The following post contains speculation, so please keep a level of skepticism.
You really shouldn't try to derive and useful information from the plasma concentrations, not in this situation anyway. The discrepancy between the routes of administration is interesting enough, but trying to compare it to endogenous levels is just going to get you nowhere. The primary reason being that the circulation of exogenous DMT is going to reach every part of the brain. However, there is no reason whatsoever that endogenous DMT would have to do the same.
This is a clunky example, but it's one of the simplest I could make. Say that you could take exogenous dopamine and have it cross the blood-brain barrier (which you cannot). That dopamine would then circulate your brain, causing effects similar to (but not entirely the same as) dopamine releasing agents. However, there are times when your endogenous dopamine might be very high in one part of the brain and very in low in another part. That would be because thinking of an entire effect spectrum as "what dopamine does" is inaccurate. Dopamine in the striatum does one thing, but dopamine in the prefrontal cortex does something entirely different, and your brain could need both of these things to be done either together or opposite of each other depending on the situation. Therefore, it is quite important to remember that saying that you have "more dopamine" does not inherently mean anything; measuring the plasma level of dopamine would tell you nothing about which parts of the brain have more or less.
I have also heard that the serotonin signal is low or mute during regular dreaming. However, serotonin receptor activation is not believed to be the point of endogenous DMT.
This paper proposes that the reason DMT could be useful at such small concentrations is because it builds up in neurons to be released on to sigma-1 receptors for its better studied effects. If that is the case, it seems even less likely to me that DMT's serotonin receptor activity would play much of a role under normal conditions. I had personally dismissed DMT a while ago for the serotonergic reason, but lately I have reconsidered at least one plausible (to me) theory....
In the striatum, a part of the brain which has been implicated in psychosis, sigma-1 receptors are known to form heteromers with both D1 and D2 receptors.
This study shows that cocaine actually creates D1-dominant dopaminergic effects by activating sigma-1 receptors which form heteromers with D1 receptors in the striatum. Activating the former potentiates the latter.
This study, in addition, shows that cocaine inhibits D2 receptor activity through a heteromer formed between sigma-1 and D2, furthering this relationship. It is known that sigma-1 receptor agonists applied exogenously have antidepressant, nootropic, and rewarding properties. The rewarding properties make sense if you consider what the cocaine studies say in that D2 in an aversion learning receptor and D1 is a reward learning receptor, since natural dopamine flow would be shifted toward the latter. The antidepressant effects could probably fall under that explanation as well. The nootropic effects I would think might come from the fact that D2 in the striatum is known to be one of the most important receptors for dopaminergic psychosis based on antipsychotic mechanisms, and D1 agonists have also been shown to have nootropic properties.
A theory of mine based on these D1, D2, and sigma-1 relationships is that D2 causes the feeling of being in a dream, and the declining lucidity that accompanies it. On the other hand, D1 would increase the feeling of alertness. This would make enough sense if you consider that there is a known relationship between D1, D2, and NMDA, wherein D1 potentiates NMDA receptor currents and D2 inhibits them. Therefore, D2 should cause some dissociative-like changes in perception, in addition to other effects. My thought is that in lower levels of dopaminergic stimulation, D1 activity probably overpowers D2 activity. That would be why moderate levels of dopamine release have a nootropic effect. Higher levels would then start to cause the dissociation that twists your mind into a dreamlike delirium. If, however, there is strong sigma-1 receptor activity at the same time, much of this dissociation should be reversed. This is interesting to think about if you consider that not all psychotic D2 activity necessarily has to take place through a D2 and sigma-1 heteromer.
This paper, which I can only link to the abstract of, highlights the importance of the heteromer formed between D2 and 5-HT2A in the striatum. It found that LSD and DOI, but not serotonin, enhanced the signaling of D2 agonists through a mechanism that was blocked by a 5-HT2A antagonist. For all intents and purposes, receptor heteromers should be thought of as serving unique, individual functions, but they of course can still be grouped to a degree by the receptor protomers that they are comprised of. Therefore, all of the D2 activities of heteromers in the striatum could be considered as a potentially important contribution to the totality of the stereotypical psychosis. My guess is that, considering that serotonin does not activate this process, the 5-HT2A/D2 heteromer might be a significant player in psychedelic entity contact. And here's why I think this would work....
