The effect of minor-groove hydrogen-bond acceptors and donors on the stability and replication of four unnatural base pairs

Article abstract of DOI:10.1021/ja034099w

The stability and replication of DNA containing self-pairs formed between unnatural nucleotides bearing benzofuran, benzothiophene, indole, and benzotriazole nucleobases are reported. These nucleobase analogues are based on a similar scaffold but have dif

Full text of DOI:10.1021/ja034099w

Published on Web 04/29/2003
The Effect of Minor-Groove Hydrogen-Bond Acceptors and
Donors on the Stability and Replication of Four Unnatural
Base Pairs
Shigeo Matsuda, Allison A. Henry, Peter G. Schultz, and Floyd E. Romesberg*
Contribution from the Department of Chemistry, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037
Received January 9, 2003; E-mail: floyd@scripps.edu
Abstract: The stability and replication of DNA containing self-pairs formed between unnatural nucleotides
bearing benzofuran, benzothiophene, indole, and benzotriazole nucleobases are reported. These nucleobase
analogues are based on a similar scaffold but have different hydrogen-bond donor/acceptor groups that
are expected to be oriented in the duplex minor groove. The unnatural base pairs do not appear to induce
major structural distortions and are accommodated within the constraints of a B-form duplex. The differences
between these unnatural base pairs are manifest only in the polymerase-mediated extension step, not in
base-pair stability or synthesis. The benzotriazole self-pair is extended with an efficiency that is only 200-
fold less than a correct natural base pair. The data are discussed in terms of available polymerase crystal
structures and imply that further modifications may result in unnatural base pairs that can be both efficiently
synthesized and extended, resulting in an expanded genetic alphabet.
Introduction
Inter-nucleobase hydrogen bonding (H-bonding) and shape
complementarity are important for the stability and faithful
replication of DNA. However, it is not clear whether Watson-
Crick base-pairing interactions are a unique solution for the
storage of genetic information or if the specific forces that
control base-pair stability in duplex DNA are the same as those
that are important for correct base pairing during replication,
where protein-DNA interactions are also critical. Evaluation
of the stability and enzymatic synthesis of DNA containing
unnatural nucleobase analogues provide insights into these
issues1-3 and is also useful for designing unnatural base pairs
to supplement the genetic alphabet.4-11 Unnatural base pairs
with novel chemical functionality, such as reactive centers or
redox-active moieties would expand the chemical potential of
DNA and allow for the synthesis of novel polymers with a high
degree of sequence and length control. An expanded genetic
Figure 1. ICS self-pair.
alphabet would also lay the foundation for the in vivo storage
and retrieval of increased information, which might eventually
be used to expand the genetic code of a living organism.11,12
In previous reports, we systematically evaluated a variety of
self-pairs formed between a single unnatural nucleobase that is
predominantly hydrophobic and does not bear an H-bonding
pattern or shape that is complementary to the natural bases.5-11
It should be emphasized that self-pairs, as opposed to hetero-
pairs, do not limit efforts to expand the genetic alphabet, but
rather facilitate the effort by minimizing the potential for
mispairing. Significant progress was made with unnatural
nucleosides derived from the isocarbostyryl (ICS) framework
(Figure 1).5-8 A variety of these self-pairs were identified, and
these self-pairs are as stable or even more stable than natural
base pairs in duplex DNA. Furthermore, several can also be
synthesized by wild-type polymerases (i.e., enzymatic extension
of a primer by incorporation of an unnatural dNTP opposite
the unnatural base in the template) with reasonable efficiency
and selectivity. However, after synthesis of any of the unnatural
base pairs, the newly formed 3′-primer terminus cannot be
efficiently extended by incorporation of the next correct natural
(1) Kool, E. T. Chem. ReV. 1997, 97, 1473-1487.
(2) Kool, E. T. Biopolymers 1998, 48, 3-17.
(3) Guckian, K. M.; Krugh, T. R.; Kool, E. T. Nat. Struct. Biol. 1998, 5, 954-
959.
(4) Lutz, M. J.; Held, H. A.; Hottiger, M.; Hu¨bscher, U.; Benner, S. A. Nucleic
Acids Res. 1996, 24, 1308-1313.
