Spike-timing-dependent plasticity enhances reliability of predictive spike coincidences

Tomoki Fukai and Katsunori Kitano

Tamagawa University & CREST, Japan Science and Technology (JST)

Pairs of premotor (PM) and primary motor (MI) neurons show significant increases of coincident spikes at those times when monkey is expecting a behavioral event (Riehle et al., 1997). Typically, the longer the preparatory period for a motor response, the higher the temporal precision of coincidences. These results provide evidence that synchrony has predictive power. To ellucidate the underlying neuronal mechanism, we model the predictive synchronous firing in a paradigm similar to classical conditioning. In this model, the firing sequence that is losely time-locked to a cue signal with 20- to 30-ms precision represents temporal information, which should be accessible to those PM/MI neurons. We examine whether the loosely timed spikes can be associated with the target events by spike-timing-dependent plasticity (STDP) to evoke the statistically significant number of spike coincidences at the expected times of the events. We find the cooperation of activity regulation function of STDP with coincidence detection essential to the emergence of predictive power. Our model accounts for the modulations of the temporal precision based on the synaptic self-organizing mechanism. Therefore, we propose that this phenomenon is independent of higher cognitive processes. We also show that the spike coincidences are easily reorganized when the to-be-predicted times of events are changed.