Uroguanylin Action in the Brain Reduces Weight Gain in Obese Mice via Different Efferent Autonomic Pathways

After the ingestion of a meal, the presence of nutrients in the gastrointestinal (GI) tract initiates complex neural and hormonal responses that send signals to the brain about the ongoing changes in nutritional status. Among the different strategies used to communicate to the brain, the gut secretes peptides that reach the central nervous system (CNS) via afferent nerve fibers or the circulation (1,2). New gut hormones are continuously discovered, and, with the exception of ghrelin, which is the only peptidic hormone favoring weight gain and adiposity, all of them are associated with a negative energy balance and are thus potential targets for the treatment of obesity (3,4).

One of the most recently discovered gut hormones is uroguanylin (UGN), a 16–amino acid peptide secreted mainly from duodenal epithelial cells. UGN is synthesized as a prohormone (pro-UGN), which, after cleavage by a still unknown enzyme, is converted to the active UGN (3,5). Both pro-UGN and UGN activate the guanilate cyclase 2C receptor (GUCY2C), which is also targeted by diarrheagenic bacterial heat-stable enterotoxins (STs) (6). The activation of GUCY2C leads to elevated intracellular levels of cyclic guanosine monophosphate (7), which in invertebrate species has been demonstrated to lead to alterations in food behavior and energy-balance regulation (8,9). Circulating UGN levels were decreased in fasted and leptin-deficient mice but recovered after refeeding or exogenous leptin infusion (10), suggesting that UGN is regulated by nutritional status in a leptin-dependent manner.

A previous study (11) proposed that the activation of GUCY2C is also relevant in the regulation of energy balance in mammals. In mice, pro-UGN is released from the gut immediately after nutrient intake and is converted into active UGN within the hypothalamus, thereby activating GUCY2C and reducing food intake (11). In agreement with these results, pharmacological stimulation of GUCY2C inhibited feeding in obese mice, and mice lacking GUCY2C are hyperphagic and more prone to develop metabolic syndrome (11).

Despite the initial enthusiasm about UGN, these novel and promising findings have been challenged recently by another report (12) indicating that the central administration of UGN or STs does not change food intake or body weight in lean rats and that GUCY2C knockout mice do not present variations in body weight or glucose metabolism when compared with the wild-type (WT) mice. In fact, that report (12) found only a modest increase in body weight, adiposity, and glucose intolerance when UGN-deficient mice were fed a high-fat diet (HFD).

Because the precise role of UGN in the regulation of energy homeostasis is controversial and the pharmacological efficiency of UGN in obesity remains unknown, the aim of the present work was as follows:

• 1) to investigate whether chronic exposure to UGN is an effective pharmacological treatment of obesity via its actions on important parameters in energy homeostasis, such as food intake, energy expenditure, or nutrient partitioning and

• 2) to dissect the molecular underpinnings involved in the metabolic action of UGN.

Swiss male mice (weight 20–25 g, age 8–10 weeks old), WT, and triple β-adrenoreceptor (AR) knockout (TKO) male mice (weight 20–25 g, age 8–10 weeks old) were housed in individual cages under controlled conditions of illumination (12-h light/dark cycle), temperature, and humidity. Mice were allowed ad libitum access to water and a standard laboratory diet (proteins 16%, carbohydrates 60%, and fat 3%; Scientific Animal Food & Engineering) or HFD (45% of calories from fat, 4.73 kcal/g; catalog #12451; Research Diets, New Brunswick, NJ) for 10 weeks. Seven to 11 animals per group were used. Ten-week-old WT and TKO mice were obtained as previously described (13–15). Briefly, β₁^+/−β₂^+/−β₃^+/− mice were crossed to generate β₁^+/+β₂^+/+β₃^+/+ and β₁^−/−β₂^−/−β₃^−/− mice. Several couples were then established from these homozygous mice at the University of Geneva, and experiments were performed on β₁^+/+β₂^+/+β₃^+/+ and β₁^−/−β₂^−/−β₃^−/− mice offspring. Food intake, body weight, and total feces mass were measured during the experimental phase in all experiments. Animals were killed by decapitation, and tissues were removed rapidly and immediately frozen on dry ice and kept at −80°C until their analysis. All experiments and procedures involved in this study were reviewed and approved by the Ethics Committee of the Universidade de Santiago de Compostela, in accordance with the European Union norm for the use of experimental animals.

