INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

A PREVIOUS STUDY (9) in baboons (Papio hamadryas sensu lato) found that naive male baboons exhibited little "need-free" intake of a 300 mM NaCl solution but developed a salt appetite with Na deficiency. The time delay before salt intake and for correction of the calculated Na deficit decreased with repeated experiences. Previous to that study, experiments in adrenalectomized and normal rhesus monkeys had found little or no evidence of salt appetite in Na deficiency (14, 16, 20). Baboons were reported to prefer the lowest dietary salt level and not to acquire a taste for salt (2). On the other hand, there is evidence that other primates in the wild, e.g., chimpanzees (13) and gorillas (18), are attracted to natural licks and other salt sources in equatorial regions where the Na content of soils and plants is low.

Against this background we decided to test for the involvement of ANG II in the salt intake of Na-deficient baboons and in the water intake of baboons that were made thirsty by a period of water restriction. The intracerebroventricular infusion of ANG II elicits water drinking in all mammalian species that have been tested (11) but has inconsistent effects on the intake of NaCl solutions. For example, intracerebroventricular infusions of ANG II at rates that cause very large water intakes may have no effect on NaCl intake in sheep, unless the ANG II causes a natriuresis (25, 26), or in cattle (3), whereas both appetites are stimulated in mice (8) and rats (1, 12).

The effects of intracerebroventricular infusions of ANG II and losartan (a specific AT₁-receptor antagonist) were tested first in normal baboons. Next, losartan was infused intracerebroventricularly at the maximal possible dosage (given solubility limitations) in 1) baboons made Na deficient by intramuscular injections of furosemide for 3 days and 2) baboons made water deficient by reducing their daily water supply to 25% of their normal daily intake for 3 days. Similar experiments with intracerebroventricular infusion of ZD-7155 (another AT₁-receptor antagonist), instead of losartan, then followed. The effect of oral captopril, an angiotensin-converting enzyme inhibitor, was also tested in Na-deficient baboons.

The results of these studies have been reported briefly (19).

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and Maintenance

All procedures and protocols were approved by the Southwest Foundation Institutional Animal Care and Use Committee.

Surgical Procedure for Lateral Ventricular Cannulation

On the day of intracerebroventricular cannula implantation, each baboon was sedated with ketamine (10 mg/kg im) with added atropine (0.54 mg im). Anesthesia was induced and maintained with isoflurane (1-2%). The baboon was placed in the primate chair, and the head was placed in the stereotaxic frame. The cranium was exposed through a midline incision, and reflection of underlying muscle and periosteum was exposed by blunt dissection. A 5-mm-diameter hole was drilled through the cranium at 19 mm anterior to the ear bars and 1.0-1.5 mm left of midline. The z-coordinate location of the dura was established by lowering the tip of the cannula until it touched the dura. A 5-ml/h infusion of sterile 0.9% NaCl through the cannula was started, and the increase in hydrostatic pressure within the cannula, using a pressure transducer, was recorded on a strip chart recorder, as the tip of the cannula was lowered through the brain tissue. The location of the lateral cerebral ventricle was established by observing the z-coordinate below the surface of the dura at which hydrostatic pressure ceased to rise but fell sharply toward zero. This location within the lateral ventricle was confirmed by raising and lowering the tip of the cannula and observing increases in pressure when the tip encountered brain tissue and decreases when the cannula tip was returned to the ventricular cavity. Cyclic variation in cerebral spinal fluid (CSF) pressure associated with respiration and a large increase in pressure with a Valsalva maneuver were additional observations used to confirm that the cannula tip was located in a lateral ventricle. Typically, this occurred at 20-25 mm below the surface of the dura.

The cannula was fixed at this location by placing dental acrylic in and around the hole in the cranium, bending the cannula to a 90° angle so that the external portion of the cannula was parallel to the surface of the skull, placing one or two self-tapping stainless steel screws in the cranium next to the cannula and adding sufficient additional dental acrylic to cover completely the screws and the cannula. A 20-gauge polyvinyl tubing (Tygon Microbore tubing; Norton Plastics) was attached to the external end of the cannula and routed subcutaneously to a subcutaneous pocket placed in the midscapular region of the back. The end of the tubing was attached to an osmotic pump filled with sterile 0.9% NaCl that was placed in the subcutaneous pocket. The cranial incision was closed with 3.0 polyglactin suture (Vicryl; Ethicon) so that skin completely covered the cannula and small block of dental acrylic.

