Synergistic Suppression of Early Phase of Adipogenesis by Microsomal PGE Synthase-1 (PTGES1)-Produced PGE2 and Aldo-Keto Reductase 1B3-Produced PGF2α

We recently reported that aldo-keto reductase 1B3-produced prostaglandin (PG) F2α suppressed the early phase of adipogenesis. PGE2 is also known to suppress adipogenesis. In this study, we found that microsomal PGE2 synthase (PGES)-1 (mPGES-1; PTGES1) acted as the PGES in adipocytes and that PGE2 and PGF2α synergistically suppressed the early phase of adipogenesis. PGE2 production was detected in preadipocytes and transiently enhanced at 3 h after the initiation of adipogenesis of mouse adipocytic 3T3-L1 cells, followed by a quick decrease; and its production profile was similar to the expression of the cyclooxygenase-2 (PTGS2) gene. When 3T3-L1 cells were transfected with siRNAs for any one of the three major PTGESs, i.e., PTGES1, PTGES2 (mPGES-2), and PTGES3 (cytosolic PGES), only PTGES1 siRNA suppressed PGE2 production and enhanced the expression of adipogenic genes. AE1-329, a PTGER4 (EP4) receptor agonist, increased the expression of the Ptgs2 gene with a peak at 1 h after the initiation of adipogenesis. PGE2-mediated enhancement of the PTGS2 expression was suppressed by the co-treatment with L-161982, a PTGER4 receptor antagonist. Moreover, AE1-329 enhanced the expression of the Ptgs2 gene by binding of the cyclic AMP response element (CRE)-binding protein to the CRE of the Ptgs2 promoter; and its binding was suppressed by co-treatment with L-161982, which was demonstrated by promoter luciferase and chromatin immunoprecipitation assays. Furthermore, when 3T3-L1 cells were caused to differentiate into adipocytes in medium containing both PGE2 and PGF2α, the expression of the adipogenic genes and the intracellular triglyceride level were decreased to a greater extent than in medium containing either of them, revealing that PGE2 and PGF2α independently suppressed adipogenesis. These results indicate that PGE2 was synthesized by PTGES1 in adipocytes and synergistically suppressed the early phase of adipogenesis of 3T3-L1 cells in cooperation with PGF2α through receptor-mediated activation of PTGS2 expression.


Introduction
Obesity contributes to insulin resistance and type 2 diabetes mellitus [1,2]. As a major target of insulin action, adipose tissue plays a critical role in the regulation of whole body metabolism and glucose homeostasis [3,4]. Adipogenesis has been extensively studied, and several key transcription factors involved in the regulation of adipogenesis have been identified [5,6]. Peroxisome proliferator-activated receptor (PPAR) c plays a central role in this regulation [7,8]. Ligand-activated PPARc regulates many genes involved in glucose and lipid homeostasis and is involved in the maintenance of insulin responsiveness [8,9,10].
Prostaglandins (PGs) and their metabolites are involved in the regulation of adipogenesis. PGD 2 [11] and its metabolite, D 12 -PGJ 2 [12], activate the middle-late phase of adipogenesis, and PGD 2 -overproducing mice become obese under the high-fat diet [13]. Moreover, prostacyclin (PGI 2 ) enhances adipogenesis through PGI 2 receptor [14,15]. In contrast, PGF 2a is produced by aldo-keto reductase (AKR) 1B3 in adipocytes; and it suppresses the early phase of adipogenesis through PTGFR receptors [16,17]. PGF 2a promotes the production of anti-adipogenic PGF 2a and PGE 2 by enhancing the expression of cyclooxygenase-2 (PTGS2; COX-2) through PTGFR (FP) receptor-activated mitogen-activated protein kinase/extracellular signal-regulated kinase kinase/ extracellular signal-regulated kinase cascade and the binding of the cyclic AMP response element (CRE)-binding protein (CREB) to the CRE of the Ptgs2 promoter [18]. Moreover, PGE 2 is known to suppress adipogenesis by acting through the PTGER4 (EP4) receptor [19], and to increase the de novo synthesis of antiadipogenic PGF 2a and PGE 2 in mouse embryonic fibroblasts [20]. These anti-adipogenic PGs repress the function of PPARc via their specific PG receptors.
Several PGE 2 synthases (PTGESs) have been identified in various tissues [21,22]. Microsomal PGES-1 (mPGES-1; PTGES1) is a member of the membrane-associated proteins in eicosanoid and glutathione metabolism (MAPEG) protein family [23], and produces PGE 2 in response to various stimuli [24]. Microsomal PGES-2 (mPGES-2; PTGES2) has also been identified and its expression is high in the heart and brain [25]. Cytosolic PGES (cPGES; PTGES3) is constitutively and ubiquitously expressed in various cells [26]. However, the PGE 2 -producing enzyme in adipocytes has never been identified; and the mechanism causing suppression of the early-phase of adipogenesis by anti-adipogenic PGs such as PGE 2 and PGF 2a remains unclear.
In this study, we demonstrate that PTGES1 was expressed in preadipocytes and that its mRNA and protein levels were consistently detected during adipogenesis. PGE 2 production was detected in preadipocytes and increased during adipogenesis with a peak at 3 h after the initiation of adipogenesis, and PTGES1 was responsible for the production of PGE 2 in adipocytes. PGE 2 elevated the production of anti-adipogenic PGF 2a and PGE 2 by enhancing the expression of PTGS2 by acting through the PTGER4 receptor, which action enhanced the binding of CREB to the Ptgs2 promoter via activation of the PTGER4 receptor/ CREB cascade in 3T3-L1 cells. Thus, PTGES1-produced PGE 2 and AKR1B3-synthesized PGF 2a synergistically suppressed the early phase of adipogenesis through elevation of PTGS2 expression in 3T3-L1 cells.

