Yehezkel Ben-Ari and his colleagues at the Institut de Neurobiologie de la Méditerranée (Inmed) in Marseille, France, showed several years ago that neurons recorded in newly born animals have elevated intracellular chloride, leading to paradoxical excitatory actions of the principal adult inhibitory transmitter gamma-aminobutyric acid (GABA)1. They showed an abrupt and brief decrease in intracellular chloride in central neurons recorded immediately after birth2. This shift is associated with an abrupt excitatory-to-inhibitory shift of the actions of GABA that is mediated by the hormone oxytocin, which also triggers labor.
This shift is important, as it exerts neuroprotective and analgesic actions on the newborn’s brain, implying that the newborn brain is more vulnerable when this mechanism does not operate. More recently, Ben-Ari’s team at Neurochlore and Inmed showed that this shift is absent in cortical neurons recorded from two animal models of autism3. One model is produced by administering the epilepsy drug valproate to the mother during pregnancy (shown in people to increase the risk of having a child with autism); the other is a genetic model of fragile X syndrome. In these models, GABA remains ‘immature’ as it excites instead of inhibiting neurons. As a result, neurons are overexcited during delivery and birth, thereby aggravating the deleterious effects of insults.
These observations suggest that restoring intracellular chloride levels might be a therapeutic strategy to restore GABA inhibition and attenuate the deleterious actions of elevated chloride. One way of doing that is to use a diuretic that lowers intracellular levels of chloride, thereby restoring efficient GABAergic inhibition. Ben-Ari and his team discovered that administration of the diuretic bumetanide to pregnant rodents during a restricted period shortly before and during delivery attenuates the behavioral and electrical features of autism in her pups, including when they become adults3,4. The action during and shortly after delivery restores intracellular levels weeks after the administration and is sufficient to exert long-term positive consequences.
This study provides experimental support to the clinical trials conducted by Eric Lemonnier, Ben-Ari and their colleagues using bumetanide in a double-blind, randomized study in which they showed that the diuretic attenuates the severity of autism5.
Collectively, these investigations raise fundamental issues in relation to environmental insults, notably during delivery, in the pathogenesis of autism. The group is pursuing their investigations both in a clinical direction (with a large multicenter trial) and with experimental investigations on the pathogenesis of autism.