Model studies indicate that copper phenanthroline induces direct strand breaks via β-elimination of the 2'-deoxyribonolactone intermediate observed in enediyne mediated DNA damage [2]

Full text of DOI:10.1021/ja9800075

J. Am. Chem. Soc. 1998, 120, 3815-3816
3815
Scheme 1
Model Studies Indicate That Copper Phenanthroline
Induces Direct Strand Breaks via â-Elimination of
the 2′-Deoxyribonolactone Intermediate Observed in
Enediyne Mediated DNA Damage
Tongqian Chen and Marc M. Greenberg*
Department of Chemistry, Colorado State UniVersity
Fort Collins, Colorado 80523
ReceiVed January 2, 1998
Oxidative damage of nucleic acids at the anomeric position of
nucleotides is effected by a variety of damaging agents and can
arise as a result of formal hydride abstraction, oxidation of the
pendant nucleobase, or hydrogen atom abstraction (1).1-7 Copper
phenanthroline (Cu(OP)2) and the enediynes (e.g., the neocarzi-
nostatin chromophore, NCS) represent two of the most well
studied families of DNA damaging agents that oxidize the C1′-
position of nucleotides in the biopolymer. Product studies and,
in the case of the enediynes, isotopic labeling experiments suggest
that the initial step in damage is hydrogen atom abstraction.5,6
Despite the formation of a common radical intermediate, Cu-
(OP)2 and the enediynes yield different products (Scheme 1). The
2′-deoxyribonolactone (2), an alkaline labile lesion, is produced
by the enediynes, whereas direct strand breaks result from Cu-
(OP)2 mediated DNA damage.
Scheme 2
Scheme 3
The cause for the apparent bifurcation in the reactivity of 1
has remained an open question. Recently, a mechanism was
tentatively put forth to explain the disparate reactivity of 1 in the
presence of these different DNA damaging agents (Scheme 2).8
Although several pathways were considered, it was suggested that
in the presence of Cu(OP)2, 1 is oxidized to the carbocation (7),
which subsequently undergoes deprotonation to the 1′,2′-dehy-
dronucleotide (8). The oxidation of 1 to 7 by one or more Cu-
(OP)2 complexes of undefined oxidation state is consistent with
the incorporation of 18O from H218O.9 The 1′,2′-dehydronucleotide
(8) is the immediate precursor to strand break formation, and it
was suggested that it gives rise to the metastable 3′-furanone (3)
and 5′-phosphate (6) containing DNA fragments via solvolysis,
obviating the need to proceed through 2. We have probed the
viability of the overall mechanism presented in Scheme 2 by
independently generating a mononucleotide analogue of 8 (11).
Based upon observations made using 11, in conjunction with
studies on 13 (a model for 2), we propose an alternative
explanation that accounts for the distinctive products formed by
Cu(OP)2 and the enediynes, such as the neocarzinostatin chro-
mophore.
In order for a 1′,2′-dehydronucleotide (8 or 11) to account for
the observed strand scission products, solvolysis must be complete
on the time scale of typical Cu(OP)2 cleavage reactions (minutes).
It is also worth noting that 11 (8) can undergo hydrolysis to yield
the free base and 13 (2) (Scheme 3).10 The 1′,2′-dehydronucle-
otide (11) was produced from phenyl selenide 9 via oxidation to
a diastereomeric mixture of selenoxides (10) by NaIO4 at 4 °C
in the probe of an NMR spectrometer (Figure 1). Upon warming
to room temperature, the major diastereomer of 10 gave rise to
11, which showed no evidence for decomposition after 48 h at
25 °C, and an additional 5 h at 50 °C.11 Purified 11 (3 mM) was
then shown to be stable in the presence and absence of Cu(OP)2
(6 mM) in HEPES buffer (pH 7.4) for 48 h at 25 °C, suggesting
that Cu(OP)2 does not accelerate its decomposition to either 12
or 13.12
(1) Neyhart, G. A.; Cheng, C. C.; Thorp, H. H. J. Am. Chem. Soc. 1995,
117, 1463.
(2) Decarroz, C.; Wagner, J. R.; Cadet, J. Free Rad. Res. Commun. 1987,
2, 295.
(3) von Sonntag, C. The Chemical Basis of Radiation Biology; Taylor &
Francis Inc.: Philadelphia, PA, 1987.
