I've mentioned my general interest in how the brain adapts or responds to variables in the local environment. One thing I think we can all identify with, at some point in our lives, is how slow we can start to feel under some extended less-than-ideal sleep circumstances. That feeling kinda sucks, doesn't it?
It's not just you feeling generally crappy and tired- your brain is actually making some changes in response to the lack of sleep. So with that I'll introduce my first literature blog post, in which I review a paper that I think is cool.
Sustained sleep fragmentation results in delayed changes in hippocampal-dependent cognitive function associated with reduced dentate gyrus neurogenesis.
N. Sportiche, N. Suntsova, M. Methippara, T. Bashir, B. Mitrani, R. Szymusiak, and D. McGinty.
Why did I pick this paper?
I really like studies that investigate not only behavior but also associated changes in biochemical goings-on in the brain. And the hippocampus is just cool. How can you not like the hippocampus?
Overall, what did they do in this study?
These researchers looked at the effects of sleep fragmentation on cognitive function by testing sleep-fragmented rats in a maze task, and then looking at how many new neurons they made in the dentate gyrus of the hippocampus during a phase of the sleep fragmentation treatment.
How did they do it?
They implanted electrodes to measure rats' sleep-wake cycles. Using the sleep-wake data, they started a treadmill to wake up the rats anytime they had been asleep for 30 continuous seconds, for 12 days. Two weeks after the rats finished the sleep fragmentation, they spent 5 days training in the Barnes maze. This is a large circular platform, with eight holes evenly spaced on the outer edge. One of the eight holes is the escape, and the rat should use visual cues around the room to navigate its way to the escape chute. (It's actually identical to and visually indistinguishable from all the other holes in the maze, I just can't resist the phrase "escape chute.") After the 5 days of Barnes maze training, they were given two days on a reversal task in the Barnes maze. The reversal task tests the rat's ability to adapt to changing locations of the escape chute. The rat had to un-learn the original escape position, and learn the new escape position.
Once the researchers observed this learning behavior, they counted new neurons that were born in the dentate gyrus during the fragmented sleep treatment.
What did they find out?
Rats in the sleep fragmentation group (black triangles) got shorter lengths of non-REM sleep at a given time, but they also got *more* fragments of sleep. This is looking like some disruption from normal.
Sleep fragmentation changes search strategy in the Barnes maze
First, the authors provide examples of search strategies that rats might have used to find the escape chute in the Barnes maze after they learn where the chute is.
1. Direct- the rat goes directly from the starting point to the correct escape point.
2. Quadrant- the rat goes to the correct quadrant of the maze, and then finds the hole that leads to the escape chute.
3. Error direct- The rat goes directly to the wrong hole- oops! - but then goes immediately to the correct hole next.
4. Serial- the rat checks every hole until it encounters the escape. This score and higher ones involve the rat not using spatial cues to navigate its way to the escape, it's just going on the premise that there is an escape and eventually it will run across it.
5. Serial-Random- the rat tries some serial, and some random wandering around the maze, to eventually run across the escape.
6. Random- the rat just wanders around until it runs across the escape.
Alright. Now while they show these nice pictures in the above figure and label them 1-6 with a decreasing sophistication of search strategy, they turn around and score these in reverse order. That was a touch confusing when I first read it. So when you look at this next graph, think backwards from the above figure. Lower score is worse performance, higher score is better. Lower = random, Higher = Direct.
The black vertical line divides the regular Barnes maze task on days 1-5 and the reversal task on days 6-7. SF rats (black triangles in the left-side graphs) are different in some ways from the control groups. They made significantly more Random-strategy attempts (top left). These scores are averaged on the top right.
The Technique score, based on the Low=Random to High=Direct scale, isn't particularly different on a given day. But when they average the scores from the entire testing session, they see a significant reduction in overall score in the SF rats.
Overall, this means that the SF rats used less sophisticated methods to find their way out of the maze, relying on the fact that they would eventually run into the exit rather than learning how to navigate using the spatial cues to take the direct way out.
As a final observation, the authors looked at the number of neurons that were made in the hippocampus on days 4-5 of the sleep fragmentation treatment. This is done using a chemical that integrates into the DNA of new cells, we're going to call it BrdU and leave it at that. BrdU labeling will show up only in cells that were made during BrdU treatment- not before, and not after. And those new labeled cells do have to stay alive between the time they are made and the time they are counted, too!
We're interested in newborn neurons in the hippocampus for many reasons, not the least of which is the vital role the hippocampus plays in cognitive performance in tasks like the Barnes maze. We have some hints that sleep disruptions might reduce neurogenesis, the process of making these new neurons, and that reduced numbers of newborn neurons might reduce our ability to adapt to and learn new things (like un-learning the first Barnes maze escape, and learning a new one in the reversal task).
It's pretty clear from this figure that the counts of newborn neurons are quite a bit lower in the SF rats than in their control groups. They show a couple of very nice example photos, a bit of double-staining to demonstrate that they are staining newborn cells and that those cells are indeed neurons (and not other interesting non-neuron type cells).
We learned some interesting things in this paper. First, rats seem to compensate for fragmented sleep by getting more fragments. I suspect people are the same, but that's just speculation. Nonetheless, the rats with sleep fragmentation tended to use a less "sophisticated" strategy to get out of the Barnes maze. Rather than using spatial cues around the room to learn which hole they should use to escape directly, they tended to use the serial and random type strategies, just checking every hole and exiting the maze when they ran across the escape. The sleep fragmented rats did have a lower number of newly made neurons in the dentate gyrus of the hippocampus compared to the control groups, as well.
Maybe you've had a few sleep-fragmented weeks- or months, or years. This doesn't mean your brain is toast, though some days it totally feels like it. In fact, the brain can be really great about bouncing back from such things, and to bring your spirits up I'll see what I can find on that topic next. But, I'm just gonna warn ya here, if you're sleep fragmented, you might reserve showing off your awesome maze escape performance skillz until you've gotten a little more rest.
Sportiche, N., Suntsova, N., Methippara, M., Bashir, T., Mitrani, B., Szymusiak, R., & McGinty, D. (2010). Sustained sleep fragmentation results in delayed changes in hippocampal-dependent cognitive function associated with reduced dentate gyrus neurogenesis Neuroscience, 170 (1), 247-258 DOI: 10.1016/j.neuroscience.2010.06.038