Robust perisomatic GABAergic self-innervation inhibits basket cells in the human and mouse supragranular neocortex

Abstract

Inhibitory autapses are self-innervating synaptic connections in GABAergic interneurons in the brain. Autapses in neocortical layers have not been systematically investigated, and their function in different mammalian species and specific interneuron types is poorly known. We investigated GABAergic parvalbumin-expressing basket cells (pvBCs) in layer 2/3 (L2/3) in human neocortical tissue resected in deep-brain surgery, and in mice as control. Most pvBCs showed robust GABAAR-mediated self-innervation in both species, but autapses were rare in nonfast-spiking GABAergic interneurons. Light- and electron microscopy analyses revealed pvBC axons innervating their own soma and proximal dendrites. GABAergic self-inhibition conductance was similar in human and mouse pvBCs and comparable to that of synapses from pvBCs to other L2/3 neurons. Autaptic conductance prolonged somatic inhibition in pvBCs after a spike and inhibited repetitive firing. Perisomatic autaptic inhibition is common in both human and mouse pvBCs of supragranular neocortex, where they efficiently control discharge of the pvBCs.

Somatic HCN channels augment and speed up GABAergic basket cell input-output function in human neocortex

“…Firstly, we speculate that in general there is evolutionary pressure toward faster and higher fidelity signaling in neuronal pathways 22 . Secondly, many human neurons including the pv+BCs exhibit highly resistive cell membrane and a relatively large cell capacitance, which together result in slow passive membrane time constant and EPSP-to-spike transformation in cell soma 19, 28, 31 . This was observed here in the experiments blocking HCN channels.…”

“…Although single-cell RNA sequencing studies denote relatively conserved gene expression patterns among GABAergic inhibitory neurons of the mammalian neocortex 24, 25 , analogous neuronal types such as pv+ GABAergic interneurons show inter-species differences in terms of neurite arborization 26 , ion channel expression 27 , as well as temporal dimensions of basal synaptic transmission 14, 28 . In general, pv+ cortical interneurons are characterized by rapid action potentials and a high-frequency firing capacity across various experimental species 10, 29, 30, 31, 32 , but action potential firing and intrinsic electrical properties of the cells differ quantitatively between species 28, 31, 33 . However, it remains unknown how these affect EPSP-spike transformation and input-to-output function of the most numerous neocortical inhibitory interneuronal type, the pv+ basket cell (BC) 34, 35 in human.…”

Influence of Autapses on Synchronization in Neural Networks With Chemical Synapses

“…Although, autaptic connections are anatomically present in vivo and in the neocortex, their functions are not completely understood (Bacci et al, 2003). Experimental and theoretical studies on excitatory and inhibitory autapses have been carried out (Tamás et al, 1997;Saada-Madar et al, 2012;Suga et al, 2014;Szegedi et al, 2020) and the results have demonstrated that autaptic connections play a significant role in normal and abnormal brain dynamics (Wyart et al, 2005;Valente et al, 2016;Wang et al, 2017;Yao et al, 2019). The effects of autapses on neural dynamics were studied for single neurons (Heng-Tong and Yong, 2015; Herrmann and Klaus, 2004;Jia, 2018;Kim, 2019) and for neural networks (HuiXin et al, 2014).…”

“…It was also found that they promote burst firing patterns (Wiles et al, 2017;Ke et al, 2019). The inhibitory autapses contribute to a negative feedback (Bacci et al, 2003;Zhao and Gu, 2017) and to the reduction of neural excitability (Bekkers, 2003;Qin et al, 2014;Szegedi et al, 2020). Guo et al (2016) analyzed chemical and electrical autapses in the regulation of irregular neural firing.…”

Probing the structure-function relationship with neural networks constructed by solving a system of linear equations

“…A total of 30 networks were generated for each target FR. Mean ± SD are shown, n = 30. b, correlation between pairs of integration neurons as a function of the scaling factor and Harvey 1999) but see 17,18 , we were interested in imposing such constrains onto the w 308 obtained by the algorithm. To do this we followed a projected gradient descent (PGD) approach 19 , 309 taking advantage of the fact that the loss function , which encloses the structural constrains, is 310 a linear function with respect to the synaptic weights, and that the matrix w can be changed 311 without changing the stimulus-response mapping (see Methods).…”