Binding of Ru(bpy)
2
(eilatin)
2+
to Matched and Mismatched DNA
Brian M. Zeglis
and
Jacqueline K. Barton
*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena,
California 91125
Abstract
The DNA-binding properties of Ru(bpy)
2
(eilatin)
2+
have been investigated to determine if the
sterically expansive eilatin ligand confers specificity for destabilized single-base mismatches in
DNA. Competitive DNA photocleavage experiments employing a sequence-neutral
metallointercalator, Rh(bpy)
2
(phi)
3+
(phi = 9,10-phenanthrenequinonediimine), and a mismatch-
specific metalloinsertor, Rh(bpy)
2
(chrysi)
3+
(chrysi = chrysene-5,6-quinonediimine), reveal that the
eilatin complex binds to a CC mismatched site with an apparent binding constant of 2.2(2) × 10
6
M
−
1
. Nonetheless, the selectivity in binding mismatched DNA is not high: competitive titrations with
Rh(bpy)
2
(phi)
3+
show that the complex binds also to well-matched B-form sites. Thus, Ru
(bpy)
2
(eilatin)
2+
, despite containing the extremely expansive eilatin ligand, displays lower selectivity
for the mismatch than does Rh(bpy)
2
(chrysi)
3+
, a metalloinsertor containing the smaller, though still
bulky, chrysene-5,6-quinonediimine ligand. In summary, the size and shape of the eilatin ligand
allow stacking with both well-matched and mismatched DNA.
Introduction
Octahedral metal complexes that bind DNA have received considerable attention based upon
their potential utility as nucleic acid probes and chemotherapeutics.
1
Our laboratory has long
focused on developing the DNA recognition capabilities of metal complexes, and recently we
have designed bulky octahedral complexes that specifically target single-base mismatches.
2
Single-base DNA mismatches occur in the cell as a result of polymerase errors, genotoxic
chemicals, or UV-induced damage.
3
Left unrepaired, mismatches can, upon replication, lead
to potentially harmful single-base mutations. In response to this continual threat to the integrity
of its DNA, the cell has evolved a complex enzymatic mismatch repair machinery that
recognizes and repairs these sites of damage as they are formed.
4
If this machinery is disabled,
however, mismatches and consequent mutations will accumulate in the genome, often with
dire consequences. Indeed, mutations in mismatch repair genes have been implicated in 80%
of hereditary nonpolyposis colon cancers.
5
Unlike sequence-specific metal complexes in which ancillary ligands bear the recognition
elements,
1
mismatch-specific metal complexes, most notably Rh(bpy)
2
(chrysi)
3+
(chrysi =
chrysene-5,6-quinonediimine) and Rh(bpy)
2
(phzi)
3+
(phzi = benzo[
a
]phenazine-5,6-
quinonediimine), depend upon the intercalating ligand for their recognition properties.
6
Aromatic steric bulk is the key. While the intercalating ligands of complexes that bind well-
matched DNA, such as Rh(bpy)
2
(phi)
3+
(phi = 9,10-phenanthrenequinonediimine), slide easily
* To whom correspondence should be addressed. E-mail: jkbarton@caltech.edu.
Supporting Information Available:
Rh(bpy)
2
(chrysi)
3+
binding constant determination gel and plot, Ru(bpy)
3
2+
/Rh
(bpy)
2
(chrysi)
3+
competition experiment gel, and Rh(bpy)
2
(chrysi)
3+
/Ru(bpy)
2
(eilatin)
2+
competition experiment plot. This material
is available free of charge via the Internet at http://pubs.acs.org.
NIH Public Access
Author Manuscript
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
Published in final edited form as:
Inorg Chem
. 2008 July 21; 47(14): 6452–6457. doi:10.1021/ic8006537.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
between the base pairs, the more sterically expansive ligands of mismatch-specific complexes
are simply too wide to fit into well-matched B-form DNA (Figure 1). The complexes bearing
these bulky ligands are, however, able to bind mismatched sites, which are thermodynamically
destabilized relative to well-matched DNA because of the impaired hydrogen bonding and
stacking of the mismatched bases. Rh(bpy)
2
(chrysi)
3+
, for example, has been shown to bind
80% of all mismatches in all possible sequence contexts. Moreover, this mismatch binding is
highly specific: the Rh complex was shown to target a single-base mismatch within a 2700 bp
DNA fragment.
