Synthesis and hydrolytic properties of thymidine boranomonophosphate

Article abstract of DOI:10.1016/j.tetlet.2013.03.110

Detailed synthetic procedures of thymidine 5′-O-(P-borano) monophosphate (TMPB) via an H-phosphonate approach are reported. The hydrolytic properties of TMPB in pH 1.8 and 6.76 buffers at 22 or 37 C were studied using LC-MS. One nucleoside product, thymid

Full text of DOI:10.1016/j.tetlet.2013.03.110

Tetrahedron Letters 54 (2013) 2882–2885
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis and hydrolytic properties of thymidine boranomonophosphate
⇑
⇑
Zhihong Xu , Zinaida A. Sergueeva, Barbara Ramsay Shaw
Department of Chemistry, Box 90346, Duke University, Durham, NC 27708, United States
a r t i c l e i n f o
a b s t r a c t
Detailed synthetic procedures of thymidine 50-O-(P-borano)monophosphate (TMPB) via an H-phospho-
nate approach are reported. The hydrolytic properties of TMPB in pH 1.8 and 6.76 buffers at 22 or
37 °C were studied using LC–MS. One nucleoside product, thymidine, was observed, supporting cleavage
at the P–O ester bond to release nucleoside and boranophosphate (O3P-BH3ꢀ3). Higher temperatures and
acidic conditions increase the hydrolysis rates.
Article history:
Received 5 January 2013
Revised 13 March 2013
Accepted 23 March 2013
Available online 2 April 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Thymidine 50-P-boranomonophosphate
Boranation
H-phosphonate
Boron-modified nucleosides
LC–MS
Stability
Nucleoside based drugs are among the first choices in the che-
motherapy of viral infections. Typically, a modified nucleoside or
prodrug is converted in vivo into its mono-, di-, and finally tri-
phosphate analog, which is then incorporated by the viral poly-
merase, resulting in chain termination of the viral DNA or RNA
synthesis. Moreover, structure modification and biological investi-
gations of non-natural nucleotide analogs have been a major fo-
cus in the drug development area. Among these analogs is the
boranophosphate diester, which is isoelectronic with normal
phosphate diester bonds in DNA and RNA but considerably more
nuclease resistant than the parent phosphate. Further, borano-
phosphate RNA and DNA are able to form normal Watson–Crick
type pairs1 and be recognized by some kinases. A salient feature
of boranophosphate ribo and deoxyribo triphosphate forms is
their ability to substitute for natural nucleoside triphosphates
(NTPs) and be incorporated into RNA and DNA with ease by
RNA and DNA polymerases, respectively. Thus, boranophosphate
NTPs are able to read and transmit genetic information and find
use in DNA sequencing, gene targeting, RNA interference (RNAi),
and aptamer therapies.2–5
deoxy oligonucleotide boranomonophosphates.7 Later this H-phos-
phonate approach was successfully applied with different
modifications to the synthesis of some nucleoside boranomono-
phosphates.8–10
Thymidine
50-O-(P-borano)monophosphate
(TMPB) can be synthesized via a phosphoramidite approach,11
H-phosphonate approach,8–10 or through other borano-phosphor-
ylating reagents.12 In the phosphoramidite approach reported by
Tomasz et al. for the first synthesis of TMPB,11 the borane com-
plex was formed via a 12 h boranation of thymidine phospho-
ramidite, followed by
a 5 h base treatment to obtain the
corresponding boranophosphoramidate, and further treatment
with trifluoroacetic acid (TFA) for 30 min at 20 °C to produce
the title compound in 52% yield. In that phosphoramidite ap-
proach, some acid hydrolysis at room temperature occurred,
which likely accounted for a lower yield because of the instability
of the final product TMPB at a low pH. A recent study showed
that the major byproduct in this approach was thymidine 50-
monophosphate (TMP)9 which is isoelectronic with the TMPB
boron analog and hence difficult to remove. Here, we report a de-
tailed modified H-phosphonate approach to obtain TMPB (4) in
better yield, together with spectroscopic and MS kinetic analyses
of this nucleoside boranophosphate, its stability in well-buffered
solutions, and hydrolytic products.10
The synthesis of nucleoside
a-P-boranomono-, di-, and tri-
phosphates involves the isoelectronic substitution of a borane
(BH3) group for one of the nonbridging oxygens in phosphate dies-
ters,3–5 and has been achieved by a number of approaches outlined
in the recent review by our group.6 Sergueev and Shaw successfully
applied the H-phosphonate approach for the synthesis of
TMPB (4) was synthesized through an H-phosphonate approach
as shown in Scheme 1 and compared with the 1H NMR and 31P
NMR data reported by Tomasz et al.11 This approach can be used
for making
a variety of nucleoside analogs. H-phosphonate
derivatives in solution have been reported to exist as an
equilibrium mixture of two tautomeric forms: a tetracoordinate
phosphonate form and a tricoordinate phosphite form.13 However,
⇑
Corresponding authors. Tel.: +1 919 660 1551; fax: +1 919 660 1605.
