Investigation of 8-Aza-7-Deaza Purine Nucleoside Derivatives

Glycosylation of 6-amino-4-methoxy-1H-pyrazolo[3,4-d]pyrimidine and its iodo- and bromo- analogues with the protected ribofuranose and 2′-deoxyribofuranose under different conditions resulted in the synthesis of N9- and N8-glycosylated purine nucleosides. Five key intermediate nucleosides, having 6-methoxy, 7-iodo, and 2-bromo groups, were further derivatized to 23 final 8-aza-7-deazapurine nucleoside derivatives. The structures of N9- and N8-glycosylated products were assigned based on UV and NMR spectra. HMBC analysis of 2D NMR spectra and X-ray crystallographic studies of the representative compounds unambiguously verified the connection of ribose ring to N9- or N8-position of the purine ring. The anticancer activity of these new compounds was evaluated.

Various 8-aza-7-deaza purine nucleosides have been synthesized in the past fewer years. However, there still remains a space for the design and development of new nucleoside-based therapeutics. We decided to investigate 8-aza-7-deaza purine nucleoside derivatives modified at the positions 6 and 7 of the purine ring. Herein, we report the synthesis of 23 modified 8-aza-7-deaza purine nucleoside

Synthesis of 8-Aza-7-deazapurine Ribo-Nucleosides
6-Amino-4-methoxy-1H-pyrazolo [3,4-d]pyrimidine (purine base A) [16], 6-amino-3-iodo-4methoxy-1H-pyrazolo [3,4-d]pyrimidine (purine base B) [17], and 3,6-dibromo-1H-pyrazolo [3,4d]pyrimidin-4(5H)-one (purine base C) [18] (Scheme 1) were made utilizing reported protocols, and selected as the heterocyclic bases of 8-aza-7-deazapurine nucleosides to be studied. These purines were chosen for a number of advantages including the switch in the nitrogen atom from position 7 to 8, which changes the binding mode and strength of purine nucleosides in the duplex nucleic acids and other biological systems, substantially altering the biological properties and application. The 2amino-group on both 8-aza-7-deazapurine bases A and B does not need to be protected during the glycosylation process. The 6-methoxy group on purine bases A and B is selected to serve as a protecting group during glycosylation process. It is stable enough not to be cleaved in subsequent processes while active enough to be substituted by strong nucleophiles, therefore, converting to the desired corresponding 2-amino-8-aza-7-deazaguanosine derivatives when needed. The 6-oxo on purine base C does not need to be protected for glycosylation. The iodination of purine bases A to B adds high versatility for further derivatization on position 7 via Heck, Stille, Suzuki, Sonogashira, and related reactions. This will expand research in the related fields. To the best of our knowledge,  Figure 1), 2-amino-6-methoxy/oxo-7-substituted purine ribonucleosides 4 and 14-17 ( Figure 1) and 2-amino-6-substituted 2′-deoxy nucleosides 18-23 ( Figure  2). Different glycosylation protocols were studied for N 9 -and N 8 -glycosylation. In addition, the glycosylation sites and anomeric configuration of the newly synthesized compounds were investigated and assigned on the basis of 1 H-NMR, 13 C-NMR, 2D NMR, UV spectra, and X-ray crystallographic analysis of the representative compounds (see Supplementary Materials). The anticancer activity of these compounds was also evaluated.