Cell Culture
Mouse 3T3-L1 cells (Health Science Research Resources Bank, Osaka, Japan) were maintained in Dulbecco's Modified Eagles Medium (DMEM; Sigma, St. Louis, MO, USA) supplemented with 10% (v/v) fetal calf serum and antibiotics. The cells were maintained in a humidified atmosphere of 5% CO 2 at 37uC.
Oil Red O staining was carried out as described previously [11]. Spectrophotometric measurement for Oil Red O staining was performed by dissolving the stained lipid droplets in the cells with isopropyl alcohol, and then the absorbance was measured at 520 nm.

RNA Preparation and Quantification of RNA
Total RNA was extracted with Sepasol-RNAI (Nacalai Tesque, Kyoto, Japan), followed by further purification with an RNeasy Purification System (Qiagen, Hilden, Germany) [17]. The firststrand cDNAs were synthesized from 1 mg of total RNA with random hexamer and ReverTra Ace Reverse Transcriptase (Toyobo, Osaka, Japan) at 42uC for 60 min after initial denaturation at 72uC for 3 min, followed by heat-denaturation of the enzyme at 99uC for 5 min. The cDNAs were further utilized as the templates for quantitative PCR analyses.
Expression levels were quantified by using a LightCycler system (Roche Diagnostics, Mannheim, Germany) with THUNDER-BIRD qPCR Mix (Toyobo) and primer sets ( Table 1). The expression level of the target genes was normalized to that of the TATA-binding protein (TBP).
For determination of the knockdown-efficiency of each siRNA, 3T3-L1 cells were transfected with each siRNA and cultured for 2 days.
For identification of the functions of PTGESs, 3T3-L1 cells were transfected with each siRNA, and caused to differentiate into adipocytes for 6 days. Transfection with siRNA was carried out every 2 days.

Measurement of PGs
PGE 2 and PGF 2a levels were measured by use of their respective enzyme immunoassay kit (EIA; Cayman Chemical) as described previously [17]. In brief, cells were treated with A23187 (5 mM; Calbiochem, San Diego, CA, USA), a calcium ionophore, for 10 min at 37uC. Medium was collected, and centrifuged at 3,0006g for 5 min to remove the cells. The resultant supernatant was then used for measurement of PGE 2 and PGF 2a by performing their respective EIA according to the manufacturer's instructions.

