*2.2. FX Enhanced Metabolic Rate*

Mice were placed in respiratory chambers at week 3 for six days to continuously monitor O2 consumption (VO2), CO2 production (VCO2). These and calculated respiratory quotients (RQ) values throughout a day and respective area under the curve (AUC) values were shown in Figure 1. At seven out of 24 hour time points, high sucrose diet supplemented with fucoxanthin (HS + F) or high-fat diet (HF + F) groups had significantly higher VO2 and VCO2, respectively, compared to HS and HF groups (Figure 1A,B). Diet did not change the AUC of VO2 (Figure 1A), but the high fat diet decreased that of the VCO2 (Figure 1B). FX increased the AUC of both VO2 and VCO2 through the dark phase and all day without an interaction with the diet. Mice fed high fat diets had significantly lower AUC of RQ 

(*p* < 0.05), indicating that these mice used more fat as their energy source. FX supplementation did not change RQ (Figure 1C), implying that FX might have enhanced the common aerobic metabolic pathway of glucose and fat metabolism. Indeed, FX was shown to enhance the utilization of glucose in skeletal muscle [32] and liver [29], as well as fatty acid Ά-oxidation in WAT [28] and liver [30]. 

**Figure 1.** O2 consumption (**A**), CO 2 production (**B**), respiratory quotient (**C**) and their area under the curve (AUC) of test mice at week 3. Mice were individually placed in metabolic chambers with free access to water and their respective diet and monitored for six days. Values are means and error bars are SD (*<sup>n</sup>* = 4). \* denotes significant difference between groups HS and HS + F, *p* < 0.05; # denotes significant difference between groups HF and HF + F, *p* < 0.05, analyzed by the Student's *t*-test. AUCs of O2 consumption (VO2), CO2 production (VCO2) were analyzed by twoway ANOVA. The AUC of RQ was analyzed by the Wilcoxon rank-sum test. 

**Figure 2.** Thermogenic (**A**) and mitochondrial homeostasis-related (**B**) mRNA levels in brown adipose tissue of mice fed test diets for five weeks. Values are means, and error bars are SEM (*<sup>n</sup>* = 4). \* denotes significant effect by either dietary factor at *p* < 0.05 analyzed by two-way ANOVA. When an interaction (*p* < 0.05) between diet and FX existed, the significance of differences among the HS, HS + F, HF and HF + F groups was further analyzed by Duncan's multiple range test. HS, HS + F, HF and HF + F: As indicated in Figure 1. Relative mRNA expression was measured by real-time qRT-PCR using Ά-actin as the internal control and normalized to group HS. 



*2.3. Effect of FX on Thermogenic and Mitochondrial Homeostasis-Related Gene Expressions in BAT and Serum Hormone Concentration* 

Both high fat diet and FX lowered Ά3-AR, Mfn2 and Fis1 mRNA levels in the BAT (*p* < 0.05). Mice fed high fat diets had significantly lower CIDEA and ERR΅ mRNA expression in the BAT (*p* < 0.05). FX also decreased Prdm16 and PPARΈ but increased Mfn1 mRNA in the BAT. Neither diet nor FX changed PGC-1΅ UCP1, PPAR΅, PPAR·, Ά2-AR, HSL, Dio2, TFAM, NRF1, NRF2 and OPA1 mRNA expressions in  BAT. Among them, diet and FX had an interaction effect on the PPAR ΅ and NRF1 mRNA expression (*p* < 0.05). FX increased PPAR ΅ mRNA expression in the high sucrose diet-fed mice, but not in the high fat diet-fed mice. FX reduced NRF1 mRNA expression in the high fat diet-fed mice, but not in the high sucrose diet-fed mice (Figure 2). 

In this study, BAT mass was significantly increased by FX as in some of previous reports [23–25], although other studies did not observe such an effect [31,38]. This discrepancy might be associated with the different mouse strain, gender and diet formula used. Although BAT mass was increased by FX and positively correlated to carbon dioxide production (*p* = 0.02) during the dark period (Supplementary Figure S1), we did not observe any thermogenic genes upregulated in BAT. Ά3-AR, Prdm16 and PPARΈ mRNA were even downregulated by FX. To validate these results, we even conducted the Housekeeping Genes PCR array (Qiagen, Germantown, MD, USA) and employed three popular software packages (GeNorm, NormFinder and BestKeeper) to confirm the use of Ά-actin as the most stable housekeeping gene. 

On the other hand, our serum hormone analysis showed that the norepinephrine concentration was decreased by FX (Table 4). Norepinephrine, the sympathetic neurotransmitter, is the main *in vivo* stimulator of the adrenergic signaling mediating the BAT thermogenic machinery. It is not known whether this low serum norepinephrine in FX-fed mice is related to the very minor change in the BAT gene expression irrespective of the enlarged mass observed in this study. Moreover, no significant correlation between BAT mass and O2 consumption was observed (Supplementary Figure S1). Taken together, it is speculated that BAT contributes little, if any, to the FX enhanced O2 consumption. 

**Table 4.** Serum thyroid hormone, (nor)epinephrine and corticosterone concentrations of C57BL/6J male mice fed test diets for five weeks. Values are the means ± SD ( *<sup>n</sup>* = 4). \* denotes a significant effect by either dietary factor at *p* < 0.05 analyzed by two-way ANOVA. T4, thyroxine; T3, triiodothyronine; NE, norepinephrine; E, epinephrine; Cort, corticosterone. 

