Flavor
In food and the food industry there is almost no more important issue than flavor. It is a fascinating topic containing elements of biology, psychology, chemistry and physics. A fascinating interdisciplinary area. If we can agree that flavor is a perception of a quality of food centered in the mouth, where does it come from? Scientists and cooks have struggled with this question for many years, and very interesting research has been done over the last few years that enables us to say that we don’t know the answer. We thought we did, but the answers we thought we had turned out to be wrong, and scientists are in the process of developing new answers (hypotheses) and testing them.
There is some agreement, now, that flavor is not a sense, but a perception that derives from a number of sensory inputs. If you are like me, you have probably heard for a number of years that some large percentage of flavor (80% or maybe, 90%) is due to smell. Thus, when we can’t smell, our food has no flavor. There is certainly some truth to this. There are at least four sensory inputs on which flavor is based. These are vision, smell, taste and touch. We will not give much emphasis to vision; it certainly influences our perception of flavor, but in more subtle and indirect ways than the other three.
Taste
Let’s start off with the obvious one. We often confound taste and flavor in speech. For the scientist, taste is the sense whose receptors are located mostly on the tongue. Flavor is the perception of which taste forms a part. Taste alone is pretty complicated. Most taste receptor cells (TRCs) are located on the tongue and organized into large, visible structures called taste buds. There are 50 to 100 TRCs per bud. The buds are organized into papillae, of which there are three types, based on morphology: fungiform, which resemble mushroom heads; filliform, which are more elongated, and circumvallate, which are quite large, and of which there are 12 at the back of the tongue. The front of the tongue contains a mixture of the other two kinds of papillae. The middle is largely covered with the filliform. Some papillae contain no taste buds at all.
As you probably know, there are five known taste receptor classes: salty, sour, sweet, bitter and umami. The first four are pretty standard. The fifth is the taste of monosodium glutamate (MSG) or other amino acids. Even though the receptors fall into these five classes or stimuli, there are roughly 100 different taste receptors. Each TRC contains thousands of receptors from more than one class. Each bud contains TRCs with receptors from, we suppose, all 5 classes. Even so, there is some degree of specialization on the tongue by area. Thus, sweet receptors are dominant on the front of the tongue, salty around the edges toward the front, sour on the sides, and bitter toward the back. TRCs are constantly being replaced and are some of the most actively dividing cells in our bodies. They are extremely sensitive, sending a signal to our brains upon exposure to just a few molecules of the tastant, at least initially. Of course, like other sensory inputs, they can become desensitized in the presence of constant stimulation.
The five classes of receptors can be grouped further into two different types. The first type are directly coupled to what are called ion channels. The uptake of the tastant, by way of protein channels specific to the given ion, alters the flow of other charged particles across the membrane of the TRC. These are the salty and sour receptors, which are responding, respectively, to Na+ and H3O+ ions. Salty taste is due almost entirely to Na+, although Li+ ions can also stimulate these receptors, and K+ a little bit.
The second type of receptors act through a general class of proteins called GPCRs, or G-protein coupled receptors. This type of receptor is found throughout the body and functions in vision, nerve transmission and hormone detection as well as in taste. In this case, the tastant binds to the receptor on the surface of the cell, but never enters the cell itself. The receptor acts as a switch or a trigger to begin a cascade of events that results in the production of a signal by the TRC connected neuron. In fact, these three receptor classes (sweet, bitter, umami) share some structural components and may exhibit what is called crosstalk. Sweet and umami receptors are composed of two protein molecules, of which one is shared between them.
The bitter receptors utilize similar but distinct proteins from those utilized by sweet and umami receptors. There are on the order of 50 different bitter receptors, responding to a variety of bitter tastants. Obviously, bitter taste is unpleasant, at least by itself, and often results in the rejection of foods. It is thought that the ability to discriminate bitter tastes is of some importance in avoiding toxic compounds in food.
It is of interest to recognize that our genetic complement determines the structures of the taste receptors, and to realize that some taste sensitivities are of genetic origin. The classic example is the ability to taste phenylthiocarbamide (PTC). The variation in the ability to taste this compound has been localized to a particular site on chromosome 7 that encodes a member of the bitter taste receptor family. Variations at this site are thought to account for the majority of the variation in PTC sensitivity, but not all. Other genetic variations in taste receptors have been found to alter the hedonic response to sweetness.