Measurement of Triglyceride Level
Cells were washed with PBS, and lysed with PBS containing 5%(v/v) Triton-X100, and then incubated at 90uC for 3 min. The supernatant was prepared by centrifugation to remove cell debris, and subsequently used for measurement of the intracellular triglyceride level by using a WAKO LabAssay Triglyceride Kit (Wako Pure Chemical, Osaka Japan) according to the manufac-turer's instructions. Absorbance was measured at 570 nm. Protein concentrations were measured as described above.

Luciferase Reporter Assay
The luciferase reporter vectors carrying the mouse Ptgs2 promoter were generated previously [18]. 3T3-L1 cells were cotransfected with each construct (0.9 mg) and pRL-SV40 (0.1 mg, Promega, Madison, WI, USA) in 6-well plates, the latter plasmid carrying the Renilla luciferase gene under the control of the SV40 promoter as the transfection control, along with FuGENE6 Transfection Reagent (Roche Diagnostics) according to the manufacturer's instructions. The cells were cultured for a further 48 h in the presence or absence of AE1-329 (1 mM) or L-161982 (10 mM), and the luciferase activities were measured by using a Dual-Glo Luciferase Assay System (Promega). The reporter activity was calculated relative to that of pGL4.10[luc2] vector (Promega), which was defined as ''1''. Data were obtained from three independent experiments, and each experiment was performed in triplicate. The relative promoter activities were presented as the mean 6 S.D.

Chromatin Immunoprecipitation (ChIP) Assay
The ChIP assay was performed as described previously [11] by using anti-CREB polyclonal antibody (H-74; Santa Cruz Biotech.). Immunoprecipitated DNA-protein complexes were reversecrosslinked, and the free DNAs were purified by ethanolprecipitation and utilized for subsequent PCR amplification with KOD FX DNA Polymerase (Toyobo) with a primer set specific for CRE at position -59 in the Ptgs2 promoter: 59-CAGAGAGGGG-GAAAAGTTGG-39 and 59-GAGCAGAGTCCTGACT-GACTC-39. PCR was conducted under the following conditions: initial denaturation at 94uC for 2 min, followed by 30 cycles of 98uC for 10 sec, 55uC for 20 sec, and 60uC for 20 sec. The amplified PCR products (expected size of 168-bp) were analyzed by performing agarose gel electrophoresis, followed by staining of the gels with ethidium bromide.
In addition, we performed the quantitative PCR analysis to measure the anti-CREB antibody-precipitated DNA level by using the same primers used in PCR analysis as described above. Briefly, the precipitated DNA level was estimated by the use of serially diluted concentration-known DNA including the Ptgs2 promoter region as the standard.

Statistical Analysis
Comparison of 2 groups was analyzed by Student's t-test. For comparison of more than 2 groups with comparable variances, one-way ANOVA and Tukey's post-hoc test was carried out. p,0.05 was considered significant.

