Metabolism of N-Acylated-Dopamine

N-oleoyl-dopamine (OLDA) is a novel lipid derivative of dopamine. Its biological action includes the interaction with dopamine and the transient receptor potential vanilloid (TRPV1) receptors. It seems to be synthesized in a dopamine-like manner, but there has been no information on its degradation. The aim of the study was, therefore, to determine whether OLDA metabolism proceeds the way dopamine proper does. We addressed the issue by examining the occurrence of O-methylation of exogenously supplemented OLDA via catechol-O-methyltransferase (COMT) under in vitro, ex vivo, and in vivo conditions using rat brain tissue. The results show that OLDA was methylated by COMT in all conditions studied, yielding the O-methylated derivative. The methylation was reversed by tolcapone, a potent COMT inhibitor, in a dose-dependent manner. We conclude that OLDA enters the metabolic pathway of dopamine. Methylation of OLDA may enhance its bioactive properties, such as the ability to interact with TRPV1 receptors.


Introduction
The subject of the present study was N-oleoyl-dopamine (OLDA), N-[2-(3,4-dihydroxyphenyl) ethyl]-9Z-octadecenamide, a novel lipid derivative of dopamine [1][2][3]. The compound is synthesized by direct N-acylation of dopamine (DA) by oleic acid followed by a standard sequence of reactions leading to the formation of a DA moiety [4]. Although OLDA is present endogenously in mammalian brain [4], neither its metabolism nor the biological role has been well explored. Exogenously applied OLDA is taken up by both central and peripheral neural tissues [5,6]. OLDA seems to be more stable in vitro than dopamine proper is [7,8] and some of OLDA major bioactivities noted in the experimental studies, such as enhancement of locomotor activity [7] or lowering the reserpine-induced muscle rigidity [9], are mediated via the dopamine pathway. OLDA also is a ligand for the transient receptor potential vanilloid V1 (TRPV1) receptors [4,10,11]. These receptors seem to favor the compounds that have the 3-methoxy-4-hydroxybenzyl (homovanillyl) and the 3,4dihydroxybenzyl (catechol) groups [12,13], which they are affiliated with. The latter group is present in OLDA and in its methylated metabolite -3-methoxy-N-oleoyl-dopamine (O-Me-OLDA). Thus, OLDA has a preferable structure regarding the affiliation with TRPV1 receptors, and indeed is a proven TRPV1 agonist [14]. The question arises of whether O-methylation of OLDA could occur in the biological setting and whether it is catechol-O-methyltransferase (COMT)-dependent. We addressed this question by examining the ability of exogenously supplemented OLDA to undergo the methylation process in in vitro, ex vivo, and in vivo conditions. The study gave a positive answer to the question posed, which gives supportive evidence for a dopamine-like metabolic pathway of OLDA and raises the possibility that pre-methylation of OLDA facilitates its affinity to TRPV1 receptors.

Ethics statement
The study was approved by the IV Warsaw Local Ethics Committee for Animal Experiments (Permit Number: 81/2009) and it was performed in accordance with the Guiding Principles for the Care and Use of Laboratory Animals of the American Physiological Society [15].

Outline of protocol and animal tissue
This basically biochemical study consisted of the in vitro, ex vivo, and in vivo experiments. Firstly, we examined whether OLDA would undergo the O-methylation in vitro catalyzed by a commercially available COMT. Secondly, we determined whether O-methylation of OLDA would be possible in an ex vivo tissue in the presence of endogenous COMT, according to the method of Brannan et al. (16]. Finally, we sought to determine the presence of O-Me-OLDA in brain tissue after intraarterial injection of OLDA in vivo. The biological material used for the assays was the rat brain tissue obtained from a total of 17 adult male Wistar rats. The animals were surgically anesthetized with intraperitoneal injection of urethane and a-chloralose (700 and 120 mg/kg, respectively). The monitoring of the state of the pupil was used as an indication of the level of anesthesia. Supplemental doses of the anesthetics, 10% of the original dose, were given when a painful stimulus, such as the piercing of a hind paw, caused a rapid reaction of the pupil. At the end of the experimental procedures, the animals were sacrificed by perfusion through the left heart with icy cold 0.9% NaCl, the brains were rapidly removed, and the homogenates were prepared as below outlined. To prove that the methylation is COMT-dependent, we performed additional experiments with and without inhibition of COMT by tolcapone. All reagents used in the study were of analytical grade and were acquired from Sigma-Aldrich (St. Louis, MO), except for tolcapone that was obtained from KeyOrganics (Camelford, UK).

Syntheses of OLDA and O-Me-OLDA
Syntheses of OLDA and O-Me-OLDA were performed according to the method of Czarnocki et al. [2]. Briefly, 0.43 g (2.27 mmol) of dopamine hydrochloride, 0.64 g (2.27 mmol) of oleic acid, 1.00 g (2.27 mmol) of benzotriazol-1-yloxytris(dimethylamino)-phosphoniumhexafluorophosphate (BOP) and 20 ml of tetrahydrofuran (THF) were introduced into a flask. The reaction mixture was cooled to 0-5uC and 0.95 ml (6.81 mmol) of triethylamine in 4 ml THF were added dropwise within 15 min, and the reaction mixture was continuously stirred for 15 h at room temperature. After evaporating the solvent, 100 ml of diethyl ether were added and the ether phase was washed three times with 30 ml of a 4% HCl solution, twice with 30 ml of a saturated NaHCO 3 solution and twice with a saturated NaCl solution. After drying over anhydrous MgSO 4 and removal of the solvent, the crude product was purified from diethyl ether/hexane by crystallization, which gave 0.90 g (92% yield) of the amide as a white solid. In the synthesis of O-Me-OLDA, dopamine hydrochloride was replaced by 0.46 g (2.27 mmol) of 3-O-methyldopamine hydrochloride.

In vitro OLDA O-methylation -commercial COMT added
The reaction was performed in a reaction buffer (RB) containing phosphate buffered saline (PBS), pH 8.0, and 10 mM Mg 2+ . Briefly, 0.5 mg of COMT, derived from porcine liver, with the activity of 150 U/mg protein (basal activity 1.107 mmol substrate per hour) in 1% bovine serum albumin (BSA), 0.87 mg (1.11 mmol) of S-adenosyl-L-methionine (SAM) and 0.46 mg (1.11 mmol) of OLDA in a drop of Tween80 were dissolved in a volume of 0.1 ml each, mixed, and then the RB was added to a total volume of 0.5 ml. The reaction mixture and the control mixture (in which COMT was replaced by RB) were kept for 75 min at 37uC. After the reaction time, both solutions were centrifuged for 10 min at 12006g. The supernatant was extracted 460.25 ml CHCl 3 giving the organic and aqueous phases. In the organic phase (containing CHCl 3 as solvent) the hydrophobic compounds OLDA and O-Me-OLDA, the supposed products of the reaction, were expected. In contrast, in the aqueous phase, all ions and hydrophilic compounds like SAM and its demethylation product, S-adenosyl-L-homocysteine (SAH), were to be found. The organic phase was dried under N 2 , the aqueous phase was dried in a vacuum evaporator, and both phases were analyzed spectrophotometrically. An additional control consisting of a reaction mixture, with the same content of ingredients, analyzed 1 min after the commencement of the reaction, showed that the UV/VIS spectra were not different from those analyzed in the control solution above outlined. Therefore, the presence of COMT had no influence on the spectra.

5.
Ex vivo OLDA O-methylation -endogenous COMT in brain tissue COMT preparations were obtained as described by Brannan et al. [16]. Briefly, three fresh rat brains were homogenized in 4 v/m of 0.15 M KCl. The homogenate was centrifuged at 150006 g for 40 min. Then, the pellet was discarded and the supernatant was used as the enzyme preparation. Five milligrams of OLDA were dissolved in one drop of Tween80; then 250 ml of the supernatant, 2 mg of SAM, 50 ml of 5 mM MgCl 2 , and 1.7 ml of calcium-free PBS were admixed. In the control solutions, OLDA was omitted. After 1 h of incubation at 37uC, the reaction was stopped by the addition of 0.4 ml of 8% trichloroacetic acid and the proteins were precipitated by centrifugation at 20006 g for 5 min at 4uC. Then, the lipophilic compounds were extracted 4 times with 1 ml of chloroform, both phases dried and analyzed by HPLC-MS.

In vivo OLDA O-methylation
Three anesthetized rats were used in this part of the study. A canula was introduced upstream into a carotid artery to inject OLDA in a dose of 40 mg/kg dissolved in 0.3 ml of dimethyl sulfoxide (DMSO). One hour later, the animals were sacrificed as above outlined. Then, brains were enucleated from the skull, weighed, homogenized in 20 v/m of chloroform:methanol 2:1 v/v and left overnight at 4uC. Later, the mixture was dried under N 2 to a volume of 3 ml and lipids were extracted 3 times with chloroform and water (1:1 v/v, 4 ml). The organic, lipidcontaining fraction was dried under nitrogen and stored at 280uC until HPLC-MS analysis. Another three anesthetized control rats received 0.3 ml of DMSO alone and otherwise underwent the same experimental procedure.

COMT inhibition by tolcapone
In the in vitro part of the study, 1 mM of tolcapone was added to the reaction mixture 20 min before OLDA and then the reaction was allowed to proceed as outlined in section 2.3. In the ex vivo part of the study, brains extracted from two rats were used. After homogenization and centrifugation, as described in section 2.4, 250 ml of the supernatant, containing endogenous COMT, 2 mg of SAM, 50 ml of 5 mM MgCl 2 , and 1.7 ml of calcium-free PBS were mixed with tolcapone at a final concentration of 0.1 and 1 mM. The mixture was incubated 20 min at 37uC. Then, 5 mg of OLDA dissolved in one drop of Tween80 were added to the reaction mixture and incubated for 1 h at 37uC. The reaction proceeded as outlined in section 2.4. In the control solutions, OLDA was omitted. Each assay was performed in triplicate.
Six anesthetized animals were used for the in vivo tolcapone part of the study. The inhibitor was injected at a dose of 15 or 30 mg/ kg, i.p., in three animals each. Two hours later, the animals were prepared surgically and 40 mg/kg OLDA was injected into the carotid artery. The remaining procedure was as above described in the in vivo O-methylation (section 2.5).

Analytical methods
8.1. UV/VIS spectrophotometry. Organic phases of the samples obtained were dissolved in 0.3 ml of CHCl 3 , whereas the aqueous phases were suspended in 0.3 ml of methanol. All samples were analyzed spectrophotometrically using a Cintra 10e UV/VIS Spectrophotometer equipped with Spectral 1.70 (GBC Scientific Equipment Pty Ltd., Victoria, Australia) in a range of 190-1000 nm, with 0.427 nm steps. Spectrophotometric measurements were performed immediately after the extraction procedures.
8.2. HPLC-MS analysis. Samples for the HPLC-MS experiments were diluted in 1 ml of dichloromethane (DCM) and 20 ml of the solution was mixed with 1 ml of methanol. A standard mix of OLDA or O-Me-OLDA was prepared by diluting 1 mg of either in 1 ml of DCM, further diluted 100 times with methanol. The HPLC-MS experiments were carried out using a highperformance liquid chromatographic prominence LC-20 (Shimadzu, Kyoto, Japan) coupled with a tandem mass spectrometer 4000 Q TRAP (Applied Biosystems, Carlsbad, CA). The mass spectrometer was equipped with an electrospray ion source (Turbo Ion Spray) and a triple quadrupole/linear ion trap mass analyzer.
HPLC separation was performed using a 4.6650 mm Luna C8 (5 mm) column (Phenomenex, Torrance, CA). A linear multistep gradient was used in all measurements. The elution started from 20% water in methanol and reached 70% water in methanol at 12 min. This composition of the eluent was kept for 5 min and during the following one minute, the amount of water in methanol was increased to 80%; the level being maintained for further 2 min. The flow rate was 2.0 ml/min.
Multiple reaction monitoring (MRM) experiments were performed in the negative ion mode. Zero air was used as a nebulizer and nitrogen was used as a curtain gas. The tip voltage was kept at 24500 V and the declustering potential was set at 60 V. For all MRM experiments, the collision energy was set to 30 eV and the dwell time to 200 ms.

In vitro OLDA O-methylation -commercial COMT added
Spectra of the organic phases had the maxima at 278-282 nm in both control mixture (OLDA with no COMT; Figure 1A dashed line) and reaction mixture (OLDA with COMT; Figure 1A -continuous line). The spectrum of O-Me-OLDA alone, a presumed reaction product, is depicted by the dotted line in Figure 1A. Localization of the minimum changed from 262 nm in the control mixture to 260 nm in the reaction mixture. The minimum of the latter matched neither that of OLDA (substrate, l min = 262 nm) nor O-Me-OLDA (product, l min = 255 nm) (Fig. 1A), but was identical with the minimum of an artificially produced 1:1 mixture of OLDA and O-Me-OLDA (Fig. 1B), which points to the occurrence of a reaction of O-methylation of OLDA via COMT, yielding both OLDA and O-Me-OLDA in the assay. Additionally, the aqueous phases were compared and no change in the spectral maxima (l max = 259 nm) was found. However, the minimum changed from 234 nm in the control to 237 nm in the reaction mixture; these minima were characteristic of SAM and SAH, respectively (data not shown).

Ex vivo OLDA O-methylation -endogenous COMT in brain tissue
The organic phase was analyzed using both UV/VIS and HPLC-MS. The UV/VIS spectra showed a maximum at 278-282 nm and a minimum at 260 nm and were identical with the spectrum of a 1:1 mixture of OLDA and O-Me-OLDA. All these spectra corresponded closely to those in in vitro O-methylation, the graphic data are therefore not shown. The

COMT inhibition by tolcapone
To demonstrate that O-methylation involves the COMT pathway, we performed experiments in which COMT was inhibited by tolcapone. In the in vitro part of the study, the spectra of control (no COMT) and reaction (with COMT and 1 mM tolcapone) mixtures were identical, with the maxima and minima at 278-282 nm and 262 nm, respectively, which were the same as those noted for OLDA. In the ex vivo part, 1 and 0.   (mean 0.316560.049)%. At the higher dose, the respective intensities were of 9.7610 4 to 1.4610 5 (mean 1.2610 5 60.2610 5 ) counts per sec and of 120 to 210 (mean 148638) counts per sec; giving a further reduced methylation yield down to 0.109-0.139 (mean 0.12260.010)%. The methylation yields obtained during in vivo O-methylation in the presence of endogenous COMT, with no tolcapone added (as outlined in Section 3.3), and after injections of two doses of tolcapone, were significantly different from one another (p,0.023).

