8/3/2019 Mark R. Segge et al- A Structure-Affinity Study of the Binding of 4-Substituted Analogues of 1-(2,5-Dimethoxyphen

1/5

1032 J. Med. Chem. 1990,33, 1032-1036A Structure-Affinity Study of the Binding of 4-Substituted Analogues of1-(2,5-Dimethoxyphenyl)-2-aminopropanet 5-HT2Serotonin ReceptorsMark R. Segge1,t Mamoun Y. Yousif,t Rob ert A. Lyon,**sMilt Titeler,t Bryan L. Roth,lI.l Eva A. Suba,"and Richard A. Glennon*JDepar tment of Medicinal Chemistry, School of Pharm acy, M edical College of Virginia, Virginia C ommonw ealth University,Richmond, Virginia 23298-0581, Department of Pharm acolog y and Toxicology, Albany M edical College,Albany, New York 12208, and N aval Med ical Research In sti tute, Bethesda, Ma ryland 20814. Received November 11, 1988

With [SH]ketanserin as he radioligand, str uctu reaf fini ty relationships (SAFIRs)for binding at central 5HTzserotoninreceptors (r at frontal cortex) were exam ined for a series of 27 4-substituted 1-(2,5dimethoxypheny1)-2-aminopropanederivatives (2,5-DMAs). Th e affinity (Ki values) ranged over a spa n of several orders of magnitude. It appearsth at th e lipophilic charac ter of th e 4-position s ubs tituen t plays a major role in determ ining the affinity of theseagents for 5-HTz eceptors, 2, 5 -D M s with polar 4-substituents (e.g. OH , NHz,COOH) display a very low affinity(Ki> 25 000 nM ) for these recep tors, whereas those with lipophilic functions disp lay a significantly higher affinity.Th e results of these stud ies promp ted us t o synthesize and evalu ate examples of newer lipophilic derivatives andsevera l of these (e.g. n-hexyl, n-octyl) bind w ith very high (Ki alues = 2.5 an d 3 nM , respectively) affinities a t central5-HT2sites. Althou gh, 2,5-DMAs are generally conside red to be 5-HTz gonists, preliminary s tudies w ith isolatedrat thoracic aorta suggest that some of the more lipophilic derivatives (e.g. the n-hexyl and n-octyl derivatives) are5-HT2antagonists.Cen tral serotonin (5-H T) receptors can be categorizedas belonging to one of thr ee ma jor classes: 5-HT1 , 5-H T2,and 5-HT3,192 Serotonin itself is nonselective and bindsa t each of these neurotran smitter sites. One of th e goalsof our work is to develop site-selective 5- H T agonists andantagonists an d, to this extent, we have demonstrated tha tcertain 4-substituted 1-(2,5-dimethoxyphenyl)-2-amino-propane (i.e., 2,5-DMA) derivatives bind a t 5-HT 2 eceptorsites.3 We have also found th at the affinity and selectivityof these agents are significantly influenced by the n atureof the 4-substituent. For example, 2,5-DMA (la) bindsa t 5 -HT2 sites with little selectivity and with ra ther lowaffinity (Ki = 5200 nM ), whereas its 4-bromo derivativeDOB (lg)binds selectively and with considerably greateraffinity (K i= 40 nM).'t* Indee d, [3H]D OB is now com-mercially available as a radioligand for use in bindingstudies. In order to determin e th e influence of the 4-

substitue nts on binding a t 5-HT 2 sites, several years agowe conducted a Hansch analysis on a series of 13 deriva-tives of 2,5-DMA th at varied in str uctu re only at th e 4-p ~ s it io n . ~ reliminary results suggested th at l ipophilicsubst ituen ts are imp ortan t for affinity. We have nowprepared an d examined a total of 27 4-substituted deriv-atives of 2,5-DMA with a broa der ran ge of substitu ents inorder to further challenge this hypothesis. We report herethe synthesis and 5-HT 2binding data for the new com-pounds and the structure-affinity relationships (SAFIR)for the e ntire series. We have also found th at some of thenew, more lipophilic 2,5-DMA derivatives act as 5-HT2antagonists.Chemistry

A total of 21 2,5-DMA deriva tives were examine d in th epresent study; some of these compounds (i.e., la-e,g-k,-m-o) re the 13 included in the original study an d theirsyntheses have been previously Th e 4-chloroderivative If was synthesized according to th e metho d ofCoutts and Malicky.s Amine I t was prepared by catalyticf Virginia Comm onwealth University.t Albany Medical College.*P res ent address: Miami Valley Labs, Proctor and Gamble11 Naval Medical Research Institute.

University Medical School, Stanford, CA 94305.Co., Cincinnati, OH .

Present address: Department of Psychiatry, Stanford

Table I. Structures and 5-HTz Binding Data for1-( ,5-Dimethoxypheny1)-2-aminopropanes"

c?$H3 R

R Ki, nM R Ki, nMla H 5200 lo Am 7 (1)lb OMe 1250 l p Hex 2.5 (0.4)IC OEt 2200 Iq Oct 3.0 (0.2)Id NO2 300 l r Ph-Pr 10 (1)le F 1100 Is Bz 7.0 (0.8)If C1 218 (18) I t NH2 26000 (900)l g B r 41 (5) l u CN 2400 (180)l h I 19 lv COOPr 2460 (90)l i Me 100 lw COOBu 1530 (125)l j Et 100 l x COEt 735 (75)l k Pr 69 1Y CONHPr 7550 (675)11 i -Pr 76 (9) l z OH > 5 mlm Bu 58 (6) laa COOH >5oooOIn t-Bu 19 (3) ketanserin 1.2 (0.2)a Affinities ( Ki values) for [3H]ketanserin-labeled 5 -HT 2 sites.Some of th e Ki values were reported Ki values are fol-lowed by i S E M only for those data not previously published.SEM not determined for l z and laa.

reduction of the corresponding 4-nitro compound by aliterature procedure.*Compounds lp-s were all prepared in a similar manner,i.e., Friedel-Crafts acylation of N-trifluoroacetyl-2,5-DMA(2 ) followed by catalytic reduction an d deprotection. Forexample, acylation of 2 with hexanoyl chloride in thepresence of TiCl., afforded ketone 3a, which was subse-Glennon, R. A. J. Med. Chem. 1987,30, 1.Fozard, J. Trends Pharmacol. Sei. 1987,8, 501.Shan non, M.; B attaglia, G.; Glennon, R. A,; Titeler, M. Eur.J. Phar maco l. 1984, 102, 23.Glennon, R. A.; McKenney, J. D.; Titeler, M. Life Sci. 1984,35, 2505.Seggel, M.; Yousif, M.; T itele r, M.; Lyon, R. A.; Glennon, R.A. Va. J. Sci. 1986, 37, 122.Glennon. R. A,: Liebowitz. S. M.: Anderson. G. M. J . Med.Chem. 1980,23; 294.Glennon, R. A.; Young, R.; Benington, F.;Morin, R. D. J.Med.Chem. 1982,25, 1163.Coutts, R. T.; Malicky, J. L. Can. J. Chem. 1973, 51 , 1402.

0022-2623/90/1833-lO32$02.50/0 0 1990 American Chemical Society

8/3/2019 Mark R. Segge et al- A Structure-Affinity Study of the Binding of 4-Substituted Analogues of 1-(2,5-Dimethoxyphen

2/5

Analogues of 1 (2,5-Dimethoxyphenyl)-2-aminopropanequently reduced to 4a and deprotected to give lp . Nitrilelu was prepared by tre atm ent of an N-phthalimido-pro-tected 4-bromo deriv ative of 2,5-DMA (i.e., 5) with cuprouscyanide followed by deprotection with hydrazine. Estersl v an d lw were obtained by esterification of acid laa, an damide l y was obtained by treatment of l v with n-propylamine . Although 4-OH analogue l z has beenmentioned in th e literature on several ocassions (e.g. seeref 9) , there is no evidence of its preparation or charac-terization. Attempts to prepare l z via hydrolysis of thediazonium salt generated from N -acetyl It were unsuc-cessful. However, 12 was prepared in low yield by Baey-er-Villiger oxid ation of th e 4-formyl deriva tive of 5 fol-lowed by d eprotection.Results

Radioligand binding data for the entire set of 27 4-substituted 2,5-DMA analogues are shown in Table I; Kivalues range from 2.5 to >50000 nM. Substitution a t the4-position by p olar/hydrophilic sub stituents such asNHz,OH, an d COOH (i.e., lt,z,aa, respectively) result in com-pounds with a lower affinity (Ki> 25 000 nM) for 5-HTzsites tha n th at of th e 4-unsubstituted 2,5-DMA itself (la,Ki = 5200 nM ). Com pounds with hydrophobic substitu-ents display high affinity for 5-H Tz sites. Th e 4-amyl (Ki= 7 nM ), 4-hexyl (K i = 2.5 nM ), 4-octyl (Ki = 3 nM),4-[3-(pheny1)propyll (Ki= 10nM ), and 4-benzyl (K i= 7nM ) d erivatives of 2,5-DMA (i.e., derivatives 10-8, re-spectively) all possess affinities (Kivalues) for the 5-HTzsites of 10 nM or less (Table I). Surprisingly, esters l v andlw , though fairly lipophilic (Csubstituen t ir values = 1.07an d 1.62, respectively),'O displa y a low affinity for 5-H Tzsites.Evaluation of 2,5-DMA Analogues as PotentialAntagonists. Overall, it seems th at affinity is explainedprimarily by th e lipophilicity of th e 4-substituent." Bu t,previous studies have shown that certain of these com-pounds behave as agonists whereas others do not (see ref1 2 for a review). For exam ple, using rats trained t o dis-crimin ate DOM from saline, ED,, values for stimulusgeneralization a re significantly correlated with 5-HT z af-finity. However, some of the com pounds included in th epresent st ud y (e.g., In and lo) possess a high affinity forthe 5-HTzsites but do no t result in stimu lus generalization.If the lipophilic contribution of the 4-sub stituent is pre-dictive of affinity a nd if all agents with high affinity ar eno t necessarily agonists, the possibility exists th at som eof the se agents m ight serve as antagonists (or at least aspartial a gonists). We examined this possibility in an iso-lated tissue assay. In ra t thoracic aorta, serotonin receptors

J o u r n a l of Medicinal Chemistry, 1990, Vol. 33, N o. 3 1033

(9) Anderson, G. M.; Castagnoli, N.; Kollman, P. A. NZDA Res.Monog. 1978,22, 199.(10) Hansch, C.; Leo, A. Substituent Constants for CorrelationAnalysis in Chemistry and Biology; John Wiley and Sons:New York, 1979.(11) A qua ntita tive structure-activity relationship analysis of th ebinding d ata shown in Table I has been conducted. The re isa significant correlation between 5-H T2affinity (Kivalues) andthe lipophilicity ( A values) of the 4-position subs tituents (n =25, r = 0.858; the 4-OH and 4-COOH derivatives l z and la a,respectively, could not be included in t he analysis due to theirlow affinity and lack of specific Ki values). If esters lv and lware neglected, for reasons discussed in the text, the correlationbetween affinity an d A values is still significant (n = 23, r =0.903). For more detail, see: Glennon, R. A,; Seggel, M. R. InProbing Bioactiue Mechanisms; Magee, P. S., Henry, D.,Block, J., Eds.; American Chem ical Society: Washington, DC ,1989; pp 264-280.(12) Glennon, R. A. In Tra nsdu ction Mechanisms of Drug Stimuli;Colpaert, F. C., Balster, R. L., Eds.; Springer -Verlag: Berlin,1988; pp 16-31.

120----I20 I

d0 100 10001201 1 120 j

U2 40

0.1

j;i00 IM1 10 I ~ i I ) l O O O 1 10 100 1000

5H T [ ~ ~ l ~ ~ l ] 5HT [ p M ]Figure 1. Antagonism of 5-HT-induced contractions of ra tthoracic aorta by the 4-benzyl (1s; A) , 4-hexyl (lp; B), 4-phenylpropyl (lr;C) , and 4-octyl (lq;D) erivatives of 2,5-DMA.%'appear to be of th e 5-HT z ype.13J4 As with central 5-H Tzsites, 5-H T activates phosphoinositol turnover in ra t ao rtal6and the physiological consequences of this 5-HT z receptoractivation appear related to tissue ont traction.'^ I t hasbeen proposed th at ao rtic receptors provide an appropriatemodel system for studying 5 -HT z receptor interactions;14J5for example, racemic DOB (lg;Ki = 41 nM) is a potentfull agonist in this system an d (R)-(-)-DOB (Ki= 24 nM)16is approximately twice as poten t as its racemate." In thepresent stu dy, we examined four agents suspected of beingpotential a ntagonists (i.e., lp-s). Preliminary evaluations(da ta not shown) revealed no consistent agon ist activity.In contrast, increasing concentrations of all four com-pounds shifted the conc entration-response curves of 5-H Tto the right a nd resulted in depressed maximal responses(Figure 1). Thus, all four comp ounds displayed antagonistcharacter; however, it appears th at the antagonism maybe of a noncom petitive nature. Because the 5-HTz an -tagonist ketanserin (PA, = 8.5) seems to produce a com-petitive antagonism,14 he present d ata m ight reflect dif-ferences between central and peripheral 5-H T2receptorsor differences in the natu re of the interactions. Never-theless, it is clear that lp-s behave as antagonists, notagonists, in this preparation.DiscussionOur preliminary studies suggested t ha t th e 5-H Tz re-ceptor affinity of 2,5-DMA analogues is related to thelipophilicity of th e 4-sub stituent and th at a s lipophilicity(13) Cohen, M.; Fuller, R. W.; Wiley, K. S. J. Pharmacol. Erp.Ther. 1981, 218, 421.(14) (a) Roth, B. L.; Nakaki, T.; Chuang, D. M.; Costa, E. Neuro-pharmacology 1984, 23, 1223. (b) Roth, B. L.; Nakaki, T.;Chuang, D. M.; Costa, E. J.Pharmacol. E r p . Ther. 1986,238,480; (c) Sub a, E. A.; Roth, B. L. Eur. J. Pharmacol. 1987,136,325.(15) (a) Roth, B. L.; Chuang, D. M. Life Sci. 1987, 41, 1051. (b)Roth, B. L.; McLean, S.; Zhu, X.-Z.; Chard, D. M. J . Neuro-chem. 1987,49, 1833.(16) Glennon, R. A.; Titeler, M.: Seggel, M. R.; Lyon, R. A. J.Med.-Chem. 1 987,30, 930.(17) Suba. E. A.: Baraban. J.: Sengel, M. R.: Glennon, R. A.; Roth,. I - .B. L.; manuscript in preparation.

8/3/2019 Mark R. Segge et al- A Structure-Affinity Study of the Binding of 4-Substituted Analogues of 1-(2,5-Dimethoxyphen

3/5

1034 Journal of Medicinal Chemistry, 2990,Vol. 33 , N O .3increa ses, affinity increase .^ Although th e initial seriesof com pounds was small ( n= 13), the results in Table Iare consistent with these prelim inary findings in that thosecompounds with th e highest affinity are those with lipo-philic 4-substituents. On th e other hand, esters lv andlw , which are alsoquite lipophilic, possess a relatively lowaffinity. Th is low affinity might be explained by th e ad-jacency of th e 4-position carbonyl group to th e 5-positionmethoxy group (i.e., a possible alteration in s ubstituen torientation might result from an electrostatic repulsionbetween the two oxygen atoms), or it may be a result ofpartial hydrolysis under th e conditions of the binding m a yto t he low-affinity 4-COOH derivative laa. Preliminarydat a showed tha t acid laa was of low affinity (i.e., Ki >5000 nM ); however, subsequent examination of this agentat higher concentrations revealed that the acid displaysminimal affinity even a t concentrations of up to 50000 nM .Hansch analysis of th e binding da ta show th at the re is asignificant correlation between affinity an d the lipophilicityof th e 4-position substituen t.llOther th an for our preliminary studies, an analysis ofthe role of the 4-position substituents of these types ofagents on 5-HT , receptor affinities has not been previouslyreported. However, we have demonstrated th at th e hal-lucinogenic potencies of phenylalkylamines are signifi-cantly correlated with their affinities for 5- HT z eceptorsand that such agents are most likely acting as 5-HT2agonist^.^ In 1975, Bar fknec ht an d co-workers'* repo rtedthat the octanol/water partition coefficients of phenyl-alkylamines may be an im portan t, though not necessarilyexclusive, determinant of their hallucinogenic potency.Using a somewhat larger da ta set, we later reported tha tthe overall lipophilicity of these agents might be important,bu t th at (a) by itself, lipophilicity could no t account forthe potencies of al l of these agents, and (b) that the overalllipophilicities of these agents m ight sim ply reflect the largelipophilic contribution of th e 4-position substituents of th emore potent agents.lg Furtherm ore, we provided evidencefor a direct receptor interaction of the 4-position sub-s t i t u e n t ~ . ' ~hulgin an d D yerz0also found, for a limitedset of 4-substituted 2,5D MA s, th at hallucinogenic potencycould be explained on th e basis of th e lipophilicity of the4-position subs titu ent . Several other studies (e.g. ref21-24) have also addr essed t he possible relationships be-tween the activities of related agents and serotonin re-ceptor interactions; however, because th e 5- HT receptorsinvolved in those studies may no t be of th e 5-H T2 ype,th e significance of thes e results is unknown.This study represents the first comprehensive investi-gation of the structure-affinity requirements for thebinding of 4-substituted phenylalkylamine derivatives at5-H Tz receptors. I t was determined th at th e lipophiliccharacter of the 4-position substituents of the 2,5-DMAsappears to be impo rtant in determining their a ffinity for

Seggel et al .central 5-HT 2 receptors. It might be hypothesized th atthe receptors possess a hydrophobic area in this region tha tcan accomm odate these 4-substituents. Th e nature of thishydrophob ic site needs to be investigated in greater detail,as does its ramifications for receptor selectivity. Never-theless, it is apparen t tha t not all agents with high affinityfor these sites behave as agonists. For example, in ratthoracic aorta, DOB (lg) is a potent 5-HT, agonist"whereas compounds 1p-s act as antagonists. Furthermore,the affinities of some of these agents at [3H]ketanserin-labeled 5-HT, sites rivals th at of the currently most pop-ular 5-HT, antagonist ketanserin (Ki = 1.2 nM) and areseveral orders of magnitude greater th an th at of serotonin(K i 500 nM).' Th us, th e possibility exists for the de-velopment of an entirely new stru ctu ral class of 5-HT2antagonists. Because [3H]ketanserin binds at neurotran-sm itter sites other than 5 -HT sites' and because it ap-parently binds to a tetrabenazine-sensitive site in brainand in the periphery,15b new str uct ura l class of 5-HTzantagonists may not s hare some of these disadv antages ofketanserin.Experimental Section

Melting points were determined on a Thomas-Hoover meltingpoint appa ratu s and are uncorrected. Microanalysis were per-formed by Atlantic Microlab (Atlanta, GA) nd determined valuesare within 0.4% of theoretical values. Pro ton NM R spectra wererecorded on a JEOL FXSOQ spectrometer, operating at 90MHzand using tetramethylsilane as an internal standard. Infraredspectra were obta ined on a Nicolet 5ZDX FT-IR spectropho-tometer. All spec tral da ta are consistent with the assignedstructures. Th e 4-isopropyl compound 11 as ita HCl sa lt, was agift from Drs. F. Benington and R. D. Morin (NeurosciencesProgram, Department of Psychiatry, University of Alabama,Birmingham, AL).l-(2,5-Dimethoxy-4-hexylphenyl)-2-aminopropaney-drochloride (lp). A suspension of 4a 0.15g, 0.4 mmol) in 15%aqueous NaOH ( 5 mL) , HzO ( 5 mL), and MeOH (10 mL) washeated a t reflux for 2.5 h. Th e solution was allowed to cool toroom tem perature and was extracted with Eh O (3 X 20mL). Thecombined extracts were dried (Na zS04 )an d ethere al HC1 (ca. 20mL) was added. After removal of the solvent unde r reducedpressure, the residue was recrystallized from EtO H/ Eh O to give0.09 g (71% ) of the title compound: m p 133-135 "C. Anal.

1-(2,5-Dimethoxy-4-octylphenyl)-2-aminopropaneydro-chloride (lq). Compound lq was prepared from 4b n a mannersimilar to that used for lp. Th e residue was recrystallized fromEtOH/EhO to give 78% of compound lq: m p 138-140OC. Anal.l-[2,5-Dimethoxy-4-(3-phenylpropyl)phenyl]-2-amino-propane Hydrochloride (lr). Compound lr was prepared from4c n a manner similar to that used for lp.Excess HCl an d solventwere removed in vacuo and th e residue was recrystallized fromEt OH /E hO to give 81% of the title compound: mp 160-162 "C.Anal. (Cz0Hz7NOz~HC1), H, N.1-(4-Benzyl-2,5-dimethoxyphenyl)-2-aminopropaney -drochloride (Is). Compound 1s was prepared from 4d in a

manner similar to tha t used for lp. Removal of the EtOH underreduced pre ssure left a wh ite solid which was recrystallized fromEtO H/E t20 , y ie ld ing 85% of th e title compound: mp 181-183"C. Anal. (C,8H23N02.HC1), C, H, N.1-(4-Cyan0-2,5-dimethoxyphenyl)-2-aminopropaney-drochloride (lu). The title compound was prepared by usingthe general procedure described by Cheng and Castagnoli.25Cup rous cyanide (0.61 g, 1.8mmol) and 5 (0.63 g, 1.57 mmol) w ereheated at reflux in DMF (10 mL) for 5 h. After cooling, th ereaction mixt ure was poured in to a solution of FeCl,-H,O (0.48g) in aqueou s HCl (0.12 mL of concentrated HCl in 0.72 mL ofHzO). Th e mixture was heated to 60 "C a nd allowed to cool. Afterdilution with HzO (50 mL ) and extraction with CHzCl2 3 X 25

(Ci,HBN02*HCl) C, H , N.

