Forebrain circumventricular organs mediate salt appetite induced by intravenous angiotensin II in rats

Brain Research 949 (2002) 42–50 www.elsevier.com / locate / bres

Research report

Forebrain circumventricular organs mediate salt appetite induced by intravenous angiotensin II in rats Michael J. Morris, Wendy L. Wilson, Elizabeth M. Starbuck, Douglas A. Fitts* Department of Psychology, University of Washington, Box 351525, Seattle, WA 98195 -1525, USA Accepted 29 April 2002

Abstract Two circumventricular organs, the subfornical organ (SFO) and organum vasculosum laminae terminalis (OVLT), may mediate salt appetite in response to acute intravenous infusions of angiotensin (ANG) II. Fluid intakes and mean arterial pressures were measured in rats with sham lesions or electrolytic lesions of the SFO or OVLT during an intravenous infusion of 30 ng / min ANG II. Beginning 21 h before the 90-min infusion, the rats were depleted of sodium with furosemide and given a total of 300 mg / kg captopril in 75 ml / kg water in three spaced gavages to block the usual salt appetite and to hydrate the rats. No other food or fluids were available for ingestion. Sham-lesioned rats drank 9.361.2 ml if 0.3 M NaCl alone was available and drank 8.961.6 ml of saline and 3.761.6 ml of water if both were available. Either SFO or OVLT lesions reduced the intakes of saline to ,5 ml in both conditions and of water to ,1 ml. Mean arterial pressure did not differ among the groups and was maintained above 100 mmHg after the depletion and captopril treatments because of the large doses of water. Thus, a full expression of salt appetite in response to an acute intravenous infusion of ANG II requires the integrity of both the SFO and OVLT.  2002 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behavior Topic: Ingestive behaviors Keywords: Angiotensin-converting enzyme inhibitor; Captopril; Drinking; Organum vasculosum laminae terminalis; Sodium appetite; Sodium depletion; Subfornical organ

1. Introduction A loss of body sodium leads rapidly to a stimulation of the renin–angiotensin–aldosterone system and eventually to an increased preference for and consumption of sodium [17]. Much evidence has accumulated to demonstrate that angiotensin (ANG) II acts on the brain to induce this salt appetite in rats. For instance, intracranial infusions of ANG II can cause euhydrated rats to increase their consumption of a normally avoided hypertonic saline solution [2,3,12], and a blockade of ANG II receptors or synthesis can disrupt salt appetite in sodium-depleted animals [22,25,45]. It is not known for certain how much or what part of the renin–ANG system is essential for salt appetite. ANG I is *Correspondence address. Tel.: 11-206-543-2440; fax: 11-206-6853157. E-mail address: [email protected] (D.A. Fitts).

synthesized in the blood from angiotensinogen secreted by the liver and from renal renin; in the brain and other tissues it is formed from locally generated substrates and renin-like enzymes. Angiotensin-converting enzyme is available both in the peripheral microcirculation and in the brain to convert ANG I to ANG II. ANG II type IA receptors are abundant in many peripheral tissues and organs as well as in brain areas known to be involved in fluid and cardiovascular homeostasis. Among these brain areas are the circumventricular organs of the brain, and in particular the subfornical organ (SFO) and organum vasculosum laminae terminalis (OVLT) [21,26]. These forebrain circumventricular organs bind circulating ANG II outside the blood–brain barrier as well as neuronally synthesized ANG II. Thus, an infusion of ANG II into the cerebral ventricles may generate a salt appetite either by acting directly on the cerebral ANG II receptors or by stimulating receptors in the circumventricular organs that

0006-8993 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 02 )02963-3

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

are normally targets of circulating ANG II. Similarly, intracerebroventricular infusions of ANG II receptor blockers could act either on central receptors or on ‘peripheral’ receptors in circumventricular organs. Early evidence seemed to suggest that only the cerebral ANG II system was important in salt appetite [23], but recent evidence demonstrates that the peripheral ANG II system is also critical [14,34,40,44]. For example, blocking the synthesis of ANG II in the circulation without blocking synthesis in the brain is sufficient to diminish salt appetite [34]. Thus, the peripheral ANG II system may activate the central system to produce salt appetite. Some data suggest that this activation across the blood–brain barrier may be mediated by circumventricular organs, but those findings are not conclusive. Infusions of ANG II into OVLT cause salt appetite, but the results from studies with SFO infusions are mixed [1,5,12,13,35]. Similarly, studies show that lesions of these circumventricular organs sometimes do and sometimes do not reduce salt appetite generated by sodium depletion [4,16,30,28,33,43], and even in successful lesion studies there is no conclusive evidence that the lesions are doing anything specific to the ANG II system. Thus, the aim of this experiment was to test the effects of electrolytic lesions of either the SFO or OVLT on salt appetite generated specifically by circulating ANG II. Unfortunately, intravenous infusions of ANG II into euhydrated rats do not create a salt appetite with a short latency [6,10,46] the way they create a thirst for water [18], and long-term infusions of ANG II could generate a salt appetite nonspecifically by causing chronic natriuresis or aldosterone secretion [6,10,46]. These complications can be overcome by doing the intravenous infusions into rats that are depleted of sodium and treated with a high dose of ANG-converting enzyme inhibitor such as captopril (CAP) [14,40]. The depletion prepares the rat for drinking sodium solutions, but the captopril blocks that appetite. An intravenous infusion of ANG II then restores the appetite in an ANG-specific fashion [14,40]. If the circumventricular organs are critical sites of detection for this circulating ANG II, then lesions should reduce this salt appetite that is specifically created by circulating ANG II binding in those areas.

