572 W.E. JOHNSON JOURNAL AND C.R. OFPROPPER EXPERIMENTAL ZOOLOGY 286:572–584 (2000) Effects of Dehydration on Plasma Osmolality, Thirst-Related Behavior, and Plasma and Brain Angiotensin Concentrations in Couch’s Spadefoot Toad, Scaphiopus couchii WILLIAM E. JOHNSON* AND CATHERINE R. PROPPER Department of Biological Sciences, Northern Arizona University, Flagstaff, Arizona 86011 ABSTRACT Under dehydrating conditions, many terrestrial vertebrates species exhibit increases in plasma osmolality and their drinking behavior. Under some circumstances, this behavioral change is accompanied by changes in plasma and central angiotensin concentrations, and it has been proposed that these changes in angiotensin levels induce the thirst-related behaviors. In response to dehydration, the spadefoot toad, Scaphiopus couchii, exhibits thirst-related behavior in the form of cutaneous drinking. This behavior has been termed water absorption response (WR) behavior. Spadefoot toads live in harsh desert environments and are subject annually to dehydrating conditions that may induce thirst-related behavior. We tested the hypothesis that an increase in WR behavior is associated with both an increase in plasma osmolality and an increase in plasma and brain angiotensin concentrations. First, we determined the degree of dehydration that was necessary to initiate WR behavior. Animals dehydrated to 85% of their standard bladder-empty weight via deprivation of water exhibited WR behavior more frequently than control toads left in home containers with water available. Next, using the same dehydration methods, we determined the plasma osmolality and sodium concentrations of dehydrated toads. Toads dehydrated to 85% standard weight also had a significant increase in plasma osmolality, but exhibited no overall change in plasma sodium concentrations, indicating that while an overall increase in plasma osmolality appears to be associated with WR behavior in S. couchii, changes in sodium concentrations alone are not sufficient to induce the behavior. Finally, plasma and brain angiotensin concentrations were measured in control toads and toads dehydrated to 85% standard weight. Plasma and brain angiotensin concentrations did not increase in dehydrated toads, indicating that dehydration-induced WR behavior that is associated with changes in plasma osmolality may not be induced by changes in endogenous angiotensin concentrations in S. couchii. J. Exp. Zool. 286:572–584, 2000. © 2000 Wiley-Liss, Inc. Because of their relatively permeable skin, anuran amphibians can become dehydrated faster than other terrestrial vertebrates (Bentley, ’66). Accordingly, anurans adapted to xeric environments have mechanisms to limit the impact of living in a dehydrating environment. One adaptive mechanism involves seasonal estivation. For most of the year, desert anurans remain underground and avoid the aboveground dehydrating conditions (Shoemaker et al., ’69). Also, desert anurans can store water in their bladders and reabsorb the water while underground (Ruibal, ’62; McClanahan, ’72). In addition, some species including the desert toad, Couch’s spadefoot toad (Scaphiopus couchii), can increase urea production in response to drying soil, increasing the plasma osmolality (McClanahan, ’72), which spadefoot toads utilize © 2000 WILEY-LISS, INC. to equilibrate with soil water potential (Shoemaker et al., ’69). As the ambient temperature rises during late spring, soil water potential diminishes, and S. couchii may experience net water loss and an increase in plasma osmolality (Shoemaker et al., ’69). Upon arrival of the first heavy summer monsoon rains, S. couchii emerge from their burrows and seek ephemeral ponds for reproduction. With the initial light rains, emergent toads have been observed rehydrating on moist substrates, utilizGrant sponsor: National Science Foundation; Grant number: IBN 93-10352. *Correspondence to: William E. Johnson, Box 5640, Dept. of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011. E-mail: email@example.com Received 1 December 1998; Accepted 27 September 1999 THIRST AND ANGIOTENSIN IN S. COUCHII ing a form of cutaneous drinking described as water absorption response (WR) behavior (Stille, ’58; Dimmitt and Ruibal, ’80; Brekke et al., ’91; Propper and Johnson, ’94). By pressing their splayed hind limbs against the moist surface, dehydrated S. couchii can regain rapidly the lost water weight lost (Claussen, ’69; Propper and Johnson, ’94; Johnson et al., ’95). Previous laboratory studies have demonstrated WR behavior in several species of desert-adapted anurans (Brekke et al., ’91; Propper and Johnson, ’94). Propper and Johnson (’94) determined that under laboratory conditions, S. couchii dehydrated for 5 hr exhibited consistent WR behavior associated with an average of 10% body weight loss for the group. These results showed that dehydration alone is sufficient to elicit WR behavior, but that the degree of dehydration necessary to elicit this behavior, as measured by changes in osmolality, remains undefined. One purpose of this study is to quantify the association between the onset of WR behavior and changes in plasma osmolality in dehydrated S. couchii. In addition, sodium concentrations were compared between hydrated and dehydrated toads, in order to test the hypothesis that an increase in WR behavior in dehydrated S. couchii is associated with an increases in plasma sodium concentrations. Increases in sodium concentrations have been associated with increases in thirst-related behavior in other vertebrates (Fitzsimons, ’79). One potential neuroendocrine mediator of WR behavior is the hormone angiotensin II (A-II). Angiotensin II has been shown to induce thirst-related behavior in many taxa of vertebrates (Rolls and Rolls, ’82; Kobayashi et al., ’79). Peripherally injected angiotensin II increases drinking behavior in teleosts (Hirano and Hasegawa, ’84; Perrott et al., ’92), elasmobranchs (Hazon et al., ’89), reptiles (Fitzsimons and Kaufman, ’77), birds (Snapir et al., ’76; Fitzsimons, ’79, ’98), and mammals (Fitzsimons, ’79, ’98). In addition, A-II injected intraperitoneally also stimulates WR behavior in two families of anurans (Hoff and Hillyard, ’91; Tran et al., ’92; Propper and Johnson, ’94). The fact that peripherally administered A-II can stimulate