The first hint of anandamide was an absence. When researchers added material extracted from pig brain to a preparation of nerve-cell membranes, less of a radioactive test compound stayed attached to those membranes. Something in the extract appeared to be competing for a binding site. At that point the scientists had a signal they could follow, but no name, structure or obvious reason why the brain would contain the substance producing it.
The experiment grew out of an old puzzle about cannabis. THC, a molecule made by a plant, has pronounced effects on the mammalian nervous system. Finding the part of a nerve cell that recognizes cannabinoids raised a further question. What molecule made by the animal itself could that same part recognize?
An address for the search
By 1988 a team including William Devane and Allyn Howlett had characterized a cannabinoid-binding site in rat brain preparations. They used a synthetic cannabinoid carrying a radioactive label. The label let them measure extremely small amounts of bound compound. Binding showed properties expected of a receptor, including a limited number of sites and competition from other cannabinoids.
Four years later, the receptor provided a way of searching a tissue extract whose components were largely unknown. Imagine sorting a box of unmarked parts when you have only one piece of equipment they might fit. Testing them one by one would take a very long time. Testing groups first gives you a reason to keep one group and set the rest aside. The chemists did something similar with fractions, mixtures separated according to their chemical properties.
A fraction that displaced the probe was worth separating again. A new set of fractions could then be tested for the same activity. This way of working reversed the usual order of discovery. The group had an experimental definition of the molecule it wanted long before it could draw that molecule on paper.
Putting a structure to the signal
In the 1992 discovery paper, Devane and his colleagues identified the active compound as arachidonylethanolamide. Its structure joins a fatty acid to ethanolamine. They named it anandamide. That long fatty-acid chain looks quite different on paper from the ring structure most people associate with THC. The receptor experiment could pick up the resemblance that mattered for binding even when the two chemical drawings looked unlike each other.
Consider what a binding measurement by itself leaves open. A molecule can occupy a receptor and prevent a second molecule from binding without producing the same biological response. And a fraction can contain a whole crowd of compounds. Chemical identification and a separate test of activity were needed to connect a particular structure with a particular effect. The team compared its isolated material with synthesized anandamide as part of that work.
A muscle twitch gives another answer
The second test involved an electrically stimulated piece of mouse tissue called the vas deferens. Under the experimental conditions it produced a reproducible twitch. Cannabinoids were known to reduce that response. Anandamide also reduced it, with the response changing as the concentration increased. The radioactive probe had answered a question about competition for a site. The twitch supplied an observation about what the new compound could do in living tissue.
Nothing about a twitch tells us what a person feels, or when a neuron in an intact brain makes anandamide. Those became questions for later experiments. A review of the early discoveries describes how the field widened to include another endogenous cannabinoid, 2-AG. Looking back, the striking feature of the pig-brain experiment is its modest starting point. The researchers didn’t need to know what anandamide looked like to begin looking for it. They needed a reliable way to notice when it was there.