Lower Protein Replacement Rates Observed in Various Methods Producing Enhanced Longevity in Mice
Today I'll point out an interesting open access paper in which the authors discuss what is probably an aspect of the observation that greater proteostasis correlates with greater longevity. Proteostasis is just a fancy way of saying the types, behaviors, and amounts of proteins produced by cells and present in tissues remain essentially the same over time, taking into account cyclic short term variations in response to repeated circumstances such as sleep, eating, and so on. As an individual ages, however, all these things change as the operation of cellular metabolism reacts to growing levels of damage. Since protein levels are the signals and controlling dials and switches of cells, ongoing age-related change in cellular behavior implies a lack of proteostasis and vice versa.
Given this, discussing proteostasis in connection with longevity and aging has always seemed a little tautological to me. It is a measure of consequences, another way of saying that aging has occurred. Aging is caused by damage, so of course there is a correlation between reactions to damage and longevity. However, it is probably the case that some measures of proteostasis will be useful as biomarkers of aging, in much the same way that some measures of epigenetic changes seem promising. A robust biomarker of aging would be a very useful tool indeed, as it could be used to rapidly test putative rejuvenation treatments, scoring their outcome on the whole of an individual's biology, rather than only their targets. For example, when senescent cell clearance therapies are deployed, determining their effects on senescent cell counts will be a part of the process, but at the moment is still necessary to wait around to see what the effects on long term health and life span will be. Even in rodents that takes years and millions of dollars. Replacing that cost with a simple test a month after treatment will greatly speed up the field.
In this paper the researchers look at the replacement rates of proteins, which is probably a function of at least the level of ongoing damage to proteins that would require them to be replaced, and rates of cellular replication, which have been observed to be lower in some forms of enhanced longevity in laboratory animals. They find that the former but not the latter correlates with life span:
Over the last 50 years, several dietary, genetic, and pharmacological interventions have been identified that extend maximum life span in laboratory animals, including mice. Understanding the molecular underpinnings of the aging process and the biochemical pathways affected by interventions that attenuate the development of age-related diseases is a high priority. In particular, identifying metrics of biological processes, or biomarkers (BMs), that are involved in the slowing of aging in laboratory mammals will be essential for guiding the future development of interventions to extend human healthspan.To identify such processes that play an etiologic role in age-related disease, our approach has been to test potential flux-based BMs of maximum life span extension. The concept underlying this strategy is that the activity of a metabolic process is best characterized by the molecular flux rate traversing the pathway and that changes in the flux rates of metabolic processes that play a causal role in the functional alterations of the condition may manifest earlier and more sensitively than static pathologic changes or complex clinical outcomes. Optimally, the functional role and measurement technique for these molecular processes will be translatable into human studies.
This rate-based BM approach may be particularly promising in combination with what is currently the most robust program for testing proposed interventions for extension of maximum life span in mammals: the National Institute on Aging's (NIA) Interventions Testing Program. These studies are the current gold standard for evaluating changes in lifespan in genetically heterogeneous mice but are time- and resource-intensive, limiting the number of interventions that can be tested each year. An initial screening strategy based on a panel of early BMs of maximum life span extension could be used to further refine which candidate interventions should be prioritized for inclusion into life span studies in mice. This represents an attractive approach for identifying interventions with the potential to extend human healthspan.
Here, we used stable-isotope mass spectrometric measurement tools to screen for BMs based on the activity of targeted physiologic pathways believed to be involved in the aging process. These measurements were performed in three different yet well-established mouse models of maximum life span extension, each on a different genetic background and each at relatively early time points in the life span of the model. The three models evaluated were Snell Dwarf mice, which are homozygous for a loss-of-function mutation in the Pit1 gene involved in anterior pituitary development; calorie-restricted (CR) mice, in which calories are reduced without malnutrition; and mice treated with rapamycin (Rapa).
A reduction in cell proliferation rates has been hypothesized to contribute to maximum life span extension by inhibiting the promotional phase of carcinogenesis and delaying cellular replicative senescence. A reduction in protein synthesis rates or slowing of protein replacement rates (RRs) (turnover) might in principle reflect preserved proteome homeostasis (proteostasis), which normally declines with age. A reduction in protein synthetic burden may preserve proteostasis by limiting the accumulation of misfolded and/or damaged proteins, possibly by increasing translational fidelity, chaperone capacity, and/or proteolytic capacity. Accordingly, the goal of the work presented here was to test the hypothesis that reduced cell proliferation rates, reduced protein synthesis rates, reduced protein RRs (prolonged half-lives), or other proteome alterations are early BMs of maximum life span extension in mice.
The major findings of this work are as follows: (i) A reduction in hepatic proteome RRs (longer half-lives) is a common feature of all three models evaluated (Snell Dwarf, CR, and Rapa-treated mice); (ii) a strong correlation exists between the degree to which hepatic proteome RRs are reduced and the degree of maximum life span extension in these models; and (iii) in vivo cell proliferation rates are not consistently reduced at early time points in all these models.
The first observation could have more than one underlying cause. We do not believe that our data suggest that a reduction in hepatic proteome RRs is an initiating factor that directly promotes maximum life span extension in the three models evaluated here. Rather, our hypothesis is that reduced hepatic proteome RRs reflect a reduced demand for protein renewal, and thus improved proteostasis in these models, likely due to reduced levels of misfolded and/or damaged proteins. The data presented here differentiate between several potential mechanisms of improved hepatic proteostasis in these models. In principle, a variety of cellular adaptations could reduce the levels of misfolded and/or damaged proteins, including an increase in proteolytic editing capacity, an increase in chaperone capacity, a reduction in the levels of damaging metabolites, or an increase in translational fidelity. The data presented here are not consistent with increased proteolytic editing as a major contributing factor to improved proteostasis, as reduced rather than increased proteome RRs (proteolytic rates) were consistently observed in all three models. We also found that the levels of the chaperones that we assessed were either unchanged or reduced in all three models, suggesting that an increase in chaperone capacity is an unlikely contributing factor to improved proteostasis. Consistent with improved proteostasis, as well as an absence of accumulated unfolded proteins in the ER in these three models, our proteomic analyses revealed that the synthesis of proteins involved in protein processing in the endoplasmic reticulum was reduced relative to the synthesis of all other proteins in the three models.