robust drinking behavior in many different species supports the view that endogenous plasma angiotensin, released under dehydrating conditions, may play a role in thirst-related behavior in vertebrates. A substantial body of correlative evidence links increases in plasma osmolality with both an increase in plasma A-II and thirst-related behavior. Dehydration-induced increases in serum osmola- 573 lity, plasma A-II, and plasma sodium concentrations are observed in mammals (Mann et al., ’80; Yamaguchi, ’81) and birds (Gray and Simon, ’85; Gray and Erasmus, ’89). Increased plasma concentrations of A-II are also associated with increased thirst-related behavior in mammals (Johnson et al., ’84), birds (Gray and Erasmus, ’88), and fish (Carrick and Balment, ’83; Perrott et al., ’92; Tierney et al., ’95). These correlative data suggest that induction of thirst-related behavior in vertebrates may be caused by increases in plasma osmolality and/or in plasma A-II concentrations. There is some preliminary evidence for a peripheral, circulating renin-angiotensin system in anurans. The presence of angiotensinogen and angiotensin-converting enzyme activity has been established in the plasma of two species of bufonid amphibians (Nolly and Fasciolo, ’71a, b; FernandezPardal et al., ’86). In addition, the A-II antagonist saralasin attenuates WR behavior in Bufo punctatus (Hoff and Hillyard, ’91), but not in S. couchii or Bufo cognatus (see Propper and Johnson, ’94). The results from these studies suggest that endogenous plasma A-II may have a role in thirst-related behavior in some species of anurans. In order to determine the relationship between dehydration, plasma A-II and thirst-related behavior in amphibians, we examined the effects of dehydration on plasma osmolality and A-II concentrations in S. couchii in the laboratory. There is also evidence that a central renin-angiotensin system may be associated with WR behavior in S. couchii. Angiotensin II injected intracerebroventricularly (icv) induced WR behavior at a 1000-fold less dose than that injected ip (10 ng/100g-animal vs. 10 µg/100 g-animal: Propper and Johnson, ’94; Propper et al., ’95). Angiotensin injected icv can also induce thirst-related behavior in mammals (Epstein et al., ’70; Wright et al., ’88), reptiles (Fitzsimons and Kaufman, ’77), and birds (Wada et al., ’75; Snapir et al., ’76; Takei, ’77), suggesting that regulation of thirst by a central renin-angiotensin system is an evolutionarily conserved phenomenon among vertebrates. Because both dehydration and A-II injected icv can stimulate drinking in rats and other species (Fitzsimons, ’79), many studies have explored the possibility that concentrations of central angiotensin increase in response to dehydration (reviewed by Saavedra, ’92). Dehydration, via 48-hr water deprivation, induced an increased release of angiotensins from rat paraventricular nuclei (Harding et al., ’92). However, Yamaguchi (’81) and 574 W.E. JOHNSON AND C.R. PROPPER Yamaguchi et al. (’81) found no changes in hypothalamic tissue A-II concentrations in water-deprived or nephrectomized rats, compared to water-replete rats. Because of the limited number of studies, the association between brain angiotensin concentrations and thirst-related behavior is not well defined in any class of vertebrates, including Amphibia. Therefore, a second goal of this study was to test the hypothesis that changes of brain angiotensin concentrations occur in S. couchii exposed to dehydrating conditions previously shown to induce WR behavior. MATERIALS AND METHODS Collection and care Male S. couchii were collected from the Buenos Aires National Wildlife Refuge in Pima County, AZ during the monsoon season in July, 1994 and July, 1995. Following transport to Northern Arizona University, all toads were housed four to five per container (53 × 41 × 18 cm) filled halfway with moistened soil. Water was provided in 10-cm water dishes, and was changed three times weekly. Toads were fed waxworms and crickets three times a week, and were maintained on a seasonal photoperiod in a temperature-controlled room maintained at 22°C. All experiments were conducted in a room maintained at a constant 22°C. Experiment 1: WR behavior This experiment was conducted in September of 1994 on toads collected in July, 1994. In order to insure that all toads were fully hydrated initially, ten male S. couchii were hydrated for 3 hr in 0.5 liter of water in a 40-liter aquarium, enough to cover the bottom of the aquarium and the ventral surface of the toads, then returned to their home containers. Twenty-four hours later and before the toad was weighed, we emptied each toad’s bladder by cannulation with a micropipette tip inserted into its cloaca. This weight is referred to as the toad’s standard weight (SW; Ruibal,’62). Toads were then randomly assigned to one of four treatments: toads were placed in a dry 40liter container then allowed to dehydrate to 95%, 90%, 85%, and 100% SW (rate of dehydration for all treatments: 0.63 ± 0.03 g/hr). Control toads (100% SW) were left in their home container with water available ad libitum. Each toad received each treatment once in random order every 4 days. The experiment room was kept at 22°C and 20– 25% relative humidity. Air flow was kept at a recommended level of 10 air changes per hour, ac- cording to Guide to the Care and Use of Laboratory Animals (National Research Council, ’96). At the end of the treatment period, toads were placed in a petri dish filled with 25 ml dechlorinated tap water. Toads were scored for behavior by the method established by Propper and Johnson (’94): toads were scored for WR behavior once every 10 min for 2 hr, with a 1 given for each observation period where WR behavior was exhibited (ventral seat patch pressed against wet substrate), and a 0 for each observation period where no WR behavior was observed. The scores were summed for the 2-hr test period for each toad’s four individual treatments. The combined scores were analyzed by Friedman’s repeated measures ANOVA by ranks, and post hoc comparisons were determined by utilizing the multiple comparison test in Chapter 7 of Siegel and Castellan (’88). Experiment 2: plasma osmolality and angiotensin This experiment was conducted in June 1995 on toads captured in July 1994. Twenty-nine male toads