journal of medicinal chemistry volume 15.pdf

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155Sixteenth Street N.W., Washington, DC 20036

Synthesis of three oxytocin analogs related to[1-deaminopenicillamine]oxytocin prossessing antioxytocic activity

Raymond J. Vavrek, Martha F. Ferger, G. Ashley Allen,Daniel H. Rich, Alfred T. Blomquist, and Vincent Du Vigneaud

J. Med. Chem., 1972, 15 (2), 123-126• DOI: 10.1021/jm00272a001 • Publication Date (Web): 01 May 2002

Downloaded from http://pubs.acs.org on April 25, 2009

More About This Article

The permalink http://dx.doi.org/10.1021/jm00272a001 provides access to:

• Links to articles and content related to this article• Copyright permission to reproduce figures and/or text from this article

http://dx.doi.org/10.1021/jm00272a001

Journal of Medicinul Chemistry 0 Coprrighr I072 bb rhe American ChemicalSocrer)

VOLUME 15, NUMBER 2 FEBRUARY 1972

Synthesis of Three Oxytocin Analogs Related to [ 1 -Deaminopenicillamine ]oxytocin Possessing Antioxytocic Activity?

Raymond J. Vavrek, Martha F. Ferger, G. Ashley Allen,.$ Daniel H. Rich? Alfred T. Blomquist," and Vincent du Vigneaud*

Department of Chemistry, Cornel1 University, Ithaca, New York 14850. Received July 21, 1971

[ 1-Mercaptodimethylacetic acidloxytocin, [l+'-mercapto-a,a-dimethylpropionic acidloxytocin, and [ 1- P-mercapto-0,P-diethylpropionic acidloxytocin, analogs of deaminopenicillamine-oxytocin, have been synthesized from a common protected octapeptide resin intermediate and purified by sequential gel filtration on Sephadex G-15 in 50% AcOH and 0.2 N AcOH. The 3 compounds were devoid of oxytocic and avian vasodepressor activities, but all showed a significant degree of inhibition of the effects of oxytocin on the isolated rat uterus and on avian blood pressure. Each compound showed approximately the same inhibitory potency in both biological systems when compared to that of deaminopenicillamine- oxytocin. [I-Mercaptodimethylacetic acidloxytocin and [ 1-0-mercapto-&,a-dimethylpropionic acidloxytocin had about 20% and 3 3 to 47%, respectively, of the inhibitory potency of deaminopenicillamine- oxytocin. [I-0-Mercapto$,P-diethylpropionic acidloxytocin had approximately twice the inhibitory potency of deaminopenicillamine-oxytocin ( [ l -P-mercapto-p,P-dimethylpropionic acidloxytocin in both systems.

In a series of studies bearing on the specificity of the half- cystine residue in position 1 of oxytocin (Figure 1) in rela- tion to the pharmacol behavior of the hormone, this residue was formally replaced with an L-penicillamine (L-P, pdi- methylcysteine) residue.' The resulting [l-~-penicillamine]- oxytocin was devoid of oxytocic activity but instead turned out to be a potent inhibitor of the oxytocic activity of oxy- t ~ c i n . ' - ~ [l-D-Penicillamine]o~ytocin'-~ was also found to possess antioxytocic activity, but to a lesser extent. Both diastereoisomers also showed an inhibitory effect on the avian vasodepressor (AVD) effect of oxytocin.

Following the unexpected finding that the presence of two Me groups on the B-carbon of the half-cystine residue converts oxytocin to an antioxytocic agent, studies were initiated on related analogs to see what modifications in the 1 position would enhance or diminish the antioxytocic activity. Since the formal replacement of the free amino group of oxytocin with H gives an analog, deamino-oxytocin ([l-0-mercaptopropionic acidloxytocin), that is even more potent in its oxytocic activity than ~ x y t o c i n , ~ , ~ the deamino analog of penicillamine-oxytocin was prepared.' Deaminopenicillamine-oxytocin ([ 1 -p-mercapto-p ,P-dimethylpropionic acidloxytocin) had as potent an antioxytocic effect as L-penicillamine-oxytocin and likewise showed an anti-AVD effect.'-' It is of interest that [l-deaminopenicil- lamined8]oxytocin, in which the deaminopenicillamine residue has been replaced with a fully deuterated deaminopenicillamine residue, has the same degree of antioxytocic activity as the protio analog.6

?This work was supported in part by U. S. Public Hea l thSeMG Grant HE 11680, by NIH Training Grant 5-T01-GM00834 and by a grant from the Mobil Foundation. All optically active amino acid residues are of the L variety unless otherwise indicated.

SNIH predoctoral fellow. SNIH training grant predoctoral fellow.

To ascertain whether both p-Me groups are necessary to produce antioxytocic activity, [1-LQ-mercapto-@methylpropionic acidloxytocin and [1-D-0-mercapto-0-methylpropionic acidloxytocin were studied.' Neither of these diastereoisomers having only one p-Me substiuent showed antioxytocic or anti-AVD activity, but both possessed 5-10% of the oxytocic and AVD potencies of deamino-oxytocin.

With the demonstration that two p-Me substituents are required for antioxytocic activity, we became interested in whether two such substituents on the a-C would have a like effect. We have therefore synthesized [l-P-mercapto-a,a-dimethylpropionic acidloxytocin, as described in the Experi- mental Section. The compd, which had no detectable oxytocic# or AVD** activity, was found to possess about half of the antioxytocic potency and about one-third of the anti- AVD potency of deaminopenicillamine-oxytocin. Although this compd is not as potent as the P-substitued analog, the results demonstrate that the antioxytocic activity is not confined to substitution on the 0-C.

It is to be noted that the analogs so far discussed all possess the 20-membered cyclic disulfide ring present in oxytocin and deamino-oxytocin. It has been found that the 19- membered ring analog of deamino-oxytocin, namely [l -mer- captoacetic acidloxytocin, possesses 25 units/mg of oxytocic and 4 units/mg of AVD activity. l3 Deamino-oxytocin possesses 803 and 975 units/mg of oxytocic and AVD activity, r e s ~ . ~ To find out whether substitution of Me groups

#The oxytocic response t o oxytocin o f isolated uteri from Sherman albino rats (200-250 g) in natural estrus was measured by the method o f Holton,'as modified by M u n ~ i c k , ~ with the use o f Mg-free van Dyke-Hastings s o h as the bathing fluid. Isotonic con- tractions were recorded with a Harvard heart/smooth muscle trans- ducer and Grass polygraph Model 5.

**Avian vasodepressor responses were measured on conscious roosters (-2500 g) by the method o f Coon," as described in "The Pharmacopeia o f the United States of America,"" as modified by Munsick, et al.

123

124 Journal of Medicinal Chemistry. 1972, Vol. 15, No. 2 I tlu tigneaud, ot ai

CaHdOH CrHs I 1

I II I I1 I CH--CH, NH, 0 CHa 0

CHr-CH-C--NH -CH--C --NH-CH

s i I 2 L o 0 0 NH

I S 1 6 I/ 6 1 1 4 I

CHI-CH-NH-C-CH-NH-C-- CH--(CHz)n-CONH*

I I C Ht

r.=n 1 - - CONHI

0 I

CHi-N, 8 I1 9

H-CH-C-NH--CH~-CONHa I

\ ' I1 CH-C-.-N

/ I C Hs-C Hr C HZ

I CH(CIIa)a

Figure 1. Structure of oxytocin, with numbers indicating the position of the individual amino acid residues.

for the hydrogens in the 1 position of the 19-membered ring analog would convert the latter to an antioxytocic agent, we have synthesized [I -mercaptodimethylacetic acid]- oxytocin. T h s analog, which had no detectable oxytocic or AVD activity, turned out to be only 20% as potent as deaminopenicillamine-oxytocin as an inhibitor of the oxytocic and AVD activities of oxytocin. Thus, the results with this compd demonstrate that antioxytocic activity is not confined to 20-membered ring analogs,

Returning, then, to a consideration of substituents in the 0 position, we decided to see whether increasing the size of the substituents would influence the degree of antioxytocic activity. We therefore synthesized the diethyl analog, [I$- mercapto-0,P-diethylpropionic acidloxytocin. T h s compd, which was devoid of oxytocic and AVD activities, possessed an enhanced antioxytocic activity which was twice that of deaminopenicillamine-oxytocin. The anti-AVD activity was enhanced to about 70% more than that of deaminopenicillamine-oxytocin. It is obvious that further increase in the size of the substituents on the 0-C might prove interesting, and furthermore that substituents on both a- and &carbons would warrant investigation.

Detailed data on the inhibitory properties of the analogs under consideration are given in Table I, along with the corresponding data for L- and D-penicillamineoxytocin and deaminopenicillamine-oxytocin, which have not previously been evaluated by the method used in this paper.

The inhibitory activities of the analogs were detd and expressed as PA, values as defined by Schild14 (see footnote a to Table I) in the antioxytocic and anti-AVD systems. Iso- lated rat uterus and conscious chicken prepns were used,

under the normal conditions of the oxytocic" and AVL)** assays, resp. A dose of synthetic oxytocin (x units) which would give a moderate, reproducible response (R) was f i rs t selected. Then 2x units were administered inimediately foi- lowing an injection of antagonist. The latter procedure was repeated until two levels of antagonist were found, one oi which would reduce the effect of 2x units of oxytocin t ( J

slightly more than and the other to slightly less than K. Concns of antagonist were calcd on the basis of ;t 1O-inl tissue bath in the antioxytocic studies and o n the basis of a n assumed blood vol of 150 in1 in the AVD studies. After plotting, according to Schild, the 2 concns of antagonist 011

a logarithmic scale against response, one interpolates between them to the molar concn (M) which corresponds to the response R. The negative logarithm of this concti to the base 10 is termed pA2. Average M ( f f ) from a given series of assays and the corresponding pA2 values are listed i n

Also presented in Table I are 2 columns ( @ ~ ~ p / i C l ~ ~ ~ l ~ , ~ ! which relate the antioxytocic and anti-AVD potencies of the different analogs to those of deaniinopenicillaminc-oxy tocin. These columns provide a profile of the group of analogs as inhibitors and make it evident that in most cases the ratios show a close parallelism between the two inhibitoi y systems studied. It is interesting that the only analogs in Table 1 which do not fit in with the parallelism shown are the two (L- and D-penicillamine-oxytocin) whicli bear a free a-NH2 group.

As earlier reported," the 22-membered ring analog of' de. amino-oxytocin, namely [I -7-niercaptovaleric acid] oxytocin, possessed no oxytocic or AVD activity, but did possess antioxytocic activity. This compd has now been found t a have anti-AVD properties also, and ava lues (expressed as in Table I) have been measured for both inhibitions: 3.22 ( 6 ) , o = 1 . I 4 (antioxytocic); 3.97 (7), o = 0.87 (anti-AVD). The corresponding pA2 values are 6.49 and 7.40, resp. Thiis for this analog also we find that its antioxytocic and anti- AVD potencies relative to the corresponding potencies ot deaniinopenicillamine-oxytocin are very nearly equal (aDAP/aarralog = 0.36 and 0.33, resp).

anti-AVD profiles of the compds studied may be significant. One may speculate that there is some basic similarity in the receptors involved j n the two tissues. Possibly further com- parisons of the relative inhibitory potencies of groups of compds in 2 or more biol systems may yield important information that can be related to the types of receptors involved.

Table I for each analog. -.

The marked parallelism between the antioxytocic and

--_ Table I. Inhibitory Properties of Oxytocin Analogs Related t o [I-Deaminopenici1lamine)oxytociri - __l-.._-l- I_-

anti- AVD _.I__ ~. . antioxytocic Analog M,a M x MijDAPIaafanalogb PA, M lo ' MDAP/fianalog - .-.- --

[ 1 -Deaminopenicillamine]oxytocin (DAP) 6.94 1.16 (36) 1 .oo 7.88 1.31 (27) 1 .oo [ I-1,-Penicillamine]oxytocin 6.86 1.39 (7) 0.83 7.50 3.18 ( 8 ) 0.4 1

[ 1-u-Penicillamine]oxytocin 6.32 4.81 (6) 0.24 6.78 16.8 (7) 0.08

[ 1-a-Mercapto-cupdimethylacetic acidloxytocin 6.16 6.91 (10) 0.1 7 7.21 6.18(6) 0.21

a = 0.51 a = 0.61

a = 0.40 = 1.05

a = 1.87

a = 3.56 [ 1-p-Mercapto-a,&-dimethylpropionic acidloxytocin 6.60 2.49 ( 9 ) 0.47 7.41 3.93 (6) 0.33

a = 0.98

a = 0.12

a = 4.0

(J = 1.71

a = 1.09

u = 0.17 [ 1-p-Mercapto-p,pdiethylpropionic acidloxytocin 7.24 0.58 ( 9 ) 2.0 8.11 0.78 (8) 1.68

___.

apA, values (see Schild") represent here the negative log to the base 10 of the av molar concn &) of an antagonist which will reduce the response of the uterine horn or the chicken t o 2x units of a pharmacologically active compd (agonist) t o the response to x units of the agonist. In these studies synt oxytocin was the agonist used. The number of individual detns is given in parentheses: and 17 is the s t d deviation. hRa t io of molar concn of [ I-deaminopenicillamine]oxytocin to molar concn of other analog.

[I -Deaminopenicillamine]oxytociiis Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 125

Experimental Section?? S-Bzl-mercaptodimethylacetic Acid (I). Benzylmercaptan (6.2

g) was added to a warm soln of Na (1.15 g) in anhyd EtOH (100 ml) and stirred under N, for 15 min. Ethyl a-bromoisobutyrate" (9.75 g) was added dropwise in anhyd EtOH (40 ml) and the soln was refluxed for 1 hr under N,. A soln of NaOH (2 g) in a mixt of H,O (20 ml) and EtOH (50 ml) was added to the hot s o h , and hydrolysis was carried out a t reflux for 6 hr. The solvents were removed in uacuo, and the residue was dissolved in H,O (150 ml). The aq soln was extd with toluene and then acidified with HCl. The ppt was collected, washed on the filter with H 0 and air-dried. Crystn from hexane give 4.4 g (needles), mp 95-%. Anal. (C,,H,,O,S)C, H. Ir and nmr data were consistent with I.

Boc-Tyr(Bzl)-Ue-Gln-Asn-Cys(Bzl>Pro-Leu-Gly Resin (11). This compd was synthesized by the solid-phase method of Merrifield.18 Boc-glycyl resin" (4 g) contg 0.76 mmole of glycine per g of resin was swelled overnight in CH,Cl, (40 ml) in a 100-mlMerrifield reaction flask. The Boc-glycyl resin was then subjected to seven 21-step cycles of deblocking, neutralization, and coupling as described in the synthesis of [8-phenylalanine]oxytocin.zo The AcOH was reagent grade which was distd through a 1.2 X 30-cm column packed with glass helices, Three equivs each of Boc-amino acids and DCI were used in the coupling steps. Boc-glutamine and Boc-asparagine were added as their p-nitrophenyl esters in 3 times the equiv amts and coupled in DMF. The DCI couplings were allowed to procede for 4 hr and the p-nitrophenyl ester reactions were allowed t o procede overnight. After the Boc-O-benzyltyrosine was coupled t o the peptide, the resin was washed with CH,Cl,, EtOH, AcOH, and finally with several portions of anhyd EtOH. The resin was sucked dry, then dried overnight in uacuo over KOH. The wt gain of the resin was 2.66 g, which represents 88% of the theoretical wt gain based on the amt of Boc-glycine esterified t o the resin.

S-Bzl-mercaptodimethylacetyl-Tyr(Bzl)-Ile-Gln-Asn-Cys(Bzl)- Ro-Leu-Gly-NH, (111). The protected octapeptide-resin compd I1 (1 g) was swelled overnight in CH,C1, (10 ml) and subjected t o one 21-step solid-phase cycle as described previously, Three equivs each of I and DCI were used in the coupling step, and after the first 3-hr coupling period the peptide-resin was subjected t o an addnl treatment with 3 equivs of the acid and of DCI, with coupling allowed to procede overnight.

Before cleavage with methanolic NH,," the substituted resin was subjected to an addnl deblocking and neutralization sequence, dried in vacuo over KOH, and then swelled for 3 hr in cold (0") anhyd MeOH. Anhyd NH,, distd from Na, was bubbled into the suspension for 3 hr at 0". The flask was stoppered and the suspension was stirred at 0" overnight. The MeOH and NH, were removed in vacuo, and the residue was stirred with DMF for 3 hr. The suspension was filtered, the resin was rinsed with several portions of DMF, and the combined DMF s o h was concd t o a small vol. This was added dropwise to a large vol of H,O with stirring, and the resulting milky suspension was stored overnight a t 5". The ppt was collected by centrifugation and washed several times with H,O in the centri- fuge tube, the H,O being removed by centrifugation and decantation. The moist, packed ppt was frozen and lyophilized, the lyophilisate was dissolved in AcOH and filtered, and the AcOH s o h was lyophilized to give 402 mg of compd 111, m p 217-218", [ a j z 2 D -32.7" (c 1, DMF). Amino acid analysis of an acid-hydrolyzed sample gave the correct ratios of amino acids and NH,. Anal. (C,,H,,N,,O,,S,. H,O) C, H, N. The material moved as one spot on silica gel tlc in Sa and sb.

[ 1-Mercaptodimethylacetic Acid]oxytocin. The protected nonapeptide amide 111 (100 mg) was dissolved in anhyd NH, (25 ml) and reduced with Na at the boiling point of NH,," allowing the blue color to persist for 15 sec before destroying excess Na with NH4C1. The NH, was removed in vacuo, and the residue was dissolved in 0.03% trifluoroacetic acid (300 ml). The soln was adjusted t o pH 6.8, and oxidn with ferricyanide4 was carried out keeping the pH at 6.8. The ferri- and ferrocyanide ions were removed with AG3-X4 resin (trifluoroacetate cycle). The clear, colorless filtrate was lyophilized, and the residue was dissolved in 50% AcOH (2 ml). This soln was applied to a 1.10 X 1 0 3 t m column of Sephadex G-15 which had been equilibrated with 50% AcOH." The analog was eluted at

j-thilelting points were detd in open capillary tubes and are uncorrected. Solvent systems:lSa, CH,OH-CHCI, (3:s); sb , n-BuOH- AcOH-H,O (4: 1: 1). Amino acid analyses were performed on a Beckman/Spinco amino acid analyzer Model 116 by the method of Spackman, er al."Where elemental analyses are indicated only by symbols of the elements, analytical results obtained for the elements, were within *0.4% of the theoretical values.

a flow rate of 9 ml/hr. E x a m of the fractions by uv absorption (275 mp) showed a sym peak centered at 46 ml. The tubes corresponding to the peak were pooled, dild with 1 vol of H,O, concd t o a small vol, and lyophilized. The lyophilisate was dissolved in 50% AcOH (2 ml) and applied to a 2.20 X 110-cm column of Sephadex G 1 5 which had been equilibrated with 0.2 N AcOH.~, The analog was eluted at a flow rate of 40 ml/hr and emerged as a sym peak centered at 400 ml, preceded by a small shoulder of uv,,,-absorbing material. The contents of the tubes corresponding t o the main fractions were pooled, concd to 5 ml, and lyophilized to give 40 mg, [aIz3D +17.8" (c 1, DMF). An acid-hydrolyzed sample of this analog gave the expected ratios of amino acids and NH,. Anal. (C,,H,,N,,O,,S,)C, H, N.

p-(S-Bzl-mercap to)-orpdimethylpropionic Acid (IV). Benzyl mercaptan (12.3 g, 0.1 mole) was added to a s o h of NaOEt (0.2 mole) in abs EtOH. After 15 min, chloropivalic acid (pchloro-a,wdi- methylpropionic acid) (13.6 g, 0.1 mole) was added; the soln was refluxed for 1 8 hr under N, and evapd to dryness. The residue was dissolved in H,O, and the basic aq layer was washed with Et,O (50 ml). The soln was adjusted t o pH 7, and the aq layer was washed with Et,O (100 ml). The soln was acidified (pH 2), and the product was extd with Et,O (160 ml). The Et,O layer was dried (Na,SO,) and evapd to dryness in vacuo. Crystn from hexane gave 13.6 g (62%) of the acid IV; mp 51-51.5" (cor). Recrystn (1.8 g) from pentane (25 ml) gave analytically pure IV (1.5 g). Anal. (C,,H,,O,S) C, H, S. Ir and nmr data were consistent with the acid IV.

p(S-Bzl-mercap to)-a,a-dimeth~lpropionyl-Tyr(Bzl>lle-Gln- Asn- Cys(Bz1)-Pro-Leu-Gly-NH, (V). The protected octapeptide-resin I 1 (1 g) was swelled overnight in CH,Cl, (10 ml) and subjected t o a sequence of deblocking, neutralization, and coupling using 3 equivs each of the acid IV and DCI as described earlier, At the end of the 4- hr coupling period, the resin was washed with CH,Cl, and a second coupling with 3 equivs of IV and DCI was allowed to procede overnight, Subsequent steps for washing, deblocking, and ammonolytic cleavage of the peptide from the resin were carried out as described for the synthesis of 111 t o give 357 mg of the protected nonapeptide amide V, m p 225", [ a I Z Z D -31.8" (c 1 , DMF). Amino acid analysis gave the correct ratios for the amino acids and NH,. The material moved as one spot on silica gel tlc plates in solvents Sa and sb . Anal. (C66H8J,101zS2) C , H, N.

[ I -p-Mercapto-a, ordimethylpropionic Acidloxytocin. The protected nonapeptide amide V (100 mg) was reduced with Na in liq NH,, the resulting product was oxidized with K,Fe(CN),, and the analog was isolated and purified as described earlier, wt 49.5 mg, [a]*'D -44.9" (c 1, DMF). The analog gave the correct ratios of amino acids and NH, on amino acid analysis. It was homogenous on silica gel tlc in solvents Sa and sb. Anal. (C45H69N1101zSZ) C, H, N.

p-(S-Bzl-mercapto)-p,p-diethylpropionic Acid (VI). A mixt of 3-pentanone (43 g), ethyl bromoacetate (83.5 g), C,H, (200 ml), and PhCH, (175 ml) was added dropwise to granular Zn (33 g) with heating. After the initial reaction had subsided, the mixt was refluxed for 4 hr, cooled, and 10% H,S04 (200 ml) was added. The org layer was dried (MgSO,), and the solvent was evapd. The residue was distd a t reduced pressure to give ethyl p-hydroxy-p,p-diethylpropi- onate (60.2 g), b p 75-76" (4 mm). This ester (46 g) was dehydrated by refluxing with P,O, (30 g) in C,H, (200 ml) for 24 hr. H,O (100 ml) was added to the cooled s o h and the org layer was sepd and dried (MgSO,). Distn of the residue after evapn of the solvent gave a mixt of 25% of the a,p-unsaturated ester and 75% of the p,yunsatu- rated ester, as detd by nmr. A portion of this mixt (15.6 g) was refluxed with benzylmercaptan (21 g) in piperidine (40 ml) for 24 hr. The s o h was cooled, acidified with HCl, and extd with Et,O. The Et,O was removed from the dried soln, and the residue was refluxed with K,CO, (25 g) in a mixt of MeOH (300 ml) and H,O (75 ml) for 24 hr. The solvent was distd from the reaction mixt until the vapors registered 100" and the distillate was clear. The residue in the flask was extd with Et,O, the org s o h was dried (MgSO,), and the Et,O was removed. Distn of the residue gave 17.3 g of a yellow oil, bp 158-167" (0.3 mm), which was crystd from hexane, m p 75-76". Anal. (C,,H,,O,S) C, H, S. Ir and nmr data were consistent with the structure VI.

p(S-Bzl-mercap to)$, p-diethylpropionyl-Tyr(Bz1)-Ile-Gln- Asn- Cys(BzI)-Pro-Leu-Gly-~, (VII) was synthesized by condensation of I1 (1 g) with VI, followed by ammonolysis and purification, all as described earlier; wt 370 mg, mp 235", [CY]~*D -43.3" (C 1, DMF). Amino acid analysis gave the correct ratios for amino acids and NH,. The protected nonapeptide amideVII moved as one spot on silica gel tlc plates in solvents Sa and sb. Anal. (C6,H,,N,,O,,SJ C, H. N: Calcd, 11.7; found, 11.2.

[ 1-p-Mercapto-p,p-diethylpropionic Acid]oxytocin. The pro-

126 Journal of Medicinal Chemistry, 1972, Vol. IS, No, 2 Hase, Schwartz, and Walter

tected nonapeptide amide VI1 was reduced with Na in liq NH,, the resulting product was oxidized with K,Fe(CN),, and the analog was isolated and purified as described earlier; wt 31 mg; [aIz3D -37.6” (c 1, DMF). Amino acid analysis gave the correct ratios of amino acids and NH,. The analog moved as 1 spot on silica gel tlc plates in the above solvents. Anal, (C4,H,3NL,0,zSz) C, H? N.

Acknowledgments. ‘The authors are greatly indebted to Dr. W. Y. Chan, Department of Pharmacology, Cornel1 University Medical College, for very helpful discussions relating to the biological studies, to Mrs. Janet Huisjen of this department for her highly capable performance in carrying out the bioassays, and to Dr. Louis L. Nangeroni, New York State Veterinary College at Cornell University, for use of his laboratory in this aspect of the work.

References (1) H. Schulz and V. du Vigneaud, J. Med. Chem., 9, 647 (1966). (2) W. Y. Chan, H. Schulz, and V. du Vigneaud, 111 International

Pharmacology Congress, Sao Paulo, Brazil, 1966, abstract 449. (3) W. Y. Chan, R. Fear, and V. du Vigneaud, Endocrinology, 81,

1267 (1967). (4) D. B. Hope, V. V. S. Murti, and V. du Vigneaud, J. Biol. Chem.,

237, 1563 (1962). ( 5 ) B. M. Ferrier, D. Jarvis, and V. du Vigneaud, ibid., 240, 4264

(1965). (6) V. du Vigneaud, J. D. Meador, M. F. Ferger, G. A. Allen, and

A. T. Blomquist, Bioorg. Chem., 1, 123 (1971).

H. Schulz and V. du Vigneaud, J. Amer. Chem. SOC., 88, 5015 (1966). P. Holton, Brit. J. Pharmcol., 3, 328 (1948). R. A. Munsick, Endocrinology, 6 6 , 4 5 1 (1960). J . M. Coon, Arch. Int. Pharmacodyn., 62, 79 (1939). “The Pharmacopeia of the United States of America,” 18th rev, U. S. P., Bethesda, Md., 1970, p 469. R . A. Munsick, W. H. Sawyer, and H. B. van Dyke, Endocri- nology, 66, 860 (1960). D. Jarvis and V. du Vigneaud, J. Biol. Chem., 242, 1768 (1967). H. 0. Schild,Erit. J. Pharmacol., 2, 189 (1947). W. Fraefel and V. du Vigneaud, J. Amer. Chem. Soc., 92, 1030 (1970). D. H. Spackman, W. H. Stein, and S. Moore, Anal. Chem., 30, 1190 (1958). J . Jacques, C. Weidmann-Hattier, and A. Marquet, Bull. SOC. Chim. Fr., 678 (1958). R. B. Merrifield, J. Amer. Chem. SOC., 85, 2149 (1963); Bio- chemistry, 3, 1385 (1964); G. R. Marshall and R. B. Merrifield, ibid.,4, 2394 (1965). H. Takashima, V. du Vigneaud, and R. B. Merrifield, J. Amer. Chem. Soc., 90, 1323 (1968). J . W. M. Baxter, M. Manning, and W. H. Sawyer, Biochemistry, 8, 3592 (1969). M. Bodanszky and J. T. Sheehan, Chem. Ind. (London), 1423 (1 964). R . H. Sifferd and V. du Vigneaud, J. Biol. Chem., 108, 753 (1935). M. Manning, 7. C. Wuu, and J. W. M Baxter, J. Chromatogr., 38, 396 (1968).

Further Studies of the Role of the Asparagine Residue in Oxytocin. Synthesis and Biological Properties of [ S-cu,y-Diaminobu tyric acid]oxytocinf $

S. Hase, I . L. Schwartz, and R. Walter*

Mount Sinai School of Medicine of the City University o f N e w York, New York, New York 10029. Received July 15, 1571

The synthesis of [ 5-cu,y-diaminobutyric acidloxytocin, an analog of oxytocin which contains an a,y-diaminobutyric acid residue in place of the asparagine residue in position 5 , is reported. On assay for biological activity this analog was found to possess 0.03 unit/mg of oxytocic activity, approximately 0.03 unit/mg of avian vasodepressor activity, and less than 0.01 and 0.001 unit/mg, respectively, of mammalian pressor and antidiuretic activities. The analog was incapable of inhibiting the oxytocin-induced responses in the 4 biological systems tested. It is concluded that [ 5-a,y-diaminobutyric acidloxytocin possesses a low affinity for the neurohypophyseal hormone receptors, and that this effect is a result of an increase in the conformational flexibility of this analog as compared with oxytocin.

All naturally occurring neurohypophyseal hormones of known amino acid sequence possess an Asp(NHz) residue in position 5. Oxytocin analogs which bear a Me,’” 0-carbox- an i idee th~ l ,~ hydroxymethyl,’ y -amin~propyl ,~ and i-Pr5 side chain in position 5 instead of the carboxamidemethyl moiety all exhibit an exceedingly low potency with respect to the activities characteristic of oxytocin-a finding which led us to focus on the importance of position 5 for the conformational stability of the hormonal molecule.678 The elucidation of the conformation of oxytocin’ confirmed the central role of the Asp(NHz) residue. It is critical for the maintenance of both /3-turns comprised of the sequences -Tyr-Ile-Glu(NHz)-Asp(NH2)- and -Cys-Pro-Leu-Gly- (Figure 1).

of [ 5-a,y-diaminobutyric acid] oxytocin and on the determination of its biological properties in order to assess

In the present communication we report on the synthesis

.-.

?This work was supported in part by United States Public Health

/Abbreviations used have been suggested in J. Med. Chem., 13(5), Service grants AM-13567 and AM-10080.

8A (1 970). The amino acids (except glycine) are of the L configuration.

the capability of qy-diaminobutyric acid (Dbu) to replace successfully the Asp(NHz). The free tetrapeptide, S-Bzl- Cys-Pro-Leu-Gly-NH, , Glu(NH2)-Nr-Pht-Dbu ( 5 ) in the presence of 1.4 equiv of DCI and 2.0 equiv of N-hydroxysuccinimide according to the procedure of Weygand, et al., l4 to yield BOC- Glu(NHz)Nr-Pht-Dbu-S-Bzl-Cys-Pro-Leu-Gly-NHz (6).

The dipeptide 5 was secured from Nr-Pht-Dbu*H,O (l), obtained after phthalylation of the dihydrochloride salt of the free acid by the general procedure of Nefkens, et al., ’’ in the following manner: the acid 1 was converted to its p- toluenesulfonate salt (2), which in turn was esterified with PhzCHN2;16-’8 the resultant ester 3 was allowed to react with BOC-G~U(NH,)-ON~’~ to give BOC-Glu(NH&Nr- Pht-Dbu diphenylmethyl ester (4), which was subsequently deesterified to give 5 . The hexapeptide (6) was elongated stepwise to yield the fully protected nonapeptide, N-ZS- Bzl-CysO- Bzl-Tyr-Ile-Glu(NH2)-Nr -Pht-DbuS-Bzl-Cys-Pro- Leu-Gly-NH, (9). Peptide 9 was successfully dephthalylated with either hydrazine or hydrazine acetate” to yield 10, which was fully deprotected by treatment with

was elongated with BOC-


Role of the asparagine residue in oxytocin. Synthesis and biologicalproperties of [5-.alpha.,.gamma.-diaminobutyric acid]oxytocin

S. Hase, I. L. Schwartz, and R. WalterJ. Med. Chem., 1972, 15 (2), 126-128• DOI: 10.1021/jm00272a002 • Publication Date (Web): 01 May 2002






126 Journal of Medicinal Chemistry, 1972, Vol. IS, No, 2 Hase, Schwartz, and Walter

tected nonapeptide amide VI1 was reduced with Na in liq NH,, the resulting product was oxidized with K,Fe(CN),, and the analog was isolated and purified as described earlier; wt 31 mg; [aIz3D -37.6” (c 1, DMF). Amino acid analysis gave the correct ratios of amino acids and NH,. The analog moved as 1 spot on silica gel tlc plates in the above solvents. Anal, (C4,H,3NL,0,zSz) C, H? N.

Acknowledgments. ‘The authors are greatly indebted to Dr. W. Y. Chan, Department of Pharmacology, Cornel1 University Medical College, for very helpful discussions relating to the biological studies, to Mrs. Janet Huisjen of this department for her highly capable performance in carrying out the bioassays, and to Dr. Louis L. Nangeroni, New York State Veterinary College at Cornell University, for use of his laboratory in this aspect of the work.

References (1) H. Schulz and V. du Vigneaud, J. Med. Chem., 9, 647 (1966). (2) W. Y. Chan, H. Schulz, and V. du Vigneaud, 111 International

Pharmacology Congress, Sao Paulo, Brazil, 1966, abstract 449. (3) W. Y. Chan, R. Fear, and V. du Vigneaud, Endocrinology, 81,

1267 (1967). (4) D. B. Hope, V. V. S. Murti, and V. du Vigneaud, J. Biol. Chem.,

237, 1563 (1962). ( 5 ) B. M. Ferrier, D. Jarvis, and V. du Vigneaud, ibid., 240, 4264

(1965). (6) V. du Vigneaud, J. D. Meador, M. F. Ferger, G. A. Allen, and

A. T. Blomquist, Bioorg. Chem., 1, 123 (1971).

H. Schulz and V. du Vigneaud, J. Amer. Chem. SOC., 88, 5015 (1966). P. Holton, Brit. J. Pharmcol., 3, 328 (1948). R. A. Munsick, Endocrinology, 66,451 (1960). J . M. Coon, Arch. Int. Pharmacodyn., 62, 79 (1939). “The Pharmacopeia of the United States of America,” 18th rev, U. S. P., Bethesda, Md., 1970, p 469. R . A. Munsick, W. H. Sawyer, and H. B. van Dyke, Endocri- nology, 66, 860 (1960). D. Jarvis and V. du Vigneaud, J. Biol. Chem., 242, 1768 (1967). H. 0. Schild,Erit. J. Pharmacol., 2, 189 (1947). W. Fraefel and V. du Vigneaud, J. Amer. Chem. Soc., 92, 1030 (1970). D. H. Spackman, W. H. Stein, and S. Moore, Anal. Chem., 30, 1190 (1958). J . Jacques, C. Weidmann-Hattier, and A. Marquet, Bull. SOC. Chim. Fr., 678 (1958). R. B. Merrifield, J. Amer. Chem. SOC., 85, 2149 (1963); Bio- chemistry, 3, 1385 (1964); G. R. Marshall and R. B. Merrifield, ibid.,4, 2394 (1965). H. Takashima, V. du Vigneaud, and R. B. Merrifield, J. Amer. Chem. Soc., 90, 1323 (1968). J . W. M. Baxter, M. Manning, and W. H. Sawyer, Biochemistry, 8, 3592 (1969). M. Bodanszky and J. T. Sheehan, Chem. Ind. (London), 1423 (1 964). R . H. Sifferd and V. du Vigneaud, J. Biol. Chem., 108, 753 (1935). M. Manning, 7. C. Wuu, and J. W. M Baxter, J. Chromatogr., 38, 396 (1968).

Further Studies of the Role of the Asparagine Residue in Oxytocin. Synthesis and Biological Properties of [ S-cu,y-Diaminobu tyric acid]oxytocinf $

S. Hase, I . L. Schwartz, and R. Walter*

Mount Sinai School of Medicine of the City University o fNew York, New York, New York 10029. Received July 15, 1571

The synthesis of [ 5-cu,y-diaminobutyric acidloxytocin, an analog of oxytocin which contains an a,y-diaminobutyric acid residue in place of the asparagine residue in position 5 , is reported. On assay for biological activity this analog was found to possess 0.03 unit/mg of oxytocic activity, approximately 0.03 unit/mg of avian vasodepressor activity, and less than 0.01 and 0.001 unit/mg, respectively, of mammalian pressor and antidiuretic activities. The analog was incapable of inhibiting the oxytocin-induced responses in the 4 biological systems tested. It is concluded that [ 5-a,y-diaminobutyric acidloxytocin possesses a low affinity for the neurohypophyseal hormone receptors, and that this effect is a result of an increase in the conformational flexibility of this analog as compared with oxytocin.

All naturally occurring neurohypophyseal hormones of known amino acid sequence possess an Asp(NHz) residue in position 5. Oxytocin analogs which bear a Me,’” 0-carbox- an i idee th~ l ,~ hydroxymethyl,’ y -amin~propyl ,~ and i-Pr5 side chain in position 5 instead of the carboxamidemethyl moiety all exhibit an exceedingly low potency with respect to the activities characteristic of oxytocin-a finding which led us to focus on the importance of position 5 for the conformational stability of the hormonal molecule.678 The elucidation of the conformation of oxytocin’ confirmed the central role of the Asp(NHz) residue. It is critical for the maintenance of both /3-turns comprised of the sequences -Tyr-Ile-Glu(NHz)-Asp(NH2)- and -Cys-Pro-Leu-Gly- (Figure 1).

of [ 5-a,y-diaminobutyric acid] oxytocin and on the determination of its biological properties in order to assess

In the present communication we report on the synthesis

.-.

?This work was supported in part by United States Public Health

/Abbreviations used have been suggested in J. Med. Chem., 13(5), Service grants AM-13567 and AM-10080.

8A (1 970). The amino acids (except glycine) are of the L configuration.

the capability of qy-diaminobutyric acid (Dbu) to replace successfully the Asp(NHz). The free tetrapeptide, S-Bzl- Cys-Pro-Leu-Gly-NH, , Glu(NH2)-Nr-Pht-Dbu ( 5 ) in the presence of 1.4 equiv of DCI and 2.0 equiv of N-hydroxysuccinimide according to the procedure of Weygand, et al., l4 to yield BOC- Glu(NHz)Nr-Pht-Dbu-S-Bzl-Cys-Pro-Leu-Gly-NHz (6).

The dipeptide 5 was secured from Nr-Pht-Dbu*H,O (l), obtained after phthalylation of the dihydrochloride salt of the free acid by the general procedure of Nefkens, et al., ’’ in the following manner: the acid 1 was converted to its p- toluenesulfonate salt (2), which in turn was esterified with PhzCHN2;16-’8 the resultant ester 3 was allowed to react with BOC-G~U(NH,)-ON~’~ to give BOC-Glu(NH&Nr- Pht-Dbu diphenylmethyl ester (4), which was subsequently deesterified to give 5 . The hexapeptide (6) was elongated stepwise to yield the fully protected nonapeptide, N-ZS- Bzl-CysO- Bzl-Tyr-Ile-Glu(NH2)-Nr -Pht-DbuS-Bzl-Cys-Pro- Leu-Gly-NH, (9). Peptide 9 was successfully dephthalylated with either hydrazine or hydrazine acetate” to yield 10, which was fully deprotected by treatment with

was elongated with BOC-

[S-cu,yDiaminobutyric acidloxytocin Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 127

n I

Figure 1. Conformation of oxytocin in solution as proposed by Urry and Walter.g

HF as described by Sakakibara, et al. 23324 The intermediary dithiol was oxidized to the disulfide by action of aq K3Fe(CN), purified by partition chromatography26 using 2 solvent systems.

found to possess 0.03 unit/mg of rat uterotonic activity. It exhibited approximately 0.03 unit/mg of avian vasodepressor activity, less than 0.01 unit/mg of rat pressor activity, and less than 0.001 unit/mg of rat antidiuretic activity. The analog did not inhibit the oxytocin-induced responses in any of these 4 biological test systems. This indicates that the greatly diminished capability of this compound to evoke the responses characteristic of oxytocin resides in its low affinity for the oxytocin receptors. The predominant factor may be that Dbu-in contrast to Asp(NH2)-lacks the carbonyl group in the side chain and thus is unable to form an H bond with the amide NH of the residue in position 8. This enhances the conformational ambiguity of the acyclic tripeptide, which factor in turn also appears to affect the conformation of the 20-membered ring adversely. It is concluded that Dbu, unlike the Asp(NH2) residue, is unable to stabilize the oxytocin conformation, and hence is not an effective substitute for the Asp(NH2) residue in neurohypophyseal hormones.

Experimental Proceduref W-Pht-Dbu*H,O (1). Dbu .2HC1(9.55 g, 0.05 mole) was dis-

solved in a mixt of H,O (100 ml) and 4 N NaOH (37.5 ml), and an aq soh (20 ml) of CuSO,.SH,O (6.25 g, 0.025 mole) was added. Carb~ethoxyphthalimide'~ (13.1 g, 0.06 mole) and NaHCO, (5 g, 0.05 mole) were added to the mixt with strong stirring. After 60 min the Cu 'salt of W-Pht-Dbu was filtered, washed successively with H,O, EtOH, CH,Cl,, and Et,O; yield, 13.0 g (92.2%). The Cu salt was sumended in a soh of EDTA (18.6 K. 0.05 mole) in H.0 (200

[ 5-a,y-Diaminobutyric acid] oxytocin was

Upon bioassay [ 5-qy-diamin~butyric acidloxytocin was

$All reactions were carried out at room temp and all products were dried at room temp over P,O, in vucuo unless otherwise noted. Evaporations were performed under reduced pressure. All melting points were detd with a Thomag-Hoover capillary melting point apparatus and are not corrected. Optical rotations were detd with a Carl Zeiss photoelectric precision polarimeter set at 0.005'. Peptide hydrolysates were chromatographed on a Beckman/Spinco Model 12OC amino acid analyzer, using Beckman custom research resin PA-28. Elementary analyses were carried out by Galbraith Labora- tories, Knoxville, Tenn. Where analyses are indicated only by symbols of the elemehts, analytical results obtd for the elements were within f0.4% of the theoretical values.

ml), and the mixt was stirred vigorously for 60 min and then re- frigerated for 2 hr.''$'* The white ppt was collected by fittration, washed successively with H,O, EtOH, and Et,O; yield, 11.55 g. The crude product was recrystd from boiling H,O (250 ml) contg EDTA (1 g); yield, 7.76 g (58.9%); mp 205-206" dec, [a]"D +14.2" (c 2, 1 N HC1). Anal. (C,,H,,N,O; H,O) C, H, N.

m-Pht-Dbu p-Toluenesulfonate (2). Compd 1 (5.32 g, 20 mmoles) was dissolved in warm H,O (15 ml) together with p-TsOH (3.99 g, 21 mmoles). The soln was concd to form a cryst residue, which was suspended in acetone and collected by filtration after cooling; yield, 7.81 g (92.9%); mp 176-177'dec; [a]'OD +9.2' (c 2.2, MeOH). Anal. (C,,H,,N,O,S) C, N; H: calcd, 4.81; found, 5.33.

of diphenyldiazomethane (3.70 g, 19 mmoles) in DMF (10 ml) was stirred into a soln of 2 (6.31 g, 15 mmoles) in DMF (5 ml) at 50'. After 10 min the reaction mixt was concd, and the resulting oil was crystd from a mixt of EtOAc (10 ml) and Et,O (100 ml). The material was filtered and washed repeatedly with EtOAc; yield, 6.86 g; mp 148-150" dec. Recrystn from CH,CN (100 ml) and then from DMF-EtOAc gave fine needles; yield 4.83 g (54.9%); mp 159-160' dec; [a]'OD +14.3' (c 1.1, MeOH). Anal. (C,,H,,N,O,S) C, H, N.

BOC-Glu(NH,)-m-Pht-Dbu Diphenylmethyl Ester (4). A soln of 3 (2.34 g, 4 mmoles) in DMF (10 ml) was allowed to react with BOC-Glu(NH,)-ONp (1.83 g, 5 mmoles) in the presence of Et3N (0.56 ml) for 24 hr. The reaction mixt was concd to an oily residue which was crystd from Et,O; yield, 2.9 g. The crude product was recrystd from EtOAc; yield, 1.3 g (50.6%); mp 148-150' (sintg 146"); [C~]'~D -12.3' (c 1, DMF). Anal. (C,,H,,N,O,) C, H, N.

mmoles) in a mixt of MeOH (70 ml) and THF (20 ml) was subjected to catalytic hydrogenation using 0.5 g of 10% Pd/C as catalyst. After 3 hr the catalyst was removed by filtration and the filtrate was concd to a residue which was crystd from Et,O. Recrystn from MeOH- Et,O gave 0.83 g (92.1%) of product; mp 170-171' dec; [a]"D +7.5" (c 1, DMF). Anal. (C,,H,,N,O,) C, H, N.

soln of DCI (346 mg, 1.68 mmoles) in DMF (5 ml) was stirred into a soln of S-Bzl-Cys-Pro-Leu-Gly-NH, (573 mg, 1.2 mmoles), 5 (572 mg, 1.2 mmoles), and N-hydroxysuccinimide (276 mg, 2.4 mmoles) in DMF (30 ml) at -22'. The reaction mixt was stirred at -22' for 1 hr and at room temp for 18 hr and then filtered and the filtrate was concd to an oily residue which solidified upon trituration with a mixt of EtOAc-Et,O (1:2); yield, 1.17 g. The crude product was repeatedly pptd from 90% EtOH; yield, 615 mg (54.9%); mp 205- 208' dec; [a]"D -70.8' (c 1, AcOH). Anal. (C,,H,,N,O,,S) C, H, Y.

BOC-lle-Glu(NH,)-~-Pht-DbuS-Bzl-Cys-Pro-Leu-Gly-NH, (7). Compd 6 (562 mg, 0.6 mmole) was treated with F3CC0,H (1 ml) for 30 min. The volatile components were removed and the ppt, which resulted upon trituration of the oil with Et,O, was filtered, washed with Et,O, and dried over NaOH; yield, 570 mg. This salt was dissolved in DMF (3 ml) and allowed to react with BOC-lle- bNpZ9 (317 mg, 0.9 mmole) in the presence of N-methylmorpholine (0.01 ml). After 24 hr a mixt of EtOAc and Et,O (l:l, 50 ml) contg AcOH (0.1 ml) was added to ppt the crude product, which was purified by repptn from 96% EtOH; yield, 395 mg (62.8%); mp 203-206" dec; [ffIz0D -73.7" (c 1, AcOH). Anal.

W-Pht-Dbu Diphenylmethyl Ester p-Toluenesulfonate (3). A soln

BOC-Glu(NH,)-W-Pht-Dbu (5). A Soh of 4 (1.22 g, 1.9

BOC-Glu(NH,)-W-Pht-DbuS-B~l-Cys-Pro-Le~-G~-~z (6). A

Gly-NH, (8). The BOC group of7 (5 25 mg, 0.5 mmold) was removed with F,CCO,H (2.5 ml) as described above. The peptide trifluoroacetate (535 mg) dissolved in DMF (5 ml) was allowed to react with N-BOC-U-Bzl-Tyr-ONpJ0 (370 mg, 0.75 mmole) in the presence of N-methylmorpholine (0.1 ml). After 23 hr EtOAc (SO ml) contg a small amount of AcOH (0.1 ml) was added to the reaction mixt to ppt the crude product; yield, 660 mg. The crude product was purified by boiling with 80% EtOH (50 ml) and collected by filtration after cooling; yield, 563 mg (86.5%); mp 236- 240" dec, [a]"D -54.1' (c 1, AcOH). Anal. (C,,H8,N,,0,,S) C, H, N.

Leu-Gly-NH, (9). The BOC group of 8 (521 mg, 0.4 mmole) was removed with F3CC0,H (2.5 ml) as described in the synthesis of 7. The resultirig trifluoroacetate (525 mg) was dissolved in DMF (5 ml) and was allowed to react with N-Z-S-Bzl-Cys-ONp" (268 mg, 0.6 mmole) in the presence of N-methylmorpholine (0.08 ml) for 20 hr. The crude product (615 mg), pptd with EtOAc (35 ml) contg AcOH (0.1 ml), was purified by boiling with 80% EtOH; yield, 572 mg (93.5%); mp 265-268" dec; [a]"D -42.8' (c 1, DMSO). Anal. (C80H96N1z015SZ) C, H, N.

N-ZS-BzlCys-O-Bzl-Tyr-Ile-Glu~H,)-~-Pht-DbuS-BzlCys-Ro-

128 Journalof Medicinal Chemistry, 1972, Vol. I S , No. 2 Hase, Schwartz, and Walter

N-%S-Bzl-Cyd)-Bzl-Tyr-Ile-Glu(NH,)-Dbu~-Bzl-Cy~~o- Leu-Gly-NH,.H,CCO,H (10). A soln of 1 M hydrazine hydrate in DMSO (0.4 ml) was added to a suspension of 9 (200 mg, 0.13 mmole) in DMSO (3 ml). The reaction mixt was stirred vigorously at 75” for 3 hr. A small amt of insol materials was removed by filtration and AcOH (0.2 ml) was added to the filtrate. The product pptd by addn of H,O (20 ml) was filtered, washed with H,O, and dried; yield, 147 mg (77%); mp 211-216” dec.

In an alternative experiment a soln of 2 M hydrazine acetate in DMF (2 ml) was used for the removal of the phthalyl group of 9 (306 mg, 0.2 mmole) and the reaction was carried out at room temp for 6 hr. The product was pptd with H,O and then applied to a short column of CM-Sephadex (H+ form) in a mixt of DMSO- MeOH-H,O (3:3:1). After washing the column well with H,O, the product was eluted with aq AcOH increasing from 50 to 80%. The eluates were dild with H,O and lyophilized; yield, 155 mg (53%); [cY]~’D -45.6” (c 0.8, DMF). Anal. (C,,H,,N,,O,,S .C,H,O,) C, H, N.

on silica gel G (Merck) in a solvent system of BuOH-AcOH-H,O (4: 1 : l ) , although in the former case a small amount of by-product was detected.

[5-a;y-Diaminobutyric acidloxytocin (1 1). Compd 10 (146 mg, 0.1 mmole) was dissolved in anhyd HF (ea. 15 ml) together with anisole (0.5 ml). The reaction mixt was stirred at 20” for 60 min. HF was evapd with N, to result in a residue, which was dried in a desiccator over NaOH for 3 hr. The dried material was dissolved in 0.1% AcOH (100 ml) and the soln was extd with Et,O (100 ml). The aq layer was passed through a column (3 X 8 cm) of Dowex 1-X2 (acetate form). Effluent and washings of the same solvent were combined (ea. 250 ml). The soln was adjusted to pH 8.2 with 1 N NH,OH and treated with excess 0.01 N K,Fe(CN),. After 15 min the pH was adjusted to 6.5 with 1 N AcOH, and ferro- and excess ferricyanide ions were removed by treatment of the soln with AG3X4 (Bio-Rad) in the CF form. Following the removal of resin the filtrate was concd to ea. 10 ml and applied to a column (3 X63 cm) of Sephadex G-25 (100-200 mesh), which had been equilibrated with the lower phase of BuOH-EtOH-H,O contg 3.5% AcOH and 1.5% pyridine (7:2:9). The column was eluted with the upper phase and 90 9.3-ml fractions were collected. Aliquots from every second fraction were taken for detn of Folin-Lowry color values.3z The fractions corresponding to the principal peak (Rf 0.36) were pooled and dild with excess H,O. The mixt was concd to a small vol and the product was isolated by lyophilization; yield, 50 mg. This material was dissolved in the upper phase (3 ml) of the solvent system BuOH-PrOH-H,O contg 3.5% AcOH and 1.5% pyridine (7:2:9). The soln was applied to the same column of Sephadex G-25 which had been equilibrated with both phases of the solvent system. After elution with the upper phase a single peak (Rf 0.21) was detected by Folin-Lowry color detn. Fractions 71-83 were combined, dild with H,O, concd and, after azeotropic removal of the org phase, lyophilized; yield of [ 5-ol,y-diaminobutyric acid] oxytocin as monoacetate, 19.3 mg (18%). A sample for analysis was dried at 100” for 8 hr; [CY]~’D -34.0” (c 0.6, 1 N AcOH). Anal.

Both methods gave identical product showing an R f 0.55 by tlc

(C,J,,N,,Oi,Sz ’ CzH,Oz) C, H, N. Amino acid analysis33 after hydrolysis in 6 N HCl at 110” for

23 hr gave the following molar ratios with glycine taken as 1.00: Cys (1.02); Tyr (0.89); Ile (0.96); Glu (1.00); Dbu (1.02); Pro (1.02); Leu (1.01);Gly (1.00); NH, (2.04).

R f values on tlc on silica gel G (Merck) with different solvent systems gave the following values: 0.34 with upper phase of BuOH- AcOH-H,O (4:1:5) (0.49 for Tyr); 0.08 with BuOH-AcOH-H,O (4 : l : l ) (0.42 for Tyr); 0.32 with upper phase of BuOH-EtOH4,O contg 3.5% AcOH and 1.5% pyridine (7:2:9) (0.46 for Tyr).

Bioassay Methods. Oxytocic assays were performed on 6 isolated uterine horns from 3 rats in natural estrus according to the method of H ~ l t o n , ~ , modified by M ~ n s i c k ~ ~ with the use of Mg-free van Dyke-Hastings soln as the bathing fluid. Avian vasodepressor assays were performed on 3 conscious chickens according to the procedure employed by Munsick, et ~ 1 . ~ ~ Assays for antidiuretic activity were performed on anesthetized, hydrated Sprague-Dawley male rats according to the method of Jeffers, et al.,” as modified by Sawyer;38 maximal depression of urine flow, in contrast to average duration of the response, was used to measure the antidiuretic activity. Assays were carried out on 6 rats; not more than 6 hormone injections were given to each animal. Rat pressor assays were carried out on 5 atropinized, urethane-anesthetized male rats as described in the United States Pharma~opeia.~’ The biological activities were measured against the USP Posterior Pituitary Refer- ence Standard.

Acknowledgments. The authors are indebted to Mr. J . Bitter, who was involved in the early phase of this work. They are also grateful for the technical assistance rendered by Mrs. I . Mintz and Mrs. A. Silverman.

References St. Guttmann and R. A. Boissonnas, Helv. Chim. Acta, 46, 1626 (1963). V. du Vigneaud, G. S. Denning, S. Dravarek, and W. Y. Chan, J. Biol. Chem., 239,472 (1964). P.-A. Jaquenoud and R. A. Boissonnas, Helv. Chim. Acta. 45. 1601 (1962). R. T. Havran. I. L. Schwartz. and R. Walter. J. Amer. Chem. SOC., 91, 1836 (1969). R. Walter and I. L. Schwartz, J. Biol. Chem., 241, 5500 (1966). R. Walter, J. Rudinger, and I . L. Schwartz, Amer. J. Med., 42, 653 (1967). R. Walter, I. L. Schwartz, J. H. Darnell, and D. W. Urry, Proc. Nat. Acad. Sei. US., 68, 1355 (1971). R. Walter, Protein and Polypeptide Horm., Proc. Int. Symp., 2nd, 1970,181 (1971). D. W. Urry and R. Walter, Proc. Nat. Acad. Sci. US., 68,956 (1971). C. Ressler and V. du Vigneaud, J. Amer. Chem. SOC., 76, 3107 (1954). M. Zaoral and J. Rudinger, Collect. Czech. Chem. Commun., 20,1183 (1955). H. C. Beyerman, J. S. Bontekoe, and A. C. Koch, Reel. Trav. Chim. Pays-Bas, 78,935 (1959). N.I.A. Overweg, I . L. Schwartz, B. M. Dubois, and R. Walter, J. Pharmacol. Exp. Ther., 161, 342 (1968). F. Weygand, D. Hoffmann, and E. Wiinsch, 2. Naturforsch. B, 21,426 (1966). G. H. L. Nefkens, G. I . Tesser, and R. J. F. Nivard? Red. Trav. Chim. Pays-Bas, 79,688 (1960). J . B. Miller, J. Org. Chem., 24,560 (1959). A. A. Aboderin, G. R. Delpierre, and J. S . Fruton, J. Amer. Chem. Soc., 87,5469 (1965). L. I . Smith and K. L. Howard, in “Organic Syntheses,” Col- lect. Vol. 111, Wiley, New York, N. Y., 1955, p. 351 . H. Zahn, W. Danko, and B. Gutte, 2. Naturforsch. B, 21,763 ( 1966). D. A. Kidd and F. E. King, Nature (London), 162, 776 (1948). J. C. Sheehan and V. S. Frank, J. Amer. Chem. Soc., 71, 1856 (1949).

(1962).

Acids, Peptides and Proteins,” B. Weinstein, Ed., Marcel Dekker, New York, N. Y., 1971, p 5 1 .

(24) S. Sakakibara, Y. Shimonishi, Y. Kishida, M. Okada, and H. Sugihara, Bull. Chem. SOC. Jap., 40,2164 (1967).

(25) D. B. Hope, V. V. S. Murti, and V. du Vigneaud, J. Biol. Chem., 237, 1563 (1962).

(26) D. Yamashiro, Nature (London), 201,76 (1964); D. Yama- shiro, D. Gillessen, and V. du Vigneaud, J. Amer. Chem. SOC.. 88, 1310 (1966).

(22) R. Schwyzer, A. Costopanagiotis, and P. Sieber, Chimia, 16, 295

(23) S. Sakakibara, in “Chemistry and Biochemistry of Amino

(27) R. Ledger and F. H. C. Steward, Aust. J. Chem., 18,933 (1965). (28) S. Kuwata and H. Watanabe, Bull. Chem. SOC. Jap., 38,676

(29) S. Hornle, Hoppe-Seyler’s Z. Physiol. Chem., 348, 1355 (1967). (30) H. C. Beyerman, C. A. M. Boers-Boonekamp, and H. M. van

den Brink-Zimmermannovi, Reel. Trav. Chim. Pays-Bas, 87, 257 (1968).

2504 (1959).

J. Biol. Chem., 193,265 (1951).

1190 (1958).

(1965).

(31) M. Bodanszky and V. du Vigneaud, J. Amer. Chem. Soc., 81,

(32) 0. H. Lowry, N. J . Rosebrough, A. L. F a r , and R. J. Randall,

(33) D. H. Spackman, W. H. Stein, and S. Moore, Anal. Chem., 30,

(34) P. Holton, Brit. J. Pharmacol., 3,328 (1948). (35) R. A. Munsick, Endocrinology, 66,451 (1960). (36) R. A. Munsick. W. H. Sawyer, and H. B. van Dyke, ibid., 66, . .

860 (1960). f37) W. A. Jeffers. M. M. Livezey, and J. H. Austin, Proc. SOC. Exp. . I

Biol. Med., 50, 184 (1942): (38) W. H. Sawyer, Endocrinology, 63,694 (1958). (39) “The Pharmacopoeia of the United States of America,” 17th

rev, Mack Publishing Co., Easton, Pa., 1965, p 749.


Synthesis and antiviral activities of adamantane spirocompounds. 1. Adamantane and analogous spiro-3'-pyrrolidines

K. Lundahl, J. Schut, J. L. M. A. Schlatmann, G. B. Paerels, and A. PetersJ. Med. Chem., 1972, 15 (2), 129-132• DOI: 10.1021/jm00272a003 • Publication Date (Web): 01 May 2002






Adamantane Spiro Compounds. I Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 129

Synthesis and Antiviral Activities of Adamantane Spiro Compounds. 1. Adamantane and Analogous Spiro3’-pyrrolidines

K. Lundahl, J. Schut, J. L. M. A. Schlatmann, G . B. Paerels,* and A. Peters

N. V. Philips-Duphar Research Laboratories, Weesp, The Netherlands. Received May 1, 19 71

Adamantanespiro-3’-pyrrolidine and several N-substituted derivatives were synthesized. In particular smaller alkyl-substituted derivatives displayed interesting antiviral action against influenza A, parainfluenza Sendai, Coxsackie A2 1, and rhinovirus. This activity was compared’with that of spiropyrrolidines derived from other alicyclic compounds. The N-methyl adamantanespiro-3 -pyrrolidine is superior to l-adamantanamine in level and spectrum of activity.

In view of the interesting antiviral properties of l-adamantanamine’ it was deemed worthwhile to study the influence of structural modifications on antiviral activity. The discovery of a suitable method for the synthesis of 2-ada- mantanone2 furthered the possibilities of substitution at a bridge atom of the adamantane nucleus and made the synthesis of secondarily substituted adamantanes, including spiro compounds, feasible. This paper deals with the synthesis and antiviral activity of a number of adamantanespiro-3’-pyrrolidines. The importance of the adamantane nucleus for antiviral activity was investigated by comparison with spiropyrrolidines derived from bornane, bicyclo [3.3.1]- nonane, perhydro-4,7-methanoindane, and cyclohexane. The synthesis and antiviral properties of related adamantane spiro heterocyclic compounds will be the subject of a further paper.3

cyanoacetic ester using Cope conditions4 to give 2 (Scheme I). Addition of HCN to 2 in strongly alkaline medium proceeded smoothly and gave, with subsequent saponification and decarboxylation, 2-cyano-2-cyanomethyladamantane (3). Hydrolysis of 3 afforded the adamantanespiro-2’-succinic anhydride (4) which was converted by treatment with NH3 into adamantanespiro-3’-pyrrolidine-2,5-dione (Sa). Compd Sa could also be synthesized by treatment of the dinitrile 3 with HBr and hydrolysis of the reaction product5 (Scheme 11). Reaction of 4 with primary amines gave the corresponding substituted pyrrolidines diones Sb-Sf (Table I). The adamantanespiro-3 -pyrrolidines 6a-6f (Table 11) were synthesized by reduction of the corresponding diones with LAH. Catalytic reduction of 3 followed by pyrolysis of the diamine dihydrochloride6 7 led to an alternate synthesis of 6a (Scheme 111). Compd 6h was formed by acetylation of 6a and subsequent reduction; 6i was formed by alkylation of 6a with acetone and subsequent reduction. Compd 6g was obtained by treatment of 6f with S0C12. The synthesis of the spiropyrrolidine derivatives of bornane, bi- cycle [3.3.l]nonane, perhydro-4,7-methanoindane, and cyclohexane has been achieved by the same methods (Table 111).

against Influenza A2 Japan are presented in Tables I1 and 111. In vitro data (Table IV) were obtained by the plaque inhibition method.’ The compounds were added just before virus inoculation. In vivo experiments were carried out using mice (Swiss SPF) weighing 19-21 g each. The substances were administered orally twice a day for 5 days. The dose was 5.1 0-4 mole/kg per day or 1.1 0-4 mole/kg per day in comparative experiments with I-adamantanamine? Virus challenge (by aerosolation) took place between the two administrations of the compounds on the first day of the experiment.

Synthesis. 2-Adamantanone (1) was condensed with

Biological Results. The in vivo antiviral activities

Scheme I COOC,H,

NCCH,COOC,H HCN NaOAc, HOAc

1 2

3 4

Sa, R = H b. R = CHI

6a, R = H b. R = C H ,

C; R = CH;CH=CH*

e , R = CHzC& e, R=CH2C6H5

c; R = CH;CH=CH, d, R = C 6 H i i d, R = C J I i i

f, R = CHZCHZOH f, R = CHzCHzOH

Scheme I1

L 3

Sa

Discussion

The antiviral activity of theN-methyl derivative 6b has been the subject of a previous communi~ation.~ Compared with 1-adamantanamine this compound is about 3 times more active in vivo against Influenza A2 Japan and A2 Hong Kong. In vitro it shows a broader antiviral spectrum: 6b was demonstrated to be active against Coxsackie A2 1 and Rhino 2 (HGP) in qualitative (Table IV) and quantitative experiments: whereas 1-adamantanamine was inactive. The adamantanespiro-3’-pyrrolidines which have smaller alkyl sub-

130 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Paerels, et al.

Table I. Adamantanespiro-3'-pyrrolidine-2',5 '-dionesa Yield,

No. R Formula Analyses Mp, "C %

Sa H C,,H,,NO, C, H, N, 0 187-189 91 b CH, C,,H,,NO, C, H, N, 0 214-216 85

e CH,C,H, C,,H,,NO, H, N, 0; Cb 138-138.5 90 f CH,CH,OH C,,H,,N03 C, H, N, 0 144-145 63 'Prepd by procedure A described in the Experimental Section.

c CH,CH=CH, C,,H,,NO, C, H, N 84.5-87.5 81 C,,H,,NO, C, H, N, 0 142.5-144 70

The synthesis of Sa has been described separately. "c: calcd, 77.63; found, 77.13.

Table 11. Adamantanesuiro-3 "molidines'

Scheme 111

3 7

6a

No. R Formula

~ ~~ ~ ~~

An tiviral Analyses MP, "C Yield, % act. in vivob

6a H C,,H,,N. HCl C, H, N, C1 b CH3 C,,H,,N hydromaleate C, H, N, 0 h C2H5C C,,H,,N. HC1 C, H, N i i-C3H7C C,,H,,N. HCI C, H, N, C1 C CH,CH=CH, Cl6HZ5N. HCl C, H, N, C1 d C,,H3,N. HCI C, H, N e CH2C6H5 C,,H,,N. HCl e f CH,CH,OH C,,H,,NO * HC1 C, H, N, 0, C1 g CH,CH,CIC C,,H,,ClN~ HCl H, N; ~f

253-254

266-268 273-276

262-264 302-306 236-237.5

144-145d

256-260 dec

254 dec

77 92 87 60 83 47 85 80 94

++ ++ ++ + + -

+ + +

'Prepd by procedure B described in the Experimental Section unless otherwise stated. b++ = activity comparable to that of l-adamanta- mine or better; + = significant activity but less than that of 1-adamantanmine; - = inactive. qrocedure given separately in the Experimental Section. dMp of hydrochloride, 266-267.5'. eAnalysis for picrate (mp 171-171.5"): C, H, N, 0. fC: calcd, 62.06; found, 61.63.

Table 111. Analogous Spiro-3'-pyrrolidines

An tiviral No. Structure Formula Analyses MP, "C Yield, % act. in vivoa

8

9

10

11 m NH ' HCIC

HCl

C,,H,,N. HCl

C,,H,,N. HCI

C,,H,,N. HCI

C,H,,N. HCl

246-247

219-221

186-188

83-8Sd

90

81.5

70.5

+

++

+

"See footnote b, Table 11. bN: calcd, 5.74; found, 6.18. %ee reference 8. dpicrate mp 148-149", litc mp 151".

stituents generally are comparable to 6b in level and spectrum of activity (Tables I1 and IV). Increasing size of the substituents at the N-atom tends tQ diminish in vivo activity against Influenza A2 Japan. A similar trend has been observed in a series of bicyclo[2.2.2]octan- and -oct-2-ena- mines." In general cytotoxicity is found with 6d-6g, making an estimation of antiviral activity in vitro impossible in the procedure used.

Of the analogous spiro-3'-pyrrolidines activity is highest for 9 structurally most closely related to adamantane.

Nevertheless 9 is less active than 6b, but comparable to 1-adamantanamine. The cyclohexane derivative 11 does not in fact possess any activity at all. This stresses the importance of the adamantane nucleus for antiviral activity in this series.** In general no activity was found against B strains of Influenza neither in vitro nor in vivo.

**Note Added in Roof. After completion of this manuscript, the diamantanespiro-3'-pyrrolidine and its N-methyl derivative were prepared. Both compounds were less active in vivo against Influenza A, Japan than 6b.

Adamantane Spiro Compounds. I Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 131

Table IV. In Vitro Antiviral Properties of the SDir0-3'-uyrrolidines Hydrochloridesa

Influenza Parainfluenza Coxsackie Rhino No. A, Japanb SendaiC A,ld 2(HGPId

6a + T - ++ 6b ++ + + ++ 6h + + f ++e 6i +e i. f T 6c ++e - +e T 8 f + f ++e 9 +e T f -

10 + - T T 11 - 1-AA ++ f

a++ = >70% inhibition; + = 40-70% inhibition; f = 20-40% inhibition; - = <20% inhibition; T = cytotoxic. Concn of compounds 10-4M. chick embryo fibroblasts. CIn calf kidney cells. M. HeLa cells. eSlightly toxic. fl-AA = 1-adamantanamine.

- f f - -

Experimental Section? Ethyl adamant-2-ylidenecyanoacetate (2) was prepd from 2-

adamantanone according to Cope, er al .:4 yield, 90%; crystd from EtOH; mp 81-82'. Anal. (C,,H,,NO,) C, H, N, 0.

2Cyano-2-cyanomethyladamantane (3). 2 (100 g, 0.41 mole) was dissolved in EtOH (800 ml); a soln of KCN (64 g, 0.98 mole) in H,O (140 ml) was added. The mixt was stirred and kept at 65" for 16 hr. After cooling, the ppt was filtered with suction. By concn of the filtrate some further crops were obtained. The crystals were stirred in a 0.2 N KOH soln (500 ml), filtered, washed with H,O, and dried: yield, 77.5 g (96%); mp 126-127". Anal. (C1,Hl6NZ) C, H, N.

Adamantanespiro-2'-succinic Anhydride (4). 3 (76 g, 0.38 mole) was dissolved at 90" in concd H,SO, (750 ml). The temp rose to approx 110". The soln was shaken for 5 min and poured on ice (approx 10 1.). The mixt was neutralized with 50% NaOH, and the ppt was filtered with suction and washed with H,O. The moist substance was heated, with stirring, in concd HCl (2.1 1.) for 2 hr at 80". After cooling, the solid was filtered off, washed, and crystd from C,H,: yield, 5.6 g (76%); mp 226-229'. Anal. (C,$I160,) C, H, 0.

Adamantanespiro-3'-pyrrolidine-2',5'dione (Sa). 4 (5 2.1 g, 0.24 mole) was melted and heated at 230" for 3.5-4 hr, NH, being passed in continuously (4 l./hr). After cooling, the residue was crystd from EtOH (0.5 1.) to give Sa (see Table I).

(5) . 4 (33 g, 0.15 mole) was dissolved in hot C,H, (50 ml) and with vigorous stirring RNH, (0.15 mole) in C,H, (30 ml) was added at 60-65". After standing for 0.5 hr at 60-65", the mixt was cooled, and the ppt was filtered and washed (C,H,). A further crop was obtained by concn of the mother liquor. The solid carbox- amidecarboxyadamantane was heated under N, for 15-30 min at 220-225'. After cooling, the residue was crystd from EtOH. A further crop was obtained by addn of H,O. Yields and properties are given in Table I.

General Procedure B. Adamantanespiro-3'-pyrrolidines (6). 5 was reduced with LAH in refluxing THF according to known methods." The bases were converted into the hydrochlorides and crystd from EtOH. Yields and properties are given in Table 11.

Adamantanespiro-3'-pyrrolidine-2',Sfdione (Sa). Procedure of Scheme II.$ 3 was converted into Sa according to ref 5 : yield, 93%.

Adamantanespiro-3'-pyrrolidine Hydrochloride (6a). Procedure of Scheme 111. 3 (10 g, 0.05 mole) in EtOH (250 ml) and concd HCl( l0 ml) was hydrogenated with PtO, (1 g) at 50". After 6-8 hr the uptake of H, ceased. The catalyst wos filtered off and a colorless soln resulted in which 6a and the diamine 7 were present ac-

General Procedure A. Adamantanespiro-3'-pyrrolidine-2',S'-dbnes

?Melting points were measured in closed capillary tubes in an electrically heated aluminum block. Temperature was indicated by a chromel-alumel couple on a Philips G. M. 6020 tube voltmeter. Ir, nmr, and mass spectroscopic data are fully in accord with the structures proposed. Microanalyses were performed by A. Bernhardt, Mikroanalytisches Lab., Elbach uber Engelskirchen, W. Germany, and F. Pascher, Mikroanalytisches Lab. Bonn, W. Germany. Where analyses are indicated only by symbols of the elements, analytical results of those elements were within *0.4% of the theoretical values.

$This synthesis was performed by Mr. J . S. Bontekoe.

cording to tlc.8 The solvent was evapd and the resulting residue was heated under N, at 320" for 5-10 min. After cooling the reaction product was converted into the free base by treatment with 2 N NaOH and was extd with CH,Cl,. The solvent was evapd, the residue dissolved in MeOH (30 ml) and the soln acidified with concd HCl(5 ml). 6a crystd after addn of Et,O: yield, 7.2 g (63%).

N-Ethyladamantanespiro-3'-pyrrolidine Hydrochloride (6h). The free base of 6a was acylated with Ac,O in a usual way. The acetamide was reduced with LAH in refluxing THF in 16 hr: yield, 87%; see Table 111.

N-Isopropyladamantanespiro-3'-pyrrolidine Hydrochloride (6i). The free base of 6a (2.8 g, 0.015 mole) was dissolved in EtOH (10 ml). Me,CO (4 ml, 0.05 mole) and PtO, (3.4 g) were added, and the mixt was hydrogenated overnight at 4.2 kg/cmz. After removal of the catalyst the soln was acidified with 1.4 N dry HC1 in EtOH (12 ml) and evapd. Crystn (EtOH-Et,O) gave 2.45 g of 6i; see Table 11.

N-p-Chloroethyladamantanespiro-3'-pyrrolidine Hydrochloride (6g). 8 (7 g, 0.026 mole) was dissolved in SOC1, (14 ml) and refluxed for 30 min. The SOC1, was then evapd and C,H, (3 X 20 ml) was added and evapd to 7.02 g of 6g; see Table 11.

acetate" was converted into 2-cyano-2-cyanomethylbornane according to the procedure described for 3: yield, 67%; mp 152-153". Anal. (C,,H,,N,) C, H, N.

2-Carboxy-2-carboxymethylbomane. Following the procedure used for the prepn of 4,2-cyano-2-cyanomethylbornane (10.5 g, 0.052 mole) was converted into 5.4 g (61%) of 2-carboxy-2-carboxy- methylbornane: mp 138-140". Anal. (C13~, ,0 , ) C, H; 0: calcd, 26.63; found, 26.03.

methylbornane (1 1 g, 0.045 mole) was heated at 180" for 20 min under N,. After the evoln of gas had ceased the product was distd in vacuo to yield 7.65 g of the anhydride as a waxy substance.

N-Methylbornane-2-spiro-3'-pyrrolidine-2',5'dione. Prepd according to method A: yield, 51%; mp 86.5-87". Anal. (C,,H,,NO,) C, H, N; 0: calcd, 13.60; found, 14.01.

N-Methylbomane-2-spiro-3'-pyrrolidine hydrochloride (8) was prepd according to procedure B (see Table 111).

Ethyl Bicyclo [ 3.3.l]non-9-ylidenecyanoacetate. Bicyclo- [3.3.1]nonan-9-0ne'~ (13.2 g, 0.096 mole) was condensed with ethyl cyanoacetate by the method used for the prepn of 2. The resulting oil was purified by dry column c h ~ o m a t o g ' ~ (Merck silica gel 0.05-0.2 mm; column 5 X 150 cm) with C6H6 yielding 10.3 g (46%) of cryst product: mp 50-51". Anal. (C,,H,,NO,) H; C: calcd, 72.07; found, 72.57.

9-Cyano-9-cyanomethylbicyclo [ 3.3.1 Inonane. According to the procedure used for the prepn of 3, ethyl bicyclo[ 3.3.llnon-9- ylidenecyanoacetate was converted into the corresponding dinitrile: yield, 96%; mp 135.5-136'. Anal. (C,,H,,N,) H; C: calcd, 76.55; found, 75.68; N: calcd, 14.88; found, 14.34.

Following the procedure of scheme 111 for the prepn of 6a, 9 was obtained from 9-cyano-9-cyanomethyl bicyclo[ 3.3. llnonane (see Table 111).

Ethyl Perhydr0-4,7-methanoind-S-ylidenecyanoacetate. Per- hydro-4,7-methanoindan-5-one# (30 g, 0.2 mole) was condensed with ethyl cyanoacetate as in the prepn of 2. The resulting red oil was purified by dry column chromatogI4 (Merck silica gel 0.05-0.2 mm, column 5 X 150 cm) with C,H, yielding 18.8 g (38%) of a colorless oil.

by the method used for the prepn of 3: yield, 64%; mp 99-99.5". Anal. (C,,H,,N,) C, H, N.

Perhydro-4-7-methanoindane-S-spiro-3'-pyrrolidine Hydro- chloride (10). By the method of Scheme 111 for the prepn of 6a, 10 was obtd from 5-cyano-5-cyanomethyl perhydro-4,7-methanoindane (see Table 111).

Plekkenpol for cooperation in the synthesis of several compounds, to Mr. F. W. van Deursen and Miss M. E. van der Heeden for measurements and interpretation of the spectra, and to Miss J. C. Sekbt for supplying the antiviral in vitro results.

2-Cyano-2-cyanomethylbornane. Ethyl born-2-ylidenecyano-

Bomane-2-spiro-2'-succinic Anhydride. 2-Carboxy-Zcarboxy-

Bicyclo [ 3.3.l]nonane-9-spiro-3'-pyrrolidine Hydrochloride (9).

5-Cyano-5-cyanomethyl perhydro-4,7-methanoindane was obtd

Acknowledgments. We are indebted to Mr. A. H. D.

fjMerck, tlc-plates silica gel F 254, solvent CH,Cl,-EtOH-25%

#Obtained from Aldrich Chemical Co. NH,OH (3:3:1 v/v).

132

References

Journal oj Medicinul Chemistry, 197.2, Vol. 15, No. 2 van Hes. et al.

(8) H. Najer, R. Giudicelli, and J. Sette, Bull. Soc. Chzm. Fr., 2572 (19641.

\ - - - .,- - _ -

W. L. Davies, K. R. Grunert, R. F. Haff, J. W. hlcGahen, E. M . (9) A . Peters, C. A. de Bock, G. B. Paerels, and J. L. M. A. Schlat- Neumayer, M. Paulshock, J . C. Watts, 'T. R. Wood, E. C. mann, Proc. International Congr. Chemother., 1969,11, A14, Hermann, and C. E. Hoffmann, Science, 144,862 (1 964). 71 (1970). H. W. Geluk and J. L. M. A. Schlatmann, Tetrahedron, 24, 5361 (10) J. G. Whitney, W. A. Gregory, J. C. Kauer, J. R. Roland, J . A. (1968). Snyder, R. E. Benson, and E. C. Hermann, J. Med. Chem., 13. R. van Hes, A . Smit, T. Kralt, and A. Peters, J. Med. Chem., 15, 254 (1970). 132 (1972). A. C. Cope, C. M. Hofmann, C. Wyckhoff, and E. Harden- J. Med. Chem., 12,126 (1969). bergh, J. Amer. Chem. Soc., 63, 3452 (1941). W. A. Nasutavicus, S. W. Tobey, and F. Johnson, J. Org. 15,381 (1950). Chem., 32,3325 (1967). A. Ladenburg, Ber., 20,442 (1887). (1968). A. Peters, and W. Th. Goedemans, Arch. Ges. Virusforsch., 19, 190 (1966). (1967).

(1 1) L. M. Rice, B. S. Sheth, K. R. Scott, and C. F. Geschickter,

(12) E. J. Cragoe, Jr., C. M. Robb. and J. M. Sprague, J. Org. Cheni.,

(13) B. Fell, W. Seide, and F. Asinger, Tetrahedron Lett., 1003

(14) B. Loev, and M. M. Goodman, Chem. Ind. (London), 2026

Synthesis and Antiviral Activities of Adamantane Spiro Compounds. 2

R. van Hes,* A. Smit, T. Kralt, and A. Peters


The synthesis of adaniantanespiro-3'-piperidine (XV) and adamantanespiro-4'-perhydroazepine (XVIII) is reported. These compounds can be prepared from the key intermediate 2-bromomethyl- 2-(&bromoethyl)adamantane (XII). With the latter several other reactions have been carried out, leading to heterocyclic spiro compounds. The antiviral activity of these compounds is discussed.

The synthesis and antiviral properties of adamantanespiro- 3'-pyrrolidine and derivatives have been described in a previous paper.' Because of the strong antiviral activities found No e. G C : O H CH COOCH, in the pyrrolidine series, we continued our investigations in this field by preparing 6- and 7-membered ring analogs.

Scheme 11

Synthesis. We first tried to obtain the 6-membered analog by an analogous route as for the adamantanespiro-3'-pyrroli- I11 dine.' For that we had to prepare 2,2-bis(cyanomethyl)- adamantane. Treatment of I with malononitrile did not give

mantylidenemalononitrile in high yield (Scheme I). We car- a Michael reaction but instead led to the formation of ada- ___, LAH FzoH Scheme I

CN IV V

VI ried out the reaction under various conditions, also with cyanoacetic ester as reagent, but failed to obtain the desired condensation product. A possible explanation may be steric hindrance. Following Scheme I1 it was possible to prepare the 6-membered ring analog.

Treatment of adamantanespiro-2'-succinic anhydride' with MeOH gave the half-ester 111. Reduction with LAH in

be converted into the lactone V with TsOH. Treatment of V with KCN in DMSO resulted in the formation of VI, which could be converted into the diacid by hydrolysis. Refluxing the diacid with Ac20 gave the anhydride VI1 in high yield. The cyclic imide VI11 was obtained by treatment of the anhydride with MeNH2 in C&IH,. Finally reduction with LAH gave the spiro compound IX. Starting with adamantanespiro-2'-succinic anhydride' we were able to prepare ada- mantanespiro4'-perhydroazepine XVIII and adamantanespiro-3'-piperidine XV via another route (Scheme 111). Re-

~ o z F J - J p =

= a, 0 CH,

Et20 yielded the hydroxycarboxylic acid IV, which could VI1 VI11

IX

duction of the anhydride with LAH in THF gave the diol X as well as 12% of the hydroxy acid IV. Treatment of the diol X with 48% HBr or with PBr3 did not produce the de-


Synthesis and antiviral activities of adamantane spiro compounds. 2R. Van Hes, A. Smit, T. Kralt, and A. Peters







132

References

Journal oj Medicinul Chemistry, 197.2, Vol. 15, No. 2 van Hes. et al.

(8) H. Najer, R. Giudicelli, and J. Sette, Bull. Soc. Chzm. Fr., 2572 (19641.

\ - - - .,- - _ -

W. L. Davies, K. R. Grunert, R. F. Haff, J. W. hlcGahen, E. M . (9) A . Peters, C. A. de Bock, G. B. Paerels, and J. L. M. A. Schlat- Neumayer, M. Paulshock, J . C. Watts, 'T. R. Wood, E. C. mann, Proc. International Congr. Chemother., 1969,11, A14, Hermann, and C. E. Hoffmann, Science, 144,862 (1 964). 71 (1970). H. W. Geluk and J. L. M. A. Schlatmann, Tetrahedron, 24, 5361 (10) J. G. Whitney, W. A. Gregory, J. C. Kauer, J. R. Roland, J . A. (1968). Snyder, R. E. Benson, and E. C. Hermann, J. Med. Chem., 13. R. van Hes, A . Smit, T. Kralt, and A. Peters, J. Med. Chem., 15, 254 (1970). 132 (1972). A. C. Cope, C. M. Hofmann, C. Wyckhoff, and E. Harden- J. Med. Chem., 12,126 (1969). bergh, J. Amer. Chem. Soc., 63, 3452 (1941). W. A. Nasutavicus, S. W. Tobey, and F. Johnson, J. Org. 15,381 (1950). Chem., 32,3325 (1967). A. Ladenburg, Ber., 20,442 (1887). (1968). A. Peters, and W. Th. Goedemans, Arch. Ges. Virusforsch., 19, 190 (1966). (1967).

(1 1) L. M. Rice, B. S. Sheth, K. R. Scott, and C. F. Geschickter,

(12) E. J. Cragoe, Jr., C. M. Robb. and J. M. Sprague, J. Org. Cheni.,

(13) B. Fell, W. Seide, and F. Asinger, Tetrahedron Lett., 1003

(14) B. Loev, and M. M. Goodman, Chem. Ind. (London), 2026

Synthesis and Antiviral Activities of Adamantane Spiro Compounds. 2

R. van Hes,* A. Smit, T. Kralt, and A. Peters


The synthesis of adaniantanespiro-3'-piperidine (XV) and adamantanespiro-4'-perhydroazepine (XVIII) is reported. These compounds can be prepared from the key intermediate 2-bromomethyl- 2-(&bromoethyl)adamantane (XII). With the latter several other reactions have been carried out, leading to heterocyclic spiro compounds. The antiviral activity of these compounds is discussed.

The synthesis and antiviral properties of adamantanespiro- 3'-pyrrolidine and derivatives have been described in a previous paper.' Because of the strong antiviral activities found No e. G C : O H CH COOCH, in the pyrrolidine series, we continued our investigations in this field by preparing 6- and 7-membered ring analogs.

Scheme 11

Synthesis. We first tried to obtain the 6-membered analog by an analogous route as for the adamantanespiro-3'-pyrroli- I11 dine.' For that we had to prepare 2,2-bis(cyanomethyl)- adamantane. Treatment of I with malononitrile did not give

mantylidenemalononitrile in high yield (Scheme I). We car- a Michael reaction but instead led to the formation of ada- ___, LAH FzoH Scheme I

CN IV V

VI ried out the reaction under various conditions, also with cyanoacetic ester as reagent, but failed to obtain the desired condensation product. A possible explanation may be steric hindrance. Following Scheme I1 it was possible to prepare the 6-membered ring analog.

Treatment of adamantanespiro-2'-succinic anhydride' with MeOH gave the half-ester 111. Reduction with LAH in

be converted into the lactone V with TsOH. Treatment of V with KCN in DMSO resulted in the formation of VI, which could be converted into the diacid by hydrolysis. Refluxing the diacid with Ac20 gave the anhydride VI1 in high yield. The cyclic imide VI11 was obtained by treatment of the anhydride with MeNH2 in C&IH,. Finally reduction with LAH gave the spiro compound IX. Starting with adamantanespiro-2'-succinic anhydride' we were able to prepare ada- mantanespiro4'-perhydroazepine XVIII and adamantanespiro-3'-piperidine XV via another route (Scheme 111). Re-

~ o z F J - J p =

= a, 0 CH,

Et20 yielded the hydroxycarboxylic acid IV, which could VI1 VI11

IX

duction of the anhydride with LAH in THF gave the diol X as well as 12% of the hydroxy acid IV. Treatment of the diol X with 48% HBr or with PBr3 did not produce the de-

Antiviral Adamantane Spiro Compounds. 2 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 133

Scheme I11

X XI XI1

(CH,),NH,. HCl 1.KOH fl 2. HCI CH,Br ___c

- PtO,-H,

I XI11

“““i MC

1 - NH ‘ HCI ’-?7

XIV XV

(CH,),NH, . HCl

CH,CH,NH; HCl - NH . HCl

Pt 0 ,-H

XVII XVIII A V 0

XIX

V

sired dibromide XI1 but the ether XI. The latter could be converted into the dibromide XI1 with Ph3P and Brz in benzonitrile.2 The reaction of the dibromide with NaCN (technical grade) in methyl Cellosolve yielded a mixture of two products, the bromocyano compound XI11 in about 30% yield and the dicyano compound XVI in 50% yield, because of the slight reactivity of one of the Br atoms (neo- pentyl structure) of the dibromide. The 2 compounds could be separated by chromatography. Refluxing of the dibromide with dried NaCN then led to the spiro compound XXI by a Thorpe-Ziegler ring closure of the intermediate dicyanide XVI (see also “Reactions with the Dibromide XII”). The bromocyano compound XI11 could be reduced with PtOz- HZ in EtOH-HCI to XIV, which cyclized to XV upon treatment with KOH in n-BuOH.

The 7-membered ring analog XVIII could be obtained from XVI by two different methods. Reduction of XVI with Pt02-Hz in EtOH-HCI resulted in the formation of XVII. Heating of the latter at about 290” caused ring closure; XVIII was obtained in low yield (16%). In the other method3 the dicyanide was treated with HBr-EtzO, followed by hydrolysis, to give the imide XX. Reduction with LAH afforded the desired compound XVIII.

Reactions with the Dibromide XII. As mentioned above, the dibromide could be cyclized to XXI by treatment with NaCN in methyl Cellosolve in a low yield. If the reaction is performed in DMSO4 with NaH, the results are much better (Scheme IV).

Treatment of a solution of the dibromide XI1 in DMSO with NaCN, followed by NaH, gave the cyclic product XXI in a yield of 75%. The yield of dinitrile, isolated in the absence of the NaH step, was about 90%. XXI on hydrolysis5 was converted in high yield to the ketone XXII. Reaction of XXII with HzNOH resulted in the formation of XXIII.

xx

By reduction of XXIII with Na in EtOH the amine XXN was obtained.

The dibromide could be cyclized with NazS in DMF to XXV, which with Me1 could be converted to the sulfonium iodide XXVI in good yield. Treatment of the dibromide with primary amines resulted in the formation of spiropyrrolidine compounds (XXVII, XXVIII). The latter com-

Scheme IV

CH,Br CH,CH,Br __c NaCN DMSO [0~~~~~2cN] D XI1 XVI

I HzNoH

2. 1. HCI N ~ - C , H ~ O H ~ ~ ~ ,

NH, . HCl

XXIV XXIIl

134 Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 van Hes, et (11

Scheme V

XXVI /

/

/

/

NCH,CH,NH, - H,NCH,CH,NH,

XXVIII

XXIX

idine derivative XV is about the same as that of I-adamantanamine. Substitution at the N atom by Me (IX) reduces the activity. The perhydroazepine compound XVIII is about twice as active as l-adamantanamine, but it still is less active than corresponding pyrrolidine compound. The cyclopentane derivative XXIV has an activity of the same order as that of the N-methyladamantanespir0-3'-pyrrolidine.' Change of the basic function to a sulfonium or hydrazino group (XXVI: XXX) destroyed, or reduced, respectively, the in vivo activity.

Experimental Section (In collaboration with Mr. H. Heym)

Melting points were determined in open capillary tubes in a Buchi apparatus and are uncorrected. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within t0 .49 of the theoretical values. Analyses were carried out in the laboratory of A. Bernhard, Elbach uber Engelskirchen, West Germany, and in the laboratory of Dornis und Kolbe, Mulheim a.d. Ruhr, West Germany. Ir and mass spectra were recorded, respectively, on a PE 337 and a MS 9 AE 1 spectrometer. Nmr spectra were measured on a Varian HA-100 instrument (Me,$). The recording and interpretation of the spectra were carried out under the supervision of Mr. ['. W. van Deursen.

2.41 g of malononitrile in 10 ml of pyridine was stirred at room temp for 23 hr. The reaction mixt was poured into 250 ml of 2 iV HCI. The ppt was ffltered and crystd from EtOH: yield 1.5 g (94%): mp 180-182". The compd was identical with an authentic sample.

2-Carboxy-2-methoxycarbonylmethyladamantane (111). A m ixt of 15.62 g of adamantanespiro-2'-succinic anhydride,' 250 ml of MeOH, and 15 ml of pyridine was boiled for 3 hr, after which the solvent was evapd in vacuo and the residue dissolved in 2 N KOH. The alkaline s o h was washed with Et,O and then acidified with 2 N HCl. The acid s o h was extd with Et,O: yield 16 g (90%) after

Adamantylidenemalononitrile (11). A mixt of 2 g of I and

Scheme VI

xv

i xxx

XXXIV XXXII

XXXI

1 H,-PtO,

XXXIII

pound was converted to the salicylideneimide XXIX (Scheme V).

The reaction of the dibromide XI1 with H2"H2 * H 2 0 gave 2 crystalline products, XXX and XXXI (Scheme VI). The ir spectrum of XXX in CC14 displayed bands at 3380, 3200, and 3135 cm-', characteristic6 of an >NNH2 group, and excluded the other possible isomer XXXIV. This also followed from the reaction of XXX with salicylaldehyde to a Schiff base. The structure of XXXI was proved by catalytic hydrogenolysis which led to the formation of adamantanespiro-3'-pyrrolidine,

Antiviral Properties. The in vivo antiviral activities against influenza A2 Japan are present in Table I. In in vitro experiments these compounds generally appeared to be cytotoxic except the two piperidine derivatives IX and XV. Esti- mation of in vitro antiviral activities was therefore not possible by the used method. The in vivo activity of the piper-

crystn from Et,O; mp 101-104"; equiv wt, found 247.5; calcd 252. 2-Carboxy-2-(p-hydroxyethyl)adamantane (IV). A mixt of

1.81 g of 111, 900 mg of LAH, and 90 ml of Et,O was refluxed 10 min, cooled, decompd with 2 N H,SO,, and extd with Et,O. The Et,O soln was extd with 2 N KOH, and then the alkaline s o h wa, acidified. The ppt was filtered and crystd from Me,CO-hexane: yield 1.28 g (80%); mp 188-191"; equiv wt, found 223; calcd 224. After evapn of the Et,O s o h and crystn of the residue from hexane, 2-hydroxymethyl-2-(phydroxyethyl)adamantane (X) was isolated as a by-product: yield 0.19 g (13%);mp 108-110".

2'-Oxoadamantanespiro-3'-tetrahydrofuran 0. A suspension of IV (9 g) and 50 mg of TsOH in 150 ml of C,H, was refluxed and the H,O formed was removed by distn. After cooling, the C,H, soln was dried (K,CO,) and evapd. Crystn from Me,CO-hexane afforded pure product: yield 7.8 g (94%); mp 174-177'; mass spectrum, M' 206; nmr spectrum (CDCl,), 2 triplets centered a t 82.34 and 4.17, respectively (CH,CH,O); ir spectrum (KBr); 1760 cm-' (7-lactone).

2-Carboxy-2-(/3-cyanothyl)adamantane (VI). A mixt of 5.5 g of KCN and 5.54 g of V in 110 ml of dry DMSO was refluxed 4 hr, cooled, and poured into 500 ml of H,O. After acidification with

Antiviral Adamantane Spiro Compounds. 2 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 135

Table I. Adamantane Spiro Compounds

No. A Formula Analyses An tiviral

act. in vivo" --

Cl,H, ,~N C, H, a, N + 0 0

IX NCH, . HC1

C14H24aN c, H, a, N ++ xv NH . HC1

C15H26CIN Cp H, N c NH . HCl XVIII

c NH, . HC1

XXIV C ,,H,,ClN~ 0.25 H,O C, H, a, N

XXVI c14H231s

XXX '1 3H23CIN2

++

++

"++ = activity comparable to that of 1-adamantanamine or better; + = significant activity but less than that of 1-adamantanamine; - = inactive. For test conditions see previous paper. k : calcd, 70.4; found, 69.7.

2 N HCl the ppt was filtered and crystd from C,H,: yield 6 g (96%); mp 150-152";equiv wt, found 232;calcd 233.

2-Carboxy-2-(p-carboxyethyl)adamantane. VI (6 g), dissolved in 250 ml of 2 N NaOH, was refluxed 18 hr. After cooling, the soln was acidified with 2 N HCI, and the ppt was filtered: yield 6 g (93%); mp 270-277" after crystn from EtOH; equiv wt, found 128; calcd 126.