7
More destabilized mismatches are bound more tightly than less destabilized
sites, with highly stable, G-containing sites escaping recognition altogether.
Importantly, recent structural studies have revealed the basis for this targeting and for the
correlation between the binding affinity and thermodynamic instability of a mismatch.
8
Interestingly, these complexes do not bind mismatches via traditional metallointercalation; Rh
(bpy)
2
(phi)
3+
, for example, binds from the major groove, inserts a ligand into the base stack,
and, in so doing, increases the base pair rise of the helix.
9
Instead, these bulkier metal complexes
bind the mismatched site through insertion of the expansive ligand into the duplex from the
minor groove so as to eject the mismatched bases into the major groove. For metalloinsertion,
the inserted ligand replaces the extruded bases in the DNA
π
stack.
8
Because this
metalloinsertion occurs via the narrow minor groove and yields no increase in the base-pair
rise, the binding site for the metal complex is sterically restrictive. Consequently, the binding
is enantiospecific for
Δ
-Rh(bpy)
2
(chrysi)
3+
: the right-handed helix can only accommodate the
right-handed enantiomer.
These bulky metalloinsertors have been employed successfully in a range of applications, yet
we continue to seek new mismatch-specific complexes in an effort to target the more stable
base mismatches.
10
,
11
We considered that augmenting the size of the bulky aromatic ligand
might provide this increase in mismatches targeted because a greater surface area for
π
stacking
might yield the boost in binding affinity required for the targeting of more thermodynamically
stable mismatched sites. Here we describe binding of Ru(bpy)
2
(eilatin)
2+
, a complex bearing
a singularly expansive ligand (Figure 1), to matched and mismatched duplex DNA.
Eilatin is a highly symmetric, heptacylic marine alkaloid first isolated in 1988 from the Red
Sea tunicate
Eudistoma
sp.
12
While the molecule itself has proven to be of significant interest
to both synthetic and biological chemists, it is, however, eilatin coordinated to an octahedral
metal complex that offers the possibility of high-affinity metalloinsertion. Ru
(bpy)
2
(eilatin)
2+
has been prepared and characterized spectroscopically by Kol and co-
workers.
13
Moreover, studies with nucleic acids by Tor and co-workers have revealed binding
to folded RNAs and nonspecific association with calf thymus DNA.
14
Our laboratory has
previously examined the binding of luminescent Ru complexes to DNA and RNA, most notably
the light-switch compound Ru(phen)
2
(dppz)
2+
(dppz = dipyridophenazine).
15
Yet here, our
interest is primarily predicated on the shape characteristics of the ligand and its potential
applications as a specific probe for mismatched DNA. Our studies show, however, that steric
bulk alone is insufficient to achieve site-specificity.
Experimental Section
Metal Complexes
Eilatin was synthesized from kynuramine (Aldrich) via the facile biomimetic pathway of
Kashman et al. and purified via preparatory thin-layer chromatography (silica, 96:4 CH
2
Cl
2
/
MeOH).
16
Ru(bpy)
2
(eilatin)
2+
was synthesized as reported and purified via an ion exchange
column (Aldrich Sephadex CM25) and reverse-phase high-performance liquid
chromatography [HP1100 HPLC system with Varian DynaMax C18 semipreparative column,
gradient of 15:85 TO 60:40 H
2
O (0.1% TFA)/MeCN (0.1% TFA) over 60 min].
13
No extinction
Zeglis and Barton
Page 2
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
coefficients in buffer have been reported. Inductively coupled plasma mass spectrometry was
employed to determine aqueous extinction coefficients of
ε
= 64 000 M
−
1
cm
−
1
(244 nm), 68
000 M
−
1
cm
−
1
(287 nm), and 38 000 M
−
1
cm
−
1
(426 nm) for [Ru(bpy)
2
(eilatin)]Cl
2
. Because
few enantioselective effects were evident in binding Ru(bpy)
2
(eilatin)
2+
to well-matched
oligonucleotides, racemates were employed in all experiments used here.
