Mapping relative firing frequency distributions during wake onto firing phase relationships during NREM sleep.

A) MODEL DATA: Calculation of mean phase of firing relative to oscillatory signal generated by inhibitory population; peaks in the inhibitory signal are used as the starting points of the phase calculation. B) MODEL DATA: Normalized phase of firing in NREM versus normalized wake frequency of excitatory neurons reveals that the neurons firing with the highest frequency align with an earlier phase of the excitatory network population whereas slower firing neurons align with later phases. C-F) Analysis of NREM firing asymmetry reveals enhanced wake frequency-dependent temporal relationships between neurons after learning. C) EXPERIMENTAL DATA: Formation of a pairwise firing asymmetry matrix of CA1 neurons’ activity recorded from a representative mouse during post-CFC NREM. Rows and columns have been arranged by wake frequency; color denotes pairwise firing asymmetry as defined in the Methods section. D) MODEL DATA: Weighted average of differences in distributions (see inset for a representative example scenario; black—baseline, red—post) of pairwise Z-scores for models with type 2 dynamics and learning (yellow) and with type 2 dynamics and no learning (black). E) EXPERIMENTAL DATA: Weighted average of differences in distributions (see inset in bottom panel for example of representative mouse; black—baseline, red—post) of pairwise Z-scores for NREM of CFC mice (top, yellow) and Sham mice (bottom, black), compared to differences in distributions calculated on random spike data following CFC (red, top) or Sham (blue, bottom) bursting architecture. F) EXPERIMENTAL DATA: Significance (Z-score) of global NREM network asymmetry vs. that of global wake asymmetry calculated for individual CFC (yellow) and sham (black) mice. Strong asymmetry is observed in NREM sleep but not in wake. Red dashed lines in (D, E) represent the significance cutoff of |Z|≥2. Error bars denote SE of repeated simulations (D) or SE across all experimental animals (E). * p < 0.05, ** p < 0.001 using a two-sided t-test.