Taking a step back, what happens when tastants enter the mouth? Any individual tastant will interact with receptors on a variety of TRCs in a number of taste buds in different parts of the mouth. Our perception of sweetness, or any taste, is the result of the combination of signals coming from all the TRCs stimulated by the tastant. I should also point out that the strength of the signal on a given nerve is indicated by the rate at which the signal is sent (how many times per second, not the speed), and that this rate depends, at least initially, on the concentration of the tastant.
Artificial sweeteners
The sweet taste is about the only taste which is known to be preferred across cultures. This natural preference has led the food industry, and cooks in general, to sweeten any number of foods in an attempt to make them more attractive. In the developed world, this has led to an increase in the consumption of simple sugars, like fructose and sucrose, with the well-known public health results (especially caries and overweight). These concerns have led, in turn, to the development of artificial sweeteners that do not contribute either to the growth of bacteria in the mouth or to the intake of energy or kcals. There are a wide variety of these artificial sweeteners, and even some sweeteners that are not classified as artificial. In order to compare them, a scale of sweetness has been devised, with sucrose as the standard. From the chemists’ point of view, these substances are interesting more for their lack of common structural features than for anything else (see next page). We learn a little bit about the interaction between the receptors and the tastants from this. We don’t know how many sweet receptors there are, but they do consist of two different protein subunits. And proteins are very large molecules, having the capacity to attach to more than one small molecule, certainly. GPCR receptors, like the sweet receptors, function by binding their target molecules and changing shape as a result. It is well known that, at least for some receptors, we can find other molecules, unrelated to their natural stimulants, that bind somewhere else on the protein and cause a similar change in shape. In fact, this appears to be how valium and its relatives work. It seems likely, then, that at least some of these artificial sweeteners work this way. In fact, there is some evidence that artificial sweeteners stimulate a different signal cascade than do the sugars. How different is not known.
Finally, let us be clear that a
sweet flavor and a sweet taste are not the same. For instance, in the
Real foods are much more complex than individual tastants, selected so that they give apparently single tastes. Many molecules in foods will bind to and stimulate more than one kind of receptor. The flavor of a given substance may include both sweet and bitter, or both sweet and umami tastes, etc. The flavor of a food, of course, arises from a great variety of taste stimuli in a rich combination of intensities and durations.
Artificial sweeteners and relative sweetness
Odor or aroma
Odors can attract or repel our interest in a given food. What is known about odorant molecules and their perception is also a rapidly developing field, full of newly discovered complexities. The detection of odors is the function of the olfactory epithelium, an extension of the brain that resides in the nasal cavity and communicates with the olfactory bulbs near the middle of the face. The olfactory epithelium is another sensory area consisting of rapidly dividing cells that are replaced continuously, at least in part because of their exposure to toxic materials. They are a kind of front line in our investigation of the environment. We have talked about odorant molecules before, and have concluded that they must be in the gas phase and thus must either be a gas or be a relatively low boiling compound or have a relatively high vapor pressure. In general, vapor pressure increases with temperature, so warmer foods will produce more detectable aroma. Since the odorant receptor neurons (ORNs) are in a tissue that is bathed in mucus, a water based solution, the odorants must be, at least to some small degree, soluble in water to be detected. Now this doesn’t have to be much, it turns out, because of the presence of odorant binding proteins (OBPs) which effectively increase the solubility of very nonpolar materials by binding to them themselves, so that water doesn’t have to. In general then, it has been found that odorants are relatively nonpolar, have slight solubility in water, have molecular weights of less than 300 (this would be about 20 carbons) and are soluble in lipid.
