Comment deleted
Expand full comment

It might interest you to know that pyrazolam is an abandoned benzo from the 1970's that recently resurfaced as a grey market designer drug. It is supposedly a2 a3 selective and experience reports confirm that it is primarily anxiolytic with minimal sedation.

Expand full comment

Baikal Skullcap is distantly related to Lavender, which has some better-researched anxiolytic effects. I bought some lavender but haven't got around to trying it, because my ashwagandha works so well (the Himalaya brand stuff from vitamin shoppe).

Expand full comment

A quick perusal of the wikipedia page for TPA-023 (https://en.wikipedia.org/wiki/TPA-023) suggests development was stopped 'due to preclinical toxicity (cataract) in long term dosing studies'. There are two references on the wiki page - one has no mention of cataracts I could find, and the other one, 'The rise of a new GABA pharmacology' published in 2011, (scihub link: https://sci-hub.hkvisa.net/10.1016/j.neuropharm.2010.10.020) also mentions cataracts on page 4, but also without a reference as far as I can see.

Does anyone have more information about this? It sounds like an interesting molecule.

Expand full comment
Dec 8, 2022·edited Dec 8, 2022

Some speculation regarding the psychoactive effects of alcohol: I don’t know anything about the specific interactions between ethanol and GABA. But I do know that ethanol increases the permeability of cell membranes. Neurons depend on maintaining particular intercellular/extracellular ion ratios at any given point in their firing cycle, and so anything which allows ions to cross the cell membrane independent of the cell’s ion channels is likely going to effect the cell’s firing threshold.

This effect on membrane permeability is not specific to ethanol; Most organic solvents such as methanol, propanol, ethylene glycol, and diethyl ether do this. Perhaps this is why all solvents have similar psychoactive effects (depending on dose): relaxation, euphoria, delirium, loss of motor function etc.

It’s also not a coincidence that while the effective dosage of most psychoactive drugs is in the milligram range or lower, many grams of ethanol must be consumed before drunkenness occurs. A line of cocaine might be 50 mg, where as a glass of wine contains about 14 g (14,000 mg!) of ethanol. Further, there is an inverse correlation between the hydrophobicity of a solvent (and therefore it’s ability permeabilise membranes) and the dose needed for psychoactive effects. For example, propanol is several times more potent that ethanol.

This leads me to conclude that ethanol and other solvents work by altering the physical properties of the membranes rather than interacting with any receptors. Or if they do interact with receptors, then the effect is caused by altering the function of the receptor on a physical level rather than specific binding and activation/blocking/regulation that we observe with other drugs.

The reason why we drink ethanol and not other solvents is because most other solvents are toxic for reasons unrelated to their psychoactive effects. Methanol causes specific toxicity to the optical nerve and ethylene glycol causes liver failure because of its downstream metabolites.*

*Edit: Methanol also cases toxicity due to downstream metabolites

Expand full comment

Regarding alcohol, I've read some very fun studies where people were given placebo alcohol in a party setting. Turns out, most of the social effects of alcohol (apart from dizziness and passing out) are essentially placebo effect.

Expand full comment
Dec 8, 2022·edited Dec 8, 2022

"Someday we will find a selective α5 blocker, gain perfect memory, and storm the gates of Heaven."

I wonder whether anyone's ever checked the handful of hyperthymesiacs we've had for some abnormality of that subunit. There's evidence for anatomical changes in regions of their brains implicated in autobiographical memory and especially for overall better connectivity between the regions (10.1080/13554794.2011.654225, 10.1016/j.nlm.2012.05.002, 10.1037/neu0000410) but what if these were adaptations rather than causes?

Sort of like the result of a built-in Anki deck from hell: being 'forced' to long-term memorise things because they keep obsessively cycling into short term memory.

Expand full comment

Deciphering the many different subunits comprising GABA receptors is crucial for understanding basic mechanisms of GABAergic inhibition, and for developing effective new drugs. However, the notion that specific receptors map directly onto specific brain functions is a hypothesis that is unreliable and often incorrect.

Expand full comment


Expand full comment

Could be some of the a5 receptors make memory better, and some make it worse. There may be a5 receptor subtypes.

Or a5 receptors in one part of the brain make it better, and others make it worse.

Or it's involved in part of the memory pathway, so it keeps you from making new memories but helps you remember old ones. Or vice versa. Or something.

Expand full comment
Dec 8, 2022·edited Dec 8, 2022

The paucity of actually robust knowledge of what exactly is going on is stunning. We are totally in infancy of understanding basic functions and the complexity.

I find the idea of trying to brain hack basically akin to alchemy. A little bit of knowledge and hubris is dangerous. How this could possible fit into a rationalist program; I am not seeing it.

