Fluorinated alcohol mediated control over cis vs trans opening of benzo[α]pyrene-7,8-diol 9,10-epoxides at C-10 by the exocyclic amino groups of O6-allyl protected deoxyguanosine and of deoxyadenosine

Article abstract of DOI:10.1021/jo070303c

(Chemical Equation Presented) A detailed study was carried out on the stereoselective control of cis- vs trans-opening of (±)-7β,8α- dihydroxy-9β,10β-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene {B[a]P DE-1 (1)} and (±)-7β,8α-dihydroxy-9α,10α-epoxy-7,8,9, 10-t

Full text of DOI:10.1021/jo070303c

Fluorinated Alcohol Mediated Control over Cis vs Trans Opening of
Benzo[a]pyrene-7,8-diol 9,10-Epoxides at C-10 by the Exocyclic
Amino Groups of O6-Allyl Protected Deoxyguanosine and of
Deoxyadenosine
Haruhiko Yagi and Donald M. Jerina*
Laboratory of Bioorganic Chemistry, National Institute of Diabetes and DigestiVe and Kidney Diseases,
The National Institutes of Health, DHHS, Bethesda, Maryland 20892
dmjerina@nih.goV
ReceiVed March 5, 2007
A detailed study was carried out on the stereoselective control of cis- vs trans-opening of (()-7â,8R-
dihydroxy-9â,10â-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene {B[a]P DE-1 (1)} and (()-7â,8R-dihydroxy-
9R,10R-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene {B[a]P DE-2 (2)} at C-10 by the exocyclic amino groups
of protected purine nucleosides in the fluorinated alcohols trifluoroethanol (TFE), hexafluoropropan-2-
ol (HFP), and perfluoro-tert-butanol (PFTB). Addition of the 2-amino group of O6-allyl-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyguanosine (3) and of the 6-amino group of 3′,5′-di-O-(tert-butyldimethylsilyl)-
2′-deoxyadenosine (4) occurs at C-10 of the epoxides. The observed cis:trans ratio for the reaction of
DE-1 (1) in the presence of 5 equiv of 3 over the range of 10-250 equiv of fluorinated alcohol varied
from 53:47 to 87:13 for TFE, 60:40 to 92:8 for HFP, and 52:48 to 73:27 for PFTB. The corresponding
ratios for DE-2 (2) varied from 22:78 to 72:28 for HFP under the same set of conditions. In contrast, the
corresponding ratios for DE-2 (2) remained unchanged (∼40:60) for TFE and for PFTB over the range
of 25-250 molar equiv. Unlike the addition of the dGuo reactant 3, the corresponding addition of the
dAdo reactant (4) to the DEs (1 or 2) in over 25 molar equiv of TFE occurred highly stereoselectively
to afford only cis adducts for both DEs. A highly efficient HPLC separation of dGuo adduct diastereomers
derived from DE-2 (2) was developed using acetone as a modifier in CH2Cl2 or in n-hexane. Through
the use of varying molar ratios of the different fluorinated alcohols described above and the newly
developed HPLC separation method, the four possible phosphoramidites (cis/trans, R/S) of the B[a]P
DE-2 N2-dGuo adducts can be prepared in an efficient fashion on gram scale for use in oligonucleotide
synthesis.
Introduction
diastereomers in which the benzylic hydroxyl group is either
cis (DE-1, 1) or trans (DE-2, 2) to the epoxide oxygen (Figure
1); each diastereomer exists as a pair of enantiomers (not
shown).2 Among these four possible stereoisomeric DEs, (+)-
(7R,8S,9S,10R)-DE-2 is the major DE metabolite formed from
The potent carcinogenic polycyclic aromatic hydrocarbon
benzo[a]pyrene (B[a]P) is a widespread environmental pollutant
that is metabolically activated to bay-region diol epoxides
(DEs).1 These DEs are formed from trans-7,8-dihydrodiols as
(1) Committee on the Biologic Effects of Atmospheric Pollutants.
Particulate Polycyclic Organic Matter; National Academy of Science:
Washington, DC, 1972.
* Address for correspondence: Donald M. Jerina, Chief, Section on Oxidation
Mechanisms, The National Institutes of Health, NIDDK, Building 8A, Room
1A01, Bethesda, Maryland 20892. TEL: 301 496 4560. FAX: 301 402 0002.
10.1021/jo070303c This article not subject to U.S. Copyright. Published 2007 American Chemical Society
Published on Web 07/03/2007
J. Org. Chem. 2007, 72, 6037-6045 6037

Yagi and Jerina
processing enzymes such as human DNA polymerases,7 topo-
isomerases I8 and II,9 vaccinia topoisomerase I,10 Werner
helicase,11 and HIV-1 integrase12 is a rapidly developing area
of research.
Direct reaction of the exocyclic amino groups of appropriately
protected purine nucleosides with the DEs provides a convenient
route to the synthesis of adducts. Such modified nucleosides
are converted to the corresponding phosphoramidites, which can
be site- and stereospecifically incorporated into the target
oligonucleotides. We have established highly efficient conditions
for the key adduct-forming reaction by utilizing O6-allyl-3′,5′-
di-TBDMS-protected dGuo (3)13 and 3′,5′-diTBDMS-protected
dAdo (4).14 Blocking of the purine hydroxyl groups as the very
nonpolar TBDMS ethers greatly increases the solubility of the
nucleosides in organic solvents. Blocking of the O6 carbonyl
group of dGuo with the allyl group not only prevents adduct
formation at this position but also enhances the nucleophilic
reactivity of the N2-amino group. The reaction of B[a]P DE-1
(1) and DE-2 (2) with 3 in dimethylacetamide (DMA) at
100 °C for 2 h formed a mixture of cis and trans adducts (∼50%
yield) with cis:trans ratio of ∼1:1.13 No reaction occurred
between 1 or 2 and the TBDMS-protected dAdo (4) in DMA,13
possibly because the N6-amino group in 4 is less nucleophilic
than the N2-amino group in 3.
FIGURE 1. Structure of benzo[a]pyrene (B[a]P) with the bay region
and key terminal benzo-ring indicated, structures of B[a]P 7,8-diol 9,-
10-epoxide DE-1 (1) and DE-2 (2) diol epoxide diastereomers, the dGuo
reactant O6-allyl-3′,5′-di-O-(tert-butyldimethylsilyl)-2′-deoxyguanosine
(3), and the dAdo reactant 3′,5′-di-O-(tert-butyldimethylsilyl)-2′-
deoxyadenosine (4). Only one of the two possible enantiomers of each
DE is shown.
