Adropin acts in brain to inhibit water drinking: potential interaction with the orphan G protein-coupled receptor, GPR19

Abstract

Adropin, a recently described peptide hormone produced in the brain and liver, has been reported to have physiologically relevant actions on glucose homeostasis and lipogenesis, and to exert significant effect on endothelial function. We describe a central nervous system action of adropin to inhibit water drinking and identify a potential adropin receptor, the orphan G protein-coupled receptor, GPR19. Reduction in GPR19 mRNA levels in medial basal hypothalamus of male rats resulted in the loss of the inhibitory effect of adropin on water deprivation-induced thirst. The identification of a novel brain action of adropin and a candidate receptor for the peptide should extend and accelerate the study of the potential therapeutic value of adropin or its mimetics for the treatment of metabolic disorders.

in 2008, Kumar et al. (6) identified the peptide hormone adropin to be encoded by the energy homeostasis-associated (Enho) gene (6). They described adropin's production in brain and liver and importantly detailed the effect of diet on Enho expression. This revealed a link between metabolic status and adropin production. The authors went further to show that transgenic overexpression of adropin and pharmacologic administration of the peptide improved insulin sensitivity in mice exposed to a high-fat diet (the diet-induced obesity model, or DIO). These effects were reported to be independent of changes in food intake, energy expenditure, or body weight. Indeed, there were no reported differences in food intake in transgenic overexpressing mice compared with controls; however, in DIO, B6 mice adropin administration (intraperitoneally twice daily for 14 days) significantly reduced food intake compared with controls. Another important observation in the transgenic mice was that the expression of genes involved in lipid synthesis was decreased, suggesting that when considered alongside the evidence that the onset of DIO in these animals was delayed, that adropin not only improves glucose homeostasis, but also protects against overaccumulation of lipids during high-fat diet exposure. The same group then reported in 2012 (7) that adropin deficiency resulted in the opposite phenotype: increased adiposity and insulin resistance. The authors concluded that adropin deficiency could predispose individuals to the development of Type 2 diabetes mellitus and the metabolic syndrome.

Shortly after the first publication on adropin (6), Lovren et al. (8) reasoned that since insulin resistance is often associated with impaired vascular function, adropin may exert protective effects in the blood vessels and, in particular, on vascular endothelial cells. In an impressive series of experiments, these authors demonstrated that adropin was expressed in human umbilical vein and coronary artery endothelial cells (ECs) and that the peptide exerted protropic effects in those cells. Adropin stimulated proliferation, migration, and capillary tube formation in cultured ECs, effects that were prevented by pretreatment with a false substrate for nitric oxide synthase (nitro-l-arginine methyl ester, l-NAME) or by acute knockdown of endogenous expression of vascular endothelial growth factor 2 receptor (VEGF2R). Indeed, adropin increased nitric oxide release via PI3 kinase-AKT and ERK1/2 pathways and upregulated VEGF2R expression in ECs. Thus, in addition to the potential therapeutic use of adropin to improve insulin sensitivity, the peptide may be a novel target for the treatment of vascular diseases, such as atherothrombosis. The data reported by Lovren et al. (8) suggested that adropin might signal via a G protein-coupled cell surface receptor.

The major problem hindering the development of novel therapeutics based on endogenous hormones is that of delivery. In the case of adropin, the mature peptide is quite large; however, a fragment, amino acids 34–76, is bioactive. Recent advances in the field of peptide biology have been made with the realization that small-molecule mimetics of endogenous peptides can be designed that have greater stability in biologic fluids and may even be orally active. To develop such mimetics, the receptor that recognizes the endogenous hormone must first be identified. Numerous “orphan” G protein-coupled receptors have been cataloged for which the endogenous ligand has yet to be identified. We have developed a “deductive reasoning strategy,” with which we have matched biologic ligands with their cognate receptors (18, 20). We used this strategy to identify a potential adropin receptor and detail here not only a novel action of the peptide, but also the loss of that action when expression of our top adropin receptor candidate was manipulated using small interfering RNA (siRNA) in vivo.

METHODS

Adult, male rats (Sprague-Dawley, Harlan/ENVIGO, Indianapolis, IN) weighing 225–275 g were housed individually (23–25°C, water and lab chow ad libitum, lights on 0600–1800) after implantation of an indwelling lateral cerebroventricle cannula (stereotaxic coordinates: A: +6.2, H: +7.4, L: −0.9) (11) under ketamine/xylazine anesthesia (Ketaset, Fort Dodge Animal Health, Fort Dodge, IA)/xylazine (TranquiVed, Vedco, St. Joseph, MO) intraperitoneally (60 mg ketamine/8 mg xylazine/ml, 0.1 ml/100 g body wt) with buprenorphine analgesia, as previously described (19). We chose to administer adropin and siRNA via the lateral cerebroventricle instead of the third ventricle to avoid damage to any midline structures, such as the subfornical organ, during cannula placement. Animals were used in only one (no repeat testing) protocol. After a minimum of 5 days of recovery from intracerebroventricular cannulation, some of the rats again were anesthetized (ketamine/xylazine), and an additional cannula (PE-50) was implanted into the left carotid artery (19). The cannula was filled with heparinized saline (200 U/ml in sterile, 0.9% NaCl) and exteriorized between the shoulder blades. Experiments in these animals were conducted the day following carotid cannulation. All procedures were approved by the Saint Louis University Animal Care and Use Committee.

Protocol 1. effect of central adropin administration on ad libitum food and water intake.

Rats were housed in metabolic cages (Nalgene, Harvard Instruments, Holliston, MA), and daily food and water intakes and body weights were monitored for 3 days. Food intake was determined by weighing the food cups and attached spill guards, and water intake was read from graduated cylinders. On the day of experimentation, food cups and water bottles were removed at 1700, and saline (2 μl, sterile 0.9% NaCl) vehicle or vehicle containing 0.1, 1.0, or 3.0 nmol adropin (rat adropin 34–76; Phoenix Pharmaceuticals, Burlingame, CA) was administered via the lateral ventricular cannula. These doses were chosen on the basis of the report of nanograms adropin per milligram of brain tissue (1) and the extensive distribution of mRNA encoding the peptide in multiple brain sites (6). Additionally, these doses are similar to those used by us previously to demonstrate the effects of a variety of peptides on drinking behaviors (12, 19, 21, 22). Ten minutes after the intracerebroventricular injection, food cups and water bottles were returned to the cages, and intake was monitored at 1800, 1830, 1900, and 1930 and again at 0800 the following morning.

Protocol 2. effect of central adropin administration on water drinking in response to overnight water restriction.

At 1800, water bottles were removed from the cages; however, food remained available. At 1000 on the following morning, saline vehicle or vehicle containing adropin (1.0 or 3.0 nmol) was injected intracerebroventricularly, and 10 min later, water bottles were returned to the cages. Water intakes were monitored 15, 30, 45, 60, 90, 120, and 180 min later, again in the following morning at 0800.

Protocol 3. targeting GPR19 mRNA levels abrogated the inhibitory effect of adropin on water drinking.

Daily water intakes were monitored in male rats before beginning treatment with siRNA constructs targeting GPR19 or as a control enhanced green fluorescent protein (20). The following siRNA constructs were designed and purchased from Integrated DNA Technologies (IDT, Coralville, IN): GPR19 siRNA: 5′-rArCrCrArArArGrArCrUrCrCrArUrCrUrArUrGrArCrUrCAT-3′, 5′-rArUrGrArGrUrCrArUrArGrArUrGrGrArGrUrCrUrUrUrGrGrUrUrA-3′, 5′-rArGrArCrGr GrUrCrArArGrArUrGrUrUrUrCrUrGrCrUrCTT-3′, 5′-rArArGrArGrCrArGrArAr ArCrArUrCrUrUrGrArCrCrGrUrCrUrUrC-3′, 5-rCrCrCrArGrUrGrGrCrUrUrArUrGrGr CrArUrCrCrUrCrCAA-3′, 5′-rUrUrGrGrArGrGrArUrGrCrCrArUrArArGrCrCr ArCrUrGrGrGrArA-3′; and eGFP siRNA 5′-rArCrCrCrUrGrArArGrUrUrCrArUrCr UrGrCr ArCrCrArCrCrG-3′, 5′-rCrGrGrUrGrGrUrGrCrArGrArUrGrArArCrUrUr CrArGrGrGrUrCrA-3′. Constructs were diluted in saline and administered intracerebroventricularly (2 μg in 2 μl) between 1700 and 1800 for two consecutive days. On the second day of treatment, water bottles were removed at 1800 (food remained present). At 1000 on the following morning, saline vehicle or vehicle containing 3.0 nmol adropin was administered intracerebroventricularly, and water bottles returned to the cages 10 min later. Intake was monitored as in Protocol 2. Animals were killed following the final time point, and brains were rapidly dissected for determination of GPR19 mRNA levels in the following tissues: medial basal hypothalamus (including the arcuate, ventromedial, and paraventricular nuclei) and the rostral hypothalamus (including the preoptic areas and the structures surrounding the lamina terminalis).

