J.Med. Chem. 1984,27, 1111-1118 room. Rats were trained on a fixed-ratio 32 (FR32) schedule. Drugs. Drugs were administered in a volume of 0.1 mL of physiologicalsaline per 100 g of body weight. LSD tartrate was obtained from the National Institute on Drug Abuse. All other drugs were used as the hydrochlorides. Data Analysis. The mean percent responding on the LSDappropriate lever was calculated for each treatment group, These data were used to construct dose-remonse curves and. where complete generalization occurred (greater than 85% responding on the drug lever), an ED50 was calculated by the method Of Litchfield and Wilco~on.*~ (19) Litchfield,J. T.; Wilcoxon, F. J.Pharmacol. Exp. Ther. 1949, 96, 99.
1111
Acknowledgment. This work was supported by funds from Grant DA02189 from the National Institute on Drug Abuse, Chemical Pharmacology ~ ~Grant GM ~ 709504 (R.A.o.), and Biomedical Research SupportGrant 2-S07-RR05586-15. We acknowledge helpful technical assistance in the drug discrimination assays provided by and Jenny Weeden* Evita Bynum, Donna Registry No. 2a, 90791-20-1;2a.HC1,90791-19-8; 2b, 9079122-3; 2b.HC1, 90791-21-2; 3, 90791-14-3; 3.HC1, 90791-13-2; 4a, 90791-15-4; 4b, 90791-16-5: 5a. 90791-17-6: 5b. 90791-18-7: 2,4,5-trimethoxybenzaldehyde, 4460-86-0;cyclopropyldiphenylsulfonium tetrafluoroborate, 33462-81-6; 2,5-dimethoxy-4methylbenzaldehyde, 4925-88-6.
Benzylamines: Synthesis and Evaluation of Antimycobacterial Properties Wolfgang R. Meindl? Erwin von Angerer,’ Helmut Schonenberger,*+ and Gotthard Ruckdeschel’ Institut fur Pharmazie der Universitat Regensburg, Lehrstuhl Pharmazeutische Chemie II, 0-8400 Regensburg, and Max-von-Pettenkofer-Institutf u r Hygiene und Medizinische Mikrobiologie der Universitat Munchen, 0-8000 Munchen 70, Federal Republic of Germany. Received December 27, 1983
The synthesis of benzylamines with various N-alkyl chains and substituents in the aromatic system as well as their evaluation on Mycobacterium tuberculosis H 37 Ra are described. The most active compounds in this test, N-methyl-3-chlorobenzylamine(19, MIC 10.2 bg/mL), N-methyl-3,5-dichlorobenzylamine (93, MIC 10.2 pg/mL), (103, MIC 6.4 pg/mL), also exhibited a marked inhibitory effect on Mycoand N-butyl-3,5-difluorobenzylamine bacterium marinum and Mycobacterium lufu used for the determinationof antileprotic properties. The combinations of 93 with aminosalicylic acid, streptomycin, or dapsone exert marked supra-additive effects on M. tuberculosis H 37 Ra. N-Alkylbenzylamines are compounds with a specific Scheme I action against myc0bacteria.l They lack any activity R2 R3 R‘ R3 R‘ R3 against fungi and Gram-positive and Gram-negative bacmethodA R ’ N H e method B L i A l H i teria.2 Structural manipulations like the replacement of or method C NoBHq one methylene hydrogen by an alkyl group (l-phenyl-l(alky1amino)alkane type),l the connection of two N-alCH CH kylbenzylamines in this position (N,N’-dialkyl-1,2-diphenylethylenediamine type),l the shortening or IengthI1 II0 Y ening of the distance between the nitrogen function and I the aromatic system (119, 120),460rthe transformation of method D HpNOH R’ the secondary amino