August 8, 2011 | 2
It’s always nice to see a paper on a subject I’m interested in appear in a high level journal. It makes you feel all warm and fuzzy inside, knowing that steps in the field are being made. And it’s even better when you know the first author! Yup, Jason of Functional Neurogenesis has a new paper on stress, depression, and hippocampal neurogenesis.
I’ve been wanting to do a somewhat in-depth post on the neurogenesis theory of depression for a while now, and this paper is the perfect opportunity to give some background and show where the field may be heading. And to do this, I’m going to let you in on two secrets (or at least, things that many people don’t necessarily know about the brain):
1) The low serotonin theory of depression is probably wrong
2) Your brain can make new neurons
Snyder et al. “Adult hippocampal neurogenesis buffers stress responses and depressive behavior” Nature, 2011.
I have actually written before on the serotonin theory of depression, and why it’s probably wrong. For a long time, the idea was that drugs that increased serotonin in the brain helped to relieve depressive symptoms, and so this must mean that depression was caused by low levels of serotonin in the brain (and this is still how you’ll see depression presented in a lot of antidepressant commercials). But it’s not that simple. Just because increasing something in the brain helps to fix a problem, it doesn’t mean that the problem is a LACK of that something. Headaches are not caused by lack of aspirin.
And decreasing serotonin doesn’t appear to CAUSE depression, though it can make you irritable. Not only that, some new drugs on the market such as Wellbutrin treat depression without touching serotonin levels. There’s no proof that people with depression have lower levels of serotonin than people without. So the serotonin theory is on the way out.
But the drugs still work (at least in a subset of patients). It turns out that these drugs which increase serotonin in the brain ALSO increase neurogenesis: the birth of new neurons. We used to think that we were born with all the neurons we would ever have, and that from birth on it was a smooth downward curve as we killed off our neurons with alcohol, head banging, and falling into things. But now we know that that isn’t true. In fact, your brain is capable of birthing new neurons and does so for most of your life.
And antidepressant drugs can INCREASE this rate of neuron birth, paprticularly in the hippocampus, and these increases produce antidepressant-like effects in animal models. Not only that, other things, like stress, can DECREASE hippocampal neuron birth rates. And this is where it begins to maybe come together. Because life stress is a major stimulus for depressive episodes in humans. Not only that, some patients with major depressive disorder show a dysregulation of the hypothalamic-pituitary axis, which controls stress responses. And if stress decreases hippocampal neurogenesis, and antidepressants increase neurogenesis…could it be that stress and hippocampal neurogenesis modulate each other and control depressive behavior?
To figure out how hippocampal neurogenesis and stress may control each other, the authors created a strain of mice v-TK. Basically, they inserted a gene in the mice that, when it’s activated (and they can activate it selectively by feeding the mice doxycycline in their drinking water) makes dividing cells (like the ones in neurogenesis) sensitive to an antibiotic. So all you have to do then is give the mice doxycycline to express the gene, and then give them the antibiotic to knock out all the dividing cells. Poof. No neurogenesis. This is a good model because, unlike many other knockout mice, this is inducible, so you don’t have the problems associated with a mouse growing up missing a gene (and missing a gene that allows cells to divide is one of those things that can make it very difficult to, you know, grow up). And it works pretty well.
You can see above the normal mice on the left, and the TK mice on the right. After getting the doxycycline and the antibiotic, the normal mice still show neurogenesis (the pretty turquoise dots) while the TK mice do not.
So now they had these mice, and they wanted to see how they responded to stress when neurogenesis was gone. This is a nice new angle to go at, several other papers have looked at how hippocampal neurogenesis responds to stress, now we’re seeing how stress responds when you take away the neurogenesis. In this case, they gave the mice restraint stress for 30 minutes (this amounts to putting the mouse in a plastic tube so it can’t turn around, and waiting). They then looked at the levels of the stress hormone corticosterone in the normal and TK mice.
