KELVIN K. OGILVIE,~ NGHE NGUYEN-BA, AND RAYMOND G. HAMILTON Department of Chemistry, McGill Universily, Montreal, P.Q.

Transcription

1 A trihydroxy acyclonucleoside series' KELVIN K. OGILVIE,~ NGHE NGUYEN-BA, AND RAYMOND G. HAMILTON Department of Chemistry, McGill Universily, Montreal, P.Q., Canada H3A 2K6 Received December 8, 1983 This paper is dedicated to Professor Gerald E. Dunn on the occasion of his 65th birthday KELVIN K. OGILVIE, NGHE NGUYEN-BA, and RAYMOND G. HAMILTON. Can. J. Chem. 62, 1622 (1984). An acyclonucleoside structure (Z) has been developed which possesses three hydroxymethyl groups attached to the carbon which is analogous to the 4'-carbon of natural nucleosides. The presence of the third hydroxymethyl group significantly reduces activity against herpesviruses as compared to the dihydroxymethyl (BIOLF-62) and monohydroxymethyl (acyclovir) models. KELVIN K. OGILVIE, NGHE NGUYEN-BA et RAYMOND G. HAMILTON. Can. J. Chem. 62, 1622 (1984). On a dcveloppc une structure acyclonucitoside (Z) comportant un carbone portant trois groupes hydroxymtthyles qui est analogue au carbone 4' des nuclcosides naturels. Si on la compare aux modkles dihydroxymcthyles (BIOLF-62) et monohydroxymcthylcs (acyclovir), on note que la prksence d'un troisikme groupe hydroxymkthyie rcduit d'une fa~on importante I'activitC contre les virus herpks. [Traduit par le journal] Introduction Nucleoside analogues represent an important group of biologically active molecules (1). We have developed an analogue structure Y (2-6, the "glycerosides"), members of which have shown pronounced antiviral activity (5, 7-12). As part of our overall program to determine the limits on structural changes to the side chain with regard to optimal biological activity, we have introduced specific changes at the 3'-position. One such structural change involves the introduction of an additional hydroxyrnethyl group at the 3'-position (analogous to the 4'-position of natural nucelosides). In this report we wish to describe the synthesis of compound Z where B = adenine, cytosine, guanine, 6-chloropurine, and thymine. When B = guanine, the series X, Y, and Z represent a progression of hydroxymethyl modification at the 4'-position. The order of activity against herpesviruses is Y > X 9 Z where X is acyclovir (1, 13). Discussion and results The key intermediate in the synthesis is compound 5. The route to compound 5 is outlined in Scheme 1. The starting material, 1,3-dibenzyloxy-2-propanol (I), is readily available (5) and is easily oxidized to the ketone 2 using N-chlorosuccinimide and dimethyl sulfide (14). Other common oxidizing agents such as pyridinium chlorochromate, pyridinium dichromate, and Jones reagent were less successful. Compound 2 was smoothly converted into the epoxide 3 in high yield using trimethylsulfoxonium iodide (15). Sodium hydride was very effective in this reaction provided that rigorously anhydrous conditions were maintained; otherwise a number of byproducts formed. However, n-butyllithium gave good yields ' Part VI in a series on antiviral and ring-open nucleoside analogues. Part V is ref. 5. Author to whom correspondence may be addressed. consistently. The epoxide ring of 3 was opened by the attack of benzylate anion at the least hindered site to produce the desired 1,3-dibenzyloxy-2-benzyloxymethyl-2-propanol (4). In preparing the glyceroside series Y we had activated the intermediate alcohol by conversion to the chloromethyl ether. This conversion