14a
,
b
Moreover,
dimerization or aggregation of Ru(bpy)
2
(eilatin)
2+
in solution over the concentrations utilized
for experiments does not appear to be an issue because, in this concentration range, no
deviations from Beer's law are found. Some aggregation associated with DNA binding may
arise however.
13
[Rh(bpy)
2
(chrysi)]Cl
3
and [Rh(bpy)
2
(phi)]Cl
3
were synthesized and purified
via published protocols
17
and also employed as racemic mixtures. All reagents and solvents
were used as received.
Oligonucleotide Synthesis and Photocleavage
Oligonucleotides were synthesized from phosphoramidites on an ABI 3400 DNA synthesizer.
Following synthesis, the oligonucleotides were purified both with and without dimethoxytrityl
protecting groups via reverse-phase HPLC [HP1100 HPLC system with Varian DynaMax C18
semipreparative column, gradient of 5:95 to 45:55 MeCN/50 mM NH
4
OAc(aq) over 30 min
for DMT-ON purification and 2:98 to 17:83 55 MeCN/50 mM NH
4
OAc(aq) over 30 min for
DMT-OFF purification]. All PAGE experiments described employed denaturing 20%
polyacrylamide gels (National Diagnostics) and were performed according to published
procedures.
17
DNA strands were radioactively labeled with [
γ
-
32
P]-ATP (MP Biomedicals)
using literature protocols and purified by 20% denaturing PAGE.
17
All irradiations were
performed using an Oriel Instruments solar simulator (300–440 nm). Gels were developed
using Molecular Dynamics phosphorimaging screens and a Molecular Dynamics Storm 820
phosphorimager and subsequently visualized and quantified with Molecular Dynamics
ImageQuant software.
Results and Discussion
Characterization of Site Selectivity of Ru(bpy)
2
(eilatin)
2+
Owing to the short excited-state lifetime of Ru(bpy)
2
(eilatin)
2+
, direct methods such as DNA
photocleavage or singlet oxygen sensitization could not be used to characterize the sites
targeted by the Ru complex within the DNA duplex.
18
Instead, competition experiments were
employed. We first utilized Rh(bpy)
2
(phi)
3+
, which binds duplex DNA with little site-
selectivity,
19
in order to probe where the Ru complex binds through competitive inhibition. A
synthetic 36-mer oligonucleotide was synthesized with complements featuring a guanine (EL-
M) or a cytosine (EL-MM) across from a central cytosine (bold) to form matched and
mismatched strands: 5
′
-CGCTACGTCTATATG
C
ATGATCCTAAGTGACAGTAC-3
′
.
After synthesis and purification, the reverse strand (not shown) was radioactively labeled
with
32
P-ATP at its 5
′
terminus via standard protocols.
17
Then, samples (1
μ
M) of radiolabeled
EL-M and EL-MM DNA in buffer (50 mM NaCl, 10 mM NaPi, pH 7.1) were incubated with
8
μ
M Rh(bpy)
2
(phi)
3+
and irradiated for 20 min using a solar simulator in the presence of
variable amounts of Ru(bpy)
2
(eilatin)
2+
. A concentration of 8
μ
M Rh(bpy)
2
(phi)
3+
provides 1
Rh molecule/4 base pairs, enough to saturate the entire oligonucleotide with Rh complexes.
Autoradiography of the resultant gel reveals that, with photoactivation, Rh(bpy)
2
(phi)
3+
promotes cleavage on the EL-M DNA at six discrete sites (with base numbers from 3
′
-end):
C19, G22, C27, C29, T32, and C33 (Figure 2). Interestingly, EL-MM DNA is cleaved at the
same locations by Rh(bpy)
2
(phi)
3+
but also displays two more cleavage bands: T13 and C16.
The C16 position is the mismatched site. The somewhat curious cleavage at T13 may result
from local conformational changes created by the nearby mismatch in the EL-MM sequence,
leading to hyperreactivity.
