a cute lil sleep-deprived mouse

On Monday, I (small and afraid) attended a Biology seminar in a building I’d never been in (Fairchild). I found a seat in the last row and got settled—but alas! Who was this daunting and impressive woman seated in front of me, peeling an orange with a special peeling instrument? Could it be? Could it really be none other than Deborah Mowshowitz, queen of Biochemistry and polarizing pedagogy? (It was.)

Unfortunately I did not show up to stargaze, so I had to temper my awe as the presentation began. Professor Darcy Kelly introduced the speaker, Dr. Wen-Biao Gan of the Skirball Institute at the NYU School of Medicine. His work centers on how learning experiences and sleep affect synaptic plasticity in the brain. Dr. Gan received his PhD from Columbia in 1995, and in her introduction, Dr. Kelly asked, flustered, “was I your thesis advisor?”

Dr. Gan’s research involved training mice with an accelerated rotarod (sort of like a treadmill, but for mice!), and analyzing both their skill level over time and their formation of dendritic spines in layer 5 pyramidal neurons in the motor cortex. That learning tasks induce dendritic spine formation is not entirely revolutionary, but Dr. Gan discovered that different types of learning tasks induce spine formation on different sets of branches (that is, spines are segregated by motor task), and that this spine formation was related to spikes of calcium activity in the dendrites. Some dendrites show calcium spikes in response to forward rotarod training, and some show spikes in response to backward rotarod training, but only a negligible amount show spikes in response to both forward and backward training.

It is important to note that there was no increase in dendritic spines until about 24 to 48 hours after the mice were trained, hinting at the possibility that sleep plays a central role in solidifying motor learning neurologically. But how exactly does that work?

To figure this out, Dr. Gan had to deprive mice of sleep (honestly, same), and compare the neurons in their motor cortices to those of healthy mice. Dr. Gan found a strong correlation between sleep and dendritic spine formation, and even noticed that allowing the sleep-deprived mice to compensate by sleeping afterwards didn’t help: it seems that sleep needs to occur soon after the learning takes place in order for it to stick. Not only did the sleep-deprived mice not exhibit the neuronal changes that are supposed to accompany learning, their performance on the same motor tasks was significantly worse than the non-deprived mice later on.

Then, Dr. Gan decided to look into the specific roles of different types of sleep, by depriving some mice of REM sleep and others of non-REM sleep. The mice deprived of REM sleep did not show decreased spine formation, but the mice deprived of non-REM sleep did. Is REM sleep just utterly pointless? Well, okay, it turns out that REM sleep is actually crucial when it comes to motor learning, just not when it comes to the formation of dendritic spines. REM sleep is responsible for pruning these newly-formed spines (which makes sense, since REM sleep usually comes after non-REM sleep). Pruning is essential because it helps facilitate the formation of new spines induced by different motor tasks later on. In addition, mice with REM sleep showed persistent strengthening of some new spines, whereas mice deprived of REM sleep did not.

Essentially, Dr. Gan’s research showed that non-REM sleep is responsible for facilitating the formation of new dendritic spines after motor learning, while REM sleep is responsible for the pruning and strengthening of existing spines. Dr. Gan lost me when he started hypothesizing about the role of microglia in the last few minutes and in general, I was pretty confused amid all the spines and spikes and tdTomato (what is that and who named it and why?). However, Dr. Gan’s insight into the impact of sleep on learning was fascinating and even inspiring. After the seminar, I took a nap.

Image via Pixabay.