Amino-acid imbalance explains extension of lifespan by dietary restriction in Drosophila

Abstract

Dietary restriction extends healthy lifespan in diverse organisms and reduces fecundity. It is widely assumed to induce adaptive reallocation of nutrients from reproduction to somatic maintenance, aiding survival of food shortages in nature. If this were the case, long life under dietary restriction and high fecundity under full feeding would be mutually exclusive, through competition for the same limiting nutrients. Here we report a test of this idea in which we identified the nutrients producing the responses of lifespan and fecundity to dietary restriction in Drosophila. Adding essential amino acids to the dietary restriction condition increased fecundity and decreased lifespan, similar to the effects of full feeding, with other nutrients having little or no effect. However, methionine alone was necessary and sufficient to increase fecundity as much as did full feeding, but without reducing lifespan. Reallocation of nutrients therefore does not explain the responses to dietary restriction. Lifespan was decreased by the addition of amino acids, with an interaction between methionine and other essential amino acids having a key role. Hence, an imbalance in dietary amino acids away from the ratio optimal for reproduction shortens lifespan during full feeding and limits fecundity during dietary restriction. Reduced activity of the insulin/insulin-like growth factor signalling pathway extends lifespan in diverse organisms, and we find that it also protects against the shortening of lifespan with full feeding. In other organisms, including mammals, it may be possible to obtain the benefits to lifespan of dietary restriction without incurring a reduction in fecundity, through a suitable balance of nutrients in the diet.

Figure 1. Amino acids mediate lifespan and fecundity changes in fly DR

(a) Adding nutrients to DR revealed that amino acids limited fecundity and their addition rescued the level for fully-fed flies (DR+Amino Acids v Fully-fed, P = 0.5288; DR v other conditions, P>0.2). (b) Adding amino acids to DR food shortened lifespan (P<0.00001) to that of fully-fed flies (P=0.102). No other nutrient additions tested changed lifespan from the DR level (P>0.1 in all comparisons). Fecundity: mean ± s.e.m.; n=10; compared using the Wilcoxon rank sum test. Survivorship: starting n=100 per treatment; compared using the log rank test.

Figure 2. Essential amino acids cause the DR effect

(a) Adding essential amino acids (EAA), but not non-essential amino acids (N-EAA), increased fecundity to the level with all amino acids (All AA) and fully-fed. (DR+EAA v Fully-fed, P=0.393; DR+EAA v DR+All AA, P=1). (b) Adding EAA or All AA to DR caused lifespan to decrease to the same extent as fully-fed (P>0.102). In contrast, N-EAA addition to DR shortened lifespan much less (P=0.011). Egg laying: mean ± s.e.m.; n=10; compared using the Wilcoxon rank sum test. Survivorship: starting n=100 per treatment; compared using the log rank test.

Figure 3. Methionine is necessary and sufficient to increase DR egg laying

(a) Adding EAA except methionine (−met) did not increase fecundity from DR (P=0.796), however EAA without tryptophan (−trp) did to the level of DR+EAA or Fully-fed (DR v DR+EAA-trp, P=0.00893; DR+EAA-trp v DR+EAA, P=0.4359; DR+EAA-trp v Fully-fed, P=0.7394; DR+EAA-met-trp v DR+EAA-met, P=0.796). (b) Methionine alone increased fecundity to that of DR+EAA and fully-fed (DR+met v DR+EAA, P=0.5288; DR+met v fully-fed, P=0.9118). (c) Fecundity increased with methionine addition (significant ≥0.35mM, P=0.02323), reaching fully-fed at 0.7mM (P=0.393). Egg laying: mean ± s.e.m.; n=10; compared using the Wilcoxon rank sum test.

Figure 4. Amino acids, insulin signaling and DR

(a) Adding methionine to DR did not shorten lifespan (P=0.683) but EAA did (DR v DR+EAA, P=0.014; DR+EAA v Fully-fed, P=0.323). (b) Adding EAA without methionine (-met) rescued DR longevity (DR v DR+EAA-met, P=0.183) but tryptophan removal (-trp) did not (DR+EAA-trp v Fully-fed, P=0.115). (c) Reduced insulin signaling (daGal4->InRDN) extended DR lifespan (v controls, P<0.00001) and reduced flies’ response to AA or yeast addition (lifespan P>0.064 and fecundity P>0.3, DR v other foods; inset: survival curves for daGal4->InRDN). Controls responded similarly to wild-types. Lines connect median lifespans. Experimental setup and statistics as previously described.