Mice were anesthetized by an intraperitoneal injection of ketamine (8 mg/kg body wt) and xylazine (3 mg/kg body wt). Intracerebroventricular cannulae were implanted stereotaxically in mice, as described previously (16). Animals received an intracerebroventricular administration of vehicle (saline), α–melanocyte-stimulating hormone (MSH; 3 µg/mouse; Sigma-Aldrich, St. Louis, MO) or UGN (10 or 25 μg/mouse; Bachem, Bubendorf, Switzerland). To evaluate the chronic central effects of UGN, we connected a catheter from the brain infusion cannula to an osmotic micropump flow moderator (25 µg/mouse/day). Using blunt dissection, we created a subcutaneous pocket on the dorsal surface, where we inserted the osmotic micropump (model 1007D ALZET Osmotic Pumps; Durect, Cupertino, CA). These pumps had a flow rate of 0.5 µL/h during 7 days of treatment. After surgery, mice were sutured and kept warm until they fully recovered. For peripheral treatments, mice received an intraperitoneal administration of UGN (25 µg/mouse).

Surgical vagotomy (VGX) was performed as previously described (17,18). Under ketamine-xylazine anesthesia, mice were placed on their backs, and a midline abdominal incision was made. The liver was then carefully moved to the right, exposing the esophagus. Dorsal and ventral branches of the vagus nerve were exposed and dissected from the esophagus. Each branch of the nerve was ligated with surgical sutures at two points as distally as possible to prevent bleeding and then was cauterized between the sutures. The abdominal muscles and the skin were then sutured with surgical silk. Sham surgeries were also performed, in which each trunk of the nerve was exposed but not tied or cauterized. The effectiveness of the VGX was assessed 3 weeks after the surgery by postmortem stomach observation, with evident increase in stomach size after VGX (due to motoric dysfunction) as an indication of the success of the therapy (data not shown). Only mice with successful therapy were included in the analysis. Osmotic minipumps and intracerebroventricular cannulae were implanted 2 weeks post-VGX.

Pharmacological inactivation of β3-AR was performed by subcutaneous administration of the specific antagonist SR59230A (Tocris Bioscience) at a dose of 3 mg/kg (19).

Body composition (fat mass) was assessed using a nuclear magnetic resonance imaging system (Whole Body Composition Analyzer; EchoMRI, Houston, TX). Measurements were performed before surgery and on the last day of the treatment. During intracerebroventricular treatment for a period of 7 days, mice were analyzed for energy expenditure, respiratory quotient (RQ), and locomotor activity using a calorimetry system (LabMaster; TSE Systems) (16).

Tissues were homogenized using the TissueLyser II (Qiagen, Tokyo, Japan) in cold radioimmunoprecipitation assay buffer containing 200 mmol/L Tris/HCl (pH 7.4), 130 mmol/L NaCl, 10% (v/v) glycerol, 0.1% (v/v) SDS, 1% (v/v) Triton X-100, and 10 mmol/L MgCl2 with antiproteases and antiphosphatases (Sigma-Aldrich). The tissue lysates were centrifuged for 30 min at 18,000g in a microfuge at 4°C. Total protein lysates from brown adipose tissue (BAT) (20 µg) and epididymal white adipose tissue (WAT) (20 µg) were run on 10% SDS-PAGE, then electrotransferred onto a nitrocellulose membrane and probed successively with the following antibodies: uncoupling protein (UCP) 1, UCP3, cell death–inducing DNA fragmentation factor α–like effector A (CIDEA), fibroblast growth factor 21 (FGF21), PR domain containing 16 (PRDM16), and hormone-sensitive lipase (HSL) (Abcam, Cambridge, U.K.); Phospho-HSL (Ser660) (Cell Signaling Technology, Danvers, MA); peroxisome proliferator–activated receptor γ coactivator 1 α (PGC1α) and HSP90 (Santa Cruz Biotechnology, Santa Cruz, CA); and β-actin (Sigma-Aldrich) after incubating the membranes with 5% BSA blocking buffer. For protein detection, we used horseradish peroxidase–conjugated secondary antibodies (Dako Denmark, Glostrup, Denmark). Specific antigen-antibody bindings were visualized using reactive chemiluminescence detection (Pierce ECL Western Blotting Substrate; Thermo Scientific). Then, the image of the membranes was made by exposing radiograph film (Super RX, Fuji Medical X-Ray Film; Fujifilm, Tokyo, Japan) and processing the film with developer and fixer liquids (AGFA, Mortsel, Belgium) under appropriate dark room conditions. The protein levels were normalized to β-actin or HSP90 for each sample.

