NUMBER OF SPIKES: A PROPER METRIC FOR PARALLEL FIBER PATTERNS RECOGNITION BY A PURKINJE CELL

Abstract

The cerebellum receives a variety of inputs and it accomplishes information processing based on its internal rules. Purkinje cells (PCs) play a major role in the information processing of the cerebellum. These cells can learn input patterns from parallel fiber (PF) by long-term depression (LTD) in PF to PC synapses, but it is not clear how PCs encode information in their firing activity. Previous studies reported different mechanisms for information processing in PCs, and some studies mentioned that PCs encode weak and strong PF input patterns using different mechanisms. Thus, it remains a controversial issue how PCs encode PF patterns in their output. Therefore, in this study, pattern recognition in PCs was explored using a multi-compartmental model of a PC. Simulation results indicated that the PC responded to PF input patterns with a short burst followed by a quiescent period, and a proper metric for patterns recognition could be the number of spikes in the burst. So, learned patterns by LTD as well as novel patterns could be discriminated based on the number of spikes in the burst, and there were no different mechanisms of information processing for weak PF input patterns versus strong patterns.