**3. Results**

#### *3.1. UCP3 Ablation A*ff*ects Resting Metabolic Rate, Energy Expenditure, and Fatty Acid Utilization in Adult Mice Acclimated at Thermoneutrality*

The resting metabolic rate (RMR), the resting energy expenditure (REE), and the respiratory quotient (RQ) were detected in 4–5 month old animals, which were acclimated at thermoneutrality for at least two weeks. RMR was significantly reduced in KO mice as compared to WT mice (−30%) (Figure 1), both when it was expressed in Litres oxygen/(hour Kg0.75) and when expressed in Litres oxygen/(hour g of lean mass). The RQ was increased in KO mice, which indicates a lower use of lipids as metabolic substrates in these animals. Indeed, in KO mice, the contribution of fatty acid oxidation to REE was reduced by about 27% relative to that in WT mice (Figure 1). In addition, in adult KO animals, food intake resulted in a reduction of about 15%, being the values 3.1 ± 0.11 and 2.63 ± 0.014 g food/day, for WT and KO mice, respectively (*n* = 6, *p* < 0.05 by Student's *t*-test).

**Figure 1.** Effect of UCP3 ablation on metabolic parameters detected in WT and KO mice housed at thermoneutrality for 2–3 weeks and fed a standard diet (**a**), (**b**) Resting Metabolic Rate, (**c**) Respiratory Quotient, (**d**), (**e**) Energy Expenditure, (**f**) contribution of fatty acid oxidation to energy expenditure (EE). Values represent mean ± SE of 6–7 animals for WT and KO mice, respectively. Statistical analyses were performed by two-tailed Student's *T*-test, \* *p* < 0.05 vs. WT.

SkM mitochondria that were isolated from WT and KO mice did not show significant differences in respiratory parameters (State 4 and State 3) when using pyruvate + malate or succinate + rotenone as substrate (Figure 2a,b). On the other hand, a significant State 3 inhibition was observed in mitochondria from KO mice when palmitoyl carnitine + malate was used as the substrate, which thus indicated the lower ability of mitochondria to oxidize fatty acids (Figure 2c).

**Figure 2.** Impact of UCP3 ablation on mitochondrial respiration rate, detected in the presence of different substrates: (**a**) Succinate (+rotenone), (**b**) Pyruvate (+malate), and (**c**) Palmitoyl-carnitine (+malate). Skeletal muscle mitochondria were isolated from WT and KO mice housed at thermoneutrality for 2–3 weeks and fed a standard diet. Values represent mean ± SE of six different animals. Statistical analyses were performed by two-tailed Student's *t*-test, \* *p* < 0.05 vs. WT.

Subsequently, we evaluated whether UCP3 ablation could influence respiratory chain complexes activity. In-gel activity of each individual respiratory complex (I, II, and IV) did not differ between WT and KO mitochondria (Figure 3). In view of the above results, we wondered whether UCP3 ablation would affect body fat accumulation in mice that were housed at thermoneutrality since weaning by determining their metabolic phenotype after feeding for 80 days, either with a standard/low fat diet (STD) or a high fat diet (HFD).

**Figure 3.** BN-PAGE-based analysis of individual respiratory complexes from dodecylmaltoside-solubilized crude mitochondria from SkM of WT and KO mice housed at thermoneutrality for 2–3 week and fed a standard diet. (**a**) Panels show representative images of Coomassie blue stained BN-PAGE gels. (**b**) Panels show representative images of histochemical staining of complex I (**I**), complex IV (**IV**), and complex II (**II**) in-gel activity. (**c**) Densitomentric quantification of band corresponding to in gel-activity of complex I, complex IV, and complex II. Protein extracts were prepared for each animal, and each individual was separately assessed. Data were normalized to the value obtained for WT animals, set as 100, and separately presented for each genotype (means ± SE; *n* = 3).

