How We Sense the Heat of Chili Peppers and the Cool of Menthol [Excerpt] (2024)

Illustrations by Joan M.K. Tycko

Here’s the plan: I’m going to give you a backpack filled with Ziploc bags—some will be filled with fresh mint leaves and others with juicy habanero chili peppers.

On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

You’ll also get a clipboard, a pencil, a spare pair of socks, and a round‑the‑world airplane ticket. Your job, should you choose to accept it, is to travel around the world and visit all sorts of places, from the biggest cities to the most remote jungle encampments. In each location you will seek out a wide variety of people—young and old, rich and poor—and then rub chopped‑up mint leaves or diced chili peppers on their skin, ask them to describe the sensation, and record their responses. Try applying the samples on both the glabrous skin of the lips and the hairy skin of the forearm. (These substances don’t have to touch the tongue for their effects to be experienced.)

If you conducted this survey where I live, in Baltimore, you’d find that the dominant word used to describe the tactile experience of the habanero, smeared on either the lips or the forearm, would be hot, while for mint it would be cool. Is this merely a convenient turn of phrase, a colloquialism? After all, if we were to use a thermometer to measure the actual temperature of mint or chili peppers, we’d find that they are not literally hot or cool. And Baltimoreans (like many others) often use these words metaphorically—to mean, for example, stylish (“the Tesla Roadster looks so cool”) or sexually attractive (“Rachel Weisz is so hot”). The use of words like cool to mean “stylish” and hot for “sexually attractive” are metaphors that are specific to a particular time and place. People in Shakespeare’s time, for example, appear not to have used either of these linguistic constructions. Are “hot chili peppers” or “cool mint” also local, culturally constructed metaphors, or do they reflect some deeper biological reality? If they are merely cultural constructions, you would expect to find groups of people in your world‑traveling survey who don’t describe the tactile sensation of chili peppers as hot or mint as cool.

If, however, your survey did indeed reveal that these figures of speech are widespread, would that constitute proof that hot chilies and cool mint are biologically determined metaphors? Not exactly. To play devil’s advocate, one could imagine that, over many years with widespread communication, the idea of hot chili peppers and cool mint originated in one place and spread around the world through cultural contact. While various species of mint are widely dispersed geographically, chili peppers originated in South America and had been carried only to Central America and the Caribbean before European colonization. They were unknown in Europe, Africa, or Asia before Columbus returned from the New World. Soon after, they were spread by European powers, notably Spain and Portugal, to their other colonies. It’s hard to imagine now, but the foods of places like India and Thailand had no fiery chili pepper before the sixteenth century.At present it’s unclear if there are any places left on earth where chili peppers have not been introduced.

So, to really do this survey properly, you’d also need a time machine to take you back to, say, Thailand in the fifteenth century and do your mint/chili survey there as well.

To my knowledge, this type of ethnographic (not to mention time‑travel) survey has yet to be done, but from a biological perspective we can predict how it would turn out. Given what we know about the biology of touch, we’d predict that nearly every person around the world would describe chili peppers as hot and mint as cool, even if he or she were experiencing these tactile sensations for the first time and had never heard others describe them. It appears that the cool‑mint and hot‑chili‑pepper metaphors are biologically hardwired from birth.

The main active ingredient in mint is menthol, while its equivalent in chili peppers is a chemical called capsaicin. Less potent chili peppers, like the Anaheim, have a low concentration of capsaicin, while very strong ones, like the Bhut Jolokia pepper, can produce about one‑thousand‑fold more.So why are we biologically predisposed to perceive menthol as cool and capsaicin as hot? One possibility is that there’s a class of nerve ending in the skin that can sense cooling and a different class that can respond to menthol. The signals conveyed by these distinct fibers could then ultimately converge in the brain: Mint and cooling might feel the same because they activate the same brain region dedicated to the sensation of cooling. In an analogous fashion, separate heat‑sensing and capsaicin‑sensing nerve fibers could ultimately send their impulses to a heat‑sensitive brain region.

