Study on the chemical reactivity of a cyano group in the 5-position of 4 and/or 6-substituted pyrrolopyrimidine nucleosides

The reactivity of the cyano group at the 5-position of the pyrrolopyrimidine nucleosides has been studied with various groups in the 4- and/or 6-positions. The susceptibility of the cyano group toward nucleophilic substitution under basic conditions as influenced by amino groups was found to be in the order: no amino group at positions 4 or 6>4-amino>6-amino>4,6-diamino. The decreased reactivity of the amino substituted derivatives is due to electron release by the amino groups. During the investigation of the reaction of 4-chloro-5-cyano-7-(B-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine with methyl hydrazine, a new tricyclic nucleoside, 6-amino-4-methyl-8-(B-D-ribofuranosyl)(4H ,8H)-pyrrolo[4,3,2-de]pyrimido[4,5-c]pyridazine, was synthesized. Subsequent studies showed that the intermediate was 5-cyano-4-N-l-methylhydrazino-7-(3-D-riboguranosyl)pyrrolo[2,3-d]pyrimidine and cyclized in situ to the tricyclic derivative.

A STUDY ON THE CHEMICAL REACTIVITY OF A CYANO GROUP IN THE 5-POSITION OF 4 AND/OR 6-SUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES by Karl Howard Schram A dissertation submitted to-the faculty of the University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Medi ci na 1 Chemi stry Department of Biopharmaceutical Sciences University of Utah June 1973 This dissertation for the Doctor of Philosophy Degree by Karl Howard Schram has been approved May 1973 ;1 I ./ member /) /�"""'/" c, ' <fl ,C! ',� (/ ', " )/�'� (/ Il : ' '"��' :,/ ,� ;"" _ _ _ , ,' ' Committee member �,/" .r ... :-..... // ',' -' , ; f " ! f ./,' • \ ; fl , ' , / ' r' ' .J / I� . --� � t ' � ��� l��-- / / � �e d u� S�c'h o o - -'G'r-a', -n -a �D at e ACKNOWLEDGMENTS I would like to express my deepest gratitude to my wife Cathy for her moral support and uncomplaining attitude during these years of schooling. I would also like to acknowledge the joy and inspiration which our daughter Lindsey has afforded us during the last couple of years. For professional guidance I would like to express my admiration to Dr. Leroy B. Townsend, who not only has directed my research but has also made this work an exciting endeavor. I would like to thank the United States Government for financial support through a National Defense Education Act, Title IV Fellowship and the G.I. Bill. Thanks are also extended to Mrs. Barbara Hinshaw and Miss Olga Leonoudakis for their assistance in the preparation of several compounds reported here;n. TABLE OF CONTENTS PAGE CHAPTER ACKNOVJLEDGt~ENTS iii LIST OF TABLES vi LIST OF FIGURES vii LIST OF REACTION SCHEMES ABSTRACT I. II. viii ;x . . . . . INTRODUCTION. 1 HISTORICAL ASPECTS. 3 Chemistry Synth(~s is of the natura 11 y occurri ng pyrro lo- pyr i nl'i d "irH~ nuc 1eo sid es • . . • • . • . . • . 3 Chemical modification of the pyrrolopyrinrid:lne nucleosides by electrophilic substitution and nucleoph~i1'ic d'isplacements G Biological Activity Biochemistry of the pyrrolopyrimidine nuclecsides. 14 Structure-activity relationships as antileukemic agents. . . . . . . . . . . . . 17 Preclinical trials using the pyrrolopyrimidine nucleosides III. 33 4,5-DISUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES . . • . . . . . . . . 40 CHAPTER PAGE 6-Amino-4-methyl-8-(s-D-ribofuranosyl) (4H,8H)pyrrolo[4,3,2-de]pyrimido[4,5-c]pyridazine . 40 Heterocyclic rings at position 5 . . . . . 45 IV. 5-SUBSITIUTED PYRROLOPYRIMIDINE NUCLEOSIDES. 55 Synthesis and reactivity of 5-cyaoo-7-(S-D- V. VI. VII. ribofuranosy1)pyrro1o[2,3-d]pyrimidine 55 Substitution at position 5 . . . . . . . 59 5,6-DISUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES. 62 4,5,6-TRISUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES 68 STRUCTURE-ACTIVITY RELATIONSHIPS 74 VIII. SUMMARY AND CONCLUSIONS IX. 82 EXPERIMENTAL 84 BIBLIOGRAPHY 104 VITA 109 . . . . . v LIST OF TABLES TABLE PAGE I. 18 I I. . 21 I I I. . 23 IV . . 25 V. • 26 VI. . 29 VI I. • 31 VIII. 35 I X. • 37 X. • 76 XI. . 79 XII .. 101 LIST OF FIGURES FIGURE PAGE 1. . . 32 LIST OF REACTION SCHEMES REACTION SCHEME PAGE 1. 5 2. 7 3. 9 4. 11 5. . 13 6. 44 7. . 48 8. 50 9. 53 10. . 58 11 . 60 12. 66 13. . 72 ABSTRACT The reactivity of the cyano group at the 5-position of the pyrrolopyrimidine nucleosides has been studied with various groups in the 4- and/or 6-positions. The susceptibility of the cyano group toward nucleophilic substitution under basic conditions as influenced by amino groups was found to be in the order: no amino group at positions 4 or 6>4-amino>6-amino>4,6-diamino. The decreased reactivity of the amino substituted derivatives is due to electron release by the amino groups. During the investigation of the reaction of 4-chloro-5-cyano-7(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine with methylhydrazine, a new tricyclic nucleoside, 6-amino-4-methyl-B-(S-D-ribofuranosyl)(4H,BH)pyrrolo[4,3,2-de]pyrimido[4,5-c]pyridaz;ne, was synthesized. Subsequent studies showed that the intermediate was 5-cyano-4-N-l-methylhydrazino-7(s-D-riboguranosyl)pyrrolo[2,3-d]pyrimidine and cyclized ~ situ to the tricyclic derivative. The synthesis of various aromatic and non-aromatic 5 and 6 membered heterocyclic rings in the 5-position of the pyrrolopyrimidine nucleosides has been achieved. Testing data indicate that a 6 membered aromatic heterocycle is the limit of size for activity to be retained. Non-aromatic rings are more active than aromatic rings, and 5 membered rings are more active than 6 membered rings. Activity was lower in this series than compounds which contain smaller groups at position 5. Testing data indicated that an amino group in the 4-position is not essential for antileukemic activity and may increase toxicity. Compounds with a thiocarboxamide or carboxamidoxime function at the 5- position possessed significant antileukemic activity. Electrophilic substitution at position 5 of the unsubstituted pyrrolopyrimidine ring by various halogenation reagents was observed. Bromine water effected substitution at position 6 of 5-cyano-7-(s-Dribofuranosyl)pyrrolo[2,3-d]pyrimidine. The 6-bromo group was displaced by various nucleophilic reagents to produce 5,6-disubstuted pyrrolopyrimidine nucleosides. Nucleophilic displacement of the 6-bromo group of 6-bromotoyocamycin by liquid ammonia was successful and various 4,5,6-trisubstituted pyrrolopyrimidine nucleosides were prepared. x CHAPTER I INTRODUCTION The isolation,I,2t3 characterization,4,S,6 and synthesis of the pyrrolopyrimidine nucleoside antibiotics toyocamycin(I), sangivamycin (II), and tubercidin(III) generated interest in this class of compounds as potential anticancer agents. Subsequent antitumor screening of these compounds confirmed that tubercidin and toyocamycin possessed significant activity against a number of experimental tumors. 9 ,lO However, both tubercidin and toyocamycin were found to be toxic to mammalian cell systems. 11 ,12 Sangivamycin also exhibits significant activity against HeLa cells grown in cell cultures and aga<inst leukemia L1210 in mice. 13 Sangivarnycin displayed no evidence of toxicity in mammalian test animals at maximally tolerated doses1 4 and is presently undergoing human clinical trial against 1eukemi a. NH2 C=N tD"'~ l::::;.,,,..-- HORo~ l-iH Toyocamyc;n(I) NH2 g-NH (JO '" I . HO r~ H20~ OH Sangivamycin(II) 2 NH2 do HO~ HO OH Tubercidin(III) Prel -j mi nary vlOrk in our 1abara tori es was des i gned to probe the structure-activity rel/;at"ionships and chemistry of these naturally occur- 2 ring nucleoside antibiotics. Of pr-imary interest was the finding that modification of the cyano group resulted in decreased toxicity and enhanced activity against leukemia L12l0 in mammalian test systems.l~ Also of interest was the finding that the chemical reactivity of the cyano group in position 5 was dependent to a large extent on the exocyclic substituent at position 4 in the pyrimidine moiety.16 In an effort to expand the knowledge of the pyrrolopyrimidine nucleosides, a study was initiated to determine the effect which various substituents at the 4- and/or 6-positions would have on the reactivity of a cyano group at position 5 and also the effect of these substituents on electrophilic substitution and nucleophilic displacement at other positions of the ring nucleus. Antitumor evaluation of these derivatives will expand the present structure-activity relationships and will include the effect of bulky substituents at the 5-position and also changes in activity due to groups at the 4- and 6-positions of the pyrrolopyrimidine nucleosides. A historical summary of the chemistry and biological activity of previously synthesized pyrrolopyrimidine nucleosides is also presented. Experimental details and physical data on all new compounds have been collected in the experimental section. Screening data, from the Drug Research and Development Branch of NCI, on compounds pertaining to this study have been summarized and compiled ;n tables found in the historical and biological activity sections. CHAPTER II HISTORICAL ASPFCTS I Chemistry Synthesis of the naturally occurring pyrrolopyrimidine nucleosides. The first pyrrolo[2,3-d]pyrimid1ne nucleoside was synthesized in 1963. 17 ,18 Condensation of 4-amino-5-(2,2-diethoxyethyl)pyrimidin6-one with 5-Q-trityl-2,3,4-tri-Q-acetyl-aldehydo-Q-ribose afforded, after six steps, the 7-ribosyl derivative of pyrrolo[2,3-d]pyrimidin4-one (IV). Deamination of tubercidin (III) with nitrous acid furnish- ed a product which was identical to the 7-ribosylpyrrolo[2,3-d]pyrimidin-4-one prepared by the condensation reaction. Attempts 18 to synthesize tubercidin from IV by treatment with phosphorous oxychloride and displacement of the chloro group by ammonia were unsuccessful. The first attempt to synthesize toyocamycin (I) was reported in 1964. 19 ,20 Based on previous studies,21 the aglycone of toyocamycin, 4-amino-5-cyanopyrrolo[2,3-d]pyrimidine (V), was prepared starting with tetracyanoethylene and hydrogen sulfide. An attempt to ribosylate V by the chloro-mercury salt method followed by deacetylation gave a product (less than 1% yield) which could not be fully characterized as toyocamycin. Further attempts to synthesize the naturally occurring pyrrolopyrimidine nucleosides are not recorded in the literature until 1967. Gerster, et al.,22 prepared a number of 4-substituted pyrrolopyrimidine nucleosides to study the structure-activity relationships 4 of the pyrro10pyrimidine nuc1eosides with respect to antitumor activity. The starting material in this study was 7-(S-D-ribofuranosy1)pyrro10[2,3-d]-4-pyrimidone (IV).17,18,23 Acetylation of IV was accomplished in good yield using a mixture of acetic anhydride and pyridine to afford the triacety1 derivative VI. Chlorination using neat phosphorous oxychloride gave a 70% yield of 4-ch10ro-7-(2,3,5-tri-~-acety1-S-D-ribo­ furanosy1)pyrro10[2,3-d]pyrimidine (VII). Previous attempts at chlorination of VI had included N,N-diethy1ani1ine as a catalyst which apparently caused decomposition and/or polymerization. Removal of the acetyl groups of VII was accomplished using methano1ic ammonia under mild conditions to furnish 4-ch10ro-7-(s-D-ribofuranosy1)pyrro10[2,3-d]pyrimidine (VIII). When VIII was heated with methano1ic ammonia at 150 0 for 3 hours in a sealed reaction vessel, nucleophilic displacement of the 4ch10ro group occurred to give a 50% yield of tubercidin (III). Other nucleophilic displacements of the 4-chloro group of VIII will be discussed later. Thus, the first synthesis of a naturally occurring pyrrolopyrimidine nucleoside had been accomplished. The preparation of toyocamycin (I) by three alternate routes was subsequently reported by Tolman, et ~.24,25, with sangivamycin (II) and tubercidin (III) then being prepared from I. Preparation of the bases IX, X, and XI was accomplished using 2-amino-5-bromo-3,4-dicyanopyrro1e 21 as the starting material. Fusion of IX or X with 1,2,3,5tetra-~-acety1ribose (TAR) furnished, after deacety1ation, 4-amino-6- bromo-5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XII). Removal of the 6-bromo group was accomplished using 5% palladium on charcoal to give toyocamycin (I) which was identical in all respects to an authentic sample. Similar fusion of XI with TAR produced toyoca- 5 REACTION SCHEME 1 o o HI' > ~. OH HO VI IV Cl CI > HOCH2 VIII VI I HO OH III 6 mycin directly. The increased yields of condensation products from IX, X, and XI (9%, 31%, and 28% respectively) compared to previous ribosylation attempts 19 ,20 was attributed to the presence of the halogen substituents in the bases 26 ,27,28 Also of interest was the apparent lack of steric effects 29 ,30,31 due to the 6-bromo group of IX and X. The authors concluded that the steric hindrance was compensated for by the electronic factors of the 6-bromo group. Preparation of sang;vamycin (II) and tubercidin from toyocamycin was then accomplished and the structural relationship of these nucleosides was established for the first time. Treatment of toyocamyc;n (I) with 30% hydrogen peroxide in concentrated ammonium hydroxide, gave a 65% yield of sangivamycin (11)25 which proved to be identical in all respects to authentic sangivamycin. Tubercidin (III) was synthesized from toyocamycin (I) by hydrolysis of the cyano group to sangivamycic acid (XIII) by refluxing I in 3 N hydrochloric acid under a nitrogen atmosphere for 12 hours. Decarboxylation of XIII by immersion in an oil bath heated to 238 0 for 10 seconds produced III in 13% yield. This work established the complete structure and achieved the total synthesis of all the known pyrrolopyrimidine antibiotics. Meanwhile, attention had already turned to the modification of the base moiety of I, II, and III in attempts to discern their chemistry and produce more active antitumor agents. Chemical modification of the pYrrolopYrimidine nucleosides Bt electrophi1ic substitution and nucleophilic displacements. The synthesis of 4-ch1oro-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (VIII)22 has been discussed earlier. The displacement of the 4-chloro group by various nucleophi1es was then undertaken. Nucleophilic displacement 7 REACTION SCHEME 2 NHAc CI Br Br H IX H XI X HO XII OH OH HO I 8 of the chloro group proceeded smoothly with sodium methox;de, dimethylamine, piperidine, hydroxylamine,32 and methylamine to afford the 4methoxy, 4-dimethylamino, 4-piperidinyl, 4-hydroxylamino, and 4-methylamino derivatives. Dechlorination was observed when VIII was hydrogenated with palladium on charcoal and produced 7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine(7-deazanebularine). All of the compounds prepared in this study had purine counterparts which possessed a wide variety of biological and chemotherapeutic activity_ It was hoped that the more stable glycosidic linkage of the pyrrolo[2,3-d]pyrimidine nucleosides 23 as compared to their purine analogs would enhance the activity compared to the purine series. The susceptibility of 4-substituted pyrrolopyrimidine nucleosides to electrophilic halogenation and the selective nucleophilic displacement of a halogen in the 4-position rather than a halogen in the 5-position was then demonstrated. 33 Treatment of 4-chloro-7-(2,3,5-tri-Q-acetyl-s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (VII) with N-bromoacetamide in methylene chloride provided a good yield of the 5-bromo derivative of VII. The position of electrophilic substitution was established using pmr. Deacetylation was accomplished without concomitant displacement of either of the halogens using methanolic ammonia at 5°. Under more stringent conditions, selective displacement of the 4-chloro group over the 5-halogen was observed in all cases. For example, treatment of 5bromo-4-chloro-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine with methanolic ammonia at 150 0 in a sealed reaction vessel gave a good yield of 5-bromotubercidin. of the 5-bromo group. There was no indication of displacement 9 REACTION SCHEME 3 > HO OH II I > HO OH XIII HO III OH 10 Bromination of toyocamycin (I) at position 6 was observed when I was treated with bromine water at room temperature. 34 ,38 This compound (XII) was identical in all respects to the 6-bromotoyocamycin produced during the total synthesis of toyocamycin. Since pyrrole and condensed pyrrole systems are poor substrates for nucleophilic displacements,3S,36 the presumption was made that nucleophilic addition would occur at the carbon of the cyano group. When XII was refluxed in an ethanolic solution of the hydroxylamine, the expected 4-amino-6bromo-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidoxime was obtained in 75% yield. However, under identical conditions, the reaction of XII with hydrazine in ethanol produced a derivative of the new tricyclic system pyrazolo [3 1 ,4 1 :5,4]pyrrolo[2,3-d]pyrimidine. 37 Careful monitoring of the reaction permitted the isolation of 4-amino5-cyano-6-hydrazino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine and established that nucleophilic displacement of the 6-bromo group by hydrazine was the initial reaction. The 6-hydrazino group then cyclized to the 5-cyano group and formed the tricyclic derivative. The final paper of interest as background to the chemistry of pyrrolopyrimidine nucleosides was a study on the relative reactivity of the 5-cyano group of toyocamycin (I) and desaminotoyocamycin under acidic and basic conditions. 