2',6'-Dioxoadamantanespiro-3'-tetrahydropyran (VII). A mixt of 1.52 g of 2-carboxy-2-(p-carboxyethyl)adamantane and 30 ml of Ac,O was refluxed for 2 hr. Then the Ac,O was evapd in vacuo: yield 1.32 g (93%);mp 126-128" after crystn from Me,CO-hexane; ir spectrum (KBr), 1770 (C=O) and 1805 cm-' (C=O).

of 0.5 g of VI1 in 60 ml of dry C,H, was refluxed and MeNH, was introduced simultaneously up to satn. The MeNH, salt was filtered, dissolved in H,O, and acidified with 2 N HCl. The ppt was filtered and crystd from EtOH-C,H,: yield 0.55 g (98%);mp 236-239'; ir spectrum (KBr), 3390 (NH), 2500 (broad, COOH), 1690 (COOH), 1615 cm-' (NC=O).

Methyl-p-carboxamidoethyl)-2-carboxyadamantane (0.5 g) was heated in a sublimation apparatus under N, a t 250-300". Sublimate and residue were suspended in CH,CI, and filtered. The undissolved product was sublimed again. The CH,CI, soln was evapd and the residue sublimed at 200-300" (15 mm). Crystn from EtOH-H,O afforded pure VIII: yield 360 mg (78%); mp 119-121'; nmr spectrum (CCI,), singlet at 63.01 (NCH,) and 2 triplets centered at 62.00 and 2.54 (CH,CP,C=O).

NMethyladamantanespim3'-piperidine Hydrochloride OX). A mixt of VI11 (0.5 g) qnd LAH (0.25 g) in 100 ml of dry Et,O was refluxed for 5 hr, then the reaction complex was decompd with H,O and extd with Et,O. EtOH-HCl was added to the dried Et,O soln. The ppt was filtered and sublimed at 250-300' (15 mm): yield

2-(N-Methyl-~carboxamidoethyl)-2-carboxyadamantane. A soln

2',6'-Dioxo-N-methyladammtanespiro-3'-piperidine (VIII). 24N-

687 mg (41.6%); mp 301-306" after recrystn from EtOH-C,H,. Anal. (C,,H,,CIN): C, H, Cl, N.

2-Hydr6~ymethyl-2-@-hydroxyethyl)adamantane (X). Ada- mantanespiro-2'-succinic anhydride (73 g) was added t o 1300 ml of dry THF, followed by slow addn of 21.3 g of LAH. The mixt was boiled for 16 hr, then cooled. Thereupon a mixt of 50 ml of H,O and 200 ml of THF was added slowly, followed by 500 ml of 2 N H,SO,. The solids were filtered with suction and washed with THF.

The THF soln was washed with a satd NaCl soln and dried (MgSO,); the THF was evapd, and the residue (69.4 g) was added to a mixt of 3800 ml of CH,CI, and 2700 ml of 1 N NaOH. After sepn, the H,O layer was extd once with 250 ml of CH,CI,. After being dried (MgSO,), the CH,CI, was evapd, yielding 59.7 g (85.6%) of pure X: mp 109-11l";nmr spectrum (CDCI,), singlet at 63.82 (CH,O) and 2 triplets centered at, respectively, 6 1.94 and 3.78 (CH,CH,O). The H,O layer was acidified with concd HC1 and the ppt filtered: yield 8.95 g (12%) of IV.

Adamantanespiro-3'-tetrahydrofuran m). A mixt of 114 ml of 48% HBr, 59.7 g of X, and 11.4 ml of 96% H,SO, was stirred and heated for 3 hr at 110-115'. After cooling, 50 ml of H,O was added and the mixt was extd with four 100-1111 portions of CH,CI,. The combined exts were washed with a 10% NaHCO, soln and dried (MgSO,). The product was a brown oil: yield 50.7 g (92.9%); nmr spectrum (CDCI,), singlet at 63.70 (CH,O) and 2 triplets centered at, respectively, 61.84 and 3.79 (CH,CH,O).

2-BrOmomethy1-2-(p-bromoethyl)adamantane (XII). A stirred soln of 62.3 g of Ph,P in 245 ml of PhCN was treated slowly at 0' with 12.3 ml of Br,. On completion of the addn, the reaction mixt was heated to 125", and then 42.2 g of XI was added in 0.5 hr, after which the temp was maintained at 125" for a further 3 hr. After cooling, the reaction mixt was chromatogd over a silica gel column, with petr ether (bp 40-60') as eluant: yield 63.5 g (86%) after crystn from hexane; mp 80-81'. Anal. (Cl,H~,,Br~): C, H, Br.

methyl-2-@-cyanoethyl)adamantane (XVI). A mixt of 16.8 g of XI1 and 5.1 g of NaCN (tech grade) in 50 ml of methyl Cellosolve was stirred and refluxed for 5 hr. Then the solvent was evapd in vacuo, and the residue was chromatogd over a silica gel column, with CH2CI, as eluant. Fraction I, 4.07 g (28.8%) of XIII, had mp 75.5-77 after crystn from hexane. Anal. (C,,H,,NBr): H, N, Br. C: calcd, 59.57: found, 60.51. Fraction II,5.61 g (49%) of XVI, had mp 116-118" after crystn from C,H,-hexane. Anal. (C15H,,,N,): C, H, N.

3'-Cyano-4'-aminoadamantanespiro-3'-cyclopentene (XXI). A mixt of 7.05 g of XI1 and 2.93 g of NaCN (dried in vacuo at 110") in 50 ml of methyl Cellosolve was refluxed for 96 hr. After evapn of the solvent under reduced pressure the residue was suspended in 100 ml of H,O and extd with four 100-ml portions of CH,CI,. After evapn of the CH,CI,, 6.9 g of solid was obtained. Crystn from C,H,

2-Bromomethyl-2-(p-cyanoethyl)adamantane (XIII) and 2-Cyano-

136 Journal of Medicrval Uiernistry, 1972, Vol. I S , Ao. 2 vat2 Hes, et al.

Jforded 1 .5 g of XXI (31%): mp 239-241"; mass spectrum; exact mass found 228.1624; calcd 228.1626. A better method for prepg XXI is the following. A mixt of 6.05 g of XII: 2.7 g of dry NaCN, and 75 ml of dry DMSO was stirred at room temp for 0.5 hrt followed by stirring at 50" for 105 min, at 80" for 75 min, and 2 hr at 110". After the mixture had cooled, 0.88 g of NaH (50% dispersion in oil) was added. The suspension was heated to 95" in 15 min and, after being stirred for another 1.5 hr, the reaction mixt was cooled and poured into 500 ml of cold H,O. The ppt was filtered, washed with H,O, and dried: yield 3.1 g (75%); mp 238-240'.

Adamantanespiro-3'-cyclopentanone (XXII). A s o h of 325 mg of XXI in 10 ml of AcOH and 0.6 ml of H,O was refluxed for 15 min; then 4 ml of 85% H,PO, was added and the mixt was refluxed under N, for 21 hr. After cooling, the reaction mixt was poured into ice-H,O and the ppt was filtered with suction: yield 250 mg (85%); mp 50.5-51.5" after crystii from EtOH-H,O. Anal. C,,H,,O: C, H, 0.

0.7 g of KOH, 1.25 g of H,NOH. HCl, and 50 ml of EtOH was stirred for 15 min, filtered, and added to a s o h of 1.1 g of XXII in 15 ml of EtOH. After refluxing for 1 hr, the solvent was evapd in vacuo and the residue was crystd from hexane: yield 1.1 g (90%); mp 127-129.5".

A refluxing soln of 0.9 g of oxime XXIII in 6 0 nil of dry EtOH was treated with 8 g of small pieces of Na in about 2 hr under N,. To the cooled soln, H,O and ice (about 300 ml) were added and the reaction mixt was extd with three 100-ml portions of CH,CI,. After drying and evapn of the CH,CI,, a yellow oil was obtained, which was treated with Et,O and EtOH-HCI: yield 0.9 g (91%); mp 281- 283". Anal. (C,,H,,CIN~ 0.25H20): C, H, Q, N.

2-(./-Aminopropyl)-2-bromomethyladamantane Hydrochloride WV). A mixt of 2.4 g of XIII, 0.7 g of FYO,, 100 ml of EtOH, and 4.25 ml of 4.5 N EtOH-HC1 was reduced in a Parr apparatus in 17 hr. The reaction mixt was filtered and most of the EtOH evapd. After addn of Et,O, the ppt was filtered and dried: yield 2.28 g (83%); mp 214-215". Anal. (C,,H,,BrCIN): C, H, Br, 0.

2-(-y-Aminopropyl)-2-(P-aminoethyl)adamantane Dihydrochlo- ride (XWI). A mixt of 2.8 g of XVI, 1.2 g of PtO,, 160 ml of EtOH, and 10 ml of 4.1 N EtOH-HCI was reduced in a Parr apparatus in 17 hr. After filtration, most of the EtOH was evapd. Then THF was added, and the ppt was filtered with suction: yield 3.58 g (94%); nip 290-792". Anal. (Ci5H30CIzN,): C. H, Cl, N.

3.23 g of XIV in 300 ml of n-BuOH was added slowly t o a stirred and refluxing s o h of 1.68 g of KOH in 200 ml of n-BuOH. After being refluxed for 12 hr the reaction mixt was cooled, filtered, and evapd to dryness in vucuo. The residue was suspended in 100 ml of H,O and extd with six 40-ml portions of CIi,CI,. After drying of the org layer (&SO,), EtOH-HCI was added, and then the reaction mixt was evapd to dryness: yield 1.63 g (68%) after crystn from EtOH-THF; mp 252-255". Anal. (C,$I,,CIN): C, H, C1, N.

Conipd XVII (190 mg) was heated in a sublimation apparatus at 290' until sublimation ceased. The sublimed product was dissolved iri 25 ml of H,O and washed with Et,O. Then the H,O layer was made alkaline and extd with three 10-ml portions of Et,O. After drying (MgSO,) and the addn of EtOH-HCI, the s o h was evapd. The residue was shaken with THF and filtered: yield 26 mg (16.5%); mp 255-262". Anal. (C,,H,,CIN): H, N. C: calcd, 70.42; found, 69.79.

2',7'-Dioxoadamantanespiro-4'-perhydroazephe (XX). HBr was introduced into a mixt of 5 g of XVI and 200 ml of Et,O at O", until the solid material had dissolved. After standing for 17 hr, the Et,O was evapd in vucuo, and the residue was stirred for 1 hr at 100" in 1 7 ml of H,O acidified with a few drops of 48% HBr. After evapn of the solvent in vacuo the residue was boiled with 65 ml of C,H, and filtered. After concn of the C,H, soln to 25 ml it was cooled. The ppr was filtered, yielding 560 mg (10.4%) of XX: mp 234-236" ; m s s spzctruni, M+ 247; ir spectrum (KRr), 3175, 3070 (NII). k715 (C'=O), 1680 cm-' (C=O).

AdaniantanespUo-4'-perhydroazepine Hydrochloride (XVIII). A mixt of 560 mg of XX and 0.4 g of LAH in 12 ml of THF was r e fluxed for 4 hr under N,, then cooled. Thereupon 1.6 ml of H,O and 0.4 ml of 15% NaOH were added slowly. The ppt was filtered and washed with THF. After addn of EtOH-HCl the THF was evapd in vacuo. The residue was dissolved in 50 ml of H,O, washed with Et,O, made alkaline, and extd with three 20-in1 portions of Et,O.

Adamantanespirocyclo3'-pentanone Oxime (XXIII). A mixt of

Adamantanespiro-3'-aminocyclopentane Hydrochloride (XXIV).

Adamantanespiro-3'-piperidine Hydrochloride (XV) . A s o h of

AdamantanespUo4'-perhydroazepine Hydrochloride (XVIII).

After drying of the combined exts, EtOH-HC1 was added. The ppt was filtered and crystd from EtOH-Et,O: yield 0.48 g (83%); mp 259-263". This product was identical with that obtd by heating of XVII.

Adaniantanespiro-3'-tetrahydrothiophene (XXV). A soln ot 8.4 g of Na,S ' 9H,O in 9 ml of H,O was treated Nith 3.9 g of XI1 and 45 ml of DMF. This mixt was maintained at room temp for 15 min and then refluxed for 1 hr. After cooling, the reaction mixt wa5 poured into 300 ml of H,O and extd with three 100-ml portions of Et,O. The combined Et,O exts were washed with H,O, dried (MgSO,), and evapd: yield 2.3 g (95%) of a yellow-orange oil; nmi spectrum (CDCI,), singlet at 62.83 (CH,S), triplet centered at 2.81 (CCH,S), and triplet centered at 2.01 (CH,CS).

S-Methyladamantanespiro-3'-tetrahydrothiophonium Iodide (XXVI). A mixt of 1.54 g of XXV and 30 ml of Me1 was stirred for 24 hr at room temp. The cryst solid was filtered and washed with Et,O: yield 2.41 g (93%); mp 181-183" (from EtOH). Anal. Ki,Hz31S): C, H? I, S.

N-Benzyladamantanespiro-3'-pyrrolidine ( W I I ) . A mixt of 320 mg of C,H,CH,NH, 300 mg of XII, 8 ml of xylene, and a few mg of NaI was refluxed for 48 hr. After the reaction mixt had cooled, 100 ml of Et,O was added, and the mixt extd with four 3 0 " portions of 0.2 N HCl. The H,O s o h was made alkaline and extd with four 30-ml portions of E,t,O. After drying and evapn of the Et,O, an oil was obtained which was chromatogd over a silica gel column, with Et,O as eluant. The oil obtained was treated with Et,O and EtOH-HCI. The ppt was filtered, yielding 50 mg of XXVII: mp 285-290". This compd was identical with an authentic sample.'

N-(p-Aminoethyl)-adamantanespiro-3'-pyrrolidine (XXVIII). A soh of 3.5 g of XI1 and 50 ml of H,NCH,CH,NH, in 300 ml of EtOH was refluxed for 68 hr. The solvent was evapd and the residuc treated with 150 ml of H,O and 20 ml of concd HCl and then extd with three 50-ml portions of CH,Cl,. The H,O s o h was made alkaline and extd with four 50-in1 portions of CH,Cl,. The CH,CI, was evapd, yielding 2.3 g (94%) of a yellow oil. Treatment of the oil with EtOH-HCl in THF resulted in the formation of the di-HC1 salt: mp

N- [ N '-Salicylidene-(P-aminoethyl)] adamantanespiro-3'pyrroli- 255-258".

dine (XXIX). A mixt of 230 mg of XXVIII, 160 mg of salicylaldehyde, and 1 ml of EtOH was refluxed for I hr. After cooling, the cryst ppt was filtered: yield 235 nig (71%);mp 93.5-94.5". Anal. (C,,H,,N,O): C, H, N, 0.

N,N'-Bis(adamantanespiro-3'-pyrrolidine) (XXXI) and N- Amino- adamantanespiro-3'-pyrrolidine (XXX). A mixt of 2.0 g of XI1 and 45 ml of H,NNH,. H,O was stirred and heated for 24 hr at 120". After evapn of most of the H,NNH,. H,O in vacuo, the residue was suspended in 50 ml of H,O and then extd with four 100-ml portions of Et,O. The combined Et,O solns were washed with a satd NaCI soh and dried (MgSO,), and then the Et,O was evapd. The residue was chromatogd over a silica gel column, with Et,O and an increasing amount of EtOH being used as eluant. Fraction I, 200 mg (17%) of compd XXXI, had mp 117-1 19" after crystn from hexane; mass spectrum, M + 380; XXXI . HCI, nip 213-215'. Anal. (C,,H,,ClN,): C, H, C1, N. The structure was proved by redn of XXXI. A mixt of 85 mg of XXXI, 40 mi of EtOH, 10 ml of H,O, 1 ml of concd HCl, and 60 mg of PtO, was treated in a Parr apparatus with H, for 4 hr. After filtration and evapn of the solvent, Et,O was added, and the ppt was filtered: yield 82 mg (80%). This compd was identical with the compd described in part 1' and prepd via a different route.

Fraction I1 was obtained as 500 mg (40%) of compd XXX (oil). With EtOH-HCI in THF the HCl salt was obtained: mp 193.5-195.5" after crystn from THF-MeOH. Anal. (C,,H,+JN,): C, H, C I , N.

N-Salicylideneamino)adamantanespiro-3 -pyrrolidine (XXXII) was prepd in the same way as XXIX: yellow needles, mp 108-108.5". Ana!, (C,,H,,N,O): C. H, N. 0.

References ( 1 ) K. Luundahl, J. Schut, J . L. M. A. Schlatmann. ti. B. Paerels,

and A. Peters, J. Med. Chem., 15, 129 (1972) (part 1). (2) A. G. Anderson and F. J. Freenor, J. Amer. Chem. Soc., 86,

5037 (1964). ( 3 ) W. A. Nasutavicus ai:d F. Johnson, J. Org Chrm., 32, 2367

(1967). (4) J . J . Bloomfield, Tetrahedron Lert.. 2273 (1964). (5) S. Baldwin, J. Org. Chem., 26,3280 (1961). ( 6 ) D. Hadzi and J. Jan, Spectrochim. Acta, 25A, 97 (1969).


Biological inactivation of pyrrolnitrin. Identificationand synthesis of pyrrolnitrin metabolites

Patrick J. Murphy, and Terry L. WilliamsJ. Med. Chem., 1972, 15 (2), 137-139• DOI: 10.1021/jm00272a005 • Publication Date (Web): 01 May 2002






Metabolism of Pyrrolnitrin Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 137

Biological Inactivation of Pyrrolnitrin. Identification and Synthesis of Pyrrolnitrin Metabolites

Patrick J. Murphy* and Terry L. Williams

The Lilly Research Laboratories, Indianapolis, Indiana 46206. Received August 18, 1971

The absorption, excretion, and metabolic conversion of [ ''C]pyrrolnitrin (1) have been examined in the rat using both in vivo and in vitro techniques. Antibiotic activity was rapidly destroyed in vivo. Although 29% of the administered radioactivity was excreted in the urine in 24 hr and 50% in the bile in the same time period there was no pyrrolnitrin-like antifungal activity found in either plasma, urine, or bile. Pyr- rolnitrin was readily metabolized in vitro using rat liver microsomes supplemented with NADPH. Three of the products have been isolated and identified by comparison to chemically synthesized standards. These compounds are 3-(3-chloro-2-nitrophenyl)succinimide (2), 4-chloro-3-(3-chloro-2-nitrophenyl)-3- pyrrolin-2-one (3), and 4-chloro-3-( 3-chloro-2-nitrophenyl)-5-hydroxy-3-pyrrolin-2-one (4). A fourth product has been identified by "trapping" the metabolite as a HS(CH2)20H adduct and identifying this adduct by comparative tlc. This product is 4-chloro-3-( 3-chloro-2-nitropheny1)maleimide (5). The chemical synthesis of each of these compds is described. Each of the compds has been found to be devoid of pyrrolnitrin-like antifungal activity.

Pyrrolnitrin, 3-chloro-4- [ 2-nitro-3-chlorophenyl] pyrrole (l), is a broad-spectrum antifungal agent first reported by Arima and coworkers.' Although this antibiotic is effective

c1 1

as a topical preparation, minimal activity is noted after oral administratione2 Gastric acidity has been mentioned as a possible cause of this decreased activity, but active metabolic degradation has not been thoroughly investigated. In the course of our studies on the pharmacology of pyrrolnitrin we have examined its metabolism in the rat using both in vivo and in vitro techniques. We have found that pyrrolnitrin disappears rapidly from plasma, and the parent compd cannot be detected in either the urine or the bile after ip administration. When radioactive pyrrolnitrin was used, a number of metabolites were detected in the plasma, none of which possessed pyrrolnitrin-like antifungal activity. In vitro studies have shown that the pyrrole ring of pyrrolnitrin is readily oxidized by enzymes having the properties of mixed-function oxidases. The oxidation products include a potent alkylating agent, 4-chloro-3-(3-chloro-2-nitro- pheny1)maleimide (5).

Results and Discussion

Excretion. The results obtained from iv administration of ['4C]pyrr~lnitrin (Table I) indicate that after 24 hr approximately 30% of the radioactivity is excreted in the urine. When [14C] pyrrolnitrin was administered to rats having cannulated bile ducts, approximately 50% of the radioactivity was excreted in the bile in the first 24 hr (Table 11).

Metabolic Conversion of [ ''C]Pyrrolnitrin in vivo. Bile. The radioactive materials excreted in the bile were extd and cxamined by tlc. Extn of the bile with PhH removed between 5 and 10% of the total radioactivity from the aq phase. The ext was concd and chromatogd by tlc. The tlc plate was analyzed by radioautography. Although a minimum of 6 radioactive materials were present, no radioactive zone corresponded to unchanged pyrrolnitrin.

indicated by miciobiological assay. A bile sample (0.8 ml) The absence of pyrrolnitrin in the bile samples was further

Table I. Urinary Excretion of Radioactive Metabolitesu Time after injection, hr Per cent radioactivity excreted

1 1.7 2 8.1 4 7.8

24 29.0

['4C]pynolnitrin (6.0 X 10' dpmlmg) in 0.2 ml of polyethylene glycol.

uEach of four 200-g male rats was given an iv injection of 2 mg of

Table 11. Biliary Excretion of Radioactive Metabolitesa % radioactivity recovered

Time interval. hr l b 2c ~~~ ~ ~~

0-1 0.4 2 . 8 1-2 2.2 5.2 2-3 2.6 6.3 3-4 5.7 6.7 4-5 6.1 12.4 5-22 35.4 15.3

Total 52.4 48.7 uThe bile ducts of two 200-g rats were cannulated. After the can-

nulae were in place and the bile flow resumed, the rats were given a single dose of [ ''C]pyrrolnitrin by injection into the stomach. Bile was collected during the indicated periods. b4.0 mg of pyrrolnitrin administered (6 X lo6 dpm/mg). c3.2 mg of pyrrolnitrin administered (6 X lo6 dpm/mg).

containing an amount of radioactivity corresponding to 100 pg of pyrrolnitrin was found to be devoid of pyrrolnitrin- like antifungal activity (level of detection >0.3 pg).

Urine. Tlc of PhH extracts of urine from rats that had been given ['4C]pyrrolnitrin by iv injection showed no unchanged pyrrolnitrin. Microbiological assay of a sample of urine (0.4 ml) containing an amount of radioactivity equivalent to 100 pg of pyrrolnitrin showed no detectable pyrrolnitrin-like antifungal activity.

Plasma. Six rats (1 50 g) were given a dose of ['4C]pyrr~l- nitrin (4 mg, 1.2 X lo6 dpm/mg) sc. Blood samples were taken at intervals from 30 min up to 24 hr. The plasma was isolated by centrifugation, and an aliquot was used for liquid scintillation counting. One third of the plasma was used for bioassay, while the remainder was extracted, and the extract was analyzed by glc. Analysis of radioactivity indicated that the extracts contained the equivalent of from 1 .O to 4.4 pg of pyrrolnitrin. However, no pyrrolnitrin could be detected by glc (limit 0.01 pg), nor was there any detectable pyrrolnitrin-like antifungal activity. The results obtained from the plasma as well as from urine and bile indicated that pyrrol-

138 Journal of Medicinal Chemistry, 1972, Vol, 15, No. 2 Murphy and Williams

nitrin is rapidly converted to a series of metabolites that are devoid of antifungal activity.

learn more about the metabolic conversion products of pyrrolnitrin in vitro incubations were conducted using various liver subcellular components. The mitochondrial and 100,OOOg supernatant fractions had no effect on pyrrolnitrin. The microsomal fraction had a minimal ability to alter the drug when incubated without cofactors, but in the presence of NADPH or (less effectively) NADH there was a significant conversion (up to 15%) of added pyrrolnitrin in 30 min. Tlc indicated that a minimum of 6 products were formed during these incubations. Because a number of these metabolites seemed to correspond chromatographically to the in vivo metabolites. further studies were undertaken using large-scale incubation of pyrrolnitrin with isolated rat liver microsomes.

The metabolites of pyrrolnitrin were isolated by column chromatography of the extracts. The distribution of the metabolites is shown in Figure 1 . The individual peak fractions were combined, concd, and analyzed.

Fraction 1 was found to contain only unchanged pyrrolnitrin. Fraction 2 contained a component, Rf (5% MeOH- i-Pr20) 0.38. This material was isolated by prep tlc. The mass spectrum of the purified material was identical with that of 3-(3-chloro-2-nitrophenyl)succinimide (2). Fractions 3 and 4 were not analyzed further. Fraction 5 was purified by prep tlc in 10% MeOH-i-Pr20. Two radioactive bands were observed (Sa, Rf -0.32; 5b, Rf -0.44). These bands were eluted separately, and the purified compds were subjected to mass spectrometric analyses. The mass spectrum of Sa was identical with that of 4-chloro-3-(3-chloro-2-nitrophenyl)-3-pyrrolin-2-one (3). The mass spectra of 5b was identical with that of 4-chloro-3-(3-chloro-2-nitrophenyl)-5- hydroxy-3-pyrrolin-2-one (4). Fractions 6 and 7 were chromatographed in a number of solvent systems and were found to contain a mixture of components. None of these metabolites has been identified.

pounds suggested that structures such as 5 and 6 might be

In Vitro Conversion of [ ''C]Pyrrolnitrin. In order to

The route of metabolism indicated by the identified com-

c1 H c1 H

5 6 readily formed by common oxidation reaction. Since both of these compds are substituted maleimides, their generation in a biological system might be expected to lead to rapid binding to proteins and/or reaction with biological SH compds.

A further study of 5 was undertaken in order to assess this possibility. When 5 was allowed to react with HS(CH&OH, a quantitative formation of 7 occurred. With the hope of trapping intermediates such as 5 a series of incubations were performed with pyrrolnitrin in the presence of 6-mercaptoethanol with and without NADPH. The ex-

f R A C l I O N

N U M B E R $. * t ' i i &A

T U B E N U M B € R

t E l ~ O . M e O H

E L U A N T t t t

Figure 1. Column chromatography of pyrrolnitrin and pyrrolnitrin metabolites. The concentrated extract was dissolved in benzene and placed atop a 240 X 22 mm column of silicic acid. The eluant was changed as indicated; each fraction was 10 ml. The radioactive fractions were combined as indicated by fraction numbers 1-7.

4 : , 0enrene 0 e n ~ e n e - E I , O E1,O

4 : l l

Rodiooct i r i ty 1:; Slondord- Compound I l l

S o l v e n t System Solvent S y s t e m S o l v e n t S y s t e m

1 2 3

A B C A B C A B C

Figure 2. Thin-layer chromatographic analyses of extracts from in vitro incubation of [ 14C]pyrr~lni t~in with rat liver microsomes: A, pyrrolnitrin + NADPH; B, pyrrolnitrin + NADPH + HS(CH,),OH; C, pyrrolnitrin + HS(CH,),OH. Solvent system 1, 15 cm in PhH followed by 10 cm in i-Pr,O-MeOH (9: l ) ; solvent system 2, 15 cm in PhH followed by 10 cm in i-Pr,O-HOAc (95:5); solve$ system 3, 15 cm in PhH followed by 10 cm in EtOAc-PhH (1:l). The Rf of pyrrolnitrin in aII 3 systems is 0.6.5.

tracts were then cochromatographed with 7 in 3 solvent systems. The results are shown in Figure 2. In the presence of P-mercaptoethanol and NADPH a number of radioactive zones are formed which are not present in the extracts from incubations containing either of these compds alone. One of the compds formed cochromatographs with 7 in each of the solvent systems tested.

type compounds are generated during the microsomal oxidation of pyrrolnitrin.

A summary of these metabolic conversions of pyrrolnitrin by rat liver microsomes is given in Figure 3. All of the metabolites identified were oxidized pyrroles. The pyrrole ring seems to be particularly swceptible to oxidatiqn, which may account for the rapid degradation of pyrrolnitrin in vivo. Compds 2 , 3 , 4 , 5, and 6 were found to have no sig nificant antifungal activity. The generation of a substituted maleimide by microsomal oxidation is most interesting in

Thus it seems that at least one, if not more, maleimide

c1 7

Metabolism of Pyrrolnitrin Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 139

R = rl- -NO2 C I

Figure 3. Metabolic conversion of pyrrolnitrin in vitro by rat liver microsomes.

light of the known reactivity of maleimides with proteins and SH groups. To our knowledge the conversion of a pyrrole to a maleimide has not been previously reported. Whether significant amounts of the types of maleimides reported herein are formed in vivo can only be assessed by further study.

Experimental Section [ 14C]Pyrrolnitrin labeled in the 2 position of the pyrrole ring

was prepd biosynthetically from [ 14C] tryptophan by the method of Hamill, et al.' The isolated material was assayed for purity by tlc and exhibited a single radioactive zone corresponding to authentic pyrrolnitrin.

All enzyme cofactors were obtained from Boehringer-Manheim Corp. Silicic acid (100-200 mesh) was purchased from Bio-Rad Laboratories.

Assays. Plasma samples were assayed for radioactivity by diges- tion with a mixt of 0.2 ml of HC10, (70%) and 0.4 ml of 30% %O, at 70" for 45 min in Teflon-capped vials. The samples were cooled rapidly in ice and dissolved in 20 ml of a soln contg 11 ml of PhMe, 9 ml of methyl Cellosolve, and 6 mg of PPO. All other samples were dissolved directly in 10 ml of dioxane-based scintillation fluid prepd by combining 104 g of naphthalene, 75 ml of Permafluor (Packard In- struments), 425 ml of PhMe, 500 ml of dioxane, and 300 ml of MeOH. Radioactivity was detd using a Packard liquid scintillation counter with external standard.

Microbiological activity was assayed using Neurospora crassa (Lilly M45-846) as the test organism and pyrrolnitrin as the standard. Glc was carried out according to the method of Hamill, et al.,

Urinary Excretion. Four male 200-g Sprague-Dawley ,rats were given a single injection of 2 mg of [ I4C] pyrrolnitrin (6 X 1 0-6 dpm/ mg) in 0.2 ml of polyethyleneglycol iv. The urine was collected for the indicated periods, and total radioactivity was detd by counting an aliquot of the sample.

Biliary Excretion. Male Sprague-Dawley rats (200 g) were anesthetized with Et,O and their bile ducts were cannulated. A single dose of [ 14C]pyrrolnitrin in 0.2 ml of PEG,,, was administered by injection into the stomach. Bile was collected over the indicated period, and then the total radioactivity was detd by analyz- ing an aliquot of the sample.

In Vitro Incubation. Rat liver microsomes were prepd by differential centrifugation of 10% liver homogenates prepd in 0.25 M s u c r o ~ e . ~ The 100,000gpellet was resuspended in 0.05 M phosphate buffer, pH 7.0. NADPH was added at 5 X 10-4M. [ I4C] Pyrrolnitrin was dissolved in 0.05 ml of 0.1% Tween 80 in acetone prior to addn to the incubation medium (final concn, 0.1 mM). Following the addn of the enzyme prepn (1.5 ml) the final vol was adjusted to 2 ml with 0.05 Mphosphate buffer, pH 7.0. After the indicated time the reactions were stopped with 2 ml of Me,CO and then were extd with three 1-ml portions of PhH. Under these condns the recovery of unchanged pyrrolnitrin averaged 90%. The recovery of the metabolites varied due to production of nonextractable material in the course of metabolic degradation.

Large-scale incubations were performed using male rats that had been treated for 5 days with phenobarbital (40 mg/kg, ip). Micro- somes were prepd in the usual manner and were resuspended in 0.05

Mphosphate buffer, pH 7.0 (200 m1/10 g of liver). Microsomal suspension (30 ml) was placed in a 125-ml Erlenmeyer flask. An NADPH- generating system consisting of isocitrate dehydrogenase, NADP+, and isocitrate was added to give a final concn of NADPH equal to 2.5 X lO-,M. [14C]Pyrrolnitrin (1 mg in 0.1 ml of 0.1% Tween 80- Me,CO) was added to each flask. The reaction mixt was incubated 45 min at 37" with vigorous shaking in air. The reaction was stopped with an equal vol of Me,CO. The protein ppt was collected by centrifugation, and the aq phase was extd with PhH. The exts were concd and used for column chromatog.

Column Chromatography. In one series of incubations a total of 100 mg (0.6 X lo6 dpm/mg) of [ ''C]pyrrolnitrin was utilized. The PhH exts contd 1 X lo' dpm. The concd PhH ext was placed atop a silicic acid column (42 g, 240 X 22 mm) packed in PhH. The column was eluted with increasing concns of Et,O and MeOH. The peak fractions were pooled and concd.

Chemical Syntheses. Oxidation of F'yrrolnitrin with 00,- HOAc. Pyrrolnitrin (2.5 g) was dissolved in 100 ml of glac AcOH. A soln of CrO, (2 g) in 80% AcOH (200 ml) was added over a 1.5- hr period with stirring at room temp. After 15 min of add1 stirring 10 ml of abs EtOH was added, and the mixt was stirred for 15 min. The dark green soln was concd in vacuo at 50". The residue was triturated with PhH and H,O, and the aq phase was extd 3 times (PhH). The extracts were washed with H,O and concd. Upon concn a ppt formed which was isolated and recrystd from EtOH to yield pure 4-chloro-3-( 3-chloro-2-nitrophenyl)-5-hydroxy-3-pyrrolin-2- one (4). The structure was confirmed by mass spectrometry, ir, nmr, uv, and elemental analyses.

acid (240 X 22 mm). The column was eluted with PhH. After elution of unchanged pyrrolnitrin a fraction was eluted which contd 2- chloro-3-(3-chloro-2-nitrophenyl)maleimide (5 ) . This material pptd upon concn and was washed (PhH) and analyzed by nmr, mass spectrometry, uv, ir, and elemental analyses.

Oxidation of Pyrrolnitrin with Chloroperbenzoic Acid. Pyrrolnitrin (10 g. 39 mmoles) was dissolved in 60 ml of CbCl , . Solid Na,HPO, (22.4 g) was added. A CH,C1, soh of m-chloroperbenzoic acid (7.8 g in 60 ml of CH,Cl,) was added slowly (75 min) to the stirred slurry at room temp. After this time the CH,Cl, slurry was treated with a 2% aq Na,S,O, until a neg peroxide test was obtd with starch-iodide paper. The soln was then extd with 3 vol of H,O and dried (MgSO,). The org layer was taken to dryness. The residue was extd with Et,O leaving a blue polymeric ppt. The Et,O ext was taken to dryness and then the residue was triturated with PhH and placed atop a column of silicic acid packed in PhH. The column was eluted with PhH followed by increasing concns of EtOAc. The initial benzene eluate contd pyrrolnitrin followed by a yellow compd, which was chromatog distinct from pyrrolnitrin. Mass spectral, nmr, ir, and uv evidence indicated that it was 343- chloro-2-nitropheny1)maleimide (6). The fraction eluted with 5% EtOAc-PhH was concd, and the crude ppt was recrystd from EtOAc- PhH to yield 3-(3-chloro-2-nitrophenyl)succinimide (2). The structure was confirmed by nmr, ir, uv, and mass spectrometric data.

The fraction eluted with 10% EtOAc-PhH was concd and a ppt appeared. This was washed (PhH) to give essentially pure 4-chloro-3- (3-chloro-2-nitrophenyl)-3-pyrrolin-2-one (3). The structure was assigned on the basis of mass spectral, ir, uv, nmr, and elemental analyses. The assignment of the carbonyl position was supported by chemical reactivity of the 4-C1 substituent. Although a number of other products were observed, none was sufficiently purified for identification.

Acknowledgment. We would like to thank Dr. R. Hamill, Mr. C. B. Carrell, Mr. J. A. Mabe, and Dr. D. R. Brannon for preparing pure [ 14C] pyrrolnitrin for use in the above studies. We would also like to thank Mr. J. Westhead for the microbiological analyses and Mr. H. Sullivan for assistance in the glc analyses.

The PhH supernatants were then placed atop a column of silicic

References (1) K. Arima, H. Imanaka, M. Kousaka, A. Fukuda, and G. Tamura,

(2) M. Nishida, T. Matsubara, and N. Watanabe, ibid., Ser. A , 18, J. Antibiot., Ser. A , 18, 201 (1965).

211 (1965). (3) R. Hamill. R. Elander. J . Mabe, and M. Gorman, Antimicrob. Ag. . ,

Chemother., 1967,388 (1968).

18,310 (1969). (4) R. L. Hamill, H. R. Sullivan, and M. Gorman, Appl. Microbiol.,

(5) G. H. Hogeboom,MethodsEnzymoL, 1, 16 (1955).


Equilibrium and kinetic effects of sixteen compoundson the forms of horse heart ferricytochrome c

George H. Czerlinski, and Rita A. GaluszkaJ. Med. Chem., 1972, 15 (2), 140-144• DOI: 10.1021/jm00272a006 • Publication Date (Web): 01 May 2002






140 Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 Cz erlinski and Galuszku

Equilibrium and Kinetic Effects of Sixteen Compounds on the Forms of Horse Heart Ferricytochrome e t

George H. Czerlinski* and Rita A. Galuszka

Department of Biochemistry, Northwestern University Medical School, Chicago, Illinois 6061i. Received July 1, 1971

Equilibria and kinetics of the forms of horse heart ferricytochrome c wzIe investigated in the pH range 7-1 0 in the presence of a variety of substances with the intent to find effectors which bind preferentially to one of the several forms. The absorption band at 695 mu was used as indicator and is attributed to the form of ferricytochrome c which participates in electron transfer. Preferential binding would then be indicated either by an increase or a decrease in the absorption. Of a total of 16 tested compds, 4 showed an absorption increase of more than 20% (ADP, NaN3, pyrazole, thiazole), while 2 showed an absorption decrease of more than 20% (aniline, imidazole), 5 compds were within 5% of the value of ferricytochrome c in buffer (AMP, benzimidazole, pyrimidine, thiazolidine-4-carboxylic acid, and thiophene- 3-carboxylic acid), 3 compounds showed an intermediate increase in absorption (ATP, barbituric acid, 2,4-dinitrophenol) and 2 compds an intermediate decrease (hydantoin-5-acetic acid, pyrrolidine). These changes were observed at pH 7.4 within a few seconds after injection of cytochrome c into the reaction mixture. If the pH of horse heart ferricytochrome c in buffer is quickly changed from 7.4 to 10, 75% of the 695 band disappears with a time constant of about 1 sec. In presence of the various test substances, the same quick pH change reduced the 695 band to an extent and with a time constant, which is characteristic of the specific compound. The changes were smallest with imidazole and largest with thiazolidine- Ccarboxylic acid. The time constant in presence of hydantoin-5-carboxylic acid probably exceeds the time resolution of the experimental arrangement (0.2 sec). The time constant for the disappearance of the 695 band is increased by about a factor of 10 in presence of AMP. The values in presence of all other compds lie in between. If the pH of horse heart ferricytochrome c in buffer is quickly changed back from 10 to about 7.4, the 695 band reappears with a time constant of close to 8 sec. The absolute value of the change back to pH 7.4 is generally somewhat smaller than the original absolute change in absorption (for pH 7.4 -+ lo), in some cases ( e . g., in presence of benzimidazole) substantially smaller, but larger in the presence of imidazole. The 695 band reappears generally slower in the presence of a test substance (slowest for thiophene-3-carboxylic acid) than in its absence; exceptions are imidazole (fastest reappearance), hydantoin-5-acetic acid, and t hiazolidine-4-carboxylic acid.

Brandt, et al.,' demonstrated that the pH dependence of the redox potential of horse heart cytochrome c is due to a rather slow interconversion of different structural forms of the protein. This interconversion is proton coupled and only the form(s) stable at pH 7 are quickly reducible by ferro- hexacyanide. For this specific electron transfer Brandt, et al., ' obtained rate constants from temperature jump experiments, which are in reasonable agreement with results of Sutin and Christman2 and of H a ~ s t e e n . ~ The proton release upon oxidation of ferrocytochrome c by ferrihexacyanide was subsequently investigated in detail by Czerlinski, ef

experiments below) led to two apparent rate constants, which suggested the reaction cycle shown in Figure 1, Scheme A.

The equilibrium constant between components 1 and 4 of this scheme of Figure 1 is 20, a value which can be derived from the earlier data of Schejter and George.6 The equilibrium constant between components 2 and 3 favors component 3 and the interconversion constant was esti- mated4 at 3. With these numbers for the 2 interconversion constants and the apparent protonic dissociation constant of the overall reaction of pK1= 9.1, one obtains pKH' =. 9.6, and pK," = 7.8 (see Figure 1, Scheme A, for definitions).

Differences between the oxidized and reduced forms of cytochrome c have been known for some time, the oxidized protein having a more open and flexible conformation than the reduced form.' Among recent reports in this connection are studies on the optical rotary dispersion*-" and on circular d i~hro i sm '~ , '~ of the protein. The above mentioned isomerization of ferricytochrome c at alkaline pH to form a species with altered electron-transfer

ations (N69-67 plus C70-66).

Subsequent "pH-switching'' experiments (see type IV

?Supported by a Grant from the Chicago and Illinois Heart Associ-

properties was also discussed by Greenwood and Palmer, '' Urry,ls and Myer and Harbury." Watt and Sturtevant16 determined the enthalpy of oxidation of ferricytochrome c over the pH range 6.0-10.9 and were also able to derive thermodynamic parameters for the interconversion between component l and 3 of Scheme A of Figure l . Since these were equilibrium experiments, they had no information as to intermediate steps betwen components l and 3.

The original intent of the present investigation was to verify Scheme A of Figure 1 by using "effectors", which would hopefully exclusively bind to either form 1 or form 3. No such exclusive binding was revealed upon studying the following possible complexing agents: AMP, ADP, ATP, hydantoin-5-acetic acid, thiophene-3-carboxylic acid, 2,4- dinitrophenol, benzimidazole, imidazole, pyrazole, thiazole, thiazolidine-4-carboxylic acid, pyrrolidine, pyrimidine, PhNH2, NaN3, barbituric acid, amytal, and antimycin A (the last 2 in 40% EtOH). Many of the listed substances are known to have medicinal effects. It was therefore also of interest to look at the effects of these substances upon the most thoroughly characterized memeber of the electron transfer chain of the inner mitochondrial membrane: ferricytochrome c. Some of the substances developed demonstrated considerable ligand-induced conformational changes (either enhancing or counteracting the previously shown pH-induced conformational change).

Experimental Section Horse heart cytochrome c Type 111 was obtd from Sigma Chem-

ical Co. and further purified according to Margoliash and Lustgar- ten," adding initially an equimolar ami of sodium ferrihexacyanide. All experiments involved the simultaneous observation of transmission changes at 633 (a high shoulder of the 695 band) characteristic of protein conformation6 and of pH changes, using a thermostated Pyrex vessel with N, blown over the s o h to diminish CO, pickup (7.0 Q pH < 10.1). The spectral change was observed

Effects on Forms of Ferricytochrome c

S O

Journal of Medicinal Chemistry, 1972 Vol. IS, No. 2 141

Figure 1. A. Reaction scheme of Czerlinski, et ~ l . , ~ depicting 4 different forms of oxidized cytochrome c. The oxidation of reduced cytochrome c was performed with sodium ferrihexacynanide. This scheme is based on both equilibrium and kinetic experiments. B. A reduced scheme, implying only oxidized cytochrome c and inhibitor I reacting with any 1 of the 3 presented forms of cytochrome c. I’ and I” are one and the same inhibitor I , but differently labeled here for later discussions.

with a small He-Ne laser, University Laboratories Model 200, utilizing a beam splitter and 2 Ge photodiodes. The pH change was observed with a combination glass electrode, Type 14073 of Instru- mentation Laboratories, and Radiometer pH-meter Model 26. The system was weakly buffered by 10 mM sodium phosphate and 4 mM glycine in all s o h , except for those where the test substance has a PKH in the range under investigation. The signals were fed into different channels of a Beckman Type RB Dynograph (with Type 9853 plug-ins). The 4 types of experiments are briefly described in the next paragraphs.

above buffer and 0.030M Na,SO, were adjusted to pH 7.40 (or 10.00) within the thermostated vessel (25”). Fifty pl of 22 mM test substance in buffer, adjusted to pH 7.40 (or to 10.00), was then injected quickly. Spectral and pH changes were recorded simultaneously. Fifty pl of 10 mM H,SO, were injected after equilibration to calibrate the pH changes. Spectral changes were calibrated by a “blank” of type IV (below).

Type 11. Test substance (4.50 ml, 0.1 1 M) (for deviating concns, see Table 11) in the above buffer and enough Na,SO, to maintain constant ionic strength (0.1 M) thorughout all types of experiments, was adjusted to pH 7.40 (or 10.00) within the thermostated vessel (25”). Oxidized cytochrome c (500 pl, 2 mM) in buffer (ionic strength adjusted to 0.1 with Na,SOJ, adjusted to pH 7.4 (or 10.00), was then quickly injected and changes recorded. In the control, Na,SO, was substituted for the test substance (compare type I experiments),

dered sodium dithionite (purified, low in Fe, of Fisher Scientific Co.) to its s o h in such an amt that twice the molarity of cytochrome c could have been reduced. The s o h was then dialyzed 3 times against the buffer mixt and spectrophotometrically tested for purity (98% or more in the reduced form). Reduced cytochrome c (5 ml, 0.2 mM), in 0.10M test substance and the above buffer with Na,SO, to adjust the ionic strength to 0.1, if necessary, was adjusted to pH 7.40 (or 10.00) at 25”. Na,Fe(CN), (22 mM, 50 ml) in the above buffer (pH 7.40 or 10.00) was quickly injected, and signal changes were recorded. The control contd Na,SO, in place of the test substance. Injection of H,SO, followed after equilibration (as for type I experiments above).

Type IV. Oxidized cytochrome c (5.00 ml of 0.2 mM) in the above buffer and 0.10M test substance (or 0.030 mM Na,SO, for the control) were adjusted to pH 7.40 (or 10.00) at 25”. NaOH (1 N , 40 pl) was injected quickly, to bring the pH to about 10. Upon equilibration, 40 pl of 1 N H,SO, was quickly injected, to bring the pH back to 7.4. To observe slow changes of pH after the rapid change in pH, proper biasing of the channel measuring pH changes is necessary. These are calibrated again by injections of 50 p l of 10 mN H,SO,.

Experiments of type IV could easily follow the other 3 types of experiments. Most frequently, we carried out first type 11, followed by type IV experiments. To illustrate the sequence of these

Type I. Oxidized cytochrome c (5.00 ml, 0.2 mM) in the

Type 111. Cytochrome c was reduced by the addition of pow-

I I , odd odd , Odd

Figure 2. Schematic demonstration of an experimental sequence, indicating 2 recorder traces and defining the various signal changes. So represents the signal before addition of cytochrome c and obtained such that the light beam through the test vessel was shut off by a piece of black paper. “Add cyt c” represents the quick injection of 500 pl of (nominally) 2 mM cytochrome c stock into 4.5 ml of a solution of the test substance (type I1 experiments). AS, represents thus the equilibrium signal change, associated with this injection (measured in mV, as all signals). “Add NaOH” represents the quick injection of enough 1 N NaOH, to reach (or exceed) pH 10 (type IV experiments); in some cases, more and/or stronger NaOH was needed to reach pH 10 (compare discussion of Table 11). AS, represents the observed equilibrium signal change, associated with this injection. “Add H,SO,” represents the quick injection of enough 1 N H,SO,, to return the pH to 7.4 (type IV experiments); the equivalents of H,SO, added generally equaled the equivalents of NaOH added before. AS3 represents the observed equilibrium signal change associated with this last injection.

experiments, Figure 2 is presented. Figure 2 introduces also the various symbols used in the evaluations.

in optical transmission are quite small, the relevant differential equations may easily be linearized. The change of the trace on the graph is given by

Evaluation of Kinetic Data, As the observed changes in pH and

d x l d t = -xX (1) with x = particular ordinate value on the graph (in mm, convertible to mV). As the abscissa is directly recorded in seconds, the apparent rate constant A is given in sec-’ . One obtains in zero approximation

x = X exp(-At) t Q

with 3 equal to the equilibrium change, extrapolated back to t = 0. The constant term a (with a < 3) takes care of any (initial) rapid change. A semilog arithmic plot results in a straight line with a slope of A. Most of our kinetic evaluations, however, were performed by using a computer program of Berman.’*

The above equations imply that the time-dependent change can be properly represented by a single exponential decay curve. This is not necessarily so, but in these surveying experiments this was assumed in first approximation. If Scheme B of Figure 1 applies, a stepwise pH change from 10 to 7 cytochrome c alone (that is without test substance present) should directly lead to the rate constant k,, as the protonic step proceeds much faster than the isomerization step.”’ However, for pH switching from component 1 to 3, in the absence of inhibitor, one could only expect to obtain a limiting value for the rate constant k,. Nevertheless,, this lower limit is of value in further discussions of the applicability of this scheme.

The rate constants, derived from a stepwise pH change for the cytochrome c system are dependent upon the final pH value. The magnitude of the total signal change upon a stepwise pH change is a function of the initial and the final pH. The signal change for the linearizable case (relatively small transmission changes) is given by

As = V(C* + c2) (3) It is assumed in this equation that a signal derives only from

components 1 and 2 and both components have the same coefficient I) (in mV/ruM).

142 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Czerlinski and Galuszka

The reaction scheme may be written in the general form (not considering effectors).

(10) -

KH' 1 1 KH"

One may then consider two cases.

k,' and k2'. For this reduced scheme one may easily derive for the relative signal change

1 - A =-

Case 1. KII' < K I I . This condition allows one to neglect K I I ' ,

c=1 t F2 6

One obtains for the apparent rate constant h = 1/7

Case 11. K p ' = K I I . This condition leads automatically to the fact that kz' /k , = K z , l . If also k,' = k , , the system of eq 4 may be treated as if the upper cycle is not present at all (assuming that com. ponent 10 has the same q as components 1 and 2). The equations for the reduced cycle are easily obtained by settingK11 = 0 in eq 5 and 6. For further details see the Appendix.

Results The experimental results are listed below in two tables. Table I summarizes the results from the control (line 1)

and 3 representative inhibitors. Controls were measured on every day on which materials were tested for preferential binding. Two to six tests of the type shown in Figure 2

Table 1. Summary of ReDresentative Result9

were generally conducted on each test substance. Only those results were used for further evaluation, where the curves indicating the pH change most closely approximated a stepwise change.

Table I1 summarizes the results obtained on all materials tested for preferential binding in an abbreviated manner. aSl'/AS,' and &,'/&,' should be compared with the corresponding values for the controls, ASl/aSo = 0.44 and AS3/ASo = 0.38. The larger the deviations from these control values are, the stronger is the effect by the test substance. Deviations within +lo% cannot be considered sig nificant. The limiting overall response time of the experimental arrangement is close to 0.2 sec, which is in good agreement with the largest AI* found experimentally. One derives, with AI* 1.04 sec-' for the control from Table I together with Al'/A1 = 4.7 from Table I1 for the hydantoin- derivative, A,,, 4.9 sec-'. Any hl'/Al between 1 and 5 may be corrected for this overall response time (omitted here).

All the compounds in Table I were tested for equivalent proton release upon injection of cytochrome c. Unfortu- nately, the experimental error could not be reduced below 0.2 equiv. Upon the injection of cytochrome c, almost all inhibitors showed a change of pH which corresponded to not more than the given error. There were 3 exceptions. Thio- phen-3-carboxylic acid, dinitrophenol, and PhNH2 showed a proton release of 1 equiv within the experimental error. However, no proton change was detectable with imidazole and pyrimidine, apparently due to the buffering capacity of these compounds since upon injection of the calibrating acid, no pH change could be detected.

The data in Tables I and I1 do not reveal that the vol of injected base (or acid) was not always the same. Generally, around 30 pl of 1 N acid or base was used, but twice this amount was needed for the 3 adenosine phosphates and for benzimidazole; 1.5 times t h s minimum amount was needed for pyrrolidine and 2.5 times this amount for aniline. NaOH (1 2 N , 15 pl) was needed for imidazole and twice this amount for hydantoin-5-acetic acid. These larger amounts of acids and bases are needed for those compounds, which have protonic dissociation constants at or near the experimental range.

dine-4-carboxylic acid undergo a slow change, which was the principal reason for our use of 0.02 M of this test substance in the experiments listed in Table 11. The apparent rate constant of this slow change was found to be a function of the analytical concn of the thiazolidine derivative and of pH. When the reaction mixture of the thiazolidine derivative with ferricytochrome c was kept long enough, the color of the reduced cytochrome c appeared. We sus- pected that the S of thiazolidine might have replaced the S

Table I1 does not reveal that the systems with the thiazoli-

Effector No. (0.1W So, mV AS", mV PH, AS. , mV AS,* , mV A,*, sec-' uH, AS*. mV AS,*, mV ha*, sec-' PH,

1 None 1535 53 7.425 28 29 f 1 1.04 t 0.03 10.049 24 22 f 0.6 0.120 f 0.007 7.460 2 Azide (Na) 1550 81 7.440 46 44 f 0.6 0.63 f 0.02 9.960 42 37 r 0.7 0.070 f 0.003 7.440 3 Benzimidazole 1550 62 7.469 28 30 0.42 f 0.02 11.150 10 1 1 * 0.5 0.067 t 0.007 7.635 4 Imidazole 1525 43 7.371 9 8.3 f 0.2 1.93 * 0.07 10.015 1 1 5 f 0.2 0.31 i 0.02 7.469 OThe initial signal S, is defined in Figure 2, as well as the ASi; however, the ASi in thistable are normalized for identical initial signal (refer-

ence So = 1500 mV) such that the experimentally obtained (ASi)obsd is multiplied with the factor (1500/So). The three pH values refer to the pH measured after the new equilibrium value is reached and are thus subscripted like the preceding AS. The pH before injection of cytochrome c is generally within 0.05 unit of pH,. Four parameters in the table are superscripted by an asterisk to indicate that the listed values have been obtained by an iterative computer program; the standard error is also listed together with these values. The subscript refers to the same (equilibrium) signal change as for Mi; Xi* represents a pHdependent (apparent) rate constant and appears as parameter in ASi(t ) = AS,* exp(-Xj*t) + a*, where a* represents a computergenerated constant to optimize the fit of the theoretical curve to the data. These results are needed for Table 11.

Effects on Forms of Ferricytochrome c Joumal of Medicinal Chemistry, 1972, Vol. 15, No. 2 143

Table 11. Summary of Evaluation Result9

Effector PHl ASl' /ASi AS,'/A$,' aS,'/AS, %'/A1 &'/A3

AMP 10.001 0.46 0.50 0.97 0.10 0.35 ADP 10.140 0.44 0.37 1-21 0.28 0.46 ATP 10.158 0.45 0.38 1.17 2.9 0.76 Aniline 11.228 0.39 0.35 0.79 2.1 0.42 Azide(Na) 9.960 0.57 0.54 1.29 0.6 1 0.58 Barbituric acid 10.110 0.44 0.50 1.16 0.48 0.37 Benzimidazole 11.150 0.45 0.16 0.99 0.40 0.5 6 2,4-Dinitrophenol 9.970 0.60 0.34 1.1 0.43 0.27 Hydantoin-5-acetic acid 10.921 0.49 0.4 1 0.8 1 4.7 1.67 Imidazole 10.015 0.21 0.26 0.69 1.8 2.5 8 Pyrazole 10.234 0.62 0.47 1.28 0.67 0.39 Pyrimidine 10.205 0.5 0 0.28 0.96 1.2 0.38 Pyrrolidine 10.110 0.46 0.34 0.91 0.33 0.57 Thiazole 9.949 0.49 0.39 1.26 0.38 0.33 Thiazolidine-4-carbox acid 10.020 0.64 0.21 1 .o 0.3 1.2 Thiophene-3-carbox acid 10.120 0.5 2 0.47 1.0 0.26 0.13

"The concns of effector were 0.1 M, except for 2,4-dinitrophenol(0.01 M ) and thiazolidine4carboxylic acid (0.02M). The ionic strength was 0.1, except for the highly charged adenosine phosphates. The initial pH, (compare Table I for definition) was 7.40 within 0.07 units in all cases, the final pH, was 7.4 within 0.1 units except for amiline (7.56), benzimidazole (7.64), hydantoin-5-acetic acid (8.531, pyrimidine (7.63), and thiophene-3-carboxylic acid (7.54). pH1 is the intermediate pH, as defined in Table I . The equilibrium signal changes in presence of effector are denoted by a prime for distinction from those in absence of effector. The subscripts are otherwise those of Table 1; for simplic- ity, a star is omitted for the h parameters. One derives easily from line 1 of Table I that for the control AS,IAS, = 0.44 and AS,/AS, 0.38. The ratio ASo'/ASo would reflect any differences in cytochrome c concentrations. However, the analytically detd concns of cytochrome C varied only from 0.167 to 0.170mM, well within the error of the detnsof AS.

of methionine 80 in the sixth ligand position. To test this idea, we conducted a series of experiments on the Cary recording spectrophotometer. The spectrum of pur compound with thiazolidine (high concentrations) was identical with that of reduced cytochrome c.

Results from experiments of types I and 111 are not explicitly tabulated, as their presentation would not provide much new information. Most ideal would have been a type I experiment as the perturbation of the system by the minute injection could certainly be neglected. However, this experiment was only comparatively successful with the thiazolidine derivative, as indicated below. Otherwise, the ob- servable effects for type I experiments were quite small and merely gave marginal confirmations of the results already listed, namely that the binding for the duration of the experiments is generally weak. Some type 111 experiments were conducted at pH 7 and also at pH 10. In most cases, essentially the same apparent rate constants were obtained except for imidazole and dinitrophenol, suggesting that the process of oxidizing reduced cytochrome c in the presence of these compounds does not proceed along a single path, as depicted thus far.

Discussion

The various materials tested can be classified into those which have practically no effect and those which do effect the red band (695-633 mp) of cytochrome c. Those which do effect the red band of cytochrome c may be divided into those which enhance the absorption band and those which result in a decrease of the absorption band. The former ones may also be labeled activators, the latter ones inhibitors, although their specific action regarding electron transfer is not quantitatively established in this paper.

It is by now well established that the 695 band is caused by the S atom of methionine 80, located in the sixth ligand position of Fe+3. Activation may thus be considered as strengthening the binding of this group to the heme Fe; inhibition may be considered as removing this S to a smaller or larger extent from this ligand position. There is also the trivial solution that the red absorption band is diminished by the reductive action of the test substance upon cytochrome c, producing the reduced form.

With respect to the results summarized in Table 11, one may distinguish compounds which seem not to effect cytochrome c from those which either enhance the red gbsorp- tion band or diminish it. The equilibrium results of the 3 adenosine phosphates are rather similar to those of the control, although aS,'/ASd is somewhat large for AMP and ASd/aSo is somewhat large for ADP. On the other hand, the kinetic parameters for these 3 adenosine phosphates differ considerably. In fact, AMP shows the slowest apparent rate constant in Table 11, if the pH is changed stepwise from 7.4 to 10.0. Next to this slowest value follow the thiophene derivatives, ADP and the thiazolidine derivatives, in that sequence (but all rather close together). The thiophene derivative is also associated with the slowest apparent rate constant for a stepwise pH change from 10 to 7.4, giving Xi&,= 0.13. Fastest for this latter change are imidazole and hydantion-5-acetic acid, the 2 compounds which also needed the largest amount of base for obtaining the proper change in pH. A secondary reaction of imidazole i s reflected in the rather small values for ASl'/AS~ and LLS,'/ASd; however, this slow reaction was not further investigated.

Earlier experiments of Czerlinski, ef al.,4,' had shown that exactly 1 H' is released at pH 10, at which point we measured a signal change of 30 mV (compared to the signal, which was present at pH 7.4, observing at 633 mp in our thermostated vessel). We changed the pH stepwise to 11, leading to decreasing increments of voltage changes, totaling 10 mV. This change is due to the second dissociation constant, labeled PKHII, which was 10.4. In crude approximation, the given ratio of voltage changes should result in the equilibrium constant between components 2 and 3. Certain- ly these are crude approximations as the protonic dissociation constants have to be taken into account quantitatively. However, the possibility of measuring such a substantial additional change suggests that a protonic dissociation is not associated with component 2 of scheme A of Figure 1, but only with component 3. This result eliminates case 11, listed above under eq 6. The pH dependence of the observables are therefore presumably given by eq 5 and 6. Effectors would influence these pH dependencies, as outlined in scheme B of Figure 1.

Scheme B of Figure 1 in fact represents a simplified mech-

144 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Craig

anism for the combination of a ligand-induced (represented by I, 1’, and Iff) with a pH-induced conformational change. Detailed measurements aiming at exact values for individual constants, are now in progress.

David Wojcinsky for technical assistance with the electronic circuits and Dr. Viera Bracokova Czerlinski for assistance in the evaluation of the data. We would also like thank Dr. Hugh J. Burford, for loaning us a Beckman Type RB Dyno- graph with plug-in units.

Appendix Equation 6 with KII = 0 was previously derived.lg Eq 6

presents thus an expression of the former derivations. More details on the derivation of eq 5 and 6 are also obtainable from a monograph on chemical relaxation” (see especially Tables 8.2 and 8.3, as well as Chapter 7). Expressions for apparent rate constants for the two-step scheme of Figure 1, heavy lines of part B (and KH’ defined as in part A) are easily obtained from eq 6 by setting k3 = k4 = 0 (as well as KU 0).

References

Acknowledgment. The authors would llke to thank Mr.

(1) K. G. Brandt, P. C. Parks, G. H. Czerlinski, and G. P. Hess, J. Bio. Chem., 241,4180 (1966).

N. Sutin and D. R. Christman, J. Amer. Chem. Soc., 83, 1773 (1 96 1). B. H. Havsteen, Acta Chem. Scand., 19, 1227 (1965). G. Czerlinski, K. Dar, and G. P. Hess, Fed. Proc., Fed. Amer. Soc. Exp. Biol., 26, 674 (1967). G. Czerlinski and K. Dar, Biochim. Biophys. Acta, 234,57 (1971). A. Schejter and P. George, Biochemistry, 3, 1045 (1964). E. Margoliash and A. Schejter, Advan. Protein Chem., 21, 113 (1966). D. D. Ulmer, Biochemistry, 4,902 (1965). D. W. Urry and P. Doty, J. Amer. Chem. Soc., 87 , 2756 (1 965). Y. P. Myer and H. A. Harbury, Proc. Nat. Acad. Se i U.S., 54, 1391 (1965). R. Musky and P. George, Proc. Nat. Acad. Sei. US. , 56, 222 (1966). Y. P. Myer, Biochemistry, 7, 765 (1968). Y. P. Myer, J. Biol. Chem., 243, 2115 (1968). C. Greenwood and G. Palmer, ibid., 240, 3660 (1965). D. W. Urry,Proc. Nat. Acad. Sei. US., 54, 640 (1965). G. D. Watt and J. M. Sturtevant, Biochemistry, 8,4567 (1969). E. Margoliash and J. Lustgarten, J. Biol. Chem., 231, 3397 (1 964). M. Berman, in “Computers in Biomedical Research,” R. Stacy and B. Waxman, Ed., Vol2, Academic Press, New York, N. Y., 1965, Chapter VII. G. Czerlinski, J. Theor. Biol., 17, 347 (1967), eq 3.17. G. Czerlinski, “Chemical Relaxation- An Introduction toTheo- ry and Application of Stepwise Perturbation,” Marcel Dekker, Inc., New York, N. Y., 1966.

Structure-Activity Correlations of Antimalarial Compounds. 1. Free-Wilson Analysis of 2-Phenylquinoline-4-carbinols

Paul N . Craig*

Smith Kline & French Laboratories, Philadelphia, Pennsylvania 19101. Received April 23, 19 71

Sixty-nine 2-phenylquinoline-4-carbinols which had been tested in the mouse for antimalarial activity were studied by the Free-Wilson method for structure-activity correlation. The results for this study significantly support the additivity concept assumed by the Free-Wilson approach. The substituent constants for groups at the para position of the 2-Ph ring were found to correlate significantly with both Hammett’s meta u constants and Hansch’s 7~ values for those substituents. Substituents on the 7 position of the quinoline ring correlate well with para u and 71 values. Substituent constants for groups at position 8 of the quinoline ring correlate with 71 values for these substituents. Substituent constants for groups at position 6 and on the meta position of the 2-Ph ring failed to correlate with u or 71 values; the substituent constants for the 16 different aminoalkyl side chains failed to correlate with 71, or 71 and 71’. The significance of these results is discussed.

The Free-Wilson method of structure-activity correlation’ has been applied with varying degrees of success in recent

This paper reports a successful application of the technique to compare the antimalarial test results in mice for 69 substituted 2-phenylquinoline carbinols of general structure I.

H-C-OH

I

The biological test reports, obtained from the Walter Reed computer record system, gave data obtained by Rane and coworkers by the reported methods6 The dose in milligrams

*Address correspondence to Craig Chemical Consulting Services, Inc., 120 Stout Rd., Ambler, Pa. 19002.

which cured 50% of the animals was obtained by extrapola- tion of the number of cures found for each of the doses tested, and was converted to the customary log l/Cvalue, where C = moles/kg test animal. For every compd, at least 3 graded doses were given to 5 animals per dose.

Results

I1 was used as input to be solved by a matrix inversion program developed by Free and coworkers.

The substituent constants (sc) whch resulted from the regression analysis are listed in Table I. The correlation coefficient for the analysis is 0.905; the standard deviation is 0.359, and the overall “average” log 1/C value is 3.39.t The

Following the reported method’ the matrix listed in Table

?This is the theoretical value for a hypothetical molecule in which none of the 6 substitutable positions contains any substituent group, including H. Substituent constants at R,, R,, R R,, and R, could be related to H as 0 by arbitrarily setting the ma& so that columns Q, U, AB, AF and AK in Table I1 are set equal to 0 , rather than 1.


Structure-activity correlations of antimalarial compounds.1. Free-Wilson analysis of 2-phenylquinoline-4-carbinols

Paul N. CraigJ. Med. Chem., 1972, 15 (2), 144-149• DOI: 10.1021/jm00272a007 • Publication Date (Web): 01 May 2002







anism for the combination of a ligand-induced (represented by I, 1’, and Iff) with a pH-induced conformational change. Detailed measurements aiming at exact values for individual constants, are now in progress.

David Wojcinsky for technical assistance with the electronic circuits and Dr. Viera Bracokova Czerlinski for assistance in the evaluation of the data. We would also like thank Dr. Hugh J. Burford, for loaning us a Beckman Type RB Dyno- graph with plug-in units.

Appendix Equation 6 with KII = 0 was previously derived.lg Eq 6

presents thus an expression of the former derivations. More details on the derivation of eq 5 and 6 are also obtainable from a monograph on chemical relaxation” (see especially Tables 8.2 and 8.3, as well as Chapter 7). Expressions for apparent rate constants for the two-step scheme of Figure 1, heavy lines of part B (and KH’ defined as in part A) are easily obtained from eq 6 by setting k3 = k4 = 0 (as well as KU 0).

References

Acknowledgment. The authors would llke to thank Mr.

(1) K. G. Brandt, P. C. Parks, G. H. Czerlinski, and G. P. Hess, J. Bio. Chem., 241,4180 (1966).

N. Sutin and D. R. Christman, J. Amer. Chem. Soc., 83, 1773 (1 96 1). B. H. Havsteen, Acta Chem. Scand., 19, 1227 (1965). G. Czerlinski, K. Dar, and G. P. Hess, Fed. Proc., Fed. Amer. Soc. Exp. Biol., 26, 674 (1967). G. Czerlinski and K. Dar, Biochim. Biophys. Acta, 234,57 (1971). A. Schejter and P. George, Biochemistry, 3, 1045 (1964). E. Margoliash and A. Schejter, Advan. Protein Chem., 21, 113 (1966). D. D. Ulmer, Biochemistry, 4 ,902 (1965). D. W. Urry and P. Doty, J. Amer. Chem. Soc., 87 , 2756 (1 965). Y. P. Myer and H. A. Harbury, Proc. Nat. Acad. Se i U.S., 54, 1391 (1965). R. Musky and P. George, Proc. Nat. Acad. Sei. US. , 56, 222 (1966). Y. P. Myer, Biochemistry, 7, 765 (1968). Y. P. Myer, J. Biol. Chem., 243, 2115 (1968). C. Greenwood and G. Palmer, ibid., 240, 3660 (1965). D. W. Urry,Proc. Nat. Acad. Sei. US., 54, 640 (1965). G. D. Watt and J. M. Sturtevant, Biochemistry, 8,4567 (1969). E. Margoliash and J. Lustgarten, J. Biol. Chem., 231, 3397 (1 964). M. Berman, in “Computers in Biomedical Research,” R. Stacy and B. Waxman, Ed., Vol2, Academic Press, New York, N. Y., 1965, Chapter VII. G. Czerlinski, J. Theor. Biol., 17, 347 (1967), eq 3.17. G. Czerlinski, “Chemical Relaxation- An Introduction toTheo- ry and Application of Stepwise Perturbation,” Marcel Dekker, Inc., New York, N. Y., 1966.

Structure-Activity Correlations of Antimalarial Compounds. 1. Free-Wilson Analysis of 2-Phenylquinoline-4-carbinols

Paul N . Craig*

Smith Kline & French Laboratories, Philadelphia, Pennsylvania 19101. Received April 23, 19 71

Sixty-nine 2-phenylquinoline-4-carbinols which had been tested in the mouse for antimalarial activity were studied by the Free-Wilson method for structure-activity correlation. The results for this study significantly support the additivity concept assumed by the Free-Wilson approach. The substituent constants for groups at the para position of the 2-Ph ring were found to correlate significantly with both Hammett’s meta u constants and Hansch’s 7~ values for those substituents. Substituents on the 7 position of the quinoline ring correlate well with para u and 71 values. Substituent constants for groups at position 8 of the quinoline ring correlate with 71 values for these substituents. Substituent constants for groups at position 6 and on the meta position of the 2-Ph ring failed to correlate with u or 71 values; the substituent constants for the 16 different aminoalkyl side chains failed to correlate with 71, or 71 and 71’. The significance of these results is discussed.

The Free-Wilson method of structure-activity correlation’ has been applied with varying degrees of success in recent

This paper reports a successful application of the technique to compare the antimalarial test results in mice for 69 substituted 2-phenylquinoline carbinols of general structure I.

H-C-OH

I

The biological test reports, obtained from the Walter Reed computer record system, gave data obtained by Rane and coworkers by the reported methods6 The dose in milligrams

*Address correspondence to Craig Chemical Consulting Services, Inc., 120 Stout Rd., Ambler, Pa. 19002.

which cured 50% of the animals was obtained by extrapola- tion of the number of cures found for each of the doses tested, and was converted to the customary log l/Cvalue, where C = moles/kg test animal. For every compd, at least 3 graded doses were given to 5 animals per dose.

Results

I1 was used as input to be solved by a matrix inversion program developed by Free and coworkers.

The substituent constants (sc) whch resulted from the regression analysis are listed in Table I. The correlation coefficient for the analysis is 0.905; the standard deviation is 0.359, and the overall “average” log 1/C value is 3.39.t The

Following the reported method’ the matrix listed in Table

?This is the theoretical value for a hypothetical molecule in which none of the 6 substitutable positions contains any substituent group, including H. Substituent constants at R,, R,, R R,, and R, could be related to H as 0 by arbitrarily setting the ma& so that columns Q, U, AB, AF and AK in Table I1 are set equal to 0 , rather than 1.

Antimalarials. Free- Wilson Analysis Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 145

F ~ta t i s t i c '~ for the correlation is 4.55; F34,34 for the 1% level is 2.4; hence the correlation is highly significant. Thus the basic assumption of the Free-Wilson method (additivity of group effects) is confirmed for these biological data.

The same antimalarial data have been analyzed by the Hansch multiple parameter method;" the results will be compared with those obtained by the Free-Wilson method in paper 2 of this series.

Discussion

Differences between the additive constants for groups at each position can be checked for statistical significance by T tests,17 which are made between each sc value (other than the reference sc) and the sc value for the reference group at that position. The sc values are listed in Table I, and the reference sc for each position is in italics. This reference group is chosen as either the most frequently studied group, or as the hydrogen analog. The absence of a T test asterisk does not signify a lack of statistical significance for the particular sc; it only means that the distinction between that sc and the reference sc is not significant. If the two sc

values are very close, it is unlikely that they will be significantly different. Also, unless the sc for a group, represented by only one or two examples, is very different from the reference sc, no significant difference should be expected.

These points are illustrated in Tables I and 11. Groups E, I, and P (identified in Table 11) gave sc values which are sig nificantly different from that of group C (the reference group), even though they were represented by only one example each. Groups F and L were represented by two examples, but because their sc values were much closer to that for group C, they were not significantly different from C. Group A is especially interesting, in that it occurred in nine examples, but its sc was so close to that for group C that it, too, was not significantly different from it. It must be em- phasized that this does not reflect upon the reliability of the sc value for group A; this can only be gauged by the overall F value and standard deviation for the entire regression.

Wilson analysis is the relative rank of substituent group constants at each position. This is summarized in Table I for

The most important information to be gained from a Free-

Table I. Substituent Constantsg

Substituent constant

Substituent (scIe na nb m - uc p - uc ERd Rl 6-Me thyl-2-piperidyl

2-Piperidyl CHZN(C4HJ 2

CH,N (i-C5H,l)2

CH,N(C8Hl 7 ) , n

CH,N(CH,) (CC,H7) CH,N(C6H1 3 2

CH,NH-CH,C$

CH,N NCH, U

CH,N(C,H,), CH,-piperid yl CH,NH(cyclopropyl) CH,N(CH,CH,OCH,CH, ), CH,N(C,H, 5) , CH,NH(l-adamantyl) CH,-morpholino

cl H

R3 OCH3

CF3 R4 I

g3

F H

CH. W H 3

0.397 0.323 0.262

-0.036 -0.077* -0.174 -0.193 -0.284

-0.43 1 * * -0.4 3 2 * * -0.45 1 * -0.573 -0.574 -0.697* -0.906** -0.973**

0.562 0.361

- 0.0554 -0.221

0.445 0.145 0.123 0.0437

-0.150 -0.203 -0.240

0.435* 0.143

-0.0641 -0.239

0.927** 0.487** 0.209

-0.168 -0.489

0.774** 0.321**

-0.0785

1 9

25 1

14 2 1 2

4

3 2 1 1 1 1 1 2 7

58 2 1 1

34 5

15 10 3 6

28 10 25

2 14 4

45 4 7

19 8

H -0.311 35 an = number of examples. bn constants for R, obtained from aliph series (ref 22); for arom substituents, n values obtd from ref 21. cu con-

stants (ref 17). d E ~ values taken from ref 20, except where noted. eone asterisk indicates significance at the 5% level. Two asterisks indicate significance at the 1% level. ~ E R values calculated by J. Rhee and C. Hansch, by CND0/2 method, private communication. gFor the overall regression, r = 0.905, s = 0.359, mp = 3.39; F = 4.55, FMaM = 2.4 for 1%.

0.88 0.38 3.3 1 1.18 5.31 4.18 2.00 7.31

-0.17

1.31 2.10 0.02 1.37 6.3 1 2.71

-0.58 0.12 0.76 0 1.16 1.26 1.16 0.70 0.15 0

-0.04 0.52 1.16 0.70

-0.04 0 1.16 0.70 0.14 0

-0.04 1.16 0.70 0.52 0

0.115 0.37 0 0.43 0.35 0.43 0.37 0.34 0 0.115

-0.07 -0.07

0.37 0.115 0 0.43 0.37 0.34 0 0.115 0.43 0.37

-0.07 0

-0.270 0.23 0 0.54 0.28 0.54 0.23 0.06 0

-0.27 -0.17 -0.17

0.23 -0.27

0 0.54 0.23 0.06 0

-0.27 0.54 0.23

-0.17 0

0.11 0.10 0 0.06f 0.12 0.06 f 0.10

0 0.11 0.03 1.16 0.10 0.11 0 0.06 f 0.10

0 0.11 0.06f 0.10 0.03 0

0.19f

0.19f


3 3 3 - 3 3 3 3 3 3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3 i i

3 3 3 i 3 -

3 3 3 3 -

3 3 3 3 3 3 3 - 3 3 3 - 3 3 -3 -

3

i 3 3 3 M i 3 3 -33 3 3 3 3 - 3 3 3.- 3 4 3 -3 - i - 3

\

Id

3 3

3 3 3 3 - 4 3 - 3

3 3 3 3 - 3

- 3 3 3 $ & 3 3-333333333 3 3 3 3 3 3 3 3 3 3 3 3 3 r id - 3 3 3 3 i i - 3

10.1

i

3

3

3

3 3

3 3 3333 3 3 3 3 3 3 3 3 3 -

3 3 3 3 3 3 43

33 3 3

3 3 3 3 3 3

Antimalarials. Free- Wilson Analysis JournaIofMedicinal Chemistry, 1972, Vol. 15, No. 2 147

* d

z 3 3 3

33 33- 33 3333 33333- 3 33v)

m

d

N

3 33 3

3 36 3

3 33 3 33 - 3 3 3 3 3 3 3 v)

d

-3 3 3

3

0 3

w , 3 3 0 0 :

N 3 33 3 3 3

33 3333 3 - 3 v ) 3 N

2 3 33 3

3 m

3

3 . r

Y 3 33 3 3 ;?;

33333

3 3

3

3 3 3333 v) 3

3

3

3 3 33

3

333 33333333333433300 v)

3 3

3 N

d 3

3

3

3

3

3 N

3

3

3 1-4

3

3 33 3 b,

3 v) N

m

333 3

3 3 3

4 3 3 m 3

148 Journal of Medicirial Chetmstry, i972, Vol. 15, No. 2 Ouig

I__ - Table Ill. Correlations Obtained by Regression - Fa, b Comments E~ua t ion P ,sa

~~

Position 1

Position 3

Position 4 (1) scc = -0.191(+0.236) + 0.978(+0.820)o-meta (2) sc = -0.041(*0.174) + 0.676(+0.635)u-para (3) sc = -0.170(+0.232) + 0.361(*0.318)n (4) sc = -0.127(+0.297) + 0.338(* 1.136)~-para + 0.220(+0.592)n (5) sc = -0.232(+0.232) + 0.626(+0.938)o-meta + 0.220(+0.353) Position 6

Position 7 (6) sc = -0.010(+0.190) + 1.811(*0.654)o-para (7) sc = -0.198(+0.397) t. 0.998(+0.651)n (8) sc = -0.457(+0.837) + 2.589(*2.787)u-meta Position 8 (9) sc = -0.395(+0.566) + 0.959(*0.780)~

0.808 0.157 9.39* 0.775 0.168 7.50* 0.794 0.162 8.53* 0.827 0.112 4.34 0.895 0.133 8.06*

0.981 0.123 77.7** 0.942 0.213 23.8* 0.863 0.322 8.75

0.966 0.151 27.9*

No significant correlations obtained between ?i or 17’ and n, and R, SL‘

No significant correlations obtained

Fl,5 = 6.61* Fl,s =6.61* Fl,s = 6.61* F,,,= 6.94* F,,,= 6.94* No significant correlations

obtained

Fl,3 = 34.1**

F, , ,= 10.1*

Fl,2 = 18.5*

aReference 17. asterisk indicates significance at the 5% level; two asterisks, at the 1% level. csc = substituent constants obtd by the regression; see Table 1.

all positions studied. However, in an attempt to gain more information from the analysis than just rank, correlations were sought by regression analysis for linear relationships between the sc values and the following parameters: Ham- mett u constants (meta and para),” Otsu’s Er constants,” and Hansch’s n values.21~22 The results are presented in Table 111.

These results point out possible linear relationships between polar effects of substituent constants in both benze- noid rings, and antimalarial activity. Also possible are relationships between the 71 values for some of the position substituents and antimalarial activity. The problem of differentiating between these possibilities when one is working with a limited set of substituents is discussed in a companion paper,23 where the judicious selection of one or two additional substituents is discussed, to help resolve which parani- eter is really the important one.

It is noteworthy that no simple dependence of antimalarial activity was found for n values of the substituent groups at R1. No correlation was found when the values for n and nz were studied; this was tried since the Hansch method has shown that such relationships are the rule, rather than the exception.

The larger coefficient for up at R7, 1.8 1, as compared to 0.98 for u,,, at R4, gives a direct comparison of the relative effect upon antimalarial activity of changes in polar effects at these two substituent positions. Thus, polar effects of substituents at R7 affect the antimalarial activity about 7 times more than do similar polar changes at R4 on the 2-Ph ring (antilog of 0.83 7). Of course it is not necessary to study the regression results in Table I11 to gain this information, since a similar conclusion is drawn from a comparison of the ranges of substituent constants at positions R7 and R4 (1.41 and 0.685, difference = 0.72, antilogg5.3), reported in Table I.

The results listed in Table I11 now allow one to predict maximum values of log l / C which might be expected for compds bearing as yet unstudied substituents. More activity would be expected from increased electron withdrawal at R7 than at any of the other aromatic ring positions. Thus, the preparation of the CF3S02 analog at R7 would be expected to result in an increased antimalarial activity of about 5-fold over the CF3 analog (a pard for CF3S02- = 0.93; for CF3 =

0.55; difference = 0.38; from eq 5, Table 111, 1.81 X 0.38 = 0.68; antilog 14.8).

The original presentation of the Free-Wilson method em- phasized the relative ranking of substituent groups at each position. The present approach allows predictions to be made outside of the matrix of compounds employed in the Free-Wilson analysis, if significant correlations with standard parameters can be obtained for the de novo sc values. Of course it is possible that too great a change in u could lead to a parabolic relationship;” hence, the prediction is to be considered as a maximum value, not necessarily fully achiev- able.

The application of eq 7, Table 111, involving 71 instead of u, would lead to a prediction that the CF3S02 analog of a CF3 member of structure I, should be approximately equivalent in activity. What is needed to resolve this problem is a substituent group such as CH3S02, which is high in u and very low in n.

Acknowledgments. This work was supported by a US. Army Research and Development Command Contract No. DADA 17-69-C-9106, and is Contribution No. 908 from the Army Research Program on Malaria. The author wishes to acknowledge helpful discussions with Dr. S. M. Free, Jr., and Dr. Corwin H. Hansch.

References (1) S. M. Free, Jr., and J. W. Wilson, J. Med. Chem., 7, 395 (1964). (2) W. P. Purcell, Biochim. Biophys. Acta, 105,201 (1965). (3) 3. G. Beasley and W. P. Purcell, ibid., 178, 175 (1969). (4) W. P. Purcell and J. M. Clayton, J. Med. Chem., 11, 199 (1968). (5) T. Ban and T. Fujita, ibid., 12, 353 (1969). ( 6 ) T. S. Osdene, P. B. Russell, and L. Rane, ibid., 10,431 (1967). (7) D. W. Boykin, Jr. , A. R. Patel, and R. E. Lutz, ibid., 11, 273

(8) W. E. Rothe and D. P. Jacobus, ibid., 11,366 (1968). (9) E. R. Atkinson and A. J. Puttick, ibid., 13,537 (1970).

(1968).

(10) A. J. Saggiomo, K. Kato, andT. Kaiya, ibid., 11, 277 (1968). (11) W. C. Campbell, J. Parisitol., 49, 824 (1963). (12) R. E. Lutz, et al., J. Amer. Chem. Soc., 68, 1813 (1946). (13) J. S. Gillespie, Jr., R. J. Rowlett, Jr., and R. E. Davis, J. Med.

(14) E. R. Atkinson and A. J. Puttick, ibid., 11, 1223 (1968). (15) F. Y. Wiselogle, Ed., “Survey of Antimalarial Drugs, 1941-

(16) M. M. Rapport, A. E. Senear, J. F. Mead, and J. B. Koepfli,

Chem., 11,425 (1968).

1945,” Edwards Bros., Ann Arbor, Mich.

Antimalarial Azaindole-3.piperidylmethanols Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 149

J. Amer. Chem. SOC., 68, 2697 (1946).

Press, Iowa City, Iowa, 1966.

(20) T. Yamamoto and T. Otsu, Chem. Ind., 787 (1967). (21) T. Fujita, J. Iwasa, and C. Hansch, J. Amer. Chem. Soc., 86,

(22) J. Iwasa, T. Fujita, and C. Hansch, J. Med. Chem., 8,150 (1965). (23) P. N. Craig, ibid., 14,680 (1971).

(17) G. W. Snedecor, "Statistical Methods," Iowa State University

(18) C. H. Hansch, Accounts Chem. Res., 2, 232 (1969). (19) H. H. Jaffe, Chem. Rev., 53, 191 (1953).

5175 (1964).

Synthesis of 1 -p-Chlorobenzyl-7-azaindole-3-cx-piperidylmethanol as a Potential Antimalarial Agentf

Anthony J. Verbiscar

Instihite of Drug Design, Sierra Madre, California 91024. Received July I , 1971

A single diastereoisomer of 1-p-chlorobenzyl-7-azaindole-3-a-piperidylmethanol was found to have antimalarial activity about 0.5 that of quinine when tested in mice against Plasmodium berghei. This is the first example of any antimalarial activity in the 7-azaindole class. None of the substituted 7-azaindole intermediates synthesized in this study showed appreciable activity. Also, the low activity of the mixed diastereoisomers indicated that one of the isomers was inactive. An intermediate in the 4-step synthesis, 7-azaindole-3-carboxaldehyde, was prepared in good yield by a Duff reaction, which represents a new and facile method for introducing the 3-formyl group on an N-unsubstituted 7-azaindole.

Very little biological activity has been published for any of the azaindoles other than 3-azaindole (benzimidazole) derivatives.' The primary interest in azaindoles has been as potential antimetabolites to naturally occurring indole derivatives. A pharmacological profile of the unsubstituted azaindoles comparing the activities of these compounds with indole, diazaindoles, and purine was reported.' Antimalarial activity in plain indoles with antimalarial type and other miscellaneous side chains has not been ~ b s e r v e d . ~ Before this study none of the reported azaindoles would be considered as candidate antimalarial target compounds. The goal of this study was to explore the potential of 7-azaindole as a new antimalarial nucleus.

The basis for exploring the 7-azaindoles rests on the ra- tionale that the structure of 1 -pchlorobenzyl-7-azaindole-3- a-piperidylmethanol is comparable to the quinoline-5-a- piperidylmethanols, and the latter compounds show antimalarial activity. In addition, 7-azaindole having a p K a = 4.59 and logP = 1.82 is isolipophilic with respect to quinoline which has a p K a = 4.95 and log P = 2.03 $

Chemistry. Unsubstituted 7-azaindole is not formylated under normal Vilsmeier reaction conditions and 7-azaindole- 3-carboxaldehyde has been made from 7-a~agramine.~$' A better method to prepare this aldehyde was discovered whereby 7-azaindole is 3-formylated with hexamethylenetetramine in refluxing aq AcOH. The yields in this reaction were consistently good and the product was pure enough for synthetic purposes. Following procedures for the plain indoles6 1 -alkylation provided 1-p-chlorobenzyl-7-azaindole- 3-carboxaldehyde plus some quaternary product.' The key aldehyde 3 was also prepared by first 1-alkylating' and then formylating under Vilsmeier reaction conditions. This experiment confirms earlier observations about Vilsmeier formylation of 7-azaindoles, i. e., that this reaction is facil- itated by 1-substitution.'

The condensation of 3 with 2-pyridyllithium gave 1-p chlorobenzyl-7-azaindole-3-a-piperidylmethanol as a labile oil which tends to decompose back to its aldehyde and

?Contribution No. 839 from the Army Research Program on Malaria; supported by the U. S. Army Medical Research and De. velopment Command Contract No. DADA1 7-69-C-9082.

$The log P of 7-azaindole and this comparison with quinoline were provided by Professor Corwin Hansch, whose assistance is gratefully acknowledged.

pyridyl components." Low pressure hydrogenation of the side chain pyridyl in its pyridinium salt form provided 1-p chlorobenzyl-7-azaindole-3-ar-piperidylmethano1 as mixed diastereoisomers. Fractional crystallization separated one of the diastereoisomers as stable microcrystals, and this isomer was tested for antimalarial activity. The second isomer was extremely hygroscopic and apparently inactive as an antimalarial agent.

Several other 3-substituted 7-azaindoles were prepared as possible intermediates with "handles" for the introduction of side chains. These included 3-bromo-7-azaindole," 3- nitro-7-azaindole," and 1 -p-chlorobenzyl-7-azagramine, Mild Mn0; ~ x i d a t i o n ' ~ of the aldehyde 3 gave 1 pchlorobenzyl-3-carboxy-7-azaindole. Only starting material was recovered when the carboxylic acid 9 was treated with 2-PyLi in an attempt to prepare the ketone.I4 n-BuLi is known to metalate indoles on the 2 position" and the reactive 2-PyLi l6

may have formed a 2-lithio-7-azaindole in our reaction sys-

Table I. Antimalarial Activity and Toxicity of 7-Azaindoles in Mice"

Compd Dose, mg/kg IMST, daysb Toxic deathsC

1 2 3 4 5 6 8

8a 9

10

11 12

640 640 640 640 640 640

10 40

160 320 640 640 640 160 320 640 640 640

0.8 0.7 0.5 0.5 0.3 0.3 0.3 0.5 0.7 3.1 3.9 0.5 0.7 0.3 0.9

1.0 0.6

-

0 0 0 0 0 0 0 0 0 0 2 0 0 0 3 5 0 0

"Test data supplied by Walter Reed Army Institute of Research. h x e a s e in mean survival time between Plasmodium berghei infected mice administered the drug sc and controls. Non-drug- treated mice survived about 6 days following infection with P. berghei. Chemical therapy was initiated 3 days postinfection. For details of this test see T. S. Osdene, et al. *' CNumber of deaths/5 mice.


Synthesis of 1-p-chlorobenzyl-7-azaindole-3-.alpha.-piperidylmethanolas a potential antimalarial agent

Anthony J. VerbiscarJ. Med. Chem., 1972, 15 (2), 149-152• DOI: 10.1021/jm00272a008 • Publication Date (Web): 01 May 2002






Antimalarial Azaindole-3.piperidylmethanols Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 149

J. Amer. Chem. SOC., 68, 2697 (1946).

Press, Iowa City, Iowa, 1966.

(20) T. Yamamoto and T. Otsu, Chem. Ind., 787 (1967). (21) T. Fujita, J. Iwasa, and C. Hansch, J. Amer. Chem. Soc., 86,

(22) J. Iwasa, T. Fujita, and C. Hansch, J. Med. Chem., 8,150 (1965). (23) P. N. Craig, ibid., 14,680 (1971).

(17) G. W. Snedecor, "Statistical Methods," Iowa State University

(18) C. H. Hansch, Accounts Chem. Res., 2, 232 (1969). (19) H. H. Jaffe, Chem. Rev., 53, 191 (1953).

5175 (1964).

Synthesis of 1 -p-Chlorobenzyl-7-azaindole-3-cx-piperidylmethanol as a Potential Antimalarial Agentf

Anthony J. Verbiscar

Instihite of Drug Design, Sierra Madre, California 91024. Received July I , 1971

A single diastereoisomer of 1-p-chlorobenzyl-7-azaindole-3-a-piperidylmethanol was found to have antimalarial activity about 0.5 that of quinine when tested in mice against Plasmodium berghei. This is the first example of any antimalarial activity in the 7-azaindole class. None of the substituted 7-azaindole intermediates synthesized in this study showed appreciable activity. Also, the low activity of the mixed diastereoisomers indicated that one of the isomers was inactive. An intermediate in the 4-step synthesis, 7-azaindole-3-carboxaldehyde, was prepared in good yield by a Duff reaction, which represents a new and facile method for introducing the 3-formyl group on an N-unsubstituted 7-azaindole.

Very little biological activity has been published for any of the azaindoles other than 3-azaindole (benzimidazole) derivatives.' The primary interest in azaindoles has been as potential antimetabolites to naturally occurring indole derivatives. A pharmacological profile of the unsubstituted azaindoles comparing the activities of these compounds with indole, diazaindoles, and purine was reported.' Antimalarial activity in plain indoles with antimalarial type and other miscellaneous side chains has not been ~ b s e r v e d . ~ Before this study none of the reported azaindoles would be considered as candidate antimalarial target compounds. The goal of this study was to explore the potential of 7-azaindole as a new antimalarial nucleus.

The basis for exploring the 7-azaindoles rests on the ra- tionale that the structure of 1 -pchlorobenzyl-7-azaindole-3- a-piperidylmethanol is comparable to the quinoline-5-a- piperidylmethanols, and the latter compounds show antimalarial activity. In addition, 7-azaindole having a p K a = 4.59 and logP = 1.82 is isolipophilic with respect to quinoline which has a p K a = 4.95 and log P = 2.03 $

Chemistry. Unsubstituted 7-azaindole is not formylated under normal Vilsmeier reaction conditions and 7-azaindole- 3-carboxaldehyde has been made from 7-a~agramine.~$' A better method to prepare this aldehyde was discovered whereby 7-azaindole is 3-formylated with hexamethylenetetramine in refluxing aq AcOH. The yields in this reaction were consistently good and the product was pure enough for synthetic purposes. Following procedures for the plain indoles6 1 -alkylation provided 1-p-chlorobenzyl-7-azaindole- 3-carboxaldehyde plus some quaternary product.' The key aldehyde 3 was also prepared by first 1-alkylating' and then formylating under Vilsmeier reaction conditions. This experiment confirms earlier observations about Vilsmeier formylation of 7-azaindoles, i. e., that this reaction is facil- itated by 1-substitution.'

The condensation of 3 with 2-pyridyllithium gave 1-p chlorobenzyl-7-azaindole-3-a-piperidylmethanol as a labile oil which tends to decompose back to its aldehyde and

?Contribution No. 839 from the Army Research Program on Malaria; supported by the U. S. Army Medical Research and De. velopment Command Contract No. DADA1 7-69-C-9082.

$The log P of 7-azaindole and this comparison with quinoline were provided by Professor Corwin Hansch, whose assistance is gratefully acknowledged.

pyridyl components." Low pressure hydrogenation of the side chain pyridyl in its pyridinium salt form provided 1-p chlorobenzyl-7-azaindole-3-ar-piperidylmethano1 as mixed diastereoisomers. Fractional crystallization separated one of the diastereoisomers as stable microcrystals, and this isomer was tested for antimalarial activity. The second isomer was extremely hygroscopic and apparently inactive as an antimalarial agent.

