INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE MOUTH IS A MAJOR DETERMINANT of whether sodium and other nutrients are accepted, but does it also control how much is ingested? Findings that sodium deficiency influences gustatory sensitivity to NaCl (2, 3) suggest a potential mechanism by which gustatory factors could modulate intake. More direct evidence of "oral metering" or "sensory-specific satiety" is provided by several early experiments. First, sodium-deficient rats drank similar amounts of sodium salts that were equated for taste, even though they differed markedly in sodium content (300 mM NaCl vs. 30 mM Na₂CO₃) (7). Second, when the postingestive effects of drinking sodium and glucose were equated by infusing into the stomach the opposite solution to the one ingested, sodium-deficient rats ingested sodium but did not ingest appreciable amounts of glucose (17). Third, and perhaps most convincingly, sodium-deprived rats did not compensate accurately for preloads of NaCl given by intubation into the stomach. That is, intake of sodium solution was greater after intubation of NaCl than after ingestion of the same amount of sodium (8, 24). The implication made from this was that oral stimulation by NaCl normally limits NaCl intake.

In contrast, several findings suggest that oral stimulation does not influence the amount of NaCl ingested by sodium-deficient rats. Sodium-deficient rats sham drank concentrated NaCl solutions with little sign of satiation (12, 20). Moreover, compensation was nearly perfect for chronic infusions of NaCl into the stomach (6) or slow infusions into the hepatic portal vein (20).

There are two explanations for these apparently contradictory series of findings. First, it is possible that oral and postingestive controls participate in the control of NaCl intake, and the conditions of each experiment selectively favor one or the other locus of control. Second, it may be that the results of studies supporting oral control of intake are misinterpreted, because the experimental manipulations have unintended effects on postingestive mechanisms.

We have used two approaches to investigate the issue of oral vs. postingestive controls of NaCl intake. First, we examined whether sodium-deprived rats reduced NaCl intake after oral experience with sodium. We found no reduction in subsequent NaCl intake in rats that sham drank NaCl for 15 min. Second, we compared the behavioral and physiological effects of NaCl ingestion with NaCl intubation. This replicated earlier findings that NaCl intake was suppressed less by intubation than by ingestion of an NaCl preload (8, 24). Work with other nutrients shows that the physiological response to, and fate of, a preload depends on its route of administration (10). To examine whether physiological events were differentially affected by ingestion and intubation of NaCl, we measured gastrointestinal sodium content and plasma concentrations of sodium and sodium-regulatory hormones at various times after sodium-deficient rats drank or were intubated with an NaCl load. One result suggested the possibility that circulating aldosterone concentrations may be influenced by oral factors, so in the final study in this series, we compared the plasma aldosterone concentrations of sodium-deprived rats that ingested or sham ingested NaCl.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and Maintenance

Induction of Sodium Deficiency

Experiment 1: Effect of Sham Ingestion of NaCl on Subsequent NaCl Intake

Gastric cannula. Each rat was implanted with a stainless steel gastric cannula (14 mm long, 6 mm ID; Cornell University Apparatus Shop, Ithaca, NY). A 16-mm-diameter flange rested inside the stomach, and a stainless steel removable machine screw prevented drainage. The cannula was implanted while the rat was anesthetized intramuscularly with a mixture of 100 mg/kg ketamine, 0.7 mg/kg acepromazine, and 2.5 mg/kg xylazine. It penetrated the greater curvature of the glandular portion of the stomach close to its border with the rumenal portion and was held in place by a purse string of 2-0 silk suture in the stomach wall around the cannula. The cannula was exteriorized through a stab wound ~1 cm to the right of the ventral midline. A stainless steel washer was screwed to the outside of the cannula to prevent it from migrating back inside the rat. This washer was removed 3-4 days after surgery to improve air circulation to the wound and, thus, speed recovery. An opioid analgesic (0.5 mg/kg sc butorphanol tartrate) and topical and systemic antibiotics (Vetropolycin ointment and 2.5 mg/kg im gentamicin, respectively) were given at 0, 24, and 72 h after surgery.

Procedure. At ~3 wk after surgery the 16 rats weighed 402 ± 13 g. At this time, the rats began a series of weekly tests. On the day after induction of sodium deficiency with furosemide (see above), the rats were prepared for sham drinking. Each rat's gastric cannula cap was unscrewed, and the stomach was thoroughly rinsed with warm water to remove its contents. A drainage tube was attached to the cannula, and the rat was placed in a Plexiglas chamber (20 cm long, 11.5 cm wide, 30 cm high). A slot running the length of the floor allowed the drainage tube to exit the chamber. The tip of the drainage tube rested in a preweighed plastic beaker under the chamber, so that drainage could be collected and weighed.

