An efficient unnatural base pair for PCR amplification

Article abstract of DOI:10.1021/ja073830m

Expansion of the genetic alphabet by an unnatural base pair system provides a powerful tool for modern biotechnology. As an alternative to previous unnatural base pairs, we have developed a new pair between 7-(2-thienyl) imidazo[4,5-b]pyridine (Ds) and 2-

Full text of DOI:10.1021/ja073830m

Published on Web 11/21/2007
An Efficient Unnatural Base Pair for PCR Amplification
Ichiro Hirao,*,† Tsuneo Mitsui,† Michiko Kimoto,† and Shigeyuki Yokoyama*,†,‡,§
Contribution from the Protein Research Group, RIKEN Genomic Sciences Center, 1-7-22
Suehiro-cho, Tsurumi-ku, Yokohama, Kanagawa 230-0045, Japan, Department of Biophysics and
Biochemistry, Graduate School of Science, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-0033, Japan, and RIKEN Harima Institute at SPring-8, 1-1-1 Kouto, Mikazuki-cho,
Sayo, Hyogo 679-5148, Japan
Received May 28, 2007; E-mail: ihirao@riken.jp; yokoyama@biochem.s.u-tokyo.ac.jp
Abstract: Expansion of the genetic alphabet by an unnatural base pair system provides a powerful tool for
modern biotechnology. As an alternative to previous unnatural base pairs, we have developed a new pair
between 7-(2-thienyl)imidazo[4,5-b]pyridine (Ds) and 2-nitropyrrole (Pn), which functions in DNA amplifica-
tion. Pn more selectively pairs with Ds in replication than another previously reported pairing partner, pyrrole-
2-carbaldehyde (Pa). The nitro group of Pn efficiently prevented the mispairing with A. High efficiency and
selectivity of the Ds-Pn pair in PCR amplification were achieved by using a substrate mixture of the
γ-amidotriphosphate of Ds and the usual triphosphates of Pn and the natural bases, with Vent DNA
polymerase as a 3′ to 5′ exonuclease-proficient polymerase. After 20 cycles of PCR, the total mutation
rate of the Ds-Pn site in an amplified DNA fragment was ∼1%. PCR amplification of DNA fragments
containing the unnatural Ds-Pn pair would be useful for expanded genetic systems in DNA-based
biotechnology.
Introduction
Ds-Pa pairing in replication; dATPN and dDsTPN reduce the
mispairings of A-Pa and Ds-Ds, respectively.
Expansion of the genetic alphabet by an unnatural base pair
system has great potential in providing a variety of novel nucleic
acids and proteins with functional components of interest.1 This
can be achieved by creating an unnatural base pair that is
compatible with the natural A-T and G-C base pairs in
replication, transcription, and translation. Many unnatural base
pairs have been synthesized and tested in replication.2 Among
them, a hydrophobic, unnatural base pair between 7-(2-thienyl)-
imidazo[4,5-b]pyridine (Ds) and pyrrole-2-carbaldehyde (Pa)
with specific shape complementation exhibits high selectivity
in PCR amplification, with the combination of the usual
triphosphate substrates for G, C, T, and Pa, and modified
γ-amidotriphosphates for A (dATPN) and Ds (dDsTPN) (Figure
1).3 Both γ-amidotriphosphates increase the selectivity of the
Although the Ds-Pa pair can be used practically for in Vitro
applications, the use of the γ-amidotriphosphates, especially
dATPN, decreases the efficiency of PCR amplification and
restricts in ViVo applications. Since the number of A bases in
DNA fragments is typically much larger than that of Ds bases,
the use of dATPN significantly reduces the amplification
efficiency. Thus, further development of another unnatural base,
instead of Pa, which more selectively pairs with Ds than A,
would bypass the need for dATPN.
Here, we describe a novel unnatural base, 2-nitropyrrole (Pn),
as an efficient pairing partner of the hydrophobic Ds base
(Figure 1A). To prevent mispairing with A, we replaced the
aldehyde group of Pa with the nitro group of Pn. The nitro
group was expected to electrostatically repel the 1-nitrogen of
A and, thus, increase the selectivity of the Ds-Pn pairing in
replication. In addition, nitro groups have been used as
components of universal bases, such as 3-nitropyrrole and
5-nitroindole, since they enhance the stacking interactions with
neighboring bases, but are not strong hydrogen-bond acceptors
with pairing bases.4 Thus, the nitro group might be a useful
component for the development of unnatural, non-hydrogen-
bonded base pairs. We chemically synthesized the Pn triphos-
† RIKEN Genomic Sciences Center.
‡ The University of Tokyo.
§ RIKEN Harima Institute at SPring-8.
(1) (a) Bain, J. D.; Switzer, C.; Chamberlin, A. R.; Benner, S. A. Nature 1992,
356, 537-539. (b) Benner, S. A.; Burgstaller, P.; Battersby, T. R.; Jurczyk,
S. In The RNA World, 2nd ed.; Gesteland, R. F., Cech, T. R., Atkins, J. F.,
Eds.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1999;
p 163. (c) Hirao, I.; Ohtsuki, T.; Fujiwara, T.; Mitsui, T.; Yokogawa, T.;
Okuni, T.; Nakayama, H.; Takio, K.; Yabuki, T.; Kigawa, T.; Kodama,
K.; Yokogawa, T.; Nishikawa, K.; Yokoyama, S. Nat. Biotechnol. 2002,
20, 177-182. (d) Benner, S. A.; Sismour, A. M. Nat. ReV. Genet. 2005, 6,
533-543. (e) Hunziker, J.; Mathis, G. Chimia 2005, 59, 780-784. (f)
Kawai, R.; Kimoto, M.; Ikeda, S.; Mitsui, T.; Endo, M.; Yokoyama, S.;
Hirao, I. J. Am. Chem. Soc. 2005, 127, 17286-17295. (g) Prudent, J. R.