This hypothesis, therefore, rests on signal convergence in the somatosensory cortex, and while it’s reasonable and appealing, it’s actually dead wrong. How do we know that? First, we can record electrical signals from single sensory nerve fibers in the arm that respond to both heat and capsaicin, and other single nerve fibers that respond to both menthol and cooling. These show that temperature and chemical signals are present in the neurons that innervate the skin long before any signals reach the brain. We also have some molecular evidence. There are free nerve endings in the epidermal layer of the skin that contain a sensor on their outer membrane called TRPV1. This single protein molecule can respond to both heat and capsaicin by opening an ion channel, a pore that lets positive ions flow inside, thereby causing the sensory neuron to fire electrical spikes. Similarly, there are free nerve endings that contain a different sensor, called TRPM8, that can respond to both menthol and cooling. The answer to our puzzle is that the metaphor is not in the culture, or even in the brain region. The metaphor is encoded within the sensor molecules in the nerve endings of the skin.

How did this molecular metaphor develop? How did thermosensors like TRPV1 and TRPM8 become sensitive to plant products like capsaicin and menthol? We can’t know for certain the sequence of evolutionary events that gave rise to these two dual‑function sensors. The best guess is that TRPV1 and TRPM8 evolved in some animals as temperature sensors and that certain plants later developed compounds that would activate them in order to deter their consumption by predators. Plants that produced menthol and capsaicin would therefore have a survival and reproductive advantage and become more prevalent in the population of that species. In this scenario it’s plant evolution that initially drove the dual‑function properties of the sensors, not animal evolution.

David Julius and his coworkers at the University of California, San Francisco, have studied the molecular properties of TRPV1 and TRPM8 by using genetic tricks to force kidney cells or frog eggs grown in a culture dish to produce great quantities of TRPV1 or TRPM8 while recording the electrical signals that pass across the cell membrane when these sensors are stimulated.These studies have revealed that features of these molecules explain aspects of our everyday tactile experiences. For example, the oil of the eucalyptus tree contains a substance called eucalyptol that, like menthol, can activate TRPM8 to produce a cooling sensation. This is why eucalyptus extract is often used in soothing skin creams, mouthwash, and throat lozenges.

TRP function can also have an impact on our experience of a summer’s day at the beach. If you’re out in the sun too long, the resulting sunburn will set in motion a cascade of inflammatory processes in your skin, including the production of compounds called prostanoids and bradykinin. These chemicals have the property of reducing the temperature threshold of TRPV1 activation from 109°F to 85°F. As a consequence, when you return home from the beach and step in the shower to rinse off the remaining sand and sunscreen, the water temperature you typically select will now be too hot, and you’ll have to reduce it to avoid a painful burning sensation.

Another example involves the bird feeder in your backyard. While mammals have the standard form of TRPV1, activated by both capsaicin and heat, birds are utterly indifferent to capsaicin, as they can’t detect it at all. (Birdwatchers often spike the seeds in their feeders with chili peppers to deter squirrels, raccoons, and other mammals while leaving the birds unaffected.) When the TRPV1 gene is extracted from a bird and expressed artificially in kidney cells, it reveals a bird‑variant form of TRPV1 that responds to heat but not capsaicin. Examination of the sequence of bird DNA can pinpoint the change to the exact spot that’s necessary for capsaicin binding, located on the inner surface of the cell’s outer membrane.

Interestingly, chili pepper plants and birds appear to have reached a satisfying sort of evolutionary détente. When mammals eat chili peppers, they tend to destroy the seeds with their molars. Birds, on the other hand, don’t have molars and so pass most of the seeds through their digestive system intact. When they defecate, they spread viable chili pepper seeds to new locations. It’s a win‑win situation for birds and chilies.

A few years after the initial identification of TRPV1, several groups used genetic engineering techniques to produce mice that lacked TRPV1 and measured their responses to capsaicin and heat. These mutant mice were found to completely lack behavioral and electrical responses to capsaicin. However, their responses to heat were diminished but not completely eliminated. For example, when their tails were placed in hot water (122°F), they eventually flicked them away, but it took four times longer than for normal mice. Likewise, the ability of inflammation to boost heat sensation was reduced but not eliminated in the mutant mice.9These results indicate that there must be other heat sensors in addition to TRPV1.