16 Treatment of toyocamycin with either hydrazine or hydroxylamine in refluxing ethanol provided the carboxamidrazone (XIV) and the carboxamidoxime (XV) derivatives in good yield. When I was treated with methanolic ammonia in an effort to prepare the carboxamidine (XVI), only I was recovered, even under drastic conditions. This indicated that although the cyano group of I was susceptible toward nucleophilic 11 REACTION SCHEME 4 XV XVI XIV I XIX XVII R_H0l>~ ~~ XVIII 12 attack, a strong nuc1eophi1e was required. The synthesis of XVI was accomplished by the reduction of XV with palladium on carbon under a hydrogen atmosphere. Reduction of XV with Raney nickel under slightly basic conditions hydrolyzed the amidoxime group to afford sangivamycin (II). Nucleophilic attack at the cyano group was also observed on treatment of I with hydrogen sulfide and triethylamine in pyridine to afford XVII. The attempted synthesis of the very reactive methyl 4-amino-7-(s-Dribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-thioformimidate (XVIII) by alkylation of XVII was unsuccessful since XVIII readily eliminated methanethiol to give toyocamycin. When I was reacted with hydrogen chloride in ethanol, ethyl 4-amino-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-formimidate hydrochloride (XIX) was isolated, but only after prolonged treatment. It was noticed that under basic conditions, I reacted readily and was initially very soluble. Under acidic conditions, however, I reacted slowly and was insoluble in the acid media. When I was deaminated to 5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-dJpyrimidin4-one (XX), the opposite reactivity was observed, i.e., under acid conditions, XX reacted readily but under basic conditions, reactions of XX were slow. XX formed the imidate XXI readily but reaction with hydraz"ine was not observed. The carboxamidrazone XXII was finally prepared by displacement of ethanol from XXI by hydrazine. Hydroxylamine did react with XX to form the carboxamidoxime but extended reaction times were necessary. Likewise, the addition of hydrogen sulfide with triethylamine in pyridine to form the thioamide XXIV was unsuccessful. The synthesis of XXIV required the treatment of XX with ethanolic sodium sulfide in a sealed vessel at 125 0 for 6 hours. The authors proposed that this tremendous difference "in reactivity of the cyano group of I versus 13 REACTION SCHEME 5 OEt I =NH > o ~NH2 " C-NH2 Hlw I ~ I XXII R XXI / I~ H I ~xxv XXIII XXIV 14 XX was due to the difference in the pyrimidine moiety. Under basic conditions, XX would be expected to lose the proton at N-3. The increased electron density of the system would then exert a deactivating influence on the cyano group. Repulsion of the incoming nucleophile by the anion in the pyrimidine ring would also affect the reactivity of the cyano group. Under basic conditions none of these effects should be seen with I. Under acidic conditions I is probably protonated at N-3. Since protonation of the cyano group would be inhibited by the positive charge in the pyr'imidine ring, reduced reactivity would be observed under acid conditions. The keto group of XX should not be protonated as readily as the amino group of I and protonation of the cyano group and reaction with a nucleophile would be more likely to occur. In summary, previous studies have shown that electrophilic attack can occur in the 5- and 6-positions by halogens. Halogens at the 4- and 6-positions can be displaced by nucleophiles, and the reactivity of the cyano group at position 5 depend~ on the group at position 4. II. Biological Activity Biochemistry of the pyrrolopyrimidine nucleosides. One of the main features which attracted interest to the pyrrolo[2,3-d]pyrimidine nucleosides was the structural similarity of the pyrrolopyrimidine ring system to the purine ring system. Comparison of the two systems reveals only one difference. In the pyrrolopyrimidines, a methine group (CH) 6 7 1~~8 2~~9 3 H 15 has been substituted for the ring nitrogen at position 7 of the purine ring. To emphasize this very close sinlilar1ty of structure, the pyrrolopyrimidines are sometimes referred to as 7-deazapurines. As one would expect, the pyrrolopyrimidines are substrates for many of the enzymes involved in purine synthesis and metabolism, protein synthesis, and DNA and RNA synthesis. The naturally occurring pyrrolopyrimidine nucleosides have also been used as biochemical tools for studying the structural requirements for interaction at the catalytic and regulatory sites of various enzymes. A study by Suhadolnik, et al.,38 on the structural requirements for interaction at the catalytic site of ribonucleotide reductase from ~ leschmannii found that the triphosphates of toyocamycin (I), sangiva- mycin (II), and tubercidin (III) were reduced to the 2 1 -deoxynucleotides. The authors concluded that C8 and N9 of the base moiety are important structural features but that N7 may be replaced by CH without affecting the binding at the catalytic site. In an extension of this work, Chassy and Suhadolnik39 studied the reduction of the pyrrolopyrimidine triphosphates by ~ coli ribonucleotide reductase. In this case, both tubercidin and toyocamycin were substrates, but sangivamycin triphosphate was not. The authors postulated that the carboxamide group, being larger than a hydrogen or cyano group, was not able to fit into the catalytic site for steric reasons. These findings substantiated the previous suggestion that N7 of the imidazole ring is not required for binding at the catalytic site of these reductases. Of particular importance to the chemotherapeutic activity of the pyrrolopyrimidines is the fact that tubercidin is completely resistant to the action of the enzyme nucleoside phosphorylase. 23 16 This enzyme, in the presence of inorganic phosphate, converts a nucleoside to the free base and ribose-1-phosphate. The stabilized glycosidic bond is important because by increasing the strength of the base-sugar bond, the nucleoside remains intact and therefore, can be converted to the mono, di, and triphosphates with the possibility of being incorporated into DNA or RNA, i.e., undergo lethal synthesis. The main pathway for the metabolism of many amino nucleosides is deamination to a keto group by the enzyme adenosine deaminase. 4o The deamination of naturally occurring purine ribonuc1eosides and analogs which have been synthesized or isolated from nature has made this enzyme appear largely as a culprit which inactivates potentially useful compounds. 41 The effect on tubercidin of a purified adenosine deaminase obtained from intestinal mucosa has been studied. 42 Results showed that tubercidin was neither a substrate nor an inhibitor of adenosine deaminase, indicating that tubercidin is not bound to any extent at the active site of this enzyme. In fact, the only enzymatic alteration observed for tuberc;din was phosphorylation by adenosine kinase_ Similar studies on toyocamycin 43 showed the same results with adenosine deaminase, i.e., toyocamycin is not deaminated by adenosine deaminase. The incorporation of tubercidin triphosphate (TuTP), toyocamycin triphosphate (ToTP), and sangivamycin triphosphate (SaTP) into RNA and DNA has been shown to occur in hcoli and !!:.-lysodeikticus_ Incorporation studies 44 have shown that the rate of formation of poly (Tu-G) directed by poly d(T-C)-d(A-G) was much lower than that of poly(A-G), but that after long incubation the final yield of poly(Tu·G) was at about the level of poly(A·G). Later results 45 indicate that TuTP, ToTP, andSaTP are all incorporated into alternating (A-U)-like ribo- 17 nucleotide polymers with ~ col B RNA polymerase and that these pyrrolo- pyr-imidine nucleotides serve as analogs of ATP. Similar results were obtained 46 using DNA dependent-RNA polymerase from ~ lysodeikticus with SaTP replacing ATP. The importance of the incorporation of analogs of the "normalll purine nucleotides into RNA is illustrated by the fact that toyocamycin and tubercidin specifically inhibit the formation of t-RNA and protein synthesis in mouse fibroblasts. 47 ,48 Both effects were attributed to an altered structure of RNA. Structure-activity relationships as antileukemic agents. The first structure activity study on toyocamycin was published soon after it was isolated. 49 The results of this study are shown in Table I. These results seemed to indicate that among the compounds related to toyocamycin, the active cancericidal action in vitro depends on the coexistence of the amino group at position 4, the cyano group at C5, and the ribofuranose moiety at position 7. Deamination of the 4-am"ino group decreased activity and toxicity. When the cyano group was converted to either the aminomethylene or carboxylic acid, a further decrease in activity was observed. Tubercidin was found to be less toxic and less active than toyocamycin. ~ vivo tests showed slight activity for desaminotubercidin but in every instance the ribofuranose moiety at position 7 was required for activity_ The inactivity of the aglycones per se was attributed 47 to their failure to be converted to nucleotides by the enzymes AMP:pyrophosphate phosphoribosyltransferase or IMP:pyrophosphate phosphoribosyltransferase. Two important conclusions can be drawn from this work. The first is that only the nucleosides of pyrrolo[2,3-d]pyrimidines possess antineoplastic activity. The second is that conversion of the amino group 18 TABLE I R' RI NH2 CN R OH CN OH I Rill in vitro LDso +++ 20 Dose(mg/kg) in vivo 1 70% R 20 no effect CH2 NH 2 R 2 no effect OH COOH R 2 no effect OH CH 3 H 10 no effect NH2 H R 27 1 24% OH H R 50 2 22% NH2 H H 300 50 no effect OH H H 400 50 no effect SH H H 350 50 no effect C1 H H 300 50 no effect R = s-D-ribofuranosyl ++ 19 at position 4 to a keto moiety decreases toxicity with a concomitant decrease in antitumor activity. Based on the findings that 9-alky1purines 5o and l-alkylpyrazolo[3,4-d]pyrimidines 51 had shown anticancer activity and that certain 9-alkylpurines inhibited the growth of cells resistant to 6-mercaptopurine,52 Montgomery and Hewson 53 prepared several 7-alkylpyrrolo[2,3-d]pyrimidines (XXVI). Derivatives of XXVI prepared were R equal to ethyl, benzyl, cyclopentyl, and 3-hydroxymethylcyclopentanyl. Testing agmnst leukemia L12l0 showed that these tubercidin analogs were cytotoxic but with no activity against leukemia. Therefore, these results confirmed the earlier findings that pyrrolopyrimidines are active antileukemic agents only when substituted at position 7 with a ribofuranosyl moiety. NH2 ~ I R 1-alkylpyrazolo[3,4-d]pyrimidines N I R 7-alkylpyrrolo[2,3-d] pyrimidines (XXVI) The effect of replacing the oxygen in the ring of the ribose moiety by a sulfur atom has been investigated. 54 A comparison of the activity against leukemia L12l0 of the 41-thio derivatives of toyocamycin versus toyocamycin was made and the results were summarized briefly by the authors who stated that toyocamycin was approximately ten times more potent than the 41-thio analog. The 6-amino and 6bromo-4 -thiotoyocamycin derivatives showed even less activity with the 1 6-bromo derivative being least active of all. Concentration for 50% I growth inhibition of leukemia L12l0 were toyocamycin, 4x 10-8;4 thio- 20 toyocamycin 4x 10-7;4 -thio-6-aminotoyocamycin, 6x 10- 7 1 41-thio-6- bromotoyocamycin, 5x 10- 6 . The antitumor activity of pyrrolopyrimidine nucleosides with groups other than amino or keto at position 4 has been described. 22 Results of testing performed by Drug Research and Development (DR&D) are shown in Table II. Although none of the tubercidin derivatives were active enough to pass Stage I of the sequential screen, a difference in toxicity is noted for the various derivatives with the thiol, chloro, and N-piperidyl being relatively non-toxic. Testing results from other studies have not been published. However, data from our files on compounds prepared in many of the investigations discussed in the chemistry section are herein compiled. It should be mentioned that some compounds have a large amount of testing data while others have only one or two data sheets. Where large amounts of data are available, as for sangivamycin and the amidoxime XV, an effort has been made to select data which are indicative of the activity and toxic levels. Testing data on the 4,5-disubstituted pyrrolopyrimidine nucleosides are shown in Tables III and IV. Since none of the 5-halotubercidin derivatives were active enough to pass Stage I of the sequential screen, it must be concluded that halogenation blocks the action of tubercidin against neoplasms. The effect of conversion of the amino group at position 4 to other groups is the same as was seen in Table II. Toxicity and activity of the 5-halo derivatives and the derivatives unsubstituted at position 5 are about the same. This would seem to indicate that halogenation in the 5-position had no effect. However, the reduced toxicity of the 5-bromodesaminotubercidin compared to desamino- 21 TABLE II x NSC No. 99439 101160 101161 107519 Dose(mgLkg) Surv. j/e! 100 25 25 12.5 6 3 0/6 4/6 6/6 6/6 6/6 6/6 126 109 106 100 OCH 3 100 50 25 0/6 4/4 6/6 101 100 CI 400 200 100 50 25 5/6 4/6 4/4 4/4 4/4 110 104 110 80 40 20 10 5 0/4 4/4 4/4 4/4 4/4 106 106 109 107 X OH H 100278 SH 500 6/6 116 105826 SCH 3 400 200 100 50 2/6 6/6 6/6 6/6 98 97 103 97 T/C= increased life span of test animal(T)/life span of control animal. 22 TABLE II (continued) NSC. No. 101159 103791 100279 X SCH2C6Hs NHCH3 N(CH 3)2 103798 NCsH 10 122816 NHOH Dose(mg/kg) Surv. T/C% 400 200 100 50 25 '1/6 6/6 6/6 6/6 6/6 116 98 107 400 200 100 50 25 1/4 0/4 1/4 4/4 6/6 125 30 7.5 0/6 1/6 6/6 120 400 200 100 50 6/6 4/4 4/4 4/4 95 98 95 93 40 20 10 5 0/6 0/6 6/6 5/6 118 110 110 97 111 23 TABLE III NSC No. X 124149 Cl 103302 113939 Dose Br I Surv. -- T/C% 20 10 5 2.5 0/6 3/6 6/6 6/6 108 100 50 10 5 2.5 0/6 3/6 6/6 6/6 100 100 80 20 10 5 0/6 6/6 6/6 6/6 103 101 102 CI 124148 Cl 400 6/6 104 103803 Br 400 200 6/6 6/6 112 107 113940 I 400 200 6/6 6/6 112 110 24 tubercidin and the inactivity of the 5-halotubercidin derivatives reveal a significant effect. The halogens may sterically block the enzymatic conversion of these compounds to their active forms. For instance, it can be postulated that halogenation at position 5 sterically prevents the derivative from acting as a substrate for a kinase which, in turn, would prevent lethal synthesis. Tables V, VI, and VII show testing data of toyocamycin and sangivamycin derivatives synthesized in our laboratory.16,38,62 The first observation of importance in perusal of this data is that conversion of the amino group at position 4 to a keto or thio group abolishes activity (see Table V). The same trend was noted in the tubercidin series as well as the 4-keto derivative of sangivamycin. The lack of activity would be explained in all of these cases if the keto and thio derivatives were not substrates for the various metabolic enzymes. The effect of modifying the cyano group and leaving the amino group intact is shown in Table VI. Of particular interest here is the significant activity shown by the cyano derivatives, except the amidrazone XIV. With no testing data on toyocamycin, it is not possible to make a comparison to toyocamycin. However, a theory has been recently proposed that attempts to correlate the antileukemic activity of a wide variety of compounds. 55 Examination of such diverse nonalkylating compounds as sangivamycin, vincristine, 5-azacytidine, camptothecin, and methotrexate showed that they all possessed a common structural feature which consisted of a triangulation composed of one nitrogen and two oxygen atoms with rather definite interatomic distances. (See Figure I.) The authors porposed that the triangulation pattern may 25 TABLE IV NSC No. 103802 X NH2 HO OH Dose Surv. T/C% 400 200 100 10 5 1 .25 0/6 0/6 0/6 3/6 6/6 6/6 100 100 107516 OH 400 200 100 50 4/4 4/4 6/6 4/6 121 112 110 107518 SH 400 200 100 6/6 4/4 4/4 128 110 104 113942 SCH 3 400 200 100 6/6 6/6 6/6 114 108 100 109312 NHCH 3 80 40 20 10 0/6 2/4 4/4 4/4 104 104 N(CH a)2 80 40 20 10 0/4 4/4 4/4 4/4 108 104 102 NHOH 80 10 5 2.5 0/4 4/4 4/4 6/6 101 103 100 109311 105825 26 TABLE V NSC No. 109310 X OH Y - CN Dose Surv. 200 100 25 6 0/6 4/4 6/6 4/6 96 95 87 T/C% 116097 SH CN 400 200 100 6/6 6/6 6/6 109 108 103 106098 SCH 3 CN 300 200 133 100 30 20 10 0/6 6/6 5/6 6/6 6/6 6/6 6/6 140 127 124 146 136 e-NH z 400 200 100 6/6 6/6 6/6 109 103 97 8-NH 2 200 100 0/6 4/6 145 116284 OH 145389 NHNH z 141 0 27 TABLE V (continued) NSC No. X - -Y Dose Surv. T/C% 117839 C1 CN 400 200 100 6/6 6/6 6/6 103 104 103 143686 N(Me)2 CN 100 50 25 12.5 6.25 0/6 6/6 6/6 6/6 6/6 100 117 115 92 400 5/6 125 100 50 35 25 12.5 6.25 3/6 6/6 6/6 6/6 6/6 6/6 100 125 150 127 116 50 25 12.5 6.25 3.12 1/6 6/6 6/6 6/6 6/6 100 137 130 138 ~ 134336 CH3NH C-NH2 136560 CH3CH2NH ~-NH2 154829 NHOH ~OH -NH2 163648 OCH3 CN 200 100 4/6 6/6 93 125 107519 H H 80 40 20 10 5 0/4 4/4 4/4 4/4 4/4 106 106 109 106 28 contribute in the bonding to one of the pertinent receptor sites in certain biopolymers involved in leukemia genesis. Although the theory is quite empirical and other factors may be involved, the data presently being discussed seem to lend some credence to the idea. It is interesting that the amino group in position 4 of the pyrrolopyrimidine ring is not included ih the triangulation pattern. There may be a slight steric effect which reduces the activity of the compounds shown in Table VI. Sangivamycin has the smallest substituent (0) on the carbon atom adjacent to the amide amine and is the most active. The next most active compound is the -imidate XIX. Although the ethoxy group is large, there is a good possibility that the imidate is hydrolyzed to sangivamycin in vivo. Activity then decreases in the order thioamide XVII, amidoxime XV, amidine hydrochloride XVI, and amidrazone XIV. This is roughly the order of increasing size. One might expect the ami dine to be more active than the amidoxime because of the smaller size. However, this is not the free amid-ine but the hydrochloride salt, and the electron density of the amine