Determination of GPR19 mRNA levels in dissected tissues.

Total RNA collected from the medial basal and rostral hypothalamus was isolated using RNeasy mini kit in accordance with the manufacturer's instructions. (Ambion, Life Technologies, Carlsbad, CA). RNA was reverse transcribed using iScript cDNA synthesis kit (Bio-Rad; Hercules, CA). Quantitative real-time PCR (qPCR) was carried out using iQ SYBRGreen Master Mix and a Bio-Rad CFX96 real-time System (Bio-Rad). The following primers were designed using PrimerBlast (NCBI, Bethesda, MD) and purchased from Integrated DNA Technologies (IDT, Coralville, IN): HPRT1 forward 5′-AGTCCCAGTGTCGTGATTAGTGAT-3′ and reverse 5′-CTCGAGCAAGTCTTTCAGTCCTGT-3′; GPR19 forward 5′-TATGACTGTGTTCCGAA GGTTT-3′ and reverse 5′-TCCTCTTAGTTCTTCATGGGGAC-3′. Changes in GPR19 mRNA expression was calculated using the ΔΔCt method and normalized to the housekeeping gene HPRT-1 (14), as we have shown previously (18, 20).

Protocol 4. effect of central administration of adropin on mean arterial pressure in conscious rats.

One day after implantation of the carotid cannula, rats were habituated to a quiet testing room for at least 2 h. The carotid catheter was connected to a pressure transducer (Digi-Med BPA, Micro-Med, Louisville, KY), and the catheter was flushed with heparinized saline (200 U/ml in sterile 0.9% NaCl). Baseline mean arterial pressures were recorded at 1-min intervals for at least 30 min. Central administration of saline vehicle (2 μl 0.9% NaCl alone or containing 1.0 or 3.0 nmol adropin) were followed by at least 30 min of undisturbed recording of mean arterial pressure (MAP).

Statistical analyses.

Differences among groups in food and water intakes (Protocols 1 and 2) were determined by ANOVA (time and treatment) with Scheffé's multiple comparison post hoc testing. For Protocol 3, differences in water intakes between the two treatment groups were determined by independent t-tests. Between-group differences in GPR19, mRNA levels were determined by nonparametric analysis (Mann Whitney U) for normalized data. For Protocol 4, data are presented as the change from preinjection baseline, calculated as the average MAP for 10 min prior to intracerebroventricular injection, to account for the natural variation in resting cardiovascular parameters between animals. Changes in MAP are shown as the area under the curve. Data were analyzed using a nonparametric test (Mann-Whitney U) because data were transformed (24). Significance was assigned to outcomes with a calculated P value of less than 0.05. All determinations were between subjects.

RESULTS

Protocol 1.

Central administration of all three doses of adropin resulted in a transient, but significant, inhibition of water drinking. There were no significant differences among the adropin treatment groups; however, compared with vehicle-injected controls, all three adropin groups consumed significantly less water at 1800, and animals in the 1.0-mol group had consumed significantly less water than controls also at 1830 and 1900 (Fig. 1A). Food intakes were not significantly different among groups, with the exception of a difference between vehicle and 1.0-nmol administered animals at 1900 (Fig. 1B). No significant differences in water or food intakes were detected at 1930 or the following morning at 0800.

Fig. 1.

The effect of intracerebroventricular administration of adropin on light-entrained water drinking (A) and food intake (B) in male rats. *P < 0.05, **P < 0.01 vs. vehicle controls (ANOVA and Scheffé's multiple comparisons).

Protocol 2.

Significantly less water was consumed by 3.0 nmol adropin-treated rats than vehicle-treated controls following overnight water restriction at all times (Fig. 2), except immediately following return of water bottles to the cages (15 min). Again, no significant differences in cumulative water intakes among the adropin treatment groups were observed. At 0800 on the following morning, no significant differences were observed in cumulative water intakes among the groups (vehicle controls: 37.7 ± 2.1; 1.0 nanomole: 35.5 ± 3.7; 3.0 nanomoles: 38.1 ± 3.8 ml water).

