Henry Gee
Monday February 12, 2001
The Guardian

The potentially-poisonous Japanese fugu fish has achieved notoriety, at least among scientists who haven't eaten any, because it has a genome that can be best described as "concise". There is no "junk" DNA, no waste, no nonsense. You get exactly what it says on the tin. This makes its genome very easy to deal with in the laboratory: it is close to being the perfect genetic instruction set. Take all the genes you need to make an animal and no more, stir, and you'd get fugu. Now, most people would hardly rate the fugu fish as the acme of creation. If it were, it would be eating us, and not the other way round. But here is a paradox. The human genome probably does not contain significantly more genes than the fugu fish. What sets it apart is - and there is no more succinct way to put this - rubbish.

The human genome is more than 95% rubbish. Fewer than 5% of the 3.2bn As, Cs, Gs and Ts that make up the human genome are actually found in genes. It is more litter-strewn than any genome completely sequenced so far. It is believed to contain just under 31,780 genes, only about half as many again as found in the simple roundworm Caenorhabditis elegans (19,099 genes): yet in terms of bulk DNA content, the human genome is almost 30 times the size.A lot is just rubbish, plain and simple. But at least half the genome is rubbish of a special kind - transposable elements. These are small segments of DNA that show signs of having once been the genomes of independent entities. Although rather small, they often contain sequences that signal cellular machinery to transcribe them (that is, to switch them on). They may also contain genetic instructions for enzymes whose function is to make copies and insert the copies elsewhere in the genome. These transposable elements litter the human genome in their hundreds of thousands. Many contain genes for an enzyme called reverse transcriptase - essential for a transposable element to integrate itself into the host DNA.

The chilling part is that reverse transcriptase is a key feature of retroviruses such as HIV-1, the human immunodeficiency virus. Much of the genome itself - at least half its bulk - may have consisted of DNA that started out, perhaps millions of years ago, as independent viruses or virus-like entities. To make matters worse, hundreds of genes, containing instructions for at least 223 proteins, seem to have been imported directly from bacteria. Some are responsible for features of human metabolism otherwise hard to explain away as quirks of evolution - such as our ability to metabolise psychotropic drugs. Thus, monoamine oxidase is involved in metabolising alcohol.

If the import of bacterial genes for novel purposes (such as drug resistance) sounds disturbing and familiar, it should - this is precisely the thrust of much research into the genetic modification of organisms in agriculture or biotechnology.

So natural-born human beings are, indeed, genetically modified. Self-respecting eco-warriors should never let their children marry a human being, in case the population at large gets contaminated with exotic genes!One of the most common transposable elements in the human genome is called Alu - the genome is riddled with it. What the draft genome now shows quite clearly is that copies of Alu tend to cluster where there are genes. The density of genes in the genome varies, and where there are more genes, there are more copies of Alu. Nobody knows why, yet it is consistent with the idea that Alu has a positive benefit for genomes. To be extremely speculative, it could be that a host of very similar looking Alu sequences in gene-rich regions could facilitate the kind of gene-shuffling that peps up natural genetic variation, and with that, evolution. This ties in with the fact that human genes are, more than most, fragmented into a series of many exons, separated by small sections of rubbish called introns - rather like segments of a TV programme being punctuated by commercials.

The gene for the protein titin, for example, is divided into a record-breaking 178 exons, all of which must be patched together by the gene-reading machinery before the finished protein can be assembled. This fragmentation allows for alternative versions of proteins to be built from the same information, by shuffling exons around. Genomes with less fragmented genes may have a similar number of overall genes - but a smaller palette of ways to use this information. Transposable elements might have helped unlock the potential in the human genome, and could even have contributed to the fragmentation of genes in the first place (some introns are transposable elements by another name). This, at root, may explain why human beings are far more complex than roundworms or fruit flies. If it were not for trashy transposable elements such as Alu, it might have been more difficult to shuffle genes and parts of genes, creating alternative ways of reading the "same" genes. It is true that the human genome is mostly rubbish, but it explains what we are, and why we are who we are, and not lying on the slab in a sushi bar.

Deep Time by Henry Gee will be published shortly in paperback by Fourth Estate. He is a senior editor of Nature. Related articles