Smell and Taste Reception in Animals
“In the land of the skunks, he who has half a nose is king.”
—Chris Farley, comedian
Most animals have evolved to be very discriminating about what they eat. If we smell or taste something incredibly rotten, for example, we avoid eating the rest of it. This response more than likely evolved as a means to deal quickly with the “I eat that” binning process I discussed in Chapter 2. All of our senses probably help us make these quick “I eat that” decisions, but how we do it has a deep evolutionary history. Remember that vertebrate brains have three basic layers of organization. The innermost, most primitive layer probably has its roots in our ancestral connection to early vertebrates, and it holds the brain stem and the cerebellum. The next layer, consisting mostly of the limbic system, adds complexity to how such information as smell and taste is interpreted. And the final layer, the cortex, adds to an even more refined way of interpreting the information from our senses.
How our sense of taste is interpreted in this layered brain is a great example of how these three layers are integrated. Taste interacts with our reward system through what neuroanatomists call the cortico-basal ganglia-thalamic loop. This loop is a set of neural tissue pathways that traverse the three major brain regions implied by its name. The most prominent reward systems in vertebrate brains are the gamma-aminobutyric acid (GABA) and dopamine wired neurons. GABA and dopamine are two small molecules that make their way to our brains and interact with receptors embedded in the membrane of neurons to trigger action potentials in the brain. The dopamine neurons in particular have a huge role in the evolution of how animals use the reward system.
Pleasure is a big part of training organisms to repeat things that are beneficial to them. It makes sense that if something is both beneficial and pleasurable, an organism will seek more of whatever triggered this reaction, such as sex, something that tastes good, or, sadly for the long run, a pleasure-inducing drug. When we taste something rotten, our brain dopamine levels fall drastically, telling us that we don’t want to taste any more of the bad stuff. But if we taste something nice and sweet or very nutritious, dopamine levels in the brain increase, and the reward system responds by saying, “I want more.” The dopamine tells us to ingest as much as we can, but because it is temporary, at a certain point we are satisfied and stop ingesting the item. Drugs such as cocaine and heroin exploit this system and hijack the brain. Instead of causing a transient dopamine concentration, the molecule establishes itself at a high plateau level, producing a craving for more of the drug on a scale that goes on a runaway course and results in addiction.
How this system starts is very similar to how odors are processed—it begins with small chemicals or molecules and a chemoreceptive sensation. Tastes emanate from the combinations of small molecules that we ingest with food or in the air or in beverages and are processed by their interaction with taste receptors in the mouth. That information is then transmitted to the brain, where the information is interpreted. The repertoire of taste receptors consists of five basic kinds: bitter, sweet, umami, sour, and salty. Carbonation and fattiness also probably qualify. Taste receptors occur predominantly on the tongue but have also been found elsewhere, for example, in the tissues of the airway and in the small intestines.
Researchers have characterized several kinds of taste receptors that influence sweet, umami, and bitter taste reception, called TASs (named so for the first three letters in the word “taste”), that act a lot like the odorant receptors for smell. TAS1s are involved in sweet and umami reception, and TAS2s are involved in bitter taste reception. Receptors for salty and sour have also been proposed, but less is known about them. There is one major candidate for sensing salty, and it is a gene that is also, oddly enough, involved in polycystic kidney disease (PKD) called PKD2L1 (2L1 indicates the kind of PKD gene involved). It is an ion channel receptor that is also involved in the sensing of acid. Ion channel receptors are proteins that reside in the membrane of the nerve cell and are responsible for moving specific kinds of ions across the cell membrane. In so doing, these ion channels start the taste response by initiating action potential as the ions move across the membrane. Other receptor molecules for both sour and salt are sodium channel (SC) receptors, called SCNNs, of which three (SCNN1a, SCNN1b, and SCNN1g) are thought to be the major conveyors of information about salt and sour to the brain. Suffice it to say that there may be more sour and salty receptors out there. Like odorant receptors, the number of genes for these signaling molecules that are found in the genomes of animals is an interesting phenomenon. Humans have about seventy of the sweet, umami, and bitter receptor genes (the TAS1s and TAS2s), and a few genes are found in the sour and salt categories. The animal world, however, is much more interesting when it comes to taste (fig. 4.1).
