by Matt Richtel
This is where Krummel began his collaboration with Allison, who won the 2018 Nobel Prize for what happened next.
Allison and Krummel decided to experiment further with CTLA-4, the other molecule that had bound with B7-1 and B7-2. They soon noticed a curious thing. When CTLA-4 was attracted to and bound to a ligand, the immune system didn’t ramp up as it had in the mouse experiment. Instead, the immune system seemed to be dampened or to have no effect at all.
“I thought, we gotta figure out what CTLA-4 does,” Allison reflected. Something about it nagged at him.
Krummel and Allison asked a question: If CD28 causes T cells to multiply, but CTLA-4 seems to have no effect, what might happen if you combined these agents?
What they discovered was a turning point. Stimulating CD28 led to an increase in T cells, and a heightened immune response. But when CTLA-4 was mixed in, it brought down the level of T cell response. Not only that: The more CTLA-4 was added, the fewer T cells proliferated. That suggested that CTLA-4, rather than causing the immune system response to increase, was causing it to turn down or even off.
They sensed they were on to something big.
Krummel devised a chemical process that would allow him to create varying levels of CD28 and CTLA-4 such that he could begin fine-tuning the amount of T cells created. The year was 1994.
“We could turn T cells up and down like turning a stereo up and down,” Krummel said. Or, if you prefer a different metaphor: “We found a hot tap and a cold tap. We immediately had this whiteboard discussion,” he said. What did this mean and what could they do with it?
They started experimenting, trying all kinds of combinations, one after the next. “In the course of nine months, we went from volume control—hot and cold—to every single animal model I could touch, pushing T cells to grow faster, watch them grow slower. That’s when Jim brought in the tumor model.”
Allison, by now steeped in this as virtually no other scientist, turned ideas over and over in his mind, trying to make sense of it all. What did these molecular interactions add up to? He joked with me about the pieces finally coming together one night in 1994, while his mind was wandering after “too much wine.” He thought he might understand how cancer was playing a trick on the immune system, allowing the disease to evade our defenses. And he had an idea how to reverse the trick.
Allison had invited a postdoc named Dana Leach into the lab. Leach brought the rodents with the tumors, which now were out of the test tube and into actual critters. The vet injected rats with several fast-growing cancers. The researchers let the cancers blossom. Then they injected the mice with a molecule—an antibody—that was aimed at disrupting any connection that the cancer cells might be trying to make to the CTLA-4.
The idea was to see if they could keep the cancer from turning on the brakes of the immune system by disrupting the communication between the cancer and the immune system.
“We were just trying things out,” Krummel said.
A few days later, Allison came in to check out the progress. “I went, ‘Holy shit! It cured all the mice.’”
The previous experiment had entailed isolating the tumor tissue in a test tube and then modifying its genetics so that it would stimulate a T cell response. This was ultimately impractical.
But in the new experiment, the researchers did nothing at all to the tumor. It was just a tumor, like one that might be growing inside any of us, eventually inside Jason. It was in its natural state.
Instead of changing the tumor, the researchers added antibodies to interrupt cancer’s trick and stimulate a response from the immune system. Specifically, they inserted an antibody that bound to the immune system so that it would take off the brakes in our elegant defense.
“What was surprising is that we hadn’t given the immune system any new information about the tumor,” Krummel said. “There was a set of preexisting cells”—the T cells—“and they were raring to go.”
When Allison looks back, he thinks of the immune system in a vastly different way than we conceived of it for the longest time. He doesn’t think of it merely as a powerful killing machine, not at all. Instead, he thinks of it as sharing killing powers with extraordinary powers of restraint. One of the chief jobs of the immune system is to shut down its attacks, hit the off button. Screeching brakes get applied to the T cells.
“They get a signal. They kill themselves. If it didn’t work, people would get diabetes, multiple sclerosis, lupus,” he said. “By far this negative selection is the central tolerance, to get rid of T cells; 90 percent of every T cell that’s developed gets killed.”
He’d figured out what CTLA-4 did. “CTLA-4 is there to protect you from being killed by your immune system.”
Whoa.
But isn’t cancer killing people? Why would our bodies allow the brakes to come on in the face of a deadly tumor?
The answer is related to a trade-off with wound healing, which is one of the most important functions of both the body and the immune system.
39
Wound Healing
If you get an acute injury—say, you step on a stick or cut your hand on the edge of a can—the event sets off an urgent survival cascade. Red blood cells rush to the scene and begin to coagulate or clot. They stop the bleeding. Cells from elsewhere move into the gap and start to divide. These include immune system cells, neutrophils and macrophages.
Sabine Werner, an expert in wound healing, describes it like emergency crews arriving on the scene. “There are very rapid events to get closure through the blood clot.” The immune cells are there to deal with “bacteria, fungi, viruses—they can all be there.”
The neutrophils produce proteases, which attentive readers now know of as enzymes. These enzymes are a bit like a grenade. They make holes in certain bacteria, “killing them actively. Bacteria are also eaten by neutrophils and by the macrophages,” Werner said. Tidy.
