P187 Climbing fiber impact on human and mice Purkinje cell spines
Stefano Masoli*1, Egidio D'Angelo1,2
1 Department of Brain and Behavioral Science, University of Pavia, Pavia, Italy 2Digital Neuroscience Center, IRCCS Mondino Foundation, Pavia, Italy
*Email: stefano.masoli@unipv.it
Introduction
Purkinje cells (PC) are one of the most complex neurons of the nervous system and can integrate multiple inputs through their dendritic tree dotted by tens of thousands of dendritic spines. Two excitatory pathways make synapses with PC spines: one is transmitted by granule cells (GrC) through their ascending axons (aa) and parallel fibers (pf), and the second one by climbing fibers (cf) originating from the inferior olive nucleus. The impact of pfs activity on PCs was studied with a multi-compartmental model [1]. It was later improved with human and mice morphologies and dendritic spines [2]. The cfs impact on PCs is still highly debated, which prompted their study using the latest PC models with the most up-to-date experimental information. Methods
A mice and human PC models with active spines [2] were expanded with five ionic channels type and location based on immunohistochemical papers. Dendritic spines were also improved based on the latest experimental data [3]. AMPA and NMDA receptors were tuned to generate a fast paired pulse depression. The cf synapses were distributed on spines belonging to the territory between pfs and the aspiny trunks. The synaptic impact was tested with cfs alone at various frequencies and with pfs too. Because of the massive number of sections involved, the simulations were performed with 48 cores on an AMD Threadripper 7980X. The simulation environment was NEURON 8.2.4 [4] and Python 3.10.16. Results The models reproduced similar intrinsic and synaptic properties showed in previous PC models [1,2,5]. The stimulations were performed with burst composed of 3 spikes every 6 ms (180Hz). The number of spines required in the generation of a complex spike was estimated to be 600 in mice and 1500 in humans. With these numbers, the mouse model showed the typical complex spike shape. Instead, the human model could not generate such a response because of the three distinct trees. A distributed approach, with a single cf for each tree [6] showed results similar to the mouse. The synchronous activation of pfs and cfs showed localized calcium increase in the spines near the stimulation sites. DiscussionValidated multi-compartmental models built in Python/NEURON can allow the exploration of behaviours that are not yet available from experimental techniques. The complex spike recorded in the mouse model matched multiple published papers. Instead, human recordings of this response is not yet viable. The model showed that the calcium in each separate trunk required a single cf to generate correct responses.This was proposed by a recent paper [6] and with a single cf terminal for each main trunk, the simulations showed results in line with the mouse model. Activation of multiple cfs at the same, on the same human morphology, in connection with the burst pause behavior, can generate an extensive parameter space.
Acknowledgements This project/research received funding from the European Union’s Horizon 2020 Framework Programme for Research and Innovation under the Framework Partnership Agreement No. 650003 (HBP FPA).
