I'm aware that NMDA antagonists interfere with LTP but is this the primary cause of psychoactive effects? If so has there been any attempt to explain why interfereing with this system causes the subjective shift in consciousness that it does. I don't understand how the effects of dissociatives could be primarily explained by an increased difficulty in forming new memories. I understand that ampakines tend to damper the effects of dissociatives which would suggest that interference with synaptic plasticity is indeed the source of the dissociative high but I have trouble making that connection. Perhaps there is some subtlty in how the brain uses these systems that I'm not understanding. Sorry if this question is a little abstract but I wasn't sure where else to ask it.
-
N&PD Moderators: Skorpio
Are the subjective effects of NMDA antagonists the result of hampered LTP?
- Thread starter Zilpe
- Start date
I don't think so?
The binding of the NMDA antagonists prevents glutamate from binding to that receptor among other things. Glutamate is considered to be one of the neurotransmitters involved with learning, memory and our cognition.
I think our perception relies on that system because our cognition is used to process and make meaning from what comes through our senses. If that processing is disrupted and distorted, the connection between what we know and understand and what we are experiencing is influenced. This can ultimately even be disconnected through loss of coherence 'completely' which is I think what we call a hole, because whatever experience and memory formation there still is may happen in 'isolation', effectively another form of sensory deprivation (which has it's own psychoactive effects by the way).
Apart from the connection to our senses, what we know and understand is also being disrupted so it can get more and more difficult to recollect how things work or what they are conceptually. Thus we feel 'stumped', and 'crossed wires' can lead to bizarre thought patterns and ideation and misinterpretation of what we are, where we are, what objects are around us and how to interact with them the usual way.
Some nootropics are primarily AMPAkines while others also have influence on NGF and LTP etc. I think the 'older' racetams mostly are the former but they still apparently attenuate dissociative effects. I think that is because they enhance glutamatergic transmission that is still left and how that is connected to cholinergic pathways. This perhaps makes a critical difference in the hippocampus, and may lessen the effects from having 'goldfish memory'.
I don't really know if synaptic plasticity and LTP even enter the picture, but my point is rather that they don't seem necessary in finding an explanation for what dissociatives do and what they stop doing when you take nootropics.
If it was mostly an effect of (hippocampal) LTP being messed with preventing formation of memories, don't you think that the effects on person would be more like those of a patient with his hippocampus removed i.e. normal behavior but just a specific kind of memory problems?
Even if that is also happening on the side, just let's not forget that a primary effect of dissociatives is scrambling how we interpret the world. Enabling us to learn and remember what happens would IMO only enhance some coping and later recollection, but not fix perceptual distortion.
Correct me if I'm wrong, I am going on recollection, not sources right now. But in my defense: I did not take any dissociatives.
Last edited:endotropic
Bluelight Crew
- Joined
- Feb 14, 2012
- Messages
- 1,140
Me neither.
You're committing a logical fallacy here. Even if NMDA antagonists prevent LTP, and Ampakines stimulate it, that doesn't mean LTP is the ONLY level of interaction between the two compounds. They could be opposing each other at multiple different levels, one of which is responsible for the NMDA antagonist subjective experience.
In some situations NMDA antagonists actually stimulate LTP, especially in the hippocampus: http://www.ncbi.nlm.nih.gov/pubmed/16183195
The leading theory for Ketamine's antidepressant effects is that it stimulates LTP (recruits AMPAr) in the hippocampus (likely in a BDNF dependent matter). Apparently in the hippocampus NMDA receptors predominate on GABA neurons, so the primary effect of NDMA antagonists in that region is disinhibition of excitatory neurons, increasing LTP.
WSH
Bluelighter
- Joined
- Nov 30, 2012
- Messages
- 360
Little correction:
The NMDA antagonists that we are talking about here (e.g. ketamine) do NOT prevent glutamate from binding to the NMDA receptor.
Glutamate can bind as much as it wants and the ketamine blocks the NMDA channel from the INSIDE, at a completely different site than the glutamate site.So do NMDA receptors have some function other than facilitating synaptic plasticity? I agree that the shift in perception seems to be similar to a type of sensory deprivation I just have trouble understanding how this could be caused by NMDA antagonism. In my mind I see the visual effects of dissociatives being the result of inhibition of feed-forward connections in the visual cortex, sensory information is no longer effectively transmitted from lower layers to higher layers. Is this something that could be caused by NMDA antagonism or do I have the wrong picture entirely?
They say the stimulation of LTP was measured 24h after treatment. Could this be the result of a rebound effect due to upregulation?