were prehydrated as above before determination of their standard weight on the treatment day. Each toad was then assigned to one of the four treatments as described above in experiment 1, for a total of seven toads per treatment group (except for eight toads in 85% SW group). We could not conduct this study concurrently on the same toads as in experiment 1, because removal of blood by heart puncture would produce a hypovolemic condition which might affect subsequent WR behavior (Taylor, Mayer, and Propper, 1999). After treatment, a 200- to 500-µl blood sample was removed from each toad by heart puncture with a heparin-coated 1-cc syringe, then prepared for radioimmunoassay as described below. The sample was transferred to a chilled (4°C) 1.5-µl polypropylene microcentrifuge tube containing 7.5 µl of a peptidase inhibitor cocktail of the following composition: 4 µg/ml enalpralat maleate (Sigma), 440 µg/ml 1,10 phenanthroline (Sigma), 40 µg/ml APMSF (Sigma), 1 µg/ml pepstatin (Sigma), 0.1%-mercaptoethanol (Sigma), and 6.25 mM EDTA (Ledwith et al., ’93). After centrifugation at 2000 g for 5 min, the plasma was decanted into a cryogenic tube for future peptide analysis. A 10-µl sample of plasma was removed at this time for measurement of osmolality with a Wescor 5500 osmometer. In a pilot study, no effect was noted for the peptidase inhibitor in these measurements (Johnson, unpublished data). Differ- THIRST AND ANGIOTENSIN IN S. COUCHII 575 ences in osmolalities among treatment groups were determined by one-way ANOVA, and post hoc comparisons were determined with StudentNeuman-Keuls multiple comparison procedure. at –20°C until protein determination. Protein concentrations were measured using a modified Bradford assay (Bradford, ’76) previously utilized by Propper and Moore (’91). Experiment 3: plasma osmolality, sodium and angiotensin, and brain angiotensin Radioimmunoassay of plasma and brain angiotensin This experiment was conducted in October–November, 1995 from toads collected in July, 1995. After prehydration and bladder cannulation as described in experiment 1, toads were either dehydrated to 85% SW (n = 10), or left to hydrate (n = 9) in 0.5 l of dechlorinated tap water in a 40-l aquarium for an equivalent length of time. Toads were then sacrificed by decapitation. Trunk blood was collected and treated with peptidase inhibitors as described previously. Plasma osmolality was measured on a Wescor 5500 osmometer, and sodium concentrations were determined by flame spectrophotometry. Brains were dissected and removed intact within 7 to 11 min of decapitation. Brains were immediately placed in Tissue Tek O.C.T. compound and frozen on dry ice. Frozen brains were then placed in cryogenic storage tubes and stored at –80°C. Cross reactivity was measured for angiotensin I, angiotensin III (A-III), and arginine vasotocin (Fig. 1A). Since angiotensin III cross-reacted 100% with the antiserum, results were reported as immunoreactive angiotensin (ir-ANG) concentrations. Plasma procedures were carried out on ice (4°C). Unextracted plasma was used because of small volumes of plasma from some animals (150–500 µl). Also, Mann, Johnson, and Ganten (’80) used unextracted plasma when measuring dipsogenic levels of A-II in rats, and Lijnen et al. (’78) found significant correlation between extracted and unextracted A-II concentrations. Parallelism was determined by comparing standard curves to serial dilutions of unextracted plasma by analysis of covariance (Fig. 1B; Dayton, ’70). There was homogeneity of regression between curves (FANCOVA 1,7 = 3.33, P > 0.10). Sixty microliters of unextracted plasma was brought up to 100 µl with the assay buffer, PEgel buffer (10 mM potassium phosphate, 1 mM EDTA, 0.1% gelatin, pH = 7.3). Standard curves were obtained by adding 100 µl of 5Val-A-II (0.25– 1000 pg; BACHEM California, Inc., Torrance, CA) in assay buffer, 100 µl of the antiserum 1:75,000 in assay buffer (RAS 7002N for experiment 2; for experiment 1 and 3, IHC-7002; Peninsula Laboratories, Inc., Belmont, CA), and 50 µl of 7000– 8000 cpm of labeled tracer (3-[125I]-(iodotyrosyl4) angiotensin II (5-leucine), Amersham Corp., Arlington Heights, IL). All standards were run in duplicates. Tubes were incubated 18 hr at 4°C. One ml of dextran-coated charcoal (1 g NORIT-A charcoal with 0.02 g Dextran [Sigma, St. Louis, MO] in 100 ml PE-Gel assay buffer) was pipetted into each tube and tubes were immediately centrifuged at 3000 rpm for 10 min. Supernatants were counted on a Beckman 5500B gamma counter. The minimum detection limit ranged from 0.39 to 1.2 pg. The intraassay coefficient of variation was 9.0 ± 1.3%, and the interassay coefficient of variation was 21 ± 0.8% for 2 pg and 20 pg replicates on five assays total with both antisera. For determination of parallelism in unextracted brain homegenate, whole brain homogenate was split into two samples, with one sample spiked Tissue preparation Brains were frozen to cryostat chucks with Tissue Tek O.C.T. compound then sectioned in 200µm thick transverse sections in a Leica cryostat kept at –17°C. Sections were thaw-mounted then frozen to microscope slides and stored at –80°C in an airtight slide box until microdissection. Eight areas of the toad brain were identified using descriptions from Northcutt and Royce (’75) and Neary and Northcutt (’83). All brain areas were removed by the micropunch technique described in Palkovitz and Brownstein (’82) and Zoeller and Moore (’85). Using a modified stainless-steel hypodermic 26-gauge needle, five to seven punches were made within each brain area and placed in a drop of 1 M acetic acid on a chilled 4°C spatula. The punches were then transferred to a chilled tube containing 250 µl of 1 M acetic acid. The brain punches were then homogenized using a Branson 450 sonifier with a double step microtip for 2 sec. (Output 1 for 10 % cycle 10 watts). The homogenized brain solutions were centrifuged at 11,000g for 20 min, and from each sample 200 µl of the supernatant was lyophilyzed and stored dessicated at –80°C until assayed for angiotensin by RIA. The remaining 50 µl was dried under airjet in a 37°C water bath and stored 576 W.E. JOHNSON AND C.R. PROPPER with 1 pg/µl A-II and one sample left unspiked. Each sample was serially diluted, then run in RIA standardization. Parallelism