Lab report

Erasing bad memories sounds like something from science fiction. Yet treatments exist and new research may improve them still further
May 3, 2009
Memories are made of this

Exposure therapy, in which bad memories or phobias are extinguished through repeated exposure to a similar but innocuous stimulus, sounds like a lazy science-fiction trope based on too much reading of A Clockwork Orange. But it seems to work, and has been used to treat post-traumatic stress disorder in war veterans. More surprising is the fact that certain drugs can have the same effect: a study in 2000 showed that rats given a stimulus to evoke an unpleasant association while receiving such a drug seem to lose the bad memory forever. The effect suggests that memories become more susceptible to erasure at the moment of recall.

There is a neurochemical basis to this. Memories become persistent by being, if you will, written in neural proteins. But these proteins are destroyed when the memory is recalled, and have to be made again to reconsolidate the memory. This process means that memories can not only be wiped but also modified during recall. In normal exposure therapies, the memory is merely suppressed by a competing state—an effect that can wear off.

The problem with current memory-erasing drugs is that they're toxic. Now a team from NYU has found a way to make exposure treatments permanent without drugs. They discovered that these treatments are more effective if administered during the "vulnerable" interval when a memory has been summoned by a prior lone stimulus and is in the process of being reconsolidated. Apparently this allows the memory to be edited to have a safe rather than a scary association. So far the effect has been seen only in rats, but without drugs, human testing should not face big hurdles.

Obama's space vision is pretty boring

The anxiety in the US space industry over Barack Obama's delay in appointing a head of Nasa seems to be about more than just the temporary lack of a helmsman. There's suspicion that dark plans are afoot, perhaps hinted in Obama's recent comment that "there's been a sense of drift to our space programme over the last several years." What doesn't he like? Perhaps it's the Constellation programme, which is developing new spacecraft to carry astronauts to the International Space Station or the moon. Constellation may have cost $44bn by the time its first rocket is sent to the space station in 2015. Many scientists feel it is industrial interests and public relations, rather than scientific inquiry, that have kept the manned space programme quids in for so long.

So far, Obama's space vision seems to prefer the boringly useful to the vacuous posturing of previous presidents. In a recent visit to Florida he said that he wants to review whether manned flights to the moon or to Mars should take precedence over "things like Hubble that yields us more information and better bang for the buck." That's music to the ears of many space scientists.

This new approach is also clear in Obama's plans for nuclear disarmament, surely one of the least fashionable issues around. Although he spiced up his initiative with talk of nuclear terrorism, his bottom line—that Iran and North Korea can hardly be expected to engage with frameworks for non-proliferation until today's nuclear powers take their own responsibilities seriously—is something western leaders have ignored for decades. And by setting the goal of "a world without nuclear weapons," he has raised the stakes in a way that could force the hand of Russia and China. Yet, like a shift of the space agenda, it could also put American jobs on the line—indeed, perhaps much the same ones.

Robot makes scientific discovery

The announcement that robots have made scientific discoveries in a British genetics lab prompted some sceptics to say this is not news—they used to be called postdocs. That's not bare cynicism; most research is uninspired, routine slog. Countless publications now stem from simply cranking the handle, while automation of repetitive lab procedures is widespread. But what distinguished the new study, by a team largely at the University of Wales, is that the robotics system (called Adam) could generate and test hypotheses, in this case about the functions of yeast genes.

It's a sensible step, since much new understanding, particularly in genetics and cell biology, relies on incremental advances in which the next question is often an obvious extension of the last. Machines can simply be more exhaustive in exploring the possibilities. For as Alan Turing pointed out in 1950, it is wrong to think that "as soon as a fact is presented to a mind, all consequences of that fact spring into the mind simultaneously with it."

That fallacy is even more pertinent in related work by two scientists at Princeton, who have devised a computer algorithm for uncovering meaningful "natural laws" from experimental data. Or, in other words, to uncover mathematical relationships that encode the physical principles at play, without being sidetracked by random noise or reducing everything to a trivial level. A system like this might, given the observations of the 16th-century Danish astronomer Tycho Brahe, have deduced from them Newton's inverse-square law of gravity. Such things don't simply fall out of the data. It may well turn out that, in genetics, we need something like this more than we do a plodding Adam, because nothing as unifying as Newton-like laws of the genome are yet in sight.