When Glenn King milks centipedes, he is not going after nutrition. He is milking their poison, and it is no simple task. “We tie them down with elastic bands, bring a pair of electrical forceps up to their pincers, apply a voltage, and they expel the venom,” says King, a biochemist at the University of Queensland in Australia.

The microliters of fluid could hold the keys to a new set of pain-relieving drugs. Venoms are natural storehouses of nerve-numbing molecules, and with 400 different types of venom in his laboratory, King is at the forefront of efforts to identify analgesics in the stings of centipedes, spiders, snails and other poisonous beasts.

Large pharmaceutical companies have been struggling to synthesize alternatives to addictive painkillers such as morphine but have had trouble making molecules that home in on the specific nerves they need to target. Venoms, however, have naturally evolved to contain molecules with this kind of specificity. In laboratory animals these molecules numb nerves without harming the rest of the body. The targets that many researchers are aiming at are called voltage-gated sodium ion channels, which are common in pain-sensing nerve cells. Plugging one particular type of channel, known as NaV1.7, keeps the cell from passing a pain message to other parts of the body, as discussed in the accompanying article.

Certain venom components have just the right shape and chemical activity to latch onto a part of the channel called a voltage sensor, and that action shuts the channel. Last year King identified a venom molecule called m-SLPTX-Ssm6a that appeared to be one of the most selective inhibitors of NaV1.7 ever seen. He found it in the venom of the Chinese red-headed centipede (Scolopendra subspinipes mutilans), which can grow up to 20 centimeters long and has a pair of vicious, pincerlike claws. “If they nail you, it'll hurt,” King says. The molecule, however, had quite the opposite effect in injured mice: in experiments, it blocked pain better than morphine. But it had no unwanted effects on blood pressure, heart rate or motor function, indicating that it was not depressing the central nervous system, as an opiate such as morphine would.

King's team produced a synthetic version to see if the molecule could be manufactured as a drug. But to the researchers' dismay, this version did not work as well. King suspects that the original preparation they had made of m-SLPTX-Ssm6a actually contained traces of another active component. He is working on a further round of centipede milking to search for the mystery ingredient.

Snake venom is also a source of selective channel blockers. Anne Baron, a pharmacologist at the Institute of Molecular and Cellular Pharmacology in France, has isolated two painkilling molecules from the venom of the black mamba. “We are nearly ready for a clinical trial,” Baron says. “We have done a lot of animal tests in rodents to assess toxicity.” The mambalgins, as the molecules are called, plug a particular set of acid-sensing ion channels in peripheral nerve cells that, like sodium channels, help the cells send pain signals. Fortuitously, the mambalgins have no effect on most other ion channels, which may explain why mice injected with the substances had no apparent side effects.

Accurate nerve cell targeting is not the only goal of venom research, says David Craik, a biochemist at Queensland. If venom molecules are to be swallowed as pain pills, they need to resist degradation by the digestive system. In 2004 the U.S. Food and Drug Administration approved a painkilling drug called ziconotide that is based on a molecule isolated from the venomous cone snail Conus victoriae. But the drug could not withstand the rigors of the stomach, so it must be pump-injected slowly into patients, a cumbersome procedure. “Ziconotide hasn't been a big seller,” Craik says.

Craik has started to reengineer painkillers derived from the cone snail toxins. His strategy is to turn the molecules, which are normally chains of amino acids, into rings. Circles are much more stable structures—enzymes in the body cannot snip off the ends. He spliced the ends together and gave oral doses of the rings to rats. The compound, dubbed cVc1.1, turned out to be 100 times more potent than gabapentin, a common treatment for nerve pain. And earlier this year at the American Chemical Society meeting in Dallas, Tex., he unveiled five more ring-shaped conotoxins that have also shown durability in early studies.

With tens of thousands of venomous species in the world, researchers think it is only a matter of time until they find a compound that hits the right target, is rugged and can be easily produced in quantity. “We perhaps know 1 percent of the products that are in these venoms,” Baron says.

SCIENTIFIC AMERICAN ONLINE See an animation of a promising drug target for pain at ScientificAmerican.com/dec2014/pain