In experiments on mice, researchers found a chemical that prevents a naturally occurring enzyme from blocking this cannabinoid receptor, called 2-arachidonoylgylcerol, or 2-AG.
Once the enzyme, known as MAGL, is deactivated, the protein is more effective in dampening pain, say the team, led by Benjamin Cravatt of the Scripps Research Institute in La Jolla, California.
The complex human cannabinoid system is thought to hold great potential for the control of chronic pain, and could also prove useful in the treatment of anxiety, depression and even obesity.
In earlier research, Cravatt and colleagues decoded the chain of chemical reactions that acted on another cannabinoid receptor, AEA, paving the way for the development of pain-relieving medications.
But finding the key for unlocking 2-AG proved more difficult. The tools - selective and efficacious MAGL inhibitors - just weren't there, said Jonathan Long, a graduate student at Scripps and lead author of the study.
The breakthrough came thanks to a new technique for rapidly testing large numbers of chemical compounds - all potential inhibitors - called Activity-Based Protein Profiling.
One of the 200 compounds the researchers created was particularly effective in blocking MAGL, and did not appear to interfere with any of several dozen other brain enzymes.
Tests on mice showed that the new molecule - dubbed JLZ184 - increased the concentration of 2-AG in the brain, significantly reducing pain in the lab animals.
The molecule did, however, have at least two drawbacks, highlighted by the complex web of reactions in neurochemical pathways: JLZ184 also caused hypothermia, a lowering of the body temperature; and reduced movement.
These side-effects would have to be managed in any treatments developed for humans, the researchers said.