After surgical removal of all salivary secretions (“desalivation”), rats increase their consumption of water while eating dry laboratory chow. In the present experiments, desalivated rats drank even more water while they ate “powdered” high-salt food (i.e., <15-mg food particles). The Na+ concentration of systemic plasma in these animals was not elevated during or immediately after the meal, which suggests that cerebral osmoreceptors were not involved in mediating the increased water intake. A presystemic osmoregulatory signal likely stimulated thirst because the Na+ and water contents of the gastric chyme computed to a solution ∼150 mM NaCl. In contrast, desalivated rats drank much smaller volumes of water while eating “pulverized” high-salt food (i.e., 60–140-mg food particles), and the fluid mixture in the gastric chyme computed to ∼280 mM NaCl solution. These and other findings suggest that the NaCl ingested in the powdered high-salt diet was dissolved in the gastric fluid and that duodenal osmoreceptors (or Na+-receptors) detected when the concentration of fluid leaving the stomach was elevated after each feeding bout, and promptly stimulated thirst, whereupon rats drank water until the gastric fluid was diluted back to isotonicity. However, when rats ate the pulverized high-salt diet, much of the NaCl ingested may have been embedded in the gastric chyme and therefore was not accessible to visceral osmoreceptors once it emptied from the stomach. Consistent with that hypothesis, fluid intakes were increased considerably when desalivated rats drank 0.10 M NaCl instead of water while eating either powdered or pulverized high-salt food.
- gastric chyme
osmoregulation in mammals is accomplished largely through the central control of thirst and neurohypophyseal secretion of vasopressin (VP), the antidiuretic hormone. Cerebral osmoreceptors that detect changes in the osmolality of systemic blood and mediate these responses have been located along the lamina terminalis in the preoptic area (7, 10, 13, 24). However, because the brain is remote from the gastrointestinal tract, it makes sense that osmoreceptors or Na+-receptors would also exist in the viscera and detect NaCl soon after it was ingested. In fact, several reports have indicated that thirst and VP secretion in rats might be stimulated by presystemic signals (2, 12, 18), based on experiments in which salt loads were administered either by intragastric intubation or by injection into the hepatic portal vein. The present investigations further pursued this issue by placing the extra salt in food and drinking fluid, the normal vehicles for NaCl loads to enter the body.
It has long been known that rats increase their daily water intake in proportion to the dietary NaCl content (6). Recent analyses indicated that rats did not drink nearly enough water to dilute ingested NaCl to isotonicity during a meal of high-salt food (22). Because those rats formed a dense gastric chyme, it seemed possible that much of the dietary salt was not dissolved in gastric fluid, perhaps because it was embedded in the chyme, and, consequently, it might not be readily detected by osmoreceptors (or Na+-receptors; Ref. 14) located in the proximal duodenum. The present study prevented a dense chyme from forming by surgically terminating the salivary flow of rats, which obliged the animals to drink large volumes of water while eating in order to swallow the dry food. The goals of the study were to determine whether desalivated rats would drink additional water during a meal of high-salt food and, if so, whether presystemic signals provided the stimulus for thirst.
The surgical procedures and all experimental protocols were reviewed by and received approval from the Institutional Animal Care and Use Committee of the University of Pittsburgh.
Ninety-three adult male Sprague-Dawley rats (300–400 g; Harlan Laboratories, Indianapolis, IN) were housed singly in wire-mesh cages located in the vivarium of the Department of Neuroscience at the University of Pittsburgh. The colony room was maintained at a constant temperature (22–23°C) and with a fixed light-dark cycle (lights off from 7 PM to 7 AM). The rats were given tap water, which was presented in water bottles mounted on the front of the cages, and pelleted food containing 1% NaCl (diet TD90229; Harlan Teklad Diets, Madison, WI), which was scattered on the cage floor.
