Arachidonic acid and prostacyclin signaling promote adipose tissue development: a human health concern

High fat intake is associated with fat mass gain through fatty acid activation of peroxisome proliferator-activated receptors delta and gamma, which promote adipogenesis. We show herein that, compared to a combination of specific agonists to both receptors or to saturated, monounsaturated, and omega-3 polyunsaturated fatty acids, arachidonic acid (C20:4, omega-6) promoted substantially the differentiation of clonal preadipocytes. This effect was blocked by cyclooxygenase inhibitors and mimicked by carbacyclin, suggesting a role for the prostacyclin receptor and activation of the cyclic AMP-dependent pathways that regulate the expression of the CCAAT enhancer binding proteins beta and delta implicated in adipogenesis. During the pregnancy-lactation period, mother mice were fed either a high-fat diet rich in linoleic acid, a precursor of arachidonic acid (LO diet), or the same isocaloric diet enriched in linoleic acid and alpha-linolenic acid (LO/LL diet). Body weight from weaning onwards, fat mass, epididymal fat pad weight, and adipocyte size at 8 weeks of age were higher with LO diet than with LO/LL diet. In contrast, prostacyclin receptor-deficient mice fed either diet were similar in this respect, indicating that the prostacyclin signaling contributes to adipose tissue development. These results raise the issue of the high content of linoleic acid of i) ingested lipids during pregnancy and lactation, and ii) formula milk and infant foods in relation to the epidemic of childhood obesity.

Obesity is associated with metabolic disorders such as dyslipidemia, diabetes, and hypertension, and fat mass ex-cess in severe obesities is typically due to an increase in adipocyte size and number.The formation of adipocytes is a critical event, as mature adipocytes do not divide in vivo and do not undergo significant turnover under physiological conditions.The capacity for proliferation of precursor cells and their differentiation into adipocytes is highest at early age and decrease thereafter in humans and rodents.A limited number of hormones can affect the adipose tissue mass and possibly its distribution (1).High dietary fat intake is now recognized to be associated with a gain of fat mass in animals and humans at all ages (2)(3)(4)(5).However, the lack of evidence of a general increase in energy intake as fat among youths, despite a striking increase in the prevalence of obesity in industrial and developing countries, may be due in part to decreased physical activity and nonexercise activity thermogenesis (6), but also to the composition of food intake in early life.The long-term relationship between the fatty acid composition of dietary fats and the development of adipose tissue in humans is difficult to assess in contrast to animals.When mother rats were fed a high-fat diet rich in linoleic acid (C18:2, -6) or saturated fatty acids, suckling pups at 17 days of age exhibited hyperplasia or hypertrophy of white adipose tissue, respectively (7).Moreover, fish oil rich in eicosapentaenoic acid (C20:5, -3, EPA) and docosahexaenoic acid (C22:6, -3, DHA) prevents obesity in rats (8,9), as well as feeding rats after weaning with dietary fats rich in ␣ -linolenic acid (C18:3, -3), the precursor of EPA and DHA, prevents excessive growth of adipose tissue (10).
The mechanisms underlying the differential adipogenic effect of -6 versus -3 polyunsaturated fatty acids suggest differences between fatty acids and/or fatty acid metabolites in promoting differentiation of adipose precursor cells into adipocytes.In vitro, at the preadipocyte stage, a member of the peroxisome proliferator-activated receptor (PPAR) family, i.e ., PPAR ␦ , and two members of the CCAAT-enhancer binding protein family, i.e., C/EBP ␤ and C/EBP ␦ , act concomitantly to upregulate the subsequent and critical expression of PPAR ␥ leading to adipogenesis (11)(12)(13)(14)(15). Natural long-chain fatty acids act in preadipocytes as adipogenic hormones, participate as transcriptional regulators of the expression of various lipid-related genes, and promote adipogenesis (16).These effects implicate PPARs that bind long-chain fatty acids and fatty acid metabolites (17).Among fatty acids, arachidonic acid (C20:4, -6, ARA), a precursor of prostaglandin I 2 (prostacyclin), synthesized and released from preadipocytes, has been identified as one of the main adipogenic components of serum.Arachidonic acid induces a rapid cAMP production.Both this effect and its long-term adipogenic effect are impaired by cyclooxygenase inhibitors such as aspirin and indomethacin (18).Consistent with an autocrine-paracrine mechanism via released prostacyclin, antibodies directed against this prostanoid and added externally decrease by half the adipogenic effect of arachidonic acid (19).Also consistent with a role of prostacyclin acting as a ligand at the cell surface, it has been shown that i ) prostacyclin and its stable analog carbacyclin mimic the effects of arachidonic acid (20) and also promote adipogenesis of clonal mouse preadipocytes and primary preadipocytes from rat and human (21), and ii ) prostacyclin binding to its cell surface receptor (IP-R) activates in preadipocytes the protein kinase A (PKA) pathway (21) and upregulates the early expression of the C/EBP ␤ and C/EBP ␦ (22).Circumstantial evidence favors the possibility that prostacyclin also binds like carbacyclin to PPAR ␦ (23), and this has been recently supported by studies on stromal cells surrounding the implanting blastocysts (24).
In vivo, invalidation of C/EBP ␤ and C/EBP ␦ genes impairs severely but does not abolish adipose tissue formation (25), whereas invalidation of PPAR ␦ gene leads to a controversial disproportionate decrease in fat mass (26,27).This suggests that prostacyclin signaling, arising from arachidonic metabolism, may play a more important adipogenic role through C/EBP ␤ and C/EBP ␦ than through PPAR ␦ in up-regulating PPAR ␥ expression.In order to estimate the relative importance of the two regulatory pathways, we have taken advantage of the recent availability of specific PPAR agonists and the generation of prostacyclin receptor-null ( ip-r Ϫ / Ϫ ) mice (28).
Our results with wild-type and ip-r Ϫ / Ϫ mice show that polyunsaturated fatty acids of the -6 and -3 series are not equipotent in promoting adipogenesis both in vitro and in vivo, and that arachidonic acid and prostacyclin signaling favor this process.In infants, given the relative enrichment of various foods in linoleic acid as precursor of arachidonic acid, its excessive consumption at a time where adipose tissue is in a dynamic phase of its development may favor childhood obesity.