(4) (a) Sigman, D. S.; Mazumder, A.; Perrin, D. M. Chem. ReV. 1993, 93,
2295. (b) Goyne, T. E.; Sigman, D. S. J. Am. Chem. Soc. 1987, 109, 2846.
(5) (a) Meschwitz, S. M.; Schultz, R. G.; Ashley, G. W.; Goldberg, I. H.
Biochemistry 1992, 31, 9117. (b) Goldberg, I. H.; Kappen, L. S.; Wu, S. H.;
Stubbe, J.; Worth, L.; Kozarich, J. W. J. Am. Chem. Soc. 1990, 112, 2797.
(c) Kappen, L. S.; Goldberg, I. H. Biochemistry 1989, 28, 1027.
(6) (a) Yu, L.; Golik, J.; Harrison, R.; Dedon, P. J. Am. Chem. Soc. 1994,
116, 9733. (b) Epstein, J. L.; Zhang, X.; Doss, G. A.; Liesch, J. M.; Krishnan,
B.; Stubbe, J.; Kozarich, J. W. J. Am. Chem. Soc. 1997, 119, 6731. (c) Xu,
Y.; Xi, Z.; Zhen, Y.; Goldberg, I. H. Biochemistry 1995, 34, 12451.
(7) Pitie´, M.; Bernadou, J.; Meunier, B. J. Am. Chem. Soc. 1995, 117, 2935.
(8) (a) Meijler, M. M.; Zelenko, O.; Sigman, D. S. J. Am. Chem. Soc. 1997,
119, 1135. (b) Zelenko, O.; Gallagher, J.; Sigman, D. S. Angew. Chem., Int.
Ed. Engl. 1997, 36, 2776.
Given the stability of 11 under aqueous conditions, we explored
an alternative explanation for the difference between Cu(OP)2
(10) (a) Haraguchi, K.; Tanaka, H.; Maeda, H.; Itoh, Y.; Saito, S.; Miyasaka,
T. J. Org. Chem. 1991, 56, 5401. (b) Robins, M. J.; Trip, E. M. Tetrahedron
Lett. 1974, 3369.
(11) The product mixture observed by 1H NMR was characterized by
electrospray mass spectrometry, which confirmed the presence of 10 and 11.
(9) Wayner, D. D. M.; Griller, D. Mol. Struct. Energy 1989, 11, 109.
S0002-7863(98)00007-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/04/1998

3816 J. Am. Chem. Soc., Vol. 120, No. 15, 1998
Communications to the Editor
Figure 2. Plot of observed rate constant for the disappearance of 13 (3
mM) as a function of Cu(OP)2 concentration. Inset: Extrapolated kobsd
and half-life as a function of Cu(OP)2 concentration.
to DNA. The effective molarity of noncovalent complexes can
be as high as 105-107 M.14 Extrapolation of the observed rate
constant to between 10 and 100 M effective concentration of Cu-
(OP)2 would result in a half-life for elimination from 13 of less
than 1 min and, if extrapolable to 2 in DNA, would explain the
formation of direct strand breaks by Cu(OP)2. Although the
oxidation state of the copper phenanthroline complex (or the
stoichiometry) responsible for this catalysis is unknown, the
stability of 11, a likely solvolysis candidate, in the presence of
Cu(OP)2 concentrations that accelerate elimination from 13,
suggests that the complex is acting as a general base catalyst.
While the monomeric compounds described above cannot
unequivocally model the reaction between Cu(OP)2 and DNA,
their reactivity leads us to propose that the differences in the
effects of the enediynes and Cu(OP)2 on nucleic acids can be
explained by applying Ockham’s razor or the principle of
mechanistic economy.15 We propose that the structurally distinct
DNA damaging agents referred to above produce 2 as a common
intermediate (possibly via different mechanisms), which when
formed in the presence of noncovalently bound Cu(OP)2 under-
goes subsequent elimination. These results suggest that similar
catalysis may be operating in other nucleic acid damage processes
mediated by coordination complexes where possible alkaline labile
lesions are not observed.16
Figure 1. 1H NMR spectra describing the generation of 11 from 9 (20
mM) via 10 in phosphate buffer (0.1 M, pD 7.4). (a) 9 prior to the addition
of NaIO4 (4 °C). (b) After the addition of 1 equiv of NaIO4 (2 h at 4 °C,
then 22 min at 25 °C). (c) After 48 h at 25 °C. 1H NMR assigments: 9;
δ 6.45 (C5), 5.51 (C1′), 5.04 (C3′). 10; δ 6.33 (C5), 5.69 (C1′), 5.29
(C3′). 11; δ 6.15 (C5), 5.56 (C2′), 5.26 (C3′).
and enediyne reactivity. We considered the possibility that 2 is
formed by both the enediynes and Cu(OP)2 but that the nonco-
valently bound copper complex catalyzes its â-elimination to 3.
Such a mechanism is consistent with the observed 18O-incorpora-
tion in the lactone from H218O by including the proposed one-
electron oxidation of the initially formed radical.8a Consequently,
the rate of decomposition of 13 was examined in HEPES buffer
over a range of Cu(OP)2 concentrations (Figure 2).13 The rate of
disappearance of 13 was quantitatively accounted for by the
appearance of phenyl phosphate and obeyed first-order kinetics
for two half-lives. Higher conversion of 13 led to a deviation
from first-order decay, which was attributed to inhibition by
coordination of Cu(OP)2 with the product phenyl phosphate. The
inhibition observed by added phenyl phosphate (3 mM) was
consistent with this hypothesis. The observed rate constant for
the disappearance of 13 varied linearly with Cu(OP)2 concentra-
tion but was unaffected by CuSO4 (which precipitates at pH 7.4)
or phenanthroline by themselves. The observed rate constant for
the disappearance of 13 approximately doubles between 0 and 6
mM Cu(OP)2, and the significance of this increase is evident when
one considers the potential effective molarity of Cu(OP)2 bound
Acknowledgment. We thank Professor David Sigman (UCLA) for
helpful discussions and Dr. Chris Rithner (CSU) for assistance with the
NMR experiments. We are grateful for support of this research by the
National Institutes of Health (GM-54996). Mass spectra were obtained
on instruments supported by the National Institutes of Health shared
instrumentation grant (GM-49631). M.M.G. thanks the Alfred P. Sloan
Foundation for a fellowship.
Supporting Information Available: Electrospray mass spectrum of
1
the mixture of 10 and 11, H NMR spectra shown in Figure 1 from δ
10.0-0.0, 1H NMR spectra of 12 and 13 in D2O (pD 7.4, 0.1 M, 25 °C),
and experimental procedure for the synthesis of 13 (8 pages). See any
current masthead page for ordering information and Web access
instructions.
(12) The dehydronucleotide proved to be stable to column chromatography
in the presence of Et3N. Analysis was carried out by HPLC using a Rainin
Microsorb-MV C18 column (5 µm, 4.6 × 250 mm). Eluent A: NH4Cl (0.1
M, pH 7.2), 5% MeOH. Eluent B: NH4Cl (0.1 M, pH 7.2), 70% MeOH.
Gradient: 0-40% B linearly over 15 min; 40-100% B linearly over 5 min,
followed by 100% B for 30 min. Ret. time: 11, 26.6 min; uridine (internal
standard), 6.6 min.
(13) Analysis was carried out by HPLC using a Waters Spherisorb S10
SAX column (5 µm, 4.6 × 250 mm). Eluent A: KH2PO4 (50 mM, pH 4.5).
Eluent B: KH2PO4 (0.1 M, pH 3.0) + KCl (0.1 M); Gradient: 0% B 7 min;
0-100% B linearly over 12 min, followed by 100% B for 30 min. Ret. time:
GMP (internal standard), 6.4 min; 13, 7.6 min; phenyl phosphate, 9.6 min.
JA9800075
(14) For a recent example, see: Hollfelder, F.; Kirby, A. J.; Tawfik, D. S.
J. Am. Chem. Soc. 1997, 119, 9578. For an overview see: Fersht, A. Enzyme
Structure and Mechanism, 2nd ed.; W. H. Freeman and Co.: New York, 1985.
(15) (a) Hoffmann, R.; Minkin, V. I.; Carpenter, B. K. Bull. Soc. Chim.
Fr. 1996, 133, 117. (b) Berson, J. A. In Rearrangements in Ground and Excited
States; de Mayo, P., Ed.; Academic Press: New York, 1980; Vol. 1.
(16) (a) Cook, G. P.; Greenberg, M. M. J. Am. Chem. Soc. 1996, 118,
10025. (b)Sitlani, A.; Long, E. C.; Pyle, A. M.; Barton, J. K. J. Am. Chem.
Soc. 1992, 114, 2303.