To study the histomorphological structure of BAT and WAT, we performed hematoxylin-eosin staining in tissue sections. Briefly, BAT and WAT samples were fixed for 24 h in 10% formalin buffer and then were dehydrated and embedded in paraffin using a standard procedure. Sections of 3 μm were made with a microtome, and staining was performed with standard Hematoxylin/Eosin Alcoholic (Bio-Optica) procedure following the manufacturer instructions (20). Sections were observed and photographed using a Provis AX70 microscope (Olympus, Tokyo, Japan). Digital images for BAT were quantified with ImageJ software (National Institutes of Health).

The detection of UCP1 in BAT and WAT was performed using anti-UCP1 (1:500; Abcam, Cambridge, U.K.), as previously described (19). Images were photographed with an XC50 digital camera (Olympus).

RNA was isolated from the adipose tissue using TRIzol Reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions (18). The extracted total RNA was purified with DNase treatment using a DNA-free kit as a template (Ambion; Thermo Fisher Scientific, Grand Island, NY) to generate first-strand cDNAs using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Quantitative real-time PCR was performed using a StepOnePlus Instrument (Applied Biosystems) with specific TaqMan quantitative RT-PCR primers and probes. The oligonucleotide-specific primers are listed in Supplementary Table 1. For the analysis of the data, hypoxanthine phosphoribosyltransferase was used as the endogenous control, and the expression levels in the sample group were expressed relative to the average of the control group.

The extraction procedure for tissue triglyceride (TG) was adapted from methods described previously (13,18). BAT was homogenized for 2 min in ice-cold chloroform-methanol (2:1, v/v). TGs were extracted during 5 h of shaking at room temperature. For phase separation, H₂O Milli-Q was added, samples were centrifuged, and the organic bottom layer was collected. The organic solvent was dried using SpeedVac and redissolved in chloroform. The TG (Randox Laboratories, Ltd., London, U.K.) content of each sample was measured in duplicate after evaporation of the organic solvent using an enzymatic method.

Results are expressed as the mean ± SEM. GraphPad Prism (version 4.0) software was used for the data analysis. Statistical analysis was performed using one-way ANOVA followed by a post hoc multiple comparison test (Bonferroni test). A P value <0.05 was considered statistically significant.

We first injected two different doses of UGN (10 and 25 μg i.c.v.) in diet-induced obese (DIO) mice subjected to overnight fasting. We found that a single intracerebroventricular injection of UGN (25 µg) significantly decreased food intake at 1, 2, 4, and 8 h after injection, and then the effect diminished (Fig. 1A). Although after 24 h there were no differences in food intake, mice treated with intracerebroventricular UGN at this dose showed a significant decrease in body weight when compared with intracerebroventricular saline–treated mice (Fig. 1B). At the lower dose, UGN did not alter food intake (Fig. 1A) or body weight (Fig. 1B). As a positive control to corroborate the efficiency of the cannulae, we also injected mice with intracerebroventricular α-MSH (3 µg), which is known to inhibit feeding and body weight (21). Next, we followed the same experimental protocol but in DIO mice fed ad libitum. Again, the higher dose of UGN (25 µg i.c.v.) decreased food intake transiently (Fig. 1C), and after 24 h intracerebroventricular UGN–treated mice showed a significant inhibition of body weight gain in comparison with intracerebroventricular saline–treated mice (Fig. 1D). As expected, intracerebroventricular α-MSH also suppressed feeding and weight gain (Fig. 1C and D) in mice fed ad libitum.