Mice
The ip-r -null mice were established by gene targeting and backcrossed with C57/BL6J mice for at least 10 generations (28), then ip-r Ϫ / Ϫ males and ip-r Ϫ / Ϫ females were bred to generate further generations.Both ip-r Ϫ / Ϫ and C57/BL6J control mice were maintained on a light/dark cycle with light from 6 AM to 6 PM at 25 Њ C. The female mice designated to be mothers were fed either a standard diet which consisted (by energy) of 7% fat, 66% carbohydrates, and 27% proteins, or a high-fat diet containing 15% corn oil (LO diet) or a mixture of 10% corn oil and 5% perilla oil (LO/LL diet).Both high-fat diets consisted (by energy) of 40% fat, 35% carbohydrates, and 25% proteins, and were supplemented with 0.04% vitamin C and 0.02% vitamin E (UAR, Carbon Blanc, France).Corn oil contained, expressed in g/100 g of total fatty acids, 13% saturated, 27% monounsaturated, 59% -6 polyunsaturated, and 1% -3 polyunsaturated fatty acids.The mixture of corn oil and perilla oil contained 10.9% saturated, 22.9% monounsaturated, 44.3% -6 polyunsaturated, and 21.9% -3 polyunsaturated fatty acids.At 8 weeks of age, female mice fed the same diet since weaning were bred to male mice and maintained throughout mating, pregnancy, and lactation on the same diet.At 18 days of age, male pups were weaned onto the same diets that their mothers had consumed and maintained thereafter.Food intake, body weight, body composition, and cellularity measurements of epididymal fat pad were performed as described previously (29).All experimental animal protocols were performed in accordance with the recommendations of the French Accreditation of Laboratory Animal Care.

Fibroblast culture
Embryos from C57 BL/6J wild-type and ip-r Ϫ / Ϫ mice at 14.5 day postcoitus were used to prepare fibroblasts after removing head, heart, and legs.Mouse embryo fibroblasts were then differentiated into adipocytes as previously described (22).