Initially it seemed as if a satisfying explanation was in hand when a family of TRPV channels was identified with a range of heat sensitivities: TRPV4 and TRPV3, when expressed in kidney cells in a culture dish, responded to warm temperatures below the range of TRPV1. Conversely, TRPV2 responded to extreme heat (>125°F), well above the threshold for TRPV1. In this way, successive activation of various TRPV channels with different thresholds could potentially detect a range of skin temperatures encountered in real life, from tepid to warm to hot to painfully hot (figure 5.1). In addition to being expressed in free sensory nerve endings, TRPV3 and TRPV4 were also found in keratinocytes, the main cell type of the epidermis where the free nerve endings terminate. This suggested that the neighboring skin cells might play a role in helping the free nerve endings to detect gentle warmth. TRPV3, one of the gentle‑warmth detectors, was also shown to be activated by compounds from a wide range of spices, including camphor, nutmeg, cinnamon, oregano, cloves, and bay leaves, some of which are associated with a perception of warmth. (As a child I was an enthusiastic consumer of Red Hots, a cinnamon‑flavored candy.)

The prediction from figure 5.1 is clear: TRPV4 and TRPV3 should be required for detecting gentle warmth, and TRPV2 for extreme heat. Taken together, these three additional TRPV sensors should account for the residual heat perception present when the TRPV1 gene is deleted or the TRPV1 protein is blocked by a drug. Surprisingly, when mutant mice were created that lacked TRPV3, TRPV4, or TRPV2, either alone or in combination, they showed no significant deficit in heat perception in a wide variety of tasks. This result strongly suggests that there are even more heat detectors in the skin that we have yet to identify, and that these are likely to be molecules that are not part of the TRPV family of genes.

Figure 5.1 A family of temperature‑sensitive TRP sensors can respond to heating, cooling, and various pungent chemicals found in plants. Here, each TRP sensor is shown at a position along the thermometer where it begins to respond to heating or cooling relative to skin temperature.

Keep in mind that while core body temperature is about 99°F, temperature in the epidermal layer of the skin is about 90°F. While the TRP sensors are not identical, they share certain properties: All of them thread their way across the cell membrane six times and all have a loop structure that dips into the membrane to form an ion channel. Each TRP sensor is drawn to show its over‑ all molecular structure. It is worth noting that the thermal activation points for all the TRP sensors are not sharp thresholds—there is quite a bit of cell‑to‑ cell variation. Adapted from L. Vay, C. Gu, and P. A. McNaughton, “The thermo‑TRP ion channel family: properties and therapeutic implications,” British Journal of Pharmacology 165 (2012): 787–801, with permission of the publisher, Wiley.

A similarly murky situation surrounds the sensation of cooling. When genetic engineering was used to create mice that lacked TRPM8, they showed a complete loss of responses to menthol and eucalyptol applied to the skin and an incomplete reduction in responses to mild cooling. In particular, their responses to gentle cooling (below 77°F) were profoundly diminished, but their responses to severe cold (less than 58°F) were normal.Similar to the partial effect of TRPV1 deletion of heat sensing, this result indicates that there must be additional molecular sensors for cold, particularly severe cold, that remain undiscovered.

When rubbed on the skin, mint feels cool and chili peppers feel hot, but what is the sensation produced by horseradish, or its Japanese cousin wasabi? It’s not exactly hot, but more like a warm pungency. Wasabi, horseradish, and yellow mustard all contain a chemical called AITC (allyl isothiocyanate), which activates a different sensor of the TRP family called TRPA1.13Another TRPA1‑activating group of compounds includes allicin and DADS (diallyl disulfide), which are found in garlic and onions and account for their effects on skin sensation, including the eye‑watering response evoked by activation of TRPA1 in the cornea.