group into a tertiary one1 led to a loss of antimycobacterial activity. However, primary benzylR‘ R3 R 2 R3 amines showed an inhibitory effect on mycobacteria (Tamethod B L i A I H 4 bles 1-111). The degree of the antimycobacterial activity of N-alkylbenzylamines is influenced by the length of the N-alkyl chain. N-Butylbenzylamine ( 5 ) proved to be the CH CH2 most active compound in a series of ring unsubstituted I ItN benzylaminesl (Table I). Ramification of the N-alkyl chain “2 in the a-position to the nitrogen atom led to a loss of I efficacy as shown by comparison of the activities of comOH pounds 14,23, and 46 with those of 13,22, and 45 (Table The in vivo activity of 15 in the M. tuberculosis H 37 11). An unequivocal correlation between the length of the Rv infected mouse was comparable with that of streptoN-alkyl chain and the antimycobacterial activity was not mycinS3Other substituents did not considerably increase found (Table I). The introduction of halogen, especially the efficacy of N-buty1benzylamine.l 4-(Aminomethy1)of fluorine, into the meta position of the most active un2-hydroxybenzoic acid (HOMO-PAS) and its N-alkyl desubstituted compound, N-butylbenzylamine, enhanced its antimycobacterial activity (N-butyl-3-fluorobenzylamine rivatives were not capable of inhibiting the growth of M . tuberculosis.4 These results are in contrast to the findings hydrochloride (15)) MIC 30 pg/mL M. tuberculosis H 37 of Kuhn et a1.: who described an effect of HOMO-PAS Ru,l MIC 25.6 pg/mL M. tuberculosis H 37 Ra (Table 11), MIC 8 pg/mL M. tuberculosis H 37 Ru (INH r e ~ i s t a n t ) , ~ MIC 32 pg/mL M . tuberculosis H 37 Rv (thiosemi(1) Ruckdeschel, G.; Stransky, D.; Schonenberger, H. Pharmazie carbazone re~istant)~). 1976, 31, 374. (2) Ruckdeschel, G.; Stransky, D.; Schonenberger, H. Pharmazie 1977, 32, 119. Institut fur Pharmazie der Universitiit Regensburg. (3) Otten, H., unpublished results. t Max-von-Pettenkofer-Institut fiir Hygiene und Medizinische (4) Meindl, W.; von Angerer, E.; Ruckdeschel, G.;Schonenberger, Mikrobiologie der Universitat Miinchen. H.Arch. Pharm. (Weinheim, Ger.) 1982, 315, 941.
9 I
P -9
f
0022-262318411827-1111$01.50/0 , , , 0 1984 American Chemical Society
i
1112 Journal of Medicinal Chemistry, 1984, Vol. 27, No. 9
Meindl et al.
Table I. Tuberculostatic N-Alkylbenzylamine Hydrochlorides Q--cHz+NHZR1
ci
-
compd R' synth methodn yield,* % mp,c "C formulad MIC,' pg/mL If H 262-263 CTHgN-HCl 76.8 2g CH3 A, B 45 178 CgH11N.HCl 102.4 3h C2H6 A, B 65 182 CSH13N.HCl 204.8 4' C3H7 A, B 73 184 CloH1SN.HCl 51.2 5k C4HB A, B 52 237 C11H1,N.HCl 51.2 6l C6Hll C12HlgN.HCl 153.6 7' C6H13 C13H21N.HCl >256.0 8' C7H16 C14H28N"Cl >256.0 9' C8H17 CiEH25N.HCl >256.0 "Capital letters refer to synthetic methods A and B in the Experimental Section. *No effort was made to optimize yields. "11 compounds were crystallized from MeOH/ether. dAll compounds were analyzed for C, H, and N within f0.40% of the calculated values. "Minimum inhibitory concentration for growth inhibition of M . tuberculosis H 37 Ra. 'Compound was purchased from EGA Steinheim, West Germany. gSee ref 7. hSee ref 8. 'See ref 9. kSee ref 10. lSee ref 11.
