Synthesis and separation of diastereomers of uridine 2′,3′-cyclic boranophosphate

Article abstract of DOI:10.1016/S0960-894X(00)00700-9

The first boron-containing 2′,3′-cyclic phosphate-modified analogue, uridine 2′,3′-cyclic boranophosphate (2′,3′-cyclic-UMPB), was synthesized. 5′-O-Protected uridine was cyclophosphorylated by diphenyl H-phosphonate to yield uridine 2′,3′-cyclic H-phosphonate, which upon silylation followed by boronation and subsequent acid treatment gave 2′,3′-cyclic-UMPB in high yield. The two diastereomers of 2′,3′-cyclic-UMPB were separated by HPLC. An alternative method for synthesis of uridine 2′,3′-cyclic phosphorothioate (2′,3′-cyclic-UMPS) via H-phosphonate was also described.

Full text of DOI:10.1016/S0960-894X(00)00700-9

Bioorganic & Medicinal Chemistry Letters 11 (2001) 615±617
Synthesis and Separation of Diastereomers of Uridine
20,30-Cyclic Boranophosphate
Kaizhang He and Barbara Ramsay Shaw*
Department of Chemistry, Gross Chemical Laboratory, Duke University, Durham, NC 27708, USA
Received 29 August 2000; accepted 7 December 2000
AbstractÐThe ®rst boron-containing 20,30-cyclic phosphate-modi®ed analogue, uridine 20,30-cyclic boranophosphate (20,30-cyclic-
UMPB), was synthesized. 50-O-Protected uridine was cyclophosphorylated by diphenyl H-phosphonate to yield uridine 20,30-cyclic
H-phosphonate, which upon silylation followed by boronation and subsequent acid treatment gave 20,30-cyclic-UMPB in high yield.
The two diastereomers of 20,30-cyclic-UMPB were separated by HPLC. An alternative method for synthesis of uridine 20,30-cyclic
phosphorothioate (20,30-cyclic-UMPS) via H-phosphonate was also described. # 2001 Elsevier Science Ltd. All rights reserved.
20,30-Cyclic phosphate esters of nucleosides are inter-
mediates in the ribonuclease-catalyzed hydrolysis of
ribonucleic acid (RNA), and are themselves substrates
for ribonucleases.1 Nucleoside 20,30-cyclic phosphoro-
thioate analogues have been exploited to study the
mechanism of ribonuclease catalysis.2 Comparison of
the con®guration of reactants and products where the
phosphate reaction center has been replaced by a phos-
phorothioate gives important information on the stereo-
chemical course of ribonuclease-catalyzed reactions.3À5
For example, pancreatic ribonuclease (RNase A) exhibits
a preference for the Rp (or endo) isomers of uridine
20,30-cyclic phosphorothioate (20,30-cyclic-UMPS) and
cytidine 20,30-cyclic phosphorothioate (20,30-cyclic-
CMPS).3 These phosphorothioate analogues have been
used to investigate the stereoselectivity of ribozyme
cleavage reactions.6 Likewise, nucleoside 20,30-cyclic
boranophosphate, in which a non-bridging oxygen of
the cyclic phosphate is replaced by an isoelectronic
borane group (BH3),7,8 should provide another useful tool
for investigating the mechanisms of the ribonuclease-
catalyzed reactions.
of other chiral phosphorothioate analogues.2 This
method involved thiophosphorylation of 50-O-acetyl4,9
or 50-O-DMT10 protected nucleoside by triimidazolyl-1-
phosphine-sul®de followed by water and ammonia
treatment (9±44% yields). Another approach utilized
2-chloro-4H-1,3,2-benzodioxa-phosphorin-4-one to phos-
phorylate the 50-O-acetyl ribonucleoside.11 Sulfurization
of the phosphorylated product followed by ammonia
treatment gave 20,30-cyclic-NMPS in 32±45% yields. In
another method, cyclization of nucleoside 30(20)-phos-
phorothioate derivatives in the presence of N,N0-dicyclo-
hexyl-carbodiimide (DCC) in pyridine or N-cyclohexyl-
N0-(3-trimethyl-ammonium-1-propyl)-carbodiimide led
to the formation of 20,30-cyclic-NMPS in 86±96%
yields.12
Here, we propose a new approach for synthesis of 20,30-
cyclic-UMPS, which involves sulfurization of the inter-
mediate uridine 20,30-cyclic H-phosphonate 4 (Scheme 1).