group may be substantially reduced. The effect of adding a bromine to position 6 of these very active compounds is shown in Table VII. None of the compounds with a bromine in position 6 is as active as those compounds with hydrogen at position 6. However, activity is still present and, in some cases, sufficient to pass Stage I of the sequential screen. The important effect of the bromo group in the 6-pos;tion is to reduce the toxicity by a significant degree. Molecular models of the interesting 4,5-diamino-8-(B-D-ribofuranosyl)pyrazolo[3~41:5,4]pyrrolo[2,3-d]pyrimidine show that the 29 TABLE VI NSC No. Cmpd. 65346 II 107512 105327 XV XVII Y lNH2 ~OH -C-NH2 -~-NH2 Dose Surv. T/C% 25 20 16 8 5 4 2 1 0.25 0.20 0/6 1/10 6/6 10/10 6/6 10/10 10/10 10/10 10/10 10/10 254 310 275 231 244 165 167 147 80 50 25 8 5 4 1 0.5 0/10 8/10 10/10 10/10 6/6 6/6 10/10 10/10 159 157 151 172 180 160 134 40 30 20 15 10 5 2 1 0.5 0/4 2/6 7/8 6/6 8/8 6/6 6/6 6/6 6/6 141 146 207 158 176 167 137 30 TABLE VI (continued) NSC No. Cmpd. 131663 XVI 107517 128664 XIV XIX Y ~H.HCl - -NH2 Dose Surv. T/C% 150 80 40 20 10 5 2 0/6 6/6 6/6 6/6 6/6 6/6 6/6 166 153 144 129 116 133 10 5 2 0.4 0/4 4/4 6/6 6/6 109 100 100 15 10 5 2 1 0.5 0.2 0/6 6/6 6/6 6/6 6/6 6/6 6/6 143 136 200 272 150 128 ~NH2 -C-NH2 ~H -C-OEt . HCl 31 TABLE VII NSC No. Y -Z - Dose Surv. T/C% 113951 CN Br 600 400 250 170 100 5/6 6/6 6/6 6/6 6/6 131 131 127 115 123 113943 ~-NH2 Br 600 400 200 100 50 6/6 6/6 6/6 6/6 6/6 129 150 136 137 105 117836 -C-NH2 Br 400 200 100 6/6 6/6 6/6 135 97 98 117838 -~-NH2 SCH 3 400 100 6/6 6/6 121 114 400 200 100 50 35 4/6 6/6 6/6 6/6 6/6 121 106 149 150 121 ~OH 117837 Ribose 32 oII C-NH II , 2 '/ Triangulation for sangivamycin ----------------N 1.08 0.56)\ Triangulation Parameters Figure 1 ~ 33 the amino group at position 5 in the pyrazole ring should still fall within the limits of the triangulation theory; this is substantiated by the activity shown for this compound. ,Also, the toxicity levels of this tricyclic nucleoside may indicate that substitution in the 6position by any group, not only bromine, may reduce toxicity. The conclusions which can be drawn from these data are that the amino group at position 4 appears to be a requirement for activity, halogenation at position 5 in the tubercidin series eliminates activity, modification of the cyano group of toyocamycin increases activity, substitution at position 6 in the toyocamycin series reduces toxicity, and only the nucleosides are active. Preclinical trials using the pyrrolopyrimidine nucleosides. Phase I studies designed to determine the safe dosage levels for humans have been performed for toyocamycin, tubercidin, and sangivamycin. Although it showed little or no antitumor effect against transplanted mouse tumors and human tumors transplanted into conditioned rats, toyocamycin was chosen for a Phase I human toxicity study because it demonstrated marked cytotoxicity in tissue culture equal to or greater than other chemotherapeutic agents. 56 Pharmacoloqical and toxicological studies in animals 57 led to a starting dose of 10 ~g/kg in humans. None of the patients could be treated by radiation or surgical procedures and most had far advanced disease, but their general condition was good. Tumor measurements, subjective, and objective antitumor evaluations were not made. A total of 23 patients were included in the study and a total of 35 courses of the drug were administered. Injections were given by syringe directly into the vein at doses up to 80 ~g/kg or above. A 5-day course was chosen in all cases. A review of the animal data in- 34 dicated that primary toxicity should occur in the gastrointestinal tract, kidneys, and liver. The major untoward effect of the drug was found to be local toxicity at the site of injection or infusion. Severe local necrosis were seen at the site of injection if the drug infiltrated into subcutaneous tissue. With higher injected doses severe phlebitis occurred without infiltration. At a dose level of 200 ug/kg the local toxic effects precluded continuation of the study. There was no evidence of other toxic reactions or antitumor effect. Clinical studies with tubercidin indicated that the significant toxic reactions are renal damage and local irritation of veins. 58 The study used 93 patients with various forms of advanced cancer with a life expectancy of more than 60 days. The initial dose was 0.025 mg/kg/day and was elevated gradually to 0.3 mg/kg/day. The most serious problem encountered was renal toxicity manifested mainly by proteinuria, azotemia, or both, with proteinurea more common. Nephrotoxic reactions were rare at doses less than 0.2 mg/kg/day. The other major toxic manifestation.was venous thrombosis which, in some cases, was prominent enough to interfere with subsequent administration. The clinical abnormalities due to tubercidin treatment are shown in Table VIII. In this study attempts were made to evaluate any possible antitumor effect wherever measurable lesions existed. In three instances there was a suggestion that this drug exerted a favorable effect on the tumor being treated. All three cases were of primary pancreatic carcinoma. One patient reported substantial subjective improvement, but no change was observed in the measurable tumor. A second course of therapy was given 35 TABLE VIII Clinical Abnormalities After Treatment With Tubercidin (Intra-venous Administration) System Mild or Moderate Severe 11 7 5 Renal Venous (thrombosis) Gastrointestinal Hematopioetic Hepatic 7 o o o 10 6 2 Summary of Toxicity: Phase I Study of Tubercidin With EDTA (Administered in Whole Blood) Dose 200mg/kg x 2 1000mg/kg x 2 1200mg/kg x 2 1500mg/kg x 8 No. Patients Liver Renal GI 8 22 7 1a la 2a o a: slight to moderate toxicity b: severe toxicity 3a 1b o o 36 without apparent benefit. Another patient evidenced moderate reduction in size of an intra-abdominal mass and experienced subjective improvement. After receiving three courses of treatment, the therapy had to be discontinued because of difficulty in find-ing patent veins. A third patient experienced both subjective and objective improvement after tubercidin therapy. The size of an intra-abdominal mass decreased after the initial treatment. After receiving seven courses the veins proximal to sites of injection had become badly thrombosed and further treatment was not possible. The tumor, however, remained in regression. It has been reported 59 that over 95% of the tubercidin added to animal or human blood in vitro is readily absorbed by the erythrocytes and is maintained in the phosphorylated form without altering the half-life of the cells. The tubercidin absorbed by red cells is released over a long period of time. Human blood cells are saturated with tubercidin at a concentration of 200 to 400 ~g/ml whole blood. Using this observation,60 tubercidin was administered in patients own blood in an effort to circumvent the venous thrombosis and necrosis of tissue caused by direct injection. Forty-five patients were involved in the study and dosages ranged from 200 ~g/kg to 1500 ~g/kg. Tubercidin was administered by withdrawing 300-500 ml of whole blood withdrawn by phlebotomy into sterile plastic bags containing 50 ml of 1.5% EDTA (to prevent coagulation). Tubercidin in the appropriate dosage was added to the blood, incubated at 37° for one hour and re-transfused into the patient. A summary of the toxic effects are shown in Table VIII. Although tumor response was not the objective of this study, four patients with evaluable lesions experienced a significant regression of their malignancy. Three of the patients had carcinoma of the pancreas 37 TABLE IX Summary of Toxic Effects of Sangivamycin Daily Administration ~osa~e mq/ g/day) 10 20 50 60 120 Duration (days) Total dose (mg/kg) 87 10 12 26 54 15 10 41 0.93 0.1 0.11 0.34 1 .24 0.73 0.54 2.83 Toxic effects () 0 0 Mild leukopenia Mild leukopenia Mild nausea 0 0 Weekly Administration 32 39 21 32 14 14 35 0.8 0.35 0.31 0.4 0.3 0.2 0.51 30 31 35 32 28 75 21 25 21 42 0.56 0.43 0.4 0.4 0.9 0.11 0.46 0.50 0.45 0.80 50-150 150 50 1 .73 0.60 150-300 30 28 O. 13 36 1 .23 100 150 0 0 Moderate hypotension 0 0 0 Flushing 0 Mild nausea 0 Moderate hypotension 0 0 Severe hypotension 0 Mild nausea 0 0 Moderate hypotension Mild hypotension, flushing Moderate hypotension 38 TABLE IX (continued) ~osage mg/kg/day) Duration (days) Total dose (mg/kg) Toxic effects Thrice-weekly Administration 21 14 12 20 0.25 0.30 0.31 0.41 0 150 44 0.12 0 50-100 46 31 1 .5 0.65 50-150 27 1 .0 0 50-250 45 18 1 .57 1 .25 0 0 50 0 0 0 Moderate hypotension 0 39 and the other had stomach cancer. The investigators concluded that with this method of administering tubercidin the problems of severe local reactions, such as venous thrombosis and tissue necrosis, seen with direct injection are eliminated. Drug toxicity in this study was infrequent, usually mild, and reversible. Initial toxicity studies using sangivamycin were reported 61 in a study involving 47 patients with various malignant diseases. Sangivamycin was administered daily, weekly, or thrice weekly by slow injection. A summary of the toxic effects is shown in Table IX. The side effects consisted primarily of hypotension (defined as falls of 20 mm/Hg or more systolic or 10 or more diastolic) and flushing at the time of injection. These effects were related to the rate of injection as well as the size of the dose and could be significantly reduced or prevented by prolongation of the infusion from 0.5 hours to one hour. Although no antitumor effects were noted, the investigators recommended a Phase II trial of sangivamycin be undertaken to evaluate the antitumor effects. In summary, Phase I toxicity studies have shown that toyocamycin and tubercidin elicit severe local reactions which in the case of toyocamycin prevented evaluation of the drug. Tubercidin can be administered in doses up to 1500 ~g/kg with only minor adverse effects. San- givamycin shows only minor toxic reactions and has been recommended for a Phase II study. CHAPTER III 4,5-DISUBSTITUTED PYRROLOPYRIMIDINES 6-Amino-4-methyl-8-(a-D-ribofuranosyl) (4H,8H)pyrro10 [4,3,2-de] pyrimido[4,5-c]pyridazine. As part of an investigation on the reactions of 4-ch1oro-5-cyano-7-(a-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine (XXVI) with various nucleophiles,62 XXVI was reacted with methylhydrazine in ethanol at reflux temperature. This reaction was expected to furnish 5-cyano-4-(N-l-methy1hydrazino)-7-(a-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine (XXVII). However, the absence of a peak in the 2200 cm- 1 region of the infrared spectrum excluded this possibility and suggested that in this instance the initial attack had occurred at the cyano group rather than at the 4-ch1oro. This prompted a more detailed study to determine the actual structure of the nucleoside. A molar equivalent of methy1hydrazine was added to a solution of XXVI in absolute ethanol and the mixture was heated at reflux temperature. A yellow solid began to separate from solution after four hours but heating was continued for a total of 12 hours. The reaction mixture was then cooled at 5° for six hours and the solid which had separated from solution was collected by filtration. The solid was air dried and thin layer chromatography revealed a highly fluorescent blue spot with an Rf of 0.3 (XXVI, 0.75). The infrared spectrum showed the absence of any peak in the 2200 cm- 1 region which indicated that a modification of the cyano group had occurred. A pmr spectrum revealed the presence of an N-methy1 group at 03.59 as a 3 proton singlet, a pattern of peaks 41 in the 03.S-S.0 region which definitely established the yellow powder as a nucleoside material, and a definite upfield chemical shift of the H6 proton from 08.38 to 07.S2 which is indicative of a modification of the cyano group in the 5-position of the pyrrolo[2,3-d]pyrimidine nucleosides. 63 That nucleophilic attack by methy1hydrazine had not occurred exclusively on the cyano group to furnish 4-chloro-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-S-N-methylcarboxamidrazone was established by elemental analysis. The product analyzed for C13H16N604·3H20 and the presence of three moles of water was corroborated by pmr spectroscopy. The addition of deuterium oxide to the sample tube furnished a spectrum which established the presence of S exchangeable protons in addition to the 6 protons for the water. Attempts to remove the water from the product by lyophilization and drying in vacuo above 100 0 were unsuccessful since on exposure to air the compound immediately rehydrated. The possibility of a direct displacement of the 4-chloro group by methylhydrazine and attack at the cyano group by another mole of methylhydrazine to furnish 4-N-methylhydrazino-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-S-N-methylcarboxamidrazone was also eliminated on the basis of the elemental analysis. Based on the above information the structure of the product was initially assigned as 6-amino-4-methyl8-(s-D-ribofuranosyl)(4H,8H)pyrrolo[4,3,2-de]pyrimido[4,S-c]pyridazine (XXVIII). This is the first report of any derivatives of this new heterocyclic ring system. The equivocal assignment of the above structure emphasized the need for elucidation of the actual pathway involved in the formation of the final product. There were two major possibilities envisioned for the formation of this tricyclic nucleoside: nucleophilic displacement of the 4-chloro 42 group of XXVI with subsequent ring annulation or initial attack at the S-cyano group to furnish a N-methylcarboxamidrazone derivative which was then followed by nucleophilic displacement of the 4-chloro group. This prompted the isolation and characterization of the intermediate in order to establish which of these two possibilities was actually occurring in this instance. A solution of XXVI in absolute ethanol was treated with a molar equivalent of methylhydrazine and stirred at room temperature and a white crystalline precipitate separated from solution after 2-3 minutes. The crystalline product was collected by filtration, recrystallized from water, and air dried. The infrared spectrum showed a strong absorption band at 2210 cm- 1 which indicated that the initial reaction must be a displacement of the 4-ch10ro group. A pmr spectrum exhibited absorbances at 08.4 (H2) and 08.23 (H6) for the aromatic protons. The anomeric proton was centered at 06.9 and other carbohydrate protons were shown in their normal positions. Superimposed on the sugar protons at oS.l was a sharp absorbance attributed to the NH2 of the hydrazino group which disappeared upon the addition of D20. At 03.38 was a sharp singlet which integrated for three protons and was assigned to the N-1 methyl group. Elemental analysis for this nucleoside was also consistent with a simple nucleophilic displacement of the 4-ch10ro group by methyl hydrazine. However, a displacement of the 4-ch10ro group by methylhydrazine could have occurred with only the secondary amine, only the pr-imary amine, or both ways. Thin layer chromatography established the presence of only one compound. If the nucleoside S-cyano-4-(2-N-methylhydrazino)7-(e-D-ribofuranosy1}pyrrolo[2,3-d]pyrimidine (XXIX) had been formed by the displacement of the 4-chloro group with the primary amine of methyl- 43 hydrazine, then the pmr spectrum should show a doublet for the methyl group. A peak in the pmr spectrum at 03.38 (3 protons) for the methyl group was observed as a singlet and on this basis the structure XXIX could be tentatively eliminated. On the basis of the above data the intermediate was assigned the structure 5-cyano-4-(N-l-methy1hydrazino)7-(e-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine (XXVII). However, this structure assignment was rather tenuous and prompted a chemical structure proof for additional corroboration. A facile reduction of heterocyclic hydrazino groups to heterocyclic amines using Raney nickel as a catalyst has been reported. 64 If the intermediate nucleoside were XXVII then reduction would afford 5cyano-4-methylamino-7-(a-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine (XXX). However, if the internlediate nucleoside were XXIX, then the reduction product would be toyocamycin (I). A 10 fold excess (w/w) of W-7 Raney nickel was added to a ref1uxing ethano1ic solution of the intermediate over a four hour period. After addition of the final portion of Raney nickel the reaction mixture was heated at reflux temperature for 12 hours. The hot reaction mixture was filtered through a ce1ite bed and the filter cake was washed with hot water. The solvents were removed under reduced pressure and the white powdery residue was recrystallized from water to yield colorless crystals. An infrared spectrum of the product revealed the presence of a cyano group (2210 cm- I ) and a mixture melting point with an authentic samp1e 62 of 4-methy1amino-5-cyano7-(e-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine (XXX) gave no depression. Therefore, the structure of the intermediate was 5-cyano-4-(N-1-methYlhydrazino)-7-(B-D-r;bofuranosyl)pyrro1o[2,3-d]py~imidine (XXVII). A solution of XXVII in water was heated at reflux temperature 44 REACTION SCHEME 6 > XXVII XXVI \ XXVIII XXX 45 for 16 hours and then allowed to stand at 5° for 12 hours to afford fine yellow crystals of XXVIII. An infrared spectrum of the product obtained from XXVII showed the absence of a cyano group at 2210 cm- 1 and the spectrum was superimposable on a spectrum of XXVIII obtained directly from XXVI. Therefore, the results of this study have established that the initial attack occurs by the methylamino moiety of methylhydrazine on the 4-chloro group of 4-chloro-5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVI) to yield the intermediate 5-cyano-4-(N-Jmethylhydrazino)-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVII). This intermediate was then cyclized in situ to form the tricyclic nucleoside 6-amino-4-methyl-8-(S-D-ribofuranosyl)(4H,8H)pyrrolo[4,3,2-de]pyrimido[4,5-c]pyridazine (XXVIII) as the first derivative of this new r"j ng system. Heterocyclic rings at position ~ Testing data from DR&D have shown that various 5-substituted derivatives of the naturally occurring pyrrolopyrimidine nucleosides possess significant antileukemic activity against L12l0 in mice. Especially active are the compounds with an amino group in the 4-position and a modified cyano group in the 5position, i.e., derivatives of sangivamycin. Therefore, it was of interest to synthesize some heterocyclic rings in the 5-position to determine the effect of steric bulk and annulation on antileukemic activity. A secondary aim of this specific study was to prepare derivatives which possessed the common receptor-complement feature proposed by Zee-Cheng and Cheng. 55 It was of interest to determine if activity remained when the environment of the lone pair of electrons on the nitrogen involved in the triangulation was changed from a primary amine, to a secondary amine, to a tertiary amine, and, finally, to an aromatic system. There- 46 fore, a series of 5- and 6-membered, aromatic and non-aromatic rings were prepared at the 5-position of the 4-aminopyrrolopyrimidine nucleoside system. Various modifications of the cyano group of toyocamycin had previously been accomplished in our laboratory. Treatment of toyocamycin with anhydrous hydrazine in ethanol afforded a good yield of 4-amino7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV).16 A recent review on the chemistry of amidrazones 65 has shown that this nitrile derivative is a versatile compound for the formation of certain heterocycles. When XIV was reacted with ninhydrin in methanol, an 81% yield of 4-amino-5-(9-oxoindeno[1 ,2-eJ-as-triazin-3-yl}-7-(S-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine (XXXI) was obtained as a yellow precipitate. The pmr showed absorbances at 08.65 and 08.01 assignable to the C2 and C6 protons, respectively,33 a broad multiplet centered at 07.83 which integrated for the 4 protons of the phenyl ring, and peaks in the 03.5 to 05 region for the carbohydrate moiety. The mode of condensation of ninhydrin with amidrazones has been shown 66 to occur by initial attack of N-1 of the amidrazone (which is most basic) with the 2-position of ninhydrin to produce the 9-oxoindeno[1,2-e]triaz~ne and not the isomeric 5-oxoindeno[2,1-e]triazine. Using similar reaction conditions, XIV was condensed with various diketones to afiQrd other as-triazinyl derivatives. With glyoxal trimer in water at reflux temperature, XIV produced 4-amino-5-( s-triazin-3-yl}~7-(B~D-ribofuranosyl}pyrrblo[2,3-d]- pyrimidine (XXXII). Pmr showed two sets of doublets centered at 09.21 and 08.85 (J = 2.5 Hz) for the C6 and C5 protons,67 two singlets at 1 1 08.53 and 08.25 (C2 and C6 protons), and peaks for carbohydrate protons between 03.5 and 05. Similar treatment of XIV with biacety1 and benzil 47 afforded good yields of the 5,6-dimethy1-as-triazine derivative (XXXIII) and the 5,6-dipheny1-as-triazine derivative (XXXIV). This series of compounds was synthesized primarily to determine the amount of steric hindrance which is permitted in the 5-position before activity is lost. Thus, the series progresses from the moderately bulky as-triaziny1 group, through the increasingly larger dimethyl and dipheny1 derivatives, to the very large ninhydrin product. CPK models show that the unshared electron pairs of the triaziny1 nitrogens are located at approximately the same distance from the 21 and 3' positions of the carbohydrate as in the amino group of the carboxamide in sangivamycin. However, in contrast to the lone pair of electrons in the amide amino group of sangivamycin which have free rotation, the electron pairs in the as-triaziny1 derivatives have a fixed orientation, because rotation is restricted by the 4-amino group. The 5-membered aromatic heterocycles were synthesized to determine the effect which a decrease in ring size would have on antitumor activity. Ring closure of XIV with formic acid at reflux temperature afforded 4-amino-5-(1 ,2,4-triazol-3-y1)-7-(s-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine (XXXV). A pmr spectrum established that ring closure had occurred due to the appearance of a aromatic proton absorbance at 88.52 in addition to the C2 and C6 protons of the parent ring system. When XIV was heated in ethanol at reflux temperature with carbon disulfide and potassium hydroxide, 4-amino-5-(1,2,4-triazol-3-yl-5-thione)7-(S-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine (XXXVI) was obtained. The pmr spectrum showed a broad absorbance centered at 813.73 which integrated for two protons. When D20 was added to the pmr sample tube, the absorbance at 813.73 disappeared. This established that the product REACTION SCHEME 7 48 I R XXXI NNH2 , II XXXII, R=H XXXIII, R=CH3 XXXIV, R=C6Hs -N H 2 / XIV H S H XXXV HO~~ R=~ I XXXVI R 49 exists in the thione rather than thio1 fornl. The infrared spectrum also possessed a strong peak at 1220 cm- 1 which could be attributed 68 to the thione function. Elemental analysis excluded the possibility that the product was the 5-[{5-mercapto)-1,3,4-thiad;azol-2-yl] compound. 69 CPK models of the triazoly1 compounds show that the ring is smaller than the as-triazinyl ring and, therefore, some indication as to the effect of ri ng size on act; vi ty may be ga.i ned from these deri vat i ves. The models also show that the electron pairs of the nitrogens should confornl to the dimensions required by the triangulation theory. To determi ne if the k"j nd of heteroatom has any effect on activity, certain 5-membered heteroaromatics containing Sand 0 in the ring were prepared. Treatment of 4-amino-7-{S-D-ribofuranosyl)pyrro10[2,3-d]pyrinlidine-5-carboxamidoxime (XV) in refluxing water with formicacetic mixed anhydride for 20 hours yielded 4-amino-5-{1,2,4-oxadiazol-3yl)-7-{S-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine (XXXVII). Product formation was evidenced by the appearance of an aromatic proton at 69.68 in addition to the two aromatic protons of the parent system. Comparison of the activity of the XV with the oxadiazoly1 derivative XXXVII should give a very clear indication as to the effect of cyclization and aromatization on activity. Along the same lines, cyclization of the active 4-amino-7-{S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-thiocarboxamide {XVII),16 using phenacyl bromide and glacial acetic acid in water at reflux temperature, afforded 4-amino-5-[{4-phenyl)thiazol-2-yl]7-{s-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine (XXXVIII). The pmr spectrum showed a singlet at 68.47 for the H5 proton and two complex 1 multip1ets due to the phenyl ring at 67.6 and 68.0. The product also gave a negative Bei1stein test 70 which indicated that ring closure had 50 REACTION SCHEME 8 NOH II C-NH2 , xv XXXVII . 5 , C-NH 2 XVII XXXVIII 51 occurred. Comparison of the testing data of the thiosangivamycinthiazolyl derivative should parallel the results of the carboxamidoximeoxadiazolyl system in indicating the effect of ring closure and aromatization of activity. The preparation of certain exocyclic 5- and 6-membered aromatic heterocycles at C5 had now been achieved, and the preparation of 5- and 6-membered non-aromatic heterocycles was undertaken. It was decided to prepare cyclized derivatives of another active compound, 4-amino-7-(aD-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidine. Addition of diamines to the cyano group of 2-cyanobenzimidazole has been shown to occur 71 with the formation of non-aromatic heterocycles. Treatment of toyocamycin with 1,2-diaminoethane at reflux temperature resulted in an addition to the cyano group to form 4-amino-5-(4H,5H-imidazolin-2-yl)7-(a-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXXIX). The infrared spectrum showed no absorbance in the 2210 cm- 1 region which indicated that addition had occurred. Absorbances integrating for 4 protons were observed in the 03.7 region of the pmr spectrum for the methylene protons. During the reaction, ammonia was evolved and elemental analysis confirmed that ring closure had occurred. Similar treatment of toyocamycin with 1 ,3-diaminopropane resulted in the formation of a six membered analog 4-amino-5-(1 ,4,5,6-tetrahydropyrimidin-2-yl)-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XL). The infrared spectrum showed no cyano absorbance and the pmr spectrum showed a complex pattern of peaks in the 03.7 region and an additional multiplet centered at 01.8 assignable to the 51 methylene protons. In both of these compounds the nitrogen atoms should fit the parameters of the triangulation as established for sangivamycin. Of special interest is the fact that a correlation may 52 also be made as to the effect of cyclization and ring size on activity. The synthesis of a compound was then initiated to deternline the effect of a primary amine at position 5 on the antitumor activity. Reduction of desaminotoyocamycin to the 5-aminomethylene derivative (which was inactive in vivo), has been reported,49 but no mention was made of the similar reduction of toyocamycin. Since the imidazolinyl (XXXIX) and the tetrahydropyrimidinyl (XL) compounds can be considered secondary aliphatic amines, the synthesis of 4-amino-5-aminomethylene-7-(S-D-ribofuranosyl)pyrrolo[2,3-dJpyrimidine (XLI) was undertaken. The use of palladium on charcoal and platinum oxide (Adam's catalyst)87 in ethanol at atmospheric or 40 psi in a Paar hydrogenator were unsuccessful and only starting material was recovered. When the reduction was run in IN hydrochloric acid using 10% palladium on charcoal at 40 psi of hydrogen4, a reduction was observed and a 52% yield of XLI was obtained. The infrared spectrum showed no cyano absorbance at 2210 cm- I . The pmr spectrum showed a sharp singlet at 02.5 (superimposed on the DMSO-d6 peak) which was assigned to the methylene group. The amino group at C4 was a broad peak centered at 08.7. The absorbances for the aliphatic amino group was obscured by the peaks due to the carbohydrate protons. Elemental analysis also confirmed that reduction had occurred. The triangulation theory does not indicate what valence state the nitrogen must be in to fulfill the con~on receptor-complement feature of the active antileukemic compounds. Possibly no one valence state is preferred over another as long as the unshared pair is at the required distance. However, since an electron pair from the nitrogen is required for activity (according to the model), it seems likely that 53 REACTION SCHEME 9 I H · 2 Her XXXIX XLI XL 54 an increase in availability of the electron pair will increase the activity of the compound. The preparation of these 5-heterocyclic substituted pyrrolopyrimidine nucleosides should give some indication as to the effect of ring size and valence state of the nitrogen on antileukemic activity. CHAPTER IV 5-SUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES Synthesis and reactivity of 5-cyano-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine. Previous studies on the chemistry and reactivity of the pyrrolopyrimidine nucleosides have been concerned with the effect on the reactivity of the cyano group by modification of the substituent at C4. It was of interest to prepare a derivative where the group at C4 would not be reactive, leaving the cyano sUbstituent as the only reactive function. One such group at C4 would be a hydrogen. Therefore, the synthesis of 5-cyano-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine was undertaken to study the reactivity of the cyano group without any competing reactions. It was also of interest to determine if an amino group was essential for antileukemic activity since this group is not involved in the triangulation theory.55 Dechlorination of 4-chloro-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine had been previously accomplished in our laboratory.22 Treatment of 4-chloro-5-cyano-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVI) with 5% palladium on charcoal under a hydrogen atmosphere effected a removal of the 4-chloro group without a concomitant reduction of the 5-cyano moiety to produce the desired 5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII). That dechlorination had occurred was evidenced by the appearance of a peak for the C4 aromatic proton at 88.9. The H2 and H6 proton absorbances were observed at 89.2 and 88.2 and peaks for the carbohydrate protons between 83.5 and 85. The relative 56 reactivity of the cyano group of XLII was then compared to that of the cyano group of the 4-amino (I) and the 4-keto (XX) analogs. It was postulated that a removal of the 4-amino group would have two major effects. First, the electron density at the exocyc1ic carbon of the cyano group should be decreased by a removal of the electron releasing amino group. This should facilitate nucleophilic attack at the cyano group. Secondly, additions to the cyano group of toyocamycin are impeded in acid solution due to protonation of the pyrimidine ring. I6 Removal of the amino group should decrease the electron density at N-3 which would decrease the protonation of the pyrimidine moiety. A greater ease of addition to the cyano group of XLII under acid conditions was therefore envisioned. Removal of the keto group should, likewise, increase the reactivity of the cyano group of XLII over that of XX under basic conditions. In summary, the reactivity of the cyano group of XLII was expected to be" greater than that of the cyano group of either I or XX under acidic or basic conditions. Treatment of XLII with a solution of dilute aqueous ammonium hydroxide on a steam bath for one hour afforded, after cooling at 50 for 12 hours, a 75% yield of". 7-(s-D-ribofuranosy1 )pyrro1o[2,3-d]pyrimidine5-carboxamide (XLIII). Evidence that the product had formed was the disappearance of the cyano absorption in the infrared spectrum and a correct elemental analysis. Under similar conditions toyocamycin was found to be unchanged. Addition of hydroxylamine to the cyano group of XLII occurred after heating at reflux temperature in ethanol for 0.5 hours to give 7-(S-D-ribofuranosy1)pyrro10[2,3-d]pyrimidine-5carboxamidoxime (XLIV) in 79% yield. The infrared spectrum showed no absorbance due to a cyano function. A pmr spectrum showed three aromatic 57 proton absorbances at 09.4 (H2), 08.9 (H4), and 08.3 (H6), the carbohydrate protons (03.5-05), a peak at 09.65 as a singlet which integrated for one proton (=NOH), and a broad peak which integrated for two protons at 05.9 (NH2). The peaks at 09.65 and 05.9 disappeared upon the addition of 020. Elemental analysis also confirmed that addition had occurred to the cyano group. The reaction of I under the same conditions requires approximately 4 hours at reflux temperature before the appearance of a product. The thiosangivamycin analog 7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-thiocarboxamide (XLV) was obtained in 57% yield by treatment of XLII with a solution of sodium methoxide in nlethanol saturated with hydrogen sulfide. Spectral confirmation that addition to the cyano group had taken place was provided by the absence of a peak in the infrared spectrum at 2210 cm- 1 and absorbance above 310 nm in the ultraviolet spectrum. Milder conditions using pyridine, triethylamine, and hydrogen sulfide produced a compound whose uv and ir spectra were identical to those of XLV. The above series of reactions indicated that a removal of the 4-amino group did enhance the reactivity of the cyano group of XLII over that of I and XX toward nucleophilic attack under basic conditions. The increased reactivity of the cyano function of XLII under acidic conditions was not as evident. Reaction of XLII with hydrogen chloride in ethanol presumably resulted in the formation of ethyl 7(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-formimidate (XLVI), but attempts to isolate the product were not successful. Thin layer chromatography indicated that the imidate hydrochloride formed (Rf, 0), but was immediately hydrolyzed by water to the amide XLIII or underwent a 58 Reaction Scheme 10 NOH 11 C-N~ ~ ~ XLIII XLVI \ / XLIV' XLV 59 Pinner cleavage 72 to form the amide. Attempts to lower the temperature of the reaction to prevent amide formation resulted in no reaction below 0°. Therefore, although no definite conclusions can be drawn as to an increase in reactivity of XLII under acid conditions, the rapid formation (less than 2 minutes) of a slow uv absorbing spot on thin layer chromatography followed by the appearance of a uv absorbing spot with an Rf equal to that of the amide, indicate that the cyano group of XLII may be more reactive than the cyano group of XX. Substitution at position ~ The synthesis of the 5-halogen deriv- atives of tubercidin 22 produced compounds which showed essentially no activity against leukemia L12l0. However, biological tests performed by Dr. E. Reich at Rockefeller University demonstrated that 5-halotubercid;n analogs showed a specificity of action. This was manifested in a striking manner by 5-bromotubercidin which stimulated certain viral growth while simultaneously inhibiting the growth of host cells. 