I've written a bunch of posts in the past on serotonin, the serotonin theory of depression (and why it's probably wrong), and some stuff on current antidepressant treatments. And I even talked before a little bit about the serotonin theory vs the BDNF theory. But I've never really COVERED what the BDNF theory IS and how it works. And then I saw this paper, and here's my opportunity!
But I have to warn you, this is a LOTTA paper. But it's ok, the point is good. ONWARD.
So what IS BDNF? And what's it got to do with depression? Let's take it from the top. BDNF stands for Brain Derived Neurotropihc Factor, and is part of cascade of proteins, produced in the brain, that promotes neuron growth and stops neurons from dying. It doesn't act alone, of course, but it's a highly useful little dude.
Here we might pause and say "neuron growth?! Wut?!" I don't know about you all, but when Sci was raised in her little public school system, she was told that we were born with all the neurons we would ever have, and that we would produce no more. But now we know that that isn't true. In fact, you produce new neurons throughout your lifetime, in two major areas of the brain, the subventricular zone and the subgranular zone. The subgranular zone is what we are concerned with right now. This is a region of the hippocampus (an area of the brain associated with learning, memory, and depression), and is found roughly here:
On the left is a human brain, on the right is a rat brain, and you can see how the hippocampus (in blue) is positioned in both species.
I'm sure you all knew that the hippocampus is associated with learning and memory, but it is ALSO associated with depression. Or, more specifically, with antidepressants. We used to think that antidepressants worked because they increased the neurotransmitter serotonin in the brain. But there was a disconnect between the way the drugs were supposed to work and the onset of the actual antidepressant effects. Traditional antidepressants like Prozac produce peak serotonin levels in the brain in within about 6-8 hours. But the antidepressant effects take longer. WAY longer, up to several weeks. So if the increases in serotonin themselves weren't causing the antidepressant effects, what was?
Well, it turns out that, if you treat an animal with a drug like Prozac for a few weeks, long enough to get a clinical effect in humans, you also get neurogenesis in the hippocampus, along with increases in BDNF. So the idea is that, somehow, antidepressants increase BDNF, which helps to increase neurogenesis, and this produces antidepressant effects.
There are other studies to support this. For example, stress and increases in the stress hormone corticosterone will cause DECREASES in BDNF, and decreases in neurogenesis, and stress itself is associated with the development of major depressive disorder. Not only that, but PEOPLE with depression actually show lower levels of BDNF in their blood than people without.
So with information that BDNF promotes neurogenesis, and stress causes low BDNF, and is associated with depressive behaviors...well, what does BDNF itself do? Scientists have tried putting BDNF directly into the hippocampus of rodents, and gotten increases in neurogenesis and antidepressant effects. But most patients wouldn't really want people all putting drugs directly into their hippocampus. So this study aimed to see if they could get antidepressant effects of BDNF, just giving it in the periphery of the body rather than the brain. Even though it's "brain derived" BDNF is in fact found all over the body, so maybe treatment there could produce effects.
But BDNF has a REALLY short activity period. REALLY short. Giving it as an injection or a pill every day wasn't going to work. So they used a nifty little gadget called an osmotic minipump. You pump these little pumps, filled with drug mixture or saline, under an animal's skin, and it'll just hang out there, with very little pain or irritation to the animal, for as long as you need. In this case, they treated the animals for two weeks with BDNF. The two week treatment is important, because antidepressant effects with normal antidepressants like Prozac normally require THREE weeks. If you can get effects in two weeks (or less than that!) it'd be better than what we have now. In this case, while they looked for the final studies at the end of two weeks, they ran behavioral tests after the FIRST week, to see if the antidepressant effect could come on that quickly.
So they treated the mice, and put them through their paces.
Up here you can see the first set of behavioral tests. The one on the far left is the forced swim test, where you look at what scientists like to call measures of "coping". Basically, you lower a mouse into a bowl of water where its feet can touch the ground. Mice are good swimmers and for a few minutes it will scrabble valiantly against the walls and swim around, trying to find a way out. When it realizes there IS no way out, the animal will stop swimming and float. We call the swimming 'active coping' and the floating 'passive coping'. We know that antidepressants decrease passive coping (floating) and increase active coping (swimming and climbing), and scientists usually take this as a sign of antidepressant activity. The forced swim test is a VERY common screen for antidepressant drugs. You can see in the graph on the far left that the animals who got either saline or the lowest dose of BDNF floated for between 250-300 seconds, while the animals that were being treated with the two highest doses floated a lot less.