Experiments 2 and 3: Behavioral and Physiological Effects of NaCl Ingestion and Intubation

Adaptation to sodium deprivation and testing. All the rats were pretested to adapt them to experimental procedures. At approximately noon on the day after induction of sodium deficiency (see above), the rats received 6 ml of 500 mM NaCl solution to drink. It is difficult to present such a small volume accurately using a sipper tube, so the NaCl was given in a cup-shaped ceramic crucible (3 cm high, 3 cm diameter). The base of the crucible was glued to the middle of a 4 × 8-cm piece of wire mesh so that the rat could not overturn it. The crucibles were left in the rats' cages until all the NaCl was consumed or overnight on the rare occasions a rat would not drink the solution immediately. Sodium-deficient diet was replaced with the rats' maintenance chow when the crucible was removed.

Experiment 2: behavioral experiment. The design of experiment 2 involved giving sodium-deficient rats an oral or intubated NaCl preload 15 or 60 min before they received an NaCl intake test. Two replications of 24 and 18 rats were conducted, involving three tests given 7 days apart. According to a counterbalanced design, each rat was tested after receiving no preload, an oral preload, or an intragastric preload. The preloads (6 ml of 500 mM NaCl) were presented according to the methods used during adaptation (see above), with the modification that, when given the oral preload, each animal was watched to ensure that it promptly drank it (they all did). Half the rats in each replication always received the preload 15 min before the test; the other half always received the preload 60 min before the test. The test consisted of free access to 500 mM NaCl solution and water (which was always available). The fluids were presented in 50-ml inverted, graduated centrifuge tubes with stoppers and drinking spouts. Intake of NaCl and water was recorded at 15, 30, 60, 120, and 240 min. At 240 min, the 50-ml tubes were replaced by larger bottles, so that a 24-h intake measurement could be made on the next day.

Experiment 3: physiological experiment. The design of experiment 3 involved comparing the physiological effects of ingested or intubated NaCl in sodium-deficient rats. Because we could find no normative data on gastric emptying of hypertonic NaCl in sodium-replete rats, the design also included a comparison of the effects of intubated NaCl in sodium-deficient or -replete rats. It was not possible to compare the physiological effects of ingested NaCl in deficient and replete rats, because replete rats do not willingly drink concentrated NaCl solutions.

Experiment 4: Influence of Sham Ingestion of NaCl on Plasma Aldosterone Concentrations

Eight of the rats used in experiment 1 were tested ~1 mo after they had completed experiment 1, at which time they weighed 453 ± 9 g. Each rat received three tests, given once a week, in counterbalanced order. At approximately noon on the day after sodium deficiency was induced (see above), the rats were removed from their cages, their gastric cannula caps were removed, and their stomachs were rinsed with warm water to remove particulate matter. They were then placed into the test boxes used in experiment 1 and allowed to drink 500 mM NaCl (i.e., their gastric cannula was closed to prevent drainage), sham drink 500 mM NaCl, or sit undisturbed for 15 min. At the end of the 15-min test, a 200-µl blood sample was collected from the tip of the tail into a chilled tube, the cannula was closed (if it was not already closed), and the rat was returned to its home cage with chow available. The blood was spun at 2,000 g for 3 min in a microcentrifuge and frozen for later analysis of aldosterone and sodium using the methods outlined above.

Statistical Analyses

Preliminary analyses of experiment 2 showed that the two replications produced very similar results, and so to simplify analyses, the data were combined. Results were analyzed using mixed-design two-way ANOVAs with independent variables of preload-test interval (15 or 60 min, between-subject) and preload type (none, oral, intragastric, within-subject). Data for each time period and for each fluid ingested (500 mM NaCl and water) were analyzed in separate analyses. Differences between the three preload conditions for each preload-test interval were determined using planned comparisons, and, when appropriate, post hoc t-tests were used to determine differences between individual pairs of means.