In all likelihood, animals are not discerning enough to place tastes into five categories and simply place them in three: Yum! (“I like that”), Yikes! (“I don’t like that”), and Ho-Hum (“I don’t care”). Those of you who are cat lovers can try to bribe your feline friends with a sweet. Try it, and you will find that your cat doesn’t give a meow about sweets. This lackadaisical attitude of felines toward sweets is not a part of their cool cat demeanor but rather the result of the fact that cats simply can’t taste sweets well. Sweet taste reception is implemented by two of those seventy or so TASs, called TAS1R2 and TAS1R3. In animals that can taste sweet (humans included), these two TASs form a coupled protein. When something sweet enters the mouth, the sugary compounds from the sweet item bind to the coupled protein, and this produces a signal that goes directly to the brain and to the cortico-basal ganglia-thalamic loop, where it is interpreted as “I like that” very likely because sweet things have lots of important carbohydrates in them. Cats have a long deletion in the TAS1R2 gene that knocks out the function of this receptor protein, causing it to be classified as a pseudogene (a gene that is present in the genome but does not make a proper protein). This simple genomic change kills most of the cats’ chance for detecting sweet tastes, although some researchers think that cats may be able to detect very large doses of sugar as sweet. Researchers know that the loss of this long stretch of the TAS1R2 gene occurred in the felines’ common ancestor, which means that the big cats, such as lions and tigers, along with domestic cats, cannot taste sweet. It turns out that canids, the closest relatives of felines in the order Carnivora, can taste sweets. Even the panda, that charismatic bamboo-loving bear, has intact sweet receptor genes. Pandas can taste sweet and prefer sugary water to tasteless water when offered such libation. And this probably explains why we hang our backpacks with sweet granola bars in them out of the reach of any species of bear when we hike in areas where these sweet-toothed carnivores live.
Figure 4.1. Bar diagram showing the number of intact, pseudo, and truncated taste receptors in a broad range of vertebrates.
An interesting aspect of the evolution of the ability to taste sweet in carnivores arises when thinking about the distribution of sweet taste ability. Is the loss of sweet taste in felines the cause or the effect of felines’ preference for meat? It is difficult to determine how the loss of function (also called pseudogenization) of TAS1R2 is involved, but the phenomenon does remind us that to jump to the conclusion that the loss of the function of the gene is adaptive for meat eating is erroneous. It might be true, but equally likely, the lost region of the TAS1R2 gene occurred through a chance event in the genome of the common ancestor of the felines and was later coopted to reinforce felines’ meat-eating characteristics. This alternate scenario is a good example of what Stephen Jay Gould and Elizabeth Vrba called an exaptation. The trait evolves for some fairly mundane reason and is later exapted or coopted by more visible trait systems (in this case, felines’ preference for meat diets).
If both smell and taste are chemosensory, then how do smelling and tasting differ? Insects are a great example to examine the differences. Besides the fact that in insects the two senses use different receptors, one can always distinguish them by the following logic: Smell is implemented by the detection of gaseous molecules—usually on the antennae. Taste is accomplished by direct bodily contact with the item being tasted. So, what body parts do the tasting in insects? Insects such as flies have mouths—almost. They have mouthparts, and that is a way of saying it’s pretty ugly in there. The mouths of those alien hunters in the Predator movies are modeled after insect mouthparts, and those aren’t pretty at all. The fly mouthparts have taste receptors on them called gustatory receptors, or GRNs (the N stands for the variant)—the fly has about seventy GRNs—but they have no relatedness to the taste receptors found in vertebrates. Clearly, though, flies can detect bitter and sweet as well as water and carbonation. It is not surprising that flies and other insects have gustatory receptors embedded in those ugly mouthparts that come into constant contact with food. And it really shouldn’t surprise us that GRNs are also placed on the wings and legs, and even on the egg-laying apparatus of females, because insects use these appendages to sense the nutrients in the surfaces they touch.
The range of number of GRNs in the genomes of insects is impressive—from eight in lice (Pediculus) to more than two hundred in the flour beetle (Tribolium). The number of odorant receptor genes is also weakly positively correlated with gustatory receptor genes in the genomes of insects, suggesting that the rich get richer with respect to such receptors and with respect to precision of those two senses. Insects that rely on one or a few sources of food would be expected to have fewer taste receptor genes. They need to know whether they are eating what they should. But insects that are a bit more adventurous with their diets (also known as polyphagous insects) might be expected to be more discerning with respect to taste. This is indeed the case with some polyphagous insects, where their genomes contain a couple hundred gustatory receptor genes. But the overall correlation is weak, and a lot more research is needed on how insects taste and how that relates to the evolution of what they eat.
The range of taste receptor genes in vertebrates is equally broad as in insects. In animals, the bitter taste receptors are perhaps the most interesting, and this makes good sense with respect to placing things into the “I eat this” bin strategy. Bitter taste is probably the most important in discerning what an animal should avoid. Sweet-tasting items are a fairly easy choice, because an organism will want to consume the carbohydrates in them. But an organism must be much more discriminating with bitter foods and can’t simply avoid them all. This may be why bitter taste receptors (TAS2s) have a large range of copy number variation in vertebrates. Because herbivores get far less nutritional value for their diet items (plants in general are less rich in calories and other nutritional components than meat), they can ill afford to reject a plant that might have nutritional value. Diyan Li and Jianzhi Zhang argue that herbivores need to be more picky about the plants they encounter. The increase in resolution of bitterness assists them in being picky about rejecting bitter things while still being able to avoid nasty-tasting foods (see box 4.1).
Cats also appear to have quite a few truncated bitter taste genes and a lot of pseudogenes, but they still retain some genes for bitter taste. This exercise of counting genes led to the fascinating discovery that marine mammals (specifically dolphins—see fig. 4.1) have no functional genes for bitter taste. In fact, all cetaceans appear to have experienced a massive loss of bitter and sweet taste genes, according to findings by Ping Feng and colleagues, who examined twelve cetacean species for the bitter and sweet receptor genes. The loss of these receptor genes means that cetaceans have lost four of their five kinds of taste—sweet, umami, sour, and bitter. Cetaceans still retain salt taste, suggesting an evolutionary mechanism for how these marine mammals taste things. However, Feng and colleagues conclude that cetaceans really don’t taste, nor do they need to. The high concentration of salt in the marine environment overwhelms most tastes, and many species swallow prey whole, without tasting their food at all: the other four senses have simply been lost through disuse. The marine environment is hard on TAS2 bitter taste genes. The manatee, the only other mammal in the study that lives in a marine environment, has had 75 percent of its bitter taste genes pseudogenized, rendering them inactive. But even though it has fewer TAS2 genes and fewer taste buds in its mouth, the manatee can still taste. Perhaps its herbivorous diet and the fact that it actually chews its food has prevented the entire gene family from being blanked out.
BOX 4.1 | TASTE RECEPTORS
Taste receptors range broadly across the vertebrates. Bitter-taste receptors, like olfactory receptors, include some pseudogenes and some truncated genes. It appears that for some vertebrates a good proportion of the genes in the genome are either pseudogenes or truncated. The number of receptors ranges from three in some birds and fish to up to seventy (only half of which are functional) in the guinea pig, and about sixty (of which more than fifty are functional) in a genus of frog called Xenopus. By analyzing the diets of various vertebrates and examining their taste receptors, Diyan Li and Jianzhi Zhang have been able to make predictions about how the number of bitter-taste receptor genes influences diet (or vice versa). Although they warn that interactions of taste receptors with the ecology of an organism are complex, they conclude that herbivores have more TAS2 receptors than omnivores and carnivores. It appears that taste has played a central role in helping vertebrates decide which items to place into the “I eat this” bin.
Birds show a paucity of taste receptor genes as all birds have very few TAS2 receptor genes (fig. 4.1). The American crow and finches have the most, at seven intact bitter taste receptor genes, and penguins have none. Again there is an ecological correlation of diet with the number of TAS2 receptor genes. Using the same reasoning of Li and Zhang, Kai Wang and Huabin Zhao point out that herbivorous (and some insectivorous) birds tend to have more TAS2 receptor genes. The penguins are another story, because they have also been examined for the loss of taste receptor genes for the four other tastes. The results of that study demonstrate that penguins have lost their sweet, bitter, and umami receptors, while they have retained sour and salty putative receptors. Other birds appear to have retained nearly all of these taste receptors, except for the loss of sweet receptors in some bird lineages. Penguins, unlike other birds, don’t have taste buds in their tongues, and they swallow their food whole, obviating the need for taste other than salty. It should be noted that these inferences are possible only because of the increased capacity to sequence whole genomes of microbes, animals, and plants. And as more and more organisms have their genomes sequenced, the prospect of uniting feeding ecology with the genetic and molecular aspects of taste will be realized.
Anyone who has been in New York City in mid-July might argue that our sense of smell is pretty good at detecting unpleasant odors. The smell is so bad in July and August in New York City that it prompted a children’s book author to call it Phew York City. One species of insect commonly seen flittering over the garbage that makes the terrible smells in Phew York City, called Drosophila melanogaster, has been the workhorse of biology for more than a century and has contributed greatly to our understanding of smell. Charles Woodworth first suggested in the early twentieth century that this tiny fly, also known as a fruit fly, would be a good experimental animal. Because it reproduces rapidly (every ten days) and is easy to grow in the lab (a little banana and apple sauce mixed with oatmeal and vinegar usually does the trick), it was touted as the ideal lab animal. The geneticist Thomas Hunt Morgan settled on it as his experimental organism of choice in the early 1900s. He was quickly rewarded for choosing this tiny fly by discovering several very visible spontaneous mutants (white eyes and curly wings, among many) that he could use to work out the rules of crossing over of genes on chromosomes. But the tiny fruit fly also has played an amazing role in understanding the mechanics of olfaction. William Morton Barrows recognized that the fly was more than likely using odors to mediate its behavior in 1907 with this statement: “The fact that the fermenting fruit upon which they feed is continually generating alcohols, and other related compounds, led me to suspect that it was these substances that served to attract the flies.”
Barrows devised ingenious experiments to pin down that the flies were responding to chemical odors. It wasn’t until almost fifty years later that more refined methods were used to generate Drosophila olfactory mutants. Several inventive devices were constructed to test mutant flies in the lab for alterations in their olfactory capacity. The most common is the “Y-shaped tube,” in which the “Y” is placed upright. The air is sucked out of the “Y” to remove any lingering odors. The odorant under study is placed at the end of one of the slanted parts of the “Y,” and the other is left unscented as a control. Flies like to climb (the technical term is that they are “negatively geotactic”), so they will climb up to the junction of the two slanted parts of the “Y.” Once there they make a decision based on whether they like or dislike—or literally can’t stand—the odorant. Very clever ways of counting and interpreting the data have been developed, and they lead to identification of flies with mutations that either lose or gain capacity to detect specific odorants (box 4.2).
Using rats, Linda Buck and Richard Axel looked at odorant phenomena in vertebrates. Their landmark study revealed a large and diverse array of odorant receptor genes in mammals. And in the end Buck and Axel realized that the odorant molecules indeed interact with the odorant receptors like locks and keys. If the odorant receptor has the right “lock” for the odorant key to fit in, then the receptor will induce further reactions in the cell that lead to neural transmission to the brain that the odorant is there.
It didn’t take long for Drosophila biologists to jump on the Buck-Axel bandwagon, but they had an advantage—the sequence of the genome of their favorite organism was nearing completion a year or two ahead of the human genome, so they were able to get a complete view of the repertoire of Drosophila odorant receptors and how they worked. They found at least sixty-one odorant receptor genes in the D. melanogaster genome. None show enough similarity to vertebrate odorant genes to warrant calling any the same as vertebrate receptors. In fact, the D. melanogaster odorant receptor genes show extreme sequence divergence with other insect genes, indicating that these receptors change rapidly over evolutionary time. The structure of the proteins coded for by these genes is very interesting, in that all follow a general theme of being embedded in the membrane of the organ that receives odors—in the case of insects, the antenna. The typical odorant receptor protein is threaded through the membrane of the odorant receiving organ’s cells with what are called transmembrane domains. The part of the odorant receptor sticking out of the cell will specifically bind compounds that then trigger intracellular reactions in the receptor cells, which then signal the brain that the specific odorant is there. Drosophila’s sixty-one odorant receptor genes are paltry compared with the more than one thousand found in the nematode, and even more paltry compared with the approximately nineteen hundred in the elephant. But counting odorant receptor genes as a measure of the smelling ability of an organism is complicated.
BOX 4.2 | FINDING ODORANT MUTANT FLIES
A simple metric is used to determine whether a fly has an odorant mutation. The metric is based on the ratio of odorant-sensitive flies to control flies (who aren’t sensitive to the odor). As described for the Y-shaped tube experiments in the text, flies are given a choice to react to specific odors and then counted on the basis of their reaction. If there is a departure from random movement of the flies (50 percent to the control side and 50 percent to the odorant side), the mutant is kept and analyzed further for its odorant capacities. John Carlson and his colleagues at Yale University built a very clever apparatus to detect the odorant sensitivities of flies in 1989 using discarded lab items such as small test tubes and pipette tips. Next, they immobilized the mutants and examined the fly organ that mediates the odorant effects to the fly brain—the antennae. In this way, they isolated six odorant mutants and were able to correlate these with alterations in the function of the antennae. It could be that every time one walks into a department store cosmetics department someone is doing the same experiment with humans, but I doubt it. The point is that the discovery of a slew of mutants in this way led to the genetic description of olfactory mutants in the tiny fruit fly, which in turn led to a more complete understanding of how smell works. Specifically, researchers reasoned that since odorants were molecules, receptors were detecting them. The mutants discovered by Drosophila workers were seminal in coming to this realization.
The first complication is that not all genes in an organism’s genome are expressed. So, although one might be able to detect the presence of sequences that are commonly found in a particular kind of gene, that doesn’t mean the gene is active. When this occurs, the genes are called pseudogenes, as I noted earlier. The lack of expression of pseudogenes usually is caused by the occurrence of a stop codon in the gene, leading to a truncated and nonfunctional gene. The stop codon is a signal to the protein translation machinery of the cell to end translating a gene into protein. The range in number of odorant receptor genes in vertebrates is impressive (see fig. 4.2 and box 4.3).
The second complication is that even the sixty-one odorant receptor proteins of Drosophila can still accomplish a lot of smelling. Many odors can be discerned quite exquisitely by a small number of receptors because of the combinatorial nature of how odorant processing occurs in the brains of animals. A single odorant can bind more than a single receptor, and hence a neuron can have multiple responses when enervated by multiple receptors. In addition, information from neurons with receptors converges to focal processing points that are called glomeruli. In this structure, multiple signals can be combined to be more discerning about odors. The combinatorial nature of odor reception means that as the number of receptors increases, the ability to sense odors also increases, but not linearly. Rather, the increase in potential smells surges exponentially as the number of genes increases.
In addition to odorant receptors, the nasal passages of vertebrates hold another kind of molecule that works in combination with the odorant receptors. These proteins are called TAARs (trace amine-associated receptors), and they function just like their names indicate—they detect trace amounts of small amine molecules. To increase their capacity to sense odorants in vertebrates, these receptors are found in different combinations on the olfactory organ. The response to these combinatorial messages sent to the brain produces behaviors that are essential to the organism’s survival. Linda Buck and colleagues have shown that, while some of the TAAR/odorant receptor responses in the brain are innate and aversive, such responses can be modulated by other odorant receptor signals to the brain. The neat part of this discovery is that the sorting out of the odorants is being done in the brain based on information from a large amount of stimulation of the nose. Our ability to detect trace amounts of a substance is nothing compared to some insects (box 4.4). Perhaps the most famous smelling feat in animals is the capacity to smell and process information from pheromones. Among the insects, Lepidoptera (moths and butterflies) are particularly good at detecting small pheromone molecules at great distances (up to two miles).
Figure 4.2. Bar diagram showing the number of intact, pseudo, and truncated odorant receptor genes in a broad range of vertebrates.
Some odorant researchers think that humans are not very good at smelling (see box 4.3) and are a mediocre species with respect to the number of odorant receptor genes in our genomes (we are in the bottom third of organisms examined at the genome level with respect to total number of odorant receptor genes). But we are actually quite typical in that about half of our odorant receptor genes are pseudogenes. In addition, the convention, since the early twentieth century, has been that humans can detect about ten thousand odorants. In 2014, however, Andreas Keller and colleagues obliterated this notion.
