Smaller, Hungrier Mice

See allHide authors and affiliations

Science  17 Mar 2006:
Vol. 311, Issue 5767, pp. 1553-1554
DOI: 10.1126/science.311.5767.1553

In their Brevia “Increase in activity during calorie restriction requires Sirt1” (9 Dec. 2005, p. 1641), D. Chen et al. report that elevation in physical activity in response to 40% food restriction is absent in mice lacking the longevity protein Sirt 1, in spite of their motor performances being preserved or even increased as assessed by the treadmill and rotarod tests. Since Sirt1 is the closest mammalian homolog of Sir 2.1, which is necessary for life-span extension by calorie restriction (CR) in lower organisms (1), this intriguing phenotype provides the first evidence that sirtuins may have conserved roles in mediating physiological effects of food restriction (including perhaps extended longevity) also in mammals.

To make the point that the behavioral response to restricted feeding, rather than the metabolic consequences of the dietary regimen per se, is modified by lack of Sirt1, the authors provide evidence for “equal food intake” (grams of chow/day) in the two strains fed ad libitum (AL) and for comparable plasma levels of glucose, triglycerides, and IGF-1 in both AL and CR conditions. The authors also claim that body weights were reduced in wildtype (WT) and knockout (KO) mice and that “inability [of KO mice] to respond to calorie restriction was also not due to them experiencing a lower degree of food restriction.”

However, in the figure legend and the Supporting Online Material, the body weight reduction is from 37.8 ± 3.8 g (AL) to 19.9 ± 1.9 g (CR) in wild-type animals and from 18.3 ± 2.7 g (AL) to 15.4 ± 1.4 g (CR) in KO mice. It is not indicated whether body weights in the two strains fed AL diverged so dramatically during the 9-month treatment or were different since the beginning.

In fact, Sirt1 KO mice are smaller than their littermates (2). Independent of the effect of the mutation, smaller mice are expected to have lower energy requirements, especially if (in CR) they have less spontaneous activity. In this perspective, the data on food intake are rather surprising, because mutant mice eat almost as much as double-sized WT mice, and their food intake is much higher when normalized for body weight. Moreover, either loss of weight in food-restricted KO mice is very limited, or they don't gain enough weight during AL feeding.

Although it is not clear why smaller mutant mice eat ad libitum as much as much bigger wild-type animals (maybe they are hypermetabolic?), 40% food restriction may still leave the smaller strain with enough energy support to prevent at least some aspects of the CR response, including the “hungry” hyperactive behavior. IGF-1 function may be altered in Sirt1 KO mice (as also suggested by their reduced size) (3), and its plasma level may not be a reliable marker of negative energy balance.


  1. 1.
  2. 2.
  3. 3.


We showed that calorie restriction (CR) promoted a large increase in physical activity in wild-type but not in Sirt1 knockout (KO) mice, demonstrating the first example in which a mammalian sirtuin is required for any phenotype induced by this dietary regimen. Pani et al. suggest that our KO mice might not have been as calorie restricted and energy limited as the wildtype (WT) mice. However, both wild-type and KO mice showed comparable physiological changes in response to CR, namely similar reductions in blood glucose, triglycerides, and IGF-1. As we stated, “body weights were also reduced by calorie restriction” in both strains, but as Pani et al. correctly point out, the KO mice did not lose as high a percentage of their body weight on the regimen.

Food intake for both wild-type and KO mice was restricted to 60% of their ad libitum food-consumption levels, which turned out to be very similar despite the smaller size of the KO mice. The KO mice showed less weight loss on the CR diet most likely because they were smaller and leaner than wild type, even though they ate comparable levels of food ad libitum. Previous studies establish the precedent that leaner mice can exhibit less weight loss on a CR diet (1). However, we concluded then and now that the KO mice were as calorie restricted as the wild type, first, because they were food-restricted to the same degree of their ad libitum levels as wild type and second, because they underwent similar changes in the physiological parameters mentioned above.

Although a comprehensive study of the Sirt1 KO mice is clearly important, we believe that the data in our Brevia do indicate important defects in the ability of the Sirt1 KO mice to respond normally to CR.


  1. 1.

Stay Connected to Science

Navigate This Article