Preadipocyte culture
Stock cultures of Ob1771 cells were maintained in Dulbecco's modified Eagle's medium (Gibco, Cergy-Pontoise, France) supplemented with biotin, pantothenate, antibiotics, and 8% (v/v) fetal bovine serum as previously described (18).Experiments were performed after growth and differentiation of confluent cells in serum-free medium as previously described (18).Fatty acids, prostaglandins (Cayman Chemicals, Montluçon, France), GW2433, and BRL49653 were dissolved in ethanol and added at a 1:100 dilution into culture media.Ethanol concentration, which did not exceed 1%, had no effect on either adipose conversion or cyclic AMP production.Oil-Red O staining was performed as described previously (18).

Biochemical assays
Glycerol-3-phosphate dehydrogenase assays were carried out in duplicate at day 7 after confluence (18).Intracellular cyclic AMP was determined with a commercial kit by radioimmunoassay, according to the manufacturer's instructions (Amersham).Before assays, cyclic AMP was extracted with 1.2 ml of ice-cold ethanol-5 mM-EDTA (2:1, v/v).After scraping the cell monolayer and centrifugation, 1 ml of the supernatant was dried in a Speed-Vac evaporator.Duplicate samples were solubilized before assays in 150 l of 50 mM Tris-HCl buffer, pH 7.5, containing 4 mM EDTA and assayed at least in duplicate.

Statistical analysis
Statistical comparisons were performed on absolute values by Fisher's PLSD using the STATVIEW software package.

-6 ARA but not -3 ARA promotes adipogenesis
We performed experiments in confluent preadipocytes exposed for 7 days to serum-free medium in the absence ( Fig. 1A ) or the presence of various effectors.Differentiation was enhanced in cells exposed to the naturally abundant -6 ARA (Fig. 1B) compared with -3 ARA, which is only present at trace amounts in dietary fat sources (Fig. 1C).Differentiation in the presence of -6 ARA was severely impaired when aspirin was included as a cyclooxygenase inhibitor (Fig. 1D).Another polyunsaturated fatty acid of the -3 series, DHA, behaved similarly to -3 ARA (Fig. 1E).In agreement with our previous observations (20), carbacyclin, a stable prostacyclin analog that binds to the cell surface prostacyclin receptor IP-R, was highly adipogenic (Fig. 1F), exhibiting greater activity than GW2433, a specific PPAR ␦ agonist (17) (Fig. 1G), or a combination of GW2433 and BRL49663, a specific PPAR ␥ agonist (30) (Fig. 1H).In order to delineate at which step -6 ARA was active in the differentiation process, we first studied adipose conversion in the presence of maximally effective concentrations of PPAR ␦ and/or PPAR ␥ agonist using glycerol-3-phosphate dehydrogenase activity, an adipocyte indicator.
The sequential addition of PPAR agonists was based upon several considerations.i ) Early markers are expressed in PPAR ␥ Ϫ / Ϫ embryonic stem cells that can be committed into the adipose lineage, i.e., lipoprotein lipase and PPAR ␦ but not the adipocyte lipid binding protein (ALBP or aP2) (31); ii ) PPAR ␦ is expressed at near maximal levels in early confluent Ob1771 cells, in contrast to PPAR ␥ , which is highly expressed later during adipogenesis (32); and iii ) activation of PPAR ␦ by activators-ligands is required to upregulate the expression of PPAR ␥ (33,34).
Consistent with these observations, addition of the PPAR ␦ agonist from day 0 to day 3, followed by its removal and addition of the PPAR ␥ agonist from day 3 to day 7, proved to be optimal.Adipogenesis was similar to that observed in the presence of both agonists between day 0 and day 7.In contrast, discontinuous or continuous exposure to either agonist for 7 days was less efficient in promoting adipogenesis ( Table 1 , part A).Of note, when the PPAR ␥ agonist was present from day 3 to day 7, the adipogenic potency of -6 ARA added during the first 3 days was 3-fold higher than that of the PPAR ␦ agonist and this effect was severely reduced after cyclooxygenase inhibition (Table 1, part A).In contrast to -6 ARA and carbacyclin, the adipogenic potencies of -3 ARA and PPAR ␦ agonist were similar.Interestingly, in the presence of the PPAR ␦ agonist between day 0 and day 3, -6 and -3 ARA exhibited an adipogenic potency similar to that of the PPAR ␥ agonist (Table 1, part A).Altogether, these results show that the substantial effect of -6 ARA takes place at early step(s) of the differentiation process, in accordance with the fact that we did not observe prostacyclin synthesis and response through IP-R in differentiating, PPAR␥-expressing cells (21).
In order to gain further insights into the role of IP-R, adipogenesis of mouse embryo fibroblasts from wild-type and ip-r Ϫ/Ϫ mice ( 22) was compared upon stimulation by a combination of BMY45778, a specific agonist of IP-R like carbacyclin but unable to activate PPAR␦ (35), and the PPAR␥ agonist.Adipogenesis was decreased 2-fold in ipr Ϫ/Ϫ fibroblasts compared with that of wild-type fibroblasts while exposure to a combination of BMY45778 and PPAR␥ agonist did not enhance adipogenesis above that observed with the PPAR␥ agonist alone (data not shown).As no expression of C/EBP␤ and C/EBP␦ was observed in ip-r Ϫ/Ϫ mouse embryo fibroblasts in response to carbacyclin (22), it appears that the remarkable adipogenic potency of carbacyclin was likely due to its dual role, first as a ligand of IP-R and activation of the PKA pathway leading in turn to C/EBP␤ and C/EBP␦ expression and up-regulation of PPAR␥ expression (36), and second as a ligand of PPAR␦.