Scheme I1
Scheme I11 W
O Z N T N O Z
HC I /EtOH
BrBr
T
N
I
z
CITNH~ CHzOH
CHZ
o
CHzOH
I H N O ? CuCIz
"2
I
42
on M . tuberculosis type gallinaceus Stamm Stein. The marked in vitro activity of 15 and of other N-alkylbenzylamines against M. marinum2 indicates the possibility of a leprostatic activity of these compounds, which can be determined using M . marinum infected mice.6 Kuhn, R.; Zilliken, F.; Trischmann, H. Chem. Ber. 1950, 83, 304. Freerksen, E.; Rosenfeld, M. 2.Tropenmed. Parasitol. 1973, 24, 17. Emde, H. Arch. Pharm. (Weinheirn, Ger.) 1909,247,351. Kraft, F. Ber. Dtsch. Chem. Ges. 1890,23, 2780. Shapiro, S. L.; Parrino, V. A.; Freedman, L. J. Am. Chern. SOC. 1959,81, 3728. Einhorn, A.; Pfeiffer, H. Justus Liebigs Ann. Chem. 1900,310, 225. Eder, M. Ph.D. Thesis, University of Munich, 1962. von Braun, J.; Michaelis, R. Justus Liebigs Ann. Chem. 1933, 507, 1. von Braun, J.; Kuhn, M.; Weismantel, J. Justus Liebigs Ann. Chem. 1926,449, 249. Pallos, F. M.; Brokke, M. E.; Arnekley, D. R. Ger. Offen. 2 218097, 1972. Schonenberger, H.; Thies, H.; Rappl, A. Arch. Pharm. (Weinheim, Ger.) 1967, 300, 483. Kindler, K. Arch. Pharm. (Weinheirn, Ger.) 1927, 265, 389. Yokohama, M.; Kawano, S.; Sugiyama, K. Eisai Co. Ltd., Japan 15674,1961. Wellcome Foundation Ltd., Brit. 881625, 1961. Ehrlich, F. Ber. Dtsch. Chem. Ges. 1901, 34, 3366. von Braun, J.; Michaelis, R.; Spanig, H. Chem. Ber. 1937, 70, 1241. Tiffeneau, M. M. Bull. SOC. Chim. Fr. 1911, 9, 825. Baltzly, R.; Phillips, A. P. J. Am. Chem. SOC. 1949, 71, 3421. Holmes, E. L.; Ingold, C. K. J. Chern. SOC. 1925, 127, 1800. Freifelder, M.; Yew Hay, N. J . Pharm. Sci. 1965, 54, 1204. Elslager, E. F.; Johnson, J. L.; Werbel, L. M. J . Med. Chem. 1981, 24, 140. DeBrunner, R. E.; Blake, E. S.; Webster, J. A. Monsanto CO., Belg. 672 659, 1966. Brabander, H. J.; Wright, W. B. J. Org. Chem. 1967,32,4053. MBtaver. M.: Dat-Xuonn, N. Bull. SOC.Chim. Fr. 1954,21,615. Surrey, A. R:; Rukwid, M,K. J. Am. Chem. SOC. 1955,77,3798.
~
Fe/HCi
"02,
9
~
0
2
CHzOH
C HzOH
1
1
1
HBF4
MnOz
c'9-N CH
II0
TF
-
Mno2
CHzGH
CH
II0
These findings prompted us to study the structure-activity relationship in the class of benzylamines in detail hoping to find drugs that can be used in combination with dapsone (DDS) or other antimycobacterial compounds in the therapy of leprosy. By appropriate variation of the N-alkyl chain and introduction of ring substituents, compounds with a strong activity against various mycobacteria were obtained (Tables I1 and 111). Chemistry. The synthesis of the benzylamines except compound 42 started from the substituted benzaldehydes. They were converted into the corresponding imines with alkylamines (method A) or into oximes with NHzOH (method D). The oximes were reduced with LiA1H4 (30) Sharadamma, H. S.; Maddi, G. B.; Kulkarni, S. N.; Sathir, P. B.; Nargund, K. S. J. Karnatak Uniu. 1967, 2, 19. (31) Angyal, S. J.; Morris, P. J.; Rassack, R. C.; Waterer, J. A. J. Chern. SOC. 1949, 2704. (32) Biel, J. H.; Nuhfer, P. A,; Drukker, A. E.; Mitchell, T. F.; Conway, A. C.; Horita, A. J . Org. Chem. 1961,26, 3338. (33) Wilcox, M. Fr. Demande 2295011, 1976. (34) Marhold, A.; Klanke, E. Ger. Offen. 2526652, 1976.