Several other 3-substituted 7-azaindoles were prepared as possible intermediates with "handles" for the introduction of side chains. These included 3-bromo-7-azaindole," 3- nitro-7-azaindole," and 1 -p-chlorobenzyl-7-azagramine, Mild Mn0; ~ x i d a t i o n ' ~ of the aldehyde 3 gave 1 pchlorobenzyl-3-carboxy-7-azaindole. Only starting material was recovered when the carboxylic acid 9 was treated with 2-PyLi in an attempt to prepare the ketone.I4 n-BuLi is known to metalate indoles on the 2 position" and the reactive 2-PyLi l6

may have formed a 2-lithio-7-azaindole in our reaction sys-

Table I. Antimalarial Activity and Toxicity of 7-Azaindoles in Mice"

Compd Dose, mg/kg IMST, daysb Toxic deathsC

1 2 3 4 5 6 8

8a 9

10

11 12

640 640 640 640 640 640

10 40

160 320 640 640 640 160 320 640 640 640

0.8 0.7 0.5 0.5 0.3 0.3 0.3 0.5 0.7 3.1 3.9 0.5 0.7 0.3 0.9

1.0 0.6

-

0 0 0 0 0 0 0 0 0 0 2 0 0 0 3 5 0 0

"Test data supplied by Walter Reed Army Institute of Research. h x e a s e in mean survival time between Plasmodium berghei infected mice administered the drug sc and controls. Non-drug- treated mice survived about 6 days following infection with P. berghei. Chemical therapy was initiated 3 days postinfection. For details of this test see T. S. Osdene, et al. *' CNumber of deaths/5 mice.

150 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Verb iscar

Scheme I

H 1

1 la, L-tartrate

/ I

R R 5 6

$H,NMe, \

H \ I X

H

11, X = NO, 12, X = Br

R R 4

I COOH

R 9

CHO

-N' equiv R R R

8 7 8a, dibenzoyl-d- tartrate

tem. This would decrease the probability for forming a di- anion at the COzH function which appears to be an intermediate in the condensation sequence leading to the ketone. § Also, simple carbcxylic acids have failed to form ketones with MeLi." Attempts to prepare 7-azaindole-3-carbinole- amines by treating 3 with dimethylsulfonium methylide to form an oxirane intermediate" were not successful.

tested for antimalarial activity against Plasmodium berghei in mice. At 320 mg/kg the single crystalline diastereoisomer of 1 -p-chlorobenzyl-7-azaindole-3a-piperidylmethanol(8) increased mean survival time by 3.1 days. This is about 0.5 the activity of quinine sulfate which at 160 mg/kg has an IMST of 3.6 days. This is, therefore, the first example of a 7-azaindole derivative with antimalarial activity. The mixed diastereoisomers of this structure were tested as the dibenzoyl-d-tartrate salt 8a, which had an IMST of 0.5 days at 640 mg/kg indicating that the second diastereoisomer is inactive. None of the other 7-azaindoles had an IMST > 1 .O day at 640 mg/kg. Except for 8 and 10 they were also not noticeably toxic at this dose level. This lack of toxicity is notable because several of the other azaindoles are quite toxic. For example, the LDS0 in mice for 5-azaindole is 16.5 mg/kg, and for 6-azaindole it is 12 mg/kg, compared to 490 mg/kg for 7-azaindole and 3 16 mg/kg for indole.*

Biological Results. All of the 7-azaindoles, 1-12, were

§Private communication f rom Professor R. E. Lutz.

Compds 1-4, 9, and 1 1 were tested vs. P. gallinaceum a t 120 mg/kg administered sc in chicks and found to be inactive. There were no toxic deaths noted for these 6 compounds.

In the mosquito screen vs. P. gallinaceum at a dose level of 0.1%, suppression of oocysts was registered by 4 (75%), 11 (50%), and 12 (50%). Compd 2 (86% deaths) was classified as toxic to the mosquitoes. Compds 1 and 9 were inactive.

ulant and possible antidepressant activity.# Compd 8 has definite CNS effects in mice showing stim-

Experimental Section** 7-Azaindole was prepd by a known methodlg in an overall

yield of 46%. A mono-L-tartrate salt formed and was recrystd from EtOH as white needles, mp 175-177", which were sol in H,O. Anal. (C,H,N,. C,H,O,) C , H, N

7-azaindole and 42 g (0.30 mole) of hexamethylenetetramine was refluxed with stirring for 6 hr in 250 ml of 33% AcOH (84 g, 1.40 mole, of AcOH and 168 ml of H,O). The resulting clear yellow s o h was dild with 500 ml of H,O, and the product was allowed to cryst in the refrigerator overnight. Recrystn of the crude product from H,O gave 14.9 g (50%) of long white needles, mp 216-218" (re-

7-Azaindole-3-carboxaldehyde. A s o h of 23.6 g (0.20 mole) of

#These effects were noted by Professor Leo Abood whose assist-

**Melting points are corrected. Where analyses are indicated only ance is gratefully acknowledged.

by symbols of the elements, analytical results obtained for those elements are within +0.40/0 of the theoretical value.

Antimalarial Azaindole-3-piperidylmethanols Journal of Medicinal Chemistry, 19 72, Vol. 15, No. 2 15 1

ported5 214.5-215'). The product is insol in dil HCl and can be crystd from this solvent. I t is sol in dil base. It formed a phenyl- hydrazone as yellow plates, mp 230-233' dec (reported5 mp 231- 232.5" dec).

reagent and solvent ratios and with quantities of 7-azaindole as small as 0.35 g. Crude yields were consistently in the range of 51-63%. However, a shorter reflux time (3 hr) and gradual addn (30 min) of hexamethylenetetramine lowered the yield to 12.5%.

1-p-Chlorobenzyl-7-aaindole. Following known procedures, * 6.47 g (0.16 mole) of 57% N& in oil was washed with petr ether and the last traces of solvent were evapd in vacuo. The flask was then cooled in an ice-n-PrOH bath and 150 ml of DMF was added, followed by a s o h of 17.7 g (0.15 mole) of 7-azaindole in 100 ml of DMF over 15 min with stirring. When the evoln of H, stopped, a soln of 24.2 g (0.15 mole) of p-chlorobenzyl chloride in 75 ml of DMF was added over 10 min and the mixt was stirred a t room temp for 3 hr. Most of the DMF was distd off in vacuo. H,O was added to the residue and the dark oil was extd into Et,O. A tlc monitor indicated the presence of 2 products. An extn of the Et,O soln with dil HOAc cleanly removed 7-p-chlorobenzyl-7H-pyrrolo- [ 2,341pyridine. The Et,O soln contg the desired product was dried (Na,SO,), the solvent evapd, and the residue distd in vacuo giving 31.1 g of crude yellow oil, bp 157-160" (0.5 mm). The oil crystd from 60-110" petr ether resulting in a 29.7 g (81%) yield of l-p- chlorobenzyl-7-azaindole, mp 35-37', as long white needles. It tends to decompose on standing for 1 month. It is insol in dil mineral acid but forms an HBr salt, mp 178-181.5", which oils out in H,O. The oil can be extd ifito Et,O. It seems likely that steric hindrance by the bulky p-chlorobenzyl group interferes with salt formation and acid solubility of this compd. Anal. (C,,H1,N,Cl) C, H, N

7-p-Chlorobenzyl-7H-pyrrolo[ 2,341pyridine. The HOAc ext from the prepn of 1-p-chlorobenzyl-7-azaindole was basified to yield 1.82 g of a yellow solid. Crystn from CC1,-petr ether gave 1.62 g (4.5%) of small yellow needles, mp 120-122". Anal. (C,,H,,N,Cl) c, H, N

l-p-Chlorobenzyl-7-azaindole-3-carboxaldehyde. Method A. To 14 ml of DMF cooled to 0" there was added over 3 min with stirring 6.95 g (0.045 mole) of POCl,. The soln was stirred outside the bath for 10 min and then a soln of 10 g (0.041 mole) of l-p-chlorobenzyl- 7-azaindole in 10 ml of DMF was added over 5 min as the temp remained below 20". The clear soln was warmed at 40-55" for 75 min and then cooled, and 70 g of ice water were added as a solid pptd. A soln of 21.6 g (0.56 mole) of NaOH in 60 ml of H,O was added and the mixt was brought to boiling for 3 min. On cooling the suspended oil crystd. It was collected, washed well with H,O, dried, and crystd from CCl, giving a 9.94 g (88%) yield of pure white needles, mp

Method B. A mixt of 10 g (0.0685 mole) of 7-azaindole-3-

The reaction as reported here was run 7 times with variations in

104-1 06".

carboxaldehyde, 11.6 g (0.072 mole) of p-chlorobenzyl chloride, and 11.3 g (0.082 mole) of K,CO, in 500 ml of Me,CO was stirred under reflux for 10 hr. After this time a tlc monitor showed that the starting aldehyde was consumed. The Me,CO was evapd and the residue was taken up in H,O and EtOAc. The EtOAc phase was extd with 200 ml of 1 N HCl in 3 portions to remove 7-pchlorobenzyl-7H-pyrrolo [ 2,341 pyridine-3-carboxaldehyde as described below. The EtOAc phase was dried (Na,SO,), and the solvent was evpd leaving a solid. This was crystd from CCl, to yield 15 g (81%) of fine white needles, mp 105-107'. I t is insol in acid probably due to both steric hindrance of the p-chlorobenzyl stibstituent and the electron-withdrawing effect of the 3-CHO which lowers basicity of the nucleus. An analytical sample was crystd to mp 107-108°. Anal. (C,,H,,N,CIO) C, H, N

7-p-Chlorobenzyl-7H-pyrrolo [ 2,341 pyridine-3-carboxaldehyde. The HCl extract of the above reaction mixt was made basic to free 3.11 g (16.7%) of a solid. This was crystd twice from EtOH to give 1.82 g of fine yellow needles, mp 142-144". This product is colorless in aq acid in which it is highly soluble. Anal, (C,,H,,N,ClO) C, H, N

(0.050 mole) of l-p-chlorobenzyl-7-azaindole-3-carboxaldehyde in 300 ml of Me,CO there was added over 15 min with good stirring a soln of 15.8 g (0.10 mole) of KMnO, in 220 ml of H,O. During this addn the temp of the reaction mixt was kept at 35-40'' using an ice bath. Stirring was then contd for another 45 min outside the bath. The mixt was filtered and the ppt was washed well with 1: 1 Me,CO- H,O. The solvent was partially removed on a rotary evaporator as an oil appeared, which was extd into Et,O. The basic aq layer was acidified with HOAc to ppt a solid. This product was crystd from

l-p-Chlorobenzyl-3-carboxy-7-azaindole. To a s o h of 13.6 g

EtOH giving 7.71 g (54%) of tan needles, mp 200-204'. Anal. (C,,H,lN,ClOz) C, H, N

1-p-Chlorobenzyl-7-aagramine Hydrochloride. A soln of 2.43 g (0.010 mole) of l-p-chlorobenzyl-7-azaindole, 0.88 g (0.016 mole) of Me,NH,*Cl-, 0.33 g (0.011 mole) of paraformaldehyde, and 45 ml of n-BuOH was refluxed for 1 hr. The solvent was evpd leaving a solid which oiled out upon addn of 50 ml of H,O. Following the addn of 1 ml of concd HCl the undissolved oil was extd into Et,O and worked up to give 1.2 g (50% recovery) of l-p-chlorobenzyl-7- azaindole. The aq acidic portion was made basic with K,CO, to free 1.24 g of a second oil which could not be induced to cryst. Addn of 1 equiv of HCl in EtOH to this oil and removal of the solvent provided a cryst salt. Crystn from EtCOMe gave 0.90 g (30%) of fine white needles, mp 202-205" dec. The dihydrochloride was hygroscopic. No attempt was made to improve the yield. Anal. (C,,HiJ'J,Clz) C, H, C1

l-p-Chlorobenzyl-7-aaindole-3-a-pyridylmethanoL A 2-PyLi reagent was generated in the usual manner in 60 ml of Et,O using 8.9 ml(O.020 mole) of 2.25 M BuLi in hexane and 3.16 g (0.020 mole) of 2-b~omopyridine.'~ After stirring the reagent at -60" for 30 min a soln of 3.52 g (0.013 mole) of 1-p-chlorobenzyl-7-azaindole- 3-carboxaldehyde in 20 ml of THF was added over 4 min as the soln became a clear orange. I t was stirred for 1 hr at -60', then allowed to come to room temp during another hour of stirring, The mixt was dild with EtOAc, washed with H,O, three 50-ml portions of 3% HOAc, Na,CO, soln, and H,O again, and then dried (Na,SO,). A tlc check at this point showed a very strong spot corresponding to about 90% of product plus about 5% of starting aldehyde as the major impurity. Evapn of the solvent gave 5 g of the carbinol as a glass that could not be induced to cryst. The oil was sol in Et,O, CCl,, and PhMe but it was insol in i-Pr,O. It was sol in dil HC1 where it gradually decompd, apparently back to the aldehyde and pyridine (tlc monitor). The hydrochlorides of the oil were cryst but extremely hygroscopic. An acetate did not cryst. Partial identification was obtd by ir analysis of the carbinol: in CHCl,: 2.94 (OH), 6.22 and 6.66 p (arom), 6.32, 6.45 p (7-azaindoles); and the a c e tate;in CHCl,: 5.74 (acetate), 6.22,6.67,6.32, 6.47 p; tlc, Merck silica gel G, 95:5 PhH-MeOH developer, carbinol Rf 0.43; acetate Rf 0.71.

enough 2-PyLi reagent must be generated to react with the aldehyde function. Excess 2-PyLi appears to react elsewhere in the 7-azaindole nucleus. The pyridylcarbinol was taken on directly to the hydrogenation step.

l-p-Chlorobenzyl-7-azaindole-3-a-piperidylmethanoL A soln of 5 g (0.013 mole) of crude l-p-chlorobenzyl-7-azaindole-3-a-pyridylmethanol in 125 ml of EtOH contg 1 molar equiv of HCl was shaken with 300 mg of PtO, under H, at 2.9 kg/cmz until 3 equiv of H, were absorbed. After removing the solvent the resulting product was taken up in 150 ml of H,O and extd 3 times with EtOAc using NH,Cl to break an emulsion. The clear light orange aq soln of acid solubles was basified to release an oil, which was extd into EtOAc and dried (NaSO,). A tlc of this soln on Merck silica gel G, 4: 1 EtCOMe-MeOH developer, showed 2 spots of about equal intensity at Rf 0.07 and 0.10, corresponding to the two diastereoisomers. After removing the EtOAc from the oil, dry Et,O was added and the slower moving diastereoisomer eventually crystd from the soln as a powder. It was crystd from PhMe to give 1.14 g (25%) of colorless microcrystals, mp 151-156'. It was homogeneous on tlc using Eastman chromagram 6060 sheets with Me,CO and also 3: 1 Me,CO-MeOH as the developer. A sample was crystd tomp 154-156" for elemental analysis. Anal. (C,,H,,N,ClO) C, H, N, C1

The mono- and dihydrochlorides were too hygroscopic to serve as derivs, but a suitable dibenzoyl-d-tartrate salt of the mixed diastereoisomers was formed.

l-p-Chlorobenzyl-7-aaindole-3-a-piperidylmethanol Dibenzoyl- d-tartrate. 1-p-Chlorobenzyl-7-azaindole-3-a-pyridylmethanol (6.7 g) was hydrogenated as in the previous expt. This pyridylcarbinol was prepd using a 2.5: 1 mole ratio of 2-PyLi to aldehyde and contd several major impurities due to excess reagent. Its punty was about 60%. After hydrogenation and work-up, tlc showed the resulting product contd the double spot corresponding to the two diastereoisomers plus several impurities. Neither diastereoisomer could be induced to cryst from this mixt. Eventually a dibenzoyl- d-tartrate salt was formed in Et,O providing 3.28 g of white powder, mp 130-145'. Crystn attempts were generally unsuccessful but the product seemed to be a reasonably pure mixt of the two diastereoisomers. Tlc on Eastman chromagram sheet 6060 using 3: 1 Me$@ MeOH developer showed 2 spots at R f 0.20 and 0.32. Anal.

A critical condn in the prepn of this pyridylcarbinol is that just

(C,,H,,N,ClO~C,,H,,O,) C,-63.90; H: 5.08; N, 5.88. Found: C, 64.11; H, 5.53;N, 5.56.

152 Journal of Medicinal Chemistry, 1972, VoE. 15, No. 2 Heman-Ackah and Garrett

Acknowledgment. We are grateful t o Drs. E. A. Steck and R. E. Strube of Walter Reed Army Institute of Research for suggestions and encouragement in this work.

References

(1) (a) R. E. Willette, Advan. Heterocycl. Chem., 9, 27 (1968); (b) L. N. Yakhontov, Russ. Chem. Rev., 37,551 (1968).

(2) T. K. Adler and A. Albert, J. Med. Chem., 6,480 (1963). (3) (a) F. Y. Wiselogle, “A Survey of Antimalarial Drugs,” Vol. 11,

Part 1, J. W. Edward Publishers, Ann Arbor, Mich., 1946, pp. 906-916; (b) M. Julia and P. Nickel, Medd. Norsk. Farm. Selsk., 28, 153 (1966); Chem. Abstr., 66,8016 (1967).

(4) W. R. N. Williamson,J. Chem. Soc., 2833 (1962). (5) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 77,

(6) (a) W. B. Whalley, J. Chem. Soc., 1651 (1954);(b) A. H. 457 (1955).

Jackson and A. E. Smith, ibid., 5516 (1964); (c) H. Pleininger, Chem. Ber., 87, 127 (1954).

(7) (a) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 77, 6554 (1955); (b) J. P. Saxena,Indian J. Chem., 5, 73 (1967).

(8) (a) M. M. Robison, F. P. Butler and B. L. Robison, J . Amer. Chem. Soc., 79,2573 (1957). (b) Merck and Co., Netherlands Application 6,510,648 (1966); Chem. Abstr., 65, 13711 (1966).

(9) (a) L. N. Yakhontov and M. V. Rubtsov, Zh. Obshch. Khim.,

34,2603 (1964); J. Gen. Chem. USSR, 34,2626 (1964); (b) L. N. Yakhontov, M. Ya. Uritskaya, and M. V, Rubtsov, Zhur. Org. Khim., 1,2032 (1965); (c) M. V. Rubtsov and L. N. Yakhontov, Authors Certificate USSR No. 162535 (1964): Byull. Izobret., No. 10, 1964.

(10) H. Bader and W. Oroshnik, J. Amer. Chem. SOC., 79, 5686 (1957); ibid., 81, 163 61959).

(11) M. M. Robison and B. L. Robison, ibid., 70, 1247 (1956). (12) M. M. Robison, B. L. Robison, and F. P. Butler, ibid., 81. 743

(1959). (13) R. V. Heinzelman, W. C. Anthony, D. A. Lyttle, and J.

Szmuszkovicz, J. Org. Chem.. 25, 1553 (1960). (14) D. W. Boykin, A. R. Patel, R. E. Lutz, and A. Burger, J. geter-

ocycl. Chem., 4,459 (1967). (15) D. A. Shirley and P. A. Roussel, J. Amer. Chem. SOC.. 75, 375

(1953). (16) J . P. Wibaut, A. P. De Jonge, H. G. P. Van der Voort, and P.

Ph. H. L. Otto, Red. Trav. Chim. Pays-Bas, 70, 1054 (1951). (17) C. Tegner, Acta Chem. Scand., 6 , 7 8 2 (1952). (18) (a) E. J. Corey and M. Chaykovsky, J. Amer. Chem. Soc., 87,

1353 (1965); (b) W. G. Duncan, W. T. Colwell, C. R. Scott. and D. W. Henry,J. Med. Chem., 11, 1221 (1968).

(19) R. R. Lorenz, B. F. Tullar, C. F. Koelsch, and S. Archer, J Org. Chem., 30, 2531 (1965).

(20) T. S. Osdene, P. B. Russel, and Leo Rane, J. Med. Ghein.~ IO, 431 (1967).

Kinetics and Mechanisms of Drug Action on Microorganisms. 13. Comparative Studies on Action of Lincomycin, Clindamycin, and U 24729A against Escherichia coli

Samuel M. Heman-Ackah and Edward R. Garret t*

The Beehive, College of Pharmacy, University of Florida, Gainesville, Florida 32601. Received June 11 , I971

Escherichia coli cultures in bro th affected by sub-completely-inhibitory concns of l incomycin ( I ) exhibit 2 phases of steady-state generation while those affected b y t h e 7( S)-halogenated compounds, clindamycin (11) and U 24729A (111), have only one phase of steady-state generation. The generation rate constants of 11- or 111-affected cultures have the same functional dependencies on drug concentrations as lincomycin- affected cultures in phase I bu t are different f rom those of I in phase I1 generation. The possibility that lincomycin blocks 2 separate sites in a metabolic sequence whereas the 7(S)-halogenated compds block only the latter in the sequential biochemical process, can rationalize these phenomena. I1 is 6.6 times and I11 is 28.5 times as active as I calcd on the basis of molar equivalency. The effects of changes in pH o n the activity of I, 11, and I11 against E. coli suggest t h a t the unprotonated fraction of t h e drug concn contributes t o the activity. The combined actions of I1 and I11 are not antagonistic at any level of activity and can be quantitatively predicted from the separate equivalent dose-response curves of either drug. However, combinations of I1 o r 111 with I show antagonistic effects which depend o n the level of activity. These can be rationalized by the one (e.g., l incomycin) allosterically modifying the receptor site for the other lincosaminide antibiotic.

Lincomycin (I) is an antibiotic produced by Streptomyces Zincolnensis. ‘9’ It has an antibacterial spectrum similar t o tha t of erythromycin and was claimed t o be ~ u p e r i o r ~ > ~ because o f its effectiveness against b o t h erythromycin-susceptible and -resistant strains of Gram-positive coccal organisms. However, i t was subsequently t h a t there was a “dissociated type” of cross-resistance in Staphylococcus aureus between erythromycin and lincomycin. Lincomycin was therefore modified’ chemically to serve as a basis for t h e understanding of structure-activity relations. Thus, analogs of enhanced potency and broadened antibacterial spectrum were prepared.

The main structural effects claimed for in vitro activity’ were (a) the variation o f the alkyl subst i tuent a t the N‘ a tom o f t h e pyrrolidine nucleus w h c h changed the antibacterial spectrum of activity; (b) the increase in size of t h e alkyl group a t the C-4’ i n t h e pyrrolidine nucleus which in-

creased lipophilicity and the activity; (c) halogen substitution of the 7-(S) configuration of l incomycin molecule which potentiated antibacterial effects; and (d) the need of

CH I R

I, lincomycin 11. clindamycin I1 U24729A [ 7(S)-chloro-7- deoxylincomycin] propyl-rt’(R) and (S )

R2 = CsH, Ri CHs n-pentylclindamycirl Xi = H Ra C,H, R, = H

[ 1 ’-demethyl-rt ‘-de- Ri = CHj

Xi OH

XL -- H x* = c1 R, = C5H,, xj = kl x , =CI


Kinetics and mechanisms of drug action on microorganisms.13. Comparative studies on action of lincomycin,

clindamycin, and U 24729A against Escherichia coliSamuel M. Heman-Ackah, and Edward R. Garrett







152 Journal of Medicinal Chemistry, 1972, VoE. 15, No. 2 Heman-Ackah and Garrett

Acknowledgment. We are grateful t o Drs. E. A. Steck and R. E. Strube of Walter Reed Army Institute of Research for suggestions and encouragement in this work.

References

(1) (a) R. E. Willette, Advan. Heterocycl. Chem., 9, 27 (1968); (b) L. N. Yakhontov, Russ. Chem. Rev., 37,551 (1968).

(2) T. K. Adler and A. Albert, J. Med. Chem., 6,480 (1963). (3) (a) F. Y. Wiselogle, “A Survey of Antimalarial Drugs,” Vol. 11,

Part 1, J. W. Edward Publishers, Ann Arbor, Mich., 1946, pp. 906-916; (b) M. Julia and P. Nickel, Medd. Norsk. Farm. Selsk., 28, 153 (1966); Chem. Abstr., 66,8016 (1967).

(4) W. R. N. Williamson,J. Chem. Soc., 2833 (1962). (5) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 77,

(6) (a) W. B. Whalley, J. Chem. Soc., 1651 (1954);(b) A. H. 457 (1955).

Jackson and A. E. Smith, ibid., 5516 (1964); (c) H. Pleininger, Chem. Ber., 87, 127 (1954).

(7) (a) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 77, 6554 (1955); (b) J. P. Saxena,Indian J. Chem., 5, 73 (1967).

(8) (a) M. M. Robison, F. P. Butler and B. L. Robison, J . Amer. Chem. Soc., 79,2573 (1957). (b) Merck and Co., Netherlands Application 6,510,648 (1966); Chem. Abstr., 65, 13711 (1966).

(9) (a) L. N. Yakhontov and M. V. Rubtsov, Zh. Obshch. Khim.,

34,2603 (1964); J. Gen. Chem. USSR, 34,2626 (1964); (b) L. N. Yakhontov, M. Ya. Uritskaya, and M. V, Rubtsov, Zhur. Org. Khim., 1,2032 (1965); (c) M. V. Rubtsov and L. N. Yakhontov, Authors Certificate USSR No. 162535 (1964): Byull. Izobret., No. 10, 1964.

(10) H. Bader and W. Oroshnik, J. Amer. Chem. SOC., 79, 5686 (1957); ibid., 81, 163 61959).

(11) M. M. Robison and B. L. Robison, ibid., 70, 1247 (1956). (12) M. M. Robison, B. L. Robison, and F. P. Butler, ibid., 81. 743

(1959). (13) R. V. Heinzelman, W. C. Anthony, D. A. Lyttle, and J.

Szmuszkovicz, J. Org. Chem.. 25, 1553 (1960). (14) D. W. Boykin, A. R. Patel, R. E. Lutz, and A. Burger, J. geter-

ocycl. Chem., 4,459 (1967). (15) D. A. Shirley and P. A. Roussel, J. Amer. Chem. SOC.. 75, 375

(1953). (16) J . P. Wibaut, A. P. De Jonge, H. G. P. Van der Voort, and P.

Ph. H. L. Otto, Red. Trav. Chim. Pays-Bas, 70, 1054 (1951). (17) C. Tegner, Acta Chem. Scand., 6,782 (1952). (18) (a) E. J. Corey and M. Chaykovsky, J. Amer. Chem. Soc., 87,

1353 (1965); (b) W. G. Duncan, W. T. Colwell, C. R. Scott. and D. W. Henry,J. Med. Chem., 11, 1221 (1968).

(19) R. R. Lorenz, B. F. Tullar, C. F. Koelsch, and S. Archer, J Org. Chem., 30, 2531 (1965).

(20) T. S. Osdene, P. B. Russel, and Leo Rane, J. Med. Ghein.~ IO, 431 (1967).

Kinetics and Mechanisms of Drug Action on Microorganisms. 13. Comparative Studies on Action of Lincomycin, Clindamycin, and U 24729A against Escherichia coli

Samuel M. Heman-Ackah and Edward R. Garret t*

The Beehive, College of Pharmacy, University of Florida, Gainesville, Florida 32601. Received June 11 , I971

Escherichia coli cultures in bro th affected by sub-completely-inhibitory concns of l incomycin ( I ) exhibit 2 phases of steady-state generation while those affected b y t h e 7( S)-halogenated compounds, clindamycin (11) and U 24729A (111), have only one phase of steady-state generation. The generation rate constants of 11- or 111-affected cultures have the same functional dependencies on drug concentrations as lincomycin- affected cultures in phase I bu t are different f rom those of I in phase I1 generation. The possibility that lincomycin blocks 2 separate sites in a metabolic sequence whereas the 7(S)-halogenated compds block only the latter in the sequential biochemical process, can rationalize these phenomena. I1 is 6.6 times and I11 is 28.5 times as active as I calcd on the basis of molar equivalency. The effects of changes in pH o n the activity of I, 11, and I11 against E. coli suggest t h a t the unprotonated fraction of t h e drug concn contributes t o the activity. The combined actions of I1 and I11 are not antagonistic at any level of activity and can be quantitatively predicted from the separate equivalent dose-response curves of either drug. However, combinations of I1 o r 111 with I show antagonistic effects which depend o n the level of activity. These can be rationalized by the one (e.g., l incomycin) allosterically modifying the receptor site for the other lincosaminide antibiotic.

Lincomycin (I) is an antibiotic produced by Streptomyces Zincolnensis. ‘9’ It has an antibacterial spectrum similar t o tha t of erythromycin and was claimed t o be ~ u p e r i o r ~ > ~ because o f its effectiveness against b o t h erythromycin-susceptible and -resistant strains of Gram-positive coccal organisms. However, i t was subsequently t h a t there was a “dissociated type” of cross-resistance in Staphylococcus aureus between erythromycin and lincomycin. Lincomycin was therefore modified’ chemically to serve as a basis for t h e understanding of structure-activity relations. Thus, analogs of enhanced potency and broadened antibacterial spectrum were prepared.

The main structural effects claimed for in vitro activity’ were (a) the variation o f the alkyl subst i tuent a t the N‘ a tom o f t h e pyrrolidine nucleus w h c h changed the antibacterial spectrum of activity; (b) the increase in size of t h e alkyl group a t the C-4’ i n t h e pyrrolidine nucleus which in-

creased lipophilicity and the activity; (c) halogen substitution of the 7-(S) configuration of l incomycin molecule which potentiated antibacterial effects; and (d) the need of

CH I R

I, lincomycin 11. clindamycin I1 U24729A [ 7(S)-chloro-7- deoxylincomycin] propyl-rt’(R) and (S )

R2 = CsH, Ri CHs n-pentylclindamycirl Xi = H Ra C,H, R, = H

[ 1 ’-demethyl-rt ‘-de- Ri = CHj

Xi OH

XL -- H x* = c1 R, = C5H,, xj = kl x , =CI

Lincomycin Antibiotics Journal of Medicinal Chemism, 1972, Vol. 15, No. 2 153

maintaining the LY configuration of the thioglycoside to maximize antibacterial activity. Clindamycin (11) and U 24729A (111) are 7(S)-C1 analogs which are claimed to be more than 4 times as active as the parent antibiotic (I)* against a variety of Gram-positive and Gram-negative organisms.'~~

Lincomycin is an inhibitor of bacterial protein synthesis.' The sequence of events occurring during protein synthesis" and the steps which might be inhibited by antibiotics have been discussed by Cundliffe and McQuillen." I has been reported to inhibit peptide bond formation by competitive binding on a ribosomal site, possibly peptidyl "transferase" which appears to be an integral part of the 50s ribosomal subunit and is also the binding site for the amino acyl-

have claimed that I binds to "translocase," a factor which promotes the move- ment of a peptidyl tRNA elongated by a single amino acid residue from its acceptor site to a donor site. Thus the translocation of tRNA from one ribosomal site to the other is presumed to be inhibited by I. The apparent conflict may be due to an attempt to correlate the results of two distinctly different experimental tests:I6 (a) the puromycin reaction which is specific for detecting an inhibition of peptide bond formation and (b) the guanosine triphosphate (GTP) dependent G-factor catalyzed reaction for deter- mining release of tRNA. An inhibition of the latter reaction does not necessarily imply inhibition of the former reaction per se, but would cause peptidyl tRNA to remain in the acceptor site from which the peptide moiety could not be removed by puromycin. Mielck and Garrett" observed from their study of I action by microbial kinetics that I, in fact, possessed 2 modes of action which they attributed to (a) an impairment in the functioning of the tRNA by binding of the drug to the 50s ribosomal site and (b) a possible interference in the synthesis and utilization of a stored metabolite.

The similarity in chemical structure of I, 11, and I11 implies that the latter 2 analogs should have mechanisms of action similar to I but with differences in intrinsic bjolog ical activities, so that the combined action of I with such analogs should be quantifiable on a kinetically equivalent basis as has been shown for chloramphenicols" and sulfon- a m i d e ~ . ' ~ Of course, it is possible that the enhanced antibacterial activity' obtained by halosubstitution at the 74s) configuration of I in the order C1< Br < I may be associated with stereoselective binding at a different site. If this were the case, it is possible that the action of I and the 74s) halo analogs might show kinetic parameters of different functional dependencies from that of either drug alone. This communication presents the results of testing this hypothesis by microbial kinetics.

In contrast, other workersI5

Experimental Section Organism. Replicate slants of E. coli ATCC 12407 (referred to

as strain B/r in previous publications1s-*' were used in all experiments. The slants had been prepared from a single colony and were stored in a refrigerator at 4".

Detroit, Mich.) was rehydrated according to the specifications of the manufacturer to peptone broth USP. The media were filtered twice through Millipore 0.45 p HA filters and autoclaved at 120' for 15 min. The pH of the media was 7.05 k0.05 with the exception of those that were used to study the antibacterial activity as a function of pH. To obtain media with a pH in the range of 5.1-8.2, various amounts of Millipore-filtered 1.7 N HC1 and 2 N NaOH, respectively, were added to the culture media aseptically before the sterilization.

II.HCl(838 pg of base equiv/mg), and III.HC1 (1000 pg of base

Culture Media. Bacto Antibiotic Medium 3 (Difco Laboratories,

Antibiotic. Assayed samples of I.HCI (895 pg of base equiv/mg),

equiv/mg) were supplied by courtesy of Dr. G. B. Whitfield, Jr., of The Upjohn Co., Kalamazoo, Mich. The references to concns of drugs throughout this paper refer t o these samples of antibiotics.

Bacterial Cultures. An aliquot (5 ml) of culture medium was inoculated from a fresh slant, and the culture was allowed to grow for 12 hr at 37.5' in an incubator. A sample of 0.5 ml was then dild 100-fold into fresh medium. The generation of the culture was followed up to 2 X lo' E. colilml. An aliquot of this culture (Le., 1 m1/100 ml of broth) was added to a bulk amount of broth contd in a Pyrex flask fitted with a 49.5-ml Calab pourer. The "seeded" broth was kept in an incubator at 37.5' for 20 min with intermit- tent shaking. Aliquots (49.5 ml) of the "seeded" broth were then aseptically transferred from the Calab pourer into replicate loosely capped erlenmeyer flasks. The flasks were maintained at 37.5' * 0.1' in a 50-gal, constant temp water bath equipped with a shaker.

Total Count Method. This method has been previously described." Samples of 1.00 ml were withdrawn a t 20-min intervals from the cultures. They were dild to obtain counts within a range of 10,000-30,000 counts per 50 pl on the Coulter Counter, Model B (Coulter Electronics Co., Hialeah, Florida). The diluent used was a Millipore 0.45 p HA-filtered aq soln of 0.85% NaCl and 1% CH,O. The instrument was equipped with a 30-p orifice. The settings were: aperture current of 5, amplification of 8, gain of 10, lower threshold of 13 , and upper threshold at maximum. The total counts were corrected for the background count of the particular batch of media used, and dild in the same way as the sample. The background counts in general did not exceed 1000 counts per 50 pl. The coincidence of total (Coulter count) and viable (colony count) numbers of E. colilml in drug-free and in sub-completely-inhibitory, lincomycin-treated cultures had been previously demonstrated." There was no significant evidence of kill superimposed on normal inhibition of generation in the presence of subcompletely-inhibitory concns of the drug.

Effect of Antibiotic Concentration on Generation Rates Fresh solns of the respective antibiotics were aseptically prepd for each experiment. They were sufficiently dild so that aliquots of 0.5 ml added to 49.5-ml culture vols yielded the desired drug concns (Table I). The solns were added to the cultures generating at 37.5" in the log phase at an organism population of about 1.0 X lo6 E. colilml. Samples were withdrawn every 20-30 min and counted by

Table I. Apparent First-Order Generation Rate Constants, kilpp in sec-', and Other Derived Constants for Generation of E. colr In Various Concentrations of Lincomycin and Its Analogs at pH 7.05 at 37.5".

Clinda- Lincomycin.HC1 mycin.HC1 111. HCI

105k, pg/ml Phase I PRase I1 pg/ml 10'kapp pg/ml 105kapp

0 62.37 0 61.55 0 62.04 30 40 60 80

100 150 200 25 0 300 350

105k,a 105kab 10'kbc 10' ka/kbd

53.64 51.10 46.21 39.80 34.18 24.66 17.11 13.06 10.30 9.34 0.28 0.63 0.9 1

68.96

33.70 28.00 24.01 15.87 10.33 7.82 3.55 2.57

0.77 1.03

14.14

5 53.97 10 45.07 20 32.26 30 20.58 40 14.69 50 10.32 60 8.61 70 8.06 80 6.62

100 6.15 1.62 3.75 5.43

68.96

i.0 54.50 2.0 49.20 2.5 44.06 3.0 40.72 5.0 29.39 7.0 20.24

10.0 14.23 13.0 9.50 15.0 8.50

7.00 15.63 22.66 68.96

'Talcd as the slope of a plot of k, YS. concn, C in accordance with the equation: kapp = ko - k , z where Cis from 0 to 100 pg/ml of I.HC1 during phase I, 0 to 20 pg/ml of II.HC1, and 0 to 4 pg/ml of III.HC1, and where kc is in ml/pg sec. bReciprocal of the intercept of a plot of C/(k , - kap ) YS. C in accordance with the equation: C/(k , - k,,,) = l/ka +%b/k,(Q for C > 100 pg/ml of I.HC1 duringphase I and for 0 < C < 350 pg/ml of I during phase 11, or for C > 16.67 pg/ml of II.HC1 or for C > 25 @g/ml III.HCl; where k, is in ml/pg sec. %e quotient of slope and intercept of such plot within the limits of footnote b in ml/pg, or the product of drug partition constant, K,, and dru affinity constant, K,, in accordance with derivations for eq 3. Bkeciprocal of the slope of such a plot within the limits of footnote b in sec-I.

154 Journal of Medicinal Chemistry, 1972, Vol. I S , No. 2

L

Drug

Heman-Ackah and Gawett

a

d 100 200 300 400

Minutes

lo9

b I

/fDruq added

100 200 300 400 0 106

Minutes

Figure 1. Generation rate curves of E. coli at 37.5' and pH 7.05 in the absence and presence of graded (a) lincomycin.HC1 and (b) clin- damycineHC1 concns. The curves are labeled according to the drug concn in pg/ml.

the Coulter method. One culture without drug was studied in each experiment as control to obtain the generation rate constant (k,) in absence of drug. The generation curves for 0-300 Mg/ml of I-HCl at pH 7.05 were obtd (Figure la). Similar experiments were performed for 0-100 pg/ml of II.HCl and 0-13 pg/ml of 1II.HCl.

Effect of pH on Drug-Affected Generation Rates, Sufficient amts of 1 N HCl and 2 N NaOH were added to broth to obtain pH values 6.10-8.20; 6 replicate 49.5-m1 vols of each pH constant broth, inoculated with E. coli culture in the log phase of generation, were maintd at 37.5" until the organism population had reached 10' E. colilml. Drug solns (0.5 ml) were added to replicate cultures to achieve the desired concns of antibiotics. The 6th replicate in each

l a

' 40

- 1

lo/ I

50 80 IOOX Lincomycin 20 0 % Clindomycin 100 80 50

Composition of Equipotent Mixtures

60

50

I 80 i 0 O X Lincomycin 50

Composition of Equipotent Mixtures

100 80 50 20 0 % U24729A

l o i - 80 100 VO Clindamycin OO 20 50

100 80 50 20 0 % U24729A Composition of Equipotent Mixtures

Figure 2. Effect of varied (a) clindamycin . HCl and lincomycin ' HCl fractions, (b) III.HCl and I.HCl fractions, and (c) 1II.HCl and 11' HCl fractions in equipotent mixts at 3 different potency levels on the apparent generation rate constants, k, 7.05 and 37.5'. (a) II.HC1 is presumed to !e 6 times as potent as 1-HCl (phase I) on a wt/wt basis. (b) III.HCl is presumed to be 25 times as potent as I*HCl (phase I) on a wt/wt basis. (c) III.HCl is presumed to be 4.16 times as potent as II.HC1 on a wt/wt basis.

(sec-I) of E. coli at pH

set contd no drug. Coulter counts were obtd from samples withdrawn every 20-30 min.

of Lincomycin and Its Analogs. Replicate 49.5-m1 samples of cultures contg 106/ml E. coli in steady-state generation were treated

Action of Equipotent Mixtures Composed of Different Fractions

Lincomycin Antibiotics Journal of Medicinal Chemistry, 1972, Vol. I S , No. 2 155

- E \

0 .- - W

lo7

0 / p DO*. , J 0 100 200 300

Minutes 0

100 200 300 Minutes

Figure 3. Effects of order of addn of equipotent (a) clindamycin.HC1 and lincomycineHC1 and (b) of equipotent III.HC1 and 11-HCl on generation rates of E. coli (a) Curve A is for generation of culture in the absence of drug. Curve B is for generation of culture in the presence of 16.67 pg/ml of 11-HCI and curve C is generated when in the presence of equipotent 100 pg/ml of 1,HCl. Curve D is generated when equipotent 1. HCI (100 pg/ml) is added to the clindamycin-affected culture of curve B and curve E is generated when equipotent 11. HCl(l6.67 pg/ml) is added to the phase I lincomycin-affected culture of curve C. Curve F is generated for a mixt of equipotent concns of II.HCl(l6.67 pg/ml) and I.HCl(100 pglml), added to the culture of curve A. Curve G is generated when 33.34 pg/ml of 11-HCl is added to the culture of curve A and curve H is generated when equipotent 200 pg/ml of I - HCl is added to the culture of curve A. (b) Curve A is for generation of culture in the absence of drug. Curve B is for generation of culture in the presence of 4 pg/ml of II.HCl or the coincident curve C is for generation of culture when in the presence of equipotent 16.67 pg/ml of IIeHCl. Curve D is generated when equipotent II.HCl(16.67 pg/ml) is added to the 111-affected culture of curve B or the coincident curve E is generated when equipotent 111. HCI (4 pg/ml) is added to the clindamycinaffected culture of curve C. Curve F is generated for a mixt of equipotent concns of III.HCl(4 pg/ml) and II.HCI(l6.67 pg/ml) added to the culture of curve A. Curve G is generated when 8 pg/ml of IIIeHCl is added to the culture of curve A or the coincident curve H is generated when equipotent 33.34 pg/ml of 11-HC1 is added to the culture of curve A.

with aliquots (0.5 ml) of equipotent mixts of antibiotics. The mixts consisted of 100,90,80,60,50,40,20, 10, and 0% concn of 1 antibiotic at 1 dose level and the residual percentage of equipotent concn of the other antibiotic to maintain the anticipated equipotency at that dose level. The equipotent concns were considered as 100 pg/ml of I'HC1, 16.67 pg/ml of II*HCl, and 4 pg/ml of III.HC1, respectively. These studies were repeated at levels of activity corresponding to the equipotencies of 50, 100, and 200 pg/ml of I. Cultures were treated with equipotent mixts of I and I1 (Figure 2a), I and 111 (Figure 2b), I1 and I11 (Figure 2c). Coulter counts were obtained from samples of cultures withdrawn every 20-30 min and apparent generation rate constants (ka p) detd.

in Combination on Microbial Generation. Replicate 49.5-m1 samples of cultures contg 106/ml of E. coli in steady-state generation (curve A in Figure 3a) were treated with aliquots (0.5 ml) of equipotent solns of I and its analogs. Equipotency of action is shown by coincident or parallel generation curves of the drugaffected cultures having the same generation rate constant. The resultant generation curves for the action of equipotent concns of 16.67 pg/ml of II.HC1 and 100 pg/ml of I*HCl (phase I) are given as curves B and C, respectively. Replicate cultures of curve A were also treated with aliquots of a mixt of equal parts of the equipotent concns of I1 and I (curve F) which was prepd to be a priori as equipotent as the corresponding I1 (curve G) or I (curve H) alone.

Fifty min after the clindamycin-affected culture of curve B had settled to a new steady-state generation, an equipotent amount of I was further added. The resultant generation is given as curve D. A similar treatment of the lincomycin-affected culture of curve C with equipotent amount of I1 resulted in ultimate generation given as curve E.

The experiment was repeated in like manner for equipotent concns of 4 pg/ml of III.HC1 and 100 pg/ml of I.HC1 or for equi-

Effect of the &der of Addition of Lincomycin and Its Analogs

potent concns of 16.67 pg/ml of II.HCl and 4 pg/ml of III.HCl (Figure 3b).

drawn every 20-30 min.

in phase I and Phase I1 Generation. Replicate 49.5-ml samples of cultures contg 106/ml of E. coli in steady-state generation (curve A in Figure 4) were treated with aliquots (0.5 ml) of solns of I and its analogs and in combinations thereof. The separate effects for equipotent concns of I1 and I (phase I) on the generation of cultures are shown, respectively, as curves B and C in Figure 4. Curve N is for the effect of a mixt of equal parts of the equipotent concns of 11 and I which was prepared to be a priori as equipotent as the corresponding I1 (curve 0) or I (curve P) alone.

When the lincomycin-affected cultures of curve Chad settled to steady-state phase I generation, Le., 80 min after addn of I in Figure 4, aliquots (0.5 ml) of graded concns of I1 were added to each of 5 replicates so that the total antibiotic concn maintd in the cultures would be a priori 1.0, 2.0, 2.2, 3.0, and 3.5 times as equipotent as the concn of the I initially present in the culture (Table 11). The r e sultant generation curves are given as D, E, F, G , and H, respectively.

Again, when the cultures of curve Chad entered into steady- state generation phase 11, i.e., 175 min after the addn of I in Figure 5 5 other replicates were treated with similar concns of the clindamycin (curves I, J, K, L, and M). Coulter counts were obtd on samples of the cultures withdrawn every 20-30 min.

phase I1 generation were detd in a similar manner and sequence (Table 11).

of I1 and I11 were detd by titration in H,O with NaOH at an ionic strength of 0.122 Mand temp of 37.5" on a Radiometer automatic titrator Model 111 Lc-SBR 2c-SBUI.

Coulter counts were obtd from samples of the cultures with-

Effects of Clindamycin and I11 on Lincomycin-Affected Cultures

The effects of I11 on lincomycin-affected cultures in phase I and

Determination of the pKa' Value of 11 and 111. The pK,' values

156 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Heman-Ackah and Garrett

1

100 20 0 300 400 500 Minutes

Figure 4. Demonstration of antagonistic action of I1 on phase I generation of lincomycin-affected cultures. Curve A is for generation of culture in the absence of drug. Curve B is for generation of culture in the presence of 16.67 pg/ml of II*HCI and curve Cis for generation when in the presence of equipotent 100 pg/ml of I*HCl. Curves D, E, F, G , and H are for generations of phase I lincomycin-affected cultures of curve C, when concns of II.HC1: 16.67, 20.0, 25.0, 33.34, and 41.67pg/ml are, respectively, added after 80 min of initial addn of the I. Curve N is for a mixt of equipotent concns of II.HC1 (16.67 pg/ml) and I.HCl(lOOpg/mI) added to the culture of curve A. Curve 0 is generated when 33.34 pg/ml of II.HC1 is added to the culture of curve A and curve P is generated when equipotent 2OOpglml of I.HCI is added to the culture of curve A.

Reversibility of Action of I and Its Analogs. Aliquots (5 and 0.5 ml) of a drug-free culture in the log phase of generation and contg 10' E. coli/ml (curve A in Figure 6 ) were added to 45 and 49.5 ml of fresh broth, respectively; so that the organism population was dild 10-fold (curve A' in Figure 6) and 100-fold (curve A" in Figure 6).

A 49.5-m1 vol of broth contg 10' E. colilml in log phase generation (curve A in Figure 6) was treated with 0.5 ml of a soln of II.HC1 to achieve a final concn of 33.34 pg/ml (curve B in Figure 6). When the clindamycin-affected culture of curve B was in a steady- state generation and had reached an organism population of lo' E. coli/ml, aliquots (5 and 0.5 ml) were added to 45 and 49.5 ml of fresh broth, respectively, so that both the organism and drug concns were dild 10-fold (curve C in Figure 6) and 100-fold (curve D in Figure 6), respectively. At the same time, aliquots of the culture of curve B were dild 10-fold (curve E in Figure 6) and 100-fold (curve F in Figure 6), respectively, in broths contg enough I.HC1 so that drug concn was restored to 16.17 wg/ml.

The experiment was repeated in like manner and sequence for cultures affected with equipotent concns of 200 pg/ml of I.HCI and 8 pg/ml of III.HC1. Coulter counts were obtd on samples withdrawn every 15 min.

Results

Shape of Generation Curves for Drug-Affected Organisms. The plots of log of numbers of E. coZi/ml vs. time, obtained

from coulter counts, are given for graded concns of I (Fig ure la), I1 (Figure lb), and 111. The curves for this latter compound were similar to those for clindamycin. Linco- mycin exhibits the characteristic 2 phases of steady-state generation previously reported" while I1 and I11 have only 1 phase of steady-state generation. The apparent first-order generation rate constants (kap in sec-') are obtd from the slopes of the linear portions of the plot of log of numbers, N of E. coli per milliliter against time, t, in accordance with the equation

In N = kappt t In No (1 1 Effect of Antibiotic Concns on Generation Rates. A plot

of kapp us. concn (Table I) is given in Figure 7. The ka values are linearly dependent on drug concns up to lor pdml of I.HC1 (phase I growth), 16 pg/ml of I1 .HC1, and 4.0 pdml of III.HC1 respectively, in accordance with the expression

kapp = k o - k c C ( 2 ) where ko is the generation rate constant for the drug-free culture, kapp is the generation rate constant for culture af-

Lincomycin Antibiotics Journal of Medicinal ChemistTy, 1972, Vol. 15, No. 2 157

fected with drug concn, C, and k, (in ml/pg sec) is the inhibitory rate constant for the drug. Above these concn ranges, the curves slowly approach asymptotes. On the other hand, nonlinear decrease of the ka with drug concn is observed for I in phase I1 generation tiroughout the concn range studied. The relative potency of the drugs on a weight basis, estimated in accordance with eq 2 is approximately of the order 1:6:25 as I.HC1 (phase I)-II~HCl-III~HC1. Thus coincidence of the plots of kaqp vs. concn of the drugs (Figure 7) is obtained by multiplying the actual concns of 11-HCl and 111-HC1, in the data of Table I by factors of 6 and 25, respectively.

Applicability of Saturable Receptor Site Model to the Action of I, 11, and 111. Figure 8 gives a plot of C/(ko - kWp) vs. C in accordance with a prev io~s ly"~ '~ derived saturable receptor site model

C / ( b - kapp = c(kb/ka) + 1 /ka (3) where ka and k b are constants of proportionality related to drug availability in the biophase and drug affinity for receptor or binding sites. Similar consideration of the potency

Table 11. Concentrations of Clindamycin' HCI or III.HC1 Added to Phase I and Phase I1 Generations of Cultures Affected by 100 pg/ml of Lincomycin' HCI

Concn (pglml) of analogs added to

phase Ib and phase Reference IIC lincomycin- Final relative equipotent CUNeSa affected cultures concnd of total antibiotics

a b Clindamycin I11 added c c 1.0 D I 16.67 4.0 2.0 E J 20.00 4.8 2.2 F K 25.00 6.0 2.5 G L 33.34 8.0 3.0 H M 41.67 10.0 3.5

addition of I. CAt 175 min after the initial addn of the I. d l O O Mg/ ml I.HCI= 16.67 pg/ml of II.HCl= 4.0 pg/ml of III.HC1).

OAs indicated in Figures 4 and 5. bAt 80 min after the initial

estimates for I.HC1 (phase I), IIeHCl, and III.HC1 in the plots of Figure 8 in accordance with eq 3 makes the plots coincident. Adherence to the model is observed from the

I "

0 100 200 300 400 500 Minutes

Figure 5 . Demonstration of antagonistic action of clindamycin on phase I1 generation of lincomycin-affected cultures. Curve A is for generation of culture in the absence of drug. Curve B is for generation of culture in the presence of 16.67 pg/ml of II.HC1 and curve C for generation of culture in the presence of equipotent 100 pg/ml of IsHCl. Curves I, J, K, L, and M are for generations of phase I1 lincomycin-affected cultures of curve C when concns of II.HC1: 16.67, 20.0,25.0,33.34, and 41.67 pg/ml are, respectively, added after 175 min of initial addn of the lincomycin. Curve N is for a mixt of equipotent concns of II.HCI (16.67 pg/ml) and I.HCI (100pg/ml) added to the culture of curve A. Curve 0 is generated when 33.34 pg/ml of II.HC1 is added to the culture of curve A, and curve P is generated when equipotent 200 pg/ml of I.HC1 is added to the culture of curve A.

158 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Heman-Ackah and Garrett

t I

- E . 5 10' W

I 3 0 0 400

""Minutes I00

Figure 6 . Semilog plots of reversibility studies of E. coli generation with time on addition of II.HC1 and diln of the cultures with broth. Curve A is without drug. Curve B is after addn of II.HC1 to the culture of curve A to a final concn of 33.34 pg/ml. Curves A' and A" are after diln of the culture of curve A, 1 : l O and 1:lOO with broth, respectively. Curves C and D are after diln of the culture of curve B, 1:lO and 1:lOO with broth to final drug concns 3.34 and 0.34 pg/ml of II.HC1, respectively. Curves E and F are after diln of the culture of curve B, 1 : l O and 1:lOOO with broth contg sufficient amts of II.HC1, respectively, so that the final drug concn is restored to 33.34 pg/ml of II.HCl.

1

i

0 00 2 0 0 3 0 0 Concentration, C x f (hq I ml I

0 00 2 0 0 3 0 0 Concentration. C x f (hq I ml I

Figure 7. Demonstration of the coincidence of the dependencies of the apparent generation rate constants, kap for E. coli cultures at 37.5' and pH 7.05 on equipotent concns, {of I*HCl, II.HC1, and 111. HCl. Curve A represents this dependence, where the open triangles are for 1. HCl (phase I), the open circles are for 11. HC1, and the closed circles are for III.HC1, and wheref= 1.0 for I.HCl,f= 6.0 for II'HCl, and f = 25.0 for III*HCl. The dashed line, curve B, demonstrates the linear dependency of k,, drug concns. Curve C represents the depenllnce of the kapp on concns, C, of I .HCl (phase 11) at 0 < C < 350 pg/ml.

on drug concn at low

linear plots obtained for concns of drugs greater than 100 pg/ml of I.HC1 (phase I), 16.0 pg/ml of II.HC1, and 4 pdml of III.HC1, respectively. Deviations occur, in all cases, in the lower concn ranges. I (phase 11) adheres to eq 3 at all concns studied. The values of the constants ka and k b calcd from the slopes and intercepts of such plots are given in Table I.

Effects of pH on Drug-Affected Generation Rates. The apparent first-order generation rate constants (kapp) were obtained at different pH values (6.10-8.20) of the culture media in the absence and presence of graded concns of I , 11,

I too zoo 300 400

0--

Concentration. C x f l p g / m l )

Figure 8. Application of saturation kinetics to the actionof 1. HCl, II.HCl, and III.HC1 at higher concns, C, on the apparent generation rate constants, kapp (sec-I) of E. coli at 37.5". Curve A represents this application, where the open circles are for concns of 1. HCl (phase I) > 100 pg/ml, the closed triangles are for concns of II.HC1 > 16.67 pg/ml, and the open triangles are for concns of 111 .HC1> 4 Mg/ml, and where f = 1.0 for I .HCl,f= 6 for I I . K l , a n d f = 25.0 for 111. HCl. Curve B represents the application +

(phase 11) at 0 < C < 350 Mg/ml. The curves are plotted from the data of Table I according to eq 3.

'mens of I ' HCl

/ /"

t I I

4 -

8.0 7,

6.0 7.0 to-

5.0 PH

Figure 9. Dependence of the log of the apparent inhibitory rate constant, ka, for the effect of I.HC1 [phase I ] (curve A), II.HC1 (curve B), and III.HC1 (curve C) on the growth of E. coli at 37.5', on the pH of the broth medium. The drawn lines are consistent with the expression: k, = k,*K,'/(K,' + [H+]), where k,* is the intrinsic inhibitory rate constant of the unprotonated drug and K,' is the dissociation constant for the protonated drug concn, C. The k, is obtd from the expression: kapp = k, - k,C/(1 + k&).

and 111. The values for I were obtained by Mielck and Gar- rett" at similar pH values using the same organism and experimental technique. The drug-free generation rate constants were invariant with pH within the range studied. The drugaffected generation rate constants at a specified concn of each drug, decrease significantly with increased pH values. Thus, larger amounts of a drug are required for the same

Lincomycin Antibiotics Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 159

fractional inhibition of microbial growth as the pH of the medium is decreased. The constants, ka, derived from plots of c/(ko - kaP) vs. C in accordance with eq 3 at different pH values are plotted as log ka against pH in Figure 9. The values of ka increase 10-fold for unit increase in the pH over the range 5.80-6.50 for I (phase I), 6.10-6.70 for 11, and 7.00-7.60 for 111, respectively. The slopes of the plots of log ka vs. pH however, tend to lessen, in all cases, as the pH approaches the respective pKa' value of the drug.

Effect of Equipotent Mixtures Composed of Different Fractions of I and Its Analogs on Generation Rates. The generation rate constants, (kapp) of cultures affected by equipotent mixtures of I and I1 are shown in Figure 2a, I and I11 in Figure 2b, and I1 and I11 in Figure 2c, respectively. The mixtures consisted of 100 to 0% of 1 antibiotic and the residual percentage of equipotent amount of the other antibiotic. The mixtures were prepared so as to be a priori equipotent in their combined action on E. coli generation in accordance with Figure 7 at 3 different levels of action. The null slopes of the plots of the kapp for all the a priori equipotent mixtures of I1 and I11 (Figure 2c) demonstrate the in- difference22 of effects. I and I1 mixtures or I and I11 mixtures, on the other hand, exhibit lessened activity as indicated by higher kapp values than the equipotent antibiotic used alone (Figures 2a and 2b). This is most readily manifested at the higher potency levels. The effect is less striking at the lowest concn studied. This unequivocally demonstrates antagonism22 of effects between I and its analogs in their combined action against E. coli.

Effect of the Order of Addition of I and Its Analogs on Microbial Generation. There are no significant differences among the generation inhibition produced by the action of 100 pg/ml of IaHCl in phase I (curve C) and the equipotent concn 16.67 pglml of II.HC1 (curve B) in Figure 3a since the initial portion of curve C has the same slope as curve B. There are however, significant differences in the effective inhibition of generation produced by the action of a mixture of 100 pglml of I.HC1 and 16.67 pg/ml of II.HC1 (curve F) and that of the a priori equipotent concn of either drug alone, i.e., 200 pg/ml of I.HC1 in phase I (curve H) or 33.34 pdml of II.HC1 (curve G). The slope of curve F is higher than that of curve G or the initial portion of curve H. This demonstrates antagonism of effects between I (phase I) and 11. Also, an equipotent amount of 11, added after 50 min to a lincomycin-affected culture of curve C, produces an ultimate steady-state generation rate (curve E) which is the same as that of the culture affected initially with I. How- ever, the addition of I after 50 min to a clindamycin-affected culture of curve B, produces an ultimate steady-state generation rate (curve D) which is the same as that of the culture affected initially with the I1 (curve B) but shows nothing of the I (phase 11) effect.

This pattern of response is likewise observed for combinations of 100 p d m l of I.HC1 with an equipotent concn Of 4 pdml of III.HC1. In both cases the order of addition of the antibiotics produces significant effects on the ultimate generation inhibition and shows antagonism of the action between I and I1 (or 111).

There are however no significant differences in the effective inhibition of generation produced by the action of a mixture of 16.67 pdml of 11-HCl and 4 pg/ml of III.HC1 (curve F) in Figure 3b and that of the a priori equipotent concn of either drug alone, Le., 33.34 pg/ml of II.HC1 (curve H) or 8 pg/ml of III.HC1 (curve G), where curves F, G, and H have the same slopes. Also, there are no significant differences in the effective inhibition of generation, produced

by an equipotent amount of the I1 added after 50 min to 111-affected culture (curve D) or by an equipotent amount of I11 added after 50 min to 11-affected culture (curve E). Curves D and E have the same slopes as curves F, G , and H. Thus, the order of addition of these 2 antibiotics produces no significant change on the ultimate generation inhibition.

Antagonistic Action of I1 or 111 on Phases I and I1 Genera- tion of Lincomycin-Affected Cultures. The equipotency of action for I1 (curve B) and I (curve C) is indicated in Figure 4 by coincident or parallel growth curves of cultures affected with the drugs. A mixture of equipotent concns of I1 and I (curve N) is less active than the a priori equipotent concn of either the I1 (curve 0) or I (curve P) alone since curve N has a higher slope than curve 0 or the initial part of curve P. Curve N possesses some elements of the biphasic characteristics observed for I action alone.

The addition of graded concns of I1 to I-affected cultures in phase I steady-state generation (curve C) results in graded responses as shown by decreasing slopes of growth curves with increasing concns of I1 (curves D, E, F, G, and H). The ultimate generation rates represented by the final slopes of curves D through H are less than would be expected a priori from the additivity of equipotent concns of either drug alone and shows antagonism. For example, the data of Table I1 show that the total drug concn in the culture of curve G is about 1.5 times as equipotent as the II in the culture of curve 0 or the I in the culture of curve P. However, curve G has about the same slope as curve 0 or the initial portion of curve P. This indicates that the combination of I and I1 in the culture of curve H has about 1/1.5 or 66.67% of its a priori potency.

On the other hand, the ultimate steady-state generation rates of lincomycin-affected cultures in phase I1 are not sig nificantly changed by addition of graded concns of I1 (curves I, J, K, L, and M in Figure 5).

Mixtures of I11 with I produce patterns that are similar to the effects of I1 on lincomycin-affected cultures. In both cases the antagonism of effects produced by combinations of I1 or I11 with the I on the generation of cultures appear to be greater in phase I1 than in phase I of the lincomycin action.

PKa' of Lincomycin and Its Analogs. The determined pKa' values at 37.5" and ionic strength of 0.122 are 7.00 for I1 and 8.05 for I11 (Table 111). The value of 7.76 has been reported for I."

Table 111. Values of pK,' and Intrinsic Inhibitory Rate Constants (ka*) for Lincomycin and Its Analogs

Cited Kineticallyb ExperimentallyC literature

105ka*a determined determined values of Drug (mlhgsec) PKa PKa ' PKa '

Lincomycin 1.43 7.57d 7.768 7.6e Clindamycin 7.14 6.94 7.00 1.6f 111 125.00 7.79 8.05 8.0f

'As detd from the slopes and intercepts of plots of l/ka vs. [H'] in accordance with the equation: l/ka = l/ka*Ka'[H+] + l/ka*, where k , is obtd from the equation fitting the apparent generation rate constant, kapp, of the drug-affected cultures at the hi er concns, C of drug, i.e., C/(k , - k ) = C(kb/k,) + l/kw g a l c d from the quotient of the slope Zfyntercept of the plot according to the equation in footnote a. CAs detd by titration. dBased on k, values at pH 5.80 and 6.10, respectively, from data of Mielck and Garrett." eReported by Herr and Bergy.2 fPersonal communi. cation from Dr. G. B. Whitfield, Jr., The Upjohn Co., Kalamazoo, Mch. &See ref 17.

160 Journal of Medicinal Chemistry, 1972, Yol. 15, No. 2 Heman-Ackah and Garrett

Reversibility of Drug Action. The equilibration of I and its analogs between nutrient medium and biophase is readily achieved. It took 15-20 min for cultures affected with equipotent concns of 33.34 pg/ml of II.HC1 (curve B in Figure 6), 200 pg/ml of I.HC1, a n d 8 pg/ml of III.HC1 to attain a new steady-state phase of generation. These latter curves were practically coincident with those of I1 so they are not presented here. The steady-state generation of the drug affected cultures reverted to new steady-state generation with predictable rate constants when diluted 10-fold and 100-fold, respectively (curves C and D, respectively in F ig ure 6 ) . However, the new steady state was attained after an interval which was dependent on the dilution factor. For instance, it took 30-40 min for the 1 : 10 dilution, but 60-80 min for the 1 : 100 diluted 11-affected cultures to attain the new steady states. Similar lag periods were observed for corresponding dilutions of I-affected and 111-affected cultures. However, the correspondingly diluted drugfree cultures showed a very short lag period, e.g., 5-10 min to revert to their new steady states.

If the time required for rejuvenation of the cells or con- sequence of shock on dilution is considered, the prolonged lag period observed for the drugaffected cultures to attain a new steady-state generation upon dilution suggests that either the rate of dissociation of drug-receptor complex in the biophase is less than the rate of formation of the complex and/or the rate of diffusion of drug from biophase to the broth medium into the biophase through the cell m e m branes.

Cultures inhibited by low concns of the drugs, e.g., 1.67 and 0.167 pg/ml of II.HC1 (curves E and F in Figure 6), 20 and 2 pg/ml of I.HC1,0.8 and 0.08 pg/ml of III.HC1 were further inhibited by addition of more drug to final concns of 16.67 pg/ml of II.HC1,200 pg/ml of I.HC1, and 8 pg/ml of I11 .HC1, respectively, at the same relatively short equilibration time of 15-20 min.

Thus no significant difference is observed between the reversible action of I, 11, and I11 against E. coli.

Discussion

The data reported in this paper show that the generation rate constants (kapp) of E. coli cultures affected by I (phase I), 11, and I11 have similar functional dependencies on drug concns (Figure 7). This suggests that the analogs have a mechanism of action similar to I (phase I). On this presumption, the ratio of the biological potencies of these drugs are 1:6:25 as I~HCl-II~HC1-III~HC1 on a weight basis at pH 7.05. The corresponding molar ratio of potencies would be 1 :6.6:28.5. These apparent differences in potency can be attributed to differences in degrees of partitioning across the bacterial cell membrane and/or to differences in the affinities of the drug for the receptor site, and/or to differences in the intrinsic efficacies of the drug bound to the site. The steady-state generation of lincomycin-affected culture changes after several generations of growth to a new steady state, Le., phase I1 (Figure la), which does not occur with cultures affected by the other two analogs (Figure Ib). The functional dependency of the k,, on concn for I (phase 11) suggests a different mechanism of I action (Figure 7) which dominates the latter stages of lincomycin-bacterial interaction.

pH Effect on Activity of Lincosaminide Antibiotics. The extent of generation inhibition by lincomycin (phase I), 11, or I11 increases as some function of pH values of the broth medium. The constants k , (ml/pg sec) as defined in eq 3

for various pH values (Figure 9) adhered to the expre~s ion '~ -7 I Ka

K a t + [H'] k , = ka*f = k,*

where, k,* is the intrinsic inhibitory rate constant of the unprotonated drug, f is the fraction of the drug concn unprotonated, Kat is the dissociation constant of the protonated base, and [H'] is the H+ concn. The plot of log k, vs. pH approaches a slope of unity when [H+] > K,' and a slope of 0 when Kat > [H'] (Figure 9). Arithmetical transformation of eq 4 yields

1 - W + I 1 + _1- k , k,*K, ka*

Values of ka* and PKa' of the drugs derived from the slopes and intercepts of linear regressions obtained by plotting l/ka vs. [H'], are given in Table 111. A reasonably good agreement is obtained between the experimentally determined PKa' values by potentiometric titrations for the pro- line rings of the drug species and the kinetically determined values from eq 5 . It is therefore concluded that eq 4 holds and that it is the unprotonated fraction of the drug concn that is responsible for the antibacterial activity of I and its analogs. The relative intrinsic activity of the uncharged drugs on a weight basis is estimated to be approximately of the order of 1 :5.0:87.4 as I~HCl-II~HC1-III~HCl. The ratios in terms of molar concns are for the uncharged drugs 1:5.2:94.4.

Equipotent mixtures of I1 and I11 are equivalent in their action against E. coli (Figure 2c) and their effects can be quantified on the basis of additivity from the separate dose- response curves of either drug alone (Figure 7). This suggests a similar mode and locus of action for the 2 drugs.*2 Furthermore, the order of addition of the drugs in a combination has no significant effect on the ultimate steady- state generation of drug-affected cultures (Figure 3b), which therefore demonstrates the lack of any significant bacterio- static antagonism or synergism (by the definitions of Gar- rett22) in the sub-completely-inhibitory range similar to t f a t observed21 for combinations of I (phase I) and erythto- mycin.

In contrast, equipotent mixtures of I (phase I) and I1 (Figure 2a) or I (phase I) and I11 (Figure 2b) demonstrate unequivocal antagonism, since the mixtures are less effective on the generation of E. coli cultures than the a priori equipotent concentration of either drug alone. In addition, the extent of generation inhibition produced by the combined action of I (phase I) with I1 (Figure 3a) or I (phase I) with I11 depends on the order of addition of the drugs to a culture. For instance, the ultimate generation inhibition produced by admixing an equipotent amount of I (phasc 1) with an equipotent amount of I1 or of I (phase I) with an equipotent amount of 111, before addition of the mixture to a culture is greater than that obtained by adding an equipotent amount of one drug after an interval of time to a culture pretreated with an equipotent amount of the other drug. In the latter case, the I (phase 11) action is eliminated. when I is added to a culture pretreated with I1 or 111.

This antagonism of effects obtained by various combinations of I with I1 (or 111) suggests, as one possibility, that: I (phase I) may have a different binding site from that of I! (or 111) and hence their mechanisms of actionz2 are not t b

same Therefore, the similarity in functional dependenc;c> of the kapp for cultures affected with I (phase I), 11, CT I11

Anomalous Antagonisms among Lincosamide Antibiotics.

Lincomycin Antibiotics Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 161

on drug concns could be attributed to the fact that their binding sites, though different, nevertheless, are functionally linked. Such would be the case, if the separate sites were engaged in “sequential” or “convergent” metabolic proc- e ~ s e s ~ ~ which lead to a common end product. However, on this presumption, the combined action of I1 (or 111) with I (phase I) would either cause a “sequential blocking”24 of the metabolic pathway to produce “synergism” or cause a “concurrent blocking”2s to produce “additivity” (or equivalence) of effects. Hence, neither of these 2 mechanisms can explain the observed antagonism between I (phase I) and I1 or 111.

Certainly, these phenomena would be more readily under- stood if the one drug inhibited the transport of the other to the receptor site where both acted. However, the fact that the separate action of each of these several antibiotics is relatively rapidly reversible on dilution denies any significant inaDility of either drug to reach or leave the biophase. This fact makes it difficult to accept the conjecture that the presence of a molecule of one drug may sterically block or selectively inhibit the entrance of a molecule of the other to the biophase and, in effect, change its apparent partition coefficient K 1 or apparent activity.

Action. One possibility for rationalizing these antagonisms is that the presence of one drug quasipermanently modifies the allosteric configuration of the receptor site so that it diminishes the effective affinity constant of the other drug for that receptor site without significantly affecting the receptor site-substrate interaction necessary for microbial growth and generation.26-28 This can be expressed sche- maticallyZ9 by

A Possible Kinetic Model to Rationalize Antagonistic

Ikm Kg’ [ 1 + (kB/ka)(A’)] metabolites K A

for the competition of 2 drugs, A and B for the same receptor site, R, where the steady state metabolic rate, dM/dr, that results in generation is proportional to the fraction of these unreacted receptor sites as per

dM/dr = k m ( l - 6A - 6,) (7) where 6A and 6 B are the fractions of total receptor sites, RT, occupied by drugs A and B, respectively. The (A’) and (B’) values are the drug concns in the biophase in equilibria with the respective extracellular (A) and (B) concns and the respective concns of bound receptors, i.e., (A’R) and (B’R). The dissociation of one receptor-drug complex; e.g., (B’R) is additionally catalyzed by the available concentration, (A’), of the other drug in the biophase. It should follow that the number of occupied receptor sites in the presence of n equipotent units of a mixture of A and B will be less than the number in the presence of n equipotent units of A or B alone. Thus, the metabolic rate which is porportional to a fraction of free receptor sites, eq 7, should be higher from the combination than would be predicted on the basis of drug equivalency. Thus, such allosteric modifications should result in antagonism for such a combination.

The equilibrium constant for the interaction of drug B with the free receptor site, R, can be implicitly defined in

B’.R/(B’R) = B’[RT - (B’R) - (A‘R)] /(B’R)

= (kB -t kB ’(A’))/k-B

= 1 -t (~B’/~B)(A’) /KB’ (8) and similarly for drug A,

(A’)*R/(A’R) = (A’)[RT - (B’R) - (A‘R)]/(A‘R)

= kA/k-A = 1 /KA ’ (9) Equation 8 can be rearranged to explicitly define the re-

ciprocal of the fraction of receptor sites occupied by the drug B as

and eq 9 can be similarly rearranged to define the reciprocal of the fraction of receptor sites occupied by the drug A as

1/6A = RT/(A’R) = 1/KA’(A’) t 1 t @%)/(A%) (1 1) Equivalent values may be obtained by solving eq 10 and

11 for R. When these two rearranged expressions are equated, the ratio (A’R)/(B’R) can be obtained and is

(A’R)/(B’R) = [ 1 t (~B’/~B)(A’)]KA) (A’)/KB’(B’)

1 1, respectively, and since

(12) This value, or its reciprocal, can be inserted into eq 10 and

(A’) = KA (A) (13)

(B’) = KB (B) (14) where (A) and (B) are the respective concns of the drugs in the media,

l / e B = [ 1 t K A K ~ (A) t ( ~ B ’ / ~ B ) K A ( A ) t

(kB’/kB)KA) (KA(A))’ -t KBKB‘(B)I /KBKB’ (B) (1 5 ) and

I /eA = [I +KAKA’@) (kB’/kB)KA(A) t

( k ~ ’ PB KA’ (KA(A))~ + KBKB ’(B)I / [ 1 -t

(kB’ /kB )KA (A) 1 KAKA’ (A) (16)

Thus, the fraction of receptor sites, fR, not reacted with either drug A or B is

fR = 1 - 6A - 6B

= [I (kB’/kB)KA(A)]/[l +KAKk(A) +KBKB’(B)+

&B ’ / k ~ )KA (A)( 1 -t KAKA ’(A))1 (17) which equation, if the dissociation of the B’R complex is not catalyzed by the drug A, Le., k ~ ’ = 0, is transformed to

fR = 1 - 6A - 6B = 1/(1 t KAKA’(A) t KBKB’(B)) (18) Thus the ratio of the fraction of free receptor sites, f ~ ,

for the case when drug A catalyzes the dissociation of receptor site com lex of drug B, to the fraction of free receptor sites, fR , when no such catalysis exists and A and B compete for such sites, i.e., their inhibitory action differs only by a potency factor, is

P

162 Joumal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Heman-Ackah and Garrett

Thus, such a model as eq 6 to represent the modification of an allosteric configuration of a receptor site for one drug by the presence of another results in an antagonism of antibacterial action; since less receptor sites are occupied by drug than would be anticipated on the basis of simple equiv. alency of action of each drug of the combina t io r~ .~~ The developments of eq 6-19 are based only on drug A catalyzing the dissociation of the (B’R) complex. Similar expressions can be developed for the simultaneous catalysis of the dissociation of the (A’R) complex by the drug B.

Since the generation rate is proportional to the metabolic rate leading to generation and the number, N, of generating organisms, then from eq 7

dN/dt = q(dM/dr)N = qkm( 1 - 8 A - 8B)N = kappN (20)

and the apparent generation rate constant, kapp, will always be greater for such allosterically modifying components of combinations (Figures 2a and 2b) than can be predicted a priori on the premise of drug equivalence of the components (Figure 2c and eq 19).

Steric configuration may play an important role in this type of drug-receptor reaction, since it has been observed7 that replacement of the 7(R)-OH group of I by a halogen substituent in the 7(R) configuration had little or no effect on in vitro antibacterial activity but a halogen substitution in the 7(S) configuration increased the activity 4- to 32-fold in the order C1< Br < I. These facts may indicate that the 7(R) and 7(S) epimers of I bind differently on the same receptor site (or enzyme). One feasible proposition could be that 2 possible binding sites exist for one lincosaminide molecule, a “catalytically active” site and an “allosterically modifying site.’rZ6-28

The high activity of I1 (or 111) relative to I (phase I) sug gests the possibility that I (phase I) may exert an allosteric effect which modifies the affinity of itself and the others for the catalytically active site. The binding of I (phase I) at the “allosteric site” possibly may produce an extensive conformational change that causes an “unfavorable” configuration of the “active” site of the receptor and thereby decreases the net reactivity of the “catalytically active” site for all of these drugs. Of course, the “unfavorable” configuration for I may be induced by the other drugs; or both may induce such configurations for the other, The fact that I does have a second mode of action is of interest with regard to this hypothesis in that it may be speculated that the phase I1 of lincomycin action may be related to its possible ability to allosterically modify the receptor site for the other lincosaminide antibiotics.

Action of Lincomycin. I (phase 11) action may be rationalized as a different mechanism from that of I (phase I) and I1 (or 111) action. It implies an action which is interdepen- dent with the initial I (phase I) action, but which becomes effective only after a finite time of drug-bacteria interaction (Figure la). The data of Table I show that the extent of generation inhibition (ko - kapp) is greater for phase I1 than for the phase I action of I at each level of concn over the range 60-350 pg/ml of I.HC1.

An explanation that has been proposedi7 for the lincomycin (phase 11) action was that I may block an additional receptor site engaged in the synthesis of an essential metabolite. The metabolite or its successors is possibly utilized by the enzyme (or receptor site) that is also the binding site for I (phase I) action, in a “sequential b l~ck ing” ’~ of protein synthesis t o enhance generation inh ib i t i~n . ’~ The depletion

A Possible Kinetic Model to Rationalize the Biphasic

of this necessary metabolite or its precursors during successive microbial generations imposes a new rate-limiting factor and a new and lessened steady-state generation (phase 11) is established. This model may be considered as similar to the synergistic action of sulfonamide-trimethoprin combinations on the tetrahydrofolate utilization sequencez9 where the prior sulfonamide action on folic acid synthesis has a 5.8 generation lag for its effect to be observed after drug addition, whereas the subsequent dihydrofolate reduc tase inhibition of trimethoprin has a relatively immediate effect on microbial generation. Consider the metabolic sequence

where the substrate, S , is acted upon by the receptor site, R1, to produce the metabolite M1 necessary for the receptor site, Rz, to transpose it to the metabolite, Mz, which is vital for microbial growth and generation. If I competes with M 1 for the receptor site Rz in phase I action and competes with S for R1 in phase I1 action, then

dMz/dt = kZ(R2M1) = k z K ~ , RzMl= k lkZK~, Rz(SR1) =

= kik2KRIKR2 SRiRz

=kikzKEIKE,S(Ri)T(R2)T(1 - ei>(l - 0 2 )

= k,(i - e,)(r - e2) (22)

(23)

where

km = k I ~ ~ K R , KR2 S(EI) T (E,) T since it may be postulated that the amount of substrate, S , and the total amounts of receptor sites ( R l h and ( R ~ ) T are constant. The 8 and tlz are the respective fractions of these sites that are treated with I, respectively, and thus unavail- able for reaction in the metabolic sequence of eq 21 leading to the Mz vital for microbial growth and generation.

These fractions of receptor sites treated with I may be defined” as

8 1 = K ~ K R , L / ( ~ -t K ~ K R , L ) (24)

where kapp, is the apparent generation rate constant in I (phase I) action definedz9 by

kapp, = qk, (1 - 6,) = ko - kaL( 1 -t kbL)

= ko - k&iKRzL/(l K&R2L)

= ko/(l -t KiKRZL) (26) where L is the I concn and ko = qkm = ka/kb.

W / d t = q W / d t = qk,(l - el)(l - 8&V= kapp,,N

and when ko = qk, and eq 26 are considered, the new generation rate constant, becomes

kapp,=ko(l - 01 - 4 + B1&)=kappI - koBi(1 - 62)

When phase I1 action comes into play, then, as in eq 18,

(27)

= ko/(1 -t KiKRIL)(l KiKR2L)

= kapp1/(1 K,KRZL) (28) and thus the apparent generation rate constant kapp for phase I1 lincomycin action will always be less than the apparent generation rate constant, k,,, for phase I lincomycin action.

Lincomycin Antibiotics Journal of Medicinal Chemistry, 1972, Vol. IS, No, 2

inhibition of “peptide bond f ~ r m a t i o n . ’ ~ ’ ~ Whatever the mechanisms of action, I does antagonize the action of its 7 4 s ) C1 analogs against E. coli.

Acknowledgments. The authors are greatly indebted to Mr. George L. Perry, Sr., for excellent technical assistance. This investigation was supported in part by Public Health Service Research Grant ROI AI 10058-01, National Institutes of Health.

163

____

100 200 300 Concentration ( / a ml-’)

Figure 10. Linearity of the plot of the ratio kappI/kapp I vs. concn for lincomycin-affected E. coli cultures a t 37.50 in accordance with the equation, kappI/kappII = k,L + b, where kapp1 is the generation rate constant for lincomycm-affected cultures in the phase I steady- state generation, kappII.is the generation rate constant for lincomycin-affected cultures m phase I1 steady-stage generation and where kc may be the product of the drug partition constant K, and the drug affinity constant K, for the lincomycin at the site of its phase I1 action. The parameter values are ki = 2.03 X ml/bg and b = 1.25.

The ratio of the apparent rate constants for phase I and phase I1 lincomycin action can be derived from eq 28

kappIlkapp, = 1 + (KIKR, >L (29) and is shown to be reasonably valid by the plots of these ratios against I concn in Figure 10 which clearly demonstrates a slope of K I K R . The intercept, however, is slightly greater than unity which may be attributed to the fact that k d k b of eq 27 and 28 may not be exactly equal to, although they are proportional to, k,,, the generation rate constant in the absence of

The precise physiological significance of these mechanisms cannot be ascertained from the available kinetic data. How- ever, it can be conjectured that the common action of I (phase I) and I1 (or 111) is related to the inhibition of the “peptide bond formation” reported by other worker^'^"^ using cell-free extracts. In this case, I (phase 11) action may be associated with the inhibition of a synthesis of some essential amino acid or derived peptide precursors not nor- mally available in the broth medium. Alternatively, the common action of I (phase I) and I1 (or 111) could be related to the inhibition of the “translocation” of peptidyl-tRNA in polypeptide synthesis and that of I (phase 11) to the

References (1) J. D. Mason, A. Dietz, and C. deBoer, Antimicrob. Ag. Chemo-

(2) R. R. Herr and M. E. Bergy, ibid., 560 (1962). (3) L. J. Hanka, D. M. Mason, R. R. Burch, and R. W. Treick,

(4) C. N. Lewis, H. W. Clapp, and J. E. Grady, ibid., 570 (1962). (5) M. Barber and P. M. Watenvorth, Brit. Med. J., 2,603 (1964). (6) J. B. R. Duncan, Antimicrob. Ag. Chemother., 723 (1967). (7) B. J. Magerlain, R. D. Birkenmeyer, and E’. Kagan, ibid., 727

(8) B. J. Magerlain, R. D. Birkenmeyer, and F. Kagan, J. Med.

(9) W. E. Herrell, “Lincomycin,” Modem Scientific Publications,

ther., 554 (1962).

ibid., 563 (1962).

(1966).

Chem., 10,355 (1967).

Inc., Chicago, Ill., 1969, p 36. (10) B. Weisblum and J. Davies, Bacteriol. Rev., 32$493 (1968). (11) E. Cundliffe and K. McQuillen, J. Mol. BioL, 30, 137 (1967). (12) F. N. Chang, C. J. Sih, and B. Weisblum, Proc. Nut. Acud. Sei.

(13) D. Vazquez and R. E. Monro, Biochem. Biophys. Acta, 142,

(14) J. M. Wilhem, N. L. Oleinick, and J. W. Corcoran, Antimicrob.

(15) R. E. Monro and D. Vazquez, J. Mol. Biol., 28,161 (1967). (16) K. Igarashi, H. Ishitsuka, and A. Kaji, Biochem. Biophys. Res.

(17) J. B. Mielck and E. R. Garrett, Chemotherapy, 14, 337 (1969). (18) E. R. Garrett, 0. K. Wright, G. H. Miller, and K. L. Smith, J.

(19) E. R . Garrett and 0. K. Wright, J. Pharm. Sci., 56, 1576

(20) E. R. Garrett and G. H. Miller, ibid., 54,427 (1965). (21) E. R. Garrett, S. M. Heman-Ackah, and G. L. Perry, ibid., 59,

(22) E. R. Garrett, Antibiot. Chemother., 8,8 (1958). (23) B. W. Lacey, Symp. Soc. Gen. Microbiol., 8,247 (1958). (24) V. R. Potter, Proc. Soc. Exp. Biol. Med., 76,41 (1951). (25) G. B. Elion, S . Singer, and G. H. Hitchings, J. Biol. Chem.,

(26) C. Frieden, J. Biol. Chem., 239,3522 (1964). (27) H. E. Umbarger, Science, 145,674 (1964). (28) J. Monod, J. P. Changeux, and F. Jacob, J. Mol. BioL, 6,306

(29) E. R. Garrett, Fortschr. Arzneimittelforsch., 15, 271 (1971).

U. S., 55,431 (1966).

155 (1967).

Ag. Chemother., 236 (1967).

Commun., 37,499 (1969).

Med. Chem., 9,203 (1966).

(1967).

1448 (1970).

208,477 (1954).

(1963).


Synthesis and biological activity of pyrazines andpyrazine ribonucleosides as pyrimidine analogs

M. Bobek, and A. BlochJ. Med. Chem., 1972, 15 (2), 164-168• DOI: 10.1021/jm00272a010 • Publication Date (Web): 01 May 2002






164 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Bobek and Bloch

Synthesis and Biological Activity of Pyrazines and Pyrazine Ribonucleosides as Pyrimidine Analogs

M. Bobek* and A. Bloch Department of Experimental Therapeutics, Roswell Park Memorial Institute, New York State Department o f Health, Buffalo, New York 14203. Received July 21, I 9 71

Because of the marked antitumor activity which derives from structural modification of the heterocycle of the natural pyrimidine nucleosides (e.g., 3-deazauridine, Sazacytidine, 6-azauridine), various pyrazine analogs of pyrimidines were prepared. Among these, the uracil analog 1,2-dihydr0-2-oxopyrazine 4-oxide, a compound which has been isolated as the antibiotic emimycin, was synthesized by an improved method, involving the oxidation of benzoyloxypyrazine. The ribonucleosides of this analog and of 1,2-dihydro- 2-oxopyrazine were prepared by ribosidation of the corresponding trimethylsilyl derivatives in the presence of TiCl4. Also synthesized were a number of derivatives of emimycin, among them the 6-carboxy-, 2-thio-, and the 1-, 5-, and 6-Me analogs. Emimycin and its ribonucleoside were equally active, each inhibiting 50% of the growth of Streptococcus faecium and Escherichia coli at 8 X 10-6M and 1 X lO-'M, respectively. The orotate analog, 1,2-dihydro-2-oxo-6-carboxypyrazine 4-oxide, interfered with the growth of these organisms at 6 X lO-'M and 4 X M, respectively. The inhibitory activity of the remaining analogs ranged from 1 o+ M to 4 X 1 o-' M. The corresponding compounds lacking the N-oxide showed no inhibitory activity in these systems. At M, none of the compounds inhibited the in vitro growth of leukemia L 12 10 or Ehrlich ascites cells in excess of 10-20%. Strains of S. faecium resistant to emimycin or its ribonucleoside were cross-resistant to each other, but not to the orotate analog. Simi- larly, a strain resistant to the latter compound retained its sensitivity to the uracil and uridine analogs. The inhibitory effects of emimycin and its ribonucleoside were reversed competitively by uracil, cytosine, and their nucleosides over the concentration range of 1 Od3 to 1 0-6 M . Over this range, the growth inhibition exerted by the orotate analog was prevented noncompetitively by these metabolites, but competitively by orotate. Because of the selectivity of the pyrazine derivatives for bacterial cells, their possible use in antibacterial chemotherapy appears worthy of exploration.

Modifications in the heterocycle of the natural nucleosides have resulted in a number of analogs, which have demonstrated marked activity against a variety of experimental tumors. Included among these are compounds such as 3-deazauridine and 3-deazacytidine,' S-a~acytidine?,~ 6- azathymidine ,4¶' and 6-azauridine !T8 An examination of possible further alterations within the pyrimidine ring led us to consider the pyrazine ring, substituted in such a way as to provide analogs of the natural pyrimidines. Investiga- tion revealed the existence of an antibiotic, emimycin,'~'~ which is the 1,2-dihydr0-2-oxopyrazine 4-oxide, and which structurally resembles uracil. Indeed, reversal of the emimycin inhibition of Escherichia coli by pyrimidines confirmed the structural analogy."

Since no data have been published on the effect of emimycin on mammalian, in particular, tumor cells, we under- took the synthesis of this compound by a new and facile route, and examined its effect upon various cell systems. We also prepared other substituted pyrazines, among which the orotic acid analog showed particularly marked biological activity.

Finally, since nucleoside derivatives frequently offer metabolic advantages over the corresponding bases, particularly with respect to drug resistance, we have prepared the nucleoside derivative of emimycin, and have compared its biological activity with that of the antibiotic. The results of these studies are provided in this paper. A preliminary account of these data has been presented.12


(a) Chemical Results. The synthesis of 1,2dihydro-2- oxopyrazine 4-oxide (emimycin) starting from pyrazine-2- carboxamide via 2-aminopyrazine 4-oxide has been reported by Terao" and independently by Palamidessi and Ber- nardi.13 Because of the low yield of the reported procedure (overall yield based on pyrazine-2-carboxamide was 4%) and

our need for the compound as a precursor of the nucleoside, an alternative route for the preparation of 1,2-dihydro-2- oxopyrazine 4-oxide was considered, starting from the readily available 1.2-dihydro-2-oxopyrazine (Scheme I). Treat- ment of Ia-d with PhCOCl in pyridine gave cryst IIa and syrupy IIb-d. The 2-benzoyloxy derivs (1Ia-d) were oxidized with m-chloroperbenzoic acid, to furnish the corresponding cryst 2-benzoyloxy 4-oxides (Ma-d). Removal of the protecting Bz group from I I I a d with MeONa in MeOH gave 1,2-dihydr0-2-oxopyrazine 4-oxides (IVa-d). Hydroly- sis of the Me ester function of IVd was accomplished by treatment with KOH soln at room temp, to give 1 ,2-dihydro-2-oxo-6-carboxypyrazine 4-oxide (V). The position of the oxide function of IVa-c and V was established by comparison of the uv absorption spectra of IVa-c and V with those of 1,2-dihydro-2-0xopyrazine 4-oxide,lo 1,2-dihydro- 2-0~0-3,6-dimethylpyrazine 4-oxide ,14 and 1,2-dihydro-2- oxo-3-isobutyl-6-sec-butylpyrazine 1 -oxide.'' 1,2-Dihydro- 2-thiopyrazine 4-oxide (VI) was prep from 2-chloropyrazine 4-oxide by treatment with NaHS at room temp.

Wagner and Frenzel16 and later Reisser and Pfleiderer" have shown that the direct glucosidation of the Ag or Hg salts of 1,2-dihydr0-2-oxopyrazine furnishes the 0-glucoside and only traces of N-glucoside. Their attempts to rearrange the 0-glucoside to the N-glucoside were unsuccessful. As a result, a different route was chosen by us for prepn of the riboside. Ribosidation of the Me& deriv of 1,2-dihydro-2-oxopyrazine in the presence of TiC1418 in boiling 1,2- dichloroethane afforded only the N-glycoside. No 0-glycoside was detected by tlc in the reaction mixt. The syrupy 1,2-dihydro-l-(2,3,5-tri-O-acetyl-/3-D-ribofuranosyl)-2-oxopyrazine (VIII) was deacetylated by MeONa in MeOH at 22' to give 1,2-dihydro-1 -(/3-D-ribofuranosyl)-2-oxopyrazine (IX), which was purified by column chromatog on silica gel. The nucleoside IX has been assigned the /3 configuration on the basis of the trans rule. Attempts to prep the N- oxide deriv of VI11 or IX by oxidn with m-ClC&CO&I or

Pyrazine Ribonucleosides Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 165

Scheme I 0 0 1 fxR - XX" -

R"O N R' 0 R' R"O R' H

0 4

LNL 0 N COOH

H H

IVa-d 1IIa-d la-d Ila-d

a-d, R" = C,H,CO a, R,R' = H b, R = CH,; R' = H

c, R = H; R' = CH, d, R = H; R' = COOCH,

V

0 I - r"]

S N H

0 ?

0 I

CH3 CH,

VI1 VI

0 Acowc Hou TiC1, _*

OAc OAc o k c OAc OH OH

VllI IX

0 1

0 1) 0 I

0 1 AcOCH, OAc

0 1

IVa -----C

(CH,),SiO

--+ Acokd Hov/ d

OAc OAc

o k c o k c OH OH

X XI

F3CC0& were unsuccessful. The trcarment of both VI11 and IX with peracids under various conditions" resulted in the gradual loss of uv absorption and yielded complex mixts. Treatment of 1,2-dihydro-l -methyl-2-oxopyrazine with m-ClC&CO& in boiling 1,2-dichloroethane afforded a low yield of 1,2-dihydro-1 -methyl-2-oxopyrazine 4-oxide (VII), demonstrating that the oxidn of N-1-substituted 1,2- dihydro-2-oxopyrazines proceeds only with difficulty. This finding prompted us to investigate the glycosidation of the 1,2-dihydro-2-oxopyrazine 4-oxide. Reaction of the Me3 Si deriv of IVa with 1,2,3,5-tetra-O-acetyl-P-D-ribofuranose in presence of TiC14 in boiling 1,2-dichloroethane produced 1,2-dihydro- 1(2,3,5-tri-O-acetyl-~-D-ribofuranosyl)-2-oxopyrazine 4-oxide (X), which was purified by column chromatog on silica gel. Deacetylation of X with MeONa in MeOH at 22" gave 1,2-dihydro-1 -(@-D-ribofuranosy1)-2-0~0-

pyrazine 4-oxide (XI). The structure of XI was established on the basis of its uv and ir absorption spectra and its elemental analysis. The spectral data pertaining to the newly prepd compds are summarized in Table I ,

(b) Biological Results. The pyrazine derivs synthesized fall into 3 categories. One comprises compds with substituents at the 2 position only, which cannot strictly be considered analogs of the natural pyrimidines. The second group includes 2-substituted derivs which, in analogy with uracil, carry an 0 atom at the 4 position. The third class of compds is made up of nucleoside derivs of some of the analogs in the first and second categories.

The compds lacking the 4-oxide did not exert any inhibitory effects on the microbiol test systems used, whereas in the presence of the N-oxide, marked inhibition of cell growth resulted (Table 11). The same relationship applies to

166 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Bobek and BIoch

Table 11. Growth Inhibitory Effect of Pyrazine Analogs of Pyrimidines0

I E ' m - 1 %

U

.- 2 e x &

- I

0oooornoo

Q a

V gi

Concn for 50% growth inhibition of

S. faecium, E. coli, _- Derivative of pyrazine 4-oxide M M

1,2-Dihydro-2-oxo 8 X 1 X 1,2-Dihydro-2-oxo-l-(p-D-ribofuranosyl) 8 X l o -" 1 x lo-' 1,2-Dihydro-2-oxo-6-carboxy 6 x lo-' 4 x 1,2-Dihydro-2-0~0-6-carbome thoxy 4 x io-s >io- 1,2-Dihydr0-2-thio 7 x 5 x 2Chloro 2 x 10-5 3 x 10-5 2-Ethoxy 2 x 1 0 - 5 1 x 10-5 1,2-Dihydr0-2-oxo-l-methyl > i o 4 >10-3

1,2-Dihydro-2-oxo-6-methyl 9 x 1 0 - ~ 1,2-Dihydro-2-oxo-5-methyl 8 X low4

aThe corresponding compds without the N-oxide function, including 1,2dihydro-l-(p-D-ribofuranosyl)-2-oxopyrazine, did not produce any growth inhibition at M.

the nucleoside derivs, the N-oxide being required for activity. At IOm4M, none of the compds which showed inhibitory activity in the bacterial systems, inhibited the growth of L-1210 cells in vitro by more than 10-20%.

