MITOAIYCINS. XXII
July 196s
chloride. However, despite its marked in vitro activity against the tetracycline-resistant S . aureus, strain Rose, and Streptococcus pyogenes, P-hemolytic strain C203, 7-methoxymitosene (11) was not effective in vivo against these organisms.6 The discovery of significant antibacterial activity for I1 was then rationally followed by the synthesis of the indoloquinone analog 111, which had a spectrum arid level of activity about the same as that observed for
T.\BLEI .kCTIVITY O F
0 Si
0
Encouraged by these results,* we undert,ook an extensive analog program in the iridoloquiriorie series, Tvhich was chosen in preference to the pyrroloindole series (see 11) for convenience of synthesis. X considerable number of pert'iuent analogs mere prepared and the structure-activit'y relationships which can be drawn from this study are discussed below. I n this discussion, act'ivities, unless otherwise noted, were obtained from a n S. aureus, st'rain Smith, assay9 in mice (subcutaneous treatment). Those compounds active by the subcut'aneous rout'e, which were also submitt'ed to t'he corresponding oral assay, were invariably active in the latter assay as well. Variations at Ni.-Optimum activit'y was recognized at this position with either the methyl or ethyl substit'uent, the activity decreasing when these substituents were replaced with propyl, isopropyl, or butyl groups (see Table 1).lo Substitut,ion a t this site with a series of P-substitut'ed ethyl groups (-CH2CH2X: X = F, C1, S a , OH, SCH,) l1 proved unrewarding, although t'he members of this series were marked by in vitro activities nearly as high as t'hat,of the corresponding ethyl compound. l1 Variations at Cz.-Only the original methyl group at, C1 appeared compatible with high potency. Modest activity was noted for t'he -CH=XiC"COXH2 group12 (estd ED;, = 128 mg/kg), but the replacement of the methyl group by hydrogen,1° ethy1,iO or certain substituted met'hyl groups (-CH2X: X = OH, OA4c, SAC, C1, F ) 1 2 did not give comOCONH2, OCH3, "2, pounds of interest. The poor in vivo showing noted for the substituted-met'hyl derivatives was disappointing because a priori these analogs appeared t o be especially pertinent since a benzylic-t'ype funct,ion at Ci conceivably affords an additional locus for biological alkylation; furthermore, this site corresponds t,o that F. Poletto, a n d B I . J . Weiss, J . A m . Chem. Sac., (7) (a) G. R. Allen, J r 8 6 , 3878 (1964); (b) G. R. Allen, Jr., a n d M. J. Weiss, J . .Wed. Ckem., 10, l(1967). (8) Also noteworthy was t h e observation t h a t , in a bonemarrow depress a n t assay in mice (see ref. 18) in xhioh t h e parent antibiotics a n d I were virulently active, I1 and I11 appeared t o show, a t most, border-line depress a n t action. (9) (a) In vivo antibacterial assays were carried o u t according t o t h e procedure of G. S . Redin a n d M. E. AlcCoy, "Antibiotics Annual, 1959-1960," Antibiotica Inc., New York, S . Y., 1960, p 213. (b) I n vitro antibacterial assay d a t a are included in t h e various synthetic papers of this series. (10) G . R. Allen, Jr., L. J. Binovi, and AI. J. Weiss, J . Med. Chem., 10, 7 (1967). (11) G. R . .Ulen, Jr., and R I . J. Weiss, i b i d . , 10, 28 (1967). (12) G . R. Allen, Jr., J. F. Poletto, a n d &I. J. Weiss, ibid., 10, 1-1 (1967).
1
R,
substituent (Ri)
Methyl Ethyl" Ethyl n-Propyl Iwpropyl n-Butyl
AHJH, I11
xi Ir.LRI L K T S IiX XICEINFECTED \\ ITH
Staphylococcus aureus, SMITH
11.7
O
743
a
Median effectii e dose range (EDso), rnglkg sc
16-32 16-32 (16-32 16-32 32-12s >128 > 12s
PO)
NH?carbamate derivative.
