(A) Comparison of mean (+/−S.E.M.) cycle duration for raga-1(ok386), raga-1(GF) and wild-type N2. (Left) Mean cycle duration on day 1. Note the cycle lengths are similar for vigorous day 1 adults of all three genotypes. (Middle) Cycle durations on day 6 of adulthood; raga-1(GF) show substantial increases compared with N2 and raga-1(ok386) (*, p<.05, Dunn's test) (Right) Cycle durations on day 16, showing lengthening of wild-type cycles while raga-1 (ok386) still exhibits comparatively short cycles, though also increased from day 6. (B) Representative curvature matrices on day 1 (left column), day 6 (middle) and day 16 (right) for raga-1(GF) (top row), wild-type (middle) and raga-1(ok386) (bottom). As early as day 6, raga-1(GF) worms show slowing cycles; bends are also deeper than wild-type, suggesting qualitative changes in swimming occurred as well. By day 16, raga-1(ok386) worms swim substantially faster than wild-type (*, p<.05, Mann-Whitney U).

(A) Representative images of animals of each genotype to illustrate size differences. (B) Sizes of adult animals derived from measurement of length in video stills used for swimming analysis. Note differences in wild-type and raga-1(ok386) by day 1 of adulthood, suggesting different larval growth properties, which persist into adulthood even though animals continue to increase in size as adults. raga-1(GF) animals do not show any adult increase in size by day 6.

(A) Collage of swimming postures from one complete swim cycle. Every frame, or every seventh frame, is shown from a day 1 and day 16 animal respectively, from 30 frames per second movies. (B) Building of curvature matrix from a spine along the worm midline. Each angle along the worm is depicted using scaled blue or red coloring for curvature in the dorsal/ventral plane. The anterior is represented at the bottom of the matrix. (C) Representative curvature matrices from the same adult on day 1 of adulthood (top) and day 16 (bottom) show dramatic slowing of the swim cycle with relative preservation of waveform. (D) Mean (+/−S.D.) swim cycle duration in seconds (n = 18, 17, 16, 14, 12, 10, and 6 for days 1, 6, 8, 10, 12, 14, and 16 respectively). The increase is curvilinear, with rapid increases following a period of relative stability earlier in adulthood (*p<.05 versus day 1; Dunn's test). (Inset) Coefficient of variation for swim cycle duration on each day tested. Variability of cycle duration is fairly stable over the lifespan after correction for increases in the mean. (E) Plot of mean cycle duration on day 1 of adulthood versus lifespan for each individual in the cohort showing no correlation or predictive value of the cycle duration on day 1. Similar plots from later ages were not predictive of lifespan either (data not shown). Two worms' natural lifespan durations were not determined (i.e., were censored) due to accidental death. (F) Age-dependent incidence of left/right bends, out of the usual dorsal/ventral swimming plane (mean+/−S.E.M; *p<.05 versus day 1, Dunn's test; #, none were observed on day 1). For this experimental cohort, the mean lifespan was 16.0+/−0.9 days (mean+/−S.E.M.; n = 16 deaths observed).

Lifespan assay results from representative experiments (see Table S1 for quantitative data and results of all experiments). (A) Comparison of raga-1(ok386) and raga-1(ok701) with wild-type N2 survival at 20°C. Strains harboring alleles ok386 and ok701 consistently showed increases in lifespan (N = 7 and 4 trials, respectively). (B) Transgenic expression of raga-1 dominant negative (DN) increased, while raga-1 gain of function (GF) decreased, lifespan versus wild-type (N = 5 and 4, respectively). (C) Effect of CeTOR (let-363) RNAi knockdown during adulthood on lifespan of raga-1(GF) and raga-1(ok386) at 25°C. See Table S3 for quantitative results of RNAi experiments. (D) Transgenic extrachromosomal expression of wild-type raga-1 under its native promoter rescues lifespan to wild-type duration in the raga-1(ok386) background.

(A–E) Survival curves at 25°C under varying dietary conditions. The bacterial density of the food source is shown on each panel. The largest increase in lifespan seen for raga-1(ok386) versus N2 wild-type control is at the highest concentration of bacterial food. (F) Mean (+/−S.E.M.) lifespan plotted versus bacterial density. Wild-type N2 shows the expected parabolic response to dietary restriction, while raga-1(ok386) is much less sensitive.