Identification of PGES in Adipocytes
At first, we investigated the suppression of adipogenesis by treatment with PGE 2 . Mouse 3T3-L1 cells were caused to differentiate into adipocytes for 6 days in the presence of various concentrations of PGE 2 . Oil Red O staining showed that intracellular lipid-droplets increased in size and number during adipogenesis, and these enhancements were repressed in a PGE 2concentration-dependent manner ( Fig. 1A and 1B). Moreover, the intracellular triglyceride level was also enhanced during adipogenesis, and when the cells were caused to differentiate into adipocytes in the presence of various concentrations of PGE 2 , its level was significantly repressed in a concentration-dependent manner (Fig. 1C). When the cells were cultured in medium containing various concentrations of PGE 2 for 3 h, followed culture for 6 days in the absence of PGE 2 , suppression effect at 3 h was slightly weaker than those for 6 days (data not shown). Furthermore, the expression level of adipogenic genes such as total Pparc (Pparc1 and Pparc2), CCAAT-enhancer binding protein (C/ ebp)a, Fabp4, and Scd1 in adipocytes was elevated approximately 2.8-, 3.6-, 59.2-, and 8.9-fold, respectively, as compared with each of those of the undifferentiated cells (Fig. 1D). Furthermore, the enhanced expression of these genes was suppressed by about 57, 74, 96, and 88%, respectively, of that of the vehicle-treated differentiated cells (Fig. 1D). These results indicate that PGE 2 suppressed adipogenesis measured in terms of the expression of adipogenic genes in 3T3-L1 cells.
Next, we examined the expression of the Ptges genes in 3T3-L1 cells. Cells were caused to differentiate into adipocytes for 6 days, and the gene expression of the three major PTGESs, i.e., PTGES1, PTGES2, and PTGES3, during adipogenesis was measured by performing quantitative PCR and Western blot analyses. All three PTGESs were expressed in preadipocytes and consistently so during adipogenesis ( Fig. 2A). The protein levels of all three PTGESs, examined by Western blot analysis, well resembled those of their mRNA expression ( Fig. 2A). However, the expression profiles of the Ptgs2 gene and protein were quite different from them; as it was transiently up-regulated at 3 h after the initiation of adipogenesis, and then decreased to the basal level ( Fig. 2A). Then, we measured the PGE 2 level during adipogenesis by EIA. PGE 2 was produced in preadipocytes, and its production level rapidly increased to a peak at 3 h after initiation of adipogenesis and then quickly decreased to a level lower than that of the undifferentiated cells (Fig. 2B). This production pattern well resembled the expression profile in the PTGS2 ( Fig. 2A).   These results reveal that all three PTGESs were consistently expressed during adipogenesis. PGE 2 production was detected in preadipocytes and transiently enhanced at 3 h after the initiation of adipogenesis, whose pattern well resembled the expression of PTGS2 in 3T3-L1 cells.

PTGES1 is Responsible for the Production of PGE 2 in Adipocytes
To identify the active PGES in adipocytes, we transfected 3T3-L1 cells separately with each of the PTGES siRNAs, and differentiated into adipocytes for 2 days. The mRNA levels of all three Ptges genes were significantly decreased more than 50% by their respective siRNAs, as compared with each of their levels when treated with N.C. siRNA (Fig. 3A). Almost the same results were obtained at the protein level by Western blot analysis; and the actin level, as the internal control, was almost the same in all samples (Fig. 3A). Each siRNA was specific for its PTGES, as it did not inhibit the expression of the other Ptges mRNAs (data not shown). Moreover, the siRNA for PTGES1 decreased the PGE 2 production to about 61.4% of that with N.C. siRNA in 3T3-L1 cells (Fig. 3B). In contrast, siRNAs for PTGES2 and PTGES3 did not have any effect on the production of PGE 2 (Fig. 3B); although the mRNA and protein levels of PTGES2 and PTGES3 were significantly decreased by their respective siRNAs (Fig. 3A).
To confirm that PTGES1 is the PGES suppressing adipocyte differentiation, we examined the role of PGES in the accumulation of intracellular lipids. As shown above, PGE 2 inhibited the accumulation of intracellular lipids of 3T3-L1 cells (Fig. 1A and  1B). When the cells were transfected with any one of the three PTGES siRNAs, the intracellular lipid level demonstrated by Oil Red O staining was increased only in the PTGES1 siRNAtransfected cells (Fig. 3C and 3D). Whereas, there were no changes in PTGES2 or PTGES3 siRNA-transfected cells, which showed almost the same lipid accumulation as the vehicle-treated cells ( Fig. 3C and 3D). In addition, the intracellular triglyceride level in PTGES1 siRNA-transfected cells was clearly increased as compared with that in the vehicle-treated or PTGES2 or PTGES3 siRNA-transfected cells (Fig. 3E), indicating that PTGES1 was involved in the accumulation of the intracellular triglyceride level in adipocytes.
Next, we investigated the expression of adipogenic genes in PTGES siRNA-transfected cells. The transcription level of the adipogenic genes such as Pparc, Fabp4, and Scd1 was enhanced approximately 1.7-, 1.4-, and 1.6-fold, respectively, by transfection with PTGES1 siRNA, as compared with those levels for cells treated with vehicle or transfected with N.C. siRNA, PTGES2 siRNA or PTGES3 siRNA (Fig. 4). Protein levels of PPARc, FABP4, and SCD1 were also up-regulated by transfection of 3T3-L1 cells with PTGES1 siRNA, but not affected in N.C., PTGES2 or PTGES3 siRNA-transfected cells (Fig. 4). These results reveal that PTGES1 acted as the PGES in adipocytes and that PTGES1produced PGE 2 suppresses adipogenesis by reducing the expression of adipogenic genes in 3T3-L1 cells.