Discussion
The major finding of this study was that exogenously supplemented OLDA underwent COMT-catalyzed O-methylation. The methylation process was confirmed both in vitro, when commercial COMT was introduced into the assay, and ex vivo, when COMT native to the animal's brain tissue was the catalyst of the reaction. After systemic OLDA administration in vivo, O-Me-OLDA was positively identified in MS spectra obtained from the rat brain; 6.4-10.5% yield of O-methylation of OLDA was obtained. Pretreatment with the COMT inhibitor tolcapone drastically reduced this yield down to 0.11-0.14% in a dosedependent manner, demonstrating the preponderance of the COMT pathway for the methylation reaction. The essential role of COMT in the metabolism of OLDA was strengthened by the apparent lack of O-Me-OLDA in the in vitro and ex vivo assays in which COMT was inhibited. The identification of O-Me-OLDA and its way of formation expands the understanding of OLDA biotransformation and its potential physiological role.
A question arises as to whether endogenous OLDA could be a substrate for COMT. Since exogenous OLDA undergoes Omethylation in vivo and is directly methylated by commercially available and ex vivo prepared COMT, it is a reasonable assumption that endogenous OLDA would also undergo this reaction. However, endogenous OLDA, present in picomolar concentrations in the brain [4], was below the detection limit in the present experiment, which made it impossible to verify whether it could be a substrate for COMT. Chu et al. (4] have identified OLDA in the bovine brain as a compound having an MS pattern identical to that of artificially produced OLDA; however, its minute concentration apparently escaped quantitative determination. On the other hand, Huang et al. [17] quantified the amount of N-arachidonoyl-dopamine (NADA) at about 7 pmol/g wet tissue in the striatum, which approximates 3 ng/g wet tissue. As OLDA is believed to be an endovanilloid, belonging to the family of N-acylated-dopamines, as does NADA, it seems a reasonable assumption that the endogenous concentration of OLDA is in the same range. With that assumption in mind, in the present study we used the concentrations of OLDA of 2.2 mM (1.1 mmol/0.5 ml) in in vitro part and 6 mM (12.0 mmol/2 ml) in ex vivo part, which were non-physiologically high. These concentrations were based on the findings of Pokorski et al. [5] that one micromole of OLDA (the approximate amount used in vitro in the present study) is the presumed quantity of the compound to be present in 1 g of rat brain tissue after systemic administration of 40 mg/kg OLDA at a penetration yield of 6%.
The only available data on O-methylation of lipid derivatives of dopamine are those of Huang et al. [17] concerning NADA. NADA was methylated ex vivo by the soluble form of COMT from the liver at about 2 pmol/min per mg protein, which seems to be low compared with the values for dopamine methylation -Vmax from 50 pmol/min per mg protein in skeletal muscles to as high as 14,690 pmol/min per mg protein in the liver [18]. A slower turnover of N-acylated dopamines might underlie a much sustained bioactivity of these compounds [19] compared with that of dopamine proper; assuming that the kinetics of exogenous and endogenous N-acylated dopamines are similar; which, however, remains to be shown.
The formation of O-Me-OLDA is congruous with the notion that OLDA enters the dopamine metabolic pathway. COMT and MAO are the two main enzymes engaged in dopamine metabolism. Metabolic reactions run in two parallel ways: COMT methylates DA to 3-methoxytyramine, which is later oxidized by MAO to homovanillic acid (HVA), or MAO acts first yielding the 3,4-dihydroxyphenylacetic acid (DOPAC), which is later methylated by COMT to HVA [20]. It is believed that COMT is the predominant metabolic enzyme for brain dopamine [21]. Therefore, OLDA, which penetrates into the brain and stays there as a stable integral compound, would not only have a DA-like functional effects, such as increased locomotion or respiratory inhibition [7], which are blocked by DA antagonists, but would also engage the DA metabolic pathway. There are other possible ways of OLDA decomposition, which, albeit not tackled in the present study, seem rather unlikely. One of those is hydrolysis of OLDA at the amide bond. However, OLDA's lack of interaction with the fatty acid amide hydrolase (FAAH) [4] and its stability in vivo up to 24 h without traces of hydrolysis or auto-oxidation [19] speak against this possibility. Likewise, oxidation of the catechol ring, which in case of DA, especially under pathological conditions, gives rise to possibly toxic dopamine quinones [22], is unlikely in case of OLDA in mammalian neural tissue in which, as opposed to invertebrates [23], N-acyl-dopamine derivatives have not been found as substrates for tyrosinase-catalysed dopamine quinone formation. Thus, the most probable metabolic pathway of OLDA, apart from the possible sulfation reported previously by Akimov et al. [24], is its O-methylation.
Pre-methylation of OLDA could also have to do with its affinity to TRPV1 receptors [4,10]. TRPV1 receptor ligands are mostly characterized by the presence of a 3-O-methylated catechol ring [25]. Capsaicin analogs, with an identical to OLDA 8-carbon lipophylic chain, containing N-methyl-amide in the B-region, and 3,4-dihydroxybenzyl (catechol) or 3-methoxy-4-hydroxybenzyl (3methoxy-catechol or homovanillyl) group in the A-region are highly active regarding the transmembrane calcium influx [12,13].The homovanillyl group is believed preferable for the interaction with the receptor. The catechol ring of OLDA is sufficient for the interaction with TRPV1, but O-methylation of OLDA would promote the interaction of its homovanillyl structure with TRPV1 receptors [26]. There seem to be satisfied experimental conditions to support this reasoning. In the present study we demonstrate the in vivo O-methylation of OLDA and both COMT and SAM, the latter a donor of the methyl group, are intracellular compounds [18], as also is the TRPV1 ligand binding structure. Despite the desirable presence of a methyl group for TRPV1 activation, the issue of the superior efficacy of O-Me-OLDA remains unresolved. There is a report showing that NADA activates TRPV1 with the potency similar to that of capsaicin [17], whereas the study examining the effect of O-Me-OLDA alone reported its potency toward TRPV1 lower than that of capsaicin [14]. It ought to be emphasized, however, that although NADA and OLDA belong to the same family of N-acylated dopamines, both compounds have different bioproperties. NADA is hydrolyzed to dopamine and arachidonic acid by FAAH [17], whereas OLDA has, in contrast, no interaction with this enzyme [4]. NADA is a ligand of CB1 cannabinoid receptors [27] and OLDA has no major activity at this receptor system. NADA synthesis proceeds via condensation of arachidonic acid and dopamine [27], rather than via enzymatic conversion of N-oleoyl tyrosine to OLDA as it is suggested in case of OLDA [4], although either way of synthesis for either compound seems a viable, albeit not clearly confirmed, possibility. The resolution of an issue of superiority of O-Me-OLDA over OLDA regarding the interaction with TRPV1 receptors ought to be based on a direct comparison of both compounds' affinities, which requires alternative study designs.
In synopsis, although physiological role of endogenous OLDA remains unclear, we believe we have conclusively demonstrated that the compound, supplemented exogenously, enters the metabolic pathway typical of dopamine and undergoes Omethylation by COMT in vivo. The study lends support for the dopamine-like properties of OLDA and suggests the possible molecular mechanism of its facilitated affinity to the methyl groupseeking TRPV1 receptors. OLDA carries a potential to incorpo-rate lipid signaling into the hydrophilic dopamine pathway; which makes the compound of interest in studying the otherwise fleeting biological effects of dopamine. OLDA holds promise as a substitute for DA in DA deficient pathological states, which makes it worthwhile to further explore the properties of OLDA in exogenously bioactive doses.