Water intake induced in rats by serotonin and 5-hydroxytryptophan: Different mechanisms?

0361-9230/87$3.00 f 40

Bruin R~sea~h BuNetin, Vol. 18, pp~501-508. 9 Pergamon Journals Ltd., 1987.Printed in the U.S.A

Water Intake Induced in Rats by S~~~t~nin and 5-Hyd~~~yt~ypt~phan: Different Mechanisms?’ NEIL E. RU~LAND,~ !3LORENCE A. CAWTU AND MELVIN f, FREGLY Departments of Psychology and Physiology, and Center for the Neurobialogical Sciences Colleges crf Liberal Arts and Sciences and Medicine, University of Florida, Gainesville, FL 3261 I

Received 14 May 1986 AND M. I. FREGLY. Water intake induced in rats by srrutunin and 7 BRAIN RES BULL 18(4) 501-508, 1987.-The present studies were designed to investigate further the mechanism by which water intake is induced in rats by peripheral administration of either serotonin (5HT) or its precursor 5-hydroxytryptophan (SHTP). Consistent with previous studies that have implicated mediation by the renal renin-angiotensin system (RAS), we now report that bitaterrd nephrectomy completely abolishes the drinking response to various doses of SHT. In contrast, nephrectomy had little effect on the drinking induced by 5HTP. Thus, SHTP may induce drinking by mechanisms other than its peripheral conversion to 5HT and subsequent activation of the RAS. The drinking responses to both SHT and 5HTP were blocked by peripheral administration of the SHT receptor antagonist, metergolioe, but the drug was at least ten-fold more potent against SHTP than 5HT. Intracerebroventricular (ICV) administration of metergoline also prevented the drinking response to peripherally-administered SHTP, The drinking responses to both 5HT and 5HTP were enhanced by peripheral admi~is~ation of low doses of an an~otensin I converting enzyme inhibitor, captoprit. Collectiveiy, these findings support previous conchrsions that SHT-induced intake of water is mediated excfusively by the renal RAS. However, SHTP-induced drinking may additionally involve a renin-independent, serotonin-media~d mechanism, possibly in the brain. ROWLAND,

N. E., F. A. CAPUTU

S-hydroxytryptophan:

Water intake Renin-angiotensin

Diffrretzt

Serotonin

mechanisms,

~-Hydroxyt~ptoph~

studies from this and other laboratories have shown that peripheral administration of either serotonin (5HT) or its precursor, 5hydroxytryptophan (SHTP), induces drinking in rats [l, 4, 6, 13, 14, 17, 231. This drinking response is dose-related and typic&y occurs within 30-60 tin of subcutaneous (SC) injection of either agent. Prior ad~nist~tion of decarboxylase inhibitors, which prevent the production of 5HT from SHTP, blocked the dipsogenic action of SHTP but not of 5HT, suggesting that the formation of SHT is crucial [14]. Moreover, the SHT receptor antagonist, methysergide, prevented the water intake to both agents [14]. This suggests that SHTP must be converted by decarboxylation to 5HT which then interacts with its receptor to induce the drinking response. However, another 5HT antagonist, cinanserin, had no inhibitory effect [ 19. It is likely that the drinking response induced by both SHTP and 5HT is mediated by the peripheral reninangiotensin system (RAS). Both agents promote renin reRECENT

Nephrectomy

Captopril

Metergoline

lease [I], and the drinking response to each is inhibited by farge doses of captopril, an angiotensin I converting enzyme inhibitor [13]. These fmdings suggest that the activation of 5HT receptors results in a release of renin from the kidneys by mechanism(s) that are not clearly understood at present, but may involve the release of catecholamines from the adrenal medulla and their subsequent stimuiation of release of renal renin. We have often noticed that 5HTP-induced drinking occurs with shorter latency than that induced by 5HT, an observation which is surprising if SHTP must be converted to 5HT for its action [6,14]. Additionally, a 5-lo-fold greater molar dose of SHTP is required compared to 5HT for an equivalent dipsogenic response [6,13]. Further, at doses with equivalent dipsogenic potency, ad~~is~tion of 5HT resulted in a reduction of arterial blood pressure and body temperature while that of SHTP did not [l]. These observations raise questions concerning the similarity of the mech-

‘Portions of thesedata were presented at the 1985 meeting of the Society for Neuroscience [ZO).This research was supported by NIH grant HLl4526. We thank S. Hoffman and G. Smith for technical assistance. 2Requests for reprints should be addressed to Dr. Neil Rowland, Department of Psychology, University of Florida, Gainesville, FL 32611.

ROWLAND.