(C19H33N02*HCl)C, H, N.

(18) Barfknecht, C. F.; Nichols, D. E.;Dunn, W. J. J . M e d . Chem .1975, 18 , 208.(19)Domelsmith, L. N.; Eaton, T. A.; Houk, K. N.; Anderson, G .A.; Glennon, R. A.; Shulgin, A. T.;Castagnoli, N., Jr.; Kollman,P. A. J . M e d . C h e m . 1981,24, 1414.(20) Shulgin, A. T.; Dyer, D. C. J . M e d . C h e m . 1975, 18, 1201.(21)Nichols, D. .; Shulgin, A. T.; Dyer, D. C. Life Sci. 1977, 21,569.(22) Gomez-Jeria, J. S.; Morales-Lagos, D. In QSAR in Design ofBioactive Compounds; Kuchar, M., Ed.; J. R. Prous Interna-tional Publishers: Barcelona, Spain; 1984; pp 145-173.(23) Gomez-Jeria,J. S.; Cassels, B. K.; Saavedra-A guilar, J. C. Eur.J . Med. Chem. 1987,22, 433.(24) For a general review: see: Gu pta , S. P.; Singh, P.; Bindal, M.C. Chem. Reu. 1983,83, 633. (25) Cheng, A. C.; Castagnoli, N., Jr. J . M e d . C h e m . 1984,27,513.

8/3/2019 Mark R. Segge et al- A Structure-Affinity Study of the Binding of 4-Substituted Analogues of 1-(2,5-Dimethoxyphen

4/5

Analogues of 1 ( 2 3 -Dimethoxyph enyl) -2-aminopropanemL), the extract was washed with 5% HC l(2 X 25 mL)and dried(MgSO,). Evaporation of the solvent and recrystallization of theresidue from 95% EtOH yielded 0.45 g (82%) of N-phthaloyl-1-(4-cyano-2,5-dimethoxypheny1)-2-aminopropane6) : m p154-156 "C. Dep rotec tion of th e amin e was accomplished byheating a solution of 6 (0.35 g, 1.0 mmol) an d hyd razine (9 7%,110 pL) a t reflux in absolute EtOH (5 mL) for 2 h. After coolingthe solution, the precipitate was removed by filtration. Th e fiitercake was washed thoroughly with absolute Et OH (30 mL) , andthe combined filtrate s were concentrated in vacuo. Th e residuewas suspended in HzO (15 mL) an d the suspension was extractedwith CHzC lz (3 x 10 mL). Th e combined organic extract waswashed with aqueous Na2C0 3 10% w/v, 3 X 15 mL). Th e organicphase was dried (Na2SO4) nd evaporated. Th e oily residue wasdissolved in absolute EtO H (5 mL) and layered with ethereal HC1.Th e resulting product was recrystallized from Et OH /Eh O to give0.15 g (58%) of the title compound: mp 236-238 "C. Anal.

n P r o p y l 2,5-Dimethoxy-4-(2-aminopropyl)benzoatey-drogen Oxala te ( lv) . The t i t le compound was prepared byfollowing the procedure for the synthesis of n-butyl ester lw.Thus , a m ixture of 1 -propanol (300 mg, 5.0 mmol), concentratedH 8 0 4 50 mg, 0.5 mmol), and laa (150 mg, 0.5 "01) in benzene,after heatin g at reflux and extraction, gave 80mg (60 %) of thepropyl ester: mp 85-90 "C. Th e hydrogen oxalate salt wasprepared as or lw : m p 169-170 "C. A nal. (C1J3,N04 ~(C02H)2)C, H, N.n B u t y l 2,5-Dimethoxy-4-(2-aminopropyl)benzoate y-dro gen Ox ala te ( lw ). 1-Butanol (650 mg, 10.0 mmol) andconcentrated H zS 04 (100 mg, 1.0 mmol) were added t o a sus-pension of laa (as he free base; 0.24 g, 1.0 mmol) in benzene (30mL). The reaction mixture was then heated under reflux for 5h with continuous removal of H 2 0 n to a Dean-Stark trap. Aftercooling, the reaction m ixture was extracted with H 2 0 (2 X 50 mL).The aqueous extract was made alkaline (pH 10) with a 10%NaZCO3solution and extracted with EGO (3 X 50 mL). Th ecombined organic extra cts were washed with HzO (2 X 50 mL)and dried (NazS 04). The solvent was removed under reducedpressure and the oily residue was dried under high vacuum (0.15mmHg) for 1h. Upon standing at 0 "C, the oil solidified to give150 mg (54% ) of the title compoun d: mp 75-80 "C.A solution of the ester (100 mg) in anhydrous EhO (10 mL)was added to a satu rated solutio n of oxalic acid in anhydrous EGO(10 mL) . Th e precipitate was collected by filtration , air dried,and recrystall ized from 2-P rO H/ Et2 0 to afford the hydrogenoxalate salt: m p 149-150 "C. A nal. (C16H~N04~(C02H)z .1 / zH20)C, H, N.1- (2,5-Dimet hoxy-4-propionylphenyl)-2-aminopropaneHydroch lor ide ( l x ) . A mixture of 7 (200 mg, 0.58 mmol), 15 %NaOH (2 mL) , H 20 (8 mL), and MeOH (5 mL) was heated a treflux for 2.5 h. Th e solution was allowed to cool to room tem -peratur e, combined with HzO (20 mL), and extracted with E hO(4 X 15 mL). Th e hydrochloride salt was prepared by addingethereal HC1 to the dried (MgSO,) extract. Removal of the solventunde r redu ced p ressure left a yellow oil, which solidified und erhigh vacuum. Recrystallization from EtO H /E hO , followed byrecrystall ization from 1-P rO H/E t20 , gave 90 mg (54 %) of thetitle compound: mp 164-166 "C. Anal. (Cl4Hz1NO3.HCl)C, H,N.N-n-Propyl-2,5-dimethoxy-4-(-aminopropy1)benzamideH y d r o g e n O x a l a t e ( l y ) . A solution of n-propyl 2,5-dimeth-oxy-4-(2-aminopropyl)benzoate lv ; 120 mg, 0.43 mmol) in n-propylam ine (5 mL) was heated und er reflux for 5 h. Excessn-propylamine was removed in vacuo and the resulting solidresidue was washed with EGO to yield 60 mg (56 %) of th e amideas a pale, buff solid: mp 110-115 "C. A solution of th e amide(50 mg) n absolute EtO H (1mL)was added to a solution of oxalicacid in anhydrous EtzO (5 mL). Th e oxalate salt was collectedby filtration and recrystallized from 2-PrOH/EGO : mp 183-184"C. Anal. (C1SHz4Nz03~(C02H)z), H, N.1-(2,5-Dimethoxy-4-hydroxypheny1)-2-aminopropaney-dro gen Ox ala te ( lz) . To a solut ion of N-phthaloyl-l - (2,5-di -metho~y-4-formylpheny1)-2-aminopropane~~0.5 g, 1.42 mmol)