When I lived in Chicago in the 1980s, there was a great Italian bar on Halsted Street that served steamed garlic that you could smear on crusty bread and wash down with Moretti beer. The dish was prepared by gently removing some of the outer papery skin and then steaming the whole garlic bulb intact. Only after it was completely cooked would the chef cut it in half, around the equator, to allow the diner to scoop out the mild soft flesh of the plant from each clove with a special tiny knife. What chefs have known for years is that the pungent chemicals in garlic and onions—the ones that cause irritation of the skin and eyes—are produced only whenthe bulb is cut or crushed. When the bulb is intact, the enzyme that produces allicin and related pungent compounds is trapped within special compartments inside the plant cells and cannot act on its substrate. Allicin and DADS are also partially degraded by the high temperatures of cooking. This means that cooking an intact onion or garlic bulb will produce a low concentration of TRPA1‑activating pungent compounds, little skin and eye irritation, and a delicious, mild appetizer.

Figure 5.2 The TRPA1 sensor, whimsically called the wasabi receptor, is activated by a wide variety of pungent compounds from plants, most notably wasabi, horseradish, and yellow mustard, as well as structurally similar products from onions and garlic and a structurally distinct compound, oleocanthol, found in extra‑virgin olive oil. It’s interesting that several different families of plants, most notably the wasabi/horseradish/mustard family (called Brassicaceae) and the onion/garlic/leek/shallot family (called Allioideae) have independently evolved chemicals to activate TRPA1, presumably to reduce predation, although these compounds also have antimicrobial properties. Adapted from L. Vay, C. Gu, and P. A. McNaughton, “The thermo‑TRP ion channel family: properties and therapeutic implications,” British Journal of Pharmacology 165 (2012): 787–801, with permission of the publisher, Wiley.

The prickly ash tree, Xanthoxylum, is also known as the tickle‑ tongue tree or the toothache tree because its sap or berries produce a numbing, tingling sensation when ingested. Indeed, the berries of Xanthoxylum, also called Szechuan peppercorns, are prized for the tingling sensation they add to spicy dishes from this region of China. These tingles suggest an interaction with sensory neurons. In both East Asia and North America, preparations of prickly ash are used as folk medicine for their anesthetic or pain‑masking properties. The active ingredient of Xanthoxylum is a chemical called hydroxyl‑alpha‑sanshool. Considering what we have learned about the actions of other plant compounds on sensory neurons that innervate the skin, an obvious guess would be that hydroxyl‑ alpha‑sanshool activates some type of TRP channel in these cells. This, however, is not the case.

Hydroxyl‑alpha‑sanshool excites sensory neurons through a novel mechanism by blocking an ion channel called the two‑pore potassium channel. This type of channel normally allows the slow leak of positive ions out of the neuron, so that when it is blocked, positive charge builds up quickly inside the cell, ultimately causing it to fire spikes and thereby send signals to the brain. The neurons that are activated by preparations of Xanthoxylum include C‑tactile fibers, which convey light pleas‑ ant touch; the caress sensors; and the Meissner fibers, which con‑ vey vibration at moderate frequencies. It’s not entirely clear why their activation produces a tingling sensation.

Vampire bats are amazing, and not only for their fanciful roles inhorror films.We’ve discussed them briefly earlier in the context of their social grooming to solicit a shared blood meal (chapter 1), but now let’s examine their tactile specializations for feeding. Vampire bats have a unique ecological niche: They are the only known mammals whose entire food supply consists of blood from warm‑blooded animals (other mammals and birds). Some species of bat will eat insects or fruits, but vampire bats can only swallow liquids and will starve to death before they will consume a nonblood meal. Vampire bats fly to select target prey and typically alight on their backs or the crests of the necks. They then proceed to search for a suitable spot to carefully bite and extract about two teaspoons of blood. They search for a place that’s not encumbered by too much hair or fur and where blood vessels run close to the surface of the skin. This search for a buried blood vessel is the moment when the ability to detect heat at a distance is particularly useful. Ludwig Kürten and Uwe Schmidt of the University of Bonn have brought vampire bats into the lab and shown that they can detect the infra‑ red radiation emitted from living human skin at a distance of about 6 inches.

