INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

INCREASES IN THE EFFECTIVE osmolality of plasma (P_osmol) are known to activate neurons in the organum vasculosum of the lamina terminalis (OVLT) in the basal forebrain (29). Lesions of this circumventricular organ markedly attenuate the stimulation of thirst and neurohypophyseal secretion of vasopressin (VP) and oxytocin (OT) by hyperosmolality in rats (17, 26, 36), indicating the importance of the OVLT in mediating these adaptive responses. However, other osmoreceptors may contribute to those effects as well. For example, the existence of peripheral osmoreceptors has been suggested by reports that intragastric intubation of hypertonic saline (ig HS) stimulates thirst and VP secretion in rats before systemic P_osmol increases to levels detectable by cerebral osmoreceptors (9, 25). Although relatively small increases in water intake and VP secretion were observed in those studies, it seems plausible that the effects were not fully expressed because euhydrated animals were used and therefore their brains received a mixed osmoregulatory message, that is, a signal of hydration from cerebral osmoreceptors may have blunted a signal of incipient hyperosmolality from putative peripheral osmoreceptors. The present experiments tested this hypothesis by determining whether the increase in water intake and plasma VP (pVP) induced by intragastric HS would be more substantial when the gastric load was given either to rats after overnight water deprivation or to rats whose P_osmol had been elevated by pretreatment with intravenous infusion of HS (iv HS). Plasma levels of OT (pOT) also were measured because this peptide is known to make important contributions to osmoregulation as a natriuretic hormone in rats (18, 19, 41).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals. Adult male Sprague-Dawley rats (Zivic Laboratory, Zelienople, PA) weighing 300-350 g were used in this study. They were housed individually in wire-mesh cages in a colony room with ambient temperature of 22-24°C and with lights on from 7:00 AM to 7:00 PM. The rats had ad libitum access to Laboratory Chow pellets (Purina no. 5001) and tap water before experiments began. All procedures for the treatment of animals were in strict compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee at the University of Pittsburgh.

Procedure. To acclimate the animals to the gastric intubation procedure, intragastric loads of isotonic saline (ig IS) were administered to all rats one time each day for 3 days before testing. A transverse V-shaped metal spring wire was placed between the rats' front teeth to keep their mouths open, and then a 10-cm length of a polyethylene feeding tube (Pharmaseal, Toa Alta, Puerto Rico) was gently inserted down the esophagus into the stomach. The 4-ml load was administered over ~30 s. Rats showed no visible signs of discomfort by the third load.

Effects of intragastric HS on drinking behavior in water-deprived rats. Three groups were studied to determine the effect of intragastric HS on drinking in water-deprived rats. One group (n = 7) was both water deprived overnight and given a 4-ml gastric load of either 0.15 or 0.5 M NaCl before the drinking test. A second group (n = 10) was not water deprived before receiving either intragastric IS or intragastric HS, whereas a third group (n = 5) was deprived of drinking water overnight but was not given a gastric load before testing. In the first and second groups, the two gastric preloads were administered to each rat in a counterbalanced order with 3-5 days separating the tests. Water was made available immediately after the loads to water-deprived rats, and 5 min after the loads to nondeprived rats, and intakes were recorded every 15 min for 1 h.

Effects of intragastric HS on neurohypophyseal hormone secretion in dehydrated rats. Two different protocols were used to examine the effects of intragastric HS on neurohypophyseal hormone secretion in dehydrated rats. In the first protocol, which mirrored the drinking study described above, rats were water-deprived overnight while food remained available. On the following day, they were given a 4-ml gastric load of either 0.15 or 0.5 M NaCl (n = 7 and 7). They were decapitated 25 min later, and trunk blood was collected in ice-cold heparinized tubes (143 USP sodium heparin). For purposes of comparison, blood samples also were taken from other rats given intragastric IS or intragastric HS treatments but not water deprived (n = 6 and 6). All samples were centrifuged, and the plasma was removed. P_osmol was measured immediately in aliquots, as above, while the remainder was frozen for later RIA of VP and OT.