BOX 4.3 | ODORANT RECEPTOR GENES IN ANIMALS
Humans have about eight hundred odorant receptor genes, but only about half of these are active. In almost all vertebrates the number of pseudogenes plus truncated genes exceeds or closely approaches the number of functional genes. Another factor involved in overall gene number is that genes can be gained or lost independently. Chimps and humans had a common ancestor more than six million or so years ago. This common ancestor had a unique combination of odorant genes that it passed on to both the chimp lineage and to the human lineage, but with some changes. For example, in going from the common ancestor of chimps and humans to our lineage, eighteen odorant receptors are gained and eighty-nine are lost. For the chimpanzee lineage, eight genes are gained and ninety-five are lost. The numbers of gains and losses for other taxa to their most recent common ancestors are similar, indicating that as species diverge, their olfactory capacity is fine-tuned by the loss and gain of genes that detect specific odorants. Odorant receptor gain and loss, then, plays a major role in the overall capacity of an organism to smell.
Keller and his colleagues started with 128 known odorants. They then mixed ten, twenty, or thirty of the common ones in jars. For a single experiment (called a discrimination test), they would make three jars—two that were identical and one with something else. The mixtures were paired up in trials such that the odd-man-out jar in some pairs had no odorants in common and others were almost identical. Each volunteer in the experiment was given 260 discrimination tests, and the results were tabulated. All that was needed was to figure out where the ability to detect different odorants in the odorant mixtures drops off. Keller and colleagues assessed this by the amount of overlap the mixtures had. So, they mixed odorants to produce, for instance, 25 percent, 50 percent, 75 percent, or 95 percent overlap in the test jars. If there was 0 percent overlap (the two test jars had no odorants in common), most respondents could easily discriminate between the two jars. If there was 97 percent overlap (one odorant difference in the thirty odorant mixtures) and no subjects could identify the difference, then the drop-off point would be 97%, and so on. The drop-off point turned out to be in the 50 to 60 percent range, meaning that if the odorants overlap by less than, say, 57 percent, most mixtures are distinguishable. Above that percentage of overlap, the odors from the jars are mostly indistinguishable. The techniques involved in interpreting these data involve complex statistical and mathematical processing and are beyond the scope of this book.
BOX 4.4 | THE LIFE OF A CATERPILLAR
The French entomologist Jean-Henri Fabre was the first to describe the pheromone response in detail in the late 1800s. He had found the cocoon of a great peacock moth (Saturnia pyri) and began a watch over it (fig. 4.3). Soon a beautiful female emerged from the cocoon. He placed the female in an enclosure to allow the eclosion process to be completed and went to bed. The next morning, he woke up to find tens of male great peacock moths clinging to the enclosure. He collected the males and left the female in the enclosure overnight again, this time becoming a peeping tom by staying up the night to observe. He continued this ritual for several days and nights, over which he collected 150 or so male great peacock moths. A prolific writer, Fabre was single-handedly responsible for a resurgence of public interest in insects and entomology in the late nineteenth century. The following quotation from his book The Life of a Caterpillar describes what he observed of the great peacock female and shows why he was so good at popularizing entomology: “As I said it was a memorable evening, this Great Peacock evening. Coming in from every direction and apprised I know not how, here are forty lovers eager to pay their respects to the marriageable bride born that morning amid the mysteries of my study. For the moment let us disturb the swarm of wooers no further.” Although he didn’t know it, Fabre was describing the pheromone attraction in insects. He even admitted that he didn’t know how the males were “apprised” that a female was nearby. Decades of research on pheromones have deciphered how the males were “apprised,” and put bluntly, they simply smelled the presence of the female. Oh, and yes, we vertebrates have and can react to pheromones, too.
Figure 4.3. Fabre’s great peacock moth.
But let’s look at what the math actually solves in these experiments, and we should be able to understand the repercussions of the study. For the thirty odorant mixtures there are 1.54 × 1029 possible combinations, and for ten odorants there are 2.27 × 1014. These are very large numbers, and not all combinations will be discerned by the human nose and brain. The trick is to see how many of these can be discriminated by the human odorant receptor apparatus. Doing the math results in the astounding inference that on average humans can discriminate 1.72 trillion different combinations when thirty odorants are combined. That’s 1,720,000,000,000 different combinations! Drosophila has been estimated to discern 65,000 different odors, and other mammals more than likely have the same odorant discrimination capacity of humans. While controversial this number of potential odorant combinations very significantly outrivals the range of sounds, taste, and sight humans can detect. Even if overestimated by several orders of magnitude the estimate implies that what at first appears as our Achilles’ heel sense turns out to be one of our best compared with our other senses.