In addition to neutrophils, there comes a second vicious killer. It has one of those impossible-to-remember names: reactive oxygen species (ROS).
Just remember this: It’s nasty. One such reactive oxygen species is hydrogen peroxide. The macrophages and neutrophils both can synthesize the chemical to kill in the area of the wound. The neutrophils and other killers didn’t just take out bacteria or other possible infections. They also killed some of the other surrounding tissue. This is the reason that often, after you experience a wound, even a minor one, the pain and inflammation are worse in the days that follow the event. Your immune system has done housecleaning with industrial-strength chemicals.
The area has been cleansed of “other,” leaving scorched earth. Then in the dead zone, the macrophages eat.
Almost as quickly, construction workers move in. In the early 1990s, Werner, researching the phenomena, noticed that a wound site, within about one to two days, exhibited a tenfold increase in growth-promoting signals. She began to vigorously pursue the question of what is happening within the body that allows such rapid healing. Where are the signals coming from that allow cells to divide quickly and replenish the tissue?
Think about how dramatic this transition is. One moment your tuna-can-slashed finger is being swept by a SWAT team, and then, within hours, an entire construction process has come in and replaced the killing machine. And what were the implications for the overall system of health? “I got really excited at how fast a wound can react to this,” she reflected.
She didn’t know yet about the dark side.
The reconstruction process has, of course, its own complicated language. The term for one of the key cell types stimulating regeneration of our tissue is fibroblast—highly versatile and hearty cells that proliferate and migrate to the site. These cells are drawn by signals sent by macrophages. This is of note in that it shows a different side of the macrophages. These “big eaters” also play a role in stimulating the growth of new tissue.
As the fibroblast cells come together, they form connective tissue, a bridge between the new and old tis
sue. At the wound site, the new tissue takes on a granular quality, hence its name granulation tissue. Crucially, these tissues are fed by blood vessels that spring up around the edge of the wound, creating veritable feeding tubes for new tissue. A kind of tenacious web forms, a fibrous matrix that, as Werner and her coauthors put it in one paper, protects against invading pathogens and “is also a reservoir of growth factors that are required during the later stages of healing and it provides a scaffold for the different cell types that are attracted to the wound site.”
In the Festival of Life, a particular party spot is imploded, then cleared away of debris. Next comes the construction of the foundation and of scaffolding, and then rebuilding starts. But as is true of many construction projects, permits must be obtained. The body must accept that what is being built is approved of as “self.” Anything seen as alien to the point of being pathogenic will be destroyed, and the site will not be rebuilt.
There is a dangerous corollary. Once permission is given, once the new cells being nourished are deemed “self,” the construction can go on with zeal. The trouble is, the new cells aren’t always self. Sometimes they are cancer.
And so the factors that promote growth of healthy tissue also appeared to promote the growth of tumors. This was an idea that had been floated since 1863, when Rudolf Ludwig Carl Virchow, a German scientist, observed: “Chronic irritation and previous injuries are a precondition of tumorigenesis.”
Werner gives talks in which she cites two other equally prescient quotes:
“Tumor production is possible overhealing,” commented Scottish physician Sir Alexander Haddow in 1972.
And then there was the observation of Harold Dvorak, a pathologist in Massachusetts, who in 1986 said, “Tumors are wounds that do not heal.”
The wisdom of these statements has been borne out in powerful lab experiments.
One decades-old telling experiment had been performed using baby chicks. The experiment, done at Berkeley, involved injecting into chickens a virus known to give them cancer. The chickens were injected under the skin or in muscle. In either case, the injection caused a minor wound.
Within one to two weeks, a tumor appeared, usually at the injection site. The chicks died within a month.
The researchers decided it was reasonable to assume that the wound itself was relevant to the growth of the tumor and came up with a second experiment to prove that point. This time, they infected a chick in the right wing but not the left wing. At the same time, though, they pierced the right wing. Lo and behold, a tumor formed at the injection site and at the site of the wound on the other wing. The tumor on the wing that was wounded but not injected took about 20 percent longer to appear.
Something about the wound clearly was playing a role in promoting the tumor.
In the 1990s, Werner started to put the pieces together. What she and others discovered begins to explain why things like smoking or coal mining or sunbathing are so carcinogenic. Each activity injures the tissue and damages the DNA. When the tissue is damaged, the immune system kicks in and cleanses the site and helps stimulate new tissue growth. The trouble is that when the DNA is damaged, the new cells that grow can be malignant cells, some that are made up of self but that are different enough to behave like a cancer. These cells aren’t playing by the normal rules of the body and staying within their boundaries. Add all this together and you can wind up with cancerous cells that are protected and even nurtured by the immune system.
This explains too some risks of cancer experienced by sufferers of certain autoimmune disorders that cause chronic tissue injury.
When a wound occurs—an insult, as it is known in scientific circles—cells divide. Of course they do. New tissue is needed. But when new cells divide, there is always a chance something can go wrong. Each cell division is an opportunity for a mistake, a mutation. A piece of DNA might get incorrectly copied, for instance. This happens all the time. Fortunately, in most cases, this mutation has no consequence because the cell dies or gets rapidly devoured. The mutation is so unusual that the cell can’t survive because essentially it lacks the genetic material to live, and the macrophages eat the refuse. Story over. At other times, the mutation is picked up by the immune system as being sufficiently foreign so as to be potentially problematic. It is bombed or blown up, destroyed and then eaten. Story over.
