Site-specific synthesis and properties of oligonucleotides containing C8-deoxyguanosine adducts of the dietary mutagen IQ

Article abstract of DOI:10.1021/ja0487022

The site-specific synthesis of oligonucleotides containing the C8-deoxyguanosine adduct of the highly mutagenic heterocyclic amine 2-amino-3-methylimidazo[4,5-f]quinoline (IQ) has been achieved, and the oligonucleotides were characterized by UV melting te

Full text of DOI:10.1021/ja0487022

Published on Web 08/21/2004
Site-Specific Synthesis and Properties of Oligonucleotides
Containing C8-Deoxyguanosine Adducts of the
Dietary Mutagen IQ
C. Eric Elmquist, James S. Stover, Zhiwei Wang, and Carmelo J. Rizzo*
Contribution from the Department of Chemistry and Center in Molecular Toxicology,
Vanderbilt UniVersity, VU Station B 351822, NashVille, Tennessee 37235-1822
Received March 6, 2004; E-mail: c.j.rizzo@vanderbilt.edu
Abstract: The site-specific synthesis of oligonucleotides containing the C8-deoxyguanosine adduct of the
highly mutagenic heterocyclic amine 2-amino-3-methylimidazo[4,5-f]quinoline (IQ) has been achieved, and
the oligonucleotides were characterized by UV melting temperature analysis, circular dichroism, and UV
absorption spectroscopy. Examination of these data indicated that the IQ-adduct is accommodated in
dramatically different environments. This sequence-dependent conformational preference is likely to play
a key role in the mutagenicity and repair of IQ-modified oligonucleotides.
Introduction
Exposure to carcinogens can occur from environmental or
work conditions, diet, smoking, and endogenous processes. In
the mid-1970s, it was discovered that the processes of cooking
beef, pork, fish, and poultry produced a variety of mutagenic
compounds in the parts per billion to parts per million
quantities.1-7 Since then, over 20 such mutagens have been
structurally characterized; these compounds are collectively
known as the heterocyclic amines (HCAs), and representative
structures (1-11) are shown in Figure 1. The systematic names
of these compounds are cumbersome, and they are more
commonly referred to by their simpler acronyms. The distribu-
tion and quantity formed is dependent upon the type of meat,
cooking method, temperature, and time. In general, 2-amino-
1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP, 8) is the most
abundantly produced, while 2-amino-3,4-dimethylimidazo[4,5-
f]quinoline (MeIQ, 2) is the most mutagenic. PhIP and 2-amino-
3-methylimidazo[4,5-f]quinoline (IQ, 1) have also been shown
to be a constituent of cigarette smoke.8
Figure 1. Structures of representative heterocyclic amines (HCAs) found
in cooked meats.
The HCAs are highly mutagenic in Ames Salmonella
reversion assays. A listing of their mutagenicities in an Ames
testers strain designed for frameshift mutations is shown in Table
1.3 The most mutagenic of these compounds, MeIQ and IQ,
are on the order of 24 and 20 times more mutagenic than
aflatoxin B1 in these assays. In bacteria, the HCAs are potent
inducers of frameshift mutations in CG repeat sequences;
however, in mammalian cells, point mutations are largely
observed.1,2,4
The HCAs, as with other arylamines, require metabolic
activation to covalently modify DNA (Scheme 1). The initial
step is the cytochrome P450 oxidation of the arylamine to the
corresponding N-hydroxylamine.9 For IQ, P450 1A2 has been
identified as the major metabolizing enzyme in humans. The
N-hydroxylamine is then esterified to an acetyl by N-acetyl
transferase enzyme (NAT) or related derivative. Solvolysis of
the N-acetyl group generates a highly reactive arylnitrenium ion,
which is the DNA modifying species. The aryl nitrenium ion
(1) Turesky, R. J. Drug Metab. ReV. 2002, 34, 625-650.
(2) Schut, H. A.; Snyderwine, E. G. Carcinogenesis 1999, 20, 353-368.
(3) Sugimura, T. Mutat. Res. 1997, 376, 211-219.
(4) Felton, J. S.; Knize, M. G.; Dolbeare, F. A.; Wu, R. W. EnViron. Health
Perspect. 1994, 102 (suppl. 6), 201-204.
(5) Felton, J. S.; Knize, M. G. In Handbook of Experimental Pharmacology;
Cooper, C. S., Grover, P. L., Eds.; Springer-Verlag: New York, 1990;
Vol. 91/I, pp 471-502.
(6) Jagerstad, M.; Skog, K.; Arvidsson, P.; Solyakov, A. Z. Lebensm.-Unters.-
Forsch. A 1998, 207, 419-427.
(7) Skog, K. I.; Johansson, M. A.; Jagerstad, M. I. Food Chem. Toxicol. 1998,
36, 879-896.
(8) Hecht, S. S. EnViron. Mol. Mutagen. 2002, 39, 119-126.
(9) Guengerich, F. P. Drug Metab. ReV. 2002, 34, 607-623.
9
10.1021/ja0487022 CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 11189-11201
11189

A R T I C L E S
Elmquist et al.
Table 1. Relative Mutagenicity of HCA in S. typhimurium Tester
group, alternative protecting group chemistry was developed
for its solid-phase DNA synthesis. Although structurally very
similar, the biological response to these lesions was dramatically
different. In bacterial mutagenesis assays, AAF gave predomi-
nantly -1 and -2 frameshift mutations, while base-pair
substitution were largely observed for AF.1,2
Strains TA 104a
2 (MeIQ)
1 (IQ)
5 (4,8-DiMeIQx)
6 (7,8-DiMeIQx)
4 (MeIQx)
7
1000
656
277
247
220
150
114
42
3 (IQx)
Results and Discussion
aflatoxin B1
11 (Cre-P-1)
8 (PhIP)
29
3
Synthesis of the C8-Deoxyguanosine Adducts of IQ and
Simple Arylamine. The Buchwald-Hartwig16-20 palladium-
catalyzed N-arylation reaction has been used recently for the
preparation of nucleoside-carcinogen adducts by Lakshman and
later Johnson, for the preparation of N6-2′-deoxyadenosine
derivatives.21,22 Hopkins and Sigurdsson, and Johnson synthe-
sized N2-deoxyguanosine-N2-deoxyguanosine and N2-deoxygua-
nosine-N6-deoxyadenosine nitrous acid cross-links as well as
other N2-aryl derivatives of deoxyguanosine via the Buchwald-
Hartwig cross-coupling reaction.23-27 It is worth noting that
Johnson’s N-arylation approach involved coupling of the
exocyclic amino group of deoxyadenosine or deoxyguanosine
with bromoarenes, while Lakshman and Hopkins and Sigurdsson
employed the corresponding bromopurine with arylamines. This
reaction has become an important method for the preparation
of deoxyadenosine and deoxyguanosine derivatives that are
modified at the exocyclic amino groups.28-31 At the time of
our initial report, we were the first to employ this strategy for
the synthesis of C8-deoxyguanosine adducts of arylamines.32
Since then, the Buchwald-Hartwig reaction has been utilized
by Meier and Scharer for the synthesis of C8-deoxyguanosine
adducts of simple arylamines and by Schoffers for the synthesis
of C8-arylamines adducts of adenosine.33-35 The coupling of a
suitably protected 8-amino-2′-deoxyguanosine derivative with
bromoarene has also been demonstrated as a feasible strategy.36
Buchwald and others have reported the N-arylation of
amides.37 Thus, conventional amide protecting groups for N2
of deoxyguanosine were judged as unsatisfactory. We found
that the tetramethyldisilylazacyclopentane (STABASE) group
developed by Magnus was a compatible N2-protecting group
10 (7,9-DiMeIgQx)
benzo[a]pyrene
9
1
1
0.003
a Reference 3.
Scheme 1
reacts with DNA to give the C8-deoxyguanosine adducts (14)
as the major product, and, in some cases, modification of the
N2-position of deoxyguanosine (15) has been unambiguously
identified as minor products.10,11 Some simple arylamines also
give C8-deoxyadenosine adducts; however, most of the evidence
indicates that deoxyadenosine adducts do not play a significant
role in HCA mutagenicity, and no HCA adducts of deoxyad-
enosine have been characterized.2,12
Perhaps the most extensively studied of the arylamine adducts
are 2-aminofluorene (AF) and N-acetyl-2-aminofluorene (AAF).13
These compounds were of commercial interest as pesticides,
but were found to be highly mutagenic and never used as
intended. Zhou and Romano reported the synthesis of phos-
phoramidite reagents of C8-deoxyguanosine adducts of 2-ami-
nofluorene and its N-acetyl analogue, thereby allowing for these
lesions to be site-specifically incorporated into oligonucleo-
tiodes.14,15 Because of the base-sensitive nature of the N-acetyl
(15) Zhou, Y.; Chladek, S.; Romano, L. J. J. Org. Chem. 1994, 59, 556-563.
(16) Yang, B. H.; Buchwald, S. L. J. Organomet. Chem. 1999, 576, 125-146.
(17) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002, 219, 131-209.
(18) Wolfe, J. P.; Wagaw, S.; Marcoux, J.-F.; Buchwald, S. L. Acc. Chem. Res.
1998, 31, 805-818.
(19) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37, 2046-2067.
(20) Hartwig, J. F. Acc. Chem. Res. 1998, 31, 852-860.
(21) Lakshman, M. K.; Keeler, J. C.; Hilmer, J. H.; Martin, J. Q. J. Am. Chem.
Soc. 1999, 121, 6090-6091.
(22) De Riccardis, F.; Bonala, R. R.; Johnson, F. J. Am. Chem. Soc. 1999, 121,
10453-10460.
(23) Harwood, E. A.; Sigurdsson, S. T.; Edfeldt, N. B. F.; Reid, B. R.; Hopkins,
P. B. J. Am. Chem. Soc. 1999, 121, 5081-5082.
(24) Harwood, E. A.; Hopkins, P. B.; Sigurdsson, S. T. J. Org. Chem. 2000,
65, 2959-2964.
(25) De Riccardis, F.; Johnson, F. Org. Lett. 2000, 2, 293-295.
(26) Bonala, R. R.; Yu, P. L.; Johnson, F. Tetrahedron Lett. 1999, 40, 597-
598.
(27) Bonala, R. R.; Shishkina, I. G.; Johnson, F. Tetrahedron Lett. 2000, 41,
7281-7284.
(28) Lakshman, M. K. J. Organomet. Chem. 2002, 653, 234-251.
(29) Johnson, F.; Bonala, R.; Tawde, D.; Torres, M. C.; Iden, C. R. Chem. Res.
Toxicol. 2002, 15, 1489-1494.
(30) Lakshman, M. K.; Gunda, P. Org. Lett. 2003, 5, 39-42.
(31) Chakraborti, D.; Colis, L.; Schneider, R.; Basu, A. K. Org. Lett. 2003, 5,
2861-2864.
(10) Turesky, R. J.; Rossi, S. C.; Welti, D. H.; Lay, J. O.; Kadlubar, F. F. Chem.
Res. Toxicol. 1992, 5, 479-490.
(11) Turesky, R. J.; Markovic, J. Chem. Res. Toxicol. 1994, 7, 752-761.
(12) Beland, F. A.; Kadlubar, F. F. EnViron. Health Perspect. 1985, 62, 19-
30.
(13) Heflich, R. H.; Neft, R. E. Mutat. Res. 1994, 318, 73-174.
(14) Zhou, Y.; Romano, L. J. Biochemistry 1993, 32, 14043-14052.
(32) Wang, Z.; Rizzo, C. J. Org. Lett. 2001, 3, 565-568.
(33) Meier, C.; Grasl, S. Synlett 2002, 802-804.
(34) Gillet, L. C. J.; Scharer, O. D. Org. Lett. 2002, 4, 4205-4208.
(35) Schoffers, E.; Olsen, P. D.; Means, J. C. Org. Lett. 2001, 3, 4221-4223.
(36) Takamura-Enya, T.; Ishikawa, S.; Mochizuki, M.; Wakabayashi, K.
Tetrahedron Lett. 2003, 44, 5969-5973.
(37) Yin, J.; Buchwald, S. L. Org. Lett. 2000, 2, 1101-1104.
9
11190 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
Scheme 2
Scheme 4
Scheme 3
that our initial assumption that the N2-amino group required
special protection to affect the Buchwald-Hartwig reaction was
incorrect. In light of these reports, we reexamined our synthetic
scheme. For the synthesis of the C8-IQ-adducted nucleoside,
the silyl protecting groups for the N2 and ribose hydroxyls could
be removed simultaneously with fluoride ion. However, for the
preparation of a phosphoramidite reagent of the C8-modified
nucleoside, it would be advantageous to have the N2-amino
group differentially protected to avoid a deprotection-repro-
tection sequence.
We attempted the Buchwald-Hartwig reaction of N2-
unprotected 8-bromo-2′-deoxyguanosine derivative 18 under
conditions we previously worked out (Scheme 4). The coupling
proceeded smoothly to give 21 in comparable yield to that
reported for the fully deprotected series. However, attempts to
protect the N2-position of 21 or the O6-deprotected analogue
22 as the corresponding acetyl or phenoxyacetyl derivative
proceeded in poor yield in our hands. Previous studies of the
conformation of the C8-IQ-adducted nucleoside (14) indicate
that the syn conformation about the glycosidic bond is preferred
for the N-arylation reaction.38 The desired substrate for the
N-arylation reaction (19) was prepared from 8-bromo-2′-
deoxyguanosine in three steps according to Scheme 2. The
optimal conditions for the cross-coupling reaction of 19 with
IQ (1) were achieved using 10 mol % catalyst and lithium
hexamethyldisilazide as the base to give 20 in 68% yield.39,40
Sequential deprotection of 20 gave the C8-IQ adduct of
deoxyguanosine 14.32
After our initial communication on the synthesis of the C8-
IQ adduct of deoxyguanosine nucleoside (14), Schoffers reported
the synthesis of C8-arylamine adducts of adenosine involving
a Buchwald-Hartwig N-arylation reaction (Scheme 3). Of
significance, it was demonstrated that the N6-amino group did
not require protection. Subsequently, Meier reported that N2-
isobutyryl-8-bromo-2′-deoxyguanosine also undergoes Buch-
wald-Hartwig reaction with arylamines. These reports indicated
(38) Djuric, S.; Venit, J.; Magnus, P. Tetrahedron Lett. 1981, 22, 1787-1790.
(39) Louie, J.; Hartwig, J. F.; Fry, A. J. J. Am. Chem. Soc. 1997, 119, 11695-
11696.
(40) Louie, J.; Hartwig, J. F. Tetrahedron Lett. 1995, 36, 3609-3612.
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11191

A R T I C L E S
Elmquist et al.
Scheme 5
Figure 2. Sequences for C8-IQ- and AFF-adducted oligonucleotides.
related protecting group for N2 may impart more desirable
solubility properties. We attempted the Buchwald-Hartwig
N-arylation reaction with IQ (1) and the N2-phenoxyacetyl
protected 8-bromoguanosine prepared by the reaction of phe-
noxyacetyl chloride with 18. However, the phenoxyacetyl group
did not survive the strongly basic conditions used previously
to couple IQ with 18 and 19. During the course of our work,
Gillet and Scharer reported the cross-coupling of 2-aminofluo-
rene with a N2-(4-isopropylphenoxy)acetyl (iPrPAC) protected
8-bromo-2′-deoxyguanosine derivative.34 In this work, strong
bases such sodium tert-butoxide lead to rapid loss of the N2-i-
PrPAC group analogous to our observations. It was also reported
that K2CO3 was effective in promoting the Buchwald-Hartwig
reaction to give the C8-aminofluorene adduct with an N2-iPrPAC
protecting group. Unfortunately, in our hands, these conditions
did not give satisfactory yields when IQ was the amine.
To examine the generality of the Buchwald-Hartwig reaction
of N2-unprotected derivative 18, we examined its reactivity with
simple arylamines. We found that simple arylamines undergo
rapid cross-coupling with 18 in 48-65% yield (Scheme 5),
which is comparable to that obtained for the coupling with IQ.
Two-step deprotection provided the adducted nucleosides 28a-d
in good overall yield (Table 2).
Site-Specific Synthesis of Oligonucleotides Containing a
C8-Deoxyguanosine Adduct of IQ. Phosphoramidite 25 is
readily soluble in methylene chloride, and we proceeded with
the oligonucleotide synthesis using a manual coupling protocol
to incorporate the IQ-modified nucleotide. For our initial
synthesis, we prepare the N-ras-12 sequence (29) shown in
Figure 2. This particular sequence was chosen as an initial target
because there is substantial experience in our department in
preparing N2-deoxyguanosine-PAH-adducted oligonucleotides
using this sequence. For the synthesis of the adducted N-ras-
12 11-mer 29, automated DNA synthesis was used for the first
six unmodified nucleotides (from the 3′-end) using standard
commercially available PAC-protected phosphoramidites. The
cassette containing the solid support was then removed from
the DNA synthesizer, and each end was fitted with a syringe,
one contained a methylene chloride solution of the IQ-adducted
phosphoramidite 25, while the other syringe contained a solution
of 1H-tetrazole. The solutions of 25 and 1H-tetrazole were
periodically drawn through the cassette over 30 min. After a
manual wash with acetonitrile, the cassette was placed back on
the DNA synthesizer and capped with phenoxyacetyl chloride.
The remainder of the synthesis was completed as usual. Johnson
had observed that the C8-aminofluorene adduct undergoes
Table 2. Synthesis of C8-Deoxyguanosine Adducts of Simple
Arylamine
due to the bulk of the C8-substituent.10 This may place the N2-
amino group in an unfavorable steric environment and hinder
its reactivity toward acetylization. We were able to protect this
position as a dimethylformamidine, which was selective for N2
even in the presence of free ribose hydroxyl groups. We found
it more convenient for purification reasons to first remove the
silyl groups of 22 with fluoride to give the adducted nucleoside
14 and subsequently treat with the dimethylformamide dimethyl
acetal to give the N2-formamidine 23, which precipitated from
the reaction in 72% yield. Conversion to the desired phosphora-
midite reagent 25 was accomplished by selective protection of
the primary 5′-hydroxy group to give the dimethoxytrityl ether
followed by phosphitylation of the 3′-hydroxyl group in 72%
yield for two steps. The synthesis of the C8-IQ-adducted
phosphoramidite required eight steps from commercially avail-
able 8-bromo-2′-deoxyguanosine and proceeded in 17% overall
yield. The Buchwald-Hartwig N-arylation reaction of IQ (1)
and N2-unprotected 8-bromo-2′-deoxyguanosine derivative 18
shortens the synthesis of the adducted nucleoside 14 to five steps
and improves the overall yield to 33% from 8-bromo-2′-
deoxyguanosine.
One potential drawback to phosphoramidite 25 is that it is
only sparingly soluble in acetonitrile, the standard solvent used
in automated DNA synthesizers. A phenoxyacetyl (PAC) or
9
11192 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
oxidative rearrangement in the presence of strong base and air,
analogous to that observed for 8-oxo-2′-deoxyguanosine.41,42 It
was found that this reaction could be suppressed by the addition
of mercaptoethanol during the deprotection of the oligonucle-
otide. To prevent complications due to oxidation of the C8-IQ-
adducted deoxyguanosine, the deprotection of oligonucleotide
25 was accomplished with degassed ammonium hydroxide under
an argon atmosphere at room temperature overnight followed
by purification by reversed-phase HPLC. The oligonucleotide
was characterized by MALDI-TOF mass spectrometry and
enzymatic digestion, and the purity was assessed by capillary
zone electrophoresis (CZE) analysis.
We also prepared an oligonucleotide containing the NarI
restriction sequence (G1G2CG3CC) site-specifically adducted
with IQ (30, Figure 2) at the G3-position. The NarI sequence
contains a CG-repeat (shown in bold in 30-32, Figure 2) and
is considered a “hotspot” for arylamine modification and is prone
to -2 frameshift mutations. Site-specific mutagenesis studies
of the C8-AF and C8-AAF adducts have been reported at all
three guanine positions of this sequence, and the NMR structural
43
analyses of the C8-AF adducts have also been disclosed.
As with oligonucleotide 29, a manual coupling protocol was
utilized to incorporate the IQ-modified base of 30, and the
adducted oligonucleotide was purified and analyzed as described
above. Enzyme digests of oligonucleotides 29 and 30 are shown
in Figure 3 along the UV spectra of the adducted base as
obtained by a diode array detector from the HPLC analysis of
the enzymatic digestion. The modified base (14) for the enzyme
digest integrated low, which we believe is due to precipitation
from its low water solubility. HPLC, CZE, and MALDI-TOF
mass spectrometric analysis indicate that the adducted oligo-
nucleotide is greater than 92% pure.
Melting Temperature (Tm), Light Absorption, and Cir-
cular Dichroism Studies. Oligonucleotides 29 and 30 were
hybridized to complementary strands, and the effect of the C8-
IQ adduct on the thermal stability of the DNA duplex was
measured through UV melting temperature analysis (Tm) by
monitoring the absorbance at 260 nm versus temperature. These
data are summarized in Table 3 along with related melting
temperatures for C8-AAF-adducted oligonucleotides at the G3-
position in the NarI sequence, 31 and 30 (Figure 2), previously
reported by Fuchs and Romano, respectively.14,44,45 In general,
modification of DNA destabilizes the duplex, resulting in a
lower Tm. In the case of the ras-12-adducted oligonucleotides
29, we observed a 9 °C decrease in the Tm versus that of the
corresponding unmodified oligonucleotides (entry a). A modest
4 °C destabilization for the IQ-adducted NarI oligonucleotide
30 was observed when compared to the unmodified duplex
(entry b, Tm ) 61 vs 65 °C). These results should be compared
with the independent observations of Fuchs and Romano who
examined oligonucleotides 31 and 32 containing a C8-AAF
adduct at the G3-position of the NarI restriction sequences.14,44
The AAF-adducted oligonucleotides 31 and 32 (Table 3, entries
Figure 3. Enzyme digests of oligonucleotides 29 (panel A) and 30 (panel
B) containing the C8-deoxyguanosine adduct of IQ. Panel C shows the
UV spectrum of the IQ-adducted base (14) from the enzyme digest taken
via diode array detector (solid line) versus that of an authentic standard
(dashed line).
c and d) were significantly more destabilized (∆Tm ) 13 and
17 °C) versus the unadducted duplex than that for the IQ-
adducted oligonucleotide 30 (entry b). We note that the sequence
of 12-mers 31 and 32 flanking the NarI restriction sequence
and the buffer conditions used for the Tm measurements by
Fuchs and Romano were slightly different than ours; however,
in all cases we are comparing the difference in Tm for the
adducted versus the unadducted oligonucleotides. Thus, we
believe the observed thermal destabilizations are due largely to
the modification, and the comparisons are qualitatively useful.
As mentioned, the G3-position of the NarI restriction sequence
is part of a CG-repeat and is a hotspot for -2 frameshift
mutations by arylamines such as AAF and IQ. One interesting
difference is that AAF blocks replication and its mutagenesis
is SOS-dependent, while IQ is by-passed and has been shown
to be SOS-independent.46 The accepted model for frameshift
mutagenesis within a repetitive sequence involves strand slip-
(41) Johnson, F.; Huang, C. Y.; Yu, P. L. EnViron. Health Perspect. 1994, 102
(suppl. 6), 143-149.
(42) Shibutani, S.; Gentles, R.; Johnson, F.; Grollman, A. P. Carcinogenesis
1991, 12, 813-818.
(43) Patel, D. J.; Mao, B.; Gu, Z.; Hingerty, B. E.; Gorin, A.; Basu, A. K.;
Broyde, S. Chem. Res. Toxicol. 1998, 11, 391-407.
(44) Koehl, P.; Valladier, P.; Lefevre, J.-F.; Fuchs, R. P. P. Nucleic Acids Res.
1989, 17, 9531-9541.
(45) Milhe, C.; Fuchs, R. P. P.; Lefevre, J.-F. Eur. J. Biochem. 1996, 235, 120-
127.
(46) Maenhaut-Michel, G.; Janel-Bintz, R.; Samuel, N.; Fuchs, R. P. P. Mol.
Gen. Genet. 1997, 253, 634-641.
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11193

A R T I C L E S
Elmquist et al.
Table 3. Tm Studies of Adducted Oligonucleotides (Temperature
vs f and entries d vs g, respectively). This led Romano to
speculate that the AAF-adducted 12•12 and 12•10 duplexes
containing oligonucleotide 32 had similar structures.14 In the
case of the IQ-adducted oligonucleotide 30, we find that the
Tm of the full length 12•12 duplex is 13 °C higher than the
12•10 duplex with the two-base deletion opposite the adducted
base (Table 3, entries b vs e), in contrast to the results reported
by Fuchs and Romano for the AAF-adducted oligonucleotide.
Fuchs examined the circular dichroism spectra of the AAF-
adducted oligonucleotides at all three deoxyguanosine positions
of the NarI restriction sequence (i.e., 31).44 It was observed that
the long wavelength absorptions from 290 to 305 nm, assigned
to the N-acetylaminofluorene group, had a negative optical
activity. This was interpreted as a local left-handed helical
structure around the adducted nucleotide of 31, which may play
a significant role in the mutagenic mechanism and polymerase
arresting properties of AAF. This local left-handed or Z-DNA
conformation is due to the AAF group inducing an anti to syn
rotation of the glycosidic bond accompanied by the AAF moiety
intercalating into the DNA helix with the adducted guanine base
displaced into the minor groove. This results in a local
denaturation of the DNA near the adducted base and an
increased susceptibility of AAF-adducted oligonucleotide to
nuclease S1, which is a single-strand-specific endonuclease. This
model has been termed base displacement or insertion-
denaturation.49,50 In addition, the positive band at 268 nm for
the unmodified duplex was slightly blue-shifted to 263 nm and
increased in amplitude by 100% upon modification by AAF at
the G3-position.
Was Raised 1 °C min-1
)
a Conditions: 100 mM NaCl, 10 mM pH 7 phosphate, 0.05 mM EDTA,
0.5 A260/mL of each oligonucleotide. b Conditions: 50 mM NaCl, 1 mM
EDTA, 10 mM Tris. c Conditions: 100 mM NaCl, 10 mM NaH2PO4, pH
7, 0.5 mM EDTA. d This work. e Reference 44. f Reference 45. g Reference
14.
The CD spectra of the IQ-adducted oligonucleotides 29 and
30 as well as their corresponding modified and unmodified
duplexes are shown in Figure 5. The duplex of the IQ-adducted
oligonucleotide at the G3-position in the NarI restriction
sequence (30) showed few of the characteristics of the AAF
spectra reported by Fuchs. The induced CD signal assigned to
the IQ group from ∼300-360 is broad, weak, and has a positive
optical activity, indicating that there is no pronounced left-
handed helical structure around the adducted base. The positive
band at 272 nm for the unadducted duplex is slightly red-shifted
upon adduction at the G3-position by IQ to 274, and its intensity
is relatively unchanged. Overall, the spectrum of the adducted
single-strand and duplex oligonucleotides strongly resembles
that of the unmodified. We contrast this spectrum to that of the
12•12 duplex involving the ras-12 sequence (29). The intensity
of the band at 270 nm is shifted to approximately 278 nm and
is significantly decreased upon adduction by IQ at the G3-
position. In the adducted single-stranded oligonucleotides, a
strong negative band is observed at 260 nm. The significant
reduction in the intensity for the duplex at 278 nm versus that
of the unmodified oligonucleotide may be due to specific
conformations of the adducted base responsible for the negative
band observed in the single-stranded adducted oligonucleotide
(29), and this conformation is largely preserved in both single-
and double-stranded DNA. We also note that the induced CD
signals assigned to the IQ moiety in 29 from ∼300 to 360 nm
are well defined and reasonably intense.
Figure 4. Possible structures for the slipped mutagenic intermediate of
C8-arylamine adducts in a CG-repeat sequence.
page during replication.47 In the present case, two bases are
extruded, leading to the -2 frameshift deletion. To simulate
the slipped-mutagenic intermediate for the observed -2 frame-
shifts, Fuchs and Romano independently examined the stability
of 31 and 32 when hybridized to a complementary 10-mer in
which two bases opposite the lesions were deleted (Table 3,
entries f and g), thereby creating a two-base bulge involving
the adducted base.14,45 In the case of the NarI CG-repeat
sequence, this two-base bulge can be accommodated in two
ways when hybridized to a complementary strand as shown in
Figure 4. NMR studies showed that structure 34 existed
predominantly or exclusively in the 12•10 duplex involving 31;
however, the NMR analysis of an aminofluorene-adducted
oligonucleotide in a related NarI sequence reached the opposite
conclusion.43,45,48 Melting temperature studies showed that the
AAF-adducted 12•10 duplex is significantly stabilized when
compared to the unadducted 12•10 hybrid (Table 3, entry f).
When the IQ-adducted oligonucleotide 30 was hybridized with
a complementary 10-mer in which the GC bases opposite the
adduct were deleted, a similar stabilization was observed versus
the unmodified 12•10 duplex (Table 3, entry e). However, the
stabilization of the IQ-adducted 12•10 duplex was less pro-
nounced than that for AAF (∆Tm ) 10 vs 15 °C). Interestingly,
Fuchs and Romano reported that the Tm value of the AAF-
adducted 12•10 duplex was essentially identical to that of the
corresponding AAF-adducted 12•12 duplex (Table 3, entries c
We believe that the dramatic differences in the IQ absorptions
between duplexes of 29 and 30 are an indication that the adduct
(47) Hoffmann, G. R.; Fuchs, R. P. P. Chem. Res. Toxicol. 1997, 10, 347-359.
(48) Mao, B.; Gorin, A.; Gu, Z.; Hingerty, B. E.; Broyde, S.; Patel, D. J.
Biochemistry 1997, 36, 14479-14490.
(49) Santella, R. M.; Grunberger, D. EnViron. Health Perspect. 1983, 49, 107-
115.
(50) Fuchs, R. P. P.; Duane, M. FEBS Lett. 1970, 14, 206-208.
9
11194 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
Figure 6. UV absorption spectra of adducted single-stranded (- - -) and
double-stranded (-) DNA in the ras-12 sequence (panel A) and NarI
restriction sequence (panel B).
Figure 5. CD spectra of unmodified duplex (-), adducted duplex (- - -
), and adducted single-strand (- - -) DNA in the ras-12 sequence (panel A)
and NarI restriction sequence (panel B).
with a base-displaced intercalated conformation where π-π
stacking interactions with neighboring DNA bases can signifi-
cantly perturb the UV absorbance of the IQ moiety. For groove-
bound conformations, the UV absorbance of the adduct is
minimally perturbed by the DNA bases. This observation is
consistent for the C8-IQ adduct in the ras-12 sequence 30 in
which we believe the IQ is accommodated in a groove-bound
conformation. As with previous studies with PAHs, we would
also expect the fluorescence of the base-displaced intercalated
adduct to be quenched and red-shifted, while the groove-bound
conformation would be less perturbed. Unfortunately, the C8-
IQ adduct of deoxyguanosine was not fluorescent.
For N2-deoxyguanosine adducts of PAHs, it was observed
that the intensity of the adduct absorbance was reduced for the
intercalated conformation. In the case of the duplex of 30, the
intensity of the IQ absorbance is barely detectable. We
speculated that this dramatic reduction in the absorbance is due
to a hydrogen-bonding interaction between the quinoline
nitrogen of IQ and the N4-amino group of the cytosine base in
the complementary strand of the duplex (Figure 7). Such a
conformation would result in an unfavorable local distortion of
the DNA duplex near the adduct site; however, this strain could
be offset by more favorable hydrophobic stacking interactions
and an additional hydrogen bond. To test this hypothesis further,
we examined the properties of adducted oligonucleotide 30 when
hybridized to a complementary strand in which a thymidine was
placed opposite the adducted base. In this case, the proposed
hydrogen-bonding interaction would be disrupted or eliminated.
The melting temperature of the thymidine mismatch duplex of
30 was not significantly lower than that with a normal
complement (entries h vs b in Table 3). G-T-mismatches can
is accommodated in dramatically different environments, the
preference of which is highly sequence-dependent. Absorbance
and fluorescence emission spectroscopy has been used to
characterize the environment of the lesion of PAH-adducted
oligonucleotides.51 For N2-PAH adducts of deoxyguanosine, it
was observed that base-displaced intercalated conformations
(termed site I) resulted in a ∼10 nm red shift in the absorbance
and a significantly quenched fluorescence emission. For groove-
bound conformations (site II) where base stacking is expected
to have less influence on the adduct chromophore, there is a
∼2 nm red-shift in the absorbance maximum and only a modest
quenching of the fluorescence emission. This analysis had been
applied to a C8-aminopyrene (AP)-adducted oligonucleotide in
the sequence 5′-d(CTCATGAPATTCC) and it was concluded
that a mixture of groove-bound and base-displaced intercalated
conformers were present.52
Absorption spectra of oligonucleotides 29 and 30 and their
duplexes are shown in Figure 6. Our interpretation of the spectra
is that the IQ-adduct of 30, which is in the CG-repeat portion
of the NarI restriction sequence, adopts a base-displaced
intercalated conformation and is groove-bound in 29. The
intensity of the IQ absorption for 30 is dramatically reduced
when the oligonucleotide is hybridized to its complementary
strand as compared to the absorbance of the adducted single-
stranded oligonucleotide. We believe this behavior is consistent
(51) Geacintov, N. E.; Cosman, M.; Mao, B.; Alfano, A.; Ibanez, V.; Harvey,
R. G. Carcinogenesis 1991, 12, 2099-2108.
(52) Nolan, S. J.; Vyas, R. R.; Hingerty, B. E.; Ellis, S.; Broyde, S.; Shapiro,
R.; Basu, A. K. Carcinogenesis 1996, 17, 133-144.
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11195

A R T I C L E S
Elmquist et al.
Figure 7. Proposed hydrogen bonding of the base-displaced intercalated
conformation of the adducted C8-deoxyguanosine with the complementary
deoxycytidine in the CG repeat of the NarI restriction sequence.
Figure 9. Averaged CD spectra of 29 and 30 with the normal complement
(-) and 30 with a T-mismatch (- - -).
small shifts in the absorbance maximum, although there appears
to be little change from the single stand, which would be more
consistent with a groove-bound adduct.
NMR solution structures of oligonucleotides containing C8-
deoxyguanosine adducts of simple arylamines have shown that
the adducted base may be in conformational exchange between
the groove bound and base-displaced intercalated conformations.
For example, solution structures of AF-adducted oligonucle-
otides at all three positions of the NarI restriction sequence have
been determined by NMR, and the equilibrium between the
groove-bound and base-displaced intercalated conformations was
highly dependent upon the sequences. At the G1 and G2
positions, the groove-bound conformation was strongly favored
in a ratio of 70:30 and 90:10, respectively, while at the G3-
position an equal mixture of the groove-bound and base-
displaced intercalated conformations was reported.43,55 Solution
structures of C8-AAF and aminobiphenyl have been determined
in other sequences and were reported to exist in equilibrium
between the groove-bound and base-displaced intercalated
conformations as well.56,57 A solution structure of C8-aminopy-
rene adducts was found to exist exclusively in a base-displaced
intercalated conformation in a sequence in which the corre-
sponding AF-adduct was found to be largely groove-bound.58
On the basis of their CD and UV spectra, we believe that the
IQ moiety of oligonucleotides 29 and 30 exists predominantly
in the groove-bound and base-displaced intercalated conforma-
tions, respectively. However, the CD spectra of 30 with a
thymidine opposite the adducted base possess features observed
in both the groove-bound and the base-displaced intercalated
spectra, which may indicate that this duplex exists in confor-
mational exchange. The CD spectra of oligonucleotides 29 and
30 (with the normal complement strand) were averaged in an
attempt to model a spectrum of an oligonucleotide in confor-
mational exchange with equal populations of base-displaced
intercalated and groove bound. This analysis assumes that the
spectra for 29 and 30 are predominantly one conformation. The
result of this analysis is shown in Figure 9. The averaged
spectrum (-) closely resembles that of 30 with the T-mismatch
(- - -) and leads us to believe that this duplex is in conformational
exchange.
Figure 8. UV (panel A) and CD (panel B) spectra of the IQ-adducted
NarI restriction sequence (30) hybridized to a complementary strand with
a thymidine opposite the adducted base (-). The UV spectrum is compared
to the adducted single-strand oligonucleotide (- - -), and the CD spectrum
is compared to the duplex with a normal complementary strand (- - -).
severely destabilize DNA duplexes, and in fact the IQ-adducted
duplex with the T-mismatch was stabilized versus the unad-
ducted G-T-mismatch.53,54 From these studies, it is difficult to
assess the degree of stabilization provided by the proposed
hydrogen bond. The UV absorption and CD spectra of this
duplex are shown in Figure 8. The CD spectrum of the duplex
of 30 with the thymidine mismatch opposite the adducted base
is similar to that of 30 with a normal complementary strand
(Figure 5, panel B) except that the IQ absorbance is clearly
visible and its intensity is only modestly reduced versus the
adducted single-stranded oligonucleotide. With the T-mismatch,
the induced CD of the IQ moiety is positive and noticeably
more intense; the band at 270 nm is blue-shifted, and its intensity
is slightly reduced as compared to that with the normal
complement (274 nm) and the unmodified duplex (272 nm).
Unfortunately, the IQ absorbance when in a duplex is broad
and rather featureless. Consequently, it is difficult to determine
(55) Mao, B.; Hingerty, B. E.; Broyde, S.; Patel, D. J. Biochemistry 1998, 37,
95-106.
(56) O’Handley, S. F.; Sanford, D. G.; Xu, R.; Lester, C. C.; Hingerty, B. E.;
Broyde, S.; Krugh, T. R. Biochemistry 1993, 32, 2481-2497.
(57) Cho, B. P.; Beland, F. A.; Marques, M. M. Biochemistry 1992, 31, 9587-
9602.
(53) Allawi, H. T.; SantaLucia, J. Biochemistry 1997, 36, 10581-10594.
(54) Isaacs, R. J.; Spielmann, H. P. J. Am. Chem. Soc. 2004, 126, 583-590.
(58) Mao, B.; Vyas, R. R.; Hingerty, B. E.; Broyde, S.; Basu, A. K.; Patel, D.
J. Biochemistry 1996, 35, 12659-12670.
9
11196 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
Our interpretation of the CD and UV data for oligonucleotide
30 is in contrast to modeling studies which predict that the C8-
IQ adduct at the G3-position of the NarI restriction sequence
would adopt a groove-bound conformation.59,60 In this work, a
nonbiased conformational search was performed on a duplex
DNA sequence 5′-G1G2CG3CCA-3′ in which G3 (designated
G4 in their manuscript) was adducted with IQ at the C8-position.
Families of structures were identified in which the IQ moiety
was bound in either the minor groove or the major groove. The
lowest energy base-displaced intercalated conformation was
approximately 10 kcal/mol higher in energy than the lowest
energy minor groove-bound conformation. It was noted that no
attempt was made to correct for non-Coulombic (hydrophobic)
forces, and it was estimated that such interaction may account
for up to 2 kcal/mol. It is possible that the hydrophobic
contribution of C8-IQ adduct may have been underestimated;
IQ itself has been shown to intercalate into CG-rich sequences
of double-stranded DNA.61 It was also noted that the lack of a
lower energy base-displaced intercalated conformation might
be due to a “multiple minimum problem,” in which the observed
groove-bound conformations are kinetically trapped structures.
Similar results were found for the C8-adduct of MeIQx.60
Recently, Cosman reported the solution structure of an
oligonucleotide containing a C8-PhIP adduct in the sequence
5′-CCATCGCTACC; this work represents the first and to date
only structural analysis of an HCA-adducted oligonucleotide.62,63
The overall structure was described as similar to the C8-AF-
adducted oligonucleotide with some key differences. The major
conformation of the C8-PhIP-adducted oligonucleotides was a
base-displaced intercalated structure in which the adducted
guanine is situated in the major groove in a syn conformation.
The C6-phenyl and N3-methyl groups of PhIP protrude into
the minor groove, causing a widening of the minor groove and
a compression of the major groove. This causes an overall
bending of the DNA structure. A slow conformational exchange
with a minor conformation was reported. Few details of this
minor conformation were disclosed; however, it was described
as a minimally perturbed B-DNA duplex with the PhIP adduct
It was noted that the C6-phenyl group of the PhIP adduct was
twisted approximately 22° from the plane of the imidazaopy-
ridine ring. Eclipsing of ortho protons about the biaryl bond
represents a significant energy barrier and this biaryl twist may
impede complete intercalation of the PhIP group into the DNA
helix; extended aromatic structures such as IQ and AF would
not be subject to such steric constraints. This may prevent PhIP
from hydrogen bonding to the complementary deoxycytidine
as proposed above for IQ.
Summary
In summary, we have developed an efficient and general
strategy for the synthesis of C8-deoxyguanosine adducts of
arylamines involving the Buchwald-Hartwig palladium-
catalyzed N-arylation reaction of a suitably protected 8-bromo-
2′-deoxyguanosine derivative 16 with various mutagenic aryl-
amines including the dietary mutagen IQ (1). The C8-IQ-
adducted nucleoside was site-specifically incorporated into
oligonucleotides using the phosphoramidite approach. We have
examined the properties of two adducted oligonucleotides by a
combination of thermal UV melting studies, UV spectroscopy,
and circular dichrosim. We believe that the C8-IQ adduct of
the N-ras-12 sequence (29) adopts a groove-bound conforma-
tion, while at the G3-position of the NarI restriction sequence
(30) it is in a base-displaced intercalated conformation. Fur-
thermore, on the basis of the strong decrease of the IQ
absorption, we believe that a hydrogen-bonding interaction
between the IQ quinoline nitrogen and the complementary
cytosine base stabilizes the base-displaced intercalated form.
This differentiates IQ and potentially the rest of the HCAs from
simple arylamines and may play an important role in their
mutagenicity.
Experimental Section
All commercially obtained chemicals were used as received. Toluene
was freshly distilled from sodium and degassed with argon. THF was
distilled from a sodium/benzophenone ketyl. Pyridine was freshly
distilled from calcium hydride. Anhydrous 1,4-dioxane and 1, 2-di-
methoxyethane (DME) were purchased from Aldrich in Sure-seal bottles
and were used as received. All reactions were performed in oven-dried
glassware and under an argon atmosphere. Melting points are uncor-
rected. Proton and carbon-13 NMR data were recorded at 300 or 400
and 75 or 100 MHz, respectively, in CDCl3 unless otherwise noted.
High-resolution FAB mass spectra were obtained from the University
of Notre Dame Mass Spectrometry Center using nitrobenzyl alcohol
(NBA) as the matrix. MALDI-TOF mass spectrometry of oligonucle-
otides was performed at the Vanderbilt University Mass Spectrometry
Resource Center using a 3-hydroxypicolinic acid (HPA) matrix
containing ammonium hydrogen citrate (7 mg/mL) to suppress sodium
and potassium adducts. Electrophoretic analyses were carried out using
a Beckman P/ACE Instrument System 5500 Series monitored at 260
nm on a 27 cm × 100 mm column packed with the manufacturer’s
100-R gel (for ss DNA) using a Tris-borate buffer system containing
7 M urea.
1
located in the major groove. Previous H and 31P NMR, UV,
and fluorescence titration studies indicated that the major
noncovalent interaction of free PhIP with DNA was as a groove
binder and not an intercalator.64 The synthesis of the PhIP-
adducted oligonucleotide was accomplished via a biomimetic
approach in which the single-stranded DNA was reacted with
the PhIP nitrenium ion. The low yield of this process may have
limited the sequence to a single deoxyguanosine. Although the
full CG repeat portion of the frameshift-prone NarI restriction
sequence was not examined, deoxycytidines were flanking the
adducted bases. Our conclusion that the C8-IQ-adducted deoxy-
guanosine of 30 adopts a base-displaced intercalated conforma-
tion is based on light absorption and CD analysis and is
consistent with the NMR analysis of the related C8-PhIP adduct.
8-Bromo-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-
â-D-2′-deoxyribofuranosyl]guanine (17). A stirred suspension of
8-bromo-2′-deoxyguanosine (16, 2.9 g, 8.37 mmol) in dry pyridine (50
mL) was cooled 0 °C, and then 1,3-dichloro-1,1,3,3-tetraisopropyl-
disiloxane (2.9 g, 9.18 mmol) was added. The reaction mixture was
warmed to room temperature and stirred overnight. The pyridine was
evaporated under reduced pressure, and the residue was poured into
water and extracted with chloroform. The combined organic extracts
were successively washed with 10% HCl, saturated sodium bicarbonate,
(59) Wu, X.; Shapiro, R.; Broyde, S. Chem. Res. Toxicol. 1999, 12, 895-905.
(60) Gauvin, J.; Broyde, S.; Shapiro, R. Chem. Res. Toxicol. 2001, 14, 476-
482.
(61) Wantanabe, T.; Yokoyama, S.; Hayashi, K.; Kasai, H.; Nishimura, S.;
Miyazawa, T. FEBS Lett. 1982, 150, 434-438.
(62) Brown, K.; Hingerty, B. E.; Guenther, E. A.; Krishnan, V. V.; Broyde, S.;
Turteltaub, K. W.; Cosman, M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
8507-8512.
(63) Brown, K.; Guenther, E. A.; Dingley, K. H.; Cosman, M.; Harvey, C. A.;
Shields, S. J.; Turteltaub, K. W. Nucleic Acids Res. 2001, 29, 1951-1959.
(64) Marsch, G. A.; Ward, R. L.; Colvin, M.; Turteltaub, K. W. Nucleic Acids
Res. 1994, 22, 5408-5415.
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11197

A R T I C L E S
Elmquist et al.
and brine, and then dried over sodium sulfate and evaporated.
Purification by flash chromatography on silica, eluting with 1-2%
was extracted with ethyl acetate. The combined organic extracts were
washed with brine, dried over sodium sulfate, filtered, and evaporated.
Purification by flash chromatography on silica, eluting with 20% ethyl
acetate in hexanes, gave 20 (170 mg, 68%) as a yellow solid. mp 233-
235 °C; [R]23D -12° (c 0.55, CHCl3); 1H NMR (CDCl3) δ 9.50-8.00
(br s, 1H), 8.85 (d, J ) 4.1, 1H), 8.60 (d, J ) 8.3, 1H), 7.92 (d, J )
8.9, 1H), 7.63 (d, J ) 8.9, 1H), 7.52 (d, J ) 7.4, 2H), 7.46-7.36 (m,
4H), 6.66 (t, J ) 7.7, 1H), 5.60 (s, 2H), 4.83-4.80 (m, 1H), 4.15 (dd,
J ) 10.5, 3.1, 1H), 4.07-4.02 (m, 1H), 3.99-3.88 (m, 2H), 3.83 (s,
3H), 2.32-2.24 (m, 1H), 1.15-1.05 (m, 28H), 0.85 (s, 4H), 0.37 (s,
6H), 0.35 (s, 6H); 13C NMR (CDCl3) δ 159.6, 154.7, 154.0, 152.6,
151.5, 148.1, 145.1, 136.9, 129.8, 129.5, 128.5, 128.4, 127.9, 127.2,
122.8, 120.6, 118.0, 112.3, 105.7, 85.5, 81.4, 74.7, 67.5, 65.1, 36.6,
28.7, 17.5, 17.4, 17.3, 17.2, 17.1, 17.0, 13.5, 13.3, 13.0, 12.7, 8.7, -0.4,
-0.6; HRMS (FAB, NBA) m/z calcd for C46H68O5N9Si4 (M + H)
938.4421, found 938.4456.
methanol in chloroform, gave 17 (4.5 g, 91% yield) as a white solid,
1
mp 194-196 °C; [R]23 +5.7° (c 2.3, CHCl3); H NMR (CDCl3) δ
D
12.00 (br s, 1H), 6.46 (br s, 2H), 6.15 (dd, J ) 8.6, 3.0, 1H), 5.17 (m,
1H), 4.03 (dd, J ) 11, 2.3, 1H), 3.93-3.84 (m, 2H), 3.28-3.22 (m,
1H), 2.54-2.45 (m, 1H), 1.14-0.93 (m, 28H); 13C NMR (CDCl3) δ
157.8, 153.5, 152.0, 121.6, 117.9, 85.2, 83.8, 72.9, 63.6, 38.0, 17.6,
17.5, 17.7, 17.3, 17.2, 17.0, 13.4, 13.2, 13.0, 12.6; HRMS (FAB, NBA)
m/z calcd for C22H39O5N5BrSi2 (M + H) 588.1673, found 588.1657.
O6-Benzyl-8-bromo-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-1,3-
disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine (18). Triphen-
ylphosphine (285 mg, 1.08 mmol) and 17 (0.58 g, 0.98 mmol) were
placed in a dry flask, and the flask was evacuated for 10 min, and then
refilled with argon. Anhydrous 1,4-dioxane (10 mL) and benzyl alcohol
(117 mg, 1.08 mmol) were added, and the resulting suspension was
cooled to 0 °C. Diethylazodicarboxylate (190 mg, 1.08 mmol) was
added dropwise via syringe. After the addition was complete, the
reaction was warmed to room temperature and stirred for 1 h, after
which time it was concentrated in vacuo to give a yellow oil.
Purification by flash chromatography on silica, eluting with 10% ethyl
O6-Benzyl-8-[(3-methyl-3H-imidazo[4,5-f]quinolin-2-yl)amino]-
N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-â-D-2′-
deoxyribofuranosyl]guanine (21). To a stirred suspension of 18 (182
mg, 0.268 mmol), 2-amino-3-methylimidazo[4,5-f]-quinoline (IQ, 106
mg, 0.536 mmol), Pd2(dba)3 (25 mg, 0.0268 mmol), and BINAP (50
mg, 0.0804 mmol) in anhydrous, degassed toluene (2 mL) under an
argon atmosphere was added lithium bis(trimethylsilyl)amide (1.0 M
in THF, 0.54 mL, 0.54 mmol). The reaction was heated to 100 °C and
stirred for 25 min, and then cooled to room temperature and quenched
with saturated sodium bicarbonate (1 mL). The layers were separated,
and the aqueous phase was extracted with ethyl acetate. The combined
organic extracts were washed with brine, dried over sodium sulfate,
filtered, and evaporated. Purification by flash chromatography on silica,
eluting with 2% methanol in methylene chloride, gave 21 (141 mg,
acetate in hexanes, gave 18 (0.51 g, 76% yield) as a white solid. mp
1
127-129 °C; [R]23 -11° (c 1.0, CHCl3); H NMR (CDCl3) δ 7.47
D
(d, J ) 7.5, 2H), 7.36-7.29 (m, 3H), 6.21 (dd, J ) 8.9, 3.4, 1H), 5.51
(s, 2H), 5.23 (dt, J ) 7.8, 6.0, 1H), 4.75 (s, 2H), 4.03-3.97 (m, 1H),
3.90-3.83 (m, 2H), 3.31-3.24 (m, 1H), 2.54-2.46 (m, 1H), 1.15-
0.97 (m, 28H); 13C NMR (CDCl3) δ 159.79, 158.6, 154.0, 136.0, 128.3,
128.2, 128.0, 125.3, 116.2, 85.0, 83.7, 72.9, 68.0, 63.6, 37.8, 17.5, 17.4,
17.3, 17.3, 17.2, 17.1, 17.0, 16.9, 13.3, 13.1, 12.9, 12.5; HRMS (FAB,
NBA) m/z calcd for C29H45O5N5BrSi2 (M + H) 678.2143, found
678.2156.
66% yield) as a yellow solid. mp 203-205 °C; [R]23 -48° (c 0.5,
D
O6-Benzyl-8-bromo-2-(2,2,5,5-tetramethyl-1-aza-2,5-disilylcyclo-
pent-1-yl)-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-
â-D-2′-deoxyribofuranosyl]inosine (19). A solution of 18 (220 mg,
0.33 mmol) in anhydrous THF (20 mL) was cooled to -78 °C, and
then lithium bis(trimethylsilyl)amide (1.0 M in THF, 0.65 mL, 0.65
mmol) was added over 5 min. The solution was stirred for 15 min,
and then 1,2-bis(chlorodimethylsilyl)ethane (72 mg, 0.33 mmol) in THF
(1 mL) was added dropwise. The reaction was warmed to room
temperature and stirred for an additional 20 min. The reaction was
quenched by the addition of saturated sodium bicarbonate (10 mL).
The layers were separated, and the aqueous phase was extracted with
ether. The combined organic extracts were washed with brine, dried
over sodium sulfate, filtered, and evaporated. Purification by flash
chromatography on silica, eluting with 3% ethyl acetate in hexanes,
CHCl3); IR (KBr) 3348, 3215, 2943, 2866, 1601, 1465, 1421, 1134,
1
1081, 1033 cm-1; H NMR (CDCl3) δ 8.87 (dd, J ) 4.18, 1.66 1H),
8.56 (d, J ) 8.20, 1H), 7.93 (d, J ) 8.86, 1H), 7.63 (d, J ) 8.92, 1H),
7.57 (d, J ) 7.16, 2H), 7.46-7.36 (m, 4H), 6.69 (dd, J ) 9.02, 3.64,
1H), 5.61 (s, 2H), 5.32 (q, J ) 12.66, 7.00, 1H), 4.65 (s, 2H), 4.13-
4.05 (m, 1H), 3.99-3.93 (m, 2H), 3.81 (s, 3H), 3.32-3.25 (m, 1H),
2.56-2.49 (m, 1H), 1.18-1.05 (m, 28H); 13C NMR (CDCl3) δ 157.67,
155.5, 153.7, 152.3, 151.7, 147.9, 144.8, 129.7, 129.0, 128.5, 128.3,
128.0, 127.9, 122.8, 120.6, 117.7, 112.4, 106.5, 84.9, 80.6, 73.8, 67.6,
64.4, 37.9, 28.6, 17.5, 17.5, 17.4, 17.3, 17.0, 13.5, 13.3, 13.0, 12.7;
HRMS (FAB NBA) m/z calcd for C40H54N9O5Si2 (M + H) 796.3786,
found 796.3760.
8-[(3-Methyl-3H-imidazo[4,5-f]quinolin-2-yl)amino]-N9-[3′,5′-O-
(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofura-
nosyl]guanine (22). A stream of hydrogen gas was bubbled through a
solution of 21 (55 mg, 0.088 mmol) and 5% Pd/C (42 mg, 0.02 mmol)
in methanol (7 mL) for 2 h. The mixture was then stirred at room
temperature overnight under 1 atm of hydrogen. The catalyst was
removed by suction filtration, and the filtrate evaporated. Purification
by flash chromatography on silica, eluting with 10% methanol in
methylene chloride, gave 22 (35 mg, 74% yield) as a light yellow solid.
mp 280 °C (dec); [R]23D -18.4° (c 0.5, MeOH:CH2Cl2 ) 1:1); IR (KBr)
3347, 3215, 3021, 2944, 2867, 1679, 1592, 1496, 1382, 1137, 1088,
gave 19 (226 mg, 85% yield) as white solid. mp 94-96 °C; [R]23
D
-13° (c 1.1, CHCl3); 1H NMR (CDCl3) δ 7.43 (d, J ) 6.9, 2H), 7.37-
7.29 (m, 3H), 6.28 (t, J ) 7.6, 1H), 5.52 (s, 2H), 4.77-4.71 (m, 1H),
4.15-4.06 (m, 1H), 3.97-3.89 (m, 2H), 3.75-3.64 (m, 1H), 2.28 (ddd,
J ) 13.5, 6.9, 3.0, 1H), 1.12-0.98 (m, 28H), 0.83 (s, 4H), 0.33 (s,
6H), 0.31 (s, 6H); 13C NMR (CDCl3) δ 160.9, 159.0, 154.7, 136.1,
128.4, 128.0, 127.7, 124.6, 116.1, 85.5, 83.8, 73.7, 68.0, 64.4, 37.4,
17.5, 17.4, 17.3, 17.2, 17.1, 17.0, 16.9, 13.4, 13.2, 13.0, 12.6, 8.2, -0.5,
-0.7; HRMS (FAB, NBA) m/z calcd for C35H59O5N5BrSi4 (M + H)
820.2777, found 820.2771.
1
1032 cm-1; H NMR (DMSO-d6) δ 8.80 (dd, J ) 4.08, 1.64, 1H),
O6-Benzyl-8-[(3-methyl-3H-imidazo[4,5-f]quinolin-2-yl)amino]-
2-(2,2,5,5-tetramethyl-1-aza-2,5-disilylcyclopent-1-yl)-N9-[3′,5′-O-
(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofura-
nosyl]inosine (20). To a stirred suspension of 19 (220 mg, 0.268 mmol),
2-amino-3H-methylimidazo[4,5-f]-quinoline (IQ, 106 mg, 0.536 mmol),
Pd2(dba)3 (25 mg, 0.0268 mmol), and BINAP (50 mg, 0.0804 mmol)
in anhydrous, degassed toluene (2 mL) under an argon atmosphere was
added lithium bis(trimethylsilyl)amide (1.0 M in THF, 0.54 mL, 0.54
mmol). The reaction was heated to 100 °C and stirred for 25 min, and
then cooled to room temperature and quenched with saturated sodium
bicarbonate (1 mL). The layers were separated, and the aqueous phase
8.66 (dd, J ) 8.26, 0.68, 1H), 7.87 (d, J ) 8.92, 1H), 7.76 (d, J )
8.84, 1H), 7.56 (dd, J ) 8.32, 4.16, 1H), 6.51-6.46 (m, 3H), 4.60-
4.58 (m, 1H), 3.97-3.92 (m, 1H), 3.87-3.81 (m, 2H), 3.74 (s, 3H),
3.47-3.41 (m, 1H), 2.16-2.10 (m, 1H), 0.94-0.87 (m, 28H); 13C NMR
(DMSO-d6) δ 154.7, 153.7, 153.0, 149.5, 149.1, 148.0, 144.9, 131.5,
129.4, 128.8, 121.9, 121.1, 1183, 113.4, 103.8, 87.2, 82.0, 71.5, 63.7,
35.8, 28.9, 17.5, 17.4, 13.3, 12.7; HRMS (FAB NBA) m/z calcd for
C33H48N9O5Si2 (M + H) 706.3317, found 706.3292.
8-[(3-Methyl-3H-imidazo[4,5-f]quinolin-2-yl)amino]-2′-deoxygua-
nosine (14). To a stirred solution of 22 (320 mg, 0.45 mmol) in THF
(8 mL) was added tetrabutylammonium fluoride (1.06 mL, 1.0 M in
9
11198 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
THF, 1.06 mmol). The reaction was stirred at room temperature for
1.5 h, after which time the solvent was removed under reduced pressure.
Purification by flash chromatography on silica, eluting with 10%
methanol in methylene chloride containing 1% NH3, gave 14 (190 mg,
methylene chloride (3 mL) was added 2-cyanoethyl tetraisopropylphos-
phorodiamidite (0.028 mL, 0.085 mmol). The reaction was stirred for
1.5 h, after which time the solvent was removed under reduced pressure
and the residue was purified by flash chromatography on silica, eluting
with 4% methanol in methylene chloride containing 0.5% triethylamine,
90% yield) as a light green solid. mp 250 °C (dec); [R]23 +16° (c
D
1
0.55, MeOH:CHCl3 )1:1); H NMR (DMSO-d6) δ 11.05 (br s, 1H),
to give 25 (85 mg, 95% yield) as a yellow solid. mp 135 °C; [R]23
D
1
8.77 (dd, J ) 4.2, 1.7, 1H), 8.66 (d, J ) 8.3, 1H), 7.88 (d, J ) 8.9,
1H), 7.77 (d, J ) 8.9, 1H), 7.57 (dd, J ) 8.4, 4.2, 1H), 6.53-6.49 (m,
3H), 4.52-4.49 (m, 1H), 3.83-3.79 (m, 1H), 3.75 (s, 3H), 3.69 (dd, J
) 11.5, 5.2, 1H), 3.54 (dd, J ) 11.5, 5.6, 1H), 3.37-3.30 (m, 1H),
2.10-2.04 (m, 1H); 13C NMR (DMSO-d6) δ 154.11, 153.38, 152.81,
149.19, 148.60, 147.78, 144.55, 130.68, 129.11, 128.46, 121.75, 120.80,
117.73, 113.17, 104.05, 87.40, 81.94, 71.32, 62.34, 35.86, 28.58; HRMS
(FAB, NBA) m/z calcd for C21H22O4N9 (M + H) 464.1795, found
464.1806.
+55.8° (c 1.0, CH2Cl2); IR (KBr) 3408, 2964, 1625, 1530, 1340; H
NMR (CD2Cl2) δ 9.23 (s, 1H), 8.84 (dd, J ) 4.14, 1.66, 1H), 8.76 (d,
J ) 8.46, 1H), 8.49 (d, J ) 8.30, 1H), 7.91 (d, J ) 8.87, 1H), 7.61
(dd, J ) 8.92, 1.91, 1H), 7.48 (q, J ) 8.40, 4.18, 1H), 7.38 (t, J )
7.07, 2H), 7.26 (t, J ) 8.47, 4H), 7.19-7.11 (m, 3H), 6.78-6.64 (m,
5H), 5.07-4.93 (m, 1H), 4.24-4.19 (m, 1H), 3.91-3.22 (m, 16H),
3.10 (s, 3H), 3.05 (s, 3H), 2.60-2.57 (m, 4H), 1.34-1.07 (m, 12H);
13C NMR (CD2Cl2) δ 159.28, 159.25, 159.22, 159.19, 159.31, 156.22,
155.61, 155.54, 153.89, 153.83, 152.30, 152.26, 149.20, 149.00, 145.86,
145.80, 136.80, 136.66, 130.83, 130.78, 130.67, 130.61, 130.30, 128.93,
128.84, 128.47, 128.43, 128.36, 127.44, 127.36, 123.91, 123.85, 121.54,
118.50, 117.85, 117.79, 113.69, 112.81, 112.61, 112.42, 86.82, 86.78,
85.42, 82.84, 82.60, 75.89, 75.63, 74.92, 74.65, 65.27, 64.82, 59.38,
59.16, 59.01, 58.77, 55.88, 55.84, 55.79, 44.19, 44.02, 41.94, 41.90,
35.65, 29.20, 25.24, 25.15, 25.09, 25.00; HRMS (FAB NBA) m/z calcd
for C54H61N12O7P (M + H) 1021.4603, found 1021.4647.
N2-[(Dimethylamino)methylene]-8-[(3-methyl-3H-imidazo[4,5-f]-
quinolin-2-yl)amino]-â-D-2′-deoxyribofuranosyl]guanine (23). To a
stirred solution of 14 (100 mg, 0.22 mmol) in anhydrous methanol (5
mL) under an argon atmosphere was added N,N-dimethylformamide
diethyl acetal (0.3 mL, 1.76 mmol). The reaction was heated to 50 °C
and stirred for 2 h. The reaction mixture was allowed to cool to room
temperature, upon which time a precipitate formed. The precipitate was
collected by suction filtration and washed with cold methanol to give
General Procedure for the Buchwald-Hartwig Palladium-
Catalyzed Cross-Coupling Reaction of 20 with Simple Arylamines.
O6-Benzyl-8-(2-naphthalenylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis(iso-
propyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine (26a).
2-Aminonaphthalene (44 mg, 0.303 mmol), Pd2(dba)3 (14 mg, 0.015
mmol), BINAP (28 mg, 0.0449 mmol), and 18 (100 mg, 0.147 mmol)
were placed in an oven-dried flask, and the flask was evacuated for 20
min and then filled with argon. Freshly distilled toluene (3 mL) was
degassed by bubbling a stream of argon through the solution for 5 min
and then was added to the flask, and the mixture was stirred for 15
min while maintaining an argon atmosphere. Lithium bis(trimethylsilyl)-
amide (1.0 M in THF, 0.3 mL, 0.30 mmol) was added, and the reaction
was heated to 100 °C for 30 min, and then cooled to room temperature
and quenched with saturated sodium bicarbonate (1 mL). The layers
were separated, and the aqueous phase was extracted with ethyl acetate.
The combined organic extracts were washed with 10% HCl and brine,
dried over sodium sulfate, filtered, and evaporated. Purification by flash
chromatography on silica, eluting with 18% ethyl acetate in hexanes,
23 (75 mg, 66% yield) as a light yellow solid. mp 280 °C (dec); [R]23
D
+12.4° (c 0.5, MeOH:CH2Cl2 ) 1:1); IR (KBr) 3346, 3220, 2924, 1666,
1
1629, 1601, 1544, 1493, 1348, 1283, 1112 cm-1; H NMR (DMSO-
d6) δ 8.81 (dd, J ) 4.16, 1.68, 1H), 8.69 (dd, J ) 4.64, 0.92, 1H), 8.43
(s, 1H), 7.86 (d, J ) 8.92, 1H), 7.78 (d, J ) 8.96, 1H), 7.58 (dd, J )
8.56, 4.16, 1H), 4.57-4.53 (m, 1H), 3.85-3.82 (m, 2H), 3.74 (s, 3H),
3.72-3.68 (m, 2H), 3.59-3.54 (m, 1H), 3.32-3.27 (m, 1H), 3.14 (s,
3H), 3.02 (s, 3H), 2.17-2.14 (m, 1H); 13C NMR (DMSO-d6) δ 158.18,
156.92, 154.33, 153.91, 150.08, 148.19, 147.59, 144.85, 130.05, 129.53,
128.64, 122.29, 121.20, 117.72, 113.48, 108.84, 87.63, 82.33, 71.55,
62.60, 55.27, 36.86, 34.94, 28.88; HRMS (FAB NBA) m/z calcd for
C24H26N10O4 (M+) 518.2138, found 518.2162.
N2-[(Dimethylamino)methylene]-8-[(3-methyl-3H-imidazo[4,5-f]-
quinolin-2-yl)amino]-N9[5′-O-[bis(4-methoxyphenyl)phenylmethyl]]-
â-D-2′-deoxyribofuranosyl]guanine (24). To a stirred solution of 23
(50 mg, 0.096 mmol) in anhydrous pyridine (1.0 mL) and freshly
distilled diisopropylethylamine (0.116 mL, 0.066 mmol) under an argon
atmosphere was added 4,4′-dimethoxytrityl chloride (98 mg, 0.29 mmol)
at 0 °C. After the addition, the reaction was allowed to warm to room
temperature and was stirred for 3 h. The solvent was then removed
under reduced pressure, and the residue was purified by flash chro-
matography on silica eluting with 1.0% methanol in methylene chloride
containing 0.5% triethylamine gave 24 (62 mg, 79% yield) as a light
yellow solid. mp 157° C (dec); [R]23D +37.6° (c 0.5, CH2Cl2); IR (KBr)
3397, 2927, 1626, 1531, 1340; 1H NMR (CD2Cl2) δ 9.31 (s, 1H), 8.81
(d, J ) 2.81, 2H), 8.77 (d, J ) 8.19, 1H), 8.37 (s, 1H), 7.84 (d, J )
8.78, 1H), 7.50 (d, J ) 8.87, 1H), 7.46 (dd, J ) 8.25, 3.94, 1H), 7.36
(d, J ) 7.92, 1.31, 2H), 7.24 (dd, J ) 7.73, 2.40, 4H), 7.14 (t, J )
7.80, 2H), 7.07 (t, J ) 7.25, 1H), 6.70-6.63 (m, 5H), 5.10 (q, J )
6.02, 3.88, 1H), 4.85 (q, J ) 7.02, 2.30, 1H), 3.71-3.65 (m, 2H), 3.62
(s, 3H), 3.58 (s, 3H), 3.51 (dd, J ) 0.024, 0.017, 1H), 3.43 (s, 3H),
3.29 (dd, J ) 9.62, 4.67, 1H), 3.03 (s, 3H), 2.97 (s, 3H), 2.47-2.41
(m, 1H); 13C NMR (CD2Cl2) δ 158.83, 158.78, 157.82, 155.55, 152.92,
151.89, 149.09, 148.47, 145.38, 145.16, 136.32, 136.27, 130.31, 130.17,
129.90, 128.40, 128.14, 128.01, 127.38, 126.98, 123.35, 121.13, 117.13,
113.24, 112.31, 86.41, 85.73, 82.40, 73.53, 65.29, 55.38, 41.51, 37.20,
35.16, 30.05, 28.50; HRMS (FAB NBA) m/z calcd for C45H44N10O6
(M + H) 821.3524, found 821.3552.
gave 26a (71 mg, 65% yield) as a yellow solid: mp 78-82 °C; [R]23
D
-44° (c 1.0, CHCl3); IR (KBr) 1680, 1589, 1364, 1038 cm-1; 1H NMR
(CDCl3) δ 8.12 (bs, 1H), 7.70-7.22 (m, 14H), 6.15 (dd, J ) 7.6, 3.8,
1H), 5.53 (s. 2H), 5.78 (dt, J ) 7.5, 7.3, 1H), 4.77 (s, 2H), 4.0-3.8
(m, 3H), 3.10 (m, 1H), 2.44-2.39 (m, 1H), 1.0-0.75 (m, 28H); 13C
NMR (CDCl3) δ 158.9, 158.3, 147.5, 183.1, 138.1, 135.0, 130.4, 129.1,
128.8, 127.9, 124.7, 120.6, 114.2, 113.1, 86.0, 84.9, 71.4, 68.6, 62.9,
39.5, 17.8, 17.6, 14.2, 13.5; HRMS (FAB, NBA) m/z calcd for
C39H53N6O5Si2 (M + H) 741.3616, found 741.3639.
O6-Benzyl-8-(9H-fluoren-2-ylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis-
(isopropyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine
(26b). Obtained in 54% as a yellow solid: mp 95-97 °C; [R]23D -24°
(c 1.0, CHCl3); IR (KBr) 1680, 1591, 1462, 1038 cm-1 1H NMR
;
(CDCl3) δ 7.9 (bs, 1H), 7.67-7.21 (m, 16H), 6.21 (dd, J ) 8.7, 3.2,
1H), 5.58 (s, 2H), 4.79 (dt, J ) 7.1, 5.7, 1H), 4.63 (s, 2H), 4.12-3.87
(m, 6H), 3.04 (m, 1H), 2.46 (m, 1H), 1.08-0.69 (m, 28H); 13C NMR
(CDCl3) δ 158.9, 158.2, 154.1, 147.7, 145.3, 143.8, 142.4, 139.6, 137.8,
136.7, 129.2, 129.2, 128.8, 127.4, 126.5, 125.6, 120.8, 119.8, 118.1,
115.8, 113.1, 85.9, 84.8, 71.5, 68.5, 62.9, 39.4, 37.86, 18.1, 17.6, 14.2,
13.3; HRMS (FAB, NBA) m/z calcd for C42H55N6O5Si2 (M + H)
779.3772, found 779.3740.
N2-[(Dimethylamino)methylene]-8-[(3-methyl-3H-imidazo[4,5-f]-
quinolin-2-yl)amino]-N9-[5′-O-[bis(4-methoxyphenyl)phenylmethyl]-
3′-O-[[bis(1-methylethyl)amino](2-cyanoethoxy)phosphino]]-â-D-2′-
deoxyribofuranosyl]guanine (25). To a stirred suspension of 24 (50
mg, 0.061 mmol) and 1H-tetrazole (5 mg, 0.07 mmol) in anhydrous
O6-Benzyl-8-[(1,1′-biphenyl)-4-ylamino]-N9-[3′,5′-O-(1,1,3,3-tet-
rakis(isopropyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]gua-
nine (26c). Obtained in 62% yield as a yellow solid: mp 86-91 °C;
[R]23 -28° (c 1.0, CHCl3); IR (KBr) 1589, 1560, 1460, 1040 cm-1
;
D
1H NMR (CDCl3) δ 7.82-7.24 (m, 18H), 6.21 (dd, J ) 6.7, 3.6, 1H),
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11199

A R T I C L E S
Elmquist et al.
5.57 (s, 2H), 4.82 (dt, J ) 7.3, 6.5, 1H), 4.64 (s, 2H), 4.1 (m, 1H),
4.1-3.9 (m, 2H), 3.0 (m, 1H), 2.47-2.39 (m, 1H), 1.0-0.85 (m, 28H);
13C NMR (CDCl3) δ 157.4, 156.5, 152.3, 145.9, 140.0, 138.0, 136.1,
133.9, 127.8, 126.7, 125.8, 117.7, 111.5, 84.3, 82.9, 70.0, 67.1, 61.4,
37.8, 16.4, 12.3, 11.6; HRMS (FAB, NBA) m/z calcd for C41H55N6O5-
Si2 (M + H) 767.3772, found 767.3776.
General Procedure for the Desilylation of 27. 8-(2-Naphtha-
lenylamino)-2′-deoxyguanosine (28a). Using a plastic syringe and
needle, pyridinium poly(hydrogen fluoride) (1.0 mL, 70% hydrogen
fluoride, ∼35 mmol of fluoride) was added to dry THF (20 mL)
containing pyridine (5.0 mL, 62 mmol), giving a stock solution of
pyridinium hydrofluoride (∼1.4 N).65 27a (29 mg, 0.045 mmol) was
dissolved in the pyridinium hydrofluoride solution (2.0 mL, ∼2.4 mmol
of fluoride), and the solution was stirred at room temperature for 24 h.
The reaction was then concentrated under reduced pressure and purified
by flash chromatography on silica, eluting with 10% methanol in
methylene chloride to give 28a (15 mg, 82% yield) as a light brown
solid: mp > 300 °C; [R]23D -15.7° (c 1.0, MeOH/CH2Cl2); IR (KBr)
3371, 2961, 1672, 1558, 1384, 1093 cm-1; 1H NMR (DMSO) δ 10.58
(bs, 1H), 8.83 (bs, 1H), 8.32 (d, J ) 2.25, 1H), 7.80-7.70 (m, 4H),
7.45 (t, J ) 7.9, 1H), 7.30 (t, J ) 7.95, 1H), 6.40-6.30 (m, 3H), 6.05
(t, J ) 4.7, 1H), 5.32 (d, J ) 3.5, 1H), 4.42 (bs, 1H), 3.95 (d, J ) 1.5,
1H), 3.78 (bs, 2H), 2.08-2.0 (m, 1H); 13C NMR (DMSO) δ 155.7,
152.8, 149.5, 143.1, 138.5, 133.8, 128.5, 128.1, 127.4, 126.6, 126.3,
123.4, 119.5, 111.7, 87.2, 82.8, 71.3, 61.3; HRMS (FAB, NBA) m/z
calcd for C20H21N6O4 (M + H) 409.1624, found 409.1643.
O6-Benzyl-8-(4-methylphenylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis-
(isopropyl)-1,3-disiloxanediyl)-â-d-2′-deoxyribofuranosyl]guanine
(26d). Obtained in a 48% yield as a white solid: mp 82-86 °C; [R]23
D
-16° (c 1.0, CHCl3); IR (KBr) 1595, 1560, 1470, 1035 cm-1; 1H NMR
(CDCl3) δ 8.12 (bs, 1H), 7.49-7.06 (m, 14H), 6.17 (dd, J ) 7.5, 4.3,
1H), 5.56 (s, 2H), 4.78 (dt, J ) 7.5, 5.8, 1H), 4.62 (s, 2H), 4.1-4.08
(m, 1H), 3.90 (m, 1H), 3.0 (m, 1H), 2.44-2.38 (m, 1H), 1.06-0.88
(m, 28H); 13C NMR (CDCl3) δ 158.6, 157.6, 153.6, 147.6, 137.4, 131.9,
129.8, 128.2, 118.8, 112.8, 85.5, 83.9, 71.6, 68.3, 62.9, 38.9, 21.2, 17.6,
17.3, 13.8, 12.8; HRMS (FAB, NBA) m/z calcd for C36H53N6O5Si2 (M
+ H) 705.3616, found 705.3634.
General Procedure for the Hydrogenolysis of 35. 8-(2-Naphtha-
lenylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-1,3-disiloxanediyl)-
â-D-2′-deoxyribofuranosyl]guanine (27a). A stream of hydrogen gas
was bubbled through a solution of 26a (39 mg, 0.052 mmol) and 10%
Pd/C (50 mg, 0.047 mmol) in methanol (10 mL) and tetrahydrofuran
(10 mL) for 2 h. The mixture was then stirred at room temperature
overnight under 1 atm of hydrogen. The catalyst was removed by
suction filtration, and the filtrate evaporated. Purification by flash
chromatography on silica, eluting with 5% methanol in methylene
chloride, gave 27a (31 mg, 86% yield) as a yellow solid: mp 178-
8-(9H-Fluoren-2-ylamino)-2′-deoxyguanosine (28b). Obtained in
77% yield as a light red solid: mp > 300 °C; [R]23 -14.7° (c 1.0,
D
MeOH/CH2Cl2); IR (KBr) 3323, 2929, 1679, 1599, 1461, 1085 cm-1
;
1H NMR (DMSO) δ 10.6 (bs, 1H), 8.96 (bs, 1H), 8.12 (d, J ) 2.25,
1H), 7.86 (m, 1H), 7.65-7.50 (m, 4H), 7.25 (m, 1H), 6.40 (bs, 2H),
6.32 (dd, J ) 5.4, 9.5, 1H), 6.05 (s, 1H), 5.32 (d, J ) 3.5, 1H), 4.42
(bs, 1H), 4.08 (dd, J ) 5.0, 10.1, 1H), 3.93 (d, J ) 2.0, 1H), 3.76 (bs,
2H), 2.06-2.0 (m, 1H); 13C NMR (DMSO) δ 193.4, 155.7, 149.5,
144.6, 142.6, 142.0, 135.8, 134.2, 133.4, 128.0, 123.8, 122.5, 121.5,
120.3, 113.1, 112.2, 87.2, 82.8, 71.3, 61.4, 48.6; HRMS (FAB, NBA)
m/z calcd for C23H23N6O4 (M + H) 447.1781, found 447.1784.
184 °C; [R]23 52° (c 1.0, CHCl3); IR (KBr) 3368, 2945, 2866, 1689,
D
1360 cm-1; 1H NMR (CDCl3) δ 7.90-7.18 (m, 6H), 6.0 (dd, J ) 7.15,
3.8, 1H), 4.66 (m, 3H), 4.0-3.71 (m, 3H), 2.81 (m, 1H), 2.33 (m, 1H),
0.99-0.83 (m, 28H); 13C NMR (CDCl3) δ 158.3, 153.3, 150.9, 145.6,
138.9, 134.6, 129.9, 129.0, 127.9, 127.4, 126.6, 124.1, 120.1, 113.7,
85.3, 83.8, 71.4, 62.7, 39.2, 30.1, 17.4, 13.5, 12.9; HRMS (FAB, NBA)
m/z calcd for C32H47N6O5Si2 (M + H) 651.3146, found 651.3132.
8-[(1,1′-Biphenyl)-4-ylamino]-2′-deoxyguanosine (28c). Obtained
in 69% yield as a brown solid: mp > 300 °C; [R]23D 2.7° (c 1.0, MeOH/
CH2Cl2); IR (KBr) 3323, 2929, 1677, 1561, 1486, 1384, 1100 cm-1
;
1H NMR (DMSO) δ 10.53 (bs, 1H), 8.75 (bs, 1H), 7.83 (m, 2H), 7.65-
7.55 (m, 4H), 7.45-7.40 (t, J ) 7.3, 2H), 7.30 (m, 1H), 6.36 (bs, 2H),
6.32 (dd, J ) 4.1, 9.8, 1H), 5.96 (t, J ) 5.0, 1H), 5.32 (d, J ) 3.5,
1H), 4.43 (bs, 1H), 3.93 (d, J ) 1.8, 1H), 3.76 (bs, 2H), 2.05-2.0 (m,
1H); 13C NMR (DMSO) δ 156.0, 153.2, 149.8, 143.5, 140.6, 140.4,
132.6, 129.2, 127.1, 126.9, 126.4, 118.1, 112.5, 87.5, 83.2, 71.6, 61.7;
HRMS (FAB, NBA) m/z calcd for C22H23N6O4 (M + H) 435.1781,
found 435.1761.
8-(9H-Fluoren-2-ylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-
1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine (27b). Ob-
tained in a 80% yield as a yellow solid: mp 182-188 °C; [R]23 15°
D
(c 1.0, CHCl3); IR (KBr) 3371, 2944, 2866, 1684, 1461, 1038 cm-1
;
1H NMR (CDCl3) δ 7.67-7.13 (m, 7H), 5.93 (dd, J ) 7.8, 3.9, 1H),
4.65 (dt, J ) 7.6, 7.4, 1H), 4.0-3.71 (m, 5H), 2.78 (m, 1H), 2.31 (m,
1H), 1.0-0.77 (m, 28H); 13C NMR (CDCl3) δ 153.2, 150.6, 146.0,
143.4, 142.12, 139.7, 136.32, 127.0, 126.1, 125.24, 120.4, 119.5, 118.0,
115.9, 113.3, 85.2, 83.7, 71.1, 62.5, 39.1, 37.4, 30.1, 17.4, 12.9; HRMS
(FAB, NBA) m/z calcd for C35H49N6O5Si2 (M + H) 689.3303, found
689.3332.
8-[(4-Methylphenyl)amino]-2′-deoxyguanosine (28d). Obtained in
70% yield as a yellow solid: mp > 300 °C; [R]23D 1.5° (c 1.0, MeOH/
CH2Cl2); IR (KBr) 3323, 2930, 1678, 1581, 1475, 1081 cm-1; 1H NMR
(DMSO) δ 10.52 (bs, 1H), 8.59 (bs, 1H), 7.60 (d, J ) 8.5, 2H), 7.05
(d, J ) 8.5, 1H), 6.38-6.28 (m, 3H), 5.89 (bs, 1H), 5.31 (bs, 1H),
4.42 (d, J ) 5.7, 1H), 3.90 (d, J ) 1.9, 1H), 3.76 (bs, 2H), 2.05-1.95
(m, 1H); 13C NMR (DMSO) δ 155.9, 153.1, 149.8, 143.9, 138.6, 129.7,
129.3, 117.8, 87.5, 83.2, 71.6, 61.6, 20.7; HRMS (FAB, NBA) m/z
calcd for C17H21N6O4 (M + H) 373.1624, found 373.1626.
Synthesis of 5′-d(GGCA-GIQ-GTGGTG) (29). The IQ-adducted
oligonucleotides were synthesized using the phenoxyacetyl-protected
dA, 4-isopropyl-phenoxyacetal-protected dG, acetyl-protected dC, and
T phosphoramidites on an Expedite 8909 DNA synthesizer (PerSeptive
Biosystems) on a 1 µmol scale using the UltraMild line of phosphora-
midites and solid supports from Glen Research. The manufacturer’s
standard synthesis protocol was followed except at the incorporation
of the modified phosphoramidite (25), which was accomplished
manually off-line. At this point, the column was removed from the
instrument and sealed with two syringes, one which contained 250 µL
of the manufacturer’s 1H-tetrazole activator solution (1.9-4.0% in
acetonitrile) and the other which contained 250 µL of 25 (25 mg, 0.098
8-[(1,1′-Biphenyl)-4-ylamino]-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopro-
pyl)-1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine (27c).
Obtained in 82% yield as a yellow solid: mp 180-184 °C; [R]23D 43°
(c 1.0, CHCl3); IR (KBr) 3376, 2948, 2868, 1694, 1594, 1038 cm-1
;
1H NMR (CDCl3) δ 7.49-7.18 (m, 9H), 6.01 (dd, J ) 7.5, 3.7, 1H),
4.67 (m, 3H), 4.01-3.71 (m, 3H), 2.76 (m, 1H), 2.32 (m, 1H), 0.98-
0.81 (m, 28H); 13C NMR (CDCl3) δ 158.3, 153.2, 150.8, 145.7, 141.2,
135.0, 129.1, 127.8, 127.0, 119.1, 113.5, 85.2, 83.6, 71.3, 62.5, 39.2,
30.1, 17.7, 17.4, 13.7, 12.8; HRMS (FAB, NBA) m/z calcd for
C34H49N6O5Si2 (M + H) 677.3303, found 689.3334.
8-(4-Methylphenylamino)-N9-[3′,5′-O-(1,1,3,3-tetrakis(isopropyl)-
1,3-disiloxanediyl)-â-D-2′-deoxyribofuranosyl]guanine (27d). Ob-
tained in 78% yield as a white solid: mp 179-185 °C; [R]23 13° (c
D
1
1.0, CHCl3); IR (KBr) 3316, 2944, 2867, 1681, 1564, 1038 cm-1; H
NMR (CDCl3) δ 7.40-7.01 (m, 5H), 6.05 (dd, J ) 7.5, 4.6, 1H), 4.70
(dt, J ) 7.4, 7.3 1H), 4.03-3.76 (m, 3H), 2.85 (m, 1H), 2.34 (m, 1H),
2.24 (s, 3H), 1.01-0.81 (m, 28H); 13C NMR (CDCl3) δ 158.3, 153.1,
150.9, 146.4, 138.3, 132.2, 129.9, 119.6, 113.2, 85.4, 83.4, 71.6, 62.8,
39.0, 30.1, 21.2, 17.7, 17.4, 13.8, 12.9; HRMS (FAB, NBA) m/z calcd
for C29H47N6O5Si2 (M + H) 615.3146, found 615.3168.
(65) Trost, B. M.; Caldwell, C. G.; Murayama, E.; Heissler, D. J. Org. Chem.
1983, 48, 3252-3265.
9
11200 J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004

Synthesis and Properties of Oligonucleotides
A R T I C L E S
M in anhydrous methylene chloride). The 1H-tetrazole and 25 were
sequentially drawn through the column, and this procedure was repeated
periodically over 30-60 min. After this time, the column was washed
with anhydrous manufacturer’s grade acetonitrile and returned to the
instrument to proceed with capping, oxidation, detritylation, and the
remainder of the synthesis. The oligonucleotide was purified by HPLC
using 0.1 M ammonium formate (solvent 1) and methanol (solvent 2)
on a C-18 reversed-phase column with UV detection. The solvent
gradient was as follows: initially 99% solvent 1, then a 40 min linear
gradient to 50% solvent 1; 5 min isochratic at 50% solvent 1, then a 5
min linear gradient to the initial conditions. The yield of 29 was 9.0
units. MALDI-TOF MS (HPA) m/z calcd for (M - H) 3647.76, found
3647.22
Synthesis of 5′-d(CTCGGC-GIQ-CCATC) (30). This synthesis
followed the same procedure as for 29, with a difference in amounts
of modified phosphoramidite and tetrazole solution, where only 150
µL of the modified phosphoramidite was used (15 mg, 0.098 M in
anhydrous methylene chloride) and 200 µL of the manufacturer’s 1H-
tetrazole/acetonitrile activator solution was used. The synthesis and
purification proceeded as described above to give 5.5 units of 30.
MALDI-TOF MS (HPA) m/z calcd for (M - H) 3776.58, found
3776.38
Thermal Melting Studies. Equal amounts of the adducted and the
complementary strands (0.3 units each) were dissolved in 0.5 mL of
buffer (100 mM NaCl, 0.05 mM EDTA, and 10 mM phosphate buffer,
pH 7.1). The UV absorption at 260 nm was monitored as a function of
temperature. The temperature was increased at a rate of 1 °C/min from
5 to 90 °C. The melting temperatures of the native and modified
oligonucleotides were calculated by determining the first derivative of
the melting curve.
Circular Dichroism Measurements. CD measurements were carried
out with the same solutions as the Tm studies and at 23 °C. Samples
were scanned from 410 to 220.0 nm at 0.5 nm intervals averaged over
1 s in a 300 µL strain-free quartz cuvette. The corresponding raw CD
spectra were subtracted from a blank, converted to molar ellipticity,
and processed using the provided manufacturer’s software package.
Acknowledgment. This paper is dedicated to Professor
Thomas M. Harris in celebration of his 70th birthday. This work
was supported by a research grant from the American Cancer
Society (RPG-96-061-04-CDD) and a center grant from the
National Institute of Environmental Health Sciences (ES000267).
J.S.S. acknowledges support from an NIEHS predoctoral
traineeship (ES007028). We thank Ms. Pamela Tamura and
Albena Kozekova for assistance with the oligonucleotide
synthesis and mass spectrometry and Dr. Ivan Kozekov and
Professors Tom and Connie Harris for many helpful discussions
and suggestions.
Supporting Information Available: C1H, 13C, and 31P NMR
spectra of all new compounds. MALDI-TOF spectra, HPLC
traces of enzyme digests, CZE, and Tm spectra of adducted
oligonucleotides. This material is available free of charge via
the Internet at http://pubs.acs.org.
Enzymatic Digestion of Oligonucleotides. Enzymatic digestion of
IQ-modified oligonucleotides was performed by treatment of purified
oligonucleotides (0.2 units) in a reaction buffer (pH 7, 20 µL, 0.1 M
Tris HCl, 10 mM MgCl2) with 4 µL of Nuclease P1 stirred for 4 h at
37 °C. This was followed by the addition of 20 µL of pH 9 buffer
solution, 3 µL of alkaline phosphatase, and 5 µL of snake venom
phosphodiesterase, and the reaction was stirred for overnight. The
reaction mixtures were filtered and injected directly on HPLC: 90%
buffer initially, 15 min linear gradient to 70% buffer, 10 min linear
gradient to 25% buffer, 10 min isochratic gradient at 25% buffer, then
5 min linear gradient to initial conditions.
JA0487022
9
J. AM. CHEM. SOC. VOL. 126, NO. 36, 2004 11201