Expert ReV. Mol. Diagn. 2006, 6, 245-252. (h) Hirao, I. Biotechniques
2006, 40, 711-717.
(4) (a) Nichols, R.; Andrews, P. C.; Zhang, P.; Bergstrom, D. E. Nature 1994,
369, 492-493. (b) Loakes, D.; Brown, D. M. Nucleic Acids Res. 1994,
22, 4039-4043. (c) Aerschot, A. V.; Rozenski, J.; Loakes, D.; Pillet, N.;
Schepers, G.; Herdewijn, P. Nucleic Acids Res. 1995, 23, 4363-4370. (d)
Loakes, D.; Brown, D. M.; Linde, S.; Hill, F. Nucleic Acids Res. 1995, 23,
2361-2366. (e) Bergstrom, D. E.; Zhang, P.; Toma, P. H.; Andrews, P.
C.; Nichols, R. J. Am. Chem. Soc. 1995, 117, 1201-1209. (f) Amosova,
O.; George, J.; Fresco, J. R. Nucleic Acids Res. 1997, 25, 1930-1934.
(2) (a) Kool, E. T. Curr. Opin. Chem. Biol. 2000, 4, 602-608. (b) Henry, A.
A.; Romesberg, F. E. Curr. Opin. Chem. Biol. 2003, 7, 727-733. (c) Hirao,
I. Curr. Opin. Chem. Biol. 2006, 10, 622-627.
(3) Hirao, I.; Kimoto, M.; Mitsui. T.; Fujiwara, T.; Kawai, R.; Sato, A.; Harada,
Y.; Yokoyama, S. Nat. Methods 2006, 3, 729-735.
9
10.1021/ja073830m CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 15549-15555
15549

A R T I C L E S
Hirao et al.
Scheme 1. Synthesis of the Nucleoside Derivatives of
2-Nitropyrrolea
a Conditions: (a) NaH, 1-chloro-2-deoxy-3,5-di-O-toluoyl-R-D-erythro-
pentofuranose, CH3CN; (b) NH3, methanol; (c) 4,4′-dimethoxytrityl chloride,
pyridine; (d) 2-cyanoethyl-N,N-diisopropylamino chloro phosphoramidite,
diisopropylethylamine, THF; (e) acetic anhydride, pyridine, then dichloro-
acetic acid, dichloromethane; (f) 2-chloro-4H-1,3,2-benzodioxaphosphorin-
4-one, dioxane, pyridine, tri-n-butylamine, bis(tributylammonium)pyro-
phosphate, DMF, then I2/pyridine, water, NH4OH. Tol: toluoyl. DMTr:
4,4′-dimethoxytrityl. Ac: acetyl.
Figure 2. Stability of the 2′-deoxyribonucleoside of Pn (3) under basic
conditions. Compound 3 was treated with concentrated ammonia at 55 °C
(2) or room temperature (b) for 1, 3, 6, and 10 h. The amounts (%) of 3
remaining were determined by HPLC using an internal standard, thymidine.
Figure 1. (A) Structures of the unnatural Ds-Pn and Ds-Pa pairs and
A-Pn and A-Pa mispairs. Space-filling models of bases alone (with methyl
in place of deoxyribose) are shown, with electrostatic potentials mapped
on van der Waals surfaces (PM3 calculations, Spartan ’06, Wavefunction
Inc.). The arrow in the A-Pn pair indicates the electrostatic repulsion
between the 1-nitrogen of A and the nitro group of Pn. (B) The structures
of usual (R ) OH) and modified γ-amidotriphosphate (R ) NH2) substrates.
anomeric configuration of 3 was identified as the â-form by
1
the H/1H NOESY spectrum. The NOESY spectrum showed
cross-peaks between H1′ and H4′ and between H5 and H3′/
H2′′.
The nucleoside was converted to the amidite (5) for DNA
chemical synthesis by conventional methods and to the triph-
osphate (dPnTP, 7) for enzymatic incorporation. The UV and
visible spectra of dPnTP in 10 mM phosphate buffer (pH 7.0)
showed absorption maxima (λmax) of 232 nm (ꢀ 3100) and 352
nm (ꢀ 12 600), which were similar to those of 2-nitropyrrole in
methanol [λmax ) 231 nm (ꢀ 4074) and 335 nm (ꢀ 17 000)],5
rather than those of 3-nitropyrrole [λmax)268 nm (ꢀ 7244) and
315 nm (ꢀ 5370)]5 and 1-(2-deoxy-â-D-ribofuranosyl)-3-nitro-
pyrrole [λmax ) 284 nm (ꢀ 4850)].4e
phate and DNA fragments containing Pn and examined the
ability of the Ds-Pn pairing in replication, in comparison to
that of the Ds-Pa pairing. We found that PCR amplification
using the Ds-Pn pair did not require dATPN for higher
efficiency and fidelity, as compared to the amplification using
the Ds-Pa pair.
Results and Discussion
We found that the nucleoside derivatives of Pn were
considerably decomposed by treatments with concentrated
ammonia at high temperature. The decomposition of 3 was
analyzed and quantified by HPLC (Figure 2). After the treatment
of 3 with concentrated ammonia at 55 °C for 6 h, 65% of 3
was decomposed. Under these conditions, the removal of
2-nitropyrrole from a short DNA fragment containing Pn,
d(TTPnT), was observed by mass spectrometry (data not
shown). Thus, for DNA synthesis involving Pn, a base-labile
protection method, using phenoxyacetyl for A, p-isopropylphe-
Chemical Synthesis. The 2′-deoxyribonucleoside of Pn (3)
was synthesized by the coupling of a fully protected ribofura-
nosyl chloride with the sodium salt of 2-nitropyrrole (1),5
followed by deprotection (Scheme 1). The structure of 3 was
confirmed by NMR and high-resolution mass spectroscopy (see
Supporting Information 1 and 2). The HMQC and HMBC
spectra revealed the N-glycoside bond between the sugar C1′
and the N-1 position of the 2-nitropyrrole base moiety. The
(5) Morgan, K. J.; Morrey, D. P. Tetrahedron 1966, 22, 57-62.
9
15550 J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007

An Unnatural Base Pair for PCR
A R T I C L E S
Table 1. Steady-State Kinetic Parameters for Insertion of Single Nucleotides into a Template-Primer Duplexa
primer
temp35N-1
5′ACTCACTATAGGGAGGAAGA
-TATTATGCTGAGTGATATCCCTCCTTCTNTCTCGA
3′
template
(N)
nucleoside
KM
M)
Vmax
(% min-
efficiency
1
c
entry
triphosphate
(
µ
)
(Vmax/KM)
1
2
3
Ds
A
G
C
T
Ds
A
G
C
T
Ds
Ds
Ds
Ds
Ds
Ds
A
A
G
G
dPnTP
dPnTP
dPnTP
dPnTP
dPnTP
dPaTP
dPaTP
dPaTP
dPaTP
dPaTP
dDsTP
dATP
dGTP
dCTP
dTTP
dDsTPN
dTTP
91 (6)b
130 (60)
80 (56)
140 (130)
140 (40)
340 (150)
330 (160)
140 (20)
n.de
170 (60)
8.0 (3.9)
150 (40)
n.d.
410 (190)
220 (20)
79 (13)
34 (5)
2.7 (1.2)
1.1 (0.4)
0.10 (0.05)
2.9 (0.6)
21 (3)
17 (7)
0.061 (0.006)
n.d
0.053 (0.016)
1.6 (0.1)
0.36 (0.09)
n.d.
0.34 (0.05)
0.41 (0.17)
0.78 (0.12)
2.8 (1.5)
2.2 (0.9)
5.5 (1.7)
0.29 (0.12)
3.7 × 105
2.1 × 104
1.4 × 104
7.1 × 102
2.1 × 104
6.2 × 104
5.2 × 104
4.4 × 102
-
4
5
6d
7d
8d
9d
10d
11d
12d
13d
14d
15d
16d
17d
18d
19d
20d
3.1 × 102
2.0 × 105
2.4 × 103
-
8.3 × 102
1.9 × 103
9.9 × 103
4.0 × 106
1.8 × 103
2.3 × 107
2.1 × 103
0.70 (0.40)
1200 (600)
0.24 (0.18)
140 (70)
dCTP
dCTP
dTTP
a Assays were carried out at 37 °C for 1.4-20 min, using 5 µM template-primer duplex, 5-50 nM enzyme, and 6-1500 µM nucleoside triphosphate
in a solution (10 µL) containing 50 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 1 mM DTT, and 0.05 mg/mL bovine serum albumin. Each parameter was
c
averaged from three to four data sets. bStandard deviations are given in parentheses. The units of this term are % min-1 M-1
.
dThese parameters were
referred from Nat. Methods 2006, 3, 729-735. eNot determined. Minimal insert products (<2%) were detected after an incubation for 20 min with 1500 µM
nucleoside triphosphate and 50 nM enzyme.
noxyacetyl for G, and acetyl groups for C,6 was used, and
deprotection was carried out at room temperature for 2-3 h
with concentrated ammonia. Under these conditions, the de-
composition of DNA fragments containing Pn was reduced to
less than 10%. No decomposition of either the Pn nucleoside
in DNA fragments or dPnTP was observed under the conditions
corresponding to 40 cycles of PCR amplification in a pH 7.5
buffer (data not shown).
were analyzed with an automated ABI 377 DNA sequencer with
the GeneScan software.9
The incorporation efficiency of dPnTP into DNA opposite
Ds in a template (Vmax/KM ) 3.7 × 105 %‚min-1‚M-1) was
6-fold higher than that of dPaTP opposite Ds (Vmax/KM ) 6.2
× 104 %‚min-1‚M-1). In contrast, the misincorporation of
dPnTP opposite A (Vmax/KM ) 2.1 × 104 %‚min-1‚M-1
)
decreased by 2.5-fold, as compared to that of dPaTP opposite
A (Vmax/KM ) 5.2 × 104 %‚min-1‚M-1). Similarly, the
incorporation efficiency of dATP opposite Pn (Vmax/KM ) 5.5
× 103 %‚min-1‚M-1) was also 7.8-fold lower than that of dATP
opposite Pa (Vmax/KM ) 4.3 × 104 %‚min-1‚M-1). As a whole,
the incorporation efficiency of dPnTP opposite Ds was more
than 154 times higher than those of the natural base substrates
opposite Ds (Table 1, entries 12-15). In addition, the incor-
poration efficiency of dDsTP opposite Pn was more than 39
times higher than those of the natural base substrates opposite
Pn (Table 2, entries 3 and 5-7).
Replication Studies. We first compared the efficiency and
selectivity of the Ds-Pn pairing in replication with those of
the Ds-Pa pairing, by analyzing the steady-state kinetics of
single-nucleotide insertion experiments7 using the exonuclease-
deficient Klenow fragment (KF exo-). The insertion experiments
using KF exo- are reliable for assessing the replication ability
of newly developed, unnatural base pairs.8 In addition, we
previously reported that the kinetics data using KF exo- are
informative for the application of unnatural base pairs to PCR
amplification.3 The experiments were carried out using a
template strand (35-mer, Tables 1 and 2) containing Pn, Pa, or
Ds, a primer labeled with 6-carboxyfluorescein (20-mer), and
each triphosphate, such as dPnTP, dPaTP, dDsTP, dDsTPN,
dATPN, and the natural triphosphates. The insertion products
The combination of the Ds γ-amidotriphosphate (dDsTPN)
and Pn also increased the selectivity of the Ds-Pn pairing in
replication, relative to that of the Ds-Pa pairing (Tables 1 and
2). Although the incorporation efficiency of dDsTP opposite
Pn (Vmax/KM ) 8.5 × 105 %‚min-1‚M-1) was slightly lower
than that of dDsTP opposite Pa (Vmax/KM ) 1.1 × 106
%‚min-1‚M-1), the efficiency of dDsTPN incorporation opposite
Pn (Vmax/KM ) 8.6 × 104 %‚min-1‚M-1) was slightly higher
than that opposite Pa (Vmax/KM ) 6.7 × 104 %‚min-1‚M-1).
Since the Pn template efficiently excluded the misincorporation
of dATP, the selectivity of Pn between dDsTPN and dATP
increased by 16-fold, as compared to that of Pa between dDsTPN
and dATP (1.6-fold). This improvement was mainly caused by
the decrease in the Vmax value of the dATP incorporation
(6) (a) Schulhof, J. C.; Molko, D.; Teoule, R. Nucleic Acids Res. 1987, 15,
397-416. (b) Chaix, C.; Duplaa, A. M.; Molko, D.; Teoule, R. Nucleic
Acids Res. 1989, 18, 7381-7393.
(7) (a) Petruska, J.; Goodman, M. F.; Boosalis, M. S.; Sowers, L. C.; Cheong,
C.; Tinoco, I. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 6252-6256. (b)
Goodman, M. F.; Creighton, S.; Bloom, L. B.; Petruska, J. Crit. ReV.
Biochem. Mol. Biol. 1993, 28, 83-126.
(8) (a) Morales, J. C.; Kool, E. T. Nat. Struct. Biol. 1998, 5, 950-954. (b)
Matray, T. J.; Kool, E. T. Nature 1999, 399, 704-708. (c) McMinn, D.
L.; Ogawa, A. K.; Wu, Y.; Liu, J.; Schultz, P. G.; Romesberg, F. E. J. Am.
Chem. Soc. 1999, 121, 11585-11586. (d) Wu, Y.; Ogawa, A. K.; Berger,
M.; McMinn, D. L.; Schultz, P. G.; Romesberg, F. E. J. Am. Chem. Soc.
2000, 122, 7621-7632. (e) Mitsui, T.; Kimoto, M.; Harada, Y.; Yokoyama,
S.; Hirao, I. J. Am. Chem. Soc. 2005, 127, 8652-8658. (f) Matsuda, S.;
Henry, A. A.; Romesberg, F. E. J. Am. Chem. Soc. 2006, 128, 6369-
6375.
(9) Kimoto, M.; Yokoyama, S.; Hirao, I. Biotechnol. Lett. 2004, 26, 999-
1005.
9
J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007 15551

A R T I C L E S
Hirao et al.
Table 2. Steady-State Kinetic Parameters for Insertion of Single Nucleotides into a Template-Primer Duplexa
Primer
Temp35N-2
5′-ACTCACTATAGGGAGCTTCT
-TATTATGCTGAGTGATATCCCTCGAAGANAGAGCT
3′
template
(N)
nucleoside
KM
M)
Vmax
(% min-
efficiency
1
c
entry
triphosphate
(
µ
)
(Vmax/KM)
1
2
3
4
5
Pn
Pn
Pn
Pn
Pn
Pn
Pn
Pn
Pa
Pa
Pa
Pa
Pa
Pa
Pa
T
dDsTP
dDsTPN
dATP
dATPN
dGTP
dCTP
20 (7)b
110 (20)
510 (210)
760 (330)
220 (30)
n.d.e
630 (80)
110 (60)
26 (12)
180 (20)
490 (260)
1200 (400)
480 (140)
n.d.
17 (7)
9.5 (2.1)
2.8 (1.6)
0.39 (0.14)
4.8 (1.2)
n.d.
1.3 (0.4)
0.74 (0.32)
28 (5)
8.5 × 105
8.6 × 104
5.5 × 103
5.1 × 102
2.2 × 104
-
6
7
dTTP
2.1 × 103
6.7 × 103
1.1 × 106
6.7 × 104
4.3 × 104
1.8 × 103
8.8 × 102
-
8
dPnTP
dDsTP
dDsTPN
dATP
dATPN
dGTP
dCTP
9d
10d
11d
12d
13d
14d
15d
16d
17d
18d
19d
12 (1)
21 (6)
2.2 (1.3)
0.42 (0.09)
n.d.
0.097 (0.025)
3.3 (1.8)
2.3 (0.8)
16 (4)
dTTP
880 (530)
1.1 × 102
4.1 × 106
4.6 × 103
7.0 × 106
2.9 × 103
dATP
dATP
dGTP
dGTP
0.81 (0.44)
500 (90)
2.3 (0.1)
420 (20)
C
C
T
1.2 (0.1)
a Assays were carried at 37 °C for 1.2-21.8 min, using 5 µM template-primer duplex, 10-50 nM enzyme, and 15-1500 µM nucleoside triphosphate in
a solution (10 µL) containing 50 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 1 mM DTT, and 0.05 mg/mL bovine serum albumin. Each parameter was averaged
from three to four data sets. bStandard deviations are given in parentheses. cThe units of this term are % min-1 M-1. dThese parameters were referred from
e
Nat. Methods 2006, 3, 729-735. Not determined. Minimal insert products (<2%) were detected after an incubation for 20 min with 1500 µM nucleoside
triphosphate and 50 nM enzyme.
opposite Pn. In addition, the misincorporation of dDsTPN
opposite Ds (Vmax/KM ) 9.9 × 103 %‚min-1‚M-1) was 37-fold
lower than that of dPnTP opposite Ds (Vmax/KM ) 3.7 × 105
%‚min-1‚M-1).
3) was higher than that of the DNA fragment containing a single
Ds-Pa pair using dATPN instead of dATP (Figure 4A, lane
6). The DNA fragment containing a single Ds-Pn pair was
approximately 500-fold amplified after 20 cycles of PCR.
Although the amplification efficiency of the DNA fragment
containing the Ds-Pa using dATP was relatively high (Figure
4A, lanes 8 and 9), the mutation rate of the Ds-Pa site also
increased, as mentioned below (Figure 4F and 4G).
Primer extension involving the Ds-Pn pair by the exonu-
clease-proficient Klenow fragment (KF exo+) also exhibited
higher selectivity, relative to that involving the Ds-Pa pair
(Figure 3). The elongation of the Ds-Pn pairing occurred
efficiently, and elongation products were obtained (33-mer in
Figure 3A, lane 2 and 34-35-mer in Figure 3B, lane 2). The
A-Pn mispairing (Figures 3A, lane 4 and 3B, lane 5) was
effectively prevented, relative to the A-Pa mispairing (Figures
3A, lane 10 and 3B, lane 6). In addition, the efficiency of the
primer extension with the combination of dDsTPN and the Pn
template (Figure 3A, lane 3) was as high as that of dDsTP and
the Pn template (Figure 3A, lane 2). Although a small amount
of Pn self-pairing was observed (Figure 3A, lane 6), on the
whole, the selectivity of the Ds-Pn pair in replication was
improved.
The selectivity of the Ds-Pn pair in PCR was assessed by
sequencing the amplified DNA fragments. As we previously
reported, DNA fragments containing Ds can be sequenced by
a conventional dideoxy-dye-terminator method with or without
a modified Pa substrate, 4-propynylpyrrole-2-carbaldehyde
(dPa′TP).3 In the sequencing with dPa′TP, only the peak
corresponding to the base opposite Ds in the templates disap-
pears. In addition, in the sequencing without dPa′TP, the
sequencing terminated at the position opposite Ds and the read-
through peaks were significantly reduced, if the DNA fragment
was faithfully amplified at the Ds-Pa position.3
PCR Amplification. We next performed PCR amplification
of a DNA fragment (174-mer) containing one Ds-Pn pair (see
the Supporting Information 3), using Vent DNA polymerase
and a substrate mixture of dDsTPN, dPnTP, dATP, dGTP,
dCTP, and dTTP. The single-nucleotide insertion experiments
using KF exo- showed that the efficiency and selectivity of
the Ds-Pn pairing were still lower than those of the natural
A-T and G-C pairings. In our experiments, the selectivity of
the cognate A-T and G-C pairings were 890-11 000-fold
higher than that of the noncognate A-C and G-T pairings
(Table 1, entries 17-20 and Table 2, entries 16-19). However,
we previously demonstrated that the 3′ to 5′ exonuclease activity
of Vent DNA polymerase increased the selectivity of the Ds-
Pa pair, for the use of the unnatural base pair in PCR
amplification.3
The high selectivity of the Ds-Pn pair was proven by a
comparison of the sequencing patterns of the 20-cycle PCR
products (Figure 4D and 4E) with those of the original DNA
fragment containing the Ds-Pn pair (Figure 4B and 4C). By
analyzing the read-through peak heights of the sequencing
without dPa′TP (Figure 4E), the mutation rate of the Ds-Pn
pair was assessed by comparison to the sequencing patterns of
the control DNA fragments (see the Supporting Information 4).3
Since the control DNA fragments contain 1-10% of the A-T
pair in place of the unnatural base pair, in the sequencing without
dPa′TP, the read-through peak heights depend on the replace-
ment rate (1-10%) of the A-T pair in the control DNA
fragments. The reliability of this method has been confirmed
by comparison to the mutation rate determined by transcription
using PCR-amplified DNA templates containing the Ds-Pa pair
with the ribonucleoside triphosphate of biotin-linked Pa.3
The amplification efficiency after 20 cycles of PCR with the
DNA fragment containing a single Ds-Pn pair (Figure 4A, lane
9
15552 J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007

An Unnatural Base Pair for PCR
A R T I C L E S
working to improve the stability of Pn. As compared with the
Ds-Pa pair, the PCR amplification involving the Ds-Pn pair
requires no dATPN and, thus, increases the amplification
efficiency. The Ds-Pn pair therefore promises to be useful in
a variety of technologies employing genetic expansion systems.
Materials and Methods
General. Reagents and solvents were purchased from standard
suppliers and used without further purification. Reactions were
monitored by thin-layer chromatography (TLC) using 0.25 mm silica
gel 60 plates impregnated with 254 nm fluorescent indicator (Merck).
1H, 13C, and 31P NMR spectra were recorded on JEOL EX270 and
BRUKER (300-AVM) magnetic resonance spectrometers. Nucleoside
purification was performed on a Gilson HPLC system with a preparative
C18 column (Waters Microbond Sphere, 150 mm × 19 mm).
Triphosphate derivatives were purified with a DEAE-Sephadex A-25
column (300 mm × 15 mm) and a C18 column (Synchropak RPP,
250 mm × 4.6 mm, Eichrom Technologies). High-resolution mass
spectra (HRMS) and electrospray ionization mass spectra (ESI-MS)
were recorded on a JEOL JM 700 mass spectrometer and a Waters
micromass ZMD 4000 equipped with a Waters 2690 LC system,
respectively. The 2-nitropyrrole (1) was synthesized according to the
previously reported method.5
1-(2-Deoxy-â-D-ribofuranosyl)-2-nitropyrrole (3). To a mixture of
15 (224 mg, 2.0 mmol) and CH3CN (20 mL) was added NaH (80 mg,
2.0 mmol, 60% dispersion in mineral oil). The resulting mixture was
stirred for 30 min at room temperature. To the reaction mixture was
added 1-chloro-2-deoxy-3,5-di-O-toluoyl-R-D-erythro-pentofuranose10
(855 mg, 2.2 mmol). After stirring for 2 h at room temperature, the
reaction mixture was separated by ethyl acetate and water. The organic
phase was washed three times with saturated NaCl, dried with Na2-
SO4, and evaporated in Vacuo. The product was purified by silica gel
column chromatography (10% hexane in CH2Cl2) to give 2 (722 mg,
78%) as a white foam. To 722 mg of 2 was added methanolic ammonia
(50 mL) that was saturated at 0 °C. The solution was stirred for 12 h
at room temperature. The solvent was evaporated in Vacuo. The residue
was purified by silica gel column chromatography (2% methanol in
CH2Cl2) and RP-HPLC to give 291 mg (82%) of compound 3.
1-[2-Deoxy-5-O-(4,4′-dimethoxytrityl)-â-D-ribofuranosyl]-2-nitro-
pyrrole 2-Cyanoethyl-N,N-diisopropylphosphoramidite (5). A por-
tion (228 mg, 1.0 mmol) of 3 was coevaporated with dry pyridine three
times. To the residue in dry pyridine (10 mL) was added 4,4′-
dimethoxytrityl chloride (373 mg, 1.1 mmol). The resulting mixture
was stirred for 1 h at room temperature. The mixture was separated by
ethyl acetate and water. The organic phase was washed three times
with 5% NaHCO3 and saturated NaCl, dried with Na2SO4, and
evaporated in Vacuo. The product was purified by silica gel column
chromatography (10% ethyl acetate in CH2Cl2) to give 4 (493 mg, 93%).
After evaporation of compound 4 (265 mg, 0.5 mmol) with pyridine
three times, the residue was dissolved in THF (2.5 mL). To the mixture
was added diisopropylethylamine (131 µL, 0.75 mmol) and 2-cyano-
ethyl-N,N-diisopropylamino chlorophosphoramidite (123 µL, 0.55
mmol). The reaction mixture was stirred at room temperature for 1 h.
After addition of methanol (50 µL), the mixture was diluted with ethyl
acetate/triethylamine (20:1, v/v) and then washed with 5% NaHCO3
and saturated NaCl. The organic phase was dried with Na2SO4 and
evaporated in Vacuo. The product was purified by silica gel column
chromatography (20% CH2Cl2 in hexane containing 2% triethylamine)
to give compound 5 (315 mg, 86%).
Figure 3. Primer extensions using natural and unnatural nucleotides in
the templates containing Pn or Pa (A) or Ds, A, or G (B). The reactions
were performed by the 3′ f 5′ exonuclease-proficient Klenow fragment
(0.1 unit/µL) with 10 µM substrates and 200 nM templates at 37 °C for 5
min. DsN means dDsTPN.
The total mutation rate of the Ds-Pn site in the amplified
DNA fragment after 20 cycles was ∼1%, when the PCR was
performed with the usual triphosphates of the natural and Pn
bases and the γ-amidotriphosphate of Ds. In the PCR amplifica-
tion of the Ds-Pa pairing, the total mutation rate of the Ds-
Pa site was 3-4%, even when the PCR was carried out with
both the γ-amidotriphosphates of Ds and A.3 When dATP was
used instead of dATPN for PCR amplification, the mutation rate
of the Ds-Pa site further increased by ∼10% (Figure 4F and
4G). Thus, the Ds-Pn pair was significantly improved for PCR
amplification, as compared with the Ds-Pa pair.
In summary, we have developed the unnatural Ds-Pn base
pair, which exhibits high efficiency and selectivity in in Vitro
replication. Even though the nitro group of Pn prevents the
mispairing with A, Pn efficiently pairs with Ds, indicating that
the nitro group of Pn also functions as a hydrogen acceptor
residue for recognition by DNA polymerases. Although the Pn
nucleoside is unstable under harshly basic conditions, it can
withstand the conditions in at least 40 cycles of PCR amplifica-
tion. We are studying the decomposition mechanism and
1-(2-Deoxy-â-D-ribofuranosyl)-2-nitropyrrole 5′-Triphosphate (7,
dPnTP). A portion (159 mg, 0.30 mmol) of 4 was coevaporated with
dry pyridine three times. The residue was added into dry pyridine (3.0
mL) and was combined with acetic anhydride (57 µL, 0.6 mmol). The
(10) Rolland, V.; Kotera, M.; Lhomme, J. Synth. Commun. 1997, 27, 3505-
3511.
9
J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007 15553

A R T I C L E S
Hirao et al.
Figure 4. PCR amplification of a DNA fragment (174-mer) containing the Ds-Pn or Ds-Pa base pair. (A) Agarose-gel analysis of original fragments and
PCR products containing one Ds-Pn pair with ATP (lanes 1-3), one Ds-Pa pair with ATPN (lanes 4-6), or one Ds-Pa pair with ATP (lanes 7-9).
(B-G) Sequencing of the original DNA (B, C), 20-cycle amplified DNA fragments containing the Ds-Pn pair using the PCR conditions in Figure 4A lane
3 (D, E), and 20-cycle amplified DNA fragments containing the Ds-Pa pair using the PCR conditions in Figure 4A lane 9 (F, G). The sequencing was
performed in the presence (B, D, and F) or absence (C, E, and G) of dPa′TP, using a BigDye Terminator v1.1 Cycle Sequencing kit. The arrow indicates
the unnatural base position.
resulting mixture was stirred for 12 h at room temperature. The mixture
was poured into 5% NaHCO3 and extracted with ethyl acetate. After
drying over Na2SO4, the solvent was evaporated in Vacuo. The residue
in dry methylene chloride (30 mL) was mixed with dichloroacetic acid
(300 µL) at 0 °C. The resulting mixture was stirred for 15 min at 0 °C.
The mixture was poured into 5% NaHCO3 and extracted with methylene
chloride. After drying over Na2SO4, the solvent was evaporated in
Vacuo. The residue was subjected to silica gel chromatography, using
methylene chloride/ethyl acetate (9:1, v/v) as an eluent, to afford 74
mg of 6 in a 91% yield. Compound 6 (41 mg, 0.15 mmol) was dissolved
9
15554 J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007

An Unnatural Base Pair for PCR
A R T I C L E S
in pyridine and evaporated to dryness in Vacuo. The residue was
dissolved in pyridine (150 µL) and dioxane (450 µL). A 1 M solution
of 2-chloro-4H-1,3,2-benzodioxaphosphorin-4-one in dioxane (180 µL,
0.18 mmol) was added.11 After 10 min, tri-n-butylamine (150 µL) and
0.5 M bis(tributylammonium)pyrophosphate in DMF (450 µL) were
added to the reaction mixture. The mixture was stirred at room
temperature for 10 min. A solution of 1% iodine in pyridine/water (98:
2, v/v) (3.0 mL) was then added. After 15 min, a 5% aqueous solution
(230 µL) of NaHSO3, followed by water (7.5 mL), was added to the
reaction mixture. The solution was stirred at room temperature for 30
min, and then concentrated ammonia (30 mL) was added. Hydrolysis
was carried out at room temperature for 2 h. After the reaction mixture
was concentrated in Vacuo, the product was purified by DEAE Sephadex
(A-25) column chromatography (eluted by a linear gradient of 50 mM
to 1 M TEAB) and then by C18-HPLC (eluted by a gradient of 0% to
30% CH3CN in 100 mM triethylammonium acetate) to give the
nucleoside 5′-triphosphate (7, dPnTP) in a 49% yield.
Stability of the Pn Nucleoside. A mixture of the Pn nucleoside
(5.7 mM) and thymidine (2.0 mM), which was added as an internal
standard, in H2O (100 µL) was treated with 28% NH4OH (2.0 mL) at
room temperature or heated at 55 °C for 1-10 h in a sample vial
(Wheaton sample vial, 4 mL, with a white-rubber lined cap). The treated
solutions were evaporated in Vacuo, and the residues were dissolved
in H2O. The solutions were analyzed by RP-HPLC (0-30% CH3CN
in 100 mM TEAA, 10 min at a flow rate of 1 mL/min, detected at 280
nm, Gilson HPLC system, Synchropak RPP, 250 mm × 4.6 mm,
Eichrom Technologies). The amounts (%) of the Pn nucleoside
remaining intact were calculated by normalization to the peak area of
the thymidine.
Steady-State Kinetics for the Single-Nucleotide Insertion Experi-
ments with KF exo-. A primer (20-mer) labeled with 6-carboxyfluo-
rescein at the 5′-end was annealed with a template (35-mer), in 100
mM Tris-HCl (pH 7.5) buffer containing 20 mM MgCl2, 2 mM DTT,
and 0.1 mg/mL bovine serum albumin. The primer-template duplex
solution (10 µM, 5 µL) was mixed with 2 µL of an enzyme solution
containing the exonuclease-deficient Klenow fragment, KF exo-
(Amersham USB). The mixture was incubated for more than 2 min,
and then the reactions were initiated by adding each dNTP solution (3
µL) to the duplex-enzyme mixture at 37 °C. The amount of enzyme
used (5-50 nM), the reaction time (1.2-21.8 min), and the gradient
concentration of dNTP (6-1500 µM) were adjusted to give reaction
extents of 25% or less. The reactions were quenched with 10 µL of a
stop solution (95% formamide and 20 mM EDTA), and the mixtures
were immediately heated at 75 °C for 3 min. The diluted products were
analyzed on an automated ABI 377 DNA sequencer equipped with the
GeneScan software (version 3.0).9 Relative velocities (V0) were
calculated as the extents of the reaction divided by the reaction time
and were normalized to the enzyme concentration (20 nM) for the
various enzyme concentrations used. The kinetic parameters (KM and
solution (400 nM, 5 µL) was mixed with 2 µL of dNTP solution (50
µM each) on ice, and the reactions were started by adding 3 µL of a
solution containing the exonuclease-proficient Klenow fragment (1U,
TaKaRa, Tokyo) diluted in distilled water. The reaction mixture was
incubated at 37 °C for 5 min and was terminated by adding 10 µL of
a dye solution (10 M urea and 0.05% BPB in 1 × TBE) and heating
at 75 °C for 3 min. The products were analyzed on a 15% polyacry-
lamide gel containing 7 M urea. The reaction extents were measured
with a bio-imaging analyzer (Fuji model BAS2500).
PCR Amplification of the 174-mer Templates. PCR was performed
in 20 mM Tris-HCl buffer (pH 8.8), with 10 mM KCl, 10 mM (NH4)2-
SO4, 2 mM MgSO4, 0.1% Triton X-100, 0.3 mM dPnTP or dPaTP,
dATP or dATPN, dGTP, dCTP, dTTP, and dDsTPN, 1 µM each Primer
1 and Primer 2, approximately 0.6 nM double-stranded DNA templates
(N1-N2 ) Ds-Pn or Ds-Pa; see Supporting Information 3), and 0.04
unit/µL Vent DNA polymerase (NEB) on a PTC-100 Programmable
Thermal Controller (MJ Research). The PCR cycle was as follows:
94 °C, 0.5 min; 45 °C, 0.5 min; 65 °C, 4 min.3 The products were
analyzed on a 4% agarose gel stained with EtBr and were quantified
by using a Bio-Rad molecular imager FX. For sequencing analysis,
the PCR products were purified by gel electrophoresis (7% polyacry-
lamide-7 M urea gel) or filtration using Microcon YM-30 (Millipore)
and Micropure-EZ (Millipore) filters.
Dye Terminator Sequencing of the 174-mer DNA Fragments.
The cycle sequencing reactions (20 µL) were performed on the PTC-
100 controller (MJ Research) with the Cycle Sequencing Mix (8 µL)
from the BigDye Terminator v1.1 Cycle Sequencing Kit (Applied
Biosystems), containing approximately 0.3 pmol of template and 4 pmol
of Primer 2, in the presence or absence of 1 nmol of dPa′TP.3 After 20
cycles of PCR (96 °C, 10 s; 50 °C, 5 s; 60 °C, 4 min), the residual dye
terminators were removed from the reaction with CENTRI-SEP
columns (Princeton Separations), and the solutions were dried. The
residues were resuspended in a formamide solution (4 µL) and were
fractionated on the ABI 377 DNA sequencer, using a 6% polyacryla-
mide-6 M urea gel. The sequence data were analyzed with the Applied
Biosystems PRISM sequencing analysis v3.2 software.
Acknowledgment. We thank Akira Sato and Tsuyoshi
Fujiwara for synthesizing the nucleoside derivatives. This work
was supported by a Grant-in-Aid for Scientific Research
(KAKENHI 15350097 (I.H.), 19201046 (I.H.), and 18710197
(M.K.)) from the Ministry of Education, Culture, Sports,
Science, and Technology and by the RIKEN Structural Ge-
nomics/Proteomics Initiative (RSGI), the National Project on
Protein Structural and Functional Analyses, Ministry of Educa-
tion, Culture, Sports, Science and Technology of Japan.
Supporting Information Available: NMR and MS data for
the nucleoside derivatives of Pn. Sequences of DNA fragments
for the PCR amplification and sequencing experiments. Se-
quencing patterns of the control DNA fragments containing
1-10% of the A-T pair in place of the unnatural base pair.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Vmax) were obtained from Hanes-Woolf plots of [dNTP]/V0 against
[dNTP].
Primer Extension Experiments with KF exo+. A primer (23-mer)
labeled with 32P at the 5′-end was annealed to a 35-mer template in an
annealing buffer (20 mM Tris-HCl (pH 7.5), 14 mM MgCl2, and 0.2
mM DTT), by heating at 95 °C and slow cooling to 4 °C. The duplex
(11) Ludwig, J.; Eckstein, F. J. Org. Chem. 1989, 54, 631-635.
JA073830M
9
J. AM. CHEM. SOC. VOL. 129, NO. 50, 2007 15555