The extent of growth inhibition produced in the microbial system varied according to the nature of the substituents on the ring, and the bacterial species used for the assay. In both test systems, emimycin and its ribonucleoside deriv were equally active, suggesting that they exert their activity via a common metabolic intermediate. This suggestion received support from the observation that strains of S. faecium and E. coli resistant to 10-3M emimycin are cross-resistant to the emimycin nucleoside and, similarly, that the strains resistant to 10-3M of the nucleoside are cross-resistant to the base,

Replacement of the 2-oxo group of emimycin by = S decreased the inhibitory activity of the antibiotic in the microbiol systems from 5- to 10-fold, whereas a 3-fold decrease in activity occurred when the hydroxyl group was replaced by C1. The 2ethoxy deriv was approx 3 times less effective than emimycin against S. faecium, but was as active as the antibiotic against E. coli.

The introduction of Me into the 1 position of emimycin abolished the inhibitory activity of the antibiotic, whereas Me at its 5 or 6 position allowed for some inhibitory activity in S. faecium, but not in E. coli. As detd in the S. faecium system, the 5-Me deriv of emimycin could not replace thymine for growth of the organism in the basic medium free of folate and contg adenine as the purine source.

of orotate was more effective than emimycin against S. faecium, but was less active than the antibiotic against E. coli. This analog retained full activity in the emimycin and emimycin ribonucleoside resistant strains of S. faecium, suggesting a different metabolic path for its activation or a different site of action.

(Table 111) that the inhibitory effects of emimycin and its ribonucleoside were prevented competitively by uracil and cytosine and their nucleosides at concns ranging from to 10-6M. Over this concn range, the inhibition exerted by the orotate analog was reversed in a noncompetitive (product) manner, indicating that the pyrazine analog of orotate does not interfere with the conversion of exogenous uracil or uridine to UMP and to further metabolic intermediates, but likely exerts its effect along the de novo path leading to UMP. Further support for this deduction comes from the

Of marked interest is the finding that the pyrazine analog

An inhibition analysis carried out with S. faecium showed

Pyrazine Ribonucleosides Journal ofMedicinal Chemism, 1972, Vol. 15, No. 2 161

Table 111. Reversal, by Pyrimidines, of the Inhibition of Growth of S. faecium by Pyrazines Inhibition indexa obtained with 1,2-Dihydro-l -@-D-ribofu-

1,2-Dihydro-2-oxopyrahe ranosyl)-2-oxopyrazie 1,2-Dihydro-2-oxo-6- Substrate 4-oxide 4-oxide carboxypyrazine 4-oxide

Uridine 2'-Deoxyuridine Cytidine 2'-deoxy cytidine Uracil Cy tosine Thymine Thymidine Orotic acid

20 10 0.9 1 0.7 0.4 2- to 3-fold reversal at lO+M

10-fold reversal at 10"M 2-fold reversal at 10-3M

20 Noncompetitive 11 Noncompetitive 0.8 Noncompetitive 2 Noncompetitive 0.9 Noncompetitive 0.5 Noncompetitive 2- to 3-fold reversal at 10-3M

2-fold reversal at 10'SM 6

2- to 3-fold reversal at lO-'M 10-fold reversal at 10-3M 10-fold reversal at 10-3M

"[I]/[ SI for 50% growth inhibition, with substrate concns ranging from lo-, to lo.-, M.

observation that orotate reversed the inhibitory effect of the orotate analog in a competitive manner, whereas, at the highest concn of lo-' M, this metabolite prevented the growth inhibition exerted by emimycin and its nucleoside by only 1.5- to 2-fold.

The fact that orotate does not reverse the emimycin or emimycin riboside action extensively, might mean that the uracil analogs act at the base or ribonucleoside level per se, or that competition for their activation via a pyrophospho- rylase or kinase is exerted more effectively by the exoge- nously supplied uracil or uridine than by UMP, obtainable in a possibly limited amount, from exogenous orotate.

At 10-3M, the highest concn used, thymine or thymidine reversed the inhibitory effect of the analogs from 4- to 15- fold in a noncompetitive manner, the orotate analog being reversed more effectively than emimycin or its ribonucleoside. This relatively small extent of reversal suggests a spar- ing effect of thymine or thymidine on de novo pyrimidine synthesis. It might be added here, that, in analogy with our observations in S. faecium, the inhibition of the growth of E. coli by emimycin was also found to be competitively reversed by uracil."

It remains to be detd why, as compared to their effect on the bacteria, the analogs are only marginally active against the mammalian cells. However, because of this pronounced selectivity, the potential use of emimycin and its derivs in antimicrobial chemotherapy appears worthy of exploration.

Experimental Section?

(4.00 g, 0.02 mole) was dissolved in dry 1,l-dichloroethane (C,H,Cl,, 50 ml). To this soln mchloroperoxybenzoic acid (85% product, 4C1C6H,C0,H, 5 .OO g) was added and the reaction mixt was heated at 55-60' for 15 hr. The soln was then cooled to 20", washed successively with a satd soln of NaHCO, (80 ml) and H,O (80 ml), and dried (Na,SO,). The solvent was removed, and the remaining syrupy product was dissolved in Et,O (15 ml). Cryst IIIa which sepd from the soln after standing for several hours at 20" was collected by filtration and recrystd from MeOH.

2-Benzoyloxy-5-methylpyazine 4-Oxide (IIIb). 5-Methyl-l,2- dihydro-2-oxopyrazine2' (2.20 g, 0.02 mole) was added to dry pyridine (20 ml). To this mixt BzCl(3.1 g, 0.022 mole) was added dropwise with stirring. The reaction mixt, after standing at 20" for 15 hr, was poured into 50 ml of icecold H,O. An amber oil sepd which was extd with PhMe (80 ml). This ext was washed with H,O (4 X 100 ml) and dried (Na,SO,). PhMe was removed by evapn and the remaining oily residue was dissolved in xylene (50 ml). This soln was concd to a thick oil which was dissolved in C,H.Cl, (80 ml). 4-

2-B enzoyloxypyrazine 4-Oxide (IIIa). 2-Benzoyloxypyrazineao

+Melting points were taken on a Fisher-Johns apparatus and are uncorrected. Optical rotation are equilibrium values and were detd on a Jasco ORD/UV Model 5 instrument at 0.2% concn in MeOH. Uv spectra were obtained on a Cary Model 14 recording spectrophotometer. The ir absorption spectra were detd in pressed KBr disks with a Perkin-Elmer Model 52 1 spectrophotometer. Solvent concn was conducted under reduced pressure in a rotary evaporator.

ClC,H,CO,H ( 5 g) was added to the soln and the work-up proceeded as described for the prepn of IIIa.

2-Benzoyloxy-6-methylpyrazine 4-oxide (IIIc) was prepd from 1,2-~ihydr0-2-0~0-6-methylpyrazine~' by the procedures described under IIIb.

2-Benzoyloxy-6-carbomethoxypyrazine 4-Oxide (IIId). BzCl (3.1 g, 0.022 mole) was added dropwise with stirring to a mixt of Idla (3.1 g, 0.02 mole) and dry pyridine (20 ml), and the mixt was worked up as above. To this soln, 4ClC,H,CO,H (4 g) was added, and the mixt was heated at reflux temp for 2 hr. The soln was cooled to 3040" and addnl 4C1C6H,C0,H (2.5 g) was added. The reaction mixt was heated at reflux temp for another 2 hr, cooled to 20", and dild with CHCl, (80 ml). The soln was washed with a satd soh of NaHCO, (80 ml) and then with H,O, and dried (Na,SO,). It was concd to an oil, which was chromatogd on a silica gel column in C,&-EtOAc (6: 1, v/v). Concn of the eluate containing Id (as shown by tlc) furnished cryst substance, which was recrystd from Me,CO- Et,O. The column was subsequently washed with Me,CO-MeOH (6: 1, v/v). The eluate was evapd to dryness and the residue was recrystd twice from EtOH furnishing 400 mg of 1,2-dihydro-2-oxo- 6-carbomethoxypyrazine 4-oxide (IVd): Vmax cm-' 3 110, 3090,

1,2-Dihydro-2-oxopyrazine 4-Oxide (IVa). IIIa (2.16 g, 0.01 mole) was dissolved in MeOH (100 ml) and a 2 N MeONa-MeOH soln was added dropwise with stirring until pH 9-10 was achieved. MeOH-washed Dowex 50 (H') resin (10 ml) was added to the soln with stirring. The resin was filtered and washed with MeOH (80 ml). The MeOH s o h was evapd, the residue suspended in EtOH (50 ml) and reevapd to a solid, which was recrystd from 96% EtOH.

1,2-Dihydro-2-oxo-5-methylpyrazine 4-oxide (IVb) was prepd from IIIb in the same manner as described for IVa.

1,2-Dihydro-2-oxo-6-methylpyrazine 4-oxide (1Vc) was prepd from IIIc (2.30 g, 0.01 mole) following the procedure described for the prepn of IV, except that a larger vol of MeOH (200 ml) was used because of the poor solubility of both IIIc and IVc.

1,2-Dihydro-2-oxo-6-carbomethoxypyrazine 4-oxide (IVd) was prepd from IIId (2.74 g, 0.01 mole) in the same manner as described for IVc: imaX 3160,3080,2940,2825 (NH, CH), 1740 (C=O car-

cm-' (N+O, COC).

g, 0.001 mole) was dissolved in a KOH (0.4 g) soln in H,O (60 ml). After 2 hr at 22", the s o h was applied to a column (1 X 20 cm) of Dowex 50 (Ht) resin. The column was washed with H,O and the acid fraction was collected and evapd. The residue was recrystd from H,O: imax 3500 (OH), 3080,2820 (broad) (NH, CH), 1730 ( G O ) , 1630 (broad) (eo, G C , G N ) , 1275, 1245 cm-' (broad) (N-tO).

1 ,2-Dihydro-2-thiopyrazine 4-Oxide (VI). 2-Chloropyrazine 4- oxide23 (1.30 g, 0.01 mole) was dissolved with stirring in a soln of NaHS, prepd by satg a soln of 0.46 g Na in EtOH (50 ml) with H,S at 5". The reaction mixt was stirred at 22" for 15 hr and evapd, the residue was suspended in EtOH-H,O (1: 1, v/v; 40 ml), and the pH of the mixt was adjusted to 4-5 with HCl. The mixt was then evapd to dryness, the residue was,extd with boiling 98% EtOH (3 X 150 ml) and filtered while hot. The filtrate was concd to 50 ml and allowed to stand at 5" for 4 hr. The product was collected by filtration and recrystd from 96% EtOH: pmaX 3125,3080,2945, 2840 (NH, Cy), 1250,1215,1130,1105 cm-' (ES,N-tO).

1,2-Dihydro-l-methyl-2-oxopyrazine 4-Oxide (VII). A soln of 1 ,2-dihydro+methy1-2-0xopyrazine~~ (0.550 g, 0.005 mole) and 4ClC,H4C0,H (1.0 g) in C,H,Cl, (60 ml) was heated at reflux temp for 2 hr and was then cooled to 30-40". Addnl4-ClC,H,CO,H (1.0

2960 (CH), 1745 (C=O), 1260 - 1250 (N-+O).

boxyl), 1660-1620 ( G O , C=C, C=N) 1295,1260,1230,1025, 825

1,2-Dihydro-2-oxo4-carboxypyrazine 4-Oxide (V). IVd (0.340

168 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Robek, Whistler, and Bloch

g) was added, and the reaction mixt was heated at reflux temp for another 2 hr. The soh was cooled to 22", concd to 5-8 ml, and chromatogd on a silica gel co!umn with PhH-Me,CO (1: 1, v/v). The fraction contg IV (as shown by tlc) was evapd, and the residue was recrystd from EtOH-Et,O.

1,2-Dihydro-l~p-D-ribofuranosyl)-2-oxopyr~~e (IX). A mixt of 1,2-dihydro-2-oxopyrazine (4.3 g, 0.045 mole), (Me,Si),NH (4 ml), and Me,SiC1(3 ml) was heated at reflux temp with stirring for 1 hr. After cooling to 22", PhMe (80 ml) was added, and the soln was concd to an oil, which was dissolved in dry C,H,Cl, (250 ml). 1,2,3,5-Tetra-O-acetylQ-D-ribofuranose (9.55 g, 0.03 mole) and TiC1, (3 ml) were added to this soln, and the reaction mixt was heated for 5 hr at reflux temp with stirring and exclusion of atm moisture. The mixt was cooled to 22" and poured slowly with vigorous stirring into a satd soln of NaHCO, (1500 ml). It was filtered through a Celite pad. which was then washed with CHC1,. The org layer was sepd, washed with H,O, and dried (Na$O,). The solvent was evapd, and the residue was dissolved in MeOH (100 ml). A catalytic amt of MeONa was added to this soln and after 0.5 hr at 22", the soln was neutralized with HCl and evapd. The syrupy residue was dissolved in MeOH (5 ml) and applied to a dry column (2.5 X 8 cm) of silica gel. A Me,CO-MeOH (4: 1, v/v) mixt was applied to the column and the eluate was evapd. The syrupy residue was dissolved in Me,CO-EtOH (1:2, v/v, 30 ml); and crystn occurred after standing of the soln overnight at 22". Recrystn from EtOH gave pure IX: [a]"D -21.2", om,, 3400,3280 (OH), 2945-2920 (CH), 1645 cm'' (GO)..

1,2-Dihydro- l-(p-D-ribofuranosyl)-2-oxopyrazine 4-Oxide (XI). A mixt of IVa (1.12 g, 0.01 mole), (Me,Si),NH (3 ml), and Me,SiCl (3 drops) was heated for 0.5 hr at 90-95" with stirring and exclusion of atm moisture. The resulting soln was cooled to 22", dild with dry PhMe (40 ml), and evapd. The remaining solid was evapd once more from dry PhMe (40 ml), and dissolved in dry C,H,Cl, (100 ml). To this soln was added 1,2,3,5-tetra-O-acetylQ-D-ribofuranose (3.20 g, 0.01 mole) and TiC1, (3 ml), and the reaction mixt was heated at reflux temp with stirring and exclusion of atm moisture for 4 hr. It was cooled to 22" and poured slowly, with vigorous stirring, into a satd s o h of NaHCO, (500 ml). The mixt was filtered through a Ce- lite pad, which was subsequently washed with CHCl,. The org layer was sepd, washed with H,O, and dried (Na,SO,). The SYNP remaining after removal of the solvent was purified by column chromatog on silica gel with CHC1,-Me,CO (7: 1, v/v) as the eluent. The syrupy 1,2-dihydro-l-(2,3,5-tri~-acetyl-p-D-ribofuranos~l)-2-oxopyrazine 4-oxide (X) (480 mg) obtd by evapn of the solvent was dissolved in MeOH (50 ml). A catalytic amt of MeONa was added to this soln and after 0.5 hr at 22" the soln was neutralized with Dowex 50 (H+) resin. The resin was filtered and washed with MeOH (20 ml). The

syrup remaining after removal of the MeOH was dissolved in 96% EtOH and purified by column chromatog on silica gel with Me,CO as the eluent. Me,CO was removed by evapn and the remaining solid was recrystd from 96% EtOH: [a]"D -204.2", V m a x 3400, 3300 (OH), 2960,2925 (CH), 1660 (GO), 1230, 840,830 cm-' (N-+O).

published p r e v i o ~ s l y . ~ ~ Biological Assays. The techniques used for these detns have been

References (1) M. J. Robins, B. L. Currie, R. K. Robins, and A. Bloch, Proc.

(2) F. Sorm, A. Piskala, A. Cihak, and J. Vesely, Experientia, 20,

(3) F. Sorm and I . Vesely,Neoplasma, 11, 123 (1964). (4) W. H. Prusoff, L. G. Lajtha, and A. D. Welch, Biochem.

(5) R. Schindler and A. D. Welch, Science, 125, 548 (1957). (6) J. J. Jaffe, R. E. Handschumacher, and A. D. Welch, Yale J .

(7) F. Sorm and H. Keilova, Experientia, 14, 215 (1958). (8) W. H. Prusoff, Biochem. Pharmacol., 2 , 221 (1959). (9) M . Terao, K. Karasawa, N. Tanaka, H. Yonehara, and H.

Umezawa, J. Antibiot. Ser. A , 13,401 (1960).

Amer. Ass. Cancer Res., 10, 73 (1969).

202 (1964).

Biophys. Acta, 20, 209 (1956).

Biol. Med., 30, 168 (1957).

(10) M. Terao, ibid., Ser. A , 16, 182 (1963). (11) J. R. DeZeeuw and E. J. Tynan, ibid., Ser. A , 22, 386 (1969). (12) M. Bobek and A. Bloch, Pharmacologist, 13,260 (1971). (13) G. Palamidessi and L. Bernardi, Gazz. Chim. Ital., 93, 343,

(14) B. Klein, E. O'Donnell, and J. M. Gordon, J. Org. Chem., 29,

(15) J. 0. Dutcher, J. Biol. Chem., 232,785 (1958). (16) G. Wagner and H. Frenzel, Z . Chem., 5,24 (1965). (17) F. Reisser and W. Pfleiderer, Chem. Ber., 99,542 (1966). (18) U. Niedballa and H. Vorbriiggen, Angew, Chem. Int. Ed., 9,

(19) E. Ochiai, "Aromatic Amine Oxides," Elsevier, Amsterdam,

(20) K. H. Schaaf and P. E. Spoerri,J. Amer. Chem. SOC., 71, 2043

(21) G. Karmas and P. E. Spoerri, ibid., 74, 1580 (1952). (22) H. Foks and J. Sawlewics, Acta Polon. Pharm., 23,437 (1966). (23) B. Klein, N. E. Hetman, and M. E. O'Donnell, J. Org. Chem,

(24) G. W. H. Cheeseman, J. Chem. SOC., 242 (1960). (25) M. Bobek, R . L. Whistler, and A. Bloch, J. Med. Chem., 13,

1963.

2623 (1964).

461 (1970).

Netherlands, 1967.

(1949).

28, 1682 (1963).

411 (1970).

Synthesis and Biological Activity of 4'-Thio Analogs of the Antibiotic Toyocamycin

M. Bobek,* R. L. Whistler, and A. Bloch Department of Experimental Therapeutics, Roswell Park Memorial Institute, New York State Department of Health, Buffalo, New York, and Department of Biochemistry, Purdue University, Lafayette, Indiana. Received July 30, 1971

The 4'4hio analog of the antibiotic toyocamycin was prepared b y condensation of 2,3,5-tri-O-acety1-4- thio-D-ribofuranosyl chloride with the chloromercuri derivative of 4-acetaminod-bromo-S-cyanopyrrolo- [ 2,3-d]pyrimidine, followed b y removal of the protecting groups with MeOH-NH3 and removal of Br with HJPd catalyst. Condensation with the chloromercuri deriIative of 4-chloro-6-bromo-5-cyanopyrrolo- [2,3-d]pyrimidine, followed b y t reatment with MeOH-NH3 a t 5 , effected removal of the protecting groups and nucleophilic substi tution of the Br group to furnish 4-chloro-6-amino-5-cyano-7-(4-thi~-/3-D- ribofuranosyl)pyrrolo[ 2,3-d]pyrimidine. When t reatment with MeOH-NH3 was carried out a t 1 2 0 , 4,6- diamino-5-cyano-7-(4-thio-~-~-ribofuranosyl)pyrrolo [ 2,341pyrimidine was formed. The 4'-thio derivatives proved to be effective inhibitors of the growth o f leukemif L- I2 1 0 cells in vitro, their concn for 50% reduction o f growth ranging f rom 4 X lo-' to 5 X 10-6M. 4 -Thiotoyocamycin retained full inhibitory activity against a strain of Streptococcus faecium resistant to I O - ~ M toyocamycin.

The antibiotic toyocamycin' is an analog of adenosine in which N-7 of the imidazole ring is replaced by C, to which is attached a CN group.' In exptl systems, this antibiotic showed marked antitumor activity: but the severe local toxoci ty in man4 which it produced l imited its clinical usefulness.

In an attempt at decreasing this toxicity, t w o structural modifications of the toyocamycin molecule were made. In one, the ring 0 of the carbohydrate moiety was replaced b y S; the other involved, in addition to this replacement, the substi tution of the 4 and 6 position of the heterocycle with C1 and amino groups, resp. The results obtained in in vitro


Synthesis and biological activity of 4'-thioanalogs of the antibiotic toyocamycin

M. Bobek, R. L. Whistler, and A. BlochJ. Med. Chem., 1972, 15 (2), 168-171• DOI: 10.1021/jm00272a011 • Publication Date (Web): 01 May 2002






168 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Robek, Whistler, and Bloch

g) was added, and the reaction mixt was heated at reflux temp for another 2 hr. The soh was cooled to 22", concd to 5-8 ml, and chromatogd on a silica gel co!umn with PhH-Me,CO (1: 1, v/v). The fraction contg IV (as shown by tlc) was evapd, and the residue was recrystd from EtOH-Et,O.

1,2-Dihydro-l~p-D-ribofuranosyl)-2-oxopyr~~e (IX). A mixt of 1,2-dihydro-2-oxopyrazine (4.3 g, 0.045 mole), (Me,Si),NH (4 ml), and Me,SiC1(3 ml) was heated at reflux temp with stirring for 1 hr. After cooling to 22", PhMe (80 ml) was added, and the soln was concd to an oil, which was dissolved in dry C,H,Cl, (250 ml). 1,2,3,5-Tetra-O-acetylQ-D-ribofuranose (9.55 g, 0.03 mole) and TiC1, (3 ml) were added to this soln, and the reaction mixt was heated for 5 hr at reflux temp with stirring and exclusion of atm moisture. The mixt was cooled to 22" and poured slowly with vigorous stirring into a satd soln of NaHCO, (1500 ml). It was filtered through a Celite pad. which was then washed with CHC1,. The org layer was sepd, washed with H,O, and dried (Na$O,). The solvent was evapd, and the residue was dissolved in MeOH (100 ml). A catalytic amt of MeONa was added to this soln and after 0.5 hr at 22", the soln was neutralized with HCl and evapd. The syrupy residue was dissolved in MeOH (5 ml) and applied to a dry column (2.5 X 8 cm) of silica gel. A Me,CO-MeOH (4: 1, v/v) mixt was applied to the column and the eluate was evapd. The syrupy residue was dissolved in Me,CO-EtOH (1:2, v/v, 30 ml); and crystn occurred after standing of the soln overnight at 22". Recrystn from EtOH gave pure IX: [a]"D -21.2", om,, 3400,3280 (OH), 2945-2920 (CH), 1645 cm'' (GO)..

1,2-Dihydro- l-(p-D-ribofuranosyl)-2-oxopyrazine 4-Oxide (XI). A mixt of IVa (1.12 g, 0.01 mole), (Me,Si),NH (3 ml), and Me,SiCl (3 drops) was heated for 0.5 hr at 90-95" with stirring and exclusion of atm moisture. The resulting soln was cooled to 22", dild with dry PhMe (40 ml), and evapd. The remaining solid was evapd once more from dry PhMe (40 ml), and dissolved in dry C,H,Cl, (100 ml). To this soln was added 1,2,3,5-tetra-O-acetyl~-D-ribofuranose (3.20 g, 0.01 mole) and TiC1, (3 ml), and the reaction mixt was heated at reflux temp with stirring and exclusion of atm moisture for 4 hr. It was cooled to 22" and poured slowly, with vigorous stirring, into a satd s o h of NaHCO, (500 ml). The mixt was filtered through a Ce- lite pad, which was subsequently washed with CHCl,. The org layer was sepd, washed with H,O, and dried (Na,SO,). The SYNP remaining after removal of the solvent was purified by column chromatog on silica gel with CHC1,-Me,CO (7: 1, v/v) as the eluent. The syrupy 1,2-dihydro-l-(2,3,5-tri~-acetyl-p-D-ribofuranos~l)-2-oxopyrazine 4-oxide (X) (480 mg) obtd by evapn of the solvent was dissolved in MeOH (50 ml). A catalytic amt of MeONa was added to this soln and after 0.5 hr at 22" the soln was neutralized with Dowex 50 (H+) resin. The resin was filtered and washed with MeOH (20 ml). The

syrup remaining after removal of the MeOH was dissolved in 96% EtOH and purified by column chromatog on silica gel with Me,CO as the eluent. Me,CO was removed by evapn and the remaining solid was recrystd from 96% EtOH: [a]"D -204.2", V m a x 3400, 3300 (OH), 2960,2925 (CH), 1660 (GO), 1230, 840,830 cm-' (N-+O).

published p r e v i o ~ s l y . ~ ~ Biological Assays. The techniques used for these detns have been

References (1) M. J. Robins, B. L. Currie, R. K. Robins, and A. Bloch, Proc.

(2) F. Sorm, A. Piskala, A. Cihak, and J. Vesely, Experientia, 20,

(3) F. Sorm and I . Vesely,Neoplasma, 11, 123 (1964). (4) W. H. Prusoff, L. G. Lajtha, and A. D. Welch, Biochem.

(5) R. Schindler and A. D. Welch, Science, 125, 548 (1957). (6) J. J. Jaffe, R. E. Handschumacher, and A. D. Welch, Yale J .

(7) F. Sorm and H. Keilova, Experientia, 14, 215 (1958). (8) W. H. Prusoff, Biochem. Pharmacol., 2 , 221 (1959). (9) M . Terao, K. Karasawa, N. Tanaka, H. Yonehara, and H.

Umezawa, J. Antibiot. Ser. A , 13,401 (1960).

Amer. Ass. Cancer Res., 10, 73 (1969).

202 (1964).

Biophys. Acta, 20, 209 (1956).

Biol. Med., 30, 168 (1957).

(10) M. Terao, ibid., Ser. A , 16, 182 (1963). (11) J. R. DeZeeuw and E. J. Tynan, ibid., Ser. A , 22, 386 (1969). (12) M. Bobek and A. Bloch, Pharmacologist, 13,260 (1971). (13) G. Palamidessi and L. Bernardi, Gazz. Chim. Ital., 93, 343,

(14) B. Klein, E. O'Donnell, and J. M. Gordon, J. Org. Chem., 29,

(15) J. 0. Dutcher, J. Biol. Chem., 232,785 (1958). (16) G. Wagner and H. Frenzel, Z . Chem., 5,24 (1965). (17) F. Reisser and W. Pfleiderer, Chem. Ber., 99,542 (1966). (18) U. Niedballa and H. Vorbriiggen, Angew, Chem. Int. Ed., 9,

(19) E. Ochiai, "Aromatic Amine Oxides," Elsevier, Amsterdam,

(20) K. H. Schaaf and P. E. Spoerri,J. Amer. Chem. SOC., 71, 2043

(21) G. Karmas and P. E. Spoerri, ibid., 74, 1580 (1952). (22) H. Foks and J. Sawlewics, Acta Polon. Pharm., 23,437 (1966). (23) B. Klein, N. E. Hetman, and M. E. O'Donnell, J. Org. Chem,

(24) G. W. H. Cheeseman, J. Chem. SOC., 242 (1960). (25) M. Bobek, R . L. Whistler, and A. Bloch, J. Med. Chem., 13,

1963.

2623 (1964).

461 (1970).

Netherlands, 1967.

(1949).

28, 1682 (1963).

411 (1970).

Synthesis and Biological Activity of 4'-Thio Analogs of the Antibiotic Toyocamycin

M. Bobek,* R. L. Whistler, and A. Bloch Department of Experimental Therapeutics, Roswell Park Memorial Institute, New York State Department of Health, Buffalo, New York, and Department of Biochemistry, Purdue University, Lafayette, Indiana. Received July 30, 1971

The 4'4hio analog of the antibiotic toyocamycin was prepared b y condensation of 2,3,5-tri-O-acety1-4- thio-D-ribofuranosyl chloride with the chloromercuri derivative of 4-acetaminod-bromo-S-cyanopyrrolo- [ 2,3-d]pyrimidine, followed b y removal of the protecting groups with MeOH-NH3 and removal of Br with HJPd catalyst. Condensation with the chloromercuri deriIative of 4-chloro-6-bromo-5-cyanopyrrolo- [2,3-d]pyrimidine, followed b y t reatment with MeOH-NH3 a t 5 , effected removal of the protecting groups and nucleophilic substi tution of the Br group to furnish 4-chloro-6-amino-5-cyano-7-(4-thi~-/3-D- ribofuranosyl)pyrrolo[ 2,3-d]pyrimidine. When t reatment with MeOH-NH3 was carried out a t 1 2 0 , 4,6- diamino-5-cyano-7-(4-thio-~-~-ribofuranosyl)pyrrolo [ 2,341pyrimidine was formed. The 4'-thio derivatives proved to be effective inhibitors of the growth o f leukemif L- I2 1 0 cells in vitro, their concn for 50% reduction o f growth ranging f rom 4 X lo-' to 5 X 10-6M. 4 -Thiotoyocamycin retained full inhibitory activity against a strain of Strep tococcus faecium resistant to I O - ~ M toyocamycin.

The antibiotic toyocamycin' is an analog of adenosine in which N-7 of the imidazole ring is replaced by C, to which is attached a CN group.' In exptl systems, this antibiotic showed marked antitumor activity: but the severe local toxoci ty in man4 which it produced l imited its clinical usefulness.

In an attempt at decreasing this toxicity, t w o structural modifications of the toyocamycin molecule were made. In one, the ring 0 of the carbohydrate moiety was replaced b y S; the other involved, in addition to this replacement, the substi tution of the 4 and 6 position of the heterocycle with C1 and amino groups, resp. The results obtained in in vitro

4’-Thio Analogs of Toyocamycin Journal ofMedicinal Chemistry, 1972, Vol. 1.5, No. 2 169

systems showed that these deriv had biol properties different from those of the parent compd. These properties and the procedures used for the synthesis of the compds are described in this paper.

used for the prepn of pyrrolopyrimidine nucleosides. Ribosidation of the chloromercuri salt of 4-amino-5-cyanopyrrolo [2,34]pyrimidine, for example, produced a nucleoside material which, because of the low yield obtained (less than l%), could not be completely characterized.’ The heavy metal salt procedure was also used in an attempt to synthesize tubercidin, but the desired nucleoside was not obtained.6 Recently, the fusion procedure has been employed for the prepn of a number of pyrrolopyrimidine nucleosides in yields varying from 3 to 20% (based on the sugar ~ o m p o n e n t ) . ~ Application of the procedure employing trimethylsilyl derivs to pyrrolo [2,3d]pyrimidines has furnished mixts of isomeric nucleosides in high yield.’

Since the chloromercuri procedure had been used by us successfully for the prepn of 4’-thio purine nucleoside^,^ we attempted to apply this method to the synthesis of 4’- thiopyrrolo [2,34]pyrimidine nucleosides (Scheme I).

Scheme I

A. Chemical. Various methods have, in the past, been

I OAc HgCl

R = N H A c or R = CI

Io R = N H A c P R=CI

- Hoed OH OH

I b R:NH2, R ’ = Br IL R = N H ~ , R I = H m R = N H ~ , R ~ = N H ~ XI R = C I , R ’ = N H e

Treatment of 2,3,5-tri-O-acety1-4-thio-D -ribofuranosyl chloride with the chloromercuri deriv of 4-acetamino-6- bromo-5-cyanopyrrolo [2,34] pyrimidine furnished a 20% yield of the acetylated syrupy nucleoside (Ia). The structure of this product was established as 4-acetamino-6- bromo-5-cyano-7-(2,3,5-tri~-acetyl-4-thio-P-D -ribofurano- sy1)pyrrolo [2 ,3d] pyrimidine on the basis of its uv, pmr, and ir spectra. On the basis of the trans rule, the anomeric configuration was presumed to be P . Deacetylation of Ia with methanolic NH3 at 5’ gave crystalline 4-amino-6- bromo-5-cyano-7-(4-thio-P-D -ribofuranosyl)pyrrolo [2 ,3d] - pyrimidine (Ib). Dehalogenation of Ib with H2/Pd gave 4-amino-5 -cyano-7-(4-thio-fl-D -ribofuranosyl)pyrrolo [ 2,341 - pyrimidine (11,4’-thiotoyocamycin). Treatment of Ib with MeOH-NH, at 1 15-120’ produced 4,6-diamino-5-cyano-7- (4-thio-0-D -ribofuranosyl)pyrrolo [2 ,34] pyrimidine (111).

The site of attachment of 4-thio-D-ribose was assigned to N-7, on the basis of uv absorption. As a result of substitution of S for 0 in the ribofuranosyl moiety, the uv absorption maxima for I1 in EtOH, pH 1 and pH 12, exhibit small bathochromic shifts relative to those obsd for toyocamycin.’ Similar bathochromic shifts were exhibited by the 9,4’-thio- (3-D -ribofuranosyladenine ring. 9,

Treatment of I1 with Me2C0, 2,2-dimethoxypropane, and TsOH produced 4-amino-5-cyano-7-(2,30-isopropylidene- 4-thio-P-D -ribofuranosyl)pyrrolo [2,34] pyrimidine (IV), indicating the presence of a cis diol.

To evaluate the influence of certain substituents in the pyrrolopyrimidine ring structure on the yield, and the site of ribosidation, 4-chloro-6-bromo-5-cyanopyrrolopyrimi- dine was used in the condensation reaction. The chloromercuri salt of 4-chloro-6-bromo-5-cyanopyrrolo [ 2,341 - pyrimidine, prepd in 85% yield, was condensed with 2,3,5- tri4-acetyl-4-thio-D -ribofuranosyl chloride in dry PhMe to give 4-chloro-6-bromo-5-cyano-7-(2,3,5-tri0-acetyl-4- thio-0-D -ribofuranosyl)pyrrolo [2 ,34] pyrimidine (V), which was isolated as a cryst substance by column chromatog. The yield of V was 25.4%. The site of glycosidation of V was assigned at N-7 on the basis of the uv spectra. Removal of the protecting groups from V was accomplished with MeOH- NH3 at 5’. Elemental analysis and spectral examn of the product isolated indicated that nucleophilic displacement of the 6-Br by NH2 had occurred, to furnish the 4-chloro-6- amino-5-cyano-7-(4-thio-P-D -ribofuranosyl)pyrrolo [ 2,341 - pyrimidine (VI). This finding was rather unexpected, since treatment of 6-bromo-4-chloro-5-cyano-7-(2,3,5-tri-O-acetyl- 0-D -ribofuranosyl)pyrrolo [2 ,34] pyrimidine with MeOH- NH3 at 110’ affected removal of the blocking groups with concomitant displacement of the 4-C1 to give 4-amino-6- bromo-5-cyano-7-(P-D -ribofuranosyl)pyrrolo [2 ,34] pyrimidine.7 The susceptibility of the 6-Br of V to nucleophilic substitution is presumably due to the ability of the S atom to accommodate both positive and negative charges.

Treatment of V with MeOH-NH, at 115” produced 4,6- diamino-5-cyano-7-(4-thio-fl-D -ribofuranosyl)pyrrolo [ 2,341 - pyrimidine (111), which was identical in every respect with the compd prepd from Ib, thus proving that the site of ribosidation of the chloromercury deriv of both 4-chloro-6- bromo-5-cyanopyrrolo [2,34] pyrimidine and 4-acetamino- 6-bromo-5-cyanopyrrolo [2 ,34] pyrimidine is the same.

B. Biological. The effect of the 4’-thionucleosides on the growth of Sfreptococcus faecium and leukemia L-12 10 cells is summarized in Table I. In the bacterial test system, 4’-thiotoyocamycin was approximately 10 times more effective an inhibitor than was toyocamycin, while the 6-amino deriv was 3 times more potent. The 6-Br deriv was inactive at 10m3M. This variance in activity constitutes a parallel to the marked growth-inhibitory activity of 8-aminoadenosine and the inactivity of 8-bromoadenosine in the S. faecium system.“

The inhibitory activity of the analogs in the mammalian cell system was found to differ from their effect on the bacteria primarily by the fact that toyocamycin was approx 10 times more potent an inhibitor of L-1210 growth than was the 4’-thio analog or its 6-amino deriv. Furthermore, unlike its inactivity in the bacterial system, 4’-thiod-bromotoyocamycin was markedly inhibitory of the tumor cell growth.

Of importance for the potential chemotherapeutic use of these analogs is the finding that the active 4’-thio derivatives retained their full activity against a strain of S. faecium resistant to lO-,M toyocamycin or to the related analog

170 Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 Bobek, Whistler, and Bloch

Table I. Inhibition of Cell Growth by 4'-Thiotoyocamycin and Derivatives

.____-. Concentration (M) for 50% growth inhibition of

S. faecium resistant to Comvound S. faecium Toyocamycin Tubercidin Leukemia L-1210

Toyocamycin 9 x 10-5 > 10-3 > 10-3 4 x 4 '-Thiotoyocam ycin 7 x 7 x 10-6 7 x 10- 4 x 10-~

4'-Thio-6-bromotoyocamycin > i o 4 >io- , > 10-3 5 x 1F6 4'-Thio-6-aminotoyocamycin 4 x 10-5 4 x 10-5 4 x 10-5 6 X lo7

tubercidin (7-deazaadenosine). This finding parallels the observation reported by us p rev io~s ly ,~ showing that the 4'-thio derivative of 6-mercapto-9-(/3-D -ribofuranosyl)purine retained its activity against a strain of S. faecium resistant to the corresponding 6-mercapto-9-(/3-D -ribofuranosyl)- purine.

Whether these observed biol differences between toyocamycin and the 4'-thio derivs result in an improvement of their toxic properties remains to be detd.


ribofuranosy1)pyrrolo [ 2,341pyrimidine (Ia). 4-Acetamino-6- bromo-5-cyanopyrrolo[ 2,3-d]pyrimidine7 (5.6 g; 0.02 mole) was added to 300 ml of H,O, and to this mixt 10% NaOH soln was added dropwise with stirring until a clear soln was achieved (approx 8 ml). A soln of 5.42 g (0.02 mole) of HgCl, in 100 ml of EtOH, and 6 g of Celite was added. The pH of the mixt was adjusted to 7.5- 8 with 10% NaOH soln, and the mixt was evapd to dryness. The residue was dried by azeotropic distn with xylene. A soln of 6.30 g (0.02 mole) of 2,3,5-tri-O-acetyl-4-thio-a,p-D-ribofuranosyl chloride" prepd from 1,2,3,5-tetra-O-acety1-4-thio-p-D-ribofuranose) in 350 ml of dry PhMe was added, and the reaction mixt was stirred under exclusion of moisture at 90-95" for 50 hr. The solids were removed by filtration and washed with 200 ml of EtOAc. This soln was washed with 30% KI (2 X 80 ml) and once with H,O (100 ml) and dried (Na,SO,). The syrup remaining after removal of the solvent was purified by column chromatog on silica gel with PhH-Me,CO (8:2, v/v) as the eluent. The syrup was dissolved in 50 ml of MeOH and evapd to yield 2.2 g (19.8% of Ia: uvmax (EtOH) 288 ( e 12,580), (pH 1) 284 (12,7901, (pH 12) 289 (14,070), 298 mp (11,510); ir (CCl,) 2235 ( G N ) , 1765 (OAc), 1735 cm-' (NHAc); pmr (CDCl,) 1.95, 2.06, 2.15 (CCH,), 6.28 d ( J , J , , = 7.5 Hz), 8.83 (C,H).

4-Amino-6-bromo-5-cyano-7-(4-thio-p- D-ribofuranosy1)pyrrlo- [2,3-d]pyrimidine (Ib). Ia (1.11 g; 0.002 mole) was covered with 50 ml of MeOH-NH, (satd at O"), allowed to stand at 5' for 24 hr, and was concd to dryness at 40-45". The remaining solid material was triturated with EtOH, and crystals were collected by filtration: yield 713 mg; mp 195-197" dec. Recrystn of this solid from EtOH- H,O mixt afforded the product in 87.5% yield (675 mg): mp 199- 201" dec; [ c ~ ] , ~ D -103.1' (solvent a); uv max (EtOH) 297 (sh) ( E 10,4201, 288 (13,510), 218 (16,800), 205 (17,180), (pH 1) 285 (12,510), 234 (14,670), 210 mp (18,530);ir (Nujol) 2235 cm-' (C=N). Anal. (C,,H,,BrN50,S~ 0.5H,O) C, H, N, S.

4-Amino-5-cyano-7-(4-thio-p-D-ribofuranosyl)pyrrolo [ 2,3-d] - pyrimidine (4'-Thiotoyocamycin) (11). Ib (505 mg) was added to a mixt of MeOH (100 ml) and Pd black (100 mg). MeOH-NH, (0.5 ml, satd at 0") was added, and the mixt was hydrogenated at 25" at atm pressure for 4 hr. After removal of the catalyst by filtration and concn of the filtrate to near dryness, EtOH (8 ml) was added, and the mixt was allowed to stand at 5' for 4 hr. Removal of the crystals by filtration gave 11: yield, 350 mg (87.2%). Recrystn from an EtOH-H,O mixt afforded an analytical sample: mp 202-205'; [a]''D -88.4" (solvent a); uv (EtOH) 293 ( E 9920), 283 (14,650), 276 (sh) (14,060), 235 (9690), 208 (22,810), (pH 1) 277 (11,8101, 237 (17,960), (pH 12) 292 (sh) (9980), 281 mp (14,860); ir (Nujol) 2235 cm-l (C=N). Anal. (C,,H,,N,O,S. 0.5H,O) C , H, N, S.

4-Acetamino-6-brmo-5-cyano-7-(2,3,5-tri-O-acetyl-4-thio-~- D-

?Where analyses are indicated by the symbols of the elements, analytical results for those elements were within *0.4% of the theoretical values. Melting points were taken on a Fisher-Johns apparatus and are corrected. Optical rotations are equilibrium values and were detd on a Jasco Model ORD/uv-S at 0.1 concn in (a) DMF- H,O (1 :4, v/v) and (b) EtOH. Ir spectra were detd on a Perkin- Elmer, Model 337 spectrophotometer. Pmr spectra were obtd on a Varian A-60 instrument (Me$). Uv spectra were taken on a Cary Model 14 spectrophotometer. Solvent concn was carried out at reduced pressure in a rotary evaporator.

4,6-D~mino-5-cyano-7-(4-thio-~~D-ribofur~osyl)py~olo [ 2,341- pyrimidine (111). Method 1. V (2.126 g; 0.004 mole) was covered with 40 ml of methanolic NH, (satd at 0") and the reaction mixt was heated in a sealed tube for 14 hr at 115-120". The cooled soh was evapd to dryness, and MeOH was added to the residue and then removed until the odor of NH, was no longer evident. Dry column chromatog utilizing silica gel and CHC1,-MeOH (4: 1, v/v) as the developing solvent resolved the nucleoside and decompn products. The solvent was evapd and the residue was recrystd from H,O: yield, 0.46 g (34.7%); mp 221-222"; [aIz5D -15.2" (solvent a); uv max (EtOH) 294 ( E 19,9801, (pH 1) 298 (15,2001, 237 (19,690), 224 mp (21,080). Anal. (C,,H,,N,O,S .0.5H20) C, H, N, S.

Method 2. Ib (60 mg) was covered with 5 ml of MeOH-NH, (satd at 0") and the reaction mixt heated in a sealed tube for 14 hr at 115-1 20". The cooled soln was evapd to dryness and the residue was recrystd from H,O: yield, 36 mg. The uv spectra and the chromatographic mobilities were identical with those of the product prepared by method 1. A mmp showed no depression.

4-Amino-S-cyan0-7-(2,3-0- isopropylidine-4-thio-p-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (IV). I1 (133 mg) was dissolved in a mixt of Me,CO (10 ml), 2,2-dimethoxypropane (0.5 ml), and TsOH ' H,O (150 mg). The reaction mixt was allowed to stand at 22" for 12 hr, then 450 mg of KHCO, was added, and the mixt was stirred for 1 hr. The solids were removed by filtration and washed with 30 ml of Me,CO. The soln was evapd to dryness and the residue was extd with 5 portions (20 ml each) of CHC1,. The solvent was evapd to dryness and the residue was recrystd from a H,O-EtOH mixt: yield, 120 mg (80%); mp 112-113'; [aIz5D -119.5" (solvent b); uv max (EtOH) 292 ( E 12,670), 282 (18,680), 275 (sh) (18,340), 234 mp (12,320). Anal. (C,,H,,N,O,S. 0.5H,O) C, H, N, S .

4-Chloro-6-bromo-5-cyano-7-(2,3,5-t~-O-acetyl-4-thio-~-D-ribofuranosyl)pyrrolo[2,~-d]pyrimidine (V). To a mixt of 4-chloro-6- bromo-5-cyanopyrrolo[ 2,3-d]pyrimidine7 (2.58 g; 0.01 mole) and H,O (150 ml), a 10% NaOH soln was added dropwise with stirring until a clear soln was achieved. Then, 2.6 g of Celite and a soln of HgC1, (2.72 g; 0.01 mole) in 40 ml of 20% EtOH were added, and the mixt was stirred for an additional 30 min. The mixt of the HgCl deriv of 4-chloro-6-bromo-5-cyanopyrrolo [ 2,3-d] pyrimidine and Celite was then collected by filtration, washed with H,O and dried in a desiccator under reduced pressure, yield 6.8 g (85.3%). To 7.7 g of this mixt, dried by azeotropic distn with xylene, was added a soln of 2,3,5-tri-0-acetyl-4-thio-cw,P-D-ribofuranosyl chloride (3.15 g; 0.01 mole) in 100 ml of dry PhMe, and the reaction mixt was stirred with exclusion of moisture at 85-90" for 30 hr. It was worked up as described for the prepn of Ia, giving a pale yellow syrup. Column chromatog on silica gel with hexane-EtOAc (6:4, v/v) as the developing solvent sepd the nucleoside from the decompn products. The syrup remaining after removal of the solvent was dissolved in 10 ml of MeOH. After 5 hr at So, crystals sepd which were collected by filtration: yield, 1.35 g (25.4%); mp 134- 136". Recrystn from MeOH afforded an analytical sample: mp 136-137.5"; uv max (EtOH) 293 (sh) ( E 13,5501, 285 (14,0901, 228 (34,000), (pH 1) 292 (sh) (14,090), 284 (14,880), 228 mp (38,280); ir (CCl,) 2235 (C=N), 1755 cm-' (OAc). Anal. (C,,H,,BrClN,O,S) C, H, N, S.

4Chloro-6-amino-5-cyano-7-(4-thio-~~D-~bofuranosy~)py~o~o- [2,3-d]pyrimidine (VI). IV (1.064 g; 0.002 mole) was dissolved in 50 ml of MeOH-NH, (satd at 0") and allowed to stand at 5" for 15 hr. The soln was then evapd to dryness, EtOH (50 ml) was added to the residue and evapd to dryness. H,O (8 ml) was added to the residue, and after 4 hr the cryst compd was collected by filtration. Recrystn from H,O afforded the product in 43.1% yield (0.31 g): mp 208-210" (H,O); [ a I z 5 D -38.3' (solvent b); uv max (EtOH) 316 ( E 11,080), 240 (19,780), 213 (21,080), (pH 1) 314 (10,7901, 239 (20,140), 213 mp (23,490); ir (Nujol) 2220 cm-' ( e N ) . Anal. (C,,H,,C1N503S ' H,O) C, H, N, C1, S.

assays have been published previo~sly.~ S. faecium was grown in a defined medium free of purines but containing 1 mg/ml of folate. The resistant strains were selected by serial transfer in increasing

Biological Assays. The techniques used for the microbial

arbocycl ic hromyc in Analog Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 171

concns of inhibitor. The in vitro antitumor assays were carried out by our microassay technique which involves the introduction of 0.5 ml aliquots of the medium (RPMI 1630 + 10% calf serum) contg the various concns of the analog into 16 X 125 mm screw cap culture tubes, followed by 0.5 ml portions of medium contg 3 X lo5 L1210 cells. The cultures are incubated at 37" for 40 hr, after which the viable cells are counted by trypan blue exclusion. During this time the cell number in the controls increases approximately eight- to ninefold, with an average viability of 99%.

Miss Ginger Dutschman and Mr. R. J . Maue is gratefully acknowledged. This study was aided by grants CA-12585 and 12422 from the U. S. Public Health Service and T 4 3 6 from the American Cancer Society.

References

Acknowledgment. The excellent technical assistance of

(1) H. Nishimura, K. Katagiri, S. Kozaburo, M. Mayama, and N. Shinaoka, J. Antibiotics, Ser. A , 9, 60-62 (1956).

(2) K. Ohkuma, ibid., Ser. A , 14,343 (1961). ( 3 ) M. Sameyoshi, R. Tokuzen, and F. Fukuoka, G a m , 56,219

(1965). (4) W. L. Wilson, Cancer Chemother. Rep., 52,301 Illus.

(1968). (5) E. C. Taylor and R. W. Hendess, J. Amer. Chem. Soc., 87, 1995

(1965). (6) C. W. Noell and R. K. Robins, J. Heterocycl. Chem., 1,34

(1964). (7) R. L. Tolman, R. K. Robins, and L. B. Townsend, J. Amer.

Chem. Soc., 91,2102 (1969). (8) R. L. Tolman, G. L. Tolman, R. K. Robins, and L. B. Town-

send,J. Heterocycl. Chem., 7,799 (1970). (9) M. Bobek, R. L. Whistler, and A. Bloch, J. Med. Chem., 13,

411 (1970). (10) E. J. Reist, D. E. Gueffroy, and L. Goodman, J. Amer. Chem.

Soc., 86,5658 (1964). (11) A. Bloch, E. Mihich, C. A. Nichol, R. K. Robins, and R. L.

Whistler, Proc. Amer. Ass. Cancer Res., 7, I (1966). (12) B. Urbas and R. L. Whistler, J. Org. Chem., 31,813 (1966).

Synthesis and Antimicrobial Activity of a Carbocyclic Puromycin Analog. 6-Dimethylamino-9- { R- [ 2R -hydroxy-3R -( p-methox yphenyl-r, -alan ylamino ) ]- cyclopentyl) purine?

Susan Daluge and Robert Vince* Department o f Medicinal Chemistry, Received July 23, I971

College of Pharmacy, University of Minnesota, Minneapolis, Minnesota 55455.

An assessment of the requirement for the furanosyl 0 and the CH'OH moiety in the puromycin molecule was undertaken by the synthesis of a novel puromycin analog. A carbocyclic analog, 6-dimethylamino- 9-(R-[ 2R-hydroxy-3R-(p-methoxyphenyl-~-alanylamino)lcyclopentyl )purine (2), was synthesized and evaluated for antimicrobial activity. The carbocyclic analog exhibited antimicrobial activity comparable to puromycin, and also circumvented the nephrotic syndrome associated with puromycin by releasing a nontoxic aminonucleoside upon hydrolysis. The diastereoisomer (19) of 2 was also isolated and found to be devoid of antimicrobial activity.

Puromycin (l), an antibiotic with antitumor activity,' has been found to inhibit protein synthesis in a wide variety of organisms. Its structure has a striking resemblance to that of the aminoacyl-adenyl terminus of aminoacyl-tRNA, and it has been demonstrated that the antibiotic causes premature release of the polypeptide chains from the ribosome.' For this reason, puromycin has been used extensively as a tool in the investigation of protein biosynthesis.

1

A variety of analogs and isomers of puromycin have been prepared to define the structural requirements for inhibition in an attempt to further understand its mode of a c t i o n . j ~ ~

?This work w a s generously supported by Grant AI 08142 from the US. Public Health Service.

However, all of these structures have been of the classical nucleoside type in which an N-substituted amino sugar is attached to a purine or pyrimidine ring through a glycosidic linkage.36 The difficulties encountered in preparing 3- aminoribosyl nucleosides have severely limited the availability of these compounds. Also, the classical nucleoside compounds introduce two undesirable structural features into the puromycin analogs which have not been demonstrated as essential for biological activity; Le., the furanosylo and the 5'-OH group. Thus, it may be possible to modify 1 within the region outlined by the dotted line and still retain the activity of the antibiotic.

Since ribonucleosides are easily cleaved hydrolytically or enzymatically, many nucleosides which may be effective chemotherapeutic agents become ineffective in vivo because they are rapidly destroyed by cleavage into a purine or pyrimidine and a carbohydrate moiety.''* This difficulty could be circumvented by replacing the furanosyl ring with a cyclopentyl system which sterically simulates the sugar moiety and provides a hydrolytically stable C-N bond. The removal of the 5'-OH group from puromycin and its analogs would be desirable from a toxicity standpoint. Toxic mani- festations, including renal lesions, have precluded the use of puromycin in the treatment of human or animal infectious diseases or neoplasm^.^ The nephrotic syndrome results from small amounts of aminonucleoside produced by the hydrolytic removal of the amino acid moiety from administered p u r ~ m y c i n . ~ Recent studies demonstrate that the aminonucleoside is first monodemethylated" and subsequently converted to the 5'-nucleotide." It has been sug


Synthesis and antimicrobial activity of a carbocyclicpuromycin analog. 6-Dimethylamino-9-[R-[2R-hydroxy-3R-(p-

methoxyphenyl-L-alanylamino)]cyclopentyl]purineSusan Daluge, and Robert Vince







arbocycl ic hromyc in Analog Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 171

concns of inhibitor. The in vitro antitumor assays were carried out by our microassay technique which involves the introduction of 0.5 ml aliquots of the medium (RPMI 1630 + 10% calf serum) contg the various concns of the analog into 16 X 125 mm screw cap culture tubes, followed by 0.5 ml portions of medium contg 3 X lo5 L1210 cells. The cultures are incubated at 37" for 40 hr, after which the viable cells are counted by trypan blue exclusion. During this time the cell number in the controls increases approximately eight- to ninefold, with an average viability of 99%.

Miss Ginger Dutschman and Mr. R. J . Maue is gratefully acknowledged. This study was aided by grants CA-12585 and 12422 from the U. S. Public Health Service and T 4 3 6 from the American Cancer Society.

References

Acknowledgment. The excellent technical assistance of

(1) H. Nishimura, K. Katagiri, S. Kozaburo, M. Mayama, and N. Shinaoka, J. Antibiotics, Ser. A , 9, 60-62 (1956).

(2) K. Ohkuma, ibid., Ser. A , 14,343 (1961). ( 3 ) M. Sameyoshi, R. Tokuzen, and F. Fukuoka, G a m , 56,219

(1965). (4) W. L. Wilson, Cancer Chemother. Rep., 52,301 Illus.

(1968). (5) E. C. Taylor and R. W. Hendess, J. Amer. Chem. Soc., 87, 1995

(1965). (6) C. W. Noell and R. K. Robins, J. Heterocycl. Chem., 1,34

(1964). (7) R. L. Tolman, R. K. Robins, and L. B. Townsend, J. Amer.

Chem. Soc., 91,2102 (1969). (8) R. L. Tolman, G. L. Tolman, R. K. Robins, and L. B. Town-

send,J. Heterocycl. Chem., 7,799 (1970). (9) M. Bobek, R. L. Whistler, and A. Bloch, J. Med. Chem., 13,

411 (1970). (10) E. J. Reist, D. E. Gueffroy, and L. Goodman, J. Amer. Chem.

Soc., 86,5658 (1964). (11) A. Bloch, E. Mihich, C. A. Nichol, R. K. Robins, and R. L.

Whistler, Proc. Amer. Ass. Cancer Res., 7, I (1966). (12) B. Urbas and R. L. Whistler, J. Org. Chem., 31,813 (1966).

Synthesis and Antimicrobial Activity of a Carbocyclic Puromycin Analog. 6-Dimethylamino-9- { R- [ 2R -hydroxy-3R -( p-methox yphenyl-r, -alan ylamino ) ]- cyclopentyl) purine?

Susan Daluge and Robert Vince* Department o f Medicinal Chemistry, Received July 23, I971

College of Pharmacy, University of Minnesota, Minneapolis, Minnesota 55455.

An assessment of the requirement for the furanosyl 0 and the CH'OH moiety in the puromycin molecule was undertaken by the synthesis of a novel puromycin analog. A carbocyclic analog, 6-dimethylamino- 9-(R-[ 2R-hydroxy-3R-(p-methoxyphenyl-~-alanylamino)lcyclopentyl )purine (2), was synthesized and evaluated for antimicrobial activity. The carbocyclic analog exhibited antimicrobial activity comparable to puromycin, and also circumvented the nephrotic syndrome associated with puromycin by releasing a nontoxic aminonucleoside upon hydrolysis. The diastereoisomer (19) of 2 was also isolated and found to be devoid of antimicrobial activity.

Puromycin (l), an antibiotic with antitumor activity,' has been found to inhibit protein synthesis in a wide variety of organisms. Its structure has a striking resemblance to that of the aminoacyl-adenyl terminus of aminoacyl-tRNA, and it has been demonstrated that the antibiotic causes premature release of the polypeptide chains from the ribosome.' For this reason, puromycin has been used extensively as a tool in the investigation of protein biosynthesis.

1

A variety of analogs and isomers of puromycin have been prepared to define the structural requirements for inhibition in an attempt to further understand its mode of a c t i o n . j ~ ~

?This work w a s generously supported by Grant AI 08142 from the US. Public Health Service.

However, all of these structures have been of the classical nucleoside type in which an N-substituted amino sugar is attached to a purine or pyrimidine ring through a glycosidic linkage.36 The difficulties encountered in preparing 3- aminoribosyl nucleosides have severely limited the availability of these compounds. Also, the classical nucleoside compounds introduce two undesirable structural features into the puromycin analogs which have not been demonstrated as essential for biological activity; Le., the furanosylo and the 5'-OH group. Thus, it may be possible to modify 1 within the region outlined by the dotted line and still retain the activity of the antibiotic.

Since ribonucleosides are easily cleaved hydrolytically or enzymatically, many nucleosides which may be effective chemotherapeutic agents become ineffective in vivo because they are rapidly destroyed by cleavage into a purine or pyrimidine and a carbohydrate moiety.''* This difficulty could be circumvented by replacing the furanosyl ring with a cyclopentyl system which sterically simulates the sugar moiety and provides a hydrolytically stable C-N bond. The removal of the 5'-OH group from puromycin and its analogs would be desirable from a toxicity standpoint. Toxic mani- festations, including renal lesions, have precluded the use of puromycin in the treatment of human or animal infectious diseases or neoplasm^.^ The nephrotic syndrome results from small amounts of aminonucleoside produced by the hydrolytic removal of the amino acid moiety from administered p u r ~ m y c i n . ~ Recent studies demonstrate that the aminonucleoside is first monodemethylated" and subsequently converted to the 5'-nucleotide." It has been sug

172 Journal of Medicinal Chemistry, I9 72, Vol. 15, No, 2 Daluge and Vince

Scheme I

dioxane EtOH 0 Et,O-H,O

3 4 5

8 9

1 . B,H, THF

12 - Ac,O-p yridine

13, R = H 13a, R = Ac

14,R= H 14a, R = Ac

CbzCl Ba(0H)

DM F l 3 Tq++

I 15 co

R H N f H

OMe

19, R = n 19a, R = Cbz

HN o n W I DMF HN)(o cbz

17

: - > HN d OH

6 I co

R H N f H

OMe

z R = H Z;, R = cbz

0 18

gested that this 5'-nucleotide is responsible for the nephrotic syndrome.

In view of these observations, we decided to synthesize the carbocyclic analog 2 which retains the structural re- auirements that have thus far been demonstrated to be es-

6 7

NMe, NMe, I I

1 . NH,OH'HCl

AcON OAc 0

HON OH

10 11 12

mycin has not been demonstrated. However, it is required for the acceptor activity of aminoacyl adenosines12 which are thought to bind at the ribosomal site in the same manner as puromycin. Thus, the antimicrobial activity of the carbocyclic analog, 2, would provide information relating to the structural requirements of the 4' and 5' oxygens of puromycin for biological activity. In addition, 2 represents a new class of puromycin analogs which should produce a nontoxic aminonucleoside upon hydrolysis of the amino acid moiety.

Chemistry. The carbocyclic analog (2) of puromycin was synthesized by the route shown in Scheme I. (Structures 3- 18 depict only one enantiomer of the racemic form actually obtained.) Cyclopent-2-enone ethylene ketal (3)13 was treated with NBS in Et,O-H,O by a modification of the procedure of Guss and RosenthalI4 in which NaHC03 was used to buffer the reaction mixture. The resulting bromoh drin 4 decomposed on storage and was not characterized.rWhen bromohydrin 4 was immediately treated with NaOH in refluxing benzene, epoxide 5 was formed in 73% yield (from 3), as a stable colorless liquid. The epoxide was opened with NaN, in dioxane by the method of Vander Werf, et al., l5

and the resulting azide reduced catalytically to crystalline trans-3-amino-2-hydroxycyclopentanone ethylene ketal (7) in 76% yield (from 5). Attack by N3- at C-3 of epoxide 5 would be expected.I6 The structure of 7 was confirmed by the nmr spectrum in which H-2 appears as a doublet at T 6.53 (J = 7.9 Hz), and H-3 appears upfield as a multiplet at T 7.3- 6.8.

The purine moiety was formed via a standard method.17 Condensation of 7 with 5-amino-4,6-dichloropyrimidine, followed by ringclosure of the resulting crude pyrimidine 8 with triethyl orthoformate in the presence of EtS03H gave the 6-chloropurine 9 in 84% yield (from 7). The 6-dimethylaminopurine 10 was formed in 75% yield when 9 was treated with refluxing Me2NJ3.

The ketal function of 10 proved unusually resistant to hydrolysis. Attempted hydrolysis by refluxing HC1 at pH 3 , refluxing aq NH4Cl, or transketalization in acetone with pTsOH gave only recovered 10. This difficulty has been encountered in the hydrolysis of other ketals and acetals of similar purine derivatives.§ The proximity of a protonated

$The corresponding bromohydrin of cyclohex-2-enone ethylene- ketal prepd by this method was a stable, cryst solid, mp 96-98', which was characterized; unpublished results.

skntial for activity., The requirement for the 2'-OH in puro- 8 Unpublished results.

Carbocyclic Puromycin Analog

amine has been found to hinder acetal hydrolysis.'* The neighboring OH would not be expected to affect the rate of hydrolysis of the ketal significantly." It thus seems likely that the protonated purine moiety accounts for the difficulty of hydrolysis of compounds such as 10.

neutralization. Similarly, treatment with Amberlite IR-120 resin at 60" for 2 hr led only to decomposition. As it a p peared that conditions necessary for hydrolysis of 10 effected decomposition of the product(s), oxime 11 was synthesized directly from 10 under what appear to be unique conditions for oxime formation. An aq soln of 10 and a 5-fold excess of HONH2 -HC1 adjusted to pH 1 with HCl was warmed (75") for 3 hr. On neutralization of the colorless reaction solution, pure 11 precipitated immediately in almost quantitative yield (solid forms as pH 3.5 is reached). However, when oxime formation was attempted by warm- ing a solution of 10 at pH 1, and then adding HONH2.HCl along with sufficient NaOH to give a pH of 6, partial recovery of 10 (69%) and considerable darkening of the reaction solution resulted. Even in the presence of HONH2*HCl, dark tar and a lowered yield of oxime (43%) were obtained if a pH slightly lower than 1 was used. Identical conditions at a pH of 3 gave only recovered 10. These results suggest that hydrolysis of the ketal of 10 requires a pH lower than 3, but that below pH 1, 10 or its hydrolysis products is un- stable and decomposes rapidly before hydroxylamine may attack. In a rather narrow pH range, HONH2 appears to prevent decomposition, probably by direct attack on the resonance-stabilized carbonium ion formed on ring .opening of the protonated ketal.

oximes. Since attempted separation by chromatography or recrystallization resulted in considerable losses, the mixture was acetylated without separation, giving a 70% yield of a chromatographically homogeneous diacetyl derivative (12). The nmr spectra of 11 and 12 confirm the structures shown. In particular, the appearance of H-2 as a doublet (J2,3 = 9.0 Hz for 11 and 9.5 Hz for 12) considerably downfield from the H-3 multiplet# indicates that the oximation conditions did not result in isomerization via an enediol.

The 0-acetyl oxime 12 was reduced to a mixture of amino alcohols with diborane in THF by a modification of the procedure used by Feuer and Braunstein for the conversion of cyclohexanone 0-acetyloxime to cyclohexylamine.20 This mixture was acetylated in Ac20 and the resulting acetamides, 13 (45%) and 14 (4%), were separated by chromatography and characterized. The AcNH and OH groups were shown to be cis in 13 and trans in 14 by acyl migration studies (see Experimental Section). The diacetyl derivatives 13a (39%) and 14a (3%) were isolated when the mixture of amino alcohols from reduction of 12 was treated with Ac20-pyridine, and 13a was converted to 13 on treatment with NH3 in MeOH. Hydrolysis of 13 with Ba(OH)2 gave a carbocyclic analog of the puromycin aminonucleoside, characterized as its AcOH salt 16. The cis stereochemistry of the H2N and OH groups was further confirmed by facile formation of the cyclic carbamate 18 on treatment of the carbobenzoxy derivative 17 with NaOMe in DMF.

The diborane reduction of 10 appears to be the first reported example of the reduction of an oxime to an amine in the presence of a purine ring. Predominant attack of

