New Radioligands for Describing the Molecular Pharmacology of MT1 and MT2 Melatonin Receptors

Melatonin receptors have been studied for several decades. The low expression of the receptors in tissues led the scientific community to find a substitute for the natural hormone melatonin, the agonist 2-[125I]-iodomelatonin. Using the agonist, several hundreds of studies were conducted, including the discovery of agonists and antagonists for the receptors and minute details about their molecular behavior. Recently, we attempted to expand the panel of radioligands available for studying the melatonin receptors by using the newly discovered compounds SD6, DIV880, and S70254. These compounds were characterized for their affinities to the hMT1 and hMT2 recombinant receptors and their functionality in the classical GTPγS system. SD6 is a full agonist, equilibrated between the receptor isoforms, whereas S70254 and DIV880 are only partial MT2 agonists, with Ki in the low nanomolar range while they have no affinity to MT1 receptors. These new tools will hopefully allow for additions to the current body of information on the native localization of the receptor isoforms in tissues.

hormone melatonin, the agonist 2-[ 125 I]-iodomelatonin. Using the agonist, several hundreds of studies were conducted, including the discovery of agonists and antagonists for the receptors and minute details about their molecular behavior. Recently, we attempted to expand the panel of radioligands available for studying the melatonin receptors by using the newly discovered compounds SD6, DIV880, and S70254. These compounds were characterized for their affinities to the hMT 1 and hMT 2 recombinant receptors and their functionality in the classical GTP S system. SD6 is a full agonist, equilibrated between the receptor isoforms, whereas S70254 and DIV880 are only partial MT 2 agonists, with K i in the low nanomolar range while they have no affinity to MT 1 receptors. These new tools will hopefully allow for additions to the current body of information on the native localization of the receptor isoforms in tissues.

Introduction
Melatonin is a neurohormone produced by the pineal gland at night [1,2] and is thought to control circadian rhythm. The activity of melatonin is relayed primarily by the seven transmembrane G protein-coupled receptors MT 1 and MT 2 [3]. The pharmacological actions of melatonin at higher concentrations (µM and above) are thought to be mediated by other protein targets, such as QR2 [3,4]. Before these receptors were cloned in the late 1990s [5,6], the measurement of melatonin binding to membranes derived from any animal organ was difficult because the level of expression is naturally very low for these receptors, and the available radioligand ([ 3 H]-melatonin) could not be synthesized with enough specific activity.
In 1984, a Finnish group described the use of a more sensible tool for binding studies: 2-[ 125 I]-iodomelatonin ([ 125 I]-2IMLT), a strongly labeled super-agonist of the receptors [7,8]. Almost immediately, all experiments reported in the literature used this tool, and the labeled natural hormone ([ 3 H]-melatonin) was not used again until a later, complete study [9]. To determine whether other tools for molecular pharmacology studies of melatonin are useful, particularly regarding the agonist nature of this ligand (2-iodomelatonin, 2IMLT), we sought an alternative to [ 125 I]-2IMLT by screening our compounds to identify antagonist(s) or partial agonists (30% and below) for any or both receptor isoforms. The aim was to broaden the panel of available tools for studying these receptors [3,10]. By conducting several series of high throughput screening HTS campaigns [11,12], we found a MT 2 -specific partial agonist, DIV880. The K i of this compound is 2 logs less potent with MT 1 than MT 2 . As a first step, we synthesized the precursors of each ligand, iodinated them, and assessed their binding characteristics with recombinant human MT 1 and MT 2 receptors.
Over the last few years, our search for new ligands has been driven mainly by the addition of more molecular tools to the available panel of molecules used to study melatonin receptors, such as ligands specific to MT 1 or MT 2 and antagonist(s) of the melatonin receptors. Such a discovery would help broaden our understanding of this system for which almost no molecules have been reported [13]. It thus follows that such a molecule would then be labeled in order to obtain new ligands.

Results and Discussion
SD6 and S70254 were the products of a rational exploration of the basic features of melatonin analogs and are structurally related to the natural hormone ( Figure 1). The influence of iodine atoms on the core structures was studied systematically, leading to interesting compounds that can be used to complete the set of labeled compounds for describing the molecular pharmacology of melatonin receptor(s). For DIV880, the process was a little different. The compound is an iodinated analog of a bromo-compound that resulted from a large screening process, similar to the compounds described elsewhere [11,12]. This compound was specific to MT 2 , with a pK i 2 logs "better" for MT 2 than for MT 1 . Therefore, the iodine equivalent was synthesized, leading to its possible use as a specific probe for MT 2 . Prior to radiolabeling the compounds, we characterized the cold compounds for recombinant human MT 1 and MT 2 receptors and compared them to 2IMLT (Table 1). 2IMLT and SD6 shared similar properties, particularly similar affinities for the receptors in the low nanomolar range, similar potency in the GTP S assay, with full agonistic effects for both receptors (E max > 100%), and almost similar pEC 50 for both receptors (with a minor discrepancy for MT 1 ) in a TR-FRET cAMP assay (E max ~100%). S70254 was a poor MT 1 ligand, in the micromolar range (pK i = 6.18), but a good MT 2 ligand (pK i = 8.73). The functionality of S72054 remained measurable but poor for MT 1 , but was a partial agonist for MT 2 (43%) in the GTP S assay and a less partial agonist (73%) in the TR-FRET cAMP assay. Finally, for DIV880, the affinity for MT 1 remained in the micromolar range, but it was in the 10 nM range for MT 2 . Similarly, the functional effect of this compound was impossible to record using both functional assays with MT 1 , but the compound behaved as a partial agonist (67%) of MT 2 in the GTPγS assay and full agonist (97%) of MT 2 in the TR-FRET cAMP assay.
We proceeded to label the compounds with [ 125 I], and characterized the ligands for both recombinant receptors. The compounds behaved properly in the binding assays after a rapid assessment of the experimental conditions. Figure 2    I]-2IMLT has been monopolizing the melatonin binding field since it was first described by Vakkuri et al. [7], and [ 3 H]-melatonin was completely described in 2000 [9] and rarely used then after. [ 125 I]-S70254 and [ 125 I]-DIV880 will permit the first specific investigation of the hMT2 receptor, as well as cellular systems overexpressing MT receptors and tissue samples (e.g., binding and autoradiography).

2-[ 125 I]-iodomelatonin and [ 35 S]-GTPγS Binding Assays
The assays were described previously [16]. Briefly, for competition experiments in CHO cells, the membranes were incubated in 250 µL binding buffer (50 mM Tris/HCl pH 7.4, 5 mM MgCl 2 ) containing 20 pM [ 125 I]-2IMLT for 2 h at 37 °C. The results were expressed as the inhibition constant Ki, taking into account the concentration of radioligand used in each experiment. Non-specific binding was defined using 10 μM melatonin. The reaction was stopped by rapid filtration through GF/B unifilters, followed by three successive washes with ice-cold buffer. The data were analyzed using the program PRISM (GraphPad Software Inc., San Diego, CA, USA). K i was calculated according to the Cheng-Prussof Equation: where IC 50 is the half maximal inhibitory concentration and L is the concentration of [ 125 I]-2IMLT [17].
For the [ 35 S]-GTPγS binding assay, the membranes and compounds were diluted in the binding buffer (20 mM Hepes pH 7.4, 100 mM NaCl, 3 mM MgCl 2 , 3 µM GDP) in the presence of 20 µg/mL saponin in order to enhance the agonist-induced stimulation [16]. Incubation was started by adding 0.1 nM [ 35 S]-GTPγS to the membranes and ligands in a final volume of 250 µL and allowed to continue for 60 min at room temperature. Non-specific binding was assessed using non-radiolabeled GTPγS (10 µM). Reactions were stopped by rapid filtration through GF/B unifilters pre-soaked with distilled water, followed by three successive washes with ice-cold buffer. The data were analyzed using the program PRISM to yield the half maximal effective concentration (EC 50 ) and maximal effect (E max ) expressed as a percentage of that observed with melatonin (1 µM = 100%). pEC 50 was calculated as pEC 50 = −log(EC 50 ).

New Ligand Binding Assays
The assays were performed in 96-well plates in 250 µL binding buffer (50 mM Tris/HCl pH 7.4, 5 mM MgCl 2 , 1 mM EDTA, plus BSA 0.1% for [ 125 I]-DIV880). The membranes, hMT 1 and hMT 2 , were used at a final concentration of 30 µg of proteins/mL for all radioactive compounds. For all protocols, the reaction was stopped by rapid filtration through GF/B unifilters (PEI 0.1% treated for [ 125 I]-DIV880), followed by three successive washes with ice-cold buffer (50 mM Tris/HCl, pH 7.4). The data were analyzed using the program PRISM (GraphPad Software Inc., San Diego, CA, USA). For the saturation assay, the binding site density (B max ) and dissociation constant for the radioligand (K d ) were calculated according to the Scatchard method.

HTRF cAMP Assay
Cellular cAMP production was measured using cAMP dynamic HTRF kits (Cisbio Bioassays, Bedford, MA, USA) according to the manufacturer's instructions. CHO-K1 cells stably expressing the hMT 1 or hMT 2 receptor were grown to confluence, harvested in PBS buffer containing 5 mM EDTA, and centrifuged at 100× g for 10 min (4 °C). The cell pellet was re-suspended in 0.5 mM HAMF12 IBMX at a concentration of 2 million cells/mL. Incubation was started by adding 5 µM forskolin (15 µL/well) to the cells (30,000 cells/well) and compounds (15 µL/well, DMSO 1.7%) in a final volume of 60 µL, and allowed to continue for 20 min at 37 °C. Next, 15 µL of cAMP-d2 conjugate and 15 µL of anti-cAMP-EuK conjugate in lysis buffer were incubated for 30 min at room temperature. The fluorescence intensity was measured at 340 nm excitation and 665 and 620 nm emission on an Envision (Perkin Elmer, Downers Grove, IL, USA). The TR-FRET 665 nm/620 nm ratio, which is inversely proportional to the production of cAMP, was used to determine the cAMP response. Non-specific binding was assessed using 100 µM non-labeled cAMP. The data were analyzed using the program PRISM to yield the EC 50 and E max .

Chemistry
In order to find and characterize new ligands for the melatonin receptors, the following strategy was used. Positive ligands bearing either an iodine or bromide atom were selected from either our chemical series or from the vast HTS campaigns we conducted, In the case of bromide, the cold iodinated compound was synthesized and tested for its characteristics at the receptors. With these results in hand, the most interesting compounds were selected based on properties such as affinity (nanomolar range), MT 1 versus MT 2 selectivity (specificity at MT1 versus MT2 receptors should be at least 2 logs in order to be usable as specific ligands, though no clear consensus exists on this particular point), functionality (agonists versus antagonists), and accessibility to the radio-iodination process.
DIV879 was discovered during a large high-throughput screening campaign as an MT 2 antagonist with at least 2 logs poorer affinity for MT 1 . This compound bears a bromide.
3.6.2. Synthesis of 6,7-dimethoxy-3-isochromanone (Compound 1, Figure 3) Ten milliliters of formaldehyde (37% in water) was added drop-wise to a solution of 10.3 g (52.4 mmol) 3,4-dimethoxyphenylacetic acid in a mixture of 10 mL of HCl 37% and 30 mL of acetic acid at room temperature. The mixture was heated to 90 °C over 1 h, cooled to room temperature, and hydrolyzed by the addition of 300 mL of water. The aqueous layer was extracted three times with dichloromethane. The organic layer was washed three times with saturated sodium bicarbonate solution, dried over magnesium sulfate, filtered, and evaporated to dryness to obtain 7.39 g of crude material. Trituration in isopropylic ether for 1 h resulted in 6.51 g (60%) of compound 1 as an off-white solid after filtration. 1H NMR (CDCl 3 ) 6.76 (s, 1H), 6.73 (s, 1H), 5.28 (s, 2H), 3.91 (s, 3H), 3.90 (s, 3H), 3.66 (s, 2H).  Figure 3) A total of 2.3 g (10.8 mmol, 1.13 eq.) of bromoveratrole was added in one portion to a solution of 2.0 g (9.6 mmol) of 6,7-dimethoxy-3-isochromanone 1 in 20 mL of formic acid. The mixture was heated to 90 °C for 2 h and then hydrolyzed by the addition of 10 mL of water. The aqueous layer was extracted three times with ethyl acetate. The organic layer was washed twice with brine, dried over magnesium sulfate, filtered, and evaporated to dryness to yield 4.18 g of crude material. Trituration in a mixture of isopropylic ether/AcOEt resulted in 3.2 g (78%) of compound 2 as an off-white solid. 1

Synthesis of Tert-butyl 2-(2-[(2-bromo-4,5-dimethoxyphenyl)methyl]-4,5-dimethoxy phenyl)
Acetate (DIV879, Figure 3) A total of 1.6 g (7.35 mmol, 3 eq.) O-tert-butyl-N,N′-diisopropylisourea was added in portions over a period of 2 h to a solution of 1.04 g (2.45 mmol) of acid 2 in 30 mL of dichloromethane at room temperature. The mixture was stirred overnight and filtered to remove insoluble materials. The filtrate was evaporated under vacuum and the residue purified twice by chromatography on silica gel (1-eluant: CH 2 Cl 2 /AcOEt 85/15, 2-eluant: heptane/AcOEt gradient) to obtain 749 mg (63%) of DIV879 as an off-white solid. 1 Figure 4) A total of 2.2 mL (17.3 mmol, 1.2 eq.) of veratrole was added to a solution of 3.0 g (14.4 mmol) of 6,7-dimethoxy-3-isochromanone 1 in 30 mL of formic acid at room temperature. The mixture was maintained at 95 °C overnight and then hydrolyzed by the addition of 30 mL of water. The aqueous layer was extracted three times with ethyl acetate. The organic layer was washed twice with brine, dried over magnesium sulfate, and evaporated to dryness to obtain 7.0 g of crude material. Chromatography on silica gel (eluant: heptane/AcOEt 1/1 + 1‰ AcOH) followed by trituration in isopropylic ether resulted in 2.2 g (44%) of compound 3 as an off-white solid after filtration. 1 Figure 4) A total of 1.7 g (4.2 mmol, 1 eq.) of compound 2 dissolved in 25 mL of acetic acid was added drop-wise to a solution of 1.77 g (6.3 mmol, 1.5 eq.) of chloramine T and 944 mg (6.3 mmol, 1.5 eq.) of sodium iodide in 25 mL of acetic acid stirred over 1 h. After two hours at room temperature, the mixture was poured into 20 mL of water and extracted three times with ethyl acetate. The organic layer was washed twice with brine, dried over magnesium sulfate, and evaporated to dryness to yield 4.0 g of crude brownish material. Purification by chromatography on silica gel (eluant: heptane/AcOEt 8/2) resulted in 1.2 g (55%) of compound DIV880 as a yellow oil. 1 Figure 5) was obtained through two routes according to the synthetic pathway illustrated in Figure 5: (A) by reaction of 5-methoxytryptamine with bromoacetyl bromide in a biphasic medium (EtOAc-H2O) according to a variant of the Schotten-Baumann reaction (K 2 CO 3 ) to obtain a bromoacetyl derivative [18]. Substitution of the bromine atom of compound 5 by refluxing in acetone with sodium iodide resulted in the iodo derivative SD6.

