Will you live to 120? · The Cells That Won’t Die

Killing the zombies

Before we try to kill these cells, we have to confront a question that should, by now, feel familiar — because it is the same question the telomere chapter asked, and the answer rhymes. If senescent cells are so destructive, why does evolution permit them at all? Why build a cellular state whose whole job in old age is to poison you?

Because in youth, senescence is not a bug. It is one of your best defenses, and it does several genuinely good jobs. It is a powerful tumor suppressor: a cell that detects it is becoming cancerous and responds by locking itself permanently out of division is a cell that cannot build a tumor — the same brake-on-cancer logic as the telomere limit in Chapter 1, now generalized. It is essential for wound healing: transient senescent cells appear at injury sites and secrete signals that orchestrate repair, then are cleared once the job is done. It even helps sculpt the embryo, with programmed senescence shaping certain tissues during development. Acute, temporary, cleared-on-schedule senescence is a tool. It is chronic, accumulated, uncleared senescence — the same mechanism, left running too long and never mopped up — that turns toxic.

This is one of the cleanest real examples of an idea at the center of why aging exists at all: antagonistic pleiotropy. A trait that helps you when you are young and reproducing can be favored by evolution even if it harms you when you are old, because natural selection cares overwhelmingly about early life — about getting you to reproduce — and is nearly blind to what happens to you decades later. Senescence pays its dividend early (no cancer, good healing) and presents its bill late (inflammaging, tissue decay), and by the time the bill arrives, selection has already gotten what it wanted from you. Aging, in this view, is not a failure of evolution. It is partly the price of mechanisms that were selected precisely because they served you when it counted.

And from that follows the intervention caution that this book keeps returning to, now in its third chapter and its third form. In Chapter 1 it was “don’t just turn on telomerase — you’d be removing a brake on cancer.” Here it is: don’t just kill every senescent cell — because some senescence is doing necessary work, and blanket clearance could cripple wound healing, impair tumor suppression, and have effects we cannot fully predict. The obvious intervention is, once again, not free. The real target is not “destroy all senescent cells” but the much harder “selectively remove the chronic, harmful ones while sparing the acute, useful ones” — and we do not yet have a clean way to tell them apart.

Deeper Water — why evolution allows aging. Antagonistic pleiotropy, proposed by George Williams in 1957, is one of three classic, complementary explanations for why aging exists, and they are worth holding together. The oldest, from Peter Medawar, is the mutation-accumulation theory: because selection weakens with age (anything that only harms you after you’ve reproduced is barely “seen” by natural selection), harmful mutations whose effects appear only in late life accumulate in the genome, simply because nothing efficiently weeds them out. Williams’ antagonistic pleiotropy goes further: late-life harm isn’t merely tolerated, it can be actively selected for when it is the flip side of an early-life benefit — the same gene doing good early and ill late. The third, the disposable soma theory (Tom Kirkwood), frames it as an energy-budgeting problem: an organism has finite resources to split between reproduction and bodily maintenance, and evolution tunes the split to maximize reproductive success, not longevity — so the body (the “soma”) is built to last just long enough, then allowed to decay. Senescence sits squarely inside all three: a late-life-harmful state, the cost of an early-life-beneficial defense, and an example of maintenance that the body could in principle keep up but doesn’t. None of these theories is “the” answer; aging is overdetermined, with several true explanations at once. But together they deliver the unsettling and clarifying punchline that aging is not a disease in the ordinary sense, nor an accident — it is, in large part, the designed-in price of being an organism that was optimized to reproduce rather than to last.

Killing the zombies

Now the exciting part — and the reason senescence became, for a while, the hottest target in all of longevity science.

The logic is irresistibly clean. If accumulated senescent cells drive aging, then removing them should reverse the damage. And the first real test of that idea is one of the most striking experiments in the field. In 2011, a team led by Jan van Deursen built mice carrying an ingenious genetic switch — a system (nicknamed INK-ATTAC) that let them kill, on command with a drug, cells expressing high levels of the senescence marker p16 — many, though not all, of which are senescent. When they threw that switch and cleared the senescent cells, the mice aged visibly better: less deterioration in muscle, fat, and eye. Five years later, the same group showed that clearing senescent cells from naturally aged mice — not an engineered fast-aging model, ordinary old mice — extended their healthy lifespan and pushed out median survival. Remove the zombie cells, and the animals lived better and somewhat longer. It is hard to overstate how electrifying that was: a single, targeted intervention against one cell state, improving aging across multiple organs at once.

All 10 chapters of Will you live to 120? · Foundations · All courses