Experiment 3 involved three independent variables: route of preload administration (ingestion or intubation), physiological state (sodium-replete or sodium-deficient), and time (0, 15, 30, 60, and 120 min). However, the design was incomplete, because sodium-replete rats do not voluntarily ingest concentrated NaCl solutions. Thus it was not possible to test rats in the sodium-replete state with ingested NaCl. Consequently, the data were subjected to two separate analyses. To determine the influence of the route of preload administration, results from the group allowed to ingest NaCl when sodium deficient were compared with those from the group given NaCl by intubation when sodium deficient. Factors in the ANOVAs were route of administration and time. To determine the influence of physiological state, results from the group given NaCl by intubation when sodium deficient were compared with those from the group given NaCl when sodium replete (factors: sodium status and time). When the ANOVAs revealed significant interactions, the source of the interaction was determined using post hoc t-tests to compare pairs of means. Additional planned comparisons were made between the control group (rats that were undisturbed, i.e., sodium replete and not given NaCl) and each experimental group using t-tests. All analyses used a criterion for significance of P < 0.05. Values are means ± SE.

Data from several rats were not included in the analyses because of technical errors. Two rats were eliminated from the experiment: one because it did not drink the NaCl preload and the other because it was inadvertently intubated twice. A third rat, which was assigned to one of the sodium-replete groups, received furosemide, so it was reassigned to a sodium-deficient group. In addition, the stomach contents of two rats were spilled. Even with these problems, group sizes were always between 8 and 11.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiment 1: Effect of Sham Ingestion on Subsequent NaCl Intake

The type of solution sham ingested had no effect on subsequent intake of 500 mM NaCl solution (Fig. 1) or water intake [data not shown; for each measurement period F(3,45) < 2.2, P = not significant (NS)].

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Intake of 500 mM NaCl by sodium-deficient rats that were previously allowed to sham drink water, 1,000 mM sucrose, 150 mM NaCl, or 500 mM NaCl. Intakes during the sham-drinking test are given in Table 1. There were no significant differences among the "preload" conditions at any time.

Experiment 2: Effect of Ingestion or Intubation of NaCl on Subsequent NaCl Intake

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Intake of 500 mM NaCl by sodium-deficient rats given no preload, 6 ml of 500 mM NaCl to drink, or 6 ml of 500 mM NaCl by intubation. The preload was given 15 min (A) or 60 min (B) before the test.

Analysis of the 60-min cumulative intakes provides a representative example of these effects. With respect to NaCl intake, there was a significant interaction of preload route and preload-test interval [F(2,80) = 3.13, P < 0.05]. This was due to 1) significantly lower NaCl intakes after the ingested and the intubated preload in the 15-min delay group relative to all four other conditions and 2) significantly lower NaCl intakes after the oral preload in the 60-min delay group than after either no-preload treatment. There was also a significant overall effect of preload route [F(2,80) = 12.6, P < 0.0001] and a significant overall effect of preload-test interval [F(1,40) = 9.15, P < 0.005]. Water intake at 60 min was significantly higher in rats tested after the 15-min than in those tested after the 60-min preload-test delay [F(1,40) = 4.73, P < 0.05], but there were no other differences in water intake among the various conditions.

Experiment 3: Comparison of Physiological Effects of NaCl Ingestion and Intubation

Comparison of Sodium-Deficient and -Replete Rats Given NaCl by Intubation

Gastric sodium. Sodium status had a dramatic effect on gastric sodium contents [group × time interaction, F(3,69) = 3.25, P < 0.05]. At 15 min after intubation, there was no significant difference between the groups in gastric sodium content. However, at later times, the sodium-deficient groups had significantly less gastric sodium than did the sodium-replete groups. The difference was largest at 60 min after intubation, when the replete group had 43% of the preload remaining in the stomach whereas the deficient group had virtually none. The pattern of emptying appeared to differ between the groups. For the replete group, there was no evidence of emptying between 15 and 60 min. Gastric sodium content between these times did not differ. However, sodium content declined significantly between 60 and 120 min. For the deficient group, emptying appeared to occur much earlier (Fig. 3).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Sodium distribution in the stomach and small intestine of sodium-deficient and -replete rats at various times after they received 3 mmol (6 ml of 500 mM) of NaCl to drink or by intubation. Dashed line, value for undisturbed rats, which were not sodium deprived and did not receive NaCl to drink.

Intestinal and total gut sodium. There were small but significant differences in intestinal sodium content between sodium-replete and -deficient rats [group × time interaction, F(3,71) = 3.24, P < 0.05]. There was less intestinal sodium in sodium-deficient than -replete rats at 15 min after intubation. At other times there was no difference. For both groups, there was more sodium in the intestine at 15 and 30 min than at 60 or 120 min.

