Journal of 'If edicinal Chemistry, 1970, Vol. I S , S o . 5 821
ANORECTIC ~-BE~YZYLCYCLOPROPYL.IJIINES
DIETHYL
TABLE I BENZYLYALONITES AXD DIETHYL BENZYL-(~-DIYETHYL.IMIXOETHYL)XM:ILOX~TES
_ _ _______RCH 2 C H ( C 0 2 E t ) p BP. 'C
Yield,
R C H ~ C ( C O * E t ) 2 ( C H z ) z ~ ~ ~ e ~ ____
I
hlethiodide,a mp, 'C
Yield.'
Salt, mp, 'C
Recrystn
7G solvent I'ormiilac C6H5 d e, f 88 2-ClCsH4 138-164g ( 1 , 2 ) 68 106-109 70 79 143-147 78 '38-100i i-PrOH-Kt2O Cl&&O4' C ~ H I & & S ~ , ~ 2->IeCsHa 137-143* ( 1 . 2 ) 2-CFaCsH4 134-141 (1.0) 70 f 88 3-ClCsH4 152-15S2 (0.8) 55 166-169 70 3->leCeH4 139-148(0.7) 54 119-121 71 3-CF3C6H4 141-147 (0.7) 34 183-185 66 4-BrC6H4 164-172" ( 0 . 4 ) 59 210-212 71 178-179 hIenCO C18H26BrNO4, HC1 4-ClCsH4 141-1470 (0.6) 39 206-207 dec 82 19C-191 EtOH-Et20 CisHnsNO4. HC1 4-1k!C& 136-14Op ( 0 . 4 ) 63 197-199 77 152-153 XeZCO-Et2O C19H2sS04.HC1 4-AleOCsHr 149-163 (0.8) 69 209-211 70 159-160 EtOH-Et20 C19H2YXO:. HC1 4-CFaC6Ha 124-131 (0.4) 57 154-156' 80 124-125 ;\TesCO CigHn6F3NO4' 0.5C4H4048 4-CFaOCsH4 136-142 ( 0 . 3 ) 58 188-190 74 130-131 EtOAc-EtnO C ~ ~ H ~ B FCaH404' ~XO~ 51 165-167 69 115-117 EtOAc-Et,zO CigHzsFaNO4S.C4H404t 4-CFaSC6H4 145-151 ( 0 . 6 ) "6-CInCsHd 164-172 ( 0 . 4 ) 98 168-170 68 124-126 EtOAc-Et20 Ci8HnsC1,NOa. C4HaO4' C~C~O-C&I 111-122r (0.5) 71 186-188 dec 73 144-145 i-PrOH Ci8HaaNOI. C4IX,Oat All methiodides were prepared in Rle&O, unless indicated otherwise. b The yield of methiodide. All compounds which were coiiverted into a salt, but not met,hiodides, were analyzed for C, H, N and the analytical values were within f 0 . 4 of the calculuted figures, unless indicat,ed otherwise. d Diethyl benzylmalonate wm obt.ained from a commercial source. e Reference 16. The methR. A. Barnes and L. iodide was prepared in EtpO; it was a viscous liquid which was used for further reaction without purification. Gordon [ J .ilmer. Chem. SOC.71, 2644 (1949)l reported bp l55-16Oo (4 mm). B. B. Elsner and K. J. Parker [ J . Chem. Soc., 592 (1957)l reported bp 133' (0.2 mm). The free base had bp 156-162" (1.3 mm). j Cyclohexylsulfamat,e. C: calcd., 58.34: J. Kenner and E. Witham [ J . Chem. Soc., 119, 1452 (1921)] reported bp 213-214" (40 mm). J. P. Trivedi and found, 57.83. J. J. Trivedi, [ J .Indian Chem. SOC.,35,687 (1958)l reported bp 180' (30 mm). J. von Braun and J. Nelles [Ber., 66, 1464 (193311 reported bp 193-198' (14 mm). J. von Braun and J. Xelles [ibid., 66, 1464 (1933)l reported bp 190-192" (14 mm). P G. Darzeiis and A. Heinz [C. R. Acad. Sci., 184, 33 (1927)l reported bp 179-180'. G. Darzens and A. LBvy [ibid.,200, 469 (1935)] reported bp 178-180" (3 mm). The methiodide was recrystallized from EtOH-EtzO. Hemimaleate. Maleate. G. S. Hiers and 1:. A d a m [ J . A m e r . Chem. Soc., 48, 2385 (1926)l reported bp 135-136" (3 mm). ("1
R
70
Q