C’APUTO AND FREGLY

TABLE. 1 WATER INTAKE OF NEPHRECTOMIZED AND SHAM-OPERATED RATS AFTER SC‘ INJECTION OF 5-HYDROXYTRYFTOPHAN (SHTP) OR SEROTONIN (5HT)

Water Intake (ml) Treatment

Mean Body Weight (g)

Study A (males) Sham-Vehicle Sham-SHTP (98 PmoUkg) Sham-SHT <10 Fmol/kg) Nephrectomy-Vehicle Nephrectomy-5HTP Nephrectomy-SHT

533 459 500 52.5

Study B (females) Sham-Vehicle Sham-SHTP (98 pmollkg) Nephrectomy-Vehicle Nephrectomy-5HTP Study C (males) Vehicle: Before nephrectomy After nephrectomy SHT i l-3 FmoUkg) Before nephrectomy After nephrectomy 5HT (4 FmoYkg) Before nephrectomy After nephrectomy 5HT (12 FmoVkg) Before nephrectomy After nephrectomy

I hr

0.1 3.8 3.5 0.9 3.7

* tlr t +

329 312 329 308

0.8 6.6 1.6 5.8

+ ti .k

466 456

0.7 -t 0.7

020 0.7 t 0.7

491 481

0.1 i 0.1 0.2 2 0.2

1.0 -ir 0.6 0.6 I 0.4

469 459

1.5 i- 0.6’F 0.1 i- 0.1

1.6 2 0.6% 0.2 t 0.2

so6 495

0.5 i 0.3

497 522

0.1 0.3* 1.0* 0.4 l.ZP I. I i 0.4

2 hr

0.7 1.6* 0.7 1.3*

O-t0

5.5

k

1.0’

0.4 7.4 7.1 2.1 4.3 1.5

-t 0.2 i 1.2* 2 1.5” + 0.4 i 1.1s i_ 0.5

1.0 7.5 2.1 6.0

-+ 1.0 i 1.8* + 0.6 i- 1.5*

10.0 5 1.7* 0.8 + 0.3

Shown are means t SE for Ns of 4-6. Group comparisons: *p
anisms for the induction of water intake by these two compounds, and have provoked a new series of studies to bear on this issue.

EXPERIMENT I: EFFECT OF ACUTE BILATE~L NEPHRECTOMY ON SHTP- AND SHT-INDUCED WATER INTAKE One of the ways to demonstrate that a renal factor is critically involved in a regulatory loop is to examine whether removal of the kidneys abolishes the regulatory responses. Thus, nephrectomy has proven to be an impo~nt method in evaluation of renal involvement in angiotensin-related drinking 141although the interpretation of the results is not always simple [ll]. In the present experiment, we investigate the effect of bilateral nephrectomy on the drinking responses to SHTP and 5HT in rats. Three studies (A-C) will be described. In Study A, both 5HT and SHTP were administered to nephrectomized and shy-operated male rats. In order to confirm and extend the unexpected results of Study A, the effect of SHTP in female rats was examined in Study B, and the effect of different doses of 5HT in male rats was examined in Study C.

METHOD

Study A: SHTP and 5NT in Male Rats Male Sprague-Dawley rats were purchased from ZivicMiller Laboratories (Allison Park, PA). These rats had been used in other drinking experiments, and at the time of this study they were about 5 months old and weighed between 450 and 550 g. The rats were housed and tested in individual stainless steel mesh cages. Purina Laboratory Chow (No. 5001) pellets and tap water were available at all times, except during tests when the food was removed. The vivarium was climate-con~olled with lights on from ~7~1~, and a temperature of 23°C. Surgery. Bilateral nephrectomy was performed on 18 ether-anesthetized rats through dorsal flank incisions about 2 cm long. The kidneys were grasped between thumb and forefinger, the renal pedicle ligated, and the kidneys removed. The porcedure, which takes about 4 min per rat, was finished by a suture in the abdomin~ wall and wound clips in the skin. Fifteen rats received a sham-operation which consisted of anesthesia, flank incisions and closing. All surgery was performed between 0800 and 1000, after which the rats were returned to their home cages without food or water.

SEROTONIN

503

AND DRINKING

Testing. Starting at noon, i.e., an average of 3 hr after surgery, the rats were administered either SHTP (25 mg/kg, or 98 kmol/kg, of the HCl salt; Calbiochem), 5HT (4 mg/kg, or 10 pmol/kg, of the creatinine sulfate salt; Sigma), or the 0.9% saline vehicle, all given SC in a volume of 1 ml/kg body weight. Tap water which had been equilibrated to room temperature, was presented immediately in calibrated tubes (20.1 ml) with metal sipper spouts. The volume consumed was recorded 30, 60 and 120 min later. Because drinking often starts around 30 min after these injections, the 30 min intakes show large individual variability and will not be presented. All nephrectomized rats were humanely sacrificed immediately after the test with an overdose of sodium pentobarbital. Study B: SHTP In Female Rats