Only -6 ARA promotes extensive adipogenesis and cyclic AMP production
The potency of various long-chain fatty acids to stimulate early events of differentiation was next examined in preadipocytes exposed subsequently to the PPAR␥ agonist (Table 1, part B).A saturated fatty acid (palmitate) or monounsaturated fatty acids (palmitoleate and oleate) was poorly adipogenic compared with -6 ARA.Furthermore, two -3 polyunsaturated fatty acids, EPA and DHA, were also poorly adipogenic.The higher adipogenic activity of -6 ARA compared with other fatty acids cannot be ascribed to its higher affinity for PPAR␦, as the latter binds arachidonic acid, saturated, monounsaturated, and -3 polyunsaturated fatty acids with similar affinity (17).As -6 ARA was unique among natural fatty acids in pro-moting extensive adipocyte differentiation, we compared cyclic AMP production after a very short exposure of preadipocytes to concentrations of the various long-chain fatty acids (Table 2).-6 ARA increased cyclic AMP production by 15-fold in 5 min in confluent preadipocytes, and this effect was abolished by indomethacin, another cyclooxygenase inhibitor.Carbacyclin increased cyclic AMP production by 13-fold and, as anticipated, indomethacin had no effect.Compared with control cells, PKA activity was increased 3-and 4-fold in the presence of -6 ARA (10 M) and carbacyclin (1 M), respectively (data not shown).-3 ARA, EPA, DHA, palmitic acid, palmitoleic acid, oleic acid, PPAR␦ agonist, or PPAR␥ agonist had no effect on cyclic AMP production.In order to investigate why -3 polyunsaturated fatty acids were not potent adipogenic agents, we attempted to inhibit PPAR␦ and/or PPAR␥ activity in the presence of -3 ARA.Adipogenesis was unaffected compared with that observed in the presence of PPAR agonists only, excluding PPARs as possible targets of -3 ARA (data not shown).Thus, we investigated in preadipocytes the possible effect on cyclic AMP production of -3 polyunsaturated fatty acids that have been reported to inhibit the cyclic AMP-dependent PKA (37).In the presence of -6 ARA (5 M), addition of -3 ARA, EPA, or DHA (10 M each) inhibited cyclic AMP production by 51%, 91%, and 9% respectively, whereas, in the presence of -6 ARA (10 M), a 33% inhibition of this production was observed by addition of EPA (10 M) (not shown).These results suggest that -3 polyunsaturated fatty acids inhibit in preadipocytes the cyclic AMP-signaling pathways triggered by arachidonic acid at level(s) upstream of PKA.

