IIarch 1966
HETEROCYCLIC SULFUR
217
llUSTA4RDS
Heterocyclic Derivatives of 2-Chloroethyl Sulfide with Antitumor Activity' RICHARD AI. PECK,ANKAP. O'COI~NELL, AND HUGHJ. CREECH The Institute for Cancer Research, Philadelphia 11, Pennsylvania Received August 25, 1965 Several monofunctional nitrogen and sulfur mustards combined with a variety of heterocyclic nuclei through aminoalkyl side c,hains have been synthesized for comparisons of their antitiimor activities. The presence of amidine, benz [ c ]acridine, and phenanthridine nuclei increased the antitumor effectiveness of both types of mustard moiety; the sulfur mustard derivatives displayed higher, but much broader, dosage ranges of activity. Since the sulfur mustards of 7-chloro- and 3,7-dichloroquinoline were conPiderably more effective against Ehrlich ascites tumors in the mouse t,han the corresponding nitrogen half-mustards, it is considered that studies utilizing sulfur mustard derivatives may be an improved procedure for the detection of pot,ential antitumor activity in the carrier portion of the molecule.
Our previous determinations of the relationships between the chemical structure and biological activity of a variety of alkylating agents showed that almo5t all the bis(2-chloroethy1)aniinoalkylamino derivatives of quinoline and acridine were highly effective against Ehrlich ascites tumors in the mouse, but that this inherent activity could be modified by the nature of the alkylamino side chain and of the heterocyclic group used as a carrier. I n fact, certain carrier structures such as 6-methoxyquinoline and various acridines were found to potentiate the activity of the nitrogen mustard n ~ o i e t y . ~ -This ~ prompted the synthesis of a wide selection of monofunctional nitrogen mustards (half-mustards), some of which displayed exceptional antitumor activity.5 For example, 2-niethoxy-9[3-(ethyl-2-chloroethyl)aminopropylamino]acridirie dihydrochloride was found to be more active on a molar basis than nitrogen mustard [methylbis(2-chloroethyljamine hydrochloride]. Several other monofunctional nitrogen mustards containing the acridine and substituted acridine nuclei were highly effective a t low molar dosage against ascites tumors, whereas the quinoline and quinazoline derivatives were either inactive or only moderately active a t extremely high molar dosage. I n our view the pronounced antitumor properties and mutagenic capabilities' of the acridine nitrogen half-mustards are due to the participation of the carrier portion of the molecule in cytotoxic activity in place of the function of the second alkylating center of the bis-mustards. Thus the half-mustards exhibit a modified form of bifunctionality. Supporting this view is the fact that acridines have been shown to have a remarkable affinity for deoxyribonucleic acid.8 These activating carriers permitted exploration of a series of sulfur mustards not otherwise possible, since the sulfur valence of two precludes attachment (1) Supported by research Grants CA 02976 and CA 06927 from t h e National Cancer Institute, National Institutes of Health, U. S. Public Health Service. (2) R. AT. Peck, R . K. Preston, and H. J. Creech, J . A m . Chem. Soc., 81, 3984 (1959). (3) H. J. Creech, E. Breuninger, R . F. Hankwitz, Jr., G. Polsky, and M. L. Wilson, Cancer Res., 2 0 , 471 (1960). (4) R . M. Peck, R. K. Preston, and H. J. Creech, J. Org. Chem., 26, 3409 (1961). (5) R. K. Preston, R. M. Peck, E. R . Breuninger, A. J. Miller, and H. J. Creech, J. Med. Chem., 7, 471 (1964). (6) R. 111. Peck, E. R . Breuninger, A. J. Miller, and H. J. Creech, ibid., 7 , 480 (1964). (7) E. A. Carlson and I. I. Oster, Genetics, 47, 561 (1962). (8) L. S. Lerman, J. M o l . B i d . , 3, 18 (1961).
