2′,3′-dideoxy-3′-thionucleoside triphosphates: Syntheses and polymerase substrate activities

Article abstract of DOI:10.1021/ol070147w

(Chemical Equation Presented) All four 2′,3′-dideoxy-3′- thio-nucleosides (ddtNTPs) function as substrates for the Y410F mutant of Deep Vent (exo^-) DNA polymerase. Not only are the ddtNTPs Incorporated to form the N + 1 product, but further elo

Full text of DOI:10.1021/ol070147w

ORGANIC
LETTERS
2007
Vol. 9, No. 6
1161-1163
2′,3′-Dideoxy-3′-thionucleoside
Triphosphates: Syntheses and
Polymerase Substrate Activities
Meena,† Mui Sam,‡ Kathryn Pierce,† Jack W. Szostak,‡ and
Larry W. McLaughlin*,†
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill,
Massachusetts 02467, and Howard Hughes Medical Institute and Department of
Molecular Biology, Massachusetts General Hospital, Boston, Massachusetts 02114
mclaughl@bc.edu
Received January 19, 2007
ABSTRACT
-
All four 2
only are the ddtNTPs incorporated to form the N
-thiol on the 5 -triphosphate. Although other polymerases are likely to differ in their use of the ddtNTPs, there does not appear to be a
′
,3′
-dideoxy-3
′
-thio-nucleosides (ddtNTPs) function as substrates for the Y410F mutant of Deep Vent (exo ) DNA polymerase. Not
+
1 product, but further elongations are observed in which the key step is attack of the
3
′
′
fundamental prohibition against using a thiol nucleophile on a phosphate anhydride electrophile. The syntheses of four ddtNTPs (C, T, A, G)
are described.
Template-directed polymerization of DNA catalyzed by
polymerases uses dNTP building blocks in which the
attacking nucleophile is a 3′-OH. Hydroxyl is not the only
functionality that could be successful in such reactions. An
amino group or thiol group might well be an effective
nucleophile when placed on the 3′-carbon and positioned to
attack a 5′-triphosphate.
The 3′-amino-2′,3′-dideoxynucleoside triphosphates or 3′-
amino-3′-deoxynucleoside triphosphates place an amino
group on the dNTP or NTP 3′-carbon. These derivatives are
reported to function as inhibitors of DNA1 or RNA2
polymerases, with the former acting largely as chain termina-
tors after the initial nucleotide elongation event.
thiol on a nucleoside 3′-p-nitrophenolphosphodiester was
very slow.3 In the only study4 we could locate where selected
ddtNTPs have been examined as polymerase substrates, they
appear to function better as chain terminators than as effective
elongation substrates for Pol 1 and HIV reverse transcriptase.
Other examples would seem to support the contention that
thiol is a poor nucleophile for phosphates. Topoisomerase
and site-specific recombinase substrates containing a 5′-
phosphorothiolate5 are cleaved readily to generate the tyrosyl-
3′-phosphodiester and a 5′-terminal thiol. To complete the
reaction, it is necessary that the liberated thiol attack the
phosphorylated tyrosine and regenerate the phosphorothiolate
diester. In this5 and a similar case6 the thiol is an incompetent
substrate relative to the native 5′-OH; the thiolate linkage
functions as a suicide substrate. In a related example, the
Although an aliphatic thiol should be partially deprotonated
at pH 7 and function better as a nucleophile than hydroxyl,
that expectation does not hold with respect to attack on
phosphates. For example, the intramolecular attack of a 2′-
(3) Dantzman, C. L.; Kiessling, L. L. J. Am. Chem. Soc. 1996, 118,
11715-11719.
(4) Yuzhakov, A. A.; Chidgevadze, Z. G.; Bibilashvilli, R. S. FEBS Lett.
1992, 306, 185-188.
† Boston College.
‡ Massachusetts General Hospital.
(5) Burgin, A. B.; Huizenga, B. N.; Nash, H. A. Nucleic Acids Res. 1995,
23, 2973-2979.
(1) Kutateladze, T.; Bebelashvili, R.; Azhaev, A.; Kraevskii, A. FEBS
Lett. 1983, 153, 420-426.
(2) Armstrong, V. W.; Eckstein, F. Eur. J. Biochem. 1976, 70, 33-38.
(6) Nunes-Duby, S. E.; Radman-Livaja, M.; Kuimelis, R. G.; Pearline,
R. V.; McLaughlin, L. W.; Landy, A. J. Bacteriol. 2002, 184, 1385-1394.
10.1021/ol070147w CCC: $37.00
© 2007 American Chemical Society
Published on Web 02/24/2007

hammerhead ribozyme uses a 2′-OH of the active-site
cytidine optimally positioned to transesterify the adjacent 3′-
phosphodiester. Replacement of this functional group with
a 2′-SH results not in transesterification of the neighboring
phosphodiester, but rather in deglycosylation (attack at C1′)
of the active-site cytidine.7
Scheme 2. Syntheses of ddtNTPs
Previous reports suggest that ddtNTPs are unlikely to be
effective polymerase substrates (at least after the incorpora-
tion of a single chain-terminating monomer),4 and that thiols
in general are poor nucleophiles for phosphates.2-7 Although
some synthetic efforts have been described for these deriva-
tives,8,9 we decided to prepare samples of all four ddtNTPs
and provide some preliminary data for the activities of these
derivatives as potential polymerase substrates or inhibitors.
The 2′,3′-dideoxy-3′-thiopyrimidine nucleosides were pre-
pared from the relevant O2-C3′ anhydronucleosides es-
sentially as described10 (Scheme 1a). After removal of the
Scheme 1. Syntheses of ddtNTP Precursors
used in a second epimerization to form the 2′,3′-dideoxy-
3′-thiopurine nucleosides (Scheme 1b). These derivatives
were stored as the thioesters (disulfide formation was less
efficient for the purine derivatives).
Formation of the corresponding nucleoside triphosphates
used the procedure described by Eckstein12 and either the
dinucleoside disulfides (Scheme 2a) or the corresponding
thiobenzoates (Scheme 2b). We were attracted to the
possibility of long-term storage of the triphosphates as the
disulfide dimers, and for 2′,3′-dideoxy-3′-thiothymidine as
well as 2′,3′-dideoxy-3′-thiocytidine, formation of the bis-
(triphosphate) occurred with high yield (Scheme 2a). Al-
though the bis(triphosphates) were not pure since triphos-
phate formation was not quantitative at both O5′-oxygens,
they could be stored in this state. The desired 3′-thio ddNTPs
were obtained after treatment of the bis(triphosphates) with
TCEP and HPLC isolation. For the thiopurines it was more
effective to convert the thiobenzoates into the desired
triphosphates followed by treatment with ammonium hy-
droxide to generate the 3′-thio NTPs.
It seemed likely that DNA polymerases would elongate a
DNA primer and form the N + 1 product with ddtNTP
substrates, since the reaction chemistry is unchanged in the
first incorporation. However, formation of the N + 2 and
longer extension products requires that the thiol take part as
the substrate nucleophile in the formation of internucleotide
phosphorothiolate diesters (Figure 1). To test these possibili-
ties, we initially screened a number of DNA polymerases
(Pol I KF (exo-), Taq, Bst, Sequenase, Superscript reverse
transcriptase, Therminator and wt and several mutants of
Deep Vent (exo-) polymerase) for primer-extension with
ddtTTP. Of these, the Y410F mutant of Deep Vent exo-
polymerase appeared the best and was chosen for further
study. We then examined all four 2′,3′-dideoxy-3′-thio-
nucleoside triphosphates for substrate activity using a DNA
primer/template complex with each template containing a
stretch of four identical residues (T, C, A, or G) at its 5′-
S-benzoyl protecting group the nucleosides could be oxidized
to the disulfides and stored in this state (see Scheme 2a).
The disulfides also functioned to protect the thiol group
during subsequent formation of the bis(triphosphates). The
purine derivatives were not so easily obtained since formation
of the corresponding anhydro derivatives cannot occur. The
most efficient procedure to epimerize the C3′ carbon and
prepare it for a simple displacement reaction with thiol is
that described by Herdewijn11 in which formation of the 5′-
benzoate and 3′-triflate results in a displacement of the triflate
ester and migration of the benzoate ester to generate the 2′-
deoxy-xylo-nucleoside (Scheme 1b). After removal of the
benzoate ester and formation of the 5′-DMT ether and 3′-
mesyl ester, the sodium salt of thiobenzoic acid could be
(7) Hamm, M. L.; Schwans, J. P.; Piccirilli, J. A. J. Am. Chem. Soc.
2000, 122, 4223-4224.
(8) Yuzhakov, A. A.; Chidzhavadze, Z. G.; Bibilashvilli, R. Sh.;
Kraevskil, A. A.; Galegov, G. A.; Dorneeva, M. N.; Nosik, D. N.; Kilesso,
T. Yu. Bioorg. Khim. 1991, 17, 504-509.
(9) El-Babary, A. A.; Khodair, A. I.; Pedersen, E. B.; Nielsen, C.
Monatsh. Chem. 1994, 125, 1017-1025.
(10) Sabbagh, G.; Fettes, K. J.; Gosain, R.; O’Neil, I. A.; Cosstick, R.
Nucleic Acids Res. 2004, 32, 495-501.
(11) Herdewijn, P. J. Org. Chem. 1988, 53, 5050-5053.
(12) Ludwig, J.; Eckstein, F. J. Org. Chem. 1989, 54, 631-635.
Org. Lett., Vol. 9, No. 6, 2007
1162

Figure 1. Elongation of a primer/template complex with ddtNTPs.
Formaton of the N + 1 product generates a conventional phos-
phodiester with a 3′-terminal thiol. Formation of the N + 2 product
and the 3′-bridging phosphorothiolate requires nucleophilic attack
by the terminal 3′-thiol on the incoming triphosphate. “B” represents
the nucleobases and the two dotted lines represent base pairing to
the template residues.
Figure 2. Nucleotide primer/template extension reactions with
varying ddtNTPs. Primer-extension reactions were performed by
using 20 nM (for A, G and T) or 40 nM (for C) Y410F mutant of
Deep Vent (exo-), 3 nM labeled primer/template duplex, and 200
µM ddtNTP in thermopol buffer at 50 °C. Extension reactions were
monitored at various time points and analyzed by 20% denaturing
PAGE.
end (Figure 2). The primer was labeled with radioactive
phosphate and phosphorimaging was used to visualize the
elongation bands. In all cases the ddtNTPs functioned as
polymerase substrates and could extend the primer by four
residues. In two cases (ddtGTP and ddtCTP) the majority
of the primer had been elongated to the +4 product after a
60 min incubation. For ddtATP and ddtTTP there were
distributions of elongation products with the +4 band in each
case representing one of the major extension products.
This study illustrates that a thiol functional group can be
used as the nucleophile in the formation of a phosphorothi-
olate linkage when primer extension is catalyzed by the
Y410F mutant of the Deep Vent polymerase. On the basis
of previous reports,4 Pol I and HIV appear to be less
accommodating in their use of such substrates. Although
different polymerases are likely to vary in the details of
substrate recognition, there does not appear to be any
fundamental prohibition of using a thiol nucleophile on a
phosphate anhydride electrophile, while corresponding cata-
lyzed reactions on phosphodiesters5,6 appear to be much less
effective. The successful elongations illustrated in Figure 2
suggest that it may be possible to synthesize longer phos-
phorothiolate sequences in enzymatic reactions by using these
monomers. If sufficiently long and accurate elongations can
be achieved with ddtNTPs, it might be possible to use in
vitro selection methods to generate functional thiolate-DNA
molecules such as aptamers and DNA thio-enzymes.
DNA linkages containing 3′-phosphorothiolates have been
used to stabilize structural motifs,13 to probe enzymatic
processes,14 and to enhance antisense complexes.15 The
phosphorothiolate DNAs used in those previous studies were
prepared by chemical syntheses.16 The work reported here
suggests that enzymatic syntheses may provide an alternate
pathway for the preparation of similar materials.
Acknowledgment. We thank Jesse Chen (Massachusetts
General Hospital) and Allen Horhota (Boston College) for
assistance with the preparation of selected derivatives. This
work was supported by awards from the NSF to L.W.M.
(MCB 0451488) and to J.W.S. (CHE 0434507).
Supporting Information Available: Experimental pro-
cedures and characterization of all compounds along with
NMR spectra. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL070147W
(13) Brazier, J. A.; Fisher, J.; Cosstick, R. Angew. Chem., Int. Ed. 2006,
45, 114-117.
(14) Elliott, S. L.; Brazier, J.; Cosstick, R.; Connolly, B. A. J. Mol. Biol.
2005, 353, 692-703.
(15) Beevers, A. P. G.; Fettes, K. J.; Sabbagh, G.; Murad, F. K.; Arnold,
R. P.; Cosstick, R.; Fisher, J. Org. Biomol. Chem. 2004, 2, 114-119.
(16) Fettes, K. J.; O’Neil, I.; Roberts, S. M.; Cosstick, R. Nucleosides,
Nucleotides & Nucleic Acids 2001, 20, 1351-1354.
Org. Lett., Vol. 9, No. 6, 2007
1163