73 Therefore, it was of interest to synthesize the 5-halopyrrolopyrimidine nucleosides to determine their effect on viral systems and for testing against leukemia L12l0. Also, although halogenation has been shown to occur at the 5-position of 4-substituted pyrrolopyrimidines and in the 6-position of 4,5-disubstituted pyrrolopyrimidines, no studies to date have been reported on the substitution reactions of the unsubstituted pyrrolopyrimidine nucleoside system. Preparation of the required starting material, 7-(2,3,5-triQ-acetyl-S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLVII), was accomplished by two different routes. Acetylation of 7-deazanebularine 22 (XLVIII) using acetic anhydride in pyridine produced the blocked nucleoside XLVII. Alternatively 4-chloro-7-(2,3,5-tri-O-acetyl-S-D-ribo- 60 REACTION SCHEME 11 / VII AcO \ XLVII ! XLVIII OAc OAc J Br CI XLIX L LI 61 furanosy1)pyrro10[2,3-d]pyrimidine (VII) was dechlorinated with 5% palladium on charcoal to give XLVII in good yield. Halogenation reactions were then studied using XLVII as starting material. Chlorination proceeded smoothly in methylene chloride using N-ch10rosuccinimide. After stirring at room temperature for 16 hours the reaction mixture was applied to a silica gel dry co1umn.74 Fractions containing the product were combined and the solvent removed under reduced pressure. The amber oil was then deacety1ated using methano1ic ammonia to afford 5-ch10ro-7-(s-D-ribofuranosy1)pyrro10[2,3-d]pyrimidine (XLIX). That chlorination had occurred was evidenced by the pmr spectrum. 7-Deazanebu1arine (XLVIII) exhibits two sets of doublets centered at 06.B and oB.O assignable to the H5 and H6 protons respective1y.22 The pmr spectrum of XLIX showed 3 aromatic protons as singlets at 09.07(H2), oB.95(H4), and oB.03(H6). Elemental analysis was also consistent with monochlorination. Similar treatment of XLVII with N-bromosuccinimide and iodine monoch10ride furnished the 5-bromo (l) and 5-iodo-7-(a-Dribofuranosy1)pyrro10[2,3-d]pyrimidine (lI) derivatives. Thus, it has been shown that the cyano group of 5-cyano-7(S-D-ribofuranosy1)pyrro10[2,3-d]pyrimidine (XLII) is more reactive under basic conditions than the nitrile function of the compounds with an amino group (I) or a keto group (XX) at position 4. No definite conclusions were reached as to the reactivity of XLII under acid conditions. Halogenation has also been shown to occur in the 5-position of 7-deazanebu1arine. CHAPTER V 5,6-0ISUBSITIUTEO PYRROLOPYRIMIOINE NUCLEOSIOES Two important findings led to the investigation of the 5,6disubstituted pyrrolopyrimidine nucleosides. The first was that antileukemic activity was not dependent on the presence of an amino group in the 4-position (vide supra). Testing data from OR&O had also indicated that significant activity was associated with the 5-substituted derivatives (see Chapters LV and VII). The 5-substituted pyrrolo- pyrimidine nucleosides also appear to be less toxic than the derivatives with an amino group in the 4-position. The second finding was that substitution in the 6-position of the pyrrolopyrimidine nucleosides caused a marked decrease in toxicity with only a minor decrease in antileukemic activity_ Therefore, it was decided to investigate the synthe- sis of derivatives which possessed a cyano group at position 5 with another substituent at the 6-position. 5-Cyano-7-(s-O-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII) underwent electrophilic attack at position 6 when treated with bromine water to afford 6-bromo-5-cyano-7-(S-O-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (LII). Pmr spectroscopy confirmed that electrophilic attack had occurred because only two aromatic peaks (89.05 and 89.30) were evident. Based on previous studies these absorbances could be assigned to the C4 and C2 protons respectively. An infrared spectrum indicated that no modification of the cyano group had occurred and elemental 63 analysis supported the proposed structure. Displacement of the 6-bromo group by various nucleophiles was then investigated. Reaction of LII with hydrazine in ethanol at reflux temperature effected displacement of the 6-bromo group to produce 5-cyano-6-hydrazino7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (LIII). The infrared spectrum of LIII showed a strong cyano absorbance in the 2200 cm- 1 region and excluded the possibility of either addition to the cyano group or cyclization to the pyrazolopyrrolopyrimid"ine LIV. The pmr spectrum showed peaks for two aromatic protons (H2 and H4) as singlets at 08.6 and 08.57, a peak for the anomeric proton (H11) as a doublet at 06.35, and peaks for the sugar protons between 03.4 and 05.6. Superimposed on the lower portion of the carbohydrate absorbances was a large peak at 05.68 (3H) which disappeared upon the addition of 020 and was assigned to the hydrazino moiety. Nucleophilic attack on the cyano group by the hydrazino moiety was effected in water at reflux temperature to produce the tricyclic pyrazolo [3 1,4 1 :5,4]pyrrolo[2,3-d]pyrimidine LIV which was isolated as the hydrochloride salt. The infrared spectrum confirmed that cyclization had occurred due to the absence of a peak in the 2200 cm- 1 region. A pmr spectrum corroborated the fact that ring closure had been effected since the absorbance at 05.7 assigned to the hydrazino group disappeared and was r'epl aced by a broad absorbance in the 08.7-07 region which integrated for 5 protons and disappeared when 020 was added. This absorbance was assigned to the amino group, the pyrazole proton, the Hel salt, and 1/2 mole of water. The aromatic protons were shown as sharp singlets at 09.45 (H2) and 09.25 (H4). Displacement of a bromo group in the 6-position of the pyrrolopyrimidine nucleosides by ammonia had previously been accomplished using 64 methanolic ammonia in a sealed reaction vessel to produce 6-aminotoyocamycin derivatives. Using the reaction conditions described,54 LII was treated with methanolic a.mmonia at 110-120° for 12 hours to afford a 33% yield of 6-amino-5-cyano-7-(a-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (LV). Thin layer chromatography of the reaction mixture, however, showed a second major uv absorbing spot which moved slower than LV. A better yield of LV was obtained with the use of liquid ammonia in a sealed reaction vessel at the same temperature and for the same reaction time. Use of this second pracedure afforded LV in 74% yield without any evidence of another compound being formed. The infrared spectrum of LV showed a strong band (eN) at approximately 2210 cm- l . The pmr spectrum evidenced 2 aromatic proton absorbances at 09.0 (H2) and 08.6 (H4), a broad peak centered at 07.9 which integrated for 2 protons (NH2) and disappeared upon the addition of 020, the anomeric proton centered at 66.35, and the remaining protons of the sugar moiety between 65.4 and 03.3. Additiona.l evidence that nucleophilic displacement had occurred was that the uv spectrum showed a large bathochromic shift· This has been previously shown to occur with 6aminopyrrolopyrimidine nucleosides. 54 Modifications were then performed on the cyano group. It has been previously shown during the studies on the reactivity of the cyano group of 5-cyano-7-(a-0-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII), toyocamycin (I), and 5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidin-4-one (XX), that an amino group or keto in the 4position has a deactivating effect toward nucleophilic addition to the cyano group under both basic and acidic conditions. Although the amino group in the pyrimidine ring influences the susceptibility of the cyano 65 group in the pyrrole ring toward nucleophilic attack by increasing the electron density at the cyano group, the electronic effects are probably more pronounced in the pyrimidine ring since the rinqs are~somewhat compartmentalized. 75 Therefore, it was expected that an amino group located in the pyrrole ring would have a greater deactivating effect on the cyano group toward nucleophilic attack than an amino group in the pyrimidine ring. Oxidation of the cyano group of LV with peroxide in base resulted in the formation of the sang;vamycin isomer 6-amino-7-(S-0ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamide (LVI). Thin layer chromatography of the reaction mixture showed the presence of at least two minor, highly fluroescent compounds. The low yield of product (39%) compared to the similar oxidation of toyocamycin to sangivamycin (65%) indicated that some side reactions were occurring. Since the amino group in the 6-position resides on an electron rich ring system,75 it is possible that N-oxidation was occurring since it has been shown that peroxide and base are capable of oxidizing pyridine to pyridine-l-Noxide and aniline to azobenzene. 76 ,77 However, that some conversion of the cyano group to the amide had occurred was evidenced by the absence of any absorbance in the 2200 cm- 1 region of the ir spectrum. A pmr spectrum exhibited two aromatic proton absorbances at 09.0 (H2) and 08.6 (H4), a broad 2 proton absorbance at 07.95 (NH2), a sharp 2 proton peak at 06.9 (CONH2), the anomeric proton absorbance centered at 06.4, and the other sugar absorbances between 05.3 and 03.6. The absorbances at 07.95 and 06.9 disappeared when 020 was added. The reduced reactivity of LV as compared to I is clearly shown in the synthesis of the thioamide derivative. Treatment of LV with 66 REACTION SCHEME 12 H LII LIII ~ LIV =N , 01 , NH2 N I R LV '\ II NH2 LVI NH2 H2 LVII 67 pyridine, triethylamine, and hydrogen sulfide for 18 hours produced no reaction. Under the same conditions, I gave a 43% yield of the 5-thioamide XVII. A recent report 78 ,79 states that thioamides can be prepared from nitriles which are not reactive to pyridine, triethylamine, and hydrogen sulfide by use of dimethylforamide, dimethyl amine, and hydrogen sulfide. Treatment of LV using reaction conditions identical to those reported effected a conversion of LV to 6-amino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-thiocarboxamide (LVII) in very low yield «10%). The infrared spectrum showed no peak in the 2200 cm- 1 region. A pmr spectrum displayed absorbances at 89.1 and 88.7 for H2 and H4 respectively, a peak which integrated for 2 protons at 89.35 (CSNH2), a peak at 88.35 (6-NH2) which integrated for 2 protons, a peak at 86.49 (Hl ' ), and absorbances for the remaining carbohydrate protons between 83.5 and 85. A higher yield of LVII (21%) was obtained when LV was treated with sodium hydrosulfide in methanol saturated with hydrogen sulfide in a sealed reaction vessel for 12 hours at 120-130°. Since the nitrile group of LV showed a reduced tendency toward nucleophilic attack under basic conditions, it would be expected that reactions carried out under acid conditions should be extremely slow. The attempted conversion of LV to the imidate derivative using hydrogen chloride in ethanol was unsuccessful even under prolonged reaction times. Only starting material was isolated with no evidence (ir) of any reaction. The above findings indicate that transposition of the amino group of position 4 to the 6-position has a definite deactivating effect on the cyano group toward nucleophilic attack, with more stringent conditions being required to modify the nitrile function. CHAPTER VI 4,5,6-TRISUBSTITUTED PYRROLOPYRIMIDINE NUCLEOSIDES The finding that substitution in the six position tends to reduce the toxicity of pyrro1opyrimidine nucleosides prompted the synthesis of certain 4,5,6-trisubstituted derivatives. By retaining the amino group at position 4 and adding various groups at the 6-position, it is postulated that the antitumor activity associated with sangivamycin and thiosangivamycin (XVII) would be retained but with reduced toxicity similar to that observed for 6-bromosangivamycin. Modifications of the cyano group at the 5-position were expected to be more difficult than similar reactions carried out on the 6amino-5-cyano, 5-cyano- and 4-amino-5-cyano derivatives because of the increased electron density at the cyano group due to the presence of two electron releasing groups. Preparation of 4,6-dianlino-5-cyano-7-(S-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine (LVIII) from 6-bromotoyocamyc;n (XII) using the same procedure which was used to prepare LV54, produced a mixture, even when run at high temperature in a sealed vessel. The use of liquid ammonia in a sealed reaction vessel at 110-120° for 12 hours, however, effected a displacement of the 6-bromo group to afford 6-aminotoyocamycin (LVIII) in 74% yield. That nucleophilic displacement had occurred was evidenced by the appearance of two broad absorption peaks in the pmr centered at 07.34 and 06.18 (2 protons each) and disappeared upon the addition of D20. The pmr spectrum also displayed peaks for the aromatic proton at C2 (08.1), the anomeric proton at 06.24, and the other carbo- 69 hydrate protons (03 to 65). The peak at 06.18, which partially obscured the anomeric proton in DMSO-~6, was assigned to the amino group at posi- tion 6. An infrared spectrum indicated the presence of a strong band (nitrile) at 2210 cm- 1 which precluded any reaction at the cyano group during the reaction. The ultraviolet spectrum was very similar to that obtained for 6-amino-4 -thiotoyocamycin. 54 Modifications of the cyano 1 group to produce the potentially more active derivatives were then attempted. Oxidation of the nitrile to the carboxamide using hydrogen peroxide and base produced a mixture of at least five compounds as indicated by thin layer chromatography. The incidence of side reactions was greater than that seen with 6-amino-5-cyano-7-(s-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine (LV), probably due to the increased electron density at the 6-amino group. Therefore, alternative methods of converting the nitrile to a carboxamide were investigated. Po1yphosphoric acid has been recommended for the conversion of unhindered nitri1es to amides. 8o For instance, when benzonitri1e is heated at 110 0 for 1 hour in po1yphosphoric acid, a 96% yield of benzamide is obtained. Although the reaction conditions may be considered somewhat extreme for nuc1eosides, it was assumed that the more stable glycosidic bond of the pyrrolopyrimidines (as compared to the purine ribosides) would withstand this treatment. However, reaction of LVIII under the above conditions resulted in a dark resinous material which was composed of several compounds as shown by thin layer chromatography. The use of ion exchange resins has also been reported to form carboxamides from nitri1es under relatively mild conditions and with no 70 hydrolysis to the carboxylic acids being observed. 81 ,82,83 When LVIII was heated at reflux with IRA-400(OH) resin in water for 20 hours and the reaction mixture was cooled to room temperature (after filtering), white crystals of starting material were obtained. The infrared spectrum showed a strong absorbance for a nitrile group. An attempt was then made to use acid conditions to hydrolyze the cyano group of LVIII. Hydrolysis using water or heating of the imidoester hydrochloride formed from nitriles when treated with hydrogen chloride in an alcohol has been recommended for the synthesis of a-hydroxy84 and a-amino 85 amides. However, treatment of LVIII with hydrogen chloride in ethanol for 1 hour, evaporation of the ethanol, and boiling of the residue in water afforded only starting material. Therefore, the hydrolysis or oxidation of the cyano group of LVIII was abandoned. An alternate route to 6-aminosangivamycin (LIX) was initiated which would involve a nucleophilic displacement of the 6-bromo group of 6-brol11osangivamycin (LX) using liquid ammonia in a sealed reaction vessel. When 6-bromosangivamycin was reacted under the conditions used for the synthesis of 6-aminot.oyocamycin (LVIII), a product was obtained which showed a bathochromic shift in the ultraviolet spectrum and indicated that nucleophilic displacement of the 6-bromo group by ammonia had occurred to produce 6-aminosangivamyc;n (LIX). A pmr spectrum showed a single aromatic proton absorbance at 09.3 (H2), three peaks which integrated for 2 protons each at 07.4 (4NH 2), 07.03 (CONH 2), and 06.37 (6NH2) which disappeared when D2 0 was added, the anomeric proton as a doublet centered at 06.3, and the other carbohydrate proton absorbances in the usual position. The reaction of LVIII with hydroxylamine in ethanol was expected 71 to yield 4,6-diamino-7-(a-0-ribofuranosy1)pyrrolo[2,3-d]pyrimidine-5carboxamidoxime (LXI). The infrared spectrum confirmed that a modification of the cyano group had occurred as the strong band (nitrile) in the 2200 cm- 1 region had disappeared. However, analysis of the product indicated that hydrolysis of the amidoxime group of LXI had occurred to produce LIX. The uv, ir, pmr, and melting point were identical to LIX, and isolation of the intermediate LXI was not achieved. Synthesis of the 6-amino analog of thiosangivamycin was then attempted. Reaction under conditions which afford thiosangivamycin (XVII), i.e., pyridine, triethylamine, and hydrogen sulfide, produced no reaction and only starting material was recovered as indicated by the strong absorption (nitrile) in the ire Treatment of LVIII with OMF, dimethylamine, and hydrogen sulfide likewise afforded only starting material. 