After 1 wk of maintenance, daily food and water intakes were measured in 34 rats. Instead of pelleted food, 16 rats were given either “pulverized diet” (i.e., 60- to 140-mg food particles obtained by pounding the pellets with a hammer) or “powdered diet” (i.e., <15-mg food particles obtained by further pounding the pulverized food), which were provided in small ceramic dishes (4 cm high × 7 cm in diameter). Drinking water was provided in graduated burettes. After 4 or 5 days of maintenance on this new form of diet, daily intakes of water (±0.1 ml) and pulverized or powdered food (±0.1 g) were measured (n = 8, 8, respectively). Food crumbs that fell beneath the cages were collected, dried, and weighed, to more accurately assess food intakes. The 18 other rats were given the pelleted form of an 8% NaCl diet (diet 92012; Harlan Teklad Diets) for an additional week before being switched to the pulverized or powdered form of the diet (n = 8, 10, respectively), and after 4 or 5 days of such maintenance, daily food and water intakes were measured in these animals, as well.
Nine of the 34 rats (from the 18 that had eaten 8% NaCl diet) were not used again in these experiments, 7 (from the 16 that had eaten 1% NaCl diet) were surgically desalivated and used in experiment 1, and the other 18 animals later were killed to provide control values in experiment 1.
Each of 66 rats was anesthetized with halothane or pentobarbital sodium (50 mg/kg), its neck was shaved on the ventral surface, and a 0.75-inch incision was made in the skin and the muscle in the midline. The submaxillary and major sublingual salivary glands, enclosed in a common connective tissue sheath just below the muscle layer, were removed together bilaterally. Then parotid salivary flow was eliminated by bilateral ligation of the parotid ducts where they pass along the surface of the masseter muscle, and the wound in the muscle and external skin was closed by sutures. [The minor sublingual glands were left intact because they contribute negligibly to salivary secretion during stimulation (15)]. This procedure will be referred to as “desalivation” throughout this report, although only four of the six salivary glands actually were removed.
The entire surgical procedure usually lasted <15 min, after which the animals were given an analgesic (ketoprophen, 2 mg/kg sc, for 5 days after surgery) and an antibiotic (30,000 U of penicillin G im). As noted previously (5, 17, 25), the rats quickly recovered from loss of salivary flow by increasing their water intake during meals of dry food, alternating intakes of food and water every 5–10 s or so, and consequently, their food intake and body weight decreased only transiently. These observations confirmed the effectiveness of the surgery because neither this unusual pattern of rapidly alternating food and water bouts, nor a substantial increase in water intake, occurs unless the salivary flow is completely interrupted (16, 17).
All 66 rats were fed the 1% NaCl diet for ∼1 wk while they recovered from surgery. Then for an additional week, 21 animals were given the same 1% NaCl diet while the other 45 rats were given the pelleted form of the 8% NaCl diet. The food was then switched from pellets to the form of the same diet that was used during the subsequent test sessions. The experiments began 4–7 days later, after daily food intakes had stabilized and water intakes measured in 33 rats (n = 17 fed 1% NaCl diet, n = 16 fed 8% NaCl diet) had indicated that much more water was consumed when rats were fed the high-salt diet.
The goal of experiment 1 was to determine whether presystemic signals could provide a stimulus for thirst when desalivated rats ate a high-salt diet. Thirty-three desalivated rats were adapted to an 18-h period of food deprivation (4:00 PM to 10:00 AM) followed by ad libitum access to high-salt diet for 6 h. On three successive days, pulverized (n = 17) or powdered food (n = 16) and drinking water were available during the 6-h period. On the fourth day, food again was removed at 4:00 PM. At 10:00 AM on the following morning, preweighed amounts of food were returned to the home cages. The water bottles were replaced with burettes containing water that was colored with a dark green food dye (McCormick, Hunt Valley, MO) that permitted the fluid to be readily visible in the small intestine (see below). The test period was terminated within 5–45 min, at arbitrary times intended to produce a broad range of intakes, and all rats were immediately killed by decapitation. Food intakes (±0.1 g) and water intakes (±0.1 ml) were recorded as was time spent feeding (±2 s).
