DNA-binding studies of a tetraalkyl-substituted porphyrin and the mutually adaptive distortion principle

Article abstract of DOI:10.1021/ic501683t

This investigation explores DNA-binding interactions of various forms of an alkyl-substituted cationic porphyrin, H₂TC3 (5,10,15,20-tetra[3-(3′-methylimidazolium-1′-yl)]-porphyrin). The motivating idea is that incorporating alkyl rather than aryl substituents in the meso positions will enhance the prospects for intercalative as well as external binding to DNA hosts. The ligands may also be applicable for photodynamic and/or anticancer therapy. Methods employed include absorbance, circular dichroism, and emission spectroscopies, as well as viscometry and X-ray crystallography. By comparison with the classical H₂T4 system, H₂TC3 exhibits a higher molar extinction coefficient but is more prone to self-association. Findings of note include that the copper(II)-containing form Cu(TC3) is adept at internalizing into single-stranded as well as B-form DNA, regardless of the base composition. Surprisingly, however, external binding of H₂TC3 occurs within domains that are rich in adenine-thymine base pairs. The difference in the deformability of H₂TC3 versus Cu(TC3) probably accounts for the reactivity difference. Finally, Zn(TC3) binds externally, as the metal center remains five-coordinate.

Full text of DOI:10.1021/ic501683t

Article
pubs.acs.org/IC
DNA-Binding Studies of a Tetraalkyl-Substituted Porphyrin and the
Mutually Adaptive Distortion Principle
Srijana Ghimire, Phillip E. Fanwick, and David R. McMillin*
Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907, United States
*
S Supporting Information
ABSTRACT: This investigation explores DNA-binding interactions of various forms of an alkyl-
substituted cationic porphyrin, H TC3 (5,10,15,20-tetra[3-(3′-methylimidazolium-1′-yl)]-
2
porphyrin). The motivating idea is that incorporating alkyl rather than aryl substituents in the
meso positions will enhance the prospects for intercalative as well as external binding to DNA
hosts. The ligands may also be applicable for photodynamic and/or anticancer therapy. Methods
employed include absorbance, circular dichroism, and emission spectroscopies, as well as
viscometry and X-ray crystallography. By comparison with the classical H T4 system, H TC3
2
2
exhibits a higher molar extinction coefficient but is more prone to self-association. Findings of
note include that the copper(II)-containing form Cu(TC3) is adept at internalizing into single-
stranded as well as B-form DNA, regardless of the base composition. Surprisingly, however,
external binding of H TC3 occurs within domains that are rich in adenine−thymine base pairs.
2
The difference in the deformability of H TC3 versus Cu(TC3) probably accounts for the
2
reactivity difference. Finally, Zn(TC3) binds externally, as the metal center remains five-
coordinate.
INTRODUCTION
telomerase inhibitors as well, cationic porphyrins also have
■
1
0,11
potential applications in anticancer therapies.
Research has
The aim of this investigation was to explore the DNA-binding
interactions of new cationic porphyrins with alkyl functional
groups in the meso positions. The present dearth of studies
dealing with alkyl-substituted forms is surprising in light of the
long history of DNA-binding studies with aryl-substituted
established that the porphyrin ligands adopt at least three
distinct binding motifs when interacting with double-stranded
1
,2,12
(
ds) DNA hosts.
The mode of binding is usually evident
from physical studies. For example, intercalation between base
pairs typically gives rise to a bathochromic shift of 10−15 nm in
the Soret region and a hypochromic response of about 30%. In
the same spectral region it also gives rise to an induced circular
dichroism (iCD) having a negative sign. In contrast, external
binding generally gives rise to a positive iCD signal, a smaller
bathochromic shift, and a much weaker, or even negative
hypochromic response. Finally, some very hydrophobic
porphyrins, such as TΘOPP in Chart 1, bind by aggregating
porphyrins, most especially the H T4 system depicted in Chart
2
1
−3
1.
Cationic porphyrins like H T4 are useful because they
2
4,5
enter cells readily. They also exhibit prominent absorption in
the red end of the visible system, which means they can
6
−8
function as sensitizers for photodynamic therapy.
application could be for topical antibacterial treatment. Being
One
9
Chart 1
1
,13,14
and/or stacking on the surface of the DNA host.
External
stacking often induces a conservative, sometimes very intense
bisignate iCD signal in the Soret region of the visible
15
spectrum. The factors that influence the choice of binding
motif are becoming clearer. The amphiphilic nature of the
ligand seems ideally suited for intercalative binding, which
allows the hydrophobic core to sandwich between DNA bases
with positively charged substituent groups of the porphyrin
extending outward toward the sugar−phosphate backbones of
the DNA host and into solution. With this mode of binding the
Watson−Crick hydrogen-bonding framework of the host also
remains intact. However, steric forces also come into play. If,
for example, the central metal of a metalloporphyrin retains one
or more axial ligands, intercalative binding is no longer
Received: July 14, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
3
possible. An X-ray structural study has also established that
host, and a ligand that can undergo a mutually adaptive
intercalation of Cu(T4) inevitably leads to clashes in the minor
distortion (MAD) enhances the induced fit.
groove involving the bulky N-methylpyridiniumyl substituents
1
6
of the porphyrin and backbone residues of the host.
EXPERIMENTAL SECTION
■
Accordingly, Cu(T4) and its analogues only intercalate into
Materials. Hexanes, dichloromethane (DCM), dimethylformamide
DMF), acetone, methanol (MeOH), toluene, acetonitrile (MeCN),
GC-rich domains where hydrogen bonding is strong, and the
(
3
,17−19
duplex structure is durable.
The alternative is external
and potassium nitrate (KNO ) were products of Mallinckrodt
3
binding, but H T4 has an extended structure with a rigidly
2
Chemicals. Macron Chemicals supplied acetic acid and nitric acid,
but the hydrochloric acid came from J. T. Baker. Silica was a product
of Sorbent Technologies, whereas alumina was from EMD Chemicals.
Integrated DNA Technologies was the provider for both single-
stranded (ss) and hairpin-forming DNA (ds) sequences. The
sequences obtained were 5′-GATTACttttGTAATC-3′ (GATTAC),
disposed charge distribution. At least in part for that reason
high-affinity external binding therefore requires significant
reorganization of the DNA structure and formation of a
1
3,20
suitable binding pocket.
As a consequence external binding
is most feasible in AT-rich domains, which have a weaker
5
′ - G A C G A C t t t t G T C G T C - 3 ′ ( G A C G A C ) , 5 ′ -
hydrogen-bonding framework and a lower melting temperature.
GCGCACttttGTGCGC-3′ (GCGCAC), and 5′-AGC-
GACttttGTCGCT-3′ (AGCGAC), where lower case letters designate
bases involved in loop formation. Similarly the ss sequences were 5′-
TCCTGCCACGCTCCGC-3′ (Puc) and 5′-TTTTTTTTTT-3′
(T ). Indium(III) chloride, trizma chloride, trizma base, pyrrole,
Many variants of H T4 have been explored. Early on, for
example, Sari et al. systematically tuned the charge by replacing
N-methylpyridiniumyl groups with phenyl groups. Not
surprisingly, they found that reducing the charge also lowers
the binding affinity. Others have kept the charge constant while
enhancing the hydrophobic character by incorporating addi-
2
2
1
1
0
tetrabutylammonium nitrate (TBAN), potassium hexafluorophosphate
(KPF ), silica thin-layer chromatography (TLC) plates, and silanizing
6
22
solution (5% dichlorodimethylsilanes in n-heptane) were products of
Sigma-Aldrich Commercial. Other materials supplied by Sigma-Aldrich
included 4-chloro-1-butanol, 1-methylimidazole, pyridinium chloro-
tional aromatic groups in the meso substitutents. Neither
approach addresses two basic limitations, however. The first is
that aromatic substituents are rigid as well as bulky, giving rise
chromate (PCC), para-toluenesulfonic acid (PTSA·H O), 2,3,5,6-
1
6
2
to previously mentioned steric clashes in the minor groove.
tetrachloro-1,4-benzoquinone (TCQ), copper acetate (Cu(OAc)2),
Zn(OAc) , KMnO , and salmon testes (ST) DNA with an average
Second, periplanar interactions involving the β hydrogens of
2
4
the porphyrin ring require the substituents to orient essentially
length of 300 base pairs.
