**4. Discussion**

In this study, we minimized thermal stress to better define the role of UCP3 in whole energy homeostasis and lipid utilization and we showed that in 4–5 month old mice acclimated for 2–3 weeks at thermoneutrality, the absence of UCP3 significantly depresses the whole animal RMR and REE, the energy intake, and the use of lipids as metabolic substrates. The last effect seems to be the result of the impaired ability of SkM mitochondria to oxidize fatty acid, which confirmed the results of previous studies employing mice that were housed at a standard temperature [11,12,34].

Interestingly, the absence of UCP3 selectively reduces the oxidation of lipid derived substrates, such as palmitoyl carnitine (State 3), while not affecting respiratory parameters (both State 4 and State 3) when pyruvate or succinate was used as substrate, i.e., a complex I- linked substrate and complex II-linked substrate, respectively. It should be considered that the control of State 4 respiration is shared between proton leak and reactions that are involved in the oxidation of substrates (among these respiratory chain). The control of State 3 respiration is shared between the reaction involved in the synthesis and export of ATP and the reaction involved in the oxidation of a substrate [35]. Thus, the absence of variations in State 4 respiration sugges<sup>t</sup> that UCP3 does not impact the respiration that is needed to balance the proton leak in basal condition. These data are in line with some studies showing no differences in basal proton-leak kinetics between WT and KO in skeletal muscle mitochondria from mice that were housed at a standard temperature [36,37] and from mice housed at thermoneutrality (our unpublished observation). They also support the concept that UCP3 could only mediate the proton leak in the presence of activators, plausibly ROS and FA [6,37]. At the same time, we did not observe any variations in State 3, in the presence of either pyruvate or succinate, between WT and KO SkM mitochondria. This finding suggests that the absence of UCP3 does not influence the activities of the reactions that are involved in the synthesis and export of ATP, and that are involved in the oxidation of the above substrates, among these the respiratory chain. The functional data, obtained in isolated mitochondria that are oxidizing pyruvate or succinate, agree with the in gel respiratory complexes activity. The reduced state 3, detected in KO mice mitochondria in the presence of palmitoyl-carnitine, suggests that it is plausibly due to the impaired activity of the reactions that are involved in the oxidation of such a substrate.

Data that were obtained in adult animals acclimated to thermoneutrality for two to three weeks sugges<sup>t</sup> that the absence of UCP3 could influence body lipid accumulation. We tested this hypothesis by chronically exposing mice to different lipid loads when housed at thermoneutrality since weaning.

The absence of UCP3 leads mice to be more susceptible to whole animal lipid accumulation only when they are subject to a HFD lipid overload. In fact, HFD-KO mice showed enhanced body energy efficiency and increased body lipid gain, which, by the end of the treatment, produced a different body composition with KO-HFD mice presenting a higher body lipid percentage and lower protein percentage when compared to HDF-WT mice. Interestingly, the evidence that KO-HFD and HFD-WT mice showed the same energy/lipid intake, but at the same time HFD-KO mice accumulated more lipid, further sustains that the absence of UCP3 reduced fat utilization and increased whole body fat storage. As a whole our data also indicates that lipids overload, as obtained by HFD administration, is important to shed light on the involvement of UCP3 in influencing body composition, since, when mice were fed with a standard/low fat diet, the absence of UCP3 does not influence the whole body lipid accumulation as well as body composition.

The ability of KO mice fed a HFD to accumulate more fat than WT mice was previously demonstrated by Costford et al. [15]. However, this accumulation was observed in mice that were housed at 23 ◦C that had received HFD for eight months, whereas the treatment was ine ffective after only four months of HFD [19]. Notably, our data indicates that, when housing mice under thermoneutral conditions, 80 days (~2.5 month) of HFD was enough to produce a higher lipid gain in KO mice when compared to wild type mice. This finding further supports the critical role of housing temperature for metabolic studies. Interestingly, despite the fact that, in young mice fed with standard/low fat diet, the absence of UCP3 does not change whole body lipid accumulation, redistribution of fat in the body does ge<sup>t</sup> a ffected, since it represses lipid storage in visceral WAT, while promoting accumulation in the liver and skeletal muscle.

Previous studies that were performed on KO mice housed at standard temperature and fed with low fat/standard diet failed to show an increase in intramuscular lipid accumulation. This was only observed when KO mice were fed with a HFD for eight months [15], thus further confirming the importance of the housing temperature for the outcome of this phenomenon in the absence of chronic lipids overload.

It should be considered that, in mice, the inability to adequately promote fatty acid utilization is associated with lipid accumulation in peripheral tissues and it contributes to the development of insulin resistance [38], a condition that is more evident when associated with enhanced WAT lipolysis. This "metabolic picture" is observed in KO mice that were acclimated to thermoneutrality, in which fatty acid oxidation is blunted, release of fatty acids by visceral adipose tissue' depots is enhanced, and the accumulation of lipid in lean tissues (liver and skeletal muscle) takes place. These data sugges<sup>t</sup> that: (i) an alteration in UCP3 activity can also a ffect the metabolism of tissues that do not express it, such as liver, presumably through changes in systemic metabolite tra fficking, and (ii) UCP3 exerts a protective role against lean tissue "lipotoxicity" and insulin resistance, by avoiding ectopic fat accumulation.

The above data are in agreemen<sup>t</sup> with our previous studies indicating that a progressive decline in insulin sensitivity in UCP3+/− heterozygous mice and UCP3−/− mice [11] and with clinical observations reporting that (i) a 50% reduction of UCP3 protein in human SkM is correlated with the incidence of T2DM [39], (ii) in humans, the UCP3 protein levels are reduced in the pre-diabetic state of impaired glucose tolerance [40,41], (iii) heterozygous mutations in the UCP3 gene (V56M, A111V, V192I, and Q252X) in children was associated with dyslipidemia and lower insulin sensitivity [42]. Nevertheless, some limitations exist in our study that are to be addressed in the future. These are:

