April 20, 1960
ALKYLATING AGENTSFROM 6-AMINO-6-DEOXY-D-GLUCOSE
2025
mole) and aqueous 5y0 potassium hydroxide (200 cc.) was refluxed for 6 hours. The resulting dark brown solution was cooled to room temperature, filtered to remove traces of unreacted starting material, and then cooled to 0". Concentrated sulfuric acid was added slowly, with vigorous shaking, to fiH 2, causing precipitation of a white solid, assumed to be the monopotassium salt since on ignition it left a residue which gave an orange flame test. Additional concentrated sulfuric acid (20 cc.) was added and the mixture was stirred for 5 minutes. The resulting solid precipitate no longer left a residue upon ignition. The mixture was extracted with ether, which dissolved the precipitatc. The ether extract was dried over calcium chloride and treated with charcoal. Most of the ether was evaporated off, light petroleum (b.p. 40-75") was added, and the solution was set aside overnight. The resulting precipitate was recrystallized from a fairly large amount of water, yielding the hydrate as white crystals (25.5 g.), which dem.p. 190-191.5". The hydrate hydrate at -120-150", shows a great tendency to oil out of concentrated water solutions and invariably oils out of mixtures of water with acetone, methanol, ethanol or dioxane. Ether-light petroleum (b.p. 40-75') may be used as a recrystallization solvent, but this combination does not produce as pure a product as does water. Drying of the hydrate a t 100' (1mm.) for 5 hours yielded 1,4,5,6,7,7-hexachloro-2,5 - norbornadiene-2,3 - dicarboxylic acid as white crystals (18.2 g., 0.0470 ?ole, 7870) m.p. 192.5-193.5", reporteds m.p. 162-163 ; Y O H -2690, 2550; YC-o 1715, 1694; vc-c.1637, 1606 cm.-l in Nujol; pKal is 3.75 and pK.2 is 6.15 in 50% (by volume) aqueous ethanol a t 32'. Anal. Calcd. for CeHzOnC16 (386.85): C, 27.94; H, 1,4,5,6,7,7-Hexachloro-2,5-norbornadiene-2,3-dicarbox0.52. Found: C, 27.97; H , 0.72. ylic Acid (XI).12-A mixture of dimethyl 1,4,5,6,7,7-hexachloro-2,5-norbornadiene-2,3-dicarboxylate(25.0 g., 0.0603 MINNEAPOLIS 14. MINN.
solution of dimethyl acetylenedicarboxylate (77.1 g., 0.543 mole) in hexachlorocyclopentadiene (292.1 g., 1.07 mole) was heated to 140°, but without further external heating the temperature rose rapidly to 230" while the light yellow solution became deep caryine red. The exothermic The solution was then reaction was cooled rapidly to 0 reheated a t 147' for 2 hours and finally briefly at 200". The solution was steam distilled until unreacted hexachlorocyclopentadiene no longer came over in the distillate. The residual aqueous mixture was extracted with ether and the ether extract was treated with charcoal and dried over calcium chloride. The dried ether solution was concentrated by evaporation and methanol was added, causing precipitation of dimethyl 1,4,5,6,7,7-hexachloro-2,5-norbornadiene2,3-dicarboxylate as white crystals (162.1 g., 0.391 mole, 720/,), m.p. 85-86'. The near ultraviolet spectrum in 9570 ethanol contained no maxima and only ill-defined inmp (log e 2.42) and 230 (3.58), with flections a t about 288 rising end absorption ( a t 220 mp log e is 3.80 and at 215, 4.00); Y C - c N 1720 in CHC13, 1725 in Nujol; vc-0 1629, 1596 ern.-' in CHCla and in Nujol. Anal. Calcd. for C I O H ~ O ~ (414.90): C~~ c , 31.84; H, 1.46. Found: C, 32.06; H , 1.57. B. By Re-esterification of the Diacid.-A solution of 1,4,5,6,7,7- hexachloro-2,5-norbornadiene-2,3 - dicarboxylic acid (2.0 g., 0.0052 mole) and boron fluoride etherate (Baker and Adamson technical grade, 5 cc.) in methanol (50 cc.) was refluxed for 5 hours. Sodium bicarbonatesolution (5y0, 30 cc.) caused the diester to precipitate. Recrystallization from aqueous 80% methanol gave white crystals (1.2 g., 0.0029 mole, 56y0), which gave no depression in mixed melting point with the sample from Part A. The infrared spectra of the two samples in Nujol were identical.
.
[CONTRIBUTION FROM
THE
DEPARTMENT O F BIOLOGICAL SCIENCES, STANFORD RESEARCH INSTITUTE]
Potential Anticancer Agents.
