CHEMMEDCHEM COMMUNICATIONS DOI: 10.1002/cmdc.201402122
Benzocyclobutane, Benzocycloheptane and Heptene Derivatives as Melatonin Agonists and Antagonists Andrew Tsotinis,*[a] Pandelis A. Afroudakis,[a] Peter J. Garratt,[b] Alina Bocianowska-Zbrog,[c] and David Sugden[c] Two series of analogues were designed, synthesised and evaluated as potential human melatonin type 1 and 2 receptor (hMT1 and hMT2) ligands. Their biological effects were assessed by a well-established, specific model of melatonin action, the pigment response of Xenopus laevis melanophores. Compounds containing a benzocyclobutane scaffold and a methoxy group in the “melatonin” orientation were found to be potent agonists, with one of the analogues exhibiting activity comparable to melatonin. In contrast, analogues with a methoxy group in non-melatonin positions or with multiple methoxy groups showed either weaker agonist activity or were antagonists. Benzocycloheptene derivatives with one methoxy group are found to be weak agonists, whereas those with two methoxy groups were found to be antagonists, as were all of the benzocycloheptane derivatives evaluated. The most active compounds were assessed in a human receptor radio ligand binding assay but showed little discrimination between MT1 and MT2. These results again show that the indole nitrogen of melatonin is not a necessary component for analogue activity and also illustrate that replacement of the indole ring with a 4membered carbocycle can provide highly active compounds when the methoxy group is in the melatonin position.
brane receptors,[3] and radioligand binding studies using 2[125I]-melatonin (1 b) have revealed a widespread, heterogeneous distribution of binding sites throughout the central nervous system (CNS).[4] Two receptor subtypes have been cloned in mammals (MT1, MT2) that, when expressed in host cells, show the general characteristics of native melatonin receptors.[5, 6] A third receptor subtype, Mel1c, has been cloned from chicken, Xenopus and zebrafish, but has not been detected in mammals.[7] Melatonin has the ability to phase-shift the circadian rhythm and has been shown to be effective in treating insomnia and jet-lag in humans. Two synthetic melatoninergic agonists are approved for use in insomnia (ramelteon) and depression (agomelatine),[2] and a third (tasimelteon) has recently been approved for non-24-hour sleep–wake disorder.[8] A variety of
The pineal hormone melatonin (N-acetyl-5-methoxytryptamine, 1 a)[1, 2] is an important neurohormone in the regulation of seasonal and circadian rhythms. In mammals, the production of melatonin is controlled by a small nucleus within the hypothalamus of the brain—the suprachiasmatic nucleus (SCN), which is the body’s biological clock. The SCN, entrained by light–dark information from the retina, instigates the production and release of melatonin during the dark period and ensures that melatonin levels in the light phase remain very low. The actions of melatonin are believed to be mediated through a family of specific, high-affinity, G protein-coupled cell-mem-
other effects of melatonin administration have been reported and additional physiological roles suggested.[1, 2] The potential of melatoninergic agents in a variety of medical conditions has driven a large number of studies directed towards understanding how melatonin binds to and activates these receptors. A substantial number of both indole and non-indole derivatives have been evaluated.[1, 2, 9] The 5-methoxy group has been shown to be important for receptor binding, but it is not essential for agonist activity,[10] and the indole nitrogen does not serve any obvious binding function and can be omitted in active analogues.[11–14] The indole ring can be replaced by 5- and 6-membered rings,[1, 2] and dibenzocycloheptene derivatives have also been reported.[15] The active conformation of the 3-ethanamine side chain has been studied in conformationally restricted indole[16, 17] and non-indole analogues.[18–22] A large number of analogues selective for the MT2 receptor have been synthesised,[23] and a much smaller number for the MT1 receptor, the latter generally being linked systems or compounds with substantial side chains.[2] Whereas the MT2-selective analogues can be highly discriminatory between the two receptor types, the MT1-selective analogues are much less so.
[a] Prof. A. Tsotinis, Dr. P. A. Afroudakis Department of Pharmaceutical Chemistry Faculty of Pharmacy, School of Health Sciences, University of Athens Panepistimioupoli-Zografou, 157 71 Athens (Greece) E-mail: [email protected] [b] Prof. P. J. Garratt Department of Chemistry, University College London 20 Gordon Street, London WC1H 0AJ (UK) [c] A. Bocianowska-Zbrog, Dr. D. Sugden Division of Women’s Health, School of Medicine, King’s College London Room 2.12N Hodgkin Building, Guy’s Campus, London SE1 1UL (UK) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201402122.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2238 – 2243
2238
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
Here, we report two types of system where the 5-membered pyrrole ring of melatonin has been replaced by either a 4- or 7-membered carbocyclic ring, and we describe the effect that these substitutions have on the receptor binding and biological activities of these molecules. Preparation of the benzocyclobutane precursors followed similar synthetic sequences (Scheme 1). The appropriately substituted bromobenzene was converted to the benzyne with sodium amide and added to 1,1-dimethoxyethene. The benzocyclobutane thus formed was treated with hydrochloric acid to remove the acetal protecting groups and give the corresponding benzocyclobutanone. In the reaction of p-bromoanisole (1 c) with 1,1-dimethoxyethene, two stereoisomers were obtained, which were hydrolysed to the respective benzocyclobutanones (3 c and 3 d). These were separated chromatographically (3 c/3 d, 1/1.5) and then under Wittig reaction conditions (Ph3P + CH2CNCl ) were converted to the corresponding vinyl cyanide, which was then reacted with Raney nickel and the appropriate acid anhydride under hydrogen at 55 psi at 60 8C. The resulting amides were purified by flash column chromatography. Benzocycloheptano 13 a–c and benzocyclohepteno amide derivatives 16 a–c and 20 a–c were prepared by one of two routes from the corresponding and readily available benzo[7]annulen-5-one derivatives. The suitable substituted benzo[7]annule-5-one was either treated with nitromethane to give the nitromethyl derivative or reacted with trimethylsilyl
