Discussion

Before we found that. hepatic SAA mRNA is induced by dietary DHA in pigs ³¹, we wanted to know that whether human hepatic SAA is induced by DHA too.

Therefore, we chosen the human hepatocarcinoma cell line, HepG2 cell, to study the effect of human SAA on human hepatic gene expression. Hepatic SREBP1c and FAS are involved in lipogenesis ^44,45 and SREBP1c up-regulates FAS expression ⁴⁶. Expression of FAS is also regulated by PPARα because the addition of PPARα ligands was shown to FAS expression in wild type mice, but not in PPARα-/- mice ⁴⁷. In our experiments, SAA down regulated the expression of PPARα which may then reduce FAS mRNA expression. The SAA treatment had no effect on SREBP1c transcription, suggesting that there is no role of this transcription factor in the SAA regulation of FAS. PPARα can regulate fatty acid oxidation by binding to the PPAR response element (PPRE) of the ACO promoter to increase the expression of ACO ⁴⁸, a key enzyme in the β-oxidation of fatty acids in peroxisomes. We found that ACO expression was down-regulated by hSAA1 treatment, speculatively from the SAA-decreased PPARα. HNFα and L-FABP, two proteins that participate in lipid transport within the cytoplasm, were decreased by hSAA1 treatment. Thus, SAA1 may decrease the transportation of fatty acids in cytoplasm and consequently decrease the utilization of fatty acids in adipocytes. L-FABP is a target gene of PPARα which suggests that the fatty acid transport ability is affected by regulation of PPARα

49.

Since DHA has been shown to increase SAA1 expression in porcine liver ³¹ and hepatic cells ⁴⁴ then we treated 50 and 100 μM DHA in human preadipocytes to study the gene expression patterns in contract to SAA effects. DHA increased expression of SAA, TNFα and IL-6, three genes considered as proinflammatory elements. However,

the DHA-induced SAA expression was about two folds in human adipocytes and four folds in porcine hepatocytes ⁴⁴. These effects are much lower those is observed in inflammatory responses (15 to 250 times) in animal studies ^50-52. When human monocytes were infected with bacteria, the concentrations of IL-6 and TNFα are increased to 30 and 80 times, respectively ⁵³. And by murine cytomegalovirus infection, TNFα and IL-6 are increased to 150 and 1500 times, respectively ⁵⁴. These are acute phase inflammatory responses. The lower SAA1, IL-6 and TNFα responses induced by DHA compared to true inflammatory responses may represent distinctive physiological responses. IL-6 has been shown to inhibit lipopolysacharide-induced TNFα production in both human monocytes and human monocytic line U937 ^55,56. Therefore, IL-6 is a likely anti-inflammatory cytokine that can stimulate the expression of other anti-inflammatory cytokines and factors, for example, IL-10, IL-1 receptor antagonist, soluable TNFα receptor and cortisol during acute inflammation in normal subjects ^57-59. It has also been reported that the administration of recombinant IL-6 to IL-6^-/- mice suppressed the expression of TNFα ⁶⁰, suggesting an anti-inflammatory effect of IL-6. Other reports also indicate that DHA treatment decreases IL-2 and interferon gamma production (both are regarded as inflammation cytokines), whereas the production of IL-10 was increased in Jurkat T cells ^61,62. Furthermore, the definition of inflammation should include several clinical signs;

such as the inflammation-related cytokine expressions, elevation of macrophage or polymorphonuclear leukocyte infiltration, increase of white blood cell accumulation, the increased expression of cyclooxygenase, eicosanoids, leukotrienes and reactive oxygen species. According to these signs, we cannot define whether DHA induces inflammation in the system, based only on the increased expression of IL-6 and TNFα.

The expression of PPARγ occurs primarily in adipocytes and PPARγ participates in the transcriptional activation of numerous adipogenic genes, including FAS, SREBP1c and LPL ^63,64. We demonstrated that PPARα, FAS and LPL expressions were decreased by hSAA1 treatment in human adipocytes, indicating that SAA1 can reduce adipogenic and lipogenic activitiess and potentially decrease accumulation of triacylglycerol in adipocytes. Moreover, DHA increased the expression of SAA and decreased PPARα, LPL and FAS mRNA. Since, DHA and SAA treatments had similar effects on adipogenic and lipogenic genes in human adipocytes. The data suggest that the mechanism for the DHA effects may be through an increase in SAA.

The direct evidence for this hypothesis has not yet been shown.