Synthesis. 4-Chlorobutanal. Commercially available 4-chloro-1-
2
3
perpendicularly with respect to the plane of the porphyrin.
That, in turn, can inhibit stacking with DNA bases, particularly
when the host site has a large footprint, as is the case with a
butanol reacts with PCC in anhydrous DCM at room temperature to
3
0
give 4-chlorobutanal. Because of poor solubility it helps to suspend
the PCC in DCM about 20 min before the addition of the alcohol.
TLC on silica by elution with 2:1 hexane/diethyl ether is a good way
of monitoring the progress of reaction. Dipping the TLC plate into a
11
leaflet of G-quadruplex DNA. Another approach involves
decreasing the number of N-methylpyridiniumyl substituents,
2
4−27
thereby reducing the steric problems.
worked quite well, and Cu(tD4) and Pd(tD4) turn out to be
This approach has
KMnO solution and drying with a heat gun reveals two spots with
4
different retention factor values. Complete conversion of the R spot of
2
6,28,29
f
universal intercalators for B-form DNA (Chart 1).
In
alcohol into the faster-moving R spot of aldehyde indicates the
f
1
other words, both ligands bind by intercalation regardless of the
base composition of the DNA. One of the most intriguing
completion of the reaction. H NMR in CDCl : 9.94 ppm (s, 1H),
3
3.73 ppm (t, 2H), 2.80 ppm (m, 2H), 2.23 ppm (t, 2H).
observations is that Pd(tD4) is superior to H T4 analogues in
H TC3. The first step in the synthesis of 5,10,15,20-tetra-(3-
2
2
29
chloropropyl)porphyrin (H TC3) is the condensation of 4-chlor-
sensitizing the formation of singlet oxygen. A downside is that
trimming the number of substituents reduces the net charge
and the solubility of the ligand in aqueous solution. In the
design adopted herein the ligand retains a net charge of 4+. The
motivation behind incorporating alkyl substituents is at once to
reduce the effective size of the porphyrin and develop a form
that is more conducive to stacking with DNA bases. In future
work it will also be possible to vary the nature of the charging
groups as well as the length of the tether. Here, the focus is on
H TC3 and metal-containing derivatives, where H TC3
2
obutanal with pyrrole in the presence of p-toluene sulfonic acid as
32
catalyst. The second step is oxidation with TCQ at ∼110 °C.
A
Dean−Stark trap is present to collect the water and maintain
anhydrous conditions. Absorption spectroscopy and TLC on silica
with 1:1 hexane/ethyl acetate are useful for following reaction
progress. Column chromatography (elution from silica with 3:1
hexane/ethyl acetate) is the main purification step. The dark pink,
1
slow-moving band is the desired product. H NMR in CDCl
: 9.53
3
ppm (s, 8H), 5.12 ppm (m, 8H), 3.87 ppm (m, 8H), 2.97 ppm (m,
8
H), −2.79 ppm (s, 2H).
Substitution of chloro groups with 1-methyl imidazole by the
2
2
denotes 5,10,15,20-tetra[3-(3′-methylimidazolium-1′-yl)]-
porphyrin; see Chart 1 for a view of the copper(II)-containing
form Cu(TC3). The binding studies reported herein employ
single- and double-stranded (ds) DNA hosts. The ds hosts
include salmon testes DNA as well as a series of hairpin-
forming oligonucleotides, vide infra, which have stems that vary
method of Wu et al. gives the desired tetra-cationic product
,10,15,20-tetra[(3-(3′-methylimidazolium-1′-yl)propyl)]porphyrin
5
33
(H TC3). The reaction requires about a week to complete. Column
2
chromatography (silica in 8% MeCN, 10% aqueous KNO , and 10%
3
deionized H O) allows separation from incompletely modified
2
3
4
porphyrin. The slowest-moving band corresponds to the desired
tetracation. After evaporation of solvents, aqueous KPF addition
in base composition. Interestingly, Cu(TC3) and H TC3 tend
6
2
precipitates the porphyrin from acetonitrile. Ion exchange to the
nitrate salt is possible by dissolving the hexafluorophosphate salt in
acetonitrile followed by precipitating with a solution of tetrabuty-
lammonium nitrate in acetone. Vapor diffusion of tetrahydrofuran into
a solution of either the nitrate in methanol or the hexafluorophosphate
form in acetonitrile gives small crystals of the desired product.
Analysis. Anal. Calcd for C H F N P : C 41.69, H 4.23, and N
to adopt different binding motifs, even though each is a
nominally planar porphyrin. With every host investigated
Cu(TC3) binds strictly by intercalation, whereas external
binding of H TC3 becomes an increasingly competitive process
2
as the percentage of AT base pairs increases in the host. The
difference in the deformability of the two forms presumably
accounts for the change in the mode of binding. High-affinity
external binding requires a significant reorganization of the
48
58 24 12 4
1
12.15%; found: C 40.81, H 4.17, and N 12.35%. H NMR in
deuterated dimethyl sulfoxide: 9.74 ppm (s, 8H), 9.22 ppm (d, 4H),
B
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
7
.96 ppm (d, 4H), 7.78 ppm (d, 4H), 5.05 ppm (m, 8H), 4.74 ppm
intensity comparisons by correcting for absorbance differences
between samples.
(m, 8H), 3.87 ppm (m, 12H), 2.93 ppm (t, 8H).
Cu(TC3). Dissolving the nitrate salt of H TC3 in 50/50 MeOH/
2
Iobs
1 − 10−
H O (v/v) and treating with dilute acid removes any traces of zinc(II)
Inorm = (
2
A)
29
from the porphyrin. Addition of aqueous KPF solution precipitates
(2)
6
the porphyrin. Metal insertion occurs open exposure to copper acetate
in DMF for 12 h. It is possible to monitor reaction progress by
absorption and emission spectroscopy. Metal insertion increases the
symmetry of a molecule and reduces the number of Q bands from four
to two. The emission signal drops to zero in DMF with the formation
of CuTC3. Anal. Calcd for C H CuF N P : C 39.91, H 3.91 and N
In eq 2, Inorm is the adjusted intensity, Iobs is the observed emission
intensity, and A is the absorbance at the exciting wavelength.
Conversion of the CD signal from millidegrees into molar absorptivity
units is possible with eq 3, where θ is the observed value, Q = 32 980, l
is the path length of the cell in centimeters, and c is the molar
concentration of the chromophore.
48
56
24 12 4
11.64%; found: C 40.01, H 4.00, and N 11.51%.
Zn(TC3). Addition of 2 equiv of zinc acetate in water to a solution of
θ
Δε =
the hexafluorophosphate salt of H TC3 in an equal volume of
2
Qlc
(3)
acetonitrile leads to the formation of the zinc derivative of the
porphyrin. Absorption spectroscopy is once again useful for
monitoring the reaction progress, and insertion is complete when
only two Q bands are present.
Sonicated ST DNA at a base-pair concentration of 70 uM was the host
used for viscometry studies. The porphyrin stock solution was made in
deionized water with no salt added. The monitoring temperature was
2
5 °C. Equation 4 gives the standard reduced viscosity ratio,
Methods. Silanization of the glassware minimized porphyrin
35
absorption on surfaces. In titrations, the porphyrin concentration
remained constant at 1.00 μM, and only the DNA concentration
changed. In the absence of DNA host, the medium used for measuring
spectra of porphyrins was pure methanol to avoid aggregation. But
when DNA was present in the sample, the only source of methanol
was a small amount introduced with the porphyrin stock solution.
Beer’s law studies yielded the molar extinction coefficients of the
porphyrins, needed for calibrating all stock solutions. For consistency
with the ST host, the unit of DNA used for the hairpins is a base pair.
Thus, the molar extinction coefficients for the hairpin hosts in Table 1
η
(t − t )
c
b
=
η0
(t − t )
d
b
(4)
where t is the flow time of buffer, t is the flow time of DNA in buffer,
and t is the flow time of DNA with porphyrin in buffer. The flow time
was determined by taking the average of three consecutive runs for
each composition.
b
d
c
Crystallography. A red needle of C H N ·4PF ·CH CN having
48
58 12
6
3
approximate dimensions of 0.22 × 0.14 × 0.10 mm was mounted on a
fiber in a random orientation to determine the crystal structure.
Preliminary examination and data collection were performed with Cu
Kα radiation (λ = 1.541 84 Å). Cell constants for data collection were
obtained from least-squares refinement, using the setting angles of 70
a
Table 1. List of Molar Extinction Coefficients
1
species
ε (260 nm, M− cm−1
)
1
36 reflections in the range of 2 < θ < 66°. The space group was
determined by the program XPREP. The structure was solved by
direct methods using SIR2004. See Table 2 for crystal data and
Figure 1 for a representation of the molecular structure of H TC3,
37
double-stranded hosts
GATTAC
38
18 300
18 300
17 200
18 600
GACGAC
GCGCAC
AGCGCA
ST
2
complete with an atom-numbering scheme.
Instrumentation. A Varian Cary 300 UV−vis spectrophotometer
yielded absorbance data. Similarly emission data came from a Varian
Cary Eclipse fluorescence spectrophotometer. Jasco J-810 and J-1500
3
6
13 200
single-stranded hosts
Puc
1
spectropolarimeters yielded circular dichroism data. H NMR data
b
130 300
came from a 300 MHz Varian Mercury Inova spectrometer. A
modified Cannon-Fenske model 25 viscometer yielded viscometric
data, and the pH meter was a Corning model 430. Midwest Microlab,
b
T10
81 600
porphyrins
c
5
5
H TC3
5.45 × 10
6.00 × 10
2
d
Cu(TC3)
Table 2. Crystal Data and Data Collection Parameters for
a
b
c
Base-pair units except for the ss hosts. Strand basis. Soret
C H N ·4PF ·CH CN
4
8
58 12
6
3
d
wavelength of 413 nm. Soret wavelength of 412 nm.
formula
C H F N P
50 61 24 13 4
formula weight
space group
a, Å
1423.99
P1 (No. 2)
̅
are the strand values quoted by the supplier divided by 8. In contrast,
for an ss host the unit of DNA is the complete strand. For each step in
a titration, q designates the ratio of the host concentration to that of
the porphyrin. To facilitate equilibration of the sample during a serial
titration, for each value of q the procedure was to add half the volume
of buffer needed, followed in order by aliquots of salt solution, DNA,
and porphyrin before adding the rest of the buffer. The buffer for the
binding studies was pH = 7.5 Tris-chloride. The solutions contained
12.1488(4)
14.8985(5)
19.1681(14)
72.763(5)
71.762(5)
70.043(5)
3026.7(3)
2
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
0.05 M chloride from the buffer and 0.10 M NaCl, or 0.05 M NaCl in
Z
the case of viscometry runs.
dcalc, g cm−3
1.562
Equation 1 yielded the percent hypochromism (% H), where A(λ)
is the absorbance at the Soret maxima of the free porphyrin and A(λ′)
is the corresponding absorbance of the bound form.
temperature, K
linear abs coef, mm−
150
1
2.264
2θ range, deg
4.97−133.15
70 136
A(λ) − A(λ′)
data collected
unique data
R(Fo)
%H =
× 100
A(λ)
(1)
9558
0.071
For luminescence studies of Cu(TC3), the slit settings were 20 nm for
excitation and emission. Normalizing the data using eq 2 facilitates
2
R (F )
0.201
w
o
C
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
DNA Titrations with Cu(TC3). Absorbance measurements
clearly signal that Cu(TC3) readily binds to DNA hosts in
aqueous solution. Moreover, data compiled in Table 3 show
that the base composition of the host makes little difference.
Illustrative results obtained with the hairpin host GATTAC
appear in Figure 3; they reveal that interaction with the host
induces a significant hypochromic shift as well as a strong
bathochromic response in the Soret band. Here, the
progression of the absorbance changes indicates that the
chromophore actually experiences a succession of binding
4
1
environments during the titration. In particular, the
hypochromic response is strongest in the early stages, prior
to attainment of the limiting spectrum, which settles in with a
bathochromic shift of Δλ = 10 nm and a hypochromic response
of H = 38% (Table 3). Band shifts of those magnitudes are
usually indicative of intercalative binding to ds DNA, both
effects resulting from coupling of porphyrin absorption with
42
electronic transitions of DNA bases. Corresponding changes
in the DNA absorbance ought to be present as well, but they
are usually harder to observe on account of the density of states
43
in the UV region of the spectrum. When the DNA takes up
Cu(TC3), however, a hyperchromic effect is clearly evident in
the 260 nm region of the spectrum (Figure 4). (A referee
points out, however, that the weak porphyrin absorbance in this
region of the spectrum, evident in Figure 2, could also be a
factor.) Reliable estimation of binding constants is problematic
because of the poor solubility of the free porphyrin and the
cooperative interactions between bound chromophores.
However, the titration results in Figure 3 suggest that the
uptake of Cu(TC3) by DNA is roughly as facile as that of
Figure 1. Representation of the cation in [H TC3)](PF ) ·CH CN
with thermal ellipsoids set at 50% probability. Hydrogen atoms
omitted for clarity.
2
6
4
3
20,41
LLC (Indianapolis, IN) carried out all microanalyses. The
diffractometer was Rigaku Rapid II equipped with confocal optics.
RESULTS
■
Cu(T4). By that reckoning, the dissociation constant K would
d
be on the order of 10− M with DNA in units of base pairs.
6
44
Characterization of Porphyrins. Unlike the more rigid
H T4 analogue, H TC3 is disadvantageously prone to
2
2
The binding interactions manifest themselves in other
physical studies as well. The observation of luminescence
from Cu(TC3) is particularly telling because the free porphyrin
is essentially nonemissive in solution, while a relatively strong
aggregation in aqueous solution, as has been reported for
tetra-substituted porphyrins with a pyridiniumyl ion extending
39
off each alkyl chain. In methanol, on the other hand, a Beer’s
law plot establishes that H TC3 exists as a monomeric ion and
emission signal is a clear indication of intercalative binding
2
5
−1
2,18,20,29
exhibits a Soret maximum at 413 nm where ε = 5.45 × 10 M
cm . Cu(TC3) behaves similarly. It has a Soret maximum at
12 nm as well as a high molar absorptivity of 6.0 × 10 M
cm . Figure 2 provides a contrast of the absorption spectra of
(Figure 5).
For perspective, note that Cu(T4) binds
−
1
externally to the AT-rich GATTAC host and is non-
5
−1
28
4
emitting, whereas Cu(TC3) is emissive because it binds by
−1
intercalation. Indeed, the signal obtained is 60% stronger than
that produced by Cu(T4) when it intercalates into GACGAC,
which contains a higher percentage of GC base pairs. Along
the same line, results in Table 3 show that the emission signal
from Cu(TC3) generally increases in intensity as the rigidity of
the host and the percentage of GC base pairs increase.
Adduct formation with a host like GATTAC also brings about
an iCD signal from otherwise CD-silent Cu(TC3). Although
the iCD signals obtained are typically bisignate, the positive
branch is comparatively weak and mostly occurs at shorter
wavelengths (Table 3), as when GACGAC acts as host. When
Cu(TC3) binds to ST DNA, the iCD signal is also bisignate;
however, the pattern inverts, and the negative branch shifts to
the shorter wavelength side.
Figure 2. Absorbance spectra of Cu(TC3) in methanol and Cu(T4) in
aqueous buffer.