Involvement of PTGER4 Receptors in the Suppression of Adipogenesis
PGE 2 exerts its action through interaction with four PGE 2 receptor subtypes; PTGER1 (EP1), PTGER2 (EP2), PTGER3 (EP3), and PTGER4 (EP4) [27]. So next, we investigated the expression of the Ptger genes during adipocyte differentiation of 3T3-L1 cells. Ptger1 mRNA was detected in preadipocytes and its expression level gradually increased after 1 day of the initiation of the adipocyte differentiation (Fig. 5). The Ptger4 gene was expressed highly in preadipocytes, and its level decreased almost 50% after the initiation of adipogenesis (Fig. 5). Whereas, the expression of Ptger2 and Ptger3 receptor genes was under the detection limit of our experimental conditions (data not shown).
Next, we examined which PTGER receptor, PTGER1 or PTGER4 was involved in the PGE 2 -mediated suppression of adipogenesis. When 3T3-L1 cells were caused to differentiate into adipocytes for 6 days in medium containing a PTGER1 receptor agonist, DI-004, the accumulation of lipid-droplet in the cells was not changed as judged by Oil Red O staining (Fig. 6A and 6B). Whereas, the amount of lipid-droplets was clearly decreased by treatment with a PTGER4 receptor agonist, AE1-329, which decrease was almost the same as that seen in PGE 2 -treated cells ( Fig. 6A and 6B). Moreover, PGE 2 -mediated suppression of lipid accumulation was cleared by co-treatment with an EP4 receptor antagonist, L-161982, but not with AH6809, a PTGER1 receptor antagonist ( Fig. 6A and 6B). Next, we measured the intracellular triglyceride level when the cells were caused to differentiate into adipocytes in medium containing PGE 2 , DI-004 or AE1-329 or PGE 2 with or without AH6809 or L-161982. Differentiationmediated enhancement of the intracellular triglyceride level was suppressed by treatment with PGE 2 or AE1-329, but not with DI-004 (Fig. 6C). Moreover, PGE 2 -mediated decrease in the intracellular triglyceride level was cleared by co-treatment with L-161982, but not AH6809 (Fig. 6C). When 3T3-L1 cells were cultured for 3 h by chemicals, followed by further cultured for 6 days without chemicals, almost the same results as those for 6 days were observed (data not shown).
Furthermore, when PTGES1 siRNA-transfected cells were caused to differentiate into adipocytes for 6 days with or without DI-004 or AE1-329, Oil Red O staining of the intracellular lipids was carried out and the intracellular triglyceride level was measured. Only AE1-329 could repress the PTGES1 siRNAmediated enhancement of adipogenesis ( Fig. 6D-F). Almost the same results were obtained when the PTGES1 siRNA-transfected cells were treated with PTGER1 or PTGER4 agonist for 3 h, followed by cultured for 6 days in the absence of PTGER1 or PTGER4 agonist (data not shown). These results, taken together, indicate that PTGES1-produced PGE 2 suppressed adipogenesis by acting through PTGER4 receptors in 3T3-L1 cells.

Activation of PTGER4 Receptor Enhances PGE 2 Production with Elevation of Ptgs2 Expression in 3T3-L1 Cells
When 3T3-L1 cells were cultured with AE1-329, we also found that the production of PGE 2 was increased approximately 2.8-fold, as compared with that obtained for the vehicle-treated cells (Fig. 7A). The expression of the Ptgs2 gene was enhanced approximately 4.3-fold at 3 h after the initiation of adipogenesis, as compared with that in the preadipocytes (Fig. 7B). Moreover, when the cells were caused to differentiate into adipocytes in the presence of AE1-329, the expression of the Ptgs2 gene was elevated about 1.3-fold at 3 h, as compared with that in vehicle-treated cells (Fig. 7B). Then the expression level quickly decreased to a level lower than that detected in preadipocytes (Fig. 7B). Furthermore, the AE1-329-mediated enhancement of the expression of the Ptgs2 gene was repressed by co-treatment with L-161982 (Fig. 7C, left  panel). In addition, PGE 2 itself was able to elevate the expression of the Ptgs2 gene in 3T3-L1 cells, and this enhancement was suppressed by co-treatment with L-161892 (Fig. 7C, right panel). These results indicate that PGE 2 enhanced its own production by acting through the PTGER4 receptor to elevate the expression of the Ptgs2 gene in an autocrine manner in 3T3-L1 cells.