Nineteen adult female Sprague-Dawley rats (272-362 g) were used in this study. Ten were nephrectomized and nine were sham-operated, as described in Study A. SHTP (25 mg/kg, SC) was then administered, water intake was recorded for 2 hr, and the rats were then killed, exactly as described in Study A. Study C: Dijjferent Doses

of SHT

Twenty-four male Sprague-Dawley rats (342-530 g) were used in this study. They were divided into four groups of 6, and each animal served as its own control on two test days. On the first test day, before any surgical procedures, the intact rats were injected with either 0, 0.5, 1.5 or 4.5 mg SHT/kg (0, 1.3, 4, or 12 pmolikg) and the drinking response was measured. Two days later, all of the rats were nephrectomized and 3 hr later were injected with the same dose of 5HT that they had received on the first day. Intakes were recorded as before. Statistics

The data were subjected to analysis of variance with main factors of surgery, drug treatment and time. Means were compared using Newman-Keuls tests (significance p
Study A

The results are shown in Table 1. At the doses used, SHTP and 5HT produced similar increases in water intake of sham-operated rats. Nephrectomized rats also drank substantial amounts after administration of SHTP, but not after 5HT. Thus, the ANOVA revealed significant main effects of surgery, F(1,28)=5.37, pcO.05, drug treatment, F(2,28)=13.21,p<0.001, and time, F(1,28)=38.24,p<0.001, two-way interactions between drug and F(2,28)=6.08, p
In the female drinking in both (Table 1). Thus, effect of surgery

rats, SHTP induced nephrectomized and after both 1 and 2 was not significant,

equivalent amounts of sham-operated groups hr, the ANOVA main F(1,15)<0.1, the main

effect of drug was significant, F(1,15)>14.8, p
On the first test day (all rats intact) water intake was stimulated above basal (0 mg/kg) by the highest dose of SHT; the dose of 1.5 mg/kg had a marginal effect (0.05
The drinking response in the first hr after administration of SHTP was unaffected by nephrectomy in both male (Study A) and female (Study B) rats. The drinking response to SHTP in the second hr was reduced in nephrectomized males, but not females. The data from the sham-operated males are unusual insofar as the drinking response to SHTP normally is complete within the first hr [6,14]. Nonetheless, it is clear that nephrectomy does not abolish the drinking response to SHTP. These findings are clearly inconsistent with previous pharmacological studies [6, 13, 141 which have suggested that the release of renal renin is the exclusive mechanism for dipsogenesis of this agent. In contrast to these findings with SHTP, the drinking response to a high dose of 5HT was completely abolished by nephrectomy (Studies A and C). This dose of 5HT causes a considerable hypotension [I] which, although not measured in this study, would be expected to be more profound in nephrectomized rats with impaired counterregulation, and so might be debilitating to them (indeed, many adopted prone postures for up to 1 hr after SHT). A similar line of reasoning was used in a demonstration that nephrectomized rats can show a drinking response to isoproterenol [ 111. Thus, Study C was designed to test lower doses of 5HT in order to see whether nephrectomized rats have a lower dose-range of responsiveness to this agent. They did not: lower doses of 5HT which were ineffective in intact rats were also without dipsogenic action after nephrectomy (Study C). These findings are consistent with previous pharmacological studies [13,14] which suggested that release of renal renin is essential for the dipsogenic action of 5HT. How can these conflicting findings with SHTP and 5HT be reconciled? It is possible that SHTP, administered to nephrectomized rats, may engage mechanisms not normally relevant in intact rats. Thus, while the present dose of SHTP has only a minimal hypotensive action in intact rats [1,5], it may be anticipated that the hypotension would be more pronounced in animals with a deficient counterregulatory system. However, it might then have been anticipated that lower doses of 5HT should have produced drinking, which was not the case. An alternative hypothesis is that SHTP, which can cross the blood-brain barrier and be decarboxylated to 5HT, engages drinking via a brain 5HT mechanism in nephrectomized rats. However, peripheral decarboxylase inhibition is sufficient to block SHTP-induced drinking in intact rats

ROWLAND,

504

TABLE EFFECT OF CAPTOPRIL (CAP; 3 OR WATER INTAKE INDUCED IN

CAPUTO

AND FREGL\’

2

30 m&g ADMINISTERED 5 OR 30 MIN BEFOREHAND) RATS BY EITHER 5-HYDROXYTRYPTOPHAN (5HTP) OR SEROTONIN (~HT)

ON

Water Intake (ml) Treatment

Mean Body Weight (g)