DNA Titrations with H TC3. There are obvious differences
2
in the binding interactions of the unmetalated form of H TC3,
2
Cu(TC3) and Cu(T4). The broadened bandwidth of the latter
presumably relates to the pyridiniumyl substituents, which can
assume a distribution of torsion angles. The same effect is
despite the fact that it, too, is a nominally planar porphyrin. For
GC-rich hosts like GACGAC, however, the spectral changes
observed with H TC3 mimic those of Cu(TC3) in that the
2
sometimes evident in the emission spectrum of H T4, albeit in
Soret band undergoes a large red shift, and the negative band of
the iCD spectrum is dominant. On the other hand, as the host
shifts to ST DNA and then GATTAC, the bathochromic shift
in the limiting spectrum becomes progressively smaller (Figure
2
40
the Q-band region of the electronic spectrum. In terms of
emission neither Cu(T4) nor Cu(TC3) exhibits a detectable
signal in methanol or methanol/water mixtures.
D
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 3. Physical Data Obtained with ds Hosts
porphyrin
DNA host
absorbance
emission
iCD
Δε, M− cm−1
1
Δλ, nm
H, %
λ, nm
Cu(TC3)
GACGAC
GATTAC
10
45
5.5
5.5
419
4
−12
7
4
27
419
25
10
38
4
−15
−23
−14
−35
8
GCGCAC
AGCGCA
ST
10
10
10
43
43
35
7.8
5.9
4.2
424
425
422
4
30
Zn(TC3)
GACGAC
GATTAC
4
2
45
40
424
425
13
6
4
32
−6
7
ST
1
9
32
43
427
416
H TC3
2
GACGAC
16
4
26
418
25
−27
26
GATTAC
6
29
4
−7
−35
−15
18
GCGCAC
AGCGCA
ST
9
9
8
54
39
43
424
425
416
4
26
−24
Figure 5. Absorbance-corrected emission spectra of Cu(TC3) in the
absence of DNA (black) and in the presence of the DNA hairpin
GATTAC at q = 20 (dashed) and 120 (gray), which are the DNA
concentrations in μM of base pairs. For the q = 0 spectrum only the
solvent is methanol.
Figure 3. Absorption spectra of 1.0 μM Cu(TC3) at q = 0 and in the
presence of the DNA hairpin GATTAC at q = 20, 40, 80, and 120,
respectively, which are the DNA concentrations in μM of base pairs.
For the q = 0 spectrum the solvent is methanol.
Figure 6. Absorption spectra of 1.0 μM H TC3 at q = 0 (black) and in
2
the presence of large excess of GATTAC (gray), ST (dashed), or
GACGAC (dotted) DNA respectively. For the q = 0 spectrum only
the solvent is methanol.
Figure 4. DNA absorbance changes during the addition of GATTAC
to 1.0 μM Cu(TC3). Each plot represents a 20 μM aliquot of DNA
and is the difference spectrum between two consecutive runs, those
being q = 0 and 20 (gray), q = 20 and 40 (black), and q = 40 and 60
(
dashed). The hyperchromic effect recedes at higher q values because
ST DNA is the host. For this reason the H% values reported in
Table 3 sometimes exaggerate the actual hypochromic
response, as the calculations ignore any change in band
shape. (If one calculates the percent hypochromicity by
comparing areas under the curves in Figure 6, the responses
no porphyrin is available to interact with the added DNA.
6
), and the positive branch of the iCD spectrum increases in
importance. The Soret band also broadens, particularly when
E
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
are 3, 14, and 27 for the GATTAC, ST, and GACGAC hosts,
respectively.)
hypochromic effect is actually a few percent larger with T10
(Table 4). Interaction with T also induces an emission signal
10
DNA Titrations with Zn(TC3). Titrations reveal that the
zinc-containing analogue Zn(TC3) binds to DNA as well. At
least two stages of binding are evident in a titration with
GACGAC. First, at a base-pair-to-Zn(TC3) ratio of q = 8, the
Soret band exhibits sizable bathochromic and hypochromic
responses (Figure 7). Addition of more host initially results in a
Table 4. Physical Data Obtained with ss Hosts
DNA
host
porphyrin
absorbance
emission iCD
λ, nm Δε, M− cm−1
1
Δλ, nm H, %
Cu(TC3)
Zn(TC3)
T10
Puc
T10
Puc
T10
Puc
8
9
3
4
5
4
33
38
30
44
33
44
3.8
2.0
421
421
421
426
431
426
21
12
6
3
6
H TC3
2
from Cu(TC3) that is ∼70% as intense as that obtained with
GATTAC. With the purine-containing sequence Puc the
bathochromic shift is a bit larger at 10 nm, but the emission
signal is weaker by about a factor of 2. In the case of H TC3 the
2
bathochromic shifts are smaller than those observed for
Cu(TC3), but the hypochromic responses are of comparable
magnitudes. The smallest Δλ values result with Zn(TC3), but
they are about equal in magnitude to those obtained with ds
DNA hosts. In comparison with results obtained with ds DNA,
the biggest departure occurs in the iCD signals. With the ss
hosts the iCD signals are essentially monosignate and strictly
positive for all three porphyrins. Cu(TC3) exhibits the most
Figure 7. Absorption spectra of 1.0 μM Zn(TC3) at q = 0 (black) and
in the presence of the DNA hairpin GACGAC at q = 8 (gray), 24
(
dashed), and 48 (dot dash), respectively. For the q = 0 spectrum only
the solvent is methanol.
weakening of the hypochromic effect, but in the presence of a
large excess of DNA it strengthens once again, before leveling
off at ∼q = 48. Figure 8 is an overlay of the iCD signals
intense iCD signals and H TC3 the weakest (Table 4).
2
Viscometry and High-Concentration Solutions. Visco-
metric data obtained for Cu(TC3) and Zn(TC3) are different.
For these experiments the porphyrin concentrations are much
higher than they are in the titration studies and range up to 42
μM. Plots in Figure 9 show how the standard reduced viscosity
Figure 8. Induced CD spectra of 1.0 μM Zn(TC3) when bound to the
DNA hairpin GACGAC at DNA-to-porphyrin ratios of q = 8 (line)
and 48 (dot dash).
Figure 9. Standard viscosity ratios of ST DNA in the presence of
Cu(TC3) (●), H TC3 (□), and Zn(TC3) (◇). The DNA
2
obtained at the two ends of the titration with GACGAC (q = 8
and q = 48). A titration with ST DNA reveals similar
complexity. Thus, the hypochromic response steadily strength-
ens until q = 30, but the bathochromic shift sets in late,
between q = 20 and q = 50. At the q = 5 stage there is no
detectable iCD signal, but a weak positive signal is apparent at q
concentration remains constant at 70 μM in base pairs, while the
porphyrin-to-DNA pair ratio, R, varies. The buffer is μ = 0.10 M pH
7.5 Tris.
−1
varies with the porphyrin-to-DNA base-pair ratio, R = q ,
when Cu(TC3) or Zn(TC3) interacts with sonicated ST DNA
in Tris buffer. The first thing to note is that the binding of
Zn(TC3) has almost no measurable impact on the flow
=
50 (Table 3). Compared with the results obtained with
Cu(TC3), the bathochromic shifts and iCD signals, respec-
tively, trend smaller and more positive with Zn(TC3), while the
calculated H% values tend to be larger. Again, however, binding
to DNA tends to increase the width of the absorption band.
Binding Studies with ss DNA. Interactions with ss DNA
hosts produce absorbance and emission spectra that in many
ways parallel those described for ds DNA hosts. Thus, in the
Soret region, Cu(TC3) once again exhibits the largest
bathochromic shifts. When Cu(TC3) binds to the ss DNA
host T , Δλ = 8 nm versus 10 nm with GATTAC, and the
characteristics of the DNA host. In contrast, η/η nearly
0
doubles with the uptake of Cu(TC3) as R ranges from 0.1−0.6.
