PROJECT SUMMARY/ ABSTRACT
REM sleep is accompanied by dreams characterized by vivid visual experience, which evidently indicates
that our brain holds a generative model of the world. Converging studies suggest that we perceive the world
through such a generative model and the abnormal expression of the model could underlie some psychiatric
disorders. Thus, it is important to understand how our brain supports the generative model. The goal of this
project is to elucidate the neurophysiological basis for the generative model by focusing on the neuronal
activity during REM sleep, during which the brain’s generative model is detached from the external world.
In the mentored phase of the award, I first propose to investigate rapid eye movements as a readout of the
generative model during REM sleep. Rapid eye movements during REM sleep are proposed to represent
active sampling of the virtual visual environment of dreams. However, this hypothesis is poorly supported by
physiological evidence. To provide physiological evidence to test the hypothesis, I will focus on the activity of
the head direction (HD) cells, which code for the animals’ HD relative to the environment. The activity pattern
of HD cells during REM sleep is similar to what is observed during wake. This makes it possible to decode
virtual HD during REM sleep. I will test whether rapid eye movements during REM sleep can predict changes
in HD decoded from HD cells recorded in the anterodorsal thalamus. To achieve this goal, I will combine large
scale electrophysiological recordings, advanced decoding methods, and miniaturized eye tracking systems.
In the R00 phase, I propose to investigate the superior colliculus (SC) as the neuronal substrate for virtual
head orienting and rapid eye movements during REM sleep. The SC is an important hub for head and eye
orienting in awake animals. I will test whether spontaneous activity in the SC during REM sleep predicts
changes in virtual HD by simultaneously recording neuronal activity in the SC and in the anterodorsal thalamus
as well as monitoring rapid eye movements. I will further test the causal impact of the SC activity on virtual
head orienting and rapid eye movements by activating and inactivating the SC during REM sleep.
In summary, this project will provide new insights into the neurophysiological understanding of how the
generative model of the world and our interaction with it is organized during REM sleep. This will constitute the
critical step to understand how we see the world through the internal model and how the abnormal expression
of the model could underlie the pathogenesis of some psychiatric disorders such as schizophrenia.
The technical and scientific expertise that I will acquire during the training period of the award will be crucial
for setting the basis of research programs in my own independent laboratory focusing on the role of the SC in
orchestrating the generative model during REM sleep. In addition to this, intense career development training,
the guidance from the mentoring team, as well as the collaboration and the rich intellectual interaction in the
UCSF neuroscience community will ensure my successful transition into an independent investigator.