1 hr

2 hr

Study A (SHTP, 91 pmol/kg) Vehicle only (no SHTP) Vehicle + SHTP CAP (3 mg/kg, 5 min) + SHTP CAP (3 mglkg, 30 min) + SHTP CAP (30 mg/kg, 5 min) + SHTP CAP (30 mg/kg, 30 min) + 5HTP

309 386 407 41 402 436

1.2 7.5 7.9 9.3 2.6 3.9

+ 0.3 i 0.8 c 1.2 ?z 1.4 r 0.8* I 2.1

1.7 9.4 12.0 13.8 4.7 5.7

2 -t -t if -t

0.4 1.4 1.3 0.81

Study B (SHT, IO Fmol/kg) Vehicle only (no 5HT) Vehicle + 5HT CAP (3 mg/kg, 5 min) + 5HT CAP (3 mg/kg, 30 min) + 5HT CAP (30 mg/kg, 5 min) + 5HT CAP (30 mg/kg, 30 min) + 5HT

309 386 407 415 402 436

1.7 6.5 7.5 6.2 1.6 2.2

2 ? 2 t t t

1.8 8.2 13.5 11.5 3.1 3.0

-c + + t? r

0.4 0.8 1.0* 1.4 1.1* 0.8*

0.4 0.5 1.3 1.7 0.7* 0.8*

1.ot 1.8

Shown are means t SE for groups 5-6. The same groups were used in Studies A and B, run 3 days apart, and the body weights at the time of Study A were used for Study B. *[1<0.01, tpcO.05 relative to Vehicle-Dipsogen condition.

[14], and there is no reason to suppose that the access of SHTP to brain is altered by nephrectomy. It is relevant to note that the plasma half-life of SHTP after IP administration to rats is of the order of 1 hr [2,18]. Concomitant levels of 5HT in plasma are only slightly elevated, but the levels of 5-hydroxyindoleacetic acid (SHIAA), the primary metabolite of 5HT, are greatly increased for several hr. The plasma half-life of 5HT is extremely short, with up to 98% removed in a single pass [25]. However, the hypotensive effect of 5HT (2 mg/kg) lasts for over 1 hr [l], suggesting either that its absorption is slow or that some of the 5HT that is taken up into tissues and blood platelets may subsequently be released. However, while these pharmacokinetic differences might explain some of the contrasts between SHTP and 5HT, they do not readily explain the present nephrectomy data. However, there is an additional possibility that distinct but functionally parallel mechanisms might be engaged by SHTP. This is not without precedent. The findings referred to above with isoproterenol [ 1I] could be interpreted in support of just such a parallelism. In that particular case both hypotension- and renin-related mechanisms might normally be engaged and either may be sufficient to induce drinking. Additionally, as we have discussed in other recent studies [7-91, the satiation of drinking must be considered when more than one dipsogenic signal is present. Thus, either the ingestion of or loading with water will potently counteract a current angiotensin-related stimulus to drink. Under appropriate conditions it may be possible to show that drinking depends on either a mechanism A or a mechanism B, but it may be more difficult to show that both A and B are operative simultaneously because the quantity of water consumed, the universal dependent measure of the strength of the “thirst signal, ” is so potently affected by the consequences of ingestion.

EXPERIMENT 2: EFFECTS OF LOW AND HIGH DOSES OF CAPTOPRIL ON WATER INTAKE INDUCED BY SHTP OR 5HT The drinking responses to stimuli that release renal renin often are increased by pretreatment with low doses of angiotensin converting enzyme inhibitors such as captopril [ 12, 16, 221. This may be because the elevated circulating levels of AI produced by these agents may result in greater access of AI to the brain where it can be converted to AII, provided that the AI-converting enzyme in the brain is not concurrently inhibited. Conversely, high doses of captopril may also inhibit converting enzyme in brain, and thus inhibit AI-related drinking. This dual effect of captopril may thus be used to give an indication of whether drinking to a given agent proceeds via an angiotensin pathway. Thus, if the drinking induced by 5HT and/or by SHTP is mediated by activation of the RAS, then it would be anticipated that drinking would be facilitated by low doses and inhibited by high doses of captopril. Conversely, for stimuli that do not engage the RAS, no such dual effect would be expected. Previous studies from this laboratory have indicated that a high dose of captopril (35 mg/kg, IP 15 min prior to the dipsogen) inhibits both SHTP- and SHT-induced water intake [6, 13, 231; lower doses of captopril were not tested in those studies. In the present study, both low and high doses of captopril are used. In addition, because the time between administration of captopril and the dipsogen may have an effect on the results, we have used delays of either 5 or 30 min between captopril and the dipsogen. METHOD

