Diet-induced developmental acceleration independent of TOR and insulin in C. elegans

Abstract

Dietary composition has major effects on physiology. Here, we show that developmental rate, reproduction, and lifespan are altered in C. elegans fed Comamonas DA1877 relative to those fed a standard E. coli OP50 diet. We identify a set of genes that change in expression in response to this diet and use the promoter of one of these (acdh-1) as a dietary sensor. Remarkably, the effects on transcription and development occur even when Comamonas DA1877 is diluted with another diet, suggesting that Comamonas DA1877 generates a signal that is sensed by the nematode. Surprisingly, the developmental effect is independent from TOR and insulin signaling. Rather, Comamonas DA1877 affects cyclic gene expression during molting, likely through the nuclear hormone receptor NHR-23. Altogether, our findings indicate that different bacteria elicit various responses via distinct mechanisms, which has implications for diseases such as obesity and the interactions between the human microbiome and intestinal cells.

Figure 1. A Comamonas DA1877 Diet Affects C. elegans Life-History Traits

(A) Developmental progression on three different diets. Synchronized N2 wild type animals (L1 stage) were grown on three different diets as indicated on the x-axis, and scored after 43 hours. Larval stage was visually determined based on the stage of vulval development (see Figure S1). (B) Developmental progression of animals at 48 hours post-L1 synchronization, OP50 to DA1877 indicates animals switched from E. coli OP50 food to Comamonas DA1877 at 24 hours. Development on Comamonas DA1877 is shown for comparison. (C) Expression of GFP in a synchronized population of Pmlt-10::GFP-pest animals fed E. coli OP50 or Comamonas DA1877 throughout development. (D) Brood size on three different diets. Wild type N2 animals were grown on three different diets indicated on the x-axis. Bars represent the average total number of progeny per animal, with the standard deviation indicated for all animals combined. (E) Post-developmental lifespan of adult animals fed each of the indicated diets. OP50 – E. coli OP50; HT115 – E. coli HT115; DA1877 – Comamonas DA1877. (F) Brood size of animals on killed E. coli OP50 or killed Comamonas DA1877. Bars represent the average total number of progeny per animal, with the standard deviation indicated for all animals combined. (G) Developmental progression of animals grown on live or killed E. coli OP50 or Comamonas 1877 at 45 hours. (H) Post-developmental survival of adult animals fed each of the indicated diets.

Figure 2. Diet-Induced Changes in Gene Expression

(A) Scatterplot indicating changes in gene expression in animals fed E. coli HT115 or Comamonas DA1877 relative to animals fed E. coli OP50 detected by microarray expression profiling. Each square indicates a gene. See also Table S1. (B) Venn diagram indicating the total number of genes changing in response to diet on E. coli HT115 and Comamonas DA1877 diets and the overlap between these, relative to standard laboratory diet of E. coli OP50. (C) Gene Ontology analysis of genes that decrease in expression on a Comamonas DA1877 diet. Numbers in parentheses indicate enrichment score. See also Table S1. (D) Venn diagrams indicating a comparison of Comamonas DA1877-responsive genes at two different stages of development. (E) Categorization of core dietary response genes. See also Table S2.

Figure 3. A Dietary Sensor in Living Animals

(A) Pacdh-1::GFP transgenic animals respond transcriptionally to the different bacterial diets and can be used as a dietary sensor in living animals. The Comamonas DA1877 fluorescence image was acquired with a 25-fold longer exposure time (exp x25). See also Figure S2. (B) Eggs display the maternal GFP expression. (C) The addition of 5 mM glucose activates the dietary sensor. (D) High levels of glucose can activate the dietary sensor on Comamonas DA1877. (E) The dietary sensor is repressed upon starvation.

Figure 4. Comamonas DA1877 Dietary Effect can be Diluted

(A) nCounter analysis of gene expression of C. elegans fed different bacterial diets. gst-4 and pqm-1 are induced by oxidative stress. cpt-3 and acs-11 are starvation responsive genes. (B) Mixing diets: numbers indicate proportion of Comamonas DA1877 to E. coli OP50. (C) Animals grown on diluted Comamonas DA1877 develop faster relative to animals grown on OP50. (D) Post-developmental survival of adult animals fed each of the indicated diets. OP50 – E. coli OP50; DA1877 – Comamonas DA1877; 1/200 and 1/1000 refers to the dilution of Comamonas DA1877 in E. coli OP50. All bacteria were seeded onto peptone-free plates to prevent bacterial growth. (E) Changes in gene expression of animals fed a diet of Comamonas DA1877 diluted in E. coli OP50 compared to Comamonas DA1877 alone diet. See also Figure S3.

Figure 5. Analysis of Known Genes and Pathways Indicates that the Comamonas DA1877 effect on Development is Independent of TOR and Insulin Signaling

(A) The dietary sensor is not affected by a loss-of-function mutation in daf-6. (B) nCounter analysis of diet-responsive genes in N2 wild type and daf-6(e1377) mutant animals. (C) daf-2(e1370) mutant animals respond to Comamonas DA1877 like wild type animals. (D) daf-16(mgDf50) mutant animals respond to Comamonas DA1877 like wild type animals. (E) Developmental progression of animals at 43 hours post-L1 synchronization of daf-16(mgDf50) and wild type (N2) animals grown on E. coli OP50 or Comamonas DA1877. (F) RNAi analysis of TOR pathway components. Animals harboring the dietary sensor were fed E. coli HT115 bacteria that express double stranded RNA for the indicated genes. See also Figure S4. (G) Developmental progression of animals at 48 hours post-L1 synchronization of rict-1(ft7) mutant animals fed the indicated diets. (H) Developmental progression of animals at 50 hours post-L1 synchronization of rsks-1(ok1255) mutant animals fed the indicated diets.

Figure 6. Two Nuclear Hormone Receptors Affect the Dietary Sensor

(A) nhr-10(tm4695) mutant animals have reduced acdh-1 promoter activity but still respond to Comamonas DA1877. (B) Developmental progression of animals at 45 hours post-L1 synchronization of nhr-10(tm4695) mutant animals fed the indicated diets. (C) Expression of GFP in a synchronized population of Pmlt-10::GFP-pest animals over time in animals fed E. coli OP50, E. coli HB101 or Comamonas DA1877. (D) Developmental rate in N2 and Pnhr-23::NHR-23::GFP animals that overexpress NHR-23 (OE). Stages are as in (B). (E) nhr-23 RNAi (diluted 1 in 20) greatly reduces acdh-1 promoter activity. (F) nhr-23 RNAi (diluted 1 in 20) reduces acdh-1 promoter activity in nhr-10(tm4695) mutant animals.

Figure 7

Model for Dietary Regulation of Developmental Rate