Kinetics of 2'-deoxyuridin-1'-yl radical reactions

Full text of DOI:10.1021/ja9839872

J. Am. Chem. Soc. 1999, 121, 2927-2928
Kinetics of 2′-Deoxyuridin-1′-yl Radical Reactions
2927
Calvin J. Emanuel,† Martin Newcomb,*,† Carla Ferreri,‡,§ and
Chryssostomos Chatgilialoglu*,§
Department of Chemistry, Wayne State UniVersity
Detroit, Michigan 48202
Dipartimento di Chimica Organica e Biologica
UniVersita` di Napoli “Federico II”
Via Mezzocannone 16, I-80134 Napoli, Italy
I.Co.C.E.A., Consiglio Nazionale delle Ricerche
Via P. Gobetti 101, I-40129 Bologna, Italy
ReceiVed NoVember 19, 1998
DNA C1′ radicals lead to abasic site damage with the formation
of 2-deoxyribonolactone residues.1 Such alkaline-labile lesions
result in strand scission1 and have been reported to be mutagenic
and resistant to repair nucleases.2 The mechanistic aspects of C1′
radical reactions under either anoxic or aerobic conditions are
currently under dispute. Figure 1 shows the reaction manifold
for C1′ radicals illustrated for the specific case of the 2′-
deoxyuridin-1′-yl radical (1). Reaction of 1 with a thiol such as
glutathione returns the initial nucleoside or its anomer, whereas
reaction with oxygen gives a C1′ peroxyl radical (3) that can
ultimately lead to 2-deoxyribonolactone (6). These reactions have
been discussed over the past two decades,3 but quantitative kinetic
measurements were not possible. Synthetic advances led to
nucleosides modified with photoreactive groups that are specific
C1′ radical precursors, and Greenberg4,5 and Chatgilialoglu6
reported product studies from radical 1, produced by photolysis
of precursor 7, that are consistent with the general pathway in
Figure 1. ESR and UV spectra of radical 1 were recently reported,
and computational results revealed structural details of this
radical.7 In this work, we report the application of laser flash
photolysis (LFP) methods for measurements of the kinetics of
reactions of radical 1 with thiols and of superoxide release from
peroxyl radical 3.
Figure 1. The 2′-deoxyuridin-1′-yl radical reaction manifold.
Figure 2. Left: Signal decay at 320 nm after photolysis of precursor 7
in (a) He-sparged water, (b) He-sparged water containing 0.1 M
glutathione, and (c) nonsparged water ([O2] ) 0.3 mM). Right: Signal
growth at 350 nm after photolysis of 7 (d) and di-tert-butyl ketone (e) in
nonsparged methanol-water (1:99) containing 1.07 × 10-4 M TNM.
Radical 1 was produced by 266-nm laser photolysis of precursor
7 as previously described.7 The initial cleavage process must
produce the pivaloyl radical, Me3CC(O)•, and 1 as the major
products because no further growth in the UV spectrum of 1 was
observed with ns-resolution after initial production by the laser
flash (Figure 2). The UV spectrum of radical 1 decays slowly in
He-sparged solutions but faster in the presence of oxygen due to
formation of peroxyl radical 3. A rate constant (kT) of 1 × 109
M-1 s-1 was reported for the reaction of 1 with O2.7
Figure 3. (A) Observed rate constants for decay of 1 in the presence of
cysteine (squares) and glutathione (circles). (B) Observed pseudo-first-
order rate constants for formation of nitroform anion in methanol-water
(12:88, v:v) at 22 °C. The line is simulated pseudo-first-order behavior
for release of superoxide radical anion with k ) 1.49 × 104 s-1 and
When radical 1 was produced in He-sparged solutions contain-
ing thiols, the rates of signal decay increased due to formation of
2 (Figure 3A). Second-order rate constants for reactions of 1 at
pH 7 and 20 °C were (2.3 ( 0.5) × 106 M-1 s-1 for 2-mer-
reaction of (O2)•- with TNM with k ) 2.3 × 109 M-1 s-1
.
captoethanol, (2.9 ( 0.4) × 106 M-1 s-1 for cysteine, and (4.4 (
0.3) × 106 M-1 s-1 for glutathione (errors at 2σ). The ratio of
absolute rate constants for reaction of 1 with 2-mercaptoethanol
and oxygen (kH/kT ) 2.3 × 10-3) is in good agreement with the
relative ratio found by Greenberg.4
Rate constants for heterolytic fragmentation of peroxyl radical
3 (kf in Figure 1) to give superoxide radical anion, (O2)•-, and
cation 5 were determined from reactions conducted in the presence
of tetranitromethane (TNM) which reacts with the superoxide
radical anion to give the nitroform anion (eq 1) with λmax at 350
nm.8,9 The TNM detection method is complicated. High concen-
† Wayne State University.
‡ Universita` di Napoli “Federico II”.
§ Consiglio Nazionale delle Ricerche.
(1) Pogozelski, W. K.; Tullius, T. D. Chem. ReV. 1998, 98, 1089-1107.
Burrows, C. J.; Muller, J. G. Chem. ReV. 1998, 98, 1109-1151.
(2) For example, see: Povirk, L. F.; Goldberg, I. H. Proc. Natl. Acad. Sci.
U.S.A. 1985, 82, 3182-3186. Povirk, L. F.; Houlgrave, C. W.; Han, Y.-H. J.
Biol. Chem. 1988, 263, 19263-19266.
(3) Goldberg, I. H. Acc. Chem. Res. 1991, 24, 191-198. Pratviel, G.;
Bernadou, J.; Meunier, B. Angew. Chem., Int. Ed. Engl. 1995, 34, 746-769.
von Sonntag, C.; Schuchmann, H.-P. In Peroxyl Radicals; Alfassi, Z. B., Ed.;
Wiley: New York, 1997; pp 173-234.
(4) Goodman, B. K.; Greenberg, M. M. J. Org. Chem. 1996, 61, 2-3.
(5) (a) Tallman, K. A.; Tronche, C.; Yoo, D. J.; Greenberg, M. M. J. Am.
Chem. Soc. 1998, 120, 4903-4909. (b) Tronche, C.; Goodman, B. K.;
Greenberg, M. M. Chem. Biol. 1998, 5, 263-271.
(8) Asmus, K.-D.; Henglein, A.; Ebert, M.; Keene, J. P. Ber. Bunsen-Ges.
Phys. Chem. 1964, 68, 657-663. Hiller, K.-O.; Asmus, K.-D. J. Phys. Chem.
1983, 87, 3682-3628.
(9) Rabani, J.; Mulac, W. A.; Matheson, M. S. J. Phys. Chem. 1965, 69,
53-70.
(6) Chatgilialoglu, C.; Gimisis, T. Chem. Commun. 1998, 1249-1250.
(7) Chatgilialoglu, C.; Gimisis, T.; Guerra, M.; Ferreri, C.; Emanuel, C.
J.; Horner, J. H.; Newcomb, M.; Lucarini, M.; Pedulli, G. F. Tetrahedron
Lett. 1998, 39, 3947-3950.
10.1021/ja9839872 CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/12/1999

2928 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Communications to the Editor
trations of TNM are desired to avoid convolution of the pseudo-
first-order rate constant for reaction of TNM with (O2)•- with
the first-order rate constant for (O2)•- release, but TNM reacts
rapidly with all reducing radicals9,10 and competes with oxygen
trapping reactions that give peroxyl radicals. An acceptable
balance was found with TNM concentrations of about 1 × 10-4
M; convolution of the TNM reaction kinetics was minor (see
below).
for which the kinetics are known.13 Control reactions with TNM-
containing blank solutions indicated that direct photolysis of TNM
did not complicate the analysis.
Whereas the rate constant for (O2)•- release from peroxyl
radical 3 found here is similar to that obtained for the related
peroxyl radical 9,12,13 it is 4 orders of magnitude larger than that
reported by Greenberg and co-workers from measurements of 18O-
label incorporation in product ribonolactone 6.5a The origin of
this large difference is not apparent.
(O2)•- + C(NO2)4 f O2 + (NO2)• + -C(NO2)3 (1)
The possible reactions of DNA C1′ peroxyl radicals are
trapping by thiol to give a hydroperoxide (such as 4), superoxide
radical anion release that produces a cation (5) and, ultimately,
2-deoxyribonolactone (6), and O2 release that returns the C1′
radical.14 The rate constant for reaction of the hydroperoxyl radical
LFP studies were conducted with precursor 7 and with di-tert-
butyl ketone (which gives pivaloyl and tert-butyl radicals upon
photolysis) in the presence of O2 and TNM, and signal growth at
350 nm was monitored. Nitroform anion was produced in two
stages (Figure 2); fast signal growth observed in the first few
microseconds from reactions of the initial radicals with TNM was
followed by slower growth. For six independent studies with di-
tert-butyl ketone in water at 25 °C,11 the weighted average pseudo-
first-order rate constant for nitroform anion formation in the slow
process was (0.95 ( 0.04) × 104 s-1. For six studies with
precursor 7 under otherwise identical conditions, the weighted
average pseudo-first-order rate constant was (1.39 ( 0.06) × 104
(HOO•) with thiol is about 120 M-1 s-1 15a
,
and an upper limit
for the rate constant for reaction of DNA-peroxyl radicals with
glutathione was e400 M-1 s-1 15b
Therefore, at physiological
.
concentrations of glutathione of about 5 mM, superoxide radical
anion release from C1′ nucleotide peroxyl radicals is orders of
magnitude faster than peroxyl trapping, and DNA C1′-hydro-
peroxides are not formed. Rate constants for loss of O2 from
allylperoxyl,16a cumylperoxyl,16b and nucleoside-C4′-peroxyl16c
radicals are on the order of 1-2 s-1, or 4 orders of magnitude
smaller than the rate constant for (O2)•- release from 3. Because
the stability of the C1′ radical is expected to be similar to that of
a C4′ radical on the basis of poor delocalization of the unpaired
electron into the base ring,7 we believe that loss of O2 from C1′-
peroxyl radicals to give the C1′ radicals will not be competitive
with (O2)•- release.
In summary, oxygen and glutathione trapping of C1′ radicals
in nature are competitive processes due to the low O2 concentra-
tion in the nucleus.17 Once formed, the C1′ peroxyl radicals expel
superoxide radical anion to give C1′ cations that lead to
2-deoxyribonolactone much faster than they can be trapped by
glutathione to give hydroperoxides. The apparent difference in
major reaction pathways for C1′ peroxyl radicals (superoxide
release) and C4′ peroxyl radicals (release of molecular oxygen)16c
is noteworthy.
s-1
.
The slow nitroform-forming reaction observed with di-tert-
butyl ketone is ascribed to release of either O2 or (O2)•- from the
pivaloylperoxyl radical, (CH3)3CC(O)OO•, because fragmentation
of the tert-butylperoxyl radical, (CH3)3COO•, will be several
orders of magnitude slower.12 The same reaction(s) occurred in
studies with precursor 7 because the pivaloylperoxyl radical again
was formed. The increased rate was due to reaction(s) of the C1′-
peroxyl radical 3. Loss of O2 from 3 is likely to be slow (see
below), and we ascribe the new reaction to superoxide release
from 3. The observed kinetics for nitroform anion formation with
7 are due to a mixture of reactions. From the amounts of signal
growth in the fast and slow stages of nitroform anion formation
(Figure 2), we conclude that oxygen trapping relative to initial
TNM trapping is somewhat less efficient for the C1′ radical 1
than for the pivaloyl radical, and we estimate that the rate constant
Acknowledgment. We are grateful for financial support from the
National Institutes of Health (GM-56511 to M.N.) and NATO (CRG-
970142 to M.N. and C.C.). C.J.E. is an NIH pre-doctoral fellow. We
thank Dr. K.-D. Asmus for suggesting studies with TNM. This paper is
dedicated to K. U. Ingold on the occasion of his 70th birthday.
for heterolysis of 3 in water is approximately 2 × 104 s-1
.
Various LFP studies conducted with precursor 7 in water, and
methanol-water solutions with varying concentrations of TNM
supported the kinetic description. Nitroform anion production rates
in the fast process were dependent on the concentration of TNM,
and photolysis of 7 in He-sparged MeOH-H2O (10:90, v:v)
containing 1 × 10-4 M TNM at 24 °C gave a second-order rate
constant for the reaction of all reducing radicals with TNM of 4
× 109 M-1 s-1, similar to rate constants reported for reactions of
R-heteroatom-substituted alkyl radicals with TNM.9,10 The pseudo-
first-order rate constant increased with increasing TNM concen-
tration in the low concentration regime but not in the high
concentration regime (Figure 3), a signature of convolution of
kinetic processes, and the observed kinetic behavior was simulated
by the model of consecutive reactions using a rate constant for
reaction of (O2)•- with TNM that is approximately equal to the
value reported in the literature (2 × 109 M-1 s-1).8 The
methodology was confirmed by studying an (O2)•- release reaction
JA9839872
(13) Photolysis of radical precursor 8 in the presence of O2 and TNM gave
the pivaloylperoxyl radical and peroxyl radical 9. The kinetic behavior in LFP
studies with varying concentrations of TNM was similar to that in Figure 3B.
For reactions in 30:70 MeOH-H2O at 24 °C, the limiting value for (O2)•-
release at high TNM concentrations was 2-3 × 104 s-1, in good agreement
with the value of 6 × 104 s-1 found in pulse radiolysis studies conducted in
water.12
(14) Bimolecular reactions of two DNA C1′ peroxyl radicals6 to produce
2′-deoxyribonolactone are unlikely due to the low probability that two
macromolecular radicals can meet.
(15) (a) Lal, M.; Rao, R.; Fang, X.; Schuchmann, H.-P.; von Sonntag, C.
J. Am. Chem. Soc. 1997, 119, 5735-5739. (b) Hildenbrand, K.; Schulte-
Frohlinde, D. Int. J. Radiat. Biol. 1997, 71, 377-385.
(10) Eibenberger, J.; Schulte-Frohlinde, D.; Steenken, S. J. Phys. Chem.
1980, 84, 704-710. Schuchmann, H.-P.; von Sonntag, C. Z. Naturforsch. 1987,
42B, 495-502.
(16) (a) Porter, N. A.; Mills, K. A.; Carter, R. L. J. Am. Chem. Soc. 1994,
116, 6690-6696. (b) Howard, J. A.; Bennett, J. E.; Brunton, G. Can. J. Chem.
1981, 59, 2253-2260. (c) Dussy, A.; Meggers, E.; Giese, B. J. Am. Chem.
Soc. 1998, 120, 7399-7403.
(11) Conditions: nonsparged 1% MeOH in water, [O2] ≈ 3 × 10-4 M,
[TNM] ) 1.07 × 10-4 M, 25 °C, data fit from 10 to 450 µs, errors at 2σ.
(12) Schuchmann, M. N.; Schuchmann, H.-P.; von Sonntag, C. J. Am.
Chem. Soc. 1990, 112, 403-407. von Sonntag, C.; Schuchmann, H.-P. Angew.
Chem., Int. Ed. Engl. 1991, 30, 1229-1253.
(17) Zander, R. Z. Naturforsch. 1976, 31C, 339-352. Zander, R. AdV. Exp.
Med. Biol. 1976, 75, 463-467.