The marine whiff of ambergris. The citrusy tang of grapefruit. The must of “corked” wine. The human nose can detect a virtually infinite palette of odors, some at vanishingly low concentrations. But puzzlingly, our bodies only use about 400 receptor proteins to interpret them. Now, fragrance researchers in Switzerland have landed on a new way to study the proteins in the laboratory—and their results, they say, challenge a foundational theory of how smell works.

For decades, scientists have struggled to get cells commonly used in laboratory settings to express the genes that encode olfactory receptors (ORs), proteins primarily found on neurons in our nasal cavities. Using a process they describe today in Current Biology, researchers at the Swiss fragrance and flavorings company Givaudan say they have tweaked lab-friendly cells into readily expressing ORs. The result was an in vitro system for identifying specific ORs, including those that strongly respond to molecules in ambergris, grapefruit, and corked wine.

The Swiss group’s discovery, other olfaction researchers say, stands to make ORs much easier to study. But more controversially, the group also claims to have observed patterns of receptor activity that call into question combinatorial coding, a long-standing hypothesis of olfaction that helped Linda Buck and Richard Axel win a Nobel Prize in 2004.

Combinatorial coding holds that multiple ORs act in concert to pick up different parts of an odorant molecule, creating patterns or codes that are recognized by the brain. Beyond that, says neuroscientist Joel Mainland of the Monell Chemical Senses Center, the model is “pretty vague on the details.”

It has been hard to test, because olfactory neurons can’t be cultured in the lab. Determining which OR detects which odorant required extensive tests in rodents, and it’s not ideal “to have to sacrifice an animal each time you want to do an experiment,” says Claire de March, a chemist at CNRS, the French national research agency. As a result, investigators were left with many so-called orphan receptors whose ligands, or binding molecules, are unknown.

Buck and Axel had cloned the genes that encode ORs, an advance published in 1991, raising hopes that ORs could be studied in lab cell lines. But researchers have had a hellish time getting human embryonic kidney (HEK) cells, a workhorse for such experiments, to express the genes.

The Swiss investigators, led by Andreas Natsch, took a new approach to the problem. Last year, Natsch and his colleague Roger Emter showed that by making changes to the end of an OR’s chain of amino acids—what’s known as the C-terminal domain—they could boost that receptor’s expression in HEK cells. Their new paper extends that concept. By modifying all the approximately 400 known human ORs with an engineered C-terminal domain, they could supercharge expression for many of them. The team monitored how the ORs responded to hundreds of mostly natural odorants whose human percepts—how they smell to us—had been previously characterized. The new study details the results for 20 receptors.

Natsch’s team found that some of the odorants preferentially activate a single OR—meaning fewer ORs, or maybe even just one, may be needed to perceive a given odorant. For instance, one formerly orphan OR detects a key molecule in ambergris, an iconic perfumery ingredient that originates in the bowels of sperm whales. Another is activated by what perfumers call woody smells, and a third by both patchoulol, the principal odorant of patchouli, and a synthetic patchouli. Two structurally distinct molecules in grapefruit turn out to ping the same receptor. Previously, “we had a very noisy signal,” Natsch says, “and now can see that it’s much more specific.”

Natsch maintains that his group is not arguing that multiple receptors can’t be involved in the smell of a molecule. Rather, their findings indicate a single OR can be responsible for a “specific odor direction.”

A new study makes use of a variant of HEK 293, a line of human embryonic kidney cells commonly used in laboratory research.Iznewton/Wikimedia Commons

Mainland says the C-terminal engineering tweak—which the Swiss team seeks to patent—is a potential breakthrough that academic labs will want to work with. “If this replicates, then it’s a pretty big jump in the ability of these assays to find ligands for receptors,” he says. The advance could also benefit perfumery. Makers want to know which molecules are essential to achieve an odor, because a given ingredient might at any time prove unsafe, costly, or hard to source.

But Mainland and de March both say the findings still need to be confirmed. For one, the C-terminal tweak didn’t work for every receptor tested. And for some important perfumery odors, such as eucalyptus and sandalwood, the search for matching ligands came up empty. Mainland notes that Natsch and his colleagues did not explain how the human percepts for each odor were measured. “If you’re gonna claim it’s a fruity receptor, block it and see if fruity disappears,” he says.

Both researchers also agree that the combinatorial code model appears safe for now. The Swiss team “is basically arguing about proportion,” Mainland says. “Are you mostly activating a small number of receptors or are you mostly activating a broad set? But maybe things are more narrowly tuned than we thought.”