Experimentally-naive

male rats, weighing between

270

SEROTONIN

505

AND DRINKING

and 490 g (mean 409 g) were used in this experiment. The general procedures were as in Experiment 1. In the first study, rats were administered captopril (SQ 14,285; 0,3 or 30 mg/kg) in a volume of 1 ml saline/kg, SC. All rats received an injection of 5HTP (20 mg/kg or 9.1 pmol/kg IP of the free base, Sigma, dissolved in dilute HCl) either 5 or 30 min later, and water intake was then recorded hourly for 3 hr. The data for the groups pretreated with saline (0 mg/kg) 5 or 30 min prior to SHTP were identical, and therefore have been combined. The second study used the same male rats and a procedure identical to that of the first study, except that the dipsogen was 5HT (4 mg/kg or 10 pmol/kg of creatinine sulfate salt, SC). RESULTS

The results are shown in Table 2. SHTP-induced drinking was affected by the main factors of captopril, F(4,24)=6.94, pO. 1. The SHTP-induced water intake was elevated after 2 hr in the groups treated with the low dose of captopril, either 5 or 30 min before, although only the latter comparison was significant (p
Captopril in a high dose (30 mg/kg) suppressed water intake induced by both 5HT and SHTP, thus confirming our previous observations [13,23]. There was a significant enhancement of water intake after a low dose of captopril with both dipsogens, but this effect occurred only during the second hr of the drinking test. This is of some interest because rats given only 5HT or SHTP normally complete most of their drinking in the first hr. As was discussed in Experiment 1, we do not know with certainty either the plasma half-lives of these agents, or the duration of any renin release that they produce, but over 90% of the time-integrated hypotensive response to 5HT occurs in the first hr. If the concentrations of renin in plasma follow a similar timecourse then, at least for 5HT, there would be little or no excess AI circulating in the second hr. In this event, the reason for the effectiveness of captopril at this time is not clear. In the case of SHTP, which does not produce marked hypotension at the dose used, but nonetheless produces a short-term release of renal renin equivalent to that produced by a ten-fold lower dose of 5HT [ 11, the levels of SHTP and SHIAA may remain elevated in plasma and tissues for several hr. Further studies will be needed to clarify the relationships between the biological half-lives of these agents, the time-course of their effects on plasma AI and AII, and the induced water intake. However, the interpretation of such studies will be complicated by the finding that the dipsogenic

effect of a SC injection of AI1 extends beyond the time at which plasma levels of AI1 should have decayed to basal [8]. In summary, the data from the present experiment support the view that AI1 mediates, at least in part, the drinking responses to both 5HT and SHTP in rats. EXPERIMENT 3: EFFECTS OF PERIPHERAL METERGOLINE ON DRINKING INDUCED BY SHTP AND SHT, AND OF CENTRAL METERGOLINE OR SHTP In Experiment 1, we suggested that the persistence of SHTP-induced drinking after nephrectomy may indicate a mechanism for drinking in addition to activation of the RAS. While the data from Experiment 2 show no differential effect of captopril on 5HT- versus SHTP-induced water intake, they do not rule out the possibility that two mechanisms may act in parallel. One clear difference between 5HT and SHTP is that only the latter is known to cross the blood-brain barrier and thus influence brain 5HT activity. We have previously shown that methysergide, given peripherally, blocks the dipsogenic effect of SHTP in intact rats [13]. In the present study we examine the effects on SHTP- and SHT-induced drinking of metergoline, a more potent and centrally-active 5HT receptor antagonist [3,10]. In order to assess whether any antagonistic action against SHTP is a result of either central or peripheral effects, metergoline was administered by both peripheral and central routes. METHOD

Adult male Sprague-Dawley rats were used in each of the several studies which comprise this experiment. Maintenance and general procedures were as described in Experiment 1. Study A: Peripheral Metergoline Drinking

and SHTP-Induced

Groups of 6 rats received SC injections (Farmitalia; 0.1, 1 or 2 mg/kg), or the dilute cle, and 15 min later received either SHTP HCI salt, SC), or its saline vehicle. Water corded 30, 60 and 120 min after the second

of metergoline acetic acid vehi(25 mg/kg of the intakes were reinjection.

Study B: Peripheral Metergoline and Drinking Induced 5HT, Angiotensin II, and Hypertonic NaCl

by

Groups of 6 rats received SC injections of metergoline (1 or 2 mg/kg, SC), followed 15 min later by either 5HT (4 mg/kg of the creatinine sulfate salt), ileu5-angiotensin II (Sigma; 200 &kg, SC), or 1 M NaCl(5 mEq/rat, IP). Water intakes were recorded as above. Study C: Central Metergoline

and ZHTP-Induced

Drinking

Rats were surgically implanted with an indwelling 23 gauge stainless steel cannula ending in the right lateral ventricle. Using a general anesthetic (Equithesin, 2.5 ml/kg) and stereotaxic coordinates of 1.5 mm lateral and 4.0 mm ventral from the flat skull on bregma, the cannula was attached to the skull with screws and dental cement. Perioperative penicillin G was administered, followed by a 1 week recovery period. It was then established that all.rats drank in excess of 2 ml of Water in response to an ICV injection of AI1 (2.5 ng). For the ICV injections, the rat was held in a towel, the wire obturator removed, and the injection made over a 10 set period through a 10.6 mm long, 30 gauge needle.