In vivo -6 polyunsaturated fatty acids are more adipogenic than -3 polyunsaturated fatty acids
To address whether the fatty acid composition of highfat diets promotes adipose tissue development, we carried out nutritional studies in wild-type and ip-r Ϫ/Ϫ mice.We chose a corn oil-supplemented diet rich in linoleic acid, a long-known essential fatty acid precursor of arachidonic acid (LO diet) compared with an isocaloric diet enriched with a mixture of corn oil and perilla oil rich in ␣-linolenic acid, another essential fatty acid precursor of EPA and DHA (LO/LL diet).The female mice designated to be mothers were fed diets containing either LO diet or LO/LL diet from 4 weeks of age.Four weeks later, these mice were bred to male mice and remained on the same diet.Pups were kept with their mothers until weaning and then maintained on the same respective diets.
From weaning to 22 weeks of age, body weight of wildtype mice fed LO diet was higher than that of animals fed LO/LL diet and this difference persisted to a large extent at the adult age (Fig. 2) despite a similar food intake (Table 3).Strikingly, the body weight of ip-r Ϫ/Ϫ mice on either type of diets was similar.As the amount of food intake was also similar despite differences in caloric intake, this suggests increased thermogenesis in ip-r Ϫ/Ϫ mice fed a high-fat diet.Moreover, body weight of ip-r Ϫ/Ϫ mice was indistinguishable when fed LO diet or LO/LL diet, demonstrating the critical role of -6 polyunsaturated fatty acids in body weight gain during pregnancy and/or the suckling period.No difference was observed in body length between wild-type and ip-r Ϫ/Ϫ mice fed either type of diets.To evaluate adiposity, we measured total fat mass and epididymal fat pad weight of wild-type and ip-r Ϫ/Ϫ mice at 8 weeks of age (Table 3).Fat mass of wild-type mice fed standard diet or LO/LL diet was identical, whereas that of animals fed LO diet was increased.In contrast, fat mass was identical in ip-r Ϫ/Ϫ mice fed either diet.Consistent with these observations, epididymal fat pad weight of wild-type mice was significantly higher in mice fed LO diet than in mice fed LO/LL diet or standard diet.Again, there was no significant difference in the epididymal fat pad weights of ip-r Ϫ/Ϫ mice fed either type of diets.Adipocyte size of epididymal fat pad was increased 1.9-fold in wild-type mice fed LO diet compared with the two other diets, and this was accompanied by a decrease of ad-Fig.2. Body weight of wild-type (A) and ip-r Ϫ/Ϫ male mice (B).During the pregnancy-lactation period, mother mice were fed a standard diet (diamond), a corn oil-enriched diet (triangle), or a mixture of corn oil and perilla oil-enriched diet (square).From weaning onward, male pups were maintained on the same diet than their mothers had consumed.(n ϭ 20), *P Ͻ 0.05 versus wild-type mice fed a standard diet.TABLE 3. Food intake, fat mass, and epididymal fat pad weight in 8-week-old wild type and ip-r Ϫ/Ϫ mice fed either a standard diet (Chow) or a high-fat diet enriched with linoleic acid (LO) or enriched with linoleic acid and ␣-linolenic acids (LO/LL) ipocyte number (Fig. 3).Compared with wild-type mice, adipocyte size was decreased and adipocyte number was increased in ip-r Ϫ/Ϫ mice fed standard diet, yet there was no difference in adipocyte size and number of the epi-didymal fat pad of ip-r Ϫ/Ϫ mice fed either diet (Fig. 3).Surprisingly, when wild-type mother mice were fed a standard diet and pups were fed after weaning a LO or LO/LL diet, body weight of animals fed LO diet was not significantly  Redundant pathways and dietary long chain-fatty acids (LCFA) implicated in adipogenesis.This scheme assumes that three ligand/ cell surface receptor systems concur to upregulate the expression of C/EBP␤ and C/EBP␦, i.e., prostacyclin/IP-R (22,39), leukemia inhibitory factor (LIF) receptor (22,39), and fibroblast growth factor (FGF)-10 receptor (48), and to promote adipogenesis.Dietary linoleic acid and/or arachidonic acid favor in preadipocytes prostacyclin synthesis and release.Thus, arachidonic acid via prostacyclin triggers a key event, plays a unique role in activating the protein kinase A pathway by means of IP-R, and enhances the differentiation process.Other dietary LCFA behave as mere activators/ligands of PPAR␦ and subsequently of PPAR␥.Upon terminal differentiation, leukemia inhibitory factor and fibroblast growth factor-10 are no longer produced.Production of prostacyclin and other prostaglandins ceases and is accompanied by reduced expression and loss of functional IP (21,49,50).Therefore, in adipocytes, PPAR␥ and possibly PPAR␦ then act as the molecular sensors of all dietary LCFA at a time where endogenous LCFA synthesis (not shown) becomes significant (51).Epidermal (keratinocyte) fatty acid binding protein in preadipocytes and adipocytes and adipocyte lipid binding protein in adipocytes are assumed to bind and transport LCFA (51,52).
higher than that of those fed LO/LL diet, emphasizing the importance of linoleic acid-enriched diet during the pregnancy-lactation period.