The baboons were then returned to their home cage and permitted to recover for a least 2 wk before collection of control observations for the initial experiments. Postoperatively, each baboon received amoxicillin, 500 mg orally, for 7 days as a prophylactic treatment. In addition, an analgesic, buprenorphine hydrochloride, 0.01 mg/kg im every 12 h (Buprenex; Norwich Eaton Pharmaceuticals), was administered for at least the first three postoperative days.

Surgical Procedure for Changing Alzet Pumps

Osmotic pumps were changed when required using standard aseptic surgical techniques. Baboons were sedated with ketamine (10 mg/kg im) with added atropine (0.54 mg im). Anesthesia was induced and maintained with isoflurane (1-2%). A 1.5- to 3.0-cm incision was made over the subcutaneous pocket containing the osmotic pump, and the pump was exposed by blunt dissection. The osmotic pump was then removed from the subcutaneous pocket and detached from the catheter. The catheter-cannula system was flushed with 2 ml NaCl followed by injection of 5 µg ANG II in 2 ml NaCl to verify that the catheter-cannula system was still accessing CSF. This was indicated by observing an increase in blood pressure and heart rate after the intracerebroventricular ANG injection. A new osmotic pump was then attached to the catheter and placed in the subcutaneous pocket, and the incision was closed with 3-0 polyglactin suture.

This procedure was completed in ~30 min, and the animal was returned to its cage. Daily measurements were made at 1200-1300 as usual, and fresh food and fluids were presented. Occasionally baboons did not eat and drink normal amounts on this first postoperative day.

Agents Used

ANG II (human octapeptide; Bachem California, Torrance, CA) was infused intracerebroventricularly at 1, 3, and 5 µg/h in early experiments and at 5 µg/h routinely. These rates of infusion are similar to the doses that are effective in other large animals, e.g., sheep (26) and cattle (3).

Losartan potassium (kindly donated by Dupont-Merck) was infused intracerebroventricularly at 300 µg/h, which was the highest dose compatible with the solubility of losartan in aCSF at room temperature [i.e., 300 µg/10 µl (rate of pump delivery) = 300 × 200 µg/2 ml (volume of 2ML1 pump) = 60 mg/pump].

ZD-7155 (kindly donated by Zeneca) was infused intracerebroventricularly at 50 µg/h based on effective rates of infusion in experiments in sheep (23). It was prepared as 50 µg/10 µl (for 2ML1 pumps) by solution in a minimum volume of ethanol (~1 mg in 50 µl) and then dilution in aCSF.

Furosemide (Lasix; Hoechst) was injected intramuscularly two times per day at 1 mg/kg for 3 days in the Na depletion experiments.

Captopril (kindly donated by Squibb) was given orally as a 1-g dose at 18 and 3 h before presentation of 300 mM NaCl solution in the Na depletion experiments.

Experimental Protocols

Experiments were conducted in the following order, except that the captopril experiments were completed before the ANG II/ANG II receptor antagonist experiments began.

ANG II (n = 5). After a control period of 7 days, the Alzet pump containing aCSF was changed at ~0900 to a pump containing ANG II. After 7 days of observation, this pump was changed at ~0900 to a 2ML2 pump containing aCSF. Observations were continued for 7 recovery days.

These experiments were repeated (n = 5) with one alteration, that the NaCl solution was not provided to the animals.

Losartan (n = 5). The procedure was the same as for ANG II except that the 2ML1 pump contained losartan (300 µg/ 10 µl).

ANG II plus losartan (n = 3). The procedure was again as for ANG II except that the 2ML1 pump contained ANG II (5 µg/10 µl) plus losartan (300 µg/10 µl).

Water restriction (n = 6). Throughout these experiments the baboons did not have access to 300 mM NaCl. After a control period of 7 days, for 3 days daily water intake was restricted to ~25% of the normal daily intake calculated for each baboon. On the 4th day, after the usual measurements, the baboons were given free access to water, and the intake was measured at 30 min, 1 h, 2 h, 4 h, and 24 h. Water intake then resumed on an ad libitum basis.

This experiment was repeated at least 2 wk later, with concurrent intracerebroventricular infusion of losartan. A 2ML1 pump containing 300 µg/10 µl losartan was put in on the morning of the 1st day of water restriction and changed back to an aCSF pump on the day after rehydration.

