Many biomarkers have been shown to be associated not only with chronological age, but also with functional measures of biological age. It is difficult to show in human populations though, whether variation in biological age is truly predictive of life expectancy, as it requires longitudinal studies over many years and possibly decades. We followed adult cohorts of 20 Drosophila Genetic Reference Panel (DGRP) strains chosen to represent the breadth of lifespan variation, obtaining estimates of lifespan, baseline mortality and rate of aging, and associated these parameters with age-specific functional traits including fecundity and climbing activity, and with age-specific targeted metabolomic profiles. We show that activity levels as well as metabolome-wide profiles are strongly associated with age, that numerous individual metabolites show a strong association with lifespan, and that the metabolome provides a biological clock that not only predicts sample age, but also future mortality rates and lifespan. This study, while relatively small in scope with 20 genotypes and 87 metabolites, establishes strong proof of principle for the fly as a powerful experimental model to test hypotheses about biomarkers and aging, and provides further evidence for the potential value of metabolomic profiles as biomarkers of aging.

Liquid chromatography-mass spectrometry (LC-MS) A C13-labeled internal standard was made by dissolving a metabolite yeast extract (Cambridge Isotopes Laboratory, ISO1) in 2.0 ml of water. 150 ul of this stock was mixed with 850 ul of 0.1% heptafluorobutyric acid (HFBA) to make the solubilization buffer. The dried metabolite samples were solubilized in 50 ul 0.1% of this solubilization buffer, and 3 ul was injected onto the LCMS system. Samples were analyzed on a Vantage triple quad mass spectrometer from Thermo Fisher using a Waters Nanoaquity HPLC system, and a Waters ACQUITY UPLC M-Class HSS T3 Column (100Å, 1.8 µm, 300 µm X 100 mm) that was operated at a constant 25°C. Solvent A was 0.02% HFBA and 0.1% acetic acid in water. Solvent B was 0.02% HFBA and 0.1% acetic acid in acetonitrile. The gradient was 0-25% B in 15 min, followed by 25-100% B in 5 min. The flow rate was 5 ul/min. The precursor and product ion m/z values, their elemental compositions, collision energies, retention times, and ion adducts are listed in Supplemental Table 6. An amino acid standard mix was acquired between each batch as a system suitability test for the LCMS system. Three additional controls were acquired at the beginning and end of the entire acquisition – a blank containing 0.1% HFBA, a blank containing only a heavy labeled yeast metabolite internal standard, and a mixture of the targeted molecules (IROA standards from Sigma-Aldrich). Chromatograms for each molecule were integrated in the program Skyline, which then created a CSV output file containing the chromatographic peak areas.

For metabolomic profiling, we collected cohorts of flies from each genotype at six time points (days 4, 10, 24, 45, 69 and 80). Samples of five females were collected into 1.5-mL Eppendorf tubes, and flash frozen in liquid nitrogen. About half of the strains did not survive to age 69 d, and only five strains survived to age 80 d. At each time point, when sample collection was complete, we selected five strains to have four biological replicates at all time points, five strains to have two biological replicates, and the remaining ten strains a single biological replicate, with a total of 182 samples for metabolomics. This design allowed us to assess the reproducibility of metabolomic data within an experiment that also included many ages and genotypes. To maximize our power to compare samples from different ages, we conditionally randomized metabolomics samples according to (Ogut et al., 2019), where each batch contained samples from 4-6 strains, and each batch contained one replicate of all ages for each strain. Biological replicates were distributed across separate batches in the same manner. Each frozen fly sample was homogenized in 200 µL H2O:PBS 9:1 in 2mL Eppendorf tubes in a Next Advanced Bullet Blender for 5 minutes. We then added 800 µL methanol to homogenized tissues, vortexed for 10 seconds, incubated at -20˚C for 30 minutes, and sonicated in an ice bath for 10 minutes. We then centrifuged the mixture at 14,000 rpm for 15 minutes at 4˚C, transferred 600 µL of supernatant to a new Eppendorf tube, and completely dried under vacuum at 30˚C for 2 hours.