Decent Newsroom
Sign In

Raising IQ: AMPA PAMs, NMDA & D-Serine

By Nootopics July 3, 2026

Research and education, not medical advice.

Glutamate forms the very basis of thought. As such, glutamatergic drugs can be some of the most potent nootropics. We saw that with TAK-653, where cognitive testing scores improved consistently for all who participated. However, these pathways are notoriously ubiquitous and nuanced, so anything targeting it should be geared towards maximum rewards. This requires rather specific mechanisms.

This page compiles my work on tuning glutamate to raise cognition in healthy people: AMPA PAMs and TAK-653, the NMDA coagonist site and D-Serine, the L-Theanine connection, my move to Neboglamine, plus Clausenamide and Sialic Acid. For the acetylcholine half of my stack see Cholinergic Cognition, and for the antidepressant overlap see Mood Beyond SSRIs.

AMPA PAMs & TAK-653

Mechanism

An AMPA PAM works by increasing the likelihood of information processing neurons, or spiking neurons, to fire electrical signals. This is a cascade set off by glutamate binding, which is a pivotal transaction in times of learning. This enhanced calcium signaling will cause long term potentiation (LTP) which strengthens memory and improves learning. [6]

However, AMPA PAMs have an interesting characteristic: in non-human primates, the increased connectivity from spiking neurons in cortical association regions then activated the precuneus when it would normally be dormant. This is a significant finding, as it indicates entirely new abilities would be possible when otherwise limited by connectivity. [6] Interestingly, the precuneus is crucial for episodic memory and human consciousness, and is normally active in a rested state. [7]

AMPA PAMs are split into two groups: low impact and high impact. Low impact AMPA PAMs preferentially block extracellular domains that deactivate the receptor, [6] while high impact AMPA PAMs may also enhance agonist binding to AMPA, as a traditional PAM would.

AMPA PAMs improve cognition in healthy people

  • Piracetam: Enhances verbal memory after 14 days. [1] Has a moderate but significant benefit to motor skills, visual acuity, working memory and generalized cortical function. [2] Decreases EEG complexity, a marker of improved brain function. [3]
  • CX516: Improves visual memory, memory of scents, spatial memory and generalized cognitive function, with the exception of verbal memory. [4]
  • Semax: Is also an AMPA PAM. [12] Improves attention, short-term memory, and decision making. [11]
  • Pesampator: Reverses ketamine-induced spatial working memory and verbal memory impairments. [5]
  • TAK-653 (new): Improves executive function in the stroop test. [10]

TAK-653

In essence, TAK-653 is a selective AMPA PAM that does not agonize resting AMPA receptors. This is important, because TAK-653 is not only safer, but it enhances cognition beyond the capacity of AMPA PAMs that act as agonists. [8] The result is an improvement to working memory and cognitive flexibility without seizures or other forms of toxicity. This is documented in TAK’s preclinical studies, but also in general with AMPA PAMs. Piracetam for instance, the first nootropic, is an AMPA PAM.

TAK-653 has went through two phase 1 clinical trials, where it was found to be safe and without side effects. It is under investigation for treatment resistant depression, after TAK-653 improved depression similarly to ketamine, but without damaging cognition. [9] In addition to the above, TAK-653 is very potent at a low dose and has a favorable half life of 10 hours.

Basically ketamine releases glutamate which binds to AMPA in the hippocampus and causes an antidepressant effect, AMPA PAMs seem to mimick this in some studies. No, AMPA PAMs are not dissociatives as you can see by them being used to reverse ketamine induced cognitive deficits. This is the same antidepressant overlap I cover in Mood Beyond SSRIs.

Why CX-class ampakines failed

There appears to be a passive aggressive feud between RespireRx (formerly Cortex Pharmaceuticals) and Takeda, with Respire popularizing the “impact/ ampakine” theory with AMPA PAMs, and Takeda saying that Respire’s AMPA PAMs failed clinical trials because they weren’t selective enough to the allosteric region. In case you haven’t read the high impact/ low impact argument, they basically state that any AMPA PAMs to enhance binding are bad, and that their ampakines are better because they only prolong AMPA currents and don’t influence binding. My take is that they both have a point, but I side with Takeda for a few key reasons:

The only promising CX candidate, CX1739, is so expensive to produce that it would cost your rent just to get the slightest effect. This doesn’t mean it’s better, it just means it’s completely unrealistic. None of Respire’s ampakines have been clinically successful, and CX717 failed phase 2 clinical trials. This was Respire’s flagship ampakine, and I can’t blame the investors for pulling out after that. They put a ton of hype behind the impact concept, only for its effects to basically scale with how little they amplify currents… Which was their main selling point. It sounds cool in theory, to prolong currents without amplifying them, but there is no proof of concept, and it’s possible this even comes as a disadvantage.