cyanide to give the corresponding nitrile. Both types of derivative could then be reduced to the corresponding amines, which were then acylated using the appropriate acid anhydride to give the desired amides (Scheme 2). Saturated derivatives 21 a–c were prepared from their corresponding congeners 20 a–c through reduction with a mixture of trifluoroacetic acid (TFA) and trimethylsilane. The results for the pigment aggregation studies for the two series of compounds are reported in Tables 1 and 2. For the benzocyclobutane derivatives (Table 1), the series of compounds 5 a–c and 7 a–c have the methoxy group in the “wrong” orientation to the side chain when compared to melatonin (Figure 1). Nevertheless, both of these series do show some agonist activity. Compounds in series 5 are all partial agonists, and 5 a and b display weak antagonist activity (IC50 ~ 5– 10 mm). No antagonist activity is observed in the series 6, 7 or 8 (IC50 > 100 mm). Compound 5 a (R = Me) has very little agonist intrinsic activity (IA = 11 %) even at high concentrations (100 mm), whereas the compounds with longer R groups (5 b: R=Et; 5 c: R = n-Pr) show improved potency and intrinsic activity (IA = 55 % and 67 %, respectively). The poor activity of 5 a suggests that the alkylamide side chain and the methoxy group cannot occupy receptor binding sites at the same time, whereas the increased potency of 5 b and 5 c suggests that these longer alkyl chains can fit into their binding site at the same time that the methoxy group is in its binding pocket. A similar observation was made in a series of simple phenylalkyamides[11] in which the position of the methoxy group on the phenyl ring was varied. Compound 7 a (R = Me), like 5 a, is a partial agonist, albeit with greater intrinsic activity (78 %), while 7 b (R = Et) and 7 c (R = n-Pr) are full agonists in the melanophore assay. Compound 7 b (R = Et) has the greatest potency in this series and is only 150-fold less potent than melatonin and only sixfold less potent than 8 b, in which the orientation of the methoxy and side chain are the same as in melatonin. Compounds 8 a and 8 c are, like 8 b, full agonists, and a clear increase in potency (~ 16-fold) is found as the amide carbonyl alkyl group is increased from methyl to n-propyl, as we and others have found in analogous series.[2] Series 6 compounds have a methoxy group in the “correct” melatonin orientation like 8 a–c but have an additional methoxy group in the same relative position as the series 5 compounds. Compound 6 a is a partial agonist (IA = 54 %), like the other analogues in which the R group is a methyl (5 a and 7 a); compound 8 a, with the single, correctly orientated methoxy group, is the only acetyl compound with full agonist activity. As occurs in the compound 7 series, changing the R group from methyl to ethyl or propyl in the 6 series gives full agonist activity, although 6 b and 6 c are less potent than their 8 b and 8 c counterparts (23-fold and 130-fold, respectively), suggesting that the second methoxy group interferes with bindScheme 1. Synthetic route to the benzocyclobutane derivatives. Reagents and conditions: ing, possibly by competition with the “correctly” oria) NaNH2, THF, 12 h, reflux; b) THF/HCl, 12 h, room temperature; c) DBU, Ph3P + CH2CNCl , entated methoxy group. Compound 8 c, with ntoluene, reflux, 12 h; d) H2, Raney-Ni, (RCO)2O, 60 8C, 8 h. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2238 – 2243
2239
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
they have not been separated. Compounds 8 b and 8 c bind with high affinity to both the human MT1 and MT2 receptors, showing a small preference for the latter. For the benzocycloheptanes and heptenes (Table 2), the methoxy groups are in the “wrong” orientation compared with melatonin (Figure 1). The main difference in this series is the presence of a double bond in the 7membered ring for compounds of type 16 and 20. The only series showing agonist activity was 16 a–c, with a double bond in the 7-membered ring, although all three compounds are weakly active (EC50 = 1–1.5 mm). As was seen for most of the benzocyclobutane derivatives (Table 1), compounds with a methyl R group were partial agonists, 16 a (IA = 54 %), whereas 16 b (R = Et) and 16 c (R = nPr) were full or nearly full agonists (IA = 102 % and 82 %, reScheme 2. Synthetic route to the benzocycloheptane and heptene derivatives. Reagents and conditions: spectively). Series 16 compounds a) Me3SiCN, ZnI2, 12 h, 40 8C; b) Me3SiCl, NaI, H2O, 6 h, room temperature; c) MeNO2, NH4OAc, 24 h, 110 8C; show little antagonist activity d) LiAlH4, 0.5 h, 0 8C–RT; e) (RCO2)O, Et3N, 15 min, 0.5 h, 0 8C–RT; f) TFA, Me3SiH, 65 8C, 2 h. (< 30 %) even at high concentrations (100 mm). By contrast, series 13 compounds, with no double bond in the 7-mempropyl R group, is comparable in effect to melatonin. Since bered ring, have no agonist activity (up to 100 mm) but all these compounds are mixtures of the R and S stereoisomers, it three analogues, 13 a–c, were full antagonists, blocking melais probable that one isomer is more potent than the other, but tonin aggregation completely although with low potency (IC50 = 2.4–4.9 mm). Compounds 20 a–c and 21 a–c, which contain two methoxy Table 1. Agonist, antagonist melatoninergic action of benzocyclobutanes 5–8 (Xenopus laevis melanophores).[a] groups, were also antagonists but were somewhat weaker (~ 2–7-fold) than series 13 compounds. The introduction of the [a] [b] [a] [b] Compd
R
Melatonin Luzindole[d] 5a Me 5b Et 5c n-Pr 6a Me 6b Et 6c n-Pr 7a Me 7b Et 7c n-Pr 8a Me 8b Et 8c n-Pr
EC50 [nm]
0.10 0.002 NA > 100 000 99 1.5 99 1.2 966 288 70 5.5 118 31 558 51 16.8 1.0 102 1 15.9 2.5 3.0 0.3 0.9 0.0
IA [%]
100 – 11 1 54 1 67 1 54 2 100 2 90 2 78 6 94 2 93 1 103 3 80 1 101 1
IC50 [nm] [c]
NA 2460 460 10 470 884 5590 380 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 NA NA NA
IA [%]
– 100 3 88 1 59 2 45 2 24 6 32 15 1 23 2 15 2 19 2 – – –
[a] Agonist (EC50) and antagonist (IC50) data on melanophores represent the mean SEM of triplicate experiments. [b] IA = intrinsic activity. Agonist—maximum activity relative to melatonin. Antagonist—maximum percentage inhibition of melatonin (1 nm) response. [c] NA = no activity detected at 100 000 nm. [d] Ref. [25].
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. Orientation in space of the methoxy groups and side chain of the new analogues relative to melatonin.
ChemMedChem 2014, 9, 2238 – 2243
2240
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
Table 2. Binding affinities of benzocyclobutanes 8 b and 8 c against human melatonin receptor subtypes hMT1 and hMT2. Compd
hMT1
Ki [nm][a] hMT2
SI[b] (MT2/MT1)
0.66 603 0.8 1.0
0.33 45 0.5 0.4
2 13.4 1.6 2.5
R
Melatonin Luzindole[a] 8b 8c
Et n-Pr
Table 4. Binding affinity of benzocycloheptene analogue 16 b against human melatonin receptor subtypes hMT1 and hMT2. Compd Melatonin Luzindole[a] 16 b
hMT1
Ki [nm][a] hMT2
SI[b] (MT2/MT1)
0.66 603 41
0.33 45 10
2 13.4 4.1
R
Et
[a] Ref. [25]. [b] Selectivity index (SI).
[a] Ref. [25]. [b] Selectivity index (SI).