XXVII. Synthesis of Alkylating Agents Derived from 6-Amino-6-deoxy-D-glucose
BY ELMERJ. REIST, ROLAND R. SPENCER AND B. R. BAKER RECEIVED AUGUST14, 1959 The synthesis of 6- [bis-(2-chloroethyl)-arnino ]-6-deoxy-~-glucosehydrochloride ( X X I I I ) , 6- [(2-chloroethyl)-ethylamino 16-deoxy-~-glucosehydrochloride (XXIV) and 6-(2-chloroethylamino)-6-deoxy-~-glucose hydrochloride (XXV) via the key intermediate 3,5-O-benzylidene-l,2-~-~sopropyl~dene-6-O-p-to~ylsulfonyl-~-glucofuranose (XI) is described.
h rationale for the design of specific irreversible enzyme inhibitors was presented in paper XVII of this series2 This rationale proposed that substrates, properly substituted by an alkylating group, could fit the specific enzyme site for the substrate, then replace a nearby active hydrogen by alkylation, thus resulting in specific irreversible inactivation of the enzyme. Recent work devoted to the synthesis of nitrogen mustards in which the carrier is a metabolite, or resembles a metabolite, has resulted in several promising anticancer compounds, such as phenylalanine mustard (sarcolysin), m-phenylalanine
mustard, 2 , 4 chloranib~cil,~ uracil mustard,6 and benzimidazole m ~ s t a r d . As ~ part of the rationale,2 it was proposed that a nitrogen mustard, attached t o a substrate as a carrier, might be a type of compound that could operate as a specific irreversible enzyme inhibitor. If such is the case, then a "onearmed" mustard should be as good as, or possibly even better than, the corresponding "two-armed" mustard as an anticancer agent. Vargha and c o - w o r k e r ~ ~have ~ ~ synthesized several nitrogen mustards in the sugar series which have some interesting biological activities. N,NBis- (2-chloroethyl)-D-glucosaminehydrochloride (I)
(1) This work was carried out under the auspices of the Cancer Chemotherapy National Service Center, National Cancer Institute, Contract No. SA-43-ph-1892. For t h e preceding paper in this series, rf. E. J. Reist, P. A. H a r t and B. R. Baker, J. Org. Chem., 24, 1G40 (1959). (2) H. F. Gram, C. W. Mosher and B. R . Baker, THIS JOURNAL, 81, 3103 (1959). (3) F. Bergel, V. C. E. Burnop and J. A. Stock, J. Chem. SOL.,1223 (1955); L. F. Larinov. A . S. Khokhlov, E. N. Shkodinskaia, 0. S. Vasina, V. I. Trusheikina and M. A. Novikova, Loncel, 269, 169 (1955).
( 4 ) T. S. Osdene, D. N. Ward, W. H. Chapman and H. Rakoff, THISJOURNAL, 81, 3100 (1959). (5) J . L. Everett, J. J. Roberts and W. C. J. Ross, J. Chenz. SOC., 2386 (1953). (6) D. A. Lyttle and H. G. Petering, THIS JOURNAL,80, 6459 (1958). (7) E. Hirschberg, A. Gellhorn and W. S. Gump, Cancer Resrnvch, 17, 904 (1957). ( 8 ) L. Vargha, L. Toldy, d. Feh& and S. Lendvai, J. Chein. Soc., 805 (1959). (9) L. Vargha, 6. Feh& and S. Lendvai, {bid., 810 (1959).
2026
ELMERJ. REIST, ROLAND R. SPENCER AND B. R. BAKER
was found to be extremely toxic, with some antitumor activity.g On the other hand, 1,6-bis-(2chloroethy1amino)- 1,6-dideoxy- D -mannitol dihydrochloride (11) was found to be strongly cytoactive and to have tumor-inhibiting activity.s It may be HOCH?