If exogenous DMT gets into the striatum and activates the 5-HT2A/D2, sigma-1/D2, and sigma-1/D1 receptor heteromers, a few different things should happen. Affinity of dopamine for D2 binding sites connected to 5-HT2A would improve, the response of D2 binding sites connected to sigma-1 would deteriorate, and the response of D1 binding sites connected to sigma-1 would improve. This is especially important because DMT increases dopamine levels in the striatum as well, likely, like all psychedelics, through various serotonin receptor activities and enhanced firing of glutamate from prefrontal 5-HT2A receptors on to ventral tegmental area dopaminergic neurons. In this case, the dopamine released by dopamine should have a theoretical shift toward the creation of dream-like perceptions like entity interactions and a shift from the abstract dissociation seen with high doses of many psychedelics towards a very intensely aware and connected nootropic state. This could logically mean that the hallucinations caused by the 5-HT2A/D2 heteromers are intensified by cognition-enhancing effects of the sigma-1/D1 heteromer without being held back by sigma-1/D2, and therefore the crispness, complexity, and rationality of the hallucinations would increase in the same was as a dream being enhanced by something like galantamine or nicotine patches, which drive cholinergic and glutamatergic systems similarly to D1 activation.
My theory about how DMT could possibly be involved in natural dream processes is this. In a normal dream state, high dopamine levels in limbic areas of the brain such as the striatum disable feedback neurotransmission to the higher cognitive areas of the brain like the prefrontal cortex, including glutamate signals. If there was going to be a dopamine receptor that created the dreamy, disconnected delirium that exists in any typical dream, wouldn't it make sense that it would be the dissociative D2 receptor? However, this level of cloudiness does vary from one dream to another, and of course, there is a state where it can be very strongly removed: lucid dreams. When you become lucid, those higher cognitive centers of your brain become active again, the fog and some of the randomness clears up, and the dream actually becomes a lot more vivid. It also suddenly feels incredibly rewarding, like a huge rush of euphoria, to the point that most people have to train themselves to not wake themselves up almost immediately. And of course, there is a remarkable feeling of clarity that comes with the whole experience, compared to the dream from before.
So what would this model look like if you proposed that a release of striatal DMT actually played a role in the transition into lucidity? As I said, high dopamine levels would be suppressing the prefrontal cortex. Suddenly, the good old attentional processes trigger some DMT to come flying their way. In normal brain states, DMT is not believed to reach the level of concentration required to reach 5-HT2A activity. So, if this is assumed to be correct, then the activity of dopamine on 5-HT2A/D2 heteromers will not change. Instead, activity of sigma-1 will disable its connected D2 activity and enhance its connected D1 activity. So, if a switch from D1 to D2 dominance at high doses of many typical psychedelics is responsible for the transition from feeling more alert to feeling like you're in a dream, then wouldn't it be logical enough to suppose that a switch from D2 to D1 dominance might cause a sudden feeling of becoming more awake? If this was the case, then striatal DMT could be the reason for this feeling of suddenly increased clarity and alertness when becoming lucid during a dream. At the same time, the increased signaling of D1 would potentiate the hallucinogenic structure of the dream, which of course also happens greatly at the moment of lucidity.
That theory works well with a dream not exactly being all that similar to a DMT trip because it actually suggests that the generation of the dream itself has nothing at all to do with the DMT, but that the clarity, complexity, and reward of a lucid dream may be a direct result of DMT release. This would then become significant when exogenous DMT is ingested, because when taken at a high enough dose it would actually cause its
own level of dreamlike activity through intense dopamine release, in addition to enhancing 5-HT2A-linked D2 activity to the point of causing the powerful, mystical hallucinations associated with psychedelic "psychosis", only at that point by this theory it would be like taking your lucidity neurotransmitter in an enormous dose, thereby blocking sigma-1-linked D2-based dissociative psychosis and shifting to an intense nootropic effect instead. That would then power the already quite intense DMT dream-like state to the level of overwhelming realism, complete immersion, feelings of sudden intense clarity, and pure, freeing euphoria of a lucid dream. The result would be taking the same world of abstractions and entity perceptions that you see on high doses of many psychedelics and supercharging it with rocket fuel and switching your perception of it from a dreamy dissociation to an intensely focused experience that makes you feel more "awake" than ever before, considering that you're activating this process very strongly while already awake in the first place. I think that would also be an important part of funneling enough processing power into entities to be able to give them actual intelligence and understanding. Just another theory....
That's my thoughts on the matter for the moment anyway. What do you guys think?