Cleavage of DNA by proton-coupled electron transfer to a photoexcited, hydrated Ru(II) 1,10-phenanthroline-5,6-dione complex

Article abstract of DOI:10.1021/ja3106863

Visible light irradiation of a ruthenium(II) quinone-containing complex, [(phen)₂Ru(phendione)]²⁺ (1²⁺), where phendione = 1,10-phenanthroline-5,6-dione, leads to DNA cleavage in an oxygen independent manner. A combination of NMR analyses, transient absorption spectroscopy, and fluorescence measurements in water and acetonitrile reveal that complex 1 ²⁺ must be hydrated at the quinone functionality, giving [(phen) ₂Ru(phenH₂O)]²⁺ (1H₂O²⁺, where phenH₂O = 1,10-phenanthroline-6-one-5-diol), in order to access a long-lived ³MLCT_hydrate state (τ ～ 360 ns in H₂O) which is responsible for DNA cleavage. In effect, hydration at one of the carbonyl functions effectively eliminates the low-energy ³MLCT_SQ state (Ru^III phen-semiquinone radical anion) as the predominant nonradiative decay pathway. This ³MLCT _SQ state is very short-lived (<1 ns) as expected from the energy gap law for nonradiative decay,(1) and too short-lived to be the photoactive species. The resulting excited state in 1H₂O²⁺* has photophysical properties similar to the ³MLCT state in [Ru(phen) ₃]²⁺* with the added functionality of basic sites at the ligand periphery. Whereas [Ru(phen)₃]²⁺* does not show direct DNA cleavage, the deprotonated form of 1H₂O ²⁺* does via a proton-coupled electron transfer (PCET) mechanism where the peripheral basic oxygen sites act as the proton acceptor. Analysis of the small molecule byproducts of DNA scission supports the conclusion that cleavage occurs via H-atom abstraction from the sugar moieties, consistent with a PCET mechanism. Complex 1²⁺ is a rare example of a ruthenium complex which 'turns on' both reactivity and luminescence upon switching to a hydrated state.

Full text of DOI:10.1021/ja3106863

Communication
pubs.acs.org/JACS
Cleavage of DNA by Proton-Coupled Electron Transfer to a
Photoexcited, Hydrated Ru(II) 1,10-Phenanthroline-5,6-dione
Complex
§
†
§
§
§
Steven A. Poteet, Marek B. Majewski, Zachary S. Breitbach, Cynthia A. Griffith, Shreeyukta Singh,
§
†
,§
Daniel W. Armstrong, Michael O. Wolf, and Frederick M. MacDonnell*
§Department of Chemistry and Biochemistry, University of Texas at Arlington, Arlington, Texas 76019, United States
†Department of Chemistry, University of British Columbia, Vancouver, BC V6T 1Z1, Canada
*
S Supporting Information
reactions such as DNA cleavage. Only a few of these transition
metal complexes are able to cleave DNA upon photoexcitation
ABSTRACT: Visible light irradiation of a ruthenium(II)
2+
quinone-containing complex, [(phen) Ru(phendione)]
without the need for O ; however, it is clear that such an
2
2
2+
(
1 ), where phendione = 1,10-phenanthroline-5,6-dione,
approach could offer some therapeutic advantages in that many
6
−9
leads to DNA cleavage in an oxygen independent manner.
A combination of NMR analyses, transient absorption
spectroscopy, and fluorescence measurements in water and
cancer cells are under hypoxic stress.
Herein, we describe the oxygen independent DNA cleavage
activity of a ruthenium(II) complex, [(phen) Ru(phendione)]-
2
2
+
acetonitrile reveal that complex 1 must be hydrated at
Cl (1Cl ) upon visible light irradiation into the MLCT band at
2
2
2
+
the quinone functionality, giving [(phen) Ru-
470 nm. Complex 1 is a commonly used synthon in
ruthenium polypyridyl chemistry
2
2+
2+
10−12
(
phenH O)] (1H O , where phenH O = 1,10-phenan-
and yet, only recently
2
2
2
throline-6-one-5-diol), in order to access a long-lived
has its unusual solvent and temperature dependent para-
3
13
MLCThydrate state (τ ∼ 360 ns in H O) which is
magnetism, and now photoreactivity, been described. An
2
responsible for DNA cleavage. In effect, hydration at one
of the carbonyl functions effectively eliminates the low-
investigation into the mechanism of DNA cleavage reveals that
2
+
2+
complex 1 is in equilibrium with a hydrated species 1H O
2
3
III
energy MLCT state (Ru phen-semiquinone radical
(see Scheme 1) and that the latter is the photoactive species.