( C I ~ H I ~ N ~ O ~ ' H C I ), H, N.

Journa l of Medicinal Chemistry, 1990, Vol. 33, No . 3 1035in CHC13 (25 mL) was added 0.42 g (2.07 mmol) of 85% 3-chloroperoxybenzoic acid. Th e solution was the n stirred at roomtemp erature for 21 h. Th e residue remaining afte r removal ofthe solvent under reduced pressure was dissolved in Eh O (50mL).This solution was washed with 4% Na HS 03 (2 X 25 mL). Thewashed ethereal solution was evaporated to dryness under reducedpressure and the residue was dissolved in absolute EtO H (50mL).HC1 (g) was bubbled through the ethanolic solution for 60 s. Theflask was sealed and allowed to stand at room temperature for18 h. Th e solvent was removed under reduced pressure and theremaining HCl was chased with 95% EtOH. Th e residue waschromatographed on silica gel (10 g, 1 2 mm i.d.) and eluted withEtOAc. Crystallization from EtO H/H zO yielded 0.57 g of a waxymaterial, which was used without fur ther p urification. A solutionof th e hydroxy compound (0.34 g) and anhy drou s hydrazine (110rL ) in absolute EtOH (5 mL) was heated at reflux for 3 h andthen stirred a t room temperature for an additional 14 h. Afterthe solvent was removed under reduced pressure, the residue waswashed with 5% H Cl(2 0 mL), and the mixture was filtered. Thefiltrate was made alkaline with 15% NaOH and extr acted withCHC13 (3 X 15 mL). Th e organic portion was dried (Naz S04),evaporated to dryness, and the residue was dissolved in a minimalamount of absolute EtOH an d was adde d dropwise to a saturatedetherea l solution of oxalic acid. T he prec ipitate was recrystallizedfrom EtO H/ EtzO to yield 25 mg (5.8%) of the t i t le compound,mp 205-210 "C dec. Anal. (C13H17 N03.(CO OH)2) , H , N.1-(4-Carboxy-2,5-dimethoxyphenyl)-2-aminopropane y-droch lor ide ( l aa ) . A solution of l g (as he free base; 1.4 g, 5.0"01) in anhydrous EhO (20mL)was added dropwise to a stirredsolution of n-butyllithium (1.3 g, 20 mmol; 12.5 mL of 1.6 Msolution) in anhydrous Et 2 0 (20 mL) a t 0 "C. Th e reactionmixture was the n stirred a t room temperature for 2 h, after whichthe mixtu re was poured over solid COP (50 g). Th e solvent wasremoved under reduced pressure an d the residue was washed withEtzO. Th e residue was dissolved in water (50 mL) and washedwith CHC13 (2 X 30 mL). The aqueous solution was acidified withconcentrated HCl and washed with CHC13 (2 x 30 mL). Theaqueous portion was evaporated under reduced pressure to givea solid residue, which was recrystallized twice from MeCN toafford 0.25 g (30% ) of a very light tan am orphous solid: mp194-196 "C (lit.27mp 196-198 "C).N-(Trifluoroacetyl)-l-(2,5-dimethoxyphenyl)-2-amino-pr op an e (2). Trifluoroacetic anhydride (20 mL) was slowly addedto 1-(2,5-dimethoxyphenyl)-2-aminopropanela ;2.7 g, 13.8 "01)while cooling in an ice bath. Th e reaction mixture was allowedto warm to room temperature, was stirred for 2 h, and was pouredover crushed ice (200 9). Th e white precipitate was collected byfiltration and was washed with a large volume of cold HzO.Recrystallization from EtO H/ Hz O afforded 3.0 g (74%) of thetitle compound as a flocculent, white solid mp 101-103 "C. Anal.( C ~ ~ H I ~ F J ' J O ~ ), H, NeN-(Trifluoroacetyl)-l-(2,5-dimethoxy-4-hexanoyl-phenyl)-2-aminopropane (3a). Titanium tetrachloride (1.4 mL,12.8 mmol) was added dropwise, under a N2 atmosphere, to asolution of 2 (2.0 g, 6.9 mm ol) in C HZCl2 60 mL) cooled t o ca.-30 "C. Hexanoyl chloride (1.28g, 9.5 mmol) in CHzClz 5 mL)was then add ed dropwise while the reaction mixture was main-tained a t ca. -30 "C. Th e mixture was stirred a t ca. -30 "C foran additional 30 min and was allowed to warm to room tem-perature. Stirring was continued for 3 days and the reactionmixtu re was cautiously poure d over 300 g of crush ed ice. Aftervigorous stirring, the layers were separated. Th e aqueous portionwas further extracted with C HzCl2 3 x 200 mL). The extractwas washed successively with HzO (200 mL) , 5% HCl(100 mL),(200 mL), saturated aqueous NaH C03 (200 mL), and saturatedaqueous NaCl (100 mL). After drying of the extract (N azS 04),the solventwas removed under reduced pressure. Rerrystallizationfrom Et O H /H 20 afforded 1.33 g (49% ) of 3a: mp 105-107 " C .Anal. (Cl9HZ6F3No4) , H, N.N-(Trifluoroacetyl)-l-(2,5-dimethoxy-4-octanoyl-phen yl)-2-a mino propa ne (3b). The t i tle compound was pre-pared in 52% yield in a manner similar to that used for the