Journal of Medicinal Chemistry, 1984, Vol. 27, No. 9 1113
Antimycobacterial Properties of Benzylamines
Table 11. Tuberculostatic Monosubstituted N-Alkylbenzylamine Hydrochlorides
z$qc ;
CI -
H N ti, R'
compd
R1
R2 synth method" yield,* % mp, OC formulaf MIC,B pg/mL CvHaFN-HCl 51.2 273' 3-F 19.2 206' 56 3-F 51.2 203' 64 3-F 25.6 217-218' 70 3-F >256.0 67 173e 3-F 25.6 3-F 51.2 58 260' 3-F 17.9 227c 341 >256.0 134' 64 2x1 10.2 170' 60 341 >256.0 198' 45 4-C1 25.6 180' 54 341 17.9 64 210' 3-C1 >256.0 17ge 50 341 >256.0 241 64.0 341 >256.0 71 246' 4-C1 51.2 55 240' 3-C1 32.0 216" 3-Br 45 19.2 50 156c 3-Br >256.0 47 219' 4-Br >256.0 2-Br 67 138' 64.0 171' 3-Br 43 51.2 3-Br 50 199' >256.0 2-Br 38.4 3-Br >256.0 4-Br 179.2 3-1 153.6 64 179' 3-1 >256.0 69 195c 3-1 256.0 51 3-1 193' 204.8 63 19oc 3-1 128.0 3-CN 54 220d 51.2 3-CN 67 156' 102.4 17lC 3-CN 70 76.8 75 210' 3-CN >256.0 3-CN 67 203c 51.2 56 210c 3-CN 76.8 3-CN 70 198-199' >256.0 4-OCH3 >256.0 4-OCH3 57 171' >256.0 4-OCH3 47 17gd 4-OCH3 45 194e >256.0 179.2 &NO2 256.0 %NO2 63 183d >256.0 %NO2 59 241d >256.0 3-NO2 63 242d %NO2 50 >256.0 223d %NO2 200c >256.0 %NO2 55 211d >256.0 3-CF3 52 >256.0 176c 3-CF3 171' 37 >256.0 3-CF3 62 189-190' >256.0 3-CF3 64 210c >256.0 3-CF3 57 169-171' >256.0 3-CF3 184' 65 >256.0 3-CF3 41 180-181' >256.0 4-CF3 30 276= >256.0 4-CF3 44 205-207' >256.0 4-CF3 46 249' >256.0 4-CF3 37 247-249' >256.0 71 4-CF3 >256.0 "Capital letters refer to synthetic methods A-D in the Experimental Section. bYield of the pure product; the yields refer to the methods mentioned under synthetic methods; no effort was made to optimize yields. Recrystallization solvents were MeOH/ether. These compounds were crystallized from EtOH. "he compound was crystallized from acetonitrile. fAll compounds were analyzed for C, H, and N within &0.4% of the calculated values. #Minimum inhibitory concentration for rowth inhibition of M. tuberculosis H 37 Ra. Compounds were purchased from EGA Steinheim, West Germany. 'See ref 1. kSee ref 12. See ref 13. mSeeref 14. "See ref 15. Osee ref 16. PSee ref 17. q See ref 18. 'See ref 9. See ref 19. :See ref 20. " See ref 21. "See ref 22. See ref 23. See ref 24. YSee ref 25. E See ref 26. On See ref 27. 1O h 11 12 13 14 15' 16 17h 18k 19k 20' 21m 22m 23 24' 25' .26" 27 28O 2gk 3w 319 32' 33 34' 35' 36' 37h 38 39 40 41' 42* 43t 44 45 46 47 48 49h 50" 51" 52 53h 54w 55 56 57 58' 59 60' 61 Y 622 63"" 64 65 66 67' 68 69 70
L
I
I" Y
f
(method B) and the imines with LiA1H4 (method B) or N&H4 (method C), depending on the substituents in the
aromatic system (Scheme I). For the synthesis of compound 42,3-(bromomethy1)benzonitrilewas treated with
1114 Journal of Medicinal Chemistry, 1984, Vol. 27, No. 9
Meindl et al.
Table 111. Tuberculostatic Disubstituted N-Alkylbenzylamine Hydrochlorides
R5
compd 72f 738 74h 75h 76' 77k 78 79 80 81 81' 83"'
R' H
R2 241 241 2x1 241 241
R3 441 4-C1 4-C1 4-C1 4-C1 541 5-C1 5-C1 5-C1 5-C1 6-C1 641 6-C1 6-C1 6-C1 441 441 441 4-C1 4-C1 5-C1 541 5-C1 541 5-C1 5-C1 541 5-F 5-F 5-F 5-F 5-F 5-F 5-F 5-F 5-F 5-F 5-F 5-NO2 &NO2
synth method"
yield) %. 50 64 51 47
mp: OC 260 185-186 185 155
formulad C,H,CloN.HCl
MIC,e pg/mL >256.0 >256.0 >256.0 >256.0 >256.0 76.8 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 >256.0 20.4 10.2 17.9 20.4 64.0 128.0 153.6 57.6 32.0 44.8 19.2 6.4 32.0 32.0 19.2 32.0 25.6 19.2 128.0 102.4 102.4 204.8 >256.0 >256.0 >256.0 >256.0
2-c1 62 250-253 2x1 171 58 2-c1 185-186 56 241 56 202 2x1 53 182 2x1 235 57 2x1 42 198 84" 2x1 62 207 85" 241 48 212 86' 241 87' 3-C1 40 234 8Sh 3-C1 54 228 89h 3-C1 230 73 90h 3-C1 232 67 91' 341 92k 341 45 264 93 3-C1 194 79 94 3-C1 248 67 95 3-C1 77 264 96 3-C1 248 75 97 3-C1 232 47 98 3-C1 47 183 99 3-F 41 270 100 3-F 231 49 101 3-F 47 209-210 102 3-F 249 48 103 3-F 281 50 104 3-F 277 45 105 41 341 211 106 3-C1 23 150 3-C1 107 45 193 51 3-C1 108 249 109 47 341 261 51 3-C1 110 173-175 111 63 3-C1 237 341 112 45 245 341 256 &NO2 113 49 441 114* 35 3-CF3 224 4-C1 115 30 3-CF3 179 47 3-CF3 176 4-C1 116 c2H5 4-C1 50 3-CF; 206 117 .. C3H7 57 179 3-CFe 4-C1 118 CA - Ho " aCapital letters refer to synthetic methods A-D in the Experimental Section. bYieldsof the pure product; the yields refer to the methods mentioned under synthetic method; no effort was made to optimize yields. "11 compounds were crystallized from MeOH/ether. dAll compounds were analyzed for C, H, and N within &0.4% of the calculated values. gMinimuminhibitory concentration for growth inhibition of M. tuberculosis H 37 Ra. f See ref 28. #See ref 29. See ref 9. 'See ref 1. See ref 30. See ref 31. "'See ref 32. "See ref 33. See ref 34. ~
hexamethylenetetramine. The ammonium salt was decomposed with HCl/EtOH (Scheme 11). 3,BDifluorobenzoic acid was synthesized according to Roe and Little.3s The reduction with LiA1H4 yielded 3,5-difluorobenzyl alcohol,36which was oxidized into the 3,5-difluorobenzaldehyde3' with use of MnOP The synthesis of 3-chloro5-nitroben~aldehyde~~ and 3-chloro-5-fluorobenzaldehyde started from 3,5-dinitrobenzyl alcohol (Scheme 111). This compound was partially reduced with ( N H 4 ) ~ S to give 3-amino-5-nitrobenzyl alcohol, which was converted into 3-chloro-5-nitrobenzylalcohol by the Sandmeyer reaction. (35) Roe, A.; Little, W. F. J. Org. Chem. 1955, 20, 1577. (36) Punja, N. Eur. Pat. Appl. 31199, 1981. (37) Prudchenko, A. T. Izv. Sib. Otd. Akad. Nauk. SSSR, Ser. Khim. Nauk. 1970,6,95. (38) Butler, G. B.; Carter, M. E. J. Am. Chem. SOC.1950, 72,2303.