Synthesis of 8-Aza-7-deazapurine Ribo-Nucleosides
6-Amino-4-methoxy-1H-pyrazolo [3,4-d]pyrimidine (purine base A) [16], 6-amino-3-iodo-4methoxy-1H-pyrazolo [3,4-d]pyrimidine (purine base B) [17], and 3,6-dibromo-1H-pyrazolo [3,4d]pyrimidin-4(5H)-one (purine base C) [18] (Scheme 1) were made utilizing reported protocols, and selected as the heterocyclic bases of 8-aza-7-deazapurine nucleosides to be studied. These purines were chosen for a number of advantages including the switch in the nitrogen atom from position 7 to 8, which changes the binding mode and strength of purine nucleosides in the duplex nucleic acids and other biological systems, substantially altering the biological properties and application. The 2amino-group on both 8-aza-7-deazapurine bases A and B does not need to be protected during the glycosylation process. The 6-methoxy group on purine bases A and B is selected to serve as a protecting group during glycosylation process. It is stable enough not to be cleaved in subsequent processes while active enough to be substituted by strong nucleophiles, therefore, converting to the desired corresponding 2-amino-8-aza-7-deazaguanosine derivatives when needed. The 6-oxo on purine base C does not need to be protected for glycosylation. The iodination of purine bases A to B adds high versatility for further derivatization on position 7 via Heck, Stille, Suzuki, Sonogashira, and related reactions. This will expand research in the related fields. To the best of our knowledge,

Synthesis of 8-Aza-7-deazapurine Ribo-Nucleosides
(purine base B) [17], and 3,6-dibromo-1Hpyrazolo [3,4-d]pyrimidin-4(5H)-one (purine base C) [18] (Scheme 1) were made utilizing reported protocols, and selected as the heterocyclic bases of 8-aza-7-deazapurine nucleosides to be studied. These purines were chosen for a number of advantages including the switch in the nitrogen atom from position 7 to 8, which changes the binding mode and strength of purine nucleosides in the duplex nucleic acids and other biological systems, substantially altering the biological properties and application. The 2-amino-group on both 8-aza-7-deazapurine bases A and B does not need to be protected during the glycosylation process. The 6-methoxy group on purine bases A and B is selected to serve as a protecting group during glycosylation process. It is stable enough not to be cleaved in subsequent processes while active enough to be substituted by strong nucleophiles, therefore, converting to the desired corresponding 2-amino-8-aza-7-deazaguanosine derivatives when needed. The 6-oxo on purine base C does not need to be protected for glycosylation. The iodination of purine bases A to B adds high versatility for further derivatization on position 7 via Heck, Stille, Suzuki, Sonogashira, and related reactions. This will expand research in the related fields. To the best of our knowledge, the direct glycosylation of 2-amino-8-aza-7-deazapurine bases A and B with routinely used l-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose (ribose I) has not been reported.
We would like to develop a high-yield protocol that would lead to the large-scale synthesis of the key intermediates and diversified final 8-aza-7-deazaguanosine derivatives. We performed the reported protocol [10] using boron trifluoride ether solution as the Lewis acid promotor at an elevated temperature. The glycosylation of ribose I with purine base A in acetonitrile worked excellently. However, the N 8 -glycosylated product 25 was isolated in 83.3% high yield, and almost no N 9glycosylated compound 24 was observed under this condition (Scheme 1). The N 9 -glycosylated product 24 was the major product under room temperature glycosylation condition while the N 8glycosylated product 25 was the thermodynamically dominant product formed at an elevated temperature. Therefore, two different compounds were made under different conditions from the same materials. A number of glycosylation protocols have been routinely used for the synthesis of different classes of nucleoside derivatives [10,19,20]. The silyl-Hilbert-Johnson glycosylation processes of silylated nucleobases with protected ribosides under Lewis acid conditions are very efficient for most cases. We glycosylated purine base A with ribose I (Scheme 1) under historic stannic chloride condition with no success. We then switched to the stronger promoter trimethylsilyl triflate (TMSOTf) and also explored silylating agents hexamethyldisilazane (HMDS) and N,O-bis(trimethylsilyl)acetamide (BSA) in acetonitrile, nitromethane and 1,2-dichloroethane. It was found that the HMDS-silylated purine base A was glycosylated with ribose I using TMSOTf in freshly distilled 1,2-dichloroethane overnight at room temperature giving the best result. The N 9 -glycosylated compound 24 (Scheme 1) was isolated in 30.3% yield as the major product, and only very trace amount of N 8 -glycosylated compound 25 was observed on TLC but could not be isolated.
We would like to develop a high-yield protocol that would lead to the large-scale synthesis of the key intermediates and diversified final 8-aza-7-deazaguanosine derivatives. We performed the reported protocol [10] using boron trifluoride ether solution as the Lewis acid promotor at an elevated temperature. The glycosylation of ribose I with purine base A in acetonitrile worked excellently. However, the N 8 -glycosylated product 25 was isolated in 83.3% high yield, and almost no N 9 -glycosylated compound 24 was observed under this condition (Scheme 1). The N 9 -glycosylated product 24 was the major product under room temperature glycosylation condition while the N 8 -glycosylated product 25 was the thermodynamically dominant product formed at an elevated temperature. Therefore, two different compounds were made under different conditions from the same materials.