Fig. 2.

Central administration of adropin 10 min prior to the return of water bottles to the cages inhibited water drinking in response to overnight water deprivation in male rats. *P < 0.05 vs. saline control (ANOVA and Scheffé's multiple comparisons).

Protocol 3.

No significant differences in 24 h, ad libitum water intakes were observed between the two treatment groups prior to overnight water restriction either before or during the period of siRNA treatment. Rats pretreated for two consecutive days with siRNA targeting GPR19 consumed significantly more water following intracerebroventricular adropin administration than rats pretreated with the control siRNA (Fig. 3). Significant differences in cumulative water intakes were observed at all time points, with more water consumed by the GPR19 siRNA-pretreated animals. Rats receiving siRNA targeted to GPR19 exhibited an ∼33 ± 7% reduction in GPR19 mRNA expression (ΔΔCt values: EGFP pretreated: 0.99 ± 0.14, GPR19 pretreated: 0.67 ± 0.07; P < 0.05) in the medial basal hypothalamus compared with those receiving control eGFP siRNA. Because of the high variability in levels of GPR19 mRNA detected in the rostral hypothalamus of both pretreatment groups, no significant difference was observed (ΔΔCt values: EGFP siRNA pretreated: 1.17 ± 0.19, GRP19 siRNA pretreated: 1.76 ± 0.29; P = 0.133).

Fig. 3.

siRNA induced reduction in GPR19 mRNA levels in medial basal hypothalamus blocked the inhibitory action of 3.0-nm adropin, given intracerebroventricular 10 min before return of water bottles to the cages, on overnight water deprivation-induced drinking in male rats. *P < 0.05, **P < 0.01 vs. EGFP siRNA-retreated controls.

Protocol 4.

Preinjection baseline MAPs did not differ significantly among groups (vehicle: 131.2 ± 2.7, 1.0 nmol; adropin: 132.9 ± 3.1, 3.0 nmol; adropin: 125.7 ± 4.8; P = 0.43 mmHg). When analyzed both as change from preinjection baseline values or as area under the curve for the 30-min observation period (Fig. 4), no significant differences in mean arterial pressure responses were observed across the three treatments (P = 0.11).

Fig. 4.

Central administration of 10 or 3.0 nm of adropin failed to significantly elevate mean arterial pressure (MAP) in conscious, unrestrained male rats. Data are presented as change from preinjection baselines in terms of area under the curve, with MAP recorded every minute for 30 min postinjection.

DISCUSSION

The broad distribution of adropin production in the brain (6) suggested to us a central nervous system site of action. The action of peripherally administered adropin to alter food intake in DIO B6 mice (7) was to some degree mirrored by the peptide's action in the central nervous system (CNS); albeit, not because of a direct effect on food intake, but, in all likelihood, secondary to an action to inhibit thirst. Although expressed in forebrain and brain stem sites known to be involved in the central control of cardiovascular function, our results did not reveal a significant action of intracerebroventricularly administered adropin to alter MAP in conscious, unrestrained male rats. This is reminiscent of the CNS action of obestatin (12, 13), another neuropeptide that alters food intake secondary to a primary action on water drinking, and does not alter MAP when administered intracerebroventricularly. Although there is a wealth of literature supporting the interdependency of water drinking and food intake, the two can be separated under a variety of circumstances (21, 22, 25).

In the case of obestatin, we were able to exaggerate water deprivation-induced drinking behavior with central nervous system administration of a neutralizing antibody. Here, we took a more direct approach, one that required that we first identified a potential adropin receptor. The work of Lovren et al. (8) strongly suggested that adropin signaled via a G protein-dependent mechanism. We employed the same strategy that we developed in our studies identifying potential receptors for neuronostatin (20) and connecting peptide (C-peptide) (18) to identify our best candidate adropin receptor from the existing list (IUPHAR) of orphan G protein-coupled receptors. Using the “deductive reasoning strategy” we first identified three cell lines that were responsive to adropin [human gastric carcinoma (KATO III), human coronary artery endothelial cells (HCAEC), and mouse fibroblasts (3T3-L1)] using cfos mRNA expression as a reporter of cellular activation. We then generated an orphan GPCR expression profile, using PCR, to identify shared populations of orphan receptors and, in doing so, identified five potential candidates; GPR151, TAAR9, GPR160, GPR19, and GPR63. On the basis of expression patterns in the published literature (5, 9) and ongoing work in the laboratory on in vitro receptor compromise, we identified GPR19 to be our best candidate receptor. We hypothesized that compromise of GPR19 expression would abrogate the action of adropin to inhibit water deprivation-induced drinking.