Statistics. The results were presented as group mean values ± SE. The effects of intravenous HS and of gastric loading were evaluated by ANOVA or t-test that compared treatment values with baseline values within groups. In experiments involving repeated blood sampling, data were analyzed by two-way ANOVA (treatment × time) with repeated measures, followed by paired t-tests for planned comparisons with a group and Tukey's honestly significant difference test for comparisons between groups. The baseline value for the intravenous HS treatment was obtained from blood taken at the end of the intravenous IS infusion period, whereas for the gastric loads it was obtained from blood taken at the end of the intravenous HS infusion period. Regression lines and correlation coefficients (r) were computed from individual values by the method of least squares. Comparisons of subsets of data with large variability were made by ² analysis. A P value < 0.05 was considered to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of intragastric HS on drinking behavior in water-deprived rats. During the hour after intragastric IS alone, rats consumed little water (0.5 ± 0.5 ml), as expected; in fact, only 3 of 10 rats drank at all. Eight of 12 rats ingested water after receiving intragastric HS alone, but the amounts consumed were not significantly larger (2.6 ± 0.8 ml; Fig. 1A). In contrast, all rats deprived of water overnight before receiving intragastric IS drank water rapidly and in large amounts (8.0 ± 1.2 ml in 15 min) similar to those consumed by rats that had been water-deprived but not given a gastric load (7.6 ± 1.5 ml). Dehydrated rats given intragastric HS drank the most water in 15 min (14.9 ± 1.0 ml), much more than the other two groups consumed (all P < 0.001). The drinking rates of rats in the three water-deprived groups then diminished comparably during the remainder of the test (Fig. 1B).

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Cumulative mean ± SE water intakes by rats given a 4-ml ig intubation of either 0.15 M NaCl [ig isotonic saline (IS)] or 0.5 M NaCl [ig hypertonic saline (HS)] alone (A; n = 10 and 12, respectively) or after overnight water deprivation (B; n = 7 and 7). Water was provided at the beginning of the test, and intakes were measured every 15 min. Dehydrated rats drank more water at all times after ig HS than after ig IS (* all P < 0.01), whereas rats that had not been water deprived drank little or no water throughout the test regardless of gastric load.

Effects of intragastric HS on neurohypophyseal hormone secretion in dehydrated rats. In one study, rats were given intragastric HS either alone or after overnight water deprivation. In both groups, the load had no statistically reliable effects on P_osmol measured 25 min later (Fig. 2). Nonetheless, intragastric HS increased pVP and pOT significantly in those animals (all P < 0.01 compared with values after intragastric IS; Fig. 2). Importantly, larger effects on pVP and pOT were observed when intragastric HS was given after water deprivation than when it was given alone (pVP, P < 0.02; pOT, P < 0.01). The effects of the gastric loads on pVP and pOT in individual rats are presented in Fig. 3; clearly, the range of P_osmol values overlapped considerably in rats given intragastric IS or intragastric HS, especially when the loads were given after overnight water deprivation.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effect of ig IS () or ig HS () on mean ± SE values of plasma osmolality (Posmol), and plasma levels of vasopressin (pVP) and oxytocin (pOT), in rats that had not (left) or had been (right) deprived of drinking water overnight. Blood samples were obtained from two groups of rats not given the gastric loads (BL; n = 10 and 6, respectively) and from two other groups 25 min after receiving the loads (n = 6 and 7, respectively). Treatment with ig HS significantly increased pVP and pOT compared with levels seen after ig IS (* all P < 0.01) and, in water-deprived rats, compared with BL levels ( P < 0.01). Note that ig HS had no significant effect on Posmol regardless of pretreatment.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Effect of ig IS () or ig HS () on plasma levels of VP and OT, plotted as a function of the associated Posmol, in rats that had not (A and C; n = 6) or had been (B and D; n = 7) deprived of drinking water overnight. Symbols represent values from individual animals, measured 25 min after the gastric loads. Mean values of Posmol, pVP, and pOT are shown in Fig. 2. Higher values of pVP and pOT were seen when rats were given ig HS instead of ig IS (all P < 0.01).

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Effect of 4-ml gastric loads of IS () or HS () on mean ± SE values of Posmol, pVP, and pOT in rats (n = 11) that had been infused with 1 M NaCl (2 ml/h iv for 2 h). Values of all three variables before the gastric loads were given (0 min) were greater than baseline values before the start of the infusion (BL; all P < 0.01). Posmol did not change significantly in either group during the 25-min test period after either load. In rats given ig IS, pVP did not change significantly, although pOT decreased at 25 min ( P < 0.01) but not at 15 min. In contrast, rats given ig HS had higher levels of pVP and pOT after the load than before the load ( all P < 0.01 except for pVP at 15 min) and higher levels of both hormones than rats did after ig IS (* all P < 0.01).