Specific viscosity increases of that magnitude are generally an
indication of intercalative binding, which induces an increase in
length as well as a decrease in flexibility of the macromolecular
17
host. At the same time absorbance studies reveal that
Cu(TC3) adopts a second binding motif in this same
concentration regime. Whereas Cu(TC3) normally exhibits a
single, red-shifted Soret band, a second Soret maximum appears
10
F
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
at the higher R values, and it is indicative of a separate bound
form (Figure 10). Likewise, the CD spectrum shows a band in
the porphyrin concentration is 15 μM. New absorbance also
grows in the vicinity of 400 nm; however, it is in the form of a
poorly resolved shoulder, rather than the clearly resolved
maximum obtained with Cu(TC3).
DISCUSSION
■
Ligand Design and Choice of Hosts. Incorporating alkyl
substituents in the meso positions of the porphyrin achieves at
least three ends. One is to isolate the π system of the porphyrin
and avoid the mesomeric interactions that typically occur with
aryl substituents. The upshot is a narrower bandwidth and a
higher molar absorptivity for the Soret absorption of H TC3 as
2
compared with H T4. While alkyl substituents clearly
2
24
contribute to the bulkiness, the flexible nature of the chains
makes for a sterically more accommodating system than H T4.
2
Finally, the four peripheral charges no longer have to extend
Figure 10. Absorbance spectra of Cu(TC3) (gray), H TC3 (dashed),
outward in a fixed plane, as is the case with H T4. An
2
2
and Zn(TC3) (black), in the presence of ST DNA. In each case the
DNA base-pair concentration is 30 μM, and the porphyrin
concentration is 10 μM. Cell path length is 2.0 mm.
unintended result of the design is that H TC3 is subject to self-
2
association in aqueous salt solutions. The same complication
46
does not occur with the H T4 system, perhaps because of the
2
bulkiness and immobility of the meso substituents.
In terms of hosts, hairpin-forming sequences most often bind
ligands in the double-helical stem domains; hence, they are
2
0,47,48
viable B-form DNA platforms.
In addition to being cost-
effective, they offer a defined length and a programmable base
composition. Including the interior sequence 5′-CttttG-3′
47,49
provides for a tight loop and thermodynamic stability.
Sonicated ST DNA serves as a useful contrast, effectively
functioning as a random sequence DNA polymer compatible
with viscometric studies. Finally, single-stranded DNA hosts
provide interesting comparisons because they are relatively
flexible, and the exposed face is more hydrophobic than is the
50
case with a ds DNA host. Like the latter, ss hosts are capable
of binding ligands externally or internally. The term often used
to describe internal binding is pseudointercalation, which
involves sandwiching the ligand between adjacent bases of the
5
1,52
host.
The lengths of T10 and Puc reflect the fact that
Figure 11. Induced CD spectrum of 10.0 μM Cu(TC3) when bound
to ST DNA at a DNA base-pair concentration of 30.0 μM. Cell path
length is 2.0 mm.
exposed runs of naturally occurring ss DNA are generally
53
short. Finally, the host compositions are complementary in
that T10 contains a single repeating pyrimidine base, while Puc
has mixed composition including both adenine and guanine
bases.
the vicinity of 400 nm (Figure 11). Emission data obtained at R
=
0.3 confirm the existence of the second binding motif because
Preferential Internalization of Cu(TC3). The results of all
of the physical studies establish that Cu(TC3) preferentially
internalizes into DNA hosts. In the first place, intercalative
binding naturally explains the observed bathochromic and
hypochromic responses, which are explicable in terms of
excitonic coupling interactions with π−π* transitions of the
the 400 nm transition does not appear in the excitation
spectrum of Cu(TC3). By R = 0.5 the 400 nm absorption
maximum is dominant, and the overall pattern of absorbance
broadens. A 400 nm band maximum also appears in spectra of
simple aqueous salt solutions containing like porphyrins.
Excitonic coupling interactions have been posited to account
for the shift of the Soret band to shorter wavelength due to
formation of H-type, face-to-face porphyrin aggregates. In
contrast, there is no evidence of a hypsochromically shifted
absorption for Zn(TC3) in the same concentration regime. The
probable explanation is that zinc(II) porphyrins show a
1
2,42
DNA bases.
Second, that kind of internalization is
necessary to account for the observation of an emission signal;
39
otherwise, associative attack by Lewis bases efficiently quenches
1
8,20,26,54
the photoexcited state.
Finally, the fact that the uptake
of Cu(TC3) induces an increase in the specific viscosity of ST
17
DNA represents classical evidence of intercalative binding.
3
,45
preference for binding axial ligands,
stacking interactions.
which interfere with
The uniformity of the absorption and emission results indicate
that the mode of binding does not depend on the base makeup
of the host because the percentage of GC base pairs varies by
a factor of 2 across ST DNA and the stem domains of
GATTAC and GACGAC. In contrast, Cu[T4] is only capable
of intercalating into only high-melting duplexes that contain at
Finally, the H TC3 system represents a third variation.
2
Figure 9 reveals that the specific viscosity of ST DNA initially
increases with the addition of H TC3, as with Cu(TC3).
However, beyond R = 0.3, η/η begins to fall off, and by R = 0.5
the specific viscosity drops below that observed for the free
DNA. In terms of absorbance, a distinct broadening of the
spectrum is evident when the ST concentration is 30 μM and
2
0
55
least 50% GC base pairs. At lower percentages local
melting of the duplex structure becomes more feasible, and
external binding of Cu(T4) becomes more favorable. The
G
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
2
0
problem posed by intercalation derives from steric clashes that
occur between the pyridiniumyl substituents of the porphyrin
the bond to the axial ligand; however, that is obviously not
the case with Zn(TC3).
16
and the sugar−phosphate backbone of the host. Cu(TC3) is
more compatible with intercalative binding because it has more
flexible substituents. The story is similar to ss DNA hosts,
which bind Cu(TC3) strictly by pseudointercalation. As before,
the evidence for assigning the binding motif comes from the
strong hypochromic shifts and emission signals generated by
uptake. Internalization of Cu(TC3) by T is so effective in that
In contrast, H TC3 is capable of intercalating because the
standard reduced viscosity of ST DNA increases with the
2
addition of H
slightly smaller than those observed with Cu(TC3), because
external binding of H TC3 is a competitive process. That there
TC3 up to a value of R = 0.3. The η/η values are
2
0
2
are two binding motifs becomes clear from a comparison of
results obtained with the three hosts GATTAC, ST, and
GACGAC, which have percentage compositions of GC base
pairs of 33, 41, and 66, respectively. The spectra presented in
Figure 6 show that the bathochromic shift Δλ increases steadily
as the percentage of GC pairs increases. The GACGAC
system is the simplest because of its relatively high GC
content. Here, the bathochromic shift is large, intercalative
binding is dominant, and the Soret band has a normal
bandwidth. With ST DNA, Δλ is smaller, and external binding
becomes more important because the host contains a higher
percentage of AT base pairs. The presence of two active
modes of binding accounts for the increase in the apparent
bandwidth because in reality the signal is the envelope of two
unresolved absorptions. Finally, when GATTAC is the host,
external binding becomes even more important, and the net
bathochromic shift is smaller. As for the iCD data, the overall
signal becomes on the whole more negative as the percentage
of GC base pairs increases. This trend is also indicative of an
increase in the fraction of intercalated porphyrin.
Mutually Adaptive Distortions. The problem that
remains is to understand why H TC3 and Cu(TC3) bind
differently to ds DNA. The zinc(II) system stands apart, of
course, because it carries an axial ligand. The issue is that
Cu(TC3) binds as an intercalator, while external binding is
much more important for H TC3 even though both porphyrins
are nominally planar ligands. The rigidity of the host is clearly a
decisive factor because the intercalation of H TC3 becomes
1
0
the emission intensity rivals that observed from Cu(T4)
intercalated into GACGAC.