#### *3.2. E*ff*ect of UCP3 Ablation on Body Weight, Energy E*ffi*ciency, and Body Composition in Mice Fed Either with a Standard Low Fat Diet or a High Fat Diet*

At weaning, representing the experimental starting condition, KO-time 0 mice tended to have a lower body weight when compared to WT-time 0. However, the two groups showed similar body composition in terms of water, lipid, and protein percentage, as well as in energy content, referring to each gram of animal (Table 1).

**Table 1.** Body weight, body composition and energy content of each gram of animal weight detected in wild type (WT) and knockout mice (KO) mice at weaning (representing the beginning of the dietary treatment, i.e., time 0 of the experimental procedure).


Values represent mean ± SE of 5 different animals.

WT and KO mice housed at thermoneutrality and fed a standard/low fat diet (STD) for 80 days since weaning (named WT-STD and KO-STD, respectively) gained the same body weight, while consuming the same amount of energy by food (Table 2), thus no differences were observed in the rough energy efficiency between WT-STD and KO-STD (Table 2).


**Table 2.** Body weight, body weight gain, food intake, body rough energy efficiency and visceral WAT weight detected in WT and KO mice, housed at thermoneutrality, and fed either a standard low fat diet (STD) or a high fat diet (HFD) for 80 days, since weaning.

Values represent mean ± SE of 7–8 different animals. Differences were evaluated for statistical significance by two-way ANOVA followed by a Tukey's post hoc multiple comparison test. Different letters indicate that differences between mean values are statistically significant, with *p* < 0.05. Two-ways ANOVA test results: Body weight gain: diet effect *p* < 0.0001, genotype effect ns interaction ns; body rough energy efficiency: diet effect ns, genotype effect ns. interaction *p* < 0.0089; Visceral WAT: diet effect *p* < 0.0001, genotype effect *p* = 0.0094, interaction ns; WAT weight/body weight \* 100: diet effect *p* < 0.0001, genotype effect *p* = 0.0223, interaction ns.

As expected, mice that were fed a HFD for 80 days since weaning (named WT-HFD and KO-HFD) gained more body weight when compared to mice under STD feeding, independent of the genotype. Body weight gain tended to be higher in KO-HFD mice than WT-HFD mice, although the food that was consumed was similar between the two groups, thus indicating that WT-HFD-mice have lower body rough energy efficiency than KO-HFD ones (Table 2). Indeed, concerning the last parameter, a two-way ANOVA test indicated the effect of the genotype and that of the diet was not statistically significant, while a significant interaction between genotype and diet was observed (see Table 2 legend). Remarkably, in both groups of mice, body rough energy efficiency changed during the treatment with either STD or HFD. Indeed, it was maximal during the first 25 days and then progressively declined (Figure 4). A difference in rough body energy efficiency between WT-HFD and KO-HFD mice was principally observed at the end of treatment, and, also in this case, a two-way ANOVA test indicated a significant interaction effect between genotype and diet (Figure 4 legend).

**Figure 4.** Variation in body rough metabolic energy efficiency during the 80 days of treatment of WT and KO mice with either STD or HFD. Values represent the mean ± SE of 7–8 different animals. The differences were evaluated for statistical significance by two-way ANOVA test, followed by a Tukey's post hoc multiple comparison test. Different letters indicate that differences between mean values are statistically significant, with *p* < 0.05. Two way ANOVA test results: diet effect *p* = 0.0162, genotype effect *p* = 0.0037, interaction *p* = 0.0081.

We then detected body energy, lipid, and protein gains, as well as the efficiency of energy, lipid, and protein deposition. Concerning energy intake (Figure 5a) and whole body energy gain (Figure 5b), no significant differences were observed between WT-STD and KO-STD. HFD feeding induced a significant increase in mouse energy gain, and, despite KO-HFD tending to have higher body energy gains when compared to WT-HFD, statistical analysis revealed a significant effect of the diet, while the genotype effect and genotype/diet interaction were not significant. When looking at the lipid gain (Figure 5b), similar values were detected in WT-STD and KO-STD. As expected, the HFD regimen induced a significant increase in mice lipid gain, which was significantly higher in KO-HFD when compared to WT-HFD. Indeed, the two-way ANOVA test revealed the significant effects of diet, genotype, as well as the interaction between genotype and diet (Figure 5 legend). No differences between WT-STD and KO-STD were observed regarding protein gain (Figure 5b). The HFD regimen significantly enhanced protein gain, and similar values were detected in WT-HFD and KO-HFD mice (Figure 5b). In fact, the two-way ANOVA test revealed the significant effect of the diet but not of genotype or genotype/diet interactions (Figure 5 legend).