Trunk blood was collected in ice-cold heparinized tubes (143 USP sodium heparin; Becton Dickinson, Franklin Lakes, NJ) and kept on ice until stomachs and intestines were obtained. The abdomen was opened and hemostats were placed at the junction of the stomach with the small intestine, at the junction of the stomach with the esophagus, and at the most distal site of visible dye in the small intestine, in that order. (Dye was rarely present in the cecum.) This portion of the surgical procedure took <2 min. Each stomach was removed from the carcass and stripped of adhering blood vessels and connective tissue; its contents were removed and placed in a beaker that was covered with Parafilm, while the intestinal distance containing the dye was measured (±1 cm). The blood was centrifuged (10,000 g for 5 min at 4°C), the plasma was harvested, the pNa was measured (±1 meq/l) using a sodium-sensitive electrode (Synchron EL-ISE model 4410; Beckman Coulter, Brea, CA), and plasma protein concentration was measured (±0.1 g/dl) using a refractometer. Finally, the stomach contents were placed in an oven and dried to constant weight at 60°C for at least 2 days. Subsequently, a fixed volume of distilled water was added to the dried stomach contents from a subset of rats (n = 17), 30–45 min were allowed for the solid matter to soften, and then the suspension was agitated until the solutes were dissolved. The Na+ concentration of the solution was measured using the electrode mentioned above, and that value was used to compute the Na+ content (in milliequivalents) of the gastric chyme.
In addition to those 33 rats, 21 other desalivated rats were treated identically except that they were given either pulverized or powdered food containing 1% NaCl rather than high-salt diet (n = 13, 8, respectively); six animals eating pulverized food were not killed after testing. A group of 18 intact rats, (i.e., rats with uninterrupted salivary flow), given pulverized food containing either 1% NaCl or 8% NaCl (n = 9, 9, respectively), also were treated similarly except they were killed just before the feeding test; measurements of their plasma provided control blood values, while measurements of their gastric contents indicated <0.10 g of solid matter.
To summarize, ingestive behavior was studied in 54 desalivated rats that had been food-deprived overnight. In 48 of them, our goal was to analyze the blood samples for plasma concentrations of Na+ and protein, gastric chyme for dry matter and water (and, in 17 rats, Na+ content), and intestinal distance traversed by the ingested food. Unfortunately, 12 of the 257 individual measurements were lost due to procedural errors. None of the 72 measurements of tissue samples from 18 control rats were lost.
The goal of experiment 2 was to determine whether NaCl ingested in food and fluid would summate to provide an enhanced osmoregulatory stimulus for thirst. The procedures were similar to those used in experiment 1 except this time, when 12 additional desalivated rats were fed powdered or pulverized 8% NaCl diet during the test sessions (n = 7, 5, respectively), 0.10 M NaCl solution was provided instead of water. The saline solution (but not water) had been available ad libitum for 4 days before experiments began. The seven rats fed powdered high-salt diet were killed by decapitation as in experiment 1, and plasma concentrations of Na+ and protein were measured, the contents in gastric chyme of dry matter, water, and Na+ were determined, and the distance traversed by the ingested fluid in the small intestine was recorded. Only one of these 42 measurements was lost. The five rats fed pulverized high-salt diet were not killed after testing.
All data are presented in scatterplots or as means ± SE values. Statistical reliability of observed differences in ingestive behavior was determined by using matched Student's t-tests of mean values. Regression equations were calculated by the method of least squares. P < 0.05 was considered to be statistically significant.
After a transient initial period, daily food intakes were not significantly affected either by surgical desalivation, by high-salt diet, or by the form of diet (Table 1). In contrast, as shown in Table 1, rats fed a diet containing 1% NaCl drank much more water daily after surgical desalivation than before, whether the food was pulverized or powdered (P < 0.01 for both). Desalivated rats drank even more water daily when fed diet containing 8% NaCl (P < 0.001 for both), like intact rats (P < 0.001 for both). The daily water intakes were not significantly different whether the diet was powdered or pulverized, with one exception: much more water was consumed when desalivated rats ate powdered rather than pulverized 1% NaCl diet (P < 0.001).