second methoxy group in series 20 effectively converts the agonist activity of the 16 series into antagonist activity, although it is weaker than that of the saturated series 21. For compounds in both series 20 and 21, a clear increase in antagonist potency was observed on increasing the length of the alkyl chain from methyl to n-propyl (Table 3). Compound 21 c has comparable potency to the melatonin antagonist, luzindole.[24] The most active cycloheptene agonist, 16 b, with COEt as the R group, exhibits significant binding affinity for the human melatonin receptor and shows a small preference for hMT2. It is interesting to note, however, that whereas 16 b binds only 41 times and 25 times less efficiently at hMT1 and hMT2 (Table 4) than 8 b (Table 2), it is approximately 1100 times less effective in the melanophore assay. Cycloheptene 16 b exists as a single isomer, whereas the cyclobutane analogues are a mixture of enantiomers, and one enantiomer may be much more potent than the other, which may distort the comparison of results between the two series of compounds. As Figure 1 indicates, the benzocyclobutane analogues that have the methoxy group in approximately the same position as melatonin are the most potent and mirror melatonin in that
Table 3. Agonist, antagonist melatoninergic action of benzocycloheptanes 13 and 21 and heptenes 16 and 20 with Xenopus laevis melanophores.[a] Compd
R
Melatonin Luzindole[d] 13 a Me 13 b Et 13 c n-Pr 16 a Me 16 b Et 16 c n-Pr 20 a Me 20 b Et 20 c n-Pr 21 a Me 21 b Et 21 c n-Pr
EC50 [nm][a]
IA [%][b]
IC50 [nm][a]
IA [%][b]
0.10 0.002 NA NA NA NA 1540 40 1100 35 1230 100 NA NA NA NA NA NA
100 – 54 2 102 2 82 2 – – – – – – – – –
NA[c] 2460 460 3790 230 4870 100 2410 80 > 100 000 > 100 000 > 100 000 27 700 2800 16 350 970 12 900 3000 10 100 1360 6560 980 3300 150
– 100 3 79 1 73 2 89 1 24 2 82 29 2 87 2 90 1 95 1 95 1 90 1 94 1
[a] Agonist (EC50) and antagonist (IC50) data on melanophores represent the mean SEM of triplicate experiments. [b] IA = intrinsic activity. Agonist—maximum activity relative to melatonin. Antagonist—maximum percentage inhibition of melatonin (1 nm) response. [c] NA = no activity detected at 100 000 nm. [d] Ref. [25].
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
lengthening the N-acetyl group to N-n-propyl increases the activity. In the case of benzocyclobutanes 7 a–c, where the methoxy group is in the “wrong” orientation with regard to the side chain, potency is at a maximum with N-ethyl, probably indicating that the n-propyl side chain further displaces the methoxy group from its binding site, although this derivative is still more potent than the N-acetyl compound. In conclusion, as previously observed, the indole nitrogen is not a necessary component of a melatonin analogue and the substitution of a 4-membered carbocyclic ring provides highly active compounds when these have a methoxy group in the melatonin position. Separation of the enantiomers of 8 c should be of interest in determining whether they have different activities, as should the introduction of an exocyclic double bond in the benzocyclobutane analogues, since this has a significant effect in increasing the potency of the benzocycloheptene derivatives.
Experimental Section Biology The biological activity of the analogues was determined in a specific model of melatonin action, the pigment aggregation response of Xenopus laevis melanophores, which we described previously.[26] In these cells, melatonin triggers a dramatic movement of melanin pigment, normally distributed throughout the cell, towards the cell centre. This response is termed pigment aggregation and can be accurately quantified in melanophores grown in 96-well plates simply by measuring the change in light (630 nm) absorbance of the cells as the pigment concentrates at the centre of each cell. Each well of the plate contains ~ 5000 cells, allowing construction of reliable concentration–response curves and determination of agonist potency (i.e., concentration giving 50 % of the maximal aggregation; EC50). Antagonist potency (i.e., concentration antagonising melatonin aggregation by 50 %; IC50) was determined by pre-incubation of cells with a series of 10-fold concentrations (1 pm to 100 mm) of the analogue 60 min before addition of melatonin (1 nm) and incubation for a further 60 min. before measurement of absorbance. Melatonin at 1 nm was just sufficient to give a full agonist response. Agonist and antagonist concentration–response experiments used nine 10-fold dilutions of analogues (1 pm to 100 mm) and were analysed using nonlinear curve fitting with GraphPad Prism. For the agonist dose–response curve to melatonin, five 10-fold dilutions (1 pm to 10 nm) were used. ChemMedChem 2014, 9, 2238 – 2243
2241
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
All EC50 and IC50 values are the mean of triplicate experiments. The melanophore line was generously provided by Dr. M. R. Lerner (University of Texas, USA). For a small number of the active analogues, the binding affinity against recombinant human MT1 and MT2 subtypes was determined in competition radioligand binding assays using 2-[125-I]-iodomelatonin.[27]
was then extracted with a mixture of MeOH/CH3Cl (5:95; 2 15 mL), and the combined organic phases were washed with water, dried (Na2SO4), and filtered. The solvent was removed in vacuo to leave an oily residue that was purified by flash column chromatography (hexane/EtOAc, 3:7) to give the desired amides 21 a–c as white solids.
Chemistry
Acknowledgements
Melting points were determined on a Bchi melting point apparatus and are uncorrected. NMR spectra were taken in CDCl3 unless otherwise stated. 1H NMR spectra were taken on a Bruker MSL 400 spectrometer, and 13C NMR spectra were recorded on a Bruker AC 200 spectrometer. Chemical shift (d) values are reported in parts per million (ppm), with coupling constants (J) given in Hertz (Hz). Standard multiplicity abbreviations are used: singlet (s), doublet (d), triplet (t), multiplet (m), broad (br). DC-Alufolein plates (Kieselgel 60 F254, Schichtdicke 0.2 mm, Merck) were used for analytical thin-layer chromatography (TLC), and plates were visualized with UV light or developed with iodine. Flash chromatography was performed using Sorbil c60-A silica as the stationary phase. Microanalyses were carried out by the Microanalytical Section of the Institute of Organic & Pharmaceutical Chemistry, National Hellenic Research Foundation, and the results obtained had a maximum deviation of 0.4 % from the theoretical value. Full characterisation data for intermediate and final compounds are given in the Supporting Information. General method for the preparation of acetonitriles 4 a–d: Ketones 3 a–d (1.48 mmol), 1,8-diazabicycloundec-7-ene (DBU) (2.22 mmol) and Ph3P + CH2CNCl (2.07 mmol) were dissolved in toluene (10 mL), and the mixture was heated to reflux for 12 h. The solvent was removed in vacuo, water (15 mL) was added to the residue, and the mixture extracted with EtOAc (2 20 mL). The combined organic extracts were washed with saturated aq NH4Cl (15 mL), dried (Na2SO4), filtered and concentrated in vacuo. Flash column chromatography (hexane/EtOAc, 9:1) gave the desired acetonitrile (4 a–d) as an off-white solid. General method for the preparation of secondary amides (5 a–c; 6 a–c; 7 a–c; 8 a–c; 13 a–c): The appropriate carboxylic acid anhydride (9.59 mmol) and Raney Ni (4.8 mmol, suspension in EtOH, 4 mL) were added to a solution of the appropriate nitrile (0.64 mmol) in dry THF (10 mL), and the mixture was hydrogenated at 55 psi at 60 8C for 8 h. The mixture was then filtered through celite, and the filtrate was concentrated in vacuo. The resulting residue was purified by flash column chromatography (hexanes/ EtOAc 9:1) to give the desired amide as a white solid. General method for the preparation of secondary amides (16 a– c; 20 a–c): A solution of the requisite amine (15 or 19) (0.32 mmol), Et3N (0.44 mmol), and the appropriate carboxylic acid anhydride (0.51 mmol) was cooled to 0 8C and stirred for 5 min. The reaction was then allowed to warm to RT and the stirring continuing for a further 10 min. CH2Cl2 (5 mL) was then added, and the mixture was washed with water (15 mL). The organic phase separated, dried (Na2SO4), filtered, and concentrated. The oily residue was purified by flash column chromatography (hexanes/EtOAc, 3:7) to give the desired amide as a white solid. General method for the preparation of secondary amides (21 a– c): A solution of the requisite amide (20 a, 21 b, 21 c) (0.22 mmol) in TFA (3 mL) was treated with Me3SiH (0.38 mmol), and the mixture was heated at 65 8C for 2 h. The solvent was evaporated in vacuo, the residue obtained was redissolved in water (5 mL), and the solution was brought to pH 8 using 2 n aq NaOH. The solution 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Prof. Andrew Tsotinis is indebted to Drs. D.-H. Caignard and P. Delagrange of the Institut de Recherches Servier (Croissey sur Seine, France) for the radioligand binding experiments. Keywords: benzocycloalkanes · drug design · melatoninergic action · melatonin receptors · structure–activity relationships [1] M. L. Dubocovich, P. Delagrange, D. N. Krause, D. Sugden, D. P. Cardinali, J. Olcese, Pharmacol. Rev. 2010, 62, 343 – 380. [2] D. P. Zlotos, R. Jockers, E. Cecon, S. Rivara, P. A. Witt-Enderby, J. Med. Chem. 2014, 57, 3161 – 3185. [3] S. M. Reppert, D. R. Weaver, C. Godson, Trends Pharmacol. Sci. 1996, 17, 100 – 102. [4] E. L. Bittman, D. R. Weaver, Biol. Reprod. 1990, 43, 986 – 993. [5] S. M. Reppert, D. R. Weaver, T. Ebisawa, Neuron 1994, 13, 1177 – 1185. [6] S. M. Reppert, C. Godson, C. D. Mahle, D. R. Weaver, S. A. Slaugenhaupt, J. F. Gusella, Proc. Natl. Acad. Sci. USA 1995, 92, 8734 – 8738. [7] S. M. Reppert, D. R. Weaver, V. M. Cassone, C. Godson, L. F. Kolakowski, Neuron 1995, 15, 1003 – 1015. [8] S. Fong, Nat. Rev. Drug Discovery 2013, 12, 731. [9] P. J. Garratt, A. Tsotinis, Mini-Rev. Med. Chem. 2007, 7, 1075 – 1088. [10] P. J. Garratt, R. Jones, S. J. Rowe, D. Sugden, Bioorg. Med. Chem. Lett. 1994, 4, 1555 – 1558. [11] P. J. Garratt, S. Travard, S. Vonhoff, A. Tsotinis, D. Sugden, J. Med. Chem. 1996, 39, 1797 – 1805. [12] A. Carocci, A. Catalano, A. Lovece, G. Lentini, A. Duranti, V. Lucini, M. Pannacci, F. Scaglione, C. Franchini, Bioorg. Med. Chem. 2010, 18, 6496 – 6511. [13] a) Y. Hu, M. K. C. Ho, K. H. Chan, D. C. New, Y. H. Wong, Bioorg. Med. Chem. Lett. 2010, 20, 2582 – 2585; b) J. Zhu, Y. Hu, M. K. C. Ho, Y. H. Wong, Xenobiotica 2011, 41, 35 – 45. [14] E. Landagaray, M. Ettaoussi, V. Leclerc, B. Traor, V. Perez, O. Nosjean, J. A. Boutin, D. H. Caignard, P. Delagrange, P. Berthelot, S. Yous, Bioorg. Med. Chem. 2014, 22, 986 – 996. [15] G. Spadoni, A. Bedini, G. Diamantini, G. Tarzia, S. Rivara, S. Lorenzi, A. Lodola, M. Mor, V. Lucini, M. Pannacci, A. Caronno, F. Fraschini, ChemMedChem 2007, 2, 1741 – 1749. [16] P. J. Garratt, S. Vonhoff, S. J. Rowe, D. Sugden, Bioorg. Med. Chem. Lett. 1994, 4, 1559 – 1564. [17] D. J. Davies, P. J. Garratt, D. A. Tocher, S. Vonhoff, J. Davies, M.-T. Teh, D. Sugden, J. Med. Chem. 1998, 41, 451 – 467. [18] S. Copinga, P. G. Tepper, C. J. Grol, A. S. Horn, M. L. Dubocovich, J. Med. Chem. 1993, 36, 2891 – 2898. [19] V. Leclerc, P. Depreux, D. H. Caignard, P. Renard, P. Delagrange, B. Guardiola-Lemaitre, P. Morgan, Bioorg. Med. Chem. Lett. 1996, 6, 1071 – 1076. [20] J. M. Jansen, S. Copinga, G. Gruppen, E. J. Molinari, M. L. Dubocovich, C. J. Grol, Bioorg. Med. Chem. 1996, 4, 1321 – 1332. [21] C. J. Grol, J. M. Jansen, Bioorg. Med. Chem. 1996, 4, 1333 – 1339. [22] M. Math-Allainmat, F. Guady, S. Sicsic, A. L. Dangycaye, S. Shen, B. Bremont, Z. Benatalah, M. Langlois, P. Renard, J. Med. Chem. 1996, 39, 3089 – 3095. [23] a) K. H. Chan, Y. H. Wong, Int. J. Mol. Sci. 2013, 14, 18385 – 18406; b) Z. Wan, F.-F. Zhang, J. Ju, D.-Z. Liu, S. Y. Zhou, B. L. Zhang, Mini-Rev. Med. Chem. 2013, 13, 1462 – 1474. [24] A. Tsotinis, P. A. Afroudakis, Lett. Org. Chem. 2008, 5, 507 – 509. [25] R. Faust, P. J. Garratt, R. Jones, L. K. Yeh, A. Tsotinis, M. Panoussopoulou, T. Calogeropoulou, M. T. Teh, D. Sugden, J. Med. Chem. 2000, 43, 1050 – 1061.
ChemMedChem 2014, 9, 2238 – 2243
2242
CHEMMEDCHEM COMMUNICATIONS [26] D. Sugden, K. Davidson, K. A. Hough, M. T. Teh, Pigm. Cell Res. 2004, 17, 454 – 460. [27] a) P. Schiess, P. V. Barve, F. E. Dussy, A. Pfiffner, Org. Synth. 1995, 72, 116; b) C. J. Bungard, J. C. Morris, Synthesis 2001, 0741 – 0744; c) D.-H. Caignard, P. Delagrange, private communication. [28] a) R. V. Stevens, G. S. Bisacchi, J. Org. Chem. 1982, 47, 2393 – 2396; b) L. S. Liebeskind, L. J. Lescosky, C. M. McSwain, J. Org. Chem. 1989, 54, 1435 – 1439. [29] B. J. Arnold, P. G. Sammes, T. W. Wallace, J. Chem. Soc. Perkin Trans. 1 1974, 415 – 420. [30] N. Negoro, S. Sasaki, S. Mikami, M. Ito, M. Suzuki, Y. Tsujihata, R. Ito, A. Harada, K. Takeuchi, N. Suzuki, J. Miyazaki, T. Santou, T. Odani, N. Kanza-
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org ki, M. Funami, T. Tanaka, A. Kogame, S. Matsunaga, T. Yasuma, Y. Momose, ACS Med. Chem. Lett. 2010, 1, 290 – 294. [31] a) R. Huisgen, G. Seidl, I. Wimmer, Justus Liebigs Ann. Chem. 1964, 677, 21 – 33; b) J. F. Bagli, H. Immer, Can. J. Chem. 1968, 46, 3115 – 3118. These references appertain to the material found in the Supporting Information.
Received: April 11, 2014 Published online on July 8, 2014
ChemMedChem 2014, 9, 2238 – 2243
2243
Benzocyclobutane, Benzocycloheptane and Heptene Derivatives as Melatonin Agonists and Antagonists Andrew Tsotinis,*[a] Pandelis A. Afroudakis,[a] Peter J. Garratt,[b] Alina Bocianowska-Zbrog,[c] and David Sugden[c] Two series of analogues were designed, synthesised and evaluated as potential human melatonin type 1 and 2 receptor (hMT1 and hMT2) ligands. Their biological effects were assessed by a well-established, specific model of melatonin action, the pigment response of Xenopus laevis melanophores. Compounds containing a benzocyclobutane scaffold and a methoxy group in the “melatonin” orientation were found to be potent agonists, with one of the analogues exhibiting activity comparable to melatonin. In contrast, analogues with a methoxy group in non-melatonin positions or with multiple methoxy groups showed either weaker agonist activity or were antagonists. Benzocycloheptene derivatives with one methoxy group are found to be weak agonists, whereas those with two methoxy groups were found to be antagonists, as were all of the benzocycloheptane derivatives evaluated. The most active compounds were assessed in a human receptor radio ligand binding assay but showed little discrimination between MT1 and MT2. These results again show that the indole nitrogen of melatonin is not a necessary component for analogue activity and also illustrate that replacement of the indole ring with a 4membered carbocycle can provide highly active compounds when the methoxy group is in the melatonin position.
brane receptors,[3] and radioligand binding studies using 2[125I]-melatonin (1 b) have revealed a widespread, heterogeneous distribution of binding sites throughout the central nervous system (CNS).[4] Two receptor subtypes have been cloned in mammals (MT1, MT2) that, when expressed in host cells, show the general characteristics of native melatonin receptors.[5, 6] A third receptor subtype, Mel1c, has been cloned from chicken, Xenopus and zebrafish, but has not been detected in mammals.[7] Melatonin has the ability to phase-shift the circadian rhythm and has been shown to be effective in treating insomnia and jet-lag in humans. Two synthetic melatoninergic agonists are approved for use in insomnia (ramelteon) and depression (agomelatine),[2] and a third (tasimelteon) has recently been approved for non-24-hour sleep–wake disorder.[8] A variety of
The pineal hormone melatonin (N-acetyl-5-methoxytryptamine, 1 a)[1, 2] is an important neurohormone in the regulation of seasonal and circadian rhythms. In mammals, the production of melatonin is controlled by a small nucleus within the hypothalamus of the brain—the suprachiasmatic nucleus (SCN), which is the body’s biological clock. The SCN, entrained by light–dark information from the retina, instigates the production and release of melatonin during the dark period and ensures that melatonin levels in the light phase remain very low. The actions of melatonin are believed to be mediated through a family of specific, high-affinity, G protein-coupled cell-mem-
other effects of melatonin administration have been reported and additional physiological roles suggested.[1, 2] The potential of melatoninergic agents in a variety of medical conditions has driven a large number of studies directed towards understanding how melatonin binds to and activates these receptors. A substantial number of both indole and non-indole derivatives have been evaluated.[1, 2, 9] The 5-methoxy group has been shown to be important for receptor binding, but it is not essential for agonist activity,[10] and the indole nitrogen does not serve any obvious binding function and can be omitted in active analogues.[11–14] The indole ring can be replaced by 5- and 6-membered rings,[1, 2] and dibenzocycloheptene derivatives have also been reported.[15] The active conformation of the 3-ethanamine side chain has been studied in conformationally restricted indole[16, 17] and non-indole analogues.[18–22] A large number of analogues selective for the MT2 receptor have been synthesised,[23] and a much smaller number for the MT1 receptor, the latter generally being linked systems or compounds with substantial side chains.[2] Whereas the MT2-selective analogues can be highly discriminatory between the two receptor types, the MT1-selective analogues are much less so.
[a] Prof. A. Tsotinis, Dr. P. A. Afroudakis Department of Pharmaceutical Chemistry Faculty of Pharmacy, School of Health Sciences, University of Athens Panepistimioupoli-Zografou, 157 71 Athens (Greece) E-mail: [email protected] [b] Prof. P. J. Garratt Department of Chemistry, University College London 20 Gordon Street, London WC1H 0AJ (UK) [c] A. Bocianowska-Zbrog, Dr. D. Sugden Division of Women’s Health, School of Medicine, King’s College London Room 2.12N Hodgkin Building, Guy’s Campus, London SE1 1UL (UK) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201402122.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2238 – 2243
2238
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
Here, we report two types of system where the 5-membered pyrrole ring of melatonin has been replaced by either a 4- or 7-membered carbocyclic ring, and we describe the effect that these substitutions have on the receptor binding and biological activities of these molecules. Preparation of the benzocyclobutane precursors followed similar synthetic sequences (Scheme 1). The appropriately substituted bromobenzene was converted to the benzyne with sodium amide and added to 1,1-dimethoxyethene. The benzocyclobutane thus formed was treated with hydrochloric acid to remove the acetal protecting groups and give the corresponding benzocyclobutanone. In the reaction of p-bromoanisole (1 c) with 1,1-dimethoxyethene, two stereoisomers were obtained, which were hydrolysed to the respective benzocyclobutanones (3 c and 3 d). These were separated chromatographically (3 c/3 d, 1/1.5) and then under Wittig reaction conditions (Ph3P + CH2CNCl ) were converted to the corresponding vinyl cyanide, which was then reacted with Raney nickel and the appropriate acid anhydride under hydrogen at 55 psi at 60 8C. The resulting amides were purified by flash column chromatography. Benzocycloheptano 13 a–c and benzocyclohepteno amide derivatives 16 a–c and 20 a–c were prepared by one of two routes from the corresponding and readily available benzo[7]annulen-5-one derivatives. The suitable substituted benzo[7]annule-5-one was either treated with nitromethane to give the nitromethyl derivative or reacted with trimethylsilyl
