We have shown that the
daf-12 and
daf-2 genes display allele-specific interactions in determination of adult life span. Mutant
daf-12 animals have a wild-type life span, mutant
daf-2 animals have a doubled life span relative to wild-type (Kenyon et al., Nature 366, 461,1993) and the life span of certain
daf-2 ;
daf-12 strains is quadrupled. This is the largest increase in life span observed to date. The same two genes interact to control larval development and to determine adult life span. In addition, these data support the mechanism of antagonistic pleiotropy in the evolution of aging, which is that a gene with an early benefit may be detrimental late and thereby limit life. The
daf-2 mutants that we used are temperature-sensitive dauer-constitutives. Hence, the brood size and the adult life span was tested at the permissive temperature (15 C). The
daf-2 (
e1370)reproductive defect is temperature-sensitive. Neither the dauer formation defect nor the reproductive defect is temperature-sensitive for the
daf-12 gene. The timing of progeny production is similar for all strains tested, including those with increased life spans. The life spans of four independent
daf-12 alleles is like that of wild type at 15 C. The
daf-2 (
e1370)is somewhat temperature sensitive for life span, and remarkably, the life span of
daf-2 (
m41)is like that of wild type. The
daf-2 (
m41);
daf-12 mutants have only a very slightly increased life span at 15 C and are therefore also temperature-sensitive. Interestingly, the super-enhanced life span observed in the
daf-2 (
e1370);
daf-12 strains at 25 C are not apparent at 15 C. This reinforces the idea that
daf-2 is the major effector of the life span extension of mutant adults, and that the
daf-12 mutations modulate mutant
daf-2 activity. Thus, the mutant phenotypes of dauer formation, life span and fertility defects are all temperature-sensitive in the
daf-2 strains that we have studied. We have conducted temperature-shift experiments to determine when
daf-2 function is necessary for wild-type life span. These experiments used the
daf-2 (
m41);
daf-12 (
m20)strain because it grows to adulthood at 25 C, and thereby permits temperature-shifts in both directions. A simultaneous wild-type control allows unambiguous subtraction of effects solely due to the changing temperatures. The life spans were determined for eight treatments of wild type and
daf-2 (
m41);
daf-12 (
m20).For four of them the animals were raised at 25 C and either left there continuously, or shifted to 15 C during the L4 stage, young adult stage, or after cessation of egg laying. For the other four treatments the animals were raised at 15 C and then subjected to shifts to 25 C timed as above for the downshifts. Even post-reproductive loss of
daf-2 activity will increase life span in a reasonable fraction of the animals. This observation suggests that life span extension in mutants is not due to a developmental or structural difference. Instead, the mechanism is pliable, perhaps involving an altered physiology under the control of
daf-2 .The
daf-12 gene, which encodes a transcriptional regulatory protein, can substantially modify this extension. Due to the parallels between dauer formation and increased life span of
daf-2 mutants we hypothesize that the adult life span extension is due to heterochronic expression of a dauer subprogram.