TAK-653 potentiates currents in valuable regions, such as the prefrontal cortex during crucial moments of learning. Due to having low intrinsic agonist activity, it evades aberrant synaptogenesis that would be prone to side effects. Takeda demonstrates TAK-653’s superiority over less selective agonists by directly comparing it to LY451646, finding only enhanced therapeutic potential, benefits to cognition and safety in TAK-653. If CX717 and LY451646 are as comparable as agonists as Takeda suggests, [9] then Respire’s interpretation of AMPA PAMs may have been flawed.

The legacy of RespireRx is depressing, and while I wish them a fast recovery, I can’t help but feel their rigidness has come at a great cost. And while I can respect them wanting to pioneer a new concept, they probably should have taken a more traditional approach, like how Takeda worked on improving selectivity and pharmacokinetics.

IDRA-21 is the most selective AMPA PAM on the market, but it is far worse in terms of selectivity, pharmacokinetics (half life and dose) and has no human studies than TAK-653. All in all, TAK-653 seems like a great candidate for a powerful nootropic, with a mechanism of action that easily translates to nootropic effects in healthy people.

The executive function study, and no “drug feeling”

A randomised, double-blind, placebo-controlled, three-way crossover (placebo, TAK-653 0.5 mg and 6 mg) study with 24 healthy volunteers was performed. It was found that TAK-653 did not affect body sway and subjective drug effects as measured by B-VAS and BL-VAS at either dose level. TAK-653 0.5 mg increased saccadic peak velocity and affected Stroop difference in reaction time between correct congruent and correct incongruent answers and number of correct responses in incongruent trials. TAK-653 6 mg improved adaptive tracking and increased saccadic peak velocity and smooth pursuit. Based on these findings it can be concluded that TAK-653 demonstrated a psychostimulant-like pharmacodynamic profile on the NeuroCart consistent with previously reported increase of cortical excitability following Transcranial Magnetic Stimulation (TMS) of the human motor cortex.

Also to note from this study was a limited psychostimulant profile and no subjective feeling of the drug, which is consistent with our trials with it.

Community IQ-test gains

We saw cognitive testing scores improve consistently for all who participated. I’ve been taking it for about 3 months daily in the range of 2-6mg. What to expect: antidepressant, potential ADHD treatment, all around improvement to cognitive ability. Perhaps unique change to thinking process.

I’m not your average vendor, my guy. I have connections and an understanding of pharmacology that others do not. It’s a relatively new substance and it’s not even listed on Wikipedia. I’m sure if the CX class compounds were cheaper to synthesize the market would be flooded with them. That’s because those Respire guys did a ton of marketing. Takeda didn’t at all. I only stumbled across the compound after my friend showed me it. So I gathered up some investors and did a custom synthesis with a lab.

I don’t think Piracetam would synergize with TAK-653 unless you were trying to lower the dose of one or the other. Both are AMPA PAMs. TAK-653 aims to do what Piracetam does but far better. Aside from the cholinergic effects, which I think Tropisetron + TAK-653 would replace, not because Tropisetron increases acetylcholine, but because Tropisetron is a partial agonist of a7 nicotinic receptors which are known to meaningfully increase cognition in healthy people as well as be some of the most important overall in terms of benefits (more on Tropisetron in Cholinergic Cognition). TAK-653 will potentiate the good effects of NMDA antagonists as well as prevent cognitive dysfunction to an extent. I think TAK-653 + Tropisetron + Neboglamine and then maybe Magtein will be the best attempt to increase IQ in a healthy individual.

D-Serine

Mechanism

In the context of neurotransmission, D-Serine serves to prime the NMDAR for activation. It does this through the NMDA glycine site, which could ironically be renamed the “D-Serine site”, as there it functions as the dominant endogenous agonist. [13] Glycine and D-Serine together are called “co-agonists”, as NMDA requires either D-Serine or glycine to fire when glutamate binds. Binding to NMDAR causes either long term potentiation (LTP) or long term depression (LTD) which is the strengthening or weakening, respectively, of a synaptic connection. This is a downstream event essential to learning and memory.