in the parent antibiotics which is most prone t o chemical solvolysis4 of the fused aziridine ring. This possibility is of interest since biochemical evidence indicates that biological alkylation may play an important role in the mechanism by which the mitomycins exert their effect.I3 However, these compounds were among the more active analogs according to the in vitro assay.12 Variations at C3.-Here, exchange of the carbamoyloxymethyl side chain for hydrogen, methyl, hydroxymethy1,7b chloromethyl, methoxymethyl, acetylthiomethyl, f0rmyl,7~or acetyl groups gave compoundsi4 inactive a t the 128-mg/kg dose level (subcutaneous), nor were a variety of carboxylic acid esters'; of the hydroxymethyl group of any particular interest. Moreover, in contrast to the Cz series, these analogs, with the exception of the carboxylates, were essentially inactive in the in vitro as well as in the i n vivo assays, although in several instances the new functions also provided good leaving groups, so that in principle a site for biological alkylation was maintained. Variations of the substituents on the carbamate nitrogen proved more fruitful and it was possible to obtain analogs'j which showed a retention and even an enhancement of activity. In the n'-monoalkyl series, the methyl and propyl members were about as active as the original unsubstituted carbamate 111. Of the N,Sdialkyl derivatives, the dimethyl analog was the most interesting, affording substantial activity, which was also observed for the piperazine derivative. Introduction of the phenyl or allyl group gave compounds of relatively modest activity. Of considerable interest was the activity observed following monosubstitution of the carbamate nitrogen by various substituted alkyl groups, particularly by the P-hydroxyethyl group which afforded the most active member of the 5-methoxy and possibly also of the 5-ethylenimino series (see below). Results with the more interesting of the carbamate nitrogen-substituted derivatives are given in Table 11. Variations at C+-Peak activity a t this position in the quinone ring apparently was obtained with the (13) (a) V. N. Iyer a n d W.Szybalski, Science, 146, 55 (1964); (b) A. Weissbach a n d A. Lisio, Biochemistry, 4 , 196 (1965); (c) for a recent review see W.Szybalski and V. N. Iyer in "Antibiotics, ' Val. 1, D. Gottlieb and P. D. Shaiv, Ed., Springer Verlap, New York. N. Y., 1967, p 211. (14) Except a s noted, for these compounds see J. F. Poletto, G. R. Allen, Jr., a n d RI. J. Weiss, J . .%led. Chem., 10, 95 (1967). (15) J. F. Poletto, G. R. Allen, Jr., a n d .\I.J. Weiss, ibicl., 11, 000 (1968).
- ---EUso range, ing kp-Srilicutaneoue Oral I
li
Nllr
SIICII, ?JIICII,CEI,CI-I, NII f CI €2)3CII3 N(C1I j ) , s( ( x 2 c I I a ) 2 A
NWNcHJ
origind methyl group, since potency \I as diminished when this group was replaced by hydrogen16 (Spropylcarbamate, estd ED,, = 32-128 mg/lig sc) or ethyllo (S-methylcarbamate, estd ED,, > 128 mg/lig sc) . Variations at Cb.-Homologatioii of the 5-niethoxy function t o ethoxyl i n one instance (S-unsubstituted carbamate) resulted in a t least a lowering of activity, hut in another instance (9-niethylcarbamate) equivalent activity \vas noted. The reverse of this situation obtained in the 5-:tmino series' 11herein substitution of nicthoxy by the primary amino group aff urded it compound of equivalent activity for the S-unsubstituted carbamate and a compound of little or no activity for the S-methylcarbamate. Several s u b d t u t e d amino derivatives !$ere prepared, but none of these, with the cxception of the bcnzplaniino aiinlog (Smethylcarbttrnate, estd EDjo = 32-128 mg: lig PC) \I x s of interest. Substitution of the 5-methoxy groul) by hydrogen l 7 gave L: compound of lessened potencj (3-methylcarbarnate, estd EDjti 128 mgikg sc), :ind sub5titution by n1ethyll7 or cliloriiicl (good lc~iviiig group) gave compounds of no particular importance. The 5-Ethylenimino Series.----\laximum activity u as observed when the methoxy group n as replaced by the cthylenimino group. This exchange afforded several compounds xvith oral activity 21s. the tetracyclineresistant S. uuwus, strain Rose. and also S . pyogenes C203 (see Table III),ga a significant forward step sirice even the most potent members of the 3-methoxy series (also 5-ainino), according t o the Staphylococcus Smith asmy, were ineffective in vivo against these two organiwis. Introduction of the 2-nicthylethj-lenimino groiiij
-
[lti) W. A. Remers and A I . J. Weiss, J. .4m. Citem. Soc., 88, 804 (1966). ( 1 7 ) It. €1. Roth, W.A . Remers, and bI. .J. Weiss, .I. Org. Chem.. 31, 1012 (1968).