of a carrier to the bifunctional alkylating moiety. Consequently, several representative heterocyclic derivatives of 2-chloroet hyl sulfide were synthesized and compared with their corresponding ethyl-2-chloroethylaniines (nitrogen half-mustards) in studies niade with mouse ascites tumors. The influence of different nuclei and side chains on the antitumor activity of the corresponding sulfur and nitrogen mustards are given in Table I along with the analytical information on these compounds. Data on the intermediates are presented in Table 11. The intermediate hydroxy derivatives of the compounds with the alkyl side chains were niade either (1) by direct condensation of the haloheterocyclic compound with the corresponding side chain or ( 2 ) by stepwise condensation with the appropriate amino alcohol followed by the replacement of the hydroxyl group b y chlorine and then the use of the resultant compound for the alkylation of the alkylaminoethanol or sodium derivative of niercaptoethanol. The two sulfur mustards with the amide side chains were prepared as described in the Experimental Section.
Experimental Section Llelting points were taken in open capillary tubes in a Hershberg apparatus using total immersion thermometers and are reported as uncorrected values. All of the 2-chloroethyl compounds in this paper were prepared by the action of excess SOCh on their hydroxy precursors.2 Where these precursors were obtained by direct condensation (method A in Table 11) the reactants have all been previously described except in the case of 3-aminopropyl-2-hydroxyethyl sulfide; preparation of this intermediate is described below in the course of the preparation of compound D-2 from Table 11. An experimental procedure utilizing method B is given below in the preparation of B-2 (Table 11). Finally, the preparation of one of the compounds containing an amide side chain is given. 3-Aminopropyl-2-hydroxyethylSulfide and 7-Chloro-4-[3-(2chloroethyl)mercaptopropylamino]quinoline Hydrochloride (D-2, Table II).-To a stirred mixture of 300 ml of ethanol containing 47 g of mercaptoethanol and 24 g of NaOH (0.6 mole each) and 300 ml of 1 N NaOH in methanol was added in portions a solution of 66 g (0.3 mole) of 3-bromopropylamine hydrobromide in 150 ml of absolute ethanol. The mixture was stirred and heated, and 400 ml of solvent was removed by distillation. After cooling, 100 ml of ether was added and the NaBr precipitate (84% yield) was removed by filtration. The filtrate was concentrated, 0.3 mole of concentrated HC1 and 100 ml of ether were added, and the precipitated XaC1 was removed. The filtrate was concentrated and the residue was distilled in vacuo in a modified yon Braun flask, yielding 38 g (937,) of material, bp 81-98' (60-100 p ) . Redistillation gave 34 g, bp 78-82" (70 p ) , of slightly impure product.
"18
I-
--.
+-
/
I-
I-
I
I
/
*
1-
1-
-I 1t
f l
J
-
-.^ .^
B
I?
?I
\ /
3 ...
I:
L -t
1I-
1?I
41 %
I-
i
-t
I?
I?
:i
:i
i i
1-
-
r-
w i Y
v M
-
v M
E n
-
z
z
HETEROCYCLIC ST-LFUR Sl r r s ~ ~ i t ~ s
March 1966 4 c
b i z "?