Plasma factors. PRA and aldosterone concentrations were much higher in sodium-deficient than -replete rats [main effect of group for PRA, F(1,71) = 172.8, P < 0.0001; main effect of group for aldosterone, F(1,71) = 41.8, P < 0.0001]. Intubation of NaCl significantly reduced PRA and aldosterone concentrations in replete and deficient conditions. Not surprisingly, given the large initial differences in PRA, the decrease was much greater in the deficient than in the replete condition [group × time interaction for PRA, F(3,71) = 12.2, P < 0.0001; group × time interaction for aldosterone, F(3,71) = 10.1, P < 0.0001; Fig. 4]. For PRA, there were significant differences between sodium-deficient and -replete rats at all times. The replete groups at 30, 60, and 120 min had significantly lower PRA levels than did the replete group at 15 min or the undisturbed control group. The deficient groups at 60 and 120 min had significantly lower PRA levels than did the deficient groups at 15 and 30 min. However, at all times, PRA levels remained significantly higher than those of undisturbed controls. For aldosterone, the replete groups at 60 and 120 min had significantly lower aldosterone concentrations than did the undisturbed control group or replete groups at 15 or 30 min. The deficient rats had significantly lower aldosterone concentrations at 60 and 120 min than at 15 and 30 min after intubation. Aldosterone levels of deficient rats were significantly higher than those of undisturbed controls at 15 and 30 min, similar to those at 60 min, and significantly lower at 120 min.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Plasma renin activity, aldosterone, and sodium of sodium-deficient and -replete rats at various times after they received 3 mmol (6 ml of 500 mM) of NaCl to drink or by intubation. Dashed line, value for undisturbed rats, which were not sodium deprived and did not receive NaCl to drink.

Comparison of NaCl Ingestion With Intubation in Sodium-Deficient Rats

Gastric and intestinal sodium. At 15 min after the preload was presented, there was significantly more sodium in the stomachs of rats given NaCl to drink than in those given NaCl by intubation. At later times there were no differences between the groups in stomach sodium content [group × time interaction, F(3,69) = 2.06, P = NS]. The group allowed to drink NaCl had changes in gastric sodium content over time that were similar to the intubated group (described above), with the exception that during the 15- to 30-min period, emptying was faster in the group that drank NaCl (53 vs. 22 µmol/min; Fig. 3).

Plasma measures. The source of the NaCl preload did not differentially affect any of the plasma measures recorded over time (all group × time interactions were not significant). The only difference between the treatment conditions was that plasma aldosterone concentrations were higher in the groups given NaCl by intubation than in those that ingested NaCl [F(1,72) = 8.29, P < 0.01]. The time between preload administration and collection of the blood sample affected PRA, [F(3,72) = 38.1, P < 0.0001] and aldosterone concentration [F(3,72) = 74.2, P < 0.0001] but not plasma sodium concentration.

Experiment 4: Influence of NaCl Ingestion and Sham Ingestion on Plasma Aldosterone Concentrations

Plasma aldosterone concentrations were significantly lower after rats drank or sham drank NaCl than after they were left undisturbed [F(2,14) = 8.58, P < 0.005; Fig. 5]. There was no difference in aldosterone concentrations between the sham drinking and drinking conditions. There was also no difference among the three conditions in plasma NaCl concentrations: 134 ± 3, 133 ± 3, and 135 ± 2 mmol/l for undisturbed, sham, and real, respectively.

View larger version (61K): [in this window] [in a new window]   Fig. 5.   Plasma aldosterone concentrations after a 15-min test in which sodium-deficient rats sat undisturbed (None), sham drank 500 mM NaCl (Sham), or ingested normally 500 mM NaCl (Drink). *P < 0.01 vs. None.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In confirmation of earlier studies (8, 24), intubation of 3 mmol of NaCl failed to reduce subsequent NaCl intake if the preload-test interval was 60 min. In contrast, intubation was almost as effective as ingestion in reducing NaCl intake if the preload-test interval was 15 min. Ingested NaCl reduced subsequent NaCl intake whether the interval between preload and test was 15 or 60 min, although it was significantly more effective after a short than a long preload-test interval. The finding that NaCl intubation was ineffective at reducing intake of NaCl given 90 min after the preload supports the interpretation that oral factors are required to determine NaCl intake. However, the finding that intubation was as effective as ingestion at reducing intake of NaCl given 15 min after the preload argues against an important contribution of oral factors. Also arguing against the contribution of oral factors to the satiation of NaCl intake was the finding that sham ingestion of NaCl had no effect on subsequent NaCl intake. Because there was a 15-min delay between the end of sham drinking and the start of the NaCl intake test, it remains possible that oral factors could influence intake over short periods, for example, within a drinking bout. However, they cannot account for the satiety seen after normal ingestion of NaCl, because this lasts 90 min (experiment 2) (18).