To rule out the possibility that centrally injected UGN leaks out of the CNS into the circulation and elicits a response by directly acting at peripheral UGN receptors, we administered UGN peripherally at the same dose injected centrally (25 µg). We were unable to detect changes in food intake (Fig. 1E) or body weight (Fig. 1F). We therefore conclude that the effects observed during intracerebroventricular administration were due entirely to UGN acting at the central level.

Central infusion of UGN (25 µg/day) for 7 days did not affect cumulative food intake (Fig. 2A), whereas body weight gain was significantly lower from day 2 to 7 of treatment (Fig. 2B). In parallel with the decreased weight gain, fat mass gain was also significantly lower in intracerebroventricular UGN–treated animals for 7 days compared with the control group (Fig. 2C). The chronic central treatment with UGN increased the fecal output corrected by grams of body weight with respect to controls (Fig. 2D). In agreement with the decreased weight gain, energy expenditure was increased when corrected by nonfat mass in intracerebroventricular UGN–treated animals versus controls (Fig. 2E and F), while locomotor activity (Fig. 2G) and RQ (Fig. 2H) remained unchanged.

Because energy expenditure was increased after chronic central UGN treatment, we next investigated whether BAT thermogenesis was responsible for the increased energy expenditure. Histological analyses revealed smaller lipid droplets in BAT of DIO mice after chronic central infusion of UGN (Fig. 3A). The results are consistent with biochemical data showing that protein levels of several thermogenic biomarkers, such as UCP1 and UCP3, PRDM16, and PGC1α, were significantly increased in the BAT of DIO mice chronically treated with intracerebroventricular UGN in comparison with vehicle-treated DIO mice (Fig. 3B). Other factors, such as CIDEA and FGF21, also tended to be increased in the BAT of UGN-treated DIO mice, although without statistical significance (Fig. 3B). The results obtained by Western blot were confirmed using different approaches. For instance, BAT UCP1 immunostaining was clearly increased in UGN-treated DIO mice in comparison with their controls (Fig. 3C). BAT gene expression of PRDM16 was also significantly increased in UGN-treated DIO mice, and PGC1α showed a clear tendency to augment when compared with vehicle-treated mice (Fig. 3D).

Because fat mass was reduced in DIO mice chronically treated with intracerebroventricular UGN, we next investigated the histological and molecular alterations within epididymal WAT. The chronic UGN treatment decreased the size of adipocytes compared with that in control animals (Fig. 3E), and, accordingly, protein levels of UCP1 (the main marker of browning) (Fig. 3F), UCP1 immunostaining (Fig. 3G), and phosphorylated HSL (pHSL), the main indicator of fatty acid oxidation, were significantly increased in the WAT of DIO mice after UGN treatment with respect to controls (Fig. 3H).

Vagal nerves regulate the physiological functions of the GI tract (22). Chronic central treatment of UGN increases fecal output in DIO mice (Fig. 2D), and we next sought to determine whether this effect involved vagal innervation. Intracerebroventricular UGN infusion was performed after dissections of both the dorsal and the ventral branches of the vagus nerve. The effectiveness of this approach was validated by assessing the expected morphological changes in the stomach (data not shown).

Central chronic UGN infusion did not affect food intake (Fig. 4A), but significantly decreased body weight (Fig. 4B) and fat mass (Fig. 4C) in both sham-operated mice and VGX mice. The fecal output was increased in sham-operated mice treated with UGN, but this effect was abolished in VGX mice (Fig. 4D). However, VGX did not blunt the effects of UGN on BAT, as demonstrated by histological analyses that revealed smaller lipid droplets (Fig. 4E and F) and decreased the levels of TGs in BAT (Fig. 4G) and increased UCP-1 protein levels (Fig. 4H) in BAT of both sham-operated and VGX mice after 7 days of UGN treatment. Therefore, the present data suggest that vagal innervation is not involved in the thermogenic effect of UGN in BAT; however, it seems to be directly related to UGN-induced increased fecal output.

