Learn about nVoke  

A Neural Mechanism for Switching between Sleep States

Posted by Kevin Chen, PhD - 03.12.2018

What is a dream? This is a question that has been asked by people for thousands of years. Dreams have played crucial roles in affecting individual’s behaviors and even affecting the human history in a “supernatural” manner. The biological understanding of dreams was not available until the last century. With the contributions by many scientists, especially Eugene Aserinsky, Nathaniel Kleitman, and William C. Dement, a sleep stage called rapid eye movement (REM) sleep was identified and was regarded as the major stage in which vivid dreams occur.

Thereafter, the regulatory mechanism of REM sleep became a major topic in sleep research. In the following two decades after REM sleep was identified, Michel Jouvet and others performed a series of lesion studies to identify the brain region responsible for generating REM sleep. Their work demonstrated that neurons in the pontine tegmentum (dorsolateral pons) enable and regulate REM sleep. Afterwards, other REM sleep regulatory nuclei were identified in the hypothalamus and other brain regions.

It was later proposed by other investigators that a flip-flop switch could be used to describe the circuit mechanism regulating REM sleep through the mutually inhibitory interaction between REM-active and REM-inactive neurons. Based on this model, each side of this model inhibits the other side. If one side’s activities are slightly higher than the other side, it will become a forward feedback loop and quickly turn the activities of the other side off.

Even though the classical electronic flip-flop switch is composed of a single electronic component at each side, which allows the circuit to switch immediately, the brain circuits are composed of a large number of neurons within different nuclei, making it difficult to use this model to simply explain the switching mechanism. Also, the long-range projection of each group of neurons may decrease the efficiency of this switching mechanism. Third, the discrete distribution of each type of neurons makes this mechanism difficult to work. The ideal REM/NREM switch, if it exists, should be able to overcome these weaknesses.

Theoretically, if scientists can identify a brain region with a small number of inhibitory neurons that are composed of both REM-active and REM-inactive neurons with distinct projection targets (projecting to REM-active and REM-inactive regions individually) and mutually inhibiting one another, this region would possibly function as the center for switching on and off REM sleep. The traditional neuroscience methods were not able to dissect circuits like this due to lack of tools to specifically manipulate and record neurons based on cell types and projection targets.


In a recent study, scientists at University of California, Berkeley, combined specific Cre mouse lines, novel viral tools and Inscopix nVista system to record the neuronal activities in a cell-type and projection-specific manner. Scientists expressed genetically-encoded calcium indicators (GCaMP) in the dorsomedial hypothalamus (DMH) of transgenic mice and performed in vivo microendoscopic calcium imaging with the nVista system to understand the correlation between DMH neuron activity and brain states. They found that there are two distinct subtypes of inhibitory neurons that express both GABA and a neuropeptide galanin within the DMH: one group has the highest activity during REM sleep, and the other group has the lowest activity during REM sleep. Using a retrograde viral tracing tool (rAAV-retro), they expressed GCaMP into selective DMH neurons based on their projections. To determine the causal functions of these neurons, they retrogradely expressed a light-gated ion channel (channelrhodopsin, ChR2) to activate each group of DMH neurons by light. They found that preoptic-area-projecting neurons have the lowest activity during REM sleep and activating them switched REM sleep to non-REM sleep, while the brainstem-projecting neurons show the opposite effect. The spatial intermingling between these two groups raises the intriguing possibility of local reciprocal inhibition, and their physical proximity could greatly improve the efficiency of the neuronal circuit in regulating the switch between REM and NREM sleep.


A Hypothalamic Switch for REM and Non-REM Sleep. Kai-Siang Chen, Min Xu, Zhe Zhang, Wei-Cheng Chang, Thomas Gaj, David V. Schaffer, Yang Dan. Neuron 2018


Recent Posts

Twitter for Academics 101

read more

Must-Read Neural Circuit Papers in April 2018

read more

Podcast: Krishna Shenoy and Brain-Machine Interfaces

read more