43

the rats were maintained on a 12 h light / 12 h dark cycle. Each rat was used only once in the following experiment.

2.2. Stereotaxic surgery Most rats (n5111) received an electrolytic lesion of either SFO or OVLT, or a sham lesion. They were first anesthetized with Equi-Thesin (Drug Services of the University of Washington Hospital, Seattle, WA; 0.35 ml / 100 g, i.p.) and secured in a Kopf stereotaxic instrument. The skull was leveled between bregma and lambda and a 2.5-mm hole was drilled near bregma. Lesions were made using a 1.0-mA current with a 31-gauge tungsten wire electrode insulated with teflon except for the cross-sectional area at the tip. SFO lesions used four penetrations of the electrode for 10 s each on the midline at anteroposterior (AP) 20.1, 20.3, 20.5, 20.7 mm and dorsoventral (DV) 25.0, 24.9, 24.7, 24.5 mm relative to the midsaggital sinus. OVLT lesions were made with one penetration for 12 s at AP 11.3 and DV 27.3. For all lesions, the electrode passed through the midsagittal sinus. Sham lesions were made by advancing the electrode through the midsaggital sinus without passing current. Bleeding resulting from the penetration of the sinus was easily controlled with direct pressure. Bone wax was applied to seal the hole in the skull before closing the wound. An intramuscular injection of 0.2 ml Gentamicin and topical Betadine were administered to control infection. The rats were allowed to recover from surgery on a warm heating pad before being returned to their home cages. Experimentation began approximately 2 weeks after surgery. Seven rats had no lesion and instead had a 23-gauge stainless steel guide cannula implanted in a lateral cerebral ventricle. The cannula was anchored to stainless steel screws in the skull with methyl methacrylate and was obturated with 31-gauge stainless steel wire except during injections. Injections or infusions were made using a 31gauge stainless steel injector that extended 0.5 mm beyond the tip of the guide. The patency of the cannulas was tested 1 week after surgery and 12 days before the experiment by injecting 40 pmol of ANG I in 2 ml of sterile isotonic saline and measuring the water drinking response for 15 min. All seven rats passed this test (mean6S.E.M., 5.861.0 ml).

2. Materials and methods

2.3. Histology 2.1. Animals The subjects were 118 male Long–Evans rats of Charles River (Wilmington, MA) origin weighing 300–550 g. They were maintained individually in hanging wire mesh cages with Harlan Teklad rat chow and tap water ad libitum. The rats also had access to both water and 0.3 M NaCl solution for drinking for at least 1 week prior to the experiments. The temperature was controlled at 23 8C and

After experimentation the lesioned rats were deeply anesthetized with pentobarbital sodium and perfused through the heart with isotonic saline for exsanguination followed by 10% formalin in saline solution for fixation. Brains were removed and stored in 10% formalin and saline until cutting on a freezing microtome in 50-mm sagittal sections. Sections were mounted on glass slides, stained with Thionine, coverslipped, and examined using a

44

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

microprojector or light microscope. Rats were considered to have good lesions of the SFO if the entire rostroventral pole containing the efferent fibers from the SFO were destroyed. OVLT lesions were considered to be complete if they destroyed no more than the most ventral third of the ventral median preoptic nucleus and the entire dorsal cap of OVLT. Complete OVLT lesions extended in a rostrocaudal dimension from the ventral surface of the brain to the third ventricle, thus destroying all dorsal connections of the OVLT. Lesions that destroyed greater than one third of the ventral median preoptic nucleus were not counted in the OVLT lesion group.

2.4. Catheterization surgery Polyethylene catheters were implanted into the left femoral vein and artery 2 days before the beginning of the experiment under halothane anesthesia in all sham-lesioned or lesioned rats for measurement of mean arterial pressure (MAP) and for i.v. infusions of ANG II. Catheters were constructed of PE-10 tubing heat-welded to a longer piece of PE-50 tubing; the latter tubing was tunneled to an exit wound between the scapulae. The catheters were filled with 100 U / ml heparin in sterile isotonic saline and obturated until the time of the experiment. Measurements of MAP and heart rate were recorded continuously by computer at 100 Hz [11].

