Figure 3. Fed and starved larvae differ in their reliance on the glyoxylate shunt and glycolysis.

(A) Starvation survival curves are plotted for wild-type (WT), daf-16, icl-1, and pck-1 mutants. Here and throughout, logistic regression is used to fit survival curves. See Supplementary file 1 for median survival and statistics. (B) Worm length after 48 hr of feeding on Escherichia coli OP50 on plates is plotted for WT, daf-16, icl-1, and pck-1 mutants. Mean and standard error of the mean of three biological replicates is shown. pck-1 and pyk-1 were significantly different than WT (p=0.001, p=0.002, unpaired t-test, n = 3). daf-16 or icl-1 were not different than WT (p=0.23, p=0.24, unpaired t-test, n = 3). **p<0.01 compared with WT. (C) Normalized dose–response curves for 2-deoxy-d-glucose in fed and starved larvae are plotted. (D) Normalized dose –response curves for 3-mercaptopicolinic acid (3 MPA) in fed and starved larvae are plotted. (C and D) Larvae were fed HB101 E. coli lysate in liquid culture and their length was measured after 48 hr (black) or they were starved and survival medians were determined from logistic regressions of survival curves (orange). These data are normalized within each replicate as a percentage of control (no drug) animals. The x-axis is log transformed.