**Figure 5.** (**a**) Total energy, lipid and protein consumed by diet (**b**) Body energy, lipid and protein gains (**c**) Body composition (**d**) Efficiency of energy, lipid, and protein deposition of WT and KO mice, housed at thermoneutrality, and fed with either a standard/low fat diet (STD) or a high fat diet (HFD) for 80 days since weaning. Body composition was detected in terms of water, lipids, and proteins percentage. Values represent mean ± SE of 5–6 different animals. Differences were evaluated for statistical significance by two-way ANOVA test followed by a Tukey's post hoc multiple comparison test. Different letters indicate that differences between mean values are statistically significant, with *p* < 0.05. Two way ANOVA test results: - Panel a, for each parameter considered: diet effect *p* < 0.0001, genotype effect ns, interaction ns. - Panel b, energy gain: diet effect *p* < 0.0001, genotype effect ns interaction ns -lipid gain: diet effect *p* < 0.0001, genotype effect *p* = 0.0074, interaction *p* < 0.0001 -protein gain: diet effect *p* < 0.0001, genotype effect ns, interaction ns, Panel c—water: diet effect *p* < 0.0001, genotype effect ns interaction ns -lipids: diet effect *p* < 0.0001, genotype effect *p* = 0.0040, interaction *p* < 0.0032 - protein gain: diet effect *p* = 0.0154, genotype effect ns, interaction ns (*p* = 0.0516), Panel d–Efficiency of energy deposition: diet effect *p* < 0.0001, genotype effect ns (*p* = 0.0913), interaction *p* = 0.013 - Efficiency of lipid deposition: diet effect *p* < 0.0001, genotype effect ns, interaction *p* = 0.0118 - Efficiency of protein deposition: diet effect *p* = 0.0016 genotype effect ns, interaction ns.

In mice that were fed a STD, the efficiency of the energy deposition (evaluated as percent of energy consumed by diet and stored in the body) was similar in WT and KO mice (Figure 5d). The HFD regimen induced a significant increase in the efficiency of the energy deposition, which was significantly higher in KO-HFD as compared to WT-HFD. Indeed, the two-way ANOVA test revealed the significant effects of diet and the interaction between genotype and diet (Figure 5 legend).

The percentage of energy that was consumed by diet as lipids that was stored in the body did not differ between WT-STD and KO-STD (Figure 5d). HFD feeding significantly reduced such a parameter and, despite KO-HFD mice tending to have higher lipid deposition efficiency than WT-HFD (+40%), differences between these two groups did not reach statistical significance. Concerning the efficiency of protein deposition, no difference was observed between WT-STD and KO-STD. The HFD regimen enhances the efficiency of protein deposition and similar values were detected in WT-HFD and KO-HFD (Figure 5d). When considering either lipids or protein deposition (Figure 5d), two-way ANOVA tests revealed a significant effect of the diet, while neither the effect of genotype or the interaction between genotype and diet were significant (Figure 5 legend).

Consistent with what is described above, after 80 days of STD no differences in mouse body composition (evaluated in terms of content of water, lipids and protein in the carcass and expressed in percentage terms) were detected between WT-STD and KO-STD (Figure 5c). As expected, HFD treatment significantly affected mouse body composition, and a reduction in water content was observed independent of the genotype. Regarding the body lipid percentage, HFD treatment significantly enhanced it, with KO-HFD showing values that were higher than WT-HFD. In this case, the two-way ANOVA test reported the significant effects of diet and genotype, as well as a significant interaction effect between diet and genotype (Figure 5 legend). Concerning the body protein percentage (Figure 5c), HFD treatment enhanced it in WT mice, but failed to affect it in KO mice. Indeed, in the WT-HFD mice the protein percentage was higher than in the WT-STD ones, while no differences between KO-HFD and KO-STD were detected. Two-way ANOVA tests revealed a significant diet effect, being that the genotype effect was not significant and the interaction between the diet and genotype effects were at the limit of significance (*p* = 0.0516).