r
-r--N
H C H 2 CI I 1 CI
@&- I
ROTH
-i '2HC1
L SHCH 1 CI-I 2 C 1
I1
I
--
."CI
__
~HCH,CH,CI 111
;e,-+ \:
1
0
IT', R-H or Ac
LS H CH 2 C H 2 C1
-
,CH, I 19,H
TsOCH,
"'fp
HO@HA
Vol. 82
01
1L
v
1' 5
1 i2C112
IiOCH
QO
01 'IP VI
chloride. Thus, it is desirable to have all the sugar hydroxyls blocked by groups which can withstand the basic conditions involved in the displacement of the 6-tosylate by amines and which can be removed by acid hydrolysis. A starting material such as 1,2 :3,5-di-O-isopropylidene-6- 0-p-tolylsulfonyl-Dglucofuranose, 13a 3,5-O-benzylidene-l,2,0-isopropylidene-G-0-methylsulfonyl- D - g l u c ~ f u r a n o s eor~ ~3,~ 5 -0-benzylidene- 1,2-O-isopropylidene-G-p-tolylsulfonyl-D-glucofuranose (XI) l4 meets these qualifications. The latter was chosen as the most convenient starting material. The synthesis of X I has been reported in several laboratories14 starting from 1,2-O-isopropylideneD-glucofuranose (VII), usually vkz 3,5-O-benzylidene-1,2-O-isopropy~idene-~-g~ucofuranose (VI1I). When this sequence was used in these laboratories,
argued that the mannitol mustard (11) meets the requirement of difunctionality which has seemed necessary for activity but it is certainly obvious that I1 is not a bis-nitrogen mustard in the normally accepted sense, in that the two arms of the mustard of I1 are not connected to the same nitrogen. It is interesting to note that l,G-bis-(2-chloroethylamino)-1,G-dideoxydulcitol(111)l o and N,N'-bis(2-chloroethyl)-1,G-hexanediamines were found to be inactive. In the light of these results, it is apparent that chloroethylamines other than the bis-chloroethylamines can have antitumor activity and that more work is necessary to correlate the activity of the bisH0(iH2 nitrogen mustards with monofunctional nitrogen mustards. Since the sugar nitrogen mustards represent a relatively unexplored territory and fit nicely into the carrier-antimetabolite concept,2s11 a program was instituted toward the synthesis of nitrogen mustards of a variety of amino sugars. The synthesis of three nitrogen mustards derived from G-amino-6-deoxy-~-glucose, namely, G- [bis(2-chloroethy~)-amino]-G-deoxy-~-g~ucose hydrochloride (XXIII), 6- [ (2-chloroethyl)-ethylamino]6-deoxy-~-glucosehydrochloride (XXIV) and G-(2chloroethylamino) -6-deoxy-D-glucosehydrochloride (XXV), is the subject of this paper. The synthesis of 6-amino-6-deoxy-~-glucoseand several of its derivatives has been accomplished by a variety of methods,12J3one of the smoothest of which utilizes the treatment of a 5,6-anhydro-~glucose or a potential 5,G-epoxide with ammonia, as illustrated in the sequence IV VI. This aptIproach suffers the disadvantage of yielding a G -Nrr2 aminoglucose derivative which possesses unblocked XI11 hydroxyls; these could easily cause undesired side reactions in a subsequent step involving the chlo- the product was nearly always contaminated rination of a 2-hydrosyethylamine with thionyl with varying amounts of by-product, probably the isomeric 5,G-0-benzylidene derivative IX1& which (10) L. Vargha, A m . N. Y.A c o d . Sci., 68, 875 (1958). (11) F. Bergel, ibid., 68, 1238 (1958). hindered crystallization of the desired VIII. Selec(12) (a) E. Fischer and K. Zach, Ber., 44, 132 (1911); (b) H. Ohle tive tosylation of VI1 to give 1,2-O-isopropylideneand L. Vargha, i b i d . , 61,79 (1923); ( c ) J. M. Grosheintz and H. 0 . L.
-
Vischer, THISJ O U R N A L , 70, 1470 (1848); (d) F. Cramer. H. Otterhach and H. Springmann, Bcr.. 92, 384 (1959). (13) (a) H. Ohle and L. Vargha, i h i d . , 6 2 , 2425 (1929); (b) B. Helferich and R. Mittag, i b i d . , 71B, 1585 (1938).
(14) (a) P. A. Levene and A. L. Raymond, ibid., 66B,384 (1933); (b) D. J. Bell, E. Friedmann and S. Williamson, J . Chem. SOC.,252 (1937); (c) A. S. Meyer and T. Reichstein, Hclo. Cham. Acta, 29, 139 (1946); (d) P. Karrer and A. Boettcher, i b i d . , SS, 570 (1953).
April 20, 1960
ALKYLATING AGENTSFROM
2027
~-A~'v~INO-~-DEOXY-D-GLUCOSE