ROWLAND.

C‘APUTO AND FKECILY

TABLE 3 EFFECT OF PERIPHERAL AND CEREBROVENTRICULAR (ICV) METERGOLINE (MET) ON WATER INTAKE INDUCED IN RATS BY’SHYDROXYTRYFI’OPHAN (SHTP), SEROTONIN (SHT). AND OTHER DIPSOGENS Water

Intake (ml)

Mean

Body Weight (g)

Treatment

1 hr

2 hr

Study A

(Peripheral metergoline) Vehicle-Saline Vehicle-SHTP (98 wmol/kg, SC) MET (0.1 mg/kg, SC)-SHTP MET (1 mg/kg, SC)-5HTP MET (2 mg/kg, SC)-SHTP

Study B (Peripheral metergoline) Vehicle-Saline Vehicle-5HT (10 pmoYkg, SC) MET (0.1 mg/kg, SC)-5HT MET (1 mglkg, SC)-5HT MET (2 mglkg, SC)-5HT

343 340 301 337 335

I.0 7.1 1.1 2.1 0.9

t- 0.4 +- 0.6* * 0.5$ 2 0.6$ i- 0.5i

262 263 293 262 213

0.2 4.0 5.3 2.5 0.6

& 0.2 2 0.7* -c 0.6* + 0.7t!: 2 0.4$

1.1 i- 0.4 8.2 +- 1.5* I.2 -+ 0.5$ 2.3 + 0.6$ 0.9 t 0.5$

0.2 IT 0.2 5.1 t 0.5* 5.3 + 0.6* 2.8 2 0.8tS 0.8 + 0.6$

Vehicle-AU (200 p/kg, SC) MET (1 mg/kg, SC)-AI1

662

5.8 + 2.5

7.0 2 2.5

648

5.5 I

2.0

5.6 * 2.0

Vehicle-NaCl (5 mEq, IP) MET (1 mg/kg, SC)-NaCI

631 653

11.4 2

I.6

12.3 2 1.7

11.2 -+ 1.0

11.3 -t I.0

Study C (Central metergoline) Vehicle-SHTP (98 PmoUkg, SC) MET (10 pg, ICY)-SHTP MET (30 pg, IC05HTP

453 460 367

3.3 t 1.6+ 3.4 t 1.21 3.8 I 2.8

Study D (Central SHTP) Saline SHTP (98 nmol, KY) 5HTP (490 nmol. ICV)

323 314 311

0.7 -c 0.4 2.2 r 0.71 0.0

11.8 i 7.2 *

1.5* 1.9*

4.3 t 2.9$.

I.0 * 0.4 2.6 f

0.51

0.5 t 0.3

Shown are means 2 SE for groups of 6-11. *p
The rats (N =611) were injected with SHTP (25 mg/kg of the HCl salt, SC) and either a simultaneous ICV bolus of metergoline (10 or 30 pg in a volume of 5 ~1) or 0.1% acetic acid vehicle. Study D: Effect of Intrucerehroventricular Drinking

SHTP an

This study was performed using rats cannulated as above and injected with SHTP (0,25 or 125 &5 ~1 of 0.15 M NaCl). RESULTS

Study A

Peripheral administration of metergoline attenuated the water intake induced by SHTP (Table 3). Two-way ANOVA with dose of metergoline and time as factors showed a main effect of dose, F(3,33)=19.95,p<0.001. The water intake of rats given SHTP and any of the doses of metergoline did not differ from the baseline intake of rats not treated with SHTP. Metergoline alone had no effect on basal intake (data not shown).

Study B

Peripheral administration of metergoline attenuated SHT-induced drinking (Table 3). Two-way ANOVA revealed main effects of dose of metergoline, F(3,20)=12.46, p
Metergoline given ICV inhibited SHTP-induced water intake after 2 hr, F(2,25)=3.45, ~~0.05, but not at 1 hr (Table 3). However, at the dose of 30 pg metergoline, the rats adopted abnormal (sometimes cataleptic) postures, and the reduction of intake may thus have been nonselective. Further studies were not performed with ICV metergoline.

