They are Captain Planet!
Ok, not quite. But, strangely, antidepressants on top of stress hormones may be stronger than they are alone. Why is this? And what's going on? Well, we're not quite sure.
David et al. "Neurogenesis-dependent and -indepdendent effects of fluoxetine in an animal model of anxiety/depression" Neuron, 2009.
The most commonly prescribed antidepressant (and antianxiety) medications out there are the SSRI, selective serotonin reuptake inhibitors. These include drugs like fluoxetine (Prozac), citalopram (Celexa), or sertraline (Zoloft). How these drugs work, however, is still up for debate. At first, everyone thought that, because these drugs increase levels of the neurotransmitter serotonin in the brain, that depression must be caused by low levels of serotonin, and the increase would make you feel better. We have since learned that this is not the case. Headaches don't result from lack of aspirin, and depression doesn't result from lack of serotonin.
The next theory for how depression, and antidepressants, might work was the neurogenesis theory. We used to believe that you were born with all the neurons you'd ever have, but we now know that neurogenesis, the birth of new neurons, occurs throughout life in areas of the brain like the hippocampus, an area usually associated with learning and memory. Antidepressants can increase neurogenesis, on a time course which matches the clinical efficacy of antidepressants.
But now we are starting to think that it might be more complicated than that. What about, for example, the role of stress? People with depression often present with elevated levels of stress hormone (cortisol in humans, corticosterone in rodents), and treating animals with corticosterone can induce symptoms that we associate with a depressive-like state. But does this affect whether, and how, antidepressant works?
To find this out, David et al treatment a bunch of mice with corticosterone in their drinking water. This is the hormone released in response to stress, and can create one very stressed out mouse. The animals became more anxious, spending less time in the center of an open area (mice prefer the dark and an open arena is a measure of anxiety), and eating less food in novel environments (another measure called novelty suppressed feeding). The mice also showed reductions in neurogenesis in the hippocampus, and a rough coat state and decreased grooming, all signs of a mouse not taking care of itself. Interestingly, though, they couldn't get effects in other antidepressant tests like the forced swim test or tail suspension test. The reason for this might be because corticosterone increases locomotor activity, and if the animals had higher locomotor activity, it would mess up some of those measures.
They then took half of those mice, and added an antidepressant (either fluoxetine or imipramine), ON TOP of the corticosterone treatment.
Doesn't that look better. You can see that adding Prozac increased the time the animals spent in the open field and made their novelty suppressed feeding measures shorter. Antidepressants were able to repair the state produced by corticosterone alone.
But what about the neurogenesis?
This is where it got funny. Corticosterone alone decreased neurogenesis in the hippocampus, like we have come to expect (lots of studies show that stress reduces the birth of new neurons). But when you added an antidepressant on top of that, you didn't just get recovery. No, instead you got AUGMENTATION. The animals had much higher neurogenesis than expected.
So what was causing these effects? How is fluoxetine producing its effects on behavior and how could corticosterone be augmenting them? David et al first looked at the role of the hippocampus. When they used X irradiation to get rid of the hippocampus in the corticosterone treated mice, they saw that some of the behavioral effects, like the novelty suppressed feeding effects, were blocked, showing that these effects were hippocampus-dependent, when the hippocampus isn't there, fluoxetine can't have its effect. But other behavioral effects of fluoxetine, like the open field effects, were NOT blocked when the hippocampus was gone. These effects appear to be independent of the hippocampus, but do depend on the present of corticosterone and fluoxetine. So where are they coming from?
To find this out, the authors looked at gene expression of various genes in different brain areas. They found that three genes, beta arrestin 1, 2, and Gi alpha 1, were changed in an area of the brain called the hypothalamus, both in response to the corticosterone alone and in the response to the combination of corticosterone and fluoxetine.
The hypothalamus is one of key players in the hypothalamic-pituitary axis, which controls the stress reponse and the secretion of corticosterone. Beta arrestin 1 and 2 and Gi alpha 1 are all genes that are involved in the control of G protein coupled receptors, and would help control how the hypothalamus responds to stress or antidepressants.
To look at this further, David et al took a bunch of beta arrestin 2 knockout mice, who had no gene for beta arrestin 2. They found that these mice had no reaction to the effects of the antidepressant fluoxetine, which suggests that this protein is necessary for antidepressant response.
So it looks like the fluoxetine is most effective when combined with corticosterone, or exposure to stress. This becomes important when you take into account that 80% of people diagnosed with major depressive disorder report symptoms after a significant life stress. It may be the stress itself that helps to make antidepressants more effective, which could become important in the clinic if we could use levels of stress hormone to determine who will respond to antidepressants and who won't. But while beta arrestin 2 might play a role in this, it is probably not the only player, and the real mechanism remains a mystery.
David, D., Samuels, B., Rainer, Q., Wang, J., Marsteller, D., Mendez, I., Drew, M., Craig, D., Guiard, B., & Guilloux, J. (2009). Neurogenesis-Dependent and -Independent Effects of Fluoxetine in an Animal Model of Anxiety/Depression Neuron, 62 (4), 479-493 DOI: 10.1016/j.neuron.2009.04.017