requires formaldehyde under strongly acidic conditions (HCl)..When applied to 4, a product was obtained which was quite unstable because of the adjacent tertiary carbon system. To avoid the acidic conditions we attempted the synthesis of the thiomethyl ether system 5. The alcohol 4 was smoothly converted to the stable product 5 using acetic anhydride in DMSO (16). The thiornethyl ether system is readily activated by iodine, allowing nucleophilic attack at the methylene position. As a result, compound 5 was coupled to purines and pyrimidines using the silylated base procedure (3, 17-19) to produce the desired compounds of general structure Z as outlined in Scheme 2. As usual, the guanine derivative (12) was produced in the lowest yield. However, the process was unexpectedly easy to carry out. The N-7 isomer is produced in significant quantity along with the N-9 isomer. Fortunately, in the method used (beginning with N-acetylguanine), the N-7 isomer crystallizes from solution after removal of the N-acetyl group from the direct condensation product. This greatly facilitates an otherwise tedious separation process. This route to the guanine derivative was prompted by a similar and earlier success in the synthesis of the guanine derivative of Y (BIOLF-62).3 The route chosen to the adenine derivative 9 involved 6-chloropurine in the condensation step. Using the conditions described (48 h condensation time), very little of the N-7 isomer is present at the end of the reaction. The chlorine at position 6 is readily displaced by ammonia to give the adenine derivative 9. All of the compounds were identified by their spectral characteristics. In this regard the I3C chemical shift data (Table 1) and the uv data (Table 2) are most useful. These identifications are confirmed by comparison of these results to those from the series Y compounds (3-6). The compounds described in this report have been subjected to testing against herpesviruses. Only the guanine derivative 12 had an ED-50 of less than 100 p,g/ml with HSV-1. The ED-50 Unpublished results.

2 OGILVIE ET AL. OH I +CH20CH2CHCH20CH24 t 4CH20 OCH2Q Toluene 2 1 A NaH ' $CH20H 4CH20 AH24 4CH20 0CH2+ OCH24 3 TABLE I. I3C nuclear magnetic resonance data (6, ppm) of acyclonucleosides" Compound Base ring C-2 C-4 C-5 C-6 C-8 C-I' C-3' C-4' -- 6 Cytosine Thymineb Chloropurine Adenine Guanine Guanine "Spectra were obtained in DMSO-d, on a Bruker WH90 spectrometer. "he C-5 methyl group appeared at ppm. - value of 12 at 54 pg/ml is substantially below that of BIOLF- spectra were recorded using a Bruker WH-90, a Varian XL-200, and 62 (Y, B = guanine) where the ED-50 vs. HSV-1 is 0.2 a Varian T60A spectrometer. Experimental General methods Products from preparative scale reactions were isolated by column chromatography using Merck silica gel 60 ( mesh) packed in glass columns (15 g of silica per gram of crude material). Thin-layer chromatographic data (R, values) are recorded from Merck Kieselgel 60F254 analytical sheets. Solvent A consists of isopropyl alcohol - concentrated ammonium hydroxide - water (7: 1 :2). Purine bases were purchased from Sigma Chemical Co. Ultraviolet spectra were recorded on a Cary 17 spectrometer. Nuclear magnetic resonance 1,3-Dibenzyloxyncetone (2) N-chlorosuccinimide (43 g) was suspended in 500 ml of toluene (Spectrograde) and the mixture was cooled in an ice bath. Dimethyl sulfide (35 ml) was added and a white precipitate formed immediately. The mixture was then cooled to -25 C in a dry ice - carbon tetrachloride bath. 1,3-Dibenzyloxy-2-propanol (60 g) in toluene (50 ml) was added to the mixture via syringe through a serum cap. The mixture was kept under nitrogen for 2 h, whereupon triethylamine (45 ml) was added and the cooling bath was removed. After a further 25 min the solution was collected by filtration. The residue was washed with ether (500 ml). The filtrate and washings were combined, neutralized with 5% aqueous HCI, washed with saturated sodium chloride