m
-
g J G g
m
k
0 N
m $ c
rz, t
2 19
Calcd for CjHlaSOS: C:, 44.31 ; H,\).ti!l: S , liireof 120-125' for :: hr, cooled, taken up in dilute acetic acid, xrid filtered. 'Tht: product was precipitated with alkali, filtered, aiid reprecipitated once more from dilute acetic- arid to give 2.6 g. m p 14'2-146'. Cryst:dliz:ition from ethariol gave 2.1% g ( i : 3 ! ; ) of D-2 ('hbI(1 11). 2-Hydroxy-2'-( 7-benz [ c ] acridiny1amino)diethylSulfide Hydrochloride (B-2, Table II).-To a stirred soliitioti of 1.55 p rif riierwptoethaiiol in 20 1111 of 1 AYniethaiiolic NaOH x i s addetl 1 .fi g (4.8 mmoles) of 7-(2-chloroeth)lamit~o)beuz[c]acridine 1iytiroc:liloride. The mixture was heated for 2 hr 011 the steam cijtie, filtered, diluted, and cooled. Since the product, resisi etl ~tallizatioii,the aqueous layer as decanted and the produvt was taken up iii ether. The solutioti w:i> rxtmcted ivith 4.5 nil of 1 S HCl and the e x t r w t was ivriic~eir1r:itedi n LUCUO :tiid t x k e ~ i up i i i a lit,tle ethanol, when the salt ( ized arid wis filtered ; it weighed 1.7 g, nip 200-204". t:illimiioti p:ive t l i r :iiialytic-al sample i n Table 11. XI-( 7-Chloro-4-quinolyl)-N2-[ S-(2-hydroxyethyl)thioglycolylj ethylenediamine (D-4, Table 11).--TI) 100 nil of 1 -Y meth:iiiolic, XaOH was added 7.8 g (0.10 riiole) of riierc:rptoetliariol. The solution was cooled arid 12.2 g (0.10 mole) of etli>-lchloroacetatr was added with swirling. The solut ion was fibered, wartlied for 2 Iir o i l the steitni cone! and refiltered. One-half of the re.ult:irit w l r t t i o r i was added t o 11 g of X-( ;-c~tilorc)-i-cl~iiiiol~l)ethylriirtli:irnitie :ind the ,-oliitioii W:LS w:rrnied o i l the steam ('IJIIC for 2 IIY i i t i d rooled. The product was reinoved fi1tr:itiorr ; it weiytietl ! ) , I g, I n p l t ~ l - . l ~ i O o .C'r? llizai,ioii f i u r i i 4 : 1 ethiirrol-watPr gave it first crop uf 4.5 g! nip 164--16ti.>', : i ~ i d:I itv.otId ( ' r ~oi4.> ~p p, rnp l6%5.5-l6So. .lncd.
s, 2 : ~ xF'otltlci:
-
Biological Results Abcites tunlor.; i n the iiiouye riniyuely -uitablc for quaiititative deteriiiiiintio~~~ of the efl'ectiveiieof analogs of thc nitrogrii :itid sulfur mustards. The :diiiiiiistrat ioii of :I. v:irictty of cheniothernpeutic agelit5 will substariti:illy reducc tlie :isrite- rell count i n the mouse, but a proiiouriced c:qxwtty to prolong the w r viva1 tinie of mice ~iiocul:itecl \VI t h :iicites tumor? I ? 1 cytricted to certain must arc The procedure5 u-ed t o d i i i ( ~aiid interpret the rcsult5 of our :tiitituiiioi 5t have becri described i i i detail 111 earlier 1)ublicati 'I hi brief, the conipounds in haline wlution or suipcns~oiiwere injected i i i t r:iperitoneallp into albino (ICH Swis.) inice which Iiatl been inocxulated the previou, (lay with 7 X 10h c*ollsof the Ehrlich a-cites iuiiior EE'. Survival data recorded daily for tlie (~xperinientnlarid control m i c ~and the antitumor :wtivity of the compound was c.:Llculnted froiii t h e v graph-. The tests on each coiii1)ouiid utilized between 100 aiid 200 mice. Uuder our c~periiiiental(-oiiditioii~, thc rolitrol ~iiice diiplayed a iiieaii survival time of 16 i1 dayq. To iiii~iiitaiiispace 111 our aiiinial qiiurteri, all experiments werc terminated a t the end of the period that was three tinlei: the mean burl i v d time of the control mice i i i each berieh of tebts. h value of 3.0 for the degree of :iiitituriior activity was assigned when every mouse 111 the experimental group of 16 given a certain dosage of ci~nipouiidlivcd to the time of sacrifice a t about 48 tlayi. Iii our experience with nitrogeii mustards, ricarlp a11 of ilie mice surxivi~igT weelis could be exlwrtetl to live inore thaii 15 \welis without signs of :iicites tunior development. X degree of activity of 1.O indicates that L: conipoinitl had 110 effect on tht. I l ? n k \ l l t 7 Tr , n J 1 i t r l ~ l o n alii1 , : m i
(L
AIarch 1966
hRSENICAL N I T R O G E N 1 1 U S T A R D S ANI) RELATED xC’ONALKYL.1TING X H S E N I C A L S
22 1
In the cabc of the sulfur mustards, however, conipounds containing these quinoline nuclei display a high order of antitumor activity, and although the molar dosage is high, that of the sulfur niustard reference compound, 3-(2-chloroethyl)mercaptopropylamirie hydrochloride, whose average activity degree is 2.0 over a range of 230-450 pmoleslkg, is even higher. Therefore potentiation by the5e less active carrier3 has been detected by ube of the 2-chloroethyl sulfide iiioiet y .
