Firing regulation of fast-spiking interneurons by autaptic inhibition

Fast-spiking (FS) interneurons in the brain are self-innervated by powerful inhibitory GABAergic autaptic connections. By computational modelling, we investigate how autaptic inhibition regulates the firing response of such interneurons. Our results indicate that autaptic inhibition both boosts the current threshold for action potential generation as well as modulates the input-output gain of FS interneurons. The autaptic transmission delay is identified as a key parameter that controls the firing patterns and determines multistability regions of FS interneurons. Furthermore, we observe that neuronal noise influences the firing regulation of FS interneurons by autaptic inhibition and extends their dynamic range for encoding inputs. Importantly, autaptic inhibition modulates noise-induced irregular firing of FS interneurons, such that coherent firing appears at an optimal autaptic inhibition level. Our result reveal the functional roles of autaptic inhibition in taming the firing dynamics of FS interneurons. In the mammalian brain, the majority of cortical inhibition is acquired by interneurons releasing gammaaminobutyric acid (GABA) on principal cells [1]. This GABA-mediated inhibition is believed to play functional roles in many important neuronal computations, such as balancing excitation [2–5], promoting neural oscillations [6–11] and gating multiple signals [12]. As the most prominent type of interneurons, cortical fast-spiking (FS) interneurons have been observed to display highly variable electrophysiological properties [13]. By integrating massive synaptic inputs, the FS interneurons process information and produce discrete trains of action potentials (so-called “spike trains”) of their own. Exploring the firing regulation of FS interneurons is therefore fundamental for the understanding of complicated inhibition-related neuronal computations in the brain. Moreover, a large proportion of FS interneurons in the neocortex and hippocampus have been identified to form self-feedbacks connections, termed as “autapses”, onto themselves [14, 15]. Previous electrophysiological recordings have revealed that most of FS interneuron autapses are mediated by GABAA receptors [16] and, importantly, their transmission is discovered to be strong and robust [17, 18]. These findings indicate that autaptic inhibition might have significant impacts on shaping the firing dynamics of FS interneurons. Such hypothesis has been confirmed by recent experimental and computational data, showing that the GABAergic autaptic inhibition might prevent repetitive firing, reduce firing rate and improve spike-timing precision [17–20]. However, so far it is still not completely established whether the autaptic inhibition participates into the modulations of input-output (IO) gain, firing pattern and firing regularity of FS interneurons. In this Letter, we address this question by considering a FS interneuron driven by both external applied and autaptic inhibition currents. The dynamics of the FS interneuron is simulated using the Wang-Buzsaki (WB) neuron [21]. As a paradigmatic model of the FS interneuron, the current balance equation of the WB neuron can be