3 CAN. J. CHEM. VOL. 62, 1984

4 OGILVIE ET AL. TABLE 2. Ultraviolet spectral data of acyclonucleosides A,,,,,, nm ( 1 Compound Base ring PH 1 PH 7 ph 14 6 Cytosine 276(10 000) 7 Thymine 265(11 500) 8 6-Chloropurine 263(7 400) 9 Adenine 257(12 400) 12 Guanine 254(11 800) 274(8 200) 13 Guanine 250(9 500) 268(7 500) in water and with water, and finally dried over magnesium sulfate. The solvents were removed at reduced pressure (liquid nitrogen trap for dimethyl sulfide) and the resulting oil was distilled in a Kugelrohr apparatus. The product (50 g, 83%) was obtained in the fraction boiling at C (0.05 Torr) (1 Torr = Pa). Normal high vacuum distillation was not satisfactory as the product decomposes at high temperature. The 'Hmr spectrum (CDCI, + TMS) showed 0 I I signals at 6 (ppm) of 4.16 (s, 4H, -C--CH2-0-), 4.50 (s, 4H, -CHz+), and 7.23 (s, ]OH, 6). This material was used directly in the subsequent experiment. 2,2-Dibenzylo.rymethylethylene oxide (3) Trimethylsulfoxonium iodide (25.3 g, mol) was dissolved in dry DMSO (350 ml). n-butyllithium (66 ml, 1.65 M in hexane) was added via syringe to the mixture kept under nitrogen. The mixture was stirred for 25 min and then cooled to room temperature, at which point compound 2 (26.7 g, mol) in DMSO (10 ml) was added via syringe. The mixture was stirred for a further 45 min and then poured into 1 L of cold water. The product was extracted into ether (3 X 300 ml). The organic layer was washed with saturated NaCl solution and dried over magnesium sulfate. Solvents were removed at reduced pressure and the residue was distilled in a Kugelrohr apparatus. The product (bp "C/0.05 Torr) was obtained in 82% yield (23 g). The 'Hmr spectrum (CDCI, + TMS) showed signals at 6 (ppm) of (s, 2H, CH~'&), 3.80 (4H, -CH,--), 4.50 (s, 4H, -CH2+), and 7.20 (s, ]OH, 4). 1,3-Dibenzyloxy-2-benzyloxymethyl-2-propanol (4) To benzyl alcohol (150 ml) in a three-neck flask (500 ml) was added sodium hydride (2.4 g, 0.1 mol obtained by washing 4 g of 60% sodium hydride dispersion in oil with hexane) in portions under a nitrogen atmosphere. The resulting mixture was stirred for 1 h. Compound 3 (22.5 g, 0.08 mol) in benzyl alcohol (20 ml) was added via syringe. The reaction mixture was heated at C for 18 h. After cooling to room temperature, the mixture was poured into 500 ml of cold water and the solution was neutralized with 5% aqueous HCI. The product was extracted into ether (2 x 300 ml) and the combined extracts were washed with water and saturated NaCl solution. After drying over magnesium sulfate the solvent was removed to leave an oil which was distilled in a Kugelrohr apparatus. The fraction boiling at C (0.1 Torr) contained the desired product (26.7 g, 86%). The 'Hmr spectrum (CDCI, + TMS) showed signals of 6 (ppm) of 2.80 (s, IH, -OH), 3.53 (s, 6H, -CH20-), 4.46 (s, 6H, -CH2+), and 7.20 (s, 15H, 4). This material was used directly in the subsequent experiment. The methylthiomethyl ether 5 Compound 4 (28 g) was dissolved in 100 ml of