Inhibition of P2Y6 Signaling in AgRP Neurons Reduces ... - Cell Press

Report

Inhibition of P2Y6 Signaling in AgRP Neurons Reduces Food Intake and Improves Systemic Insulin Sensitivity in Obesity Graphical Abstract

Authors Sophie Marie Steculorum, Katharina Timper, Linda Engstro¨m Ruud, ..., Stephan Bremser, Peter Kloppenburg, € ning Jens Claus Bru

Correspondence [email protected]

In Brief P2Y6 signaling was recently identified as regulator of AgRP neuron activity and food intake. Steculorum et al. now find that P2Y6 signaling in AgRP neurons remains functional in obese mice and that AgRP-neuron-restricted P2Y6 deficiency reduces food intake and fat mass, and improves systemic insulin sensitivity in obese mice. Their work thus defines P2Y6 as a target to regulate both hyperphagia and adiposity as well as systemic insulin resistance in obesity.

Highlights d

CNS P2Y6 signaling retains its orexigenic effect in obese mice

d

P2Y6 deficiency reduces fat mass and improves insulin sensitivity in obesity

d

AgRP-neuron-restricted P2Y6 deficiency reduces adiposity and feeding in obesity

d

AgRP-neuron-restricted P2Y6 deficiency improves systemic insulin sensitivity in obesity

Steculorum et al., 2017, Cell Reports 18, 1587–1597 February 14, 2017 ª 2017 Max Planck Institute for Metabolism Research. http://dx.doi.org/10.1016/j.celrep.2017.01.047

Cell Reports

Report Inhibition of P2Y6 Signaling in AgRP Neurons Reduces Food Intake and Improves Systemic Insulin Sensitivity in Obesity Sophie Marie Steculorum,1,2,3 Katharina Timper,1,2,3 Linda Engstro¨m Ruud,1,2,3 Nadine Evers,1,2,3 Lars Paeger,3,4 € ning1,2,3,5,6,* Stephan Bremser,3,4 Peter Kloppenburg,3,4 and Jens Claus Bru 1Department of Neuronal Control of Metabolism, Max Planck Institute for Metabolism Research, Gleueler Strasse 50, 50931 Cologne, Germany 2Center for Endocrinology, Diabetes and Preventive Medicine (CEDP), University Hospital Cologne, Kerpener Strasse 26, 50924 Cologne, Germany 3Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases (CECAD) and Center of Molecular Medicine Cologne (CMMC), University of Cologne, Joseph-Stelzmann-Strasse 26, 50931 Cologne, Germany 4Institute for Zoology, University of Cologne, Zu €lpicher Strasse 47b, 50674 Cologne, Germany 5National Center for Diabetes Research (DZD), Ingolsta ¨ dter Land Strasse 1, 85764 Neuherberg, Germany 6Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.celrep.2017.01.047

SUMMARY

Uridine-diphosphate (UDP) and its receptor P2Y6 have recently been identified as regulators of AgRP neurons. UDP promotes feeding via activation of P2Y6 receptors on AgRP neurons, and hypothalamic UDP concentrations are increased in obesity. However, it remained unresolved whether inhibition of P2Y6 signaling pharmacologically, globally, or restricted to AgRP neurons can improve obesity-associated metabolic dysfunctions. Here, we demonstrate that central injection of UDP acutely promotes feeding in diet-induced obese mice and that acute pharmacological blocking of CNS P2Y6 receptors reduces food intake. Importantly, mice with AgRP-neuron-restricted inactivation of P2Y6 exhibit reduced food intake and fat mass as well as improved systemic insulin sensitivity with improved insulin action in liver. Our results reveal that P2Y6 signaling in AgRP neurons is involved in the onset of obesity-associated hyperphagia and systemic insulin resistance. Collectively, these experiments define P2Y6 as a potential target to pharmacologically restrict both feeding and systemic insulin resistance in obesity. INTRODUCTION Controlling the obesity and type 2 diabetes mellitus (T2DM) epidemic has become a major challenge considering the billions of people suffering from these metabolic diseases and their associated disorders (Geiss et al., 2014; WHO, 2006). Over the last decades, our understanding of the fundamental homeostatic

processes governing energy balance and glucose homeostasis has largely evolved and pinpointed a pivotal role of the CNS and more particularly the arcuate nucleus of the hypothalamus (ARH) (Sohn et al., 2013; Varela and Horvath, 2012). The ARH notably contains orexigenic neurons co-expressing neuropeptide Y (NPY) and agouti-related peptide (AgRP) (Sternson and Atasoy, 2014). In addition to potently modulating feeding, AgRP neurons have been identified as key player in the acute regulation of systemic insulin sensitivity (Aponte et al., 2011; Gropp et al., 2005; Ko¨nner et al., 2007; Krashes et al., 2011; Luquet et al., 2005; Steculorum et al., 2016). Here, acute activation of AgRP neurons promotes not only hyperphagia but also glucose intolerance and insulin resistance (Steculorum et al., 2016). The ability of AgRP neurons to modulate both food intake and systemic insulin sensitivity, in concert with the welldescribed involvement of the melanocortin circuitry downstream to AgRP neurons in the etiology of obesity and T2DM in humans, highlights AgRP neurons as key target for the treatment of metabolic diseases (Cone, 2005; Farooqi and O’Rahilly, 2008). The idea of pharmacologically targeting AgRP neurons has been challenged by the discovery of the onset of neuronal resistance to their main endogenous regulators (i.e., insulin and leptin) in € ning, 2012; Vogt and obesity (Friedman, 2004; Ko¨nner and Bru € ning, 2013). Bru Lately, the discovery of regulators of AgRP neurons identified these pathways as putative targets for the control of this neuronal population in disease. We recently demonstrated that AgRP neurons express the purinergic receptor 6 (P2Y6) and that activation of P2Y6 signaling by its ligand uridine-diphosphate (UDP) in lean mice increases their firing rate and subsequently promotes feeding (Steculorum et al., 2015b). Furthermore, we revealed that hypothalamic UDP contents increased in obese and diabetic mice (Steculorum et al., 2015b). However, this study left the important questions unaddressed, whether P2Y6 signaling remains functional in obesity and, if so, whether it is involved in the onset of obesity associated metabolic disturbances.

Cell Reports 18, 1587–1597, February 14, 2017 ª 2017 Max Planck Institute for Metabolism Research. 1587 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Thus, in the current study, we investigated whether inhibition of UDP-dependent P2Y6 activation might indeed serve as a target for pharmacological intervention in obesity and/or obesity-associated insulin resistance. We demonstrate that the ability of centrally applied UDP to acutely promote feeding is retained in obese mice. Conversely, pharmacological blocking of P2Y6 activation in the CNS via intracerebroventricular (i.c.v.) application of a P2Y6 antagonist inhibits feeding in obese mice. Moreover, both conventional and AgRP-restricted P2Y6deficient animals exhibit reduced obesity as well as improved whole-body insulin sensitivity when exposed to high-fat-diet (HFD) feeding. Collectively, our results provide evidence that inhibition of P2Y6 signaling, particularly through its action on AgRP neurons, can provide a strategy for the treatment of obesity, obesity-associated hyperphagia and systemic insulin resistance. RESULTS AND DISCUSSION UDP/P2Y6 Signaling Promotes Feeding in Obese Mice We have previously shown that P2Y6 is highly enriched in the ARH in lean mice (Steculorum et al., 2015b). As P2Y6 is also expressed on activated microglia in the hippocampus (Koizumi et al., 2007) and in light of the notion that obesity promotes hypothalamic inflammation, including the activation of microglia (Gao et al., 2014), we investigated whether the expression pattern of P2Y6 in the hypothalamus of obese mice might differ from that previously reported in lean mice (Steculorum et al., 2015b). To study the expression pattern of P2Y6 in obese animals, we characterized the anatomical distribution of P2Y6 receptor in the hypothalamus of HFD-fed reporter mice, which express GFP under transcriptional control of the endogenous P2Y6 promoter (Figures 1A and 1B). This analysis revealed an unaltered regional distribution of P2Y6-dependent GFP expression in the hypothalamus of HFD-fed animals, where P2Y6 expression is highly enriched in the ARH as compared to other hypothalamic regions (Figure 1A). Moreover, as previously reported in lean mice, P2Y6 expression remains restricted to neurons in the ARH in obese mice as revealed by its colocalization with NeuroTrace-immunoreactive neurons (Figure 1B). This is consistent with our previous report that hypothalamic mRNA expression of P2ry6 remains unaltered both in obese and diabetic mice compared to control mice as assessed by real-time PCR analysis (Steculorum et al., 2015b). Here, we further reveal that P2ry6-mRNA expression in specific microdissected hypothalamic nuclei remains unchanged in obese mice fed a HFD as compared to control mice fed a normal chow diet (NCD) (Figure S1A). Since we had previously demonstrated that UDP can promote feeding in lean mice and that hypothalamic UDP concentrations are increased in obese animals, we aimed to investigate whether UDP indeed can also increase food intake in obese mice. Therefore, we assessed the food-intake stimulatory effect of i.c.v. applied UDP as the bona fide P2Y6 agonist in HFD-fed obese mice. This analysis revealed that, similar to what was observed in lean mice, also in HFD-fed animals UDP significantly increased food intake by z30% (Figures 1C, S1B, and S1C). To assess the contribution of UDP-dependent P2Y6 signaling on the steady-

1588 Cell Reports 18, 1587–1597, February 14, 2017

state regulation of feeding in obesity, we moreover investigated the effect of i.c.v. injection of the characterized P2Y6 antagonist MRS 2578 on food intake in HFD-fed obese mice. This analysis revealed that i.c.v. MRS 2578 injection suppressed spontaneous feeding by z40% (Figures 1D, S1D, and S1E). Collectively, these experiments indicate that the ability of UDP to promote feeding is retained in obesity and that the increased hypothalamic UDP content in obese mice potentially contributes, at least in part, to their hyperphagic phenotype. Validation of Mouse Models with Global or AgRP NeuronRestricted P2Y6 Deficiency Given that acute pharmacological inhibition of P2Y6 in the CNS acutely reduces steady-state food intake in obesity and that AgRP neurons regulate not only food intake but also systemic glucose homeostasis, we aimed to investigate whether chronic abrogation of UDP-dependent P2Y6 signaling in obesity may affect energy and glucose homeostasis in the long term. To this end, we compared the effect of HFD feeding in control animals to those lacking P2Y6 either globally (P2Y6D/D mice) or restricted to AgRP neurons (P2Y6DAgRP mice). Real-time PCR analyses of multiple tissues confirmed the expected absence of P2ry6-mRNA expression in all investigated tissues of conventional P2Y6D/D mice (Figure 2A). To obtain mice lacking functional P2Y6 expression only in AgRP neurons, we crossed mice carrying a loxP-flanked P2Y6-allele with mice expressing the Cre-recombinase under control of the AgRP-promoter. Further intercrosses yielded P2Y6flox/flox (control) and P2Y6flox/floxAgRPCre+/ mice, i.e., AgRPDP2Y6 mice. While P2ry6-mRNA expression in peripheral tissues of AgRPDP2Y6 mice remained unaltered compared to control mice, its hypothalamic expression displays a significant reduction as compared to that of control littermates (Figure 2B). These findings further support the notion that, among all hypothalamic neuronal cell type, expression of P2Y6 is highly enriched in AgRP neurons given the minor contribution of AgRP neurons to the total number of hypothalamic cells. To confirm that this decreased P2ry6-mRNA expression was indeed associated with a deletion of P2Y6 specifically in AgRP neurons, we further investigated the P2ry6-mRNA expression in control and AgRPDP2Y6 mice via fluorescent in situ hybridization (RNAscope). These analyses confirmed the expression of P2ry6 mRNA in AgRP-expressing neurons (Figure 2C). However, while dual-labeled cells for AgRP and P2ry6 mRNA as well as single-labeled, non-AgRP P2ry6-expressing cells were observed in the ARH of control animals, P2Y6D/D mice displayed a complete lack of P2ry6-mRNA expression. To the contrary, residual non-AgRP P2ry6-mRNA expressing cells were visualized in the arcuate nucleus of P2Y6DAgRP mice but no colocalization with AgRP expression was observed in P2Y6DAgRP mice (Figure 2C). To further provide functional validation of the deletion of P2Y6-sigaling in P2Y6DAgRP mice, we genetically marked AgRP neurons in AgRPDP2Y6 mice by crossing them to Cre-dependent dtTomato-reporter mice and investigated the ability of UDP to modulate action potential firing of identified AgRP neurons in control and AgRPDP2Y6 mice (Figure 2D). This analysis revealed that, similar to what we had previously reported, UDP

Figure 1. UDP/P2Y6 Signaling Promotes Feeding in Obese Mice (A) Representative microphotographs of GFP immunostaining (P2Y6-EGFP, green) and corresponding nuclear counterstaining (DAPI, blue) in the medio-basal hypothalamus containing the arcuate nucleus of the hypothalamus (ARH), the ventromedial nucleus of the hypothalamus (VMH), the dorsomedial nucleus of the hypothalamus (DMH), and the lateral hypothalamic area (LHA) of diet-induced obese P2Y6-EGFP mice. (B) Representative microphotographs of ARH depicting co-immunostaining of P2Y6-EGFP (P2Y6-EGFP, green) and fluorescent Nissl staining (NeuroTrace, neurons, red). (C and D) 2-hr food-intake measurement (depicted as food intake to body weight relative to control) after intracereborventricular (i.c.v.) administration of (C) UDP (100 mM) or vehicle (saline) (n = 18 versus 18) and (D) of the P2Y6-antagonist MRS 2578 (1 mM) or vehicle (DMSO 0,1%) (n = 14 versus 14). Scale bars, 150 mm. 3V, third ventricle. Data are presented as mean ± SEM, *p % 0.05 as determined by unpaired Student’s t test. See also Figure S1.

significantly increased action potential firing in AgRP neurons of control mice (Figure 2D). However, the ability of UDP to elicit increased action potential firing of AgRP neurons was abrogated in AgRPDP2Y6 mice (Figure 2D). Collectively, these data unequivocally demonstrate that (1) UDP’s ability to increase AgRP neuron activity depends on P2Y6 receptors expressed in these neurons and (2) AgRPDP2Y6 mice represent an appropriate model

to study the physiological consequences of AgRP-neuronrestricted P2Y6 deficiency. AgRP-Neuron-Restricted P2Y6 Deficiency Limits HFDInduced Hyperphagia and Adiposity To determine the long-term contribution of UDP-dependent P2Y6 activation to the manifestation of HFD-elicited alterations

Cell Reports 18, 1587–1597, February 14, 2017 1589

Figure 2. Mouse Models for the Study of Global or AgRP Neuron-Restricted P2Y6 Deficiency (A and B) Real-time qPCR analysis of pyrimidinergic receptor P2Y, G protein coupled, 6 (P2ry6) mRNA expression in hypothalamus (Hyp), white adipose tissue (WAT), liver and skeletal muscle (SKM) of (A) whole-body P2Y6-deficient mice (P2Y6 D/D) and their littermates control (n = 7–10 versus 9) and of (B) P2Y6 DAgRP and their littermates control (n = 6–17 versus 6–18).

(legend continued on next page)


Figure 3. AgRP Neuron-Restricted P2Y6 Deficiency Limits High-Fat-Diet-Induced Adiposity and Hyperphagia (A–H) Analysis of high-fat diet (HFD)-fed P2Y6 D/D (left panels) and P2Y6 DAgRP (right panels) and their respective control littermates. (A and B) Body weight curves from 3 to 18 weeks (A, n = 13 versus 12 and B, n = 11 versus 9) and naso-anal length at sacrifice (A, n = 12 versus 12 and B, n = 11 versus 9; 20 weeks). (C and D) Body-fat composition in percentages of lean and fat mass (C, n = 12 versus 12 and D, n = 11 versus 9; 20 weeks). (E and F) Quantification of adipocytes area in perigonadal white adipose tissue (E, n = 14 versus 17 and F, n = 13 versus 7; 20 weeks) and representative images of H&E staining (scale bar, 50 mm). (G and H) Cumulative 24-hr food intake (G, n = 13 versus 20 and H, n = 11 versus 17; 16 weeks). Data are represented as presented as mean ± SEM, p % 0.05 and ***p % 0.001 as determined by two-way ANOVA followed by Bonferroni post hoc test (A, B, G, and H) or two-tailed unpaired Student’s t test (C, D, E, and F).

in energy and glucose homeostasis, we next exposed both conventional and AgRP-neuron-restricted P2Y6-deficient mice and their respective control littermates to HFD feeding. Monitoring body weight development of these mice revealed no alterations of overall body weight between control and P2Y6D/D mice (Figure 3A), or between control and AgRPDP2Y6 mice (Figure 3B). Similarly, body length was indistinguishable between the different groups of mice, showing that P2Y6 deficiency does not alter somatic growth (Figures 3A and 3B). In contrast, when we assessed body composition of these animals

via magnetic resonance (MR) spectroscopy, this analysis revealed a significant reduction of fat mass in both P2Y6D/D mice and AgRPDP2Y6 mice compared to their controls (Figures 3C and 3D). Quantitative assessment of adipocyte size distribution in P2Y6D/D mice revealed a significant decrease of large adipocyte as compared to control (Figure 3E). Similar investigation in AgRPDP2Y6 mice showed an increased number of adipocytes of smaller size and a reduced number of large adipocytes (Figure 3F). Collectively, these experiments revealed that either whole-body or AgRPneuron-restricted P2Y6 deficiency protects from diet-induced adiposity. Next, we aimed to define through which alterations in energy balance P2Y6 deficiency protects from diet-induced adiposity. To this end, we assessed parameters of food intake, energy expenditure, and locomotor activity during an indirect calorimetry. This analysis revealed a tendency towards a decreased food intake in P2Y6D/D mice that did not reach statistical significance (Figures 3G and S2A). However, a significant decrease of steady-state food intake of z25% was detected in AgRPDP2Y6

(C) Representative microphotographs of RNAscope in situ hybridization of P2ry6-mRNA (red) and AgRP-mRNA (green) and of corresponding nuclear counterstaining (DAPI, blue) in the arcuate nucleus (ARH) of top row, control mice; middle row, P2Y6 D/D mouse; bottom row, P2Y6 DAgRP mouse. Arrows depict doublelabeled cells positive for P2ry6- and AgRP-mRNA and stars highlight cells that are P2ry6-mRNA positive but AgRP-mRNA negative. (D) Left: representative traces of AgRP neurons firing frequency in response to the application of 3 mM UDP in control mouse (upper panel) and in P2Y6 DAgRP mouse (lower panel). Middle: quantification of action potential (AP) frequency (control, upper panel n = 21; P2Y6 DAgRP bottom panel, n = 20; responsive and nonresponsive pooled). Light-gray and dark-gray circles mark, respectively, single recordings responding with a significant increase or decrease in AP frequency and open circles are non-responsive neurons. Right: overall responsiveness to 3 mM UDP of the recorded AgRP neuron population. All recordings have been conducted in synaptically isolated neurons. Scale bars, 50 mm. Data are represented as presented as mean ± SEM and *p % 0.05 and ***p % 0.001 as determined by unpaired Student’s t (A and B) and as boxplots generated according to the ‘‘Tukey method’’ (mean: ‘‘+,’’ median: horizontal line) (D), *p % 0.05 as determined by paired Student’s t test.


mice compared to their controls (Figures 3H and S2B). However, we found that these changes in food intake were not associated with significant alterations in mRNA expression of key hypothalamic neuropeptides (Figures S2C and S2D). In addition, to complement the characterization of the conditional P2Y6 deletion in AgRP neurons, we also investigated the influence of P2Y6 deletion in this neuronal population on AgRP fiber density innervating the paraventricular nucleus of the hyphothalamus (PVH). These analyses revealed that P2Y6 deletion does not influence the AgRP fiber density projecting to the PVH (Figure S2E). While food intake was decreased in those mice, on the other hand locomotor activity, energy expenditure, and respiratory exchange ratio remained unaltered both in conventional and AgRP-neuron-specific P2Y6-deficient mice (Figures S3A–S3H). Taken together, these experiment clearly indicate that inhibition of P2Y6 action in AgRP neurons is sufficient to reduce hyperphagia associated with high-fat-diet feeding. P2Y6 Action in AgRP Neurons Contributes to HFDInduced Deterioration of Peripheral Insulin Sensitivity Having identified a critical role for P2Y6-dependent activation of AgRP neurons for the development of HFD-induced hyperphagia and adiposity, we next aimed to investigate the effect of global or AgRP-neuron-restricted P2Y6 deficiency on systemic glucose homeostasis in obesity. Therefore, we first performed glucose tolerance tests in P2Y6D/D mice and AgRPDP2Y6 mice compared to their controls. This analysis revealed a similar degree of glucose intolerance between the different groups of animals (Figures 4A and 4B). In contrast, when we assessed systemic insulin sensitivity in the different groups of mice, both P2Y6D/D mice and AgRPDP2Y6 mice exhibited a similar degree of significantly improved insulin sensitivity compared to their HFD-fed control littermates (Figures 4C and 4D). To determine in which target tissue insulin action improved upon AgRP-neuron-restricted P2Y6 deficiency, we next investigated insulin’s ability to activate its downstream signaling pathway in skeletal muscle (SKM), brown adipose tissue (BAT), and liver of HFD-fed AgRPDP2Y6 mice and their control littermates (Figures 4E and S5). This analysis revealed that the ability of insulin to phosphorylate AKT was not altered in SKM and BAT of AgRPDP2Y6 mice as compared to littermate controls. However, a clear increase in the ability of insulin to promote AKT-phosphorylation was observed in the liver of AgRPDP2Y6 mice compared to control mice (Figure 4E). Collectively, our experiments indicate that P2Y6 acts in AgRP neurons to contribute not only to obesity-induced hyperphagia and adiposity, but also to the deterioration of systemic insulin resistance primarily via inhibiting hepatic insulin sensitivity. Conclusions Herein, we demonstrate the critical importance of P2Y6 signaling in the onset of diet-induced obesity and associated metabolic outcomes. Extending on our previous studies, which revealed an acute food-intake regulatory function of UDP in the hypothalamus, here we report that selective P2Y6 deficiency in AgRP neurons prevents diet-induced hyperphagia, adiposity, and insulin resistance in the long term (Figure S4). The fact that AgRP-specific P2Y6 deletion recapitulates all the metabolic


improvements observed in whole-body P2Y6-deficient mice provides a clear evidence of the fundamental role of P2Y6-dependent signaling in AgRP neurons in the onset of obesity-associated metabolic alterations, such as obesity-induced adiposity, hyperphagia, and also insulin resistance. Indeed, in line with a recent report highlighting the key role of AgRP neurons in the control of systemic insulin sensitivity (Steculorum et al., 2016), the present study demonstrates that, in addition to modulating feeding, P2Y6 signaling in AgRP neurons also plays a critical role in peripheral insulin sensitivity. Protection against dietinduced systemic insulin resistance in AgRP-specific P2Y6-deficient mice is associated with an improvement of hepatic insulin sensitivity, which is in line with the well-described role of chronically altered insulin action in AgRP neurons to control hepatic insulin action and glucose production (Ko¨nner et al., 2007). Although previous studies largely conducted in cultured mouse skeletal muscle cells and adipocytes in vitro had indicated that synthetic P2Y6-agonists can stimulate glucose uptake in myocytes and adipocytes (Balasubramanian et al., 2014), the clear improvement of systemic insulin sensitivity of conventional P2Y6-deficient mice in the setting of HFD-induced obesity argues against a physiologically relevant role for UDP to control peripheral glucose uptake in a P2Y6-dependent manner in obesity. We previously reported the existence of a UDP-uridine loop in which hypothalamic UDP contents are increased in obesity as a direct consequence of elevated uridine supply from the periphery (Steculorum et al., 2015b). Consistent with this model, chronic uridine supplementation in mice leads to glucose intolerance associated with decreased insulin signaling in liver and increase hepatic gluconeogenesis (Urasaki et al., 2016). These data in conjunction with the results of the present study suggest that increased hypothalamic UDP levels, as a result of elevated peripheral supply of uridine, directly act on AgRP neurons to control hepatic insulin sensitivity and to deteriorate systemic insulin sensitivity and glucose homeostasis. Interestingly, circulating uridine concentrations correlate with insulin resistance in humans (Dudzinska et al., 2013; Hamada et al., 2007), providing evidence that similar regulatory mechanism may also exist in humans. In addition to the beneficial effects on obesity-associated adiposity and hyperphagia reported here, P2Y6 deficiency has been shown previously to blunt macrophage-inflammatory responses (Bar et al., 2008; Riegel et al., 2011) and notably to limit atherosclerosis development (Garcia et al., 2014; Guns et al., 2010; Stachon et al., 2014). Collectively, these results define P2Y6 as potential therapeutic targets to tackle several aspects of the metabolic syndrome such as hyperphagia, adiposity, and systemic insulin resistance as well as low-grade inflammation and cardiovascular inflammation. EXPERIMENTAL PROCEDURES Animal Care All animal procedures were conducted in compliance with protocols approved by local government authorities (Bezirksregierung Ko¨ln) and were in accordance with NIH guidelines. Mice were housed at 22 C–24 C using a 12-hr-light/12-hr-dark cycle (dark cycle at 06:00 p.m.). Unless otherwise stated, animals were fed with a maintenance diet (R/M-H; Ssniff Diet) containing 58% carbohydrates, 33% protein, and 9% fat. Animals had ad libitum

Figure 4. AgRP-Neuron-Restricted P2Y6 Deficiency Improves Obesity-Associated Insulin Resistance (A and B) Glucose tolerance tests (GTT) performed in 12-week-old HFD-fed (A) P2Y6 D/D and control mice (n = 15 versus 12) and (B) P2Y6 DAgRP mice and control mice (n = 14 versus 16). (C and D) Insulin tolerance tests (ITT) in 13-week-old HFD-fed (C) P2Y6 D/D and control mice (n = 16 versus 21) and (D) P2Y6 DAgRP mice and control mice (n = 16 versus 17). Area under the curves (AUCs) of each GTT and ITT are plotted next to the corresponding curves. (E) Quantification and representative immunoblots of phosphorylated and total AKT as well as respective calnexin loading controls in skeletal muscle (SKM; n = 4 versus 4), brown adipose tissue (BAT; n = 5 versus 6), and liver (n = 5 versus 6) of 11-week-old mice fed a HFD. Data are represented as mean ± SEM. *p % 0.05 and **p % 0.01 as determined by two-way ANOVA followed by Bonferroni post hoc test for the curves or unpaired one-tailed Student’s t test.


access to water. Mice were housed in groups of three to five animals unless stated otherwise. C57BL/6N mice were purchased from Charles River Laboratories. All experiments were performed in adult mice at the indicated ages. Unless otherwise stated, experiments have been performed in male animals. Mouse Models Genetic Mouse Models P2Y6-EGFP and P2Y6-deficient mice have been previously described (Bar et al., 2008). P2Y6-EGFP were bred and used for experiments homozygously. P2Y6 conventional knockout mice (P2Y6 D/D) were mated heterozygous (P2Y6 D/WT), therefore allowing us to perform experiments in P2Y6 D/D and their control wild-type littermates (P2Y6 WT/WT, denoted as control). To generate AgRPDP2Y6 mice, we crossed mice carrying a loxP-flanked P2Y6allele (P2Y6flox/flox) described previously (Bar et al., 2008) with mice expressing the recombinase under the control of the AgRP-promoter (AgRP-IRESCre mice, AgRPCre, (Tong et al., 2008). Mice used for experiments were P2Y6flox/flox (control) and P2Y6flox/floxAgRPCre+/ mice, i.e., AgRPDP2Y6 mice. To generate P2Y6flox/floxAgRPCre+/ mice expressing the tdTomato reporter mice homozygously carrying the B6;129S6-Gt(ROSA)26Sortm9(CAG-tdTomato)Hze/J (The Jackson Laboratories) allele were crossed to AgRPDP2Y6 mice. Used mice were P2Y6flox/flox; tdTomatotg/WT, AgRPCretg/wt. Diet-Induced Obese Mice Mice were fed a high-fat diet (HFD; D12492-(I); Ssniff Diet) containing 26.2% carbohydrates, 26.3% protein, and 34.9% fat (60% of calories from fat) or corresponding normal chow diet (NCD; D12450 (B); Ssniff Diet) containing 67.3% carbohydrates, 19.2% protein, and 4.3% fat (9% of calories from fat). C57BL/6N male mice were fed a HFD or a NCD starting at 8 weeks of age, and, for phenotyping experiments, P2Y6 D/D, AgRPDP2Y6 mice and their respective control littermates were weaned on HFD (i.e., 3 weeks of age). Intracerebroventricular Injections Intracerebroventricular Cannulation Intracerebroventricular (i.c.v.) cannulations were performed as described elsewhere (Steculorum et al., 2015b). Briefly, adult mice were anesthetized using a ketamine-dexmedetomidine combination and were placed into a stereotaxic apparatus. 26G cannulas were implanted to the left ventricle using appropriate coordinates (bregma, –0.2 mm; midline, +1 mm; dorsal surface, –2.1 mm). Dental acrylic was used to secure the cannula to the skull. Animals were allowed to recover for a week prior to experiment. After experiments, stereotaxic implantations were anatomically verified. Compound Administration i.c.v. injections of UDP were performed by administration of 2 mL of 100 mM UDP (Sigma) or vehicle (saline). MRS 2578 (Sigma) was dissolved in dimethyl sulfoxide (DMSO) (final concentration: 0.1% DMSO in saline). For i.c.v. administration of MRS 2578, mice received either MRS 2578 (2 mL, 1 mM) or vehicle (0.1% DMSO, 2 mL). Food-Intake Measurement following i.c.v. Administration Mice were single caged and acclimated to food racks for a week prior to measurements. Four hours before the onset of the dark phase, mice were provided with fresh cages to avoid leftover of food spilling in the bedding. Food intake was measured 2, 4, 12, and 24 hr post-administration from pre-weighed food racks. All injections have been performed 30 min before the onset of the dark phase. Food-intake measurements were performed in a crossover fashion, in which each animal received a different treatment weekly. Each animal received a maximum of four different treatments, and a washout period of 1 week was allowed between each experiment. Body weight was recorded weekly. Metabolic Phenotyping Body Weight, Length, and MR Spectrometry Body weight was assessed weekly from 3 to 18 weeks of age. Naso-anal length was measured at sacrifice (i.e., 20 weeks). Nuclear magnetic resonance was employed to determine whole-body composition of 20-week-old live animals using the NMRAnalyzer minispeq mq7.5 (Bruker Optik).


Glucose and Insulin Tolerance Test Glucose tolerance tests were carried out on 12-week-old mice after a 6-hr fast during the morning. After determination of fasted blood glucose levels, animals were injected i.p. with a bolus of 2 mg/g body weight of glucose (20% glucose, Delta Select), and blood glucose levels were monitored 15, 30, 60, and 120 min after injection. Insulin tolerance tests were performed with 13-week-old animals fed ad libitum. After measuring basal blood glucose levels, each animal was intraperitoneally (i.p.) injected with 0.75 U/kg body weight of insulin (Actrapid; Novo Nordisk A/S). Blood glucose levels were recorded 15, 30, and 60 min after injection. Metabolic Chambers For basal food intake, locomotor activity, and energy expenditure, 16-weekold mice were placed in the PhenoMaster System (TSE systems) as previously described (Mauer et al., 2014). Data regarding energy expenditure and respiratory exchanges were corrected for lean mass. Insulin Signaling Insulin Administration 11-week-old mice were deeply anesthetized and the abdominal cavity of the mice was opened, and 100-mL samples containing 5 U regular human insulin (Actrapid, Novo Nordisk) diluted in 0.9% saline were injected into the inferior vena cava. Sham injections were performed with 100 mL of 0.9% saline. Samples of liver, skeletal muscle (SKM), and brown adipose tissue (BAT) were harvested 5 min after injection and snap frozen in liquid nitrogen. Western Blot Proteins were extracted from tissues for western blot analysis using a standard western blot protocol. Briefly, after defrosting of the samples on ice, samples were homogenized in protein lysis buffer (20 mM Tris-HCL [pH 7.4], 140 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 10% glycerol, 10% Np-40, and protease and phosphatase inhibitors cocktail from Roche). Determination of the protein concentrations was performed by measuring the absorbance at 562 nm using the Pierce BCA Protein Assay Kit (Thermo Scientific) according to the manufacturer’s instructions. Lysates were diluted to equal concentrations micrograms per microliters in a total volume of 100 mL 1 3 Laemmli buffer (Bio-Rad). Equal protein amounts (i.e., 10 mg) from each sample were loaded onto SDS-PAGE gels and run at 100 V for approximately 15 min and then at 200 V for 45 min. Gels were blotted onto polyvinylidene fluoride (PVDF) membranes using a semi-dry transfer cell (Bio-Rad) and 75 mA per gel for 60 min. Membranes were blocked with 1% blocking solution (1% western blocking reagent, Roche, 1 3 TBS) for 1 hr and incubated with primary antibody overnight at 4 C (pAKT Ser 473, 1:1,000, Cell Signaling Technology #9271; AKT, 1:1,000 Cell Signaling Technology #11E/7; Calnexin, 1:1,000, Cell Signaling Technology #208880). After primary incubations, blots were washed four times for 20 min 0.1% Tween 20 in 1 3 TBS and incubated with the secondary antibody for 1 hr. After washing four times for 5 min, signal was detected using the Pierce ECL western blotting kit and an automatic developer (Chemiluminescence Imaging Fusion SOLO; Vilber Lourmat). Analysis of Gene Expression For gene analysis, all mice were sacrificed by decapitation. P2Y6 DAgRP, P2Y6 D/D mice fed a HFD and their respective control littermates were sacrificed 20 weeks in a random fed state. Microdissection analysis was performed as described previously (Steculorum et al., 2015b). Briefly, 20-week-old mice fed an HFD or NCD were sacrificed, and the brain was quickly removed. 1-mm sections were prepared using Brain Matrices (Braintree Scientific) and PVH, ARH, ventromedial nucleus of the hypothalamus (VMH), dorsomedial nucleus of the hypothalamus (DMH), and lateral hypothalamic area (LHA) were carefully microdissected under visual guidance using a binocular microscope. Microdissected regions obtained from two animals were pooled. White adipose tissue (WAT), liver and skeletal muscle (SKM) from P2Y6 D/D, AgRPDP2Y6 mice and their respective control littermates were collected after sacrifice. Isolated mRNA from tissues was analyzed using real-time qPCR. mRNA of microdissected regions was isolated using the Arcturus Picopure kit (Applied Biosystems), the mRNA of the liver was isolated using the QIAGEN RNeasy Kit (QIAGEN), and the mRNA of the hypothalamus was isolated using the miRVana kit, whereby all were combined with the RNase-Free DNase Set (QIAGEN). RNA

was reversely transcribed with high-capacity cDNA RT Kit and amplified using TaqMan Universal PCR-Master Mix, NO AmpErase UNG with TaqMan Assayon-demand kits (Applied Biosystems). Relative expression of target mRNAs was adjusted for total RNA content by Hprt (Mm01545399_m1). The following inventoried TaqMan probe were used: P2ry6 (Mm02620937_s1), Agrp (Mm00475829_g1), Npy (Mm03048253_m1), Pomc (Mm00435874_m1), Cart (Mm00489086_m1), Pmch (Mm01242886_g1), Hcrt (Mm01964030_s1), Trh (Mm01182425_g1), and Crh (Mm01293920_s1). Calculations were performed by a comparative method (2 DDCT). qPCR was performed on an AB-QuantStudio7Flex (Applied Biosystems). Histology P2Y6-EGFP Expression Expression pattern of P2Y6 was performed as described elsewhere (Steculorum et al., 2015b). Briefly, adult mice were anesthetized and perfused transcardially with 4% paraformaldehyde. Brains were frozen and sectioned at 30 mm on a freezing sliding microtome and processed for immunofluorescence as follow: after an overnight blocking step (2% serum, 0.3% triton), sections were incubated for 48 hr in primary antibody (chicken anti-GFP 1:1,000, Abcam #13970). The primary antibodies were localized with corresponding affinity-purified immunoglobulin Gs (IgGs) conjugated with Alexa fluorophores (Life Technologies, 1:400). Sections were incubated with NeuroTrace (1:300) and were mounted in DAPI-containing Vectashield (Vector Laboratories). Pictures were acquired using a confocal Leica TCS SP-8-X microscope equipped with a 203 objective. RNAscope-In Situ Hybridization Mice were deeply anesthetized and perfused transcardially with 0.9% saline followed by ice-cold phosphate buffered 4% paraformaldehyde (PFA, pH 7.4). Brains were removed, post-fixed for 18 hr at room temperature in 4% PFA and then transferred to 20% sucrose in 0.1 M PBS (pH 7.4) for 24 hr at 4 C. 14-mm-thick sections were cut on a freezing microtome and subsequently stored at 20 C in sterile antifreeze solution (30% ethylene glycol and 20% glycerol in PBS). Every 12th section throughout the ARH was processed for immunohistochemistry as described below. Fluorescent in situ hybridization for the simultaneous detection of the P2ry6 and the AgRP transcripts was performed using RNAscope (ACD; Advanced Cell Diagnostics). A custom-designed probe was made targeting the floxed region (exon 3 [Bar et al., 2008]) of the P2ry6 transcript (target region: 401–1379, accession number NM_183168.2; ACD). The AgRP probe targeted the region 11–764 (accession number NM_001271806.1; ACD). Negative and positive control probes recognizing dihydrodipicolinate reductase, DapB (a bacterial transcript) and cyclophilin and PolR2A, respectively, were processed in parallel with the target probes to ensure tissue RNA integrity and optimal assay performance. All pretreatment reagents, detection kit, and wash buffer were purchased from ACD. All incubation steps were performed at 40 C using a humidified chamber and a HybEz oven (ACD). On the day before the assay, sections were mounted on SuperFrost Plus Gold Slides (Thermo Scientific), allowed to dry, and then were briefly dipped in diethyl-pyrocarbonate (DEPC)-treated Millipore water and dried at 60 C overnight. On the day of the assay, slides were first submerged in Target Retrieval (ACD) heated to 98.5 C–99.5 C for 8 min, followed by two brief rinses in DEPC water. The slides were quickly dehydrated in 100% ethanol and allowed to air dry for 5 min. The sections were then incubated with Protease III for 30 min (ACD), the protease was renewed after this time, and slides were further incubated for 15 more minutes. The subsequent steps, i.e., hybridization of the probe, the amplification steps, and detection of the probe, were performed according to the online protocol for RNAscope Multifluorescent Assay. In brief, the procedure included the following steps: the P2ry6 probe (channel 1) and the AgRP probe (channel 2) were mixed according to the manufacturer’s instructions and hybridized to the sections for 2 hr, followed by 2 3 2 min washes in wash buffer (ACD), incubation with Amp1 for 30 min, two washes, Amp2 for 15 min, two washes, Amp3 for 30 min, two washes, and finally Amp4C for 15 min followed by two washes. Sections were then immediately coverslipped with ProLong Gold Antifade Mountant with DAPI and stored in dark at 4 C until imaging. The P2ry6 probe was detected with Atto 550 and the AgRP probe with Atto 647.

Images were captured using a confocal Leica TCS SP-8-X microscope, equipped with a 403 objective. Maximum intensity projections were made in FIJI (NIH), and images were adjusted for brightness and contrast. White Adipose Tissue Histology Perigonadal fat depots were dissected from 20-week-old animals after decapitation. Dissected tissue samples were fixed in 4% paraformaldehyde, embedded in paraffin (FFPE), and sectioned as previously described (Mauer et al., 2010). H&E staining was carried out after deparaffination, and images were captured using the Zeiss Imager M2 microscope equipped with a 103 objective. Quantification of adipocyte size was performed using an automated software Adiposof plug-in form FIJI as described elsewhere (Galarraga et al., 2012). AgRP Fibers Quantification AgRP Immunohistochemistry Quantification of AgRP fiber density was performed as previously described (Steculorum et al., 2015b). Briefly, P2Y6 DAgRP mice fed a HFD and their respective control (9–12 weeks, random fed, n = 6 per group) were perfused transcardially with 4% paraformaldehyde; the brains were then frozen and sectioned at a 30-mm thickness and processed for immunofluorescence using standard procedures. The primary antibodies used for immunohistochemistry (IHC) included rabbit anti-AgRP (1:2,000, Phoenix Pharmaceuticals), which was visualized with Alexa Fluor 488 goat anti-rabbit IgGs (1:400, Invitrogen). Quantitative Analysis of Fiber Density For the histological experiments, two sections at the same rostro-caudal level of the PVH from animals of each experimental group (n = 6 animals per group) were acquired using a LeicaSP8 confocal system equipped with a 203 objective. Quantification of AgRP fiber analysis was performed as described previously (Steculorum et al., 2015a). Image analysis was performed using ImageJ analysis software (NIH). For the quantitative analysis of fiber density each image plane was binarized to isolate-labeled fibers from the background and to compensate for differences in fluorescence intensity. The integrated intensity, which reflects the total number of pixels in the binarized image, was then calculated for each image. The resulting value is an accurate index of the density of the processes in the volume sampled. Electrophysiology Animals and Brain Slice Preparation UDP response experiments were performed on brain slices from adult female and male P2Y6flox/flox; tdTomatotg/WT, AgRPCretg/wt, and AgRPtdTomato. The animals were anesthetized with halothane and subsequently decapitated. The brain was rapidly removed and a block of tissue containing the hypothalamus was immediately cut out. Coronal slices (275 mm) containing the arcuate nucleus of the hypothalamus were cut with a vibration microtome under cold (4 C), carbogenated (95% O2 and 5% CO2), glycerol-based modified artificial cerebrospinal fluid (GaCSF; Ye et al., 2006) to enhance the viability of neurons. GaCSF contained (in mM): 250 glycerol, 2.5 KCl, 2 MgCl2, 2 CaCl2, 1.2 NaH2PO4, 10 HEPES, 21 NaHCO3, 5 glucose adjusted to pH 7.2 (with NaOH) resulting in an osmolarity of 310 mOsm. Brain slices were transferred into carbogenated artificial cerebrospinal fluid (aCSF). First, they were kept for 20 min in a 35 C ‘‘recovery bath’’ and then stored at room temperature (24 C) for at least 30 min prior to recording. aCSF contained (in mM): 125 NaCl, 2.5 KCl, 2 MgCl2, 2 CaCl2, 1.2 NaH2PO4, 21 NaHCO3, 10 HEPES, and 5 glucose adjusted to pH 7.2 (with NaOH) resulting in an osmolarity of 310 mOsm. Perforated Patch Recordings Slices were transferred to a recording chamber (3 mL volume) and continuously superfused with carbogenated aCSF at a flow rate of 2.5 mL 3 min–1. Perforated patch recordings were performed using protocols modified from Akaike and Harata (1994) and Horn and Marty (1988). Electrodes with tip resistances between 5 and 7 MU were fashioned from borosilicate glass (0.86-mm inner diameter; 1.5-mm outer diameter) with a vertical pipette puller (PP-830; Narishige). Perforated patch recordings were performed with ATP and GTP free pipette solution containing (in mM): 128 K-gluconate, 10 KCl, 10 HEPES, 0.1 EGTA, 2 MgCl2, and adjusted to pH 7.3 (with KOH) resulting in an osmolarity of 300 mOsm. To label single cells, 1% (w/v) biocytin was added to the pipette solution. ATP and GTP were omitted from the intracellular solution to prevent uncontrolled permeabilization of the cell membrane (Lindau and


Fernandez, 1986). The patch pipette was tip filled with internal solution and back filled with amphotericin B (150–200 mg 3 mL–1)-containing internal solution to achieve perforated patch recordings. Amphotericin B was dissolved in DMSO (Rae et al., 1991). Either substance was added to the modified pipette solution shortly before use. The DMSO concentration had no obvious effect on the investigated neurons. During the perforation process access resistance (Ra) was constantly monitored, and experiments were started after Ra had reached steady state (15 – 20 min) and the action potential amplitude was stable. To check the integrity of the perforated patch, we monitored Ra tdTomato fluorescence of the neurons and added Dextran-Fluorescein (0.02% [w/v]; 3,000 molecular weight [MW]) to the pipette solution. Patch-Clamp Recordings Neurons in the ARH were visualized with a fixed-stage upright microscope (BX51WI; Olympus), using a 603 water immersion objective (LUMplan FI/IR; 603; 0.9 numerical aperture; 2-mm working distance; Olympus) with infrared-differential interference contrast and fluorescence optics (Dodt and Zieglga¨nsberger, 1990). AgRP neurons were identified by their tdTomato expression. Patch-clamp recordings were performed with an EPC10 patch-clamp amplifier (HEKA) controlled by the program PatchMaster (v.2.52; HEKA) running under Windows. Data were sampled at 10 kHz and low-pass filtered at 2 kHz with a four-pole Bessel filter. The calculated liquid junction potential of 14.6 mV between intracellular and extracellular solution was compensated (calculated with Patcher’s Power Tools plug-in from http://www3.mpibpc. mpg.de/groups/neher/index.php?page=software for Igor Pro 6 [Wavemetrics]). After the electrophysiological characterization in the perforated patchclamp configuration, we established the whole-cell configuration to load the neuron with biocytin. Standard histological and immunohistochemical methods were used to label the biocytin-loaded neuron and the tdTomato-expressing neurons. Drugs For current-clamp recordings drugs were bath-applied at a flow rate of 2.5 mL 3 min–1. 3 mM UDP was added to the normal aCSF shortly before the experiments. To block synaptic currents and exclude network effects, the aCSF contained 10–4 M pertussis toxin (PTX), 5 3 10–5 M D-AP5, an 10–5 M 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX). Data Analysis of the Patch-Clamp Recordings Data analysis was performed with Spike2 (CED) and GraphPad PRISM (GraphPad). We found that the basic firing properties of AgRP neurons and their responsiveness to UDP were not homogeneous. Therefore, we used the ‘‘three times SD’’ criterion, and a neuron was considered UDP responsive if the change in action potential firing frequency induced by UDP was three times larger than the SD (Dhillon et al., 2006; Kloppenburg et al., 2007). For each neuron, time frames from 6 to 0 min and 9 to 15 min after UDP application were each divided into 12 bins of 30-s intervals to calculate means and respective SDs of action potential frequencies. To determine differences in means between two treatments, paired t tests were performed. For more than one treatment, one-way ANOVA with post hoc pairwise comparisons (t tests with the Newman-Keuls method for p value adjustment) were performed. Boxplots are generated according to the ‘‘Tukey method,’’ and means are reflected by ‘‘+’’ and medians by horizontal line, respectively. Statistical Analyses All values were expressed as the means ± SEM. Statistical analyses were conducted using GraphPad PRISM (v.5.0a). Datasets with only two independent groups were analyzed for statistical significance using unpaired two-tailed Student’s t test unless otherwise stated. Datasets subjected to two independent factors were analyzed using two-way analysis of variance (ANOVA) followed by Bonferroni post hoc test. All p values below or equal 0.05 were considered significant. *p % 0.05, **p % 0.01, and ***p % 0.001.

SUPPLEMENTAL INFORMATION Supplemental Information includes five figures and can be found with this article online at http://dx.doi.org/10.1016/j.celrep.2017.01.047.


AUTHOR CONTRIBUTIONS S.M.S. and J.C.B. conceived the project, designed the experiments, analyzed the data, and wrote the manuscript with input from the other authors. K.T. performed surgeries and contributed to the food-intake experiments. L.E.R. performed RNAscope experiments. N.E. provided technical assistance. L.P., S.B., and P.K. performed and analyzed electrophysiological recordings. ACKNOWLEDGMENTS We are extremely grateful to Marko Idzko for providing us with the P2Y6 wholebody and P2Y6-flox mice. We thank A¨nne Lautenschlager and Jens Alber for outstanding technical assistance. This work was supported by a grant from the DFG (BR 1492/7-1) to J.C.B., and received funding by the DFG within the framework of the TRR134 and within the Excellence Initiative by German Federal and State Governments (CECAD). This work was funded (in part) by the Helmholtz Alliance ICEMED -Imaging and Curing Environmental Metabolic Diseases, through the Initiative and Networking Fund of the Helmholtz Association. Moreover, the research leading to these results has received funding from the European Union Seventh Framework Program (FP7/2007-2013) under grant agreement no. 266408. S.M.S. was funded by the Humboldt-Bayer program of the Alexander von Humboldt Foundation and received a grant from CECAD. Received: October 24, 2016 Revised: January 6, 2017 Accepted: January 19, 2017 Published: February 14, 2017 REFERENCES Akaike, N., and Harata, N. (1994). Nystatin perforated patch recording and its applications to analyses of intracellular mechanisms. Jpn J Physiol. 44, 433–473. Aponte, Y., Atasoy, D., and Sternson, S.M. (2011). AGRP neurons are sufficient to orchestrate feeding behavior rapidly and without training. Nat. Neurosci. 14, 351–355. Balasubramanian, R., Robaye, B., Boeynaems, J.M., and Jacobson, K.A. (2014). Enhancement of glucose uptake in mouse skeletal muscle cells and adipocytes by P2Y6 receptor agonists. PLoS ONE 9, e116203. Bar, I., Guns, P.J., Metallo, J., Cammarata, D., Wilkin, F., Boeynams, J.M., Bult, H., and Robaye, B. (2008). Knockout mice reveal a role for P2Y6 receptor in macrophages, endothelial cells, and vascular smooth muscle cells. Mol. Pharmacol. 74, 777–784. Cone, R.D. (2005). Anatomy and regulation of the central melanocortin system. Nat. Neurosci. 8, 571–578. Dhillon, H., Zigman, J.M., Ye, C., Lee, C.E., McGovern, R.A., Tang, V., Kenny, C.D., Christiansen, L.M., White, R.D., Edelstein, E.A., et al. (2006). Leptin directly activates SF1 neurons in the VMH, and this action by leptin is required for normal body-weight homeostasis. Neuron 49, 191–203. Dodt, H.U., and Zieglga¨nsberger, W. (1990). Visualizing unstained neurons in living brain slices by infrared DIC-videomicroscopy. Brain Res. 537, 333–336. Dudzinska, W., Lubkowska, A., Jakubowska, K., Suska, M., and Skotnicka, E. (2013). Insulin resistance induced by maximal exercise correlates with a postexercise increase in uridine concentration in the blood of healthy young men. Physiol. Res. 62, 163–170. Farooqi, I.S., and O’Rahilly, S. (2008). Mutations in ligands and receptors of the leptin-melanocortin pathway that lead to obesity. Nat. Clin. Pract. Endocrinol. Metab. 4, 569–577. Friedman, J.M. (2004). Modern science versus the stigma of obesity. Nat. Med. 10, 563–569. Galarraga, M., Campioń, J., Munõz-Barrutia, A., Boque´, N., Moreno, H., Martıńez, J.A., Milagro, F., and Ortiz-de-Solo´rzano, C. (2012). Adiposoft: Automated software for the analysis of white adipose tissue cellularity in histological sections. J. Lipid Res. 53, 2791–2796.

