The human genome gets more and more complicated
Published on The Economist, Oct. 15, 2009.
IT WAS, James Watson claimed, something even a monkey could do. Sequencing the human genome, that is. In truth, Dr Watson, co-discoverer of the double-helical structure of DNA back in the 1950s, had a point. Though a technical tour-de-force, the Human Genome Project was actually the sum of millions of small, repetitive actions by cleverly programmed robots. When it was complete, so the story went, humanity’s genes—the DNA code for all human proteins—would be laid bare and all would be light.
It didn’t quite work out like that. Knowing the protein-coding genes has been useful. It has provided a lexicon of proteins, including many previously unknown ones. What is needed, though, is a proper dictionary—an explanation of what the proteins mean as well as what they are. For that, you need to know how the genes’ activities are regulated in the 220 or so different types of cell a human body is made from. And that is the purpose of the American government’s Roadmap Epigenome Programme, results from which are published this week in Nature by Ryan Lister and Mattia Pelizzola of the Salk Institute in California, and their colleagues …
… Their first discovery was that the stem cells were more methylated than the lung cells—5.8% of cytosines, compared with 4.3%. Moreover, the difference was largely accounted for by something strange. Previous studies have shown that methylated cytosines are usually next to a letter called guanine (G). It is a common characteristic of the so-called promoter regions of genes, where transcription begins, that they contain long, repetitive sequences of alternating Cs and Gs. If these areas become methylated, it tends to suppress transcription of the gene in question. A quarter of the methylated cytosines in stem cells, however, were not followed by guanines. Nor were they found in the promoter regions of genes, but rather in the transcribed parts of the genes themselves. They also had the opposite effect from methylated cytosines found in promoter regions. The genes they occurred in tended to be transcribed more than usual, not less. In particular, a lot of genes involved in processing RNA were activated in the stem cells in this way.
One unexpected discovery made during the decade since the genome project was finished is that there are thousands of small genes whose RNA copies are not translated into proteins. Instead, the RNA acts in its own right. In plants, for example, it is one of the things that switches other genes on and off at their promoter sites. Whether it does so in mammals has yet to be established. But it might. In any case, unusual patterns of RNA processing in stem cells are something that will require further examination.
The complexities of methylation, then, are myriad—as are the complexities introduced by all these unexpected small genes. Reading the human genome in the first place may, indeed, have been work for mechanical monkeys. Interpreting the result will require the finest minds that humanity can muster. (full text).