2.5. Sodium depletion and captopril blockade On the day of the experiment, all rats received a subcutaneous injection of furosemide (20 mg / kg) to induce diuresis and natriuresis and were placed into individual metabolic cages without access to food, water, or saline. At the end of the 1-h diuresis period, urine was collected, and the volume was measured to the nearest 1 ml. Samples were frozen in sealed vials for later determination of sodium and potassium concentrations by flame photometry. After the 1 h diuresis period, rats were given an intragastric (i.g.) load of 100 mg / kg CAP diluted in 30 ml / kg of tap water. Rats were then returned to metabolic cages without food, water, or saline. Urine was again collected at 17 h after the furosemide injection. Rats were given another i.g. load of CAP at 18 h (100 mg / kg in 30 ml / kg tap water) and another at 20 h (CAP 100 mg / kg in 15 ml / kg tap water). Thus, the total captopril given was 300 mg / kg in 18 h. The total water load was 75 ml / kg (30 ml for a 400 g rat), and 45 ml / kg of the load was administered within 2 h of the infusion. The purpose of the large volume of loaded water was to prevent any effect on salt intake by an interaction of the circumventricular organ lesions with osmotic dehydration [30]. This total volume was selected because it is approximately what a sodiumdepleted rat will voluntarily drink overnight when food is withheld [28].

2.6. ANG II infusion and drinking test The following procedure applies to all sham-lesioned and lesioned rats. Immediately after the last i.g. load of captopril and water, rats were transferred to cylindrical plastic cages for measurement of MAP and infusions of ANG II. The arterial and venous catheters were hooked up to the blood pressure transducers and to infusion pumps, respectively. Rats were allowed to recover from handling for approximately 15 min, with the last minute of recovery recorded as a baseline blood pressure. The syringe pumps were then turned on for 1 min at a rate calculated to clear the dead space and then for 90 min to deliver ANG II intravenously at 30 ng / min in 0.6 ml / h isotonic saline vehicle. Five minutes into the infusion rats were given either tap water and 0.3 M NaCl or saline solution alone in glass burettes that allowed for recording intakes to the nearest 0.1 ml. Latency to drink either fluid was recorded and intakes were recorded at 15, 30, 45, 60, 75, and 90 min. Rats were considered to have taken a drink if they licked the drinking spout sufficiently to cause a bubble to rise in the burette. If a rat did not drink a fluid in 90 min it received a latency score of 90 min for that fluid. At the end of the 90-min infusion, the pumps were turned off and blood pressures were recorded for 10 min postinfusion. Blood pressure in the last minute of the postinfusion period was used for the analysis of the recovered MAP. The seven rats with lateral ventricular cannulae were treated on the experiment days identically to the lesioned rats except that they were not fitted with vascular catheters and therefore were not hooked up to the blood pressure and i.v. infusion apparatus. Instead, an injector was inserted into the guide cannula in the lateral ventricle, and isotonic saline was infused continuously at 2 ml / h beginning 5 min before the presentation of the water and saline drinking tubes. Thus, these rats were treated similarly to the lesioned animals in that they had received some prior surgery and were lightly instrumented during the experiment, but they never received an exogenous treatment with ANG II. The group was originally intended as a control group for a different experiment, but they are included here instead as an informal comparison because it was, unfortunately, not logistically possible to include formal controls having i.v. infusions of vehicle instead of ANG II. The data for the group are included in Fig. 3 but were not included in the statistical analysis.

2.7. Statistics Data were analyzed using analysis of variance (ANOVA). Planned comparisons used Fischer’s least significant difference test following a significant F ratio. Unplanned post-hoc comparisons used Tukey’s Honestly Significant Difference (HSD) test. Lesion and fluid availability (saline only versus saline and water for choice drinking) were used as between-subject factors. Non-

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

45

parametric tests were used for latencies because the data grossly violated the assumptions of homogeneity and normality for the ANOVA. A probability of less than 0.05 was required for significance.

3. Results Histological analysis showed that 17 rats had SFO lesions (SFOX), 15 rats had OVLT lesions (OVLTX), 25 rats had sham lesions (SHAM), and 33 rats composed a control lesion group that included missed SFO and OVLT lesions (ControlX). There were 21 rats that had either partial damage to SFO or partial damage to OVLT and were eliminated from the study. Thus, 90 rats were included in the statistical analysis. The sample sizes of each group are shown in Table 1. Photomicrographs of various lesions are illustrated in Fig. 1. No significant differences were found among the groups with respect to volume of urine excreted as measured after the 1 h diuresis period (SHAM, 13.662.50 ml; SFOX, 13.262.9 ml; OVLTX, 12.261.7 ml; ControlX, 12.562.1 ml) or the following morning (SHAM512.062.5 ml; SFOX512.763.3 ml; OVLTX510.761.9 ml; ControlX5 12.062.9 ml). The amount of urinary sodium excreted also did not differ among the groups at either time, thus assuring that all rats were equally depleted at the start of the infusions. The total amounts of sodium excreted by the beginning of the infusions were: SHAM, 1.5260.06 mmol; SFOX 1.4360.10 mmol; OVLTX 1.4160.04 mmol; and ControlX, 1.4160.04 mmol. Cumulative 90-min saline intake was analyzed by between-subjects ANOVA with the four lesion groups as one factor and the availability of water during the test as the other factor (available or not). The data are displayed in Fig. 2 for the groups without water available and in Fig. 3 for the groups with water available. The main effect for lesion was significant for saline intake, F3,82 57.61, P, Table 1 Median latency in minutes to drink water or saline during intravenous infusions of ANG II in rats with lesions of the SFO or OVLT SHAM

