Arachidonic Acid Added during the Differentiation Phase of 3T3-L1 Cells Exerts Anti-Adipogenic Effect by Reducing the Effects of Pro-Adipogenic Prostaglandins

A linoleic acid (LA) metabolite arachidonic acid (AA) added to 3T3-L1 cells is reported to suppress adipogenesis. The purpose of the present study aimed to clarify the effects of AA added during the differentiation phase, including adipogenesis, the types of prostaglandins (PG)s produced, and the crosstalk between AA and the PGs produced. Adipogenesis was inhibited by AA added, while LA did not. When AA was added, increased PGE2 and PGF2α production, unchanged Δ12-PGJ2 production, and reduced PGI2 production were observed. Since the decreased PGI2 production was reflected in decreased CCAAT/enhancer-binding protein-β (C/EBPβ) and C/EBPδ expression, we expected that the coexistence of PGI2 with AA would suppress the anti-adipogenic effects of AA. However, the coexistence of PGI2 with AA did not attenuate the anti-adipogenic effects of AA. In addition, the results were similar when Δ12-PGJ2 coexisted with AA. Taken together, these results indicated that the metabolism of ingested LA to AA is necessary to inhibit adipogenesis and that exposure of AA to adipocytes during only the differentiation phase is sufficient. As further mechanisms for suppressing adipogenesis, AA was found not only to increase PGE2 and PGF2α and decrease PGI2 production but also to abrogate the pro-adipogenic effects of PGI2 and Δ12-PGJ2.


Introduction
White adipose tissue is a major metabolic organ responsible for energy homeostasis [1]. The status of adipocytes significantly affects the development of obesity and diabetes [2], two typical noncommunicable diseases that have recently become major problems in developed countries. Preadipocytes in the immediate vicinity of adipocytes in white adipose tissue can alter adipose tissue function by differentiating into adipocytes. Established immortal preadipocyte cell lines such as 3T3-L1 cells have facilitated studies of adipogenesis [3,4]. In general, adipogenesis is initiated in vitro by incubating confluent 3T3-L1 preadipocytes in a medium containing a mixture of 3-isobutyl-1-methylxanthine (IBMX), dexamethasone, and insulin (MDI). MDI triggers the post-confluence mitotic clonal expansion of cells at G0/G1 growth arrest, forcing the cells to exit the cell cycle into terminal differentiation. IBMX and dexamethasone rapidly induce the expression of CCAAT/enhancer-binding protein-β (C/EBPβ) and C/EBPδ, respectively, during the from Promega (Madison, WI, USA). Oligonucleotides for real-time quantitative (RT-q) PCR amplification were provided by Sigma Genosys Japan (Ishikari, Japan).

Culture of 3T3-L1 Cells and Induction of Adipogenesis
Mouse 3T3-L1 pre-adipogenic cells (JCRB9014; JCRB Cell Bank, Osaka, Japan) in the growth phase were seeded at a density of 1 × 10 5 or 2 × 10 5 in 35 or 60 mm dishes containing 2 or 4 mL, respectively, of growth medium (GM; DMEM-HEPES supplemented with 10% FBS, penicillin G [100 units/mL], streptomycin sulfate [100 µg/L], and ascorbic acid [200 µM]), then incubated at 37 • C under 7% CO 2 until they reached confluence. Confluent monolayers were incubated with differentiation medium (DM; GM supplemented with dexamethasone [1 µM], IBMX [0.5 mM], and insulin [10 µg/mL]) for 6-48 h to induce differentiation into adipocytes. The cells were then incubated for 6-10 days in maturation medium (MM; GM supplemented with insulin [5 µg/mL]). The medium was replaced with fresh MM every 2 days to promote the accumulation of fat in 3T3-L1 cells during maturation. We examined the effects of various agents added during differentiation on adipogenesis by incubating confluent cell monolayers in DM supplemented with test compounds for 6-48 h, followed by the standard maturation protocol. The test compounds were dissolved in ethanol and added to the DM to a final ethanol concentration of 0.2%.

