Thirst and Water Balance
Summary and Keywords
Maintaining water balance is critical for survival, but our bodies are constantly losing more water than we produce. Consuming water, therefore, is needed to restore what is lost by sweating, bleeding, vomiting, urinating, even breathing. Because the fluid in the body is divided into intracellular and extracellular compartments, and because depletion can happen in one compartment without affecting the other, separate detection mechanisms for losses in each are required. Moreover, the relatively high concentration of sodium in the extracellular space means that sodium loss accompanies extracellular dehydration. Accordingly, the behavioral response to loss of fluid from the extracellular space needs to include sodium intake. Activity of osmoreceptors (in the case of intracellular loss), or baroreceptors and the renin-angiotensin system (in the case of extracellular loss), underlies the responses to perturbations of fluid balance, and promotes the appropriate behaviors needed to restore balance to the system. The peptide angiotensin II (AngII) is a key component of these responses. Studies of AngII in drinking have been critical in our understanding of how a peripherally derived peptide can act in the brain without transport across the blood–brain barrier, and AngII-induced drinking has served as an important model for the study of intracellular signaling pathways that affect behavior. Although much has been discovered about these systems and how they respond to fluid deficits, the precise means by which the systems generate a behavioral response and the mechanism that mediates satiety remains poorly understood. Nevertheless, ongoing experiments on these open questions have already started to provide a new perspective on the negative reinforcement that is provided by drinking under conditions of thirst.
“Thirst belongs to humanity, everywhere, in all ages … One could never remember himself in eternity by the mere fact of having loved or hated any more than by that of having thirsted”—Oliver Wendell Holmes, from the Breakfast-Table Series—the Autocrat of the Breakfast Table, London, George Routledge and Sons, 1887.
Thirst is incredibly powerful. Most of us know what it feels like to be thirsty. We know the dry mouth, the longing for drink that comes with that sense. Water is a critical element of life, and makes up the majority of the human body. The water that we comprise is under routine threat. We face the loss of water because of injury that makes us bleed, because of pathologies that cause vomiting or diarrhea. We expel water in urine and sweat. We even lose water when we breathe. We experience insensible loss of almost a liter of water from our skin and lungs each day. Just being alive costs water, and although the chemical reactions of metabolism produce some water from within, our best way to maintain the water we lose is by eating and drinking. Likewise, sodium is critical for our survival. As important as water is, the balance of water inside and outside of our cells requires the osmotic pressure from sodium. Sodium is critical for neural and muscle function, and a true sodium deficiency is deadly. Given the importance of water and sodium, it should not be surprising that we have important regulatory mechanisms that drive thirst and sodium appetite responses, helping to maintain the needed levels of water and sodium in our bodies.
What Are Thirst and Salt Appetite?
Scientists are often forced to make a choice. One of those choices is whether to use human participants or laboratory animals as experimental subjects. Both come with advantages and disadvantages. The issues facing scientists who study thirst are illustrative in this respect. Most humans experience thirst, and those who do, recognize the experience and know to call it thirst. When humans experience thirst, and have the opportunity to drink water, that drinking generally happens. In this sense, the act of drinking water can be seen as evidence of thirst. But is drinking perfectly correlated with need? Of course not. Humans drink because of an anticipated need that has not yet occurred, and humans drink in the absence of thirst because of advice from health professionals. Humans drink to help swallow something dry, even without any sensation of thirst. If given the opportunity to win a contest by drinking a large volume of water, many would drink a large volume of water, but very likely would not feel the sensation of thirst while doing so. As such, the act of drinking does not perfectly represent the subjective sensation of thirst. Thus, when scientists study thirst in laboratory animals, it is often difficult, if not impossible, to know if the subject is experiencing thirst. If a pharmacological manipulation is made to a rat, for instance, and that rat responds by drinking copious amounts of water, can the scientist know if the rat is experiencing thirst? Unfortunately, the answer is generally no. Scientists can measure how much the animal drinks. Scientists can measure the pattern and timing of the drinking. Scientists can measure how hard the animal would work for the fluid. Yet none of these measurements provides any certainty that the rat is experiencing thirst, at least not in the way humans experience thirst. While this is a clear disadvantage of using non-human subjects, the questions scientists can ask and answer in laboratory animals provide important insight into the physiology that underlies what may or may not be genuine thirst. It could be argued fairly that the burden of proof is on those who posit that the animal is not feeling thirst in the way humans do. If a stimulus causes humans to report feeling thirsty, and the stimulus causes both humans and laboratory animals to drink, claiming that the animal is not experiencing thirst requires the belief that thirst is unique to humans. Nevertheless, it seems reasonable to approach the issue with some caution, and to recognize that when scientists discuss thirst in a non-human subject, or in any experiment where subjective measures of thirst are not recorded or reported, it includes ascribing a sensation that may or may not exist in the laboratory animal.
The same is true for sodium hunger. Many humans experience cravings for sodium, and describe their subjective experience as such (e.g., craving salt). In general, humans like the taste of sodium, and use table salt (NaCl) as a means to enhance and add flavor to food. This liking of salt certainly helps humans consume the sodium needed for survival, but most Americans eat far more sodium than necessary (Jackson, Coleman King, Zhao, & Cogswell, 2016; U.S. Department of Health and Human Services and U.S. Department of Agriculture, 2015). A strong liking of sodium may provide a meaningful glimpse into human evolutionary history. Indeed, salts (including sodium salts) were an important and widely used method to preserve food. There is evidence of salt-curing of meat as early as 1000 bce (Kaloyereas, 1950). Thus, it is not difficult to imagine that ancestors who did not like the taste of salt would have been less likely to consume preserved foods, and, therefore, more likely to get sick. Those who liked salty food would have had more to eat, and it is not hard to see how being able to eat more food would provide a reproductive advantage, thereby selecting for a taste for salt. However, as is the case for thirst, it is difficult, if not impossible, to conclusively attribute sodium intake as being driven by a sodium appetite, at least as it might be described in humans (e.g., salt cravings). There is, nevertheless, a bit more evidence for similar experiences between humans and laboratory animals when it comes to salt appetite than there might be when it comes to thirst.