WSH
Bluelighter
- Joined
- Nov 30, 2012
- Messages
- 360
I think you MEAN the right thing, you just used the wrong word, ketamine does NOT block or prevent the NMDA RECEPTOR (= glutamate site) from being activated, but it DOES block the NMDA CHANNEL (the NMDA receptor is only a small part of the NMDA channel).
The term "uncompetitive channel blocker" says it all: they don't compete against glutamate + they block the CHANNEL (and not the receptor)
There are also competitive NMDA antagonists, but I'm pretty sure Zilpe is talking about the UNcompetitive ones, right?I was talking about uncompetetive ones. This got me thinking about the actual effect of NMDA antagonists on synaptic plasticity. We know that synaptic plasticity is mediated by NMDA and AMPA receptors so it makes sense that an NMDA antagonist would somehow interfere with this process but I've been wondering about exactly how it would do this. I found some models of synaptic plasticity involving NMDA receptors, one is detailed here
http://icwww.epfl.ch/~gerstner/SPNM/node74.html#SECTION04142000000000000000
The idea is that NMDA receptors have three states NMDAup, NMDArest, and NMDAdown, this is just a model so the biophysical interpretation of these states isn't clear but the proposed explanation is that a receptor in the up state is saturated by glutamate whereas the down state doesn't refer to the state of the receptor but rather a downregulation of receptor expression. One other interpretation I suppose is that an NMDA receptor in the rest state is bound by Mg which is expelled to allow receptor activation which we consider the up state. A presynaptic spike sends receptors to the up state and a post synaptic spike sends receptors to the down state, when there are receptors in the up state a post synaptic spike will signal a secondary messanger Sup to increase the synaptic efficacy, since there is a decay term for NMDAup this gives a window of time following a presynaptic spike for which a postsynaptic spike will cause an increase in synaptic efficacy this is LTP. NMDAdown and Sdown act in a similar manner but with a postynaptic spike triggering a window of time for which a presynaptic spike will cause a decrease in synaptic efficacy, this is LTD.
Now it's not entirely clear how the dynamics would be effected by the prescence of an uncompetetive NMDA antagonist. One might expect this is equivalent to an overall decrease in the concentration of NMDA receptors for NMDArest, this immediately makes the given biophysical interpretation of NMDAdown suspect as NMDArest might become equivalent to NMDAdown and all post-synaptic spikes would in effect cause LTD. somehow this seems like long term dissociative use would then have a much more profound impact on the brain. Ignoring the interpretation we can assume any shift in relative concentration of NMDAup/NMDAdown/NMDArest states causes an equilibrium shift, this will cause a shift in dynamics which depens on how the drug interacts with the up/down/rest states that will presumably settle to a new equilibrium. If the drug binds to all three states of the receptor equally well then we might not see any effect as shifts between the states wouldn't cause an equilibrium shift. If, however, the drug bound better to the up and down states then that would increase the decay term resulting in a quicker return of up and down states to the rest state. This in effect would decrease the window of time for which LTP and LTD could take effect reducing the overall effect of synaptic plasticity.
In terms of the effect of the secondary messanger we might assume a receptor in the up or down state that's bound by an NMDA antagonist is not able to signal the secondary messanger in which case the increase in Sup and Sdown following a correct series of signals would be less effective in changing the synaptic efficacy. At any rate using this purely heuristic reasoning about the model it seems like NMDA antagonists would infact decrease the overall synaptic plasticity of the brain. This however is predicated on quite a few assumptions about the various interactions between the NMDA antagonist and the various hypothetical receptor states. Furthermore if receptor expression actually was an intermediate variable used to determine LTD then the story would be different. I think this would actually be an interesting project to work on for someone in mathematical biology/neuroscience, unfortunately I already have my thesis to work on so I probably won't be able to look into it.
Regarding the premise I guess I just don't see how these effects on plasticity could ellicit any sort of high as I don't see the change of synaptic efficacy as something we experience but rather something that happens over a much longer time frame. Sorry for the long post but I realize not everybody has a background in mathematics so I wanted to try and explain the model in a way that was understandable.
St3ve
Bluelighter
- Joined
- Mar 15, 2013
- Messages
- 621
Not remembering what happened during the peak of a dissociative high undoubtedly has something to do with its LTP inhibiting properties. Apart from inducing LTP I think NMDA receptors also have a role in maintaining it. The general idea is that sufficient activation of a cell in synchrony with NMDAR activation will induce LTP by increasing the amount of AMPA receptors on the cell, sensitising it to glutamatergic stimulation at those sites. The NMDA channel; however, is a non-selective cation channel though and will let in Na+ as well as Ca2+ and can contribute to cell depolarization (and hereby increasing probability of an action potential being fired). By blocking this component, cells will not be stimulated as much from glutamatergic activation as they would normally. This would in turn mess with the input the higher visual areas get from the lower ones.