TABLE V INCREASE IN RESISTANCE TO IV
--
Initial MIC, pg/ml
Organisms
Compound I V S . aureus Smith S. aureus Rose S . pyogenes XY5 S. pyogenes C203 PIIitomycin C S. uureus Smith S. aureus Rose S. pyogenes NY5 S. pyogenes C203
AKD
lI1,romcm C
Fold increase in resistant---No. of transfers7 9 11 13 15
1
2
3
4
5
1.25 2 . 3 - 5 .0 1,25-2. .i 0.62-1.23
8 2 4 8
16 16 16 8
64 64 32 16
125 64 64 16
125 32
125 64
125 64
125 64
125 64
32
32
230
125
0.16-0.31 0,16-0.31 0.04 0.02
2 2 0 2
2 2 0 2
4 4 2 0
4 4 2 2
4
4 4 32 4
16 4 32 8
32 8 64 4
T.~BLE VI TOXICITY OF COMPOUND I V .4cute toxicity, --dead/totai'' Subcutaneous Oral
-
Dose, mg/kg
745
~IITOMYCINS. XXII
July 1968
I N LIlCE
Marrow nucleated cell count, 7controls/treatedbSubcutaneous Oral
312 2,'2 2/2 9.2 2.2 256 1/ 2 1j 2 6.3 2.0 128 0 p 012 2.1 1.o 64 0/2 0/2 1.3 1.1 0 Fourteen days after administration of I T'. * The count was made on the third day after administration of I V ; two mice per gioiip. The femoral marrow was used.
mas apparent, 2.5 pg/nil, more than 97% of the microorganisms introduced with the inoculum were no longer viable. At higher concentrations the bactericidal activity was even more pronounced. A study of resistance development by selected organisms to compound IV and mitomycin C was also carried out (see Table V). Toward IV, resistance developed more rapidly than toward mitomycin C, and a plateau was reached a t about the fourth or fifth transfer for the staphylococci and a t about the ninth transfer for S. pyogenes (2203. In contrast, the resistance developed to mitomycin C began to increase in a stepwise pattern for three of the organisms ranging from the sixth to thirteenth transfer, whereas for S.aureus Rose there was only a 16-fold increase a t the end of the 19th transfer. Note, too, that the initial minimal inhibitory concentrations for IV were about ten to fifty times as great as that observed for mitomycin C. A preliminary acute toxicity determination with IV and a bone-marrow depression study'g gave the results shown in Table VI. Gross examination of the contents of the abdominal cavity revealed gelatinous, toneless intestines in the dead mice. However, all survivors appeared normal and, upon sacrifice, gross examination of the abdominal cavity contents revealed nothing unusual. According to the data of Table VI, bone-marrow depression would appear to be definitely established by the subcutaneous route. However, the oral data, in view of the limited number of animals, is of questionable significance. I n any case, these toxicity manifestations are sufficient to preclude any interest in IV as a possible clinical antibacterial agent. Thus, the development of a useful antibacterial agent based upon the mitomycin class of (18) For t h e assay procedure see E. J. Kirsch, A. C. Dornbush, a n d E. J. Backus, ref 9a, p 205. (19) For t h e assay procedure see A. W. Vogel, Cancer Res., 21, 636 (1961).
4 16 2
17
19
250 125
125 125
125
250
500
250
230 8 125 32
123 16 250 32
500 16 500 500
16 500 2000
antibiotics, one of the most potent groups of orally effective antibacterial substances known, remains an achievement yet to be accomplished. The antibacterial activity manifested by the various indoloquinone carbamates leads to an observation of some interest to the question of the biochemical mechanism by which the mitomycins exert their biological effect. There is accumulated evidence that these antibiotics can cross-link DNA and that this ability to cross-link depends upon an initial biological red ~ c t i o n . ' ~I n the mitomycins the two alkylating sites required for cross-linking can be conceived as provided by the asiridine and the carbamoyloxymethyl functions; a third possibility, the methoxy- or amino-substituted carbon in the quinone ring, is inactivated by reduction to the hydroquinone state. It is, of course, quite possible that the antibacterial activity, as well as the toxicity manifestations, of these synthetic analogs and the mitomycins does not result from a common mechanism, or, if it does, this mechanism does not involve a cross-linking process. However, in the event of a common cross-linking mechanism, the observed activity of the desaziridino synthetic analogs requires that, in the mitomycins, the aziridino function, usually considered the biological alkylating group par excellence, in fact does not participate in the cross-linking process. Cross-linking still could be achieved by use of the carbamate and quinone sites. However, inasmuch as an initial biological reduction is r e q ~ i r e d , ~ ~aJ Osubsequent biological reoxidation, presumably after alkylation by the carbamate function, would then be necessary in order to reestablish the quinone site. Unfortunately, we can offer no biochemical evidence concerning these speculations. Finally, we would note that a broad selection of the indoloquinone carbamates prepared in the course of this study was submitted to an antitumor assay2' using the 72j mammary adenocarcinoma in C3H mice. In view of the important antitumor properties manifested by mitomycin C, i t was disappointing that none of the indoloquinones could be found active a t a nontoxic dose. Acknowledgment.-We wish to acknowledge the support and encouragement of Drs. J . J. Denton and J . A I . Smith, Jr., throughout the course of this investigation. (20) H. S. Sohwartz. J. E. Sodergren, a n d F. S. Philips, Sczence, 14, 1181 (1963). (21) A. W. Vogel a n d J. D. Haynes, Cancer Chemotherapy R e p t . , No. 22, 23 (1962).
