Do we finally now know how all antidepressants work?
For decades many within and without of psychiatry have successfully propagated the notion that depression is caused by not enough of the neurochemical serotonin in the brain. This myth has survived, to a greater or lesser degree, despite there being a considerable lack of evidence to the contrary. It seems to make so much sense so has refused to go away, and over time it has morphed and become increasingly complex [1]. It made sense because the main drug treatments for depression, antidepressant drugs, seemed to fairly universally increase the amount of serotonin in the brain. If the drugs worked by increasing serotonin, the story went, depression must be due to not having enough serotonin in the first place. It’s continuation undoubtedly was also helped by the lack of a decent alternative hypothesis.
Now, over 50 years after the first anti-depressant drugs started to be used, it seems we may be able to put a final nail in the coffin of the so called serotonin hypothesis of depression. At the same time, perhaps we can also challenge what seems to be the most unchallengeable of these assumptions: that antidepressants actually work through increasing serotonin.
This seems unchallengeable when the most common antidepressants are actually called selective serotonin reuptake inhibitors (SSRIs): the name says it all. They must work through changing serotonin. However, a fascinating article published this month from a diverse group of researchers from England, Germany, France, Brazil and the US has presented an attractive alternative model, one if correct, will banish the serotonin connection to depression into the dustbin of history [2].
The paper in question describes a complex series of thoroughly conducted and connected experiments developing the idea that the antidepressant activity of a variety of drugs, including classical so called serotonergic antidepressants, as well as novel drugs like ketamine, may act through a completely different mechanism. One involving a chemical in the brain called brain derived neurotrophic factor (BDNF). BDNF is a so-called neurotrophic factor. Neurotrophic factors are chemicals in the brain that promote various aspects of the growth and adaption of nerve cells. Clearly these are pretty important.
The idea that BDNF may be relevant to understanding depression is not new. Studies have previously demonstrated that antidepressant medications tend to cause an increase in BDNF in the brain and BDNF seems to have antidepressant effects in animal models of depression. It was frequently presumed, however, that changes in BDNF activity was secondary to the primary action of antidepressant drugs on serotonin although the nature of this link was never clear. This new paper, proposes a radically different hypothesis: that antidepressant drugs work directly on the BDNF system and their effects on serotonin maybe totally unrelated to their therapeutic value.
Understanding the findings of this paper requires exploring a little bit of the neurobiology of the BDNF system. The receptor for BDNF in the brain, where it acts, is the complicatedly named neurotrophic tyrosine kinase receptor 2 (TRKB). When BDNF finds to the TRKB receptor, this stimulates a variety of actions in brain cells but generally these are related to aspects of brain plasticity, including the strengthening or weakening of connections between nerve cells. BDNF binding to the TRKB receptor looks like it is good for the nerve cells in your brain and the way they adapt during learning.
There is also an important relationship between action at the TRKB receptor and cholesterol in the brain. Although most people will be familiar with cholesterol as the troublesome thing we need to limit in our diets, cholesterol is a critical substance in the brain as it is needed to help make up the membranes — the outer layers — of our brain cells. When the TRKB receptor is stimulated, this promotes production of cholesterol in brain cells. However, to make things even more complex, cholesterol itself has effects on the TRKB receptor itself.
So what has this all got to do with antidepressants?
First, the authors of the paper conducted a series of experiments that demonstrated that antidepressant drugs bind to the TRKB receptor itself. Effects of antidepressants on BDNF were not connected to serotonin at all: the drugs tested, from several antidepressant classes, all bound directly to a specific site on the TRKB receptor. This binding resulted in changes in the TRKB receptor that actually would increase signalling at this receptor: this would enhance the effects of BDNF, promoting neural plasticity. The binding of antidepressants to the receptor seems to increase signalling at the receptor as well as increasing the number of receptors on the cell surface able to be stimulated by BDNF itself.
The authors then showed that these effects of antidepressants were functional. They increased aspects of cellular plasticity and antidepressant effects in an animal model. They also showed that these antidepressant effects in an animal model did not happen if the receptor for the antidepressant on the TRKB receptor was disrupted by a genetic mutation. Interestingly, the antidepressant effects were also disrupted in the presence of unusually high or low levels of cholesterol.
