Synthesis of the minor acrolein adducts of 2′-deoxyguanosine and their generation in oligomeric DNA

Article abstract of DOI:10.1016/S0045-2068(02)00525-4

Acrolein, a known mutagen, undergoes reaction in vitro under physiological conditions with both 2^′-deoxyguanosine and native DNA to give rise to exocyclic adducts of the 5,6,7,8-tetrahydropyrimido[1,2-a]purine-10(3H)-one class having an hydroxy group at either the 6 or the 8 position. Previously we have shown that the 8-hydroxy derivative in a bacterial system has very low mutagenicity probably because in double-stranded DNA this residue exists in the open-chain aldehydic form [N²-(3-oxopropyl)-2^′-deoxyguanosine] (3). To continue our investigation in this area, we needed ample supplies of the 6-hydroxy isomers. This current paper describes high-yield simple methods for the synthesis in bulk of the 6-hydroxy adduct 1 and its incorporation into DNA oligomers. The basic methods for the synthesis of the adduct 1, involve 1-substitution of dG derivatives with a 3-butenyl group, dihydroxylation of the olefin with osmium tetroxide and N-methylmorpholine N-oxide, then diol cleavage with periodate ion after incorporation of the 1-(3,4-diacetoxybutyl)-2^′-deoxyguanosine into oligomeric DNA.

Full text of DOI:10.1016/S0045-2068(02)00525-4

BIOORGANIC
CHEMISTRY
Bioorganic Chemistry 31 (2003) 136–148
www.elsevier.com/locate/bioorg
Synthesis of the minor acrolein adducts of
20-deoxyguanosine and their generation
in oligomeric DNAq
Yanhe Huang, M. Cecilia Torres, Charles R. Iden,
and Francis Johnson*
Department of Pharmacological Sciences, State University of New York at Stony Brook,
Stony Brook, NY 11794, USA
Received 22 July 2002
Abstract
Acrolein, a known mutagen, undergoes reaction in vitro under physiological conditions
with both 20-deoxyguanosine and native DNA to give rise to exocyclic adducts of the
5,6,7,8-tetrahydropyrimido[1,2-a]purine-10(3H)-one class having an hydroxy group at either
the 6 or the 8 position. Previously we have shown that the 8-hydroxy derivative in a bacterial
system has very low mutagenicity probably because in double-stranded DNA this residue ex-
ists in the open-chain aldehydic form [N2-(3-oxopropyl)-20-deoxyguanosine] (3). To continue
our investigation in this area, we needed ample supplies of the 6-hydroxy isomers. This current
paper describes high-yield simple methods for the synthesis in bulk of the 6-hydroxy adduct 1
and its incorporation into DNA oligomers. The basic methods for the synthesis of the adduct
1, involve 1-substitution of dG derivatives with a 3-butenyl group, dihydroxylation of the ole-
fin with osmium tetroxide and N-methylmorpholine N-oxide, then diol cleavage with perio-
date ion after incorporation of the 1-(3,4-diacetoxybutyl)-20-deoxyguanosine into oligomeric
DNA.
Ó 2003 Elsevier Science (USA). All rights reserved.
Keywords: Acrolein; Mutagenicity; 20-Deoxyguanosine adducts; 1-Alkylation; Osmium tetroxide; Oligo-
meric synthesis; Diol cleavage; Xenodeoxynucleotides
q This work was supported by a United States Public Health Service-National Cancer Institute, Grant
CA47995.
* Corresponding author. Fax: 1-631-632-7394.
E-mail address: francis@pharm.sunysb.edu (F. Johnson).
0045-2068/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0045-2068(02)00525-4

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
137
1. Introduction
Acrolein is a substance ubiquitous in the human environment [1]. It is regarded as
genotoxic and mutagenic because it reacts readily with nucleophilic biological mac-
romolecules. However, evaluation of the possible carcinogenicity of acrolein has
been complicated by its irritant properties, toxicity, and especially the difficulty of
introducing its nucleoside adducts, site-specifically, into DNA [2].
The major reaction products of acrolein with 20-deoxyguanosine1 are adducts,
which are formed through a pair of regio-isomeric Michael additions with initial
bond formation happening either at the 1 or N2 position, followed by ring closure
to form three exocyclic hydroxy-isomeric adducts. These have been characterized
by Chung et al. [3] as the 6- and 8-hydroxy-tetrahydropyrimido-[1,2-a]purine-
10(3H)-one nucleosides adducts 1 and 2 (Scheme 1). Although it has been shown that
the 8-hydroxy adduct is formed when DNA is treated acrolein, the 6-hydroxy ad-
ducts (1) have not been reported [3]. Nevertheless, given the chemical reactivity of
acrolein it seems likely that 1 is also generated in DNA in small amounts. Such exo-
cyclic nucleic acid adducts, because they have the potential to disrupt base-pairing in
DNA, probably play an important role in mutagenesis and possibly carcinogenesis
[4,5]. However, it is difficult to evaluate how any one adduct alters the structure
or coding properties of the damaged DNA.
Recently we devised [6] an original strategy for the generation in DNA of alkali-
labile adducts derived from a;b-unsaturated aldehydes. This uses an alkali-resistant
terminal homo-1,2-diol as a masked form of the aldehyde. The diol is easily pro-
tected as its diacetyl derivative during oligomer synthesis but is regenerated during
the subsequent ammonia deprotection step. Treatment with periodate then generates
the desired aldehyde adduct in essentially quantitative yield together with one equiv-
alent of formaldehyde. From mass special evidence it does not appear that the form-
aldehyde interferes further with the DNA. This method was applied by our group to
the synthesis of several oligomers containing 2 as a nucleotide residue [6] and was
later adopted by Nechev et al. [7] for the same purpose and for the generation in
DNA of the closely related crotonaldehyde adducts [8]. More recently the method
has been applied by Wang and Rizzo [9] to the synthesis of the 1,N2-deoxyguanosine
adducts of the lipid peroxidation product, trans-4-hydroxynonenal.
In our own investigations [10] of the structural aspects of the DNA lesion derived
from 2, we found that it exists in both the monomeric nucleoside and in single-
stranded DNA as the closed-ring form depicted in 2, but in double-stranded DNA
it exists dominantly, if not exclusively, in the ring-opened form 3 (or its hydrate).
A similar observation was made by Marnett and his associates [11] in the case of
the exocyclic adduct (M1G) derived from malondialdehyde and 20-deoxyguanosine.
It was also found [12,13] that this lesion is non-mutagenic in bacterial systems, a
finding that is in accord with previous work by Lloyd [14] who demonstrated that
the N2-monoalkyl adducts of 20-deoxyguanosine derived from butadiene
1
Abbreviations used: dG, 20-deoxyguanosine.

138
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
Scheme 1.
mono-epoxide also were only marginally mutagenic. This may be a general rule in
prokaryotic systems for relatively small N2-alkyl substituents. However in a mamma-
lian system (COS-7 cells) this lesion has been shown to be significantly mutagenic
[15]. Nevertheless it is not clear if the mutagenicity of acrolein is entirely due to this
adduct (i.e. 2) or if other adducts are involved.
In the light of these uncertainties we decided that it would be important to study
the isomeric adduct 1 and its incorporation into DNA oligomers. We therefore
turned our attention to preparative studies to try to develop a high yield method that
would allow the synthesis of 1 in bulk and thereafter the preparation of homoge-
neous oligomeric DNA containing 1, located site-specifically within the linear
array.
2. Results and discussion
In a previous publication [16], we described a method for the synthesis of 3-(30;
50-O-bis(tert-butyldimethylsilyl)-20-deoxy-b-
D-erythro-pentofuranosyl)-6-hydroxy-5,
6,7,8-tetrahydropyrimido[1,2-a]purine-10(3H)-one (4) in high yield. It was important
to test the stability of compound 4 under the conditions needed for deprotection and
release of synthetic oligodeoxynucleotides from the control pore glass (CPG) sup-
port [11], since our aim was the site-specific incorporation of this monomer into olig-
omers. Compound 4 was treated with 28% NH4OH containing a small amount of
MeOH for 20 h at room temperature in sealed condition. The analysis by HPLC

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
139
showed that about 13% of the starting material 4 decomposed. After treatment with
80% HOAc for 40 min at room temperature, about 18% of starting material 4 hydro-
lyzed. It was clear that monomer 4 is not stable to the conditions needed to obtain a
completely deprotected oligomer. Attempts to protect the OH group of 4 by acety-
lation led principally to 6 and only a small amount of the desired 5 (Scheme 1).
Further work on this approach was abandoned when it was found that a key in-
termediate 9 could be prepared directly (Scheme 2) and in exceptionally high yield
(94%) by the reaction in THF solution of N2-dimethylaminomethylene-30; 50-di-t-bu-
tyldimethylsilyl-20-deoxyguanosine (8) with 4-bromobutene in the presence of the
mild base Cs2CO3 [17]. That substitution at the 1-position of the dG derivative
had occurred was confirmed chemically. Oxidation of 9 with N-methylmorpholine-
N-oxide in the presence of a catalytic quality of osmium tetroxide at room
temperature in t-BuOH gave 30; 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybu-
tyl)-N2-dimethylaminomethylene-20-deoxyguanosine (10). The free hydroxy groups
in diol 10 were then protected by acetylation to give 11 using acetic anhydride in pyr-
idine. The 1H NMRspectrum of 11 showed the presence of two new peaks at d 1.961
and 1.965 each representing an acetyl group thus substantiating structure (11).
Scheme 2.

140
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
The diacetyl compound 11 was then treated with methanolic ammonia solution
which led to 30; 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-20-deoxygu-
anosine (12). When the latter was oxidized by sodium periodate in aqueous
CH3CN, the known [16] cyclic 6-hydroxy derivative 4 was obtained. When the oxi-
dation was run in aqueous MeOH, two products namely cyclic 6-hydroxy and 6-
methoxy acrolein adducts of dG (4 and 13) were formed as expected. Both adducts
1
showed H NMR, mass and UV spectra identical with previously synthesized [16]
samples. In addition their HPLC behavior was identical.
The DMT-phosphoramidite 16 (Scheme 3), the building block required for the in-
troduction of the nucleoside lesion into DNA, was then synthesized as follows from
Scheme 3.

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
141
the diacetyl compound 11. Selective removal of the silyl protecting groups to give
compound (14) was accomplished with HF in excess pyridine, a reaction that leaves
intact both the nucleoside linkage and the N2-dimethylaminomethylene moiety. The
conversion of compound 14 into the corresponding 50-O-dimethoxytrityl derivative
was then achieved by treatment with 4; 40-dimethoxytrityl chloride in pyridine under
standard conditions. The resulting 50-O-(4,40-dimethoxytrityl-1-(3,4-diacetylbutyl))-
N2-dimethylaminomethylene-20-deoxyguanosine (15) was then allowed to react with
3.3 eq of 2-cyanoethyl diisopropylchlorophosphoramidite in the presence of 6 eq of
triethylamine to obtain the required product 16, whose structure was confirmed by
1
mass spectral, H NMRand 31P NMRdata. Having shown that the cyclic acrolein
adducts can be formed by cleavage of the precursor the diol under conditions men-
tioned above, we now directed our attention to the generation of lesion 14 in DNA.
The experimental heptamer 17a [50-TTCXATT-30, where X ¼ 1-(3,4-diac-
etoxybufyl)-N2-(dimethylaminomethylene) dG] was synthesized using standard
phosphoramidite chemistry on an Applied Biosystems 394 synthesizer (Foster City,
CA). The coupling time allowed for the modified base was 5 min which gave a cou-
pling efficiency of 99%. The oligomer was treated with 28% NH4OH at 55 °C over-
night to cleave it from the (CPG) support and to remove the protecting groups. A
Waters HPLC system (Milford, MA) consisting of a 600 E Multisolvent Delivery
System, a U6K Injector, and a 996 Photodiode Array Detector was used to purify
the modified oligomer. The dimethoxytrityl capped (DMT-on) oligonucleotide was
purified on a Luna 5l Phenyl-Hexyl (250 Â 10 mm) column (Phenomenex, Torrance,
CA) run at 4 mL/min. A gradient of 16–36% acetonitrile (35 min) in 0.1 M TEAA
buffer, pH 6.8, was used to collect the DMT-on oligomer. The DMT group was re-
moved with 80% HOAc, and the pure DMT-off product (18a) was obtained using a
gradient of 0–20% acetonitrile over 40 min. ESI-MS was used to confirm the mass of
both, the DMT-on and the DMT-off oligomers. The observed masses were
2476:5 Æ 0:9 and 2173:3 Æ 0:4 Da, respectively (the corresponding calculated masses
are 2476.4 and 2174.4 Da). Heptamer 18a (50-TTCXATT-30 where X is the 1-(3,4-di-
hydroxybutyl) dG moiety (20), was treated with 100 mM sodium periodate in 0.5 M
sodium acetate buffer, pH 6.0, for 5 min at room temperature [6] to generate the 6-
hydroxy-1,N2-propanodeoxyguanosine (6-OH-PdG) moiety. A final HPLC purifica-
tion (Fig. 1a) was carried out under the same conditions described above for the
DMT-off oligodeoxynucleotides. Two partially resolved peaks could be seen corre-
sponding to the two possible diastereomers (6-hydroxy isomers). The peaks were col-
lected as a single fraction and the mass determined by electrospray ionization mass
spectrometry (ESI-MS). The observed mass for oligodeoxynucleotide (19a) was
2142:2 Æ 0:1 (Fig. 1b, the calculated m/z is 2142.40 Da). In a separate experiment,
each of the fractions corresponding to the two observed peaks was collected sepa-
rately and dried under vacuum. Each fraction was redissolved in water and subjected
to an additional HPLC pass. Two peaks were again observed for each fraction, in-
dicating an active equilibrium between the two isomeric oligodeoxynucleotides.
The presence of the intact 6-hydroxy isomer (1) was confirmed by enzymatic anal-
ysis of the oligodeoxynucleotide 19a using nuclease P1 followed by treatment with
snake venom phosphodiesterase, and bacterial alkaline phosphatase [18]. Subsequent

142
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
Fig. 1. Data related to the incorporation of 6-hydroxy-1,N2-propanodeoxyguanosine (6-OH-PdG) into
oligomeric DNA under normal DNA synthesizer condition: (a) the HPLC of the post-synthetically mod-
ified heptamers after oxidation with NaIO4 and their UV spectra; (b) the electrospray mass spectrum of
the purified heptamers; (c) the HPLC of the enzymatic digestion products of the site-specific modified hep-
tamers.
HPLC analysis of the resulting nucleoside mixture and comparison with a prepared
standard of 6-hydroxy isomer 1 [16] corroborated the structure of the desired oligo-
mer. The 6-hydroxy isomer 1 gives two peaks, corresponding to the two possible dia-
stereomers, one of which elutes at almost the same retention time as dA under the
HPLC conditions used (Fig. 1c).
Modified nucleoside 14 was then introduced into a series of DNA oligomers. By
way of examples, the 13-mer (17b) (50-CTCCTC-X-ATACCT-30) and the 28-mer

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
143
(17c) (50-CTGCTCCTC-X-ATACCTACACGCTAGAAC-30) were synthesized and
purified as described for the heptamer above. When the purified oligomers (18a
and b) containing the 1-(3,4-dihydroxybutyl) dG moiety were treated with an excess
of sodium periodate in aqueous solution (0.1 M in CH3COONa buffer, pH 6.8), until
no starting material remained (15 min), they afforded the corresponding single-
stranded DNA oligomers (19a and b) containing the desired adduct 1. After a further
HPLC purification (DMT-off conditions) both oligomers were characterized by ESI-
MS. Observed m=z values for the 19a and 19b, respectively, were 3900:2 Æ 0:7 (calcd.
3901.5) and 8508:0 Æ 1:3 Da (calcd. 8509.6).
In conclusion, we have developed a synthetic method that allows the generation of
the alkali-sensitive acrolein adduct 1 in DNA. We believe that this methodology has
the general potential to allow the synthesis of DNA containing other alkali-sensitive
lesions. Work is now in progress to determine the mutational spectrum of adduct 1 in
both bacterial and mammalian cell systems, and to determine the effect of this lesion
on DNA structure, using NMRmethods.
After this work was completed, a publication by Nechev et al. [19] appeared
describing a similar but not identical approach for the synthesis of DNA oligo-
mers containing the 6-hydroxy acrolein adduct 1. However our method has the
advantage of using an alkylation procedure at the 1-position of 20-deoxyguanosine
that gives the desired products in high yield and in an excellent state of
purity.
3. Experimental procedures
All reagents and solvents were of commercial grade and were used as such un-
1
less otherwise specified. The H NMRspectra were recorded on a Bruker AC-250
or a General Electric QE-300 spectrometer. Fast atom bombardment (FAB) mass
spectra were recorded on a micromass Trio 2000, glycerol or nitrobenzyl alcohol,
being used as the matrix. Electrospray ionization mass spectra were obtained
from a micromass Quattro LC. Flash column chromatographic separations were
carried out on 230–400 mesh silica gel (TSI Chemical Company, Cambridge,
MA). The purities of all new products were confirmed by TLC analysis in two
solvent systems (EtOAc:hexane) and (CH2Cl2:MeOH). In all cases only a single
spot was observed.
3.1. 30; 50-O-Bis(tert-butyldimethylsilyl)-1-(but-3-enyl)-N2-dimethylaminomethylene-
20-deoxyguanosine (9)
To a suspension of 30; 50-O-bis(tert-butyldimethylsilyl)-N2-dimethylaminomethyl-
ene-20-deoxyguanosine (2.205 g, 4 mmol) in THF (75 mL) was added Cs2CO3
(3.258 g, 10 mmol), and the mixture was stirred at room temperature for 10 min. 4-
Bromo-1-butene (1.6 mL, 16 mmol) was then added and the mixture was heated
at 70 °C under a nitrogen atmosphere overnight. The reaction mixture was then

144
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
evaporated under reduced pressure to remove solvent, and the residue was stirred
with CH2Cl2 (200 mL). The organic extract was then washed with water and brine,
dried with anhydrous Na2SO4, filtered, and evaporated. The residual material was
purified by silica gel chromatography (CH2Cl2/MeOH/concentrated ammonia in
water: 20/1/0.1–10/1/0.1) to afford 9 as a white solid (2.194 g, 94%). MP: 84–85 °C;
1H NMR(CD OD, 300 MHz) d, ppm: 0.072 (6H, s, Si–(CH3Þ ), 0.102 (6H, s, Si–
3
2
(CH3Þ ), 0.904 (9H, s, C–(CH3Þ ), 0.911 (9H, s, C–(CH3Þ ), 2.340 (1H, m, H02)
2
3
3
2.446 (3H, m, C@C–CH2–C–N, H200), 3.118 (3H, s, NCH3), 3.188 (3H, s, NCH3),
3.750 (2H, d, J ¼ 3:6 Hz, H05), 3.962 (1H, dd, J ¼ 3:6, 6.6 Hz, H04Þ, 4.378 (2H, t,
J ¼ 7:2 Hz, C@C–C–CH2–N), 4.564 (1H, m, H03) 4.963 (1H, d, J ¼ 10:2 Hz,
HC@C–C–C–N), 5.024 (1H, dd, J ¼ 1:8, 13.5 Hz, HC@C–C–C–N), 5.860 (1H,
m, C@CH–C–C–N), 6.344 (1H, t, J ¼ 6:7 Hz, H01), 7.834 (1H, s, H8), 8.563 (1H, s,
2-N@CHN); FAB-MS: m=z 627 (M + Naþ), 605 (M + Hþ).
3.2. 30; 50-O-Bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-N2-dimethylaminom-
ethylene-20-deoxyguanosine (10)
To a solution of osmium tetroxide (OsO4, 2 mg) in 0.3 mL tert-butanol was added
N-methylmorpholine N-oxide (1.38 mL, 50% in water, 6.66 mmol), water (280lLÞ,
and acetone (1.7 mL), and the mixture was stirred at room temperature for 10 min.
30; 50-O-Bis(tert-butyldimethylsilyl)-1-(but-3-enyl)-N2-dimethylaminomethylene-20-
deoxyguanosine (1.815 g, 3 mmol) was then dissolved in THF (4.5 mL) and the solu-
tion was cooled in an ice bath. The two solutions were mixed together and cooling
was continued for a further 15 min, then the ice bath was removed and the reaction
mixture was allowed to stand at room temperature for 20 h. The mixture was treated
with 3 mL of 20% Na2SO3 and concentrated at reduced pressure.
The resulting residue was diluted with 50 mL ethyl acetate and the organic extract
was washed with saturated NaHCO3 solution, then water and brine, and finally dried
over anhydrous Na2SO4. Concentration in vacuo afforded the desired product 10 as
1
a white solid (1.815 g, 94.6%). MP: 90–92 °C; H NMR(CDCl , 300 MHz) d, ppm:
3
0.067 (6H, s, Si–(CH3Þ ), 0.093 (6H, s, Si–(CH3Þ ), 0.895 (9H, s, C–(CH3Þ ), 0.900
2
2
3
(9H, s, C–(CH3Þ ), 1.676 (1H, m, 1-C–C–CH–C–N), 1.939 (1H, m, 1-C–C–CH–C–
3
N), 2.380 (2H, m, H02Þ, 3.118 (3H, s, NCH3), 3.203 (3H, d, J ¼ 1:2 Hz, NCH3),
3.488 (3H, m, O–CH2–C–C–C–N, O–C–C(O)H–C–C–N), 3.750 (2H, d, J ¼ 3:6 Hz,
H05Þ, 3.956 (1H, dd, J ¼ 3:4, 6.6 Hz, H04Þ, 4.366 (H, m, C–C–C–CH–N), 4.563 (2H, m,
0
0
H3, C–C–C–CH–N), 6.323 (1H, t, J ¼ 6:6 Hz,H1 ), 7.878 (1H, s, H8), 8.576 (1H, s,
2-N@CHN); FAB-MS: m=z 661.3 (M + Naþ), 639.3 (M + Hþ).
3.3. 30; 50-O-Bis(tert-butyldimethylsilyl)-1-(3,4-diacetoxybutyl)-N2-dimethylaminom-
ethylene-20-deoxyguanosine (11)
To a solution of 30; 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-N2-
dimethylaminomethylene-20-deoxyguanosine (10) (3.2 g, 5 mmol) and a catalytic
amount of dimethylaminopyridine (DMAP) in pyridine (70 mL) was added acetic
anhydride (Ac2O, 3 mL, 29.5 mmol). The reaction mixture was stirred at room

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
145
temperature for 15 h, then poured over ice to produce a yellow solid. The solid was
filtered and washed with water to obtain the desired product 11 as a yellow powder.
This was dried under vacuum overnight (3.37 g, 93%). 1H NMR(CDC1 3, 250 MHz)
d, ppm: 0.137 (12H, s, Si(CH3Þ ), 0.933 (18H, s, C–(CH3Þ ), 1.961 (3H, s, CH3COO–),
2
3
1.965 (2H, m, 1-C–C–CH2–C–N), 2.004 (3H, s, CH3COO–), 2.346 (H, m, H02), 2.707
(H, m, H020Þ, 3.129 (3H, s, NCH3), 3.221 (3H, s, NCH3), 3.717 (2H, m, H05Þ, 3.962 (1H,
m, H04), 4.070 (1H, m, O–C–C(O)H–C–C–N), 4.261 (1H, m, O–C–C–C–CH–N),
0
4.307 (2H, m, O–CH2–C–C–C–N), 4.653 (1H, m, H3 ), 5.078 (1H, m, O–C–C–C–
0
CH–N), 6.306 (1H, t, J ¼ 6:8 Hz, H1 ), 8.020 (1H, s, H8), 8.539 (1H, s, 2-N@CHN);
FAB-MS: m=z 723.3 (M + HþÞ.
3.4. 1-(3,4-Diacetoxybutyl)-N2-dimethylaminomethylene-20-deoxyguanosine (14)
To a solution of 30; 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-diacetoxybutyl)- N2-
dimethylaminomethylene-20-deoxyguanosine (11, 1.464 g, 2 mmol) in pyridine
(12 mL) cooled in an ice bath was added a solution (70%) of HF/pyridine
(1.2 mL) and the ice bath was then withdrawn. The mixture was stirred at room
temperature for 3 h, then was neutralized with NaHCO3. An air stream was then
used to remove the solvent. The residual solid was then purified by direct appli-
cation to a silica gel column. Elution with CH2Cl2/MeOH (10/1) gave the product
1
14 as a fluffy yellow foam (0.810 g, 82%). H NMR(CD OD, 250 MHz) d, ppm:
3
1.905 (3H, s, CH3COO–), 1.960 (3H, s, CH3COO–), 2.032 (2H, m, 1-C–C–CH2–
C–N), 2.388 (H, m,H02), 2.726 (H, m, H200), 3.161 (3H, s, NCH3), 3.235 (3H, s,
NCH3), 3.737 (2H, m, H05), 3.993 (1H, m, H04), 4.065 (1H, m, O–C–C(O)H–C–
C–N), 4.278 (1H, m, O–C–C–C–CH–N), 4.360 (2H, m, O–CH2–C–C–C–N),
0
4.546 (1H, m, H3 ), 5.115 (1H, m, O–C–C–C–CH–N), 6.346 (1H, t, J ¼ 6:8 Hz,
0
H1 ), 8.044 (1H, s, H8Þ, 8.600 (1H, s, 2-N@CHN); FAB-MS: m=z 495.2
(M + HþÞ.
3.5. 50-O-(4; 40-Dimethoxytriphenylmethyl)-1-(3,4-diacetoxybutyl)-N2-dimethylami-
nomethylene-20-deoxyguanosine (15)
1-(3,4-Diacetoxybutyl)-N2-dimethylaminomethylene-20-deoxyguanosine (14) (494
mg, 1 mmol) was co-evaporated with dry toluene (3 Â 4 mL) and pyridine
(3 Â 4 mL). The anhydrous residue was then redissolved in pyridine (15 mL) and
4; 40-dimethoxytritylchloride (DMT-Cl, 407 mg, 1.2 mmol) was added. The mixture
was stirred at room temperature for 1 h under nitrogen and then quenched with
0.75 mL MeOH, and the volatiles were removed under reduced pressure. Methylene
chloride (50 mL) was then added to the residue and the organic extraction was
washed with saturated NaHCO3 solution then water, dried (Na2SO4Þ, and concen-
trated in vacuo. The residue was purified by silica gel column chromatography
(CH2Cl2/MeOH/TEA: 25/1/1%) to obtain the desired product 15 (600 mg, 75%) as
1
a yellow solid. H NMR(CDCl , 250 MHz) d, ppm: 1.961 (2H, m, 1-C–C–CH2–
3
C–N), 1.965 (3H, s, CH3COO–), 1.985 (3H, s, CH3COO–), 2.535 (2H, m, H02),
3.037 (3H, s, NCH3), 3.069 (3H, s, NCH3), 3.294 (2H, m, H05), 3.543 (1H, m, H40 ),

146
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
3.572 (6H, s, CH3OAr), 4.023 (1H, m, O–C–C(O)H–C–C–N), 4.197 (2H, m, O–
0
CH2–C–C–C–N), 4.322 (1H, m, O–C–C–C–CH–N), 4.601 (1H, m, H3 ), 5.600
0
(1H, m, O–C–C–C–CH–N), 6.364 (1H, t, J ¼ 6:8 Hz, H1 ), 6.728 (4H, d, J ¼ 8:7 Hz,
ArH), 7.183 (8H, m, ArH), 7.348 (1 H, d, J ¼ 7:3 Hz, ArH), 7.664 (1H, s, H8), 8.550
(1H, s, 2-N@CHN); FAB-MS: m=z 797 (M + HþÞ.
3.6. 30-O-[N,N-Diisopropylamino-(2-cyanoethoxy)phosphinyl]-50-O-(4,40-dimethoxy-
triphenylmethyl)-1-(3,4-diacetoxybutyl)-N2-dimethylaminomethylene-20-deoxyguano-
sine (16)
50-O-(4,40-Dimethoxy-triphenylmethyl)-1-(3,4-diacetoxybutyl)-N2-dimethylami-
nomethylene-20-deoxyguanosine (15) (398 mg, 0.5 mmol) was co-evaporated with dry
toluene (3 Â 4 mL) and pyridine (3 Â 4 mL). The residue was re-dissolved in dry
CH2Cl2 (10 mL) and treated with triethylamine (0.56 mL) and in one portion 2-cy-
anoethyl diisopropylchlorophosphoramidite (208 lL). The reaction mixture was stir-
red at room temperature for 2 h under nitrogen and then diluted with CH2Cl2
(40 mL). The resulting solution was washed with saturated NaHCO3 solution then
water, dried (Na2SO4Þ, and concentrated in vacuo to give the desired product 16
1
1
(423 mg, 85%). H NMR(CDC1 , 250 MHz) d H NMR(CDC1 , 250 MHz) d,
ppm: 1.205 (12H, m, N(C(CH3Þ Þ3 ), 1.981 (2H, m, 1-C–C–CH2–C–N), 2.004 (3H,
3
2
2
s, CH3COO–), 2.017 (3H, s, CH3COO–), 2.478 (2H, m, H02), 2.592 (2H, m,
CH2CN), 3.105 (3H, s, NCH3), 3.141 (3H, s, NCH3), 3.303 (2H, m, H05), 3.659
(5H, m, N(CH (CH3Þ Þ , H04), P–O–CH2CCN), 3.752 (6H, s, CH3OAr), 4.067 (1H,
2
2
m, O–C–C(O)H–C–C–N), 4.252 (2H, m, O–CH2–C–C–C–N), 4.379 (1H, m, O–C–
0
C–C–CH–N), 4.669 (1H, m, H3 ), 5.267 (1H, m, O–C–C–C–CH–N), 6.336 (1H, t,
0
J ¼ 6:8 Hz,H1 ), 6.771 (4H, m, ArH), 7.197 (8H, m, ArH), 7.361 (1H, d, J ¼ 7:3 Hz,
ArH), 7.691 (1H, s, H8), 8.534 (1H, s, 2-N@CHN); 31P NMR(CDC1 3, 250 MHz) d,
ppm: 149.952, 150.070; FAB-MS: m=z 997.5 (M + HþÞ.
3.7. 30; 50-O-Bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-20-deoxyguanosine
(12)
A mixture of 30; 50-O-bis(tert-butyldimethylsilyl)-1- (3,4-diacetoxybutyl)-N2-dime-
thylaminomethylene-20- deoxyguanosine (11) (73 mg, 100 lmol) and 2 mL 20% meth-
anolic ammonia was sealed in a flask and heated at 50 °C overnight. The reaction
mixture was evaporated under reduced pressure, and the residue was purified by
1
chromatography to give the desired product 12 as a yellow solid (54 mg, 93%). H
NMR(CD 3OD, 300 MHz) d, ppm: 0.077 (6H, s, Si–(CH3Þ ), 0.095 (6H, s, Si–
2
(CH3Þ ), 0.905 (9H, s, C–(CH3Þ ), 0.913(9H, s, C–(CH3Þ ), 1.686 (1H, m, 1-C–C–
2
3
3
CH–C–N), 1.924 (1H, m, 1-C–C–CH–C–N), 2.389 (2H, m, H02), 3.491 (3H, m,
O–CH2–C–C–C–N, O–C–C(O)H–C–C–N), 3.761 (2H, d, J ¼ 3:6 Hz, H05), 3.983
(1H, dd, J ¼ 3:4, 6.6 Hz, H04), 4.373 (H, m, C–C–C–CH–N), 4.559 (2H, m, H03, C–
0
C–C–CH–N), 6.321 (1H, t, J ¼ 6:6 Hz, H1 ), 7.876 (1H, s, H8), 8.581 (1H, s,
2-N@CHN); FAB-MS: m=z 606.2 (M + Naþ), 584.2 (M + Hþ).

Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
147
3.8. 3-(30; 50-O-Bis(tert-butyldimethylsilyl)-20-deoxy-b-
hydroxy-5,6,7,8-tetrahydropyrimido[1,2- a]purine-10(3H)-one (4) and 3-(30,50-O-
bis(tert-butyldimethylsilyl)-20-deoxy-b-
-erythro-pentofuranosyl)-6-methoxy-5,6,7,
D-erythro-pentofuranosyl)-6-
D
8-tetrahydropyrimido-[1,2-a]purine-10(3H)-one (13)
To a solution of 30; 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-20-
deoxyguanosine (12) (20 mg, 0.034 mmol) in MeOH (1.2 mL) was added a solution
of NaIO4 (500 lL, 0.1 M) in water. The reaction mixture was shaken at room tem-
perature for 1 h and then diluted with CH2Cl2 (10 mL). The organic layer was sep-
arated, washed with water (5 mL), dried (Na2SO4Þ, and concentrated in vacuo.
The residue was purified by silica gel column chromatography (CH2Cl2/MeOH:
25/1) to afford the products 4 (5 mg, 26.7%) and 13 (10 mg, 52%). Compound
1
(4): H NMR(DMSO-d , 300 MHz) d, ppm: 0.02 (6H, s, Si–(CH3Þ ), 0.08 (6H,
6
2
s, C–(CH3Þ ), 0.92 (18H, s, Si–(CH3Þ ), 1.733 (H, m, H7aÞ, 1.950 (H, m, H7b),
3
2
0
0
0
2.215 (1H, m, H2 a), 2.633 (1H, m,H2 b), 3.455 (H, m, H8aÞ, 3.651 (2H, m, H5 )
0
3.793 (H, m, H4 ), 4.408 (H, m, H8b), 4.470 (H, m,H3 ), 4.951 (H, m, H6), 5.927
0
(H, dd, J ¼ 1:5, 4.8 Hz, disappears after D2O is added, 6-OH), 6.096 (H, m,
þ
0
H1 ), 7.888 (H, S, H2), 8.410 (H, m, H5). FAB-MS: m=z 574 (M + Na ), 552
þ
1
(M + H Þ; compound (13): H NMR(DMSO-d , 300 MHz) d, ppm: 0.024 (6H,
6
s, Si–(CH3Þ ), 0.089 (6H, s, C–(CH3Þ ), 0.852 (18H, s, Si–(CH3Þ ), 1.743 (H,
2
3
2
0
m, H7a), 2.116 (H, m, H7b), 2.236 (1H, m, H2 a), 2.635 (1H, m, H2 b), 3.291
0
0
(3H, m, OCH3), 3.306 (1H, m, H8a), 3.653 (2H, m, H5 ), 3.801 (H, m, H4 ),
0
0
0
4.470 (H, m, H8b), 4.486 (H, m, H3 , 4.496 (H, m, H6), 6.111 (H, m, H1 ), 7.923
(H, S, H2), 8.824 (H, d, J ¼ 3:9 Hz, NH5). FAB-MS: m=z 566 (M + Hþ). The com-
pounds 4 and 13 were identical in spectroscopic data with the previously prepared
sample [16].
3.9. 3-(30,50-O-Bis(tert-butyldimethylsilyl)-20-deoxy-b-
D-erythro-pentofuranosyl)-
6-hydroxy-5,6,7,8-tetrahydropyrimido-[1,2-a]purine-10(3H)-one (4)
To a solution of 30, 50-O-bis(tert-butyldimethylsilyl)-1-(3,4-dihydroxybutyl)-20-de-
oxyguanosine (12) (20 mg, 0.034 mmol) in CH3CN (1.2 mL) was added a solution of
NaIO4 (500 lL, 0.1 M) in water. The reaction mixture was shaken at room temper-
ature for 1 h and then CH2Cl2 (10 mL) was added. The organic layer was separated,
washed with water (5 mL), dried (Na2SO4), and concentrated in vacuo. The residue
was purified by silica gel column chromatography (CH2Cl2/MeOH: 25/1) to afford
product 4 (15.3 mg, 82%).
Acknowledgments
This work was supported by a grant (CA47995) from the National Cancer Insti-
tute, National Institutes of Health. We thank Mr. Robert Rieger and Ms. Avalyn
Lewis for mass spectral data, and Maryann Wente for oligomer purification.

148
Y. Huang et al. / Bioorganic Chemistry 31 (2003) 136–148
References
[1] R.O. Beauchamp Jr., D.A. Andjelkovich, A.D. Kligerman, K.T. Morgan, H.dÕA. Heck, A critical
review of the literature of acrolein toxicity, in: CRC Critical Reviews in Toxicology, vol. 14, CRC
Press, Cleveland, 1985, pp. 309–380.
[2] International Agency for Research on Cancer, IARC Monograph on the Evaluation of the
Carcinogenic Risk of Chemicals to Human Beings, vol. 26, IARC, Lyon, France, 1981, pp. 165–202.
[3] F.L. Chung, R. Young, S.S. Hecht, Formation of cyclic 1,N2-propanodeoxyguanosine adducts in
DNA upon reaction with acrolein or crotonaldehyde, Cancer Res. 44 (1984) 990–995.
[4] R. Shapiro, Reactions with purines and pyrimidines, Ann. N. Y. Acad. Sci. 163 (Art. 2) (1969)
624–632.
[5] (a) B. Singer, H. Bartsch, The Role of Cyclic Nucleic Acid Adducts in Carcinogenesis and Mutagene,
IARC Scientific Publications 70, International Agency for Research on Cancer, Lyon, France, 1986;
(b) B. Singer, D. Grunberger, Molecular Biology of Mutagens and Carcinogens, Plenum Press, New
York, 1983;
(c) A.K. Basu, J.M. Essigmann, Site-specifically modified oligodeoxynucleotides as probes for the
structural and biological effects of DNA-damaging agents, Chem. Res. Toxicol. 1 (1988) 1–18.
[6] S. Khullar, C.V. Varaprasad, F.J. Johnson, Postsynthetic generation of a major acrolein adduct of 20-
deoxyguanosine in oligomeric DNA, J. Med. Chem. 42 (1999) 947–950.
[7] L.V. Nechev, C.M. Harris, T.M. Harris, Synthesis of nucleosides and oligonucleotides containing
adducts of acrolein and vinyl chloride, Chem. Res. Toxicol. 13 (2000) 421–429.
[8] L.V. Nechev, I. Kozekov, C.M. Harris, T.M. Harris, Stereospecific synthesis of oligonucleotides
containing crotonaldehyde adducts of deoxyguanosine, Chem. Res. Toxicol. 14 (2001) 1506–1512.
[9] H. Wang, C.J. Rizzo, Stereocontrolled syntheses of all four stereoisomeric 1,N2-deoxyguanosine
adducts of the lipid peroxidation product trans-4-hydroxynonenal, Org. Lett. 3 (2001) 3603–3605.
[10] C. De los Santos, T. Zaliznyak, F. Johnson, NMRcharacterization of a DNA duplex containing the
major acrolein-derived deoxyguanosine adduct-c-OH-1,N2-propano-20-deoxyguanosine, J. Biol.
Chem. 276 (12) (2001) 9077–9082.
[11] H. Mao, N.C. Schnetz-Boutaud, J.P. Weisenseel, L.J. Marnett, M.P. Stone, Duplex DNA catalyzes
the chemical rearrangement of a malondialdehyde deoxyguanosine adduct, Proc. Natl. Acad. Sci.
USA 96 (1999) 6615–6620.
[12] I.Y. Yang, M. Hossain, H. Miller, S. Khullar, F. Johnson, A. Grollman, M. Moriya, Responses to
the major acrolein-derived deoxyguanosine adduct in Escherichia coli, J. Biol. Chem. 276 (2001)
9071–9076.
[13] L.A. VanderVeen, M.F. Hashim, L.V. Nechev, T.M. Harris, C.M. Harris, L.J. Marnett, Evaluation
of the mutagenic potential of the principal DNA adduct of acrolein, J. Biol. Chem. 276 (2001)
9066–9070.
[14] J.R. Carmical, M. Zhang, L. Nechev, C.M. Harris, T.M. Harris, R.S. Lloyd, Mutagenic potential of
guanine N2 adducts of butadiene mono- and diolepoxide, Chem. Res. Toxicol. 13 (2000) 18–25.
[15] M. Kanuri, I.G. Minko, L.V. Nechev, T.M. Harris, C.M. Harris, R.S. Lloyd, Error prone translesion
synthesis past-hydroxypropano deoxyguanosine, the primary acrolein-derived adduct in mammalian
cells, J. Biol. Chem. 277 (2002) 18257–18265.
[16] Y. Huang, F. Johnson, Regioisomeric synthesis and characteristics of the a-hydroxy-1,N2-
propanodeoxyguanosine, Chem. Res. Toxicol. 15 (2) (2002) 236–239.
[17] Y. Huang, F. Johnson, Regioselective 1-alkylation of 20-deoxyguanosine, Nucleosides and Nucleo-
tides 21 (2002) 435–447.
[18] S.J. Kim, H.K. Jajoo, H.Y. Kim, L. Zhou, P. Horton, C.M. Harris, T.M. Harris, An efficient route to
N6ꢀ deoxyadenosine adducts of diol epoxides of carcinogenic polycyclic aromatic hydrocarbons,
Bioorg. Med. Chem. 3 (1995) 811–822.
[19] L.V. Nechev, I.D. Kozekov, A.K. Brock, C.J. Rizzo, T.M. Harris, DNA adducts of acrolein: site-
specific synthesis of an oligodeoxynucleotide containing 6-hydroxy-5,6,7,8-tetrahydropyrimido[1,
2-a]purin-10(3H)-one, an acrolein adduct of guanine, Chem. Res. Toxicol. 15 (2002) 607–613.