Many species of bat have a facial structure called a nose leaf, which is thought to aid in the echolocation of prey, but only vampire bats have a set of three nasal pits surrounding the nose leaf (figure 5.3). The skin of these pits is thin, hairless, and devoid of glands, making it an ideal location to house infrared sensors. The pits are also separated from the surrounding parts of the face by a layer of dense connective tissue that serves as a thermal insulator. As a result, the temperature of the nasal pits is about 84°F, substantially cooler than the 99°F temperature of the surrounding skin. This allows the heat sensors in the nasal pits to distinguish between the heat of prey and the heat of the bat’s own face.

So what sensor does the vampire bat use to detect infrared radiation?

Figure 5.3 A modified supersensitive form of TRPV1 allows vampire bats to detect infrared radiation. (A) Vampire bats, which can detect infrared radiation, have nasal pits, indicated by arrows, while fruit bats, which cannot detect infra‑ red radiation, do not. (B) The amino acid sequence of the carboxy‑tail‑end of two alternatively spliced forms of TRPV1, the super‑heat‑sensitive short form and the normally heat‑sensitive long form. The short form is highly expressed in neurons from the trigeminal ganglion that innervate the face (including the na‑ sal pits) but not in the dorsal root ganglion neurons that innervate the body of the vampire bat. (C) When artificially expressed in kidney cells grown in a dish, the supersensitive short form of TRPV1 begins to be activated at 86°F while the long form is activated only at temperatures above 109°F. Adapted from E. O. Gracheva, J. F. Cordero‑Morales, J. A. Gonzales‑Carcacia, N. T. Ingolia, C. Manno, C. I. Aranguren, J. S. Weissman, and D. Julius, “Ganglion‑specific splicing of TRPV1 underlies infrared sensation in vampire bats,” Nature 476 (2011): 88–91, with permission of Nature Publishing Group.

We already know that TRPV1 in humans and mice can detect temperatures greater than 109°F, but clearly that’s insufficiently sensitive. To identify the infrared sensor in vampire bats, David Julius, Elena Gracheva, and their colleagues performed a clever experiment. They gathered vampire bats and fruit bats (which can’t sense infrared radiation). Then they carefully dissected out the clusters of neuronal cell bodies that innervate the face (the trigeminal ganglion) and analyzed their expression of the TRPV1 gene. They found that there are actually two different forms of TRPV1 expressed in trigeminal ganglion sensory neurons: a long form, which has the conventional heat threshold of 109°F, and a short form, which is activated at a much lower temperature, about 86°F, just above the resting temperature of the nasal pits. Fruit bats have only the long form in their trigeminal ganglia, while vampire bats have both, in roughly equal measure.

The lovely result is that the vampire bat has evolved a supersensitive form of TRPV1 to enable it to detect infrared radiation for feeding. But what does this mean for the rest of its body? After all, the vampire bat needs to detect heat in other body parts as well. When dorsal root ganglia—clusters of neurons that innervate non‑ face areas—were examined, they showed only trace amounts of the supersensitive short form of TRPV1. This explains why vampire bats can maintain normal thermal sensitivity in other body regions that are not used for locating a blood meal.

I’m sure that you’ve been lying awake pondering this question: If you blindfold a rattlesnake, can it still accurately strike its prey? Thanks to a group of intrepid researchers led by Peter Hartline at the University of Illinois at Urbana‑Champaign, we know the answer. These investigators (carefully) blindfolded their rattlesnakes and placed them on a pedestal at the center of a circular enclosure. Then they induced them to strike by jiggling a heat source (the hot tip of a soldering iron) in an enticing fashion to mimic the movements of a warm‑blooded animal. The soldering iron was placed at various angles in the direction the snake was facing and just outside of striking range (about three feet away). Even with both eyes completely covered, the snake was able to strike accurately, within five degrees of the target. As the authors of this study noted, “This is very impressive, and for a mouse it is deadly.”