dry DMSO (100 ml) followed by the addition of acetic anhydride (100 ml). The solution was stirred at room temperature for 48 h. The solvents were removed at reduced pressure and the residue was dissolved in 300 ml of ether. After washing with water and saturated NaCl solution, the ether solution was dried over magnesium sulfate and again concentrated to an oil at reduced pressure. The product (27.3 g, 84%) was collected in the fraction boiling at C (0.1 Torr) on distillation in a Kugelrohr apparatus. The 'Hmr spectrum (CDCI, + TMS) showed signals at 6 (ppm) of 2.20 (s, 3H, -SCH3), 3.70 (s, 6H, -CH,O), 4.53 (s, 3H, -CH&), 4.83 (s, 2H, -CH2S-), and 7.30 (s, 15H, +). The product 5 was stable on standing and was used directly in the condensation reactions described below. Standard procedure for the condensation of 5 with purines and pyrimidines. Reaction condition 1 The purine or pyrimidine base (10 mmol) was heated at reflux in HMDS (50 ml) to which ammonium sulfate (200 mg) had been added. Reflux was continued until the solution became clear (30 min for 6-chloropurine; 3 h for cytosine and thymine; 18 h for N- acetylguanine). At this point the excess HMDS was removed by distillation at reduced pressure. Atmospheric pressure was restored using nitrogen and the solid residue was dissolved in 75 ml of freshly distilled THF. Compound 5 (4.52 g, 10 mmol) dissolved in dry THF (15 ml) is added, followed by 5 g of oven activated molecular sieves (type 4A). Iodine (2.54 g) was added and the solution was stirred at room temperature for 48 h. (A nitrogen atmosphere was maintained at all times. For the N-acetylguanine reaction, the solution was stirred at room temperature for 24 h, followed by heating at a gentle reflux for 6 h.) Sodium sulfite solution (150 ml, 5% aqueous) was added and the resulting solution was extracted with methylene chloride (3 x 200 ml). The combined extracts were washed with saturated NaCl solution and water and then dried over magnesium sulfate. The solvent was removed at reduced pressure and the residue was flashchromatographed on silica gei. The specific solvents for each base and the properties of the product are described below. Standard procedure for the removal of the benzyl groups. Reaction condition 2 The product (2 g) from reaction condition 1 (the tribenzyl product) was dissolved in methylene chloride (50 ml) and the solution was coolled to -78OC. A 1 N solution of BCI, in methylene chloride (23 ml) was added via syringe and under a nitrogen atmosphere. The mixture was stirred at -78'C for 3 h. A solution of methanol and methylene chloride (1 : 1, 50 ml) was added and the cooling bath was removed. The clear solution was neutralized with triethylamine or sodium bicarbonate and stirred for an additional 35 min. The solvent was removed at reduced pressure, water (20 ml) was added, and again the solvent was removed at reduced pressure. This process was repeated three times, followed by the evaporation of absolute ethanol (30 ml). At this point variations in the work-up occurred. The remaining procedures are described for the individual compounds listed below. I-[[I,3-Dihydroxy-2-(hydro.rymethy1)-2-propo.~ 6 Reaction condition 1 (from cytosine) The product was isolated by flash chromatography using chloroform - ethyl acetate - ethanol (4:5: 1) as eluant. The yield of the

