A high-sugar diet induces more severe hyperglycemia than do other high-calorie diets. Canton-S animals were fed a control diet or one of three high-calorie diets from embryo hatch until the post-feeding wandering L3 stage. The amount of food eaten by larvae on control, high-sugar, high-fat and high-protein foods was used to estimate the relative kilocalorie intake using the calorie data found in supplementary material Table S1. (A) Mass of each food consumed per larva from late L2 instar to the onset of metamorphosis on all foods. Developmental stage is used instead of time to account for the longer feeding time in high-sugar-fed larvae. (B) Number of mouth hook contractions per second in L3 larvae raised on each food. (C) Calories consumed on control, high-sugar, high-fat and high-protein diets. (D) Hemolymph glucose concentrations on all four diets in wandering L3 larvae; n>32. (E) Hemolymph trehalose levels; n≥3. (F) Whole body glycogen levels; n≥5. Error bars are ± s.e.m. An unpaired two-tailed t-test was used to derive all P-values.

High-sugar-fed insulin-resistant Drosophila accumulate stored fat. Wandering L3 larvae raised on either control or high (1 M)-sucrose diets were analyzed for fat content. (A) Total TAG levels were assayed enzymatically in control and high-sugar-fed larvae, and normalized to weight; n≥11. (B) Nile Red staining of control and high-sugar larval fat bodies. (C) Quantification of lipid droplet area in two-dimensional confocal slices; n≥15. (D) Lipid droplet numbers in control and high-sugar-fed larval fat bodies; n≥15. (E) Free fatty acid (FFA) levels in control and high-sugar-fed larvae; n≥3. Error bars are ± s.e.m. An unpaired two-tailed t-test was used to derive P-values.

A high-sugar diet induces insulin resistance. Canton-S larvae were reared on control and high-calorie diets. (A) Wandering L3 weights; n≥15. (B) Developmental time course to pupariation on all four diets; n=3. (C) Mono- and disaccharide sugars similarly elicit developmental delay. Sugar concentrations were 0.15 M (low sugar) and 1.0 M (high sugar) for disaccharides sucrose and maltose, and 2.0 M (high sugar) for monosaccharides fructose and glucose; n=3. (D) Young adult brains harboring DILP2-GAL4>UAS-GFP are highlighted in a 3D reconstruction of confocal images; n>10 brains per condition. (E) Quantitative RT-PCR of DILPs in 2- to 4-day-old adult male heads after rearing on both diets. DILP2, DILP3 and DILP5 transcripts were measured using actin5c as a control. P=0.06 for DILP2, P=0.04 for DILP3 and P=0.05 for DILP5 between control and high sugar; n≥7. A two-tailed t-test was used to derive P-values. (F) Western blot of hemolymph DILP2-FLAG from control- and high (0.7 M)-sugar-reared DILP2-GAL4>UAS-DILP2-FLAG wandering L3 larvae. Equal amounts of hemolymph were loaded in each lane. Canton-S (CS) served as a negative control. (G) Quantitation of high-sugar and control D2-FLAG signal on n=4 western blots. (H) Dissected organs from L3 male larvae raised on control or high-sugar diets were treated with insulin (0.5 μM) or vehicle and visualized using antibodies against Drosophila PO₄-Akt, pan-Akt or syntaxin. (I) Bands from five western blot experiments were quantified, and PO₄-Akt was normalized to syntaxin as a loading control; n=5. A one-tailed t-test was used to derive P-values. All error bars are ± s.e.m.

Expression analysis reveals potential mechanisms of insulin resistance at the transcriptional level in high-sugar-fed Drosophila. RNA was isolated from male control and high-sugar-fed Canton-S L3 larvae. Affymetrix GeneChips (Drosophila 2.0) were used to analyze gene expression. (A) Gene expression that changed more than 1.3-fold upon high-sugar feeding. The intersection of 12-hour and chronically high-sugar-fed datasets was modest, but significant for both up- and down-regulated genes; P<0.001 by a Pearson’s χ² test; n=3 for each condition. (B) Gene expression changes that are consistent with known mechanisms of metabolic homeostasis or insulin resistance. Fold changes are only reported if statistically significant probe intensity differences exist; P<0.05 via unpaired two-tailed t-test. nc, no change.