side chains although the required dosages were unusually high with both types of mustard (D-3 and D-4). Even higher dosages were needed to obtain a display of activity with the mustards combined with the 3,Tdichloroquinoline nucleus (E-1 and E-2). Since the iiitrogen half-mustard reference compound had an activity degree of 2.3 at 23-60 pmolei/kg, it i3 obvious that 7-chloroquinoline and 3,’i-dichloroquinoliiie do iiot cause ariy activation of the nitrogen half-mustard moiety, but, in fact, depress its ailtiturnor eff ectiveneis.
Synthesis, Chemistry, and Preliminary Pharmacology of Arsenical Nitrogen Mustards and Structurally Related Nonalkylating Arsenicals1 THOMAS J. BARDOS, XIRMALENDU
DATT.4-GUPT.4,
AND P E T E R
HEBBORN
Departments of Medicinal Chemistry and Biochemical Pharmacology, School of Pharmacy, State University of New York at Buffalo, Buffalo, New York Received Azcgust 26, 1965 The synthesis of phenyl nitrogen mustards substituted in the para position with arsonic acid, arsenoso, arseno, and dithiarsenolane groups and the synthesis of the corresponding arsenic derivatives of diethylaniline and bis(P-hydroxyethyl)aniline, representing nonalkylating structural analogs of the former, are described. In relation to this work, some novel aspects of the chemistry of organic arsenicals are discussed. All of the arsenical nitrogen mustards synthesized show very low chemical alkylating activities; this fact, in corroboration with the nmr spectra, indicates that not only the arsonic acid group but also the trivalent arsenic groups have strong electronattracting character. All arsenical nitrogen mustards, except the arseno compound, are highly toxic in mice, but their toxicities are apparently due to their arsenic groups rather than their alkylating moieties. Preliminary screening results indicate that the arsenoso and arseno mustards have significant activities in the Ehrlich ascites test system, both showing complete inhibition a t nontoxic dose levels.
Previous work directed toward the synthesis of compounds designed t o incorporate the biologically essential structural features of two different but synergistic inhibitors into a single molecule resulted in several types of new antineoplastic agents,2 some of which proved to be of clinical i n t e r e ~ t . ~The rationale of this so-called “dual antagonist” approach to chemotherapy has been d i s ~ u s s e d . ~Since cyst’eine, glutat hione, and several other sulfhydryl compounds are known to antagonize the toxic effects of ionizing radiation as well as of some “radiomimetic” alkylating agents (e.g., HX2 and aromatic nhrogen mustards5), it was thought that “sulfhydryl inhibitors” such as organic arsenicals6 could, conversely, potentiate the effects of alkylating agents.’ It appeared possible that new types of dual antagonists could be designed by combining the structural features of “sulfhydryl inhibitors” with those of alkylating agents. The synthesis and ( 1 ) (a) This investigation was supported by Public Health Service Research Grants KO. Ch-06695 a n d CA-06645 from the National Cancer Institute. (b) A preiiminary report, covering a small portion of this work, was presented before the Division of Xedicinal Chemistry, 147th Piational Meeting of the American Chemical Society, Philadelphia. Pa., April 1964; Abstracts, p 26RI. (2) T. J. Uardos, A. I