SFOX

OVLTX

ControlX

17 3.0 0–62

9 34.5 3–90

8 6.6 0–90

28 10.6 0–90

Water and saline groups n 8 Water Median 38.7 Range 6–90 Saline Median 6.9 Range 0–30

8 56.2 † 7–90 36.9* 1–90

7 90.0* † 13–90 42.4* † 3–90

5 7.1 0–10 9.5 7–24

Saline alone groups n Saline Median Range

Sham5sham lesion; SFOX, OVLTX5SFO or OVLT lesion; ControlX5 missed lesions with no damage to SFO or OVLT. *P,0.05, lesion vs. sham; † P,0.05, lesion vs. ControlX.

Fig. 1. Sagittal photomicrographs of lesions in or near the SFO (A, B) or OVLT (C, D). (A) Destruction of rostral pole of SFO including afferents and efferents; (B) complete lesion of SFO; (C) lesion rostral and dorsal to OVLT omitted from analysis; (D) lesion of about 90% of OVLT including all dorsal connectivity — note that the damage extends to the optic recess. Scale bar in B50.5 mm.

0.001, and planned comparisons revealed that both SFOX and OVLTX rats drank less saline than SHAM or ControlX rats. The interaction of lesion and fluid availability was not significant, F3,82 50.54, indicating that the effectiveness of the lesions was about equal in the two wateravailability conditions despite the apparent difference between the two lesion groups in Fig. 2. Cumulative 90-min water intake was analyzed by oneway ANOVA for the four lesion groups that had water available. The data are shown in Fig. 3. The SFOX and OVLTX groups appeared to drink less than the control groups, but the difference was not significant, F3,24 52.86, P50.058. Fig. 3 also illustrates the 90-min saline and water intakes of rats that were similarly treated except that they had no lesion and received lateral ventricular vehicle infusions instead of i.v. ANG II infusions. Because these rats drank little water or saline, the results confirm that the captopril was effective in suppressing, but not abolishing, salt appetite in this procedure. The median latencies to drink water or saline for the SHAM, SFOX, OVLTX, and ControlX rats are given in Table 1. The latency data violated the assumptions for ANOVA and therefore the data for each row in the table

46

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

Fig. 2. Mean cumulative intakes of 0.3 M NaCl solution by different lesion groups receiving only the saline solution to drink during an intravenous infusion of ANG II. The groups with lesions of the SFO (SFOX) or OVLT (OVLTX) drank significantly less than either the sham-lesioned group or the control group having missed lesions (ControlX).

were analyzed by the nonparametric Kruskal–Wallis test followed by pairwise Mann–Whitney tests if the global test was significant. In the saline alone condition, the latencies did not differ significantly, H3 56.69, 0.05.P.0.10. In the group with access to both fluids, the four lesion groups differed in latency to begin drinking both water, H3 5 12.12, P50.007, and saline, H3 513.32, P50.004. For water drinking, both SFOX and OVLTX groups had significantly longer latencies than the ControlX group, and the OVLTX group had a longer latency than the SHAM group. For saline drinking, both SFOX and OVLTX groups had significantly longer latencies than the SHAM group, and the OVLTX group had a significantly longer latency than the ControlX group. Thus, in the groups receiving both fluids, both the SFO and OVLT lesions caused the rats to wait significantly longer before drinking either water or saline than one or other of the control groups. It should be noted that a short drinking latency for a fluid did not imply a large intake. Occasionally rats, especially in the lesioned groups, would sample the fluid soon after presentation without drinking a substantial volume. The data for mean arterial blood pressure and heart rate throughout the experiments are given in Fig. 4. The data were analyzed by ANOVA with both lesion and fluid availability as factors. The data were virtually identical between the fluid availability conditions, and these have been combined in the graph for convenience of presentation. The arterial pressures of SFOX and OVLTX groups did not differ at any time from that of the two

Fig. 3. Mean cumulative intakes of 0.3 M NaCl or water by different lesion groups receiving both fluids to drink during an intravenous infusion of ANG II. The groups with lesions of the SFO (SFOX) or OVLT (OVLTX) drank significantly less saline than either the sham-lesioned group or the control group having missed lesions (ControlX). The lesion effect on water intake was not significant (P,0.058). The NO ANG symbol represents the intakes of seven rats that were depleted and given captopril without receiving an intravenous infusion of ANG II.

control groups. A lesion-group by time interaction was significant for the heart rates, F12,292 53.32, P,0.001. The post-hoc HSD tests revealed two main sources for this interaction including an elevated preinfusion heart rate in the missed lesion group and, more interesting, an elevated heart rate at the end of the postinfusion recovery in both the SFOX and OVLTX groups. The latter two lesioned groups had higher heart rates at the end of the postinfusion period than their own preinfusion baselines or than the mean rates of the other two ControlX or SHAM groups.