SQ
anion) as the predominant nonradiative decay pathway.
This MLCTSQ state is very short-lived (<1 ns) as
3
Scheme 1. Hydration Equilibrium for Complex 12+ in Water
expected from the energy gap law for nonradiative
1
decay, and too short-lived to be the photoactive species.
2
+
The resulting excited state in 1H O * has photophysical
2
3
2+
properties similar to the MLCT state in [Ru(phen) ] *
with the added functionality of basic sites at the ligand
periphery. Whereas [Ru(phen) ] * does not show direct
DNA cleavage, the deprotonated form of 1H O * does
3
2
+
3
2
+
2
via a proton-coupled electron transfer (PCET) mechanism
where the peripheral basic oxygen sites act as the proton
acceptor. Analysis of the small molecule byproducts of
DNA scission supports the conclusion that cleavage occurs
via H-atom abstraction from the sugar moieties, consistent
with a PCET mechanism. Complex 1 is a rare example of
a ruthenium complex which ‘turns on’ both reactivity and
luminescence upon switching to a hydrated state.
2+
The excited state photophysics of 1 and 1H O and the
2
mechanism by which DNA cleavage is induced were probed by
a combination of transient absorption and luminescence
spectroscopy, electrochemistry, and product analysis as
described below.
The DNA photocleavage activity of 1 under aerobic and
anaerobic conditions was examined using a common plasmid
cleavage assay. Because of the hydration reaction (vide infra),
2
+
2
+
2+
2+
1
is present as an equilibrium mixture of 1 and 1H O
2
number of transition metal complexes have attracted
which is collectively referred to as RuPD in the following
section. Supercoiled pUC18 plasmid DNA was incubated with
RuPD (loading ratio: 1 Ru complex per approximately 6 DNA
base pairs) and irradiated with 470 nm light for various lengths
of time in the presence or absence of O . The plasmid cleavage
products were visualized using agarose gel electrophoresis to
separate supercoiled (Form I), circular (Form II) and linear
A
attention as potential photodynamic therapy (PDT)
agents for cancer treatment, as they often have a high affinity
for DNA and accessible photoexcited states in the visible
2
region. Turro, Brewer, Dunbar, and others have shown that
2
complexes of Ru(II), Os(II), and dinuclear Rh(III) cores can be
3
−5
made into effective DNA photocleavage agents.
Most
commonly, the PDT agent works by activation of cellular O2
1
to form reactive oxygen species (ROS), such as O or ·OH,
Received: October 30, 2012
Published: January 27, 2013
2
which are responsible for inducing apoptosis via damaging
©
2013 American Chemical Society
2419
dx.doi.org/10.1021/ja3106863 | J. Am. Chem. Soc. 2013, 135, 2419−2422

Journal of the American Chemical Society
Form III) plasmid DNA, as shown in Figure 1. As can be seen
Communication
(
Integration leads to an estimated equilibrium constant of 0.5
and as expected, addition of DCl shifts the equilibrium toward
the nonhydrated species (Figure S4). The observation of two
distinct species (noncoalescence) and line broadening suggests
that the hydration/dehydration process is on the slow end of
in lane 2, no DNA cleavage is observed upon incubation of
24
the NMR time scale (ca. 0.25 s), in good agreement with that
estimated by Anson and co-workers using electrochemical
16
modeling for the Os complex process (kdehyd = 0.5 s).
To determine the photophysical properties of the two
complexes, we used a combination of transient absorption
(TA) and fluorescence techniques in acetonitrile or water. TA
Figure 1. Ethidium bromide stained 1% agarose gel of photocleavage
products of RuPD and pUC18: 1 μg/1 μL, 0.154 mM pUC18 DNA
base pairs (4 mM phosphate, 50 mM NaCl, pH = 7.4) and 27 μM of
the chloride salt of the ruthenium complex, irradiated (when
applicable) at 470 nm for 2 h; lane 1, plasmid only; lane 2, RuPD
dark in presence of O ; lane 3, RuPD irradiated in presence of O ; lane
2
+
data is perhaps the most revealing. In MeCN, where 1 is
undoubtedly the bulk species in solution, we observe a weak
signal consistent with bleaching of the ground state MLCT
bands (400−500 nm) in the 0.05−1 ns time regime (see Figure
S5) but not at longer times, indicating a subnanosecond
lifetime. A more pronounced MLCT bleach is observed in the
2
2
2
+
4
, RuPD irradiated, deaerated; lane 5, [Ru(phen) ] irradiated,
3
2+
deaerated; lane 6, [(phen) Ru(dppz)] irradiated, deaerated.
2
2
+
excited state difference spectrum of [Ru(phendione)3]
Figure S5) and a monoexponential fit of this decay data
gave τ = 60 ps (Figure S6). The very short lifetime of the
(
DNA with RuPD in the dark, but upon irradiation, DNA
cleavage to circular form DNA is observed under both aerobic
3
2+
MLCTSQ state in [Ru(phendione) ] * is anticipated on the
(
lane 3) and anaerobic (lane 4) conditions, with somewhat
3
1
basis of the energy gap law and, by analogy, is anticipated for
the MLCT state in 1 * (although we were not able to
measure this directly). In water, the difference TA spectrum of
better cleavage yields observed with O2 present. In the
anaerobic experiment (lanes 4−6), samples were subjected to
three consecutive freeze−pump−thaw cycles prior to exper-
3
2+
SQ
2
+
2+
the mixture of 1 * and 1H O * (Figure S7) shows a bleach in
imentation in an O -free glovebox. To accentuate the unusual
2
2
the ground state MLCT region which can be fit (mono-
exponentially) to τ = 334 ns, a lifetime that correlates well to
the observed luminescence decay data (Figure S8. λem = 610
nm; τem = 359 ns). Given the very short lifetime of the
nonhydrated complex, this longer-lived excited state behavior
anaerobic photocleavage activity of RuPD, two closely related
2
+
ruthenium complexes, [Ru(phen) ] , which lacks the
3
2
+
phendione ligand, and [(phen) Ru(dppz)] (dppz = dipyrido-
2
[
(
3,2-a:2′,3′-c]phenazine), which is known to bind DNA tightly
7
−1 14
K ∼10 M ) via intercalation, were examined for anaerobic
b
2
+
can be assigned to the hydrate 1H O *. Hydration is reported
photocleavage activity in the same assay. Neither [Ru-
(
2
phen)3]2 (lane 5) nor [(phen) Ru(dppz)] (lane 6) is
+
2+
to render the quinone function electrochemically inactive in
some studies or sluggish in others, as it is proposed that it is
necessary to dehydrate and reform the quinone for electro-
2
seen to cause any cleavage under these conditions, indicating
the essential role of the phendione ligand.
2
1,23,25
chemical activity.
As this hydration-dehydration process
The enhanced photocleavage activity of RuPD in the
presence of oxygen (compare lanes 3 vs 4 in Figure 1) shows
that photoexcited RuPD can also induce DNA damage by the
generation of ROS, but the data in lane 4 reveals that an O2
independent pathway also exists. Because it is very difficult to
completely remove O2 from aqueous solutions, several
anaerobic DNA photocleavage experiments were also per-
formed in the presence of ROS quenchers. Superoxide
dismutase (SOD) effectively quenches superoxide while sodium
is very slow compared to the lifetimes of the excited state
species, it is clear that the lowest energy excited state in
2
+
2+
1
H O * is somewhat similar to that in [Ru(phen) ] *, which
2
3
is reflected both in the emission energy and the lifetime.
This model is shown graphically in the Jablonski diagram
2
+
2+
1
(
Figure 2). Both 1 and 1H O absorb in the MLCT band
2
3
2+
and intersystem cross to the MLCT
rapidly deactivates to the MLCTSQ state. From ground state
state. For 1 , this
phen
3
.‑
5
azide is known to quench both singlet oxygen and OH .
Addition of either SOD or sodium azide did not show any
attenuation of the photocleavage reaction. (see Figure S3).
However, addition of 2,2,6,6-tetramethylpiperidine-1-oxyl
(
TEMPO) does attenuate the cleavage (Figure S3). TEMPO
is a good quencher for metal- and carbon-based radical
species,
1
5−18
suggesting that one or both of these species are
generated upon photoexcitation of RuPD and are responsible
for the cleavage activity.
The question now arises, what is the photoreactive species
and what is the mechanism of DNA cleavage? NMR analysis of
complex 12 in CD CN reveals a spectrum consistent with the
+
3
proposed structure; however, the NMR spectrum in D O
2
shows substantial line broadening and complicated speciation
(
see Figure S4). The hydration reaction shown in Scheme 1
was implicated as the enhanced hydration of carbonyl moieties
on coordinated ligands is a known process and frequently
19
20−23
2+
reported for phendione ligands in particular.
co-workers report an equilibrium constant of 0.9 for hydration
21
Anson and
Figure 2. Proposed Jablonski diagram for 1 in MeCN (red box) or
1H O in water (blue box). Deprotonation of 1H O to 1(OH)+
2
+
2+
2
2
2
+
of [(bpy) Os(phendione)] at pH 5.2 in aqueous solution.
shifts the emission maxima from 624 to 602 nm.
2
2
420
dx.doi.org/10.1021/ja3106863 | J. Am. Chem. Soc. 2013, 135, 2419−2422

Journal of the American Chemical Society
Communication
redox potentials and an estimate of 0.5 eV energy loss due to
aqueous photochemistry of 12+ is dominated by formation of
3
intersystem crossing, the energy of the MLCT is estimated
the hydrate 1H O, which has the effect of turning off the redox
SQ
2
2
+
at 0.8 eV. For 1H O , the SQ state is not available (on the
activity of the quinone function leading to an excited state
2
2+
time scale of the excited state process) and the ‘bipyridine-like’
manifold similar to [Ru(phen) ] *, except that this complex
3
portion of the phenH O ligand is the site of electron storage.
also possesses peripheral basic sites. These two features
2
3
This MLCT
state is 1.9 eV above the ground state (from
combine to lead to effective DNA cleavage via a PCET
‘
bpy’
luminescence data) which is consistent with the considerably
longer excited state lifetime.
mechanism. We note that when O is not removed from the
solution, DNA photocleavage is enhanced, presumably through
2
In an effort to further establish the mechanism of action, we
scaled up the anaerobic reaction and extracted the aqueous
phase with CH Cl to look for neutral, small-molecule DNA
the formation of ROS in addition to the O independent
2
mechanisms.
2
2
cleavage products characteristic of H-atom abstraction from
either C1′ or C5′ of the deoxyribose units. The presence of
furfural in the irradiated samples was confirmed by HPLC
ASSOCIATED CONTENT
Supporting Information
■
*
S
Experimental details, tables, and figures regarding spectroscopic
properties and HPLC chromatograms of the DNA small
molecule products. This material is available free of charge via
the Internet at http://pubs.acs.org.
(
Figure S9) and is indicative of H-atom abstraction from the
2,26,27
C5′ position.
Collecting all this data, we propose the mechanism shown in
Scheme 2. It is clear that the oxygen independent cleavage of
AUTHOR INFORMATION
Corresponding Author
macdonn@uta.edu
■
+
Scheme 2. Proposed PCET Mechanism between 1(OH) *
and DNA
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank NSF 0911720 (F.M.M.) and the Welch Foundation
Y-1301, F.M.M.), NSERC (M.O.W.) and Mountain Equip-
■
(
ment Co-op (M.B.M.) for funding and LASIR (UBC) for TA
and emission lifetimes. We also acknowledge the Genetics and
Core Facilities at UT Arlington for biological materials and
instrumentation, David Boston for the photoreactor design and
actinometry, and Jonathan Smuts for help purifying our
complexes.
2
+
DNA by 1H O * must involve the peripheral oxygens as
2
otherwise we would expect to see similar photocleavage by
complexes such as [Ru(phen) ] * and [(phen) Ru(dppz)] *.
Fluorescent pH titration (see Figure S10) established the pK
2
+
2+
3
2
a
REFERENCES
2
+
2+
■
and pK * for 1H O and 1H O * at 5.8 and 7.1, respectively,
a
2
2
+
(1) Kober, E. M.; Caspar, J. V.; Lumpkin, R. S.; Meyer, T. J. J. Phys.
Chem. 1986, 90, 3722−3734.
indicating that the deprotonated complex, 1OH , is the proton
acceptor, as shown in Scheme 2. We note that some of the
increase in cleavage may also be due to an increase in the
(
(
2) Armitage, B. Chem. Rev. 1998, 98, 1171−1200.
3) Miao, R.; Mongelli, M. T.; Zigler, D. F.; Winkel, B. S. J.; Brewer,
+
concentration of 1OH as high pH favors formation of this
K. J. Inorg. Chem. 2006, 45, 10413−10415.
species. Upon excitation, localization of the excited electron in
the ‘bpy-like’ MO on the hydrated ligand would boost the
basicity of the oxygen base which accepts a proton from the
sugar C−H bond. At the same time, an electron is transferred
to the Ru(III) site in a PCET mechanism, resulting in a Ru(II)
semiquinone species (after dehydration) and a deoxyribose
radical species. Formation of deoxyribose radical at any of the
deoxyribose carbons is known to lead to strand scission by
subsequent decay reactions. The Ru(II) semiquinone species
is likely to be unstable to disproportionation and yields 1 and
(phen) Ru(phendiol)] as terminal products. We note that as
the complex is bound to DNA, the short-lived MLCT state
cannot be ruled out as the reactive species at this point.
The PCET reaction between [(bpy) Ru(phendione)] * and
tetrachlorohydroquinone has been reported by Meyer and co-
workers in a preliminary report, but no details or mention of
the role of the hydrated species was included.
(4) Sun, Y.; Collins, S. N.; Joyce, L. E.; Turro, C. Inorg. Chem. 2010,
49, 4257−4262.
(5) Sun, Y.; Joyce, L. E.; Dickson, N. M.; Turro, C. Chem. Comm
2010, 46, 2426−2428.
(6) Wang, J.; Higgins, S. L. H.; Winkel, B. S. J.; Brewer, K. J. Chem.
Commun. 2011, 47, 9786−9788.
(7) Bradley, P. M.; Angeles-Boza, A. M.; Dunbar, K. R.; Turro, C.
Inorg. Chem. 2004, 43, 2450−2452.
(
(
8) Swavey, S.; Brewer, K. J. Inorg. Chem. 2002, 41, 6196−6198.
22
9) Joyce, L. E.; Aguirre, J. D.; Angeles-Boza, A. M.; Chouai, A.; Fu,
2
+
P. K. L.; Dunbar, K. R.; Turro, C. Inorg. Chem. 2010, 49, 5371−5376.
(10) MacDonnell, F. M.; Kim, M.-J.; Bodige, S. Coord. Chem. Rev.
1999, 185−186, 535−549.
2
+
[
2
3
SQ
(11) Hartshorn, R. M.; Barton, J. K. J. Am. Chem. Soc. 1992, 114,
5919−5925.
2+
2
(12) Bolger, J.; Gourdon, A.; Ishow, E.; Launay, J.-P. J. Chem. Soc.,
Chem. Comm 1995, 1799−1800.
28−31
(13) We were fortunate to receive a preprint of a recent paper by
Schmidt, R. D.; Kent, C. A.; Concepcion, J. J.; Lin, W.; Meyer, T. J.;
Forbes, M. D. E. entitled “A Little Spin on the Side: Solvent and
T e m p e r a t u r e D e p e n d e n t P a r a m a g n e t i s m i n
Similarly, Ru
complexes with peripheral basic sites such as [Ru-
2+
(
phen) (bpz)] (bpz = 2,2′-bipyrazine) are also known to
2
react with hydroquinones via essentially the same PCET
mechanism as shown in Scheme 2. This is the first report to
delineate the role of ligand hydration in this chemistry and to
apply it toward a biological substrate. It is now clear that the
II
2+
[
Ru (bpy) (phendione)] ” (submitted for publication) which
2
describes this unexpected behavior.
(14) Ossipov, D.; Pradeepkumar, P. I.; Holmer, M.; Chattopadhyaya,
J. J. Am. Chem. Soc. 2001, 123, 3551−3562.
2
421
dx.doi.org/10.1021/ja3106863 | J. Am. Chem. Soc. 2013, 135, 2419−2422

Journal of the American Chemical Society
Communication
(
15) Step, E. N.; Buchachenko, A. L.; Turro, N. J. J. Am. Chem. Soc.
994, 116, 5462−5466.
16) Mohler, D. L.; Dain, D. R.; Kerekes, A. D.; Nadler, W. R.; Scott,
T. L. Bioorg. Med. Chem. Lett. 1998, 8, 871−874.
17) Tenhaeff, S. C.; Ceovert, K. J.; Castellani, M. P.; Grunkemeier,
J.; Kunz, C.; Weakley, T. J. R.; Koenig, T.; Tyler, D. R. Organometallics
993, 12, 5000−5004.
18) Connolly, T. J.; Baldova, M. V.; Mohtath, N.; Scaiano, J. C.
Tetrahedron Lett. 1996, 37, 4919−4922.
19) Velarde, E. L.; Stephen, R. A.; Mansour, R. N.; Hoang, L. T.;
Burkey, D. J. J. Am. Chem. Soc. 2003, 125, 1188−1189.
20) Hilt, G.; Jarbawi, T.; Heineman, W. R.; Steckhan, E. J. Eur.
1
(
(
1
(
(
(
Chem. 1997, 3, 79−88.
(
(
21) Lei, Y.; Anson, F. C. J. Am. Chem. Soc. 1995, 117, 9849−9854.
22) Gillard, R. D.; Hill, R. E. E. J. Chem. Soc., Dalton. Trans. 1974,
1
(
(
(
217−1236.
23) Goss, C. A.; Abrun
̃
a, H. D. Inorg. Chem. 1985, 24, 4263−4267.
24) Bryant, R. G. J. Chem. Educ. 1983, 60, 933−935.
25) Eckert, T. S.; Bruice, T. C. J. Am. Chem. Soc. 1983, 105, 4431−
4
441.
(
(
26) Kumar, A.; Sevilla, M. D. Chem. Rev. 2010, 110, 7002−7023.
27) Pogozelski, W. K.; Tullius, T. D. Chem. Rev. 1998, 98, 1089−
1108.
(
28) Weinberg, D. R.; Gagliardi, C. J.; Hull, J. F.; Murphy, C. F.;
Kent, C. A.; Westlake, B. C.; Paul, A.; Ess, D. H.; McCafferty, D. G.;
Meyer, T. J. Chem. Rev. 2012, 112, 4016−4093.
(
(
29) Huynh, M. H.; Meyer, T. J. Chem. Rev. 2007, 107, 5004−5064.
30) Concepcion, J. J.; Brennaman, M. K.; Deyton, J. R.; Lebedeva,
N. V.; Forbes, M. D. E.; Papanikolas, J. M.; Meyer, T. J. J. Am. Chem.
Soc. 2007, 129, 6968−6969.
(
31) Lebedeva, N. V.; Schmidt, R. D.; Concepcion, J. J.; Brennaman,
M. K.; Stanton, I. N.; Therien, M. J.; Meyer, T. J.; Forbes, M. D. E. J.
Phys. Chem. A 2011, 115, 3346−3356.
2
422
dx.doi.org/10.1021/ja3106863 | J. Am. Chem. Soc. 2013, 135, 2419−2422