The Discovery of the Endocannabinoid System

A system hiding in plain sight

Forty years of cannabis research produced almost nothing useful, and then , suddenly , it produced everything. That's not quite fair as a summary, but it's close enough to be worth sitting with. The discovery of the endocannabinoid system in the late 1980s and early 1990s wasn't a single eureka moment. It was a slow accumulation of molecular detective work, a lot of scepticism from mainstream pharmacology, and at least one researcher who simply refused to accept that a plant compound could act on the brain without there being a biological reason for it.

Understanding how that discovery unfolded tells you a lot about how science actually works. Not the sanitised textbook version, but the messy, contested, politically inconvenient version.

The problem of THC

By the mid-20th century, cannabis had been used by humans for thousands of years, but its chemistry was poorly understood. That changed in 1964 when Raphael Mechoulam and Yechiel Gaoni at the Weizmann Institute of Science in Israel isolated and characterised delta-9-tetrahydrocannabinol , the primary psychoactive constituent of cannabis. This was genuinely significant work. But it immediately raised a harder question: how, exactly, was THC affecting the brain?

The dominant hypothesis at the time was that THC was simply dissolving into cell membranes; a non-specific lipid interaction, nothing more sophisticated than that. It was a tidy answer that also happened to make the whole area feel intellectually uninteresting. If a drug works by sloshing around in lipids, there's not much of a lock-and-key story to tell. Most pharmacologists moved on to other things.

Mechoulam did not move on.

Finding the receptor: the work of Allyn Howlett and William Devane

The real breakthrough came from a different direction. In 1988, pharmacologist Allyn Howlett and her graduate student William Devane, working at Saint Louis University, published a paper demonstrating that THC binds to a specific G-protein-coupled receptor in rat brain tissue. This was not membrane dissolution. This was a receptor, discrete, specific, saturable, doing exactly what receptors are supposed to do. The paper appeared in Molecular Pharmacology and, if you follow the history carefully, marks the moment when everything changed.

But why would the mammalian brain have a receptor that perfectly accommodates a compound from a flowering plant? The obvious answer; the one Mechoulam in particular pushed hard, was that it wouldn't. Not unless there was an endogenous molecule that normally occupied that receptor. The body doesn't build locks without keys.

The receptor itself was cloned and formally characterised in 1990 by Lisa Matsuda and colleagues at the National Institute of Mental Health, and is now known as the CB1 receptor. A second receptor, found in higher concentrations in immune tissue, was identified shortly after and designated the CB2 receptor. Two receptors. Now, where were the endogenous ligands?

Anandamide: the body's own cannabinoid

In 1992, Mechoulam's team, including William Devane, who had moved to Jerusalem; isolated exactly that molecule from porcine brain tissue. It was an endogenous lipid that bound to the CB1 receptor. They named it anandamide, from the Sanskrit word ānanda, meaning bliss or joy. I've always found that naming choice quietly charming; scientists don't usually get poetic about their ligands.

A second major endocannabinoid, 2-arachidonoylglycerol (2-AG), was identified by Mechoulam's group and independently by Shimon Ben-Shabat in 1995. These two molecules, anandamide and 2-AG, form the core of what we now call the endocannabinoid system. They are produced on demand in cells, travel backwards across synapses (retrograde signalling, which is unusual), and are rapidly broken down after they act. The machinery of synthesis and degradation has its own enzymes: FAAH breaks down anandamide; MAGL handles 2-AG.

This wasn't a niche curiosity. Endocannabinoid receptors turned out to be among the most abundant G-protein-coupled receptors in the entire central nervous system.

What the system actually does; mechanistically

Here's where I want to be careful, because this is a section of cannabis science that gets mangled constantly. The endocannabinoid system is a neuromodulatory system. Its role, broadly characterised in the peer-reviewed literature, involves retrograde inhibition of neurotransmitter release, meaning it helps regulate the signalling activity of other neurotransmitter systems, including glutamatergic and GABAergic pathways.

CB1 receptors are found in high density in the brain, particularly in areas associated with memory, motor control, and pain processing. CB2 receptors are expressed predominantly in peripheral tissues and immune cells, though more recent research has found CB2 expression in the brain as well. The system also interfaces with the pharmacokinetic fate of cannabinoids after administration, how they're absorbed, distributed, and cleared.

None of this, by itself, tells you what happens when something goes wrong with this system, or what happens when you introduce a plant-derived cannabinoid. Those are separate questions, and the research answering them is far more variable in quality than the foundational receptor science.

The politics that slowed everything down

Honestly, the history here is frustrating. Cannabis was classified as a Schedule I substance in the United States under the Controlled Substances Act of 1970; meaning no accepted medical use and high abuse potential, by federal definition. Federally funded research required special permissions. The compound that had unlocked an entire signalling system in the human body was, administratively speaking, treated as if it had no scientific relevance whatsoever.

Australian scheduling followed a similar trajectory. Cannabis and its derivatives sat firmly in Schedule 9 of the Poisons Standard (the prohibited substances category) for decades, with no therapeutic scheduling framework. The regulatory environment actively constrained research and any clinical application. It wasn't until 2016 that Australia established a legislative framework for medicinal cannabis access, and even now, products containing THC remain Schedule 8 controlled drugs under the Poisons Standard, with access requiring either the Special Access Scheme or an Authorised Prescriber pathway via the TGA.

The endocannabinoid system, meanwhile, had been sitting in our brains the whole time, entirely indifferent to scheduling decisions.

Mechoulam's legacy and the open questions

Raphael Mechoulam, who passed away in March 2023 at the age of 92, is rightly regarded as the central figure in cannabinoid science. His career spanned the isolation of THC in 1964 through to active research in his final years, including work on what he termed "abnormal cannabidiol" and other non-classical cannabinoid pathways. The field he helped build now encompasses not only the classical CB1/CB2 system but a range of additional receptors; GPR55, TRPV1, GPR18, that interact with cannabinoid-like molecules.

I'll admit I spent a bit of time last winter, between cold morning swims at Sandy Bay, going back through the original 1992 Science paper on anandamide. It holds up remarkably well. The methodology is precise, the claims are tight, and the authors are careful not to overreach, which is more than you can say for a lot of the secondary literature that's built up around this field since.

Questions that remain genuinely open in the peer-reviewed literature include: the full extent of CB2's role in the central nervous system; the physiological significance of lesser-studied endocannabinoids beyond anandamide and 2-AG; and how the entourage effect hypothesis; the idea that cannabinoids and terpenes interact in ways that modify their individual activity, should be evaluated against controlled clinical data rather than in vitro observations.

That last question, in my view, is where the field most needs more methodological rigour. The entourage hypothesis is plausible. It is not yet robustly demonstrated in human trials. There's a difference between those two things, and the tendency to conflate them is one of the more persistent problems in popular cannabinoid science writing. But that's a longer argument for another article.

What isn't in dispute is the core story: a specific receptor was found, its endogenous ligands were isolated, and a signalling system was described that had been operating in vertebrate biology for at least 500 million years. That's not a small thing to have discovered inside a few decades of difficult, contested, politically hamstrung research.

Sources

• Howlett AC et al., "The cannabinoid receptor: biochemical, anatomical and behavioral characterization", NCBI / Trends in Neurosciences
• Mechoulam R & Parker LA, "The Endocannabinoid System and the Brain"; NCBI / Annual Review of Psychology
• Medicinal Cannabis Guidance Documents, Therapeutic Goods Administration (TGA)
• Lu HC & Mackie K, "An Introduction to the Endogenous Cannabinoid System", NCBI / Biological Psychiatry

, Hannah Bui, Evidence & Research-Literacy Writer