music // Title card reads Back to Basic Toronto, 2014. Polly Matzinger PhD. The Danger Model: Update to immune system theory. Polly Matzinger stands alone on the main stage at the Back to Basic conference. An on-screen graphic identifies her as Chief, T-Cell Tolerance and Memory Section Laboratory of Immunogenetics. As she speaks, she draws on an overhead projector.
Polly Matzinger: I’m going to take you through 65 years of immunological theory in one diagram, to show you how we differ from the older models, and how we stand on top of some of them, all right?
Matzinger draws two circles, one labelled “B” and one labelled “Killers.”
So, the self-non-self model was invented in 1953 by a guy named Macfarlane Burnet in Australia. And, if you look at that model in modern terms, what he said was that the B-cells and the killer cells — the effector arm of the immune system — work by recognizing non-self.
Matzinger draws several arms on the surface of the B-cell circle.
What he said was that a B-cell has on its surface a whole bunch of copies — we now know that it’s about a hundred thousand — of the antibody that it’s going to later make. And, early in life, all the B-cells, whose antibody can see one of your own components, are deleted.
Matzinger covers the B-cell with her hand.
So, the only ones that are left are the B-cells that can see foreign.
Matzinger lifts her hand and draws a long, wavy shape next to the killer cell, representing a virus. Arms reach out from the killer cell toward pieces of virus eminating from the virus itself.
Killer cells are the same. They don’t recognize viruses purely, but what they recognize are virus components. Pieces displayed on the surface of infected cells. And he said that those killers are also — the ones that can see self, like your own kidney, your own heart — are deleted early in life, and the only ones that are left are the ones that see foreign.
That was ‘53. That got a Nobel Prize, and it lasted until 1969, when Peter Bretscher and Mel Cohn, at the Salk Institute in San Diego, modified it. And they modified it by adding another cell and another signal.
Matzinger adds the number 1 to the inside of the B-cell. She draws a wavy virus near the B-cell.
The signal that Burnet had suggested, we call Signal 1, boringly. And that is that, when a B-cell sees, say, a virus, the binding of its antibodies gives it a Signal 1.
And Burnet said that was enough to get it to divide and start making antibody. Peter Bretscher and Mel Cohn said, “No, no. This isn’t going to work.” And the reason they said that was that, in the mid-60s, it was discovered that, when B-cells do see a virus — flu, HIV, whatever — not only do they divide to make a small army and start pumping out 2,000 antibody molecules per second, they also mutate. In fact, they hypermutate. They mutate a lot. And we think the reason is that a B-cell that has very weak binding to a virus, when it mutates, will generate a few B-cells that have very high binding — strong binding — and therefore better antibodies.
But, in the process of all that mutation, you can imagine that some B-cells would now mutate and be able to see self. So, if the immune system discriminates self from non-self, how can you do that if you remove all your non-self reactive B-cells but then, by hypermutation, creates some more?
So, what they suggested was that it takes two cells to create an immune response. If you have two cells, both seeing the same virus, and both mutate, it will be a rare day when they both mutate to see the same self-protein. So, if you require two cells to make an immune response, you would very rarely have autoimmunity.
Actually, they wanted to require three. But, the problem with that is that the frequency of B-cells for any one foreign protein is about one in a million. So, if you needed three cells to start an immune response, you would need 10^9 — which is a million times a million, times a million — B-cells to make an immune response. We don’t have that many cells. So they settled on two, all right?
So, they invented — this is how immunological theory happens, hey? — they invented another cell that they called a “helper,” that we we, today, call a “T helper.”
Matzinger draws a new circle labelled “T helper.”
And they suggested that the T helper needs to see the same foreign body as the B-cell…
Matzinger gestures to the virus near the B-cell. She draws a line from the T-helper to the B-cell and labels it with the number 2 and the word “help.”
… and that, in fact, if the B-cell sees, let’s say, the virus and gets Signal 1 without a second signal from the helper, Signal 2, that they called “help,” the B-cell would die.
Matzinger covers the T helper cell with her hand.
So, a B-cell that mutates to become autoreactive, if it doesn’t get help, will die.
Matzinger lifts her hand to uncover the T helper cell and draws pieces of virus on the surface of the B cell.
