Just How Dynamic are Cellular Senescence Levels in Old Tissues?

Accumulation of senescent cells is one of the root causes of aging. Based on the comparatively few measures established in old tissues, the proportion of cells that are senescent does not rise to more than a few percent of all cells even in very old individuals. That few percent is enough to wreak havok, however. Senescent cells actively secrete a mix of signals that promote chronic inflammation, destructively remodel tissue structure, and change the behavior of surrounding cells for the worse. They are harmful enough to be a significant direct contribution to many age-related diseases. Data exists for their baleful influence to produce osteoarthritis, fibrosis of the lung and other organs, and many other conditions.

Given all of this, there is considerable enthusiasm for the development of means to selectively destroy these cells: small molecule senolytic drugs, immunotherapies, and suicide gene therapies are all under development, the first now in human trials. Interestingly, despite some years of this active development, the ability to accurately and usefully measure the count and life span of senescent cells in tissue has lagged behind. There are methods that work well enough in animal studies, but few approaches that are useful in human medicine, and none of them are yet widely used. So there is really very little data on the degree to which senescent cell counts rise and fall over time, in response to environmental circumstances. All that is known for certain is that old people have more senescent cells. Are those senescent cells lasting for years? Are they created at a small rate and linger for decades? Is there are a rapid turnover in most tissues, and the increasing number is a function of dysfunction in the processes of removal?

In this context, the research reported in the open access paper here is most interesting. The authors show quite large short-term variations in cellular senescence in muscle tissue in response to strength training in young people. Even given the youth of the subjects, taken on its own it suggests a cautious reevaluation of the idea that all senescent cells accumulate slowly and last for a long time, and thus that senolytic therapies would have to be undertaken only infrequently. (Not to mention posing the question of how much of the way in which strength training improves health in older individuals is due to eliminating senescent cells).

Yet this must be balanced with the established evidence for significant lasting benefits to result from a single senolytic treatment in mice, which seems only possible if senescent cells arrive at a slow rate and linger for a long time following creation. It is possible that there are different populations and types of senescent cells, some dynamic, some not. It is also possible that the standard senescent markers show up in cells that are not senescent in some circumstances. It is likely that senescence dynamics are quite different in different tissue types. Whatever the answers, it seems clear that assessment of senescent cell counts and dynamics is overdue a greater level of attention.

Aged cells in human skeletal muscle after resistance exercise

Most of the cells in the human body are continuously aging, dying and regenerating to gradually evolve a fairly stable size of multicellular system with a wide range of cell ages. Skeletal muscle is the largest tissue of the human body, in which cell lifespan varies considerably among different cell types. For example, myofibers are long-lived, whereas endothelial cells in capillary surrounding myofibers age rapidly with a short half-life around 2 weeks. Selective elimination of senescence cells in skeletal muscle and other tissues has been shown to increase lifespan in mice, suggesting a promising approach for anti-aging intervention. The protein p16Ink4a, a cyclin-dependent kinase inhibitor CDKN2A, is a widely used senescence marker expressed specifically in aged cells. However, p16Ink4a+ senescence cells in human skeletal muscle are rarely studied. It is currently unclear whether senescent cells are accumulated in human skeletal muscle at young age and whether exercise has significant influence on its number.

Senescent cells can be selectively recognized and rapidly cleared by phagocytic macrophages. One way to direct macrophages into skeletal muscle is resistance exercise. After weight loading, phagocytic macrophages (M1 phenotype) infiltrated into damaged sites, followed by protracted presences of regenerative macrophages (M2 phenotype). The cell turnover process instantly demands nitrogen sources from amino acids or proteins for nucleotide synthesis and DNA replication. A delayed protein supplementation after resistance training can significantly undermine muscle hypertrophy, suggesting a far-reaching impact of protein availability in time around exercise challenge on long-term muscle adaptation. It remains uncertain whether protein availability influences macrophage presences and senescent cell clearance in exercising skeletal muscle.

In this study, senescent cell distribution and quantity in vastus lateralis muscle were examined in young human adults after a single bout of resistance exercise. To determine the effects of dietary protein availability around exercise on senescent cell quantity and macrophage infiltration of skeletal muscle, two isocaloric protein supplements (14% and 44% in calorie) were ingested before and immediately after an acute bout of resistance exercise, in a counter-balanced crossover fashion. An additional parallel trial was conducted to compare the outcome of muscle mass increment under the same dietary conditions after 12 weeks of resistance training.

The main findings of the study are as follows: 1) No senescent myofibers are detected in the skeletal muscle of young men aged between 20-25 y; 2) Most of the senescent cells found around muscle fibers are endothelial progenitor cells; 3) A single bout of resistance exercise reduces the senescent endothelial progenitor cells by 48% in challenged muscle and maintains at low levels for 48 hours; 4) Resistance exercise with low protein availability is associated with greater increases in macrophage infiltration and further depletion of senescent endothelial progenitor cells in muscle tissue during recovery, but prevents muscle hypertrophy for a long term. Taken together, these data suggest that senescence cell clearance and muscle mass increment are associated with the magnitude of muscle inflammation after resistance exercise, which can be influenced by protein supplementation around exercise.


If weight training can reduce the SC numbers for only 48 hours that would mean a very quick turnover. Of course , there could be a special type of SC to promote growth , which gives a whole new meaning to the muscle soreness and pain after exercise.

Then what if the SC promote growth by their inflammation factors and they sacrifice themselves in the process? Seems a bit wasteful as am approach , but hey. No pain, no gain.

Can specialized non-progenitor/stem cells become senescent?

Can the age of the SC be assessed by isotope composition in an already aged body? Can we use some other technique or a combination of approaches?

