Feature: A modern history of immune tolerance

This feature appeared in the November/December 2010 issue of Australian Life Scientist. To subscribe to the magazine, go here.

In 1960, Australian scientist, Sir Frank Macfarlane Burnet, was awarded the Nobel Prize in Physiology or Medicine, alongside Peter Brian Medawar, for their discovery of acquired immunological tolerance, a finding the magnitude of which cannot be overstated.

At last year’s Australasian Society for Immunology conference in Perth, a modern day pioneer in immunology, Professor Chris Goodnow, recent recipient of the Ramaciotti Medal, gave the Burnet Oration. Fittingly, he devoted it to our understanding of immunological tolerance, looking at what we know and what we don’t know about the phenomenon 50 years on from Burnet’s Nobel.

A key part of Burnet’s theory was that the immune system learns to recognise its own tissue as ‘self’, and so does not attack it. It can also recognise and fight off foreign antigens such as those on bacteria and viruses. The first part – a process Burnet called clonal selection – argued that rare cells in the blood that were reactive to a virus would proliferate and increase in frequency so they could mount a much better and rapid response against that virus next time around.

“Many beautiful experiments in the 50 years since make it absolutely clear and uncontroversial that this is how we acquire immunity,” says Goodnow. “No question about that. And, of course, Sir Gus Nossal, one of Australia’s other leading lights, really provided the first definitive evidence for that part of Burnet’s theory during his PhD.”

The second part of Burnet’s theory, which he called clonal deletion, was actually first articulated in a 1949 paper with Frank Fenner, another great Australian virologist. The theory said that rare cells with the potential to react against a body’s own tissue would commit suicide and be deleted from the repertoire before birth, and so the only cells circulating post-development of the immune system should not react to self, but might react with foreign antigens.

“The evidence that tolerance could be learned was made concrete by a series of experiments by Medawar in the late 1950s, demonstrating that if genetically dissimilar cells were put into a newborn mouse, that mouse would then happily accept skin grafts for the rest of its life that would normally be rejected.” Medawar thus did the experiments that validated Burnet’s concept, culminating in their shared Nobel prize.

It is this second part of Burnet’s theory – self-reactive clones nobly bowing out during development – that is still, in some people’s minds, controversial, even 50 years on, says Goodnow. His oration will discuss the theory and subsequent experimental evidence from then to now, and put it in context with his own findings. Essentially, he wants to examine how certain the immunology field is that the ‘forbidden clones’, as Burnet called them, are deleted before birth, or are they dealt with in some other way?

---PB---

Controversy

Goodnow explains how everyone loved the concept in 1960, when Burnet and Medawar won the prize, but by the mid-1960s some studies were showing what appeared to be forbidden clones circulating in the blood of adults. “The controversy got thicker in the ‘70s and people really started to become less and less convinced by Burnet’s idea of clonal deletion.”

Around that time, Nossal published a paper adding a new branch to Burnet’s theory. Backed up by some very good experimental evidence, he argued that instead of being physically deleted, these clones became anergic, so they were floating around but unable to mount an immune response, like “zombies”, says Goodnow. “Nossal’s results in a series of papers on clonal anergy could explain some of the reports of self antigen-binding cells circulating in healthy people.

“Then, other people in the ‘70s started to come up with more Byzantine ideas, not as simple as Burnet’s,” says Goodnow. One of these was that a very complex computer-like network of interactions is set up among lymphocytes based on idiotypes to prevent self responses. “These cells would come to a decision by committee as to whether to respond. It was very different from Burnet’s idea, which was quite Darwinian: all about survival of the fittest and killing off the forbidden clones.”

Another model that also emerged around that time for how tolerance works is referred to by Goodnow as the police-state scenario, whereby law-enforcement cells in the immune system would respond to self in a way that would stop forbidden clones from responding.

“This suppressor-cell theory became really popular in the late ‘70s and early ‘80s and generated a whole experimental industry purporting to demonstrate the existence of these suppressor cells in ever-greater detail. But in the end it all turned out to be basically rubbish, using up a lot of research money and time in the process.”

So by 1983, Burnet’s idea of clonal deletion had virtually been discarded by the immunology community. Nossal was about the only voice in the wilderness still arguing that tolerance might be done in a Darwinian way as opposed to the socialist or police-state way.

“The trouble is that when the theory is being pushed by these hyper-intelligent people, but the tools are not there to do the definitive tests, science gets bogged down in a lot of political stuff and it can be very counter-productive. And this is really what happened in the ‘70s,” Goodnow says. “Without the tools to get an absolutely clear result, you are working in this twilight area, and then people start to get comfortable working in that dim light and their vision starts to get a bit fuzzy.”