At intermediate loadings (q = 8−16), another effect comes
into play, as Figure 3 reveals that the hypochromic effect is
higher in the initial stages of a titration. This effect is readily
understandable in terms of dipole−dipole coupling between the
transition moments of neighboring, intercalatively bound
4
1
porphyrins. This observation suggests that uptake may be
somewhat cooperative because there are more host molecules
than porphyrins in solution. In view of the absorption strength
of Cu(TC3), the separation between bound chromophores
5
6,57
could easily be as much as a few base pairs, however.
At
higher overall concentration the binding picture becomes even
more complex. Indeed, as noted above, results obtained with
ST DNA show that another type of cooperative association
occurs as Cu(TC3) begins to aggregate in H-type fashion on
the surface of the host. The signature for that binding motif is
the Soret band, which starts to grow in at 400 nm, beyond the
normal red-shifted band found in the vicinity of 420 nm
2
(
Figure 10). That the H-form porphyrin is a DNA-bound form
follows from the fact that the 400 nm transition also shows an
induced CD signal. It is, however, a separate fraction of
porphyrin because the 400 nm band is absent in the emission
excitation spectrum.
2
External Binding of Zn(TC3) and H TC3. The most
2
2
definitive information about the binding of Zn(TC3) comes
from viscometric data, which show that it does not intercalate
into ST DNA. It therefore must bind externally. Because the
viscometry data relate to relatively high-loading conditions, the
discussion that follows strictly pertains to the early phases of
titrations, in the q = 8−16 region; however, there are no
indications of a qualitative change in the mode of binding when
the DNA is in large excess. For all the ds hosts, including ST
DNA, the small bathochromic shifts (Δλ values) are consistent
with external binding. For the sake of reference, Δλ = 2 nm
more important as the GC content increases. Hence it is
logical to infer that the rigidity of the porphyrin has an
influence as well. In particular, H TC3 is amenable to
2
undergoing an out-of-plane distortion, whereas studies have
established that relatively large metal ions like Cu(II) fill the
porphyrin cavity and reinforce planarity. Experimental evidence
for the latter conclusion comes from studies of the rates of
58
racemization of planar chiral porphyrins, as well as kinetic
59,60
studies of atropisomerization processes.
To see why
distortion of the porphyrin might influence the mode of
binding, it is helpful to consider how ligand binding affects the
structure of the DNA host.
when Zn(T4) binds externally to [poly(dA−dT)] , whereas the
2
shift is 6 times greater when Zn(tD4) intercalates into the same
2
5
host. The iCD data also imply that the zinc and copper forms
of the porphyrin bind very differently to DNA. In contrast to
the results obtained with Cu(TC3), the interaction of Zn(TC3)
with ds DNA produces iCD signals that are biphasic, but
predominantly positive. The most curious finding is that
external binding of Zn(TC3) produces such strong hypo-
chromic responses, although it is important to note that
binding to DNA dramatically enhances the width of the Soret
absorption band. Heterogeneity may partly explain the
broadening effect if there is no one preferred binding sequence
or adduct structure. Distortion of the porphyrin is another
possibility, vide infra. The most likely explanation for the
change of binding motif is that, in contrast to copper(II)3
analogues, zinc(II) porphyrins prefer to bind an axial ligand.
Intercalation of a Zn(II) porphyrin would still be possible if the
process were exothermic enough to compensate for the loss of
The first thing to note is that high-affinity binding of a
cationic porphyrin to DNA necessarily involves an induced fit.
For example, for a porphyrin to intercalate, the DNA host has
61
to unwind and create a cavity to house the ligand. In favorable
circumstances the host maintains base pairing and its double-
helical structure; however, the uptake of a very bulky ligand
may disrupt Watson−Crick base pairing, at the same time
forcing a base to extend or “flip”out into the solution
1
6,62
environment.
In contrast, the host structure or structures
that support high-affinity external binding in solution remain to
be identified. Solid-state structures of externally bound
1
1,63
porphyrins are available,
but the relevance to solution
work is unclear because the porphyrin sandwiches between
neighboring hosts in the crystal lattice. A complicating factor is
that porphyrins are not natural groove binders because they do
64
not have the crescent shape of netropsin. Accordingly, many
H
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
investigators agree that the DNA molecule almost certainly has
to distort from its canonical structure and form a suitable
binding pocket to bind a bulky porphyrin ligand with high
the model, the GACGAC host is relatively rich in GC base
pairs and not very compatible with external binding due to a
high ΔG barrier. In contrast, external binding becomes the
R
2
0,31,65,66
affinity.
Formation of the pocket presumably involves
preferred motif with the GATTAC host due to fewer hydrogen
bonds in the stem, a lower melting temperature, and a reduced
generating a suitable hydrophobic surface while at the same
time promoting Coulombic contacts. From that point of view
the term “external binding” may be a bit of a misnomer; indeed,
recent quenching studies of externally bound Pd(T4) suggest
ΔG barrier. The Cu(TC3) ligand, on the other hand,
R
preferentially intercalates into the GATTAC host, presumably
because of reduced strain in the minor groove domain and a
more favorable ΔGB term. The bigger challenge is under-
29
that the host largely envelops the ligand.
A sympathetic or compensatory distortion of the porphyrin
ligand could certainly bolster the interaction with a reorganized
host, and the most likely motion within the core is an out-of-
standing why the unmetalated porphyrin H TC3 preferentially
2
binds externally. Most likely, H TC3 is uniquely capable of
2
undergoing a sympathetic distortion that enhances the induced
6
7,68
plane distortion.
The impetus for the distortion can come
fit (ΔGB term). As Cu(TC3) and H TC3 carry identical
2
from internal forces, such as cadmium(II) insertion (induces
substituents, in all likelihood the distortion involves a core
rearrangement, that is, an out-of-plane distortion. The
possibility of a propeller distortion is particularly intriguing in
this regard because it would render the porphyrin intrinsically
doming) or protonation of core nitrogens (results in
6
7,69
saddling).
Alternatively, incorporating bulky alkyl or aryl
substituents on the periphery of the porphyrin can also lead to
6
7
72,73
saddling. Host−guest interactions are another possibility.
Thus, Yatsunyk and co-workers have published an X-ray
structure that suggests the domed structure of N-methyl
mesoporphyrin IX predisposes it to bind to human telomeric
G-Quadruplex DNA because the 3′-G-tetrad of the host
naturally presents a complementary inverted domelike surface
chiral and directly impact the iCD signal.
The above results illustrate that ligand modification is an
effective strategy for tuning the reactivities of metal-containing
complexes with biomolecules. Further options to explore with
M(TC3) systems include terminating the alkyl substituents
with amine groups so that introducing charge is possible by
protonation or conjugation with therapeutically active metal
7
0
geometry. In addition, they propose that additional
distortions in the host and guest mutually reinforce each
other in the course of adduct formation. Host−guest
74
ions. Shortening the chain length is also possible and could
encourage dissolution as a monomer in water. In closing it is
worth noting that ligand modification is equally profitable as a
tuning strategy when the metal complex has a nonplanar
coordination geometry. To cite a few examples, Barton and co-
workers have studied ruthenium(II) polypyridine complexes
and have shown that elongating one of the conjugated ligands
interactions similarly influence the binding of H TC3. When
2
H TC3 binds to a flexible DNA host like T , the half width of
2
10
the Soret band increases noticeably, consistent with induce-
71
ment of a distortion. In contrast, incorporating copper(II)
stiffens the porphyrin, and the half width of the Soret band of a
DNA-bound form generally remains narrow. It is worth noting
75
promotes intercalative binding. The same group has also
shown that introducing an oversized ligand destabilizes
intercalative binding in favor of insertion at sites where there
that the H T4 system is also relatively inflexible in that its aryl
2
substituents constrain the positive charges to extend outward.
