HYPOGLYCEMIC AGENTS
Sept. 5 , 1959
4635
gave a melting point and mixed melting point with authentic tion without heating gave a solid and a yellow liquid. The solid was sublimed (5 hours, llOo, 50 p ) to give m.p. and acetaldehyde 2,4-dinitrophenylhydrazone of 167-168". mixed m.p. with starting material of 185-187'. The The refluxing reaction mixture was allowed to distil slowly liquid, upon distillation (b.p. 50" (80 p ) ) , gave four drops a t atmospheric pressure into a receiver. The distillate of an oil which solidified on standing. Recrystallization (1830 ml.) was extracted with ether to give 1.2 g. of an Oil from ethanol gave crystals which did not depress the melting upon evaporation. This was eventually combined with the point of the monobromide obtained by bromination in ethmain batch of oil obtained below. anol. T h e degraded material remaining in the pot was acidified (3) Methylation.-Slightly moist diazomethane ilt 50 by the dropwise addition, with stirring, of 400 ml. of conml. of ether from 42 millimoles of N - n i t r o s o m e t h y l ~ r e a ~ ~ centrated hydrochloric acid. Overnight sweeping produced no additional solid in the derivatizing solution. The acidic was added to a solution of 2.04 g. (12 millimoles) of the acid in a mixture of 50 ml. of ether and 8 ml. of absolute methether was washed with water, and the washings brought anol. rifter standing for one hour the yellow color had disto pH 7 and re-extracted. The combined ether layers were appeared. The solution was washed with 5% NaHCOa dried with Na2S0, and evaporated t o give an oil weighing 12.4 g. This oil was combined with that obtained from the solution; the washings yielding no solid upon acidification. This was followed by further washing with 10% hydroextraction of the aqueous distillate above and distilled chloric acid, then with water until the washings were neuthrough a micro-still: cut 1, 1.5 g . , b.p. i9-81" (20 mm.), ItZ5D 1.4182; cut 2, 2.2 g., b.p. 81-86" (20 mm.),n Z 6 D tral. The average yield of methylated acid from two identical runs was 485%. Distillation gave 1.5 g. of thc 1.4172. Values obtained for authentic triniethylpyruvic ~ methyl ether, b.p. 78.6-81.0' (0.16 mm.), n Z 61.4685. acid under the same conditions are: b.p. 84-85' (20 mm.), A1zal.*0 Calcd. for C10HI603: C, 65.19; H , 8.76; OMe, n Z 6 1.4168. ~ Samples of the fractions solidified when 16.8. Found: C, 64.39; H, 8.96; OMe, 1i.2. Theprodallowed t o stand exposed to the atmosphere, as did triuct reverted to starting matertal after one month as shown methylpyruvic acid. \Vhen heated with 2,4-dinitrophenylhydrazine and sulfuric acid in ethanol, the fractions gave by recrystallization and mixed m.p. determination. For comparison with the above results, a-ethyltetronic ethj-1 trimethylpyruvate 2,4-dinitrophenylhydrazone, idenacid mas methylated. Diazomethane from 0.101 mole of tical with material obtained under the same conditions from N-nitrosomethylurea in 125 ml. of ether was poured into ii authentic trimethylpyruvic acid: fine canary yellow needles (from alcohol-water), m.p. 163-164", in agreement with the cold solution of 4.0 g. (31.2 millimoles) of a-ethyltetroriic acidI3 dissolved in a mixture of 50 ml. of ether and 5 ml. of literature.** The dktihtiori pot residue was a brown solid, methanol. The color of the diazomethane was discharged which upon recrystallization from a 4 : 1 mixture of petroafter 15 minutes. Extraction and washing as before, leum ether-ether \vas found tci be trimethyllactic acid, m.p. followed by distillation, gave 700 mg. ( 1 8 7 0 of the methyl and mixed m . p . 86-87', ether of a-ethyltetronic acid, b.p. 75-79' (60 p ) , n Z 61.4833. ~ ( 2 ) Bromination. ( a ) In Ethanol.-Two grams (11.8 Anal. Calcd. for C?HloO3: OMe, 21.9. Found: OMe, millimoles) of the sanie acid esainined above &is dissolved in 10 ml. of ethunol and nionobrominated by the rapid addi- 20.1. On standing, a solid was obtained which did not detion of 23.3 inl. of j C 1 bromine in carbon tetrachloride press the melting point of the parent acid. (4) Attempted Hydrogenation.--So uptake of hydrogen (11.8 millimoles of B r ? ) . After 0.5 hour the discharge of gas was observed in 12 hours a t atmospheric pressure when the bromine color w a s alinost complete. Solvent removal was accompanied by the evolution of hydrogen bromide, as a solution of the acid in glacial acetic acid was shaken with detected by moist blue litmus paper. The oily residue was .$dams catalyst. Starting material was recovered. ( 5 ) Attempted Ozonolysis.-.A sample of the methyl distilled: cut 1 , 0.9 g., b.p. 78-81' (0.3 mm.); cut 2, 1.1 ether of the acid was dissolved in ethyl acetate and treated g., b.p. 81-83' (0.3 xnni.). Both fractions solidified upon with ozone from a discharge generator for four hours. Atstanding overnight. These were combined and recrystaltempted hydrogenation of the product a t atmospheric T h e monobromide lized from 95% ethanol. m.D. 5ti-57'. js insoluble in Gater, but dissolves with slight effervescence pressure using Pd(OH)2 on CaC03 gave no discernible presui %5CG SaHCOa solution. Anal.*2 Calcd. for C8Hx~03Br: sure drop after shaking for 10 hours. Starting material was reclaimed. C, 43.39; H, 5.26; Br, 32.08 Found: C, 43.41; H, 5.17; Br, 32.33. Acknowledgment.-G. H. D., Jr., wishes to ac(b) In Chloroform.-One gram (5.9 millimoles) of the acid in 180 ml. of purified chloroform a t ice temperature knowledge a grant-in-aid from the Hynson, Westcott, and Dunning Fund. mas treated rapidly with 1.0 g . (6.3 millimoles'l of bromine in 10 ml. of chloroform. After two days a t room tempera(29) F. A m d t , "Organic Syntheses." Coll. Vol. 11, John Wiley and ture the color of the bromine was not discharged. EvaporaSons, Inc., Xew York, X. Y., 1943, p. 166. BALTIMORE 18, MD.
( 2 8 ) P. Clarke, Ckrinislry 3 177dUT/l'y, 12633 (1954).
[COS1R I H C T I O S
FROM 7"E O K G A Y I C R E S E A R C H
Hypoglycemic Agents. BY
SEYMOUR
I,.
IV.
'-'j
LAHORATOKIES
OF 'THE
I:.
s. \.ITAMIN
? PIIARMACEUTICAL i
CORP.]
N1,N5-Alkyl- and Aralkylbiguanides
SHAPIRO, VINCEKT
A.
PARRINO A N D r,OUIs
FREEDMAN
RECEIWIDFEBRUARY 25, 1959 .4 series o f X1,T%ubstituted biguaiiides hare been prepared by condensation of the amine hydrochloride, RIRwTH.HCI, with the RirRI-substituteddicyandiamide. Reversal of the substituents on t h e initial reactants afforded t h e same biguanide. A greater bulk or number of substituents on the biguanide was associated with the formation of monopicrates rather thaii dipicrates and has been interpreted as restricting hydrogen bond formation and failure t o afford t h e cyclic cation I. Hyp?glycemic effects with these biguanides are noted, particularly when Y-methyl or Nb-dimethyl substituents are introduced 111 physiologically active N'-substituted biguanides.
I n proposing the intramolecular hydrogen-bonded, six-membered ring structure (I) for ,8-phenethyl-
biguanide and related N1-substituted biguanides, several assumptions had been made.4 With N1,N5-
(1) S . L. Shapiro, V. A. Parrino and L. Freedman, THISJ O U R N A L , 81, 2220 (19>9). ( 2 ) S. L. Shapiro, V. A. Parrino, E. R o g o a and I,. Freedman, i b i d . , 81, 3725 (1959). (3) S. L. Shapiro, V. A. Parrino and L. Freedman, i b i d . , 81, 3996 (1959).
( 4 ) Assuming t h a t within this series, peak hypoglycemic activity would be associated with similar molecular forms of the active species, the noted activity wherein RIand Ra are hydrocarbon radicals requires that the 5 2 of the biguanide be an imino nitrogen. Indications for a conjugated double bond system1 complete the bond distribution pattern a5 shown for I, with the four hydrogen atoms a t the h" and Ne
l l 3
4636
SEYMOUR
L. SHAPIRO,
\'INCENT
-1. PARRIXV
AND
LOUISFREEDMAN
Vol. 81
TABLE I H
I
SO.
R1
1 CHa 2 CHr 3 CZH54 CzHs5 n-CaHt6 i-CaHi7 i.C3H;8 C3HP 9 C3HVe 10 w G H r 11 i-CIHo12 i-CrH913 I I - C ~ H I I 14 i-CbHii15 i-C5Hli16 i-CaHI117 C6HaCHr 18 CsHsCHr 19 CsHaCHz20 CsHaCHz. 21 CsHaCHr 22 CsHsCHa23 CsHsCHr 24 CsHsCHz 25 CeHsCHp 26 CsHaCHr 27 CsHaCHr 28 CsHsCHr 28a CsHsCHz 29 CaHsCHr 30 CsHaCHr 31 CaHsCHr 32 CsHaCHr 33 CsHsCHaCHr 34 CaHsCHzCHz35 CsHsCHzCHr 36 CsHsCHzCHr 37 CsHsCHzCHr
KZ H H H H H H H H H H H H H
R4 CHsCHaCHr CHPCHSCHr CHr CHa CH3CHaCHr CHZCHr CHa H
R3
CH3CHr CH3CHr CHr CH3CHaCHr CHs C H3CHr CHr CHs I€ CH3H i-CsH~iH i CaHi; E€ H CH9 H H i-CsHuH H CHaCH3H CsHsCHr H H CsHaCHr H H CtiHjCHzCHr H CsHsCHz H CHa . CH3H C H r CHiH C H3r C H r CHs H CHa- i-CbHuH C H r CsHsCHr H C H s CsHaCHr CHI- CsHaCHr CHrCH3- C s H i C H r CHr C H s CsHsCHzCHr H C~HICHZCHZ- H CIhCH3H H CH2H H CHsH CHr CoHsCHzCH2- €1 H CsHoCHzCHr H H
HX HSOI 2HPic.d HC1 2HPkd HC1 HCl 2HPkd IICl ZHPic." HC1 HCI HPic." HCl HC1 HCl HPic.d €IC1 HC1 HCl HC1 HP~c. HCl "21 HCI ~ H P ~ €IC1 HC1 HCl HPic.' HC1 HPic.d HC1 HPkd HCl 2HPic.d HCl HCI HPic.d
-
M.p. ,a OC.
139-143 170-171 171- 17-1 166-1b7 166-1ti7 202-20:3 158- 139 169-171 148-149 107-109 178- 182 lil-172 157-159 1G9-1 i o 183-154 159-1 1;O 145-150 18'2-18 4 181-184 196-197 ~16%-1G3 113-211 152-153 170-174 C125-126 . ~ 193-194 188-190 2 15-2 17 147-148 207-209 99-101 199-201 1G5-166 133-140 174-175 201-203 17s-180 148-149
KSL A A
B
x I3
B A
R A L3
R A B
c B D E
4 F G A
G '4
n c A G
F A G
dnalysesc Carbon, yo Hydrogen, Yo Calcd. Found Calcd. Found 29.1 29.1 6.8 6.8 3.9 32.7 3.2 3.0 . 1 . 2 37.1 8.3 7.9 3Ll 31.9 3.4 3.3 4 0 . 5 40.2 8.8 8.9 40.5 41.3 8.8 8.8 36.5 3.7 3.7 36.2 40.9 41.0 7.8 7.5 36.7 30.4 3.4 3.6 -13.3 $3.0 9.1 9,l 9.1 9.0 43.3 43.2 40.6 5.4 5.7 41.0 46.9 46.1 9.4 9.2 45.9 46.1 9.4 9.3 51.6 10.2 9.9 51.9 4G.0 6.4 6.3 46.4 6.7 6.G 49.7 49.9 8.1 8.2 j(i.5 5O.G 7.1 7.4 51.7 51.9 6.3 6.3 50.4 fi0.5 51.1 51.1 4.4 4.3 6.7 61.6 8.7 61.5 6.7 7.1 61.7 61.5 7.4 7.1 48.3 18.8 3.4 3.6 40.6 40.8 7.5 7.3 53.4 53.2 57.8 57.8 8.4 8.3 61.5 61.G 6.7 6.8 52.7 52.9 4.6 4.G 62.5 62.7 7.0 7.0
A
F
c
lj2.5 53.5
62.2 53.6
7.0 4.9
7.0 4.9
C Z ~ H Z S N I I O I ~4 0 . 8 CnHrClNa 53.4 63.5 CisHzaClSs CaHiaXsO; 53.5
40.8 53.4 82.3 53.9
3.4 7.5 7.0 4.9
3.3 7.6 7.0 5.1
n
F F
I' G
-
Nitrogen, Yo Calcd. Found 40.8 40 3
25.0 33 7 :?:1 7 2+..-> 34.1
33.8 31.2 2k.l 33.9
31.6 31.G 27.0 29.7 29.7 2j.2
31.7 31.8 27.1 29.8 29.7 24.!4
29.0 23.5 27.4 22.0 21.9 21.1 21.1 25.6 22.7 26.0 22.5 21.1
28.Y 2.1.0
27.2 21.8 21.7 21.0 21.3 25.8 22.5 26.2 22.3 20.8
20.1 20.3
20.3 20.0
27.4
27.1
26.0 20.2 20.8
26.4 19.8 20.8
29.0
A = ethanol-hexane, B = acetonitrile, c = ethaa Melting points are not corrected. b RS = recrystallizing solvent: Analyses by Weiler and Strauss, oxnol, D = propanol-hexane, E = ethyl acetate, F = isopropyl alcohol, G = water. ford, England. HPic. = picric acid. CaHs = allyl. / Crystallizes as monohydrate.
unsymmetrically substituted biguanides, however, isomeric forms for I as typified by I11 and I V or N
"H'
Ri
Synthesis of the compounds was effected by fusion of the amine hydrochloride with the substituted dicyandiamide (see Table 11) following methods
11
mixtures of these are possible. Inspection of molecular models of such compounds suggests steric inhibition of hydrogen bond formation. In addition to examination of these factors, assessment of hypoglycemic effects with K1,N6substituted biguanides was indicated, and the compounds prepared are described in Table I. of the biguanide satisfying the site of high electron density a t the imino nitrogen a t NZ t o complete the sterically favorable six-menhered hydrogen-bonded ring. The dominant protonated species was assigned with the H + added to the RiRz-bearing nitrogen as a consequence of electron enrichment by the A groups.5 The structure I1 having only the two hydrogens on N4 available for hydrogen bonding is sterically less favorable. ( 5 ) hl. hi. Tuckerman, J. R. Mayer ahd 1'. C. Kachod. THISJ U U R NAL, 81, 92 (1959).
previously In three systems cxamined wherein the RIRz groups mere initially on the amine reactant and the R3R4 groups on the reactant dicyandiamide, when the substituent dispositioii was reversed, only a single compound
4637
HYPOGLYCEMIC AGENTS
Sept. 5, 1959
TABLE I1 SUBSTITUTED
No.
RI
RS
1 2 3"
CHr CHr i-CbH11H -(CHs)s4d CoH5CH2H 5 CsHsCH2CHr 6 CeHsCHiCHr H Melting points are not corrected. benzene; xylene. Ref. 28 reports
M . p.,'~ b OC.
3
DICYANDIAMIDES NCNHC1-I
Formula
I/
?r"
Carbon, % Calcd. Found
AnalyseHydrogen, To Calcd. Found
-. h'itrogen, Yo Calcd. Found
165-169 C4HsN:r 42.8 42.7 7.2 7.4 50.0 49.4 79-81*" C7HIIN( 54.5 54.8 9.2 9.1 36.3 36.4 166-16Sbb C,HI2S, 55.2 55.6 8.0 8.2 36.8 37.1 61.8 5.8 5.7 CgH10N4 62.1 101-104 63.1 6.4 6.5 29.8 30.2 CIOHI~N~ 63.8 107-109 63.8 6.4 6.5 29.8 29.8 CIOH~~N~ 63.8 114-115 b The recrystallizing solvent was ethyl acetate-hexant: unless otherwise shown; m.p. 166-168". Ref. 28 reports m.p. 108-109".
was isolated.6 This would indicate that a particular structural form for I would predominate, and whichever synthetic combination is used equilibration proceeds to yield this form. I t was noted that the N1,N6-substituted biguanides having N5-benzyl-h'5-methyl substitution were much more readily isolated than N5-methylbiguanides. Accordingly, the potential of debenzylations by hydrogenolysis was explored, and N1-benzyl-N1-methylbiguanidewas successfully reduced to N '-methylbiguanide, isolated as the dihydrochloride. However, with the more complex biguanide, compound 26, the hydrogenolysis failed, as i t did with N',N5-dibenzylbiguanide.g For compounds with particular group bulk or disposition, steric hindrancelo to hydrogen bond formation (I) was possible and is suggested by the results below. Throughout this work many of the biguanides have been characterized as picric acid derivatives. AS a rule, the only product isolated was the dipicrate." I n the present series (Table I ) many of the compounds yielded monopicrates, particularly (6) Reaction of the amine hydrochloride with the dicyandiamide proceeds through attack a t the cyano group as is shown in Scheme I and could conceivably yield the isomers shown.
as the bulk of the substituent increased on N5 (compounds 16, 21, 28, 30, 32, 38), sugggesting a structural influence. While the biguanides are diacid bases, i t has been shown' for the dihydrochloride, that the second mole of hydrogen chloride is not retained, as a consequence of the occupation of the second basic site by hydrogen bond formation. It would then be unlikely that formation of the dipicrate vs. the monopicrate would be a function of a locus of the second proton, l 2 or a function of atomic distances13 between nitrogen atoms. Our experimental pattern is just the reverse of that of Knott and Bre~kenridge'~ i n that with increasing bulk of the substituents, monopicrates are obtained, and the majority of the striictures which reflected less steric hindrance yielded dipicrates insoluble in water. lb The observations suggest that the formation of the dipicrate involves addition of one mole of picric acid in salt formation, and that the second mole of picric acid adds to the hydrogen bonded cation I as a molecular ~ o m p l e x . ~ ~Formation J~ of the monopicrate serves as an indicator for structures with steric inhibition of hydrogen bond formation as shown for I.lS
(12) T. R. Harkins and H. Freiser, TxIs JOURNAL, 77, 1374 (19551, discuss compounds which possess two basic nitrogen atoms t h a t combine with only one proton in basic solution. R1RzNH.HCl NCNHCNRaRa I11 (13) R. A. Abramovitch, J . Chcm. .Sot.. 3839 (1954). (14) R. F. K n o t t and J. G. Breckenridge, Can. J. Chem., Sa, 512 h" (1954), in an investigation of 2,2'-bipyridyl analogs obtained only R&T\".HCl iYCliHCNR1Hz IV monopicrates with their compounds, suggesting formation of a hydrogen bond between nitrogen atoms by the first proton which was sufSH ficiently strong to hold the molecule in cis configuration. AlternaRecently: isomeric forms for triguanides have been implied and detively, when they employed substituents which afforded sufficient scribed. steric repulsion t o such hydrogen bond formation, the two ring systems were free to rotate about the connecting bond and presumably as(7) K. N. Nandi and hf. A. Phillips, Chemisfuy b Indtlsfry. 719 sumed a coplanar trans conEguration and afforded dipicrates. (1958). (8) W.H. Hartung and R. Siminoff. "Organic Reactions," Vol. VII, (15) A single exception in the compounds studied was isolation of John Wiley and Sons,Inc., New York, N. Y., 1953, p. 263. the monopicrate of N'-isobutyl-N'.N,'-dimethylbiguanide (compound (9) Debenzylations of tertiary amine groups proceed far mnre 12). Under the usual experimental conditions, the dipicrate of this readily than secondary amines, ibid., p. 276. compound was apparently soluble in aqueous picric acid medium, and on addition of excess biguanide, the reaction mixture afforded the pre(10) (a) W. F. Forbes and J F. Templeton, Can. J . Chem., 36, 180 (1958); (b) S. Searles, M. Tamres and G . M. Barlow, THISJOURNAL, cipitate of the monopicrate. 7 5 , 71 (1953); (c) W. S. Frye, J . Chem. Phys., 21, 2 (1954); (d) A. H. (16) R. D. Kross and V. A. Fassel, THISJOURNAL, 79, 38 (1957). Rlatt,J.Org. Chem.,20,591 (1955); (e) G. J.Brealy and M . K a s h a , T ~ r s (17) Some measure of the difference in the mode of linkage in dipicrates of biguanides is to be inferred from T. Callan and N. Strafford, J O U R N A L , 7 7 , 4462 (1955); (f) K. Nakamoto, M. Margoshes and R. E. J . SOC.Chmm. I n d . (London), I S , 1 (1024), who observed that with the Rundle, ibid., 77, 6840 (1955); (g) C. G. Cannon. Speclrochim. A d a , dipicrate of o-tolylhiguanide, the second molecule of picric acid is only 10, 341 (1958); (h) E. W. Gill and E. D. Morgan, Nafure, 133, 248 (1959). loosely combined and loses some picric acid on recrqstallization from alcohol. I t should be pointed out t h a t many of our dipicrates were (11) I n one instance,' N'-benzyl-N'-phenylbiguanide afforded a monopicrate. T h e bulky character of the substituent groups sugstable upon recrystallization from organic solvents, including alcohol. jested a steric influence, and it was established that (2,8-dimethpl(18) Although none of the hibuanides derived from the compounds phenyl)-higuanide would afford a monopicrate by controlling the characterized as monopicrates (except compound 11) afforded practicable hypoglycemic activity, t h e structureactivity relationships quantity of picric acid used' although the dipicrate formed under the would suggest that the sum of the carbon content of the substituents usual conditions. I n contrast, using aqueous picric acid with 8was too high for effective hypoglycemia. phenethylbiguanidel only the dipicrate was isolated.
SCHEME I
+
+
II II
+
+
4G3S
SEYMOUR
L.
SH.\PIRO, V I N C E N T
/I.P A R R I N O -IKLI LOUISFREEDXW
Pharmacology.-The compounds were evaluated for hypoglycemic activity by methods previously described3and results are noted in Table 111. TABLE 111 HYPOGLYcRMIc
ACITTVITY
OF
S
1
,
S
6
-
S RI~
GUANIDES" " 4+
3t
1-
JL
0
13 3 , 20,29 :L5 24 13, 22, X 8,m 26 11,2:3 :i 1 18,27 a The compounds were evaluated subcutaneously in guinea pigs a t test levels corresponding t o '/j t o 1 / 1 ~of the LD,,,. (minimum lethal dose subcutaneous in mice). Numbers correspond with compound numbers in Table I and compounds are classified according to hypoglycemic response in terms of percentage rpduction of blood sugar from the normal blood sugar of the animal; 0 = less than 10% reduction; If = 10-2070 reduction; 2f = 21-35% reduction; 3 f = 36-607, reduction and 4+ = over 607, reduction. The following compounds were evaluated orally: 1,O; 10,2+; 1 7 , 3 + .
10 1;
1, 11 6,19
1
The data indicate that substantial hypoglycemic activity may be attained with methyl or dimethyl substitution in the N5-position in biguanides having active IC1-substituents. The use of larger substituents in the Nj-position in similar systems is associated with considerable diminution or disappearance of activity. Insofar as i t has been explored, the active N1,N5-substituted biguanides are orally absorbed less effectively than the corresponding N1-substituted biguanides. The fact that no practicable hypoglycemia is noted with tetrasubstituted compounds would dis'. count the significance of such structures as 1
x I-I
N
,/e\ R?
I'
H'
1.
1701,Sl
chloropheny1)-biguanide and K'-ethyl-N'-(Z-methylpheny1)-biguanide. It is of interest that these biguanides all manifested "biguanide" type resonance' in contrast to the "acetanilide" type resonance obtained with many of the other arylbigua~ nides. ~ ~ ~ ~ ~ ~ ~ ~ The designation of the hydrogen-bondedZucyclic2' cation'? I (protonated a t the lilR2-bearing nitrogen1,23in equilibrium with the dibasic cation' " \ a t the fiH of gastric juice and stomach contents) as the physiologically active form of these biguanide oralz6 hypoglycemic agents may be of c w siderable importance in pharmaoclogical studies on the mode of action of these compounds. Such studies, particularly with phenethylbiguanide (DBI), are being pursued in our laboratories'6 and elsewhere. Experimentalz7 N'-Benzyl-Nl-methyldicyandiarnide (Table 11, Compound 5) .-The dicyandiamides were prepared following the procedure of Redmon and Nagy'a and were typified by the preparation below. A mixture of 79.3 g. (0.5 mole) of benzylniethrlarnine hydrochloride, 55 g. (0.5 mole) of sodium dicyanamide in ,500 ml. of 1-butanol and 40 ml. of water was stirred and heated under reflux for 7 hours. T h e formed sodium chloride was separated and the filtrate concentrated to dryness. T h e residue was granulated under 200 ml. of water and filtered, yielding 8i.O g. (93%) of product. The biguanides described in Table I were prepared by the same general procedure and typical examples are giveii below. N1,N5-Dibenzyl,N1-methylbiguanide Hydrochloride (Table I, Compound 28).-.4 mixture of 16.2 g . (0.093 mole) of benzyldicyandiamide (Table 11, compound 4) and 11.4 g. (0.093 mole) of N-methylbenzylamine hydrochloride was heated gradually with stirring in an oil-bath. The mixture began t o melt at 55' and fused completelv a t 110' (bath 116'). Heating was continued over 30 minutes with gradual rise of bath temperature t o 149'. The cooled fusion product was recrystallized from water and yicldcd 18.0 g. (58%). The same rompound was obtained (no depression in mixed m .p.) by treating benzylamine hydrochloride with Pi-methylbenzyldicyandiamide (Table 11, compound 5 ) . In a similar fashion i t was established t h a t reaction of Pphenethyldicyandiamide and benzylamine hydrochloride, or reaction of benzvldicyaiidiamide with @-phenethylamitie hydrochloride, afford the same compound, N1-bcnzyl-S6(@-phenethyl)-bigunniclehydrochloride (Table I, c o m ~ x ~ u i i d 23).
The substituted dicyandiamides showed little or no hypoglycemic effect with the exception of P-phenethyldicyandiamide which was slightly active. Certain of the other previously described deriva120) Other publications involving consideration of hydrogen b m d tives' of phenethylbiguanide including 0-phen- formation in biological systems; (a) I,. Pauling and R. B. Corey, A ~ c h . ethylamidinourea, N-amidino-N'-P-phenethylurea, Biochem. R i o p h y s . , 66, 164 (1956); (b) W. Traub, Science, 129, 210 (1959); ( c ) E. R . Garrett, T K I SJ O U R N A L , 80, 1049 (1958): (d) I. AI. 8-phenethylurea, p-phenethylamine, 2-amino-4-P- Klotz and J. Ayers, i b i d . , 79, 4078 (1957). phenethylamino-s-triazine,2-arnino-4-P-phenethyl( 2 1 ) I n the anti-malarial arylbiguanide, the existence of a cyclic amino-6-methyl-s-triazineand 2-amino-4-P-phen- form of the molecule has been implied to have biological significance.' ( 2 2 ) T h e fact that the metabolically active form is an ion is of interethylamino - 6 - (a- hydroxyethy1)- s -triazine were in view of B. D. Davis, Avclt. Biochetn. Biophrs., 78, 197 (1958), without significant hypoglycemic activity. Pharma- est who states "a survey of metabolic reactions discloses that all known low cological studies of the in vitro hydrolysis product, molecular weight and water-soluble biosynthetic intermedicites Iiossess 0-phenethylguanidine, have been described in yroups that are essentially completely ionized a t neutral PH." ( 2 3 ) Protonation a t this site should substantially reduce f a t soludetaillg and show a complicated pattern involving bility; W. D. Dettbarn, I. B. Wilson and D. Sachmansohn. ,Sfit'ufe, hyper- and hypoglycemia. ias, 1275 (isxt). Among the arylbiguanides, the most active com(21) Recent work has indicated t h a t basic drugs are concentrated in pound was (2,4,6-trimethylphenyl)-biguanide which the gastric juice; P. A. Shore, B. B. Brodie and C. A. hi. Hogben. .i. Phnvmacol. E x p . T h e v a p . , 119, 361 (1957). showed 4+ hypoglycemia a t 50 mg./kg. (sc.) (25) While the hypoglycemic effect was noted in a broader structural ( L D m i n . 200 mg./kg.), although i t proved to be inrange, the additional requisite of oral effectiveness has restricted the active when tested per os a t SO mg./kg. Other structural scope of active compounds largely to those wherein RI is arylbiguanides which showed 2+ activity were CrCa alkyl and w-CaC8 aralkyl. (26) G. Ungar,S. Psychoyos and H. Hall, Metabolism, in press (1959). (2-methyl-4-chlorophenyl)-biguanide,(2-methyl-5(19) G. Kroneberg and K. Stoepel, Arzneamitlel-Forsch,. 8, 470 11958).
(27) Descriptive data shown in the tables are not herein reproduced. (28) B. C. Redmon and D . E. Nagy, U.S . Patent 2,455,807 fDec. 7, 1948).
DIAZOACETYL A$NALOCS
Sept. 5, 1959
The alternative synthetic paths also yielded the same compound in the preparation of N1-bemzyl-K1-methyl-N6(6-phenethy-1)-biguanide hydrochloride (Table I, compound 31). Methylbiguanide Dihydrochloride (by Catalytic Debenzylation of N1-Benzyl-N1-methylbiguanide Hydrochloride) .A solution of 12.0 g. (0.05 mole) of W-benzyl-W-methylbiguanide hydrochloridea in 120 ml. of ethanol and 50 ml. of water was treated with a suspension of palladium-oncarbon (from 1.0 g. of palladium chloride in 35 ml. of water and 0.5 ml. of 3 N hydrochloric acid, and 10.0 g. of carbon (Darco)) and hydrogenated a t 45’ under 50 lb. hydrogen pressure for 22 hours in the Parr hydrogenator. A t this point two equivalents of hvdrogen was absorbed. The catalyst was removed and the filtrate (strong toluene odor) was evaporated t o dryness. The residue of 6.45 g. was leached with 900 ml. of acetonitrile and the insoluble portion of 5.65g. was recrystallized (ethanol) to yield 3.23 g. of a mixture of mono- and dihydrochloride of the product. One gram of this mixture was dissolved in 20 ml. of methanol and 2.5 ml. of 3 S hydrochloric acid and was evaporated to dryness. The residue recrystallized (ethanol-hexane) gave 0.92 g. of product which melted at 225” dec. Anal. Calcd. for C3HllC12N5:C , 19.2;H, 5.9; S , 37.2. Found: C, 19.2; H,5.9;5,363. The dipicrate melted a t 198-199’. Anal. Calcd. for ClsHiiS11014: C, 31.5; H,2.7. Found: C, 31.0; H, 2.9. Similar attempted debenzylations with compounds 20 and 26 of Table I were unsuccessful and afforded only recovery of the initial reactant.
OF
4639
CHLORAMBUCIL TABLE IV SPECTRA OF BIGUAYIDES~ KO.b
17 19 20 24 26 28 29
a
Am..,
ma
e
x
236
10-8
16.5
236 237 238 240 240 244
14.6 17 8 15.5 17 9 17.4 18 9 The spectra were determined in water in the Beckman
DK recording spectrophotometer, using 1-cm. cells. * The compounds correspond to compound numbers in Table I. Ultraviolet Absorption Spectra.-The spectra of some of the compounds have been described in Table IV. Relative to the spectrum of p-phenethylbiguanide hydrochloride,l it is noted that increasing substitution in the N1,N6-substituted biguanides is associated with bathochromic and hyperchromic effects.
Acknowledgment.-The authors are grateful to Dr. G. Ungar for the results of the hypoglycemic tests reported herein, and to M. Blitz for the ultraviolet absorption spectra. YOSKERS1, 1;.Y.
[COTTRIBUTIOT FROM
THE
DEPARTMENT OF BIOLOGICAL SCIEXCES, STANFORD RESEARCH IXSTITUTE I
Potential Anticancer Agents.l
XX. Diazoacetyl Analogs of Chlorambucil
BY K. *4. SKINNER, HELENF. GRAM,CAROL W. MOSHERAND B. R. BAKER RECEIVED FEBRUARY 12, 1959 3-(p-Diazoacetylphen~1)-propionic acid (1-1)was synthesized from methyl hydrocinnamate ( I ) via the key intermediate, 3-(p-glycylphen7.l)-propioriic acid hydrochloride (V). Attempts to synthesize 3-(p-diazoacetamidophenyl)-propionic acid (XV) by two different routes failed on the last step. However, the corresponding methyl ester XVI was synthesized from 3-(p-aminophen~-l)-1,ropionic acid I1XI) TJZOmethyl 3-(p-glycylarninophenyl)-propionatehydrochloride (XI\’).
Interest in diazo derivatives of amino acids has Everett, et nl.,6 found that a series of p-[bis-(2been stimulated by the discoveries that azaserine? chloroethyl) -amino ] -phenylcarboxylic acids in(0-diazoacetyl-L-serine) and DON3 (6-diazo-5-oxo- hibit the growth of the transplanted Walker rat L-norleucine) show activity in inhibiting the growth Sarcoma-256, the most active compound being the of the Crocker Sarcoma-180 tumor. butyric acid derivative (chlorambucil). The Sarcolysine (3-{ p - [bis-(2-chloroethyl)-amino]- methyl and ethyl esters were also active. phenyl] -DL-alanine, “phenylalanine mustard”) has The hypotheis has been proposed7 that azaserine, been shown to be one of the more active nitrogen sarcolysine and chlorambucil might be considered mustard compounds. In addition, Luckj has found members of a broad class of anticancer agents conthat the two next higher homologs of phenylalanine sisting of metabolites bearing an alkylating group mustard are also effective on the Cloudman malig- that function by irreversible inhibition of the cornant melanoma (S-91). responding enzymes. This hypothesis suggests that 3-phenylpropionic acid or 4-phenylbutyric (1) This program is under the auspices of the Cancer Chemotherapy National Service Center, National Cancer Institute, Contract No. acid be considered as carrier (metabolite) groups for SA-43-ph-1892. For the preceding paper in this series cf. C. D. the diazoalkyl grouping characteristic of azaserine. Anderson, L. Goodman and €3. H. Baker, paper X I X of this series, This paper describes the synthesis of two such comTHISJ O U R N A L , 81, 3967 (1959). pounds, namely, 3-(p-diazoacetylphenyl)-propionic (2) C. C. Stock, D. A . Clarke, H . C. Reilly, S. M . Buckley and C. P. Rhoads, X a f u r e , 173, 71 (1954). acid (VI) and methyl 3-(p-diazoacetamidophenyl)(3) D. A. Clarke, H. C. Reilly and C. C. Stock, Abstracts of Papers, propionate (XVI). 129th Meeting, American Chemical Society, Dallas, Texas, April, 3 - ( p - Diazoacetylphenyl) - propionic acid (VI) JOURNAL, 80, 1956, p. 12-hf; H. A . DeWald and A . h f . Moore, THIS was synthesized in five steps from methyl hydrocin3941 (1958). (4) F. Bergel, V. C. E . Burnop and J. A. Stock, J . Chem. Soc., 1223 namate (I). A Friedel-Crafts reaction of acetyl (19.55); F. Bergel and J. A. Stock, ibid., 2-109 (1954); L. F. Larinov, A . S . Khokhlov, E. N.Shkodinskaia, 0. S. Vasina, V. I. Trusheikina and M. A. Novikova, Loncel, 269, 169 (1955). ( 5 ) J. M. Luck, Cancer Res., 17, 1071 (1957); H. E . Smith and J. M. Luck, J . Org. Clzem., 23, 837 (1958).
(6) J. L. Everett, J. J. Roberts and W. C. J. Ross, J . Ckem. Soc.. 2386 (1953). (7) H. F. Gram, Carol W. Mosher and B , P.Baker, paper XVIII of this series, THIS JOURNAL, 81, 3103 (1859).
