly, GX^−/− mice have ∼80% higher plasma corticosterone concentrations compared to GX^+/+ mice under both basal and ACTH-induced stress conditions (22). Given that glucocorticoids exert pleiotrophic effects on whole body metabolism and energy partitioning, we cannot rule out the possibility that alterations in adrenal function may contribute to the increased adiposity in GX^−/− mice. However, we found no evidence that GX^−/− mice had altered food intake or energy expenditure, as would be expected if increased corticosterone levels were having a central effect on adiposity (52). In addition, data from our in vitro gain-of-function and ex vivo loss-of-function studies provide strong evidence that reduced TG accumulation in adipose tissue is due, at least in part, to a direct effect of GX sPLA₂ on adipocytes.

In a recent study, it was reported that oral administration of the nonselective PLA₂ inhibitor methyl indoxam suppressed diet-induced obesity and glucose intolerance in C57BL/6 mice (53). The authors attributed this effect to the inhibition of group 1B phospholipase A₂ (GIB sPLA₂) in the intestinal lumen. GIB sPLA₂ is a pancreatic enzyme that contributes to diabetes and obesity by generating the digestive product lysophospholipid, which when absorbed postprandially contributes directly to glucose intolerance and hyperglycemia (54). Our findings with GX sPLA₂ are in interesting contrast to recent studies investigating another adipocyte-expressed phospholipase A₂, AdPLA. AdPLA is a membrane-associated intracellular PLA₂ that is highly enriched in WAT and induced during adipocyte differentiation (55). AdPLA-deficient mice have significantly reduced adipose tissue mass and TG content but apparently normal adipogenesis. The phenotype of AdPLA-deficient mice was attributed to increased lipolysis in adipose tissue through a mechanism involving the downstream AA metabolite prostaglandin E2 (48). In the current study, we provide strong evidence that GX sPLA₂, through a different mechanism, negatively regulates the adipogenic program and suppresses the ability of adipocytes to store lipid. Future studies will define the role of GX sPLA₂ in the metabolic consequences of obesity.