When 10 was refluxed in HC1 at pH 1, black tar formed on

Tlc of 11 indicated it t o be a mixture of syn and anti

Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 113

~~~ ~~~

#In the nmr (DMSO-d,) of trans-2-acetoxy-3-(6-dimethylamino-9- puriny1)cyclohexanone 0-acetyloxime, the cyclohexyl analog of 12, H-2 also appears as a doublet (7, 3.77, .I?,B = 10.5 Hz) downfield from the H-3 multiplet at 7 5.1; unpublished results.

hydride from the purine side of the cyclopentyl ring, if general, could be useful in the synthesis of amino sugar nucleosides. It is hoped that further studies now in progress on the diborane reduction of analogs of 10 will provide information on the mechanism of this highly stereospecific reduction.**

The carbocyclic aminonucleoside analog 16 was coupled to N-benzyloxycarbonyl-pmethoxyphenyl-~-alanine~ by 2 methods: A, the dicyclohexylcarbodiimide-N-hydroxysuccinimide method21'22 and B, a modification of the mixed anhydride method suggested by Anderson, et a123 The resulting carbobenzoxy blocked diastereomers 2a and 19a (97% by method A, 77% by method B) could not be separated. The mixture of amino alcohols from the reduction of 12 was also coupled to N-benzyloxycarbonyl-pmethoxyphenyl- L-alanine by method A, giving a yield of 2a and 19a, after chromatography, comparable with the overall yield via 13 and 16. Following hydrogenolysis of the Cbz group, separation of diastereomers 2 and 19 by chromatography was possible. Structure 2 is assigned to the diastereomer having [aI2'D of -83" and structure 19 to the diastereomer having [(uI2'D of -8" (see Results and Discussion). The two coupling methods resulted in samples of 2 and 19 with identical optical purities.

slightly only in the relative intensities of some ions. The molecular ion (m/e 439) is relatively small, and there is a minor (M - 18) peak due to loss of H20 involving the OH group. As with puromycin,24 the fragmentation is dominated by the aminoacyl moiety. Fission of the bond a to the C=O

The mass spectra of 2 and 19 are almost identical, differing

B B I

HNH OH HN OH I . I c=o+ e= +

I/ m/e 289 (M - 150)

m/e 439 ( M I )

B B B

I H i OH HN OH HN

I I I

I I CH-NH, CH CH

?=O c=o c=o .+

u

m/e 439 (M,) m/e 318 (M - 121) m/e 300 (M - 121 - 18)

in the molecular ion MI accounts for the m/e 289 (M - 150) peak. Cleavage of the benzylic bond followed by loss of H20 accounts for the base peak at m/e 300 - 121 - 18) and prominent peaks at m/e 318 (M - 121) and m/e 121. There is a metastable peak for the transition 3 18' + 300+ t 18. The (M - 12 1 - 18) ion is somewhat less abundant for puromycin. This would be expected as the puromycin (M - 121 __ _______--

**Further details concerning this reaction will be published.

114 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2

,g _ - ;-:-&- - t - -: - - - - -& - 0 24dnU

Daluge and Vince

mle 245

c

mle 228

mle 271

- 18) ion has several favorable routes for further fragmentation which these carbocyclic analogs lack, e.g., loss of the elements of CHzO involving the 5’-OH. Fission of the m/e 300 ion on either side of the amide carbonyl accompanied by proton transfers accounts for the m/e 245 and 271 peaks. Fission of the bond between the aminoacyl group and the cyclopentyl ring accounts for the m/e 228 ion. There is a metastable peak for the transition 300’ + 228’ t 72. As with puromycin, the m/e 300 peak further fragments to give a prominent m/e 164 (B t 2H) peak characteristic of the dimethylaminopurine moiety, as indicated by an appropriate metastable peak. Other fragmentations directed by the purine moiety account for ions of m/e 206 (B t 44), 190 (B t 28), 163 (B t H), 162 (B), and 134 (B t H - CH3N). These ions, or corresponding ones, are prominent in the spectra of adenosines” and are also abundant in the spectrum of puromycin. It is consistent with the structure proposed for the (B t 30) ion in adenosines, which incorporates the ribose ether that replacement of 0 by CH2 results in a shift of 2 mass units to (B t 28).


Antimicrobial testing of diastereomers 2 and 19 revealed that 1 isomer was completely inactive while the other exhibited growth inhibition on the same order of magnitude as puromycin. The absolute stereochemistry of the active isomer has tentatively been assigned that of structure 2 on the basis of its biological activity and in accordance with the stereochemistry of puromycin. The minimum inhibitory concns by a 2-fold serial dilution test in broth for puromycin and the carbocyclic analog 2, respectively, are as follows (mM): Staphylococcus aureus (NRRL B-3 13), 0.244 and 0.244; Bacillus subtilis (NRFX B-549, 0.030 and 0.060; Klebsiella pneumoniae (NRRL B-l17), 0.485 and 0.485; Escherichia coli (NRRL B-2 10) 0.060 and 0.120. A growth curve for S. aureus in the presence of different concns of puromycin or the carbocyclic compound is illustrated in Figure 1. A lag period is observed with both compounds when the concns are lower than those required for complete inhibition. Such a lag period is consistent with the mechanism of action of puromycin2 since the antibiotic would be expected to be consumed as it is incorporated into the grow- ing peptide chains.

The aminonucleoside 16 was tested for nephrotoxicity in rats at a dose of 33 mg/kg under the same conditions that are required for puromycin aminonucleoside to cause severe nephrotic syndrome at 15 mg/kg.26 No nephrotoxicity was observed even after 17 days of treatment with 16.

The novel puromycin analog 2 provides a molecule with the structural features required for puromycin-like antimicrobial activity. Thus, the ribofuranosyl ring can be replaced with the more hydrolytically stable cyclopentane ring without loss of activity. In addition, the removal of the CHzOH moiety is not detrimental to activity and at the same time provides 2 with a resistance to kinase activity upon re-

0 0 6 l n M

~ - 0- r

/ 0 4

lease of the aminonucleoside. This resistance to phosphoryla- tion circumvents the nephrotic syndrome associated with puromycin aminonucleoside. This carbocyclic analog and others which are under preparation are being evaluated for in vitro inhibition of protein biosynthesis in an attempt to explore the requirements for binding to ribosomes. Pre- liminary studies with 2 and 19 on an E. coli ribosomal system are consistent with the antimicrobial activities. The details of these experiments will be the subject of a future paper.

Experimental Section?? 6-Oxabicyclo[ 3.1.0]hexan-2-one Ethylene Ketal (5 ) . The

method of prepn of 4 is a modification of that of Guss and Rosen- thal.’, Cyclopent-2-enone ethylene ketal (3)’, (37.85 g, 0.300 mole), NBS (53.40 g, 0.300 mole), NaHCO, (4.20 g, 50.0 mmoles), Et,O (240 ml), and H,O (240 ml) were stirred vigorously for 6.5 hr, at which time all of the solid had disappeared and the pH was approx 7. The aq layer was satd with NaCl, and the Et,O layer then sepd. The aq layer was extd with addnl Et,O (2 X 100 ml). The combined Et,O layers were washed with satd NaCl and dried (CaSO,). Evapn left crude bromohydrin 4 as a yellow oil (70 g). In another run, this oil was partially solidified from petr ether (bp 30-60°), giving gummy white crystals: mp 42-47”; ir (Nujol) 3450 (OH), 1200-1040 cm-’ (CO). This solid could not be recrystd and darkened on standing.* Best overall yields of epoxide were obtd by using the crude oil immediately. Crude 4 was dissolved in PhH (600 ml) and refluxed with powd NaOH (36 g) for 1.0 hr. The mixt was filtered, and the black solid washed with addnl PhH (200 ml). Evapn of the combined PhH soln and wash left a pale yellow liq which was distd, giving a forerun whiqh ir showed to be a mixt of 3 and unketalized material (1.36 g),

_- -~ ??Melting points were detd on a Mel-Temp apparatus and are

corrected. Optical rotations were measured at ambient temps with a Perkin-Elmer 141 automatic polarimeter; nmr, with a Varian A-60D spectrometer; ir, with a Perkin-Elmer 237B spectrophotometer; uv, with a Cary 14 recording spectrophotometer; low-resolution, 50-eV mass spectra, with a Hitachi Perkin-Elmer RMU-6D mass spectrometer (ion source temperature 250°, accelerating potential ISOOV), equipped with a direct inlet probe. TIC was run on silica gel (East- man chromagram sheets with fluorescent indicator) in these solvent systems: A, 2% MeOH-CHCI,; B, 5 % MeOH-CHCl,; C, 10% MeOH- CHCI,; D, 15% MeOH-CHCI,; E, 20% MeOH-CHCI,. Prep tlc was done on 20 X Z O cm glass plates coated with 2 mm of silica gel F 254 (E. Merck, Darmstadt) and column chromatog on silica gel (Baker, AR, 60-200 mesh). Evapns were carried out in vacuo with a bath temp of less than 45’ unless otherwise noted. Solid samples were dried in vacuo (<1 mm) at 56’ before analysis. Analytical results are within f0.4% of the calcd values. Celite is a diatomaceous earth product of Johns-Manville.

Carbocyclic Puromycin Analog Joumal ofhfedicinal Chemistry, 1972, Vol. 15, No. 2 175

bp 28-86' (7 mm), followed by 5 (31.28 g, 73%), bp 86.5-88" (7 mm); ir identical with that of the analytical sample. Redistn of such a sample gave the analytical sample of 5 as a colorless liq, bp 98-99' (10 mm), nZoD 1.4698; ir and nmr were as expected. Anal. (C,Hi,O,) C, H.

Samples of 5 (distd once) were stored for months at 5' without deterioration.

trans- 3-Aminp-2-hydroxycyclopentanone Ethylene Ketal (7). The method of prepn of 6 is that of Vander Werf, et al. 5 (2.84 g, 20.0 mmoles) in dioxane (40 ml) was brought to reflux. A soln of NaN, (1.63 g, 25.0 mmoles) in H,O (10 ml) was added to the refluxing soln over 1.0 hr. The resulting mixt was refluxed with vigorous stirring for 48 hr. The dioxane layer was sepd from the aq layer, and the aq layer was extd with addnl dioxane (50 ml). The combined dioxane layers were washed with satd NaCl(50 ml) and dried (CaSO,). Evapn (25', 0.5 mm) left 6 as a pale yellow gummy solid (3.46 g); ir (Nujol) 3300 (OH), 2093 cm-' (N,). This material was dissolved in abs EtOH (75 ml) and shaken with PtO, (100 mg) under H, (50 psi) for 20 hr. After filtration, the EtOH was evapd, leaving pale yellow solid (2.91 g) which ir showed to contain no azide. Crystn from F%H gave 7 as pale yellow crystals (2.41 g, 76%); mp 97-98'; ir identical with that of the analytical sample. Sublima- tion of such a sample at 90-100" (0.20 mm) gave the analytical sample of 7 as white needles: mp 99-100'; ir as expected; nmr (CDCl,) 7 8.9-7.8 (m, 4, 2 CH,), 7.68 (s, 3, NH, and OH), 7.3-6.8 (m, 1, H-3), 6.53 (d, l , J , , = 7.9 Hz, H-2), 6.1-5.9 (m, 4, ethylene ketal). The singlet at 7 7.68 disappeared on addn of D,O. Anal.

mole of 5 gave a 54% yield of 7 (after crystn from PhH).

Ketal(9). A soln of 7 (14.04 g, 88.20 mmoles), 5-amine4,6-dichloropyrimidine (14.47 g, 88.20 mmoles), and Et,N (37 ml, 265 mmoles) in n-BuOH (160 ml) was refluxed under N, for 44 hr. Evapn (SO", 0.2 mm) left a brown oil (42 g) contg 8 which was shaken with triethyl orthoformate (200 ml) for a few min. The white solid EtgH'Cl- (10.82 g, 89%) did not dissolve and was removed by filtration. EtS0,H (1.0 g) was added to the filtrate, and it was stirred at ambient temp for 15 hr. At this time solid had formed. Hexane (200 ml) was added, and the mixt was cooled. The tan solid was collected and washed with hexane (100 ml), giving crude 9 (24.42 g). Crystn from EtOAc (500 ml) gave tan needles (22.03 g, 84% from 7), mp 165-167'; ir identical with that of the analytical sample. Such a sample was recrystd twice from EtOAc, giving the analytical sample of 9 as white needles, mp 164.5-166.5"; Rf 0.50 in solvent B; t t ir, uv, and nmr as expected. Anal. (C,,H,,N,O,Cl) C, H, N.

trans- 3-(6-Dimethylamino-9-purinyl)-2-hydroxycyclopentanone Ethylene Ketal (10). A soln of 9 (3.85 g, 13.0 mmoles) in aq 25% Me,NH (100 ml) was refluxed 3.0 hr. The soln was concd to 20 ml and extd with EtOAc (3 X 100 ml). The EtOAc ext was dried (CaSO,) and concd to 50 ml. White crystals of 10 were collected (2.99 g, 75%), mp 156-158'; ir identical with that of the analytical sample. Recrystn of such a sample from EtOAc gave the analytical sample of 10 as white crystals, mo 156-158" : Rf 0.54 in solvent B:

A soln of

(C,H,,NO,) C, H, N. When run on a larger scale, the yield of 7 was lower, e.g., 0.18

trans- 3-(6-Chlor~9-purinyl)-2-hydroxycyclopentanone Ethylene

uv, and ir as expected; nmr (DMSb-d,) 7 6.54 (i, 6, N(CH,),). Anal. (C.,H,,N.OJ C. H. N. .- 1.

trans-~-~6-Dimethylam~o-9-purinyi)-2-hy~oxy~clopentanone Oxime (11). To a stirred mixt of 10 (10.00 g, 32.8 mmoles), NH,OH.HC1(11.38 g, 164 mequiv), and H,O (140 ml) was added 2 N HCl (approx 28 ml) dropwise until the pH was 1.0 (measured on a meter), and all solid had dissolved. The soln was stirred at 70-75' for 3.0 hr. The pH of the hot soln was then adjusted to 6.5 with 6 N NaOH, and white solid began to ppt. The mixt was cooled (5") for several hours. The solid was collected, washed with H,O (50 ml), and air-dried, giving 11 (8.95 g, 99%), mp 202-204" dec; tlc indicates a mixt of syn and anti oximes, Rf 0.47 and 0.39 in solvent C; ir differs slightly in the fingerprint region from the analytical sample; nmr identical. Such samples of 11 were sufficiently pure for use. The analytical sample of 11 was prepd by crystn from EtOAc- hexane, giving white needles, mp 208-209' dec; tlc still shows 2 spots, with somewhat different relative intensities; ir as expected; nmr (DMSO-d,) 7 8.0-7.3 (m, overlaps DMSO-d,, 2CH,), 6.52 (s, 6, N(CH,),), 5.7-5.1 (m, 1, H-3), 5.1-4.8 (m, 1, H-2), 4.3 (m, 1, OH), 1.77 (s, 2, purine H-2' and H-8'), -0.76 (s, 1, C=NOH). The multiplet a t 7 4.3 and the singlet at 7 -0.76 disappeared, and the multiplet at 7 5.1-4.8 sharpened to a doublet at 7 4.98, J % , = 9.0 Hz, on addn of D,O. Anal. (C,,H,,N,O,) C, H, N.

trans-2-Acetoxy-3-(6dimethylamino-9-purinyl)cyclopentanone 0-Acetyloxime (12). A soln of 11 (8.95 g, 32.4 mmoles) in Ac 0 (200 ml) was stirred at 60-65' for 4.25 hr. Evapn left a tan solia

which was cryst from EtOAc, giving 12 as white crystals (8.16 g, 70%), mp 152-154';Rf 0.45 in solvent A; ir differs slightly in the fingerprint region from that of the analytical sample. Recrystn of such a sample 3 times from FtOAc gave 12 as white prisms, mp 162.5-164'; Rf 0.45 in solvent A; ir as expected; nmr (DMSO-d,) 7 8.01 (s, 3, COCOCH,), 7.82 (s, 3, NOCOCH,), 7.7-6.8 (m, overlaps DMSO-d,, 2CH,), 6.53 (s, 6, N(CH,),), 5.2-4.4 (m, 1, H-3), 3.50 (d, l,J, , = 9.5 Hz, H-2), 1.74 and 1.66 (both s, 2, purine H-2' and H-81. Anal. (C,,H,$,O,) C, H, N.

Diborane Reduction of 12; Separation of (+-)-9-[@(3a-Acet- amido-2a-hydroxy)cyclopentyl].6.dimethylaminopur~e (1 3) qnd (*)-9- [ a-(3a-Acetamido-2@hydroxy) cy clopenty I] 4-dimethy lamino- purine (14). To a stirred, cooled (0-5') soln of 12 (4.37 g, 12.1 mmoles) in dry THF (100 ml) was added a 1 M soln of BH, in THF (44,O ml, approx 130 mequiv of hydride) over a period of 1.0 hr under N,. The soln was stirred for an addnl 3.0 hr at 0-So, and then for 12.0 hr at ambient temp. After cooling the reaction soln (ice bath), H,O (8.3 ml) was added. The THF was evapd (25', 0.5 mm), and the residue was stirred with 2 N HCl(110 ml) at ambient temp for 3.5 hr. After evapn, the residue was dissolved in MeOH (50 ml) and passed slowly through a column of 100 ml of Amberlite IRA- 400 resin (OH-) packed in MeOH. The basic eluent (500 ml) was evapd, leaving a yellow, gummy solid (3.26 g), which tlc in solvent E indicated to be largely a mixt of amino alcohols (major spot a t Rf 0.30, smaller spot at Rf 0.33) and numerous minor contaminants having greater Rf values; ir (KBr) 3250 broad (OH, NH,), 1600, 1555 cm-' (purine). The amino alcohols could not be separated by prep tlc and appeared to carbonate on standing. This gummy solid was dissolved in Ac,O and stirred at 60' for 1.75 hr. Evapn left a brown glass (3.74 g) which was chromatogd on a column of silica gel (200 g) packed in CHCI,. Elution with 2% MeOH-CHCIt (2 1.) gave a pale yellow glass (290 mg), Rf 0.46 in solvent B, which crystd from EtOAc-hexane to a white solid (64 mg), mp 181-183".$$ Elution with 3% MeOHCHC1,(3 1.) gave 13 as a pale yellow glass (2.09 g), Rf 0.35 in solvent B, which crystd from EtOAc (1.65 g, 45% from 12), mp 15 1-152'; ir identical with that of the analytical sample. Recrystn of such s sample from EtOAc gave the analytical sample of 13 as white crystals, mp 151.5-152.5' ; uv as expected; ir (KBr) 3300-3050 (OH, NH), 1660,1640 (amide C=O), 1610, 1555, 1535 (purine, amide); ir (CHC1,) 1660 (amide C=O), 1600 (purine), 1562 cm-' (amide); nmr (DMSOd,) 7 8.5-7.6 (m) overlapping 8.08 (s) and DMSO-d, (7.3,2CH2 and CH,CO), 6.49 (s, 6, N(CH,),), 6.0-4.9 (m, 3, H-1', H-2', and H-3'),4.63 (d, 1 , J = 4.8 Hz, OH), 2.5-2.1 (m, 1, NHC=O), 1.73 and 1.75 (both s, 2, purine H-2 and H-8). The doublet at 7 4.63 and the multiplet at 7 2.5-2.1 disappeared within 5 min of addn of D,O. Anal. (C,,~,N,O,) C, H, N.

Contd elution with 3% MeOH-CHCl, (1.5 1.) gave fractions contg both 13 and 14 (124 mg, 3%), followed by fractions contg only 14 as an amorphous white solid (247 mg), Rf 0.28 in solvent B. Crystn of this solid from EtOAc gave 14 as white crystals (159 mg, 4%), mp 171-171.9 ;uv as expected;ir (KBr) 3250,3200, 3100 (OH, NH), 1638 (amide C=O), 1600,1550 (purine, amide); ir (CHCI,): 1660, 1600, 1562 cm-'; nmr (DMSO-d,) 7 8.5-7.6 (m) overlapping 8.08 (s) and DMSO-d, (7.5, 2CH, and CH,C=O), 6.49 (s, 6, N(CH,),), 6.4-5.0 (m, 3, H-1', H-2', and H-3'), 4.57 (d, 1, J = 4.0 Hz, OH), 1.82 and 1.72 (both s) overlapped by 1.8-1.7 (m, 3, purine H-2 and H-8 and NHC=O). The doublet at 7 4.57 disappeared immediately on addn of D,O. Complete exchange of the amide NH required several hours, in contrast to the rapid exchange noted for 13. Anal. ( C ~ 4 ~ o N 6 0 z ) C, H, N.

The diacetyl derivatives 13a and 14a were obtd when the mixt from the BH, redn of 1.00 g (2.78 mmoles) of 12 was stirred in Ac.0 (5 ml)-Dyridine (10 ml) at ambient temo for 18 hr. Evaun left

$$Spectra of this solid (uv, ir, nmr), a mass spectrum molecular ion of 332, and elemental analysis (C, H, N) suggeSt a molecular formula of C,,%4N,q and the structure a. An 0 --t N acyl migration during the BH, redn, followed by redn of the N-acyl moiety and subsequent acetylation of treatment with Ac,O could account for this minor product.

176 Journal ofMedicina1 Chemistry, 1972, Vol. 15, NO. 2 Daluge and Vince

a brown glass (800 mg) which was chromatogd on a column of silica gel (50 g) packed in CHC1,. Elution with 1% MeOH-CHC1, (700 ml) gave 14a as a colorless glass (154 mg), Rf 0.52 in solvent B. Crystn from EtOAc-hexane gave white solid (32 mg, 3%), mp 207.5- 210" dec; ir as expected. Anal. ~C,,H,,N,O,) C, H, N.

Elution with 2% MeOH-CHCl, (850 ml) gave a yellow glass which crystd from EtOAc giving 13a as tan crystals (377 mg, 39%), mp 194.5-196", Rf 0.41 in solvent B. Recrystn of such a sample from EtOAc gave the analytical sample of 13a as white needles, mp 194.5-196.5' ; ir and nmr as expected. Anal. (C,,HZ2N,O,) C, H, N.

In the chromatog of 13 and 14, as well as 13a and 14a, dark materials remained on the silica gel columns after all of the acetylated compds had been eluted.

The diacetyl derivative 13a was converted to 13 in 85% yield on treatment with a satd s o h of NH, in MeOH, confirming the assignment of structure of 13a and 14a

Acyl Migration Studies. Condns were chosen which have been shown to give 63% migration with cis-2-benzamidocyclopentanol and no migration with trans-2-benzamidocyclopentanol.2'

A. With 13. TheN-Ac derivative 13 (50 mg, 0.164 mmole) was dissolved in 0.444 N ethanolic HC1 (7.43 ml, 3.30 mequiv) and stirred at 25" in a stoppered flask for 18 hr. Evapn left a white solid, mp 100-110" with frothing; ir (KBr) 1737 cm-' (acetate C=O indicates presence of 15).

B. With 14. Treatment of 14 with ethanolic HC1 as in part A gave, on evapn of EtOH, a white solid, mp 95-100" with frothing; ir (KBr) shows no acetate C=O band.

purine Acetate (16). The N-Ac deriv 13 (730 mg, 2.40 mmoles) was refluxed in a satd aq soln of Ba(OH), (approx 0.5 N, 40 ml) for 3.0 hr. The soln was dild with EtOH (50 ml) and treated with excess Dry Ice. The BaCO, was removed by filtration through Celite. Evapn left a white solid (740 mg), mp 180-182" dec, which was re- solidified from EtOH-Et,O, giving 16 as a white powder (685 mg, 89%), mp 184-186" dec; uv and ir as expected; nmr (DMSO-d,) 7 8.18 (s) overlapped by 8.8-7.7 (m, 7, CH,CO,-and 2CH,), 6.5 1 (s, 6, N(CH,),), 5.8-4.2 (m, 3, H- l f , H-2', and H-37, 4.5-4.2 (m, 4, NH: and OH), 1.77 (s, 2, purine H-2 and H-8). The multiplet at 7 4.5-4.2 disappeared on addn of D,O. Anal. (C,,H,,N,O .CH,CO,H) C, H, N.

(+)-9- [p-( 3cl-Amino-2cl-hydroxy)cyclopentyl]-6dimethylamino- purine 2', 3'-Carbamate (18). The procedure is that used by Baker and Joseph" to prepare the 2',3'-carbamate of the puromycin aminonucleoside. A s o h of 16 (100 mg, 0.310 mmole) and Et,N (0.13 ml, 0.93 mmole) in DMF (5 ml) was cooled to 5 " . Carbobenzoxy chloride (0.05 ml, approx 0.5 mmole) was added, and the s o h was stirred at ambient temp for 1.0 hr. H,O (20 ml) was added, and the oil which formed was extd into CHC1, (3 X 20 ml). The CHCI, exts were dried (CaSO,) and evapd, leaving a colorless glass (96 mg), which tlc in solvent B showed to be a mixt of 17 (Rf 0.50) and 18 (Rf 0.28) in approx equal amts. When the mixt of 17 and 18 was treated with NaOMe in DMF as described below, 18 was isolated in 74% yield (from 16). In another run 17 and 18 were separated on a silica gel column eluted with 3% MeOH-CHCl,. The Cbz derivative 17 was eluted as a glass which crystd from abs EtOH, giving white crystals, mp 163-164.5'; ir and nmr as expected. Anal. (C,,H,,N,O,) C, H, N.

Contd elution of the column gave 18 as a white solid, mp 223- 224" ; ir identical with that of the analytical sample prepd as described below.

A sample of 17 purified by chromatog (890 mg, 2.25 mmoles) was dissolved in DMF (10 ml), and 0.30 ml(0.4 mmole) of an approx 1.4 N s o h of NaOMe in MeOH was added. The soln was stirred at 100" for 2.0 hr. At this time tlc showed complete conversion to 18. Evapn left 18 as a glass which solidified from abs EtOH (10 ml), giving a white powder (452 mg, 70%), mp 218-220'. Crystn from abs EtOH gave white crystals, mp 226-227" ; ir (KBr) 3275 broad (NH), 1757 and 1733 (carbamate C=O), ir (CHCI,) single carbamate band at 1760; nmr as expected. Anal. (C,,H,,N,O,) C, H, N.

9- {R- [ 3R-(Benzyloxycarbonyl-p-methoxyphenyl-L-alanylamino)- 2R -hydroxy]cyclopentyl}-6dimethylaminopurine (2a) and 9- {S- [ 3S-(Benzyloxycarbonyl-p-methoxyphenyl-L-alanylamho)- 2S-hydroxy]cyclopentyl}-6dimethylaminopurine (19a). Method A.2"zz .4 s o h of 16 (145 mg, 0.450 mmole) in MeOH was passed slowly rhrough a column of 10 ml of Amberlite IRA-400 resin (OH-) packed in MeOH. Evapn left the free amine as a white solid (118 mg, 0.450 mmole), mp 1.53-154", which was dissolved immediately in DMF (5 ml), along with N-benzyloxycarbonyl-p-methoxyphenyl-i-ala- nine4 (155 mg, 0.472 mmole) and N-hydroxysuccinimide (54.3 mg, 0.472 mmole). The soln was cooled to 0", and DCC (97.4 mg, 0.472 mmole) was added. After stirring at 0" for 30 min. the soln

(+-)-9- [p-( 3a-Amino-2ol-hydroxy)cyclopentyl]-6dimethylamho-

was allowed to stir at ambient temp for 20 hr. The mixt was filtered, the dicyclohexylurea was washed with EtOAc (20 ml), and the combined filtrate was evapd. A soln of the residue in EtOAc (5 ml) was cooled a t 0" and then filtered. The filtrate was dild to 15 ml with EtOAc, extd with H,O (2.5 ml), then one-half satd NaHCO, (2.5 ml), then H,O (2 X5 ml). Evapn of the dried EtOAc soln left a mixt of 2a and 19a as a white solid foam (250 mg, 0.436 mmole, 97%), R f 0.52 in solvent B. The diastereomers could not be sepd by crystn or chromatog. An analytical sample of the mixt was prepd by solidification from MeOH, giving white powder; mp: part at 170-172" and the rest at 200-205" ; ir and nmr as expected. The purine and methoxy nmr resonances indicate a mixt. Anal. (C,,H,,N,O,) C, H, N.

THF (10 ml) was cooled to -10". Ethyl chlorocarbonate (72.4 mg, 0.667 mmole) was added. The soln was stirred for 1 min, and then N-benzyloxycarbonyl-p-methoxyphenyl-L-alanine, (200 mg, 0.606 mmole) was added and the mixt stirred at - 10" for 10 min. A sample of the free amine obtd by resin treatment of 16 as in method A (159 mg, 0.606 mmole) was dissolved in DMF, and the soln was cooled t o -10" and added to the mixed anhydride. The resulting mixt was stirred at -10" for 1.0 hr and then stored at 4" for an addnl 24 hr. It was filtered, and the solid collected was washed with addnl DMF (5 ml). Evapn of the combined filtrates left a colorless glass which was triturated with H,O (5 ml), evapd to dryness, and chromatogd by prep tlc on a plate developed in solvent C. Extn of the major band with 20% MeOH-CHCI, (other bands were unreacted starting materials) gave a mixt of 2a and 19a as a solid foam (266 mg, 77%), Rf, ir, and nmr identical with those of the mixt obtained from method A.

redn of 3.00 g (8.32 mmoles) of 12 was treated directly with the mixed anhydride of N-benzyloxycarbonyl-p-methoxypheny1-L- alanine in this manner, giving a mixt of 2a and 19a, after chromatog, in 30% yield (from 12);Rf, ir, nmr identical with those of the samples described above.

alanylamino)] cyclopentylburine (2) and 6-Dimethylamino-9- {S- [ 2S-hydroxy-3S-~-methoxyphenyl-L~alanylam~o)]cyclopentyl 1- purine (19). A mixt of 2a and 19a prepd by method A (250 mg, 0.436 mmole) was dissolved in glacial AcOH (15 ml) and shaken with 10% Pd/C (125 mg) under H, (1 atm) for 10 min, by which time H, uptake had ceased. The mixt was filtered through Celite and the Celite was washed with addnl AcOH (10 ml). Evapn of the combined filtrate and wash (35", 0.5 mm) left a colorless glass. A s o h of this glass in MeOH was passed slowly through a column of 10 ml of Amberlite IRA400 resin (OH-) packed in MeOH. The basic MeOH eluent (250 ml) was evapd, leaving a mixt of 2 and 19 as a white solid (176 mg, 92%), mp 194-196"; tlc gave 2 spots with Rf 0.41 and 0.51 in solvent C; tlc and ir identical with those of the analytical sample. Recrystn from MeOH (5 ml) gave a 1 : 1 mixt of 2 and 19 as white crystals, mp 195-196'; [a]589 -45.6", [a]436 -103.8" ( c 0.26, CHCI,); uv max (0.1 N HCl) 270 mp (log E 4.307); (H,O) 277 (4.314); (0.1 N NaOH) 277 (4.312); ir (KBr) 3420, 3310 (OH, NH,, NH), 1670 and 1655 (2 amide C=O), 1600 cm-' (arom, NH,); ir (CHCI,) 1660 (2 amide, C=O): 1605 (arom, NH,); nmr identical with that of pure 19. Anal. (C,,H,,N,O,) C, H, N. Further recrystn of such a mixt did not change the ratio of 2 to 19.

The diastereomers 2 and 19 were sepd by prep tlc (50-70 mg per plate) in solvent D. The 2 bands were each stirred for 18 hr with 20% MeOH-CHCI,, filtered, and evapd, giving almost quant recovery of the pure diastereomers as colorless glasses. The glass having Rf 0.5 1, assigned structure 19, crystd from abs EtOH, giving white needles, mp 161-161.5"; -&lo, [a]436 -16.2" (c 0.43, CHCI,); uv max (0.1 N HC1) 270 mp (log E 4.301); (H,O) 277 (4.323); (0.1 N NaOH) 277 (4.322);ir (KBr) 3417, 3309,3300-3100 (OH, N€&, NH), 1655 (amide C=O), 1600 cm-' (aromatic, NH,); ir (CHCI,) 1650, 1600;nmr (DMSO-d,) 7 8.7-8.2 (m, 2 , NH,), 8.2-7.0 (m, 6, 3 CH,), 6.52 (s, 7, N(CH,), and NCHC=O), 6.25 (s, 3, OCH,), 6.0-5.0 (m, 3, H-l', H-2', and H-3'), 4.7-4.2 (m, 1, OH), 2.95 (q, 4, OC,H4), 2.3- 1.9 (m, 1, NHC=O), 1.73 (s, 2, purine H-2 and H-8). The multiplets at 7 8.7-8.2, 4.7-4.2, and 2.3-1.9 disappeared on addn of D,O; mass spectrum (probe temp ca. 250") m/e above 80 (relative intensity) 439 (0.6), 422 (0.8), 421 (1.2), 420 (0.6),419 ( O S ) , 418 ( O S ) , 405 (0.7), 404 (1,8), 403 (0.9), 318 (23.6), 301 (21.4), 300 (loo), 271 (8.6), 228 (11,9), 190 (12.5), 165 (8.1), 164 (82.4), 163 (16.5), 150 (14.8), 148 (9.4), 134 (23.3), 122 (10.3), 121 (59.6), 120 (5.6), 109 (5.7), 82 (9.8); metastable transitions: 283.0 (283.0 calcd for 318t + 300+), 173.5 (173.3 calcd for 300++ 228+), l lO(110.2 calcd for 163+ + 134'), 89.7 (89.7 calcd for 300'- 164'). Anal. (C,,H,,N,O,) C, H, N.

Method B." A s o h of Et,N (61.3 mg, 0.606 mmole) in dry

The crude mixt of amino alcohols resulting from the diborane

6-Dimethylamino-9-{R- [2R-hydroxy-3R-(p-methoxyphenyl-~-

The glass having Rf 0.41, assigned structure 2, could not be

Metabolism of Puromycin Aminonucleoside JournalofMedicinal Chemistry, 1972, Vol. 15, No. 2 177

crystd. Drying at 56’ (0.05 mm) for 24 hr gave a white solid foam; [a] ,8 , -82.6”, [a]436 -188.4” (c 0.14, CHC1,);uv max (0.1 N HCI) 270 mp (log E 4.308); (H,O) 277 (4.317); (0.1 N NaOH) 277 (4.317); ir (KBr) 3375 broad (OH, NH,, NH), 1650 (amide C=O), 1590 cm-’ (arom, NH,); ir (CHCI,) 1650, 1597; mass spectrum (probe temp ca. 260’), m/e above 80 (relative intensity) 439 (l.O), 422 (1.6), 421 ( l S ) , 420 (1.8), 419 (3.7), 418 (2.7) ,405 (1.0) ,404 (1.8), 403 ( l . l ) , 319 (6.5), 318 (34.2), 301 (22.61, 300 ( loo) , 289 (6.5), 271 (8.8), 228 (11.8), 190 (17.5), 165 (7.7), 164 (76.81, 163 (17.5), 150 (13.2), 148 (12.3), 134 (23.21, 122 (6.7), 121 (39.0), 120 (5.41, 109 (4.1), 82 (9.8), metastable transitions same as those of 19. Anal. (C,,H,,N,O,) C, H, N.

Hydrogenolysis of a mixt of 2a and 19a prepd by method B gave a 91% yield of 2 and 19 which, after sepn by chromatog, had [a] within experimental error of those of samples prepd by method A.

Maxine Palm is acknowledged. The authors also are indebted to Dr. Herb Nagasawa for providing the data on nephro- toxici ty studies.

Acknowledgments. The excellent technical assistance of

References (1) B. L. Hutchings, Chem. Biol, Purines, CIBA Found. Symp.,

(2) D. Nathans,Proc. Nut. Acad. Sci. U.S., 5 1 , 5 8 5 (1964). (3) D. Nathans and A. Neidle, Nature (London), 197, 1076 (1963). (4) B. R. Baker, J. P. Joseph, and J. H. Williams, J. Amer. Chem.

1956,777 (1957).

SOC., 7 7 , l (1955). (51 L. V. Fisher. W. W. Lee. and L. Goodman. J. Med. Chem., 13, . ,

775 (1970). (6) J. P. H. Verheyden, D. Wagner, and J. G. Moffatt, J. Org.

Chem., 36,250 (1971), and references therein. (7) P. M. Roll, H. Weinfeld, E. Carroll, and G. B. Brown, J. Biol.

Chem., 220,439 (1956). (8) B. R. Baker, “Design of Active-Site-Directed Irreversible

Enzyme Inhibitors. The Organic Chemistry of the Active Site,” Wiley, New York, N. Y., 1967, pp 79, 93.

(9) D. Nathans in “Antibiotics I, Mechanism of Action,” D. Gott- lieb and P. D. Shaw, Ed., Springer-Verlag, New York, N. Y.,

(10) R. F. Derr, C. S. Alexander, and H. T. Nagasawa, Proc. SOC. Exp. Biol. Med., 125,248 (1967).

(11) E. Kmetec and A. Tirpack, Biochem. Pharmacol., 19, 1493 (1970).

(12) I. Rychlik, J. Cerna, S . Chladek, J. Zemlicka, and Z. Haladova, J. Mol. BioL, 43, 13 (1969).

(13) E. W. Garbisch, Jr. ,J . Org. Chem., 30, 2109 (1964). (14) C. 0. Guss and R. Rosenthal, J. Amer. Chem. SOC., 77,2549

(15) C. A. Vander Werf, R. Y. Heisler, and W. E. McEwen, ibid.,

(16) R. E. Parker and N. S . Isaac, Chem. Rev., 59, 737 (1959). (17) H. J. Schaeffer and C. F. Schwender in “Synthetic Procedures

in Nucleic Acid Chemistry,” W. W. Zorbach and R. S . Tipson, Ed., Vol. 1, Interscience Publishers, New York, N. Y., 1968,

1967, pp 259-277.

(1955).

76, 1231 (1954).

PP 6-7: (18) S . David and A. Veyrieres, Carbohyd. Res., 10, 35 (1969). (19) T. C. Bruice and D. Piszkiewicz, J. Amer. Chem. SOC., 89,

(20) H. Feuer and D. M. Braunstein, J. Org. Chem., 34, 1817

(21) J. E. Zimmerman and G. W. Anderson, J. Amer. Chem. Soc.,

(22) W. W. Lee, G. L. Tong, R. W. Blackford, and L. Goodman,

(23) G . W. Anderson, J. E. Zimmerman, and F. M. Callahan, J.

(24) S. H. Eggers, S . I . Biedron, and A. 0. Hawtrey, Tetrahedron

(25) S . J. Shaw, D. M. Desiderio, K. Tsuboyama, and J. A. Mc-

(26) H. T. Nagasawa, C. S . Alexander, and K. F. Swingle, Toxicol.

(27) G. Fodor and J. Kiss, J. Chem. SOC., 1589 (1952). (28) B. R. Baker and J. P. Joseph, J. Amer. Chem. SOC., 77, 15

3568 (1967).

(1969).

89 ,7151 (1967).

J. Org. Chem., 35, 3808 (1970).

Amer. Chem. Soc., 89,5012 (1967).

Lett . , 3271 (1966).

Closkey, J. Amer. Chem. Soc., 92,2510 (1970).

Appl. Pharmacol., 11,336 (1967).

(1955).

Identification and Synthesis of the Major Nucleoside Metabolite in Rabbit Urine after Administration of Puromycin Aminonucleoside’ ?- Herbert T. Nagasawa,* Frances N. Shirota, and Carl S. Alexander Medical Research Laboratories, Minneapolis Veterans Hospital, and the Departments of Medicinal Chemistry and of Medicine, University of Minnesota, Minneapolis, Minnesota, 5541 7. Received June 4, I971

9~3’-Amino-3’-deoxy;O-D-ribofuranosyl)-6-dimethylamino-9H-purine ( l a ) , t he aminonucleoside of puromycin, when administered to rabbits is monodemethyla ted at t h e 6-N posit ion to give 943‘- am~no-3’-deoxy-~-D-ribofuranosyl)-6-methylamino-9H-purine (9) , t he la t ter consti tuting t h e major nucleoside metabolite of l a in the urine. The 3’-N-acetylated derivative of t h e metabolite, 9, i.e., 94 3‘-acetamido-3’-deoxy-~-D-ribofuranosyl)-6-methylamino-9~-punne ( 8 ) was identical in all respects ( t lc patterns, degradation products, mass spectral fragments) t o 8 synthesized chemically by methylation of 9~3’-acetamido-3’-deoxy~-D-ribofuranosyl)-6-amino-9H-purine ( 6 ) o n t h e 1 posi- t i on with MeI, followed b y rearrangement in dil NH40H. Contrary to earlier speculations, rabbits d o not metabolize l a by methyla t ion o n the 3’-amino group of the amino ribose moiety, as shown by comparison of t h e urinary metabolites of l a with chemically synthesized 3’-N-methylated derivatives of l a, v iz . , 94 3’-met hylamino-3’-deoxy-~-D-ribofuranosyl)-6-dimethylamino-9H-purine (1 c), and 9~3’-dimethylamino-3’-deoxy-~-D-ribofuranosyl)-6-dimethylamino-9H-purine ( 1 b).

The aminonucleoside l a produced by the Edman degrada- t ion of the antibiotic puromycin and independent ly synthesized, b y Baker, e t aZ.,2y3 exhibits trypanocidal as well as an t i tumor proper tie^.^" The appearance of massive, though reversible, proteinuria frustrated clinical trials of l a as a t u m o r chemotheraoeut ic agent in mane6 When ad-

?This work was supported in part by Grant HE-04983, United States Public Health Service, and Grant 65G078, American Heart Association, and was presented in part at the 52nd Annual Meeting of the Federation of American Societies for Experimental Biology, Atlantic City, N. J . , April 18, 1968.

ministered to rats by oral , sc, or i p routes, l a elicits a nephro t ic syndrome characterized by hypoproteinemia, hy- perlipidemia, hypercholesterolemia, proteinuria, edema, and ascites-a syndrome that is clinically indistinguishable f rom the kidney disease frequently observed in children.’ l a has since been utilized extensively for the experimental induct ion of this disease in rats.*

The striking species susceptibility to toxicity b y l a manifested b y i ts lack of nephro toxic i ty in mice, guinea pigs, or rabbits, does no t appear to be reflected in differential


Identification and synthesis of the major nucleoside metabolite inrabbit urine after administration of puromycin aminonucleoside

Herbert T. Nagasawa, Frances N. Shirota, and Carl S. AlexanderJ. Med. Chem., 1972, 15 (2), 177-181• DOI: 10.1021/jm00272a013 • Publication Date (Web): 01 May 2002






Metabolism of Puromycin Aminonucleoside JournalofMedicinal Chemistry, 1972, Vol. 15, No. 2 177

crystd. Drying at 56’ (0.05 mm) for 24 hr gave a white solid foam; [a] ,8 , -82.6”, [a]436 -188.4” (c 0.14, CHC1,);uv max (0.1 N HCI) 270 mp (log E 4.308); (H,O) 277 (4.317); (0.1 N NaOH) 277 (4.317); ir (KBr) 3375 broad (OH, NH,, NH), 1650 (amide C=O), 1590 cm-’ (arom, NH,); ir (CHCI,) 1650, 1597; mass spectrum (probe temp ca. 260’), m/e above 80 (relative intensity) 439 (l.O), 422 (1.6), 421 ( l S ) , 420 (1.8), 419 (3.7), 418 (2.7),405 (1.0),404 (1.8), 403 ( l . l ) , 319 (6.5), 318 (34.2), 301 (22.61, 300 ( loo) , 289 (6.5), 271 (8.8), 228 (11.8), 190 (17.5), 165 (7.7), 164 (76.81, 163 (17.5), 150 (13.2), 148 (12.3), 134 (23.21, 122 (6.7), 121 (39.0), 120 (5.41, 109 (4.1), 82 (9.8), metastable transitions same as those of 19. Anal. (C,,H,,N,O,) C, H, N.

Hydrogenolysis of a mixt of 2a and 19a prepd by method B gave a 91% yield of 2 and 19 which, after sepn by chromatog, had [a] within experimental error of those of samples prepd by method A.

Maxine Palm is acknowledged. The authors also are indebted to Dr. Herb Nagasawa for providing the data on nephro- toxici ty studies.

Acknowledgments. The excellent technical assistance of

References (1) B. L. Hutchings, Chem. Biol, Purines, CIBA Found. Symp.,

(2) D. Nathans,Proc. Nut. Acad. Sci. U.S., 51 ,585 (1964). (3) D. Nathans and A. Neidle, Nature (London), 197, 1076 (1963). (4) B. R. Baker, J. P. Joseph, and J. H. Williams, J. Amer. Chem.

1956,777 (1957).

SOC., 7 7 , l (1955). (51 L. V. Fisher. W. W. Lee. and L. Goodman. J. Med. Chem., 13, . ,

775 (1970). (6) J. P. H. Verheyden, D. Wagner, and J. G. Moffatt, J. Org.

Chem., 36,250 (1971), and references therein. (7) P. M. Roll, H. Weinfeld, E. Carroll, and G. B. Brown, J. Biol.

Chem., 220,439 (1956). (8) B. R. Baker, “Design of Active-Site-Directed Irreversible

Enzyme Inhibitors. The Organic Chemistry of the Active Site,” Wiley, New York, N. Y., 1967, pp 79, 93.

(9) D. Nathans in “Antibiotics I, Mechanism of Action,” D. Gott- lieb and P. D. Shaw, Ed., Springer-Verlag, New York, N. Y.,

(10) R. F. Derr, C. S. Alexander, and H. T. Nagasawa, Proc. SOC. Exp. Biol. Med., 125,248 (1967).

(11) E. Kmetec and A. Tirpack, Biochem. Pharmacol., 19, 1493 (1970).

(12) I. Rychlik, J. Cerna, S . Chladek, J. Zemlicka, and Z. Haladova, J. Mol. BioL, 43, 13 (1969).

(13) E. W. Garbisch, Jr. ,J . Org. Chem., 30, 2109 (1964). (14) C. 0. Guss and R. Rosenthal, J. Amer. Chem. SOC., 77,2549

(15) C. A. Vander Werf, R. Y. Heisler, and W. E. McEwen, ibid.,

(16) R. E. Parker and N. S . Isaac, Chem. Rev., 59, 737 (1959). (17) H. J. Schaeffer and C. F. Schwender in “Synthetic Procedures

in Nucleic Acid Chemistry,” W. W. Zorbach and R. S . Tipson, Ed., Vol. 1, Interscience Publishers, New York, N. Y., 1968,

1967, pp 259-277.

(1955).

76, 1231 (1954).

PP 6-7: (18) S . David and A. Veyrieres, Carbohyd. Res., 10, 35 (1969). (19) T. C. Bruice and D. Piszkiewicz, J. Amer. Chem. SOC., 89,

(20) H. Feuer and D. M. Braunstein, J. Org. Chem., 34, 1817

(21) J. E. Zimmerman and G. W. Anderson, J. Amer. Chem. Soc.,

(22) W. W. Lee, G. L. Tong, R. W. Blackford, and L. Goodman,

(23) G . W. Anderson, J. E. Zimmerman, and F. M. Callahan, J.

(24) S. H. Eggers, S . I . Biedron, and A. 0. Hawtrey, Tetrahedron

(25) S . J. Shaw, D. M. Desiderio, K. Tsuboyama, and J. A. Mc-

(26) H. T. Nagasawa, C. S . Alexander, and K. F. Swingle, Toxicol.

(27) G. Fodor and J. Kiss, J. Chem. SOC., 1589 (1952). (28) B. R. Baker and J. P. Joseph, J. Amer. Chem. SOC., 77, 15

3568 (1967).

(1969).

89,7151 (1967).

J. Org. Chem., 35, 3808 (1970).

Amer. Chem. Soc., 89,5012 (1967).

Lett . , 3271 (1966).

Closkey, J. Amer. Chem. Soc., 92,2510 (1970).

Appl. Pharmacol., 11,336 (1967).

(1955).

Identification and Synthesis of the Major Nucleoside Metabolite in Rabbit Urine after Administration of Puromycin Aminonucleoside’ ?- Herbert T. Nagasawa,* Frances N. Shirota, and Carl S. Alexander Medical Research Laboratories, Minneapolis Veterans Hospital, and the Departments of Medicinal Chemistry and of Medicine, University of Minnesota, Minneapolis, Minnesota, 5541 7. Received June 4, I971

9~3’-Amino-3’-deoxy;O-D-ribofuranosyl)-6-dimethylamino-9H-purine ( l a ) , t he aminonucleoside of puromycin, when administered to rabbits is monodemethyla ted at t h e 6-N posit ion to give 943‘- am~no-3’-deoxy-~-D-ribofuranosyl)-6-methylamino-9H-purine (9), t he la t ter consti tuting t h e major nucleoside metabolite of l a in the urine. The 3’-N-acetylated derivative of t h e metabolite, 9, i.e., 94 3‘-acetamido-3’-deoxy-~-D-ribofuranosyl)-6-methylamino-9~-punne ( 8 ) was identical in all respects ( t lc patterns, degradation products, mass spectral fragments) t o 8 synthesized chemically by methylation of 9~3’-acetamido-3’-deoxy~-D-ribofuranosyl)-6-amino-9H-purine ( 6 ) o n t h e 1 posi- t i on with MeI, followed b y rearrangement in dil NH40H. Contrary to earlier speculations, rabbits d o not metabolize l a by methyla t ion o n the 3’-amino group of the amino ribose moiety, as shown by comparison of t h e urinary metabolites of l a with chemically synthesized 3’-N-methylated derivatives of l a, v iz . , 94 3’-met hylamino-3’-deoxy-~-D-ribofuranosyl)-6-dimethylamino-9H-purine (1 c), and 9~3’-dimethylamino-3’-deoxy-~-D-ribofuranosyl)-6-dimethylamino-9H-purine ( 1 b).

The aminonucleoside l a produced by the Edman degrada- t ion of the antibiotic puromycin and independent ly synthesized, b y Baker, e t aZ.,2y3 exhibits trypanocidal as well as an t i tumor proper tie^.^" The appearance of massive, though reversible, proteinuria frustrated clinical trials of l a as a t u m o r chemotheraoeut ic agent in mane6 When ad-

?This work was supported in part by Grant HE-04983, United States Public Health Service, and Grant 65G078, American Heart Association, and was presented in part at the 52nd Annual Meeting of the Federation of American Societies for Experimental Biology, Atlantic City, N. J . , April 18, 1968.

ministered to rats by oral , sc, or i p routes, l a elicits a nephro t ic syndrome characterized by hypoproteinemia, hy- perlipidemia, hypercholesterolemia, proteinuria, edema, and ascites-a syndrome that is clinically indistinguishable f rom the kidney disease frequently observed in children.’ l a has since been utilized extensively for the experimental induct ion of this disease in rats.*

The striking species susceptibility to toxicity b y l a manifested b y i ts lack of nephro toxic i ty in mice, guinea pigs, or rabbits, does no t appear to be reflected in differential

178 Journal ofhfedicinal Chemistry, 1972, Vol. 15, No. 2 Nagasawa, Shirota, and Alexander

CH CH3 \; /

HOCH 0 lul /N\ OH

R I R 2

I

metabolic handling of the drug by the different species. Thus, the metabolic fate of l a follows similar courses, qualitatively and quantitatively, in both susceptible (rat) and nonsusceptible species (guinea pigs, mi~e).~3'' Wilson, et aZ., have suggested that rabbits, which are refractory to toxicity by la, metabolize this compound differently from rats, and speculated that perhaps rabbits detoxified la by methylation of the 3'-amino group on the amino- ribose moiety, although direct chemical evidence was not presented. This possibility is not without merit as rabbits are known to possess an enzyme system uniquely capable of N-methylating a variety of structurally unrelated amines quite nonspecifically, and this N-methyl transferase, whch is found predominantly in lung tissue, is absent in the lung fractions of the rat, mouse, or guinea pig.'* In the present study we have attempted to clarify this point by administration of l a to rabbits and examining their urines for the 3'-N-methylated products, l b and IC.

Chemistry. Synthetic 3'-N-substituted dimethyl ( lb) and monomethyl (IC) derivatives of l a were required for comparison with the metabolites of l a isolated from rabbit urine. The preparation of l b by reductive methylation

Scheme I CH C H 3 c q 3 ,"3 \" / F!

N OH NH OH

CH

' 6 %

II I

1 2 TH2 3 '6'5 / CH3 CH3

Scheme I1

NH2 On

5 on J NH on

cow, cow, 6 7

I

'\c-J * z

9

b-u' NH OH

C O W )

8

of l a with formaldehyde under catalytic hydrogenation condition^'^ led to mixtures of the 3'-N-dimethyl (lb) and 3'-N-monomethyl (IC) derivatives as revealed by tlc, unless the catalyst was prereduced prior to addition of the Schiff base. Eschweiler-Clarke methylation of l a also gave l b in good yield and free from IC.

An unambiguous synthesis of the 3'-N-monomethylated compd IC is depicted in Scheme I. The benzilidine derivative 2 of the aminonucleoside l a was reduced to the benzyl analog 3 with NaBH4 in MeOH, and 3 methylated with CH20-HC02H to the benzyl methyl derivative 4. Catalytic hydrogenolysis of the benzyl group gave IC.$

9-(3 '-Amino-3~-deoxy-~-D-ribofuranosyl)-6-methylamino- 9H-purine (9), the major nucleoside metabolite of la in the rat, guinea pig, the mouseg-" has previously been prepared by solvolysis of the halogen in the corresponding blocked 6-chloropurineaminonucleoside with MeNHz fol- Iowed by debl~cking'~-" and by enzymatic demethylation of la." We have prepared 9 acetylated on the 3'- amino N, viz., 8, by an alternate method (Scheme 11) starting from 3'-amino-3'-deoxyadenosine (S), the antitumor principle isolated from the broth of Helminrhospori- um sp. No. 215.19 The latter has been synthesized previously by Baker, et al. 2o Treatment of 5 in aq soln with Ac10 selectively acetylated the 3'-amino group to give 6 . Methylation of 6 with Me1 in DMA gave the 1-methiodide (7) which rearranged8 in dil NH40H to 8.

Metabolism and Degradation Studies. After administration of l a t o rabbits, the nucleoside fraction isolated from urine by ion-exchange chromatography contained 3 aminonucleoside components separable and detectable by tlc (Figure 1A). One of these was unchanged la, but none of the metabolites corresponded in Rf to the 3'-N-methylated derivatives l b and IC prepared above. Although one of the two metabolites appeared to have low mobility such as would be expected for 3'-amino-3'deoxyinosine, this product was present only in small amounts and our inter-

$Lee, et aL," have recently reported the synthesis of l b and I C by somewhat different procedures.

§This rearrangement very likely proceeds by opening of the pyrimidine ring followed by recyclization in a manner similar to the rearrangement of 1-methyladenosine and 1-methyl-2'-deoxyadenosine to 6-N-methyladenosine and 6-N-methyl-2'-deoxyadenosine, respect ively.21

Metabolism of Puromycin Aminonucleoside Journal ofMedicinal Chemistry, 1972, Vol. 15, No. 2 179

I / e .......,. . . . . . . .

(a) (b) ic) id) (e) if) (gi ih) 0) (1) la) (b) (c ) id) (e) if) -1 1- Figure 1. Tlc patterns of I a metabolites and their degradation products. o = Fluorescence-quenching under 2.5374 lamp; = red after spraying with aniline phthalate reagent and heating at 110" for 0.5 hr. A. Silica gel HF,,,, CMA-11. (a) Ib; (b) IC; (c) crude nucleoside fraction from rabbit urine after administration of la; (d) la; ( e ) 5 ; (f) 3'-amino-3'-deoxyinosine. The patterns were essentially reversed on Silicar tlc 7-GF with 88%HCO,H- abs EtOH-H,O (3:16:3). B. Silicar tlc 7GF, CMA-I. (a) 6-Methyl- aminopurine; (b) major nucleoside metabolite (hydrolyzed = hyd); (c) acetylatedmetabolite(hyd); (d) synth 8 (hyd), ( e ) l a (hyd); (0 acetylated la (hyd); (g) 6-dimethylaminopurine; (h) synth 3- amino-3-deoxy-D-ribose; (i) D-ribose; 6 ) adenine. C. Silicar tlc 7- GF, 88% HC0,H-abs EtOH-H,O (3:32:6). (a) Synth 3-amino-3- deoxy-D-ribose; (b) l a (hyd); (c) major nucleoside metabolite (hyd); (d) D-ribose; ( e ) 6-methylamino purine; (f) 6-dimethylaminopurine.

est was focused on the other metabolite which predomi- nated. This major nucleoside metabolite, which was isolated by prep tlc, exhibited a uv absorption maximum at 266 nm (minimum, 236 nm), and gave a positive test with ninhydrin under alk conditions. Acetylation gave a product no longer chromogenic with ninhydrin reagent. Hy- drolysis of either the metabolite or its acetylated derivative gave a sugar moiety which tested positive with ninhydrin, aniline phthalate, and alkaline AgN03 reagents, and a purine base whose uv absorption spectra in acid and al- kali resembled that for 6-methylaminopurine. The tlc patterns of the hydrolysate compared against 6-methylaminopurine and synthetic 3-amino-3-deoxy-D-ribose in several solvent systems (Figures 1 A and 1 B) confirmed their identity, leading to the conclusion that this metabolite was 9- (3'-amino-3' -deoxy-fl-D-ribofuranosyl)-6-met hylamino-9i- purine, i.e., 9.

Further evidence was adduced from analysis of the mass spectra of the metabolite and its acetylated derivative (Table I). The aminonucleoside metabolite exhibited a molecular ion at m/e 280, while the acetylated product showed the expected molecular ion at m/e 322. The presence of unsubstituted 5'-OH groups in both compounds was indicated by the (M - 30) peaks (loss of CH20) at m/e 250 and mle292. The (M - 59) peak (loss of CH3CONH2)22y23at m/e 263 in the acetylated derivative verifies the chemical evidence that the acetylation occurred on the 3'-amino N. The series of ions (Table I) which include the purine base fragment b, the base plus various portions of the sugar skeleton, and the ions m/e 121 (b t 2H - 29) and m/e 120 @tH-29), clearly establish the purine base moiety of the molecule as 6-methylaminop~rine.~~ Synthetic 8 had essentially identical mass spectral fragmentation patterns (Table I) as well as identical Rf values on tlc in several solvent systems as this acetylated metabolite, thereby fully identifying the latter as 8. It follows that the metabolite of l a isolated from rabbit urine is, in fact, 9.

Thus, the early metabolic transformations of l a in the rabbit appear to be similar to that observed in the other

180 JournalofMedicinal Chemistry, 1972, Vol. 15, No. 2 Nagasawa, Shirota, and Alexander

species examined ,9310 the 6-N-mono demet hylat e d product of la, viz. 9, predominating, and no evidence for the presence of 3'-N-methylated metabolites of l a could be found in the urine of rabbits dosed with la. Both New Zealand white and Shingler strain rabbits gave the same results, and no qualitative differences in the metabolic handling of l a were observed in young or mature rabbits. Furthermore, incubation of la with S-adenosylmethionine and the soluble fraction from rabbit lung containing the N-methyl transferase described by Axelrod'* failed to give l b or IC as determined by tlc.

Toxicity Studies. The nephrotoxicity of some of the intermediates and products synthesized in the course of this work was evaluated in rats by measuring their daily urinary protein excretion after administration of the compounds according to previously established protocol^.^^^# The benzylidine derivative 2 (dose: 5 1 pmoles/kg per day for 24 days, sc) to male albino rats weighing 45-50 g caused the appearance of mild proteinuria commencing at day 14, which is 7-10 days later than the usual onset of massive proteinuria following treatment with equimolar doses of la. Ascites and edema were absent in these animals even at day 25; 2 is therefore comparable in nephrotoxicity to puromycin itself.26 No evidence of nephrotoxicity to rats was observed with l b (dose: 5 1 pmoles/kg per day for 23 days); however, rats given IC (dose: 51 pmoles/ kg per day for 27 days) developed slight proteinuria commencing on day 23 which persisted until the drug was withdrawn on day 28. Proteinuria gradually subsided thereafter and disappeared by day 41. Rats sacrificed on days 29 and 30 had 1-2 g of ascitic fluid in the peritoneal cavity.

Experimental Section* *

l a as a 1% aq soln was administered sc to male New Zealand white rabbits housed in metabolism cages (dose: 30 mg/kg per day in divided doses to 2 different sites) for 6 days. The daily urine collec- tions were adjusted to pH 8.0 t 0.5, pooled, and stored refriger- ated under toluene. Control urine samples collected for 6 days prior t o administration of l a were separately pooled and treated similarly in the following description. After centrifugation of the urine at 2500 rpm at +5" for 20 min, the clear amber supernatant was dild five- t o tenfold with distd H,O, adjusted to pH 1 with concd HC1, and charged on a 3.6 X 14.5 cm column of Bio-Rad Ag 50-X4 cation-exchange resin (100-200 mesh, NH:). The column was washed with 500 ml of 0.01 N HCl and then approx 6 0 ml of the discolored resin at the topmost portion of the column was withdrawn and transferred to a smaller column contg about 20 ml of fresh resin, The new column was washed with 3 1. of 0.05 M ammonium formate buffer (pH 5.9) t o remove the free purine bases. The desired nucleoside fraction was obtained by elution with 3 1. of 0.2 M ammonium formate buffer (pH 7.9). This latter fraction was adjusted to pH 1 and desalted by (a) charging onto a fresh column of AG 5@X4 (NH;), (b) washing the column with distd H,O, and (c) eluting the nucleoside metabolites with 1.0 N NH,OH. Eluates having uv absorbance of >0.1 were pooled and lyophilized. This crude isolate contd 3 uv-absorbing compds as detd by tlc, one of which was unchanged l a (Figure 1A). None of these spots corresponded in R f to synthetic l b and IC, and control

Isolation of Nucleoside Metabolites of l a from Rabbit Urine.

#Doses are based on the weights of the animals on day 1 . **Melting points were taken on a Fisher-Johns mp apparatus and

are corrected; optical rotations were detd in a Perkin-Elmer Model 141 polarimeter. Spectrophotometers used were: uv, Beckman DK- 2A; ir, Beckman 1R-10; ms, Hitachi-Perkin Elmer RMU-6 (ionization energy, 70 eV; ion source temps as indicated). Microanalyses by Schwarzkopf Microanalytical Laboratory, Woodside, N. Y., or Gal- braith Laboratories, Knoxville, Tenn. Where analyses are indicated only by symbols of the elements, analytical results obtained for these elements were within *0.4% of the theoretical values. TIC solvents: CHCl,-MeOH-2.5 NNH, (50:20:3) (CMA-I); CHC1,- MeOH-1 N NH, (5 :2: 1, lower phase) (CMA-11).

urine samples did not show the presence of any of these components.

Purification and Acetylation of the Major Nucleoside Metabo- lite. The metabolite with Rf -0.15 was sepd from l a and from the small amount of unknown product near the origin by preparative tlc twice on silica gel with CMA-I1 and eluted with 0.1 N NH,. The eluate was concd to dryness, and the residue was recrystd from MeOH to give colorless crystals, mp 238-241", uv max (H,O) 266 nm, min 236 nm. This product on tlc or paper chromatog gave a blue-violet color with ninhydrin reagent at a level of 5 pg, as did 1, 5, and 3'-amino-3'-deoxyinosine; but lb, IC, or acetylated l a did not. This metabolite was therefore unsubstituted on the 3'- amino group of the amino sugar moiety. Treatment of this metabolite in H,O with Ac,O gave a product, mp 236-237" after recrystn from MeOH-Et,O, which was no longer chromogenic with ninhydrin indicating that the 3'-amino group was now acetylated. The mass spectra of these products are recorded in Table I .

Hydrolytic Degradation Studies. The metabolite was hydrolyzed in 0.5 ml of 2 N HC1 for 3 hr at 100". For comparative purposes, the following compds were also subjected to the conds of this hydrolysis, viz., la, acetylated la, acetylated metabolite, 1 b, IC, and 8 prepd synthetically (vide infra). l b was not completely hydrolyzed under these conds, and some unchanged nucleoside was still detectable by tlc after the prescribed period (Silicar tlc 7- GF; 5% (NHJ,CO,-95% EtOH, 90:7 or 90:3). The purines liberated from this hydrolysis were compared with 6-dimethylaminopurine, 6-methylaminopurine, and adenine, and the results (Figure 1B) showed the product from the metabolite to be 6-methylaminopurine, confirmed by comparison of its uv spectrum in 0.01 N HC1 and 0.02 N NaOH with an authentic sample. The 3-dimethylamino- 3-deoxy-D-ribose and the 3-methylamino-3-deoxy-D-ribose resulting from the hydrolysis of I b and IC, respectively, were not chromogenic with ninhydrin but gave the characteristic red color for pentoses with aniline phthalate reagent. The sugars liberated from all of the above nucleosides gave positive alk AgNO, tests, but only the sugar resulting from la, acetylated la, the metabolite, acetylated metabolite, and synthetic 8 had identical Rf values in two solvent systems as synthetic 3-amino-3-deoxy-D-ribosez7 (Fig- ures 1B and 1C).

94 3'-Dimethylamino-3'-deoxy-~D-ribofuranosyl)-6-dimethylamino-9H-purine (Ib). A. By Reductive Methylation. PtO, (500 mg) was prereduced in 50 ml of 95% EtOH-HOAc (1:l) at 1.05 kg/cm2 for 30 min. The mixt was carefully equilibrated to the atm (caution: f i e hazard) and 594 mg (2.02 mmoles) of l a in 5 ml of 37% aq CH,O was added followed by an addnl50 ml of 95% EtOH-HOAc. The mixt was hydrogenated at room temp until H, uptake ceased (5 hr), the catalyst then was removed by filtration (Celite), the filtrate was concd to dryness, the residue was dissolved in abs EtOH, and the solvent was evapd in vacuo. This was repeated until the odor of HOAc and CH,O was no longer detectable. Recrystn of the crude product from EtOAc gave 519 mg (80.6%) of lb, mp 184-185" (lit.I4 mp 184.5-186"). A second crop, 106 mg, mp 183-1234", was obtained by concn of the mother liquor; total yield, 625 mg (97%); uv max (MeOH) 274 nm ( E 19,700); (cyj*'D -60.1' (c 1.60, Py). Anal. (C,,H,,N,O,) C. H, N. If the catalyst was not activated by prereduction, or if the catalyst was Pd black, mixts of the mono (IC) and dimethyl (1 b) derivatives were formed requiring extensive chromatog manipula- tion for sepn.

mmole) in 2.0 ml of 88% HCO,H and 0.5 ml of 37% aq CH,O was heated under reflux for 2 hr. Tlc indicated that l a had completely reacted by this time. The reaction mixt was dild with 2 vols of abs EtOH and concd to incipient dryness in vacuo, and the excess HCO,H and CH,O were removed as azeotropes by successive addn of PhMe and EtOAc followed by evapn. The glassy residue which was contaminated with the 5'-formyl derivative of I b [ir (KBr) 1730 cm-' (ester); m/e 350 (M')] was dissolved in MeOH contg 1.0 ml of Et,NH and heated under reflux for 1 hr. Evapn of the solvent and recrystn of the residue from CH,CN gave 122 mg (76% yield) of lb , mp 184-185" (lit.', 184.5-186'). An analytical sample recryst from CH,CN gave the same mp. Anal. (C,,H,,N,O,) C, N, N.

9-(3'-Benzal im~o-3'-deoxy-p-D-ribofu~no~l)~dimethyl~~ no-9H-purine (2). A mixt of 1.00 g (3.40 mmoles) of l a and 1.3 ml of freshly distd PhCHO in 30 ml of abs EtOH was heated on the steam bath under reflux for 2 hr. The solvent was evapd to incipient dryness in vacuo and the glassy residue was crystd from EtOAc-petr ether t o give 1.20 g (92% yield) of 2, colorless crystals, mp 162.5-164.5"; uv max (MeOH) 270 nm ( C 23,800). Anal.

B. By Eschweiler-CLatke Methylation. l a (147 mg, 0.50

Metabolism of Puromycin Aminonucleoside Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 181

(C,&I,,N,O,) C, H, N. A second crop, 75 mg, obtained from the mother liquor melted at 158-161".

no-9H-purine (3). To a stirred soln of 1.23 g (3.22 mmoles) of 2 in 50 ml of MeOH was added slowly at room temp, 600 mg of cryst NaBH,. The reaction mixt was stirred at room temp for 15 min, then heated under reflux on a steam bath for 2.5 hr and filtered. The filtrate was concd to incipient dryness, and the solid residue was dissolved in 25 ml of H,O. After multiple extns with EtOAc, the combined EtOAc exts were washed (H,O), dried (Na,SOJ, and reduced in vol when 3 crystd spontaneously, 1.06 g (86% yield), mp 165-167" (lit.I3 162-164'; uv max (MeOH) 274 nm ( E 19,400); [a]''D -49.2" (c 1.60, Py) (lit." [or]"D -44", Py). Anal. (C1$24N60J C, H, N. In some runs recrystn of the crude product from EtOAc-petr ether or MeOH-Et,O gave crystals melting at 122-125" or slightly lower. That this was a dimorphic cryst form of the higher melting product was indicated by (a) the identity of their Rf values on tlc (CMA-11); (b) the identity of their ir spectra in CHCl,, and (c) the elemental analysis. Anal.

9 4 3'-Benzylamino-3'-deoxy-~D-ribofurano~l~6-dimethylami-

(C,$,4N,O,) C, H, Ne 9- [ 3'-(N-Methyl~-benzyl)amino-3'-deoxy-p-D-ribofuranosyl]-

6-dimethylamino-9H-purine (4). 3 (384 mg, 1.0 mmole) in a mixt of 0.6 ml of 88% HCOOH, 0.3 ml of 37% aq CH,O, and 3.0 ml of 95% EtOH was heated under reflux until tlc showed no trace of remaining 3 (10 hr). The solvent was evapd, and the residual glass was subjected to the action of a mechanical vac pump for 18 hr. Abs EtOH was added to dissolve the residue, and the EtOH was evapd in vacuo. This process was repeated until the odor of HCOOH was no longer detectable. The fluffy white solids were recrystd from EtOAc-petr ether, 345 m (86.6%), mp 137-139";

Anal. (C,&,,N,O,) C, H, N. 9 4 3'-Methylamino-3' deoxy-p-D-ribofuranosyl)-6dimethylamino-

9H-purine (IC). A soln of 154 mg (0.462 mmole) of 4 in 6 ml of 95% EtOH-HOAc (1: 1) was heated on the steam bath for a few min with a spatulaful of Pd black, and the catalyst removed by filtration. This procedure was necessary to remove traces of impurities which poisoned the catalyst. Fresh Pd black catalyst (78 mg) was added to the filtrate, and the mixt was hydrogenated at room temp in an all-glass micro hydrogenator until tlc (CMA-11) indicated complete removal of the benzyl group (3 hr). After removal of the catalyst, the solvent was evapd to dryness in vacuo, and the residue was triturated with ether to give 127 mg of crude product, mp 208-212" dec. Recrystn from abs EtOH gave 117 mg (82%) of IC, mp 212-214" (lit." 216.5-218"); uv max (MeOH) 247 nm ( E 19,600); ir (KBr) 3275 (OH), 3125 (NH), 1595 cm-' (C=N). Anal. ( C ~ & O N ~ O ~ ) c, H, N.

9 4 3'-Acetamido-3'-deox~p-D-ribofurano~l)-6-amino-9~-purine (6) . A 100-mg (0.37 mmole) sample of 5"in 1.0 ml of H,O was acetylated in the cold by addn of 200 pl of freshly distd Ac,O. The reaction, monitored by tlc (CMA-I), was found to be complete within 1 hr. The mixt was evapd to dryness in vacuo and the residue recrystd from abs EtOH to give 57 mg (50% yield) of 6, mp 250-252"; ir (KBr) 1690 (amide I), 1660 (C=N), 1565 cm-' (amide 11). Anal. (C,,HI6N,O?) C, H, N. The highest melting sample obtd was 258-259" dec 263-265'; 247" dec), but the mp (dec pt) appears to be a poor criterion for assessing purity, and tlc was relied upon. Crop 2 (33 mg) was obtd by diln of the mother liquor with Et,O, homogeneous by tlc.

9 4 3'-Acetamido-3'-deoxy-p-D -ribofuranosyl)-6-methylamino- 9H-purine (8). To 74 mg (0.24 mmole) of 6 dissolved in 5 ml of di- methylacetamide was added 500 pl of Mel, and the reaction mixt was stirred at room temp until 6 disappeared (19 hr) as revealed by tlc (CMA-11). The reaction mixt was poured into 250 ml of petr ether, and the yellow oil that settled out was taken up in Me,CO. The white hygroscopic product which pptd on addn of Et,O was collected, washed (petr ether), and dried to give 11 mg of crude 7. This product was dissolved in 1 N NH, and passed through a column containing 10 g of 50-100 mesh Dowex-3 ( O H ) resin at a rate of 12-15 drops/min collecting 10-ml fractions. The tubes having uv max of 265 nm and min below 235 nm were combined and lyophilized. The residue was dissolved in abs EtOH and the EtOH evapd to remove traces of H,O. The residue was

uv max (MeOH) 274 nm (E 19,800); [ a ] 1 5 D -32.1" (c 0.93, Py).

then extd several times with hot EtOAc and the combined EtOAc extracts concd to give 58.2 mg (73%) of crude 8, mp 219-223' dec. Recrystd from MeOH-Et,O, mp 235-237" 230" and 233" for compd with 0.25 mole of H,O). Anal. (C13H,,N60, *H20) c, H, N.

Acknowledgments. We are indebted to Dr. Nancy N. Gerber for an inoculum of Helminthosporium sp No. 21 5 as well as a sample of 5 isolated from this organism and to Dr. Elmer Reist for experimental quantities of chemically synthesized 5. Both samples behaved similarly in the tlc system described in the Experimental Section. We also thank K. F. Swingle and M. Coleman for technical assistance in the toxicity screening and J. McMahon for the mass spectra. Lederle Laboratories Division, American Cy- anamide Co., provided generous samples of puromycin aminonucleoside .

229-

References (1) H. T. Nagasawa, C. S. Alexander, and R. F. Derr, Fed. Proc.,

(2) B. R. Baker, J. P. Joseph, and J. H. Williams, J. Amer. Chem Fed. Amer. SOC. Exp. Biol., 27,465 (1968).

Soc.. 77. l(195.5). , ~ .~ ...,_ (3) B. R. Baker, R. E. Schaub, J. P. Joseph, and J. H. Williams,

ibid., 77,12 (1955). (4) R. I. Hewitt, A. R. Gamble, W. S. Wallace, and J. H. Williams,

(5) S. L. Halliday, P. L. Bennett, and J. J. Oleson, Cancer Res., 15,

(6) J. H. Burchenal, Current Res. Cancer Chemother., 4 , 1 6 (1956). (7) S. Frenk, S. Antonowicz, J. M. Craig, and J. Metcalf, Proc. Soc.

(8) U. C. Dubach, Prog. Drug. Res., 7, 356 (1964). (9) H. T. Nagasawa, K. F. Swingle, and C. S . Alexander, Bwchem

Antibiot. Chemother., 4, 1222 (1954).

693 (1955).

Exp. Biol. Med., 89,424 (1955).

PharmcoL, 16, 2211 (1967).

Nagasawa, ibid., 17, 265 (1968).

25, 228 (1960).

(10) R. F. Derr, D. K. Loechler, C. S. Alexander, and H. T.

(11) S. G. F. Wilson, W. Heymann, and D. A. Goldthwait, Pediatrics,

(12) J. Axelrod, J. Pharmcol. Exp. Ther., 138, 28 (1962). (13) British Patent 782,440 (Sept 4, 1957); Chem. Abstr., 52,

2938s (1958). (14). W. W. Lee, G. L. Tong, R. W. Blackford, and L. Goodman, J.

Org. Chem., 35,3808 (1970). (15) B. R. Baker, J. P. Joseph, and R. E. Schaub, U. S. Patent

2,852,505 (Sept 16,1958); Chem. Abstr., 53,8175 (1959). (16) L. Goldman and J. W. Marsico, U. S. Patent 2,852,506 (Sept

16, 1958); Chem. Abstr., 53,8174 (1959). (17) L. Goldman and J. W. Marsico, J. Med. Chem., 6 , 413 (1963). (18) N. Dickie, L. Norton, R. F. Derr, C. S. Alexander, and H. T.

Nagasawa, hoc. Soc. Exp. Biol. Med., 123, 421 (1966). (19) N. N. Gerber and H. A. Lechevalier, J. Org. Chem., 27, 1731

(1962). (20) B. R. Baker, R. E. Schaub, and H. M. Kissman, J, Amer. Chem

Soc., 77,5911 (1955). (21) J. W. Jones and R. K. Robins, J. Amer. Chem. Soc., 85, 193

( 1963). (22) R. J. Suhadolnik, B. M. Chassy, and G. R. Waller, Biochim

Biophys. Acta, 179, 258 (1969). (23) S. H. Eggers, S . I. Biedron, and A. C. Hawtrey, Tetrahedron

Lett., 3271 (1966). (24) S. J. Shaw, D. M. Desiderio, K. Tsuboyama, and J. McCloskey,

J. Amer. Chem. Soc., 92,2510 (1970). (25) H. T. Nagasawa, C. S. Alexander, and K. F. Swingle, Toxicol.

Appl. Pharmcol., 11, 336 (1967). (26) B. A. Borowsky, D. M. Kessner, and L. Recant, Proc. Soc. Exp.

Biol. Med.,97, 857 (1958). (27) B. R. Baker, R. E. Schaub, and J. H. Williams,J. Amer. Chem.

SOC., 7 7 , 7 (1955).


Nucleosides of 2-azapurines and certain ring analogsJohn A. Montgomery, and H. Jeanette Thomas







182 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Montgomery and Thomas

Nucleosides of 2-Azapurines and Certain Ring Analogs? John A. Montgomery* and H. Jeanette Thomas Kettering-Meyer Laboratory, Southern Research Institute, Birmingham, Alabama 35205. Received August 20, 19 71

A convenient synthesis of 2-azapurine nucleosides (7-glycosylimidazo[ 43-d 1-v-triazines) in 5 steps from the corresponding purine nucleosides is described. 2-Azaadenosine, 2 -deoxy-2-azaadenosine, and 4-amino-1 -P-D-ribofuranosylpyrazolo-4-carboxamidine are all cytotoxic at low concns, and 2-azaadenosine has shown consistent activity against L1210 leukemia on both chronic and single-dose schedules.

Azapurines have shown anticancer activity and other interesting biological properties.' The ribonucleosides of certain 8-azapurines (v-triazolo [4,5-d]pyrimidines) have been

and have shown biological activity as the intact nucleoside^,'^^ 3 but the synthesis of a 2-azapurine (imidazo [4,5-&-triazine) nucleoside has not previously been described.$ Stevens, et al., described the preparation of 2- azaadenosine l-oxide but were unable to reduce this material to 2-azaadenosine (4-amino-7-0-D -ribofuranosylimid- azo [4,5-d]-v-triazine, 13b). It occurred to us that an adap- tation of the procedure developed by us for the synthesis of

5-amino- 1-0 -D -ribofuranosylimidazole4carboxamide (15b)9 might provide an intermediate suitable for the preparation of nucleosides of 2-azaadenine. However, such an approach to the preparation of imidazolecarboxamidines from adenine nucleosides was complicated by the ease of reclosure of the proposed intermediates of the Dimroth rearrangement of 1 -substituted adenines. Thus Taylor," and later Leonard," observed that 1-substituted adenines rearrange quantitatively in boiling water to N-substituted adenines; no imidazole intermediates were isolated.

Our initial studies were carried out on a model c o m

14 15 16

a series, R = cyclopentyl b series, R = @-Pnbofuranosyl c series, R = 2-deoxy-pD-ribofuranosyl d series, R = p-Parabinofuranosyl e series, R = p-Pxylofuranosyl

?This work was supported by funds from the C. F. Kettering Foundation, Chemotherapy, National Cancer Institute, National In- stitutes of Health, Contract No. NIH-71-2021.

$A preliminary communication describing part of this work has appeared.8

pound, 9-cyclopentyladenine (la), which was benzylated in the usual manner12 to give 1 -benzyl-9-cyclopentyladenine hydrobromide (2a) and converted to its I-oxide (5a) by the procedure of Stevens, et al. l3 Treatment of 2a with reflux-

2-Azapurine Nucleosides Joumal ofMedicinal Chemistry, 1972, Vol. 15, No. 2 183

ing aq ethanolic NaOH (excess) caused the Dimroth rearrangement to occur in agreement with previous observations” giving N-benzyl-9-cyclopentyladenine (3a), although a better yield was obtained if the reaction was carried out in aq EtOH containing only 1 equiv of NaOH. In either case, no intermediate was detected. Base hydrolysis of the 1 -oxide Sa gave 1 cyclopentyl-5-formamidoimidaz- ole4carboxamidoxime (6a), which was deformylated in acid to himino- 1 -cyclopentylimidazole4carboxamidox- ime (7a). 7a could also be produced directly from 6a by acid treatment.14 Catalytic reduction of the carboxamidoxime 7a with Ra Ni gave 5-amino-1 eyclopentylimidazole-4- carboxamidine (9a), which was isolated as the HCl salt and converted to 9-cyclopentyl-2-mercaptoadenine (4) by treatment with CS2 in DMF and to 9-cyclopentyl-2-azaadenine (4-amino-7-cyclopentylimidazo [4,5-d]-v-triazine, 13a) by treatment with aq NaN02. Hot, dil aqNaOH caused cleavage of both rings of 13a to give approximately equal amounts of 5 -amino-1 -cyclopentylimidazole-4-carboxamide (1 Sa)9 and 4,5-diamino-6-cyclopentylamino-v-triazine (16a). Since aq base is known to convert adenosine to 4,5,6-triaminopyrimidine (16, R = H) exclusively, these results indicate that the v-triazine ring is more labile to base than the pyrimidine ring. Since previous work from this laboratory showed that imidazo[4,5-e]-s-triazolo [ 1 ,SI-v-triazine (17) exists in equilibrium with 3- [ 5(4)-diazoimidazol-4(5)-yl]-s- triazole (18) and that only 18 was detected in CF3C02H solution of 17,”we examined the ir spectrum of 13a in CF3C02H for the presence of a diazo band; its absence indicated that if there is an equilibrium between 13a and the corresponding diazo compound, the equilibrium favors 13a almost exclusively.

17 18

Although the route described above provided the desired model compounds for our nucleoside work, it is not amen- able to nucleosides since the acidic cleavage of the pyrimidine ring would also result in cleavage of the glycosyl linkage. Furthermore, in contrast to the results of Stevens, et a1.,16 we found that aq base treatment of Sa opened the pyrimidine ring but did not cause deformylation of the resultant formamido compound 6a. Also, as mentioned above, Stevens, et al., were unable to reduce 2-azaadenosine 1- oxide to the desired 2-azaadenosine (13b).I6 Benzylation of the 1-oxide Sa gave, in agreement with the results of F ~ j i i , ’ ~ 1 -benzyloxy-9-cyclopentyladenine bromide (loa), an un- stable compound that could be purified by careful recrystallization from EtOH. Boiling in EtOH or H20, however, caused debenzylation back to the 1 -oxide. In boiling H20 the free base from 10a was converted toN-benzyloxy-9cy- clopentyladenine (14a). If, however, the crude bromide was carefully neutralized, and the neutral EtOH-H20 solution was allowed to stand 2 days at room temp, ring opening occurred to give N-benzyloxy-1 -cyclopentyl-5-formamidoimidazole-4-carboxamidine (1 la), a result also in general agreement with the findings of Fujii.17 An attempt to de- formylate l l a with methanolic HCl gave only l-benzyloxy- 9-cyclopentyladenine (loa), in contrast to the Dimroth rearrangement obtained in boiling H20 (loa or l l a + 14a). Hydrogenolysis of 1 l a with Ra Ni gave primarily 9-cyclo-

Table 1. Cytotoxicity of 2-Azaadenine Nucleosides

ComDound ED.,. umolelLa 2-Azahypoxanthine 1.5 2-Azaadenine 4.0 9-Cy clopent y l-2-azaadenine 49 2-Azaadenosine 0.22 2’-Deoxy-2-azaadenosine <2.0 9-p-D-Arabinofuranosyl-2-azaadenine >loo 9-0-D-Xylofuranosyl-2-azaadenine >loo

[3,4+-triazine 20 4-Amino-7-p-D-ribofuranosylpyrazolo-

“The concn required to inhibit the growth of treated cells to 50% of that of untreated controls as measured by colony counts.’

pentyladenine (la), whereas hydrogenolysis with Pd/C gave primarily 9-cyclopentyladenine 1 -oxide (Sa), resulting from hydrogenolysis of the N-0 bond with Ra Ni and of the C-0 bond with Pd/C with concomitant ring closure in both cases. It was not possible to prevent cyclization of l l a or its reduction products (6a and Sa). It seemed possible that deformylation of 1 l a might be accomplished by a transamidation reaction that could compete effectively with the intramolecular ringclosure reaction. Treatment of 1 la with methanolic NH3 (satd at 0”) at 80” for 2 days gave the desired 5-amino-N-benzyloxy-1 -cyclopentylimidazole-4- carboxamidine (12a) in good yield. Hydrogenolysis of the benzyloxy group of 12a with Ra Ni then gave 5-amino-l-c~- clopentylimidazole-4-carboxamidine (sa). It seemed prob- able that, despite the precedent of the Dimroth rearrangement, 1-benzyloxy-9-cyclopentyladenine hydrobromide (loa) could be converted directly to 12a by the methanolic N H 3 treatment, and this turned out to be the case. The conditions employed in this route to 9a are completely compatible with nucleosides.

1-Benzyloxyadenosine hydrobromide (lob), prepared by the method described above, behaved exactly as the cyclopentyl compound; refluxing a neutralized EtOH-H20 solution of 10b gave N-benzyloxyadenosine (14b), while standing 3 days at room temp gave the imidazole l l b , which could be deformylated with methanolic NH3 to 12b. It was also possible to convert 10b directly to 12b. In order to establish whether the particular conditions of this ring-opening procedure or the benzyloxy group permitted ring-opening followed by deformylation rather than reclosure (Dim- roth rearrangement), we prepared 1-benzyladenosine and subjected it to the methanolic NH3 treatment, which gave N-benzyladenosine exclusively.§ Thus, the decreased ba- cisity of the N-benzyloxyamidine (1 lb) compared to the N-benzylamidine is essential for successful deformylation, but the use of methanolic NH3 is also necessary (see above). Ra Ni-catalyzed hydrogenolysis of the benzyloxy group of 12b then gave 5-amino-1 -P-D -ribofuranosylimidazole-4-carboxamidine (9b). Thus, the 5-amino-1 -glycosylimidazole-4- carboxamidines requisite for conversion to 2-azaadenine nucleosides (13) can now be obtained from the corresponding adenine nucleosides (1) by a relatively simple 4-step procedure. Treatment of the aminoamidine (9b) with NaN02 in aq AcOH gave 2-azaadenosine (13b). This procedure was also used to prepare 2’-deoxy-2-azaadenosine (4-amino-7- (2-deoxy-0-D -erythro-pentofuranosylimidazo [4 ,541-V- triazine, 13c), 9-/3-D-arabinofuranosyl-2-azaadenine (4- amino-7-0-D -arabinofuranosylimidazo[4,5-d]-v-triazine, 13d), 9-/3-D-xylofuranosyl-2-azaadenine (4-amino-7-0-D- xylofuranosylimidazo[4,5-d]-v-triazine, 13e), and 4-amino-

§This result was anticipated, since treatment of 1-benzyl-9-P-D ribofuranos lpurine-6(1H)-thione in a similar manner gave N-benzyladenosine. ,H

184 Journal ofMedicinal Chemistry, 1972, Vol. IS, No. 2 Montgomery and Thomas

HO OH HO OH HO OH

19 20 21

a series, X = CH b series, X = N

7-~-D-ribofuranosyl-7H-pyrazolo[ 3 ,4-d]-v-triazine (21a) from 4-amino- 1 -0 -D -ribofuranosy lpyrazolo [ 3,4- d ] pyrimidine ( 19a)19 and 2,8-diazaadenosine (7-amino-3-P-D-ribofuranosyl-v-triazolo[4,5-d]-v-triazine, 21b) from 8-azaadenosine (19b)20 via the corresponding pyrazole (20a) and triazole (20b).

rines to human epidermoid carcinoma cells No. 2 in cul- Biologic Evaluations. The cytotoxicities of the 2-azapu-

ture' are given in Table I. 2-Azaadenosine (13b) is 5 times as toxic as 8-azaadenosine,'20 times as toxic as 2-azaadenine, and about 7 times as toxic as 2-azahypoxanthine, indicating that neither enzymatic sugar cleavage nor deamina- tion followed by sugar cleavage appear to be important to the activity of this nucleoside. Of the other nucleosides of 2-azaadenine, only 2'-deoxy-2-azaadenosine (1 3c) appears to be cytotoxic, but its EDS0value has not been accurately determined. 4-Amino-7-fl-D -ribofuranosylpyrazolo[ 3,4-d]- v-triazine (21a), the 2-aza analog of the ribonucleoside of 4-aminopyrazolo[ 3,4d]pyrimidine (4-APP) is only 0.005 as toxic, but the synthetic intermediate to 21a, 4-amino-l- p-D-ribofuranosylpyrazolo-4-carboxamidine (20a), is quite cytotoxic, having an ED5,, value of 1.9 pmoles/l.

2-Azaadenosine (13b) given at levels of 75-400 mg/kg on day 1 only increased the life-span of BDF, mice injected ip with lo5 leukemia L1210 cells by 30-35%. It was about equally active at levels of 8-23 mg/kg per dose when given chronically qd 1-9. None of the other 2-azapurines or the intermediates leading to them that have been tested has shown any significant activity against leukemia L 1 2 10.

Experimental Section

corrected. Uv spectra were detd in aq s o h with a Cary Model 14 Melting points were detd with a Mel-Temp apparatus and are not

Table 11. 9-Pentofuranosyl-2-azapurines

Crude yield, Recrystn

- R % solvent MP, "C Formula"

A. 9-Pentofuranosyladenine 1-oxides (5b-e) 233-235 1 $1 3N 5'3

218-220 CloHl,N504. 0.25H20

173-175 C,,Hl,N505* 0.4H.20

p-D-Ribofuranosylb 30 HZO 2-Deoxy-p-D-ribofuranosylC 64 H2O

p- D-Xylofuranosyld 74 HZO 2 20-2 2 2 1 OH 1 3N50$ p-D- Arabinofuranosyld 67 95%EtOH

B. 5-Amino-N-benzyloxy-l-pentofuranosylimidaole-4-carboxamidines ( 12b-e) p- D-Ribofuranosyled 97 H,Og 146-150h C16H21N505

2-Deoxy-p-D-ribo furanosy li 31 MeOHi C16H21N504

p-D- Arabinofuranosyle 100 H,Og 175-177h C 1 6 H 2 1 N ~ 0 5

p-D-Xylofuranosyle 54 H,Og 166-16gh 1 6H?,1N SO5

p-D-Ribofuranosyle 48 H,Og 192-194'' 'QH1 5N504

2-Deoxy-p-D-ribofuranosyle 25 H,Og 150h C9H I 5N 5'3

p-D- Arabinofuranosyl' 57 MeOHl C9H 1 SN 5'4

p-D-Xylofuranosyli 33 H,Oi 'gH1 5N504

C. 5-Amino- 1-pentofuranosylimidazole-4-carboxamidines (9b-e)

D. 9-Pentofuranosyl-2-azaadenines (13b-e) C9H 1 Z N 6 0 ,

203-204 C9H12N603

p-D- Arabinofuranosyl 43 H ,Ok 236-238 'gH lZN6O4

240-241 p-D-Ribofuranosyl 64 HZO 2-Deoxy-p-D-nbofuranosyl 31 H,Ok

C9H12N604 - p-D-Xylofuranosyl 28 H20k 234-235

UAnalyzed for C, H, N. bSee Ref 13. CH. Klenow and S. Frederiksen,Biochim. Biophys. Acta, 52, 384 (1961). dE. J. Reist, D. F. Calkins, and L. Goodman, J. Med. Chem., 10, 130 (1967). eAnalytica1 sample of picrate obtd. fAlso prepd by deformylation in MeOH-NH, at 78" of N-benzyloxy-5-formamido-l-~-D-ribofuranosylimidazole-4-carboxamidine. gRecrystn of picrate. hMp of picrate. fAnalytica1 sample was not obtd. iDid not cryst. Sample was obtd by evapn of a s o h of the compd in this solvent. kIsolated by chromatog on a thick silica gel plate before recrystn.

Table 111. 4-Amino-7-p-D-ribofuranosyl-7H-pyrazolo[3,4il]-v-triazine (21a) Crude yield, Recrystn

Compound % solvent MP, "C Formula"

4-Amino-l-~-D-ribofuranosylpyrazolo[ 3,4-d]pyrimidine 5- oxide 74 HZO 233-234 C10H13N505 5-Amino-N-benzyloxy-l-p-D- ribofuranosylpyrazole-4- carboxamidine 69 MeOH-H,O 190-1 9 1 C16H,,N,0,* 0.25H.20

5-Amino- 1-p-D-ribofuranosyl-

4-Amino-7-P-D-ribofuranosyl- p yrazole-4-carboxamidineb 60 H,OC 17 5- 17 6d C9H 1 SN5'4

pyrazolo [3,4-d]-v-triazine 32 H P 207-208 C9H12N604 __ "Anaiyzed for C, H, N. bAnalyzed as the picrate. CRecrystn of picrate. dMp of picrate. eIsolate as the picrate. Free base obtd by treat-

ment of the picrate with Dowex 1-X8 (carbonate) ion-exchange resin.

2-Azapurine Nucleosides Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 185

Table 1V. 2.8-Diazaadenosine (21b)

Compound

Crude yield, Recr y ste

% solvent MP, 'C Formulas

9Q-D-Ribofuranosyl-8-aza-

5-Amino-N-benzyloxy-1-p-D-ribofuranosyl-l,2,3-triazole-4-carboxamidine 54 MeOH 138-140 CI6H,,,N,O, * 0.5H20

adenine 1-oxide 78 MeOH 208-210 'gH 1 a N 6 0 5

5-Amino-l$-D-ribofuranosyl-l,2,3-

2,8-Diazaadenosine 7.5 H,Oe

before recrystn.

1 8 1-1 8 2d C8H14N604

193-195 C8H11N,04 triazole-4-carboxamidineb 40 H P

UAnalyzed for C, H, N. bAnalyzed at the picrate. CRecrystn of picrate. dMp of picrate. eIsolated by chromatog on a thick silica gel plate

spectrophotometer. Ir spectra were detd in pressed KBr discs with Perkin-Elmer Models 221-G, 521, and 621 spectrophotometers, pmr spectra in DMSO-d, (TMS) with a Varian A-60A spectrometer; chemical shifts quoted in the case of multiplets are measured from the approximate center. Chromatog analyses were carried out on tlc plates of silica gel H (Brinkmann). The spots were detected by uv light after spraying the plates with Ultraphor (WT, highly concentrated). Most of the chromatog purifications were carried out on Mallinckrodf SilicAR-7 with the solvents indicated. The analytical samples were dried over P,O, (0.07 mm) for 16-20 hr a t the temps given.

9-Cyclopentyladenine (la). A. A soln of 5-amino-1-cyclo- pentylimidazole4carboxamidine (9) . 1.4HC1(64 mg, 0.26 mmole) in 98% formic acid (2 ml) was refluxed for 4 hr and then evapd to dryness. The pH of a soh of the residue in H,O (15 ml) was raised to 10 with concd NH40H before it was extd with CHCI, (2 X 20 ml). The CHC1, ext was dried (MgSO,) and evapd to dryness in vacuo. The residue crystd from C,H, as a white solid: yield, 30 mg (57%). This material was identical in all respects (mp, ir and uv spectra) with an authentic sample of 9-cyclopentyladenine."

B. A soln of l-cyclopentyl-5-formamido-N-benzyloxyimidaz- ole-4-carboxamidine (11,327 mg, 1.0 mmole) in abs EtOH contg sponge N i catalyst (50 mg) was hydrogenated at atm pressure for 24 hr. The catalyst was removed, fresh catalyst (50 mg) was added, and hydrogenation was resumed for another 24 hr. The process was repeated. The catalyst was then removed, and the soln was evapd to dryness. The residue crystd from MeOH as a white solid: yield, 68 mg (31%). The material was identical with an authentic sample of 9-cyclopentyladenine .*

9-cyclopentyladenine (203 mg, 1.0 mmole) and benzyl bromide (342 mg, 2.0 mmoles) in DMA (20 ml) was heated for 20 hr at 110" and evapd to dryness in vacuo. After trituration with ether, the residue crystd from EtOH: yield, 241 mg (66%).

The analytical sample was obtd by recrystn from EtOH, and dried at 78": mp 256-257"; hmax nm (E X lo-,) 0.1 N HCl, 263 (13.1); pH 7 , 262 (13.2); 0.1 N NaOH, 261 (13.5), 268 (sh) (12.11, 298 (sh) (3.52); omax (cm-') 3300-2600 (br) (NH+, CH); 1685 ( G N ) ; 1630, 1575, 1510 (purine ring stretch); 1495,730,690 (C,H,). Anal. (C,,H,,N,.HBr) C, H, N.

N-Benzyl-9-cyclopentyladenine (3a). A. A soln of 2a (150 mg, 0.4 mmole) in 95% EtOH (20 ml) and 1 N NaOH (2 ml) was refluxed for 3 hr, neutralized with 1 N HC1, and evapd to dryness in vacuo. The combined CHCl, ext of the residue was dried (MgSO,) and evapd to dryness in vacuo. The residue crystd from EtOH as a white solid: yield, 60 mg (53%); mp 102-103"; Amax nm (E X lo-')

(cm-') 3265 (NH); 3125,3075,3050,3020 (arom CH); 2950,2865 (aliph CH); 1625, 1580, 1530, 1490, 1485 (purine and C,H, ring stretch); 730,695 (C,H,). Anal. (C,,H,,N,) C, H, N.

B. A soln of 2a (374 mg, 1.0 mmole) in EtOH (75 ml) andH,O (100 ml) was neutralized with 1 ml of 1 N NaOH, refluxed for 24 hr, and evapd to dryness in vacuo. The residue crystd from aq EtOH (80%): yield, 246 mg (87%). This material was identical in all respects with that obtd from reaction A.

9-Cyclopentyl-2-mercaptoadenine (4a). A soln of 9.1.4HC1 (230 mg, 0.94 mmole) in DMF (25 ml) contg a suspension of anhyd K,CO, (276 mg, 2.0 mmoles) was stirred for 10 min before the addn of CS, (5 ml), after which it was stirred for 20 hr, and then poured into H,O (150 ml). Evapn of the aq mixt gave a residue that crystd from 20% aq MeOH as a yellow solid: yield, 100 mg (45%). The analytical sample from DMF was dried at 100": mp 321-323'; hmax nm (E X lo-,) EtOH, 234 (19.3), 273 (10.8); Ymax (cm-l) 3310,3160 (NH,); 2950,2865 (aliph CH); 1640,1580 (purine ring

I-Benzyl-9-cyclopentyyladenine Hydrobromide (2a). A soln of

0.1 N HCl, 268 (19.9); pH 7 ,271 (20.1); 0.1 N NaOH (20.1); BmaX

stretch); 1450 (CH). Anal. (CI,Hl,N5S) C, H, N. 9-Cyclopentyladenine 1-Oxide (Sa). A. A s o h of 9-cyclo-

pentyladenine (2.03 g, 10.0 mmoles) in glacial AcOH (50 ml) and 30% aq H,O, (5 ml) was kept for 10 days with exclusion of light and then evapd to dryness in vacuo below 40". The residue crystd from EtOH (125 ml) as a white solid: yield, 1.70 g (78%). The analytical sample, obtd from a previous run by recrystn from EtOH, was dried at 78'; mp 292' dec; hmax nm (E X lo-'): 0.1 N HCI, 216 f28.6),260 (12.0); pH 7, 233 (43.2), 263 (7.94), 295 (2.11); 0.1 N NaOH, 232 (27.2), 268 (8.45), 305 (3.99): Bmax (cm-'): 3365 3260,3190,3080 (NH); 2950,2870 (aliph CH); 1660,1565,1490 (purine ring stretch); 1225 (N-.O). Anal. (Cl,H,,N50), C, H, N.

B. A soln of 1-cyclopentyl-N-benzyloxy-5-formamidoimidazole-4-carboxamidine (327 mg, 1.0 mmole) in EtOH (150 ml) contg 5% Pd/C (100 mg) was hydrogenated at room temp and atm pressure for 20 hr during which the theoretical uptake of H, was oh- served. The soln was filtered, and the fiitrate was evapd to dryness in vucuo. The residue crystd from EtOH: yield, 144 mg (65%) The uv, ir, and pmr spectra of this material indicate that it is impure Sa.

5-Amino-l-cyclopentyl-4-imidazolecarboxamidoxime (7a). A. A soln of Sa (16.05 g, 73.6 mmoles) in 3 N HCl(322 ml) was refluxed for 10 min and evapd to dryness in vacuo. When a soln of the residue in H,O (200 ml) was neutralized with 1 N NaOH, the product pptd as a white solid: yield, 10.37 g (67.5%). The analytical sample was obtd in a previous run by recrystn from EtOH, and dried at 78": mp 192-194"; Amax nm (E X lo-') 0.1 N HCl. 278 (9.63); pH 7, 257 (9.60); 0.1 N NaOH, 260 (9.50); Bmax (cm-') 3435,3140,3295 (NH); 3140 (arom CH); 2960,2870 (aliph CH); 3300-2200 (br) (OH); 1630 ( G N ) ; 1565,1495 (imidazole ring stretch); 1450 (CHI. Anal. (C&,N,O) C, H, N.

B. A soln of Sa (1.00 g, 4.58 mmoles) in 1 N NaOH (10 ml! was refluxed for 30 min, neutralized with concd HCl, and extd with CHCl, (100 ml). After drying (MgSO,), the ext was evapd to dryness. The residue crystd from EtOH as a white solid: yield, I88 mg (17%). Spectral data indicated that this material is J -cyclopentyl-5- formamidoimidazole-4-carboxamidoxime (6a). From the filtrate, 391 mg (39%) of starting compd was obtained. The product, without purification, was refluxed for 10 min in 3 N HCl(4 ml). The soln was neutralized with dil NaOH and extd with CHCl, (4 X 25 ml). The CHCl, ext was dried (MgSO,) and evapd to dryness. The residue crystd from EtOH as a white solid: yield, 22 mg. Further CHCl, extn of the residue from the aq soln gave another 20 mg: total yield, 37%. This material was identical in ir and uv spectra with an authentic sample of 5-amino-l-cyclopentyl-4-imidazolecarboxamidoxime (7).

5-Amino-l-cyclopentyli1nidazole-4-carboxamidine (9a). A. A soln of 7a (2.09 g, 10.0 mmoles) in MeOH (350 ml) contg Ra Ni catalyst (2 g) was hydrogenated for 24 hr at room temp and atm pressure. The catalyst was removed by filtration and washed with MeOH. The combined fiitrate and washes were evapd to dryness in vacuo. A soln of the residue in MeOH was acidified with concd HCl and evapd to dryness. The residue crystd from EtOH as an HCl salt: yield, 2.18 g (90%). The analytical sample was obtd in a previous run by recrystn from EtOH and dried at 100" (0.07 mm) over P,05: mp 265-266'; hmax nm ( E X lo-') 0.1 N HCI, 283 (11.4); pH 7, 284 (12.1); 0.1 N NaOH, 267 (9.60); Omax (cm-') 3295, 3170 (NH); 2960,2870 (aliph CH); 1560,1645,1575 (C=NH$ and imidazole ring stretch). Anal. (C&,N,. HCl . 0.67HZO) C, H, C1, N. A second crop was obtd from the soln of the analytical sample and dried at 100': hmax nm (E X lo-') 0.1 N HCI, 283 (11.0); pH 7, 284 (11.8); 0.1 N NaOH, 266 (9.45). Anal. (CpH,,N,.HC1.0.44H20) C, H, N.

The identity of the fractional hydrates was further confirmed by conversion to 9-cyclopentyladenine (see above).

B. A soh of 7a .HC1(970 mg, 2.90 mmoles) in H,O (60 ml)

186 Journalof Medicinal Chemistry, 1972, Vol. 15, No. 2 Montgomery and Thomas

was made basic (pH 9) with 1 N NaOH. Repeated CHC1, extn of the resulting cloudy s o h gave a clear aq layer. The combined CHCl, ext was dried (MgSO,) and evapd to dryness in vacuo. The residue, a white cryst solid, was dissolved in EtOH (50 ml) and hydrogenated at room temp and atm pressure for 20 days in the presence of Ra Ni catalyst. The soln was filtered and evapd to dryness. An EtOH s o h of the residue was acidified with concd HCl and evapd to dryness. The residue crystd in 2 crops from EtOH as a white solid (1.4 HCl): yield, 163 mg (23%); the second crop was a 1.2HC1: yield, 143 mg (21%). Both ir and uv data of these two crops are in agreement with those given in A above.

I-Benzyloxy-9-cyclopentyyladenine Hydrobromide (loa). A s o h of 9-cyclopentyladenine 1-oxide (218 mg, 1.0 mmole) in DMA (20 ml) contg PhCH,Br (684 mg, 4.0 mmoles) was kept for 4 days at 25" and evapd to dryness in vacuo without heating. The residue, after 2 triturations with Et,O, crystd from EtOH as a light yellow solid: yield, 303 mg (78%). The analytical sample was obtd by careful recrystn from EtOH. It was dried at 78": mp 177-182"; hmax nm ( E X NaOH, 258 (11.9), 266 (sh) (10.9); omax (cm-') 3400 (OH); 3400- 2400 (br) (NH:); 1680 (C=NH:); 1610, 1495 (C,H,); 1565 (purine ring stretch); 1450 (CH); 740, 690 (C,H,). Anal. (C,,H2,,BrN,0~ 0.5C,H50H) C, H, N.

l-Cyclopentyl-5-formamido-N-benzyloxy~idazole-4-car~x- amidine (1 la). A s o h of 10a (413 mg, 1.0 mmole) in H,O (40 ml) and EtOH (15 ml) was neutralized with 1 N NaOH, kept for 2 days at 25", and evapd in vacuo without heating to remove the alc. The ppt that formed in the remaining aq s o h was collected as a white yolid: yield, 268 mg (82%). The analytical sample was obtd by recrystn from EtOH and dried at 78": mp 132-133"; hmax nm ( E X

0.1 N HCl, 262 (12.3); pH 7, 262 (12.1); 0.1 N

0.1 N HC1, 255 (8.84); pH 7, 255 (sh) (6.92); 0.1 N NaOH, 256 (12.6); YmaX (cm-') 3490,3400,3345,3300 (NH); 3160,3100 (arom CH); 2940,2920,2865 (aliph CH); 1695 (C=O); 1690, 1680 (amide I); 1630 (C=N); 1600, 1490 (C,H,); 1530 (amide 11); 735, 695 (C,H,). Anal. (C,&N,02) C, H, N.

5-Amino-N-(benzyloxy)- l-cyclopentyliidazole-l-carboxmi- dine (12a). A. A soln of l l a (6.54 g, 20.0 mmoles) in methanolic NH, (satd at 0") (800 ml) was heated in a stainless steel reaction vessel a t 80" for 2 days and then evapd to dryness. A s o h of the residue in EtOH was acidified with concd HCl and evapd to dryness. The residue crystd from EtOH. The analytical sample, obtd by recrystn from EtOH, was a white solid: yield, 4.02 g, (60%); mp 179-181" dec; kmax nm ( E X lo-'): 0 . 1 N HC1, 279 (10.0); pH 7, 264 (10.9); 0.1 N NaOH, 264 (10.7); Dmax (cm-') 3450,3320 (NH); 3190,3140,3080,3020,3000 (arom CHI; 2945,2865 (aliph CHI; 3200-2200 (br) (NH;); 1650 (CH=NH$); 1590,1500 (C,H,); 1525, 1460 (imidazole ring stretch); 735,685 (C,H,). Anal. (C,,H,,N,O~ HC1) C, H, N. An 18% recovery of starting material was obtained from the filtrate.

at 0") (200 ml) was heated in a stainless steel reaction vessel at 80" for 2 days and evapd to dryness in vacuo. A soln of the residue in EtOH was acidified with concd HCl and evapd to dryness. Two recrystn from EtOH gave the product as a white solid: yield, 528 mg (79%). This material was identical in all respects with that from the previous run.

2-Aza-9-cyclopentyladenine (13a). To a cold, stirred s o h of 5- amino-1-cyclopentylimidazole4carboxamidine . 1.4HC1 (1.15 g, 4.7 mmoles) in H,O (40 ml) was added slowly a soh of NaNO, (690 mg, 10.0 mmoles) in H,O (15 ml). The ppt that formed was collected after standing 2 hr in the cold: yield, 703 mg (73%). The analytical sample was obtd in a previous run by recrystn from H,O and dried at 78": mp 212-215"; hmax nm ( E X lo-)) 0.1 N HCl, 254 (7.88), 293 (4.27); pH 7, 257 (7.37), 297 (6.20); 0.1 N NaOH,

2870 (aliph CH); 1700, 1645, 1615, 1510 (azaadenine ring stretch). Anal. (CyHl,N,) C, H, N.

N-Benzyloxy-9-cyclopentyladenine (14a). A. A s o h of 1 l a (327 mg, 1.0 mmole) in H,O (60 ml) and EtOH (10 ml) was refluxed for 27 hr. The white solid that pptd on cooling was collected by filtration: yield, 204 mg (66%); mp 169-171'. The analytical sample was obtd by recrystn from EtOH and dried at 78": hmax nm ( E X i o n 0 . 1 N HCI. 274 (14.3); pH 7 ,270 (17.1); 0.1 N NaOH, 287 (13.2): omax (cm-') 3100 (br), 3025 (NH); 2950, 2910, 2860 (aliph CH); 1645, 1525 (purine ring stretch); 1590, 1500, 730,690 (C,H,). Anal. (C,$,,N,O) C, H, N.

B. A s o h of 10a (390 mg, 0.95 mmole) in H,O (60 ml) was neutralized with 1 N NaOH (0.95 ml), dild with EtOH (30 ml) to give a clear s o h , and then refluxed for 8.5 hr. Upon cooling, a cryst solid pptd: yield, 107 mg (37%). This product was identical in all respects with that from the previous run. Evapn of the filtrate to 10

B. A s o h of 10a (827 mg, 2.0 mmoles) in methanolic NH, (satd

257 (7.37), 297 (6.17); Dmax (cm-') 3400-3000 (br) (NH); 2945,

ml caused the pptn of another cryst solid that was identified by its uv and ir spectra as 1cyclopentylJ-formamido-N-(benzyloxy)imidazole-l-carboxamidine: yield. 111 mg (36%).

N-Benzyloxyadenosine (14b). A s o h of 1-benzyloxyadenosine hydrobromide (908 mg, 2.0 mmoles) in H,O (150 ml) to which 2 ml of 1 N NaOH was added was refluxed 9 hr and then evapd to dryness. The product was isolated by chromatog on a silica gel plate as a white glass that crystd from H,O: yield, 118 mg (16%); mp 129- 130"; Amax nm ( E X lo-,) 0.1 N HCl, 269 (17.0); pH 7, 269 (16.9); 0.1 N NaOH, 284 (13.4); vmax (cm-I) 3380 (OH); 2920,2860 (CH); 1650, 1590,1530,1505 (C=N, C=C); 1205 (COC); 1075,1040 (CO); 830,790 (C,H,). Anal. (C,&,N,O,) C, H, N.

Effect of Base on 2-Aza-9-cyclopentyladenine. 5-Amino-1-cyclopentylimidazole-4-carboxamide (15) and 4,5-Diamino-6-cyclopentylamino-v-triazine (16). A s o h of 2-aza-9-cyclopentyladenine (375 mg, 1.84 mmoles) in 0.1 N NaOH (30 ml) was refluxed with stirring for 20 hr, then cooled, and the white cryst solid that pptd was collected by filtration (27 1 mg). Purification of this material was effected by chromatog on a thick plate using CHC1,-MeOH (9: 1) as developing solvent; 3 bands were obtd. Each was extd with hot MeOH. The slower moving band gave a white solid; yield, 125 mg (36%). The analytical sample of this material, identified as 4,5- diaminod-cyclopentylamino-v-triazine (16), was obtd by recrystn from EtOH and dried at 78": mp 275" dec; hmax nm ( E X 0.1 N HCl, 204 (13.8), 252 (18.8), 289 (9.30), 374 (5.72); pH 7, 235 (25.4), 286 (7.06), 318 (4.85); 0.1 N NaOH, 236 (25.7), 286 (7.12), 318 (4.90); Omax (cm-') 3430, 3380, 3340, 3190 (NH); 2950,2860 (aliphatic CH); 1680, 1610, 1575 (ring stretch); 6 in ppm, 1.7 (m, CH,); 4.4 (m, CH); 5.0,5.7 (m, NH and NH,). Anal. (C.H,.NJ C, H, N. - ._ "

The middle band also gave a white solid: yield, 145 mg (41%). It was identified by spectral and analytical data as 5-amino-1-cy- clopentylimidazole4carboxamide ( 15).y

as 9-cyclopentyl-2-azaadenine (13).

amino-7-0-D-ribofuranosylpyrazolo [ 3 ,I-d]-v-triazine (2 la), and 2,8-diazaadenosine (21b) were all prepd in essentially the same manner as 9-cyclopentyl-2-azaadenine ( l a + 5a + 10a + 12a + 9a +

13a). The details for each compd are given in Tables 11-IV.

C. Coburn, Jr., and members of the Molecular Spectroscopy Section of Southern Research Institute for the spectral and most of the microanalytical data reported, to Mrs. Martha Thorpe for her help in the interpretation of the pmr spectra, to Mrs. Margaret H. Vail for the cytotoxicity data reported, and to Dr. W. R. Laster for the leukemia L1210 screening data.

The fast moving band gave a solid (6 mg, 1%) that was identified

Nucleosides. The 9-pentofuranosyl-2-azapuMes (1 3b-e), 4-

Acknowledgment. The authors are indebted to Dr. W.

References (1) J. A. Montgomery, h o g . Med. Chem., 7,69 (1970). (2) J. Davoll, J. Chem. SOC., 1593 (1958). (3) W. W. Lee, A. P. Martinez, G. L. Tong, and L. Goodman, Chem.

(4) G. L. Tong, W. W. Lee, L. Goodman, and S. Frederikson, Arch.

(5) J. A. Montgomery and H. J. Thomas, J. Org. Chem., 36, 1962

(6) J. A. Montgomery, F. M. Schabel, Jr., and H. E. Skipper, Can-

(7) L. L. Bennett, Jr., M. H. Vail, S. Chumley, and J. A. Mont-

(8) J. A. Montgomery and H. J. Thomas, Chem. Commun., 458

(9) H. J. Thomas and J . A. Montgomery, Abstracts, 19th South-

Ind. (London), 2007 (1963).

Biochem. Biophys., 112,76 (1965).

(1971).

cerRes., 22,504 (1962).

gomery, Biochem. Pharmacol., 15,1719 (1966).

(1969).

east Regional Meeting of the American Chemical Society, At- lanta, Georgia, Abstr 74 (1967).

(10) E. C. Taylor and P. K. Loeffler, J. Amer. Chem. SOC., 82, 3147 (1960).

(11) N. J. Leonard, S. Achmatowicz, R. N. Loeppky, K. L. Carra- way, W. A. H. Grimm, A. Szweykowska, H. Q. Hamzi, and F. Skoog, Proc. Nat. Acad. Sci. U. S., 56, 709 (1966).

(12) J. A. Montgomery and H. J. Thomas, J. Heterocycl. Chem., 1, 115 (1964).

(13) M. A. Stevens, D. I. Magrath, H. W. Smith, and G. 8. Brown, J. Amer. Chem. SOC., 80,2755 (1958).

(14) M. A. Stevens and G. B. Brown, ibid., 80,2759 (1958).

Nb-Substituted Adenosines Journal ofMedicina1 Chemistry, 1972, Vol. 1.5, No. 2 187

(15) C. Temple, Jr., C. L. Kussner, and J. A. Montgomery, J. Org.

(16) M. A. Stevens, H. W. Smith, and G. B. Brown, J. Amer. Chem.

(17) T. Fujii, C. C. Wu, T. Itaya, and S . Yamada, Chem. Znd. (Lon-

(18) J. A. Montgomery and H. J. Thomas, J. Org. Chem., 28,2304

(1963).

J. Heterocycl. Chem., 1,215 (1964).

215 (1970).

80,409 (1958).

Chem., 32,2241 (1967).

SOC., 82, 3189 (1960).

don), 1598, 1967 (1968).

(19) J. A. Montgomery, S. J. Clayton, and W. E. Fitzgibbon, Jr.,

(20) J. A. Montgomery, H. J . Thomas, and S . J. Clayton, ibid., 7 ,

(21) J. A. Montgomery and C. Temple, Jr., J. Amer. Chem. Soc.,

N6-Substituted Adenosines: Synthesis, Biological Activity, and Some Structure-Activi ty Relationships

M. H. Fleysher Department o f Experimental Therapeutics, Roswell Park Memorial Institute, Buffalo, New York 14203. Received July I 7, I 9 71

Nucleosides of N6-substituted adenines, which possess cytokinin activitz, inhibit the growth of tumor cells, while the corresponding adenines are relatively inactive. Addnl N -substituted adenosines have been prepd and tested to secure information regarding structure-activity relationships, if any. The new compds include N6-butyl-, N6-n-2-propoxyethy1, N6-n-2-butoxyethyl-, N6-cyclohexyl-, N6-cyclopropylmethyl-, N6-tetrahydrofurfuryl-, N6-geranyl-, N6-farnesyl-, and @-a-pyridoxyladenosine. They were prepd from 6-chloropurine riboside by nucleophilic substitution with the appropriate amine. The known cytokinin compds, 2-methylthio-N6-isopentenyladenine, cis-6<~-chloro-2-butenylamino)purine, and trans-6-(~-chloro-2-butenylamine)purine and their ribosides, and trans-zeatin riboside were examd for other biological activity. The alkylated adenosines show optimal cytokinin activity when the N6- substituent contains a double bond. In Escherichia coli the compds were active at 10-6-10-4M with the trans isomers showing greater activity than the cis compds. As inhibitors of mouse adenocarcinoma cells (TA-3) in culture, some of the compds were active at 10-5-10-6M but in sarcoma $180 cells in culture they were all less active. Redn of the double bond in the side chain lowered activity of these compds in the tumor cell cultures. The trans isomers are more active against tumor cells in vitro than the cis analogs, paralleling their activity as cytokinins. The presence of an OH group in the side chain diminished antitumor activity. A moderate increase in survival time of mice bearing leukemia L-1210 was produced by the compds bearing an ether linkage in the side chain.

We have previously reported',' the synthesis of a series of N6 -substituted adenine ribosides which are potent cytokinins. Many of these compds inhibited the growth of neoplastic cells in vitro at concns of about 1 OW6 M. Most of these adenosine analogs had different effects on various leukemic cells and no effects on lymphocytes in vitro. It was also noted that at lower concns (e.g, 10-8-10-7M) some stimulation of human leukemic (line 6410) cell growth took place in contrast to the inhibitory effects that occurred at higher concns. The corresponding free adenines were relatively inactive against tumor cells. The more active tumor inhibitory nucleosides were found to be N6- benzyl-, N6-furfuryl-, N64hoxyethyl-, N6-phenyl-, and N6- thenyladenosines. Like N6-(3-methyl-2-butenyl)adeno~ine'~ (IPA), these analogs are also active as cytokinins.

The present communication? is an extension of our previous work. In order to secure further information on the structure-activity relationships in this series of adenosine derivatives, addnl analogs were prepd and examd for their biol properties. These include N6-n-Bu- (I), N6-n-2-propoxyethyl- (11), N6-n-2-butoxyethyl- (111), N6-cyclohexyl- (IV), fl-cyclopropylmethyl- (V), N6-tetrahydrofurfuryl- (VI), N6geranyl- (VII), N6-farnesyl- (VIII), and N6-cu4-pyridoxyl- adenosines (IX). The compds were tested for cytokinin activity in the tobacco callus bioassay, and in microbial and tumor systems. This series of compds was augmented by truns-N64-hydroxymethyl-2-butenyladenosine (zeatin riboside), a potent cytokinin that was prepd in this work by the method of Shaw, e t aL4 The compds were chosen for synthesis and examn of properties for the following reasons.

t A portion of this material was presented by the author at the 160th National Meeting of the American Chemical S ~ c i e t y . ~

(I) The n-Bu fragment represents the shortest C chain among a series of N'mbstituted adenines resulting in really good cytokinin a ~ t i v i t y . ~ (11,111) The propoxyethyl and butoxyethyl compds are homologs of N6-2-ethoxyethyladenosine, a compd with antitumor activity in vivo.' (IV, VI) The cyclohexyl and tetrahydrofurfuryl compds are satd derivatives of unsatd analogs which showed biol and antitumor activity. (V) The cyclopropylmethyl derivative was prepd to examine the biol effect of the smallest satd ring structure. (VII, VIII) The geranyl and famesyl side chains were chosen to det the effects of multiple isopentenyl fragments on a single side chain. (IX) The cu4-pyri- doxy1 side chain was included to evaluate the effect of a strong hydrophlic substituent which contains a biol active moiety.

Zeatin riboside (trans-N6-(4-hydroxymethyl-2-buteny1)- adenosine was included because it is about the most active known cytokinin of the N6-adenosine ~ e r i e s . ~ , ~ The compds were prepd by condensing 6-chloropurine riboside (6-chloro- g-P-D-ribofuranosyl-9H-purine) with the corresponding amines by nucleophilic substitution in boiling EtOH, using CaC03 or E t a as ancillary acid acceptors.' The compds were isolated and purified by crystn. The purity of the products was confirmed by chromatog, elemental analyses, and by uv spectra. The physical data are given in Tables I and 11.

The new compds were examd for cytokinin activity in a tobacco pith assay system by the method of Murashige and Skoog.8 As shown in Table 111, the compds vary in activity. The N6-butyl and propoxyethyladenosines have good activity, although somewhat less than that of zeatin riboside. The other compds displayed marginal or no cytokinin activity noted since high concns had to be used for initial response.


N6-Substituted adenosines. Synthesis, biologicalactivity, and some structure-activity relations

M. H. FleysherJ. Med. Chem., 1972, 15 (2), 187-191• DOI: 10.1021/jm00272a015 • Publication Date (Web): 01 May 2002






Nb-Substituted Adenosines Journal ofMedicina1 Chemistry, 1972, Vol. 1.5, No. 2 187

(15) C. Temple, Jr., C. L. Kussner, and J. A. Montgomery, J. Org.

(16) M. A. Stevens, H. W. Smith, and G. B. Brown, J. Amer. Chem.

(17) T. Fujii, C. C. Wu, T. Itaya, and S . Yamada, Chem. Znd. (Lon-

(18) J. A. Montgomery and H. J. Thomas, J. Org. Chem., 28,2304

(1963).

J. Heterocycl. Chem., 1,215 (1964).

215 (1970).

80,409 (1958).

Chem., 32,2241 (1967).

SOC., 82, 3189 (1960).

don), 1598, 1967 (1968).

(19) J. A. Montgomery, S. J. Clayton, and W. E. Fitzgibbon, Jr.,

(20) J. A. Montgomery, H. J . Thomas, and S . J. Clayton, ibid., 7 ,

(21) J. A. Montgomery and C. Temple, Jr., J. Amer. Chem. Soc.,

N6-Substituted Adenosines: Synthesis, Biological Activity, and Some Structure-Activi ty Relationships

M. H. Fleysher Department o f Experimental Therapeutics, Roswell Park Memorial Institute, Buffalo, New York 14203. Received July I 7, I 9 71

Nucleosides of N6-substituted adenines, which possess cytokinin activitz, inhibit the growth of tumor cells, while the corresponding adenines are relatively inactive. Addnl N -substituted adenosines have been prepd and tested to secure information regarding structure-activity relationships, if any. The new compds include N6-butyl-, N6-n-2-propoxyethy1, N6-n-2-butoxyethyl-, N6-cyclohexyl-, N6-cyclopropylmethyl-, N6-tetrahydrofurfuryl-, N6-geranyl-, N6-farnesyl-, and @-a-pyridoxyladenosine. They were prepd from 6-chloropurine riboside by nucleophilic substitution with the appropriate amine. The known cytokinin compds, 2-methylthio-N6-isopentenyladenine, cis-6<~-chloro-2-butenylamino)purine, and trans-6-(~-chloro-2-butenylamine)purine and their ribosides, and trans-zeatin riboside were examd for other biological activity. The alkylated adenosines show optimal cytokinin activity when the N6- substituent contains a double bond. In Escherichia coli the compds were active at 10-6-10-4M with the trans isomers showing greater activity than the cis compds. As inhibitors of mouse adenocarcinoma cells (TA-3) in culture, some of the compds were active at 10-5-10-6M but in sarcoma $180 cells in culture they were all less active. Redn of the double bond in the side chain lowered activity of these compds in the tumor cell cultures. The trans isomers are more active against tumor cells in vitro than the cis analogs, paralleling their activity as cytokinins. The presence of an OH group in the side chain diminished antitumor activity. A moderate increase in survival time of mice bearing leukemia L-1210 was produced by the compds bearing an ether linkage in the side chain.

We have previously reported',' the synthesis of a series of N6 -substituted adenine ribosides which are potent cytokinins. Many of these compds inhibited the growth of neoplastic cells in vitro at concns of about 1 OW6 M. Most of these adenosine analogs had different effects on various leukemic cells and no effects on lymphocytes in vitro. It was also noted that at lower concns (e.g, 10-8-10-7M) some stimulation of human leukemic (line 6410) cell growth took place in contrast to the inhibitory effects that occurred at higher concns. The corresponding free adenines were relatively inactive against tumor cells. The more active tumor inhibitory nucleosides were found to be N6- benzyl-, N6-furfuryl-, N64hoxyethyl-, N6-phenyl-, and N6- thenyladenosines. Like N6-(3-methyl-2-butenyl)adeno~ine'~ (IPA), these analogs are also active as cytokinins.

The present communication? is an extension of our previous work. In order to secure further information on the structure-activity relationships in this series of adenosine derivatives, addnl analogs were prepd and examd for their biol properties. These include N6-n-Bu- (I), N6-n-2-propoxyethyl- (11), N6-n-2-butoxyethyl- (111), N6-cyclohexyl- (IV), fl-cyclopropylmethyl- (V), N6-tetrahydrofurfuryl- (VI), N6geranyl- (VII), N6-farnesyl- (VIII), and N6-cu4-pyridoxyl- adenosines (IX). The compds were tested for cytokinin activity in the tobacco callus bioassay, and in microbial and tumor systems. This series of compds was augmented by truns-N64-hydroxymethyl-2-butenyladenosine (zeatin riboside), a potent cytokinin that was prepd in this work by the method of Shaw, e t aL4 The compds were chosen for synthesis and examn of properties for the following reasons.

t A portion of this material was presented by the author at the 160th National Meeting of the American Chemical S ~ c i e t y . ~

(I) The n-Bu fragment represents the shortest C chain among a series of N'mbstituted adenines resulting in really good cytokinin a ~ t i v i t y . ~ (11,111) The propoxyethyl and butoxyethyl compds are homologs of N6-2-ethoxyethyladenosine, a compd with antitumor activity in vivo.' (IV, VI) The cyclohexyl and tetrahydrofurfuryl compds are satd derivatives of unsatd analogs which showed biol and antitumor activity. (V) The cyclopropylmethyl derivative was prepd to examine the biol effect of the smallest satd ring structure. (VII, VIII) The geranyl and famesyl side chains were chosen to det the effects of multiple isopentenyl fragments on a single side chain. (IX) The cu4-pyri- doxy1 side chain was included to evaluate the effect of a strong hydrophlic substituent which contains a biol active moiety.

Zeatin riboside (trans-N6-(4-hydroxymethyl-2-buteny1)- adenosine was included because it is about the most active known cytokinin of the N6-adenosine ~ e r i e s . ~ , ~ The compds were prepd by condensing 6-chloropurine riboside (6-chloro- g-P-D-ribofuranosyl-9H-purine) with the corresponding amines by nucleophilic substitution in boiling EtOH, using CaC03 or E t a as ancillary acid acceptors.' The compds were isolated and purified by crystn. The purity of the products was confirmed by chromatog, elemental analyses, and by uv spectra. The physical data are given in Tables I and 11.

The new compds were examd for cytokinin activity in a tobacco pith assay system by the method of Murashige and Skoog.8 As shown in Table 111, the compds vary in activity. The N6-butyl and propoxyethyladenosines have good activity, although somewhat less than that of zeatin riboside. The other compds displayed marginal or no cytokinin activity noted since high concns had to be used for initial response.

188 Journalof Medicinal Chemistry, 1972, Vol. 15, No. 2 Fleysher

Table I. Paper Chromatography and R f Values of N'-Substituted Adenosines

Solvent sys te f i Compound A B C D E

6-Chloropurineriboside 0.53 0.68 0.72 0.57 0.41 N6-n-Butyladenosine 0.74 0.80 0.87 0.72 0.67 N 6-n-2-Propoxyethyladenosine 0.72 0.81 0.83 0.63 N 6-n-2-Butoxyethyladenosine 0.77 0.86 0.86 0.79 0.72 N 6-Cyclohexyladenosine 0.81 0.87 0.83 0.72 0.70 N 6Cyclopropylmethyladenosine 0.66 0.81 0.76 0.60 0.59 N 6-Tetrahydrofurfuryladenosine 0.51 0.72 0.80 0.67 0.48 N 6-Geranyladenosine 0.90 0.89 0.89 0.79 0.79 N 6-Farnesyladenosine 0.89 0.86 0.90 0.85 0.83 N 6a4-Pyridoxyladenosine 0.29 0.52 0.73 0.6 3 N6-(4-Hydroxy-3-methyl-trans-butenyl)- 0.46 0.72 0.68

adenosine (trans-zeatin riboside)b uThe solvent systems used for descending chromatog (Whatman

No. 1 paper) (measured by vol): A, EtOAc-n-PrOH-H 0 (4: 1:2) (upper phase); B, i-PrOH-H 0-NH,OH (7:2: 1); C, i-PrbH-1% aq (NH BuO~)H~ACOH-H,O (100032.5:300) (cloudy upper phase). b e e reference 4.

SO, (2: l), D, i-PrOHkoncd Ha-H,O (680:170: 144). E, n-

adenines and adenosines were investigated by Hecht, et al. lo They concluded that planar side chains, i.e., trans isomers impart the highest order of cytokinin activity; whereas the cis isomers, possessing more steric bulk, were less active. Because of these cytokinin variations some of these compds were tested in this work in addn to the new ones prepd here. $

The activities of all the compds discussed as inhibitors of the growth of E. coli are summarized in Table IV. The most active inhibitors were N6-2-propoxyethyladenosine and trans-zeatin riboside, followed by the n-butyl-, n-2-butoxyethyl-, cyclohexyl-, isopentenyl-, and tetrahy- drofurfuryladenosines. The a4-pyridoxyl-, farnesyl-, geranyl-, and cyclopropylmethyladenosines were inactive.

The low water solubility shown by N6-isopentenyl-2- thiomethyl compds possibly accounts for their minimal activity in this system. The N6-3-chloro-2-butenyladenosines and adenines are less active than the N'4sopentenyl analogs but their trans isomers are distinctly more active than the

Table 11. Uv Absorption Spectra of N6-Substituted Adenosines pH 1.0 pH 7.0 pH 12.0

Compound A max, mp e x lo-' h max, mp E X 10-3 h max, mp E x 10-3

N6-n-Butyladenosine N6-n- 2-propox yethyladenosine N6-n-2-Butoxyethyladenosine N W y clohex yladenosine "Cy clopropylmethyladenosine N6-Tetrahydrofurfuryladenosine N6-Geranyladenosine N6-Farnesyladenosine N6-(r4-Pyridoxyladenosine

26 3 26 3 263 265 263 26 5 266 266 274

19.3 18.3 16.4 19.4 18.6 17.9 19.1 17.7 20.3

Table 111. Cytokinin Activity of the New N6-Substituted Adenosines on Tobacco Bioassay4

Relative growth compared to control as 1.00 at

concentrations

Compound 10 ve/l 25 ~ d l 200 udl N h B u t y ladenosine N6-n-2 Propoxyethyladenosine N6-n-2 Butoxyethyladenosine N W y clohexyladenosine N6Cyclopropylmethyladenosine N6-Tetrahydrofurfuryladenosine N6-Geranyladenosine N6-Farnesyladenosine N 6 e 4 Pyridoxyladenosine N6-(4-Hydroxy-3 methyl-trans-buteny1)-

usee reference 8. bSee reference 4. adenosine (zeatin riboside)*

2.17 3.80 2.90 1.73 7.24 7.48 1.00 1.87 4.00 1.00 1.00 2.41 1.28 1.40 5.13 1.13 1.20 4.84 1.41 3.12 1.00 1.59 2.72 1.05 1.07 1.52

2.95 5.62 4.72

The results show the effects of N6 side chain satn and of very low water solubility (geranyl- and famesyladenosines). The N6-a4-pyridoxyladenosine is inactive.

An @-adenosine analog isolated from sol RNA of E. coli, wheat germ, and yeast, having cytokinin activity has recently been described by Leonard and his coworkers' and identified as N6-isopentenyl-2-methylthioadenosine. This nucleoside and other 2-substituted N6-isopentenyl and zeatin ribosides have been synthesized and examd for cytokinin activity! While 2-methylthio, 2-amino, and 2-chloro substituents in the original compd had very little influence on cytokinin effectiveness, the 2-hydroxy substituents greatly lowered cytokinin activity6 showing that an OH group in the structure has marked effects. In another study, the effects of side-chain planarity in a series of "hbs t i tu ted

267, 210 267,210 267, 210 268, 211 267,210 267,210 268 26 9 268,325

18.0, 19.1 18.0, 19.4 16.5, 17.4 18.2, 17.7 18.0, 19.8 18.0, 18.1 19.0 17.3 20.9, 4.7

Table IV. Growth Inhibition in E. coli

267 267 267 268 26 7 26 7 26 8 269 270,309

18.4 18.0 16.7 18.6 18.2 18.1 18.8 17.3 20.6, 9.5

Molar concn for 50% N6€ompound growth inhibition

p-n-But y ladenosine 5 x ~-n-2-Propoxyethyladenosine 5 x N6-n-2-Butoxyethyladenosie 5 x N'-Cyclohex yladenosine 4 x 10-5 fl-Cyclopropylmethyladenosine None N'-Tetrahydrofurfuryladenosine 8 X N'Geranyladenosine 8 X (in suspension) N6-Farnesyladenosine 3 x (in suspension) N"-a4-Pyridoxyladenosine N'-trans-Zeatin riboside 5 x N6-Isopentenyl-2-thiomethyladeno~ine~ 5 X (in suspension)

N6-3Chloro-trans-2-butenyladenosine* 7 X Nb-Isopentenyl-2-thi~methyladenine~ 5 X (in suspension) N6-3Chloro-cis-2-butenyladenineb None P-3-Chloro-frans-2-butenyladenineb 8 X

1 x 10-3

N6-3-Chloro4s-2-butenyladenosine* 1 x lo-%

"-1 sopen tenyladenine4d I x 10-3 N6-IsopentenyladenosineCtd 9 x 10-5

%ee reference 9. *See reference 10. CSee reference 11. dSee ref- ence 12.

cis compds paralleling their cytokinin activity. The adenines are less active than the corresponding ribosides, a fact which has been obsd previously.2

Against sarcoma S-180 cells (Table V) most of the new compds other than N6-isopentenyladenosine were inactive paralleling their cytokinin effects. Zeatin riboside is not active in this system, however, this effect may be due to the presence of an OH group in the side chain causing loss of antitumor activity. Pyridoxyladenosine is inactive, perhaps for the same reason. The propoxyethyl- and butoxy-

~~~~

$We are indebted to Dr. S. M. Hecht and Professor N. J. Leonard for generous gifts of these materials.