As for the LTP measured after 24hours, it might be due a rebound effect, or it may be caused by other NMDA-independent LTP mechanisms... Not all hippocampal cells require NMDA activation for LTP induction.Very interesting, and thanks for the explanations. Seems like some of these complex mechanisms could be useful in explaining subtle parts of dissociative ASC phenomenology. Not remembering the peak perhaps doesn't have to be an implicit effect of a strong dissociative experience if it is true that some people are using combinations of drugs successfully to retain more of the peak of the experience.
Similarly I wonder what is suggested by the apparent ability of nootropic ampakines to negate most of the dissociative effects, maybe it would pay off to look more closely at which dissociative effects are attenuated or quelled the most by nootropics. Unfortunately it may get extra complicated if we have to explain that by the (often multiple) MOAs of those nootropics, but who knows they may have influence on modulating the LTP-dependent effects much more than the others.
Yes I didn't take this into account, activation of NMDA receptors is not only a step in the process of LTP/LTD but contributes to the depolarization of a cell. Perhaps the overall inhibition of the glutamatergic system is responsible for the subjective effects and the reduced effectiveness of LTP/LTD is merely a side effect, this is supported by what Solipsis says about people using drug combinations to remember the dissociative experience without dampening it (very interested in these combinations by the way). It also wouldn't contradict the dampening of the experience by nootropics as an AMPA modulator would somewhat attenuate the overall effects of NMDA antagonism and not just increase the effects of LTP.
In terms of AMPA I thought it was the other way around, activation of AMPA receptors by presynaptic spikes caused Mg to be expelled from NMDA receptors allowing them to be activated. My only source for this is wikipedia so I could easily be wrong, if AMPA receptor expression is the "up" state in the process of synaptic plasticity this might have an effect on the model I was describing as NMDAup would have to be reinterpreted as the concentration of AMPA receptors, which aren't effected by the presence of an NMDA antagonist. This wouldn't decrease the window of oppurtunity for LTP to occur since starting in an initial state for a certain concentration of AMPA receptors the dynamics of the putative secondary messanger are unchanged. However the reduced equilibrium concentration of NMDAdown would result in a filtering out of weaker presynaptic signals. In any case it seems the conclusion is unavoidable, NMDA antagonists interfere with synaptic plasticity, the only question that arises looking at the simplified model is whether that's via a reduced time window or some other mechanism.
Do we know of any selective AMPA modulators that could be used to facilitate these experiments?
St3ve
Bluelighter
- Joined
- Mar 15, 2013
- Messages
- 621
You are correct that in order for LTP induction to occur you first need AMPA activation to expel the Mg of the NMDAR before it can be activated. The downstream effects of NMDAR (Ca2+ second messengers) activation leads to production of extra AMPARs which are then inserted into the cell membrane at the relevant site (look up "synaptic tagging and capture" by Frey and Morris if you want to get into the details on how we think that works).
Jonneh
Bluelighter
- Joined
- Jul 29, 2014
- Messages
- 274
We should remember that LTP does not directly correspond to memory at the behavioural level. Rather, synaptic plasticity (of which LTP, particularly in in the hippocampus, is a well-studied paradigm) is a form of cellular memory (lasting change in activity), which, in turn, is a process that could conceivably underlie virtually every adaptive function the brain has. Of course, even a static architecture of brain-level complexity could exhibit some extraordinarily dynamic behaviour, but it appears that fluctuating patterns of activity leave their mark on the connective efficacy of the cells in which they occur. Zilpe's very reasonable doubts about the relevance of these plastic processes to the drug high highlight the importance of temporal scale: long term changes in efficacy, which occur over minutes to hours, are quite different from the fastest changes, like the paradigmatic presynaptic modifications of post tetanic potentiation and paired pulse facilitation, which could account for computationally relevant processes such as input frequency filtering (e.g. low, high and band pass), potentially germane to aspects of the drug high.
Karl Friston, a hero of mine, has a very recent paper (pre-published last week; http://www.ncbi.nlm.nih.gov/pubmed/25053181) on the very issue of ketamine and its effect on top-down control (read: the ego), which is of great help in explaining its dissociative effects.
Optimising DCMs of theta and gamma frequency range responses, model comparisons suggest that both enhanced gamma and depressed theta power result from a reduction in top-down connectivity from mPFC to the hippocampus, mediated by postsynaptic NMDA receptors (NMDARs). This is accompanied by an alteration in the bottom-up pathway from dCA1 to mPFC, which exhibits a distinct asymmetry: here, feed-forward drive at AMPA receptors increases in the presence of decreased NMDAR-mediated inputs. Setting these findings in the context of predictive coding suggests that NMDAR antagonism by ketamine in recurrent hierarchical networks may result in the failure of top-down connections from higher cortical regions to signal predictions to lower regions in the hierarchy, which consequently fail to respond consistently to errors.