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Effect of ig IS ( and ) or ig HS ( and ) on plasma levels of VP (A) and OT (B), plotted as a function of the associated Posmol, in rats (n = 11) that had been infused with 1 M NaCl (2 ml/h iv for 2 h). Symbols represent values from individual animals at 15 ( and ) and 25 ( and ) min after the gastric loads. Mean values of Posmol, pVP, and pOT are shown in Fig. 4. Higher values of pVP and pOT were seen when rats were given ig HS instead of ig IS (all P < 0.001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It is well recognized that thirst and VP secretion are stimulated in rats and other animals by increases in the osmolality of blood in the general circulation. An increase in systemic P_osmol also stimulates neurohypophyseal secretion of OT in rats (34). Cerebral osmoreceptors located in the basal forebrain are known to mediate those adaptive osmoregulatory responses (35). The present results indicate that the same responses occurred in euhydrated rats after small intragastric HS loads were administered, before an increase in systemic P_osmol was seen, although the effects obtained were relatively small. These findings are consistent with previous observations (9, 25). The main point of the present report is that a more substantial stimulation of water intake and neurohypophyseal VP and OT secretion occurred when, before the intragastric HS treatment, rats either were water deprived overnight or received intravenous HS to increase P_osmol. Thus coactivation of cerebral osmoreceptors appears to potentiate the effects of peripheral osmoreceptors (or Na⁺ receptors) to stimulate thirst and neurohypophyseal secretion in rats.

Kraly et al. (25) reported that rats began to drink water 10-15 min after they received gastric loads of hypertonic NaCl solution, by which time P_osmol had not yet changed. In fact, it can be calculated that the smallest load given in that study, 2 ml of 0.3 M NaCl, would not have increased P_osmol in adult rats by 2%, the apparent threshold for thirst (15), even after the entire load had been absorbed and equilibrated with body fluids. Thus it seems unlikely that cerebral osmoreceptors were responsible for stimulating the drinking response. A similar conclusion may be drawn from the results of the present study, in which rats were deprived of water overnight before receiving a 4-ml gastric load of 0.5 M NaCl solution. Those animals drank much more water in 15 and 30 min than water-deprived rats given either intragastric IS or no gastric preload even though significant changes in P_osmol were not observed during this time period. Note that, in water-deprived rats given intragastric HS treatment, it can be calculated that the P_osmol would have increased by ~13 mosmol/kgH₂O if the entire load had been absorbed, equilibrated with body fluids, and retained. Because the observed increase in P_osmol was only ~3 mosmol/kgH₂O at 25 min after the load was administered, and only ~8 mosmol/kgH₂0 at 55 min, it seems likely that the load was absorbed very slowly. It can be calculated that the absorption of ~1 ml (i.e., ~25% of the 4-ml load) of 0.5 M NaCl would raise P_osmol by ~3 mosmol/kgH₂O in these animals, which is consistent with previous observations of gastric emptying in water-deprived rats (28).

Although these data suggest that a peripheral signal of thirst resulting from the intragastric HS treatment was potentiated in dehydrated rats, there is an alternative explanation of the findings. The presence of HS in the stomach should increase the concentration of fluid passing to the intestines after rats drank water, thereby reducing the effectiveness of ingested water in providing hydrational signals that inhibit water intake. To determine whether the observed effects of intragastric HS resulted from an extra excitatory signal or from a diminished inhibitory signal, it was necessary to study this phenomenon under circumstances in which drinking water was not available. That was the protocol used to investigate neurohypophyseal hormone secretion, and the results indicate that intragastric HS treatment caused a substantial increase in plasma levels of VP and OT without significantly affecting P_osmol in rats pretreated either with overnight water deprivation or with intravenous HS. These findings suggest that an extra excitatory signal was present, which also may have contributed to the increased water intake observed in the drinking experiment.

In a similar experiment, Carlson et al. (6) gave euhydrated rats 2.9 ml of 0.3 M NaCl intragastrically, and pVP increased by ~2.5 pg/ml when measured 10 min later. This effect was associated with a significant increase in the P_osmol of blood in the hepatic portal vein but not in systemic blood. In the present experiments, euhydrated rats were given 4 ml of 0.5 M NaCl intragastrically, and pVP increased by ~4 pg/ml at 25 min after the intubation. In contrast, when rats were water deprived before receiving intragastric HS, pVP increased by ~15 pg/ml when measured 25 min later. Values of pOT paralleled those changes, increasing by ~10 and by ~30 pg/ml in the same blood samples. Similarly, when rats were pretreated with intravenous HS, pVP increased by ~12 pg/ml when measured 25 min after intragastric HS, whereas pOT increased by ~24 pg/ml. Thus greater increases in neurohypophyseal secretion were observed in response to the same intragastric HS treatment when rats were dehydrated rather than euhydrated, despite the absence of further increases in P_osmol. The relative magnitude of the neurohypophyseal secretions stimulated by intragastric HS resembles the reported effects in rats of HS or hypertonic mannitol solution administered systemically, or of hypovolemia; each of these treatments evokes increases in pOT that are approximately two times greater than the increase in pVP (20, 34).