During the test sessions, desalivated rats ate the powdered diets at a steady rate, whereas they ate the pulverized high-salt diets at a logarithmic rate. As shown in Fig. 1, desalivated rats ate powdered 8% NaCl diet much more slowly (63 ± 5 mg/min) than pulverized 8% or 1% NaCl diet (185 ± 11, 140 ± 8 mg/min, respectively; P < 0.001 for both) but not more slowly than powdered 1% NaCl diet (59 ± 5 mg/min, P = ns).
Previous studies (22) indicated that intact rats drank relatively little water (∼1.3 ml/g food) during an ingestive episode when eating a largely pulverized diet containing 1% NaCl, whereas they drank much more water (∼2.5 ml/g of food eaten) when the diet contained 8% NaCl. Desalivated rats drank considerably more water than that (Fig. 2). They consumed 2.9 ± 0.3 ml/g food while eating pulverized 1% NaCl diet and 4.8 ± 0.2 ml/g food while eating pulverized 8% NaCl diet (P < 0.001). In addition, they consumed 4.1 ± 0.3 ml/g food while eating powdered 1% NaCl diet and 8.8 ± 0.5 ml/g food while eating powdered 8% NaCl diet (P < 0.001). The latter water intakes were significantly larger than those observed when rats ate a high-salt diet that was pulverized (P < 0.001).
An 8% NaCl diet contains ∼1.37 meq Na+/g of food. This value was used to compute the concentrations of the NaCl solutions derived from the dietary NaCl and water that were consumed by desalivated rats during the test sessions (shown in Fig. 2), assuming that all of the ingested NaCl was dissolved in all of the ingested water. Desalivated rats fed powdered 8% NaCl diet drank almost enough water to dilute the ingested NaCl to an isotonic solution (162 ± 8 mM). In contrast, desalivated rats fed pulverized 8% NaCl diet drank much less water per milliequivalent Na+ ingested, and therefore the computed fluid mixture was equivalent to a NaCl solution with a much higher concentration (289 ± 10 mM; P < 0.001).
Figure 3 presents a scatterplot of gastric solids as a function of ingested food in individual animals. It is apparent that rats emptied similar amounts of food from their stomachs regardless of whether they were fed 1% or 8% NaCl diet and whether the diet was in pulverized or powdered form (r = 0.97, P < 0.001). Figure 4 presents a similar plot of gastric water as a function of ingested water by the same animals. Gastric emptying of water was influenced by the form of the diet (i.e., pulverized or powdered) but not by its NaCl content (i.e., 1% or 8%). More specifically, rats emptied 64 ± 4% of ingested water from their stomachs, while they ate powdered food but only 42 ± 2% while they ate pulverized food (P < 0.001).
Figure 5 presents a scatterplot of gastric water as a function of gastric solids. Desalivated rats clearly had more water in the gastric chyme than intact rats are reported to have (∼50%; Ref. 22), and, as expected, the water content was even greater when the rats ate a high-salt diet. The water content of the gastric chyme in desalivated rats eating powdered 8% NaCl diet (80.9 ± 1.1%) was comparable to that in desalivated rats eating pulverized 8% NaCl diet (80.5 ± 0.5%), yet both values were greater than the water content of gastric chyme in desalivated rats eating 1% NaCl diet (74.8 ± 1.0%; P < 0.001 for both). However, these differences were not apparent upon visual inspection of the chyme; the slurry of masticated food and water seemed similar whether rats ate the powdered or pulverized form of either diet.
The Na+ content in the gastric chyme of these rats indicated that NaCl represented less than 8% of the gastric solids, which implies that some of the ingested Na+ was leached from the chyme during the testing period. Figure 6 presents a scatterplot of gastric Na+ as a function of gastric solids in the desalivated rats fed powdered or pulverized 8% NaCl diet. In both groups, gastric Na+ was highly correlated with gastric solids (r = 0.98, 0.99, respectively; P < 0.001 for both). Furthermore, it is evident that more Na+ was leached from the gastric chyme when they ate powdered 8% NaCl diet instead of the pulverized diet (0.56 ± 0.07, 0.30 ± 0.06 meq, respectively; P < 0.02).