The latter was oxidized with MnOz to form 3-chloro-5nitroben~aldehyde~~ or reduced with Fe/HC1 to give 3amino-5-chlorobenzyl alcohol; the amino group was replaced by a fluoro atom by thermal decomposition of the diazonium tetrafluoroborate to afford 3-chloro-5-fluorobenzyl alcohol. Oxidation with Mn02 yielded 3-chloro-5fluorobenzaldehyde. Biological Properties. The first step to improve the efficacy of benzylamines against mycobacteria concerned the kind and position of Substituents at the aromatic ring. We chose substituents that differ considerably in their lipophilic (a)and electronic (a) parameters: F, C1, Br, I, CF, (+a/+a); OH,OCH,, NH2 (-a/-u); CH3 (+a/-6); NOz, CN (-a/+u) (see ref 1 and Table 11). None of the compounds with substituents in the ortho and para positions was more active than the corresponding unsubstituted primary and secondary benzylamines; most of the
Antimycobacterial Properties of Benzylamines
Journal of Medicinal Chemistry, 1984,Vol. 27, No.9 1115
Table IV. Antimycobacterial Activity of the Benzylamines 19, 93, and 103 (MIC ~ g / m L ) ~ ~ M. tuberculosisa~c M. auiumavc M. marinuma*c M. smezmatis'pc M. 1ufubVd compd H 37 Ru SN 304 SN 1254 ATCk 607 L 209 19 12.0 16.0 24.0 32.0 32.0-64.0 32.0 4.0 4.0 4.0 93 3.0 32.0 8.0 3.0 6.0 103 4.0 "Lockeman medium with 5% bovine serum. bDuboe medium with 5% bovine serum. CFortest method, see ref 2. dFor test method, see ref 39. ~~
compounds have to be considered as inactive (MIC > 250 pg/mL, ref 1and Table 11). We suppose that substitution in the ortho and para position leads to a strong steric hindrance of the benzylamine receptor interaction. Accordingly, the benzylamines with F, C1, Br, NO2, and CH3 residues in the aromatic ring exhibited the highest activity for the meta isomers (ref 1 and Table 11). Only primary and secondary benzylamines bearing the +a/+u substituents F, C1, and Br in the 3-position exceeded the antimycobacterial activity of the corresponding unsubstituted compounds (Tables I and 11). The influence of the N-alkyl chain length (one to five C atoms) on the activity was relatively weak (Table 11). However, their analogues with branched N-alkyl chains in the a-position to the nitrogen atom lacked any antimycobacterial effects (Table 11). The most active compounds were N-methyl-3-fluorobenzylamine hydrochloride (11, MIC 19.2 pg/mL, M. tuberculosis H 37 Ra), N-methyl-3-chlorobenzylamine hydrochloride (19, MIC 10.2 pg/mL, M . tuberculosis H 37 Ra), and N-methyl-3-bromobenzylamine hydrochloride (29, MIC 19.2 pg/mL, M. tuberculosis H 37 Ra). In the series of benzylamines with the most interesting +a/+u substituents in the 3-position, the following order of activity was observed: C1> F > Br > I > CF3. The 3-I-substituted benzylamines are already less active than the corresponding unsubstituted compounds (Tables I and 11). 3-CF,-substituted benzylamines are inactive a t a concentration of 256 pg/mL (Table 11). The u values of the halogen residues are comparable to each other, but the a values increase in the order F