cyanide to give the corresponding nitrile. Both types of derivative could then be reduced to the corresponding amines, which were then acylated using the appropriate acid anhydride to give the desired amides (Scheme 2). Saturated derivatives 21 a–c were prepared from their corresponding congeners 20 a–c through reduction with a mixture of trifluoroacetic acid (TFA) and trimethylsilane. The results for the pigment aggregation studies for the two series of compounds are reported in Tables 1 and 2. For the benzocyclobutane derivatives (Table 1), the series of compounds 5 a–c and 7 a–c have the methoxy group in the “wrong” orientation to the side chain when compared to melatonin (Figure 1). Nevertheless, both of these series do show some agonist activity. Compounds in series 5 are all partial agonists, and 5 a and b display weak antagonist activity (IC50 ~ 5– 10 mm). No antagonist activity is observed in the series 6, 7 or 8 (IC50 > 100 mm). Compound 5 a (R = Me) has very little agonist intrinsic activity (IA = 11 %) even at high concentrations (100 mm), whereas the compounds with longer R groups (5 b: R=Et; 5 c: R = n-Pr) show improved potency and intrinsic activity (IA = 55 % and 67 %, respectively). The poor activity of 5 a suggests that the alkylamide side chain and the methoxy group cannot occupy receptor binding sites at the same time, whereas the increased potency of 5 b and 5 c suggests that these longer alkyl chains can fit into their binding site at the same time that the methoxy group is in its binding pocket. A similar observation was made in a series of simple phenylalkyamides[11] in which the position of the methoxy group on the phenyl ring was varied. Compound 7 a (R = Me), like 5 a, is a partial agonist, albeit with greater intrinsic activity (78 %), while 7 b (R = Et) and 7 c (R = n-Pr) are full agonists in the melanophore assay. Compound 7 b (R = Et) has the greatest potency in this series and is only 150-fold less potent than melatonin and only sixfold less potent than 8 b, in which the orientation of the methoxy and side chain are the same as in melatonin. Compounds 8 a and 8 c are, like 8 b, full agonists, and a clear increase in potency (~ 16-fold) is found as the amide carbonyl alkyl group is increased from methyl to n-propyl, as we and others have found in analogous series.[2] Series 6 compounds have a methoxy group in the “correct” melatonin orientation like 8 a–c but have an additional methoxy group in the same relative position as the series 5 compounds. Compound 6 a is a partial agonist (IA = 54 %), like the other analogues in which the R group is a methyl (5 a and 7 a); compound 8 a, with the single, correctly orientated methoxy group, is the only acetyl compound with full agonist activity. As occurs in the compound 7 series, changing the R group from methyl to ethyl or propyl in the 6 series gives full agonist activity, although 6 b and 6 c are less potent than their 8 b and 8 c counterparts (23-fold and 130-fold, respectively), suggesting that the second methoxy group interferes with bindScheme 1. Synthetic route to the benzocyclobutane derivatives. Reagents and conditions: ing, possibly by competition with the “correctly” oria) NaNH2, THF, 12 h, reflux; b) THF/HCl, 12 h, room temperature; c) DBU, Ph3P + CH2CNCl , entated methoxy group. Compound 8 c, with ntoluene, reflux, 12 h; d) H2, Raney-Ni, (RCO)2O, 60 8C, 8 h. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2238 – 2243
2239
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
they have not been separated. Compounds 8 b and 8 c bind with high affinity to both the human MT1 and MT2 receptors, showing a small preference for the latter. For the benzocycloheptanes and heptenes (Table 2), the methoxy groups are in the “wrong” orientation compared with melatonin (Figure 1). The main difference in this series is the presence of a double bond in the 7membered ring for compounds of type 16 and 20. The only series showing agonist activity was 16 a–c, with a double bond in the 7-membered ring, although all three compounds are weakly active (EC50 = 1–1.5 mm). As was seen for most of the benzocyclobutane derivatives (Table 1), compounds with a methyl R group were partial agonists, 16 a (IA = 54 %), whereas 16 b (R = Et) and 16 c (R = nPr) were full or nearly full agonists (IA = 102 % and 82 %, reScheme 2. Synthetic route to the benzocycloheptane and heptene derivatives. Reagents and conditions: spectively). Series 16 compounds a) Me3SiCN, ZnI2, 12 h, 40 8C; b) Me3SiCl, NaI, H2O, 6 h, room temperature; c) MeNO2, NH4OAc, 24 h, 110 8C; show little antagonist activity d) LiAlH4, 0.5 h, 0 8C–RT; e) (RCO2)O, Et3N, 15 min, 0.5 h, 0 8C–RT; f) TFA, Me3SiH, 65 8C, 2 h. (< 30 %) even at high concentrations (100 mm). By contrast, series 13 compounds, with no double bond in the 7-mempropyl R group, is comparable in effect to melatonin. Since bered ring, have no agonist activity (up to 100 mm) but all these compounds are mixtures of the R and S stereoisomers, it three analogues, 13 a–c, were full antagonists, blocking melais probable that one isomer is more potent than the other, but tonin aggregation completely although with low potency (IC50 = 2.4–4.9 mm). Compounds 20 a–c and 21 a–c, which contain two methoxy Table 1. Agonist, antagonist melatoninergic action of benzocyclobutanes 5–8 (Xenopus laevis melanophores).[a] groups, were also antagonists but were somewhat weaker (~ 2–7-fold) than series 13 compounds. The introduction of the [a] [b] [a] [b] Compd
R
Melatonin Luzindole[d] 5a Me 5b Et 5c n-Pr 6a Me 6b Et 6c n-Pr 7a Me 7b Et 7c n-Pr 8a Me 8b Et 8c n-Pr
EC50 [nm]
0.10 0.002 NA > 100 000 99 1.5 99 1.2 966 288 70 5.5 118 31 558 51 16.8 1.0 102 1 15.9 2.5 3.0 0.3 0.9 0.0
IA [%]
100 – 11 1 54 1 67 1 54 2 100 2 90 2 78 6 94 2 93 1 103 3 80 1 101 1
IC50 [nm] [c]
NA 2460 460 10 470 884 5590 380 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 > 100 000 NA NA NA
IA [%]
– 100 3 88 1 59 2 45 2 24 6 32 15 1 23 2 15 2 19 2 – – –
[a] Agonist (EC50) and antagonist (IC50) data on melanophores represent the mean SEM of triplicate experiments. [b] IA = intrinsic activity. Agonist—maximum activity relative to melatonin. Antagonist—maximum percentage inhibition of melatonin (1 nm) response. [c] NA = no activity detected at 100 000 nm. [d] Ref. [25].
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. Orientation in space of the methoxy groups and side chain of the new analogues relative to melatonin.
ChemMedChem 2014, 9, 2238 – 2243
2240
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
Table 2. Binding affinities of benzocyclobutanes 8 b and 8 c against human melatonin receptor subtypes hMT1 and hMT2. Compd
hMT1
Ki [nm][a] hMT2
SI[b] (MT2/MT1)
0.66 603 0.8 1.0
0.33 45 0.5 0.4
2 13.4 1.6 2.5
R
Melatonin Luzindole[a] 8b 8c
Et n-Pr
Table 4. Binding affinity of benzocycloheptene analogue 16 b against human melatonin receptor subtypes hMT1 and hMT2. Compd Melatonin Luzindole[a] 16 b
hMT1
Ki [nm][a] hMT2
SI[b] (MT2/MT1)
0.66 603 41
0.33 45 10
2 13.4 4.1
R
Et
[a] Ref. [25]. [b] Selectivity index (SI).
[a] Ref. [25]. [b] Selectivity index (SI).