Involvement of CREB in the PGE 2 -mediated Activation of Ptgs2 Gene Expression
CREB has been identified as the activator for the transcription of the Ptgs2 gene in 3T3-L1 cells [18]. So, we investigated whether the CREB was involved in the PGE 2 /PTGER4 receptor-elevated Ptgs2 gene expression by performing a luciferase reporter assay. The transcription initiation site of the mouse Ptgs2 gene has been determined [28]. When the construct carrying the promoter region from 2300 to +124, named 2300/+124, was used for the transfection, efficient reporter activity was detected (Fig. 8A). Moreover, when the 2300/+124 construct-transfected cells were  treated with AE1-329, the luciferase reporter activity was enhanced to become approximately 151% (black column) of that of the vehicle (white column); and this AE1-329-activated Ptgs2 promoter activity was suppressed by the co-treatment with L-161982 (gray column) to become about 78% of the promoter activity of the AE1-329-treated cells (Fig. 8A). Furthermore, when the region from 2300 to 250 was deleted, the luciferase reporter activity was significantly decreased, and the responses to AE1-329 and L-161982 disappeared (Fig. 8A). To confirm the importance of the CRE at position 259 in the PGE 2 -derived elevation of Ptgs2 gene expression in 3T3-L1 cells, we introduced a mutation at this position in the 2300/+124 construct; 2300/+124(mu) [18]. When the cells were transfected with this 2300/+124(mu) construct, the responsiveness to AE1-329 and L-161982 was lost; triglyceride level. 3T3-L1 cells were cultured as described in the legend of Fig. 7A. Data are presented as the mean 6 S.D. from 3 independent experiments. *p,0.01, as indicated by the brackets. D. Accumulation of lipid-droplet by PTGES1-produced PGE 2 acting via the PTGER4 receptor. 3T3-L1 cells (undifferentiated cells: U) were transfected with PTGES1 siRNA (m-1 siRNA) and caused to differentiate into adipocytes (D) for 6 days in medium containing either DI-004 (1 mM) or AE1-329 (1 mM). Bar = 50 mm. E. Measurement of Oil Red O dye extracted from lipid droplet-laden cells. F. Effects of PTGES1 siRNA and PTGER agonists on intracellular triglyceride level. 3T3-L1 cells were cultured as described in the legend of Fig. 6D. Data are presented as the mean 6 S.D. from 3 independent experiments. *p,0.01, as indicated by the brackets. doi:10.1371/journal.pone.0044698.g006 although the basal promoter activity was not altered (Fig. 8A). These results indicate that PGE 2 activated Ptgs2 gene expression through the CRE at position 259 of the mouse Ptgs2 promoter in 3T3-L1 cells.
Next, we examined the binding of CREB to the CRE at 259 of the Ptgs2 promoter by performing a chromatin immunoprecipitation (ChIP) assay. The expected size (168-bp; Fig. 8B, left panel) of an amplicon containing the CRE at 259 was detected in the formaldehyde-fixed DNA-protein complexes immunoprecipitated with anti-CREB antibody (Fig. 8B, right panel). Moreover, when the cells were treated with AE1-329, the binding efficiency was enhanced about 4.7-fold as compared with that of the untreated cells (Fig. 8B, right panel), and the AE1-329-derived increase in the efficiency of binding of CREB to the CRE was clearly suppressed by co-treatment with L-161982 (Fig. 8B, right panel). On the contrary, there was no detectable signal when rabbit normal IgG was added (Fig. 8B, right panel). These results indicate that PGE 2mediated upregulation of Ptgs2 gene expression occurred by enhancing the binding of CREB to the CRE of the Ptgs2 gene promoter in 3T3-L1 cells.  