Serotonin, Synapses, and Setting: Gitte Moos Knudsen on the Neurobiology of Psychedelic Action

How Do Classic Psychedelics Work in the Brain?

“We have a huge potential for doing novel, groundbreaking work by looking at synaptic density—a molecular target that has recently become available for in vivo imaging. This hasn’t really been exploited yet, but…seems to be a very interesting avenue”

Josh Hardman, Psychedelic Alpha: What do we know about mechanisms of psychedelic action, and what we can see through imaging work?

Dr. Gitte Knudsen: Through imaging work, we know for a fact that it’s the stimulation of the serotonin 2A (5-HT2A) receptor that leads to psychedelic experiences. We also know this from behavioural studies where people have administered 5-HT2A receptor antagonists and shown that they can completely, or almost completely, block the psychedelic experience.

So that is the primary pharmacology of the classic psychedelics: psilocybin, psilocin, LSD, DMT, etc. We also know that the 5-HT2A receptors are primarily located in the pyramidal neurons in the cortex and that when sufficiently stimulated by serotonergic psychedelics, the experiential effects occur.

What happens alongside the stimulation, and the resulting psychedelic experience, becomes a little bit more uncertain. We believe, based on animal work and to some extent imaging work, that the pharmaceutical agonist effect, and perhaps also the psychedelic experience itself, is associated with neuroplasticity. Here, we have huge potential for doing novel, groundbreaking work by investigating synaptic density via a molecular target, SV2A, that has recently become available for in vivo imaging. This hasn’t really been fully exploited yet, but looking at changes in synaptic density after stimulation of the 5-HT2A receptors seems to be a very interesting avenue to pursue. We’ve done some of it, but maybe I can get back to that later.

The neuroplasticity phenomenon has also been associated directly with serotonergic psychedelics. Animal studies, particularly from the Castrén group in Finland, that seem to suggest these serotonergic psychedelics can stabilise the receptor associated with brain-derived neurotrophic factor (BDNF). Their finding suggests is that if you can stabilise the receptor that BDNF is interacting with, you can also augment neuroplasticity. I think this is an interesting theory, but we don’t yet know for sure if it’s the case or not.

And of course, there’s all the animal work where people have looked at the synaptic density, and the dendrites, and how the axons begin sprouting after doses of serotonergic psychedelics. So this is where we are now, on the molecular level.

When it comes to other aspects of the brain—meso level or macro level—people have also studied the circuit levels, where they look at networks across the brain. Here, the excitation of the serotonergic psychedelics shows consistently, that there is increased global integration across the brain and, in parallel, a reduced integrity of high-level association networks such as the default mode network (DMN). So there seems to be a decrease in within-network connectivity and an increase in between-network connectivity. This has been shown mostly for psilocybin and, to a lesser extent, for LSD.

Limitations of Current Neuroimaging Methods

“One important caveat with fMRI is the assumption that the BOLD response is preserved while under the influence of serotonergic psychedelics. Here, we are making an implicit assumption that the neurovascular coupling is unaltered but since the serotonergic psychedelics stimulate the vasculature 5-HT2A receptors this assumption may not be valid. To me it is currently unclear whether global functional connectivity changes that have been reported can – at least to some extent – be ascribed to neurovascular uncoupling effects.”

There is one important caveat to keep in mind that I think is a bit understudied now. When we use fMRI and BOLD response, we are making an implicit assumption about neurovascular coupling. Neurovascular coupling is based on a fundamental assumption: When you increase regional brain metabolism, there is a direct effect on blood flow to that region. When metabolism goes up, blood flow goes up, and vice versa.

However, we also know that some of the serotonergic psychedelics stimulate the vasculature—we can see there is a vasoconstriction response to psilocybin. This represents an underexplored area which needs to be better established before we can firmly say there are circuit effects that are purely mediated and matter in themselves, rather than just being an epiphenomenon to these neurovascular changes.

Hardman: There was a big paper about this last October from Padawer-Curry and colleagues, right? What do you think the implications are? I know your group did some work on psilocybin and ketanserin’s effects on cerebral blood flow as well.

Knudsen: Yes, that paper is another example of how careful one should be when analyzing and interpreting MR-based measures of psychedelics. We also have data where we contrast psilocybin and LSD. Because LSD is slower in its pharmacokinetics, it takes longer to set in and induce psychedelic effects, presumably because the affinity is very high. So, it seems to take a very long time for LSD to find and bind efficiently to its target, the 5-HT2A receptor and. I don’t think it has anything to do with a slow blood-brain barrier passage of LSD.