6-AminothioSangivamycin (LXII) was finally obtained when LVIII was added to a methanolic solution of sodium hydrosulfide which had been saturated with hydrogen sulfide for 45 minutes and the mixture was heated in a sealed vessel at 110-120° for 12 hours. That addition to the cyano group had occurred was established by the infrared spectrum which showed no absorbance in the 2200 cm- 1 region. A pmr spectrum corroborated the ir data in that two broad absorbances at 08.3 and 06.4 showed the exchange of 6 protons when 020 was added to the pmr sample tube. The downfield peaks (4 protons) were assigned to the amino group in the 4-position and the thioamide amino group at position 5. The absorbance at 06.3 was attributed to the 6-amino group. The ultraviolet spectrum was characteristic of a thioamide function in that a large bathochromic shift was observed with absorbance above 346 nm. The above findings indicate that the presence of two amino 72 REACTION SCHEME 13 =N Br NH2 XII LXII 1 NOH II , -NH 2 Br LX LVIII 1 / :~~ NNH HH LXIII LXI HO LIX OH 73 groups on the pyrrolopyrimidine ring system substantially deactivates the cyano group toward substitution by nucleophilic reagents as compared to systems with no exocyclic amino groups or when the amino function is contained in the pyrimidine ring at position 4. One other area was of interest in the 4,5,6-trisubstituted pyrrolopyrimidine nucleosides and required the synthesis of 6-hydrazinosangivamycin. As was discussed in the historical section, modification of the cyano group results in increased antitumor activity and substitution in the 6-position decreases toxicity. Therefore, it was of interest to determine if the antitumor activity of 4,5-diamino-8-(SD-ribofuranosyl )pyrazolo[3 1 ,4 1 :5,4]pyrrolo[2,3-d]pyrimidine was inherent to the compound or to a combination of the above factors. The pyrazolopyrrolopyrimidine can be viewed as a 6-substituted sangivamycin derivative. 6-Hydrazinosangivamycin possesses the modified cyano group and a 6-substituent but lacks the cyclic structure. Therefore, comparison of the activity of 6-hydrazinosangivamycin versus the tricyclic pyrazolopyrrolopyrimidine should show the importance of the tricyclic structure on antileukemic activity. Reaction of 6-bromosangivamycin (LX) with hydrazine in ethanol effected a nucleophilic displacement of the hydrazinosangivamycin (LXIII). 6~bromo group to yield 6- A pmr spectrum showed broad absorbances at 68.15 (CONH2), 66.95 (NH 2), and 05.85 (NHNH 2) which disappeared upon addition of 020. H2 was shown as a singlet at 68.0, the anomeric absorbance was centered at 66.2, and the remaining carbohydrate-peaks were located between 63 and 65. Testing data on this compound should give a clear indication as to the importance of the tricyclic structure on antileukemic activity. CHAPTER VII STRUCTURE-ACTIVITY RELATIONSHIPS Testing data on the compounds synthesized in the studies described in this thesis are displayed in Table X and Table XI. It should be emphasized that, during the time these compounds were prepared, there was a turn around time of about six months between the time a compound was submitted to DR&D for primary screening and receipt of the initial testing data. No data have as yet been received on the 5,6-disubstituted and 4,5,6-trisubstituted derivatives so that, obviously, no conclusions can be made as their antileukemic efficacy. Data received on the other compounds, however, show some interesting structure-activity relationships. It should be pointed out that conclusions drawn by comparing these series of compounds are probably valid even though the amount of data is, in some cases, small. Comparison of specific compounds, however, mayor may not be valid. Therefore, it should be noted that comparisons made between specific compounds are tentative and more data is needed to form firm comparisons. Table XI shows the data which have been received on the 5substituted pyrrolopyrimidine nucleosides. The first, and probably most important finding, is that the amino group at position 4 is not essential for a high level of antitumor activity. The toxic levels of compounds shown in Table VI appears to be around 20 mg/kg in most cases, while the compounds without the amino group at position 4 appear to be 75 about five times less toxic, i.e., toxicity decreases at around 100 mg/kg. A comparison of thiosangivamycin (XVII) with XLV shows the latter to be nontoxic at 200 mg/kg with a T/C of 167. 6-Thiosangivamycin is toxic until a level of about 20 mg/kg with a T/C range of about 137 to over 200. The activity levels of the compounds shown in Table VI appear to be slightly higher than the compounds shown in Table XI, but more testing data are required to determine the optimum dose for the 5-monosubstituted derivatives. The 5-monosubstituted pyrrolopyrimidine nucleosides display the same trend as was previously seen in the toyocamycin series, i.e., the compounds with a modified cyano group show the greatest activity. It is interesting that even the 5-cyano compound XLII shows good activity and is relatively non-toxic. The compounds with halogens in the 5-position are, as expected, not effective antileukemic agents. The two most important conclusions which can be drawn from the testing data obtained for the 5-substituted pyrrolopyrimidine nucleosides are that the amino group at position 4 is not essential for antitumor activity and removal of the amino group appears to decrease toxicity. Preliminary testing data on the pyrrolopyrimidopyridazine compound XXVIII is very encouraging. Below the toxicity level (approximately 75 mg/kg) very significant activity is displayed over a wide range of dosages. The highest T/C is 215% at a dosage of 15 mg/kg. Especially significant was the report of 2 cures out of the six mice involved in the study. Although the amount of testing data is limited on XXVIII, the cures are significant because only two other pyrrolopyrimidine nucleosides have evidenced cures, sangivamycin (II) which has undergone 76 TABLE X 5-Substituted Pyrrolopyrimidine Nucleosides y NSC No. Compound 138279 XLII Y CN Dose(mg/kg) Survivors T/C 200 100 0/6 4/6 145 200 0/6 200 100 6/6 6/6 167 164 140859 XLIII ~-NH2 143687 XLV ~-NH2 145386 XLIV C-NH2 200 100 50 25 2/6 5/6 5/6 6/6 120 168 159 166603 XLIX Cl 200 100 0/6 5/6 107 1477787 L Br 200 1/6 163650 LI I 100 50 25 12.5 2/6 6/6 6/6 6/6 ~OH 106 106 102 77 clinical trial and thiosangivamycin, which is being considered for preclinical trial (Stage lib). One of the criteria for advancement to Stage lib is unique structure. Since XXVIII shows significant activity and possesses a unique structure, the possibility of advancement to Stage lib is very good. As was the case with the tricyclic pyrazo10pyrro10pyrimidine, XXVIII may be active due to the inherent activity of the tricyclic system or may only be acting as a 4-substituted pyrro10pyrimidine with a modified cyano group at position 5. If XXVIII is acting as a 4-substituted pyrro10pyrimidine with a modified cyano group, the same level of activity would be expected from a pyrro10pyrimidine nucleoside substituted in the 4-position with a methy1amino group and with a carboxamidine function in the 5-position. However, as was shown in the study involving the effect of modifying the amino group at position 4, a methy1amino group at position 4 drastically reduces activity. Also, comparison of the activity of the carboxamidine (hydrochloride) XVII, shows that the amidine has the lowest level of activity. Therefore, based on previous studies it appears that the antitumor activity of XXVIII is inherent to the tricyclic system. Another way of viewing XXVIII is as a 4-hydrazino-5-carboxamide derivative. The compound, 4-hydrazino-7-(e-D-ribofuranosy1)pyrro10[2,3-d]pyrimidine-5-carboxamide (N.S.C. No. 145389) has been prepared in our laboratory (Table V). Only two lines of testing data are available and at a dose of 100 mg/kg (toxic at 200) a TIC of 145% is reported. Therefore, without more data on NSC 145389, a definite conclusion cannot be drawn. In any case, the unique structure of XXVIII- increases its chances of passage to Stage lIb according to DR&D protoco1. 86 78 The effect of exocyclic heterocyclic rings at the 5-position of the pyrrolopyrimidine nucleosides is shown in Table XI. The first observation of consequence is that the 6-membered aromatic heterocycles are inactive and non-toxic. The large ring most likely prevents binding at an essential enzyme site and therefore, the compounds are probably not inhibitors or substrates for enzymes involved in the metabolism nucleic acids. Since all of the five membered aromatic heterocyclic rings show some activity and toxicity, it may be that the 6-membered aromatic rings are simply too large. However, effects other than size may also be influencing the activity because the phenylthiazole derivative XXXVIII shows very good activity and yet this molecule is larger than the dimethyltriazine XXXIII. However, rotation of the phenyl group may permit XXXVIII to bind to the enzyme site. Therefore, a reasonable conclusion as to the size of ring allowable for activity is that a 6-member heterocycle is on the border line and any substitution of methyl, ethyl, etc., will render the compound inactive. All of the five-membered rings show some activity and the conclusion can therefore be made that a reasonable amount of steric bulk is tolerated in the 5-position before activity is lost. The exception to this generalization is the trtazole derivative XXXVI. The steric contribution of the sulfur atom is probably not sufficient to explain the inactivity of XXXVI as compared to XXXV. XXXVI has two more tautomers than does XXXV and both of these tautomers reduce the electron density at the ring nitrogens. Since XXXV shows only borderline activity, the decrease in electron density of the nitrogens may be great enough to eliminate activity. A very definite increase in antitumor activity is noted when 79 TABLE XI 4-Amino-5-heterocyclic Pyrrolopyrimidine Nuc1eosides y Dose Surv. XXXI 200 100 6/6 6/6 100 100 XXXIV 400 200 100 6/6 6/6 6/6 40 102 200 100 6/6 6/6 96 97 0/6 6/6 6/6 6/6 6/6 100 128 NSC No. Compound 150421 154825 150417 XXXII 150422 XXXV ,(1 1 .56 0.78 0.39 0.19 0.09 86 131 119 80 TABLE XI (continued) NSC No. Compound Dose Surv. 400 200 100 3/6 6/6 6/6 108 106 H 147785 XXXVI J[>S H 154824 XXXVIII 200 100 6/6 6/6 148 135 147786 XXXIX 200 100 50 25 6/6 6/6 6/6 6/6 105 150 155 144 200 100 50 25 12.5 5/6 6/6 4/6 5/6 6/6 127 135 133 127 117 200 0/7 100 0/7 50 6/6 30 6/6 25 6/6 20 6/6 15(2 cures)6/6 12.5 6/6 8.20 6/6 5.50 6/6 134 159 174 159 215 160 139 124 .t) H 158899 XL D H 154020 XXVIII C~nNH2 ~~ I Ribose 81 non-aromatic ring is compared to an aromatic ring. The imidazolinyl derivative XXXIX shows very good activity, and decreased toxicity, compared to the triazole XXXV. Likewise, the pyrimidinyl compound XL is active whereas the triazine compounds are not. The effect is most probably due to an increased electron density at the ring nitrogen atoms caused by elimination of delocalization due to aromaticity. Finally, the increase in size from a 5-membered non-aromatic ring to a 6-membered non-aromatic ring causes a slight decrease in activity as is shown by comparing the imidazolinyl derivative XXXIX to the pyrimidinyl compound XL. This is most likely a steric factor. In summary, this study has shown that a 5-membered ring has a greater likelihood of showing activity than a 6-membered ring and increasing the electron density at the ring nitrogen, i.e., going from aromatic to non-aromatic systems, has a significant effect on antitumor activity. It should be noted, however, that none of the derivatives with exocyclic rings at the 5-position are as active as the compounds with less bulky substituents in the 5-position. CHAPTER VIII SUMMARY AND CONCLUSIONS The effect of various substituents at the 4- and/or 6-positions on the susceptibility of a cyano group at the 5-position to nucleophilic addition was studied. Nucleophilic attack at the cyano group was facil- itated when a hydrogen resided at the 4-position. An amino group at the 6-position was found to have a greater deactivating effect on the cyano group than an amino group at the 4-position. The presence of amino groups at both the 4- and 6-positions greatly deactivates the cyano group toward nucleophilic attack and high temperatureswlth long reaction times were necessary to effect any addition to the cyano group. The effect on antileukemic activity of heterocyclic rings in the 5-position of the pyrrolopyrimidine nucleosides was studied. Ring closures on the cyano group of toyocamycin were effected using various 1,3-dipolar addition reactions to form 5 and 6 membered heterocyclic rings. Conden- sation of tubercidin-5-carboxamidrazone with various diketones led to substituted as-triazines and with aldehydes furnished certain 1,2,4-triazoles. Preparation of a 1 ,2,4-oxadiazole was achieved using tubercidin-5-carboxamidoxime. Ring annulation of tubercidin-5-thiocarboxamide with phenacyl bromide produced a thiazole. Testing data indicate that a six membered aromatic ring is the largest group which can be accommodated without loss of activity. All of the 5-membered heterocyclic rings showed some activity. Activity was increased when the ring was non-aromatic and a 5membered non-aromatic ring was more active than a 6-membered non-aromatic 83 ring. None of these derivatives were as active as the compounds with smaller groups at position 5. Testing data on compounds with a cyano group or modified cyano group at position 5 indicate that an amino group at the 4-position is not essential for activity and a decrease in toxicity is observed upon removal of the 4-amino group. The derivatives which showed the best antileukemic activity were the 5-carboxamidexime and the·5~thiocarbox­ amide pyrrolo[2,3-d]pyrimidine nucleosides. Electrophilic halogenation of 7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine at the 5-position was observed with N-chlorosuccinimide, Nbromosuccinimide, and iodine monochloride. Treatment of 5-cyano-7-(S- D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine with bromine water produced the 6-bromo derivative. Nucleophilic displacement of the 6-bromo group with hydrazine and ammonia was achieved. A new tricyclic nucleoside, 6-amino-4-methyl-8-(S-D-ribofuranosyl)(4H,BH)pyrrolo[4,3,2-de]pyrimido[4,5-c]pyridazine, has been prepared by reaction of methylhydrazine with 4-chloro-5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine. The reaction pathway was discerned by the isolation of 5-cyano-4-N-l-methylhydrazino-7-(S-D-ribofuranosyl)pyrr010[2,3-d]pyrimidine. Testing data shows that this tricyclic compound possesses significant activity against leukemia L12l0. In conclusion, thas work has shown that the best antileukemic activity for a pyrrolopyrimidine nucleoside can be expected from compounds possessing either an amino group or hydrogen at position 4 and a carboxamide or thiocarboxamide at position.5. An area which deserves more study is the tricyclic derivatives of the pyrrolopyrimidine nucleosides. CHAPTER IX EXPERIMENTAL SECTION I. Instrumentation and Procedures Elemental analyses were performed by Heterocyclic Chemical Corp., Harrisonville, Missouri. Melting Points were deternlined on a Thomas-Hoover melting point apparatus and are uncorrected. Ultraviolet spectra were determined on a Beckman DK-2 Spectrophotometer. Infrared spectra were determined on a Beckman IR-5A using pressed potassium bromide pellets. Pmr spectra were obtained on a Varian A-60 high resolution spectrometer utilizing tetramethylsilane as an internal standard and the chemical shifts are expressed as 8, parts per million, from tetra- methyl silane. Thin layer chromatography used Mallinkrodt SilicAR 7GF spread at 0.25 mm thickness on glass plates. Development was accomplished by the ascending technique using ethyl acetate/~-propanol/water (upper phase) in the ratio 4:1:2 unless otherwise noted. Toyocamycin was purchased from Koninklyke Nederlandsche Gist. and Spiritusfabriek N.V. II. Organic preparations 5-Cyano-4-(N-l-methylhydrazino)-7-(a-D-ribofuranosyl)pYrrolo- 85 [2,3-d]pyrimidine (XXVII). To 310 mg of 4-chloro-5-cyano-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVI) in 10 ml of ethanol was added, in one portion, 150 mg of methylhydrazine in 15 ml of ethanol. The reaction mixture was stirred at room temperature for 15 minutes. After 5 minutes a white precipitate began to form. The white precipitate was collected by filtration and recrystallized from a minimum amount of water to yield 166 mg (52%) of product. A form change occurred at 173-175° with a final melt at 205°. Anal. Calcd. for C13H16N604: C, 48.73; H, 4.69; N, 26.25. Found: C, 48.49; H, 4.81; N, 26.21. 6-Amino-4-methyl-8-(~-D-ribofuranosyl)(4H,8H)pyrrolo[4,3,2-de]­ pyrimido[4,5-c]pyridazine (XXVIII). Method A To 310 mg of 4-chloro-5-cyano-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVI) in 10 ml of ethanol was added 50 mg of methylhydrazine and the mixture heated at reflux temperature for 12 hours. After cooling at 5° for six hours the reaction mixture was filtered and the filter cake was recrystallized from a minimum amount of water to give 110 n1g (34%) of XXVIII. (mp, 205°). The white crystals were filtered and air dried. Anal. Calcd. for C13H16N604·3H20: C, 41.82; H, 5.88; N, 22.47. Found: C, 41.70; H, 5.78; N, 23.01. Method B A solution of 5-cyano-4-(N-l-methylhydrazino)-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVII) in water was heated at reflux temperature for 16 hours and then allowed to stand at 5° for 12 hours to afford fine yellow crystals of XXVIII. An infrared spectrum of the 86 product was superimposable on a spectrum of XXVIII obtained directly from XXVI. There was no depression on a mixture melting point with XXVIII from method A. 5-Cyano-4-methylamino-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXX). One hundred milligrams of 5-cyano-4-(N-l-methylhydrazino)-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVII) and 20 ml of water were mixed and heated to reflux temperature. One gram of wet W-7 Raney nickel was added portion-wise over a two hour period. The reaction was heated at reflux temperature for 12 hours. The hot reaction mixture was filtered through a celite bed. A uv spectrum of the filtrate was identical to the uv spectrum of 5-cyano-4-methylamino7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine. 62 The filtrate was concentrated to approximately 20 ml and cooled at 50 for 12 hours. The colorless crystals were filtered and air dried. A mixture melting point with an authentic sample of 5-cyano-4-methylamino-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine 62 showed no depression (179-180 0). The infrared spectra were essentially superimposable. 