The Chemistry of Smell and Taste

In the spring of 1991, on the upper floors of the Hammer Health Sciences Center at Columbia University in New York, two researchers were staring at a gel. Linda Buck and Richard Axel had run a clever variation on the polymerase chain reaction, using degenerate primers (short DNA probes deliberately built with a bit of slop in their sequence so they could latch onto many related genes at once) against complementary DNA from rat nasal tissue. What came off that gel was the first glimpse of a large, previously unknown family of genes for olfactory receptors, the protein machines that let an animal smell. In that moment, the chemistry of smell stopped being a vague mystery and became a tractable problem in molecular biology.

That shift is the subject of this article. We tend to treat smell, taste, vision, and the burn of a chili pepper as separate and slightly magical experiences, but underneath each sits a piece of well-understood chemistry: a molecule, or a particle of light, meeting a protein and changing its shape. The question this article answers is how a world made of molecules gets translated into the electrical language of the brain, and why the answer turns out to be roughly the same story told four different ways.

What a Sense Receptor Actually Does

Every one of your senses depends on a particular kind of protein called a sensory receptor. A sensory receptor is a protein that converts a physical or chemical stimulus into an electrical signal inside a neuron. The stimulus might be a molecule of coffee aroma drifting up your nose, a sodium ion landing on your tongue, or a single particle of light striking the back of your eye. Whatever the input, the receptor's job is the same, namely to produce a change in voltage across a nerve cell membrane, because voltage is the only currency the nervous system trades in.

The receptors that read chemistry (the ones for smell, for taste, and for the heat of capsaicin) work by binding. A molecule, called the ligand, fits into a pocket on the receptor and sticks there for an instant through ordinary intermolecular forces, the same hydrogen bonds and van der Waals attractions that govern any molecular encounter. Vision works slightly differently, reading light through photochemistry rather than binding. But in both cases the next step is identical in spirit. The binding event, or the absorption of a photon, causes the protein to change its three-dimensional shape, a process called a conformational change. That shape change is the switch. In some receptors it directly tugs open an ion channel, letting charged particles flood across the membrane. In others it sets off a relay called a G-protein cascade, a chain of molecular messengers that amplifies the original tiny signal into something the cell cannot ignore. Either way, a chemical event has become an electrical one.

Olfaction and the Combinatorial Code

The discovery Buck and Axel made in 1991, which won them the Nobel Prize in Physiology or Medicine in 2004, was startling for its scale. They found not one or two smell receptors but an entire family of them. In humans, about 400 of these genes are functional, making olfactory receptors one of the largest gene families in the genome. Each one encodes a particular kind of protein, a seven-transmembrane G-protein-coupled receptor, meaning a single chain of protein that threads back and forth through the cell membrane seven times and signals through a G-protein on the inside.

Here is the reason a few hundred receptors can do so much. A given odor molecule does not have its own private receptor. Instead, each odorant activates some combination of receptors, switching a handful on while leaving the rest quiet. One smell might light up receptors number 12, 88, and 301; another might light up 12, 88, and 412. The brain does not read any single receptor as meaning "rose" or "gasoline." It reads the overall pattern, the particular chord struck across the whole array. This is called a combinatorial code, and combinatorics is exactly why it is so powerful, because the number of possible combinations grows explosively with the number of receptors. By current estimates the human nose can distinguish on the order of a trillion distinct odors, all from roughly 400 receptor types. It is the same trick that lets a few dozen letters spell every word in a language.

Four Chemistries Drawn on a Single Page

One of the quiet pleasures of this corner of chemistry is that four different senses can be laid out side by side on a single diagram, because they are variations on one theme. In the nasal epithelium, the lining high inside the nose, sit the olfactory receptors. On the tongue sit the taste receptors, divided into two families with the technical names T1R and T2R. In the retina, at the back of the eye, sit the visual pigments, rhodopsin in the rod cells and the cone opsins that handle color. And distributed across many tissues, including the mouth, sits a channel called TRPV1, which responds to capsaicin and to heat.

Four locations, four classes of protein, but one underlying logic. In each case a stimulus arrives, a protein changes shape, and a current flows. Keeping all four in view is a useful corrective to the way we usually learn about the senses, one at a time and in isolation, as though smelling and seeing had nothing to do with each other. Chemically, they have a great deal to do with each other.

Taste: Five Basic Modalities, Several Mechanisms

Taste is more modest than smell. Where olfaction juggles hundreds of receptors, taste recognizes just five basic modalities, namely sweet, salty, sour, bitter, and umami. The last of these, umami, is the savory, brothy taste of glutamate, and it is the most recent addition to the list. The Japanese chemist Kikunae Ikeda isolated and named it in 1908, identifying the compound responsible in the seaweed broth that flavors so much of Japanese cooking. Umami is simply the Japanese word for a pleasant savory taste, and it took most of a century for it to be fully accepted in the West.

What makes taste a nice teaching example is that its five modalities do not all use the same machinery. Sweet and umami are detected by receptors in the T1R family, which are G-protein-coupled receptors much like the smell receptors, sensing whole molecules of sugar or glutamate. Bitter is handled by the T2R family, a set of about twenty-five G-protein-coupled receptors, a number that makes evolutionary sense, since bitterness often signals a poisonous plant compound and it pays to detect many different toxins. Sour and salty, by contrast, skip the G-protein machinery entirely and work through ion channels directly. Sourness is the detection of acidity, and it is sensed by channels responsive to hydrogen ions, the very ions that define an acid. Saltiness is sensed by sodium-sensitive channels of a type called ENaC, which simply let sodium ions in when they are abundant. Two GPCR families and a couple of ion channels, between them, account for everything your tongue can taste.

Vision: One Molecule Catches a Photon

Vision belongs in a discussion of chemical senses even though its stimulus is light, because the first event in seeing is a genuine chemical reaction, one of the fastest in biology. Inside every rod and cone cell sits a small molecule called retinal, derived from vitamin A, covalently bonded to a large protein called opsin. In its resting state the retinal is bent into a shape called the 11-cis configuration. When a photon strikes it, the molecule absorbs that energy and straightens out, snapping into the all-trans configuration. This is an isomerization, a change in the geometry of a molecule without any change in its atoms, and it happens with astonishing speed, in around 200 femtoseconds, where a femtosecond is a millionth of a billionth of a second.

That tiny flick of a molecule is the entire trigger for sight. The change in retinal's shape forces a change in the surrounding opsin protein, which sets off a G-protein cascade that, in this case, hyperpolarizes the cell, pushing its voltage in the negative direction and so signaling that light has arrived. Color vision comes from having three versions of cone opsin, each tuned to absorb most strongly at a different wavelength, with peaks at about 420, 530, and 560 nanometers, corresponding loosely to blue, green, and red light. The biochemist George Wald worked out this chemistry across a long career running from 1933, when he found vitamin A in the retina, to the 1960s, and received the Nobel Prize in 1967 for it.

When Hot Means Heat: Capsaicin and TRPV1

Now for the sense that is not really a sense. When you bite into a chili pepper, the burn you feel is not a taste at all. There is no "spicy" receptor among the five tongue modalities. The molecule responsible, capsaicin, instead activates a channel called TRPV1, a non-selective cation channel (one that lets various positive ions through rather than selecting for just one) that the physiologist David Julius cloned in 1997, work recognized with the Nobel Prize in 2021.

The revealing detail is what else opens that same channel. TRPV1 is, first and foremost, a heat detector. It opens when the temperature climbs above about 43 degrees Celsius, right around the threshold where genuinely hot things start to feel painful. Capsaicin works by latching onto this channel and tricking it into opening at ordinary body temperature, so that your brain receives the exact signal it would get from real heat and pain. This is why a chili pepper and a sip of too-hot coffee can feel uncannily alike, and why we describe the sensation as "hot" in both senses of the word. The language was right all along, because chemically there is one channel doing both jobs.

A Century of Untangling, and One Persistent Myth

The full molecular picture came together slowly, across what we might call a long twentieth century of sensory chemistry. Ikeda named umami in 1908. Wald found vitamin A in the retina in 1933 and pieced out the photochemistry of vision over the following decades. Buck and Axel cloned the olfactory receptor family in 1991. Julius cloned TRPV1 in 1997. And the T1R and T2R taste-receptor families were identified between 2000 and 2002, completing the molecular roster of the chemical senses only a couple of decades ago.

It is worth ending with a correction, because one of the most widely taught "facts" about taste is simply false. You may have seen the tongue map, the diagram claiming that sweet is detected at the tip of the tongue, salty along the front edges, sour further back, and bitter at the very rear. It is a teaching myth. It descends from a 1901 German study whose data were modest and easily misread, and it spread through English-language textbooks largely because of a 1942 mistranslation by the psychologist Edwin Boring, who turned tentative regional differences into hard zones. In reality all five basic tastes can be detected across the whole tongue. The honest version is less tidy than a colored map, but it is correct, and accepting it is a small lesson in how a clean diagram can outlive the evidence that should have killed it.

Key Takeaways

The chemical senses all run on one logic, in which a stimulus meets a protein, the protein changes shape, and an electrical signal results, with smell, sweet, umami, and bitter routed through G-protein-coupled receptors while sour, salty, and the heat of capsaicin work through ion channels. Smell is the showpiece, using about 400 olfactory receptors discovered by Buck and Axel in 1991 (Nobel 2004) to distinguish roughly a trillion odors through a combinatorial code rather than one receptor per smell. Taste recognizes five basic modalities, sweet, salty, sour, bitter, and umami, the last named by Ikeda in 1908, split between the T1R and roughly twenty-five T2R receptor families and a pair of ion channels. Vision turns on the 11-cis to all-trans isomerization of retinal in about 200 femtoseconds, worked out by Wald (Nobel 1967), with three cone opsins peaking near 420, 530, and 560 nanometers. Chili heat is not a taste but the heat-and-pain channel TRPV1 (cloned by Julius in 1997, Nobel 2021) being tricked into opening below its normal 43-degree threshold. And the familiar tongue map is a myth born of a 1942 mistranslation, since every basic taste is detectable everywhere on the tongue.