(5) McMinn, D. L.; Ogawa, A. K.; Wu, Y.; Liu, J.; Schultz, P. G.; Romesberg,
F. E. J. Am. Chem. Soc. 1999, 121, 11585-11586.
(6) Ogawa, A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P. G.; Romesberg,
F. E. J. Am. Chem. Soc. 2000, 122, 3274-3287.
(7) Berger, M.; Ogawa, A. K.; McMinn, D. L.; Wu, Y.; Schultz, P. G.;
Romesberg, F. E. Angew. Chem., Int. Ed. 2000, 39, 2940-2942.
(8) Yu, C.; Henry, A. A.; Schultz, P. G.; Romesberg, F. E. Angew. Chem., Int.
Ed. 2002, 41, 3997-4000.
(9) Ogawa, A. K.; Wu, Y.; Berger, M.; Schultz, P. G.; Romesberg, F. E. J.
Am. Chem. Soc. 2000, 122, 8803-8804.
(10) Wu, Y.; Ogawa, A. K.; Berger, M.; McMinn, D. L.; Schultz, P. G.;
Romesberg, F. E. J. Am. Chem. Soc. 2000, 122, 7621-7632.
(11) Tae, E. L.; Wu, Y.; Xia, G.; Schultz, P. G.; Romesberg, F. E. J. Am. Chem.
Soc. 2001, 123, 7439-7440.
(12) Bain, J. D.; Switzer, C.; Chamberlin, A. R.; Benner, S. A. Nature 1992,
356, 537-539.
9
6134
J. AM. CHEM. SOC. 2003, 125, 6134-6139
10.1021/ja034099w CCC: $25.00 © 2003 American Chemical Society

Stability and Replication of Unnatural Base Pairs
A R T I C L E S
Deprotection of silyl groups by treatment with n-TBAF gave
9, which was converted to the free nucleoside 10 with heating
under basic conditions (K2CO3/MeOH/H2O/reflux). The un-
natural nucleoside that had a BTz base moiety was synthesized
according to a previously reported procedure.16 The phosphora-
midites for solid-phase DNA synthesis were obtained by
standard dimethoxytritylation and phosphitylation. The assign-
ment of â-stereochemistry at C1′ for BFr, BTp, and IN was
based on COSY and NOESY experiments. The corresponding
triphosphates were obtained as described previously.6 The
oligonucleotides were synthesized on a DNA synthesizer, using
standard solid-phase â-cyanoethyl phosphoramidite chemistry.
After deprotection under basic conditions, the crude oligonucle-
otides were purified with gel electrophoresis and confirmed by
matrix-assisted laser desorption ionization-time-of-flight mass
spectroscopy.
Figure 2. Unnatural bases BFr, BTp, IN, and BTz (from left to right).
dNTP. In general, the rate of extension is 3 or 4 orders of
magnitude slower than natural base-pair extension. Thus, we
have inferred that there are different and more-selective criteria
governing extension than those critical for self-pair stability or
synthesis. A thorough examination of the nucleobase properties
that contribute to efficient extension is critical for the design
of self-pairs that will expand the genetic alphabet.
To determine whether the unnatural self-pairs affect the
overall structure of duplex DNA, each was incorporated into
the complementary oligonucleotides 5′-GCGATGXGTAGCG-
3′ and 5′-CGCTACYCATCGC-3′ at positions X and Y, and
circular dichroism (CD) spectra were measured (Figure 3). The
spectra are similar and indicate a right-handed helix with a
structure resembling that observed in the same duplex with a
natural base pair (X ) dG, Y ) dC). The rotational strength of
the duplex containing the BTz self-pair is most similar to that
for the fully native natural duplex, whereas the duplexes
containing the other self-pairs were slightly stronger. Thus, the
unnatural bases do not appear to induce major structural
distortions and are able to form self-pairs within the constraints
of a B-form duplex. Therefore, any differences in base-pair
stability or polymerase recognition do not appear to result from
an inability of the self-pairs to be accommodated in duplex
DNA, but rather must result from more local structural or
electronic effects that are unique to each self-pair.
It is commonly accepted that an important criterion for
efficient base-pair extension is the presence of an H-bond
acceptor positioned appropriately in the developing minor
groove of the nascent DNA duplex.13 Formation of an H-bond
between the primer terminus and the polymerase may be critical
for achieving the requisite geometry at the primer terminus for
efficient continued synthesis. In this manner, the polymerase
ensures that primers terminating in a correct base pair are
efficiently extended while exonuclease activity is competitive
with extension for primers terminating in a mispair. Although
the ICS analogues have a carbonyl group that might act as an
H-bond acceptor in the developing minor groove, interbase
hydrophobic interactions within the self-pair may result in a
3′OH orientation that is less than optimal. To evaluate these
issues, we synthesized unnatural nucleoside analogues with
differently oriented H-bond acceptors and donors. To examine
both H-bond donors and acceptors, while maintaining an
aromatic glycosidic linkage, a [5 + 6] ring fusion was chosen
as the unnatural nucleobase scaffolding (Figure 2). We now
report the stability and replication properties of the self-pairs
of benzofuran (BFr), benzothiophene (BTp), indole (IN), and
benzotriazole (BTz).
To evaluate the stability of the unnatural base pairs in duplex
DNA, UV melting studies were performed using the same 13
mer oligonucleotides (Table 1). The duplex melting temperature
(Tm) ranged from 59.2 °C (X ) dA, Y ) dT) to 61.8 °C (X )
dC, Y ) dG) for duplexes containing natural base pairs, and
from 44.8 to 55.4 °C for duplexes containing mispairs between
two natural bases. For duplexes containing unnatural self-pairs,
the value of Tm was almost independent of the atom that was
proposed to be located in the minor groove. The BTp self-pair
was relatively more stable (52.2 °C), and IN and BFr self-
pairs were relatively less stable (51.7 and 50.7 °C, respectively).
In addition, neither the presence of an imine moiety nor the
increased polarity/polarizability of BTz significantly affected
duplex stability (Tm ) 51.2 °C). This independence of Tm on
the minor-groove group (ether, amine, thioether, or imine)
implies that duplex formation does not strongly alter the
solvation of the unnatural base pairs. This likely results from
the minor-groove “spine of solvation”,17 which preserves the
accessibility of minor-groove moieties to water upon duplex
formation.
Results and Discussion
The BFr and BTp nucleosides were synthesized as shown
in Scheme 1. Lithiation of 2,3-benzofuran (1a) and ben-
zothiophene (1b), followed by coupling with 3,5-O-((1,1,3,3-
tetraisopropyl)disiloxanediyl)-2-deoxy-D-ribono-1,4-lactone,14 af-
forded the corresponding hemiacetals, which were subsequently
reduced with an excess of Et3SiH and BF3‚OEt2 and then
deprotected with n-TBAF to provide the desired free nucleosides
2a (36%, 3 steps) and 2b (19%, 3 steps), respectively. For the
synthesis of the unnatural nucleoside (10) that has an IN base
moiety, 1-(phenylsulfonyl)indole (7) was lithiated with n-
butyllithium and added to 2-deoxy-3,5-O-(tetraisopropyldisi-
loxane-1,3-diyl)-D-erythropentofuranose.15 This coupling reac-
tion gave the corresponding diol in 35% yield and the
subsequent ring-closure reaction, which was effected by treat-
ment with 1,1′-azobis(N,N-dimethylformamide) and n-tribu-
tylphosphine, afforded the â-anomer of 8 in 57% yield.
In all cases, the stability of the unnatural self-pair was greater
than all possible mispairs in the sequence context that was
examined. With the exception of those involving dC, mispairs
(13) Guo, M.-J.; Hildbrand, S.; Leumann, C. J.; McLaughlin, L. W.; Waring,
M. J. Nucleic Acids Res. 1998, 26, 1863-1869.
(14) Wichai, U.; Woski, S. A. Org Lett. 1999, 1, 1173-1175.
(15) Yokoyama, M.; Ikeue, T.; Ochiai, Y.; Momotake, A.; Yamaguchi, K.; Togo,
H. J. Chem. Soc., Perkin Trans. 1 1998, 2185-2191.
(16) Kazimierczuk, Z.; Seela, F. HelV. Chim. Acta 1990, 73, 316-325.
(17) Saenger, W. Principles of Nucleic Acid Structure, 2nd ed.; Springer-
Verlag: New York, 1983.
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6135

A R T I C L E S
Matsuda et al.
Scheme 1 a
a (a) n-BuLi, -78 °C, then 6, THF; (b) BF3‚OEt2/Et3SiH/CH2Cl2, -78 °C; (c) n-TBAF/THF; (d) DMTrCl, DMAP, Et3N, pyridine; (e) NC(CH2)2OP(NiPr2)2,
1H-tetrazole, CH3CN, (f) POCl3, n-Bu3N, proton sponge, trimethyl phosphate, (Bu3NH)3(H4P2O7).
Figure 3. CD spectra of duplex DNA containing self-pairs or dG:dC. See text for details.
between an unnatural base and a natural base gave similar Tm
values, differing by only 2.5 °C (47.3 °C, compared to 49.8
°C). In general, the mispairing of the unnatural bases with
natural purine bases resulted in slightly more-stable duplexes.
This likely reflects increased intrastrand aromatic stacking
interactions. The mispairs with dC were significantly less stable,
with Tm values of 44.0-47.0 °C. This may result from the
diminished aromatic surface area and increased hydrophilicity
of dC, relative to the other natural nucleobases.18
The unnatural nucleobases were also evaluated as substrates
for the exonuclease-deficient Klenow fragment of Escherichia
coli DNA polymerase I (KF). Steady-state kinetic constants were
determined by measuring the initial velocities (<15% conver-
sion) of radiolabeled primer extension reactions.6 The reactions
were analyzed by denaturing polyacrylamide gel electrophoresis
and quantified with a PhosphorImager (Molecular Dynamics).
The initial velocities were plotted versus [dNTP] and fit to the
Michaelis-Menten equation to determine kcat and KM. Unnatural
nucleobases were assayed both in template DNA and as
triphosphates. The steady-state kinetic constants for single
nucleotide incorporation are reported in Table 2.
The efficiency of self-pair synthesis was only modestly
dependent on the unnatural base. Of the four self-pairs, BTp is
the most efficiently synthesized with a specificity constant (kcat/
KM) of 3.3 × 105 min-1 M-1, 136 times less efficient than the
synthesis of a dA:dT base pair in the same sequence context
(∼4.5 × 107 min-1 M-1). The self-pair synthesis rate of BFr
is 2.8 × 105 min-1 M-1, 1.2 times lower than that of BTp. The
(18) Shih, P.; Pedersen, L. G.; Gibbs, P. R.; Wolfenden, R. J. Mol. Biol. 1998,
280, 421-430.
9
6136 J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003

Stability and Replication of Unnatural Base Pairs
A R T I C L E S
Table 1. Denaturation Temperatures for Duplex DNA Containing
BFr, BTp, IN, and BTza
triphosphate interacts with the template base as well as a water
molecule. The observed independence of the insertion rates on
unnatural dNTP implies that these interactions are not significant
for unnatural base-pair synthesis, or that they are sufficiently
satisfied by bound water molecules.
X
Y
Tm, °C
X
Y
Tm, °C
In three of the four cases, the unnatural self-pairs are
synthesized selectively, relative to all possible mispairs, albeit
with fidelities that are compromised, relative to the natural base
pairs. For each unnatural base in the template, dATP is inserted
most competitively, with rates varying by a factor of 10, from
3.2 × 104 min-1 M-1 (for IN) to 4.2 × 105 min-1 M-1 (for
BTz). In the case of BTz, dATP is inserted 5.6 times faster
than dBTzTP. These efficiencies result in part from KM effects,
because dATP is consistently the most tightly bound dNTP. This
likely results both from the hydrophobicity of adenine,18 which
favors packing opposite the predominantly hydrophobic un-
natural base, and from the heteroatom substitution of the purine
scaffolding, which favors packing with the base at the primer
terminus.10,21,22
BFr
BFr
BFr
BFr
BFr
BFr
C
G
T
A
50.7
44.0
48.7
47.7
48.7
IN
IN
IN
IN
IN
IN
C
G
T
51.7
45.8
49.7
48.8
49.8
A
BTp
BTp
BTp
BTp
BTp
BTp
C
G
T
A
52.2
47.0
49.1
48.2
48.2
BTz
BTz
BTz
BTz
BTz
BTz
C
G
T
A
51.2
44.3
48.7
47.3
48.3
a Determined in 10 mM PIPES, 10 mM MgCl2, 100 mM NaCl, pH 7.
For reference, Tm ) 59.2 °C for X ) dA, Y ) dT and Tm ) 61.8 °C for
X ) dC, Y ) dG.
Table 2. Efficiency of Self-Pair Synthesis and Selectivity Opposite
Unnatural Bases in the Template
The next most efficiently formed mispair is that resulting from
dTTP insertion opposite each unnatural base in the template.
The mispairs were synthesized, relative to the correct self-pairs,
with 3-, 10-, 30-, and 31-fold reduced efficiency for the BTz,
BTp, IN, and BFr, respectively. In the case of dTTP insertion,
an order-of-magnitude gain in selectivity is observed in the IN
and BFr bases, compared to that of the BTz base, and there is
no significant difference between IN and BFr. Against each
unnatural base in the template, dGTP is the next most efficiently
incorporated, although less so than dATP or dTTP, and 4-, 88-,
140-, and 275-fold less efficiently than correct self-pair synthesis
for BTz, BFr, IN, and BTp, respectively. In this case, a 20-
fold gain in selectivity is observed for BFr, compared to BTz;
a 35-fold gain is observed for IN, and a 70-fold gain is observed
for BTp. Finally, dCTP is the triphosphate that is least efficiently
incorporated opposite each unnatural base in the template, with
self-pair synthesis selectivities of 16-, 140-, 175-, and 330-fold
for BTz, IN, BFr, and BTp, respectively. In the case of dCTP
insertion, the selectivities represent gains of 9- to 20-fold for
IN, BFr, and BTp, compared to BTz.
The ability of KF to extend a primer that terminates at each
of the unnatural self-pairs (by insertion of dCTP opposite dG)
was also examined (Table 3). Unlike with stability and synthesis,
the unnatural bases show a marked variation in extension rates,
depending on the nature of the atom disposed toward the
developing minor groove. Sulfur substitution results in a BTp
self-pair that is extended with an efficiency of only 7.4 × 102
min-1 M-1, which results from both a low kcat value (0.17
min-1) and a large KM value (230 µM). Oxygen and NH
substitution result in moderate increases in kcat but have little
effect on KM. Remarkably, the BTz self-pair is extended with
a kcat/KM value of 1.7 × 105 min-1 M-1, which is 60-230 times
more efficient than the other members of the family. The
increase in rate is due to a large increase in kcat (by a factor of
17-82) and a more modest decrease in KM (by a factor of
X
Y
kcat (min-1
)
KM (µM)
kcat/KM (min-1 M-1
)
BFr
BFr
BFr
BFr
BFr
BFr
A
19 ( 6
67 ( 5
15 ( 5
283 ( 105
86 ( 13
147 ( 18
2.8 × 105
1.0 × 105
1.6 × 103
3.2 × 103
9.1 × 103
1.5 ( 0.5
0.4 ( 0.1
0.28 ( 0.06
1.3 ( 0.6
C
G
T
BTp
BTp
BTp
BTp
BTp
BTp
A
19 ( 5
2.9 ( 0.4
nda
0.17 ( 0.08
4.9 ( 0.6
57 ( 30
41 ( 5
3.3 × 105
7.0 × 104
<1.0 × 103
1.2 × 103
3.2 × 104
C
nda
G
141 ( 15
154 ( 56
T
IN
IN
IN
IN
IN
IN
A
C
G
T
10 ( 4
0.9 ( 0.2
nda
69 ( 5
27 ( 7
nda
1.4 × 105
3.2 × 104
<1.0 × 103
<1.0 × 103
4.6 × 103
nda
nda
0.9 ( 0.7
197 ( 56
BTz
BTz
BTz
BTz
BTz
BTz
A
16 ( 1
209 ( 93
11 ( 1
7.5 × 104
4.2 × 105
4.7 × 103
1.8 × 104
2.5 × 104
4.8 ( 0.6
1.0 ( 0.3
1.6 ( 0.2
3.8 ( 0.4
C
219 ( 44
91 ( 8
G
T
153 ( 21
a Rates too slow for determination of kcat and KM independently.
IN self-pair is synthesized with an efficiency (1.4 × 105 min-1
M-1) that is reduced by a factor of 2.4, relative to that of the
BTp self-pair. Finally, the BTz self-pair is the least efficiently
synthesized, with a rate of 7.5 × 104 min-1 M-1, which is 4.4
times lower than that of the BTp self-pair, and 600 times lower
than that of natural synthesis. The moderate variation in self-
pair synthesis rates demonstrates that, as in base-pair stability,
the nature of the H-bond donor/acceptor at the position in the
nucleobase R to the C-glycosidic bond is not critical for
unnatural base-pair synthesis.
The crystal structures of two type I DNA polymerases are
available (from Bacillus stearothermophilus and Thermus
aquaticus).19,20 Both polymerases are highly homologous to KF,
in regard to sequence, structure, and function. The structures
show that the heteroatom H-bond acceptor of the bound
(19) Kiefer, J. R.; Mao, C.; Braman, J. C.; Beese, L. S. Nature 1998, 391, 304-
307.
(20) Kim, Y.; Eom, S. H.; Wang, J.; Lee, D.-S.; Suh, S. W.; Steitz, T. A. Nature
1995, 376, 612-616.
(21) Bugg, C. E. 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.
(22) Dahl, T. Acta Chem. Scand. 1994, 48, 95-106.
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6137

A R T I C L E S
Matsuda et al.
Scheme 2 a
a (a) n-BuLi, -78 °C, then 14, THF; (b) 1,1′-azobis(N,N-dimethylformamide), n-Bu3P, benzene; (c) n-TBAF/THF; (d) K2CO3, MeOH, H2O, reflux; (e)
DMTrCl, DMAP, Et3N, pyridine; (f) NC(CH2)2OP(NiPr2)2, 1H-tetrazole, CH3CN; (g) POCl3, n-Bu3N, proton sponge, trimethyl phosphate, (Bu3NH)3(H4P2O7).
Table 3. Correct Extension of Unnatural Self-Pairsa
lenged24), whereas extension of a mispair is limited by the rate
of the bond formation step (nucleophilic displacement of
pyrophosphate from the incoming dNTP by the primer 3′OH).
The difference in the rate-limiting step for correct or mispair
extension results in different elemental effects: kcat(dNTP)/kcat-
(R(S)dNTP) is fully manifested for mispair extension (∼17) but
partially masked for correct pair extension (∼2). To determine
which step limits the rate of BTz self-pair extension, the kinetic
experiments previously described with dCTP substrates were
repeated with a mixture of R and S diastereomers of 5′-O-1-
thiotriphosphate (Trilink Biotechnologies). Extension of a correct
dA:dT base pair showed an R-thiotriphosphate elemental effect
of 2.8, in good agreement with previous reports. Extension of
the BTz self-pair showed an elemental effect of 34. This
observation implies that nucleophilic attack of the 3′OH of the
unnatural nucleotide at the primer terminus on the incoming
dNTP is fully rate-limiting. Thus, even with the most efficiently
extended unnatural base pair, the primer terminus is still not
optimally oriented for continued synthesis. Importantly, this also
implies that further self-pair derivatization that optimizes
positioning of the 3′OH group should directly translate to more-
efficient self-pair extension.
unnatural
self-pair
kcat (min-1
)
KM (µM)
kcat/KM (min-1 M-1
)
BFr
BTp
IN
0.81 ( 0.05
0.17 ( 0.04
0.4 ( 0.1
14 ( 3
294 ( 9
230 ( 49
207 ( 30
85 ( 20
2.8 × 103
7.4 × 102
1.8 × 103
1.7 × 105
BTz
a dCTP incorporation opposite G in the template. Extension with an
incorrect natural dNTP was, in every case, <3 × 102 min-1 M-1
.
2-3.5). This self-pair is extended only 200-fold less efficiently
than a correct natural base pair. In the context of the high
extension rate, it is interesting to recall that, although each self-
pair appears to be accommodated in B-form duplex DNA, the
CD spectrum of DNA containing the BTz self-pair was the most
similar to that of fully natural DNA. Thus, the increased
extension rate may have its origin in subtle structural effects
that are also reflected in the duplex CD strength.
To help understand what limits the rate of BTz self-pair
extension, we evaluated the R-thiotriphosphate elemental ef-
fect.23 In general, KF-mediated extension of a primer that
terminates at a correct pair between two canonical bases is rate-
limited by a nonchemical step (which is proposed to be a protein
conformational change,23 although this has been recently chal-
The polymerase crystal structures show that the nucleobase
at the primer terminus packs against the nucleobase of the
incoming triphosphate and, along with the template nucleobase,
is tightly packed by hydrophobic side chains of the protein.19,20
This binding pocket is thought to exert a steric selection against
the aberrant structures of mispaired bases. The structures also
reveal sequence-independent H-bonds between the polymerase
(Arg573) and the purine N3 or pyrimidine O2 atoms of the
(23) Mizrahi, V.; Henrie, R. N.; Marlier, J. F.; Johnson, K. A.; Benkovic, S. J.
Biochemistry 1985, 24, 4010-4018.
(24) Showalter, A. K.; Tsai, M.-D. Biochemistry 2002, 41, 10571-10576.
9
6138 J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003

Stability and Replication of Unnatural Base Pairs
A R T I C L E S
natural bases at the primer terminus. Failure of the unnatural
DNA base pair to engage the protein suitably via any of these
interactions may result in a misaligned 3′OH primer terminus
that is unable to act as an efficient nucleophile during the
extension step. Although the overall structure and accommoda-
tion in B-form duplex DNA is similar for each self-pair, the
BTz self-pair differs in at least two ways. BTz presents a
natural-like endocyclic imine functionality in the developing
minor grove at the primer terminus, as opposed to an ether
(BFr), amine (IN), or thioether (BTp), to the developing minor
groove at the primer terminus. The imine moiety may form a
more favorable H-bond with Arg573 in KF and, therefore, orient
the primer terminus for continued extension.25 The overall dipole
and polarizability are also significantly different for BTz, relative
to the other unnatural nucleobases. This difference may alter
packing with the adjacent primer base, or the incoming dNTP,
in a manner favorable for extension.10,21
structure/function studies. For example, we have initiated NMR
solution structure studies to address the orientation and packing
of the unnatural nucleobases in duplex DNA. Nonetheless, the
BTz self-pair shows extension rates that are reduced by a factor
of only ∼102, relative to natural DNA in the same sequence
context. Primer extension after synthesis of the unnatural base
pair is consistently the step that limits the synthesis of DNA
containing unnatural bases. The sensitivity of this step to
derivatization of the unnatural base suggests that modifications
may be found that result in unnatural base pairs that can be
both efficiently synthesized and extended, resulting in an
expanded genetic alphabet.
Acknowledgment. Funding was provided by the National
Institutes of Health (No. GM 60005 to F.E.R.), the Skaggs
Institute for Chemical Biology (F.E.R. and P.G.S.), and the JSPS
Research Fellowships for Young Scientists (S.M.).
Supporting Information Available: Details of synthetic
procedures and characterizations (PDF). This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
Further clarification of the physical origins of the increased
rates for continued primer extension will require additional
(25) (a) Monge, M. A.; Puebla, G.; Gallego, M. M. R.; Pardo, C.; Elguero, J.
J. Heterocycl. Chem. 1992, 29, 499-502. (b) Buchet, R.; Dion, A. J.
Biomol. Struct. Dyn. 1986, 4, 231-241.
JA034099W
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6139