When evaluating if laboratory rats feel a hunger for sodium under conditions of need, similar to that felt by humans during states of salt craving, taste reactivity testing can be particularly informative. The taste reactivity test, pioneered by Grill and Norgren (1978a, 1978b), allows an investigator to probe the affective reaction to a taste experience (Figure 2). Addition of a bitter taste to the mouth of a rat elicits a stereotyped pattern of mouth and limb movements that includes gapes, chin rubs, head shakes, and forelimb flailing, collectively called aversive responses. In contrast, addition of a sweet taste to the mouth leads to a pattern of movements including rhythmic mouth movements, tongue protrusions, and paw licking, which look markedly similar to the movements made when licking a spout. These movements are classified as ingestive responses. In the case of sodium, the ratio of aversive to ingestive responses in a laboratory rat when strong sodium stimuli are delivered to the mouth is highly dependent on the internal state of the animal. Aversive responses decrease, and ingestive responses increase, after a rat has been deprived of sodium (Figure 2) (Berridge, Flynn, Schulkin, & Grill, 1984). This, coupled with demonstrations of rats consuming relatively large amounts of sodium after sodium depletion, shows strong evidence of a hunger for salt that appears similar to that experienced by humans in cases of sodium need.
Using the Double-Depletion Model as a Framework
Since 1973, the study of thirst and sodium appetite has been largely guided by the double-depletion model of thirst and sodium appetite. This model, described originally by Fitzsimons (1973) and Epstein (1973), is based on several key elements. Firstly, the fluid in the body is distributed in two main compartments: intracellular and extracellular (Figure 3). Although the extracellular compartment can be further subdivided into fluid in circulation, fluid between cells (interstitial), and other minor fluid components (e.g., lymph, bladder), the distribution of fluid inside and outside of cells of the brain appears to be the critical driver of behavior. Because these compartments are separate, and because there are differences in the nature of fluid from each, the types of fluid loss need separate detection mechanisms as well as separate means to induce behavior. Some convergence on a final common pathway is expected, but also must be modulated to reflect differences in need under different circumstances. For instance, as will be described in more detail in the Intracelluar Dehydration and Extracellular Dehydration sections, loss of fluid from the extracellular space requires intake of both water and sodium, whereas loss of fluid from the intracellular space requires intake of water, without intake of sodium that could further exacerbate the loss of fluid from the intracellular space. Accordingly, both of these may feed through a final common pathway governing water intake, because both require intake of water, but they must diverge with respect to a pathway controlling sodium intake, because intake is important in one condition, while problematic in the other. As such, the two parts of the double-depletion model can certainly be considered overlapping and interactive, but also separable.
Our cells are water-filled building blocks of our tissues. The amount of water in them is a function of several factors, including the osmotic pressure that draws water into or out of the cells, depending on the relative concentration of solutes in the intracellular and extracellular spaces. When the concentration of sodium, for example, increases in the extracellular space, it produces osmotic pressure that draws water from the inside of the cell to the outside of the cell. This movement of water triggers a type of thirst called “osmotic thirst” based on the stimulus that triggers it. Osmotic thirst is the result of changes in the extracellular space that either reduce the amount of water, without reducing the solute, or add additional solutes, without changing the volume of water. The former is something fought on a regular basis because every single breath causes exhalation of water, without exhaling the solutes dissolved in the water when it is inside our body. Although this may seem surprising, it is easily demonstrated by breathing on glass. This produces a fog on the glass surface that dries without leaving behind a film of salts, such as what may be seen after letting sweat or tears dry on glass. Although a very small amount of water is lost with each breath, after breathing many times each day, it can accumulate to a noticeable amount of water loss. In addition, solutes are added to the extracellular space by eating or drinking things other than pure water. The effects of this tend to be more rapid when the meal is salty, but consumption of anything that is not isotonic has the potential to affect the osmotic balance between the intracellular and extracellular space.
Embedded within various brain structures are specialized cells called osmoreceptors that detect changes in osmolality. The movement of water between the intracellular and extracellular space impacts all cells, but osmoreceptors are specialized in that they change their activity based on these changes, and signal the change to other cells in the nervous system. These cells, first hypothesized and named by E. B. Verney (1946, 1947), have since been found in several brain structures including the organum vasculosum of the lamina terminalis (OVLT), which has received a great amount of attention as an osmosensing structure (see Bourque (2008) for a review).
The idea of an osmoreceptor originated in studies of diuresis. These classic studies by Verney measured diuresis in response to treatments that increased plasma osmolality. Increased osmolality was associated with reduced urine, leading to several hypotheses about the underlying physiology. Without the assays or methods needed to test these hypotheses, Verney hypothesized the existence of cells that could detect the change in osmolality (osmoreceptors), as well as a hormone that carried the signal from the central osmoreceptors to the kidney to affect urine output. This anti-diuretic hormone would later come to be known as vasopressin, although the name anti-diuretic hormone persists in some disciplines. In this respect, osmoreceptors engage magnocellular cells of the paraventricular nucleus (PVN) and the supraoptic nucleus (SON) of the hypothalamus. These cells release vasopressin into circulation, thereby causing the kidney to retain water, preventing loss of water that would exacerbate the hypertonic extracellular environment.
Osmoreceptors play a critical role in the stimulation of drinking under conditions of osmotic thirst. Early studies focused on regions of the hypothalamus, including the lateral hypothalamus and lateral preoptic area (LPO). In a set of experiments by Blass and Epstein (1971), destruction of the LPO reduced drinking after thirst was stimulated by hypertonic saline. Furthermore, the report shows that injection of hypertonic saline directly into the LPO stimulated drinking, and that injection of water into the LPO reduced drinking after thirst was stimulated by systemic hypertonic saline. Although this report strongly suggested the existence of osmoreceptive elements in the LPO, it is notable that the manipulations of the LPO were not completely effective (e.g., some drinking occurred after destruction). Accordingly, it became clear that other areas likely participated in the response, and decades of research between now and then point to the OVLT as a more critical site in this respect.
