If you have even the slightest interest in the biology of depression (or if you’ve spent any time treating depression), you’ve heard about the connection between stress and depressive illness. There does seem to be a biological—maybe even a causative—link, and in many ways, this seems intuitive: Stressful situations make us feel sad, hopeless, helpless, etc—many of the features of major depression—and the physiological changes associated with stress probably increase the likelihood that we will, in fact, become clinically depressed.
To cite a specific example, a steroid hormone called cortisol is elevated during stress, and—probably not coincidentally—is also usually elevated in depression. Some researchers have attempted to treat depression by blocking the effects of cortisol in the brain. Although we don’t (yet) treat depression this way, it is a tantalizing hypothesis, if for no reason other than the fact that it makes more intuitive sense than the “serotonin hypothesis” of depression, which has little evidence to back it up.
A recent article in Molecular Psychiatry (pdf here) adds another wrinkle to the stress hormone/depression story. Researchers from King’s College London, led by Christoph Anacker, show that antidepressants actually promote the growth and development of new nerve cells in the hippocampus, and both processes depend on the stress hormone receptor (also known as the glucocorticoid receptor or GR).
Specifically, the group performed their experiments in a cell culture system using human hippocampal progenitor cells (this avoids some of the complications of doing such experiments in animals or humans). They found that neither sertraline (Zoloft) alone, nor stress steroids (in this case, dexamethasone or DEX) alone, caused cells to proliferate, but when given together, proliferation occurred—in other words, the hippocampal progenitor cells started to divide rapidly. [see figure above]
Furthermore, when they continued to incubate the cells with Zoloft, the cells “differentiated”—i.e., they turned into cells with all the characteristics of mature nerve cells. But in this case, differentiation was inhibited by dexamethasone. [see figure at right]
To make matters more complicated, the differentiation process was also inhibited by RU486, a blocker of the receptor for dexamethasone (and other stress hormones). 
What’s amazing is that RU486 prevented Zoloft-induced cell differentiation even in the absence of stress hormones. (However, it did prevent the damaging effects of dexamethasone, consistent with what we might predict.) [see figure at left]
The take-home message here is that antidepressants and dexamethasone (i.e., stress hormones) are required for cell proliferation (first figure), but only antidepressants cause cell differentiation and maturation (second figure). Furthermore, both processes can be inhibited by RU486, a stress hormone antagonist (third figure).
All in all, this research makes antidepressants look “good.” (Incidentally, the researchers also got the same results with amitripytline and clomipramine, two tricyclic antidepressants, so the effect is not unique to SSRIs like Zoloft.) However, it raises serious questions about the relationship between stress hormones and depression. If antidepressants work by promoting the growth and development of hippocampal neurons, then this research also says that stress hormones (like dexamethasone) might be required, too—at least for part of this process (i.e., they’re required for growth/proliferation, but not for differentiation).
This also raises questions about the effects of RU486. Readers may recall the enthusiasm surrounding RU486 a few years ago as a potential treatment for psychotic depression, promoted by Alan Schatzberg and his colleagues at Corcept Pharmaceuticals. Their argument (a convincing one, at the time) was that if we could block the unusually high levels of cortisol seen in severe, psychotic depression, we might treat the disease more effectively. However, clinical trials of their drug Corlux (= RU486) were unsuccessful. The experiments in this paper show one possible explanation why: Instead of simply blocking stress hormones, RU486 blocks the stress hormone receptor, which seems to be the key intermediary for the positive effects of antidepressants (see the third figure).
The Big Picture: I’m well aware that this is how science progresses: we continually refine our hypotheses as we collect new data, and sometimes we learn how medications work only after we’ve been using them successfully for many years. (How long did it take to learn the precise mechanism of salicylic acid, also known as aspirin? More than two millennia, at least.) But here we have a case in which antidepressants seem to work in a fashion that is so different from what we originally thought (incidentally, the word “serotonin” is used only three times in their 13-page article!!). Moreover, the new mechanism (making new brain cells!!) is quite significant. And the involvement of stress hormones in this new mechanism doesn’t seem very intuitive or “clean” either.