The middle graph shows similar effects in a test for antidepressant activity called novelty induced hypophagia (NIH). This test, and the one on the far right (Chronic mild stress, which I'll get to in a minute) are important because they measure the effects of CHRONIC antidepressant treatment, while the forced swim test usually measures the effects of a single injection (though it can also work for chronic treatment as we saw here).
Anyway, NIH test tests to see how long it takes an animal to approach a tasty food in a novel environment. Antidepressants make them approach it faster. Think of it like this: every day, you come home from work or school or whatevs and I give you a tasty cake. You get cake every day, and pretty soon, you come home from work every day being like "WHERE IS MY CAKE!!?!?!" Then, on the last day, I throw you out in the middle of the ocean on a life raft, with the cake. Presumably you'd be a bit more worried about being on a raft in the middle of the ocean, and it would take a little while to get to your cake. With mice, it's just like that, only with something mice find tasty, like peanut butter chips or Froot Loops. You can see that the mice take a lot longer to approach the tasty food in the novel environment, and treatment with the higher doses of BDNF actually decreases how long they take, which is interpreted as both antidepressant and an antianxiety response.
Finally, on the far right up there we have chronic mild stress. This involves treating the animals for a while with...chronic mild stress. This varies so the animals don't get too used to it, but involves some mixture of chilly rooms, strobe lights, rock and roll music, rocking the animals back and forth, and other little things designed to discomfit the animals. After chronic mild stress animals show signs of depression, like drinking less tasty sugary drink when given the opportunity. You can see there that treatment with BDNF made them drink more sugar water after stress than when they didn't get treated (though they still didn't do as well as unstressed animals).
This figure shows the effects of chronic BDNF (I think they picked the middle dose as the best one) on measures of anxiety. Both the elevated plus maze and the open field test work in the same way. Mice like the dark and nice covered tight corners. Light and open spaces stress them out. So how much they will go in to a lighted area or an open area is a measure of anxiety. You can see that BDNF helped with anxiety in the elevated plus maze, but didn't show any difference in the open field. The authors say maybe this is because the open field is less stressful, but it's hard to tell. These tests are VERY sensitive, and honestly it could have been anything.
So ALL these graphs put together basically show that chronic BDNF treatment produces antidepressant-like (we always have to say antidepressant-like until we've seen it in a human) effects. And the cool thing is that they saw these effects after only a WEEK of treatment, rather than the usual THREE WEEKS required for most antidepressants.
And behavioral tests are fine, but what about neurogenesis? After all, antidepressants produce neurogenesis, right?
Well, so does this:
These graphs show increases in neurogenesis and survival. You can see that while the drug treatment had no effect on neurogenesis (it didn't help produce MORE neurons), it DID have an effect on survival. Normally, there is turnover of produced neurons, with some of them dying off. Giving BDNF prevented that, so allowed more neurons to stay around. So while it's not producing neurogenesis, it has a similar effect (though I think it could possibly have produced neurogenesis itself if maybe they had given it longer).
Other experiments they did in this study confirmed that BDNF was working through specific cellular machinery specific to BDNF, and not through other ways (this included differences in cellular machinery with names like pCREB and ERK), but the finding is clear, giving BDNF produced antidepressant effects in stressed mice, and helped neurogenesis by increasing cell survival. This means that drugs which target BDNF, rather than the antidepressants we have now which target serotonin, could be used to maybe combat depression (though whether this is specific to stressed individuals, who can say). Not only that, it may also means that we can use blood levels of BDNF to measure depression in people (though I am personally not to sure about this, there have been mixed findings on this one).
Now, I would like to point out that none of this PROVES that low levels of BDNF cause depression. In fact, decreasing BDNF artificially in animals doesn't itself cause changes in the depressive behaviors I described above (though it makes them more susceptible to stress). So the reality may be that low levels of BDNF and decreases in neurogenesis may be only part of the problem, or may be a symptom rather than the problem itself. But just because it's not the cause, doesn't mean we can't use advances in the study of BDNF to try and find new treatments. And now that we know that BDNF itself can be administered and have antidepressant effects in animals, perhaps we can use that knowledge to make new drugs which target BDNF, producing better antidepressant effects than what we have now. Sure, BDNF right now can't be taken as a pill, but you'd be amazed at the stuff we can come up with.
Schmidt HD, & Duman RS (2010). Peripheral BDNF Produces Antidepressant-Like Effects in Cellular and Behavioral Models. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology PMID: 21085113