Gao, Y., Ottaway, N., Schriever, S.C., Legutko, B., Garcıá-Caćeres, C., de la Fuente, E., Mergen, C., Bour, S., Thaler, J.P., Seeley, R.J., et al. (2014). Hormones and diet, but not body weight, control hypothalamic microglial activity. Glia 62, 17–25. Garcia, R.A., Yan, M., Search, D., Zhang, R., Carson, N.L., Ryan, C.S., SmithMonroy, C., Zheng, J., Chen, J., Kong, Y., et al. (2014). P2Y6 receptor potentiates pro-inflammatory responses in macrophages and exhibits differential roles in atherosclerotic lesion development. PLoS ONE 9, e111385. Geiss, L.S., Wang, J., Cheng, Y.J., Thompson, T.J., Barker, L., Li, Y., Albright, A.L., and Gregg, E.W. (2014). Prevalence and incidence trends for diagnosed diabetes among adults aged 20 to 79 years, United States, 1980–2012. JAMA 312, 1218–1226. Gropp, E., Shanabrough, M., Borok, E., Xu, A.W., Janoschek, R., Buch, T., Plum, L., Balthasar, N., Hampel, B., Waisman, A., et al. (2005). Agouti-related peptide-expressing neurons are mandatory for feeding. Nat. Neurosci. 8, 1289–1291. Guns, P.J., Hendrickx, J., Van Assche, T., Fransen, P., and Bult, H. (2010). P2Y receptors and atherosclerosis in apolipoprotein E-deficient mice. Br. J. Pharmacol. 159, 326–336. Hamada, T., Mizuta, E., Yanagihara, K., Kaetsu, Y., Sugihara, S., Sonoyama, K., Yamamoto, Y., Kato, M., Igawa, O., Shigemasa, C., et al. (2007). Plasma levels of uridine correlate with blood pressure and indicators of myogenic purine degradation and insulin resistance in hypertensive patients. Circ. J. 71, 354–356. Horn, R., and Marty, A. (1988). Muscarinic activation of ionic currents measured by a new whole-cell recording method. J Gen Physiol. 92, 145–159. Kloppenburg, P., Zipfel, W.R., Webb, W.W., and Harris-Warrick, R.M. (2007). Heterogeneous effects of dopamine on highly localized, voltage-induced Ca2+ accumulation in identified motoneurons. J. Neurophysiol. 98, 2910– 2917. Koizumi, S., Shigemoto-Mogami, Y., Nasu-Tada, K., Shinozaki, Y., Ohsawa, K., Tsuda, M., Joshi, B.V., Jacobson, K.A., Kohsaka, S., and Inoue, K. (2007). UDP acting at P2Y6 receptors is a mediator of microglial phagocytosis. Nature 446, 1091–1095. €ning, J.C. (2012). Selective insulin and leptin resistance Ko¨nner, A.C., and Bru in metabolic disorders. Cell Metab. 16, 144–152. Ko¨nner, A.C., Janoschek, R., Plum, L., Jordan, S.D., Rother, E., Ma, X., Xu, C., Enriori, P., Hampel, B., Barsh, G.S., et al. (2007). Insulin action in AgRP-expressing neurons is required for suppression of hepatic glucose production. Cell Metab. 5, 438–449. Krashes, M.J., Koda, S., Ye, C., Rogan, S.C., Adams, A.C., Cusher, D.S., Maratos-Flier, E., Roth, B.L., and Lowell, B.B. (2011). Rapid, reversible activation of AgRP neurons drives feeding behavior in mice. J. Clin. Invest. 121, 1424– 1428. Lindau, M., and Fernandez, J.M. (1986). IgE-mediated degranulation of mast cells does not require opening of ion channels. Nature 319, 150–153. Luquet, S., Perez, F.A., Hnasko, T.S., and Palmiter, R.D. (2005). NPY/AgRP neurons are essential for feeding in adult mice but can be ablated in neonates. Science 310, 683–685. Mauer, J., Chaurasia, B., Goldau, J., Vogt, M.C., Ruud, J., Nguyen, K.D., Theurich, S., Hausen, A.C., Schmitz, J., Bro¨nneke, H.S., et al. (2014). Signaling by

IL-6 promotes alternative activation of macrophages to limit endotoxemia and obesity-associated resistance to insulin. Nat. Immunol. 15, 423–430. €her, M., Kolanus, Mauer, J., Chaurasia, B., Plum, L., Quast, T., Hampel, B., Blu €ning, J.C. (2010). Myeloid cell-restricted insulin W., Kahn, C.R., and Bru receptor deficiency protects against obesity-induced inflammation and systemic insulin resistance. PLoS Genet. 6. http://dx.doi.org/10.1371/journal. pgen.1000938. Rae, J., Cooper, K., Gates, P., and Watsky, M. (1991). Low access resistance perforated patch recordings using amphotericin B. J Neurosci Methods 37, 15–26. Riegel, A.K., Faigle, M., Zug, S., Rosenberger, P., Robaye, B., Boeynaems, J.M., Idzko, M., and Eltzschig, H.K. (2011). Selective induction of endothelial P2Y6 nucleotide receptor promotes vascular inflammation. Blood 117, 2548–2555. Sohn, J.W., Elmquist, J.K., and Williams, K.W. (2013). Neuronal circuits that regulate feeding behavior and metabolism. Trends Neurosci. 36, 504–512. Stachon, P., Peikert, A., Michel, N.A., Hergeth, S., Marchini, T., Wolf, D., Dufner, B., Hoppe, N., Ayata, C.K., Grimm, M., et al. (2014). P2Y6 deficiency limits vascular inflammation and atherosclerosis in mice. Arterioscler. Thromb. Vasc. Biol. 34, 2237–2245. Steculorum, S.M., Collden, G., Coupe, B., Croizier, S., Lockie, S., Andrews, Z.B., Jarosch, F., Klussmann, S., and Bouret, S.G. (2015a). Neonatal ghrelin programs development of hypothalamic feeding circuits. J. Clin. Invest. 125, 846–858. Steculorum, S.M., Paeger, L., Bremser, S., Evers, N., Hinze, Y., Idzko, M., €ning, J.C. (2015b). Hypothalamic UDP increases in Kloppenburg, P., and Bru obesity and promotes feeding via P2Y6-dependent activation of AgRP neurons. Cell 162, 1404–1417. Steculorum, S.M., Ruud, J., Karakasilioti, I., Backes, H., Engstro¨m Ruud, L., Timper, K., Hess, M.E., Tsaousidou, E., Mauer, J., Vogt, M.C., et al. (2016). AgRP neurons control systemic insulin sensitivity via myostatin expression in brown adipose tissue. Cell 165, 125–138. Sternson, S.M., and Atasoy, D. (2014). Agouti-related protein neuron circuits that regulate appetite. Neuroendocrinology 100, 95–102. Tong, Q., Ye, C.P., Jones, J.E., Elmquist, J.K., and Lowell, B.B. (2008). Synaptic release of GABA by AgRP neurons is required for normal regulation of energy balance. Nat. Neurosci. 11, 998–1000. Urasaki, Y., Pizzorno, G., and Le, T.T. (2016). Chronic uridine administration induces fatty liver and pre-diabetic conditions in mice. PLoS ONE 11, e0146994. Varela, L., and Horvath, T.L. (2012). Leptin and insulin pathways in POMC and AgRP neurons that modulate energy balance and glucose homeostasis. EMBO Rep. 13, 1079–1086. €ning, J.C. (2013). CNS insulin signaling in the control of enVogt, M.C., and Bru ergy homeostasis and glucose metabolism—from embryo to old age. Trends Endocrinol. Metab. 24, 76–84. WHO (2006). Obesity and Overweight (Fact sheet 311), World Health Organization (http://www.who.int/mediacentre/factsheets/fs311/en/index.html). Ye, J.H., Zhang, J., Xiao, C., and Kong, J.Q. (2006). Patch-clamp studies in the CNS illustrate a simple new method for obtaining viable neurons in rat brain slices: glycerol replacement of NaCl protects CNS neurons. J Neurosci Methods 158, 251–259.
