A Start on Mapping Biomarkers of Cellular Senescence by Tissue and Age
Cellular senescence is one of the root causes of aging. Cells enter a senescent state in response to damage or the end of their replicative life span, and near all quickly self-destruct or are destroyed by the immune system. Others enter senescence to assist in regenerative processes following wounding, again being destroyed soon afterwards. Senescent cells that linger are a real problem, however. They generate harmful signaling that produces chronic inflammation, destructively remodels tissue structures, and changes the behavior of nearby cells for the worse. The accumulation of senescent cells over the years directly contributes to the progression of age-related dysfunction, disease, and risk of death.
Just how many senescent cells is any given individual burdened with, however? What should we expect from this cause of aging at a specific age? Is it negligible at 40 or 50, with a sudden leap to significant levels at 60? Does the answer vary by tissue type? How do the usual health-associated lifestyle choices affect these numbers? Are senescent cells significantly different from tissue to tissue in terms of the signals they generate and the harm done?
The answers to these questions are not yet established in any robust way, but the development of therapies capable of destroying senescent cells is proceeding regardless - there is plenty of evidence to show that removing these cells is beneficial, even without the greater insight into the fine details. This more detailed information is important, however, when it comes to the energy with which any particular individual should pursue access to the first generation of senolytic therapies capable of destroying senescent cells, and where those groups involved in therapeutic development should focus most of their attention.
One of the paths to a better understanding of how the burden of cellular senescence progresses with age, and how that progress varies by tissue type, is the production of a more detailed mapping of biomarkers of senescence. The open access paper here is an example of this sort of work, initially focused on mice. Better and more discerning markers of cellular senescence and the harms it creates will help to validate existing senolytic therapies and steer the development of new and better approaches.
Age- and Tissue-Specific Expression of Senescence Biomarkers in Mice
Cellular senescence plays a complex role, both beneficial and deleterious, in biological processes such as embryonic development, wound healing, tissue regeneration, and tumor suppression, as well as age-related disorders. Senescent cells accumulate within aged tissues and at sites of age-related pathology in vivo, and potentially contribute to the age-related decline of tissue function by affecting the growth, migration and differentiation, of neighboring cells, impacting overall tissue architecture, and promoting chronic inflammation. Indeed, studies on both progeroid and naturally aged mice showed that selective elimination of p16Ink4a-expressing senescent cells increased healthspan and lifespan. Thus, the selective elimination of senescent cells (senolytics) or the disruptions of the senescence-associated secretory phenotype (SASP) program have been developed as potential therapeutic strategies against aging.
However, while p16Ink4a expression has been used as a classical senescence biomarker, no biomarker of senescence identified thus far is entirely specific to the senescent state. Thus, due to the lack of robust biomarkers of cellular senescence in vivo, the precise extent of senescent cell accumulation in aged animals and the functional outcome of such an accumulation, along with the exact target cells of, and removal by, senolytics, remain unclear. Surprisingly, a systematic multi-tissue in vivo study of senescence markers during aging has not been conducted in wild-type animals.
In the era of senolytics, it becomes imperative to develop robust biomarkers of senescence in vivo for preclinical trials, especially with several senolytics now nearing human clinical studies. As a first step, in this study we profiled the expression of a panel of known molecular markers of senescence in multiple tissues in mice at multiple ages, ranging from young (4 months) to very old (30 months). The results demonstrate that the secretory profiles and classical hallmarks of cellular senescence in aged tissues are highly variable and complex, suggesting that a systematic and concerted effort is needed to develop robust biomarkers of senescence for the identification, quantification, and monitoring of senescent cells in vivo.
The wide diversity in tissue-specific profiles we observed was striking. Nevertheless, the matrix metalloproteinase Mmp12 represents a robust SASP factor that showed consistent age-dependent increases in expression across all tissues analyzed in this study. It has been demonstrated that mice lacking Mmp12 are protected from vascular injury, M2 macrophage accumulation, and perivascular heart fibrosis. Together with our data, this finding suggests that Mmp12 upregulation with age has a deleterious impact on heart function.
In this study, we did not observe significant age-dependent upregulation of the prominent SASP cytokine Il6 in any tissue, although an upward trend was observed that was consistent in magnitude with previous observations in the heart and kidney. This modest age-related upward trend could be explained by a previous report which demonstrated that senescent cell-secreted IL-6 acts in an autocrine manner, reinforcing the senescent state, rather than inducing senescence or promoting dysfunction in neighboring cells.
The decreased expression of Il6 with age we observed in the hypothalamus could be indicative of a lack or loss of senescent cells in that tissue with age. In support of this interpretation, p16Ink4a expression was non-detectable in the hypothalamus at any age. Taken together, these results suggest that some other age-related process results in the increased expression of the pro-inflammatory factors Il1b, Mmp12, Cxcl1, and Cxcl2 observed in the aged hypothalamus. Conversely, p16Ink4a expression was upregulated with age in all other tissues analyzed, consistent with previous reports, and thus reinforcing the importance of p16Ink4a as a biomarker of tissue aging.
Questions still remain, however, regarding the ultimate identity of the cells targeted for senolytic elimination in previous studies, as it has been demonstrated repeatedly that p16Ink4a expression is not exclusive to senescent cells, and thus does not represent an unequivocal target for senolytic therapies. Interestingly, however, CDKN2A (the gene that encodes p16Ink4a) was one of the top human genes that exhibited elevated expression with age, in 6 out of 9 tissues, including subcutaneous adipose, tibial artery, lung, skeletal muscle, tibial nerve, and whole blood, as detected by RNA-seq analysis. Thus, utilizing p16Ink4a-expressing cells as a biomarker of tissue aging and a target of senolytic therapies could still prove to be an effective strategy in the future treatment of age-related diseases in humans.
The CDK2NA gene SNP rs10757278 AA alleles is protective of heart attacks, CAD, strokes, aneurisms, etc., while the GG alleles results in increased levels of these diseases with age. I happen to have the AA alleles that is carried by about 25% of Caucasians. A good reference for the increase of p16Ink4a that is encoded by this gene at increased ages is: Tsygankov, 2009, Proc. Natl. Acad. Sci. USA, A quantitative model for age-dependent expression of the p16INK4a tumor suppressor. If we could get the protective alleles expressed in all the tissues cited in the article that have higher expression of p16INK4A with age, we might do wonders for reduction of many diseases of aging, and thus greatly benefit longevity.
rs10757278 is probably an optimization polymorphism. G allele is very evenly spread for CEU population (to the point that it could be named wild variant), but almost nonexistent in african YRI population. My guess is that its connected to Nordic genotype and probably an optimization to the ancestral diet(and goes the wrong way for the modern diet).
@Andey: The longevity gene FOXO3A SNP rs18002292 T allele shows a similar geographic distribution between CEU European population and YRI West Africans that the CDKN2A gene SNP rs10757278 G allele has. Whether this has anything to do with ancestral diet in the respective areas as you suggest for the latter SNP allele or maybe it is just coincidental.