ANG II and ANG II plus ZD-7155 (n = 6). In view of the observation that losartan at 300 µg/h did not abolish NaCl and water intake in response to intracerebroventricular ANG II at 5 µg/h, the potency of ZD-7155 was tested by intracerebroventricular infusion of ANG II at 5 µg/h and later concurrent infusion of ANG II at 5 µg/h plus ZD-7155 at 50 µg/h. The protocols for these experiments were similar to the protocols used in ANG II and ANG II plus losartan, which tested losartan.

Na depletion 1 (effect of losartan, n = 6). After a control period of 7 days, the baboons were given two times daily for 3 days an intramuscular injection of furosemide (1 mg/kg). They did not have access to 300 mM NaCl during this period. Cumulative net Na loss during the 3 days of furosemide injection was calculated from daily Na balances, which included the Na intake in food. On the 4th day, after the usual measurements of food and water intake, the baboons were given free access to 300 mM NaCl solution, and the intake was measured at 30 min, 1 h, 2 h, 4 h, and 24 h. NaCl intake then resumed on an ad libitum basis.

This experiment was repeated at least 2 wk later with intracerebroventricular infusion of losartan. A 2ML1 pump containing 300 µg/10 µl losartan was put in on the morning of the 1st day of Na depletion and changed back to aCSF on the day after the animals voluntarily corrected their Na deficits.

Na depletion 2 (effect of ZD-7155, n = 6). The protocol was the same as in Na depletion 1, except that the repeat experiment was done using intracerebroventricular infusion of ZD-7155 at 50 µg/h during the period of Na depletion and repletion instead of intracerebroventricular infusion of losartan.

Na depletion 3 (effect of captopril, n = 8). These experiments were performed on baboons that had not been used for the other experiments. The protocol was the same as in Na depletion 1 except that in the repeat experiment the baboons were given 1 g captopril orally 18 h and again 3 h before the access to 300 mM NaCl solution. Sham experiments, 4 wk later, involved intramuscular injection of normal saline for 3 days instead of furosemide.

Statistical Analysis

	RESULTS

Top Abstract Introduction Methods Results Discussion References

ANG II

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Effect of icv infusion of ANG II for 7 days at 3 µg/h (1 baboon) or 5 µg/h (4 baboons) on daily intakes of 300 mM NaCl and water. Relationship between daily total Na intake (food and NaCl solution; circles) and urinary Na output (triangles) is also shown in bottom. Values are means ± SE. * P < 0.05 comparing ANG II with baseline.

In separate experiments, baboons that had only water available for drinking fluid during intracerebroventricular infusion of ANG II infusion (n = 6) increased mean water intake from 2,101 ± 650 ml/day to 3,893 ± 621 ml (P < 0.05) on the 1st day of infusion. The mean daily intake for the group remained at a high level for the next 6 days of infusion, but only the mean intake on the 5th day was statistically significant (P < 0.05).

Losartan

ANG II Plus Losartan

Water Restriction

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Cumulative water intakes during 24 h by 6 baboons on 1) a baseline day, 2) a day that followed 3 days of restriction of water intake to 25% of the baseline intake, and 3) as in 2 except that losartan was infused intracerebroventricularly at 300 µg/h from the 1st day of water restriction to the day after free access to water. Values are means ± SE. * P < 0.05.

The recovery of daily water intake to the initial value was delayed after the losartan infusion. In the control experiment the mean daily water intake was back to normal on the day after rehydration, but in the losartan experiment recovery was not complete until 3 days after the day of free access to water. The daily food intake was significantly reduced on the 3rd day of water deprivation in both the control and the losartan experiment. It was also reduced during the 3 days of recovery after losartan when daily water intake also was low.

ANG II and ANG II plus ZD-7155

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Infusions (icv) of ANG II for 7 days at 5 µg/h alone (circles) or with concurrent icv infusion of ZD-7155 at 50 µg/h (squares) in 6 baboons. Effects on daily intakes of 300 mM NaCl and water. Values are mean ± SE. * P < 0.05 comparing the 2 infusions.

The concurrent infusion of ZD-7155 abolished the responses to ANG II so that NaCl and water intakes for 7 days were not different from the baseline values (Fig. 3). Food intakes during ZD-7155 infusion were not reduced relative to baseline.

