Advertisement

Global deletion of NTPDase3 protects against diet-induced obesity by increasing basal energy metabolism

  • Author Footnotes
    1 Equal contributions.
    Bynvant Sandhu
    Footnotes
    1 Equal contributions.
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Author Footnotes
    1 Equal contributions.
    Maria C. Perez-Matos
    Footnotes
    1 Equal contributions.
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Stephanie Tran
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Garima Singhal
    Affiliations
    Division of Endocrinology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Ismail Syed
    Affiliations
    Division of Endocrinology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Linda Feldbrügge
    Affiliations
    Department of Surgery, Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, Germany
    Search for articles by this author
  • Shuji Mitsuhashi
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Julie Pelletier
    Affiliations
    Centre de recherche du CHU de Québec – Université Laval, Québec City, QC G1V 4G2, Canada
    Search for articles by this author
  • Jinhe Huang
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Yusuf Yalcin
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Eva Csizmadia
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Shilpa Tiwari-Heckler
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Keiichi Enjyoji
    Affiliations
    Division of Hemostasis and Thrombosis, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Jean Sévigny
    Affiliations
    Centre de recherche du CHU de Québec – Université Laval, Québec City, QC G1V 4G2, Canada

    Département de microbiologie-infectiologie et d'immunologie, Faculté de Médecine, Université Laval, Québec City, QC G1V 0A6, Canada
    Search for articles by this author
  • Eleftheria Maratos-Flier
    Affiliations
    Division of Endocrinology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Simon C. Robson
    Correspondence
    Correspondence to: S. C. Robson, Division of Gastroenterology and Hepatology, Departments of Anesthesia and Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Office E/CLS 612, 3 Blackfan Circle, Boston, MA 02215, USA.
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA

    Department of Anesthesiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Z. Gordon Jiang
    Correspondence
    Correspondence to: Z. G. Jiang: Division of Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Liver Center, 110 Francis St. Suite 8E, Boston, MA 02115, USA.
    Affiliations
    Division of Gastroenterology & Hepatology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Author Footnotes
    1 Equal contributions.
Published:February 22, 2021DOI:https://doi.org/10.1016/j.metabol.2021.154731

      Abstract

      Background

      Ecto-nucleoside triphosphate diphosphohydrolase 3 (NTPDase3), also known as CD39L3, is the dominant ectonucleotidase expressed by beta cells in the islet of Langerhans and on nerves. NTPDase3 catalyzes the conversion of extracellular ATP and ADP to AMP and modulates purinergic signaling. Previous studies have shown that NTPDase3 decreases insulin release from beta-cells in vitro. This study aims to determine the impact of NTPDase3 in diet-induced obesity (DIO) and metabolism in vivo.

      Methods

      We developed global NTPDase3 deficient (Entpd3−/−) and islet beta-cell-specific NTPDase-3 deficient mice (Entpd3flox/flox,InsCre) using Ins1-Cre targeted gene editing to compare metabolic phenotypes with wildtype (WT) mice on a high-fat diet (HFD).

      Results

      Entpd3−/− mice exhibited similar growth rates compared to WT on chow diet. When fed HFD, Entpd3−/− mice demonstrated significant resistance to DIO. Entpd3−/− mice consumed more calories daily and exhibited less fecal calorie loss. Although Entpd3−/− mice had no increases in locomotor activity, the mice exhibited a significant increase in basal metabolic rate when on the HFD. This beneficial phenotype was associated with improved glucose tolerance, but not higher insulin secretion. In fact, Entpd3flox/flox,InsCre mice demonstrated similar metabolic phenotypes and insulin secretion compared to matched controls, suggesting that the expression of NTPDase3 in beta-cells was not the primary protective factor. Instead, we observed a higher expression of uncoupling protein 1 (UCP-1) in brown adipose tissue and an augmented browning in inguinal white adipose tissue with upregulation of UCP-1 and related genes involved in thermogenesis in Entpd3−/− mice.

      Conclusions

      Global NTPDase3 deletion in mice is associated with resistance to DIO and obesity-associated glucose intolerance. This outcome is not driven by the expression of NTPDase3 in pancreatic beta-cells, but rather likely mediated through metabolic changes in adipocytes.

      Abbreviations:

      AUC (area under the curve), CLAMS (comprehensive lab animal monitoring system), Cpt1a (carnitine palmitoyl transferase 1A), Dio2 (iodothyronine deiodinase 2), FFPE (formalin-fixed paraffin-embedded), GSIS (glucose-stimulated insulin secretion), HFD (high-fat diet), ITT (insulin tolerance test), OGTT (oral glucose tolerance test), NTPDase 3 (ecto-nucleoside triphosphate diphosphohydrolase 3), RER (respiratory exchange ratio), SEM (standard error of the mean), UCP-1 (uncoupling protein-1), WAT (white adipose tissue), WT (wildtype)

      Keywords

      1. Introduction

      Extracellular ATP modulates diverse cellular responses via purinergic receptors and plays an important role in host homeostasis [
      • Eltzschig H.K.
      • Sitkovsky M.V.
      • Robson S.C.
      Purinergic signaling during inflammation.
      ]. Nucleoside triphosphate diphosphohydrolases (NTPDases) are a class of plasma-membrane bound ecto-enzymes responsible for the hydrolysis of extracellular ATP to ADP and AMP [
      • Robson S.C.
      • Sévigny J.
      • Zimmermann H.
      The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance.
      ]. Amongst this family of enzymes, NTPDase1, 2, 3 and 8 are expressed on the plasma membrane, hence are most important in modulating signaling via extracellular purines and metabolites [
      • Zimmermann H.
      • Zebisch M.
      • Strater N.
      Cellular function and molecular structure of ecto-nucleotidases.
      ]. As examples: NTPDase1 (CD39), the prototype of the family, is expressed by endothelium and immune cells;[
      • Robson S.C.
      • Sévigny J.
      • Zimmermann H.
      The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance.
      ,
      • Antonioli L.
      • Pacher P.
      • Vizi E.S.
      • Hasko G.
      CD39 and CD73 in immunity and inflammation.
      ] NTPDase2 (CD39L1) expression is found in mesenchymal cells such as pericytes, myofibroblast, and glial cells [
      • Braun N.
      • Sévigny J.
      • Robson S.C.
      • Hammer K.
      • Hanani M.
      • Zimmermann H.
      Association of the ecto-ATPase NTPDase2 with glial cells of the peripheral nervous system.
      ,
      • Feldbrugge L.
      • Moss A.C.
      • Yee E.U.
      • Csizmadia E.
      • Mitsuhashi S.
      • Longhi M.S.
      • et al.
      Expression of Ecto-nucleoside triphosphate Diphosphohydrolases-2 and -3 in the enteric nervous system affects inflammation in experimental colitis and Crohn’s disease.
      ,
      • Dranoff J.A.
      • Kruglov E.A.
      • Robson S.C.
      • Braun N.
      • Zimmermann H.
      • Sévigny J.
      The ecto-nucleoside triphosphate diphosphohydrolase NTPDase2/CD39L1 is expressed in a novel functional compartment within the liver.
      ]; and NTPDase8 expression is noted in the canaliculi of the liver [
      • Fausther M.
      • Lecka J.
      • Kukulski F.
      • Levesque S.A.
      • Pelletier J.
      • Zimmermann H.
      • et al.
      Cloning, purification, and identification of the liver canalicular ecto-ATPase as NTPDase8.
      ]. The expression of NTPDase3 (CD39L3) has been noted in peripheral nerves, the central nervous system, the epithelium of digestive tract, as well as the islet of Langerhans. [
      • Lavoie E.G.
      • Gulbransen B.D.
      • Martin-Satue M.
      • Aliagas E.
      • Sharkey K.A.
      • Sévigny J.
      Ectonucleotidases in the digestive system: focus on NTPDase3 localization.
      ,
      • Lavoie E.G.
      • Fausther M.
      • Kauffenstein G.
      • Kukulski F.
      • Kunzli B.M.
      • Friess H.
      • et al.
      Identification of the ectonucleotidases expressed in mouse, rat, and human Langerhans islets: potential role of NTPDase3 in insulin secretion.
      ] Notably, NTPDase3 has been proposed as a marker for beta-cells in human due to its robust expression in the pancreatic islets [
      • Saunders D.C.
      • Brissova M.
      • Phillips N.
      • Shrestha S.
      • Walker J.T.
      • Aramandla R.
      • et al.
      Ectonucleoside triphosphate Diphosphohydrolase-3 antibody targets adult human pancreatic beta cells for in vitro and in vivo analysis.
      ]. However, the biological function of NTPDase3 is not well characterized at any of these sites.
      Ectonucleotidases impact purinergic signaling, which in turn modulates various physiological pathways including autonomic regulation, immunity, thrombosis, tissue repair, as well as metabolism. [
      • Eltzschig H.K.
      • Sitkovsky M.V.
      • Robson S.C.
      Purinergic signaling during inflammation.
      ,
      • Zimmermann H.
      • Zebisch M.
      • Strater N.
      Cellular function and molecular structure of ecto-nucleotidases.
      ,
      • Deaglio S.
      • Robson S.C.
      Ectonucleotidases as regulators of purinergic signaling in thrombosis, inflammation, and immunity.
      ] The deficiency of NTPDase1 has been linked to hepatic insulin resistance [
      • Enjyoji K.
      • Kotani K.
      • Thukral C.
      • Blumel B.
      • Sun X.
      • Wu Y.
      • et al.
      Deletion of cd39/entpd1 results in hepatic insulin resistance.
      ]. In the islet of Langerhans, NTPDase3 is the most abundant ectonucleotidase expressed in islet endocrine cells, while NTPDase1 is found on the endothelium and NTPDase2 is expressed in the islet capsule [
      • Lavoie E.G.
      • Fausther M.
      • Kauffenstein G.
      • Kukulski F.
      • Kunzli B.M.
      • Friess H.
      • et al.
      Identification of the ectonucleotidases expressed in mouse, rat, and human Langerhans islets: potential role of NTPDase3 in insulin secretion.
      ]. The expression pattern of NTPDase3 raised the possibility it might be involved in the regulation of glucose and energy metabolism. Importantly, ATP is co-released when insulin is secreted from beta-cells and serves as a potent amplifier of insulin release in beta-cells via P2Y and P2X nucleotide receptors.[
      • Petit P.
      • Hillaire-Buys D.
      • Manteghetti M.
      • Debrus S.
      • Chapal J.
      • Loubatieres-Mariani M.M.
      Evidence for two different types of P2 receptors stimulating insulin secretion from pancreatic B cell.
      ,
      • Loubatieres-Mariani M.M.
      • Chapal J.
      Purinergic receptors involved in the stimulation of insulin and glucagon secretion.
      ] The activation of P2Y receptors activates the inositol triphosphate pathway, while the activation of inotropic P2X receptors results in the influx of Ca2+ and the opening of voltage-dependent calcium channels. Both mechanisms can trigger and/or modulate the release of insulin. NTPDase3 converts extracellular ATP to ADP and ADP to AMP, thus may effectively abrogate this autocrine loop. It has been shown that pharmacological inhibition of NTPDase3 by ARL67156 effectively augments insulin secretion from MIN6 cells, an insulinoma cell line, in vitro [
      • Syed S.K.
      • Kauffman A.L.
      • Beavers L.S.
      • Alston J.T.
      • Farb T.B.
      • Ficorilli J.
      • et al.
      Ectonucleotidase NTPDase3 is abundant in pancreatic beta-cells and regulates glucose-induced insulin secretion.
      ].
      These observations indicate the potential pharmacological value of targeting NTPDase3 in the treatment of hyperglycemia associated with obesity and type II diabetes mellitus. Herein, we generated global and then targeted pancreatic beta-cell specific knockout mice to test the hypothesis that global deletion or targeted genetic inhibition of NTPDase3 could attenuate or preclude hyperglycemia associated with diet-induced obesity (DIO) in mice.

      2. Methods

      2.1 Animals and nutritional interventions

      2.1.1 Generation of transgenic animal models

      The generation of global NTPDase3 knockout (Entpd3−/−) mice was accomplished by deleting exon 3 of the Entpd3 locus on chromosome 9, as previously described (Fig. 1A ) [
      • Feldbrugge L.
      • Moss A.C.
      • Yee E.U.
      • Csizmadia E.
      • Mitsuhashi S.
      • Longhi M.S.
      • et al.
      Expression of Ecto-nucleoside triphosphate Diphosphohydrolases-2 and -3 in the enteric nervous system affects inflammation in experimental colitis and Crohn’s disease.
      ]. Similarly, NTPDase3 floxed (Entpd3flox/flox) mice were generated by inserting loxP sequences flanking exon 3 (Fig. 4B). Knockout and floxed transgenic mice were backcrossed to C57BL6 mice for five generations. The Entpd3flox/flox mice were bred with Ins1-Cre mice (Jackson Laboratory, Bar Harbor, ME) to produce beta-cell specific knockouts (Entpd3flox/flox,InsCre+). All mice used in experiments were littermates bred from either Entpd3+/− mice or male Entpd3flox/flox,InsCre+ and female Entpd3flox/flox,InsCre- mice. Genotyping was confirmed using PCR at Transnetyx (Transnetyx, Cordova, TN). All procedures were conducted in accordance with National Institute of Health Guidelines for the Care and Use of Animals and were approved by the Beth Israel Deaconess Medical Center Institutional Animal Care and Use Committee.
      Fig. 1
      Fig. 1Generation of global NTPDase3 deficient mice
      A. Scheme for the generation of global NTPDase3 deficient mouse (Entpd3−/−). B. Representative immunohistochemistry of NTPDase3 expression in various tissues from WT and Entpd3−/− mice. Similar results are observed in both male and female mice. Arrows indicate location of expected NTPDase3 expression in WT. Scale bar represents 50 μm.

      2.1.2 Murine models of DIO

      Mice were housed under standard conditions at 21 °C with 12-h alternating light and dark cycles and ad libitum access to food and water. In all DIO experiments, 10–12 mice per gender and group were initially fed with Teklad standard chow rodent diet until 6–8 weeks of age (Envigo, Franklin, NJ). Experimental groups were either continued on the chow diet or on a high-fat diet (HFD, #TD93075 from Envigo) for a total of 20 weeks.
      At the conclusion of the study, mice were euthanized using Ketamine/Xylazine and tissues (fat, liver) were harvested for histology and direct measurement of body composition. The relative weights of brown adipose tissue (BAT), inguinal white adipose tissue (WAT), and liver tissue to whole body weight were measured following 20 weeks on standard chow or HFD. The right tibia was collected for assessment of skeletal size as a measurement of growth rate. The soft tissue was cleared and tibial length (proximal articulating surface to medial malleolus) measured using a caliper. All animal experiments were conducted in duplicates.

      2.2 Metabolic analysis

      2.2.1 Body composition analysis

      Body composition was determined using an EchoMRI 3-in-1 quantitative nuclear magnetic resonance (qNMR) system (Echo Medical Systems, Houston, TX) at baseline and 20 weeks. Body fat, lean mass, and total body water were measured in live conscious mice with ad libitum access to chow or HFD.

      2.2.2 Measurement of calorie intake

      Mice were switched to AlphaPad (LBS Biotech, United Kingdom) bedding for measurement of food intake and fecal collection. Food intake measurement and mouse body weight were recorded weekly for three weeks, following a one-week acclimatization period. Feed efficiency was calculated as body weight gain divided by total caloric intake in the defined assessment period.

      2.2.3 Oxygen bomb calorimetry

      Energy assimilation was assessed using a Parr 6725EA Semimicro Calorimeter with 1107/22 ml Oxygen Bomb (Preiser Scientific, Louisville, KY) following the manufacturer's guidance to determine the energy content per gram of feces. Fecal lipid excretion was analyzed following lipid extraction using the chloroform:methanol (2:1) method, as per published protocol [
      • Kraus D.
      • Yang Q.
      • Kahn B.B.
      Lipid extraction from mouse feces.
      ].

      2.2.4 Indirect calorimetry

      Mice were maintained on a 12:12-h light-dark cycle and metabolic rate was measured by indirect calorimetry using a Comprehensive Lab Animal Monitoring System (CLAMS) (Columbus Instruments, Columbus, OH). Sample air was passed through an O2 sensor (Columbus Instruments) for determination of O2 content. O2 consumption was calculated by examining the difference of O2 concentration of air entering the chamber compared with air leaving the chamber and heat production on per-animal basis was calculated from the following equation: (3.82 + 1.23 × RER) × VO2, where the respiratory exchange ratio (RER) is the volume of CO2 produced/volume of O2 consumed per hour. The sensor was calibrated against a standard gas mix containing defined quantities of O2, CO2, and nitrogen. Food and water were available ad libitum. The measurement of 24 h of data collection was averaged and binned to create day and night depictions of metabolic rate and normalized to effective mass. The analysis of CLAMS data was conducted in CalR [
      • Mina A.I.
      • LeClair R.A.
      • LeClair K.B.
      • Cohen D.E.
      • Lantier L.
      • Banks A.S.
      CalR: a web-based analysis tool for indirect Calorimetry experiments.
      ].

      2.2.5 Home cage activity

      Spontaneous locomotor activity was recorded continuously using the OptoM3 apparatus beam breaks (Columbus Instruments).

      2.2.6 Measurement of glucose and insulin metabolism

      Baseline metabolic studies were performed on mice aged 6 to 8 weeks before initiation of HFD. Subsequent studies were performed at 12 and 20 weeks. Oral glucose tolerance tests (OGTT), insulin tolerance tests (ITT) and glucose-stimulated insulin secretion (GSIS) studies were performed following a 12-h overnight fast. For OGTT, Glucose (Gibco, Waltham, MA) was delivered at 1 g/kg orally through a gavage syringe. Blood glucose levels were measured from the tail vein of unrestrained mice at 0, 15, 30, 60, and 120 min with an OneTouch Ultra 2 glucometer. ITT was performed by intraperitoneal administration of insulin at 1 U/kg and measurement of glucose levels at 0, 15, 30, 60 and 120 min. For GSIS studies, whole blood was collected in EDTA tubes (Starstedt, Nümbrecht, Germany) at 0, 15, and 30 min and then centrifuged to remove cells and platelets. Plasma insulin concentration was measured with the Ultra Sensitive Mouse Insulin Elisa kit (Crystal Chem, Elk Grove Village, IL). The area under the curve (AUC) was calculated by the trapezoidal method.

      2.3 Biochemical analysis

      2.3.1 Histology and immunohistochemistry

      All tissues for histology were in either fixed in formalin or flash-frozen in 2-methylbutane for histology or immunohistochemistry. Formalin-fixed paraffin-embedded (FFPE) or frozen tissue were prepared into 5 μm sections, and stained with the following primary antibodies: NTPDase1, −2, −3, insulin, glucagon, somatostatin, CD45, synaptophysin, and UCP-1 (Suppl. Table 1). Only FFPE-processed WAT was used. Tissue samples were typically stained with the primary antibody overnight at 4 °C. Avidin and biotin blocking were each performed for 15 min at room temperature. The corresponding secondary antibodies were then applied for 1 h at room temperature. The color was developed with the Vectastain Elite ABC HRP Kit (Vector Laboratories, Burlingame, CA) and the ImmPACT DAB Substrate Kit (Vector Laboratories). Slides were then counterstained with hematoxylin for 30 s.

      2.3.2 Adipocyte size measurement

      FFPE inguinal WAT obtained from female mice at 20 weeks on HFD was stained with hematoxylin & eosin (H&E) for the determination of adipocyte size. Analyses were performed on ten random high power fields per section (sections per mouse, n = 8–12 mice per group), with the Adiposoft Plugin on ImageJ (National Institute of Health) [
      • Galarraga M.
      • Campion J.
      • Munoz-Barrutia A.
      • Boque N.
      • Moreno H.
      • Martinez J.A.
      • et al.
      Adiposoft: automated software for the analysis of white adipose tissue cellularity in histological sections.
      ].

      2.3.3 Quantitative RT-PCR

      RNA was isolated from tissue flash-frozen in liquid nitrogen using an RNAeasy mini kit (QIAGEN) according to the manufacturer's instructions. A deoxyribonuclease (QIAGEN) step to digest the genomic DNA was included. cDNA was made from isolated RNA using oligo(dt) and random hexamer primers and reverse transcriptase (QuantiTech RT kit; QIAGEN). Quantitative PCR was performed using the 7800HT (Applied Biosystems) thermal cycler and SYBR Green master mix (Applied Biosystems). Relative mRNA abundance was calculated and normalized to levels of TATA box-binding protein. Primers are available in Suppl. Table 2.

      2.3.4 Immunoblotting

      In brief, BAT tissues were homogenized in lysis buffer containing 50 mM Tris/HCl, 500 mM NaCl, 1% NP40, 20% Glycerol, 1 mM DTT, and 10 mM nicotinamide at pH 7.4. Protein concentrations were determined with a Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Equivalent amounts of 10 μg protein were loaded onto a 4–12% Criterion XT Bis-Tris Protein Gel (Bio-Rad) for electrophoresis and transferred to an Immobilon-FL PVDF membrane (Millipore, Temecula, CA). UCP1 was detected with rabbit anti-UCP1 polyclonal antibody (Abcam, Cambridge, MA) and IRDye 800CW goat anti-rabbit IgG (Li-Cor, Lincoln, NE). β-actin was detected with mouse anti-β-actin monoclonal antibody (Biolegend, Dedham, MA) and IRDye 680RD goat anti-mouse IgG (Li-Cor). Relative protein concentration was determined by the fluorescent signal digitally processed under ImageStudio (Li-Cor).

      2.4 Statistical analysis

      GraphPad Prism (La Jolla, CA) was utilized to analyze statistical differences and generate plots. Student's t-test was used to compare two groups and the analysis of variance (ANOVA) test was used to compare three or more groups. Data are displayed as the mean ± standard error of the mean (SEM).

      3. Results

      3.1 NTPDase3 deletion protects from diet-induced obesity

      We generated NTPDase3 knockout mice by the deletion of exon 3 of Entpd3, which encodes the transmembrane domain of the protein (Fig. 1A). Global NTPDase3 deficient (Entpd3−/−) mice demonstrated normal embryonic development and breeding. Breeding of heterozygous Entpd3+/− mice produced off-spring with an expected Mendelian ratio. No differences in sex ratio or survival rate were noted, suggesting that the deletion of NTPDase3 does not impact fertility or reproductive fitness. Immunohistochemistry comparing wildtype (WT) and Entpd3−/− confirmed the expression of NTPDase3 in the salivary gland, epithelial cells of the esophagus, enteric nerves of the digestive tract, and the islet of Langerhans (Fig. 1B). The expression of NTPdase3 in the brain was diffuse and varied across various anatomical regions (Suppl. Fig. 1).
      Female Entpd3−/− mice exhibited normal feeding and behavior compared to WT mice. On a chow diet, Entpd3−/− mice displayed a similar growth curve to WT mice (Fig. 2A ). At eight weeks of age, female Entpd3−/− mice had no change in growth, as measured by tibial length (Fig. 2B). However, when fed the HFD, female Entpd3−/− mice gained less weight than WT mice. Growth curves for female WT and Entpd3−/− mice started to separate from week 5 on HFD (Fig. 2A, upper). The growth curve for the female Entpd3−/− mice on HFD approximated their counterparts on standard chow. The percentage in body weight gain at 20 weeks was 41% less in female Entpd3−/− mice compared to WT mice (Fig. 2A, lower). This difference was not due to altered growth rates, as both groups had similar tibial length after 20 weeks of HFD (Fig. 2B). Rather, female Entpd3−/− mice exhibited a lean phenotype on HFD with a lower whole-body fat mass ratio and a higher whole-body lean mass ratio compared to WT mice (Fig. 2C). The inguinal WAT depots were significantly decreased in Entpd3−/− mice compared to WT mice, but interscapular BAT mass showed no difference (Fig. 2D). Liver weight increased in both strains under HFD compared to chow, but the liver-to-body weight ratio remained unchanged with HFD (Fig. 2D).
      Fig. 2
      Fig. 2Comparisons of weight and body composition between WT and NTPDase3 deficient mice
      A. Growth curves (above) and weights of female Entpd3−/− (KO) and WT mice at baseline and 20 weeks (below) on standard chow and HFD (n = 14 for KO and WT on HFD, and n = 11 at 20 weeks). B. Comparison of female tibial bone length. C. Body composition of female WT and KO mice at 20 weeks measured by NMR. D. Weights for the female adipose tissue and liver. E. A representative picture of female inguinal WAT from Entpd3−/− and WT mice at 20 weeks on HFD. Female KO and WT mice were littermates and cohoused throughout the experiment. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01 using two-tailed Student's t-test.
      A similar trend of reduced weight gain on HFD was noted in male Entpd3−/− mice, although the difference was smaller compared to that of the females (Suppl. Fig. 2). This change was associated with a higher number of injuries and aggressive behavior in male mice, when compared to the females. This prompted us to focus our investigation on the female mice to understand the mechanism behind this weight difference on HFD.

      3.2 NTPDase3 global deficiency protects from DIO and glucose intolerance

      As NTPDase3 is highly expressed in beta cells of the pancreatic islet of Langerhans, we evaluated whether the weight difference of Entpd3−/− mice on HFD was associated with changes in glucose homeostasis. At baseline before diet induction, female Entpd3−/− and WT mice exhibited no differences in fasting glucose levels or glucose tolerance (Fig. 3A ). However, at 12 weeks, female Entpd3−/− mice on the HFD exhibited lower blood glucose levels at 15 min after an oral glucose challenge, when compared to WT counterparts (Fig. 3A). There was no evidence that Entpd3−/− mice produced more insulin than the WT mice as measured by the glucose-stimulated insulin secretion test, rather there were increasing trends toward lower rate of insulin secretion in Entpd3−/− mice at 12 and 20 weeks (Fig. 3B). After 20 weeks on HFD, Entpd3−/− mice had both lower fasting glucose and smaller AUC compared to WT counterparts (Fig. 3A, B). Furthermore, we compared insulin tolerance between Entpd3−/− and WT mice, and found that Entpd3−/− had the same insulin sensitivity at baseline (Fig. 3C). At 20 weeks, the glucose curve of Entpd3−/− was lower than that of the WT, but the slope remained unchanged indicating similar insulin sensitivity.
      Fig. 3
      Fig. 3Impact of NTPDase3 on glucose and insulin homeostasis
      A. Oral glucose tolerance test (OGTT). Glucose tracing over time (upper row) and area under the curve (AUC) in bar representation (lower row) were shown. B. Oral glucose stimulated insulin secretion test (GSIS). Insulin tracing over time (upper) and AUC (lower) were shown. C. Insulin tolerance test (ITT). Experiments conducted at baseline (left), 12 weeks (middle) and 20 weeks (right). Unadjusted glucose measurements in mg/dL shown in A and C. Data are presented as mean ± SEM. n = 14 for female KO and WT at baseline, n = 11 for female KO and WT on HFD at 12 and 20 weeks. All female KO and WT mice were littermates and cohoused throughout the experiment. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 using two-tailed Student's t-test.

      3.3 Beta-cell specific NTPDase3 deletion has no impact on diet-induced obesity or glucose intolerance

      To test whether the favorable metabolic phenotype in global NTPDase3 deficient mice was mechanistically mediated by beta cells in pancreatic islets of Langerhans, we generated beta-cell specific knockout of NTPDase3 by breeding the Entpd3flox/flox with Ins-Cre mice (Fig. 4A ). NTPDase3 was absent in the beta-cells of the Entpd3flox/flox,InsCre+ mice, with some residual protein expression in the periphery presumably in alpha- or delta- cells (Fig. 4B, C). As expected, Entpd3flox/flox,InsCre+ mice had normal expression of NTPDase3 in peripheral nerves compared to WT mice (Figs. 1B, 4B).
      Fig. 4
      Fig. 4Generation of beta-cell specific NTPDase3 deficient mouse
      A. A scheme for the generation of beta-cell specific NTPDase3 KO mice (Entpd3flox/flox,InsCre+). B. Immunohistochemistry of NTPDase3 in various tissues in Entpd3flox/flox,InsCre+ and control (Entpd3flox/flox,InsCre neg). Arrows indicate location of expected NTPDase3 expression in WT. C. Immunohistochemistry for the pancreatic expression of NTPDase3, 2, 1, glucagon, insulin, somatostatin, CD45 and synaptophysin. Scale bar represents 50 μm.
      When compared to Entpd3flox/flox Cre-negative controls, Entpd3flox/flox, InsCre+ mice did not demonstrate the observed resistance to diet-induced obesity as seen in global NTPDase3 deficient mice. The weight curves remained the same throughout the 20 weeks of HFD (Fig. 5A ). In both Cre control and Entpd3flox/flox,InsCre+, HFD induced significant changes in glucose tolerance as measured by OGTT at 20 weeks (Fig. 5B). There were no significant differences in oral glucose tolerance, oral glucose stimulated insulin secretion, or insulin tolerance between Cre control and Entpd3flox/flox,InsCre+ mice at either individual time points or area under the time-course curve.(Fig. 5B, C, Suppl. Fig. 3).
      Fig. 5
      Fig. 5Impact of NTPDase3 from beta-cells on glucose and insulin homeostasis
      A. Growth curves of both female and male Entpd3flox/flox,InsCre+ (Cond KO) and Entpd3flox/flox,InsCre+ (Cre Ctrl) mice. B. OGTT curve over time. C. GSIS curve over time. For both B, and C, n = 25 for Cond KO and Cre Ctrl at baseline, n = 15 for Cond KO and Cre Ctrl on HFD at 12 weeks, n = 8 for Cond KO and 6 for Cre Ctrl on HFD at 20 weeks. Data for female mice presented as mean ± SEM. For comparison between HFD and chow groups in B using Student's t-test: *, p < 0.01 for Cre Ctrl, p < 0.001 for Cond KO; **, p < 0.01 for Cre Ctrl, p < 0.05 for Cond KO; ***, p < 0.05 for Cond KO. No difference is significant between Cre Ctrl and Cond KO in either OGTT or GSIS.
      We went on to confirm that this observation was not due to compensatory expression of other CD39 isoforms. Indeed, no differences in the expression of NTPDase1 or NTPDase2 were noted in the pancreatic islets between Entpd3flox/flox,InsCre+ and Entpd3flox/flox Cre-neg control mice to indicate the presence of compensatory expression of other ectonucleotidases (Fig. 4C). Similarly, there were no difference in the expression of insulin, glucagon or somatostatin to indicate major changes in the distribution of pancreatic islet cells, nor was there evidence of insulitis, as the number of CD45 positive cells did not alter in the islets of Entpd3flox/flox,InsCre+ mice. Taken together, these findings led us to conclude that the favorable metabolic phenotype seen in global NTPDase3 deficient mice was not mediated by NTPDase3 expressed in pancreatic beta cells.

      3.4 NTPDase3 regulates basal energy expenditure

      Body weight is determined by the balance between calorie intake and energy expenditure. To determine the mechanism underlying the resistance in DIO in global NTPDase3 deficient mice, we measured energy consumption and expenditure. Surprisingly, despite resulting in a lower body weights, Entpd3−/− mice consumed significantly more food, when the comparison was made to WT controls (Fig. 6A ). Furthermore, the residual fecal calorie in Entpd3−/− mice was also significantly less than that from WT controls indicating better absorption (Fig. 6B). The stool fat content was similar between the two groups (Fig. 6B).
      Fig. 6
      Fig. 6Impact of NTPDase3 on energy balance and metabolic rate
      A. Daily calorie intake of female mice on HFD. B. Residual fecal calorie and lipid content. C. Measurement of O2 consumption and CO2 production on comprehensive lab animal monitoring systems. Representative 24-h tracing shown above and mean values shown in bottom panels. D. Respiratory exchange ratio (RER) during light and dark cycles. E. Locomotor activity measured on laser bream breaker cages reported in counts/h. n = 8 for WT and global null mice, in all experiments. Data are presented as mean ± SEM *P < 0.05, **P < 0.01 for comparison between KO and WT on same diet using two tailed Student's t-test.
      Remarkably, Entpd3−/− mice exhibited a significant increase in basal oxygen consumption and carbon dioxide production compared to their WT counterparts at 20 weeks on HFD (Fig. 6C). This difference was significant both at night and during the day. Respiratory exchange ratios (RER) were similar between these two groups in keeping with the same diet they were consuming (Fig. 6D). The locomotor activity was not significantly different between the knockout and WT mice, although the Entpd3−/− mice trended toward less activity at night (Fig. 6E). We then evaluated mice on chow prior to HFD induction when no significant weight difference was detected between Entpd3−/− and WT mice. Although the means of VO2 and VCO2 averaged over a 12-h period were not significantly different between female WT and Entpd3−/− mice, there was a clear trend toward higher VO2 and VCO2 in Entpd3−/− compared to WT mice, suggesting a small increase in basal metabolic rate even in the absence of significant weight differences (Suppl. Fig. 4).

      3.5 NTPDase3 deficiency induces BAT activation and WAT ‘browning’

      The difference in basal metabolic rate between Entpd3−/− mice and WT controls led us to investigate the adipose tissues, the key regulator of basal metabolic rate through thermogenesis. IHC comparing Entpd3−/− and WT control mice showed that NTPDase3 is expressed in the nerve bundles of BAT (Fig. 7A ). Further examination showed that the interscapular BAT of Entpd3−/− mice displayed higher UCP1 protein expression as measured on western blot after normalization by the β-actin control (Fig. 7B, Suppl. Fig. 5). The slight variation in the β-actin levels likely reflected tissue-to-tissue heterogeneity. The mRNA levels of UCP1, as well as other genes involved in BAT metabolism were similar between Entpd3−/− and WT mice on HFD (data not shown). The inguinal WAT mass in Entpd3−/− mice was significantly less than that of the WT controls (Fig. 2D). Furthermore, inguinal WAT from Entpd3−/− mice exhibited notable difference on histology in comparison to those from WT controls. The adipocytes from Entpd3−/− mice from WAT were significantly smaller, organized in clusters of multilocular cells, suggesting browning (Fig. 7C). When stained for UCP1, higher UCP1 staining was seen in smaller adipocytes (Fig. 7D). This drove the impression that the adipose tissue from Entpd3−/− mice demonstrated higher UCP1 expression compared to WT mice. In addition, the expression of beige adipocyte marker genes, such as carnitine palmitoyl transferase 1 A (Cpt1a) and iodothyronine deiodinase 2 (Dio2), were upregulated in inguinal white adipose tissue of Entpd3−/− mice, when compared to WT mice in keeping with the process of active browning (Fig. 7E). This change was not associated with a significant upregulation of representative genes in the mitochondria electron transport chain complex in adipose tissue (Suppl. Fig. 6). Nor were there changes in the PPARγ, PRDM16, transcriptional regulator in browning, or FGF21, an autocrine/paracrine stimulator of browning in WAT (Suppl. Fig. 7).
      Fig. 7
      Fig. 7Impact of NTPDase3 on function of BAT and browning of WAT
      A. Immunohistochemistry of NTPDase3 in adipose tissue showing the positive staining of nerve fibers. B. The relative expression of UCP1 in interscapular BAT measured in female Entpd3−/− (KO) and WT mice at 20 weeks on HFD (n = 8 for WT, 9 for KO) as measured by intensity on western blot (Suppl. Fig. 5). C. H&E staining of inguinal WAT (left) showing the differences in the size of adipocytes. Adipocyte size was analyzed using ImageJ (right). D. Expression of UCP1 in female KO and WT mice at 20 weeks on HFD. E. Comparison of genes involved in browning and thermogenesis in inguinal WAT of female KO and WT mice at 20 weeks on HFD (n = 10 for WT, 11 for KO). Data are presented as mean ± SEM *P < 0.05 for KO vs WT on same diet by two-tailed Student's t-test. Scale bar represents 50 μm.

      4. Discussion

      Amongst plasma membrane-bound ectonucleotidases, NTPDase3 is the most highly expressed in beta cells of Langerhans in the pancreas [
      • Lavoie E.G.
      • Fausther M.
      • Kauffenstein G.
      • Kukulski F.
      • Kunzli B.M.
      • Friess H.
      • et al.
      Identification of the ectonucleotidases expressed in mouse, rat, and human Langerhans islets: potential role of NTPDase3 in insulin secretion.
      ]. NTDPase3 shares high level of homology to other surface NTPDases, i.e. NTPDase 1, 2, and 8 [
      • Robson S.C.
      • Sévigny J.
      • Zimmermann H.
      The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance.
      ]. NTPDase3 has both ATPase and ADPase activity, although its ADPase activity is weaker than that of NTPDase1 [
      • Robson S.C.
      • Sévigny J.
      • Zimmermann H.
      The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance.
      ]. It has been suggested that NTPDase3 might regulate glucose-induced insulin release in beta-cells by attenuating extracellular ATP mediated insulin release; as such pharmacological inhibition of NTPDase3 has been shown to increase insulin secretion in vitro [
      • Syed S.K.
      • Kauffman A.L.
      • Beavers L.S.
      • Alston J.T.
      • Farb T.B.
      • Ficorilli J.
      • et al.
      Ectonucleotidase NTPDase3 is abundant in pancreatic beta-cells and regulates glucose-induced insulin secretion.
      ]. These observations suggest that NTPDase3 could be a druggable target to augment insulin secretion. Based on this premise, we studied the impact of NTPDase3 deletion on glucose homeostasis in vivo. Our data suggest that the deletion of NTPDase3 reduces weight gain in experimental models of DIO. Somewhat surprisingly, this benefit is not mediated by the expression of NTPDase3 on pancreatic islet cells, but rather appears to be driven by its impact on the basal metabolic rate dictated by the adipose tissue.
      Global deficiency of NTPDase3 did not result in a basal phenotype. With exposure to HFD, NTPDase3 deficient mice exhibited a much slower rate of weight gain and better glucose tolerance, as compared to WT mice. However, to our surprise, improvements in glucose tolerance were not linked to increases in insulin secretion from the beta-cells in the Entpd3−/− mice. Rather, the differential effects observed are likely secondary to the relative decrease in body weight in the global NTPDase3 null mice while on HFD. Although we did not detect a difference in insulin tolerance via ITT, the ITT used a total body weight-adjusted insulin dosing, which may underestimate the insulin resistance in the more obese group of mice. We also noted that the beta-cell specific NTPDase3 deficient mice on HFD exhibited no differences in weight and glucose homeostasis when compared with the control mice on the same diet. This observation suggested that the modulation of the insulin-dependent pathway through NTPDase3 was relatively minor in vivo.
      Glucose homeostasis is regulated by a multiple neuronal and hormonal pathways. Hence, compensatory mechanisms might be responsible for the lack of expected changes in glucose-induced insulin secretion in vivo. While our data do not support the previous findings that pharmacological inhibition of NTPDase3 promotes insulin secretion, we have noted important impacts of NTPDase3 on basal energy metabolism in vivo. This is supported by the significantly higher basal metabolic rate in global NTPDase3 deficient mice, as measured using comprehensive lab animal monitoring systems (CLAMS). Furthermore, NTPDase3 deficient mice have higher daily calorie intake, lower excreted calories in the stool, and similar locomotor activity, when compared to WT mice. These observations suggested impacts of NTPDase3 on thermogenesis, the key regulator of basal metabolic rate. Indeed, there was an increase in the expression of UCP1 in BAT of NTPDase3 deficient mice, as measured by Western blot. Moreover, evidence of heightened browning was noted in WAT in NTPDase3 deficient mice, with observations of smaller adipocytes, higher expression of UCP1 as well as other genes relevant to thermogenesis.
      Regulation of thermogenesis occurs on several levels. Centrally, the hypothalamus plays an important role. The expression of NTPDase3 in the hypothalamus is lower compared to other anatomic regions of the brain, such as the paraolfactory diagonal band nucleus (NDB) (Suppl. Fig. 1B). Previous in situ hybridization studies also suggested that NTPDase3 transcript is less abundant in the hypothalamus compared to brainstem regions such as the pons and medulla in mice [
      • Lein E.S.
      • Hawrylycz M.J.
      • Ao N.
      • Ayres M.
      • Bensinger A.
      • Bernard A.
      • et al.
      Genome-wide atlas of gene expression in the adult mouse brain.
      ]. Nonetheless, a mechanism that involves the hypothalamus cannot be ruled out. Despite a decreased locomotor activity, there were no apparent muscle abnormality by histology or behavioral observations. Sympathetic nerves are the major neuronal regulators of both WAT and BAT. This innervation regulates the lipolysis of WAT and the thermogenesis from BAT under cold exposure.[
      • Morrison S.F.
      • Madden C.J.
      • Tupone D.
      Central neural regulation of brown adipose tissue thermogenesis and energy expenditure.
      ,
      • Bartness T.J.
      • Liu Y.
      • Shrestha Y.B.
      • Ryu V.
      Neural innervation of white adipose tissue and the control of lipolysis.
      ] In addition to norepinephrine, sympathetic nerves release ATP, which acts as a co-transmitter. While β-adrenergic signaling is the main driver for the thermoregulation, purinergic and adenosinergic signaling have been shown to also play complementary roles. Adipocytes can release ATP directly through pannexin-1 channels [
      • Adamson S.E.
      • Meher A.K.
      • Chiu Y.H.
      • Sandilos J.K.
      • Oberholtzer N.P.
      • Walker N.N.
      • et al.
      Pannexin 1 is required for full activation of insulin-stimulated glucose uptake in adipocytes.
      ]. It has been shown that adenosine, the downstream ectonucleotidase product of ATP, could stimulate the activation of BAT and concurrent browning of WAT [
      • Gnad T.
      • Scheibler S.
      • von Kugelgen I.
      • Scheele C.
      • Kilic A.
      • Glode A.
      • et al.
      Adenosine activates brown adipose tissue and recruits beige adipocytes via A2A receptors.
      ]. More recently, Razzoli and colleagues demonstrated that mice deficient of all β-adrenergic receptors (β-less) still can upregulate UCP1 in BAT under cold exposure [
      • Razzoli M.
      • Frontini A.
      • Gurney A.
      • Mondini E.
      • Cubuk C.
      • Katz L.S.
      • et al.
      Stress-induced activation of brown adipose tissue prevents obesity in conditions of low adaptive thermogenesis.
      ]. Compared to WT mice, these β-less mice also show an upregulation of P2X7 and P2Y12 receptors in BAT, but no changes in A2A or A1 receptors. Furthermore, the authors have shown that ATPγS, an ATP analogue that is resistant to phosphohydrolysis, can stimulate UCP1 expression in both β-less and WT mice, and the effect in the β-less mice is double that of WT mice. This observation suggests a direct impact of purinergic modulation of thermogenesis and could be attenuated by NTPDase3 deficiency.
      NTPDase3 regulates the relative concentrations of ATP vs. AMP in the extracellular milieu. The global deficiency of NTPDase3 would be expected to shift the balance toward the accumulation of ATP and relative decrease in the levels of adenosine. Our observation therefore provides further evidence for purinergic regulation in the thermogenesis of BAT and browning of WAT. These findings also suggest that alterations in purinergic regulation, as in NTPDase3 deficiency per se, may outweigh the loss of stimulation from the adenosinergic pathway in adipose tissues.
      Although mRNA transcripts of NTPDase3 are present in human adipose tissues [
      • Fagerberg L.
      • Hallstrom B.M.
      • Oksvold P.
      • Kampf C.
      • Djureinovic D.
      • Odeberg J.
      • et al.
      Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics.
      ], no significant NTPDase3 protein expression has been detected in WAT or BAT adipocytes. However, NTPDase3 is highly expressed by visceral autonomic nerves, and purinergic regulation may occur at sympathetic nerve endings, where ATP is released. We have previously reported that NTPDase2 or 3 in enteric nerves can modulate intestinal inflammation [
      • Feldbrugge L.
      • Moss A.C.
      • Yee E.U.
      • Csizmadia E.
      • Mitsuhashi S.
      • Longhi M.S.
      • et al.
      Expression of Ecto-nucleoside triphosphate Diphosphohydrolases-2 and -3 in the enteric nervous system affects inflammation in experimental colitis and Crohn’s disease.
      ]. Alternatively, NTPDase3 is also found in central nervous system, such as the hypothalamus, spinal cord, and dorsal root ganglia, and modulate energy homeostasis through central mechanisms and manifest via sympathetic innervation [
      • McCoy E.
      • Street S.
      • Taylor-Blake B.
      • Yi J.
      • Edwards M.
      • Wightman M.
      • et al.
      Deletion of ENTPD3 does not impair nucleotide hydrolysis in primary somatosensory neurons or spinal cord.
      ]. Further studies are required to dissect these mechanisms potentially through neuronal and other targeted and specific knockouts of NTPDase3.
      Our study has limitations and also raised several additional questions. First of all, the conclusion is limited by the experimental setting. IHC has showed that the Ins-Cre knockout has significantly decreased NTPDase3 expression in beta cells, but residual protein expression cannot be ruled out. Secondly, we have focused studies using female mice, as the impact of NTPDase3 on DIO appeared to be more robust in females than males. Whether this difference is pathophysiologically relevant needs to be further investigated and if so, the mechanism behind this difference could be interesting. Third, the changes in glucose homeostasis in NTPDase3 deficient mice could be entirely due to a leaner body weight. We have not identified an alternative explanation. Indeed, the metabolic role of NTPDase3 in vivo remains incompletely understood, as gene deletion paradoxically resulted in a favorable phenotype with no obvious deleterious consequences noted. These metabolic changes may in part involve the pro-opiomelanocortin and Agouti-related protein expressing neurons in the hypothalamus, due to their importance in energy metabolism [
      • Boswell T.
      • Dunn I.C.
      Regulation of agouti-related protein and pro-Opiomelanocortin gene expression in the avian Arcuate nucleus.
      ]. It is possible that NTPDase3 plays a role in energy conservation under stress situations, such as starvation and cold exposure. The roles of NTPDase3 in the adipocyte differentiation and mitochondria energetics require further characterization. Similarly, this study indicates that NTDase3 does not alter the absorptive function of the digestive tract, despite the high level of expression in the enteric nerves. The lack of phenotype relevant to insulin and glucose homeostasis in the Entpd3−/− and beta-cell targeted deletion argues for a different pathophysiological function such as neuroimmune regulation of metabolism.
      In summary, we found that the global deficiency of NTPDase3 increases basal energy metabolism, at least in part, mediated by alterations in BAT and browning of WAT. These changes result in protection from DIO. Unlike as previously suspected, NTPDase3 in the pancreatic beta cells does not significantly contribute to this phenotype and we propose the involvement of autonomic nerves in the modulation of BAT bioenergetics. This regulatory role of NTPDase3 in energy metabolism could be targeted for new treatments of obesity-related metabolic diseases.

      Grant support

      This work is in part supported by the Fulbright Scholarship to BS; DFG FE1434/1-1 , FE1434/2-1 from German Research Foundation to LF; PJT-156205 from Canadian Institutes of Health Research , “Chercheur National” Scholarship from the Fonds de Recherche du Quebec-Sante to JS; 1R21CA221702 and R01HL094400 from National Institutes of Health (NIH) to SCR; NIH K08DK115883 , a Clinical Research Award from ACG and the Alan Hofmann Clinical and Translational Research Award from AASLD to ZGJ.

      Declaration of competing interest

      The authors declare no conflict of interest relevant to this study.

      Appendix A. Supplementary data

      References

        • Eltzschig H.K.
        • Sitkovsky M.V.
        • Robson S.C.
        Purinergic signaling during inflammation.
        N Engl J Med. 2012; 367: 2322-2333
        • Robson S.C.
        • Sévigny J.
        • Zimmermann H.
        The E-NTPDase family of ectonucleotidases: structure function relationships and pathophysiological significance.
        Purinergic Signal. 2006; 2: 409-430
        • Zimmermann H.
        • Zebisch M.
        • Strater N.
        Cellular function and molecular structure of ecto-nucleotidases.
        Purinergic Signal. 2012; 8: 437-502
        • Antonioli L.
        • Pacher P.
        • Vizi E.S.
        • Hasko G.
        CD39 and CD73 in immunity and inflammation.
        Trends Mol Med. 2013; 19: 355-367
        • Braun N.
        • Sévigny J.
        • Robson S.C.
        • Hammer K.
        • Hanani M.
        • Zimmermann H.
        Association of the ecto-ATPase NTPDase2 with glial cells of the peripheral nervous system.
        Glia. 2004; 45: 124-132
        • Feldbrugge L.
        • Moss A.C.
        • Yee E.U.
        • Csizmadia E.
        • Mitsuhashi S.
        • Longhi M.S.
        • et al.
        Expression of Ecto-nucleoside triphosphate Diphosphohydrolases-2 and -3 in the enteric nervous system affects inflammation in experimental colitis and Crohn’s disease.
        J Crohns Colitis. 2017; 11: 1113-1123
        • Dranoff J.A.
        • Kruglov E.A.
        • Robson S.C.
        • Braun N.
        • Zimmermann H.
        • Sévigny J.
        The ecto-nucleoside triphosphate diphosphohydrolase NTPDase2/CD39L1 is expressed in a novel functional compartment within the liver.
        Hepatology. 2002; 36: 1135-1144
        • Fausther M.
        • Lecka J.
        • Kukulski F.
        • Levesque S.A.
        • Pelletier J.
        • Zimmermann H.
        • et al.
        Cloning, purification, and identification of the liver canalicular ecto-ATPase as NTPDase8.
        Am J Physiol Gastrointest Liver Physiol. 2007; 292: G785-G795
        • Lavoie E.G.
        • Gulbransen B.D.
        • Martin-Satue M.
        • Aliagas E.
        • Sharkey K.A.
        • Sévigny J.
        Ectonucleotidases in the digestive system: focus on NTPDase3 localization.
        Am J Physiol Gastrointest Liver Physiol. 2011; 300: G608-G620
        • Lavoie E.G.
        • Fausther M.
        • Kauffenstein G.
        • Kukulski F.
        • Kunzli B.M.
        • Friess H.
        • et al.
        Identification of the ectonucleotidases expressed in mouse, rat, and human Langerhans islets: potential role of NTPDase3 in insulin secretion.
        Am J Physiol Endocrinol Metab. 2010; 299: E647-E656
        • Saunders D.C.
        • Brissova M.
        • Phillips N.
        • Shrestha S.
        • Walker J.T.
        • Aramandla R.
        • et al.
        Ectonucleoside triphosphate Diphosphohydrolase-3 antibody targets adult human pancreatic beta cells for in vitro and in vivo analysis.
        Cell Metab. 2019; 29 ([e744]): 745-754
        • Deaglio S.
        • Robson S.C.
        Ectonucleotidases as regulators of purinergic signaling in thrombosis, inflammation, and immunity.
        Adv Pharmacol. 2011; 61: 301-332
        • Enjyoji K.
        • Kotani K.
        • Thukral C.
        • Blumel B.
        • Sun X.
        • Wu Y.
        • et al.
        Deletion of cd39/entpd1 results in hepatic insulin resistance.
        Diabetes. 2008; 57: 2311-2320
        • Petit P.
        • Hillaire-Buys D.
        • Manteghetti M.
        • Debrus S.
        • Chapal J.
        • Loubatieres-Mariani M.M.
        Evidence for two different types of P2 receptors stimulating insulin secretion from pancreatic B cell.
        Br J Pharmacol. 1998; 125: 1368-1374
        • Loubatieres-Mariani M.M.
        • Chapal J.
        Purinergic receptors involved in the stimulation of insulin and glucagon secretion.
        Diabete Metab. 1988; 14: 119-126
        • Syed S.K.
        • Kauffman A.L.
        • Beavers L.S.
        • Alston J.T.
        • Farb T.B.
        • Ficorilli J.
        • et al.
        Ectonucleotidase NTPDase3 is abundant in pancreatic beta-cells and regulates glucose-induced insulin secretion.
        Am J Physiol Endocrinol Metab. 2013; 305: E1319-E1326
        • Kraus D.
        • Yang Q.
        • Kahn B.B.
        Lipid extraction from mouse feces.
        Bio Protoc. 2015; 5
        • Mina A.I.
        • LeClair R.A.
        • LeClair K.B.
        • Cohen D.E.
        • Lantier L.
        • Banks A.S.
        CalR: a web-based analysis tool for indirect Calorimetry experiments.
        Cell Metab. 2018; 28 ([e651]): 656-666
        • Galarraga M.
        • Campion J.
        • Munoz-Barrutia A.
        • Boque N.
        • Moreno H.
        • Martinez J.A.
        • et al.
        Adiposoft: automated software for the analysis of white adipose tissue cellularity in histological sections.
        J Lipid Res. 2012; 53: 2791-2796
        • Lein E.S.
        • Hawrylycz M.J.
        • Ao N.
        • Ayres M.
        • Bensinger A.
        • Bernard A.
        • et al.
        Genome-wide atlas of gene expression in the adult mouse brain.
        Nature. 2007; 445: 168-176
        • Morrison S.F.
        • Madden C.J.
        • Tupone D.
        Central neural regulation of brown adipose tissue thermogenesis and energy expenditure.
        Cell Metab. 2014; 19: 741-756
        • Bartness T.J.
        • Liu Y.
        • Shrestha Y.B.
        • Ryu V.
        Neural innervation of white adipose tissue and the control of lipolysis.
        Front Neuroendocrinol. 2014; 35: 473-493
        • Adamson S.E.
        • Meher A.K.
        • Chiu Y.H.
        • Sandilos J.K.
        • Oberholtzer N.P.
        • Walker N.N.
        • et al.
        Pannexin 1 is required for full activation of insulin-stimulated glucose uptake in adipocytes.
        Mol Metab. 2015; 4: 610-618
        • Gnad T.
        • Scheibler S.
        • von Kugelgen I.
        • Scheele C.
        • Kilic A.
        • Glode A.
        • et al.
        Adenosine activates brown adipose tissue and recruits beige adipocytes via A2A receptors.
        Nature. 2014; 516: 395-399
        • Razzoli M.
        • Frontini A.
        • Gurney A.
        • Mondini E.
        • Cubuk C.
        • Katz L.S.
        • et al.
        Stress-induced activation of brown adipose tissue prevents obesity in conditions of low adaptive thermogenesis.
        Mol Metab. 2016; 5: 19-33
        • Fagerberg L.
        • Hallstrom B.M.
        • Oksvold P.
        • Kampf C.
        • Djureinovic D.
        • Odeberg J.
        • et al.
        Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics.
        Mol Cell Proteomics. 2014; 13: 397-406
        • McCoy E.
        • Street S.
        • Taylor-Blake B.
        • Yi J.
        • Edwards M.
        • Wightman M.
        • et al.
        Deletion of ENTPD3 does not impair nucleotide hydrolysis in primary somatosensory neurons or spinal cord.
        F1000Res. 2014; 3: 163
        • Boswell T.
        • Dunn I.C.
        Regulation of agouti-related protein and pro-Opiomelanocortin gene expression in the avian Arcuate nucleus.
        Front Endocrinol (Lausanne). 2017; 8: 75