Synthesis and enzymatic incorporation of modified deoxyadenosine triphosphates

Article abstract of DOI:10.1002/ejoc.200800381

Several deoxyadenosine triphosphates containing modifications at the 8-position have been synthesized. Suitably protected 8-bromodeoxyadenosines were coupled with five imidazole-containing moieties by nucleophilic aromatic substitution or Sonagashira coup

Full text of DOI:10.1002/ejoc.200800381

FULL PAPER
DOI: 10.1002/ejoc.200800381
Synthesis and Enzymatic Incorporation of Modified Deoxyadenosine
Triphosphates
Curtis Lam,[a] Christopher Hipolito,[a] and David M. Perrin*[a]
Keywords: Nucleotides / Heterocycles / SELEX / Bioorganic chemistry / DNA / Enzymes
Several deoxyadenosine triphosphates containing modifica-
mercially available DNA polymerases, and it was found that
tions at the 8-position have been synthesized. Suitably pro- two of the modified dATPs could be effectively taken up as
tected 8-bromodeoxyadenosines were coupled with five
imidazole-containing moieties by nucleophilic aromatic sub-
stitution or Sonagashira coupling to give modified nucleo-
sides that were then triphosphorylated. Incorporation assays
were performed for these modified residues with many com-
substrates by Sequenase V2.0. These two residues are candi-
dates for substrates in combinatorial selections in the search
for improved catalysis from DNAzymes.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
have identified binding or catalytic activities that address a
quantitative improvement in activity that cannot be iden-
tified by an unmodified selection. One case in which modifi-
cation has demonstrably improved catalysis is the case of
divalent metal-independent ribonuclease activity.
Appending extra chemical functionality to nucleic acids
has led to the discovery of a sequence-specific RNase A
mimic, 925-11.[12] RNase A is an endonuclease that cleaves
RNA to afford a 2Ј–3Ј cyclic phosphate intermediate prod-
uct. Two histidine residues in the active site catalyze this
reaction and several nearby lysine residues help to stabilize
the reaction intermediate.[17] The combinatorially selected
DNAzyme 925-11 contains two modified residues: 8-[2-(4-
Introduction
In vitro selection techniques[1–3] (e.g., SELEX) have en-
abled the systematic screening of libraries of up to 1015 ran-
dom nucleic acid sequences for binding and catalysis. In the
past 15 years, catalytic oligonucleotides have been selected
for reactions ranging from the sequence-specific cleavage of
RNA[4,5] to the formation of peptide bonds.[6,7] Aptamers
to a broad collection of targets have been discovered as
well, and these have been reviewed.[8,9]
Compared with proteins, nucleic acids lack many of the
functional groups that protein enzymes employ for catalysis.
This deficiency has in part limited the types of reactions
that can be catalyzed and the catalytic rates that can be
observed for nucleic acid catalysts. To overcome this appar-
ent shortcoming, several groups have synthesized modified
deoxyribonucleoside triphosphates (dNTPs) that can be in-
corporated into oligonucleotides by polymerases. The func-
tionalities that have been synthetically attached include flu-
orescent groups,[10] boronic acids[11] and nearly every one
of the amino acid side-chains.[12–16] Most of these modified
nucleosides are substrates for at least one polymerase and
many of them have also been successfully used in ex-
ponential amplification (PCR). Successful incorporation is
a necessary precondition for combinatorial selections using
modified dNTPs in place of their natural counterparts, but
not a sufficient one for obtaining activity, which will depend
on the modified dNTPs.
imidazoyl)ethylamino]-2Ј-deoxyadenosine
and
5-allyl-
amino-2Ј-deoxyuridine. The added imidazoles appear to fa-
cilitate general acid–general base catalysis, whereas the
amines provide electrostatic stabilization. Together, these
two modifications gave rise to a DNAzyme that could facili-
tate the cleavage of a phosphodiester bond downstream of
an embedded ribocytidine in the absence of a divalent metal
ion. The observed catalytic rate was substantially superior
to divalent metal-independent RNA-cleaving DNAzymes
that contain only unmodified nucleotides.[18–20]
Although triphosphate 1 (Figure 1) has been successfully
used in a combinatorial selection,[12] the achieved rate for
self-cleavage was still significantly less than what is typically
observed for several divalent metal-dependent, unmodified,
self-cleaving catalysts, not to mention protein enzymes. One
hypothesis to account for this observation is that the linker
between the imidazole and the nucleotide that is part of the
catalytic scaffold was not ideal. Because nucleic acids are
much more flexible polymers than proteins, a more rigid
and perhaps shorter linker may allow the catalyst to orient
the substrate in a more ordered fashion. In addition, the
exact composition of the linker may also be very important.
Despite a plethora of reports detailing the substrate
properties of various modified dNTPs, far fewer reports
[a] Department of Chemistry, University of British Columbia,
2036 Main Mall, Vancouver, BC V6T 1Z1, Canada
Fax: +1-604-822-2847
E-mail: dperrin@chem.ubc.ca
Supporting information for this article is available on the
WWW under http://www.eurjoc.org/ or from the author.
Eur. J. Org. Chem. 2008, 4915–4923
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4915

C. Lam, C. Hipolito, D. M. Perrin
FULL PAPER
Triphosphate 1 contains an amine at the 8-position that will enable the evaluation of the dependence of catalytic
could participate in enhanced Hoogsteen base-pairing.[21] rate on linker length. Triphosphate 4 is the same length as
The presence or absence of this interaction might have an 1, but contains an extra carboxylate group. This carboxylate
effect on the three-dimensional structure of a DNAzyme.
group could participate in catalysis or chelate metal ions.
The remaining two linkages (5 and 6) do not possess an
amino moiety and have been designed to assess the extent
of hydrogen-bonding of the 8-amino functionality. Triphos-
phate 6 contains a flexible linker similar to 2, and 5 con-
tains a rigid alkynyl unit.
Results and Discussion
Figure 1. Imidazole-containing 8-modified deoxyadenosine triphos-
phates.
Design and Synthesis of Modified dATPs
Thus far, most efforts towards modified nucleotides have
been focused on tethering the functional groups using longer
linkers, which for reasons that remain unclear enhance
polymerase uptake.[11,13] Such nucleotides have for the most
part been good substrates for polymerases in both primer
extensions and PCR. When short linkers are used, the func-
tional groups may clash with the polymerase scaffold that
makes contacts with the growing strand such that incorpo-
ration of these modified residues becomes much more diffi-
cult. In an attempt to assess the effect of the linker length/
composition on the substrate properties of these triphos-
phates with the long-term goal of testing the hypothesis that
shorter linkers provide better catalysts, we have synthesized
five imidazole-modified deoxyadenosine triphosphates con-
taining linkers of various lengths, compositions, and flexi-
bilities. These triphosphates have been studied for their rela-
tive ease of enzymatic incorporation.
The synthesis of the modified nucleosides is outlined in
Scheme 1 and Scheme 2. To synthesize the nucleosides con-
taining amine linkers, N6-benzoyl-8-bromo-5Ј-(dimeth-
oxytrityl)deoxyadenosine[25,26] (10) was heated with the ap-
propriate imidazole-containing dihydrochloride salt in tri-
ethylamine and ethanol. Both 11 and 12 were readily pre-
pared in this fashion. Displacement of the bromine substit-
uent with histidine methyl ester, however, required pro-
longed heating. By using methanol as solvent to prevent
transesterification, the desired product was isolated in low
yield, giving rise to several spots in the TLC analysis. The
most significant byproduct was isolated and found to be
debenzoylated 10. Attempts with other solvents such as
DMF, DMSO, or pyridine showed little improvement.
Recently, an investigation of the enzymatic incorporation
of dATPs modified at the 7- and 8-positions has been pub-
lished.[22] This study found that 8-substituted dATPs could
not be readily incorporated by polymerases. However, when
the same modifications were tethered from the 7-position
of 7-deazadATP, the modified nucleotides could be taken
up as substrates. The results of Capek et al.,[22] along with
studies from other groups who have synthesized 7-deaza-
dATPs,[23,24] have all shown that modifications at this posi-
tion can be tolerated very well by polymerases. In light of
these findings, it is important to note that 8-histaminyl-
dATP (1) is a poor substrate that cannot be readily incorpo-
rated more than three times in a row at best. Nevertheless,
despite poor substrate properties, triphosphate 1 was se-
lected in the context of the 925-11 DNAzyme and gave rise
to the first M2+-independent RNaseA-like catalyst. This
would suggest that although incorporation is necessary for
selection, highly efficient incorporation may not be. Thus,
the modified dATPs in this study were chosen largely be-
cause they are somewhat analogous to 1 and will provide a
more comprehensive assessment of the limitation of incor-
porating 8-modified adenosines with short linkers.
