I am not a betting man, but I am reasonably confident-and would be prepared to wager a decent bottle-that this is the first article you will have read on saliva. And if, by chance, it isn’t, I am almost certain it is the first article on saliva that you’ll have encountered in the pages of a wine magazine. The subject of the inside of our mouths (perhaps more formally referred to as the oral environment) is one that has been neglected in discussions about the taste of wine.1 But the mouth is where we interact with wine most closely, at least until it is swallowed.
Most wine drinkers with more than a passing interest in what is in their glass usually smell before they taste, but then any preliminary judgment from the aroma is supplanted by a more concrete judgment made once the wine is in their mouth. The way we experience the wine once we have begun to interact with it in this way is multimodal in that it involves the senses of smell, taste, and touch. For this reason, saliva is a vital component that mediates how we experience wine. It is, therefore, surprising that the role of saliva in the taste of wine has been so neglected. This article is my attempt to redress such neglect.
Saliva is a watery secretion from three different glands: the parotid (in the cheek, under the ear), the submandibular (under the jaw), and the sublingual (under the tongue). Without any stimulation, there is a low but steady flow of saliva. Try not to swallow for a few minutes, and you will feel your mouth begin to fill up with liquid. This unstimulated flow is important to keep the mouth moist and for protecting the teeth and surfaces inside the oral cavity.
Most saliva, however, is secreted because of some sort of stimulation. The stimulation can come from smell or taste, or from a mechanical stimulus, such as chewing. As an example, when the dentist fiddles inside your mouth, the stimulated salivary flow is sufficient that a suction tube is needed to keep your mouth from filling up. Salivary flow can even be stimulated by an association-remember Pavlov’s dogs, who began slobbering at the ringing of a bell, even in the absence of a food stimulus. The average person secretes 17-50 fl oz (0.5-1.5 liters) of saliva a day, most of which is swallowed. It is thought that around 80-90 percent of this is produced by stimulated flow. Flow decreases at nighttime (during sleep) to virtually zero and can be affected by medication or dehydration.
So, what is special about the composition of saliva? First, it contains proteins called mucins, which have a high carbohydrate content and are very slippery, capable of absorbing large amounts of water into their structure. In the mouth, these mucins form a covering over the oral tissues that helps lubricate the mouth (useful for talking and for masticating food) and also protect against irritation and unwanted microbes. The ability of mucins to absorb quite a bit of water into their structures helps keep this protective lubricating layer reasonably thick and, in concert with the liquid nature of saliva, clears unwanted microorganisms and food debris from the mouth.
Saliva also contains high concentrations of calcium and phosphate ions, which help protect dental enamel and allow the remineralization of teeth. It also protects the teeth by washing away or diluting any potentially harmful chemicals and buffering against pH changes, such as those caused by acids in the diet. Wine, of course, is high in acidity. The teeth are further protected by a covering layer of salivary proteins called a pellicle. A range of different proteins is involved in the formation of this coating, which controls the colonization of the tooth surface by bacteria (encouraging good ones and discouraging bad ones) and helps protect against acid.
All this protection is necessary because the inside of our mouths is a highly vulnerable place. It’s warm and wet: ideal conditions for harmful microbes to grow in. Teeth are sensitive to acidity, which can destroy them. Saliva is, therefore, playing an essential protective role-one that we take for granted until salivary flow is diminished or stops altogether, as occurs in some diseases, with some medications, or after radiotherapy treatments for cancer. It takes someone to have the condition xerostomia (a dry mouth due to reduced salivary secretion) to realize just how important saliva is. If saliva flow is severely diminished, people are forced to use artificial saliva, which must be sprayed into the mouth at regular intervals. These artificial salivas, however, are relatively unsophisticated and don’t carry out all the functions of normal human saliva.
PRPs, HRPs, and tannins
The saliva components most of interest to us in a wine context are two groups of proteins known as the prolinerich proteins (PRPs) and histatins, or histadine-rich proteins (HRPs). The PRPs make up about 70 percent of salivary proteins overall and have a high proportion of the amino acids proline, glycine, and glutamine. There are three subtypes: acidic, basic, and glycosylated PRPs. Altogether, the three groups have around 20 different members. Acidic PRPs are unique to saliva. They bind calcium strongly and are important in forming the tooth pellicle layer and making sure there is enough calcium present for tooth remineralization. This is because they bind calcium when it is present at high levels and are then able to release it gradually during conditions when there is less of it around. Glycosylated PRPs are lubricants and also interact with microbes. In contrast, basic PRPs have only one role: to bind to tannins, forming precipitates. HRPs are small proteins, rich in histidine and found only in saliva. Twelve of them are known in humans, and they account for just 2.5 percent of saliva protein. They have antibacterial and antifungal properties, but crucially they are also very good at binding tannins. The ability of PRPs and HRPs to bind with tannins is where the real interest, from a wine perspective, comes in.