Another interesting piece of the puzzle related to the concentrations of the medications required to produce these effects. The effects were produced at relatively low concentrations of ketamine, consistent with the amount of ketamine that would presumably get to the receptor relatively quickly after receiving a therapeutic dose in human subjects. Notably, this is consistent with the observation that when ketamine works, it works quite rapidly.
However, significant concentrations of traditional antidepressants were required to achieve these effects. These concentrations would only achieved in the human brain after antidepressant drugs have been administered for several weeks. Note, it is known that these drugs accumulate at the relevant sites in the brain, plateauing after several weeks. This time course is consistent with how long it takes these traditional antidepressants to work. In contrast, they bind to the traditional serotonin pathways relatively immediately and there is always been a question as to how these immediate serotonin effects are linked to the slow onset of antidepressant actions.
So according to this theory, traditional antidepressants would attached quickly to their target in the serotonin system but this effect would be of no relevance (except maybe from causing side effects). The levels of these antidepressants will then gradually accumulate in the brain over several weeks to the point that they can start to stimulate the TRKB receptor, producing antidepressant effects. Ketamine would stimulate the TRKB receptor pretty much straight away, explaining its fast onset of action.
So in summary, this study found that a variety of antidepressant medications from different classes bind to a specific site on the TRKB receptor for BDNF. This binding seem to produce relevant neuroplastic and antidepressant effects in animal models and these effects were not present if there was an abnormality in the site where the antidepressant bound to the TRKB receptor. These effects also occurred at doses of antidepressant drugs that appear to be consistent with the time course of the effects of these drugs in the brain of humans undergoing antidepressant therapy.
This finding provides a consistent and plausible alternative hypothesis to explain the antidepressant effects of pretty much all the known forms of antidepressant medications. This hypothesis is more consistent than the existing serotonin related hypotheses with the time course of medication related clinical benefits.
There are also variety of interesting connections between these findings and observations around the potential interaction between cholesterol, depression and antidepressant drugs. There have been reports of low cholesterol levels in patients with suicidal ideation and the possibility of increased rates of suicide on some drugs which reduce cholesterol, consistent with the notion that cholesterol is required for the activation of the TRKB receptor. Low cholesterol would reduce the capacity of BDNF to stimulate its receptor and reduce neuroplasticity. In contrast, there is also some evidence that very high levels of cholesterol may also have disruptive effect in this process, which provides a potential explanation of the common co-occurrence of metabolic syndromes where people have high levels of cholesterol and depression.
Why is this all-important? Clearly we need a much better understanding of the pathophysiology of depression itself, most importantly to guide the development of new therapeutics. If we continue to try and develop drugs targeting the serotonin system we are just playing Russian roulette hoping that they may work or not if in fact their action is fundamentally through different mechanisms.
These findings can directly lead to the development of new molecules specifically targeting this binding site on the TRKB receptor, which may well be a much more focused and successful pathway to identify novel and potentially better tolerated antidepressant drugs. For example, it may well be possible to develop rapidly acting antidepressants like ketamine — which don’t bind to the other ketamine targets such as the NMDA receptor — and therefore avoid some of ketamine’s really problematic side-effects. Drugs that specifically target the TRKB receptor may also avoid a lot of the standard antidepressant side effects which presumably are caused through effects on serotonin system.
As a final comment, it is possible that the BDNF system will prove to be a final common pathway for an even broader class of treatments than just antidepressant medications. There is already some evidence that brain changes in response to repetitive transcranial magnetic stimulation and electroconvulsive therapy may be dependent on variations in the BDNF system.
It has taken 50 years but we finally have a meaningful alternative neurochemical hypothesis for the cause of depression and one which can directly lead to new family of potentially more effective and well-tolerated treatments. This research clearly needs replication and extension but is truly an exciting finding and one that we may ultimately look back on as having been quite revolutionary.
1. Fakhoury, M., Revisiting the Serotonin Hypothesis: Implications for Major Depressive Disorders. Mol Neurobiol, 2016. 53(5): p. 2778–2786.
2. Casarotto, P.C., et al., Antidepressant drugs act by directly binding to TRKB neurotrophin receptors. Cell, 2021.