An examination of the human genome has detected about 1000 genes for odorant receptors (ORs). About 350 of these are functional; the others are what are referred to as pseudogenes, apparently not active. For comparison, mice have 1300 OR genes, with relatively fewer pseudogenes. This represents some 1-2% of the genome capacity, which is quite a lot for a single function. These ORs are GPCRs like the sweet, bitter and umami taste detectors, and so they bind the odorant molecules on the outside of the cell and do not transport them inside. Rather, the OR initiates a signal cascade that results in the firing of the neuron. These ORs are located in tiny cellular projections that increase the surface area available for odor detection. Each ORN has only one kind of receptor on its surface, unlike the TRCs. Every ORN with a given type of OR extends to a single center in the olfactory bulb, of which there are thousands. There are only 3 OBPs in humans (compared to a few dozen in flies). Thus, the OBPs are not very specific for particular structures. Their function seems to be to transport the odorant through the mucus to the OR. It is an interesting fact that 10% of the human OR genes are active in testes. Some of these are unique to the testes and some are shared with the olfactory epithelium. One final point about the genetic basis of olfaction: the OR genes are related to the bitter taste receptor genes.
Detection of odor is complicated. Unlike early theories suggesting that there were only a few basic odors (like taste), we now know that these 350 odor receptors are largely different, and are organized neurologically by receptor. What seems to happen, then, is that any given odorant molecule will bind to a range of ORs, depending on particular structural features. It may bind more tightly to some and less tightly to others, thus triggering different strength signals in different parts of the olfactory bulb. It seems to be this combination of signals that we perceive as an odor. Since we are combining a few signals out of 350 there are thousands of different signals that can be detected, and thus, thousands of different individual odors.
On the other hand, a given OR is likely to be able to bind more than one odorant molecule, so that there is overlap between the different signals. This just emphasizes the idea that what we perceive is a whole complex of signals making up the property or quality we term odor. Of course, we don’t generally encounter odors of pure substances, but rather odors of mixtures, and our neurons don’t know anything about the chemists distinction between pure substances and mixtures. Thus, odors of mixtures can and often are perceived as unitary.
Tactile sense or chemesthesis
Besides the olfactory and taste neurons in the oral cavity, there are also pain, cold and heat receptors, or neurons. That is, there are also nerves in the mouth that respond to these stimuli. These nerves feed into three of the spinal nerves, rather than directly to the brain, so this is sometimes called trigeminal sensing, because the trigeminal nerve collects many, if not most of these sensations. Flavors like hot peppers, mustard, carbonated beverages and mint are detected to a large extent by way of these tactile sensations. Mint, or menthol, stimulates cold-sensitive neurons, so that we perceive a lowered temperature, even though there hasn’t really been any change in temperature. Likewise, capsaicin and related “hot” flavors stimulate the heat sensitive nerves, and pain reporting nerves. The sensitivity to these components is distributed throughout the oral cavity and in the throat. That is, it is not concentrated on the tongue. Carbonation has been found to stimulate pain and cold receptors. Cinnamic aldehyde (from cinnamon) stimulates heat sensitive nerves, isothocyanates stimulate pain receptors, giving a burning and a stinging sensation, respectively.
At the same time, taste and olfactory receptors may be stimulated. Capsaicin (in 50% of subjects) and menthol stimulate bitter receptors. So we see that “flavor” is made up of a variety of sensory stimuli, integrated by the brain into a unitary feature for a given substance. The flip side is that, even when the entire range of stimuli is not present, our memories of the flavor may fill in the gaps and we may perceive the whole spectrum. At this point, it appears that “sweet” may be the only pure taste sensation. Even so, adding sweet odors, like pineapple, raspberry and caramel, to sugar water makes those solutions smell “sweeter” to subjects. Other flavors not perceived as sweet do not (for instance, peanut butter). On the other side, addition of caramel flavor to citric acid makes it appear less sour. We can be, and probably have been, trained to associate certain flavors, odors and tastes.
Finally
Maybe I should have titled this section miscellaneous, a few odd facts about flavor perception. For example, most sensory inputs are susceptible to desensitization. That is, continued stimulus at a given level causes a reduced signal. For instance, chewing gum typically loses its flavor rather quickly. Direct measurement of what is in the mouth shows that menthol and peppermint, and presumably other flavors persist in the mouth at relatively high levels long after we stop tasting them. Capsaicin shows an interesting pattern. If stimulation by capsaicin is halted for several minutes, desensitization occurs. Continuous restimulation at the same level resensitizes the nerves continually. Thus, if eating very hot foods, a pause of a few minutes is more likely to reduce the intensity of the flavor than continual consumption.