If there is no difference between senescent cells who promote growth and the SC that promote inflammation why aren't the old people admirable bodybuilders?

Let's hope that senescent cells removal didn't turne out to be yet another promising therapy that doesn't translate will from the mice model.

Posted by: Cuberat at July 2nd, 2018 7:39 PM

Low protein availability enhanced senescent cell clearance.

Low protein also enhanced phagocyte and leukocyte penetration into muscle.

This might be a reason why physical activity and a low protein diet is common in the well studied centenarians.

Posted by: Lee at July 3rd, 2018 6:48 AM

I think we have to be very skeptical of this finding for now. Their only basis for claiming a reduction in senescent EPCs is p16 expression, but the leaders in the field (Campisi, Kirkland, Van Deursen, etc) constantly emphasize that no one marker is sufficient evidence of senescence: in the case of p16, it is also expressed in activated macrophages and possibly other cell types. If you read any of their papers, they (almost?) without exception assess multiple markers (such as senescence-associated β-galactosidase, SASP factors, γ-H2AX, SAHF, downregulated Lamin B1, etc) in any given tissue or cell type, particularly when evaluating the effects of interventions. Additionally, it is particularly important not to rely on p16 expression alone without getting protein levels, not only because changes in mRNA expression do not always lead to a corresponding change in protein levels, but also because several genes (p14, p12, and p16-γ) are also encoded on the CDKN2A locus, leading to potential false positives if the primer is not very specific.

Additionally, their interpretation rests on the idea that senescent cells are cleared by macrophages, for which they cite as evidence this paper - but these are not actually senescent cells per se, but merely "old" red blood cells. From extensive work on immune surveillance of senescent cells, it is evident that the lead role is played by natural killer cells, with a role for macrophages primarily in cleaning up the debris after such cells are destroyed.

I think that also largely refutes the idea that the skewing of macrophage polarization with age is responsible for the accumulation of senescent cells with age, for which there are multiple alternative explanations (increased generation due to rising oxidative stress and replicative exhaustion; persistence of a subset of senescent cells that evade innate immune surveillance due to engaging known mechanisms; the ratcheting effect of established senescent cells inducing neighboring cells to become senescent via exosomes; alterations in innate immunity; etc).

Posted by: Michael at July 3rd, 2018 6:59 PM

Macrophages and Neutrophils clear cellular debris and possibly other immune cells too. There is a lot of redundancy built into the immune system. The polarization of macrophages and other immune cells is influenced strongly by inflammation.

I also highly recommend people investigate the role of the gut microbiome, butyrate production, and loss of gut membrane integrity as the potential starting point of inflammaging.

Posted by: Steve Hill at July 4th, 2018 7:29 AM

Here we see how neutrophils clear cell debris to facilitate nerve repair. It was assumed that macrophages were responsible for engulfing debris and clearing up but this study shows otherwise. They screened the various types of immune cells and bingo the found neutrophils also clear up.

Lindborg, J. A., Mack, M., & Zigmond, R. E. (2017). Neutrophils are critical for myelin removal in a peripheral nerve injury model of Wallerian degeneration. Journal of Neuroscience, 2085-17.

Posted by: Steve Hill at July 4th, 2018 7:33 AM


One thing to note regarding van Duersen, et. al., in their 2016 paper (Nature 530, 184-189) they demonstrate that p16 positive cells shorten healthy lifespan. They are targeting senescent cells for removal by way of p16 expression through their INK-ATTAC genome modification and AP20187 administration. Even though they may say that multiple markers are needed to definitively identify senescent cells they, as well as Oisin Biotechnologies (and others), are primarily targeting for removal cells expressing p16. Interestingly, targeting p16 cells for removal does not seem to improve muscle strength, coordination or balance in the mice used in the 2016 van Duersen study. I am not implying that more sophisticated methodologies that use combinations of markers to target senescent cells for removal are not needed.

I need to read this paper carefully to ascertain whether or not their claim that macrophages are responsible for removing the senescent endothelial progenitor cells is reasonable. This observation could possibly be tested in a murine model using clodronate to deplete macrophages, but I am unsure of methods to do resistance training with mice. There may also not be a similar effect as seen in humans.

Posted by: aaron at July 6th, 2018 1:07 AM


Thanks for the link.


I read the paper and their assertion that macrophages are clearing the senescent endothelial progenitor cells is severely lacking. In fact, they state "Senescent cells can be selectively recognized and rapidly cleared by phagocytic macrophage[7]" and reference the Kay paper from 1975. The 1975 Kay paper showed that IgG attachment to senescent red blood cells makes them 'vulnerable to phagocytosis by macrophages.' It is known that macrophages express FcγRI which binds to human IgG1 and IgG3 and can result in phagocytosis when IgG1/3 are bound to a target. However, the authors of the paper in question never mention the antibody-dependent cell mediated cytotoxicity mechanism. This sort of points to a bit of sleight of hand on their part. My guess is that they had to have some semblance of a mechanism to explain their observations and found this paper by way of Google. The reviewers of this paper did a poor job with letting that slide, but thanks to you for calling that reference out as being poor. The observations in the paper are interesting, but the mechanism they put forth is tenuous at best and highly circumstantial.

@Reason mentioned dynamics of senescence:
Does the presence of these p16+/CD34+ cells serve a similar function to that seen in wound healing (e.g., secreting PDGF-AA)? Meaning are they serving a positive function with respect to promotion of muscle repair? Additionally, what is the mechanism of their removal (if they are in fact being eliminated)?

Posted by: aaron at July 8th, 2018 2:02 AM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.