GABA (γ-aminobutyric acid), the main inhibitory neurotransmitter in the vertebrate nervous system, has many modes of action. For one of its receptors GABAB — a heterodimer consisting of GABAB1 and GABAB2 subunits — functional diversity is achieved by different synaptic localization and physiological action of its subunit isoforms GABAB1a and GABAB1b, according to two studies published in Neuron.

To uncover the role of different GABAB1 versions in inhibitory neurotransmission, Bernhard Bettlers and colleagues used an ingenious approach to generate knockout mice in which GABAB1a and GABAB1b were inactivated one at a time. Electron microscopy studies showed that, in the hippocampus, GABAB1a was predominantly localized at glutamatergic terminals, whereas GABAB1b was mainly found at dendritic spines opposite to the glutamate release sites. This differential localization of GABAB1 isoforms correlated with their functional differences: at hippocampal CA3-to-CA1 synapses, GABAB1a assembled receptors that blocked presynaptic glutamate release, while GABAB1b was involved in postsynaptic inhibition of neuronal firing. Intriguingly, the constitutive absence of GABAB1a, but not GABAB1b, resulted in impaired synaptic plasticity and hippocampus-dependent memory formation. This suggests that different GABAB receptor compositions might mediate distinct neural functions.

In a companion paper, Perez-Garci and colleagues report similar results of distinct GABAB1 receptor functions in pyramidal neurons residing in layer 5 (L5) of the neocortex. These L5 neurons are innervated by numerous inhibitory inputs, including those from interneurons in L1. The researchers first established that extracellular stimulation of L1 inhibited the dendritic Ca2+ spikes in L5 pyramidal neurons. Further, the short-lasting component of this inhibitory effect was mediated by GABAA receptors, whereas the long-term counterpart was mediated by GABAB receptors. Using the knockout mice generated by Bettler et al., the researchers found that L5 pyramidal neurons from GABAB1a-deficient mice showed normal short- and long-lasting inhibitory components in response to L1 stimulation. By contrast, the absence of GABAB1b completely abolished the long-term inhibitory response of L5 pyramidal neurons, but the GABAA-mediated short-term inhibition was unaffected.

These two studies shed fresh light on the apparent contradiction between functional diversity and molecular simplicity of the GABAB receptor system. It will be interesting to test whether other GABA receptors deploy similar strategies to generate distinct activity under different physiological conditions.