Senescent cells accumulate throughout the body with age. They are constantly created and destroyed throughout life, but the balance between creation and destruction is upset with age, leading to an accumulated burden of cellular senescence. These cells secrete a potent mix of signals that produce chronic inflammation, disrupt tissue structure and cell function, and encourage other cells to also become senescent. The more senescent cells, the worse the impact. They are an important contributing cause of aging.

Targeted removal of senescent cells has been shown to meaningfully reverse the progression of age-related disease for numerous conditions in animal studies. Further, it extends life span in mice. These results are generally easily replicated and quite robust; we should take it as well settled that clearance of senescent cells in mammals produces literal rejuvenation. Even so, some researchers have suggested that senescence in old tissues might be in some way adaptive, preserving cells that would otherwise not be replaced. Perhaps the balance of negative impacts favors their retention rather than destruction, even though these cells cause harm in their senescent state. This argument has been put forward for senescent T cells of the adaptive immune system, for example, given that new T cells are barely created at all in late life, and even though there are plenty of concrete examples of senescent T cells causing harm.

The proposition of cellular senescence being, on balance, better than the alternative of losing the cells in question entirely cannot be universally (or even broadly) true throughout the body, given the existing animal data on clearance, but perhaps it is true for smaller populations of senescent cells in specific tissues. In the paper I’ll point out today, researchers suggest that a class of liver endothelial cells are one such population. This must still be balanced with the greater weight of research suggesting that global clearance of senescent cells is unambiguously beneficial, but as noted here, there are questions as to just how global that global clearance is for various approaches. Populations that are beneficial may be skipped by one or another type of therapy, a situation that could lead to roadblocks in the development of therapies down the line. We shall see.

Defined p16High Senescent Cell Types Are Indispensable for Mouse Healthspan


Substantial evidence has demonstrated that the accumulation of senescent cells can drive many age-associated phenotypes and pathologies. For example, senescent cells accumulate in adipose tissue of patients with diabetes and age-related metabolic dysfunction, in osteoarthritic joints, in the aorta in vascular hyporeactivity and atherosclerosis, and in the lungs in idiopathic pulmonary fibrosis. While selective elimination of these senescent cells confers notable benefits in some tissues, recent studies have also described beneficial roles for senescent cells, raising the question of the differential roles of these cells in various tissues.

Senescence-ablator mouse models have pioneered the field of in vivo senescence studies. With the use of one such model, known as the INK-ATTAC mouse, that is based on a 2,617 base pair fragment of the p16Ink4a gene promoter, it has been proposed that removal of p16-expressing cells results in life extension in mice. The concern, however, is whether the reporter construct with a part of the p16 genomic sequence fully resembles endogenous p16 gene expression, especially with aging. This is further supported by the fact that some p16-expressing cells are not efficiently removed by the INK-ATTAC system in several tissues, including the liver, colon, and T lymphocytes. Thus, it is unclear whether there are important senescent cell types in tissues where the INK-ATTAC system does not work and what impact their removal has on health span.

Here, we generated two knock-in mouse models targeting the best-characterized marker of senescence, p16Ink4a. Using a genetic lineage tracing approach, we found that age-induced p16High senescence is a slow process that manifests around 10-12 months of age. The majority of p16High cells were vascular endothelial cells mostly in liver sinusoids (LSECs), and to lesser extent macrophages and adipocytes. In turn, continuous or acute elimination of p16High senescent cells disrupted blood-tissue barriers with subsequent liver and perivascular tissue fibrosis and health deterioration. Our data show that senescent LSECs are not replaced after removal and have important structural and functional roles in the aging organism. In turn, delaying senescence or replacement of senescent LSECs could represent a powerful tool in slowing down aging.