6-O-p-to~y~su~fony~-~-g~ucofuranose (X) followed by treatment of the 6-tosylate X with benzaldehyde, a sequence described by Meyer and Reichstein,'" obviated this difficulty and gave a product (XI) which was easily purified by recrystallization. It was found that the crude tosylate X was satisfactory for the reaction with benzaldehyde to give V I 01 \' X I ; the over-all yield (3oYO)of X I from VI1 was IP IP higher from the crude tosylate X than from reXlV. R = CH,CH20H XVII.R= CHzCHZCl crystallized X (1470). XVIII R = CH2CH3 XV, R = CHZCH, There has been no mention in the literature conX\'I,H=H XIX. R = H cerning the displacement of the p-tolylsulfonyl group of X I with ammonia or amines. However, 6 , C H t O H the 6-amine XI1 has been synthesized by treatment CH,SCH2C€I,Cl of the 6-0-mesylate of VI11 with liquid ammonia for three weeks13b; displacement of the tosyl from OH @OH,HCl 1,2 :3,5-di-O-isopropylidene-6-0 -9 - toly~sulfony~-Dglucofuranose by methanolic ammonia a t 100' for HO €10 4 hours has also been r e ~ 0 r t e d . l ~Ammonolysis ~ OH OH of the 3,5-benzylidene-B-tosylate X I with methXX,R-=CHzCH2Cl XXIII R = CH>('IlzCl anolic ammonia a t 100' gave a S i % yield of the XXI, R=CHzCHIOII 130FH, XXIV R - CHzCH3 crude p-toluenesulfonic acid salt of 6-amino-3,5-0XXV R = H benzylidene-G-deoxy-1,2-O-isopropylidene - D - gluco(? furanose (XII). Conversion to the free base and recrystallization from ethanol gave pure X I 1 in FI 0 5270 yield; melting point and rotation were in C 1c' 5 k < H ,CI 12CI satisfactory agreement with those reported by C€IZ-CHZ Helferich and hlittag.13b Removal of the blockXXII ing groups with 3 N hydrochloric acid gave a quantitative yield of crude 6-amino-6-deoxy-~-glucose (XIII) hydrochloride, which was recrystallized in whereas more dilute acid gives the mustard contaminated with a second material, as shown by 41% yield, m.p. 163-165" decal5 Treatment of the tosylate X I with the various paper chromatography. This second material, ethanolamines a t 140-160° for 3 hours was adequate with Rg 0.95, is thought to be the partially hyto bring about the displaccnient of the 6-tosylate. drolyzed N - hydroxyethyl - N - chloroethylamine Thus, 2,2'-iminodiethanol gave an 81% yield of 3,5- (XX), since the dihydroxyamine X X I is an even 0-benzylidene-6-deoxy- 6 - [bis - (2 - hydroxyethy1)- slower moving material in this paper chromatogamino] - 1,2-0- isopropylidene - D - glucofuranose raphy system, having Rg 0.58. Although the bis-mustard hydrochloride XXIII, (XIV), 2-ethylaminoethanol gave 64% of 3,5-0benzylidene-6-deoxy - 6- [ethyl - (2 - hydroxyethy1)- as an anhydrous sirup, appears to be relatively amino] - 1,2 - 0 - isopropylidene - D - glucofuranose stable, an aqueous solution of X X I I I underwent (XV), and 2-aminoethanol gave 81% of 3,5-0- alteration over a period of days a t room temperabenzylidene-6-deoxy-6-(2-hydroxyethylamino) - 1,2- ture. Paper chr~matography'~of this solution 0-isopropylidene-D-glucofuranose (XVI), all as showed the conversion of X X I I I to be a material with R, 0.95 which was presumably the same as the crystalline solids. Chlorination of the bis-(hydroxyethy1)-amine second component of the dilute acid hydrolysis, and XIV with thionyl chloride in dichloromethane16 was assumed to be the partially hydrolyzed musgave crystalline 3,5-0-benzylidene-6- [bis-(2-chloro- tard XX. It should be noted that Varghag obethyl) -amino]-6-deoxy-1,2-O-isopropylidene- D - glu- served a similar behavior by aqueous solutions of cofuranose (XVII) in 70% yield. Hydrolysis of glucosamine mustard (IV), in that about half of the blocked bis-mustard XVII with 6 N hydro- the chlorine of IV was converted to chloride ions. He attributed this change to the formation of the chloric acid gave 6- [bis-(2-~hloroethyl)-amino]-6deoxy-D-glucose hydrochloride (XXIII) as an aziridinium cation XXII. Although i t seems likely uncrystallizable sirup which was homogeneous, that solutions of IV and X X I I I suffer the same as shown by paper c h r ~ m a t o g r a p h y ,with ~ ~ Rg type of decomposition, no further work has been 1.28, and was analytically pure. The use of 6 N done to determine the product of this decomposihydrochloric acid is a critical point. More con- tion. Treatment of the N-hydroxyethylamine XV centrated acid causes excessive decomposition, with thionyl chloride in dichloroniethane gave (15) M. J. Cron, 0. B. Fardig, D. L. Johnson, H.Schmitz, D. F. crystalline 3,5 - 0 - benzylidene - 6 - [(2 - chloroWhitehead, I. R. Hooper and R. U. Lemieux, THISJOURNAL, 80, 2342 ethyl) - ethylamino] - 6 deoxy - 1,2 - 0 - isopro(1958). pylidene-D-glucofuranose (XVIII) in 60% yield. (1G) The use of dichloromethane in place of chloroform as solvent Hydrolysis of the blocked mustard XVIII with 6 gave much less darkening during the chlorination; cf. ref. 4. (17) Paper chromatograms were run by the descending technique on N hydrochloric acid gave a quantitative yield of Whatman No. 1 paper with n-butyl alcohol-acetic acid-water (4: 1:5). 6 - [ ( Z - chloroethyl) - ethylamino] - 6 - deoxy - DSpots were detected by an aniline citrate spray. Glucose was used BS a glucose hydrochloride (XXIV) as an uncrystalstandard and spot locations were expressed as R, units, with glucose at Rs 1.00. lizable sirup which, however, was homogeneous as
4 \I
'3
?
/e
etoH
-
2028
ELMERJ. REIST,ROLAND R. SPENCER AND B. R. BAKER
VOl. 82
3,5-0-Benzylidene-6-deoxy-6[bis-(2-hydroxyethy1)shown by paper chromatography,I7 with R, 0.91, amino] -1,2-0- isopropylidene - D - glucofuranose (XIV).-A and was analytically pure. solution of 0.50 g. (1.1 mmoles) of 3,5-O-benzylidene-1,2-0It is interesting to note that whereas an aqueous ~sopropyl~dene-6-O-p-tol~lsulfonpl-~-glucofuranose ( X I ) in 5 solution of the bis-mustard XXIII had decomposed ml. of 2,2'-iminodietlianol was heated a t 150-160" for 3 in three to four days, as shown by paper chroma- hours. The solution was cooled to room temperature, the11 tography,'7 an aqueous solution of the "one- diluted with 15 ml. of chloroform. The chloroform solution washed with two 5-ml. portions of water. The water armed" mustard XXIV showed no evidence of de- was washes were back extracted with 3 ml. of Chloroform. The composition and, even after several weeks a t room combined chloroform layers were dried over magnesium sultemperature, the solution maintained its horno- fate, then concentrated to dryness in aacuo to give 0.39 g. (93%) of a white 120-123°. Recrvstallization ~ ~solid. . . ~m.n. ~ . .~~. . geneity, as shown by paper chromatography. frogethanol gave 0.21 g. (it%)of white crystal;, m.p. 127.5 Treatment of the blocked monohydroxyethyl- -128.5", 2.90 (OH), +19.23 (0.57, in water); ?A: amine XVI with thionyl chloride in dichloro- 7.25 (CH3), 9.25, 9.80 (C-0-C and C-OH), 13.14, 14.28 J! methane gave an uncrystallizable oil which was pre- (monosubstituted phenyl). A n d . Calcd. for C20H2~N0.i: C, 60.7; H , 7.39; N , sumably the blocked chloroethylamine XIX. Hydrolysis of XIX with G IV hydrochloric acid gave 3.54. Found: C,60.9; H,7.15; N,3.59. scale, 50 g. of tosylate XI gave 33 g. (81%) of the sirup G-(2-chloroethylamino)-G-deoxy-~-g~u-On a large product, m.p. 127.5-128.5", that was suitable for cose hydrochloride (XXV), which was essentially crystalline the next step. homogeneous on paper chromatography, l7 with 3,5-0-Benzylidene-6-deoxy-6[ ethyl-( 2-hydroxyethy1)amino] -1,Z-0-isopropylidene-D-glucofuranose(XV) was prcR, 1.00, and was analytically pure. ~
~~~
Dared in the manner described for XIV from 2.5 E. of 3,5-0benzylidene-1,2-O-isopropylidene - 6 - 0 - p - tolyliulfonyl - Dglucofuranose ( X I ) and 5 ml. of 2-( ethy1amino)-ethanol to 3,5-O-Benzylidene-l,2-O-isopropylidene-6-O-p-tolylsul-give, after two recrystallizations from Skellysolve C, 1.31 g. fonyl-D-glucofuranose (XI) was prepared by a slight modi- (64%) of crystalline material, m.p. 101-102". fication of the procedure described by Meyer and ReichA similar preparation had m.p. 105.5-106.5' [au]241) in which 45 g. of 1,2-O-isopropylidene-D-glucofur1-14.4' (0.570in ethanol); XE;:2.90 (OH), 7.22 (CHs), 8.88, anosele was converted to 47.5 g. (6l0/,) of the crude chloro9.24,9.82 (C-0-C, C-OH), 13.24, 14.32 p (monosubstituted form extracted tosylate as a tan-colored sirup. Treatment phenyl) and was analyzed. of the crude tosylate X with 150 ml. of freshly distilled benC, 63.3; H, 7.i0; N, 3.69. Anal. Calcd. for CZOHZ8NOB: zaldehyde and 40.0 g. of zinc chloride gave after recrystalFound: C, 63.3; H , 7.29; N, 3.76. lization from methanol, 28.5 g. (30y0 based on VII) of X I , On a large scale, 60 g. of tosylate X I gave 42.0 g. (767,) of m.p. 110-112", [ a ] D + l 5 ' (1% in chloroform); Ai:: 7.28 (CHs, OSOz), 8.50 .(OSOn), 9.20, 9.82 (C-0-C), 13.20, product, m.p. 104-105", that was suitable for the next step. 3,5-0-Benzylidene-6-deoxy-6-( 2-hydroxyethylamino)-1,2report 14.28 p (monosubstituted phenyl). Bell, et 0-isopropyfidene-D-glucofuranose (XVI) was prepared from m.p. 118" and [a10 +14.2" (chloroform). 60 g. of 3,5-O-benzylidene-l,2-O-isopropylidene-~-O-~6-Ammo-3,5-O-benzylidene-6-deoxy1,2-O-isopropylideneD-glucofuranose (XII).-A solution of 6.0 g. (13.0 mmoles) tolylsulfonyl-D-glucofuranose ( X I ) and 900 ml. of 2-amino0 - p - tolylsul- ethanol in the same manner as XI!' to give, after two recrysof 3,5-O-benzylidene-l,2-O-isopropylidene-6fOnyl-D-glUCOfUranOSe ( X I ) in 50 ml. of absolute methanol, tallizations from SFllysolve C, 33.9 g. (8lY0) of white cryswhich had been previously saturated with anhydrous am- tals, m.p. 137-139 . h similar preparation formed white needles upon recrysmonia a t O " , was heated in a Parr bomb a t 100" for 24 hours. At the end of this time, the reaction mixture was cooled, then tallization from ethanol-Skellysolve C, m.p. 138.5139.5" [ a I z 4 D +26" (0.5% in ethanol); ;:A : 2.91 (OH), 7.25 (CH3), poured into 100 ml. of a 1: 1 mixture of ice and saturated 9.30, 9.80 (C-0-C and C-OH), 13.19, 14.28 p (inonosubaqueous sodium bicarbonate and extracted with three 50-1111. portions of chloroform. The chloroform extracts were stituted phenyl) and was analyzed. A n d . Calcd. for C18H2sNOG.1/2C2HsOH:C, 61.0; washed with two 25-ml. portions of mater, then combined, dried over magnesium sulfate and evaporated to dryness i n H , 7.48: N, 3.75. Found: C, 60.9; H , 7.12; N, 4.12. mcuo t o yield 2.68 g. (67%) of a pale yellow, gummy solid. 3,5-O-Benzylidene-6- [bis-(2-chloroethyl)-amho] -6-deoxyRecrystallization from 6 ml. of absolute ethanol gave 2.07 gd 1 ,2-O-isopropylidene-D-glucofuranose (XVII).-A sohtion (52%) of white crystals, m.p. 120-125", [a"]D $24.5 of 1.0 g. (2.52 mmoles) of 3,5-0-benzylidene-6-deoxy-6(0.66% in chloroform). Helferich and hlittag13breported [bis-(2 - hgdroxr ethyl)-amino] -1,2-O-isopropylidene-~-glt1co~ D (chloroform). m a p . 127" and [ C Y ] ~+25.4" furanose (XIV) in 20 ml. of dry dichloromethane was treated 6-Amino-6-deoxy-~-glucose (XIII) Hydrochloride.--A with 3.0 ml. (26.2 mmoles) of thionyl chloride, then heated solution of 300 mg. (3.3 mmoles) of 6-amino-3,5-O-benzyli- under reflux for 1 hour. The reaction mixture was cooled, dene-6-deoxy-l,2-O-isopropy~ideiie-~-g~ucofuranose ( X I I ) in then poured onto 50 g . of ice. The resulting mixture was 1 nil. of methanol was heated on a steam-bath with 2 ml. of adjusted to pH 6 with saturated aqueous sodium carbonittc 3 Ar hydrochloric acid for 15 minutes. After being cooled, (about 40 ml. was required); then the dichloromethane was the aqueous solution was extracted with two 5-nil. portions separated, dried over magnesium sulfate and evaporated to of chloroform. The aqueous layer was lyophilized t o give dryness in zucuo to yield 1.04 g. (%yo)of a yellow sirup, 210 mg. of a white foam which gave a positive test with Bene- which crystallized slowly. Recry-;tallizatic,rl from 4 nil. of dict reagent for reducing sugars and traveled as a single spot absolute ethanol gave 0.76 g. (7070)of white needlcs, 1n.p. with XI 0.07 on paper ~hromatograpliy.'~ 108-109', La]% +S.9" (0.9% i n chloroform); A% no OH From a similar run, recrystallization of 700 nig. of the a t 2.9, 7.29 (CHa), 9.2-1, 9.85 p (C-0-C), 13.25, 14.33 w foam from 4 ml. of methanol gave 290 nig. (4170) of a white (monosubstituted phenyl). solid, n1.p. 172-176'. After two more recrystallizations Anal. Calcd. for C 2 0 ~ ~ 2 ~ C 1 2C, ~ \ ;55.5; o s : 11, 6.23; C1, ~ .--t from methanol, it had xx1.p. 163-165' dec., [ a ] +18" 16.4; N,3.24. Found: C, 55.5; II,G.38; C1,16.2; N,3.22. $33" (0.36O,i,in water); ::A: 2.98 ( O H ) ,6.20,6.62 (hTH3+),' 6- [Bis-( 2-~hloroethyl)-amino] -6-deoxy-~-glucoseHydro9.07, 9.45, 9.82 w (C-0-C and C-OH). Cron, et aZ.,ls 50.1" chloride (XXIII).-A solution of 610 mg. (1.4 mmoles) of reported m.p. 161-162" dec., [ a I z s ~ 23.0" -t 3,5-O-benzylidene-6- [bis-(2-chloroethyl)-amino~ -6-deoxy-1 ,1.0% in water), Rr 0.06." ( S V I I ) in 6 ml. of 6 1V Anal. Calcd. for C&IlsNOa.HC1: C, 33.4; H , 6.54; C1, 2-O-isopropylidene-~-glucofuranose acid was heated on a steam-bath for 30 minutes. 16.4; N, 6.49. Found: C, 33.4; H, 6.65; C1, 10.1; N, hydrochloric The reaction mixture was cooled, then extracted with 4 ml. of 6.39. chloroform. Lyophilization of the aqueous layer after treatment with Norit gavc 320 xng. (67cjC) of a white foam. Papcr (18) hlelting points were tokcti 011 a l;isher--Johus apparatus and are chromatography17 shonwl a strong spot a t I Q 1.35. uncorrected. Optical rotations were measured with a Standard l'olarornetcr mr,del L ) attachment to the Beckmau D U spectruphotom. Anal. Calcd. for C11,€I~1C12NOs~l-ICI: C, 35.3: H, 5.92; cter calibrated with standard s i i c n w suliitions. C1, 31.2; 3,4.12. I'ouiitl: C, 35.2; H , 5.8; C1, 31.4; N , 4.27. ( l Y ) S. C. Laland, r l i l a C h c m S c o i i d . . 8 , S(i6 ( I Y S ) .
Experimental17
+
+
April 20, 1960
ORGANOMAGNESIUM COMPOUNDS O N 3-ARYLRHODANINES
2029
3,5-0-Benzylidene-6- [ (2-chloroethyl)-ethylamino] -6-de- dene-6-deoxy-6-(2 - hydroxyethyl-amino) - 1,2 - 0 - isopropyli(XVIII).-A solu- dene-D-glucofuranose (XVI) and 2.53 ml. (35.2 mmoles) oxy-I ,2-O-isopropylidene-~-glucofuranose tion of 2.0 g. (5.26 mmoles) of 3,5-0-benzylidene-6-deoxy- of thionyl chloride in 20 ml. of dry dichloromethane was heated a t reflux for 1 hour. The reaction mixture was al6-[ethyl-(Z-hydroxyethyl)-amino]-1,2- 0 - isopropylidene - Dglucofuranose (XV) in 50 ml. of dry dichloromethane was lowed to cool to room temperature, then it was diluted with 10 ml. of dry dichloromethane and added dropwise with treated with 2.0 ml. (27.7 mmoles) of thionyl chloride at reflux temperature for 1.5 hours. The solution was cooled, stirring to 75 ml. of saturated aqueous sodium carbonate. diluted with 20 ml. of dry dichloromethane, then added drop- The dichloromethane layer was separated and the aqueous wise with stirring to 200 ml. of saturated aqueous carbonate. layer was extracted with 20 ml. of dichloromethane. The The dichloromethane layer was separated, washed with 25 combined dichloromethane layers were dried over magnernl. of water, dried over magnesium sulfate, then evaporated sium sulfate, then evaporated to dryness in vacuo to yield to dryness in vacuo to yield 1.82 g. (86%) of a tan sirup. the blocked mustard X I X as an oil. The crude blocked mustard X I X was heated on a steamRecrystallization from absolute ethanol gave 1.26 g. (607?) of white c stals, m.p. 65-67', [ c Y ] * ~ D+4.8' (0.5% in bath with 10 ml. of 6 N aqueous hydrochloric acid for 30 ethanol); no OH a t 2.9, 7.25 (CHa), 9.12, 9.26, 9.84 minutes. After being cooled to room temperature, the acid solution was extracted with two 10-ml. portions of chloro(C-0-C), 13.20, 14.30 p (monosubstituted phenyl). Anal. Calcd. for Ct0H&1N05: C, 60.3; H , 7.04; C1, form, then the aqueous layer was lyophilized to give 0.96 g. of a brown foam which contained a major spot on paper 8.92; N, 3.52. Found: C, 60.3; H, 7.30; C1, 8.86; N, c h r ~ m a t o g r a p h y with , ~ ~ R, 1.02 plus a trace component with 3.51. R, 0.60. 6- [ ( 2-Chloroethyl)-ethylamino]-6-deoxy-~-glucosehydrochloride (XXIV) was prepared from 3,5-0-benzylidene-6Anal. Calcd. for CLH1GClNO~~HCl:C, 34.6; H , 6.12; [ (2-chloroethyl)-ethylamino] -6-deoxy-1,2-0 - isopropylideneC1, 25.5; N, 5.03. Found: C,35.2; H,6.21; C1, 25.2; N, D-glUCOfUranOSe (XVIII) by the procedure used for the prep- 5.04. aration of 6-[bis-(2-chloroethyl)-amino]-6-deoxy-~-glucose Acknowledgments.-The authors are indeb led to hydrochloride ( X X I I I ) . The "one-armed" mustard was obtained in quantitative yield as a pale brown sirup which Dr. Peter Lim and staff for the chromatograms and ~, R, optical rotations, as well as the interpretation of the was homogeneous on paper c h r ~ m a t o g r a p h y ~with 0.91. infrared spectra; and to Mr. 0. P. Crews, Jr., and Anal. Calcd. for C I ~ H 2 ~ C 1 N O 6 . H C l ~ 1 / 3 HC,2 0 38.5; : staff for large-scale preparations of intermediH , 6.94; C1, 22.8. Found: G38.7; H , 7.22; C1,22.3. ates. 6 4 2-Chloroethylamino)-6-deoxy-~-glucose Hydrochloride (XXV).-A solution of 2.53 g. (7.2 mmoles) of 3,5-O-benzyli- MENLOPARK,CALIF.