was determined by comparing regressions of serial dilutions of whole brain homogenates with standard curves by analysis of covariance (Dayton, ’70). There was no difference between slopes (Fig. 2; FANCOVA 1,11 = 0.40, P > 0.5), indicating that there were no extraneous agents in the unextracted brain homogenate. Lyophilyzed brain homogenates were reconstituted in 100 µl PE-gel assay buffer (10 mM potassium phosphate, 1 mM EDTA, 0.1% gelatin, pH = 7.3). Standard curves were obtained by similar methods used in determining plasma standards. The minimum detection limit ranged from 0.39 to 1.2 pg. The intraassay coefficient of variation was 7.6%, and the interassay coefficient of variation was 12.5%. As noted earlier, the antiserum cross-reacts with A-III 100% (Peninsula Laboratories, personal communication) and the results are reported as immunoreactive angiotensin (ir-ANG). Eight brain areas from each toad were sampled in the experiment. All results were expressed as pg angiotensin/µg protein. Because many samples had protein concentrations below detectable limits, the angiotensin concentrations of five of the brain areas were not included due to insufficient sample size for statistical analysis. Fig. 1. (A) Competitive binding curves of angiotensins I, III, and arginine vasotocin against the standard curve for AII. Angiotensin II and III competed similarly with labelled AII for the antisera; the other peptides compete at much higher concentrations. (B) Parallelism among regression lines of serial dilutions (60, 30, 15, and 7.5 µl) of two plasma samples from different toads compared to the standard curve for A-II, determined by analysis of covariance (Dayton, ’70). There was no difference among slopes of the plasma samples and the standard curve (FANCOVA 1,7 = 3.33, P > 0.10). Fig. 2. Serial dilutions of whole brain homegenates run in RIA against standard curves. There was no difference among slopes (FANCOVA 1,11 = 0.40, P > 0.5). THIRST AND ANGIOTENSIN IN S. COUCHII All results, including osmolalities, sodium concentrations, plasma angiotensin concentrations, and brain angiotensin concentrations, were analyzed by Student’s t-test for independent samples. Angiotensin concentrations determined to be outliers by statistical methods described in Sokol and Rohlf (’81) were removed from the analysis. 577 RESULTS Experiment 1: WR behavior There was a significant effect of dehydration on WR behavior in S. couchii (Fig. 4A). Toads dehydrated to 85% SW showed significantly more WR behavior than 0% SW toads (χ2= 13.82, P = 0.003). Angiotensin III and WR behavior Because A-III can comprise 10–20% of the circulating angiotensin in mammals (Semple et al., ’76; Hanssens et al., ’90), and our assay could not distinguish between A-II and A-III, a repeated measures ANOVA experiment was designed to examine the induction of WR behavior by A-III. On the first experimental day, hydrated, bladder-emptied toads were injected intraperitoneally with either 0, 10, and 100 µg A-III/100 g-wt toad. Toads were scored for the total number of 10 min periods spent in WR behavior after each treatment (see experiment 1). The experiment was repeated twice four days apart so each toad received each treatment in random order. Results were analyzed by Friedman’s repeated measures ANOVA by ranks. The median number of time periods all toads spent in WR behavior after each treatment is shown in Figure 3. Angiotensin III did not induce an increase in WR behavior in hydrated toads (P = 0.91). Fig. 3. Angiotensin III injected intraperitoneally at 0, 10, or 100 µg/100 g-wt toad in hydrated toads did not induce an increase in WR behavior. Vertical lines through bar graphs represent intraquartile ranges for WR behavior in the three treatments. Figure 4. 578 W.E. JOHNSON AND C.R. PROPPER According to the results of the post hoc test, the 90% and 95% SW groups were not different from either the 85% or 100% SW groups. To determine whether toads showing WR behavior were more effective in rehydration than toads not showing the behavior, a Spearman’s correlation analysis was used. The toads that spent more time in WR behavior gained more weight (Fig. 4B; r = 0.65, P < 0.005). Experiment 2: plasma osmolality and angiotensin from heart puncture There was a significant effect of dehydration on the plasma osmolalities of the toads (Fig. 5A). When dehydrated to 85% SW, toads had significantly higher plasma osmolalities than when left in home containers (F3,6 = 3.51, P = 0.031). According to the results of the post hoc test, the 90% and 95% SW groups were not different from either the 85% or 100% SW groups. There was no change in plasma angiotensin concentrations with dehydration treatment (Fig. 5B; F3,26 = 0.56, P = 0.65). There was also no correlation between angiotensin concentrations and plasma osmolality (Fig. 6; Pearson’s r = 0.174, P = 0.396), indicating that increases in plasma osmolality do not appear associated with circulating angiotensin concentrations. Experiment 3: plasma osmolality, plasma sodium concentrations, and plasma angiotensin concentrations from trunk blood; brain angiotensin As in experiment 2, toads dehydrated to 85% SW had significant increases in plasma osmolality (Fig. 7A; t = 3.51, P = 0.003). However, the overall increase in osmolality was not augmented by a concurrent increase in plasma sodium concentrations (Fig. 7A; t = 0.136, P = 0.193). Fig. 4. (A) Experiment 1: Bar graphs represent median (± intraquartile ranges) amount of time periods spent by S. couchii in water absorption response (WR) behavior after four treatments: dehydration to 95%, 90%, or 85% standard weight (SW), or kept in a container as a control (100% SW). Bars with different letters above them are significantly different. Toads dehydrated to 85% SW spent more time in WR behavior than control toads. (B) Percent weight change during WR behavior test plotted against median WR behavior score for each toad in all treatments. The different dehydration treatment groups are represented by different symbols. Results from the correlation test indicated that toads regained most, if not all, of the water that was lost during dehydration treatment, and toads that spent more time in WR behavior gained more weight. Fig. 5. (A) Experiment 2: Bar graphs represent mean (± SEM) plasma osmolalities for toads