Synergistic PGE 2 and PGF 2a -mediated Suppression of Adipogenesis PGF 2a and PGE 2 suppress the progression of adipogenesis through their specific PG receptors, i.e., PTGFR and PTGER4, respectively [18,20]. Moreover, PGF 2a induces the production of anti-adipogenic PGE 2 and PGF 2a by triggering the PTGFR receptor/MEK/ERK cascade in 3T3-L1 cells [18]. PGE 2 also enhances the production of anti-adipogenic PGE 2 and PGF 2a in mouse embryonic fibroblasts [20].
We examined whether PTGER4 receptor-mediated activation would enhance PGF 2a production in 3T3-L1 cells. PGE 2 production was increased by the treatment with AE1-329, and this enhancement was lost by co-treatment with L-161982, (Fig. 9A). Furthermore, PGF 2a production was also enhanced by treatment with AE1-329, and the co-treatment with L-161982 blocked this increase (Fig. 9A). These results reveal that AE1-329derived activation of the PTGER4 receptor enhanced de novo synthesis of anti-adipogenic PGE 2 and PGF 2a in 3T3-L1 cells.
As both PGE 2 and PGF 2a act as anti-adipogenic PGs in adipocytes, we investigated their suppression effects on adipogenesis in 3T3-L1 cells. When the cells were caused to differentiate into adipocytes for 6 days in medium containing either PGF 2a or PGE 2 along with NS-398, which is an inhibitor of PTGS2 and thus suppresses the de novo synthesis of PGs including the antiadipogenic PGE 2 or PGF 2a , the accumulation of the intracellular lipids shown by Oil Red O staining was decreased as compared with that in the vehicle-treated cells (Fig. 9B and 9C). In addition, greater suppression was observed when the cells were cultured in medium containing both PGF 2a and PGE 2 ( Fig. 9B and 9C). The intracellular triglyceride level was elevated during adipogenesis, and the enhanced triglyceride level was suppressed by cotreatment with either of PGE 2 and PGF 2a . Moreover, when the cells were caused to differentiate into adipocytes in medium containing both PGE 2 and PGF 2a , the intracellular triglyceride level was lower than that in PGE 2 or PGF 2a -treated cells (Fig. 9D).
Next, we measured the expression level of adipogenic genes in PGE 2 -and/or PGF 2a -treated cells. When 3T3-L1 cells were caused to differentiate into adipocytes, the expression levels of Pparc, Fabp4, and Scd1 genes were enhanced approximately 3.1-, 7.1-, and 3.3-fold, respectively, as compared with those in the undifferentiated cells (Fig. 9E). In addition, the expression levels of these genes were enhanced even more in adipocytes cultured in medium containing NS-398. When the cells were caused to differentiate into adipocytes in medium containing both PGE 2 and PGF 2a , the expression levels of the genes were decreased to a greater extent than when the cells were cultured in medium containing PGE 2 or PGF 2a (Fig. 9E). Furthermore, the suppression effect on adipogenesis by PGE 2 was weaker than that by PGF 2a (Fig. 9E). These results indicate that PGE 2 or PGF 2a synergistically suppressed adipogenesis in 3T3-L1 cells.