How does the rattlesnake accomplish this? It’s not by the sense of smell: The snakes will strike accurately at a warm, odorless object or one that is completely encased in an odor‑blocking shield. However, if you place the warm object behind a special pane of glass that blocks infrared radiation, it can no longer strike accurately. Like vampire bats, rattlesnakes can sense the infrared radiation emitted from warm objects. Rattlesnakes, however, are much more sensitive and are able to detect warm objects at a maximum distance of about 39 inches, compared to 6 inches for vampire bats. The structure that confers the infrared sensitivity of the rattlesnake is the pit organ, a small cavity located between the eye and the nostril (figure 5.4). If the pit organs on each side are covered or damaged, then the rattlesnakes can no longer strike accurately when they are blindfolded or placed in the dark. Rattlesnakes are not the only type of snake with an infrared‑detecting pit organ. They are one species of a group of related snakes called pit vipers (subfamily Crotalinae), which include moccasins, lanceheads, and bushmasters in the Americas and temple vipers and hundred‑pace vipers in Asia.

The pit organ functions like a crude pinhole camera. There is a small aperture at the front, and a thin infrared‑sensitive membrane at the rear, stretched so that there is an air space on either side of it. Figure 5.4B shows how the aperture of the pit organ restricts infra‑ red radiation so that a source at a particular point in space will strike only a small section of the pit membrane, enabling the pit organ to form a low‑resolution picture of the infrared world. The pit membrane is innervated by about seven thousand sensory fibers from the snake’s trigeminal ganglion that then carry information encoding the rattlesnake’s infrared map of the world to a part of the brain called the optic tectum, where it is combined with visual information in a fashion that aligns the visual and infrared maps.

Figure 5.4 The rattlesnake can detect infrared radiation using its pit organ, which contains a modified temperature‑sensitive form of TRPA1. (A) The pit organ is located between the eye and the nostril. (B) A cross‑section of the pit organ shows how it functions as a crude pinhole camera to localize prey. Nerve fibers from cells in the trigeminal ganglion branch in the pit membrane, which is stretched like a drumhead over an air‑filled apace. (C) An artist’s rendering of the visual (top) and infrared (bottom) sensory worlds of the rattlesnake. These two streams of information are aligned and combined in the snake’s brain. Note that the snake can detect the vague outline of the warm rabbit with infrared sense even when it is hidden in a bush (or it is dark outside). The snake’s infrared sense can be used to detect not only warm objects against a cooler background but also cool objects against a warmer background, like the frog emerging from a pond onto sun‑warmed grass. (D) TRPA1 from a rattlesnake is genetically modified so that it can be activated at temperatures above 86°F. TRPA1 from a rat snake, a species without infrared sense, is only weakly activated by warm‑ ing, and human TRPA1 is not activated at all. Panels A, B, and D are adapted from E. O. Gracheva, N. T. Ingolia, Y. M. Kelly, J. F. Cordero‑Morales, G. Hol‑ lopeter, A. T. Chesler, E. E. Sánchez, J. C. Perez, J. S. Weissman, and D. Julius, “Molecular basis of infrared detection by snakes,” Nature 464 (2010): 1006–11, with permission of Nature Publishing Group. Panel C is adapted from E. A. Newman and P. H. Hartline, “The infrared ‘vision’ of snakes,” Scientific American 246 (1982): 116–27, with permission of Macmillan Publishers.

One might guess that the molecular esensor that detects infrared radiation in the pit organ of the rattlesnake is the same supersensitive form of TRPV1 that’s used by vampire bats. However, when David Julius and his colleagues examined the trigeminal ganglia (where the sensory neurons that innervate the pit organ reside), they found that TRPV1 was not enriched there, as one would expect if it were the pit‑organ infrared sensor. However, they surprisingly did discover that the wasabi receptor, TRPA1, was elevated four‑hundred‑fold in the snake’s trigeminal ganglion. This was an odd finding, because mammalian TRPA1 is not activated by heat at all. When human and rattlesnake TRPA1 were expressed in kidney cells, heat was shown to activate rattlesnake TRPA1 at temperatures above 86°F; however, human TRPA1 was almost entirely insensitive to heat. Rat snakes, which do not possess facial infrared‑sensing organs, have a form of TRPA1 that is only weakly heat‑sensitive.23If we have come to think of TRPA1 as the wasabi sensor, it is only because we happened to study mammalian TRPA1 first. If we had a more vipercentric world‑ view, we’d say that TRPA1 is a heat sensor that can also be activated by wasabi and garlic.