---PB---

Genetics to the rescue

Around this time in the 1980s, molecular biology came on the scene like a big broom that basically cleaned out the tolerance closet. “Genetics got rid of all the woolly twilight rubbish, all those artefacts of low-quality non-specific reagents and marginal effects, like suppressor cells, and it also gave us the tools to get a more specific readout,” Goodnow says.

“A basic issue with Burnet’s idea was that testing it definitively was like asking, ‘was there a needle in the haystack and has it gone now?’, because we knew that the frequencies of these forbidden clones would be incredibly low.” Genetic technologies provided the tools to increase the frequencies of very specific needles in the haystacks, so that they could be accurately counted and their fate traced.

It was about this point that Goodnow entered the tolerance story, at the time when people worked out how to make transgenic mice. “It was immediately clear that with these new tools, we could start to address some of these questions about tolerance.

We could engineer mice such that we could track needles, we could control their frequency and we could mark them in very precise ways to ask things like ‘if these needles are self-reactive what happens to them?’” This was a huge step forward in immunology and particularly in settling the clonal deletion controversy.

And what did Goodnow find in the haystack? “These marked cells get anergised, as Nossal had proposed, and then they are deleted as Burnet proposed, and this was a really stunning result.” Then genetics also revealed a gene called FOXP3, which was missing in a mouse strain that had terrible autoimmune disease, revealing a kind of enforcer cell (now called a natural T regulatory cell) that is self-reactive: it knows what self is and stops any other cells becoming dangerous to self.

So, it seems that the last 20 years of mouse transgenics and genetics has proved that the Darwinian mechanism of Burnet and Nossal exists, but then layered over the top is the police-state idea of a suppressor-type cell.

Of course, the next question for the field was what happens in people: is it the same as in a genetically engineered mouse? “And that is pretty well where we are now, trying to go back from the transgenic universe to the real-life human situation. One of the ways to do this is still genetics and, of course, that is my driving thing.”

About 10 years ago, Goodnow was intrigued with some work being done in the US and Europe on a gene associated with a very rare but devastating disease in people called autoimmune polyendocrine syndrome. “In this condition, the immune system starts attacking endocrine organs in a very curious fashion, and it was clear that whatever was defective would tell us something very fundamental about how your system learns not to attack your own organs.”

Goodnow’s group studied transgenic mice carrying the causative AIRE gene mutation and found something spectacular: these mice showed no clonal deletion of T cells in the thymus during development. Thus, AIRE appeared to support thymic deletion, and when defective, the good old forbidden clones are allowed to leave the thymus and go out to where they can recognise and respond to self antigens in peripheral organs, leading to disease.

“For me, that was extremely exciting because we now had a bridge between the mouse experiments and the human setting where we can’t measure clonal deletion directly, so the genetic support for the human condition,” says Goodnow. Thus genetics, in particular mouse genetics, has allowed us to say that Burnet was right.

---PB---

Twists and turns

However, at the International Congress of Immunology in Japan in August, Goodnow listened to past mentor, Professor Mark Davis present some startling new results. Davis originally cloned the T cell receptor gene, and has since developed a very powerful way to measure the frequency of T cells that recognise different antigens. That is, needles in haystacks.

“Mark presented some preliminary data at the meeting that went right back to the ‘60s. Using this very sophisticated assay to detect the frequency of forbidden clones in blood, he argued that clonal deletion does not occur in humans and, in fact, he ended his talk by saying that maybe the mouse had misled us and that Burnet was wrong. And so the pendulum swings!”

So Goodnow’s ASI fable of Burnet and the forbidden clones has two threads woven through the great tapestry of scientific process. The first is that doing a definitive experiment is always going to be very, very hard, and doing them in humans is harder still.

The second, and perhaps deeper issue according to Goodnow “is that underlying tension between experimental and descriptive work, and it comes back to what was so great about Burnet and why he had such a lasting impact on Australian science. He was an absolute believer in the critical role of the experiment,” he says.

“I see this theme as a very big issue facing biomedical science right now, that these incredibly important and time-tested principles of science are being eroded by the new kind of science. We have such amazing technology available for collecting complex biomedical data, and sometimes we might be tempted to think that having enough descriptive data will substitute for a proper experiment and answer all our questions.

“In this age of complexity, it is even more critical to stick to the old mantra: ‘every conclusion has to be based on good experimental data’. And the immune tolerance field is a great example of why this is an absolute truth. As a country, we cannot fall into that fuzzy-vision trap.

“Every scientist and clinician we train must understand that description is the start of the scientific process. But, in the end, it means nothing without the experiment.” In the words of the Royal Society motto, Nullius in verba, take no-body’s word for it!

This feature appeared in the November/December 2010 issue of Australian Life Scientist. To subscribe to the magazine, go here.