D-Serine is synthesized by an enzyme called Serine Racemase, which converts L-Serine to D-Serine. This enzyme and process is also stimulated by magnesium. [54]

Early in life, glycine is used as the primary co-agonist, but it quickly transitions to D-Serine with age. [13] Crosstalk between glycine and D-Serine “fine-tunes” the NMDAR, [19] and glycine inhibits D-Serine synthesis and release. Unlike glycine, D-Serine causes internalization of NR2B, and this catalyzes an important developmental process called the “synaptic shift”. [11] Genetic removal of D-Serine prevents the synaptic shift [22] and this results in strange social behavior, [23] reminiscent of Schizophrenic phenotypes. Furthermore, Schizophrenics quite literally have less D-Serine [24] [25] and more glycine. [26]

D-Serine has the stereotypical benefits of both NMDA antagonists and glutamatergic drugs. D-Serine also stimulates adult neurogenesis [31] in regions vulnerable despite spatial constraints. [43]

Nootropic effect in humans

  • Young, healthy people: Reduces sadness and anxiety. Improves attention, learning performance and information retention. [1]
  • Old, healthy people: Improves spatial memory, learning and problem solving. Didn’t change mood. [17]
  • An outlier: Surprisingly, D-Serine failed to improve cognition in different tests that were emotionally charged, suggesting its nootropic effect may not be universally applicable. [18]
  • PTSD: Improves anxiety, depression and general PTSD symptoms. [15]
  • Parkinson’s: Significantly improves symptoms in parkinson’s patients. [16]
  • Schizophrenia: Significantly improves Positive, Negative and cognitive symptoms of Schizophrenia. Meta analysis. [8]

One should expect mild anti-anhedonic effects, a reduction in anxiety, improved attention and better recall. There may also be anti-addictive effects.

Ketamine-like antidepressant action

D-Serine has identical mechanisms to ketamine in treating depression, [21] logically through releasing glutamate by preferentially internalizing NR2B [11] which then binds to AMPA to stimulate BDNF. This triggers adult neurogenesis. [31]

Like other disorders, depression can be looked at as a learning impairment. And ironically, this is how NMDA antagonists help. D-Serine has identical mechanisms to ketamine in this regard, [21] and this can be summarized by synaptic changes and increased BDNF in the hippocampus, decreased BDNF in the nucleus accumbens. [34] D-Serine’s efficiacy as an antidepressant is shown both acutely and chronically when supplied exogenously.

Since D-Serine is so capable of enhancing learning, it can facilitate a phenomena called “fear extinction”. [32] Basically, anxiety can be looked at as a learning disorder, in where the victim is unable to draw a non-threatening association to new circumstances. By extension, PTSD would be a severe example of this.

Dosing

For a healthy person, a reasonable dose of D-Serine is 2-5g. For a Schizophrenic person, 5-9g. It has a half life of 4 hours. More recent evidence suggests that D-Serine is both safe and more effective at higher doses (~8g vs. common 2g). [10] Take 4g, once or twice a day with Magtein and L-Serine, maybe Piracetam.

Stack it with Magtein

It is my personal opinion that D-Serine should be consumed alongside Magnesium L-Threonate (Magtein), as L-Threonate reliably enhances magnesium influx through the blood brain barrier [36] which primarily inhibits extrasynaptic NMDA receptors through increased extracellular magnesium, and would target the problem at its source to offer protection as well enhance learning further. [37] Furthermore it appears the antidepressant mechanisms of magnesium are blocked by exogenous D-Serine administration [38], bolstering the argument that they are in direct competition at that site, thus supporting a need for supraphysiological levels of magnesium in the brain.

For long term users of D-Serine, it is advisable to take it alongside L-Serine and Magtein. L-Serine is also a precursor to D-Serine in the brain, however this effect is mainly seen with long-term chronic use. [50] Note: L-Serine may be sedating. A 2:1 ratio of D/L-Serine may be more desirable for daytime users.

Why it beats Sarcosine

There have been doubts about its efficiacy in comparison to Sarcosine by one Taiwanese researchers [6] [7], but the strongest form of evidence, a meta-analysis, does not reciprocate this, [8] and Sarcosine sometimes fails when used alone. [12] And strangely, Sarcosine is incorrectly given credit for D-Serine’s success on the Serine wikipedia. [9] D-Serine is anything but a failed drug. The rumors claiming Sarcosine to be a superior drug are false. If Sarcosine increases glycine, and glycine inhibits D-Serine, then perhaps that could have some unforeseen consequences.