If there is a helper that can see that same foreign protein, you will get an immune response. In fact, we know, now, that that’s true, and that, in fact, the B-cell takes in the virus, breaks it up into little pieces, displays it on the surface in special schlepper molecules called MHC…
Matzinger draws arms reaching from the T helper toward the virus presented on the surface of the B-cell.
… and that is seen by the T helper, that gives the T helper Signal 1, it then gets activated, gives the B-cell Signal 2, and voila! You get an immune response, okay?
That was 1969. That lasted until 1973.
The video jumps forward to another point in Matzinger’s presentation.
In 1973, back in Australia, Lafferty and Cunningham modified the model, yet again, by adding another cell and another signal. And the problem they were dealing with was that T helpers seemed to be more responsive to the molecules of another human than they are to the molecules of a mouse, or a dog, or a monkey. And yet, a mouse, a dog, or a monkey ought to be more foreign to a human helper than another human. So, there seemed to be some kind of species-specificity going on, here.
And so, to deal with that, they added another cell and another signal.
Matzinger draws a wavy shape near the T helper, labelled as “APC.” She draws pieces of a foreign body on the surface of the APC. The T helper reaches toward the APC, and Matzinger connects the APC to the T helper with a line that she labels “2-costimulation.”
They added what they called an “accessory cell,” and today we call an “antigen-presenting cell,” “macrophage,” or “dendritic cell.” And they suggested that the helper cell itself needed a second signal, and that, without that second signal, it couldn’t make a response. And that it got the second signal from the accessory cell. So, what they said was, the macrophage — or the dendritic cell; the APC — breaks up the foreign body into little pieces, displays it on its surface where it can be seen by the helper, that gives it Signal 1; if it also gets a second signal that they called “co-stimulation,” you get an immune response.
Now, that was 1973. What happened for the next 13 years was that “help” was studied. You could get NIH grants to study help. I bet you could get MRC grants to study help. And presenting cells were totally ignored, until 1986, when Jenkins and Schwartz, at the National Institutes of Health in DC, found by accident that, if you allowed the presenting cells to degrade the virus and put its pieces on the surface, and then you fixed it into a little billiard ball, it would not stimulate an immune response. You needed an active, happy, presenting cell in order to give the second signal, co-stimulation.
That was ‘86. And, one then has the question, why did it take 13 years? After 1986, presenting cells and co-stimulation became a huge topic in immunology. I would say a third of the immunology labs around the planet studied it. Why did it take 13 years?
Here’s the reason. We scientists have our own failings, right? If something doesn’t fit into your model, you ignore it. And co-stimulation didn’t fit into the self-non-self model, and here’s why. If you think that the immune system functions by discriminating self from non-self, cells that are specific, meaning they can see one thing, are useful. The helper cell can be the cell that governs immunity, because you can remove the helpers that see self. The presenting cell doesn’t discriminate between self and non-self. It picks up anything. They clean up wounds, they pick up anything. They pick up viruses, bacteria, fungus, allergens, everything. The presenting cell cannot discriminate self from non-self. So, if that’s the cell that starts an immune response, how does the immune system recognize self from non-self?
So, it was ignored for 13 years. And then it was rediscovered. And then immunologists had a problem. And then, for three more years, they had a problem, and then Charlie Janeway solved it, sort of, by suggesting that the presenting cell itself can discriminate self from non-self in a funny way. That bacteria, not viruses, bacteria, are so evolutionarily distant from humans that we could have genetically-encoded, selected over time, receptors for bacterial proteins.
The most important thing he said was the presenting cell is normally off.
The video jumps forward to another point in Matzinger’s presentation.
You turn it on when it sees a bacterium — an evolutionarily distant non-self — and that was cute, because it put co-stimulation, which is only made by activated APCs, back under the control of a self-non-self signal.
But Ephraim and I said to Charlie, “But, wait a minute, Charlie, the immune system rejects transplants. Most well done heart transplants are not covered in bacteria, right? It can cause autoimmunity, and it can sometimes reject a tumor. So, there no bacteria there. How does the immune system work against those things?” And Charlie said, “Oh, no problem. Transplants are a modern invention. We didn’t evolve to deal with those. And tumors and autoimmunity tend to kill you late enough in life that you’ve already had your kids, and so evolution doesn’t care.”