4-Amino-5-[9-oxoindeno-(1 ,2-e)-as-triazin-3-Yl]-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXXI). One gram of 4-ai11~no .. 7-(s-D- ribofuranosyl)pyrrolo[2,3-d]pyrimid;ne-5~carboxamidr~zone(XIV),600 mg of ninhydrin, and 50 ml of methanol were mixed, and heated for 5 minutes to effect solution. The reaction mixture was filtered to remove insoluble material. The filtrate was stirred for 18 hours at room temperature. The bright yellow product was collected by filtration and dried in vacuo to yield 1.1 g.{81%) of XXI which melted at 284-285°. Anal. Calcd. for C2oH17N70S'1/2H20: C, 54.05; H, 4.05; N, 22.25. Found': C, 53.85; H, 4.09; N, 22.03. 87 4-Amino-5~(as-triazin-3-yl)-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]­ pyrimidine (XXXII). A mixture of 960 mg of 4-amino-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV), 210 mg of glyoxal trimer, and 45 ml of water was heated at reflux temperature for 12 hours. The water was removed under reduced pressure on a steam bath and the pale brown residue was extracted with 80 ml of hot diethyl ether. The brown powder was air dried to give 900 mg (88%) of product, mp 254-255°. Anal. Calcd. for C14HlSN704l/2H20: C, 46.45; H, 4.52; N, 28.40. Found: C, 46.83; H, 4.76; N, 28.49. 4-Amino-5-(5,6-dimethyl-as-triazin-3-yl)-7-(e-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXXIII). One gram of 4-amino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV) was dissolved in 50 ml of boiling water. To this solution was added 0.35 ml of biacetyl and heating was continued for 0.5 hours. The reaction mixture was allowed to stand at 5° for 5 hours. The pale yellow solid was collected by filtration and the solid was washed with 50 ml of ethyl ether. The product was recrystallized from methanol to yield 850 mg (74%) of pale yellow crystals, mp, 280-281°. Anal. Calcd. for C16H19N704: C, 51.44; H, 5.09; N, 26.27. Found: C , 50.98; H, 5.27; N, 26.29. 4-Amino~5-(5,6-diphenyl-as-triazin-3-yl)-7-(S-D-ribofuranosyl)­ pyrrolo[2,3-d]pyrimidine (XXXIV). Three hundred and twenty milligrams of 4-amino-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV) and 210 mg of benz;l were mixed in 15 ml of ethanol-water (1:2, v/v). The mixture was heated at reflux temperature for 10 minutes and the yellow solid was collected by filtration. The filter 88 cake was washed with 50 ml of diethyl ether and the yellow filter cake was air dried to yield 430 mg (87%) of product melting at 274-275°. Anal. Calcd. for C26H23N704: C, 62.78; H, 4.63; N, 19.92. Found: C, 62.80; H, 4.82; N, 19.92. 4-Amino-5-{1 ,2,4-triazol-3-yl)-7-{S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXV). To 640 mg of 4-amino-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV) was added 20 ml of 97% formic acid and the mixture was heated at reflux temperature for 20 hours. The excess formic acid was removed in vacuo and the residue was coevaporated with 15 ml portions of ethanol until a white solid was obtained. The solid was recrystallized from methanol to give 350 mg (53%) of XXXV, mp, 240-241°. Anal. Calcd. for C13HlSN704·1/2H20: C, 45.75; H, 4.66; N, 28.73. Found: C, 45.63; H, 4.67; N, 28.67. 4-Amino-5-(1 ,2,4-triazol-3-yl-5-thione)-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXXVI). One gram of 4-amino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamidrazone (XIV), 40 ml of ethanol, 200 mg of potassium hydroxide (solid), and 8 ml of carbon disulfide were mixed and heated at reflux temperature for 23 hours. The reaction mixture was cooled to room tempearture and the precipitate was collected by filtration. The filter cake was dissolved in 50 ml of water and IN hydrochloric acid was added dropwise until no more precipitate formed. The pale yellow product was collected by filtration, washed with hot water, (20 ml), and air dried. The product did not melt below 300° and weighed 950 mg (83%). Anal. Calcd. for C13H17N70SS·H20: C, 40.73; H, 4.04; N, 25.35. Found: C, 40.79; H, 4.02; N, 25.42. 89 4-Amino-5-{1 ,2,4-oxadiazol-3-yl)-7-(S-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine (XXXVII). A mixture of 320 mg of 4-amino-7{s-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine-5-carboxamidoxime (XV), 25 m1 of water, and 10 m1 of formy1acetic anhydride (distilled at 38-39° and 50 mm) were heated at reflux temperature for 20 hours. The solvents were removed in vacuo and the oily residue was triturated with hot ethanol (30 m1). The white solid which formed was collected by filtration and air dried to yield 100 mg (33%) of XXXVII, mp, 226-228°. Anal. Ca1cd. for C13H14N60S"1/2H20: C, 45.45; H, 4.39; N, 24.42. Found: C, 45.45; H, 4.51; N, 24.39. 4-Amino-5-[(4-pheny1)thiazol-2-y1]-7-(~-D-ribofuranosy1)­ pyrro1o[2,3-d]pyrimidine (XXXVIII). One gram of 4-amino-7-{S-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine-5-thiocarboxamide (XVII) and 600 mg of phenacyl bromide were added to 24 m1 of glacial acetic acid-water (3:1, v/v) mixture and heated at reflux temperature for one hour. The reaction mixture was cooled to room temperature, the precipitate was collected by filtration, and transferred to a beaker containing 50 m1 of water. Concentrated ammonium hydroxide was then added dropwise until the pH had reached 9. The mixture was heated on a steam bath for 10 minutes, filtered, and the crude product was recrystallized from methanol to yield 800 mg (62%) of XXXVIII, mp, 230-232°. Anal. Ca1cd. for C20Hl~Ns04S·1/2H20: C, 55.30; H, 4.62; N, 16.11. Found: C, 55.33; H, 4.63; N, 16.12. 4-Amino-5-(4H,5H-imidazo1in-2-yl)-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXXIX). A mixture of 1.1 g of toyocamycin and 40 m1 of ethylenediamine was heated at reflux temperature for 24 hours. The excess ethylenediamine was renloved in vacuo and the residue was 90 coevaporated with ethanol until a white solid remained. The crude product was recrystallized from a minimum amount of water to yield 1.2 g (90%) of XXXIX, mp, 260-261°. Anal. Calcd. for C14HlSN604: C, 50.30; H, 5.39; N, 25.15. Found: C, 50.35; H, 5.69; N, 25.41. 4-Amino-5-(1 ,4,5,6-tetrahydropyrimidin-2-yl)-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XL). To one gram of toyocamycin (1) in 60 ml of ethanol was added 6 ml of 1,3-diaminopropane and the mixture was heated at reflux temperature for 18 hours. The solvents were removed under reduced pressure and the residue was coevaporated with ethanol until a white solid remained. Recrystallization from water afforded 1.1 g (98%) of XL, mp, 261-262°. There was no absorbance in the infrared spectrum for a cyano group. Anal. Calcd. for ClsH20N604·l/2H20: C , 50.42; H, 5.88; N, 23.53. Found: C, 50.51; H, 5.84; N, 23.40. 4-Amino-5-aminomethy1ene-7-(S-D-ribofuranosyl)pyrro10[2,3-d]pyrimidine dihydroch10ride (XLI). Two grams of toyocamycin were dissolved in 100 ml of IN hydrochloric acid. To the solution was added 1 g of 10% palladium on charcoal and the mixture was hydrogenated at 40 p.s.i. for 5 hours at room temperature, and the reaction mixture was then filtered through a celite bed. The celite bed was washed with 50 ml of boiling water, the filtrates were combined, and taken to dryness under reduced pressure on a steam bath. The residue was coevaporated with ethanol (3 x 25 ml), triturated with 50 ml of hot ethanol (3 x 25 ml), and filtered. The residue was recrystallized from 50 ml of hot ethanol with enough water added to effect complete solution. The solution was allowed to stand at 5° for 12 hours to 91 afford 1.3 g (52%) of XLI, mp, 234-235°. Anal. Calcd. for C12H19C12Ns04 'H 20: C, 37.31; H, 5.44; N, 18.08. Found: C, 37.38; H, 5.50; N, 18.07. 5-Cyano-7-(S-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine (XLII). Five hundred milligrams of 4-ch10ro-5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XXVI) was dissolved in 50 ml of ethanol. To this solution was added 300 mg of 10% palladium on charcoal and 100 mg of solid sodium bicarbonate. The mixture was hydrogenated at 40 psi for 4 hours at room temperature. The reaction mixture was filtered through a celite bed to remove the charcoal and the filter cake was washed with ethanol (100 m1). The filtrate and wash were combined and the ethanol was removed under reduced pressure using a steam bath. The foam was coevaporated with ethanol (2 x 50 ml) to afford a hard yellow foam which weighed 320 mg (68%), mp, 175°. Anal. Calcd. for C12H12N404: C, 52.17; H, 4.38; N, 20.29. Found: C, 52.50; H, 4.54; N, 20.69. 7-(S-D-Ribofuranosy1)pyrrolo[2,3-d]pyrimidine-5-carboxamide (XLIII). Five hundred milligrams of 5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII), 5 ml of concentrated ammonium hydroxide, and 10 ml of water were mixed and heated on a steam bath for 3 hours. The solvents were removed in vacuo and the residue was triturated with ethanol at room temperature and filtered. The filter cake was recrystallized from ethanol with water added dropwise to effect solution. After cooling at 5° for 12 hours, the white crystalline XLIII was collected by filtration and dried under reduced pressure at 100° to give 300 mg (56%) of product, mp, 248-249°. Anal. Calcd. for C12H14N40S: C, 49.02; H, 4.80; N, 19.07. 92 Found: C, 48.98; H, 4.82; N, 19.15. 7-(B-D-Ribofuranosyl)pYrrolo[2,3-d]pYrimidine-5-carboxamidoxime (XLIV). One gram of 5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII), 550 mg of solid hydroxylamine, and 50 ml of ethanol were mixed and heated at reflux temperature. After 0.5 hour a white precipitate formed which was collected by filtration and washed with 50 ml of hot ethanol. Recrystallization from water gave, after air drying, 930 mg (83%) of XLIV, mp, 192-193°. Anal. Calcd. for C12HlSNSOs·H20: C, 44.04; H, 5.24; N, 21.40. Found: C, 43.93; H, 5.12; N, 21.21. 7-(S-D-Ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-thiocarboxamide (XLV). Method A Seven hundred and fifty milligrams of 5-cyano-7-(S-D-ribofurano- syl)pyrrolo[2,3-d]pyr"imidine (XLII) was added to a solution of 100 I11g of sodium in 40 ml of methanol and hydrogen sulfide was passed through the solution for 2 hours. The reaction mixture was tightly stoppered and stirred at room temperature for 12 hours. The pale yellow precipitate was collected by filtration and recrystallized from a minimum amount of boiling methanol to which water was added to the cloud point. Four hundred and eighty milligrams (57%) of XLV crystallized after cooling at 5° for 12 hours, mp, 243-244°. Anal. Calcd. for C12H14N404S·1/2H20: C, 45.28; H, 4.74; N, 17.55. F0 und: C, 45. 42; H, 4. 84; N, 17 . 65 . Method B Two hundred mg of 5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLII) was dissolved in 10 ml of pyridine containing 0.2 ml 93 of triethylamine. Hydrogen sulfide was passed through the solution for 6 hours at room temperature. The reaction mixture was tightly stoppered and stirred at room temperature for 1B hours. The solvents were then removed under reduced pressure, the residue dissolved in a minimum amount of methanol, filtered, and cooled at 5° for 14 hours. The pale orange powder was collected by filtration and air dried to give 120 mg (55%) of product whose ir, uv, and mp were identical to XLV prepared by Method A. 7-(2,3,5-Tri-0-acety1-e-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine (XLVII). Method A Two grams of 7-deazanebu1arine (XLVIII), 30 ml of dry pyridine, and 15 ml of acetic anhydride were mixed and stirred at room temperature for 20 hours. The pyridine and excess acetic anhydride were removed n vacuo. The oil which remained was dissolved in 100 m1 of methylene chloride, the methylene chloride was washed with 1N hydrochloric acid (2 x 30 m1), and water (2 x 50 m1). The methylene chloride layer was dried over anhydrous sodium sulfate. The methylene chloride was then removed under reduced pressure to give 1.15 g (39%) of a pale yellow syrup. The syrup was dissolved in methylene chloride and used without further purification in the halogenation reactions. Method B To a solution of 17.1 g of 4-ch10ro-7-(2,3,5-tri-Q-acety1-BD-ribofuranosyl)pyrro10[2,3-d]pyrimidine (VII) in 250 ml of ethanol was added 3.4 g of sodium bicarbonate and 6.B g of 5% palladium on charcoal. The mixture was hydrogenated at room temperature at 40 psi in a Paar hydrogenator for 4 hours. The reaction mixture was filtered 94 through a celite bed and the filtrate was taken to dryness under reduced pressure. The oily residue was dissolved in lOa ml of methylene chloride, washed with water (2 x 50 ml), and dried over anhydrous sodium sulfate. The methylene chloride was removed under reduced pressure to yield 11.4 g (71%) of XLVII as a syrup. The acetylated 7-deazanebularine was used without further purification in the halogenation reactions. 5-Chloro-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLIX). Two grams of 7-(2,3,5-tri-~-acetyl-s-D-ribofuranosyl)pyrrolo[2,3-d]­ pyrimidine (XLVII) and 800 mg of N-chlorosuccinimide were dissolved in 20 ml of methylene chloride. The reaction mixture was stirred at room temperature for 16 hours. Ten milliliters of Silica gel (Baker AR) was added to the reaction mixture and the methylene chloride was removed under reduced pressure. The Silica gel containing the reaction mixture was added to the top of a Silica gel dry column (5 x 38 cm). The Silica gel column was not deactivated. The column was eluted with acetone-chloroform (1 :4; v/v) and fifty milliliter fractions were collected. Fractions 1-3 were combined and the solvents were removed under reduced pressure to afford 1.9 g of an amber oil. The oil was covered with 50 ml of methanolic ammonia (previously saturated at -10°), tightly stoppered, and left at room temperature for 18 hours. The solvent was removed under reduced pressure and the residue was extracted with 50 ml of boiling ethanol, filtered, and air dried to yield 750 mg (72%) of XLIX, mp, 198-200°. Anal. Calcd. for CIIH12C1N304: C, 46.15; H, 4.20; N, 14.65. Found: C, 46.03; H, 4.25; N, 14.71. 5-Bromo-7-(s~D-ribofuranosyl)pyrrolo[2,3-dJpytimid;ne ~ 95 To 3 g of 7-(2,3,5-tri-Q-acetyl-s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLVII) in 100 ml of methylene chloride was added 1.15 g of N-bromosuccinimide and the mixture was stirred at room temperature for 16 hours. Column chromatography was performed as described for XLIX. Fractions 1-5 were conlbined, taken to dryness under reduced pressure, and deacetylated as described for LXIX. The residue, after removal of the methanol, was extracted with 50 ml of hot ethanol, filtered, and the filter cake air dried to give 1.28 g of L (49%), mp, 193-195°. Anal. Calcd. for CIIH12BrN304: C, 40.00; H, 3.66; N, 12.71. Found: C, 39.89; H, 3.72; N, 12.67. 5-Iodo-7-{S-D-ribofuranosyl)pyrrolo[2,3-dJpyrimidine (LI). To 2.3 g of 7-(2,3,5-tri-Q-acetyl-S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (XLVII) in 50 ml of methylene chloride was added 0.78 ml of iodine monochloride, and the mixture was stirred at room temperature for 20 hours. The reaction mixture was washed with 10% sodium thiosulfate (4 x 25 ml), and dried over sodium sulfate. Column chromatography was performed as described for XLIX. The first fraction was discarded. Fractions 2-5 were combined, taken to dryness under reduced pressure, and the carbohydrate moiety was deacetylated using methano1ic ammonia (100 ml). After removal of the methanol under reduced pressure, the reddish-brown residue was dissolved in a minimum amount of methanol and water was added dropwise until the solution became cloudy. The solution was cooled at 5° for 12 hours. The brown-red crystals were collected by filtration and air dried to yield 600 mg (26%) of LI, mp,216-218°. Anal. Calcd. for CIIH12IN304: C, 35.01; H, 3.18; N, 11.14. Found: C, 35. 18; H, 3.34; N, 11.01. 96 6-Bromo-5-cyano-7-(S-D-ribofuranosy1}pyrro1o[2,3-d]pyrimidine (LII). Two hundred milligrams of 5-cyano-7-(S-D-ribofuranosy1)pyrro1o[2,3-d]pyrimidine (XLII) was added to 10 m1 of water and stirred at room temperature. Water saturated with liquid bromine was added dropwise until a precipitate began to form (approximately 10 minutes). Three 10 m1 portions of bromine water were then added at 10 minute intervals. The reaction was stirred for an additional 30 minutes after the addition of the final portion of bromine water. The yellow precipitate was collected by filtration using an aspirator vacuum, washed with approximately 50 m1 of acetone, and air dried to give 200 mg (56%) of LII melting with decomposition at 243°. Anal. Calcd. for C12HIIBrN404: C, 40.56; H, 3.10; N, 15.77. Found: C, 40.52; H, 3.27; N, 15.81. 5-Cyano-6-hydrazino-7-(s-D-ribofuranosy1)pyrrolo[2,3-dJpyrimidine LIII . To 800 mg of 6-bromo-5-cyano-7-(s-D-ribofuranosy1}pyrroloL2,3-d]pyrimidine (LII) in 40 ml of ethanol was added 2.4 m1 of 97% hydrazine. The mixture was heated at reflux temperature for 0.5 hour and the white precipitate which had formed was collected by filtration and air dried to give 410 mg (60%) of LIII, rnp, 264-265° with decomposition. Anal. Calcd. for C12H14N604: C, 47.06; H, 4.51; N, 27.45. Found: C, 47.12; H, 4.49; N, 27.38. 5-Amino-8-(S-D-ribofuranosyl)pyrazolo[3 ,4 -5,4]pyrrolo[2,3-d]1 pyrimidine (LIV). 1 Three hundred milligrams of 5-cyano-6-hydrazino-7- (S-D-ribofuranosyl}pyrrolo[2,3-dJpyrinlidine (LIII) and 30 m1 of water were mixed and heated at reflux temperature for 6 hours. The reaction mixture was cooled to room temperature and acidified to pH 2 with 3~ 97 hydrochloric acid. The solvent was removed under reduced pressure on a steam bath. The residue was triturated with 30 ml of boiling ethanol, filtered, and the product was air dried to afford 210 mg (57%) of LIV, mp, >300°. Anal. Calcd. for C12HloC1N604: C, 41.02; H, 4.50; N, 23.93. Found: C, 41. 