the observed formation by HPLC of the mixed chalcogenide
(GSSePh) derived from glutathione and the phenyl selenide
(2), which was independently prepared.20 On the basis of
literature values for bimolecular rate constants of related
reactions, we expect most of the alkyl radicals derived from
VA-044 to be scavenged by GSH.16,19 Although we tenta-
tively suggest that the thiyl radical reacts with 2 to produce
1 and GSSePh (eq 4, Scheme 3), further investigation is
Table 2. Reaction of Phenyl Selenide (2) with GSH under
Aerobic Conditions at 37 °Ca
% yieldb
VA-044
7
8
10
mass balance
+
-
6.9 ( 1.2
6.3 ( 1.4
57.9 ( 0.7
60.2 ( 1.3
5.9 ( 2.9
7.3 ( 2.8
85.0 ( 1.5
86.8 ( 2.7
a [2] ) 50 µM, [GSH] ) 5 mM, [VA-044] ) 25 µM, 3 h reaction time.
b Based upon unrecovered starting material. Yields determined using
deoxyadenosine as an internal standard.
Scheme 3
The lack of a need for azo initiator to induce decomposi-
tion of 2 was attributed to trace metal catalysis, a known
pathway for O2-dependent thiol oxidation.18 This proposal
is supported by the effect of added disodium EDTA (1 mM)
on the reaction. Less than 1% reaction is observed over the
course of 3 h in the absence of VA-044 when the chelator is
added (Figure 2). The overall decomposition rate of 2 is
warranted. A mechanism to explain the formation of 1 under
aerobic conditions is also uncertain. In the absence of EDTA,
trace metals are expected to produce thiyl radicals, whereas
deazatization of VA-044 is the major pathway when EDTA
is present. Thiyl radicals are rapidly, but reversibly trapped
by O2.21 The formation of GSSePh is consistent with the
reaction between thiyl radical and phenyl selenide to form
1. Although the sulfur radicals are typically poor chain
propagators, the present reaction conditions for producing 1
do not necessitate a chain mechanism.22
In summary, we find that phenyl selenide 2 enables us to
generate 5-(2′-deoxyuridinyl)methyl radical (1) under milder
conditions than 3 or 4. Furthermore, 1 can be produced from
the phenyl selenide under biologically relevant conditions
(O2, GSH, 37 °C) without UV photolysis. The mechanism
and scope of this reaction warrants investigation. Regardless,
thermolytic formation of 1 should prove to be useful for
studying the reactivity of this radical in duplex DNA.
Figure 2. Reaction of phenyl selenide 2 (50 µΜ) with GSH (5
mM) in the presence of EDTA (1 mM) under aerobic conditions
with and without VA-044 (25 µM) at 37 °C.
considerably slower in the presence of EDTA, even when
VA-044 is added. However, there is a clear increase in phenyl
selenide decomposition when the initiator is present.
Acknowledgment. We are grateful for support of this
research by the National Institute of General Medical
Sciences (NIGMS-054996).
Regardless of whether EDTA is present, hydroxymethyl-
2′-deoxyuridine (8) is the major product of the “thermolyses”,
suggesting that a common intermediate is formed. We believe
that the peroxyl radical (11) derived from O2 trapping of
5-(2′-deoxyuridinyl)methyl radical (1) is the common inter-
mediate (Scheme 2). Literature precedent suggests that 1
could form under anaerobic conditions by SH2 reaction
between 2 and the alkyl radical resulting from deazatization
of VA-044.19 However, this mechanism is inconsistent with
Supporting Information Available: Experimental pro-
cedures for the synthesis of 2 and spectroscopic characteriza-
tion of all new compounds. This material is available free
OL047754T
(20) Engman, L.; Andersson, C.; Morgenstern, R.; Cotgreave, I. A.;
Andersson, C.-M.; Hallberg, A. Tetrahedron 1994, 50, 2929-2938.
(21) Razskazovskii, Y. V.; Becker, D.; Sevilla, M. D. In S-Centered
Radicals; Alfassi, Z. B., Ed.; John Wiley & Sons: Chichester, 1999.
(22) Keck, G. E.; Byers, J. H. J. Org. Chem. 1985, 50, 5442-5444.
(18) Bagiyan, G. A.; Koroleva, I. K.; Soroka, N. V.; Ufimtsev, A. V.
Russ. Chem. Bull., Int. Ed. 2003, 52, 1135-1141.
(19) Curran, D. P.; Martin-Esker, A., A.; Ko, S.-B.; Newcomb, M. J.
Org. Chem. 1993, 58, 4691-4695.
Org. Lett., Vol. 6, No. 26, 2004
5013