It’s not in contradiction, Sarcosine enhances NMDAR currents through the same mechanisms. More is better than none, but that’s not my point here. Mainly I’m saying instead of relying on glycine, it’d be more efficient to just enhance D-Serine activity.

The walk-back

I’ve moved on to neboglamine instead. A while back I did a guide on D-Serine, but since then I have decided it is not good enough. That is despite it doing some very cool things. In general I’ve found out more since writing the D-Serine post.

L-Theanine + D-Serine

L-Theanine inhibits D-Serine uptake, which plays a role in its nootropic effects. Extracellular D-Serine concentration is modulated by ASCT neutral amino acid transporters. L-Theanine, a neutral amino acid component of green tea, competitively inhibited D-Serine uptake into SH-SY5Y cells with an IC50 value of 9.68 mM. It was found to be a weak competitive inhibitor of the ASCT transporters.

In plain terms: L-theanine enhances extracellular D-serine which allows more to bind to its respective NMDA site, which is nootropic.

Neboglamine

A while back I did a guide on D-Serine, but since then I have decided it is not good enough. But for a year I have been planning to make Neboglamine, and I think this will be the answer to it all.

The most interesting part about Neboglamine is that it is a NMDA glycine site positive allosteric modulator (PAM). In practice, it enhances the binding of endogenous D-Serine which is important because D-Serine is released regionally and during critical periods of learning. In theory, this more dynamic mechanism should translate to better nootropic effects. This is supported by TAK-653 being a superior AMPA PAM due to being the most selective of its class.

Why it’s the upgrade past D-Serine

At a ~50mg human equivalent dose, it would appear that Neboglamine improves learning acquisition in healthy rats, [1] [4] much like how D-Serine improved areas of short term memory in healthy young [2] and old people. [3] Since recent data is suggesting D-Serine should be dosed at over 8g, this is a big improvement. So far there has only been one comparison between Neboglamine and D-Serine, wherein a large dose of Neboglamine increased neuronal activation in similar regions as a low dose of D-Serine, but with twice the potency. [5] Due to the dose discrepancy, however, this data can’t be extrapolated.

One legitimate caveat I encountered with D-Serine was that it caused oxidative stress, even in small amounts, and that it wasn’t reversed by L-Serine in vitro. [16] On the topic of Neboglamine and excitotoxicity, it seems like it’s neuroprotective rather, based on what I’ve read about it. No signs of toxicity are found in the brain or body within dose range.

The 7→15 IQ-point combination claim

Glutamate fine tuning is basically the dynamic strengthening and weakening of synapses to form the most accurate memories. Sound complicated? That’s because it is. The dynamics between AMPA and NMDA governing thought have tons of overlap, and cannot be easily stereotyped. However, given what we know about D-Serine and AMPA PAMs, it is not a stretch of the imagination to say that a PAM of the glycine site would have added benefit. Additionally, TAK-653 and Neboglamine could even be combined, perhaps bringing a 7 point IQ increase to 15 points. This I hope to explore by following through on creating Neboglamine.

Neboglamine enhances the binding of D-Serine in the brain, which could be used as an alternative strategy to AMPA PAMs for cognition enhancement. In short Neboglamine could be used alone or alongside TAK-653 to improve executive function, with all data pointing towards less addictive tendencies, higher IQ and better mental stability. It is the only drug with this mechanism.

Clausenamide

Clausenamide is basically “herbal piracetam”. It’s a natural AMPAergic and cholinergic nootropic extracted from the fruit tree Clausena Lansium, currently being studied for cognitive decline in China. There is a language barrier, and the Chinese translation spells “Xanthamide” or “Levoxanthamide” in English, which is likely why not many people know about it.

It shows promise as a nootropic: “The synthetic single enantiomer Clausenamide could improve Ab induced impairment of spatial discrimination, potentiate basic synaptic transmission and HFS-induced LTP on either anesthetized or freely moving rats, and increase cortical ChAT activity, hippocampal synapses and Mossy fiber sprouting. In addition, was 50–100 times more active than the known nootropic drug, Piracetam, and 5–10 times more active than the racemic Clausenamide.”

It being “more active” than Piracetam really means nothing special, but ChAT, the acetylcholine synthesizing enzyme, being enhanced, is unique. In addition to its neurogenic effects, of course. Clausenamide increases thickness of the cerebral cortex and synapses density in the hippocampal CA3 region (like a 10% increase in cortical thickness in mice) and of course improves learning and memory.

Sialic Acid

Sialic acid is an essential endogenous compound that enhanced cognition in healthy rodents and is a neurogenic compound.

Sources

Write a comment