January 1969
NOTES
phenyl Sulfone.-A solution of 2.5 g (5.7 mmoles) of N-(3-chloronaphthoquinonyl) diaminodiphenyl sulfone in 40 ml of 97% HC02H JTas heated under reflux for 2.0 hr, an equal volume of H20was added to the hot solution, and the bright orange crystalline precipitate was collected arid washed (HzO, MeOH, EtsO); yield 2.6 g, mp 244-248’ dec. The compound was recrystallized by solutioii in 100 ml of DMF (dissolved a t room temperature) arid addiiig ail equal volume of HZO. The product (2.4 g ) separated as glisteiiing oraiige platelets, mp 244-248”. See Table I. N-Acetyl-”- (3-chloro-1,4-naphthoquinonyl) Diaminodiphenyl Sulfone.--S-(3-Chloronaphthoq~iinonyl) diaminodiphenyl sulfone (0.3 g, 1.25 mmoles) was suspended in 20 ml of 4 c 2 0 and then warmed 011 the .-team bath. The red crystalline suspension dissolved and ail orange crystalliiie product precipitated. After cooling to room temperature, the .tals were collected by filtratiori sild washed (MeOII, EtaO); yield 0.29 g. The crystals were recrystallized from a T H F aiid HZO mixture to give 0.27 g of product. See Table I.
They further suggested that it might be an axiridinium, piperaxinium, or v-triazolinium salt. We wish to report the identification of the transformation product, by X-ray crystal structure analysis as well as nmr and mass spectral studies, as 1-(2-chloroethyl)-3-(~-carbamoylimidazol-4-yl)-A2-1,2,3-triazoliniumchloride (111). This shows that the transformation reaction is a novel means of formation of the v-triasoline ring system. Bond lengths and angles between nonhydrogen atoms a t the present stage of refinement ( R = 0.10) are shown in Figure 1. The two rings, amide group, and a carbon of the chloroethyl group are roughly coplanar, the maximym deviation of any atom from the plane being 0.22 A. The bond lengths and angles in the triaxolinium ion show that i t has the symmetrical resonance-stabilized form. The chloroethyl group has a gauche conformation. The hydrogen atoms have been located in a three-dimensional diff ererice synthesis, and their positions confirm the tautomeric form abcribed to the imidazole ring, and the existence of an intra9olecular S-H. . .S hydrogen bond of length 2.98 A between the amide S and the K2 of the triazoline ring.
Acknowledgment.-The authors are grateful to Dr. AI. E. Wall, Director of this laborat,ory, for his kind encouragement and support, of this nork.
Single Crystal Studies of Chemotherapeutic Agents. I. The Structure of 1- (2-Chloroethyl)-3-(5-carbamoylimidazol-4-y1)A2-1,2,3-triazolinium Chloride1
189
\
&,3
no
D. J. A4U1L\H.\M,2 J. S.~ ~ C T H E I L F O I I D. ,~ N D11. I).1tOSI:SSjTEIN Deparhkent of ,lfedicinal Chemistry, School of Pharmacy and the Crystallography Laboratory, rniversity of Pittsburgh, Pittsburgh, Pennsylvania 16213
CI
I81
Received September 13, 1968
The nitrogen niustard .5(4)-[3,3-bis(2-chloroethyl)l-triazeno]imidazole-4(3)-carboxaniide (I) is a promising anticancer agent.3 Other related triaaenes, particularly the dimethyl compound 11, also show anticancer activity to a lesser e ~ t e n t . Unfortunately, ~ the nitrogen mustard I undergoes a spontaneous
CH9-kH,
I, R = CHPCHZCI
C
C
111
II,R=CH3>
0
Iv transformation ut room temperature in the solid state to form an inactive isomer, which was recognized as a quaternary ammonium chloride by Shealy, et a1.j (1) This research was supported by contract PH43-67-1188 from t h e Cancer Chemotherapy Kational Service Center, Kational Cancer Institute, National Institutes of Health, Bethesda, Md. (2) Joint member of both t h e Department of Medicinal Chemistry and t h e Crystallography Laboratory. ( 3 ) T.F. Piiealq- and C. .I,Krauth, Suture. 210, 208 (1966). (4) T. F. Sheals, J. A . Montgomery, and JV. R. Lafter, Jr., Biochem. f’harmacol., 11, 6 i 4 (1962). ( 5 ) Y . F. Hhealy, C. A. Krauth, L. B. Holurn, and JV. E. Fitayibbun, J . Pkarm. Sci., 57, 83 (1968).
Figure 1.-Bond lengths arid aiigles ill the 1-(‘2-chIoroethyl)-3(3-carbamoylimidazol-4-yl)-~~-l,2,3-tria~olirii~im ion.