Sometimes, though, the mutation is extremely subtle. The cell has the genetic material sufficient to survive, and it is sufficiently like “self” that it isn’t recognized as problematic by the immune system. In some cases, the immune system tests the material but decides that it is more likely self than not and leaves it alone.
This doesn’t mean that such a cell is necessarily cancerous. A cell with a single mutation is highly unlikely to be cancer. Werner explained to me that a cell that turns cancerous needs to undergo at least five to ten different genetic changes. Not just that, to be a “perfect bad cancer cell,” the random genetic occurrence needs specific changes in different regions of its DNA. For instance, a mutated cell that is likely to survive and become cancer has chanced upon the ability to send signals to immune cells with the instruction: Don’t attack me; protect me and nurture me.
“They secrete factors which change the immune cells,” Werner told me. For example, “the macrophages are not inflammatory anymore, but rather, they protect the cancer cells and stimulate the formation of blood vessels.”
This is a juncture where cancer takes advantage of the immune system. The cancer grows and grows, quietly protected, fed by blood vessels, even guarded by fibrous networks. The tumor “is cruising along, invisible and growing,” said Allison, the pioneering researcher of CTLA-4.
But then “at some point, [tumors] reach a certain size and they can’t get enough oxygen, enough food,” Allison explained; they have gotten too big for their environment. “They start dying,” and the macrophages come in, phagocytosis takes place, the tumor debris gets cleaned up, and then the immune system starts providing more growth infrastructure, as it would in a healing wound, and simultaneously CTLA-4 shuts down the attack dogs.
It’s a vicious cycle perpetuated by the immune system. Full stop. The immune system starts to feed and nurture the cancer. Your elegant defense has turned against you.
What this adds up to is that the likelihood of getting cancer depends in large part on how often a person experiences injury or certain types of injury. This is just math. More injury means more cell division and, simply, more opportunity for dangerous mutation to occur.
Enter one of the world’s biggest killers.
When someone smokes a cigarette, tiny little wounds are created inside the fragile pink tissue of the lung. Into the lungs pour several thousand chemicals, including a number of them that not only damage the DNA but that interfere with repair of DNA. Meanwhile, the police and fire brigade of the immune system shows up, and the process of wound healing begins. New cells are created. Over and over and over, cigarette after cigarette, year after year. (Smoking is a chronic activity, as opposed to, say, inhaling fewer chemicals less directly at the occasional campfire.) In the case of smoking, the malignant cells are fed and protected by the same system that cleaned out the wound in the first place and made sure that no pathogens were there to inflict harm.
Some of those new cells are mistakes and are recognized as nonself. And some have the right combination of random mutations to live and look a lot like self, so much so that the immune system, the very system created to defend us, becomes the promoter and protector of the tumor.
Again, in key respects, cancer is just a numbers game. The more wounds you get, the more mutations and inflammatory events, the greater the likelihood of cancer. That’s what makes things like smoking so potent. The risks grow with each puff. Similarly, sun exposure, absent sunscreen, presents another opportunity for a wound and an inflammatory response, which, combined with mutations directly induced by UV irradiation, enhance the risk of the development of skin cancer, including the
particularly dangerous melanoma. Other toxins that come into the body, whether food toxins or chemical ones, can also create wounds, places of insult, even minor, that require repair, inflammation, rebuilding. Each minor assault is a chance for cell division and an immune system response that, while intended to cleanse, might also lead to cancer. Smokers are almost certain to get cancer at some point, due to the very certitude of math. If you’re a smoker, you might have cancer right now. In fact, you probably have cancer right now. Most likely, though, it lacks the precise types of genetic changes that will let it proliferate by, in particular, co-opting the immune system. Just because cancer exists doesn’t mean it will take hold.
Those of us who do not engage in such high-risk behaviors are much less likely to get cancer, or rather, we’re not as likely to get it as quickly. But if we live long enough, math will catch up with us too.
The fact that you’re likely to get cancer eventually, and that it will take hold eventually, says a mouthful about the trade-offs being made by your immune system. It has evolved so that it allows for the possibility, even likelihood, of cancer taking root. The reason is simple: It is willing, in the short term, to risk a mutation taking hold in order to allow for the immediate rebuilding of tissue. What, after all, would be the alternative? Leave holes in your tissue? Allow your body to be chipped away at, one nick and cut at a time?
Cell division is a must. Mutation, cancer, is a by-product of cell division. It is one reason your death is preordained. This dynamic, though, also holds the keys to combating cancer. That’s what Allison began to exploit with CTLA-4. A second major conceptual discovery has helped tinker with the immune system to turn the odds in favor of life.
40
Programmed Death
Recall that James Allison had discovered we could modulate the immune system by playing with CTLA-4. That’s the molecule on T cells that helped dampen or kill an immune system response.