Figure 7 presents a scatterplot of the measured Na+ and water contents of gastric chyme in desalivated rats eating 8% NaCl diet. Assuming that all of the Na+ was dissolved in all of the water, it is noteworthy that the fluid mixtures based on these values approximated an isotonic NaCl solution when the rats ate powdered high-salt diet (154 ± 6 mM). However, the computed mixtures were much greater when the rats ate the pulverized diet (282 ± 23 mM, P < 0.002).
Subtracting the amounts of NaCl and water in the gastric chyme from the amounts of NaCl and water that were ingested allows the calculation of the NaCl and water that emptied from the stomachs of these desalivated rats. Assuming that all of the emptied NaCl was dissolved in all the emptied water, the mixture was equivalent to a solution of 151 ± 15 mM NaCl in rats eating powdered high-salt diet and 273 ± 31 mM NaCl in rats eating pulverized high-salt diet (P < 0.01).
The green food dye was found to have traveled deep in the small intestine when the rats were decapitated; observed values of 60–120 cm represent 50–100% of the entire length of the small intestine. Figure 8 indicates that the distance (from the pylorus) was logarithmically related to the time spent eating in all desalivated rats (r = 0.75, P < 0.001).
As shown in Table 2, the Na+ concentrations of systemic plasma were not significantly altered when desalivated rats ate powdered 8% NaCl diet and drank water, whereas they were slightly but significantly lower when desalivated rats ate either pulverized 8% NaCl diet or powdered 1% NaCl diet (P < 0.01, P < 0.001, respectively). Plasma protein levels were elevated only after desalivated rats ate powdered or pulverized 1% NaCl diet (P < 0.001 for both).
An additional group of seven desalivated rats were given 0.10 M NaCl to drink instead of water while eating powdered 8% NaCl diet. These animals ate food half as quickly as rats that ate the same food and drank water (30 ± 3, 63 ± 5 mg/min, respectively; P < 0.001), and they drank considerably more fluid (14.7 ± 1.2, 8.8 ± 0.5 ml/g food, respectively; P < 0.002). Consequently, although much of that fluid emptied quickly from the stomach (see Fig. 4), significantly more water was present in the gastric chyme of desalivated rats when they ate powdered 8% NaCl diet and drank 0.10 M NaCl instead of water (Fig. 5; 85 ± 1%, 81 ± 1%, respectively; P < 0.05).
As indicated in Fig. 9, the relation between food and saline intake was logarithmic, not linear, when rats ate powdered high-salt diet and drank 0.10 M NaCl. Closer inspection of these data indicates that five rats ate 1.3 g or less of the powdered high-salt diet, whereas two animals ate 2.0 g or more, and the fluid intakes of the two subgroups (per gram of food intake) were significantly different (16.1 ± 0.8, 11.0 ml, respectively; P < 0.005). Thus further analysis of the data from this group considered the two subgroups separately.
Assuming that all the Na+ derived from ingested NaCl and 0.10 M NaCl was dissolved in all of the ingested fluid, the concentration of the computed fluid mixture was similar to, but slightly higher than, the values observed when desalivated rats ate powdered 8% NaCl diet and drank water. More specifically, the above-mentioned subgroup of five rats that drank 0.10 M NaCl, while eating a powdered high-salt diet had a calculated fluid mixture equivalent to 186 ± 5 mM NaCl. The fluid mixture from measured gastric Na+ and water contents in those animals (shown in Fig. 7) computed to a similar value, 180 ± 8 mM NaCl (P < 0.05 compared with rats eating powdered high-salt diet and drinking water), as did the calculated mixture of NaCl and water that emptied from their stomachs, which was equivalent to 186 ± 4 mM NaCl. Note that the two rats excluded from these calculations drank the least amount of fluid among rats in this group when expressed per gram of food intake (see Fig. 9); consequently, the computed concentrations of the fluid mixture in their stomachs (211, 231 mM) were the highest in the group (see Fig. 7) as were the computed concentrations of the emptied fluid (213, 244 mM, respectively).
