Muscle wasting affects millions of people around the world, including the elderly, people with illnesses, and people who are immobilised for long periods of time. The loss of muscle mass leads to a decline in independence, promotes disease, increases resistance to treatment, and is associated with increased mortality. As a result, muscle wasting is a major public health problem. Many biomolecular mechanisms have been documented to explain the occurrence of muscle wasting, mainly through the use of laboratory rodent models. However, no treatment is really effective and/or adaptable for all today. The main objective of this thesis was to find new underlying mechanisms that could become therapeutic targets to combat muscle atrophy in humans. We chose an approach based on biomimicry. Our strategy was (1) to perform a comparative physiology study between the brown bear model naturally resistant to atrophy during hibernation and the unloading mouse sensitive to muscle atrophy, (2) to study the role of the ATF4 and TGF-β/BMP signalling pathways in these two models, and finally (3) to initiate studies on human muscle cells to validate the hypotheses from the first two studies. In our first study, the strategy was to identify genes differentially regulated in brown bear muscle between the hibernation and active periods. Then we compared them to those differentially regulated in the muscles of the unloading mouse versus the control mouse. We showed that the concomitance of inhibition of TGF-β signalling and induction of BMP signalling appeared to be crucial for the maintenance of muscle mass under conditions of prolonged physical inactivity. In our second study, we showed that the induction of the ATF4 signalling pathway in muscle was uncoupled from muscle atrophy in healthy and physically inactive mice when previously treated with the halofuginone molecule, and also in hibernating bears. In all three situations, the maintenance of muscle mass was associated with both the induction of ATF4 and BMP signalling and the inhibition of TGF-β. Finally, preliminary results obtained by cultivating human muscle cells with hibernating brown bear serum suggest the presence of a circulating active compound that may mimic some of the characteristics observed in atrophy-resistant hibernating brown bear muscle. In conclusion, this work provides numerous perspectives in the modulation of the balance of TGF-β and BMP signalling pathways in situations of prolonged physical inactivity. In addition, it opens up new research on the identification of active compounds in bear serum that could be used in the human clinic to limit or prevent the onset of muscle atrophy during immobilisation or in other pathophysiological conditions.