Intermediate 4 was obtained by condensation (cyclization)
of a mixture of uridine 30- and 20-H-phosphonates13 (2
and 3) by pivaloyl chloride. The condensation (cycliza-
tion) step was completed within 15 min, as suggested by
31P NMR (CDCl3) spectra of the reaction mixture in
which two signals at d 7.2 and 6.6 (for compounds 2 and
3) were shifted to d 21.0 and 19.0. Without puri®cation,
the reaction mixture containing intermediate 4 was oxi-
dized with elemental sulfur in carbon disul®de to yield
50-O-DMT-uridine 20,30-cyclic phosphorothioate 5. The
formation of intermediate 5 was con®rmed by the
appearance of the two single peaks at d 80.0 and 78.7 in
31P NMR (CDCl3) spectra of the reaction mixture. The
The ®rst synthesis of a 20,30-cyclic phosphate-modi®ed
analogue, 20,30-cyclic-UMPS, was described by Eckstein
in 1968.9 The Rp (or endo) isomer was isolated by frac-
tional crystallization of the triethylammonium salt3 and
used as reference to determine the absolute con®gurations
*Corresponding author. Tel.: +1-919-660-1551; fax:+1-919-660-1618;
e-mail: brs@chem.duke.edu
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00700-9

616
K. He, B. R. Shaw / Bioorg. Med. Chem. Lett. 11 (2001) 615±617
Scheme 1.
Scheme 2.
®nal deprotection of the 50-O-DMT group with 3%
dichloroacetic acid in dichloromethane gave the desired
20,30-cyclic-UMPS 6 (40% overall, 2/3!6).14 The Sp
(exo)- and Rp (endo)-diastereomers of 20,30-cyclic-
UMPS 6 were separated by HPLC (Fig. 1A).15
dichloromethane with water) failed because of its
instability and reactivity. Therefore, we developed an
alternative way to prepare intermediate 4 as shown in
Scheme 2.
We found that the reaction of diphenyl H-phosphonate
with 50-O-DMT-uridine 1 for 20 min gave the desired
intermediate 4 (31P NMR (CDCl3) d 27.8 and 23.9).
Without puri®cation, silylation of intermediate 4 with
BSA (5 min) yielded the phosphite triester 7, as indi-
cated by two signals at d 140.1 and 136.6 in 31P NMR
(CDCl3) spectra of the reaction mixture. Subsequent
boronation of the phosphite triester 7 and simultaneous
removal of the trimethylsilyl group were achieved by
addition of borane-N,N-diisopropylethyl-amine com-
plex.16À18 After 4 h, two broad peaks centered at d 127.6
and 123.9 appeared in 31P NMR (CDCl3) spectra of the
reaction mixture, con®rming the formation of 50-O-
DMT-uridine 20,30-cyclic boranophosphate 8. After
Synthesis of the uridine 20,30-cyclic boranophosphate
analogue 9 (Scheme 2) was not as straightforward as
that of the phosphorothioate analogue 20,30-cyclic-
UMPS 6. The boronation requires the intermediate
formation of a cyclic phosphite triester 7 by silylation of
H-phosphonate 4 using N,O-bis(trimethylsilyl)acetamide
(BSA).16À18 However, the silylation of intermediate 4
(prepared by cyclization of compound 2 and 3, Scheme
1) did not give the phosphite triester intermediate 7,
presumably because intermediate 7 could react with the
excess pivaloyl chloride and water formed in the con-
densation step (Scheme 1, 2/3!4). Attempts to purify
intermediate 4 using an extraction work up (ethyl acetate or
Figure 1. Isocratic separation of the two diastereomers of (A) 20,30-cyclic-UMPS 6 and (B) 20,30-cyclic-UMPB 9 by HPLC. Elution was carried out
on a Waters Delta Pak C18 column (15 m, 300 A, 7.8Â300 mm) with 6% (for 6) or 5% (for 9) methanol in 100 mM triethylammonium acetate
(TEAA, pH 6.8) at a ¯ow rate of 3.0 mL/min.

K. He, B. R. Shaw / Bioorg. Med. Chem. Lett. 11 (2001) 615±617
617
deprotection of the 50-O-DMT group by acid treatment
and ion-exchange chromatography puri®cation, the
desired 20,30-cyclic UMPB 9 was obtained in good yield
(70% overall, 1!9).19 The two diastereomers of 20,30-
cyclic UMPB 9, arbitrarily named as isomer I and II,
were separated by HPLC (Fig. 1B).20
13. 50-O-DMT-uridine 30-2 and 20-H-phosphonates 3 (72%,
31P NMR (CDCl3) d 8.5 (s), 7.2 (s); MS (FABÀ) 610.08 (calcd
610.56 for C30H31N2O10P)) were synthesized by phosphoryla-
tion of 50-O-DMT-uridine 1 by 4-chloro-4H-1,3,2-benzodioxa-
phosphorin-4-one and subsequent treatment with triethyl-
ammonium bicarbonate (TEAB) (Marugg, J. E.; Tromp, M.;
Kuyl-Yeheskiely, E.; van der Marel, G. A.; van Boom, J. H.
Tetrahedron Lett. 1986, 27, 2661).
14. 20,30-Cyclic-UMPS 6 (40%): 31P NMR (D2O) d 77.6 (s),
76.3 (s); 1H NMR (D2O) d 7.56, 7.55 (2d, J=8.0 Hz, 1H, H-6),
5.83, 5.75 (2d, J=3.2 Hz, 1H, H-10), 5.70, 5.69 (2d, J=8.0 Hz,
1H, H-5), 5.08 (m, 1H, H-20), 4.88±4.78 (m, 1H, H-30), 4.26,
4.17 (2m, 1H, H-40), 3.76±3.63 (m, 2H, H-50); MS (FABÀ)
320.98 (calcd 321.20 for C9H10N2O7PS).
During the preparation of this manuscript, a new e-
cient method utilizing the same intermediate 20,30-cyclic
H-phosphonate 4 to synthesize 20,30-cyclic-NMPS 6 has
been reported.21 The reaction of 50-O-DMT protected
nucleosides with diphenyl H-phosphonate in pyridine led
to the formation of intermediate 4, which upon sulfuriza-
tion and the subsequent removal of 50-DMT, gave 20,30-
cyclic-NMPS in 78±93% yields.21
15. 20,30-Cyclic-UMPS 6, isomer I (Sp, exo): rt=6.58 min
1
(33%); 31P NMR (D2O) d 77.6 (s); H NMR (D2O) d 7.54 (d,
J=8.4 Hz, 1H, H-6), 5.74 (m, 1H, H-10), 5.67 (d, J=7.6 Hz,
1H, H-5), 5.05 (m, 1H, H-20), 4.80 (m, 1H, H-30), 4.15 (m, 1H,
H-40), 3.73±3.60 (m, 2H, H-50); MS (FABÀ) 321.01 [M]À (calcd
321.20 for C9H10N2O7PS). 20,30-Cyclic-UMPS 6, isomer II
(Rp, endo): rt=10.24 min (47%); 31P NMR (D2O) d 76.1 (s);
1H NMR (D2O) d 7.52 (d, J=8.0 Hz, 1H, H-6), 5.79 (m, 1H,
H-10), 5.64 (d, J=8.0 Hz, 1H, H-5), 5.01 (m, 1H, H-20), 4.78
(m, 1H, H-30), 4.20 (m, 1H, H-40), 3.70±3.58 (m, 2H, H-50);
MS (FABÀ) 321.00 [M]À (calcd 321.20 for C9H10N2O7PS).
16. Sergueev, D.; Hasan, A.; Ramaswamy, M.; Shaw, B. R.
Nucleosides Nucleotides 1997, 16, 1533.
In summary, the ®rst 20,30-cyclic boranophosphate ana-
logue, 20,30-cyclic-UMPB, has been synthesized by an
H-phosphonate approach, which involves the silylation
of a 20,30-cyclic H-phosphonate intermediate followed
by boronation.16À18 The availability of the two diastereo-
mers of 20,30-cyclic-UMPB should be useful for deter-
mining the absolute con®gurations of other borano-
phosphate analogues.22À28 The potential of cyclic
boranophosphates as substrates or inhibitors for ribo-
nucleases should also provide valuable information
about the mechanisms of such ribonuclease-catalyzed
reactions.
17. Sergueev, D .; Shaw, B. R.J. Am. Chem. Soc. 1998, 120, 9417.
18. He, K.; Sergueev, D.; Sergueeva, Z.; Shaw, B. R. Tetra-
hedron Lett. 1999, 40, 4601.
19. 20,30-Cyclic-UMPB 9 (70%): 31P NMR (D2O) d 120.7 (q,
J=125.85 Hz), 116.10 (q, J=124.88 Hz); 1H NMR (D2O) d
7.48 (d, J=8.0 Hz, 1H, H-6), 5.74 (2d, J=3.2 Hz, 1H, H-10),
5.64 (m, 1H, H-5), 4.98±4.92 (m, 1H, H-20), 4.79 (q, J=5.6 Hz,
1H, H-30), 4.17, 4.01 (2q, J=4.0 Hz, 1H, H-40), 3.68±3.56
(m, 2H, H-50); MS (FABÀ) 303.07 (calcd 303.00 for
C9H13BN2O7P).
Acknowledgements
This work was supported by grant GM-57693 from the
NIH to B.R.S. Helpful discussions with Dr. Dmitri
Sergueev, Dr. Zinaida Sergueeva, and Mr. Jinlai Lin are
greatly appreciated. The authors thank Dr. Zinaida
Sergueeva for reading the manuscript and Dr. George
Dubay for doing mass spectroscopy.
20. 20,30-Cyclic-UMPB 9, isomer I: rt=9.40 min (39%); 31P
NMR (D2O) d 120.8 (q, J=145.71 Hz); 1H NMR (D2O) d 7.38
(d, J=7.6 Hz, 1H, H-6), 5.75 (m, 1H, H-10), 5.60 (d,
J=7.2 Hz, 1H, H-5), 5.03 (m, 1H, H-20), 4.79 (m, 1H, H-30),
4.19 (q, J=4.0 Hz, 1H, H-40), 3.74±3.62 (m, 2H, H-50); MS
(FABÀ) 303.0 [M]À (calcd 303.00 for C9H13BN2O7P). 20,30-
Cyclic-UMPB 9, isomer II: rt=19.11 min (43%); 31P NMR
1
References and Notes
(D2O) d 116.1 (q, J=130.17 Hz); H NMR (D2O) d 7.41 (d,
J=7.6 Hz, 1H, H-6), 5.68 (d, J=2.8 Hz, 1H, H-10), 5.62 (d,
J=7.6 Hz, 1H, H-5), 5.05 (m, J=3.2 Hz, 1H, H-20), 4.87 (q,
J=6.0 Hz, 1H, H-30), 4.02 (q, J=3.2 Hz, 1H, H-40), 3.75±3.63
(m, 2H, H-50); MS (FABÀ) 303.0 [M]À (calcd 303.00 for
C9H13BN2O7P).
1. Perreault, D. M.; Anslyn, E. V. Angew. Chem., Int. Ed.
Engl. 1997, 36, 432.
2. Eckstein, F. Annu. Rev. Biochem. 1985, 54, 367.
3. Saenger, W.; Eckstein, F. J. Am. Chem. Soc. 1970, 92, 4712.
4. Eckstein, F.; Schulz, H. H.; Ruterjans, H.; Haar, W.;
Maurer, W. Biochemistry 1972, 11, 3507.
21. Jankowska, J.; Wenska, M.; Popenda, M.; Stawinski, J.;
Kraszewski, A. Tetrahedron Lett. 2000, 41, 2227.
22. Tomasz, J.; Shaw, B. R.; Porter, K.; Spielvogel, B. F.;
Sood, A. Angew. Chem., Int. Ed. Engl. 1992, 31, 1373.
23. Sood, A.; Shaw, B. R.; Spielvogel, B. F. J. Am. Chem.
Soc. 1990, 112, 9000.
24. Krzyzanowska, B. K.; He, K.; Hasan, A.; Shaw, B. R.
Tetrahedron 1998, 54, 5119.
25. Sergueeva, Z. A.; Sergueev, D. S.; Shaw, B. R. Tetra-
hedron Lett. 1999, 40, 2041.
5. Heaton, P. A.; Eckstein, F. Nucleic Acids Res. 1996, 24,
850.
6. Koizumi, M.; Ohtsuka, E. Biochemistry 1991, 30, 5145.
7. Shaw, B. R.; Sergueev, D.; He, K.; Porter, K.; Summers, J.;
Sergueeva, Z.; Rait, V. Methods Enzymol. 1999, 313, 227.
8. He, K.; Hasan, A.; Krzyzanowska, B.; Shaw, B. R. J. Org.
Chem. 1998, 63, 5769.
9. Eckstein, F.; Gindl, H. Chem. Ber. 1968, 101, 1670.
10. Oivanen, M.; Ora, M.; Almer, H.; Stromberg, R.; Lonn-
berg, H. J. Org. Chem. 1995, 60, 5620.
26. He, K.; Porter, K.; Hasan, A.; Briley, J. D.; Shaw, B. R.
Nucleic Acids Res. 1999, 27, 1788.
11. Ludwig, J.; Eckstein, F. J. Org. Chem. 1989, 54, 631.
12. Holy, A.; Kois, P. Collect. Czech. Chem. Commun. 1980,
45, 2817.
27. Lin, J.; Shaw, B. R. Chem. Commun. 1999, 1517.
28. He, K.; Shaw, B. R. Nucleic Acids Symp. Ser. 1999, 41,
99.