20
Zeglis and Barton
Page 3
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 2 also shows the effect of increasing Ru(bpy)
2
(eilatin)
2+
concentrations on Rh
(bpy)
2
(phi)
3+
photocleavage by gel autoradiography and a corresponding line plot. With
increasing concentrations of Ru(bpy)
2
(eilatin)
2+
, on both the matched and mismatched
duplexes, all of the Rh(bpy)
2
(phi)
3+
cleavage bands lessen in intensity, indicating that Ru
(bpy)
2
(eilatin)
2+
is competing with, and eventually inhibiting, Rh binding at all sites. At these
Ru concentrations, this nonspecific inhibition of Rh photocleavage cannot be accounted for
primarily through light absorption by the Ru complex but instead must reflect competitive
binding of the Ru complex to well-matched DNA sites. Increasing concentrations of Ru
(bpy)
3
2+
, a metal complex that binds DNA very weakly and has extinction coefficients similar
to those of Ru(bpy)
2
(eilatin)
2+
over the spectral range of interest, have no effect on the
photocleavage intensities of Rh(bpy)
2
(phi)
3+
and Rh(bpy)
2
(chrysi)
3+
in the salient
concentration range.
15
Moreover, this Ru(bpy)
3
2+
control also excludes the likelihood that Ru
(bpy)
2
(eilatin)
2+
inhibits Rh photocleavage by quenching of the Rh excited state. For the well-
matched duplex, photocleavage with 8
μ
M Rh(bpy)
2
(phi)
3+
is fully inhibited at
∼
15
μ
M Ru
(bpy)
2
(eilatin)
2+
. Nonspecific duplex binding occurring in the micromolar range is therefore
comparable for the two complexes.
Interestingly, however, on the mismatched duplex, site preferences for both Rh(bpy)
2
(phi)
3+
and Ru(bpy)
2
(eilatin)
2+
are evident. In the absence of Ru, Rh photocleavage on the mismatched
duplex is most intense at the mismatched site, C16. However, with increasing Ru, it is
photocleavage at this mismatched site that is preferentially inhibited, and cleavage at the
mismatch site is competed out at noticeably lower concentrations of Ru(bpy)
2
(eilatin)
2+
(
∼
5
μ
M). This differential inhibition is most evident in the line plot and gel quantitation of the
titration (Figure 2). The higher photocleavage for Rh(bpy)
2
(phi)
3+
in the absence of Ru actually
reflects a slightly higher affinity for the mismatched versus matched site, a common
characteristic of classical intercalators.
7b
,
21
Preferential inhibition of Rh photocleavage by Ru
(bpy)
2
(eilatin)
2+
may similarly reflect this preferential stacking with a mismatched site. Indeed,
the gel quantitation shows that binding to the mismatch is less than an order of magnitude
tighter than to matched sites. Curiously, the T13 cleavage site is also competed out well before
the other matched locations by the Ru complex. Because hyperreactivity of Rh(bpy)
2
(phi)
3+
at T13 depends on the nearby C16 mismatch, it appears it is similarly affected in competition
with Ru(bpy)
2
(eilatin)
2+
.
Determination of the Binding Affinity of Ru(bpy)
2
(eilatin)
2+
to the Mismatch
While competition with a nonspecific intercalator provides qualitative information about site
preference, quantitative data regarding site-specific affinity can be determined by competition
with the mismatch-specific metalloinsertor, Rh(bpy)
2
(chrysi)
3+
. For this second competition
experiment, a similar but shorter 5
′
-
32
P-labeled oligonucleotide was synthesized to minimize
binding to matched DNA. Complements containing a guanine and cytosine across from the
central cytosine (bold) were also synthesized to afford matched (ES-M) and mismatched (ES-
MM) duplexes: 5
′
-
32
P-TTAGGATCAT
C
CATATA-3
′
. A titration employing 1
μ
M
mismatched DNA in buffer (50 mM NaCl, 10 mM NaPi, pH 7.1) and variable Rh
(bpy)
2
(chrysi)
3+
was first used to obtain a mismatch-specific binding constant for the Rh
complex of 1.7(2) × 10
6
M
−
1
.
