Forward genetics identifies a novel sleep mutant with sleep state inertia and REM sleep deficits

Abstract

Switches between global sleep and wakefulness states are believed to be dictated by top-down influences arising from subcortical nuclei. Using forward genetics and in vivo electrophysiology, we identified a recessive mouse mutant line characterized by a substantially reduced propensity to transition between wake and sleep states with an especially pronounced deficit in initiating rapid eye movement (REM) sleep episodes. The causative mutation, an Ile102Asn substitution in the synaptic vesicular protein, VAMP2, was associated with morphological synaptic changes and specific behavioral deficits, while in vitro electrophysiological investigations with fluorescence imaging revealed a markedly diminished probability of vesicular release in mutants. Our data show that global shifts in the synaptic efficiency across brain-wide networks leads to an altered probability of vigilance state transitions, possibly as a result of an altered excitability balance within local circuits controlling sleep-wake architecture.

Somatic genetics analysis of sleep in adult mice

“…Moreover, the semi-automated sleep staging software only has ∼90% accuracy and requires intensive efforts to manually correct the annotation of several days of EEG/EMG data per mouse. On the other hand, a number of non-invasive sleep monitoring systems have been developed and utilized for mouse sleep analysis, including the infrared video recording system (Banks et al, 2020; Fisher et al, 2012; Pack et al, 2007), the piezoelectric system tracking mouse movement (Flores et al, 2007; Yaghouby et al, 2016) and the plethysmography system monitoring the respiration of mouse (Sunagawa et al, 2016).…”

“…Secondly, to determine whether the mouse is active or not, the algorithm calculates a high confidence difference score for the mouse between “t” and “t+m” frames by integrating the network prediction score (predict), the mouse mask area (Intersection of Union, IoU), the center of mass of the mask (activity), and the gray information (color) within the detected mask. Finally, SleepV defines the sleep state as ≥ 40- s of continuous immobility similar to previously reported video-based sleep monitoring systems (Banks et al, 2020; Fisher et al, 2012; Pack et al, 2007).…”

“…Moreover, there was a 95-99% epoch-by-epoch agreement of sleep/wake staging between the two methods, although SleepV cannot distinguish between NREMS and REMS (Figure 1F). The accuracy of our SleepV system was significantly higher than the 88-94% accuracy of previously reported video-based sleep monitoring systems (Banks et al, 2020;Fisher et al, 2012;Pack et al, 2007). Next, we used SleepV to record the sleep/wake cycles of fifty-nine C57BL/6J adult mice for three consecutive days.…”

“…Although SleepV could not distinguish between NREMS and REMS, there was a remarkable 95-99% epoch-byepoch agreement between the two methods, with ~96% in sensitivity, specificity and accuracy of sleep/wake staging by SleepV (Figure 2C and 2D). The mean accuracy (96.4%) of SleepV is significantly higher than that (92-94%) of previous reported video-based sleep monitoring systems (Banks et al, 2020;Fisher et al, 2012;Pack et al, 2007). Next, we used SleepV to record the sleep/wake cycles of fifty-nine C57BL/6J adult mice for three consecutive days.…”

Machine learning approaches to identify sleep genes

“…Many key features of sleep are conserved from invertebrates to vertebrates (13). Large-scale forward genetics screens in flies (14)(15)(16) and mice (17,18) have identified several genes whose alteration impacted sleep regulations. The twoprocess model (19,20) proposed that both circadian clocks and sleep homeostasis drive the sleepwake cycle.…”