We developed with the assistance of IDT a triple cocktail of siRNA with which we could reduce GPR19 mRNA levels in cultured cells and turned to our in vivo model of adropin's CNS action, the inhibition of water deprivation-induced water drinking, to study the potential physiological relevance of adropin production in the brain. Administration of GPR19 siRNA intracerebroventricular for 2 days significantly lowered receptor mRNA levels in medial basal hypothalamus of male rats. The inhibitory action of adropin on water deprivation-induced drinking behavior was prevented when mRNA levels of GPR19 had been reduced significantly by pretreatment with siRNA targeting the receptor. Thus, we feel that we have identified a potential receptor for adropin, a step necessary before the physiological relevance of the pharmacologic effects of the peptide can be fully understood.

An important question when using in vivo siRNA is how much knockdown of the target is required to observe significant effects. For in vitro studies, it is generally accepted that the goal is to use the lowest concentration of siRNA/antisense/ribozyme that yields an ∼70% reduction in gene expression. Gene silencing in vivo using small mRNA-interfering molecules has had varying efficacy, depending upon the target protein type (receptor vs. secreted hormone, etc.), endogenous expression level, and distribution of the target, delivery method, and the tissue of interest. Different groups have reported in vivo knockdown ranging from 20 to 90% (16), but in our experience (18–20) and as reported by others (10), a measured knockdown of 20–30% is more than enough to observe measurable differences in a physiological process in vivo, as is the case in the present experiments. Finally, we do not know whether administration of the siRNA into the third cerebroventricle would provide different results from those reported here, with regard to efficiency of target compromise; however, as mentioned above, we have chosen in all of our protocols to cannulate the lateral ventricle to avoid damaging midline structures at the transition between the ventricles or along the wall of the third ventricle itself.

Clearly, we must now demonstrate a physical association of adropin with GPR19 itself, and those studies are ongoing at this time. Similarly, we need to determine whether adropin has a broader complement of actions related to fluid and electrolyte homeostasis. Does it interact with mechanisms controlling vasopressin secretion? What other thirst-producing stimuli can be interrupted by adropin administration and is there a unique pattern of ad libitum water drinking that is regulated by endogenous levels of adropin (e.g., during intermeal intervals)? Is there a sex bias in the action(s) of the peptide and might this explain differences in thirst mechanisms observed in male vs. female laboratory animals? Just the same, the demonstration of a novel CNS action of adropin and the identification of a potential receptor for the peptide merit announcement, so that ongoing studies in other laboratories may benefit from these discoveries.

Perspectives and Significance

Can adropin's CNS action to inhibit water drinking be related to the published, peripheral actions of the peptide? Perhaps adropin plays a central role in the control of body water content by balancing the water derived endogenously from lipid metabolism (6, 7) with a central action to prevent the ingestion of additional fluid. Adropin is produced in the kidney (1) and circulates in plasma with levels changing under a variety of physiological/pathological states (2–4, 15, 17, 23). Is it possible that adropin also regulates water handling in the kidney, where expression of GPR19 has been reported (5)? Our identification of a candidate adropin receptor should facilitate the search for those answers. Additionally, the identification of a potential adropin receptor will speed the development of small molecules with enhanced adropin-like actions on lipogenesis and vascular function, and perhaps facilitate the development of novel therapeutic agents for use in the treatment of metabolic disease.

GRANTS

The work was funded in part by National Institutes of Health Grant 1 RO1 HL-121456 to W. K. Samson. L. M. Stein was supported by National Institute of General Medical Sciences Grant T32-GM008306.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: L.M.S., G.L.C.Y., and W.K.S. conception and design of research; L.M.S., G.L.C.Y., and W.K.S. performed experiments; L.M.S., G.L.C.Y., and W.K.S. analyzed data; L.M.S., G.L.C.Y., and W.K.S. interpreted results of experiments; L.M.S. and W.K.S. prepared figures; L.M.S. and W.K.S. drafted manuscript; L.M.S., G.L.C.Y., and W.K.S. edited and revised manuscript; L.M.S., G.L.C.Y., and W.K.S. approved final version of manuscript.