Next, we set out to study the glycosylation of purine base B, which introduces a 7-iodo moiety that enables further derivatization. The glycosylation of iodo-purine base B with ribose I under both SnCl 4 and TMSOTf conditions resulted in the complicated mixtures, and only trace amount of N 9 -glycosylated product 26 could be detected. We then utilized boron trifluoride-ether as the catalyst for this glycosylation in acetonitrile at room temperature. The desired N 9 -glycosylated product 26 was isolated in about 28% yield (Scheme 1). The purine base C, without 6-oxo-protection, was also glycosylated with ribose I, resulting in the protected nucleoside 27. The two bromo-atoms on this molecule are expected to have different reactivity allowing for chemo-selective functionalization towards the synthesis of novel nucleoside derivatives.
The glycosylated products 24, 25, and 26 thus obtained were deprotected giving the corresponding 8-aza-7-deazapurine nucleosides 1-3. These 6-methoxy derivatives were heated with ammonium hydroxide solution overnight providing the corresponding 2-amino-8-aza-7-deazaadenosine derivatives 5, 8, and 11 in excellent yields (Scheme 2). The 6-methoxy compounds 1, 2, and 3 were also treated with aqueous potassium hydroxide solution to undergo deprotection, yielding the corresponding 8-aza-7-deazaguanosine derivatives 7, 9, and 10. Similarly, key intermediates 1 and 2 were heated with hydrazine solution to give the corresponding 6-hydrozino-purine nucleosides 6 and 12. The 6-hydroxyamino purine nucleoside 13 was obtained by treatment of compound 2 with hydroxylamine at 60 • C. Therefore, the 6-methoxy group was effectively used as the protecting group during the glycosylation process, and it can be further substituted with various nucleophiles to form 6-modified purine nucleoside derivatives. When compound 27 was deprotected with NH 3 /MeOH at 120 • C, the bromo-atom at position 2 was substituted with NH 2 group during the process, resulting in the corresponding 7-bromo-8-aza-7-deazaguanosine analogue 4. Therefore, the bromo-atom at position 2 is more reactive against nucleophile than the bromo-atom at position 7. Next, we set out to study the glycosylation of purine base B, which introduces a 7-iodo moiety that enables further derivatization. The glycosylation of iodo-purine base B with ribose I under both SnCl4 and TMSOTf conditions resulted in the complicated mixtures, and only trace amount of N 9glycosylated product 26 could be detected. We then utilized boron trifluoride-ether as the catalyst for this glycosylation in acetonitrile at room temperature. The desired N 9 -glycosylated product 26 was isolated in about 28% yield (Scheme 1). The purine base C, without 6-oxo-protection, was also glycosylated with ribose I, resulting in the protected nucleoside 27. The two bromo-atoms on this molecule are expected to have different reactivity allowing for chemo-selective functionalization towards the synthesis of novel nucleoside derivatives.
The glycosylated products 24, 25, and 26 thus obtained were deprotected giving the corresponding 8-aza-7-deazapurine nucleosides 1-3. These 6-methoxy derivatives were heated with ammonium hydroxide solution overnight providing the corresponding 2-amino-8-aza-7deazaadenosine derivatives 5, 8, and 11 in excellent yields (Scheme 2). The 6-methoxy compounds 1, 2, and 3 were also treated with aqueous potassium hydroxide solution to undergo deprotection, yielding the corresponding 8-aza-7-deazaguanosine derivatives 7, 9, and 10. Similarly, key intermediates 1 and 2 were heated with hydrazine solution to give the corresponding 6-hydrozinopurine nucleosides 6 and 12. The 6-hydroxyamino purine nucleoside 13 was obtained by treatment of compound 2 with hydroxylamine at 60 °C. Therefore, the 6-methoxy group was effectively used as the protecting group during the glycosylation process, and it can be further substituted with various nucleophiles to form 6-modified purine nucleoside derivatives. When compound 27 was deprotected with NH3/MeOH at 120 °C, the bromo-atom at position 2 was substituted with NH2 group during the process, resulting in the corresponding 7-bromo-8-aza-7-deazaguanosine analogue 4. Therefore, the bromo-atom at position 2 is more reactive against nucleophile than the bromo-atom at position 7. The site of glycosylation and anomeric configuration of 24, 25, and 26 was assigned on the basis of 1 H-NMR, 13 C-NMR, 2D NMR, UV spectra, and X-ray crystallographic analysis of the representative compounds.
