What can we learn about aging from mutations that alter longevity? Rather than examining single biomarkers of aging to compare the effects of mutations, in today's open access paper the authors instead report on a comparison of the shift in all gene expression. The researchers measured the whole transcriptome across the life span of nematode worms, comparing worms with long-lived and short-lived gene variants, commenting on the results. The data supports the view that favorable gene variants literally slow aging: the same transcriptomic trajectory is observed in all worms, but stretched out over a longer period of time in the longer-lived individuals.
Is it actually helpful to examine the biochemistry underlying natural variations in life span? If the goal is to chip away at the massive undertaking of developing an complete understanding of how aging progresses at the detail level, then most likely yes. All data is useful. From the point of view of building meaningful therapies to treat aging, then most likely no. Researchers have a good map of important mechanisms in aging. The best way forward to is target those mechanisms, producing results that have no analogy in natural variations in aging. Looking at differences in how humans age will not inform us as to how good it is to entirely remove senescent cells, or entirely replace damaged mitochondria with fresh mitochondria - only actually building the therapies will answer the question of how good they are, how many years of healthy life might be added via their use.
In contrast to chronological age, which is measured simply by the ticking of a clock, defining physiological age is a much more complex task. To deal with this complexity, we use biomarkers of aging as proxy measurements to determine how physiologically young or old an individual is. However, this requires that we understand what "young" and "old" look like for a given biomarker. Using the average biomarker levels over chronological time to build a trajectory from a youthful to aged state, we can then place an individual on this trajectory and compare whether it is physiologically older or younger than its actual chronological age would suggest. This works because biomarkers of aging vary over time in each individual at a faster or slower rate depending on each individual's rate of aging. Here, we expanded this logic by building a trajectory of aging using the whole transcriptome and comparing the transcriptomes of subpopulations predicted to be long- or short-lived by the expression of four different biomarkers of aging. In doing so, we identified a class of genes which separate along this trajectory of physiological age and another which separates orthogonally to it.
That biomarkers of aging correlate with a common physiological age state is also consistent with results suggesting that various interventions which affect population longevity, such as long-lived mutants, "rescale" lifespan and healthspan relative to the wildtype population. Researchers have shown, for instance, showed that several interventions which lengthen or shorten lifespan rescale the hazard curve of the wildtype population, and more recently it was shown that several long-lived mutants have proportionally-scaled healthspan relative to wildtype controls. Other researchers performed a meta-analysis of several RNA-seq studies of long-lived mutants and similarly found that the transcriptomic age of these populations was scaled primarily along a single axis in a manner correlated with lifespan extension. Our results suggest a similar phenomenon occurring among untreated individuals of the same population, whereby long- and short-lived individuals undergo, in large part, temporally-scaled versions of the same transcriptional trajectory. While our interpretations are limited by sorting and sequencing populations only at one time-point, future work could confirm whether differently-fated individuals continue to follow this common trajectory by sequencing at later timepoints post-sorting.
While the finding that a difference in predicted lifespan largely resembles a difference in apparent age may be intuitive, the consistency of this signature across each biomarker tested is notable. One could imagine an alternate model in which each biomarker correlates with a specific age-related etiology, resulting in several different ways to be healthy or unhealthy - instead, we find the transcriptomic differences underlying high versus low expression of each biomarker tested to be remarkably similar. This result lends further support to previous findings that certain transcriptional biomarkers of aging, even when expressed in different tissues, appear to correlate with some common underlying state related to future lifespan.
View the full article at FightAging