Na Depletion 1 (Effect of Losartan)

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Results of 3 sets of Na depletion experiments. Left: cumulative intakes during 24 h of 300 mM NaCl solution by 6 baboons after 3 days without access to that solution, in the following 3 situations: situation 1) no treatments (squares), situation 2) a day that followed 3 days of intramuscular injections of furosemide (1 mg/kg, 2 times daily; circles), and situation 3) as in situation 2 except that losartan was infused icv at 300 µg/h from the 1st day of furosemide to the day after free access to 300 mM NaCl (triangles). Values are means ± SE. Middle: same design as on left except that in situation 3 ZD-7155 was infused icv at 50 µg/h instead of losartan in 6 baboons. * P < 0.05 comparing situation 2 with situation 3. Right: same design as in left and middle except that in situation 1 physiological saline was injected intramuscularly daily instead of no treatments, and in situation 3 1 g of captopril was given orally 18 and 3 h before access to NaC1 in 8 baboons. * P < 0.05 comparing situation 2 with situation 3 and situation 3 with situation 1.

Concurrent infusion of losartan during furosemide treatment (cumulative net Na loss 103 ± 15 mmol) did not significantly reduce the intakes of NaCl solution at any time during the 24-h period of access to NaCl solution (Fig. 4), but the intake of water during access to NaCl was significantly less during the 1st h in the losartan experiments. Food intake was unaltered by losartan infusion during furosemide treatment or subsequent access to NaCl. Daily water intake was unaltered by furosemide treatment alone (mean prefurosemide 2,491 ± 359 ml, mean furosemide 2,400 ± 69 ml) and was unaltered through the furosemide treatment (day 1 2,285 ± 330 ml; day 3 2,522 ± 333 ml). Water intake did fall from 2,745 ± 485 to 1,826 ± 68 ml from day 1 to day 3 of furosemide treatment when losartan was concurrently infused.

Na Depletion 2 (Effect of ZD-7155)

Daily water intake was unaltered during 3 days of furosemide injections alone (day 1 3,188 ± 428 ml; day 3 2,944 ± 540 ml) but fell from 3,017 ± 765 ml on the 1st day to 1,959 ± 289 ml on the 3rd day of furosemide when ZD-7155 was concurrently infused. Food intake during these experiments was not affected by furosemide treatment alone or throughout the period of furosemide plus ZD-7155 infusion.

Na Depletion 3 (Effect of Captopril)

Administration of captopril on the day before and the day of access to NaCl solution (cumulative net Na loss 114 ± 8 mmol) significantly reduced the intake of NaCl throughout the first 4 h of access but did not significantly reduce the total intake: 224 ± 24 mmol over 24 h (Fig. 4, right).

Daily water and food intakes were not altered by the 3 days of furosemide treatment alone or in the experiments with addition of captopril on the 3rd day.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The major findings of these experiments in baboons were that the infusion of ANG II into the brain ventricles stimulated both water intake and NaCl intake, that the infusion of the AT₁-receptor antagonist losartan by the same route inhibited the water intake after water deprivation, and that the infusion of ZD-7155 inhibited the NaCl intake after Na depletion. The peripheral administration of captopril also reduced the NaCl intake after Na depletion.

Evidence of consistent and substantial positive Na balance on day 1, continuing into day 2, of the ANG II infusion experiments indicates that the large increase in Na intake on those days was not secondary to urinary loss of Na caused by possible cardiovascular or other central actions of ANG II. The observation that the increased water intake in response to intracerebroventricular infusion of ANG II still occurred, but to a lesser extent, when 300 mM NaCl solution was not consumed is consistent with evidence that intracerebroventricular infusion of ANG II is dipsogenic in all species tested so far (11). Water intake in response to bolus intracerebroventricular injections of 5 µg of ANG II has been shown previously in baboons (15), but the water intake diminished with repeated injections of ANG II. Such diminution was not observed in the present experiments. In other primates, intracerebral injection of 10¹⁰ mol of ANG II induced water drinking in rhesus monkeys (21), and subcutaneous injection of ANG II at 50 µg/kg elevated water intake in African green monkeys (27).

The Na deficits resulting from 3 days of furosemide injections were not altered by the infusions of losartan or ZD-7155 (or captopril intake) during the Na depletion period. However, these infusions of each antagonist did reduce the daily water intakes on the 3rd day of depletion, suggesting that the water intake in these circumstances too was at least partly due to brain ANG II. However, losartan was found not to reduce daily water intake in nondepleted baboons.