Quantification of Intracellular TAGs and Cellular Proteins
Cultured mature adipocytes were harvested, suspended in phosphate-buffered saline (PBS) without Ca 2+ and Mg 2+ (PBS [-]) supplemented with 0.05% trypsin and 0.53 mM EDTA, and incubated at 37 • C for 5 min. The resulting cell suspensions were washed with PBS (-), divided into two portions, and homogenized in 25 mM Tris-HCl buffer (pH 7.4) containing 1 mM EDTA and 1 N NaOH. Amounts of intracellular TAGs were quantified in one portion using Triglyceride E-Test Kits (Wako). Cellular proteins were precipitated in the other portion with ice-chilled 6% trichloroacetic acid to remove interfering substances and then quantified using the Lowry method with fatty acid-free bovine serum albumin as a standard. Amounts of intracellular TAGs were normalized to the protein content and are expressed as relative amounts of accumulated intracellular TAGs.

Quantification of PGs by ELISA
Levels of PGE 2 , PGF 2α , ∆ 12 -PGJ 2 , and 6-keto-PGF 1α were measured in the DM containing IBMX in the presence 50 µM AA for 48 h. We quantified PGE 2 by ELISA specific for PGE 2 as described [32]. A PGE 2 -conjugate and fatty acid-free bovine serum albumin were immobilized in 96-well microplates. Immobilized antigen in standards or test samples was competitively incubated with diluted mouse monoclonal antibody specific for PGE 2 . The resultant immunocomplex was detected spectrophotometrically by monitoring the peroxidase activity using o-phenylenediamine as a substrate after binding to biotin-conjugated rabbit anti-mouse IgG antibody and ExtrAvidin peroxidase conjugate as described [33]. We determined the amounts of ∆ 12 -PGJ 2 using monoclonal antibodies specific for ∆ 12 -PGJ 2 . Polyclonal mouse antisera specific for PGF 2α and 6-keto-PGF 1α that reflect PGI 2 biosynthesis were used to develop solid-phase ELISA for the corresponding immobilized antigens as described [34]. Standard curves were generated in fresh DM containing IBMX to quantify PGs including PGE 2 , PGF 2α , ∆ 12 -PGJ 2 , and 6-keto-PGF 1α , which were biosynthesized during differentiation.

Quantification of Gene Expression
Total RNA (1 µg) extracted from the cells after 6, 24, and 48 h of the differentiation phase, and on day 6 of the maturation phase using acid guanidium thiocyanate/phenol/ chloroform was reverse transcribed (RT) using M-MLV reverse transcriptase (Point mutation without Ribonuclease H activity). Single-stranded cDNA was synthesized using oligo-(dT) 15 and a random 9-mer (Promega) as primers in the RT reaction. Transcript levels were determined by RT-qPCR using TB GreenTM Premix Ex TaqTM II (Tli RNaseH Plus) kits (Takara Bio Co., Inc., Kusatsu, Japan) and a Thermal Cycler Dice TM Real Time System (Takara Bio Co., Inc.) according to the threshold cycle (CT) and ∆∆ CT methods described by the manufacturer. Table 1 shows the oligonucleotides used herein. The cycling program comprised 95 • C for 30 s, 40 cycles at 95 • C for 5 s and 60 • C for 30 s, followed by 95 • C for 15 s and 60 • C for 30 s. Levels of target gene transcripts were determined and normalized to those of β-Actin. The accession numbers of the target genes are as follows: C/Ebpβ, NM_009883; C/Ebpδ, NM_007679; Pparγ, NM_011146; C/Ebpα, NM_001287523; Lpl, NM_008509; Glut4, AB008453; Leptin, NM_008493; β-Actin, NM_007393.

Statistical Analyses
All results are expressed as means ± standard error of the mean (SEM). Data were statistically analyzed by Dunnett tests, Student t-tests, and Tukey-Kramer tests using Excel 2011 (Microsoft Corp., Redmond, WA, USA) and Statcel 4 (OMS Publishing Co., Saitama, Japan). Values with p < 0.05 were considered statistically significant.

