Orexin deficiency modulates the dipsogenic effects of angiotensin II in a sex-dependent manner

The orexin (hypocretin) neuropeptide system is an important regulator of ingestive behaviors, i.e., it promotes food and water intake. Here, we investigated the role of orexin in drinking induced by the potent dipsogen angiotensin II (ANG II). Specifically, male and female orexin-deficient mice received intracerebroventricular (ICV) injections of ANG II, followed by measuring their water intake within 15 min. We found that lower doses of ANG II (100 ng) significantly stimulated drinking in males but not in females, indicating a general sex-dependent effect that was not affected by orexin deficiency. However, higher doses of ANG II (500 ng) were sufficient to induce drinking in female wild-type mice, while female orexin-deficient mice still did not respond to the dip-sogenic properties of ANG II. In conclusion, these results suggest sex-dependent effects in ANG II-induced drinking and further support the sexual dimorphism of orexin system functions.


Introduction
Polydipsia is defined as a pathologically high intake of water and is frequently expressed in neuropsychiatric disorders [1].Preclinical studies have broadened the understanding of its neurobiological basis using models of schedule-, stress-, or drug-induced polydipsia [2][3][4][5].One example of a highly potent dipsogen is angiotensin II (ANG II) [6], which is the main effector peptide of the renin-angiotensin system and critically involved in regulating cardiovascular and body fluid homeostasis [7,8].Indeed, many studies have shown that intracerebroventricular (ICV) administration of ANG II causes rats and mice to promptly drink high amounts of water within a short period of time [8][9][10][11][12].These dipsogenic responses to ANG II are commonly used as a method to verify ICV cannula implantation sites [cf.13].
Another important regulator of ingestive behaviors is the orexin (hypocretin) system.This neuropeptide system is comprised of two peptides, orexin A and orexin B, that are exclusively expressed in neurons of the lateral and dorsomedial hypothalamus and perifornical area [14][15][16].These peptides bind to their G protein-coupled receptors, the orexin 1 (OX1R) and orexin 2 receptor (OX2R), which are widely expressed throughout the brain [17][18][19].The orexin system has been reported to play an important role in many different functions, and may hence be implicated in various human disorders [20].However, it was originally identified for its promoting effects on ingestive behaviors [14].For instance, chemogenetic stimulation of orexin neurons or central administration of orexin A were shown to increase the intake of food and water [14,21,22].Interestingly, the dipsogenic effects of ICV orexin A were described to be almost as strong as those of ICV ANG II [21].It is unclear, however, whether the orexin system is involved in mediating the dipsogenic effects of ANG II.
To address this question, we investigated the effects of orexin deficiency on drinking induced by ICV injection of ANG II in male and female mice.Given orexin's stimulating effects on water intake, we hypothesized that the dipsogenic responses to ANG II treatment would be reduced in orexin-deficient mice.

Treatment and behavioral testing
ANG II (human angiotensin II, Sigma-Aldrich, Steinheim, Germany) was dissolved in physiological saline.In the first set of mice, we injected a dose of 100 ng ANG II in a volume of 1 µl [13].However, since we observed that the efficacy of this dose was sex-dependent, we treated another set of mice with a higher dose of ANG II (500 ng/1 µl).Each mouse was injected with ANG II (in the respective dose) and saline (vehicle) in a balanced order.There was at least one day of break between the two injections.
For details on injection procedures and behavioral testing, see Faesel et al. [13].Briefly, ICV injections were performed in freely moving mice using a system consisting of custom-made injectors (length: 7 mm, diameter: 0.3 mm), that were connected to microliter syringes via tubes, and a microinjection pump (CMA/100, CMA Microdialysis, Kista, Sweden).Following injections, the mice's drinking behavior was recorded for 15 min.For that, they were placed in small cages (29 cm × 19 cm × 15 cm) that were each equipped with bedding material, a wire lid, and a bottle filled with tap water.Due to leakage issues when turning the bottles, the water intake was determined by weighing the whole cage (without the mouse) before and after testing.Notably, the dipsogenic responses to ANG II treatment were reported to precede significant increases in urination [27].Therefore, it was unlikely that this parameter affected the cage weight during the short duration of testing (15 min) [13].Moreover, the latency to drink and the total time spent drinking were assessed.
At the end of the experiment, ICV injection sites were verified by ink injection and histological inspection as described previously [13].Fig. 1 shows a representative example of a properly placed ink injection in the right lateral ventricle.Mice with ink injections outside the lateral ventricle (n = 7) were not included in the analysis.