What we see is this delay between plasma, LSD, and when the brain effects start to set in—a kind of hysteresis effect, whereby over time we do not have a linear relationship between plasma, LSD, and the effect we see. I think this could be why we see a less pronounced neurovascular coupling effect; we do see opposite effects of psilocybin and LSD when it comes to global connectivity.

This is a very interesting area. There have been assumptions made for many of these connectivity papers that warrant further understanding, and that means more in vivo research must go into the neurovascular coupling.

Hardman: Some of the figures from neuroimaging studies have become very popular, even among the non-scientific community. A figure from Petri et al.’s 2014 paper, which aims to visualise functional connectivity under psilocybin versus placebo, was graffitied on a wall by a canal!

Knudsen: Yes, it’s super appealing and very intuitive, but I think there are certainly caveats. We’ve also produced similar graphs in our papers but I’m starting to worry that maybe it’s been over-emphasised and doesn’t really have much to do with reality. The only way you can investigate this in people would be to do simultaneous PET FDG with pharmacological interventions while you’re also measuring cerebral blood flow.

Hardman: So that you’re controlling for it?

Knudsen: Yes.

“There is a delay between plasma LSD, and when the brain effects start to set in—a kind of hysteresis effect, whereby over time we do not have a linear relationship between plasma, LSD, and the subjective psychedelic effect. This is in contrast to the effects of psilocybin where we see a close correspondence between plasma psilocin and psychedelic effects. When it comes to global functional connectivity, we also see opposite effects of the two drugs, something that makes you wonder if there are differences in the neurovascular coupling between the two drugs.”

Are Neurobiological Explanations Reductive?

Hardman: Interesting. Some people in the broader psychedelics field worry that neurobiological explanations of drug action are reductive, arguing that the setting and non-pharmacological factors are just as important.

I’m interested to hear your thoughts on this because your group published a preprint last year on psilocybin’s setting-dependent effects. I think you used SV2A, that measure of synaptic density, and then put healthy volunteers in an MRI and looked at how people fared in an enriched environment versus the MRI setting. Maybe you could explain this further?

Knudsen: Our data from awake pigs shows an increase in frontal cortex synaptic density one week after psilocybin, but not one day after. To examine synaptic density in humans, we made use of that information to design a clinical study where we used PET SV2A, done before the intervention and then one week after an intervention with a psychedelic dose of psilocybin.

For various reasons, it happened to be so that some of the individuals we had included in our study were scanned while they were tripping in the MR scanner, and others were in a quiet room.

When we analysed the data, we found an interaction effect. Those who were in the therapeutic room (calm environment, music playing, somebody comforting them) had greater increases in synaptic density than those in the MR scanner, where it was super noisy and disruptive. That paper is now in review. I think this speaks to this interaction between the setting and the neurobiology.

I wouldn’t think of neurobiology as being reductionistic. I think these things must work hand in hand.

“Those who were in the therapeutic room (calm environment, music playing, somebody comforting them) had greater increases in synaptic density than those in the MR scanner…I think this speaks to this interaction between the setting and the neurobiology.”

We do know for a fact that the self-reported degree of mystical experience matters. This has been a relatively reproducible finding, a good predictive marker for how depressed people benefit from the intervention. But also in healthy people, we see a consistent association between those who have a good mystical experience and those who have better long-term outcomes.

Is this because there is something about the experience itself, or is it engraved in the neurobiology? I don’t know, but I think they go hand in hand.

Predicting Response; Dosing

“Ultimately, I think what really rules the outcome is dosing—there is nothing that beats plasma concentration of these compounds; you need to get the dose right.”

Hardman: Another thing I’m curious about is the idea of predicting response in patients, or identifying biomarkers of response. Are you convinced by any biomarkers?

Knudsen: This idea of finding out who is going to benefit from these therapies really is the holy grail. Some people have investigated personality features, but the findings don’t show any specific effects. You would imagine that perhaps those who were scoring higher on creativity or openness would be more likely to respond. But you could also hypothesise the opposite, that if you have low openness, then you might be in a better position for improvement.

We’ll soon know more about these kinds of things. Important data sets are coming out of the large phase 2b/3 trials and we are currently running a large-scale study of 120 healthy individuals randomised to psilocybin or placebo, with and without music. Do we believe that music is critical for a good psychedelic experience, or does it not matter at all? Either way, I would imagine that if the outcome shows that music doesn’t matter, people will say ‘you used the wrong playlists’.