The nmr spectrum agrees well with this structure, both in the ratios of the various types of hydrogens and in the observed splittings. The fact that all CH, protons occur downfield from the usual CH2range is in accord with this structure, since b and c are next to a positive N, while protons of type a are adjacent to C1. This investigation further allows us to draw inferences concerning the mode of action of the drug, based on the stereochemistry of the inactive compound. The substituent groups on the imidazole ring are almost coplanar with it, and so arranged that the amide ?; and S 2 of the tri%zolinering face each other at a distance of about 3 A. An intramolecular H bond between these two atoms holds the molecule in this configuration. If we assume that this arrangement also occurs in the parent compound, the structural similarity t o guanine (IV) becomes immediately apparent. The C (2)X(1) bond of guanine has been replaced by this S-H. . I\’ hydrogen bond. The peripheral pattern of Hbond donors and acceptors, required by guanine’s role in the nucleic acids and determining the specificity of enzymes, remains unchanged except for the modification of Y z to a nitrogen mustard. The fact that the anticancer activity remains, although reduced, when these 2-chloroethyl groups are replaced by other substituents, supports the conclusion that the activity is related to the structure of the remainder of the molecule. The over-all implication is that I may bind, 1
and subsequently bond, by nieniis of the alliylatirig groups, to, for example, the active site of eiizymes for which gutninc is a substrate. Experimental Section Nmr.----The nmr of TI1 \vas determined iii DllSO-ds w i t h >le,% as the internal standard. I I were asigiied as show1 i l l Table I. Upon addition of DSO the peaks at, 6 7.50 arid 8.0 disappe:tred. T.\HI.E . S S S I G S ~ l E S T OF TSJP I [
L
I) I’
d e f
6 raiue
I IIYuiu~c;Esa
13 ill i u
Splitting
)
Distorted triplet Dishrted triplet 1)istorted triplet Singlet Broad huitqis Broad hiunps
-
: , !I.; 4 . :2,j’,
(i
4.601 7.60 7 .. i O or S,0 7 .,50 or 8.0
I
, >
Mass Spectrum.--The ,spectr:il fnignietitat i o i i pat terii diowed the molecular ion to o r c ~ i rat n i / e 2.1‘2, yhich curre..poiids t o the positive ion 111. Crystal Data and Structure Determination.--The (rawformai i c i i i product was re( ,tallized (AIeOII) t o give thick tabulEr yellow cryntals (001’) t h a = 12.80, b = 7.41, c = 12.42 A ; p = 90’44‘; I‘ = 1178 &; D, (flotation) 1.5Y;Z = 4; U , 1.573; absorption coefficient f i x C u K a radiatiort, 49.3 c111-l. The crystal size xva 0.2 x 0.2 x 0.2 mni. Iiiteiisity data were measured with a Picker four-circle aut,crr 8-20 mode to a maximum 28 of 130”. for Oh-0 ( k odd) arid 1101 il odd) iiidicated, unambiguously, the rnoiioclinic space groiip PL’l,~c. The structure x a s solved using I’att,ersoii aiid heav>--atom method3 aiid refined by block diagcxial least squares to give an agreement fartor X of 0.20. Further anisotropic full-matrix leajt-squares refinernelit has reduced the /L’ factor to 0.10 and permitted the determination of 11-atom pcisitioris by the tiifli’ereiice Fourier method.
3-Aza-A-homoandrostenes
Oiic of the vie\\-s on the mechanism of steroid hormone action is that steroids exert their effect at t,he level of DKA coiitrol of RSA synthe seriger R S h molecules thus arc the tem enzyme syiithesis arid it is these eiizymes which regulate the process re~ult~ing in the observed physiological effects.2 The interaction bet\veen steroid aiid DSd3-; and steroid and protein6 is well cst,ablished. In fact, ‘l’s’o and Lu7 have demonstrated that the coil form of DNA has a higher affinity for steroids than the helical form. These types of findings prompted IZingolds to postulate an a-face adsorption for androgen molecules. I I ) hI. Zalokar in “Control LIeclianimis in Cellrilar Processes,” D. >I. Honner, Ed., The Ronald Press Co.. S e w York, K . Y . , 1960. ( 2 ) P. liarlson, Perspwtiues Bioi. M e , / . , 6, 203 (1963). 13) A . ;\Iuiick, J. F. Scott, arid I,. L. Engel, Riochzm. Biophy.Acta, 26, 397 (IWi). nd I,. L. E n g d , d i d . , 23, 665 (1957). n d I,. I,. I,:nprl i n “Alecliaiiism of Action of Steroid blurlei’ m i d 1,. I,. J:rig