Given a known binding constant for Rh(bpy)
2
(chrysi)
3+
, the competition experiment yields
the quantitative binding affinity for the mismatched site. The competition experiment was
performed using 3
μ
M ES-MM DNA and 3
μ
M Rh(bpy)
2
(chrysi)
3+
in 50 mM NaCl, 10 mM
NaPi, and pH 7.1 along with increasing concentrations of Ru(bpy)
2
(eilatin)
2+
(0–20
μ
M). The
samples were then irradiated for 15 min on the solar simulator and subsequently eluted through
a denaturing polyacrylamide gel. The resultant gel clearly shows initially strong Rh
(bpy)
2
(chrysi)
3+
photocleavage at the mismatch site that is inhibited by increasing
Zeglis and Barton
Page 4
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
concentrations of Ru(bpy)
2
(eilatin)
2+
(Figure 3). From these titration data, we can extract a
CC mismatch-specific binding constant for Ru(bpy)
2
(eilatin)
2+
of 2.2(2) × 10
6
M
−
1
.
22
It is
interesting that the Ru affinity for this mismatched site is comparable to that of Rh
(bpy)
2
(chrysi)
3+
. Note that this value reflects binding to a 15-mer that contains additional
matched sites to which the Ru complex may also bind (albeit likely at higher Ru
concentrations). As a result, the binding affinity for the mismatched site must be considered
in the context of competition also with matched sites.
Implications for the Design of Bulky Metalloinsertors
Taken together, the two competition experiments described clearly indicate that while Ru
(bpy)
2
(eilatin)
2+
does show some preference in binding the CC mismatch, the bulky complex
also displays significant binding to well-matched B-form DNA sites. The site specificity of
the Ru complex for a mismatch is therefore significantly less than that of Rh
(bpy)
2
(chrysi)
3+
.
7
A comparison of the measured mismatched-site dissociation constant [
K
D
= 460(9) nM] to those obtained for matched sites supports this assertion; with matched DNA,
binding is in the low micromolar range.
14
Thus, the selectivity of the complex for mismatched
sites is modest (ratio of binding mismatched versus matched
≤
10). It is noteworthy that earlier
it was suggested that Ru(bpy)
2
(eilatin)
2+
may bind preferentially to large structural motifs in
folded RNAs.
14
Binding of the hydrophobic and large cationic Ru complex may arise with a
range of nucleic acid structures.
The ability of Ru(bpy)
2
(eilatin)
2+
to bind both matched and mismatched DNA prompts the
consideration of how the Ru complex may interact structurally with matched and mismatched
sites. Figure 4 shows schematic illustrations of Ru(bpy)
2
(eilatin)
2+
bound to mismatched DNA
in comparison to Rh(bpy)
2
(chrysi)
3+
and to matched DNA in comparison to Rh
(bpy)
2
(phi)
3+
. Binding of the Rh complexes to their target sites is based upon crystal
structures
8
,
9
and shows access from the minor groove side for metalloinsertion into a
mismatched site and from the major groove side for access by metallointercalation. For the Rh
complexes, it is apparent that these binding modes permit complete stacking of the inserting
ligand between the base pairs. Moreover, the ancillary ligands of the octahedral complexes
provide a barrier to both deeper insertion and significant rotation in the pocket. The complexes
are bound so that the dyad axis of the base pairs bisects the immine–Rh–immine angle. In this
mode, binding of the complexes is optimized for stacking, at both the mismatched and matched
sites. As illustrated in Figure 4, the Ru complex is also well situated within a mismatched site
for substantial stacking overlap, despite the large size of the eilatin ligand. Overlap with the
base pairs is quite comparable for the chrysi and eilatin ligands, consistent with their similar
binding affinity for the CC mismatch. Significantly for Ru(bpy)
2
(eilatin)
2+
, however, the
complex can still stack well within a matched site although rotated relative to the bound Rh
complex. The eilatin ligand is sufficiently expansive that substantial stacking is available
between the base pairs without a straight-on orientation of the complex. It is noteworthy that
we have seen previously for Ru(bpy)
2
dppz
2+
fluorescence and NMR results that are consistent
with a mixture of straight-on and side-on orientations in matched duplex DNA.
23
Here, at the
matched site, the eilatin complex can easily rotate within the intercalation site and maintain
significant overlap with the bases above and below. Indeed, the stacking area appears
comparable to that of the phi complex, just as their binding affinities for matched sites are
similar. The great expanse of the eilatin permits this significant stacking without the axial
ligands serving as a barrier to rotation. Thus, while binding to a mismatched site by Ru
(bpy)
2
(eilatin)
2+
is possible, binding to the matched site is not precluded.
These studies show that simply increasing the expanse of a metalloinsertor is not sufficient to
gain an increase in specific binding to a mismatched site in duplex DNA. While binding to the
mismatched site is well accommodated by a bulkier ligand, the increased expanse also provides
Zeglis and Barton
Page 5
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
a stacking area for the complex at a matched site if the ligand is particularly expansive. In a
comparison of the family of bulky metal complexes, similar affinities in binding-mismatched
DNA are observed for those bearing the phzi and chrysi ligands versus that containing the even
more expansive eilatin ligand. The stacking interaction is defined by the surface area of the
base pairs above and below the metalloinsertor, not just the size of the metalloinsertor. With
the eilatin complex, a substantial affinity and stacking area are also available at the matched
site. Thus, for this expansive complex, instead, we see that specificity for a single base-pair
mismatch is lost. That the eilatin ligand extends considerably from the metal center in two
directions is likely responsible for this loss in specificity for mismatched sites and gain in
affinity for matched DNA; these structural characteristics of Ru(bpy)
2
(eilatin)
2+
allow the
complex to bind matched DNA in a manner that Rh(bpy)
2
(chrysi)
3+
and Rh(bpy)
2
(phzi)
3+
cannot. As a consequence, then, these experiments teach us something simple about the design
of mismatch-recognition ligands: bulky is good, but bulkier is not necessarily better.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
We thank the NIH (Grant GM33309) for their financial support. We also thank Drs. Y. Kashman and Y. Tor for helpful
correspondence.
References
1. Erkilla KE, Odom DT, Barton JK. Chem Rev 1999;99:2777. [PubMed: 11749500]
2. Zeglis BM, Pierre VP, Barton JK. Chem Commun 2007:4565.
3. (a) Modrich P. Annu Rev Genet 1991;25:229. [PubMed: 1812808] (b) Kolodner R. Genes Dev
1996;10:1433. [PubMed: 8666228]
4. (a) Harfe BD, Jinks-Robertson S. Annu Rev Genet 2000;34:359. [PubMed: 11092832] (b) Modrich P.
J Biol Chem 2006;281:30305. [PubMed: 16905530]
5. (a) Kolodner RD. Trends Biochem Sci 1995;20:397. [PubMed: 8533151] (b) Arizamanoglou II, Gilbert
F, Barber HR. Cancer 1998;82:1808. [PubMed: 9587112] (c) Loeb LA, Loeb KR, Anderson JP. Proc
Natl Acad Sci U S A 2003;100:776. [PubMed: 12552134]
6. (a) Jackson BA, Barton JK. J Am Chem Soc 1997;119:12986. (b) Junicke H, Hart JR, Kisko J, Glebov
O, Kirsch IR, Barton JK. Proc Natl Acad Sci U S A 2003;100(7):3737. [PubMed: 12610209]
7. (a) Jackson BA, Alekseyev VY, Barton JK. Biochemistry 1999;38:4656. (b) Jackson BA, Barton JK.
Biochemistry 2000;39:6176. [PubMed: 10821692]
8. (a) Pierre VP, Kaiser JT, Barton JK. Proc Natl Acad Sci U S A 2007;104:429. [PubMed: 17194756]
(b) Cordier C, Pierre VP, Barton JK. J Am Chem Soc 2007;129:12287. [PubMed: 17877349]
9. Kielkopf CL, Erkilla KE, Hudson BP, Barton JK, Rees DC. Nat Struct Biol 2000;7:117–121. [PubMed:
10655613]
10. (a) Hart JR, Johnson MD, Barton JK. Proc Natl Acad Sci U S A 2004;101:14140. (b) Hart JR, Glebov
O, Ernst RJ, Kirsch IL, Barton JK. Proc Natl Acad Sci U S A 2006;103:15359. [PubMed: 17030786]
(c) Brunner J, Barton JK. J Am Chem Soc 2006;128:6772. [PubMed: 16719441] (d) Zeglis BM,
Barton JK. J Am Chem Soc 2006;128:5654. [PubMed: 16637630] (e) Schatzschneider U, Barton JK.
J Am Chem Soc 2004;126:8631.
11. (a) Ruba E, Hart JR, Barton JK. Inorg Chem 2004;43:4570. [PubMed: 15257584] (b) Petitjean A,
Barton JK. J Am Chem Soc 2004;126:14728. [PubMed: 15535691]
12. Rudi A, Benayahu Y, Goldberg I, Kashman Y. Tetrahedron Lett 1988;29:6655.
13. (a) Rudi A, Kashman Y, Gut D, Lellouche F, Kol M. Chem Commun 1997:17. (b) Gut D, Goldberg
I, Kol M. Inorg Chem 2003;42:3483. [PubMed: 12767184] (c) Bergman SD, Gut D, Kol M, Sabatini
C, Barbieri A, Barigelletti F. Inorg Chem 2005;44:7943. [PubMed: 16241144] (d) Gut D, Goldberg
Zeglis and Barton
Page 6
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
I, Kol M. Inorg Chem 2003;42:3483. [PubMed: 12767184] (e) Gut D, Rudi A, Kopilov J, Goldberg
I, Kol M. J Am Chem Soc 2002;124:5449. [PubMed: 11996586]
14. (a) Luedtke NW, Hwang JS, Glazer EC, Gut D, Kol M, Tor Y. ChemBioChem 2002;3:766. [PubMed:
12203975] (b) Luedtke NW, Hwang JS, Nava E, Gut D, Kol M, Tor Y. Nucl Acids Res 2003;31:5732.
[PubMed: 14500837]
15. (a) Friedman AE, Chambron JC, Sauvage JP, Turro NJ, Barton JK. J Am Chem Soc 1990;112:4960.
(b) Jenkins Y, Friedman AE, Turro NJ, Barton JK. Biochemistry 1992;31:10809. [PubMed: 1420195]
16. Gellerman G, Babad M, Kashman Y. Tetrahedron Lett 1993;34:1827.
17. Zeglis BM, Barton JK. Nat Prot 2007;2:357.
18. Chow CS, Barton JK. Methods Enzymol 1992;212:219. [PubMed: 1381460]
19. Pyle AM, Long EC, Barton JK. J Am Chem Soc 1989;111:4520.
20. Lim MH, Lau IH, Barton JK. Inorg Chem 2007;46:9528. [PubMed: 17918931]
21. Jackson, BA. PhD Dissertation. California Instititue of Technology; Pasadena, CA: 2000.
22. (a) Garbett NC, Chaires JB. Methods Cell Biol 2008;84:3. [PubMed: 17964926]Hart, JR. PhD
Dissertation. California Institute of Technology; Pasadena, CA: 2006.
23. Dupureur CM, Barton JK. J Am Chem Soc 1994;116:10286.
Zeglis and Barton
Page 7
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 1.
Chemical structures of eilatin, 9,10-chrysene-5,6-quinone diimine,
Δ
-Rh(bpy)
2
(chrysi)
3+
,
Δ
-
Rh(bpy)
2
(phzi)
3+
,
Δ
-Rh(bpy)
2
(phi)
3+
, and
Δ
-Ru(bpy)
2
(eilatin)
2+
. Width approximations made
using ChemDraw 3D with energy-minimized structures.
Zeglis and Barton
Page 8
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 2.
Competitive binding of Ru(bpy)
2
(eilatin)
2+
to matched and mismatched DNA monitored using
Rh(bpy)
2
(phi)
3+
photocleavage. Left: Denaturing polyacrylamide gel showing the competition
of Rh(bpy)
2
(phi)
3+
and Ru(bpy)
2
(eilatin)
2+
for matched (left) and mismatched (right) DNA of
the sequence 3
′
-GCGATGCAGATATAC
C
TACTAGGATTCACTGTCATG-
32
P-5
′
. All
samples were prepared with 1
μ
M DNA, 50 mM NaCl, 10 mM NaPi, and pH 7.1 and, unless
otherwise stated, irradiated for 20 min on a solar simulator. Left and right AG and CT lanes
are Maxam–Gilbert sequencing reactions for matched and mismatched DNA, respectively.
Lanes 1–10 employ matched DNA and lanes 11–20 mismatched DNA. Sample conditions:
lanes 1 and 11, 1
μ
M Rh(bpy)
2
(chrysi); lanes 2–10 and 12–20, 8
μ
M Rh(bpy)
2
(phi)
3+
. Lanes
3–10 and 13–20 also contain increasing amounts of Ru(bpy)
2
(eilatin)
2+
, beginning with 2.5
μ
M Ru(bpy)
2
(eilatin)
2+
in lanes 3 and 13 and increasing in increments of 2.5–22.5
μ
M in lanes
10 and 20. The arrow marks the mismatch site. Right top: Line plots of lanes 14, 16, 18, and
20 in gel. The arrow marks the mismatched site. Right bottom: Quantitation of the Rh
(bpy)
2
(phi)
3+
cleavage band intensity as a function of the Ru(bpy)
2
(eilatin)
2+
concentration.
Symbols: filled square, C16; empty square, C29; filled triangle, C27; empty triangle, C19.
Zeglis and Barton
Page 9
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 3.
Competitive binding of Ru(bpy)
2
(eilatin)
2+
to mismatched DNA monitored using Rh
(bpy)
2
(chrysi)
3+
photocleavage. Top: Denaturing polyacrylamide gel of a competition
experiment between Rh(bpy)
2
(chrysi)
3+
and Ru(bpy)
2
(eilatin)
2+
for a CC mismatch in the
oligonucleotide 5
′
-
32
P-TTAGGATCAT
C
CATATA-3
′
. AG and CT lanes are Maxam–Gilbert
sequencing reactions. All samples contained a 3
μ
M mismatched duplex in a buffer of 50 mM
NaCl, 10 mM NaPi, and pH 7.1 and were irradiated for 10 min using a solar simulator unless
otherwise stated. Sample conditions: lane 1, DNA only irradiated without Rh; lane 2, 3
μ
M Rh
(bpy)
2
(chrysi)
3+
without irradiation; lane 3, 3
μ
M Ru(bpy)
2
(eilatin)
2+
irradiated without Rh;
lanes 4–16, 3
μ
M Rh(bpy)
2
(chrysi)
3+
and increasing concentrations of Ru(bpy)
2
(eilatin)
2+
, 0,
Zeglis and Barton
Page 10
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
0.1, 0.33, 0.66, 1, 2, 3, 5, 10, 20, 33, 66, and 100
μ
M, respectively. Wide photocleavage bands
do not reflect nonspecific photocleavage at more than one site but rather the multiple products
produced by hydrogen abstraction upon photoactivated cleavage at the mismatched site.
Bottom: Plot of a fraction cleaved against the Ru(bpy)
2
(eilatin)
2+
concentration for four trials
of the competition experiment.
Zeglis and Barton
Page 11
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 4.
Schematic illustrations of Ru(bpy)
2
(eilatin)
2+
(right) bound to mismatched (top) and matched
(bottom) DNA sites based on the crystal structures of chrysi (top left) and phi (bottom left)
complexes of Rh bound to mismatched and matched DNA, respectively. For binding to the
mismatched site, the metal complexes are oriented from the minor groove side, whereas for
binding to the matched site, the association is from the major groove side.
Zeglis and Barton
Page 12
Inorg Chem
. Author manuscript; available in PMC 2009 September 21.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript