HIGH EFFORT MEMANTINE GUIDE

[tension]

Note: This is an example of a mechanistic synthesis based on previously published pharmacology and pharmacokinetic extrapolations; this is not a clinical trial for this particular drug regimen. Therefore, use this information as examples of logic, not established fact

1- Introduction​

---

​

Hey guys, I’ve seen a lot of people using memantine on discord or tiktok stacks, however they are usually at doses that are way too high and poorly timed. So I thought it was worth putting together a proper breakdown on how to actually use it. There is a lot of noise around this drug and most of the discussion gravitates towards the approved Alzheimer’s dosing, which is, to put it plainly, the wrong frame of reference for what most people here are trying to achieve when using to only prevent excitotoxicity.


The goal with memantine is to preferentially block tonic activity of extrasynaptic NMDA receptors, specifically targeting the GluN2B subtype of these receptors while leaving the synaptically located NMDA receptors in tact which are critical for memory formation (consolidation), LTP (long term potentiation) and normal cognitive functioning.
If you appropriately select the appropriate target with the correct amount of drug, you may be able to inhibit excessive glutamatergic signaling. However, if you fail to do so by using too much of the drug or because you have no knowledge of the drugs' pharmacokinetics, you will start to degrade the very same systems you intended to preserve.

In addition to explaining how to determine the appropriate concentration of drug required for selective blockade of the GluN2B subtype of NMDA receptors without reaching a CSF level equal to or greater than IC50 concentrations, the authors explain how to prevent unwanted spillover into D2 receptors, α7 nicotinic acetylcholine receptors and also avoid blocking synaptic NMDA receptor

2- Memantine IC50 Target Values​

---

​

Memantine is not a pure, single target drug. It binds to many receptor systems in the body. The plasma levels vary with each system, and therefore each has a unique range for IC50 values. Thus, there exists a narrow plasma concentration window on which low-dose every-other-day (EOD) dosing is based. The colors used in these tables represent the same meanings as they do universally.

Receptor	IC50 (nM)	IC50 (μM)	Significance
Extrasynaptic GluN2B (NMDA)	~22 nM	0.022	PRIMARY TARGET pathological tonic activation
Synaptic GluN2A (NMDA)	~800 nM	0.80	AVOID essential for LTP and memory
Dopamine D2 (High affinity)	137–917 nM	0.14–0.92	Avoid dopaminergic side effects, confounds mechanism
α7 nicotinic AChR	340–5,000 nM	0.34–5.0	Unreachable at any standard dose
5-HT3 receptor	~1,000 nM	~1.0	Unreachable at any standard dose

The pharmacologically relevant difference between the extracellular IC50 for GluN2B receptors (approximately 22 nM), and the closest receptor of interest (synaptic GluN2A receptors) at approximately 800 nM, represents a 36 fold pharmacological window. This represents the pharmacologically relevant window we are working in. (Limapichat et al., 2013; Traynelis et al., 2010):

3- Why 2.5 mg Every Other Day Emerges as the Preferred Regimen AND Why Standard Clinical Dosing Is the Wrong Reference Point​

---

​

Efficacy vs. Selectivity - proposed minimal dose of 2.5mg every other day (EOD) has been suggested as the lowest effective dose that provides a balance between significant activation of the receptors with maximum selectivity toward those receptors for glutamate binding without being efficacious.

The standard dosage of Memantine includes administration of Memantine in amounts ranging from 10-20mg per day. With xr formulations, it may be increased up to an amount of 28mg. When the drug was first developed for treatment of moderate to severe Alzheimer's disease, by the time patients had progressed to this stage in their disease they had typically lost almost all or all of their GluN2B NMDA receptors present on their synaptic surfaces through pathology mediated by amyloid beta & tau.

Therefore when most of the NMDA synaptic signaling is gone it is safe to block NMDA receptors with high amounts of Memantine. Your case isn't like that. You want to reduce pathological signals from NMDA receptors outside of synapses while still keeping your synaptic function intact. Taking 10 – 20 mg of Memantine will actually hurt your cause.


What Standard Doses Actually Hit

Regimen	SS Plasma (nM)	Extrasyn. GluN2B	Synaptic GluN2A	D2 Receptorv
2.5 mg EOD	~27.5 nM	Targeted	Spared	Not engaged
5 mg QD (steady state)	~110 nM	HIT	Still spared	Borderline
10 mg BID (steady state)	~609 nM	HIT	Affected	Engaged
28 mg XR QD (steady state)	~909 nM	HIT	Blocked	Fully engaged

Aside from 2.5 mg which will be calculated in Section 4, there is a considerable amount of clinical evidence demonstrating that standard dose memantine can have adverse cognitive effects in mild Alzheimer’s patients including impairing memory consolidation and reducing long-term potentiation (LTP). These failed drug trials in mild Alzheimer’s (MEM-MD-22, NMDAT) patients were almost certainly due to high-dose NMDA blockade blocking the remaining functional synaptic GluN2B in the patients. The dose used in the failed trials was inappropriate for the stage of the disease, not the mechanism.

4- Pharmacokinetics & The Source Data​

---

​

Free base formulation all conversions were made using a molecular weight of Memantine free base = 179.30 g/mol (the form of Memantine found in plasma after dissociation from its hydrochloride salt). Although the articles used the hydrochloride salt of Memantine (MW= 215.76 g/mole) when describing the tablets administered to patients, they reported plasma levels as the equivalent of the free-base of Memantine; therefore, the molecular weight of the free base of Memantine should be used as the conversion factor.


Conversion formula: nM = (ng/mL ÷ 179.30) × 1000

Regimen	Cmax (ng/mL)	Cmax (μM)	Cmax (nM)
Single 2.5 mg (extrapolated)	~3.06	0.017	~17.0
2.5 mg EOD steady state (extrapolated)	~4.93	0.0275	~27.5
Single 5 mg	6.20	0.035	34.6
Single 10 mg	11.60	0.065	64.7
Single 20 mg	25.34	0.141	141.3
5 mg QD steady state (Day 14)	19.69	0.110	109.8
10 mg IR BID steady state (Day 29)	109.19	0.609	609.0
28 mg XR QD steady state (Day 29)	163.06	0.909	909.4

5- Extrapolating 2.5 mg EOD The Math (skip if not interested)


Memantine demonstrates proven linear pharmacokinetics across a wide dose-range (5–20 mg) such that Cmax and AUC increase proportionally with increasing dose.

5.1 Source parameters​

The 0.25 × weekly-dose ratio gives us the steady-state average concentration (~27.5 nM plasma); note this is the steady-state average, not the peak. The peak/trough is dependent on the frequency of administration (48 hours), and an EOD schedule allows for a greater degree of drainage of drug compared to daily dosing

Single Dose	Cmax (ng/mL)	ng/mL per mg
Half-life (t½)	65 h	midpoint of published 60–70 h range
Elimination rate constant (kₑ)	0.693 / 65 = 0.01066 h⁻¹	first-order
Dosing interval (τ)	48 h	every other day (EOD)
Single 2.5 mg dose peak (C₀)	~17.0 nM	extrapolated (see §2)
Weekly dose ratio	8.75 mg/wk ÷ 35 mg/wk = 0.25×	2.5 mg EOD vs 5 mg QD
Weekly dose ratio	~0.52	published (~0.46–0.56)
Average	-	1.222

5.2 Step 1: Single Dose Cmax​

The right way to extrapolate down to 2.5 mg is to utilise all three published single-dose points, not simply half of one point. To effectively estimate a reduction to 2.5 mg based on per milligram Cmax averages across the dataset provides an advantage over utilising only one measured point:

Single Dose	Cmax (ng/mL)	ng/mL per mg
5 mg	6.20	1.240
10 mg	11.60	1.160
20 mg	25.34	1.267
Average	-	1.222

Single 2.5 mg = 2.5 × 1.222 = ~3.06 ng/mL

3.06 ÷ 179.30 × 1000 = ~17.0 nM

The use of a single-dose IC50 (22 nM) is less relevant than the steady-state IC50 when administered once per day. Halve the single dose to get ~3.10 ng/mL = 17.3 nM or ~2%, therefore either method can be used; however, the average for all data points will yield the most accurate steady-state value.

5.3 Step 2: EOD Steady State​

Average steady-state plasma concentrations of the 5 mg QD anchor (Day 14) = 109.8 nM
Weekly dose comparison:
The weekly doses are: 2.5 mg every other day (EOD), with a total of 8.75 mg/week, vs. 5 mg daily (QD), for a total of 35 mg/week

Ratio = 8.75 ÷ 35 = 0.25×

C_ss,avg = anchor (5mg QD steady state)x weekly ratio / 2.5 mg EOD steady state = 0.25 × 109.8 = ~27.5 nM


~27.5 nM steady-state average, > than the IC50 value of 22 nM; thus, receptor interaction becomes pharmacologically feasible.

5.4 Step 3: Trough (48h dosing interval)​

t½ = 65 hours implies that steady-state will be achieved in about three half-lives, or roughly two to three weeks (~195 hours).
Gradual titration is embedded within the drug's pharmacokinetics.

Reference: Theoretical accumulation factor R at EOD dosing


Trough = 27.5 × e^(−0.693 × 48 ÷ 65) = 27.5 × 0.600 = ~16.5 nM


Accumulation factor (R): R = 1 / (1 − e^(−kₑτ)) = 1 / (1 − e^(−0.01066 × 48)) = 1 / (1 − 0.600) 2.5×

Peak (C_max,ss): C_max,ss = C₀ × R = 17.0 nM × 2.5 = ~42.5 nM (plasma)


Trough (C_min,ss): C_min,ss = C_max,ss × e^(−kₑτ) = 42.5 nM × 0.600 = ~25.5 nM (plasma)


The plasma concentration decreases below 22nm. As you will learn about CSF in the next section, this does not matter because it still functions. Steady-state conditions are achieved approximately 3 half lives = 195 hours (approximately 2-3 weeks) which is consistent with the duration of the "slow accumulation window" described in the article.


Internal Consistency Check: Peak − Trough = 42.5 − 25.5 = 17.0 nM = precisely one single 2.5 mg dose. At steady state, the body clears out precisely enough of drug to equal one complete dose per 48 hour interval; thus with each subsequent dose administered, the concentration is restored back up to oscillate about the 27.5 nM average between 25.5 nM and 42.5 nM. The above equation verifies that the numerical mathematics presented herein is internally consistent.

Week	% Steady State	Approx Plasma (nM)	Above IC50?
Week 1	~40%	~11 nM	Not yet
Week 2	~70%	~19 nM	Borderline
Week 3	~95%	~26 nM	Yes

5.5 Step 4: Translation to CSF (the receptor compartment)​

Metric	Plasma	CSF (× 0.52)
Peak	42.5 nM	~ 22.1 nM
Average	27.5 nM	~ 14.3 nM
Trough	25.5 nM	~ 13.3 nM

6- Full Dose Comparison Table​

Dose	Single Cmax (ng/mL)	Single Cmax (nM)	SS Peak (nM)	CSF Peak (nM)	SS Trough (nM)	CSF Trough (nM)	vs IC50 22nM (SS)
2.5 mg EOD	~3.1	~17.3	~27.5	~14.3	~16.5	~8.6	1.25
5 mg QD SS	19.69	109.8	~109.8	-	-	~57.1	5
10 mg BID SS	109.19	609.0	~609.0	-	-	~316.7	27.7
28 mg XR QD SS	163.06	909.4	~909.4	-	-	~472.9	41.3

7- Why We Avoid D2 and What Hits It​

7.1 D2 Receptor IC50 = 137–917 nM​

At standard therapeutic doses (10–20 mg/day), memantine plasma levels reach 609–909 nM at steady state well within D2 range. This produces:

• An increase in dopaminergic activity within both pre-frontal cortex and striatum.

• Potential inhibition of prolactin release via D2 agonist action.

• Variable neuropsychiatric side effects among sensitive individuals.

• Interference with glutamate-based mechanisms being targeted.

7.2 What Dose Starts Hitting D2?​

Regimen	SS Cmax (nM)	D2 IC50 Low End (137 nM)	D2 Status
2.5 mg EOD	~27.5 nM	5× BELOW low end	NOT ENGAGED
5 mg QD (steady state)	~110 nM	Near low end	BORDERLINE
10 mg BID (steady state)	~609 nM	4.4× ABOVE low end	ENGAGED
28 mg XR QD (steady state)	~909 nM	At high end of range	FULLY ENGAGED

Key threshold: D2 engagement begins at approximately 5 mg QD (borderline) and is firmly engaged at 10 mg BID standard dosing. 2.5 mg EOD maintains a ~5× safety margin below D2 low end.

8- What Would Hit the α7 nAChR​

8.1 α7 nAChR IC50 = 340–5,000 nM (0.34–5.0 μM)​

You may have seen on social media the fearmongering memantine antagonising the α7 nAChR receptor. However this receptor range is so high that even standard clinical doses barely reach the low end.

8.2 What Dose Starts Hitting α7 nAChR?​

Regimen	SS Cmax (nM)	vs α7 Low End (340 nM)	α7 Status
2.5 mg EOD	~27.5 nM	12× BELOW	NOT REACHED
5 mg QD (steady state)	~110 nM	3x BELOW	NOT REACHED
10 mg BID SS	~609 nM	1.8× ABOVE low end	PARTIAL
28 mg XR QD SS	~909 nM	2.7× ABOVE low end	PARTIAL
Hypothetical ~60 mg+ dose	~5,000+ nM	At high end	FULL BLOCK

The α7 nAChR is not meaningfully engaged at any standard clinical dose of memantine. Full blockade would require supratherapeutic dosing well above 20 mg. At 2.5 mg EOD, it is not a clinical consideration.

9- Receptor Selectivity, Why 2.5 mg EOD Hits Only Glutamatergic​

The therapeutic window for 2.5 mg EOD is ~17–28 nM (single dose to steady state peak). The following table maps every relevant receptor IC50 against this window:

Receptor	IC50 Range (nM)	2.5mg EOD SS (nM)	5mg ED SS (nM)	10mg BID SS (nM)	Status @ 2.5mg
Extrasynaptic GluN2B (NMDA)	~22 nM	27.5	110.0	609	HIT
Synaptic GluN2A (NMDA)	~800 nM	27.5	110.0	609	SAFE
Dopamine D2 (High)	137–917 nM	27.5	110.0	609	SAFE
α7 nicotinic AChR	340–5,000 nM	27.5	110.0	609	NOT REACHED
5-HT3 receptor	~1,000 nM	27.5	110.0	609	NOT REACHED

10- The Plasma:CSF Ratio​

---

Technically, the receptor compartment is interstitial fluid (ISF), and not CSF. However, CSF serves as a reasonable proxy for ISF due to an exchange-free assumption that is relatively valid for small lipophilic molecules like memantine but has yet to be validated. Furthermore, this is particularly true in cases where localised clearance failure occurs and is modeled


In order to determine whether this lower-bound threshold exists within central compartments, peripheral PK data need to be scaled into CNS distribution using established partition coefficients (Valis et al., 2019) [7]. Clinical studies examining simultaneously collected human plasma and CSF memantine concentrations at steady state demonstrate a consistent central/peripheral ratio of approximately 45% to 56%. Additionally, they report an approximate mean distribution coefficient of 0.52 (Valis et al., 2019) [7]


CSF concentration ≈ plasma x 0.52

At 2.5 mg EOD steady state average (~27.5 nM plasma):

CSF peak ≈ 42.5 x 0.52 = ~22.1 nM

CSF average ≈ 27.5 x 0.52 = ~14.3 nM

CSF trough ≈ 16.5 x 0.52 = ~8.6 nM


Thus CSF fluctuates between ~13 and ~22 nM, while averaging ~14.3 nM. While projected peak, average and trough central values are technically at or below the population average baseline IC50 (~22 nM), modeling this threshold as a static "pharmacologic failure" fundamentally misrepresents the complex dynamics and state-dependent biophysics of open channel blockade (Rammes et al., 2008) [1].

11- Mechanistic Evaluation of 2.5 mg EOD Memantine For Anti-Excitotoxicity​

A mechanistic hypothesis, presented with its supporting biophysics and its explicit, unmeasured boundaries.

11.1 Overview​

In moderate-to-severe Alzheimer’s Disease, Memantine is taken at 10–20 mg per day for a steady-state concentration in Cerebrospinal Fluid (CSF), approximately 0.5–1.0 µM (Rammes et al., 2008; Valis et al., 2019).

To provide a protective barrier (a "shield") against the effects of excessive excitatory neurotransmitter activity that does not arise from Alzheimer’s Disease but may produce neurotoxicity, this study examines whether a much lower concentration range of Memantine could be used.

This study describes a model based upon the proposed administration regimen of 2.5 mg every other day (EOD), designed to create an anti-excitotoxic filter in the central nervous system with a low nanomolar (~14 nM CSF) concentration of Memantine, rather than presenting measured or empirical results. By scaling human peripheral pharmacokinetics, we estimate that the steady-state plasma concentration of Memantine would be approximately ~27.5 nM. Using the previously determined CSF/plasma partition ratio of ~0.52 we then predict that this will yield an approximate steady-state central nervous system (CNS) concentration of approximately ~14 nM. This framework relies on state-dependent receptor kinetic processes as opposed to dose dependent saturation in the periphery: We use the rapid off rate properties of Memantine combined with its strong voltage dependence to allow for discrimination between prolonged pathological stimulation and transient physiological signaling. Further, because the dose is below the level where synaptic trapping and cognitive impairment occur at higher doses (Parsons et al., 1999; Blanpied & Johnson, 1997; Wise & Lichtman, 2007).

11.2 The Biophysical Rationale: Retention of Block During Magnesium Collapse​

Memantine and magnesium will both be removed from the channel when the membrane depolarises transiently. However, the primary factor separating them is their kinetic properties. Unlike magnesium, memantine does not rapidly exit the channel due to the sustained and moderate degree of depolarisation associated with long-term excitotoxicity (Parsons et al., 1999). Quantitatively, memantine experiences ~70–80% of the transmembrane field (δ = 0.80 ± 0.02; Parsons et al., 1999, Fig 4B), and less than 6 % of this block is independent of voltage – compared to 47 % for the toxic blocker MK-801 (δ = 0.46). Memantine’s selectivity results from its rapid release after depolarisation as well as strong voltage dependence.



Voltage dependence of block (Woodhull model, verified Parsons 1999 parameters). Memantine (δ=0.80, 6%
voltage-independent) recovers toward full current on depolarization, sparing synaptic transmission; MK-801 (δ=0.46) retains a ~47% block floor it cannot shed; the mechanistic basis of its toxicity. Shown at each drug's characteristic concentration to display the voltage shape; this illustrates the block's voltage behaviour, distinct from occupancy at the 14 nM target.



This schematic for the pathological extrasynaptic leak shows a result of a “reduced (partial)” which reflects a submaximal block. This degree is the concentration and IC50-dependent and unmeasured at low nanomolar levels


11.3 Rejecting a Static Occupancy Figure: The Kinetic Filter Model
A naïve Hill-equation calculation, f = [D]/([D]+IC₅₀), with a "negative potential" extrasynaptic IC50 of about 22 nM (Rammes et al., 2008, referencing Zhao et al.) estimates approximately 39% occupancy at 14 nM. This static estimate is denied in this study for three reasons:

1. The IC₅₀ is not a constant. it can shift based on the applied voltage, presence of magnesium or other agonists (Parsons et al., 1999), it can shift based on the applied voltage, presence of magnesium or other agonists Phillips et al. (2026) demonstrated that memantine does so in a manner distinct from ketamine, with memantine acting to stabilise a calcium dependent desensitisation of NMDARs which results in increased efficacy during high Ca2+ states associated with pathological activity. Validated IC50 values range from approximately 22 nM to less than micromolar. Using multiple validated IC50 values in the above described Hill Equation will result in estimated percentages ranging from very few percent to near 40% occupancy.
2. Trapping decouples membrane-bound drug concentration from bulk concentration. Memantine is a type of uncompetitive open channel blocking agent with trapping characteristics. When the agonist dissociates from its binding site within the receptor complex, the gates shut over the trapped memantine molecule. Memantine has been shown to have limited slow release (~1/6 on washout) (Blanpied & Johnson, 1997).
3. It is the wrong model. The honest characterisation is qualitative: at ~14 nM, memantine acts as a sub-saturating kinetic filter, throttling rather than plugging the chronic calcium trickle.

"memantine blocks extrasynaptic NMDARs (IC50 = 22 nM) 100 times more potently than the synaptic NMDA receptors at negative membrane potentials (IC50 = 2.5 µM) and the block of both types of NMDA receptor was strongly voltage-dependent"

- Rammes, Danysz & Parsons, 2008 (via Zhao et al.)

"around 15% of channels released memantine following agonist removal"
- Parsons, Danysz & Quack, 1999 (confirming Blanpied & Johnson, 1997)



See graph below. Occupancy/evidence context, which is a rendered occupancy vs IC50 plot, goes here: same 14 nM gives ,39% occupancy at IC50=22 nM down to 2% at 800 nM, the quantitative backbone of rejecting a single fixed percentage.


shows occupancy at 14 nM swinging from ~39% to ~2% across the valid IC₅₀ range.



Because the gate can close over bound memantine (~1/6 released on washout; Blanpied & Johnson, 1997), channel occupancy is not a clean equilibrium function of bulk CSF concentration.

11.4 Why Low Dose Is the Right Strategy: Selectivity, Not Just Magnitude​

The major reason for advocating a low nanomolar drug level is based on selectivity rather than potency; hence, 2.5 mg every other day (EOD) may provide the optimal balance of being insufficiently engaged versus overly blocked. This is a conscious decision and should be evaluated independently from the issue of whether or not 14 nM provides adequate protection (See Section 4.6). These two issues are independent: this section advocates using a lower drug level as the right approach; however, whether or not the specific endpoint is high enough to protect against disease is an open question.


This position aligns with how the developers have framed memantine's mechanism. Parsons, Stöffler & Danysz (2007) n describing what they termed a "restoration of glutamatergic homeostasis" with respect to the glutamate system through a "signal-to-noise ratio," noted that "a little stimulation is good but too much is bad" when discussing memantine's ability to selectively suppress pathological tonic "noise" in glutamate systems, while simultaneously conserving phasic synaptic "signals." Thus, based upon this perspective the goal of therapy is clearly defined as restoring balance and not achieving maximal blockade of glutamate receptors; i.e., the same selective over magnitude rationale that this section presents.


In support of the rationale stated above, increasing the dose of memantine will destroy its value. Glutamate transmission has been described as being part of a biphasic, inverted-U relationship: as the concentration of glutamate increases, so does the degree to which memantine occupies NMDA receptors within synapses involved in normal transmission and thereby also those NMDA receptors essential for long-term potentiation and learning. As such (Wise & Lichtman, 2007) demonstrated in their studies of glutamatergic transmission in animal models that increased doses produce impairments in both spatial memory and performance in mazes. Therefore, at higher concentrations, memantine can be said to represent "dirtier" treatment since it only adds a greater blockade of the pathological pool of glutamate receptors at the expense of the physiological pool it was intended to preserve


"memantine elicited an inverted U-shape dose-response relationship, with low doses (0.3 and 0.56 mg/kg) reducing the number of errors committed during the retrieval test, while high doses (3 and 10 mg/kg) disrupted maze running"

- Wise & Lichtman, 2007

Going low inverts this trade-off. Because memantine's action is voltage-dependent and kinetically fast (its developers note that a true agonist-concentration/use-dependence of equilibrium block is disputed and was likely an early experimental artifact Gilling et al., 2007 so the load-bearing properties are voltage-dependence and fast kinetics, not use-dependence), lowering the ambient concentration concentrates what limited occupancy the drug achieves onto the channels that remain open longest the tonically-active extrasynaptic pool while leaving the briefly-opening synaptic population almost untouched:

• High-voltage synaptic clearance: a physiological burst delivers a sharp, intense voltage transient; memantine's low affinity and fast off-rate let that transient eject the drug, so the learning signal passes uninhibited.
• Selective low-frequency attenuation: the tonic extrasynaptic leak lacks that sharp voltage spike, so the drug is not ejected there and quietly filters the background activity.

The lower the concentration, the cleaner this partition becomes. The objective is not maximal blockade but the highest achievable ratio of pathological-to-physiological block, and that ratio improves as dose falls, the opposite of a conventional "more is better" antagonist. This is why a low dose is good: it is the strategy that preserves the synaptic network the treatment exists to protect.

"memantine blocks extrasynaptic NMDAR-mediated currents induced by bath application of 100 µM NMDA/10 µM glycine with a twofold higher potency than its blockade of the NMDAR component of evoked EPSCs; this effect persists under conditions of pathological depolarization in the presence of 1 mM extracellular Mg2+"

- Xia, Chen, Zhang & Lipton, 2010



A secondary, disease-tuned layer (calcium-dependent potency)

Beyond the kinetic/voltage mechanism, memantine carries a second selectivity feature that operates specifically under pathology. Phillips et al. (2026) show that memantine potency increases with intracellular Ca2+, its IC₅₀ falls from 2.76 µM at resting calcium to 0.70 µM under high (pathological) calcium, via a Ca2+-dependent desensitised state. Critically, this potentiation is not significant across the physiological calcium range (P = 0.09) and becomes significant only at pathological calcium (P < 0.01), making it genuinely disease-selective. The effect is GluN2A-specific. It is supporting context, not evidence for the low-dose target: even maximally potentiated, the IC₅₀ (0.70 µM) remains ~50× above 14 nM.

"inhibition of recombinant and native NMDARs by memantine, but not ketamine, increases with increasing intracellular Ca2+ concentration"

- Phillips et al., 2026

"A modest increase in [Ca2+]i resulted in a noticeable, but nonsignificant, decrease of memantine IC50 (P = 0.09). Larger increases consistent with pathological conditions resulted in larger and highly significant increases in memantine potency (P < 0.01)"

- Phillips et al., 2026

"memantine inhibition of GluN1/2B receptors was entirely insensitive to [Ca2+]i"

- Phillips et al., 2026



Cleanest proof the calcium-potentiation is GluN2A-specific: IC₅₀ 2.76–>0.69 µM for GluN1/2A vs 0.90–>0.83 for GluN1/2B (flat); 2C/2D flat.


This figure displays the <1 nM calcium IC₅₀ (2.76 µM) as the tallest bar, i.e. it visually leads with memantine's weakest potency, the number least favourable to the low-dose case. Prefer Table 3 above, which carries the same GluN2A-specificity point as numbers without that visual emphasis. If you do use Fig 1C–D, frame it explicitly as showing the calcium-tuning range (the gap from 2.76 to 0.70 µM = the disease-selectivity), not absolute potency.

11.5 Extrasynaptic NMDARs Are a Druggable Target, Not Only a Genetic One​

The premise that extrasynaptic NMDAR (eNMDAR) over-activation drives excitotoxic cell death is well established: synaptic NMDAR activity is anti-apoptotic and pro-survival (via nuclear Ca2+/CREB signalling), whereas extrasynaptic activity triggers CREB shut-off, mitochondrial collapse, and death (Hardingham, Fukunaga & Bading, 2002; Hardingham & Bading, 2010). This pathological tonic activation arises when glutamate clearance fails, loss of astrocytic GLT-1/EAAT2 produces large rises in extracellular glutamate and excitotoxic neurodegeneration (Rothstein et al., 1996; Tanaka et al., 1997).

"Synaptic NMDA receptors have anti-apoptotic activity, whereas stimulation of extrasynaptic NMDA receptors caused loss of mitochondrial membrane potential and cell death"

- Hardingham, Fukunaga & Bading, 2002

"The loss of glial glutamate transporters GLAST or GLT-1 produced elevated extracellular glutamate levels, neurodegeneration characteristic of excitotoxicity, and a progressive paralysis"

- Rothstein et al., 1996

"EAAT2 takes up released glutamate to maintain the extracellular glutamate concentration below neurotoxic levels. In excitotoxicity glutamate is released in higher concentration and glutamate clearance is decreased, causing excessive activation of glutamate receptors resulting in Ca2+ influx and activation of a cascade of phospholipases, endonucleases, and proteases such as calpain that can lead to apoptotic or necrotic cell death"

- Fontana, 2015

Crucially, much of this foundational evidence is genetic or relies on bath-applied glutamate. Savchenko et al. (2016) provide pharmacological proof-of-concept: a memantine–gold-nanoparticle conjugate (AuM), engineered larger than the synaptic cleft, selectively inhibited eNMDARs while sparing synaptic transmission, and was neuroprotective in NMDA cytotoxicity, oxygen–glucose deprivation, and Aβ models, outperforming free memantine.

"efficiently and selectively inhibited extrasynaptic NMDARs (eNMDARs), while having no effect on sNMDARs and synaptic transmission"

- Savchenko, Braun & Molokanova, 2016

11.6 Methodological Limitations (explicit, unmeasured inferences)​

1. The concentration extrapolation. The approximately 14nM target is a model: the approximately .52 partition ratio was used with an estimated plasma concentration of about 27.5nM. This is NOT a direct measure of central microdialysis concentration under the conditions of 2.5mg of memantine given every other day. Data for concentrations measured in human cerebrospinal fluid (CSF) typically fall below the IC₅₀ even when dosed at 20mg per day (Valis et al., 2019).
2. The cell-survival inference. The assertion that reducing tonic currents by less than full saturation can prevent apoptosis is entirely based on a hypothesis. All measured concentrations at which memantine neuroprotection has been demonstrated are in the low micromolar range: (Parsons et al., 1999) reported 1.1-1.4µM;(Rammes et al., 2008) reported about 1µM; (Ferrer-Acosta et al., 2022) reported 1-3µM; Phillips et al. (2026) found 3µM as the half maximal effective dose. Therefore, the 14nM target is approximately 50-200x lower than all concentrations tested that provided protective effects. ~50–200x below any concentration at which protection has been demonstrated.
3. The mechanism may not be through NMDA receptors at all in some cases. In those contexts where benefits were seen from low doses of memantine but outside of excitotoxic disease, the observed effects appear to be nicotinic (through α7 nAChR /PI3K), rather than mediated through NMDA receptors (Bali et al., 2019; Bruszt et al., 2021; Ferrer-Acosta et al., 2022)

"memantine is neuroprotective when it targets α7 nACh-R at concentrations below 3 µM. At higher concentrations, memantine inhibits the neuroprotective NMDA-R and becomes neurotoxic"


- Ferrer-Acosta et al., 2022

12- Summary​

Low-dose memantine is framed not as a non-selective blocker but as a use- and voltage dependent "kinetic shield" because of its high degree of fast-off kinetics (i.e., it rapidly releases from the receptor) and a significant degree of voltage dependence (δ = .80). As such, it remains bound to the receptor during prolonged periods of pathological depolarisation while it is released into solution after transient periods of physiological activity, thereby discriminating between these two patterns based upon the temporal characteristics of the activation event as opposed to the subunit composition or mass concentration of the drug. The specific 2.5 mg every other day / 14nM formulation is presented as an educated guess based upon mechanistic considerations alone. That is, the predicted peak concentration of the drug is calculated using standard pharmacokinetic principles (e.g., half-life, clearance rate, etc.) and therefore represents an estimate, rather than an actual measurement. Moreover, the notion that this concentration would be sufficient for neuroprotection has not been demonstrated empirically, nor does it appear to be within the range of concentrations that have been shown to produce therapeutic effects.

13- Conclusion​

Using published information regarding the pharmacokinetics of memantine, estimated potencies, and logical constructs related to selective extrasynaptic modulation, I suggest that the theoretically optimal dosing regimen for memantine would be 2.5mg administered every other day. This particular regimen represents a theoretical maximum, i.e., a dosage level that exists at a point in time where additional increases would result in sufficient broad blockade to preclude meaningful extrasynaptic modulation.

While there are many limitations inherent in what I present here, including my reliance on modelled predictions (as opposed to direct measurements) and my recognition that virtually all evidence demonstrating neuroprotective effects associated with memantine have utilized significantly higher concentrations, I believe that my analysis provides a valid framework for testing a clearly articulated hypothesis: namely, that administering 2.5 mg of memantine every other day will enable a unique form of extrasynaptic modulation that cannot be achieved without appreciable broad blockade.

14- Sources​

• Rammes G, Danysz W, Parsons CG (2008). Pharmacodynamics of memantine: an update. Curr Neuropharmacol 6(1):55–78.
• Parsons CG, Danysz W, Quack G (1999). PMID 10462127.
• Phillips MB, Povysheva NV, Neureiter EG, Nigam A, Harnett-Scott KA, Hell JW, Aizenman E, Johnson JW (2026). State-specific inhibition of NMDA receptors by memantine… Science Advances 12(22):eaec3154. DOI 10.1126/sciadv.aec3154.
• Blanpied TA, Boeckman FA, Aizenman E, Johnson JW (1997). Trapping channel block… J Neurophysiol 77(1):309–323. PMID 9120573.
• Wise LE, Lichtman AH (2007). The uncompetitive NMDA receptor antagonist memantine prolongs spatial memory… Eur J Pharmacol 575(1–3):98–104. PMID 17850786.
• Xia P, Chen HV, Zhang D, Lipton SA (2010). Memantine preferentially blocks extrasynaptic over synaptic NMDA receptor currents… J Neurosci 30(33):11246–11250.
• Valis M, et al. (2019). Front Pharmacol 10:943.
• Ferrer-Acosta Y, et al. (2022). Toxicology in Vitro 84:105453.
• Bali ZK (2019); Bruszt N (2021). Rat ultra-low-dose α7-nicotinic cognitive enhancement.
• Hardingham GE, Fukunaga Y, Bading H (2002). Nat Neurosci 5(5):405–414. DOI 10.1038/nn835.
• Hardingham GE, Bading H (2010). Nat Rev Neurosci 11:682–696.
• Rothstein JD, et al. (1996). Neuron 16(3):675–686. PMID 8785064.
• Tanaka K, et al. (1997). Science 276:1699–1702.
• Savchenko A, Braun GB, Molokanova E (2016). PMID 27490923.
• Fontana ACK (2015). J Neurochem (review).
• Chen HS, Lipton SA (1992). PMID 1432103.
• Parsons CG, Stöffler A, Danysz W (2007). Neuropharmacology 53 (2007) 699e723

 

[tension]

what the fuck bro

 

[Zaied]

Nigga nobody is reading allat

 

[BatemanNigga]

What does it do

 

[tension]

What does it do

idk

 

[BatemanNigga]

idk

Bruh