Dr. Barondes, you were a researcher at the NIH during the period when the earliest research on neurotransmitters and antidepressants was conducted. What happened while you were there?When I arrived at the NIH in 1960, I was interested in the way the brain controls the pituitary gland and how this influences behavior. We already knew that there was serotonin and norepinephrine in the hypothalamus, the brain region that controls the pituitary. So I began working on the effects of these monoamines on the pituitary, and this brought me into contact with Julius Axelrod and Bernard Brodie, pioneers in molecular pharmacology who were both at NIH. I was soon lured away from work with the amines when I joined Marshall Nirenberg’s lab to learn molecular biology and to work on the genetic code. But I kept an eye on the exciting work that Axelrod and others were doing.

When were norepinephrine and serotonin discovered?Norepinephrine was discovered at the beginning of 20th century in extracts of the adrenal medulla and then recognized as a substance that is released by sympathetic nerves. Serotonin was discovered in the 1940s in blood platelets. It is called serotonin because platelets release it into the serum (sero) during clotting and because it increases the tone (tonin) of blood vessels by promoting contraction of smooth muscle. By the 1950s both of these amines had been found in the brain. Brodie, who had been Axelrod’s mentor, became particularly interested in serotonin, and Axelrod, who had his own lab at NIMH, focused on norepinephrine.

At what point was it suspected that these neurotransmitters might have something to do with depression?The tip-off came from observations in the 1950s of the effects of drugs on depression. The first observation was that an antituberculosis drug, iproniazid, is an antidepressant; the second was that imipramine, an antihistamine, is also an antidepressant. These discoveries were made on the basis of clinical observations and with no understanding of the mechanisms of action of the drugs. But it was soon found that iproniazid blocks an enzyme called monoamine oxidase that degrades a variety of amines, including norepinephrine and serotonin, meaning that iproniazid was the first MAOI. Once its mechanism of action became clear, it raised the possibility that norepinephrine and serotonin are natural mood elevators. Support for this idea came from observations about reserpine, a drug that was then widely used to treat hypertension.

How does reserpine fit into the neurotransmitter story?When I was in medical training, reserpine was a commonly used antihypertensive, but one of the side effects was depression. People in Brodie’s lab found that the brains of animals that were given reserpine have very low levels of serotonin and norepinephrine. We now know that reserpine has this effect because it blocks the vesicular storage of these neurotransmitters, leading to their degradation, which in turn might be responsible for the depression that the drug produces.

What about imipramine’s mechanism of action? Who figured that out?This took a little more time to figure out. The first guess was that imipramine is also an MAOI, but a fairly easy experiment showed that it is not. And this is where Julius Axelrod comes in. He had been studying the accumulation, release, and metabolism of radioactively labeled norepinephrine in nerve terminals. At one point, he added imipramine along with radioactive norepinephrine to the fluid around neurons and found that the drug prevented the amine from being taken up by the terminals. This suggested that there is a reuptake pump in nerve terminals that has two purposes: to terminate the chemical signal and to recycle the norepinephrine so it can be used again. So neurons have two ways of terminating a norepinephrine signal – they can destroy the neurotransmitter using monoamine oxidase, or more conservatively, they can terminate the signal by pumping NE back into the terminal. Ultimately, imipramine has the same functional effect as an MAOI – prolonging the action of norepinephrine (and serotonin) at synapses.

And is that why Axlerod got the Nobel Prize?He got the Nobel Prize for a cumulative body of work, but I think that if you had to pick a single experiment that was most important, it was the imipramine work. Axelrod also made many other big discoveries. For example, he is the one who discovered the cytochrome P-450 enzymes that are involved in drug metabolism. He also showed that imipramine inhibits not only the reuptake of norepinephrine but also the reuptake of serotonin. So Axlerod showed that imipramine was the original dual reuptake inhibitor, long before Cymbalta (duloxetine) and Effexor (venlafaxine) came along.

Now move us up to Prozac (fluoxetine).Prozac, as you know, is a selective serotonin reuptake inhibitor with little effect on norepinephrine, and the public now believes that it is serotonin rather than norepinephrine that is the brain’s natural antidepressant. But it was clear from the start that both amines influence mood. And in the early days norepinephrine was king. In fact one of the most famous papers in the history of psychiatry is “The Catecholamine Hypothesis of Affective Disorders” (Am J Psychiatry 1965;122:509-522), in which Joseph Schildkraut, who was then at NIMH, argued that depression is caused by a deficiency in norepinephrine. But serotonin always had its fans, the most notable of whom is Arvid Carlsson, who, like Axelrod, had worked with Brodie and also went on to win a Nobel Prize.

What was his story?Carlsson is a Swedish neuropharmacologist who got interested in serotonin while he was working with Brodie at NIH. After returning home, he persuaded Astra, the Swedish drug company (now AstraZeneca), to find drugs that block serotonin reuptake. This was not much of a stretch – after all, Axlerod had shown that a known antidepressant, imipramine, blocked the reuptake of both NE and serotonin. And desipramine, a metabolite of imipramine, was already available as a pretty selective norepinephrine reuptake inhibitor. So in some ways it was natural to consider making a drug that was selective for serotonin.

And was this what would eventually become Prozac?Not exactly. The first marketed serotonin reuptake inhibitor was, indeed, developed by Astra and was called Zelmid (zimelidine). It was marketed in Europe and was prescribed to many thousands of people, and it appeared to be an excellent antidepressant. Furthermore, Lilly, which had become interested in serotonin reuptake inhibitors and had already started working on fluoxetine, was so impressed by it that it considered dropping fluoxetine and partnering with Astra to be the American purveyor of Zelmid – just as Forest Laboratories much later partnered with Lundbeck to sell Celexa in the U.S. This type of arrangement is very common and is a way for a drug company to avoid the development costs associated with creating a new compound. 

So what happened to Zelmid?Although Zelmid worked great, it had a terrible side effect: about a half dozen patients who took it developed Guillain-Barre syndrome, and because of this it was taken off the market. So Lilly went back to working on fluoxetine. And years later, they came out with Prozac, which, of course, transformed psychiatry.

How do we know that blocking the reuptake of serotonin is really the thing that makes SSRIs work as antidepressants?That’s an interesting question. We have pretty good evidence that blocking the reuptake of serotonin is involved, because a variety of drugs that share this property are all antidepressants. Because SSRIs such as Prozac and Zoloft and Paxil and Celexa have such different chemical structures, it is unlikely that they also share another unknown property that is really responsible for the antidepressant effect. But it has been apparent since the discovery of imipramine that there may be a lag of weeks before the antidepressant effect kicks in, whereas the effect on serotonin reuptake is established immediately. So something else must be going on that takes weeks to develop, and there has been a great deal of speculation about what that might be. One idea that is now popular – and that is supported by some experimental observations – is that the SSRIs stimulate neurogenesis in certain brain regions such as the hippocampus. This may have a therapeutic effect by replacing neurons that were killed by the corticosteroids that are released in large amounts during the course of severe depression. In this view, the increased serotonin in synapses helps repair that damage by promoting neurogenesis. 

Thanks for this fascinating history lesson, Dr. Barondes. I want to make sure readers know that they can get the fuller version of this information in your book, Better than Prozac (2003, Oxford University Press).