N 6-Su bstituted Adenosines JournalofMedicinaI Chemistry, 1972, Val. I S , No. 2 189

Table V. Growth Inhibitory Activity against Sarcoma 180 Cells and Carcinoma TA-3 Cells in Vitro Molar concentration for 50% growth inhibition

Compound Sarcoma 180 Cells TA-3 Cells

N6-n-Butyladenosine > 10-4 >10-4 N6-n-2-propox yethyladenosine None at low4 >10-4 N6-n-2-Butoxyethyladenosine None at >io-' N6Cy clohexyladenosine None at 2.3 x 1 0 - ~

N6-Tetrahydrofurfuryladenosine None at lo-* >io-'

N6-d Pyridoxyladenosine None at 7.5 x N6-trans-~4-Hydroxymethyl-2-butenyladenosine (zeatin riboside)

N6-3 Chlorosis-2-butenyladenosineb 6 X 3.3 x

N6€yclopropylmethyladenosine 9 x 10-5 2.5 X lo-'

N6-(kranyladenosine None (suspension) None at lo-' (suspension) N6-Farnesyladenosine None (suspension) None at (suspension)

N6-Isopentenyl-2-methylthioadenosine' None at 2 X lo- ' (suspension) Slight at IO-'

N6-3 chlorotrans-2-butenyladenosineb 1.2 x 10-5 2.5 X N6-Isopen tenyl-2-thi~methy ladenine~ None at (suspension) None at lo-' (suspension) N6-3 Chlorosis-2-butenyladenineb None at

1 x 1 0 - ~ 1.2 x 10"

N6-3 Chloro-trans-2-butenyladenineb None at 7.5 x 10-5 N6-Isopentenyladenosinecd (IPA) 1.25 X lo-' 7 x N6-IsopentenyladenineGd 2.5 x 10-4 1.9 X lov4

a-dSee footnotes a-d, Table IV.

-- Table VI. In Vivo Activity against Mouse Leukemia L-1210 No. of Mean life % increased

Compound mice span, life-span

Control 39 7.4 100 n-Bu 5 75 9.0 122

8 100 10.0 135 3 200 8.3 112

n-2-Propoxethyl 8 100 9.3 125 11 200 9.4 128 5 300 3.8 Toxic

n-2-butox yethy 8 100 8.6 116 8 200 6.5 Toxic

Cyclohexyl 6 5 5.3 Toxic 5 2 6.6 Inactive

Cyclopropylmethyl 5 75 9.2 124 3 100 8.5 115

Tetrahydro furfuryl 11 100 9.2 125 8 200 9.3 125

Geranyl 6 100 6.9 Inactive Farnesyl 6 100 7.0 Inactive a4-Pyridoxyl 6 100 6.9 Inactive trans-4-Hydroxy-methyl-2-butenyl (zeatin riboside) 11 100 8.0 108 frans-3Chloro-2-butenyl 5 75 6.2 Inactive

8 100 8.6 117 3 200 3.0 Toxic

(N6-adenosine analog) used Dose, mg/kg per day days (over control)

ethyladenosines are inactive in vitro, as was found previously in the case of ethoxyethyladenosine which was active in vivo. The 2-thiomethyl-IPA possibly because of its very low water solubility is inactive in this system, but the cis and trans chlorobutenyl adenosines are active, paralleling their high cytokinin activity. The adenosine nucleosides are far more active than the corresponding bases as obsd before' in this tumor system. Also in this case the trans analogs are more active than their cis isomers paralleling their cytokinin effects. The geranyl- and farnesyladenosines have very low solubility in aq systems which may be the reason why they are inactive. The satd chain compds have lower effectiveness than their unsatd analogs paralleling their cytokinin effects.

The compds were also tested against mouse mammary carcinoma cells (TA-3) in culture, a more sensitive system, and were found to be more active than against sarcoma 180 (Table V). Cyclohexyladenosine and cyclopropylmethyladenosine are slightly active in this system. Zeatin riboside, as in sarcoma 180, is inactive in this tumor system. The isopentenyl analog (IPA)" is more active in this cell line than in sarcoma 180. The sparingly soluble thiomethyl-IPA is

slightly active while the cis and trans chlorobutenyl analogs are quite active, paralleling their cytokinin activity. The corresponding adenine bases are relatively inactive, as previously detd.

The effects of these compds on mice injected with lo6 cells of leukemia L 1210 are shown in Table VI. The compds were given ip daily for 6 days at the dosage indicated. Of the compds tested, N6-butyl (135% increased life- span), N6-2-propoxyethyl (1 28% increased life-span), N6- cyclopropylmethyl(l24% increased life-span), and N 6 - tetrohydrofurfuryl(l25% increased life-span) analogs have the best activities, paralleling their cytokinin activity. The N6-2 butoxyethyl derivative is less effective. It may be mentioned that the ether side chain analogs (N6-alkoxyethyl- adenosines) show far more antitumor activity in vivo than in vitro.8

The cyclohexyl analog proved to be extremely toxic. When the dose was reduced to 2 mg/kg per day the compd was inactive.

trans-Zeatin riboside, a very potent cytokinin and mi-

§See also the case of Nb-2-ethoxyethyladenosine (reference 2).

I90 Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 Fleysher

crobial inhibitor, is virtually inactive in the case of this tumor as in the case of the in vitro tumor cell tests. It appears that the introduction of an OH group in the side chain of IPA sharply lowers antitumor activity. Since it was noted above that 2-hydroxy-N6-IPA had sharply curtailed cytokinin activity vs. that of IPA, hydroxylation may effect an unexpected behavior.

a'-Pyridoxyladenosine contains 2 OH groups in the side chain and is inactive. Geranyl- and farnesyladenosines are too insoluble at levels needed to be active. trans-3-Chloro- 2-butenyladenosine shows slight activity at 100 mg/kg per day and reduction of dosage to 7 5 mg/kg per day renders it inactive.

To summarize it may be said that in Nbsubstituted adenosine analogs with side chains containing 4-7 C atoms: (1) cytokinin activity is optimal when a double bond is present in the side chain; (2) high cytokinin activity is paral- lelled by E. coli growth inhibition; (3) tumor cell growth inhibition follows cytokinin activity with an exception that the presence of an OH group in the side chain diminishes antitumor effects.

Experimental Section Mps were detd on a Mel-Temp mp apparatus and are not cor. Uv

spectra were obtained on a Cary Model 14 recording spectrophotometer. Optical rotation was measured on a Jasco Model ORD-UV5 optical rotatory dispersion recorder. The solvent systems used for descending chromatog are given in Table I. Whatman paper No. l was used. 6-Chloro-9p-D-ribofuranosyl-9H-purine (6-chloropurine riboside) was purchased from K & K Laboratories, Inc., Plainview, N. Y. The amines employed are identified under each prepn. Where analyses are indicated only by symbols of the elements, the analytical results obtained for those elements are within +0.4% of the theoretical values.

N'-n-Butyladenosine (I). To 100 ml of abs EtOH was added 2.00 g (7 mmoles) of 6-chloro-9-~-D-ribofuranosyl-9H-purine (6- chloropurine riboside), 1.4 g (14 mmoles) of CaCO,, and 1.54 g (21 mmoles) of distilled n-BuNH, (Eastman Kodak Co.). The reaction mixt was refluxed with stirring for 18 hr after which chromatog controls indicated no further presence of 6-chloropurine riboside in the reaction. The mixt was filtered hot to remove Ca salts and the product deposited from the filtrate on cooling. The crystals were filtered off and washed with cold EtOH then with Et,O and dried; yield, 1.90 g (84%); mp 176"; [aIzSD -71.4" (c 0.216, 95% EtOH). Anal. (C,,H,,N5~J C, H, N.

added 2.00 g (7 mmoles) of 6-chloropurine riboside, 1.44 g (14 mmoles) of n-2-propoxyethylamine,# and 1.415 g of Et,N (14 mmoles), and the mixt was refluxed 6 hr, when chromatog controls indicated that 6-chloropurine riboside was no longer present in the mixt. The reaction mixt was evapd and redissolved in hot abs EtOH. The product, which crystd on cooling, was filtered, washed with chilled EtOH, and dried: yield, 2.15 g (87%); mp 125"; [cyIzSD --59.5" (c 0.39, 95% EtOH).AnaZ. (C,,H,,N,O,) C, H, N. Use of CaCO, as in the preceding prepn, gave the same results.

N6-n-2-Butoxyethyladenosine (111). A mixt of 100 ml of abs EtOH, 2.00 g (7 mmoles) of 6-chloropurine riboside, 1.40 g (14 mmoles) of CaCO,, and 2.46 g (21 mmoles) of n-2-butoxyethyla- mine* * was refluxed and stirred as above for 18 hr after which the reaction chromatog controls indicated the absence of 6-chloropurine riboside. The hot reaction mixt was filtered through Celite and the product crystd from the filtrate on cooling, The crystals were filtered, washed with cold EtOH, and dried: yield, 2.4 g (93%); mp 126.5'; [aJZ5D --56" (c 0.318,95% EtOH). Anal. (Cl,H,5Ns0,) C, H, N.

N6-Cyclohexyladenosine (IV). The prepn of this compd by the nucleophilic substitution method using either Et,N or CaCO, as the ancillary acid binder, presents some difficulty in isolation of the desired product.

N6-n-2-Propoxyethyladenosine (11). To 100 ml of abs EtOH were

#Supplied by K tk K Laboratories Inc. The material had a bp 1 2 9 O (750 mm) and nZsD 1.4127 comparing with the same lit. values of Harder, er aLI4 The b p w a s detd by Siwoloboff's method.',

**Supplied by K .& K Laboratories, The material had bp 1 5 2 O (750 mm) and nZSD 1.4196 comparing with the same lit. values of IXarder, et al. l 4 boiling point determined by Siwoloboff's method.

A mixt of 100 ml of abs EtOH, 2.00 g (7 mmoles) of 6-chloropurine riboside, 1.39 g (14 mmoles) of cyclohexylamine (nzSD 1.4587, Aldrich Chemical Co.) and 1.4 g (14 mmoles) of CaCO, was refluxed with stirring 18 hr when chromatog controls indicated absence of 6-chloropurine riboside. The Ca salts were removed by filtration of the hot soln. Since no product crystd out on cooling, the filtrate and wash were evapd to about 0.2 of their vol and cooled, whereupon cyclohexylamine hydrochloride pptd in cryst form. It was filtered and dried (0.76 g, mp 208") and analyzed cor- rectly for C,H,'NH; HC1. On adding Et,O to the mother liquor the product crystd out, was filtered, and washed with Et,O. It was recrystd from EtOH-Et,O and dried: yield, 2.06 g (84%); mp 185"; [alzsD -59.0" (c 0.312, EtOH),Anal. (C,,H,,N,O,) C, H. N.

An alternate method of isolation is to take up the evapd filtered material (after removal of cyclohexylamine hydrochloride) in 100 ml of H,O and ext 3 times with 100-ml portions of CHC1, (adding 2 ml of HOAc to the first CHCI, portion to take up any amine). The combined CHCI, fractions were washed with H,O and dried (Na,SO,). The CHC1, was evapd, and the residue was crystd from EtOH-Et,Q, as above, yielding the same results. This compd cannot be made vin the N'-quaternization procedure' since adenosine does not react with iodocyclohexane in DMF nor in N,N-dimethylaceta- mide.

N6-CyclopropyImethyladenosine (V). (a) Cyclopropylmethyla- mine. This amine has been made by Roberts and MazuP by the redn of cyclopropyl cyanide with Na in abs EtOH followed by isolation of the hydrochloride, basifying, ether extn of the amine, and fractionation. The redn of the nitrile could also be effected with LAH in Et,O. Cyclopropyl cyanide?? (25 g, 373 mmoles) was dissolved in 250 ml of anhyd Et,O. The soln was added dropwise to a stirred and cooled suspension of 14.17 g of LAH (373 mmoles) in 200 ml of anhyd Et,O in 3 hr. The suspension was stirred 18 hr allowing the temp to rise to room temp. To the stirred and cooled suspension, cold H,O (26.9 ml, 1.492 moles) was added dropwise, stirring contd for I hr and the mixt allowed to stand for 2 hr. The white suspension was filtered, and the ppt was washed with Et,O. The combined Et,O filtrates were sepd and dried (Na,SO,). The Et,O was evapd, and the amine was collected as the fraction which distd at 83-81": izZ5D 1.4250; yield, 5.50 g (20.7%).

(b) N6-Cyclopropylmethyladenosine (V). To 100 ml of abs EtOH was added 3.73 g (13 mmoles) of 6-chloropurine riboside, 3.90 g (38.7 mmoles) of Et,N, and 2.75 g (38.7 mmoles) of the cy- clopropylmethylamine. The reaction mixt was stirred and refluxed for 17 hr when chromatog controls indicated the absence of 6-chloropurine riboside. On cooling to room temp the product crystd out. It was filtered and washed with cold EtOH followed by Et,O and then dried: yield, 3.84 g(9270); mp 171-172". The material was recrystd from H,O to yield 80% silky needles melting at 176": (CY]~'D -64.5" (C 0.217 95% EtOH). Anal. (C,,H,,N,O,) C, H, N.

W-Tetrahydrofurfuryladenosine (VI). To 100 ml of abs EtOH was added 2.0 g (7 mmoles) of 6-chloropurine riboside, 1.414 g (14 mmoles) of tetrahydrofurfurylamine,$$ and 1.41 g (14 mmoles) of Et,N. The reaction mixt was refluxed with stirring for 4 hr at which time chromatog controls indicated the absence of 6chloropurine riboside. After filtration the material was crystd from EtOH and dried: yield, 1.98 g (81%); mp 148"; [a]2sD --59" (c 0.306 95% EtOH). Anal. C,,H,,N,O,) C, H, N.

N6-Geranyladenosme (VII). To 50 ml of abs EtOH was added 1.0 g (3.5 mmoles) of 6-chloropurine riboside, 1.07 g (7 mmoles) of geranylamine,S§ and 1.41 g (14 mmoles) of Et,N. The reaction mixt was refluxed under N, with stirring 6 hr when chromatog controls showed the absence of 6-chloropurine riboside. It was cooled, and the product was filtered off and washed with chilled EtOH then Et,O and dried. The material was recrystd from MeOH: yield, 1.04 g (74%); mp 108"; [aIZsD --55" (c 0.342, 95% EtOH).Anal. (C,,H,,N,O,) C, H, N. The use of CaCO, as an ancillary condensing agent practically abolished reactivity.

N6-Farnesyladenosine (VIII). To 30 ml of abs EtOH were added 0.644 g (2.25 mmoles) of 6-chloropurine riboside, 0.50 g (2.26 mmoles) of farnesylamine,@ and 0.455 g (4.50 mmoles) of Et,N, and the mixt was refluxed under N, with stirring for 5 hr when chromatog controls showed the absence of 6-chloropurine riboside. The reaction mixt was filtered and evapd to dryness. The solids were crystd from MeOH-H,O (14:1), filtered, and washed with chilled MeOH-H,O and sucked dry. Recrystn from MeOH yielded

_______.___.__I.̂ _.__.- I.- .. ??Supplied by Aldrich Chemical Co., nZ5D 1.4207 $$Supplied b y Aldrich Chemical Co., r l Z S D 1.45'70.

SSThis amine was generously donated by Hoffman-I.aHoche, Inz., through the courtesy of I k . W. E. Scott.

N6-Substituted Adenosines Joumal of Medicinal Chemistry, 1972, Vol. 15, No. 2 191

0.447 g (42%): mp 98"; [aIZ5D -63.1" (c 0.646,95% EtOH). Anal. (C,,H,,NSO,) C, H, N. Use of CaCO, or Li,CO, as ancillary condensing aEntsresulted in poor reactivity.

N6-Pyxidoxyladeno~ine(a4-pyridoxyl-N 6-adenosine) (IX). To 100 ml of abs EtOH was added 2.86 g (10 mmoles) of 6chloropu- rine riboside, 3.00 g (12.5 mmoles) of pyridoxamine dihydrochloride (Sigma Chemical Co.), and 6.06 g (60.0 mmoles) of Et,N. On refluxing a white ppt deposited initially which later dissolved leaving an opalescence in the reaction mass. With further refluxing a white solid deposited out of soh. After 20 hr at reflux temp, the reaction mass was filtered hot, and the ppt was washed with EtOH and dried at room temp: yield, 3.90 g (93.5%); mp 249"; [a]"D -51.6' (c 0.155,35% EtOH). Anal. (C,,H,,O, .H,O) C, H, N.

compds was assayed with the callus derived from the pith from api- cal stems of Wisconsin 38 tobacco plants using the method and media of Murashige and Skoog.' Results are given Table 111.

Microbial Assay Procedure. The microbial assays were carried out according to techniques previously published.I6 E. coli (K-12) was grown in the synth medium of Gray and Tatum.', The compds were added to the medium at lO-'-lO-' M. Results are given in Table IV.

Effects on the Growth of Sarcoma-180 and Carcinoma TA-3 Cells in Vitro. (a) Sarcoma-180 Cells. The compds were added in aq soln (or as a suspension if the solubility was low) to cultures of S-180 cells grown as monolayers in T-15 flasks in Eagle's'' medium containing 5% horse serum. The cells were exposed to the compds for 6 days during which time there were 3 changes of medium. The control cultures increased by 10- to 15-fold. The quantity of cells was estimated by the crystal violet method of Grady, et a1.,I9 and by protein detns." (Table V).

(b) Carcinoma TA-3 Cells. This mammalian cell line (mouse mammary adenocarcinoma TA-3) was grown in RPMI 1640medi- umzl supplemented with 10% horse serum. The test system consisted of tube cultures (12 X 75 mm) inoculated with 5 X lo4 cells in 1 ml of medium, supplemented with another milliliter of medium containing the compd and then incubated at 36" in upright position. All the concns of the compds were tested in 5 tubes each. The medium was replaced with fresh medium on the second day by centrifug- ing the cells for 10 min at 500 rpm and by aspirating off the supernatant. The growth of the controls as estd by protein assayzz was 12- to 15-fold in 3 days (Table V).

Effect of NWubstituted Adenosines on Leukemia L 1210, in vivo. Female DBA,/HA mice (6-8 weeks old) (18-20 g) were obtd from the RPMI breeding colony. Each animal was injected ip with 1 x lo6 cells of leukemia L 1210, and treated with the drug once daily for 6 consecutive days starting the day after tumor inoculation. The compds were administered in a homogenized 1% CMC emulsion. Sur- vival time was noted on each animal (Table VI).

research grants (CA-11047) from the National Cancer Insti- tute, U. S. Public Health Service, and (T-436) from the American Cancer Society. The author wishes to express his thanks to Dr. Sahai Srivastava for help with the cytokinin

Assay of Cytokinin Activity. The cytokinin effect of the new

Acknowledgment. This work was supported in part by

assays, to Dr. M. T. Hakala and Miss A. I. Mulhern for cell culture determinations, to Dr. Alexander Bloch, Miss Ginger Dutschmann, and Dr. Gerald Grindey, for help with the mouse study, and Mr. Robert Maue for the microbial assay procedures.

References (1) M. H. Fleysher, M. T. Hakala, A. Bloch, and R. H. Hall, J. Med.

Chem., 11,717 (1968). (2) M. H. Fleysher, A. Bloch, M. T. Hakala, and C. A. Nichol, ibid.,

12, 1056 (1969). (3) M. H. Fleysher, Abstracts of the 160th National Meeting of

the American Chemical Society, Chicago, Ill., Sept 1970, MEDI 54.

(4) G. Shaw, B. M. Smallwood, and D. V. Wilson, J. Chem SOC. C, 921 (1966).

(5) F. Skoog, H. Q. Hamzi, A. M. Szweykowska, N. J. Leonard, K. L. Carraway, T. Fujji, J. P. Halgeson, and R. N. Loeppky, Phytochemistry, 6,1169 (1967).

(6) S. M. Hecht, N. J. Leonard, R. Y. Schmitz, and F. Skoog, ibid., 9, 1173 (1970).

(7) S. M. Hecht, N. J. Leonard, R. Y. Schmitz, and F. Skoog, Isr, J. Chem., 6, 539 (1968).

(8) T. Murashige and F. Skoog, Physwl. Plantarum 15,3345 (1966).

(9) W. J. Burrows, D. G. Armstrong, F. Skoog, S. M. Hecht, J. T. A. Boyle, N. J. Leonard, and J. Occolowitz, Science, 161, 961 (1968);Biochemistry, 8, 3071 (1969);Science, 166, 1272 (1969).

(10) S. M. Hecht, N. J. Leonard, R. Y. Schmitz, and F. Skoog, Phytochemistry, 9, 1907 (1970).

(11) R. W. Hall, M. J . Robins, L. Stasiuk, and R. Thedford, J. Amer. Chem Soc., 88,2614 (1966).

(12) M. J. Robins, R. H. Hall, and R. Thedford,Biochemistry, 6, 1837 (1967).

(1 3) A. A. Morton,"Laboratory Technique in Organic Chemistry," 1st ed, 1938, p 51, McGraw Hill Book Co., New York, N. Y.

(14) U. Harder, E. Pfeil, and K. Zenner, Chem Ber., 97,510 (1964). (15) J . D. Roberts and R. H. Mazur, J. Amer. Chem. Soc., 73, 2509

(16) A. Bloch and C. Coutsogeorgopoulos, Biochemistry, 5 , 3345 (195 1).

(1966); A. Bloch, M. H. Fleysher, R. Thedford, R. J . Maue, and R. H. Hall, J. Med. Chem., 9,886 (1966).

(17) C. H. Gray and E. L. Tatum,Proc. Nat. Acad. Sci. U. S., 30, 404 (1944).

(18) H. Eagle, Science, 130, 132 (1959). (19) J. E. Grady, W. L. Lummis, and C. G . Smith, Cancer Res., 20,

(20) 0. H. Lowry, N. J. Rosenbrough, A. L. Fan, and R. J . Randall,

(21) G. E. Moore, R. E. Gemer, and H. A. Franklin, J. Amer. Med.

1114 (1964).

J. Biol. Chem., 192, 265 (1951).

Ass., 199,s 19 (1967). (22) V. I. Oyama and H. Eagle, Proc. SOC. Exp. Biol. Med., 91,

305 (1956).


Potential antitumor agents. 5. Methylated.alpha.-(N)-heterocyclic carboxaldehyde thiosemicarbazones

Krishna C. Agrawal, Robert J. Cushley, Seymour R. Lipsky, J. Randall Wheaton, and Alan C. SartorelliJ. Med. Chem., 1972, 15 (2), 192-195• DOI: 10.1021/jm00272a016 • Publication Date (Web): 01 May 2002






192 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Agrawal, et al.

Potential Antitumor Agents. 5. Methylated a-(N)-Heterocyclic Carboxaldehyde Thiosemicarbazones?

Krishna C. Agrawal,* Robert J. Cushley,* Seymour R. Lipsky,* J. Randall Wheaton, and Alan C. Sartorelli Department of Pharmacology and Section of Physical Sciences, Division of Health Science Resources, Yale University School of Medicine, New Haven, Connecticut 06510. Received August 16, 1971

Several monomethylated derivatives of a-(N)-heterocyclic carboxaldehyde thiosemicarbazones were prepared to define the molecular dimensions compatible with antineoplastic activity. These were the thiosemicarbazones of 3-, 4-, and 5-methylisoquinoline-l-carboxaldehyde and of 3-, 4-, 5-, and 6-methyl- pyridine-2-carboxaldehyde. Tests for tumor-inhibitory potency indicate that in general the pyridine derivatives were better inhibitors of the growth of the L1210 lymphoma than were the isoquinolines. In- troduction of a 6-Me group in the pyridine ring and of an analogous 3-Me substituent in the isoquinoline structure resulted in compounds with no antineoplastic activity indicating an apparent intolerance to substitution at the a' position to the heterocyclic N atom for inhibitory action. Me substituents at the other positions of these ring systems did not significantly increase the carcinostatic potency of the parent compounds.

A variety of thiosemicarbazones of a-(N)-heterocyclic carboxaldehydes have been shown to be potent inhibitors of transplanted rodent neoplasms,' spontaneous lymph- omas of dogs,2 and DNA viruses of the Herpes fa mil^.^ Investigation of the structure-activity relationships required for antineoplastic activity4 have indicated that certain of the overall molecular dimensions of 1 -formylisoquinoline thiosemicarbazone (IQ-1), one of the most active members of this series,' could be modified with the retention of high biological activity. Thus, substituents such as 5-OH and 5-OAc were found to confer therapeutic indices to the resultant derivatives that were greater than for the parent compound.4a Substitution of these groups (Le., OH and OAc) in the 4 position of the isoquinoline ring resulted in compounds which were both less active and less toxic than the parent compound against sarcoma 180 ascites cells in mice; the sodium salt of the 4-OH derivative, however, was considerably more efficacious than the parent compound on the L1210 lymphoma.& Sodium salts of several other OH derivatives of both isoquinoline and pyridine ring systems have also been prepared as a means of solubilizing for parenteral administration these extremely insoluble compounds; in many instances such solubilization conferred greater therapeutic gain.6 The structural specificity of the formyl thiosemicarbazone side chain was found to be critical, since modifications made at the various positions of the side chain resulted in either a decrease or a complete loss of antitumor activity.4b

The biochemical basis for the growth-inhibitory activity of IQ-1 has also been studied in our laboratory.' The investigations indicate that in mammalian cells the primary site of action of these agents is the biosynthesis of DNA, with the location of the metabolic lesion being the conversion of ribonucleotides to deoxyribonucleotide forms.' Blockade of the formation of RNA and protein also occurs, but these pathways are considerably less susceptible to drug induced inhibition. A similar mechanism of action appears to be operative with both 3-hydroxy-2-formylpyridine thiosemicarbazone and 5-hydroxy-2-formylpyridine thiosemicarbazone ,9 two thiosemicarbazone derivatives of the pyri-

dine ring system with relatively great therapeutic indices as antineoplastic agents. lo

It was deemed desirable to further define the molecular dimensions compatible with the biological activity of this relatively new class of antineoplastic agents. To this end a number of monosubstituted alkylated derivatives of the two most active heterocyclic ring systems in this series, the isoquinoline and pyridine, have been synthesized and bio- logically evaluated.

Chemistry. The syntheses of Me-substituted derivatives of isoquinoline- and pyridinecarboxaldehyde thiosemicarbazones were initiated from corresponding dimethyl- substituted heterocyclic ring systems. Lutidines were obtained commercially and dimethylisoquinolines were synthesized utilizing the Bischler-Napieralski reaction. Con- siderable difficulties were encountered in the dehydrogenation with Pd of 3,4-dihydroisoquinolines when prepared according to a published procedure. '* Dehydrogenation of 1,3-dimethyl-3,4-dihydroisoquinoline or 1,4-dimethyl-3,4- dihydroisoquinoline with Pd at 200" produced in part unknown compounds which were found to have no aromatic protons in nmr; the identification of these compounds is under investigation. Dehydrogenation was carried out in good yield, however, by heating the dihydroisoquinolines with Ph2S2 and removing the formed thiophenol by dis- tillation.

Direct oxidation of 1,3- and 1,4-dimethylisoquinolines (1 and 7) with SeOz resulted in poor yields of the respective 1 carboxaldehydes. These compounds were therefore synthesized by rearrangement of their N-oxides with Ac2O

Scheme I

C H 3

1

CH3

2

I AczO 2 d i l H C i

OH

CHZOH cn3

3 4

I CHO

'\

?Presented in part before the Division of Medicinal Chemistry at 5, S

. , \ y H : N k ~ ~ c n 3

k q h 5

the 157th National Meeting of the American Chemical Society, Minneapolis, Minn., April, 1969; this work was supported by Grant CA-02817 from the National Cancer Institute, U. S. Public Health Service, and Grant T-23 from the American Cancer Society and the Biotechnology Resources Branch of the National Institutes of Health (FR 00356). CH-NNHCNHz

$Section of Physical Sciences. 6

Anitumor 7Rwsemicarbazones Journal ofhfedicinal Chemistry, 1972, Vol. IS, No. 2 193

Scheme II

(Schemes . _ ~ I and 11). The rearrangement of l-methylisoquinoline N-oxide has been reviously reported and a possible mechanism proposed.& A similar mechanism would be expected to be operative in the transformation reactions of 1 (Scheme I). Two compounds, 3-methyl-1 -hydroxymethylisoquinoline (3) and 1,3-dimethyl-4-hydroxyisoquinoline (4), were isolated after acid hydrolysis of the mixture of esters obtained by heating 1,3-dimethylisoquinoline N- oxide (2) with Ac20. Oxidation of 3 with Mn02 followed by reaction with thiosemicarbazide gave the desired thiosemicarbazone 6.

Treatment of 1,4-dimethylisoquinoline N-oxide (8) with Ac20 (Scheme 11) produced mainly 4-methyl- 1 -acetoxymethylisoquinoline (1 1). A minor isomer, possibly l-methyl- 4-acetoxymethylisoquinoline, was present as a contaminant. Positive identification of this trace material was not accomplished. Compound 11 was hydrolyzed with dil HCl to yield 4-methyl- 1 -hydroxymethyBsoquinoline (1 2) and then purified by alumina column chromatography. The reaction of 8 with Ac20 contrasts with the attempted rearrangement of 4-methylisoquinoline N-oxide by Robison and Robison12 who could not isolate a discrete chemical material from the polymeric mass which was produced. In the case of 9, with a Me group in position 1, the rearrangement presumably proceeds by an intramolecular mechanism through intermediates 9 and 10 in a manner analogous to that proposed for 2-picoline N - 0 ~ i d e . l ~ Nucleophilic attack of the N- oxide O on Ac20 results in a formation of intermediate 9. Abstraction of a proton from 9 by AcO- follows to produce the anhydro base 10 which undergoes an intramolecular rearrangement. Compd 12 was then oxidized with Mn02 to the corresponding carboxaldehyde 13 which on reaction with thiosemicarbazide yielded the desired product 14.

Selective oxidation of the 1-Me group in 1,5-dimethylisoquinoline (15) to the corresponding 1-carboxaldehyde (16) was achieved in fair yield with Se02 (Scheme 111). Compd

Scheme III

16 on reaction with thiosemicarbazide produced the desired derivative (17). Methylated pyridine-2-carboxaldehydes, the precursors of the synthesis of thiosemicarbazones,

were prepared by previously published procedure^'^ utilizing the rearrangement of corresponding lutidine N-oxides with Ac~O.

Nmr Studies. Nmr spectral parameters for the compds prepared were in accord with the structures proposed; dif- ferentiation of the various isomers from nmr results has been possible. All peaks assigned to replaceable protons were confirmed by addition of CF3C02H.

gated in the present study, as well as previously,& some useful empirical results have been obtained which can be used to characterize compds in the isoquinoline series. As an example, unless C-8 or the ring N contains substituents (e.g., N + 0), the chemical shift for a 1-Me group occurs between 6 2.81 and 2.92. Similarly, doubly bonded electro- negative substituents on the C directly bonded to C-1 cause a paramagnetic shift of the proton resonance at C-8 into the region 6 9.13-9.26. This C-8 proton multiplet lies to lower field than the other ring protons.

An interesting long-range coupling in isomers I and I1 of 3,4-dihydroisoquinolines (Figure 1) has been used to establish the cyclized nature of the compounds. In I, the 1-Me substituent (6 2.32) is a triplet,J1-a,,+3 = 1.5 Hz, due to long-range coupling to the two H-3's (dark lines). On the other hand, in I1 the Me is a doublet showing a 5-bond long range coupling, J l - a ,H-3 = 1.9 Hz, due to coupling with the single proton on &-3. Such longrange coupling must in- volve the heterocyclic N atom. Double irradiation of the 1-Me resonance lines in I reduced the multiplet structure at 6 3.42 to an octet which is the AB subspectrum of the ABMXB spin system. In the case of the 5-Me compounds, only in 16 was a splitting resolved between the 5-Me group and the protons at C-4 and C-6. The 5-Me group in this compound was a triplet, J = 0.7 Hz.

Biological Results and Discussion. The antitumor activity of methylated a-(N)-heterocyclic carboxaldehyde thiosemicarbazones against the L12 10 lymphoma in mice is given in Table I. The findings indicate that, in general, the pyridine derivatives were better inhibitors of the L1210 lymphoma than the isoquinoline series. Although the parent compound, IQ-1, doubled the average life-span of tumor-bearing mice as compared to untreated controls, the Me-substituted derivatives 6, 14, and 17 increased the average survival time to a lesser extent. In the pyridine series 2-formylpyridine thiosemicarbazone (PT) was the most active agent causing more than a 2-fold increase in the average survival time. However, this compound was relatively toxic as shown by a 14.4% loss in body weight at the optimal dose level of 5 mg/kg administered twice daily. Introduction of a 3-Me group (18) did not result in therapeutic improvement over PT; however, 19, with a 4-Me substituent, was not only found to be equally active with PT as a tumor inhibitor, but was also much less toxic causing only 4.3% loss in body weight. Compd 20 having a 5-Me substituent was relatively nontoxic as shown by a minimal loss in body weight, but still increased the average survival time to 18.2 days. Intro-

From the relatively abundant number of compds investi-

I

Figure 1.

194 Journal of Medicinal Chemistry, 19 72, Vol. 15, No. 2 Agrawal, et al.

Table I. Effect of Me-Substituted a-(N)-Heterocyclic Carboxaldehyde Thiosemicarbazones on the Survival Time of Mice Bearing the L1210 Lymphoma

Maximum effective Average A Compound daily dose, mg/kgn weight, %b

None + 5.7 IQ-1 20 - 0.9

6 20 + 8.5 14 20 +11.5 17 20 + 8.0 PT 5 -14.4 18 15 -10.5 19 5 - 4.3 20 15 - 0.5 21 40 + 1.6

Average survival, days

8.0 16.6 10.8 13.6 8.4

21.6 17.4 20.6 18.2 12.2

'Administered twice daily for 4 consecutive days at 12-hr intervals beginning 24 hr after tumor transplantation. bAverage wt change from onset to termination of drug treatment.

the corresponding amides by published procedures" according to the Bischler-Napieralski reaction.

1,3-Dimethylisoquinoline (1). An equimolar mixture of 1,3-dimethyl-3,4-dihydroisoquinoline (9.27 g) and 12.71 g of Ph,S, was heated to 225" in an oil bath permitting the by-product C,H,SH to distill off. The residue was acidified with 10% HC1, the impurities were extd with Et,O, and the aq layer was basified with NaOH soln. The mixt was then extd with Et20, dried (NaJO,), filtered, and flash evapd, and the residual oil was distd at 90-93" (0.4 mm) to give 7.95 g (89%) of pure colorless product: nmr S 2.85 (s, 3 H, l-CH,), 2.55 (s, 3 H, 3-CH3).

1,4-Dimethylisoquinolinine (7). The procedure employed was similar to the prepn of 1. The residual oil was distd at 100-105" (0.3 mm) to yield 7.3 g (80%) of pure colorless compd: nmr 6 2.86 (s, 3 H, 1-CH,), 2.48 (s, 3 H, 4-CH,), 7.39 (broad s, 1 H, H-3).

1,s-Dimethylisoquinoline (15). This compd was synthesized according to Spath, et al.:" mp 92-93" (reported 97-98"); nmr 6 2.89 (s, 3 H, l-CH,), 2.61 (s, 3 H, 5-CH,), 8.40 (d, 1 H, J = 7.0 Hz, H-3).

1,3-Dimethylisoquinoline N-Oxide (2). A 40% s o h of AcO,H

- Table 11. Methylated a-(N)-Heterocyclic Carboxaldehyde Thiosemicarbazones Me Substitution

Compound position Mp, "C dec Yield, % Crystn solvent Formula Analyses Isoquinolines

6 3 207-208 87 THF-cyclohexane C I Z H l 2N4S C, H, N, S 14 4 214-21 5 70 EtOH-H,O '1 2H1,N4S C, H, N, S 17 5 233-235 83 a CIZHl 2N4S C, H, N, S

18 3 214-2 16& EtOH C8H10N4S C, H, S 19 4 193-195' EtOH-KO C8H10N4S C, H, S 20 5 224-22Sd EtOH C8H1CDN4S C, H, S 21 6e

aRecrystn was not necessary. bReported mp 209°.14a 'Reported mp 194-196°.14b dReported mp 220.5°.14a eThis derivative was generously supplied by MI. Frederic A. French, Mount Zion Hospital and Medical Center, Chemotherapy Research Laboratory, Palo Alto, Calif.

Pyridines

duction of a 6-Me group in PT (21) resulted in a compound with only minimal biological activity. A similar result was obtained in the isoquinoline series where a 3-Me substituent, which may be considered to be analogous to the 6-Me group of pyridine series, resulted in a compd (6) with much lower antineoplastic activity. Compd 17 which contains a 5-Me group in the isoquinoline ring also showed no antineoplastic activity. This derivative (17) has been reported to be equal in activity to 14 as an inhibitor of the target enzyme ribonucleoside diphosphate reductase in vitro. l5 This difference might be explained by the extreme water insolubility of 17 which leads to poor uptake of the compound by neoplastic cells.


melting point apparatus and are uncor. Elemental analyses8 were performed by the Schwarzkopf Microanalytical Laboratory, Wood- side, N. y., and the Baron Consulting Co., Orange, Conn. Nmr spectra were detd with a Bruker HFX-3 spectrometer operating at 90 MHz and fitted with an H-P frequency counter. Chemical shifts (6) are given in ppm downfield from TMS. Spectra were obtd in DMSO-d, (s, singlet; d, doublet; m, multiplet; dd, doublet of doub- lets, etc). Only those resonance signals necessary for differentiating the various compds are described.

Antitumor Screening. The thiosemicarbazones were tested for antineoplastic activity in mice bearing the L1210 lymphoma. Com- plete details of the screening procedure have been described earlier.6 The ascites tumor was transplanted by inoculating BDF, mice ip with approximately 4 X lo6 tumor cells. Drugs were administered ip in fine suspension beginning 24 hr after tumor implantation and treatment was continued twice daily at 12-hr intervals for 4 consecutive days. Determination of antineoplastic activity was based upon the prolongation of survival time afforded by the drug treat- mert.

Melting points were determined with a Thomas-Hoover capillary

Dimethyl-3,4dihydroisoquinolines were synthesized by cyclizing

§Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements are within 0.4% of the theoretical value.

(6.1 ml) was mixed with 0.46 g of NaOAc and added slowly to 4.74 g of 1 preheated at 80". The mixt was heated for 4 hr with stirring and then allowed to remain at room temp overnight. The AcOH was removed by flash evapn, 2 was then extd with Et ,O, solvent was removed, and the compd was distd at 153-155" (0.2 mm) to yield 4.07 g (78%) of pure material which crystd on standing: mp 50-51"; nmr 6 2.76 (s, 3 H, 1-CH,), 2.50 (s, 3 H, 3-CH,). Anal. (C,,H,,NO) C, H, N.

1,4-Dimethylisoquinoline N-oxide (8) was synthesized according to the ptocedure of Robison and Robison" for the synthesis of isoquinoline N-oxide. Recrystn from EtOAc and cyclohexane (Norit A) gave pale, tan-colored crystals: yield 58%; mp 150-151" ; nmr 6 2.68 (s, 3 H, 1-CH,), 2.50 (s, 3 H, 4-CH3), 8.10 (broad s, 1 H, H-3). Anal. (C,,H,,NO) C, H, N.

Rearrangement of 1,3-Dimethylisoquinoline N-Oxide. Compd 2 (4.01 g) washeated at 120" in 40 ml of Ac,O for 2 hr. The mixt was concd to give a dark oil which distd at 130-150" (0.35 mm) yielding 4.4 g (88%) of an oily mixt of esters. To 4.28 g of this mixt of esters 100 ml of 10% HCl was added and heated for 1 hr at 100". The soln was made alk with NaOH soln (pH 11.0) and 3-methyl-1- hydroxymethylisoquinoline (3) was extd with Et,O, dried (NaZS04 ), and concd. The residual solid was recrystd from petr ether to give 1.26 g (37%) as colorless crystals: mp 103-104"; nmr 6 2.60 (s, 3 H, CH,), 5.05 (s, 2 H, CH,). Anal. (C,,H,,NO) C, H, N.

The alk layer was neutralized to pH 7.0 with 10% HCI, whereupon 4-hydroxy-1,3-dimethylisoquinoline (4) pptd. The ppt was filtered, washed with H,O and dried. Recrystn from EtOAc contg a small amt of MeOH yielded 0.7 g (20%) of pure material: mp 169- 170';nmr 6 2.84 (s, 3 H, 1-CH,), 2.77 (s, 3 H, 3-CH,), 3.86 (broad s, 1 H, OH). Anal. (C,,H,,NO) C, H, N.

Rearrangement of 1,4-Dimethylisoquinoline N-Oxide. Compd 8 (1.5 g) was heated at 110" in 1.5 ml of Ac,O for 2 hr. Excess Ac,O was removed leaving a dark oil which was extd with Et,O. The Et,O was removed and then the oil was distd at 167-170" (0.5 mm). The pale yellow oil crystd upon cooling: yield 1.14 g (61%); mp 73-76". Several recrystns from petr ether raised the mp to 76-77". The 4- methyl-1-acetoxymethylisoquinoline (1 1) was possibly contaminated with the 4-acetoxymethyl isomer and was purified after acid hydrolysis.

1 hr. The soln was basified with NaOH soh and extd with Et,O. Compd 11 (0.8 g) was heated at 100" in 20 ml of 10% HC1 for ~.

The Et,O layer was dried (Na,SO,), filtered, and concd to give 4- methyl-1-hydroxymethylisoquinofine (12). The crude 12 was

Coenzyme Q-Enzyme Inhibitors Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 195

chromatogd on 6 0 g of silica gel (100-200 mesh) and eluted with EtOAc (500 ml). The solvent was removed and the residue was recrystd from hexane to give colorless crystals: 0.27 g (42%); mp 87-88"; nmr 6 2.56 (s, 3 H, 4-CH,), 5.05 (d, 2 H, CH,). Anal. (CllH1,NO) C, H, N.

2.2 mmoles) was dissolved in 20 ml of C,H, and 0.35 g (4.0 mmoles) of MnO, was added. The mixt was refluxed for 4 hr and fitered, and the C,H, was removed to yield 5. Recrystn from petr ether gave colorless needles: 0.27 g (72%); mp 71-72'; nmr 6 10.22 (s, 1 H, CHO), 2.71 (s, 3 H, 3-CH,). Anal. (CliHJO) N.

the same procedure as 5, oxidizing 12 with MnO, . Recrystn from petr ether yielded colorless crystals in 70% yield: mp 59-60"; nmr 6 10.21 (s, 1 H, CHO), 2.68 (s, 3 H, 4-CH3). Anal. (C,,%NO) N.

g, 2 mmoles) was dissolved in 25 ml of dioxane and 0.22 g (2 mmoles) of SeO, was added slowly. The mixt was refluxed for 2.5 hr and filtered, dioxane was removed, and the residue was extd with dil HCl. The acid layer was filtered and made alk with NaHCO, , and the resulting ppt was collected, washed (H,O), and dried. Crystn from hexane (Norit) yielded colorless fibrous material: 0.25 g (72%); mp 108-109";nmr 6 10.28 (s, 1 H, CHO), 8.81 (d, 1 H , J = 5.7 Hz,

Anal. (C,,H,NO) N.

treating alcoholic solns of the corresponding carboxaldehydes with an aq soln of thiosemicarbazide contg a few drops of dil AcOH. Relevant data concerning these compds are listed in Table 11.

Corby Morton, Miss Lynn A. Bon Tempo, and Miss Annette F. Gwardyak for excellent assistance.

3-Methylisoquinoline-1-carboxaldehyde (5). Compd 3 (0.38 g,

4-Methylisoquinoline-1-carboxaldehyde ( 13) was synthesized by

5-Methylisoquinoline-1-carboxaldehyde (16). Compd 15 (0.32

H-3), 8.18 (dd, 1 H, J = 5.5 and 0.8 Hz, H4), 2.69 (t, 3 H, S-CH,).

Thiosemicarbazones. The thiosemicarbazones were prepd by

Acknowledgment. The authors wish to thank Mrs. Jill

References (1) (a) R. W. Brockman, J. R. Thomson, M. J. Bell, and H. E.

Skipper, Cancer Res., 16,167 (1956); (b) F. A. French and E. J. Blanz, Jr., J. Med. Chem., 9 , 5 8 5 (1966).

Sartorelli, Fed. Proc., Fed. Amer. Soc. Exp. BioL, 2 9 , 9 0 8 (1970).

(3) W. Brockman, R. W. Sidwell, G. Arnett, and S. Shaddix,Proc. Soc. Exp. Biol. Med., 133,609 (1970).

(4) (a) K. C. Agrawal, B. A. Booth, and A. C. Sartorelli, J. Med. Chem., 1 1 , 7 0 0 (1968); (b) K. C. Agrawal and A. C. Sartorelli, ibid., 12,771 (1969); (c) K. C. Agrawal, R. J. Cushley, W. J. McMurray, and A. C. Sartorelli, ibid., 13,431 (1970).

( 5 ) F. A. French, and E. J. Blanz, Jr., CancerRes., 25,1454 (1965). (6) K. C. Agrawal and A. C. Sartorelli, J. Pharm. Sei., 57,1948

(1968). (7) (a) A. C. Sartorelli,Biochem. Biophys. Res. Commun., 27, 26

(1967); (b) A. C. Sartorelli, Pharmacologist, 9, 192 (1967). (8) E. C. Moore, M. S. Zedeck, K. C. Agrawal, and A. C. Sartorelli,

Biochemistry, 9 , 4 4 9 2 (1970). (9) (a) B. A. Booth, E. C. Moore, and A. C. Sartorelli, Cancer Res.,

31,228 (1971); (b) E. C. Moore, B. A. Booth, and A. C. Sartorelli, ibid., 31,235 (1971).

(10) (a) F. A. French and E. J. Blanz, Jr., ibid., 26, 1638 (1966); (b) F. A. French and E. J . Blanz, Jr., Gann, 2 , 5 1 (1967).

(11) E. Spath, F. Berger, and W. Kuntara, Ber., 63B, 134 (1930). (12) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 80,

(13) V. J. Traynelis and R. F. Martello, ibid., 80,6590 (1958). (14) (a) W. Mathes and W. Sauermilch, Ber., 90, 758 (1957); (b) S.

(15) A. C. Sartorelli, K. C. Agrawal, and E. C. Moore, Biochem.

(2) W. A. Creasey, K. C. Agrawal, K. K. Stinson, and A. C.

3443 (1958).

Furukawa and Y . Kuroiwa, Chem. Pharm. Bull., 3 , 2 3 2 (1955).

Pharmacob, 20,3119 (1971).

Specificity of Inhibition of Coenzyme Q-Enzyme Systems by Lipoidal Benzoquinone Derivatives?

Ronald S. Pardini, Joseph C. Catlin, James C. Heidker, and Karl Folkers" Stanford Research Institute, Menlo Park, California 94025, and Institute for Biomedical Research, The University o f Texas at Austin, Austin, Texas 78712. ReceivedAugust 9, 1971

5-Substituted 2,3-dimethoxy-6-phytyl-l,4-benzoquinones were found to inhibit mitochondrial NADH- oxidase and succinoxidase systems from beef heart. The most effective group in the 5 position was OH for the 6-phytyl derivatives. The 5 4 1 and 5-Br derivatives were less inhibitory than the 5-OH derivatives, and in diminishing degree. The 5-Me0 derivative was essentially noninhibitory. 6-Alkyl- and 6-isoprenyl-2,3- dimethoxy-5-hydroxy-1,4-benzoquinones were similarly evaluated. Inhibition of the NADH-oxidase system was greatest when the hydroxyquinone possessed a side chain of from 16 and 17 C. Inhibition of the succinoxidase system was relatively nonspecific in respect t o the side chain. The succinoxidase system was generally more sensitive to most of the benzoquinones tested than was the NADH-oxidase system.

Mitochondrial reconstruction'Y2 and spectrophotometric investigations on the kinetics of coenzyme Q turnover during electron transport3 are responsible for the view that coenzyme Q participates in the primary electron transport sequence. Coenzyme Q has a widespread distribution in biological systems4 including the malarial parasite.'-' Mammalian succinoxidase and NADH-oxidase systems have been extensively studied and may be considered representative of the coenzyme Q electron transport sequences.

Preliminary structure-activity investigations demonstrated that mitochondrial succinoxidase activity was inhibited by various antimalarial naphthoquinone analogs, and the activ- i ty was restored by coenzyme Q and its derivative^.^ Next, chloroquine and a new napthoquinone antimalarial, 2-w- cyclohexyloctyl-3 -hydroxy- 1,4-naphthoquinone, were

'Author to whom correspondence should be addressed at the

"Coenzyme Q. 1 4 1 . University of Texas.

shown to inhibit beef heart mitochondrial succinoxidase systems." Again, the inhibitory action of antimalarial agents was reversed by coenzyme Q.

These findings link one kind of inhibition of mitochondrial electron transport a t the coenzyme Q loci to chemotherapy of malaria. CoQs is the dominant CoQ of Plasmodium '-' A series of benzoquinones structurally related to known antimalarial naphthoquinones, which were also succinoxidase inhibitor^,^ have been synthesized." These benzoquinones also have structural resemblance to coenzyme Q. They represent potential antagonists of coenzyme Q function in mitochondrial electron transport; hence, they also represent potential antimalarial activity.

Separate sites for the function of coenzyme Q in the succinoxidase and NADH-oxidase systems in beef heartl2'l3 and yeast13>14 mitochondria have been implicated by studies on the structural specificity of coenzyme Q. These investigations demonstrated that the function of coenzyme Q in the


Specificity of inhibition of coenzyme Q-enzymesystems by lipoidal benzoquinone derivatives

Ronald S. Pardini, Joseph C. Catlin, James C. Heidker, and Karl FolkersJ. Med. Chem., 1972, 15 (2), 195-197• DOI: 10.1021/jm00272a017 • Publication Date (Web): 01 May 2002






Coenzyme Q-Enzyme Inhibitors Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 195

chromatogd on 6 0 g of silica gel (100-200 mesh) and eluted with EtOAc (500 ml). The solvent was removed and the residue was recrystd from hexane to give colorless crystals: 0.27 g (42%); mp 87-88"; nmr 6 2.56 (s, 3 H, 4-CH,), 5.05 (d, 2 H, CH,). Anal. (CllH1,NO) C, H, N.

2.2 mmoles) was dissolved in 20 ml of C,H, and 0.35 g (4.0 mmoles) of MnO, was added. The mixt was refluxed for 4 hr and fitered, and the C,H, was removed to yield 5. Recrystn from petr ether gave colorless needles: 0.27 g (72%); mp 71-72'; nmr 6 10.22 (s, 1 H, CHO), 2.71 (s, 3 H, 3-CH,). Anal. (CliHJO) N.

the same procedure as 5, oxidizing 12 with MnO, . Recrystn from petr ether yielded colorless crystals in 70% yield: mp 59-60"; nmr 6 10.21 (s, 1 H, CHO), 2.68 (s, 3 H, 4-CH3). Anal. (C,,%NO) N.

g, 2 mmoles) was dissolved in 25 ml of dioxane and 0.22 g (2 mmoles) of SeO, was added slowly. The mixt was refluxed for 2.5 hr and filtered, dioxane was removed, and the residue was extd with dil HCl. The acid layer was filtered and made alk with NaHCO, , and the resulting ppt was collected, washed (H,O), and dried. Crystn from hexane (Norit) yielded colorless fibrous material: 0.25 g (72%); mp 108-109";nmr 6 10.28 (s, 1 H, CHO), 8.81 (d, 1 H , J = 5.7 Hz,

Anal. (C,,H,NO) N.

treating alcoholic solns of the corresponding carboxaldehydes with an aq soln of thiosemicarbazide contg a few drops of dil AcOH. Relevant data concerning these compds are listed in Table 11.

Corby Morton, Miss Lynn A. Bon Tempo, and Miss Annette F. Gwardyak for excellent assistance.

3-Methylisoquinoline-1-carboxaldehyde (5). Compd 3 (0.38 g,

4-Methylisoquinoline-1-carboxaldehyde ( 13) was synthesized by

5-Methylisoquinoline-1-carboxaldehyde (16). Compd 15 (0.32

H-3), 8.18 (dd, 1 H, J = 5.5 and 0.8 Hz, H4), 2.69 (t, 3 H, S-CH,).

Thiosemicarbazones. The thiosemicarbazones were prepd by

Acknowledgment. The authors wish to thank Mrs. Jill

References (1) (a) R. W. Brockman, J. R. Thomson, M. J. Bell, and H. E.

Skipper, Cancer Res., 16,167 (1956); (b) F. A. French and E. J. Blanz, Jr., J. Med. Chem., 9 ,585 (1966).

Sartorelli, Fed. Proc., Fed. Amer. Soc. Exp. BioL, 29 ,908 (1970).

(3) W. Brockman, R. W. Sidwell, G. Arnett, and S. Shaddix,Proc. Soc. Exp. Biol. Med., 133,609 (1970).

(4) (a) K. C. Agrawal, B. A. Booth, and A. C. Sartorelli, J. Med. Chem., 11 ,700 (1968); (b) K. C. Agrawal and A. C. Sartorelli, ibid., 12,771 (1969); (c) K. C. Agrawal, R. J. Cushley, W. J. McMurray, and A. C. Sartorelli, ibid., 13,431 (1970).

( 5 ) F. A. French, and E. J. Blanz, Jr., CancerRes., 25,1454 (1965). (6) K. C. Agrawal and A. C. Sartorelli, J. Pharm. Sei., 57,1948

(1968). (7) (a) A. C. Sartorelli,Biochem. Biophys. Res. Commun., 27, 26

(1967); (b) A. C. Sartorelli, Pharmacologist, 9, 192 (1967). (8) E. C. Moore, M. S. Zedeck, K. C. Agrawal, and A. C. Sartorelli,

Biochemistry, 9 ,4492 (1970). (9) (a) B. A. Booth, E. C. Moore, and A. C. Sartorelli, Cancer Res.,

31,228 (1971); (b) E. C. Moore, B. A. Booth, and A. C. Sartorelli, ibid., 31,235 (1971).

(10) (a) F. A. French and E. J. Blanz, Jr., ibid., 26, 1638 (1966); (b) F. A. French and E. J . Blanz, Jr., Gann, 2 , 5 1 (1967).

(11) E. Spath, F. Berger, and W. Kuntara, Ber., 63B, 134 (1930). (12) M. M. Robison and B. L. Robison, J. Amer. Chem. Soc., 80,

(13) V. J. Traynelis and R. F. Martello, ibid., 80,6590 (1958). (14) (a) W. Mathes and W. Sauermilch, Ber., 90, 758 (1957); (b) S.

(15) A. C. Sartorelli, K. C. Agrawal, and E. C. Moore, Biochem.

(2) W. A. Creasey, K. C. Agrawal, K. K. Stinson, and A. C.

3443 (1958).

Furukawa and Y . Kuroiwa, Chem. Pharm. Bull., 3 ,232 (1955).

Pharmacob, 20,3119 (1971).

Specificity of Inhibition of Coenzyme Q-Enzyme Systems by Lipoidal Benzoquinone Derivatives?

Ronald S. Pardini, Joseph C. Catlin, James C. Heidker, and Karl Folkers" Stanford Research Institute, Menlo Park, California 94025, and Institute for Biomedical Research, The University o f Texas at Austin, Austin, Texas 78712. ReceivedAugust 9, 1971

5-Substituted 2,3-dimethoxy-6-phytyl-l,4-benzoquinones were found to inhibit mitochondrial NADH- oxidase and succinoxidase systems from beef heart. The most effective group in the 5 position was OH for the 6-phytyl derivatives. The 5 4 1 and 5-Br derivatives were less inhibitory than the 5-OH derivatives, and in diminishing degree. The 5-Me0 derivative was essentially noninhibitory. 6-Alkyl- and 6-isoprenyl-2,3- dimethoxy-5-hydroxy-1,4-benzoquinones were similarly evaluated. Inhibition of the NADH-oxidase system was greatest when the hydroxyquinone possessed a side chain of from 16 and 17 C. Inhibition of the succinoxidase system was relatively nonspecific in respect t o the side chain. The succinoxidase system was generally more sensitive to most of the benzoquinones tested than was the NADH-oxidase system.

Mitochondrial reconstruction'Y2 and spectrophotometric investigations on the kinetics of coenzyme Q turnover during electron transport3 are responsible for the view that coenzyme Q participates in the primary electron transport sequence. Coenzyme Q has a widespread distribution in biological systems4 including the malarial parasite.'-' Mammalian succinoxidase and NADH-oxidase systems have been extensively studied and may be considered representative of the coenzyme Q electron transport sequences.

Preliminary structure-activity investigations demonstrated that mitochondrial succinoxidase activity was inhibited by various antimalarial naphthoquinone analogs, and the activ- i ty was restored by coenzyme Q and its derivative^.^ Next, chloroquine and a new napthoquinone antimalarial, 2-w- cyclohexyloctyl-3 -hydroxy- 1,4-naphthoquinone, were

'Author to whom correspondence should be addressed at the

"Coenzyme Q. 1 4 1 . University of Texas.

shown to inhibit beef heart mitochondrial succinoxidase systems." Again, the inhibitory action of antimalarial agents was reversed by coenzyme Q.

These findings link one kind of inhibition of mitochondrial electron transport a t the coenzyme Q loci to chemotherapy of malaria. CoQs is the dominant CoQ of Plasmodium '-' A series of benzoquinones structurally related to known antimalarial naphthoquinones, which were also succinoxidase inhibitor^,^ have been synthesized." These benzoquinones also have structural resemblance to coenzyme Q. They represent potential antagonists of coenzyme Q function in mitochondrial electron transport; hence, they also represent potential antimalarial activity.

Separate sites for the function of coenzyme Q in the succinoxidase and NADH-oxidase systems in beef heartl2'l3 and yeast13>14 mitochondria have been implicated by studies on the structural specificity of coenzyme Q. These investigations demonstrated that the function of coenzyme Q in the

196 Journal of Medicinal Chemistry, 1972, Voi. 15, No. 2 Folkers, et 01.

succinoxidase system has little specificity for the various isoprenologs, but the activity of coenzyme Q in the NADH- oxidase system is directly proportional to the length of the isoprenoid side chain. This different organic structural specificity for the succinoxidase and NADH-oxidase systems will surely influence the capacity of various coenzyme Q analogs to inhibit each of the enzyme systems. Consequently, an investigation of the structure-inhibition relationships was conducted on the recently synthesized benzoquinones, and the data are herein described. This investigation was designed to evaluate the inhibition of various 2,3-dimethoxy-5-hydroxy-6-alkyl- 1,4-benzoquinones, particularly the analogous 5-Me0-, 5-C1-, and 5-bromo-6-phytyl-substituted quinones, on beef heart mitochondrial succinoxidase and NADH- oxidase activities.


lyophilized as described.’ Portions of the lyophilized HBHM were extd with pentane’ to remove the coenzyme Q. The coenzyme Q deficient mitochondria were reconstructed by adding coenzyme Q and asolectin (a soybean phospholipid). The activities of the lyophilized unextd and the extd-reconstructed mitochondrial succinoxidase and NADH-oxidase enzyme systems were detd mano- metrically in the absence and presence of the various benzoquinones to be tested. Coenzyme Q and the various compds were added in EtOH, and the concn of EtOH was maintained in the flasks at 0.1 ml of EtOH per 3 ml of reaction mixt. Asolectin was added to serve as a carrier for the quinones. The mitochondrial protein was detd by the biuret methodI6 and was maintained at 0.5-1.0 mg per flask.

The differences in Tables I and I1 for inhibition of the extracted and reconstructed NADH-oxidase and succinoxidase systems between the specific activity (patoms of 0 consumed/min per mg

“Heavy beef heart mitochondria” (HBHM) were i~olated’~ and

of protein) of the coenzyme Q,, supplemented (100 nmoles/flask) systems and the extd control was defined as 100% activity. The per cent activity of the assay systems (specific activity for the compd minus the specific activity of the extd control) was calcd from the uninhibited specific activity representing 100%.

The per cent activities in Tables I and I1 for the lyophilized and unextd NADH-oxidase and succinoxidase systems were detd from the specific activity of the unextd control which was defmed as 100%. The flasks also contd 100 nmoles of added coenzyme Q,,.

The compd (100 nmoles) was tested in the presence of 100 nmoles of coenzyme Q,, to obtain the data in Tables I and 11. Each value in Table I and in Table I1 represents the average of from 2 to 6 assays. The specific activities of the uninhibited extd-reconstructed NADH-oxidase, unextd NADH-oxidase, extd-reconstructed succinoxidase, and unextd succinoxidase ranged from 0.250 to 0.350,0.400 to 0.650,0.300 to 0.450, and 0.350 to 0.450 lratom of 0 consumed per min per mg of protein, respectively.


The effect of various 5-substituted 2,3-methoxy-6-phytyl- 1,4-benzoquinones on mitochondrial electron transport systems was determined, and the data are in Table I. The 5-OH, 5-C1,5-Br, and 5-methoxy-6-phytyl derivatives inhibited 96, 52, 24, and 2% of the controls, respectively. The lyophilized unextracted NADH-oxidase system was similarly affected, but the magnitude of the inhibitions was less.

The extracted-reconstructed succinoxidase system was depressed to 17,49,79, and 100% of the controls in the presence of the 5-OH, 5-C1,5-Br, and 5-methoxy-6-phytyl derivatives, respectively. Similar results were obtained with the unextracted succinoxidase system, although the magnitude of the inhibitions was less. These data demonstrate that the most significant structural change for inhibition of coenzyme Q-enzyme systems is the replacement of the 5-Me group of

Table I. Inhibition of Mitochondrial Succinoxidase and NADH-oxidase by 6-Substituted 2,3-Dimethoxy-6-phytyl-l ,4-benzoquinones Inhibition of NADH-oxidase Inhibition of succinoxidase

R E and Ra L non-Eb E and Ra L non-Eb

7% c H 3 0 ~ ~ H z C H = C ~ C H 3 ~ C H z ( C H ~ C H z C H C H z CH,O ),H

0 OH cl Br OCH.

96 52 24

2

62 36 6 8

83 5 1 21

0

78 25 24

0

aE and R = extracted and reconstructed. bL non-E = lyophilized and nonextracted.

Table 11. Inhibition of Mitochondrial Succinoxidase and NADH-oxidase by 6-Substituted 2,3-Dimethoxy-5-hydroxy-l,4-benzoquinones Inhibition of NADH-oxidase Inhibition of succinoxidase

E and Ra L non-Eb E and Ra L non-Eb R ~~~~~ ~~ ~

CH,O c H 3 0 ~ 0 H 0 R

(CH,CH=C(CH&CH,),H (geranyl) (CH,CH=C(CH,)CH,),H (famesyl) (CH,CH=C(CH,)CH, 14H (tetraprenyl) (CH,CH=C(CH,)CH,),H (solanesyl)

- 7%

0 41

100 15

9 23 62 39

56 16

100 91

45 32 85 69

CH,CH=C(CH,)CH,(CH,CHzCHCHz),H (phytyl) 96 62 83 78 (CH,CH,C(CH,)HCH,),H (dihydrophytyl) 92 49 89 19

100 65 100 65 76 67 41 90 54 (CH,),C,H,, (w-cyclohexyloctyl) 76 39 86

(CHz)i&& (CHz),,CH,

‘E and R = extracted and reconstructed. bL non-E = lyophilized and nonextracted

Coenzyme Q-Enzyme Inhibitors Journal ofhfedicinal Chemism, 1972, Vol. 15, No. 2 197

coenzyme Q with 5-OH. The 5-OH substituted quinone was the most effective inhibitor of all 4 of the CoQ-enzyme systems. The 5-C1 derivative was a more effective inhibitor than the 5-Br derivative. The 5-Me0 derivative was essentially noninhibitory in all forms of the enzyme systems.

Since the derivative with the 5-OH group was the most effective inhibitor of these 5-substituted derivatives, the effect of various 6-alkyl-substituted 2,3-dimethoxy-S-hydroxy-1,4- benzoquinones on the activities (HBHM) of NADH-oxidase and succinoxidase was assessed; and the data are in Table 11. These data show that the extracted-reconstructed mitochondrial NADH-oxidase activity was not inhibited by the 5-hydroxybenzoquinone containing a 6-geranyl (Clo) side chain, but those compounds containing the 6-farnesyl (CIS), 6-cyclohexyloctyl (C 14), 6-nonadecyl (C 19), and 6-solanesyl (C45) groups depressed the enzyme activity to 25-65% of the controls. However, those 5-hydroxybenzoquinones containing the 6-tetraprenyl (CZO), 6-phytyl (CZO), 6-dihydrophytyl (C,,), and 6-heptadecyl (C groups inhibited the extracted-reconstructed mitochondrial NADH-oxidase system to 0,4,8, and 0% of the controls, respectively.

Similar structure-inhibition relationships were observed in the experiments conducted with preparations of lyophilized HBHM which was not extracted with pentane to remove CoQlo. Again, the most effective inhibitors were the 6-tetraprenyl (C,,), 6-phytyl (C,,), and 6-heptadecyl (Cl,) derivatives which inhibited the enzyme activity to 35-40%. The other derivatives caused less than 50% inhibition.

On the basis of these data, an alkyl side chain of 17 or 20 C atoms was more effective for the inhibition of the NADH- oxidase system than were side chains of 14 or 19 C atoms. Branching or degree of unsaturation of the side chain appeared to be of less structural significance than the general length of the side chain.

The specificity for the length of the alkyl side chain which was observed in the studies in NADH-oxidase inhibition was not apparent in the succinoxidase inhibition. The 6-tetraprenyl, 6-solanesyl, 6-phytyl, 6-dihydrophytyl, 6-heptadecyl, 6-nonadecyl, 6-cyclohexyloctyl, and 6-farnesyl derivatives inhibited the activity of the extracted-reconstructed succinoxidase to 0-25% of that of the controls. The 6-geranyl analog inhibited the enzyme activity to 50% of that of the controls. Apparently, the 6-geranyl derivative is too hydrophilic to be functional in electron transfer in the hydrophobic en- vironment associated with the activity of coenzyme Qlo.

With the exception of the 6-phytyl and the 6-dihydrophytyl analogs, the extracted-reconstructed succinoxidase system was more sensitive to the various 5-hydroxybenzoquinone analogs than was the comparable NADH-oxidase system.

In the lyophilized and unextracted HBHM, the succinoxidase system appeared to be more sensitive to the various antimetabolites of coenzyme Q than was the NADH-oxidase system.

The finding that the extracted-reconstructed mitochondrial enzyme systems are more sensitive to the analogs of coenzyme Q than are their respective unextracted preparations implies that the site of inhibition of these hydroxybenzoquinones is at the region of coenzyme Q in the electron transport chain, since the inhibitor and coenzyme Q were both present during reconstruction at equal concentration. This inte retation is consistent with the findings of Castelli, et d.,' that the inhibition by hydroxybenzoquinones of a mitochondrial NADH-oxidase system was reversed by coenzyme Qz .

The observed greater sensitivity and the lower structural specificity of the succinoxidase as compared to the NADH- oxidase system for the various h droxybenzoquinones also support the previous con~lusion'~'~ that coenzyme Q has separate locations in complexes I and I1 and, therefore, is situated in two different molecular environments. The greater sensitivity of the succinoxidase system to inhibition may be explained by a greater affinity of the CoQ-site associated with succinoxidase for the hydroxybenzoquinones. These relative inhibitions could also be influenced by a greater affinity for coenzyme Q in the NADH-oxidase than the succinoxidase system.

E. Harris Postdoctoral Fellowship in Biomedical Research. He and Dr. Karl Folkers express their gratitude to Mr. Harris of Woodside, Calif., for this fellowship.

References

Acknowledgments. Dr. Ronald S. Pardini held the Stanley

(1) R. L. Lester and S. Fleischer, Biochim. Biophys. Acta, 47,358

(2) L. Szarkowska, Arch. Biochem. Biophys., 113,519 (1966). (3) M. Klingenberg and A. Kroger, Biochem. Mitochondria, Colloq.,

(4) F. L. Crane, Biochem. Quinones, 183 (1965). (5) P. J. Rietz, F. S. Skelton, and K. Folkers, Int. 2. Vitaminforsch.,

37,405 (1967). (6) F. S. Skelton, P. J. Rietz, and K. Folkers, J. Med. Chem., 13,

602 (1970). (7) F. S. Skelton, K. D. Lunan, K. Folkers, J. V. Schnell, W. A.

Siddiqui, and Q. M. Geiman, Biochemistry, 8, 1284 (1969). (8) J. V. Schnell, W. A. Siddiqui, Q. M. Geiman, F. S . Skelton,

K. D. Lunan, and K. Folkers, J. Med. Chem., 14, 1026 (1971). (9) E. G. Ball, C . B. Anfinsen, and 0. Cooper, J. Biol. Chem., 168,

(1961).

1966,ll (1967).

. . 257 (1947).

110) F. S. Skelton. R. S. Pardini. J. C. Heidker, and K. Folkers, J. .- -, Amer. Chem.'Soc., 90,5334 (1968).

K. Folken, ibid., 90,3572 (1968).

Biophys., 120,539 (1968).

ibid., 142,407 (1971).

Commun., 34,200 (1968).

(11) J. C. Catlin, R. S. Pardini, G. D. Daves, Jr., J. C. Heidker, and

(12) G. Lenaz, G. D. Daves, Jr., and K. Folkers, Arch. Biochem.

(13) G. Lenaz, A. Castelli, G. P. Littarm, E. Bertoli, and K. Folkers,

(14) A. Castelli, G. Lenaz, and K. Folkers, Biochem. Biophys. Res.

(15) P. B. Blair,Methods Enzymol., 10,78 (1967). (16) E. Layne, ibid., 3,47 (1957). (17) A. Castelli, E. Bertoli, G. P. Littarru, G. Lenaz, and K. Folkers,

Biochem. Biophys. Res. Commun., 42,806 (1971).


.beta.-Adrenergic blocking agents of the chromone and xanthone groupsP. Da Re, P. Valenti, A. Borraccini, and G. P. Primofiore







198 Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 Notes

0-Adrenergic Blocking Agents of the Chromone and Xanthone Groups P. Da Re,* P. Valenti, A. Borraccini, and G. P. Primofiore Institute of General Chemishy, University of Pisa, 561 00 Pisa, Italy. Received March 23, I971

The search for new P-adrenergic blocking drugs with enhanced therapeutic properties is still very active' and there has recently been an increasing interest in heterocyclic (e.g., b e n ~ o d i o x a n e , ~ , ~ indole,4,' benzofuran,6 dibenzofuran,' quinoline, etc.8) derivatives. We now report the preparation of some new &adrenergic blocking derivatives of the chromone and flavone group, with which we wish to illustrate further the versatility of the benzo-y-pyrone molecule as a carrier moiety in medicinal research. We have also prepared 2 analogous xanthone derivatives (19,20) on the basis of the results obtained among the CNS stimulants of the same

The basic chain was located, according to the synthetic possibilities, in position 6 for 4, 6, 10, and 12

R'

4, R = CH3; R' = i-C3H7 6, R = C6H,; R' = i-C3H7

19, R' = i-CJr, 20, R' = tert-C4H9

10; R = CH3; R' = tertIC4H9 12, R = C6HS; R' = tert-CJI,

and in the 2 position for 19 and 20. The synthesis of these compds was carried out from the corresponding Ac derivatives, prepared by standard methods, by bromination, amination, and reduction of the final amino ketone intermediates. As an alternative, the reductive amination of the glyoxalyl derivatives (obtained by SeOz oxidation of the same starting materials) according to Fodor and Kovacs" gave unsatisfactory results. In Table I all the chromone and flavone derivatives prepared are shown, while the xanthone

analogs are described in detail in the Experimental Section for illustrative purposes.

Compds 4,6, 10, 12, 19, and 20 have been evaluated for their fl-adrenergic blocking activity in a number of tests at the Institute of Pharmacology of the University of Padua. The N-isopropyl derivatives were significantly more active than the tert-Bu analogs; the activity of the former group decreases in the order 4 > 19 > 10. In this group, 6 - [ 1- hydroxy-2-isopropylaminoethyl] -2,3-dime thylchromone (4), was the most active in all tests employed.

(a) Lipomobilizing Activity. l2 The norepinephrine-induced lipid mobilization in rats is competitively blocked by equimolecular doses of 4, down to 2 X IO-'M.

of the short-circuited current in isolated frog skin is an- tagonized (60%) by 4 at 10-6M.

(c) Contraction Rates of Isolated Right Atrial Strips. The chronotropic effect of 0.001 pg/ml of isoproterenol is reduced to about half and completely abolished with doses of 5 and, respectively, 10 pg/ml of 4.

(d) Lengthening of the Refractory Period of the Isolated Guinea Pig Auricles.'6 An average 20% (ECZo) reduction in the maximal driven rate (MDR) at which the isolated guinea pig auricles respond to electrical stimulation is produced by 10 pg/ml of 4. On the basis of the ECzo values, propranolol is approximately 10 times more active.

(e) Isolated Guinea Pig Tracheal Chain." The isoproterenol reduction of the contraction of the isolated guinea pig tracheal chain caused by carbachol (0.5 pg/ml) is an- tagonized by very small doses of 4. Figure 1 reports cumulative log concn response curves for the agonist (isoproterenol) in the presence of various concns of the antagonist

(b) Isolated Short-circuited Frog Skin. 13,14 The increase

(4). ( f ) Blood Pressure in Anesthetized Dogs." Iv adminis-

tration of 1 and 5 mg/kg of 4 causes a 100% and 87% reduction of the hypotensive response to 0.2 and, respectively. I pg/kg of iv isoproterenol. The same doses of 4 cause a slight and transient hypotensive response of the order of -5 and - 10 mm.

Compd 4 is therefore a selective P-adrenergic blocking

I__.____ ____-_- Table I. Chromone and Flavone Derivatives Compdd R R' X MP, "C Formula Analyses

--___ _I__ -_______--

RfCHZX&t3

1 CH, H co 136-138' I 3H1203 C, H 2 CH3 3 CH3 4 CH3

NH-tert-C,H,. HC1 co 5 CH3 6 CH, 7 C6H5 8 C6H5 Br CO

Br co 143-146b 13H1 lBr03 C, H, Br

NH-CC,H, . HC1 CHOH

NH-tert-C,H, . HC1 H co

C, H, C1, N C, H, C1, N C, H, C1, N C, H, C1, N

NH-i-C,H, ' HCI co 220-223' 1 6H20CM03

233-235' 1 6H22C1N03

200-202' c l,H,2C1NO, CHOH 244-247' c132,C1NO,

154-156' C18HL403 C, H 2 16-2 17 C18H13Br03 C, H, Br

C, H, C1, N 202-204' c21H24CM03 C, H, C1, N

co 257-259' C22H24C1N03 C, H, C1, N C6H5 C, H, C1, N

NH-i-C3H, . HC1 co 137-139' c2 lH2ZClN0 3 9 C6H5

10 C6H5 NH-CC3H7 HCI CHOH NH-tert-C,H,. HC1 11

12 C6H5 NH-tert-C,H, ' HCl CHOH 269-270c C22H26C1N03 I

Crystn solvent: 'ligroin, bEtOAc, CMeOH-Et20. dThe bases corresponding to compds 4,6, 10, and 12 melted respectively (ligroin) at 105-107", 157-159", 121-122", 146-148".

Notes Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 199

90

" 1 " 1 " 1 M

I I 7 6 q -*

1 0 1 0 1 0 ID Io 1 0 ' - 6 .o

1 0

Figure 1. Cumulative log concentration-response curves for isoproterenol in the presence of various concns (a, 0 M;A, 3.2 X lO-'M;

, 3.2 x 10-6M;v, 3.2 x 10-SM) of 6-[l-hydroxy-2-isopropylaminoethyl] -2,3-dimethylchromone. Abscissa: molar concns of isoproterenol. Ordinate: per cent of maximal release.

agent of the propranolol type,I8 with membrane activity (test d) and devoid of intrinsic sympathomimetic activity (test f). In comparison with propranolol, 4 has a potency ratio of 0.1 (test d), but its LDS0 [223 mg/kg (205.5-242.0) iv in albino mice] is 2.5 times lower.

The data reported seem to confirm the pharmacological potentialities of the benzo-y-pyrone molecule also in this field of medicinal chemistry.

Experimental Section All melting points were detd in open glass capillaries, using a

Biichi apparatus, and are uncor. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within +0.4% of the theoretical values.

The general methods of synthesis described are illustrative of those of analogous compounds.

2-Propionyl-4-ace~lphenol(l3). To a soln of 25 ml of o-hy- droxypropiophenone and 25 ml of AcCl in 100 ml of CSd, 95 g of AlC1, was added in 1 hr, and the mixt was kept at 50-60 for 1 hr. After removal of the solvent the residue was treated with ice and HCl and extd with CHCl,. The CHC1, layer was dried and filtered, and the solvent was evapd. The residue was distd at 175-180' (4-5 mm); the collected oil solidified on cooling. On crystg from ligroin, 23 g of white product (mp 64-65') was obtd. Anal. (C,,H,,O,) C, H.

pionyl-4-acetylpheno1(13), 6 g of anhyd NaOAc, and 10 ml of Ac,O was heated in an oil bath at 170-180' for 7 hr. The fused mixt was taken up in H,O and the sepd solid was collected, washed (H,O), and dried. On crystg from ligroin, 8 g of yellow product, mp 136-138", was obtd. Anal. (C,,H,,O,) C, H.

acetylphenol(13), 30 g of BzCl, and 45 g of PhCOONa was heated in an oil bath at 180-190" for 7-8 hr. The reaction mixt was taken up in H,O, washed (NaOH-H,O), and extd with CHCl,. Removal of the solvent left a residue which on crystg from ligroin, gave 15 g of yellow solid, mp 154-156". Anal. (C,,H,,O,) C, H.

0-chlorobenzoic acid, 6 g of p-hydroxyacetophenone, 4 g of NaOH, and 0.5 g of Cu powder was heated in an oil bath at 200" for 0.5 hr. The mixt was poured into ice water and acidified with dil HCl, and the sepd product was isolated by filtration, dissolved (NaHCO,), and repptd with dil HC1. The crude product, on crystg from EtOH- H,O, gave 6 g of white cryst solid, mp 185-186". Anal. (C,,H,,OJ C, H.

2-Acetylxanthone (15). To a soh of polyphosphoric acid (from 100 g of P,O, and 100 g of 85% %PO, ), 9.3 g of 14 was added in small portions. The reaction mixt was kept on a steam

6-Acetyl-2,3-dimethylchromone (1). A mixt of 10 g of 2-pro-

6-Acetyl-3-methylflavone (7). A mixt of 15 g of 2-propionyl-4-

2-Carboxy-4'-acetyldiphenyl Ether (14). A mixt of 7.8 g of

bath for 1.5 hr and then poured into ice H,O. The sepd solid was collected, washed (NaHC0,-%O), and dried. On crystg from EtOH, 5.5 g of white solid, mp 199-200", was obtd. Anal. (C,,H,,O,) C, H.

2-Bromoacetylxanthone (16). To a soh of 4.8 g of 15 in 300 ml of CHCl,, a soln of 3.2 g of Br, in 75 ml of CHCl, was added, with stirring, in 2 hr. The reaction mixt was transferred to a separ- atory funnel, washed (dil NaOH-H,O), and dried, and the solvent was evapd. The residue, after 2 crystns from EtOAc, gave 3 g of white solid, mp 194-196". Anal. (C,,H9Br0,) C, H, Br.

of 17 g of 16 in 1.5 1. of PhH, a slight excess of i-PrNH, was added, and the mixt was kept at room temp with stirring for 4 hr. The soln was washed (H,O) and dried. The PhH layer, treated with HC1 gas, gave the amino ketone as the HCl salt. On crystallg the crude product from MeOH-Et,O, 8 g of white solid, mp 210-211", was obtd. Anal. (C,8H,,ClN0,) C, H, Cl, N.

2-tert-Butylaminoacety~anthone Hydrochloride (1 8). In a similar manner, starting from 7.3 g of 2-bromoacetylxanthone (16), 3 g of the corresponding amino ketone hydrochloride (18), mp 189-193" dec (MeOH-Et,O), was obtained. Anal. (C,,H,,ClNO,) C, H, C1, N.

2- [ l-Hydroxy-2-isopropylaminoethyl] xanthone Hydrochloride (19). A soln of 4.1 g of 17 in 40 ml of MeOH was hydrogenated over 10% Pd/C until H, uptake ceased. The soln was filtered from the catalyst and evapd to dryness. The residue, on crystg from MeOH-Et,O gave 2.5 g of white product, mp 224-226" [Anal. (C,,H,,ClNO,) C, H, C1, N] ; base, white cryst solid, mp 108-1 10" (ligroin) [Anal. (C,,H,,NO,) C, H, N].

2- [ I-Hydroxy-2-tert- butylaminoethyl]xanthone Hydrochloride (20). With the same procedure 3 g of rert-butylaminoacetyl- xanthone hydrochloride (18) gave 1.7 g of the corresponding amino alcohol hydrochloride as a white solid, mp 230-233" (MeOH-Et ,O>; [Anal. (C,,H,,CINO,) C, H, C1, N] ;base, white cryst solid, mp 112-114' (ligroin) [Anal. (C,JI,,NO,) C, H, N].

R. Santi, Director of the Institute of Pharmacology, Uni- versity of Padua, for his kind permission to report some biological data.

2-Isopmpylaminoacetylxanthone Hydrochloride (17). To a soln

Acknowledgment. The authors are indebted to Professor

References (1) M. S. K. Ghouri and T. J. Haley, J. Pharm. Sei., 58,511 (1969). (2) G. Marchetti, L. Merlo, and V. Noseda, Arzneim.-Forsch., 18,

(3) R. Howe, B. S. Rao, and M. S. Chodnekar, J. Med. Chem., 13,

(4) J. H. Biel and B. K. B. Lum, Fortschr. Arzneimittelforsch., 10,

( 5 ) R. C. Hill and P. Turner, Brit. J. Pharmacol., 36,368 (1969). (6) R. C. Hill and P. Turner, Brit. J. Pharmacol. Chemother., 32,

663 (1968). (7) R. Wandestrick, C. Goldenberg, F. Binon, and R. Charlier,

Chim. Ther., 1970,285. (8) W. Hepworth, A. Mitchell, N. S. Chodnekar, and R. Howe,

French Patent 3697 (Dec 27, 1965); Chem. Abstr. 82119r (1967).