Fluorinated alcohols such as trifluoroethanol (TFE), hexafluo-
ropropan-2-ol (HFP), and perfluoro-tert-butyl alcohol (PFTB)
have a dramatic catalytic effect on the addition reactions between
blocked purines and DEs. Their combination of low nucleo-
philicity, high hydrogen bonding donor strength, high polarity,
and high ionizing power make them ideal solvents for addition
reactions that proceed through carbocation intermediates.15 Thus,
reactions of B[a]P DE-1 (1) and DE-2 (2) with 3 in a large
excess (1040 molar equiv) of TFE16 proceeded at room
temperature to give mixtures of cis and trans adducted products
with cis:trans ratios of 85:15 in an overall yield of 65% (DE-1)
and 40:60 in an overall yield of 43% (DE-2). Unlike the
corresponding reaction in DMA, which yielded no adducts,
addition of 4 to 1 or 2 afforded cis-opened adducts, albeit in
B[a]P in mammals2 and is by far the most carcinogenic in
several animal models.3 The major binding site of this ultimate
carcinogen to DNA occurs at the exocyclic amino group of
deoxyguanosine (dGuo) and to a lesser extend at the exocyclic
amino group of deoxyadenosine (dAdo).4 Adduct formation is
thought to be a key step in tumor initiation. Since even minor
adducts may have significant biological effects, there has been
a continued need for small oligonucleotides containing minor
as well as major adducts of B[a]P DEs at specific sites in defined
sequence contexts. These DE adducted oligonucleotides have
been used in structural studies by two-dimensional NMR5 and
X-ray crystallographic analysis.6 Use of such modified oligo-
nucleotides as probes for the study of the mechanism of DNA
(7) (a) Frank, E. G.; Sayer, J. M.; Kroth, H.; Ohashi, E.; Ohmori, H.;
Jerina, D. M.; Woodgate, R. Nucleic Acid Res. 2002, 30, 5284-5292. (b)
Graziewicz, M. A.; Sayer, J. M.; Jerina, D. M.; Copeland, W. C. Nucleic
Acids Res. 2004, 32, 397-405.
(8) Pommier, Y.; Laco, G. S.; Kohlhagen, G.; Sayer, J. M.; Kroth, H.;
Jerina, D. M. Proc. Acad. Sci. U.S.A. 2000, 97, 10739-10744.
(9) Khan, Q. A.; Kohlhagen, G.; Marshall, R.; Austin, C. A.; Kalena,
G. P.; Heiko, K.; Sayer, J. M.; Jerina, D. M.; Pommier, Y. Proc. Acad. Sci.
U.S.A. 2003, 100, 12498-12503.
(2) Thakker, D. R.; Yagi, H.; Levin, W.; Wood, A. W.; Conney, A. H.;
Jerina, D. M. In BioactiVation of Foreign Compounds; Anders, M. W., Ed.;
Academic Press: New York, 1985; pp 177-242.
(3) (a) Buening, M. K.; Wislocki, P. G.; Levin, W.; Yagi, H.; Thakker,
D. R.; Akagi, H.; Koreeda, M.; Jerina, D. M. Proc. Natl. Acad. Sci. U.S.A.
1978, 75, 5358-536. (b) Slaga, T. J.; Bracken, W. J.; Gleason, G.; Levin,
W.; Yagi, H.; Jerina, D. M.; Conney, A. H. Cancer Res. 1979, 39, 67-71.
(4) (a) Jerina, D. M.; Chadha, A.; Cheh, A. M.; Schurdak, M. E.; Wood,
A. W.; Sayer, J. M. In Biological ReactiVe Intermediates IV; Witmer, C.
M., Snyder, R., Jelow, D. J., Kalf, G. F., Kocsis, J. J., Sipes, I. G., Eds.;
Molecular and cellular effects and their impact on human health. Plenum:
New York, 1991; pp 533-553. (b) Dipple, A. In DNA Adducts: Identifica-
tion and Biological Significance; Hemminiki, K., Dipple, A., Shuker, D.
E. G., Kadlubar, F. F., Segerbaeck, D., Bartsch, H., Eds.; IARC Publications
No. 125: Lyon, 1994; pp 107-129.
(5) For leading references see: (a) Lin, C. H.; Huang, X.; Korbanovskii,
A.; Hingerty, B. E.; Amin, S.; Broyde, S.; Geacintov, N. E.; Patel, D. J. J.
Mol. Biol. 2001, 306, 1059-1080. (b) Pradhan, P.; Tirumala, S.; Liu, X.;
Sayer, J. M.; Jerina, D. M.; Yeh, H. J. C. Biochemistry 2001, 40, 5870-
5881.
(10) (a) Yakovleva, L.; Tian, L.; Sayer, J. M.; Kalena, G. P.; Kroth, H.;
Jerina, D. M.; Shuman, S. J. Biol. Chem. 2003, 278, 42170-42177. (b)
Yakovleva, L.; Handy, C. J.; Sayer, J. M.; Pirrung, M.; Jerina, D. M.;
Shuman, S. J. Biol. Chem. 2004, 279, 23335-23342. (c) Yakovleva, L.;
Handy, C. J.; Yagi, H.; Sayer, J. M; Jerina, D. M.; Shuman, S. Biochemistry.
2006, 45, 7644-7653.
(11) (a) Driscoll, H. C.; Matson, S. W.; Sayer, J. M.; Kroth, H.; Jerina,
D. M.; Brosh, M., Jr. J. Biol. Chem. 2003, 278, 41126-41135. (b)
Choudhary, S.; Doherty, K. M.; Handy, C. J.; Sayer, J. M.; Yagi, H.; Jerina,
D. M; Brosh, R. M., Jr. J. Biol. Chem. 2006, 281, 6000-6009.
(12) Johnson, A. A.; Sayer, J. M.; Yagi, H.; Kalena, G. P.; Amin, R.;
Jerina, D. M.; Pommier, Y. J. Biol. Chem. 2004, 279, 7947-7965.
(13) Kroth, H.; Yagi, H.; Seidel, A.; Jerina, D. M. J. Org. Chem. 2000,
65, 5558-5564.
(14) Pilcher, A. S.; Yagi, H.; Jerina, D. M. J. Am. Chem. Soc. 1998,
120, 3520-3521.
(15) Eberson, L.; Hartshorn, M. P.; Persson, O.; Radner, F. Chem.
Commun. 1996, 2105-2112.
(6) Ling, H.; Sayer, J. M.; Polsky, B. S.; Yagi, H.; Boudsocq, F.;
Woodgate, R.; Jerina, D. M.; Yang, W. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 2265-2269.
(16) Ramesha, A R.; Kroth, H.; Jerina, D. M. Tetrahedron Lett. 2001,
42, 1003-1005.
6038 J. Org. Chem., Vol. 72, No. 16, 2007

Opening of Benzo[a]pyrene-7,8-diol 9,10-Epoxides
TABLE 1. Fluorinated Alcohol Mediated Synthesis of Cis- and
Trans-Opened N2-dGuo Adducts from B[a]P DE-1 (1)a
cis:trans
adduct
ratiob
yield
dGuo
yield
molar
equiv
temp
(°C)
time
(min)
solvent
solvent
TFE
adductsc
adductsc
250
200
150
100
50
rt
rt
rt
rt
rt
70
70
rt
rt
rt
rt
rt
70
70
rt
rt
rt
70
70
70
70
10
10
15
15
30
30
30
10
15
15
30
30
30
30
10
10
30
30
30
30
30
87:13
83:17
79:21
71:29
65:35
58:42
53:47
92:8
68
64
62
63
54
50
45
70
68
75
68
62
60
60
70
72
68
68
65
63
60
4.0
2.5
1.5
1.7
2.0
10
12
16
18
15
16
20
15
15
ndd
nd
nd
nd
nd
nd
nd
25
10
HFP
250
200
150
100
50
92:8
90:10
85:15
75:25
68:32
60:40
73:27
73:27
72:28
69:31
62:38
56:44
52:48
25
10
PFTB
250
200
150
100
50
FIGURE 2. Stereoselectivity of fluorinated alcohol-mediated cis- and
trans-opening of B[a]P DE-1 (1) by the dGuo reactant O6-allyl-3′,5′-
O-(tert-butyldimethylsilyl)-2′-deoxyguanosine (3) in a molar ratio of
1:5: b, TFE; O, HFP; 1, PFTB.
25
10
low yield (22-33%). The major side products were from solvent
addition and accounted for as much as 50% of the starting
material.
a All reactions were run with a molar ratio of the B[a]P DE-1 (1) to
dGuo nucleoside (3) of 1:5. b The ratio was determined by HPLC (280 nm)
as disilyl trihydroxy derivatives. c Actual isolated yield of combined cis
and trans adducts (10-50 mg scale). d Not detectable.
A detailed study of the addition reaction of 3 with the less
reactive DEs of benzo[c]phenanthrene (B[c]Ph) revealed that
the cis:trans adduct ratios were highly dependent on the amount
of solvent HFP used, when less than ∼200 molar equiv of TFE
and 5 molar equiv of 3 were used.17 In the present report, we
describe optimization of this addition reaction in fluorinated
alcohols for B[a]P DEs (1 and 2) with 3 and 4 to reduce
undesired solvent adduct formation and to control the cis vs
trans stereochemistry of addition over a wide range of cis/trans
product ratios. We also report a new and highly efficient HPLC
separation of the 10R and 10S diastereomeric adducts that are
produced on reaction of racemic B[a]P DEs with 3.
for the solvent-free reaction for 1 was 50:50.20 This strongly
contrasts with the corresponding solvent-free reaction of DEs
of B[c]Ph with 3, which produced the corresponding trans
adduct almost exclusively.17 In all cases, the reaction was
complete at room temperature within 10-30 min at high molar
ratios of fluorinated alcohol to DE-1 (1) (>50 molar equiv of
TFE or HFP or >150 molar equiv for PFTB). When less than
the above amount of solvent was used, the reaction required
heating at 70 °C for 30 min in a sealed glass tube to obtain a
clear solution, presumably due to the reduced solubility of the
reactants. Under these conditions, the yields of the desired
adducts were as high as 75%, and the side product of solvent
addition was reduced to less than 12% for the reaction in TFE
and less than 20% in HFP. No solvent adducts were formed in
PFTB (Table 1).
B[a]P DE-2-dGuo Reactions. Quite intriguingly, unlike the
reaction of DE-1 (1), no cis/trans ratio change was observed
for the corresponding addition reaction of 3 with DE-2 (2) in
TFE (cis/trans ratio ∼40/60) and in PFTB (cis:trans ratio ∼
40:60) over the range of 25-250 molar equiv (Figure 3 and
Table 2). However, a distinct dependency was observed for the
reaction of 3 with 2 in HFP over a range of 10-250 molar
equiv: the cis:trans ratio varied from 22:78 to 72:28 with
increasing HFP. The cis/trans ratio previously observed for the
solvent-free reaction20 was 15:85, which is compatible with the
curve obtained in this study, as shown in Figure 3.
Results and Discussion
Direct opening of B[a]P DEs (1 and 2) was investigated using
TFE (pKa ) 12.4),18 HFP (pKa ) 9.3),18 and PFTB (pKa )
5.2)18b as solvents. In all cases, the DEs were added to a clear
solution of the protected dGuo 3 in the solvent. The addition
reaction was complete as soon as the DEs dissolved completely,
and a clear solution was obtained. This order of addition of the
substrate DEs to the reactant 3 is very important to reduce
competitive formation of solvent adducts.19 Unnecessarily
prolonged (overnight) treatment in TFE as previously reported16
also decreases the yield considerably due to decomposition of
the desired products.
B[a]P DE-1-dGuo Reactions. The observed cis:trans adduct
ratio for DE-1 (1) varied, as shown in Figure 2 and Table 1
over the range of 10-250 molar equiv of each of the three
alcohols. The greatest ratio change was observed for HFP (60:
40 to 92:8) followed by TFE (53:47 to 87:13) and PFTB (52:
48 to 73:27). Interestingly, the previously observed cis/trans ratio
The reactivity of B[a]P DE-2 (2) with 3 in fluorinated
alcohols under comparable conditions is appreciably lower than
that of DE-1 (1). Thus, reaction of DE-2 (2) with 3 was complete
at room temperature within 6 h in the presence of >150 molar
equiv of TFE, within 1 h for >50 molar equiv of HFP, or within
2 h for > 100 molar equiv of PFTB. When less solvent was
used, the reaction required heating at 45-70 °C for 1 to 21 h
(17) Yagi, H.; Ramesha, A. R.; Kalena, G.; Sayer, J. M.; Kumar, S.;
Jerina, D. M. J. Org. Chem. 2002, 67, 6678-6689.
(18) (a) Matesich, M. A.; Knoefel, J.; Evans, D. F.; J. Phys. Chem. 1973,
77, 366-369. (b) Filer, R.; Schure, R. M. J. Org. Chem. 1967, 32, 1217-
1219.
(19) In the previous paper16 DEs and the reactant 3 or 4 were added
simultaneously.
(20) Ramesha, A. R.; Kroth, H.; Jerina, D. M. Org. Lett. 2001, 3, 531-
533.
J. Org. Chem, Vol. 72, No. 16, 2007 6039

Yagi and Jerina
opened adducts. As the amount of solvent is decreased relative
to the other reactants, the medium becomes less polar, and the
portion of the reaction that proceeds through an SN1 mechanism
must compete with that through an SN2 mechanism, which
increases the amount of trans-opened adducts. Unlike the
solvent-free reaction of the less reactive B[c]Ph DEs, which
produced almost exclusively the trans-opened adduct through
an SN2 mechanism,17 the corresponding solvent-free reaction
of B[a]P DEs proceeds substantially through an SN1 mechanism.
This is consistent with formation of a much more stable
carbocation at the C-10 position of B[a]P DEs compared with
a much less stable carbocation at the C-1 position of the B[c]-
Ph DEs.21 A plausible mechanism for the observed dependency
of adduct ratio on the amount of solvent observed for B[a]P
DEs in this study is similar to that given for the corresponding
reaction of B[c]Ph DEs.17 Since exclusive formation of cis-
opened adducts was observed for the reaction of 4 with B[a]P
DEs, as will be discussed later, the rate of capture of the two
possible carbocation conformations might also be dependent on
the nature of the nucleophile used.
FIGURE 3. Stereoselectivity of fluorinated alcohol-mediated cis- and
trans-opening of B[a]P DE-2 (2) by the dGuo reactant O6-allyl-3′,5′-
O-(tert-butyldimethylsilyl)-2′-deoxyguanosine (3) in a molar ratio of
1:5: b, TFE; O, HFP; 1, PFTB.
Separation and Identification of B[a]P DEs-dGuo Reac-
tion Products. The solvent adducts (9a, 10a, and 11) were also
acetylated. Relative cis stereochemistry of the 9,10-substituents
in the acetylated adducts 9b, 10b, and 12 was assigned by
TABLE 2. Fluorinated Alcohol Mediated Synthesis of Cis- and
Trans-Opened N2-dGuo Adducts from B[a]P DE-2 (2)a
cis:trans
adduct
ratiob
yield
dGuo
yield
1
comparison of their H NMR spectra with those of a known,
molar
equiv
temp
(°C)
time
(h)
solvent
solvent
TFE
adductsc
adductsc
related compound.22
We have developed a highly efficient separation of the
diastereomeric adducts by HPLC on a silica column (16 × 250
mm). As shown in Figure 4, using 3% acetone in methylene
chloride as solvent, a baseline separation of a mixture of 7a
and 7b (50 mg/injection) was easily achieved. An even better
separation was achieved for 8a and 8b. Thus, a complete
baseline separation of as much as 200 mg/injection of 8a and
8b was achieved using 12% acetone in n-hexane as solvent. A
previous separation of these diastereomers using CH2Cl2-
MeOH (98:2) on an Axxiom silica column (9.5 × 250 mm)
allowed a maximum of 3 mg of sample per injection.13 Acetone
has rarely been used as an HPLC solvent, since it has rather
strong UV absorption below 300 nm. Fortunately, all of the
present adducts have a strong UV absorption at ∼340-370 nm
due to the pyrene chromophore, which can be used to monitor
the HPLC effluent in the presence of acetone. The high solubility
of these adducts in acetone is another advantage in that it
minimizes the volume of injections.
250
200
150
100
50
Rt
Rt
Rt
70
70
70
Rt
Rt
Rt
Rt
Rt
45
70
Rt
Rt
Rt
Rt
45
70
2
4
6
1
5
7
1
1
1
1
1.5
5
12
1
1
1
2
4
21
38:62
38:62
37:63
36:64
38:62
37:63
72:28
72:28
70:30
59:41
43:57
30:70
22:78
40:60
40:60
39:61
40:60
38:62
40:60
59
60
58
50
48
43
78
75
78
80
80
65
50
80
80
85
87
90
45
10
9
7
5
7
25
3
HFP
250
200
150
100
50
ndd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
25
10
PFTB
250
200
150
100
50
25
a All reactions were run with a molar ratio of the B[a]P DE-2 (2) to
dGuo nucleoside (3) of 1:5. b The ratio was determined by HPLC (280 nm)
as disilyl tracetoxy derivatives for the reaction in TFE and as disilyl
trihydroxy derivatives for the reaction in HFP and in PFTB. c Actual isolated
yield of combined cis and trans adducts (10-50 mg scale). d Not detectable.
1
A comparison of benzo-ring H NMR data (600 MHz) for
the triacetates of the O6-allyl protected cis- and trans-opened
DE-2 N2-dGuo adducts (7a,b and 8a,b) is given in Table 3.
Previously unassigned J9,10 for 8a,b13 is now clearly assignable
to be 5.5 Hz, the value of which is consistent with the previously
observed coupling constant for the structurally related acetylated
trans-opened DE-2 N2-dGuo adducts.22
in a sealed glass tube. Under these conditions, yields of desired
adducts were 43∼90% and formation of the solvent adduct 11
was observed only for the reaction in TFE (Table 2). For the
preparative scale reactions (1.2 g of 2; see Experimental
Section), the best yield (90%) was obtained when a 1:8 ratio of
2 to 3 was heated at 45 °C in PFTB (50 molar equiv) for 4-6
h in a sealed glass tube. The cis and trans mixed diastereomers
and byproduct solvent adducts (9-11) were separated by a
combination of column chromatography and HPLC, and the
individual acetylated diastereomers (5a, 5b, 6a, 6b, and 7a, 7b,
8a, 8b) were separated by HPLC (Scheme 1).
dAdo Adducts. Unlike the addition reaction of 1 and 2 with
the O6-allyl dGuo di-TBDMS ether (3), the corresponding
reaction of these DEs (1 and 2) with the dAdo di-TBDMS ether
(4) in TFE proceeded highly stereoselectively to give only cis
dAdo adducts 13a,b and 14a,b, respectively (>25 molar equiv
of TFE as solvent, Scheme 2). Despite the relatively low yields
(21) Jerina, D. M.; Lehr, R. E. In Microsomes and Drug Oxidation;
Ullrich, V., Roots, I., Hindebrandt, A. G., Estabrook, R. W., Conney, A.
H., Eds.; Pergamon Press: Oxford, 1977; pp 709-720.
(22) (a) Yagi, H.; Thakker, D. R.; Hernandez, O.; Koreeda, M.; Jerina,
D. M. J. Am. Chem. Soc. 1977, 99, 1604-1611. (b) Szeliga, J.; Dipple, A.
Chem. Res. Toxicol. 1998, 11, 1-11.
Mechanism of Addition Reaction. At solvent ratios where
the cis:trans adduct ratio remains constant, the reaction is
presumed to proceed largely by an SN1 mechanism, which is
consistent with the formation of substantial amounts of cis-
6040 J. Org. Chem., Vol. 72, No. 16, 2007

Opening of Benzo[a]pyrene-7,8-diol 9,10-Epoxides
SCHEME 1a
a Key: (i) fluorinated alcohols; (ii) Ac2O-DMAP-pyridine; (iii) HPLC. Only one of the two enanthiomers is shown for 9a, 9b, 10a, 10b, 11, and 12.
previously reported (22% for 1 and 33% for 2, respectively),16
this method provides an attractive source for the cis-dAdo
adducts, which are often difficult to synthesize. Surprisingly,
no reaction occurred between the DEs and 4 in HFP or in
PFTB.16
(1) via a multiple step synthesis.14 Although the yield for 2 in
the present study is comparable with that previously obtained
through an aminohydroxylation approach,14 the procedure and
workup for the present addition reaction are much simpler, and
the reaction is more robust.
Phosphoramidite Synthesis. In a previous study23 we
reported the synthesis of 5′-dimethoxytrityl (DMT)-protected
3′-phosphoramidite building blocks of N2-dGuo adducts derived
from cis- and trans-opening of B[a]P DE-1 (1) and DE-2 (2) at
C-10. Since the starting material for these syntheses is a racemic
DE, each of the cis- and trans-phosphoramidites consists of a
mixture of diastereomers. Their use in modified oligonucleotide
synthesis results in a mixture of 10S/10R constructs. Although
separation of the two diastereomeric adducted oligonucleotides
has generally been achieved, use of single phosphoramidites
diastereomers of known configuration greatly simplifies the
HPLC purification and characterization of the modified oligo-
In the present study, we attempted to modify previous reaction
conditions to see if it is possible to increase the yield of the
objective dAdo adducts. Thus, the reaction was carried out by
portionwise addition of DEs (1 or 2) to a solution of 4 (6 molar
equiv) in a much smaller amount of TFE (∼1/7 of the amount
previously used) at 45-60 °C. The reaction mixture was
acetylated and the products (Scheme 2, 13a,b and 9b from 1,
14a,b and 12 from 2, respectively) were separated by HPLC.
Under these conditions, we could successfully reduce the
formation of the unwanted solvent adducts (9b and 12) to less
than half (25%) and double the yield of the objective dAdo
adducts (40% for 1 and 60% for 2, respectively). The direct
addition reaction between DE-1 (1) and 4 is by far the simplest
method to obtain the cis-opened dAdo adduct. An earlier method
required the use of the cis-opened aminotriol derived from DE-1
(23) Kroth, H.; Yagi, H.; Sayer, J. M.; Kummer, S.: Jerina, D. M. Chem.
Res. Toxicol. 2001, 14, 708-719.
J. Org. Chem, Vol. 72, No. 16, 2007 6041

Yagi and Jerina
SCHEME 2a
a Key: (i) CF3CH2OH; (ii) Ac2O-DMAP-pyridine; (iii) HPLC. Only one of the two enanthiomers is shown for 9b and 12.
1
TABLE 3. Comparison of Benzo-Ring H NMR Data for the
nucleotides. In addition, if only one isomer is desired, the other
diastereomer reduces the effective yield and essentially wastes
half of the starting material. In general, separation of R/S pairs
of adducted oligonucleotides tends to become increasingly more
difficult as the length of the sequence increases. Thus, use of
diastereomerically pure phosphoramidites as building blocks for
O6-Allyl Protected Cis- and Trans-Opened N2-dGuo Adducts as
Their Disylyl Triacetates (600 MHz, Acetone-d6)a
H7 J7,8 H8 J8,9
H9
J9,10
H10
compound
(ppm) (Hz) (ppm) (Hz) (ppm) (Hz) (ppm)
(+)-7a (10S)
6.64
6.64
6.81
6.81
5.74
5.72
5.97
5.97
5.94
5.96
6.12
6.12
6.85
6.85
6.32b
6.32b
cis-dGuo O6-all-DE-2
(-)-7b (10R)
3.4
3.4
9.4
8.9
2.3
2.3
2.3
2.3
5.4
5.4
2.6
2.6
(-)-8a (10S)
trans-dGuo O6-all-DE-2
(+)-8b (10R)
a Assignments were confirmed by decoupling and COSY experiments.
b Previously reported13 chemical shifts for the H10 proton for (+)-8a (7.10
ppm) and for (+)-8b (7.23 ppm) were in error.
oligonucleotide synthesis is desirable for synthesis of longer
oligonucleotides.
Major improvements have been achieved for the desilylation
of the sugar hydroxyl groups in 7a, 7b, 8a, and 8b and the
successive introduction of the 5′-DMT protecting group on the
sugar (Scheme 3). Thus, a complete and nearly quantitative
desilylation (>97%) was achieved using 7% HF in pyridine
for 12 h at room temperature to give 15a,b and 16a,b instead
of using 2% HF in a mixture of acetonitrile and pyridine (4:
1).23 The assignment of the relative stereochemistry and the
absolute configuration of (+)-trans-16b were recently confirmed
by X-ray crystallographic analysis.24 The newly developed
method for the DMT protection of the 5′ sugar hydroxyl group
involved replacing THF with CH2Cl2 in the presence of DMT
fluoroborate.25 As before, formation of any side products was
prevented.17 Thus, 17a,b and 18a,b were obtained from 15a,b
and 16a,b, respectively, in nearly quantitative yield (>95%).
Finally, introduction of the phosphoramidite function at the
3′-sugar hydroxyl group in the 5′-DMT derivatives (17a,b and
18a,b) was achieved with 2-cyanoethyltetraisopropyl phos-
FIGURE 4. HPLC separation of the diastereomeric O6-allyl-protected
dGuo adduct disilyl triacetates derived from B[a]P DE-2 (2) on a
LiChrosorb-60 Si (16 × 250 mm) column. (A) Separation of the early
eluting (-)-cis-7b (10R) and the late eluting (+)-cis-7a (10S). (B)
Separation of the early eluting (-)-trans-8a (10S) and the late eluting
(+)-trans-8b (10R). See Supporting Information for details.
(24) Karle, I. L.; Yagi, H.; Sayer, J. M.; Jerina, D. M. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 1433-1438.
(25) Bleasdale, C.; Ellwood, S. B.; Golding, B. T. J. Chem. Soc. Perkin
Trans. 1990, 1, 803-805.
6042 J. Org. Chem., Vol. 72, No. 16, 2007

Opening of Benzo[a]pyrene-7,8-diol 9,10-Epoxides
SCHEME 3a
a Key: (i) 7% HF/pyridine; (ii) DMT+ BF4-CH2Cl2; (iii) 2-cyanoethyltetraisopropyl phosphoramidite.
phoramidite and N,N-diisopropylammonium tetrazolide in CH2-
Cl2 to give 19a,b and 20a,b, respectively, in 80-90% yield.
Each of the diastereomeric phosphoramidites thus obtained
consisted of a mixture of two peaks due to the asymmetry of
the phosphorus. Although these phosphorus diastereomers could
be separated by HPLC (see Experimental Section in the
Supporting Information), such separation is not necessary prior
to their use in oligonucleotide synthesis, as the resulting
phosphates are no longer chiral at phosphorus.
concentrations,26a where the two carbocation conformers are in
rapid equilibrium. This observation is consistent with the
speculation that the carbocation for DE-2 (2) undergoes con-
formational equilibration in these two fluorinated alcohols faster
than it is captured by the weakly nucleophilic O6-allyl dGuo
di-TBDMS ether (3). The only exception to this explanation is
found for the cis:trans N2-dGuo adduct ratio observed for the
reaction of DE-2 (2) (72:28, Table 2) in the presence of high
concentration of HFP (>200 molar equiv). The mechanism for
this unusually high cis adduct formation is presently unknown
and may involve solvent effects on the equilibrium between
the two carbocation conformations and/or the relative rates of
their capture by 3. It is also interesting to point out that exclusive
cis adduct formation was observed for the solvent adducts and
for the reaction of both DEs 1 and 2 with the dAdo di-TBDMS
ether (4) in TFE.
Conclusions
The present observations on the addition reactions of the
B[a]P DEs (1 and 2) are mechanistically similar to results
obtained for the hydrolysis of these DEs in water.26 The
predominant cis-N2-dGuo adduct formation from DE-1 (1) (87%
for TFE, 92% for HFP, 73% for PFTB, respectively, Table 1)
in the presence of high amounts of solvent (250 molar equiv)
is consistent with the predominant formation of cis-tetraol (85%)
on acid-catalyzed hydrolysis of DE-1 (1).26b This result suggests
that the same conformation of the carbocation intermediate from
DE-1 (1) is trapped by solvent water and by the O6-allyl dGuo
di-TBDMS ether (3) in the fluorinated alcohols used in the
present study. The cis:trans ratio of the N2-dGuo adducts
obtained from DE-2 (2) in TFE and PFTB (∼40:60, Table 2)
is also very similar to the cis:trans tetraol ratio (35:65) formed
from its acid-catalyzed hydrolysis at high chloride ion
Experimental Section
Caution: Benzo[a]pyrene-7,8-dihydrodiol and diol epoxides
DE-1 and DE-2 are mutagenic and carcinogenic and must be
handled carefully in accordance with NIH guidelines.27
Reaction of B[a]P DE-1 (1) with O6-Allyl-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyguanosine (3) in TFE. To a solution
of the O6-allyl dGuo di-TBDMS ether (3) (5 molar equiv) in TFE
(10-250 molar equiv) was added B[a]P DE-1 (1) (1 molar equiv),
and the reaction mixture was stirred at room temperature or at
(26) (a) Doan, L.; Yagi, H.; Jerina, D. M.; Whalen, D. L. J. Am. Chem.
Soc. 2002, 124, 14382-14387. (b) Doan, L.; Yagi, H.; Jerina, D. M.;
Whalen, D. L. J. Org. Chem. 2004, 69, 8012-8017.
(27) NIH Guidelines for the Laboratory Use of Chemical Carcinogens;
NIH Publication No. 81-2385; U. S. Government Printing Office:
Washington, DC, 1981.
J. Org. Chem, Vol. 72, No. 16, 2007 6043

Yagi and Jerina
70 °C (bath temperature) in a sealed glass tube until a clear solution
was obtained (for the conditions, see Table 1). After the reaction
was complete, the mixture was evaporated in vacuo to remove TFE,
and the residue was subjected to HPLC on two coupled Axxiom
silica columns (9.5 × 250 mm) using 25% n-hexane in EtOAc
(containing 0.4% THF) at a flow rate of 6 mL/min (detected at
280 nm). Evaporation of the fraction corresponding to the peak at
tR ) 5.8 min gave recovered 3 (∼75% recovery). Evaporation of
the fraction corresponding to the peak at tR ) 8.8 min gave (()-
10â-trifluoroethyl-7â,8r,9â-trihydroxy-7,8,9,10-tetrahydroben-
zo[a]pyrene (9a) as a colorless solid. 1H NMR and HRMS of this
solvent adduct and its triacetate (()-10â-trifluoroethyl-7â,8r,9â-
triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (9b) are given in
the Supporting Information. Diastereomeric N2-{10-(7,8,9-trihy-
droxy-7,8,9,10-tetrahydrobenzo[a]pyrenyl)}-O6-allyl-3′,5′-di-O-
(tert-butyldimethylsilyl)-2′-deoxyguanosine adducts derived from
cis- and trans-opening of (()-DE-1 (1) had tR ) 13.4 and 14.3
min, respectively. The cis/trans ratios obtained are shown in Table
1 and Figure 2. Each of the diasteromeric mixtures of cis- and trans-
adducts was separately acetylated and separated by HPLC into
diastereomerically pure cis- and trans-triacetates (5a,b), and (6a,b),
respectively. Their structures were established by comparison of
1H NMR and CD spectra with those of the authentic samples.13
Reaction of B[a]P DE-1 (1) with O6-Allyl-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyguanosine (3) in HFP. Reaction of
1 with 3 in HFP was carried out essentially the same way as above
for the corresponding reaction in TFE (for the conditions see Table
1). Separation of the unreacted 3 and products by HPLC was carried
out as above. Evaporation of the fraction corresponding to the peak
at tR ) 5.8 min gave unreacted 3 (∼75% recovery). Evaporation
fore, the combined fractions corresponding to the peak at tR ) 12.3
and 16.1 min were evaporated, the resulting mixture was acetylated
and further separated by HPLC on two coupled Axxiom silica
columns (9.5 × 250 mm) using 14% acetone in n-hexane at a flow
rate of 6 mL/min (detected at 345 and 358 nm). The cis TFE solvent
adduct triacetate 12 [(()-10r-Trifluoroethyl-7â,8r,9r-triacetoxy-
7,8,9,10-tetrahydrobenzo[a]pyrene, mp 233-234 °C (EtOAc-
n-hexane)] had tR ) 14.4 min. See the Supporting Information for
1H NMR and HRMS. The cis-(10S)-(7a) and cis-(10R)-dGuo adduct
triacetate (7b) had tR ) 19.2 and 20.1 min, respectively. The trans-
(10S)-(8a) and trans-(10R)-dGuo adduct triacetate (8b) had tR
)
24.4 and 29.9 min, respectively. The structures of these dGuo adduct
triacetates were identified by comparison of HPLC retention times,
CD, 1H NMR, and mass spectra with those of the authentic
samples.13
(+)-N2-{10S-(7S,8R,9S-Triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrenyl)}-O6-allyl-3′,5′-di-O-(tert-butyldimethylsilyl)-2′-deox-
yguanosine (7a). [R]D ) +145° (c ) 4.9, CH2Cl2). (-)-N2-{10R-
(7R,8S,9R-Triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyrenyl)}-O6-
allyl-3′,5′-di-O-(tert-butyldimethylsilyl)-2′-deoxyguanosine (7b).
[R]D ) -111° (c ) 5.5, CH2Cl2). (-)-N2-{10S-(7R,8S,9R-
Triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyrenyl)}-O6-allyl-3′,5′-
di-O-(tert-butyldimethylsilyl)-2′-deoxyguanosine (8a). [R]D
)
-10° (c ) 6.4, CH2Cl2). (+)-N2-{10R-(7S,8R,9S-Triacetoxy-
7,8,9,10-tetrahydrobenzo[a]pyrenyl)}-O6-allyl-3′,5′-di-O-(tert-bu-
tyldimethylsilyl)-2′-deoxyguanosine (8b). [R]D ) + 16° (c ) 6.0,
CH2Cl2). For 1H NMR (600 MHz, acetone-d6) spectra of compounds
7a, 7b, 8a, and 8b, see the listings of tetrahydrobenzo-ring protons.
An Optimized Large-Scale Reaction of B[a]P DE-2 (2) with
3 in PFTB. To a solution of 3 (17.1 g, 32 mmol) in PFTB (28 mL,
200 mmol) was added B[a]P DE-2 (2) (1.21 g, 4 mmol), and the
reaction mixture was stirred in a sealed tube under Ar gas at 45 °C
(bath temperature) until a clear solution was obtained (4-6 h).
Evaporation of the solvent in vacuo gave a hard oil that was
subjected to column chromatography on silica gel (4 × 20 in.) eluted
with of 30% EtOAc in n-hexane to give unreacted 3 (14.5 g, 84.7%
recovery). The column was washed with 40% EtOAc (2000 mL)
and 50% EtOAc in n-hexane (200 mL) and then eluted with 65%
EtOAc in n-hexane. Three fractions were collected. The first 2000
mL of eluent contained the diasteromeric mixture of the trans-
dGuo adducts (1.5 g, >98% pure). The second 3000 mL afforded
a mixture of trans/cis-dGuo adduct diastereomers (635 mg), and
the third 200 mL of eluent gave a diasteromeric mixture of cis-
dGuo adducts (900 mg, >98% pure). The mixture obtained from
the second fraction above was further separated by HPLC on two
coupled Axxiom silica columns (9.5 × 250 mm) as described above
using 25% n-hexane in EtOAc (containing 0.4% THF) into the
trans-dGuo adduct diastereomers (250 mg) and the cis-dGuo adduct
diastereomers (360 mg). Thus, the combined yield for the trans-
and the cis-dGuo diastereomers was 52.2% (1.75 g) and 37.6%
(1.26 g), respectively. Thus, the combined yield for the total dGuo
adducts was 89.8%.
of the fraction corresponding to the solvent adduct peak at tR
)
8.8 min gave (()-10â-hexafluoroisopropyl-7â,8r,9â-trihydroxy-
7,8,9,10-tetrahydrobenzo[a]pyrene (10a) as a colorless solid. 1H
NMR and HRMS of this solvent adduct and its triacetate (()-10â-
hexafluoroisopropyl-7â,8r,9â-triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrene (10b) are given in the Supporting Information. Diaster-
eomeric N2-{10-(7,8,9-trihydroxy-7,8,9,10-tetrahydrobenzo[a]-
pyrenyl)}-O6-allyl-3′,5′-di-O-(tert-butyldimethylsilyl)-2′-deoxyguano-
sine adducts derived from cis- and trans-opening of (()-DE-1 (1)
had tR ) 13.4 and 14.3 min, respectively, and were characterized
as above.
Reaction of B[a]P DE-1 (1) with O6-Allyl-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyguanosine (3) in PFTB. The reaction
of 1 with 3 in PFTB was carried out essentially as above for the
corresponding reaction in TFE (for the conditions, see Table 1).
Separation of unreacted 3 (∼75% recovery) and the cis- and the
trans-adducts by HPLC was also carried out as above.
Reaction of B[a]P DE-2 (2) with O6-Allyl-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyguanosine (3) in Fluorinated Alco-
hols. To a solution of the O6-allyl dGuo di-TBDMS ether (3) (5
molar equiv) in TFE, HFP, or PFTB (10-250 molar equiv) was
added B[a]P DE-2 (2) (1 molar equiv), and the reaction mixture
was stirred at room temperature or heated at 45-70 °C (bath
temperature) in a sealed glass tube until a clear solution was
obtained (for the conditions, see Table 2). After reaction, the mixture
was evaporated in vacuo to remove solvent, and the residue was
subjected to HPLC on two coupled Axxiom silica columns (9.5 ×
250 mm) using 25% n-hexane in EtOAc (containing 0.4% THF)
at a flow rate of 6 mL/min (detected at 280 nm). Evaporation of
the fraction corresponding to the peak tR ) 5.8 min gave unreacted
3 (∼75% recovery). Diastereomeric mixtures of the trans-dGuo
and the cis-dGuo adducts eluted, respectively, as single peaks at tR
) 12.3 and 16.1 min. In the case of the reaction in TFE, the solvent
adduct (11) cochromatographed with the cis-dGuo adducts. There-
Reaction of B[a]P DE-1 (1) with 3′,5′-Di-O-(tert-butyldim-
ethylsilyl)-2′-deoxyadenosine (4) in TFE. To a solution of 3′,5′-
di-O-(tert-butyldimethylsilyl)-2′-deoxyadenosine (4) (2.95 g, 6
mmol) in TFE (10.9 mL, 150 mmol) was added B[a]P DE-1 (1)
(302 mg, 1 mmol) portionwise (∼20 mg at a time over a period of
10 h with sonication at 45 °C bath temperature. When the reaction
mixture became a clear solution, an additional ∼20 mg of 1 was
added. After the addition of 1 was complete, the reaction mixture
was further sonicated at 45 °C for 4 h to ensure complete reaction.
The solvent was removed in vacuo, and the residue was acetylated
by the standard procedures to give a hard oil that was subjected to
column chromatography on a silica gel column (2.5 × 40 cm)
developed with 20% EtOAc in n-hexane to give a solid (520 mg)
that was further separated by HPLC on an Axxiom silica column
(2.5 × 250 mm) eluted with 24% EtOAc in n-hexane at a flow
rate of 10 mL/min (detected at 295 nm). Evaporation of the
combined fractions corresponding to the peak at tR ) 3.26 min
(28) Steinbrecher, T.; Wameling, C.; Oesch, F.; Seidel, A. In Polycyclic
Aromatic Comouunds: Synthesis, Analytical Measurements, Occurrence
and Biological Effects; Garrigues, P., Lamotte, M., Eds.; Gordon and
Breach: Amsterdam, 1993; pp 223-230.
6044 J. Org. Chem., Vol. 72, No. 16, 2007

Opening of Benzo[a]pyrene-7,8-diol 9,10-Epoxides
gave (()-10â-Trifluoroethyl-7â,8r,9â-triacetoxy-7,8,9,10-tet-
rahydrobenzo[a]pyrene (9b) (132 mg, 25%). Evaporation of the
combined fractions corresponding to the peak at tR ) 10.4 min
and the peak at tR ) 12.3 min afforded N6-{10S-(7S,8R,9S-
triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyrenyl}-3′,5′-di-O-(tert-
butyldimethylsilyl)-2′-deoxyadenosine (13a) (180 mg, 20%) and
N6-{10R-(7R,8S,9R-triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyre-
nyl)}-3′,5′-di-O-(tert-butyldimethylsilyl)-2′-deoxyadenosine (13b)
(182 mg, 20%), respectively. The structures of these adducts were
5.40 (d, 1H, Hc allyl, J ) 10.5), 5.44 (d, 1H, NH, J ) 10.4), 5.53 (d,
1H, Ht allyl, J ) 17.6), 5.63 (dd, 1H, H8, J ) 3.1, 2.2), 6.00 (dd,
1H, H9, J ) 5.6, 2.2), 6.19 (m, 1H, H1′), 6.27 (m, 1H, Hv allyl), 6.64
(d, 1H, H7, J ) 3.1), 6.97 (dd, 1H, H10, J ) 10.4, 5.6), 7.69 (s, 1H,
H
G8), 7.98-8.26 (m, 8H, pyrene aromatic protons). HRMS
(FAB+): calcd for C39H37O10N5Cs (M+ + Cs) 868.1595, found
868.1590.
(-)-N2-{10R-(7R,8S,9R-Triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrenyl)}-O6-allyl-2′-deoxyguanosine (15b). [R]D ) -63° (c
1
1
established by comparison of the H NMR, mass, and CD spectra
) 6.6, CH2Cl2). H NMR (300 MHz, CDCl3) δ: 1.90, 2.08, 2.20
with those of the authentic samples.14
(each s, each 3H, 3 × CH3CO), 2.12 (m, 1H, H2′′), 2.83 (m, 1H,
H2′), 3.59 (d, 1H, H5′, J ) 11.3), 3.84 (d, 1H, H5′′, J ) 11.3), 4.14
(m, 1H, H4′), 4.52 (d, 1H, H3′, J ) 4.7), 5.17 (m, 2H, CH2 allyl),
5.40 (d, 1H, Hc allyl, J ) 12.3), 5.51 (d, 1H, NH, J ) 10.4), 5.58 (d,
1H, Ht allyl, J ) 12.3), 5.61 (dd, 1H, H8, J ) 3.1, 2.2), 6.00 (dd,
1H, H9, J ) 5.6, 2.2), 6.24 (m, 1H, H1′), 6.29 (m, 1H, Hv allyl), 6.64
(d, 1H, H7, J ) 3.1), 6.95 (dd, 1H, H10, J ) 10.4, 5.6), 7.95-8.30
(m, 9H, HG8 and 8 pyrene aromatic protons). HRMS (FAB+): calcd
for C39H37O10N5Cs (M+ + Cs) 868.1595, found 868.1588.
(-)-N2-{10S-(7R,8S,9R-Triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrenyl)}-O6-allyl-2′-deoxyguanosine (16a). Colorless needles,
Reaction of B[a]P DE-2 (2) with 3′,5′-Di-O-(tert-butyldim-
ethylsilyl)-2′-deoxyadenosine (4) in TFE. To a solution of 3′,5′-
di-O-(tert-butyldimethylsilyl)-2′-deoxyadenosine (4) (12.7 g, 26.5
mmol) in TFE (52.2 mL) was added B[a]P DE-2 (2) (1.34 g, 4.42
mmol) portionwise, ∼200 mg at a time over a period of 10 h under
sonication at 60 °C. As soon as the reaction mixture became a clear
solution, another 200 mg was added. After completion of the
addition of 2, the reaction mixture was further sonicated at 60 °C
for 10 h to ensure complete reaction. The reaction mixture was
evaporated in vacuo to remove solvent, and the residue was
acetylated by the standard procedures to give a hard oil that was
subjected to column chromatography on 500 mL of silica gel eluted
with 20% EtOAc in n-hexane (2400 mL) to give a solid (3.2 g)
that was further separated by HPLC on a Vertex LiChrosorb Si-60
column eluted with 24% EtOAc in n-hexane at a flow rate of 25
mL/min (detected at 295 nm). Evaporation of the combined fractions
corresponding to the peak at tR ) 8.0 min gave (()-10r-
trifluoroethyl-7â,8r,9r-triacetoxy-7,8,9,10-tetrahydrobenzo[a]-
pyrene (12) (592 mg, 24.5%). Evaporation of the combined
fractions corresponding to the peak at tR ) 12.8 min) and the peak
at tR ) 15.3 min afforded N6-{10S-(7S,8R,9S-triacetoxy-7,8,9,-
10-tetrahydrobenzo[a]pyrenyl)}-3′,5′-di-O-(tert-butyldimethyl-
silyl)-2′-deoxyadenosine (14a) (1.17 g, 30.1%) and N6-{10R-
(7R,8S,9R-triacetoxy-7,8,9,10-tetrahydrobenzo[a]pyrenyl)}-3′,5′-
di-O-(tert-butyldimethylsilyl)-2′-deoxyadenosine (14b) (1.17 g,
30.1%), respectively. The structures of these adducts were identified
1
mp 184-185 °C (EtOH); [R]D ) -64.5° (c ) 6.6, CH2Cl2). H
NMR (300 MHz, CDCl3) δ: 2.05, 2.11, 2.31 (each s, each 3H, 3
× CH3CO), 2.30 (m, 1H, H2′′), 2.94 (m, 1H, H2′), 3.59 (m, 1H,
H5′,), 3.72 (m, 1H, H5′′), 4.07 (m, 1H, H4′), 4.66 (m, 1H, H3′), 5.07
(m, 2H, CH2 allyl), 5.28 (m, 1H, Hc allyl), 5.44 (d, 1H, Ht allyl, J )
17.5), 5.45 (d, 1H, NH, J ) 6.3), 5.91 (dd, 1H, H8, J ) 9.0, 2.1),
6.10 (dd, 1H, H10, J ) 6.3, 3.5), 6.15 (m, 1H, Hv allyl), 6.16 (app t,
1H, H9), 6.31 (m, 1H, H1′), 6.78 (d, 1H, H7, J) 9.0), 7.95-8.37
(m, 9H, HG8 and 8 pyrene aromatic protons). HRMS (FAB+): calcd
for C39H37O10N5Cs (M+ + Cs) 868.1595, found 868.1592.
(+)-N2-{10R-(7S,8R,9S-Triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrenyl)}-O6-allyl-2′-deoxyguanosine (16b). Colorless prisms,
mp 185-187 °C (EtOH); [R]D ) +31°(c ) 0.9, CH2Cl2). 1H NMR
(300 MHz, CDCl3) δ: 2.03, 2.09, 2.29 (each s, each 3H, 3 × CH3-
CO), 2.20 (m, 1H, H2′′), 3.01(m, 1H, H2′), 3.63 (m, 1H, H5′′), 3.90
(m, 1H, H5′′), 4.07 (m, 1H, H4′), 4.63 (m, 1H, H3′), 5.00 (m, 2H,
CH2 allyl), 5.24 (m, 1H, Hc allyl), 5.35 (d, 1H, Ht allyl, J ) 17.5),
5.40 (d, 1H, NH, J ) 6.3), 5.91 (dd, 1H, H8, J ) 9.0, 2.1), 6.10
(app. t, 1H, H9), 6.15(m, 1H, Hv allyl), 6.18 (dd, 1H, H10, J ) 6.7,
3.5), 6.20 (m, 1H, H1′), 6.74 (d, 1H, H7, J ) 9.0), 7.69 (s, 1H,
1
by comparison of H NMR, mass, and CD spectra with those of
authentic samples.14
Synthesis of Diastereomerically Pure O6-Allyl-dGuo Building
Blocks for Phosphoramidite Derived from B[a]P DE-2 (2). The
diastereomerically pure O6-allyl-protected (+)-cis-(10S)-7a or (-)-
cis-(10R)-7b and (-)-trans-(10S)-8a or (+)-trans-(10R)-8b, derived
from B[a]P DE-2 (2) (each on a 200 mg scale) were separately
dissolved in 7% HF-pyridine (15 mL) under cooling at 5 °C in a
polyethylene vial. The reaction mixture was stirred at room
temperature for 18 h. To the reaction mixture was added 5%
NaHCO3 solution under cooling (ice bath) until evolution of CO2
gas subsided. The mixture was evaporated in vacuo to remove
pyridine, and the residue was diluted with water (100 mL) and
extracted with EtOAc (250 mL). The extract was washed with water
(50 mL), dried (Na2SO4), and evaporated to give a colorless solid
(>95% yield).
H
G8), 7.90-8.64 (m, 8H, pyrene aromatic protons). HRMS
(FAB+): calcd for C39H37O10N5Cs (M+ + Cs) 868.1595, found
868.1616.
Acknowledgment. The authors are greatly indebted to Dr.
Jane M. Sayer at this Institute for helpful discussion and
suggestions. This research was supported by the Intramural
Research Program of the NIH (National Institute of Diabetes
and Digestive and Kidney Diseases).
Supporting Information Available: Synthetic procedures for
the preparation of dimethoxytrityl and phosphoramidite derivatives,
characterization of solvent adducts, a highly improved HPLC
separation of acetylated cis and trans dG adducts, as well as proton
NMR spectra of the new compounds 9a,b, 10a,b, 12, 15a,b, and
16a,b are provided. This material is available free of charge via
the Internet at http://pubs.acs.org.
(+)-N2-{10S-(7S,8R,9S-Triacetoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrenyl)}-O6-allyl-2′-deoxyguanosine (15a). [R]D ) +105° (c
1
) 6.6, CH2Cl2). H NMR (300 MHz, CDCl3) δ: 1.94, 2.11, 2.24
(each s, each 3H, 3 × CH3CO), 2.16 (m, 1H, H2′′), 2.83 (m, 1H,
H2′), 3.54 (d, 1H, H5′, J ) 12.6), 3.84 (d, 1H, H5′′, J ) 12.6), 4.12
(m, 1H, H4′), 4.50 (d, 1H, H3′, J ) 4.7), 5.13 (m, 2H, CH2
)
JO070303C
allyl ,
J. Org. Chem, Vol. 72, No. 16, 2007 6045