March 1970
abJTITU310R ,%CTIVITY OF
ALKYLliTISG A G E S T s
TABLE I HALOXETHYL L4ROM.iTIC HYDROC.LRBOSS 7 -
4-Chloromethyl-l-methylnaphthalenec 10-Chloromethyl-9-methylphenanthrened 9-Chloromethylanthracenee 10-Chloromethyl-9-methylanthracene'
10-Iodomethyl-9-methylanthraceneg 7-Ch1oromet)hylberiz[a] anthracene* 7-Chloromethyl-12-methylbenz[a] anthracenei 7-Iodomethyl-12-methylbenz [a]anthracenej 1-Chloromethy1p)rene"
Antitumor activitya--Range, pmol/kg
-\. llonofunctional 60-300 35-300 1-12j 4-100 20-300 0.8-20 2-175 20-350 0.8-35
Degree
"0 2.3 2.1 2.4 2.6 2.0 2.2 2.3 2.3
Mp,
"C
155,.5-156.5 142-144 193-197 dec Dec 190-192, s 185 144-147 Dec 138-142 dec
B. Bifunctional 1,4-Bis (ch10romethyl)benzene~ Inactive; toxic a t 100 CJLC12 3,6-Bis ( chloromethyl)durenec 50-180 2.3 CirHi6C12 C 8 H 4 c16 a,a'-2,3,~5,6-Hexachloro-p-xylenec Inactive; toxic at 30 9,10-Bis(chloromethyl)anthracenec 0.3-30 2.4 Ci6Hi?Clz h Although all of these compounds have been reported in the literature, crystallization to cona See section on biological resiilts. stant composition was carried out on all the compounds in part A apart from the first compound (commercially available) and t,he iinstable iodomethyl derivatives. I n part B, the first three were commercial products and used directly. The last commercial product S. Haiiptmann, Chem. Ber., was found to be impure and was crystallized three t,imes from benzene. c Commercially available. 93, 2604 (1960). e R. T. Hunter, et al., J . Org. Chem., 21, 1512 (1956). f J. L. hdelfang and G. H . Ilaith, J . .inier. Chem. Soc.,SO, Although this compound was described by Badger and G. 11.Badger and R. S.Pierce, J . Chem. SOC., 2314 (1950). 1405 (1958). Cook, ibid., 802 (1939), we found that, the action of dry ethanolic hydrogen chloride on the methylol gave a cleaner prodrict. J. Pataki, R. Wlos, and Y. Cho, J . Z e d . Chem., 11, 1083 (1968). i W. E. Bachmann and AI. Carmack, J . A m e r . Cheni. Soc., 63, 2494 ( 1941) . Q
DxRIv.yrIvEs
OF
T.\HLEI1 IICHINHCH~CH~SCH~CH?CI HC1
--Antitumor activity'Range, pmol/kg Degree
Yield,
'3 Np, o c Formula" 9-Anthryl 40-1.50 2.5 41 18.5-187 C19H2&1K;8 HCl 83 140-140.5 C1gH2oClNS HC1 9-Phenanthryl 60-275 2.4 10-hIethyl-9-an thryl': 1.6-90 2.6 .8i 194-196.5 CnoHprClKS.HC1 10-Ethyl-9-anthryld 80-500 2.1 70 183-183 dec C?1Hz,ClNS.HC1 4, 10-Dimethyl-9-anthry1" 30-180 2.1 76 1!10-192 dec CyLH2rClNS HC1" 4-Fluoro-10-methyl-9-anthryl 1-15 2.3 86 207-209 C2oH21ClFNS HCl 194-195.5 CPoH~iCI&S'HClu 4-Chloro-10-methyl-9-anthryl 2.5-100 2.3 62 7-Benz [alanthryl 12-160 2.5 79 193-197 CniHrpClNY . HC1 12-AIethyl-7-benz [ a ]anthryl 3-250 2.6 61 183-186 dec C21H2aClNS HCl By methods I and 11. I n t'he course of the See section on biological results. h ,411 compounds were analyzed for C, H, S, C1. animal tests it was found that the two samples of t,his compound, apparently chemically identical, gave differing antitumor results. We could find no rationale for this discrepancy. d Bv methods I and 11. e By methods I1 and 111. f C : calcd, 63.94; formd, 64.39, 64.37. g Cl: calcd, 23.65; found, 24.84, 24.95. R
Q
blocking group adjacent to both meso positions, failed to yield a methyliodomethylanthracene. Replacement of OH by C1 to produce the compounds in Table I1 was carried out by the method routinely used in the past, i.e., excess SOClz as a solvent a t 5-30' (method I in the Experimental Section). Where the terminal ring(s) carried an alkyl substituent however, only intractable tars were obtained, and two other methods were found to yield clean products. AIethod I1 consisted of a slight molar excess of SOClz in dioxane solution on the steam cone7 and method 111, successful only with certain hydroxyethyl sulfides, utilized a large excess of 12 N HC1 as a solvent and dilution after standing overnight. Two of the unstable iodomethyl hydrocarbons not of analytical purity5s6 and a variety of corresponding chloromethyl compounds (Table I ) were included in the antitumor screen. For comparison with these unexpectedly potent compounds, four commercially available bis(chloromethy1) hydrocarbons were tested. (7) W.T. Hunter, J. S. Buck, F. W. Gubita, and C . H. Bolen, J . O w . Chem., 21, 1512 (1956).