β-ARs represent a key link involved in the regulation of adipose tissue metabolism by the sympathetic nervous system (SNS) (19). To determine whether the central UGN action on BAT was mediated by SNS, TKO (β1-AR, β2-AR, and β3-AR) mice were centrally infused with UGN for 7 days. UGN did not affect food intake in WT or TKO mice (Fig. 5A). UGN significantly decreased body weight gain (Fig. 5B) and fat mass gain (Fig. 5C) in WT mice, but not in TKO mice. The lack of β-ARs did not appear to affect GI function, as represented by the fact that UGN increased the fecal output in both WT and TKO mice (Fig. 5D).

In BAT, the implication of the adrenergic innervation in the mediation of the effects of UGN on BAT was corroborated by immunohistochemistry showing that central UGN reduced the size of the lipid droplets in WT mice, but not in TKO animals (Fig. 5E and F), and the reduced content of TGs in the BAT of WT mice treated with UGN was blunted in TKO mice (Fig. 5G). Consistently, the thermogenic effect of central UGN treatment was present in the WT mice, but not in the TKO mice, as demonstrated by the abolition of UGN-induced UCP-1 levels in BAT (Fig. 5H). The lack of effects of central UGN on WAT of TKO mice was also evident when assessed by morphological approaches (Fig. 5I) and by the lack of changes in UCP-1 protein levels in epididymal WAT of UGN-treated TKO mice (Fig. 5J).

The results obtained in TKO mice were corroborated by performing an acute pharmacological inactivation of the β3-AR with the specific antagonist SR59230A, which blunted the effect of intracerebroventricularly administered UGN on body weight 12 h after its administration in DIO mice without affecting feeding behavior (Fig. 6A and B). Histological analyses showed that 12 h after the central injection of UGN, the size, but not the number, of lipid droplets was reduced (Fig. 6C and D) and the content of TGs in BAT was also significantly lower than in intracerebroventricular vehicle–treated DIO mice (Fig. 6E). However, all the UGN-induced actions were blocked when animals were also treated with SR59230A. In line with this, the activation of BAT UCP-1 after central administration of UGN was also prevented by SR59230A (Fig. 6F).

The present work demonstrates that central UGN regulates energy balance by directly controlling peripheral metabolism in obese mice. Specifically, we show that chronic central infusion of UGN stimulates the thermogenic program in BAT, induces lipid mobilization and browning in WAT independent of feeding behavior, and augments fecal output. Those actions occur through effects on two different pathways: the SNS mediates the effects of central UGN on adipose tissue metabolism, whereas the parasympathetic vagus nerve controls the effects of central UGN on GI motility.

UGN has been proposed as a novel factor of the gut-brain axis involved in energy homeostasis regulation (5). However, its pharmacological long-term efficiency in obese animal models has remained unknown. The current study is, to the best of our knowledge, the first to show that central UGN infusion reduces body weight and adiposity in DIO mice independent of food intake. Indeed, many peripheral signals have been attributed to the regulation of energy expenditure via neuronal pathways, and increasing evidence indicates that BAT activity (23,24) and activation of beige/brown-in-white adipocytes in WAT, a process known as browning (25–27), are controlled by the CNS. Our findings suggest that the mechanisms by which central UGN infusion reduces body weight involve increased energy expenditure through promotion of BAT thermogenesis and the browning of WAT. In addition to playing a key role in the control of energy expenditure, the CNS can also modulate peripheral lipid metabolism (28,29). In this regard, we found that central UGN administration stimulates lipid mobilization, as shown by the increased levels of pHSL in WAT.

It is well known that the CNS control of BAT and WAT metabolism is directly mediated by the SNS (28,30,31). Thus, it was reasonable to hypothesize that the effects of central UGN on BAT and WAT involved this canonical pathway. Indeed, in TKO mice with TKO for the β1-AR, β2-AR, and β3-AR, CNS UGN failed to affect body weight and fat mass. Consistently, central UGN did not activate the thermogenic program in BAT, did not induce browning of WAT, and did not stimulate lipid mobilization in WAT of TKO mice. These findings observed in TKO mice were corroborated in WT mice fed an HFD, where pharmacological inhibition of the β3-AR blunted the acute effects of central UGN on body weight and BAT. Therefore, these results indicate that the sympathetic disruption abolishes the effects of CNS UGN on body weight and adiposity.