The green food dye was found to have traveled deep in the small intestine when these rats were decapitated. The intestinal distance (from the pylorus) followed the same logarithmic relation to the time spent eating seen in all the other animals (Fig. 8). As shown in Table 2, the Na+ levels of systemic plasma were significantly elevated when desalivated rats ate powdered 8% NaCl diet and drank 0.10 M NaCl instead of water (P < 0.001).
Finally, five other desalivated rats were given 0.10 M NaCl to drink instead of water while eating pulverized 8% NaCl diet. These animals ate more slowly than rats that ate the same high-salt food and drank water (101 ± 9, 185 ± 11 mg/min, respectively; P < 0.001) and, as indicated in Fig. 9, they drank significantly more fluid (6.9 ± 0.3, 4.8 ± 0.2 ml/g food, respectively; P < 0.001). However, they did not drink nearly as much saline as rats eating powdered 8% NaCl diet did (P < 0.001). The rats that drank 0.10 M NaCl while eating pulverized high-salt diet had a computed fluid mixture equivalent to 299 ± 10 mM NaCl, which was similar to the computed mixture when rats ate the same food but drank water (289 ± 10 mM NaCl, P = ns).
The goal of the present experiments was to determine whether putative visceral osmoreceptors (or Na+-receptors; Ref. 14) could provide an early signal of thirst in rats. The obtained results appear to support this possibility. In short-term tests (i.e., up to 45 min) conducted after 18 h of food deprivation, desalivated rats increased their already elevated water intakes when fed 8% NaCl diet, especially when the food was powdered. In fact, they increased water consumption so much that the computed concentration of the ingested amounts of Na+ and water was approximately isotonic to body fluids, as was the computed concentration of the gastric Na+ and water contents. Those findings suggest that the rats were able to rapidly detect the NaCl ingested in powdered high-salt diet, that thirst was stimulated in consequence, and that the rats drank water until the ingested NaCl was diluted to isotonicity and the thirst stimulus was thereby removed. Each of these hypotheses will be considered, in turn.
How did the rats detect the dietary NaCl load? Although there is no evidence for appropriate receptors in the stomach, several studies have suggested that visceral osmoreceptors can play a role in the stimulation and inhibition of VP secretion in rats. For example, VP secretion was stimulated when hypertonic NaCl solution was given by gavage, before systemic plasma osmolality (pOsmol) was increased (2). This effect was even more pronounced when rats were dehydrated before the gastric NaCl load was administered (18). Conversely, an established VP secretion was inhibited in dehydrated rats by a gastric water load before systemic pOsmol was lowered, and it occurred even when concentrated NaCl solution was infused intravenously to prevent osmotic dilution (1). The physiological significance of the latter observations was demonstrated in recent reports that water consumption by thirsty rats decreased VP secretion before elevated pNa had returned to normal (9, 21).
Visceral osmoreceptors also appear to provide an early stimulus for thirst in rats. For example, Kraly et al. (12) found that euhydrated rats given a gastric load of 0.5 M NaCl began to drink water before systemic pOsmol became elevated. As might be expected, considerably more water intake was stimulated when dehydrated rats received a gastric NaCl load (18). In the present experiments, the amount of water consumed (per gram of food eaten) by desalivated rats eating powdered 8% NaCl diet was significantly greater than the amount consumed when powdered 1% NaCl diet was eaten (Fig. 2). If visceral osmoreceptors were present in the proximal duodenum, they would be ideally situated to detect ingested NaCl soon after it left the stomach and to stimulate water intake when the concentration of the emptied fluid was hypertonic. For this stimulation to occur and a concentration of 150 mM to be produced in the stomach, none of the ingested NaCl must be retained in the food or embedded in the gastric chyme, but instead, all of it must be dissolved in the gastric fluid so that it could be accessible to the putative osmoreceptors. Presumably, that is what happened when desalivated rats ate powdered 8% NaCl diet; each feeding bout increased the NaCl concentration of the fluid emptying from the stomach, and each drinking bout restored it to isotonicity. In contrast, when the rats ate pulverized 8% NaCl diet, the ingested NaCl may not have been fully accessible until the chyme was digested in the distal small intestine. That hypothesis is consistent with observations that the ingested food and fluid did, in fact, reach the distal small intestine while the rats were eating (Fig. 8). According to this arrangement, systemic pOsmol would not be raised until after a meal of pulverized 8% NaCl diet, which would then allow cerebral osmoreceptors to stimulate thirst and contribute to the increased daily water intakes.