dehydrated to 95% SW, 90% SW, or 85% SW, or kept in a container as a control (100% SW) as in the behavioral experiment. Bars with different letters above them are significantly different. At 85% SW loss, toads had higher plasma osmolalities than toads in 100% SW treatment but not different from either 95% or 90% SW. Error bars are standard error of the mean. (B) Bar graphs represent mean (± SEM) plasma angiotensin (ir-ANG) concentration from the toads’ blood collected by heart puncture. There was no difference in ir-ANG concentrations among the SW groups. THIRST AND ANGIOTENSIN IN S. COUCHII 579 Fig. 6. Experiment 2: Plasma osmolalities versus plasma ir-ANG concentrations. There was no significant correlation between the variables, indicating that there was no association between dehydration-induced increases in plasma osmolalities and plasma ir-ANG concentrations. As with experiment 2, plasma angiotensin concentrations from trunk blood did not increase with dehydration (Fig. 7B; t = 0.72, P = 0.49), and there was no correlation between plasma osmolality and angiotensin concentrations (Fig. 8A; r = –0.22, P = 0.40). There was no difference between the dehydrated and hydrated groups (Fig. 8B) in ir-ANG concentrations of the anterior pituitary (T = 76, P = 0.74), posterior pituitary (t = 0.94, P = 0.37), and preoptic area (t = 0.89, P = 0.40). DISCUSSION Plasma osmolality and WR behavior This study demonstrates that dehydration to 85% SW is sufficient to consistently induce WR behavior in S. couchii. These results concur with the previous results with S. couchii by Propper and Johnson (’94), except that in the 1994 study toads were not prehydrated or bladder cannulated. Since the toads in the 1994 study showed WR behavior at 10% mass loss, some toads may have started dehydration treatment already at a water deficit. Brekke et al. (’91) and Tran et al. (’92) also demonstrated that dehydrated bladder-emptied Bufo punctatus and Bufo woodhouseii exhibit WR behav- Fig. 7. (A) Experiment 3: Bar graphs indicate mean plasma osmolalities (± SEM for total osmolality) for toads dehydrated to 85% SW or hydrated (100% SW). Toads dehydrated to 85% SW had significantly higher plasma osmolalities than toads in 100% SW treatment. Plasma sodium concentrations (shaded areas) were not different between the two treatments. (B) Bar graphs represent mean (± SEM) plasma ir-ANG concentrations from the toads’ trunk blood. There was no difference in plasma ir-ANG concentrations between the 100% SW and 85% SW groups. 580 W.E. JOHNSON AND C.R. PROPPER Fig. 8. (A) Experiment 3: Plasma ir-ANG concentrations plotted against plasma osmolalities. There was no correlation between plasma osmolalities and plasma ir-ANG concentrations. (B) Bar graphs represent mean (± SEM) ir-ANG concentrations in three brain areas from dehydrated (85% SW) or hydrated (100% SW) toads. There was no effect of dehydration to 85% SW on ir-ANG concentrations in the anterior pituitary, posterior pituitary, and preoptic area. Numbers in parentheses above bar graphs indicate sample sizes. ior at varied levels of dehydration. Plasma osmolality was not measured in Propper and Johnson (’94) or Brekke et al. (’91), therefore it is difficult to determine the dehydration level for toads in those experiments. Though plasma osmolality was not measured in toads exhibiting WR behavior (experiment 1), dehydration to 85% SW resulted in an increase in plasma osmolality for toads in experiments 2 and 3. Considered together, these findings support the hypothesis that an increase in thirst-related behavior is associated with an increase in plasma osmolality in S. couchii. In other vertebrate taxa, increases in plasma osmolality are also associated with increases in thirst-related behavior (Fitzsimons and Kaufman, ’77; Fitzsimons, ’79; Carrick and Balment, ’83; Perrott et al., ’92). Thus, an increase in plasma osmolality appears to be associated with increased drinking behavior in vertebrates in general. Because sodium comprises a large portion of the total plasma solute concentration (McClanahan, ’67, ’72), changes in plasma osmolality may greatly affect sodium concentrations and thereby induce thirst-related behavior. The proposed mammalian thirst-related osmoreceptors may be sodium-sensitive (Andersson and Westbye, ’70). Rats given icv injections of hypertonic saline show increased drinking of both water and saline solutions (Buggy, ’77). Emergent S. couchii have elevated plasma sodium levels compared to foraging and breeding pond toads (McClanahan, ’67), thus it is possible that plasma sodium concentration changes may be associated with the induction of WR behavior in S. couchii. Although spadefoot toads injected ip with hypertonic saline showed WR behavior more often than toads injected with isotonic saline (Taylor, Mayer, and Propper, ’99), plasma osmolality did not increase (Mayer and Propper, 2000). In addition, plasma sodium levels in our study did not increase significantly when toads were dehydrated to 85% SW. In light of this recent evidence, sodium concentration changes in peripheral circulation may not be involved in the induction of WR behavior in S. couchii. Alternatively, overall osmolality may be more important in the induction of WR behavior rather than any one solute. Injections of hypertonic sucrose solutions induced WR behavior in S. couchii, an indication that induction of WR behavior may be associated with cellular dehydration (Taylor et al., ’99). Toads in our study had overall increases in osmolality, although when considered separately, sodium concentrations did not increase. Therefore, increases in osmolality may be due to increased concentrations of other osmolytes. In addition, hemorrhaging, a form of extracellular dehydration, also induced WR behavior in S. couchii (Taylor et al., ’99), indicating that changes in plasma volume may have significant impact on WR behavior as well. THIRST AND ANGIOTENSIN IN S. COUCHII Another factor that underscores the complexity of thirst-related behavior in anurans is the observation that toads become opportunistic, anticipatory drinkers in response to regular exposure to dehydrating conditions. Jorgensen (’94) found that Bufo bufo maintained a steady hydrated state when water was available. How anticipatory drinking is related to the water deprivation-induced