How, precisely, osmoreceptors signal changes in extracellular concentration has been an exciting field of research. Mammalian cells, which lack a cell wall, will swell or shrink depending on the relative concentrations of the intracellular and extracellular spaces. Osmoreceptors are typical, therefore, in the shrinking that occurs when they are in a hypertonic environment, but are specialized in that they are equipped to alter their activity in response to that shrinking. Indeed, osmoreceptor cells in the OVLT are depolarized and show increased frequency of action potentials when in a hypertonic environment. This specialization is currently thought to be the function of specialized transient receptor potential vanilloid 1 (Trpv1) channels that function as mechanoreceptors (Ciura & Bourque, 2006; Zaelzer et al., 2015), whereas another type of Trpv channel, Trpv4, inhibits cellular activity in hypotonic environments (Ciura et al., 2018). Accordingly, cells in the OVLT are able to alter their responses to either increased or decreased concentration of the extracellular solution.
Although much has been learned about the stimulation of drinking by osmoreceptors, the termination of drinking after osmotic thirst conditions is far less understood. Indeed, the perturbation in fluid osmolality that leads to drinking affects osmoreceptors rapidly, and the drinking-inducing change in osmolality persists beyond the termination of drinking. In other words, drinking stops before the osmotic pressure at the osmoreceptors is restored to normal. In spite of the persistent osmotic pressure, drinking itself rapidly suppresses vasopressin secretion and terminates further drinking. This has been shown in a variety of species including rats (Stricker & Hoffmann, 2005), sheep (Blair-West, Gibson, Woods, & Brook, 1985), dogs (Thrasher, Nistal-Herrera, Keil, & Ramsay, 1981), non-human primates (Arnauld & du Pont, 1982), and humans (Geelen et al., 1984; Rolls et al., 1980). These rapid effects of drinking on behavior and physiology are similarly shown in studies of sweating in humans (Takamata, Mack, Gillen, Jozsi, & Nadel, 1995) and panting in sheep (McKinley, Weissenborn, & Mathai, 2009). These findings suggest that the system can predict the restoration of normal osmolality, and once enough fluid has been consumed, can ignore the osmotic stimulus that remains present until well after the drinking has stopped. Studies using a combination of transgenic approaches and the development of calcium-sensing proteins have been used to monitor activity-induced florescence in vasopressin cells of the hypothalamus (Mandelblat-Cerf et al., 2017). These studies clearly showed that drinking induces rapid decreases in activity of vasopressin cells. Moreover, when a water-deprived mouse was given an empty water bowl, the initial response was similar to that seen when water was given, but within 30 seconds, activity of vasopressin cells returned to the activity level associated with dehydration. This shows a remarkable ability of the system to adapt to the expectation of water, and to adjust when that expectation is not met. How this happens remains to be determined, but studies have shown that the median preoptic nucleus (MnPO) plays a key role. Indeed, chemogenetic suppression of cells in the MnPO suppresses drinking behavior in water-restricted mice (Augustine et al., 2018). Moreover, a subset of these cells appear to track liquid intake. Monitoring of calcium increases in the brains of behaving mice showed that these cells responded to licking at liquid, but not solid, substances. Specifically, these cells were activated by licking at spouts delivering water, isotonic saline, or a non-aqueous silicone gel, but not by licking at an empty spout. Moreover, in food-restricted mice, licking sucrose solutions similarly activated these cells, but licking a solid substance (peanut butter) did not stimulate activity in these cells, suggesting that these MnPO cells track consumption of fluid, rather than licking itself, or reward. How this is transmitted from orogastric signals remains to be determined.
At the core of the double-depletion model is that intracellular and extracellular compartments can be individually dehydrated, triggering disparate detection systems, and separate (but overlapping) behavioral responses. Indeed, when osmotic thirst occurs, the key effect is not a change in overall body volume, but rather a movement of the volume from the inside to the outside of cells. But loss of volume is also possible, and produces a powerful thirst that is referred to as “hypovolemic” thirst. Bleeding, vomiting, diarrhea, or excessive perspiration all reduce the volume of the extracellular space (hypovolemia), but because both the water and solutes dissolved in the water are lost, there is no change to the overall concentration of the fluid, and thus no impact on osmoreceptors. Accordingly, a different system is needed to detect and respond to this kind of loss. This is accomplished largely through two main limbs: baroreceptors that monitor blood vessel stretch, and the renin-angiotensin system. Moreover, because the loss that occurs in these cases is a loss of both water and the salts dissolved in the water, restoring water alone does not sufficiently address the deficit. Thus, increases in sodium intake accompany the water intake that results from hypovolemia. Both baroreceptors and the renin-angiotensin system play roles in stimulating these motivated behaviors.
Baroreceptors are critical for the detection of changes in blood volume and pressure. Most of the early evidence for baroreceptor involvement in thirst, however, was largely circumstantial. For instance, treatments that induced volume depletion caused drinking that was correlated with the magnitude of the volume effect, and hypotension was shown to inhibit drinking (Fitzsimons, 1961; Stricker, 1966). Subsequent studies provided more direct evidence by showing that drinking was stimulated by reducing venous return to the heart (Fitzsimons & Moore-Gillon, 1980; Thrasher, Keil, & Ramsay, 1982), and that this could be prevented by surgical denervation of cardiopulmonary and sinoaortic baroreceptors (Quillen, Reid, & Keil, 1988).
The renin-angiotensin system is responsible for a constellation of effects that occur in concert with baroreceptors. When a loss of volume occurs, the enzyme renin is released from the juxtaglomerular cells of the kidney (Catanzaro, Mullins, & Morris, 1983). In this context, renin has the critical job of being the rate-limiting step in the biosynthetic pathway that generates angiotensin II (AngII) (Figure 4). The main role of renin is the cleavage of the large glycoprotein, angiotensinogen, which is produced by the liver and is abundant in circulation. The cleavage of angiotensinogen produces circulating angiotensin I, which is largely inert. Angiotensinogen serves as a substrate for the enzyme angiotensin converting enzyme (ACE), which removes two amino acids from the carboxyl end of angiotensin I to produce AngII. The effects of AngII are numerous, but most serve important roles in compensating for the lost fluid, and promoting behaviors that replace water and sodium, thereby defending fluid homeostasis. These actions include vasoconstriction, decreased vagal tone, increases in circulating vasopressin, and promotion of thirst and salt appetite. Whereas the study of the cardiovascular effects of AngII have made it a common target for treatments of hypertension (e.g., ACE inhibitors and angiotensin receptor blockers), the study of AngII effects on thirst and salt appetite also has served as a critical model of motivated behavior.