I find the report of trying doses of baikal skullcap disturbing. It conjures up Russian roulette and the case of David Hahn, a boy scout in the 90s who thought he could build a nuclear reactor in his backyard.

Or maybe let's give junior high school kids CRISPR technology and see what they come up with.

"The first principle is that you must not fool yourself and you are the easiest person to fool." ~ Richard P. Feynman

Expand full comment

I feel like this post should come w/ a codicil advising extreme caution when it comes to GABAergic drug/supplement use. Basically two classes of drugs will kill you on withdrawal: alcohol and benzos. Compensatory down-regulation and de-sensitization of specific GABA(a) receptor subunits is comparatively quick and long-lasting. That is also somewhat true with the non-ionotropic GABA(b) receptor, so messing around w/ stuff like phenibut/GHB/baclofen long-term really isn't advisable.

GABA-A receptor modulation is one area where I have comparative expertise, and a few notes I'd add are:

-The δ subunit-containing receptors are mostly extrasynaptic and modulate tonic inhibition (the baseline propensity towards excitability) and the subunits 1, 2, 3 and 5 (though a5 are more often associated with tonic inhibition) are mostly synaptic and mediate phasic (acute) inhibition.

-Ethanol is generally active at the a4/delta-containing subunit as both an agonist and positive allosteric modulator (PAM). Chronic use can make these neurons (mostly located in thalamic tracts) more prone to excitation, w/ all the attendant problems like insomnia, anxiety, &, most notably, issues w/ smooth pursuit eye movements and a shift towards saccadic vision. Curiously, some inhibitory neuroactive steroids like allopregnanolone also act as PAMs at this specific subunit type (mostly a4b3delta). In a weird way, chronic alcohol use could theoretically improve memory consolidation, but that is probably not the case given its toxic metabolites, broad and systemic effects, and pro-inflammatory nature.

-Comments on the a5 subunit seem confused. Blocking any GABA-A subunit type would probably engender hypermnesic effects but also raise the potential for excitotoxicity b/c you are effectively altering the intracellular chloride- concentration. A5 subunits are mostly located extrasynaptically (tonic inhibition) in the hippocampus, so that is why they might be especially relevant for memory.

Expand full comment

Two things that weren’t mentioned here: at the neurobiological level, a1 and a2 are localized to synapses, so they contribute to what we call “phasic inhibition” — i.e. they dampen acute bursts of activity. a4 and a5 are extrasynaptic, so they provide a kind of low-grade, constant level of “tonic inhibition”, probably by catching GABA that spills over from synapses. I have no idea how this translates to their psychiatric effects, as I am a lowly biologist, sorry.

Also, I’m surprised you didn’t mention that most GABA drugs bind not to a single subunit, but to the interface of two subunits, usually two different ones. For instance, diazepam has to bind at an alpha-gamma interface. This probably contributes to some sort of combinatorial code for the effects of different drugs (depending on which subunit combos they bind to and where in the nervous system those receptors are located). This is almost certainly why selective modulation of a single subunit type never seems to work as well as expected.

Expand full comment

I'm trying to find an Phenibut equivalent that I can take more often (really scared about tolerance/addiction), and I've found that taurine seems to work great (though less potent and empathogenic). What I'm wondering is if cycling between taurine and phenibut will help with phenibut tolerance, or will only make it worse. It seems that they're both GABA receptor agonists, but Phenibut seems to be a GABA-b agonist, while Taurine seems to be a GABA-a agonist. I'm a big noob so I don't really know how this works.

Expand full comment

In the spirit of "Fictional Drugs Banned by the FDA" I propose that TPA-023 dropped off the face of the earth because in early human trials someone lied on their screening questionnaire and took it at the same time as Ambien. In the investigation of their death and the complete trashing of their apartment it was discovered that if you combine these medications it makes the hallucinatory walruses real! The entire incident and the drug itself were immediately classified by the military so that they could investigate the combat potential of a drug that physically manifests hallucinations.

Expand full comment

Interesting that Scott got no effect from Scutellaria baicalensis. Should have tried S. lateriflora “Virginia Skullcap”.

I made a tea from just 600mg of fresh leaf of some S. lateriflora I grew and can attest to strong effects. Like getting my brain activity dial turned down to about 10%. Made it hard to read. Neither pleasant nor unpleasant. I could see how it helps in cases of panic or other extreme emotional state. Don’t understand how people find the effects desirable as a departure from ordinary consciousness however.

Expand full comment

I had insomnia and anxiety issues about decade ago. My doctor prescribed Trazodone. I'm not sure what interaction it has with the GABA areas (if any), but it's been working beautifully for me all this time.

Expand full comment

After reading this email, I felt more confused than enlightened. Here is some additional information that may clariify certain aspects of the topic. GABAergic neurons are mostly interneurons that release synaptic GABA onto postsynaptic neuronal GABA receptors. GABA receptors are ligand-gated ion channels that when activated permit a Cl- (chloride ion) to cross the cell membrane. Cl- currents are mostly inward (i.e. Cl- ions move from outside to inside the cell) and act in opposition to inward sodium (Na+), calcium (Ca+2) currents that are responsible for the action potentials (a transient change in membrane potential) that encodes neuronal signaling). Thus, GABA(a) agonists/positive allosteric modulators down-modulate neuronal signaling and have clinical manifestations as sedatives and hypnotics Conversely, GABA (a) antagonists/negative allosteric modulators up-modulate neuronal signaling and have clinical manifestations as pro-convulsants. As pointed out all GABA receptors are heteromeric pentamers, that share ~30-40% amino acid identity. There are six families of receptor subtypes, several of which have variants: α1- α6, β1-β6, γ1 –γ3, δ1, ε1, ρ1 – ρ3. As found in the brain the α1βγ2 receptor (ratio 2α1: 2β (mixed variants): 1γ2) is the most common receptor. Different receptor subtypes have different pharmacological and electrophysiological profiles. For example, the GABA binding site is located at the α:ß subunit interface, whereas the benzodiazepine binding site requires an α:γ subunit interface. By combining pharmacology, binding stoichometry and electrophysiological profiles, GABA subunit topologies can be predicted.

Expand full comment

An excellent article on Gaboxadol, the perfect sleep aid: https://harpers.org/archive/2013/08/gaboxadol

It was ignored because the owners didn't realize there was a market for sleep until Ambien came along, but then soon after they were perhaps spooked by the strange side effects of Ambien and didn't want to enter that market (summary from my hazy memory of the article).

Wikipedia states that it operates on GABA but in a way different than benzos or Z-drugs.

Expand full comment
Dec 8, 2022·edited Dec 8, 2022

"These are baicalein and baicalin, flavanoids derived from the Baikal skullcap. I tried very high doses of both of these and can report that they had absolutely no effect on me whatsoever, sorry. I’m not sure how to update my credulity about their selective GABA modulation based on this experience."

You were taking the wrong variety? 😁 I had some vague memory of reading that skullcap was used medicinally, so I looked it up. Seemingly there are two varieties, American and Chinese (the Chinese is the Baikal skullcap version) and it's the *American* one that is used as a sedative and mood-booster:


"American skullcap (Scutellaria lateriflora) is a perennial herb native to North America. In bloom, the plant is covered in tiny, tubular blue flowers, although color can vary.

The leaves of American skullcap have been used in traditional herbal medicine as a sedative and to treat conditions like anxiety and convulsions. The plant was prized by Native Americans for its powerful medicinal properties.

Chinese skullcap (Scutellaria baicalensis) is native to several Asian countries, as well as Russia.

The dried roots of this plant have been used for centuries as a traditional Chinese medicine known as Huang Qin to treat diarrhea, insomnia, dysentery, high blood pressure, hemorrhaging, respiratory infections, and inflammation.

In Asia, Huang Qin is used in herbal remedies, such as Xiao Chai Hu Tang or Sho-saiko-to (SST), a popular formulation used to treat conditions like fevers, gastrointestinal issues, and liver disease.

May boost mood and reduce anxiety

American skullcap has been shown to boost mood and reduce symptoms of anxiety.

A study in 43 people found that those who received 1,050 mg of American skullcap daily for 2 weeks reported significant enhancements in mood compared to a placebo group.

It’s thought that American skullcap positively impacts mood and reduces anxiety by stimulating gamma-aminobutyric acid (GABA), a neurotransmitter that helps calm nerves.

Notably, this plant was used in traditional medicine practices as a sedative and treatment for conditions like insomnia and anxiety."

The site does warn that you shouldn't just go out and guzzle a handful of supplements, since there can be serious side-effects. So - Chinese for inflammation, American for anxiety!

Expand full comment

Can you please comment on Gabapentin and Pregabalin?

Expand full comment

Phenibut makes me feel quite drunk, and Wikipedia says it's a GABA-B receptor agonist. I wonder if anyone has checked if alcohol affects GABA-B? The Wikipedia page for that doesn't contain GABA-B anywhere on it. Also interesting is that phenibut comes with a pleasant 'afterglow' effect the next day, which alcohol notably does not.