Synthesis of N-[2-(5-methoxy-1H-indol-3-yl)ethyl]iodoacetamide (compound SD6,
Method A: A solution of 3.11 g (0.01 mol) of compound 5 in 50 mL of anhydrous acetone was treated with 1.5 g (0.01 mol) of sodium iodide and the mixture heated at reflux for 2 h. After cooling, the reaction mixture was filtered and evaporated. The residue was then crystallized from toluene, resulting in 2.5 g (70%) of compound 5.
Method B: A solution of iodoacetic acid (1.85 g, 0.01 mol) in 50 mL of methylene chloride was stirred at −10 °C for 20 min. Triethylamine (1.62 mL, 0.012 mol), EDCI (1.86 g, 0.0012 mol), and HOBt (1.62 g, 0.0012 mol) were added, and the mixture stirred at -10 °C for 30 min. A solution of 5-methoxy tryptamine (1.9 g, 0.01 mol) in 10 mL of methylene chloride was cooled at -10 °C and added drop-wise. After 6 h of stirring at room temperature, the reaction mixture was washed with water, a 1M HCl solution, water, a 10% NaOH solution, and water until a pH of 7 was reached. The organic phase was dried over MgSO 4 , filtered, and concentrated under reduced pressure. The residue was then crystallized from toluene, obtaining 2.  Figure 6) [19,20] (1.21 g, 3.4 mmol) was solubilized in a 1:10 water/methanol mixture. Potassium hydroxide (6.61 g, 118 mmol) was added and the reaction mixture refluxed for 87 h. After cooling, methanol was evaporated under reduced pressure and water (100 mL) added. The product was extracted with ethyl acetate (3 × 50 mL) and the organic phases combined and washed with a saturated aqueous solution of ammonium chloride (50 mL). The organic phase was dried over magnesium sulfate, filtrated, and evaporated under reduced pressure. The crude product was purified by silica gel column chromatography (ethyl acetate 84/methanol 15/ammoniac 1) to obtain amine 8 as a colorless oil (0.62 g).

Selectivity Studies for Melatonin Receptor Ligands
In order to evaluate ligand selectivity, the cold compound S70254, SD6, and DIV880 were submitted to our standard selectivity procedure. The specificity of the compounds was assessed by testing a standard set of receptors and a small number of enzymatic targets:  19 ]-MCH. Inhibition of the activity of the following enzymes was also tested: caspase-3(h); EGFR kinase(h); and PKCα(h). For all of the details and protocols, go to www.cerep.org. Our compounds did not exhibit any activity towards these targets; they were mainly inactive or barely active (less than 20% effect) at 10 µM. This margin was thought to be enough to consider the compounds specific for the melatonin receptor(s).

Conclusions
Though the description of SD6 did not result in a new tool for studying melatonin receptors, the present work permitted the discovery, synthesis, and characterization of two specific ligands for the MT2 isoform: S70254 and DIV880. Complete characterization of the kinetics of ligand binding has started in order to determine if the behavior of these two ligands is different from [ 125 I]-2IMLT and [ 3 H]-melatonin using the available melatonin receptors from various species [14,16,22,23]. The next steps will comprise a comparison of the pharmacology of these ligands using a rather large set of compounds already described by our group, as well as binding and autoradiography using native organ membrane preparations or slices. We are still searching for a MT 1 -specific ligand. Unfortunately, the melatonin receptor ligand literature is sparse in regards to compounds with specificity better than 2 logs between receptors, despite several attempts by others [24,25] and our own group [26][27][28].