established by comparison with an unambiguously synthesized sample, was obtained. Hydrogenolysis of this cyclopropane derivative is another example of reductive ring-opening of an NH-substituted cyclopropane by LAH, possibly by a mechanismy resembling one postulated by Hochstein and Brownlo for LAH reduction of cinnamyl alcohol. Secessity for initial formation of a N anion, as suggested by this mechanismlY~'O is supported by the observation that LAH reduction of several tertiary N-(1-benzylcyclopropy1)amides to amines occurred without cleavage of the cyclopropane ring. Additionally, in analogy with LAH reduction of the double bond of allyl alcohol, which proceeds relatively more slowly than that of cinnamyl a l c ~ h o l , ' ~ hydrogenolysis of the cyclopropane ring might be anticipated to occur less readily with XI than with 2-arylcyclopropylamine derivatives in which the anion can achieve benzylic stabilization.y Hydrogenolysis was not observed upon LAH reduction of l-benzylcyclopropyl isocynnates a t 25". Also, reductive cleavage of XI could be avoided by a shorter reduction time (30 min) which gives 1-benzyl-N-ethylcyclopropylamine (XIV, R = Ph) in low yield. N-Ethyl 1-substituted cyclopropylamines XIV were obtained in higher yield from corresponding formamides XI11 via ethylation and subsequent acid hydrolysis of the formyl group, as indicated in Chart 111. Structure-Activity Relationships.-Anorectic activity of 1-benzylcyclopropylamines\\as determined in the rat and dog anorexia tests described in the Experimental Section. The parent compound of the series, l-benzylc ~ c l o p r o ~ ~ ~ l a n(l), i i nise a potent anorectic agent n hich (9) C
J
haioer, h I3urger I 27, 768 (1962)
ZiIngibl, C S Daiis, a n d C. L Zirhle,
0 , g Chem
(101 I. (1948)
1 Hoclistein and 11. G. Uronn, J. Amer Chem Soc , 7 0 , 3484
RCH,C( COLEt),(CH,),SMe,
I1
I
IICH-C-CH,
\CH,/
I. hfci -----+
2. OH-Ex. Resin EtOH
OH-
I RCH-C-CH
111
\CH. IV
CUI i I U 5
-
sco I
RCH~C-CH
\CH. /
VI
RCH.~-CH
\CH/
VI1
reduces food consumption of rats, dogs, and cats. The main side effects of 1 are those of mild CNS stimulation as indicated by slight restlessness and occabional increased motor activity observed in dogs. Cardiovascular effects in dogs treated with 1 are comparable to those noted with equivalent anorectic doses of dextroamphetamine.
822 Journal of Ncdicinal Chemastry, 1970, Vol. 13, S o . 6
Itil lii2
Journal of Medicinal Chemistry, 1970, Vol. 13, No. 6 823
ANORECTIC ~-BENEYLCYCLOPROPYLAMINES
TABLE IV N-METHYL ~-BENZYLCYCLOPROPYLAM~XES
NHMe
Recr) s t n solvent
balt,
5-0,
LIP, o
17
160-162'
Et OH-Et
18 19 20 21 22 23 24
1.io- 15% 174-176 178-180 132-134 98-100 82-85 147- 149
z-PrOH 7-PrOEI-Et20 z-PrOH-Et20 2-PrOH Et OA4c-Et2 0 Et OAc-Et p 0
c
2 0
,--El)io
Yield.
? ai 38
Formulah
CiiHijN.HC1
.\nurectic acti\ itj.'"' ( P O ) mg lig (fidncial litnits)--Rate I>U..i
1.1
6.,5 = 3.5';, 8 . 2 = -86'; 10 = -36'3 16.7 (9,9-28 4 ) 2 1 . 6 = -38 12.6 -80 1.5.1 = -24";
=
1 4
16.6 =
R . X ( T , 9-11.6)
I
(j(j=?
-.x((,
8.7 = 21.7 = -ox:; 6 . 4 = -7X';;, 2.5 2-P1 O H 1.52-1.54 17 = -99r; 5 . 2 (2.7-17.6) z-PrOH-Et 2 0 1.52- 1.51 26 21 = -8.jG 27 z-PrOH-EtgO 136-138 16.6 = --.5,icC LIe2C0 144-146 28 10.7 (6.0-1.5. 8) ?rIeSCO-EtrO 29 117-118.5 11 = +lcc ?rIe2CO-Et20 :10 123-12.5 1 6 . 6 = -175; i-PrOH 16C162 31 15 = +16'C i-Pr OH-Et SO 32 137-139 All .l--meth\laniiiie ,alts ueie a n d \ /ed Oveiall yield from appropliate 1-benqlcyclopropanecarbonyl chloiide See footnote See footnote c, Table 111. and the analytical values were within 2 ~ 0 . 4 %of the calculated figure.. e See footnote e , Table 111. J Reference 1. Maleate.