DISCUSSION
The development of adipose tissue depends on the existence of redundant pathways that ultimately upregulate the expression of PPAR␥.This expression relies predominantly on the expression of C/EBP␤ and C/EBP␦ (36) and disruption of both genes severely impairs the formation of adipose tissue (25).We had identified cues that trigger the expression of C/EBP␤ and C/EBP␦ in preadipocytes, i.e., leukemia inhibitory factor and its cognate cell surface receptor activating the ERK pathway (38) and the prostacyclin/IP-R system activating the PKA pathway (22).Our present results show that polyunsaturated fatty acids of the -6 series are more adipogenic both in vitro and in vivo compared with their -3 counterparts.On one hand, -6 ARA is unique via prostacyclin in activating the PKA pathway.On the other hand, -3 polyunsaturated fatty acids do not affect this pathway as well as prostaglandin I 3 , a product synthesized from EPA (not shown).Figure 4 summarizes our views on redundant pathways and on the role of arachidonic acid and other long-chain fatty acids in promoting adipogenesis.
In light of our results from in vitro studies reported herein, a high-fat diet rich in linoleic acid could stimulate arachidonic acid formation and, through prostacyclin synthesis, activate cyclic AMP-dependent signaling pathways in preadipocytes.This would favor the formation of mature adipocytes whose lipid filling is then increased to cope with the high exogenous levels of fatty acids.In contrast, in wild-type mice fed a high-fat diet containing a mixture of linoleic and ␣-linolenic acids, the lowered production of cyclic AMP may limit the formation of adipocytes, leading to hyperplasia to accommodate the fatty acid supply.This alteration of cyclic AMP production could be due to a decreased arachidonic acid synthesis from linoleic acid through inhibition of ⌬6 desaturase activity by ␣-linolenic acid and its metabolites (39).In ip-r Ϫ/Ϫ mice, no activation of the PKA pathway occurs through the prostacyclin receptor, and the adipogenic effect of -6 and -3 polyunsaturated fatty acids is similar.Despite prostacyclin synthesis in ip-r Ϫ/Ϫ mice, no difference in adipose tissue mass was observed between mice fed LO or LO/LL diet, suggesting that the activation of PPAR␦ by prostacyclin, if any, had no significant impact on the overall differentiation process.Our finding that inclusion of ␣-linolenic acid in an isocaloric diet rich in linoleic acid prevents the enhancement of fat mass is consistent with our in vitro observations.Varying the proportion of these essential fatty acids thus should alter the proportion of arachidonic acid versus EPA and DHA (39) and could lead to changes in the pattern of adipose tissue development which occurs during pregnancy and the suckling period.
Remarkably, although no difference in the litter size was observed in wild-type mice fed either diet, pups from mother mice fed LO diet were, at weaning, 50% heavier than those from mice fed LO/LL diet.This raises an important issue in humans, as adipose tissue is being formed during the third trimester of gestation and sensitive periods of its development take place before 1 year of age (1).In this respect, the overweight prevalence among US children of 6 to 11 months of age increased in boys from 4.0% in 1976 through 1980 to 7.5% in 1988 though 1994, and in girls from 6.2% to 10.8% during the same periods of time (40).It is of interest to observe that the content of linoleic acid of mature breast milk of US women, a reflection of their fat intake, has increased from 5% to 17% between 1945 and 1995 (r ϭ 0.85, P Ͻ 0.001, n ϭ 29).Furthermore, despite the fact that the ratio of linoleic acid to ␣-linolenic acid in mature breast milk of US and European women is similar, it is also interesting to note that the ratio of arachidonic acid to DHA is 1.8-fold higher in the milk of US women due to its low DHA content (41)(42)(43).It has been reported that breast feeding may help decrease the prevalence of overweight and obesity in childhood (44,45).Although the enhanced fatness was attributed to the greater energy intake of formula-fed infants (46), the fatty acid composition should be considered, as the percentage of linoleic acid is increased by approximately 50% in infant formula compared with breast milk (47).We suggest that, at a very early age where energy expenditure appears rather similar between individuals, large amounts of linoleic acid consumed during pregnancy, suckling period, and early infancy are important determinants of physiological events implicated at a time when adipose tissue is in a dynamic phase of its development, and that could lead to childhood obesity.

