Norepinephrine hydrochloride, nisoxetine hydrochloride, semicarbazide hydrochloride, and catalase were purchased from Sigma-Aldrich (St. Louis, MO). Pargyline hydrochloride for the contractility experiments was purchased from Cayman Chemical (Ann Arbor, MI). Corticosterone was purchased from Tocris Bioscience (United Kingdom). Pargyline and clorgyline used in the oxidase assay experiments were supplied within the Amplex® Red Monoamine Oxidase Assay Kit (cat# A12214, ThermoFisher Scientific, Grand Island, NY USA).

Male Sprague-Dawley rats (225–275 g or ~8–10 weeks of age, Charles River, Indianapolis, IN USA) were used. All protocols were approved by the MSU Institutional Animal Care and Use Committee and follow the “Guide for the Care and Use of Laboratory Animals,” 8th edition, 2011. Rats were anesthetized with sodium pentobarbital (60–80 mg/kg, IP). Anesthesia was verified by lack of paw pinch and eye blink reflexes. Death was assured by pneumothorax and exsanguination after which tissues were removed for one of the following protocols.

The liver was collected, snap frozen in liquid N₂ and saved to serve as a positive control for the RT-PCR and Western blot assays. Mesentery, brain and aorta were collected in physiological salt solution (PSS); in mM; 130 NaCl; 4.7 KCl; 1.18 KH₂PO₄; 1.17 MgSO₄ 7H₂O; 14.8 NaHCO₃; 5.5 dextrose; 0.03 CaNa₂ ethylenediaminetetraacetic acid; 1.6 CaCl₂ (pH 7.2). A portion the rat brain corresponding to the midbrain and pons was dissected and the aorta was cleaned of PVAT. Both samples were then saved in potassium phosphate buffer (50 mM) for the oxidase activity assay. The same tissues (brain and aorta without PVAT) from a separate set of animals were dissected, frozen in liquid N₂ and saved for RT-PCR. MPVAT and the associated mesenteric resistance arteries and veins (MRV) were separated out in a Sylgard® coated petri dish in PSS with the aid of a stereomicroscope. The MPVAT was either used for fractionation, frozen in liquid N₂ for protein isolation, or saved in potassium phosphate buffer (50 mM) for the oxidase assay. The MRV were either frozen in liquid N₂ for protein or RNA isolation, or saved in potassium phosphate buffer (50 mM). Samples for the oxidase activity assays were saved at −20°C and used within 1 week. Mesenteric resistance arteries (2 mm in length) with or without PVAT for the use in isometric contraction experiments were dissected out in a Sylgard® coated petri dish in PSS with the aid of a stereomicroscope.

MPVAT was added to 1 mL of PSS with 1 mg/mL collagenase from Clostridium histolyticum type IA (cat# C9891, Sigma) and incubated at 37°C with gentle agitation until fully digested. The sample was centrifuged at 200 × g for 5 min after which the SVF was transferred into a separate tube. The fractions were then washed six times by adding 1 mL of PSS and centrifuging at 200 × g for 10 min. This protocol is one used to routinely isolate adipocytes and that has been verified by other groups to satisfactorily exclude other cell types (Vargovic et al., 2013). Purity of the isolation (>95% adipocytes) was verified by counting the adipocytes vs. non-adipocytes present with a hemocytometer. Phase contrast images of the fractions were taken with a 20 × objective (Hi PLAN I 20X/ 0.30 PH1) on an inverted microscope DMi1 (Leica, Buffalo Grove, IL, USA) using Leica Application Suite (LAS). The PSS was then removed and the samples were placed in 50 mM potassium phosphate buffer to be used in the oxidase assay before freezing or snap frozen in liquid N₂ for protein isolation.

All tissues (brain, liver, MRV, MPVAT, and aorta) were homogenized using the Bead Ruptor 24 (Omni International, NW Kennesaw, GA). RNA was extracted with the Quick RNA MiniPrep kit (cat# R1054, Zymo Research Corporation, Irving, CA USA) and purity (260/280 and 260/230 ratios ≥ 1.8) was verified using a Nanodrop 2000C spectrophotometer (Thermo Scientific, Wilmington, DE USA). The mRNA (1 μg) was reverse transcribed with the High-Capacity cDNA Reverse Transcription Kit (cat# 4368814, ThermoFisher Scientific). RT-PCR was performed using PerfeCTa FastMix II, ROX (cat# 95119, Quanta Biosciences, Gaithersburg, MD USA) on the ABI 7500 Fast Real Time PCR system (Life Technologies, Carlsbad, CA USA) with the following parameters: 95°C for 20 s, 95°C for 1 s and 60°C for 20 s for 40 cycles. Taqman Primers were purchased from ThermoFisher Scientific. The sequences are proprietary. Thus, we have listed the catalog numbers which are as follows: Aoc3 (cat# 4448892, assay ID: Rn01452826_m1), Comt (cat#4448892, assay ID: Rn01404927_g1: Actb (cat#4448892, assay ID: Rn00667869_m1), Maoa (cat#4448892, assay ID: Rn01430950_m1), Maob (cat#4448892, assay ID: Rn00566203_m1). Measures were normalized to β-actin (Actb) and expressed as fold change relative to the positive control tissue as described by Livak and Schmittgen (2001).

MPVAT and MRV protein was isolated in phosphate buffered saline (PBS) with protease inhibitors [sodium orthovanadate (1 mM), aprotinin/leupeptin (100 μg/ml) and phenylmethylsulfonyl fluoride (1 mM)] and homogenized using the Bead Ruptor 24 (Omni International). The protein from the positive controls was isolated as follows. Stomach fundus and gut mucosa were dissected from the rat and were placed into RIPA buffer (cat# R3792, Teknova, Hollister, CA) with the above protease inhibitors before homogenizing with the Bead Ruptor 24 (Omni International) and centrifuging at 10,000 rpm for 10 min. Protein from the aorta was isolated in 1X lysis buffer [Tris HCl (62.5 mM) pH 7.8, 2% SDS, 10% glycerol] with the above protease inhibitors, frozen in liquid N₂, homogenized with mortar and pestle followed by a centrifugation at 11,000 rpm for 10 min. Supernatants were then separated. The Jurkat whole cell lysate was purchased from Santa Cruz Biotechnology (cat# SC-2204, Dallas, TX). The protein concentrations were determined using the Bicinchoninic Acid Protein Assay Kit (cat# BCA1, Sigma-Aldrich). The protein samples (50 μg) were separated on 10% SDS polyacrylamide gels using the Bio-Rad Mini Protean 3 system. Protein was transferred to PVDF-FL (MAO-A) or nitrocellulose (MAO-B, COMT, VAP-1) and blocked for 3 h at 4°C in 4% chicken egg ovalbumin (MAO-A, β-actin), LI-COR Odyssey Blocking Buffer (MAO-B, COMT) or 5% bovine serum albumin (VAP-1). Primary antibody [MAO-A, 1:200 (epitope corresponds to amino acids 458–527 of MAO-A of human origin; Santa Cruz Biotechnology, cat#SC-20156); MAO-B, 1:200 (epitope near the C-terminus of MAO-B of human origin; Santa Cruz Biotechnology, cat#SC-18401); COMT, 1:200 (epitope corresponding to amino acids 1-271 of COMT of human origin; Santa Cruz Biotechnology, cat#SC-25844); VAP1, 1:200 (epitope near the C-terminus of VAP-1 of human origin; Santa Cruz Biotechnology, cat#SC-13741); or β-actin, 1:2000 (Sigma, cat#A3854)] was incubated overnight in blocking buffer at 4°C. Antibody was recovered and blots were washed with TBS + 0.1% Tween 20 (TBS-T) for 10 min at 4°C (three times). Blots were then incubated with species-specific LI-COR IRDye 800 secondary antibody (1:1000, MAO-A, MAO-B, COMT, VAP1) or LI-COR IRDye 700 secondary antibody (1:1000, β-actin) in LI-COR Odyssey Blocking Buffer for 1 h at 4°C, followed by washes with TBS-T for 10 min at 4°C (three times). Experiments using a competing peptide (MAO-B: cat#SC-18401 P, VAP-1: SC-13741 P, Santa Cruz Biotechnology) were performed as above except for that before the primary antibody was added to the blot it was incubated overnight at 4°C with the competing peptide at 5x the concentration of the antibody. Bands were visualized using the LI-COR Odyssey Classic or the LI-COR Odyssey CLx. Densitometry was completed with Image J.

Samples were homogenized with the Bead Ruptor 24 (Omni International). The samples were centrifuged at 600 × g for 10 min and the supernatant transferred to new tubes. The protein concentration of an aliquot of the sample was measured before the oxidase assay using a Bicinchoninic Acid Protein Assay Kit (cat# BCA1, Sigma-Aldrich) to inform us on the volume of the sample to load into the assay. MPVAT and the adipocyte fraction (AF) samples were loaded into a black clear-bottom 96-well plate at 20 mg of protein per well. Twice this protein, 40 mg, was loaded for the MRV, the SVF, brain (positive control for MAO) and aorta (positive control for SSAO). More protein was used for non-adipocyte containing tissues because the baseline H₂O₂ production in these extracts was less. Thus, more was loaded to ensure H₂O₂ would be detected. Samples in 1X reaction buffer were added into each well of a 96-well plate with an inhibitor of metabolism/or vehicle. The plate was incubated at room temperature for 30 min. Then, the substrate was added (either tyramine or benzylamine; 1 mM), with 100 μL of the Amplex® Red Reagent to start the reaction. The plate was incubated for another 30 min at room temperature after which the fluorescence was read at 530–560 nm excitation and 590 nm emission on the Infinite M1000 PRO (Tecan Group Ltd, Männedorf, Switzerland). Measures were compared to a resorufin standard curve and expressed as pmol/min/mg of protein.

Rat third-order mesenteric resistance arteries cleaned of fat (−PVAT) or with fat intact (+PVAT) were mounted into a Multi Wire Myograph System 620M (Danish Myo Technology, Denmark). Data were acquired using a PowerLab Data Acquisitions unit (ADInstruments, Colorado Springs, CO, USA). Baths contained warmed, oxygenated PSS. Rings were pulled to optimum resting tension (13.3 kpa) with the aid of the DMT Normalization Module for LabChart software and PowerLab (Danish Myo Technology and ADInstruments; http://cdn.adinstruments.com/adi-web/brochures/DMT-Normalization-2011.pdf) and equilibrated for 1 h with washes every 20 min. The arteries were exposed to an initial concentration of 60 mM KCl to test viability. Tissues were washed and tone returned to baseline and exposed to another concentration of 60 mM KCl to elicit the maximum contraction. The 60 mM KCl maximum contraction was what we normalized the force of contraction to. Tissues were then washed and returned to baseline. Either vehicle or inhibitor was added for 1 h without washing. NE was then added in a cumulative fashion, with significant time necessary for a response to plateau prior to the next addition. If no response was recorded within 3 min, the next concentration of NE was applied. Tissues were washed and a final 60 mM KCl addition was performed to test for tissue viability at the end of the experiment. Data were expressed as percent maximum contraction. The final contraction to 60 mM KCl was compared to the initial contraction (Data Supplementary Figures 4, 5). If the tissue did not contract to the final concentration of 60 mM KCl, the data for that experiment was not used.

Data are reported as means ± SEM for number of animals indicated by N or near each bar within the graphs. Statistical analyses were performed with GraphPad Prism 6.0 (GraphPad Software, Inc., La Jolla, CA). Outliers were identified and removed following a Grubb's test. Contraction was reported as means ± SEM as a percentage of the initial contraction to 60 mM KCl. Potency means (−logEC₅₀, M) were calculated using GraphPad Prism 6.0. Where a maximum was not achieved, the values are estimated and true potencies equal or greater than that reported. P < 0.05 was considered statistically significant.

MPVAT and the underlying artery-vein pair (Data Supplementary Figure 1) were dissected from male Sprague-Dawley rats and analyzed for the expression of amine metabolizing enzyme genes. Relative Maoa expression in the MPVAT was similar to that in the brain (Figure 1A), the positive control for Maoa and Maob (Jahng et al., 1997). However, MRV expression of Maoa was significantly lower than in the brain but not the MPVAT. Maob expression was significantly lower in both the MRV and in the MPVAT vs. the brain positive control (Figure 1B). When compared to the liver, a positive control for Comt expression (Karhunen et al., 1994), MRV and MPVAT had low levels of Comt expression (less than 0.01-relative expression from liver for both; data not shown). By contrast, Aoc3 (amine oxidase, copper containing 3; the gene for SSAO) was expressed at higher levels in the MPVAT (Figure 1C) vs. the MRV when compared relative to the aorta, a positive control for Aoc3 expression (Wanecq et al., 2006).

Western blot analysis of protein isolated from the MRV and associated MPVAT, revealed presence of MAO-A, MAO-B and SSAO (Figure 2A). Consistent with there being no difference between the mRNA expression for Maoa in the MRV and MPVAT, there was no statistical difference in relative protein signal between the MRV and MPVAT (p = 0.29; Figures 2A,B). MAO-B signal was lower the MPVAT than in the MRV (Figures 2A,C). Bands for COMT were not detected in either the MRV or the MPVAT (Figure 2A; densitometry not shown). SSAO signal was higher in the MPVAT vs. the MRV (Figures 2A,D), also consistent with Figure 1C.

Oxidation of amines in the MRV and in the MPVAT could alter vascular tone through the removal of vasoactive amines and through the release of the reaction product H₂O₂. To quantify the oxidase activity in these tissues, the MRV and MPVAT were carefully separated. MPVAT was split into its constituent fractions, the AF and the SVF. A microscopic image of the separated fractions is shown in Data Supplementary Figure 2, which represents the AF and SVF used in this study. Tyramine, a substrate for MAO-A, MAO-B and SSAO, was added to the sample extracts to drive oxidase activity. Addition of tyramine increased the oxidase activity (H₂O₂ produced) in the MRV, MPVAT, the AF, and the SVF vs. vehicle (H₂O) (Figure 3). The brain contains both MAO-A and B activity but low SSAO activity (Kalaria et al., 1988; Castillo et al., 1999). Thus, the brain served as a positive control for the MAOs. The brain had higher oxidase activity in response to tyramine vs. vehicle. The aorta, the positive control for SSAO, also had higher activity to tyramine vs. vehicle.

To determine which amine oxidase(s) contributed to the tyramine-driven H₂O₂ production in the mesentery, the MRV, MPVAT, AF and the SVF were incubated with inhibitors to each of the amine oxidases (MAO-A, MAO-B, and SSAO) before adding tyramine. Clorgyline, an irreversible inhibitor of MAO-A, was used due to its specificity to MAO-A (IC₅₀ of 0.03 μM and 8 μM at MAO-B (Ozaita et al., 1997) and its lack of inhibition of SSAO (Clarke et al., 1982). Pargyline (1 μM) was used to specifically inhibit MAO-B (K_i 0.5 μM) and a higher concentration (10 μM) was used to inhibit both MAO-A and B (K_i for pargyline at MAO-A = 15 μM Fowler et al., 1982). Semicarbazide was used to specifically inhibit SSAO, at a concentration (1 mM) routinely used in monoamine assays to selectively inhibit SSAO activity (K_i 15 μM Lizcano et al., 1996; Repessé et al., 2015). Oxidase activity was quantified as the amount of H₂O₂ produced per minute normalized to the amount of protein.

Tyramine-driven oxidase activity in the MRV was not reduced by the addition of clorgyline or pargyline (Figure 4A). However, there was a non-statistically significant reduction with 1 mM semicarbazide (Figure 4A). On the other hand, semicarbazide significantly reduced (almost by 100%) the tyramine-driven oxidase activity in the MPVAT and the AF (Figures 4B,C).

Oxidase activity in the SVF was not significantly reduced by inhibition of the MAOs or SSAO vs. vehicle (Figure 4D). However, each inhibitor (clorgyline, pargyline, and semicarbazide) caused large non-significant reductions in oxidase activity. No one enzyme had activity that was higher over the other enzymes. Thus, each amine oxidase (MAO-A, MAO-B, and SSAO) may contribute to oxidase activity in the SVF. The brain had reduced oxidase activity to clorgyline (1 μM- specific for MAO-A), pargyline (1 μM- specific for MAO and 10 μM- inhibits both MAO-A and MAO-B), but no reduction with the SSAO inhibitor semicarbazide (Figure 4E). Oxidase activity in the aorta was abolished by semicarbazide (1 mM); by contrast, there was not reduction in oxidase activity by clorgyline (1 μM) or pargyline (1 μM) (Figure 4F). The results for these controls support that the assay can detect and distinguish between SSAO and MAO activity.

Benzylamine differs from tyramine in that it is a substrate for MAO-B and SSAO only (and not MAO-A). Addition of benzylamine increased oxidase activity vs. vehicle in the MPVAT, AF, SVF, brain, and aorta (Figure 5). Amine oxidase activity in the MRV was abolished by the SSAO inhibitor, semicarbazide, and was reduced in the MPVAT, AF, and the SVF (Figures 6A–D). This is different from what was observed with tyramine as the substrate, where the reduction with semicarbazide was not significant. Tyramine is a poor substrate for SSAO compared to benzylamine with a K_m of 17.6 mM. Whereas, the K_m of benzylamine for SSAO is 161 μM (Precious and Lyles, 1988). Some of the tyramine-driven oxidase activity in the MRV could have been due to other amine oxidases other than SSAO. Oxidase activity in the brain was completely abolished by inhibition of MAO-B with pargyline, but was not affected by inhibition of SSAO (Figure 6E). Oxidase activity in the aorta was not affected by pargyline, but was reduced significantly by semicarbazide (Figure 6F).

Third-order mesenteric resistance arteries with (+) or without (−) PVAT were mounted on a 4-channel wire myograph. The force of contraction to 60 mM KCl of the arteries with and without PVAT were not different (Data Supplementary Figure 3). The arteries were incubated with either vehicle or an inhibitor prior to the generation of a NE concentration-response curve. The presence of PVAT gave a significant right-shift of the NE-response curve (Figure 7). NE had a −logEC₅₀ [M] of 5.64 ± 0.05 in mesenteric resistance—PVAT and a −logEC₅₀ [M] of 5.08 ± 0.06 in arteries +PVAT (Figure 7). Inhibition of MAO-A and B by 10 μM pargyline, or SSAO by semicarbazide (1 mM) did not shift the NE concentration-response curve of arteries + or −PVAT vs. vehicle (Figures 8A,B). However, a left-shift of the +PVAT curve was observed with incubation with semicarbazide + pargyline (10 μM; SP) suggesting that multiple NE metabolizers may contribute to PVAT's anti-contractile effect on mesenteric resistance arteries to NE (Figure 8C). To understand whether the shift in the NE-curve with SP was due to H₂O₂ being produced, we incubated +PVAT arteries with vehicle, SP, catalase (2000 U/ml), or SP + catalase (2000 U/ml) for 1 h prior to the addition of NE. The arteries that were incubated with SP and catalase had a left-shift in their response-curve to NE vs. those incubated with catalase only (Figure 8D). However, no shift was observed in the arteries that were exposed to SP vs. vehicle suggesting that inhibition of metabolism alone is not always sufficient to affect arterial response to NE and that both PVAT's metabolism of NE and production of H₂O₂ are involved in decreasing contraction to NE. The pharmacological parameters for Figures 8B–D are listed in Table 1.

Arteries incubated with the organic cation transporter 3 (OCT3) inhibitor corticosterone (100 μM), had NE-response curves (+ and −PVAT) shifted to the left from vehicle (Figure 9A). However, the −logEC₅₀ of the two curves were not statistically different (Table 1). Nisoxetine (1 μM), an inhibitor of the norepinephrine transporter (NET), shifted the NE-curve of +PVAT arteries significantly to the left vs. vehicle (Figure 9B; Table 1). We then tested whether transport of NE potentiated metabolism of NE in PVAT. Inhibition of SSAO [semicarbazide (1 mM)], MAO-A and B [pargyline (10 μM)], and OCT3 [corticosterone (100 μM); SPC] did not alter PVAT's anti-contractile effect when compared to vehicle (Figure 9C). The curve for the arteries +PVAT incubated with inhibitors was shifted slightly to the left from arteries +PVAT with vehicle but this shift was not significant (Table 1). It appears that inhibiting the entry of NE through OCT3 does not decrease NE removal. By contrast, inhibition of metabolism and uptake through NET by SPN [SPN = semicarbazide (1 mM), pargyline (10 μM), and nisoxetine (1 μM)] shifted the +PVAT curve to the left (Figure 9D) reflected by an increase in potency of NE on +PVAT arteries (Table 1).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.