The alcohols corresponding to the nitrogen mustards in Table I11 were synthesized from these halomethyl compounds by condensation with t'he necessary 2[ (w-aminoalkyl) ethylamino ]ethanol
Experimental Section Melting points were taken in open capillary tubes in a Welshberg apparatus using total immersion thermometers and are reported as uncorrected values. Where analyses are indicated only by symbols of the elements, analytical results obtained for these elements were within 410.4% of the theoretical values. 2- [Z-[ (9-Anthrylmethyl)amino]ethylthio] ethanol (Method A). -A mixture of 4.12 g (0.02 mol) of 9-anthraldehyde and 2.7 g (0.022 mol) of 2-(2-aminoethylthio)ethanol in 20 ml of EtOH was refluxed gently for 2 hr, diluted with hexane, cooled overnight, and filtered. The product was recrystallized from EtOH-H?O to give 4.9 g of the product in Table IVY(79%). This was dissolved in 50 ml of warm C6H6 and added in portions over a period of 0.5 hr to a stirred solution of 1 g of LiBHd in 1 1. of dry Et20. The resulting suspension was stirred for an additional 3.5 hi- and decomposed with ice and AcOH. The aqueous layer was filtered and made alkaline to yield 1.9 g of product. Further extraction of the EtqO with 100 ml of 20yo AcOH gave an additional 0.25 g. The combined product was reprecipit'ated from solution in dilute -4cOH in the presence of a little petroleum ether (bp 30-60") to give 2.1 g of the product iii Table V.
,.t h e rcssitiiie w:i.
hlieirtl i ' i ~ ~ i i 13;tOH ii h:izO t o give (i.5 g ~)f'procliic.r. tallizatioii iEf,011--Et2(1 j gave 3.5 g (83"; ), nip 20W 201 '. 'I-Chloromethyl-12-methylbenz [a]anthracene.". TI) 20 nil I SOCl,, -timed i i i a i i i1.e bath, iyas added 700 mg of 7-methosymethyl-12-mrthylbeiiz [trjarithr.ace~ie.~ The soliitioii was allowed to i t a n d 2 h r a t 0" : t i i d 1 hr ;it room remperatiiw and then tonj f
ceirti,ated i n w c i r r i . The re-itliie was decomposed with EtOH, again taken trr I ~ I ~ I I P: i -i i ~d ,filtered fromfresh Et013 t l J give 0.60 g, mp 1:3S-140.tSc. I- aii esc.eption, the following method, I, was the mean> of ~ynIhe.is of a11 the compoiuids iii Tal)le 11. The otily variatioii i n cvtiditioris ~ ( I I ,thoye compound> in Table I11 was that the overiiight iwctioti prrioil wn-:c:irritd o i i t a t room temperature. S- [2-(2-Chloroethylthio)ethyl] -10-methyl-9-anthracenemethylamine Hydrochloride.