EtOH-Et20
('
6.4=
::
8.3 = 3 4 8 4 = J : 6.X
I
f o 1 C, H , S tl, Table I11
Q
cyclopropylamine increased anorectic potency slightly whereas others decreased it. Results of anorectic testing of a series of substituted l-benzylcyclopropylamines and their S-methylated derivatives in rats and dogs are tabulated in Tables I11 and IV. Substitution of the ortho position of 1-benzylcj-clopropj-lamine with Me (3) reduced anorectic potency in the rat by 0.5, whereas similar substitution with CI (2) or CE's (4) resulted in an even greater decrease of potency. Substitution of the meta position with CI ( 5 ) , ?\le (6), or CFs (7) also decreased anorectic potency; however, the decrease waq less thanwith the ortho isomers. Introduction of substituents into the para position of l-benzylcyclopropylamine had a variable influence on anorexia. h p-C1 derivative (9) was somewhat more potent than the parent 1 in both rats and dogs. Likewise, the p-\le congener 10 was more potent than 1 in producing anorexia in rats, but it was less potent in dogs. -4 p-Br (8) derivative was approximately equipotent with 1 in rats, but it was less potent in dogs. The p-lIeO (11) analog was only slightly less potent than the parent, whereas p-CFs (12), CFsO (13), and CF3S (14) derivatives were less than one-half as potent in rats. Only one disubstituted compound, a 2,6-dichloro derivative 15, was studied; it was about one-sixth a\ potent as 1 in the rat anorexia test. A cyclohexyl congener 16 was less than 0.1 as effective as 1-brnzylcyclopropylamine in producing anorexia in rats. Although the N-methylated derivatives (17-32) were generally less potent than the corresponding primary amines, in several instances, e.g., m-CFa (7,23) and p-CF,O (13, 29), the substituted derivatives and their unsubstituted analogs were approximately equipotent as anorectics in rats arid dogs.
3.3
PhCH2CH(CO& )CH2CH=CH2
+
IX
CH m r I11 NH,
NHAc 4L 0
RCHLC-CHL
\I CH.
\/
RCH-C-CH
CH-
LAH
' ( 6 hri
NH Et
I
RCH,CHCH,CH, XII, R = Ph
SI. R = P h E t 0 CH
YHCHO
XI11
XI\'
Rebults of anorectic examination of variouh other S-substituted 1-benzylcyclopropylamines in rats and dogs are tabulated in Table V. N-Ethyl-1-berixylcyclopropylamine (33) and its 0-C1 (42),p-C1 (45), m-Ne (44), and p-Ne (51) derivatives were slightly less potent than the corresponding primary amines or their N-Ale counterparts; however, in the ni-CF3 series an N-Et deriwtive 43 "a5 somewhat more potent than either thc primary amine or its N-Me homolog. The only other Yec-
824
Journal of Jfedicinal Chemtstry, 1970, V o l . l s l j .Yo.b
-
I .
U
I
P
x
,
c
ANORECTIC ~-BEXYLCYCLOPROPYLAMISI:S
ondary amine studied was N - (2-hydroxyethyl)-l-benzylcyclopropylamine (34) ; it was approximately equipotent to the corresponding N-Et congener 33. S,X-Disubstitution of I-benzylcyclopropylamine and bubstituted analogs with Ale or Et, e.g., 35-36,46,47,52, generally afforded compounds with anorectic potency approximately equivalent to the corresponding N-monoethyl derivatives. An exception is N,N-diethgl-l-(4methylbenzy1)cyclopropylamine (53) ; it is much less potent than either its N-Et (51) or N,N-A\Le2(52) counterpart. Two N-aralkyl-N-methyl-1-benzylcyclopropylamines (38, 39) exhibited markedly reduced potency in the rat. Several K-acylated 1-benzylcyclopropylamines were also examined for anorectic activity in rats and dogb. -4n N-carbethoxy derivative (40) of l-benzglcyclopropylamine was less than half as potent ab the primary amine in the rat anorexia test mhereasthe corresponding benzyl carbamate 41 was onlyx-eakly active at 50 mg/kg. N-Formyl (48) and N-carbethoxy (50) derivatives of 1-(4-chlorobenxyl)cyclopropylamine,although less potent than the primary amine, retained significant anorectic activity; however, an K-acetylated congener 49 failed to produce significant anorexia in rats a t 25 mg/kg. Similarly, N-Ac (54) and N-carbethoxy (55) derivatives of 1-(4-methylbenzyl)cyclopropylaminehad only weak anorectic activity at the doses studied.
Experimental Section Pharmacology. Methods. A. Rat Anorexia Te~t.~-Anorectic activity iri the rat was determined by measuring changes in food corisumption of male, albino rats (\%tar strain) specially trained t o consume their normal 24-hr food intake in only 6 hr. Compounds were administered orally by gastric intubation (8 rats per treatment group). Food (powdered Purina Laboratory Chow) was presented in tared cups a t the predetermined time of peak effect. After 1 hr the cups were weighed and the mean food consumption of each drug-treated group was compared to controls. A doseresponse curve was obtained by plotting mean food consumption against the log of the dose. The ED&,is the dose that produces a 50%,reduction in food consumption of drugtreated rats relative t o controls. Fieller's Theorem" was used t o calculate 95% fiducial limits. I n cases where EDsO's were not determined, the results (Tables 111-V) are expressed as per cent reduction of food consumption of rats treated with the indicated dose, relative t o controls. B. Dog Anorexia Test.-Anoretic activity was determined in groups of dogs fasted for 18 hr prior to oral administration of various doses of test compounds. About 0.1 of the daily food ration (a commercial canned meat preparation) was offered to t.he dogs a t 0.5-hr intervals following administration of the drug. Anorexia is defined as failure t o eat for two successive feedings. The dose of drug effective in causing anorexia in 50% of the dogs (EDj")and 95cofiducial limits were calculated by the method of Litchfield and Wilcoxon.12 For compounds whoye ED6,'s were not determined, results are ureseiited (Tables 111-V) as the number of dogs displaying anorexia, the total number tested at the indicated dose. Chemistry.13 Benzyl Halides.-Both benzyl chlorides and (11) D. J. Finney, "Probit Analysis," 2nd ed, Cambridge University Press, London, 1952. (12) J. T. Litchfield, Jr. a n d F. Wilcoxon, J . Pharmacul. E z p . Ther., 96, 99 (1949). (13) All melting points were determined with a capillary melting point apparatus a n d are uncorrected. Boiling points are also uncorrected. Where analyses are indicated only by symbols of t h e elements, analytical results obtained for those elements were within f 0 . 4 Y o of t h e theoretical values. Elemental analyses were performed by Miss Margaret Carroll and coworkers of t h e Bnalytical a n d Physical Chemistry Section, Research and Development Dirision, Smith Kline a n d French Laboratories. I r spectra were as anticipat.ed for all compounds. They were determined as Xujol mulls or natural films and were recorded on a Perkin-Elmer Infracord spectrophotometer. N m r analyses !yere carried out on a Varian T60 nnir spectrometer a n d are reported in 6 (ppm) calibrated against TMS.
Journal of Medicinal Chemistry, 1970, Vol. 13,
A-0.
5 825
bromides were employed for alkylation of diethyl malonate. With the exception of 2- and 4-CF3-, 4-CF30-, and 4-CFsS-benzyl chlorides, the required benzyl halides were obtained from commercial sources and were distilled prior t o reaction. 4-Trifluoromethylbenzyl chloride was prepared according t o literat,ure direction^.'^ The same met.hod was employed for synthesis of 2-CFa-, 4-CF30-, and 4-CFaS-benzyl chlorides. Requisite substituted benzoic acids were reduced with LAH and the resulting alcohols were treated with SOC12 t o afford 2-CF&6H4CH~C1 (76.5% yield), bp 103-110" (25 mm), 4-CF30C6H4CH2C1 (86'h b bp 8G-90" (25 mm), and 4-CF,SC6H4CH2C1 (55yo),bp 86-92 (9 mm). General Methods. A. Diethyl Benzy1malonates.-Diethyl malonate (80.1 g, 0.5 mole) was added dropwise at 20-25" to a stirred suspension of 20.5 g (0.5 mole) of a 58.57, dispersion of S a H iri mineral oil in 200 ml of DMSO. The mixture was stirred at this temperature until H2 evolutioii Tvas completed, then 0.5 mole of the appropriate benzyl chloride or bromide in 100 ml of DlISO was added slowly. After the mixture wm heat,ed for 30 min on the st'eam bath, it was poured into 1-1, of ice-H20. The precipitated oil was ext,racted into Et&. The E h O extracts were dried (A2gSO4) and concentrated. Uist'illation of t'he residual liquids gave the diethyl benzylmalonates listed in Table I. B. Diethyl Benzyl(dimethylaminoethy1)malonates.-Alkylat'ion of diethyl benzylmalonates with 2-dimethylaminoethyl chloride was carried out with S a H in DIMS0 by method A. In most instances, diethyl benzyl(dimethylaminoethy1)malonates were converted into quaternary methiodides by treatment' with ai1 excess of Ale1 in hIe2CO-Et20. Several of the esters were characterized as salts (see Table I). C. Ethyl 1-Benzylcyclopropanecarboxylates.--il1i approximately threefold excess of anion-exchange resin (OH- forni)'j was washed several times with EtOH. The resin (ca. 0.6 mole) was suspended in 500 ml of EtOH, and 0.2 mole of the appropriate diethyl benzyl(dimethylaminoethy1)malonate methiodide was added. The mixt'ure was stirred at 25" for 1 hr; then it was filtered and the filtrate was conceiitrated in Z J U C ! ( ~ . The residue was gradually heated to 13O-20O0 at water aspirat,or pressure (10-23 mm). After decomposition, as evidenced by gas evolution, was completed the residual liquid was distilled t o give the ethyl 1-benzylcyclopropanecarboxylates tabulated in Table 11. Ethyl 1-benzylcyclopropanecarboxylate was also prepared (25% yield) by thermal decomposition of diethyl benzyl(2dimethylaminoethy1)malonate methiodide16 at 200" (10 mm.). D. 1-Benzylcyclopropanecarboxylic Acids.-Ethyl l-benzylcyclopropanecarboxylates were hydrolyzed with a 207, excess of KOH in EtOH-H2O. After distillation of the EtOH, the crystalline acids (Table 11) were precipitated by acidification of the aq solution with 3 N HC1. The st,ructure of each of the 1-benzylcyclopropanecarboxylic acids was confirmed by their nmr spect,ra. Typically, the nmr (CDC13) had peaks at 6 0.7-1.1 (m, cyclopropyl CHZ), 1.251.55 (m, cyclopropyl CHZ),2.9-3.1 (s, benzyl CHz), 7-7.4 (aromatic H's), and 12.0-12.2 (s, COOH). 1-Cyclohexylmethylcyclopropanecarboxylic acid gave an nmr (CDC13) in which the peaks for t,he cyclopropyl H's (AzB2 pattern at 6 0.78 and 1.25) were superimposed with the cyclohexylmethyl H peaks (m, 1.02.0); it also showed a peak at S 12.0 (s, COOH). E. 1-Benzylcyclopropanecarbonyl Chlorides.-A mixture of 1 mole of I-berizylcyclopropanecarboxylicacid and 2 moles of SOC12 was allowed to stand at 25" for 24 hr (or until gas evolution wad completed). a f t e r excess SOC1, was removed in vucuo, the acid chlorides wwe distilled. For boiling points and yields of 1berizylcyclopropanecarbonyl chlorides see Table 11. F. 1-Benzylcyclopropyl Isocyanates.-1-Benzylcyclopropanecarbonyl chlorides lvere converted into 1-benzylcyclopropyl isocyanates by the 'k-et K a S s procedure" as described previously.' The isocyanates were obtained in nearly quantitative yield by this procedure. They were employed for further reaction without additional purification. G. Acid Hydrolysis of Isocyanates.-1-Benzylcyclopropyl isocyanates were hydrolyzed to primary amines by the same procedure described for "2-Substituted Cyclopropylamines. Acid (14) J. R. O n e n and \I-. H. Saunders, .Jr., J . Amer. Chem. SOC., 88, 5809 (1966). (15) I strongly basic polystyrene alkyl quaternary amine (hydroxide form) of medium porosity was employed. Research grade Rexyn 201 (OH), purchased from t h e Fisher Scientific Co., and Amberlite IR.\, purchased from Mallinckrodt Chemical TVorke, were found eclually satisfactory. (16) C. K. Ingold and hl. A. T. Rogers, J . Chcm. Soc., i 2 2 ( I Y 3 5 ) .