Desalivated rats must drink water to eat dry food. When they ingest standard chow containing 1% NaCl, their increased water intakes do not reflect increased thirst but a need for water to lubricate the mouth and throat, which enables them to swallow the food (11). Evidently, they need much more water to consume food when the 1% NaCl diet is powdered rather than pulverized (Fig. 2). In either case, the amounts of water ingested ordinarily are much greater than the volumes needed for water balance, and the excess is excreted in urine. However, that extra water could be used to support osmoregulation when the rats eat a high-salt diet and have a large osmotic load. In fact, the amount of water they drank to swallow powdered 1% NaCl diet is sufficient to dilute to isotonicity the NaCl load in ∼4% NaCl diet. Thus, while eating powdered 8% NaCl diet, the rats had to drink even more water than usual to neutralize the osmotic effect of the entire dietary NaCl load. That is what they did. The absence of a significant increase in systemic pNa suggests that cerebral osmoreceptors did not participate in the stimulation of thirst.
The same considerations can be applied to the food and fluid intakes observed when desalivated rats ate pulverized 8% NaCl diet. Those animals increased their water intake significantly above the amounts consumed by desalivated rats eating pulverized 1% NaCl diet, which is consistent with observations that some NaCl had been leached from the food (Fig. 6). On the other hand, those volumes of water were much smaller than the amounts needed to dilute the entire ingested NaCl load to isotonicity. By explanation, we propose that NaCl consumed by rats ingesting pulverized high-salt diet is not fully dissolved in gastric fluid and therefore cannot fully stimulate visceral osmoreceptors soon after it empties into the duodenum. In this regard, note that systemic pNa was not elevated while the rats were eating pulverized high-salt diet (Table 2), which suggests that most of the chyme had not yet been digested and influenced body fluids.
If this hypothesis is correct, then it should be possible to increase fluid intake by adding more NaCl to the gastric fluid whether desalivated rats eat powdered or pulverized 8% NaCl diet. Experiment 2 was designed to test that hypothesis, not by further increasing the NaCl content of the diet but by adding NaCl to the drinking fluid. The goal was to determine whether NaCl derived from the drinking fluid would summate with NaCl derived from the food to provide an augmented stimulus for thirst. It certainly did when rats ate powdered 8% NaCl diet. Rats drank much more fluid per gram of consumed food when 0.10 M NaCl was available to drink instead of water (14.7, 8.8 ml/g food, respectively). Curiously, the computed intakes of the five rats that drank the most saline per gram of food eaten were equivalent to 186 mM NaCl solution, which approached, but did not reach, isotonicity. Similarly, the fluid mixture from measured gastric Na+ and water contents in those five animals computed to 180 mM NaCl. Why didn't they drink enough fluid to reduce the concentration of the gastric fluid mixture to 150 mM? One possibility is that the volumes required for such dilution (∼27 ml/g food) were so large that before those amounts could be consumed, gross distension of the stomach and small intestine provided a signal inhibiting thirst (8) but not hunger in these food-deprived rats. Consequently, their fluid intakes were reduced to the amounts (4.1 ml/g for food intakes >1.3 g; see Fig. 9) consumed by desalivated rats while eating powdered 1% NaCl diet (4.1 ± 0.3 ml/g food; see Fig. 2). In this regard, the measured gastric water contents, which generally were very large in desalivated rats while eating, were especially pronounced when they ate powdered 8% NaCl diet and drank 0.10 M NaCl (Fig. 5), and their small intestines were visibly swollen with fluid. Furthermore, their systemic pNa was significantly elevated (Table 2), which is consistent with calculations that a hypertonic solution had emptied from the stomach. In addition, it is possible that a portion of the gastric Na+ remained associated with the ingested food and was not in solution.