--To 60 in1 of ? t i r i d SOC12cooled in nii ic,e bat,li was added, in portiona, i1.1 g of 2-[2-[(lO-niethyI-!)arittirylmethyl)arniiio]ethylthio] e t h a n ~ 11 f'l's hie T: ). . i f I er soli It i U t l \vas cvmplet,e, the flask \\-as stored overnight a t ahoil1 ,So arid at room temperature fur 4 hr: the11 volatile material was removed a t water pump vacuum. The re;idue was decomposed with 5-10 nil of EtOH, from which the p r * d u c t trjxtallized. Solvent, was again removed in im-cm, the proditct was stirred with more EtOH, ether wasadded, :urd the product was filtered and washed (EtnO) to yield 2.8 g, nip 193--194dec. Recrystallizatiori (EtOH-EtzO) gave 2.1 g of the product in Table 11. 1Let.hod I1 was devised for compoiiiids unstable to the a b ~ ~ v e conditions. It was also applied to a resynthesis of the above compound, as follows. To a w1iitioii of 5.55 g of 2-[2-[(10niethyl-9-anthrylmet hyl)aniii~o]etli~lthio]et hanol in pure dry dioxane was added a solution of 2.5 g (2Af;, molar excess) of t;OCl? i n 23 ml of dioxane, with stirring. The solution was immediately heated on the steam cone, aiid crystal., soon began to lorrn. Although the mass became quite thick, heating was coiltiniietl 20 niin; then volatile material im.: removed in z'ocuo. ~
1Iarch 1970
287
h T I T U 3 1 0 R A C T I V I T Y OF A L K Y L A T I N G A G E N T S
TABLE V HYDROXY PRECURSORS OF ~\ICSTARDS Yield,
7” Rlp. o c Formula Fiom Table 11: R C H ~ N H C H ~ C H ~ S C H ~ C H Z O H h 42 111 5-113 CdziNOS A 36 68 5 7 0 .i Ci9HnNOS Bh 42 185-186 -5 CzoH22iOS. HC1 A 36 10-5-106 CiiH25NOS B 52 84 5-86 CslH2sNOS CmH2zFXOS B 41 113-114 3 79-80 Cx,Hz2ClKOS B 29 h 2.j 112-113 C23H23NOS Be 47 74-76 CaHxYOS
Methoda
R
A. O-;\iithryl 9-Phenanthryl l(J-liethyl-9-aiithryl 10-E thyl-9-anthryl 4,lO-l)imet hyl-9-anthryl 4 - F l u o ~10-methyl-9-anthrj-1 4-Chloro-10-methyl-9-anthryl 7-Beuz [ a ]aiithryl 19-llethyl-7-henz [ a ] anthryl
iinrtlyses
c, H, s C, H, S C, H, S, C1 C, H, N, s C, H, N, S C, H, s C,d H, S, C1 c, H, N, s C, H, s
B. From Table 111. RCH~XH(CHI),K[C~H,)CH~CH~OH C, H, N, C1 B 69 188 5-190 C22H28PIjLO.2HC1 9-.iiithiylj n = 3 C, H,f N, Clf 123-127 dec C22H28N20.2HCl.O 3HiO B 20 10-llethyI-9-anthry1, n = 2 C,h H, Tu’, Ih BQ 2.j 165-166 3 C23H30N20.2HI 10-lIethyl-9-anthr,-1, n = 3 C,H, X, C1 C26H30N20.2HC1 B 43 230-233 dec 7-Benzialanthrl-1. n = 3 12-~1ethyl-7-henz[a]anthryl,n = 3 B 23 138-162 CnH32N20.2HI C J iH, Ii Also by method A. a Pee Experimental Section; method B1 utilized the chloromet,hyl hydrocarbon at 70-90’ without solvent. C: calcd, 66.80: found, 67.48. e First preThe hydrochloride was also prepared, mp 174-177”, and analyzed for C, H, S, C1. cipitated as the hydriodide from ilcOH extracts and recrystallized from EtOH-H20, mp 137.5-139’. f Calcd: H, 7.48; C1, 16.94; fulllid: H, 7.95: C1, 16.42. Also prepared by reaction of 3-(l0-methyl-9-anthrylmethyl)amino-l-chlo~opropane4with excess ethglCalcd: C, 49.40; I, 38.68: found: C, 48.82, 49.03; aminoethafiol. ‘L Calcd: C, 43.57; I, 41.82; found: C, 44.73; I, 42.33. I, 38.31, 38.21. .