5 1626 CAN. J. CHEM. tribenzylated product was 3.5 g (58%). The product had an R,of 0.31 in chloroform-methanol (9: 1). The 'Hmr spectrum (CDCI, + TMS) showed signals at 6 (ppm) of 3.66 (s, 6H, -CH20-), 4.50 (2, 6H, -CH24), 5.40 (s, 2H, -NCH20-), 5.73 (d, lh, 5 = 6 HZ, H-5), 7.30 (s, total 16H, 15 from 4 and H-6 buried under aromatic H's). Reaction condition 2 The above product was treated as described in the general procedure. The residue so obtained was extracted with chloroform (30 ml) under stirring and the solution was filtered through Celite. The Celite was washed twice with chloroform (2 x 20 ml) which removed the triethylamine hydrochloride. The Celite was then washed with 150 ml of hot methanol. The methanol solution was concentrated at reduced pressure and the residue was crystallized from methanol-ether to give the desired product 6 (72%, mp "C, Rr in solvent A of 0.53). The 'Hmr spectrum (DMSO-d6 + TMS) showed signals at 6 (ppm) of 3.50 (s, 6H, -CH2-OH), 5.23 (s, 2H, -NCH20-), 5.80 (d, lh, H-5, 55.6 = 6 HZ), 7.73 (d, lh, H-6, 55.6 = 6 HZ), and 7.73 (broad, 2H, -NH2). The 13C chemical shifts are recorded in Table 1. The uv characteristics are collected in Table 2. Anal. calcd. for C9HI5N3O5-1/2H20: C 42.52, H 6.29, N 16.53; found: C 42.62, H 6.15, N [[I,3-Dihydroxy-2-(hydroxytnethyl)-2-propoxy]methyl]thymine 7 Reaction condition 1 (from thymine) The tribenzyl product was recovered as an oil in 74% yield from flash chromatography using ethyl acetate - hexane (1 : 1) as eluant. The 'Hmr spectrum (CDC13 + TMS) showed signals at 6 (ppm) of 1.70 (s, 3H, -CH3), 3.63 (s, 6H, -CH,O-), 4.46 (s, 6H, -CH24), 5.30 (s, 2H, -NCH20-), 7.10 (m, lh, H-6), and 7.20 (s, 15H, 4). The compound had A,,, of 263 nm in methanol. This intermediate was subjected directly to the general reaction condition 2. Reaction condition 2 The residue was stirred with 30 ml of hot methanol and filtered. Silica gel (5 g) was added and the resulting slurry was concentrated in a rotary evaporator and the silica was dried under high vacuum overnight. This silica was placed on top of a prepared silica gel column which was then eluted with methylene chloride - methanol (4: 1). The desired product 7 was obtained in 70% yield (mp C). The product had an Rr of 0.48 in methylene chloride - methanol (4: 1). The 'Hmr spectrum (DMSO-d6 + TMS) showed signals at 6 (ppm) of 1.80 (s, 3H, -CH3), 3.55 (s, 6H, -CH20-), 4.50 (broad, 3H, -OH), 5.23 (s, 2H, -NCH20-), and 7.53 (s, lh, H-6). The I3C spectral properties and the uv data are collected in Tables 1 and 2, respectively. Anal. calcd. for C lahi6n2o5: C 46.16, H 6.19, N 10.76; found: C 46.00, H 6.08, N [[1,3-Dihydroxy-2-(hydroxymethyl)-2-propoxy]methyl]-6- chloropurine 8 Reaction condition 1 (from 6-chloropurine) The residue was flash chromatographed on silica gel using ethyl acetate - hexane (1: 1). The tribenzyl product was obtained as an oil in 55% yield with Rr of 0.38 in the eluting solvent. The uv spectrum (methanol) showed A,,, at 263 nm. The '~mr spectrum (CDC13 + TMS) showed signals at 6 (ppm) of 3.63 (s, 6H, -CH20-), 4.43 (s, 6H, -OCH24), 5.90 (s, 2H, -NCH20-), 7.20 (s, 15H), 8.20 (s, lh, H-2), and 8.66 (s, lh, H-8). A trace amount of another product, assumed to be the N-7 isomer, was also obtained. This material had Rr 0.22 in the eluting solvent and a A,,, of 267 nm in methanol. The 'Hmr spectrum of this product (CDCI, + TMS) showed signals at 6 (ppm) of 3.70 (s, 6H, -CH20-), 4.40 (s, 6H, -OCH24), 6.0 (s, 2H, -NCH20-), 7.30 (s, 15H, +), 8.30 (s, lh, H-2), and 8.80 (s, lh, H-8). Reaction condition 2 The residue (obtained by starting with the desired N-9 product from reaction condition 1) was extracted with warm methanol. The solution was filtered and treated with 5 g of silica gel as described for compound 7. The column was eluted with methylene chloride - methanol (4: 1). The desired product 8 was crystallized from methanol-ether in 65% yield (mp C) and had R, of 0.58 in the eluting solvent. The 'Hmr spectrum (DMSO-d6 + TMS) showed signals at 6 (ppm) of 3.66 (s, 6H, -CH20H), 4.66 (t, 3H, -OH), 6.00 (s, 2H, -NCH20-), 8.83 (s, lh, H-2), and 8.86 (s, IH, H-8). The I3C spectral data and the uv data are collected in Tables 1 and 2 respectively. Anal. calcd. for CloH13C1N404. 1/2H20: C 40.34, H 4.74, N 18.82; found: C 40.72, H 4.38, N [[I,3-Dihydroxy-2-(hydroxymethy1)-2-prop- 9 The N-9 product of reaction condition 1 from the 6-chloropurine condensation was heated in a steel bomb at 125 C for 17 h in saturated (at 0 C) methanolic ammonia (15 ml/g of starting material). The bomb was cooled to room temperature, the solvents were evaporated, and the product obtained by chromatography on silica gel using methylene chloride - methanol (9: 1) as eluant. The product was obtained in 88% yield and had A,,, (methanol) at 257 nm and an Rf of 0.68 in the eluting solvent. The 'Hmr spectrum (CDCl, + TMS) showed signals at 6 (ppm) of 3.70 (s, 6H, -CH20-), 4.40 (s, 6H, +CH24), 5.80 (s, 2H, -NCH20-), 6.20 (b, 2H, -NH2), 7.30 (s, 15H, +), 8.0 (s, lh, H-2), and 8.30 (s, lh, H-8). This material was subjected to reaction condition 2 and the product was obtained by silica gel chromatography using methylene chloride - methanol (4: 1). The product 9 (76%) had mp C and an R, of 0.17 in the eluting solvent. The 'Hmr spectrum (DMSO-d6 + TMS) showed signals at 6 (ppm) of (bs, 6H, -CHIOH), 5.80 (s, 2H, -NCH20-), 7.40 (b, 2H, -NH2), 8.20 (s, IH, H-2), and 8.30 (s, lh, H-8). The I3C spectral data and the uv data are collected in Tables 1 and 2 respectively. Anal. calcd. for C 15N /2H20: C 43.16, H 5.79, N 25.17; found: C 43.41, H 5.81, N [[I,3-Dihydroxy-2-(hydroxymethyl)-2-propoxy]methyl]guanine (11) and 7-[[1,3-dihydroxy-2-(hydroxymethy1)-2-propoxy]- methyl]guanine (12) Reaction condition 1 The starting material in this condensation was N-acetylguanine. The residue from reaction condition 1 was flash-chromatographed on silica gel using chloroform - ethyl acetate - ethanol (4: 5 : 1). An oil was obtained (43%) which according to 'Hmr was a mixture of the N-9 and N-7 isomers (3: 2). The mixture was dissolved in dry methanol (50 ml) and this solution was added to methanol (100 ml) containing sodium (500 mg). This solution was heated at reflux for 3 h. After cooling to room temperature the solution was carefully neutralized using glacial acetic acid. White crystals formed during the neutralization and these were collected by filtration and found to be the pure N-7 isomer (11) in 9% overall yield (mp C). The filtrate was concentrated and chromatographed on silica gel using methylene chloride - methanol (9: 1). The N-9 isomer (10) was obtained in 15% overall yield (mp 185 C (dec.)). The 'Hmr spectrum of N-9 isomer 10 (DMSO-d6 + TMS) showed signals at 6 (pprn) of 3.6 (s, 6H, -CH,O-), 4.53 (s, 6H, -CH24), 5.60 (s, 2H, -NCH20-), 6.46 (b, 2H, -NH2), 7.3 (s, 15H, +), 7.66 (s, 1H, H-8), and 9.14 (b, lh, H-1). The 'Hmr spectrum of the N-9 isomer (11) under the same conditions showed signals at 6 (ppm) of 3.63 (s, 6H, -CH20-), 4.53 (s, 6H, -CH24), 5.86 (s, 2H, -NCH20-), 6.23 (b, 2H, -NH2), 7.33 (s, 15H, +), and 7.96 (s, IH, H-8). Reaction condition 