Self & Other
3. The Specialists
Sometime in the early 1950s, a question started bothering a small group of biologists. The question went like this:
How can your body make antibodies for things it has never seen?
The puzzle was real. By the 1950s, immunologists had figured out, more or less, what antibodies were. They were Y-shaped proteins floating in the blood. They stuck to specific things — the protein on the surface of a measles virus, say, or the toxin produced by a tetanus bacterium — and once stuck, they marked those things for destruction. One antibody, one target. That was the deal.
The trouble was that you could trick the body into making antibodies against almost anything. Researchers in this era loved to do exactly that. They would take a totally artificial chemical compound — something that had never existed in nature, that no human cell or animal had ever encountered — and inject it into a rabbit. A few weeks later, the rabbit’s blood would contain antibodies that recognized that artificial compound with exquisite precision.
This was very strange.
Think about it from the rabbit’s point of view. Evolution gives you immune systems that can fight the threats your ancestors fought. A rabbit ancestor a million years ago presumably never met a synthetic chemical synthesized in a Cambridge laboratory in 1952. And yet there it is — antibodies, perfectly fitted, ready to go.
You could explain this if a rabbit had a specific gene for every possible antibody. But how many would that be? Researchers tried to estimate. The number you needed, to cover every possible molecule the immune system might be asked to recognize, came out at something like ten billion. Maybe a thousand times more.
A rabbit does not have ten billion genes. A rabbit has, roughly, twenty thousand genes — the same as you, the same as a mouse, the same order of magnitude as basically every animal on earth.
The math did not add up. There weren’t enough genes to write the answers in advance.
So how was the body doing it?
The Wild Idea
In 1957, an Australian immunologist named Frank Macfarlane Burnet proposed something so strange that, for a few years, most of his colleagues thought he had simply gone wrong somewhere.
Burnet’s idea was this. The body does not pre-write the antibodies. The body pre-writes the cells.
Imagine, he said, that very early in your development — before you’re even born — your body makes a vast number of immune cells. Billions of them. Each one has a slightly different receptor on its surface, generated more or less randomly. Most of them will never do anything in your entire life. They drift around your bloodstream and lymph nodes, waiting, holding their unique little receptors out to whatever passes by, like fishermen with billions of differently shaped hooks.
When some foreign molecule shows up — a virus, a bacterium, a synthetic chemical from a Cambridge lab — most of the receptors don’t fit. But somewhere, in that vast lottery of cells, a few will fit. Just a few. Maybe one in a hundred thousand. And that one cell, the one whose random receptor happens to match the new threat, gets activated. It starts dividing. Each daughter cell carries the same receptor. Within a week, you have millions of cells, all making antibodies against that one specific thing.
Burnet called this clonal selection. The body doesn’t design the right antibody when it sees a threat. The body has every possible antibody already, distributed across billions of cells, and the threat itself selects which cells get to multiply.
It’s the same principle, in miniature, as evolution by natural selection. The variation is generated randomly and in advance. The environment chooses the winners.
When Burnet first proposed this, he had almost no direct evidence. It was a guess. A spectacularly good guess, but a guess. Most senior immunologists found the whole thing hard to swallow. The idea that your body would make billions of cells, most of which would never be useful — that seemed, biologically, wasteful. Unbiological. Suspicious.
Burnet won the Nobel Prize in 1960 (technically for related work on immunological tolerance, which we’ll get to in Chapter 7). His clonal selection idea would take another decade or so to be fully accepted. The mechanism that made it work — the actual chemistry of how a body could generate billions of different receptors from twenty thousand genes — would have to wait until 1976.
DNA Scrambling
In 1976, a young Japanese researcher named Susumu Tonegawa was working at the Basel Institute for Immunology in Switzerland. He had a simple, audacious experimental plan: he was going to compare the DNA of an antibody-making cell to the DNA of, say, a liver cell, and see if they were the same.
This was, for the time, a strange thing to do. Every biologist learned in school that the DNA in every cell of your body is identical. (Almost identical, anyway — there are small caveats for sperm and eggs.) Liver cells use some of that DNA. Brain cells use other parts of it. But the DNA itself, the underlying text, was supposed to be the same in all of them.
Tonegawa was not so sure. He had a hunch — based on Burnet’s idea, and on some other clues — that whatever made the antibody system so flexible might involve actually changing the DNA inside antibody-producing cells. Permanently. In a way no one had ever documented in any cell, anywhere.