Highly efficient synthesis of 2,2prime;-anhydro-1-(3prime;-bromo-3prime;-deoxy-5prime;-O-trityl-beta;-D-arabinofuranosyl)thymine and its derivatives from an unsaturated thymine nucleoside

摘要

2289J. CHEM. SOC. PERKIN TRANS. 1 1994 Highly Efficient Synthesis of 2'2'-Anhydro-I -(3'-bromo-3'-deoxy-5'-O-trityl +-D-arabinofuranosy1)thymine and its Derivatives from an Unsaturated Thymine Nucleoside Katsumaro Minamoto,*na Masataka Oishi,a Akikazu Kakehi,b Naoki Ohta? lsamu Matsuda/ Kenji Watanabe,a Kazufumi Yanagihara,a Toyohide TakeuchiC and Keizo Tanigawad a Department of Applied Chemistry, School of Engineering, Nag0 ya University, Furo-cho, Chikusa -ku, Nagoya 464, Japan Department of Material Chemistry, Faculty of Engineering, Shinshu University, Wakasato, Nagano 380, Japan Faculty of Engineering, Gifu University, I -7 62I -7 Yanagido, Gifu 507 -I I, Japan Synthesis Research Dept., Central Research Institute, Nissan Chemical Industries, Ltd., 722- I, Tsuboi-cho, Funabashi-shi, Chiba 274, Japan Reaction of 5'-0-trityl-2',3'-thymidinene 1 with hypobromous acid gave (5R,6R) -2,2'-anhydro-5-bromo-1 -(3'-bromo-3'-deoxy-5'-~-trityl-~-~-arabinofuranosyI)-6-hydroxy-5,6-dihydrothymine3a and its (SS,GS)-trans isomer 4a.Similarly, 6-methoxy analogues (3b and 4b) and 6-acetoxy analogues (3c and 4c) of 3a and 4a were synthesized. Compounds 3a and 4a were converted into the corresponding 5,6-epoxy derivatives, 5 and 6. Deoxygenation of oxiranes 5 and 6 with Ph,P gave 2,2'-an hydro- 1-(3'- bromo-3'-deoxy- 5'-0-trityl-P-D-arabinofuranosyl) thymine 7, which was also obtainable in excellent yields from compounds 3a, b or/and 4a, b by treatment with Ph,P-NaHCO,, or directly from unsaturated furanose 1 by one-pot synthesis via methyl ethers 3b and 4b or acetates 3c and 4c. Compound 7 was deprotected to give the mother compound 8 and was also converted into the 2,3-lyxo epoxy thymine furanosides, 11and 12, in highyields.Since the finding that dideoxynucleosides such as 2',3'-dide- halogen atom other than iodine was desirable as a more oxycytidine (ddC), 2',3'-dideoxyinosine (ddI) and 3'-azido- appropriate intermediate. This paper describes the results of the 3'-deoxythymidine (AZT) 'are potentially effective anti-AIDS reaction of compound 1 with hypobromous acid generated in agents, much effort has been directed toward the effective situ from N-bromoacetamide (NBA). deoxygenation of natural nucleosides to 2',3'-dideoxynucleo- The reaction of compound 1 with 2.4 mole equivalents of sides.3 3'-Deoxythymidine, 2',3'-didehydro-3'-deoxythymidine NBA in a mixture of acetone and water gave (5R,6R)-2,2'- On the an hydro- 5-bromo-1-(3'-bromo-3'-deoxy-5'-0-trit yl- P-D-arab- (d4T) and others are also under clinical in~estigation.~ 3a (53%) and its other hand, considerable effort has been devoted to the inofuranosyl)-6-hydroxy-5,6-dihydrothymine synthesis of nucleosides carrying a sugar moiety modified in a (5S,6S)-trans analogue 4a (32%) (Scheme 1).The 2,2'-anhydro variety of ways. However, from the viewpoint of synthesis, the structures of these compounds are in accord with the 'H NMR range of sugar modifications of the 2'-deoxynucleosides is data lacking an N3-H signal and showing an abnormally notably limited by the absence of a 2'-hydroxy group as deshielded 2'-H signal (S5.70 in each case) as well as the large compared with that of ribo or arabino nucleosides. Hence, the JIg,,.-values (6.50 and 5.75 Hz, respectively) (Table 1). Both chemistry of thymine furanosides involving the 2'-functional- compounds displayed only inflections in the UV spectra (see ization has developed starting from the ribosyl or arabinosyl Experimental section).The (5R,6R)-trans structure of com- thymine.5c36 pound 3a was confirmed by X-ray analysis (vide infra, Fig. 1). Largely owing to the current interest in modified thymidine Hence the counterpart 4a should be a (5S,6S)-trans analogues as potential anti-AIDS agents and the far easier diastereoisomer.commercial availability of thymidine as compared with the When fewer than 2 mole equivalents of NBA were used, a 5,6-other thymine furanosides (thymine ribo- or arabino-sides) $ or bromohydrinated derivative of compound 1 was isolated as a general 2'-deoxynucleosides, we recently exploited a method for very minor product.