A basic understanding of the predictive coding approach is useful here. The idea is that every level of the grand hierarchy that is the brain has a model (expectations) of what it expects from lower levels. The lowest level makes predictions (which result from its model) about the causes of sensory data (i.e. how the world [the level below - there is no strictly demarcated boundary] works). If experience differs from expectation (prediction error), the model of the world is modified and the updated model is shown to the next level up in the hierarchy, which compares it with its own model. The process continues, onwards and upwards, until all levels have taken the new sensory data (experience) into account (minimised prediction error). The end product of this cascade, then, is a percept - effectively a final judgement regarding the likely causes of sensory data, experienced consciously. The mechanism by which different hierarchical levels come to an agreement on how to reconcile discrepancies between prediction and experience consists in taking into account the precision (synaptic gain) of the model (variance of the probability distribution of the causes of sensory data) - the more precise a model, the less it accommodates data that don't agree with it. At the higher level, changes in a model are changes in perception. Therefore, aberrant precision (overly precise models) results in hallucinations - the realising of expectations. When Friston and company state that prediction and error signalling is impaired, they describe a disconnect between lower and higher levels, between the body and the self (dissociation).
Essentially, these predictive models are encoded in the synaptic strengths of the connections between neurons in different hierarchical levels. When they change, it is likely that some of the well-described mechanisms of long-term synaptic plasticity (receptor delivery/removal and structural changes) are recruited. During the dissociative experience, however, NMDA atagonism goes some way towards simulating receptor removal, and given that the experience is transient, it is doubtful to what extent longer lasting changes, such as those concerning synaptic receptor content, take place.
Perhaps some creations of Cortex Pharm, but only time will tell...
I am aware of the predictive coding approach and was trying to fit the dissociative high into that framework based on what I knew about the pharmacological mechanisms of dissociatives, that is NMDA antagonism, which to my understanding was primarily something that effected synaptic plasticity which is where my confusion came from. Interesting to see an actual paper on this though, I think the relation between drug activity in the brain and cognitive state is one that really needs to be explored more formally.
I also find quite interesting the idea that such shifts in global brain funciton could have long-term impacts on how the brain processes information since all of the learning done in the brain is "online." I wonder if HPPD is actually just a result of synaptic plasticity in the visual cortex adjusting the overall model. In this sense maybe the dampening of synaptic plasticity from dissociatives is actually a good thing, preventing the brain from adjusting itself to it's new state. HPPD does seem to be less common from dissociative use although things like depersonalization and presistent dissociation are reported. Surely such questions would require an empirical approach.
Finally I think it's interesting to note that recently some artificial neural networks(deep belief networks) used in machine learning have come to resemble the heirarchical nature of the visal cortex and have been given a similar interpretation. Maybe in the future we will be able to conduct in silico experiments on perception.
Jonneh
Bluelighter
- Joined
- Jul 29, 2014
- Messages
- 274
Yes, I quite agree. In a sense, any long-lasting changes induced by perception-altering drugs (including beneficial changes) must be the result of memory of that transient shift. What makes these changes last in some cases (such as in HPPD) and not in others is indeed a very interesting question, and factors involved in cementing these changes through long-term plasticity could be good candidates.
Regarding the mechanism of NMDAR blockade by ketamine, there is some controversy over the requirement for ion flow through the NMDA receptor for NMDAR-dependent LTD. The claim by those who doubt it is that a metabotropic function of the NMDA receptor (abolished by glutamate binding site blockers, such as APV, but preserved by ion channel blockers such as ketamine and MK-801) is involved in LTD. Certainly the subjective effects of different NMDA antagonists differ wildly, and this could be a mechanistic basis for this difference. The point is that we don't know whether ketamine's NMDA antagonism is actually blocking certain forms of long term plasticity, but it's still a reasonable speculation.
I find these approaches magnificently broad in their scope of explanation, and we're surely getting there. The architecture of these artificial neural networks is still imposed in a top-down manner (by the programmer) so it's not as if evolution and artificial networks that optimise their own structure have converged on the same solution, but once set free they certainly use similar strategies. We can still learn an awful lot from how the brain is organised, and as messy as it is often described to be, it is, in some romantic sense, perfect, and its illusions and biases are increasingly found to reflect trade-offs at deeper levels that we aren't going to be able to circumvent easily in ANNs.