[CONTRIBUTIONFROM
THE
DEPARTMENT OF CHEMISTRY, FACULTY O F SCIENCE, CAIRO UNIVERSITY]
Action of Grignard Reagents. XV1I.l Action of Organomagnesium Compounds on 5-Arylidene Derivatives of 3-Arylrhodanines, of 3-p-Tolyl-2,4-thiazolidinedione and on 2-Arylidene-3 (2H)-4,5-benzthianaphthenone-l, 1-dioxides BY AIIMEDMUSTAFA, WAFIA~ S I < E R ,SAMIR RHATTAB, MOHAMED Ezz EL DIN SOBHY, ABDALLAH MoNAMED FLEIFEL AND KAMAL ABU-ELAZAYEM RECEIVED AUGUST14, 1959 Treatment of 5-arylidene derivatives of 3-arylrhodanines and of 3-fi-tolyl-2,4-thiazolidinedione with Grignard reagents does not effect hetero-ring opening and only the double bond of the lateral chain of I1 and IV enters into reaction, yielding colorless products, believed to have structures 111 and V, respectively. 3-Phenylrhodanine and 3-p-tolyl-2,4-thiazolidinedione were proved to be stable when treated with phenylmagnesium bromide under similar conditions. When I I I a was treated with aqueous sodium hydroxide solution, or-mercapto-@,@-diphenylpropionicacid was obtained. Conjugate addition, without any indication of cleavage, now has been observed when the newly prepared 2-arylidene-3(2H)-4,5-benzthianaphthenone-1,l-dioxides(Xa-b) are allowed to react with organomagnesium halides, yielding colorless reaction products, believed to have structures XI.
Recently12 in conjunction with a study of the pharmacological and toxicological properties of rhodanine derivatives] Mustafa and co-workers have prepared derivatives of 5-methylrhodanine and 5 - methyl - 3 - phenyl - 2,4 - thiazolidinedione, through the action of Grignard reagents on 5-aralkylidene rhodanines and 5-aralkylidene-3phenyl-2,4-thiazolidinediones, re~pectively.~ (1) A. Mustafa and bl. M . Sallam, THIS JOURNAL, 81, 1980 (1059). (2) S. Tawab, A. Mustafa and A. F. A. Shalaby, h'ature, 183, 607
(1950). (3) I n view of t h e marked interest in many derivatives of thiazolidone which proved t o be useful a5 anesthetics (A. R Surrey, THIS J O U R N A L , 71, 3354 (1949)), anti-convulsants (H. D. Troutman and L. M. Long, ibid., 70, 3436 (1948)) a n d amebacidal agents (A. R. Surrey and R. A. Cutler, ibid., 7 6 , 578 (1954)), t h e presence of a thiazolidine moiety in penicillin, t h e fungi-toxic or bacteria-toxic activity shown by many derivatives of rhodanines (H. K. Pujari and M. K. Rout, J . Sci. ZiLd. Res. ( I n d i a ) , 14B,398 (1965)); F. C. Brown a n d C. K. Bradsher, Nature, 168, 171 (1951); F. C. Brown, C. K. Bradsher, E. C. Morgan, M. Tetenbaum and P. Wilder, T H i s J O U R N A L , 78, 384 (1956).
We now would like to report our extension of this investigation for the aim of preparation of a number of new derivatives needed for the pharmacological studies. I n this investigation, the action of Grignard reagents on 5-arylidene-3-arylrhodanines (ITa-h) and on 5-arylidene-3-p-toly1-2,4-thiazolidinediones (IVa-e) has now been undertaken with the formation of the reaction products IIIa-l and Va-f, respectively (cf. Scheme A and B). The Grignard reagents do not effect the lieteroring opening and only addition of organomagnesium compounds to the conjugation created by attachment of an exocyclic double bond in the 5position of a heterocyclic ring having a carbonyl function I1 and IV takes place. The structure of IIIa, which is takeii :is an (4) A. Mustafa, W. Asker, A I(, A. Shdlnby and hl. B.:Sobhy,
J. Org. Chetn., 23, 1992 (1958).
