The transience of interneuron circuit diversity just “sped” up

Understanding the diversity of inhibitory neurotransmission is a central issue in neuroscience. GABA-mediated neurotransmission takes place across a broad time scale (Fig. 1). Fast GABA_A receptor-mediated phasic inhibition decays within milliseconds, whereas tonic GABA_A receptor-mediated inhibition persists for seconds and longer. Between these ends of the spectrum, there are the more enigmatic slow GABA_A-mediated phasic inhibition [GABA_A,slow (1)] and inhibition mediated by the GABA_B receptor. GABA_A,slow is poorly understood. The time course of decay is on a time scale of tens of milliseconds, intermediate to fast phasic inhibition and GABA_B-mediated inhibition. The work of Szabadics et al. (2) in a recent issue of PNAS sheds light on the mechanism of GABA_A,slow in cortical pyramidal cells (PCs) and also provides evidence for additional diversity of interneuron circuits in the cortex.

Fig. 1.

The time scale of GABA-mediated inhibition spans several orders of magnitude. GABA receptor diversity allows for a broad spectrum from fast phasic GABA_A inhibition to moderate slow phasic GABA_A and GABA_B inhibition and then to a persistent tonic GABA_A inhibition. In a recent issue of PNAS, Szabadics et al. (2) add the unique morphology of the neurogliaform–pyramidal cell (NGFC-PC) synapse as a contributing factor to diversity of inhibitory neurotransmission.

In 2003, Tamás et al. (3) began to fill in our understanding of inhibition that acts within the tens to hundreds of milliseconds range, setting the stage for Szabadics et al. (2). In this 2003 study, paired recordings demonstrated GABA_B-mediated inhibition in postsynaptic PCs after a single action potential in neurogliaform cells (NGFCs). NGFCs are multipolar cells with dendrites that branch close to the cell body and have a dense axonal plexus, and they are characterized by a late-spiking firing pattern (4, 5). GABA_B receptors are often located perisynaptically, and their activation after NGFC activation suggests spillover from the synapse after a single action potential. This was the first hint at the unique nature of the NGFC–PC synapse. The inhibitory postsynaptic current (IPSC) generated by NGFC stimulation had slow kinetics that were not due solely to the GABA_B component, and Szabadics et al. (2) identified this as the enigmatic GABA_A,slow (1). Mediators of tonic GABA_A inhibition, such as receptors containing the δ and α5 subunits, are located outside the synapse and are activated by spillover and ambient GABA. Szabadics et al. used modulators of extrasynaptic GABA_A receptors, in addition to alternations in GABA uptake and diffusion away from the synapse, to demonstrate a role for these tonic receptors in the NGFC-evoked GABA_A,slow. However, they were only a minor factor in the slow kinetics.

Classically, variation in kinetics of GABA_A-mediated inhibition is largely associated with the stochastic nature of postsynaptic GABA_A receptors that form different functional subtypes with regional and cellular specificity (6, 7). Receptor gating, unbinding, and desensitization kinetics of channels made by individual subunits contribute to current duration (7). Generally, neurons with faster IPSC decay kinetics usually express high levels of α1 subunits, as opposed to other α subunits (8). As a result, IPSCs recorded in different brain regions and at different times of development display different decay kinetics (8, 9). Interestingly the pharmacology of these NGFC–PC synapses is that of α1-containing GABA_A receptors, which are typically associated with “fast” GABA_A receptor-mediated neurotransmission in the cortex (8). Diversity is the rule in inhibition, and another source of variability is the GABA transient, which itself is affected by a variety of factors (10, 11). Through a series of clever experiments, including using the low-affinity competitive antagonist TPMPA [(1,2,5,6-tetrahydropyridine-4-yl)methyl phosphinic acid], Szabadics et al. (2) showed that the concentration of GABA is lower at the NGFC synapse than at the fast-spiking basket cell (FSBC) synapse but that this concentration difference likely does not mediate the slower kinetics. Szabadics et al. also ruled out asynchronous transmitter release or dendritic filtering as causes of the slow kinetics. It appears that it is instead the morphology of the NGFC synapse that provides a GABA transient that produces the slow kinetics. The authors note that the NGFC synapse junctional area is relatively small and that this small area, if coupled with a relatively large cleft distance between pre- and postsynaptic cell, would explain their many experimental observations and the slow kinetics. Neurotransmitter transients and the morphology of the synapse take over where GABA receptor subtype diversity leaves off to provide a full spectrum of duration to GABA-mediated inhibition.

The evidence for a slower time course of IPSCs by the GABA transient in the NGFC–PC synapse presented by Szabadics et al. (2) adds another level of complexity of the diversity of inhibitory neurotransmission. From this diversity stems a specialization of inhibitory function for processes including sensory processing, experience-dependent plasticity, and learning and memory (12). Specialized circuits or “GABA_A specific circuits” (Fig. 2) are composed of (i) a particular presynaptic inhibitory neuron [characterized by morphology, biochemistry, and action potential firing pattern (12)], (ii) the subcellular positioning of the presynaptic contact on the target neuron (soma, dendrite, or axon initial segment), and (iii) the subunit composition of GABA_A receptors in the synapse of the target neuron (13–15). The findings in Szabadics et al. show that in addition to subcellular location, the shape of the presynaptic terminal and the synaptic cleft are also critical. Therefore, this additional mechanism of slow phasic GABA_A receptor-mediated neurotransmission in the cortex just sped up our understanding of interneuron circuit specificity.

Fig. 2.

GABA_A-specific circuits in the CNS. The previous view of GABA_A-specific circuits consisted of two defined circuits. Circuit 1: FSBC synapsing onto pyramidal cell (PC) soma with α1-containing GABA_A receptors (15). This circuit may be important for somatic feedforward inhibition and critical period plasticity (16, 17), where fast and reliable α1-subunit-mediated IPSCs are required. Circuit 2: fast spiking chandelier cell (FSCC) synapsing onto axon initial segments of PCs with α2-containing receptors (13, 14). This circuit provides α2-mediated slow IPSC decay onto the initial segment and may be important for regulating neuronal firing (16). The data in Szabadics et al. (2) provide an additional circuit that takes into account the synapse structure. Circuit 3: late-spiking NGFC that synapses onto PC dendrites with α1-containing receptors. This circuit produces a slow IPSC in PCs that is dictated by the GABA transient because of the unique structure of the NGFC–PC synapse. Note that the presence of α1 subunits in this synapse does not determine IPSC kinetics but rather allows for a more faithful response to the transient.