It makes me wonder (yet again) what the heck these drugs are doing. I’m not suggesting we call a moratorium on the further use of antidepressants until we learn exactly how they work, but I do suggest that we practice a bit of caution when using them. At the very least, we need to change our “models” of depression. Fast.
Overall, I’m glad this research is being done so that we can learn more about the mechanisms of antidepressant action (and develop new, more specific ones… maybe ones that target the glucocorticoid receptor). In the meantime, we ought to pause and recognize that what we think we’re doing may be entirely wrong. Practicing a little humility is good every once in a while, even especially for a psychopharmacologist.
Thought Broadcast 
Exciting new knowledge, Steve, and I have not read the article, but we should avoid jumping to conclusions. You allude to cell growth as a new antidepressant mechanism, but although I read there’s an association, I see no evidence to establish that cell growth reduces depression. But what do you mean by depression anyway? Depressed mood? Dysthymic disorder? Major Depressive Disorder/Episode?
We are dealing with mental illness here, but when you think about it isn’t it normal to feel down when something bad happens, especially if it was loss of a loved one (Bereavement)? To me the greater the loss, the more normal it is to experience “symptoms.”
Kenneth Kendler in Richmond has studies the relationship between adverse life events and (I believe) major depressive episodes. He only found about a 30% correlation, but time is a critical variable. If the episode starts before the event, of course it cannot have been caused by it, then here comes post hoc ergo propter hoc. Suppose it starts one second after the event. One week. One year. There is no accepted incubation period, and the longer it takes the greater the likelihood of other events or even factors or just chance.
Kendler says we won’t be in a position to identify a particular event or situation as causing an episode until we have fully elucidated the mechanism by which it occurs. This research may help, but we’re still a long way off. In the meantime it’s trial and error.
moviedoc, yes I should point out that the authors made it clear that antidepressants cause neurogenesis but said nothing about the role of neurogenesis in depression per se. (One of the benefits of working with cell culture– none of those nasty symptoms to deal with!) In fact, they state that the role of neurogenesis in reversing depressive symptoms (note I don’t say “treating depression”) is being “intensely debated.”
If glucocorticoids are involved in the stress response and in the onset of depression (and possibly also required for cell proliferation as a way to “recover” from prior insults), they may represent the missing link between normal stress/sadness and clinical depression. Of course, this has been hypothesized for years; we just have to figure out where to draw the “treatment threshold.”
If you haven’t done so yet, check the first linked article for further roles of the stress steroids.
As a chronic pain patient I have to wonder if the fact that antidepressants have a place in treating chronic pain may give indications as to the mechanism in the brain. Chronic pain and depression are two very different problems. How is it that a medication that is psychotrophic also works on (some) physical pain and pain disorders.
Carol
http://apainedlife.blogspot.com/
Yeah well I wonder if placebos promote neurogenesis. After all, they have been shown to be as effective as antidepressants in “treating” depression!
I wonder if psychotherapy, ECT, rTMS, VNS promote neurogenesis.
BTW: stressful events can also make us happy, and still increase the risk of imminent physical illness. But how do we usually know an event or situation is “stressful?” By the fact that it makes us sad, depressed, anxious, etc. We’re going around in circles.
BTW2: I like the way engineers use the term: Stress is the externally applied force. STRAIN is the resulting deformation of the object or material. We can carry the analogy even farther: There is complete recovery from elastic deformation (resilience). Plastic deformation is recognized by permanent change.
I am not sure why this is being considered “new news” here. The positive effects of increased serotonin levels to neurogenesis in the hippocampus has been known for years. The mechanism has not been clearly known, and is still not clearly known, but it appears to be that increased serotonin levels in the synapse turns on BDNF to the extent that it can assist in the healing of “burned” hippocampus nerve cells. Check my somewhat old PPT presentation found at http://www.hippocampus.us for more on what ssri’s (well, at leaset prozac) do for the hippocampus.
I read this post completely concerning the difference of most recent and earlier technologies,
it’s remarkable article.
If you want to get much from this paragraph then you have
to apply such strategies to your won website.
After going over a number of the articles on your web site, I truly like your way
of writing a blog. I saved as a favorite it to my bookmark website
list and will be checking back soon. Take a look at my
web site too and tell me your opinion.
Remarkable! Its really remarkable paragraph, I have got much clear idea regarding from this post.