Organic Cation Transporter 3 Contributes to Norepinephrine Uptake Into Perivascular Adipose Tissue

Am J Physiol Heart Circ Physiol 309: H1904–H1914, 2015. First published October 2, 2015; doi:10.1152/ajpheart.00308.2015.

Organic cation transporter 3 contributes to norepinephrine uptake into perivascular adipose tissue

X Nadia Ayala-Lopez,1 William F. Jackson,1 Robert Burnett,1 James N. Wilson,2 Janice M. Thompson,1 and Stephanie W. Watts1 1Department of Pharmacology and Toxicology, Michigan State University, East Lansing, Michigan; and 2Department of Chemistry, University of Miami, Coral Gables, Florida Submitted 30 April 2015; accepted in ﬁnal form 30 September 2015

Ayala-Lopez N, Jackson WF, Burnett R, Wilson JN, Thompson fourth layer of the blood vessel, the “tunica adiposa” (14). JM, Watts SW. Organic cation transporter 3 contributes to norepi- Beyond providing structural support, PVAT has many roles in nephrine uptake into perivascular adipose tissue. Am J Physiol Heart modulating blood vessel function (68). The release of vasoac- Circ Physiol 309: H1904–H1914, 2015. First published October 2, tive molecules from PVAT influences vascular function by 2015; doi:10.1152/ajpheart.00308.2015.—Perivascular adipose tissue (PVAT) reduces vasoconstriction to norepinephrine (NE). A mecha- altering the proliferation, migration, inflammation, and con- nism by which PVAT could function to reduce vascular contraction is traction of vascular smooth muscle (9, 20–22, 24, 46, 68, 72). by decreasing the amount of NE to which the vessel is exposed. Interestingly, a releasable pool of catecholamines is present in PVATs from male Sprague-Dawley rats were used to test the hypoth- PVAT (5, 67). Although both contractile and anticontractile esis that PVAT has a NE uptake mechanism. NE was detected by substances can be released from PVAT (9, 22, 24, 25, 72), the HPLC in mesenteric PVAT and isolated adipocytes. Uptake of NE (10 presence of PVAT on blood vessels generally reduces vessel ␮M) in mesenteric PVAT was reduced by the NE transporter (NET) contraction in response to various agonists, including norepi- inhibitor nisoxetine (1 ␮M, 73.68 Ϯ 7.62%, all values reported as nephrine (NE) (64). Knowledge on how these mechanisms percentages of vehicle), the 5-hydroxytryptamine transporter (SERT) interact to influence the anticontractile properties of PVAT in inhibitor citalopram (100 nM) with the organic cation transporter 3 (OCT3) inhibitor corticosterone (100 ␮M, 56.18 Ϯ 5.21%), and the NE-induced contraction is not complete (36). NET inhibitor desipramine (10 ␮M) with corticosterone (100 ␮M, The anticontractile effect of PVAT is lost in obesity and 61.18 Ϯ 6.82%). Aortic PVAT NE uptake was reduced by cortico- hypertension, implicating PVAT as an integral link between sterone (100 ␮M, 53.01 Ϯ 10.96%). Confocal imaging of mesenteric both of these diseases (3). Over one-third of all adults in the PVAT stained with 4-[4-(dimethylamino)-styrl]-N-methylpyridinium United States are hypertensive (13), a condition that signifi- iodide (ASPϩ), a fluorescent substrate of cationic transporters, de- cantly increases the risk of death from myocardial infarction or ϩ ϩ tected ASP uptake into adipocytes. ASP (2 ␮M) uptake was stroke (42). A major risk factor for hypertension is obesity reduced by citalopram (100 nM, 66.68 Ϯ 6.43%), corticosterone (100 (29). Globally, 13% of adults are obese (have a body mass ␮M, 43.49 Ϯ 10.17%), nisoxetine (100 nM, 84.12 Ϯ 4.24%), citalo- Ն ␮ Ϯ index of 30) (73a), and in the United States, the number is pram with corticosterone (100 nM and 100 M, respectively, 35.75 higher, with 34.9% of adults classified as obese (53). In 4.21%), and desipramine with corticosterone (10 and 100 ␮M, respectively, 50.47 Ϯ 5.78%). NET protein was not detected in mes- obesity, dysfunction of the anticontractile effect of PVAT is enteric PVAT adipocytes. Expression of Slc22a3 (OCT3 gene) observed along with overall changes in adipocyte function (3). mRNA and protein in PVAT adipocytes was detected by RT-PCR and Thus, the relationship between the adipocytes within PVAT immunocytochemistry, respectively. These end points support the and blood vessel function is of interest. presence of a transporter-mediated NE uptake system within PVAT A dynamic adrenergic system that affects blood vessel with a potential mediator being OCT3. contraction exists in PVAT (5). Catecholamines are released 4-[4-(dimethylamino)-styrl]-N-methylpyridinium iodide; norepineph- from PVAT upon the addition of tyramine, a sympathomimetic rine uptake; high-performance liquid chromotography; cat- drug, leading to contraction of the rat aorta and superior echolamines; perivascular adipose tissue; adipocytes mesenteric artery (5). Moreover, pharmacological inhibition of NE transporters (NET) reduces the PVAT-dependent vascular contraction to tyramine (5). Soltis and Cassis (64) discovered that inhibition of NE uptake in the rat aorta abolished the NEW & NOTEWORTHY anticontractile effect of PVAT. Collectively, this work led us to Perivascular adipose tissue (PVAT) reduces vasoconstriction investigate the presence of a NE uptake system in PVAT. to norepinephrine (NE); thus, we tested the hypothesis that The present study tested the hypothesis that PVAT takes up PVAT has a NE uptake mechanism. Functional NE uptake NE through molecular transporters and aims to identify the assays, fluorescence imaging, and PCR of mesenteric and transporters that transport NE in PVAT. Our interest in study- aortic PVAT revealed the potential role of organic cation ing mesenteric PVAT is guided by the knowledge that con- transporter 3 in NE uptake. traction in mesenteric resistance arteries increases peripheral PERIVASCULAR ADIPOSE TISSUE (PVAT) closely envelops many resistance, a contributing event toward the elevation of blood blood vessels of the body (65). This relationship between pressure. Furthermore, this adipose depot is important for PVAT and the blood vessel has earned PVAT its place as the cardiovascular risk. Individuals with large masses of visceral fat have a higher risk of cardiovascular disease than individuals Address for reprint requests and other correspondence: N. Ayala-Lopez, with large masses of subcutaneous fat (40). Dept. of Pharmacology and Toxicology, Michigan State Univ., 1355 Bogue PVAT of mesenteric resistance arteries most closely resem- St., Rm B445, East Lansing, MI 48824 (e-mail: [email protected]). bles white adipose tissue in that it contains adipocytes that have

H1904 0363-6135/15 Copyright © 2015 the American Physiological Society http://www.ajpheart.org NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE H1905 large unilocular lipid droplets (10). We focused on adipocytes allowed to equilibrate to temperature for 15 min, after which a background image was captured. For the ASPϩ concentration uptake from normal rats as studies of NE transport in adipocytes are ϩ sparse and none have been performed on PVAT adipocytes experiment, the tissue was superfused with PSS containing ASP (1 ␮ specifically. To test our hypothesis, we used PVAT from nM–10 M) for 10 min and imaged. To test each concentration, a new normal male Sprague-Dawley rats for HPLC measures of NE section of tissue was used from the same animal and the order in which the concentrations were tested was randomized. Each tissue in PVAT and isolated adipocytes. We also measured uptake of was only used for one condition. For ASPϩ uptake experiments in NE and used pharmacological inhibitors to transporters to which inhibitors or NE were used, the tissue was superfused with an reveal the main transporters that transport NE in PVAT. inhibitor of transport, NE (1 mM), or vehicle in PSS for 10 min, and Confocal microscopy of PVAT was used using the fluorescent an image was captured to assess background fluorescence. ASPϩ (2 ϩ NE transport substrate dye methylpyridinium ASP (61) in ␮M) was added for 10 min, and the tissue was imaged again. For the addition to immunohistochemistry, immunocytochemistry, and ASPϩ concentration uptake experiments and ASPϩ uptake experi- gene expression analysis of mesenteric and aortic PVAT to ments where nisoxetine or citalopram were used, tissue imaging was reveal the role of organic cation transporter 3 (OCT3). performed with a Leica DMLFSA confocal microscope (Leica Mi- crosystems, Wetzlar, Germany) equipped with a Yokogawa CSU10 MATERIALS AND METHODS spinning disk confocal head (Yokogawa, Tokyo, Japan) coupled to a XR-Mega10 intensified charge-coupled device (Stanford Photonics, Chemicals. Chemicals were purchased from Sigma-Aldrich (St. Palo Alto, CA) using a ϫ40 water-immersion objective. Illumination Louis, MO). The pharmacological inhibitors nisoxetine (inhibits was provided by an X-cite Exacte Illuminator (Excelitas Technolo- NET), citalopram [inhibits the 5-hydroxytryptamine (serotonin) trans- gies, Waltham, MA). Images were recorded with Piper-Control (Stan- porter (SERT)], corticosterone (inhibits OCT3), and desipramine ford Photonics) and analyzed using ImageJ (NIH). For the corticoste- [inhibits NET and SERT at the concentration used (10 ␮M)] were rone, citalopram with corticosterone, desipramine, desipramine with purchased from Bio-Techne (Minneapolis, MN). Pargyline (a mono- corticosterone, and NE experiments, the protocol was the same as amine oxidase inhibitor), Ro 41-0960 (a catechol-o-methyltransferase above except that a solid-state 488-nm laser was used for illumination inhibitor), and NE were purchased from Sigma-Aldrich. ASPϩ was and a TurboEX ICCD camera (Stanford Photonics) controlled by synthesized and provided by James N. Wilson (University of Miami, ␮Manager (66a) was used for image acquisition. Images were cap- Miami, FL) (73). tured as stacks of 50 TIFF (16-bit) images, which were then combined Animals. Male Sprague-Dawley rats (225–275 g or ϳ8–10 wk of with the average z-projection function in ImageJ. Fluorescence inten- age, Charles River, Indianapolis, IN) were used. All protocols were sity was quantified in relative fluorescent units. approved by the Institutional Animal Care and Use Committee of Sample preparation of the mesenteric PVAT, mesenteric resistance Michigan State University and followed the National Institutes of vessels, adipocytes, and stromal vascular fraction. Mesenteric PVAT Health (NIH) Guide for the Care and Use of Laboratory Animals (8th and mesenteric resistance vessels were dissected in a Sylgard-coated ed., 2011). Rats were anesthetized with pentobarbital sodium (60–80 petri dish in PSS with the use of stereomicroscope. Images of the mg/kg ip). Anesthesia was verified by lack of paw pinch and eye blink whole mesentery were captured with a Lumix DMC-ZS25 camera reflexes. Death was assured by pneumothorax and exsanguination, (Panasonic, Osaka, Japan) and processed using Adobe Photoshop CC after which tissues were removed for one of the following protocols. 2014 (Adobe Systems, San Jose, CA). PVAT was either flash frozen NE uptake. Mesenteric and aortic PVATs were dissected, and for whole PVAT measurements or digested to obtain separate cellular 20–100 mg of tissue were placed in microcentrifuge tubes containing fractions by the following protocol. PVAT was added to 1 ml PSS physiological salt solution (PSS) [containing (in mM) 130 NaCl, 4.7 with 1 mg/ml collagenase from Clostridium histolyticum type IA KCl, 1.8 KH2PO4, 1.7 MgSO4·7H2O, 14.8 NaHCO3, 5.5 dextrose, (catalog no. C9891, Sigma-Aldrich) and incubated at 37°C with slow 0.03 CaNa2 EDTA, and 1.6 CaCl2 (pH 7.2)] within 30 min of tissue rotation until fully digested (ϳ1 h). PVAT was centrifuged at 200 g removal from the rat. Pargyline (10 ␮M) and Ro 41-0960 (1 ␮M) for 5 min, and the stromal vascular fraction (SVF), which pellets to were added to the PSS to inhibit NE metabolism. Vehicle or a the bottom, was transferred to a separate tube. Adipocytes and the transporter inhibitor [nisoxetine (1 ␮M), citalopram (100 nM), corti- SVF were washed three times with PSS and centrifuged at 200 g for costerone (100 ␮M), citalopram (100 nM) with corticosterone (100 10 min. For immunocytochemistry, mesenteric PVAT adipocytes ␮M), desipramine (10 ␮M), or corticosterone (100 ␮M) with desip- were resuspended in PSS and centrifuged onto CellTak (catalog no. ramine (10 ␮M)] was added for 30 min at 37°C. The concentrations 54240, BD Biosciences, Bedford, MA)-coated slides using a Cytospin were selected based on their specificity for the transporter in question. 4 cytocentrifuge (700 g for 2 min), and an aliquot was saved to assess NE (10 ␮M) or vehicle (either H2O or ethanol) was added for another purity using a hemacytometer. For Western blots, the mesenteric 30 min. Tissues were rinsed four times in drug-free PSS and then three PVAT, mesenteric resistance vessels, adipocytes, and SVF were times in tissue buffer (0.05 mM sodium phosphate and 0.03 mM citric added to RIPA buffer solution (catalog no. R3792, Teknova, Hollister, acid buffer, pH 2.5, in 15% methanol). Samples were saved in tissue CA) with protease inhibitors (0.5 mM PMSF, 1 mM orthovanadate, 10 buffer and kept at Ϫ80°C until assay. The day of the assay, samples ␮g/ml aprotinin, and 10 ␮g/ml leupeptin) into a 2-ml bead tube were thawed and sonicated for 3 s. Samples were centrifuged at (Omni, Kennesaw, GA). Tissues were homogenized using the Omni 18,000 g for 15 min at 4°C, and the supernatant was transferred to new Bead Ruptor Homogenizer (Omni) and centrifuged for 15 min at tubes for HPLC analysis. Tissue pellets were dissolved in 1.0 N 18,000 g, and supernatants were saved for Western blot analysis. NaOH and assayed for protein using a Bicinchoninic Acid Protein Supernatants were quantified for protein content using a Bicin- Assay Kit (catalog no. BCA1, Sigma-Aldrich). choninic Acid Protein Assay Kit (catalog no. BCA1, Sigma-Aldrich). ASPϩ uptake. The mesenteric arcade was dissected from Sprague- For mRNA isolation and HPLC analysis, the adipocytes and SVF Dawley rats and stored in PSS without calcium [containing (in mM) were placed into separate tubes with PSS and centrifuged 200 g for 10 140 NaCl, 5 KCl, 1 MgCl2·7H2O, 10 HEPES, and 10 glucose; pH 7.4] min, after which the supernatant was removed. The tissue was then at 4°C until use, for up to 5 h. Immediately before experiments, flash frozen in liquid nitrogen and saved at Ϫ80°C until assay. Images mesenteric resistance arteries with associated PVAT were dissected of the isolated adipocytes were taken on a Nikon TE2000 inverted and pinned onto the Sylgard-coated bottom of an imaging chamber microscope with MMI Cell Tools (Molecular Machines & Industries, (volume ϭ 1 ml) with the use of a stereomicroscope. Experiments Zurich, Switzerland). were performed in the dark or under safe lights at 37°C. The tissue Western blot analysis for NET. Fifty micrograms of protein from was superperfused with PSS with calcium (1.8 mM CaCl2·2H2O) and the mesenteric PVAT, mesenteric resistance vessels, adipocytes, SVF,

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org H1906 NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE and vena cava (positive control) were separated on a 10% SDS gel and Mann-Whitney test was used when the variances were different (as transferred to a polyvinylidene difluoride membrane. The membrane verified by the F-test). When more than two groups were compared, was blocked in 4% (wt/vol) chicken egg ovalbumin in Tris-buffered ANOVA with a Newman-Keuls test was used. With non-normally saline and Tween 20 (TBST) for3hat4°Candthen incubated with distributed data, Kruskal-Wallis ANOVA was used followed by a primary antibody [mouse anti-NET (1:500, NET05-2, MAb Technol- Dunn’s test for multiple comparisons. The tests were unpaired. P ogies, Stone Mountain, GA) and mouse anti-␤-actin (1:2,000, A3854, values of Ͻ0.05 were considered statistically significant. Means Ϯ SE Sigma)] diluted in blocker overnight at 4°C. The blot was washed with are reported where appropriate. To calculate percent NE uptake, the TBST (10 min each, three times) and then incubated with IRDye concentration of NE in tissues incubated with NE and the pharmaco- anti-mouse secondary antibody (1:1,000, no. 926-32210, Li-Cor, logical inhibitor were divided by the concentration of NE in tissue Lincoln, NE) diluted in Odyssey Blocking Buffer (no. 927-40000, incubated with NE and the vehicle. Image contrast for the ASPϩ Li-Cor) for1hat4°C. The blot was washed with TBST (10 min each, experiments was normalized to the brightest image recorded in the three times) and developed on the Li-Cor Odyssey. Densitometric data set. Pseudocolorization of ASPϩ-stained images was performed analysis was done using ImageJ. using the “Fire” lookup table and constructed surface plots by apply- Preparation of aortic PVAT for PCR and immunohistochemistry. ing the “surface plot” function in ImageJ (version 1.48). All image To obtain samples of aortic PVAT, the thoracic aorta was removed adjustments in brightness and contrast were made to the whole panel from the rat and placed into PSS. PVAT was then dissected from the of an image, not a portion. To calculate percent ASPϩ uptake, the aorta on a Sylgard-coated petri dish with the use of a stereomicro- fluorescence intensity ratio (F/F0) was used, where F is the fluores- scope. PVAT was removed and snap frozen in liquid nitrogen for cence intensity after incubation with the inhibitor and ASPϩ minus the ϩ RNA extraction. The Investigative Histopathology Laboratory at background intensity and F0 is the fluorescence intensity of ASP Michigan State University prepared the fresh frozen rat aorta slides. incubated with vehicle minus the background intensity. Real-time PCR. Tissue was homogenized using an Omni Bead Ruptor (Omni). RNA was extracted with the Quick RNA MiniPrep kit RESULTS (catalog no. R1054, Zymo Research, Irving, CA), and purity (260-to- 280 and 260-to-230 ratios Ն 1.8) was verified using a Nanodrop NE is present in rat mesenteric PVAT. A large proportion of 2000C spectrophotometer (Thermoscientific, Wilmington, DE). NE found in PVAT is in the adipocyte fraction (Fig. 1A). mRNA was reverse transcribed with qScript cDNA SuperMix Measures of NE by HPLC in mesenteric PVAT and isolated (Quanta Biosciences, Gaithersburg, MD). RT-PCR was performed mesenteric PVAT adipocytes were similar (P Ͼ 0.05). Repre- using FAST SYBR Green MasterMix (catalog no. 4385612, Applied sentative images of each sample are shown in Fig. 1A, bottom Biosystems, Foster City, CA) on the ABI 7500 Fast Real Time PCR left and bottom right. The white box in Fig. 1A, bottom left, system (Life Technologies, Carlsbad, CA) using the following param- highlights a representative area of tissue that was used. One eters: 95°C for 20 s, 95°C for 1 s, and 60°C for 20 s for 40 cycles. The primer sequences for Slc22a3 (OCT3 gene) amplification were as mechanism by which NE could localize to adipocytes in PVAT follows: forward 5=-TATGCAGCGGACAGATACGG-3= and reverse is through transporter-mediated uptake of extracellular NE; 5=-AAAATTCGGTGCAAACGCCA-3= (Integrated DNA Technolo- thus, we investigated this further. gies, Coralville, IA). Measures were normalized to ␤2-microglobulin NE uptake occurs in PVAT. We used a pharmacological (RT2 qPCR Primer Assay, catalog no. PPR42607A, Qiagen, Valencia, approach to test the hypothesis that uptake of NE into PVAT is CA). A melt curve was performed to verify the presence of one PCR Ϫ⌬⌬ product after the amplification. Data were analyzed using the 2 CT Mesenteric PVAT Mesenteric PVAT method (where CT is threshold cycle) (45). A B ␮ 25 Immunocyto/histochemistry. Fresh frozen 8- m aortic tissue sec- N=11 tions and adipocyte slides (described above) were fixed in acetone and * immunostained using a VECTASTAIN ABC kit (rabbit: catalog no. 20 PK-4001, Vector Laboratories, Burlingame, CA) and an Avidin- Biotin Blocking Kit (SP-2001, Vector Laboratories). Slides were 15 g protein) incubated for 24 h with anti-OCT3 antibody (1:100, catalog no. µ orb107605, Biorbyt, San Francisco, CA) or without primary antibody 10 at 4°C. Slides were developed using 3,3-diaminobenzidine (catalog no. SK-4100, Vector Laboratories) and counterstained with hematox- NE (pg/ 5 ylin (catalog no. H-3401, Vector Laboratories) for 30 s. Imaging was performed on a Nikon TE2000 inverted microscope with MMI Cell 0 Tools (Molecular Machines & Industries). HPLC. PVATs were weighed and homogenized in four times their vehicle weight of 0.1 M perchloric acid and centrifuged at 15,000 g for 10 10 µM NE min, and the supernatant was analyzed by HPLC. Supernatants from uptake experiments were diluted 1:10 in tissue buffer before analysis. The HPLC system consisted of a Coulochem III electrochemical detector set at Ϫ350 mV with a HR-80 reverse-phase column with Cat-A-Phase II mobile phase (Thermoscientific). The separation column was maintained at 35°C with a flow rate of 1.1 ml/min. Quan- Fig. 1. Norepinephrine (NE) is present in perivascular adipose tissue (PVAT) tification was performed by comparing sample area measurements adipocytes, and PVAT can take up extracellular NE. A: mesenteric PVAT with a calibration curve. Standards were run every fifth sample to adipocytes were isolated, and NE content was measured by HPLC. Measures verify the identity of our peaks of interest on the chromatogram. The were normalized to tissue weight. Bottom left, representative image of the mesenteric PVAT used. The white box highlights the portion of PVAT used in limit of detection was 0.1 ng/ml, and NE content was either expressed the experiments in this study. Bottom right, representative image of adipocytes per weight or by protein content. isolated from mesenteric PVAT. B: addition of NE (10 ␮M) to PVAT in Data analysis. Statistical analyses were performed using GraphPad physiological saline solution (PSS) for 30 min increased NE accumulation, as Prism 6.0 (La Jolla, CA). When two groups were compared, either an measured by HPLC and normalized to protein content (*P Ͻ 0.05). Values are unpaired Student’s t-test was used with similar variances and the means Ϯ SE for the number of animals (N) shown.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE H1907

NE (10 µM) periphery of the adipocyte, where the cytoplasm is located 150 (Fig. 4A, top). The surface plot image of the ﬂuorescence intensity levels with pseudocolorization showed this more 7 clearly (Fig. 4A, bottom). Confocal imaging at the level of the adipocyte nucleus revealed intense staining around the adi- 100 10 10 pocyte nuclei, an area rich in mitochondria in white adipose 16 tissue adipocytes (Fig. 4B, with the top showing an image * 9 11 without color information and the bottom showing the image * * with pseudocolor; arrows point to perinuclear staining) (19). Perinuclear ASPϩ ﬂuorescence in PVAT adipocytes suggests 50 transport of ASPϩ into the adipocyte as opposed to only

NE uptake (% vehicle) surface binding. Preincubation of PVAT with an excess of NE (1 mM) reduced ASPϩ uptake (Fig. 5). Any residual binding of ASPϩ 0 in the presence of this saturating concentration of NE can be considered nonspecific to NE transport mechanisms. To identify which transporters were involved in ASPϩ uptake, mesenteric PVAT was incubated at 37°C in PSS containing an ϩ 1 µM nisoxetine inhibitor of transport or vehicle for 10 min followed by ASP 100 nM citalopram 100 nM citalopram10 µM desipramine10 µM desipramine (2 ␮M) and imaged (Fig. 5). Representative images are shown 100 µM corticosterone ϩ + 100 µM corticosterone + 100 µM corticosterone of the fluorescence obtained with vehicle only with ASP (2 ␮M), background (no ASPϩ), and each experiment after pre- Fig. 2. PVAT accumulates NE through transporter-mediated uptake. Mesen- ϩ teric PVAT was incubated for 30 min with transporter inhibitors before the incubation with NE or an inhibitor of transport (Fig. 5). ASP addition of 10 ␮M NE. Data are reported as percent uptake from vehicle. uptake was significantly reduced by inhibition of NET by Values are means Ϯ SE. Numbers above the bars are N values for each nisoxetine (100 nM and 10 ␮M), SERT by citalopram (100 inhibitor. *P Ͻ 0.05 vs. vehicle. nM), OCT3 by corticosterone (100 ␮M), SERT and OCT3 by citalopram with corticosterone (100 nM and 100 ␮M, respec- mediated by transporters. First, we established that NE uptake tively), and NET and OCT3 by desipramine with corticoste- occurs in PVAT (Fig. 1B). PVAT was then incubated with NE rone (10 and 100 ␮M, respectively). However, uptake was not (10 ␮M) or vehicle in PSS for 30 min, the tissue was washed significantly reduced by desipramine alone (10 ␮M). Binding to remove excess NE, and NE in the tissue was measured by and transport of ASPϩ in PVAT adipocytes thus may be HPLC. As shown in Fig. 1B, mesenteric PVAT NE content mediated by NET, SERT, and OCT3, consistent with our was significantly increased after the addition of NE (10 ␮M) experiments of NE transport into PVAT (Fig. 2). compared with the addition of vehicle (18.87 Ϯ 2.42 vs. 4.48 Ϯ 1.98 pg/␮g protein, respectively, P Ͻ 0.05). This NE uptake could be reduced by the inhibition of NE transport by prein- A cubation with inhibitors of NET, SERT, and OCT3: nisoxetine (1 ␮M), citalopram (100 nM) with corticosterone (100 ␮M), and desipramine (10␮M) with corticosterone (100 ␮M). De- sipramine, citalopram, or corticosterone alone did not significantly reduce NE uptake compared with vehicle (Fig. 2). These 1 nM 10 nM 100 nM 1 µM 2 µM 10 µM data support transporter-mediated uptake of NE in mesenteric B PVAT through NET, SERT, and OCT3. 250 ϩ ASP fluorescently labels monoamine transporters on PVAT N=3-4 adipocytes. The transporter substrate ASPϩ was used to iden- 200 ϩ tify the presence of NET on PVAT adipocytes. ASP fluo- 150 resces upon binding and is transported and accumulates in mitochondria (61). To determine if ASPϩ would bind to PVAT 100 ϩ RFU * adipocytes, we added increasing concentrations of ASP to 50 mesenteric PVAT and imaged the tissue by fluorescence microscopy, and a graph of the concentration-fluorescence inten- 0 sity relationship was constructed (Fig. 3A). Quantification of -9 -8 -7 -6 -5 ϩ -50 the intensity of ASP fluorescence and comparison of fluores- log[ASP+], M cence intensity between vehicle (water) and each concentration Fig. 3. ASPϩ, a fluorescent substrate of cation transporters, binds to mesenteric revealed a significant increase in fluorescence from vehicle ϩ ϩ ϩ PVAT adipocytes. ASP fluorescence in PVAT was tested by adding ASP at starting at 1 ␮M ASP . Fluorescence intensity saturated the different concentrations for 10 min and then imaging by fluorescence micros- camera at 10 ␮M ASPϩ (Fig. 3B). A concentration of 2 ␮M copy. A: representative images of ASPϩ at each concentration tested. B: ϩ ASPϩ was chosen for subsequent experiments to achieve a quantification of ASP fluorescence intensity at each concentration. *P Ͻ 0.05 indicates the first concentration with significantly increased fluorescence vs. detectable fluorescence signal while avoiding camera satura- background. The fluorescence intensity was expressed in relative fluorescence tion. By imaging the adipocyte at the focal plane that trans- units (RFU). Values are means Ϯ SE for the number of animals (N) shown. ϩ verses the adipocyte, ASP fluorescence was localized to the Samples were imaged with a ϫ40 objective.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org H1908 NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE A B 2 µM ASP+ 2 µM ASP+

Fig. 4. ASPϩ is transported into mesenteric PVAT adipocytes. ASPϩ (2 ␮M) was added to mesenteric PVAT for 10 min in PSS and imaged by confocal microscopy. A: pseudo- colored representation of the fluorescence intensity of ASPϩ (2 ␮M) binding revealing that ASPϩ binds along on the periphery of the adipocyte in a punctate pattern (top). The surface plot of the same image shows this more clearly (bottom). Refer to the RFU scale to the right. Lighter (white-yellow) colors indicate higher fluorescence intensi- ties as measured by RFU, and darker (violet- black) colors indicate lower fluorescence in- tensities. B, top: image of a PVAT adipocyte imaged at the level of the nucleus. Arrows point to areas of perinuclear staining. Pseudocolorization of the images (bottom) shows that intense staining was present in the perinuclear region. Images are representative of six animals. Scale bars ϭ 50 ␮m. Samples were imaged with a ϫ40 objective.

NET is not present in mesenteric PVAT adipocytes. The staining for OCT3 was located to the tunica media (labeled M presence of NET in the mesentery was investigated because of in Fig. 8A) but not the tunica adventitia (labeled V in Fig. 8A). the modest effect of nisoxetine on ASPϩ uptake. Protein from Immunostaining for OCT3 was present on the periphery of the mesenteric PVAT, mesenteric resistance vessels, mesenteric adipocytes in both aortic and mesenteric PVATs (Fig. 8, A and PVAT adipocytes, and mesenteric PVAT SVF were assayed B). Immunostaining was not present when the primary anti- for NET by Western blot analysis. We did not observe bands body was excluded (Fig. 8, C and D). Inhibition of OCT3 with for NET in adipocytes or the SVF but did observe some faint corticosterone (100 ␮M) reduced NE uptake in aortic PVAT by bands for NET in the mesenteric PVAT, resistance vessels, and 47.0 Ϯ 11.0% (Fig. 8E). These data support the presence of vena cava, our positive control (Fig. 6), indicating that the NET OCT3 on adipocytes and the potential for OCT3 to transport is most likely not the main transporter that is mediating uptake NE in PVAT. of NE in PVAT. Corticosterone and corticosterone with cita- ϩ lopram caused the greatest reduction in ASP uptake (Fig. 5); DISCUSSION therefore, we focused on OCT3 for the rest of the experiments. OCT3 is present in PVAT adipocytes. PVAT expression of PVAT modulates blood vessel function (5, 9, 14, 67, 68), Slc22a3 (the gene for OCT3) was compared with its relative and, in the current study, we present evidence that at least part expression in the heart, a positive control (15) (Fig. 7A). of this can occur through direct NE uptake. The novel discov- Mesenteric and aortic PVATs expressed Slc22a3 at higher ery of NE uptake by PVAT could present a mechanism by relative expression than in the heart (Fig. 7, A and B). Mesen- which adipose tissue reduces the local concentration of NE, teric PVAT was separated into adipocytes and the SVF to thereby reducing the ability of NE to interact with vascular allow for the assay of Slc22a3 expression in each fraction. smooth muscle to induce contraction and vasoconstriction. Relative expression of Slc22a3 was higher in adipocytes than This could be a physiologically relevant mechanism by which in the SVF (Fig. 7C). We used the reference gene ␤2-micro- PVAT modulates vascular tone. In the present study, we globulin because it gave us the most similar expression among discuss NE uptake in PVAT with the consideration that most all of the sample types compared with other housekeepers we NE that the blood vessel is exposed to is not circulating NE but assayed (data not shown). The cycles at which each sample rather NE released from sympathetic nerve boutons in and reached threshold for ␤2-microglobulin were similar between around the blood vessel. Autonomic nervous system interac- the heart and mesenteric PVAT (CT: 20.0 and 19.4, respec- tions with PVAT and its effects on the blood vessel have been tively) as well as the heart and aortic PVAT (CT: 17.8 and 17.5, recognized (12). NE released from the nerves that innervate respectively) but were dissimilar for the heart, adipocytes, and PVAT and the vascular smooth muscle cell could be taken up SVF (CT: 18.4, 16.7, and 19.9, respectively). This would affect by PVAT, thus reducing vascular contraction. In other words, our calculations for relative expression for the last group (CT PVAT might serve as a sink or source of NE. for Slc22a3: 26.9 in the heart, 23.0 in adipocytes, and 30.7 in PVAT can take up NE. An adrenergic system exists in the SVF). Immunostaining revealed OCT3 protein in aortic adipose tissue, as evidenced by the discovery that mesenteric PVAT (Fig. 8A) and mesenteric PVAT adipocytes (Fig. 8B) adipose tissue adipocytes synthesize NE and serotonin (66, 69). using the aorta as a positive control for OCT3 (70). Aortic We previously demonstrated that PVATs possess measurable

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE H1909

150 ASP+ (2 µM)

125

10 100 10 6 * 11 ** 75 *** 6 7 6

uptake (of vehicle) 6 ** + *** ** 50 # ϩ *** Fig. 5. ASP uptake is reduced by transporter inhibitors. Mesenteric PVAT was in- % ASP % cubated with inhibitors of transport or vehi- 25 cle for 10 min followed by the addition of ASPϩ (2 ␮M). Data are reported as percent uptake from vehicle. Numbers above the bars 0 indicate numbers of animals (N) used in each experiment. Values are means Ϯ SE. *P Ͻ 0.05, **P Ͻ 0.005, and ***P Ͻ 0.001 vs. vehicle; #P Ͻ 0.05 vs. corticosterone. Rep- 1 mM NE resentative images from each experiment are located beneath the corresponding bar.

10 µM nisoxetine 100 nM nisoxetine 100 nM citalopram 100 nM citalopram10 µM desipramine10 µM desipramine Vehicle (ASP+ only) 100 µM corticosterone + 100 µM corticosterone + 100 µM corticosterone

No ASP+ (background) Representative images correspond to the figure above.

catecholamines (5). The source of these catecholamines may [100 ␮M, OCT3 IC50: 120–290 nM (39)]. Desipramine, a NET be from the adipocyte fraction (Fig. 1A). Other sources of NE inhibitor, at higher concentrations can also inhibit SERT [ki: could be found in the SVF of PVAT, such as macrophages 129.00 nM (54)] and OCT3 [IC50 of 700 nM (75)]. Since our (11), lymphocytes (58), and neurons. The catecholamines pres- experimental samples were intact tissues, not isolated trans- ent in PVAT are releasable by tyramine and support contrac- porters or membranes, each inhibitor was used at a concentration in the rat superior mesenteric artery independent of sym- tion that was ϳ50–100 times above the ki values to assure the pathetic innervation (5). PVAT-dependent contraction to ty- inhibitors reached their target, considering they can be metab- ramine was reduced by the NET inhibitor nisoxetine (5), olized or bound. directing us further into the investigation of NE transport in Uptake of NE was reduced upon the addition of nisoxetine, PVAT. In the present study, we used the transporter inhibitors citalopram with corticosterone, and desipramine with cortico- desipramine [10 ␮M, inhibitory constant (ki) for NET: 7.36 nM sterone. While the high concentration of nisoxetine reduced (55) and for SERT: 228 nM (16)], nisoxetine [1 ␮M, ki for uptake on its own, there are two possibilities for why inhibition NET: 0.46 nM and for SERT: 158 nM (16)], and corticosterone with two drugs (desipramine/corticosterone and citalopram/

Rat 1 Rat 2 Rat 3 Rat 4

NET

-actin

Fig. 6. NE transporter (NET) is not located in mesenteric PVAT (mPVAT) adipocytes. Whole PVAT was assayed for NET by Western blot along with the mesenteric resistance vessels (MRV), PVAT adipocytes (adip), and PVAT stromal vascular fraction (SVF). The arrow points to the band of interest. ␤-Actin was used as the loading control, and the vena cava was used as a positive control. N ϭ 4.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org H1910 NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE A B C 8.0 N=4 ** 16 N=6-7 6.0 N=4-5 ** 14 ** 5.0 6.0 12 4.0 10 3.0 4.0 8 6 2.0 normalized by B2m normalized by B2m normalized by B2m Relative expression Relative expression

2.0 4 Relative expression 1.0

2 0.0 0.0 0 heart SVF

heart adipocytes heart aortic Mesenteric PVAT mes PVAT PVAT Fig. 7. Slc22a3 mRNA is expressed in mesenteric and aortic PVAT. A: relative mesenteric (mes) PVAT expression of Slc22a3 mRNA was measured and compared with the heart as a positive control. B: whole aortic PVAT relative expression of Slc22a3 was measured and compared with the heart as a positive control. C: PVAT adipocyte relative expression of Slc22a3 mRNA was higher than that in the SVF (**P Ͻ 0.005) compared with the heart as a positive control. Measures were normalized to B2M. Values are means Ϯ SE for the number of animals (N) shown. **P Ͻ 0.005. corticosterone), as opposed to either of them alone, was nec- OCT3. Homodimerization of rat OCT1 and human OCT2, the essary to observe a reduction in uptake. First, redundancy of other OCT isoforms, is required for the transporter to be placed NE transport may exist through different transporters in PVAT. on the plasma membrane (38). While oligomerization would be In the brain, the uptake-2 system (another term for OCT3- interesting to study, it is not the focus of this work. It is a mediated transport) has been implicated in limiting the reduc- possibility that different PVAT depots may contain a different tion of NE, dopamine (DA), and serotonin uptake by specific distribution of transporters of NE. Therefore, the applicability inhibitors (30). Therefore, in our system, to significantly re- of these findings to other PVAT depots is not known outside of duce NE uptake in PVAT, multiple transporters may have to be rat aortic and mesenteric PVATs. targeted. Second, it is possible that transporters on adipocytes The cation transporter substrate ASPϩ is taken up by PVAT heteroligomerize (18, 32, 37). NET and SERT can heteroli- adipocytes in a NET-, SERT-, and OCT3-dependent manner. gomerize, but it is debated whether heteroligomerization af- ASPϩ is a useful experimental tool for probing NE transport, fects their function. Less is known about oligomerization of as previously validated in NE uptake assays using radiolabeled

Aorta with PVAT Mesenteric PVAT adipocytes A B E Aortic PVAT P 150

V Ld N=11

Positive 100 M L * C D 50 NE uptake (% vehicle) 0 e on vehicle Negative ster o ortic c

100 µM Fig. 8. Immunohistochemical and immunocytochemical staining revealing the presence of organic cation transporter 3 (OCT3) in aortic PVAT and mesenteric PVAT adipocytes and that aortic PVAT NE uptake is reduced by inhibition of OCT3. A: the aorta was used as the positive control for OCT3. Aortic staining for OCT3 protein was evident in the media (M) but not in the adventitia (V). PVAT (P) around the aorta also stained for OCT3. L, lumen. B: mesenteric PVAT adipocytes were isolated and stained for OCT3 protein. The black arrow points to the location of adipocyte nuclei, and the red arrow points to the location of the cytoplasm of the adipocyte. Ld, lipid droplet. C: the aorta with PVAT was stained without the inclusion of the primary antibody to serve as a negative control. D: mesenteric PVAT adipocytes were stained without the inclusion of the primary antibody as a negative control. Adipocyte images are representative of four animals. Samples were imaged with a ϫ40 objective. Scale bar ϭ 50 ␮m. E: NE uptake of aortic PVAT was assayed after incubation with vehicle or corticosterone (100 ␮M). Data are reported as percentages from vehicle, and measures are expressed as NE concentration to protein content. Values are means Ϯ SE for the number of animals (N) shown. *P Ͻ 0.05.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE H1911 3[H]NE (28). ASPϩ fluorescence detection permits its use for be taken when interpreting findings from ASPϩ-binding stud- live cell imaging to identify potential transporters of NE. ies to mechanisms of specific NE transport. Both ASPϩ and Confocal imaging of ASPϩ-stained PVAT adipocytes allowed NE uptake experiments pointed to OCT3 as being important, us to visualize ASPϩ taken into the adipocyte via the obser- and, thus, the was the focus in our final experiments. vation of bright halos around the adipocyte nuclei. This pattern OCT3 is highly expressed in mesenteric and aortic PVATs. of perinuclear ASPϩ fluorescence was observed in all six of the OCT3 is a low-affinity, high-capacity uptake transporter, for- tissues imaged. The punctate pattern of ASPϩ fluorescence in mally known as extraneuronal monoamine transporter and also the adipocyte is strikingly similar to that observed when termed as the uptake-2 system, and is broadly expressed in adipocyte mitochondria were stained using rhodamine 123 non-neuronal cells (51). OCT3 expression on PVAT adi- (19). This previous study reported nuclear “haloing” when pocytes has not yet been investigated. We were surprised to mitochondria were stained using the fluorescent dye similar to find a higher expression of OCT3 in mesenteric and aortic what we observed when we applied ASPϩ to PVAT. Further- PVATs than in the heart. It is interesting to speculate on the more, ASPϩ accumulation in mitochondria after transport into function NE transport molecules on adipocytes when NE the cell has been shown previously (8), additionally supporting activates lipolysis. Adipocytes contain levels of monoamine our observation that ASPϩ was able to bind and be transported oxidase activity that are comparable to that of the liver, an into the cell. The finding that ASPϩ was citalopram sensitive is organ high in monoamine metabolizing activity (57). This in line with the finding that adipocytes express functional would support the idea that the function of NE transporters is SERT (66). Moreover, a study by Pizzinat et al. (57) found that to deliver NE into the cell to be metabolized. [3H]NE uptake in isolated human adipocytes obtained from We observed high expression of OCT3 in aortic PVAT and abdominal or mammary lipectomies could be reduced by inhibition of OCT3 reduced NE uptake. The presence of OCT3 inhibiting the uptake-2 system (OCT3) with disprocynium 24. in the aorta, a conduit artery, would lead one to question the In our study, nisoxetine reduced ASPϩ uptake at both concen- role of OCT3 in blood pressure regulation. Ultimately, the trations tested, and it inhibited NE uptake in PVAT at 1 ␮M, main role of OCT3 in PVAT is unknown. Uptake of NE by a concentration that would be nonspecific for NET. Although aortic PVAT was reduced by inhibition of OCT3 with corti- convincing evidence that NET is present on mesenteric PVAT costerone. Thus, it could serve to remove excess NE (57), but adipocytes was not found (Fig. 6), it is possible that ASPϩ is there may be other physiological roles for OCT3 that may not a more sensitive tool for the detection of NET. We did not involve blood pressure regulation, such as polyamine transport investigate the presence of DA transporters (DAT) further due (60) and clearance of toxins (35). Therefore, OCT3 may be to the finding that GBR-12935 [100 nM, a DAT inhibitor, ki: more necessary in the aortic PVAT due to its other roles versus 3.7 nM (59)] did not reduce ASPϩ fluorescence, and mRNA mesenteric resistance PVAT. This would have to be investi- for DAT could not be detected by PCR in mesenteric PVAT in gated further. 40 cycles (data not shown). Therefore, it is unlikely that DAT Limitations. Although in vitro NE transport influences the plays a role in NE uptake in PVAT. contractility of blood vessels to NE (5, 64), there is no The “anticontractile” effect of PVAT to NE in the rat confirmation that this occurs in vivo. The mesentery and thoracic aorta is attenuated by desipramine plus deoxycortico- omentum are considered “visceral fat” (33, 52), and the adi- sterone (64), and this observation was the impetus to study pose tissue around mesenteric resistance arteries is most spe- PVAT NE transport. Since the anticontractile effect of PVAT cifically referred to as mesenteric PVAT. We showed an image due to NE transport at least in the rat aorta has already been of mesenteric PVAT in Fig. 1 to clarify this point. Mesenteric shown (64), we did not pursue these experiments in the present PVAT is a common PVAT depot that is studied for its study. Instead, we set out to find the mechanism by which relevance to alterations in vascular response and blood pressure PVAT can take up NE. We also used desipramine and corti- (5, 23, 43, 49, 71, 72), whereas PVAT around skeletal muscle costerone (similar to deoxycorticosterone in that it inhibits arteries is more associated with mechanisms of insulin resis- OCT3) in this study to investigate transport. Desipramine alone tance (7, 27, 47, 48, 74). Therefore, mesenteric PVAT was the did not have an effect on ASPϩ or NE uptake, but, when added most relevant PVAT for us to study. in conjunction with corticosterone, we observed a significant Aortic PVAT was included in the immunohistochemistry, reduction in both assays. We observed similar general patterns NE uptake, and RT-PCR analyses as this is the most discrete of inhibition in NE uptake and ASPϩ experiments. Interest- and widely studied PVAT depot and was therefore useful when ingly, in contrast to our NE uptake experiment, which required we compared our findings with those of other studies. None- both desipramine and corticosterone or citalopram and corti- theless, it should be noted that mesenteric PVAT mechanisms costerone to reduce uptake, ASPϩ fluorescence was reduced by of NE uptake are more physiologically relevant with regard to preincubation with corticosterone or citalopram alone. This blood pressure regulation. The lack of specific antibodies could be due to a difference in transporter affinity for ASPϩ against NET and SERT in which we have confidence for use in versus NE. ASPϩ has been used as a surrogate for [3H]NE, due adipose tissue has been an experimental limitation and why a to the similarity in their pharmacological profiles, but there are pharmacological approach using well-characterized transporter differences in their affinity to certain transporters (28). This inhibitors was used to characterize monoamine transporters. was also evidenced by the observation that adding a high The Western blot analysis for NET lacked strong reactivity to concentration of NE to saturate NET failed to abolish ASPϩ the positive control (Fig. 6), indicating that a lack of signal in staining, indicating the presence of ASPϩ fluorescence non- our samples could be attributed to the low affinity of the specific to NE transport mechanisms. Schwartz et al. (61) antibody and not that the transporter is not present. This was observed nonspecific fluorescence of ASPϩ in experiments the best NE antibody that we had available. Studies using using human embryonic kidney cells. Therefore, care needs to knockout animals and/or small interfering RNA toward differ-

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org H1912 NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE ent molecular transporters could also be helpful to elucidate study of adipose tissue. Experiments using ASPϩ may be monoamine transport mechanisms. However, this approach is extended to compare NE transport in disease models of obesity accompanied with upregulation of transporters to take up NE, and hypertension to investigate the regulation and dysfunction facilitated by the promiscuous nature of neurotransmitter trans- of NE transport. porters (17). This has been observed in the brains of NET The present study identified NE uptake into PVAT as part of knockout mice (63), which exhibit increased SERT and DAT a larger project defining the physiology and pathophysiology of expression. In addition, mice with reduced SERT expression an endogenous adrenergic system in PVAT. Adipose tissue overexpress OCT3 and exhibit increased serotonin clearance maintains high expression of the amine-metabolizing enzymes through OCT3 (6). These points have been discussed in a monoamine oxidase and semicarbazide-sensitive amine oxi- previous report as it pertains to SERT knockout rats (44). This dase (1). Therefore, it is likely that one way PVAT interacts upregulation of transporters in genetic models of deficient with NE is by breaking it down with amine oxidases. Relative transport would thereby make the interpretations of experimen- to NE storage, tyramine (a sympathomimetic drug)-induced tal results arising from these techniques difficult and is why we PVAT-dependent contraction of the rat thoracic aorta and rat did not use the OCT3 knockout rodent in this study. superior mesenteric artery was reduced by tetrabenazine (a The anticontractile effect of PVAT due to NE transport at vesicular monoamine transporter inhibitor). This finding, pub- least in the rat aorta was discovered by Soltis and Cassis in lished by our group, supports that local stores of NE in PVAT 1991 (64), and we present the first biochemical measures to could contribute to vascular contraction (5). Collectively, these investigate the mechanism by which PVAT can take up NE. findings will define an adrenergic system in PVAT that we can This study suggests that NE transport may be involved in the then investigate in obesity. anticontractile effect of PVAT. We fully recognize that other This study sheds light on the interaction between PVAT and mechanisms exist that reduce NE-induced contraction, such as the blood vessel within a local adrenergic system. The role of the release of adiponectin from PVAT adipocytes (26). The PVAT reducing vascular contraction in health could be, in part, contribution of these mechanisms in a physiological system due to NE uptake into PVAT, and this mechanism of NE may dictate the pathology observed in different vascular dis- removal may be dysfunctional in diseases of altered vascular orders. A decline in NE transport capacity of PVAT may tone. The scarcity of information on transporters of NE and NE exacerbate loss of adiponectin followed by adipocyte hyper- uptake in PVAT led us to study normal (nondisease model) rats trophy and dysfunction in disease. In addition, in situations of to test whether uptake is an important physiological mecha- dysfunctional adiponectin release, PVAT NE transport may nism in PVAT. Understanding the normal functional charac- become a more important mechanism of reducing vascular teristics of NE uptake in the nondiseased rodent allows us to tone. These are questions we will pursue in the future. know what to look for when investigating a disease model. In The method in which we euthanized the animal before tissue this study, we developed an assay (confocal microscopy of collection could have a potential effect on adrenergic system ASPϩ staining of adipose tissue) that we and/or other investi- activation; thus, these factors and how they would affect our gators could use to investigate PVAT mechanisms in an obese end points were considered. Hirota et al. (31) reported inhibi- and/or hypertensive model organism. The present study is a tion of NE (noradrenaline) and DA release after exposure of rat first step in this direction. brain striatal slices to barbiturates, including pentobarbital. Other groups have similarly reported an inhibition of NE ACKNOWLEDGMENTS release (34, 50, 62) or no effect (41, 56) by pentobarbital in The authors thank Cassandra LaMarche for technical assistance. their studies. Our laboratory has previously euthanized rats using isoflurane or pentobarbital, and we have not observed a GRANTS difference in tissue catecholamine content with either of them, This work was supported by National Institutes of Health Grants P01-HL- suggesting similar functions of uptake and release. In the 70687 and 5-T32-GM 92715-4. present study, most of the experiments allow comparisons within each animal by reporting percent NE uptake (opposed to DISCLOSURES absolute NE values) and every animal was euthanized the same No conflicts of interest, financial or otherwise, are declared by the author(s). way (with pentobarbital). Thus, we hope that small changes in baseline NE would not affect our results. AUTHOR CONTRIBUTIONS Great care was taken to clean our adipocyte fraction from any contaminating cells. A possibility is that there was some Author contributions: N.A.-L., W.F.J., J.N.W., and S.W.W. conception and design of research; N.A.-L. and J.M.T. performed experiments; N.A.-L. and contamination of nerve fibers and SVF cells in the adipocyte R.B. analyzed data; N.A.-L. interpreted results of experiments; N.A.-L. pre- isolates, and this is why it was especially revealing to observe pared figures; N.A.-L. drafted manuscript; N.A.-L., W.F.J., R.B., J.N.W., fluorescent labeling of PVAT adipocytes with ASPϩ. With J.M.T., and S.W.W. edited and revised manuscript; N.A.-L., W.F.J., R.B., confocal microscopy, it was possible to confirm that ASPϩ was J.N.W., J.M.T., and S.W.W. approved final version of manuscript. labeling adipocytes and could readily distinguish if blood vessels or other structures that were in the visual field were REFERENCES ϩ exhibiting fluorescence with ASP . 1. Abella A, Garcia-Vicente S, Viguerie N, Ros-Baro A, Camps M, Conclusions, novelty, and significance. The present study Palacin M, Zorzano A, Marti L. Adipocytes release a soluble form of identified NE transport in PVAT and found OCT3 to be the VAP-1/SSAO by a metalloprotease-dependent process and in a regulated manner. Diabetologia 47: 429–438, 2004. prime candidate transporter of NE within the PVAT adipocyte. 3. Aghamohammadzadeh R, Heagerty AM. Obesity-related hypertension: This is the first report of NE transport in PVAT adipocytes. epidemiology, pathophysiology, treatments, and the contribution of ϩ This is also the first time that ASP has been applied to the perivascular adipose tissue. Ann Med 44, Suppl 1: S74–S84, 2012.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE H1913

5. Ayala-Lopez N, Martini M, Jackson WF, Darios E, Burnett R, Seitz B, tile properties of perivascular fat in obese patients. Circulation 119: Fink GD, Watts SW. Perivascular adipose tissue contains functional 1661–1670, 2009. catecholamines. Pharmacol Res Perspect 2: e00041, 2014. 27. Hariya N, Mochizuki K, Inoue S, Morioka K, Shimada M, Okuda T, 6. Baganz NL, Horton RE, Calderon AS, Owens WA, Munn JL, Watts Goda T. Insulin resistance in SHR/NDmc-cp rats correlates with enlarged LT, Koldzic-Zivanovic N, Jeske NA, Koek W, Toney GM, Daws LC. perivascular adipocytes and endothelial cell dysfunction in skeletal mus- Organic cation transporter 3: Keeping the brake on extracellular serotonin cle. J Nutr Sci Vitaminol (Tokyo) 60: 52–59, 2014. in serotonin-transporter-deficient mice. Proc Natl Acad Sci USA 105: 28. Haunso A, Buchanan D. Pharmacological characterization of a fluores- 18976–18981, 2008. cent uptake assay for the noradrenaline transporter. J Biomol Screen 12: 7. Bakker W, Eringa EC, Sipkema P, van Hinsbergh VW. Endothelial 378–384, 2007. dysfunction and diabetes: roles of hyperglycemia, impaired insulin sig- 29. Henry SL, Barzel B, Wood-Bradley RJ, Burke SL, Head GA, Armit- naling and obesity. Cell Tissue Res 335: 165–189, 2009. age JA. Developmental origins of obesity-related hypertension. Clin Exp 8. Blakely RD, Defelice LJ, Galli A. Biogenic amine neurotransmitter Pharmacol Physiol 39: 799–806, 2012. transporters: just when you thought you knew them. Physiology 30. Hensler JG, Artigas F, Bortolozzi A, Daws LC, De Deurwaerdere P, (Bethesda) 20: 225–231, 2005. Milan L, Navailles S, Koek W. Catecholamine/serotonin interactions: 9. Brandes RP. The fatter the better? Perivascular adipose tissue attenuates systems thinking for brain function and disease. Adv Pharmacol 68: vascular contraction through different mechanisms. Br J Pharmacol 151: 167–197, 2013. 303–304, 2007. 31. Hirota K, Kudo M, Kudo T, Kitayama M, Kushikata T, Lambert DG, 10. Brown NK, Zhou Z, Zhang J, Zeng R, Wu J, Eitzman DT, Chen YE, Matsuki A. Barbiturates inhibit Kϩ-evoked noradrenaline and dopamine Chang L. Perivascular adipose tissue in vascular function and disease: a release from rat striatal slices–involvement of voltage sensitive Ca2ϩ review of current research and animal models. Arterioscler Thromb Vasc channels. Neurosci Lett 291: 175–178, 2000. Biol 34: 1621–1630, 2014. 32. Horschitz S, Hummerich R, Schloss P. Functional coupling of serotonin 11. Brown SW, Meyers RT, Brennan KM, Rumble JM, Narasimhachari and noradrenaline transporters. J Neurochem 86: 958–965, 2003. N, Perozzi EF, Ryan JJ, Stewart JK, Fischer-Stenger K. Cat- 33. Ibrahim MM. Subcutaneous and visceral adipose tissue: structural and echolamines in a macrophage cell line. J Neuroimmunol 135: 47–55, 2003. functional differences. Obes Rev 11: 11–18, 2010. 12. Bulloch JM, Daly CJ. Autonomic nerves and perivascular fat: interactive 34. Ida Y, Tsuda A, Tsujimaru S, Satoh M, Tanaka M. Pentobarbital mechanisms. Pharmacol Ther 143: 61–73, 2014. attenuates stress-induced increases in noradrenaline release in specific 13. Centers for Disease Control and Prevention. Vital signs: prevalence, brain regions of rats. Pharmacol Biochem Behav 36: 953–956, 1990. treatment, and control of hypertension–United States, 1999–2002 and 35. Jonker JW, Schinkel AH. Pharmacological and physiological functions 2005–2008. MMWR Morb Mortal Wkly Rep 60: 103–108, 2011. of the polyspecific organic cation transporters: OCT1, 2, and 3 (SLC22A1- 14. Chaldakov GN, Tonchev AB, Stankulov IS, Ghenev PI, Fiore M, Aloe 3). J Pharmacol Exp Ther 308: 2–9, 2004. L, Rancic G, Panayotov P, Kostov DD. Periadventitial adipose tissue 36. Kang YS. Obesity associated hypertension: new insights into mechanism. (tunica adiposa): enemy or friend around? Arch Pathol Lab Med 131: Electrolyte Blood Press 11: 46–52, 2013. 1766; author reply 1766–1767, 2007. 37. Kocabas AM, Rudnick G, Kilic F. Functional consequences of homo- 15. Chen L, Pawlikowski B, Schlessinger A, More SS, Stryke D, Johns SJ, but not hetero-oligomerization between transporters for the biogenic Portman MA, Chen E, Ferrin TE, Sali A, Giacomini KM. Role of amine neurotransmitters. J Neurochem 85: 1513–1520, 2003. organic cation transporter 3 (SLC22A3) and its missense variants in the 38. Koepsell H. Polyspecific organic cation transporters and their biomedical pharmacologic action of metformin. Pharmacogenet Genomics 20: 687– relevance in kidney. Curr Opin Nephrol Hypertens 22: 533–538, 2013. 699, 2010. 39. Koepsell H, Lips K, Volk C. Polyspecific organic cation transporters: 16. Davids E, Zhang K, Kula NS, Tarazi FI, Baldessarini RJ. Effects of structure, function, physiological roles, and biopharmaceutical implica- norepinephrine and serotonin transporter inhibitors on hyperactivity in- tions. Pharm Res 24: 1227–1251, 2007. duced by neonatal 6-hydroxydopamine lesioning in rats. J Pharmacol Exp Kotchen TA. Ther 301: 1097–1102, 2002. 40. Obesity-related hypertension: epidemiology, pathophysiol- 17. Daws LC. Unfaithful neurotransmitter transporters: focus on serotonin ogy, and clinical management. Am J Hypertens 23: 1170–1178, 2010. uptake and implications for antidepressant efficacy. Pharmacol Ther 121: 41. Kubota T, Anzawa N, Hirota K, Yoshida H, Kushikata T, Matsuki A. 89–99, 2009. Effects of ketamine and pentobarbital on noradrenaline release from the 18. De Felice LJ, Adams SV. Serotonin and norepinephrine transporters: medial prefrontal cortex in rats. Can J Anaesth 46: 388–392, 1999. possible relationship between oligomeric structure and channel modes of 42. Lawes CM, Vander Hoorn S, Rodgers A; International Society of conduction. Mol Membr Biol 18: 45–51, 2001. Hypertension. Global burden of blood-pressure-related disease, 2001. 19. DeMartinis FD, Ashkin KT, Lampe KT. Fluorescence detection of Lancet 371: 1513–1518, 2008. mitochondrial clusters in mammalian white fat cells in vivo. Am J Physiol 43. Li R, Andersen I, Aleke J, Golubinskaya V, Gustafsson H, Nilsson H. Cell Physiol 253: C783–C791, 1987. Reduced anti-contractile effect of perivascular adipose tissue on mesen- 20. Dubrovska G, Verlohren S, Luft FC, Gollasch M. Mechanisms of teric small arteries from spontaneously hypertensive rats: role of Kv7 ADRF release from rat aortic adventitial adipose tissue. Am J Physiol channels. Eur J Pharmacol 698: 310–315, 2013. Heart Circ Physiol 286: H1107–H1113, 2004. 44. Linder AE, Diaz J, Ni W, Szasz T, Burnett R, Watts SW. Vascular 21. Fernandez-Alfonso MS, Gil-Ortega M, Garcia-Prieto CF, Aranguez I, reactivity, 5-HT uptake, and blood pressure in the serotonin transporter Ruiz-Gayo M, Somoza B. Mechanisms of perivascular adipose tissue knockout rat. Am J Physiol Heart Circ Physiol 294: H1745–H1752, 2008. dysfunction in obesity. Int J Endocrinol 2013: 402053, 2013. 45. Livak KJ, Schmittgen TD. Analysis of relative gene expression data Ϫ⌬⌬ 22. Fesus G, Dubrovska G, Gorzelniak K, Kluge R, Huang Y, Luft FC, using real-time quantitative PCR and the 2 CT method. Methods 25: Gollasch M. Adiponectin is a novel humoral vasodilator. Cardiovasc Res 402–408, 2001. 75: 719–727, 2007. 46. Lohn M, Dubrovska G, Lauterbach B, Luft FC, Gollasch M, Sharma 23. Galvez B, de Castro J, Herold D, Dubrovska G, Arribas S, Gonzalez AM. Periadventitial fat releases a vascular relaxing factor. FASEB J 16: MC, Aranguez I, Luft FC, Ramos MP, Gollasch M, Fernandez 1057–1063, 2002. Alfonso MS. Perivascular adipose tissue and mesenteric vascular function 47. Meijer RI, Bakker W, Alta CL, Sipkema P, Yudkin JS, Viollet B, in spontaneously hypertensive rats. Arterioscler Thromb Vasc Biol 26: Richter EA, Smulders YM, van Hinsbergh VW, Serne EH, Eringa 1297–1302, 2006. EC. Perivascular adipose tissue control of insulin-induced vasoreactivity 24. Gao YJ, Lu C, Su LY, Sharma AM, Lee RM. Modulation of vascular in muscle is impaired in db/db mice. Diabetes 62: 590–598, 2013. function by perivascular adipose tissue: the role of endothelium and 48. Meijer RI, Serne EH, Korkmaz HI, van der Peet DL, de Boer MP, hydrogen peroxide. Br J Pharmacol 151: 323–331, 2007. Niessen HW, van Hinsbergh VW, Yudkin JS, Smulders YM, Eringa 25. Gao YJ, Takemori K, Su LY, An WS, Lu C, Sharma AM, Lee RM. EC. Insulin-induced changes in skeletal muscle microvascular perfusion Perivascular adipose tissue promotes vasoconstriction: the role of super- are dependent upon perivascular adipose tissue in women. Diabetologia oxide anion. Cardiovasc Res 71: 363–373, 2006. 58: 1907–1915, 2015. 26. Greenstein AS, Khavandi K, Withers SB, Sonoyama K, Clancy O, 49. Mendizabal Y, Llorens S, Nava E. Vasoactive effects of prostaglandins Jeziorska M, Laing I, Yates AP, Pemberton PW, Malik RA, Heagerty from the perivascular fat of mesenteric resistance arteries in WKY and AM. Local inflammation and hypoxia abolish the protective anticontrac- SHROB rats. Life Sci 93: 1023–1032, 2013.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org H1914 NOREPINEPHRINE TRANSPORT IN PERIVASCULAR ADIPOSE TISSUE

50. Mizuno T, Ito E, Kimura F. Pentobarbital sodium inhibits the release of regulates dopamine and serotonin transporters in the mouse brain. Neuro- noradrenaline in the medial preoptic area in the rat. Neurosci Lett 170: chem Int 59: 185–191, 2011. 111–113, 1994. 64. Soltis EE, Cassis LA. Influence of perivascular adipose tissue on rat aortic 51. Mooney JJ, Samson JA, Hennen J, Pappalardo K, McHale N, Alpert smooth muscle responsiveness. Clin Exp Hypertens A 13: 277–296, 1991. J, Koutsos M, Schildkraut JJ. Enhanced norepinephrine output during 65. Stern N, Marcus Y. Perivascular fat: innocent bystander or active player long-term desipramine treatment: a possible role for the extraneuronal in vascular disease? J Cardiometab Syndr 1: 115–120, 2006. monoamine transporter (SLC22A3). J Psychiatr Res 42: 605–611, 2008. 66. Stunes AK, Reseland JE, Hauso O, Kidd M, Tommeras K, Waldum 52. Nguyen NL, Randall J, Banfield BW, Bartness TJ. Central sympathetic HL, Syversen U, Gustafsson BI. Adipocytes express a functional system innervations to visceral and subcutaneous white adipose tissue. Am J for serotonin synthesis, reuptake and receptor activation. Diabetes Obes Physiol Regul Integr Comp Physiol 306: R375–R386, 2014. Metab 13: 551–558, 2011. 53. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood 66a.Stuurman N, Edelstein AD, Amodaj N, Hoover KH, and Vale RD. and adult obesity in the United States, 2011–2012. JAMA 311: 806–814, Computer Control of Microscopes Using ␮Manager (online). http:// 2014. www.ncbi.nlm.nih.gov/pmc/articles/PMC3065365/ [15 October 2015]. 54. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB. Neurotransmitter 67. Szasz T, Bomfim GF, Webb RC. The influence of perivascular adipose receptor and transporter binding profile of antidepressants and their me- tissue on vascular homeostasis. Vasc Health Risk Manag 9: 105–116, tabolites. J Pharmacol Exp Ther 283: 1305–1322, 1997. 2013. 55. Paczkowski FA, Bryan-Lluka LJ, Porzgen P, Bruss M, Bonisch H. 68. Szasz T, Webb RC. Perivascular adipose tissue: more than just structural Comparison of the pharmacological properties of cloned rat, human, and support. Clin Sci (Lond) 122: 1–12, 2012. bovine norepinephrine transporters. J Pharmacol Exp Ther 290: 761–767, 69. Vargovic P, Ukropec J, Laukova M, Cleary S, Manz B, Pacak K, 1999. Kvetnansky R. Adipocytes as a new source of catecholamine production. 56. Pashkov VN, Hemmings HC Jr. The effects of general anesthetics on FEBS Lett 585: 2279–2284, 2011. norepinephrine release from isolated rat cortical nerve terminals. Anesth 70. Verhaagh S, Schweifer N, Barlow DP, Zwart R. Cloning of the mouse Analg 95: 1274–1281, 2002. and human solute carrier 22a3 (Slc22a3/SLC22A3) identifies a conserved 57. Pizzinat N, Marti L, Remaury A, Leger F, Langin D, Lafontan M, cluster of three organic cation transporters on mouse chromosome 17 and Carpene C, Parini A. High expression of monoamine oxidases in human human 6q26-q27. Genomics 55: 209–218, 1999. white adipose tissue: evidence for their involvement in noradrenaline clearance. Biochem Pharmacol 58: 1735–1742, 1999. 71. Verlohren S, Dubrovska G, Tsang SY, Essin K, Luft FC, Huang Y, 58. Qiu YH, Peng YP, Jiang JM, Wang JJ. Expression of tyrosine hydrox- Gollasch M. Visceral periadventitial adipose tissue regulates arterial tone ylase in lymphocytes and effect of endogenous catecholamines on lym- of mesenteric arteries. Hypertension 44: 271–276, 2004. phocyte function. Neuroimmunomodulation 11: 75–83, 2004. 72. Watts SW, Dorrance AM, Penfold ME, Rourke JL, Sinal CJ, Seitz B, 59. Rothman RB, Baumann MH, Dersch CM, Romero DV, Rice KC, Sullivan TJ, Charvat TT, Thompson JM, Burnett R, Fink GD. Carroll FI, Partilla JS. Amphetamine-type central nervous system stim- Chemerin connects fat to arterial contraction. Arterioscler Thromb Vasc ulants release norepinephrine more potently than they release dopamine Biol 33: 1320–1328, 2013. and serotonin. Synapse 39: 32–41, 2001. 73. Wilson JN, Brown AS, Babinchak WM, Ridge CD, Walls JD. Fluo- 60. Sala-Rabanal M, Li DC, Dake GR, Kurata HT, Inyushin M, Skatch- rescent stilbazolium dyes as probes of the norepinephrine transporter: kov SN, Nichols CG. Polyamine transport by the polyspecific organic structural insights into substrate binding. Org Biomol Chem 10: 8710– cation transporters OCT1, OCT2, and OCT3. Mol Pharm 10: 1450–1458, 8719, 2012. 2013. 73a.World Health Organization. Obesity and Overweight (Fact Sheet 311) 61. Schwartz JW, Blakely RD, DeFelice LJ. Binding and transport in (online). http://www.who.int/mediacentre/factsheets/fs311/en/ [15 Octo- norepinephrine transporters. Real-time, spatially resolved analysis in sin- ber 2015]. gle cells using a fluorescent substrate. J Biol Chem 278: 9768–9777, 2003. 74. Yudkin JS, Eringa E, Stehouwer CD. “Vasocrine” signalling from 62. Shimokawa A, Jin QH, Ishizuka Y, Kunitake T, Takasaki M, Kannan perivascular fat: a mechanism linking insulin resistance to vascular dis- H. Effects of anesthetics on norepinephrine release in the hypothalamic ease. Lancet 365: 1817–1820, 2005. paraventricular nucleus region of awake rats. Neurosci Lett 244: 21–24, 75. Zhu HJ, Appel DI, Grundemann D, Richelson E, Markowitz JS. 1998. Evaluation of organic cation transporter 3 (SLC22A3) inhibition as a 63. Solich J, Faron-Gorecka A, Kusmider M, Palach P, Gaska M, Dziedz- potential mechanism of antidepressant action. Pharmacol Res 65: 491– icka-Wasylewska M. Norepinephrine transporter (NET) knock-out up- 496, 2012.

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00308.2015 • www.ajpheart.org