In summary, surgically desalivated rats, like intact rats, increased daily water intake when fed 8% NaCl diet. Of more significance, desalivated rats drank much larger amounts of water during a meal when the high-salt diet was powdered than when it was pulverized. It seems likely that ingested NaCl was readily leached from powdered food into the watery gastric fluid and thus could be detected by duodenal osmoreceptors (or Na+ receptors) when the fluid emptied from the stomach, which stimulated increases in water intake that were sufficient to dilute the gastric NaCl to isotonicity. This rapid leaching of NaCl may not occur when desalivated rats ate the larger particles of pulverized high-salt diet, hence the smaller water intakes in the latter group. These and other findings suggest that visceral osmoreceptors (or Na+ receptors) detect hypertonicity in the fluid that leaves the stomach and stimulate thirst, as if the imminent rise in systemic pOsmol was anticipated. The precise nature and location of these receptors are uncertain and remain to be determined.
Osmoreceptors in the lamina terminalis (including the organum vasculosum, a circumventricular organ in the preoptic area) are well known to detect increases in systemic pOsmol and to initiate thirst and VP secretion (23). These receptors evidently are necessary for the two responses because both are eliminated by destruction of the ventral portion of the lamina terminalis in rats (7, 10). The cerebral osmoreceptors are stimulated whenever systemic pOsmol is elevated above normal levels, as happens when body water is lost during dehydration. An increase in systemic pOsmol also occurs naturally when dry food is consumed. However, an earlier, presystemic signal would result if visceral osmoreceptors could detect hyperosmolal fluid passing through the stomach and small intestine. Those are the receptors that are proposed to stimulate VP secretion after a NaCl load is administered to rats by gavage (2), and those same receptors also might mediate thirst in the present study when desalivated rats were fed powdered high-salt diet.
The value of an early osmoregulatory signal of thirst is illuminated in the response of rats after that signal has been eliminated. Rats often drink water and concentrated NaCl solution within the same ingestive episode when salt appetite is induced by experimental treatment (19, 20). This rapid message of thirst likely is mediated by vagal afferent nerves that project from the stomach and small intestine to the caudal brain stem. Destruction of those sensory fibers by the neurotoxin capsaicin, or of their projection sites in the caudal brain stem by surgical lesions, cause rats to drink concentrated NaCl solution for a longer period of time, and therefore in much larger amounts, than control rats do before switching to water (3, 4). Thus these rats behave as if they could not detect the concentrated fluid leaving the stomach until it elevated systemic pNa and thereby activated cerebral osmoreceptors.
On the other hand, it should be emphasized that both surgical desalivation and powdered high-salt diet were required for rats to drink large amounts of water to osmoregulate during meals of dry laboratory food. The fact that these experimental conditions are rather unusual implies that rats ordinarily are very well buffered against the early stimulation of thirst while they eat. In other words, the putative visceral osmoreceptors are not fully stimulated by an ingested NaCl load unless the load is already in solution or soon will be, and therefore cerebral osmoreceptors probably play an important role in detecting the load and stimulating thirst and VP secretion despite their distance from the gastrointestinal tract.
This research was supported in part by a grant from the U.S. National Institute of Mental Health (MH-25140).
The authors are grateful for the expert technical assistance of Jamie Spicer and April Protzik, and the helpful comments of Michael Bushey, Michael Bykowski, Carrie Smith, and Jennifer Vaughan.
Preliminary versions of this report were presented at the annual meeting of the Society for the Study of Ingestive Behavior, in Cincinnati, OH, in July 2004 and in Pittsburgh, PA, in July 2005.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society