Ultimately, I think what really rules the outcome is dosing—there is nothing that beats plasma concentration of these compounds; you need to get the dose right. And that’s what molecular imaging with PET is so good at: determining which level would be spot on. You need enough to get the full psychedelic experience, but not so much to induce adverse events and negative experiences.

Hardman: Do you think that would be indication-dependent? Some people have said that in certain conditions, you might need a higher dose. Do you think that could be the case, or is there a fixed dose that would work for anyone with any condition?

Knudsen: I think the neurobiology is the same. There is not much difference between depressed individuals and healthy controls in brain density of 5-HT2A receptors, and even if there was, I am not convinced it would matter. A dose corresponding to roughly 65% occupancy is where you get these effects. We can measure this for psilocybin and for LSD, and we’ve also done some peak studies looking at this. I think it is pretty much the same range for everyone.

“I think for all people a dose around 65% occupancy at the 5-HT2A receptors is where you get profound psychedelic effects”

Non-Hallucinogenic Psychedelics

Hardman: Earlier you were saying there’s a likely correlation between mystical experience and psychedelic outcomes. Generally speaking, what’s your view on the endeavour of trying to separate out the hallucinogenic effects?

Knudsen: I’m pretty agnostic to the outcome. Whatever works, works. I know there is some resistance against the concept, because people believe that you need to have a true psychedelic experience or you need to have talk therapy afterwards to integrate what has happened.

My experience is that, yes, some people really do benefit from having the opportunity to talk through what they experienced and to interpret this in the context of their own lives. Other people walk out the door and say, ‘This was a very revealing experience. I feel totally fine now, I have no need to talk to anybody about it, and I’ve had my own insights that were super useful’.

“Some people really do benefit from having the opportunity to talk through what they experienced and to interpret this in the context of their own lives. Other people walk out the door and say, ‘This was a very revealing experience. I feel totally fine now, I have no need to talk to anybody about it”

The question about how much integration is needed is also scientifically very interesting. I would love to see some more data on that. After all, there are the short-lived compounds, like 5-MeO-DMT, where virtually no talking therapy is provided afterwards, but they still seem to work. I think there needs to be some way whereby you can follow-up to ensure that people are fine and well, but it won’t be everyone that needs a super thorough integration and follow-up.

One complicating aspect in animal studies is obviously how to assess “psychedelic” effects and the quality of the experience. Preclinical researchers have generally used doses guided by the head twitch response, but not by receptor occupancy studies.

We are now examining dose-5-HT2A receptor occupancy relationships in live pigs. Once we have established which doses are required to get 60% occupancy, head twitch studies. We now see is that some of the compounds termed non-hallucinogenic psychedelic, when given in appropriate doses, are resulting in head shakes. When you start to ask around, people who have been using these compounds for recreational purposes will tell you they clearly have psychedelic effects.

“Some of the compounds termed non-hallucinogenic psychedelic, when given in appropriate doses, do result in head shakes. When you start to ask around, people who have been using these compounds for recreational purposes will tell you they clearly have psychedelic effects.”

Hardman: Do you think that’s an indictment of the head twitch response as an assay? Or do you think it’s a dosing issue whereby, at a therapeutically relevant dose, it will have some psychoactive effects?

Knudsen: The latter.

Unanswered Questions and Future Studies

Hardman: Can you tell us more about your research agenda? What are the big questions that you want to answer? And, if funding were unlimited, what do you think would be the most useful study for the field?

Knudsen: I already mentioned neurovascular coupling—I think that is something we need to understand, specifically whether it is something we can dismiss when doing fMRI studies.

But what I would really like to do would be to have a large-scale study on SV2A methodology and get the temporal window right, because we don’t think that synaptic density will increase and increase and increase after a psychedelic dose. I think there’s more likely a window where things are being optimised, and then the “wrong” synapses are pruned.

I think what the field needs is funding to replicate things across laboratories and to standardise the way we do these studies, so that it’s not so much a question of getting papers published. We’ve seen a lot of repurposed data use, which is welcome, makes sense, and is ecological, but it also carries a risk of redundancy or bias because you’re using a small data set, and you use that for publications that don’t replicate that well. So as a field, we really need to get together and carry out these larger-scale studies.

Finally, it would be fantastic if one could start to pinpoint some predictive factors of the therapeutic outcome of psychedelics. Which patients are most likely to get the therapeutic benefit? That’s also a place where the field really is missing something.

“The field needs funding to replicate findings across laboratories and to standardise the way we do these studies, so that it’s not so much a push for getting papers published in high-impact journals. We’ve seen a lot of reusing data, which is welcome, makes sense, and is ecological, but it also carries a risk of redundancy or bias”