Sunday, December 21, 2008

Dark matter and DIY genomics

[This is my column for the January 2009 issue of Prospect.]

Physicists’ understandable embarrassment that we don’t know what most of the universe is made of prompts an eagerness, verging on desperation, to identify the missing ingredients. Dark energy – the stuff apparently causing an acceleration of cosmic expansion – is currently a matter of mere speculation, but dark matter, which is thought to comprise around 85 percent of tangible material, is very much on the experimental agenda. This invisible substance is inferred on several grounds, especially that galaxies ought to fall apart without its gravitational influence. The favourite idea is that dark matter consists of unknown fundamental particles that barely interact with visible matter – hence its elusiveness.

One candidate is a particle predicted by theories that invoke extra dimensions of spacetime (beyond the familiar four). So there was much excitement at the recent suggestion that the signature of these particles has been detected in cosmic rays, which are electrically charged particles (mostly protons and electrons) that whiz through all of space. Cosmic rays can be detected when they collide with atoms in the Earth’s atmosphere. Some are probably produced in high-energy astrophysical environments such as supernovae and neutron stars, but their origins are poorly understood.

An international experiment called ATIC, which floats balloon-borne cosmic-ray detectors high over Antarctica, has found an unexpected excess of cosmic-ray electrons with high energies, which might be the debris of collisions between the hypothetical dark-matter particles. That’s the sexy interpretation. They might instead come from more conventional sources, although it’s not then clear whence this excess above the normal cosmic-ray background.

The matter is further complicated by an independent finding, from a detector called Milagro near Los Alamos in New Mexico, that high-energy cosmic-ray protons seem to be concentrated in a couple of bright patches in the sky. It’s not clear if the two results are related, but if the ATIC electrons come from the same source as the Milagro protons, that rules out dark matter, which is expected to produce no such patchiness. On the other hand, no other source is expected to do so either. It’s all very perplexing, but nonetheless a demonstration that cosmic rays, whose energies can exceed those of equivalent particles in Cern’s new Large Hadron Collider, offer an unparalleled natural resource for particle physicists.

A Californian biotech company is promising, within five years, to be able to sequence your entire personal genome while you wait. In under an hour, a doctor could deduce from a swab or blood sample all of your genetic predispositions to disease. At least, that’s the theory.

Pacific Biosciences in Menlo Park has developed a technique for replicating a piece of DNA in a form that contains fluorescent chemical markers attached to each ‘base’, the fundamental building blocks of genes. Each of the four types of base gets a differently coloured marker, and so the DNA sequence – the arrangement of bases along the strand – can be discerned as a string of fairy lights, using a microchip-based light sensor that can image individual molecules.

With a readout rate of about 4.7 bases per second, the method would currently take much longer than an hour to sequence all three billion bases of a human genome. And it is plagued by errors – mistakes about the ‘colour’ of the fluorescent markers – which might wrongly identify as many as one in five of the bases. But these are early days; the basic technology evidently works. The company hopes to start selling commercial products by 2010.

Faster genome sequencing should do wonders for our fundamental understanding of, say, the relationships between species and how these have evolved, or the role of genetic diversity in human populations. There’s no doubt that it would be valuable in medicine too – for example, potential drugs that are currently unusable because of genetically based side-effects in a minority of cases could be rescued by screening that identifies those at risk. But many researchers admit that the notion of a genome-centred ‘personalized medicine’ is easily over-hyped. Not all diseases have a genetic component, and those that do may involve complex, poorly understood interactions of many genes. Worse still, DIY sequencing kits could saddle people with genetic data that they don’t know how to interpret or deal with, as well as running into a legal morass about privacy and disclosure. At this rate, the technology is far ahead of the ethics.

Besides, it is becoming increasingly clear that the programme encoded in genes can be over-ridden: to put it crudely, an organism can ‘disobey’ its genes. There are now many examples of ‘epigenetic’ inheritance, in which phenotypic characteristics (hair colour, say, or susceptibility to certain diseases) can be manifested or suppressed despite a genetic imperative to the contrary (see Prospect May 2008). Commonly, epigenetic inheritance is induced by small strands of RNA, the intermediary between genes and the proteins they encode, which are acquired directly from a parent and can modify the effect of genes in the offspring.

An American team have now shown a new type of such behaviour, in which a rogue gene than can cause sterility in crossbreeds of wild and laboratory-bed fruit flies may be silenced by RNA molecules if the gene is maternally inherited, maintaining fertility in the offspring despite a ‘genetic’ sterility. Most strikingly, this effect may depend on the conditions in which the mothers are reared: warmth boosts the fertility of progeny. It’s not exactly inheritance of acquired characteristics, but is a reminder, amidst the impending Darwin celebrations, of how complicated the story of heredity has now become.

No comments: