The Neurogenesis theory of depression and a little guy called CREB

Sci wishes she could begin this post with something clever. But she has a cold. Suffice it to say that this paper is cool and interesting. And also, as Sci has a cold, I expect all of you to read this post out loud to yourselves in suitably stuffy, gluey Sci-voices.

(*sniff*) Gur et al. "cAMP Response Element-Binding Protein Deficiency Allows for Increased Neurogenesis and a Rapid Onset of Antidepressant Response" The Journal of Neuroscience, 2007.
(Yeah, yeah, the title is long and scary. Worry not, Sci will 'splain.)
And this paper is especially good because it allows Sci to write a post on a topic she's been meaning to get to even since she did a depression series way back when: the neurogenesis theory of antidepressant responses.
So here we go. And a new neuron is born.

(From Bumpy Brains. Sci thinks the rendition of diapers as glia is hilarious.)

So some of you might think that by the time you're grown up, you'll have all the neurons you're every going to have. WRONG! We used to think that this was the case, but it turns out that there are two areas of your brain where neurogenesis (neuro = neuron, genesis = birth, neuronbirth!) continues to happen throughout adulthood. One of those areas is a place called the subgranular zone of the hippocampus.
The hippocampus (comes from a word for seahorse) is a little...seahorse shaped...area of your brain. Located here:

(yeah, ok, from there it doesn't look like a seahorse. But if you look at it from the front, it looks kind of squiggly like...sort of. Look, I didn't name it.)
This is an area of your brain that scientists have classically associated with processes like learning and memory, but it's also an area that people are studying for various other things, including (for the purposes of this post) depression.
Why depression? Well, scientists have been studying depression, and the way antidepressants work, for some time now. And they've noticed a few things. First, studies of humans with depression show that these people may have less neurogenesis in the hippocampus. Whether this is a result of being depressed, or whether it causes depression, is still up for grabs. New neurons are just not being born at the same rate as in healthy brains, and people with severe depression (or at least, studies done in suicide cases) have shown decreased brain VOLUME in the hippocampus.
Second, as you probably all know from commercials for antidepressants and from stuff you've read in the news, antidepressants don't work instantly. They take several weeks, which seems rather odd. Until you notice that antidepressants like Prozac actually cause increases in neurogenesis in the hippocampus after you give them for a few weeks. Sure enough, antidepressant treatment increases neuron birth and survival in the hippocampus (though not in all cases, and not to the same degree, which might go a long way to explaining why some people do not respond to antidepressants).
Put these things together, and the neurogenesis hypothesis was born. The hypothesis is that some people (not necessarily all people) with depression have decreases in neurogenesis in the hippocampus. When you give these people antidepressants, their neurogenesis increases, and it is this increase which treat depression long term.
Obviously this is still a hypothesis, there's a lot of ongoing studies to determine HOW this happens. For example: HOW do selective serotonin reuptake inhibitors, like Prozac, cause increases in neurogenesis? Why does this only work for some people and not others? How come it takes so long? And so on and so on. And so a lot of studies are looking into mechanisms behind neurogenesis, trying to figure out how it may produce antidepressant effects. Like this study here.
Enter the molecule CREB. CREB (cAMP response element binding protein) is a protein which binds to particular DNA sequences. Therefore, when it's activated, CREB can increase or decrease the activity of certain genes. If those genes were, say, involved in neurogenesis, and if, say, antidepressant treatment increased CREB, then maybe CREB is involved in the hippocampal neurogenesis that occurs following antidepressant treatment. Conversely, if CREB is involved in neurogenesis, and antidepressants DECREASE CREB, maybe CREB modulates neurogenesis. Either way, there's a mechanism.
And to test this idea, enter the CREB-deficient mouse. And Reese's Pieces.

So we'll start with the CREB-deficient mouse. It's a mutant mouse that produces less CREB than normal (it produces low levels, but some production is required for the mouse to live successfully). And the first thing the authors did was to hang these mouse up by the tail and then make them swim.
It sounds kind of odd, sure, but the tail suspension test, and the forced swim test, are both tests for antidepressant efficacy. The idea is that mice will give up struggling or swimming after a while (usually less than 5 minutes), and just hang or float. But this test is very sensitive to antidepressants, and if you give a mouse an antidepressant beforehand, they will keep struggling or swimming for longer. Are the mice depressed? Probably not, but the test has what we scientists like to call "predictive validity". It doesn't mimic a disease phenotype, but it's a very good screen for what drugs are going to be effective in treating it. Drugs that act as antidepressants work in the forced swim and tail suspension tests. Drugs that don't act as antidepressants don't. There you go.
Back to the CREB-deficient mice. It turns out that these mice exhibit an antidepressant effect in the tail suspension test and the forced swim test. Basically, they don't give up easily, and act like they are on antidepressants, even though they aren't. So the question was, does this antidepressant effect mean neurogenesis?

Indeed it does. You can see up there the number of BrdU labeled cells in the hippocampus of normal and CREB-deficient mice. BrdU is a way to measure baby cells, so labeling a brain area with it can show how many new cells are being born. And you can see that the CREB-deficient mice (the grey bars) have much higher levels of neurogenesis than the normal mice (white bars). In fact, it's about like what you get when you treat a normal mouse with an antidepressant (the pale hashed bars), though the antidepressants didn't seen to have an effect on the CREB-deficient mice (the dark hashed bars).
So CREB-deficient mice have more neurogenesis. These cells aren't only born, they stay around for a while.
So the question becomes, what makes these mice have all this neurogenesis? Well, since we know that many antidepressants act on serotonin, the scientists here decided to deplete serotonin in the CREB-deficient mice. Not only did getting rid of serotonin cause these mice to behave like normal mice in the depression tests (they gave up just like normal mice), it also caused DECREASED neurogenesis. This means that serotonin is required for CREB-deficient mice to show antidepressant effects like they do. Interestingly, though, CREB-deficient mice have the same levels of serotonin that normal mice do (if the serotonin is required for the antidepressant effect, you might think it would be higher), but it could be that CREB-deficient mice have difference sensitivity to their levels of serotonin than normal mice.
So, there's another test used for antidepressants, that tends to be thought more valid, because it takes a while to kick in. This is called novelty-induced hypophagia, and this is where we talk about Reese's Pieces. Basically, mice love Reese's Pieces (or peanut butter chips, which is what they used, which sounds like Reese's to me. Sci knows she would respond very well for Reese's). And so they'll gobble them up whenever they get them. BUT, if you put the mouse in a new environment, they get freaked out. The environment is unfamiliar, and until they are sure it's not threatening, those Reese's are going to stay right where they are. They call this novelty-induced hypophagia (novelty=new, hypophagia=not eating).
It turns out, that if you treat a mouse with antidepressants, they will feel better about chowing down on tasty food in a novel environment. BUT, until the forced swim test and the tail suspension test, this effect doesn't happen instantly. It take a few weeks of treatment, just like people on antidepressants take a few weeks of treatment to feel the effects. So this test seems like a good one for measuring antidepressant (and anxiety) effects that correlate ("correlate" not are caused by because we can't really really say that yet) with neurogenesis.
And it turns out that CREB-deficient mice kind of fail at the novelty-induced hypophagia. Part of this may be that these mice are much more anxious than normal mice. What was cool, though, was that BOTH the CREB-deficient mice and the normal mice responded to antidepressant treatment in the Reese's pieces test. Not only that, while normal mice responded only with chronic antidepressant treatment, CREB-deficient mice responded instantly.
So what does all this mean? So far, it appears to mean that CREB does affect neurogenesis in mice, and that lower levels of CREB increase neurogenesis (which may mean that higher levels of CREB decrease it, though that would have to be tested separately), and may provide a link for how antidepressants act. However, this paper (like many good papers do) leaves far more questions than answers. What to higher levels of CREB do? Why does serotonin depletion affect these mice? Are their serotonin receptors different? What about the fact that the mice had an instant reaction to antidepressants in novelty tasks? What is the cause of that effect? What is the precise effect of CREB on neurogenesis?
So many experiments, so little time! Probably some of these have been written, and Sci will try to check them out and get back to you. Cause I'm in it for the science!
Gur, T., Conti, A., Holden, J., Bechtholt, A., Hill, T., Lucki, I., Malberg, J., & Blendy, J. (2007). cAMP Response Element-Binding Protein Deficiency Allows for Increased Neurogenesis and a Rapid Onset of Antidepressant Response Journal of Neuroscience, 27 (29), 7860-7868 DOI: 10.1523/JNEUROSCI.2051-07.2007

7 responses so far

  • ENT-TT says:

    Thanks Sci, that made my day!
    I've been curious about variable neurotransmitter sensitivity in humans for a while now, and this opens up all kinds of cool possibilities. Excellent post.

  • Katie says:

    Neat! My research focuses on the neurogenesis theory of depression and norepinephine so this was really fun to read. Great post!

  • BG says:

    There is a typo in the 4th to the last paragraph:
    "But until the forced swim test..."
    Should probably be:
    "But UNLIKE the forced swim test..."
    which makes a bit of difference in the meaning of the sentence.
    Unless you really do have to do the other two tests before the third test shows results, in which case, please disregard the comment.

  • Scicurious says:

    BG: you're right, Typo. My bad.

  • fog says:

    That is really cool. I read the whole thing I might have actually understood it too. Thanks!

  • Andreas Johansson says:

    And also, as Sci has a cold, I expect all of you to read this post out loud to yourselves in suitably stuffy, gluey Sci-voices.

    I normally "hear" your prose in an upbeat and distinctly girly voice, both of which characteristics I'm poorly placed to emulate. Stuffy and gluey I think I can manage.

  • Passing through says:

    Isn't it the case that caffeine interacts with cAMP?
    A lot of intellectuals and grad student types habitually take in much more caffeine than can be warranted for stimulant effects alone (and usually have sufficient tolerance that they exhibit only minimal physical effects like heart rate or blood pressure).
    Connection? Self medication in pursuit of increased neurogenesis?