Tannins are formed by plants as defense molecules, defending against microbial attack and also acting as antifeedants. Plants are extremely vulnerable to being eaten, because they are literally rooted in place, and so they have gone to great lengths to make themselves unpalatable, acting as chemical factories to produce a wide range of toxic defensive secondary metabolites, as well as developing physical defenses such as thorns or stings. It is only a relatively restricted set of plant species that is suitable for human consumption: In many cases, the only bit of the plant we are able to eat is the bit that it wants us (or other animals) to eat-fruits produced to aid seed dispersal. In the case of grapes, they are camouflaged green and made unpalatable by high tannins, high acidity, and no sugar until the seeds are mature enough to be dispersed, at which point the ripening process has made them attractive to eat and easy to find.
Chemically, it is hard to give a strict definition of what constitutes a tannin. The term is used as a catch-all for a group of rather different polymers (chemicals made from repeated subunits, which, in the case of tannins, consists of phenolic groups) that historically were used to convert animal skins into leather in the process of tanning. There are two types found in wine: condensed and hydrolyzable. Condensed tannins are extracted from grapes during red-wine making and are then modified further during the winemaking process. Hydrolyzable tannins come from oak used in winemaking and also the stems and seeds of grapes. Tannins in wine are found in a variety of states: They are intrinsically “sticky” molecules and join up with several other wine components such as anthocyanins (the colored pigments found in grape skin) to form pigmented polymers, or can combine with other chemicals. The pigmented polymers in particular are important in forming stable red-wine color. Tannin size is referred to by the term DP (degree of polymerization), and this indicates the number of tannin subunits that are joined together. Seed and wood tannins are typically smaller than skin tannins, and it is thought that these smaller tannins possess a bitter taste rather than express astringency. (More on this later.) A chief goal of the process of élevage is to manage tannins in such a way, by means of limited oxygen exposure, to produce a wine that is harmonious. It can be speculated that the ability of tannins to modify themselves in the presence of oxygen is because this is one of their roles in plant defense; when you bite into an apple and leave it, after a while the “wound” area becomes brown, with the tannins in the tissue interacting with oxygen to make the damaged area less inviting for colonizing microbes.
One of the key roles of salivary PRPs and HRPs is to protect us from the harmful effects of tannins, by binding to them and precipitating them before they reach the gut. This makes the plants more edible than they otherwise would be, neutralizing one of their defenses. If the salivary PRPs didn’t cause this precipitation, the tannins would interact with digestive enzymes (which are also proteins) in our gut and render them ineffective. This would reduce the palatability of plant components by making them much less digestible. The aversive taste of unripe fruits is, in part, due to high tannin concentrations, with the plant using this as a way of keeping the fruit from being consumed before the seeds are ready for dispersal, along with color changes and high acid/low sugar. We find the bitter taste and astringent sensation of tannin aversive, and as with such unpleasant oral sensations, the aversion can protect us from harmful consumption. Thus, the PRPs and HRPs are potentially filling two roles: allowing us to detect tannins in food and to reject the food if the concentrations might be dangerous, and also helping to neutralize any tannins present in food to be ingested.
The affinity of tannins for proteins is the basis for the winemaking use of proteins such as the albumin in egg whites as fining agents in red wines. They help precipitate excess tannin out of the wine, to make it taste more appealing and less aggressively tannic.
