The influence of bacterial diet on fat storage in C. elegans

Abstract

Background: The nematode Caenorhabditis elegans has emerged as an important model for studies of the regulation of fat storage. C. elegans feed on bacteria, and various strains of E. coli are commonly used in research settings. However, it is not known whether particular bacterial diets affect fat storage and metabolism.

Methodology/principal findings: Fat staining of fixed nematodes, as well as biochemical analysis of lipid classes, revealed considerable differences in fat stores in C. elegans growing on four different E. coli strains. Fatty acid composition and carbohydrate levels differ in the E. coli strains examined in these studies, however these nutrient differences did not appear to have a causative effect on fat storage levels in worms. Analysis of C. elegans strains carrying mutations disrupting neuroendocrine and other fat-regulatory pathways demonstrated that the intensity of Nile Red staining of live worms does not correlate well with biochemical methods of fat quantification. Several neuroendocrine pathway mutants and eating defective mutants show higher or lower fat storage levels than wild type, however, these mutants still show differences in fat stores when grown on different bacterial strains. Of all the mutants tested, only pept-1 mutants, which lack a functional intestinal peptide transporter, fail to show differential fat stores. Furthermore, fatty acid analysis of triacylglycerol stores reveals an inverse correlation between total fat stores and the levels of 15-methylpalmitic acid, derived from leucine catabolism.

Conclusions: These studies demonstrate that nutritional cues perceived in the intestine regulate fat storage levels independently of neuroendocrine cues. The involvement of peptide transport and the accumulation of a fatty acid product derived from an amino acid suggest that specific peptides or amino acids may provide nutritional signals regulating fat metabolism and fat storage levels.

Figure 1. Dietary E. coli influences fat storage in C. elegans.

(A) Ancestral relationship between four E. coli strains used in this study. OP50 and DA837 are derived from E. coli strain B, HB101 is a B x K12 hybrid, and HT115(DE3) is derived from E. coli K-12. (B) Fat stores in C. elegans depend on the dietary E. coli strain. Young adults were fixed with paraformaldehyde and stained with Nile Red. (C) Lipids were extracted from young adult C. elegans grown on four E. coli strains. Total protein was determined using an aliquot of the nematode pellet used for lipid extraction. Lipid extracts were separated into phospholipids and TAGs using thin layer chromatography and quantified using gas chromatography. Error bars are SEM, n = 3−5 independent growths. The fatty acids in phospholipid fractions did not vary but the relative amount of fatty acids found in TAG fractions varied up to 2 fold depending on the dietary E. coli strain.

Figure 2. Comparison of live Nile Red staining, fixed Nile Red staining, and triacylglycerol stores in wild type and mutants.

Triacylglycerol (TAG) stores were determined by TLC/GC of lipid extracts. %TAG refers to the percentage of total fatty acid detected in the TAG fraction. Anterior is to the left. Nile Red fed to live worms accumulates in gut granules, which are lysosome-related organelles. After worms are fixed in paraformaldehyde, the intensity and size of Nile Red staining-droplets correlates well with the biochemical determinations of TAG stores.

Figure 3. Characterization of cell number, dry weight, and macronutrient composition of E. coli lawns.

Error bars are SEM, n = 3−5 independent growths. (A) Average number of viable bacterial cells in 3-day old lawns of E. coli washed off of 6 cm NGM plates. (B) Protein levels (normalized per cell) in four strains of E. coli. (C) Fatty acid levels (normalized per cell) in four strains of E. coli. (D) Average dry weight of bacterial lawns (normalized per cell) in four strains of E. coli. (E) The amounts of carbohydrate/cell vary significantly in the four E. coli strains. Total sugars in hydrolyzed E. coli lawns were determined by the Anthrone method. Error bars are SEM, n = 3−5 independent growths. (F) Addition of 5% glucose to plates leads to increased carbohydrate accumulation in bacteria and changes the morphology of bacterial lawns. (G) Growing bacteria on high glucose media increases carbohydrate levels in bacteria but does not cause decreased triacylglycerol stores in C. elegans. Error bars are standard deviation, n = 2−5 independent growths.

Figure 4. Fatty acid composition differences in E. coli strains and C. elegans.

(A) The relative proportions of saturated, monounsaturated (MUFA), and cyclopropane fatty acids in four E. coli strains. E. coli lawns were washed off of NGM plates. Pelleted bacteria were derivatized to produce fatty acid methyl esters (FAMEs) for gas chromatography analysis. Error bars are standard deviation, n = 4−5 independent growths. (B) The relative proportion of saturated (SAT), monomethyl branched chain (MMBA), monounsaturated (MUFA), and polyunsaturated (PUFA) fatty acids in wild-type C. elegans raised on four E. coli strains. Error bars are standard deviation, n = 4−5 independent growths.

Figure 5. Relationship between branched chain fatty acid C17iso levels in triacylglycerol stores and total fat stores.

(A) The percent of total fatty acids in triacylglycerol (TAG) fractions of wild-type young adults C. elegans feeding on four E. coli strains. Error bars are SEM, n = 3−5 independently grown samples. (B) The % of C17iso in triacylglycerol (TAG) fractions of wild-type young adults C. elegans feeding on four E. coli strains. Error bars are SEM, n = 3−5 independently grown samples. (C) The % of total fatty acids in TAG fractions measured in young adult wild type (WT) and various mutants. Although fat storage in many strains is greater or less than wild type, fat stores in most strains were reduced when grown on E. coli HB101 vs. OP50. Only pept-1 mutants, defective in intestinal peptide transport, show no significant difference in fat stores when grown on OP50 and HB101. Error bars are standard deviation, n = 2−4 independently grown samples. (D) The % of C17iso in TAG fractions in various mutant C. elegans grown on OP50 and HB101. Error bars are standard deviation, n = 2−4 independently grown samples. (E) Inverse correlation between %C17iso in TAG and % of total fatty acids in TAG. Data points shown are 3–4 independent growths of wild-type worms on OP50, HB101, DA837, HT115 in addition to pept-1 grown on OP50 and HB101. (F) Brood size is reduced in pept-1 animals growing on HB101 compared to OP50, while brood size does not depend on dietary E. coli in wild type, daf-2 or eat-2 mutants. Error bars are standard deviation, n = 15 individuals of each genotype on each food source.