The effects of AngII on water intake are largely due to primary actions on circumventricular structures in the brain. These structures, particularly the subfornical organ (SFO) and OVLT, respond to circulating AngII and relay the signal to other areas of the brain that are part of a drinking-related distributed circuit. Interestingly, many of the nodes of this circuit respond to AngII, but in the absence of laboratory experiments, likely are stimulated by AngII acting as a transmitter, rather than the endocrine effects that occur at the SFO and OVLT, because circulating AngII does not appear to cross the blood–brain barrier. Thus, the circuit includes nodes that respond to endocrine actions of AngII, that then use AngII in a transmitter-like manner to engage downstream nodes. As such, experimental injections into the ventricle, such as those commonly used in the laboratory, likely activate a combination of nodes, including those behind the blood–brain barrier that do not normally see circulating AngII.
The roles of the SFO and OVLT in drinking behavior were established by a series of experiments using straightforward, traditional approaches, but the path to understanding the system was somewhat less straightforward. Of course, the initial step in this discovery was the finding that AngII worked in the brain to cause drinking behavior. Although this seems like an obvious conclusion today, at the time, many were skeptical about the ability of a circulating peptide to act in the brain. Indeed, early studies found that renal extracts elicited a drinking response (Asscher & Anson, 1963; Linazasoro, Jimenez Diaz, & Castro Mendoza, 1954; Nairn, Masson, & Corcoran, 1956). More than a decade after the first of these studies, a hint of the dipsogenic effect of centrally administered AngII appeared in the literature, but was presented as a minor component of a larger study focusing on hypothalamic injections of norepinephrine (Booth, 1968). The next year, another report of drinking after central AngII appeared, but this was a smaller part of a series of studies that were more focused on pituitary responses to central injections of AngII (Daniels, Ogden, & Vernikos-Danellis, 1969; Daniels-Severs, Ogden, & Vernikos-Danellis, 1971) (note that the first author, Anne Daniels, is not related to the present author). At the same time, members of Epstein’s group were performing similar experiments, but with a primary focus on the drinking effects of brain injections of AngII (Epstein, Fitzsimons, & Rolls, 1970; Epstein, Fitzsimons, & Simons, 1969), and soon thereafter after Simpson and Routtenberg reported their studies showing that direct injections of AngII into the SFO caused drinking behavior (Simpson & Routtenberg, 1973). Thus, a story emerged that painted the SFO as a primary site of action for circulating AngII. This not only answered the question of where AngII could act to cause drinking, but the site being the SFO, which lacks a blood–brain barrier, also provided a needed solution to the problem of how a circulating peptide could access the brain. The story did not stop there, however, and subsequent attempts to discover other sites of action in the brain have become a useful teaching tool for scientists attempting to make injections of virtually any substance into specific brain regions. Specifically, in a set of studies that made small injections of AngII into a number of brain regions, testing for dipsogenic responses, Johnson and Epstein clearly showed that virtually any injection target was effective, as long as the cannula traveled through a forebrain cerebral ventricle to the targeted structure (Johnson & Epstein, 1975). This was elegantly confirmed using radiolabeled AngII, and clearly showed that when a cannula passed through the lateral ventricle en route to a targeted underlying structure, the injected AngII filled the cannula tract and the brain ventricles. This provided strong evidence that the ventricle was a potent avenue for the responses to AngII, and served as a lesson to other scientists to be careful of ventricle efflux when making brain injections through a cannula that passes through a ventricle. Accordingly, drawing a conclusion about a site being responsive to AngII requires that the same dose of AngII is ineffective when injected directly into the ventricle. This was true in the earlier studies of the SFO, and was also shown to be true in subsequent studies testing the effect of AngII into the OVLT and surrounding anteroventral third ventricle (AV3V) region (see Vento and Daniels (2014) for example).
Lesion studies were also informative in the path to understanding central effects of AngII on drinking behavior. Simpson and Routtenberg found that lesions made in the SFO attenuated drinking after intracranial injection of AngII (Simpson & Routtenberg, 1973), but the results were somewhat variable, suggesting cooperation of another structure. Later studies suggested that the effect of the SFO lesion was dependent upon an indirect effect of the lesion, specifically debris or edema blocking the ventricular foramen, thereby preventing AngII from reaching critical periventricular structures (Buggy & Fisher, 1976; Buggy, Fisher, Hoffman, Johnson, & Phillips, 1975). The drinking deficit was more pronounced when lesions were made in the AV3V region (Buggy & Johnson, 1978), and experiments designed to obstruct and prevent access to the AV3V after injection into the ventricle provided strong evidence that injected AngII must reach this region of the periventricular space to stimulate water and sodium intakes (Buggy et al., 1975; Buggy & Fisher, 1976; Buggy & Johnson, 1978; Hoffman & Phillips, 1976).
Although the importance of the SFO and OVLT in the drinking response have been established for decades, a 2015 report describes experiments that helped tease apart a heterogeneous population of cells in the SFO that control drinking behavior. Specifically, activation of light-sensitive ion channels expressed in SFO cells with Ca2+/calmodulin-dependent kinase II (CamKII) resulted in drinking, whereas activation of light-sensitive channels expressed in cells with the vesicular GABA transporter (Vgat) suppressed drinking in mice that had been water deprived (Oka, Ye, & Zuker, 2015). Moreover, the CamKII cells (which also express ETV-1) were found to also express AngII type I (AT1) receptors, highlighting them as the specific targets of AngII within the SFO. As such, a more complex circuit within the SFO is emerging that contains not only stimulatory elements, but also those that can suppress drinking. How these interact, and how they are engaged, perhaps selectively, remains an open avenue of study.