Effects of AA or LA Added during Only the Differentiation Phase of 3T3-L1 Cells on MDI-Induced Adipogenesis
The present study examined whether AA added during the differentiation phase affects the anti-adipogenic effects following the procedure in Figure 1A. As an indicator for adipogenic differentiation of 3T3-L1 cells, we evaluated intracellular TAG accumulation. AA added up to 100 µM during the differentiation phase of 3T3-L1 cells dose-dependently reduced the accumulation of intracellular TAG in MDI-induced mature adipocytes ( Figure 1B). Due to the fact that 50 µM AA was sufficient to be effective, we used this concentration in subsequent analyses. We next checked whether LA, like AA, also suppresses MDI-induced adipogenesis following the procedure in Figure 1A because AA in humans is largely produced from LA ingested in the diet. Figure 1C shows that LA, unlike AA, added during the differentiation phase did not inhibit the accumulation of intracellular TAG in MDI-induced mature adipocytes. These results indicated that the metabolic conversion of LA to AA is needed to suppress the accumulation of intracellular TAG during maturation induced by MDI. In addition, if 3T3-L1 cells are exposed to AA during the differentiation phase, its anti-adipogenic effects can be sufficiently exerted even without exposure to AA during the maturation phase.  We evaluated what PGs were produced by adding AA during the differentiation phase. Analysis performed according to the procedure in Figure 2A confirmed the elevated PGE 2 and PGF 2α production, similar to previous reports ( Figure 2B,C) [28,31]. Since the production of other PGs is unknown, we checked whether other PGs are also synthesized. The production of ∆ 12 -PGJ 2 was not altered under these conditions ( Figure 2D), while that of PGI 2 -derived 6-keto-PGF 1α was decreased, suggesting decreased PGI 2 production ( Figure 2E). As PGI 2 is rapidly degraded to stable 6-keto-PGF 1α , the amount of 6-keto-PGF 1α reflects that of PGI 2 [34]. Based on the results of Figure 2, we examined how MDI-induced adipogenesis would be affected if 3T3-L1 cells responded to PGs during only the differentiation phase ( Figure 3A). PGE 2 and PGF 2α added during the differentiation phase inhibited MDI-induced adipogenesis during the maturation phase ( Figure 3B). On the other hand, ∆ 12 -PGJ 2 , and PGI 2 added during the differentiation phase promoted MDI-induced adipogenesis during the maturation phase ( Figure 3B). Together with the results in Figure 2, the suppression of MDI-induced adipogenesis during the maturation phase by AA added during the differentiation phase may be due not only to increased anti-adipogenic PGE 2 and PGF 2α but also to reduced pro-adipogenic PGI 2 production during the differentiation phase.

Adipogenesis-Related Gene Expression in Response to AA during the Differentiation Phase
Although endogenous PGI 2 has a very short chemical lifespan [35], we found that PGI 2 induces adipogenesis by activating IP receptors in an autocrine fashion before spontaneous degradation [30]. Furthermore, PGI 2 added during the differentiation phase significantly promoted the adipogenesis of 3T3-L1 cells ( Figure 3B). Therefore, since the results in Figure 2E suggested that AA decreased PGI 2 production, we predicted that induction of the MDI-induced adipogenic program would be suppressed, starting with activated IP receptor to increase C/Ebpβ and C/Ebpδ expression [20]. The expression of C/Ebpβ and C/Ebpδ was analyzed as described in Figure 4A. The RT-qPCR findings show that AA added during the differentiation phase reduced the expression of C/Ebpβ and C/Ebpδ which are critical for the progression of the early phase of adipogenesis ( Figure 4B,C). Since C/Ebpβ and C/Ebpδ levels affect the expression of Pparγ and C/Ebpα, master regulators of adipogenesis [5][6][7][8][9], we analyzed their expression as described in Figure 5A. The RT-qPCR revealed reduced Pparγ and C/Ebpα expression, which should result from the downregulation of C/Ebpβ and C/Ebpδ ( Figure 5B,C). Due to the fact that AA decreased the expression of these transcription factors involved in adipogenesis, we further analyzed the expression of the established adipocytespecific marker genes, lipoprotein lipase (Lpl), glucose transporter 4 (Glut4), and Leptin by RT-qPCR as described in Figure 6A-D show downregulated expression of Lpl, Glut4, and Leptin. Taken together, these results suggest that AA added during the differentiation phase may attenuate the activation of the IP receptor by PGI 2 and prevent the progression of the early phase of adipogenesis triggered by C/EBPβ and C/EBPδ increased otherwise.

