Could our highly developed immune system be the result of an ancient infection? Alka Agrawal reports.
What do we have in common with sharks? At first glance, not much. But we both have jaws - and it turns out that all jawed animals have other, less obvious similarities beneath the skin.
Like sharks, humans have a sophisticated immune system capable of recognising specific pathogens such as bacteria and viruses. It allows us to fight the continually changing array of disease-causing microbes or pathogens that challenge us, and it remembers those we have encountered before.
So what gives animals as diverse as sharks, humans, mice and frogs - jawed vertebrates - such a remarkable system? These animals generate millions of unique proteins, the best known being antibodies, that pick out pathogens. Yet the genome carries far too few genes to explain how so many different antibodies could be generated. So how are we making them? In the 1970s, Susumu Tonegawa and colleagues showed that we inherit antibody genes, but they are split into many pieces. Throughout life, two or three of these pieces are reshuffled and joined together - usually sloppily - to create the enormous range of custom-made antibodies that protect us against infection. The discovery of this genetic reshuffling, known as V(D)J recombination, won Tonegawa a Nobel prize in 1987.
Then a new mystery presented itself. This recombination occurs only in jawed vertebrates, which diverged from their jawless cousins over a period of about 50-75 million years, some 450 million years ago. How did such an elegant system arise so suddenly by the standards of geological time? Some clues were suggested in the 1970s. Bordering the antibody gene segments were DNA sequences known as recombination signals. They are common to jawed vertebrate species and presumably target recombination to the gene segments they sit by. Scientists noticed that these sequences looked like those found at the ends of mobile DNA elements known as transposons, otherwise called "jumping DNA".
The goal of a transposon is to remove itself from one location in the genome and insert itself into another, a process called transposition. Transposons are responsible for the spread of antibiotic resistance between bacteria, for example, while retroviruses such as the human immunodeficiency virus are also a kind of transposon. The human genome is riddled with their remnants.
This idea got another boost in 1990, when David Schatz and colleagues isolated the genes responsible for V(D)J recombination. They found that the pair - RAG1 and RAG2 - were very close together and had suspiciously compact structures like those of transposons and retroviruses, where being compact is essential for mobility. Later in that same decade, Martin Gellert and colleagues showed that the chemical mechanism used by the proteins produced by these genes to cut DNA is reminiscent of the enzymes that catalyse transposition and integration of retroviral DNA.
But despite the mounting evidence, no one had been able to show that the RAG proteins could produce a transposition reaction or cause DNA to move from one place to another. While I was working in Schatz's laboratory as a graduate student, he and I observed an unexpected result that proved to be the final nail in the coffin of the transposon origin of V(D)J recombination. The RAG proteins made an unexpected product from a piece of DNA that carried the recombination signals at its ends and resembled a transposon. It turned out that the product was the result of a transposition reaction - the final proof that the RAG proteins had retained the ability to catalyse transposition. But rather than jumping into another piece of DNA, it resulted from the transposon-like piece of DNA jumping into itself. The RAG proteins could also, it transpired, cause this transposon-like DNA to jump into another piece of DNA.
Transposons have usually been thought of as the ultimate form of "selfish DNA", which care only for their own survival and do nothing useful for the host. So the transposon origin of the RAG genes, and ultimately of V(D)J recombination, is a spectacular example to the contrary. About 450 million years ago, a transposon carrying the RAG1 and RAG2 genes, with recombination signals on its ends, jumped into the genome of an ancestor of all jawed vertebrates. It evolved into V(D)J recombination over the course of millions of years. This boosted the immune system, allowing specific recognition and memory of pathogens, and may even have allowed us to crawl out of the ocean and evolve into human beings.
Alka Agrawal received her PhD from Yale University in 1999 and was the grand prizewinner of the Amersham Pharmacia Biotech and Science Prize for Young Scientists in 2000.