It is likely that the high intakes of NaCl solution during ANG II infusion were contributory to the water intakes by sustaining or even increasing plasma Na concentration. Unfortunately, changes in plasma composition could not be measured during these experiments without a tether arrangement or anesthetizing the animals. There are few data on long-term effects of intracerebroventricular ANG II on water intake, but such an infusion did increase daily water intake during 4 days in mice on a low-Na diet and no access to NaCl solution (8). However, this increase appeared slower than in experiments in which NaCl solution was also available.

The finding that intracerebroventricular infusion of ANG II stimulates both water intake and NaCl intake is not unique in that similar results have been obtained in mice (8) and rats (11). However, the finding is not in accord with the results of some of the studies in larger mammals, e.g., sheep (25, 26) and cows (3). The intracerebroventricular infusion of ANG II for several hours in these larger animals stimulated water intake only, when Na depletion was averted, whereas intravenous infusions of ANG II affected Na appetite only, this pattern building toward an argument that ANG II or AT₁-receptor antagonists in the brain ventricles accessed brain pathways subserving thirst while ANG II in blood accessed brain pathways subserving Na appetite (4, 5, 23). This dichotomy was supported by the potent inhibitory effects of converting-enzyme inhibitors (CEIs) administered intravenously on Na appetite in sheep (25) and cattle (3) and the absence of any such effect from intracerebroventricular administration of CEIs.

These findings with exogenous ANG II reveal what ANG II can do in a primate brain, but they cannot reveal what ANG II does in physiological situations. However, the findings with the AT₁ antagonists losartan and ZD-7155 revealed that inhibition of ANG II action reduced thirst or Na appetite in situations of physiological water deficiency or physiological Na deficiency, respectively, pointing to a physiological role.

It is intriguing that the intracerebroventricular dose of losartan that was insufficient to block the actions of exogenous ANG II, and insufficient to reduce NaCl intake after furosemide treatment, was sufficient to antagonize the action of ANG II on brain sites concerned with water intake in water-deprived baboons. Possibly, the level of ANG II in the brain caused by water restriction was less than the levels caused by intracerebroventricular infusion of ANG II or by Na depletion. It was not possible to increase the dose of losartan (due to considerations of solubility in the small volumes in the miniosmotic pumps), but it was possible to switch to the more potent AT₁-receptor antagonist ZD-7155 (23). This antagonist then abolished the actions of exogenous ANG II on water and NaCl intake and the NaCl intake in response to Na depletion caused by furosemide.

It may not have been only greater receptor antagonist potency that accounted for the positive response to ZD-7155 compared with the failure of response to losartan. Looking for reasons for this difference, it may have been that ZD-7155 had greater penetrability in brain tissue than losartan. Also, the two AT₁-receptor antagonists may not be equally effective in inhibiting the "a" and "b" variants of the AT₁ subtype. There are no data known to the present authors on these aspects of action of these two agents. Aside from unresolved questions of potency, penetration, and specificity of the AT₁ antagonists, the findings strongly indicate that brain ANG II appears to be involved in both thirst and Na appetite in baboons.

Further work needs to be done with this preparation to compare the effects of intravenous administration of AT₁-receptor antagonists and to test the effects of intracerebroventricular infusion of AT₂-receptor antagonists. At the present stage, our results from experiments with orally administered captopril indicate that reduction of peripheral blood ANG II concentration reduced the NaCl intake of Na-depleted baboons. This is consistent with similar findings in several other species, rats, sheep, cows, and mice (3, 10, 17, 22, 24, 25). It cannot be excluded that captopril at the doses used in all of these experiments could have crossed the blood-brain barrier to inhibit the formation of brain ANG II. However, in sheep (6, 23) and cows (3), high doses of captopril and other CEIs infused directly into the brain ventricles did not inhibit water intake in water-deprived animals or NaCl intake in Na-depleted animals. This suggests that endogenous ANG II formation in the brain was not influenced by inhibition of the converting enzyme step in brain extracellular fluid, so that the ability of captopril to cross the blood-brain barrier may be irrelevant.

Finally, the AT₁-receptor antagonists used in these experiments were observed to be almost free of other effects on behavior in experiments lasting several days. Losartan infused alone was observed to reduce daily food intake in some baboons, but when it was used in experiments with ANG II infusion or water or Na deficiency no substantial effects were observed. The ZD-7155 infusion alone or in combination with other treatments did not alter daily food intake.

Perspectives