You can see above that at time 0 (just after the stress) the mice showed no difference in their corticosterone levels. But 30 minutes AFTER the stress, the TK mice showed higher levels, meaning they were not recovering from the stress as quickly as normal mice. They then tried to see if the effect would continue if they stressed the mice repeatedly for a long time, but even after 17 days of stress, the TK mice showed elevated corticosterone. Without neurogenesis, then, TK mice can’t control their corticosterone levels as well. They even got this effect in another model of decreased neurogenesis (X-irradiation of the dentate gyrus, the part of the hippocampus responsible for neurogenesis).
So we’ve got increased stress responses, and we’ve got decreased hippocampal neurogenesis, but how does this impact behavior? To look at this, the authors ran a series of tests which evaluate depressive-like behavior in mice. We can’t ever SAY that a mouse is depressed (can’t put it on a couch and ask about its self-esteem), but we can say that some of these behaviors are associated with decreased feelings of pleasure (anhedonia) in mice, and some are associated with despair, and all of them are improved when you give mice antidepressants.
What you can see up here are the results of the behavioral tasks. The top set is a test called novelty suppressed feeding. You take a hungry mouse, put him in a scary new environment, put a food pellet in the center, and see how long it takes him to give in. Mice with depressive-like behaviors will eat less and take longer to approach the food pellet (even though they are equally hungry) than mice without. You can see on the top left that the normal and TK mice showed normal levels of feeding in the task, until they were STRESSED. The application of restraint stress rattled the TK mice much more than the normal mice, and they took longer to approach the food and each it.
The second test (middle set of graphs) is the forced swim test, where you take a mouse and put it in a beaker of water where it can’t touch the bottom. Mice are good swimmers and will paddle around for a few minutes, trying to find a way out. But after a little while, they will start to just float at the surface with their noses out of the water (don’t worry, the test only lasts 6 minutes). When a mouse is showing depressive-like behavior, it will float more and sooner than a mouse that is feeling happy go lucky. While the TK mice showed some depressive-like effects at baseline, they showed no differences in forced swim test following stress. This could be because the normal mice showed a strong effect of stress, or it could be for other reasons.
The final set of graphs up there shows the sucrose preference test, where you give a mouse a choice between a bottle of water and a bottle of sugar water. Mice LOVE sugar and will go to down on the sugar bottle. UNLESS they are showing depressive-like behavior. You can see that the TK mice drank less sugar water both at baseline and in response to restraint stress. while the normal mice were gulping down the sugar water under both conditions. Combined together, the authors think that the three behaviors indicate that TK mice have enhanced depression-like behavior in response to stress, linking stress with neurogenesis.
So what does this mean? It means that not only does stress regulate neurogenesis, neurogenesis regulates stress, and the interactions between these two could hold a key to some symptoms of depression. It also means that normal levels of neurogenesis may help “buffer” stress responses, making them a little easier to handle.
This is a pretty good paper, and I really like the model of using inducible TK, and the angle of going after neurogenesis first and stress second. But I am surprised that the behavioral testing didn’t come out…bigger. I’m surprised that the forced swim test didn’t show differences and that the TK mice didn’t drink even LESS sucrose following stress. But remember, this is after only ONE episode of restraint stress, not a long period of stress. It makes me wonder whether they would have got stronger effects if they had used a chronic stress paradigm, exposing the mice to stress for a long period of time. I think they might.
And of course, all of this begs the question: chicken or egg? Stress and decreased neurogenesis combine to produce depressive effects and regulate each other, but what is CAUSING the decreased neurogenesis in the first place? In this case, it was a mutant mouse, but what about in humans? ARE there the same effects in humans? If so, do some humans have naturally decreased neurogenesis and this makes them more sensitive to the effects of stress? Or are some people just more sensitive to the effects of stress on the brain? Is it a priming effect, where an initial stress decreases neurogenesis, and then FURTHER stress creates a depressive episode? Or is it something else all together?
No matter what, it looks like theories of depression will continue to get more complicated. It’s FAR more than serotonin, and now it’s more than neurogenesis, and it may even be more than neurogenesis and stress combined. But in the end, every research step brings us closer to an answer.
Jason Snyder et al. “Adult hippocampal neurogenesis buffers stress responses and depressive behaviour” Nature, 2011. DOI: 10.1038/nature10287