SEROTONIN

507

AND DRINKING

Study D

ICV injection of 25 pg SHTP increased the intake of water relative to vehicle-injected controls (Table 3). The significant p
The present results with metergoline complement OUT previous observation that methysergide inhibits the drinking induced by both SHTP and 5HT [13]. The failure of cinanset-in in this regard [13] thus seems anomalous. Since 5HT itself does not cross the blood-brain barrier, peripherallyadministered 5HT should engage a peripheral mechanism. However, our previous findings that SHTP-induced drinking was completely blocked by pretreatment with carbidopa [14], a drug which ostensibly prevents decarboxylation of SHTP to 5HT only in the periphery [26], suggests that any central conversion of SHTP to 5HT is not important for the drinking response. Carbidopa has no effect on water intake induced by 5HT, which by-passes the enzymatic blockade [ 141. Nonetheless, the present findings that (1) SHTP given ICV can induce drinking, and (2) that ICV metergoline can inhibit SHTP-induced drinking are suggestive preliminary evidence that a brain 5HT mechanism may be engaged. The finding that peripherally-administered metergoline was more effective in attenuating SHTP- than SHT-induced drinks (Studies A and B) has some interesting implications. One might argue that SHTP given SC is converted to 5HT in plasma at a relatively slow rate [2]. This would be consistent with the blockade with low doses of metergoline (Study A) as well as with the minimal effects of SHTP on blood pressure [ 11. In contrast, one might argue that 5HT injection achieves higher and more transient plasma elevations in 5HT, and this would be consistent with the necessity for higher doses of antagonist (Study B) and the pronounced acute hypotensive effect [l]. This kinetic model predicts that, compared with 5HT, SHTP should produce a slower and longer-lasting drinking response, and a correspondingly slower renin release. Neither of these predictions is borne out by data past [I, 6, 13, 141 or present. GENERAL

DISCUSSION

The present experiments examined the hypothesis that activation of the peripheral RAS is the exclusive mediator of the water intake induced by SHTP and by 5HT. The effects of nephrectomy (Experiment 1) differed markedly between the two dipsogens, with SHTP-induced drinking largely surviving removal of the kidneys. However, both our previous studies 113,231, and the data from Experiment 2 showing that low doses of captopril enhance and high doses inhibit both SHTP- and SHT-induced drinking, argue in favor of a major

contribution of the RAS. It therefore seemed possible that SHTP, which crosses the blood-brain barrier, may engage a brain drinking circuit which normally works in parallel with peripheral angiotensin-related mechanisms. A possible substrate for this is suggested in recent work by Lind [16a] who has found a dense serotonergic fiber plexus within the subfomical organ. This is known to be one of the brain regions critical to the normal expression of AIIrelated drinking. Depending upon the nature of the bloodbrain barrier at this plexus, SHTP may preferentially gain access to this site and may indeed modulate the actions of AI1 in the organ. Further studies will be needed to address these questions. The possible contribution of brain 5HT mechanisms was investigated in Experiment 3, with mixed results. SHTP injected ICV was capable of inducing a small intake of water. Further, ICV injection of a 5HT antagonist was able to reduce drinking induced by peripheral SHTP. However, the specificity of action of that dose of metergoline (30 pg) was questioned because it produced motor side-effects and also because the same quantity of the antagonist given peripherally (0.1 mgikg) was effective in blocking SHTP-induced drinking. Thus, the ICV metergoline might have leaked out of the brain and blocked 5HT receptors in the periphery. The results are thus inconclusive with regard to the possible participation of a brain 5HT mechanism in SHTP-induced drinking. However, as was discussed in Experiments 1 and 3, some unanswered questions remain concerning the half-lives of SHTP and 5HT, and how this relates to the physiological and behavioral effects of these agents. We have attempted to address these questions in other ways including studies of drinking following administration of the 5HT agonists, quipazine and p-chloroamphetamine [20]. Despite the fact that these agents promote renin release [24], and thus should at least give rise to an AII-related drink, we have been unable to elicit drinking reliably with them. Another approach that we have used is to make brain lesions (with 5,7_dihydroxytryptamine) that deplete brain 5HT and may produce 5HT receptor supersensitivity. These rats drank slightly more than intact rats following administration of SHTP, SHT, or AII; thus, because the SHTPinduced intake was not enhanced selectively, this study provided no evidence in favor of a brain 5HT action. However, it should be noted that the SHTP may be decarboxylated at different site(s) in brains of lesioned rats, and so the negative result is not easy to interpret. It has been suggested that endogenous 5HT may be released during meals and contribute to postmeal drinking, insofar as this latter is partially blocked by methysergide [15]. It is possible that this mechanism, together with other mealrelated fluid and hormonal shifts (each of which may be subthreshold) summate, possibly via AII, to produce postmeal drinkiug. We note that subdiaphragmatic vagotomy attenuates the water intake induced by peripheral administration either 5HT or AI1 [ 19,211, although the reason for this interesting finding is unknown at this time.

of

REFERENCES 3. cox,

1. Barney, C. C., R. M. Threatte, D. C. Kikta and M. J. Fregly. Effects of serotonin and L-Shydroxytryptophan on plasma renin activity in rats. Pharmacol Biochem Behav 14: 895-900, 1981. 2.