§ Since this compound disappeared from synthesizing 2,2'-anhydro- 1-(3'-deoxy-3'-iodo-5'- 0-trityl-P-D- the reaction mixture on addition of further NBA, and no TLC arabinofuranosy1)thymine 2 'from 5'-0-trityl-2',3'-thymidin-spot corresponding to bromide 7 was observed, the 5,6-bromo- ene 1,' which is easily available from thymidine. Although hydrination appears to have preceded the 2,2'-cyclization. compound 2 was readily converted into another series of Treatment of the alcohol 3a with an excess of triethylamine in important intermediates, 2,3-anhydrolyxofuranosylderivatives acetone under reflux gave a high yield of (SS,6R)-2,2'-anhydro- of thymine and other modified thymine^,^ the 3'-iodo group 1-(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arabinofuranosyl)-5,6-of compound 2, especially in its detritylated form, was found to epoxy-5,6-dihydrothymine5, while similar epoxidation of be quite susceptible to various nucleophiles and protonic acids to regenerate unsaturated furanose 1.Light-induced, gradual 0 This product melted between 146 and 151 "C (Found: C, 61.6; H, 5.0;decomposition of iodide 2 was also observed, especially in N, 4.95. C,,H,,BrN,O, requires C, 61.82; H, 4.83; N, 4.97%);solution.Therefore, an analogue of compound 2 having a 3'- G,(CDCl,) 1.56 (3 H, s, 5-Me), 3.32 (I H, dd, J,,, 10.2, J5,a,4r3.8, 5'-Ha), 3.46(1 H,dd, J,,, 10.2,J5*b,4*4.4,5'-Hb),4.93(1H, ddd, J4,,1*4.0, J4,,21 .6,J4,,3r2.4,4'-H),5.1 6( 1H,S, 6-H), 6.07( 1 H,ddd,J, ,,1 * 1.4,J3S.2f6.0, t AIDS: Acquired Immune Deficiency Syndrome. J3q.4,2.4, 3'-H), 6.31 (1 H, dt, J2,,1,2.0, J2*,3,6.0, J2*,4*1.6,2'-H), 6.87 (1 11 -P-D-Arabinosyl-or ribosyl-thymine is several tens of times more H, ddd, J,.,,.2.0, J,.,,, 1.4, J1,,4f4.0, 1'-H), 7.28-7.46 (16 H, m, ArH expensive than thymidine. and 6-OH) and 8.32 (1 H, s, N3-H). w w W Table 1 'H NMR resonances of compounds 310"" ~ ~~~ Compd. 5'-H 4'-H 3'-H 2'-H 1'-H 5-Me 6-H Others 3a 3.22 (1 H, dd, 4.63 (1 H, dd, 5.70 (1 H, dd, 6.10 (1 H, d, 1.74 (3 H, s) 5.16 (1 H, 7.51 (1 H, d, J6.40, Jgem 1 1 *O, J3t.2' 2.50, J2!,1* 6.50, J,,,2,6.50) d, J6.40) 6-OH), 7.25-7.36 J5ra,4, 5.0, 5'-Ha) J3,,4,4.5) J2?,3r 2.50) (ArH)3.27 (1 H, dd, Jgcm 11-0, J5flb.4, 7.20, 5'-Hb) 4a 2.96 (1 H, dd, 4.43 (1 H, ddd, 4.74 (1 H, dd, 5.70 (1 H, dd, 6.1 1 (1 H, d, 1.68 (3 H, s) 4.67 (1 H, 7.91 (I H, d, J 6.36,Jgem 3.97, J3rs2r 1.59, J2,,1, 5.75, J,r,2.11.12, J4,,3, 5.75) d,J6.36) 6-OH), 7.247.44 7.15, J3f.43J5'a,4' J4,,5,a 3.97) J23.3, 1.59) (ArH)7.15, 5'-Ha) J4,,5*b4.77) 3.20 (1 H, dd, Jge,11.12, J5fb,4, 4.77, 5'-Hb) 3b 3.18 (1 H, dd, 4.32 (1 H, m) 4.72 (1 H, dd, 5.77 (1 H, dd, 6.28 (1 H, d, 1.77 (3 H, s) 5.26 (1 H, s) 3.50 (3 H, s, 6-OMe), Jgem 10-8, J3,,2,2.0, J23.1' 5.4, J,.,,,5.4) 7.28-7.36 (15 H, J5ra,4, 5.6, 5'-Ha) J3V.4, 3.8) J2',3' 2.0) ArH)3.27 (1 H, dd, Jgem 10.8, J5Pb.4, 6.8, 5'-Hb) 4b 3.07 (1 H, dd, 4.45 (1 H, m) 4.81 (1 H, dd, 5.78 (1 H, dd, 6.23 (1 H, d, 1.79 (3 H, s) 5.36 (1 H, s) 3.41 (3 H, s, 6-OMe), Jgem 10.6, J3,*2,2.1, J2.,,.5.6, J1,,2,5.6) 7.25-7.38 (1 5 H, J5ra,4, 6.9, 5'-Ha) J3',4v4.3) J2f.3-2.1) ArH)3.19 (1 H, dd, Jgem 10.6,Js'b.4'4.6, 5'-Hb) 3c 3.19 (1 H, dd, 4.48 (1 H, ddd, 4.59 (1 H, dd, 5.36 (1 H, dd, 6.05 (1 H, d, 1.84 (3 H, s) 6.16 (1 H, s) 2.1 1 (3 H, S, 6-OAc),2.8, J3,,2, 1.4, J2,,1,Jgem 10.2, J4,,3, 5.6, J,cs2p 5.6) 7.27-7.37 (15 H, J5ca,4, 9.2, 5'-Ha) J4',5'a 9.23 J3r.4, 2.8) J2,,3* 1.4) ArH)3.45 (1 H, dd, J4',5'b 4-8) Jgem 10.2,J.jsb.4~4.8, 5'-Hb) 4c 3.07 (1 H, dd, 4.47 (1 H, ddd, 4.55 (1 H, dd, 5.41 (1 H, dd, 6.02 (1 H, d, 1.75 (3 H, s) 6.48 (1 H, s) 1.81 (3 H, S, 6-OAc),2.4, 5.0,Jgem 10.2, J4a,3# J33,2,0.8, J2!,18 J1#,2#5.0) 7.27-7.39 (15 H, J5ra,4, 8.0, 5'-Ha) J4'.5a 8.0, J3r,4,2.4) J2,,3,0.8) ArH)3.21 (1 H, dd, J4'.5'b 5.6)Jgcm10.2 J5pb,4s5.6, 5'-Hb) W \o P J.CHEM. soc. PERKIN TRANS. I 1994 2291 n x 3 W-c? Y 4 ah a a-? $ T 5 ;o 00 ElIn 2292 J. CHEM. SOC. PERKIN TRANS. 1 1994 Me 1 C" 0 a; R=H b; R=Me C; R=Ac 6r 6r 4c' 3a-c 4a-c 1 \ 0 0 TrO ""v-'"vBr Br ir 5 7R=Tr 6 8R=H /'c 0 Tr = Trityl 11 12 0 3a -7+8+ Br Br Br Br 9 10 Scheme 1 In contrast with these results, ourcompound 4a proceeded more smoothly at room temperature 2,2'-azoi~obutyronitrile.~ to give the (5R,6S)analogue 6 quantitatively. Similar synthesis initial trials of debromohydrination of compounds 3a and 4a to and reactions of 5,ti-brornohydrins of thymidine were 2,2'-anhydro-1-(3'-bromo-3'-deoxy-5-U-trityl-~-~-arabino-extensively studied recently by Yoneda and co-workers in furanosy1)thymine 7by these means were all unsuccessful, no connection with the oxidative damage and repair of pyrimidine change having been observed.Clearly, the present 2,2'-cyclized bases in DNA.' This research demonstrated the easy repair of form, 3a or 4a, is unable to generate a bromo radical.' 