Many flavor molecules to which we are most sensitive are only generated by cooking our food and do not exist as such “in nature.” Why we should have neurons and receptors so sensitive to such flavors is not understood.
The olfactory epithelium is elaborated by about 3 months post-conception. Thus the fetus is stimulated by flavors from an early age. Most of the flavor molecules that reach the fetus are those derived from the mother’s diet and circulating in her blood. There is clear evidence that food and odor preferences are begun in utero.
It has also been discovered recently that olfaction is more sensitive to the rate of chane of the level of an odorant, than it is to the level itself. What this means, is that continual exposure to a given amount of odorant will not stimulate the olfactory nerves as much as a brief, rapid increase in the amount of odorant. This has led to the strategy of including particles in food that, when chewed, release a burst of flavor and/or odor molecules. Of course, this idea of having concentrated particles of flavor in a food has been used by cooks for some time.
Flavor molecules
There are a wide variety of specific flavor/aroma molecules that fall into several common functional groups. I will survey some of those here. The leffingwell site has a much more complete survey.
Esters are a common class of aromas that are found largely in fruits, but also in a variety of other foods. It is interesting, and gives some insight into how flavor perception works, that some very similar esters are perceived differently. For instance, adding two carbons to the ester that we perceive as “pear” gives a “banana” aroma. Adding three more gives a rum aroma. These are not really large changes, and it suggests that our odorant receptors have significant sensitivity for mere physical size. Likewise, rum and pineapple esters differ by only one carbon. These fine distinctions are unlikely to be possible for single ORs, but require some differential binding by a set of ORs stimulating different parts of the olfactory bulb to different extents. It is also of some interest to note that the perception can vary with the amount or concentration of the odorant. Given the discussion above, this is not surprising, as higher concentrations may be presumed to increase the range of ORs stimulated by a given odorant. Cyclic esters are known as lactones, and many of these are common flavor molecules. They evoke notes of sweetness, creaminess, butter, coconut, some smell and taste like peach.
Another common odorant class is the aldehydes. FDA has a huge list of GRAS (generally regarded as safe) aldehyde flavors and aromas. Among these are the aromatic aldehydes distinctive of almond and cinnamon. Other simple aldehydes have nutty or citrus-like aromas. Somewhat similar structures, chemically, are the methyl ketones. Blue cheese aroma is attributable in part to a methyl ketone. Cyclic ketones often have aromas that we associate with burnt or cooked sugar. Some of these are caramel, cotton candy, maple sugar, cooked fruits such as strawberry & pineapple.
Alkenols are alcohols with an unsaturation. They commonly give herbaceous, leafy, green aromas. The pyrazines, which contain nitrogen are known for their nutty odors, some smelling like peanuts and chocolate. Breakdown products of carotene-like compounds have distinctive odors, generally of fruit or vegetable.
A large class of odorants is the terpenes. These are derived from a branched 5 carbon hydrocarbon, called isoprene, and the terpenes are part of a larger class called the isoprenoids. Cholesterol is assembled from isoprene units, as is rubber. So this is a very diverse group of compounds. Among these terpenes are the characteristic odor molecules of coniferous bark and needles, of citrus fruits and of many flowers. Menthol, peppermint, and spearmint (carvone) are terpene odorants. These are rather volatile, and reactive compounds, and are often diminished or lost upon heating and cooking.
The phenolics are a little more water soluble, and so persist a little longer. These include clove, cinnamon, anethole, vanilla, anise, thyme, oregano and tarragon. Obviously many spices are among these types of compounds. Phenols are aromatic alcohols.
The thiocyanates are pungent aromas that cause irritation mostly in the olfactory region. Horseradish, mustard and garlic are in this class. These are very volatile aromas, which is why they have such a strong nasal response.
Finally the alkyl amides include capsaicin and other hot or “spicy” flavors. These are much larger molecules, not so volatile, so there is less nasal component, and they are more persistent.