Discussion
PGs are known to be involved in the regulation of adipogenesis. PGD 2 is synthesized by lipocalin-type PGD synthase in adipocytes and accelerates the mid-late phase of adipogenesis [11]. PGI 2 is involved in the activation of preadipocytes to adipocytes through PGI 2 receptor [14,15]. In contrast, PGF 2a and PGE 2 suppress the progression of adipogenesis [17,18,19,20]. PGF 2a is synthesized by AKR1B3 in adipocytes and represses the early phase of adipogenesis by engaging the PTGFR receptor [17]. Moreover, PGF 2a enhances the production of itself and PGE 2 by enhancing the expression of the COX-2 gene via activation of the PTGFR receptor-ERK/CREB cascade [18]. PGE 2 also acts as antiadipogenic factor, by acting through the PTGER4 receptor [19]. A recent study demonstrated that PGE 2 -PTGER4 signaling suppresses adipocyte differentiation by negatively affecting Pparc expression in an autocrine manner in adipocytes [20]. However, the PGE 2 -producing enzyme in adipocytes and the precise mechanism regulating the suppression of adipogenesis by PGE 2 have not been fully understood. Here, we found that PTGES1 synthesized PGE 2 in 3T3-L1 cells, which PG then suppressed the early phase of adipogenesis via the PTGER4 receptor. Moreover, PGE 2 enhanced Ptgs2 gene expression through the positive feedback loop via PTGER4 receptor, and the elevated PGE 2 and PGF 2a production. Furthermore, AKR1B3-produced PGF 2a suppresses the early phase of adipogenesis through PTGFR receptor [17], and increased the expression of the Ptgs2 gene [18], like PGE 2 . Thus, PTGES1-produced PGE 2 and AKR1B3synthesized PGF 2a synergistically suppress the progression of the early phase of adipogenesis (Fig. 10).
Until now, three major enzymes that catalyze the production of PGE 2 from PGH 2 have been identified [21,29]: PTGES1 [23], PTGES2 [25], and PTGES3 [26]. PTGES1 has been identified as the member of the MAPEG family [30]. These three PTGESs were constitutively expressed during adipocyte differentiation of 3T3-L1 cells (Fig. 2). When the expression of PTGES1 was suppressed by its siRNA, the PGE 2 level was significantly decreased (Fig. 3A and 3B), indicating that PTGES1 was the PGES active in adipocytes. There are two different papers in the literature concerning the expression of PTGES1 in adipocytes. Hetu et al. reported that PTGES1 levels in obese fats are significantly lower than those in lean animals [31]. However, the other report by Xie et al. indicated that PTGES1 is enhanced during differentiation of 3T3-L1 cells [32], thus differing from our results. At the present time, there is no clear explanation for this discrepancy. Further precise studies of the in vitro and in vivo functions of PTGES1 in adipocytes are needed to solve this problem. In addition, we have to also elucidate the effects of GST activity of PTGES1 in the regulation of adipogenesis, because PTGES1 also carries GST activity [33].
PG synthesis is coordinately regulated through the coupling of terminal PG synthases with each or both of PTGS1 (COX-1) and PTGS2 [34]. Both PTGSs were expressed in the undifferentiated 3T3-L1 cells (data not shown), indicating that both PTGSs would probably have the ability to couple with PTGES1 for the production of PGE 2 . PTGES1 is co-localized with both PTGS isozymes in the perinuclear envelope [35]. However, PTGES1 is functionally coupled with PTGS2 to produce PGE 2 [35]. In fact, the PGE 2 production profile well resembled the expression profile of the Ptgs2 gene (Fig. 2). PGs are known to be associated with Ptgs2 gene expression in an autocrine manner in a variety of cells including adipocytes [18,20,36,37]. PGF 2a increases the expression of the Ptgs2 gene via the PTGFR/ERK/CREB cascade, which increase is followed by elevation of PGF 2a and PGE 2 production in 3T3-L1 cells [18]. PGE 2 also enhances the production of PGE 2 and PGF 2a through the PTGER4 receptor in mouse embryonic fibroblasts [20] and suppresses the progression of adipogenesis [19]. Anti-adipogenic PGF 2a and PGE 2 increased themselves to enhance the suppression of adipogenesis in the early phase of adipogenesis. However, the suppression of adipogenesis by these anti-adipogenic PGs was terminated within several hours after the initiation of adipogenesis. Therefore, the molecular mechanism underlying this termination needs be further elucidated.
In summary, we identified PTGES1 (mPGES-1) as the PGES active in adipocytes, whose expression was detected in both preadipocytes and adipocytes during adipogenesis. PTGES1synthesized PGE 2 suppressed the early phase of adipocyte differentiation via the PTGER4 (EP4) receptors by downregulating adipogenic gene expression. Furthermore, PGE 2 enhanced the expression of the Ptgs2 (COX-2), and induced the production of itself and PGF 2a . Both anti-adipogenic PGs synergistically suppressed the progression of adipogenesis in 3T3-L1 cells. Further studies will be needed to elucidate the in vivo functions of PGE 2 and PGF 2a in the suppression of obesity.