(9) P. Da Re, V. Mancini, E. Toth, and L. Cima, Arzneim.-Forsch., 18,718 (1968).

J. Med. Chem., 13,527 (1970).

(1949).

43 (1968).

169 (1970).

46 (1966).

(10) P. Da Re, L. Sagramora, V. Mancini, P. Valenti, and L. Cima,

(11) G. Fodor and 0. Kovacs, J. Amer. Chem. Soc., 71, 1045

(12) G. Fassina, Arch. Inf . Pharmacodyn., 166,281 (1967). (13) H. H. Ussig and K. Zerahn, ActaPhysiol. Scand., 23, 110

(14) G. Fassina, F. Carpenedo, and G. Fiandini, J. Pharm. Pharma-

(15) J. W. Blank, W. A. M. Duncan, and R. G. Shanks, Brit. J.

(16) F. P. Luduena, J. H. Howard, and J. K. Borland, Arch. Int.

(17) P. N. Patil, J. Pharmacol. Exp. Ther., 160,308 (1968). (18) J. D. Fitzgerald, Clin. Pharmacol. Ther., 10, 292 (1969).

(1951).

col., 20,240 (1968).

Pharmacol., 25,577 (1965).

Pharmacodyn., 107,335 (1956).


Synthesis of trimethoprim variations.Replacement of methylene by polar groupings

Eugene L. StogrynJ. Med. Chem., 1972, 15 (2), 200-201• DOI: 10.1021/jm00272a019 • Publication Date (Web): 01 May 2002






200 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 Notes

Synthesis of Trimethoprim Variations. Replacement of CHp by Polar Groupings? Eugene L. Stogryn Government Research Laboratory, Esso Research and Engineering Company, Linden, New Jersey. Received February 22, I971

The antimalarial efficacy of 5-ary1-2,4-diaminopyrimi- dines, as a function of the electronic demands of aryl substituents, was defined by the extensive studies of Hitchings' and coworkers. Although their work also indicated an activity dependence on the nature of the functionality uniting the aryl and pyrimidine rings, comparatively few studies have dealt with this structural parameter.

An examination of this parameter in trimethoprim [2,4- diamino-5-(3,4,5-trimethoxybenzyl)pyrimidine] congeners appeared particularly pertinent in light of its recently reported activity against drugresistant strains of malaria.' We wish to describe structural variations of trimethoprim in which the CH2 group that connects the aryl and pyrimidine rings has been replaced by polar functionality.

The target compounds listed in Table I were prepared by

was a competing nucleophile. Displacement of both C1 required high temp and pressure amination techniques.

with 2,4,5-triaminopyrimidine to yield 7 was intolerably slow and isolation of product difficult. These synthetic de- ficiencies were alleviated by an initial preparation of the Schiff s base from 3,4,5-trimethoxybenzaldehyde and i-PrNH2. An amine exchange between this Schiff's base and 2,4,5-triaminopyrimidine occurred readily, giving good yields of 7.

Compds 9 and 10, obtained virtually quantitatively. appear to undergo facile thermal alteration.$ Purification by sublimation or recrystallization failed. Attempts to sub- lime either 9 or 10 resulted in the isolation of 3,4,5-trimethoxyaniline, with destruction of 9 and 10.

Biological Tests.§ The target structures did not affect the mean survival time of Plasmodium berghei infected mice sufficiently, at the maximum dose levels (640 mg/kg), to be classified as active. Compds 2, 3, and 6-9 in the avian activity screen (P. gallinaceum) failed to improve the survival time of the test animals beyond day 1, at a dose level of 320 mg/kg. At a drug concn of 0.1%, 1 produced 100% abnormal

The direct condensation of 3,4,5-trimethoxybenzaldehyde

-. Table I. Antimalarial act,*Vc Recrystn

solvent Formula Analysisa A ST, days No. X MP, "C

1 2 3 4 5 6 7 8 9

10 -

0 N = N ~ C O N H ~

co, - C H , N H ~ ~ ~

NHCOg

CH=N

CH,NH NHCONHg NHCSNH

204-205 270-272 140-142 226-228 222-224 185-185.5 169-172 193.5-195 i i

OMe

NHl THF '1 ,H16N404

H,O C13H,8N607S

HlO C14H19NS0S

HlO C14H1$505

EtOH 14H16N405

PhH 7N503

C18H27N507

PhH C14H19N503

C14H20N605

14H I sN6O 3'

e 0.2 0.1 3.3 0.2 2.5 0.7 0 0.5 0.6

a m e r e analyses are indicated by symbols of elements analytical results were obtained for those elements within ?0.4% of the theoreticai value. bMice were treated 3 days postinfection sc with a single dose of the compd being screened. The change in survival time (AST) is an indication of activity against P. berghei (see ref 6). CAST at 640 mg/kg. dC, calcd, 53.42; found 53.85. eNot evaluated in the rodent screen. fH,S04 salt. ghlonohydrate. hN, calcd, 20.77; found 21.33. iHOA~ salt. iDecompd on heating.

the synthetic techniques detailed in the Experimental Sec- tion. Several of these syntheses were atypical and merit comment. Although aminolysis of chloropyrimidines' are generally high temp processes, the known promotional influence of carbonyl functionality in nucleophilic displace- ments presented the possibility of obtaining 4 by treating

Cl 4a

4a with NH,-EtOH at atm pressure. Reaction of 4a with refluxing NH,-EtOH over extended time periods failed to effect complete amination of 4a. Spectral data of the reaction product suggested that under these conditions, EtOH

tThis work was supported by the U. S. &my Medicinal Research and Development Command under Contract No. DADA1 7-68-C- 8103. This is Contribution No. 901 from the Army Research Pro- gram on Malaria.

oocyst and 100% suppression of oocyst in the mosquito (Aedes aegypti) screen. However, the next lower dose level (0.01%) failed to produce a response.

Experimental Section 2,4-Diamino-5-(3,4,5-trimethoxyphenoxy)pyrimidine (1). A

mixt of 16.3 g of 3,4,5-trimethoxyphenol, 6.6 g of chloroaceto- nitrile, and 12.2 g of K,C03 in 60 ml of dry Me,CO was refluxed overnight. The mixt was added to 140 ml of 3% NaOH and extd with Et,O. Acidification of the aq phase yielded 3 g of 3,4,5-trimethoxyphenol. Evapn of the Et,O gave 14.3 g (88%) of 3,4,5- trimethoxyphenoxyacetonitrile, mp 65.5-66' (PhH-hexane).

To a slurry of NaH (3.7 g of 50% NaH) in 130 ml of PhH was added a soln of 16.7 g of the phenoxyacetonitrile and 11.2 g of ethyl formate. After stirring at ambient temp overnight 100 ml of H,O was added. Neutralization of the aq phase and extn with Et,O gave 10.5 g of a viscous oil. Treatment of this oil in 1 1. of Et,O

$Intramolecular cyclization of 5-ureido and -thioureidopyrimi- dines with expulsion of amines, are known synthetic entries into purines. For discussion of purines see ref 3, p 339 and ref 4.

#The rodent and avian activity screens were performed by Dr. L Rane. The mosquito activity screen was conducted by Dr. E. Ger- berg. The significance of this screen is described in ref 5 .

Notes Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 201

with 5.3 g of CH,N, gave 8.9 g of a dark viscous oil which gradually solidified, mp 110-1 11" (MeOH).

A soh of guanidine (from 9.7 g of guanidine hydrochloride and 5.4 g of NaOMe) and 8.9 g of the above described methylated aldehydonitrile in 125 ml of MeOH was refluxed for 18 hr. The MeOH was evapd and the residue extd with hot THF. The residue from the THF soh was chromatogd through base-treated silica gel. The product, 1.2 g, was found in the second Me,CO eluate.

2,4-Diamino-5-(3,4,5-trimethoxyphenylazo)pyrimidine Sulfate (2). To a soh of 20.6 g of 3,4,54rimethoxyaniline in 250 ml of 1 N HCl at 0" was added 8 g of NaNO, in 40 ml of H,O. This soln was added to 11 g of 2,4-diaminopyrimidine in 110 ml of H,O at 0". After 5 min sufficient NaHCO, (about 35 g) was added to raise the pH to 8. The red-brown solids, formed after standing overnight, were filtered and extd with 500 ml of boiling 1 N H,SO,. On cooling the %SO, soln pptd 3 g of 2 as a yellow solid.

2,4-Diamino-5-(3,4,5-trimethoxybenzamido)pyrimidine Mono- hydrate (3). To a cold soln of 2,4,5-triaminopyrimidine7 (1 g) and 1.5 g of Et,N in 20 ml of H,O was added 1.95 g of 3,4,5-trimethoxybenzoyl chloride in 20 ml of THF. The mixt was stirred for 20 min and concd. The pptd solid was filtered and recrystd, yield 2 g.

2,4-Dichloro-S-pyrimidinecarbonyl Chloride. 2,4-Dihydroxy- 5-pyrimidinecarboxylic acid was converted into the titled compound by the procedure Gershon* used to prep the analogous 6-pyrimidinecarbonyl chloride, bp 90-91" (0.3 mm), 50% yield. Anal. (C,HCl,N,O) N, Cl.

Et,N (1.62 g) and 2.84 g of 3,4,5-trimethoxyaniline in 100 ml of THF was added to a cold soln of 3.3 g of 2,rldichloro-5-pyrimidinecarbonyl chloride in 50 ml of THF. After standing ovemight the reaction was filtered and the THF evapd. The residue was washed with cold dil NaHCO, and recrystd from PhH-hexane, mp 196-198". Anal. (C,,H,,Cl,N,O,) N, C1.

2,4-Diamino-5-pyrimidinecarbox-3,4,5-trimethoxyanilide Monohydrate (4). A mixt of 5 g of 4a, 13 ml of NH,OH, and 66 ml of liquid NH, was heated in an autoclave for 8 hr at 180". The bomb was cooled, and the contents were filtered and recrystd, yield 3 g.

2,4-Diamino-S -( 3,4,5 -trimethox ybenzoy loxy )p yrim idine (5). To a cold soln of 1 g of 2,4diamino-S-hydroxypyrimidine hydro- chlorideQ and 3.1 g of Et,N in 20 ml of THF-H,O was added 1.36 g of 3,4,5-trimethoxybenzoyl chloride in 15 ml of THF. After 20 min the reaction mixt was concd, poured into a large vol of H,O, filtered, H,O washed, and recrystd, yield, 1.8 g.

3,4,5-Trimethoxybenzylideneisopmpylamine. A soln of 40 g of 3,4,5-trimethoxybenzaldehyde and 35 ml of i-PrNH, in 200 ml of THF was refluxed for 5 hr. Evapn of solvent left 48 g of a white solid, mp 76577 .5" (MeOH-H,O). Anal. (C,,H,$rlO,) C, H, N.

2,4-Diamino-5-( 3,4,5-trimethoxybenzylideneamino)pyrimidine (6). An EtOH soln, 350 ml, contg 15.4 g of 3,4,5-trimethoxyben- zylideneisopropylamine and 8.1 g of 2,4,5-triaminopyrimidine was refluxed for 20 hr and then allowed to stand at room temp for 2 days. The pptd yellow solid was filtered and recrystd. Add1 quants of 6 could be obtd by concn of the filtrate.

2,4-Diamino-5-(3,4,5-trimethoxybenzylamino)pyrimidine Acetate Salt (7). Low pressure hydrogenation of 2.5 g of 6 in 140 ml of HOAc over PtO, occurred in 15 min. The reaction mixt was filtered, poured into a large vol of Et,O, filtered, and repeatedly washed with Et,O, yield 2 g.

2,4-Diamino-5-( 3,4,5-trimethoxybenzylamino)pyrimidine (8). An H,O soln of 7 was made basic with NH,OH, cooled, filtered, and recrystd.

g in 125 ml of PhMe was added 23.1 g of 3,4,5-trimethoxybenzoyl chloride in 250 ml of PhMe. Stirring was contd until the characteristic acyl halide ir absorption was absent. The mixt was then heated on a steam bath until N, evoln ceased. It was cooled, filtered, and distd (14.4 g), bp 120-140" (0.1 mm). This product solidified on standing, mp 43-43.5". Anal. (C,JI,,NO,) C, H, N.

3,4,5-Trimethoxyphenyl Isothiocyanate. To CSCI, (25 g) suspended in 180 ml of ice H,O was added 30.7 g of 3,4,5-trimethoxyaniline in 300 ml of CHCl,, maintg the temp below 5". After 15 min the CHC1, layer was sepd and dried over CaCl,. Distn gave 24.8 g of product, bp 145" (0.5 mm). The distillate solidified on standing, mp 63" (hexane). Anal. (C,,H,,NO,S) C, H, S.

Monohydrate (9). To a soln of 2 g of 2,4,5-triaminopyrimidine in 150 ml of 50% EtOH was added 2.8 g of 3,4,54rimethoxyphenyl isocyanate. After several hours stirring at room temp the ppt was filtered and washed d t h H,O and EtOH, yield 4 g.

2,4-Dichloro-5-pyrimidinecarbox-3,4,5-tnmethoxyanilide (4a).

3,4,5-Trimethoxyphenyl Isocyanate. To a slurry of NaN,, 19.5

1-[ 5-(2,4-Diaminopyrimidiny1)]-3-(3,4,5-trimethoxyphenyl)urea

1- [ 5-(2,4-Diaminopyrimidinyl)] -3-(3,4,5-trimethoxyphenyl)- thiourea (10). Powdered 3,4,5-trimethoxyphenyl isothiocyanate (2 g) was added to a soh of 2,4,5-triaminopyrimidine in 40 ml of 75% EtOH. After 1 hr stirring the ppt was filtered and washed thoroughly with H,O and EtOH.

References (1) (a) E. A. Falco, P. B. Russell, and G. H. Hitchings, J. Amer.

Chem. Soc., 73, 3753 (1951); (b) E. A. Falco, S. DuBreuil, and G. H. Hitchings, ibid., 73, 3758 (1951); (c) P. B. Russell and G. H. Hitchings, ibid., 73, 3763 (1951); (d) F. R. Gerns, A. Perrotta, and G. H. Hitchings, J. Med. Chem., 9,108 (1966).

(2) D. C. Martin and J. D. Arnold, J. Amer. Med. Ass., 203,476 (1968).

(3) D. J. Brown, "The Pyrimidines," Interscience, New York, N. Y.,

(4) R. K. Robins, Heterocycl. Compounds, 8,232 (1967). (5) E. Gerberg, L. T. Richard, and J. T. Poole, Mosquito News, 26,

(6) T. S. Osdene, P. B. Russell, and L. Rane, J. Med. Chem., 10,431

(7) D. J. Brown,J. Appl. Chem., 112 (1957). (8) H. Gershon, J. Org. Chem., 27, 3507 (1962). (9) R. Hul1,J. Chem. Soc., 2033 (1956).

1962, pp 153-154.

359 (1966).

(1967).

Preparation and Antitumor Activity of Derivatives of 1 -Phenyl-3,3-dimethyltriazeneP Ying-Tsung Lin, Ti Li Loo,* Department of Developmental Therapeutics, The University of Texas at Houston, M. D. Anderson Hospital and Tumor Institute, Houston, Texas

Srikrishna Vadlamudi, Microbiological Associates, Inc., Bethesda, Maryland

and Abraham Goldin National Cancer Institute, National Institutes of Health, Bethesda, Maryland. Received May 20, 19 71

A number of 5-triazenoimidazoles are active against experimental tumors.' An excellent and comprehensive review of this subject has recently appeared.2 Of the 5-triazenoimidazoles, 5-(3,3-dimethyl- 1 -triazeno)imidazole-4-carboxamide (DIC) in particular, is clinically useful for the induc- tion of temporary remission in malignant m e l a n ~ m a . ~ The mechanism of the antitumor action of DIC remains obscure. However, on the basis of studies of the biotransformation of DIC4>' and the carcinogenesis of phenyltriazenes, it has been proposed that the triazenes may act as alkylating agents through the in vivo generation of carbonium ions.6 Nevertheless, it is entirely possible that as a derivative of 5- aminoimidazole-4-carboxamide (AIC), the precursor of the purine base, 5-triazenoimidazole may somehow-interfere with imidazole and purine metabolism.'

of phenyltriazene.' To extend these observations and, above all, to elucidate the structural requirements for antitumor activity in the triazenes and specifically to ascertain whether an imidazole ring with a carboxamide moiety ortho to the triazeno side chain is indispensable in DIC, we have synthesized 6 derivatives of 1 -phenyl-3,3-dimethyltriazene (la-If). Except for

Antitumor activity has been observed in several derivatives

these compounds have not been

?Supported in part by Contracts PH 43-66-1 156 and PH 43-68- 1283 with Chemotherapy, National Cancer Institute, National In- stitutes of Health, U. S. Public Health Service.


Preparation and antitumor activity ofderivatives of 1-phenyl-3,3-dimethyltriazene

Ying-Tsung Lin, Ti Li Loo, Srikrishna Vadlamudi, and Abraham GoldinJ. Med. Chem., 1972, 15 (2), 201-203• DOI: 10.1021/jm00272a020 • Publication Date (Web): 01 May 2002






Notes Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 201

with 5.3 g of CH,N, gave 8.9 g of a dark viscous oil which gradually solidified, mp 110-1 11" (MeOH).

A soh of guanidine (from 9.7 g of guanidine hydrochloride and 5.4 g of NaOMe) and 8.9 g of the above described methylated aldehydonitrile in 125 ml of MeOH was refluxed for 18 hr. The MeOH was evapd and the residue extd with hot THF. The residue from the THF soh was chromatogd through base-treated silica gel. The product, 1.2 g, was found in the second Me,CO eluate.

2,4-Diamino-5-(3,4,5-trimethoxyphenylazo)pyrimidine Sulfate (2). To a soh of 20.6 g of 3,4,54rimethoxyaniline in 250 ml of 1 N HCl at 0" was added 8 g of NaNO, in 40 ml of H,O. This soln was added to 11 g of 2,4-diaminopyrimidine in 110 ml of H,O at 0". After 5 min sufficient NaHCO, (about 35 g) was added to raise the pH to 8. The red-brown solids, formed after standing overnight, were filtered and extd with 500 ml of boiling 1 N H,SO,. On cooling the %SO, soln pptd 3 g of 2 as a yellow solid.

2,4-Diamino-5-(3,4,5-trimethoxybenzamido)pyrimidine Mono- hydrate (3). To a cold soln of 2,4,5-triaminopyrimidine7 (1 g) and 1.5 g of Et,N in 20 ml of H,O was added 1.95 g of 3,4,5-trimethoxybenzoyl chloride in 20 ml of THF. The mixt was stirred for 20 min and concd. The pptd solid was filtered and recrystd, yield 2 g.

2,4-Dichloro-S-pyrimidinecarbonyl Chloride. 2,4-Dihydroxy- 5-pyrimidinecarboxylic acid was converted into the titled compound by the procedure Gershon* used to prep the analogous 6-pyrimidinecarbonyl chloride, bp 90-91" (0.3 mm), 50% yield. Anal. (C,HCl,N,O) N, Cl.

Et,N (1.62 g) and 2.84 g of 3,4,5-trimethoxyaniline in 100 ml of THF was added to a cold soln of 3.3 g of 2,rldichloro-5-pyrimidinecarbonyl chloride in 50 ml of THF. After standing ovemight the reaction was filtered and the THF evapd. The residue was washed with cold dil NaHCO, and recrystd from PhH-hexane, mp 196-198". Anal. (C,,H,,Cl,N,O,) N, C1.

2,4-Diamino-5-pyrimidinecarbox-3,4,5-trimethoxyanilide Monohydrate (4). A mixt of 5 g of 4a, 13 ml of NH,OH, and 66 ml of liquid NH, was heated in an autoclave for 8 hr at 180". The bomb was cooled, and the contents were filtered and recrystd, yield 3 g.

2,4-Diamino-S -( 3,4,5 -trimethox ybenzoy loxy )p yrim idine (5). To a cold soln of 1 g of 2,4diamino-S-hydroxypyrimidine hydro- chlorideQ and 3.1 g of Et,N in 20 ml of THF-H,O was added 1.36 g of 3,4,5-trimethoxybenzoyl chloride in 15 ml of THF. After 20 min the reaction mixt was concd, poured into a large vol of H,O, filtered, H,O washed, and recrystd, yield, 1.8 g.

3,4,5-Trimethoxybenzylideneisopmpylamine. A soln of 40 g of 3,4,5-trimethoxybenzaldehyde and 35 ml of i-PrNH, in 200 ml of THF was refluxed for 5 hr. Evapn of solvent left 48 g of a white solid, mp 76577 .5" (MeOH-H,O). Anal. (C,,H,$rlO,) C, H, N.

2,4-Diamino-5-( 3,4,5-trimethoxybenzylideneamino)pyrimidine (6). An EtOH soln, 350 ml, contg 15.4 g of 3,4,5-trimethoxyben- zylideneisopropylamine and 8.1 g of 2,4,5-triaminopyrimidine was refluxed for 20 hr and then allowed to stand at room temp for 2 days. The pptd yellow solid was filtered and recrystd. Add1 quants of 6 could be obtd by concn of the filtrate.

2,4-Diamino-5-(3,4,5-trimethoxybenzylamino)pyrimidine Acetate Salt (7). Low pressure hydrogenation of 2.5 g of 6 in 140 ml of HOAc over PtO, occurred in 15 min. The reaction mixt was filtered, poured into a large vol of Et,O, filtered, and repeatedly washed with Et,O, yield 2 g.

2,4-Diamino-5-( 3,4,5-trimethoxybenzylamino)pyrimidine (8). An H,O soln of 7 was made basic with NH,OH, cooled, filtered, and recrystd.

g in 125 ml of PhMe was added 23.1 g of 3,4,5-trimethoxybenzoyl chloride in 250 ml of PhMe. Stirring was contd until the characteristic acyl halide ir absorption was absent. The mixt was then heated on a steam bath until N, evoln ceased. It was cooled, filtered, and distd (14.4 g), bp 120-140" (0.1 mm). This product solidified on standing, mp 43-43.5". Anal. (C,JI,,NO,) C, H, N.

3,4,5-Trimethoxyphenyl Isothiocyanate. To CSCI, (25 g) suspended in 180 ml of ice H,O was added 30.7 g of 3,4,5-trimethoxyaniline in 300 ml of CHCl,, maintg the temp below 5". After 15 min the CHC1, layer was sepd and dried over CaCl,. Distn gave 24.8 g of product, bp 145" (0.5 mm). The distillate solidified on standing, mp 63" (hexane). Anal. (C,,H,,NO,S) C, H, S.

Monohydrate (9). To a soln of 2 g of 2,4,5-triaminopyrimidine in 150 ml of 50% EtOH was added 2.8 g of 3,4,54rimethoxyphenyl isocyanate. After several hours stirring at room temp the ppt was filtered and washed d t h H,O and EtOH, yield 4 g.

2,4-Dichloro-5-pyrimidinecarbox-3,4,5-tnmethoxyanilide (4a).

3,4,5-Trimethoxyphenyl Isocyanate. To a slurry of NaN,, 19.5

1-[ 5-(2,4-Diaminopyrimidiny1)]-3-(3,4,5-trimethoxyphenyl)urea

1- [ 5-(2,4-Diaminopyrimidinyl)] -3-(3,4,5-trimethoxyphenyl)- thiourea (10). Powdered 3,4,5-trimethoxyphenyl isothiocyanate (2 g) was added to a soh of 2,4,5-triaminopyrimidine in 40 ml of 75% EtOH. After 1 hr stirring the ppt was filtered and washed thoroughly with H,O and EtOH.

References (1) (a) E. A. Falco, P. B. Russell, and G. H. Hitchings, J. Amer.

Chem. Soc., 73, 3753 (1951); (b) E. A. Falco, S. DuBreuil, and G. H. Hitchings, ibid., 73, 3758 (1951); (c) P. B. Russell and G. H. Hitchings, ibid., 73, 3763 (1951); (d) F. R. Gerns, A. Perrotta, and G. H. Hitchings, J. Med. Chem., 9,108 (1966).

(2) D. C. Martin and J. D. Arnold, J. Amer. Med. Ass., 203,476 (1968).

(3) D. J. Brown, "The Pyrimidines," Interscience, New York, N. Y.,

(4) R. K. Robins, Heterocycl. Compounds, 8,232 (1967). (5) E. Gerberg, L. T. Richard, and J. T. Poole, Mosquito News, 26,

(6) T. S. Osdene, P. B. Russell, and L. Rane, J. Med. Chem., 10,431

(7) D. J. Brown,J. Appl. Chem., 112 (1957). (8) H. Gershon, J. Org. Chem., 27, 3507 (1962). (9) R. Hul1,J. Chem. Soc., 2033 (1956).

1962, pp 153-154.

359 (1966).

(1967).

Preparation and Antitumor Activity of Derivatives of 1 -Phenyl-3,3-dimethyltriazeneP Ying-Tsung Lin, Ti Li Loo,* Department of Developmental Therapeutics, The University of Texas at Houston, M. D. Anderson Hospital and Tumor Institute, Houston, Texas

Srikrishna Vadlamudi, Microbiological Associates, Inc., Bethesda, Maryland

and Abraham Goldin National Cancer Institute, National Institutes of Health, Bethesda, Maryland. Received May 20, 19 71

A number of 5-triazenoimidazoles are active against experimental tumors.' An excellent and comprehensive review of this subject has recently appeared.2 Of the 5-triazenoimidazoles, 5-(3,3-dimethyl- 1 -triazeno)imidazole-4-carboxamide (DIC) in particular, is clinically useful for the induc- tion of temporary remission in malignant m e l a n ~ m a . ~ The mechanism of the antitumor action of DIC remains obscure. However, on the basis of studies of the biotransformation of DIC4>' and the carcinogenesis of phenyltriazenes, it has been proposed that the triazenes may act as alkylating agents through the in vivo generation of carbonium ions.6 Nevertheless, it is entirely possible that as a derivative of 5- aminoimidazole-4-carboxamide (AIC), the precursor of the purine base, 5-triazenoimidazole may somehow-interfere with imidazole and purine metabolism.'

of phenyltriazene.' To extend these observations and, above all, to elucidate the structural requirements for antitumor activity in the triazenes and specifically to ascertain whether an imidazole ring with a carboxamide moiety ortho to the triazeno side chain is indispensable in DIC, we have synthesized 6 derivatives of 1 -phenyl-3,3-dimethyltriazene (la-If). Except for

Antitumor activity has been observed in several derivatives

these compounds have not been

?Supported in part by Contracts PH 43-66-1 156 and PH 43-68- 1283 with Chemotherapy, National Cancer Institute, National In- stitutes of Health, U. S. Public Health Service.

202 Journal of Medicinal Chemistry, I9 72, Vol. IS, No. 2 Notes

la, R = NH2, o- b, R = NH,, m-

e, R = OCH3, m.

C, R = NHz, p- d, R = OCHj, O-

1 f , R = O C H j , p

N=NN(CHJ,

described previously. They were readily prepared by treating the corresponding diazotized methyl aminobenzoates and m- and p-aminobenzamide, respectively, with MezNH in Na2C03 ~ o l n . ~ However, because the diazotization of o-benzamide directly afforded 1,2,3-benzotriazin-4(3H)- one instead of the required diazo intermediate, l a was synthesized by a modified procedure. An ethyl benzoyl carbonate was prepared from o-(3,3-dimethyl-l -triazeno)- benzoic acid and ethyl chloroformate. Ammonolysis of the ethyl benzoyl carbonate which was not isolated gave la.

Antitumor Activities. In Table I we compare the anti-

dicates that the CONHz group in the phenyltriazenes may be replaced with C02CH3 with no adverse effect. Further, the orientation of either the CONH2 or the COzCH3 substituent with respect to the dimethyltriazeno side chain is not critical, as all 3 derivatives, (ortho, meta, and para) show comparable antileukemic activity. Finally, because these phenyltriazenes and DIC are equally active and because the pyrazole analog of DIC likewise displays antileukemic property,"'" it appears that the imidazole ring, the Ph ring, and the pyrazole ring are all equivalent therapeutically in the triazenes.

Experimental Section$

(0.13 mole) was added to a mixt of 100 g of crushed ice and 33 ml of concd HCl with vigorous stirring at 0-5". Diazotization was achieved by the slow addn, accompanied by thorough agitation, of NaNO, (9 g, 0.13 mole) dissolved in 25 ml of H,O. After standing

Derivatives of l-Phenyl-3,3dimethyltriazene. The arom amine,

Table I. Comparison of the Antitumor Activity of Phenyl Analogs of DIC with DIC Itself in Mouse Leukemia L1210a

Dose,b DIC l a l b IC Id l e If mg/kg MSTC I L S ~ MST ILS MST ILS MST ILS MST ILS MST ILS MST ILS

Treatment Day 1 only 833 12 (3) 20 5 00 11 (2) 10 2 0 2 0 2 0 10 (2) 0 7 (6) 0 300 11 (2) 10 12.5 (3) 25 6 (6) 0 10.5 (5) 5 10.5 (5) 5 11 (0) 10 14 ( 3 ) 40 180 10 (2) 0 11 (2) 10 1 3 (4) 30 12.5 (3) 25 12 (3) 20 11 (0) 10 1 1 (0) 10 108 11 (1) 10 11 (2) 10 12 (0) 20 10 (0) 0 10.5 ( 1 ) 5 10 (0) 0 1 1 (0) 10 65 10 (0) 0 10.5 (0) 5 10 (0) 0 11 (0) 10 10 (0) 0 10 (0) 0

Treatment Days 1 , 5 , and 9 180 12 (0) 20 10.5 (4) 5 10.5 (4) 5 12.5 (0) 25 15 (2) 50 108 11 (0) 10 10.5 (2) 5 13.5 (4) 35 14 (2) 40 1 3 (2) 30 11 (0) 10 12.5 (0) 25 65 11 (0) 10 10 (2) 0 12.5 (3) 25 13 (1) 30 11.5 (0) 15 10.5 (0) 5 13 (0) 30

Treatment Days 1-9 108 11.5 (3) 15 6 (3) 0 6 (3) 0 8 ( 3 ) 0 10 (2 ) 0 65 14 (2) 40 11 (2) 10 9.5 (2) 0 10.5 (2) 5 13 (3) 30 14 (1) 40 39 12 (0) 20 10 (1) 0 15 (2) 50 14 (1) 40 1 3 (3) 30 14 (1) 40 23 12.5 (0) 25 10 (1) 0 1 3 ( 1 ) 30 13 .5 (1) 35 11 (0) 10 13.5 (0) 35

=BDF, mice inoculated ip with lo5 L1210 leukemic ascites cells; 8 mice per group and 46 untreated controls. Median survival time of untreated controls: 10 days. bTreatment ip CMedian survival time, days. Number in parenthesis represents the average body loss in grams on the 6th day. %crease in life span, %. ILS of 20% or greater is suggestive of reproducible activity.

Table 11. Yield, Molecular

Compound % MP, "C formula" l a 56 127-13 1 'QH1 Z N 4 0 l b 73 145-146 ZN4' I C 78 176-178 CQH12N40 le 70 43-45 I$, 3N302

If 80 103-104 Cldr,,N,O, aAll compounds analyzed within *0.3% for C, H, N.

leukemic activity of the 1 -phenyl-3,3-dimethyltriazene derivatives with that of DIC. From these results, it is clear that although the tolerated dose range is lower for the Ph analogs, they are nevertheless as effective as DIC in increasing the survival time of the leukemic mice. In fact, at least one of these, If, may have a more favorable therapeutic index than DIC. The determination of the activity of each compound against a variety of transplantable tumors is currently in progress.

The fact that l-phenyl-3,3-dimethyltriazene with no annular substituent does not increase the survival of mice bearing L1210 leukemia (sensitive to methotrexate)8 would suggest that the carboxamide moiety is necessary for antileukemic activity. However, our present work in-

at 0-5" for 20 min, the above cold s o h contg the diazo compound was carefully added to Me,NH (40% in H,O, 15.8 g, 0.14 mole) and ice (75 g) with stirring, the temp being kept between 5 and 15". At the end of the mixing, the stirring was contd at room temp for an add1 2-3 hr. The light brown crude phenyltriazene derivative was collected by filtration and dried over P,O, in vacuo; yields of the crude product were 94-100%.

The crude compound was dissolved in abs MeOH and treated with activated charcoal. The clarified MeOH soln was dild with H,O to ppt the purified product. However, l e was recrystd from anhyd Et,O instead.

0-(3,3-Dimethyl-l-triazeno)benzamide, la. To a soln of 0-(3,3- dimethyl-1-triazeno)benzoic acid (2 g, 10.35 mmoles in 30 ml of an equivol mixt of anhyd Et,O and p-dioxane) was added Et,N (1.39 ml, 10 mmoles) and ethyl chloroformate (1 ml, 10.3 mmoles, dild with 20 ml of anhyd Et,O). After stirring at room temp for 30 min, the mixt was made ammoniacal with concd NH,OH (2.1 ml, 15 mmoles) while stirring was contd for another 3.5 hr. The solid formed was removed by filtration. A second crop of the crude product was obtd by extg the filtrate with PhH followed by concn. The combined crude product was recrystd from PhH. An analytical sample was further recrystd from PhH-Et,O (4: l by vol, with the addn of a few drops of EtOH).

- SMp's were detd with a Fisher-Jones apparatus. Microanalyses

were performed by Dr. William C. Alford and associates of the Na- tional Institute of Arthritis and Metabolic Diseases, National Insti- tutes of Health, to whom we wish to express our gratitude.

Notes Joumal of Medicinal Chemistry, 1972, Vol. 15, No. 2 203

References (1) Y. F. Shedy, C. A. Krauth, and J. A. Montgomery, J. Org.

(2) Y. F. Shealy, J. Pharm. Sei., 59,1533 (1970). ( 3 ) J. K. Luce, W. C. Thurman, B. L. Issacs, and R. W. Talley,

Cancer Chemother. Rep., 54,119 (1970). (4) J. L. Skibba, G. Ramirez, D. D. Beal, and G . T. Bryan, Biochem.

PharmacoL, 19,2043 (1970). ( 5 ) G. E. Housholder and T. L. Loo, J. Pharmacol. Exp. Ther., in

press. (6) R. Reussmann, A. von Hodenberg, and H. Hengy, Biochem.

Phamacol., 18,l (1969). (7) Y. F. Shealy, J. A. Montgomery, and W. R. Laster, Jr., ibid.,

11,674 (1962). (8) J. H. Burchenal, M. K. Dagg, M. Beyer, and C. C. Stock, Proc.

Sot. Exp. Biol. Med., 91,398 (1956). (9) J. Elks and D. H. Hey, J. Chem. Soc., 441 (1943).

Chem., 27,2150 (1962).

(10) C. W. Noell and C. C. Cheng, J. Med. Chem., 12,545 (1969). (11) Y. F. Shealy and C. A. ODell,J. Pharm. Sci., 60,554 (1971).

Biochemical Studies on Drugs and the Central Nervous System. 1 . Synthesis and Activity of Pyridoxal Derivatives (Studies on the Syntheses of Heterocyclic Compounds. 438l) T. Kametani,* M. Koizumi, Pharmaceutical Institute, Tohoku University, Aobayama, Sendai, Japan

K. Okui, Y. Nishii, and M. Ono Research Laboratories, Chugai Pharmaceutical Co. Ltd. , Tokyo, Japan. Received May 12, 1971

It is well known that vitamin B6 has significant action on the CNS, and that it is of importance in the metabolism of brain cells. A number of studies on the pharmacological action of vitamin B6 and its derivatives have been reported.2 ,3

We had found a method of synthesizing the various 1,2,3, 4tetrahydroisoquinoline derivatives by cyclization of the corresponding carbonyl compounds with 3-hydroxyphenethylamine derivatives 111-VI without acid4-' as a catalyst. We wish to report the synthesis of 1,2,3,4-tetrahydro-l-py- ridoxylisoquinoline derivatives (VIII-XI), 4-pyridoxyl-4,5,- 6,7-tetrahydro-3H-imidazo[4,5-c]pyridine (VII), and their Ac derivatives, by the applkation of this method using pyridoxal with several 3-hydroxyphenethylamines and hista-

CHO

I

R

" f l N H 1 R' \

111, R = H; R' = OH IV, R = OH; R' = OH V, R = OH; R' = H

VI, R = H; R' = OMe

H

VI1

R

R' \

VIII. R = H: R' = OH

< T N C O C H 3 H

IX; R = OH; R' =OH X, R = OH; R' = OH

XI, R = H; R' = OMe

R

XIII, R = H; R' = OAc XIV, R = OAc ; R' = H

xv

Table 1. Products from Pyridoxal and 3-Hydroxyphenethylamines and with Histamine Starting materials Product

Compd Pyridoxal, mg Amine, mg Yield, mg (%) Mp, dec, "C Appearance Formulae

VI I 92 I1 150 160 (82) 252-254b Colorless plates C13H1SN,01f

VI11 420 IIIa-HCl 470 400 (53) 243-24F Colorless needles C,,H,,NzO,f IX 836 IV 845 810 (51) 270-273 Colorless needles C16H18N20S X 1050 V-HC1 1020 630 (33) 243-245 Pale brown needles C,,H,,N,O, .O.sH,Od XI 410 VIa-HCI 500 120 (16) 198-200 Pale yellow needles Cl,HIJVIO,. 0.5H10d

'The free base was prepd as usual. bLit.10 mp 252-253" dec. CLit.ll mp 242-244" dec. dDried over PzOs at 120" (1 mm) for 24 hr. eC, H, N anal. fNot analyzed.

Table 11. The Properties of Acetyl Derivatives of mine, respectively. Their activity on the CNS has also been 1,2,3,4-Tetrahydro-6-hydroxy4-pyridoxylisoquinolines and examined. 4-Pyridoxyl4,S ,6,7-tetrahydro-3H-imidazo [4,5-c]pyridine Chemistry. 3-Hydroxyphenethylamine derivatives (III- No. MP, "C Recrystn solvent Formulau VI), prepared by the usual method, were cyclized with py- XI1 181-183 MeOH-El,Ob c..H,,N.o. ridoxal to give the cyclized compounds listed in Table I. XI11 264-206 MeOH-EtiOC C,H ;:N~O; Although the structure of the above products could be XIV 241-243 MeOHb C&mNlO, thought to be that of Schiff bases, this was ruled out by the

following evidence. Treatment of VII-XI with dil HCl led 'C, H, N anal. bColorless prisms. CColorless needles.


Synthesis of heterocyclic compounds. 438. Biochemical studies on drugs andthe central nervous system. 1. Synthesis and activity of pyridoxal derivatives

T. Kametani, M. Koizumi, K. Okui, Y. Nishii, and M. OnoJ. Med. Chem., 1972, 15 (2), 203-204• DOI: 10.1021/jm00272a021 • Publication Date (Web): 01 May 2002






Notes Joumal of Medicinal Chemistry, 1972, Vol. 15, No. 2 203

References (1) Y. F. Shedy, C. A. Krauth, and J. A. Montgomery, J. Org.

(2) Y. F. Shealy, J. Pharm. Sei., 59,1533 (1970). ( 3 ) J. K. Luce, W. C. Thurman, B. L. Issacs, and R. W. Talley,

Cancer Chemother. Rep., 54,119 (1970). (4) J. L. Skibba, G. Ramirez, D. D. Beal, and G . T. Bryan, Biochem.

PharmacoL, 19,2043 (1970). ( 5 ) G. E. Housholder and T. L. Loo, J. Pharmacol. Exp. Ther., in

press. (6) R. Reussmann, A. von Hodenberg, and H. Hengy, Biochem.

Phamacol., 18,l (1969). (7) Y. F. Shealy, J. A. Montgomery, and W. R. Laster, Jr., ibid.,

11,674 (1962). (8) J. H. Burchenal, M. K. Dagg, M. Beyer, and C. C. Stock, Proc.

Sot. Exp. Biol. Med., 91,398 (1956). (9) J. Elks and D. H. Hey, J. Chem. Soc., 441 (1943).

Chem., 27,2150 (1962).

(10) C. W. Noell and C. C. Cheng, J. Med. Chem., 12,545 (1969). (11) Y. F. Shealy and C. A. ODell,J. Pharm. Sci., 60,554 (1971).

Biochemical Studies on Drugs and the Central Nervous System. 1 . Synthesis and Activity of Pyridoxal Derivatives (Studies on the Syntheses of Heterocyclic Compounds. 438l) T. Kametani,* M. Koizumi, Pharmaceutical Institute, Tohoku University, Aobayama, Sendai, Japan

K. Okui, Y. Nishii, and M. Ono Research Laboratories, Chugai Pharmaceutical Co. Ltd. , Tokyo, Japan. Received May 12, 1971

It is well known that vitamin B6 has significant action on the CNS, and that it is of importance in the metabolism of brain cells. A number of studies on the pharmacological action of vitamin B6 and its derivatives have been reported.2 ,3

We had found a method of synthesizing the various 1,2,3, 4tetrahydroisoquinoline derivatives by cyclization of the corresponding carbonyl compounds with 3-hydroxyphenethylamine derivatives 111-VI without acid4-' as a catalyst. We wish to report the synthesis of 1,2,3,4-tetrahydro-l-py- ridoxylisoquinoline derivatives (VIII-XI), 4-pyridoxyl-4,5,- 6,7-tetrahydro-3H-imidazo[4,5-c]pyridine (VII), and their Ac derivatives, by the applkation of this method using pyridoxal with several 3-hydroxyphenethylamines and hista-

CHO

I

R

" f l N H 1 R' \

111, R = H; R' = OH IV, R = OH; R' = OH V, R = OH; R' = H

VI, R = H; R' = OMe

H

VI1

R

R' \

VIII. R = H: R' = OH

< T N C O C H 3 H

IX; R = OH; R' =OH X, R = OH; R' = OH

XI, R = H; R' = OMe

R

XIII, R = H; R' = OAc XIV, R = OAc ; R' = H

xv

Table 1. Products from Pyridoxal and 3-Hydroxyphenethylamines and with Histamine Starting materials Product

Compd Pyridoxal, mg Amine, mg Yield, mg (%) Mp, dec, "C Appearance Formulae

VI I 92 I1 150 160 (82) 252-254b Colorless plates C13H1SN,01f

VI11 420 IIIa-HCl 470 400 (53) 243-24F Colorless needles C,,H,,NzO,f IX 836 IV 845 810 (51) 270-273 Colorless needles C16H18N20S X 1050 V-HC1 1020 630 (33) 243-245 Pale brown needles C,,H,,N,O, .O.sH,Od XI 410 VIa-HCI 500 120 (16) 198-200 Pale yellow needles Cl,HIJVIO,. 0.5H10d

'The free base was prepd as usual. bLit.10 mp 252-253" dec. CLit.ll mp 242-244" dec. dDried over PzOs at 120" (1 mm) for 24 hr. eC, H, N anal. fNot analyzed.

Table 11. The Properties of Acetyl Derivatives of mine, respectively. Their activity on the CNS has also been 1,2,3,4-Tetrahydro-6-hydroxy4-pyridoxylisoquinolines and examined. 4-Pyridoxyl4,S ,6,7-tetrahydro-3H-imidazo [4,5-c]pyridine Chemistry. 3-Hydroxyphenethylamine derivatives (III- No. MP, "C Recrystn solvent Formulau VI), prepared by the usual method, were cyclized with py- XI1 181-183 MeOH-El,Ob c..H,,N.o. ridoxal to give the cyclized compounds listed in Table I. XI11 264-206 MeOH-EtiOC C,H ;:N~O; Although the structure of the above products could be XIV 241-243 MeOHb C&mNlO, thought to be that of Schiff bases, this was ruled out by the

following evidence. Treatment of VII-XI with dil HCl led 'C, H, N anal. bColorless prisms. CColorless needles.

204 Journal of Medicinal Chemistry, 19 72, Vol. IS, No. 2 Notes

Table 111. Pharmacological Activities of Pyridoxal Derivativesa ~

Barbiturate Traction test potentiating action

Behavioral Relative 30 min 90 min Analgetic action Compds observation activity Judgement TTb FTC TTb FTc 45 min 90” Judgement

Aminophylline 1.63 1.25 ++ 0.82 VI1 -

Pain response Flexor reflex 0 - 0 1.7 1.5 +++

.- - 0.07 - 0 1.0 0.9 VI11 No 0.90 - - 0 0 0.9 1.1 1X No 0.82 - *

X No 0.91 - - 0 - 0 1.6 1.1 ++ XI No 1.46 - 0 0.07 1.6 1.4 XI1 No 1.23 - - 0 -. 0 1.8 1 .o +++

0 0 1.0 0.8 XI11 No 1.13 - -

XIV No 1.36 - 0 - 0 0.9 0.7 -

-

j: - ++ -

2

a[Each number shows the mean value for 5 mice (ddy strain, d , 23-28 g)] . bTT = tranquilizing tendency. CFT = fallen tendency.

to recovery of the starting material (VII-XI); the lack of absorption due to C=N was observed in the ir spectrum; and, although the H, proton of XV was resonant at r 1 there appeared no proton at a lower field than r 2.1 in our product. Moreover, the uv spectrum showed the maximum characteristic of pyridoxal derivatives at 320 nm,’ and of tetrahydroisoquinoline derivatives of 280 nm.4-7 The compounds, VII, VIII, and X, were acetylated in the usual way to afford the Ac derivatives, which showed the absorption band attributable to NHC=O at around 1650 cm-’ in the ir (KBr).8 Compounds VI1 and VIII were identical with the authentic sample prepared by Heyl and his coworkers.”>’’ These facts are consistent with the cyclic structures presented here.

Pharmacology. The compounds so obtained were tested for analgetic effect, traction, and hypnotic action using mice as described later in the Experimental Section. The results are listed in Table 111. Compounds VI1 and XII, which are derivatives of histamine, were found to have slightly more analgetic activity than aminophylline; X and XI have the same analgetic effects as aminophylline, used as a control.

Experimental Section? Cyclization of Pyridoxal ( 1 ) with Amines (Table I). A mixt of

150 mg (0.9 mmole) of pyridoxal, 92 mg (0.9 mmole) of histamine, and 5 ml of EtOH was heated on a water bath for 7 hr. Evap of the solvent gave a pale yellow powder, which was dissolved in 5 ml of 5% NaOH. After filtration, followed by neutralization with 5% HCI, the sepd crystals were collected by filtration. Since the compd was insol in all the solvents, purification was done by repptn.

Acetylation of Pyridoxal Derivatives (Table 11). A mixt of 100 mg (0.5 mmole) of 4-pyridoxyl-4,5,6,7-tetrahydro-3H-imidazo[ 43- cjpyridine (VII), 2 ml of Ac,O, and 1 ml of pyridine was heated on a water bath for 3 hr, and the excess reagents were distd off to leave the Ac derivative XU, recrystn of which from MeOH-Et,O gave 120 mg (61%) of colorless prisms, mp 181-183”.

Pharmacological Tests (Table 111). To male mice (ddy strain, 23-28 g), 100 mg/kg of each compd suspended in 1% arabic gum was administered per os. Barbiturate potentiating action, tractive action, and analgetic action were tested by the methods of Kuhn and coworkers,” Co~rvosier’~ and Burn,’‘ respectively. After administration, general behavior was observed for 90 min and classified into 43 types.

Acknowledgments. The authors are indebted to the Psy- chopharmacology Division of Research Laboratories, Chugai Pharmaceutical Co. Ltd., for the pharmacological data presented.

References (1) T. Kametani, K. Kigasawa, M. Hiragi, and T. Aoyama, J. Med.

(2) P. A. G. Dereymaeker,Med. Pharmacol. Exp., 17, 333 (1967). (3) R. Koch, Strahlentherapie, 113,89 (1960). (4) T. Kametani, K. Fukumoto, H. Yagi, K. Kigasawa, H. Sugahara,

M. Hiiragi, T. Hayasaka, and H. Ishimaru, J. Chem. Soc. C, 112 (1968).

Chem., 14,1235 (1971).

(5) T. Kametani. S. Shibuva. and M. Satoh. Chem. Pharm. Bull.. . I

16,953 (1968). T. Kametani, H. Yagi, and K. Fukumoto, ibid., 16, 1285 (1968). T. Kametani, S. Takano, and S. Hibino, Yukugaku Zasshi, 88, 1123 (1968). W. Kuryinyk, H. Ahrens, and N. Angelino, Tetrahedron, 26, 5445 (1970). S. F. Mason, J. Chem. Soc., 1253 (19.59). D. Heyl, E. Luz, S. A. Harris, and K. Folkers, J. Amer. Chem. Soc., 70, 3669 (1948). D. Heyl, E. Luz, S. A. Harris, and K. Folkers, ibid., 74, 414 (1 95 2). W. L. Kuhn and E. F. Van Marren, J. Pharmacol. Exp. Ther., 134,60 (1961). S . Courvosier, “Psychotropic Drugs,” Garanttini and Ghetti, Ed., Elsevier Publishing Co., Amsterdam, p 373. J. H. Burn, “Biological Standardization,” Oxford University Press, London, 1950.

Antimalarial Compounds.?. 12. Guanidine Derivatives of Diphenyl Sulfone and Related Compounds W. Peters, Liverpool School of Tropical Medicine, Liverpool L3 SQA, England

H. Piotrowska, B. Serafin, and T. Urba6ski Institute of Organic Chemistry and Technology, Technical University (Politechnika), Warsaw 10, Poland. Received August 31, I971

Continuing our research on potential antimalarial agents,2 a series of biguanide and amidineurea derivatives of phalo- diphenyl sulfones, p-nit rodiphenyl ether, pnitrodiphenyl- methane, and pnitrodiphenylamine was prepared. In the present work we tried to investigate to what extent the nitro and sulfo group were responsible for the antimalarial activity of pnitrodiphenyl sulfone derivatives.2

Chemistry. The starting amines I-VI were prepared according to the l i t e r a t ~ r e ~ ’ ~ and subsequently caused to react

tMelting points were determined on a Yanagimoto microappara. tus (MP-S2) and uncorrected.

?The financial support of this work from the World Health Organization is gratefully acknowledged.


Antimalarial compounds. 12. Guanidine derivativesof diphenyl sulfone and related compounds

W. Peters, H. Piotrowska, B. Serafin, and T. UrbanskiJ. Med. Chem., 1972, 15 (2), 204-206• DOI: 10.1021/jm00272a022 • Publication Date (Web): 01 May 2002







Table 111. Pharmacological Activities of Pyridoxal Derivativesa ~

Barbiturate Traction test potentiating action

Behavioral Relative 30 min 90 min Analgetic action Compds observation activity Judgement TTb FTC TTb FTc 45 min 90” Judgement

Aminophylline 1.63 1.25 ++ 0.82 VI1 -

Pain response Flexor reflex 0 - 0 1.7 1.5 +++

.- - 0.07 - 0 1.0 0.9 VI11 No 0.90 - - 0 0 0.9 1.1 1X No 0.82 - *

X No 0.91 - - 0 - 0 1.6 1.1 ++ XI No 1.46 - 0 0.07 1.6 1.4 XI1 No 1.23 - - 0 -. 0 1.8 1 .o +++

0 0 1.0 0.8 XI11 No 1.13 - -

XIV No 1.36 - 0 - 0 0.9 0.7 -

-

j: - ++ -

2

a[Each number shows the mean value for 5 mice (ddy strain, d , 23-28 g)] . bTT = tranquilizing tendency. CFT = fallen tendency.

to recovery of the starting material (VII-XI); the lack of absorption due to C=N was observed in the ir spectrum; and, although the H, proton of XV was resonant at r 1 there appeared no proton at a lower field than r 2.1 in our product. Moreover, the uv spectrum showed the maximum characteristic of pyridoxal derivatives at 320 nm,’ and of tetrahydroisoquinoline derivatives of 280 nm.4-7 The compounds, VII, VIII, and X, were acetylated in the usual way to afford the Ac derivatives, which showed the absorption band attributable to NHC=O at around 1650 cm-’ in the ir (KBr).8 Compounds VI1 and VIII were identical with the authentic sample prepared by Heyl and his coworkers.”>’’ These facts are consistent with the cyclic structures presented here.

Pharmacology. The compounds so obtained were tested for analgetic effect, traction, and hypnotic action using mice as described later in the Experimental Section. The results are listed in Table 111. Compounds VI1 and XII, which are derivatives of histamine, were found to have slightly more analgetic activity than aminophylline; X and XI have the same analgetic effects as aminophylline, used as a control.

Experimental Section? Cyclization of Pyridoxal ( 1 ) with Amines (Table I). A mixt of

150 mg (0.9 mmole) of pyridoxal, 92 mg (0.9 mmole) of histamine, and 5 ml of EtOH was heated on a water bath for 7 hr. Evap of the solvent gave a pale yellow powder, which was dissolved in 5 ml of 5% NaOH. After filtration, followed by neutralization with 5% HCI, the sepd crystals were collected by filtration. Since the compd was insol in all the solvents, purification was done by repptn.

Acetylation of Pyridoxal Derivatives (Table 11). A mixt of 100 mg (0.5 mmole) of 4-pyridoxyl-4,5,6,7-tetrahydro-3H-imidazo[ 43- cjpyridine (VII), 2 ml of Ac,O, and 1 ml of pyridine was heated on a water bath for 3 hr, and the excess reagents were distd off to leave the Ac derivative XU, recrystn of which from MeOH-Et,O gave 120 mg (61%) of colorless prisms, mp 181-183”.

Pharmacological Tests (Table 111). To male mice (ddy strain, 23-28 g), 100 mg/kg of each compd suspended in 1% arabic gum was administered per os. Barbiturate potentiating action, tractive action, and analgetic action were tested by the methods of Kuhn and coworkers,” Co~rvosier’~ and Burn,’‘ respectively. After administration, general behavior was observed for 90 min and classified into 43 types.

Acknowledgments. The authors are indebted to the Psy- chopharmacology Division of Research Laboratories, Chugai Pharmaceutical Co. Ltd., for the pharmacological data presented.

References (1) T. Kametani, K. Kigasawa, M. Hiragi, and T. Aoyama, J. Med.

(2) P. A. G. Dereymaeker,Med. Pharmacol. Exp., 17, 333 (1967). (3) R. Koch, Strahlentherapie, 113,89 (1960). (4) T. Kametani, K. Fukumoto, H. Yagi, K. Kigasawa, H. Sugahara,

M. Hiiragi, T. Hayasaka, and H. Ishimaru, J. Chem. Soc. C, 112 (1968).

Chem., 14,1235 (1971).

(5) T. Kametani. S. Shibuva. and M. Satoh. Chem. Pharm. Bull.. . I

16,953 (1968). T. Kametani, H. Yagi, and K. Fukumoto, ibid., 16, 1285 (1968). T. Kametani, S. Takano, and S. Hibino, Yukugaku Zasshi, 88, 1123 (1968). W. Kuryinyk, H. Ahrens, and N. Angelino, Tetrahedron, 26, 5445 (1970). S. F. Mason, J. Chem. Soc., 1253 (19.59). D. Heyl, E. Luz, S. A. Harris, and K. Folkers, J. Amer. Chem. Soc., 70, 3669 (1948). D. Heyl, E. Luz, S. A. Harris, and K. Folkers, ibid., 74, 414 (1 95 2). W. L. Kuhn and E. F. Van Marren, J. Pharmacol. Exp. Ther., 134,60 (1961). S . Courvosier, “Psychotropic Drugs,” Garanttini and Ghetti, Ed., Elsevier Publishing Co., Amsterdam, p 373. J. H. Burn, “Biological Standardization,” Oxford University Press, London, 1950.

Antimalarial Compounds.?. 12. Guanidine Derivatives of Diphenyl Sulfone and Related Compounds W. Peters, Liverpool School of Tropical Medicine, Liverpool L3 SQA, England

H. Piotrowska, B. Serafin, and T. Urba6ski Institute of Organic Chemistry and Technology, Technical University (Politechnika), Warsaw 10, Poland. Received August 31, I971

Continuing our research on potential antimalarial agents,2 a series of biguanide and amidineurea derivatives of phalo- diphenyl sulfones, p-nit rodiphenyl ether, pnitrodiphenyl- methane, and pnitrodiphenylamine was prepared. In the present work we tried to investigate to what extent the nitro and sulfo group were responsible for the antimalarial activity of pnitrodiphenyl sulfone derivatives.2

Chemistry. The starting amines I-VI were prepared according to the l i t e r a t ~ r e ~ ’ ~ and subsequently caused to react

tMelting points were determined on a Yanagimoto microappara. tus (MP-S2) and uncorrected.

?The financial support of this work from the World Health Organization is gratefully acknowledged.


Table I. Toxicity, mg/kg Antimalarial activity (parasitaemia

(mice)

LD,, LDo/da No. PO sc PO

VI1 1166 100 200 VI11 2500 100 200 IX 100 X < 100 XI > 300 XI1 450 100 17 XI11 660 100 17 XIV 830 100 75 XV > 300 XVI > 300 XVII > 100

relative to controls) mg/kg per day sc

1 3 10 30 100 300 93.7 89.8 41.4 100 55.9 55.9

73.1 52.8 Inactive Inactive

100 71.6 18.7 0 100 85.1 77.8 58.7 9.0

83.3 83.5 51.5 h a c tive Inactive

Inactive 'The highest dose, administered on 4 consecutive days, that pro-

duced no deaths.

Table 11.

VII-IX and XII-XIV showed a similar level of subacute toxicity sc but the latter were more toxic orally, probably due to better oral absorption than VII-IX. XI and XV-XVII were significantly less toxic sc while X was the most toxic.

Antimalarial Activity.$ The antimalarial activity of VII- XVII was tested against Plasmodium berghei in mice sc (Table I). Compounds VII-IX and XII-XIV showed some antimalarial action, XI1 being the most active. In no case was complete activity obtained at the LD0p or below. No activity was detected at the maximum doses tested in X, XI or XV-XVII.

Experimental Section All analytical data of the new compounds were in agreement

with the calcd ones for the expected structures which were also confirmed by ir spectra.

Yield Cyano- 35% alc EtOH, Reaction Amine % Recrystn solvent No. Wt, g guanidine, g HC1, ml ml time, min BG Mp,"C g

I1 7.5 5 .O 10.0 30.0 60 VI1 178-179 3.8 40 40% EtOH 111 6.0 6.0 7 .O 25 .O 60 VI11 175-176 3.5 35 40% EtOH IV 7.5 5.0 5 .O 50.0 60 Ix 180-181 7.0 65 50% EtOH v 7.5 5 .O 7.5 20.0 120 X 174-175 6.1 65 MeOH + EtOH

Table 111.

10% HC1, Biguanide No. Wt, g ml VI1 2.0 10 VI11 2.0 10 IX 1 .o 20 X 1 .o 20 XI 1.5 20

Yield Reaction time, min Amidineurea MP, "C g % Recrystn solvent

15 XI11 190-191 1.1 52 60% EtOH 15 XIV 187-188 1.2 60 60% EtOH

80% EtOH 90 XVI 206-208 0.2 15 80% EtOH 120 xv 215-217 0.2 20

15 0 X V I I >260 0.7 37 DMF

with cyanoguanidine in alcoholic HC1. p-Nitro-p'-biguanido- diphenylamine (XI) was prepared from VI and cyanoguanidine in the presence of pyridine hydrochloride. Amidine- ureas XII-XVII were prepared by hydrolysis of the corresponding biguanides. For details see Experimental Section.

Scheme 1

BG = -NHCNHCNH, AU = -NHCONHCNH, II II II

NH NH NH

Amine BG AU X Y I XI1 SO, F

c1 Br

I1 VI1 XI11 SO2 I11 VI11 XIV SO2 IV IX xv 0 NO

NO; 2 NO. V X XVI VI XI XVII

Toxicity.* Acute toxicity of VII, VIII, and XII-XIV was tested by oral administration and subacute toxicity of VII- XVII (4 daily doses) by oral and sc administration (Table I).

$Tests were carried out partly at the Institute of Drugs, Warsaw (acute and subacute oral tests) and the Liverpool School of Tropical Medicine (subacute sc tests).

___

Biguanides VII-X were prepared as follows. The amine was dissolved in alc HCl, cyanoguanidine was added, and the mixt was refluxed. The resulting hydrochloride was made alk with 5% NaOH, dried, and boiled with PhMe to remove the unreacted starting amine (Table 11).

The biguanide XI was prepared by refluxing 2.3 g of VI, 1.2 g of pyridine hydrochloride, and 1.0 g of cyanoguanidine in 10 ml of pyridine for 4 hr. The soln was poured into H,O and made alk with 5% NaOH. The crude product upon washing with acetone crystd from 50% EtOH: mp 201-202"; yield, 1.1 g (48%).

respectively, on heating in dil HCl (Table 111).

white mice (Porton breed) in groups of 20. The compds were administered by gavage in a 5% suspension of aq gum arabic in a vol of 0.7-0.8 m1/20 g of body wt. The LD,, was calcd graphically by Litchfield and Wilcoxon's method as modified by Roth. The animals were observed for 7 days. Subacute toxicity was assessed by de$ the highest dose administered on 4 consecutive days orally (as above) or sc (in a 4% soln of Tween 80), that produced no deaths in groups of 10 or 15 mice. Doses (sc) were contained in a vol of 0.2 m1/20 g of body wt (Table I).

Antimalarial Activity.$ All tests were carried out in white mice (CFl line) infected ip on day 0 with donor blood contg approximately lo' parasitised red blood cells. Animals received 4 consecutive daily doses of drug in 0.2 ml of soln or suspension sc from day 0 through day +3. The percentage of parasitised red blood cells was counted in treated groups of animals on day +4 and compared with the percentage in saline-treated controls.

Acknowledgments. We wish to express our gratitude to Dr. T. Lepes and Dr. J. Haworth of the World Health Organ- ization for their kind interest and aid. The technical assistance of Mrs. K. Grzelecka in preparation of the samples

Amidineureas XIII-XVII were obtained from biguanides VII-XI,

Toxicity.$ Acute toxicity on oral administration was tested in

$Tests were carried out by the method of Peters,' at the Liverpool School of Tropical Medicine.

206 Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 hrotes

for biological screening and of Mrs. J. Portus and Mr. B. Robinson for antimalarial screening is also gratefully acknowledged.

References (1) B. Serafin and M. H. Aldridge, J. Med. Chem., submitted for pub-

(2) B. Serafin, T. Urbahski, D. C. Warhurst, ibid., 12,33 (1969). (3) C. Richter and W. Frey, Swiss Patent No. 278939 (1952). (4) L. C. Raiford and J. C. Colbert, J. Amer. Chem. Soc., 48,2660

( 5 ) L. H. Litvinenko and K. Levchenko, Zh. Obshch. Khim., 29,

(6) F. Ullman and K. Dahmen,Ber., 41, 3753 (1908). (7) W. Peters, Exp. Parasitol., 17, 80 (1965).

lication (part 11).

(1926).

1970, 3079 (1959).

Synthesis and Enzymological Activity of 3-Hydroxy-2-n-propy1-4,S-pyridinedimethanol Paul Melius* and Joseph L. Greene, Jr. Department of Chemistry, Auburn University, Auburn, Alabama. Received July 15, 1971

The conversion of 3-hydroxy-2-ethy1-4,5-pyridinedimeth- an01 and 3-hydroxy-2-isopropyl-4,5-pyridinedimethanol to their corresponding aldehydes by yeast pyridoxine dehydrogenase was described by Melius and Marshall.' Also 3-hydroxy-2-methyl-6-chloro-4,5-pyridinedime than01 was found to be oxidized by the enzyme with an activity of the order of that for the i-Pr analog. The rate of reaction for the Me and Et analogs was about 4 times that for the i-Pr analog and the hydroxychloro compound. In the present report the synthesis and yeast pyridoxine dehydrogenase action on 3-hydroxy-2-n-propyl-4,5-pyridinedimethanol (VIII) is described.

The synthesis of VI11 involved an initial condensation of ethyl-n-butyropyruvate and cyanoacetamide to form 4- carbethoxy-3-cyano-6-n-propyl-2-pyridone (I).2 I was then carried through a sequence of reactions involving n i t r a t i ~ n , ~ chlorination, reduction with reduction with Pd-H2,' hydrolysis with HCl,3 diazotization, and reduction with NaBh,6 to give finally VIII. Thus a modification of the reduction of the NOz group was utilized here, in which SnC12 was used in place of Fe which had been utilized in the preparation of the i-Pr analog.

The pyridoxine dehydrogenase enzyme used here was a preparation described by Morino and Sakamoto.2 The assay of enzymatic activity toward VI11 is given in Table I and compared with the activities of the compds prepared and studied by Melius and Marshall.'


fluxing a soln of the Na salt of ethyl n-butyr~pyruvate~ (208 g; 1.0 mole) and cyanoacetamide (92 g; 1.1 moles) in abs EtOH (1400 ml) for 3 hr. After standing at room temp overnight, the reaction mixt was chilled and treated with an ice cold soln made up by dilg concd HCl(200 ml) to 1200 ml with ice and H,O. The crude

4Carbethoxy-3-cyano-6-n-propyl-2-pyridone (I) was prepd re-

?Melting points are corrected and were determined in a Mel-Temp apparatus (Laboratory Devices, Cambridge, Mass.) Microanalyses were by Galbraith Laboratories, Knoxville, Tenn.

Table I. Enzyme Activity of Analogs

PDH Compound R, R, R, activity, %

1 Pyridoxine Me OH H 100 2 w-Methylpyridoxine' Et OH € I 95 3 &Rando$ i- PI OH 11 25 4 C1 anal08 Me OH e1 22 5 n - R analog VI11 n-Pr OH H 5

'Data for these PDH activity estimates were obtd from Melius and Marshall. *

product thus pptd was washed thoroughly with H,O before being crystd from aq EtOH (3000 ml; (60:40) EtOH-H,O) to give 152 g (65%) of I, mp 146-148'.

4Carbethoxy-3-cyano-6-n-propylJ-nitro-2-pyridone (11). Compd I (23.5 g; 0.1 mole) was nitrated with HN0,-Ac,O, essentially as described by W ~ e s t , ~ to give, after recrystn from 50% aq EtOH, 18.5 g (66.3%) of 11, mp 163-164".

4Carbethoxy-2-chloro-S-nitro-6-n-propylnicotinonitrile (111). Compd I1 (27.9 g; 0.1 mole) and PCl, (22.9 g; 0.11 mole) were mixed and heated together at 125 +5" for 2 hr. POCl, was removed in vacuo before the residue was triturated with crushed ice until solidification was complete. Recrystn of the crude product from abs EtOH gave 21.2 g (71%) of 111, mp 48-50'.

Compd 111 (29.8 g; 0.1 mole), suspended in Et,O was treated with a freshly filtered s o h of SnCl, (78 g) in concd HCl(l65 ml) in a manner analagous to that described by Greene and Montgomery.4 The crude product was recrystd from abs EtOH (625 ml) to give 23 g (86%) of IV, mp 168-169".

genation of IV (26.8 g; 0.1 mole) over 5% Pd-BaCOS3 and work-up of the reaction mixt gave 10.5 g (45%) of V after recrystn from abs EtOH, mp 132-133'.

3-Amin0-2-n-propylpyridine-4,S-dicarboxylic Acid (VI). Hydrolysis of V (15.6 g; 0.067 mole) with concd HCl' gave 9.0 g (60%) of the dicarboxylic acid VI, mp 215-216' dec.

3-Hydroxy-2-n-propylpyridine-4,5dicarboxytic Acid (VII). Compd VI (8.3 g; 0.037 mole) was diazotized at 80" in aq s o h to give 4.1 g (49.4%) of VIII, mp 230-232".

3-Hydroxy-2-n-propyl-4,5-pyridinedimethanol Hydrochloride (VIII). Reduction of the 2 CO,H groups of VI1 (2.25 g; 0.01 mole) with NaBH,-AlCl, as described by Blackwood6 gave the n-Pr analog of pyridoxine hydrochloride (0.94 g; 41%); mp 210- 212' dec.Anal. (C,,H,,NO;HCl) C, H, N.

shall. l The yeast pyridoxine dehydrogenase (PDH) activity was measured by the reaction of phenylhydrazine with the aldehyde produced from the pyridoxine and its analogs.'

Acknowledgment. We wish to thank Miss Barbara K. Newton for her technical assistance.

5-Amino-2-chloro-4-carbethoxy-6-n-pmpylnicotinonitr~e (IV).

5-Ammo-4-carbethoxy-6-n-propylnicotinonitrile (V). Hydro-

Enzymatic assays were carriedout as described by Melius and Mar-

References (1) P. Melius and D. L. Marshall, J. Med. Chem., 10, 1157 (1967). (2) Y. Morino and Y. Sakamoto, J. Biochem. (Tokyo), 48,733

(1960). (3) H. M. Wuest, J. A. Bigot, Th. J. DeBoer, B. van der Wal, and

J. P. Wibaut, Reel. Trav. Chim. Pays-Bas, 78,226 (1958). (4) J. L. Greene, Jr., and J. A. Montgomery, J. Med. Chem., 6, 294

(1963). (5) C. S. Marvel and E. E. Dreger in "Organic Synthesis," Collect.

Vol. I , H. Gilman, Ed., Wiley, New York, N. Y., 1946, p 238. (6) R. K. Blackwood, G. B. Hess, C. E. Larrabee, and F. J. Pilgram,

J. Amer. Chem. Soc., 80,6244 (1958). (7) H. Wada and E. E. Snell, J. Biol. Chem., 236, 2089 (1961).


Synthesis and enzymological activity of3-hydroxy-2-propyl-4,5-pyridinedimethanol

Paul Melius, and Joseph L. Greene Jr.J. Med. Chem., 1972, 15 (2), 206-206• DOI: 10.1021/jm00272a023 • Publication Date (Web): 01 May 2002






206 Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 hrotes

for biological screening and of Mrs. J. Portus and Mr. B. Robinson for antimalarial screening is also gratefully acknowledged.

References (1) B. Serafin and M. H. Aldridge, J. Med. Chem., submitted for pub-

(2) B. Serafin, T. Urbahski, D. C. Warhurst, ibid., 12,33 (1969). (3) C. Richter and W. Frey, Swiss Patent No. 278939 (1952). (4) L. C. Raiford and J. C. Colbert, J. Amer. Chem. Soc., 48,2660

( 5 ) L. H. Litvinenko and K. Levchenko, Zh. Obshch. Khim., 29,

(6) F. Ullman and K. Dahmen,Ber., 41, 3753 (1908). (7) W. Peters, Exp. Parasitol., 17, 80 (1965).

lication (part 11).

(1926).

1970, 3079 (1959).

Synthesis and Enzymological Activity of 3-Hydroxy-2-n-propy1-4,S-pyridinedimethanol Paul Melius* and Joseph L. Greene, Jr. Department of Chemistry, Auburn University, Auburn, Alabama. Received July 15, 1971

The conversion of 3-hydroxy-2-ethy1-4,5-pyridinedimeth- an01 and 3-hydroxy-2-isopropyl-4,5-pyridinedimethanol to their corresponding aldehydes by yeast pyridoxine dehydrogenase was described by Melius and Marshall.' Also 3-hydroxy-2-methyl-6-chloro-4,5-pyridinedime than01 was found to be oxidized by the enzyme with an activity of the order of that for the i-Pr analog. The rate of reaction for the Me and Et analogs was about 4 times that for the i-Pr analog and the hydroxychloro compound. In the present report the synthesis and yeast pyridoxine dehydrogenase action on 3-hydroxy-2-n-propyl-4,5-pyridinedimethanol (VIII) is described.

The synthesis of VI11 involved an initial condensation of ethyl-n-butyropyruvate and cyanoacetamide to form 4- carbethoxy-3-cyano-6-n-propyl-2-pyridone (I).2 I was then carried through a sequence of reactions involving n i t r a t i ~ n , ~ chlorination, reduction with reduction with Pd-H2,' hydrolysis with HCl,3 diazotization, and reduction with NaBh,6 to give finally VIII. Thus a modification of the reduction of the NOz group was utilized here, in which SnC12 was used in place of Fe which had been utilized in the preparation of the i-Pr analog.

The pyridoxine dehydrogenase enzyme used here was a preparation described by Morino and Sakamoto.2 The assay of enzymatic activity toward VI11 is given in Table I and compared with the activities of the compds prepared and studied by Melius and Marshall.'


fluxing a soln of the Na salt of ethyl n-butyr~pyruvate~ (208 g; 1.0 mole) and cyanoacetamide (92 g; 1.1 moles) in abs EtOH (1400 ml) for 3 hr. After standing at room temp overnight, the reaction mixt was chilled and treated with an ice cold soln made up by dilg concd HCl(200 ml) to 1200 ml with ice and H,O. The crude

4Carbethoxy-3-cyano-6-n-propyl-2-pyridone (I) was prepd re-

?Melting points are corrected and were determined in a Mel-Temp apparatus (Laboratory Devices, Cambridge, Mass.) Microanalyses were by Galbraith Laboratories, Knoxville, Tenn.

Table I. Enzyme Activity of Analogs

PDH Compound R, R, R, activity, %

1 Pyridoxine Me OH H 100 2 w-Methylpyridoxine' Et OH € I 95 3 &Rando$ i- PI OH 11 25 4 C1 anal08 Me OH e1 22 5 n - R analog VI11 n-Pr OH H 5

'Data for these PDH activity estimates were obtd from Melius and Marshall. *

product thus pptd was washed thoroughly with H,O before being crystd from aq EtOH (3000 ml; (60:40) EtOH-H,O) to give 152 g (65%) of I, mp 146-148'.

4Carbethoxy-3-cyano-6-n-propylJ-nitro-2-pyridone (11). Compd I (23.5 g; 0.1 mole) was nitrated with HN0,-Ac,O, essentially as described by W ~ e s t , ~ to give, after recrystn from 50% aq EtOH, 18.5 g (66.3%) of 11, mp 163-164".

4Carbethoxy-2-chloro-S-nitro-6-n-propylnicotinonitrile (111). Compd I1 (27.9 g; 0.1 mole) and PCl, (22.9 g; 0.11 mole) were mixed and heated together at 125 +5" for 2 hr. POCl, was removed in vacuo before the residue was triturated with crushed ice until solidification was complete. Recrystn of the crude product from abs EtOH gave 21.2 g (71%) of 111, mp 48-50'.

Compd 111 (29.8 g; 0.1 mole), suspended in Et,O was treated with a freshly filtered s o h of SnCl, (78 g) in concd HCl(l65 ml) in a manner analagous to that described by Greene and Montgomery.4 The crude product was recrystd from abs EtOH (625 ml) to give 23 g (86%) of IV, mp 168-169".

genation of IV (26.8 g; 0.1 mole) over 5% Pd-BaCOS3 and work-up of the reaction mixt gave 10.5 g (45%) of V after recrystn from abs EtOH, mp 132-133'.

3-Amin0-2-n-propylpyridine-4,S-dicarboxylic Acid (VI). Hydrolysis of V (15.6 g; 0.067 mole) with concd HCl' gave 9.0 g (60%) of the dicarboxylic acid VI, mp 215-216' dec.

3-Hydroxy-2-n-propylpyridine-4,5dicarboxytic Acid (VII). Compd VI (8.3 g; 0.037 mole) was diazotized at 80" in aq s o h to give 4.1 g (49.4%) of VIII, mp 230-232".

3-Hydroxy-2-n-propyl-4,5-pyridinedimethanol Hydrochloride (VIII). Reduction of the 2 CO,H groups of VI1 (2.25 g; 0.01 mole) with NaBH,-AlCl, as described by Blackwood6 gave the n-Pr analog of pyridoxine hydrochloride (0.94 g; 41%); mp 210- 212' dec.Anal. (C,,H,,NO;HCl) C, H, N.

shall. l The yeast pyridoxine dehydrogenase (PDH) activity was measured by the reaction of phenylhydrazine with the aldehyde produced from the pyridoxine and its analogs.'

Acknowledgment. We wish to thank Miss Barbara K. Newton for her technical assistance.

5-Amino-2-chloro-4-carbethoxy-6-n-pmpylnicotinonitr~e (IV).

5-Ammo-4-carbethoxy-6-n-propylnicotinonitrile (V). Hydro-

Enzymatic assays were carriedout as described by Melius and Mar-

References (1) P. Melius and D. L. Marshall, J. Med. Chem., 10, 1157 (1967). (2) Y. Morino and Y. Sakamoto, J. Biochem. (Tokyo), 48,733

(1960). (3) H. M. Wuest, J. A. Bigot, Th. J. DeBoer, B. van der Wal, and

J. P. Wibaut, Reel. Trav. Chim. Pays-Bas, 78,226 (1958). (4) J. L. Greene, Jr., and J. A. Montgomery, J. Med. Chem., 6, 294

(1963). (5) C. S. Marvel and E. E. Dreger in "Organic Synthesis," Collect.

Vol. I , H. Gilman, Ed., Wiley, New York, N. Y., 1946, p 238. (6) R. K. Blackwood, G. B. Hess, C. E. Larrabee, and F. J. Pilgram,

J. Amer. Chem. Soc., 80,6244 (1958). (7) H. Wada and E. E. Snell, J. Biol. Chem., 236, 2089 (1961).


Antimalarial activity and conformation of erythro- and threo-.alpha.-(2-piperidyl)-3,6-bis(trifluoromethyl)-9-phenanthrenemethanol

R. E. OlsenJ. Med. Chem., 1972, 15 (2), 207-208• DOI: 10.1021/jm00272a024 • Publication Date (Web): 01 May 2002






Notes Journal ofhfedicinal Chemistry, 1972, Vol. 15, No. 2 207

Antimalarial Activity and Conformation of ery thro- and threo-cw-( 2-Piperidyl)-3,6-bis( trifluoromethy1)-9- phenanthrenemethanolt R. E. Olsen Aerojet Solid Propulsion Company, Sacramento, California 95813. Received June 10, 1971

A convenient route to antimalarials containing a 2-piperidinemethanol group involves, as the final reaction step, a catalytic hydrogenation of a 2-pyridyl ketone,' thus producing a racemic mixture of erythro and threo amino alcohols. Previous piperidinemethanol preparations were accomplished either by a different synthetic scheme? usually yielding only one isolated diasteromeric form,$ or on a scale insufficient to permit isolation and characteriza- tion of the isomer formed in lesser amounts. There thus appears to be limited data available as to the effect of conformation on the antimalarial activity of piperidinemethanols.

2-Pyridyl 3,6-bis(trifluoromethyl)-9-phenanthryl ketone was prepared by the procedure of Nodiff, et al.,' and hydrogenated at atm pressure to give about an 85: 15 mixture (analysis via nmr and tlc) of epimeric piperidinemethanols. The hydrochloride of the major epimer (I) was sparingly MeOH soluble and was obtained in pure form after 2 evaporative recrystallizations; concn of the mother liquors gave a residue containing roughly equal amounts of the diastereomers. Upon conversion to the TsOH salt, the minor epimer (11) could be isolated in pure form after 2 recrystallizations from MeOH. The amino alcohols were treated with methanolic CH20 and converted to the corresponding 1 - [3,6-bis(trifluorome thyl)-9-phenanthryl] hexahydro-3H- oxazolo[3,4a]pyridire derivative (I + 111, I1 + IV).

I I1

Ha Hb

111 IV

AI = 3,6-Bis(trifluoromethy1)-9-phenanthyl

The difference in chemical shift between the two C3 CH2 of 1 -substituted-hexahydro-3H-oxazolo [ 3,4u] pyridines and the geminal coupling constant for the same 2 protons can be employed to assign configuration to the amino alcohol precursors.6 Thus, by examination of the nmr spectra of 111 and IV, we have assigned the configuration of I and I1 as erythro and threo, respectively. The data for 111 and IV are shown in Table I; for comparative purposes, data6 for the racemic epimers of 1 -phenylhexahydro-3H-oxazolo- [3,4a] pyridine are included.

An nmr study of I and I1 and their protonated forms was conducted since it is possible their pharmacological activities are related to conformational preferences. The erythro (I) and threo (11) amino alcohols can lie in 3 possible staggered

--__- - ?This work was supported by the U. S. Army Medical Research

and Development Command under Contract DA-49-193-MD-2891. This is Contribution No. 947 from the Army Research Program on Malaria.

$H. Sargent3 reported the isolation of both piperidinemethanol epimers from the catalytic reduction of 2-piperidyl 6-methoxy-4- quinolyl ketone. The materials were tested as SN2157 and SN8279 against P. lophurae in ducks and found to be of comparable a ~ t i v i t y . ~

Table I. Nmr Data for l-Aryl-hexahydro-3H-oxazolo[ 3,4-a]pyridinesU

Ar Of HaHb c3 protonsb *H,H,, JH,GemC Configuration Chemical shift

Phenyls Cisd 3.78 4.70 0.92 -1.3 e (111) 4.17 5.09 0.92 -1.5

e (IV) 4.52 4.95 0.43 --2.6 Phenyls Transd 4.21 4.64 0.43 -2.7

'Spectra obtd on a Varian Model DP-60 hi h resolution spectrometer in CDCl, at room temp at 10% concn. gI i 6 values, relative to TMS. CAn average of 6 runs. dConfign of phenyl 2-piperidylmethanols has also been established by Dudas and Weisz' and by Kovar, et al. e3,6-Bis(trifluoromethyl)-9-phenanthryl.

Table 11. JAB Values for Epimeric (~-(2-~per~dyl)-3,6-bis(trifluoro~ethyl)-9-phenanthrenemethanol and Their HCl Saltsn

JAR, Hzb Solvent Temn OC I I1 I.HC1 II.HCI

DMSSd, 100 4.66c 7.31 2.65c 7.55 TFA 60 2.69 6.98 Ace tone -d 50 4.09 6.91 CDCl, 50 2.70e 6.58 C6D6 50 2.68e 6.41

'Spectra were obtd at indicated temp on a Varian Model DP-60 high-resolution spectrometer at a concn of 10%. bAn average of 6-10 runs; values accurate to an estimated k0.2 Hz. COH resonance eliminated by addn of D,O. dF,CCO,H. eExchangeable protons replaced with D by shaking overnight with D,O.

conformations (A, B, and C) with the vicinal coupling constant, JAB, giving an estimate of the rotamer populations. In similar compounds, the pure gauche and trans coupling constant values have been suggested as 2.8 and 10S9 and 2.6 and 10.3 Hz." TheJAB values for I and I1 and their HCl salts in solvents of differing polarity are shown in Table 11.

These data suggest that in nonpolar solvents the erythro epimer resides almost entirely in the gauche forms IA and IB, with IA favored over IB since on inspection of molecular models the latter shows severe nonbonded interactions. Based on these steric interactions alone, IC should be the preferred rotamer. But in view of JAB values, intramolecular H bonding must significantly influence the conformational preferences. The increase in JAB of 1.39 and 1.96 h z (compared with the CDCIB value), respectively, in Me&O and DMSO are consistent with the importance assigned to intramolecular H bonding as in these solvents, one would expect an increase in intermolecular H bonding" [OH---O=C(CH,), and OH---O=S(CH&] with a decrease in intramolecular H bonding (OH-N).

In the threo amino alcohol, in which nonbonded interactions enhance intramolecular H bonding, the JAB values did not show as marked a solvent dependence. It is expected the population of IIC would be minimal, since it is incapable of intramolecular H bonding and has a high degree of nonbonded interaction. The nmr data indicates that I1 exists as a mixture of IIA and IIB, with IIA favored as it has the lesser steric interactions. The increase in JAB observed in MezCO and DMSO (compared to CDC13) is consistent with the increased importance of steric factors in these solvents.

The coupling constants for the HC1 salts of I and I1 (Table 11) indicate the preferred protonated rotamers are IA and a mixture of IIA and IIB. It is of interest to compare the JAB


?H

IA IB IC

IIA IIB IIC

Ar = 3,6-Bis(trifluoromethyl)-9-phenanthryl

coupling constants observed in DMSO for the free bases and their protonated forms. Evidently, on protonation, the bulk of the solvated quaternary N in I . HCl is significantly increased by H bonding with DMSO. This would enhance the population of IA at the expense of IB and IC. Similarly, protonation of I1 should increase the population of IIA with a concomitant decrease in IIB, due to the increased steric requirements of the solvated ammonium group.

a-(2-piperidyl)-3,6-bis(trifluoromethyl)-9-phenanthrenemethanol were tested12 for antimalarial activity as their HCl and TsOH salts and as their oxazolopyridine derivatives against Plasmodium berghei in mice and P. gallinaceum in chicks by Dr. Leo Rane at the University of Miami. The test results, furnished to us through the Walter Reed Army Institute of Research, show these materials to be highly active against P. berghei, giving 5 cures out of 5 infected mice at dosages of 40 mg/kg for the threo epimers and 5 cures at 80 mg/kg for the erythro compds.§ No toxic deaths were reported up to dosages of 640 mg/kg. Against P. gal- Zinaceum, the minimum dosage showing activity was 20 and 160 mg/kg for the threo and erythro compds, respectively.

To attempt a general correlation of epimer conformation with antimalarial activity on the above limited data is premature.

Biological Data. The dl-erythro and dl-threo isomers of

Experimental Section# dl-erythro-ol-(2-Piperidyl)-3,6-bis(trifluoromethyl)-9-phenan-

threnemethanol Hydrochloride (I HCI). H, was passed through a mixt of 115 g (0.275 mole) of 2-pyridyl3,6-bis(trifluoromethyl)-9- phenanthryl ketone, 5.0 g of PtO, (Engelhard 85%), 4.2 1. of MeOH, and 40 ml of concd HCl for 16 hr. Darco was added and, after filtration, the filtrate was evapd in vacuo to 10% the original vol, pptg a mass of white crystals. The solids were dissolved in MeOH (Darco) and again concd to 10% the original vol to give 96.0 g (82.3%) of I .HCl, mp 331-332" dec. Anal. (C,,H,$JOF,. HC1) C, H, C1, N.

methanol Hydrochloride (I1 HCI). The mother liquors from the above recryst were concd to dryness, treated with dil K,CO, soln, and dried. The mixt of I and 11 (15.0 g, 35 mmoles) in 200 ml of MeOH was treated with 4.9 g (25 mmobs) of TsOH . H,O, refluxed

dl-t hreo-or-(2-Piperidyl)-3,6-bis(~uorome~yl)-9-ph~n~~ene-

§The synthesis and activity of I ' HC1 has been previously reported by Nodiff, et al.'

#All melting point (uncorr) were taken 01% a Buchi apparatus. In- struments employed were: Beckman IR-9 infiared spectrophotometer. Varian Model DP-60 high-resolution nmr spectrophotometer and Beckman DK-2 uv spectrophotometer. Elemental anal. were correct (*0.3%) and were performed b y Galbraith Laboratories, Inc., Knoxville, Tenn.

for 5 min, and cooled to ppt I1 . TsOH. Recrystn (2x, MeOH, Darco) gave an analytical material, mp 269-270". Anal. (C,,H,,NO,F,S) C, H, N, F.

I1 ' TsOH (50 g, 0.1 mole) was neutralized by stirring overnight with dil aq NaOH soh. I1 was dissolved in anhyd Et,O and satd with HCl g a s to ppt 11. HCl. Refluxing with CCl, removed a yellow impurity to leave 11. HC1 as a white powder, mp 284-285" (44.2 g, 95%). Anal. (C,,H,,NOF,. HCl) C , H, N, C1.

I- [ 3,6-Bis(trifluoromethyl)-9-phenan~l] hexahydro-3H-o~- azolo[3,46 ]pyridine. 111. A mixt of 4.9 g (1 1 mmoles) of I, 2 ml of (CH,O), soln, and 50 ml of MeOH was refluxed 8 hr. Add1 (C%O), s o h (2 ml) was added, the reflux was contd overnight. The mixt was cooled and filtered, and the product was recrystd (EtOH, Darco) to give 3.0 g (63%) of I11 as white flakes, mp 167-168'. Anal. (C,J-L.NOF,) C, H, N. _- _I

IV. A scmilar reaction of I1 gave 4.5 g of crude product. The solid was dissolved in CHCl,, poured onto a silica gel H column (75 g), and eluted with CHCl,. Recrystn (MeOH-H,O) gave 4.0 g (84%) of IV as a white powder, mp 181-182'. Anal. (C,,H,JVOF,) C, H, N.

Putnam for the nmr spectra. Acknowledgment. The author wishes to thank Roy

References (1) D. W. Boykin, A. R. Patel, R. E. Lutz, and A. Burger, J. Hetero-

cycl. Chem., 4,459 (1967). (2) A. D. Ainley and H. King, Proc. Roy. Soc., Ser. B, 125,60

(1938); R. F. Brown, et al., J. Amer. Chem. Soc., 68, 2705 (1946); E. R. Buchman and D. R. Howton, ibid., 68,2718 (1946); E. R. Buchman, H. Sargent, T. C. Meyers, and D. R. Howton, ibid., 68,2710 (1946); E. R. Buchman, H. Sargent, T. C. Meyers, and J. A. Seneker, ibid., 68,2697 (1946); J. B. Koepfli, M. M. Rapport, A. E. Senear, and J . F. Mead, ibid., 68,2697 (1946); R. E. Lutz, et al., ibid., 68, 1813 (1946); J. F. Mead, A. E. Senear, and J. B. Koepfli, ibid., 68,2708 (1946); R. A. Seibert, T. R. Norton, A. A. Benson, and F. W. Bergstrom, ibid., 68,2721 (1946); A. E. Senear, H. Sargent, J. F. Mead, and J. B. Koepfli, ibid., 68, 2605 (1946); S. Win- stein, T. L. Jacobs, E. F. Levy, D. Seymour, G. B. Linden, and R. B. Henderson, ibid., 68, 27 14 (1946).

(3) H. Sargent, ibid., 68,2688 (1946). (4) F. Y. Wiselogle, "A Survey of Antimalarial Drugs, 1941-

1945," J. W. Edwards, Ann Arbor, Mich., 1946. (5) E. A. Nodiff, K. Tanabe, C. Seyfried, S. Matsuura, Y. Kondo,

E. H. Chen, and M. P. Tyagi, J. Med. Chem., 14,921 (1971). (6) T. A. Crabb and R. F. Newton, J. Heterocycl. Chem., 3,418

(1966). (7) A. Dudas and I. Weisz, Chem. Ber., 94,412 (1960). (8) J. Kovar, J. J a y , and K. Blaha, Collect. Czech. Chem. Com-

(9) P. S. Portoghese, J. Med. Chem., 10, 1057 (1967). mun., 28,2199 (1963).

(10) M. E. Munk, M. K. Meilahn, and P. Franklin, J. Org. Chem., 33,3480 (1968).

(11) 0. L. Chapman and R. W. King, J. Amer. Chem. Soc., 86, 1256 (1964).

(12) T. S. Osdene, P. B. Russell, and L. Rane, J. Med. Chem., 10, 431 (1967).

N-Demethylation of Morphine and Structurally Related Compounds with Chloroformate Esterst M. M. Abdel-Monem and P. S. Portoghese* Department of Medicinal Chemistry, College of Pharmacy, University of Minnesota, Minneapolis, Minnesota 55455. Received August 5, I971

Hundreds of modifications of morphine and structurally related compounds have been performed and the compounds tested in an effort to analyze the relationship between structure and analgetic activity.ly2 The most common of these modifications is replacement of the Me group attached to the basic N with some other substituent. Thus,

?This investigation was supported by NIH Grant NS 08738.


N-Demethylation of morphine and structurallyrelated compounds with chloroformate esters

M. M. Abdel-Monem, and P. S. PortogheseJ. Med. Chem., 1972, 15 (2), 208-210• DOI: 10.1021/jm00272a025 • Publication Date (Web): 01 May 2002







?H

IA IB IC

IIA IIB IIC

Ar = 3,6-Bis(trifluoromethyl)-9-phenanthryl

coupling constants observed in DMSO for the free bases and their protonated forms. Evidently, on protonation, the bulk of the solvated quaternary N in I . HCl is significantly increased by H bonding with DMSO. This would enhance the population of IA at the expense of IB and IC. Similarly, protonation of I1 should increase the population of IIA with a concomitant decrease in IIB, due to the increased steric requirements of the solvated ammonium group.

a-(2-piperidyl)-3,6-bis(trifluoromethyl)-9-phenanthrenemethanol were tested12 for antimalarial activity as their HCl and TsOH salts and as their oxazolopyridine derivatives against Plasmodium berghei in mice and P. gallinaceum in chicks by Dr. Leo Rane at the University of Miami. The test results, furnished to us through the Walter Reed Army Institute of Research, show these materials to be highly active against P. berghei, giving 5 cures out of 5 infected mice at dosages of 40 mg/kg for the threo epimers and 5 cures at 80 mg/kg for the erythro compds.§ No toxic deaths were reported up to dosages of 640 mg/kg. Against P. gal- Zinaceum, the minimum dosage showing activity was 20 and 160 mg/kg for the threo and erythro compds, respectively.

To attempt a general correlation of epimer conformation with antimalarial activity on the above limited data is premature.

Biological Data. The dl-erythro and dl-threo isomers of

Experimental Section# dl-erythro-ol-(2-Piperidyl)-3,6-bis(trifluoromethyl)-9-phenan-

threnemethanol Hydrochloride (I HCI). H, was passed through a mixt of 115 g (0.275 mole) of 2-pyridyl3,6-bis(trifluoromethyl)-9- phenanthryl ketone, 5.0 g of PtO, (Engelhard 85%), 4.2 1. of MeOH, and 40 ml of concd HCl for 16 hr. Darco was added and, after filtration, the filtrate was evapd in vacuo to 10% the original vol, pptg a mass of white crystals. The solids were dissolved in MeOH (Darco) and again concd to 10% the original vol to give 96.0 g (82.3%) of I .HCl, mp 331-332" dec. Anal. (C,,H,$JOF,. HC1) C, H, C1, N.

methanol Hydrochloride (I1 HCI). The mother liquors from the above recryst were concd to dryness, treated with dil K,CO, soln, and dried. The mixt of I and 11 (15.0 g, 35 mmoles) in 200 ml of MeOH was treated with 4.9 g (25 mmobs) of TsOH . H,O, refluxed

dl-t hreo-or-(2-Piperidyl)-3,6-bis(~uorome~yl)-9-ph~n~~ene-

§The synthesis and activity of I ' HC1 has been previously reported by Nodiff, et al.'

#All melting point (uncorr) were taken 01% a Buchi apparatus. In- struments employed were: Beckman IR-9 infiared spectrophotometer. Varian Model DP-60 high-resolution nmr spectrophotometer and Beckman DK-2 uv spectrophotometer. Elemental anal. were correct (*0.3%) and were performed b y Galbraith Laboratories, Inc., Knoxville, Tenn.

for 5 min, and cooled to ppt I1 . TsOH. Recrystn (2x, MeOH, Darco) gave an analytical material, mp 269-270". Anal. (C,,H,,NO,F,S) C, H, N, F.