second methoxy group in series 20 effectively converts the agonist activity of the 16 series into antagonist activity, although it is weaker than that of the saturated series 21. For compounds in both series 20 and 21, a clear increase in antagonist potency was observed on increasing the length of the alkyl chain from methyl to n-propyl (Table 3). Compound 21 c has comparable potency to the melatonin antagonist, luzindole.[24] The most active cycloheptene agonist, 16 b, with COEt as the R group, exhibits significant binding affinity for the human melatonin receptor and shows a small preference for hMT2. It is interesting to note, however, that whereas 16 b binds only 41 times and 25 times less efficiently at hMT1 and hMT2 (Table 4) than 8 b (Table 2), it is approximately 1100 times less effective in the melanophore assay. Cycloheptene 16 b exists as a single isomer, whereas the cyclobutane analogues are a mixture of enantiomers, and one enantiomer may be much more potent than the other, which may distort the comparison of results between the two series of compounds. As Figure 1 indicates, the benzocyclobutane analogues that have the methoxy group in approximately the same position as melatonin are the most potent and mirror melatonin in that
Table 3. Agonist, antagonist melatoninergic action of benzocycloheptanes 13 and 21 and heptenes 16 and 20 with Xenopus laevis melanophores.[a] Compd
R
Melatonin Luzindole[d] 13 a Me 13 b Et 13 c n-Pr 16 a Me 16 b Et 16 c n-Pr 20 a Me 20 b Et 20 c n-Pr 21 a Me 21 b Et 21 c n-Pr
EC50 [nm][a]
IA [%][b]
IC50 [nm][a]
IA [%][b]
0.10 0.002 NA NA NA NA 1540 40 1100 35 1230 100 NA NA NA NA NA NA
100 – 54 2 102 2 82 2 – – – – – – – – –
NA[c] 2460 460 3790 230 4870 100 2410 80 > 100 000 > 100 000 > 100 000 27 700 2800 16 350 970 12 900 3000 10 100 1360 6560 980 3300 150
– 100 3 79 1 73 2 89 1 24 2 82 29 2 87 2 90 1 95 1 95 1 90 1 94 1
[a] Agonist (EC50) and antagonist (IC50) data on melanophores represent the mean SEM of triplicate experiments. [b] IA = intrinsic activity. Agonist—maximum activity relative to melatonin. Antagonist—maximum percentage inhibition of melatonin (1 nm) response. [c] NA = no activity detected at 100 000 nm. [d] Ref. [25].
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
lengthening the N-acetyl group to N-n-propyl increases the activity. In the case of benzocyclobutanes 7 a–c, where the methoxy group is in the “wrong” orientation with regard to the side chain, potency is at a maximum with N-ethyl, probably indicating that the n-propyl side chain further displaces the methoxy group from its binding site, although this derivative is still more potent than the N-acetyl compound. In conclusion, as previously observed, the indole nitrogen is not a necessary component of a melatonin analogue and the substitution of a 4-membered carbocyclic ring provides highly active compounds when these have a methoxy group in the melatonin position. Separation of the enantiomers of 8 c should be of interest in determining whether they have different activities, as should the introduction of an exocyclic double bond in the benzocyclobutane analogues, since this has a significant effect in increasing the potency of the benzocycloheptene derivatives.
Experimental Section Biology The biological activity of the analogues was determined in a specific model of melatonin action, the pigment aggregation response of Xenopus laevis melanophores, which we described previously.[26] In these cells, melatonin triggers a dramatic movement of melanin pigment, normally distributed throughout the cell, towards the cell centre. This response is termed pigment aggregation and can be accurately quantified in melanophores grown in 96-well plates simply by measuring the change in light (630 nm) absorbance of the cells as the pigment concentrates at the centre of each cell. Each well of the plate contains ~ 5000 cells, allowing construction of reliable concentration–response curves and determination of agonist potency (i.e., concentration giving 50 % of the maximal aggregation; EC50). Antagonist potency (i.e., concentration antagonising melatonin aggregation by 50 %; IC50) was determined by pre-incubation of cells with a series of 10-fold concentrations (1 pm to 100 mm) of the analogue 60 min before addition of melatonin (1 nm) and incubation for a further 60 min. before measurement of absorbance. Melatonin at 1 nm was just sufficient to give a full agonist response. Agonist and antagonist concentration–response experiments used nine 10-fold dilutions of analogues (1 pm to 100 mm) and were analysed using nonlinear curve fitting with GraphPad Prism. For the agonist dose–response curve to melatonin, five 10-fold dilutions (1 pm to 10 nm) were used. ChemMedChem 2014, 9, 2238 – 2243
2241
CHEMMEDCHEM COMMUNICATIONS
www.chemmedchem.org
All EC50 and IC50 values are the mean of triplicate experiments. The melanophore line was generously provided by Dr. M. R. Lerner (University of Texas, USA). For a small number of the active analogues, the binding affinity against recombinant human MT1 and MT2 subtypes was determined in competition radioligand binding assays using 2-[125-I]-iodomelatonin.[27]
was then extracted with a mixture of MeOH/CH3Cl (5:95; 2 15 mL), and the combined organic phases were washed with water, dried (Na2SO4), and filtered. The solvent was removed in vacuo to leave an oily residue that was purified by flash column chromatography (hexane/EtOAc, 3:7) to give the desired amides 21 a–c as white solids.
Chemistry
Acknowledgements
Melting points were determined on a Bchi melting point apparatus and are uncorrected. NMR spectra were taken in CDCl3 unless otherwise stated. 1H NMR spectra were taken on a Bruker MSL 400 spectrometer, and 13C NMR spectra were recorded on a Bruker AC 200 spectrometer. Chemical shift (d) values are reported in parts per million (ppm), with coupling constants (J) given in Hertz (Hz). Standard multiplicity abbreviations are used: singlet (s), doublet (d), triplet (t), multiplet (m), broad (br). DC-Alufolein plates (Kieselgel 60 F254, Schichtdicke 0.2 mm, Merck) were used for analytical thin-layer chromatography (TLC), and plates were visualized with UV light or developed with iodine. Flash chromatography was performed using Sorbil c60-A silica as the stationary phase. Microanalyses were carried out by the Microanalytical Section of the Institute of Organic & Pharmaceutical Chemistry, National Hellenic Research Foundation, and the results obtained had a maximum deviation of 0.4 % from the theoretical value. Full characterisation data for intermediate and final compounds are given in the Supporting Information. General method for the preparation of acetonitriles 4 a–d: Ketones 3 a–d (1.48 mmol), 1,8-diazabicycloundec-7-ene (DBU) (2.22 mmol) and Ph3P + CH2CNCl (2.07 mmol) were dissolved in toluene (10 mL), and the mixture was heated to reflux for 12 h. The solvent was removed in vacuo, water (15 mL) was added to the residue, and the mixture extracted with EtOAc (2 20 mL). The combined organic extracts were washed with saturated aq NH4Cl (15 mL), dried (Na2SO4), filtered and concentrated in vacuo. Flash column chromatography (hexane/EtOAc, 9:1) gave the desired acetonitrile (4 a–d) as an off-white solid. General method for the preparation of secondary amides (5 a–c; 6 a–c; 7 a–c; 8 a–c; 13 a–c): The appropriate carboxylic acid anhydride (9.59 mmol) and Raney Ni (4.8 mmol, suspension in EtOH, 4 mL) were added to a solution of the appropriate nitrile (0.64 mmol) in dry THF (10 mL), and the mixture was hydrogenated at 55 psi at 60 8C for 8 h. The mixture was then filtered through celite, and the filtrate was concentrated in vacuo. The resulting residue was purified by flash column chromatography (hexanes/ EtOAc 9:1) to give the desired amide as a white solid. General method for the preparation of secondary amides (16 a– c; 20 a–c): A solution of the requisite amine (15 or 19) (0.32 mmol), Et3N (0.44 mmol), and the appropriate carboxylic acid anhydride (0.51 mmol) was cooled to 0 8C and stirred for 5 min. The reaction was then allowed to warm to RT and the stirring continuing for a further 10 min. CH2Cl2 (5 mL) was then added, and the mixture was washed with water (15 mL). The organic phase separated, dried (Na2SO4), filtered, and concentrated. The oily residue was purified by flash column chromatography (hexanes/EtOAc, 3:7) to give the desired amide as a white solid. General method for the preparation of secondary amides (21 a– c): A solution of the requisite amide (20 a, 21 b, 21 c) (0.22 mmol) in TFA (3 mL) was treated with Me3SiH (0.38 mmol), and the mixture was heated at 65 8C for 2 h. The solvent was evaporated in vacuo, the residue obtained was redissolved in water (5 mL), and the solution was brought to pH 8 using 2 n aq NaOH. The solution 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Prof. Andrew Tsotinis is indebted to Drs. D.-H. Caignard and P. Delagrange of the Institut de Recherches Servier (Croissey sur Seine, France) for the radioligand binding experiments. Keywords: benzocycloalkanes · drug design · melatoninergic action · melatonin receptors · structure–activity relationships [1] M. L. Dubocovich, P. Delagrange, D. N. Krause, D. Sugden, D. P. Cardinali, J. Olcese, Pharmacol. Rev. 2010, 62, 343 – 380. [2] D. P. Zlotos, R. Jockers, E. Cecon, S. Rivara, P. A. Witt-Enderby, J. Med. Chem. 2014, 57, 3161 – 3185. [3] S. M. Reppert, D. R. Weaver, C. Godson, Trends Pharmacol. Sci. 1996, 17, 100 – 102. [4] E. L. Bittman, D. R. Weaver, Biol. Reprod. 1990, 43, 986 – 993. [5] S. M. Reppert, D. R. Weaver, T. Ebisawa, Neuron 1994, 13, 1177 – 1185. [6] S. M. Reppert, C. Godson, C. D. Mahle, D. R. Weaver, S. A. Slaugenhaupt, J. F. Gusella, Proc. Natl. Acad. Sci. USA 1995, 92, 8734 – 8738. [7] S. M. Reppert, D. R. Weaver, V. M. Cassone, C. Godson, L. F. Kolakowski, Neuron 1995, 15, 1003 – 1015. [8] S. Fong, Nat. Rev. Drug Discovery 2013, 12, 731. [9] P. J. Garratt, A. Tsotinis, Mini-Rev. Med. Chem. 2007, 7, 1075 – 1088. [10] P. J. Garratt, R. Jones, S. J. Rowe, D. Sugden, Bioorg. Med. Chem. Lett. 1994, 4, 1555 – 1558. [11] P. J. Garratt, S. Travard, S. Vonhoff, A. Tsotinis, D. Sugden, J. Med. Chem. 1996, 39, 1797 – 1805. [12] A. Carocci, A. Catalano, A. Lovece, G. Lentini, A. Duranti, V. Lucini, M. Pannacci, F. Scaglione, C. Franchini, Bioorg. Med. Chem. 2010, 18, 6496 – 6511. [13] a) Y. Hu, M. K. C. Ho, K. H. Chan, D. C. New, Y. H. Wong, Bioorg. Med. Chem. Lett. 2010, 20, 2582 – 2585; b) J. Zhu, Y. Hu, M. K. C. Ho, Y. H. Wong, Xenobiotica 2011, 41, 35 – 45. [14] E. Landagaray, M. Ettaoussi, V. Leclerc, B. Traor, V. Perez, O. Nosjean, J. A. Boutin, D. H. Caignard, P. Delagrange, P. Berthelot, S. Yous, Bioorg. Med. Chem. 2014, 22, 986 – 996. [15] G. Spadoni, A. Bedini, G. Diamantini, G. Tarzia, S. Rivara, S. Lorenzi, A. Lodola, M. Mor, V. Lucini, M. Pannacci, A. Caronno, F. Fraschini, ChemMedChem 2007, 2, 1741 – 1749. [16] P. J. Garratt, S. Vonhoff, S. J. Rowe, D. Sugden, Bioorg. Med. Chem. Lett. 1994, 4, 1559 – 1564. [17] D. J. Davies, P. J. Garratt, D. A. Tocher, S. Vonhoff, J. Davies, M.-T. Teh, D. Sugden, J. Med. Chem. 1998, 41, 451 – 467. [18] S. Copinga, P. G. Tepper, C. J. Grol, A. S. Horn, M. L. Dubocovich, J. Med. Chem. 1993, 36, 2891 – 2898. [19] V. Leclerc, P. Depreux, D. H. Caignard, P. Renard, P. Delagrange, B. Guardiola-Lemaitre, P. Morgan, Bioorg. Med. Chem. Lett. 1996, 6, 1071 – 1076. [20] J. M. Jansen, S. Copinga, G. Gruppen, E. J. Molinari, M. L. Dubocovich, C. J. Grol, Bioorg. Med. Chem. 1996, 4, 1321 – 1332. [21] C. J. Grol, J. M. Jansen, Bioorg. Med. Chem. 1996, 4, 1333 – 1339. [22] M. Math-Allainmat, F. Guady, S. Sicsic, A. L. Dangycaye, S. Shen, B. Bremont, Z. Benatalah, M. Langlois, P. Renard, J. Med. Chem. 1996, 39, 3089 – 3095. [23] a) K. H. Chan, Y. H. Wong, Int. J. Mol. Sci. 2013, 14, 18385 – 18406; b) Z. Wan, F.-F. Zhang, J. Ju, D.-Z. Liu, S. Y. Zhou, B. L. Zhang, Mini-Rev. Med. Chem. 2013, 13, 1462 – 1474. [24] A. Tsotinis, P. A. Afroudakis, Lett. Org. Chem. 2008, 5, 507 – 509. [25] R. Faust, P. J. Garratt, R. Jones, L. K. Yeh, A. Tsotinis, M. Panoussopoulou, T. Calogeropoulou, M. T. Teh, D. Sugden, J. Med. Chem. 2000, 43, 1050 – 1061.
ChemMedChem 2014, 9, 2238 – 2243
2242
CHEMMEDCHEM COMMUNICATIONS [26] D. Sugden, K. Davidson, K. A. Hough, M. T. Teh, Pigm. Cell Res. 2004, 17, 454 – 460. [27] a) P. Schiess, P. V. Barve, F. E. Dussy, A. Pfiffner, Org. Synth. 1995, 72, 116; b) C. J. Bungard, J. C. Morris, Synthesis 2001, 0741 – 0744; c) D.-H. Caignard, P. Delagrange, private communication. [28] a) R. V. Stevens, G. S. Bisacchi, J. Org. Chem. 1982, 47, 2393 – 2396; b) L. S. Liebeskind, L. J. Lescosky, C. M. McSwain, J. Org. Chem. 1989, 54, 1435 – 1439. [29] B. J. Arnold, P. G. Sammes, T. W. Wallace, J. Chem. Soc. Perkin Trans. 1 1974, 415 – 420. [30] N. Negoro, S. Sasaki, S. Mikami, M. Ito, M. Suzuki, Y. Tsujihata, R. Ito, A. Harada, K. Takeuchi, N. Suzuki, J. Miyazaki, T. Santou, T. Odani, N. Kanza-
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org ki, M. Funami, T. Tanaka, A. Kogame, S. Matsunaga, T. Yasuma, Y. Momose, ACS Med. Chem. Lett. 2010, 1, 290 – 294. [31] a) R. Huisgen, G. Seidl, I. Wimmer, Justus Liebigs Ann. Chem. 1964, 677, 21 – 33; b) J. F. Bagli, H. Immer, Can. J. Chem. 1968, 46, 3115 – 3118. These references appertain to the material found in the Supporting Information.
Received: April 11, 2014 Published online on July 8, 2014
ChemMedChem 2014, 9, 2238 – 2243
2243