2 on 10 The N-9 isomer from above (10) was subjected to the standard conditions. The residue obtained crystallized from 9590 ethanol to give pure 12 in 77% yield (mp >30O0C). The R, in solvent A on tlc was The 'Hmr spectrum of 12 (DMSO-d6 + TMS) showed signals at 6 (ppm) of 3.60 (s, 6H, -CH20H), 5.63 (s, 2H, -NCH20-), 6.53 (b, 2H, -NH2), 7.90 (s, lh, H-8), and (b, lh, H-1). The I3C and uv data are collected in Tables 1 and 2 respectively. Anal. calcd. for CloHI5N5O5: C 42.10, H 5.26, N 24.56; found: C 41.89, H 5.19, N Reaction condition 2 on 11 The N-7 isomer from above (11) was subjected to the standard

6 OGILVIE ET AL i I I conditions. The dry white residue from the reaction was treated with chloroform and filtered through Celite. The Celite was washed with chloroform (2 x 20 rnl). The Celite was then poured into 200 rnl of water which was brought to the boiling point and filtered. The filtrate was lyophilized to yield a white powder (78% yield, RI on tlc in solvent A of 0.53). The 'Hrnr spectrum of 13 in (DMSO-d6 + TMS) showed signals at 6 (pprn) of 3.50 (s, 6H, -CH20H), 5.70 (s, 2H, -NCH20-), 6.40 (b, 2H, -NHZ), and 8.10 (s, lh, H-8). The I3C and uv data are collected in Tables 1 and 2 respectively. Anal. calcd. for CIOHISN505: C 42.10, H 5.26, N 24.56; found: C 42.39, H 5.40, N Acknowledgements - We are grateful for financial support from Bio Logicals Inc. and the Natural Sciences and Engineering Research Council of Canada. We are indebted to Dr. K. 0. Smith for the antiviral tests. 1. R. T. WALKER, E. DE CLERCQ, and F. ECKSTEIN (Editors). Nucleoside analogues. Plenum Press, New York K. K. OGILVIE and M. F. GILLEN. Tetrahedron Lett. 21, 327 (1980). 3. K. K. OGILVIE, U. 0. CHERIYAN, B. K. RADATUS, K. 0. SMITH, K. S. GALLOWAY, and W. L. KENNELL. Can. J. Chern. 60,3005 (1982). 4. K. K. OGILVIE, R. G. HAMILTON, M. F. GILLEN, B. K. RA- DATUS, K. 0. SMITH, and K. S. GALLOWAY. Can. J. Chern. 62, 16 (1984). 5. K. K. OGILVIE, N. NGUYEN-BA, M. F. GILLEN, B. K. RADATUS, U. 0. CHERIYAN, H. R. HANNA, K. 0. SMITH, and K. S. GAL- LOWAY. Can. J. Chern. 62, 241 (1984). 6. K. K. OGILVIE, D. M. DIXIT, B. K. RADATUS, and K. 0. SMITH. Nucleosides Nucleotides, 2, 147 (1983). 7. K. 0. SMITH, K. S. GALLOWAY, W. L. KENNEL, K. K. OGILVIE, and B. K. RADATUS. Antirnicrob. Agents Chernother. 22, 55 (1982). 8. K. 0. SMITH, K. S. GALLOWAY, K. K. OGILVIE, and U. 0. CHERIYAN. Antirnicrob. Agents Chernother. 22, 1026 (1982). 9. K. 0. SMITH, K. S. GALLOWAY, S. L. HODGES, K. K. OGILVIE, B. K. RADATUS, S. S. KALTER, and R. L. HEBERLING. Am. J. Vet. Res. 44, 1032 (1983). 10. A. K. FIELD, H. PERRY, R. LIOU, H. BULL, R. L. TOLMAN, and J. D. KARKAS. Biochern. Biophys. Res. Cornrnun. 116, 360 (1983). 11. Y.-C. CHENG, E.-S. HUANG, J.-C. LIN, E.-C. MAR, J. S. PA- GANO, G. E. DUTSCHMAN, and S. P. GRILL. Proc. Natl. Acad. Sci. U.S.A. 80, 2767 (1983). 12. J. C. MARTIN, C. A. DVORAK, D. F. SMEE, T. R. MATTHEWS, and J. P. H. VERHEYDEN. J. Med. Chern. 26, 759 (1983). 13. H. J. SCHAEFFER, L. BEAUCHAMP, P. DE MIRANDA, G. B. ELION, D. J. BAUER, and P. COLL~NS. Nature, 272, 583 (1978). 14. E. J. COREY andc. U. KIM. J. Am. Chern. Soc. 94,7586 (1972). 15. E. J. COREY and M. CHAYKOVSKY. Org. Synth. 49, 78 (1969). 16. K. YAMADA, K. KATO, H. NAGASE, and Y. HIRATA. Tetrahedron Lett. 17, 65 (1976). 17. T. NISHIMURA, B. SHIMIZU, and I. IWAI. Chern. Pharm. Bull. Jpn. 11, 1470 (1963). 18. E. WITTENBURG. Z. Chern. 4, 303 (1964). 19. G. RITZMANN and W. PFLEIDERER. Chern. Ber. 106, 1401 (1973).