I1 ' TsOH (50 g, 0.1 mole) was neutralized by stirring overnight with dil aq NaOH soh. I1 was dissolved in anhyd Et,O and satd with HCl g a s to ppt 11. HCl. Refluxing with CCl, removed a yellow impurity to leave 11. HC1 as a white powder, mp 284-285" (44.2 g, 95%). Anal. (C,,H,,NOF,. HCl) C , H, N, C1.

I- [ 3,6-Bis(trifluoromethyl)-9-phenan~l] hexahydro-3H-o~- azolo[3,46 ]pyridine. 111. A mixt of 4.9 g (1 1 mmoles) of I, 2 ml of (CH,O), soln, and 50 ml of MeOH was refluxed 8 hr. Add1 (C%O), s o h (2 ml) was added, the reflux was contd overnight. The mixt was cooled and filtered, and the product was recrystd (EtOH, Darco) to give 3.0 g (63%) of I11 as white flakes, mp 167-168'. Anal. (C,J-L.NOF,) C, H, N. _- _I

IV. A scmilar reaction of I1 gave 4.5 g of crude product. The solid was dissolved in CHCl,, poured onto a silica gel H column (75 g), and eluted with CHCl,. Recrystn (MeOH-H,O) gave 4.0 g (84%) of IV as a white powder, mp 181-182'. Anal. (C,,H,JVOF,) C, H, N.

Putnam for the nmr spectra. Acknowledgment. The author wishes to thank Roy

References (1) D. W. Boykin, A. R. Patel, R. E. Lutz, and A. Burger, J. Hetero-

cycl. Chem., 4,459 (1967). (2) A. D. Ainley and H. King, Proc. Roy. Soc., Ser. B, 125,60

(1938); R. F. Brown, et al., J. Amer. Chem. Soc., 68, 2705 (1946); E. R. Buchman and D. R. Howton, ibid., 68,2718 (1946); E. R. Buchman, H. Sargent, T. C. Meyers, and D. R. Howton, ibid., 68,2710 (1946); E. R. Buchman, H. Sargent, T. C. Meyers, and J. A. Seneker, ibid., 68,2697 (1946); J. B. Koepfli, M. M. Rapport, A. E. Senear, and J . F. Mead, ibid., 68,2697 (1946); R. E. Lutz, et al., ibid., 68, 1813 (1946); J. F. Mead, A. E. Senear, and J. B. Koepfli, ibid., 68,2708 (1946); R. A. Seibert, T. R. Norton, A. A. Benson, and F. W. Bergstrom, ibid., 68,2721 (1946); A. E. Senear, H. Sargent, J. F. Mead, and J. B. Koepfli, ibid., 68, 2605 (1946); S. Win- stein, T. L. Jacobs, E. F. Levy, D. Seymour, G. B. Linden, and R. B. Henderson, ibid., 68, 27 14 (1946).

(3) H. Sargent, ibid., 68,2688 (1946). (4) F. Y. Wiselogle, "A Survey of Antimalarial Drugs, 1941-

1945," J. W. Edwards, Ann Arbor, Mich., 1946. (5) E. A. Nodiff, K. Tanabe, C. Seyfried, S. Matsuura, Y. Kondo,

E. H. Chen, and M. P. Tyagi, J. Med. Chem., 14,921 (1971). (6) T. A. Crabb and R. F. Newton, J. Heterocycl. Chem., 3,418

(1966). (7) A. Dudas and I. Weisz, Chem. Ber., 94,412 (1960). (8) J. Kovar, J. J a y , and K. Blaha, Collect. Czech. Chem. Com-

(9) P. S. Portoghese, J. Med. Chem., 10, 1057 (1967). mun., 28,2199 (1963).

(10) M. E. Munk, M. K. Meilahn, and P. Franklin, J. Org. Chem., 33,3480 (1968).

(11) 0. L. Chapman and R. W. King, J. Amer. Chem. Soc., 86, 1256 (1964).

(12) T. S. Osdene, P. B. Russell, and L. Rane, J. Med. Chem., 10, 431 (1967).

N-Demethylation of Morphine and Structurally Related Compounds with Chloroformate Esterst M. M. Abdel-Monem and P. S. Portoghese* Department of Medicinal Chemistry, College of Pharmacy, University of Minnesota, Minneapolis, Minnesota 55455. Received August 5, I971

Hundreds of modifications of morphine and structurally related compounds have been performed and the compounds tested in an effort to analyze the relationship between structure and analgetic activity.ly2 The most common of these modifications is replacement of the Me group attached to the basic N with some other substituent. Thus,

?This investigation was supported by NIH Grant NS 08738.


the secondary amine serves as an important intermediate in the synthesis of such compounds. Although there are several methods reported"' for the demethylation of tertiary amines, they suffer from the disadvantage of the toxicity of reagents employed and/or the low yields of demethylated product.

One approach to this problem which seemed worthwhile exploring involved demethylation of analgetics with a chloroformate ester' followed by hydrolysis of the resulting carbamate to the secondary amine as outlined below.

Cleavage of tertiary amines with chloroformate esters was first reported in 191 1.8 Gadamer and Knoch' studied the effect of ethyl chloroformate on a variety of cyclic tertiary amines and observed that bulbocapnine, corydine, and laudanosine were converted to the corresponding ethyl carbamates, whereas compounds such as morphine, codeine, heroin, and tropine were not cleaved. Several examples of this reaction have been published since,'@l3 and in this regard, phenyl chloroformate was found to be superior to both benzyl and ethyl chloroformate in the cleavage of tertiary amines.14 We wish to describe the utilization of this reaction as a means of conveniently demethylating, in high yield, structures related to morphine.

the presence of KHC03 in boiling CHC13 and subsequent treatment of the product with a mild base afforded, after purification, a nonbasic crystalline material which was assigned structure 2, based on its spectral properties. This assignment was confirmed further by LAH reduction of 2 to morphine and by its hydrolytic conversion to normorphine (3) with KOH. It is noteworthy that the reaction of morphine with phenyl chloroformate was unsuccessful in the absence of base, possibly because esterification of the 3- or 6-OH generates HCl which may protonate the basic N and hence diminish its reactivity.

Codeine (4) was demethylated with ethyl chloroformate in a two-phase system containing aq KOH and CHC13. The spectral characteristics of the neutral intermediate appear to be in harmony with 5. The ethyl carbonate group of 5 could be hydrolyzed selectively to 6 which then was cleaved to norcodeine (7) under more vigorous conditions.

Treatment of morphine (1) with phenyl chloroformate in

Ron RoR 1, R = R' = H; R" = Me 2, R = R' = H; R" = COOPh 3, R = R ' = R " = H 10. R = M e : R ' = H

8, R = R' = M e 9, R = Me; R' = COOPh

4; R = Me;R' = H ; R" =Me 5, R = Me; R' = R" = COOEt 6, R = Me; R' = H; R" = COOEt 7, R = Me; R' = R" = H

Reaction of 3-methoxy-N-methylmorphinan (8) with phenyl chloroformate at room temp afforded carbamate ester 9. The carbamate was obtained in crystalline form by selectively hydrolyzing the phenyl chloroformate contaminant with aq K2C03. Structure 9 was corroborated by its reconversion to 8 with LAH. Hydrolysis of 9 with KOH gave the desired secondary amine (10). Demethylation of 8 also was carried out with ethyl chloroformate. Hydrolysis

11; R = Me: R ' = COOEt 12, R = R' = H

of this product (1 1) with HBr-HOAc afforded 3-hydroxymorphinan (12).

It is noteworthy that the present procedure also can be employed as a convenient and relatively inexpensive method for introducing radioactivity into a N-Me group. Thus, [3H]LAH reduction of the carbamate intermediate obtained from the demethylation step would afford a 3H- labeled Me group.$ A 14C label may be incorporated by demethylating with [ ''C]C1COOR followed by r ed~c t ion . '~

Experimental Section! N-Carbophenoxynormorphine (2). A stirred suspension of 2.5

g (0.0088 mole) of morphine (1) and 15 g (0.15 mole) of KHCO, in CHCI, (250 ml) was treated with 11.5 g (0.077 mole) of phenyl chloroformate and refluxed for 60 hr. The mixt was treated with H,O (100 ml) and the CHCI, phase was sepd and concd in vacuo. The residue was dissolved in MeOH (150 ml), treated with an aq soln (100 ml) contg 5.6 g of KOH and 10.0 g of KHCO, , and stirred under N, for 24 hr. After the mixt was acidified with concd HCl, the MeOH was removed and the residue was dild with H,O and extd (Et,O). The Et,O was removed and the oily residue was chromatogd on a column of 45 g of silica gel (E. Merck AG, 70-325 mesh) and eluted with Et,O. Fractions 6-12 (50 ml each) were pooled, and the solvent was removed in vacuo to afford 3.12 g of 2: mp 127-130' (yield 91%); mass spec (70 eV) m/e 391. Anal. (C,,H,,NO,) C, H, N.

LAH Reduction of Carbophenoxynormorphe (2). A THF soln (50 ml) contgO.8 g (0.002 mole) of 2 was added slowly to a cooled, stirred suspension of LAH (0.2 g) in THF (150 ml). The mixt was refluxed under N, for 20 hr, cooled, treated with EtOAc (15 ml), and refluxed for 30 min. The mixt was treated with 2 N HCl(50 ml) and sodium potassium tartrate (6 g) and refluxed for 4 hr. This was concd in vacuo to remove THF and extd with Et,O. The acid phase was adjusted to pH 8.5 and extd with a mixt of CHC1,-i-PrOH (3:l). The exts were concd in vacuo, and the residue was crystd from 50 ml of MeOH-H,O (1:4) to afford 0.45 g of morphine: mp 252- 254" (reported 254-256")16 (yield 76%).

Normorphine (3). A soln of 0.4 g (0.001 mole) of 2 in a mixt of EtOH (80 ml) and 50% KOH (20 ml) was refluxed under N, for 24 hr. The soh was treated with concd HCl(20 ml), the EtOH was removed in vacuo, and the residue was extd with Et,O. The Et,O ext contd some unreacted 2 as shown by tlc. The acid phase was filtered, adjusted to pH 8.5, and extd with CHC1,-i-PrOH (3:l). The solvent was removed in vaeuo to afford 0.12 g of 3: mp 277' dec (reported 276-277");" 3 . HCl, mp 307"dec (reported 305' dec);I7 yield, 43.5%.

(0.01 1 mole) of 8 and 2.0 g (0.013 mole) of phenyl chloroformate in CH,C1, (30 ml) was maintd at 25' for 24 hr. The CH,CI, was removed in vacuo, and the residue was suspended in 1 N HCl (50 ml) and extd with Et,O. Basification of the aq soln afforded 0.86 g of unreacted 8. The Et,O soln was concd under reduced pressure and the oily residue was dissolved in MeOH (60 ml) and treated with 2% aq K,CO, (40 ml). The mixt was maintd at room temp for 24 hr and the MeOH removed in vacuo. Crystn (MeOH-H,O) afforded 2.7 g (yield, 65%) of 9, mp 100-102". Anal. (C,.,H,,NO,) C, H, N.

soln (80 ml) contg 2.5 g (0.007 mole) of 9 was treated with 50% aq KOH (20 ml) and the reaction mixt was refluxed under N, for 24 hr. The soln was dild with HzO (20 ml) and the EtOH was partially removed under reduced pressure. The aq suspension was extd with Et,O and the Et,O layer was extd with 1 N HC1. Removal of E t20 afforded 0.8 g of 9. The aq acid soln was filtered, basified, and extd with Et,O. The Et,O was dried (MgSO,) and treated with ethanolic HC1 to give 1.45 g of crude 10. HCl: yield, 74.5%; recrystd 10. HCl (MeOH-EtOAc), mp 253.5-255'. Anal. (C]&$JO*HCl) C, H, N.

LAH Reduction of 9. A THF soln (100 ml) contg 0.8 g (0.002 mole) of 9 was added slowly to a cooled, stirred suspension of 0.2 g of LAH in THF (100 ml). The mixt was refluxed under N, for 4 hr,

(+)-3-Methoxy-l7-carbophenoxymorphinan (9). A soln of 3.0 g

(+)-3-Methoxymorphinan Hydrochloride (10 HC1). An EtOH

SA. E. Takemori and J . Wang, Dept. of Pharmacology, University of Minnesota, have successfully prepared 'H-labeled morphine b y employing a modification of our procedure.

!Melting points were detd in open capillary tubes using a Thomas- Hoover melting point apparatus and are uncorrected. Microanalyses were performed by M-H-W Laboratories, Garden City, Mich. The ir spectra were obtd with a Perkin-Elmer 237B spectrophotometer in CHCI, solution or KBr disk. The nmr spectra were obtd with a Varian A-60D spectrometer (CDCl,, TMS). Mass spectra were obtd with a Hitachi Perkin-Elmer RMU-CD mass spectrometer.

210 Journal ofMedicinal Chemistry, 1972, Vol. 15, No. 2 Notes

cooled, treated with EtOAc (10 ml), and refluxed for 30 min. After addn of H,O and 1 N NaOH altemately, the mixt was filtered and concd in vacuo t o remove THF. The residue was acidified with 6 N HCl and extd with Et,O. The aq acid soln was basified and extd with Et,O, and the dried Et,O ext was treated with ethanolic HC1 to afford 0.51 g (yield, 78%) of 8 * HC1.

(+)-3-Hydroxymorphinan (12). A CHCl, soln (25 ml) contg 4.0 g (0.0015 mole) of 7 was treated with ethyl chloroformate (20 ml) and anhyd K,C0,,(2 g). The reaction mixt was stirred for 24 hr and the CHC1, then removed in vacuo. The residue was suspended in 1 N HCl(50 ml) and extd with Et,O. Removal of Et,O afforded an oil (2.5 g) which was dissolved in a mixt of glacial AcOH (10 ml) and 48% HBr (10 ml). After refluxing under N, for 6 hr, the reaction mixt was poured in ice water (300 ml), extd with Et,O, and basified to afford 1.1 g (31%) of 12, mp 260-263" [reported for (-)-3-hydroxymorphinan, mp 260-262'1 . I *

CHCl, (50 ml) was treated with ethyl chloroformate (5 ml) and 15% aq KOH (50 ml). The 2-phase reaction mixt was shaken for 24 hr. Five successive 1-ml portions of ethyl chloroformate were added to the reaction mixt at 1-hr intervals, and KOH soh was added when necessary to maintain a pH > 10. At the end of 24 hr, the CHCl, layer was sepd and extd with 1 N HCl. The CHC1, was removed in vacuo to afford an oil (3.2 g) whose ir spectrum included characteristic absorptions at 1690 cm-' (C=O of N-C0,Et) and 1740 cm-' (C=O of 06-C02Et). This was treated with a mixt of MeOH (90 ml) and 10% aq K,CO,(10 ml) for 2 hr, concd in vacuo to remove MeOH, extd with Et,O, and the Et,O was extd with 1 N HC1. Removal of Et,O afforded 2.5 g of an oily uncrystallizable material. A soln of 2.0 g of the oil in 95% EtOH (80 ml) was treated with 50% aq KOH (20 ml) and refluxed under N, for 24 hr. The soln was dild with H,O (20 ml) and the EtOH was removed under reduced pressure. The aq acid soln was basified and extd with Et,O. The Et,O was removed under reduced pressure to give 0.7 g (yield, 43%) of 7: mp 183-185" (reported mp 185");" 7.HC1, mp 309- 311" dec (reported mp 309" dec).Ig

Norcodeine (7). A soln of 3.8 g (0.012 mole) of codeine (4) in

References (1) P. S . Portoghese, J. Pharm. Sei., 5 5 , 865 (1966). (2) "Analgetics," G. deStevens, Ed., Academic Press, New York,

(3) H. A. Hageman, Org. React., 7, Chapter 4 (1953). (4) 0. Diels and E. Fischer, Ber., 47,2043 (1914). (5) A. Pohland and H. R. Sullivan, Jr., U. S. Patent 3,342,824

(6) E. Speyer and L. Walther, Ber., 69B, 852 (1930). (7) E. Spath, L. Marion, and E. Zajic, ibid., 69B, 251 (1936). (8) F. Bayer and Co., German Patent 255,942 (1911). (9) J. Gadamer and F. Knoch, Arch. Pkarm. (Weinheim), 259, 135

N. Y., 1965.

(1967).

(1921).

Chem. Soc., 71, 3104 (1955). (10) E. H. Flynn, €I. W. Murphy, and R. D. McMahon, J. Amer.

(11) J. A. Campbell, J. Org. Chem., 22, 1259 (1957). (12) M. M. Abdel-Monem and T. 0. Soine, J. Pkarm. Sei., 56,976

(13) G. Kraiso and L. Nador, Tetrahedron Lett., 57 (1971). (14) J. D. Hobson and J. B. McCluskey, J. Chem. Soc., 15 (1967). (15) R. E. McMahon in "Advances in Tracer Methodology," Vol. 4,

S. Rothchild, Ed., Plenum Press, New York, N. Y., 1968, p 29. (16) Merck Index, 8th Ed., D. G. Stecher, Ed., Merck and Co., Inc.,

Rahway, N. J., 1968, p 702. (17) Reference 16, p 749. (18) 0. Schnider and A. Grussner, Helv. Chim. Acta, 34, 221 1

(19) Reference 16, p 747.

(1967).

(1951).

Synthesis of 4(3H)-Pteridinones Ernst Felder,* Davide Pit&, and Sergio Boveri Research Laboratories, Bracco Industria Chimica, Milan, Italy. Received May 24, 1971

The sedative-hypnotic activity of some 4(3H)-quinazo- lones' prompted us to synthesize a number of the isosteric 4(3H)-pteridinones and to investigate their hypnotic and

sedative activities. The existing literature gives only a few examples of preparation of 3-alkyl or 3-aryl substituted 4(3H)-pteridinone~~-~ and no data at all on their pharmacology. The synthesis of the title compounds involved the intermediate 3-aminopyrazinecarboxamides, described in Table I , which were obtained, in good yield, from 3-aminopyrazinoic acid (I),' via the mixed anhydride (11)6 and reaction of the latter with appropriate amines (RzNHz).

R , ~ x y COOH

I

R y ~ y C O O C O O C , H , RIY\YCONHRz

I1 I11

0 I1

IV

The 3-aminopyrazinecarboxamides (111) could be cyclized to the desired 4(3H)-pteridinones (IV) by condensation with an ortho ester R3C(OC2H5)3 in AczO solution. The amides (III), in contrast t o the 4(3H)-pteridinones (IV), reveal a characteristic fluorescence under uv light, which is helpful for their identification by chromatography.

In preliminary CNS screening the majority of the compounds were found to be without hypnotic or sedative activity. Compds 23,30, and 31 showed a slight sedative activity at 300 mg/kg (mouse) and with 30,31, and 38 some analgetic activity was observed at 150-250 mg/kg (mouse; phe- nylbenzoquinone test) and 150-500 mg/kg (mouse; hot- plate test), but all the compounds showed too low a therapeutic index, the LDS0 (mg/kg; mouse; Litchfield and Wil- coxon) being 1250 (1042-1500), 575 (483-684), 1220 (1070-1391), and 1750 (1400-2187) for 23,30,31, and 38, respectively.


a Mettler FP-1 apparatus, all the others with a Buchi apparatus, and are uncorrected. Uv and ir spectra were measured for some typical compds and were as expected. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within +0.4% of the theoretical value.

3-Aminopyrazinecarboxamides (1-20). A mixt of 5.56 g (0.04 mole) of 3-aminopyrazinoic acid, 7.4 g (0.04 mole) of Bu,N, and 50 ml of dioxane was stirred at room temp until a clear soln resulted. This soln was cooled to 7-8" and 4 ml (0.04 mole) of EtOCOCl was added dropwise, keeping the temp at 11-12'. After cooling again to 7-8', 0.04 mole of the appropriate amine hydrochloride was added, and the reaction was allowed to proceed at room temp for 3 hr. The solvent was removed on a rotatory evaporator under reduced pressure and the residue was triturated for 30 min with 50 ml of H,O, filtered, dried, and recrystd. Recrystn solvents and physical data are given in Table I.

4(3H)-Pteridinones (2147). A mixt of 0.01 mole of 111, 25 ml of ortho ester, and 20-30 ml of Ac,O was refluxed for 5 hr and then concd on a rotatory evaporator at room temp in vacuo. The residue was triturated with 20 ml of EtOH, and, after evapn of the solvent, washed with Et,O, filtered, dried, and recrystd. Recrystn solvents and physical data are given in Table 11. For 23 the reaction was carried out in anhyd HCO,H, and for 24 in 1 : 1 anhyd HC0,H-Ac,O

The melting points of all but four compds 21-47 were taken with


Synthesis of 4(3H)-pteridinonesErnst Felder, Davide Pitre, and Sergio Boveri








cooled, treated with EtOAc (10 ml), and refluxed for 30 min. After addn of H,O and 1 N NaOH altemately, the mixt was filtered and concd in vacuo t o remove THF. The residue was acidified with 6 N HCl and extd with Et,O. The aq acid soln was basified and extd with Et,O, and the dried Et,O ext was treated with ethanolic HC1 to afford 0.51 g (yield, 78%) of 8 * HC1.

(+)-3-Hydroxymorphinan (12). A CHCl, soln (25 ml) contg 4.0 g (0.0015 mole) of 7 was treated with ethyl chloroformate (20 ml) and anhyd K,C0,,(2 g). The reaction mixt was stirred for 24 hr and the CHC1, then removed in vacuo. The residue was suspended in 1 N HCl(50 ml) and extd with Et,O. Removal of Et,O afforded an oil (2.5 g) which was dissolved in a mixt of glacial AcOH (10 ml) and 48% HBr (10 ml). After refluxing under N, for 6 hr, the reaction mixt was poured in ice water (300 ml), extd with Et,O, and basified to afford 1.1 g (31%) of 12, mp 260-263" [reported for (-)-3-hydroxymorphinan, mp 260-262'1 . I *

CHCl, (50 ml) was treated with ethyl chloroformate (5 ml) and 15% aq KOH (50 ml). The 2-phase reaction mixt was shaken for 24 hr. Five successive 1-ml portions of ethyl chloroformate were added to the reaction mixt at 1-hr intervals, and KOH soh was added when necessary to maintain a pH > 10. At the end of 24 hr, the CHCl, layer was sepd and extd with 1 N HCl. The CHC1, was removed in vacuo to afford an oil (3.2 g) whose ir spectrum included characteristic absorptions at 1690 cm-' (C=O of N-C0,Et) and 1740 cm-' (C=O of 06-C02Et). This was treated with a mixt of MeOH (90 ml) and 10% aq K,CO,(10 ml) for 2 hr, concd in vacuo to remove MeOH, extd with Et,O, and the Et,O was extd with 1 N HC1. Removal of Et,O afforded 2.5 g of an oily uncrystallizable material. A soln of 2.0 g of the oil in 95% EtOH (80 ml) was treated with 50% aq KOH (20 ml) and refluxed under N, for 24 hr. The soln was dild with H,O (20 ml) and the EtOH was removed under reduced pressure. The aq acid soln was basified and extd with Et,O. The Et,O was removed under reduced pressure to give 0.7 g (yield, 43%) of 7: mp 183-185" (reported mp 185");" 7.HC1, mp 309- 311" dec (reported mp 309" dec).Ig

Norcodeine (7). A soln of 3.8 g (0.012 mole) of codeine (4) in

References (1) P. S . Portoghese, J. Pharm. Sei., 5 5 , 865 (1966). (2) "Analgetics," G. deStevens, Ed., Academic Press, New York,

(3) H. A. Hageman, Org. React., 7, Chapter 4 (1953). (4) 0. Diels and E. Fischer, Ber., 47,2043 (1914). (5) A. Pohland and H. R. Sullivan, Jr., U. S. Patent 3,342,824

(6) E. Speyer and L. Walther, Ber., 69B, 852 (1930). (7) E. Spath, L. Marion, and E. Zajic, ibid., 69B, 251 (1936). (8) F. Bayer and Co., German Patent 255,942 (1911). (9) J. Gadamer and F. Knoch, Arch. Pkarm. (Weinheim), 259, 135

N. Y., 1965.

(1967).

(1921).

Chem. Soc., 71, 3104 (1955). (10) E. H. Flynn, €I. W. Murphy, and R. D. McMahon, J. Amer.

(11) J. A. Campbell, J. Org. Chem., 22, 1259 (1957). (12) M. M. Abdel-Monem and T. 0. Soine, J. Pkarm. Sei., 56,976

(13) G. Kraiso and L. Nador, Tetrahedron Lett., 57 (1971). (14) J. D. Hobson and J. B. McCluskey, J. Chem. Soc., 15 (1967). (15) R. E. McMahon in "Advances in Tracer Methodology," Vol. 4,

S. Rothchild, Ed., Plenum Press, New York, N. Y., 1968, p 29. (16) Merck Index, 8th Ed., D. G. Stecher, Ed., Merck and Co., Inc.,

Rahway, N. J., 1968, p 702. (17) Reference 16, p 749. (18) 0. Schnider and A. Grussner, Helv. Chim. Acta, 34, 221 1

(19) Reference 16, p 747.

(1967).

(1951).

Synthesis of 4(3H)-Pteridinones Ernst Felder,* Davide Pit&, and Sergio Boveri Research Laboratories, Bracco Industria Chimica, Milan, Italy. Received May 24, 1971

The sedative-hypnotic activity of some 4(3H)-quinazo- lones' prompted us to synthesize a number of the isosteric 4(3H)-pteridinones and to investigate their hypnotic and

sedative activities. The existing literature gives only a few examples of preparation of 3-alkyl or 3-aryl substituted 4(3H)-pteridinone~~-~ and no data at all on their pharmacology. The synthesis of the title compounds involved the intermediate 3-aminopyrazinecarboxamides, described in Table I , which were obtained, in good yield, from 3-aminopyrazinoic acid (I),' via the mixed anhydride (11)6 and reaction of the latter with appropriate amines (RzNHz).

R , ~ x y COOH

I

R y ~ y C O O C O O C , H , RIY\YCONHRz

I1 I11

0 I1

IV

The 3-aminopyrazinecarboxamides (111) could be cyclized to the desired 4(3H)-pteridinones (IV) by condensation with an ortho ester R3C(OC2H5)3 in AczO solution. The amides (III), in contrast t o the 4(3H)-pteridinones (IV), reveal a characteristic fluorescence under uv light, which is helpful for their identification by chromatography.

In preliminary CNS screening the majority of the compounds were found to be without hypnotic or sedative activity. Compds 23,30, and 31 showed a slight sedative activity at 300 mg/kg (mouse) and with 30,31, and 38 some analgetic activity was observed at 150-250 mg/kg (mouse; phe- nylbenzoquinone test) and 150-500 mg/kg (mouse; hot- plate test), but all the compounds showed too low a therapeutic index, the LDS0 (mg/kg; mouse; Litchfield and Wil- coxon) being 1250 (1042-1500), 575 (483-684), 1220 (1070-1391), and 1750 (1400-2187) for 23,30,31, and 38, respectively.


a Mettler FP-1 apparatus, all the others with a Buchi apparatus, and are uncorrected. Uv and ir spectra were measured for some typical compds and were as expected. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within +0.4% of the theoretical value.

3-Aminopyrazinecarboxamides (1-20). A mixt of 5.56 g (0.04 mole) of 3-aminopyrazinoic acid, 7.4 g (0.04 mole) of Bu,N, and 50 ml of dioxane was stirred at room temp until a clear soln resulted. This soln was cooled to 7-8" and 4 ml (0.04 mole) of EtOCOCl was added dropwise, keeping the temp at 11-12'. After cooling again to 7-8', 0.04 mole of the appropriate amine hydrochloride was added, and the reaction was allowed to proceed at room temp for 3 hr. The solvent was removed on a rotatory evaporator under reduced pressure and the residue was triturated for 30 min with 50 ml of H,O, filtered, dried, and recrystd. Recrystn solvents and physical data are given in Table I.

4(3H)-Pteridinones (2147) . A mixt of 0.01 mole of 111, 25 ml of ortho ester, and 20-30 ml of Ac,O was refluxed for 5 hr and then concd on a rotatory evaporator at room temp in vacuo. The residue was triturated with 20 ml of EtOH, and, after evapn of the solvent, washed with Et,O, filtered, dried, and recrystd. Recrystn solvents and physical data are given in Table 11. For 23 the reaction was carried out in anhyd HCO,H, and for 24 in 1 : 1 anhyd HC0,H-Ac,O

The melting points of all but four compds 21-47 were taken with

Notes Journal of Medicinal Chemistry, I9 72, Vol. 15, No. 2 21 1

Table I. 3-Aminopyrazinecarboxamides (111) No. R, R, Formula MP, "C Crystn solvent Anal.

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20

CH3 CH(CH3), Cyclo-C5H9 Cyclo-C,H,, C6H5 o-CH3C6H4 m-CH3C6H4

o-(i-Pr)C6H4 p-(i-Pr)C6H4

m-EtOC6H4 p-EtOC,H.

p-CH3C6H4

O-EtOC6H4

H H H H H H H H H H H H H H H H H H Br Br

C6H8N40 C8H, ,N,O * HCl * H,O C10H14N40

'1 lHl 6N40

CI lHl,N,O c 1 ,HI 8 4 0

CiaHiiN40 C11H12N40

14H1 6N40

16N40

I 3H14N40,

I ,H14N402

'1 3H14N402

C15H18N402

C19H18N402

leH, C,,H9CIN40

C, ,H9BrN40 C10H9N50

Cl I B I N 8

134 134 73

105 136 112 158 154 125

107-108

134-135 104-105 139-140 134 180 136 190-191 155-156 157 176

33% EtOH EtOH

33% EtOH 50% EtOH 50% EtOH 50% EtOH EtOH 95% EtOH 95% EtOH 65% EtOH 65% EtOH 65% EtOH EtOH n-BuOH 95% EtOH Dioxane-%O EtOH 95% EtOH 95% EtOH

EtOH-H,O

Table 11. 4(3H)-Pteridinones (IV) No. R, R3 Rl Formula MP, "C Crystn solventa Anal.

2 1 CHI H H C7H6N40 294.5-296.5 204.5-207

23 o-CH3C6H4 H H C13H10N40 166.5-172.5 24 m-CH3C6H4 H H 1 3H10N40 235-236.5 25 PC2H30C6H4 H H C14H11N402 239-24 1.5

22 CH(CH3), H H 'SH1 ON,'

26 o-CH3C6H4 H Br C, ,H,BrN40 -147d H C7H6N40 >299

202-205.5 H C12H14N40 164.5-167 H C13H16N40 203-204.5 H C14H11N40 188.5-190.5

201-204.5 33 PCH3C6H4 CH3 H IsH, ZN4O 226-227.5 34 o-(i-Pr)C6H4 CH3 H 6N40 179.5-1 82.5

H C 1 6 H 1 3 4 0 177.5-180.5 H C15H14N402 240.5 -24 3.5

H C I O H l l N 4 0

27 H CH3 28 CH(CH3)z CH3 29 Cyclo-C,H, CH3 30 Cyclo-C6Hll CH3 31 o-CH,C6H4 CH3 32 m-CH3C6H4 CH3

35 p-(i-h)C6H4 CH3 36 O-EtOC6H4 CH3

H a N 4 0

31 m-EtOC,H, CH3 H 1 5H14N401 148d

39 p-C4H90C6H4 CH3 H C17H1$r1402 -144d 38 p- EtOC 6H4 CH3 H ClSH14N40, 196-198.5

40 C6H5(CH,)@C& CH3 H C21H18N401 155-159.5 C H C H

41 -CH-CHl CH3 H C21H 19 8.5 -204 l 6 l 6

200.5 dec 42 3-Py CH3 H 1,H9NS0 Br C,,H,BrN,O -195d H 14H 1 E N 4 0 147.5-150.5

45 p- (i-Pr)C6H4 C2H5 H Cl,Hl8N,O 160-163 46 pEtOC,H, C2H5 H C16H16N402 165.5-167.5 41 pEtOC6H4 CH,C6H5 H C21H18N402 117-121.5

43 C6H5 CH, 44 Cyclo-C,H,, C2H5

aP = petr ether. bCalcd, 62.67; found, 62.15. CCalcd, 63.82; found, 64.35. dBiichi apparatus.

MeCN EtOH C6H6-P CHCl -Et 2 0

MeCN EtOH MeCN THF n-Bu,O C6H61P

EtOH C,H6-P

EtOH EtOH EtOH-i-Pr,O MeOEtOH EtOH EtOH EtOH EtOH

C6H6-P

C6H6-P EtOH

EtOH EtOH EtOH EtOH-Et aO

C; Hi N H. N. Cb

Acknowledgment. We thank Dr. G. Rosati, of these Re- 43,637,(1955). search Laboratories, for the pharmacological screening. (2) E. C. Taylor, Jr., J. Amer. Chem. SOC., 74,1651 (1952).

(3) J. H. Jones, J. B. Bicking, and E. J. Cragoe, Jr., J. Med. Chem., 10.899 (1967).

(4) A.'Albert and D. J. Brown, J. Chem. SOC., 74 (1953). (5) S. Gabriel and A. Sonn, Chem. Ber., 40,4850 (1907). (6) D. Pit&, S. Boveri, and E. Grabitz, ibid., 99,364 (1966).

References (1) M. L. Gujral, P. N. Saxena, and R. S. Tiwari, Indian Med. Res.,


Hepatocarcinogenicity of the trimethylhomologs of 4-dimethylaminoazobenzene

Ellis V. Brown, and Alice KruegelJ. Med. Chem., 1972, 15 (2), 212-212• DOI: 10.1021/jm00272a027 • Publication Date (Web): 01 May 2002







Hepatocarcinogenicity of the Trimethyl Homologs of 4-Dimethylaminazobenzene Ellis V. Brown* and Alice Kruegel Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506. Received August I I , 19 71

Definite to strong carcinogenic activity has been shown by some monosubstituted and disubstituted derivatives of 4-dimethylaminoazobenzene (DAB). 1-3 The only active tri- substituted-DAB tested has been the 2’,4’,6‘-trifluoro deriv- a t i ~ e . ~ It seemed of interest to synthesize and test for rat hepatocarcinogenic activity all of the trimethyl homologs of DAB with Me groups in the primed positions only. These are all new compounds and can be prepared by the diazotization of the proper trimethylanilines followed by coupling with PhNMez. The new azo compounds are listed in Table 1.

Experimental Section All melting points were detd on a Fisher-Johns apparatus and

are uncorrected. The C, H, N analyses were performed in this department on an F and M Model 185 analyzer by MI. Daryl Sharp. Where analyses are indicated only by symbols of the elements analytical results obtained for those elements were within +0.4% of the theoretical values.

Trimethylmilines. Me~idine,~ bp 224-228”, was prepd from nitromesitylene6 by reduction with Sn-HCl. 5-Aminopseudocumene, mp 62-63’, was prepd from 5-nitropseud~cumene~ in the same way. 4-Amin0hemimellitene,~ mp 29-27”, was obtd from 5-nitrohemi- mellitene,’ 6-aminopse~documene~ was prepd from 6-nitropseudo- cumene,” and 5-aminohemimellitene,” mp 75-78’, was prepared from 5-nit~ohemimellitene~ by reduction (Fe-AcOH). 3-Amino- pseudocumeneg was produced by Fe-AcOH reduction of 3-nitro- pseudocumene which in turn was produced by the hypophosphorus acid reduction of the diazonium salt from 3-nitro-6-aminopseudocumene.’2

2’,4’,6’-TrimethyLDAB, Mesidine (60 g) was dissolved in a mixt of 113 ml of concd HC1 and 376 ml of H,O and diazotized at 0” using 30.6 g of NaNO, in 150 ml of H,O. One-half hr after the final addn, a s o h of 54 g of C,H,NMe,, 552 ml of 95% EtOH, 264 ml of H,O, and 109 g of NaOAc was added, and the soh was stirred for 24 hr. Extn with PhH and evapn of PhH left a semisolid material which was submitted to column chromatog over alumina in toluene-heptane soh. The first orange band was eluted, and the solvent was removed to give 19.9 g of the dye as bright orange needles which were recrystd from abs EtOH, mp 104.5-106” (see Table I).

Biological Properties. Young male rats of the Sprague-Dawley strain, approximately 8 weeks old and weighing 150-200 g, were dis- tributed as equally as possible in initial body weight into groups of 10 animals each. Each group was fed a diet, patterned after the “low protein, low riboflavin” diet of Miller‘ to which had been added one of the azo compounds at a level of 0.06%. The composition of the basal diet per kilogram was as follows: crude casein, 120 g; Cerelose, 770 g; Osborne and Mendel salt mixt, 40 g; corn oil, 50 g; Vitab (rice bran concentrate, from Charles Bowman Co.), 20 g; riboflavin, 0.5 mg; vitamin A palmitate, 67,500 IU.

received only the basal diet. All the rats were kept individually in screen-bottomed cages and were offered food and water ad libitum. Laparotomies were performed at the indicated times and micro-

A group received DAB at the 0.06% level while the control group

Table I. Trimethvl4-dimethvlaminoazobenzenes ~~

Yield“ Mp, “C recryst % Formulab - Compd

-

2’,4’,6’-TrimethyLDAB 104.5-1 06 17 C17H21N3

2’,3’,4’-TrimethyLDAB 147-151 26 C17H21N3

3‘,4’,5’-Trimethyl-DAB 131.5-133 57 C17H21N3

2’,3’,6’-Trimethyl-DAB 132.5-134 8 C17H21N3

2’,4’,5‘-Trimethyl-DAB 145-146.5 55 C,,H,IN,

2‘,3’,5‘-Trimethyl-DAB 101.5-103 25 C’,H21N,

“All samples recrystd from 95% EtOH after chromatog on alumina from toluene-heptane. bAll compds were analyzed for C. €1. N

scopic examinations were made whenever an animal died or at the end of the experiment.

Results

DAB gave tumor incidences of 6/ 10 at 4 months and 9/ 10 at 6 months, 3’,4’,5‘-DAB gave 9/10 at 4 months and 10/10 at 6 months. None of the other trimethyl-DAB homologs produced tumors at 9 months at which time the experiment was terminated. 3’,4’,5-Trimethyl-DAB appears to be about as hepatocarcinogenic as DAB.

L. Weiss and Dr. T. Yoneyama, Department of Pathology, University of Kentucky College of Medicine, for the micro- scopic evaluation of the tumors.

Acknowledgment. The authors are indebted to Dr. Daniel

References (1) J. A. Miller and E. C. Miller,Aduan. CancerRes., 1,339 (1953). (2) E. V. Brown and A. A. Hamdan, J. Nat. CancerInst., 27,663

(3) E. V. Brown, J. Med. Chem., 11, 1234 (1968). (4) J. A. Miller, E. C. Miller, and R. W. Sapp, Cancer Res., 11, 269

(5) A. Landenburg, Justus LiebigsAnn. Chem., 179, 163 (1875). (6) J. E. Purvis, J. Chem. Soc., 97,1546 (1910). (7) M. Dolinsky, J . H. Jones, C. D. Ritchie, R. L. Yates, and M. A

(8) C. H. Fisher and C. T. Walling, J . Amer. Chem. Soc., 57, 1700

(9) K. Sato, Y . Fujima, and A. Yamada, Bull. Chem. SOC. Jap., 41,

(1961).

(1951).

Hall, J. Ass. Offic. Agr. Chem., 42,709 (1959).

(1935).

442 (1968). (10) H. Hock and H. Kropf, Chem. Ber., 89, 2436 (1956). (11) F. M. Beringer and I. Ugelow, J. Amer. Chem. Soc., 75, 2635

(12) A. Huender, Reel. Trav. Chim. Pays-Bas., 34, 1 (1915). (1953).

Some Aliphatic Amines as Antipityrosporum Agents

Harold Bxon , Norman J. Van AbbB, Beecham Products (U. K.) , Brentford, England

and John K. Sugden* School of Pharmacy, City of Leicester Polytechnic, Leicester, England. Received May 26, I9 71

Pityrosporuni ovale and orbiculare, budding yeasts o f the family Cyptococcacae’ are not usually regarded as patho- gens. Van Abbe2 has discussed the possible relationship between dandruff and the presence of P. ovale on the scalp. Tinea versicolor was formerly attributed to the presence of Malassazia furfur,3 but more recent work has shown that M. furfur and P. orbiculare are probably the same organisms in different phases of g r ~ w t h . ~ It is claimed that P. ovale and P. orbiculare are only pathogenic in susceptible persons. In spite of this limited pathogenicity or perhaps limited virulence, these yeasts do seem to be implicated in the causa- tion of skin disorders and their suppression is clinically desirable. Much of the doubt as to the pathogenicity of Pityro- sporum species has arisen from the difficulties encountered in artificial culture, since isolation and maintainance in vitro of these lipophilic organisms are markedly influenced by variations in the constituents of the media. This sensitivity makes assessment of antipityrosporum agents difficult and earlier reports of measurements suspect. Developments in the cultural technique’*’ have much reduced the problems of assessment of this class of compound. Recent work in pyrrolidine chemistry has shown that ethyl AT-alkyl-4-hydroxy-


Aliphatic amines as antipityrosporum agentsHarold Dixon, Norman J. Van Abbe, and John K. Sugden








Hepatocarcinogenicity of the Trimethyl Homologs of 4-Dimethylaminazobenzene Ellis V. Brown* and Alice Kruegel Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506. Received August I I , 19 71

Definite to strong carcinogenic activity has been shown by some monosubstituted and disubstituted derivatives of 4-dimethylaminoazobenzene (DAB). 1-3 The only active tri- substituted-DAB tested has been the 2’,4’,6‘-trifluoro deriv- a t i ~ e . ~ It seemed of interest to synthesize and test for rat hepatocarcinogenic activity all of the trimethyl homologs of DAB with Me groups in the primed positions only. These are all new compounds and can be prepared by the diazotization of the proper trimethylanilines followed by coupling with PhNMez. The new azo compounds are listed in Table 1.

Experimental Section All melting points were detd on a Fisher-Johns apparatus and

are uncorrected. The C, H, N analyses were performed in this department on an F and M Model 185 analyzer by Mr. Daryl Sharp. Where analyses are indicated only by symbols of the elements analytical results obtained for those elements were within +0.4% of the theoretical values.

Trimethylmilines. Me~idine,~ bp 224-228”, was prepd from nitromesitylene6 by reduction with Sn-HCl. 5-Aminopseudocumene, mp 62-63’, was prepd from 5-nitropseud~cumene~ in the same way. 4-Amin0hemimellitene,~ mp 29-27”, was obtd from 5-nitrohemi- mellitene,’ 6-aminopse~documene~ was prepd from 6-nitropseudo- cumene,” and 5-aminohemimellitene,” mp 75-78’, was prepared from 5-nit~ohemimellitene~ by reduction (Fe-AcOH). 3-Amino- pseudocumeneg was produced by Fe-AcOH reduction of 3-nitro- pseudocumene which in turn was produced by the hypophosphorus acid reduction of the diazonium salt from 3-nitro-6-aminopseudocumene.’2

2’,4’,6’-TrimethyLDAB, Mesidine (60 g) was dissolved in a mixt of 113 ml of concd HC1 and 376 ml of H,O and diazotized at 0” using 30.6 g of NaNO, in 150 ml of H,O. One-half hr after the final addn, a s o h of 54 g of C,H,NMe,, 552 ml of 95% EtOH, 264 ml of H,O, and 109 g of NaOAc was added, and the soh was stirred for 24 hr. Extn with PhH and evapn of PhH left a semisolid material which was submitted to column chromatog over alumina in toluene-heptane soh. The first orange band was eluted, and the solvent was removed to give 19.9 g of the dye as bright orange needles which were recrystd from abs EtOH, mp 104.5-106” (see Table I).

Biological Properties. Young male rats of the Sprague-Dawley strain, approximately 8 weeks old and weighing 150-200 g, were dis- tributed as equally as possible in initial body weight into groups of 10 animals each. Each group was fed a diet, patterned after the “low protein, low riboflavin” diet of Miller‘ to which had been added one of the azo compounds at a level of 0.06%. The composition of the basal diet per kilogram was as follows: crude casein, 120 g; Cerelose, 770 g; Osborne and Mendel salt mixt, 40 g; corn oil, 50 g; Vitab (rice bran concentrate, from Charles Bowman Co.), 20 g; riboflavin, 0.5 mg; vitamin A palmitate, 67,500 IU.

received only the basal diet. All the rats were kept individually in screen-bottomed cages and were offered food and water ad libitum. Laparotomies were performed at the indicated times and micro-

A group received DAB at the 0.06% level while the control group

Table I. Trimethvl4-dimethvlaminoazobenzenes ~~

Yield“ Mp, “C recryst % Formulab - Compd

-

2’,4’,6’-TrimethyLDAB 104.5-1 06 17 C17H21N3

2’,3’,4’-TrimethyLDAB 147-151 26 C17H21N3

3‘,4’,5’-Trimethyl-DAB 131.5-133 57 C17H21N3

2’,3’,6’-Trimethyl-DAB 132.5-134 8 C17H21N3

2’,4’,5‘-Trimethyl-DAB 145-146.5 55 C,,H,IN,

2‘,3’,5‘-Trimethyl-DAB 101.5-103 25 C’,H21N,

“All samples recrystd from 95% EtOH after chromatog on alumina from toluene-heptane. bAll compds were analyzed for C. €1. N

scopic examinations were made whenever an animal died or at the end of the experiment.

Results

DAB gave tumor incidences of 6/ 10 at 4 months and 9/ 10 at 6 months, 3’,4’,5‘-DAB gave 9/10 at 4 months and 10/10 at 6 months. None of the other trimethyl-DAB homologs produced tumors at 9 months at which time the experiment was terminated. 3’,4’,5-Trimethyl-DAB appears to be about as hepatocarcinogenic as DAB.

L. Weiss and Dr. T. Yoneyama, Department of Pathology, University of Kentucky College of Medicine, for the micro- scopic evaluation of the tumors.

Acknowledgment. The authors are indebted to Dr. Daniel

References (1) J. A. Miller and E. C. Miller,Aduan. CancerRes., 1,339 (1953). (2) E. V. Brown and A. A. Hamdan, J. Nat. CancerInst., 27,663

(3) E. V. Brown, J. Med. Chem., 11, 1234 (1968). (4) J. A. Miller, E. C. Miller, and R. W. Sapp, Cancer Res., 11, 269

(5) A. Landenburg, Justus LiebigsAnn. Chem., 179, 163 (1875). (6) J. E. Purvis, J. Chem. Soc., 97,1546 (1910). (7) M. Dolinsky, J . H. Jones, C. D. Ritchie, R. L. Yates, and M. A

(8) C. H. Fisher and C. T. Walling, J . Amer. Chem. Soc., 57, 1700

(9) K. Sato, Y . Fujima, and A. Yamada, Bull. Chem. SOC. Jap., 41,

(1961).

(1951).

Hall, J. Ass. Offic. Agr. Chem., 42,709 (1959).

(1935).

442 (1968). (10) H. Hock and H. Kropf, Chem. Ber., 89, 2436 (1956). (11) F. M. Beringer and I. Ugelow, J. Amer. Chem. Soc., 75, 2635

(12) A. Huender, Reel. Trav. Chim. Pays-Bas., 34, 1 (1915). (1953).

Some Aliphatic Amines as Antipityrosporum Agents

Harold Bxon, Norman J. Van AbbB, Beecham Products (U. K.) , Brentford, England

and John K. Sugden* School of Pharmacy, City of Leicester Polytechnic, Leicester, England. Received May 26, I9 71

Pityrosporuni ovale and orbiculare, budding yeasts o f the family Cyptococcacae’ are not usually regarded as patho- gens. Van Abbe2 has discussed the possible relationship between dandruff and the presence of P. ovale on the scalp. Tinea versicolor was formerly attributed to the presence of Malassazia furfur,3 but more recent work has shown that M. furfur and P. orbiculare are probably the same organisms in different phases of g r ~ w t h . ~ It is claimed that P. ovale and P. orbiculare are only pathogenic in susceptible persons. In spite of this limited pathogenicity or perhaps limited virulence, these yeasts do seem to be implicated in the causa- tion of skin disorders and their suppression is clinically desirable. Much of the doubt as to the pathogenicity of Pityro- sporum species has arisen from the difficulties encountered in artificial culture, since isolation and maintainance in vitro of these lipophilic organisms are markedly influenced by variations in the constituents of the media. This sensitivity makes assessment of antipityrosporum agents difficult and earlier reports of measurements suspect. Developments in the cultural technique’*’ have much reduced the problems of assessment of this class of compound. Recent work in pyrrolidine chemistry has shown that ethyl AT-alkyl-4-hydroxy-

Notes JoumalofMedicinal Chemistry, 1972, Vol. IS, No. 2 213

Table I

No. R X

P. orbiculare

8 14

P. ovule strains strains Mp or bp Yield,

Y (mm), "C Formulaa Form % 1 2 3

Zinc pyriihGne as a control

134-136

179-180 78-80

(0.8)

38-40 64-65

171-172 3-5

200 (11) 135-136 182-184 180 (0.5) 176 (0.8) 220 (0.8) 231-232 245 -24 7

Base 48

HC1 Salt 56 Base 26 Base 81 Acetyl deriv of 4 55 HBr salt 50 Acetyl deriv of 6 50 Acetyl deriv of 1 55

HBr salt 16 HBr salt 30 Base 45

Base 64

Base 63

HCl salt 25 HBr salt 80

>loo0 >loo0 >lo00 128 64

125 125 125 125 125 500 500 500 >lo00 >lo00 500 500 500 500 500

>loo0 >loo0 >loo0 >loo0 >loo0 31 31 31 125 125

>loo0 >loo0 >loo0 >loo0 >loo0 500 500 500 >lo00 >lo00

125 125 125 256 256b 62.5 62.5 62.5>1000 >lo00

500 500 500 >lo00 >lo00

31 31 31 64 64

31 31 31 128 128

31 31 31 500 1000 62.5 125 62.5>1000 >loo0 16 16 16 8 8

aAll compds analyzed for C, H, N. bPartial inhibition at 128 pg/ml.

5-oxo-3-pyrroline-3-carboxylates have moderate activity against P. ovule and P. orbicul~re.~ The object of the present work is to investigate the antipityrosporum activity of smaller molecules than those described before' and to indicate some structure-activity relationship with regard to P. ovale and P. orbiculare.

activity against P. ovale and P. orbiculare. P. orbiculare is a much less vigorous organism in artificial culture and more exacting in its growth requirements than is P. ovale. The selectivity of 1, 10, and 14 does tend to justify the classifica- tion of these yeasts as distinct species. In the case of P. ovule there is some correlation between the tests on malt extract Tween agar medium (without added fatty acids or bile salts) and the suspension assay technique.

Optimal activity against P. ovale of amines 1 on malt extract Tween agar medium was observed in those compounds where R was Clo or more. The nature of the functional group, X, was significant, X = C02Et showing maximal inhibition, X = C02Me being slightly less active, with X = CONH, much less active, and X = CN having minimal activity of the groups tested.

The presence of a secondary amino group was found to be essential for the activity of amines 1 in both assays. The nature of the side chain Y had considerable influence on the activity of the compounds against P. ovule in the malt extract Tween agar medium assay (1 1 and 9 being less active than 6), but possibly less effect on activity when tested by the suspension assay.

that the length of the aliphatic chain R is critical for inhibitory action. The secondary amino group appears to be essential, since acetylation abolishes activity (7 and 8). Apparently optimal activity is found when the side chain, Y, is (CH,),, 6 being more active than 9 and much more active than 1 1. The nature of the functional group, X, in- fluences the activity of amines 1 against P. orbiculare, X = C02Et being the most active and X = CONH, being the least

Examination of Table I shows that amines (1) have modest

With regard to P. orbiculare (Table I) there is no indication

active, with X = CN having an intermediate level of activity amongst the groups tested, in contrast with the results of tests on P. ovule. The latter observation adds some weight to these yeasts being regarded as separate species.


as expected. Melting points were taken on a Biichi apparatus and are uncorrected.

Alkyl N-Substituted-3-aminopropionates (1). Primary amines (0.1 mole) and ethyl or methyl acrylate, or ethyl methacrylate (0.1 mole) were heated under reflux in abs EtOH (100 ml) for 1 hr, and the product was isolated by the method described by Sugden6 and characterized as the base or converted into an appropriate salt. Acetyl derivatives were prepd by conventional methods and recrystd from petr ether, bp 60-80" (see Table I).

N-Hexadecyl-3-aminopropionitrile was prepd by the method of Caldo' from hexadecylamine (24.2 g, 0.1 mole) and acrylonitrile (5.3 g, 0.1 mole) and recrystd from petr ether: bp 40-60"; mp 39- 40"; yield, 12.8 g (81%).

N-hexadecyl-3-aminopropionitrile (12.1 g, 0.08 mole) in Ac,O (20 ml) by a conventional method: mp 64-65"; yield, 13.5 g (50%). Anal. (C,,H4,N,0) C, H, N.

N-Hexadecyl-3-aminopropionamide was prepd from hexadecylamine (12.1 g, 0.05 mole) and acrylamide (3.5 g, 0.05 mole) and catalyzed by Triton B s o h (0.05 ml) in the usual manner: mp 78- 80"; yield, 4.0 g (26%). Anal. (CI9H4,N,O) C, H, N.

Ethyl N-Dodecylaminoacetate. n-Dodecylamine (1 8.0 g, 0.1 mole), and ethyl chloroacetate (12.2 g, 0.1 mole), and anhyd Na,C03 (10.6 g, 0.1 mole) were heated in abs EtOH (100 ml) for 18 hr, and the product was isolated in theusual way: bp 180-190" (0.5-0.7 mm); yield, 6.0 g (22%).Anal. (CI6H,,NO,) C, H, N.

Microbiological Testing. Minimum Inhibitory Concentration. The technique for detg the MIC involved making serial dilns of the compd in agar and surface inoculating the growth medium with test organisms. The growth medium was Dixon's formula' without modification for tests against P. orbiculure but for studies with P. ovule, the ox bile and glyceryl monooleate were omitted since these appeared to inactivate the test materials and were not essential for the growth of P. ovule, except under condns of primary isolation; they are, however, needed for P. orbiculare.

After dispersing the compd in 2% Tween 40, two-fold dilns were prepd in measured amt of liquefied culture medium.' The final concn of compd tested in agar was within the range 1000-8

Uv, ir, and nmr spectra were measured for all compds and were

Acetamido-N-hexadecylamino-3-propionitrile was prepd from

214 Journal of Medicinal Chemistry, 1972, Vol. 15, No. 2 New Compounds

pg/ml. The medium was poured into petri dishes and left to harden ovemight at room temp. The surface of the agar was then inoculated with the test organisms (0.02 ml of standard suspension). The inoculated plates together with the appropriate organism controls were incubated for 3 days at 37" in the case of P. ovule and up to 5 days in the case of P. orbiculare. MIC's were detd by .observing the lowest concn which inhibited growth under the prescribed condns.

Suspension Technique. Each test compd (0.1 g) was dissolved or suspended in Tween 40 (2 ml) and the vol made up to 100 ml with sterile dist H,O. A sample was inoculated with P. ovule (0.1 ml of standard suspension contg lo6 organisms/ml) and stored at room temp for 1 hr. The test samples and appropriate controls were plated out on petri dishes of Dixon's medium' and incubated at 37"

for 3 days. Activity of the compds was assessed on the growth observed.

References (1) J. Lodder, "The Yeasts; A taxonomic study," North Holland

Publishing Co., Amsterdam, 1370, pp. 1166-1186. (2) N. J. Van Abbb, J. SOC. Comet . Chem., 15,609 (1964). (3) T. H. Sternberg and F. M. Keddie, Arch. Dermafol., 84,999

(1961). (4) R. C. Burke, J. Invest. Dermafol., 36, 389 (196 1). (5) J. E. Hogan, Ph.D. Thesis, University of London, 1970. (6) J. K. Sugden, Ph.D. Thesis, University of London, 1964. (7) C. Caldo, Chim. Ind. (Milan), 44, 753 (1962).

New Compounds

A Rapid, Convenient Preparative Procedure for Phenethylamines Edgar F. Kiefer Department of Chemistry, University of Hawaii, Honolulu, Hawaii 96822. Received August 6, 19 71

In view of the very broad pharmacological utility of substituted 2-phenylethylamines, we wish to contribute a synthetic procedure which, because of its versatility and con- venience, may find considerable use. Although based entirely on standard synthetic methods, the overall scheme is specifically tailored to the properties of the benzylic intermediates involved, and eliminates the need for isolation of intermediates and other time-consuming operations. The procedure is described for the pmethoxy derivative; it is also applicable without substantive modification to other ring alkoxy-, alkyl-, and halogen-substituted phenethylamines.


(100 g, 0.725 mole) was shaken with 500 ml of concd HCl for 2 min. The org phase was washed with %0,5% NaHCO,, and H,O, then added over 40 min to a stirred slurry of 49 g (1.0 mole) of NaCN in 400 ml of DMSO,' with icawater cooling to maintain the temp at 35-40". After addn was complete, the cooling bath was removed, the mixt was stirred for 90 min and then added to 300 ml of H,O, and the small upper phase sepd. The aq DMSO layer was extd with two 100-ml portions of Et,O, which were combined with the product layer, and the whole was washed once with H,O and dried (MgSO,).

A dry flask was charged with ca. 600 ml of abs Et,O and chilled in ice as 80 g (0.6 mole) of anhyd AlC1, was added portionwise, followed by 23 g (0.6 mole) of LAH.'? The dried Et,O soln of crude p-methoxyphenylacetonitrile was added at such a rate as to maintain gentle reflux without extemal heat (cu. 1 hr). The mixt was stirred for 2 hr, then chilled in ice, and treated dropwise with 25 ml of H,O followed by 250 ml of 20% of aq NaOH, with periodic addn of Et,O through the condenser to replenish losses and facilitate stirring. The resulting voluminous, granular ppt of NaCl and LiCl and alumi- nate was removed by filtration, washed well with Et,O, and dis- carded. The filtrate was mixed with one-third its vol of abs EtOH and 60 ml of concd HCl was added slowly with continuous swirling and ice cooling. After chilling to Oo, the cryst amine hydrochloride was collected, 101 g, mp 212-214", identified by mass spectroscopy [ m / e 122, 30, 121, 28, 151 (M+)]. The overall yield was 75%

4-Methoxyphenylethylamine Hydrochloride. p-Anisyl alcohol

tLAH alone and other metal hydride reagents are unsatisfactory for the reduction of benzylic nitriles to amines.

from anisyl alcohol. The hydrochloride may be recrystd from Et,O- EtOH or i-PrOH.

N-Methyl-p-methoxyphenylethylamine Hydrochloride. p-Me th- oxyphenethylamine, generated from 100 g (0.536 mole) of the hydrochloride by stirring with concd aq NaOH, was treated with 100 ml of PhH and 70 g (0.66 mole) of PhCHO. A mildly exothermic reaction began at once. The mixt was heated under reflux until no more H,O was present in the condensate (ea. 1 hr), then, without cooling, an attached Dean-Stark trap was removed and a soln of 82 g (0.65 mole) of Me,SO,' in 200 ml of PhH was added through the condenser at such a rate as to maintain reflux (15 min). The 2-phase mixt was heated for 90 min on the steam bath, cooled slightly, treated with 200 ml of H,O, and heated for an add1 20 min. After cooling in ice, the aq layer was washed twice with Et,O to remove unreacted PhCHO and made strongly basic with 50% aq NaOH. Two Et,O exts of the basic aq phase were added to the amine layer which sepd, and the resulting soln was evacd at the aspirator for 30 min, leaving 90 g (102%) of crude N-methyl-p-methoxyphenethyl- amine. This material was dissolved in 500 ml of 20% abs EtOH-Et,O and treated with 50 ml of concd HCl with swirling and cooling to yield the white, cryst hydrochloride, which was washed thoroughly with ice-cold 20% EtOH-Et,O and dried, mp 185.5-186.5', identified by mass spectroscopy [ m / e 121, 44, 165 (M+)]. The yield was 83 g (77%).

References (1) R. A. Smiley and C. Arnold, J. Org. Chem., 25, 257 (1960). (2) R. F. Nystrom,J. Amer. Chem. Soc., 77,2544 (1955). (3) J. J. Lucier, A. D. Harris, and P. S . Korosec, Org. Syn., 44,72

(1964).

Synthesis of 2-Fluoro-9-0-D -rib0 furanos ylpurine ( 2-Fluoronebularine) Masajiro Kawana, Robert J. Rousseau,* and Roland K. Robins ICN Nucleic Acid Research Institute, Irvine, California 92664. Received August 16, I9 71

The antibiotic nebularine (9-p-D -ribofuranosylpurine) has shown tuberculostatic,' antimitotic,2 and anticancer activ- it^.^,^ The mode of action has been proposed to be in the purine biosynthetic pathway!,' It has limited usefulness because of its high t o ~ i c i t y . ~ ~ ~ , ' '

We wish to report the synthesis of 2-fluoronebularine (2a). Synthesis of the title compound 2a was accomplished by removal of the benzylthio group from 6-benzylthio-2-fluoronebularine ( 1)8 with Raney Ni.


Rapid, convenient preparative procedure for phenethylaminesEdgar F. Kiefer














New Compounds









N-Methyl-p-methoxyphenylethylamine Hydrochloride. p-Me th- oxyphenethylamine, generated from 100 g (0.536 mole) of the hydrochloride by stirring with concd aq NaOH, was treated with 100 ml of PhH and 70 g (0.66 mole) of PhCHO. A mildly exothermic reaction began at once. The mixt was heated under reflux until no more H,O was present in the condensate (ea. 1 hr), then, without cooling, an attached Dean-Stark trap was removed and a soln of 82 g (0.65 mole) of Me,SO,' in 200 ml of PhH was added through the condenser at such a rate as to maintain reflux (15 min). The 2-phase mixt was heated for 90 min on the steam bath, cooled slightly, treated with 200 ml of H,O, and heated for an add1 20 min. After cooling in ice, the aq layer was washed twice with Et,O to remove unreacted PhCHO and made strongly basic with 50% aq NaOH. Two Et,O exts of the basic aq phase were added to the amine layer which sepd, and the resulting soln was evacd at the aspirator for 30 min, leaving 90 g (102%) of crude N-methyl-p-methoxyphenethyl- amine. This material was dissolved in 500 ml of 20% abs EtOH-Et,O and treated with 50 ml of concd HCl with swirling and cooling to yield the white, cryst hydrochloride, which was washed thoroughly with ice-cold 20% EtOH-Et,O and dried, mp 185.5-186.5', identified by mass spectroscopy [m/e 121, 44, 165 (M+)]. The yield was 83 g (77%).


(1964).





Synthesis of 2-fluoro-9-.beta.-D-ribofuranosylpurine (2-fluoronebularine)Masajiro Kawana, Robert J. Rousseau, and Roland K. Robins














New Compounds









N-Methyl-p-methoxyphenylethylamine Hydrochloride. p-Me th- oxyphenethylamine, generated from 100 g (0.536 mole) of the hydrochloride by stirring with concd aq NaOH, was treated with 100 ml of PhH and 70 g (0.66 mole) of PhCHO. A mildly exothermic reaction began at once. The mixt was heated under reflux until no more H,O was present in the condensate (ea. 1 hr), then, without cooling, an attached Dean-Stark trap was removed and a soln of 82 g (0.65 mole) of Me,SO,' in 200 ml of PhH was added through the condenser at such a rate as to maintain reflux (15 min). The 2-phase mixt was heated for 90 min on the steam bath, cooled slightly, treated with 200 ml of H,O, and heated for an add1 20 min. After cooling in ice, the aq layer was washed twice with Et,O to remove unreacted PhCHO and made strongly basic with 50% aq NaOH. Two Et,O exts of the basic aq phase were added to the amine layer which sepd, and the resulting soln was evacd at the aspirator for 30 min, leaving 90 g (102%) of crude N-methyl-p-methoxyphenethyl- amine. This material was dissolved in 500 ml of 20% abs EtOH-Et,O and treated with 50 ml of concd HCl with swirling and cooling to yield the white, cryst hydrochloride, which was washed thoroughly with ice-cold 20% EtOH-Et,O and dried, mp 185.5-186.5', identified by mass spectroscopy [ m / e 121, 44, 165 (M+)]. The yield was 83 g (77%).


(1964).




New Compounds Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 215

SCH,C,H, I

HO OH 1

RO OR 2

a , R = H b, R = COCH,

(3) G. B. Brown and V. S. Weliky, J. Biol. Chem., 204,1019 (1953). (4) M. P. Gordon, D. I. Magrath, and G. B. Brown, J. Amer. Chem.

Soc., 79,3256 (1957). (5) R. J. Winzler, W. Wells, J. Shapiro, A. D. Williams, I. Bomstein,

M. J. Burr, and W. R. Best, CancerRes., 19,377 (1959). (6) A. P. Truant and H. E. D'Amato, Fed. Proc., Fed. Amer. Soc.

Exp. Biol., 14,391 (1955). (7) M. P. Gordon and G. B. Brown. J . Biol. Chem., 220.927 (1956). (8) J. F. Gerster and R. K. Robiqs,'J. Amer. Chem. Soc:, 87,3752

(9) W. W. Zorbach and R. S. Tipson, "Synthetic Procedures in (1965).

Nucleic Acid Chemistry," Interscience Publishers, New York, N. Y., 1968, p 244.

Experimental Section? 2-Fluoro-9-(p-D-ribofuranosyl)purine (2a). To a boiling soln of

1.64 g (4 mmoles) of 1 H,O in 30 ml of EtOH was added a suspension of 14 g of Raney Ni in 60 ml of EtOH. The mixt was refluxed for 40 min with stirring. The catalyst was removed by filtration with a Celite pad and washed thoroughly with boiling EtOH. The combined filtrate and washings were concd to ea. 10 ml in vacuo. The undissolved material was removed by fitration with charcoal and the filtrate was evapd to dryness in vacuo to give 590 mg (55%) of crude 2a as a foam. The product was chromatogd on a silica gel column with EtOAc-EtOH (955, v/v). Evapn of the solvents gave a colorless foam, which was crystd from EtOAc contg a small amt of MeOH: mp 144.5-146O; [a ]"D -32.3' (c 1, H 0);uv Aky: 263 nm (E 8000), hkyil 264 (7500), h z F H 264 (80bO); nmr (DMSO-d,- D,O) 6 6.00 (d, J,,., 5.4 Hz, 1 H, HI,), 8.83 ( s , 1 H, HR), 9.06 (4 JH F = 1.2 Hz; 'i H, H,); nmr for F (DMSO-dI,) -24.8 ppm. Anal. Calcd for C,,H~,FN,O, : C, H, N.

9-(2,3,5-Tr1-O-acetyl-~-D-ribofuranosyl)-2-fluoropur~e (2b). To a stirred s o h of 50 ml of 48-50% HBF, was added 3.9 g (0.01 mole) of 9-(2,3,5-tri-O-acetyl-p-D-ribofuranosyl)-2-aminopurine9 at -20 to -25". To this mixt was added a soln of 2.1 g (0.03 mole) of NaNO, in 4 ml of H,O over a period of 10 min. The reaction mixt was stirred at the same temp for another 15 min and 50 ml of EtOH (precooled below -20') was then added. The mixt was neutralized with ea. 26 ml of concd NH,OH to pH 6 below -15'. The resulting ppt was removed by filtration and washed with 50 ml of cold EtOH. The combined filtrate and washings were concd to ea. 50 ml at 30-35" in vacuo, and the soln was extd with two 150-ml portions of CH,Cl,. The exts were washed with, successively, 50 ml of H,O, 50 ml of 1% NaHCO,, and two 50-ml portions of H,O, and then dried (MgSO,). Evapn of the solvents gave 2.9 g (73%) of crude 2b as a gummy material. This product was chromatogd on a silica gel column (130 g, 4 X 30 cm) using EtOAc-heptane (7:3, v/v), and 250-ml fractions were collected. Fractions 26-35 contd 960 mg (24%) of 2b, which was contaminated with a trace of impurity. Fractions 36-67 were combined and evapn of the solvents gave 1.07 g (27%) of analytically pure 2b as a lass. [a]"D -3.5" (c 2.88, CHCl,); uv 264 nm (E 7300), k&il 265 (72001, nmr (DMSO-d,) 6 6.33 (d, J I t , , * = 4.8 Hz, 1 H, HI,), 8.83 (s, 1 H, He), 9.13 (d, JH F = 1.2 Hz, 1 H, H,); nmr for F (DMSO-d,) -27.7 ppm. Anal. Cafid for C,,H,,FN,O,: C, H, N.

NH, at 4' occurred with concommitant displacement of the F at and formation of 9-p-D-ribofuranosyl-2-aminopurine.

Attempts to remove the Ac blocking groups of 2b with EtOH-

References (1) N. Lofgren and B. Luning, Acta Chem. Scand., 7,225 (1953). (2) J. J . Biesele, M. C. Slautterback, and M. Margolis, Cancer (Phila-

delphia), 8, 87 (1955).

4-[ (Aminooxy )methyl] thiazole Dihydrochloride Glenn H. Hamor Department of Biomedicinal Chemistry, School of Pharmacy, University of Southern California, Los Angeles, California 9000 7. ReceivedAugust 27, 1971

Recent reports showing the title compound (I) to possess potent in vitro and in vivo inhibition of specific histidine decarboxylase' and to markedly lower rat brain histamine,2y3 prompt me to report its synthesis. By literature methodsF6 alkylation of N-hydroxyphthalimide with 4-chloromethyl- thiazole, followed by hydrazinolysis, gave I.


(0.214 mole) ofN-hydroxyphthalimide in 250 ml of MeCN and 43.3 g (0.428 mole) of Et,N was added 36.3 g (0.214 mole) of 4-chloro- methylthiazole hydrochloride' and refluxed for 6 hr. After cooling, the cryst ppt was filtered, washed with a little MeCN and thoroughly with H,O, and dried, to yield 24 g (43%) of cryst product, mp 158- 159"(EtOH). Anal. (C,,H,N,O,S) C, H, N.

0.05 mole) was refluxed for 2 hr with 2.5 g (0.05 mole) of hydrazine hydrate, 99-loo%, in 150 ml of anhyd EtOH. After cooling and removal by filtration of pptd phthalhydrazide, ethanolic HC1 was added to the filtrate, and the cryst solid was filtered off and dried. Recrystn from EtOH-Et,O gave 8.5 g (83%) of white solid, mp 176- 177' dec. Anal. (C,H,N,OS. 2HC1) C, H, N.

N-(4-Thiazolylmethoxy)phthalimide (11). To a s o h of 34.9 g

4-[(Aminooxy)methyl]thiazole Dihydrochloride (I). I1 (13 g,

References D. Aures, G. H. Hamor, W. G. Clark, and S. S. Laws, 4th Inter- national Pharmacology Congress, Basel, 1969, p 179. M. K. Menon, D. Aures, and W. G. Clark, Pharmacologist, 12, 205 (1970). K. M. Taylor and S. H. Snyder, Science, 172,1037 (1971). A. F. McKay, D. L. Garmaise, G. Y. Paris, and S. Gelblum, Can. J. Chem., 38,343 (1960). D . G. Martin, E. L. Schuman, W. Veldkamp, and H. Keasling, J. Med. Chem., 8,455 (1965). D. J. Drain, J. G. Howes, and H. W. R. Williams, British Patent 984,305 (1965); Chem. Abstr., 62, 14572 (1965). W. T. Caldwell and S. M. Fox, J. Amer. Chem. Soc., 73,2935 (1951).

?Melting points were taken on a Thomas-Hoover capillary melting point apparatus and are uncorr. Microanalyses were performed by Heterocyclic Chemical Corp., Harrisonville, Mo. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within *0.4% of the theoretical values. The F nmr spectra were run with 1% CF,COOH as an external standard.

?Melting points were detd with a Fisher-Johns app and are uncorr. Ir spectra (Nujol mull) were measured on a Perkin-Elmer infracord 137 spectrometer. Absorption bands were as expected. Elemental anal. were performed by Elek Microanalytical Laboratories, Tor- rance, Calif. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within +0.40/0 of the theoretical values.


4-[(Aminoxy)methyl]thiazole dihydrochlorideGlenn H. Hamor







New Compounds Journal ofMedicina1 Chemistry, 1972, Vol. 15, No. 2 215

SCH,C,H, I

HO OH 1

RO OR 2

a , R = H b, R = COCH,

(3) G. B. Brown and V. S. Weliky, J. Biol. Chem., 204,1019 (1953). (4) M. P. Gordon, D. I. Magrath, and G. B. Brown, J. Amer. Chem.

Soc., 79,3256 (1957). (5) R. J. Winzler, W. Wells, J. Shapiro, A. D. Williams, I. Bomstein,

M. J. Burr, and W. R. Best, CancerRes., 19,377 (1959). (6) A. P. Truant and H. E. D'Amato, Fed. Proc., Fed. Amer. Soc.

Exp. Biol., 14,391 (1955). (7) M. P. Gordon and G. B. Brown. J . Biol. Chem., 220.927 (1956). (8) J. F. Gerster and R. K. Robiqs,'J. Amer. Chem. Soc:, 87,3752

(9) W. W. Zorbach and R. S. Tipson, "Synthetic Procedures in (1965).

Nucleic Acid Chemistry," Interscience Publishers, New York, N. Y., 1968, p 244.

Experimental Section? 2-Fluoro-9-(p-D-ribofuranosyl)purine (2a). To a boiling soln of

1.64 g (4 mmoles) of 1 H,O in 30 ml of EtOH was added a suspension of 14 g of Raney Ni in 60 ml of EtOH. The mixt was refluxed for 40 min with stirring. The catalyst was removed by filtration with a Celite pad and washed thoroughly with boiling EtOH. The combined filtrate and washings were concd to ea. 10 ml in vacuo. The undissolved material was removed by fitration with charcoal and the filtrate was evapd to dryness in vacuo to give 590 mg (55%) of crude 2a as a foam. The product was chromatogd on a silica gel column with EtOAc-EtOH (955, v/v). Evapn of the solvents gave a colorless foam, which was crystd from EtOAc contg a small amt of MeOH: mp 144.5-146O; [a ]"D -32.3' (c 1, H 0);uv Aky: 263 nm (E 8000), hkyil 264 (7500), h z F H 264 (80bO); nmr (DMSO-d,- D,O) 6 6.00 (d, J,,., 5.4 Hz, 1 H, HI,), 8.83 ( s , 1 H, HR), 9.06 (4 JH F = 1.2 Hz; 'i H, H,); nmr for F (DMSO-dI,) -24.8 ppm. Anal. Calcd for C,,H~,FN,O, : C, H, N.

9-(2,3,5-Tr1-O-acetyl-~-D-ribofuranosyl)-2-fluoropur~e (2b). To a stirred s o h of 50 ml of 48-50% HBF, was added 3.9 g (0.01 mole) of 9-(2,3,5-tri-O-acetyl-p-D-ribofuranosyl)-2-aminopurine9 at -20 to -25". To this mixt was added a soln of 2.1 g (0.03 mole) of NaNO, in 4 ml of H,O over a period of 10 min. The reaction mixt was stirred at the same temp for another 15 min and 50 ml of EtOH (precooled below -20') was then added. The mixt was neutralized with ea. 26 ml of concd NH,OH to pH 6 below -15'. The resulting ppt was removed by filtration and washed with 50 ml of cold EtOH. The combined filtrate and washings were concd to ea. 50 ml at 30-35" in vacuo, and the soln was extd with two 150-ml portions of CH,Cl,. The exts were washed with, successively, 50 ml of H,O, 50 ml of 1% NaHCO,, and two 50-ml portions of H,O, and then dried (MgSO,). Evapn of the solvents gave 2.9 g (73%) of crude 2b as a gummy material. This product was chromatogd on a silica gel column (130 g, 4 X 30 cm) using EtOAc-heptane (7:3, v/v), and 250-ml fractions were collected. Fractions 26-35 contd 960 mg (24%) of 2b, which was contaminated with a trace of impurity. Fractions 36-67 were combined and evapn of the solvents gave 1.07 g (27%) of analytically pure 2b as a lass. [a]"D -3.5" (c 2.88, CHCl,); uv 264 nm (E 7300), k&il 265 (72001, nmr (DMSO-d,) 6 6.33 (d, J I t , , * = 4.8 Hz, 1 H, HI,), 8.83 (s, 1 H, He), 9.13 (d, JH F = 1.2 Hz, 1 H, H,); nmr for F (DMSO-d,) -27.7 ppm. Anal. Cafid for C,,H,,FN,O,: C, H, N.

NH, at 4' occurred with concommitant displacement of the F at and formation of 9-p-D-ribofuranosyl-2-aminopurine.

Attempts to remove the Ac blocking groups of 2b with EtOH-

References (1) N. Lofgren and B. Luning, Acta Chem. Scand., 7,225 (1953). (2) J. J . Biesele, M. C. Slautterback, and M. Margolis, Cancer (Phila-

delphia), 8, 87 (1955).

4-[ (Aminooxy )methyl] thiazole Dihydrochloride Glenn H. Hamor Department of Biomedicinal Chemistry, School of Pharmacy, University of Southern California, Los Angeles, California 9000 7. ReceivedAugust 27, 1971

Recent reports showing the title compound (I) to possess potent in vitro and in vivo inhibition of specific histidine decarboxylase' and to markedly lower rat brain histamine,2y3 prompt me to report its synthesis. By literature methodsF6 alkylation of N-hydroxyphthalimide with 4-chloromethyl- thiazole, followed by hydrazinolysis, gave I.


(0.214 mole) ofN-hydroxyphthalimide in 250 ml of MeCN and 43.3 g (0.428 mole) of Et,N was added 36.3 g (0.214 mole) of 4-chloro- methylthiazole hydrochloride' and refluxed for 6 hr. After cooling, the cryst ppt was filtered, washed with a little MeCN and thoroughly with H,O, and dried, to yield 24 g (43%) of cryst product, mp 158- 159"(EtOH). Anal. (C,,H,N,O,S) C, H, N.

0.05 mole) was refluxed for 2 hr with 2.5 g (0.05 mole) of hydrazine hydrate, 99-loo%, in 150 ml of anhyd EtOH. After cooling and removal by filtration of pptd phthalhydrazide, ethanolic HC1 was added to the filtrate, and the cryst solid was filtered off and dried. Recrystn from EtOH-Et,O gave 8.5 g (83%) of white solid, mp 176- 177' dec. Anal. (C,H,N,OS. 2HC1) C, H, N.

N-(4-Thiazolylmethoxy)phthalimide (11). To a s o h of 34.9 g

4-[(Aminooxy)methyl]thiazole Dihydrochloride (I). I1 (13 g,

References D. Aures, G. H. Hamor, W. G. Clark, and S. S. Laws, 4th Inter- national Pharmacology Congress, Basel, 1969, p 179. M. K. Menon, D. Aures, and W. G. Clark, Pharmacologist, 12, 205 (1970). K. M. Taylor and S. H. Snyder, Science, 172,1037 (1971). A. F. McKay, D. L. Garmaise, G. Y. Paris, and S. Gelblum, Can. J. Chem., 38,343 (1960). D . G. Martin, E. L. Schuman, W. Veldkamp, and H. Keasling, J. Med. Chem., 8,455 (1965). D. J. Drain, J. G. Howes, and H. W. R. Williams, British Patent 984,305 (1965); Chem. Abstr., 62, 14572 (1965). W. T. Caldwell and S. M. Fox, J. Amer. Chem. Soc., 73,2935 (1951).

?Melting points were taken on a Thomas-Hoover capillary melting point apparatus and are uncorr. Microanalyses were performed by Heterocyclic Chemical Corp., Harrisonville, Mo. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within *0.4% of the theoretical values. The F nmr spectra were run with 1% CF,COOH as an external standard.

?Melting points were detd with a Fisher-Johns app and are uncorr. Ir spectra (Nujol mull) were measured on a Perkin-Elmer infracord 137 spectrometer. Absorption bands were as expected. Elemental anal. were performed by Elek Microanalytical Laboratories, Tor- rance, Calif. Where analyses are indicated only by symbols of the elements, analytical results obtained for those elements were within +0.40/0 of the theoretical values.


Contractile Proteins and MuscleAlfred Burger

J. Med. Chem., 1972, 15 (2), 216-216• DOI: 10.1021/jm00272a900 • Publication Date (Web): 09 January 2004






216 Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 Book Reviews

Book Reviews

Contractile Roteins and Muscle. Edited by Koloman Laki, with 14 other contributors. Marcel Dekker, New York, N. Y. 1971. xii + 606 pp. 16 X 23.5 cm. $42.50

Muscle transduces chemical energy into mechanical work. It contains a succession of segments of thin and thick filaments. The thin filaments appear to make contact with the headpiece of myosin which bends out from the thick filaments, and these sections shorten while the unengaged filaments slide. The cycle is concluded by disengagement of the filaments and elongation of the shortened fibriles. We are at the verge of recognizing the role of individual proteins in these changes; this book is an account of the history of muscle research from A. Szent-Gyorgyi’s early observations on the chemistry of muscle contraction to the present. The evolution of muscles may date back to ciliary and flagellar apparatus which are treated as suitable objects of study. Flagellin, actin, tropomysin A and B, and myosin are discussed in individual chapters. The decomposition of ATP is tied to myosin and muscle work, and critical examination is accorded to these reactions. For the anatomist, chapters on the functional morphology and the development of muscle and the excitable membrane will be of value. Biochemists will profit from reading about the chemical changes in contracting muscle and about muscle as a fibrous protein system. Medicinal chemists will welcome an outline of muscle diseases (atrophies, dystrophies, myotonias, myasthenias, glycogenoses, etc.), and a chapter on the uterus as a model for medical understanding and therapy. The area of muscular disorders, treatable mostly only by methods which suppress symptoms, should now open up to causative cures. This book will do much to make available the literature and expert judgment of the background and status in this field. University of Virginia Charlottesville, Virginia

Alfred Burger

Annual Reports in Medicinal Chemistry, 1970. Edited by C . K. Cain and 6 section editors, 49 contributors. Academic Press, New York, N. Y. 1971. ix + 300 pp. 25.3 X 17.8 cm. paper-back. $9.00.

The 6 major topics reviewed in this volume are CNS agents, phar- macodynamic agents (antihypertensives, platelet aggregation inhibitors, gastrointestinal, antiarrhythmic, and diuretic drugs), chemo- therapeutics, and metabolic disease and endocrine function drugs. Drug metabolism and adenylcyclase and CAMP analogs are covered in Topics in Biology. The annual feature, Topics in Chemistry, contains a review of linear free energy and MO hypotheses in drug design, as well as lesser articles on pharmaceutic subjects.

date. The selection of the topics reflects the general activity in medicinal research, particularly in the industry. Editor Cain now pub- lically avows the position this reviewer has taken for several years when he writes in the preface “. . . there were no real breakthrough3 reported during 1970. This may well characterize the whole area of Medicinal Chemistry at present. Where are the new drugs?”

Many major companies now have promising novel drugs in clinical trial. If nothing happens to them on the way to the FDA, a reversal of the present pessimistic trend may be expected around 1976. How- ever, the intellectual progress of drug discovery and design appears to be in a stalemate. The high hopes statistical and quantitative methods have held out are beginning to fade, and regression analysis aided by MO calculations now looks only like another tool of limited applicability which will save some time in routine molecular modification. No other exciting ideas have appeared on the horizon.

All the reviews are well edited, well documented, and quite up-to-

University of Virginia Charlottesville, Virginia

Alfred Burger

Narcotic Drugs. Biochemical Pharmacology. Edited by Doris H. Clouet, with 29 contributors. Plenum Press, New York, N. Y. 1971. 16.5 X 23.5 cm. xxii + 506 pp. $28.00.

The introductory chapters of this book deal with the structure of narcotic analgetics, their methods of chemical analysis, and structure-activity relationships among these drugs. They represent author- itative, stereochemically oriented, and up-todate factual reviews of

the clinically used and a few major experimental compounds and offer historical views of the intellectual processes that led to their development. The chapter on analysis contains all modern spectroscopic methods used in the analysis of minute quantities of foreign substances in body fluids and tissues and should be useful far beyond the scope of this specific class of drugs.

But from Chapter 4 on, the subtitle of the volume, biochemical pharmacology, takes over and encounters all the vicissitudes of such studies. Metabolites of alkaloidal and synthetic potent analgetics have been identified, to be sure, but the explanation of the action of the drugs on tissues and enzymes is almost as hazy and inconclu- sive as it was 50 years ago, in spite of the herculean amount of work that has been done on these questions. Each major biosynthetic pathway (carbohydrates, proteins, lipids, catecholamines, ACh, eti . J is presented to the reader as background information, and these condensed reviews are followed by careful compilations of the effects of narcotic analgetics upon these processes and the enzymes involved. It is obvious that the potent analgetics disrupt many physiological functions, but it has not yet been possible to single out any specific enzyme system, CNS location, or interference with a metabolite bjo- synthesis as a decisive step. These difficulties are multiplied in the attempts to explain tolerance, dependence, and the withdrawal syndrome. The many conflicting hypotheses concerning these phenor- ena are reviewed skillfully but the authors have been careful not to propose experimentally unwarranted conclusions.

We have today analgetics thousands times more potent than morphine, others almost free from dependence liability, and some virtually free from side-effects at therapeutic doses. We have a dim outline of some three-dimensional regions of the analgetic receptors. we know the major metabolites of several of the drugs, and we PO’: sess compounds that can counteract heroin dependence or at least modify withdrawal therapeutically. But the intrinsic mode of action of analgetics still eludes us. This book is a capable review o f all these facets and is full of suggestions for future researches. University of Virginia Charlottesville, Virginia

Alfred Biirger

Lipid Metabolism. Edited by Salih J. Wakil, with 12 contributors. Academic Press, New York, N. Y. 1970. xi + 613 pp, 16.S Y 27 h cm. $28.50.

This is a timely and well-written book which is a formidable addi. tion to the literature of lipid biochemistry. S. Wakil has done an excellent job in assembling a group of highly respected scientists to write on various facets of lipid metabolism and has provided a well- integrated book on topics ranging from fatty acid metabolism to the biosynthesis of polyisoprenoid quinones. Essentially the book can be divided into two major parts, the first of which encompasses fa-tty acid metabolism (S. Wakil), physiological-chemical aspects of fatty acid oxidation (R. Bressler), fatty acid metabolism of plants (P. R Stumpf), phospholipid metabolism (E, E, Hill and W. E. M. Lands), and glyceride metabolism (G. Hubscher). The other major division, though not as closely interrelated as the first, centers on the follov, ing important topics: prostaglandins (B. Samuelson), bacterial lipids (W. J. Lennarz), steroid metabolism (P. W. Holloway), biosynthesis of polyisoprenoid quinones and related compounds (R. Bentley ). These chapters are well developed and provide an in-depth description of progress in these fields to 1969. The authors are all expert.; in their respective areas of interest and this is certainly indicated by depth of understanding and coverage in each chapter. This is one of the few books recently published on lipids to which the term “significant” can be applied. Undoubtedly, it will prove to be an important guide-post for several years to come. Perhaps the only de. tracting factor is the cost of the book, which at $28.50 will rule out many graduate students and postdoctorates and even some of the less affluent professors. It is hoped that the publishing companies in the near future will take a serious look at the best manner in which to provide texts and reference books at a reasonable price and also consider a new mechanism by which publication can be achieved in a shorter period of time. Department of Biochemistry Universily of Arizona Tucson , Arizona

Donald J. Itanahan


Narcotic Drugs. Biochemical PharmacologyAlfred Burger

J. Med. Chem., 1972, 15 (2), 216-216• DOI: 10.1021/jm00272a902 • Publication Date (Web): 09 January 2004






216 Journal of Medicinal Chemistry, 1972, Vol. IS, No. 2 Book Reviews

Book Reviews

Contractile Roteins and Muscle. Edited by Koloman Laki, with 14 other contributors. Marcel Dekker, New York, N. Y. 1971. xii + 606 pp. 16 X 23.5 cm. $42.50

Muscle transduces chemical energy into mechanical work. It contains a succession of segments of thin and thick filaments. The thin filaments appear to make contact with the headpiece of myosin which bends out from the thick filaments, and these sections shorten while the unengaged filaments slide. The cycle is concluded by disengagement of the filaments and elongation of the shortened fibriles. We are at the verge of recognizing the role of individual proteins in these changes; this book is an account of the history of muscle research from A. Szent-Gyorgyi’s early observations on the chemistry of muscle contraction to the present. The evolution of muscles may date back to ciliary and flagellar apparatus which are treated as suitable objects of study. Flagellin, actin, tropomysin A and B, and myosin are discussed in individual chapters. The decomposition of ATP is tied to myosin and muscle work, and critical examination is accorded to these reactions. For the anatomist, chapters on the functional morphology and the development of muscle and the excitable membrane will be of value. Biochemists will profit from reading about the chemical changes in contracting muscle and about muscle as a fibrous protein system. Medicinal chemists will welcome an outline of muscle diseases (atrophies, dystrophies, myotonias, myasthenias, glycogenoses, etc.), and a chapter on the uterus as a model for medical understanding and therapy. The area of muscular disorders, treatable mostly only by methods which suppress symptoms, should now open up to causative cures. This book will do much to make available the literature and expert judgment of the background and status in this field. University of Virginia Charlottesville, Virginia

Alfred Burger

Annual Reports in Medicinal Chemistry, 1970. Edited by C . K. Cain and 6 section editors, 49 contributors. Academic Press, New York, N. Y. 1971. ix + 300 pp. 25.3 X 17.8 cm. paper-back. $9.00.

The 6 major topics reviewed in this volume are CNS agents, phar- macodynamic agents (antihypertensives, platelet aggregation inhibitors, gastrointestinal, antiarrhythmic, and diuretic drugs), chemo- therapeutics, and metabolic disease and endocrine function drugs. Drug metabolism and adenylcyclase and CAMP analogs are covered in Topics in Biology. The annual feature, Topics in Chemistry, contains a review of linear free energy and MO hypotheses in drug design, as well as lesser articles on pharmaceutic subjects.

date. The selection of the topics reflects the general activity in medicinal research, particularly in the industry. Editor Cain now pub- lically avows the position this reviewer has taken for several years when he writes in the preface “. . . there were no real breakthrough3 reported during 1970. This may well characterize the whole area of Medicinal Chemistry at present. Where are the new drugs?”

Many major companies now have promising novel drugs in clinical trial. If nothing happens to them on the way to the FDA, a reversal of the present pessimistic trend may be expected around 1976. How- ever, the intellectual progress of drug discovery and design appears to be in a stalemate. The high hopes statistical and quantitative methods have held out are beginning to fade, and regression analysis aided by MO calculations now looks only like another tool of limited applicability which will save some time in routine molecular modification. No other exciting ideas have appeared on the horizon.

All the reviews are well edited, well documented, and quite up-to-

University of Virginia Charlottesville, Virginia

Alfred Burger

Narcotic Drugs. Biochemical Pharmacology. Edited by Doris H. Clouet, with 29 contributors. Plenum Press, New York, N. Y. 1971. 16.5 X 23.5 cm. xxii + 506 pp. $28.00.

The introductory chapters of this book deal with the structure of narcotic analgetics, their methods of chemical analysis, and structure-activity relationships among these drugs. They represent author- itative, stereochemically oriented, and up-todate factual reviews of

the clinically used and a few major experimental compounds and offer historical views of the intellectual processes that led to their development. The chapter on analysis contains all modern spectroscopic methods used in the analysis of minute quantities of foreign substances in body fluids and tissues and should be useful far beyond the scope of this specific class of drugs.

But from Chapter 4 on, the subtitle of the volume, biochemical pharmacology, takes over and encounters all the vicissitudes of such studies. Metabolites of alkaloidal and synthetic potent analgetics have been identified, to be sure, but the explanation of the action of the drugs on tissues and enzymes is almost as hazy and inconclu- sive as it was 50 years ago, in spite of the herculean amount of work that has been done on these questions. Each major biosynthetic pathway (carbohydrates, proteins, lipids, catecholamines, ACh, eti . J is presented to the reader as background information, and these condensed reviews are followed by careful compilations of the effects of narcotic analgetics upon these processes and the enzymes involved. It is obvious that the potent analgetics disrupt many physiological functions, but it has not yet been possible to single out any specific enzyme system, CNS location, or interference with a metabolite bjo- synthesis as a decisive step. These difficulties are multiplied in the attempts to explain tolerance, dependence, and the withdrawal syndrome. The many conflicting hypotheses concerning these phenor- ena are reviewed skillfully but the authors have been careful not to propose experimentally unwarranted conclusions.

We have today analgetics thousands times more potent than morphine, others almost free from dependence liability, and some virtually free from side-effects at therapeutic doses. We have a dim outline of some three-dimensional regions of the analgetic receptors. we know the major metabolites of several of the drugs, and we PO’: sess compounds that can counteract heroin dependence or at least modify withdrawal therapeutically. But the intrinsic mode of action of analgetics still eludes us. This book is a capable review o f all these facets and is full of suggestions for future researches. University of Virginia Charlottesville, Virginia

Alfred Biirger

Lipid Metabolism. Edited by Salih J. Wakil, with 12 contributors. Academic Press, New York, N. Y. 1970. xi + 613 pp, 16.S Y 27 h cm. $28.50.

This is a timely and well-written book which is a formidable addi. tion to the literature of lipid biochemistry. S. Wakil has done an excellent job in assembling a group of highly respected scientists to write on various facets of lipid metabolism and has provided a well- integrated book on topics ranging from fatty acid metabolism to the biosynthesis of polyisoprenoid quinones. Essentially the book can be divided into two major parts, the first of which encompasses fa-tty acid metabolism (S. Wakil), physiological-chemical aspects of fatty acid oxidation (R. Bressler), fatty acid metabolism of plants (P. R Stumpf), phospholipid metabolism (E, E, Hill and W. E. M. Lands), and glyceride metabolism (G. Hubscher). The other major division, though not as closely interrelated as the first, centers on the follov, ing important topics: prostaglandins (B. Samuelson), bacterial lipids (W. J. Lennarz), steroid metabolism (P. W. Holloway), biosynthesis of polyisoprenoid quinones and related compounds (R. Bentley ). These chapters are well developed and provide an in-depth description of progress in these fields to 1969. The authors are all expert.; in their respective areas of interest and this is certainly indicated by depth of understanding and coverage in each chapter. This is one of the few books recently published on lipids to which the term “significant” can be applied. Undoubtedly, it will prove to be an important guide-post for several years to come. Perhaps the only de. tracting factor is the cost of the book, which at $28.50 will rule out many graduate students and postdoctorates and even some of the less affluent professors. It is hoped that the publishing companies in the near future will take a serious look at the best manner in which to provide texts and reference books at a reasonable price and also consider a new mechanism by which publication can be achieved in a shorter period of time. Department of Biochemistry Universily of Arizona Tucson , Arizona

Donald J. Itanahan

journal of medicinal chemistry volume 15.pdf

Documents

deaminopenicillamine

deaminopenicillamine

oxytocin figure

related analogs

oxytocic activity of

devoid of oxytocic activity

inhibitory effect

antioxytocic agent