[@niethard_sleep-stage-specific_2016]

Background

##

Sleep

{ width=60% }

##

Sleep Rhythms

• Slow-Wave Sleep (SWS)
  • Slow Waves/”Delta” waves: 0.5-4 Hz
  • Sleep spindles: 10~15 Hz
• Rapid-Eye Movement (REM) Sleep
  • Theta waves: 4-7 Hz

Results: Dorsal PY Activity

##

Results: Dorsal PY Activity

{ width=60% }

##

Results: Dorsal PY Activity

• (Show Movie 1)

##

Results: Dorsal PY Activity

{ width=70% }

##

Results: Dorsal PY Activity

##

Results: Dorsal PY Activity

• Main point: activity was maximal in wake, reduced in SWS, and, contrary to expectations, minimal in REM

Results: Deep/Superficial Injections

##

Results: Superficial Injections

{ width=70% }

##

Results: Deep Injections

{ width=70% }

##

Results: Deep/Superficial Injections

• Main point: the same decreasing activity level through wake->SWS->REM was seen in both superficial and deep layers of sensorimotor cortex.

Results: PV-/SOM-Cre vs Total Activity

##

Results: PV-Cre Mouse

{ width=70% }

##

Results: PV-Cre vs Total Activity

{ width=70% }

##

Results: PV-/SOM-Cre vs Total Activity

{ width=60% }

##

Results: PV-/SOM-Cre vs Total Activity

“These data demonstrate the specific activation of PV-INs during REM sleep, which is accompanied by the suppression of both pyramidal neural and SOM-IN activity…[This] suggests that the shift in the overall excitation/inhibition balance toward predominant inhibition during REM sleep is conveyed by a subset of REM sleep-active PV-INs.”

##

The elephant in the room: REM should be MORE active than SWS!

##

Results: PV-/SOM-Cre vs Total Activity

{ width=60% }

##

Results: PV-/SOM-Cre vs Total Activity

• They blame selection bias: “neurons that are active during the wake period might be more likely to be selected for extracellular recordings…and only those neurons might also show higher activity during REM sleep than SWS.”
• “Overall, these data indicate that neurons with higher activity during wake also show higher activity during REM sleep, compared to SWS” – this includes unlabeled cells, probably including PYs!

Results: Exc/Inh Co-activation

##

Results: Exc/Inh Co-activation

{ width=80% }

##

Results: Exc/Inh Co-activation

• They claim this is accounted for by lower firing rates in REM because the rank correlation coefficients for the most active 20% of PY cells also showed significantly reduced correlations in SWS and REM compared to wake (data not shown).
• “PV-INs and SOM-INs tend to be co-activated with pyramidal cell during wake, [but] during sleep…activity of inhibitory and excitatory cells becomes dissociated.”

Conclusions

##

Conclusions

• They found global cortical activity decreases progressively from wake, to SWS, to REM.
  • This was the case for PY and SOM-IN cells in general, and was the case for cells in both superficial and deep layers.
• “PV-INs with the highest activity during wakefulness also showed highest activity during REM sleep, and the same relationship was observed for PY cells and SOM-INs, although less pronounced.”
  • Therefore, it’s possible that prior REM studies indicating greater activity than SWS are biased in favor of these 80~95% percentile-active cells

##

Conclusions

• They theorize that both
  • The balance of excitation/inhibition shifts towards inhibition in sleep, particularly in REM
  • Inhibitory cells become much more “dissociated” from excitatory activity in sleep, particularly in REM

References

##