(26) Ho, B. T.; Tansey, L. W. J . Me d . Chem. 1971, 14 , 156. (27) Matin, S. B.; Callery, P. S.; Zweig, J. S.; OBrien, A.;Rapoport,R. ; Castagnoli, N., Jr. J . Med. Chem. 1974, 17 , 877.

8/3/2019 Mark R. Segge et al- A Structure-Affinity Study of the Binding of 4-Substituted Analogues of 1-(2,5-Dimethoxyphen

5/5

1036 Journal of Medicinal Chemistry , 1990, Vol. 33, No. 3synthesis of 3a; recrystallization from E tO H /H 20 afforded thepro duc t as an off-white solid: m p 104-105 "C. Anal. (C21H30-

N - ( T r i f l u o r o a c e t y 1 ) -1- (4-benzoyl-2,5-dimethoxy-phenyl)-2-am inopropane (3c) . Compound 3c was prepared ina manner similar to that used for the synthesis of 3a. Recrys-tallization from Et OH /H 20 gave 48% of the title compound mp134-136 "C. Anal. (C20H zoF3N 04) , H , N.N - T r i f l u o r o a c e t y 1 ) - 1-(2,5-dimethoxy-4-hydro-cinnamoylphenyl)-2-aminopropane 3d). Compound 3d wasprepared by using the same method used for the prepa ration of3a. Th e material remaining after removal of the solvent underreduced pressure was recrystallized from E tOH /H zO to give 1.3g (44%) of the title compou nd: mp 131-133 "C. Anal. (CZ2-&$'&OI) C, H , N.N-(Trifluoroacetyl)-l-(2,5-dimethoxy-4-hexylphenyl)-2-aminopro pane (4a). A solution of 3a (0.22 g, 0.6 mmol) in glac ialacetic acid (60 mL) containing ca. 0.3 mL of 60% aqueous HC lO,was hydrogenated over 0.1 g of 10% P d/C for ca.6 h. The catalystwas removed by filtration an d washed with CHZClz 25mL). Afteraddition of HzO (150mL ) to the fiitrate, the product was extractedwith CH2C12 4 X 75 mL) . Th e organic portion w as washed withHzO (200 mL), saturated aqueous NaH C03 (200 mL), and againwith H 20 200mL), dried (MgSO,), and evaporated under reducedpressure. Th e residue was recrystallized from EtO H/H zO, yielding0.15 g (66%) of compo und 4a: mp 119.5-120.5 "C. Anal.N-(Trifluoroacetyl)-l-(2,5-dimethoxy-4-octylphenyl)-2-

aminopropane (4b). Th e title compound was prepared by usingthe procedure described for 4a. Recrystallization of the residuefrom 95% E tO H gave 56% of the title compound: mp 109-111"C. Anal. (C21H 32F3N 03) , H, N.N-(Trifluoroacety1)-1-( 4-benzyl-2,5-dimethoxyphenyl)-2-am in op ro pa ne (4c). Th e title compound was prepared with theproc edur e used for th e syn thesis of 4a. Recrystallization fro mE tO H /H 20 gave 61% of 4c: mp 147-149 "C. Anal. (C20H22-F 3 N 0 3 )C, H, N.N - T r i f l u o r o a c e t y 1 ) -1 [2 , 5 - d im e t h o x y - 4 - ( 3 - p h e n y l -propyl)phenyl]-2-aminopropane 4d). Th e title compound wasprepar ed by using the proce dure described for the synthesis of4a. Th e residue was recrystallized from Et OH /Hz O to give 63%of 4d: m p 125-126.5 "C. Anal. (C22H26 F3N03) , H , N.N-Phthaloyl-l-(2,5-dimethoxy-4-bromophenyl)-2-amino-p r o p a n e (5). A solution of 48% HBr (0.80g) in glacial aceticacid (10mL) was added to a stirred solution of N-phthaloyl-l-(2,5-dimethoxypheny1)-2-aminopropanez31.5 g, 4.0 mmol) inglacial acetic acid (10 mL) at 0 C. A solu tion of Br2 (0.8 g, 4.2mmol) in glacial acetic acid (10 mL) was then added dropwise.Th e reaction mixture was then allowed to stir at room temperaturefor 3 h. Th e solvent was removed under reduced pressu re andthe residue was recrystallized from 95% EtO H to give the titlecompound as a white, crystalline solid in qua ntitativ e yield: mp111-112 "C. Anal. (ClsHl8BrN O4) C, H , Br.N - T r i f l u o r o a c e t y 1 ) -1 ( 2 ,5 - di m e t h o x y - l - p r o p i o n y l -phenyl ) -2 -aminopropane(7). Titaniu m tetrachloride (1.4 mL,12.8 mmol) was added dropwise under a N z atmosphere to asolu tion of 2 (2.0 g, 6.9 mmo l) in CH2C12 70 mL) cooled t o ca.-30 "C. P ropio nyl chloride (0.88g, 9.5 mmol) in CH2C12 5 mL)wa s added dropwise to the d ark brown solution while the reactionmixture was maintained at ca. -30 "C. Th e mixture was stirredat ca. -30 "C for an additional 30 min an d was allowed to warmto room temperature. Stirring was continued for 3 days and thereaction mixture was cautiously poured over 300 g of crushed ice.Th e deep red dissipated upon vigorous mixing. After separatingthe layers, the aqu eous portion was extracted with CH2C12 4 X175 mL) an d th e organic portion was washed successively withH 20 (200 mL) , 5% HCl (100 mL), H zO (200 mL), saturatedaqueous NaH C0 3 (200 mL), and satu rated aqueous NaCl (200mL). After drying (Na2S O4), he solvent was removed underreduced pressure. Recrystallization from Et O H /H 20 gave 1.63g (68% ) of 7: mp 143-145 "C. Anal. (C16H20F3N04) , H, N.

F,NO*) C, H, N.

(Ci&&3N03) C, H , Ne

Seggel et al.Radio l igand Bind in g Studies. Radioligand binding assayswere conducted in essentially the same manner as previouslyreported.28 Briefly, Taconic Farm s male Sprague-Dawley rats(ca. 220 g) were decapitated and the brains were immediatelyplaced in ice-cold 0.9% saline. Dissecting over ice, th e frontalcorticies were removed with th e anterior bo rder of the corpuscallosum a s a landmark. Tissue was either used immediately orstored a t -30 "C until needed ( no differences were noted betweenpreparations). Mem brane homogenate5 were prepared in a 50mM T ris HCl buffer ( pH 7.4 at 37 "C) containing 10mM MgS04an d 0.5mM Na2EDTA. Assays were performed in 2.0 mL of thissame buffer to which was added to 0.1% ascorbate and 10 pMpargyline; membranes (3-mg wet weight) were added last. Dis-placement experiments at 11concentrations of nonlabeled d rugwere performed in triplicate with tritiated ketanserin (90.4 Ci/"01) (New England Nuclear). Specific binding was de fi ed with1pM cinanserin. Solutions of all test compounds were made freshdaily. Following incubation at 37 OC for 15min, membranes wererapidly filtered over glass-fiber filters th at h ad been presoakedin 0.1% polyethyleneimine, followed by a 10-mL wash with ice-cold buffer. Following a 6-h equilibration in Scin tiven e (Fisher),samples were counted in a Bec kman 3801 counter with an effi-ciency of 45%. IC50 values were determ ined an d Ki values werecalculated with th e Cheng-Prusoff equation.29Isola ted Tissue Stu die s. Th e isolated tissue studies with rataorta were performed as previously described in detai1.14J7 Inbrief, using male Sprague-Dawley rats (250-300 g), 4-mm ringsof r at thoracic aorta were dissected a nd su spended between two

stainless steel hooks connected to a force displacement transducer.Th e organ bath was continuously bubbled with 95% 02/5% COzin a Krebs-bicarb onate buffer mainta ined at 37 "C. Restingtension was set at 2.5 g and, after a 1-h equilibration period,concentration-response stu dies were performed as previouslydescribed.', Th e aortic segments were incubated with varyingconcentrations of antagonist for 10min at 37 OC prior to itrationwith graded concentrations of 5-HT. All results represent themean, f tandard error of the mean, of five or six individualexperiments.Acknowledgment. This work was supported in partby PH S grants DA 01642 and NS 23520. BLR and EASare members of the Naval Medical Research CommandWork Unit #MRO 4120.05-1004. The opinions and as-

sertions contained herein are private ones and are not tobe construed as reflecting the views of th e Navy D epart-ment, t he Naval Service at large, or the Dep artment ofDefense. The ex periments reported herein were conductedaccording to th e principles set forth in th e Guide for theCare and Use of Laboratory Animal Resources; NationalResearch Council. Department of Health, Education, andWelfare. U. S. Gove rnment Printin g Office: W ashington,DC, 1985, Publication No. (NIH ) 78-23.(28) Lyon, R. A.; Titeler, M.; Seggel,M.R.; Glennon,R. A. Eur. J .

Pharmacol. 1988, 145, 291.(29) Cheng, Y. C.; Prusoff, W. H. Biochem. Pharmacol. 1973,22,3099.(30 ) ECm ( fSEM ) values were calculated for 5-HT by using logit-lo g transformations of the cummulative concentration-re-sponse curves and linear least-squares regreasion analysis. TheStudent's t test for unpaired sam ples was used for evaluationof significant differences (p < 0.05*). For each compound,concentrations (and pM ECw values of 5-HT) are as follows:Is,0 pM (6.3 f 0.8), 1 pM (7.7 f 1.71, 10pM (20.5 f .8*), 30pM (12.4 f 1.8*); l p ,OpM (5.8 f .5), 1pM (5.4 h 1.3),10pM(12.8 f 0.8*), 50 pM (9.5 f 1.4*); lr , 0 pM (4.8 f 0.6), 1 pM(13.8 f .4*), 10pM (18.2 f 0.4*), 100pM (8.4 f .2); lq , 0pM(5.4f 0.9), 1 pM (4.1 i .8), 10pM (7.4 f 0.9), 80 pM (9.8 f3.2).