Adipogenesis-Related Gene Expression in Response to AA during the Differentiation Phase
Although endogenous PGI2 has a very short chemical lifespan [35], we found that PGI2 induces adipogenesis by activating IP receptors in an autocrine fashion before spontaneous degradation [30]. Furthermore, PGI2 added during the differentiation phase significantly promoted the adipogenesis of 3T3-L1 cells ( Figure 3B). Therefore, since the results in Figure 2E suggested that AA decreased PGI2 production, we predicted that induction of the MDI-induced adipogenic program would be suppressed, starting with activated IP receptor to increase C/Ebpβ and C/Ebpδ expression [20]. The expression of C/Ebpβ and C/Ebpδ was analyzed as described in Figure 4A. The RT-qPCR findings show that AA added during the differentiation phase reduced the expression of C/Ebpβ and C/Ebpδ which are critical for the progression of the early phase of adipogenesis ( Figure 4B,C). Since C/Ebpβ and C/Ebpδ levels affect the expression of Pparγ and C/Ebpα, master  the downregulation of C/Ebpβ and C/Ebpδ ( Figure 5B,C). Due to the fact that AA decreased the expression of these transcription factors involved in adipogenesis, we further analyzed the expression of the established adipocyte-specific marker genes, lipoprotein lipase (Lpl), glucose transporter 4 (Glut4), and Leptin by RT-qPCR as described in Figure 6A-D show downregulated expression of Lpl, Glut4, and Leptin. Taken together, these results suggest that AA added during the differentiation phase may attenuate the activation of the IP receptor by PGI2 and prevent the progression of the early phase of adipogenesis triggered by C/EBPβ and C/EBPδ increased otherwise.

Effects of AA Coexistent with Pro-Adipogenic PGs during the Differentiation Phase of 3T3-L1 Cells on MDI-Induced Adipogenesis
Since the production of PGI2 during the differentiation phase was reduced by the addition of AA ( Figure 2E), we examined whether the inhibition of MDI-induced adipogenesis during the maturation phase by AA is alleviated by the coexistence of PGI2 with AA during only the differentiation phase ( Figure 7A). Although addition of PGI 2 during the differentiation phase promoted MDI-induced adipogenesis during the maturation phase ( Figure  3B), the addition of PGI 2 with AA during the differentiation phase was not able to rescue AAstimulated inhibition of adipogenesis ( Figure 7B). In addition, although the production of PPARγ activator, Δ 12 -PGJ2, during the differentiation phase was not changed by the addition of AA ( Figure 2D), we also evaluated whether pro-adipogenic PGs other than PGI2 influence the AA-elicited inhibition of MDI-induced adipogenesis during the maturation phase ( Figure 7A). As the case of PGI 2 , pro-adipogenic Δ 12 -PGJ 2 ( Figure 3B) was not able to rescue AA-stimulated inhibition of adipogenesis when Δ 12 -PGJ 2 coexisted with AA during the differentiation phase ( Figure 7B). Taken together, these results indicate that pro-adipogenic PGs production during the differentiation phase does not influence the suppression of MDI-induced adipogenesis during the maturation phase by adding AA during the differentiation phase.