Data analyses
Data are presented as individual values with the median.Due to the differential variability and violation of normality in most data sets, statistical analyses relied on non-parametric within-subject comparisons (ANG II vs. vehicle) using two-tailed Wilcoxon matched-pairs signed rank tests (with Pratt's method) [cf. 13].This means that we report on whether or not ANG II affected the water intake (raw values, not corrected for body weight), time spent drinking, and latency to drink compared to control treatment within each respective group.The analyses were performed using GraphPad Prism 8 (GraphPad Software, San Diego, USA).Values of p < 0.05 were defined to be significant.
Expectedly, all three parameters of drinking behavior significantly correlated with each other.That is, the sooner an animal started to drink after ANG II injection, the longer and more it drank overall (latency vs. time: R 2 = 0.38, p < 0.0001; latency vs. water intake: R 2 = 0.42, p < 0.0001; pooled analyses for all genotypes, sexes, and ANG II doses).Also, the more time an animal spent drinking in response to ANG II, the more water it ultimately consumed (time vs. water intake: R 2 = 0.66, p < 0.0001).

Discussion
In this study, we sought to determine the effects of orexin deficiency on drinking stimulated by ICV ANG II in male and female mice.We report on two main findings: First, lower doses of ANG II induced dipsogenic responses in male but not in female mice, irrespective of the orexin genotype.Second, higher doses of ANG II were sufficient to cause drinking in female wild-type mice, but not in female orexin-deficient mice.Hence, our results indicate both a general sex-dependent effect as well as a sex-dependent effect of orexin deficiency in ANG II-induced drinking.
In line with previous studies [11,12], both doses of ICV ANG II potently stimulated drinking in male mice (Fig. 2A, Fig. 3A).These dipsogenic effects were observed on a group level, but not in each individual mouse.This indicates a biological variability of ANG II effects in mice and emphasizes the necessity of reevaluating the validity of ANG II-induced drinking as a method to verify ICV cannula implantations, at least in mice [cf.13].In fact, the dipsogenic effects of ANG II were described to be most potent in rats, while mice seem to be less responsive to ANG II's dipsogenic properties [7,28].
Moreover, we found that 100 ng ANG II did not significantly enhance the water intake of female mice of either genotype (Fig. 2B).While sex differences in ANG II-induced increases in blood pressure are well established, i.e., pressor responses to ANG II are stronger in male than in female rodents [29,30], sex differences in ANG II-induced drinking are less understood as most studies were performed in males only.One recent exception is a study conducted by Santollo et al. [31], in which male and female rats were compared regarding their dipsogenic responses to ICV ANG II.Similar to our results in mice, the authors reported the increase in water intake to be greater in male than in female rats [31].Collectively, these findings suggest that females are less susceptible to the pressor and dipsogenic properties of ANG II, presumably due to protective effects of estrogens [32,33].
Given this sex-dependent effect, we next asked whether higher doses of ANG II would be effective in stimulating water intake in females.Indeed, 500 ng ANG II was sufficient to increase drinking in female wildtype mice (Fig. 3B).Furthermore, and in line with orexin's stimulating effects on water intake [21,22], we found that orexin deficiency modulated the dipsogenic responses to ANG II in a sex-dependent manner.Specifically, female orexin-deficient mice did not significantly increase their water intake even when treated with the higher dose of ANG II (Fig. 3B), whereas male orexin-deficient mice expressed drinking responses similar to their wild-type littermates with either ANG II dose (Fig. 2A, Fig. 3A).These results may indicate sex-specific interactions of the renin-angiotensin and orexin systems in the regulation of water intake.Moreover, our findings add to accumulating evidence showing that the orexin system is sexually dimorphic with respect to a variety of functions, e.g., feeding behavior [34][35][36], stress and cognition [37][38][39], and social interaction [25].Orexin expression as well as orexin neuron activity were reported to be naturally higher in females than in males [37,40,41].These sex differences at basal conditions may explain why the lifelong orexin deficiency had a stronger impact on females [see also : 25,36,38].Furthermore, the orexin system has been shown to interact with gonadal hormones.For example, orexin modulates the secretion of luteinizing hormone, partly in an estrogen-dependent manner [42][43][44].Moreover, the expression of orexin peptides and receptors in the hypothalamus is affected by the estrous cycle [45][46][47][48].Therefore, and since estrogens were suggested to contribute to sex differences in ANG II effects [32,33], it is plausible to speculate that the present results were influenced by the estrous cyclicity in females.In this study, we did not Fig. 2. Drinking behavior in response to 100 ng ANG II.Water intake, time spent drinking, and latency to drink (note: 15 min latency = no drinking during the test) of (A) male and (B) female OX +/+ , OX +/− , and OX − /− mice after ICV injection of 100 ng ANG II or vehicle.Overall, males of each genotype showed ANG IIinduced drinking behavior, but females did not.Data are depicted as individual values with the median (black line).Within-subject data points for vehicle and ANG II treatment are connected by the colored lines (note: subjects with identical value pairs are summarized as one line).* p < 0.05, as indicated, Wilcoxon matched-pairs signed rank tests.