Boas and pythons are from a snake lineage that is approximately 30 million years older than the pit vipers. They also have infrared‑ sensing pits, typically thirteen per side, located in two rows, one above and one below the mouth. The openings of these pits are not constricted and so they do not function like pinhole cameras. Rather, each pit has a slightly different field of view based upon its position on the snake’s face. From behavioral tests we know that pythons and boas are not as sensitive to infrared radiation as rattlesnakes are, so it was not entirely surprising to find that TRPA1 from pythons was less heat‑sensitive than that of a rattlesnake but more sensitive than rat snake TRPA1. When the sequences of the TRPA1 genes for humans, pythons, and rattlesnakes are compared, one can see that modification of TRPA1 to render it heat‑sensitive evolved twice in the snake lineage: once in the ancient boas and pythons, and then again in the more modern pit vipers.Sometimes the process of random mutation and natural selection will yield a related molecular and structural solution to a problem (like infrared sensing) in different organisms, millions of years apart. This is the wonderful process of convergent evolution.

Not all creatures use their infrared detectors to find prey. For example, most animals run or fly away from forest fires, but the fire beetles of North America (called Melanophila) are drawn toward them. It’s not a desire for self‑immolation that compels them, however. The beetles arrive at the site of a fire just as the flames have subsided and then copulate in the comfortably still‑warm ashes. The female then deposits her eggs under the charred bark of newly burned pine trees. When the fire beetle larvae hatch the following summer, they can feed on the charred wood. (Living wood has chemical defenses that make it inedible to the larvae.) In some cases fire beetles have been drawn to other hot sites, including factories and even a football game held in a stadium where lots of the spectators were smoking cigarettes. Perhaps the most dramatic infestation of this type occurred in the Central Valley of California in August 1925. When a huge fire consumed an oil tank near the town of Coalinga, huge numbers of fire beetles began to swarm toward it. Newspaper reports of the time estimated that millions of fire beetles converged on Coalinga and remained for several days after the fire was extinguished.

Because Coalinga is situated in an arid valley, the best guess is that these beetles came from a site in the western foothills of the Sierra Nevada mountains, approximately eighty miles away. Melanophila beetles have a single infrared‑detecting pit on each side of the abdomen. Many years later, when Helmut Schmitz and Herbert Bousack of the University of Bonn performed calculations to estimate the amount of infrared radiation that would have fallen on these sensors at a distance of eighty miles, they found that it was so small,it was embedded within ambient thermal noise produced by the fire beetle’s body. The nervous system of the fire beetle has a difficult engineering task to extract this tiny signal and use it to trigger migratory behavior. To date we do not know if the infrared sensors of the fire beetles use TRPV1 like vampire bats, TRPA1 like rattlesnakes, or an entirely different mechanism—perhaps not even a member of the TRP family at all.

If you type the word paradise into a Google Images search, the result will be a screen filled with one hundred different images, each one of which will be a view of a tropical beach. What’s the explanation for that? In part—at least for people living in affluent societies—a tropical beach suggests a leisurely vacation. But why, then, doesn’t the search term paradise bring up pictures of other popular vacation spots, like New York City, or a ski resort, or Disneyland? The reason is the weather: Paradise is a place where our bodies don’t have to work very hard to maintain our core temperature of approximately 99°F. We humans and other homeothermic animals (mammals and birds) cannot tolerate deviations of our core temperature of more than a few degrees. If it’s hot, we engage in both reflexive and voluntary activities: We sweat, vasodilate, have a cold drink, or jump in a swimming pool to cool our core. If it’s cold, we shiver, vasoconstrict, and put on a sweater.