We sense tannins in wine largely as an astringent sensation, but there can also be a contribution from taste in some circumstances. Astringency is not principally a taste, in the sense that it is not one of the primary taste modalities of sweet, sour, bitter, salty, umami. Instead, it is chiefly detected by the sense of touch in our mouths. (There is still some discussion in the scientific literature about whether or not astringency is tasted.) The feeling of the wine in the mouth, referred to commonly as mouthfeel, is detected by mechanoreceptors in the mouth. These are neurons specializing in sensing touch, including Ruffini endings, Merkel cells, Meissner cells, and free nerve endings, and they are found throughout the oral cavity. Anyone who has had dental work that has modified the inside of their mouth in some way will know that we are very sensitive to changes in the mouth environment. We have a very well-developed sense of touch, and the tongue especially is good at exploring the inside of the mouth in great detail. Interestingly, one of the things the sense of touch does is to localize tastes and smells to where the food or drink is within the mouth. This makes us ascribe the properties of smell and taste to the food or drink, even though they might be sensed elsewhere in the mouth. It is useful if we need to remove the food from our mouths when it tastes bad. It is also important to clear the mouth of any remaining food after eating. Dietary tannins entering the mouth are bound by proteins present in saliva and form precipitates. These proteins include the PRPs and HRPs, whose role is to carry out this binding and protect us from the potentially harmful effects of tannins in inhibiting digestive enzymes. But it also includes one of the other important protein types in saliva, the mucins. As mentioned earlier, these are involved in forming a lubricated, protective layer over the internal surface of the mouth. Tannins remove this lubrication, causing a sense of dryness, puckering, and loss of lubrication in the mouth. This is what we describe as “astringent.”
Related to astringency is the taste of bitterness. The majority of tannins are chiefly sensed as astringent, but they can also be tasted as “bitter” when they are small enough to interact with bitter receptors on the tongue. Tannins seem to reach their most bitter taste at a DP of 4 (four subunits joined together, which is small) and then decrease in bitterness and increase in astringency, with this astringency peaking at a DP of 7, before becoming steadily less astringent as they become larger. The astringent nature of tannins can be moderated by the presence of polysaccharides (sugars) or other wine components. It is also modified by the chemical adornments that tannins can grab, and there are many of these. In wine, tannins are continually changing their length (DP) and adding things to their structure. So, structurally, wine tannins can be incredibly complicated, and researchers are still trying to correlate mouthfeel properties with structure.
Interestingly, tannins are more astringent with lower pH (that is, wines with higher acidity taste more astringent, even with the same tannin content) and less astringent with increasing alcohol. However, the bitterness of tannins rises with the alcohol level and is unaffected by pH changes.
It is worth noting that acid stimulates salivary flow. If salivary flow is increased, there is more protein present in the saliva to form precipitates with tannins. The implication from this is that two red wines with identical tannin composition but different pH levels will have a different mouthfeel. This could be part of the explanation for the observation that lowering pH increases the sense of astringency. But it could be that there is some sort of additive effect between sensing acidity and astringency.
As we drink wine, the wine itself will increase the flow of saliva, which, in itself, will change the perception of the wine. The binding of tannins may well reduce their ability to reach the bitter receptors, and thus their bitterness may decrease and their astringency increase at the same time.
Implications for wine tasting
This is all well and good, but what are the implications for wine tasting? Next time you taste and spit red wine, take a look at the spittoon. It’s actually quite an unpleasant sight, with large strings of congealed, red/purple/black strings of saliva. This is the result of interactions between wine and saliva-chiefly the binding of salivary proteins by tannins to form precipitates. The mucins in saliva do their bit by helping create these viscoelastic strings of colored spit.
In the normal situation of drinking wine, it seems likely that the production of saliva is able to keep pace with the rate at which the wine is consumed. With red wine, the challenge to the palate is the repeated exposure to tannins. With white wine, the tannic content is much lower, and the challenge will be the acidity, which is usually much higher than in red wines (that is, the pH of whites is lower). With Champagne and sparkling wine, acidity is higher still. None of these should provide a major challenge to the perception of wine-unless exposure is repeated rapidly over a short time.
Just such an exposure takes place in many situations where wine is tasted professionally. Whether the situation is a trade tasting or competition judging or a critical assessment of a region’s wines, it is common to find professionals tasting upward of 100 samples a day. And this often includes repeated assessments of samples, pushing the figure up considerably. Though I am not aware of any scientific studies that have investigated this sort of scenario, it is possible to predict what might be happening in terms of salivary flow.
In the first instance, with red wines, tannins will be interacting with salivary proteins, precipitating out and causing a sensation of astringency and mouth drying. The initial layer of mucins lubricating the mouth will be stripped. Then, with repeated exposure, the deeper layer of mucins will be stripped; this is something that would not normally happen in typical wine drinking.
Usually, repeated exposure to the same smell or taste will result in a degree of adaptation. With astringency, however, repeated exposure results in the sensation of astringency increasing. Wine ingestion stimulates saliva production, but this is not sufficient to deal with repeated samples of red wine in close succession, in that it fails to replenish the lubricating layer of salivary mucins on the mouth surfaces. The result is that the sense of astringency increases with each fresh sample, to the point where it can become uncomfortable. The last thing I tend to feel like after a day of thorough tasting is another glass of wine-my mouth is feeling utterly fatigued.