N. Faesel et al. determine estrous cycle phases, but future research on orexin's (sex-specific) role in fluid intake should take this into consideration.
Of note, a few studies have shown that the orexin system interacts with neural substrates that play an important role in regulating fluid intake.For instance, orexin neurons project to the subfornical organ (SFO) [21] and orexin A excites neurons in this structure, likely via OX1R [49].The SFO and organum vasculosum of the lamina terminalis are circumventricular organs that, together with the median preoptic nucleus, form the structures of the lamina terminalis (LT) [50].These forebrain structures are essentially involved in mediating the dipsogenic effects of ANG II [7,27,51,52].Interestingly, Hurley and colleagues [53] recently demonstrated that orexin neurons receive projections from the LT and, in turn, send projections to dopaminergic neurons of the ventral tegmental area.This suggests that the orexin system may be an interface linking neural systems that control fluid balance with those that mediate motivated and reward-related behaviors [53].The finding that ANG II does not effectively stimulate increased drinking in female orexin-deficient mice provides preliminary support for this proposed pathway.However, as these constitutive knockout mice lack orexin throughout the lifespan, it should be noted that our results could be confounded by compensatory mechanisms.Hence, a limitation of this study is that we cannot conclude whether the observed effects were solely caused by orexin deficiency.Future studies are needed to specify our results by using acute manipulations of orexin signaling (e.g., pharmacological blockade of orexin receptors) in animals of both sexes.
Polydipsia describes the excessive intake of fluids and is particularly common in patients with schizophrenia, which is thought to be caused to some extent by chronic treatment with typical neuroleptics [54][55][56].Specifically, it has been suggested that chronic dopamine D2 receptor blockade generates a hypersensitivity to dopamine that, in turn, further increases ANG II function and ultimately promotes polydipsia [54][55][56][57].Moreover, clinical and animal studies provided evidence indicating that the orexin system, perhaps via its interactions with dopamine [cf.58,59], may also be involved in schizophrenia-related polydipsia [5,56,60].For instance, studies in schizophrenic patients revealed significant associations between OX1R gene polymorphisms and polydipsia [56,60].Furthermore, polydipsia induced by increased dopaminergic tone, i.e., via repeated quinpirole injections, was shown to be potentiated by OX1R antagonism in male rats [5].Given these findings and our results of orexin deficiency hampering ANG II-induced drinking only in female mice, it would be interesting to measure acute effects of OX1R blockade on (pharmacologically induced) polydipsia in female animals.
In conclusion, we found both dose-related sex-dependent effects as well as sex-dependent effects of orexin deficiency in mice's dipsogenic responses to ICV ANG II.These results extend our knowledge about the physiological regulation of drinking behavior and may contribute to the understanding of orexin's potential role in polydipsia expressed by schizophrenic patients.Furthermore, this study emphasizes the importance of investigating the role of sex in general and with respect to orexin system function in particular.

Declaration of Competing Interest
None.

Fig. 1 .
Fig. 1.Histological verification of ICV cannulation.At the end of the experiment, blue ink (1 µl) was injected through the guide cannulas and the brains were cut at the level of cannula implantation.The image shows a representative example of a correctly placed ink injection in the right lateral ventricle (R = right hemisphere).

Fig. 3 .
Fig.3.Drinking behavior in response to 500 ng ANG II.Water intake, time spent drinking, and latency to drink (note: 15 min latency = no drinking during the test) of (A) male and (B) female mice of each orexin genotype following ICV injection of 500 ng ANG II or vehicle.Males of each genotype and female OX +/+ mice showed significantly increased drinking after ANG II compared to vehicle treatment.These dipsogenic effects of ANG II were not observed in female OX +/− and OX − /− mice.Data are depicted as individual values with the median (black line).Within-subject data points for vehicle and ANG II treatment are connected by the colored lines (note: subjects with identical value pairs are summarized as one line).* p < 0.05, ** p < 0.01, as indicated, Wilcoxon matched-pairs signed rank tests.