Fig. 3 .
Fig. 3. Adipocyte number and size in epididymal fat pad of 8-week-old male wild-type and ip-r Ϫ/Ϫ mice (n ϭ 3 in each group) fed after weaning to a standard diet (SD), LO diet, or LO/LL diet.Values are expressed as mean Ϯ SEM; * P Ͻ 0.05 versus wild-type fed a standard diet.O, P Ͻ 0.05 versus wild-type fed LO diet; X, P Ͻ 0.01 versus wild-type fed LO/LL diet.

Fig. 4 .
Fig.4.Redundant pathways and dietary long chain-fatty acids (LCFA) implicated in adipogenesis.This scheme assumes that three ligand/ cell surface receptor systems concur to upregulate the expression of C/EBP␤ and C/EBP␦, i.e., prostacyclin/IP-R(22,39), leukemia inhibitory factor (LIF) receptor(22,39), and fibroblast growth factor (FGF)-10 receptor (48), and to promote adipogenesis.Dietary linoleic acid and/or arachidonic acid favor in preadipocytes prostacyclin synthesis and release.Thus, arachidonic acid via prostacyclin triggers a key event, plays a unique role in activating the protein kinase A pathway by means of IP-R, and enhances the differentiation process.Other dietary LCFA behave as mere activators/ligands of PPAR␦ and subsequently of PPAR␥.Upon terminal differentiation, leukemia inhibitory factor and fibroblast growth factor-10 are no longer produced.Production of prostacyclin and other prostaglandins ceases and is accompanied by reduced expression and loss of functional IP(21,49,50).Therefore, in adipocytes, PPAR␥ and possibly PPAR␦ then act as the molecular sensors of all dietary LCFA at a time where endogenous LCFA synthesis (not shown) becomes significant(51).Epidermal (keratinocyte) fatty acid binding protein in preadipocytes and adipocytes and adipocyte lipid binding protein in adipocytes are assumed to bind and transport LCFA(51,52).

TABLE 1 .
Comparative stimulation of adipogenesis by PPAR agonists and long-chain fatty acids Values are expressed as mean Ϯ SEM of experiments performed on 3-10 independent series of cells.

TABLE 2 .
Comparative cyclic AMP production by arachidonic acid and various effectors of adipogenesis a Values are expressed as mean of duplicate values representative of experiments performed on three independent series of cells.Fatty acids and indomethacin (added 15 min prior to fatty acid) were present at 10 M; ␦ agonist, GW 2433 (1 M); ␥ agonist, BRL 49653 (0.5 M); carbacyclin (0.2 M).