WR behavior is not clear. Therefore, as with other vertebrates, regulation of thirst-related behavior in S. couchii probably involves a combination of factors including overall osmolality and volume level, rather than any one factor. Angiotensin and WR behavior The results from our present studies indicate that although dehydration causes an increase of plasma osmolality in S. couchii, there is no correlation between changes in osmolalities and concentrations of plasma angiotensin. One possibility is that endogenous plasma angiotensin alone may not be a stimulus for WR behavior in S. couchii. Nolly and Fasciolo (’71b) found that in the toad, Bufo arenarum, acclimation to hypertonic NaCl solutions increased plasma osmolalities but caused little change in plasma renin activity. In one study in rats (Abdelaal et al., ’76), subcutaneous injection of hypertonic saline induced thirst but caused no change in plasma angiotensin. In addition, plasma A-II concentrations did not increase in S. couchii injected ip with hypertonic saline (Mayer and Propper, 2000). Our initial studies have shown that dehydration to a threshold of 85% SW mass loss is sufficient to induce WR behavior and an increase in osmolality, therefore, in S. couchii, osmotically-induced thirst-related behavior may not be associated with changes in plasma angiotensin concentrations that may invoke thirst. Another alternative hypothesis to consider is that plasma angiotensin concentrations may increase to concentrations sufficient to induce thirst, then rapidly diminish before onset of observed behavior. Hoff and Hillyard (’91) have noted that the time course of angiotensin activity in toads is largely undefined. In our present study, angiotensin II concentrations did not increase with 85% SW dehydration, but an A-II signal to induce thirst may have already diminished by sample collection time. Angiotensin is a relatively fast-acting compound; four classes of vertebrates respond to angiotensin injections with drinking after 5– 15 min (reviewed by Fitzsimons, ’79; Mann et al., ’80). Angiotensin II injected ip induces thirst-re- 581 lated behavior in S. couchii within 15 min (Propper and Johnson, ’94). There are also reported differences in the length of time necessary to dehydrate rats in order to produce dipsogenic concentrations of angiotensin (12 vs. 48 hr; Mann et al., ’80; Van Eekelen and Phillips, ’88). A time course study with increased sample sizes could address the timing in dehydration-induced angiotensin release. The lack of dehydration-induced angiotensin response in S. couchii may provide some evidence why both saralasin and captopril had no effect in attenuating WR behavior in S. couchii (Propper and Johnson, ’94; Propper, unpublished data). In another anuran species, Bufo arenarum, saralasin did not block hydrosomotic water uptake responses (Reboreda and Segura, ’89). However, dehydrated Bufo punctatus respond to saralasin, a nonselective angiotensin antagonist, by decreasing WR behavior (Hoff and Hillyard, ’91). The desert dwelling toad B. punctatus also had a 10fold greater sensitivity to exogenously administered A-II in the induction of WR behavior than S. couchii, and a 100-fold greater sensitivity than Bufo cognatus (Hoff and Hillyard, ’91; Propper and Johnson, ’94). The WR response for B. punctatus is also at a lower mass loss threshold than for S. couchii (Brekke et al., ’91). Hoff and Hillyard (’91) and Hillyard et al. (’98) have suggested that the interspecific difference in WR behavior may be due to adaptation to environments where water availability differs. Alternatively, there may be a central A-II response involved with thirst-related behavior in S. couchii and other anurans. The colocalization of A-II and atrial natriuretic hormone (ANH) receptors in several brain regions of Xenopus laevis (Kloas and Hanke, ’92) supports the view that AII may have central effects in anuran fluid homeostasis. Although dehydration to 85% SW in the laboratory did not increase ir-ANG concentrations significantly in the brain regions (anterior and posterior pituitary, preoptic area) examined in this study, increases in ir-ANG concentrations in the periventricular nuclei of the hypothalamus have been found in toads exhibiting WR behavior in the field (Johnson, Propper, and Hillyard, ’98). In addition, angiotensin receptors are present in the ventral hypothalamus of S. couchii near ventricular areas (Propper and Hillyard, ’96). In rats, both drinking behavior and A-II concentrations in the hypothalamus increased significantly after hemorrhage (Phillips, Heininger, and Toffolo, ’96). The combined evidence supports the possibility of 582 W.E. JOHNSON AND C.R. PROPPER a central renin-angiotensin thirst-related system in S. couchii. Similar to our peripheral A-II study, where toads were sampled only once after dehydration, changes in thirst-related central angiotensin could have occurred prior to sampling. Harding et al. (’92) found significant amounts of angiotensin released over several hours from the paraventricular nuclei of water-deprived rats. In rats, a single bolus of angiotensin infused icv has been estimated to have a half-life of approximately 30 sec before degradation by peptidases (Harding et al., ’86). Therefore, central angiotensin in amphibians may be produced, bound, and degraded too rapidly for a single measurement in an individual animal to accurately reflect prior changes in angiotensin concentrations related to the elicited thirst-related behavior. A third possibility is that toads were not dehydrated sufficiently to produce an increase in central angiotensin concentrations that would invoke thirst, especially if the thirst-related behavioral response is elicited by hypovolemia. Spadefoot toads dehydrated to 70% SW show significant decreases in blood pressure (Hillman, ’80). Central angiotensin has been demonstrated in mammals to play a simultaneous role in blood pressure regulation and in thirst behavior (Unger et al., ’88; Evered, ’91). In support of this hypothesis for anurans, microinjection of the hypotensive agent αadrenergic antagonist phenoxybenzamine (POB) into the midbrain tegmentum of the toad B. arenarum increases water uptake (Segura et al., ’82). Therefore, if a thirst-related response in S. couchii is associated with central angiotensin control of blood pressure, then a central angiotensin response may not occur until the toad is exposed to severe dehydration and resultant hypotension. In summary, S. couchii exhibits an increase in plasma osmolality and in thirst-related behavior when dehydrated to 85% SW in the laboratory, but does not show a concurrent increase in plasma angiotensin during the sampling intervals used in these experiments. Immunoreactive angiotensin can be found in the brain of S. couchii, but concentrations did not increase in the brain areas examined after exposure to dehydrating conditions in the laboratory. Therefore, the thirstrelated behavior induced in S. couchii in these studies is probably not associated with an increase in plasma angiotensin concentrations or in changes in brain angiotensin concentrations, but may be induced by an increase in plasma osmolality. ACKNOWLEDGMENTS We would like to thank Stan Hillyard of UNLV for his field help and intellectual guidance; Loretta Mayer, Randy Wade, and Lisa Rania Gibbons for laboratory assistance and animal collection help; and Dr. Robert Manicke of the University of Maryland for statistical advice. We would like to especially thank the staff at the Buenos Aires National Wildlife Refuge for our housing, collecting permits, and general assistance. LITERATURE CITED Abdelaal AE, Mercer PF, Mogenson GJ. 1976. Plasma angiotensin II levels and water intake following β-adrenergic stimulation, hypovolemia, cellular dehydration and water deprivation. Pharmac Biochem Behav 4:317–321. Andersson B, Westbye O. 1970. Synergistic action of sodium and angiotensin on brain mechanisms controlling fluid balance. Life Sci 9:601–608. Bentley PJ. 1966. Adaptations of amphibia to arid environments. Science 152:619–623. Bradford MM. 1976. A rapid and sensitive method for for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248–254. Brekke DR, Hillyard SD, Winokur RM. 1991. Behavior associated with the water absorption response by the toad, Bufo punctatus. Copeia 1991:393–401. Buggy J. 1977. Drinking elicited by angiotensin or hyperosmotic stimulation of the rat antero-ventral third ventricle: single or separate neural substrates? In: Buckley JP, Ferrano CM, editors. Central actions of angiotensin and related hormones. New York: Pergamon Press, Inc. p 315–319. Carrick S, Balment RJ. 1983. The renin-angiotensin system and drinking in the euryhaline flounder Platichthys flesus. Gen Comp Endocrinol 51:423–433. Claussen DL. 1969. Studies of water loss and rehydration in anurans. Phys Zool 42:1–14. Dayton CM. 1970. The analysis of covariance. In: The design of educational experiments. New York: McGraw-Hill. p 304–337. Dimmitt MA, Ruibal R. 1980. Environmental correlates of emergence in spadefoot toads (Scaphiopus). J Herpetol 14:21–29. Epstein AN, Fitzsimons JT, Rolls BJ. 1970. Drinking induced by injection of angiotensin into the brain of the rat. J Physiol London 210:457–474. Evered MD. 1991. Investigating the role of angiotensin II in thirst: interactions between arterial pressure and the control of drinking. Can J Physiol Pharmacol 70:791–797. Fernandez-Pardal J, Perseguino JC, Berman DM, Coviello A. 1986. Angiotensin converting enzyme in the toad Bufo arenarum. Comp Biochem Physiol 83A:331–332. Fitzsimons JT. 1979. The physiology of thirst and sodium appetite. New York: Cambridge University Press. Fitzsimons JT. 1998. Angiotensin, thirst, and sodium appetite. Physiol Rev 78:583–686. Fitzsimons JT, Kaufman S. 1977. Cellular and extracellular dehydration, and angiotensin as stimuli to drinking in the common iguana, Iguana iguana. J Physiol 265: 443–463. Fitzsimons JT, Kucharczyk J, Richards G. 1978. Systemic an- THIRST AND ANGIOTENSIN IN S. COUCHII giotensin-induced drinking in the dog: a physiological phenomenon. J Physiol 276:435–448. Gray DA, Simon E. 1985. Control of plasma angiotensin II in a bird with salt glands (Anas platyrhynchos). Gen Comp Endocrinol 60:1–13. Gray DA, Erasmus T. 1988. Plasma arginine vasotocin and angiotensin II in the water deprived kelp gull (Larus dominicanus), cape gannet (Sula capensis), and jackass penguin (Spheniscus demersus). Comp Biochem Physiol 91A: 727–732. Gray DA, Erasmus T. 1989. Plasma arginine vasotocin, angiotensin II, and salt gland function in freshwater- and seawater-adapted kelp gulls (Larus dominicanus). J Exp Zool 249:138–143. Hanssens M, Kierse MJNC, Lijnen PJ, Kucharski M-J, Luyten C, Van Assche FA. 1990. Measurement of individual angiotensin peptides by HPLC-RIA. J Biochem Biophys Meth 21:103–113. Harding JW, Yoshida MS, Dilts RP, Woods TM, Wright JW. 1986. Cerebroventricular and intravascular metabolism of [125I]angiotensins in rat. J Neurochem 46:1292–1297. Harding JW, Jensen LL, Hanesworth JM, Roberts KA, Page TA, Wright JW. 1992. Release of angiotensins in paraventricular nucleus of rat in response to physiological and chemical stimuli. Am J Physiol 262:F17–F23. Hazon N, Balment RJ, Perrott M, O’Toole LB. 1989. The renin-angiotensin system and vascular and dipsogenic regulation in elasmobranchs. Gen Comp Endocrinol 74:230–236. Hillman SS. 1980. Physiological correlates of differential dehydration tolerance in anuran amphibians. Copeia 1980: 125–129. Hillyard SD, Hoff KVS, Propper CR. 1998. The water absorption response: a behavioral assay for physiological processes in terrestrial amphibians. Physiol Zool 71:127–138. Hirano T, Hasegawa S. 1984. Effects of angiotensins and other vasoactive substances on drinking in the eel, Anguilla japonica. Zool Sci 1:106–113. Hoff KVS, Hillyard SD. 1991. Angiotensin II stimulates cutaneous drinking in the toad Bufo punctatus. Physiol Zool 64:1165–1172. Johnson AK, Robinson MM, Mann JF. 1984. The role of the renal renin-angiotensin system in thirst. In: De Caro G, Epstein AN, Massi M, editors. The physiology of thirst and sodium appetite. Vol. 105. New York: Plenum. p 161–180. Johnson WE, Propper CR, Harvey L. 1995. Osmotic threshold for water absorption response (WR) behavior in Scaphiopus couchii. Am Zool 35:94A. Johnson WE, Propper CR, Hillyard SD. 1998. Effects of dehydration on plasma and brain angiotensin in anurans: association with thirst-related behavior. Physiologist 41:370. Jorgensen CB. 1994. Water economy in a terrestrial toad (Bufo bufo), with special reference to cutaneous drinking and urinary bladder function. Comp Biochem Physiol 109(A):311–324. Kloas W, Hanke W. 1992. Localization of binding sites for atrial natriuretic factor and angiotensin II in the central nervous system of the clawed toad Xenopus laevis. Cell Tiss Res 267:365–373. Kobayashi H, Uemura H, Wada M, Takei Y. 1979. Ecological adaptation of angiotensin-induced thirst mechanism in tetrapods. Gen Comp Endocrinol 38:93–104. Ledwith BJ, Cahill MK, Losse LS, Satiritz SM, Eydelloth RS, Dallob AL, Tanaka WK, Galloway SM, Nichols WW. 1993. Measurement of plasma angiotensin II: purification by cation-exchange chromatography. Anal Biochem 213:349–355. 583 Lijnen PJ, Amery AK, Fagard RH, Katz FH. 1978. Radioimmunoassay of angiotensin II in unextracted plasma. Clin Chim Acta 88:403–412. Mann JFE, Johnson AK, Ganten D. 1980. Plasma angiotensin II: dipsogenic levels and angiotensin-generating capacity of renin. Am J Physiol 238:R372–R377. Mayer LP, Propper CR. 2000. Intra-and extracellular dehydration has no effect on plasma levels of angiotensin II in an amphibian. J Exp Zool 286:343–349. McClanahan L Jr. 1967. Adaptations of the spadefoot toad, Scaphiopus couchii, to desert environments. Comp Biochem Physiol 20:73–99. McClanahan L Jr. 1972. Changes in body fluids of burrowed spadefoot toads as a function of soil water potential. Copeia 1972:209–216. National Research Council. 1996. Guide to the care and use of laboratory animals. Washington, DC: National Academy of Sciences, National Academy Press. Neary TJ, Northcutt RG. 1983. Nuclear organization of the bullfrog diencephalon. J Comp Neurol 213:262–278. Nolly H, Fasciolo JC. 1971a. The renin-angiotensin system in Bufo arenarum and Bufo paracnemis. Comp Biochem Physiol 39A:823–831. Nolly H, Fasciolo JC. 1971b. Renin-angiotensin system and sodium homeostasis in Bufo arenarum. Comp Biochem Physiol 39A:833–841. Northcutt RG, Royce GJ. 1975. Olfactory bulb projections in the bullfrog Rana catesbeiana. J Morph 15:251–268. Palkovitz M, Brownstein MJ. 1982. Microdissection of brain areas by the punch technique. Proc Ann Meet Soc Neurosci 60–71. Perrott MN, Grierson CE, Hazon N, Balment RJ. 1992. Drinking behaviour in sea water and fresh water teleosts, the role of the renin-angiotensin system. Fish Physiol Biochem 10:161–168. Phillips IM, Heininger F, Toffolo S. 1996. The role of brain angiotensin in thirst and AVP release induced by hemorrhage. Regul Pept 66:3–11. Propper CR, Hillyard SD. 1996. Distribution of angiotensin II receptors in the skin and brain of the spadefoot toad Scaphiopus couchii. Am Zool 36:30A. Propper CR, Moore FL. 1991. Effects of courtship on brain gonadotropin hormone-releasing hormone and plasma steroid concentrations in a female amphibian (Taricha granulosa). Gen Comp Endocrinol 81:304–312. Propper CR, Johnson WE. 1994. Angiotensin II induces water absorption behavior in two species of desert anurans. Horm Behav 28:41–52. Propper CR, Hillyard SD, Johnson WE. 1995. Central angiotensin II induces thirst-related responses in an amphibian. Horm Behav 29:74–84. Reboreda JC, Segura ET. 1989. Water balance effects of systemic and intracerebroventricular administration of angiotensin II in the toad Bufo arenarum. Comp Biochem Physiol 93A:505–509. Rolls BJ, Rolls ET. 1982. Thirst. New York: Cambridge University Press. Ruibal R. 1962. The adaptive value of bladder water in the toad, Bufo cognatus. Phys Zool 35:218–223. Saavedra JM. 1992. Brain and pituitary angiotensin. Endocr Rev 13:329–380. Segura ET, Bandsholm UC, Bronstein A. 1982. Role of the CNS in the control of water economy of the toad Bufo arenarum Hensel: II. Adrenergic control of water uptake across the skin. J Comp Physiol 146:101–106. 584 W.E. JOHNSON AND C.R. PROPPER Semple PF, Boyd AS, Dawes PM, Morton JJ. 1976. Angiotensin II and its heptapeptide (2–8), hexapeptide (3–8), and pentapeptide (4–8) metabolites in arterial and venous blood of man. Circ Res 39:671–678. Shoemaker VH, McClanahan L Jr, Ruibal R. 1969. Seasonal changes in body fluids in a field population of spadefoot toads. Copeia 1969:585–591. Siegel S, Castellan NJ. 1988. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, Inc. Snapir N, Robinzon B, Godschalk M. 1976. The drinking response of the chicken to peripheral and central administration of angiotensin II. Pharmacol Biochem Behav 5:5–10. Sokol RR, Rohlf FJ. 1981. Biometry. New York: WH Freeman and Co. Stille WT. 1958. The water absorption response of an anuran. Copeia 1958:2177–2178. Takei Y. 1977. Angiotensin and water intake in the Japanese quail (Coturnix cotrunix japonica). Gen Comp Endocrinol 31:364–372. Taylor K, Mayer LP, Propper CR. 1999. Intra- and extracellular dehydration induce water absorption response behavior in an amphibian. Physiol Behav 65:717–721. Tierney ML, Luke G, Cramb G, Hazon N. 1995. The role of the renin-angiotensin system in the control of blood pressure and drinking in the european eel, Anguilla anguilla. Gen Comp Endocrinol 100:39–48. Tran DY, Hoff KS, Hillyard SD. 1992. Effects of angiotensin II and bladder condition on hydration behavior and water uptake in the toad, Bufo woodhouseii. Comp Biochem Physiol 103A:127–130. Unger T, Badoer E, Ganten D, Lang RE, Rettig R. 1988. Brain angiotensin: pathways and pharmacology. Circulation 77:I40–I54. Van Eekelen JAM, Phillips MI. 1988. Plasma angiotensin II levels at moment of drinking during angiotensin II intravenous infusion. Am J Physiol 255:R500–R506. Wada M, Kobayashi H, Farner DS. 1975. Induction of drinking in the white-crowned sparrow Zonotrichia leucophrys gambelii, by intracranial injection of angiotensin II. Gen Comp Endocrinol 26:192–197. Wright JW, Roberts KA, Harding JW. 1988. Drinking to intracerebroventricularly infused angiotensin II, III, and IV in the SHR. Peptides 9:979–984. Yamaguchi K. 1981. Effects of water deprivation on immunoreactive angiotensin II levels in plasma, cerebroventricular perfusate and hypothalamus of the rat. Acta Endocrinol 97: 137–144. Yamaguchi K, Sakaguchi T, Kamoi K, Hama H, Ito S. 1981. Effect of nephrectomy on angiotensin II concentrations in plasma and hypothalamus of the dehydrated rat. Brain Res 224:431–435. Zoeller RT, Moore FL. 1985. Seasonal changes in luteinizing hormone-releasing hormone concentrations in microdissected brain regions of rough-skinned newt (Taricha granulosa). Gen Comp Endocrinol 58:222–230.