Synthesis and properties of a novel phosphodiester analogue, nucleoside boranophosphorothioate

Article abstract of DOI:10.1039/a901157i

The first boranophosphorothioate [(RO)₂P(S)(BH₃)^-] mimic of a phosphodiester compound, dithymidine boranophosphorothioate, has been synthesized; while it is water soluble, this new analogue is more lipophilic and nuclease

Full text of DOI:10.1039/a901157i

Synthesis and properties of a novel phosphodiester analogue, nucleoside
boranophosphorothioate
Jinlai Lin and Barbara Ramsay Shaw*
Paul M. Gross Chemical Laboratory, Department of Chemistry, Duke University, Durham, North Carolina
27708-0346, USA. E-mail: brs@chem.duke.edu
Received (in Corvallis, OR, USA) 8th February 1999, Accepted 28th April 1999
The first boranophosphorothioate [(RO)2P(S)(BH3)2]
mimic of a phosphodiester compound, dithymidine bor-
anophosphorothioate, has been synthesized; while it is water
soluble, this new analogue is more lipophilic and nuclease
resistant than natural nucleoside phosphodiesters
[(RO)2P(O)(O)2] and phosphorothioates [(RO)2P(S)(O)2].
4 with 31P NMR signal at dP 116.6 (br). Dry 4 was reacted with
Li2S to give 5 (broad 31P NMR signal at dP 161.0) which was
converted to 6 with conc. NH4OH–MeOH (1:1) at room
temperature. The crude mixture was purified by ion-exchange
column chromatography on QA-52 (HCO32) cellulose to give 6
as the ammonium salt and isolated by HPLC. The overall yield
of dithymidine boranophosphorothioate 6 (TSpBT) was about
28%. Successful separation of the two diastereomers (RP and
SP) of 6 was achieved by reverse-phase HPLC. The first eluted
isomer TSpBT I (6a) and the second eluted isomer TSpBT II (6b)
Novel oligonucleotide analogues are currently attracting atten-
tion as probes in biochemistry and molecular biology1 and as
possible therapeutic agents against cancer and viral diseases.2
An impressive variety of these analogues3 have been developed
as potential therapeutic drugs for ‘antisense’ and ‘antigene’
targeting of specific genes to modulate their expression.4 Of
these modified oligonucleotides, the nucleoside phosphoro-
thioates5 and nucleoside boranophosphates6 (Fig. 1) are among
the most promising because they are resistant to nucleases and
support RNase H induced cleavage7 of the complementary
messenger RNA. Nearly a dozen phosphorothioate oligonucleo-
tides are now in clinical trials.8 By structurally combining the
phosphorothioate and boranophosphate backbones, we have
created a new phosphodiester analogue, the boranophosphoro-
thioate, [SNP–BH3]2, wherein the two nonbridging oxygen
atoms of a phosphodiester group are replaced with a sulfur atom
and borane group.
1
were characterized by 31P and H NMR.10 The Li2S method
used above should be applicable to the synthesis of other
boranophosphorothioates including [35SNP–BH3]2 phospho-
diesters.
In oligodeoxynucleotides (ODN), the replacement of a
nonbridging oxygen atom in the natural phosphodiester linkage
by S2, Me or BH32 imparts resistance to nucleases that cleave
DNA.6f,g,11 For instance, the non-ionic methylphosphonates
(Me-ODN) are highly resistant11c to phosphodiesterases; the
Here, we report the first example of a novel boranophosphoro-
thioate compound, specifically the dithymidine boranophos-
phorothioate, its synthesis and properties.
The general procedure for the synthesis of dinucleoside
boranophosphorothioates is outlined in Scheme 1. 5A-O
Fluorenylmethoxycarbonyl (Fmoc)-thymidine9 was phosphityl-
ated with (Pri2N)2PCl catalyzed by DMAP to give 1. Phosphite
1 was treated in situ with 3A-O-acetylthymidine and tetrazole in
DMF to give 2. To the above mixture, 4-nitrophenol and
tetrazole in DMF were added to yield 3, which was then treated
with excess BH3·SMe2 complex to afford the phosphite–borane
Boranophosphate
O
Normal Phosphate
O
Boranophosphorothioate
S
O
P
O
O
H:B:H
P
..
O
–
O
H:B:H
P
O
–
..
..
:O:
..
..
–
..
H
H
(b)
(a)
(c)
Phosphorothioate
O
Methylphosphonate
O
Phosphoramidate
O
O
P
O
O
H:C:H
P
O
N
P
O
..
..
..
H
:S:
:O:
..
–
..
..
–
H
(d )
(f )
(e)
Fig. 1 Structurally and/or electronically similar internucleotide linkages and
their abbreviations: (a) normal phosphate [ONP–O2], (b) boranophosphate
[ONP–BH32], (c) boranophosphorothioate [SNP–BH32], (d) phosphor-
othioate [ONP–S2], (e) methylphosphonate [ONP–Me], (f) phosphor-
amidate. Placement of the negative charge does not necessarily reflect the
actual spatial location of charge in the molecule.
Scheme 1
Chem. Commun., 1999, 1517–1518
1517

Table 1 Special properties of [SNP–BH3]2 backbone relative to natural and other phosphodiester backbone analogues
Property
[ONP–O]2
[ONP–S]2
ONP–Me
[ONP–BH3]2
[SNP–BH3]2
Nuclease resistance
Lipophilicity
RNase H activity
—
—
+
+
+
+
++
+++
—
+
+
+
++
++
N.D.
anionic phosphorothioates (S2-ODN) can be hydrolyzed by
phosphodiesterase I, but with a much lower rate11d than O2-
ODN. Boranophosphates (BH32-ODN) are even more re-
sistant6g to certain phosphodiesterases than S2-ODN. This
nuclease resistance, together with the ability to form stable
duplex structures with DNA or RNA, has led to broad trials of
S2- and Me-ODN as agents for regulating gene expression in
vitro and in vivo.4 By combining aspects of both the charged
[ONP–S]2 and the non-ionic ONP–Me into a [SNP–BH3]2
hybrid phosphodiester linkage, it was anticipated that the
resulting compound may exhibit greater nuclease resistance and
other unique or potentially useful properties. The BH3 moiety in
[SNP–BH3]2 is isoelectronic with oxygen and isosteric with the
Me group in ONP–Me, but imparts a negative charge to the
backbone, like S2 in [ONP–S]2. By virtue of the larger volume
of the BH3 group and its lack of lone pair electrons, we expect
that [SNP–BH3]2 oligonucleotides should be more lipophilic
than [ONP–S]2 oligonucleotides. The new [SNP–BH3]2-ODN,
which is a hybrid backbone of S2-ODN and BH32-ODN, has
some special properties summarized in Table 1.
The boranophosphorothioate group is very stable towards
basic or acidic hydrolysis. The diastereomers, TSpBT I and
TSpBT II, were each dissolved in 100 mM AcOH–NH4OH pH
3 or pH 11 buffer and incubated for 24 h at 37 °C. No
hydrolyzed or degraded products were detected via HPLC.
The boranophosphorothioate internucleotide linkage in dimer
6 is quite stable towards cleavage by both snake venom
phosphodiesterase (SVPDE) and bovine spleen phosphodiester-
ase (BSPDE). Under conditions where the natural dithymidine
phosphate (TpT) was > 99% cleaved by SVPDE, both TSpBT I
and TSpBT II were > 99% stable. Similarly, with BSPDE, while
TpT was > 96% cleaved, TSpBT I and TSpBT II were > 98 and
97% stable, respectively.
intriguing class of compounds, including modified nucleotides
and nucleic acids. Their similarity to natural nucleic acids and
unique properties such as high lipophilicity and resistance to
enzymatic cleavage, in conjunction with their potential utility as
molecular probes for the study of stereochemical aspects of
enzymatic and nonenzymatic reactions and as carriers of 10B in
boron neutron capture therapy (BNCT)14 for the treatment of
cancer, make the [SNP–BH3]2 linkage a promising candidate
for further mechanistic, diagnostic and therapeutic applica-
tions.
We thank Drs Dmitri S. Sergueev, Vladimir Rait and Zinaida
Sergueeva and Mr Kaizhang He for their suggestions and help.
This work was supported by grants 1R01-GM57693-01 from
NIH and DE-FG05-97ER62376 from DOE to B. R. S. and is in
partial fulfillment of requirements (J. L.) for a PhD at Duke
University.
Notes and references
1 F. Eckstein, Angew. Chem., Int. Ed. Engl., 1983, 22, 423.
2 Antisense Therapeutic: Progress and Prospect, ed. G. L. Trainor,
ESCOM Science Publishers, The Netherlands, 1996, pp. 1–85.
3 M. Egli, Angew. Chem., Int. Ed. Engl., 1996, 35, 1894; E. Uhlmann and
A. Peyman, Chem. Rev., 1990, 90, 543.
4 S. T. Crooke and C. F. Bennett, Annu. Rev. Pharmacol. Toxicol., 1996,
36, 107.
5 F. Eckstein, Annu. Rev. Biochem., 1985, 54, 367.
6 (a) A. Sood, B. R. Shaw and B. F. Spielvogel, J. Am. Chem. Soc., 1990,
112, 9000; (b) J. Tomasz, B. R. Shaw, K. Porter, B. F. Spielvogel and
A. Sood, Angew. Chem., Int. Ed. Engl., 1992, 31, 1373; (c) B. R. Shaw,
J. Madison, A. Sood and B. F. Spielvogel, Methods Mol. Biol., 1993, 20,
225; (d) H. Li, C. Hardin and B. R. Shaw, J. Am. Chem. Soc., 1996, 118,
6606; (e) K. W. Porter, J. D. Briley, B. R. Shaw, Nucleic Acids Res.,
1997, 25, 1611; (f) F. Huang, A. Sood, B. F. Spielvogel and B. R. Shaw,
J. Biomol. Struct. Dynam., 1993, 10, a078; (g) D. S. Sergueev and B. R.
Shaw, J. Am. Chem. Soc., 1998, 120, 9417.
7 V. Rait and B. R. Shaw, Antisense Nucleic Acid Drug Dev., 1999, 9, 53;
B. R. Shaw, in Papers from the 5th Annual International Conference on
Antisense: DNA and RNA Based Therapeutics. IBC, Southborough,
MA, 1998, pp. 1–8; RNase H activity was mentioned by Higson et al.
(A. P. Higson, A. Sierzchala, H. Brummel, Z.-G. Zhao and M. H.
Caruthers, Tetrahedron Lett., 1998, 39, 3899), but no data were
shown.
The [SNP–BH3]2 dimers carry a full negative charge and are
water soluble, yet are intermediate between normal phosphates
and methylphosphonates in lipophilicity. In partitioning experi-
ments,12 TSpBT was 320- and 18-fold more lipophilic than
natural TpT and TpBT (dithymidine boranophosphate) ac-
cordingly.
The [SNP–BH3]2 nucleotidic linkage is the only non-
bridging disubstituted chiral phosphodiester with a negative
charge. This property coupled with ready synthesis of isotopic
[35SNP–BH3]2 compounds from Li2S* (S* = 35S) could make
this linkage very useful for elucidating the stereochemical
course of phosphoryl and nucleotidyl transfer reactions and at
the same time probing whether one or two non-bridging
oxygens are necessary in these reactions.
8 S. Agrawal, Trends Biotechnol., 1996, 14, 376.
9 C. Lehmann, Y.-Z. Xu, C. Christodoulou, Z.-K. Tan and M. J. Gait,
Nucleic Acids Res., 1989, 17, 2379.
10 Selected data for 6 dH(D2O, 400 MHz) 7.56, 7.52, 7.50, 7.47 (4s, 1H,
H6), 6.18–6.05 (m, 2H, H1A), 4.85–4.72 (m, 1H, H3A), 4.44–4.34 (m, 1H,
H3A), 4.00 (m, 2H, H4A), 3.94–3.59 (2m, 4H, H5A), 2.38–2.14 (2m, 4H,
H2A), 1.77, 1.75, 1.74, 1.71 (4s, 6H, 5-CH3), 0.68–20.44 (br, 3H, BH3);
dP(D2O) 160.1 (br); lmax/nm 267; m/z (FAB2) 559.16 (M2) [Calc. for
To summarize, we have synthesized a totally new type of
modified phosphodiester analogue, in which the two non-
bridging oxygen atoms of a phosphodiester group have been
replaced with a sulfur atom and borane group. The analogue has
been placed in a nucleic acid and the resulting dithymidine
boranophosphorothioate diastereomers shown to be stable
under a broad range of pH conditions and highly resistant to
enzymatic cleavage relative to natural DNA. Based on partition-
ing into octanol, the boranophosphorothioates may exhibit a
greater membrane permeability than the O-oligonucleotides, yet
maintain nuclease resistance like the methylphosphonates. The
novel combination of high lipophilicity, reasonable water
solubility and nuclease resistance could be extremely useful for
drug design;13 the [SNP–BH3]2 linkage instead of [ONP–O]2
may enable the compounds to penetrate the plasma membrane
and to enter cells.
C
20H29O10N4BPS, 559.1435 (M2), found, 559.1421]; HPLC condi-
tions: Waters Delta Pak C18-300 Å, 15 m, 7.8 3 300 mm column;
eluants were 18% MeOH and 82% 20 mM KH2PO4 (pH 7.0); flow rate,
3.0 ml min21; tR (6a) = 28.69 min, tR (6b) = 32.52 min.
11 (a) P. S. Sarin, S. Agrawal, M. P. Civéira, J. Goodchild, T. Ikeuchi and
P. C. Zamecnik, Proc. Natl. Acad. Sci. USA, 1988, 85, 7448; (b) D. M.
Tidd, Anticancer Res., 1990, 10, 1169; (c) K. L. Agarwal and F. Riftina,
Nucleic Acids Res., 1979, 6, 3009; (d) P. M. J. Burgers and F. Eckstein,
Biochemistry, 1979, 18, 592.
12 Partition coefficients, defined as the ratio of concentration in octan-1-ol
to that in water, were 2.8 3 1022, 1.6 3 1023 and 8.7 3 1025 for TSpBT
(mixture of diastereomers), TpBT (dithymidine boranophosphate), and
TpT respectively. The partition coefficient of TSpBT was 320 times
greater than that of TpT.
13 C. McGuigan, D. Cahard, A. Salgado, E. De Clercq and J. Balzarini,
Antiviral Chem. Chemother., 1996, 7, 31.
14 M. F. Hawthorne, Angew. Chem., Int. Ed. Engl., 1993, 32, 950.
Thus, synthesis of the first [SNP–BH3]2 phosphodiester
analogue offers the possibility of preparing an entirely new and
Communication 9/01157I
1518
Chem. Commun., 1999, 1517–1518