With acidity, frequent exposure to a high acid stimulus is likely to overwhelm the buffering and dilution capacity of the saliva. This can leave the mouth feeling sensitive to subsequent samples and might lead to acidity being misjudged. However, my experience as a taster is that it is less fatiguing to taste many white wines in a single session than it is to taste many red wines.
This is not meant to read as a counsel of despair for professional wine tasting. But it is an observation that should encourage us to approach tasting with a degree of humility. Tasting lots of wines in succession carries with it risks-not just of palate fatigue, but also of the effects of presentation order. The perception of any one wine can be influenced by the nature of the preceding wine. For this reason, it is good practice to have different tasters on a panel taste in different order, even if it is as simple as some tasting in reverse order. In sensory analysis at an academic level, presentation order is randomized.
Given that saliva is unable to cope well with the sort of frequency of wine tasting typically carried out by professionals, what can we do to help? At the most basic level, we need to hydrate properly. Dehydration reduces saliva flow. If we spit, we eject not only the wine but also the saliva our mouths have produced. If we are producing around a liter of saliva a day, and then spitting and not swallowing it, this fluid shortfall needs to be made up. Often tasters clear their palates with water, plus solids such as crackers, bread, or black olives. This may help by absorbing some of the tannins that have built up and not been cleared by the overworked salivary flow, but it is far from being a sophisticated solution. The effectiveness of palate cleansers in restoring the palate to baseline conditions has been examined. One study compared astringency build-up using a number of different cleansers: deionized water, a 1g/l pectin solution, a 1g/l CBMC (carboxymethylcellulose) solution, and unsalted crackers. Subjects tried the same wine six times, with a cleansing process after the third wine. The unsalted cracker was most effective at reducing astringency build-up, while water alone was the least effective, but astringency built up whatever cleanser was used. Another study showed that pectin rinse was the most effective cleanser, followed by unsalted crackers. CBMC can be effective in some instances.
In sensory analysis work, the 100+ samples that are regularly taken on daily by professional tasters would not be tolerated. The “noise” produced from this sort of palate fatigue would likely render any statistical analysis non-significant. But as a wine judge, I have frequently taken part in competition scenarios that would have sensory analysts tearing their hair out. My experience has been that experienced, competent judges are still able to make good decisions under these circumstances, such as occur in large wine competitions. But they would find making the sorts of fine discriminations that are important in judging among top-quality wines more difficult with fatigued palates. Certainly, results would be cleaner and better if fewer wines were tasted and gaps between flights were sufficient to allow palate recovery. That said, in my experience the chief problem in wine competitions is the “noise” produced by less competent judges. Good judges, able to perform well in blind-tasting settings, are rare, but they do exist; and I’d rather have the verdict of a competent but palate-fatigued judge than that of an incompetent judge with a fresh palate.
For fine wines, small differences in quality are significant. And for top red wines, mouthfeel is one of the key components of the wine. Elegance and harmony-much prized in older wines, in particular-depend in large part on mouthfeel. For assessing these sorts of wines, palate fatigue is near fatal. For this reason, the number of samples that can be assessed reliably is much reduced.
There is one further point that must be made in relation to saliva and wine tasting, and this concerns inter- and intra-individual differences in saliva production. People differ in the composition and production of saliva, and each person’s salivary flow rate will change with a number of factors, including hydration state, time of day, emotional state, and the influence of medication. In addition, 10-15 percent of the population breathe largely through their mouths, and this will result in significant evaporation of saliva (estimated to be a loss of 12 fl oz [350ml] per day). Both inter- and intra-individual differences in saliva are likely to affect the mouthfeel of red wines. This adds another level of inter-individual variation to wine tasting, in addition to factors such as taste-bud density, olfactory receptor repertoire, and knowledge and experience. The extra level of intra-individual variation is also something we must be aware of as tasters. The fact that our experience of wine can differ from day to day, or even from hour to hour, should help us remain humble in the face of wine.
In conclusion, we encounter wine most profoundly when it is in our mouths. And the internal environment of the mouth clearly has a significant impact on wine perception. Saliva is vital in mediating our experience of wine, so any attempt to understand wine appreciation needs to consider saliva and the mouth environment as intrinsic components in the perception of flavor and texture.
1. Here, I am using the term “taste” in its broader sense: a combination of the senses of smell, taste, sight, and touch, as perceived consciously only after some hidden higher-order processing in the brain, influenced unconsciously by other factors such as learning, expectation, knowledge, and context.