4. Discussion This experiment demonstrates for the first time that the salt appetite response induced by an intravenous infusion

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

Fig. 4. Mean arterial pressure (MAP) and heart rate (HR) during preinfusion (Pre), infusion, and postinfusion periods in all rats of the study. ANG II infusion immediately elevated mean arterial pressure and suppressed heart rate throughout the infusion. No differences among groups occurred for MAP. The SFOX and OVLTX groups developed a slight but significant tachycardia during the postinfusion period relative to the other control groups or to their own preinfusion baselines.

of ANG II in rats is mediated by two forebrain circumventricular organs, the SFO and OVLT. A lesion of either circumventricular organ reduced the salt appetite of infused rats to the approximate level of uninfused rats. This reduction in intake by lesioned rats could not be explained by differences in blood pressure or heart rate between the lesion groups. The first demonstration that a lesion of a circumventricular organ could abolish the water intake caused by an intravenous infusion of ANG II was published more than 25 years ago [27]. That experiment was aided by the fact that intravenous ANG II reliably induces water drinking in normally-hydrated rats. For this reason, the experiment could be conducted straightforwardly by infusing ANG II intravenously into SFO-lesioned rats and observing the decline in intake. The fact that intravenous ANG II does not also reliably cause salt intake with short latency in rats

47

raised some concern that perhaps circulating ANG II did not contribute to salt appetite [10,23,46]. This concern was based partly on the fact that the salt drinking obtained with very long-term infusions of ANG II (i.e. days) could result from non-ANG-specific consequences of the infusion such as a pressor-induced natriuresis or aldosterone secretion [10,46]. It was also based on the fact that intravenous infusions of ANG II receptor blockers did not decrease salt appetite whereas centrally-infused blockers did reduce the appetite in a dose-dependent fashion [23]. Salt appetite with intravenous infusions of ANG II in the rat was made possible by the recognition that an intravenous infusion of ANG II into a normally hydrated animal produces both powerful excitatory and inhibitory influences on salt appetite and water drinking [7,9,36]. The inhibitory influences, which may result from hypertension, activity of cardiac volume receptors [38] or of natriuretic and vasodilator hormones such as atrial natriuretic peptide [15], are not present in sodium-depleted rats. Experiments that successfully controlled these inhibitory events during ANG II infusions resulted in salt appetite and much larger water intakes compared with experiments that did not [7,9,14,36,37,40]. Once it became possible to generate a salt appetite with intravenous ANG II, it immediately became of interest to determine if saline intake induced by circulating ANG II is dependent on the forebrain circumventricular organs. The present experiment now demonstrates that both the SFO and OVLT are critical for a normal salt appetite in response to intravenous ANG II. The dose of captopril used in this study is larger than a dose that is known to block the central synthesis of ANG II after acute administration. That is, peripheral administration of captopril (5–50 mg / kg) effectively blocked drinking in response to ANG I injected into the cerebral ventricles [8]. Because the central ANG II system may be critical for the development of salt appetite after sodium depletion [23,25], it is curious that our peripheral ANG II infusions could generate a salt appetite if that central system was blocked by the high dose of peripheral captopril. This would seem to argue that a stimulation of the peripheral ANG II system can create a salt appetite even when the central system is blocked [41,42]. Alternatively, the results from the acute captopril experiments [8] might not generalize to longer-term blockade. Chronic administration of captopril induces the synthesis of new brain angiotensin-converting enzyme [39], and this may counteract the effects of any captopril that circumvents the blood–brain barrier. Some previous studies found that lesions of SFO or OVLT reduced salt appetite after sodium depletion in rats [4,16,33,43], but other studies of SFO lesions found little effect [28,30]. In addition, we recently found no effect of either SFO or OVLT lesions on daily salt intake induced by adrenalectomy (Wilson, Starbuck and Fitts, unpublished data), a form of salt appetite that is suspected to rely exclusively on ANG II [24]. A major difference between