E-mail addresses: zhihong.xu@duke.edu (Z. Xu), barbara.r.shaw@duke.edu (B.R.
Shaw).
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.03.110

Z. Xu et al. / Tetrahedron Letters 54 (2013) 2882–2885
2883
O
N
hydroxide for 2–3 h at room temperature. The reaction mixture
was separated over a QA52 ion exchange column, eluting with a
5–200 mM gradient of ammonium bicarbonate in water. One ma-
O
N
NH
NH
O
P O
H
SalPCl
-O
HO
O
O
DMF/Pyr
jor peak of compound 4 appearing at
a retention time of
O
O
ꢁ60 min, with a NH4HCO3 concentration at ꢁ90 mM, was sepa-
rated and freeze dried. Compound 4 was further purified with
HPLC by using a 0–10% acetonitrile gradient over 20 min in
10 mM triethylammonium acetate (TEAA) buffer (pH 7) as eluent.
Data for 4: C10H17BN2O7P. LC–MS: 319 (Mꢀ); 1H NMR (D2O, water
presaturated, 500 MHz, 25 °C) d 7.71 (s, 1H, H-6), 6.24 (t, J = 7.0 Hz,
1H, H-10), 4.47 (m, 1H, H-30), 4.06 (m, 1H, H-40), 3.90 (m, 2H, H-50),
2.25 (m, 2H, H-20), 1.83 (s, 3H, CH3-5), +0.57 to ꢀ0.05 (dq, J = 17.5,
TEAB
OAc
OAc
1
2
1. BSA
2. (CH3)2S:BH3
3. TEAB
O
O
N
NH
84.5 Hz, 3H, BH3); 31P NMR (D2O, 162 MHz, 25 °C) d 86.41 (q, JP–B
=
NH
O
O
O
NH4OH
164 Hz, 1P); UV (H2O) kmax 267 nm. The overall yield of 4 was
ꢁ65%; the purity was examined by LC–MS (Fig. 1) with isocratic
1% acetonitrile in 10 mM TEAA as eluent, and a single LC peak
(Fig. 1B) with a molecular ion mass 318.8 (TMPB) and a dimer mass
peak (at 638.7) at ꢁ4 min was detected (Fig 1C). However, after the
sample was freeze dried and redissolved in aqueous solution for
further stability studies, some decomposition was observed at
time = 0 (see Supplementary data), likely due to the instability of
the sample in water and upon lyophilization.
-O P O
-O P O
N
O
O
O
-
-
BH3
BH3
OH
OAc
4
3
Scheme 1. Synthesis of thymidine 50-O-(P-borano)monophosphate 4.
the equilibrium can be shifted to the P(III) state by fixing the phos-
phite form with N,O-bis(trimethylsilyl)acetamide (BSA), a silylat-
ing reagent.7 Therefore, the nucleophilic nature of the
phosphorus center becomes predominant and acts as an electron
donor to the BH3 group.
It is important to know the stability of nucleoside a-P-borano-
phosphate analogs in biologically relevant conditions and applica-
tions. Previously, Tomasz reported preliminary studies on TMPB
stability between pH 1 and 13 and showed that thymidine was
the sole UV-absorbing hydrolysis product.11 Huang showed that
the hydrolysis rate constant varied with pH, especially below pH
1 and above pH 7 at 37.2 °C,14 although the rate constant difference
between pH 7 and pH 2 was about 0.5-fold. In order to examine the
fate of the boron atom by 31P and 11B NMR studies in non-buffered
solutions, limited hydrolysis of concentrated TMPB in H2O and D2O
was conducted at 50 and 60 °C by Li et al. in 1996,15 which together
with Huang’s14 studies revealed that boranophosphate (BP, i.e.,
O3P-BH3ꢀ3) (observed by NMR15) and thymidine (studied by HPLC
in varied pH solutions14) were the main products (Scheme 2).
However, as the pH values of the hydrolysis solutions dropped
from pH 6–5, the chemical shifts in NMR spectra changed slightly
in Li’s study.15 Therefore, it was necessary to conduct and report
hydrolytic studies in well-buffered solutions. Our present work
by LC–MS is focused on quantifying the products and rates of time
dependent hydrolysis of TMPB in buffers under physiologically rel-
evant conditions: pH 6.76 and 1.8, and at 37 and 22 °C.
In order to increase the yield of TMPB (4), all reagents and sol-
vents were dried (unless specified), and the progress of the phos-
phorylation and boranation reactions was monitored by 31P NMR.
Compound 30-O-acetylthymidine 50-O-H-phosphonate (30-O-acet-
yl-TH-P, 2) was synthesized in a round bottomed flask under nitro-
gen, by injecting 4 mL DMF/pyridine (1.5 mL pyridine) into 585 mg
30-O-acetylthymidine (1, 1 equiv, 2.4 mmol), and stirring at room
temperature (rt, 24 °C) until the solid was dissolved, followed by
the addition of freshly prepared 2-chloro-4H-1,3,2-benzodioxa-
phosphorin-4-one (SalPCl, 542 mg, ꢁ1.3 equiv) in 1 mL DMF. After
stirring the mixture for an additional 20 min at rt, the reaction was
stopped by the addition of 4 mL of 1 M triethylammonium bicar-
bonate (TEAB) aqueous solution. About 924 mg of the H-phospho-
nate 2 (ꢁ95% purity by NMR) was obtained after separation on
silica gel (HPFC, Biotage HorizonTM) by eluting with a solvent of
CH2Cl2/MeOH (7:1). The fraction collection was monitored by UV
and normal phase silica gel TLC (60 F254, Merck) with CH2Cl2/
MeOH (7:3) as a developing solvent. The overall yield of 2 was
ꢁ90% by weight. Data for 2: C12H16N2O8P. LC–MS: 347 ([MꢀH]ꢀ);
1H NMR (DMSO, 400 MHz, 25 °C) d 11.33 (s, 1H, N–H), 7.75 (s,
1H, H-6), 6.61 (d, J = 592.3 Hz, 1H, P–H), 6.19 (dd, J = 8.8, 6.0 Hz,
1H, H-10), 5.19 (m, 1H, H-30), 4.07 (s, br, 1H, H-40), 3.84 (s, br, 2H,
H-50), 3.03 (q, J = 7.2 Hz, 6H, CH2 on NEt3H+), 2.32 and 2.21 (m,
2H, H-20), 2.04 (s, 3H, CH3 on AcO), 1.78 (s, 3H, CH3-5), 1.17 (t,
J = 7.2 Hz, 9H, CH3 on NEt3H+); 31P NMR (DMSO, 162 MHz, 25 °C)
d 2.71 (s, 1P, H-decoupled). UV (H2O) kmax 267 nm.
The title compound was obtained upon boranation of com-
pound 2 followed by deprotection of 30-O-acetylthymidine 50-O-
(P-borano)monophosphate (3) under basic conditions. To a dried
round bottomed flask with compound 2 (650 mg, 1.45 mmol) in
1 mL DMF, N, O-bis(trimethylsilyl)acetamide (BSA, 1.94 mL,
ꢁ5 equiv) was added. After 10 min, ꢁ1.28 mL boranation reagent
dimethylsulfide–borane complex, (CH3)2S/BH3 (2 M in THF,
ꢁ10 equiv), was added and stirred for 15 min, followed by the
addition of 5 mL TEAB (1 M), upon which the silylated intermedi-
ate was hydrolyzed by water to give ꢁ70% yield of 3 (ꢁ1 mmol).
The organic solvent in the reaction mixture was then evaporated
under vacuum, and TMPB 4 was produced quantitatively by depro-
tection of the 30-O-acetyl group with concentrated ammonium
For kinetic analysis, the freeze dried TMPB 4 was freshly dis-
solved in Milli-Q deionized water, and diluted to 100
lM final con-
centration in the 50 L hydrolysis solutions immediately prior to
l
the incubation at 22 or 37 °C. All buffers used in this study were
self-made,10 and buffer concentrations in the final incubation mix-
tures were 50 mM. Before and during the incubation of TMPB,
either 5 or 6 lL hydrolysis mixture was injected into the Agilent
1100 Series LC–MS system and eluted with 0–10% acetonitrile/
10 mM TEAA buffer in 10 min at 0.3 mL/min before the column
was washed with 60% acetonitrile. The first injection time of the
sample was referred to as t = 0.
Over several days at pH 6.76, the main peak TMPB 4 was slowly
converted into thymidine (T) in an aqueous solution. Time depen-
dent hydrolysis data of 4 by LC–MS in buffers of different pHs and
temperatures are available in Figure 2 and supporting material.
Under all conditions, the major hydrolysis product observed by
LC–MS was thymidine ([MꢀH]ꢀ: 241), which is in agreement with
the unbuffered NMR analysis by Hong Li15 (Scheme 2) for the
hydrolysis of a boranophosphoester bond and formation of inor-
ꢀ3
ganic boranophosphate (BP). Further, the reaction of (O3P-BH3
)
would result in some production of boric acid, which is expected
to be tolerable to cells, and possibly important in prebiotic condi-
tions. At time zero, two major peaks, namely T and TMPB, and one

2884
Z. Xu et al. / Tetrahedron Letters 54 (2013) 2882–2885
Figure 1. TMPB 4 purity detection by LC–MS. Panel A: MS negative ion chromatogram. Panel B: LC chromatogram. Conditions: isocratic 1% acetonitrile in 10 mM TEAA buffer,
pH 7 at 40
lL/min. Column: Polaris 3 C18-A 150 ꢃ 1.0 mm. Panel C: MS Spectrum.
O
N
O
N
half-lives (t1/2) and the pseudo-rate constants (k) of 4 in three buf-
fers are listed in Table 1. As expected, with different incubation
conditions, the hydrolyses proceeded with different rates. A differ-
ence of 15 °C in temperature caused a ꢁ7-fold increase in the pseu-
do-rate constant k (Table 1). At the same temperature (22 °C), the
overall hydrolysis rate was 1.5-fold faster in more acidic condition.
Under all conditions, the concentration (ln[TMPB])-time data
(Fig. 2) were fit to the pseudo-first-order reaction model, as de-
picted in Eq. 1, where [TMPB]0 is the initial concentration of 4
(starting concentration at t = 0), [TMPB]t is the concentration of 4
remaining in the hydrolysis mixture at time t, and k is the first-or-
der pseudo-rate constant.
NH
NH
O
O
P
H2O
-O
P
O
-
O-
-
+
-O
O
HO
O
O
O
BH3
BH3
OH
OH
TMPB
T
BP
Scheme 2. Proposed hydrolysis of TMPB 4 in H2O at 50 °C by Hong Li et al.15
ln½TMPBꢂt ¼ ꢀkt þ ln½TMPBꢂ0
ð1Þ
5
4.5
4
In summary, thymidine boranomonophosphate was synthe-
sized through an H-phosphonate approach with high yield. No acid
treatment was applied for this approach. Our results indicate that
under different incubation conditions, the hydrolyses of TMPB 4 fit
well to the first-order reaction equation with different rates. Our
experiments, together with the 31P and 11B NMR analysis of phos-
3.5
3
2.5
2
phate products by Li et al.15 are consistent with cleavage occurring
first at the P–O ester bond to release boranophosphate (O3P-BH3
and thymidine, and thereby pre-empting P–B bond cleavage (deb-
oranation) that might generate TH-P. While stable for days at neu-
tral pH, it is recommended that low temperatures and avoiding
acidic pH conditions be considered when handling boranomono-
phosphate type compounds.
ꢀ3
)
1.5
1
0.5
0
0
10
20
30
t(hr)
40
50
60
pH1.8 22 ºC
pH6.76 22 ºC
pH6.76 37 ºC
Figure 2. Hydrolysis kinetics of 100 lM TMPB in buffers.
Table 1
minor peak (<4%) of thymidine H-phosphonate (TH-P), were de-
tected (see Supplementary data). However, as no detectable
change in the peak area of TH-P was observed throughout the
experiments, its presence was attributed to a small amount of
background TH-P contained in the original TMPB product, but
not detectable earlier in the diluted HPLC fractions. The hydrolytic
Kinetic studies of 0.1 mM TMPB stabilities in buffers
Incubation solution (50 mM) pH
T (°C) t1/2 (h) k (sꢀ1
)
KH2PO4/H3PO4
K2HPO4/KH2PO4
K2HPO4/KH2PO4
1.80 22
26.0
39.0
5.0
7.38 ꢃ 10ꢀ6
6.76 22
6.76 37
(4.93 0.01) ꢃ 10ꢀ6
(3.74 0.03) ꢃ 10ꢀ5

Z. Xu et al. / Tetrahedron Letters 54 (2013) 2882–2885
2885
Symposium Series; Tocik, Z., Hocek, M., Eds.; Academy of Sciences of the Czech
Republic: Prague, 2002; Vol. 5, pp 169–180.
Acknowledgments
3. Shaw, B. R.; Madison, J.; Sood, A.; Spielvogel, B. F. Synthesis and Properties In
Methods in Molecular Biology Series; Agrawal, S., Ed.; Humana Press: Clifton, NJ,
1993; Vol. 20, pp 225–243.
4. Shaw, B. R.; Sergueev, D.; He, K.; Porter, K.; Summers, J.; Sergueeva, Z.; Rait, V.
Part A, Antisense Technology In Methods in Enzymology; Phillips, M. I., Ed.;
Academic Press: Orlando, FL, 2000; Vol. 313, pp 226–257.
5. Shaw, B. R.; Dobrikov, M.; Wang, X.; Wan, J.; He, K.; Lin, J. L.; Li, P.; Rait, V.;
Sergueeva, Z. A.; Sergueev, D. Ann. N. Y. Acad. Sci. 2003, 1002, 12–29.
6. Li, P.; Sergueeva, Z. A.; Dobrikov, M.; Shaw, B. R. Chem. Rev. 2007, 107, 4746–
4796.
NIH Grants (R01-AI-52061 and 2R01-GM-57693) to B.R.S.; LC–
MS data were collected at Duke University on the Agilent 1100 Ser-
ies departmental instrument and also in Professor John Simon’s lab
supported by the MFEL program administered by the AFOSR; NMR
data were collected at Duke NMR Center. We thank Prof. Don M.
Coltart for stimulating discussions.
7. Sergueev, D. S.; Shaw, B. R. J. Am. Chem. Soc. 1998, 120, 9417–9427.
8. He, K. Z., Ph.D. Dissertation; Duke University: Durham, NC, 2000.
9. Li, P., Ph.D. Dissertation; Duke University: Durham, NC, 2000.
10. Xu, Z. Ph.D. Dissertation, Duke University, Durham, NC 2008. Buffer
preparations: 1 M pH 1.8 buffer was prepared by titration of 1 M KH2PO4
with 1 M H3PO4 while pH was externally measured; 1 M pH 6.67 buffer was
prepared by mixing 1 M K2HPO4 and 1 M KH2PO4. All 1 M stock solutions were
Supplementary data
Supplementary data (hydrolysis data) associated with this arti-
cle can be found, in the online version, at http://dx.doi.org/
10.1016/j.tetlet.2013.03.110.
0.2 lm filtered (pore size 0.2 lm, 13 mm diameter, Whatman GD/X).
11. Tomasz, J.; Shaw, B. R.; Porter, K.; Spielvogel, B. F.; Sood, A. Angew. Chem., Int.
Ed. Engl. 1992, 31, 1373–1375.
References and notes
12. Lin, J. L. Ph. D. Dissertation; Duke University: Durham, NC, 2001.
13. Doak, G. O.; Freedman, L. D. Chem. Rev. 1961, 61, 31–44.
14. Huang, F. Q., Ph.D. Dissertation; Duke University: Durham, NC, 1994.
15. Li, H.; Hardin, C.; Shaw, B. R. J. Am. Chem. Soc. 1996, 118, 6606–6614.
1. Johnson, C. N.; Spring, A. M.; Sergueev, D.; Shaw, B. R.; Germann, M. W.
Biochemistry-US 2011, 50, 3903–3912.
2. Shaw, B. R.; Wan, J.; Wang, X.; Dobrikov, M.; He, K.; Porter, K.; Lin, J.-L.; Rait, V.;
Sergueev, D.; Sergueeva, Z. Chemistry of Nucleic Acid Components In Collection