Bogdanski, D. F., H. Weissbach and S. Udenfriend. Pharmacological studies with the serotonin precursor, 5-hydroxytryptophan. J PharmacoL Exp Ther 122: 182-194, 1958.

B. and C. Ennis.

tryptaminergic

autoreceptors

Characterization

of Shydroxyin the rat hypothalamus. J Phorm

Pharmacol34: 438-441, 1982. 4. Fitzsimons, J. T. The Physiology of Thirst and Sodium Appe-

rite. Cambridge, MA: Cambridge University Press, 1979.

ROWLAND.

SO8

Freed, C. R.. H. Echizen and D. Bhaskaran. Brain serotonin and blood pressure regulation: studies using in \jirr, electrochemistry and direct tissue assay. Liti, SC+ 37: 1783-1793. 1985. 6. Fregly, M. J., T. M. Connor, D. C. Kikta and R. M. Threatte. Dipsogenic effect of I-5-hydroxytryptophan in rats. Bruin Rcs Bull 5: 719-724, 1980. I. Fregly, M. J., J. E. Greenleaf and N. E. Rowland. Effect of intraperitoneal and intragastric loading with water and isosmotic solutions of saline and glucose on water intake of dehydrated rats. Bruin Rrs Bull 16: 415-420, 1986. 8. Fregly, M. J. and N. E. Rowland. Do peripheral and cerebroventricular injections of angiotensin II act at the same site? Studies on the additivitv of drinking. Bruin Res Bull 16: 24%257. 1986. 9. Fregly, M. .I. and N. E. Rowland. Effect of intragastric and intraperitoneal water and saline loads on the pharmacologic induction of drinking in rats. Brain Rrs Bull 16: 407-414, 1986. 10. Garattini, S. and R. Samanin. Drugs affecting serotonin: survey. In: .S<,rotonin in Heulth and Di.tc~r.vr, Vol 2, edited by W. B. Essman. New York: Spectrum, 1978, pp. 247-293. 11. Hosutt, J. A., N. Rowland and E. M. Stricker. Hypotension and thirst in rats after isoproterenol treatment. Physiol Brhnv 21: 593-598, 1978. 12. Katovich, M. J., C. C. Barney, M. J. Fregly and R. E. McCaa. Effect of an angiotensin converting enzyme inhibitor (SQ 14,225) on /3-adrenergic and angiotensin-induced thirsts. Eur J Phurr?~crc~ol56: 123-130, 1979. 13. Kikta, D. C., C. C. Barney, R. M. Threatte, M. J. Fregly, N. E. Rowland and J. E. Greenleaf. On the mechanism of serotonininduced dipsogenesis in the rat. Pharmuc~d Biochern Brhuia 19: 5 19-525. 1983. 14. Kikta. D. C., R. M. Threatte, C. C. Barney, M. J. Fregly and J. E. Greenleaf. Peripheral conversion of L-5-hydroxytryptophan to serotonin induces drinking in rats. Phurmuc~rd Biochrm 5.

Rchu~~ 14: 889-893,

1981.

15. Kraly, F. S., S. M. Specht, S. Geasey and J. E. McOsker. Examining a role for serotonin in the control of drinking elicited by eating. Eostrm Psycho/ Assoc. Ahstr 57: 40, 1986.

CAPUTO

AND FRL(;LY

16 Lehr, D., H. W. Goldman and P. Casner. Rerun-angiotensin role in thirst: paradoxical enhancement of drinking by angiotensin converting enzyme inhibitor. SCit,,,r (’ 182: 103I -1034. 1973. Iha.Lind, R. W. Bi-directional, chemically specified neural connections between the subfomical organ and the midbrain raphe system. Brain Rcs 384: 250-261, 1986. 17. Meyer, D. K.. M. Abele and G. Hertting. Influence of serotonin on water intake and the renin-angiotensin system in the rat. Awh

lrlt Phurmuc~ody,~

212: 130-140,

1974.

18. Moir, A. T. B..and D. Eccleston. The effects of precursor loading in the cerebral metabolism of 5-hydroxyindoles. .I N<>rrrr~rhrrt7 15: 1093-1108, 1968. 19. Rowland, N. Impaired drinking to angiotensin I1 after subdiaphragmatic vagotomy in rats. Physiol Brhtrl, 24: 1177-1180, 1980. 20. Rowland, N. E. and M. J. Fregly. Serotonin-induced drinking in rats: further analysis. Sl E.rp Ther 197: 545-555. 1976.