And we said, “You know what? If you want to make a model that describes what the immune system does, you have to describe what it does. Not just what you think it evolved to do, but what it does. And it does reject transplants, and it does give us autoimmunity, and it can reject a tumor. So how do you deal with that?”
And so, what we did is we followed tradition. We added another cell and another signal, and this is it. There are no more cells you can add, all right?
Matzinger continues to write on the overhead, but the projection does not appear on screen.
Because what we brought into this conversation was every tissue of the body. And we said, if you have a tissue made up of cells, and the cells are healthy, and happy, or if they die a normal death, and get scavenged by their neighbors — which is what happens when cells die a normal death — everything’s fine. But, should a cell get virus-infected, or damaged and blown open, it will release alarm signals. And that those alarm signals are what the presenting cells are really waiting for.
And that that’s what starts an immune response. It’s the alarm signals from damaged tissue, not the recognition of foreignness. The bacterium comes in and does no damage, you don’t respond. We have plenty of bacteria in our guts, our skin, etc, that we don’t make immune responses to. They’re not doing damage.
So, here we follow tradition. We’ve added another cell and another signal, and it turns out that if you take that one more step with me, you fall off a cliff. And you stand in a different place. And you look at the immune system from a different point of view. And when you look at the immune system from that point of view, it turns out that you can explain almost everything it gets right and almost everything it gets wrong.
So, here’s a list. We’re not gonna have time to do the whole list.
Matzinger refers to a per-prepared list she has placed on the overhead projector.
You can explain why you don’t reject yourself at puberty, why lactation is not a problem, why fetuses are not rejected — and we’ll come back to that — why transplants are rejected, why tumors are not rejected, why you get graft-versus-host disease, why there are parasites like Filaria, which is the worm that makes elephantiasis. It lives in the lymphatic vessels in intimate contact with the immune system, and, usually, you don’t make an immune response.
And it doesn’t explain why you get allergy or asthma, and that doesn’t bother me. And, the reason it doesn’t bother me is that allergy and asthma are a problem in a different universe of question. What do I mean by that? Here’s what I mean. So, the immune system has to deal with two questions: 1) do I respond or not, when faced with something? That’s what we’re talking about here. Do I respond or not? The second question, once you decide to respond, what kind of response do I make? How do I know to make the right response to fight a worm, or the right response to fight a virus? And allergy and asthma are a problem in that question. Allergens are dangerous. Der p 1, which is the main allergen in house dust mite, is a protease that attacks the surface epithelium of the lung, and also attacks B-cells. Bee venom is not innocuous. So, the danger model says that the reason an allergen is an allergen is that it either is dangerous itself, is packaged with something that’s dangerous, or mimics an endogenous alarm signal. Why some people make IgE to allergens and other people make IgG, and are not allergic, is a question of the second question, which is what kind of response do I make? What kind, rather than whether.
So, let’s stay with whether, because we don’t have time to do everything. But the short answer is, you don’t reject your fetuses because they don’t look dangerous. You do reject transplants because they look dangerous. And you don’t reject tumors because they don’t look dangerous. That’s Immunology 101-A as seen from the danger model.
Now, one minute, when I teach this to grammar school kids, I give them the following scenario. I say, if you think about the body as a community, the self-non-self model suggested that the white blood cells that are part of your immune system are like cops — American cops — they go around shooting any foreigner they meet, and they define a foreigner as anybody they hadn’t already met by the time they finished high school. The danger model says that, no, no. If you think of the body as a community, right, the immune system is more like firemen. They sit in their fire houses playing cards until somebody rings an alarm. And it doesn’t matter if the alarm is wrung by a member of the community or a traveling salesman, which is not allowed in the other community, or an immigrant — they only respond when there’s an alarm. And, unlike the cops, they have more than one way of responding. If it’s a cat up a tree, they bring a different truck from if it’s a three alarm fire.
So, that’s the difference between the self-non-self model and the danger model. I hope you enjoyed the talk. Thank you very much.
applause followed by music
Note on video accessibility: For a fully keyboard-accessible alternative to this video, view it in Chrome or on any Android or iOS device, view it in Firefox with the YouTube ALL HTML5 add-on installed, or disable Flash in Internet Explorer.