Some insight into the behavioral results can be gleaned from the comparison of the physiological effects of ingested and intubated NaCl, particularly the differences in early rates of gastric emptying. At 15 min after the rats were given NaCl, the intubated group had less sodium in their stomachs and more in their small intestines than did the group allowed to drink it. If, as is quite likely, gastric capacity limits NaCl intake of sodium-deficient rats (see Ref. 1 for discussion), subsequent intake would be expected to be higher in the condition with the emptiest stomachs, which is what we found. On the other hand, at 60 min after preload administration, virtually all sodium had emptied from the stomach, whether this was ingested or intubated. Thus there could not have been any differential influence of gastric capacity on drinking at this time. Moreover, at this time, most of the preload had been absorbed, so the contribution of postabsorptive mechanisms would be expected to dwarf any remaining putative oral contributions. This is consistent with our finding that route of preload delivery did not affect intake of NaCl given 60 min after the preloads.

To discuss gastric emptying, it is convenient to distinguish between effects during the first 15 min after intubation and those during the remaining 105 min of the observation period. The distinction between an initial rapid phase and a later restrained phase of emptying has been made for other nutrients (5) and appears to be appropriate here too. During the restrained phrase, there was no evidence that physiological sodium status influenced the rates of emptying or clearance. However, the onset of gastric emptying was accelerated in sodium-replete relative to sodium-deficient rats. Sodium clearance from the gut appeared to be unaffected by deficiency in either phase. However, these conclusions are based on comparisons of the effects of NaCl that was intubated, so it is possible that different patterns of emptying and absorption would be present if the NaCl was ingested.

Circulating aldosterone concentrations dropped more rapidly after ingestion than after intubation of NaCl. Sodium appetite is thought to be stimulated in part by the central action of aldosterone, such that higher peripheral aldosterone levels would be expected to contribute to greater NaCl intake (23). It is thus conceivable that the rats consumed more NaCl after the preload was intubated than after it was ingested because of higher circulating aldosterone levels. However, we consider this unlikely, because it would imply an immediate, nongenomic action for aldosterone on NaCl intake, and, in any case, plasma aldosterone concentrations can be dissociated from satiation of NaCl intake (18). Instead, we suspect that the same neural or physiological events that blunt aldosterone secretion also inhibit NaCl intake.

The findings that intubated NaCl was less effective than ingested NaCl in lowering plasma aldosterone levels and that sham ingestion of NaCl lowered plasma aldosterone concentrations to the same extent as did normal ingestion (experiment 4) argue that taste (or the act of drinking) may influence plasma aldosterone concentrations. It is tempting to speculate that these results reflect a cephalic-phase influence on aldosterone secretion. Nicolaidis (9) reported that oral stimulation with NaCl induced natriuresis in sodium-replete rats, but as far as we know, this is the first evidence for a cephalic-phase hormonal response related to mineral balance. Certainly, a decrease in circulating aldosterone is a useful anticipatory response to counteract the increase in plasma sodium that follows ingestion of hypertonic NaCl, just as the cephalic-phase release of insulin anticipates the increase in blood glucose after ingestion of sugar.

Taken together, the findings of these series of experiments lead to two possible conclusions: 1) NaCl satiation requires the combined action of oral and postingestive stimulation. A similar conclusion has been drawn for the control of sugar intake (4). The alternative is that 2) normal ingestion is not adequately replicated by the combination of oral stimulation and intubation. Our demonstration of marked physiological differences in the effects of ingestion and intubation of NaCl argues that intubation does not mimic the effects of normal ingestion. We suspect that differences in the intestinal or postabsorptive delivery of sodium are responsible for the route-dependent behavioral effects of sodium administration. These factors were not measured in earlier studies (7, 17, 24), and the failure to consider them may have led to the incorrect conclusion that oral metering controls NaCl intake.

Perspectives

Previous work has emphasized the involvement of oral factors in the satiation of sodium appetite (2, 3, 7, 8, 17, 24). Our report suggests this may be an overemphasis, because differences in behavior observed after ingestion or intubation of NaCl that others have attributed to oral factors could be explained by differences in the gastrointestinal handling of sodium. This is not to say that the mouth is unimportant for the expression of sodium appetite. Clearly, it directs consumption to the appropriate, salty target. Our work also raises the possibility that sodium receptors in the mouth initiate hormonal responses that prepare the body to deal with the incoming sodium load. More research will be required to determine the significance of these preparatory responses for the control of NaCl intake.