Effects of AA Coexistent with Pro-Adipogenic PGs during the Differentiation Phase of 3T3-L1 Cells on MDI-Induced Adipogenesis
Since the production of PGI 2 during the differentiation phase was reduced by the addition of AA ( Figure 2E), we examined whether the inhibition of MDI-induced adipogenesis during the maturation phase by AA is alleviated by the coexistence of PGI 2 with AA during only the differentiation phase ( Figure 7A). Although addition of PGI 2 during the differentiation phase promoted MDI-induced adipogenesis during the maturation phase ( Figure 3B), the addition of PGI 2 with AA during the differentiation phase was not able to rescue AA-stimulated inhibition of adipogenesis ( Figure 7B). In addition, although the production of PPARγ activator, ∆ 12 -PGJ 2 , during the differentiation phase was not changed by the addition of AA ( Figure 2D), we also evaluated whether pro-adipogenic PGs other than PGI 2 influence the AA-elicited inhibition of MDI-induced adipogenesis during the maturation phase ( Figure 7A). As the case of PGI 2 , pro-adipogenic ∆ 12 -PGJ 2 ( Figure 3B) was not able to rescue AA-stimulated inhibition of adipogenesis when ∆ 12 -PGJ 2 coexisted with AA during the differentiation phase ( Figure 7B). Taken together, these results indicate that pro-adipogenic PGs production during the differentiation phase does not influence the suppression of MDI-induced adipogenesis during the maturation phase by adding AA during the differentiation phase.  Figure 8 summarizes the present findings. In line with previous findings reported by other groups [28,29,31], the present study confirmed that AA added during the differentiation phase of 3T3-L1 cells inhibited MDI-induced adipogenesis during the maturation phase and elicited anti-adipogenic PGE2 and PGF2α production. The novel findings are as follows: 1) LA added during only the differentiation phase did not inhibit MDI-induced adipogenesis during the maturation phase, even though LA is a precursor of AA; 2) Δ 12 -PGJ2 production was not affected by AA added during the differentiation phase; 3) PGI2 production was decreased by AA added during the differentiation phase, which in turn suppressed the progression of the adipogenic program initiated by increased C/Ebpβ and  Figure 7. Effects of PGI 2 or ∆ 12 -PGJ 2 on 3T3-L1 cells incubated with AA during the differentiation phase of 3T3-L1 cells. (A) Experimental procedure. We seeded and incubated 3T3-L1 cells (1 × 10 5 /dish) in 35 mm dishes in 2 mL of GM until they reached 100% confluence. The cells were then incubated for 48 h with 2 mL of DM containing vehicle, AA (50 µM), and PGI 2 .Na (100 nM) or ∆ 12 -PGJ 2 (1 µM). The differentiation medium was replaced with 2 mL of fresh MM every 2 days thereafter. Intracellular TAG levels were analyzed in terminally differentiated mature adipocytes on day 10. (B) Intracellular TAG levels in cultured adipocytes. Data are shown as means ± SEM of n = 3 experiments for (B). * p < 0.05 vs. control (vehicle in DM) (Tukey-Kramer tests). GM, growth medium; DM, differentiation medium; MM, maturation medium; AA, arachidonic acid; PGI 2 .Na, prostaglandin I 2 .Na; ∆ 12 -PGJ 2 , ∆ 12 -prostaglandin J 2 ; TAG, triacylglycerol; SEM, standard error of the mean. Figure 8 summarizes the present findings. In line with previous findings reported by other groups [28,29,31], the present study confirmed that AA added during the differentiation phase of 3T3-L1 cells inhibited MDI-induced adipogenesis during the maturation phase and elicited anti-adipogenic PGE 2 and PGF 2α production. The novel findings are as follows: (1) LA added during only the differentiation phase did not inhibit MDI-induced adipogenesis during the maturation phase, even though LA is a precursor of AA; (2) ∆ 12 -PGJ 2 production was not affected by AA added during the differentiation phase; (3) PGI 2 production was decreased by AA added during the differentiation phase, which in turn suppressed the progression of the adipogenic program initiated by increased C/Ebpβ and C/Ebpδ expression; (4) the addition of PGI 2 or ∆ 12 -PGJ 2 with AA during the differentiation phase did not influence the AA-induced inhibition of MDI-induced adipogenesis during the maturation phase. Taken together, in addition to the alteration of PGs production, the inhibitory effects of AA on the actions of pro-adipogenic PGs during the differentiation phase may play an important role in the suppression of MDI-induced adipogenesis during the maturation phase.

Discussion
Life 2022, 12, x FOR PEER REVIEW 13 phase did not influence the AA-induced inhibition of MDI-induced adipogenesis du the maturation phase. Taken together, in addition to the alteration of PGs production inhibitory effects of AA on the actions of pro-adipogenic PGs during the differentia phase may play an important role in the suppression of MDI-induced adipogenesis ing the maturation phase. Ingested LA must be metabolized to AA to affect white adipocytes in vivo [28]. present and other studies have shown that AA participates in the inhibition of adipo esis [28,29,31]. However, whether LA affects adipocytes in the same way as AA rem unclear. Our findings indicated that LA, unlike AA, did not inhibit adipogenesis indu by MDI, even though it is a precursor of AA. Thus, the present findings indicate th reduction in white adipose tissue mass associated with ingesting LA from a hightein/low-carbohydrate diet is due to the action of AA, and not LA. Furthermore, the sent study indicates that the anti-adipogenic effects of AA are fully exerted when ad cytes are exposed to AA during the differentiation phase.
When AA is added to 3T3-L1 cells during the differentiation phase without IB production of PGE2 and PGF2α was decreased, while that of PGI2 increases in the ph promoting the accumulation of intracellular TAG levels after maturation [30]. In cont the present study found that AA added during the differentiation phase with IBMX pressed the accumulation of intracellular TAG levels after maturation. We also foun creased PGE2 and PGF2α, and decreased PGI2 production in the differentiation phase. only difference between the previous and the present studies is the presence or abs of IBMX added during the differentiation phase. Since IBMX is a cAMP-elevating ag the difference in the types of PGs produced in the previous and current studies shoul produced by cAMP-activated protein kinase A (PKA) or exchange protein dir  Figure 8. Schematic representation of arachidonic acid function during the differentiation phase of 3T3-L1 cells and subsequent effects on adipogenesis induced by 3-isobutyl-1-methylxanthine, dexamethasone, and insulin. Increased PGE 2 and PGF 2α production, whereas the reduced PGI 2 production together with the suppression of PGI 2 and ∆ 12 -PGJ 2 actions by arachidonic acid during the differentiation phase inhibited MDI-induced adipogenesis during the maturation phase. C/EBPα, CCAAT/enhancer-binding protein-α; C/EBPβ, CCAAT/enhancer-binding protein-β; C/EBPδ, CCAAT/enhancer-binding protein-δ; DEX, dexamethasone; Ins, insulin; IBMX, 3-isobutyl-1-methylxanthine; PPARγ, peroxisome proliferator-activated receptor-γ; LPL, lipoprotein lipase; GLUT4, glucose transporter 4; PG, prostaglandin.
Ingested LA must be metabolized to AA to affect white adipocytes in vivo [28]. The present and other studies have shown that AA participates in the inhibition of adipogenesis [28,29,31]. However, whether LA affects adipocytes in the same way as AA remains unclear. Our findings indicated that LA, unlike AA, did not inhibit adipogenesis induced by MDI, even though it is a precursor of AA. Thus, the present findings indicate that a reduction in white adipose tissue mass associated with ingesting LA from a high-protein/low-carbohydrate diet is due to the action of AA, and not LA. Furthermore, the present study indicates that the anti-adipogenic effects of AA are fully exerted when adipocytes are exposed to AA during the differentiation phase.
When AA is added to 3T3-L1 cells during the differentiation phase without IBMX, production of PGE 2 and PGF 2α was decreased, while that of PGI 2 increases in the phase, promoting the accumulation of intracellular TAG levels after maturation [30]. In contrast, the present study found that AA added during the differentiation phase with IBMX suppressed the accumulation of intracellular TAG levels after maturation. We also found Life 2023, 13, 367 13 of 16 increased PGE 2 and PGF 2α , and decreased PGI 2 production in the differentiation phase. The only difference between the previous and the present studies is the presence or absence of IBMX added during the differentiation phase. Since IBMX is a cAMP-elevating agent, the difference in the types of PGs produced in the previous and current studies should be produced by cAMP-activated protein kinase A (PKA) or exchange protein directly activated by cAMP (Epac). Activation of both Epac and PKA during the differentiation phase contributes to MDI-induced adipogenesis [36,37], whereas activation of only PKA suppresses it [38]. Since MDI-induced adipogenesis was suppressed in the present study, only PKA may have been activated by adding IBMX in the presence of AA. In fact, the coexistence of H89, an inhibitor of PKA, with AA during the differentiation phase is reported to suppress the anti-adipogenic effects of AA [28]. Thus, during the differentiation phase, the change in PG production by adding AA depends on the presence of IBMX, suggesting the effect of PKA activation.
Since reduced PGI 2 production leads to decreased C/Ebpβ and C/Ebpδ expression [15], we speculated that not only the rise in both anti-adipogenic PGs, PGE 2 , and PGF 2α , but also the reduction in PGI 2 production, by AA added during the differentiation phase was also responsible for the suppression of MDI-induced adipogenesis during the maturation phase. However, the inhibition of MDI-induced adipogenesis by AA during the maturation phase was not alleviated by the coexistence of PGI 2 with AA during the differentiation phase. Based on these results, we also examined other pro-adipogenic PG, ∆ 12 -PGJ 2 , and found similar results. These results proposed that the property of AA to nullify the effects of adipogenic PGs is the most significant factor leading to anti-adipogenic effects. Therefore, we speculate that the decreased Pparγ expression observed in the present study was primarily caused by the reduced PPARγ transcriptional activity rather than the reduced PGI 2 production induced by AA. The reason supporting the speculation that decreased PPARγ transcriptional activity leads to decreased Pparγ expression is that PPARγ has an auto-loop mechanism with C/EBPα, and their transcriptional activity and expression levels correlate with each other [10]. In fact, this mechanism may be at work as well since a reduction in both Pparγ and C/Ebpα expression was observed in the present study. Reduced PPARγ transcriptional activity is induced by its phosphorylation through MAPK activation, and PGF 2α triggers this mechanism [39]. The observed increase in PGF 2α production induced by AA added during the differentiation phase in the present study may cause the decreased Pparγ expression through this mechanism. Furthermore, PGE 2 also contributes to PGF 2α production by increasing COX-2 expression [40]. Based on these results, the increased PGE 2 production induced by AA added during the differentiation phase may suppress MDI-induced adipogenesis during the maturation phase through the same mechanism as PGF 2α described above. A morphological evaluation of the effect of the addition of AA to the 3T3-L1 cells during the differentiation phase on adipogenesis, such as the size of intracellular lipid droplets as well as the shape and size of adipocytes, could provide us with a better understanding of the molecular mechanism governing the anti-adipogenic function of AA during the early phase of adipocyte differentiation. Since the above action mechanisms of AA are our speculation, further study will be needed in the future.

Conclusions
The present findings indicate that differentiation of cultured 3T3-L1 pre-adipocytes to matured adipocytes is attenuated by the addition of AA, but not LA, during the differentiation phase. The anti-adipogenic effects of AA could be explained, at least partially, by alteration in PG synthesis during the differentiation phase. The addition of AA increased anti-adipogenic PGE 2 and PGF 2α , while decreasing or not affecting pro-adipogenic PGI 2 and ∆ 12 -PGJ 2 . The addition of AA also decreased the expression of the differentiationinitiating genes, C/Ebpβ and C/Ebpδ, which could, by cooperating with the alteration in PG synthesis, decrease the expression of the following master-regulator genes, Pparγ and C/Ebpα, and finally the adipocyte-specific marker genes, Lpl, Glut4, and Leptin. AA should affect the very initiating steps in the differentiation as confirmed by the results that coexisting PGI 2 or ∆ 12 -PGJ 2 did not cancel the AA-induced suppression of adipogenesis. PG synthesis is regulated by PKA: the presence of cAMP-increasing IBMX increases antiadipogenic PGs, while its absence increases the production of pro-adipogenic PGs [30]. Therefore, controlling PKA activity by feeding, for example, the balance of carbohydrates and proteins in the diet, could be a determining factor for the effects of AA on the development of adipose tissue.