What causes autism is a mystery. One theory is that a phenomenon called the cellular-danger response lies at the root of it. The CDR makes cells put their ordinary activities on hold and instead switch on their defence systems, in reaction to high levels in the bloodstream of chemicals called purines.
These are important and widespread substances: ATP, a molecule that shuttles energy around cells, is a purine; so are half the "genetic letters" in DNA. Cells under viral attack tend to shed them.
Too many of them in the blood can thus be a signal of viral infection. In that case activating the CDR makes perfect sense. But studies have shown that people with autism (and also those with some other brain conditions, such as schizophrenia) often seem to have chronic CDR. The purine signal has somehow got stuck in the "on" position.
Why this happens is obscure. But it has occurred to Robert Naviaux of the University of California, San Diego, that once the signal is stuck in this way, chronic CDR might, by subverting the function of crucial brain cells, be the immediate cause of the symptoms of autism. In a series of experiments, the latest of which has just been published in Translational Psychiatry, he makes a plausible case that this is exactly what is happening--and he also illuminates a route to a possible treatment.
Dr Naviaux's experiments start by injecting pregnant mice with viral genes. This stimulates the CDR in both mother and fetus. The offspring of such pregnancies often show behaviors, such as fear of strangers, fear of novelty and poor co-ordination, that would be interpreted as autistic in people, and this procedure is thus used as a model of autism.
Dr Naviaux reasoned that, if chronic CDR is the cause of autism's symptoms, abolishing it should abolish the symptoms. He further reasoned that one way this might be done is with a drug that binds to a cell's purine receptors, so that they cannot react when purines come along. One such drug exists. It is called suramin, and is used to treat sleeping sickness. Dr Naviaux has therefore been trying it out on the model mice.
Once they are weaned, he puts these mice through a series of behavioral tests and compares the results with those of similar mice whose mothers were injected with saline rather than viral genes. One test measures whether they have trouble interacting with a stranger mouse. A second studies how physically co-ordinated they are. A third runs them through mazes to see if they have a preference for "sameness", as many autistic people do. At the age of six months, he then gives them either an injection of suramin or an injection of saline solution, and tests them again. Finally, five weeks after the injection, when the level of suramin in their bodies has fallen back to zero, he tests them once more.
The upshot, he has found, is that suramin treatment abolishes some of the animals' autistic symptoms. Mice which once showed an aversion to novelty, and considerable anxiety when forced to meet a stranger, show neither of these after treatment with the drug--though their co-ordination is not improved. In his latest series of experiments he has also looked at the drug's metabolic effects. He finds that suramin treatment returns to normal 17 of the 18 metabolic pathways that are noticeably different in autistic and non-autistic mice.
The effect is not permanent. The experimental mice's autistic behaviours come back once the suramin has disappeared from their bodies. But these are interesting results. They suggest CDR really is an important part of autism, and that it may be a tractable one. Dr Naviaux is not suggesting using suramin itself to treat people with autism.
Though appropriate for an acute, life-threatening illness, which sleeping sickness is, its side effects (the most serious of which is damage to the adrenal cortex) make it unsuitable for chronic prescription to someone whose life is not in danger. But if his work stands up to the scrutiny of others, then looking for another purine-receptor blocker would surely be a worthwhile endeavour.
Click here to subscribe to The Economist