5,6-bromohydrins of 1,3-dimethylthymine as well as those of Deoxygenation of epoxides 5and 6 with the use of tri-thymidine by treatment with heat, sunlight or a radical initiator, phenylphosphine was therefore tried and gave compound 7in 2293J.CHEM. SOC. PERKIN TRANS. 1 1994 H Fig. 1 X-Ray molecular structure of compound 3a 0 0 Br hr 3a; R=H 4a b; R=Me 0 24% 4b Fig. 2 'H NMR NOE measurements on compounds 3a, b and 4a, b moderate yields. Selected experiments are given in the Experimental section [experiment (A-I) and (A-2)]. In view of the rather unacceptable yields of compound 7 from substrates 5 and 6, the common debromohydrination of compounds 3 by using zinc powder in dimethylformamide (DMF)-AcOH or DMF-MeOH (room temperature) was tried. However, this reagent proved to be inappropriate, since the unsaturated furanose 1 was reproduced, probably due to attack of zinc on the 3'-bromo group to cause eliminative ring opening (process 2 -1).It was finally found that compounds 3a and 4a can be smoothly converted into compound 7 with the use of a combination of Ph,P and sodium hydrogen carbonate. Thus, treatment of compound 3a with 1.9 mol equiv. of Ph,P in the presence of an excess of NaHCO, gave compound 7 in 80% yield [experiment (B-1)], while similar treatment of compound 4a gave compound 7 in 86% yield after a shorter reaction time [experiment (B-2)]. Application of the same reaction to an equimolar mixture of isomers 3a and 4a afforded compound 7 in 65% yield when a 3-fold excess of Ph,P was used [experiment (B-3)]. The reason for the retardation of the reaction and the rather low yield in this case is unclear at present.However, this point suggested a possible intermolecular association between isomers 3a and 4a through hydrogen bonding by the 6-hydroxy group of compound 3a or 4a, sterically disturbing the access of the bulky triphenylphosphine. Accordingly, we decided to prepare some analogues of the alcohols 3a and 4a carrying a 6-methoxy or 6-acetoxy group instead of the 6-hydroxy group and to examine their chemical behaviour toward triphenyl- phosphine. Additionally, the difference in electronegativity of alkoxy and ester oxygens might shed some light on the mechanism. Thus, compound 1was treated with 2 mol equiv. of NBA in methanol containing acetone to afford (5R,6R)-2,2'- anhydro-5-bromo-1-(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arab-inofuranosyl)-6-methoxy-5,6-dihydrothymine3b and (5S,6S)- 2,2'-anhydro-5-bromo-1-(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arabinofuranosyl)-6-methoxy-5,6-dihydrothymine4b in 48 and 36% yield, respectively.The UV and 'H NMR data of structural importance are quite similar to those for the parent alcohols 3a and 4a. The stereochemistry at C-5 and C-6 of the base was deduced from a comparison of the results of 'H NMR nuclear Overhauser effect (NOE) measurements of compounds 3a, b and 4a, b (Fig. 2): SR,6R-compounds 3a, b showed no NOE between 1'-H and 6-H, while SS,6S-cornpounds 4a and 4b showed the corresponding NOE of 4and 2-3%, respectively.A molecular-model study also indicated the closer proximity of 1'-H and 6-H in the (5S,6S) series. Similarly, treatment of compound 1with NBA in the presence of 2 mol equiv. of acetic acid gave (5R,6R)-6-acetoxy-5-bromo-2,2'-anhydro-1-(3'-bromo-3'-deoxy-5'-O-trityl-~-~-arabinofuranosyl)-5,6-dihydro-thymine 3c and its (5S,6S)-trans analogue 4c in similar yields. The latter compound was spectroscopically identical with a product obtained by acetylation of the alcohol 4a and hence compound 3c should have the (5R,6R)-trans structure. The notably deshielded 6-H signals in the 'H NMR spectra of these compounds confirmed the presence of the acetoxy group at C-6. Interestingly, the (SS,6S)-trans isomers 4a-c are all less polar than the corresponding 5R,6R counterparts 3a-c in TLC (see Experimental section).In the acetylation of the alcohol 4a, another product, 4c', having a polarity between those of acetates 3c and 4c was obtained. On the basis of elemental analysis, UV and 'H NMR data (Experimental section) compound 4c' seems to be a structural isomer of acetates 3c and 4c.* Attempted X-ray analysis failed owing to decomposition of compound 4c' on X-ray irradiation and hence its structural elucidation was abandoned. Debromomethoxylation of the ethers 3b and 4b with the use of Ph,P-NaHCO, proceeded extremely rapidly (10-20 min) to give an excellent yield of compound 7 in each case. Finally, one- pot synthesis of compound 7 from compound 1via the ethers 3b and 4b was tried [experiment (C-3)] to give compound 7 in 85% overall yield.Also, one-pot synthesis of compound 7 viu acetates 3c and 4c was successful [experiment (D-1)], giving compound 7 in 81% yield. Compound 7 could be easily deprotected to 2,2'- anhydro-1-(3'-bromo-3'-deoxy-~-~-arabinofuranosyl)thymine 8 with the use of boron trifluoride-diethyl ether or 80% acetic acid in contrast to compound 2, which tended to decompose or/and regenerate unsaturated substrate 1 when treated with 80% acetic acid.7 In these repair reactions, it is imperative to intercept the released hydrogen bromide molecules instantane- ously by NaHCO, or other appropriate bases. Otherwise, hydrogen bromide attacks the 2,2'-anhydro bridge to cause ring opening and/or 5'-deprotection.Thus, the reaction of compound 3a with Ph,P in the absence of sodium hydrogen carbonate gave 1-(2,3-dibromo-2,3-dideoxy-5-O-trityl-P-~-ribofuranosy1)thymine 9 as a major product, together with *On the basis of the upfield shift of the 6-H signal (S 5.98) of compound 4c' as compared with that of compounds 3c and 4c (86.05 and 6.02, respectively), we propose compound 4c' to be a trans-5-acetoxy-6-bromo analogue formed from acetate 4c through an ortho- ester intermediate. compound 8. Compound 9 was deprotected to crystalline 1- (2,3-dibromo-2,3-dideoxy-~-~-ribofuranosyl)thymine10 for analysis. Compound 7 was converted into the known 1-(2,3- anhydro-5-O-trityl-~-~-lyxofuranosyl)thymine11 7g10and its 2-0-methylthymine analogue 127 in 93 and 85% yield, respectively: these results are far better than those obtained starting from compound Z7 Experimental M.p.s were recorded on a Yanagimoto micro melting point apparatus and are uncorrected.UV spectra were measured on a JASCO Model V-560 spectrophotometer. The 200 MHz 'H NMR spectra of compounds 3b, c, 4b, c, 5-10 were recorded on a GEMINI-200 FT NMR spectrometer, and the 500 MHz 'H NMR spectra of compounds 3a, 4a and 4c' on a JEOL-JNM- GS500 FT NMR spectrometer in the laboratory of the Daiichi Seiyaku Co., Ltd. J-Values are in Hz. Elemental analyses were conducted using a Perkin-Elmer 240B elemental analyser. For preparative-scale thick-layer chromatography (PLC), glass plates coated with a 2 mm thick layer of Wakogel B-5F silica gel were used after activation at 100°C for 10-12 h.All evaporations were carried out under reduced pressure at or below 40 "C. Crystallography of Compound 3a.-A single crystal (0.12 x 0.20 x 0.68 mm), which was grown from acetone solution, was used for the unit-cell determinations and the data collections of a Rigaku AFC5S four-circle diffractometer, with graphite-monochromated Mo-Ka radiation (A = 0.710 69 A). This crystal was orthorhombic, space group P212,2,, Z = 4 with a = 14.918(8), b = 18.560(6), c = 10.962(6) A, V = 3035(2) A3, and Dcalc= 1.533 g cm-,. The computer program used was TEXSAN," the structure was solved by direct methods (MITHRIL),', and the non-hydrogen atoms were refined isotropically. The hydrogen atoms were included in the structure-factor calculations in idealized positions.The final R-factor after the full-matrix least-squares refinement was 0.09 1 for 1084 observed reflections. The PLUTO ' drawing of compound 3a is shown in Fig. 1. (SR,6R)-2,2'-Anhydro-5-bromo-1-(3'-bromo-3'-deoxy-5'-0-trityl-~-~-arabinofuranosyl)-6-hydroxy-5,6-dihydrothymine3a J. CHEM. SOC. PERKIN TRANS. 1 1994 infl.) and 250.3 (10400, infl.) (Found: C, 53.6; H, 4.4; N, 4.1. C,,H,,Br,N,O,-MeOH requires C, 53.43; H, 4.48; N, 4.15%). (SR,6R)-2,2'-Anhydro-5-bromo-1-(3'-bromo-3'-deoxy-5'-0-trityl-~-~-arabinofuranosyl)-6-methoxy-5,6-dihydrothymine3b and (SS,6S)-2,2'-Anhydro-5-bromo-1-(3'-bromo-3'-deoxy-5'-0-trityl-P-D-arabinofuranosyl)-6-methoxy-5,6-dihydrothymine 4b.-To a stirred, ice-cold solution of compound 1 (1.O g, 2.1 mmol) in a mixture of acetone (4 cm3) and MeOH (1 cm3) was added NBA (580 mg, 4.2 mmol).After 4 h, the mixture was stirred at room temperature for 2 h. TLC monitoring at this stage showed that the starting material was consumed and two less polar products had been formed. The mixture was evaporated and the residue was partitioned between EtOAc (30 cm3) and water (10 cm3). The separated organic layer was dried over sodium sulfate and evaporated, and the residue was fractionated on 2 sheets of silica plates [20 x 20 cm; CHC1,- EtOAc (5 :l), developed twice] to give, from the polar band compound 3b (660 mg, 48%) as crystals, m.p. 214-215 "C (from EtOAc): A,,,(MeOH)/nm (E) 230.4 (14 000, infl.) and 245.6 (9600, infl.) (Found: C, 55.1; H, 4.3; N, 4.05.C,OH,,Br,N,O, requires C, 54.89; H, 4.30; N, 4.27%). Elution of the