1
I
G,
,
Q
carbon- ; 4 in these studies, the alkylating function con-ihted of a 2-chloroethylamine or a 2-chloroethyl dfide. In addition to the sulfur mustards, which are injected in animal, as homogenized suspensions, where a sub.tantia1 partial solubility nevertheless exists, data on 5oluble nitrogen mustard derivatives and on 13 halomethyl hydrocarbons, whose suspensions contain e--entiall> no dissolved material, are given (Tables I and 111). The length of the side chain and the nature of the aniine joining the arjlalkyl portion of the molecule to the niu-tard moiety mere found to be important. AS I\ a. the general cabe with the structures studied to date, compound- containing a “propyl” side chain diplayed antitumor activity a t lower dosage levels than tho-e with the “ethyl” side chain. The latter mere considerably lebs toxic than the propyl analogs, however, with the result that a more favorable therapeutic irides wa- .shown by the ethyl series of K- and Smu-t ard.;. uggebted the use of the side chain in Table I1 for evaluating the structure-activity relationships of wriou- condensed polynuclear hydrocarbons. & i n iniprebsive illustration of a large biological effect due to a small alteration in chemical structure is shown by the first, third, and fourth compounds in which the nzeso wbstituent changes from H to RIe to E t . Blthough an additional Cl or F substituent a t the 4-po-ition of the anthracene nucleus had little influence on the antitumor activity of the S mustard derivative, a Me in this position caused a significant decrease in activity. Enlargement of the condensed ring System by one benzenoid nucleus resulted in the lowering of the dosage needed to demonstrate activity (line3 8 and 1). A 1 I e a t the other vzeso position had a lesq pronounced effect in the S mustard of benz[a]anthracene than of anthracene in causing increased antitumor effectiveness (lines 9 and 3 ) . These data confirm and extend the previous observations* with
beveral polycyclic hydrocarbon derivatives containing the ((propyl”side chain and again indicate the existence of a parallelism between the carcinogenic and carcinostatic potentialities of the aromatic hydrocarbon moieties. The antitumor activities of the S mustard derivatives of several polynuclear aromatic hydrocarbons are listed in Table 111. Since these compounds, in contrast to their corresponding -ulfur mustards, have appreciable solubilities in aqueous media, they were administered to the mice ah *ohition.; in saline. Comparisons of the dosage range. of the t x o types of mustards show that the K mustard.; are toxic a t much lower dosage levels than the S mustards which exhibit superior therapeutic indices. The importance of a Me on the other meso position of anthracene and of benz[alanthracene is again evident +ince this substitution results in an appreciably greater degree of antitumor activity a t a somewhat lower do.;age level and over a broader dosage range. The length of the alkyl side chain is of significance, with the “ethyl” form showing lower activity than the “propyl” form, but a higher therapeutic index. Table I lists alkylating derivatives which have essentially no solubility in H 2 0 ; the high potency of a number of these compounds appears remarkable in the light of this property. All of these compounds have been described in the literature; it was found, however, that exhaustive recrystallization of the products was usually necessary to obtain a sample that was pure by the criteria of both elementary composition and biological activity. The procedure used to obtain the fourth compound in particular a t first gave samples displaying greater potency than that listed in the table. The presence of a second CH&1 as a n impurity was suspected; this led to the testing of the extremely potent final compound in Table I, a bifunctional alkylating agent. An indication pertinent to the great importance of geometric configuration of the aromatic moiety is given by the comparative biologic properties of three
Rp&
R
'
@ \
\
,
('H-C'I
CH-CI
I
11
ot1it.r p-bis(chloromethy1) model..
.ill ma) be reprewited by I , the simple benzene derivative ( I , R = Hj ii iiiactive but the durene derivative. ( 1 , I{ = CH,) ha' enough .teric re5emblarice to the highl). active II to impart activity t o the hiniple benzene ring. In the hex:tchloro compound (I, R = C1) the bull+ halogen atonii have a11 equivocal cffrct : t h e trcnd t o higher toxicity may mask any latent :intitunnor activit) Tn-o additionnl points of iiiterebt conceriiirig the> propertie. of the compouiid' in T h l c I x i h e . i I ) the ICH? derivative. are le- xctive bio1ogic:illj thaii t h r corre\ponding ClCH, hydrocaihn. under t h r te-ting coridition., although niucli inore rcnctivc chemically a. alkylnting agent*; ( 2 ) the ClCH, coinpounds of Tablth I (A) :I\ well it< thc 1a.t compouritl I I I Table I (13) -ho\v .trong :ictivit> the S - 3 i :I.cit