Scheme 1.
The synthesis of 15 was initially attempted unsuccessfully
The five linkages that were synthesized (Figure 1) are with 4-ethynylimidazole[27] under Sonagashira condi-
aminomethyl (2), aminopropyl (3), (S)-1-amino-1-carboxy- tions.[28] It was found that protected [1-(4,4Ј-dimethoxytri-
ethyl (4), ethynyl (5), and ethyl (6). Triphosphates 1 and 2 tyl)-1H-4-imidazolyl]ethyne (14) would undergo this coup-
4916
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4915–4923

Enzymatic Incorporation of Deoxyadenosine Triphosphates
The nucleosides were then converted into triphosphates
by applying the multistep “one-pot” synthesis of Ludwig
and Eckstein.[29] Briefly, this procedure involved treating
the 3Ј-protected nucleoside with salicyl phosphorochlorid-
ite. The addition of pyrophosphate displaces the salicylate
and the phosphite is oxidized to the phosphate with I2.
Aqueous ammonia is used for global deprotection. Triphos-
phates 2, 3, 5, and 6 were successfully isolated in this fash-
ion. For 4, the methyl ester of compound 19 was first re-
moved with lithium carbonate at slightly elevated tempera-
tures before deprotection with aqueous ammonia. The
crude product mixtures were first purified by preparative
TLC and then by HPLC to afford the final products.
MALDI-TOF MS was used to verify the identity of each
modified dATP.
Scheme 2.
The final yields of the modified dATPs were quite low
(3–23%). The initial steps of the reaction were monitored
by 31P NMR, and the spectra obtained did not reveal any
irregularities. Possible reasons for the low yields include in-
terference from the unprotected imidazoles and incomplete
removal of the benzoyl protecting group. As the quantities
isolated were more than what was required for enzymatic
incorporation and combinatorial selection, the reactions
were only carried out once or twice with little optimization.
ling much more efficiently. As the dimethoxytrityl group
was compatible with our synthesis, no additional deprotec-
tion steps were required. Nucleoside 16 was obtained by
the hydrogenation of 15 in the presence of platinum oxide,
pyridine, and CH2Cl2. The pyridine/CH2Cl2 ratio of 1:60
was found to be quite important. When the amount of pyr-
idine was increased, the reaction became sluggish, and
when this ratio was decreased to 1:1000, significant depro-
tection of the dimethoxytrityl group was observed.
Attempts to reduce the alkyne to the cis-alkene under vari-
ous conditions, including the use of Lindlar’s catalyst and
palladium over barium sulfate, were unsuccessful and led to
either unreacted starting materials or complete reduction to
16.
The dimethoxytrityl- and benzoyl-protected nucleosides
were then methoxyacetylated with methoxyacetic anhydride
in pyridine and detritylated in the presence of acetic acid
and water (4:1) (Scheme 3). This protocol was effective for
all of the modifications except for 15, for which the depro-
tection step gave poor yields. It was later found that a mix-
ture of dichloromethane and acetic acid (3:1) gave much
better yields.
Enzymatic Incorporation of Modified dATPs
Incorporation assays of the unnatural nucleotides (in-
cluding 1) were attempted with several DNA polymerases.
The previously studied triphosphate 1 was included both as
a control and to study its incorporation more extensively.
The incorporations were carried out by extending a 30 nu-
cleotide primer along several 55 nucleotide template oligo-
nucleotides (Figure 2). The enzyme Sequenase V2.0 has pre-
viously been used successfully with 1 as a substrate, so ini-
tial incorporation assays were performed with this enzyme.
For the most part, the incorporation gels did not include 4
as this nucleoside triphosphate was found to photodecom-
pose during HPLC purification to give 8-amino-dATP and
a fragment that appears to be urocanic acid (data not
shown).[30]
Figure 2. Template sequence regions used for the primer extension
experiments (the primer binding region has been omitted).
Figure 3 shows a Sequenase V2.0 primer extension using
the T1 template. This template required a total of eight
dATPs to be incorporated, and in two locations along the
template, two dATPs must be incorporated consecutively.
Scheme 3.
Eur. J. Org. Chem. 2008, 4915–4923
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4917

C. Lam, C. Hipolito, D. M. Perrin
FULL PAPER
Under these conditions, both 1 and 2 were taken up ef-
The enzymatic uptake of the modified dATPs by Dpo4
ficiently. Their corresponding full extension products mi- in the absence of other natural nucleotides is shown in Fig-
grated somewhat more slowly than the product band corre- ure 4. By using the T2 template, the enzyme was able to
sponding to unmodified DNA, probably due to the extra incorporate 1 and 3 with the extension going to completion
molecular bulk/size of the modifications. In contrast, Se- for 1. For the other three modified dATPs, Dpo4 was only
quenase V2.0 had difficulty incorporating both 6 and 3 as able to incorporate one to two residues before stopping.
substrates. The bands corresponding to fully extended ma- These results initially looked very promising as Dpo4
terial were very faint. Their intensities suggest that 3 was seemed to be more capable of tolerating the modified
tolerated slightly more than 6. Lane 10, which contains sub- dATPs than Sequenase V2.0. However, when the other
strate 5, looked identical to lane 5 in which no dATP was three natural dNTPs were included in the incorporation ex-
used and hence 5 is not a substrate for Sequenase V2.0.
periments, the results were very different; the extension
went no further than what was observed in Figure 4. In
other words, the enzyme was able to incorporate up to five
modified dATPs in a row, but could not accept any natural
substrates afterwards.
Figure 4. Dpo4 primer extensions of modified dATPs at 37 °C using
the T2 template. Lane 1: primer; lane 2: dATP (5 µ); lane 3: 2; lane
4: 3; lane 5: 6; lane 6: 5; lane 7: 1. Modified dATPs were used at a
concentration of 100 µ.
Figure 3. Sequenase V2.0 primer extensions of modified dATPs at
37 °C using the T1 template. Lane 1: primer only; lane 2: primer and
enzyme; lane 3: dCTP and dGTP; lane 4: dCTP, dGTP, and dATP;
lane 5: dCTP, dGTP, and dTTP; lanes 6–11 all contain dCTP, dGTP,
and dTTP; lane 6: dATP; lane 7: 2; lane 8: 3; lane 9: 6; lane 10: 5;
lane 11: 1. All nucleotides were used at a concentration of 50 µ.
With this result in mind, full extensions using Dpo4 and
the T1 template were attempted (Figure 5). As this template
only had regions that required up to two modified dATPs
to be incorporated in a row, it was thought that 1 and 3
would both be easily fully extended. Unfortunately, no
In an effort to improve the incorporation of the modified bands corresponding to the fully extended products could
dATPs, the pH of the buffers that were used was raised or be identified.
lowered by one pH unit. When a different template, T2, was
To understand this result, the incorporation of 1 using
used, which requires five dATPs to be incorporated one af- Dpo4 was attempted with two other templates (Figure 6).
ter another, the enzymatic uptake of 3 was somewhat im- The T3 template contains the same base composition as T1.
proved and the enzymatic uptake of the other modified It also requires up to two modified dATPs to be incorpo-
dATPs was unaffected. This observation initially looked rated in a row, but the locations of the modifications are
promising. However, when the T1 template was used under somewhat more spaced apart. The T4 template, which re-
full extension conditions at the elevated pH, 5, 6, 2, and quired 10 modified dATPs to be incorporated sequentially,
3 produced no fully extended product. Modified dATP 1 was also used to gauge how many modified nucleotides
produced some fully extended product, but the band was could be incorporated sequentially one after another. The
very faint.
full extension with T3 was successful at both 37 and 70 °C.
Following the examination of Sequenase V2.0, other In the case of the reaction at 70 °C, a slightly longer prod-
polymerases were investigated. We opted for enzymes that uct was produced as well. Use of T4 under the same condi-
lacked 3Ј–5Ј exonuclease activity. The enzymes tested in- tions gave products with five to six modified nucleotides in
cluded Taq, Vent exo-, Bst large fragment, Phusion, and a row.
Dpo4. Of these enzymes, only Dpo4 was able to incorporate
Through extensive incorporation studies, it was found
the modified dATPs. Dpo4 is a DNA polymerase that be- that the 8-modified dATPs studied were somewhat tolerated
longs to the Y family, a group of enzymes that are capable as substrates for DNA polymerases. Sequenase V2.0 was
of bypassing lesions. Isolated from Sulfolobus solfataricus, found to be the enzyme that could most tolerate these
the enzyme is thermostable, but is also active at physiologi- modifications and could consecutively incorporate up to
cal temperatures.[31]
three residues of 1. When the linker region between the
4918
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4915–4923

Enzymatic Incorporation of Deoxyadenosine Triphosphates
atoms as the linker of 2. A more plausible explanation for
this would be that the amine can potentially hydrogen bond
to the 5Ј-oxygen on the sugar. In double-stranded DNA,
nucleotides adopt an anti conformation in which the 8-posi-
tion is on the same side as the phosphates. When modifica-
tions were added to this position, the extra steric bulk
would tend to favor the syn conformation. The presence of
a hydrogen bond between the linker and the sugar would
have helped to stabilize the anti conformation and allow the
amine containing modified dATPs to be relatively better
substrates for polymerases.
Even though the 8-modified dATPs were to some extent
difficult to incorporate, several of them may still be used
for in vitro selection. Earlier, Roth and Breaker carried out
a selection that employed free histidine as a cofactor for the
self-cleavage of an embedded RNA residue.[32] Their results
show that imidazoles are probably not essential to the fold-
ing of a catalytic motif, and hence a DNAzyme should only
require a very small number of properly positioned catalytic
residues within the random region of an oligonucleotide li-
brary. As long as this requirement can be met, the modified
nucleotide will be a useful substrate. The DNAzyme 925-11
was the most active sequence that arose from a selection
using Sequenase V2.0 and 1, and this sequence only re-
quired up to two residues of 1 to be incorporated in a row.
Consequently, both 2 and 3 will serve as candidates for the
selection of more catalytically active DNAzymes. The use
of 6 may be possible as well.
Figure 5. Dpo4 primer extension with the T1 template at 37 °C. Lane
1: primer only; lane 2: primer and Dpo4; lane 3: dCTP and dGTP;
lane 4: dCTP, dGTP, and dATP; lane 5: dCTP, dGTP, and dTTP;
lanes 6–12 all contain dCTP, dGTP, and dTTP; lane 6: dATP; lane
7: 2; lane 8: 3; lane 9: 6; lane 10: 5; lane 11: 4; lane 12: 1. Nucleotides
were used at a concentration of 50 µ.
Conclusions
Five imidazole-containing modified dATPs have been
successfully synthesized. The process involved coupling of
the imidazole-containing moieties to the 8-position of aden-
osine followed by triphosphorylation. In general, the modi-
fied dATPs that contain alkyl, aminomethyl, and ami-
nopropyl linkages could be incorporated by Sequenase V2.0
sufficiently when the template did not require very many
modified residues to be incorporated. As a DNAzyme will
only require a small number of imidazoles for catalysis, at
least two of these nucleotides are candidates for the in vitro
selection of improved catalytic systems.
This work was undertaken with a view to developing
novel DNA polymerase substrates for use in combinatorial
selections. Some 8-substituted dATPs are not competent as
substrates for most DNA polymerases and thus they will
not find use in such selections. It is nevertheless essential to
report these findings in order to characterize the current
limitations of enzymatic incorporation for those who are
pursuing the use of modified dNTPs in combinatorial selec-
tions and for those who are seeking to evolve new DNA
polymerases with relaxed substrate specificities.
Figure 6. Dpo4 primer extension of 1 using the T3 (A) and T4 (B)
templates. Samples used in the incorporation experiments were incu-
bated for 2 h at 37 °C (lanes 1–5) or at 70 °C (lane 6). Lane 1: primer
and enzyme; lane 2: dATP; lane 3: 1; lane 4: dATP, dGTP, dTTP,
and dCTP; lanes 5 and 6: 1, dGTP, dTTP, and dCTP. Natural nucleo-
tides were used at a concentration of 50 µ and modified nucleotides
were used at a concentration of 100 µ.
imidazole and adenine was elongated or shortened by one
methylene unit, only two modified residues in a row could
be incorporated. Similarly, Dpo4 also preferred 1 as a sub-
strate over 2 and 3. These observations suggest that the
linker length in 1 may be optimal for incorporation. For the
two nucleotides that did not contain an amine in the linker,
incorporation was found to be much more difficult. This
observation was probably not due to linker length as the
Experimental Section
General: Starting materials were purchased from Sigma–Aldrich
linkers of both 5 and 6 contained the same number of and Fisher Scientific. 4-Aminomethylimidazole dihydrochloride (7)
Eur. J. Org. Chem. 2008, 4915–4923
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4919

C. Lam, C. Hipolito, D. M. Perrin
FULL PAPER
was prepared according to the literature.[33] Flash chromatography
was performed on silica gel (230–400 mesh) from Silicycle. Thin-
layer chromatography (TLC) and preparatory TLC were performed
on precoated glass-backed plates of silica gel 60 F254 from EMD
Chemicals. Natural dNTPs and Klenow Fragment exo– were pur-
chased from Fermentas Life Sciences. T4 polynucleotide kinase,
Therminator DNA polymerase, Bst DNA polymerase, large frag-
ment, Vent exo–, Taq DNA polymerase, and Phusion DNA poly-
merase were purchased from New England Biolabs. Sequenase Ver-
sion 2.0 T7 DNA polymerase, inorganic pyrophosphatase and sin-
gle-stranded DNA binding protein (SSB) were purchased from GE
Healthcare. DNA polymerase IV was purchased from Trevigen.
NMR spectra were obtained by using a Bruker AV-300 or AV-400
was then suspended in aq. HCl (25 mL, 3 ) and acetone was
added until all solid components dissolved. The contents were re-
fluxed for 1 h. After removal of the acid, the product was recrys-
tallized with EtOH/EtOAc to give 216 mg of white crystals. 1H
NMR (300 MHz, D2O, 25 °C): δ = 8.44 (s, 1 H, 2-H), 7.12 (s, 1 H,
5-H), 2.91 (t, J = 7.8 Hz, 2 H, 8-H), 2.70 (t, J = 7.6 Hz, 2 H, 6-H),
1.91 (dt, J = 7.8, 7.6 Hz, 2 H, 7-H) ppm.
[1-(4,4Ј-Dimethoxytrityl)-1H-4-imidazolyl]ethyne (14): 4-(2,2-Di-
bromovinyl)-1-(4,4Ј-dimethoxytrityl)imidazole (2.68 g, 4.8 mmol)
was suspended in THF (100 mL) and cooled to –78 °C. tert-Butyl-
lithium (20 mL, 34 mmol) was added to this solution. The cold
bath was then removed. After 5 min of stirring, MeOH (10 mL)
was slowly added followed by the addition of H2O (30 mL). The
THF was evaporated. The resulting crude product was extracted
with CH2Cl2 (60 mL), dried with sodium sulfate (anhydrous), and
purified by flash chromatography. An off-white solid (1.53 g, 81%)
was eluted with EtOAc/hexanes (2:3). 1H NMR (300 MHz, CDCl3,
25 °C): δ = 7.36 (s, 1 H, 2-H Imid), 7.36–7.24 (m, 3 H, 3-H Phe, 4-
H Phe), 7.20–7.15 (m, 3 H, 5-H Imid, 2-H Phe), 7.00 (d, J = 9.0 Hz,
4 H, 2-H PheOMe), 6.82 (d, J = 9 Hz, 4 H, 3-H PheOMe), 3.79 (s,
6 H, OMe), 3.03 (s, 1 H, HCC) ppm. LRMS (ESI+): calcd. for
C26H22N2O2Na+: 417.2; found 417.1.
1
instrument. 13C, 31P, and H NMR spectra were calibrated to the
signal of the deuterated solvent. ESI-MS were obtained with a
Micromass LCT, Bruker Esquire-LC or Waters LC/MS instrument
in positive ion mode. MALDI-TOF MS were obtained with a
Bruker Biflex instrument in positive ion mode. Solvents were dried
by distillation or with molecular sieves (4 Å).
HPLC of Triphosphates: HPLC purification was carried out with
an Agilent 1100 system using a Phenomenex Jupiter 10µ C4 300A
column. The flow rate was set at 1 mL/min and the eluents con-
tained 0.05  ammonium acetate (pH 7). Because the absorption of
imidazole is very far from the absorbance maxima of the modified
adenosines, the εmax value for dATP (15400 –1 cm–1) was used to
quantify the synthesized triphosphates. The HPLC gradient pro-
grams used are shown in Table 1.
N6-Benzoyl-5Ј-O-(4,4Ј-dimethoxytrityl)-8-[(4-imidazolyl)methylami-
no]deoxyadenosine (11): 4-(Aminomethyl)imidazole dihydrochlo-
ride (7) (460 mg, 2.72 mmol), was suspended in Et3N (1.52 mL,
10.9 mmol) and EtOH (10 mL). This mixture was added to a solu-
tion of N6-benzoyl-8-bromo-5Ј-O-(4,4Ј-dimethoxytrityl)-2-deoxya-
denosine (10) (1.00 g, 1.36 mmol) in EtOH (10 mL). After 4.5 h of
stirring at 65 °C, the solvent was removed and the reaction was
redissolved in CH2Cl2 and washed with aq. NaHCO3 (5%). The
organic extracts were combined and dried with Na2SO4 (anhy-
drous) and purified on silica gel. The desired product was eluted
with MeOH/CHCl3 (5:95) to give 650 mg (63%) of a slightly off-
white solid. 1H NMR (400 MHz, CD2Cl2, 25 °C): δ = 8.38 (s, 1 H,
2-H), 8.05 (d, J = 7.3 Hz, 2 H, 2-H Bz), 7.61 (s, 1 H, 1-H Imid),
7.61 (t, J = 7.4 Hz, 1 H, 4-H Bz), 7.53 (dd, J = 7.4, 7.3 Hz, 2 H,
3-H Bz), 7.40 (d, J = 6.7 Hz, 2 H, 2-H Phe), 7.35–7.25 (m, 7 H, 3-
H Phe, 4-H Phe, 2-H PheOMe), 6.82 (d, J = 8.8 Hz, 4 H, 3-H
PheOMe), 6.47 (dd, J = 8.9, 5.7 Hz, 1 H, 1Ј-H), 6.39 (s, 1 H, 5-H
Imid), 4.75–4.70 (m, 1 H, 3Ј-H), 4.15–4.00 (m, 2 H, CH2-Imid, 4Ј-
H), 3.77 (s, 6 H, OMe), 3.60–3.50 (m, 1 H, CH2-Imid), 3.50 (dd, J
= 10.6, 2.5 Hz, 1 H, 5Ј-H), 3.41 (dd, J = 10.6, 3.0 Hz, 1 H, 5Ј-H),
2.88 (ddd, J = 13.5, 8.9, 6.3 Hz, 1 H, 2Ј-H), 2.30 (ddd, J = 13.5,
5.7, 1.9 Hz, 1 H, 2Ј-H) ppm. 13C NMR (100 MHz, CD2Cl2, 25 °C):
δ = 166.5, 159.5, 154.7 (C-8), 154.4 (C-4), 148.8 (C-2), 144.7, 143.9
(C-6), 135.9 (C-2 Imid), 135.7, 134.1 (C-4 Imid), 133.2, 130.9,
130.8, 129.3, 129.0, 128.6, 128.6, 127.9, 124.1 (C-5), 113.8, 87.6,
Table 1. HPLC gradients.
Time [min] Program A: ACN/H2O [%] Program B: ACN/H2O [%]
0
1
5
25
50
50
0
0
4
10
18
19
24
25
30
0
MALDI-TOF MS of Triphosphate Products: The matrix solution
was prepared by suspending 3-hydroxypicolinic acid (30 mg) in ace-
tonitrile (0.25 mL) and H2O (0.25 mL) and mixing this solution
with aq. ammonium citrate (2.4%) in a ratio of 4:1. Prior to sample
loading, a few cation-exchange beads (Bio-Rad AG50W-X8 resin
converted into the ammonium form) were added to the nucleoside
triphosphate samples (2 µL, 100 µ) and left at room temperature
for at least 15 min. Matrix solution (1 µL) was first applied to the
MALDI target followed by the nucleoside triphosphate sample
(1 µL) and the solutions were mixed thoroughly. Care was taken to
ensure that a portion of the cation-exchange beads were transferred
as well. The samples were analyzed in the positive ion mode with
a reflectron detector.
86.8 (C-4Ј), 84.1 (C-1Ј), 71.9 (C-3Ј), 63.5 (C-5Ј), 55.8, 39.3 (C-2Ј),
37.3 (CH2-Imid) ppm. HRMS (ESI+): calcd. for C42H41N8O6
753.3149; found 753.3160.
:
+
N6-Benzoyl-5Ј-O-(4,4Ј-dimethoxytrityl)-8-[3-(4-imidazolyl)propyl-
amino]-2Ј-deoxyadenosine (12): Compound 12 was synthesized in
the same way as 11. 4-(3-Aminopropyl)imidazole dihydrochloride
(8) (182 mg, 0.92 mmol) and 10 (450 mg, 0.61 mmol) were con-
verted into 190 mg (40%) of a flaky, slightly off-white solid. 1H
NMR (400 MHz, CD3OD, 25 °C): δ = 8.28 (s, 1 H, 2-H), 8.04 (d,
J = 7.3 Hz, 2 H, 2-H Bz), 7.62 (t, J = 7.3 Hz, 1 H, 4-H Bz), 7.51
(t, J = 7.3 Hz, 2 H, 3-H Bz), 7.49 (s, 1 H, 2-H Imid), 7.31 (d, J =
6.8 Hz, 2 H, 2-H Phe), 7.25–7.13 (m, 7 H, 3-H Phe, 4-H Phe, 2-H
PheOMe), 6.78–6.68 (m, 5 H, 3-H PheOMe, 5-H Imid), 6.26 (t, J
= 6.4 Hz, 1 H, 1Ј-H), 4.78 (ddd, J = 6.5, 5.0, 3.9 Hz, 1 H, 3Ј-H),
4.15–4.05 (m, 1 H, 4Ј-H), 3.70 (s, 3 H, OMe), 3.69 (s, 3 H, OMe),
4-(3-Aminopropyl)imidazole Dihydrochloride (8): 3-(N-Trityl-4-
imidazolyl)acrylonitrile[34] (1.81 g, 5.0 mmol) was converted into 8
according to the procedure of Adger and Surtees.[35] However,
recrystallization of the product could not be performed as de-
scribed. The crude product and trityl chloride (2.79 g, 10 mmol)
were suspended in Et3N (5.6 mL, 40 mmol) and DMF (50 mL).
After stirring at room temperature for 12 h, the solvent was re-
moved. The crude product was redissolved in CH2Cl2, washed with
aq. NaHCO3 (5%), and dried with sodium sulfate (anhydrous). The
bis-tritylated intermediate was purified on silica gel. A white solid
(1.03 g, 34%) was eluted with EtOAc/hexanes (3:2). This product
4920
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4915–4923

Enzymatic Incorporation of Deoxyadenosine Triphosphates
3.52 (ddd, J = 13.3, 6.5, 6.4 Hz, 1 H, 2Ј-H), 3.45–3.25 (m, 4 H, 5Ј-
H, H2C-NH), 2.58 (t, J = 7.3 Hz, 2 H, CH2-Imid), 2.32 (ddd, J =
13.3, 6.4, 5.0 Hz, 1 H, 2Ј-H), 1.93 (q, J = 7.3 Hz, 2 H, CH2CH2C-
2H) ppm. 13C NMR (100 MHz, CD3OD, 25 °C): δ = 168.0, 160.1,
160.0, 156.1 (C-8), 154.6 (C-4), 149.0 (C-2), 146.1, 144.4 (C-6),
7.3, 5.8 Hz, 1 H, 1Ј-H), 5.02–4.96 (m, 1 H, 3Ј-H), 4.14–4.08 (m, 1
H, 4Ј-H), 3.82 (s, 6 H, OMe), 3.75 (s, 3 H, OMe), 3.74 (s, 3 H,
OMe), 3.52–3.35 (m, 3 H, 5Ј-H, 2Ј-H), 2.42 (ddd, J = 13.0, 7.3,
5.0 Hz, 1 H, 2Ј-H) ppm. 13C NMR (100 MHz, CD2Cl2, 25 °C): δ
= 164.6, 159.7, 158.8, 152.9 (C-2), 151.3 (C-4), 149.4 (C-6), 145.4,
137.1, 136.9, 135.7 (C-2 Imid), 135.4 (C-4 Imid), 133.7, 131.3, 142.7, 140.2 (C-2 Imid), 137.8 (C-8), 136.4, 136.3, 134.3, 133.0,
131.2, 129.7, 129.4, 129.2, 128.7, 127.8, 124.9 (C-5), 118.0 (C-5 132.3, 131.3, 130.4, 130.2, 129.7, 129.2, 128.9, 128.5, 128.4, 128.1,
Imid), 114.0, 87.6 (C-4Ј), 87.5 (C-1Ј), 85.2, 72.6 (C-3Ј), 64.7 (C-5Ј), 128.0, 126.9, 123.7 (C-5), 120.9 (C-5 Imid), 113.8, 113.2, 92.0 (CC-
55.7, 43.4 (CH2-NH), 38.2 (C-2Ј), 30.0 (CH2-CH2-CH2), 24.7 Imid), 86.7, 86.4 (C-4Ј), 85.5 (C-1Ј), 78.8 (CC-Imid), 75.9, 72.8 (C-
(CH2-Imid) ppm. HRMS (ESI+): calcd. for C44H44N8O6
781.3462; found 781.3459.
:
3Ј), 64.5 (C-5Ј), 55.7, 55.5, 37.8 (C-2Ј) ppm. HRMS (ESI+): calcd.
+
for C64H55N7O8+: 1072.4010; found 1072.4011.
N6-Benzoyl-5Ј-O-(4,4Ј-dimethoxytrityl)-8-[(S)-1-methylcarboxy-2-
(4-imidazolyl)ethylamino]-2Ј-deoxyadenosine (13): -Histidine
methyl ester dihydrochloride (9) (0.48 g, 2 mmol) and Et3N
(1.1 mL, 7.9 mmol) were dissolved in MeOH (4 mL). Nucleoside
10 (735 mg, 1 mmol) was added and the mixture was stirred at
50 °C for 2 d. Following the removal of solvent, the crude product
was redissolved in CH2Cl2 and washed with aq. NaHCO3 (5%).
Sodium sulfate (anhydrous) was used to dry the product. Purifica-
tion of the product was carried out on silica gel. The product was
eluted in MeOH/CHCl3 (6:94) to give 131 mg (16%) of a slightly
off-white powder. A yellow foam (212 mg, 33.6%) was isolated and
was identified as the debenzoyled product of 7. 1H NMR
N6-Benzoyl-5Ј-O-(4,4Ј-dimethoxytrityl)-8-{2-[N-(4,4Ј-dimethoxytri-
tyl)-4-imidazolyl]ethyl}deoxyadenosine (16): Compound 15 (500 mg,
0.48 mmol) and PtO2 (250 mg, 1.10 mmol) were suspended in pyr-
idine (2.5 mL) and CH2Cl2 (150 mL). This mixture was stirred at
room temperature under H2 (1 atm) for 4 h. Following the removal
of solvent, the crude product was dissolved in MeOH. Filter paper
was used to remove the catalyst. The MeOH was then evaporated.
The resulting material was redissolved in CH2Cl2, washed with aq.
NaHCO3 (5%), and dried with Na2SO4 (anhydrous). Purification
by silica gel flash chromatography using Et3N/MeOH/CHCl3
(1:20:980) gave 265 mg of 16 (53%). 1H NMR (400 MHz, CD2Cl2,
25 °C): δ = 9.00 (s, 1 H, NH-CO), 8.46 (s, 1 H, 2-H), 7.93 (d, J =
(400 MHz, CD2Cl2, 25 °C): δ = 8.97 (s, 1 H, CO-NH), 8.31 (s, 1 7.5 Hz, 2 H, 2-H Bz), 7.60 (t, J = 7.4 Hz, 1 H, 4-H Bz), 7.51 (dd,
H, 2-H), 7.96 (d, J = 7.4 Hz, 2 H, 2-H Bz), 7.59 (t, J = 7.4 Hz, 1
H, 4-H Bz), 7.50 (t, J = 7.4 Hz, 2 H, 3-H Bz), 7.45–7.15 [m, 9 H,
2-H Phe, 3-H Phe, 4-H Phe, 2-H PheOMe], 7.39 (s, 1 H, 2-H Imid),
6.85–6.70 (m, 5 H, 3-H PheOMe, 5-H Imid), 6.23 (t, J = 6.6 Hz, 1
H, 1Ј-H), 4.82 (dd, J = 6.2, 4.0 Hz, 1 H, CH-CO2), 4.61 (m, 1 H,
J = 7.5, 7.4 Hz, 2 H, 3-H Bz), 7.36 (d, J = 6.7 Hz, 2 H, 2-H Phe),
7.32 (s, 1 H, 2-H Imid), 7.30–7.12 (m, 10 H, 4-H Phe, 3-H Phe, 4-
H Phe, 3-H Phe, 2-H PheOMe), 7.03 (d, J = 6.9 Hz, 2 H, 2-H Phe),
6.96 (d, J = 8.1 Hz, 4 H, 2-H PheOMe), 6.80–6.68 (m, 8 H, 3-H
PheOMe), 6.56 (s, 1 H, 5-H Imid), 6.39 (t, J = 6.8 Hz, 1 H, 1Ј-H),
3Ј-H), 4.11 (m, 1 H, 4Ј-H), 3.72 (2s, 6 H, OMe), 3.64 (s, 3 H, OMe), 4.85–4.79 (m, 1 H, 3Ј-H), 4.15–4.07 (m, 1 H, 4Ј-H), 3.76 (s, 6 H,
3.4–3.1 (m, 5 H, CH2-Imid, 5Ј-H, 2Ј-H), 2.32 (ddd, J = 13.5, 6.6, OMe), 3.74 (s, 3 H, OMe), 3.73 (s, 3 H, OMe), 3.57–3.49 (m, 1 H,
4.0 Hz, 1 H, 2Ј-H) ppm. 13C NMR (100 MHz, CD2Cl2, 25 °C): δ
= 172.3 (CO2Me), 165.7, 158.9, 153.6 (C-8), 153.3 (C-4), 148.6 (C-
2Ј-H), 3.39–3.27 (m, 4 H, CH2CH2-Imid, 5Ј-H), 3.15–3.00 (m, 2 H,
CH2-Imid), 2.23 (ddd, J = 13.4, 6.8, 4.0 Hz, 1 H, 2Ј-H) ppm. 13C
2), 145.3, 144.2 (C-6), 136.2, 135.5 (C-2 Imid), 134.4 (C-4 Imid), NMR (100 MHz, CD2Cl2, 25 °C): δ = 164.4, 159.0, 158.4, 156.1
132.9, 130.4, 130.3, 129.1, 128.4, 128.2, 128.1, 127.1, 123.2 (C-5), (C-8), 152.3 (C-4), 151.3 (C-2), 148.2 (C-6), 144.7, 142.9, 139.0 (C-
118.6 (C-5 Imid), 113.3, 86.6 (C-4Ј), 86.4 (C-1Ј), 84.7, 72.4 (C-3Ј), 4 Imid), 138.4 (C-2 Imid), 136.0, 135.9, 134.6, 134.1, 132.5, 130.9,
64.2 (C-5Ј), 56.1 (CH-CO2Me), 55.5, 52.8, 38.0 (C-2Ј), 29.4 (CH2-
Imid) ppm. HRMS (ESI+): calcd. for C45H45N8O8+: 825.3360;
found 825.3367.
N6-Benzoyl-5Ј-O-(4,4Ј-dimethoxytrityl)-8-{2-[N-(4,4Ј-dimethoxytri-
tyl)-4-imidazolyl]ethynyl}deoxyadenosine (15): This reaction was
performed with the total exclusion of O2 and H2O. Catalysts were
transferred in a glove box and solvents were degassed by three
freeze–thaw cycles under vacuum. A catalyst suspension was pre-
pared by adding cuprous iodide (92 mg, 0.48 mmol) and tetrakis(-
triphenylphosphane)palladium (280 mg, 0.24 mmol) to DMF
130.0, 129.9, 129.4, 128.8, 128.2, 127.9, 127.7, 126.7, 113.2, 113.0,
86.2, 86.1 (C-4Ј), 84.4 (C-1Ј), 74.4, 73.0 (C-3Ј), 63.9 (C-5Ј), 55.3,
55.2, 36.8 (C-2Ј), 28.1 (CH2CH2-Imid), 26.6 (CH2CH2-Imid) ppm.
HRMS (ESI+): calcd. for C64H59N7O8+: 1054.4503; found
1054.4507.
N6-Benzoyl-8-[(4-imidazolyl)methylamino]-3Ј-O-(methoxyacetyl)de-
oxyadenosine (17): Compound 11 (200 mg, 266 µmol) was dried by
co-evaporation from pyridine under vacuum overnight. After dis-
solving the sample in pyridine (2.0 mL), 75% methoxyacetic anhy-
dride[36] (234 µL, 0.98 mmol) was added. The reaction was stirred
(40 mL). This mixture was added to a solution of 10 (1.77 g, at room temperature for 1 h. The reaction contents were concen-
2.4 mmol) and 1-[1-(4,4Ј-dimethoxytrityl)-1H-4-imidazolyl]ethyne trated to give a gum and then resuspended in acetic acid/H2O
(14) (2.00 g, 5.07 mmol) in DMF (80 mL) and Et3N (20 mL). After (2.0 mL, 80%). After 2.0 h at room temperature, the reaction was
the reaction was stirred at 45 °C for 22 h, the solvent was removed.
The crude product was redissolved in CH2Cl2, washed with aq.
NaHCO3 (5%), and dried with NaSO4 (anhydrous). Purification
by silica gel flash chromatography with Et3N/EtOAc (1:1000) as
eluent afforded 1.98 g (78%) of compound 15 as a yellow-green
foam. Excess starting material [14; 0.78 g (1.12 mmol)] was isolated
on passing through an additional silica column using Et3N/EtOAc/
concentrated and purified on silica gel. The product was eluted in
MeOH/CHCl3 (7:93) to give 106 mg (76%) of a light, pale-yellow
solid. H NMR (400 MHz, CDCl3, 25 °C): δ = 8.40 (s, 1 H, 2-H),
8.02 (d, J = 7.3 Hz, 2 H, 2-H Bz), 7.66 (s, 1 H, 2-H Imid), 7.58 (t,
J = 7.3 Hz, 1 H, 4-H Bz), 7.48 (t, J = 7.3 Hz, 2 H, 3-H Bz), 6.99
(s, 1 H, 5-H Imid), 6.57 (dd, J = 10.1, 5.4 Hz, 1 H, 1Ј-H), 5.50 (d,
J = 6.4 Hz, 1 H, 3Ј-H), 4.51 (d, J = 15.2 Hz, 1 H, 2-H Imid), 4.41
1
hexanes (0.1:60:40) as eluent. 1H NMR (400 MHz, CD2Cl2, 25 °C): (d, J = 15.2 Hz, 1 H, 2-H Imid), 4.15–4.11 (m, 1 H, 4Ј-H), 4.06 (s,
δ = 8.94 (s, 1 H, 17-H), 8.53 (s, 1 H, 2-H), 7.98 (d, J = 7.3 Hz, 2 2 H, OCH2CO2), 3.96 (d, J = 11.6 Hz, 1 H, 5Ј-H), 3.86 (dd, J =
H, 2-H Bz), 7.70–7.12 [m, 15 H, 2-H Phe, 3-H Phe, 4-H Phe, 4-H
11.6, 2.0 Hz, 1 H, 5Ј-H), 3.43 (s, 3 H, OMe), 2.75 (ddd, J = 14.2,
Bz, 3-H Bz, 5-H Imid, 2-H Imid], 7.25 (d, J = 8.9 Hz, 4 H, 2-H 10.1, 6.4 Hz, 1 H, 2Ј-H), 2.26 (dd, J = 14.2, 5.4 Hz, 1 H, 2Ј-H) ppm.
PheOMe), 7.06 (d, J = 9.0 Hz, 4 H, 2-H PheOMe), 6.88 (d, J =
9.0 Hz, 4 H, 3-H PheOMe), 6.75 (d, J = 8.9 Hz, 2 H, 3-H
PheOMe), 6.72 (d, J = 8.9 Hz, 2 H, 3-H PheOMe), 6.64 (dd, J =
13C NMR (100 MHz, CDCl3, 25 °C): δ = 169.9, 166.1, 153.6 (C-
4), 153.0 (C-8), 148.0 (C-2), 143.1 (C-6), (C-2 Imid), 133.4, 132.5,
132.0 (C-4 Imid), 128.5, 127.8, 123.0 (C-5), 119.4 (C-5 Imid), 85.2
Eur. J. Org. Chem. 2008, 4915–4923
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4921

C. Lam, C. Hipolito, D. M. Perrin
FULL PAPER
(C-4Ј), 83.5 (C-1Ј), 76.3 (C-3Ј), 69.4, 61.4 (C-5Ј), 59.2, 37.1 (CH2- temperature, the solvent was removed. The reaction contents were
Imid), 35.3 (C-2Ј) ppm. HRMS (ESI+): calcd. for C24H27N8O6
523.2054; found 523.2043.
:
again resuspended in CH2Cl2, washed with aq. NaHCO3 (5%), and
dried with Na2SO4 (anhydrous). Purification by silica gel
chromatography with MeOH/CHCl3 (5:95) as eluent gave 70.5 mg
+
N6-Benzoyl-8-[3-(4-imidazolyl)propylamino]-3Ј-O-(methoxyacetyl)-
2Ј-deoxyadenosine (18): Compound 18 was prepared in the same
way as 17. Compound 12 (150 mg, 192 µmol) was converted into
1
(57%) of a yellow solid. H NMR (300 MHz, CDCl3, 25 °C): δ =
8.64 (s, 1 H, 2-H), 7.96 (d, J = 7.4 Hz, 2 H, 2-H Bz), 7.59 (s, 1 H,
2-H Imid), 7.52 (t, J = 7.4 Hz, 1 H, 4-H Bz), 7.46 (s, 1 H, 5-H
Imid), 7.42 (t, J = 7.4 Hz, 2 H, 3-H Bz), 6.60 (dd, J = 9.2, 5.6 Hz,
1 H, 1Ј-H), 5.59 (d, J = 6.5 Hz, 1 H, 3Ј-H), 4.23–4.19 (m, 1 H, 4Ј-
H), 4.05 (s, 2 H, OCH2CO2), 3.90 (d, J = 12.5 Hz, 1 H, 5Ј-H), 3.83
(dd, J = 12.5, 1.6 Hz, 1 H, 5Ј-H), 3.38 (s, 3 H, OMe), 3.14 (ddd, J
= 14.1, 9.2, 6.5 Hz, 1 H, 2Ј-H), 2.46 (dd, J = 14.1, 5.6 Hz, 1 H, 2Ј-
H) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ = 169.9, 165.4,
152.3 (C-2), 150.3 (C-4), 149.8 (C-6), 136.9 (C-8), 133.2, 132.8,
128.6, 128.0, 123.4 (C-5), 86.9 (C-1Ј), (C-4Ј), 76.2 (C-3Ј), 69.5, 62.7
(C-5Ј), 59.2, 36.8 (C-2Ј) ppm. HRMS (ESI+): calcd. for
C25H23N7O6+: 540.1608; found 540.1611.
1
103.3 mg (98%) of a light, pale-yellow solid. H NMR (400 MHz,
CDCl3, 25 °C): δ = 8.47 (s, 1 H, 2-H), 8.02 (d, J = 7.2 Hz, 2 H, 2-
H Bz), 7.66 (s, 1 H, 2-H Imid), 7.53 (t, J = 7.3 Hz, 1 H, 4-H Bz),
7.45 (dd, J = 7.3, 7.2 Hz, 2 H, 3-H Bz), 6.77 (s, 1 H, 5-H Imid),
6.63 (dd, J = 10.3, 5.3 Hz, 1 H, 5Ј-H), 5.54 (d, J = 6.5 Hz, 1 H, 3Ј-
H), 4.20–4.16 (m, 1 H, 4Ј-H), 4.07 (s, 2 H, OCH2CO2), 3.99 (d, J
= 10.9 Hz, 1 H, 5Ј-H), 3.90 (dd, J = 10.9, 1.7 Hz, 1 H, 5Ј-H), 3.55–
3.35 (m, 2 H, NHCH2), 3.44 (s, 3 H, OMe), 2.77 (ddd, J = 14.0,
10.3, 6.5 Hz, 1 H, 2Ј-H), 2.70–2.60 (m, 2 H, H2-Imid), 2.26 (dd, J
= 10.3, 5.3 Hz, 1 H, 2Ј-H), 2.12–2.00 (m, 1 H, CH2CH2CH2), 1.90–
1.80 (m, 1 H, 12-H) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ
= 169.9, 165.9, 153.0 (C-8), 152.8 (C-4), 148.1 (C-2), 143.3 (C-6), N6-Benzoyl-8-[(4-imidazolyl)ethyl]-3Ј-O-(methoxyacetyl)deoxyadeno-
133.9, 132.3, 128.5, 127.8, 121.7 (C-5), 114.7 (C-5 Imid), 85.2 (C- sine (21): Compound 21 was prepared in the same way as 17. Com-
4Ј), 83.4 (C-1Ј) 76.5 (C-3Ј), 69.5, 61.3 (C-5Ј), 59.3, 41.6 (CH2NH), pound 16 (150 mg, 0.14 mmol) was converted into 45.4 mg (61%) of
35.2 (C-2Ј), 28.4 (CH2CH2CH2), 22.6 (CH2-Imid) ppm. HRMS a light orange solid that was eluted from a column of silica gel with
(ESI+): calcd. for C26H31N8O6+: 551.2367; found 551.2362.
MeOH/CHCl3 (6:94) as eluent. 1H NMR (300 MHz, CDCl3, 25 °C):
δ = 8.64 (s, 1 H, 2-H), 8.09 (d, J = 7.2 Hz, 2 H, 2-H Bz), 7.72 (s, 1
H, 2-H Imid), 7.65 (t, J = 7.3 Hz, 1 H, 4-H Bz), 7.56 (dd, J = 7.3,
7.2 Hz, 2 H, 3-H Bz), 6.89 (s, 1 H, 5-H Imid), 6.39 (dd, J = 8.6,
6.1 Hz, 1 H, 1Ј-H), 5.63 (d, J = 6.4 Hz, 1 H, 3Ј-H), 4.20–4.15 (m, 1
H, 4Ј-H), 4.17 (s, 2 H, OCH2CO2), 3.89 (dd, J = 12.3, 3.1 Hz, 1 H,
5Ј-H), 3.82 (dd, J = 12.3, 3.7 Hz, 1 H, 5Ј-H), 3.46 (s, 3 H, OMe),
3.45–3.30 (m, 3 H, CH2CH2-Imid, 2Ј-H), 3.19 (t, J = 6.7 Hz, 2 H,
CH2-Imid), 2.37 (ddd, J = 14.0, 6.1, 1.4 Hz, 1 H, 2Ј-H) ppm. 13C
NMR (75 MHz, CDCl3, 25 °C): δ = 169.5, 165.9, 154.2 (C-4), 152.7
(C-8), 150.8 (C-2), 148.7 (C-6), 132.9, 128.6, 127.9, 123.0 (C-5), 86.6
(C-1Ј), 85.4 (C-4Ј), 76.2 (C-3Ј), 69.4, 62.5 (C-5Ј), 59.2, 36.4 (C-2Ј),
29.4 (CH2CH2-Imid), 28.1 (CH2-Imid) ppm. HRMS (ESI+): calcd.
for C25H27N7O6+: 522.2101; found 522.2103.
N6-Benzoyl-3Ј-O-(methoxyacetyl)-8-[(S)-1-methoxycarbonyl-2-(4-
imidazolyl)ethylamino]deoxyadenosine (19): Compound 13 (36 mg,
43 µmol) was dried by co-evaporation from pyridine under vacuum
overnight. After dissolving the sample in pyridine (0.6 mL), 90%
methoxyacetic anhydride (30.8 µL, 174 µmol) was added. The reac-
tion was stirred at room temperature for 100 min. The reaction was
concentrated to a gum and redissolved in CH2Cl2. The organic
solution was washed with aq. NaHCO3 (5%) and dried with so-
dium sulfate (anhydrous). Following removal of the solvent, the
crude product was resuspended in acetic acid/H2O (0.6 mL, 80%).
After 2.25 h at room temperature, the reaction was concentrated
and purified on silica gel. The product was eluted in MeOH/CHCl3
(8:92) to give 20.6 mg (80 %) of a white powder. 1H NMR
(400 MHz, CDCl3, 25 °C): δ = 8.74 (s, 1 H, NHCO), 8.47 (s, 1 H,
2-H), 7.88 (d, J = 7.6 Hz, 2 H, 2-H Bz), 7.54 (t, J = 7.4 Hz, 1 H,
4-H Bz), 7.50 (s, 1 H, 2-H Imid), 7.44 (dd, J = 7.6, 7.4 Hz, 2 H, 3-
H Bz), 6.88 (s, 1 H, 5-H Imid), 6.60 (dd, J = 10.0, 5.4 Hz, 1 H, 1Ј-
H), 5.63 (d, J = 6.4 Hz, 1 H, 3Ј-H), 4.85–4.75 (m, 1 H, NHCH-
COO), 4.12 (s, 1 H, 4Ј-H), 4.09 (s, 2 H, OCH2CO2), 4.02 (d, J =
8-[2-(4-Imidazolyl)ethynyl]deoxyadenosine Triphosphate (5): Com-
pound 20 (26 mg, 50 µmol) was dried by co-evaporation from pyr-
idine under vacuum overnight in a 5 mm diameter NMR tube. The
reaction was monitored by 31P NMR spectroscopy. The starting ma-
terial was suspended in dioxane (0.3 mL) and pyridine (0.1 mL). Fol-
lowing the addition of a solution of 2-chloro-4H-1,3,2-benzodiox-
12.1 Hz, 1 H, 5Ј-H), 3.98 (d, J = 12.1 Hz, 1 H, 5Ј-H), 3.70 (s, 3 H, aphosphorin-4-one (11.1 mg, 55 µmol) in dioxane (55 µL), the reac-
CO2Me), 3.47 (s, 3 H, CH2OCH3), 3.27 (d, J = 13.3 Hz, 1 H, CH2-
Imid), 3.15–3.00 (m, 2 H, CH2-Imid, 2Ј-H), 2.30 (dd, J = 14.2,
5.4 Hz, 1 H, 2Ј-H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
172.4 (CO2Me), 169.9, 165.3, 152.7 (C-4), 152.5 (C-8), 148.4 (C-2),
tion was vortexed thoroughly and then periodically over 15 min. Bi-
s(tri-n-butylammonium) pyrophosphate (35.6 mg, 75 µL) in DMF
(150 µL) and tri-n-butylamine (50 µL) were mixed thoroughly and
added to the reaction. Following another 15 min of vortexing, I2
143.8 (C-6), 133.9, 132.3, 128.6, 127.6, 122.0 (C-5), 115.0 (C-5 (20 mg, 79 µmol) in H2O (20 µL) and pyridine (980 µL) was added.
Imid), 85.4 (C-1Ј), 83.2 (C-4Ј), 76.0 (C-3Ј), 69.7, 60.6 (C-5Ј), 59.5, After 20 min of periodic vortexing, the reaction contents were trans-
57.6 (CHCO2Me), 52.5, 35.3 (C-2Ј), 29.7 (CH2-Imid) ppm. HRMS ferred to a 25 mL flask. Aq. NaHSO3 (5%) was used to quench
(ESI+): calcd. for C27H31N8O8+: 595.2264; found 595.2265.
excess I2. Once the reaction had been evaporated to dryness, it was
left to stand at room temperature in H2O (5 mL) for 30 min. Then
concentrated NH3 (10 mL) was added. The flask was stoppered and
while stirring was heated at 50 °C for 1 h. Following evaporation of
the reaction, the contents were purified by preparative TLC using
dioxane/H2O/NH4OH (6:4:1) as eluent. The appropriate bands were
eluted with H2O and further purified by HPLC. Triphosphate 5
eluted at 8.0 min in program A. Yield 23%. UV: λmax = 312 nm; λmin
= 263 nm. MS (MALDI+): m/z = 582.1. 31P NMR (121.5 MHz,
D2O, 25 °C): δ = –9.9 (m, 1 P, Pγ), –10.6 (m, 1 P, Pα), –22.3 (m, 1 P,
Pβ) ppm.
N6-Benzoyl-8-[2-(4-imidazolyl)ethynyl]-3Ј-O-(methoxyacetyl)deoxy-
adenosine (20): Compound 15 (250 mg, 0.240 mmol) was dried by
co-evaporating with pyridine overnight. The nucleoside was dis-
solved in pyridine (2.5 mL) and 90% methoxyacetic anhydride
(150 µL, 625 mmol) was added to this solution. After 45 min of
stirring at room temperature, the solvent was removed. The re-
sulting gum was resuspended in CH2Cl2, washed with aq. NaHCO3
(5%), and dried with Na2SO4 (anhydrous). Following the removal
of the solvent, pyridine (1 mL) and H2O (1 mL) were added to the
crude product, and the mixture was left under vacuum until dry.
The crude product was then dissolved in CH2Cl2 (5.4 mL), acetic
acid (1.8 mL), and H2O (0.2 mL). After 16 h of stirring at room
8-[(4-Imidazolyl)methylamino]deoxyadenosine Triphosphate (2): Com-
pound 2 was synthesized in the same way as 5 and eluted at 5.4 min
4922
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 4915–4923

Enzymatic Incorporation of Deoxyadenosine Triphosphates
[3] C. Tuerk, L. Gold, Science 1990, 249, 505.
in program A. Yield 8.6%. UV: λmax = 276 nm; λmin = 238 nm. MS
(MALDI+): m/z = 587.2. 31P NMR (121.5 MHz, D2O, 25 °C): δ =
–8.3 (m, 1 P, Pγ), –10.6 (m, 1 P, Pα), –21.3 (m, 1 P, Pβ) ppm.
[4] S. W. Santoro, G. F. Joyce, Proc. Natl. Acad. Sci. USA 1997, 94,
4262.
[5] J. W. Liu, A. K. Brown, X. L. Meng, D. M. Cropek, J. D. Istok,
D. B. Watson, Y. Lu, Proc. Natl. Acad. Sci. USA 2007, 104, 2056.
[6] B. L. Zhang, T. R. Cech, Nature 1997, 390, 96.
[7] P. A. Lohse, J. W. Szostak, Nature 1996, 381, 442.
[8] D. S. Wilson, J. W. Szostak, Annu. Rev. Biochem. 1999, 68, 611.
[9] M. Famulok, J. S. Hartig, G. Mayer, Chem. Rev. 2007, 107, 3715.
[10] G. Giller, T. Tasara, B. Angerer, K. Muhlegger, M. Amacker, H.
Winter, Nucleic Acids Res. 2003, 31, 2630.
[11] N. Lin, J. Yan, Z. Huang, C. Altier, M. Y. Li, N. Carrasco, M.
Suyemoto, L. Johnston, S. M. Wang, Q. Wang, H. Fang, J. Ca-
ton-Williams, B. H. Wang, Nucleic Acids Res. 2007, 35, 1222.
[12] D. M. Perrin, T. Garestier, C. Helene, J. Am. Chem. Soc. 2001,
123, 1556.
[13] M. Kuwahara, K. Hanawa, K. Ohsawa, R. Kitagata, H. Ozaki,
H. Sawai, Bioorg. Med. Chem. 2006, 14, 2518.
[14] M. Kuwahara, J. Nagashima, M. Hasegawa, T. Tamura, R. Kita-
gata, K. Hanawa, S. Hososhima, T. Kasamatsu, H. Ozaki, H.
Sawai, Nucleic Acids Res. 2006, 34, 5383.
[15] T. Gourlain, A. Sidorov, N. Mignet, S. J. Thorpe, S. E. Lee, J. A.
Grasby, D. M. Williams, Nucleic Acids Res. 2001, 29, 1898.
[16] K. Sakthivel, C. F. Barbas, Angew. Chem. Int. Ed. 1998, 37, 2872.
[17] R. T. Raines, Chem. Rev. 1998, 98, 1045.
[18] C. R. Geyer, D. Sen, Chem. Biol. 1997, 4, 579.
[19] M. A. Carrigan, A. Ricardo, D. N. Ang, S. A. Benner, Biochemis-
try 2004, 43, 11446.
[20] D. Faulhammer, M. Famulok, J. Mol. Biol. 1997, 269, 188.
[21] E. Cubero, A. Avino, B. G. de la Torre, M. Frieden, R. Eritja,
F. J. Luque, C. Gonzalez, M. Orozco, J. Am. Chem. Soc. 2002,
124, 3133.
[22] P. Capek, H. Cahova, R. Pohl, M. Hocek, C. Gloeckner, A.
Marx, Chem. Eur. J. 2007, 13, 6196.
[23] S. E. Lee, A. Sidorov, T. Gourlain, N. Mignet, S. J. Thorpe, J. A.
Brazier, M. J. Dickman, D. P. Hornby, J. A. Grasby, D. M. Wil-
liams, Nucleic Acids Res. 2001, 29, 1565.
[24] S. Jager, G. Rasched, H. Kornreich-Leshem, M. Engeser, O.
Thum, M. Famulok, J. Am. Chem. Soc. 2005, 127, 15071.
[25] M. Ikehara, M. Kaneko, Tetrahedron 1970, 26, 4251.
[26] D. Singh, V. Kumar, K. N. Ganesh, Nucleic Acids Res. 1990, 18,
3339.
8-[3-(4-Imidazolyl)propylamino]deoxyadenosine Triphosphate (3):
Compound 3 was synthesized in the same way as 5 and eluted at
7.1 min in program A. Yield 3.3%. UV: λmax = 279 nm; λmin
=
239 nm. MS (MALDI+): m/z = 615.1.
8-[(S)-1-Carboxy-2-(4-imidazolyl)ethylamino]deoxyadenosine Triphos-
phate (4): Compound 4 was synthesized in the same way as 5 except
the deprotection was performed differently. After the reaction was
left to stand in H2O (5 mL) for 30 min, aq. Li2CO3 (0.2 , 10 mL)
was added. The reaction was stirred at 50 °C for 5.5 h and at room
temperature overnight. The contents were once again dried. H2O
(2.5 mL) and concentrated aqueous ammonia (10 mL) were then
used for deprotection. This was carried out at 50 °C for 3 h. Com-
pound 4 eluted at 4.1 min in program B. Yield 2.6%. UV: λmax
=
277 nm; λmin = 239 nm. MS (MALDI+): m/z = 644.9. 31P NMR
(121.5 MHz, D2O, 25 °C): δ = –8.7 (m, 1 P, Pγ), –10.2 (m, 1 P, Pα),
–21.6 (m, 1 P, Pβ) ppm.
8-[2-(4-Imidazolyl)ethyl]deoxyadenosine Triphosphate (6): Compound
6 was synthesized in the same way as 5 and eluted at 5.6 min in
program A. Yield 8.3%. UV: λmax = 262 nm; λmin = 236 nm. MS
(MALDI+): m/z = 586.2.
DNA Templates: (5Ј to 3Ј) P1 ATTAGCCCTTCCAGTCCCCCC-
TTTTCTTTT. T1 GGAGCTGTAGATCTTAGTTACTGGCC-
AAAAGAAAAGGGGGGACTGGAAGGGCTAA. T2 GGAGCTGT-
CCATCTTAGTTACTTTTTAAAAGAAAAGGGGGGACTGGAAG-
GGCTAA. T3 GCAGCTGTAGATCTTAGCCAGGCCTTAAAA-
GAAAAGGGGGGACTGGAAGGGCTAA. T4 GGAGCTGTCCAT-
CACATTTTTTTTTTAAAAGAAAAGGGGGGACTGGAAGGGCT-
AA.
Primer Extension Protocol: Typically, a 32P-labelled primer (P1) was
annealed with the appropriate template in the presence of enzyme
buffers. Single-stranded binding protein (SSB) (0.2 µL/reaction), pyr-
ophosphatase (0.3 µL/reaction), and dithiothreitol (DTT) (0.5–1 µL/
reaction) were added to the annealed oligomers as a cocktail. SSB
and pyrophosphatase were omitted at elevated incubation tempera-
tures. The enzymes were added last; 1–6 units of enzyme were used.
Reactions were prepared with the primer (ca. 1 pmol) in a final incor-
poration volume of 10 µL. Following incubation, formamide/aq.
EDTA solution (30 µL) was added to each incorporation assay and
denaturing PAGE was carried out.
[27] J. G. Phillips, C. E. Tedford, N. C. Chaturvedi, S. M. Ali, US
Patent 6166060, 2000; Chem. Abstr. 2001, 134, 56668.
[28] M. T. Tierney, M. W. Grinstaff, Org. Lett. 2000, 2, 3413.
[29]
[30]
J. Ludwig, F. Eckstein, J. Org. Chem. 1989, 54, 631.
UV irradiation from the photodiode array of the HPLC resulted
in the formation of a small amount of byproduct that interfered
with enzymatic incorporation. The byproduct was a much better
substrate for polymerases and thus incorporation of 4 could not
be observed.
Supporting Information (see also the footnote on the first page of this
article): Synthesis of 4-(2,2-dibromovinyl)-1-(4,4Ј-dimethoxytrityl)-
imidazole, 1H and 13C NMR spectra of the 3Ј-protected triphosphate
precursors, and HPLC traces for the synthesized triphosphates.
[31]
[32]
[33]
F. Boudsocq, S. Iwai, F. Hanaoka, R. Woodgate, Nucleic Acids
Res. 2001, 29, 4607.
A. Roth, R. R. Breaker, Proc. Natl. Acad. Sci. USA 1998, 95,
6027.
R. A. Turner, C. F. Huebner, C. R. Scholz, J. Am. Chem. Soc.
1949, 71, 2801.
Acknowledgments
This work was supported by NSERC Operating Funds. We wish to
thank Dr. Richard Ting for expert synthetic advice in preparing these
derivatives.
[34]
[35]
[36]
R. K. Griffith, R. A. Dipietro, Synth. Commun. 1986, 16, 1761.
B. M. Adger, J. Surtees, Synth. Commun. 1987, 17, 223.
B. V. L. Potter, F. Eckstein, B. Uznanski, Nucleic Acids Res. 1983,
11, 7087.
[1] A. D. Ellington, J. W. Szostak, Nature 1990, 346, 818.
[2] D. L. Robertson, G. F. Joyce, Nature 1990, 344, 467.
Received: April 16, 2008
Published Online: September 5, 2008
Eur. J. Org. Chem. 2008, 4915–4923
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4923