Although AngII plays a clear and critical role in the stimulation of water intake, it is also important for the sodium appetite that is needed after extracellular loss. Sodium appetite is critical in this event because the extracellular loss includes water and the sodium that is lost with the water. Whereas AngII injection into the brain causes mild sodium intake (Buggy & Fisher, 1974), treatment with AngII alone is not sufficient to cause a robust sodium appetite such as that illustrated in Figure 2. Moreover, a sodium appetite can be elicited by treatment with adrenal steroids such as aldosterone, but only at doses higher than would be found in the physiological range. In this respect, it is important to note that under non-experimental conditions, a circumstance that elevates AngII would also elevate adrenal steroid secretions. Indeed, a hypovolemic insult or sodium deprivation, both of which cause robust sodium appetite, cause the concurrent elevation of both AngII and aldosterone (Epstein, 1982, 1984; Fluharty & Epstein, 1983; Fluharty & Sakai, 1995; Fregly & Rowland, 1985; Sakai, 1986; Sakai, Nicoladidis, & Epstein, 1986). As such, the elevation of one without the other is largely limited to experimental conditions, suggesting some cooperation of the two. In support of the hypothesis that AngII and adrenal steroids work together to generate sodium appetite, exogenous mineralocorticoids DOCA or aldosterone, and ICV AngII, all at doses that were insufficient to elicit salt intake when each was given alone, produced a robust salt appetite when given together (Fluharty & Epstein, 1983). Conversely, blockade of either AngII or aldosterone alone partially reduces deprivation-induced sodium appetite, but simultaneous inhibition of both abolishes the appetite (Sakai, 1986; Sakai, Nicolaidis, & Epstein, 1986).
The study of AngII in thirst and water balance has also been a useful tool for helping to understand the roles played by individual intracellular signaling pathways in the control of behavior. There is strong evidence that the effects of AngII on water and sodium intake occur through the AT1 receptor subtype (Beresford & Fitzsimons, 1992; Li et al., 2003; Sakai et al., 1994; Sakai, Ma, He, & Fluharty, 1995; Weisinger, Blair-West, Burns, Denton, & Tarjan, 1997). This receptor is a prototypical G protein-coupled receptor (GPCR). AT1 receptors exert their effects by acting through G protein pathways, and the AT1 receptor also activates members of the mitogen activated protein (MAP) kinase family (de Gasparo, Catt, Inagami, Wright, & Unger, 2000; Sadoshima, Qiu, Morgan, & Izumo, 1995). The development of modified AngII peptides that act as biased agonists have allowed investigators to engage MAP kinase, in the absence of signaling that is downstream of G protein activation. Experiments using cell culture models indicate that one such biased agonist has antagonist-like effects on water intake stimulated by AngII; it has no effect on its own, but blocks the dipsogenic effect of AngII (Daniels, Yee, Faulconbridge, & Fluharty, 2005). On the other hand, when measuring salt intake, the same drug shows agonist-like effects, stimulating intake of saline. This finding gave rise to the “divergent signaling hypothesis of angiotensin-induced water and saline intake” (Daniels, 2010; Daniels et al., 2005; Daniels, Mietlicki, Nowak, & Fluharty, 2009; Daniels, Yee, & Fluharty, 2007). According to this hypothesis, AngII stimulates water intake through G protein-dependent signaling pathways, whereas saline intake is more dependent on MAP kinase-mediated signaling. In support of this hypothesis, selective inhibitors parts of the G protein pathway selectively reduce water intake, without affecting saline intake, whereas inhibition of the relevant MAP kinases affects only saline intake, without affecting water intake (Daniels et al., 2009). Additional support for the hypothesis is found in studies showing that other AngII-related peptides that selectively affect MAP kinase activation, without affecting AngII activation of G protein pathways, have an effect on AngII-induced saline intake, without affecting water intake (Liu et al., 2014).
Thirst as an Unpleasant State
When thinking about motivated behaviors, it may be helpful to consider the nature of the drive. In this sense, comparing and contrasting thirst and hunger, and drinking and eating, may provide important lessons to help us understand these complex behaviors. In this respect, hunger and thirst appear similar in that they are both unpleasant, and we are therefore driven to eat or drink to remove that unpleasantness. Thus, the eating or drinking serves as negative reinforcement by removing the unpleasant state of hunger or thirst. But food and water appear different in this respect, in that some eating occurs independent of hunger, or even in the face of fullness. How many of us, for instance, have somehow found room for dessert in spite of having just eaten well beyond satiety, exchanging the discomfort of hunger for a discomfort from overeating? In contrast, how often do we drink so much water that we reach a comparable point of discomfort? Clearly less often for most people. These comparisons may reveal important differences between the drive states that accompany hunger and those that accompany thirst, and may offer excellent opportunities to study the learning that occurs with each. Studies in recent years have found strong evidence that negative reinforcement plays a key role in drinking (Betley et al., 2015; Leib et al., 2017). Moreover, studies using AngII have found that repeated exposure to AngII increases the drinking response, and that this requires both AngII and drinking behavior, suggesting that the negative reinforcement provided by drinking can be strengthened by repeated trials (Postolache, Santollo, & Daniels, 2017). Additional research is needed, but these findings suggest several potentially fruitful lines of research to help understand the neural control of these separable motivated states and the learning that may impact them.
Although many scientists have no problem justifying the study of thirst and salt appetite as an interest in understanding the phenomena themselves, studies of thirst and salt appetite are further justified because the phenomena serve as important models for other critical features of life. The principles discovered in these studies may apply to and could inform studies on other motivated behaviors such as hunger, and use of alcohol and other drugs of abuse. Indeed, it is very likely that many of the central targets of drugs of abuse overlap with those associated with thirst and salt appetite. Indeed, some of the changes in gene expression observed after sodium depletion include changes to gene products otherwise associated with addiction-like conditions (Liedtke et al., 2011). This may not be surprising given that drugs of abuse are known to take advantage of neural substrates involved in the response to so-called natural rewards such as food and water. In this respect, understanding fluid intake is important not only to understand and prevent conditions associated with fluid imbalance (e.g., hypertension), but also to understand the motivated states that are associated with drug use and abuse. In this sense, exploring systems known to be involved with drug taking has revealed interesting connections with the internal state of the animal. Specifically, studies of dopamine signaling revealed that dopamine responses to water or sodium are highly dependent on the internal state of the animal. In these studies, Fortin and Roitman (2018) found that increases in phasic dopamine were evoked only by the needed stimulus: saline in a sodium-deprived animal and water in a water-deprived animal. This serves as a powerful example of the ability of internal state to affect responses to external stimuli.
