Your body makes its own signaling molecules that calm neural activity down after it's been dialed up — anandamide and 2-AG are the two best studied. But a signal that never switches off isn't a signal, it's noise. Just as important as making these molecules is clearing them once they've done their job. That clearing work is done by a small set of enzymes, and the research on what happens when you speed them up or slow them down is more complicated, and more honest, than a simple "more is better" story.
Finding the off switch
The enzyme responsible for breaking down anandamide, fatty acid amide hydrolase (FAAH), was identified and cloned back in 1996. That foundational study showed FAAH converts anandamide into arachidonic acid, effectively ending its signal. It sounds like a small technical detail, but it opened three decades of research into what happens when this switch runs too fast, too slow, or gets blocked entirely.
What happens when you block it in people
The most direct human test came from a 2026 randomized clinical trial in 100 adults with post-traumatic stress disorder. Participants took a FAAH inhibitor for four weeks, which did raise anandamide levels as expected. But brain imaging found no difference in functional connectivity or emotional task activation compared to placebo. This is a genuinely useful null result: raising the level of a signaling molecule in the blood doesn't automatically translate into a measurable change in how the brain is functioning on a scan. The relationship between blood-level chemistry and brain-level function is not as direct as it's often assumed to be.
A second molecule, and a real cost when you block its enzyme
The other major endocannabinoid, 2-AG, is cleared mainly by a different enzyme, monoacylglycerol lipase (MAGL). Here the animal research shows something worth taking seriously: a 2019 study in mice found that blocking MAGL did raise 2-AG as intended, but it also triggered reactivity in the brain's immune cells in the cerebellum and impaired fine motor coordination, through a separate inflammatory pathway. Turning up one signaling molecule came with a measurable structural cost in a different system. That's a clear example of why "raise the calming chemical" is not automatically the same as "improve the outcome." Not every attempt to block MAGL runs into that problem, though. A 2023 study designed an inhibitor that worked only outside the brain, raising 2-AG in peripheral tissue and engaging its receptor there, without producing the central nervous system effects seen with a whole-body blocker. Where the enzyme is blocked, not just whether it's blocked, appears to matter.
A backup system, and a genetic variation that changes appetite
2-AG isn't only cleared by MAGL. A 2010 study identified a second enzyme, ABHD6, that also limits how much 2-AG accumulates at neurons and immune cells in the brain, shaping both the "brake pedal" signaling at synapses and a separate role in cell migration. Having two enzymes doing overlapping cleanup work suggests this system is built with redundancy, not a single point of failure. The clearance rate of these molecules isn't fixed from person to person, either. A 2023 study in mice carrying a genetic variant that reduces FAAH activity found it amplified feeding driven by stress hormones and the hunger hormone ghrelin, through a hypothalamic pathway, while also weakening the signal that normally tells the brain "you're full." A slower off switch, in other words, can shift how strongly other hormonal systems get expressed.
The honest summary
These enzymes are the reason endocannabinoid signaling is a pulse, not a constant hum. Speeding up or slowing down that pulse changes things, but not always in the direction you'd predict, and not always with a matching change in how the brain behaves. Where the raw material for these molecules comes from in the first place is its own question, and it turns out to be shaped by what's on your plate.
Related reading
How the receptors these molecules act on are built and where they sit, and why omega-3 fats specifically change how much raw material is available, are the next two pieces of this system worth exploring. The underlying studies are also browsable, filtered to this mechanism.
Key Takeaways
- FAAH and MAGL are the enzymes that break down anandamide and 2-AG — clearing the signal is just as important as making it, and these molecules work as a pulse, not a constant hum.
- In a 2026 randomized trial in people with PTSD, a FAAH inhibitor raised anandamide as expected but produced no measurable change in brain connectivity or emotional task activation on imaging.
- Blocking MAGL in mice raised 2-AG but also triggered immune-cell reactivity in the cerebellum and impaired motor coordination through a separate inflammatory pathway — turning up a calming molecule can carry a measurable structural cost elsewhere.
- A newer MAGL inhibitor designed to act only outside the brain raised 2-AG in peripheral tissue without central nervous system side effects — where an enzyme is blocked appears to matter as much as whether it's blocked.
- A genetic variant that slows FAAH activity amplified stress- and ghrelin-driven feeding in mice — clearance speed varies person to person and can shift how other hormonal systems express themselves.
Sources
Powered by CellWell.life- 1.Molecular characterization of an enzyme that degrades neuromodulatory fatty-acid amides (1996)
- 2.Functional neuroimaging of fatty acid amide hydrolase inhibition in posttraumatic stress disorder: a randomized clinical trial (2026)
- 3.Monoacylglycerol lipase blockade impairs fine motor coordination and triggers cerebellar neuroinflammation through cyclooxygenase-2 (2019)
- 4.A monoacylglycerol lipase inhibitor showing therapeutic efficacy in mice without central side effects or dependence (2023)
- 5.The serine hydrolase ABHD6 controls the accumulation and efficacy of 2-AG at cannabinoid receptors (2010)
- 6.A genetic variant of fatty acid amide hydrolase (FAAH) exacerbates hormone-mediated orexigenic feeding in mice (2023)
Related