Determinants of unnatural nucleobase stability and polymerase recognition

Article abstract of DOI:10.1021/ja035398o

Six new unnatural nucleobases have been synthesized and characterized in terms of stability and selectivity of self-pairing in duplex DNA and efficiency and fidelity of self-pairing during polymerase-mediated replication. Each nucleobase has a conserved r

Full text of DOI:10.1021/ja035398o

Published on Web 07/18/2003
Determinants of Unnatural Nucleobase Stability and
Polymerase Recognition
Allison A. Henry, Chengzhi Yu, and Floyd E. Romesberg*
Contribution from the Department of Chemistry, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California, 92037
Received March 31, 2003; E-mail: floyd@scripps.edu
Abstract: Six new unnatural nucleobases have been synthesized and characterized in terms of stability
and selectivity of self-pairing in duplex DNA and efficiency and fidelity of self-pairing during polymerase-
mediated replication. Each nucleobase has a conserved ring structure but differs from the others in its
specific pattern of substitution with oxygen and sulfur atoms. Heteroatom derivatization within the conserved
scaffold is shown to have only moderate effects on unnatural self-pair synthesis by the polymerase; larger
effects were observed on the thermal stability and polymerase-mediated extension of the self-pairs. The
largest effects of heteroatom substitution were on the stability and synthesis of mispairs between the
unnatural and natural bases. Certain heteroatom substitutions were found to have a general effect while
others were found to have effects that were specific for a particular unnatural or natural base. The data are
useful for designing stable and replicable third base pairs and for understanding the contributions of
nucleobase shape, polarity, and polarizability to the stability and replication of DNA.
1. Introduction
In characterizing the stability and replication of these non-
H-bonding unnatural nucleobases, we identified several promis-
ing base pairs formed between two of the same bases. In
principle, an unnatural self-pair, as opposed to a heteropair (like
the natural base pairs), is not a limitation on expansion of the
genetic alphabet as the addition of a single self-pair would result
in 64 new codons. In practice, a self-pair is advantageous
because the potential for mispairing with dA, dC, dG, and dT
is reduced. Previously, the most interesting unnatural nucleo-
bases were based on the azaindole, isocarbostyril, or naphthyl
scaffolds: 7AI:7AI,13,15 ICS:ICS,12 PICS:PICS,11 and 3MN:
3MN14 (Figure 1). These self-pairs are stable in duplex DNA
and enzymatically synthesized with reasonable efficiency and
selectivity; however, continued primer extension downstream
of the unnatural self-pair is inefficient in each case.
The storage and replication of biological information is based
on complementary interactions between the natural nucleobases
dA:dT and dG:dC. The identification of enzymatically replicable
unnatural base pairs with thermal stabilities comparable to those
of dA:dT and dG:dC would significantly expand the biological
and chemical potential of DNA. While the hydrogen-bonding
(H-bonding) patterns of the natural base pairs are an obvious
source of stability and specificity, efforts to design unnatural
bases with different H-bonding patterns have proven difficult.1-5
An alternative unnatural nucleobase design strategy is based
on the observation that H-bonds are not an absolute requirement
for the stability or replication of DNA.6-10 Thus, we have
explored the utility of hydrophobic and van der Waals forces
in developing additional base pairs to store and replicate
increased genetic information.11-17
There are at least two possible reasons for the observed poor
unnatural self-pair extension rates. One obvious feature of nat-
ural DNA is the presence of heteroatoms in both the major and
minor grooves. These atoms are not only involved in interbase
H-bonding but also contribute to the physical properties of the
nucleobases, including dipole moment and polarizability.18-22
(1) Bain, J. D.; Switzer, C.; Chamberlin, A. R.; Benner, S. A. Nature 1992,
356, 537-539.
(2) Horlacher, J.; Hottiger, M.; Podust, V. N.; Hu¨bscher, U.; Benner, S. A.
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 6329-6333.
(3) Lutz, M. J.; Held, H. A.; Hottiger, M.; Hu¨bscher, U.; Benner, S. A. Nucleic
Acids Res. 1996, 24, 1308-1313.
(4) Lutz, M. J.; Horlacher, J.; Benner, S. A. Bioorg. Med. Chem. Lett. 1998,
8, 1149-1152.
(13) Wu, Y.; Ogawa, A. K.; Berger, M.; McMinn, D. L.; Schultz, P. G.;
Romesberg, F. E. J. Am. Chem. Soc. 2000, 122, 7621-7632.
(14) Ogawa, A. K.; Wu, Y.; Berger, M.; Schultz, P. G.; Romesberg, F. E. J.
Am. Chem. Soc. 2000, 122, 8803-8804.
(5) Lutz, S.; Burgstaller, P.; Benner, S. A. Nucleic Acids Res. 1999, 27, 2792-
2798.
(6) Moran, S.; Ren, R. X.-F.; Rumney, S. I.; Kool, E. T. J. Am. Chem. Soc.
1997, 119, 2056-2057.
(15) Tae, E. L.; Wu, Y. Q.; Xia, G.; Schultz, P. G.; Romesberg, F. E. J. Am.
Chem. Soc. 2001, 123, 7439-7440.
(7) Moran, S.; Ren, R. X.-F.; Kool, E. T. Proc. Natl. Acad. Sci. U.S.A. 1997,
94, 10506-10511.
(16) Berger, M.; Luzzi, S. D.; Henry, A. A.; Romesberg, F. E. J. Am. Chem.
Soc. 2002, 124, 1222-1226.
(8) Morales, J. C.; Kool, E. T. Nat. Struct. Biol. 1998, 5, 950-954.
(9) Kool, E. T. Biopolymers 1998, 48, 3-17.
(17) Yu, C.; Henry, A. A.; Romesberg, F. E.; Schultz, P. G. Angew. Chem., Int.
Ed. 2002, 41, 3841-3844.
(10) Kool, E. T.; Morales, J. C.; Guckian, K. M. Angew. Chem., Int. Ed. 2000,
39, 990-1009.
(18) Bergman, E. D.; Weiler-Feilchenfeld. In The Dipole Moments of Purines;
Bergman, E. D., Pullman, B., Eds.; The Israel Academy of Sciences and
Humanities: Jerusalem, 1971; pp 21-28.
(11) McMinn, D. L.; Ogawa, A. K.; Wu, Y.; Liu, J.; Schultz, P. G.; Romesberg,
F. E. J. Am. Chem. Soc. 1999, 121, 11585-11586.
(12) Ogawa, A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P. G.; Romesberg,
F. E. J. Am. Chem. Soc. 2000, 122, 3274-3287.
(19) Lister, J. H. The Purines. Supplement 1; John Wiley and Sons: New York,
1996; Vol. 54.
9
9638
J. AM. CHEM. SOC. 2003, 125, 9638-9646
10.1021/ja035398o CCC: $25.00 © 2003 American Chemical Society

Nucleobase Stability and Polymerase Recognition
A R T I C L E S
has reduced surface area, compared to the aforementioned bases,
and may be derivatized with heteroatoms likely to be located
in the minor groove (position 1), at the nucleobase interface
(position 2) or in the major groove (position 3). This provides
an opportunity to vary the heteroatom substitution pattern in a
controlled manner and is expected to help deconvolute the
contributions of nucleobase shape, hydrophobicity, dipole
moment, and polarizability to the stability and replication of
the unnatural self-pairs, as well as the mispairs between
unnatural and natural bases. Herein, we show that heteroatom
substitution has variable effects on the stability and polymerase-
catalyzed synthesis of the unnatural self-pairs and mispairs.
Studies of this nature not only aid in the design of unnatural
nucleobases to expand the genetic alphabet but also contribute
to a general understanding of the determinants of DNA stability
and replication.
2. Results
Figure 1. 7AI, ICS, PICS, 3MN, and SNICS self-pairs.
2.1. Base Pair Design and Synthesis. Using the modeling
program Insight II (Biosym/MSI), we predicted that self-pairs
formed between unnatural nucleobases comprised of six- and
five-membered ring fusions would be well accommodated in
B-form DNA. For the six-membered ring component, pyridone
and thiopyridone were chosen to give an N-glycoside with either
an oxygen or a sulfur atom positioned in the minor groove. For
the five-membered ring component, furan and thiophene were
chosen in order to allow manipulation of the physical properties
of the unnatural nucleobases, such as their dipole moment and
polarizability. These elements lead to the six unnatural nucleo-
base analogues (see Figure 2): furo[3,2-c]pyridin-4(5H)-one
(4OFP), thieno[3,2-c]pyridin-4(5H)-one (4OTP), furo[3,2-c]-
pyridine-4-thiol (4TFP), furo[2,3-c]pyridin-7(6H)-one (7OFP),
thieno[2,3-c]pyridin-7(6H)-one (7OTP), and furo[2,3-c]pyri-
dine-7-thiol (7TFP). To facilitate the following discussion, we
refer to heteroatom substitution at the positions labeled 1, 2,
and 3 (Figure 2). Substitution at position 1 is expected to orient
the heteroatom in the minor groove of duplex DNA (or in the
developing minor groove during replication); substitution at
position 2 is expected to bury the heteroatom within the duplex;
and substitution at position 3 is expected to position the
heteroatom in the major groove. These structural predictions
are based on an assumption that the unnatural nucleobases will
adopt a pyrimidine-like conformation in duplex DNA.
Unnatural bases 1-3 and 4-6 were synthesized as shown in
Schemes 1 and 2, respectively. Starting materials 1A, 2A, 4A,
and 5A were prepared using a literature protocol.24 Analogues
3B and 6B, which contain a thiocarbonyl group, were synthe-
sized from the corresponding lactams using Lawesson’s chem-
istry, as previously reported.17 Briefly, formation of the N-gly-
cosidic bond was effected by silyl-Hilbert-Johnson methodology
under Vorbruggen conditions.25 After removal of the toluoyl
protecting groups and 5′-tritylation of the free nucleoside, the
desired â-anomer was easily separated from the R-anomer by
silica gel column chromatography. The assignment of â-stere-
ochemistry at C1′ for each dimethoxytrityl (DMTr)-protected
nucleoside was based on COSY and NOESY spectra, which
showed cross-peaks between H1′ and both H4′ and R-H2′. The
free nucleosides to be used in the syntheses of the corresponding
Figure 2. Unnatural nucleobases. Derivatization with an oxygen or sulfur
heteroatom is indicated by the letter X and differentiated by a numerical
subscript corresponding to positions 1, 2, and 3, assumed to be minor groove,
self-pair interface, and major groove, respectively (see text).
It is not simple to evaluate how these properties affect stability
and replication, but it seems possible that the general absence
of these heteratoms from the previously reported unnatural bases
could contribute to their poor extension efficiency.23 For
example, ICS lacks heteroatoms except for the carbonyl that is
expected to be oriented toward the minor groove. By compari-
son, its analogue, SNICS (Figure 1), with a nitrogen and a sulfur
atom that may be positioned in the major and minor grooves,
respectively, forms self-pairs that are extended with improved
efficiency.17 The second possible origin of the poor extension
rates may be the large aromatic surface area of the unnatural
nucleobases. This could cause steric repulsion or interstrand
intercalation between the unnatural bases that results in distortion
of the primer terminus to a geometry that is inappropriate for
continued primer extension.
To systematically evaluate the effects of surface area and
heteroatom derivatization, we synthesized and characterized six
nucleobases that share a scaffold based on the fusion of a
pyridone or thiopyridone with a furan or thiophene (Figure 2).
These base analogues can be coupled with 2′-deoxyribofuranose
by an N-glycosidic linkage while retaining aromaticity. Each
(20) Hurst, D. T. An Introduction to the Chemistry and Biochemistry of
Pyrimidines, Purines and Pteridines; Page Bros.: Norwich, U. K., 1980.
(21) Gilchrist, T. L. Heterocyclic Chemistry; Longman Scientific & Technical:
Essex, 1985.
(22) Bugg, C. E. In Solid-state packing patterns of purine bases; Bergmann, E.
D., Pullman, B., Eds.; The Israel Academy of Sciences and Humanities:
Jerusalem, 1971; pp 178-204.
(23) Guo, M.-J.; Hildbrand, S.; Leumann, C. J.; McLaughlin, L. W.; Waring,
M. J. Nucleic Acids Res. 1998, 26, 1863-1869.
(24) Shiotani, S.; Morita, H. J. Heterocyclic Chem. 1982, 19, 1207-1209.
(25) Niedballa, U.; Vorbruggen, H. J. Org. Chem. 1974, 39, 3654-3660.
9
J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003 9639

A R T I C L E S
Henry et al.
Scheme 1 a
Table 1. Denaturation Temperatures for Duplex DNA Containing
Unnatural Base Pairsa
5′-dGCGTACXCATGCG
3′-dCGCATGYGTACGC
X
Y
Tm, °C
X
Y
Tm, °C
4OTP
4OTP
dA
dT
dC
dG
4TFP
dA
dT
dC
dG
7OTP
dA
dT
dC
dG
57.9 ( 0.04
51.5 ( 0.10
50.6 ( 0.02
49.0 ( 0.03
49.8 ( 0.06
53.7 ( 0.14
50.1 ( 0.03
50.2 ( 0.10
46.2 ( 0.13
47.0 ( 0.11
52.6 ( 0.07
51.2 ( 0.21
50.7 ( 0.24
45.7 ( 0.13
47.5 ( 0.30
4OFP
4OFP
dA
dT
dC
dG
7OFP
dA
dT
dC
dG
7TFP
dA
dT
dC
dG
55.0 ( 0.07
51.7 ( 0.27
47.2 ( 0.19
48.9 ( 0.04
50.7 ( 0.33
53.3 ( 0.10
50.3 ( 0.11
46.6 ( 0.14
48.6 ( 0.31
49.5 ( 0.16
52.0 ( 0.01
48.2 ( 0.02
47.4 ( 0.02
44.4 ( 0.03
46.3 ( 0.04
4TFP
7OTP
7OFP
7TFP
Scheme 2 a
a See text for experimental details.
where 4OFP was found to be 1.7 °C more stable than 7OFP
and more pronounced with sulfur as the heteroatom, where
4OTP was found to be 5.3 °C more stable than 7OTP. In fact,
the 4OTP self-pair was the most stable of all, only 1.3 °C less
stable than a dA:dT pair in the same sequence context (Tm
)
57.9 and 59.2 °C for 4OTP:4OTP and dA:dT, respectively).
In contrast, in the case of heteroatom placement at position 2,
the duplex containing an unnatural self-pair was slightly more
stabilized by oxygen than sulfur (compare 7OFP to 7OTP).
Each of the unnatural nucleobases was strongly selective for
self-pairing (Table 1). This thermal selectivity was greatest for
4OTP, a self-pair that was 6.4 °C to 8.9 °C more stable than
any mispair, comparable to the thermal selectivity observed
among the natural base pairs. Stabilities of the mispairs between
4OTP and the natural bases varied over a range of only 2.5 °C
but generally increased with the hydrophobicity of the natural
base, dA being the most hydrophobic and forming the most
stable mispair. The 4OFP self-pair was the next most selective,
with mispairs destabilized by 3.3 °C to 7.8 °C, and again, the
relative stabilities paralleled the hydrophobicity of the natural
base. It is apparent that the slightly reduced selectivity of the
4OFP self-pair, relative to 4OTP, resulted from a decrease in
the stability of the self-pair while the absolute stabilities of the
mispairs remained the same, except in the case of dT. In this
case, substitution of a sulfur atom (4OTP:dT) with an oxygen
atom (4OFP:dT) resulted in a 3.4 °C decrease in duplex Tm. A
comparison of 4OFP with 4TFP reveals that substitution of
the minor groove oxygen atom with sulfur resulted in decreased
stability of the 4TFP self-pair and its mispairs to give an overall
selectivity that was virtually identical to that observed with
4OFP; however, the mispair with dT was again an exception
as the presence of sulfur at position 1, like at position 3,
stabilized the mispair with dT, by 3.0 °C compared to 4OFP:
dT. The 7OFP, 7OTP, and 7TFP self-pairs, with oxygen or
sulfur at positions 1 and 2, were 1.4 to 6.9 °C more stable than
all mispairs with natural bases. The thermal selectivities of
7OFP, 7OTP, and 7TFP were the result of decreases in self-
pair stability as well as decreases in the stabilities of the
mispairs.
a General conditions: (a) Bis-TMS acetamide, rt, 7, then SnCl4, 0 °C;
(b) 0.5 M NaOMe, MeOH, rt; (c) DMTrCl, pyridine, rt; (d) LiBF4/CH3CN,
MeOH-CH2Cl2 (1:9), rt; (e) Proton Sponge, POCl3, Bu3N, Bu3NPPi,
(MeO)3P, DMF, 0 °C; (f) 2-Cyanoethyl diisopropylchlorophosphoramidite,
DIEA, CH2Cl2, 0 °C; (g) Lawesson’s reagent, toluene, reflux.
triphosphates were obtained by treatment of the DMTr-protected
nucleoside intermediates with LiBF4.26 In all cases, preparation
of phosphoramidites from the corresponding DMTr-protected
nucleosides followed reported procedures.11
2.2. Unnatural Self-Pair Stability and Selectivity in Duplex
DNA. The stabilities of the unnatural self-pairs and the mispairs
with natural bases were determined by thermal denaturation of
duplex DNA. The duplex melting temperature (Tm) was
measured using the complementary oligonucleotides 5′-GC-
GATGXGTAGCG-3′ and 5′-CGCTACYCATCGC-3′, with
either an unnatural or a natural base incorporated at positions
X and Y (Table 1).
Heteroatom substitution was found to have significant effects
on self-pair stability in duplex DNA. For example, comparison
of 4OFP to 4TFP and 7OFP to 7TFP shows that in both cases
an amide conferred greater self-pair stability than a thioamide,
by 1.3 °C. Heteroatom substitution in the five-membered ring
had less general effects on duplex stability, though in all cases
placement of the heteroatom at position 3 resulted in more stable
self-pairs than placement of the heteroatom at position 2. This
effect was less pronounced with oxygen as the heteroatom,
2.3. Replication of DNA Containing the Unnatural Base
Pairs. In general, the two components of DNA replication are
(26) Chen, A.; Zheng, Y.; Zhou, X. Synth. Commun. 1999, 29, 3421-3423.
9
9640 J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003

Nucleobase Stability and Polymerase Recognition
A R T I C L E S
Table 2. Incorporation of Natural and Unnatural Triphosphates
Table 3. Incorporation of Unnatural Triphosphates Opposite
Natural Templating Basesa
Opposite Unnatural Templating Basesa
k
cat/KM
k
cat/KM
template base
triphosphate
k
cat (min-1
ndb
)
KM (µM)
ndb
(min-1 M-1
)
template base
triphosphate
k
cat (min-1
)
KM (µM)
(min-1 M-1
)
4OFP
d(4OFP)TP
dATP
dCTP
dGTP
dTTP
d(4OTP)TP
dATP
dCTP
dGTP
dTTP
d(7OFP)TP
dATP
dCTP
dGTP
dTTP
d(7TFP)TP
dATP
dCTP
dGTP
dTTP
d(7OTP)TP
dATP
dCTP
dGTP
dTTP
d(4TFP)TP
dATP
dCTP
dGTP
dTTP
<1.0 × 103
3.4 × 103
<1.0 × 103
8.8 × 103
1.2 × 104
5.9 × 103
<1.0 × 103
<1.0 × 103
4.8 × 103
1.0 × 104
5.2 × 103
2.5 × 103
<1.0 × 103
1.5 × 106
9.1 × 103
1.1 × 104
<1.0 × 103
<1.0 × 103
<1.0 × 103
<1.0 × 103
1.6 × 104
4.9 × 103
<1.0 × 103
1.9 × 104
1.2 × 105
9.3 × 104
4.2 × 105
<1.0 × 103
2.7 × 106
4.7 × 104
A
d(4OFP)TP
d(4OTP)TP
d(7OFP)TP
d(7TFP)TP
d(7OTP)TP
d(4TFP)TP
d(4OFP)TP
d(4OTP)TP
d(7OFP)TP
d(7TFP)TP
d(7OTP)TP
d(4TFP)TP
d(4OFP)TP
d(4OTP)TP
d(7OFP)TP
d(7TFP)TP
d(7OTP)TP
d(4TFP)TP
d(4OFP)TP
d(4OTP)TP
d(7OFP)TP
d(7TFP)TP
d(7OTP)TP
d(4TFP)TP
1.25 ( 0.03
2.5 ( 0.5
2.7 ( 1.0
2.7 ( 1.6
0.21 ( 0.02
0.4 ( 0.2
ndb
0.13 ( 0.04
0.11 ( 0.04
0.08 ( 0.05
0.11 ( 0.01
0.27 ( 0.08
ndb
176 ( 47
48 ( 12
40 ( 12
16 ( 7
12 ( 4
14 ( 5
ndb
30 ( 10
23 ( 6
7.1 ( 4.3
15 ( 9
8.3 ( 3.5
ndb
15 ( 1
17 ( 5
12 ( 5
2.6 ( 2.0
8.4 ( 2.6
25 ( 13
42 ( 25
31 ( 12
17 ( 9
15 ( 8
25 ( 4
7.1 × 103
5.3 × 104
6.7 × 104
1.7 × 105
1.8 × 104
3.2 × 104
<1.0 × 103
4.2 × 103
4.7 × 103
1.2 × 104
7.4 × 103
3.3 × 104
<1.0 × 103
2.8 × 106
2.2 × 106
2.0 × 106
3.8 × 106
1.5 × 106
6.2 × 104
1.0 × 105
1.8 × 105
2.3 × 105
3.5 × 104
1.8 × 105
0.09 ( 0.01
28 ( 20
ndb
ndb
1.4 ( 0.8
2.0 ( 0.8
0.19 ( 0.09
ndb
157 ( 63
178 ( 76
32 ( 6
ndb
4OTP
7OFP
7TFP
7OTP
4TFP
C
G
T
ndb
ndb
0.6 ( 0.4
1.8 ( 1.4
0.16 ( 0.03
0.10 ( 0.04
ndb
130 ( 23
181 ( 122
32 ( 4
42 ( 29
ndb
53 ( 24
1.7 ( 0.3
0.06 ( 0.01
ndb
35 ( 6
188 ( 14
5.3 ( 0.2
ndb
41 ( 7
37 ( 11
24 ( 8
10 ( 2
ndb
ndb
13 ( 2
ndb
ndb
1.5 ( 0.4
4.2 ( 0.2
5.5 ( 1.3
3.8 ( 1.3
0.54 ( 0.04
4.5 ( 1.3
ndb
ndb
0.28 ( 0.08
0.20 ( 0.06
ndb
2.8 ( 0.8
14 ( 5
0.56 ( 0.07
1.4 ( 0.6
ndb
1.8 ( 0.6
2.4 ( 1.4
17 ( 4
41 ( 12
ndb
146 ( 63
118 ( 48
6.0 ( 3.5
3.4 ( 0.9
ndb
a See text for experimental details. b Rates too slow for determination
of kcat and KM independently.
Table 4. Correct Single Nucleotide Extension of
0.7 ( 0.5
51 ( 27
Unnatural Self-Pairsa
k
cat/KM
(min-1 M-1
a See text for experimental details. b Rates too slow for determination
unnatural self-pair
k
cat (min-1
)
KM (µM)
)
of kcat and KM independently.
4OFP
4OTP
7OFP
7TFP
7OTP
4TFP
7.4 ( 0.1
2.4 ( 0.1
8.4 ( 0.4
ndb
1.6 ( 1.2
1.0 ( 0.6
183 ( 12
181 ( 23
80 ( 3
4.0 × 104
1.3 × 104
1.1 × 105
<1.0 × 103
9.4 × 103
2.3 × 105
base pair synthesis (i.e., selective incorporation of the correct
triphosphate opposite a template base) and extension (synthesis
of the next base pair). Each unnatural and natural base was
incorporated into the primer-template substrates shown in Tables
2-4. The efficiency and selectivity of unnatural DNA synthesis
by KF exo- were examined both in the context of self-pair
synthesis (Tables 2 and 3) and self-pair extension (Table 4).
2.3.1. Efficiency of Unnatural Self-Pair Synthesis. Self-
pair synthesis efficiencies for the six unnatural nucleobases vary
over 2 orders of magnitude (Table 2). The most efficiently
formed is the 4TFP self-pair, with a specificity constant (kcat/
KM) of 9.3 × 104 min-1 M-1, which is 500-fold reduced relative
to a dA:dT in the same sequence context (4.7 × 107 min-1 M-1).
The next most efficient are the 7OTP and 7TFP self-pairs,
which are synthesized 5.8- and 8.4-fold slower than the 4TFP
ndb
171 ( 21
4.5 ( 2.0
a See text for experimental details. b Rates too slow for determination
of kcat and KM independently.
triphosphate is found to be inserted opposite 7TFP with a
detectable rate. Thus, it is not possible to measure the selectivity
of 7TFP self-pair synthesis but is only possible to estimate that
it is greater than 11-fold. In all other cases, the unnatural self-
pair is synthesized slower than its most efficiently synthesized
mispair. Despite a somewhat reduced rate of self-pair synthesis
compared to 7TFP, 4OTP is the next most selective unnatural
base in the template. The 4OTP self-pair is synthesized more
efficiently than all its mispairs except that with dT. dTTP is
inserted opposite 4OTP 1.7-fold more efficiently than d(4OTP)-
TP, despite a high KM (181 µM) and because of a relatively
large kcat (1.8 min-1). Relative to dTTP and d(4OTP)TP, dGTP
is inserted opposite 4OTP less efficiently, due to a decreased
self-pair (1.6 × 104 min-1 M-1 and 1.1 × 104 min-1 M-1
,
respectively). Following 7OTP and 7TFP, the next most
efficiently synthesized self-pairs are those of 4OTP and 7OFP,
with rates of 5.9 × 103 min-1 M-1 and 5.2 × 103 min-1 M-1
,
respectively (each approximately 16-fold reduced relative to the
4TFP self-pair). Finally, the 4OFP self-pair is the least
efficiently synthesized of the group with an almost undetectable
synthesis rate, at least 2 orders of magnitude reduced compared
to 4TFP.
k
cat relative to dTTP and an increased KM relative to d(4OTP)-
TP. dATP and dCTP are inserted opposite 4OTP at undetectable
rates. The third most selective self-pair is 7OTP, which is
synthesized from 3.4-fold to at least 16-fold more efficiently
than its mispairs with dA and dC, respectively; however, 7OTP
mispairs with dTTP 7.3-fold more efficiently than it correctly
pairs with d(7OTP)TP. This is the result of an increased kcat
(14 min-1) that overcomes the reduced binding of dTTP relative
to the unnatural triphosphate. 7OTP also directs KF to insert
dGTP with a rate approximately equal to that for insertion of
2.3.2. Fidelity of Unnatural Self-Pair Synthesis. We
determined the fidelity of self-pair synthesis in the context of
single incorporation of an unnatural or a natural triphosphate
opposite the unnatural base in the template (Table 2). Despite
an only modest rate of self-pair synthesis, the 7TFP self-pair
is replicated with the greatest selectivity. In fact, no natural
9
J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003 9641

A R T I C L E S
Henry et al.
the unnatural triphosphate. Despite a comparatively efficient rate
of self-pair synthesis (9.3 × 104 min-1 M-1), the 4TFP self-
pair is not selectively formed due to the remarkably efficient
incorporation of both dGTP (2.7 × 106 min-1 M-1) and dATP
(4.2 × 105 min-1 M-1). In this case, the efficiency of dGTP
insertion results largely from very tight binding of the triphos-
phate (KM ) 0.7 µM). The pyrimidine triphosphates dTTP and
dCTP are inserted less efficiently opposite 4TFP than is
d(4TFP)TP, with rates of 4.7 × 104 min-1M-1 and <1.0 ×
103 min-1M-1, respectively. 4OFP is not selective for self-pair
synthesis primarily because of inefficient self-pair synthesis;
dTTP and dGTP are inserted against 4OFP 9-fold and 12-fold
faster than d(4OFP)TP, respectively. Finally, the least selective
unnatural base, in the template, is 7OFP against which dGTP
is inserted 290-fold more efficiently than d(7OFP)TP. Remark-
ably, the kcat for insertion of dGTP against 7OFP (53 min-1) is
only about 3-fold reduced compared to that for natural base
pair synthesis.12 The 7OFP self-pair is synthesized 1.8-fold less
efficiently than the 7OFP:dT mispair, but 2- and 5-fold more
efficiently than the mispairs with dA and dC, respectively.
the previous first generation unnatural base pairs, such as the
ICS self-pair.
3. Discussion
Despite sharing the same nucleobase scaffold, the six un-
natural bases were found to have very different properties with
respect to self-pair stability and recognition by the polymerase.
This demonstrates the importance of heteroatom substitution,
which affects nucleobase hydrophobicity, polarity, and polariza-
bility.18-22 Several trends in the thermodynamic and kinetic data
are now discussed.
Stability. In all cases, the unnatural self-pairs are more stable
than all possible corresponding mispairs with natural bases. This
thermodynamic selectivity illustrates the strength of the hydro-
phobic effect in duplex DNA, as has been observed with other
unnatural base pairs.27 The most stable self-pair was that of
4OTP, which was only 1.3 °C less stable than a dA:dT pair in
the same sequence context. The 4OTP self-pair was also as
selective for “correct” pairing as the natural bases, with each
mispair between 4OTP and a natural base conferring a Tm of
6.4 °C to 8.9 °C below that of the self-pair.
We also characterized the efficiency of insertion of the
unnatural triphosphates against each natural base in the template
(Table 3). The rates were found to depend most strongly on
which natural base was in the template. The greatest rates were
consistently seen with dG in the template, followed by dT, dA,
and finally dC. The decrease in rates was largely due to kcat
effects. Each unnatural triphosphate, with the exception of
d(4OFP)TP, is efficiently inserted opposite dG with rates that
vary only between 1.5 × 106 min-1 M-1 and 3.8 × 106 min-1
M-1. The specificity constants are only about 20-fold less than
that for natural DNA synthesis. While there was some variation
in the rates of unnatural triphosphate insertion opposite dT, dA,
and dC, this variation was less than 10-fold in all cases. The
only exception was the reasonably efficient insertion of d(4OFP)-
TP opposite dT; d(4OFP)TP was inserted opposite dG, dA, and
dC with efficiencies that were reduced by up to 3 orders of
magnitude compared to the other five unnatural bases.
The absolute stabilities of the mispairs were more varied than
those of the self-pairs, indicating that the unnatural-natural
interbase interactions are more specific than simple hydropho-
bicity. Comparison of 4OFP to 4TFP and 7OFP to 7TFP
reveals the effect of the atom at position 1, namely oxygen or
sulfur, which is presumably disposed in the minor groove. For
each of these four unnatural bases, the self-pair was most stable,
followed by the mispairs with dA, dG, and dC, in that order.
The greater stability of the purine mispairs indicates that packing
interactions may be important. Both the self-pairs and mispairs
(except the mispair with dT) were destabilized by the substitu-
tion of oxygen with sulfur. Since sulfur is expected to be more
polarizable than oxygen, a property favorable for packing
interactions, sulfur’s destabilizing effect relative to oxygen likely
results from unfavorable steric repulsion caused by its greater
size. The four mispairs with dT behaved very differently from
the other mispairs in that sulfur stabilized the mispair. In fact,
the 4TFP:dT and 7TFP:dT mispairs were virtually as stable as
the corresponding mispairs with dA. It is interesting that the
dT mispair behaved so differently from the dC mispair,
considering the similar size and minor groove functionality of
the two pyrimidines (see below).
The effect of heteroatom substitution is also informative in
the series 4OFP, 4OTP, 7OFP, and 7OTP. In this case, each
unnatural nucleobase has an amide carbonyl group oriented
toward the minor groove, with either an oxygen or a sulfur atom
at position 2 or 3. Substitution at position 3 is expected to orient
the heteroatom toward the major groove, which is available for
solvation by water, while substitution at position 2 results in a
different overall dipole and a heteroatom that is expected to be
packed within the duplex. Placement of a heteroatom at position
3 was more stabilizing for the unnatural self-pairs than place-
ment of a heteroatom at position 2. Compared to an oxygen
atom, a sulfur atom at position 3 significantly stabilized the
self-pair, by about 3 °C, while a sulfur atom at position 2 slightly
destabilized the self-pair. The stabilities of mispairs between
the unnatural bases and the natural bases were also sensitive to
2.3.3. Extension of Self-Pair by One Natural Nucleotide.
There is no obvious correlation between self-pair synthesis
efficiency and self-pair extension efficiency except that the most
efficiently synthesized self-pair of this family, 4TFP, is also
the most efficiently extended self-pair, with a specificity constant
of 2.3 × 105 min-1 M-1 (Table 4). Thus, 4TFP exhibits one of
the best rates of single nucleotide extension of an unnatural self-
pair that we have observed, and it is only 200-fold reduced
relative to the rate of extension of a dA:dT in the same sequence
context.12 The second most efficiently extended self-pair of the
group is 7OFP; with a rate of 1.1 × 105 min-1 M-1, the 7OFP
self-pair is extended only 2-fold less efficiently than the 4TFP
self-pair. The 4OFP self-pair is the third most efficiently
extended with a specificity constant of 4.0 × 104 min-1 M-1
,
,
5.8-fold less than 4TFP. With a rate of 1.3 × 104 min-1 M-1
the fourth most efficiently extended self-pair is that of 4OTP.
The least efficiently extended self-pairs are those of 7OTP and
7TFP, with specificity constants of 9.4 × 103 min-1 M-1 and
<1.0 × 103 min-1 M-1, respectively. These rates represent a
more than 25-fold reduction compared to the extension rate of
4TFP. With the exception of 7TFP, all of the self-pairs in this
family are extended at least 5-fold more efficiently than any of
(27) Berger, M.; Ogawa, A. K.; McMinn, D. L.; Wu, Y.; Schultz, P. G.;
Romesberg, F. E. Angew. Chem., Int. Ed. 2000, 39, 2940-2942.
9
9642 J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003

Nucleobase Stability and Polymerase Recognition
A R T I C L E S
the identity of the heteroatom at these two positions. Surpris-
ingly, specific stabilization of the dT mispair by sulfur substitu-
tion, mentioned above for position 1, was also observed at both
positions 2 and 3 (7OTP:dT and 4OTP:dT). The remaining
mispairs, with dA, dG, and dC, were less sensitive to the identity
of the heteroatom at position 3 and were affected differently
by the identity of the heteroatom at position 2 (sulfur destabilized
the mispairs with dC and dG by 2 °C to 3 °C and stabilized the
mispair with dA by 0.9 °C).
polymerase (Tyr766 in KF). Packing between the terminal base
of the primer and Tyr766 may be particularly favorable for the
base of the unnatural triphosphate due to its aromatic surface
area and heteroatoms, which may act to increase its polariz-
ability.13 This tight binding is most pronounced for the 7OTP
triphosphate, which is bound opposite dG with an affinity similar
to that of dCTP. Opposite each of the other three natural bases,
d(7OTP)TP is bound with a Michaelis constant only 4-fold
lower than that for the correct natural triphosphate.
As discussed previously with respect to the unnatural self-
pair extension, tight binding of the unnatural triphosphates does
not produce ternary complexes that lead to efficient catalytic
turnover, as evidenced by the small kcat values for insertion
opposite the natural bases in the template, except in the case of
dG. Opposite dG, the unnatural triphosphates are apparently
tightly bound in a geometry that is suitable for efficient catalysis;
the observed kcat values are only 3-fold to 10-fold smaller than
those typical for correct insertion of a natural dNTP.12 With
the exception of d(4OFP)TP, the specificity constants for
unnatural triphosphate insertion opposite dG vary by less than
It is remarkable that sulfur substitution at each of the three
positions examined results in selective stabilization of the
mispair between the unnatural base and dT. This is unlikely to
be the result of a nonspecific increase in the hydrophobic force,
as dT is not the most hydrophobic of the natural bases. The
only other strong isotropic force that could be increased by sulfur
substitution is polarizability. Thus, it may be that sulfur
substitution results in increased polarizability that favors in-
trastrand and interstrand packing interactions in the unnatural
mispair with dT.18-22 As discussed below, this effect is not
apparent in the kinetic data, implying that different forces control
stability and replication of DNA.28
Replication. Self-pair synthesis efficiency was not strongly
sensitive to the specific heteroatom substitution pattern of the
unnatural base; the Michaelis constants of the unnatural tri-
phosphates varied by only 6-fold, and the value of kcat varied
by only 10-fold. In general, the specificity constants for self-
pair synthesis are very different from those reported for other
unnatural self-pairs indicating that nucleobase “shape” may be
important for efficiency.
Compared to self-pair synthesis, the self-pair extension rates
are more sensitive to heteroatom substitution. With the exception
of the 4TFP self-pair, the presence of any of the unnatural self-
pairs at the primer terminus appears to interfere with the binding
of dCTP. In the case of 4TFP, dCTP seems to pack on the
unnatural self-pair, opposite dG in the template, well enough
to give a Michaelis constant typical of natural synthesis.
However, as the catalyzed chemistry proceeds with a reduced
kcat, it may be inferred that the ternary complex with an unnatural
primer-template is structurally or dynamically different from
the fully natural complex. In this case, tight binding may actually
inhibit replication by stabilizing the complex in a form that
requires rearrangement prior to the chemical step. Nonetheless,
the 4TFP self-pair is extended with an efficiency only 200-
fold reduced relative to natural synthesis.12 This indicates that
unnatural nucleobases with a conserved shape may be optimized
for extension efficiency by variation of the heteroatom substitu-
tion pattern.
Heteroatom substitution has its greatest effect on the poly-
merase-catalyzed synthesis of mispairs between the unnatural
and natural bases. In general, the unnatural triphosphates
described herein are tightly bound opposite all natural bases in
the template. Therefore, the unnatural triphosphates must interact
strongly, but nonspecifically, with the polymerase and the natural
primer-template. Based on the various crystal structures available
for KF and homologous polymerases (type I DNA polymerases
from B. stearothermophilus and T. aquaticus29,30), the base of
the incoming triphosphate packs between the natural base at
the primer terminus and the side chain of a tyrosine in the
2.5-fold, from 1.5 × 106 min-1 M-1 to 3.8 × 106 min-1 M-1
,
within 10-fold to 20-fold of the rate for dCTP insertion.
The specific dipole moment and heteroatom substitution
pattern of the unnatural base had little effect on its overall
efficiency of insertion against dG, with the sole exception of
d(4OFP)TP, which was not inserted at a rate sufficient for
detection (<1 × 103 min-1 M-1). The inefficiency of d(4OFP)-
TP incorporation with dG in the template must not be based on
the shape of the nucleobase but rather on its specific heteroatom
substitution pattern. Interestingly, of all the unnatural bases
evaluated, the heteroatom substitution pattern of 4OFP is most
similar to a natural base, specifically thymidine. Perhaps KF
accepts or rejects unnatural triphosphate substrates based not
on their engagement of favorable interactions but rather based
on their avoidance of unfavorable interactions. This would
explain the specific rejection of d(4OFP)TP, which may be
recognized by the polymerase as dTTP whose insertion opposite
dG the polymerase evolved to avoid.
The unnatural mispairs with dG were synthesized in the
opposite sense by insertion of dGTP opposite the unnatural bases
in the template, with a much broader distribution of rates,
ranging from <103 min-1 M-1 for insertion opposite 7TFP to
>106 min-1 M-1 for insertion opposite 7OFP and 4TFP. The
insertion of dGTP opposite 7OFP and 4TFP is efficient for
different reasons. Against 7OFP in the template, dGTP is
inserted with a remarkably high kcat (53 min-1, only about 3-fold
reduced relative to natural synthesis) and a moderate KM (35
µM). In contrast, against 4TFP in the template, dGTP is inserted
with a relatively low kcat (1.8 min-1) and the remarkably low
KM of 0.7 µM, which is similar to that for insertion of dGTP
opposite dC. This tight binding in the polymerase active site is
not correlated with greater thermal stability as 4TFP:dG was
one of the least stable mispairs in duplex DNA.
Conclusion. In duplex DNA, all six unnatural self-pairs are
stable and selective, demonstrating the suitability of non-H-
bonding forces for the storage of genetic information. Sulfur
substitution at each position examined stabilized mispairs
(29) Kiefer, J. R.; Mao, C.; Braman, J. C.; Beese, L. S. Nature 1998, 391, 304-
307.
(28) Petruska, J.; Goodman, M. F.; Boosalis, M. S.; Sowers, L. C.; Cheong, C.;
(30) Kim, Y.; Eom, S. H.; Wang, J.; Lee, D.-S.; Suh, S. W.; Steitz, T. A. Nature
1995, 376, 612-616.
Ignacio Tinoco, J. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 6252-6256.
9
J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003 9643

A R T I C L E S
Henry et al.
and brine. The organics were dried over anhydrous Na2SO4, and solvents
were removed in Vacuo. Purification via flash column chromatography
on silica gel (30-50% ethyl acetate in hexanes) afforded the desired
product as a mixture of two anomers.
between the unnatural bases and dT. These stabilizing interac-
tions were not apparent in the KM values determined in the gel-
based single nucleotide incorporation assay. In fact, as has been
observed with natural DNA,28 there is no correlation between
thermal stability and the efficiency and fidelity of polymerase-
catalyzed synthesis, in this case of the unnatural self-pairs. This
implies either that the rate-limiting transition state is relatively
early or, more likely, that interactions between the bases of the
nascent pair at a primer terminus in a polymerase active site
are significantly different from those of a base pair in duplex
DNA. Heteroatom substitution of the unnatural nucleobases had
a small effect on self-pair synthesis efficiency, a larger effect
on self-pair extension efficiency, and the largest effect on
polymerase-mediated pairing of an unnatural base with a natural
base. This information is expected to aid in the future design
of unnatural bases that are efficiently and selectively replicated.
Moreover, the data imply that dipole and other electrostatic
effects play an important role in the fidelity of replication of
natural DNA.
General Toluoyl Deprotection Procedure. To a stirred solution
of B (1 equiv) in methanol (0.2 M) was added dropwise NaOMe (2.4
equiv, 0.5 M in CH3OH). After the reaction was complete, it was
quenched by addition of saturated aqueous NH4Cl (excess) and
concentrated. Purification via flash column chromatography on silica
gel (10-40% methanol in ethyl acetate) afforded free nucleoside.
General Phosphoramidite Synthesis Procedure. To a solution of
free nucleoside (1 equiv), azeotroped from a small amount of toluene,
in pyridine (0.2 M) was added a solution of DMTrCl (1.4 equiv) in
pyridine (0.1 M) via syringe pump. After being stirred for 1 h at room
temperature, the reaction mixture was quenched by saturated aqueous
NH4Cl. The reaction mixture was partitioned between ethyl acetate and
brine. The organics were dried over anhydrous Na2SO4 and solvents
were removed in Vacuo. Purification via flash column chromatography
on silica gel (30-70% hexanes in ethyl acetate) afforded the tritylated
â-anomer, which was dissolved in CH2Cl2 (0.1 M) and cooled to 0 °C.
Diisopropylethylamine (1.1 equiv) was added, followed by 2-cyanoethyl
diisopropylaminochlorophosphoramidite (1 equiv). After 15 min, the
reaction mixture was partitioned between ethyl acetate and brine. The
layers were separated, and the aqueous layer was extracted once with
ethyl acetate. The combined organic layers were dried over anhydrous
Na2SO4, filtered, and concentrated. Purification via flash column
chromatography on silica gel (15-60% ethyl acetate in 5% triethyl-
amine/hexanes) afforded the desired phosphoramidite as a mixture of
two diastereomers.
4. Experimental Section
General Methods. All reactions were carried out in oven-dried
glassware under inert atmosphere, unless otherwise noted. All solvents
were dried over 4 Å molecular sieves except dichloromethane, which
was freshly distilled from CaH2. Compound 1A was purchased from
Ubichem PLC (UK). All other reagents were purchased from Sigma-
Aldrich. T4 polynucleotide kinase and Klenow fragment exo- were
purchased from New England Biolabs. Redivue [γ-33P]-ATP was
purchased from Amersham Biosciences. Oligonucleotides were syn-
thesized using an Applied Biosystems Inc. 392 DNA/RNA synthesizer.
DNA synthesis reagents were purchased from Glen Research, Stering,
VA.
Oligonucleotide duplex denaturation temperature measurements were
made in buffer containing 10 mM PIPES, 10 mM MgCl2, and 100 mM
NaCl with a final double-stranded oligonucleotide concentration of 3
µM, using a Cary 300 Bio UV-visible spectrophotometer. Measure-
ments were taken over the range 16-80 °C at 0.5 °C/min intervals.
Melting temperatures were obtained from the derivative method in the
Cary WinUV thermal application software.
Oligonucleotide primers for the kinetic assays were 5′-radiolabeled
with T4 polynucleotide kinase and [γ-33P]-ATP. Primer-templates were
annealed in the reaction buffer by heating to 95 °C followed by slow
cooling to room temperature. Assay conditions included 40 nM primer-
template, 0.1-1.3 nM enzyme, 50 mM Tris-HCl (pH 7.5), 10 mM
MgCl2, 1 mM DTT, and 50 µg/mL acetylated BSA. The reactions were
carried out by combining the DNA-enzyme mixture with an equal
volume (5 µL) of 2X dNTP stock solution, incubating at 25 °C for
1-10 min, and quenching by addition of 20 µL of loading dye
containing 95% formamide and 20 mM EDTA. The reaction mixtures
were resolved by 15% polyacrylamide gel electrophoresis, and the
radioactivity was quantified by means of a PhosphorImager (Molecular
Dynamics) and the ImageQuant program. A plot of kobs versus [dNTP]
was fit to a Michaelis-Menten equation using the program Kaleida-
graph (Synergy Software). The data presented are averages of three
independent determinations.
General Procedure for Bistoluoyl Nucleoside Synthesis. To a
stirred solution of substrate A (1 equiv) in acetonitrile (0.25 M) at room
temperature under argon was added bis(trimethylsilyl)acetamide (1
equiv). After stirring at ambient temperature for 40 min, 3,5-bis-
(toluoyl)-2-deoxyribosyl chloride (1 equiv) was added. The reaction
was brought to 0 °C. To the reaction mixture was added dropwise SnCl4
(20 mmol %). After 1 h, complete dissolution of the 2′-deoxyribofura-
noside had occurred. To the reaction was added EtOAc, and the
resulting solution was successively extracted with saturated NaHCO3
General Triphosphate Synthesis Procedure. The tritylated â-ano-
mer nucleoside was treated with 20% trichloroacetic acid in dichlo-
romethane (0.2 M) for 30 min. The resulting red solution was quenched
by solid sodium bicarbonate (excess) and concentrated in Vacuo.
Purification via flash column chromatography afforded the desired
â-free nucleoside. Proton sponge (1.5 equiv) and the free nucleoside
(1 equiv) were dissolved in trimethyl phosphate (final concentration
0.3 M) and cooled to 0 °C for 5 min. POCl3 (1.05 equiv) was added
dropwise, and the purple slurry was stirred at 0 °C for an additional 2
h. Tributylamine (4 equiv) was added, followed by a solution of
tributylammonium pyrophosphate (2.5 equiv) in DMF (final concentra-
tion 0.15 M). After 1 min, the reaction was quenched by addition of 1
M aqueous triethylammonium bicarbonate (20 vol equiv). The resulting
solution stood for 40 min at 0 °C and was lyophilized and purified by
reverse-phase (C18) HPLC (4-35% CH3CN in 0.1 M Et3N-HCO3,
pH 7.5) to afford the triphosphate as a white solid.
1
Compound 1B. H NMR (300 MHz, CDCl3) δ 7.90 (2H, d, J )
7.8 Hz), 7.83 (2H, d, J ) 7.8 Hz), 7.56 (1H d, J ) 5.6 Hz), 7.48 (1H,
d, J ) 7.4 Hz), 7.18-7.08 (m, 5H), 6.75 (1H, t, J ) 6.6 Hz), 6.52
(1H, d, J ) 7.4 Hz), 6.05 (1H, d, J ) 6.0 Hz), 4.67 (s, 2H), 4.54 (1H,
d, J ) 2.4 Hz), 2.88 (1H, dd, J ) 14.2, 4.8 Hz), 2.40-2.20 (7H, m,
singlets at 2.32, 2.29, 3H each). 13C NMR (100 MHz, CDCl3) δ 154.1,
153.6, 146.5, 135.8, 132.2, 118.7, 117.8, 117.5, 117.4, 117.2, 116.9,
114.8, 114.7, 112.9, 112.7, 112.3, 89.4, 76.4, 73.5, 63.1, 52.3, 27.3;
HRMS (MALDI-FTMS): calcd for C28H25NO6SNa (MNa+), 526.1295;
found, 526.1302.
1
Compound 1C. H NMR (300 MHz, CDCl3) δ 7.81 (1H, d, J )
7.4 Hz), 7.62 (1H, d, J ) 5.6 Hz), 7.45 (1H, d, J ) 6.6, 2.1 Hz), 7.36-
7.22 (11H, m), 6.90-6.70 (5H, m), 6.53 (1H, d, J ) 7.6 Hz), 4.56
(1H, brs), 4.45 (1H, d, J ) 5.0 Hz), 4.22 (1H, d, J ) 7.6 Hz), 3.76
(6H, s), 3.52-3.48 (2H, m), 2.85-2.71 (1H, m), 2.46-2.45 (1H, m).
13C NMR (100 MHz, CDCl3) δ 158.8, 153.3, 149.1, 148.8, 143.3, 136.0,
133.3, 130.5, 129.5, 129.3, 127., 127.2, 113.5, 107.5, 100.9, 87.5, 85.5,
72.3, 63.7, 55.5, 42.6; HRMS (MALDI-FTMS): calcd for C33H32-
NO6S (MH+), 570.1945; found, 570.1942.
9
9644 J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003

Nucleobase Stability and Polymerase Recognition
A R T I C L E S
Compound 1D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -6.20 (d, J ) 17.2 Hz), -10.71 (d, J ) 15.8 Hz),
-21.82 (t, J ) 17.1 Hz).
4.80-4.59 (m, 1H), 4.13 (brs, 1H), 3.73 (s, 6H), 3.60-3.31 (m, 6H),
2.90-2.81 (1H, m), 2.63 (1H, t, J ) 6.2 Hz), 2.48-2.46 (2H, m), 1.21-
1.15 (12H, m). 31P NMR (140 MHz, CDCl3) δ 150.6, 149.7. HRMS
(MALDI-FTMS): calcd for C42H48N3O7PSNa (MNa+), 792.2848;
found, 792.2832.
Compound 4B. 1H NMR (300 MHz, CDCl3) δ 8.00-7.89 (3H, m),
7.77 (0.5H, d, J ) 1.8 Hz), 7.74 (0.5H, d, J ) 2.0 Hz), 7.63 (0.5H, d,
J ) 6.2 Hz), 7.52 (1H, t, J ) 7.2 Hz), 7.30-7.21 (4H, m), 7.11 (1H,
d, J ) 7.8 Hz), 6.89 (0.5H, dd, J ) 5.6, 8.2 Hz), 6.72-6.62 (1H, m),
6.53 (0.5H, d, J ) 7.0 Hz), 6.40 (0.5H, d, J ) 7.4 Hz), 5.64 (1H, t, J
) 6.2 Hz), 4.95 (0.5H, brs), 4.73 (1H, brs), 4.60 (1H, d, J ) 2.0 Hz),
4.59-4.57 (0.5H, m), 3.18-2.88 (1H, m), 2.62-2.40 (1H, m), 2.26
(3H, s), 2.17 (1.5H, s), 2.04 (1.5H, s). HRMS (MALDI-FTMS): calcd
for C28H25NO7Na (MNa+), 510.1523; found, 510.1509.
1
Compound 1E. H NMR (300 MHz, CDCl3) δ 7.84 (0.5H, d, J )
7.8 Hz), 7.76 (0.5H, d, J ) 7.4 Hz), 7.64 (1H, d, J ) 5.0 Hz), 7.44
(2H, brs), 7.41-7.18 (8H, m), 6.85-6.79 (4H, m), 6.72 (1H, d, J )
5.8 Hz), 6.48 (1H, d, J ) 7.4 Hz), 4.78-4.60 (1H, m), 4.21 (1H, brs),
3.81-3.35 [12H, m, singlet at 3.78 (6H)], 2.74-2.60 (1H, m), 2.60
(1H, t, J ) 6.2 Hz), 2.42 (1H, t, J ) 6.2 Hz), 2.40-2.20 (1H, m),
1.40-1.11 (9H, m), 1.02 (3H, d, J ) 6.9 Hz). 31P NMR (140 MHz,
CDCl3) δ 150.3, 149.6. HRMS (MALDI-FTMS): calcd for C42H48N3O7-
PSNa (MNa+), 792.2843; found, 792.2840.
1
Compound 2B. H NMR (300 MHz, CDCl3) δ 7.98 (2H, d, J )
6.0 Hz), 7.90 (1H, d, J ) 6.0 Hz), 7.61 (1H, d, J ) 5.7 Hz), 7.48 (1H,
d, J ) 3.0 Hz), 7.28 (2H, d, J ) 6.6 Hz), 7.22 (2H, d, J ) 6.0 Hz),
6.97 (1H, dd, J ) 3.0, 0.6 Hz), 6.79 (1H, dd, J ) 12, 3.9 Hz), 6.46
(1H, d, J ) 5.7 Hz), 5.65-5.62 (1H, m), 4.78-4.60 (2H, m), 4.60
(1H, t, J ) 2.1 Hz), 2.97-2.91 (1H, m), 2.44 (6H, s), 2.39-2.26 (1H,
m). 13C NMR (100 MHz, CDCl3) δ 166.2, 166.1, 152.9, 148.6, 144.4,
144.2, 143.2, 132.8, 129.9, 129.6, 129.3, 129.2, 16.7, 126.4, 126.3,
107.3, 100.8, 85.1, 82.9, 77.4, 77.1, 76.7, 75.1, 64.3, 39.3, 21.7, 21.6.
HRMS (MALDI): calcd for C28H25NO7Na (MNa+), 510.1523; found,
510.1518.
1
Compound 4C. H NMR (400 MHz, CDCl3) δ 7.89 (1H, d, J )
2.0 Hz), 7.80 (1H, d, J ) 7.4 Hz), 7.43 (2H, d, J ) 6.6 Hz), 7.65-
7.28 (7H, m), 6.80 (4H, d, J ) 8.6 Hz), 6.72 (1H, d, J ) 2.0 Hz), 6.65
(1H, t, J ) 6.4 Hz), 6.37 (1H, d, J ) 7.4 Hz), 4.52 (1H, d, J ) 5.8
Hz), 4.06 (1H, d, J ) 3.6 Hz), 3.71 (6H, s), 3.48-3.42 (3H, m), 2.48-
2.41 (1H, m), 2.28-2.17 (1H, m). 13C NMR (100 MHz, CDCl3) δ
158.8, 153.3, 149.0, 148.8, 144.8, 143.3, 136.0, 135.9, 133.3, 130.3,
128.4, 128.2, 127.8, 127.7, 127.1, 113.4, 107.4, 107.3, 100.8, 86.9,
86.7, 85.4, 85.3, 72.3, 63.7, 55.5, 55.4, 42.7. HRMS (MALDI-
FTMS): calcd for C33H31NO7Na (MNa+), 576.1993; found, 576.1973.
1
Compound 2C. H NMR (300 MHz, CDCl3) δ 7.84 (1H, d, J )
6.0 Hz), 7.38 (2H, d, J ) 3.8 Hz), 7.27-7.24 (6H, m), 7.19 (1H, t, J
) 5.4 Hz), 6.83 (4H, d, J ) 6.0 Hz), 6.26 (1H, d, J ) 4.5 Hz), 5.54
(1H, d, J ) 6.3 Hz), 4.49 (1H, dd, J ) 8.1, 3.6 Hz), 3.97 (1H, dd, J )
5.7, 2.4 Hz), 3.35 (2H, d, J ) 4.8 Hz), 2.44-2.38 (1H, m), 2.19-2.12
(1H, m). 13C NMR (100 MHz, CDCl3) δ 160.3, 159.7, 159.0, 145.0,
144.6, 135.7, 135.6, 130.2, 129.9, 128.2, 127.7, 126.8, 115.4, 112.9,
106.5, 96.1, 86.9, 86.6, 85.7, 70.7, 63.0, 54.5, 41.8. HRMS (MALDI):
calcd for C33H31NO7Na (MNa+), 576.1993; found, 576.1988.
Compound 2D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -6.66 (d, J ) 17.2 Hz), -10.76 (d, J ) 17.1 Hz),
-21.98 (t, J ) 17.4 Hz).
Compound 4D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -6.03 (d, J ) 17.1 Hz), -10.73 (d, J ) 17.2 Hz),
-21.85 (t, J ) 17.2 Hz).
1
Compound 4E. H NMR (300 MHz, CDCl3) δ 7.79 (0.5H, d, J )
7.4 Hz), 7.74 (0.5H, d, J ) 7.4 Hz), 7.69 (1H, s), 7.46 (2H, d, J ) 7.0
Hz), 7.40-7.18 (6H, m), 6.87-6.70 (5H, m), 6.59 (1H, s), 6.26 (1H,
d, J ) 7.4 Hz), 4.69-4.60 (1H, m), 4.20 (1H, s), 3.77 (6H, s), 3.80-
3.37 (7H, m), 2.80-2.59 (1H, m), 2.63 (1H, t, J ) 6.7 Hz), 2.42 (1H,
t, J ) 6.7 Hz), 2.40-2.21 (1H, m), 1.30-1.11 (9H, m), 1.10 (3H, d,
J ) 6.9 Hz). 31P NMR (140 MHz, CDCl3) δ 150.1, 149.6. HRMS
(MALDI-FTMS): calcd for C42H48N3O8PNa (MNa+), 776.3071;
found, 776.3065.
1
Compound 2E. H NMR (300 MHz, CDCl3) δ 7.91 (0.5H, d, J )
7.8 Hz), 7.83 (0.5H, d, J ) 7.8 Hz), 7.47-7.28 (9H, m), 6.96 (1H, s),
6.86 (2H, d, J ) 2.8 Hz), 6.81 (2H, d, J ) 3.0 Hz), 6.70 (1H, dd, J )
25.2, 6.2 Hz), 6.30 (1H, d, J ) 7.8 Hz), 4.68-4.59 (1H, m), 4.21 (1H,
brs), 3.79 (6H, s), 3.78-3.35 (7H, m), 2.80-2.62 (1H, m), 2.61 (1H,
t, J ) 6.4 Hz), 2.43 (1H, t, J ) 6.2 Hz), 2.30-2.18 (1H, m), 1.30-
1.21 (9H, m), 1.19 (3H, d, J ) 6.9 Hz). 31P NMR (140 MHz, CDCl3)
δ 150.2, 149.6. HRMS (MALDI): calcd for C42H48N3O8PNa (MNa+),
776.3071; found, 776.3058.
1
Compound 5B. H NMR (300 MHz, CDCl3) δ 7.96 (d, J ) 8.0
Hz, 2H), 7.90 (d, J ) 7.9 Hz, 2H), 7.67 (d, J ) 5.2 Hz, 1H), 7.55 (d,
J ) 7.6 Hz, 1H), 7.26-7.10 (m, 5H), 6.85 (dd, J ) 8.0, 5.6 Hz, 1H),
6.55 (d, J ) 7.4 Hz, 1H), 5.64 (d, J ) 6.2 Hz, 1H), 4.72 (brs, 2H),
4.61 (s, 1H), 2.91 (dd, J ) 14.0, 5.4 Hz, 1H), 2.37 (s, 3H), 2.33 (s,
3H), 2.40-2.36 (m, 1H). 13C NMR (100 MHz, CDCl3) δ 166.4, 165.9,
158.4, 145.6, 144.5, 144.4, 133.9, 130.3, 129.9, 129.6, 129.3, 127.8,
126.9, 126.5, 124.3, 102.8, 88.5, 85.6, 78.0, 77.4, 76.8, 75.3, 64.6, 39.6,
21.9. HRMS (MALDI-FTMS): calcd for C28H25NO6SNa (MNa+),
526.1300; found, 526.1283.
Compound 3B. To a stirred solution of 2B in toluene (0.1 M) was
added Lawesson’s reagent (1.5 equiv) at room temperature. The
resulting solution was refluxed until the reaction was complete. After
concentration in Vacuo, the residue was purified via flash column
chromatography on silica gel (20-40% ethyl acetate in hexanes) to
afford 3B as a yellow foam. HRMS (MALDI-FTMS): calcd for
C28H25NO6SNa (MNa+), 526.1300; found, 526.1272.
1
Compound 5C. H NMR (300 MHz, CDCl3) δ 7.86 (d, J ) 7.8
Hz, 1H), 7.65 (d, J ) 3.0 Hz, 1H), 7.50-7.11 (m, 11H), 6.83 (d, J )
8.6 Hz, 4H), 6.50 (d, J ) 7.6 Hz, 1H), 4.61 (brs, 1H), 4.38-4.20 (m,
1H), 3.77 (s, 6H), 3.54-3.41 (m, 2H), 2.78-2.72 (m, 1H), 2.36-2.26
(m, 1H). 13C NMR (100 MHz, CDCl3) δ 158.7, 158.5, 145.6, 144.8,
135.9, 135.8, 134.0, 130.4, 129.8, 128.4, 128.2, 127.2, 124.4, 113.4,
103.7, 86.9, 86.8, 85.7, 72.2, 63.7, 55.5, 42.7. HRMS (MALDI-
FTMS): calcd for C33H31NO6SNa (MNa+), 592.1764; found, 592.1744.
Compound 5D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -7.64 (d, J ) 17.2 Hz), -10.60 (d, J ) 17.2 Hz),
-22.14 (t, J ) 17.1 Hz).
Compound 5E. 1H NMR (300 MHz, CDCl3) δ 7.73 (dd, J ) 14.0,
7.0 Hz, 1H), 7.38 (d, J ) 5.2 Hz, 1H), 7.42-7.05 (m, 10H), 6.76 (d,
J ) 8.6 Hz, 4H), 6.72-6.60 (m, 1H), 6.37 (d, J ) 7.4 Hz, 1H), 4.62-
4.50 (m, 1H), 4.14 (brs, 1H), 3.70 (s, 6H), 3.65-3.23 (m, 6H), 2.70-
2.53 (m, 2H), 2.38 (t, J ) 6.7 Hz, 1H), 2.30-2.12 (m, 1H), 1.25-1.15
(m, 9H), 1.10 (d, J ) 7.0 Hz, 3H). 31P NMR (140 MHz, CDCl3) δ
150.15, 149.48. HRMS (MALDI-FTMS): calcd for C42H48N3O7PSNa
(MNa+), 792.2848; found, 792.2827.
1
Compound 3C. H NMR (300 MHz, CDCl3) δ 8.42 (d, J ) 7.4
Hz, 1H), 7.44 (d, J ) 2.0 Hz, 1H), 7.37-7.04 (11H, m), 6.75 (d, J )
9.0 Hz), 6.49 (d, J ) 7.4 Hz, 1H), 4.53 (d, J ) 5.8 Hz, 1H), 4.02 (d,
J ) 7.2 Hz, 1H), 3.69 (s, 6H), 3.60-3.31 (m, 2H), 2.87 (brs, 1H),
2.8-2.62 (m, 1H), 2.30-2.21 (m, 1H). 13C NMR (100 MHz, CDCl3)
δ 174.2, 158.8, 154.5, 145.6, 144.7, 135.7, 135.6, 133.1, 130.4, 130.3,
128.4, 128.3, 127.4, 113.5, 110.3, 101.1, 90.7, 87.2, 86.5, 78.0, 77.4,
76.7, 70.1, 62.3, 55.6, 42.2. HRMS (MALDI-FTMS): calcd for C33H31-
NO6SNa (MNa+), 592.1764; found, 592.1733.
Compound 3D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -9.25 (d, J ) 17.1 Hz), -10.79 (d, J ) 17.0 Hz),
-22.56 (t, J ) 17.2 Hz).
Compound 3E. 1H NMR (300 MHz, CDCl3) δ 8.50 (dd, J ) 18.0,
7.4 Hz, 1H), 7.48 (d, J ) 1.8 Hz, 1H), 7.40-7.05 (m, 12H), 6.79 (d,
J ) 3.6 Hz, 2H), 6.75 (d, J ) 3.5 Hz, 2H), 6.44 (d, J ) 5.8 Hz, 1H),
9
J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003 9645

A R T I C L E S
Henry et al.
Compound 6B. To a stirred solution of 4B in toluene (0.1 M) was
added Lawesson’s reagent (1.2 equiv) at room temperature. The
resulting solution was refluxed until the reaction was complete. After
concentration in Vacuo, the residue was purified via flash column
chromatography on silica gel (10-30% ethyl acetate in hexanes) to
afford 6B as a yellow foam. 1H NMR (300 MHz, CDCl3) δ 8.18-7.88
(5H, m), 7.67 (1H, d, J ) 8.4 Hz), 7.60 (1H, dd, J ) 7.7, 0.9 Hz),
7.40-7.28 (3H, m), 7.19 (1H, d, J ) 8.1 Hz), 7.02 (1H, d, J ) 7.4
Hz), 6.90-6.80 (1H, m), 5.74-5.68 (1H, m), 5.17 (0.5H, t, J ) 4.3
Hz), 4.92-4.82 (0.5H, m), 4.80-4.62 (2H, m), 3.45-3.28 (1H, m),
2.50 (3H, s), 2.49 (1.5H, s), 2.43 (1.5H, s), 2.50-2.31 (1H, m). HRMS
(MALDI-FTMS): calcd for C28H25NO6SNa (MNa+), 526.1300; found,
526.1291.
(100 MHz, CDCl3) δ 164.2, 158.8, 154.4, 140.9, 144.7, 135.9, 135.8,
130.9, 130.3, 129.3, 128.4, 128.3, 128.2 128.1, 127.9, 127.2, 113.5,
113.4, 107.4, 105.5, 93.2, 89.6, 86.8, 73.1, 64.3, 55.5, 42.9, 29.4. HRMS
(MALDI-FTMS): calcd for C33H31NO6SNa (MNa+), 592.1764; found,
592.1764.
Compound 6D. 31P NMR (140 MHz, 50 mM Tris, 2 mM EDTA,
pH 7.5 in D2O) δ -7.54 (d, J ) 18.5 Hz), -10.59 (d, J ) 17.2 Hz),
-22.04 (t, J ) 17.2 Hz).
1
Compound 6E. H NMR (300 MHz, CDCl3) δ 8.50 (1H, d, J )
7.8 Hz), 7.90 (1H, s), 7.47 (1H, d, J ) 7.9 Hz), 7.39-7.30 (9H, m),
6.84-6.81 (4H, m), 6.69 (1H, s), 6.54 (1H, d, J ) 7.0 Hz), 4.78-4.72
(1H, m), 4.22 (1H, d, J ) 4.1 Hz), 3.81 (6H, s), 3.67 (1H, dd, J )
10.8, 2.6 Hz), 3.60-3.52 (4H, m), 3.47 (1H, dd, J ) 10.5, 2.6 Hz),
2.90-2.85 (1H, m), 2.63 (2H, t, J ) 6.2 Hz), 2.48-2.46 (1H, m), 1.21-
1.15 (12H, m). 31P NMR (140 MHz, CDCl3) δ 149.9, 148.8. HRMS
(MALDI-FTMS): calcd for C42H48N3O7PSNa (MNa+), 792.2848;
found, 792.2836.
1
Compound 6C. H NMR (300 MHz, CDCl3) δ 8.47 (1H, dd, J )
11.0, 5.4 Hz), 7.96 (1H, d, J ) 9.9 Hz), 7.52 (2H, dd, J ) 10.7, 6.0
Hz), 7.44-7.31 (8H, m), 6.95-6.84 (4H, m), 6.78 (1H, d, J ) 10.2
Hz), 6.70 (1H, dd, J ) 11.1, 5.4 Hz), 4.66 (1H, t, J ) 5.1 Hz), 4.18
(1H, d, J ) 6.6 Hz), 3.89 (6H, s), 3.73 (2H, t, J ) 8.7 Hz), 3.57 (1H,
t, J ) 8.1 Hz), 2.92-2.87 (1H, m), 2.42 (1H, t, J ) 4.5 Hz). 13C NMR
JA035398O
9
9646 J. AM. CHEM. SOC. VOL. 125, NO. 32, 2003