We collected 60 age-dependent transcriptomes for C. elegans strains including four exceptionally long-lived mutants (mean adult lifespan extended 2.2- to 9.4-fold) and three examples of lifespan-increasing RNAi treatments. Principal Component Analysis (PCA) reveals aging as a transcriptomic drift along a single direction, consistent across the vastly diverse biological conditions and coinciding with the first principal component, a hallmark of the criticality of the underlying gene regulatory network. We therefore expected that the organism’s aging state could be characterized by a single number closely related to vitality deficit or biological age. The “aging trajectory”, i.e. the dependence of the biological age on chronological age, is then a universal stochastic function modulated by the network stiffness; a macroscopic parameter reflecting the network topology and associated with the rate of aging. To corroborate this view, we used publicly available datasets to define a transcriptomic biomarker of age and observed that the rescaling of age by lifespan simultaneously brings together aging trajectories of transcription and survival curves. In accordance with the theoretical prediction, the limiting mortality value at the plateau agrees closely with the mortality rate doubling exponent estimated at the cross-over age near the average lifespan. Finally, we used the transcriptomic signature of age to identify possible life-extending drug compounds and successfully tested a handful of the top-ranking molecules in C. elegans survival assays and achieved up to a +30% extension of mean lifespan.
The largest relative lifespan extension yet recorded has been in C. elegans, and corresponds to an almost 10-fold increase with the mg44nonsense mutation in the age-1 gene1,2. However, this hyperlongevity requires homozygosity of the mutation for two generations, resulting in total pre-embryonal genetic disruption. In human subjects, sensible anti-aging therapies would instead be applied in adulthood, ideally at advanced ages. Sadly, the best C. elegans models of therapeutic interventions yield significant, but considerably smaller reported increases in life-span by treatments begun even as early as embryonic development (e.g., up to roughly +90% by let-363 RNAi3). Late-life pharmacological interventions yielded even smaller effects on lifespan in flies4,5, nematodes6,7 and mice8,9,10. It is not fully understood why a single nonsense mutation can dramatically extend animal lifespan, while an RNAi block of the same gene does not produce a comparable effect, especially when administered later in life. In many cases, temperature-sensitive mutations extend lifespan without completely eliminating the biosynthesis of the gene product, so the difference is unlikely to be incomplete suppression of transcripts by RNAi. Neuronal resistance to RNAi is another likely explanation for the reduced impact of some RNAi constructs. Perhaps the mutation dramatically changes the molecular machinery of the whole organism during development such that the course of aging of the super-long-living strains is qualitatively different both regarding rates and form, and hence could not be easily reproduced therapeutically. Alternatively, perhaps the gene regulatory network is sufficiently robust that a therapy can reduce the pace of aging without qualitative alterations of the relevant molecular mechanisms.
To address these alternatives, we compiled an RNA-seq dataset of age-dependent transcriptomes produced from C. elegans isogenic strains and populations that have vastly different lifespans. Among them are three long-lived isogenic strains carrying mutations: daf-2(e1370), age-1(mg44) [at the first and second generations of homozygosity], and daf-2(e1391); daf-12(m20) double mutant1,11; three RNAi treatments (daf-4, che-3 and cyc-1); and two independent control runs represented by wild-type (Bristol-N2, DRM stock). The overall range of adult lifespans across all the experiments extends from 17 to 160 days. For each of the mutants or interventions, we measured gene-expression levels over time, across their lifespans, collecting 60 transcriptomes in total (9 different biological time-series, each in duplicate).
Principal Component Analysis (PCA) of gene expression reveals aging in all strains and treated groups as a transcriptomic drift in a single direction, consistent across the vastly diverse biological conditions and coinciding with the first principal component of the combined dataset, which is a hallmark of the criticality of the underlying regulatory network12. We therefore expected that the organism’s physiological aging state can be characterized by a single stochastic variable having the meaning of biological age and coinciding approximately with the first principal component score. The aging trajectory, i.e., the dependence of the biological age variable on chronological age, is then universally determined by the underlying regulatory interactions and the experimental conditions through a single phenomenological property describing the effective regulatory-network stiffness. The quantity imposes a natural time scale proportional to the mortality rate doubling time, the fundamental dynamic characteristic of the aging process12.
To identify a set of genes universally associated with aging across many different biological conditions, the aging signature, and to evaluate the theoretical model, we performed a meta-analysis of publicly available gene-expression measurements in C. elegans (more than 4000 samples in total). The identified aging signature comprises a set of genes, many of which have no known role in the regulation of aging or longevity. We used the same data to introduce a robust transcriptomic biomarker of aging, as a read-out or predictor of “biological age”, and demonstrated its utility across the datasets. The biological age dynamics in our experiments reveal a universal “aging trajectory”: the rescaling of age by lifespan simultaneously brings together the time-dependent trajectories of the transcriptomic biomarker on age and the survival curves. Throughout the paper, “age” means the chronological adult age (post-L4/adult molt for C. elegans). Therefore, the universality of aging trajectories may provide a natural molecular basis for the scaling universality of survival curves recently observed13 and independently confirmed in the survival data of all the strains and treatments in our experiments. We investigated the relationship between the stochastic evolution of the biological age variable and mortality using the survival data from an independent experiment. We also experimentally confirmed the model prediction of the equivalence between the mortality rate doubling exponent (inferred at the cross-over age, corresponding to the average lifespan) and the limiting mortality value (corresponding to the mortality plateau). Finally, we used the transcriptomic signature of age to identify possible life-extending drug compounds and successfully tested a handful of them in C. eleganssurvival assays.