Maintenance of body fluid balance is critical. Although much is known about the stimuli that elicit thirst and salt appetite, and some of the relevant brain areas, much remains to be discovered. We are just beginning, for instance, to understand how water is detected, and how this feeds back to the brain to modify behavioral responses. A better understanding of this will require considerable research investment. Moreover, it remains unclear how the engaged circuits cause changes in behavior. Part of the problem, in this respect, is that we lack a clear end point to use in ascribing a circuit. Whereas we know that the SFO and OVLT are connected with, and engage, the MnPO, and we know many of the outputs of the MnPO, we are left with an unclear understanding of what cells in the MnPO do with that information. We are left uncertain where to even begin. Do we look for circuits that engage the limbs and direct them to a source of water? Do we look for circuits that engage the tongue and create a licking pattern? Do we look for circuits that engage the visual system to increase salience of visual cues for water? These basics of motivational states remain poorly understood, not just for thirst and salt appetite, but also for food intake and drug use. Fortunately, the work over the past half-century on thirst provides an excellent model system that can be used to explore these complex unanswered questions.
Arnauld, E., & du Pont, J. (1982). Vasopressin release and firing of supraoptic neurosecretory neurones during drinking in the dehydrated monkey. Pflugers Archiv: European Journal of Physiology, 394, 195–201.Find this resource:
Asscher, A. W., & Anson, S. G. (1963). A vascular permeability factor of renal origin. Nature, 198, 1097–1098.Find this resource:
Augustine, V., Gokce, S. K., Lee, S., Wang, B., Davidson, T. J., Reimann, F., . . . & Oka, Y. (2018). Hierarchical neural architecture underlying thirst regulation. Nature, 555, 204–209.Find this resource:
Beresford, M. J., & Fitzsimons, J. T. (1992). Intracerebroventricular angiotensin II-induced thirst and sodium appetite in rat are blocked by the AT1 receptor antagonist, Losartan (DuP 753), but not by the AT2 antagonist, CGP 42112B. Experimental Physiology, 77, 761–764.Find this resource:
Berridge, K. C. (2000). Measuring hedonic impact in animals and infants: Microstructure of affective taste reactivity patterns. Neuroscience and Biobehavioral Reviews, 24, 173–198.Find this resource:
Berridge, K. C., Flynn, F. W., Schulkin, J., & Grill, H. J. (1984). Sodium depletion enhances salt palatability in rats. Behavioral Neuroscience, 98, 652–660.Find this resource:
Betley, J. N., Xu, S., Cao, Z. F., Gong, R., Magnus, C. J., Yu, Y., & Sternson, S. M. (2015). Neurons for hunger and thirst transmit a negative-valence teaching signal. Nature, 521, 180–185.Find this resource:
Blair-West, J. R., Gibson, A. P., Woods, R. L., & Brook, A. H. (1985). Acute reduction of plasma vasopressin levels by rehydration in sheep. American Journal of Physiology, 248, R68–71.Find this resource:
Blass, E. M., & Epstein, A. N. (1971). A lateral preoptic osmosensitive zone for thirst in the rat. Journal of Comparative and Physiological Psychology, 76, 378–394.Find this resource:
Booth, D. A. (1968). Mechanism of action of norepinephrine in eliciting an eating response on injection into the rat hypothalamus. Journal of Pharmacology and Experimental Therapeutics, 160, 336–348.Find this resource:
Bourque, C. W. (2008). Central mechanisms of osmosensation and systemic osmoregulation. Nature Reviews Neuroscience, 9, 519–531.Find this resource:
Buggy, J., & Fisher, A. E. (1974). Evidence for a dual central role for angiotensin in water and sodium intake. Nature, 250, 733–735.Find this resource:
Buggy, J., & Fisher, A. E. (1976). Anteroventral third ventricle site of action for angiotensin induced thirst. Pharmacology Biochemistry and Behavior, 4, 651–660.Find this resource:
Buggy, J., Fisher, A. E., Hoffman, W. E., Johnson, A. L., & Phillips, M. I. (1975). Ventricular obstruction: Effect on drinking induced by intracranial injection of angiotensin. Science, 190, 72–74.Find this resource:
Buggy, J., & Johnson, A. K. (1978). Angiotensin-induced thirst: Effects of third ventricle obstruction and periventricular ablation. Brain Research, 149, 117–128.Find this resource:
Catanzaro, D. F., Mullins, J. J., & Morris, B. J. (1983). The biosynthetic pathway of renin in mouse submandibular gland. Journal of Biological Chemistry, 258, 7364–7368.Find this resource:
Ciura, S., & Bourque, C. W. (2006). Transient receptor potential vanilloid 1 is required for intrinsic osmoreception in organum vasculosum lamina terminalis neurons and for normal thirst responses to systemic hyperosmolality. Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 26, 9069–9075.Find this resource:
Ciura, S., Prager-Khoutorsky, M., Thirouin, Z. S., Wyrosdic, J. C., Olson, J. E., Liedtke, W., & Bourque, C. W. (2018). Trpv4 mediates hypotonic inhibition of central osmosensory neurons via taurine gliotransmission. Cell Reports, 23, 2245–2253.Find this resource:
Daniels, A., Ogden, E., & Vernikos-Danellis, J. (1969). Effects of centrally administered angiotensin II in the unanesthetized rat. Physiologist, 12, 205.Find this resource:
Daniels, D. (2010). Alan N. Epstein award: Intracellular signaling and ingestive behaviors. Physiology & Behavior, 100, 496–502.Find this resource:
Daniels, D. (2017). Neuropeptides and regulation of water intake. Reference Module in Neuroscience and Biobehavioral Psychology: Elsevier.Find this resource:
Daniels, D., & Fluharty, S. J. (2009). Neuroendocrinology of body fluid homeostasis. In D. W. Pfaff, A. P. Arnold, A. M. Etgen, S. E. Fahrbach, & R. T. Rubin (Eds.), Hormones brain and behavior (2nd ed., Vol. 1, pp. 259–288). San Diego, CA: Academic Press.Find this resource:
Daniels, D., Mietlicki, E. G., Nowak, E. L., & Fluharty, S. J. (2009). Angiotensin II stimulates water and NaCl intake through separate cell signalling pathways in rats. Experimental Physiology, 94, 130–137.Find this resource:
Daniels, D., Yee, D. K., Faulconbridge, L. F., & Fluharty, S. J. (2005). Divergent behavioral roles of angiotensin receptor intracellular signaling cascades. Endocrinology, 146, 5552–5560.Find this resource:
Daniels, D., Yee, D. K., & Fluharty, S. J. (2007). Angiotensin II receptor signalling. Experimental Physiology, 92, 523–527.Find this resource:
Daniels-Severs, A., Ogden, E., & Vernikos-Danellis, J. (1971). Centrally mediated effects of angiotensin II in the unanesthetized rat. Physiology & Behavior, 7, 785–787.Find this resource:
Epstein, A. N. (1973). Epilogue: Restrospect and prognosis. In A. N. Epstein, H. R Kissileff, & E. Stellar (Eds.), The neuropsychology of thirst: New findings and advances in concepts (pp. 315–332). Washington, DC: V.H. Winston & Sons.Find this resource:
Epstein, A. N. (1982). Mineralocorticoids and cerebral angiotensin may act together to produce sodium appetite. Peptides, 3, 493–494.Find this resource:
Epstein, A. N. (1984). The dependence of the salt appetite of the rat on the hormonal consequences of sodium deficiency. Journal de physiologie, 79, 496–498.Find this resource:
Epstein, A. N., Fitzsimons, J. T., & Rolls, B. J. (1970). Drinking induced by injection of angiotensin into the brain of the rat. Journal of Physiology, 210, 457–474.Find this resource:
Epstein, A. N., Fitzsimons, J. T., & Simons, B. J. (1969). Drinking caused by the intracranial injection of angiotensin into the rat. Journal of Physiology, 200, 98P–100P.Find this resource:
Fitzsimons, J. T. (1961). Drinking by rats depleted of body fluid without increase in osmotic pressure. Journal of Physiology, 159, 297–309.Find this resource:
Fitzsimons, J. T. (1973). Some historical perspectives in the physiology of thirst. In A. N. Epstein, H. R Kissileff, & E. Stellar E (Eds.), The neuropsychology of thirst: New findings and advances in concepts (pp. 3–33). Wasthington, DC: V.H. Winston & Sons.Find this resource:
Fitzsimons, J. T., & Moore-Gillon, M. J. (1980). Drinking and antidiuresis in response to reductions in venous return in the dog: Neural and endocrine mechanisms. Journal of Physiology, 308, 403–416.Find this resource:
Fluharty, S. J., & Epstein, A. N. (1983). Sodium appetite elicited by intracerebroventricular infusion of angiotensin II in the rat: II. Synergistic interaction with systemic mineralocorticoids. Behavioral Neuroscience, 97, 746–758.Find this resource:
Fluharty, S. J., & Sakai, R. R. (1995). Behavioral and cellular analysis of adrenal steroid and angiotensin interactions mediating salt appetite. In S. J. Fluharty, A. R. Morrison, J. Sprague, & E. Stellar (Eds.), Progress in psychobiology and physiological psychology (Vol. 16, pp. 177–212). San Diego, CA: Academic Press.Find this resource:
Fortin, S. M., & Roitman, M. F. (2018). Challenges to body fluid homeostasis differentially recruit phasic dopamine signaling in a taste-selective manner. Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 38, 6841–6853.Find this resource:
Fregly, M. J., & Rowland, N. E. (1985). Role of renin-angiotensin-aldosterone system in NaCl appetite of rats. American Journal of Physiology, 248, R1–11.Find this resource:
de Gasparo, M., Catt, K. J., Inagami, T., Wright, J. W., & Unger, T. (2000). International union of pharmacology. XXIII. The angiotensin II receptors. Pharmacology Reviews, 52, 415–472.Find this resource:
Geelen, G., Keil, L. C., Kravik, S. E., Wade, C. E., Thrasher, T. N., Barnes, P. R., . . . & Greenleaf, J. E. (1984). Inhibition of plasma vasopressin after drinking in dehydrated humans. American Journal of Physiology, 247, R968–R971.Find this resource:
Grill, H. J., & Norgren, R. (1978a). The taste reactivity test. I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Research, 143, 263–279.Find this resource:
Grill, H. J., & Norgren, R. (1978b). The taste reactivity test. II. Mimetic responses to gustatory stimuli in chronic thalamic and chronic decerebrate rats. Brain Research, 143, 281–297.Find this resource:
Hoffman, W. E., & Phillips, M. I. (1976). Regional study of cerebral ventricle sensitive sites to angiotensin II. Brain Research, 110, 313–330.Find this resource:
Jackson, S. L., Coleman King, S. M., Zhao, L., & Cogswell, M. E. (2016). Prevalence of excess sodium intake in the United States—NHANES, 2009–2012. Morbidity and Mortality Weekly Report, 64, 1393–1397.Find this resource:
Johnson, A. K., & Epstein, A. N. (1975). The cerebral ventricles as the avenue for the dipsogenic action of intracranial angiotensin. Brain Research, 86, 399–418.Find this resource:
Kaloyereas, S. A. (1950). On the history of food preservation. Scientific Monthly, 71, 422–424.Find this resource:
Leib, D. E., Zimmerman, C. A., Poormoghaddam, A., Huey, E. L., Ahn, J. S., Lin, Y. C., . . . & Knight, Z. A. (2017). The forebrain thirst circuit drives drinking through negative reinforcement. Neuron, 96, 1272–1281 e1274.Find this resource:
Li, Z., Iwai, M., Wu, L., Shiuchi, T., Jinno, T., Cui, T. X., & Horiuchi, M. (2003). Role of AT2 receptor in the brain in regulation of blood pressure and water intake. American Journal of Physiology–Heart and Circulatory Physiology, 284, H116–121.Find this resource:
Liedtke, W. B., McKinley, M. J., Walker, L. L., Zhang, H., Pfenning, A. R., Drago, J., . . . & Denton, D. A. (2011). Relation of addiction genes to hypothalamic gene changes subserving genesis and gratification of a classic instinct, sodium appetite. Proceedings of the National Academy of Sciences of the United States of America, 108, 12509–12514.Find this resource:
Linazasoro, J. M., Jimenez Diaz, C., & Castro Mendoza, H. (1954). The kidney and thirst regulation. Bull Inst Med Res Kuala Lumpur, 7, 53–61.Find this resource:
Liu, J., Yosten, G. L., Ji, H., Zhang, D., Zheng, W., Speth, R. C., . . . & Sandberg, K. (2014). Selective inhibition of angiotensin receptor signaling through Erk1/2 pathway by a novel peptide. American Journal of Physiology–Regulatory, Integrative and Comparative Physiology, 306, R619–626.Find this resource:
McKinley, M. J., Weissenborn, F., & Mathai, M. L. (2009). Drinking-induced thermoregulatory panting in rehydrated sheep: Influences of oropharyngeal/esophageal signals, core temperature, and thirst satiety. American Journal of Physiology–Regulatory, Integrative and Comparative Physiology, 296, R1881–1888.Find this resource:
Mandelblat-Cerf, Y., Kim, A., Burgess, C. R., Subramanian, S., Tannous, B. A., Lowell, B. B., & Andermann, M. L. (2017). Bidirectional anticipation of future osmotic challenges by vasopressin neurons. Neuron, 93, 57–65.Find this resource:
Nairn, R. C., Masson, G. M., & Corcoran, A. C. (1956). The production of serous effusions in nephrectomised animals by the administration of renal extracts and renin. Journal of Pathology and Bacteriology, 71, 155–163.Find this resource:
Oka, Y., Ye, M., & Zuker, C. S. (2015). Thirst driving and suppressing signals encoded by distinct neural populations in the brain. Nature, 520, 349–352.Find this resource:
Postolache, M., Santollo, J., & Daniels, D. (2017). Associative learning contributes to the increased water intake observed after daily injections of angiotensin II. Physiology & Behavior, 179, 340–345.Find this resource:
Quillen, F. W., Jr., Reid, I. A., & Keil, L. C. (1988). Cardiac and arterial baroreceptor influences on plasma vasopressin and drinking. In A. W. Cowley Jr., J. F. Liard, & D. A. Ausiello (Eds.), Vasopressin: Cellular and integrative functions (pp. 405–411). New York, NY: Raven Press.Find this resource:
Rolls, B. J., Wood, R. J., Rolls, E. T., Lind, H., Lind, W., & Ledingham, J. G. (1980). Thirst following water deprivation in humans. American Journal of Physiology, 239, R476–482.Find this resource:
Sadoshima, J., Qiu, Z., Morgan, J. P., & Izumo, S. (1995). Angiotensin II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90-kD S6 kinase in cardiac myocytes. The critical role of Ca(2+)-dependent signaling. Circulation Research, 76, 1–15.Find this resource:
Sakai, R. R. (1986). The hormones of renal sodium conservation act synergistically to arouse sodium appetite in the rat. In G. de Caro, A. N. Epstein, & M. Massi (Eds.), The physiology of thirst and sodium appetite (pp. 425–430). New York, NY: Plenum.Find this resource:
Sakai, R. R., He, P. F., Yang, X. D., Ma, L. Y., Guo, Y. F., Reilly, J. J., . . . & Fluharty, S. J. (1994). Intracerebroventricular administration of AT1 receptor antisense oligonucleotides inhibits the behavioral actions of angiotensin II. Journal of Neurochemistry, 62, 2053–2056.Find this resource:
Sakai, R. R., Ma, L. Y., He, P. F., & Fluharty, S. J. (1995). Intracerebroventricular administration of angiotensin type 1 (AT1) receptor antisense oligonucleotides attenuate thirst in the rat. Regulatory Peptides, 59, 183–192.Find this resource:
Sakai, R. R., Nicolaidis, S., & Epstein, A. N. (1986). Salt appetite is suppressed by interference with angiotensin II and aldosterone. American Journal of Physiology, 251, R762–768.Find this resource:
Simpson, J. B., & Routtenberg, A. (1973). Subfornical organ: Site of drinking elicitation by angiotensin II. Science, 181, 1172–1175.Find this resource:
Stricker, E. M. (1966). Extracellular fluid volume and thirst. American Journal of Physiology, 211, 232–238.Find this resource:
Stricker, E. M., & Hoffmann, M. L. (2005). Inhibition of vasopressin secretion when dehydrated rats drink water. American Journal of Physiology–Regulatory, Integrative and Comparative Physiology, 289, R1238–1243.Find this resource:
Takamata, A., Mack, G. W., Gillen, C. M., Jozsi, A. C., & Nadel, E. R. (1995). Osmoregulatory modulation of thermal sweating in humans: Reflex effects of drinking. American Journal of Physiology, 268, R414–422.Find this resource:
Thrasher, T. N., Keil, L. C., & Ramsay, D. J. (1982). Hemodynamic, hormonal, and drinking responses to reduced venous return in the dog. American Journal of Physiology, 243, R354–362.Find this resource:
Thrasher, T. N., Nistal-Herrera, J. F., Keil, L. C., & Ramsay, D. J. (1981). Satiety and inhibition of vasopressin secretion after drinking in dehydrated dogs. American Journal of Physiology, 240, E394–E401.Find this resource:
U.S. Department of Health and Human Services and U.S. Department of Agriculture. (2015). 2015–2020 Dietary Guidelines for Americans. 8th Edition.Find this resource:
Vento, P. J., & Daniels, D. (2014). The anteroventral third ventricle region is critical for the behavioral desensitization caused by repeated injections of angiotensin II. Behavioural Brain Research, 258, 27–33.Find this resource:
Verney, E. B. (1946). Absorption and excretion of water: The antidiuretic hormone. Lancet, 2, 739–744.Find this resource:
Verney, E. B. (1947). The antidiuretic hormone and the factors which determine its release. Croonian lecture. Proceedings of the Royal Society of London B: Biological Sciences, 135, 25–106.Find this resource:
Weisinger, R. S., Blair-West, J. R., Burns, P., Denton, D. A., & Tarjan, E. (1997). Role of brain angiotensin in thirst and sodium appetite of rats. Peptides, 18, 977–984.Find this resource:
Zaelzer, C., Hua, P., Prager-Khoutorsky, M., Ciura, S., Voisin, D. L., Liedtke, W., & Bourque, C. W. (2015). DeltaN-TRPV1: A molecular co-detector of body temperature and osmotic stress. Cell Reports, 13, 23–30.Find this resource: