Coordination Mechanism Between LFP Network And Spike Network During Working Memory

Objective Working memory(WM) refers to the short-term maintenance and manipulation of information necessary for higher cognitive functions. There is a growing interest in the investigation of working memory mechanism from the view of neural signal network connectivity. The aim of this study is to investigate the mechanism of coordination between LFP network and spike network in rat prefrontal cortex and is expected to provide new insights for the investigation of working memory mechanism.Methods: 1. Y-maze working memory task Laboratory animals: 6 Male SD rats, weighted 300-350 g. Rats were trained in a Y-maze working memory task until the rats can reach 85% correct rate on two consecutive days. 2. Acquisition of LFPs and spikes during working memory Multi-channel LFPs and spikes were obtained from rat prefrontal cortex(PFC) through the implanted microelectrode array during a Y-maze working memory task. 3. Time-frequency analysis for working memory related frequency bands Time-frequency analysis was applied to analyze the power changes of WM related frequency bands(theta band) in the LFPs. 4. Construction of LFPs network Functional connectivity in the LFPs(theta band) was estimated by directed transfer function(DTF). LFPs network was constructed based on the connectivity metrics. 5. Construction of spikes network Functional connectivity in the spikes was estimated by maximum likelihood estimation(MLE). Spikes network was constructed based on the connectivity metrics. 6. Canonical correlation between LFPs network and spike network Canonical correlation(CC) between LFPs network and spikes network was quantified via canonical correlation analysis(CCA). The dynamic changes of CC was estimated during the working memory task to investigate the mechanism of network coordination in working memory.Results: 1. Power changes of LFPs during the Y-maze task The theta band power was higher during the working memory task. The theta power peaked before the rats arrived the reference point(RP). 2. Dynamic changes of LFPs network during the working memory task The connectivity metrics in LFPs varied during the working memory task. The mean of matrix elements(DTF) increased and peaked before the rats arrived the RP. 3. Dynamic changes of spikes network during the working memory task The connectivity metrics in spikes varied during the working memory task. The mean of matrix elements(GC) increased and peaked before the rats arrived the RP. 4. Dynamic changes of Canonical correlation(CC) between LFPs network and spikes network during the working memory task The Canonical correlation(CC) between LFPs network and spikes network varied during the working memory task. The CC-curve increased and peaked before the RP. The peak values of CC from 6 rats were respectively 0.7265 ± 0.0106、0.6998 ± 0.0129、0.7150 ± 0.0207、0.6620 ± 0.0126、0.6602 ± 0.0088、0.6231 ± 0.0121(mean ± SEM). The CC value in working memory(0.8436 ± 0.0102) showed a significant higher level than that in resting state(0.3291 ± 0.0121)(t-test,p<0.01).Conclusions: 1. The power of theta band was higher during the working memory task. 2. The connectivity in the LFPs network and spikes network enhanced greatly during the working memory task. The connectivity in the two networks reached the maximum before the rats arrived the RP. 3. The Canonical correlation between the LFP network and spike network enhanced during the working memory task and reached the maximum before the RP.It may provide a potential network mechanism for working memory: the connectivity in the LFP network and spike network as well as the coordination between the two networks enhance during working memory.