less polar fraction with acetone gave compound 4b as a homogeneous foam (495 mg, 36%); &,,,,(MeOH)/nm (E) 230.6 (1 1 000, infl.) and 246.4 (6400, infl) (Found: C, 54.93; H, 4.53; N, 4.05%). (5R,6R)-6-Acetoxy-2,2'-anhydro-5-bromo-1-(3'-bromo-3'-de-oxy-5'-O-trityl-~-~-arabinofuranosyl)-5,6-dihydrothymine3c and (5S,6S)-6-Acetoxy-2,2'-anhydro-5-bromo-l-(3'-bromo-3'-deoxy-5'-0-trityl-~-~-arabinofuranosyl)-5,6-dihydrothymine4c. -NBA (1.1 1 g, 8.06 mmol) was added to a stirred, ice-cooled solution of compound 1 (2.0 g, 4.30 mmol) in a mixture of acetone (15 cm3) and acetic acid (0.50 cm3, 8.77 mmol).After 5 h, more NBA (160 mg, 1.16 mmol) was added. After 23 h, further NBA (106 mg, 0.77 mmol) was added (total 1.38 g, 9.99 mmol) and the mixture was stirred for an additional 25 h under ice-cooling. The solvent was evaporated off and the residue was partitioned between CHCl, (40 cm3) and water (10 cm3). The separated organic layer was dried over sodium sulfate and evaporated, and the residue was chromatographed on a silica gel column (3.5 x 40 cm) with CHC1,-EtOAc (5:l) to give, from the faster running fractions, compound 4c(1.032 g, 35.1%) and (5S,6S)-2,2'-Anhydro-5-bromo-l-(3'-bromo-3'-deoxy-5'-0-as crystals, m.p. 180-1 82 "C (after recrystallization from trityl-~-~-arabinofuranosyl)-6-hydroxy-5,6-dihydrothymine EtOAc); A,,,(MeOH)/nm (E) 230.4 (13 500, infl.) and 244.8 4a.-To a stirred, ice-cold solution (or suspension) of com-pound 1 (2.0 g, 4.28 mmol) in a mixture of acetone (8.0 cm3) and water (2.0 cm3) was added NBA (717 mg, 5.2 mmol). After the mixture had been stirred for 5 h at 0 "C, more NBA (700 mg, 5.1 mmol) was added and the mixture was stirred at 0 "C for a further 35 h to give a TLC-pure solid precipitate, which was more polar than substrate 1 in TLC [silica; CHC1,-EtOAc (1 :l)]. The solid was collected by suction and was recrystallized from MeOH to give compound 3a (1.45 g, 5373, which gradually decomposed at between 145 and 167 "C; ;1,,,(MeOH)/nm (E) 233.7 (6000, infl.) and 250.4 (3500, infl.) (Found: C, 54.0; H, 4.2; N, 4.4.C,9H,,Br,N,0, requires C, 54.22; H, 4.08; N, 4.36%). TLC monitoring of the filtrate separated from the solid showed another, less polar, product together with a negligible amount of R,R-compound 3a and no starting material. The filtrate was evaporated and the residue was partitioned between EtOAc (30 cm3) and water (10 cm3). The separated organic layer was dried over sodium sulfate, and evaporated, and the residue was left with MeOH (4 cm3) to give TLC-homogeneous crystals, which were collected, and recrystallized from MeOH to afford compound 4a (914 mg, 32%), which gradually melted between 120 and 129 "C; A,,,(MeOH)/nm (E) 232.2 (15 200, (9300, infl.); v,,,(KBr)/cm-' 1770 (6-acetoxy) (Found: C, 54.4; H, 4.3; N, 3.9. C31H28Br2N206-$EtOA~ requires C, 54.41; H, 4.33; N, 3.93%).The slower running fractions gave compound 3c (1.037 g, 35.3%) as crystals, m.p. 212-213 "C (after recrystallization from MeOH at room temperature); A,,,(MeOH)/nm (E) 23 1.2 (17 500, infl.) and 243.6 (12 600, infl.); v,,,(KBr)/cm-' 1780 (Found: C, 54.3; H, 4.1; N, 4.1. C,,H,,Br,N,O, requires c, 54.40; H, 4.12; N, 4.09%). Acetylation of Alcohol 4a.-To a solution of compound 4a (methanolate) (200 mg, 0.3 mmol) in pyridine (1 cm3) was added acetic anhydride (0.13 cm3, 1.4 mmol) and the mixture was stirred at room temperature overnight. TLC monitoring [silica; CHC1,-EtOAc (3 :1 or 5 :l)] showed the formation of two less polar, major products, the faster running of which coincided with acetate 4c in terms of TLC mobility. The mixture was treated with MeOH (1 cm3) at room temperature for 1 h, evaporated, and repeatedly co-evaporated with MeOH.The residue was partitioned between EtOAc (15 cm3) and water (5 cm3). The separated organic layer was dried over sodium sulfate and evaporated, and the residue was fractionated on a silica plate [20 x 20 cm; CHC1,-EtOAc (3:1), developed 3 J. CHEM. SOC. PERKIN TRANS. I 1994 times] to give, from the least polar fraction, compound 4c (63 mg, 29.8%), identical with an authentic sample in terms of IR and UV spectroscopic data. The more polar fraction gave compound4c' (80 mg, 39.5%) as crystals, m.p. 214-216 "C (after recrystallization from MeOH at room temperature); I,,,(MeOH)/nm (E) 229 (1 3 400); G,(CDCl,) 1.68 (3 H, s, 5-Me), 2.18 (3 H, s, AcO), 3.07 (1 H, dd, Jgem 10.33, J5ta,4r7.15,5'-Ha), 3.28 (1 H, dd, Jgem10.33,J5,,,4* 4.77, 5'-Hb), 4.50 (1 H, complex m, 4'-H), 4.60 (1 H, m, 3'-H), 5.40(1 H, dd, J2s.3, 1.59, J2*,1,4.77,2'-H), 5.98(1 H,s,6-H), 5.99 (1 H, d, J19,2f4.77, 1'-H) and 7.24-7.37 (1 5 H, m, ArH) (Found: C, 54.4; H, 4.1;N, 4.1%).