48

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

the present study and those lesion studies is the timing of the onset of the stimulus for salt appetite: i.e. rapid onset for ANG II infusions and gradual onset for adrenalectomy and sodium depletion. In addition, the adrenalectomy study allowed a full day for expression of the behavior, whereas the ANG II infusion study and sodium depletion studies demanded a relatively rapid response (90–120 min). It is possible that behavior requiring a rapid response to a small or rapidly developing stimulus is more susceptible to a lesion of an individual circumventricular organ. Behavior resulting from a large, long-term stimulus that allows plenty of time for the rat to respond may be fully supported by mechanisms other than that one lesioned circumventricular organ. Several other examples of this rapid / acute vs. slow / chronic distinction exist for circumventricular organs. First, water intake is elevated both acutely and chronically after treatment with the diuretic / natriuretic furosemide [28]. A lesion of the SFO blunts the acute intake occurring immediately after water is returned an hour or so after furosemide injection [33]. Overnight water intake may be elevated 50–100% because of the sodium depletion [28], but this intake is only minimally affected by a lesion of SFO [33,43]. Second, water intake is elevated by treatment with colloids such as polyethylene glycol that slowly draw isosmotic extracellular fluid out of the circulation and into an edema at the site of injection. A lesion of SFO reduces water intake more after a 20% dose than a 30% dose of colloid [19]. It also has a greater effect on water intake when water is provided immediately after administration of the colloid instead of after a delay of 5 h [19]. Thus, the effects of the lesion were greater with smaller levels of stimulation and with more immediate response requirements. A third example is the finding that SFO lesions increase the latency to drink and amount of water consumed acutely after a meal, but do not affect the total daily intakes of food or water [29]. In addition, several of these examples in which SFO lesions did not reduce intake, including the sodium intake after adrenalectomy, allowed the rats to respond not only over many hours but also as a part of their normal nocturnal cycle instead of in the daytime. One alternative interpretation of our data is the possibility that some interaction of SFO or OVLT lesions and sodium depletion or captopril administration made the animals sick. This malaise may then have prevented the appropriate water and salt intakes. We consider this highly unlikely, because rats with these lesions under other circumstances show perfectly normal ingestive responses [16,28–30], and we have no knowledge of any demonstration of malaise among the published studies of lesions of different parts of the lamina terminalis. The rats all appeared alert and healthy as they were connected to the instruments and introduced into the infusion cages, and their blood pressures and heart rates were all within the normal range. However, we have not conducted experi-

ments in sodium-depleted, captopril-treated, lesioned rats that would detect more subtle signs of malaise, so this interpretation cannot be ruled out. Arterial blood pressure was measured in this study for two reasons. Lesions of circumventricular organs may affect the pressor response to intravenous ANG II by altering the central component of the response including increased sympathetic outflow and vasopressin secretion [20,26]. Also, a blockade of circulating ANG II during sodium deficiency might inhibit drinking responses nonspecifically by reducing blood pressure enough to compromise the behavioral competency of the rats [31,32]. In the present study, we did not find a significant difference in mean arterial pressure between the lesion treatments at any time before, during, or after the ANG II infusions. Therefore, differences in pressor responses could not account for the differences in drinking. Furthermore, the average mean arterial pressures of the depleted, captoprilblocked rats were always above 100 mmHg because of the large loads of water given by gavage. In a previous study with the same equipment [28], unlesioned rats that had been depleted of sodium and injected with 100 mg / kg captopril in a small volume had mean arterial pressures around 90 mmHg, which was not low enough to cause complete anuria or a large elevation of blood urea nitrogen [28]. We did not measure urine volume during the infusions in the present study, but we did observe urine stains in the bedding beneath the cages after the infusions. indicating that the large volumes of loaded water probably prevented anuria. Collectively, these facts argue against the interpretation that ANG II infusions restored blood pressure and the ability to respond behaviorally in the shamlesioned groups but not in the lesioned groups. We conclude that the reduction in intake by lesioned rats during intravenous infusions of ANG II results from a specific interference in the rat brain’s ability to detect circulating ANG II. Heart rate was measured and analyzed in this study in addition to arterial pressure because it was convenient to do so and not because we had any specific hypotheses about it. Nevertheless, this variable clearly distinguished the rats with lesions of the SFO or OVLT from those with control or sham lesions. Specifically, both circumventricular organ-lesioned groups developed a slight tachycardia at the end of the 10-min postinfusion period relative either to the rates of the control groups or to their own rates during the preinfusion baseline period. This may indicate some difficulty in maintaining arterial pressure after the abrupt withdrawal of the ANG II infusion in these lesioned animals resulting in a compensatory increase in heart rate. Water intake should decline as a result of a SFO lesion during an intravenous infusion of ANG II [26,27], but we know of no previous studies of water intake during such infusions using rats having small specific lesions of OVLT. The present data demonstrate that more than half of the rats with OVLT lesions drank no water at all (Table 1) and

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

the median latency to begin drinking water was longer in OVLT-lesioned rats than in either control group. The absolute intakes by the two lesioned groups were not quite significantly reduced because of the relatively low water intakes by the control groups, and this may have resulted in part from the large volumes of water given by gavage prior to the infusions. Thus, the data suggest that the OVLT may be required for a rapid water drinking response during an intravenous infusion of ANG II. However, a definitive answer to this question requires a separate study without the confounding influences of osmotic dilution, sodium depletion and salt intake.

Acknowledgements Supported by NIH grant NS-22274 to D.A.F.

References ˜ Saad, L.A. de Arruda ´ Almeida, V.R. Antunes, W. Abrao [1] N.A. Araujo Camargo, Effects of the alpha antagonists and agonists injected into the lateral hypothalamus on the water and sodium intake induced by angiotensin II injection into the subfornical organ, Brain Res. Bull. 48 (1999) 513–520. [2] D.B. Avrith, J.T. Fitzsimons, Increased sodium appetite in the rat induced by intracranial administration of components of the renin– angiotensin system, J. Physiol. (Lond.) 301 (1980) 349–364. [3] R.W. Bryant, A.N. Epstein, J.T. Fitzsimons, S.J. Fluharty, Arousal of a specific and persistent sodium appetite in the rat with continuous intracerebroventricular infusion of angiotensin II, J. Physiol. (Lond.) 301 (1980) 365–382. [4] E. Chiaraviglio, M.de F. Perez Guaita, Anterior third ventricle (A3V) lesions and homeostasis regulation, J. Physiol. (Paris) 79 (1984) 446–452. [5] D.S.A. Colombari, J.V. Menani, A.K. Johnson, Forebrain angiotensin type 1 receptors and parabrachial serotonin in the control of NaCl and water intake, Am. J. Physiol. 271 (1996) R1470–R1476. [6] A.D. Dalhouse, H.G. Langford, D. Walsh, T. Barnes, Angiotensin and salt appetite: physiological amounts of angiotensin given peripherally increase salt appetite in the rat, Behav. Neurosci. 100 (1986) 597–602. [7] M.D. Evered, Investigating the role of angiotensin II in thirst: interactions between arterial pressure and the control of drinking, Can. J. Physiol. Pharmacol. 70 (1992) 791–797. [8] M.D. Evered, M.M. Robinson, M.A. Richardson, Captopril given intracerebroventricularly, subcutaneously or by gavage inhibits angiotensin-converting enzyme activity in the rat brain, Eur. J. Pharmacol. 68 (1980) 443–449. [9] M.D. Evered, M.M. Robinson, P.A. Rose, Effect of arterial pressure on drinking and urinary responses to angiotensin II, Am. J. Physiol. 254 (1998) R67–R74. [10] A.L.R. Findlay, A.N. Epstein, Increased sodium intake is somehow induced in rats by intravenous angiotensin II, Horm. Behav. 14 (1980) 86–92. [11] D.A. Fitts, BPHR: A C-language program to analyze mean arterial blood pressure and heart rate from a digitized pressure recording, in: P. Wellman, B. Hoebel (Eds.), Ingestive Behavior Protocols, Society for the Study of Ingestive Behavior, New York, 1997, pp. 195–213. [12] D.A. Fitts, D.B. Masson, Preoptic angiotensin and salt appetite, Behav. Neurosci. 104 (1990) 643–650.

49

[13] D.A. Fitts, E.M. Starbuck, A. Ruhf, Circumventricular organs and ANG II-induced salt appetite: blood pressure and connectivity, Am. J. Physiol. 279 (2000) R2277–R2286. [14] D.A. Fitts, R.L. Thunhorst, Rapid elicitation of salt appetite by an intravenous infusion of angiotensin II in rats, Am. J. Physiol. 270 (1996) R1092–R1098. [15] D.A. Fitts, R.L. Thunhorst, J.B. Simpson, Diuresis and reduction of salt appetite by lateral ventricular infusions of atriopeptin II, Brain Res. 348 (1985) 118–124. [16] D.A. Fitts, D.S. Tjepkes, R.O. Bright, Salt appetite and lesions of the ventral part of the ventral median preoptic nucleus, Behav. Neurosci. 104 (1990) 818–827. [17] J.T. Fitzsimons, Angiotensin, thirst, and sodium appetite, Physiol. Rev. 78 (1998) 583–686. [18] J.T. Fitzsimons, B.J. Simons, The effect on drinking in the rat of intravenous infusions of angiotensin, given alone or in combination with other stimuli of thirst, J. Physiol. 203 (1969) 45–57. [19] J.A. Hosutt, N. Rowland, E.M. Stricker, Impaired drinking responses of rats with lesions of the subfornical organ, J. Comp. Physiol. Psychol. 95 (1981) 104–113. [20] M.L. Mangiapane, J.B. Simpson, Subfornical organ lesions reduce the pressor effect of systemic angiotensin II, Neuroendocrinology 31 (1980) 380–384. [21] M.J. McKinley, R.M. McAllen, F.A.O. Mendelsohn, A.M. Allen, S.Y. Chai, B.J. Oldfield, Circumventricular organs: neuroendocrine interfaces between the brain and the hemal milieu, Front. Neuroendocrinol. 11 (1990) 91–127. [22] K.D. Moe, M.L. Weiss, A.N. Epstein, Sodium appetite during captopril blockade of endogenous angiotensin II formation, Am. J. Physiol. 247 (1984) R356–R365. [23] R.R. Sakai, S.Y. Chow, A.N. Epstein, Peripheral angiotensin II is not the cause of sodium appetite in the rat, Appetite 15 (1990) 161–170. [24] R.R. Sakai, A.N. Epstein, Dependence of adrenalectomy-induced sodium appetite on the action of angiotensin II in the brain of the rat, Behav. Neurosci. 104 (1990) 167–176. ¨ [25] R.R. Sakai, S. Nicolaıdis, A.N. Epstein, Salt appetite is suppressed by interference with angiotensin II and aldosterone, Am. J. Physiol. 251 (1986) R762–R768. [26] J.B. Simpson, The circumventricular organs and the central actions of angiotensin, Neuroendocrinology 32 (1981) 248–256. [27] J.B. Simpson, A. Routtenberg, Subfornical organ lesions reduce intravenous angiotensin-induced drinking, Brain Res. 88 (1975) 154–161. [28] E.M. Starbuck, D.A. Fitts, Influence of salt intake, ANG II synthesis, and SFO lesion on thirst and blood pressure during sodium depletion, Appetite 31 (1999) 309–331. [29] E.M. Starbuck, D.A. Fitts, Influence of the subfornical organ on meal-associated drinking in rats, Am. J. Physiol. 280 (2001) R669– R685. [30] E.M. Starbuck, J.R. Lane, D.A. Fitts, Interaction of hydration and subfornical organ lesions in sodium-depletion induced salt appetite, Behav. Neurosci. 111 (1997) 206–213. [31] E.M. Stricker, The renin–angiotensin system and thirst: a reevaluation. II. Drinking elicited in rats by caval ligation or isoproterenol, J. Comp. Physiol. Psychol. 6 (1977) 1220–1231. [32] E.M. Stricker, The renin–angiotensin system and thirst: some unanswered questions, Fed. Proc., Fed. Am. Soc. Exp. Biol. 37 (1978) 2704–2710. [33] R.L. Thunhorst, K.J. Ehrlich, J.B. Simpson, Subfornical organ participates in salt appetite, Behav. Neurosci. 104 (1990) 637–642. [34] R.L. Thunhorst, D.A. Fitts, Peripheral angiotensin II causes salt appetite in the rat, Am. J. Physiol. 267 (1994) R171–R177. [35] R.L. Thunhorst, D.A. Fitts, J.B. Simpson, Separation of captoril effects on salt and water intake by subfornical organ lesions, Am. J. Physiol. 252 (1987) R409–R418. [36] R.L. Thunhorst, A.K. Johnson, Renin–angiotensin, arterial blood pressure, and salt appetite in rats, Am. J. Physiol. 266 (1994) R458–R465.

50

M. J. Morris et al. / Brain Research 949 (2002) 42 – 50

[37] R.L. Thunhorst, S.J. Lewis, A.K. Johnson, Role of arterial baroreceptor input on thirst and urinary responses to intracerebroventricular angiotensin II, Am. J. Physiol. 265 (1993) R591–R595. [38] E. Toth, J. Stelfox, S. Kaufman, Cardiac control of salt appetite, Am. J. Physiol. 252 (1987) R925–R929. ¨ ¨ T. Yukimura, W. Rascher, R.E. [39] T.A. Unger, D. Hubner, B. Schull, Lang, D. Ganten, Dissociation between in vivo and in vitro measurements of converting enzyme activity after chronic oral treatment with captopril in rats, Am. J. Cardiol. 49 (1982) 1530– 1532. [40] R.S. Weisinger, J.R. Blair-West, P. Burns, D.A. Denton, M.J. McKinley, E. Tarjan, The role of angiotensin II in ingestive behaviour: a brief review of angiotensin II, thirst and Na appetite, Regul. Pept. 66 (1996) 73–81. [41] R.S. Weisinger, J.R. Blair-West, P. Burns, D.A. Denton, E. Tarjan, Role of brain angiotensin in thirst and sodium appetite of rats, Peptides 18 (1997) 977–984.

[42] R.S. Weisinger, J.R. Blair-West, D.A. Denton, E. Tarjan, Na depletion-induced Na appetite of sheep is independent of brain angiotensin II, Physiol. Behav. 62 (1997) 43–51. [43] R.S. Weisinger, D.A. Denton, R. Di Nicolantonio, D.K. Hards, M.J. McKinley, P.J. Oldfield, P.J. Osborne, Subfornical organ lesion decreases sodium appetite in the sodium-depleted rat, Brain Res. 526 (1990) 23–30. [44] R.S. Weisinger, D.A. Denton, R. Di Nicolantonio, M.J. McKinley, A.F. Muller, E. Tarjan, Role of angiotensin in sodium appetite of sodium-deplete sheep, Am. J. Physiol. 253 (1987) R482–R488. [45] M.L. Weiss, K.E. Moe, A.N. Epstein, Interference with central actions of angiotensin II suppresses sodium appetite, Am. J. Physiol. 250 (1986) R250–R259. [46] Z.F. Yang, A.N. Epstein, Blood-borne and cerebral angiotensin and the genesis of salt intake, Horm. Behav. 25 (1991) 61–476.