*2.4. FX Induced Thermogenic-Related Gene Expressions in eWAT and iWAT*

While the high fat diet decreased CIDEA mRNA expression, FX significantly increased mRNA expressions of CIDEA, PGC-1 ΅, ERR ΅, PPAR·, Ά3-AR, Dio2, PPAR ΅ and HSL in eWAT (Figure 3A). The high fat diet decreased UCP1, CIDEA, Prdm16, ERR ΅, PPAR·, Dio2 and PPAR ΅ mRNA expression, but FX increased CIDEA, PGC-1 ΅, ERR ΅, Ά3-AR, Dio2 and PPAR ΅ mRNA expression in iWAT (Figure 3B). FX did not significantly affect mRNA expressions of these genes in the abdominal rWAT (data not shown), except for upregulating PGC-1 ΅. 

Expressions of UCP1, Prdm16, CIDEA and PGC-1 ΅ in WAT are indicators for the "browning" of WAT, which could reduce the adverse effects of WAT and help to improve metabolic health [34,39]. In this study, we observed increases in the mRNA of CIDEA and PGC-1 ΅, but not UCP1 and Prdm16 in eWAT and iWAT in mice fed the FX diets. The expression of UCP1 is transcriptionally regulated by the adrenergic signaling ( Ά-AR, cAMP-dependent protein kinase A (PKA), *etc.*) coupled to PGC-1 ΅ and PPAR [40]. PGC-1 ΅ has been shown to be required for exercise-induced UCP1 expression in WAT [41]. Prdm16, by co-activating PGC-1 ΅ increased mitochondrial content, as well as enhanced uncoupled respiration. Prdm16 activates a robust brown fat phenotype, including the induction of PGC-1 ΅, UCP1 and Dio2 [42–44]. In this study, although UCP1 mRNA in these WATs of FX-fed mice was slightly higher than that of the respective controls, the difference did not reach statistical significance, due to the low number of animals per group ( *n* = 4), as well as the large individual variations. Like our study, Woo *et al.* did not observe increased UCP1 in eWAT of mice fed the 0.2% FX diet, although mice fed the 0.05% FX diet did show an increase of UCP1 in eWAT [31]. CIDEA has been shown to inhibit uncoupling activity of UCP1 when co-expressed in yeast and increases mitochondrial coupling by suppressing UCP1 expression [45]. Therefore, the unchanged UCP1 mRNA level in this study could also be related to the dose of FX and the elevated CIDEA mRNA expression in WAT. Induction of CIDEA has been used as a marker for the emergence of brown adipocyte-like cells in WAT [46]. As we found that only two of the four "beige" adipocyte-specific genes were upregulated, whether FX induces the whole program of WAT "browning" cannot be confirmed. 

On the other hand, we observed that Ά3-AR, Dio2, PGC-1 ΅, ERR ΅ and PPAR ΅ mRNA in both eWAT and iWAT and HSL in eWAT were elevated by FX. The adrenergic signaling ( Ά3-AR, PKA, *etc.*) coupled to PGC-1 ΅ and PPAR regulates UCP1 expression [40]. PGC-1 ΅ also controls mitochondria biogenesis and respiratory function through targeting multiple transcription factors, like NRF1, NRF2 and ERR ΅. Ά3-AR is expressed abundantly and predominantly in brown and white adipocytes. Treatment of mice with Ά3-AR agonists increases oxygen consumption [47]. Under Ά3-AR stimulation, PKA is activated and further turns on the transcription of Dio2 and PGC-1 ΅ and phosphorylation of HSL [48,49]. Dio2 catalyzes the conversion of T4 to T3 (the active form of thyroid hormone) in most  peripheral tissues and is essential for the adrenergic receptor mediated thermogenesis in BAT [50]. T3 increases oxygen consumption, which is mediated through PGC-1΅ and NRF1 [51]. In this study, increased Dio2 mRNA in WATs is speculated to increase local T3 content and further stimulate PGC-1΅ mRNA expression and the metabolic rate of mice. 

Fucoxanthinol and amarouciaxanthin A are two major metabolites of FX in mouse plasma and liver [52]. In mouse adipose tissue, amarouciaxanthin A is the major metabolite of FX [53]. *In vitro*, amarouciaxanthin A showed a higher activity in suppressing 3T3-L1 adipocyte differentiation than FX, fucoxanthinol and amarouciaxanthin B [54]. The suppression is associated with downregulations of PPAR· and CCAAT-enhancer-binding protein ΅ (C/EBP΅) mRNA levels. In the present *in vivo* study, however, PPAR· mRNA expression in eWAT was upregulated by FX, implying that inhibition of adipocyte differentiation might have a minor role on the low WAT mass observed in our *in vivo* study. 

**Figure 3.** Thermogenic gene expressions in epididymal (**A**) and inguinal (**B**) white adipose tissue. Values are means, and error bars are SEM (*<sup>n</sup>* <sup>=</sup> 4). \* denotes significant effect by either dietary factor at *p* < 0.05 analyzed by two-way ANOVA. HS, HS + F, HF and HF + F: as indicated in Figure 1. The measurement of relative mRNA expression is as indicated in Figure 2.

*2.5. FX Increased Mitochondrial Biogenesis and Fusion-Related Gene Expressions in eWAT and iWAT* 

FX increased NRF1 and NRF2 mRNA expressions in eWAT, but not iWAT. Diet did not affect TFAM, NRF1 and NRF2 mRNA expressions in eWAT and iWAT (Figure 4). The high fat diet decreased Mfn1, Mfn2, OPA1 and Fis1 mRNA levels in iWAT, but not eWAT. In contrast, FX increased mitochondrial fusion-related genes, including Mfn1, Mfn2 and OPA1, but not fission-related gene Fis1 and mRNA expressions in both eWAT and iWAT (Figure 5). Mitochondria provide ~90% of the cellular energy supply and are also the most important organelle of metabolism. PGC-1΅ is a strong promoter of mitochondrial biogenesis and oxidative metabolism through NRF1, NRF2, ERR΅, PPAR· and PPAR΅ [13–15]. PGC-1΅ targets NRF1 and NRF2 directly or indirectly through ERR΅ in stimulating nuclear genes required for mitochondrial biogenesis and respiration function [14]. PGC-1΅ interacts with PPAR΅ in the transcription control of genes encoding mitochondrial fatty acid oxidation enzymes [55]. PGC-1΅ acts as a partner of PPAR· in the induction of adaptive thermogenesis in brown fat and adipocyte differentiation [14]. The increased metabolic rate of mice fed the FX diets in this study coincides with increased PGC-1΅ networks. ERR΅ further activates the transcriptional activity of the Mfn2 promoter, and the effect was synergic with those of PGC-1΅ [16]. In addition to Mfn2, other mitochondrial fusion-related genes (Mfn1 and OPA1) were also increased by FX in this study. In contrast, mitochondrial fission-related gene Fis1 in WATs was not affected by FX. Mitochondrial fusion is frequently found in metabolically active cells [56], and elevated mitochondrial fusion genes agreed with enhanced VO2 and VCO2 and the decreased WAT mass of mice in the present study. Among these genes, Mfn2 was shown to be stimulated under cold, adrenergic agonist or exercise treatment through the regulation of PGC-1΅ and ERR΅ [16,57]. Therefore, the elevated mitochondrial fusion genes by FX were in accordance with increased PGC-1΅ and ERR΅ mRNA in WATs. 