76
That may explain why H T4 exclusively intercalates into the
is a base-pair mismatch. Working with the same metal center,
Glazer and co-workers have shown that introducing a bulky
2
GC-rich host GACGAC, while external binding remains a
77
competitive process for the more fluid H TC3 porphyrin.
ligand induces light-activatable anticancer activity. Finally,
utilizing dirhodium frameworks, Dunbar and Turro have shown
that varying the number of coordinated polypyridines can affect
2
VISTA
■
78
the binding motif, while the aspect ratio of the active ligand
The striking finding that Cu(TC3) intercalates into the AT-
79
impacts the mode of uptake as well as the binding affinity.
rich host GATTAC, while H TC3 binds externally, is an
2
indication that a number of forces jointly influence the binding
2
0
ASSOCIATED CONTENT
Supporting Information
CIF file showing crystallographic data and structural
information. This material is available free of charge via the
Internet at http://pubs.acs.org.
motif.
Studies involving Cu(T4) and the 5′-GAA-
■
GACt GTCTTC-3′ hairpin illustrate the impact the base
*
S
4
19
composition has. Whereas Cu(T4) intercalates into the
hairpin, the same porphyrin shifts to external binding when
base replacement results in the loss of even one hydrogen bond
in the stem. Fortunately, the above results find a simple
explanation in the context of a simple two-step thermodynamic
AUTHOR INFORMATION
Corresponding Author
E-mail: mcmillin@purdue.edu.
■
*
2
0
model. Step one takes account of the structural reorganization
required of the DNA host along with the associated free energy
change ΔG . Step two incorporates the binding energy released
R
Notes
(
ΔG ) upon adduct formation with the preorganized host
B
The authors declare no competing financial interest.
(
DNA*).
DNA → DNA*
ACKNOWLEDGMENTS
■
The NSF funded this research via Grant Nos. CHE 0847229
and CBET-11097833.
porphyrin + DNA* → adduct
The elementary version of the model being invoked assumes
any reorganization energy associated with the porphyrin is
small, although at a minimum desolvation must occur. As one
application, recall that Cu(T4) intercalates into the GACGAC
REFERENCES
■
(
1) Fiel, R. J. J. Biomol. Struct. Dyn. 1989, 6, 1259−1275.
(2) McMillin, D. R.; McNett, K. M. Chem. Rev. 1998, 98, 1201−
2
0,28
hairpin but binds externally to GATTAC.
In the context of
1219.
I
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
2
(
3) Pasternack, R. F.; Gibbs, E. J.; Villafranca, J. J. Biochemistry 1983,
2, 2406−2414.
4) Snyder, J. W.; Lambert, J. D. C.; Ogilby, P. R. Photochem.
Photobiol. 2006, 82, 177−184.
5) Tada-Oikawa, S.; Oikawa, S.; Hirayama, J.; Hirakawa, K.;
(37) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112−122.
(38) Burla, M. C.; Caliandro, R.; Camalli, M.; Carrozzini, B.;
Cascarano, G. L.; De Caro, L.; Giacovazzo, C.; Polidori, G.; Spagna, R.
J. Appl. Crystallogr. 2005, 38, 381−388.
(
(39) Kano, K.; Fukuda, K.; Wakami, H.; Nishiyabu, R.; Pasternack, R.
F. J. Am. Chem. Soc. 2000, 122, 7494−7502.
Kawanishi, S. Photochem. Photobiol. 2009, 85, 1391−1399.
(
(
6) Milgrom, L.; MacRobert, S. Chem. Ber. 1998, 34, 45−50.
7) MacDonald, I. J.; Dougherty, T. J. J. Porphyrins Phthalocyanines
(40) Vergeldt, F. J.; Koehorst, R. B. M.; Vanhoek, A.; Schaafsma, T. J.
J. Phys. Chem. 1995, 99, 4397−4405.
(41) Shelton, A. H.; Rodger, A.; McMillin, D. R. Biochemistry 2007,
2
(
001, 5, 105−129.
8) Agostinis, P.; Berg, K.; Cengel, K. A.; Foster, T. H.; Girotti, A.
4
(
6, 9143−9154.
W.; Gollnick, S. O.; Hahn, S. M.; Hamblin, M. R.; Juzeniene, A.;
Kessel, D.; Korbelik, M.; Moan, J.; Mroz, P.; Nowis, D.; Piette, J.;
Wilson, B. C.; Golab, J. CaCancer J. Clin. 2011, 61, 250−281.
42) von Holde, K. E.; Johnson, W. C.; Pui, S. H. Principles of Physical
nd
Biochemistry, 2 ed.; Prentice Hall: Upper Saddle River, NJ, 2006.
43) Cantor, C. R.; Schimmel, P. R. Biophysical Chemistry, Freeman:
San Francisco, CA, 1980; Vol. 2, Ch. 7.
44) Strickland, J. A.; Marzilli, L. G.; Wilson, W. D. Biopolymers 1990,
(
(
9) Hamblin, M. R.; O’Donnell, D. A.; Murthy, N.; Contag, C. H.;
Hasan, T. Photochem. Photobiol. 2002, 75, 51−57.
10) Han, F. X. G.; Wheelhouse, R. T.; Hurley, L. H. J. Am. Chem.
Soc. 1999, 121, 3561−3570.
11) Parkinson, G. N.; Ghosh, R.; Neidle, S. Biochemistry 2007, 46,
(
(
2
(
(
9, 1307−1323.
45) Collins, D. M.; Hoard, J. L. J. Am. Chem. Soc. 1970, 92, 3761−&.
(
46) Ito, A. S.; Azzellini, G. C.; Silva, S. C.; Serra, O.; Szabo, A. G.
2
(
390−2397.
Biophys. Chem. 1992, 45, 79−89.
47) Lugo-Ponce, P.; McMillin, D. R. Coord. Chem. Rev. 2000, 208,
69−191.
48) Rentzeperis, D.; Alessi, K.; Marky, L. A. Nucleic Acids Res. 1993,
1, 2683−2689.
49) Antao, V. P.; Lai, S. Y.; Tinoco, I. Nucleic Acids Res. 1991, 19,
901−5905.
50) Jain, R. K.; Sarracino, D. A.; Richert, C. Chem. Commun.
Cambridge, U.K.) 1998, 423−424.
51) Pasternack, R. F.; Brigandi, R. A.; Abrams, M. J.; Williams, A. P.;
Gibbs, E. J. Inorg. Chem. 1990, 29, 4483−4486.
52) Biancardi, A.; Biver, T.; Marini, A.; Mennucci, B.; Secco, F. Phys.
12) Biver, T. Appl. Spectrosc. Rev. 2012, 47, 272−325.
13) Mukundan, N. E.; Petho, G.; Dixon, D. W.; Marzilli, L. G. Inorg.
(
1
(
2
(
5
(
(
Chem. 1995, 34, 3677−3687.
(
Am. Chem. Soc. 1992, 114, 7575−7577.
(
14) Marzilli, L. G.; Petho, G.; Lin, M. F.; Kim, M. S.; Dixon, D. W. J.
15) Pasternack, R. F. Chirality 2003, 15, 329−332.
16) Lipscomb, L. A.; Zhou, F. X.; Presnell, S. R.; Woo, R. J.; Peek,
(
M. E.; Plaskon, R. R.; Williams, L. D. Biochemistry 1996, 35, 2818−
823.
17) Strickland, J. A.; Marzilli, L. G.; Gay, K. M.; Wilson, W. D.
Biochemistry 1988, 27, 8870−8878.
18) Hudson, B. P.; Sou, J.; Berger, D. J.; McMillin, D. R. J. Am.
Chem. Soc. 1992, 114, 8997−9002.
19) Tears, D. K. C.; McMillin, D. R. Chem. Commun. (Cambridge,
U.K.) 1998, 2517−2518.
20) McMillin, D. R.; Shelton, A. H.; Bejune, S. A.; Fanwick, P. E.;
Wall, R. K. Coord. Chem. Rev. 2005, 249, 1451−1459.
21) Sari, M. A.; Battioni, J. P.; Mansuy, D.; Lepecq, J. B. Biochem.