#### *3.3. E*ff*ect of UCP3 Ablation on Visceral Adipose Tissue and Lipid Accumulation in Lean Tissue in Mice Fed Either with a Standard*/*Low Fat Diet or a High Fat Diet*

The contribution of metabolically active tissues (liver, heart, gastrocnemius skeletal muscle, brown adipose tissue) to the weight of mice was similar when comparing the WT-STD, KO-STD, WT-HFD, and KO-HFD mice (data not shown). Interestingly, the contribution of total visceral WAT (vWAT) to body weight was lower in KO-STD than in WT-STD mice (about −25%) (Table 2). As expected, in mice under HFD, the contribution of vWAT weight to whole body animal was significantly higher than in mice under STD. Indeed, in WT-HFD the contribution was 56% higher than that of WT-STD, while in KO-HFD it was 97% higher than that observed in KO-STD (Table 2). In this case, a two-way ANOVA test indicated the genotype effect and the diet one as significant, but the interaction effect between genotype and diet as not significant (Table 2 legend).

Variations in vWAT mass, which were detected between WT-STD and KO-STD, were not associated with change in adipocytes size (Figure 6a–c), while WAT basal lipolysis was almost doubled in KO mice as compared to WT (Figure 7). Histological analysis of lean tissues (liver and skeletal muscle) revealed the presence of some lipid droplets in liver from WT-STD exposed at thermoneutrality, in accordance with data present in literature. Interestingly, the same analysis indicated an ectopic accumulation of fat that is associated with the absence of UCP3 in liver and skeletal muscle (Figure 6a). Indeed, H&E staining of skeletal muscle and liver sections from KO-STD at 100× magnification showed many large intracellular lipid droplets (LD) as uncoloured circles. Numerous large (~2 μm) intramyocellular lipid droplets (IMLDs) were only present in the skeletal muscle of KO-STD mice as well as a massive lipid accumulation being shown in the cytoplasm of all the hepatocytes. Mice under HFD presented an accumulation of lipids in liver whatever the genotype, and lipid accumulation in skeletal muscle was more evident in the KO-HFD mice then in the WT-HFD ones.

**Figure 6.** Effect of UCP3 ablation on lipid accumulation in WT and KO mice housed at thermoneutrality and fed with either a standard/low fat diet (STD) or high fat diet (HFD) for 80 days, since weaning. (**a**) Representative histological analysis of visceral WAT, liver and gastrocnemius skeletal muscle isolated from WT and KO mice, housed at thermoneutrality since weaning and fed with a standard/low fat diet for 80 days. Insets: high magnification images (100x) to visualize intracellular LD as uncoloured circles. In liver, arrowheads indicated small LD whereas arrows showed very large LD. (**b**) Mean Surface Area of adipocytes (μm2), (**c**) frequency distribution for surface area of adipocyte Values represent mean ± SE of 3 different animals. Differences were evaluated for statistical significance by two-way ANOVA test followed by a Tukey's post hoc multiple comparison test. Different letters indicate that differences between mean values are statistically significant, with *p* < 0.05. Two-way ANOVA test results for adipocyte mean area: diet effect, genotype effect and interaction ns.

**Figure 7.** Effect of UCP3 ablation on visceral WAT lipolysis detected in WT and KO mice housed at thermoneutrality and fed with a standard/low fat diet (STD) for 80 days, since weaning. A glycerol release was detected on small pieces of epididymal WAT from WT-STD and KO-STD mice in the absence and in the presence of isoprotenerol. Values represent mean ± SE of 6 different animals. Statistical analyses were performed by two-tailed Student's *t*-test, \* *p* < 0.05 vs. WT.