The function of perilipin protein is to prevent untimely lipid mobilization and maintain the structure of lipid droplets. When energy is needed, the PKA-phosphorylated perilipin allows PKA-activated (phosphorylated) HSL to hydrolyze triacylglycerols and release free fatty acids and glycerol ^65,66. There is no glycerol kinase to metabolize glycerol in adipocytes so glycerol is released into the circulation in vivo or into the culture medium in vitro. Hence, glycerol release is an indication of lipolysis. In the current study, both hSAA1 and DHA treatments increased the glycerol concentration in culture medium, indicating elevated lipolysis.

Because IL-6 increases the release of glycerol ⁶⁷ and stimulates fatty acid oxidation ^67,68 in human adipocytes, our finding that hSAA1 and DHA treatments increased the expression of IL-6 suggest that at least part of the effect was mediated by IL-6. In 3T3L1 adipocytes, TNFα down-regulates the expression of LPL and increases glycerol release ^69,70. TNFα also can inhibit the activity of perilipin by phosphorylation though mitogen activated protein kinases ⁷⁰ and activate PKA to phosphorylate perilipin ⁷¹ to promote lipolysis. The current study found that hSAA1

and DHA treatments increased the expression of TNFα so we speculate that hSAA1- and DHA-induced TNFα reduced the expression of perilipin. Moreover, perilipin expression can be stimulated by PPARγ to accumulate lipid ⁷². Our current study showed that both DHA and hSAA1 treatments reduced the expression of PPARα mRNA to potentially reduce expression of perilipin. The combined mechanisms to reduce perilipin and increase HSL expression coupled with accentuated phosphorylation potential may all work in concert to stimulate lipolysis.

DHA may inhibit or increase TNFα and IL-6 expression or concentration depending on the cell type or culture medium conditions. Dietary DHA-rich fish oil supplementation up-regulates serum IL-6 and TNFα concentrations in human leukocytes ⁷³. Treatment with DHA increases the production of TNFα and IL-6 in macrophages ⁷⁴. In contrast, 50 or 500 μM of DHA had no effect on TNFα in murine 3T3-L1 adipocytes ⁷⁵. Furthermore, TNFα and IL-6 secretion by human mononuclear cells is inhibited by dietary fish-oil supplementation ⁷⁶. In the current study, we demonstrated that DHA increased the expression of TNFα and IL-6. These increments may then modify the cellular function of human adipocytes.

The DHA treatment promotes PPARα mRNA expression ^77,78. The PUFA are natural PPARα ligands, with simultaneous stimulation of fatty acid oxidation and inhibition of fatty acid synthesis ⁷⁹. Incubation of human adipocytes with DHA increased the expression of PPARα whereas incubation with hSAA1 decreased the expression of PPARα so that the effects of DHA and SAA were opposite. Which regulatory mechanisms are associated with these divergent effects of DHA and SAA are not clear now. In liver cells (HepG2), SAA reduced both PPARα and ACO mRNAs suggesting inhibition of fatty acid oxidation.

Although conjugated linoleic acid (CLA) and DHA are different in the position

and number of double bonds, they have similar effects on alleviating obesity and the expression of lipid metabolism related genes ⁸⁰. Microarray profiling of white adipose tissue (WAT) of mice fed trans-10, cis-12 CLA indicated that CLA reduced the expression of genes involved in lipogenesis and adipogenesis such as LPL, FAS, fatty acid binding protein 4 and PPARγ⁸¹. In the current study, DHA treatment also reduced LPL, FAS and PPARγ in human adipocytes and increased transcript levels for IL-6, TNFα. Similar to our findings, CLA reduces the expression of lipid synthesis genes, such as FAS,HSL, LPL ⁸², and perilipin ⁸³. Moreover, CLA increases the gene expression of IL-6 in human adipose tissue ⁸³. DHA treatment reduces body weight and energy intake in human subjects ^21,84 and promotes gene expression of TNFα and reduces PPARγ. However, CLA has no effect on these genes and body weight in humans ⁸³. Although there are similar effects between CLA and DHA on prevention or improvement of obesity, the role and function of different fatty acids can be very different.

In conclusion, we have demonstrated that DHA increases the expression of SAA1 in human adipocytes and that the effects of SAA1 and DHA on expression of the lipolytic genes, TNFα, IL-6, HSL and perilipin in human adipocytes are parallel.

In contrast, PPARγ, LPL, FAS mRNAs were decreased by the SAA1 and DHA treatments, suggesting that DHA and SAA1 also inhibit adipogenesis in human adipocytes. These results suggest that DHA may enhance lipolytic activity and decrease lipogenic and adipogenic activity by regulating the expression of SAA1.

Such DHA function will be useful for developing new approaches to reduce body fat deposition.

Chapter 3 Tetranectin promotes adipogenesis and