15; H, 4.52; N, 24. 19. 6-Amino-5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (lV). Method A One gram of 6~bromo-5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]­ pyrimidine (lII) and 75 ml of methanolic ammonia (saturated at _5°) were mixed in a sealed vessel and heated at 110-115° for 16 hours. Thin layer chromatography showed a mixture of two major compounds, Rf 0.8 and 0.5, which were bright blue under shortwave (254 nm) uv light. The mixture was dissolved -in a minimum amount of boiling methanol and cooled at 5° for 18 hours. The tan crystals were filtered and air dried to yield 250 mg (30%) of lV melting at 260-261°. Anal. Calcd. for C12H 13 Ns04: C, 49.14; H, 4.47; N, 24.05. Found: C, 49. 19; H, 4.57; N, 24. 1O. Method B To 680 mg of 6-bromo-5-cyano-7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (lII) in a sealed reaction vessel was added approximately 3 ml of liquid ammonia. The reaction vessel was sealed, placed in an oil bath which had been preheated to 100-110°, and heated for 12 hours. The residue, after evaporation of the excess ammonia, was triturated with 20 ml of methanol at room temperature. The insoluble yellow solid was collected by filtration, and air dried to give a 70% 98 yield of LV which displayed identical ir, uv, Rf, and mp to that prepared by method A. The second spot with an Rf of 0.5 was not observed in this case. 6-Amino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5carboxamide (LVI). Two hundred mg of 6-amino-5-cyano-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (LV), 10 ml of concentrated ammonia, and 2 011 of 30% hydrogen peroxide were mixed and stirred at room temperature for 2 hours and then allowed to cool at 5° for 12 hours. The white precipitate which had formed was collected by filtration and recrystallized from a minimum amount of water to give 120 mg (38%) of LVI, mp, 257-258°. Anal. Calcd. for C12HlSNsOs: C ,46.60; H, 4.85; N, 22.65. Found: C, 46.40; H, 4.84; N, 22.60. 6-Amino-7-(e-D-r1bofuranosyl)pyrrolo[2,3-d]pyrimidine-5thiocarboxamide (LVII). Five hundred milligrams of 6-amino-5-cyano7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyr-imidine (LV) was added to a solution containing 200 mg of sodium hydrosulfide and 20 ml of methanol which had been saturated with hydrogen sulfide for 40 minutes. The mixture was placed in a sealed reaction vessel, and the vessel was then heated at 110-120° for 12 hours, cooled to room temperature, and the methanol removed under reduced pressure. The residue was dissolved in a minimum amount of water and allowed to stand at 5° for 12 hours. The off-white precipitate was filtered and air dried to give 120 mg (21%) of LVII, mp, 237°. Anal. Calcd. for C12HlSNs04S'1/2H20: C , 43.11; H, 4.79; N, 20.96. Found: C, 42.99; H, 4.78; N, 20.83. 5-Cyano-4,6"diam;no~7-(S-D-ribofuranosyl)pyrrolo[2,3-d]- 99 pyrimidine(6-aminotoyocamycin) (LVIII). Two hundred milligrams of 6bromotoyocamycin (XII) was added to approximately 5 ml of liquid ammonia in a small sealed reaction vessel. The reaction mixture was placed in a preheated (lOO-llOO) oil bath, and heated for 22 hours. The reaction vessel was cooled in an ice water bath and the ammonia allowed to evaporate. The residue was triturated with 20 ml of methanol at room temperature. The white, methanol insoluble powder was collected by filtration and recrystallized from water. The white crystals were collected by filtration, and dried in vacuo at 80° for 12 hours to give 120 mg (72%) of LVIII, mp, 258-260° dec. Anal. Calcd. for ClzH14N604·HzO: C, 44.44; H, 4.94; N, 25.93. Found: C, 44.38; H, 4.94; N, 25.87. 4,6-Diamino-7-(S-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine-5carboxamide(6-aminosangivamycin) (LIX). Method A A mixture of 1.1 g of 6-aminotoyocamycin (LVIII) and 1.1 g of hydroxylamine in 80 ml of ethanol was heated at reflux temperature for 16 hours. The white solid was collected by filtration and recrystallized from a minimum amount of water. The colorless crystals were collected by filtration and dried under reduced pressure at 80° for 5 hours to give 880 mg (75%) of LIX, mp, 268-269°. Anal. Calcd. for ClzH 1S N6 06: C, 44.44; H, 4.94; N, 25.92. Found: C, 44.19; H, 4.73; N, 25.78. Method B One gram of 6-bromosangivamyc1n (LX) was added to approximately 5 ml of liquid ammonia and the mixture was heated at 110-120° in a sealed vessel for 16 hours. After removal of the excess ammonia by 100 evaporation, the residue was triturated with 20 ml of ethanol and filtered. Recrystallization of the solid from water gave a 59% yield of LIX which was identical (uv, ir, mp) with LIX obtained by method A. 4,6-0iamino-7-(s-O-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5thiocarboxamide(6-am1nothiosangivamyc1n) (LXII). A mixture of 300 mg of 6-aminotoyocamyc;n, 100 mg of sodium hydrosulfide, and 20 ml of methanol (saturated for 45 minutes with hydrogen sulfide) was heated at 120-130° for 16 hours in a sealed reaction vessel, and then allowed to cool to room temperature. The pale yellow crystals which had formed were collected by filtration and recrystallized from a minimum amount of water. The pale yellow crystals were collected by filtration and dried under reduced pressure at BO° for 12 hours to give 240 mg (73%) of LXII, rnp, 246-247°. Anal. Calcd. for ClzH16N604S: C, 42.35; H, 4.71; N, 24.71. Found: C, 42.23; H, 5.02; N, 24.50. 4-Amino-6-hydrazino-7~-D-ribofuranosyl}pyrrolo[2,3-dJ­ pyrimidine-5-carboxamide (LXIII). A mixture of 1.1 g of 4-arnino-6bromo-7-(B-D-ribofuranosyl}pyrrolo[2,3-dJpyrimidine-5-carboxamide (LX) and 3 rnl of 97% hydrazine in 20 ml of ethanol was heated at reflux temperature for 45 minutes. The solvent was then removed under reduced pressure. The remaining residue was dissolved in 20 ml of hot water and allowed to stand at 5° for 12 hours. The precipitate was recrystallized from a minimum of methanol to give 600 mg (63%) of LXII, mp, 195-196°. Anal. Calcd. for C12H17N70s·1/2H20: C, 41.38; H, 5.18; N, 2B.44. Found: C, 41.29; H, 5.21; N, 2B.25. TABLE XII Ultraviolet absorption spectra of some pyrrolo[2,3-d]pyrimidines Number Name of Compound A. pHl max 274 9.3 X 10- 3 A. Ale. max fa' Ale. ., _3 A. pH 11 max X ,0 max 6.9 27 .1 15.7 12.5 287 283 (235) 10.5 275 (241) 236 15.6 4-Ch1oro-5-cyano-7(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine XXVII 5-Cyano-4-(N-l-methylhydrazino)- 278 238 7-(S-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine XXVIII 6-Amino-4-methyl-8-(S-D-ribofuranosy1)[4,3,2-de]pyrimido[4,5-c]pyridazine XXX 5-Cyano-4-N-methylamino7-(s-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine XXXVI max (288) 274 223 XXVI XXXIII fa' pH 1 14.3 E pH11 10- 3 max X 274 8.4 18.0 286 18.4 (310) 292 11 .0 (310) 290 286 12.0 19.0 (293) 284 (276) 237 16.3 15.9 7.6 (292) 283 (276) 237 7.6 4-Amino-5-{5,6-dimethylas-triazin-3-yl)-7-(S-D-ribofuranosyl)pyrrolo[2,3-dJ pyrimidine 295 266 15.7 15.3 302 280 11 .3 16. 1 (300) 280 16. 1 4-Amino-5-[(4-pheny1)thiazol-2-yl-5-thione]-7-(S-D-ribofuranosy1)pyrro1o[2,3-d] pyrimidine 326 252 11 .6 21 .6 330 266 10.8 18.6 324 264 10.0 18.2 0 XXXIX 275 4-Amino-5-(4H,5H-imidazo1in-2y1) -7 s-D-ribofuranosy1)pyrro1o- 232 [2,3 12.4 15.4 279 230 14.0 10.3 278 231 14.3 11 .3 XLII 5-Cyano-7-(B-D-ribofuranosy1)pyrrolo[2,3-d]pyrimidine 265 223 3.3 21 .0 271 220 5.2 21 .5 270 224 5.0 12.4 XLIII ibofuranosy1)pyrroi ,3-dJpyrimidine5-carboxamide (300) 267 224 3.5 29. 1 275 220 5.9 29.2 274 6.2 XLIV 7-(S-0 i bofuranosy1 )pyrrolo[2,3-d]pyrimidine5-carboxamidoxime (290) (260) (310) 227 32.0 (295) 228 31 .9 7- (S-D-R i bofuranosy1 )pyrrolo[2,3-d]pyrimidine5-thiocarboxamide (320) 260 224 13.0 31 .0 316 (280) 266 223 12.4 24.8 (318) 267 226 13.8 23.5 5-Bromo-7-(S-D-ribofuranosy1)pyrroloL2,3-d]pyrimidine (288) 267 234 2.6 23.1 (295) 264 228 3.6 25.1 XLV L 7-( 10.5 (29B) 270 229 2.6 19.1 LII 6-Bromo-5-cyano-7-(S-Dbofuranosyl)pyrro1o[2,3-d]pyrimidine 274 226 9.0 17.4 274 11 .0 282 10. 1 LITI 5-Cyano-6-hydrazino-7(S-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine 303 245 21 . 1 B.O 314 240 5.5 11 .3 312 232 15.0 5-Amino-B-(S-D-ribofuranosyl)- 320 261 . 7.3 16.9 (320) 248 9.0 (314) 246 17.3 LTV pyrazo1o[3~4~5,4]pyrro1o- [2,3-d]pyrimidine B.O 0 N 6-Amino-5-cyano-7-(S-Dbofuranosyl)pyrro1o[2,3-d]pyrimidine LVI 6-Amino- -( ,3 pyrrol carbOXi.lmi de bofuranosyi)pyrimidine-5- 299 252) 239 302 9.6 .2 311 279 (2S2) 236 10.3 10.7 26.1 304 (274) 2S0 232 23.0 (310) 281 234 17 .1 25. 1 9.3 10.0 242 11 .8 16.8 31 .0 312 282 267 236 14.3 21 .4 330 277 246 18.4 15.4 28.0 338 270 238 .5 12.0 24.6 338 268 238 20.6 1S .4 45.0 260 4.9 LVII 6-Amino-7-{S-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidineS-thiocarboxamide LVIII 5-Cyano-4,6-diamino-7-( D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (6-Aminotoyocamycin) 293 (230) 1S.0 292 24.2 290 19.9 LIX 4,6-Diamino-7-(S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-S-carboxamide (6-Aminosangivamycin) 294 (235) 15.3 293 17.9 293 18.2 LX 4-Amino-6-bromo-7-{S-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine-5-carboxamide 281 229 11 .3 12.2 (294) 286 11 .6 (294) 285 11 .6 368 14.6 13.0 18.1 363 (298) 273 (238) 16.0 10.2 287 233 II LXIII 4,6-Diamino-7-(S-D-ribofuranosyl)pyrro1o[2,3-d]pyrimidine-S-thiocarboxamide 4-Amino-6-hydrazino-7-(s-Dribofuranosy1)pyrro1o[2,3-d]pyrimidine-S-carboxamide 304 278 (249) 280 (230) 13.6 22.8 362 (297) 271 235 10.2 10.9 280 233 11 .9 11 .2 21 .2 16.4 0 w BIBLIOGRAPHY 1. H. Nishimura, K. Katagiri, K. Sato, M. Mayama, and N. Shimaoka, 2. K. V. Rao and D. W. Renn, ~ntimj.9:'..Qbi~J !19.ents_~he!!l_'!.~!l_?I~ 77 (1963) . 3. K. Anzai, G. Nakamura, and S. Suzuki, J. ,~_~_ ~!!!.ib;o!:.. (Tokyo), gAl 60(1955). 20"( (1957) . 4. K. Ohkuma, J. An_~~tQ.h (Tokyo), l_~~_!. 343 ( , 961 ) . 5. K.V. Rao, Abstr. 150th Amer. Chem. Soc. Meeting, Atlantic City, 6. S. Suzuki and S. Mar-'umo, ~ Antibiot. (Tokyo), 14AL 34(1961). 7. R.L. Tolman, R.K. Robins, and L.B. Townsend, New Jersey, September 1965, 24P. ~. ~ He~erocyc'. C~~m~~ 230(1967). ~ Amer~ Ch_~!!!:_ S0.f~...!.. 8. R.L. Tolman, R.K. Robins, and L.B. Townsend, ~QL 524(1968). 9. A. Bloch, M.T. Hakala, E. Mihich, and C.A. Nichol, Proc. Amer. _ . , . _ . _ .•_ _ • _0 _. ~a!!.~~_!:. ~_e~_=.i _~_ 6 (1964) . ------ - - - 10, s. P. 11. W. L. Wi 1son ,~!'!.~_~, ~l1~_mot~_~~QY_ g.~P!~_l. 52_!. 301 (1968) . 12. C. G. Zubr'od, Ca._~_~~ ~_h~lnQ~~.~X~y'. 13 C IL Cavins, _...~ . _......___ fh~!!L~_!!ler~~_ ._.. 14. C. A "15" Drug Owen, C. G. SOli th, f.anc~!,_ C~~!l!Q!~~rap.l .~~.E!.:..,_. ~~!t_ 19 (1964) . r: avins!I _><_,._._.,,"_.:.. 50 496 (1966) . ~ §IJ.. 197(1967). _J, .. _ _ •••• .. • ..............".. _ •. ~..!!!~r.:.. A~~Q5:_: . "__ '_"_"'~_"'" ~_~s"':.1. '12(1966) ., and Deve 1opment Di v'j s i on of Cancer Treatment, National Cancer Institute, National Institutes of Health, unpublished data. 16. B.C. !lin \J. l.l.r. st.er) R.K. Hohins, and L B, . ~~em., ~§._, 236 ('1970) • rovmsend~ 17. Y. M. Ikehara, K. Watanabfl, and S. Suzaki, ,J. Chern., .~_?., 3331 (1963) . 18. Y. Mizuno, M. Ikehara) K. Watanabe, and S. Suzaki, Chern. J:..~it!-,!,_.~ull.· (Tokyo), J1.' 109(1963). Mi7tJnO, QT~t. 105 19. E" C. Taylor and R. W. Hendess, ~ !\mel'..!- Chem. ~<?~:...!_ §..~_!.. 951 (1965) . 20. E.C. Taylor and R.W. Hendess, ~ Arner-=-. f.he~~.._ ~Q~":_~._~Z.!.. 1995(1965). 2'1. \~.d. Middleton, 22. J.F. Gerster, B. Carpenter, R.K. Robins, and L.B. Townsend, J. ~~~_~_ ~hem. !.. .1Jh 326 (1967) . 23. lLF. Pike, L. S)lechta, and P.F. Wiley, ~_._ !!et~.r.Q£Y..(J. ~~~rn~_.1.. _t~_ 159 (1964 . 24. R.L. Tolman, R.K. Robins, and L.B. Townsend, ~Am~~.f..t:!.~~:..?.pc., ~Q1. 524(1968). 25. R.L. Tolman, R.K. Robins, and L.B. Townsend, ~_~ ~~.r.:. fl1_~~_~_ ?oC~ ~1-1.. 2102(1969) . 26. Y. IShidO,(A. Ma)ruyama, and Y. Kikuchi, Bull._ Chem._ ?O~..!- Jap., 40, 1389 1967 . 27. R.L. TOlman,(R.K.)RObins, and L.B. Townsend, ~_!.. 230 1967 . 28. W. Pfleiderer and R.K. Robins, Chem._ Ber., 98.!_ 1511(1965). 29. T. Sato, T. Shim)adaChi, and R. Ishido, ~~_ Chem-=- SoS.-=- ~_~~~ 3~_1. V.A. Engelhardt, and B.S. Fisher, J. ~oc_=_1.. ~O! 2822 (1958) . E:!!!.~.!_ ~.hern~ ~~eterocy'cl_:. Ch~m'-L 1440(1960 . 30. Y. Ishido, H. Hosono, S. Isome, A. Maruyama, and T. Sato, J. Chern. 31 . H Tanaka, K. Iwabuchi, S. Isome, A. Maruyama, Y. Kikuchi, and T. ~_Q~.!._ Jap. , Sato, R'L ~ ~hem ._ 1389 ( 1964 ) . Soc. --.---- ._, ~~£.:_~_ 3th 2019 (1965) . 32. B.C. Hinshaw, R.K). Robins, and L.B. Townsend, ~...:.. !!ete_roc.Y~J_,~_ £t!~m:,.2. 33. B C. H'inshaw, J.F. Gerster, R.K. Robins, and L.B. Townsend, J. .?~!. 885 ( 1968 . _~.~.!~.r:o_sy..sJ_~ ~ he.!!1...:~~ §..!. 21 5 ( 1969) . . ,_..- 34. R.l. To 1man, R. K. Robins, and L.B. Townsend, ~~_ HeteIg_~~~.~. Chern.-!.. ?" 703 ( 1971 ) . 35. A. H. 36. K. 37. ILL. Tolman and L.B. Townsend, Tetrahedron ~ett.!..!~ 4815(1968). II Heterocyc 11 c Compounds, Ii R. C. E1 der"fi e'J d, ed., Wi 1ey, New York, N.Y., 1950, pp.300~·317. CO("W"j n, Schofield, IIHetero~aromatic Nitrogen Compounds," Plenum Press, New York, N.Y., 1967, p. 60. 106 38. R.J. Suhadolnik, S.!. Finkel, and B.M. Chassy, J. Biol. Chern., 243, --- 3532(1968). 39. B.M. Chassy and R.J. Suhadolnik, 40. H.M. Kalckar, 41. M.E. Balis, "Antagonists and Nucleic Acids," Wiley, New York, N.Y., 1968, p. 61. 42. A. Bloch, R.J. Leonard, and C.A. Nichol, Biochim. Biophys. Acta, 138, 10(1967). ---- 43. Y. Fukagawa, T. Sawa, T. Takeuchi, and H. Umezawa, J. Antibiot. -(Tokyo), ~, 191(1965). 44. S. Nishimura, F. Harada, and M. Ikehara, 129,301(1966). 45. D.C. Ward, A. Cerami, E. Reich, N. Mendelsohn, and G. Acs, Chern., 245,1843(1970). 46. R.J. Suhado1nik, T. Uematsu, H. Uematsu, and R.G. Wilson, Chern., 243, 2761 (1968). 47. G. Acs, E. Reich, and A. Mori, Proc. Natl. Acad. Sci. (U.S.), 493(1964). 48. A. Tavitian, S.C. Uretsky, and G. Acs, Biochim. Biophys. Acta, -157, 33(1968). 49. M. Saneyoshi, R. Tokuzen, and F. Fukuoka, Gann., 50. F.M. Schabel, Jr., J.A. Montgomery, H.E. Skipper, W.R. Laster, Jr., and J. R. Thomson, Cancer Res., .?l, 690 (1961 ) . 51. H.E. Skipper, R.K. Robins, J.R. Thomson, C.C. Cheng, R.W. Brockman, and F.M. Schabel, Jr., Canc~r ~es., lL, 579(1957). 52. G.G. Kelley, G.P. Wheeler, and J.A. Montgomery, Cancer Res., 22, - - - - - -22, 329(1962). 53. J.A. Montgomery and K. Hewson, 54. M. Bobek, R.L. Whistler, and 'A. Bloch, 55. K. Zee-cheng and C.C. Cheng, 56. William Wilson, Cancer 57. J.E. Willson, D.E. Brown, D. Reynolds, D. Jaeger, and E. Timmens, Toxicological and Pathological Studies with BA-181008. Brochure supplied by John L. Smith Memorial for Cancer Research, Chas. Pfizer and Co., Inc. ~ ~ Biol. Chern., 243,3538(1968). Biol. Chern. ,167, 461 (1947). ~~ed. ~ Biochim~.Biophys. ~, Acta, ~ ~ Biol. Biol. ~, 219 (1965) . Chern., lQ, 665(1967). ~ Med. Chern., .12, 168(1972). Pharm. Sci., 59, 1630(1970). Chemothe~ Rept., 52,301(1968). 107 58. II.F. Bisel, F.J. Ansfield, J.H. Mason, and W. W'ilson, ~_Q. ' ~.E!)~.e_Y'_ 76 ( 1970) . Bes_., -- 59. C.G. Smith, l.M. Reineke, H. Harpootlian, M.R. Burch, A.M. Shefner, and E. C. NU'i rhea,d, f_~~~~.r. 3.~_~.!., 30 69 (1970) . 60. T.B. Grage, D.B. Rochlin, A.J. Weiss, and W.L. Wilson, Cancer Res., 19., 79(1970). - _. 61. J.A. Cavins, T.C. Hall, K.B. Olson, C.L. Khung, J. Horton, J. Co1f~~ (1 ;~~) ~. K. Shadduc k, Can_~~r f_~~TTl~_~!!~.~?,P.Y_ Jl~.~t~,_, ~J_, 62. B.C. Hinshaw, O. Leonoudakis, K.H. Schram, and L.B. Townsend, 162nd American Chemical Society National Meeting, September 13-17, 1971, Washington, D.C., Medicinal Chem'istry Section, Paper No. 16" 63. K.H. Schram and L.B. Townsend, 26th Annual Northwest Regional Meeting of the American Chemical Society, June 16-18, 1971, Montana State University, Bozeman, Montana. Abstract p. 77. 64. R.A. Long, R.K. Robins, and L.B. Townsend, J. Q.!:9.:._ 275(1967). 65. 0 G. Nelson, R. Roger, J.W. Heathe, and L.R. Newlands, 66. W. Ried and P Schomann, ~.nn .. , 67. L.'iter'atut'e values for as-triazine are 89.36 for H6 and 68.69 for H5. J,2 :lllZ. W. w. -Paudl er and J. M. Barton, J. g!.'~_ f}~~m .. , }l, 7_0_, 151 (1970) . 17200 68. f.~~_~, 3?_, C~~ Re~, ?Ji" 128(1968). ). 1<. Na.kanishi, IIInfrared Spectroscopy·~Practical , Holden-Day, San II Franci sco, 1962. fi9. 1\. Dornow 70. ILL. K Fischer, Chern ~ ~_~l"~_,~.~.' 72 (1966) . Slu'iner, R.C. Fuson, and D.Y. Curtin, "The Systematic Ident'ification of Organic Compounds," Wiley, New York, N.Y., 1964, p. 64. 71. [,W. Berndt, Fratzke, and B.G. Held, -..J._- .,Heterocycl. Chern., :.\ ('j H.A. 972) . ....- ..----.--.. -..... __ . ._.--- 72. F.C. Schaefer, liThe Chemistry of the Cyano Group,!! Z. Rappoport, ed., Interscience Publishers, New York, N.Y., 1970, p. 264. 73. F Reich, A Symposium on Recent Advances in Nucleoside Chemistry at the l56th National Meeting of the American Chemical Society, Atlantic City, N.J., 1968, MEDI 30. ~, 108 Jn~~ 74. B. Loev and M.M. Goodman, Chenl. 75. A. Albert, IIHeterocyclic Chemistry,'1 Oxford University Press, New York, N.Y., 1968. 76. G.B. Payne and P.H. Williams, J. 77. G.B. Payne, P.H. Deming, and P.H. Williams, 659(1961). 78. E.E. Gilbert and E.J. Rumanowski, U.S. Pat., 3,336 381(1967); Chern. ~,~?.!!'..:._, 68, 68747 (1968) . 79. E.E. Gilbert, E.J. Rumanowski, and P.E. Newa11is, Data, 11., 130(1968). 80. H.R. Snyder and C.T. Elston, 81. J.M. Bobbitt and D.A. Scola, J. 82. J.M. Bobbitt and R.E. Doolittle, 83. A. Galat, 84. H.E. Johnson and D.G. Crosby, 85. S.M. McElvain and BeE. Tate, h Amer. 86. R.I. Geran, N.H. Greenberg, M.M. Macdonald, A.M. Schumacher, and B. J. Abbott, Cancer Chen~C?therapy ~!. ~, Pa rt 3, l, No.2, September 1972. 87. R. Adams, V. Voorhees, and R.G. Shriner, 463(1944) . ~~ Amer. Chern. ~~ ~oc., (London), ~ fhe~, Amer. ~ ~. (1967) . 26, 651 (1961). i:_ QrlL:_. C~~rn. ,~§.., Chem~ .~oc._, Chern., ~, ~ Chern. ~ 76, 3039(1954). 561(1960). Org. Chern., 29,2298(1964). 70, 3945(1948). ~ ~ Che~, 28, 3255(1963). Che~ ~oc., 7~, 2233(1951). ~~, ~oll. Vol. 1,