Biophys. Res. Commun. 1986, 141, 643−649.
22) Manono, J.; Marzilli, P. A.; Fronczek, F. R.; Marzilli, L. G. Inorg.
Chem. 2009, 48, 5626−5635.
23) Fiel, R. J.; Howard, J. C.; Mark, E. H.; Dattagupta, N. Nucleic
Acids Res. 1979, 6, 3093−3118.
24) Andrews, K.; McMillin, D. R. Biochemistry 2008, 47, 1117−
125.
25) Wall, R. K.; Shelton, A. H.; Bonaccorsi, L. C.; Bejune, S. A.;
Dube, D.; McMillin, D. R. J. Am. Chem. Soc. 2001, 123, 11480−11481.
26) Bejune, S. A.; Shelton, A. H.; McMillin, D. R. Inorg. Chem. 2003,
2, 8465−8475.
27) Wu, S.; Li, Z.; Ren, L. G.; Chen, B.; Liang, F.; Zhou, X.; Jia, T.;
Cao, X. P. Bioorg. Med. Chem. 2006, 14, 2956−2965.
28) Briggs, B. N.; Gaier, A. J.; Fanwick, P. E.; Dogutan, D. K.;
McMillin, D. R. Biochemistry 2012, 51, 7496−7505.
29) Bork, M. A.; Gianopoulos, C. G.; Zhang, H.; Fanwick, P. E.;
(
2
(
(
(
(
Chem. Chem. Phys. 2011, 13, 12595−12602.
(53) Wadkins, R. M. Curr. Med. Chem. 2000, 7, 1−15.
(54) Chirvony, V. S. J. Porphyrins Phthalocyanines 2003, 7, 766−774.
(55) Eggleston, M. K.; Crites, D. K.; McMillin, D. R. J. Phys. Chem. A
(
(
1
(
998, 102, 5506−5511.
56) Lewis, F. D.; Zhang, L. G.; Liu, X. Y.; Zuo, X. B.; Tiede, D. M.;
(
Long, H.; Schatz, G. C. J. Am. Chem. Soc. 2005, 127, 14445−14453.
(
(
57) Lewis, F. D. Photochem. Photobiol. 2005, 81, 65−72.
58) Konishi, K.; Miyazaki, K.; Aida, T.; Inoue, S. J. Am. Chem. Soc.
(
1
(
(
990, 112, 5639−5640.
(
59) Gottwald, L. K.; Ullman, E. F. Tetrahedron Lett. 1969, 3071−&.
60) Freitag, R. A.; Whitten, D. G. J. Phys. Chem. 1983, 87, 3918−
(
1
(
3
925.
(
61) Calladine, C. R.; Drew, H. R. Understanding DNA; Academic:
New York, 1997.
62) Norden, B.; Lincoln, P.; Akerman, B.; Tuite, E. Probing of
(
(
4
(
Nucleic Acids by Metal Ion Complexes of Small Molecules. In Metal
Ions in Biological Systems; Sigel, A., Sigel, H., Eds.; CRC Press: Boca
Raton, FL, 1996: Vol. 33, pp 177−252.
(63) Bennett, M.; Krah, A.; Wien, F.; Garman, E.; McKenna, R.;
(
Sanderson, M.; Neidle, S. Proc. Natl. Acad. Sci. U. S. A. 2000, 97,
9
(
(
476−9481.
(
64) Neidle, S. Nat. Prod. Rep. 2001, 18, 291−309.
Choi, J. H.; McMillin, D. R. Biochemistry 2014, 53, 714−724.
(
(
65) Gibbs, E. J.; Maurer, M. C.; Zhang, J. H.; Reiff, W. M.; Hill, D.
30) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647−2650.
31) Fiel, R. J.; Jenkins, B. G.; Alderfer, J. L. In Molecular Basis of
T.; Malickablaszkiewicz, M.; McKinnie, R. E.; Liu, H. Q.; Pasternack,
R. F. J. Inorg. Biochem. 1988, 32, 39−65.
Specificity in Nucleic-Acid-Drug Interactions; Pullman, B., Jortner, J.,
Eds.; Kluwer Academic: Dordrecht, The Netherlands, 1990; pp 385−
(66) Sehlstedt, U.; Kim, S. K.; Carter, P.; Goodisman, J.; Vollano, J.
F.; Norden, B.; Dabrowiak, J. C. Biochemistry 1994, 33, 417−426.
399.
(67) Sparks, L. D.; Medforth, C. J.; Park, M. S.; Chamberlain, J. R.;
(
32) Kadish, K. M.; Ou, Z. P.; Zhan, R. Q.; Khoury, T.; Wenbo, E.;
Crossley, M. J. J. Porphyrins Phthalocyanines 2010, 14, 866−876.
33) Wu, X. E.; Ma, L.; Ding, M. X.; Gao, L. X. Chem. Lett. 2005, 34,
12−313.
34) Batinic-Haberle, I.; Spasojevic, I.; Hambright, P.; Benov, L.;
Crumbliss, A. L.; Fridovich, I. Inorg. Chem. 1999, 38, 4011−4022.
35) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning a
Ondrias, M. R.; Senge, M. O.; Smith, K. M.; Shelnutt, J. A. J. Am.
Chem. Soc. 1993, 115, 581−592.
(68) Senge, M. O. Chem. Commun. (Cambridge, U.K.) 2006, 243−
256.
(
3
(
(69) Cheng, B. S.; Munro, O. Q.; Marques, H. M.; Scheidt, W. R. J.
(
Am. Chem. Soc. 1997, 119, 10732−10742.
Laboratory Manual, 2nd ed.; Cold Spring Harbor Press: Cold Spring
Harbor, NY, 1989; Vol 3.
(
(70) Nicoludis, J. M.; Miller, S. T.; Jeffrey, P. D.; Barrett, S. P.;
Rablen, P. R.; Lawton, T. J.; Yatsunyk, L. A. J. Am. Chem. Soc. 2012,
134, 20446−20456.
36) Felsenfeld, G.; Hirschman, S. Z. J. Mol. Biol. 1965, 13, 407−427.
J
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
71) D’Souza, F.; Villard, A.; Vancaemelbecke, E.; Franzen, M.;
Boschi, T.; Tagliatesta, P.; Kadish, K. M. Inorg. Chem. 1993, 32, 4042−
048.
72) Shelnutt, J. A.; Song, X. Z.; Ma, J. G.; Jia, S. L.; Jentzen, W.;
Medforth, C. J. Chem. Soc. Rev. 1998, 27, 31−41.
73) Choi, J. K.; D’Urso, A.; Balaz, M. J. Inorg. Biochem. 2013, 127,
−6.
74) Naik, N.; Rubbiani, R.; Gasser, G.; Spingler, B. Angew. Chem.,
Int. Ed. 2014, 53, 6938−6941.
75) Friedman, A. E.; Chambron, J. C.; Sauvage, J. P.; Turro, N. J.;
Barton, J. K. J. Am. Chem. Soc. 1990, 112, 4960−4962.
76) Zeglis, B. M.; Pierre, V. C.; Barton, J. K. Chem. Commun.
Cambridge, U.K.) 2007, 4565−4579.
77) Howerton, B. S.; Heidary, D. K.; Glazer, E. C. J. Am. Chem. Soc.
012, 134, 8324−8327.
78) Angeles-Boza, A. M.; Bradley, P. M.; Fu, P. K. L.; Shatruk, M.;
Hilfiger, M. G.; Dunbar, K. R.; Turro, C. Inorg. Chem. 2005, 44, 7262−
264.
79) Aguirre, J. D.; Angeles-Boza, A. M.; Chouai, A.; Pellois, J. P.;
Turro, C.; Dunbar, K. R. J. Am. Chem. Soc. 2009, 131, 11353−11360.
4
(
(
1
(
(
(
(
(
2
(
7
(
K
dx.doi.org/10.1021/ic501683t | Inorg. Chem. XXXX, XXX, XXX−XXX