Highlights Housing at 30°C blunts several adaptations to exercise training in mice Exercise-sensitive protein induction is dampened at 30°C in skeletal muscle 30°C-housing blunts training-induced increase in insulin-stimulated glucose uptake 105 day and night in untrained mice, in contrast with previous reports 20,27 , with no effect of ET on resting 106 RER (Fig. 2C). Chronically housing mice at thermoneutrality increased habitual (+90%) activity 107 during the night in UT mice (Fig. 2D) underlining the importance of adequate acclimatization time 108 Raun et al. Housing temperature influences exercise training adaptations in mice 2019 page 8 of 34during temperature changes, as acute temperature did not change habitual activity (Suppl. Fig. 1g).
109Increased habitual activity was despite lower food intake compared to UT mice at ambient 110 temperature (Suppl. Fig. 1i). This contrasts with the nightly running volume that tended to remain 111 lower (p=0.064, -35%, Fig. 2E) in mice housed at 30°C compared to ambient temperature, while 112 running volume during daytime was increased at 30 o C (Fig. 2F). Overall mice housed at 30°C ran 113 35% less than when housed at 22°C (p=0.06, Fig. 2G) accompanied by a tendency to a lower maximal 114 (p=0.054, Fig. 2H) and decreased average ( Fig. 2I) running speed. To test if reduced running volume 115 could be due to over-heating during exercise, we measured core temperature during the day (when 116 the mice are resting/inactive) and in the early dark period (when the mice are running the most). UT 117 mice displayed 0.6°C lower core temperature during the inactive period when housed at 22 o C 118compared with 30 o C, highlighting the mild cold stress inflicted by 22°C housing (Fig. 2J). Core 119 temperature increased to the same absolute values in all groups during the dark period (Fig. 2J). Thus, 120 reduced running of mice housed at 30°C is likely not due to overheating. 121 Overall, mice in 30°C displayed lower energy consumption, lesser improvements in body 122 composition, as well as no improvement in glucose tolerance following the standard laboratory 123 exercise model of voluntary wheel running, despite similar improvements in running performance. 124 Adaptations in glucose tolerance, but not performance or body composition, are ascribed to 125 training volume. 126 Having established remarkable differences in exercise training adaptations with this model in mice 127 housed at different temperatures, we next sought to determine if these alterations were due to the 128 lower running volume in 30 o C housed mice. Thus, we restricted voluntary running in mice housed at 129 22°C (Fig. 3A) to mimic the training volume of thermoneutrally housed mice ("paired 22°C ET", Fig. 130 3B).131 Raun et al. Housing temperature influences exercise training adaptations in mice 2019 page 9 of 34Body weight (Fig. 3C) and food intake (Fig. 3D) were unaffected by paired 22°C ET. Similar to 22°C 132 housing ( Fig. 1D), paired 22°C ET reduced fat mass gain (Fig. 3E). As such, mice trai...