The present results suggest that intragastric NaCl loads generate peripheral signals that precede significant absorption of NaCl into the general circulation and its detection by cerebral osmoreceptors. Postgastric receptors in the splanchnic area would be ideally located to sample solutions and influence ongoing behavioral and physiological responses before gastric NaCl loads enter the general circulation. In this regard, the hepatic portal vein already has been implicated as a site of osmo- or Na⁺ receptors (27, 39). Vagal afferent nerves responsive to HS infused in the hepatic portal vein are known to project to the nucleus tractus solitarius (NTS) subadjacent to the area postrema (AP) in the brain stem (22, 24). If this neural pathway mediates an early osmoregulatory signal that affects fluid intake and neurohypophyseal hormone secretion, then that signal should be eliminated after destruction of this pathway or its projection sites in the brain stem. Consistent with this expectation, vagotomized rats drank larger volumes of concentrated saline solution than control rats did in response to various stimuli (39), as if they were not receiving an early signal of imminent hyperosmolality. Rats with lesions of the AP/NTS similarly drank larger amounts of concentrated saline solution (10, 33), as did rats with damage to peripheral sensory fibers caused by systemic administration of the neurotoxin capsaicin (11). An analogous effect of AP/NTS lesions to attenuate the increased VP and OT secretion in response to intravenous HS has been reported (20), although such lesions did not blunt the early stimulation of VP secretion by intragastric HS treatment (7, 8, 40).

Previous reports indicate that thirst and secretion of pituitary VP and OT are stimulated in an approximately additive fashion in rats when P_osmol is elevated while plasma volume is reduced (32, 34). Those findings have been interpreted to signify that the neural circuits involved in the control of thirst and neurohypophyseal secretion respond to multiple sensory signals with little interaction (also see Ref. 16). The same functional arrangement does not describe the present findings. Although the effects of intragastric HS treatment are relatively small when the animals are well hydrated, they are more substantial when animals are dehydrated or when P_osmol already is elevated. In other words, the osmoregulatory system appears to operate as if there was a gating mechanism that inhibits the peripheral signals when cerebral osmoreceptors detect euhydration and disinhibits them when cerebral osmoreceptors detect dehydration. Whatever the mechanism, the adaptive significance of this functional arrangement is plain. When rats are dehydrated and consume osmolytes in concentrated NaCl solution or in food, it is useful for them not to wait for P_osmol to increase before secreting OT and VP and increasing water consumption, so that they avoid becoming too dehydrated.

When rats are dehydrated and consume water, it is similarly useful for them not to wait for P_osmol to decrease before terminating VP secretion and ongoing water consumption, thereby avoiding overhydration. An early signal of hydration was discussed many years ago (1, 4) to explain why drinking by dehydrated dogs stopped well before the ingested water was absorbed. This anticipatory element in the control of water intake was clarified subsequently in a series of elegant investigations reported by Appelgren et al. (2) and Thrasher et al. (37, 38). Briefly, an early inhibition of thirst and VP secretion was observed after water consumption by dogs fitted with a gastric fistula, which drained the stomach and thereby prevented the possibility of rehydration. The same rapid inhibitory effects occurred when dogs drank HS solution (although, ultimately, when the saline was absorbed and P_osmol was elevated, the dogs became even thirstier and secreted more VP than before, as might be expected). These observations highlight the importance of an early inhibitory signal in the control of water intake and suggest its basis: a neural input to the brain from the oropharynx, associated with rapid swallowing during the act of drinking. This signal, which essentially allowed thirsty dogs to meter their intake, had a rapid but temporary inhibitory effect on thirst and VP secretion. When ingested water was subsequently absorbed and P_osmol was diluted back to normal levels, a more sustained termination of thirst and VP secretion was produced because of systemic rehydration.

Several other species, including humans (31), also use oropharyngeal signals to inhibit thirst and neurohypophyseal secretion during water consumption. Similarly, in rats pretreated with intravenous HS, water drinking provided a rapid stimulus to inhibit VP and OT secretion before a decrease in P_osmol was seen in systemic blood (21). When thirsty rats drank IS instead of water, however, no change in pVP or pOT was observed (21). Thus the concentration of the fluid that rats consume, not its volume, seems to be the critical variable in providing early inhibitory signals (3, 5). Furthermore, vagotomized rats (23, 25), capsaicin-treated rats (11), and rats with lesions of the AP/NTS (12, 14) all drank much more water than control rats did in response to various thirst stimuli, as if they were not receiving an early satiety signal. The present findings, together with those previous observations, therefore support the hypothesis that in rats peripheral osmo- or Na⁺ receptors detect the osmotic consequences of fluid ingestion, whether increases or decreases in P_osmol, and inform the brain of imminent changes in systemic P_osmol before they could be detected by cerebral osmoreceptors. Remarkably, this early signal of hydration reduces thirst and neurohypophyseal secretion despite the continued stimulation of cerebral osmoreceptors by elevated P_osmol. These rapid and potent feedforward effects prepare animals for similar but more gradual signals, much like the familiar autonomic reflex involving the taste and smell of food stimulates insulin secretion before ingested food has been absorbed from the gastrointestinal tract.