Expand full comment

If anyone has insomnia issues, I recommend trying oleamide. That stuff absolutely knocks me out, has a stronger effect than Ambien, and does not cause headaches next day. I use it sparingly, only when I need it. I don't know how well it work in general. Here are two interesting papers about it:



Expand full comment

I keep wanting something like, a 'users manual for my brain' that says, basically:

- when you feel X

- it is caused by systems A, B, C

- the following are things you can do to reduce/increase the extent of feeling X

Does such a thing exist? Posts like this one make me think... we barely understand these 'brain' things, and mostly know the aggregate statistical results of poking them in various places and with various chemicals, and then we kinda how how things are connected, but not really how it all works.

Expand full comment

I've read that benzo addiction can be assigned to the effect on the a1 subunits:


The sedative effect on a1 increases dopamine levels in the mesolimbic dopamine system, which is basically the definition of an addictive drug. Anecdotally, I've talked to a few former heroin addicts who subsequently got hooked on high doses of ambien, perhaps because it's also a fast acting sedative that's rapidly calming and rewarding in a similar way.

The first a2, a3 specific benzo (MRK-409) still had weak activity at a1, which might still be enough to cause the effect. TPA023 seemed more promising because it was an agonist of the a2 and a3 receptors but an antagonist of the a1 receptors. So it should have been a non-sedating, non-addictive anxiolytic. I have no idea what went wrong in the trials.

That said, I'm sure TPA023 would still cause tolerance and dependence, since most psych meds do. It's not just an issue of dopamine, there's also the basic principle of homeostasis -- provide too much of one chemical and the body/brain gradually downregulates the receptors for that. Many different drugs give unpleasant withdrawal, whether that's benzos or SSRI's or beta blockers or zyrtec.

Withdrawal is abnormally bad for benzos, for some reason, but that's not a function of the psychological addictiveness, it's probably more because GABA/glutamate balance is so important to normal brain function. I suspect that quitting TPA023 wouldn't be as bad as normal benzo withdrawal, but you'd probably have weeks or months of horrific rebound anxiety.

Zopiclone/Lunesta do seem to be more benzo-like than the other Z drugs. At least, the withdrawal from them is often benzo level bad, while the withdrawal from ambien usually is not. That might have something to do with the GABA receptor subunits, but it's also just a function of the half life of the drug -- zopiclone has a half-life of 4-6 hours, ambien is closer to 2 hours. So if you take lunesta nightly, you're stimulating the gaba receptors somewhat, at all times. If you take ambien nightly, the receptors return to baseline for most of the day. Halcion is the shortest acting benzodiazepine, with a half-life similar to ambien. The Ashton manual for benzodiazepine withdrawal says that one also has lower dependence than most, and can generally be quit without tapering, unlike the longer acting benzos.

Expand full comment

>there are botanicals claiming to have α2- and α3- selective GABA modulation. These are baicalein and baicalin, flavanoids derived from the Baikal skullcap. I tried very high doses of both of these and can report that they had absolutely no effect on me whatsoever, sorry. I’m not sure how to update my credulity about their selective GABA modulation based on this experience.

Flavonoids are notorious for having poor bioavailability as well as a tendency to nonspecifically mess up biochemical assays, so I tend to be skeptical of any research involving them.

Expand full comment
Dec 8, 2022·edited Dec 8, 2022

The effect on a5 makes sense. For months I was taking a now-out-of-stock GABA-melatonin-l-theanine-CBD pill. During that time, I suddenly felt that my memory was the greatest it has ever been (32-year-old male); it was shocking to me how smooth and automatic my recall was. When I ran out of the pill, I switched to a CBD-melatonin pill, and soon my memory went back to its baseline (or became worse) — to the point of making me feel brain-fogged and frustrated. GABA gave me what felt like a clean 5.5-7 pt IQ increase.

Expand full comment

I always thought that alcohol is an NMDA antagonist, like ketamine.

Expand full comment

I'm not sure that perfect memory it is the power necessary to storm the gates of heaven :)

Expand full comment
Dec 12, 2022·edited Dec 12, 2022

I take Escitalopram (SSRI) since two years now, prescribed for helping me getting to sleep more easily. It improved a lot of my decades long personal issues like angst, endless thought cycles, bad self confidence etc. Then a year ago because of a very toxic work enviroent I started to take Zolpidem for getting even a little sleep at all. Initially I took the full prescribed 10mg dose which played havoc with my mood and temper the next days. I got anxious, my brain worked pretty bad, I was searching a lot for words or concepts, got extremely impatient and showed a lot of nerves the following day. I reduced the dose step by step to 2.5mg which now is working extremely well for me.

I'm not sure if I am a normal guy or an ADHD or an Autistic or a hyper sensible person, but the combination of the above things did virtually change my life.

The experiment is not exactly a RCT and there is quite some coffeeine and ethanol at play as well but just saying...

Expand full comment

Wouldn’t α1β2γ2 have 1 alpha not two and 2 gamma not one?

Expand full comment