(SS,6R)-2,2'-Anhydro-1 -(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arabinofuranosyl)-5,6-epoxy-5,6-dihydrothymine5.-A mixture of the alcohol 3a (500 mg, 0.78 mmol) and triethylamine (0.50 cm3, 3.6 mmol) in acetone (8 cm3) was heated to reflux for 5 h. After cooling, the solid precipitate of triethylamine hydro- bromide was filtered off and the filter-cake was washed with a small volume of EtOAc.The filtrate was evaporated to give a gum, which was subjected to PLC [silica; 20 x 20 cm; CHC1,- EtOAc (3:l), developed twice] to afford epoxide 5 (358 mg, 82%) as a homogeneous foam; A,,,(MeOH)/nm (E) 232 (18 500, sh) (Found: C, 62.25; H, 4.5; N, 4.7. C,,H,,BrN,O, requires C, 62.04; H, 4.49; N, 4.99%). (SR,6S)-2,2'-Anhydro-1 -(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arabino furanosyl)-5,6-epoxy- 5,6-dihydrothymine 6.-A mixture of the alcohol 4a (623 mg, 0.92 mmol) and triethylamine (0.58 cm', 4.2 mmol) in acetone (12 cm3) was stirred at room temp. for 1 h. The precipitate of triethylamine salt was filtered off and the filtrate was evaporated. The residue was partitioned between EtOAc (1 3 cm3) and water (3 cm3). The separated organic layer was treated with Norit and evaporated to give a crystalline solid, which was again dissolved in acetone and then the solvent was removed under reduced pressure to afford a foam.Trituration of the foam with MeOH (4 cm3) gave pure crystals (453 mg). A further crop (58 mg) of compound 6 was obtained from the filtrate (total 511 mg, 99%); A,,,(MeOH)/nm (E) 231.6 (8400, sh) (Found: C, 62.0;H, 4.5;N, 5.0%). 2,2'-Anhydro-1 -(3'-bromo-3'-deoxy-5'-O-trityl-P-~-arabino-juranosy1)thymine 7.-(A-2) Deoxygenation of compound 5. To a solution of compound 5 (358 mg, 0.64 mmol) in EtOAc (8 cm3) was added Ph,P (184 mg, 0.70 mmol) and the mixture was heated at 80°C for 20 h. After cooling, the mixture was evaporated and the residue was fractionated on a silica plate [20 x 20 cm; CHC1,-EtOAc (3 :l)] to give, from the major band, compound 7 (166 mg, 48%) as crystals, m.p. 2 10-2 12 "C; imaX(MeOH)/nm(E) 225 (16 200, infl.) and 253 (7600, infl.) (Found: C, 63.9; H, 4.5; N, 5.2.C2,H,,BrN204 requires C, 63.86;H, 4.62;N, 5.14%). (A-2)Deoxygenation of compound 6.A mixture of compound 6 (200 mg, 0.36 mmol) and Ph,P (1 2 1 mg, 0.46 mmol) in EtOAc (5 cm3) was heated to reflux for 14.5 h, during which time substrate 6 disappeared. The mixture was concentrated, and fractionated on a silica plate [20 x 20 cm; CHC1,-EtOAc (l:l), developed twice] to give, from the major band, compound 7 (58 mg, 30%), identical with the above obtained product 7 in terms of IR, UV and 'H NMR data.(B-1)Debromohydrination of compound 3a. To a vigorously stirred mixture of compound 3a (642.4mg, 1 mmol) and finely pulverized NaHCO, (227 mg, 2.7 mmol) in DMF (10 cm3) was added Ph,P (510 mg, 1.94 mmol). After 3.5 h, TLC showed 3 spots corresponding to Ph,P=O, compound 7 and Ph,P. The inorganic salts were filtered off and the filtrate was evaporated. The residue in CHCl, (10 cm3) was filtered with Norit, the filtrate was concentrated, and the residue was fractionated on silica plates [20 x 20 cm, 2 sheets; CHC1,-EtOAc (l:l), developed 3 times] to give, from the most polar fraction, compound 7 (436 mg, 80%), identical with an authentic specimen in every respect. (B-2) Debromohydrination of compound 4a. To a vigorously stirred mixture of compound 4a (methanolate) (100 mg, 0.148 mmol) and finely ground NaHCO, (34 mg, 0.4 mmol) in DMF (1.5 cm3) was added Ph,P (58 mg, 0.22 mmol).After 70 min, the mixture was worked up as in experiment (B-1) to give compound 7 (69 mg, 86%) after PLC [silica, 20 x 20 cm; CHCl3-EtOAc (3 l)]. (B-3)Debromohydrination of a mixture of alcohols 3a and 4a. Triphenylphosphine (1 15 mg, 0.44mmol) was added to a stirred mixture of compounds 4a (methanolate) (100 mg, 0.148 mmol), 3a (95.1 mg, 0.148 mmol) and fine NaHCO, powder (67.2 mg, 0.8 mmol) in DMF (3.0cm3). After 4 h, the inorganic salts were filtered off and the filtrate was evaporated. The residue was partitioned between EtOAc (20 cm3) and water (7 cm3). The separated organic layer was dried over sodium sulfate and concentrated, and the residue was fractionated on a silica plate [20 x 20 cm; CHC1,-EtOAc (1: l)] to give compound 7 (105 mg, 65%), identical with the above obtained authentic sample.(C-2) Debromomethoxylation of the ether 3b. A vigorously stirred mixture of compound 3b (200 mg, 0.305 mmol) and powdered NaHCO, (69 mg, 0.82 mmol) in DMF (3.1 cm3) was treated with Ph,P (152 mg, 0.58 mmol). After 10 min, the mixture was worked up as above to give compound 7 (142 mg, 85.3%) after PLC [silica, 20 x 20 cm; CHC1,-EtOAc (3 :l)]. (C-2) Debromomethoxylation of the ether 4b. A stirred mixture of compound 4b (300 mg, 0.46 mmol) and finely pulverized NaHCO, (77 mg, 0.92 mmol) in DMF (4.0cm3) was treated with Ph,P (1 80 mg, 0.69 mmol) under argon for 20 min ,and was then worked up as above to give compound 7 (202mg, 81%).(C-3)One-pot synthesis from compound 1 via ethers 3b and 4b.A mixture of compound 1 (200 mg, 0.42 mmol) and NBA (1 16 mg, 0.84 mmol) in a mixture of EtOAc (0.8 cm3) and MeOH (0.1 cm3) was stirred overnight. TLC monitoring indicated that substrate 1 was completely consumed and that compounds 3b and 4b had formed. After addition of EtOAc (2 cm3) and powdered NaHCO, (106 mg, 1.26 mmol), Ph,P (1 32 mg, 0.50 mmol) was added in 3 portions to the vigorously stirred mixture. After 1 h, the inorganic material was filtered off. The filtrate was directly partitioned between EtOAc (20 cm3) and water (5 cm3), and the separated organic layer was worked up as above to give compound 7 (195 mg, 85.2%) after PLC [silica, 20 x 20 cm; CHC1,-EtOAc (I :l)].(D-2) One-pot synthesis from compound 1 via acetates 3c and 4c.To a stirred, ice-cold solution of compound 1 (500 mg, 1.07 mmol) and acetic acid (0.3cm3, 2.58 mmol) in acetone (10 cm3) was added NBA (350 mg, 2.54 mmol). The reaction mixture was allowed to warm up gradually to room temperature during 4 h before being evaporated, and then co-evaporated with acetone a couple of times, and the residue was taken up into EtOAc (15 cm3). Finely powdered NaHCO, (500 mg, 5.95 mmol) was added and the mixture was vigorously stirred to remove the residual acetic acid. After 30 min, Ph,P (400 mg, 1.53 mmol) was added and the mixture was stirred for another 1.5 h.The inorganic material was filtered off and the filtrate was evaporated. The residue was fractionated on a silica plate [20 x 20 cm; CHC1,-EtOAc (3 :l)] and the most polar fraction was eluted with acetone to give compound 7 (472 mg, 81%) after crystallization from MeOH. 2,2'-Anhydro-1-(3'-br.omo-3'-deoxy-P-~-arabinofuranosy1)-thymine &-(A). To a stirred solution of compound 7 (500 mg, 0.917 mmol) in chloroform (20 cm3) was added boron trifluoride-diethyl ether (0.06 cm3, 0.48 mmol). After 2 h, more boron trifluoride-diethyl ether (0.02 cm', 0.16 mmol) was added and the mixture was stirred for additional 1 h. The mixture was thoroughly evaporated and the residue, dissolved in acetone, was applied to a silica plate (20 x 20 cm).After development with CHC1,-MeOH (85 :19, the desired fraction was eluted with acetone and recrystallized from the same solvent to give compound 8 (187 mg, 67.3%), m.p. 224-226 "C; R,,,(MeOH)/ nm (E) 229.0 (2900) and 252.4 (5800) (Found: C, 39.75; H, 3.5; N,9.3. C,,H, ,BrN,O, requires C, 39.62; H,3.60; N,9.24%). (B).A solution of compound 7 (200 mg, 0.37 mmol) in 80% acetic acid (40 cm3) was left at room temperature for 40 h and was then evaporated below 35 "C. The residue was repeatedly co-evaporated with MeOH to remove the residual acetic acid. The obtained solid residue was digested with diethyl ether (20 cm3) and the sparingly soluble solid was collected by suction. The filter-cake was washed with diethyl ether (5 x 2 cm'), dried over silica gel under high vacuum, and recrystallized from acetone to give compound 8 (76 mg, 68%), identical with the above obtained product in terms of mixed m.p.and IR spectroscopy. Debromohydrination of Compound 3a in the Absence of NaHCO,.-To a suspension of the alcohol 3a (200 mg, 0.31 mmol) in EtOAc (8 cm3) was added Ph3P (87 mg, 0.33 mmol). The mixture gradually became clear. after 3 h, the mixture was concentrated and subjected to PLC [silica, 20 x 20 cm; CHC1,- MeOH (9 :l), developed twice] to give, from the most polar fraction, a tiny amount of compound 8, which was identical with an authentic sample by IR and UV spectroscopy after crystallization from acetone. The intermediate fraction was eluted with acetone to give compound 7 (23 mg, 14%), identical with an authentic sample after recrystallization from EtOAc.The highly mobile fraction was again fractionated on a silica plate [lo x 20 cm; CHC1,-EtOAc (3: l)] to give dibromide 9 (85 mg, 44%) as a homogeneous foam; A,,,(MeOH)/nm 266 (E 2600). Compound 9 was analysed after deprotection to compound 10. 2',3'-Dibromo-3'-deoxythymidine 10.-A stirred solution of compound 9(1 11 mg, 0.18 mmol) in CHCl, (5 cm3) was cooled to -10 "C and boron trifluoride-diethyl ether (0.01 cm', 0.08 mmol) was added. After the mixture had been left at 0 "C for 20 h, the precipitate was collected by suction, washed with CHCl,, and recrystallized from MeOH to give compound 10 (40 mg, 5973, m.p. 206208 "C; R,,,(MeOH)/nm 265 (E 12 100) (Found: C, 31.3; H, 3.7; N, 6.95. ClOHl2Br2N2O4- $MeOH requires C, 31.52; H, 3.53; N, 7.00%).Conversion of Bromide 7 into 1-(2,3-Anhydro-5-0-trityl-P-~-1yxofuranosyl)thymine 11.-To a stirred solution of compound 7 (150 mg, 0.275 mmol) in acetone (1.4 cm3) was added 1 mol dm-, NaOH (0.63 cm3, 0.63 mmol). After 20 min, the mixture was neutralized with 1 mol dmP3 AcOH-EtOH and evaporated, and the residue was partitioned between EtOAc (20 cm3) and water (7 cm3). The separated organic layer was dried over sodium sulfate and concentrated to give compound 11(146 mg, 93%) as the mono ethyl acetate solvate, identical with an authentic sample in terms of IR and UV spectroscopic data. Conversion of Bromide 7 into 1-(2,3-Anhydro-5-O-trityl-P-~-lyxofuranosyl)-2-O-methylthymine12.-To a stirred solution of J.CHEM. SOC. PERKIN TRANS. 1 1994 compound 7 (200 mg, 0.366 mmol) in a mixture of acetone (1 cm3) and MeOH (1 cm3) was added NaOMe (60 mg, 1.1 mmol). After 2 h, further NaOMe (50 mg, 0.93 mmol) was added (total 2.03 mmol). After 2 h, TLC monitoring [silica; CHC1,- MeOH (9:1)] showed that the starting material had disappeared and that a less polar product had formed with a negligible amount of a faster running by-product. The total product was neutralized with 1 mol dm-, AcOH-EtOH, the mixture was evaporated, and the residue was partitioned between EtOAc (30 cm3) and water (10 cm3). The separated organic layer was dried over sodium sulfate and evaporated, and the residue was fractionated on a silica plate [20 x 20 cm; CHC1,-MeOH (9:1), developed twice].Elution of the major band with acetone gave compound 12 (155 mg, 85%) as a foam, identical with an authentic sample7 in every respect. References 1 H. Mitsuya and S. Broder, Proc. Natl. Acad. Sci. USA, 1986,83,1911. 2 H. Mitsuya, K. J. Weinhold, P. A. Furman, M. H. St. Clair, S. N. Lehrman, R. C. Gallo, D. Bolognesi, D. W. Barry and S. Broder, Proc. Natl. Acad. Sci. USA, 1985,82, 7096. 3 K. E. Pfitzner and J. G. Moffatt, J. Org. Chem., 1964, 29, 1508; M. J. Robins and J. S. Wilson, J. Am. Chem. Soc., 1981, 103, 932; R. A. Lessor and N. J. Leonard, J. Org. Chem., 1981, 46, 4300; D. G. Norman and C. B. Reese, Synthesis, 1983, 304; M. J. Robins, J.S. Wilson and F. Hansske, J, Am. Chem. SOC., 1983, 105, 4059; B. Doboszewski, C. K. Chu and H. V. Halbeek, J. Org. Chem., 1988, 53,2777; V. Nair and G. S. Buenger, J. Am. Chem. SOC., 1989,111, 8502; H. Rosemeyer and F. Seela, Helv. Chim. Acta, 1989,72, 1084; P. Serafinowski, Synthesis, 1990, 41 1; M. Sekine and T. Nakanishi, J. Org. Chem., 1990, 55, 924; S. Czernecki and J.-M. Valery, Synthesis, 1991,239. 4 D. M. Huryn and M. Okabe, Chem. Rev., 1992,92, 1745. 5 T. Nishimura and B. Shimiz, Chem. Pharm. Bull., 1965, 13, 803; K. H. Scheit, Chem. Ber., 1966, 99, 3884; K. A. Watanabe and J. J. Fox, J. Heterocycl. Chem., 1969, 6, 109. 6 J. J. Fox, N. Yung and A. Bendich, J. Am. Chem. Soc., 1957,79,2775; I. L. Doerr, J. F. Codington and J. J. Fox, J. Med. Chem., 1967, 10, 247; C. Nakayama, H. Machida and M. Saneyoshi, J. Carbohydr., Nucleosides, Nucleotides, 1979,6, 295. 7 K. Minamoto, Y. Hamano, Y. Matsuoka, K. Watanabe, T. Hirota and S. Eguchi, Nucleosides, Nucleotides, 1992, 11,457. 8 J. P. Horwitz, J. Chua, M. A. Da Rooge, M. Noel and I. L. Klundt, J. Org. Chem., 1966,31, 205. 9 T. Harayama, R. Yanada, M. Tanaka, T. Taga, K. Machida and F. Yoneda, J. Chem. SOC., Perkin Trans. I, 1988,2555; R. Yanada, T. Akiyama, T. Harayama, K. Yanada, H. Meguri and F. Yoneda, J. Chem. SOC.,Chem. Commun., 1989,238. 10 M. Ashwell, A. S. Jones and R. T. Walker, Nucleic Acids Rex, 1987, 15, 2157; J.-T. Huang, L.-C. Chen, L. Wang, M.-H. Kim, J. A. Warshaw, D. Armstrong, Q.-Y. Zhu, T.-C. Chou, K. A. Watanabe, J. Matulic-Adamic, T.-L. Su, J. J. Fox, B. Polsky, P. A. Barton, J. W. M. Gold, W. D. Hardy and E. Zuckerman, J. Med. Chem., 1991,34, 1640. 11 TEXSAN TEXRAY, Structure Analysis Package, Molecular Structure Corporation, 1985. 12 C. J. Gilmore, J. Appl. Crystallogr., 1984, 17,42. 13 S. Motherwell and W. Clegg, PLUTO Program for plotting molecular and crystal structures, University of Cambridge, England, 1985. Paper 4/0 1 3991 Received 9th March 1994 Accepted 9th May 1994