Jan van Deursen: "Link between Senescence, Aging, and Diseases of Aging" (Unity Biotechnology R&D Day, pt. 3)

Part 3 of a series of articles covering Unity Biotechnology’s inaugural R&D Day, a special presentation to investors and analysts highlighting the company’s recent progress and future plans. An index of our comprehensive coverage of the event appears in Part 1.

Prof. Jan van Deursen of the Mayo Clinic.

Prof. Jan van Deursen of the Mayo Clinic.

At the beginning of Unity’s 2018 R&D Day, Prof. Judith Campisi’s introduction to cellular senescence brought us from the founding observations of the field to our modern understanding of senescent cells as inflammatory drivers that persist in tissues and damage cells in their environment. Following Campisi’s talk, she turned the microphone over to Prof. Jan van Deursen of the Mayo Clinic, who explained the fundamental experiments in mice that have convinced us that senescent cells are responsible for multiple aspects of aging.

van Deursen’s approach to the field was somewhat indirect, beginning with his efforts to explore a very old hypothesis in cancer biology. “What I was trying to do 20 years ago is to test a theory posed by Theodore Bovary in 1914,” he began. “His theory was that cancer cells become aneuploid, that is, they acquire an abnormal number of chromosomes, and he was proposing or posing that this abnormal number of chromosomes was a driver in the formation of tumors. But tools to actually test that hypothesis were not available for a long, long time for technical issues.”

Fast forward a hundred years, when genome engineering technology had enabled targeted knockouts in mice. To explore the role of aneuploidy in cancer, van Deursen chose to eliminate BubR1, a gene involved in ensuring accurate chromosome segregation. “I was applying that technology to address this century-old question in cancer research. But then something happened. We started to know the genes that are necessary for accurate segregation of chromosomes, and we started to manipulate them, but we did not get a tumor phenotype—we got a premature aging phenotype.”

“We started to know the genes that are necessary for accurate segregation of chromosomes, and we started to manipulate them, but we did not get a tumor phenotype—we got a premature aging phenotype.”

In other words, the BubR1 knockout (KO) mice were not cancer-prone, but progeroid, developing aging-related ailments much earlier than their wild-type counterparts—and a closer examination revealed that they also accumulated senescent cells much more rapidly.

“What we observed in the cells that we cultured from these animals is that markers of senescence (p53, p21, p19, p16) were all elevated. And when we did a staining for senescent cells that Judy developed, senescence-associated beta-gal staining, there seemed to be a lot of these cells, for instance, in the kidney.”

The evidence, though circumstantial, strongly implicated cellular senescence in aging-related change. But diving into the field to follow this up would not be without risks. “This is now not in the days that senescence was very fashionable and everybody was looking at this,” he explained, “This was when senescence was very obscure, when prominent scientists in the field, mostly cancer biologists, were saying that senescence really was perhaps only a tissue culture artifact that may not even be relevant in vivo.” Nonetheless, van Deursen and his colleagues forged ahead.

To critically test whether senescence was contributing aging, they first knocked out a key regulator of senescence in the BubR1 knockouts. “We basically inactivated the p16 gene in this BubR1 progeroid mouse and asked the question, can we prevent the accumulation of senescent cells in this model? And if so, are these age-related phenotypes now attenuated? And the answer, as you can see, is that these aging-related phenotypes did not occur at the accelerated rate anymore.”

“We inactivated the p16 gene in this BubR1 progeroid mouse…And these aging-related phenotypes did not occur at the accelerated rate anymore.”

Promising, but this approach would not work in wild-type mice aging at the normal rate. “You cannot just proceed by eliminating the genes that drive cells into senescence, because as Judy already explained, those genes are important for preventing cancer. So by the time your animals become aged, they have tumors, and they will die. So you cannot just knock out p16 in an animal to prove this concept that senescent cells are driving aging. So we had to come up with another strategy.”

van Deursen then outlined the clever system his team used to conditionally eliminate senescent cells in aging mice: “What we did, basically, was to express a suicide gene under the control of the p16 promoter. If a cell becomes senescent and the p16 gene becomes active, this protein also gets expressed, but you need to then provide a drug, a dimerizer that links these two molecules together and then activates them. And this triggers the activation of apoptosis and the removal of senescent cells.”

“The median lifespan extension was about 25 to 30 percent, indicating that these senescent cells, while they are accumulating, take away years of our life.”

They used their system to eliminate senescent cells starting at the middle of the mouse lifespan, at a time when senescent cells have accumulated at detectable levels in tissues and organs throughout the body. The results were dramatic: “The median lifespan extension was about 25 to 30 percent, indicating that these senescent cells, while they are accumulating, take away years of our life. Also, a whole series of age-related diseases were attenuated by the removal—so, kidney disease, lipodystrophy, formation of cataracts. Also, the animals were a lot more exploratory and active. Cancer incidence was delayed, and osteoarthritis and sarcopenia—and very surprisingly, cardiac diseases were delayed.”

Importantly, this major intervention had no deleterious consequences. “We were removing these senescent cells for more than a year, let's say a year and a half, and we could not find any negative side effects. So, we are aware that senescent cells also have roles in tissue regeneration and repair, but we did not see any pathologies associated with the clearance of senescent cells that could be linked to that.” This implies that if we could devise a pharmaceutical means capable of eliminate senescent cells as cleanly as van Deursen’s suicide gene approach, it would be free of adverse side effects—a key consideration when developing a therapy meant to be administered chronically (or acutely on a regular basis) in elderly human beings.

“You're lucky if a tissue in some disease condition has a percent or so of senescent cells, so that makes it really hard to imagine how a few cells can wreak such havoc in a tissue.”

Having shown that senescence was necessary for the normal rate of aging, the team then did the converse experiment to determine whether they were sufficient. There was reason for healthy skepticism on this front: senescent cells are never more than a small minority of cells in a tissue. “You're lucky if a tissue in some disease condition has a percent or so of senescent cells,” van Deursen explained, “so that makes it really hard to imagine how a few cells can wreak such havoc in a tissue. So we wanted to kind of have more evidence that they could do that.”

For this purpose, they experimentally induced senescence in a small number of cells in the pericardium, a thin sac that surrounds the heart, amounting to no more than 0.01% of cells in the organ overall. Within a few weeks, the hearts of the mice doubled in size, a phenomenon known as cardiomegaly, and also exhibited fibrotic changes normally only seen in very aged animals or in models of myocardial infarction—clear evidence that senescent cells were actively driving age-related changes in an essential organ. “This illustrates how tremendously potent these senescent cells are in changing the tissue architecture,” he emphasized.

Van Deursen closed his talk by highlighting three diseases that we now believe to be caused or exacerbated by senescence, based largely on mouse experiments of the kind described above.

  • The pathogenesis of atherosclerosis involves the formation of plaques that are capped by collagen- and elastin-rich structures that prevent them from bursting. The proteases secreted by senescent cells weaken these caps, increasing the likelihood of a catastrophic collapse leading to heart attack or stroke. “We found that if you cure senescent cells while these mice are developing these plaques, you see less plaques and the plaques that you do that are very small… And if you remove senescent cells that accumulate in these plaques, even at the end stage, you are basically restoring these caps so they become thicker again, and the vascular smooth muscle cells increase again in number.”

  • In osteoarthritis, degeneration of cartilage in the joints leads to pain and physical disability. Senescent cells accumulate around sites of cartilage destruction, and SASP factors both degrade structural proteins directly and stimulate infiltration by immune cells, leading to the sterile inflammation mentioned in Judy’s talk. Elimination experiments revealed the causal role of senescent cells: “If you remove senescent cells, the cartilage almost looks like in a young animal. And that was demonstrated here in almost 100 percent of the animals. If you have an injury model that induces cartilage degeneration, you basically get the same result: removal of senescent cells really prevents the degradation, and also promotes the regeneration of cartilage. So that's all good, and that's the reason why Unity has this as one of their leading disease candidates.”

  • As we discussed here previously, Alzheimer’s disease may be driven by senescence in astrocytes and microglia, cells that support and maintain the neurons in the central nervous system. In particular, the secretory program of senescent glia creates inflammation and contributes to the formation of neurofibrillary tangles, a key pathological hallmark of multiple neurodegenerative diseases. “But with the transgenic model… you can eliminate these senescent cells. [Darren Baker found that] if you do that while this disease is developing, you found that you basically can largely prevent the formation of these tangles. And with the prevention of the formation of these tangles, these mice were able to fully preserve their cognitive ability.”

Over the course of his wide-ranging talk, van Deursen reviewed the large body of evidence that senescence drives both normal aging and age-related disease in multiple organs throughout the body. Accordingly, senescent cells represent a uniquely valuable target for drugs aimed at treating, curing, or even preventing these illnesses altogether.

Next, scientists from Unity reviewed the company’s ongoing and future research program. That article will be linked here as soon as it is complete.