The Endocannabinoid System in Six Papers, 1964 to 1995

Hashish, 1964, one structure

In 1964, Gaoni and Mechoulam reported the isolation, the structure and a partial synthesis of the active constituent of hashish [1]. The compound was delta-9-tetrahydrocannabinol. THC, for short.

It would be easy to file that away as one more entry in plant chemistry. Look instead at what it did to the question.

Before 1964, the object of study was the plant [1]. Cannabis, hashish, resin, preparations that differ from one another by origin and by how they were made. Studying a plant means studying a mixture. You can compare samples. You cannot point at a shape.

After 1964, the object of study was a molecule with a known geometry [1]. That is a different kind of object entirely. A structure can be drawn on paper. It can be built in a laboratory instead of pulled out of resin, which is what partial synthesis describes [1]. And it can travel between research groups without anyone arguing about what was in the sample.

From there, the question turns mechanical. If a molecule with this exact shape does something in a body, then something in that body has a surface the shape fits. So: what does this shape touch.

Notice the quality of that question. It is answerable in principle. Not by debate, not by consensus, but by locating a thing. Either there is a protein with a matching surface, or there is not. The 1964 structure turned a botanical subject into a search with a defined target [1].

The same work set the naming habit that everything afterwards inherited. THC was called a cannabinoid because it came out of cannabis [1]. At that point the word was a statement about origin, nothing more.

What was missing in 1964 was the other half. No receptor. No molecule made inside the body. Just a shape from a plant, and a well-founded suspicion that the shape mattered because it fitted something.

Twenty-six years passed between that structure and the first report of the surface it binds to [1][2]. Mechoulam's name turns up again in the work on the body's own cannabinoids three decades later, in 1995 [6]. Long careers are part of this story.

The gap is what makes the sequence odd, and it is also why a signalling network inside the human body ended up carrying the name of a plant grown in a field. Chemistry arrived first. Biology caught up afterwards [1][2].

1990: inference becomes object

Matsuda and colleagues reported the structure of a cannabinoid receptor together with functional expression of the cloned cDNA [2]. It is the receptor now known as CB1.

That one word, cloning, carries more weight than it looks like it should.

Before, a cannabinoid receptor was an inference. Binding experiments point towards something; the something stays unnamed and unowned. After 1990, there was a sequence [2]. A sequence can be put into cells, which then build the receptor to order, in any laboratory that asks for the material [2]. Inference had become an object, and the object was transferable.

The phrase functional expression is worth unpacking, because it does the heavy lifting. Cells given the cloned sequence did not merely carry it. They built a receptor that worked in binding experiments [2]. Structure and function, in the same report.

That matters more than it sounds. Two groups on different continents could now work on the same receptor and know it was the same receptor [2]. Results become comparable. Disagreements become testable.

Two years later came the other half. Devane and colleagues reported the isolation and structure of a brain constituent that binds to the cannabinoid receptor [3]. The molecule was named anandamide. The body makes its own compound for that site [3].

One detail in the 1992 work explains a good deal about the delay. Anandamide does not belong to the class of messengers that fills the standard textbook chapter on neurotransmission, the water-soluble ones [3]. It sits in a different chemical family, a fatty, lipid-type signal [3].

Spend years hunting for a water-soluble messenger while the messenger is a lipid, and the hunt takes longer. The extraction chemistry differs. The methods differ. Where you look differs [3].

So, the sequence put plainly. Plant molecule, 1964 [1]. Receptor, 1990 [2]. The body's own molecule for that receptor, 1992 [3]. Twenty-eight years from the shape to the native signal that shares it [1][3].

With anandamide on the table, the rationale changed. A receptor plus a molecule the body produces itself is a signalling system in its own right, whether or not anyone ever cultivated the plant [2][3]. The plant becomes what it always was here: the thing that pointed at the door.

CB1 also did not need a number in 1990. It got one later, once a second cannabinoid receptor turned up [2][4].

Six reports, six dates

Cannabinoid from within

By the close of 1995 the parts list was short, and it hung together. Two receptors [2][4]. Two molecules made inside the body [3][5][6]. And a reason for the whole arrangement to exist that owes nothing to anyone growing hemp.

Back up two years. Munro and colleagues published the molecular characterisation of a second cannabinoid receptor, called peripheral at the time and CB2 ever since [4]. Its sequence is related to the first receptor, but not identical [4]. Read that closely and it says something structural: cannabinoid signalling does not run on one circuit. It runs on at least two, at separate addresses in the body [4].

Then 1995 delivered the second native molecule, twice over. Sugiura and colleagues put forward 2-arachidonoylglycerol as a possible endogenous ligand for the cannabinoid receptor in brain [5]. The same year, Mechoulam and colleagues reported a compound of the same type [6].

2-AG is built from arachidonic acid, joined to a chemical partner that is not the one anandamide uses [6]. Same family, different construction. Two body-made molecules, not one molecule counted twice.

Now the word itself. Endocannabinoid, read literally, means cannabinoid from within. That sounds like a claim about origin. It isn't one.

Anandamide and 2-AG were filed as cannabinoids because their shape resembles the plant compound described in 1964 [1][3][5]. Not because of where they come from. They come from us. The label points at a constituent of hashish because that constituent was on the table first [1].

So the vocabulary works like a fossil record. Read it in order and you get the sequence of discovery rather than the biology. Plant compound, 1964 [1]. Receptor, 1990 [2]. First molecule made in the body, 1992 [3]. Second receptor, 1993 [4]. Second such molecule, 1995 [5][6].

A system named after a plant it does not need. Had the receptors been found first, the words would look different: the plant compound would be the one described by resemblance, and cannabis would be the visitor that happens to fit.

One more detail worth keeping. The 1990 receptor and the 1993 receptor came from separate groups, and only together did they force the numbering into CB1 and CB2 [2][4]. What those six reports set down is a parts list, with names and dates attached [1][2][3][4][5][6]. Thirty-one years of chemistry to get the naming straight. The functional questions carried on well past 1995, and a fair number of them are still open.