The nucleoside products 1 and 3 have very similar UV absorption compared to the corresponding free purine bases A and B (Table 1). This indicates that the ribofuranosyl group was glycosylated at the N9 position of purine bases A and B because the structure still maintains the same aromatic conjugate system [1]. However, the N 8 -glycosylated nucleosides showed very different UV spectra because the ribose-base bond connection dramatically altered the aromatic conjugate system. Careful UV spectral comparison of the N 9 -and N 8 -glycosylated nucleoside pairs shows the difference clearly. The N 8 -glycosylated products 2, 10, 11, and 12 have UV maximum absorbances λ max at longer wavelength than corresponding N 9 -glycosylated products 1, 7, 5, and 6 ( Figure 3). The site of glycosylation and anomeric configuration of 24, 25, and 26 was assigned on the basis of 1 H-NMR, 13 C-NMR, 2D NMR, UV spectra, and X-ray crystallographic analysis of the representative compounds.
The nucleoside products 1 and 3 have very similar UV absorption compared to the corresponding free purine bases A and B (Table 1). This indicates that the ribofuranosyl group was glycosylated at the N9 position of purine bases A and B because the structure still maintains the same aromatic conjugate system [1]. However, the N 8 -glycosylated nucleosides showed very different UV spectra because the ribose-base bond connection dramatically altered the aromatic conjugate system. Careful UV spectral comparison of the N 9 -and N 8 -glycosylated nucleoside pairs shows the difference clearly. The N 8 -glycosylated products 2, 10, 11, and 12 have UV maximum absorbances λmax at longer wavelength than corresponding N 9 -glycosylated products 1, 7, 5, and 6 ( Figure 3).  We studied the 2D NMR (HMBC heteronuclear multiple bond correlation) of N 9 -glycosylated 8aza-7-deazaguanosine 7 and N 8 -glycosyalted 8-aza-7-deazaguanosine derivatives 10 ( Figure 4). 1′-H proton (5.88 ppm) on the ribose ring has a long-range coupling correlation with C4 (157.7 ppm) of the purine base for compound 7 while compound 10 does not have this coupling correlation. 1′-H (5.66 ppm) on the ribose ring has a long-range coupling correlation with C7 (128.2 ppm) of the purine base for compound 10 but compound 7 does not have the same coupling correlation. In addition, the 7-H on the purine ring of compound 10 is closer to the 1′-C of the ribose ring comparing to the distance We studied the 2D NMR (HMBC heteronuclear multiple bond correlation) of N 9 -glycosylated 8-aza-7-deazaguanosine 7 and N 8 -glycosyalted 8-aza-7-deazaguanosine derivatives 10 ( Figure 4). 1 -H proton (5.88 ppm) on the ribose ring has a long-range coupling correlation with C4 (157.7 ppm) of the purine base for compound 7 while compound 10 does not have this coupling correlation. 1 -H (5.66 ppm) on the ribose ring has a long-range coupling correlation with C7 (128.2 ppm) of the purine base for compound 10 but compound 7 does not have the same coupling correlation. In addition, the 7-H on the purine ring of compound 10 is closer to the 1 -C of the ribose ring comparing to the distance of compound 7. Therefore, the 7-H (8.525 ppm) on the purine ring of compound 10 has the long-range coupling correlation with 1 -C (94.057 ppm) of the ribose ring. However, compound 7 does not have the same coupling. Therefore, the 1 -C of the ribose ring is connected to N9 of the purine base on compound 7 (1 -H vs C4 coupling) while 1 -C of the ribose ring is connected to N8 of the purine base on the compound 10 (1 -H vs. C7 coupling and 7H vs 1 -C coupling). The data clearly further verified that compound 7 is the N 9 -glycosylated 8-aza-7-deazaguanosine, and compound 10 is the N 8 -glycosylated 8-aza-7-deazaguanosine derivative.
Molecules 2019, 24, x FOR PEER REVIEW 6 of 17 of compound 7. Therefore, the 7-H (8.525 ppm) on the purine ring of compound 10 has the long-range coupling correlation with 1′-C (94.057 ppm) of the ribose ring. However, compound 7 does not have the same coupling. Therefore, the 1′-C of the ribose ring is connected to N9 of the purine base on compound 7 (1′-H vs C4 coupling) while 1′-C of the ribose ring is connected to N8 of the purine base on the compound 10 (1′-H vs. C7 coupling and 7H vs 1′-C coupling). The data clearly further verified that compound 7 is the N 9 -glycosylated 8-aza-7-deazaguanosine, and compound 10 is the N 8glycosylated 8-aza-7-deazaguanosine derivative.

X-ray Crystallographic Study
Slow crystallization of compounds 7 and 8 from water gave X-ray quality crystals. A suitable crystal was selected and mounted on a Xcalibur, Eos, Gemini diffractometer. The crystal was kept at 293(2) K during data collection. Using Olex2, the structure was solved with the ShelXS structure solution program using Direct Methods and refined with the ShelXL refinement package using Least Squares minimization. The resulted crystal structures are shown in Figures 5 and 6. The results confirmed the β-D-anomeric configuration and the site of glycosylation as N9. Consequently, the structures of the corresponding products 7 and 8 were further verified.

X-ray Crystallographic Study
Slow crystallization of compounds 7 and 8 from water gave X-ray quality crystals. A suitable crystal was selected and mounted on a Xcalibur, Eos, Gemini diffractometer. The crystal was kept at 293(2) K during data collection. Using Olex2, the structure was solved with the ShelXS structure solution program using Direct Methods and refined with the ShelXL refinement package using Least Squares minimization. The resulted crystal structures are shown in Figures 5 and 6. The results confirmed the β-D-anomeric configuration and the site of glycosylation as N9. Consequently, the structures of the corresponding products 7 and 8 were further verified.
Furthermore, the iodinated intermediate 3 is a valuable starting point for the introduction of aromatic or alkynyl substituents by the Pd catalyzed cross-coupling reaction. The 7-iodine intermediate 3 was coupled with 2-(tributylstannyl)furan catalyzed by bis(triphenylphosphine) palladium(II) chloride at 90 • C in DMF providing derivative 14 in 90% yield. Compound 3 was reacted with phenylboronic acid catalyzed by tetrakis(triphenylphosphine)palladium (Pd(PPh 3 ) 4 ) in DME-water (2:1) mixture resulting in the desired 7-substituted product 15 in 80% yield. The 7-alkynylated products 16 and 17 were obtained in high yields by Sonogashira coupling reaction of 3 with alkynyl reagents (Scheme 3). The 8-aza-7-deazaguanine derivatives having 7-iodo (compound 9) and 7-bromo (compound 4) gave extremely low yields under Pd-catalyzed cross-coupling reaction conditions. Therefore, protection of 6-oxo group by methoxy group is required to activate the 7-halo atom for C-C coupling reactions. 293(2) K during data collection. Using Olex2, the structure was solved with the ShelXS structure solution program using Direct Methods and refined with the ShelXL refinement package using Least Squares minimization. The resulted crystal structures are shown in Figures 5 and 6. The results confirmed the β-D-anomeric configuration and the site of glycosylation as N9. Consequently, the structures of the corresponding products 7 and 8 were further verified.   Furthermore, the iodinated intermediate 3 is a valuable starting point for the introduction of aromatic or alkynyl substituents by the Pd catalyzed cross-coupling reaction. The 7-iodine intermediate 3 was coupled with 2-(tributylstannyl)furan catalyzed by bis(triphenylphosphine) palladium(II) chloride at 90 °C in DMF providing derivative 14 in 90% yield. Compound 3 was reacted with phenylboronic acid catalyzed by tetrakis(triphenylphosphine)palladium (Pd(PPh3)4) in DME-water (2:1) mixture resulting in the desired 7-substituted product 15 in 80% yield. The 7alkynylated products 16 and 17 were obtained in high yields by Sonogashira coupling reaction of 3 with alkynyl reagents (Scheme 3). The 8-aza-7-deazaguanine derivatives having 7-iodo (compound 9) and 7-bromo (compound 4) gave extremely low yields under Pd-catalyzed cross-coupling reaction conditions. Therefore, protection of 6-oxo group by methoxy group is required to activate the 7-halo atom for C-C coupling reactions. DME-water (2:1) mixture resulting in the desired 7-substituted product 15 in 80% yield. The 7alkynylated products 16 and 17 were obtained in high yields by Sonogashira coupling reaction of 3 with alkynyl reagents (Scheme 3). The 8-aza-7-deazaguanine derivatives having 7-iodo (compound 9) and 7-bromo (compound 4) gave extremely low yields under Pd-catalyzed cross-coupling reaction conditions. Therefore, protection of 6-oxo group by methoxy group is required to activate the 7-halo atom for C-C coupling reactions.

Biological Evaluation
Newly synthesized compounds were tested for inhibitory activity against human lung carcinoma cell line A549 and the human breast cancer cell line MDA-MB-231 (Table 2). We found that the 7-iodine substituted derivative 8 showed the best inhibitory activity against A549 with an IC50

Biological Evaluation
Newly synthesized compounds were tested for inhibitory activity against human lung carcinoma cell line A549 and the human breast cancer cell line MDA-MB-231 (Table 2). We found that the 7-iodine substituted derivative 8 showed the best inhibitory activity against A549 with an IC 50 value of 7.68 µM. Compounds 14 and 16 showed slightly better activity than other modified derivatives, which did not show detectable activity against the tested tumor cell lines. By comparing the activity of tested analogs with different substituents, it suggested that electron-donating group at 6-position of the base may improve the activity. Further biological evaluation of these modified nucleosides is in progress and will be reported in due course.

Procedure for Preparation of Glycosylated Products 24-26 and Key Intermediate 3
Synthesis of 6-amino-4-methoxy-1-(2,3,5-tri-O-benzoyl-β-D-ribofuranosyl)-1H-pyrazolo [3,4-d]pyrimidine (24): A suspension of 6-amino-4-methoxy-1H-pyrazolo [3.4-d]pyrimidine (purine base A, 9.0 g, 54.5 mmol, 1.0 eq) and a catalytic amount of ammonium sulfate in hexamethyldisilazane (HMDS, 150 mL) was refluxed for 6 h. The excess hexamethyldisilazane was removed by evaporation under reduced pressure, and the residue was dissolved in 1,2-dichloroethane (200 mL). l-O-Acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose (ribose I) (35.7 g, 70.8 mmol, 1.3 eq) was added at room temperature. The reaction mixture was cooled to 0 • C, and trimethylsilyl trifluoromethanesulfonate (TMSOTf, 29.6 mL, 163.5 mmol, 3.0 eq) was added dropwise for 30 min with stirring. The reaction mixture was stirred at room temperature overnight. Upon completion of the reaction as monitored by TLC, the mixture was diluted with dichloromethane (200 mL) and washed with saturated sodium bicarbonate solution. The aqueous layer was extracted with dichloromethane. The combined organic layer was dried over anhydrous sodium sulfate and concentrated under reduced pressure. The resulting residue was purified by column chromatography to afford 10.0 g main product N 9 β-isomer 24 as a white solid in 30.3% yield with an HPLC purity of 96%; R f = 0.3 (petroleum ether-ethyl acetate = 1:1). was added in 30 min with stirring at room temperature, and the reaction mixture was stirred at the same temperature for 5 h. Upon the completion of the reaction as monitored by TLC, the reaction mixture was poured into 500 mL of saturated sodium bicarbonate solution and extracted with ethyl acetate. The organic phase was separated, and the aqueous phase was extracted with ethyl acetate. The combined organic phases were dried over anhydrous sodium sulfate. The drying agent was filtered off, and the filtrate was concentrated under reduced pressure to afford crude product 26, which was used directly without further purification. R f = 0.2 (dichloromethane-methanol = 30:1). The crude compound 26 was dissolved in 100 mL of MeOH and 10 mL of THF solution. Then sodium methoxide (2.8 g, 51.6 mmol, 3.0 eq) was added, and the mixture was stirred at room temperature for 5 h. Upon completion of the reaction as monitored by TLC, the mixture was neutralized with 2 N HCl solution and evaporated under reduced pressure. The resulting residue was purified by column chromatography to afford 1.8 g desired product 3 as a white solid in 25% overall yield with an HPLC purity of 98.5%. R f = 0.  [3,4-d]pyrimidin-4(5H)-one (purine base C; 14.2 g, 48.4 mmol, 1.0 eq) was suspended in 300 mL dry acetonitrile and 1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose (ribose I) (36.6 g, 72.6 mmol, 1.5 eq) was added. The reaction mixture was heated to reflux at 95 • C, and the freshly distilled BF 3 ·OEt 2 (12.2 mL, 96.8 mmol, 2.0 eq) was then added with stirring. The reaction mixture became clear immediately and then slowly became dark. The mixture was kept stirred at this temperature for 20 min. Upon the completion of the reaction as monitored by TLC, the reaction mixture was cooled to room temperature and concentrated under reduced pressure. The resulting residue was purified by column chromatography to afford 33 g product 27 as a white solid in 92% yield with an HPLC purity of 98%.