Although the chronic effects of central UGN are independent of food intake, previous reports showed controversial results regarding the effects of the UGN-GUCY2C system on feeding. Whereas one report (11) showed a decreased food intake 2 h after the central injection of STs, a subsequent study (12) failed to detect changes in food intake after the administration of ST or UGN. Similar to the first report (11), we found that UGN (25 µg i.c.v.) transiently decreased feeding and that this effect was amplified by fasting. Moreover, we also found that chronic central infusion of UGN decreases weight gain in mice fed a chow diet or an HFD. The discrepancies between the studies might be related to the different methodologies used. For instance, our study uses both lean and obese mice, whereas the study by Valentino et al. (11) used obese mice and the study by Begg et al. (12) used lean rats.

Another very important aspect that deserves attention is the potential link between the UGN-GUCY2C system and a diarrheagenic function (32). The reason for this is that GUCY2C is also the receptor for heat-ST that is responsible for acute diarrhea (6). For instance, linaclotide, a synthetic peptide that is structurally related to the endogenous guanylin peptide family, activates GUCY2C and is clinically used for chronic constipation (33). Consistent with this, our results indicate that central UGN increases fecal output. The subdiaphragmatic vagal nerve and its branches innervate the GI tract, and the extensive distribution of vagal nerves to the gut plays an important role in the regulation of many digestive functions, including gastric accommodation and GI motility (22). Therefore, we assessed the contribution of UGN-induced GI motility to the reduced weight gain and adiposity. Treatment with CNS UGN was performed in animals subjected to surgical VGX, and we found that UGN failed to affect fecal output. In VGX mice treated with central UGN, the effects on body weight and fat mass were only partially abolished, likely because its thermogenic effect was still active. Thus, we can conclude that the parasympathetic nervous system is mediating the GI effects of CNS UGN, but that it is not involved in the effects on adipose tissue metabolism.

In summary, the present work reveals for the first time that chronic central UGN infusion decreases body weight and adiposity in DIO mice independent of food intake via the following specific efferent autonomic pathways:

• 1) central UGN induces BAT thermogenesis through the SNS, as well as browning of and lipid mobilization in WAT and

• 2) CNS UGN augments fecal output through the parasympathetic nervous system.

These findings are of special relevance as they suggest that the beneficial metabolic actions of UGN through the SNS are clear-cut and still evident in animals exposed to an HFD.

Acknowledgments. The authors thank Cecilia Castelao (CIBER Fisiopatología de la Obesidad y Nutrición) for her excellent technical assistance.

Funding. This work has been supported by the Ministerio de Economía y Competitividad (grant BFU2011-29102 to C.D. and grant BFU2012-35255 to R.N.), Xunta de Galicia (grant 2012-CP070 to M.L. and grants EM 2012/039 and 2012-CP069 to R.N.), Fundación Mutua Madrileña, and Fondo de Investigaciones Sanitarias del Instituto de Salud Carlos III and cofunding by FEDER (European Regional Development Fund) (PI12/01814 to M.L. and PI12/02021 and PI15/01272 to L.M.S.) and CIBER Fisiopatología de la Obesidad y Nutrición (CIBERobn). CIBERobn is an initiative of the Instituto de Salud Carlos III (ISCIII) of Spain, which is supported by FEDER funds. The research leading to these results has also received funding from the Seventh Framework Programme of the European Community under the following grants: grant FP7/2007-2013 to C.D. and grant 245009, NeuroFAST, and European Research Council grants StG-2011 and OBESITY53-281408 to R.N. C.F. is funded by IDIS (Instituto de Investigación Sanitaria de Santiago de Compostela), and S.B.-F. is supported by Xunta de Galicia. O.A.-M. is funded by the ISCIII/SERGAS through research contract “Sara Borrell” (CD14/00091). L.M.S. is a researcher under grant I3SNS-SERGAS/ISCIII.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. C.F., D.B., A.C., O.A.-M., S.B.-F., and A.S. contributed to the experiments, the data analysis, the development of the analytical tools, and the discussion. J.F., M.L., C.D., F.F.C., and F.R.-J. contributed to the development of the analytical tools and the discussion. L.M.S. and R.N. contributed to the experimental design, the development of the analytical tools, the discussion, and the writing of the manuscript. L.M.S. and R.N. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.