These homeostatic reflexes and behaviors require that we constantly monitor our internal core temperature and the temperature of the outside world as sensed through the skin. We need to know when our skin is cold or hot enough to require a physiological response to maintain our core temperature within a narrow range. The thresholds of human TRPM8 and TRPV1 are well calibrated for this task: TRPM8 is activated at temperatures below 78°F, and TRPV1 is activated at temperatures above 109°F.

If the TRPM8 and TRPV1 activation thresholds for detecting cooling and warming are truly designed to help us maintain our core temperatures, then we should expect these thresholds to be different in animals with core temperatures unlike our own. Indeed,when DNA encoding TRPM8 from a chicken, a rat, and a frog (the clawed frog Xenopus laevis) was used to artificially express TRPM8 channels, it was shown that chicken TRPM8 was tuned to have a warmer threshold of activation: around 86°F, befitting defense of its core temperature of 107°F. The frog, which is not homeothermic and hence needs only to sense extreme cold, has its TRPM8 tuned cooler, activating only at temperatures below 66°F (figure 5.5).26 On the heat‑sensing side, thresholds also appear to be set by core temperature. For example, human TRPV1 is activated at temperatures above 109°F, but the zebra fish, which is not homeothermic, has TRPV1 that activates at about 91°F. The bottom line: The thresholds for hot and cold detection in various animals are not random. Rather, they make sense in terms of the temperature regulation each animal needs to achieve in order to function physiologically.

Figure 5.5 The threshold for TRPM8 response to cooling is correlated with core body temperature. These temperature response curves were generated through artificial expression of TRPM8 from a frog, a rat, and a chicken. Adapted from B. R. Myers, Y. M. Sigal, and D. Julius, “Evolution of thermal response properties in a cold‑activated TRP channel,” PLOS One 4 (2009): e5741, an open‑access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Years ago, I went on a family visit to my former sister‑in‑law’s house in Ohio in the late fall. When I awoke in the basem*nt guest room, it seemed more than a little chilly. After bundling up, I headed upstairs and checked the thermostat, which was set to 52°F. That was just right for her, for she was standing happily in the kitchen in shorts and a T‑shirt. The following year I made sure to bring along a portable space heater for the guest room. I have another friend who heats his house to the point where the doorframes start to warp, and you almost expect to see the sun‑bleached skull of a coyote lying on the hardwood floor.

What accounts for such extremes in individual temperature preference in humans? We know that people with a thinner layer of body fat tend to prefer warmer temperatures, which makes sense in terms of core temperature regulation. We also know that those who are more physically active, even if that activity is mere fidgeting, produce more heat through muscular contraction and thereby prefer cooler temperatures. This may partly explain why young children and teenagers are often resistant to putting on a coat. There is also a daily cyclic variation in core body temperature, and hence external temperature preference. But are there actual differences in the temperature‑sensing molecules or circuits of the skin and brain that might explain some of these individual variations? Do temperature preferences run in families? At present, the best answer to these questions is maybe.

We know that different species of animal can have TRPV1 and TRPM8 variants that are tuned to different temperatures. It has also been shown in both rats and humans that drugs that block TRPV1 can produce hyperthermia, a rise in core body temperature, and that drugs that activate TRPM8 can do the same. There is also a rare recessive mutation in humans called WNK1/HSN2. Two copies of this mutant gene produce severe degeneration of sensory neurons, but single‑copy carriers of the mutation are unaffected. However, careful measurement of heat and cool detection thresholds showed that the WNK1/HSN2 carriers have their warm threshold shifted slightly cooler and their cold threshold shifted slightly warmer than age‑ and sex‑matched controls.27But the simple result that one might expect— that genetic variation in human TRPV1 or TRPM8 could account for a portion of individual temperature preference—has yet to emerge. I strongly suspect that my former sister‑in‑law has a froglike TRPM8 gene, but that has yet to be confirmed.